A Summary of Changes to the HS2 Economic Case

Transcription

1 A Summary of Changes to the HS2 Economic Case April 2011

2 Contents 1 Introduction 4 2 Cost Changes 6 3 Appraisal Changes 7 4 Summary of Changes 9 Annex 1: Capital Costs 10 Annex 2: Operating Costs 13 Annex 3: Differences in Approach for the Y Network 18 List of tables Table 1: Modelling and Appraisal Framework Changes since March Table 2: Impact of Cost Changes on the HS2 Economic Case 9 Table 3: Impact of Appraisal Changes on the HS2 Economic Case 9 Table 4: Impact of Modelling Changes on the HS2 Economic Case 9 Table A1: Total Capital Cost 12 Table A2: Total Operating Costs (HS2 and Classic Network) 17 List of figures Figure 1: Construction Cost Profile 10 Figure 2: Rolling Stock Purchase Cost Profile 12 2

3 List of Acronyms ADM BCR MVA-MM NMF PLD PS PM PVB PVC SCM WCML WEI - Air Demand Model - Benefit Cost Ratio - MVA and Mott MacDonald - Network Modelling Framework - PLANET Long Distance - PLANET South - PLANET Midlands - Present Value of Benefits - Present Value of Costs - Station Choice Model - West Coast Main Line - Wider Economic Impacts 3

4 1. Introduction This document is part of a set of documents that describe the changes to modelling, cost estimation and appraisal that have occurred since our March 2010 report 1, and which have been implemented between then and the publication of the Economic Case for HS2: The Y Network and London West Midlands 2 in February Since March 2010 we have taken the opportunity to review and refine the modelling and analytical approach used to estimate the economic case. Further details of the changes in the modelling approach can be found in reports by Atkins 3 (relating to developments up to October 2010) and MVA Consultancy and Mott MacDonald (MVA-MM) 4 (relating to developments since October 2010) that are published separately The work undertaken and reported in the Atkins and MVA-MM reports is mostly related to the development and updating of the PLANET model suite, which we use to calculate the benefits of the scheme. Over the same period, we have undertaken work to improve and refine our approach for calculating costs and undertaking the appraisal. Further details on these changes are included as annexes to this document Table 1 summarises all of the changes that have impacted on the economic case for HS2 since the March 2010 report, and indicates the relevant document where each of these changes is described in detail. 1 High Speed Rail: London to the West Midlands and Beyond. A Report to Government by High Speed Two Limited, ghspeedrail/hs2ltd/hs2report/ Modelling and Appraisal Updates and their impact on the HS2 Business Case A Report for HS2, by Atkins 4 HS2 London West Midlands Consultation Model Development and Baseline Report, by MVA-MM 4

5 Table 1: Modelling and Appraisal Framework Changes since March 2010 Report Author and Changes implemented and documented Title WS Atkins Inclusion of business non-car available rail benefits Modelling and Appraisal Updates Revision of weightings of generalised journey cost and their impact on components the HS2 Business Case - A Report for Revision of interaction between PLD and PS HS2 SCM changes: - address incomplete capture of local leg benefits - application of behavioural weighting to London local leg benefits - remove double weighting of local leg time transferred to PLD - apply local leg times and station shares on a production/attraction basis rather than OD - revised London local leg costs for London/W. Midlands movements - disaggregate London local leg benefits for economic appraisal - add W. Midland local leg costs to non-london movements Updates to ADM MVA-Mott MacDonald Model Development and Baseline Report (This Report includes some further work undertaken by Atkins not documented in above report) Updates to future year demand matrices to take account of: - revised short-medium term economic growth forecasts - impact of Coalition government s policy on regulated rail fares (RPI+3% for three years) - the use of DfT s unconstrained air demand forecast rather than constrained for GB internal air demand - revised forecasts for Heathrow throughput for the ADM Further changes to the SCM: - calibration of the parameter controlling users sensitivity to generalised cost - adjustments to access times to reflect the relative ease of interchange at Old Oak Common compared to Euston 5

6 Changes to behavioural values of time. Corrections to coding of certain rail services on the West Coast Main Line (WCML) Addressing convergence problems through changes to the do-minimum scenario HS2 Ltd A summary of changes to the HS2 Economic Case Changes to construction costs Changes to operating costs Changes to treatment of indirect tax Impacts of connection to HS1 Corrections to discounting The remainder of this report is structured as follows: Section 2 outlines the changes that have been made to costs estimated in the appraisal. Section 3 outlines some other technical changes, as well as additional benefits included for the HS1 link Section 4 then summarises the impact on the business of all of the changes to the modelling and appraisal methodology since March Cost Changes 2.1. Capital Costs Since March 2010 we have refined and improved our approach to calculating capital costs. This has involved: Recalculation of construction costs to reflect revised assumptions on tunnelling costs, as well as the inclusion of Heathrow stubs 5 and the HS1 link Changing the profile of rolling stock purchase costs, allowing rolling stock to be purchased later, reflecting the fact that with 5 Heathrow stubs are the infrastructure required to build a spur to Heathrow at a later date without disruption to HS2. 6

