The Water Industry Commission for Scotland Scottish Water Scottish Water

Transcription

1 The Water Industry Commission for Scotland Scottish Water Scottish Water Reporter s Report SR10 2 ND draft Business Plan Appendix D Maintaining the Supply Demand Balance April 2009

2 SCOTTISH WATER REPORTER S REPORT SR10 2 nd DRAFT BUSINESS PLAN APPENDIX D MAINTAINING SUPPLY DEMAND BALANCE CONTENTS D1 INTRODUCTION 1 D1.1 General...1 D1.2 Scope of the audit...1 D2 SECTION 1: STRATEGY 3 D2.1 Key points to our findings...3 D2.2 Water resource planning and supply demand balance for SR D2.3 Specific questions arising from the Reporting Requirements...6 D2.4 SW s methodology for calculating its SRELL...9 D2.5 SW s methodology for calculating its supply demand balance...9 D3 SECTION 2: EXPENDITURE IMPLICATIONS OF MAINTAINING THE SUPPLY DEMAND BALANCE 20 D3.1 WG4 Increase security of supply in 15 water resource zones...20 D4 SW S METHODOLOGY FOR CALCULATING ITS LRELL 25 D4.1 Introduction...25 D4.2 Approach to the audit...26 D4.3 The methodology...26 D4.4 Final analysis...34 D4.5 Conclusions...36 (i)

3 Document issue details: B&V project no Client s reference no. SR10 2 nd draft Business Plan Version no. Issue date Issue status Distribution Issue 01 14/03/09 Draft Scottish Water Issue 02 02/04/09 Final WICS and Scottish Water Notice: This report was prepared by Black & Veatch Limited (BVL) solely for use by Scottish Water and the Water Industry Commission for Scotland (WICS). This report is not addressed to and may not be relied upon by any person or entity other than Scottish Water and WICS for any purpose without the prior written permission of BVL. BVL, its directors, employees and affiliated companies accept no responsibility or liability for reliance upon or use of this report (whether or not permitted) other than by Scottish Water and WICS for the purposes for which it was originally commissioned and prepared. In producing this report, BVL has relied upon information provided by others. The completeness or accuracy of this information is not guaranteed by BVL. (ii)

4 SCOTTISH WATER REPORTER S REPORT SR10 2ND DRAFT BUSINESS PLAN APPENDIX D MAINTAINING THE SUPPLY DEMAND BALANCE D1 INTRODUCTION D1.1 General This report on Scottish Water s SR10 Supply Demand Balance has been prepared in response to the Reporter Guidance set out in the SR10 guidance issued by WICS on 15 October The Reporting Requirements ask the Reporter s report to follow the same format as Scottish Water s submission, namely: Section 1: strategy; and Section 2: expenditure implications of maintaining the supply demand balance (SDB). Under the section on strategy we first answer the specific questions asked in the Reporting Requirements (RR). We then give a more detailed description on SW s methodology. We then discuss the expenditure implications of the strategy in Section 2. Commentary on SW s methodology for calculating its economic level of leakage is provided as a separate section. For the water service, in this Appendix we have limited our comments to work needed to improve the supply demand balance in deficit zones following a re-examination of yields and demands (Driver WG4). There is no essential expenditure related to reinstating any reduction in the available resources as a result of current examinations under the Water Framework Directive (Driver WR1 in the 1st draft Business Plan). For the wastewater service, SW is not proposing any works under SDB alone other than for development constraints ( M). As for the water service, this is discussed under the Enhancement Programme. D1.2 Scope of the audit Two meetings were held with the staff of SW s Water Resource Planning/Water Framework Directive Team (on 20 and 21 January 2009). We understood the basis of the current situation with water resource planning in Scotland, and the changes made to the 2 nd draft Business Plan since the audit of the 1 st draft Business Plan. Final versions of the tables were not available during the audits, although a complete version of Table 3 tables were available in draft only. Generally, therefore, it was only possible to understand and audit principles, (D-1)

5 methodologies and assumptions. Information was provided on the likely changes that would occur between draft and final version of tables. Although the audit was based on the latest snapshot position and not the final submitted tables, fifteen schemes have been agreed with DWQR (Technical Expressions) to restore the supply demand deficit in the zones. These schemes are 15 of the worst deficit zones identified using SOSI prioritisation and are included in the Essential programme. A further 17 Desirable schemes have been identified and confirmed by final SDB analysis for the 2 nd draft Business Plan. For 1 st draft Business Plan, schemes were included under drivers CS14 and WR1. However, for this plan expenditure under WR1, control of abstraction to protect downstream water bodies under the Water Framework Directive has not been included in the draft Ministerial Directives as an Essential objective and, therefore, no costs or outputs have been included in this draft. (D-2)

