Appendices. July Network Rail Scotland Route Study 89

Transcription

1 Appendices Network Rail Scotland Route Study 89 Appendix 1 - Long Term Planning Process 90 Appendix 2 - Scotland Market Study 95 Appendix 3 - Freight Market Study Conditional Outputs 153 Appendix 4 - Long Distance Market Study Conditional Outputs 155 Appendix 5 - Cross-Boundary Analysis 158 Appendix Option Identification and Development 159 Appendix 7 - Appraisals - CP6/CP7 Choices for Funders 199 Appendix 8 - Glossary 206

2 Appendix 01 Network Rail Scotland Route Study 90 The Long Term Planning Process Background to the development of the Long Term Planning Process In June 2005 the Office of Rail Regulation (ORR) 1 modified Network Rail s network licence to require the establishment and maintenance of Route Utilisation Strategies (RUSs), for the use and development of the network consistent with the funding that is, or is likely to become, available. This modification to the Network Rail network licence followed the rail review in 2004 and the Railways Act The RUS programme, led by Network Rail on behalf of the industry, started in late 2004 and culminated with the publication of the establishment of the West Coast Main Line RUS in August As the network licence requires the maintenance of RUSs, the completion of the initial programme of geographic RUSs gave the opportunity to review how best to discharge this requirement in the future. The review took into account: changes in administrations in England, Wales, and Scotland, together with very significant changes in planning policy long term strategic investments in the rail network, examples include the development of a high speed line between London and Birmingham and beyond to Leeds and Manchester (HS2) decisions to install electrification on significant route mileage of the rail network changes to signalling technology through deployment of the European Rail Traffic Management System (ERTMS) and progression of the Network Rail Operating Strategy a need to inform maintenance and renewal strategies for the rail network changes to funders objectives in the light of the significantly tighter fiscal environment, including a clear policy shift towards revenue generation and making best use of the existing railway the conclusions from the Rail Value for Money report by Sir Roy McNulty in May The need was clear for the industry to consider network-wide long term infrastructure development, rather than as now plus isolated enhancements, to the rail network. Network Rail and the industry worked together to develop a revised methodology to the RUS process to take the changes set out above into account, to continue to develop the long term strategic direction of the rail network. This successor programme, the Long Term Planning Process (LTPP) was endorsed by the ORR in April The LTPP will consider such changes and is designed to enable the industry to take account, and advantage, of long term strategic investment being made in Great Britain s rail network. The planning horizon for the LTPP is over a 30-year context. This clearly involves uncertainties, however, the approach is intended to adapt to potential structural changes in the economy, and the approach to social and environmental responsibility, so that the rail industry can respond to change over the long-term life of the assets used to operate the rail network. Due to the uncertainties of a 30-year horizon, the LTPP will be iterative; future planning cycles will enable an updated view to take into account the changing context and requirements of the industry and economy. An objective of the LTPP is to understand the longer-term context whilst creating a prioritised view of requirements for the next Control Periods ( ), in order to present a clear strategy for funding as part of the industry process. This commences with the submission of the Industry Advice later in Future iterations of the LTPP will evolve, identifying requirements for future Control Periods as part of this on-going process. 1 The Office of Rail Regulation changed its name from 1st April 2015 to the Office of Rail and Road

3 Appendix 01 Network Rail Scotland Route Study 91 LTPP Governance Arrangements The Long Term Planning Process is designed to be as inclusive as possible with contributions encouraged both from the rail industry and wider stakeholders. Overall governance responsibility for the process lies with the Rail Industry Planning Group (RIPG) whose membership comprises: Transport Scotland Freight Operators Network Rail Office of Rail and Road Transport Focus Passenger Transport Executive Group Rail Delivery Group Rail Freight Group Railway Industry Association Rail Freight Operators Association Rolling Stock Leasing Companies Department for Transport London Travel Watch Transport for London Welsh Government. RIPG meets once every two months and provides strategic direction and endorsement of the constituent publications of the LTPP process. The LTPP consists of a number of different elements, which, when taken together, seek to define the future capability of the rail network. These elements: - Market Studies, Route Studies, Cross- Boundary Analysis and the Network RUS are considered in more detail. Market Studies In October 2013, Network Rail published four Market Studies: Long Distance passenger, London & South East passenger, Regional Urban passenger and Freight. All four have been established by the ORR and are available on the Network Rail website. The three passenger Market Studies have clear connections to the three sectors in which passenger train services are often divided. It is important to emphasise that each Market Study considers a particular market, rather than a particular set of train services. The passenger Market Studies have three key outputs: identification of the long term strategic goals which define the successful provision of rail services to each of the three market sectors. These are based on the aspirations of current and likely future industry funders Demand forecasts for the sector, over a 10 and 30-year planning horizon. s are used to reflect key uncertainties, where appropriate conditional outputs for the sector. The conditional outputs are aspirational levels of service (in terms of, for example, frequency journey time and/or passenger capacity on key flows in the sector). The conditional outputs reflect stakeholder views of how rail can support delivery of their strategic goals, and opportunities created by planned investments, as well as reflecting current service levels and forecast future demand. The aim of the Market Studies is to provide demand forecasts, and conditional outputs, that are consistent across the Route Studies. For freight the conditional outputs are intended to meet the forecast level of freight set out in the Freight Market Study in 2023/24 and The Freight Market Study produced demand forecasts over a 10 and 30 year planning horizon, with preferred routeing of services and the implied requirements in terms of network capacity and capability. s were used to reflect key uncertainties in particular markets. Conditional outputs should be viewed as aspirations for the future rather than recommended investment decisions.

