REPORT NO. AUTHOR DATE : : X : 15 TH TH JANUARY 2010 2010 Interfleet Technology Ltd Interfleet House Pride Parkway Derby DE24 8HX Registered in England No. 3062722
TITLE : LEN ENGTH EPORT NO. : REPORT ISSUE SSUE : 1 REV DATE ATE : 15 TH EV 1 TH JANUARY 2010 2010 ORIGINATOR : DATE X INTERFLEET TECHNOLOGY LIMITED ATE 15/01/2010 HEAD OF RAILWAY SYSTEMS AND A STRATEGY CHECKED BY : DATE X INTERFLEET TECHNOLOGY LIMITED PRINCIPAL CONSULTANT ATE 15/01/2010 APPROVED BY : DATE X INTERFLEET TECHNOLOGY LIM IMITED ITED PRINCIPAL CONSULTANT ATE 15/01/2010 DISTRIBUTION : NAME X TITLE PROJECT SPONSOR COMPANY DEPARTMENT OF TRANSPORT INTERFLEET TECHNOLOGY LTD TD. 2010 ALL LL RIGHTS RESERVED No part of this work may be reproduced or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, or stored in any retrieval system of any nature, without the written permission of Interfleet Technology Ltd, application for which shall be made to the Managing Director, Interfleet Technology Ltd, Interfleet House, Pride Parkway, Derby DE24 8HX. ASSIGNMENT NUMBER UMBER : T22898 ASSIGNMENT MANAGER ANAGER : X TEL EL: +44 +44 (0) 1332 223088
AMENDMENTS ISSUE REVISION DESCRIPTION DISTRIBUTION DATE 1 0 First Issue X 12/01/10 1 1 2 nd issue X 15/01/10 Page No. 3 of 69
CONTENTS PAGE 1 EXECUTIVE SUMMARY... 5 1.1 1.1 APPROACH 1.2 1.2 RESULTS 1.3 PPROACH...... 5 ESULTS......... 6 1.3 CONCLUSIONS 1.4 ONCLUSIONS...... 9 1.4 RECOMMENDATIONS ECOMMENDATIONS...... 9 2 GLOSSARY OF TERMS... 11 3 INTRODUCTION...... 11 4 METHOD...... 12 4.1 4.1 CAPACITY AND CONFIGURATION 4.2 4.2 ALTERNATIVE ROLLING STOCK CONFIGURATIONS 4.3 4.3 OPERATIONAL EVALUATION 4.4 4.4 TRAINCREW REQUIREMENTS 4.5 ONFIGURATION... 13 STOCK CONFIGURATIONS... 20 VALUATION...... 25 EQUIREMENTS...... 33 4.5 ROLLING STOCK MAINTENANCE AND RELIABILITY 4.6 4.6 MAINTENANCE AND RELIABILITY COSTS OSTS, RESULTS 4.7 6 AND 4.8 DEPOT 4.9 AND 8 CAR SENSITIVITY ANALYSIS EPOT, STABLING AND INFRASTRUCTURE IMPACT 4.9 FINANCIAL MODEL ELIABILITY... 45 ESULTS... 50 NALYSIS... 52 MPACT... 55 ODEL...... 59 5 CONCLUSIONS...... 67 6 RECOMMENDATIONS... 68 7 APPENDICES...... 69 Page No. 4 of 69
1 EXECUTIVE SUMMARY 1.1 The DfT wished to undertake a review of Thameslink unit configurations in order to determine if the original, base case, 12 and 8 car concept fleets still represent the most appropriate solution for delivery of the service specification. The original decision to specify fixed formation units was based on the need to meet high passenger demand in the core peak by operating a 24 tph high integrity service. Fixed formation units also offered an elegant solution to PRM boarding and alighting which supports the object of short dwell times during peak services. When considering other fleet configurations these, and other, important issues must be taken into consideration to ensure that the full impact of any proposed change is considered from a system wide perspective. The options that were considered and benchmarked against the base case are: 8 and 4 car units 8 and 6 car units 4 car units The basis for assessing these options is that they can each operate in multiple formation to form a 12 car service train. Savings will be made as a result of the ability to split the train at convenient times resulting in lower mileage and maintenance cost, reduced energy consumption, reduced VTAC and possibly reduced cost associated with the depot and stabling provision. Clearly additional cost will be also be incurred in associated with increased rolling stock cost, splitting and joining infrastructure, PRM and operational staff costs. The objective of this study is to assess these issues in detail for each option, compare the result against the base case fixed formation units and consider whether the cost savings that emerge are sufficiently robust to warrant changes to the rolling stock strategy. 1.1 APPROACH Our approach to this study has been to assess the demand figures for the peak shoulder and off peak both in the core and outside of the core and then consider which fleet options can accommodate the required peak, shoulder peak and off peak demand. For each option we have considered how equipment will be redistributed throughout the train and determined the impact on unit capacity, cost, weight and reliability. We have taken the Version 1e Thameslink diagrams and amended them to include splitting and joining at the terminus stations associated with each service group. This has enabled us to determine the required fleet size and mix, unit mileage and number of splits / joins required for each option. Using the revised fleet diagrams we have developed crew diagrams which have enabled us to estimate the additional cost of drivers associated with splitting and joining. For each option we have calculated a revised pence per mile maintenance cost. This cost takes into consideration the additional maintenance cost associated with more cabs, auto couplers and reconfigured train systems, the overall reduction in vehicle mileage Page No. 5 of 69
1.2 due to splitting and joining and the depot operating costs that must be amortised across the fleet mileage. The three optional unit configurations all weigh more than the base case as a result of additional cabs, PRM vehicles and associated equipment. This change in weight results in the train (when in 12 car operation) drawing additional energy (EC4T) from the traction supply and incurring a higher Variable Track Access Charge (VTAC). The revised fleet configurations and diagrams were supplied to Arup and Network Rail so that the impact on depots, sidings, stations and infrastructure capital and operational costs can be determined. The outputs of these studies are included within our cost model. The result of each workstream has been captured within a cost model which shows the NPV for all capital and operational cost deltas for each option compared against the base case and appraised over a project life of 30 years. Where the model considers costs or savings which are based on values which have not been substantiated we have undertaken a sensitivity analysis which tests the impact of a range of values on the overall cost model. 