CCT Appraisal of the Thames Water draft Water Resource

Management Plan 2019

Contents:

1. Executive Summary …………………………………………………………………………………... Page 2

2. The TW dWRMP Preferred Programme as it affects the Cotswold Canals .. Page 3

3. The Alternatives ........................................................................….. Page 4

3.1 Teddington Direct River Abstraction ......................................….. Page 43.2 Abingdon Reservoir ................................................................. Page 63.3 Water Re-use/Desalination ....................................................... Page 7

4. Severn – Thames Transfers …….....................................................…. Page 7

4.1 How it works ......................................................................….. Page 74.2 The Cotswold Canals Severn-Thames Transfer (CCSTT) ............….. Page 8

5. National Water Grid – Alternative Concept ....................................…... Page 13

Ken Burgin BScChief Executive – Cotswold Canals Trust

Draft 1.1 20th March 2018

Cotswold Canals Trust TW dWRMP 2019 Briefing - Rev 1.1 Page 1 of 14

1.Executive Summary

Thames Water(TW) have launched a period of public consultation on their draft WaterResource Management Plan 2019 (dWRMP19) and are inviting comments until 29 th

April 2018. This briefing has been produced to provide insight into some of the issuesand to encourage the support of a Cotswold Canals Severn – Thames Transfer(CCSTT) scheme to convey raw water from the west of the country to the drier south-east.

TW has evaluated a wide range of potential solutions to address the increasingvulnerability of London and the South-East to serious water shortages in droughtconditions. This work has included looking at Severn-Thames Trasfers (STT), includingusing the Cotswold canals for this, as well as many other options. The outcome of thishas been a range of feasible options from which a preferred programme is beingproposed. Although the canal option has been established a feasible, the preferredprogramme does not include this option and we think it should.

There are no easy options and those chosen in TW's preferred programme are notwithout their own difficulties and some of these are outlined later. Information,particularity relating to costs, has been withheld to an extent that seems to go wellbeyond that which could reasonably be justified on the grounds of commercialconfidentially. Cost extrapolations based on the information that is available suggeststhat the CCSTT option could be very cost effective compared with the best of the otheroptions with a similar water delivery capacity.

Add to this the huge benefits that the CCSTT scheme would deliver in terms of theenvironment (including biodiversity connectivity), recreation, heritage, economic andwell-being, the case for a rethink is very strong.

Responses to the draft WRMP need to be sent by email to:consultations@thameswater.co.uk

Or by post to:

WRMP c/o Water Resources Policy, Area 3D Nobel House, 17 Smith Square, LondonSW1P 3JR

Cotswold Canals Trust TW dWRMP 2019 Briefing - Rev 1.1 Page 2 of 14

2. The TW dWRMP Preferred Programme as it affects the CotswoldCanals

The Thames Water draft Water Resource Management Plan was released for publicconsultation between 5th February until 29th April 2018.

The draft plan and and its many supporting documents can be found at:

https://www.thameswater.co.uk/sitecore/content/Your-Water-Future/Your-water-future/Providing-enough-water/Our-draft-Water-Resources-Management-Plan-2019

This is a vast pool of information and only a limited amount needs to be understood toenable meaningful comment on the plan insofar as it relates to using the CotswoldCanals for transferring water from the Severn to the Thames.

The main thing to understand is that London and the south east has growing waterdemands and some existing supplies are going to become reduced or unavailable –mainly due to environmental damage.

The other thing to understand is that problems occur when there are periods of lowrainfall extending over more than one year. Were Thames Water to have to curtail thenormal continuous availability of water, the cost to the economy would be of the orderof £330 million per day. Such disruption, were it to occur, is unlikely to last for a dayor two but possibly for weeks or even months. The stakes are therefore high anddoing nothing and hoping for the best is no longer a valid option.

