Introduction

1. Thames Water propose a huge reservoir near Abingdon to provide about 300 million litres a day (Ml/d) of water, mainly for London but also Swindon and Oxfordshire (SWOX), costing about £1billion.

2. The aim of this summary is to inform recipients by pointing out the most important technical flaws in the TWdWRP so that these can be understood and transmitted to DEFRA by interested parties in the consultation phase.

3. The Group Against Reservoir Development (GARD) bases its arguments on strictly factual scrutiny of all the data available and a study of the text. It has engaged, and worked with, a consultant, Mr Chris Binnie, who is an acknowledged expert on water resources. He has produced a detailed report on the TWdWRP which reveals that some of the arguments, figures and detail in the plan have serious limitations, and its conclusions are not robust. There are also omissions from the range of options considered by Thames Water (TW) and in the information provided which together undermine the plan as a comprehensive and balanced process for selecting the best solution. In purely business terms from the viewpoint of a monopoly provider the Abingdon reservoir is attractive, but for the customer it is by far the most expensive choice and it is certainly not in the best interests of the consumers who will have to fund it. TW have not justified the need, and GARD has submitted more appropriate and cheaper alternatives. Allowing for demand overestimates and for reasonable rather than excessive contingency, the reservoir option would not be needed, at least for the span of the next 25 years.

4. Once a commitment to build a reservoir is made the expenditure cannot be prevented, and the consumer will have to pay; so any plan to go ahead must be fully proven. A repeat of the Kielder reservoir which, some 25 years after completion, still only operates at 25% capacity, must be avoided. Responsibility for scrutiny lies with the Environment Agency (EA) and OFWAT, with the ultimate decision being made by DEFRA. TW cannot lose, but customers will be facing about £20 pa on bills for all time.

5. The 900 page TWdWRP contains a huge amount of detail and tables covering demand, supply and options. However the capital costs are not broken down or explained and differ significantly from those published less than two years ago some viable alternative schemes have been ruled out.

Analysis.

6. Demand. TW has applied 165 litres per head per day (l/h/d) for its customers whereas the government expectation is 130 l/h/d, with new homes aiming for 115 l/h/d. This illustrates the cautious approach taken by TW. However a figure which is clearly not credible is the garden watering demand estimate (which happens during peak summer). The annual average assessed by other water companies for measured households is about 5 to 10 l/h/d. The TW figure, for London which is about half flats, rises to more than 4 times this figure. This gross overestimate of demand by TW amounts to about 140Ml/d! There are other small but significant overestimates in household demand, but a further marked anomaly is the non-household demand, which TW predicts as falling only slowly compared with recent appreciable falls.

7. Reducing Demand.
The main features are leakage reduction and metering.
a. Leakage. The largest reduction in demand can be achieved by TW tackling its leakage, on which it had a bad record until 2006. In the 7 years to 2010, leakage will be cut by 270 Ml/d assuming the target of 690 Ml/d is achieved. They plan to reduce this to 520 Ml/d by 2020 (a reduction of 170 Ml/d over 10 years). Thereafter leakage hardly changes so it is hard to avoid the implication that by then TW thinks it will not need more water as it will have its reservoir! See also “b” below. TW have stated they aim to be industry average by 2020, however by then the industry average will have reduced, and TW would need to achieve about 460 Ml/d (ie. a further 60Ml/d reduction and 230Ml/d in total)..

b. Metering. The only really new aspect of the dRWP is the introduction of compulsory metering which is welcomed, and which will draw consumers’ attention to the need to reduce their consumption or pay the price. Although the EA have been pressing for universal metering with 80% completed by 2015, the dWRP gives 2020 as its 80% target. They state in their data that leaks from customer supply pipes are assessed as 26% of total leakage, and that 75% of these should be located by metering. This results in 115 Ml/d reduction, and is part of the total 170 Ml/d leakage reduction between 2010 and 2020 mentioned above, thus leaving a puny 55 Ml/d reduction from other repairs and replacement over a full 10 years which is a surprisingly low target. Beyond 2020 their plan is even less ambitious.

c. Building Regulations, water efficiency, customer education, advice and audits etc.
TW efforts at customer demand management are meagre considering they could postpone the construction of a large reservoir and save some £45m in interest charges for each year the scheme is postponed.

8. Contingency, called Headroom.
This is the factor of safety applied to demand figures to cover contingencies such as climate change variation. TW have taken a risk averse view by comparison with other companies previously. The EA Water resources Planning Guidelines state: “We will not expect water companies to include allowances in future headroom to cope with uncertainties that can either be reduced or dealt with as time progresses”. WRPs are re-submitted every 5 years. There are a number of schemes which could be implemented in 5 years. Yet TW expect customers to pay for construction to cover 70% of the risk in 25 years time. Other experts suggest a 50% risk for 2030 would be appropriate indicating a reduction in TW’s contingency of about 60 Ml/d.