1.2 RESULTS The early assessments of peak and off-peak passenger demand revealed that that the 4 car unit option would not be able to deliver the capacity required. The saloon space required for additional cabs and PRM toilets significantly reduces passenger seating and standing capacity and would cause unacceptable overcrowding. Further to this, the capital cost of the additional equipment required (including cabs, couplers, pantographs, PRM toilets) more than off-set any operational saving opportunities through fleet mileage reduction. As such the 4 car unit option was not considered further in the analysis. The 8 & 4 car units and 8 & 6 car units were both able to deliver satisfactory passenger loadings in spite of slight reduction in capacity due to cabs and PRM toilets. Both of these options also give comparable vehicle capital cost increase. However, operational cost savings due to fleet mileage reduction through splitting and joining are not as prominent with the 8 & 4 car unit option as mileage savings are achieved only with the 4 car unit. Also, due to the inherent asymmetry of the 8 & 4 car option when formed in a 12 car consist, there are likely to be operational complications due to unpredictable train orientation. These will impact on dwell times (PRM and first class locations) and stabling. Therefore, the 8 & 6 car unit options was the only practical alternative to the base case. The comparative analysis of the base case and the 8 & 6 car unit option is demonstrated below in the summarised cost model analysis. In order to demonstrate the sensitivities of the 30 year NPV cost savings to the capital lease cost and operational costs, the best, mid, worse and most likely case results are shown in tables 1-1, 1-2, 1-3 and 1-4. It should be noted that the most likely case reflects the view of Interfleet formed with input from the Thameslink stakeholders. The most likely case has been formed by flexing the principal inputs to the cost model to determine a most likely 30 year NPV cost delta. Page No. 6 of 69
The most likely case considers a mid range capital cost for the 8 & 6 car unit rolling stock when benchmarked against other modern rolling stock vehicle prices. The additional cost associated with splitting and joining of multiple units (8 & 6 car option) assumes worst case costs as the detailed scope of any platform/infrastructure works has not been developed and there therefore a risk here that the costs could be towards the top end of the range. Additional PRM costs are assumed to be near the middle of the range as there is flexibility in whether the optimum 8 & 6 car solution could be train based or platform based. Cost estimates for each are close to the mid range for the PRM solution. The mileage savings gained through splitting and joining of the 8 & 6 car units are assumed to be mid range. This reflects splitting and joining during the interpeak and before the evening peak but none before the morning peak. The mileage saving in turn drives the impact on EC4T, VTAC and mileage driven maintenance costs. The best case has been assumed for impact on reliability of single pantographs on 6 car units and additional auto couplers. Interfleet and stakeholders agreed that the reliability analysis undertaken was conservative and as such the best case was adopted. Table 1-1 1 Best Case Scenario Measures 12 + 8 6 + 8 Deltas Percentage Variance Fleet Size (Service Vehicles) 1,116 1,116 0 0% Annual Mileage (inc ECS) 202,305,864 154,603,498-47,702,366 Annual Kilometres (inc ECS) 325,579,728 248,810,212-76,323,786-24% 30 year NPV: Ops - -11% 30 year NPV: Infrastructure 30 year NPV: Capital Lease +6% 30 year NPV total - -3% Table 1-2 1 Mid Case Scenario Measures 12 + 8 6 + 8 Deltas Percentage Variance Fleet Size (Service Vehicles) 1,116 1,116 0 0% Annual Mileage (inc ECS) 202,305,864 157,741,150-44,564,714 Annual Kilometres (inc ECS) 325,579,728 253,859,773-71,303,542-22% 30 year NPV: Ops - -11% 30 year NPV: Infrastructure 30 year NPV: Capital Lease +9% 30 year NPV total - -1% Table 1-3 1 Worst Case Scenario Measures 12 + 8 6 + 8 Deltas Percentage Variance Fleet Size (Service Vehicles) 1,116 1,116 0 0% Annual Mileage (inc ECS) 202,305,864 178,464,144-23,841,720 Annual Kilometres (inc ECS) 325,579,728 287,210,199-38,146,752-12% 30 year NPV: Ops - -3% 30 year NPV: Infrastructure 30 year NPV: Capital Lease +12% 30 year NPV total +5% Page No. 7 of 69
Table 1-4 1 Most Likely Scenario Measures 12 + 8 6 + 8 Deltas Percentage Variance Fleet Size (Service Vehicles) 1,116 1,116 0 0% Annual Mileage (inc ECS) 202,305,864 157,741,150-44,564,714 Annual Kilometres (inc ECS) 325,579,728 253,859,773-71,303,542-22% 30 year NPV: Ops - -11% 30 year NPV: Infrastructure 30 year NPV: Capital Lease +7% 30 year NPV total - -2% The results demonstrate that the best case scenario for 8 & 6 car fleet configuration delivers a 3% saving compared to the base case. The worst case scenario delivers a 5% cost increase compared to the base case. The most likely scenario when compared to the base case gives a 2% saving, which corresponds to an estimated saving of (30 year NPV). Table 1.5 below shows the discrete out turn costs (and savings) used within the model. The two dominant elements that impact on the life cycle cost model are the increased capital costs associated with reconfigured rolling stock which must be rentalised over the life of the asset and the principal saving that is made as a result of accumulation reduced vehicle mileage which delivers savings in maintenance cost, EC4T and VTAC. Table 1.5 V1e Base Case Best case 8s and a 6s Mid case 8s and 6s Worst Case 8s and 6s Most Likely 8s and 6s Capital Expenditure (Assumed one off cost not reantalised) Capital Expenditure + + + + Stabling infrastructure 0 0 0 0 capital al cost Splitting & Joining + + + + Locations Capital Cost Depot Capital Cost 0 0 0 0 Allowance for PRM + + + + Annual Costs (Annual cost over 30 years) EC4T per annum - - - - VTAC per annum - - - - Operations per annum + + + + Maintenance per - - - - annum Coupling & Pantograph Reliability Cost per annum + + + + The key advantage of the 8 and 6 car unit configuration is the opportunity to reduce fleet mileage through splitting and joining in order to better meet fluctuations in passenger demand throughout the day (peak and off-peak). The reduced mileage drives Page No. 8 of 69
