OUR PLAN Representations on the Draft Determination October 2014 I NN

Transcription

1 OUR PLAN Representations on the Draft Determination October 2014 I NN OVA TI ON ON ATI OR AB LL CO TRAN SFORMATION

2

3 Executive Summary Executive Summary Our Representations Cost threshold adjustments 2.1 Water cost threshold Unmodelled costs Supply demand balance inputs Third party costs 2.2 Wastewater cost threshold Ofwat's modelling approach Wastewater Growth Customer bill profile and financeability Retail costs CEF engagement Cost of Capital Response to Ofwat's Specified Requirements Risk and reward 7.1 ODIs 7.2 RORE FD09 service standards Serviceability Customer engagement 10.1 Non-household retail control 10.2 Financeability tools 10.3 In-period bill variation Retail costs Wholesale charges

4

5 1 1 Executive Summary Introduction Throughout the Price Review process we have been keen to understand the needs and priorities of our customers and to reflect their views in the building of our plans - customers told us that they expect a safe, clean water supply and a reliable wastewater service. Ofwat has recognised our customer engagement, and development of outcomes, as being exceptional. The views of our customers and other stakeholders have been fundamental in the development of our outcomes for customers and the environment. Our draft plan, published in July 2013, proposed keeping bills flat before inflation, while investing to maintain and improve services, meet legal obligations and tackle future challenges. Over 90% of customers surveyed said that they found the plan acceptable. We could have stopped there, but customers also told us they expect to see reduced leakage, and improved security of future water supplies to meet the challenge of population growth and changing, more extreme weather patterns as well as meeting their expectation that bills will be fair and affordable. In arriving at our December Business Plan we did not take away or reduce any of the outcomes or commitments that customers had valued, but we reduced profits, reduced headroom against financial ratios and challenged ourselves to deliver more efficiencies, both in the wholesale business and retail services. Increases in bills were held to 1.8% p.a. below inflation (1.55% p.a. after adjusting for the abatement of K in ) - the largest real terms reduction of any water and sewerage company. We could have reduced customers' bills in the short term even further by investing less to reduce leakage and improve resilience of supplies - but customers are very clear that they want us to take a long term approach to making the most of our water resources and to ensuring a resilient service. Ofwat's feedback from its Risk Based Review was very positive about all material aspects of our plan except for our proposed costs, which were higher than Ofwat's predicted costs. In our revised June plan we made some changes but we remained true to ensuring that we were still able to deliver in the areas that mattered most to customers, while continuing to reduce bills, in real terms, by more than any other water and sewerage company. We provided evidence to demonstrate that the remaining gap between our costs and Ofwat's predicted costs were clearly attributable to key elements of our plan where we were going further than other companies, and beyond what we have done previously in three key areas: Meeting new legal obligations for ecological improvements; Reducing leakage below the economic level to meet customers' clear expectations; and Increasing investment in resilience. And we also provided evidence where Ofwat had not fully reflected the impact of: Delivering our final water resource management plan, going beyond what other companies are planning; and Investing in sewerage infrastructure and growth to meet the needs of growth in our region.

6 2 Executive Summary Executive Summary We are pleased that Ofwat has responded positively to our revised plan and the evidence that we have provided in arriving at the Draft Determination. We think there are still some gaps to be filled and we have provided further evidence to enable Ofwat to make further adjustments in arriving at their Final Determination. As discussed in our recent meeting with Sonia Brown and the Ofwat team, in making these representations, our approach has been to focus on those areas that are company specific and material to us delivering what matters most to customers and to meet legal obligations: Remaining differences on cost thresholds A small number of specific points on Outcome Delivery Incentives Retail cost base The profile of customer bills and financeability Considerations for Ofwat's review of the cost of capital. We have carried out significant further customer engagement and are responding to specific requests from Ofwat for further information. Cost thresholds We are pleased that Ofwat's Draft Determination recognises the very high quality of the evidence that we provided, including the additional external assurance, to support the cost of delivering our plans and that Ofwat increased its cost thresholds as a result. However, as the Draft Determination notes, there remains a gap in relation to wholesale water costs and the evidence that we had provided on the impact of growth on sewerage costs was only partially accepted. We are providing further evidence to enable Ofwat to further close the gap on water totex costs and to further increase the threshold on sewerage. In particular: PR14 Representations on the Draft Determination anglianwater.co.uk The adjustments in the Draft Determination reflected special cost claims that fully met Ofwat's four evidence tests. Ofwat fully accepted the evidence for ecological improvements and investment in dual supplies for populations above 30,000 and an adjustment was made for the full claim in the "unmodelled cost model". This resulted in an increase in the cost threshold of one third of the additional cost above the implicit allowance. In our meeting with Ofwat in April and in a subsequent letter (1) we were assured that where costs are "significantly different from business as usual and supported by compelling evidence then companies can make a full claim for the extra costs involved." Following discussion at our most recent meeting with Sonia Brown and the Ofwat team, we are providing additional evidence to demonstrate that these additional costs are not reflected in either the "unit cost model" or the "full top down totex model" as the investment is of a different scale to that seen in the past or proposed by other companies. Our evidence supports the remaining two thirds of the additional costs, of 31m, being reflected in the water cost threshold. In reflecting our Final Water Resource Management Plan (FWRMP) in the Draft Determination cost thresholds, Ofwat has explicitly excluded from its calculations enhanced metering (2) and our proposed increase in the number of water efficiency audits and retrofitting of water efficient devices. Ofwat also assumed a level of meter 1 Letter from Sonia Brown to Peter Simpson, 8 May 2014, in response to John Earwaker's comment that Ofwat's cost assessment methodology has been set up in such a way as to pass a maximum of 33p on to consumers. 2 Enhanced metering is our terminology for our strategy of compulsorily fitting meters and then giving customers the choice to switch to a metered basis of charge or switching on change of occupancy. Our plan is to increase the number of households that have a meter fitted to 95% by 2020, with an expectation that 88% of customers will pay a measured charge.

7 3 optants that was inconsistent with excluding the enhanced metering programme. Metering and water efficiency have been key elements of our water supply strategy since privatisation and were at the centre of our Strategic Direction Statement and our strategies for dealing with climate change and growth. We are providing evidence that both metering and water efficiency are a key part of our FWRMP, are cost beneficial and have strong support from Government, customers and the independent Customer Engagement Forum and Advisory Panels. Reflecting the water demand strategies in the "unit cost model" and "full top down totex model" adds 18m to the water cost threshold. Ofwat's assessment of water third party costs is based on an historical analysis of costs as a percentage of third party revenue which included revenue from non-potable supplies. Our forecasts for total third party income in to excluded non-potable revenue as we assumed it would be part of the wholesale revenue and controlled by default tariffs. For the purposes of making an accurate estimate of our AMP6 third party costs, we have now added the revenue from non-potable supplies to our AMP6 third party revenue forecasts. Applying the historical cost recovery percentage to the consistent forecast increases the amount included in the cost threshold for third party costs from 9.9m to 25.2m, an increase of 15.3m. Ofwat fully accepted the need for additional investment in sewage treatment as a result of forecast growth. The cost benefit of proposed solutions and robustness of costs was only partially accepted because the unit cost per increase in population is higher than the industry average. We are providing evidence that a more appropriate basis for calculating a unit cost benchmark is the increase in population equivalent capacity provided and on this measure our unit costs are well below average. Our evidence supports an increase in the sewerage cost threshold of 40m. We present reasons why our wastewater threshold should not be reduced by the unmodelled adjustment of 18.6m. With the introduction of an overall 5% cap on variance from thresholds, it is no longer appropriate to also have a cap within the assessment of unmodelled costs. Overall our representations support an increase in the water cost threshold of 64m and the wastewater cost threshold of 59m. Outcome Delivery Incentives Ofwat intervened in the plan to adjust Committed Performance Levels (CPLs) and associated Outcome Delivery Incentives (ODIs). Some of the interventions are based on new evidence of other companies' performance in and consistency checks carried out by Ofwat across companies that was not available to us at the time we submitted our Revised Plan in June. We do not think that new CPLs and ODIs imposed for drinking water quality, pollution incidents and environmental performance are necessary for the protection of customers because there are clear enforcement mechanisms available to the quality regulators and in any case there are quality metrics embedded within our serviceability measures. However, we understand that Ofwat wishes to see some consistency in quality measures and as this is not a company specific issue we are not making representations on these. We have made a proposal on the scale of rewards and penalties for the new measures.

8 4 Executive Summary Executive Summary Supply interruptions We agree that supply interruptions should be a focus for the ODIs and that rewards should only be earned for upper quartile performance. However, we do not consider that the proposed performance commitment and design of the incentive is appropriate. There is significant variability in the performance of individual companies between years: Ofwat's proposed required performance level of 10 minutes is not based on the consistent performance across three years of a single company. Instead Ofwat has taken the upper quartile performance in each year irrespective of the average performance of that company. Based on consistent performance across three years the resulting performance commitment would be 12.3 minutes. Taking Ofwat's approach, the company (Dee Valley) with an average three year performance at the upper quartile over three years of 12.3 minutes would have incurred a maximum penalty in two of the three years. Thames with a consistent performance at 12.5 minutes and minimal variability would incur a maximum penalty in every year. The assessment of upper quartile performance level needs to be calibrated to reflect the nature of a company's network. There is a clear correlation between supply interruptions (minutes) per property and length of main per property: PR14 Representations on the Draft Determination anglianwater.co.uk

9 5 This should not be a surprise - it is well established, in the way that Ofwat has previously assessed service associated with burst mains, that there is a correlation between bursts and length of main (3). Companies that have longer lengths of main per property would be expected to have more bursts per property and consequently more interruptions per customer. In the Draft Determinations Ofwat has applied an overly simplistic performance level. As illustrated in the chart above, the ten minutes performance achieved by the "upper quartile" companies is largely a factor of their networks. We propose that the performance level should be based on the upper quartile performance adjusted for length of main per property. This would result in a committed performance level for Anglian Water of 17 minutes. This performance level remains challenging, especially given the normal level of variability seen in this measure year on year. Nevertheless, if this adjusted performance level is reflected in the Final Determination then the challenge could be set of using it as a basis for rewards and penalties from (i.e. no glidepath) if using an annual target or, as we propose, through an interim target based on a three year average, which implies performance at 17 minutes from onwards. The Draft Determination provides a very narrow range over which the rewards and penalties apply within a single year assessment. We believe that this will result in perverse incentives and be to the detriment of customers. Assuming the run-rate for supply interruptions can be reduced to meet the performance commitment, then a single one-off event can result in the performance level being missed and the question then is whether the penalty is proportionate? To put this into context a single interruption of 12 hours, affecting 3,700 customers, could result in the full penalty of over 8m being incurred. If such an event occurs at the beginning of a year, then there could also be a temptation for companies that are less focused on doing the right thing for customers to reduce the response to interruptions to save cost in the rest of the year. We propose the use of a three year average to better maintain the incentives to perform. Retail costs In the Draft Determinations, Ofwat reduced the proposed costs of providing retail services by deflating costs by the increase in the RPI index between and This seems inconsistent with Ofwat's premise that RPI is not an appropriate index of costs for retail services and it results in a cost base lower than actual nominal cost of providing services in any year between and (or expected in ). To be consistent with Ofwat's stated methodology, which we applied in developing our plan, nominal costs in should be used as the basis for the determinations. This will add 9.4m to the Household Retail Control and 1.2m to the Non-household Retail default tariffs. Profile of customers' bills and financeability Customers have consistently told us that they prefer a stable bill profile compared with volatility year on year. In particular they do not like bills going down in one year for them to increase in the following year. This clear message underpinned the profile of bills in both our original and updated plans. One of the ways that we achieved this was through the even phasing of the CIS outperformance adjustment across the five years of AMP6, on an NPV neutral basis. Our approach was consistent with the approach taken in recovering capex outperformance on 3 Ofwat's measure of service performance in June Returns (JR11, Table 11, line 12) was bursts per 1000km of main

10 6 Executive Summary Executive Summary a rolling basis in previous price reviews, as well as resulting in a steady state in bills moving in to AMP7. This was a particular focus for the Customer Engagement Forum. In assessing our tests for financeability in our plan, we adjusted PAYG to balance the profile of customers' bills to ensure our financial ratios were only just acceptable. We highlighted in our plan that any changes to the plan would mean that we would need to reconsider both the level and profile of the ratios. In addition, our financeability testing and resulting PAYG requirement was applied having already established a smooth profile for the CIS adjustment. In the Draft Determination Ofwat made a "midnight" adjustment to reflect CIS outperformance. The effect of this is to reduce revenues in AMP6 by 82m but to increase RCV at the end of the period by 98m, resulting in current customers paying less but future customers paying more. While there could be a technical argument about whether the CIS adjustment should be applied as a midnight adjustment or spread over the period, the real question is about inter-generational equity and whether customers prefer a bill profile that is stable both within and between AMP periods. The effect of Ofwat's approach is also to further tighten financeability. We have re-tested the profile of bills, and impact on financeability, with the Customer Engagement Forum and the four Independent Advisory Panels. Together we accept the initial reduction in bills in as this seems to be a feature of all of the Draft Determinations. We then explored the options of either (1) setting bills as low as possible between 2015 and 2020 in line with Ofwat's Draft Determination or (2) avoiding sharp increases in bills in 2021 and setting bills so costs are spread fairly between current customers (who pay their bills in 2015 to 2020) and future customers (after 2020). We also illustrated the potential impact on bills of the additional cost of financing that would arise in the event of a further one notch downgrade in credit rating. PR14 Representations on the Draft Determination anglianwater.co.uk Average Household Bills 2015 to The conclusion of the CEF and the Advisory Panels was that the second option was preferable and that this should be tested with a survey of customers. The survey demonstrated clear support (72% of customers surveyed) for option 2. To achieve the preferred profile we are therefore proposing that the RCV run-off rate is increased from 3.95% to 4.30% for water and from 3.85% to 4.20% for sewerage, resulting in an increase in revenues of 82m and a reduction in closing RCV of 98m. Review of cost of capital In publishing the Draft Determinations, Ofwat has emphasised that it may review the cost of capital for Final Determinations to take account of market and regulatory developments. Year Risk Option 1 Option 2

11 7 We do not believe that there is any evidence to support a change in the cost of equity assumption and Ofwat's assumptions are very much in line with recent regulatory precedent. Whilst there has been a fall in the cost of raising debt over this year we doubt whether this will be sustained. There has been a notable upturn in gilt rates since August 2014 which tends to at least cast doubt on whether the lower rates in the first part of 2014 will endure. Investors broadly welcomed the early clarity on Ofwat's views on the cost of capital in the Risk and Reward Guidance, and noted that these assumptions remained unchanged seven months later at the Draft Determinations. For the benefit of future customers it is vital that long term stability is not jeopardised in seeking a marginal additional benefit based on short-term volatility in the market. Ofwat has to balance the benefits that a lower assumed cost of capital in the short term delivers in immediate bill reductions for customers, against ensuring that companies can remain financeable, and satisfy debt covenants and credit ratio hurdles required by investors and rating agencies. A reduction in the cost of capital at this late stage discourages companies from taking a long term view and taking prudent action to manage their costs. In preparing our business plans in 2013, we locked into forward market rates to cover a significant proportion of our debt-raising for AMP6. A further fall in interest rates could not have been predicted at the time we submitted both our December and June submissions, and when we prudently hedged our forward debt issuance to mitigate the risks of rates increasing. We provide further analysis in our representations and question whether reducing the cost of capital is in the long term interests of customers and the efficient financing of the sector. Additional information and further customer engagement Ofwat has set out specific areas where we need to provide more information: Respond to the ODI interventions, including Ofwat's suggestions for additional performance commitments, and consider the effect on RORE: we have provided, as requested, updated tables A20 and A20a and the Risk Assessment Tool. Provide assurance over the cost allocation of network and non-network calls for the household retail control: PricewaterhouseCoopers (PwC) has provided additional assurance on the analysis of underlying data that was used to allocate calls between the retail price controls. Provide an explanation why non-household opex is higher than the preceding two years. The increase results from a change in our policy towards non-household bad debt relating to the treatment of credits on account. Our change in approach applies from onwards and the figures in that year are therefore an appropriate baseline for these costs going forward. Provide evidence for the delivery of FD09 service standards for resilience and odour: evidence is provided that demonstrates service standards have been achieved. Provide an update on performance for burst mains and flooding other causes, to demonstrate stable serviceability: evidence is provided that supports a continuing stable assessment.

12 8 Executive Summary Executive Summary Provide views on the length and form of the non-household control: we are confident that the allocation of costs we have provided is robust but we are implementing new systems that will allow an increase in the direct charging of costs and on that basis we are proposing that we have the option to re-open default tariffs in 2017, if justified by new information. Re-consider the RPI sharing mechanism: we considered the position in June and withdrew the mechanism in the light of the reduction in the financial headroom and a one-notch lower target credit rating. The Draft Determination further increases the risk to the business compared with our June plan as a result of the exclusion of our proposed risk mechanism for growth and further NEP obligations; and new/amended ODIs that are significantly more likely to result in penalties and a much lower prospect of financial rewards. We expect that Ofwat will respond positively to our representations and that revenues will be higher in the Final Determination but the increased risk associated with growth and the NEP programme will remain; as well as some increased risk remaining from changes in ODI rewards and penalties. On this basis the Board is not proposing to re-instate the RPI sharing mechanism. Complete the excel wholesale charges template and highlight any remaining aspects of wholesale charges that do not fit into the charging structure: We have completed the spreadsheet and there are no material issues to highlight. A copy of a letter to Ofwat on our charging proposals is included for ease of reference. We have also engaged further with customers: On the use of financeability tools. We have engaged with customers on the profile of bills within AMP5 and into AMP6, having used depreciation and PAYG tools to derive options for bill profiles: 72% of customers support a smooth profile of bills between and Re-engage with customers over the principle of in-period bill changes, and provide an updated draft licence modification to enable this to happen: 80% of customers support in period adjustments, and 71% support the scale required to deliver our proposed leakage reduction of 172 Ml/d. We have also provided a draft licence modification to support the necessary changes. PR14 Representations on the Draft Determination anglianwater.co.uk Demonstrate engagement with customers on the re-opening of default tariffs: we have engaged with the Customer Engagement Forum and Independent Advisory panels, which include Business customers, and there is clear support for the option to re-open tariffs. Our Draft Determination has been discussed in detail with the Anglian Independent Customer Engagement Forum and Advisory Panels. They will be submitting a separate report to Ofwat setting out their views on the Draft Determination. Assurance, monitoring and reporting obligation Ofwat has set out its plans for assurance, monitoring and reporting obligations for companies in its Draft Determination. We support the principle that assurance should be appropriate and tailored to each company, and that this leads to a proportionate approach.

13 9 We were pleased that the quality of our Board assurance was recognised in the Risk Based Review and attracted one 'exceptional' and two 'acceptable' ratings. We are also pleased that in its Draft Determination Ofwat is '...satisfied with the company's proposals for self-reporting.' and believes that we should be able to provide self assurance for outcome delivery. Board Assurance The Board of Anglian Water Services has reviewed and approved these representations for submission to Ofwat. The Board has also assured itself that the business remains financeable in the light of the Draft Determination, assuming that Ofwat reflects our representations in the Final Determination. Summary In summary, our representations provide evidence to support an increase in revenues of 207m compared with the Draft Determination: Chapter Summary of representation Impact on cost thresholds (where applicable) Impact on revenue requirement for Final Determination Impact on 2020 closing RCV Water Wholesale price control Allow 100% of the costs of ecological improvements and network resilience: evidence provided that the two top-down totex models do not reflect additional costs because the obligations did not exist in AMP5 or the scale of expenditure has increased m + 19m + 8m Evidence provided that enhanced metering programme and increased water efficiency are a key part of Final WRMP and are supported by customers and government: programmes should be reflected in the supply demand unit cost model and totex model C m + 11m + 5m

14 10 Executive Summary Executive Summary Chapter Summary of representation Impact on cost thresholds (where applicable) Impact on revenue requirement for Final Determination Impact on 2020 closing RCV Third party costs to be updated to reflect third party revenue, which includes revenue from non-potable supplies, to be consistent with the basis on which costs as a percentage of revenue was determined m + 9m + 4m 7.1 Outcome Delivery Incentives: Supply Interruptions: the upper quartile should be adjusted to reflect a single company performance over three years and adjusted for network density. The range over which rewards and penalties apply should be widened. n/a None None Wastewater wholesale price control The introduction of an overall 5% cap on variance from thresholds means it is no longer appropriate to also have a cap within the assessment of unmodelled costs m + 11m + 6m PR14 Representations on the Draft Determination anglianwater.co.uk Allow additional costs of growth: unit cost comparison with other companies should be based on unit cost per increase in PE capacity rather than growth in population in period m + 24m Profile of customer bills and financeability (affects both water and wastewater wholesale controls) 3 Bills should be smoothed between AMP6 and AMP7, and within AMP6 between and This is achieved by increasing the old RCV run-off rates. n/a + 82m + 12m - 98m

15 11 Chapter Summary of representation Impact on cost thresholds (where applicable) Impact on revenue requirement for Final Determination Impact on 2020 closing RCV Retail price controls 4 Retail costs should be set at cost base. Increase in: Household costs 9.4m Non-household costs 1.2m n/a + 10m n/a Taxation N/a We have estimated the impact of the above representations on the revenues allowed for taxation. n/a + 41m n/a Total m + 207m - 63m The impact on customer bills will be as set out in the table below. The bill is higher in than proposed in our June plan in large part as a result of the step reduction in bills in and the resulting change in the profile of bills Ofwat's view of our June Business Plan Ofwat's Draft Determination After adjusting for representations

16 12 Executive Summary Executive Summary PR14 Representations on the Draft Determination anglianwater.co.uk

17 13 2 Cost threshold adjustments Introduction We welcome the changes Ofwat has made to its cost models and to the cost thresholds for Anglian Water. However, we do not consider that Ofwat has adequately explained the reasons for the discontinuity in its PR14 modelling compared with previous relative efficiency assessments (see pages and of our June submission). For the purposes of these representations there are still a small number of areas, which we believe are material, where we propose that cost thresholds should be adjusted. The following two tables summarise the impact of these representations, and describe where in the following sections the detail supporting our representations can be found. Summary of proposed changes to Water threshold m June Business Plan Draft Determination BP v DD Section Starting point 1, , % Representations regarding RSA, Eels & mains resilience Section Unmodelled costs Representations regarding SDB enhancements & meters Section Supply demand balance inputs Representations regarding third party costs Section Third party costs Proposed cost threshold 1, % Summary of proposed changes to Wastewater threshold m June Business Plan Draft Determination BP v DD Section Starting point 2, , % Representations regarding 5% cap Section Ofwat's modelling approach Representations regarding wastewater growth Section Wastewater growth Proposed cost threshold 2, %

18 14 Our Representations Cost threshold adjustments 2.1 Water cost threshold Unmodelled costs Ofwat's Draft Determination In the Draft Determination (DD) of our wholesale water price control, Ofwat accepted the arguments we put forward in the June Business Plan (BP) submission regarding investment for restoring sustainable abstractions (RSA), to comply with the requirement of the Eels Directive and providing resilient water supplies for populations over 30,000. For the bottom up modelling, these two areas of cost fell into the category referred to by Ofwat (in relation to the bottom up modelling) as unmodelled. These summed to an additional 46.0m more than the Implicit Allowances already made for the relevant drivers. The additional allowances relevant to the bottom up model were then subject to the 'one third rule' on the basis that allowances for the expenditure had already been made in the two top down totex models that contribute to the threshold. As a consequence, only a third of the aggregate additional allowances ( 15.3m) were added to the cost threshold. Summary of our Representations We present evidence to support that 100% of the additional allowances for RSA, Eels and network resilience be added to the water cost threshold. These drivers are not captured in the two top-down totex models because the obligation did not exist in AMP5 or the scale of expenditure was materially lower. Thus the amount to be added to the water threshold should be 30.7m over and above the 15.3m already included, or 46.0m in total. PR14 Representations on the Draft Determination anglianwater.co.uk Ofwat's approach The diagram below shows the logic followed by Ofwat in arriving at the wholesale water cost threshold. Essentially there are three totex models: A top-down full model (model C) A top-down refined model (averaged output of models B & D) A bottom-up model consisting of: Base totex botex (averaged output of models E & F) Three unit cost models (covering supply demand balance, lead removal and new development investment). These were developed by analysing industry AMP5 spend on those items Unmodelled Allowance (to cover items which Ofwat found impossible to model effectively in its bottom up model). This was computed by looking at the share of

19 15 totex these items represented across the industry in AMP5 and applying the same percentage to AMP6. ic Totex Models Model C (full COLS) Calculate modelled totex forecasts One modelled totex estimate (top-down) full Model B (refined RE) Calculate modelled totex forecast Triangulate refined models One modelled totex estimate (top-down refined) Model D (refined COLS) Calculate modelled totex forecasts Base model and separate enhancement models Weight three modelled totex estimates Initial average basic cost threshold Calculate basic cost threshold: frontier, UQ Average LQ Calculate unmodelled uplift Calculate enhancement forecasts from unit cost models: 1. Supply demand balance 2. Lead reduction Weight four unit cost approaches Add together estimate of base costs, unit cost and unmodelled costs One modelled totex estimate (bottom-up) 3. New development Calculate base totex forecasts: Model E (base RE) Model F (base COLS) Weight two base model estimates The three totex models (two top-down, one bottom-up) were all accorded the same weights. As a result, the triangulated value for the modelled totex amounts to one third of each of the three models. Implicit in Ofwat s approach is that all relevant costs are captured in each of the three models. More specifically, any cost captured by a unit cost model or the unmodelled allowance in the bottom-up model is also captured in the two top down totex models. Consequently, if one considers the unit cost models and the unmodelled allowance within the bottom-up model, only one third of their value is included in the triangulated totex value. By extension, only a third of any increment to the value of the unit cost models or the unmodelled allowance feeds through to the triangulated value. This is what we refer to as the 'one thirds rule'. We expressed our concern to Ofwat in April that where there are (in terms of the bottom up model) unmodelled items of expenditure in AMP6 that were not materially present in AMP5, no allowance could have been made for such expenditure in the top-down totex models (dependent as they are on historic cost data as inputs). Ofwat wrote to us in early May to confirm that where there are items of expenditure which were not materially present in AMP5, then those amounts will be included at 100% rather than according to the one third rule. (1) In the following sections we set out why we believe the additional allowances that Ofwat has made in the draft determination should be allowed at 100% rather than be subject to the one thirds rule. 1 Quoting from a letter to Peter Simpson from Sonia Brown dated 8 May 2014: Where outcomes are significantly different from business as usual and supported by compelling evidence then companies can make a full claim for the extra costs involved.