4 Appendix 01 Network Rail Scotland Route Study 92 It is also important to state that the conditional outputs shown are dependent on affordability, fundability, and a value for money business case for current and potential future rail industry funders being made for any interventions that this, and subsequent Route Studies, in the LTPP may consider as a way to deliver them. Equally the conditional outputs will need to be technologically, operationally and physically deliverable. Route Studies There is generally one Route Study for each of Network Rail s devolved routes. The full Route Study programme setting out current timescales for each of the forthcoming Route Studies can be found on the Network Rail website. The Scotland Route Study has undertaken a Market Study for passenger demand wholly within Scotland and incorporates relevant outputs from the Long Distance Market Study and Freight Market Study. A Route Study develops and assesses choices for the long term use and development of the network. Its starting point is to determine whether the conditional outputs from the relevant Market Studies can be accommodated on the existing network, with committed enhancements. It then develops train service options, corresponding to different uses of the network, (and hence to different trade-offs between stakeholders strategic goals. Only then will consideration be given to choices involving infrastructure investment. These choices were assessed against funders decision making criteria. This includes quantitative assessment as in the previous RUS process. It will also, where appropriate, include a wider assessment against factors such as strategic fit, wider economic impacts and affordability. Choices for Funders identified within this and other route studies, are intended to inform the development of proposals to consider within rail industry funding discussions for Control Period 6/7. Equally, other potential rail industry funders, for instance Local Authorities, may wish to consider the information this Route Study contains, when taking forward their own plans and proposals which may impact upon the rail network. Route Study Governance Arrangements A three tier structure for rail industry and wider stakeholder dialogue has been established to oversee and help produce this Scotland Route Study. First, a Route Study Board, chaired by Network Rail with senior level representation from passenger and freight train operating companies, Rail Delivery Group, Transport Scotland and the Office of Rail and Road provides a high level review and a forum to resolve any significant issues which the Working Group wish to remit to the board for decision. Second, a Working Group, chaired by Network Rail had a mandate to discuss the study on behalf of the rail industry with other stakeholders and a review of the ongoing work to develop them. The Working Group is where stakeholders met to determine how the conditional outputs from the Market Studies could be accommodated, including identification of service specifications and options with the aim of developing choices to funders for Control Period 6/7 and for 2043 through publication of the Route Study. The Working Group comprises representatives from the current Train Operating Companies (both passenger and freight) who operate on the route, Rail Delivery Group, Transport Scotland, Network Rail, and the Office of Rail and Road as an observer. Thirdly, a Regional Working Group convened and chaired by Network Rail, provided location specific oversight as well as an opportunity to collaborate in the production of the Route Study with wider stakeholders. The Regional Working Group membership comprised Regional Transport Partnerships and some Local Authorities. Network Rail managed the development of the work through an internal Technical Working Group to deliver the information necessary to support the deliberations of the Working Group. Where industry input has been required, this has been augmented by attendance or discussions with rail industry stakeholders.

5 Appendix 01 Network Rail Scotland Route Study 93 Scotland s Route Study Governance Arrangements Regional Working Group Route Study Board Route Study Working Group Route Study Technical Working Group Cross-Boundary Working Group Wider Stakeholders Figure shows the governance arrangements. Cross-Boundary Analysis Services that run across more than one Route Study area are considered in a separate cross-boundary workstream. This workstream led by a cross-boundary working group has developed and assessed options for cross-boundary services (passenger and freight) to deliver the connectivity conditional outputs articulated in the Market Studies. The output from the cross-boundary analysis is a cross-boundary indicative train service specification which provides a set of common assumptions that Route Studies should adopt regarding cross-boundary services. Network RUSs In addition Network Rail facilitates the production of Network RUSs. These strategies look at network-wide issues and look to future capacity and technology related issues for the railway. Current studies being undertaken under the auspices of the Network RUS include:- Network RUS: Electrification: An Electrification Route Utilisation Strategy was established in 2009 and had a number of the electrification schemes proposed within that have been funded for delivery. In the future, Network Rail aims to issue a new Electrifictation Study to consider future options for electrfication in the longer term following the completion of publiciy committed schemes. The study will look at the case for further opportunities to develop the electrified network in Control Period 6/7 and beyond. Network RUS: Freight: The Freight RUS was established in 2007 and a number of its recommendations to develop the Strategic Freight Network have been implemented. The Network RUS: Freight is looking at the future capability requirements of the network to accommodate freight growth and will consider: longer and heavier trains efficient operating characteristics and freight network performance 7-day and 24-hour capability including diversionary routes and resilience loading gauge strategic rail freight interchanges and terminals axle load freight train journey time and maximum speed. When considering capacity the Network RUS: Freight will primarily focus on capacity for cross-boundary freight flows. Where different commodities on a corridor have specific requirements this will be identified. It will bring together Route Study strategies on major cross-boundary freight programmes, for example: Anglo Scottish traffic Felixstowe to Nuneaton Southampton to the West Coast Main Line Northern Ports and Trans Pennine Freight Study South West and Wales to the Midlands. Route specific capacity issues are being considered by the individual Route Studies. Network RUS: Interoperability: the Department for Transport (DfT) has requested Network Rail to lead a cross-industry review of Interoperability on the UK rail network. The outcome of this review is expected to be the publication of the Network RUS: Interoperability in The aim is to establish this RUS in sufficient time to inform the Initial Industry Advice in late The content of the RUS will complement the Scotland and other Route Studies. The Railways (Interoperability) Regulations 2011 and associated Technical Specifications for Interoperability (TSI) apply to the entire UK rail network with the exception of the exclusions defined on the DfT web-site 1. Network Rail, along with other Infrastructure Managers in the UK, is legally obliged to comply with the Interoperability Regulations when the nature of the works being undertaken so requires. 1