The new dWRMP, like the last one in 2014, relies on reducing demand and fixing leaksto cope in the short term (i.e. next 5 years) but it recognises that work has to startvery soon on planning the delivery of an additional resource of about 250 – 300 Ml/d.This additional resource needs to be working within the next 10 years or so. Afterthat, a second new resource of a similar size is needed in about 25 years time andthen further ones beyond that. The plan extends to 2100.

The Preferred Programme in dWRMP19 identifies the first resource as the TeddingtonDirect River Abstraction and this is described later in section 3.1.

The next resource would be the building of a huge new reservoir at Abingdon 50%larger than the very controversial one turned down by the inspectors in the 2010pubic inquiry into the WRMP09.

Beyond that, water reuse (cleaning up sewage and putting it back into the watersupply) is brought into play.

As the WRMPs are revisited every 5 years, it is almost inevitable that the longer termpreferred resources will change in the light of increases or decreases in demandand/or changes of technology and/or environmental issues.

The issue for supporters of the Cotswold Canals needs to be focussed on whether aCotswold Canals Severn-Thames Transfer (CCSTT) scheme is better than theshort(ish) term planned Teddington DRA option currently favoured and/or the longerterm Abingdon Reservoir. It also needs to have more merit than a Severn-Thamestransfer using just a pipeline from Deerhurst near Tewkesbury.

Cotswold Canals Trust TW dWRMP 2019 Briefing - Rev 1.1 Page 3 of 14

It is important to understand that what is being discussed and consulted on is a draftWRMP. It may well get changed in response to public input or changes happening inthe background which may change the basis upon which the current plan is based.After this, a revised WRMP (rWRMP) will be produced and that could well go to apublic inquiry which might chuck it out as happened in 2010 when the CCSTT optionwas viewed as having enough merit by the Inspectors for them to explicitlyrecommend that it be properly investigated.

The next sections look at the alternatives to the CCSTT and why the Trust thinks theyare are not as good as the CCSTT option(s) based on the information available and,where it is not, our assessment of the situation.

3. The Alternatives

3.1 Teddington DRA

The thinking behind this scheme is that the amount of water that Thames Water cantake from the Thames upstream of London is controlled by the amount of water thatflows over Teddington Weir. This supposed to ensure that a certain amount of freshwater continues to flow down the river and down the estuary.

The preferred scheme involves redirecting water from a treated sewage outfall from 4miles below Teddington Weir and building a new pipeline to discharge about270Ml/day of water above it. The logic is that 270Ml/d extra can be taken out of theThames further upstream.

The potential flaw in this logic is that whilst the flow over the weir would be the same,the river upstream of the pipeline would now have a flow reduced by 270Ml/d and theflow into the estuary below where it otherwise would have been discharged has alsolost 270Ml/d of flow (see next page).

If the reserve flow over Teddington Weir was designed to protect flows down thewhole estuary rather than just locally to the weir itself and a short section of riverdownstream of it, it is very difficult to see how this option is somehow acceptable onenvironmental grounds.

Under drought order conditions allowing additional abstractions from the river aboveTeddington Weir, the existing outflow from the Mogden Sewage Treatment Workswould constitute a very high proportion of the fresh water flows passing down into theestuary. If this scheme goes ahead, much, if not most of this residual flow will bediverted and will be abstracted upstream. What is left will further reduce as customersare encouraged to save water and flows into the Mogden STW reduce. The overalleffect will be that the river downstream of Teddington will become more saline withpotentially serious ecological consequences.

The capital cost of what seems a relatively simple scheme on paper is given in thedWRMP Appendix A as £1.56 Billion which is more than a CCSTT scheme is likely tocost to build. These costs presumably reflects the cost of getting a pipeline underurban London and the cost of extra cleaning of the effluent to a higher standard.

Cotswold Canals Trust TW dWRMP 2019 Briefing - Rev 1.1 Page 4 of 14

Cotswold Canals Trust TW dWRMP 2019 Briefing - Rev 1.1 Page 5 of 14

3.2 The Abingdon Reservoir

A very large reservoir to be located near Abingdon has been proposed for manydecades and has a long history of going nowhere.