1.3 reductions in most elements of the operating costs shown in table 1.5 as annual costs. In NPV terms, the most likely case (table 1-4) estimates a 30 year operational saving of when compared to the fixed formation base case. However, in order to enable this flexibility, it is necessary to buy a different train. The different train is the 8 and 6 car unit fleet. Simplistically, this requires additional cabs and auto-couplers, additional PRM toilets, additional facilities for splitting and joining and additional PRM facilities at the platform/train interface. These are reflected as capital expenditure in table 1-5, and drive a 30 year NPV cost increase of + giving a cost increase of. The resultant saving for the most likely case is therefore X giving a saving of in (approximately 2%). 1.3 CONCLUSIONS 1.4 8 and 6 car Fleet The results of this study have shown that the 8 and 6 car fleet meets the loading requirement for the route and offers the greatest potential life cycle cost saving when compared against the base case V1e fixed formation fleet. PRM provision is made more complex resulting in platform ramps being difficult to implement at all locations in the core but particularly St Pancras and Farringdon. 8 and 4 car Fleet The 8 and 4 car fleet meets the loading requirement for the route and offers a marginal life cycle cost saving (in the best case scenario) when compared to the base case V1e fixed formation fleet. As with the 8 and 6 car fleet PRM provision is made more complex resulting in platform ramps being difficult to implement in the core. 4 car Fleet The 4 car fleet cannot meet the loading requirements for the route and, because of the additional cabs, PRM vehicles and associated systems and equipment does not offer any cost saving when compared to the V1e base case fixed formation fleet. 1.4 RECOMMENDATIONS A 4 car unit fleet configuration is not suitable for Thameslink services because it fails to meet passenger loading requirements, offers no life cycle cost benefits and is not compatible with the infrastructure. This option should therefore be dismissed. The 8 and 4 car fleet configuration offers only a marginal cost saving (in the best case scenario) when compared to the base case fixed formation fleet. The benefits of switching to this fleet configuration is not sufficiently great to warrant the risk associated with making the change at this late stage in the project. This option should not be considered further. Page No. 9 of 69
The 8 and 6 car fleet configuration delivers the greatest life cycle cost saving and operational flexibility, however the percentage savings achieved compared to the base case fixed formation units are relatively small over the 30 year project life and therefore there is a risk that savings may not be achieved if, for example, the additional capex associated with multiple unit rolling stock is at the higher end of our tolerance. Furthermore there are a number of areas of risk that require further analysis and may as a result reduce the benefits estimated so far. The risks are: Agreement that a higher loading level and therefore off crowding level is acceptable when 6 car units are operating the off and interpeak for some of the busier routes. The required level of Network Rail infrastructure reconfiguration and associated costs is not fully detailed. A station capacity study may need to be undertaken at locations where splitting and joining are likely to take place in order to confirm. The ability of the 12 car train (made up of two 6 car units) to meet the dwell time targets has been tested / modelled and found to be acceptable, however the increased complexity associated with PRM has the potential to adversely affect dwell time therefore further modelling is required to specifically address this issue. It is clear that the financial case is not strong enough to justify recommending 8 and 6 car units for the V1e service specification. Based on this analysis it is therefore recommended that this option is not considered further and that fixed formation 8 and 12 car units are retained. Page No. 10 of 69
2 GLOSSARY OF TERMS TTS - Train Technical Specification TIIS - Train Infrastructure Interface Specification VTISM- Vehicle Track Interface Strategic Model VTAC - Variable Track Access Charge FCC - First Capital Connect ECS - Empty Coaching Stock EC4T - Energy Cost for Traction TOCs - Train Operating Company PIS - Passenger Information System CRS - Computerised Reservation System DfT - Department for Transport 3 INTRODUCTION The DfT wish to revalidate the strategy that led to the specification of 12 and 8 car fixed formation units for the Thameslink service. To do this a detailed evaluation of other unit configurations must be undertaken. The options that will be considered and benchmarked against the 12 and 8 car fixed formation base case are: 8 and 4 car units 8 and 6 car units 4 car units It is a pre-requisite that a 24tph service is maintained. In order to determine the most appropriate fleet configuration for Thameslink it will be necessary to consider all aspects of the service in a quantified manner and then benchmark each option against each other. The study consequently evaluates a range of issues examples of which are show below: Operations Issues o Diagrams, mileage and fleet size o Crew numbers o Split and joining support staff Train Technical Issues o Performance, weight, VTAC, EC4T, Train Maintenance o Maintenance change for a reconfigured unit o Impact of mileage change on maintenance cost Station Infrastructure Issues o Signalling changes Customer Service Issues o Capacity, provision of first class, PRM Page No. 11 of 69
The analysis and results are captured within a comparative cost model within which life cycle cost is developed and sensitivity analysis is applied. The results are then benchmarked against the base case fixed formation units. 4 METHOD Our approach to this study was to form a team which has the skills to undertake detailed analysis across a range of disciplines as depicted in figure 4-1below. Figure 4-1 4 As can be seen in the chart above all activities undertaken are linked, therefore it is necessary to undertake the work in a logical sequence so that the output of one workstream can inform the activity of the next, resulting ultimately in a complete model. It should be noted that Arup and Network Rail undertook the depot and stabling analysis and consequently the diagrams associated with each option were supplied. The results of the Arup / Network Rail work can, when available, be input into our overall cost model. Page No. 12 of 69