6 Appendix 01 Network Rail Scotland Route Study 94 European and UK legislation defining objectives for Interoperability and the Trans European Transport Network (TEN-T) will be taken into account in the development of this Route Study. Network Rail and the wider rail industry have sound practical experience in applying the respective regulations and associated TSIs. The experience has been used to good effect to: demonstrate legal compliance with the requirements and provide feedback to government and the European Railway Agency on practical issues of application leveraging the benefits associated with the Interoperability principles developing plans to assess the full potential of an interoperable network, including connectivity with continental Europe. There are a number of established Network RUSs that are considered as part of the development of this Route Study, these include: Network RUS: Passenger Rolling Stock: This strategy explores the potential for greater efficiency in the purchase of new rolling stock to replace the existing fleet and accommodate growth in demand. The RUS identified the potential for significant economies of scale through procuring a smaller, standardised range of stock types targeted at specific market sectors. Network RUS: Stations: This strategy provides guidance about potential interventions to relieve congestion at stations. Network RUS: Alternative Solutions: This strategy describes the circumstances in which it may be appropriate to consider the operation of alternative forms of public transport than conventional diesel or electric heavy rail. This includes the use of tram trains, the conversion of sections of the network for use by trams, the use of battery powered vehicles, hybrid light vehicles, personal rapid transit (such as the pods which serve business car parks from Terminal 5 at Heathrow), bus rapid transit and guided bus. It also considers the circumstances in which discontinuous electrification and coasting could reduce the cost of electrification schemes, and when the designation of a route or service to community rail status could increase patronage on the railway in a cost effective manner. Further information on the Long Term Planning Process, can be found on Network Rail s website.

7 Appendix 02 Network Rail Scotland Route Study 95 Scotland Market Study Modelling Documentation - Background Background to Scotland Market Study and Definition of Markets The Long Term Planning Process (LTPP), described in Appendix 1, is the rail industry s planning cycle, of which the Scotland Route Study is an important element. The Route Study programme is the successor to the programme of Route Utilisation Strategies (RUSs) that informed the enhancement programme up to and including Control Period 5 (CP5). In regulatory terms, the Route Studies contribute to satisfying Network Rail s obligation under the Network Management and Planning terms of its Network Licence to: plan the means by which it will [ ] over the short, medium and long term [ ] meet reasonably foreseeable future demand for railway services. The key difference between the LTPP and the RUS programmes is that the LTPP covers a far longer (30 year) timeframe than the RUSs. The reason for this extended forecasting window is that an incremental approach to infrastructure planning is unlikely to deliver the network that long term passenger growth may require in an efficient way given the strong growth that has been experienced since the mid-1990s. The LTPP takes the view that taking a longer, 30-year window - which is consistent with infrastructure asset lives - is the appropriate way to plan investment in railway infrastructure. The initial outputs of the LTPP so far have been four Market Studies covering the key market served by the GB rail network, together with the Route Studies. The Market Studies published to date are as follows: Regional Urban Market Study Long Distance Market Study London and South East Market Study Freight Market Study The Scotland Route Study draws upon evidence from the Long Distance and Freight Market Studies. Domestic demand in Scotland was not covered by the Market Study programme. As a result of this, a Scotland Market Study has been produced to provide the demand forecasting evidence to support the Scotland Route Study. The passenger markets considered are: Morning Peak Commuter Interurban Rural. This document describes the methodology employed to produce demand forecasts for these markets, including: assumptions made sources of data processes used to validate the forecasts. Scotland Market Study Demand Forecasting Outputs The purpose of the Route Study programme is to provide options to funders in order to inform and influence the content of both the Industry Advice and the High Level Output Specifications (HLOSs) issued by the Department for Transport and Transport Scotland. The remit for the Scotland Market Study is to develop demand forecasts for the final year of CP6 (2023/24) and for twenty years beyond this (2043). The remit agreed by the Working Group was as follows: Establish an estimate of actual rail demand for 2014 Develop baseline forecasts for 2012 and 2018, reflecting current rail investment plans for CP5 committed obligations for the new ScotRail franchise and Transport Scotland s road investment plans Produce scenario-based forecasts for 2023/24 and 2043 drawing on the framework and evidence developed for the Regional Urban Market Study. The forecasts were developed to (a) set capacity-related Conditional Outputs where network capability is unlikely to meet the requirements of the market and (b) inform the appraisal of schemes developed to meet these Conditional Outputs.

