Something we should all agree on.

Pic above: The River Ver in the drought of 2022. And the million dollar question: how much flow do we get back if we turn off abstraction?

Making Chalk Streams First a reality: how do we overcome the uncertainties?

A few months ago the Chalk Streams First group commissioned a follow-on report from John Lawson, to investigate and collate evidence of flow recoveries following abstraction reductions in the Chilterns and Herts chalk streams. The report is now finished and we will publish it soon. 

The blog post below introduces some of the complex discussions in the report. It leads to an important proposal for a type of scheme which traditionally the conservation community has been wary of, but which is well worth considering in the light of the debate over flow recovery. 

It could help to address water-supply resilience and ecological restoration and thus allow the full delivery of Chalk Streams First within a few years as opposed to a few decades.


If I was to say “abstraction has a much smaller impact on chalk stream flows at low flows and in times of drought, than it does at high flows and in winter” you’d probably furrow your brow. 

Everyone does.

The explanation is actually found in simple physics. But it’s counter-intuitive and so not easily understood, and because of that, not generally known.

But it is really important that we get our heads around the idea, because otherwise we’re in danger of allowing the Chalk Streams First project, and especially the potential pace at which it could be realised, to be derailed. And we’re in danger of missing the clear benefits of a groundwater scheme that could bring all parties together.

With a surface abstraction from a river, it’s easy to see the link between abstraction and flow. The river is flowing at X Ml/d. You insert a pump into the river and pump at a rate of Y Ml/d and the flow d’stream of the pump will become X – Y Ml/d. 

As the flow in the stream diminishes through the summer, Y will become a larger and larger proportion of that flow. And vice versa, it will become a smaller and smaller proportion of that flow through the winter. Therefore constant surface water abstraction has a BIGGER impact on low flows than high flows.

And yet most groundwater abstractions – and certainly groundwater abstractions at a catchment scale – seem to have the opposite impact: a smaller impact on low flows than high flows.


The reason is bound up with the fact that once you abstract water from the ground and not directly from the river, the impact on flows is no longer direct either, but occurs via the impact on groundwater level. Thus the impact is bound to the relationship between groundwater level and flow, because it is groundwater level – the head of the groundwater level above the bed of the river – that drives flows into chalk streams.

Just as the rate of flow out of a hole at the bottom of a bucket increases as the level of water in the bucket rises, so the rate of flow from an aquifer into a chalk stream rises as the groundwater level rises. The force that drives the flow is gravity or hydraulic head.

But it’s not just hydraulic head that places a cog in the link between groundwater level and flow. There are the properties of the rock matrix of the aquifer – how much water it holds and where and how quickly that water can move – as well as the topographical shape of the valley. And there’s time too. 

All of these combine to create a non-linear relationship between groundwater level and flow. Which means that for each unit rise in groundwater level, you get an exponential increase in flow. The aquifer, its topographic features and time provide gearing, if you like. 

So, if a rise in groundwater level from, say, 98 to 99 AOD gives a much greater increase in flow than a rise from 95 to 96 AOD, it stands to reason that a reduction in groundwater level from 99 to 98 AOD, will cause a much greater reduction in flow than a reduction in groundwater level from 96 to 95 AOD. 

Same unit rise or fall, but a bigger or smaller net reduction in flow.

Abstraction has an overall lowering effect on groundwater level. Therefore, one can see that the overall effect will have a bigger impact on flow when groundwater levels and flows are high, than when they are low.

The implications of this, when combined with water resource considerations, are significant.

We’ve always said, when pushing the case for Chalk Streams First, that if you switch off abstraction you get, on average, about 80% of the water back as surface flow. We’ve also always said that % flow recovery varies through the year, with over 100% in winter and as low as 30% or so in late summer (when groundwater levels are lower). In very low flows (Q99) this recovery might drop to only 20%.

This is where TIME comes in. The abstraction rate was constant, but because of aquifer gearing, the “less” you get back in the summer, actually comes back as “more” in the winter and you have to take a long view to see that over the full year the amount of water you get back is about 80% of that former abstraction rate (the remaining 20% passes through the aquifer under the ground).

Obviously these figures vary stream to stream.

It is the figure at low flows that is key, because it is low flows and especially 18-month droughts, that threaten water supplies in and around London. 

In the current drafts of Affinity Water’s & Thames Water’s water resource management plans, and the WRSE national framework plan, the estimate of flow recovery at low flows is a conservative 17%, where we think the figure should be over 50%.

The critical droughts used for planning purposes are the 1921 and 1933/34 droughts. The WRMPs are built around the basis that flows in these droughts averaged at the 98th percentile. In fact, from the beginning to the ends of these droughts flow averaged the 95th and 89th percentiles respectively, a combined average of the 92nd percentile, at which flow recovery from the Chilterns chalk streams – in our estimation – would be over 50% of the abstraction reduction.

The knock-on effect of planning according to that – in our view – irrationally low figure of 17% is that the lion’s share of the Colne / Lea chalk-stream abstraction reductions have to wait for a very large strategic resource option such as Severn-to-Thames Transfer or Abingdon reservoir. That pushes them back until after 2040. 

Can we really wait two more decades – and probably longer – for the full restoration of flows to our beleaguered chalk streams? In our view we could see most of the very much needed abstraction reductions far sooner. 

Part 2. The way round this which we could all agree on and act now.

Despite our differences of opinion over flow recovery, we can probably all agree that there is indeed uncertainty over the % flow recovery we will see at low flows. 

