After another summer of drought, hosepipe bans and depleted reservoirs, it is perhaps unsurprising that Britain is talking seriously about building more reservoirs.
After all, if we don't have enough water when we need it, surely the answer is to store more when we do?
There is considerable logic in that argument.
But then somebody points out that our water companies lose billions of litres every day through leakage. Somebody else suggests we should take shorter showers. Others argue that we should capture more rainwater from our roofs, recycle wastewater or stop building homes in areas already struggling to supply the population.
And suddenly, what appeared to be a straightforward engineering problem becomes rather more complicated.
The question is no longer simply whether Britain needs more reservoirs.
It is whether we have been thinking about water in the wrong way.
Five billion litres is a lot of water
The Environment Agency estimates that, without further action, England could face a public water supply deficit of approximately five billion litres every day by 2055.
That's roughly a third of the water currently supplied for public consumption.
It is a startling number, but understanding where the shortfall comes from is important.
Population growth is one factor. Climate change is another. Longer periods of dry weather and changing rainfall patterns are making water resources less predictable.
But the largest component of the projected deficit is something that rarely features in the headlines: the need to take less water from the natural environment.
For decades, we have relied on rivers, lakes and aquifers to supply our homes and businesses. In some catchments, abstraction places significant pressure on wildlife and habitats, particularly during dry periods.
The Environment Agency's modelling attributes approximately 2.6 billion litres a day of the projected 2055 deficit to reductions in abstraction intended to protect the environment.
In other words, more than half the challenge isn't necessarily that Britain will suddenly need more water.
It is that we need to stop taking so much from places that cannot sustainably afford to lose it.
That changes the conversation considerably.
The return of the reservoir
Britain's reservoirs are a vital part of our water infrastructure, storing water when it is plentiful for use when supplies come under pressure. But as demand grows and our climate changes, is building more reservoirs enough, or do we also need to rethink how we use the water we already have?
Britain is planning a substantial programme of new water infrastructure.
The Government has identified nine new reservoirs in addition to Havant Thicket in Hampshire, which is already under construction. Alongside these are plans for water transfers, new supply sources and improvements to existing infrastructure.
One of the most ambitious proposals is the White Horse Reservoir near Abingdon in Oxfordshire, formerly known as the South East Strategic Reservoir Option.
The proposed reservoir would store around 150 billion litres of water. Its purpose is relatively straightforward: take water from the River Thames during periods when sufficient water is available, store it, and release it when river flows or supplies come under pressure.
But the project has also become a powerful illustration of the difficulties involved in building new infrastructure.
Its estimated cost has risen dramatically, with recent figures placing it in the region of £5.5 billion to £7.5 billion. It would occupy a substantial area of land, alter the local landscape and require major engineering works.
Residents have raised concerns about flooding, environmental impacts and the wider consequences for surrounding communities.
Those concerns deserve serious consideration.
Equally, doing nothing has consequences. Water shortages carry economic, social and environmental costs of their own.
The important question is whether such projects represent the most effective way of securing the water we need, and what alternatives would be required if they were not built.
Before we build more, shouldn't we stop losing what we have?
This is perhaps the most obvious challenge to the reservoir argument.
According to Ofwat, water companies in England and Wales lost an average of 46.4 litres per person per day through leakage in 2024–25.
Residents of Banbury will remember the dramatic scenes earlier in 2026 when a burst water main sent a spectacular jet of water into the air beside the town's historic statue. While such incidents are particularly visible, much of the treated drinking water lost through leakage disappears unnoticed beneath our streets. Every litre lost is water that has already been abstracted, treated and pumped, but never reaches the customer. SOURCE: Banbury Guardian
Across the population, that represents an enormous volume of treated water disappearing from the distribution network before reaching customers.
The industry has committed to halving leakage by 2050 compared with 2017–18 levels, with further reductions required during the current investment period.
Progress has been made, but not quickly enough.
And it is reasonable for customers to question why they should be asked to reduce consumption, or pay for expensive new infrastructure, while so much water continues to escape from ageing pipes.
However, there is an important complication.
Eliminating leakage entirely is neither technically straightforward nor economically realistic. Finding and repairing increasingly small leaks becomes progressively more expensive, particularly where pipes are buried beneath roads, buildings and other infrastructure.
That doesn't excuse poor performance. It means we need to judge leakage reduction against the costs and benefits of other options.
A litre saved through repairing a leak is a litre that doesn't need to be abstracted, treated and pumped.
But a litre saved in one region isn't necessarily available to a community experiencing shortages elsewhere.
Location matters just as much as quantity.
What if our buildings became part of the solution?
There is another approach that receives considerably less attention than major reservoirs.
Instead of treating homes, schools, commercial buildings and industrial premises simply as consumers of water, what if we designed them to manage more of their own requirements?
Rainwater harvesting is an obvious example.