6 D2 SECTION 1: STRATEGY D2.1 Key points to our findings The key findings arising from our audits were: We believe that the information presented by SW gives a reasonable representation of the water resource situation in Scotland in 2007/08. Reductions in abstractions and increases to compensation flows under the Water Framework Directive have not been taken into account in the Water Resource Zone plans completed in 2008 and submitted to SEPA for approval. Expenditure under WR1 included in 1st draft Business Plan to address this environmental driver in nine named zones and study the problem in a further 27 named zones has been recategorized as desirable by SEPA and excluded from the second draft plan. Achieving SRELL across all Water Resource Zones together with 15 WG4 schemes which deliver additional leakage reduction towards LRELL and supply side interventions will significantly improve the Security of Supply Index by the end of the SR10 period. SW has updated WRP08 to the 2009 Water Resource Plan using the methodology agreed with SEPA. The guidelines are broadly in line with those for companies in England and Wales. Draft 2008 plans were issued to SEPA in May We understand that to date SEPA has not formally approved the WRP 2008 plan or provided comprehensive comments on WRP08 although they have stated in principle that they support the plan and the approach taken. Overall SW has applied a thorough and consistent approach to the analysis of the Supply Demand Balance, SOSI Index calculation and in developing the cost estimates for the 15 essential schemes. The audit reviewed the methodology and changes to it since 1st draft Business Plan. Since the 2nd draft Business Plan audit, SW has completed a sensitivity analysis of treatment works and trunk main losses assumptions. We agree that for the 31 zones analysed, including the 15 WG4 schemes, the supply/demand balance and proposed investments are insensitive to the assessed losses. The implications for the remaining zones in deficit will be reviewed during the AR09 audit. Four schemes were audited in detail from the SOSI calculation through engineering options to the preferred scheme and investment. SW are to be commended for the work, its thoroughness and the availability of the data and information during the audits and for the approach to providing additional information to support their plan between the 1st and 2nd draft Business Plans. As one would expect, methodology changes and data enhancements have tended to be focused on zones where investment has been identified to resolve a specific issue. The majority of factors used for investment zones analyses and some of the other zones, for example climate change, deployable output and treatment works losses, have been specifically derived from contractors reports. Some reports indicate that generic company wide conclusions should not be drawn from the study results, for other factors, the size of zones or plants in the study group may not be representative of the characteristics and mix of components. We, therefore, have a number of concerns relating to the application of the findings from reports which were commissioned for a specific set of assets generically to the 198 zones not included in the specific studies. Specific examples of mixtures of site specific and generic factors and two methodologies for the same component are discussed in the relevant sections below. In the longer term however, we recommend that SW extends the methodology and uses zone and site specific factors from detailed studies to all 230 zones. (D-3)

7 We recognise that there is a need for a wealth of detailed data required for the SoSI calculation and consequential prioritization analyses for investment and that collecting the additional data will take time. Since the 1st draft Business Plan, SW has increased the coverage of some studies and intends continuing towards 100% coverage. This is to be commended. However, until there is 100% coverage, issues relating to zone comparisons, prioritization and uncertainty will remain. SW believes that irrespective of the approach and methodology adopted the SR10 list of 32 priority deficit zones as classified by SoSI would not have varied substantially as demonstrated over the last year where data improvements did not change significantly the relative ranking of the zones. For some of the 230 zones, deficits are small and it is important that raw water mains and WTW losses and distribution system leakage are accurately calculated or assessed as they may be material to longer term investment decisions. We understand that SW is currently installing flow measurement on many of their raw water supplies. Establishing the combined raw water and treatment works losses by monitoring the difference between these meters and DI meters will better quantify the overall losses (raw water mains and WTW losses). However, it is equally important to understand the components of the losses, where loss reduction can be achieved and thereby savings can be achieved. It is, therefore, important to be able to monitor the two components independently by metering at the intake, inlet to the WTWs and the DI. However, due to practical installation issues, we understand that many of the meters will be located at the works inlet, thereby only allowing WTW losses to be measured. SW has prepared detailed studies for 15 water resource zones in supply demand deficit which are categorized as Essential and agreed with DWQR for inclusion in this Business Plan. For the 2nd draft Business Plan, SW has estimated a long run economic level of leakage (LRELL) which has been incorporated into the current SDB submission. This is discussed in more detail below. In calculating deployable output, SW has interpreted the definition of water resource system to include the capacity of the WTW. Thus, in some resource zones available headroom is limited by WTW capacity. We are unsure if WICS wants WTW constraints to be included in this table. However, SW s interpretation does give a true picture of its ability to supply water to its customers. D2.2 Water resource planning and supply demand balance for SR10 Since the 1 st draft Business Plan, SW has rationalised its company area into 230 water resource zones (WRZs) (from 239 in 2008), supplied by more than 500 sources. Some of these comprise single sources supplying just a few properties in remote areas. This is double the total number of WRZs in England and Wales and, therefore, makes the work of producing water resource zone plans onerous for a single company. Since SR06, SEPA has provided guidelines for the development of water resource plans. The specification is similar to that being adopted for plans being produced in England and Wales. Completed plans are submitted to SEPA for approval. (D-4)

