Aiming high to set new targets

What is the right level of leakage and has an economic frontier has been reached on leakage reduction? David Howarth of the Environment Agency aims to find out

Zero leakage from company and customer pipes is impossible to achieve and to even approach such a level of leakage would be prohibitively expensive, resulting in considerable additional costs for customers. But ‘high’ (and let us not define high for the moment) leakage is also unacceptable, imposing costs on the customer in waste of product, unnecessary abstraction and the need for new resources to keep supplying the leaks. Somewhere between these two extremes is a level of leakage that is the right one. But what is that level of leakage?

Value for money is important, so the economics do matter, but should this ‘right’ level be determined solely by economics or by a combination of economics and judgement, taking into account social, technical, environmental and even political influences?

This is a question that needs to be asked and answered because the industry and its regulators have currently reached a leakage barrier – which could be termed an
economic frontier.

How have we reached this position? In May 1997, within three weeks of taking office, the Labour government (largely in response to the drought of 1995) in the form of deputy prime minister John Prescott drew up a ten-point plan for a better water industry in England and Wales.

Starting with point number one: The director general of Water Services will set tough mandatory targets for total leakage, which will enforce a substantial reduction in leakage over the next five years.

This would in part be facilitated by point number two: The government expects all water companies to provide a free leakage detection and repair service for supply pipes owned by domestic customers.

Ofwat initially set targets pragmatically (the higher your leakage in one per property per day and the smaller your water resource margin, the greater would be the expected reduction) and subsequently based them on the companies’ own calculations of their economic level of leakage (ELL).

Suddenly, water company ELLs began shifting significantly downwards, converging towards the pragmatic targets set by Ofwat. Total leakage fell from 5112Ml/d in
1994-1995 to 3245Ml/d in 2000-2001 – a decline of 35% and a success story claimed by regulators and companies alike. However, in the last two years, this six-year trend of declining leakage has seen a reversal (see Figure 1) despite the fact it is government policy that leakage should not be allowed to rise.

The 209Ml/d rise in the last year is largely attributable to two companies – Severn Trent and Thames. With Severn Trent, it is primarily due to a re-estimation of their unmeasured per-capita consumption (pcc) from 141 l/d to 127 l/d, increasing the residual term leakage in the water balance.

The poor performance of Thames Water is posing a continual challenge to regulators. Draft water companies’ water resources plans looking forward 25 years submitted to the Environment Agency (EA) in August 2003 showed the companies have taken a conservative view of their future progress in reducing leakage.

Companies currently above their ELL have defined a glide path to achieve it but most predicted little change from the current position.

This seems to neglect the fact the economics are dynamic. In many parts of England and Wales, water scarcity is set to increase and a potential need for expensive water resource investments will change the economics of leakage control, as will likely future technological development of leakage control methods and tools. The economic level of leakage has become the holy grail of water company leakage managers and regulators. The concept, where the cost of saving 1m3 of water by leakage control is equal to the cost of providing 1m3 by other means (new resources or demand management), is relatively simple to understand. The practice, however, is another matter.

Although the regulators got together in 2002 to produce Future Approaches to Leakage Target Setting for Water Companies in England and Wales (the tripartite study), which sets out how the ELL calculation should be approached, considerable leeway still remains in interpretation and assumptions. For example, costs, unmeasured pcc, projected savings from leakage policies and, although water companies and regulators are signed up to the principle of incorporating social costs and benefits into their ELL calculation, the reality is that it is the costs and benefits to the company that take centre stage with social and environmental costs relegated to the margins.

capital expenditure

The ELL calculation also works on the assumption companies will save on capital expenditure on resource developments, despite the fact in most cases, they are yet to be awarded this capital expenditure through the periodic review process.
So what does the ELL represent? Is it a genuine boundary between what is economic and what is not, or merely a convenient theoretical construct on which to hang a number? At the Society of British Water and Wastewater Industries (SBWWI) leakage conference on December 18, Mick Merrick of Radcom technologies stated “most water users can’t understand ELL and most would not regard it as justifiable”. So, has an economic frontier been reached? The question can be interpreted in two ways – have companies reached their ELLs and to go beyond is undesirable because it is uneconomic (one sort of frontier) or is this method of setting leakage targets approaching the end of its useful life? (a different frontier) or perhaps both.