A rainwater harvesting tank being installed beneath a property. By capturing rainfall for uses such as toilet flushing and garden irrigation, buildings can reduce their reliance on treated drinking water. Incorporating these systems during construction is generally easier and less disruptive than retrofitting them later.
Rain falling onto roofs can be collected, stored and used for activities that don't require drinking-quality water, including toilet flushing, garden irrigation and some commercial processes.
Greywater recycling can potentially reuse water from showers and washbasins for appropriate non-potable applications.
Water-efficient appliances and fittings can reduce consumption without requiring people to make substantial changes to their daily routines.
And smart meters can help households identify unusual consumption and leaks that might otherwise go unnoticed.
None of these measures will eliminate the need for public water supplies.
Nor are they suitable or economically justified in every building.
But together they challenge a fundamental assumption that has shaped our infrastructure for generations: that virtually every litre of water used in a building must first be treated to drinking-water standards.
Consider a new housing development.
We might install efficient toilets and showers, but still connect every property to a system supplying potable water for every purpose.
We then design drainage to move rainwater away from the site as quickly as practical.
During summer, the development may require additional water for gardens and other outdoor uses. During winter, its drainage system may contribute to pressure on downstream infrastructure.
There is an obvious opportunity to think differently.
And importantly, these decisions are easiest to make when buildings and developments are designed, rather than years later when retrofitting becomes more difficult and expensive.
We also need to rethink the landscape
The same principle applies beyond buildings.
For generations, much of Britain's drainage infrastructure has been designed around getting water away.
That remains essential when protecting people and property from flooding.
But there is a difference between safely managing excess water and automatically treating rainfall as something to dispose of.
Sustainable drainage systems, wetlands, restored floodplains, healthy soils and carefully managed catchments can all influence how water moves through the landscape.
Some approaches slow runoff. Others improve infiltration, reduce local flood peaks or support groundwater recharge where geology allows.
Not every measure creates additional water that can be used for public supply. Nor does every litre retained in the landscape eventually reach a reservoir or aquifer.
Nevertheless, managing water closer to where it falls can deliver benefits that conventional supply infrastructure doesn't always provide.
It can support biodiversity, reduce flood risk, improve water quality and help landscapes withstand periods of drought.
Wetlands and floodplains are part of nature's water management system. By slowing the movement of water through the landscape, they can help reduce flood peaks, support biodiversity and, in suitable conditions, encourage groundwater recharge. Managing water where it falls can be just as important as deciding where to store it.
A reservoir stores water in one place.
A well-managed catchment can influence how water is retained, released and used across an entire landscape.
These are complementary approaches, not necessarily competing ones.
So which delivers the best value?
This is where I think the debate needs to become more sophisticated.
We often compare the cost of building a reservoir with the cost of fixing leaks or installing water-efficient technology.
But those measures don't necessarily provide the same benefits.
A reservoir can supply substantial volumes of water during prolonged drought, provided it has been filled and is connected to the places that need it.
Leakage reduction saves treated water every day.
Rainwater harvesting can reduce demand for drinking-quality water at individual properties.
Catchment management can support river health, groundwater and flood resilience.
Water transfers can help connect regions with different supply conditions.
Each addresses a different part of the problem.
The real question is how much resilience each investment delivers, where it delivers it, and what other benefits or costs come with it.
A £6 billion reservoir may prove worthwhile if it provides decades of reliable supply to millions of people.
Equally, it would be difficult to justify such an investment if more effective, lower-cost measures could meet the same need with fewer environmental consequences.
That is why we need transparent comparisons based on whole-life costs, environmental impacts, drought performance and benefits to customers.
And we need to be prepared to change our plans as technology, demand and climate projections evolve.
The answer isn't either/or
There is something rather British about our approach to water.
When it rains, we worry about flooding.
When it stops raining, we worry about drought.
We spend billions moving water away from places where it causes problems, then contemplate spending billions more to bring it back when we need it.
Of course, the geography and engineering are considerably more complicated than that.
But the underlying contradiction is worth examining.
The Government already recognises the need for a twin-track approach, combining new supply infrastructure with reduced leakage and more efficient water use.
I would argue that we should go further.
We need to think about water as a resource that moves through connected systems, rather than simply something supplied through a pipe and disposed of through a drain.
That means building reservoirs where the evidence supports them.
It means holding water companies accountable for reducing leakage.
It means designing new developments to use water more intelligently.
And it means recognising that rivers, soils, wetlands and aquifers are themselves part of our national water infrastructure.
The five-billion-litre challenge is real. We cannot wish it away with shorter showers, nor can we necessarily solve it by pouring concrete into the countryside.
Britain almost certainly needs more water storage.
But perhaps the more important ambition is to build a country that makes better use of the water it already receives.
The question isn't simply how much more water we can store. It is how much less water we need to take, waste and move in the first place.