8 The plans take account of the Water Framework Directive, the Birds and Habitats Directive and the CAR licensing system that has been implemented since the last Strategic Review. Licence issues have identified over abstraction at a number of sites. Seventy-eight zones are being funded during SR06 to achieve good ecological status and of these approximately 18 may require a physical intervention to be compliant. Using the SEPA guidelines, SW has assessed that 132 zones are currently in deficit according to supply-demand balance calculations based on Critical Period Demand and national criteria. 92 are calculated to be in deficit based on Dry Year Annual Average Demand. Draft 2009 Water Resource Plans have been updated from those prepared for 2008 and submitted to SEPA for approval. By 2010, SW reports that SR06 schemes and leakage reduction activities will reduce these zones to 48 supplying a population of approximately 893,000. By 2014, SW assesses that through achieving ELL but without investment to maintain the supply demand balance 32 zones supplying a population of approximately 822,000 will be in deficit based on dry year annual average calculations. The SR10 programme is intended to reduce the deficit in 15 of these zones (Essential expenditure) with a further 7 zones categorized as Desirable. The Security of Supply Index (SOSI) schemes to be implemented during SR10 are intended to move the Critical Period Index from -12 (in 2007/08) to 89 (in 2013/14) and the Dry Year Annual Average Index from -2 to 91 over the same period. This target includes planning assumptions on zonal leakage targets and delivery of quality and growth investment. The draft Ministerial Objective is to Ensure compliance with the duty imposed by Section 6 of the Water (Scotland) Act 1980 to provide a supply of wholesome water, by providing a level of service in water supply zones to be agreed with the Drinking Water Quality Regulator, such that Drought Orders in these zones are no more frequent that once in every 40 years. The planned output will be an improved SOSI to Band B for the Dry Year Annual Average Demand at the end of the investment period. The SOSI expenditure along with reduction in leakage to ELL will reduce the population in deficit to about 60,000. The deficits in the 15 zones have been investigated in detail, and a range of more than 70 options developed 28 of which were used to provide incremental cost information for use in the LRELL model. During the audit, draft tables were audited that indicated a SOSI of 90 in However, the tables were subject to finalization after the audit including minor changes and errors identified during the audit leading to a SoSI in 2014 of 91. For 1 st draft Business Plan, SW identified 36 zones for investigation and nine for implementation under the driver WR1, Water Framework Directive. As directed by SEPA there is no essential expenditure associated with this driver in the 2 nd draft Business Plan. The 15 WG4 schemes rely upon achieving SRELL in each zone together with a mixture of supply side capital schemes and further operational leakage expenditure towards LRELL and overlap between four expenditure programmes, SOSI, water quality, growth and capital maintenance. The SOSI capital expenditure is 61.0m, significantly lower than the sum in 1 st draft Business Plan of 145.6m, although the latter did include 33.6m for driver WR1 schemes. Supply demand calculations include a contribution from leakage reduction to SRELL in all zones plus a further contribution to LRELL to improve the balance for five zones; Black Esk WRZ, Fife WRZ, Inveraray WRZ, Inverness and Nairn WRZ and Afton, Braden and Penwhapple WRZ. Additional Opex costs to reduce leakage to the LRELL are 1.5m. With (D-5)

9 leakage there is always a risk that Active Leakage Control may not be able to deliver the required water savings within the ELL for the zone with a consequential potential need for an additional capital scheme to reduce the deficit. However, the current plan largely closes the SDB gap with supply side schemes. In the section on LRELL below, we note that additional leakage control might have been found to have been economic if the LRELL optimiser had been allowed to choose the optimal size of supply side projects. SW states that cost data for additional incremental options considered only 3 of the 15 WG4 zones indicated a potential for further long run leakage reduction. SW reports that this has not yet been validated and the results reported in the LRELL of 31 December 2008 updated. D2.3 Specific questions arising from the Reporting Requirements D2.3.1 Consistency of the information submitted with the ELL appraisal We confirm that the work on the supply demand balance for the 2 nd draft Business Plan took account of the output from the SRELL and LRELL analyses. D2.3.2 Consistency of the information submitted and explanations given with the draft water resource plan For this audit, draft 2009 WRPs were used as the base data. These plans were updated from the 2008 drafts prepared and submitted in 2008 to SEPA for approval. We confirm that information used in both plans is consistent. D2.3.3 Basis of population changes We give information on population projections in Section D2.3.4 Economic justification for any proposed changes in levels of service Previously, SW had different target levels of service in large and small water resource zones (1.50 yrs between drought orders in large zones and 1.30 years in small zones). For this business plan, SW made a business policy decision to adopt a single level of service of 1:40 years with the knowledge of SEPA. SW considers the target LoS to be a reasonable balance of environmental and customer service issues. D2.3.5 Any issues associated with the draft water resources management plan that remain unresolved We are not aware of any significant issues remaining on the draft WRMP, but have not discussed with SEPA any comments that it may have. SEPA have not informed SW of any significant issues. D2.3.6 Economics and practicality of the metering strategy In the SR06 period, SW is currently universally metering the majority of its non domestic customers. It has no plans to meter domestic customers. (D-6)