Looking forward 25 years, what should the level of leakage be? Assuming the government does not change its position, it will not be higher than current levels.
About the same is the company view, but this takes no account of changing economics and politics, so is also unlikely. This leaves leakage reducing further as the only possible outcome, but by how much should it be reduced? We have seen the first generation of leakage targets were extremely successful, but where will the next generation come from? There is no international best practice to call on – where targets have been set, these are usually in (now discredited) percentages.
One of the outcomes of the tripartite study was that the ELL approach to leakage target setting would remain for at least the next five years, while in parallel some alternative approaches could be explored.

The EA has expressed an interest in exploring two alternative approaches to leakage target setting:

  • targets based on system characteristics – such an approach would consider the characteristics of a company’s system (for example, age of mains, burst rate, number of connections, operating pressure, length of mains, etc) and estimate what the level of leakage would be for a given policy. Hence, if there was a regulatory expectation of a given policy being implemented this could, in theory, be translated to an absolute leakage level target for each company. But such an approach should not assume these characteristics remain static – pressure, system age, burst rate can all be improved by pressure management and mains replacement,
  • targets set on abstraction levels – this approach recognises it is abstraction that damages the environment rather than leakage per se. Abstraction targets would be based on environmental limits. Abstraction targets could then be met by water companies (and other abstractors) by whatever demand management policy is most cost-effective for them, which may or may not be an increased leakage control effort.

The EA is proposing to investigate these options further as part of the implementation of one of the 30 actions from its national and regional water resources strategies.

The system for setting annual leakage targets should be maintained and developed.
In addition to monitoring how close a water company is to its ELL, the Ofwat report Security of Supply, Leakage and the Efficient Use of Water 2002-2003 also quotes leakage levels in l/property/d and m3/km/d, allowing comparisons to be made.
But neither of these measures take much account of differing system characteristics. The International Water Association (IWA) in its Blue Pages, Losses from Water Supply Systems: Standard Terminology and Recommended Performance Measures introduces the infrastructure leakage index (ILI).

With this approach, each system has a value for unavoidable average real
losses (UARL), which can be calculated if the following variables are known – burst rate, length of mains, number of connections, length of service pipes and pressure.


The technical indicator for real losses (TIRL) is simply the current annual volume of real losses divided by the number of connections (not the number of properties).
The ILI is then simply the ratio of TIRL to UARL. An ILI of 1.0 represents a well-managed system in good condition. It is important to note that achieving an ILI of 1.0 or below may not be economic, from either the company or society’s standpoint.

For the first time this year, the EA collected from the water companies sufficient data in order to be able to calculate individual water company (and water resource zone) ILIs. The results of this exercise are shown in Figure 2 for the water companies of England and Wales.

ILIs from 22 water companies (one water company did not provide sufficient data) range from 1.1 to 6.4. By making allowances for significant system characteristics, as the ILI calculation does, a different view is obtained of the best and worst performing companies. This range of ILIs can be compared with the IWA’s international dataset that ranged from 0.7 to just less than 11.0. The majority of water companies are within the bottom half of that international dataset. One criticism of the ILI is that by allowing for pressure, it doesn’t drive a pressure management solution – hence there is an ongoing need to assess separately the scope for pressure management. Establishing an ILI value that companies would be expected to meet could be one way of setting a target based on system characteristics. The EA believes leakage control will continue to play an important role in the management of the supply-demand balance. With the majority of the companies claiming to have achieved their ELL and predicting few reductions in future, a period of apparent stagnation has been reached, whereas the EA would expect to see water company leakage declining into the future.

One way of achieving this might be a new, open, transparent approach for setting leakage targets to be applied consistently and fairly.

Perhaps a new method of leakage target setting should take account of the economics as opposed to being entirely driven by them – the challenge then being to achieve the target economically, through a properly funded price setting process.
One other recommendation from the tripartite study was that the industry should develop a suite of leakage performance indicators (LPIs) to explain variability in both targets and performance.

The industry has taken up this challenge and the EA looks forward to playing a part in this process.

Comparing LPIs across companies should increase the dialogue leading to the promotion and sharing of good practice and drive the next round of technological development.

At the SBWWI conference Merrick also stated “there wasn’t a wide discourse between the industry and its suppliers and it was becoming increasingly difficult for manufacturers to take the risk”.

This does not sound like the route to the next technological advance. When 700 tankers transported water around Yorkshire in the 1995 drought, leakage hit the headlines.

While it has not remained there, it has been prominent on the water company/regulatory agenda since then and is likely to remain there for the foreseeable future.

When the next drought comes, leakage will once again come under public and political scrutiny

Action inspires action. Stay ahead of the curve with sustainability and energy newsletters from edie