7 lower demand forecasts not all of the rolling stock would need to be operational on the opening day The net impact of these changes is to reduce capital costs by 0.1bn Operating Costs on HS2 and Classic Lines We have taken the opportunity to further review operating costs and in particular how these may change over time. Previous analysis focused on fixed unit rates of costs that did not change over time. The outcome of this review: Adjusted the likely cost of testing the line prior to opening (now assumed to be 50% of the normal cost of operating the line) Reviewed likely changes in real wage costs and productivity, particularly for train drivers where scope for productivity growth was felt to be less likely Incorporated the latest forecasts on electricity prices from the Department of Energy and Climate Change to better reflect future traction power costs. This nearly doubles energy costs by The net effect of these improvements in methodology is to increase the estimated operating costs of HS2 over the 60 year appraisal period. However it also has knock on effects of increasing the savings made on the classic network We have also taken the opportunity to correct two errors in our estimation of operating costs. The first was a misinterpretation of modelled data that resulted in under-estimating the reduction in classic line train km. The second was an error in the application of optimism bias in operating costs. The effect of both of these corrections has been to reduce the overall operating costs in the appraisal. 3. Appraisal Changes 3.1. Discount Rate Following the March 2010 report, HS2 Ltd clarified the approach to discounting outlined in WebTAG. Guidance on discounting suggests a lower discount rate should be applied after 30 years, and then again after a further 45 years. However it was unclear when the 30 years should be taken from. 7

8 Following discussions with DfT it was agreed the correct approach would be to apply the 30 years from the appraisal base year in this case Indirect Taxation DfT have changed the way in which the BCR is calculated to reflect the new method which has been used for the Comprehensive Spending Review. Under the previous system used in the March 2010 report, the loss of indirect taxes paid to government that the scheme would lead to was subtracted from the revenue it would generate. This was then subtracted from the total capital and operating costs to give a total cost to government known as the Present Value of Costs (PVC) This was altered for the spending review and we expect this revision to be incorporated into WebTAG very shortly. The loss of indirect taxes paid to government is now subtracted from the total Present Value of Benefits (PVB), rather than from the revenue. The net result of this is that the indirect tax loss is now part of the numerator rather than the denominator of the BCR calculation, as illustrated below: Old calculation: New calculation: BCR = PVB / (PVC Revenue Indirect Tax Change) BCR = (PVB + Indirect Tax Change) / (PVC Revenue) 3.3. HS1 Link The HS1 link is now included in the Day 1 scheme. We have appraised this slightly differently to the rest of the network Analysis by SKM and Atkins 6 suggested that in the absence of an HS1 link, HS2 could offer benefits of around 350m to international passengers who might use HS2 to access Eurostar services at St Pancras. With an HS1 link and the opportunity to interchange at Old Oak Common, Atkins estimated these passengers would gain a further 350m benefits (i.e. 700m in total) Since the HS2 proposal now includes a link to HS1 from Old Oak Common we have included some of these benefits in the economic case. However we have been conservative by only including the 350m benefits of an interchange at Old Oak Common. 6 International Connections - A Report for HS2 Ltd by Atkins 8

9 4. Summary of Changes Tables 2-4 below show the effects of the changes that have been undertaken since the March 2010 report The changes documented in this report are shown in tables 2 and 3, whilst table 4 shows the changes included in Atkins and MVA-MM s documents The annexes to this report follow, and explain in detail the methodology we use in costs and appraisal. Table 2: Impact of Cost Changes on the HS2 Economic Case March 2010 Report Revised Capital Costs Revised Operating Costs Revised Classic Line Op Costs Revised Application of Optimism Bias Impact of Cost Changes PVB 28,708 28,708 PVC 11, , ,517 BCR* * excludes WEIs Table 3: Impact of Appraisal Changes on the HS2 Economic Case Change in BCR metric Correction of Discount Rate Benefits and Costs of HS1 Link Impact of Appraisal Changes PVB 28,708-1, ,130 PVC 9,517-1, ,675 BCR* * excludes WEIs Table 4: Impact of Modelling Changes on the HS2 Economic Case Model Updates to Dec 2010 Model Updates since Dec 2010 Demand Forecast Updates Impact of Modelling Changes PVB 28,130-6, ,912 16,544 PVC 8,675 1, ,320 BCR* * excludes WEIs 9