10 D2.3.7 Water efficiency strategy Water efficiency is a growing issue in England and Wales and companies have to develop water efficiency plans and targets. Ofwat has stated that it will impose water efficiency targets as part of the PR09 process. Similar water efficiency plans are not yet required in Scotland. SW states that their WRP09 which is due to be released at the end of March 2009 as a supplementary document to the 2 nd draft Business Plan contains a summary of their water efficiency strategy and other ongoing activity. D2.3.8 The robustness of SW s leakage strategy including its ELL assessment Historically, SW and its precursor water authorities had high leakage levels and only had limited leakage control infrastructure. Currently it has just completed a near universal coverage of DMAs (approximately 96.7%) and is developing its capability in active leakage control and pressure control. We are pleased to see that SW now has a dedicated leakage control group, and the necessary tools to rapidly reduce current leakage levels. We understand that currently SW is on target to meet or exceed its leakage target for 2008/9. This is an encouraging sign. In December 2008, SW completed an assessment of its LRELL. This followed best practice as set down in recent UKWIR guidance. As serious leakage control has only recently commenced base data for the analysis remains limited, although it is a significant improvement from the calculation of its SRELL in December We believe that the resulting estimate of LRELL is likely to be reasonable; although SW s resubmission in 2010 based on 2009/10 data is likely to be more robust as SW will then have 12 months data from all its DMAs. We give more information on SW s LRELL analysis in Section D2.4 below. D2.3.9 Consistency of the proposed supply demand outputs with the strategy We believe that the outputs proposed by SW under driver WG4 is consistent with its strategy. SW states that the draft Water Resource Plan 2008 has not been formally approved by SEPA and that, in discussions with SEPA on the plan and approach to the WRP09 plan, SEPA have not indicated any cause for concern or reasons for approval to be withheld. The WRP09 will be issued to SEPA on 31 March 2009 as a supplementary document to 2 nd draft Business Plan. D Sustainability reductions in the plan Driver WR1 which was included in the 1 st draft Business Plan has been reclassified by SEPA from Essential expenditure to the Desirable category for 2 nd draft Business Plan. Sustainability reductions have therefore not been included in the core plan. D Forecast of water delivered to different customer groups Forecasts of leakage are based on the LRELL as submitted to WIC in December (D-7)

11 SW has assumed that domestic unit water consumption will remain flat until 2018, growth in domestic demand being directly related to population growth. Although domestic consumption surveys commenced in 2008, the preliminary results have not been included in current forecasts as a longer period of stable data is required before data can be used with confidence. SW has assumed there will be no significant changes in the consumption of large nondomestic consumers beyond estimated reductions in consumption due to the current economic downturn as reported in the updated Experian Report (November 2008). From 2011/12 onwards, Experian suggest that consumption will revert to their earlier report profile. By the end of SR06, there will be approximately 125,000 non-domestic customer meters installed as a result of the metering policy being implemented during the period. For SR10 forecasts, SW has adopted an average unit annual consumption based on available readings from newly installed meters, provided by the retailer. The implication of the switch is to reduce consumption by this category of (unmetered) non-domestic consumer from an assessed 62 Ml/d in 2007/08 to about 28Ml/d from 2010 derived from a sample set of meter readings taken during 2008/09. The set of readings may not be statistically significant, but the conclusion is unlikely to be a gross underestimate of consumption. No growth has been assumed in demand during the SR10 period. For the SR10 period, we consider these assumptions to be reasonable, until conclusions from the domestic consumption survey and the longer term metering of the un-metered nondomestic category of consumers become available. D Baseline supply demand balance We believe that the baseline supply demand balance is consistent with ongoing leakage and demand management strategies. Current levels of leakage are calculated as the balancing item between distribution input, consumption and other uses. SW is on target to meet its leakage target for 2008/09. SW s assumed leakage levels are given in Table Appendix D line 32. D The effects of climate change At SEPAs request the effects of climate change have not been incorporated into deployable output calculations. This is discussed in more detail in Section D Climate change has been included in the calculations of target headroom, although inconsistently. SW has incorporated climate change uncertainty in the 20 WRZ where the 2003 UKWIR methodology has been used. In other zones the 1998 methodology has been used which makes an allowance for climate change but not to the same level of detail (See Section D2.5.8). SW state that for each investment zone site specific calculations have been used to assess the sensitivity and robustness of each zone to climate change. SW concluded that climate change is not materially significant to the SR10 investment for WR4. These analyses were not audited. (D-8)