8 Appendix 02 Network Rail Scotland Route Study 96 Passengers Weekday Passengers Saturday Passengers Sunday Passengers Figure 2.1 Profile of demand throughout the day at Glasgow Central Morning Peak Commuter Markets The market where train capacity is likely to be most constrained is the Morning Peak Commuter market. This section sets out the approach to forecasting demand in the Morning Peak Commuter market. For the purposes of the Route Study, the Morning Peak Commuter markets are defined as existing in and around cities where there are significant commuting flows. The Working Group agreed that for the purposes of the Scotland Market Study the Morning Peak Commuter market should be defined as consisting of: Glasgow Edinburgh Aberdeen. Using Figure 2.1 as an example, commuter demand tends to be more concentrated in the morning peak than in the evening peak, the morning peak is therefore where any capacity constraint is likely to initially apply. It is for this reason that demand in the morning peak has been modelled. One output of the Morning Peak Commuter market element of the Scotland Route Study is the identification of current and future capacity constraints on the key routes into city centres. It therefore needs to know which services are likely to become capacity constrained and which year the crowding constraint is likely to first apply. Interurban Markets For the purposes of the Route Study, Interurban markets are defined as the demand for travel between the seven cities in Scotland (Glasgow, Edinburgh, Aberdeen, Inverness, Perth, Dundee and Stirling). The key journey purposes in this market are business and leisure travel, but commuting will also be important. The routes covered will be as follows reference Figure 2.2: Glasgow-Aberdeen via Stirling, Perth and Dundee Glasgow-Inverness via Stirling and Perth Edinburgh-Inverness via Perth Edinburgh-Aberdeen via Dundee Aberdeen-Inverness The scope of the Morning Peak Commuter market demand study means that there is a potential overlap with the Interurban market. It is important to note, however, that the purpose of the Morning Peak Commuter market demand forecast is to identify routes where capacity is constrained in the morning peak because of commuting demand. With the Interurban market, demand is less peaked and the factors that are likely to influence investment are broader in scope than they are in the Morning Peak Commuter market. So, while on-train capacity is an issue in the Interuban market, other factors such as journey times and opportunities to travel may be equally important. Some corridors serve both the Interurban and Morning Peak Commuter markets. Where this is the case, both demand forecasts will be applied based on the balance of 2014 revenue. Rural markets The markets where capacity is most likely to be most constrained in terms of infrastructure capacity are the Morning Peak Commuter and the Interurban markets. However, although demand in Rural markets tends to be significantly lower than in other markets, it can also be subject to significant variation depending on the time of day, week or season that the service is operating. Localised demand can also be significant for instance schools traffic or localised employment while capacity is often constrained by asset capability. In terms of journey purpose, leisure travel is the primary driver of rail use but commuting will also be important in some markets. Extended journey times on many rural routes mean that rail often cannot compete with road on journey times and car ownership tends to be higher in rural markets than it is in urban markets. These factors point towards a bespoke approach being suitable for rural

9 Appendix 02 Network Rail Scotland Route Study 97 routes than for Morning Peak Commuter and Interurban Markets. For the purposes of the Route Study, rural markets are defined as the following routes: Glasgow to Oban/Fort William/Mallaig Glasgow to Stranraer Glasgow to Carlisle via Dumfries Inverness to Kyle of Lochalsh/Wick/Thurso. Inverness Aberdeen It was originally intended to treat the Borders Railway between Edinburgh and Tweedbank as a Rural Market, but this has instead been considered as part of the Morning Peak Commuter Market in order to reflect the passenger numbers experienced to date. Perth Dundee Glasgow Edinburgh Figure 2.2 Diagram of the Interurban network in Scotland