20 16 Our Representations Cost threshold adjustments Ecological improvements at abstractions The following table, taken from the August 2013 datashare, shows the extent of ecological improvements at abstractions in AMP5 for all WASCs (W3001 in table W3). WASC expenditure on ecological improvements at abstractions in AMP5 Company m m m m m AMP5 m ANH WSH NES SVT SRN SWT TMS NWT WSX YRK TOTAL 65.5 The following table shows the same figures for AMP6 for all WASCs using data from the individual company feeder models published by Ofwat. Proposed WASC expenditure on ecological improvements at abstractions in AMP6 PR14 Representations on the Draft Determination anglianwater.co.uk Company ANH WSH NES SVT SRN SWT TMS NWT m m m m m AMP6 m

21 17 Company m m m m m AMP6 m WSX YRK TOTAL It can be seen that: i. Apart from Anglian Water, only two other companies, Southern and United Utilities, had more than 10m of capex spend on ecological improvements at abstractions in AMP5. In AMP6, only four companies propose spend below 10m (although, interestingly, they include South West and Northumbrian) ii. The capex spend for all WASCs in AMP 5 was 65.5m. The comparable figure for AMP6 is 279.7m, an uplift of 327%. iii. Our proposed spend on this item in AMP6 is 777% of what we spent in AMP5 iv. Our AMP5 figure represents 8% of total WASC AMP5 capex on this item. For AMP6, our proportion of the total has risen to 15%. These figures show that at an industry level AMP6 expenditure for ecological improvements at abstractions is materially greater than the AMP5 spend. Furthermore, our AMP6 expenditure is materially greater than our AMP5 spend, both on its own and in the context of all WASCs. There are two drivers for our proposed expenditure in AMP6 for ecological improvements at abstractions: the Restoring Sustainable Abstractions (RSA) programme and the Eels Directive. The driver for RSA is the Habitats Directive, which was incorporated into UK law by Statutory Instrument 290 of 2010 and which came into effect in March 2010, that is, right at the start of AMP5. It is notable that of Anglian Water's RSA expenditure in AMP6, 84% relates to one particularly large scheme, that for the Costessey relocation. The remaining elements of the company's RSA expenditure in AMP6 (schemes for Matishall and Hunstanton) amount to 3.6m pre efficiency, a similar level to the AMP5 expenditure. We discussed the Eels Directive in our June BP: "All Anglian Water s surface water intake structures have been assessed as non compliant with the Eel (England and Wales) Regulations (2009) and will require investment to meet the requirements of the legislation. The Regulations seek to reverse the severe decline of eel stocks in the UK and are effective from January 2010 for new intake structures and from 1 January 2015 for existing structures [our emphasis]. Time limited exemptions will be issued by the Environment Agency for non-compliant intakes to allow time for investment to be made." We created no new intake structures during AMP5 and therefore had no applicable investment under this driver. However, companies' existing intakes become subject to the requirements of the Directive from the start of AMP6. The Environment Agency's summary report on

22 18 Our Representations Cost threshold adjustments the companies' statutory requirements and environmental obligations confirm that these are "significant new requirements" requiring investment at 216 sites across England. Given this, the Eels Directive can be clearly seen as a new driver to AMP6. In conclusion, there are strong arguments that the main elements of expenditure driven by RSA and the Eels Directive in AMP6 were not present in AMP5: The total level of spend is significantly higher than in AMP5 For investment on Eels, the requirement to implement the proposed projects is the result of a new regulatory obligation which cannot be modelled by reference to previous behaviour as there was no material expenditure driven by the Eels Directive in AMP5. Consequently, the 14.8m pre-efficiency capex had no parallel in AMP5. Even if there were a history of regulatory interventions to analyse, none of the drivers in the totex models captures the drivers to this expenditure. (A relevant driver for RSA expenditure might, for example, be percentage of abstraction points in areas subject to the Habitats Directive.) Costessey RSA scheme ( 21.7m capex pre-efficiency) is of a different order of magnitude to the other AMP6 schemes and to the AMP5 schemes. Proposed adjustment to costs In Anglian Water's June BP, the post-efficiency capex for these two items under ecological improvements at abstractions was 39.0m. In the RBR in March, Ofwat calculated an Implicit Allowance (IA) of 7.3m for the ecological improvements at abstractions. In the DD, the IA has been restated to be 8.1m. In the DD Ofwat accepted in full the case for these items. Subtracting the 8.1m implicit allowance from the total left a balance of 31.0m, which - after application of the thirds rule - translated into an upwards adjustment of 10.3m. We believe that the full 31.0m should be added to the threshold rather than only one third of the amount. PR14 Representations on the Draft Determination anglianwater.co.uk Resilience Connectivity The following table, taken from the August 2013 datashare, shows the extent of resilience expenditure in AMP5 for all WASCs (W3011 in table W3). WASC expenditure on Resilience in AMP5 Company ANH WSH NES SVT SRN m m m m m AMP5 m

23 19 Company m m m m m AMP5 m SWT TMS NWT WSX YRK TOTAL The following table shows the same figures for AMP6 for all WASCs. Proposed WASC expenditure on Resilience in AMP6 Company m m m m m AMP6 m ANH WSH NES SVT SRN SWT TMS NWT WSX YRK TOTAL Three facts are immediately apparent from the above two tables: 1. In AMP6, Anglian and Severn Trent account for 99% of the proposed resilience expenditure. In AMP5, they accounted for only 61% of the total spent on resilience. 2. In AMP5, five WASCs accounted for more than 5% of total resilience expenditure. In AMP6, only two WASCs had more than 5% of the total. 3. Overall, proposed resilience spend in AMP6 is more than two times higher that in AMP5.

24 20 Our Representations Cost threshold adjustments Both the level and the dispersion of resilience expenditure in AMP6 are clearly very different to that in AMP5. We set out the case for investment in resilience (mains connectivity) in our June BP: Since privatisation, we have invested continuously to increase the resilience of our supply systems. Originally these investments were made to manage the effects of severe drought; since AMP4 we have also invested to mitigate the impact of catastrophic water treatment works failure. Our strategy for non-drought resilience prioritises investment on the basis of the size of the population at risk. For affordability, the programme is then spread over several asset management periods (AMPs). In AMP4 and AMP5, we delivered schemes for vulnerable populations greater than 50,000. In AMP6 we are targeting populations between 30,000 and 50,000. Populations of less than 30,000 will be targeted from AMP7 onwards. The 30,000 to 50,000 vulnerable population thresholds align with guidance from Defra on required SEMD planning (Advice Note 9, (AN/9 Ed. 3) February 2005). The advice note details our 30,000 to 50,000 planning assumptions and shows the long term strategy for resilience planning to provide alternative supplies for 15,000 to 25,000. Included within resilience expenditure in AMP5 are three mains resilience projects: Anglian Water expenditure on mains resilience in AMP5 Project Grafham to Hannington main West Pinchbeck Barrow Total Expenditure 3.7m 9.5m 1.0m 14.1m PR14 Representations on the Draft Determination anglianwater.co.uk In our June BP the pre-efficiency capex for the mains resilience projects which were assessed by Ofwat was 37.2m. Post efficiency this equates to 36.0m. At RBR, this category of capex was rejected due to lack of evidence. At the same time, the Grafham Hannington Main overlap project (worth 22.5m) was accepted (this was not part of the 36.0m). At the DD Ofwat allowed a further 1.4m for this project, reflecting that PR14 s efficiency assumptions should not have been applied to that project given efficiencies had already been applied at PR09. The DD allocated an Implicit Allowance of 20.9m for mains resilience. In the DD Ofwat accepted the additional evidence we provided in June for the remaining 15.1m but then subjected this uplift to the thirds rule, with the effect that only 5.0m was added to our threshold. Mains resilience is a category of expenditure which reaches back across AMP5 and indeed across AMP4. The total value of our AMP6 resilience work, including the Grafham to Hannington main overlap project (Business case ), is ( )m = 61.1m. This is 333% higher than the 14.1m we spent in AMP5.

25 As Anglian Water's AMP6 mains resilience capex is materially larger than the AMP5 figure, we believe that the scale of the increase represents compelling evidence that a step change has occurred in such investment. Furthermore, the expenditure is on a different scale compared to other companies. This investment is not reflected in the two top down totex models, consequently, Ofwat s approach of only adding a third of the additional allowance, i.e. 5.0m, onto the threshold is not a valid application of its stated approach and the full amount of 15.1m should be added to the threshold. 21

26 22 Our Representations Cost threshold adjustments Supply demand balance inputs Ofwat's Draft Determination In its bottom-up assessment of efficient water costs,ofwat calculates an allowance for investment to maintain the supply demand balance (SDB) through a unit cost model. In the April risk-based review (RBR), Ofwat over-wrote Jacobs assumptions regarding our leakage and SDB enhancements with the figures from our Draft Water Resources Management Plan (DWRMP). In our June submission, we argued that Ofwat s model should be updated to reflect the Final Water Resources Management Plan (FWRMP). In the Draft Determination (DD) Ofwat has partially accepted our argument. However, it has not reflected in the thresholds our plans from our FWRMP on enhancement and selective metering and water efficiency 'as no deficit appears to exist in their absence'. Ofwat has also disallowed the reduction of 12 Ml/d from our SELL of 211 Ml/d to 199 Ml/d as we have already achieved leakage levels below 199 Ml/d and 'we therefore do not consider this [achievement of 199 Ml/d] to be an AMP6 enhancement'. As well as reducing the Ml/d enhancements that go into the unit cost model calculation, Ofwat also reduced the number of selective meters that drive Totex model C to ensure consistency with its decision on the enhanced metering programme. These changes contribute to reductions in the overall water cost threshold of 9.9m via the unit cost model and 8.2m via Totex Model C. In our June submission we also argued that the supply demand unit cost model ignored the impact of water stress on companies' SDB costs. Ofwat has not commented on or appeared to consider these arguments. Summary of our Representations PR14 Representations on the Draft Determination anglianwater.co.uk Analysis of our 2015 Water Resources Management Plan shows that without the supply-demand metering and water efficiency programmes we will have a number of deficits in our supply system by These will be in areas vulnerable to growth and climate change. Using 73.78Ml/d instead of 61.25Ml/d as an input to the supply demand unit cost model increases the threshold as computed by Ofwat s model by 9.9m. The following adjustments should be made: Ml/d - enhanced metering Ml/d - selective metering Ml/d - 'Bits and Bobs' Ml/d - 'Drop by 20' Ml/d - Costessey difference Ml/d - Mattishall, omitted in error Ml/d - total

27 23 Reflecting enhanced metering also requires an adjustment to the inputs to totex model C. Doing so increases the threshold as computed by Ofwat s model by a further 8.2m. Thus the overall effect of reflecting the Final WRMP, as approved by the Secretary of State, instead of Ofwat's alternative view is to increase the water threshold by 18.1m. There is good evidence that the supply demand unit cost model should be adjusted to reflect the impact of water stress on companies' SDB costs. We are not repeating these arguments in detail here but we consider they add weight to the need to reflect all aspects of our Final WRMP in the cost threshold. Introduction Reducing the demand for water is a key element of our long-term strategy for managing the security of supply risk from growth and climate change. This approach is consistent with: our approach since privatisation, reinforced in our 2010 Strategic Direction Statement Government policy for the water sector the Guiding Principles of the Water Resource Planning Guideline Defra s Strategic Policy Statement to Ofwat the 2014 Water Act the expectations of our customers and other stakeholders Analysis of our 2015 Water Resources Management Plan shows that without these programmes we will have a number of deficits in our supply system by , and for this reason these programme should be included in the water cost threshold. These will be in areas vulnerable to growth and climate change As well as avoiding these deficits, the schemes: Are cost-beneficial Reduce greenhouse gas emissions, and Reduce customer bills In the long-term, we are exposed to supply-demand risk equivalent to 567 Ml/d, approximately half our current level of distribution input. We aim to manage around 25% of this risk by reducing demand and our AMP6 metering and water efficiency programmes are a crucial, practical and minimum regret element of this strategy Our business plan proposals on supply demand balance enhancement In Table W4 of our December business plan, we set out the following cost drivers for SDB enhancement. None of these changed in our June plan; the numbers are taken from our FWRMP, which we were directed to publish by Defra in August The third line of the table adopts Ofwat's 'rule' for translating figures from W4 into model inputs. When the peak figure is zero in a year, then the combined figure for that year is the annual average figure. When there is a peak figure, then that is the number used for the combined figure.

28 24 Our Representations Cost threshold adjustments Enhancements to the supply demand balance in Ml/d (from AW's business plan) Line description Total W4006 Total enhancements to the SDB (dry year, critical peak conditions) W4007 Total enhancements to the SDB (dry year, annual average conditions) Combined peak & enhancement figure (If critical/peak>0, use critical peak figure; if critical peak =0, use average year figure) In turn, these lines were made up of the following elements: Components of the supply demand enhancements in Ml/d - dry year, critical / peak conditions Line description Total W4006 Costessey (Heigham replacement scheme) PR14 Representations on the Draft Determination anglianwater.co.uk Components of the supply demand enhancements in Ml/d - dry year, average conditions Line description Meter Optants Installations pa (x000) Per HH saving inc CSPL effects (l/d) Aggregate saving inc CSPL effects (Ml/d) Total 3.55

29 25 Line description Total Enhanced Domestic Water metering Installations pa (x000) Switchers pa (x000) Per HH saving inc CSPL effects (l/d) Aggregate saving inc CSPL effects (Ml/d) Selective Water Metering Installations pa (x000) Per HH saving inc CSPL effects (l/d) Aggregate saving inc CSPL effects (Ml/d) Leakage to 199Ml/d Bits & Bobs Audits pa (x000) Per HH saving (l/d) Aggregate saving (Ml/d) Drop by 20 Audits pa (x000) Per HH saving (l/d) Aggregate saving (Ml/d)

30 26 Our Representations Cost threshold adjustments Line description Total Sustainability Reductions Fenland (Hunstanton) Costessey (DYAA) Mattishall W4007 Total enhancements to the SDB (dry year, annual average conditions) There is an important point of terminology to be made here. Anglian Water uses selective meters to refer to those that are fitted compulsorily for customers because they are large water users (e.g. after installing a swimming pool). Enhanced meters are ones which are installed in a blanket meter installation programme. Customers whose properties are fitted with an enhanced meter can continue to pay on an unmeasured basis or choose to switch to measured charges. Ofwat uses the term selective meters to refer to both of these categories. Our approach to enhanced metering is well established in Anglian but unique in the industry. Our use of it since the mid-90s has helped us achieve industry leading levels of meter penetration without compulsorily requiring customers to switch. It is a more cost efficient approach to meter installation than relying on optants, which was recognised in our PR09 final determination. PR14 Representations on the Draft Determination anglianwater.co.uk Selective (compulsory) meters are comparatively few in number and relatively insignificant. The thrust of our arguments in this section relate to our enhanced metering programme which we regard as critical to achieving our strategic objective of moving substantially all customers to measured charging in the medium term. Ofwat's Draft Determination At DD, Ofwat has made the following judgements in determining the input to the supply demand unit cost model: Summary of Ofwat's treatment of our AMP6 supply demand enhancements (Ml/d) Line description Costessey scheme Meter optants Critical / peak AW BP 58.0 DD 57.0 Annual average AW BP 3.55 DD 3.55

31 27 Line description Critical / peak Annual average Enhanced domestic water metering Selective water metering Leakage Bits and Bobs Drop by Sustainability reductions TOTAL In summary, of the 83.51Ml/d of SDB enhancements included in our June Business Plan (BP) Ofwat has used a figure of 61.25Ml/d in the Draft Determination (DD). (These numbers are not immediately apparent from the table above because they depend on the application of the 'rule' on annual average and critical / peak described earlier). Reviewing in detail what Ofwat reflected in the threshold: On Critical / Peak, Ofwat used 57 Ml/d instead of Anglian Water's figure of 58 Ml/d. There is no obvious reason for this. On Annual average: Meter optants were reflected in full. Enhanced domestic water metering and selective water metering were rejected outright. Bearing in mind that there were peak figures in the final two years of the AMP, this removed 1.14 Ml/d (2.02 Ml/d 0.88 Ml/d, being the aggregate volume in and for those two categories of metering) across the AMP from the computation. Ofwat rejected in its entirety our proposed demand reduction schemes within 'Bits & Bobs' and 'Drop by 20', our water efficiency programmes. In terms of the impact on the unit cost computation, the removal of 'Bit & Bobs' lowers the figure by 7.05 Mld for the AMP as a whole (8.46 Ml/d Ml/d from the last two years of the AMP when there was a Peak enhancement). For 'Drop By 20', the volume excluded was 0.86 Ml/d (1.31 Ml/d 0.45 Ml/d). Ofwat reflected the one year of the Fenland scheme where there was no overlap with Peak volumes. This was stated as 0.7 Ml/d. Ofwat appears to have overlooked the Mattishall sustainability reduction total of 1.6 Ml/d. No reason was given for this omission. Meter numbers While the supply demand unit cost model is used in Ofwat's bottom-up approach to cost assessment, in its top-down totex models the cost driver is the number of meters to be fitted. Inputs to both models should be consistent within the determination. Thus, when striking out the Ml/d contributions from enhanced and selective metering for the unit cost model Ofwat also removed these meters from the F inputs with adjustments sheet of the basic cost threshold model, with an impact on the cost allowance calculated by totex model C.

32 28 Our Representations Cost threshold adjustments In Anglian Water's RBR, for meter optants, Ofwat used a total figure for AMP6 of 145,478 and a figure for selective meters of 2,678. These were the figures from our DWRMP. Anglian Water's Final WRMP included details of the enhanced metering programme which was still being consulted on at the time of the draft plan. Anglian Water recognised that the enhanced programme would cover a lot of customers who might have otherwise switched as optants. Consequently, the optant numbers in the final plan were reduced to avoid the double-count and Ofwat selectively used this lower figure of 75,860 in its DD even though it was clear that the lower number only applied in the event of an enhanced metering programme. For selective metering, in the DD Ofwat has set the total across AMP6 to zero. This therefore ignores both the small number of selective (compulsory) meters and the much larger numbers of enhanced meters that were added in finalising the WRMP. Summary of metering numbers DWRMP FWRMP AW BP RBR AW June plan DD Metering programme - Household Selective Meters (including enhanced metering) 2,678 87,120 87,120 2,678 87,120 0 Metering programme - Household Optional Meters 145,478 75,858 75, ,478 75,858 75,860 Our representations on Ofwat's draft determination on supply demand investment PR14 Representations on the Draft Determination anglianwater.co.uk Our main concern with Ofwat's draft determination of our supply demand investment is the exclusion of the majority of our demand side activities, namely our enhanced metering and water efficiency programmes ('Bits and Bobs' and 'Drop by 20'). In this section we reinforce the critical importance of these demand reduction activities and the need for them to be included in our cost thresholds at Final Determination. Our (supply-demand) demand management programme includes: Bits and Bobs water efficiency campaign (180,000 audits with water savings of 8.5 Ml/d) Drop by 20 water efficiency campaign (65,440 audits with water savings of 1.3 Ml/d) Enhanced metering (85,595 meter installations with water savings in AMP6 of 1.9 Ml/d and total savings in of 4.3 Ml/d), and Selective metering (1,500 meter installations with AMP6 water savings of 0.1 Ml/d). Over the period to , these schemes will deliver total water savings of 14.2 Ml/d. From the figure below, it can be seen that we put the same amount of water into supply now as we did in 1999, even though the population in our region has increased by 20%. High levels of metering and water efficiency activity are partly responsible for this and remain key elements in our strategy for managing demand.

33 29 Strategic Direction Statement Our Strategic Direction Statement (SDS) was published in In this, we described the long-term strategy for our business and customers. The SDS showed that climate change and growth are among the main challenges we face. In surveys completed for the SDS, stakeholders told us that dealing with these were their two highest priorities. In response, we said that we plan to: Increase the resilience of our supply systems, and Secure and conserve water resources. Reducing the demand for water is a key element of our strategy for meeting these objectives. In the SDS, we stated that we will do this through increased levels of household metering and water efficiency activity. The next two figures show the distribution of our climate change and growth vulnerabilities. These are the areas that will be targeted for demand management:

34 30 Our Representations Cost threshold adjustments Figure 2 - Climate change vulnerabilities PR14 Representations on the Draft Determination anglianwater.co.uk

35 31 Figure 3 - Planning enquiries and development areas Policy Framework Government policy objectives for the water sector are given in the water white paper Water for Life (December, 2011). They include promoting growth and protecting the environment.