12 D2.4 SW s methodology for calculating its SRELL Our commentary on SW s methodology is presented in Section D4 below. D2.5 SW s methodology for calculating its supply demand balance D2.5.1 SW s levels of service For SR10, SW has adopted a target level of service for a drought order of once every 40 years in any water resource zone. This is similar to the reference level of service in England and Wales, although companies can have different levels of service. This has decision has been arrived at in the knowledge of SEPA. SW has defined that hosepipe bans will be a precursor to a drought order, and only intends to implement rota cuts in extreme drought conditions. Therefore, there is no target return period for either measure. SW has not supported the change with any economic justification, but regards it as a rationalisation of its previous practice of having different levels of service in large and small WRZs to be more consistent with practices in England & Wales. We believe that a common level of service is equitable. D2.5.2 Methodology for Yield Assessment General In broad terms, the methodology for assessing water resources in Scotland follows the principles set out for England and Wales in the following documents: Water Resources and Supply: Agenda for Action (DoE, 1996); and A unified methodology for the determination of deployable output from water sources Volume 2, UKWIR (2000). For groundwater sources, assessments are based on analyses of source performance in historic droughts and the derivation of relationships between source output and aquifer drawdown. For surface water reservoir systems and conjunctive use schemes, the recommended approach involves behavioural analysis (i.e. simulation of the operation of a water resources system as currently configured using long time-series of naturalised historic data). In Scotland, the need for simulation modelling was addressed by the 2001 SNIFFER project. This is described in Surface water yield assessment ; volume 1 Final report, volume 2 Trial Sites Report, SNIFFER (2001). This project produced a software tool (Aquator-HYSIM) which enables the operation of a wide range of water resource systems to be simulated. The main features of the system such as reservoirs, pumping stations, treatment works, abstractions and transfers are represented within Aquator. The network links that connect these features are also represented although they are not modelled hydraulically. The model is driven by demand data supplied to model (D-9)

13 nodes that represent the main demand centres. Water availability is defined by long time series of river flow supplied to inflow nodes or reliable output figures for groundwater sources. The transfer of water from source to demand centre may be subject to constraints and control rules which mimic the realistic operation of the system. Where available, inflow sequences are provided by long time series of observed river flows. These are scarce in Scotland, and so the HYSIM rainfall-runoff model is used to either extend short observed flow sequences or derive new flow sequences from long rainfall records. Nevertheless, there are numerous catchments and sources in Scotland where neither suitable river flow nor rainfall records are available, and statistical techniques for estimating runoff from ungauged catchments have to be employed. In the past, CEH Report 108 (Gustard A., Bullock A. and Dixon J.M., Low flow estimation in the United Kingdom, Report no. 108, Institute of Hydrology, Wallingford, UK, December 1992) has generally been used for this. The report was tailored for use in Scotland by re-calibrating the flow prediction equations using only data for Scotland (WHS (2007), Recast of the Report 108 linking equation for predicting MAM7 from Q95 and SAAR, Wallingford HydroSolutions Ltd., July 2007). Recently, SEPA has obtained the Low Flow 2000 software package from CEH, which enables them to estimate low flows of any duration and return period for any location in Scotland. SEPA has recently produced guidance on the assessment of yields in Scotland: SEPA (2006), Water Resource Plan Guidance for Scottish Water, Final draft Version 1.0, 31st January The guidance is closely aligned to that issued by the Environment Agency in England and Wales, but has been tailored specifically for Scottish circumstances and legislation. The Guidelines state that for yield assessment, the Water Resource Plan should contain: a clearly documented recent assessment of deployable output for each resource zone, covering surface and groundwater sources and any conjunctive use; a clear statement of the level of service and associated operating rules; surface sources should normally include 30 days provision for emergency storage unless otherwise agreed with SEPA; and a statement of the yield assessment method used in each case, together with its limitations and uncertainties, particularly in respect of data. Application of the methodology by SW SW offices in Edinburgh were visited in April 2008 and January 2009 to discuss the yield assessment methodology used by SW in formulating its Water Resources Plan (WRP). Interim plans were produced in 2006 and 2007, but the current (2008) WRP is the first official plan produced by SW. The document itself was not seen. Nevertheless, the principles and methods of yield estimation used by SW staff were discussed to check their conformity to standard methods and SEPA guidance. (D-10)

14 Yield assessment methodology The majority of SW s supply comes from surface water sources, although groundwater is important, especially in Fife and Howdenwells WRZs. Before the Water Framework Directive gave rise to CAR (Controlled Activities Regulations) licences in 2005, groundwater abstractions were not licensed in Scotland but were subject to Water Orders.. Nevertheless, groundwater resources have been assessed using conventional measures such as pump tests and many sources have SRO (Source Reliable Output) diagrams as described in UKWIR s Unified Methodology. Most major demand centres are served by large surface water reservoirs and behavioural analysis in the form of HYSIM-Aquator modelling has been used to evaluate these. SW has developed 43 HYSIM-Aquator models. These cover 251 sources in more than 43 Resource Zones and are said to account for approximately 80% of SW s total output. Where insufficient data are available to provide model inputs or where system complexity does not merit modelling, source yields have been estimated using CEH Report 108 as modified for Scottish catchments. In previous assessments, the yield of some sources had been classified as unknown but in the current assessment the yield of all sources has been evaluated. It is understood that the original Report 108 can produce results that in some case are not consistent with Low Flows However, the revised Report 108 equation modified for Scottish catchments, which is now used, takes Low Flows 2000 data as an input thus resolving the inconsistency. There are still inconsistencies between HYSIM and Low Flows 2000 and further studies will be required in due course to iron out these inconsistencies. Climate change For behavioural analysis systems, SW has investigated the effect of climate change on yield using Method 2B as recommended by UKWIR, and also the Environment Agency s method 2X which requires fewer model runs. The rainfall and evaporation sequences used in the HYSIM modelling have been perturbed to produce new flow sequences. The 2X method works satisfactorily for run of river intakes but not so well for sources with storage (because all months in the sequence are perturbed and the effect is cumulative which overestimates the effect). As a result, 2B has been chosen as the preferred method. The effect of climate change is not included in the final supply-demand tables because SEPA wants DO to be quoted for non-climate change scenarios. In Scotland generally, climate change is predicted to result in drier than current conditions in only two months of the year (July and August). Thus, only the DO of river intakes of small storages with very short critical periods will be reduced. The yield of large storages may increase. So too will groundwater recharge. Climate change is not therefore regarded as a big issue in Scotland in this connection. For catchments using Report 108, information from UKWIR method 1B has been used. The Q95 flow is assumed to occur in either July or August, and so the August flow perturbation factor has therefore been applied to the Q95 estimate. (D-11)