10 Appendix 02 Network Rail Scotland Route Study 98 Establishing a picture of actual demand in 2014 City Glasgow Edinburgh Aberdeen Gatelined City Centre Stations Glasgow Central (High Level and Low Level) Glasgow Queen Street (High Level and Low Level) Argyle Street Charing Cross Waverley Haymarket Aberdeen Table 2.1 Gated stations in key urban areas As with the other Route Studies carried out throughout CP5, the most recent demand year at the date the study commenced (2014) has been used as the point at which base demand is estimated. In order to estimate base year demand, three potential approaches were identified each with advantages and disadvantages: MOIRA 1 demand forecasts for 2014 (based on 2014 LENNON 2 ticket sales data) Ticket gateline count data On-train count data. MOIRA demand forecasts MOIRA is the industry demand forecasting model, the base data of which is derived from the LENNON ticket sales database. LENNON records nearly all tickets sold on the rail network in Great Britain with the exception of certain operator-specific tickets and some types of promotional fares. Given these are low in volume, the LENNON database can be treated as being a valid representation of origins and destinations across the network, albeit at an extremely aggregate level. The advantage of using MOIRA is that it has universal coverage by route and by train. It is also based on actual ticket sales so it should be accurate at a population level. The disadvantages of using MOIRA are as follows: MOIRA is not a primary data source. It is itself a model based on LENNON ticket sales. As such, it is subject to modelling error The size of the London and South East commuter market (where, for reasons of crowding, demand is pushed into the shoulder peak) arguably distorts the results that MOIRA predicts for services outside the South East. Demand in Scotland tends to be peakier than MOIRA predicts 1 MOIRA is an industry demand forecasting tool which applies LENNON base data to individual train services. 2 LENNON is the rail industry ticket sales database, which reflects most journeys that are made on the GB rail network. MOIRA does not reflect demand resulting from some types of operator-specific ticket sales, principally the multi-journey Flexi Pass sold on the Edinburgh-Glasgow via Falkirk High route. Ticket gateline count data Transport Scotland has funded the installation of ticket gatelines at most of the key stations on the network in Scotland, primarily for revenue protection purposes. As it stands, approximately 80 per cent of journeys within Scotland have a gatelined station at one end of the journey, and this is likely to be far higher in the urban markets. The ticket gatelines capture a breadth of travel data including the journey origin, destination, time at gateline and ticket type (peak, off-peak, season). They do not assign any of these journeys to individual trains. Nearly all stations defined as city centre stations in and around Glasgow, Edinburgh and Aberdeen have had gatelines installed. The exceptions to this are High Street, Partick and South Gyle. Edinburgh Waverley is partially gated: the long distance platforms are not gated, but most platforms served by domestic traffic are. Table 2.1 lists the gated stations within each of the city centres. Gateline data has four principal limitations as far as estimating base demand is concerned. They are: Gatelines will only record limited origin data on other ticket types (e.g. Strathclyde Partnership for Transport (SPT) Zone Cards) and, with regard to season tickets, there may be some instances (particularly in Glasgow), where origin stations being located close to one another may mean that a season ticket at one station may well be used frequently on other services. Gatelines will not pick up origins where the destination is not a city centre station (e.g. South Gyle to the west of Edinburgh). In addition, gateline data does not pick up failures with the ticket or with the gateline equipment. However, these failures are not thought to be correlated with the journey taken, and can potentially be either stripped out completely or repaired using the data that does exist.

11 Appendix 02 Network Rail Scotland Route Study 99 At Edinburgh Waverley the long distance platforms are not gated but are staffed. This is problematic as suburban services also use these platforms during peak periods and long distance services are an important source of passenger capacity during peak periods. On-Train count data On-train count data provides exactly the right type of data for this study. On-train counts record passenger loadings on specific trains at specific points on their journey. ScotRail carries out on-train passenger counts using the following two methodologies: Automatic counts, which are carried out on the Class 380 trains. The only route where these trains are diagrammed 100 per cent of the time is the Edinburgh-North Berwick service and may also be useful for many (but not all) Ayrshire Coast services Manual on-train counts, which are carried out on an annual or biannual basis. These counts are relevant for all services. However, across most of the network on-train counts are only carried out periodically (once or twice a year), so they cannot be considered to be statistically robust. Some automatic train counts are carried out, recording loads on every train, but the coverage of these counts is poor at present. Approach followed The coverage and quality of both the on-train count data and the gateline were adjudged to be of insufficient quality to form the basis of base demand for all corridors. Therefore, the decision was made to use MOIRA/LENNON as the primary source of data for estimating baseline (2014) demand. Where the quality of on-train counts and gateline data was deemed to be good, these data sources have been used to validate MOIRA/ LENNON results, particularly in profiling the peak an area where MOIRA is known to have limitations and to capture the full range of tickets issued on the Scottish network. This has enabled the 2014 baseline to be validated by ScotRail. Adjustments to baseline demand The consultation process identified two areas for improving the base data. The first area was that rail journeys undertaken on multi-modal Zonecard tickets promoted by (SPT) are not included in the LENNON database. An exercise was therefore undertaken based on assigning the total number of rail journeys to corridors using SPT travel diaries. The second area of improvement relates to the Edinburgh-Glasgow via Falkirk route, where MOIRA differed significantly from the count data. LENNON does not include the ScotRail Flexipass ticket that is a popular product on that route. Because the count data was strong on this particular route, it was preferred to the MOIRA estimate on this corridor.