4.1 4.1 CAPACITY AND CONFI ONFIGURATION The first step in determining the unit configuration is to understand the percentage utilisation of the passenger capacity in the peak, shoulder peak and off peak for each configuration. To make this evaluation it is necessary to understand the passenger demand for Thameslink services throughout the day. Figure 4.1-1below illustrates how loading changes throughout the day: Figure 4.1-1 It was agreed with the Department that the approaching the core future demand projections used for the original unit configuration in the June 2008 executive paper would be used to determine the ability of the alternative configurations to meet the peak loading requirements. 2026 Approaching the CoreC Passenger Demand Forecasts. The approaching the core demand forecast for 2026 indicates that 43000 passengers will use the 18tph (240m services) approaching London Bridge in the morning peak. This represents an average three hour morning peak load of approximately 800 passengers per train. As only morning peak demand forecasts are available it is necessary to apply adjustments factors to determine high peak, shoulder peak and off peak as follows; Page No. 13 of 69
Table 4.1-1 Adjustment Average High Peak Shoulder Peak Off Peak from Average 3 hour morning peak morning 3 hour peak load 800 150% 1200 75% 600 35% 280 Unit Capacity. Using the seating and standing capacities specified by the Train Technical Specification as a guide the seating and standing capacities of the unit configurations under consideration are given below. Table 4.1-2 Unit Length 1 Class Seating Standard Class Standing Capacity Capacity Seating Capacity 12 cars 48 572 1128 8 cars 48 364 732 6 cars 24 276 542 4 cars 24 164 258 The capacities above are for units without end gangways. For the 12 car and 8 car units, bidder designs meet the TTS capacity requirements. For 6 car and 4 car units the capacities have been determined with reference to bidders designs and making appropriate configuration adjustments to accommodate the necessary cabs, toilets and PRM facilities. Train Loading Approaching the Core. The level of train loading is measured as a percentage of seat utilisation; greater than 100% indicates passengers are standing. The percentage train loadings for the unit configurations are given in the tables below. It has previously been suggested that for off peak services crowding off occurred above loading levels of 60%. We consider this to be an unrealistically low level that does not make efficient use of the train or fleet capacity, particularly approaching the core (the basis for this initial review). We believe that off peak loading levels of 85% to 100% are acceptable approaching the core area provided they provision for appropriate projected growth levels. The train capacity loading results are presented as a percentage of standard seating capacity. A 100% loading means all standard seats occupied. Total train capacity (i.e. crush laden), with all seats occupied and all standing space occupied equates to approximately a 300% loading. The train capacities of each unit configuration are given below. Page No. 14 of 69
Table 4.1-3 Train Configuration Standard Class Seating Capacity Train Capacity Crush Laden 12 car unit 572 1700 6 car unit + 6 car unit 556 1606 8 car unit+ 4 car unit 522 1646 8 car unit 364 1096 6 car unit 276 818 Red Amber Green Loading Table Key Heavily over crowded unit with insufficient capacity. Marginal capacity with higher levels of crowding than is optimal for the service time. Sufficient capacity or acceptable level of crowding given the service time (e.g. peak, shoulder). 240m Unit ConfigurationsC Loadings. Table 4.1-4 Train configuration 1 st Class Seating Capacity Standard Class Seating Capacity Standing Capacity High Peak Loading Shoulder Peak Loading Off Peak Loading 12 car fixed unit 48 572 1128 210% 105% 49% 8 car unit + 4 48 556 1090 car unit 216% 108% 50% 2 x 6 car units 48 522 1084 217% 109% 51% 3 x 4 car units 72 492 1074 244% 122% 57% Initial Conclusion for peak service operation tion. For high peak services both 8 car + 4 car and 2 x 6 car train configurations have a marginally lower capacity than the 12 car fixed configuration, however the level of crowding remaining within the acceptable range for this type of service. For the 3 x 4 car configuration the reduction in capacity caused by the additional cabs push the level of crowding into the marginally acceptable range and could become unacceptable as there is limited margin for further growth beyond 2026. Train loading based on uplifted TOC passenger numbers Off Core. In order to determine whether a 6 car unit can meet the demand requirements during the off, inter and should peak further analysis was undertaken using measured data of today s passenger numbers provided by the TOCs. It must be noted that passenger load measurements from today s services are not totally representative of future Thameslink passenger loads. In addition to organic growth of passenger numbers there will be transfer of passengers to Thameslink from other Page No. 15 of 69
services and also a merging of services currently provided by different TOCs. To make an allowance for this growth we have uplifted the data by 25%, 50% and 75%. Whilst this is a crude method of passenger load forecasting it is the best option available to us within the timescales. It should also be noted that systems like Planet or Rail Plan modelling are used to forecast peak demand whilst we are specifically interested in the off / interpeak loadings in this instance. Based on this data we have identified average passenger loads for the AM and PM high peaks, AM and PM shoulder peaks and off peak for service groups 1, 3, 4, 5, 6, and 7 (see section 4.3 for identification of service groups) outside the core. Service groups not assessed are those that only operate a peak service. For illustration purposes service group 1, Bedford to Brighton, is shown within this section of the report. Appendix 1 however contains the results for all service groups evaluated. Services from the key stations below operate north and south bound. The results presented below show the worst case direction for each station. Page No. 16 of 69