9. Supply
a. The Proposed Abingdon Reservoir Scheme. TW project that there will be shortfall in supply to meet demand by about 2020. However this is only to cover projected demand plus the contingency. Without the contingency there would be no need for another big source until about 2027/8. Thus there is the likelihood, if the reservoir was to be built, that it would not be used for some 7 years at least.. The reservoir is the favoured TW option. It is impossible, using only the TWdWRP data to make a fair and detailed financial comparison of this proposal as against other options because the table of capital costs provided does not include re-use or other transfer options. However there is a surprising divergence in cost, measured in £/Ml, in the figures that have been published. The TW Stage 1 Report of 2006 quoted the cost at £ £454/Ml, but the TWdWRP, less than 2 years later, now quotes £276/Ml. Such variations are difficult to believe and need to be challenged. There is no explanation for the difference and an oral response by TW was that a “full use” had been assumed as other water companies would be using the resource. This is not so! From the dWRPs of neighbouring companies the first time there would be a net requirement would be 2030 but only in small measure. No evidence is provided to show that the flow in the Thames at the intake near Abingdon would be sufficient to maintain the supply proposed. The reservoir is by far the most expensive option with the longest construction time and involves serious local disruption. Its claimed environmental and recreational benefits are peripheral and not relevant to a robust case, and in any event would only be realized in decades to come.

b. Water Transfer. One alternative is to store water in the Severn catchment and to pump it over to the Thames. Previously the reservoir site was Criag Goch in mid Wales but this is now an Specially Protected Area, so not available. The best option appears to be a reservoir near the lower Severn valley which has for years been identified as a suitable site, and merits examination. It was considered by Severn Trent but is ignored by TW. GARD would regard this as a suitable long term project with some strategic flexibility. Other schemes are to pump water from the Severn rail tunnel, or from under Birmingham by backpumping the canal network. All would allow the water to be used in Oxfordshire and then the returning effluent travel downriver to London.

c. Water Re-use. Virtually all sewage is now treated to high standards. Inland, treated effluent is discharged into rivers where it forms part of the water to be treated in water treatment works to the exacting standards of the Drinking Water Directive. Unlike inland towns, London’s sewage treatment works discharge into the saline Thames estuary. A very satisfactory scheme at Langford, in Essex has been operating for 5 years treating such waste water to a higher standard, discharging it into the nearby river from which the combined flow is then abstracted and treated to drinking water standards. Anglia Water proposes in their dWRP to adopt this system for some new supplies. In contrast TW have a sewage works in the Lee Valley which has both the river Lee nearby and a reservoir in which to mix the water and the Coppermills advanced water treatment works set up to treat such water. The economics, engineering, proven technology and sustainability of such schemes is clear. Such re-use schemes can also be implemented relatively quickly when compared to construction times for a reservoir. However TW assume that much research and a pilot plant using the very expensive reverse osmosis process are required. They do not include this as a supply in their 25 year plan. Of note is that a waste water re-use plant is a guaranteed source of supply which is not dependent on rainfall or climate change and is close to the centre of demand thus requiring little transfer energy.

Deductions.

10. From the above it can be seen that the TW stated requirement for the reservoir in 2020 is much too early; their garden watering is overstated by 140Ml/d and their commercial demand is overstated by about 80 Ml/d. To meet their own leakage target would require reducing leakage by another 60 Ml/d and their contingency is 60 Ml/d too high. On these assumptions the reservoir, even if built, would not be needed until after 2035.
There are also spare resources not included in TWdWRP including the Kennet valley of 15 Ml/d, bulk supplies from other water companies of about 20Ml/d, the Deephams re-use full scale plant 25 Ml/d, and the decommissioning of the Didcot A power station releasing about 40 Ml/d of licence. Of all the new sources the reservoir is also the most expensive option, and the table below gives some of the alternatives (not all of which are explained in this brief summary for lack of space) which are cheaper and can be implemented more quickly.

Schemes marked with* amount to 481 Ml/d and could replace the need for a reservoir, The lower Severn Transfer Scheme also needs to be re-examined.

Conclusions. 11.

GARD believes that the TWdWRP is seriously flawed because:

a. The errors and omissions pointed out above make the validity and technical credibility of the TWdWRP as a whole highly suspect. TW’s case is not robust.
b. TWdWRP figures, when corrected, demonstrate that an expensive reservoir is not needed at least for the next 25 years. Leakage targets beyond 2020 are far too modest. c. TW have failed to consider several attractive and viable water resource options. Specifically they have not included the lower Severn scheme, or Mogden re-use which can be brought on in stages to reflect demand increases. They have not included effluent re-use in their preferred plan.
d. TW have not provided a breakdown or explanation of capital costs which are appreciably different from those published less than two years ago. It impossible judge the full financial implications from their tables.
e. The Abingdon reservoir is by far the costliest option with the longest construction time. It is not financially in the customer’s best interests. No evidence has been provided to show that a reservoir on this scale is sustainable without water transfer as well.
f. A decision should be deferred for fuller assessment of the alternatives.

12. GARD has put forward water re-use schemes which eliminate the longer term need for a reservoir either for London or for SWOX. Diagrammatic and explanatory details are enclosed for those wishing to study them.

13. It is hoped that this summary of GARD’s full appraisal will inform non-technical readers and provide sound information to assist them in their responses to the TWdWRP consultation. GARD appears to be the only group carrying out a technical scrutiny apart from the EA. GARD will be happy to brief those interested further.

24 June 2008