12 Appendix 02 Network Rail Scotland Route Study 100 Modelling demand growth High level modelling approach After considering a number of potential alternatives, the Working Group supported the development of a specific model to examine the Morning Peak Commuter and Interurban Markets in Scotland and to use the ScotRail demand forecast as the basis of the Rural Market forecast. The decision to construct bespoke models to estimate demand in the Morning Peak Commuter and Interurban markets reflected: Consistency of approach across Market Studies: The aim of the Market Studies has been to present evidence on demand as consistently as possible across the network independent of the identity of the primary funder Transparency - As the Market Study is part of an industry process, the modelling that informs it has to be transparent in terms of the assumptions made so that it has credibility within the industry. All models require careful interpretation, and transparency of what assumptions have and have not been made is important in enabling this Flexibility The 30-year time frame of the Market Study means that the forecasting approach is scenario based. The modelling had to be flexible enough to allow for carrying out a large number of scenario runs Using the best data available to reflect a rail-specific demand forecasting requirement. This has utilised the wealth of information contained within TMfS 1 /CSTM 2 whilst deviating from it in certain areas where it was necessary to do so (e.g. city centre parking). The modelling framework adopted follows a market size/market share approach developed for the Regional Urban Market Study (RUMS). The key features of this approach are as follows: Market size model This estimates the size of the total market for travel in Scotland based on the key drivers for growth that were identified in the Regional Urban Market Study(RUMS). For the Morning Peak Commuter market, the RUMS identified city centre employment growth as being the principal driver of the overall demand for travel into city centres in the morning peak. In the Interurban market, the evidence in the Long Distance Market Study (LDMS) identified population growth as being the principal driver of overall demand for travel. Total market size is estimated for each origin-destination pair (although in the Morning Peak Commuter model, only city centre destinations are selected). Market share model The market share model estimates rail market share for each origin and destination pair based on the relative attractiveness of rail relative to other modes. Generalised cost is used as a proxy for the perceived attractiveness of each mode. Market shares are estimated using a hierarchical logit approach Due to the size of the TMfS/CSTM data sets approximately 0.5 million lines of data in TMfS and 1.1 million lines of data in CSTM), the model could not be constructed in Excel. It was instead constructed in Access Key Assumptions The main data source for the Morning Peak Commuter and Interurban models is TMfS and CSTM respectively. Cost and demand skims were supplied by Transport Scotland s consultants (Jacobs) for 2012 and Forecasts were completed for 2012 and 2018 based on this data, with scenarios being applied thereafter. The data received from Jacobs reflected a number of infrastructure and policy assumptions. 1 TMfS is a multimodal transport model covering most of Scotland 2 CSTM is a multimodal transport model covering the Central Belt of Scotland

13 Appendix 02 Network Rail Scotland Route Study 101 The infrastructure assumptions are as follows: Assumed Infrastructure Committed Road schemes are: M8 completion A9 Upgrade A90 Balmeddie to Tipperty Dualling Aberdeen Western Peripheral Route (AWPR) Queensferry Crossing (Forth Replacement Crossing) M74 Raith Interchange Inveramsey Bridge Improvement. Committed Rail schemes are: Borders Rail completed 2015 Aberdeen to Inverness Phase 1 Highland Main Line Phase 2 EGIP (Key outputs 1, 2, 3 and 4) Glasgow to Paisley service improvements. Rail Fare Assumptions The rail policy assumptions relate to fares, as per the ScotRail franchise agreement. They are: Peak fares: increase in line with inflation as measured by the Retail Price Index (RPI) Off-peak fares: increase in line with inflation, minus one per cent (RPI-1 per cent). Bus Fare and Journey Time Assumptions Bus fare and journey times are as per TMfS/CSTM for 2012 and Assumed Car Use Costs Generalised Journey Times (GJTs), Car distances, and rail/bus fares were provided by Transport Scotland/Jacobs from CSTM and TMfS. Rail and bus GJTs and fares were estimated by Jacobs. Rail and bus GJTs included in-vehicle time and access/interchange times. All other explicit assumptions are derived from the Scottish Transport Appraisal Guidance (STAG) Technical Database Economy Spreadsheet ( These consist of: Vehicle Operating Costs (VOC) Fuel-related car costs: Average Vehicle Fuel / Energy Cost Formulae Parameter Values (p, 2010 prices): Work/Non-Work (Excluding VAT) Non-Fuel car costs: Forecast Non-Fuel Resource Vehicle Operating Costs (2010 prices) Values of time (Perceived cost values, per hour (2010 prices)). As per TMfS, car driver values of time are assumed. Parking Cost Assumptions Parking costs are derived from TMfS for 2012, but separate assumptions have been developed for The reason for this is that TMfS assumes no real-term increase in parking costs, which doesn t reflect the limited evidence we have available. The following assumptions have been made with respect to parking: In-work parking costs are assumed to be zero, reflecting an assumption that employees can claim costs back and will therefore not perceive the full cost of parking Parking costs are applied in defined city centre zones. The mapping set out in TMfS is out of date, but it is understood that the zones assumed to be parking zones approximately match the parking zones defined in TMfS. Parking costs are assumed for non-work commute and non-work other trips. In the Morning Peak Commuter model the TMfS assumption that 15 per cent of trips do not result in a car being