It is the subject of considerable opposition and has its own dedicated and wellresourced group known as GARD set up specifically to resist any attempts to build it.

A 100 billion litre incarnation of this reservoir was rejected at the Public Inquiry intoThames Water's WRMP09 and specific recommendations were made to look into aSevern – Thames transfer as an alternative with a specific recommendation that theone using the Cotswold Canals should be investigated properly. At the time, the costof the CCSTT was assessed by Jacobs, TW's own consulting engineers, at less than40% of the capital cost of the reservoir (£390 million vs £1 billion) – even taking intoaccount the full restoration of the canal for navigation as part of the process.

The new preferred programme in the dWRMP19 plan proposes a 150 billion litrereservoir (e.g. 50% bigger) and at a capital cost of £1.76 billion. At this size, itdelivers a comparable amount of water to a 300 Ml/d water transfer.

This figure seems very suspicious when compared with CCSTT increases.

Project 2010 Capital Cost 2017 Capital Cost Increase %

Abingdon Reservoir £1.0 billion £1.76 billion (for a 50%bigger reservoir

76% (or just 17%pro-rated)

Cotswold CanalsSevern-ThamesTransfer

£0.39 billion(240 Ml/d)

£1.84 billion*(300Ml/d)

470% (or 377%pro-rated)

* Deehurst Pipeline @ £1.26 billion (Appendix A) x 146% as given in table 6.3 on page 68 ofthe Raw Water Transfers Feasibility Report dated February 2018 which is the increase claimedif the restored canal is used instead of the pipeline .

For the pro-rated figures, this seems to suggest that the cost increases attributed tothe CCSTT are about 22 times higher than that applied to the reservoir. If thepositions were to be reversed, the reservoir would come out at about £5.7 billion andthe CCSTT scheme about £570 million.

This is a fairly crude analysis but is offered in the context of Thames Water redactingall the key figures needed for a better evaluation.

The reservoir suffers from a number of drawbacks in addition to the objections totaking up an area of land similar to the footprint of Heathrow Airport in theOxfordshire countryside. One of the main ones is delivery time. Although it is secondon the list in the preferred programme, it could not be brought forward quicklyenough to replace the Teddington DRA scheme were this to prove unacceptable.

GARD suggest that TW's keenness for the reservoir option is due to attractive taxbreaks and other financial advantages which can be maximised through high CAPEX(capital expenditure) projects. Another is that everything to do with it is within TW'sown geographical area which could encourage a bias in favour of convenience.

Cotswold Canals Trust TW dWRMP 2019 Briefing - Rev 1.1 Page 6 of 14

3.3 Water Reuse and Desalination

Another strand to the large scale water resource options is water reuse anddesalination. Both methods use reverse osmosis as a means of cleaning up the waterand are energy intensive.

There is a further technical issue in that the plant used to process the water cannotsimply be switched on when needed but deteriorates if not in near constant use.

The various water reuse/desalination schemes provide water for London but do notaddress the shortfall in the SWOX (Swindon & Oxfordshire) area. There is no realadvantage in these schemes unless they operate on water that otherwise would bedischarged below the tidal limit of the Thames.

TW has already built a desalination plant but it is understood that this has had its ownproblems.

Desalination results in an increase in salinity within the water body from which thebrackish or sea water is taken. Environmental issues are expected to limit thelocation(s) and potential capacity of these schemes.

Water reuse schemes generally clean up the output from sewage treatment worksrather than using sea or estuary water. This is easier than processing seawater.

The capital and operating costs of such schemes are high and they do not form part ofTW's preferred options this time round although they were more favoured in WRMP14.

The CAPEX for a 300Ml/d reuse scheme at Beckton is shown as £2.5 billion.

Were the Teddington DRA option to be dropped, one or more of these options could bepromoted in favour of a Severn-Thames Transfer (STT).

4. Severn-Thames Transfers

4.1 How it works

The basic concept is simple enough. The Severn is a big river and the west of thecountry gets more rainfall than the south-east and London. There are two keyquestions:

Is there enough water?