Service Group 1 Bedford Brighton Uplift on 2009-2026 AM High Peak Bedford Brighton (Service Group 1) 6+6 Car Units 25% 50% 75% Bedford 23 mins 32% 38% 44% Luton St Albans 13 mins 66% 79% 92% St Pancras International Farringdon City TL 20 mins 96% 135% 3 mins 50% 116% 162% 60% 135% 189% 70% 32% 39% 45% Blackfriars London Bridge 5 mins 12 mins 115% 139% 111% 133% 162% 155% Seats Used > 100 % East Croydon 80 100% 42 mins 80% 96% 113% 50 80% Brighton < 50% No information Uplift on 2009-2026 AM High Peak Bedford Brighton (Service Group 1) 8+4 Car Units 25% 50% 75% Bedford 23 mins 31% 38% 44% Luton St Albans 13 mins 65% 78% 91% St Pancras International 20 mins 96% 115% 134% Farringdon 3 mins 134% 50% 161% 60% 188% 70% City TL 32% 39% 45% Blackfriars 5 mins 115% 138% 161% London Bridge Seats Used 12 mins 110% 132% 154% > 100 % East Croydon 80 100% 42 mins 80% 96% 112% 50 80% Brighton < 50% No information Page No. 17 of 69
Uplift on 2009-2026 AM High Peak Bedford Brighton (Service Group 1) 12 Car Units 25% 50% 75% Bedford 23 mins 30% 37% 43% Luton St Albans 13 mins 63% 76% 89% St Pancras International 20 mins 93% 112% 130% Farringdon 3 mins 130% 49% 157% 58% 183% 68% City TL 31% 38% 44% Blackfriars 5 mins 111% 134% 156% London Bridge Seats Used 12 mins 107% 128% 149% > 100 % East Croydon 80 100% 42 mins 78% 93% 109% 50 80% Brighton < 50% No information Uplift on 2009-2026 Off peak Bedford Brighton (Service Group 1) 6 Car Units 25% 50% 75% Bedford 23 mins 15% 18% 21% Luton St Albans 13 mins 29% 35% 41% St Pancras International 20 mins 50% 60% 70% Farringdon 3 mins 61% 55% 73% 66% 86% 77% City TL 61% 73% 85% Blackfriars 5 mins 86% 103% 120% London Bridge Seats Used 12 mins 83% 99% 116% > 100 % East Croydon 80 100% 42 mins 50 80% Brighton < 50% No information Page No. 18 of 69
Uplift on 2009-2026 Off peak Bedford Brighton (Service Group 1) 8 Car Units 25% 50% 75% Bedford 23 mins 11% 14% 16% Luton St Albans 13 mins 22% 26% 30% St Pancras International 20 mins 37% 45% 52% Farringdon 3 mins 46% 41% 55% 49% 64% 58% City TL 46% 55% 64% Blackfriars 5 mins 64% 77% 90% London Bridge Seats Used 12 mins 62% 75% 87% > 100 % East Croydon 80 100% 42 mins 50 80% Brighton < 50% No information Conclusion. The results of the analysis confirms that the 2 x 6 car units and 8 + 4 car units can comfortably meet all peak demand levels. The analysis also shows that 6 car units cannot operate effectively in the shoulder peak because the relative loading would exceed that of a 12 car unit in the peak. However 6 car units can operate effectively during the off and interpeak albeit that loading levels would be reasonably high at key locations e.g. East Croydon and London Bridge experience all seats occupied conditions, but in our view this is acceptable for stations near the core and shows efficient use of rolling stock. A fleet made up entirely of 4 car units should not be considered further at this time because seating capacity is considerably reduced and a 4 car unit operating in the off and interpeak will experience unacceptably high loading. Page No. 19 of 69
4.2 4.2 ALTERNATIVE ROLLING STOCK CONFIGURATIONS Over and above the two base options of a fixed consist 12 car and a fixed consist 8 car a number of alternatives have been developed as a desk exercise in order to assess the relative operational merits. These options are:- 1) 12 car trains comprising an 8 car supplemented by a 4 unit 2) 12 car trains comprising two identical 6 car trains that might individually operate the shoulder and off peak services. 3) 12 car trains comprising 3x4 car units that can be coupled as 8 cars for other than the high peak services. For each of these alternatives there has been a need to examine length, weight, capacity, cost, dwell time and compatibility with PRM initiatives within the Thameslink project. Over and above all these has been a consideration of whether a change in concept could affect the viability of the envisaged 24TPH service. The findings are presented essentially by diagrams and tables; see Appendix 2 - Unit Configuration, from which conclusions can be drawn. To aid the reading of these it is essential to understand the pre-conceptions within the Thameslink project which emerged after many hours of conceptual studies. These basic tenets can be summarised as follows:- a) Trains were to be either 8 or 12 car fixed formation, the longer units being necessary for the high peak and the shorter units for the shoulder peak and off peak services. The 12 car fixed consist trains would only be required for about 3 hours per day and during the off peak periods would be stabled or undergo maintenance. Wherever practical, 12 car trains would not be deployed on off peak services in order to reduce vehicle mileage thus minimising wear on vehicles and track and optimising energy consumption. There was nothing particularly novel in the fixed 8 car concept except that most current 8 car trains comprise two 4 cars units. These 4 car units can be maintained in very compact depot buildings whereas fixed formation units, comprising either 8 or 12 cars require very long workshops. b) These 12 and 8 car fixed consist trains would have the PRM facilities (toilet / wheel chair position) located in one of the centre two vehicles. By stopping trains centrally it was thus possible, with a 40 m long platform hump, to provide level access to the doorways designated for PRM passengers regardless of train orientation. It had been shown possible to incorporate such humps at the middle 40m of each of the core stations. In order to minimise lateral stepping distances, such humps are best located on straight sections of the platform. A survey of Thameslink platforms had shown that the centre sections were more often than not straight and fortuitously this supported the concept of just one central PRM universal toilet per 8 or 12 car train. Page No. 20 of 69