36 32 Our Representations Cost threshold adjustments Government strategy for achieving these goals is based on managing the risk which is posed by climate change and population growth. Key elements include: Preparing for change Increasing sustainability, and Maintaining affordability. "Water for Life" makes clear that "water companies have a critical role in planning for the future they must help drive change in the way we use water through the advice and support they offer to customers." In respect of maintaining the security of supply, specific guidance on water resource planning and investment is given in the Water Resource Planning (WRP) Guideline Guiding Principles (Environment Agency, Ofwat, Defra and the Welsh Government, 2012). In this, water companies are directed to: Take a long-term perspective Take better account of the value of water Consider all options for maintaining the supply-demand balance Reduce the demand for water, and Ensure that customer views are taken into account. These requirements are also evident in the Strategic Policy Statement to Ofwat (Defra, 2013). In this Ofwat is directed to recognise the need for companies to take a long term approach to delivering changes in customer attitudes and behaviour on water conservation: Ofwat should encourage companies to demonstrate how they will reduce water use per person, and recognise in its decisions the need for companies to take a long term approach to delivering changes in customer attitudes and behaviour on water conservation. The 2014 Water Act gives further direction in respect of securing the resilience of supply systems as regards environmental pressure, population growth and consumer behaviour. In the Act, increasing efficiency and reducing demand are specified as key measures for reducing pressure on water resources. Under Section 22 of the Act, Ofwat has a new primary duty to secure the resilience of water supply systems. PR14 Representations on the Draft Determination anglianwater.co.uk In conclusion, in respect of this well-established and robust policy framework: Our SDS is consistent with the priorities of Government as expressed in Water for Life and the Strategic Policy Statement to Ofwat, and Our 2015 Water Resources Management Plan (WRMP) has been prepared in accordance with the requirements of the WRP Guiding Principles. On this basis, it includes cost-beneficial demand management schemes that are necessary to mitigate risk from growth and climate change. We are under obligations to put in place demand management programmes. Security of Supply Assessment To determine the value of the AMP6 demand management programme, we have assessed its effect on our Security of Supply Index (SoSI) in The assessment is based on the final planning solution (FPS) tables in our 2015 WRMP. We were directed to publish this by the Secretary of State in August 2014.

37 33 SoSI is a function of the supply-demand balance at resource zone level; the size of any deficits and the size of the population which is affected by those deficits. There are four SoSI categories: Table 1 - SOSI categories Band A B C D Description No deficit against target headroom in any resource zone Marginal deficit against target headroom Significant deficit against target headroom Large deficit against target headroom Index Score Below 50 The results of the assessment are given below: Table 2 - SOSI assessment 2015 WRMP final planning solution Distribution Input (Ml/d) Water Available for Use (Ml/d) SoSI SoSI band Comment Without Demand management programme 1,122 1, B Marginal deficit against target headroom With Demand management programme 1,108 1, A No deficit against target headroom Without the demand management programme a number of resource zones (RZs) go into deficit in From Figure 4: The RZs benefitting from the programme are concentrated in the south and east of our system From Figure 2, these areas are vulnerable to the effects of climate change, and From Figure 3, these areas contain a large number of our growth hotspots including: Milton Keynes and Bedford Northampton Peterborough Ipswich Colchester

38 34 Our Representations Cost threshold adjustments Norwich King s Lynn and Wisbech. Table 2 and Figure 4 show how the excluded programme contributes directly to our strategy for mitigating risk from climate change and growth. In AMP6, this involves driving efforts to achieve a permanent change in the attitudes and behaviours of customers about the use of water. This is consistent with our SDS and our Love Every Drop campaign and supports the policy objectives of Government. Figure 4 - RZs with deficits if the AMP6 demand management programme is not delivered PR14 Representations on the Draft Determination anglianwater.co.uk

39 35 It is critically important to view the SDB programme in general, and the demand management programmes in particular, in the context of the 25 year view to , set out in the WRMP. Cost Benefit Assessment In common with the rest of our PR14 BP submission, the costs and benefits of the excluded programme have been assessed using Asset+. In this, the benefits are a function of customer willingness to pay. This has been determined from surveys completed during the development of the PR14 Business Plan. The results are given below and show that each scheme is cost-beneficial: Table 3 - cost benefit analysis Programme Equivalent annualised benefits ( k) Equivalent annualised costs ( k) Equivalent annualised benefit ( k) 'Bits and Bobs' 51, ,417 'Drop by 20' 7, ,608 Enhanced metering 18,115 1,722 16,393 *Note: no cost benefit assessment has been completed for selective (compulsory) metering as meters installed under this programme are installed compulsorily. In addition: Based on water savings of 14.2 Ml/d and 446 tco 2 e per Ml/d of water into supply, the programme will reduce greenhouse gas emissions by approximately 6,000 tco 2 e per year, and We estimate that customers switching from unmeasured to measured supplies typically save around 100 per year on their water bill. Given 85,595 installations, our enhanced metering programme has the potential to save 8.6m on customer bills. For measured customers, savings will also result from our water efficiency programme. Given 245,440 audits, these will be comparable in value to the savings from enhanced metering. The proposed metering and water efficiency programme target areas where, in AMP6, we are planning river restoration works or investigations linked to the National Environment Programme (NEP). In these areas, the water savings from our combined demand management activities are sufficient to offset the AMP6 effects of growth. This means that during the period of the investigations and the implementation of the restoration works, we can mitigate related environmental risk.+ The three programmes in question are strongly cost beneficial. Aggregate net benefits are more than 36 times the aggregate annual costs. Moreover, the programmes contribute significantly to increasing customer affordability and will have further ancillary benefits.

40 36 Our Representations Cost threshold adjustments Customer Support and Outcomes Customer views about our PR14 Business Plan are summarised in our Outcomes in Detail document. This also shows how customer views have been central to the development of the plan and the outcomes we are committed to delivering. Outcomes that are linked to the excluded programmes include Supply Meets Demand, Fair Charges and Caring for Communities. Our customers recognise that there is increasing pressure on water supply systems in the UK and that there are particular vulnerabilities in our region which are linked to growth and changing weather patterns. They are clear that severe water restrictions are unwanted and they expect us to be taking action now to avoid problems in the future. In respect of metering and water efficiency: Most customers regard meters as the fairest way to pay for water, and There is strong support for campaigns to help customers save water including more advice from us; low cost or free water saving devices and greater evidence of pay-back on bills. In response, we have committed to increasing levels of domestic meter penetration in AMP6 to 95% and to reducing levels of per consumption. The excluded AMP6 metering and water efficiency activities are critical to achieving these outcomes. Since customers reduce their water and energy bills when they switch to measured supplies or install water saving devices, these programmes are also a key element of our strategy for maintaining affordability. The programmes not included in Ofwat's cost threshold are strongly supported by our customers. Having sought their views and crafted plans to meet their legitimate requirements, it is disappointing that Ofwat has not reflected these commitments in assessing cost thresholds. Water Resources East Anglia PR14 Representations on the Draft Determination anglianwater.co.uk We estimate that the total supply-demand risk that we will be exposed to in is 567 Ml/d, approximately half of our current level of distribution input. This assumes worst case impacts from growth, climate change and the reductions in deployable output that are needed to restore abstraction to sustainable levels. Our strategy for managing this risk combines demand management with a potentially large increase in supply. This could include building new reservoirs; desalination plants; water re-use plants and strategic transfers. The strategy is summarised in Figure 5 and shows that the contribution from demand management is significant (around 25%). Included in this are reductions in leakage and household consumption, with most (around 18%) from the effects of household metering and water efficiency activity. In respect of increasing supply, detailed planning is needed to make sure that the right assets are delivered at the right time. Without this, there are unacceptable risks to affordability (building too much, or too soon) and levels of services (building too little, or too late). This planning work will be completed in AMP6 as part of our Water Resources East Anglia (WREA) project.

41 37 Since demand management will always be needed as part of our overall strategy for maintaining the security of supply, there are no affordability or level of service risks for our AMP6 metering and water efficiency programmes. Instead, delivering these is a 'minimum regret' strategy for: Increasing the resilience of our supply systems Mitigating our medium to long-term climate change and growth related risks, and Increasing the affordability of our long-term supply-demand programme. Figure 5 - The scale of our long-term supply-demand challenge and the role of demand management in mitigating the related risk In addition to our case for the demand management programme we also make the following representations on supply demand: We can see no reason why the peak figure should be 57 Ml/d instead of 58 Ml/d for the Costessey scheme. There is no obvious reason apart from oversight for ignoring Mattishall. It cannot be on the grounds of materiality as the Fenland scheme was included despite being less than half the size of the Mattishall scheme. In our June submission concerning the quality of modelling, we recalculated the inputs for the SDB unit cost model for water stressed and non water stressed companies. This showed that the uplift to the water cost threshold should be 57.5m when the Final WRMP inputs and the revised model drivers taking into account water stress are both

42 38 Our Representations Cost threshold adjustments used. This figure is 20.2m higher than that accepted by Ofwat in the absence of taking water stress into account. This supports the case for adding 18.1m to reflect the Final WRMP. Conclusion We consider that the correct combined peak / average enhancement total across the AMP should be as follows: Ml/d Starting point (Ofwat in DD) Add back in enhanced metering Add back in selective metering Add back in 'Bits & Bobs' Add back in 'Drop by 20' Add back in Costessey difference Add back in Mattishall Revised total Using Ml/d instead of Ml/d as an input to the supply demand unit cost model increases the threshold as computed by Ofwat s model by 9.9m. PR14 Representations on the Draft Determination anglianwater.co.uk Accepting our case for enhanced metering also requires an adjustment to the inputs to totex model C. If one includes the 87,120 enhanced and selective meters in 'F_inputs with adjustments', item BN1711_W003, then this increases the threshold as computed by Ofwat s model by a further 8.2m. Thus the overall effect of reflecting the Secretary of State-approved Final WRMP for Anglian Water is to increase the water threshold by 18.1m.

43 Third party costs Ofwat's Draft Determination Ofwat included an allowance of 9.9m for third party costs as part of the policy additions to our water cost threshold. "We excluded third party operating costs from our totex modelling because they are unlikely to be properly captured by our totex models. We therefore need to separately calculate a forecast of third party opex. Our approach is to calculate for water and wastewater separately the percentage of third party revenue recovered relative to third party costs over the period to (the most recent 5 years we have matching costs and income). We have then applied this percentage to your forecast wholesale third party income reported in business plans for each of water and wastewater, and used the resulting estimates as estimates of third party costs for each of water and wastewater." Summary of our Representations Ofwat's calculation of historical cost recovery has been performed using figures for total third party revenue, which include revenue from non-potable supplies. Our forecasts for total third party income in to excluded non-potable revenue as we assumed it will be part of the wholesale revenue and controlled by default tariffs. For the purposes of making an accurate estimate of our AMP6 third party costs, we have added the revenue from non-potable supplies to our AMP6 third party revenue forecasts. Applying the historical cost recovery percentage to this consistent forecast increases the amount included in the cost threshold for third party costs from 9.9m to 25.2m, an increase of 15.3m. In the table below we have replicated the calculation which Ofwat performed to obtain an estimate of the percentage of third party revenue recovered relative to third party costs over the period to In this calculation Ofwat used figures for total third party income from the regulatory accounts. In accordance with regulatory accounting guidelines of the time, this sum comprised three components bulk supplies, rechargeable works and sales of non-potable water. The table shows that over this period, third party costs comprised 29% of total third party revenue. Third party costs as a percentage of third party income to Water Total Third party income (incl non-potable income)( m) Third party costs ( m)

44 40 Our Representations Cost threshold adjustments Water Total Costs as a percentage of income (%) 27% 34% 20% 24% 38% 29% The table below replicates our forecasts for third party income in to from Table W9 line 9 of our business plan. These forecasts included only bulk supplies and rechargeable works. Non-potable revenue was excluded from this line but included in Line 1 as we assumed it will be part of the wholesale revenue and controlled by default tariffs. Copy of Table W9 Line 9 from AW's June submission Total Third party revenue (excl non-potable income)( m) The amount included in our draft determination cost threshold of 9.9m was calculated as 29% of this forecast third party income. It is clear that there is a mismatch between the historical figures which have been used to calculate the third party costs percentage and the future third party revenues to which the percentage has been applied. In the table below we show a calculation to remedy this misunderstanding by adding forecast revenues from non-potable supplies to the other third party revenue figures we provided in Table W9 Line 9. Total third party revenue for , including from non-potable supplies Total Third party revenue (excl non-potable income)( m) W9 L PR14 Representations on the Draft Determination anglianwater.co.uk Non-potable income, incl special agreements ( m) Total third party revenue (incl non-potable income)( m) The table shows that our total forecast third party revenue for the period , including from non-potable supplies, is 88.2m which, as can be seen from the tables above, is consistent with similar revenues of 85m in the period 2006 to Using the historical recovery rate calculated from previous years (29%) gives an allowance for third party costs in AMP6 of 25.3m. This sum should be added to our water cost threshold in place of 9.9m as part of the policy additions, an increase of 15.3m. We have not resubmitted Table W9 as we assume that the way we completed it in our June submission is consistent with Ofwat's intention for the treatment of non-potable revenue as part of the wholesale revenue control.

45 Wastewater cost threshold Ofwat's modelling approach Ofwat's Draft Determination In publishing its Draft Determinations (DDs), Ofwat made a change to its methodology, capping at 5% the extent by which a company's baseline will exceed its planned totex where planned totex undershoots Ofwat's forecast threshold Summary of our Representations We present reasons why our wastewater threshold should not be reduced by the unmodelled adjustment of 18.6m. Taking the lower of the company's forecast and Ofwat's allowance for unmodelled items is inconsistent with introducing a 5% cap at the total threshold level. Policy implication of introducing the 5% cap on company undershoot of Ofwat totex threshold In announcing the Draft Determinations, Ofwat made a policy shift affecting companies whose business plan forecasts of totex are significantly below initial draft determination thresholds (section A3.3.2 of Draft price control determination notice: technical appendix A3 wholesale water and wastewater). The policy shift is to cap at 5% the extent to which Ofwat s menu baseline exceeds the company s proposed totex. Six reasons were given by Ofwat for this policy change: 1. The impact of the approaches adopted for the two enhanced companies and the two early draft determinations and viewing these as effectively boundaries to what could be allowed (an upper bound of 8% and a lower bound of 4%). 2. The need for a consistent approach across companies (excluding the enhanced ones, where it is important to retain a higher cap to reflect the benefit of having achieved enhanced status). 3. Providing a strong protection for customers from excessive costs and possible shortcomings in Ofwat's modelling approach. 4. The difference between the upper quartile and average efficiency levels. 5. Any specific known modelling impacts, such as scale/density, which, on analysis, could have influenced the modelling results for one or more of the companies. 6. The importance of maintaining incentives for companies to submit lean plans and make real efforts to secure efficiencies over and beyond those expected by Ofwat.

46 42 Our Representations Cost threshold adjustments This new policy is designed to ensure that companies which submit lean business plans will not be penalised and thus given a perverse incentive not to submit lean proposals in future, and we support these aims. However, it renders the current approach to unmodelled costs inconsistent. Currently, irrespective of a company's totex total vis-à-vis the Ofwat threshold, if the aggregate value of a company's plans for unmodelled items is below Ofwat's unmodelled allowance, then the value of the undershoot is deducted from the Ofwat threshold in its entirety. In the light of Ofwat's new approach, the deduction should only be applied if the company's totex is more than 5% below the Ofwat threshold. The impact of the unmodelled cost allowance limit for Anglian Water is that: Ofwat calculated Anglian s pre-efficiency wastewater unmodelled allowance as 89.0m ( 79.8m post-efficiency) Anglian s wastewater unmodelled costs are 68.3m pre-efficiency ( 61.2m. post efficiency) Consequently, the post efficiency amount removed from the threshold was 18.6m ( 79.8m- 61.2m) With a wastewater totex proposal from Anglian of 2,517.6m, our threshold would be 2.1% below the Ofwat threshold if the 18.6m adjustment were not made. As the extent of the undershoot is less than 5%, following the methodological approach now in place, the 18.6m adjustment should be reinstated. PR14 Representations on the Draft Determination anglianwater.co.uk

47 Wastewater Growth Ofwat's Draft Determination In our wastewater cost threshold Ofwat reflected a sum for growth at sewage treatment works (STWs) using its unit cost model. The volume driver in this unit cost model was change in population from to ; this period was used rather than the full AMP5 period used for other unit cost models to avoid any step change in data as a result of the impact of the 2011 census. At the Risk-Based Review (RBR) Ofwat used Jacobs' estimate of our AMP6 population growth, which was 179,539. For the draft determination (DD) Ofwat has replaced this number with our own estimate of 297,500. This adjustment added another 23m to our threshold but still leaves a gap of 39m to the investment we proposed in our business plan. Ofwat said 'ANH has not attempted to explain why the cost it proposes is significantly greater than the unit cost per increase in the industry'. Summary of our Representations Ofwat's unit cost model is driven by population increase expected in AMP6 but companies' investment is driven by the need to increase STW capacity. An increase in population will not require investment in STWs if either there is existing headroom from investment in a previous period or headroom is created in the period by declines in non-domestic load. Conversely, investment may be required where forecast population increases are comparatively modest. We propose a benchmark unit cost model based on population equivalent treatment capacity provided rather than population increase. On this basis, Anglian's costs in AMP5 were slightly more efficient than the industry average and a forecast for AMP6 based gives a benchmark that is higher than the 106m included in Anglian's business plan. Our proposed expenditure for STW growth should therefore be reflected in full and 39.8m added to the wastewater cost threshold. The unit cost model used by Ofwat to estimate companies' AMP6 investment needs for STW growth uses industry expenditure in to and growth in population over the same period. On the face of it, growth in population should be a good driver of the need to invest at STWs but in practice there are a number of reasons why this is not the case: Firstly, an increase in connected population will not necessarily give rise to a need to invest if there is sufficient existing capacity, or headroom, from earlier investment. Conversely, investment may be required in response to only a small population increase if there is a deficit in capacity because of load increases that have been absorbed in previous periods. The ability to absorb increases in load will also be affected by the sensitivity of receiving water courses and the effect of the Water Framework Directive No Deterioration requirements. While the load increase over time may be relatively smooth, changes in capacity tend to be

48 44 Our Representations Cost threshold adjustments stepped because growth schemes at any individual site are infrequent and treatment units are large. The alternative - to make frequent returns to sites with small additions to capacity in response to incremental increases in load - would be inefficient. Secondly, domestic population is only one component of load, the balance being load from industrial and commercial customers. Should this load decline it will free up STW capacity, potentially enabling a company to absorb additional population without having to make any investment at all. All increases in load are therefore net of any load reductions. The industry developed the 'population equivalent' (p.e.) concept to enable the two types of load to be considered together. Population equivalent acts as a common currency of the demand on a STW and is a better driver of STW growth unit costs than increases in population. Thirdly, population estimates can be significantly affected by census results. Uncertainty about the application of 2011 census data during AMP5 forced Ofwat to use the four year period to in deriving its sewage treatment unit cost growth model. The p.e. capacity of works is much more stable than population numbers that may be affected by census data. What drives investment at a STW, therefore, is not the short-term increase in the population it serves, but its capacity to treat its receiving load. We argue that Ofwat's unit cost model should be benchmarked against the additional STW capacity that is being provided. We recognise that an assumption underlying this approach is that companies will not undertake unnecessary investment to increase capacity. In our June submission we sought to address this concern by clearly setting out the increased capacity required on a works by works basis. The design horizon chosen for assessing additional capacity at each works was chosen to be least cost given the future growth plans in the relevant catchment. PR14 Representations on the Draft Determination anglianwater.co.uk Using numbers from the 2013 August Submission we have calculated the industry weighted and unweighted sewage treatment growth unit costs on the basis of population enhancement treatment capacity provided by the water and sewerage companies in AMP5. (We have replicated Ofwat's approach for this purpose, applying BCIS regional wage factors and excluding Southern as an extreme outlier). We find that Anglian's unit cost, at 422 per unit of population equivalent treatment capacity provided, is lower than the industry level value of 448. This is in contrast to Ofwat's finding that our unit costs, calculated on a population basis, are relatively inefficient. We consider that the population equivalent approach is more valid for the reasons set out above. It is also more consistent with the findings of Franklin and Andrews, whose independent benchmarks (described on page 248 of our June submission) were close to our business plan proposals. We set out our calculations below: Calculation of STW growth unit cost Company Anglian Welsh Northumbrian Population equivalent treatment capacity enhancement ('000) Expenditure on growth at STWs incl. BCIS ( m) Unit cost ( /p.e ,290

49 45 Company Population equivalent treatment capacity enhancement ('000) Expenditure on growth at STWs incl. BCIS ( m) Unit cost ( /p.e. Severn Trent Southern South West Thames ,105 United Utilities Wessex Yorkshire Weighted average (excl Southern) 403 Unweighted average (excl Southern) 494 Triangulated average (excl Southern) 448 Furthermore we can use this industry unit cost to forecast our expenditure on STW growth in AMP6. In our June submission (pages 263 onwards) we set out the capacity increase we proposed to provide at the 24 STWs in our plan. Capacity increase at STWs in AMP6 to provide for growth STW Brackley Raunds Barton on Humber Broadholme Buckingham Cambridge Chelmsford Colchester Dereham AMP6 capacity increase (p.e.) 5,630 7,140 2,670 16,970 4,700 19,000 11,940 16,900 4,770

50 46 Our Representations Cost threshold adjustments STW Dunstable Gt Dunmow Leighton Linslade Long Buckby Marston North Hykeham Poringland Ramsey Sawtry Silverstone Sleaford Stowmarket Whitlingham Wickford Wymondham Total AMP6 capacity increase (p.e.) 25,860 5,400 16,150 2,680 37,480 8,590 5,020 6, ,740 6,850 5,910 61,230 9,190 6, ,780 PR14 Representations on the Draft Determination anglianwater.co.uk On the basis that we plan to provide additional treatment capacity at STWs of 294,780 p.e. and using the AMP5 unit cost of 448 per p.e. The "unit cost" benchmark should be 132m. As this alternative approach results in a benchmark that is 25% higher than the 106m included in our business plan we do not consider that there is valid justification for scaling back the costs included in our plan. Our proposed expenditure for STW growth should therefore be reflected in full and 39.8m added to the wastewater cost threshold.

51 47 3 Customer bill profile and financeability Ofwat's Draft Determination Revenues in the Draft Determination are 82m lower than in our June Business Plan submission as a result of Ofwat's approach to the CIS Midnight Adjustment. This has resulted in a different profile of bills compared to our June Submission, with a sharp decrease in and then further falls. A rise in prices in would ensue. Based on the DD we would not be able to provide assurance that our Business Plan is financeable without further adjustment. Ofwat accepted our proposed PAYG and Depreciation parameters but asked us to engage further with customers on our proposals. Summary of our Representations Results of a new survey of customers and engagement with our Customer Engagement Forum clearly support our proposed approach to profiling of bills and revenues over the period to % of customers surveyed were in favour of our proposals compared to a profile similar to the Draft Determination. The Draft Determination is not financeable without further adjustment. Part of the reason for this is due to the tightness of the FFO/Debt ratio. We consider it unlikely that we will be able to convince Standard and Poor's to de-emphasise this ratio for water companies. We therefore consider we are at significant risk of ratings down grade should no adjustment be made to the plan. We propose an amendment to RCV Run Off rates to bring foward 82m of revenues to AMP6 and changes to PAYG rates to deliver a smooth profile of bills from for the remainder of AMP6 and through into AMP7. We propose to use Run Off Rates rather than making a further adjustment to PAYG rates because it seems more appropriate in principle. The resulting parameters are also more consistent with figures used in the industry generally. Customer bills profile and financeability When we submitted our Business Plan in December 2013 we sought to balance affordability for customers and financeability. We were able to deliver some of the biggest bill reductions in the industry and give customers the smoothly declining bill profile they tell us they want. In our extensive customer research it was extremely clear that affordability for customers was a key priority but also there was clear customer support, based on specific research, for the profile of bills through the AMP period to be smooth.