10 Annex 1 - Capital Costs Construction Costs 1. Our central estimate of construction costs for HS2 is 16.8 billion (2009 factor prices) including quantitative risk assessment and optimism bias of 34%, which takes account of the scope of work being fixed, and a limited amount of innovation being required. This reflects two major changes since the March 2010 report: the cost of the HS1 link is now included; and tunnelling costs have been reduced in light of the work undertaken by Infrastructure UK. 2. A spur linking HS2 to Heathrow is being considered as part of ongoing work for the design of the wider network. We have included the cost of stubs to allow this to be constructed easily at the same time as the Y network in the Day 1 costs, which is 50m (2009, factor price). 3. For the purposes of appraisal we add a further adjustment to convert these costs into market prices. This adjustment is required to ensure that costs and benefits are on a comparable basis. 4. Including these adjustments brings the total construction cost to 20.3bn in market prices. Figure 1 shows the profile of the construction costs. Discounting this profile of costs, the 2009 present value of all construction costs is 13.9 billion. Figure 1 - Construction Cost Profile bn Year 10

11 Renewals 5. In addition to the initial costs of construction, some elements of HS2 infrastructure would need substantial renewal and reconstruction during the 60 year period over which we appraise HS2. Civils and structures Asset life greater than 60 years, and therefore no renewal costs during the appraisal period Railway systems 100% renewal after 30 years Control systems 100% renewal after 15 years Stations complete renewal of some elements of construction after 40 years Depots 50% renewal after 30 years 6. The overall impact of the these assumptions is a renewal cost for the scheme of 1.1bn (PV, 2009 prices) Rolling Stock Purchase 7. Table 8 of the Economic Case for HS2 provides the cost of rolling stock including risk and optimism bias as 2.84billion, spread over the 6 years to The rolling stock is expected to have a useful life of 35 years, and in line with the approach to construction costs and renewals we have assumed that the full stock will be replaced after 35 years at the same cost in real terms. 8. As a 2009 present value converted into market prices rolling stock capital costs add a total of 2.8 billion, bringing the total Capital Costs to 17.8bn (PV, 2009 prices). 9. Figure 2 shows the profile of rolling stock purchase costs: 11

12 Figure 2 - Rolling Stock Purchase Cost Profile bn Year Total Capital Cost 10. The total capital cost is shown in the table below: Table A1: Total Capital Cost billion (2009 PV) Construction 13.9 Renewals 1.1 Rolling Stock Purchase 2.8 TOTAL

13 Annex 2 - Operating Costs 1. A full explanation of the derivation of individual operating costs by category is detailed in paragraphs 4-23 below. These categories are: Rolling Stock Maintenance Rolling Stock Traction Power Train Crew Track Maintenance Station Operating Costs Station Maintenance Classic Line Operating Costs 2. The key drivers of most operating costs are the number of train sets required, and the annual train km covered. These are: Number of HS captive sets: 16 Number of classic compatible sets: 45 Annual train km of HS captive sets: 8,856,750 Annual train km of classic compatible sets: 29,745, The annual operating costs are implemented every year from 2025 to the end of the appraisal period, with the following exceptions: We treat 2025 as a test year. Since passenger services would not be operating during this period, we assume that operating costs will be around half, as a full timetable would not be run. We assume no Station Operating Costs or Station Maintenance during this period, which we would expect to have no cost during the test year Our cost assumptions reflect the likely profile of demand in the future. It is unlikely that a full capacity of 400m trains during peak periods would be required before Up to this point some peak services would only be 200m in length, reducing operating costs by around 24% in the opening year. Operating costs then increase linearly to 100% in the year We assume that traction costs and driver costs will increase above inflation these issues are covered in more detail below. 13

14 Rolling Stock Maintenance 4. The cost and risk model 7 estimates the maintenance costs of rolling stock at 2.80/km for HS captive stock and 3.50/km for hybrid stock. It also suggests the annual train km run by each type of stock (see above). Using these figures and adjusting for market prices suggests the annual cost of maintenance is 220m (2009 market price) or 2.4bn present value over 60 years. Rolling Stock Traction Power 5. We have taken the opportunity to significantly improve our estimate of traction power costs, drawing on the latest evidence. Instead of a single cost estimate, we have used real data on costs of energy used in the rail industry and increased these annually in line with the latest projections of energy costs from DECC. 6. This means: We now base the cost of energy on the actual 2010 cost of energy for TOCs as advised by Network Rail (7.3p / kwh) We now grow this cost in line with DECC s projections 8 of the cost of energy. These project that energy costs will nearly double by the year 2030 In addition we have corrected the treatment of hybrid stock running on the classic network. Any train on the High Speed network is assumed to use 28 kwh / km (based on detailed modelling of the proposed HS2 route). Beyond this, we assume consumption will be similar to existing Pendolinos (14.3 kwh / km). We do not assume any increase in efficiency in train technology, which is a conservative assumption. 7. The annual cost of traction power is 150m (2009 market price) from 2030 onwards, and slightly less than this in the preceding years. We discount this to give a PV cost of 2.2bn. 7 HS2 Cost and Risk Model ghspeedrail/hs2ltd/riskmodel/pdf/report.pdf