c) Fixed train consists could be inherently more reliable by not needing to couple and uncouple and could have their traction systems configured on an equal split basis offering protection against single source failures. Methodology Each alternative train, i.e. 8+4, 2x6 and 4+4+4 configuration has been examined with these tenets, or the desires behind them, firmly in mind. Comparisons of the merits and demerits of these options have centred round the following considerations: Train Length Train weight Train capacity Train cost PRM provision Dwell time Train Performance The findings have been tabulated in Appendix 1 for each of 5 basic alternative train formations namely:- Illustration 1; the base case 8 and 12 car fixed formations as specified in the TTS. Illustration 2; 8 car plus 4 car concept for 12 car trains. Illustration 3; two 6 car units with one PRM universal toilet in the centre of each. Illustration 4; three identical 4 car units making one 12 car. Illustration 5; two 6 car units with a PRM universal toilet at the trailing end of each driving car. TRAIN Cabs within a train consist add to the overall train length. The 8+4 and 2x6 is 2m longer than a fixed 12 car consist and a 3 x 4 car more than 3m longer. Although the TTS specified length limit of 243m will be exceeded by both the 8+4 and the 2x6 consists, checks have been made by Network Rail and even the 3.2 m exceedance of the 3x4 car would be acceptable. However, there is a further TIIS requirement relating to the spacing between the drivers door and the rearmost passenger door. This should not exceed 237.5m The figure for the 12 Car fixed consist will be 234.145 and for the 8+4 or 2x6 configurations will be 236.299 i.e. acceptable. For a 3x4 configuration the figure is 238.45 which would be unacceptable. TRAIN WEIGHT A cab without a gangway adds an estimated 2 tonnes to a vehicle weight and if a through gangway is required this adds a further 1.5 tonnes. The percentage increase on the tare weight of an 8+4 or a 2x6 will be 1.3% without a gangway or 2.8% with a through gangway. The loss of seats due to cabs accounts for a reduction in the Page No. 21 of 69
passenger crush load of 28 passengers at 80kg = 2.2t thus offsetting part of the weight increase due to cabs. The decision on whether to opt for full or half width cabs must be take into consideration the impact that each has on the train in terms of cost, capacity, and operational flexibility. In making this decision the views of First Capital Connect, Southern, South Eastern and East Midland Trains were sought. The general consensus was that full width cabs would be preferred because the following benefits result: Lower weight. Easier application of crashworthiness. Increased seating capacity. Better driver environment. Easier accommodation and use of equipment and systems. As a result of the decision to eliminate gangwayed cabs the weight increase of additional cabs will be 4 tonnes per unit. It is considered that this weight increase should not require any rating uplift of the traction packages proposed for the train; the bidders will have made some provision for failing to meet their weight target and the desired, rather than minimum, passenger capacity. They are in the best position to run their simulation to ascertain what action, if any, needs to be taken and this could be a viable question to ask of them. TRAIN CAPACITY Although cabbed vehicles are longer than intermediate vehicles they consume in effect one row of seats and the standing space between them. This accounts for 4 seats and an estimated 4 standees. If a through gangway is included then it is assessed that the cab back wall on the driver s side will need to go back one more seat row losing two more seats. Although the detailed figures are in Appendix 1 it should be noted that for the 2x6 formation with through gangway a total of 24 standard and 4 first Class seats are lost which is an overall reduction in seating capacity of 4.5%. This includes the loss of standard class seats caused by having a PRM universal toilet in each 6 car unit. TRAIN COS OST Modern cabs contain a lot of sophisticated equipment and the two additional intermediate gangwayed cabs of the 2x6 consist and an additional PRM universal toilet adds between and per train. This is equivalent to an increase of between 9% and 16% over the base 12 car cost. For the 8 +4 this increase is between and which equates to an increase of between 10% and 17%. PRM PRM PROVISION In order to still deliver 24TPH through the core the only way to utilise the original concept of a central 40m hump with the split train concept is to fit the 2x6 car units with a PRM universal toilet at each end of each unit. See Illustration 5. This is not attractive since it loses 7.7% of total seating capacity. For the other 2x6 consist (illustration 3) there will be two PRM universal toilets per 12 car train and to align just one of these with the hump would need the hump to be much closer to the southern end of the platforms where the platform curvature at St Pancras Page No. 22 of 69
and Farringdon would present unacceptable lateral stepping distances. This would not suit an 8 car unit unless the hump was 60m long. For the 8+4 and 2x6 (Illustration 3) options, financial provision has been made within the financial model for either platform humps, manually operated ramps which are to be built into each unit at the PRM designated doorways (both sides) and improved telematics. Without further analysis and modelling it is not possible to determine the optimum solution for PRM boarders and alighters, it is clear thought that this issue has the potential to severely impact on dwell time and potentially compromise the 24tph service. DWELL TIME The concept of a pair of cabs in the middle of a train has been appraised using Legion modelling. The cabs do two things, they reduce capacity within the saloon area behind the cab and they create a slightly longer walk to the nearest doorway for passengers waiting in the exact middle of the platform (in the case of the 2 x 6 car consist). A further problem was foreseen insofar that if the two 6 cars become coupled First Class end to First