52 48 Our Representations Customer bill profile and financeability We continued to discuss options for profiling of bills with our Customer Engagement Forum throughout the period up to our June Submission. These discussions reinforced these priorities and so we maintained our approach to the profiling of revenues and bills and to maintaining financeability. In order to achieve this we needed to deal with a number of specific issues, namely: Lower revenues as a consequence of our accepting Ofwat's Risk Reward Guidance The impact of AMP5 capital outperformance Tight actual FFO/Debt ratios Ofwat's Risk Reward Guidance We accepted Ofwat's Risk Reward Guidance and as a consequence revenues in AMP6 were substantially lower than we had assumed in December. We looked at various options for ensuring that our plan was financeable on both an actual and notional basis. Our June submission sought to resolve this issue by bringing forward revenues to AMP6 from future years by means of a higher PAYG ratio. We characterised this as an uplift of c5 percentage points to the "natural" PAYG rate for both the water and sewerage services. In our December Plan we set our PAYG ratio at its "natural" rate. We considered that the natural rate should be defined to be the proportion of forecast "operating costs" as a percentage of totex over the 5 years of our plan. However, it is worth noting that result of this calculation can vary depending on how one defines "operating costs": Natural PAYG rates implied by our June Business Plan submission How is "Opex" Defined UKGAAP (all IRE treated as capex) IFRS (IRE split between opex and capex according to IFRS accounting standards) Water PAYG % Sewerage PAYG % PR14 Representations on the Draft Determination anglianwater.co.uk UKGAAP Opex + all IRE (all IRE treated as "opex") June Submission and Draft Determination Our June Submission was based on the second of these three definitions. There is a reasonable argument, however, that the third approach would be an appropriate measure since IRE expenditure has Pay as You Go characteristics and was effectively treated as an operating cost by the rating agencies in previous reviews. Our Business Plan proposals fell between these two alternative measures of the natural rate which we continue to believe was a reasonable position, given affordability concerns for customers

53 49 The impact of AMP5 capital outperformance We delivered substantial outperformance on capital expenditure for customers in AMP5 which needs to be reflected as reductions to our RCV in AMP6 under the Capital Incentive Scheme mechanism. Because the outperformance is material, at around 334m, reflecting our successful delivery of capital efficiencies, it potentially has a large impact on both revenues and RCV. We therefore considered two options for dealing with this adjustment: a single midnight adjustment on 1st April 2015, reducing RCV by c 334m and hence lowering both depreciation and returns over the five years five equal adjustments of c 75m made at the beginning of each year of AMP6 (total 372m over the AMP), with the amount calculated to ensure that the present value of these adjustments equalled a single adjustment of 334m at the beginning of AMP6. Compared to the previous period this results in higher revenues during AMP6 of around 82m but RCV at the end of the AMP lower by 98m. We decided that the second of these approaches was aligned with the views of our customers because it spread the adjustment more evenly over AMP6. Moreover it had a beneficial impact on our overall financeability, avoiding the need for a higher PAYG ratio. FFO/Debt Ratio To be assured that our Plan was financeable we needed to ensure that we met threshold levels on key financial ratios. Of these the FFO/Debt ratio was particularly challenging both on a notional and actual basis. This ratio is a "core" ratio for Standard and Poor's. Throughout the period covering our December and June submissions we have engaged with S&P to discuss their approach to this ratio in the case of water companies and the consequences for their assessment of our plans both on an actual financing basis and on a notional financing basis. In considering our actual credit ratings, Standard & Poor's applies a threshold level of 6% for AW, given its risk characteristics, taking into account fundamentals such as capital structure and the protections provided by our securitisation arrangements. As a result of our engagement with them, we were able to gain sufficient assurance from Standard & Poor's about its view of our proposals to be confident that the June Plan was financeable. Standard and Poor's, in common with other rating agencies are understandably less concerned with the notional capital structure used by Ofwat to determine prices. However, as a result of our discussions with S&P, we concluded that on a notional basis, with a lower level of gearing but with a conventional financing structure, the relevant threshold level would have been higher at 9%. The ratios implied by our plan were lower than this, at around 7-8% but in discussions with S&P the benefit of our actual capital structure with its securitisation covenants were recognised and we were therefore of the view that the Plan was financeable on a notional basis. However, in both cases the ratios were seen as very tight. We concluded that any further reduction in planned revenues would substantially increase the chances of an actual credit rating downgrade either as a consequence of the Draft Determination itself or as a consequence of further downside risks crystallising under a more challenging determination.

54 50 Our Representations Customer bill profile and financeability The Draft Determination In its Draft Determination Ofwat made a number of adjustments to our June Plan that reduced revenues, including an adjustment that reversed our proposal for dealing with the CIS adjustment for capex outperformance so that it is made as a single midnight adjustment. This single change has reduced revenues by 82m compared to the plan, while at the same time increasing the level of the RCV by the end of the period by around 98m. The changes have meant that without further adjustment we would be unable to provide assurance that the Plan is financeable on an actual basis. The reduction in revenues has meant in particular that the key FFO/Debt ratio is even more challenging. Whilst the publication of the Draft Determination has not resulted in any ratings action in our case, it has meant that there is substantially less flexibility to deal with any further downside shocks. We have looked at a number of downside scenarios and have concluded that we would find it difficult to maintain an acceptable FFO/Debt ratio under a number of scenarios. Further, recent engagement with Standard & Poor's suggest to us that the prospects of persuading S&P to change its methodology to take account of the particular circumstances of water companies have receded. The fact that it has recently placed Thames Water on negative watch appears to support this. We therefore no longer think that there is a realistic prospect and therefore face the real potential for a ratings downgrade unless there is some adjustment to the DD. On a notional basis, although Ofwat's calculations show FFO/Debt ratios which are similar to those we proposed in the June Plan, they remain below the appropriate threshold level (9%) on average and, as a consequence of the profile of revenue and bills in the Draft Determination, they are somewhat low in the second half of the AMP period. Ofwat did not intervene to adjust our PAYG ratios in the DD on the basis of financeability but expects us to engage further with customers on alternative scenarios in relation to revenue recovery at AMP6. PR14 Representations on the Draft Determination anglianwater.co.uk Ofwat has highlighted its calculation that Anglian Water has passed through c30% of the reduction in WACC between our December and June Business Plans. The Draft Determination does not state how this figure is calculated but it draws comparison with others who have passed through the reduction to different extents. We believe that this figure should be seen in the context of our approach to affordability and financeability in our plan as whole. We set out to respond to customers views on the level and profile of bills over the period. This resulted in one of the largest decreases in customer bills and in consequence limited headroom on some ratios. Other companies may have been in very different situations in respect of financeability in their December plans and we are sceptical about what conclusions should be drawn from this comparison. The views of our customers We have engaged with our customers in three ways since the Draft Determinations were published: Initial discussions with key members of our Customer Engagement Forum: to clarify the Draft Determination and our response to it and to plan how we would engage with the wider group of stakeholders. A day-long workshop attended by our Customer Engagement Forum and members of our four Independent Advisory Panels at which six key issues raised in the Draft

55 51 Determination were discussed in depth. Bill profile and financeability was one of these issues. A survey of customers seeking views on alternative profiles of bills and hence revenue recovery over AMP6 and beyond. The design of the survey was informed by and agreed with members of the Customer Engagement Forum. Customer Engagement Forum Discussions The members of the Customer Engagement Forum (CEF) and all four Independent Advisory Panels were invited to a workshop on 17 September which discussed the implications of the Ofwat Draft Determination for the Anglian Water plan for 2015 to 2020 to help inform the company and CEF representations to Ofwat. The majority of CEF and panel members, 23 in total, were able to attend on the day representing a good cross section of the panels and different stakeholder interests. Bringing everyone together was also valuable in exposing any different perspectives that might exist between members. We circulated a detailed draft briefing paper which explained the bill profiling options and the issues associated with ensuring financeability for regulated water companies, Anglian Water's approach and Ofwat's views. A copy of this briefing note is available. CEF members commented favourably on the clarity of this note in exposing the issues concerned. It asked the CEF to consider the following key questions: 1. Do you understand the approach we adopted to ensuring our Plan is financeable both for the December Plan and in June? 2. What might we consider in balancing recovery of costs between current and future customers? 3. What is your view about how bills reductions should be profiled across 2015 to 2020? 4. Do you have any views on how we should approach the issue of financeability in canvassing the views of customers? The key outcomes of the day are summarised below. These findings have been confirmed as an accurate record with workshop attendees. Outcomes of Customer Engagement Workshop Overall, the CEF felt that more stable bills were preferable for customers and that they would prefer to avoid price hikes even if this meant forgoing a short term reduction. It was particularly concerned about the uplift in bills after 2020 and prefers to see a long term framework for investment and planning. The CEF recommend that Anglian Water includes bill figures in its survey when consulting customers and illustrates the impact post 2020.

56 52 Our Representations Customer bill profile and financeability The Customer and Communities panel felt Ofwat had probably made up its mind on this and the issue would be non-negotiable. It felt that customers would not notice lower bills too much now but when they increase at the end of the period this would be more noticeable. It believes an increase in bills later on will make things more difficult and customers want a smoother flight line. A further conversation is needed with customers about changes in bills. It recommends that when the issue is discussed with customers in the survey, indicative bill impacts are given in figures and the impact of future bills is included. It also noted that media interest and customer perception etc. of increase in later years was likely. The Economic panel s main concern when discussing financeability was how the possible uplift in bills in 2021 would be communicated to customers. There was strong feeling among attendees that this bill impact would need to be made clear in advance in order for customers to make an informed decision. Another element discussed was what possible impact changing the financial mechanisms in the plan may have. There was concern about the possible negative impact on the company s credit rating with agencies already putting Thames Water on a downgrade warning. The panel would support the company s suggested approach to making a representation on the financeability element of the plan, but would like to see more information on possible AMP7 bill impacts. A further point was made that the company's proposals in relation to leakage could also lead to bill variability within the period, although it was accepted that the proposal seemed a pragmatic approach. The Environment panel commented that there must be a long term (multi decade) framework for investment and planning to achieve water quality and habitat objectives. The AMP process should have limited influence on this. PR14 Representations on the Draft Determination anglianwater.co.uk The Hartlepool panel felt that this issue had less of an impact from its perspective. It was against burdening future generations with investment that should take place now rather than later. It felt that customers want value for money now with stable and predictable future bills to enable budgeting, prevent shocks and protect against future increases. Customers would forego a dip this year if it means no future price hikes. The panel felt that 5 years is long term for many domestic customers and many small commercial companies may not plan for longer than 3 years which makes it difficult to gauge a view on altruism for the future and financing. So, it believes Anglian Water needs to ask customers what they want and the panel recommend that when Anglian Water consults customers it shows figures rather than graphs. The overall messages from the meetings and the workshops with CEF and Independent Panel members were therefore: There is a broad understanding of the tension between affordability for customers now and the requirement to be able to finance our business plan. It would be concerning if a consequence of the Final Determination were an actual ratings downgrade. A smooth profile of bills remains preferable for customers and this preference would apply between AMP periods as well as within them. Sudden price hikes are very undesirable. There should not be an unfair burden on customers over time. The initial fall in bills in was probably something that was a given in the Final Determination. However, there are options which should be explored with customers

57 53 around the recovery of costs over time. In particular the profile of bills implied by the Draft Determination should be explored against alternative options where the profile of bills is smoother over AMP6 and AMP7. In engaging with customers directly the focus should be on bill profiles over time and the bill levels using both graphs and figures. Attempting to engage customers directly on technical regulatory parameters such as PAYG percentage will not be effective. The company needs to take responsibility for effective communication of bill changes. Customer Surveys Following the CEF workshop we undertook a short survey of 200 domestic customers which dealt specifically with the issue of bill profiles. The survey approach and questionnaire were discussed and agreed with members of the CEF, who suggested useful changes to make the survey more accessible and understandable to customers. The survey asked customers about the relative importance of the following four factors when we set customer bills: 1. Setting bills as low as possible between 2015 and Avoiding sharp increases in bills in Setting bills so costs are spread fairly between current customers (who pay their bills in 2015 to 2020) and future customers (after 2020) 4. Reducing the risk of the company being unable to raise finance as easily (and therefore raising customers bills) in the future It then asked customers to choose between two bill profile options as set out in the following graph: Average Household Bills 2015 to Year Risk Option 1 Option 2 The outcome of the survey is clear. 72% of customers surveyed preferred Option 2 which offers a flatter profile of bills in AMP6 but then avoids a price increase at the beginning of AMP7. Separately we have also undertaken a survey of 200 customers to gauge views about the acceptability of varying bills within period in the context of our plans to use an Outcome Delivery Incentive to fund leakage investment. This is described elsewhere but it concluded that while customers preferred to see a smooth and predictable profile of bills, 80% said they would accept variability of bills within period within a reasonable limit. Around 71% of those surveyed would be comfortable with annual bill variations of 5 or more.

58 54 Our Representations Customer bill profile and financeability Options We have considered a number of options for smoothing the bill profile within and between periods, as well as resolving potential financeability concerns arising from the Draft Determination. Broadly we considered both: How much revenue ought to be brought forward to AMP6 to ensure that our plan is financeable and how that revenue should be profiled over time What method we should use to effect it: PAYG ratios, depreciation of the 2015 RCV or depreciation rates applied to new totex added to the RCV How much revenue to bring forward The Draft Determination resulted in a number of changes to the revenue requirement compared to our June submission, for example as a result of Ofwat's totex baseline compared with our plan. As we set out above, 82m relates to Ofwat's decision to apply a single midnight adjustment to make the adjustment to RCV relating to AMP5 capital outperformance, rather then spreading this adjustment evenly over the five years as we had proposed. We considered what levers were available to us to manage any financeability issue and have concluded that because there are a number of levers available to us to manage financeability over the AMP that it would be inappropriate, given overall concerns of customers to maintain affordable bills, to seek to bring forward the whole of the revenues that had been removed. We therefore looked at three other options in detail Bringing forward an amount of 82m into AMP6 - equivalent to the impact of re-instating the midnight adjustment proposal in our business plan. Bringing forward an amount less than this. Bringing forward a greater amount. In all cases we further profiled the revenue recovery to ensure a smooth profile of bills for the remaining years of AMP6. PR14 Representations on the Draft Determination anglianwater.co.uk Bringing forward 82m would allow us to deliver a financeable business plan on both an actual and notional basis although financeability would remain tight. Moreover it would allow us to avoid a jump in bills in , which customers are keen to avoid. Customer feedback is in favour of the profile of bills which would be delivered by this option. In contrast, bringing forward a materially lower amount of revenue would also result in the likelihood of a sharp increase in bills in 2021 which would not fit with what customers have told us about their preferences for profiling of bills over time. We therefore concluded that we should seek to bring forward an amount of 82m and further that we should profile these revenues within the period to ensure that bills after the initial fall in bills in are profiled smoothly. How to bring the revenue forward We considered a number of options: The re-instatment of our approach to the CIS Midnight Adjustment. An adjustment to PAYG rates. An adjustment to RCV Run Off Rates.

59 55 An adjustment to the rate of Depreciation of New Assets. Combinations of the above methods. We also considered the idea of simply requesting that our method of smoothing the adjustment for capital outperformance should be re-instated. Whilst this would resolve the financeability issue satisfactorily, it would not deliver an initial reduction in bills in that seems to be a feature of all of the draft determinations. It would also mean asking Ofwat to apply a different methodology to our Final Determination compared to other companies. We therefore concluded that while this would be a sufficient measure it was unlikely that this would be an acceptable solution to Ofwat. We therefore looked in detail at the remaining options. Each of these parameters can be used alone or in combination to deliver the profile of revenues and bills which both ensures financeability and meets customer expectations. In considering which of these would be appropriate, we considered a number of factors The theoretical basis for the adjustment. The extent to which an adjustment to these parameters was consistent with our Business Plan and with any underlying characteristics of our business. The value of the specific parameter compared to others in the industry. The implications of customer feedback. The impact on financial ratios. We concluded that using depreciation rates on new totex was arguably least justifiable on a theoretical basis as the rates applied to this new spend ought to bear some relation to the economic life of the underlying assets represented by the spend even though this spend passes through the totex mechanism. In addition the adjustment required to the rate at which new totex would be depreciated needed to be large to achieve the desired effect. We considered that this would result in depreciation rates that were too high. Members of the Customer Engagement Forum were also uncomfortable with using this method. In terms of PAYG, we concluded that the PAYG rates we assumed ought to be reasonable when compared to the natural rates as discussed above. We note particularly that some of the rating agencies attach a degree of importance to the PAYG rate in evaluating creditworthiness. Fitch Ratings, in particular, looks for a PAYG rate which is consistent with underlying spend planned on operating costs and infrastructure renewals expenditure and would view an acceleration of revenues using PAYG at a level higher than this as a negative. We also noted that if we used PAYG rates then the rates we would need to apply would be high compared to others in the industry, especially for wastewater. While this may not be a reason in itself to reject this method, it suggests that the lever should be used cautiously. We show this in the following graphs. In these graphs and in the similar ones that follow: The blue bars are from the Draft Determinations of other companies The lilac bar - AW (DD) - is from Anglian Water's Draft Determination The purple bar - AW (Reps) - is the value proposed in these representations The pink bar - AW (Nat BP) - is the natural rate quoted in the June Submission and calculated based on IFRS accounting standards The dark red bar - AW (Nat IRE) - is the natural rate calculated assuming that all IRE is treated as pay as you go type expenditure

60 56 Our Representations Customer bill profile and financeability The graphs below show what PAYG rates (averaged over 5 years) would be required for the two services if the amount recovered were recovered wholly by adjusting PAYG rates. These show that the resulting PAYG rates are higher than those we assumed in our June Submission and as a result are at or near the top of the range of rates across the industry. They are also at or above the "natural" PAYG rate that is calculated by assuming that all IRE expenditure (even that which is capitalised under IFRS) should be considered as Pay As You Go type expenditure. The PAYG rates here are shown as an unweighted average to enable comparison with the PAYG of other companies from their Draft Determinations. 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% PAYG % - Water 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% PAYG % - Sewerage By contrast, if we used adjusted the Run Off rate for old RCV then the resulting PAYG rates are more in line with those used at the June Submission and appear to be more consistent with others in the industry. Some adjustment compared to the June Submission is needed to ensure a smooth profiling of revenues. The following two graphs show this. PR14 Representations on the Draft Determination anglianwater.co.uk 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% PAYG % - Water 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% PAYG % - Sewerage We also concluded that the use of RCV Run-Off was most attractive from a theoretical perspective, partly because it operates more "cleanly" on the balance between revenues and RCV and is less inherently related to characteristics of the underlying plan. The change in RCV Run-Off rates required is moderate, requiring an increase of c35bps compared to the June Submission. We have applied the same increase to both services so Run Off Rates are now 4.3% for Water (compared to 3.95% in June) and 4.2% for Wastewater (compared to 3.85% in June).

61 57 The graphs below show that under this scenario, the revised rates are broadly in the middle of the range of rates in Ofwat's Draft Determinations. 7.00% 6.00% 5.00% 4.00% 3.00% 2.00% 1.00% 0.00% RCV Run Off - Water 7.00% 6.00% 5.00% 4.00% 3.00% 2.00% 1.00% 0.00% RCV Run Off - Sewerage We discussed the options of using PAYG or RCV Run-Off with informed members of our Customer Engagement Forum. They were in favour of our using RCV Run Off rates. In terms of financial ratios, the use of RCV Run-Off rates results in Adjusted Interest Cover Ratios which are marginally tighter then using PAYG rates. This was a factor in our change in PAYG rates in June compared to our December Business Plan but is not a critical factor in the decision at this stage. Both options deliver similar FFO/Debt ratios. On balance therefore we have concluded from these graphs that an approach which alters RCV Run Off would be more appropriate. Our proposal for the Final Determination We propose that an adjustment is made to the RCV Run-Off rate is made for the Final Determination. Specifically we propose the following rates. Proposed RCV Run Off Rates Water Service Sewerage Service 4.3% 4.2% PAYG rates should be set as follows Proposed PAYG Rates Year Water 56.7% 50.0% 58.3% Sewerage 64.6% 50.2% 56.4%

62 58 Our Representations Customer bill profile and financeability Year Unweighted average Weighted average (by totex spend) Water 66.5% 69.2% 60.1% 59.4% Sewerage 54.3% 54.4% 56.0% 55.6% The outcome of this adjustment would be to recover an additional 82m of revenues in AMP6 compared to the Draft Determination and RCV will be lower at the end of the period by 98m. The resulting profile of bills over the period is as follows: Average Household Bills 2015 to Year Risk Option 1 Option 2 We calculate that this would result in the following key ratios, calculated on a notional basis, which we believe are acceptable given our target credit rating at Baa2, which is unchanged from the June Submission. We calculate the ratios on two bases: (i) that the only change to the DD figures is to alter the PAYG and Run Off Rate assumptions as set out above, and (ii) that our other representations are accepted. Implied Notional Financial Ratios - Average PR14 Representations on the Draft Determination anglianwater.co.uk Ratio Cash Interest Cover Adjusted Cash Interest Cover FFO/Debt Net Debt/RCV PAYG and Run Off Rate Change Only % 62.5% Other Representations Accepted % 62.1% The Draft Determination signals that Ofwat may reconsider its assumptions in relation to the cost of debt. We have argued that this would be not be appropriate. If a material change in the allowed cost of debt were to be made, however, there could be a knock on impact on our ability to finance our plans. We do not believe it is reasonable to expect companies to provide assurance about the financeability of their plans in advance. Our arguments on Ofwat's consideration of further reducing the cost of capital are set out in section 6.

63 59 4 Retail costs Ofwat's Draft Determination The base year for setting the HH and NHH Retail Controls is However, Ofwat required us to submit our plan in prices. Ofwat has said that we have failed to apply the indexation correctly and as a result has reduced our allowed costs in the retail price controls by around 10m in total. Summary of our Representations The impact of the change is material. Ofwat has been inconsistent in its treatment of inflationary cost pressures by applying RPI deflation to the base year for the control but not allowing for it during the control period. By doing so it has essentially placed an additional efficiency challenge on the retail businesses by means of a methodological change which was not adequately signalled. As a consequence the allowances for retail costs in all years of AMP6 are lower than in any year in AMP5. This is not reasonable. Ofwat should base our AMP6 retail allowances on our outturn costs. In its DD Ofwat has said that we failed to deflate our retail costs to prices and as a consequence has reduced the figure for by an amount which would represent the indexation of costs by the Retail Price Index from to The relevant figures relating to retail operating costs from our June submission are as follows: Impact of applying general RPI deflation to retail operating costs retail operating costs per June Submission Tables (Line 1 of Table R3/R4) Restated for RPI deflation applied in the Draft Determination ( / ) Reduction in retail costs Impact of reduction over AMP6 Household m Non Household m We consider the impact of the adjustment to be material in the context of the retail price controls. For example, for the Retail Household control the reduction is around 19% of the net margin allowed for in the DD.

64 60 Our Representations Retail costs Ofwat has been consistent in its methodology in stating that it does not consider water companies retail costs to be subject to inflationary cost pressure. Or rather, it has maintained the stance that it would not be appropriate to adjust either HH or NHH costs for such pressures without further evidence. At the same time it offered companies the opportunity to provide such evidence in order to demonstrate the existence of inflationary cost pressures which should be allowed for in retail price controls. A large majority of companies took up this opportunity. In one case, Yorkshire Water, Ofwat appears to have accepted the companies arguments and therefore presumably agreed that in principle such inflationary cost pressures could exist. Anglian Water did not propose any inflationary increase in retail costs in its Business Plan. Like Ofwat we were doubtful that RPI was a particularly good measure of inflationary pressures applied to the specific basket of inputs required to deliver retail services. That is not to say that we believe there are no such pressures. Rather we consider that where such pressures exist, then we could commit to the significant challenge of absorbing such pressures by means of better efficiency in that part of our business. In other words we chose to accept an efficiency challenge. To be precise the scale of the efficiency challenge we chose to accept in our Business Plan was the difference between the retail costs and the efficient level of costs in each year of AMP6. The company therefore took on the risk that cost pressures could not be controlled. In other words we accepted Ofwat's methodology. However, in our view the implication is that any adjustments to express figures in a different price base should be made using a consistent view of inflationary pressures. We therefore believe that the appropriate adjustment to make to reflect retail cost in prices is zero. This is the basis on which we submitted our plan. We are therefore disappointed that, in deflating costs by RPI in the Draft Determination, Ofwat appears to have ignored its own views. We believe that the fact that Ofwat has made this adjustment implies that PR14 Representations on the Draft Determination anglianwater.co.uk between and Ofwat believes that RPI is indeed a good measure of inflationary cost pressure if it does not then it believes it is in any case appropriate to make a further downwards adjustment to Anglian Water's cost base which is, essentially, an additional efficiency challenge. If the first of these holds, then it appears to raise a strange inconsistency in Ofwat's logic: it would mean that Ofwat believes that RPI is a good measure of inflationary cost pressure up to 31st March 2014, but not thereafter. This would seem a difficult position to hold, in any case, and particularly given Ofwat's clear statements on the topic in its methodology. However, if the second position holds, it means that Ofwat has, essentially, applied an additional efficiency challenge on the retail businesses which has never been clearly signalled in its methodology. Specifically it has never been open that it sees removing indexation from the retail price controls as an efficiency challenge. Moreover it has been clear that companies should see as being the key base year for the determination of retail costs and companies have planned on this basis. At a practical level the change that Ofwat suggests seems out of line with historic evidence of our actual retail costs. In the table below we set out actual costs in price of the day over AMP5.