15 Conclusions on Yield Assessment SW has made considerable progress in producing formal, quantitative estimates of source yield at all of their sources. Where possible, SW has employed behavioural analysis in the form of HYSIM-Aquator models to evaluate yield, in line with SEPA guidance. These models use long sequences of inflow data as required by the guidance; in some cases as long as 84 years ( ), but where long rainfall records are sparse, gridded rainfall data provided by the Meteorological Office has been used for the period CEH Report 108, as modified for use in Scotland, has been used on ungauged catchments where suitable observed data are not available. Further work may be needed where HYSIM flow estimates are not consistent with SEPA s Low Flows Although the best available techniques have been used on ungauged catchments where no observed data currently exist, the results should be regarded as approximate at best, and efforts made to install adequate monitoring as soon as possible. It is understood that SW are currently planning to extend their own monitoring network to supplement the gauges already operated by SEPA. D2.5.3 Deployable output and water available for use (WAFU) Water Available for Use (WAFU) is defined as: Deployable Output Outage Deployable output is either constrained by licences set by SEPA, by hydrological considerations (calculated by Aquator or the Low Flow Estimation System) or by raw water infrastructure both with an allowance for raw water transmission losses and WTW losses or by WTW capacity. Yield and licence with an allowance for losses, is reviewed against available treatment capacity and the lowest figure taken. This is then further reduced for outages if the WRZ is considered to have an annual average critical period. All agreed SR06 reductions in abstraction licences as a result of the Water Framework Directive have been incorporated into the SOSI calculation. This has no impact on the SOSI calculation as investment to restore the impact of lost Deployable Output is included in SR06. Any further investment to restore any further reduction in WAFU as a result of Water Framework Directive has not been included in the 2 nd draft Business Plan as the WR1 driver has been classified as Desirable by SEPA. There are likely to be further changes as abstraction monitoring becomes more comprehensive and over abstraction is addressed. In calculating deployable output, SW has interpreted the definition of water resource system to include the capacity of the WTW. Thus, in some resource zones available headroom is limited by WTW capacity. All 15 WG4 zones are yield constrained. As previously commented, we are unsure if WIC wants WTW constraints to be included in this table. However, SW s interpretation does give a true picture of its ability to supply water to its customers. (D-12)

16 D2.5.4 Outages Outage allowance has been applied to 43 Water Resource Zones, where the WRZ is considered to have an annual average critical period. In WRP06, outage allowance was assessed by adopting a percentage allowance based on consultants experience of working with Water Companies in England and Wales. In WRP08, SW sought to improve the estimation of the outage allowance and adopted the approach outlined in the UKWIR report Outage Allowances for Water Resource Planning, For the Draft Water Resource Plans, SW commissioned a contractor to analyse outage data for 38 WRZs in pilot studies across 7 geographical areas. The report, dated February 2008, concluded that 95%ile outage allowances ranged between 0 and 4.38% compared with the WRP06 assumed values of between 5% and 10%. Of the 38 zones, 9 zones have deployable output greater than 10 Ml/d and 2 zones over 100Ml/d. The outage allowances range from 0.06% to 1.06% with the values for the two large zones being 0.05% and 0.06%. The WRP06 assumed values for these zones were all 5%. Treatment works outage for 30 works within the zones was also estimated at between 0% and 1.55%. Seven treatment works in the study have deployable output greater than 10 Ml/d, one being over 100Ml/d; the majority of which the outage allowance was between 0% and 0.23%. The WRP06 assumed values for these works were between 3% and 5%. As there was limited reliable outage data available, SW commissioned a set of studies to collect the data. The analyses were based on workshops in compliance with UKWIR guidance on outage assessment. There were 7 workshops in total, covering multiple zones. Whilst expert knowledge of local operational staff was captured to determine the parameters required to estimate outage, a degree of subjectivity was also inherent in the method. We are aware that this approach has been used elsewhere in England and Wales. For SR10, SW has assumed that: The estimated outage value will be used for zones/works included in the study For other zones: o o 0.5% outage value for WRZs less than 1 Ml/d DO, and 3.0% outage value for WRZs greater than 1 Ml/d DO. These values represent a reduction from the 5% to 10% of WAFU used previously. In response to our comments in the 1 st draft Business Plan concerning the lack of statistical data to derive a significant value for zones with DO greater than 1 Ml/d, SW has carried out additional studies in 6 zones where the DO is greater than 1Ml/d. Three zones had DO less than 10 Ml/d and 3 zones had DO greater than 10Ml/d. The results of the studies are shown in the table below: WTW DO (Ml/d) % Outage Bradan Afton Glenlatterach (D-13)