15 Train Crew 8. The train crew costs are the cost of the driver and conductor. These have been estimated to be 0.78 / km for drivers and 0.56 / km for conductors. 9. Since 400m services to Birmingham will be made up of 2x200m sets with no connecting corridor, such services will require one driver and two conductors. 10. We have made one adjustment to the way in which these costs were calculated in the March 2010 report for the consultation: we now assume that driver costs will increase by RPI + 1.5% per annum, whereas previously we assumed they would increase by RPI. Conductor costs continue to rise at RPI. This reflects the relative scope for productivity gains from these two roles. 11. The annual cost of train crew would be 69m in This rises as driver wages increase. We discount this to give a PV cost of 1.2bn. Track Maintenance 12. We calculate the cost of track maintenance by multiplying the length of track by the cost per kilometre of maintaining the track. The cost and risk model document gives the cost per kilometre as 180k / km (based on the observed cost of maintenance on HS1), with 207km of track planned for HS2. Adjusting for market prices, this gives a cost of 45m per annum and a PV cost of 0.7bn. Station Operating Costs 13. We calculate Station Operating Costs by using bottom up estimates of the number of staff required for different types and sizes of station. We benchmark the assumptions for this against similar stations for example, Old Oak Common is benchmarked against Stratford International station on HS This gives a market price cost of 15m per annum. We discount this to give a PV cost of 0.2bn. 15. There has been no change to the method of calculation for this between the March 2010 report and the consultation document. 15

16 Station Maintenance 16. We calculate Station Maintenance on the basis of observed data for a four-platform station, with this scaled up for the larger stations on the HS2 route. 17. This gives a market price cost of 4m per annum. We discount this to give a PV cost of less than 0.1bn 18. There has been no change to the method of calculation for this between the March 2010 report and the consultation document. Classic Line Operating Costs 19. The derivation and calculation of classic line savings has been substantially updated and improved since the March 2010 report. There are two important factors in calculating classic line savings: the change in train kilometres on classic rail; and the cost per train kilometre on classic rail. Both of these are calculated for individual stock types, since the cost per kilometre varies widely between different types of rolling stock. 20. There are four types of rolling stock affected by the change in the classic line timetable that has been modelled for HS2. These are: Pendolino Voyager / Super Voyager Desiro Class We have used PLANET Long Distance, PLANET Midland and PLANET South to determine the change in train kilometres that will occur as a result of introducing HS2. In revisiting the calculations that were made for the March 2010 report, we discovered an error in the calculation of this change in train kilometres. This has been corrected in our latest estimates. 22. In addition we reviewed the modelling of the classic line services to Glasgow. Previously we had erroneously modelled an increase in the number of classic line services running to Glasgow after the introduction of HS2. This has been corrected so that the classic line service level to/from Glasgow is now the same after the introduction of HS2 as before. 23. To calculate the costs per kilometre for different types of rolling stock we gathered data from the DfT s Network Modelling Framework (NMF) to estimate fuel, lease and maintenance costs of rolling stock, as well as 16

17 staff and overhead costs. The latter two costs were allocated according to an average cost per train km rather than specific types of rolling stock. This is broadly the same process as the March 2010 report, however electricity costs (which were previously erroneously excluded) have now been incorporated into costs using the same approach as for HS2, described in paragraph 6 of this annex. Additionally some minor corrections have been made to the interpretation of data from the NMF. Total Operating Costs 24 The total operating costs for HS2 and the classic network are shown in the table below: Table A2: Total Operating Costs (HS2 and Classic Network) billion (2009 PV) Rolling Stock Maintenance 2.4 Rolling Stock Traction Power 2.2 Train Crew 1.2 Track Maintenance and OpEx 0.7 Station OpEx 0.2 Station Maintenance <0.1 Classic Line Operating Costs -2.3 TOTAL (excluding Optimism Bias) 4.4 TOTAL (including Optimism Bias)