65 61 AMP5 Retail Costs POTD Household Household Adjusted Non Household deflated Household retail costs in and are high because of peaks in our work load on demand-led water efficiency initiatives (driven by dry conditions) and because of customer-side leak work, which from is no longer offered free of charge to customers. Excluding for these effects our costs are remarkably consistent in nominal terms - see "Household Adjusted" column. This provides some support for the the view that inflationary pressure on retail costs can be offset by efficiencies, at least to an extent. However, it also means that by applying deflation to costs the resulting allowance for expenditure in the Retail Household Control is substantially lower than the spend actually incurred in outturn prices in any year of AMP5. We believe that this makes the adjustment unjustifiable. We have been asked by Ofwat to explain the movement in Non Household Retail costs between and We provide details on this in Section 11. In summary it relates to a change in approach on dealing with credits on account which impacts the bad debt charge. This means that the figures in and are not representative of the level of costs from and future years. We consider that Ofwat's position on this is both inconsistent from a theoretical and unreasonable in its impact on the actual allowances. On this basis, therefore we believe that Ofwat should re-instate the indexation adjustment and so increase allowances for costs in the Draft Determinations.

66 62 Our Representations Retail costs PR14 Representations on the Draft Determination anglianwater.co.uk

67 63 5 CEF engagement Summary of our Representations We have continued to engage customers and the CEF on the implications of the Draft Determination for our plan. This section summarises our engagement and how the role of the CEF will evolve post PR14 to hold us to account in delivering the final plan. Overview We have conducted further engagement with our Customer Engagement Forum and four Independent Advisory Panels on the Draft Determination and commissioned two new customer surveys. This has helped us ensure that the representations we make reflect customers' views and priorities as well as provide the additional evidence that the Draft Determination requested. Engagement with CEF We held conference calls with the Independent Chair of the CEF and the Deputy Chair to discuss the Draft Determination and its implications for our plan, and to agree how best to explore areas of key challenge with customers and the full CEF. Following these calls, six topics were identified where further engagement was either requested by Ofwat or considered important to ensure the CEF could adequately advise and challenge us on our proposed representations as well as make its own representations to Ofwat. These are summarised in the table below. Topic/key challenge for engagement Outcome Delivery Incentives Leakage Outcome Delivery Incentive and principle of in period rewards/penalties Agreed approach for engagement Discuss with CEF/panels Discuss with CEF/panels Additional customer survey Dealing with uncertainty Enhanced metering and water efficiency Non household default tariffs Bill profiles and financeability Discuss with CEF/panels Discuss with CEF/panels Discuss with CEF/panels Discuss with CEF/panels Additional customer survey

68 64 Our Representations CEF engagement Joint CEF/Independent Advisory Panel meeting We had already invited all CEF and panel members to a meeting on 17 September to review the DD and explore the evolution of the CCG after the end of PR members were able to attend on the day representing a good cross section of the panels and different stakeholder interests. Bringing everyone together in this way also helped to expose any different perspectives that might exist between members. The briefing notes on each topic were also circulated to all members before the meeting so those unable to attend could also share any thoughts. At the meeting Peter Simpson, Chief Executive, Anglian Water explained the implications of the Draft Determination for customers' and CEF priorities in the company plan and Jean Spencer, Regulation Director, Anglian Water gave an overview of the topics of challenge in the Draft Determination. Members then spent time discussing these issues in more depth to come to a conclusion on their views for each. Anglian Water experts for each topic were on hand to answer any queries during these discussions and hear stakeholder views first hand. The afternoon was spent reflecting with members on the CEF and panels process throughout the PR14 process to date; how best to evolve the CEF and panels in the future to have a role in holding Anglian Water to account for the delivery of the plan; and continuing to advise and challenge on future issues. Proposals from the CEF for a revised role, structure and format were shared in advance with all members and Dame Yve Buckland led the session to explore and develop these with panel members. Customer surveys We have conducted two additional surveys with customers about issues raised in the Draft Determination. 1) Leakage ODI and principle of in period rewards/penalties PR14 Representations on the Draft Determination anglianwater.co.uk This survey explored customers' attitudes to funding some improvements, for example leakage, through ODIs so that customers only pay for the improvements when (and if) Anglian Water has delivered them with bills changing in period. It asked customers about the acceptability of this new approach and explored how this varied from acceptability of the current approach. It also explored attitudes to what is an acceptable amount of bill fluctuation in any year. McCallum Layton conducted the survey on our behalf, interviewing 400 household customers in an online survey. The sample was selected to ensure representativeness to the population. The survey was developed with input from the CEF Customer Engagement Champion who reviewed and fed back comments on the questionnaire. McCallum Layton's report summarising this research is found in an appendix to these Representations. The results are discussed in section 10.

69 65 2) Financing the plan This survey explored customers' attitudes to the impact on bills of the choices that Anglian Water can make when seeking the most sustainable way to finance the plan. It explained how different choices in the financial assumptions can affect the plan and customers' bills and it presented two scenarios for customers to consider and choose which they preferred (one based on the Ofwat DD and one based on a proposed alternative). McCallum Layton conducted the survey on our behalf, interviewing 200 household customers in an online survey. We asked the CEF for their advice on the best way to explore this topic with customers and their advice informed the way that we presented the choices to customers. The CEF recommended that we included supporting text to set out bill impacts and took the forecast through to beyond A copy of the final graph presented to customers and extract of some of the supporting narrative on the two options is shown below. Scenarios presented to customers in the survey Option 1 lower bill in 2015 to 2020 (purple line) In this option, average bills will reduce steeply from 408 in 2015 to 380 in 2016 and will then continue to fall through to 363 by Bills would then increase steeply to 386 in 2021 before rising slightly to 389 in 2025 after which they would fall to 383 in 2026 (the same level as option 2).

70 66 Our Representations CEF engagement Option 2 smoother long term bill (orange line) In this option, average bills also reduce steeply from 408 in 2015 to 380 in 2016 as for Option 1. They then reduce only slightly to 375 in 2020 before rising gradually to 383 in 2026 (the same level as option 1). The questionnaire set out the advantages and disadvantages to both these options including how bills might be impacted from 2020 if the cost of finance increased (shown as the dotted purple risk line). It also asked customers which scenario they preferred and why. The questionnaire was developed with input from the CEF Customer Engagement Champion who reviewed and feedback comments on the questions. A copy of the final version is found in McCallum Layton's report summarising this research in an appendix to these Representations. The results are discussed in section 10. Findings of our engagement The results of our engagement with CEF and customers are summarised in the relevant section of our representation: Topic/key challenge for engagement Outcome Delivery Incentives Leakage Outcome Delivery Incentive and principle of in period rewards/penalties Dealing with uncertainty Enhanced metering and water efficiency Non household default tariffs Customer bill profile and financeability Section (see below) and 10 PR14 Representations on the Draft Determination anglianwater.co.uk There is one area that we have explored in detail with the CEF but have decided not to make a representation on ourselves - dealing with uncertainty. We believe this is an important area and have set out in other submissions (for example, in our revised Business Plan submission from June 2014 pages and in the commentary supporting our data tables in both the December 2013 and June 2014 submissions) our reasoning for retaining logging as an uncertainty mechanism for growth and environmental obligations. We continue to believe that such a mechanism would protect both customers and companies in the event of material variances beyond the company's control in those areas. However, it is clear from the Draft Determinations that Ofwat has issued that this is not a mechanism that Ofwat is prepared to consider, and so on that basis, we have decided not to include representations to retain logging. Nevertheless, the CEF has made comment that in their view the logging mechanism should be retained. The CEF representation includes details on their view and a summary is provided below.

71 67 Mechanisms for dealing with uncertainty The CEF concluded: "We would like to see a logging mechanism retained for the impact of growth on wastewater costs and the additional environmental improvements under the National Environment Programme (NEP). Without such a mechanism we would be concerned that efforts to control the scope of requirements and their costs could lead to a less collaborative relationship between Anglian Water and the Environment Agency and may impact on delivery of River Basin Management Plans and compliance with the Water Framework Directive or that the company seeks an IDOK which would lead to greater uncertainty in customers bills. We would also be concerned that there is an increased risk that water and wastewater infrastructure becomes a brake on growth." CEF representation p3. Ongoing assurance role for CEF At our joint CEF and panel meeting we discussed reflections on how the process had gone so far and proposals for how the group role and structure would evolve post PR14. Dame Yve Buckland shared the proposals on evolving the CEF that the forum had agreed earlier in the year and as set out in our June supplementary report. The outcome of the discussions was the approach proposed below. Proposed CEF revised structure Summary of key features of possible future structure: Customer Engagement Forum: A new independent chair will be appointed. CCWater has proposed that it manages the process for recruiting chairs for the next price review to help ensure their independence. For the Anglian CEF its suggested that Bernard Crump takes over from Dame Yve Buckland as an interim chair until the recruitment process is

72 68 Our Representations CEF engagement agreed nationally with Ofwat. Bernard is currently a member of the CEF and Economic Development Panel and Regional Chair of Consumer Council for Water Central and Eastern. New members. New members will include the Citizens Advice Bureau and CBI, as well as consideration being given to also including the federation for Small Businesses and / or a Chamber of Commerce, to help strengthen customer representation. This would replace the current business representative. An expert with a background in economics and regulation would also become a core member of the CEF. Panels and wider engagement: The proposal aims to find a more flexible mixture of engagement methods that can be tailored to specific topical issues and make engagement via existing forums and tools more visible. Organisation and focus: The CEF will meet three times a year. At its first meeting each year in Q1, it will review the progress of the company for the previous year in delivering against the plan. Its view of performance will be publicly reported in the company annual report and accounts each July. The CEF will also agree the agenda for the annual regional customer and stakeholder forums at this meeting. Around September, stakeholders will be invited to a series of regional customer and stakeholder forums to review the company progress against delivery of the plan and explore key issues. The CEF will review the outcomes of these to agree requirements for any working groups or further research to explore the issues in more depth. Working group chair(s) will be appointed and will join the CEF for the period of their working group activity. The CEF will meet in Q4 to review progress. More use will be made of facilities such as Skype for meetings - where appropriate. PR14 Representations on the Draft Determination anglianwater.co.uk The Environment panel is excited to continue setting collaborative agendas for future working, and making an early start on engagement for the next AMP. It intends to review its membership to ensure it has a good representation of different organisations. The Environment and climate change panel will meet annually. The Hartlepool group, comprising a cross section of stakeholders will also continue to meet on a regular basis to consider the particular needs for this part of our region. PR19 is anticipated to begin around 2017 and there will be a check-point at this stage to review how best to engage in the next price review for example continue the working group format, reinstate panels or another approach. Overall, members felt the process had been useful and they felt they had had a positive influence on the company plan and strategy. They felt it was important to continue the engagement and were supportive of the proposed new role and structure with some suggestions for developing it further.

73 We are delighted that the CEF is keen to continue to play an important role in holding us to account in the delivery of our plan. We will continue to refine the proposals for the CEF post PR14 through discussions with Dame Yve Buckland and the CEF. The next meeting of the CEF is in January 2015, where the Final Determination will be reviewed and the CEF will move into its assurance role. 69

74 70 Our Representations CEF engagement PR14 Representations on the Draft Determination anglianwater.co.uk

75 71 6 Cost of Capital Ofwat's Draft Determination Ofwat has signalled that it may review its cost of capital assumptions at the Final Determination. Summary of our Representations There appears to be little justification for any reassessment of the cost of equity. Ofwat's assumption is not substantially different from other recent regulatory assessments. Whilst interest rates have moved below market expectations since the start of the year, this should be considered against the guidance given by the Governor of the Bank of England, and short term geo-political issues. Economic forecasts still point to short term interest rate increases in the UK and the US in Consequently, we regard the lower rates seen recently to be short term, and unlikely to be sustained during the period to Regulatory risk is a substantial concern for investors, as evidenced by recent Water UK surveys. Changing a well-signalled cost of capital at the last stage of the price review will tend to increase perceptions of regulatory risk. A further reduction in the cost of capital at this late stage also discourages companies from taking a long term view and taking prudent action to manage their costs. Even if a modest reduction in the assumed cost of debt could be objectively justified, investors are already unsettled by the changes throughout this price review process, and further changes would impact investors perceptions of risk in the water section and may well result in disbenefit to customers in the longer term with additional returns demanded for regulatory risk. Ofwat should therefore maintain the assumptions which were signalled in its Risk And Reward Guidance. In publishing the Draft Determinations, Ofwat has emphasised that it may review the cost of capital for Final Determinations to take account of market and regulatory developments. It is clearly reasonable to consider such market developments but we would urge a balanced view of any recent market developments. We believe that recent falling interest rates may well reverse in the short term. Moreover, any change in cost of capital assumption at this late stage of the review could further damage investor confidence in the water sector, to the real detriment of future customers.

76 72 Our Representations Cost of Capital Market Evidence There appears to be little justification for any reassessment of the cost of equity. There are a number of reasons for this. Firstly, recent proposals/determinations from Ofgem, the CMA and Ofcom sit broadly in line with Ofwat s calculations of the cost of equity (allowing for the differences in gearing assumptions). Secondly, the cost of equity estimate should ultimately be forward-looking, and it is hard to argue that the market outlook for the period to has changed materially in the last eight months. Third, there is no evidence of a change in the expected market return or beta parameters. In relation to the cost of debt, Ofwat point to a fall compared to market expectations between January and August Clearly, however, the market may continue to change over the next few months in response to developments in, for example, European responses to stagnation, as well as the impact on debt markets of unrest in the Ukraine and the Middle East. Economic forecasts still point to short term interest rate increases in the UK and the US in Recent guidance given by the Governor of the Bank of England would tend to support this view. There is some evidence that the market is already responding to these pressures. The graph below shows yields on 10 year gilts over the year. The inflection point in August 2014 is striking and at least throws into doubt whether the falls up to August 2014 will prove to be enduring. Gilt yields PR14 Representations on the Draft Determination anglianwater.co.uk GBP10y On balance we believe that it is more likely than not that the lower rates seen this year will prove to be short term and unlikely to be sustained during the period to On this basis we would not agree that the market evidence as it stands justifies any change in the cost of debt assumption. Long term impact of changing the cost of capital assumption now The long-term stable regulatory environment has given investors confidence to finance over 100bn of investment since privatisation, with capital provided at low rates, a benefit which is being fully passed on to customers in a lower cost of capital. For the benefit of future customers it is vital that this long term stability is not jeopardised in seeking a marginal additional benefit based on short-term volatility in the market. Ofwat has to balance the

77 73 benefits that a lower cost of capital in the short term delivers in immediate bill reductions for customers, against ensuring that companies can remain financeable, and satisfy debt covenants and credit ratio hurdles required by investors and rating agencies. Indepen carried out a survey (1) of investors in May 2014, which highlighted that Ofwat's Risk and Reward guidance had made the sector less attractive. The majority of investors see the sector as being at least as risky as, or more risky than, electricity and gas transmission and distribution. The top risk for all investment types was regulatory risk. The views of equity, and in particular unlisted equity were the most negative of all investors. These views might be dismissed by Ofwat as to be expected in the light of a well trailed tightening of the regulatory settlement. But the views of bondholders cannot be so easily ignored: 86% of bondholders said that Ofwat's risk and reward guidance made the sector "significantly less attractive or less attractive" and nobody thought it was "more attractive"; similarly 86% thought that regulatory risk had increased since March 2013; 72% of bondholders had regulatory risk and 86% risks associated with financing amongst their top three risks; 78% thought that Ofwat's guidance on financeability made the sector "significantly less attractive or less attractive" than previously. Ofwat has highlighted the extensive level of engagement with investors throughout the the price review process and this is recognised in the Indepen survey. However, almost 80% of bondholders disagreed with the statement that Ofwat is listening to investors. Bondholders do not have to invest in the water sector and there is a real risk if Ofwat reduce the cost of capital further then bondholders will see that as a further increase in regulatory risk with the consequence that, other things being equal, the cost of debt will increase to the detriment of customers in the longer term. A further reduction in the cost of capital at this late stage also discourages companies from taking a long term view and taking prudent action to manage their costs. In preparing our business plans in 2013, we locked (2) into forward market rates to cover a significant proportion of our debt raising for AMP6. The locked-in cost of debt was reflected in our proposed cost of capital. Our approach was in tune with Ofwat's guidance at the time that guided companies to seek to lock in low rates (3). A further fall in interest rates could not have been predicted at the time we submitted both our December and June submissions, and when we prudently hedged our forward debt issuance to mitigate the risks of rates increasing. On that basis we locked in future debt costs, at the low market rates available, in the interest of mitigating risks for company and in the interests of customers. In summary, we do not consider it is in the long term interest of customers to further reduce the cost of capital. 1 Water UK Survey of Investors. Indepen, published June In response to recent market movements, we have put in place financial instruments which partially protect against future rises in debt raising costs, which in turn allow us to limit our cost of debt assumption in the Plan", Our Plan , page 41 3 See for example Sonia Brown's address to Sustainable Water 2014 on 11 September 2013

78 74 Our Representations Cost of Capital PR14 Representations on the Draft Determination anglianwater.co.uk

79 75 7 Risk and reward 7.1 ODIs Ofwat's Draft Determination Ofwat has made a number of additions to and removals from Anglian Water's suite of ODI measures. Company specific appendix, pages Summary of our Representations We are not making representations on most of Ofwat's interventions on our package of ODIs in the Draft Determination. We are proposing a change to Ofwat's Determination with respect to supply interruptions, on the basis that the Draft Determination over-simplifies what should be considered upper quartile performance, and the resulting incentive is not consistent with the variability inherent in the measure. We have included our revised proposal for this ODI. We have provided additional information where requested in the Draft Determination, including proposals for new ODIs. We are not making representations on most of the interventions made by Ofwat to our package of ODIs, and we are providing additional information or designed new ODI penalties and rewards where requested. In principle, we understand Ofwat's desire to use ODIs as a disruptive force to stimulate improvement over a range of measures. However, no single company is upper quartile across all of the measures where this approach has been taken. This means that most companies will find the Draft Determination extremely challenging in one or more respects with their package of ODIs. In making the representations below, we have been careful to consider the Draft Determination in the round. This has led us to make only one significant representation with respect to the ODI for supply interruptions. CEF views on the ODIs in the Draft Determination We do not wish to see changes to the ODI make the company more risk adverse or stifle innovation. We support the principle behind upper quartile performance for ODIs before a company can earn a reward. However, in some instances the threshold chosen may be overly simplistic and may lead to perverse implications. An example would be supply interruptions where we support the company s recommendation to set targets that reflect an upper quartile position adjusted to take account of the nature of the company s networks. CEF Representation p3

80 76 Response to Ofwat's Specified Requirements Risk and reward Representations on Outcome Delivery Incentives Supply interruptions We agree that supply interruptions should be a focus for the ODIs and that rewards should only be earned for upper quartile performance. However, we do not consider that the proposed performance commitment and design of the incentive is appropriate. Upper quartile calculation In calculating upper quartile performance for this measure, we note that Ofwat has calculated the upper quartile level in each of three years and then taken an average. This approach is inconsistent with that taken for other measures, such as water quality complaints, where instead an average company performance across three years is first calculated and then the upper quartile of these averages is used. This makes a significant difference to the assessment of upper quartile as shown below (12.3 minutes against 10.4 minutes). Company year average ANH WSH NES SVT SWT SRN TMS UU WSX YKY AFW (VCE in and ) PR14 Representations on the Draft Determination anglianwater.co.uk BRL DVW PRT SBW SEW SSC (SST in and ) SES Upper quartile Average of upper quartile (Ofwat calculation)

81 77 This is an inappropriate use of the upper quartile, and does not represent a real company's sustained performance. For example, Dee Valley Water would be considered upper quartile against a 10 minute commitment in , but be 2 minutes above this in and 5.6 minutes above this in In both cases this is a greater variance than the penalty collar proposed by Ofwat in the Draft Determination. In other words, an 'upper quartile' performer would have incurred the maximum penalty two out of three years. We suggest that this indicates that not only should the upper quartile calculation be based on average company performance, but also that inherent variability in this measure needs to be addressed through either averaging or deadbands (in all years, not just the glidepath to 'upper quartile'). Variability in supply interruptions relative to the upper quartile The assessment of upper quartile performance level also needs to be calibrated to reflect the nature of a company's network. There is a clear correlation between supply interruptions (minutes) per property and length of main per property:

82 78 Response to Ofwat's Specified Requirements Risk and reward Supply interruptions correlation to length of main per property This should not be a surprise - it is well established, in the way that Ofwat has previously assessed service associated with burst mains, that there is a correlation between bursts and length of main (1). Companies that have longer lengths of main per property would be expected to have more bursts per property and consequently more interruptions per customer. In the Draft Determinations Ofwat has applied an overly simplistic performance level. As illustrated in the chart above, the ten minutes performance achieved by the "upper quartile" companies is largely a factor of their networks. We propose that the performance level should be based on the upper quartile performance adjusted for length of main per property. Such an approach would be consistent with Ofwat's approach for pollution incidents, which is normalised by length of sewer. PR14 Representations on the Draft Determination anglianwater.co.uk To assess a normalised upper quartile level, we shifted the trend line down until four companies were below the trend. This approximates upper quartile performance, given the length of main per property served. This implies a committed performance level for Anglian Water of 17 minutes. This performance level remains challenging, especially given the normal level of variability seen in this measure year on year. Nevertheless, if this adjusted performance level is reflected in the Final Determination then the challenge could be set of using it as a basis for rewards and penalties from (i.e. no glidepath) if using annual targets, or as we propose below, through an interim target based on a three year average, which implies performance at 17 minutes from onwards. Economic assessment In our original business plan submission, we included evidence of the economic level for supply interruptions, according to customer Willingness to Pay. We accept that this analysis takes a static view of efficiency and that in setting targets at upper quartile performance, 1 Ofwat's measure of service performance in June Returns (JR11, Table 11, line 12) was bursts per 1000km of main

83 79 Ofwat seeks to stimulate dynamic efficiency improvements which might justify the improvement. However, by our analysis, incremental costs would need to be halved in order to justify an economic level of supply interruptions of 10 minutes. We do not consider this to be a realistic expectation to be achieved by Supply interruptions: Economic analysis Variability and annual assessments The Draft Determination provides a very narrow range over which the rewards and penalties apply within a single year assessment. We believe that this will result in perverse incentives and be to the detriment of customers. As shown above, almost all companies experience a wide range of performance which exceeds the range of penalties and rewards in the incentive. This measure is susceptible to weather effects, but can also be affected by one-off incidents. As with leakage, basing a target on a single year's performance means that we would need to aim significantly below the target in order not to miss the target if such an event or extreme weather were to occur. This would require significant investment in what would in effect be redundancy in our operations, 'just in case'. Assuming the run-rate for supply interruptions can be reduced to meet the performance commitment, then a single one-off event can result in the performance level being missed and the question then is whether the penalty is proportionate? To put this into context a single interruption of 12 hours, affecting 3,700 customers could result in the full penalty of over 8m being incurred. If such an event occurs at the beginning of a year, then there could also be a temptation for companies that are less focused on doing the right thing for customers to reduce the response to interruptions to save cost in the rest of the year. We propose the use of a three year average to better maintain the incentives to perform.