14 Appendix 02 Network Rail Scotland Route Study 102 Decision Level 1 Car PT Decision Level 2 parked (i.e. the purpose of the journey is to drop someone off) and that another 40 per cent of trips result in private, nonresidential parking at zero cost is followed. Parking costs are therefore spread over all trips at the rate of 45 per cent. In the Interurban model (which uses inter-peak data), it is assumed that short stay parking is used. It is assumed (following TMfS) that 80 per cent of trips will incur parking charges and so the cost of parking is applied at this rate. TMfS assumes no real term growth in the cost of parking, however the model does not include this assumption because it does not reflect the evidence (from City Parking detailed 10 per cent increases in parking costs in FY2013/14). It is recognised that the evidence base for parking is poor, and highly reliant on anecdotal evidence. To reflect this, a real terms growth rate in the cost of parking of 3 per cent per annum has been applied between 2012 and 2018, with different growth rates applied thereafter depending on which scenario is assumed. This divergence from TMfS reflects: The requirement on local councils to reduce vehicle emissions in line with European Law. It is understood that much of the policy response to this is likely to focus on buses and Heavy Goods Vehicles. However, it is understood that a reduction in overall traffic levels is likely to be pursued in Glasgow The increasing importance of parking charges as a source of revenue for local councils. It may be that charges are capped to marginal cost of providing parking, but where free parking is currently provided (e.g. at stations) it is understood that many councils are looking at charging at previously uncharged locations Increased competition for development land in city centre locations will inevitably increase the return on capital required from private car parks. No explicit assumption is made for parking search time. Parking search time is included in the assumptions made for GJT. Population and Employment Assumptions Population and employment forecasts are derived from TELMoS 3 up to 2018, with the 2018 data being interpolated from 2017 and 2022 data. No assumptions have been made to attempt to disaggregate forecast employment into different employment types (which may have different propensities to travel by rail). However, scenario employment growth rates are based on employment types that are associated with city centre employment activities. Market size model The market size model sums the total demand for travel for 2012 and Post-2018 demand is estimated by applying growth to the 2018 total demand based on the key drivers identified for each market under consideration, following RUMS and the LDMS. The factors are: City centre employment growth as the driver of the Morning Peak Commuter Market Population growth as the driver of the Interurban market. These assumptions are based on the findings of the Regional Urban Market Study and the Long Distance Market Study which found that modelling for other potential demand drivers (demographics and car ownership) did not materially affect the forecasts given the additional complexity they introduce into the modelling. Total transport demand estimates were taken from TMfS/CSTM forecasts for 2012 and 2018 and growth factors for both variables were contained in the scenarios developed for 2023 and Total demand is estimated at a zone-to-zone level of granularity. Market Share model Overview Bus Figure 2.3 Market Share model structure Rail 3 TELMoS (Transport and Economic Land-use Model of Scotland), provides independant demographic planning and economic forecasts which form the basis for future travel demands. It underpins many of the assumptions on which both CSTM and TMfS are based.

15 Appendix 02 Network Rail Scotland Route Study 103 Total Market Size X Forecast Rail Demand Rail Market Share The market share element of the model considers two decisions on the part of transport users. The first (Level 1) decision is the decision between car and public transport (PT). The second (Level 2) decision is, for PT users, between rail and bus. Rail market share is estimated by multiplying level 1 market share (car versus PT) by the level 2 market share (rail versus bus). PT costs in the level 1 market share model were estimated based on implied market shares from the level 1 decision. The model structure is displayed in Figure 2.6. Functional Form The functional form of the market share -αgci model is as follows: ii Where: MSi = MSi is the market share of mode i GCi is the generalised cost of a journey by mode i consisting of: In vehicle time / access time Interchange time Ticket price/marginal cost of motoring α is the degree of responsiveness to differences in generalised cost (known as the Mode Share Parameter or MSP). As Figure 2.4 illustrates, the value of the Mode Share Parameter is important for determining rail market share: a higher α value suggests that the market is more responsive to cost variances than a lower α value. The model is calibrated by adjusting the Level 1 and Level 2 MSPs. e e -αgcijii Total Rail Demand Forecast Total rail demand is estimated by multiplying the modelled estimate of rail market share by the modelled estimate of total transport demand for each forecast year. For the Interurban market, all station origin-destination (O-D) pairs were utilised for the forecast; for the Morning Peak Commuter market, the dataset was filters so that only those O-D pairs where the destination zone was in a defined city centre zone were considered. City centre zones were defined as follows: For Aberdeen and Edinburgh, a 2km cordon was drawn around the main stations (Aberdeen, Waverley and Haymarket) and major rail-served development areas outside these cordons (Dyce/South Gyle/Edinburgh Park/Edinburgh Gateway) were also included For Glasgow, where there are more stations in and around the city centre core, a cordon was drawn linking those stations where current (2014) demand on trains is at its peak. For instance, loads on North Electric services accessing Glasgow Queen Street Low Level from the west tend to be at their peak loading at Partick or Charing Cross. Assigning demand to corridors The model forecasts demand for station pairs. It does not assign demand to corridors. Assigning demand to corridors is essential as it is only then that demand can be compared to on-train capacity. Assignment of zones to stations is carried out on the basis of geographical proximity. This part of the process was spreadsheet-based and utilised the data tables prepared for developing the connectivity-based conditional outputs. Figure 2.4 Deriving Forecast Rail Demand