class end then standard class passengers in the centre of the platform would be faced with an even longer walk. The Legion report is attached as Appendix 3 and basically concludes for the worst case loading at St Pancras am peak that the required dwell time of 27 seconds can still be met by a 2x6 consist but if the first class ends both occur at the centre a further 1 second is required. Though not specifically modelled it can be inferred that the 8+4 consist will also meet the required dwell times. The First/First coupling increase by 1 second would not occur with the 8+4 consist since it would seem wise in this scenario to only have standard class seating in the 4 car unit. TRAIN PERFORMANCE As presently configured by the bidders the 12 and 8 cars have split traction systems such that they are already 2x6 and 2x4 respectively. However they both have through train control and monitoring systems and no requirement to be able to multiple with each other except mechanically for rescue. All cab ends will require full auto couplers and the trains will need to be configured to automatically recognise the second unit after coupling and to monitor and control the whole train when coupled. Since the trains are not designed in detail yet, and their predecessors were configured for multiple operation it has been assessed that an average cost per cab of should cover this. Electrical heads for mechanical couplers have been advised by one manufacturer at max leaving for system reconfiguration. These figures have been used within the cost model and have then been subject to sensitivity analysis. Although once a source of unreliability, modern auto couplers are now offering acceptable levels of reliability. Examination of failure statistics for both Scharfenberg and Dellner couplers has substantiated this. Traction performance with an estimated increase in total train weight of 1.3% is considered to be only marginally adversely affected and bidders should be questioned on how to tweak their equipment to ensure the specified performance. A similar argument applies to braking performance. Page No. 23 of 69
As outlined above, dwell time has been demonstrated to be unaffected with coupled units but it should be recognised that neither the bidder s work nor this most recent work has attempted to quantify the effects of PRM passengers. This is not just wheelchair users since the modelling has focussed only on able bodied commuters and not those with luggage or children and other mobility reducing characteristics. This must represent a risk to the 24 TPH operation and the original concept of raised humps for the PRM designated doors, whilst it may have helped ingress and egress for wheelchairs and PRM, still leaves over 80% of the doorways with far from level access. It is incumbent upon the operator to manage dwell time and if some combination of manual ramps and humps are to be deployed then the application of telematic systems should be able to aid platform staff to position themselves to assist those who need help. Devices as simple as a RADAR key operated switch adjacent to each wheelchair position that lights a light on the outside of the car just prior to the destination station could help platform staff to see when and where they are needed. Station CIS might also be configured to advise where the PRM doorway will occur for any arriving train. First class passengers could well expect to be advised where their accommodation is; indeed they might expect it even with the 8 or 12 car operation. Conclusions The only real problem with 8+4 or 2x6 configurations, given that the extra cost can be justified, is the management of PRM. Although the centre platform hump went a long way to address this, no work has been done with Legion to confirm or otherwise that 24TPH is practicable with a given percentage of PRM passengers. The manual devices that might be deployed, at a cost, on the 8+4 or 2x 6 consists could be just as effective on the fixed consists and would have the benefit of working throughout the network without recourse to infrastructure humps or local mobile ramps. It is suggested that bidders are asked to quote for an integral manually operated device at the PRM designated doorways, with an option to offer a powered device if practicable. We could then use Legion or other techniques to identify the best way to protect the 24TPH objectives. Page No. 24 of 69
4.3 4.3 OPERATIONAL EVALUATION Background The current V1e unit diagrams were originally produced by X, X in September 2008 in connection with the stabling and depot work for the Thameslink Project. The diagrams were compiled against a service specification for a standard peak and off peak hour in the core section of the route between St Pancras International and Blackfriars. In addition, the timetable was expanded either side based on comparative sectional running times between the stations destined to be served by the new Thameslink services. It should be noted that as far as we are aware, no validation of the timetable paths has been undertaken. Validation will be subject to the usual train planning rules and processes in conjunction with other train operating companies. The outline timetable as devised by FCC forms the basis of this report. This timetable was used by FCC as a basis to develop the diagrams. The timetable was developed using existing sectional running times of existing rolling stock operating on the routes in question together with station dwell times in accordance with the Network Rail Rules of the Plan (ROTP). In addition, robust turnaround times were incorporated at the start and end locations to link in with return workings. A major consideration in the construction of the diagrams was to ensure overnight balancing of units took place at depot and stabling locations in connection with the proposed Depot and Stabling Plan, but also, importantly, to optimise the fleet size in an attempt to reduce the overall fleet requirement from the original plan for approximately 1275 vehicles; V1e reduced this requirement to 1116 vehicles required for daily service in either 8 or 12 car fixed formations. In addition, FCC were instructed to diagram units to dedicated service groups throughout