84 80 Response to Ofwat's Specified Requirements Risk and reward Proposed incentive We propose the following: Performance should be based on a three year average, not a single year. Performance can be assessed in and again in , with no glidepath, which implies the commitment should be met (on average) in each year from The assessment in to be worth one fifth of the incentive, and that in to be four fifths (in the same way as proposed in the Draft Determination for low pressure). Upper quartile should be determined relative to km main per property. We assess this to equate to a target of 17 minutes, given our network. Whilst we have retained a relatively narrow range of rewards and penalties, by taking a three year average, our proposal reduces fluctuation in penalties or rewards due to isolated events. Supply interruptions: Performance commitments Unit Starting level Committed performance levels PC Mins/prop/year Penalty collar Mins/prop/year Penalty deadband Mins/prop/year Reward deadband Mins/prop/year Reward cap Mins/prop/year Supply interruptions: Incentive rates Incentive type Performance levels (obligations delivered) Lower Upper Incentive rate ( m/incident/year) PR14 Representations on the Draft Determination anglianwater.co.uk Penalty Reward

85 81 Supply Interruptions ODI Interventions where no change to the Draft Determination is sought Mean Zonal Compliance At the request of Ofwat, we included an incentive relating to Mean Zonal Compliance in our June submission. We are not seeking to make representations on the interventions made on this incentive in the Draft Determination, as we understand that Ofwat seeks a consistent measure of water quality across companies. We also understand the DWI view that any target on Mean Zonal Compliance should be 100%. This does not necessarily need to translate into a commitment of 100% for the purposes of an economic incentive. We are concerned that a commitment of 100% is unlikely to be delivered (given that some failures are attributable to customer fittings, for example), and that not meeting the commitment may prove a difficult message for customers to understand. The power of a commitment is undermined if there is no realistic means of achieving it. Low Pressure We are not making any representation on the change to the reward rate in the Draft Determination, the introduction of an interim commitment of 361 by , and the associated penalty collar and deadband which accounts for one-fifth of the overall incentive. We also note the change of the reward deadband to 230 properties. All of the assessment of rewards in our incentive is made in and applies to five years of revenue in AMP7. Value for money We are not making any representation on the condition that rewards for the water and sewerage versions of this incentive rely on our absolute score improving, as well as relative score.

86 82 Response to Ofwat's Specified Requirements Risk and reward Leakage We are not making any representation on the increased penalty rate as proposed in the Draft Determination. Internal Flooding We are not making any representation on the reward deadband, cap and rate as proposed in the Draft Determination. Calculation of reward rates We note Ofwat's interventions to change our reward rate calculations for some incentives, and Ofwat's response to our query on this matter. We do not agree that companies should only be able to earn rewards if they are able to reduce costs to justify additional delivery. There are other reasons why a commitment might be set at a level below the economic level at a given cost, such as the practicality of achieving the economic level in a single AMP. Innovation in the period may not just affect costs, but might also provide alternative approaches which allows a company to deliver more than was originally considered possible. However, we are not making any representation on the change, on the basis that in a number of areas, Ofwat's interventions depart from our economic analysis and customer willingness-to-pay, making the point somewhat academic. Additional information requested in the Draft Determination Water quality complaints We have accepted Ofwat's proposed incentive for water quality complaints, although we consider the assessment of upper quartile performance to be an extremely challenging commitment. PR14 Representations on the Draft Determination anglianwater.co.uk We propose that a reward is justified in this case on the basis that a step change in performance which exceeds this challenging commitment should be rewarded, as it is unlikely to be achieved without considerable effort and expense, and we have not included any expenditure in our plan to achieve this. We have proposed that the reward rate should be equal to the penalty rate proposed by Ofwat. The reward cap is set at 0, since this is the best performance that can be achieved. Like the setting of the penalty rate and collar by Ofwat in the Draft Determination, this is not based on economic analysis.

87 83 Proposed incentive Water quality complaints: Performance commitments Unit Starting level Committed performance levels PC No. / 1000 customers Penalty collar No. / 1000 customers Penalty deadband No. / 1000 customers Reward deadband No. / 1000 customers Reward cap No. / 1000 customers Water quality complaints: Incentive rates Incentive type Performance levels (obligations delivered) Lower Upper Incentive rate ( m/incident/year) Penalty Reward Water quality complaints ODI Pollution incidents (Category 3) As we have consistently stated since our original business plan submission, we believe that there are adequate incentives for companies to deliver their statutory obligations including: Enforcement powers of the quality regulators: the Environment Agency and DWI

88 84 Response to Ofwat's Specified Requirements Risk and reward Reputational incentives Serviceability (which includes a measure of category 1, 2 and 3 pollution incidents). Nevertheless, we recognise that Ofwat seek to apply an additional incentive to improve performance in this area. We therefore are not making any representation on the proposals for an incentive as laid out in the Draft Determination, and propose the following as required by the determination. Upper quartile calculation In principle we agree with Ofwat's approach to setting the upper quartile, normalised by length of sewers to be reasonable. However, there are significant differences in the extent to which companies self-report pollution incidents. Self reporting of pollution incidents (Source: Environmental performance of the water and sewerage companies in 2013, EA, August 2014 ) Company Anglian Water Northumbrian Water Severn Trent Water Southern Water South West Water Thames Water United Utilities Wessex Water Yorkshire Water Sector Self reporting 76% 39% 69% 77% 55% 58% 69% 59% 80% 66% PR14 Representations on the Draft Determination anglianwater.co.uk Self-reporting has increased significantly in recent years, but we are concerned that an ODI on pollution incidents could potentially discourage this trend, especially if too heavy a financial penalty is applied to each incident. We also note that all category 3 incidents are included, regardless of whether they are consented. Differences in company consents is not considered in Ofwat's assessment of upper quartile. We are concerned that this could encourage unnecessary investment to avoid penalties. This is another factor which should discourage overly aggressive incentives. Another alternative might be to concentrate the incentive on un-consented pollution incidents. Penalty collar and reward cap The Draft Determination requested that we suggest a penalty collar and reward cap. We consider that it is appropriate to set a collar at 500 incidents and a reward cap at 135 incidents. This means that a penalty would apply if performance deteriorated. Rewards are not extended to reach all the way to 0 incidents to maintain a higher downside risk than upside reward. A reward deadband is introduced at 293 incidents in all years as required by the Draft Determination.

89 85 Penalty and reward rate This incentive was not part of our original business plan, and is therefore not supported by the same level of economic analysis as our original proposals. This ODI is based on cross industry comparison, and so we have also used cross industry analysis to select a suitable incentive rate. Most companies appear to have incentives which include Category 1 and 2 pollution incidents. The incentive rate for Category 3 incidents should therefore be lower than these incentives. We have taken the incentive rate from the Draft Determination for South West Water at 0.009m per incident as both the penalty and reward rate. We propose this as a pragmatic solution to introduce an incentive at this stage in the determination. Proposed incentive If there is a substantial change to how the EA classifies pollution incidents, these commitments would need to be revised. Pollution incidents (category 3): Performance commitments Unit Starting level Committed performance levels PC Incidents Penalty collar Incidents Penalty deadband Incidents Reward deadband Incidents Reward cap Incidents Pollution incidents (category 3): Incentive rates Incentive type Performance levels (obligations delivered) Lower Upper Incentive rate ( m/incident/year) Penalty Reward

90 86 Response to Ofwat's Specified Requirements Risk and reward Pollution incidents ODI Environmental compliance The Draft Determination requires us to develop ODIs for environmental compliance relating to investment under Eel Regulations and RSA (water wholesale) and UWWTD and WFD (wastewater). We are proposing two penalty only ODIs based on delivering obligations under these drivers that are detailed in our business cases. A penalty would be incurred if any of these obligations are not delivered by the end of AMP6, unless they have already been adjusted for through an IDoK, or if it is as a result of agreement with the Environment Agency. The incentive rate is based on the Equivalent Annualised Cost and the number of obligations in the following business cases: Environmental compliance (water): relevant business cases PR14 Representations on the Draft Determination anglianwater.co.uk Business case Eel Regulations Sustainability reductions (RSA) Environmental compliance (water) Obligations EAC ( m) Average incremental cost ( m) Notes Excludes 9 reservoir studies Hunstanton RZ, Norwich & the Broads RZ, Mattishall WTW

91 87 Environmental compliance (wastewater): relevant business cases Business case Obligations EAC ( m) Average incremental cost ( m) Notes UWWTD site storm tanks (2000 pe), 4 sites 10,000 pe GWWD Redgrave, Castle Acre, Balsham and Neatishead King St No deterioration (2015) obligations at 21 Water Recycling Centres (WRCs) WFD Ammonia & BOD Enhanced secondary treatment at 3 WRCs WFD Phosphorous Improvements at 28 WRCs No deterioration (2021) Improvement at 6 WRCs Environmental compliance (wastewater) Proposed incentives These incentive details are below, with the penalty rate set to equal the average incremental cost above. Environmental compliance (water): Performance commitments Unit Starting level Committed performance levels PC Obligations delivered Penalty collar Obligations delivered Penalty deadband Obligations delivered Environmental compliance (water): Incentive rates Incentive type Performance levels (obligations delivered) Lower Upper Incentive rate ( m/obligation/year) Penalty

92 88 Response to Ofwat's Specified Requirements Risk and reward Environmental compliance (wastewater): Performance commitments Unit Starting level Committed performance levels PC Obligations delivered Penalty collar Obligations delivered Penalty deadband Obligations delivered Environmental compliance (wastewater): Incentive rates Incentive type Performance levels (obligations delivered) Lower Upper Incentive rate ( m/obligation/year) Penalty Proposed reference level and control limits for AMP6 reactive mains bursts serviceability measure In our December Business Plan we proposed a new serviceability measure for burst water mains. This measure would be designed to remove the effect of extraneous factors on the total number of burst mains currently recorded for serviceability. We demonstrated in our June business plan resubmission that the two extraneous factors that most affected the previous measure for mains bursts were variation in prevailing annual weather conditions and the number of bursts detected proactively through leakage reduction efforts. Our proposed measure therefore counts reactive bursts only and incorporates an allowance for weather variations. At the time of our June submission we had not completed the analysis necessary to make a proposal on reference level and control limits for reactive mains bursts and in the table on page 451 left these as to be confirmed. We have completed this analysis and now submit our proposal. PR14 Representations on the Draft Determination anglianwater.co.uk In order to reduce our leakage beyond the SELL of 211 during AMP5, we have substantially increased the number of bursts we have proactively detected and fixed. Removing these bursts from the count of total bursts is straightforward as we have recorded these separately for many years. Year Number of proactively detected bursts Percentage of bursts detected proactively 5% 11% 19% 13% 14% 15%

93 89 Year Number of proactively detected bursts 1,199 1,322 1,317 1,434 Percentage of bursts detected proactively 20% 22% 27% 29% Removing the effect of extraneous weather factors is more complicated. To understand the impact of weather on mains bursts in AMP5 we developed a model that correlated the number of reactive bursts against certain key meteorological observations: Minimum temperature (oc) Cumulative days of temperature below 1oC (days) Maximum fluctuation in temperature (oc) Cumulative days of air frost (days) Maximum soil moisture deficit (mm) Change in SMD (mm/d) This model enables us to calculate, after the end of the year, the number of reactive bursts we should have expected to observe given the meteorological conditions that obtained. We call this RBm. Our proposal for AMP6 is that we use RBm as an annual benchmark for determining our performance against the mains bursts serviceability measure. In this approach we define our reference level and control limits as percentage variations of actual bursts against RBm. For example, if RBm for a given year's conditions is 4000 and the observed number of reactive bursts (RBo) is 4100 the percentage variation would be 2.5% above RBm. In proposing the reference level and control limits we have applied the methodology used by Ofwat for this purpose, as set out in PR09/38. The key features of this are that Unless demonstrably sub-optimal or atypical, the reference level is based on the best historical performance achieved by the company (the base year) The reference level is calculated from the average of the selected base year and its next lowest neighbour Control limits are drawn at three times sigma above and below the reference level (where sigma is half the difference between the base year and the next best neighbouring year). The graph and table below show the number of observed and modelled reactive bursts for Anglian since The best year in this time series is when the number of reactive bursts was 7.7% below RBm. In accordance with Ofwat methodology, we have therefore chosen as our base year.

94 90 Response to Ofwat's Specified Requirements Risk and reward Observed (RBo) Modelled (RBm) Variation (RBo-RBm) Variation % PR14 Representations on the Draft Determination anglianwater.co.uk The only neighbouring year to is so we have used the performance in this year, at 2.9% below RBm, to calculate the reference level. The average of 7.7 and 2.9 is 5.3, so we propose that our reference level for AMP6 is that the number of observed reactive burst mains is 5.3% below the number calculated by our model after year end (RBm 5.3%). Under Ofwat methodology, the control limits are drawn at three times sigma above and below the reference level, where sigma is calculated as half the difference of the two numbers. Under this calculation sigma is 2.4 (7.7 less 2.9, divided by two) and three sigma is, after rounding, 7.1. The upper control limit is therefore set at 1.8% above the modelled level (RBm + 1.8%) and the lower control limit at 12.4% below (RBm 12.4%)

95 91 The graph below shows our performance in AMP4 and AMP5 based on this proposed methodology. There were two years where performance was above the upper control limit. The most serious was , which was the first year where we experienced an extended period of cold weather after a decade of benign winters. This hard weather found out weak sections of pipe that had survived previous years, thus adding a proportion of backlog to the underlying performance. From this graph we draw reassurance that the reference level and control limits we have proposed for reactive mains bursts are reasonable. We will continue to use the same bursts model during AMP6 to provide an unchanging reference level. However, to improve the accuracy of the model for AMP7 we will continue to develop a version of the model by adding additional data with each successive year. There may also be the opportunity to include longer term weather trends.

96 92 Response to Ofwat's Specified Requirements Risk and reward Table A20a commentary In each case, P10 and P90 are assessed as the penalty collar and reward cap with the following exceptions: Environmental compliance: it is extremely unlikely that we will not deliver our obligations under this measure unless the obligation changes in agreement with the EA. Therefore, 10% of the maximum penalty has been included in this assessment. Serviceability: we have estimated P10 to be the penalty that would apply if one measure were at RED and another at AMBER for the sub-service. Water wholesale Incentive Penalty ( m) - Low case, avg year Reward ( m) - High case, avg year 1 Water supply interruptions averaged over three years Properties at risk of persistent low pressure Water quality complaints Value for money perception - variation from baseline against WASCs (water) Mean Zonal Compliance % of population supplied by single supply system Per Property Consumption (PPC) reduction Leakage - three year average Environmental compliance (water) Serviceability: Water infrastructure Serviceability: Water non-infrastructure Total PR14 Representations on the Draft Determination anglianwater.co.uk

97 93 Wastewater wholesale Incentive Penalty ( m) - Low case, avg year Reward ( m) - High case, avg year 1 Properties flooded internally from sewers - 3 yr avg (reduction) Properties flooded externally from sewers - 3 yr avg (reduction) Value for money perception - variation from baseline against WASCs (sewerage) % of bathing waters attaining excellent status Pollution incidents (category 3) Environmental compliance (sewerage) Serviceability: Sewerage infrastructure Serviceability: Sewerage non-infrastructure Total Retail Incentive Penalty ( m) - Low case, avg year Reward ( m) - High case, avg year 1 Fairness of bills perception - variation from baseline against WASCs Affordability perception - variation from baseline against WASCs Total

98 94 Response to Ofwat's Specified Requirements Risk and reward 7.2 RORE Ofwat's Draft Determination "We expect companies to consider the impact of the interventions and provide further information on the impact on the RoRE range in response to draft determinations." Draft price control determination notice: technical appendix A6 risk and reward page 5 Summary of our Representations We welcome Ofwat's views expressed in the Draft Determination that the approach we have taken to A20 and our scenarios work as being robust. We have updated our table A20, A20a and the Risk Assessment Tool (RAT) to reflect the Draft Determination and the proposed revisions to the ODIs. The table and graph below shows the average annual RoRE range of our base case and demonstrates how the various incentive and risk sharing mechanisms contribute. These values take into consideration the views expressed by Ofwat in Draft Determination, specifically: The removal of the uncertainty mechanisms for Household Growth and the National Environment Programme; The updates to the ODIs as captured in the commentary to section 7.1; and The impact of representations on revenues required and totex thresholds. We have made no other changes to the high and low scenarios in Table A20 and the RAT. PR14 Representations on the Draft Determination anglianwater.co.uk The values in this section are presented as post-tax values to align with the output of Ofwat's Risk Assessment Tool: RoRE Range (post-tax values) Cost performance ODI SIM Financing Overall RoRE variance Reward (post-tax) +1.75% +0.70% +0.21% +0.51% +3.16% Penalty (post-tax) -2.27% -1.51% -0.41% -0.18% -4.38%

99 95 This demonstrates that for the base case our proposed incentives are consistent with Ofwat's expectations. On the basis of the high and low cases from table A20, we have also derived comparisons to the average annual RoRE for the appointed business compared to the these alternative scenarios. The table below presents the impact of the our representations on this range. Average annual RoRE Range - Comparison Base case High case Low case Return on Equity - June Submission (post queries) 5.97% 9.03% 1.78% Return on Equity - Representations 5.66% 8.71% 1.63% Consistent with our approach to the June Submission, we have used Ofwat's Risk Assessment Tool (RAT) to populate a number of the lines in table A20. These are: Return on regulated equity (RORE) impact (high and low cases) Wholesale water - RoRE (base, high and low cases) Wholesale wastewater - RoRE (base, high and low cases) Household Margin (high and low cases) Non-Household Margin (high and low cases) RoRE values for the base case, high and low cases are captured in the table below:

100 96 Response to Ofwat's Specified Requirements Risk and reward Table A20 - Company RoRE values (%) Average Base case 5.7% 5.7% 5.7% 5.7% 5.6% 5.7% High case 7.6% 8.3% 8.5% 8.6% 10.5% 8.7% Low case 2.9% 2.5% 2.1% 2.1% -1.4% 1.6% Overall we are satisfied that this comparison shows the appropriate comparable nature of our RoRE calculations with those from the RAT. PR14 Representations on the Draft Determination anglianwater.co.uk

101 97 8 FD09 service standards Ofwat's Draft Determination "We have considered applying shortfalls equal to the cost of the FD09 project(s) with defined service standards. However, in many cases there is some evidence that the projects and activities have been delivered, but there is a lack of compelling evidence that the service standards specified have been achieved. For the purposes of these draft determinations, we will not applying shortfalls on this issue conditional upon this information being provided as part of companies draft determination representations. We would expect companies to respond to this issue in their representations. If they do not provide adequate evidence to demonstrate achievement of the service standards set out, then they should assume that we will apply a shortfall equal to the costs assumed for the project(s) at FD09 within our final determinations in December 2014." Company specific appendix page 64 Summary of our Representations We present evidence to support delivery of service standards in relation to Enhanced Service Levels - wastewater odour and water resilience. Odour - reduced nuisance for 71,338 properties against a target of 67,000 Resilience - reduced the risk of supply interruption from catastrophic failure of water treatment works for 613,167 consumers against a corrected FD target of 575,445; reduced the risk of interruptions as a result of fluvial flooding of water treatment works for 1,040,007 consumers against a target of 1,023,615. Company specific information in relation to service standards for Anglian Water We provided responses to two queries: rfbp/anh/legacy/001 and ANH/LEG/002, which requested additional information on service level outputs achieved in respect of FD09 tables 2.5.2b and In August 2014 we received a request to provide further evidence that the service standards have been achieved in our response to our Draft Determination. The table summarising the company specific information request is replicated below.

102 98 Response to Ofwat's Specified Requirements FD09 service standards AMP5 Service Standards FD09 Table Ref Table description FD09 Expenditure ( m) Evidence provided of service standard delivery? (y/n) Is the evidence compelling? (y/n) Has a query been raised to request evidence? (y/n) Did query response provide compelling evidence? Has service standard been met? 2.5.2b Sewerage Enhanced Service Levels - Odour 9 N N Y N Unknown Resilience 40.3 N N Y N Not clear Total 49.3 Odour from sewage (Other ESL) In the FD09 Ofwat outlined the requirement for us to deliver the outputs as replicated in the table below, which represented its understanding of our proposals in terms of activity and service standards. The service standard that we proposed in our FBP was to reduce persistent odour nuisance for 67,000 properties. FD09 Table 2.5.2b Other ESL sewerage service outputs Description Activity output Service standard output Completion date Odour New Canwick Odour Control Unit Install an Odour Control Unit Reduce nuisance from odour 31/12/12 Odour New Odour Control Unit Solution Bedford Install an Odour Control Unit Reduce nuisance from odour 31/03/15 PR14 Representations on the Draft Determination anglianwater.co.uk Odour control chemical dosing Chemical septicity control at 89 sites Reduce nuisance from odour 31/03/15 Since our PR09 submission there have been advances in odour nuisance evaluation. In adopting these developments we have re-assessed the catchments to better target investment to reduce odour nuisance. We therefore re-prioritised the catchments targeted as part of the programme of work to ensure that the maximum benefit to customers from the solutions we deliver. As we said in our response to query rfbp/anh/legacy/001, the number of catchments differs from the number in our 2009 FBP as we have have focused on properties at greatest risk to odour nuisance. The number of properties benefiting from reduced nuisance from odour has increased from 67,000 properties to 71,338. Source of odour nuisance Odour from wastewater treatment is generated when organic matter is stored or processed in anaerobic conditions. The absence of oxygen allows bacteria to break down organic compounds, producing hydrogen sulphide, carbon dioxide, water, energy and other

103 99 fermentative side products. These all have the potential to generate malodours. When the matter moves through the treatment process, odorous compounds can be released to the air at a high emission rate, particularly if allowed to vent, untreated to the atmosphere. There are two main sources of odour nuisance arising from wastewater treatment; network septicity and sludge handling, storage and treatment. Network septicity: Hydrogen Sulphide (H 2 S) is generated when sewage is retained in the network for any length of time in anaerobic conditions. Typically sewage will become septic within four hours in a pumped rising main or around 20 hours in a gravity sewer. The potential for network septicity is affected by a number of factors, including dry weather flow, diameter and length of pipe, organic load and temperature. Sludge handling, storage and treatment: Sludge handling and storage or primary / raw sludge can often be a prevalent source of malodours. The causes of these odours include open-topped sludge holding tanks, lack of storage capacity, backlogs of sludge, tankering, temporary over-pumping, poor settlement of sludge and less efficient desludging of Primary Settlement Tanks (PST). If thickening liquors from septic sludge are returned to the treatment process, there is also potential for malodours to be generated. Mitigation of odour nuisance risk As part of our AMP5 odour reduction programme, we considered a number of approaches to either treat the biological condition of the effluent directly or to reduce the probability that odours will be released in the first instance. The following are examples of the general approaches we have considered / adopted: Installing new nitrate dosing plant on the sewerage network Enhancing existing nitrate dosing plan on more complex sewerage networks (particularly if affected by seasonal variations) Enclosing sludge storage and processing activities Up-rating existing or providing new OCUs Optimising flow control to reduce retention times Auto desludging to prevent build-up of septic sludge in PSTs Improved throughput / efficiency of sludge treatment processes to reduce backlogs Chemical dosing to inhibit septicity in sludge return liquors. Activity output We have completed work in 66 catchments / Sewage Treatment Works (STWs) in total, reducing odour nuisance from 71,338 properties. A further six catchments, with an associated property count of 5,841, are currently being investigated and additional solutions may be implemented at these before the end of March 2015.

104 100 Response to Ofwat's Specified Requirements FD09 service standards Summary of solutions Primary solution type STW size band Installation of nitrate dosing Sites Properties Sites Structural alterations Properties New / modified sludge treatment process Sites Properties Enhancement / optimisation of odour control unit Sites Properties Sites Other Properties , , , , , , ,800 Total 43 28, , , , ,421 Total sites 66 Total properties 71,338 Assessing delivery of the service standard output Our standard approach During AMP5 we have developed a more sophisticated and accurate method of calculating / assessing the number of properties at risk of odour nuisance. We were able to apply this method consistently to the catchments on the original PR09 FBP list, as well as any new catchments where a potential for odour nuisance had been established subsequently. The method that we now use is an accepted industry approach, based on a number of factors, to calculate the number of properties likely to be affected by odour at a level that may be sufficient to prompt complaints. The inputs to the calculations include: PR14 Representations on the Draft Determination anglianwater.co.uk The resident population served by the STW The complaint range derived from total emissions, using reference to 'Odour Control - A concise guide' published by the Warren Spring Laboratory in 1980, for the then Department of Environment The size and complexity of the associated STW Consultants Ove Arup & Partners Ltd developed a methodology to assess the odour nuisance risk posed by a STW based on process size, complexity, complaints history and atmospheric dispersion modelling. Using this methodology we can categorise our STWs to identify their odour potential A proportional number of addresses within a modelled complaints range This is derived from postal addresses within the complaints range for a sample of 16 catchments, representing all size / complexity bandings and a spread of urban, suburban and rural locations This approach produces a close correlation with the modelled number of properties included in our PR09 FBP, so is therefore comparable.