16 Appendix 02 Network Rail Scotland Route Study 104 Passenger Growth vs Available Seats % % % 125% 116% 114% 105% % 67% Aberdeen/Dundee/Perth to Glasgow Figure 2.5 Sample output from Capacity Analysis Spreadsheet Tool Once a station O-D pair has been assigned to a corridor, a corridor growth factor was then applied to a MOIRA download for May 2014 in the Capacity Analysis Spreadsheet Tool (adjusted from the 2012-based growth rates), to which rolling stock capacities had been applied. This enabled the identification of which corridors are likely to become capacity constrained in different time bands. Limitations of the Market Share / Market Size Approach The approach to modelling that was ratified by the Working Group represents a significant step forward for forecasting rail demand in Scotland. The models for the Morning Peak Commuter and Interurban markets are able to forecast demand on a corridor-bycorridor basis something which has not been feasible before. It needs to be recognised, though, that the approach selected has limitations of its own. These can be categorised as follows: Limitations of scope The models have been developed to be as parsimonious as possible (i.e. to explain as much variation as possible with the minimum number of explanatory variables). This is generally good modelling practice, but it increases the risk of omitting variables that could explain some variation in demand. Potential omitted variables include car ownership, demographics and destination choice. In each of the markets that it covers, the model focuses on either the origin (in the case of the Interurban market) or the destination (in the case of the Morning Peak Commuter markets) but not both. For instance (using the Morning Peak Commuter model for Aberdeen), the model will grow the number of trips according to employment growth at the destination. What it won t do is capture any changes in the distribution of population at the origin: the distribution of trip originations is held constant. The model does not include data for tram or subway use. The model was constructed to allow for these modes, but the dataset for them was not complete so they were switched off in the model. Future versions of the model could include these if the database is amended. Weak assumptions/data The model takes account of parking charges in city centres. Both TMfS and CSTM include an allowance for parking costs in city centre zones, but both TMfS/CSTM assume that the supply of parking is infinitely elastic i.e. that parking costs remain the same in real terms in future years. This assumption has been amended in our modelling because the evidence strongly suggests that this may not be realistic. The evidence base on parking, however, is recognised to be weak. The use of a single driver to grow overall market size simplifies the reasons people travel where and when they do i.e. accessing employment not the only source of commuting (studying, access to services etc). TMfS/CSTM zoning issues: TMfS and CSTM are both built up from trip matrices between origin and destination zones. The size and shape of these zones, together with the nodes that sit within them, can have an impact on the model results and require careful interpretation. This is particularly the case where the zones are large (e.g. around Aberdeen and Inverness). Cost data is estimated from the node, and this can lead to demand being assigned that does not reflect the reality of where people live. Modelling limitations The model uses two mode share parameters. The Rail vs Bus Mode Share Parameter is -0.6 and the car vs Public Transport parameter has been set to following the calibration exercise. This implies (a) that people are more sensitive to cost variability in rail vs bus decision than they are when deciding whether the drive or use public transport and (b) that their sensitivity to cost is independent of their journey purpose. Both of these are simplifying assumptions, but both could be easily amended given how the model is structured The model includes all journey purposes in its estimate of total market size. This could be amended to filter out non-applicable journey purposes It is important to note that the models intentionally make no allowance for crowding. The reason for this is that the purpose of the Route Study is to start with an unconstrained vision of conditional outputs (including demand) which have further assumptions applied (including crowding) at the appraisal stage

17 Appendix 02 Network Rail Scotland Route Study 105 This assignment of origin and destination zones is potentially problematic (a) when zones are large and (b) when there is a high concentration of stations The Interurban forecasts are based on single leg journeys. In principle, the decision to travel will factor in the return leg. This could be factored in to the analysis by adding costs for the two journey legs together and then applying the market share decisions to the costs. Model Calibration The purpose of the model is to estimate the growth in rail demand given changes in the explanatory variables identified in the Regional Urban Market Study. Because of this, although the growth factors produced are applied to actual data (LENNON or on-train counts), the overall rail demand forecasts cannot be calibrated to these. The reasons are that: The mode choice decision does not factor in qualitative aspects The mode choice decision does not factor in mode choice inertia (i.e. habit). For this reason, the model has been calibrated against TMfS/CSTM which make broadly similar assumptions (although TMfS/CSTM are considerably more complex models). These have been compared to ScotRail forecasts and to MOIRA as a sense check. Having compared the model forecast with forecast rail demand from CSTM/TMfS, an exercise was carried out to explain the differences between the model forecast and the rail forecast in TMfS/CSTM The principal differences related to The mode share parameter values The extent to which certain parts of the market are captive to public transport The extent to which people change their choice of destination on the basis of the transport offer. Mode share parameter The MSP is the key element to LOGIT modelling. The MSP amplifies user responses to difference in overall (generalised) costs. A high MSP (approaching -1) implies highly responsive markets; a low MSP (approaching zero) implies highly unresponsive markets. In general, MSPs for choices between PT modes are higher than MSPs for choices between PT and car. In principle, different MSPs can be applied for different times of day and different journey purposes. They are also likely to be higher where costs explain a high level of demand variability. TMfS/CSTM has a suite of MSPs, but for reasons of simplicity two values have been used in our models one for PT versus Car and one for Rail versus Bus. PT Captive Market In practice, mode choice is not only driven by marginal transport costs. Investment in a car is a major cost for households and car ownership is likely to vary significantly depending on parking availability, household income, residential density and other factors. All these factors mean that PT demand from some zones is likely to be far higher than would be implied by cost alone. TMfS/CSTM attempts to capture this by taking a certain proportion of journeys out of the car vs PT mode choice model. This process is not done within the model, and this explains some of the difference between them. Destination Choice For some journey purposes (for instance commuting) journey destinations are fixed in the short run. For other journey purposes, however, this is not necessarily the case, and changes in journey times may lead to different destinations being selected if it takes longer to get to their preferred destination. This has not been modelled. Overall Results The overall calibration process suggested that model forecast approximately 10 per cent more rail demand in both 2012 and 2018 than TMfS/CSTM. This is acceptable because the purpose of the calibration exercise was to achieve consistency of outcome rather than to estimate actual flows.