the day if possible but with the avoidance of creating inefficient diagrams; this consideration was incorporated for reasons relating to performance. The basis for unit allocation ensured that in both the morning (0745-0915) and afternoon (1645-1815) high peak 90 minute periods, based on the current FCC peaks at St Pancras International, full length trains of 12 cars were provided. However, on service group 4 (St Albans-Orpington/Sevenoaks) and service group 8 (Royston- Maidstone East), the maximum formation of 8 cars applies due to platform length limitations en-route. There was also a requirement to provide at least 8 car formations in the shoulder peak periods either side of the 90 minute high peak. This resulted in at least 8 cars being provided on every service operating through St Pancras International between 0700-0959 and 1600-1859, these being the standard peak periods under the railway industry PIXC (Passengers In Excess of Capacity) standards applicable to the London suburban train operating companies. The resultant diagram set V1e comprised 110 unit diagrams, 51 x 8 car and 59 x 12 car (1116 vehicles). The 8 car formations are further divided into Inner Suburban (15 sets) and Outer Suburban (43) variants. It should be noted that the numbers quoted are the units/diagrams required to operate the service; maintenance or spare cover will be over and above this. Page No. 25 of 69
All the diagrams, without exception, operate through the core section between St Pancras International and Blackfriars at some point during the morning or afternoon peak periods. Due to the limited time available for FCC to develop the unit diagrams, very few empty coaching stock (ECS) movements were included in V1e. Assumptions were made regarding the start and finish depot/stabling point locations depending on the start and finish locations of the first and last passenger carrying services. Services The train service specification is split into the following 8 service groupings with unit resources as allocated in V1e: 1 Bedford to Brighton 2 Bedford to Paddock Wood & Tunbridge Wells (Peak only) (38 diagrams, service group 1&2 combined) 3 Luton/Brent Cross to East/West Croydon (14 diagrams) 4 St Albans to Sevenoaks & Orpington (8-car only) (15 diagrams) 5 Peterborough to Horsham (12 diagrams) 6 Peterborough to East Grinstead (Peak only) (10 diagrams) 7 Cambridge to Three Bridges (11 diagrams) 8 Royston to Maidstone East (8 car only) (10 diagrams) Other than those service groups specified above as operating with maximum 8 car formations, all other service groups contain a mix of 8 and 12 cars: Table 4.3-1 Service Group 8 Car Units 12 Car Units 1&2 11 27 3 4 10 4 15 5 6 6 6 4 6 7 1 10 8 10 Total 51 59 Service Groups 1&2 are combined due to a significant number of those diagrams operating on both service groups each day. Timetable and Diagram Analysis In order to gain a measure of the efficiency of the unit diagrams in V1e, it was first necessary to identify all passenger carrying services and form these into a separate listing. Page No. 26 of 69
Our initial approach was to diagram formations to specific service groups as far as possible but also being prepared to cycle diagrams between service groups if this enabled efficiencies to be made. The 8 separate service groups do not lend themselves well to the operation of cyclical diagrams as most start and end locations are a large distance apart, often on completely different routes. The only location where this is possible is either Peterborough or Bedford where more than one service group start or end their journeys. However, as only one service group at each of these locations operates throughout the day, the others operating peak time only, the only benefit to be achieved in cycling diagrams is replacing a longer formation with a shorter one or to reduce turnaround times. Consideration was given to the benefits of swapping formations between service groups to better match capacity to demand. This produced large inefficiencies as it would require two ECS movements over some distance to take place simultaneously in order to swap formations and pick up the naturally occurring return working. This also requires additional crew resource and could import a performance risk should one of the workings be delayed. The timetable devised means that a high level of efficiency can be achieved by allocating rolling stock to diagrams dedicated to specific service groups that operate throughout the day; turnaround times at destination are such that very efficient use can be made of the rolling stock. However, it is apparent that large inefficiencies arise when allocating rolling stock to the service groups that only operate during peak periods; Peterborough- East Grinstead, Bedford-Paddock Wood/Tunbridge Wells. The service specification and end to end journey times are such that a large number of vehicles are required to operate these services, the majority of which are stabled during the day waiting to form later peak workings. As the all day service groups work extremely well using self contained diagrams, there is no benefit achieved in swapping out these formations with others in the self contained groupings apart from a few at Bedford or Peterborough. After allocating rolling stock resource to specific service groups so that efficiencies in diagrams could be achieved, a list of other services remained to link together. This was largely achieved but did require some integration and swapping between service groups, primarily at Bedford, in order to achieve the optimum set of diagrams. Following the above analysis and actions, it became apparent that our ultimate aim of reducing the number of diagrams, and therefore a possible reduction in rolling stock resource, could not be achieved as we could only match the 110 diagrams in V1e; we could not achieve a reduction. In addition, the revised diagram set we devised was produced solely to seek a reduction in unit/vehicle requirement and did not take into account depot balancing requirements as occurred in V1e. It was deduced therefore that as we could not achieve a saving in rolling stock resource, the use of V1e would achieve the ultimate in diagram efficiency. Page No. 27 of 69