105 101 The calculated number of properties that were at risk of odour nuisance for each of the 66 catchments included in our programme of work totalled 71,338 properties. These have all benefited from reduced risk of nuisance from odour as a result of the mitigation work. Supplementary approaches Other approaches that we can use to provide additional evidence of the reduction in nuisance from odour includes the analysis of complaints, which is outlined in more detail below. Complaints Complaints are a potentially good example of direct evidence that the service standard output has been achieved. However, there are a number of drawbacks which means we do not use them as the primary means of assessing the service standard we have delivered. Some of the drawbacks are outlined below. Small numbers: Complaint numbers are often very small Human behaviour: Complaints often reflect past instances of annoyance and a complainant may continue to reference an odour source that ihas been mitigated Contributing factors: Factors such as rate of emission, time of day, prevailing meteorology and psychological responses, will influence the potential for complaints Complaints versus Complainants: It is likely that a single complainant will complain on numerous occasions, while the remaining complaints will arise from a very small number of complainants Time lag: There is often a perceptible lag in the response to improvements that may not manifest in reduced complaints for a period of time Analysis of complaints data While acknowledging the drawbacks with complaints data, nevertheless, at a region-wide level, we have seen a steady decline in the number of written complaints (letters, s and faxes) in AMP5 that we have received that relate to odour. In total there have been 940 complaints in the 53 months between April 2010 to the end of August The spikes in the summer months reflect the increased odour potential of drier, warmer weather.

106 102 Response to Ofwat's Specified Requirements FD09 service standards Written complaints in AMP5 We received 16,900 unwanted telephone contacts between April 2009 and March 2014 that related to odour complaints. These were telephone contacts that were coded in our systems as either 'investigate odour suspected blockage', 'investigate odour' or 'treatment site odour'. We have screened this complaints data to define more precisely to include only those in catchments where we are carrying out mitigation work in AMP5. 4,455 complaints were received between April 2009 and March Split down by year, the data is showing a decreasing trend. PR14 Representations on the Draft Determination anglianwater.co.uk

107 103 Odour complaints for mitigated catchments Complaints: Case studies Below are a couple of examples where we have used complaints data to test the success of a solution. Ingoldmells Ingoldmells STW serves a widely dispersed catchment, including the resort of Skegness. In the summer the population increases significantly as a result of a large number of holiday-makers visiting the area. Because of the location of the works, extensive use of pumping is required to transfer sewage to the STW. The limitations to pumping means that the potential for septicity is high. The solution we implemented here was: Improved control of nitrate dosing in the sewerage network New system of ferrous chloride dosing to sludge liquors Following this intervention, the average septicity levels logged at the inlet works reduced from peaks in excess of 300ppm H 2 S to a baseline average of 20ppm. Septicity in the centrate discharge also reduced from regular peaks of 50ppm to occasional peaks not exceeding 10ppm. By monitoring the odour complaints received relating to Ingoldmells, we were able to see a marked reduction following the implementation of the solution. Although figures for 2014 are not yet complete, the number of complaints is lower, despite including the summer peak time period. The graph below shows the change in complaints levels from 2010 to 2014 to date.

108 104 Response to Ofwat's Specified Requirements FD09 service standards Ingoldmells complaints data Great Billing Great Billing STW is a large site with complex processes and although located in an industrial area there is some residential encroachment close to the site. The main issues at the site which led to a complex range of odours were: PR14 Representations on the Draft Determination anglianwater.co.uk A lack of buffer capacity which led to a backlog of sludge which compromised the treatment processes The inlet works was susceptible to sewage septicity from both the Terminal Pumping Station (TPS) and incoming sewers Co-composting of digested sludge and household garden waste. The solution at Great Billing included the following elements: A covered silo to store imported sludge Relocation of a centrate liquor return to reduce the likelihood of odours Nitrate dosing at Earls Barton TPS to remove septicity Ferrous chloride dosing point just inside the STW boundary but upstream of the inlet works. We began implementing these measures in late The increase in complaints relating to odour at this site in 2012 reflected some variability in prevailing conditions as well as the effects of the issues at the site. The warm dry summer of 2013 would have increased the odour potential from the site, and we would have expected to see an increase in the number of odour related complaints. We actually saw an overall reduction in the number of complaints

109 105 received. The partial figures for 2014 cover the year to mid September but include the high risk, summer period and show that complaints relating to odour have dropped substantially. The graph below shows the change in complaints over a four year period to date. Great Billing complaints data AMP5 Resilience Since AMP4, we have invested to increase the resilience of our supply system to catastrophic water treatment works failure. Our strategy has been to target systems based on the total number of potentially vulnerable customers and to provide these with an alternative source of supply. The PR09 Business Plan Final Determination made an allowance for three such schemes. Details are summarised below (from Table 2.5.3): Resilience enhanced service outputs from FD09 Scheme Activity Output Service Standard Output Completion Date Grafham WTW Improved connectivity between Hannington and Grafham 280,016 properties (613,592 consumers) benefiting from a reduction in risk of less of supply from 1/1000 to

110 106 Response to Ofwat's Specified Requirements FD09 service standards Scheme Activity Output Service Standard Output Completion Date West Pinchbeck WTW Extension to Covenham to Boston supply-demand scheme 65,902 properties (100,713 consumers) benefiting from a reduction in risk of less of supply from 1/1000 to Barrow WTW Network reinforcement 47,841 properties (62,281 consumers) benefiting from a reduction in risk of less of supply from 1/1000 to The number of customers benefiting from each scheme was estimated through failure-consequence modelling. In this, a model of our supply system was used to estimate the supply-demand deficit, in Ml/d, that would result from catastrophic failure of a water treatment works. Based on generic per property consumption estimates, this deficit was converted to an equivalent population (consumers) affected. In respect of Table the following is noted: 1. At PR09 the Ml/d deficits were converted to population equivalent using a regional consumption estimate of 257 l/person/day. This was estimated from PZ level distribution input (DI) and population estimates. We have used the same approach for determining the number of consumers who have benefitted from the schemes we have delivered and used an updated consumption estimate of 243 l/person/day, based on DI and population data from The property estimates are the total number of household properties in the PZs affected by the resilience event while the consumer estimates are the number of people who will experience a loss of supply. The two numbers are not directly comparable since it is possible that a resilience event will only affect a small proportion of the total properties in a PZ, with the remainder receiving supplies from an alternative source. We have calculated our service standard outputs on consumer numbers. PR14 Representations on the Draft Determination anglianwater.co.uk 3. The activity and service standard outputs for Grafham were incorrectly matched in Ofwat's Final Determination. Our PR09 Final Business Plan (Part B6, Table 6.10) showed a resilience deficit for Grafham of 158 Ml/d, with an equivalent population exposed of 613,592. We proposed (para ) to provide a partial solution through the Grafham to Hannington main, which we said would meet a resilience deficit of 106 Ml/d. (Other schemes proposed in our Draft Business Plan which would have addressed the remaining deficit had been rejected by Ofwat on cost benefit grounds). Using our PR09 methodology, this would have provided a service standard output of 412,451 consumers. That the Grafham-Hannington scheme would not reduce the resilience risk for all the population exposed was confirmed in table 6.13, which showed a remaining deficit of 52 Ml/d after completion of the scheme. Our final determination confirmed the Grafham-Hannington scheme as the only activity output but failed to adjust the service standard output accordingly for the activities which had been deemed non cost beneficial and excluded from the FD. Relevant extracts from our PR09 FBP are shown in the box below:

111 107 Extract from Anglian Water PR09 FBP Grafham WTW is located in Ruthamford and serves a population of 829,000. We estimate that approximately 614,000 of these would be affected by a major outage. The resilience solution involves improving the connectivity between the systems supplied by the Wing WTW and the Grafham WTW. This will be achieved by installing a 106 Ml/d gravity main between Hannington reservoir and the Grafham WTW at a cost of 55.5m Table 6.13 'Impacts of our proposed resilience investment' demonstrates how the proposed investments will benefit our customers. From this, we will deliver resilience for the majority of our vulnerable customers in the 50,000 population category. However, as we have removed the non-infrastructure elements of the Grafham solution, a residual vulnerable population will exist in this system. In AMP6, this is likely to reduce through the delivery of additional supply-demand schemes, including both the non-infrastructure elements referred to above and any related infrastructure schemes. Table 6.13 Impacts of our proposed resilience investments Scheme Pre-investment Post-investment Deficit (Ml/d) No. Households exposed Deficit (Ml/d) No. Households exposed Barrow WTW resilience , Grafham WTW resilience , ,562-92,125 Newton WTW resilience , West Pinchbeck WTW resilience , Range of possible deficits in the Grafham supply zone Ml/d depending on demand at time of incident and the volume of additional resource that is developed in Ruthamford in AMP6 The Barrow and West Pinchbeck schemes have been delivered. The Grafham scheme will be delivered early in AMP6 as planned at PR09. The service standards delivered, or to be delivered, by these schemes are shown in the table below.

112 108 Response to Ofwat's Specified Requirements FD09 service standards Summary of resilience service standards delivered against FD09 targets Scheme Grafham West Pinchbeck Barrow TOTAL Target service standard (consumers) 412, ,713 62, ,445 Service standard delivered (consumers) 436, ,111 65, ,167 Grafham WTW Resilience The failure consequence modelling for Grafham WTW showed a resilience deficit of up to 158 Ml/d. Of this, 106 Ml/d will be avoided through improved connectivity between Hannington and Grafham. Details of the connectivity scheme are given below: PR14 Representations on the Draft Determination anglianwater.co.uk The process followed to ensure that the Grafham scheme will deliver the expected benefits may be summarised as follows:

113 109 Validation of the failure-consequence modelling through confirmation of the demand met from Grafham WTW. This involved collation and review of works output and flow data in the Ruthamford system using sources normally used for regulatory reporting, such as ARTS and SWORPS Where the validation exercise identified critical uncertainties, such as the capacity of key infrastructure under resilience conditions, field trials were completed. These involved inspection and testing local infrastructure as well as testing the performance of the system at a larger scale The development and testing of alternative schemes for meeting the (validated) demand from Grafham WTW in the event that the works suffered a catastrophic failure. This step was completed in support of a risk and value process and involved detailed hydraulic modelling to simulate the effect of a sudden outage and determine system requirements for avoiding any disruption to supply. To improve the accuracy of the modelling, the hydraulic models were calibrated using data from the field trials, and From the above, confirmation of the infrastructure and non-infrastructure required to deliver the expected outputs. This modelling confirms that our scheme will reduce the resilience deficit by an amount equivalent to 106 Ml/d, providing, at 243 Ml/person/day, benefit to 436,214 consumers, compared with equivalent service standards in our FBP for this scheme of 412,451. Following completion of the scheme early in AMP6 trials will be completed to validate the modelling and confirm that the completed scheme has delivered the required outputs. As stated in our PR09 FBP, completion of Grafham-Hannington scheme provides only partial solution to the Grafham resilience deficit. The remaining resilience deficit will be eliminated through the delivery of additional supply-demand schemes in the south Ruthamford system. In AMP6 the development of extra capacity at Clapham WTW and Pulloxhill WTW has provided 7.4 Ml/d at peak output. West Pinchbeck WTW Resilience The failure consequence modelling for West Pinchbeck WTW showed a resilience deficit of 25.9 Ml/d in the Boston supply system. In the event of a catastrophic failure at the works, the scheme which has been delivered will provide 15.6 Ml/d of supplies from Covenham WTW and 11.4 Ml/d of supplies from Etton WTW and Wilsthorpe WTW. With these, there will be no disruption to supply for customers. Details are given below:

114 110 Response to Ofwat's Specified Requirements FD09 service standards The process used to develop this solution and to confirm that it has delivered the expected outputs may be summarised as follows: PR14 Representations on the Draft Determination anglianwater.co.uk Validation of the supply-demand and resilience needs for the Boston system using works output and flow data normally used for regulatory reporting. This included data from AW Regional Telemetry System (ARTS), and Source Works Output Reporting System (SWORPS) The development and testing of alternative schemes for meeting the combined resilience and supply-demand needs. This involved detailed hydraulic modelling in support of a risk and value process and resulted in the selection and design of the preferred option, and Following delivery and commissioning of the preferred solution, capacity testing using field trials. This modelling confirms that our scheme has eliminated the resilience deficit of 27 Ml/d. At 243 Ml/person/day, this provides benefit to 111,111 consumers, compared with a service standard of 100,713. Barrow WTW Resilience The failure consequence modelling for Barrow WTW showed a resilience deficit of 16 Ml/d in the Barrow supply system and adjacent parts of the Scunthorpe supply system. In the event of a catastrophic failure of Barrow WTW, the scheme which has been delivered will: Support the Barrow supply system via a new main from Elsham WTW, and Support the Scunthorpe supply system through opening of a closed valve cross-connection between the Elsham and Scunthorpe systems. Details are given below:

115 111 The process used to develop this solution and to confirm that it has delivered the expected outputs may be summarised as follows: Validation of the resilience needs using works output and flow data normally used for regulatory reporting The development and testing of alternative schemes for meeting those needs. This involved hydraulic modelling in support of a risk and value process and resulted in the selection and design of the preferred option, and Following delivery of the preferred solution, capacity testing using field trials. This is planned for completion in the near future. This modelling confirms that our scheme has eliminated the resilience deficit of 16 Ml/d. At 243 Ml/person/day, this provides benefit to 65,842 consumers, compared to a service standard of 62,281. Fluvial Flooding Resilience Schemes Fluvial flooding resilience schemes for 20 water treatment works were included in our PR09 final determination. The schemes will protect customers from a loss of supply under conditions where the works output would previously have been lost due to fluvial flooding. These schemes are due for completion by March 2015; 14 have been delivered to date, with the remainder due for completion before the end of December. The service standards proposed at PR09 were as follows: 1,023,615 customers to benefit from an increase in the security of their water supply Protection against floods of less than 1:100 return period In our June 2014 business plan submission, we confirmed the delivery of the activity output would be achieved and confirmed our intention to meet the service standard outputs. Here we are able to provide further evidence that the service standard outputs will be achieved by reviewing the number of customers that will be protected and providing evidence that the flood design levels have been adhered to.

116 112 Response to Ofwat's Specified Requirements FD09 service standards We have assessed the number of customers to benefit using the same methodology that generated the figures reported in our PR09 business plan. This is carried out using the average modelled volume of deficit during a flooding event and the regional average distribution input per customer. For the latter we have used the same updated figure as we did for calculating the service standard at Barrow, West Pinchbeck and Barrow (243 Ml/person/day). We have also provided information to compare the 1:100 flood level (including climate change) and the level of flood protection provided at each site. The additional level of protection varies depending on confidence in the flood level provided and site specific technical issues. The table below shows the service standard delivered by each scheme in the programme. It confirms a total service standard of 1,040,007 customers to benefit from an increase in the security of their water supply. Service standards met by our PR09 fluvial flooding resilience programme Works Name Target service standard - customers benefiting Actual service standard - customers benefiting 1:100 level (maod) Protection level (maod) Barrow Heath 35,476 37, Barrow 62,281 65, Cadney Bridge - River Intake 169, , Covenham 53,800 57, Harborough 12,680 13, Healing 12,680 13, Hillington - Greensand 54,776 58, PR14 Representations on the Draft Determination anglianwater.co.uk Littlecoates Maltby Le Marsh Marham - River Nar Intake Mumby Newton On Trent Petches Bridge Raithby Saltersford Tetney 53,800 12,300 31,876 39, ,647 33,101 45,624 40,900 53,800 57,035 13,034 33,778 41, ,548 35,076 48,346 43,341 57,

117 113 Works Name Target service standard - customers benefiting Actual service standard - customers benefiting 1:100 level (maod) Protection level (maod) Weelsby 53,800 57, West Pinchbeck (1) 100, , / / 2.80 Whitton 23,689 25, Wixoe (2) 18,573 0 N/a N/a Cadney Reservoir 0 0 (3) Total population 1,023,615 1,040, There are two flood levels at this site due to it being spread over two levels of the Fenland drainage system 2. Wixoe removed as no longer a source 3. Work identified at this site contributes to ensuring benefit for customers is realised from Cadney Bridge work

118 114 Response to Ofwat's Specified Requirements FD09 service standards PR14 Representations on the Draft Determination anglianwater.co.uk

119 115 9 Serviceability Ofwat's Draft Determination "For the purposes of the draft determination we have assumed no intervention. This is conditional upon the performance in being improved to a position such that it could be considered as stable. We may consider a shortfall adjustment if this is not achieved. We would expect the company to demonstrate its latest performance as part of its representations and in advance of the final determination. Burst Mains: The company has had two breaches of the upper control limit in and However, performance has recently recovered to just below the upper control limit. The company has provided evidence to show that adverse weather impacted on performance in those years. Considering the weather impacts, only remains above the upper control limit. We require the company to demonstrate stable serviceability in , if this is not achieved, we may consider a shortfall adjustment. Flooding other causes: The company has had three breaches of the upper control limit between and The company has provided evidence to show severe weather impact on performance in and performance recovered to near the reference level. However, we require the company to demonstrate stable serviceability in If this is not achieved, we may consider a shortfall adjustment." Company specific appendix page Summary of our Representations We have provided an update of our performance to date in , together with a forecast for the year. Both of these show that our performance is stable. Water Infrastructure: Burst Mains In to the end of July we have recorded 1,230 verified burst mains. This translates into a forecast of 4,616 at the end of the report year. Our forecast is based on trends in previous years and as shown in the graph below it takes into account higher winter burst rates.

120 116 Response to Ofwat's Specified Requirements Serviceability When we incorporate this with our previous performance this shows a trend continuing to move back towards the reference level. PR14 Representations on the Draft Determination anglianwater.co.uk As a result of this, we can confirm that our performance is stable.

121 117 Sewerage Infrastructure: Flooding Other Causes In to the middle of September we have recorded 61 verified flooding incidents due to other causes. This translates into a forecast of 133 at the end of the report year. When we incorporate this with our previous performance this shows a trend moving below the reference level. As a result of this, we can confirm that our performance is stable.

122 118 Response to Ofwat's Specified Requirements Serviceability PR14 Representations on the Draft Determination anglianwater.co.uk

123 Customer engagement Introduction We have conducted further customer engagement in response to Ofwat's Draft determination, and this section includes the results of that engagement. We have engaged with customers on: The form and duration of the non-household retail control. See section 10.1 for more detail. Use of financeability tools and profiling of customers' bills. See section 10.2 for more detail. In-period adjustments due to company performance. See section 10.3 for more detail Non-household retail control Ofwat's Draft Determination In any representations on our draft determinations we invite companies to consider whether they would prefer for us to: set their non-household retail price control for five years, as set out in our final methodology statement; or change the form of the control in some way that would allow them greater time to consider and address any issues we have highlighted for example, through a shorter control (for example, one to two years) or a form of reopener to the price control within the five years. We expect companies to provide evidence of engagement with their customers about the possibility of reopening their non-household retail price controls within the next five years in their representations to us. We anticipate that our invitation may result in different requests from individual companies across the sector. We are not expecting a common, sector-wide solution to emerge from this feedback. As a minimum, we would expect to see evidence that: each company has engaged with its customer challenge group (and better still, local non-household customer groups as well), on potential options to change the length or form of the control; and there is customer support for the structure of its proposed average revenue controls and associated default tariffs; and revenue levels appear acceptable to those customers. Draft price control determination notice: technical appendix A5 non-household retail page 19

124 120 Response to Ofwat's Specified Requirements Customer engagement Summary of our Representations We have engaged with the CEF and Independent Advisory Panels (which both include business customers and non-household customer groups) about our non household retail tariffs. They support our recommendation to have the option to re-open the non household default tariff for the charging year , and that this should apply to the industry as a whole, to allow retail charges to be re-set as appropriate based on new cost data. Background The default tariffs set out in our Business Plan submission replicated our current non-household tariff structure covering water, wastewater (foul and surface) and trade effluent. No new structures were proposed in relation to default tariffs. The calculation of our default tariffs was based on the allocation of retail costs between household and non-household customers in line with Ofwat s regulatory accounting guidelines, and the allocation of non-household retail costs across services and customer bands in line with Ofwat s guidance issued in April this year. We are confident that our allocation is robust to the extent allowed in taking this approach. The PR14 process has been the first time that companies have been required to undertake a detailed retail cost allocation exercise. As Ofwat noted in its consultation document on charges for , capturing new data can take a great deal of time to ensure a reasonable level of robustness and consistency. PR14 Representations on the Draft Determination anglianwater.co.uk As part of our plans to reflect the Level Playing Field requirements for the new contestable market and move to functional separation of the Non-Household Retail operation (Anglian Water Business, or AWB ) from the Wholesale/Household Retail operation, new financial management and billing systems are being installed in AWB during These new systems will allow for a greater degree of accuracy in the allocation of retail costs between customer groups, with more direct attribution through activity based costing, providing better information in order to re-set default tariffs prior to market opening in As a result, we would expect that whilst: (a) the total level of non-household retail costs is unlikely to change (based on costs); and (b) we expect to maintain a straight line allocation of the residual margin (being the element to remunerate risk and encourage efficient entry into the market) as a constant percentage of the default tariff per customer band,the future allocation of costs between tariffs might vary with better data allowing direct attribution of costs. Engagement on this issue As agreed with the Chair of the CEF, we therefore engaged with the forum and the Independent Advisory Panels on the issue of changing the form of the non-household retail control.