Therefore, from all points of view it is surely best to investigate ideas that overcome this uncertainty, ideas that might ensure the resilience of public water supply – as that is the issue of greatest concern to Affinity and their primary statutory duty – and also fully re-naturalise flows in the chalk streams, within a reasonable time-frame, not 20+ years.

The original Chalk Streams First idea was designed as a pragmatic solution to the formerly irreconcilable tension between public water supply and the ecological health of chalk streams near London. It always depended on storage because of the differing levels of flow recovery you get through the year. Now, with flow recovery at the very lowest of flows identified by planning constraints as the limiting factor, it is surely worth exploring additional ways to insure against these uncertainties.

Minworth – GUC transfer.

Water transfer from the Minworth sewage outfall via the Grand Union Canal is one such idea that has been universally accepted as worthwhile: it is in the WRSE and WRMP plans and of itself covers off a proportion of the abstraction reductions needed to restore natural flows to the Chilterns and Herts chalk streams. But it isn’t enough.

Licence Relocation

Rolling the groundwater abstractions down the catchment is another such idea: this came from Affinity and is a version of the Chalk Stream First concept. If the abstractions are sited down-catchment, where there is always perennial flow that vastly exceeds the pumping rate, much supported by discharges, then it is likely that stream-side groundwater abstractions will have an almost 1:1 impact in the same way surface abstractions do.

Groundwater Insurance Scheme.

In spite of both ideas above being no-brainers and enjoying universal support, the full-fat version of the Chalk Streams First proposal is presented as dependent on another large strategic scheme, such as Abingdon Reservoir or the Severn-to-Thames transfer. The Chalk Streams First group sees both as potentially important components to help ensure resilient water-supplies, but we would be extremely and justifiably disappointed if the full realisation of Chalk Streams First was made contingent on these either or both of these schemes. It would push the ecological recovery of these chalk streams decades down the line.

Therefore a third idea, tried and tested elsewhere, and worth exploring in the context of the Colne and Lea, would be a version of the West Berkshire Groundwater scheme (WBGWS).

Once it is understood that abstraction at low flows has a much smaller proportional impact at the time*, then it can be understood that something like the WBGWS has the potential to guarantee drought supply with a minimal ecological impact. This guarantee then underwrites the whole Chalk Streams First concept because the limiting low-flow recovery is underwritten.

(*Important to contextualise this “smaller” impact: where abstraction runs at 25 – 50% of recharge, as it does in the chalk streams around London, this “smaller” difference can still be the difference between flow and a dry river).

The West Berkshire Groundwater Scheme (WBGWS) was constructed in the 1970s to augment London’s water supplies during severe droughts – its planned use is about once in 25 years. The scheme abstracts water from boreholes in the chalk aquifer in the upper Lambourn, Pang, Enbourne and Loddon valleys, discharging water into those rivers from where it flows down into the River Thames for later abstraction to fill London’s reservoirs. It contributes about 90 Ml/d to London’s deployable output. 

The groundwater outfall on the Winterbourne tributary of the Lambourn, part of the West Berkshire Groundwater Scheme. A scheme like this on the Colne / Lea could underwrite all the abstraction reductions suggested in Chalk Streams First.
© Des Blenkinsopp

The WBGWS concept could be used in the Colne and Lea chalk tributaries, in combination with current proposal for reduced abstractions for day-to-day supplies. Replacement supplies would be transferred from the London supply system using the Thames to Affinity transfer and the ‘Connect 2050’ pipe network. 

Our initial assessment of the WBGWS concept in the Chilterns chalk streams has shown that on the River Ver a reduction of abstraction from the current 28 Ml/d to about 8 Ml/d, combined with WBGWS-type drought support of up to 25 Ml/d, would almost re-naturalise River Ver flows and also give a net increase in London supplies of about 9 Ml/d. 

If the concept was adopted in all the Colne and Lea chalk streams, abstraction could be reduced by 150 Ml/d to meet EFIs throughout the catchment. The 50 Ml/d first phase of the GUC transfer is a no-brainer, so only 100 Ml/d of replacement needs to come from the connection to London’s supplies. 

If the deployable output recovery of London’s supplies from the 150 Ml/d reduction is the 50+% that we predict, the concept gives a 50-60 Ml/d gain in London’s supplies. If the DO recovery is only 17%, the concept would still allow the 150 Ml/d chalk stream reductions to take place without impacting London’s supplies.

The drought support would only be needed about once in 25 years. Drought flows in the chalk streams would be increased by the WBGWS-type releases and would be slightly less in the following year but importantly, they would still much more than with abstraction at recent levels. 

This would remove much of the doubt that currently exists over the amount of flow recovery from abstraction reductions. A net gain in deployable output of 55-60 Ml/d could make this a significant new water resource in its own right. 

Why have schemes like this traditionally been seen as a “bad thing” by conservation groups? An incomplete appreciation of the gearing of impact must be one reason. Another may be an ideological aversion to the active management of a natural system. But all aquifers in south east England are managed to a degree: better by far to manage them intelligently to suit all desired outcomes. A cynicism about schemes which are usually proposed by water companies might be another reason!

Let’s put all these aside. Surely we need to engage in serious discussion about this idea?

Chalk Streams First appears to have brought all parties together: Ofwat, the water companies, EA and the NGO’s. But in the current draft plans it’s pace of instigation is still too slow / and the full version is at best uncertain.

The GUC transfer is a must have – so please support that in your responses – but let’s ask for a serious investigation of the groundwater insurance scheme too.

Much credit is due to John Lawson for the many, many hours of investigative work he put into the ideas I have summarised above.

One thought on “Something we should all agree on.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s