17 WTW DO (Ml/d) % Outage Amlaird Penwhapple Coursehouse Benchmarking with other companies published information indicates a wide range of figures from a low of around 0.5% to a high of 8% or more. We commend SW for extending the works outage analyses to include larger zones and works and for implementing the full scale outage date collection trial in the Tayside/Fife zones. However, we still recommend that the programme of site specific studies be extended to all the larger sites, and that once the trial is complete a company wide outage information acquisition procedure and database be established to support future analyses. D2.5.5 Critical Period analysis The analysis approach has not changed from that used for the 1 st draft Business Plan. SW has completed critical period analyses for 20 WRP08 zones in line with SEPA s guidance for 2006 and the EA s Water Resources Planning Guidance April Currently, the critical period for the majority of Water Resource Zones is either the Dry Year Annual Average Demand (DYAA) or the Average Day Peak Week Demand (ADPW), the category being determined by the availability of raw or treated water storage. For three zones where SW assessed there to be sufficient storage to move away from ADPW, but not sufficient to allow use of DYAA, the Average Day Peak 3 Months was selected as the critical period. Where the critical period has been assessed as the ADPW (35 zones in supply/demand deficit), SW has assumed that planned and unplanned outages would not occur. In the case of planned outages, we concur as maintenance would be planned around critical periods. Previously, we have commented on the application of unplanned outages occurring in the period of the peak week demand. As detailed above, SW has completed a study of outage that confirms the low incidence of unplanned events. While it is true that unplanned outages are unlikely to occur during such a short critical period, it is still possible. D2.5.6 Measuring raw water mains and treatment works losses There has been no change to the methodology or supporting data used to estimate raw water mains and treatment works losses between the 1 st and 2 nd draft Business Plans. The raw water abstraction flow meter installation programme (WR5 driver) is due to be completed in 2009, after which SW proposes to evaluate transmission and treatment works losses from: Metered abstraction Distribution Input = (Transmission pipe + Works losses) SW reports that all survey work has been completed and installation work on the WR5 meter programme in SR06 will be complete by March The majority of meters will be located at the works inlet close to a power supply and telemetry network. Although these meters will be useful for assessing works losses, they will not improve the estimated losses in the transmission mains upstream. (D-14)

18 No new data are available to improve the estimates for the 2008 water resource plan. Although the meters should allow much better estimates to be made of overall raw water losses, there will still be calculation uncertainty in the split between transmission and works losses; with the potential for assumptions relating to say treatment works losses resulting in over or underestimates of transmission mains losses. See also our observations in the following sections. Raw water transmission losses SW has no information on losses from the raw water transmission mains and for SR10 water balance and SOSI calculations continues to apply an assumed overall average leakage of 21 m 3 /km/day; the figure having been derived historically for potable systems. The data for the water balance for AR08 implies that trunk main leakage is about 10 m 3 /km/day. From our experience, we would expect generic figures for trunk or transmission mains leakage to be less than from a distribution network. The calculation includes a length factor (increase) and pipe size and works capacity factor (increase or decrease). The estimated leakage from transmission mains must be considered uncertain. The estimates imply significant leakage in relation to the Distribution Input for some smaller resource zones that may be overly pessimistic and thereby distorting the water balance. The only way to estimate transmission mains losses with any certainty is to meter flows at either end of the pipe. If permanent meters can not be installed at both ends, a fitting should be provided for a temporary flow meter installation so that periodic measurements can be made. Otherwise, raw water transmission mains losses will continue to be uncertain with consequential impact on the water balance calculation. SW state that following the audit checks were carried to demonstrate the insensitivity of the SDB deficit to halving the raw water mains losses. As discussed above, we were given the opportunity to review the calculations, and we concur with their conclusions. Water treatment works losses Water Treatment Works (WTW) losses are assessed either using a table of losses for a range of generic treatment processes, or from treatment works specific measured losses, or following a works visit by an asset planner for a zone plan. The generic values are similar to those used in 2006, but with losses from spiral membranes and tubular membranes reduced. SW has presented evidence to support treatment works losses for coagulated filtration of 10% (range of works DO, 0.4Ml/d to 19 Ml/d), spiral membranes of 28% (range of works DO, Ml/d to Ml/d) and tubular membranes of 29% (range of works DO, to 0.45 Ml/d). The data for the membrane losses relate to very small capacity works. Taking into account works capacity, the weighted percentage losses for the three categories would be 6%, 19% and 28% respectively. All the generic figures appear high and are greater than we have observed elsewhere, although the weighted percentage losses for coagulation filtration and spiral membranes are more in line with our information. As for outage allowances, published data is variable. For a sample of companies, losses varied between zero and 11% with most WRZ in the band zero to 3%. We accept that losses from membrane plants are high. (D-15)