125 121 Our high level principle is not to reopen price controls unless absolutely necessary and, where it is necessary, to reopen as few times as possible. The options we proposed to the CEF and Independent Advisory Panels therefore were as follows: 1. Not to reopen the default tariff control at all; 2. To reopen the default tariff control once for charges; or 3. To reopen the default tariff controls for charges and each year thereafter until the start of the next control period in We provided a briefing note to all members to explain the background to the topic, and reasons for changing (or not) the length of the control before Experts were also available on the day to help with any queries. The briefing note explained that: 1. Not to reopen the default tariff control could mean that some customers pay more than is justified by the updated cost data because the company is unable to make changes within the existing controls as more accurate cost data becomes available. We did not recommend this option. 2. Although logic follows that if more accurate cost data continues to become available over time then the default tariff controls should be reopened each year, the degree to which data will become more accurate over time is expected to diminish and in line with our principal to reopen controls as few times as possible, we did not recommend this option. 3. As a result, we believe the best option is to review and allow changes to the tariff once for charges, based on new cost data. The note provided an illustration of the typical split of total bills between retail and wholesale charges for customers on different tariff bands. On the day, the panels focusing on this topic in depth were the Economic Development panel and the Hartlepool panel, whose membership includes a range of businesses such as Travis Perkins, Lark Energy, Lincolnshire County Council and organisations representing businesses such as Hartlepool Business Forum and Consumer Council for Water. We asked members for their views on this issue and which option they preferred. The Economic panel did not see any downsides to re-opening the non-household retail tariffs in 2017 alongside the introduction of the retail market. There was discussion about whether this approach should perhaps be applied consistently across the industry and the group agreed that this should be the case. The panel agreed that they would be willing to support the company s approach to non-household retail tariffs going forward. The Hartlepool panel felt that a single review for was most appropriate. It wanted to see consistency in the approach across the industry to avoid companies reacting to changes made by competitor companies. It could see no real down side from a customer perspective. It felt the one off change would enable market stability and ensure a better understanding of costs and implementation of a competitive market. Overall,the CEF concluded:

126 122 Response to Ofwat's Specified Requirements Customer engagement "We support Anglian Water s recommendation that tariffs should be reopened once in 2017/18 alongside the introduction of the retail market. We believe this approach should be introduced for all water companies to ensure consistency and market stability." CEF Representation p4. Conclusion On this basis, we would propose that the non-household retail control is re-opened once for the industry as a whole, to allow retail charges to be re-set as appropriate based on new cost data for the charging year PR14 Representations on the Draft Determination anglianwater.co.uk

127 Financeability tools Ofwat's Draft Determination We are also asking the following companies to re-engage with their customers where they have provided satisfactory evidence on the benefits to customers from the use of the financeability tools but have not provided satisfactory evidence on engagement with customers: Anglian Water, Severn Trent Water, Southern Water, United Utilities and Wessex Water. Draft price control determination notice: technical appendix A7 financeability and affordability page 10 Summary of our Representations We have engaged with the CEF, Independent Advisory Panels and directly with customers through a survey on the implications of different approaches to financeability tools on average household bills. The feedback from this has been clear that customers prefer smooth bill profiles and do not like 'price hikes' even if this means foregoing short term reductions in bills. 72% of customers surveyed preferred our proposed approach that achieves a smoother profile of average household bills than the Draft Determination and avoids a steep increase in Background One of the biggest changes in the DD to how we finance our plan, has been to how capital efficiencies made over the last five years are dealt with. Our plan was to phase the return of these efficiencies to customers over 2015 to 2020 helping to keep a smooth profile for customers bills. Instead, Ofwat has asked us to return all the efficiencies to customers in the first year This has three main implications: For our investors - The change reduces revenues collected from customers in 2015 to 2020 by 82m and at the same time it increases Regulatory Capital Value by 98m. Overall this adjustment has no absolute impact on returns to our investors in net present value term customers in this period pay less but customers in later periods will pay more. For the company. Delaying revenues means that we would need to borrow more in the period. It also means that some important financial ratios deteriorate and increases the risk that we will be seen as less creditworthy and potentially suffer a ratings downgrade. As a result the costs of debt finance would rise which would impact customer bills in future. For customers - Customers bills will see a bigger drop in than we had planned and will continue to fall but this could mean that they would increase more steeply from 2020 onwards. Again, there is no absolute impact on customers bills in net present value terms but what has changed is the profile of customer bills and the balance between how costs are recovered between current and future customers. However, if the company is impacted by a rating downgrade this would increase debt finance costs

128 124 Response to Ofwat's Specified Requirements Customer engagement which would impact customer bills in the future and customers would overall pay higher charges. As agreed with the Chair of the CEF, to re-engage customers on this issue we discussed it with the CEF and Independent Advisory Panels at the joint meeting on 17 September and conducted a customer survey. Engagement with CEF and panels We provided a briefing note to all members to set out the background to the topic and potential implications of the DD for customers, the company and investors. We explained that customers have so far told us that they prefer a smooth bill profile and to this end our business plan sought to fairly balance customer bills between current and future customers, as well as trying to minimise volatility of bills and the potential for a sharp increase in bills at the end of a regulatory period. The Draft Determination has both lower bills overall in and a bigger reduction in the first year, reflecting the change above. This potentially increases bill volatility due to the likelihood of a bill spike at the start of the next regulatory period. Reversing the change to the treatment of efficiencies is not the only way of dealing with the issue. We could also adjust depreciation rates, for example. Before deciding exactly how we would achieve this we plan to engage with customers on two key issues: how we should balance recovery of costs between current and future customers the sort of profile that would be acceptable for bills in the current period given other changes to the plan. While we have previously discussed these issues with customers, the revisions in the Draft Determination mean that it is important to do so again, particularly given the high level of bill reductions in our plan. PR14 Representations on the Draft Determination anglianwater.co.uk We included a graph to shows Household (HH) bills over 2014 to 2020 for our June plan compared to the Ofwat Draft Determination (DD). We illustrated with the green line an example of the kind of bill profile that we think would find a good balance between Ofwat s expectations for an early bill reduction and our requirements to achieve a financeable plan that has more balance of bills between current and future customers. Illustrative graph for CEF We asked members for their view on this issue including: The balance of recovering costs between current and future customers

129 125 How bill reductions should be profiled across 2015 to 2020 Advice on the best way to explore this with customers in the forthcoming survey. The Customer and Communities panel felt Ofwat had probably made up its mind on this and the issue would be non-negotiable. It felt that customers would not notice lower bills too much now but when they increase at the end of the period this would be more noticeable. It believes an increase in bills later on will make things more difficult and customers want a smoother flight line. A further conversation is needed with customers about changes in bills. It recommends that when the issue is discussed with customers in the survey, indicative bill impacts are given in figures and impact of future bills is included. It also noted that media interest and customer perception etc. of increase in later years was likely. The Economic panel s main concern when discussing financeability was how the possible uplift in bills in 2021 would be communicated to customers. There was strong feeling among attendees that this bill impact would need to be made clear in advance in order for customers to make an informed decision. Another element discussed was what possible impact changing the financial mechanisms in the plan may have. There was concern about the possible negative impact on the company s credit rating; with agencies already putting Thames Water on a downgrade warning. The panel would support the company s suggested approach to making a representation on the financeability element of the plan, but would like to see more information on possible AMP7 bill impacts. The Environment panel was not originally asked to review this area but would like to make the comment that there must be a long term (multi decade) framework for investment and planning to achieve water quality and habitat objectives. The AMP process should have limited influence on this. The Hartlepool panel felt that this issue had less of an impact from its perspective. It was against burdening future generations with investment that should take place now rather than later. It felt that customers want value for money now with stable and predictable future bills to enable budgeting, prevent shocks and protect against future increases. Customers would forego a dip this year if it means no future price hikes. The panel felt that 5 years is long term for many domestic customers and many small commercial companies may not plan for longer than 3 years which makes it difficult to gauge a view on altruism for the future and financing. So, it believes Anglian Water needs to ask customers what they want and the panel recommend that when Anglian Water consults customers it shows figures rather than graphs. Overall, the CEF concluded: "We consistently hear from customers that they prefer stable bills and we are concerned that the changes to financeability in the DD and reduction in bills over 2015 to 2020 will lead to an uplift in bills after We support Anglian Water s recommendation for a smoother bill profile from in order to ensure that bills remain more stable over the long term." CEF Representation p4. Customer survey Details of the survey methodology can be found in section 5.

130 126 Response to Ofwat's Specified Requirements Customer engagement The survey explored customers' views on two options for average household bills depending on how we finance the plan. Option 1 was based on the Draft Determination and Option 2 was a proposed new approach to financing the plan which we believe is more sustainable and achieves a better balance between current and future customer bills with more stable bills over 2015 to 2020 and beyond. Showcard for customer research 72% of those surveyed preferred option 2. The most common reasons given for preferring option 2 to option 1 were that it is easier to budget, no sudden increases, less volatile. PR14 Representations on the Draft Determination anglianwater.co.uk We have discussed the implications of these results with the CEF. It said "We are confident that the questions were phrased in such a way as to not introduce bias or present options as preferred or otherwise by the company. The response to the question asking which option customers preferred triangulates well with responses to other questions." CEF Representation p12 Conclusion As a result of our engagement with customers and CEF, we propose changing the RCV run-off rate to ensure we can achieve the preferred profile of average household bills in the period and beyond. Further details on this are found in section 3 Customer bill profile and financeability.

131 In-period bill variation Ofwat's Draft Determination "We also require the company to demonstrate that its customers support the idea of their bills varying during the period to reflect the company s performance. If we accept the company s proposed approach to its leakage performance commitment and incentives we will formally consult on a licence change before we make our final determination." Company specific appendix page 33 Summary of our Representations The CEF and panels believe it is very important for us to continue to deliver the proposed leakage reduction of 172 Ml/d and they support the proposed funding through an ODI mechanism with reward/penalties applying in period. When we surveyed customers, there was no difference in levels of support for funding improvements through ODI with in period adjustments compared to the current approach and significant (80%) support for the new approach if the alternative was that the improvements would not be introduced. 71% are prepared to see a fluctuation in bills of up to 5.00, which is more than the maximum increase of 4.80 in a single year ( ) associated with our leakage proposals. Given the support from the CEF and customers, we believe the Final Determination should reflect our proposed leakage ODI and in-period adjustments. We have provided an updated draft licence modification. Background In our updated plan submitted in June, following feedback from Ofwat, we removed around half our planned leakage expenditure relating to the customer driven leakage reduction beyond the Sustainable Economic Level of Leakage. We proposed instead to fund this using a combination of an Outcome Delivery Incentive (ODI) and the totex incentive mechanism. Unlike the other ODIs, where the incentives/penalties apply to customer bills from 2020, performance against the leakage ODI will affect bills from April 2017 (i.e. performance for is assessed in , and the reward or penalty is applied to charges in and subsequent years.). We discussed this change with the CEF before making our June submission and they would have preferred the original approach set out in our December 2013 plan but, '...given the very direct and clear message from customers that the company should be investing in more than the economic level of leakage... have accepted the new proposals based on an ODI mechanism. p3&4. CEF supplementary report June 2014.

132 128 Response to Ofwat's Specified Requirements Customer engagement Engagement with CEF and panels As agreed with the Chair of the CEF, we engaged with the CEF and Independent Advisory Panels on this topic at our joint meeting on 17 September and conducted a customer survey. We provided a briefing note to all members to explain the background to the topic including an indication of likely bill impacts and explanation of other factors that influence bills. Year Performance range Ml/d 211 to to to to to Penalty / Reward earned -7.8m to +0.5m -7.8m to +2.5m -7.8m to +5.1m -7.8m to +7.6m -7.8m to m - - Impact on bills to to to to to We explained that our preferred approach has always been to reduce leakage to the ambitious target funded through the plan (as submitted to Ofwat in December). However, we see this as the next best option, which enables us to deliver what customers want but they only pay for it once it has been achieved. We believe that a small degree of bill variability is preferable to ignoring customers clear preference to reduce leakage significantly. We asked the CEF their views on our proposal to use in period reward/penalties mechanisms. PR14 Representations on the Draft Determination anglianwater.co.uk The Customer and Communities panel felt that leakage could be viewed as the day job and Anglian Water should be working towards this anyway. It noted that surveys showed leakage reduction as a high priority for customers. Overall, it felt customers would prefer a more even spread of bill impacts and so it preferred the original plan. It felt that it was important to educate customers to take their part in water efficiency measures in the home which can be greater than leakage reduction. It would like to have a water element to money saving advice and see more PR in water saving and the impact customers can have. The Environment panel felt it was very important that Anglian Water continued to deliver its ambitious leakage programme and it is not so concerned with the mechanism by which this is achieved. It was not opposed to the approach of using ODI to fund the additional investment beyond the Sustainable Economic Level of Leakage and thought that the level of instability in bills would not be sufficient to cause customers concern. It reinforced that it was essential that this was part of a coherent package of response to growth, climate change and water efficiency. It noted that media coverage on leaks and any future drought may shift the external context in terms of public perception. The Economic panel did not have enough time to fully explore the leakage outcome delivery inventive. However, local authority representatives noted that they were encouraged with the current position of the company with regards to leakage and their approach for reducing this further in the future.

133 129 Overall, the CEF concluded: "We feel it is very important for Anglian Water to continue to deliver its ambitious leakage programme in response to customers priorities and we support the proposals of funding through an ODI mechanism with reward/penalties applying in period." CEF Representation p3 Customer survey The methodology for the survey is summarised in section 5. If the proposed new approach was the only way in which improvements could be introduced, 80% said they thought the new proposed approach is very/fairly acceptable. When we surveyed customers, there was no difference in levels of support for funding improvements through ODI with in period adjustments compared to the current approach and significant (80%) support for the new approach if the alternative was that the improvements would not be introduced. 71% are prepared to see a fluctuation in bills of up to 5.00, which is more than the maximum increase of 4.80 in a single year ( ) associated with our leakage proposals. From this, we conclude that the majority of customers support our proposals to fund our ambitious leakage programme through an ODI and the totex incentive mechanism. We have explored the implications of these results with the CEF and their view was "Given the results of the customer survey and the importance of the leakage programme, we continue to support the new proposals of funding through an ODI mechanism with in-period rewards/penalties." CEF Representation p9. Licence Modification Ofwat requested that we provide a draft licence modification as part of our Representations. We believe there are two possible ways of approaching the modification, as the concept of in-period rewards and penalties may become a regular feature of future price reviews. The first approach is to take a general approach in order to enable it to apply to other, future, rewards and penalties in-period, and the second approach is to draft the modification to specifically allow only the leakage ODI which is currently proposed to operate. These two options are set out below. Option 1 - general approach 9.4A. Additional Adjustment 9.4A.1 In respect of each of the subsequent Charging Years (not including the first or second Charging Years for the period of the Price Control) the Water Services Regulation Authority shall determine whether the Price Control in respect of the Appointed Business s Wholesale Activities in respect of its water services or its sewerage services or both, as appropriate, should be adjusted by an Additional Factor, which may be a positive number or a negative number, and, if so, the amount of such Additional Factor in accordance with paragraph 9.4A.3 below.

134 130 Response to Ofwat's Specified Requirements Customer engagement 9.4A.2 In order that the Water Services Regulation Authority can make the determination in respect of each such Charging Year in accordance with paragraph 9.4A.1 above, the Appointee shall provide to the Water Services Regulation Authority no later than 16 September (or such other date as the Water Services Regulation Authority may notify to the Appointee) in the Prior Year: 1. such details as the Water Services Regulation Authority may reasonably require as to the Appointed Business s performance in respect of the Wholesale Activities to which the Additional Factor is related in the year preceding the Prior Year; and 2. a report verifying the accuracy of the details provided under sub-paragraph (1) commissioned by the Appointee from a company that is independent of the Appointee and has previously been approved in writing by the Water Services Regulation Authority (such approval not to be unreasonably withheld or delayed). 9.4A.3 In determining the Additional Factor pursuant to paragraph 9.4A.1 above, the Water Services Regulation Authority shall consider whether or not and, in either case, how the Appointee s actual performance in the year preceding the Prior Year measures against the base performance targets set by the Water Services Regulation Authority and the enhanced performance targets proposed by the Appointee in the Appointed Business s business plan submitted to the Water Services Regulation Authority prior to the Determination of the relevant Price Control at the last Periodic Review provided that: 1. If the Appointee s actual performance in the year preceding the Prior Year meets the enhanced performance targets proposed by the Appointee, the Additional Factor shall be such positive number as the Water Services Regulation Authority shall determine having regard to the proposals made by the Appointed Business in the business plan submitted to the Water Services Regulation Authority prior to the Determination of the relevant Price Control at the last Periodic Review; 2. If the Appointee s actual performance in the year preceding the Prior Year meets the base performance targets set by the Water Services Regulation Authority but fails to meet the enhanced performance proposed by the Appointee, the Additional Factor shall be zero; and PR14 Representations on the Draft Determination anglianwater.co.uk 3. If the Appointee s actual performance in the year preceding the Prior Year fails to meet the base performance targets set by the Water Services Regulation Authority, the Additional Factor shall be such negative number as the Water Services Regulation Authority shall determine having regard to the proposals made by the Appointed Business in the business plan submitted to the Water Services Regulation Authority prior to the Determination of the relevant Price Control at the last Periodic Review. 9.4A.4 The Water Services Regulation Authority shall make its determination under paragraph 9.4A.1 above no later than 15 December in the Prior Year. Option 2 - specific approach 9.4A. Additional Adjustment 9.4A.1 In respect of each of the subsequent Charging Years (not including the first or second Charging Years for the period of the Price Control) the Water Services Regulation Authority shall determine whether the Price Control in respect of the Appointed Business s Wholesale

135 131 Activities in respect of water services should be adjusted by an Additional Factor, which may be a positive number or a negative number, and, if so, the amount of such Additional Factor in accordance with paragraph 9.4A.3 below. 9.4A.2 In order that the Water Services Regulation Authority can make the determination in respect of each such Charging Year in accordance with paragraph 9.4A.1 above, the Appointee shall no later than 16 September (or such later date as the Water Services Regulation Authority may notify to the Appointee) in the Prior Year: 1. publish details of its performance against its base and enhanced leakage targets in the Charging Year preceding the Prior Year; 2. publish details of its average performance against its base and enhanced leakage targets for the last three Charging Years (namely, the Prior Year and the two previous Charging Years); and 3. provide to the Water Services Regulation Authority a report verifying the accuracy of the details published under sub-paragraphs (1) and (2) above commissioned by the Appointee from a company that is independent of the Appointee and has previously been approved in writing by the Water Services Regulation Authority (such approval not to be unreasonably withheld or delayed). 9.4A.3 In determining the Additional Factor pursuant to paragraph 9.4A.1 above, the Water Services Regulation Authority shall consider whether or not and, in either case, how the Appointee s actual performance measures against the base and enhanced leakage performance proposed by the Appointee in the Appointed Business s business plan which was submitted to the Water Services Regulation Authority prior to the Determination of the Price Control at the last Periodic Review provided that: 1. If the Appointee s actual performance in the year preceding the Prior Year meets the enhanced performance targets proposed by the Appointee, the Additional Factor shall be such positive number as the Water Services Regulation Authority shall determine having regard to the proposals made by the Appointed Business in the business plan submitted to the Water Services Regulation Authority prior to the Determination of the Price Control at the last Periodic Review; 2. If the Appointee s actual performance in the year preceding the Prior Year meets the base performance targets set by the Water Services Regulation Authority but fails to meet the enhanced performance proposed by the Appointee, the Additional Factor shall be zero; and 3. If the Appointee s actual performance in the year preceding the Prior Year fails to meet the base performance targets set by the Water Services Regulation Authority, the Additional Factor shall be such negative number as the Water Services Regulation Authority shall determine having regard to the proposals made by the Appointed Business in the business plan submitted to the Water Services Regulation Authority prior to the Determination of the relevant Price Control at the last Periodic Review. 9.4A.4 The Water Services Regulation Authority shall make its determination under paragraph 9.4A.1 above no later than 15 December in the Prior Year. In both cases, there would be a need to make a minor amendment to paragraph 9.4(1) as follows:

136 132 Response to Ofwat's Specified Requirements Customer engagement In sub-paragraph (a) delete and at the end. In sub-paragraph (b) add and at the end. After sub-paragraph (b) add (c) the Additional Factor determined under paragraph 9.4A, if appropriate. Conclusion As a result of our engagement with customers and CEF which has showed their continued support to reducing leakage beyond the SELL and support for the proposed funding route, we believe the Final Determination should reflect our proposed leakage ODI. PR14 Representations on the Draft Determination anglianwater.co.uk

137 Retail costs This section covers two topics related to retail costs that Ofwat has asked us to include in our representations. First, we address the requirement to provide further assurance over a sample of network and non-network calls used in our retail cost allocation, and second, we explain why non-household retail costs in are higher than the preceding two years. Ofwat's Draft Determination "On the whole we were satisfied with the external assurance that the company had obtained over its cost allocations. However we did note one gap in the external assurance provided. The company had undertaken sampling to allocate its network and non-network enquiries and complaints between household and non-household but had not obtained external assurance on this sample as required by our guidance. This was because the company did not believe that these costs were material. We disagree with this. We consider these costs to be material (as they represent 13.2% of retail costs and 115% of non-household retail costs) and so conclude that external assurance should have been obtained over this sample. We expect the company to provide us with external assurance over this sample with its draft determination representations on 3 October." Company specific appendix page 33 Summary of our Representations We engaged PwC to conduct further assurance on the sample and include the assurance letter below. We engaged our financial auditors, PwC, to carry out some further assurance over the sample used to determine the allocation of costs to network and non-network enquiries. The results of that assurance are set out in the letter below.

138 134 Response to Ofwat's Specified Requirements Retail costs PR14 Representations on the Draft Determination anglianwater.co.uk

139 135

140 136 Response to Ofwat's Specified Requirements Retail costs PR14 Representations on the Draft Determination anglianwater.co.uk

141 137 Ofwat's Draft Determination "Overall the company s proposed costs do not increase by more than our non-household retail materiality threshold of 5.3% between 2015 and We have therefore accepted the company s cost proposals (post the pension deficit recovery adjustment). In total, this resulted in the company s proposed costs being adjusted from million over the control period to million. However, we also note that the company s opex in is significantly higher than the preceding two years. As part of its representations we request for the company to provide us with a clear explanation as to the cost increase, and to explain why the increase should not be treated as an exceptional one-off event." Company specific appendix page 40 Summary of our Representations The increase in non household retail costs relates partly to our revisions to cost allocation and partly to a change in policy towards the treatment of unallocated credits on account from non-hh customers. Non household retail costs in therefore reflect our expectation of baseline costs as we enter AMP6 and are not atypical. The increase in non household costs in above the previous two years is due to an increase in in customer services costs (in particular network and non network queries c. 0.6m pa) and an increase in customer debt charges (c. 1.2m pa). The increases partly reflect the ongoing improvements in our cost allocation methodology which we would expect to continue beyond In relation to the bad debt charge, prior to , the total value of unallocated credits over one year old were netted off our required bad debt provision with the result that our reported charge in those years did not reflect the underlying cost base. Our policy is now to net off the in year movement only as there is a reasonable expectation that new unallocated credits will continue to be received each year. Even with this change, our charge is lower in than in previous years adjusted to reflect the change in policy (row 2), and effectively assumes a run rate of unallocated credits of c. 970k p.a. being received (although in the year to date, the level of unallocated credits has fallen due to refunds being made). Non household retail costs in therefore reflect our expectation of baseline costs as we enter AMP6 and are not an atypical event. We decided at the time of our business plan submission not to adjust historic costs to reflect the updated methodology, but as outlined below, our costs in are well below the previous two years when re-stated (row 2).

142 138 Response to Ofwat's Specified Requirements Retail costs Row Description March 2014 March 2013 March 2012 March Bad debt charge included in Table R4 2,341,000 1,072,000 1,172,000 1,150,000 2 Bad debt charge using PR14 methodology in and ,341,000 3,770,106 3,177,008 n.a. 3 Underlying bad debt charge, before allowing unallocated credits 3,311,321 3,679,023 3,770,106 3,055,008 4 Annual change in the value of unallocated credits ( 970,321) ( 8,917) ( 693,098) n.a. PR14 Representations on the Draft Determination anglianwater.co.uk

143 Wholesale charges Ofwat's Draft Determination "Alongside these draft determinations we have published a spreadsheet template for companies to complete and submit to us by 3 October As part of their 3 October responses, companies should highlight any remaining aspects of the standard wholesale charges schedule that do not fit their charging structures, including clear proposals to address those issues." Draft price control determination notice: technical appendix A8 charging page 4 Summary of our Representations We provide the spreadsheet and a copy of a letter to Ofwat regarding the our charging proposals. We have completed the spreadsheet as requested and we also provide a copy of a letter to Ofwat setting out our charging proposals. We can confirm that there are no material issues to highlight.

144 140 Response to Ofwat's Specified Requirements Wholesale charges Simon Smith Head of Charging Reform Tom Rogers Principal Analyst Ofwat, Centre City Tower, 7 Hill St, Birmingham B5 4UA Anglian Water Services Limited Anglian House Ambury Road Huntingdon Cambs. PE29 3NZ Tel October 2014 Dear Simon, Tom Charges Proposals The purpose of this letter is to bring to your attention notable features of our charges proposals for next year. These proposals are consistent with the completed charges spread sheet we have submitted as part of our company representations on the draft determination, and provide an accompanying commentary on the notable features of that submission. They reconcile to the retail revenue controls set out in the draft determination and the wholesale revenue figures with their implied menu choice (less capital contributions from connection charges and revenue from infrastructure charges) at 2012/13 base inflated to 2015/16 prices i.e. by 9.6%. PR14 Representations on the Draft Determination anglianwater.co.uk Consistent with our practice in previous years, I am copying this letter to CC Water to give sufficient time for both yourself and CC Water to respond with comment and questions before we finalise our proposals in time for the charges scheme submission on or before 16 January, in line with IN 14/15. Please note that these proposals should be considered provisional for three reasons: 1. They have not yet been considered by our Board;