The short answer is yes, a lot of the time, but the longer answer is morecomplex. If there is a drought in the south-east, it is possible that rainfall hasbeen low in the west of the country as well and when the water is needed most,it may not be available from the Severn.

The River Severn is a regulated river. At times of low flow, it receives additionalwater from groundwater and reservoir sources. Work has confirmed that capacityis available, or can be made available, to supply additional water for transfer tothe Thames.

In addition to this, there are some other sources that could be brought into play

Cotswold Canals Trust TW dWRMP 2019 Briefing - Rev 1.1 Page 7 of 14

through additional development or redirecting water to the Severn that currentlyflows into other catchments. Some of this work has not been concluded and maybe one reason why the estimated contribution of the Severn-Thames transferschemes appears so low compared with the conveyance capacity of 300 Ml/d.

How do you get it to London?

There are two main contenders for the conveyance mechanism. TW's preferredone is a pipeline from Deerhurst near Tewkesbury to the Thames at Culham. Theother is using the Cotswold Canals.

TW have concluded that 300 Ml/d is the most appropriate conveyance capacitytaking into account supply and environmental constraints. Both the canal and apipeline can deliver this.

Both schemes have been costed with extensive water treatment included. This ismainly in an attempt to avoid the transfer of invasive species (nearly all of whichare both in the Thames and the Severn already anyway) or algae. In the case ofthe pipeline, it is also necessary to remove the potentially high silt load for whichthe River Severn is well known.

The schemes have also been loaded with the additional cost of piping the wateras far downstream as Culham downstream of Oxford. Earlier studies concludedthat the water could enter the Thames at Radcot (where the Deerhurst pipelinereaches the Thames). The reasons originally given for extending the pipe toCulham seem fairly straightforward to mitigate at a cost saving of perhaps £100million (base cost) in the case of the canal option and £70 million for the pipelineoption. Details of this alternative were provided to TW in May 2017 but not actedon.

The water once in the Thames can be abstracted downstream to fill the Londonreservoirs or put into supply. In the scheme(s) CCT are proposing, it would alsobe available to fill Farmoor Reservoir to help resolve the shortages in theSwindon & Oxford area (SWOX).

4.2 The Cotswold Canals Severn-Thames Transfer (CCSTT)

TW have carried out a lot of detailed work investigating the potential use of theCotswold Canals for a Severn-Thames transfer. The key conclusion is that it can bedone and all the potential issues are capable of mitigation. The detailed report into theCCSTT scheme can be found as Appendix G in the Raw Water Transfer FeasibilityReport February 2018 but all of the important costing information has been redacted.

Water coming down the Severn (supported by reservoirs or other sources asnecessary) would be pumped into Gloucester Docks and pass down the Gloucester &Sharpness to a point near Saul Junction (by which time, any silt will have settled out).It is important to understand that the G&S Canal is already used to supply up to about240 Ml/d for Bristol Water so there is nothing at this end that has not been donebefore.

Cotswold Canals Trust TW dWRMP 2019 Briefing - Rev 1.1 Page 8 of 14

Once at Saul, it would them be pumped up to the summit level of the T&S Canal intwo stages: Saul to Newtown where it would enter the canal, and then from Ryefordto the summit. The 4km pound between Newtown and Ryeford is the only one longenough for it to make it worthwhile using the canal channel on the western side. TheTW's published report advocates four pumping stations which are financially inefficientand increases the risk of interruption through breakdowns.

Rationalised pumping system on western side of the canal

Once at the summit, the water would flow under gravity to Inglesham. In the TWscheme, it would be cleaned up before entering a 32Km pipeline to Culham on theThames downstream of Oxford. CCT has provided a detailed report suggesting a farless expensive alternative to address the main reasons TW think this pipeline isneeded.