18 Annex 3 - Differences in Approach for the Y Network 1 HS2 Ltd has been asked by Government to develop a detailed assessment of the business case for extending the proposed HS2 line from the West Midlands to Manchester and Leeds. We are due to report on this at the end of However as part of the consultation, HS2 Ltd has made an initial assessment of the potential economic case. The approach to this is outlined below. Costs 2 We have used estimates of the unit cost of track in different situations (tunnel, at surface and viaduct) to provide a high level assessment of the likely costs of extending the line to Manchester and Leeds. Based on this assessment, we estimate that the total infrastructure capital cost of the Y network including a link to Heathrow, would be around 32.2 billion (Quarter 3, 2009 prices), including risk allowances and optimism bias. Renewals of infrastructure are included in the same proportion as the London to West Midland scheme. 3 The cost of purchasing, operating and maintaining rolling stock similarly uses a higher level approach. We have not yet calculated diagrams an assessment of how an indicative timetable might be served by individual trains and therefore we use an algorithm based on the distance and journey time of trains on a particular route to estimate the distance travelled, and the number of train sets that might be required. An additional allowance is made for maintenance and lengthening of trains. 4 This gives an estimate of the number of trains and the annual distance travelled by HS2 trains to which we apply exactly the same methodology as in Annex 2. 5 We estimate Track Maintenance costs, and Station Maintenance and Station Operating Costs in the same fashion as in Day 1. Appraisal 6 The appraisal of the full Y network is complicated slightly by the staggering of opening years. Appraising the scheme over 60 years from opening of the first phase would capture the full costs of the scheme, without comparable benefits. An alternative and perhaps more balanced approach would be to appraise the incremental benefits beyond the first phase. However this would implicitly treat the first phase of the scheme as committed, which is not appropriate given the current consultation and prior to a decision on the first phase. 18

19 7 The key factor in appraising the scheme is, however, that it is assessed on a comparable basis to the strategic alternatives, and that costs and benefits are appraised over the same period. We have therefore agreed with the Department that the appraisal should take accounts of all costs and benefits up to a point 60 years beyond the opening of the full Y network. Additional Impacts 8 Our assessment of the economic case for the Y is based on the model and the methodology used in our London to West Midlands work. However there are a number of additional impacts that we have not been able to incorporate in our modelling at this stage. To reflect this, a series of off-model estimates of the potential scale of these impacts has been made: Released Capacity - we looked at the modelled benefits from released capacity on the WCML in our London to West Midlands work, and made a deliberately cautious assumption that we would receive only half that level of benefit from capacity released on the Midland and East Coast Main Lines combined. Wider Economic Impacts - If these are in the same proportion to transport benefits as for the London to West Midlands scheme, the Y network would deliver a further 4.7 billion of benefits. Indeed the proximity of Leeds, South Yorkshire and the East Midlands may mean agglomeration benefits are stronger than those observed in the London to West Midlands scheme. However, to be prudent, for these estimates we have only taken half the proportionate benefits. Cost savings on the classic network We estimated the potential for classic line operating cost savings by estimating the total train km that would be saved by removing some of the long distance services currently operating on the Midland Mainline and East Coast Mainline that the Y network would be expected to replace. Our central estimate includes only half of this potential saving. 9 We believe these are cautious assessments of the potential impact on the overall economic case. Further analysis will be conducted during the course of detailed work during

20 10 Given the reduced level of certainty when calculating benefits and costs for the Y extensions, we have produced low, central and high estimates for: Benefits from released capacity Wider Economic Impacts (WEIs) Classic line operating costs 11 The high estimate for benefits from released capacity is equal to the benefits from released capacity calculated for Day 1. Whilst we expect the Y extensions to give a similar benefit from released capacity as Day 1, we decided to be cautious and make our central estimate 50% of this. Our low estimate is for there to be no benefits from released capacity realised by the Y extensions. 12 We derived the high estimate for WEIs on the Y extensions by assuming that they would be the same proportion of total benefits as they were in Day 1. Due to the closer proximity of the areas served by the Y extensions in comparison with Day 1, there is reason to believe that WEIs from the Y extensions will actually be proportionately larger than those from Day 1. However, we have cautiously made our central assumption of these 50% of our high estimate, with our low estimate being that the Y extensions provide no WEIs. 13 We derived the high estimate for classic line operating cost changes by estimating the total train km that would be saved by removing some of the long distance services currently operating on the Midland Mainline and East Coast Mainline that the Y network would be expected to replace. This was then treated in the same manner as for Day 1 (see Annex 2) to calculate a total costs saving. Again, we made our central estimate 50% of this. Our low estimate is for no classic line operating cost savings. 20