19 The asset register reports 333 treatment works of which 18 are abandoned, 114 are less than 0.2 Ml/d capacity and 52 are greater than 10 Ml/d capacity. Generic values have been used for the majority of works, based on the treatment processes on the site. For seven works, site specific losses were used and for a further 9 works the generic value was modified based on information available at the time. Twelve of the 16 adjustments reduced the assumed losses including at four works supplying Edinburgh and Lothians. The relatively high generic figures could result in the Deployable Output being underestimated which could be significant, specifically for the zones supplied from the 52 larger capacity works. Given the impact of these figures on zonal deficits we recommend that SW examines the basis of the figures in Table 3-2 Water Treatment Works Losses in its report Water Resource Plan April 2006 and the three modified figures used for SR10 (based on 2004/05 data). In particular, we suggest that SW analyses data for the range of works capacities and processes within their asset register and assesses whether the use of outliers and site or capacity specific data should be included in any calculation of average losses used in other zones. The omission of the outliers would more reasonably reconcile SW s estimation of outage with our information. We believe that further objective data and analysis is required on this important parameter. D2.5.7 Dry year distribution input Average daily distribution input (DI) is used as the denominator in the headroom calculation. As for previous returns, SW has taken the DI increased by 3% for a dry year as its denominator. This is consistent with the water resource plans. The dry year average day peak week is calculated by applying a peak factor to the dry year average DI. The peak factor has been calculated from DI flow records for each treatment works. We note that SW commissioned their consumption survey which will be used to update this figure in the longer term. In the meantime, SW has used an estimate based on the average England and Wales figure. We propose to audit SW s detailed analysis of the Dry Year Factor as part of the AR process. D2.5.8 Target headroom For the 2008 Water Resource Plan, SW commissioned a contractor to report on headroom uncertainty for 20 water resource zones. The zone studies adopted the 2003 UKWIR Methodology (An Improved Methodology for Assessing Headroom) Target headroom uncertainty including and excluding supply side climate change both vary between 4% and 12% of WAFU. The higher factors tend to be for the smaller and rural zones. Target headroom for the remaining zones has not been reassessed, and the factors derived using the 1998 UKWIR methodology at the megazone level have been used for SR10. The mega-zone analysis results in a 2010 target headroom range of between 6% of WAFU for the larger water resource zones of the Central Belt to 9% for the rural and smaller zones. These factors are generally slightly lower than those calculated by the 2003 methodology. The contractor recommended that a higher level of risk be accepted for future years, 90%ile from 2007 to 2015, reducing to 70%ile by This is a similar approach to those that we have seen elsewhere. Limited comparison can be made between the 1998 mega-zones and (D-16)

20 2003 zones. However, the 2003 methodology for Fife zone increases the uncertainty factor slightly. As we have previously commented, from our experience these factors are not unusual, although we would not expect all small zones necessarily to have high headroom factors. Overall, we believe that the chosen headrooms are acceptable for the current calculations. The UKWIR 1998 methodology incorporates an allowance for climate change. However, apart from where the sensitivity of supply side climate change has been considered for the 2003 factors, SW has stated that the supply and demand calculations do not include a specific allowance for climate change. PR09 business plans for companies in England and Wales are including the impact of climate change in water resources assessments. Although the impact of climate change for Scotland may be very different to that expected in areas of England and Wales, this omission represents an uncertainty for all resource assessments. It is important that target headroom is carefully considered. Where demand is flat it may be the sole driver for investment. The 15 WRZs in the 2 nd draft Business Plan are in current deficit now even without the addition of headroom. The considerations above are not therefore relevant to those zones. As discussed in D above, SW states that for each investment zone site specific calculations have been used to assess the sensitivity and robustness of each zone to climate change. SW concluded that climate change is not materially significant to the SR10 investment for WR4. Although we accept SW s statement, these analyses were not audited. D2.5.9 Population and household projections SW has used a consistent methodology for population forecasting for all sections of the 2 nd draft Business Plan. The approach uses data from: GROS - Projected Population of Scotland (2006-based) GROS - Projected private household population, GROS based household projections by household type and age group for all 32 local authority areas GROS based projected population by sex and single year of age, Scotland Government Actuary s Department WIC4 Reports Billing and Household data (September 2008) and 2004 C-tax base reports, Growth rates 07 to 17 Experian Water Demand Forecasts (February 2008) Experian Water Demand Forecasts - Autumn 2008 (26 November 2008) NHBC New House Building Statistics, Annual Returns 2007/08 (D-17)