CCT would prefer that all of the canal is restored as an integral part of the project. TWargue that work to restore the canal would add costs and have a “do minimum” optionthat restores the channel where it is used (including bridges to navigation standards)but not the locks. Their level of restoration appears to be gold plated including therelining of the canal channel with a membrane/concrete liner. What is extraordinary isthat losses of something like 20Ml/d have then been assumed based on average canallosses. (The T&S Canal seems to have operated successfully with a pumped supply atthe summit of no more than 15 Ml/d over the drier summer months for much of itslife).

Another more recent variant would use a pipeline from Deerhust to Coates and restorethe eastern half of the canal. On the basis of the figures we have extrapolated, thiswould appear to have a less expensive CAPEX than TW's much longer pipeline route.However, it loses the advantage of the much lower elevation crossing of the Cotswoldsthat Sapperton Tunnel affords.

Cotswold Canals Trust TW dWRMP 2019 Briefing - Rev 1.1 Page 9 of 14

Cost

CCT were given a cost breakdown of CAPEX elements of the canal project and fromthis it is possible to extrapolate actual costs to some degree.

The absolute cost figures used in the Thames Water studies have not been provided soit has been necessary to extrapolate them. This suggests that the approximate baseCAPEX costs of the alternatives are as follows:

Option: Do All Do Minimum Deerhurst Pipeline

Extrapolated TW Cost:

£600 million £510 million £400 million

A review of the extrapolated costs and the design upon which these have been basedhas revealed some significant design optimisations. Whilst most of the individual costshave not been modified, some costed items seem to have been significantlyoverstated, a few have been understated and some are for work that we believe is notneeded. TW has been provided with our comments but there has been insufficienttime for these to be fully worked through.

The results with the design changes and the other adjustments yield the following:

Option: Do All Do Minimum New Canal Option

After CCT Changes: £405 million £325 million £290 million

The new canal option involves a pipeline from Deerhurst to Coates and the use of the

Cotswold Canals Trust TW dWRMP 2019 Briefing - Rev 1.1 Page 10 of 14

Cotswold Canals

eastern half of the canal. This does not provide all the energy efficiencies of theoptions which utilise Sapperton Tunnel but appears to result in a capital cost whichcompetes with the current much longer Deerhurst Pipeline proposal. In itself, thisoption does not create a new route for invasive species although CCT is scepticalabout the perceived risks of such as most are already present in both the Thames andthe Severn.

Underpinning a significant proportion of the reduction in the cost is an alternativemitigation scheme which would enable the navigable Thames from Inglesham nearLechlade to Culham to convey the transfer water volumes without adversely affectingthe more sensitive weir pools associated with the locks. The differences in cost aredramatic with 32000m of pipeline being replaced by simple and localised physicalmitigation where necessary and the sympathetic operation regime of existinginfrastructure at other sites.

The “Do All” and “Do Minimum” canal options can use only two pumping stations andpipelines instead of four at a significant reduction in cost. This also addressesconcerns about too many critical pieces of infrastructure in series.

There remain potential enhancements to the canal options to improve resilience andreduce the net OPEX (operating expenditure) costs/carbon footprint but these notcovered in any of the TW reports as the whole canal based scheme has not beenoptimised but only taken to a stage where TW felt they could drop the option in favourof the pipeline.

How TW have derived the costs they show in their Appendix A remains a mystery butthe base costs have been multiplied by risk and “optimism bias” figures. These are nottransparent and potentially provide a mechanism to create whatever outcome isdesired from the process.

As has been seen in section 3.2, the reservoir option seems to have been subject to a17% prorated increase since 2010 and the CCSTT option 377%.

If we use the £400 million extrapolated figure for the Deerhurst pipeline option andcompare that with the CAPEX figure of £1.26 billion given in Appendix A for thisscheme, the comparable increase is about 300%.

TW could overcome the obvious suspicion that the figures have been manipulated byreleasing, rather than redacting, those which they have used with full transparency.TW has chosen to not do so.

The OPEX figure attributed to the Severn-Thames pipeline transfer option is alsoextremely high at about £1.5 billion NPV (net present value). There is no breakdownof this available but it would be expected to include maintenance and pumping costsas well as any contributions paid to other water companies in connection withsupporting the additional flows in the River Severn. TW are suggesting that the latterare very high.

Previous studies have indicated that the CCSTT scheme would use far less energy thanthe pipeline – perhaps half as much. This is due to the lower elevation of SappertonTunnel and the much lower frictional losses associated with the much shorter length ofpipelines.

Cotswold Canals Trust TW dWRMP 2019 Briefing - Rev 1.1 Page 11 of 14

Without TW's figures, it has not been possible to prove if the reduction in energy costsoffsets, or more than offsets, the alleged additional cost of using the canal instead ofthe pipeline. However, the very high OPEX costs attributed to the pipeline STT optionstrongly suggest that the CCSTT option even including the fully restored canal, isactually a lower cost option than the TW favoured pipeline and those schemes with asimilar capacity currently within TW's preferred programme.

Benefits of the Canal Scheme

What is clear is that the CCSTT scheme offers far greater collateral benefits than anyof the other options able to deliver 250 – 300 Ml/d of water.

In addition to the obvious recreational benefits, the canal provides an environmentalcorridor which includes biodiversity connectivity (the ability of wildlife to travelbetween pockets of otherwise isolated habitat).

There are also economic and well-being benefits as well as rescuing many structuresof heritage importance including the canal itself.

TW's own consultations have shown that their customers like to see such collateralbenefits. Spending lots of money on more pipes buried in the ground does not seemto inspire.

The restoration of the Cotswold Canals is a project of national importance and itsrestoration features in all of the local authority plans along its route. This means that,rather than being opposed and frustrated at every step as has happened with theAbingdon Reservoir, the project is likely to be broadly welcomed so making its deliveryfairly straightforward.

Cotswold Canals Trust TW dWRMP 2019 Briefing - Rev 1.1 Page 12 of 14

5. National Water Grid – Alternative Concept

Water is an increasing valuable and limited resource which has resulted in many callsfor a national water grid. To implement this in an unsophisticated way would requirethe laying of extensive pipelines all over the country at a very high cost and involvinga great deal of disruption.

There is an alternative which would be both effective and far more affordable. TheCotswold Canals Severn – Thames water transfer scheme would form part of this.

It relies on the individual water companies using their resources available within theirown areas flexibly and intelligently so that raw water surpluses can be used in otherareas where resources are more limited.

The maps above illustrate how water resources can be distributed from one end of thecountry to the the other without extensive new pipelines.

The map on the left shows how two water resources are currently deployed and themap on the right shows how these might be redeployed to provide water into the driersouth-eastern part of the country including London.

The first resource to be identified is Lake Vyrnwy which was built in the 1880s tosupply Liverpool with clear water. This is now under the control of United Utilitieswhich operates in the north-west of England supplying places like Manchester andLiverpool. United Utilities has access to many resources of water including ones in theLake District which has a relatively high annual rainfall. It is relatively easy to supportflows in the River Severn using Lake Vyrnwy and much of the water currently used tosupply Liverpool can be replaced by other water resources available to United Utilities.

The second example is the outfall of the Minworth Sewage Treatment Works processesthe waste from the equivalent of 1.7million people. The cleaned up water from thisprocess eventually flows into the River Trent. However, the works is more or less on

Cotswold Canals Trust TW dWRMP 2019 Briefing - Rev 1.1 Page 13 of 14

the dividing line between the catchment area of the Trent and the Severn (via theAvon). Redirecting some of this water to the Severn would further support the transferto the south-east.

These are both resources that have been identified during Thames Water'sinvestigations into a Severn – Thames raw water transfer; there are others.

A Severn-Thames transfer has the potential to help resolve shortfalls likely to beencountered by other water companies in the south-east.

However, the development of strategic cross water company solutions is at risk ofbeing hampered by parochial interests and real or perceived commercial exploitationissues. This is something that OFWAT and the Government need to address if inferiorsolutions are to be avoided.

Cotswold Canals Trust TW dWRMP 2019 Briefing - Rev 1.1 Page 14 of 14