Scottish Government

The Ability of Public Transport to Cope withTarget Passenger Increases: Final Report

5. ANALYSIS OF RAIL CAPACITY UTILISATION

5.1 Introduction

5.1.1 This section identifies the rail capacity currently provided in each of the four cities under consideration and assesses how much spare capacity, if any, is found on existing services. It also consolidates the existing data relating to latent infrastructure capacity within the system that could be used to provide additional capacity where required in the future. The key objectives for this analysis of the rail network are:

• to identify the existing rail infrastructure utilisation factors for the relevant rail routes, particularly at peak periods but also in general terms
• to determine the approximate level of spare capacity on existing train services, again primarily at peak periods
• to assess the extent to which target passenger increases for each of Glasgow, Edinburgh, Aberdeen and Dundee are likely to be able to be catered for by the current (or planned) service provision
• where additional train or infrastructure capacity is likely to be required, to broadly identify what resource implications would be anticipated, particularly where proposals to enhance capacity are not currently in progress.

5.1.2 Rail capacity, in terms of the number of people that can be carried by a route or network, is governed by the following factors:

• capacity of the track and signalling infrastructure, which is influenced primarily by the number of tracks, the number and nature of junctions and the characteristics of the signalling system. There may also be specific infrastructure features or different mixes of train types and/or operating characteristics that influence route capacity. In combination, these attributes govern the number of train paths per hour
• platform lengths, since these generally dictate the maximum length of train that can be operated
• platform availability and occupancy, particularly at terminal stations
• actual train lengths operated, which may be less than that allowed by the platform lengths and may be dictated by factors such as the train fleet available and its service availability
• internal layouts of the trains used, since they may differ in terms of numbers of seats, amount of standing space, etc.
• in certain circumstances, the capacities of stations to be able to accommodate the numbers of passengers arriving, waiting and/or departing. This is not believed to be a critical factor on any part of the Scottish rail network, but will be discussed with rail companies.

5.1.3 In determining the number of train paths per hour, Railtrack studies tend to focus upon the track and signalling infrastructure to identify a theoretical capacity limit. To allow for different train types, recovery time to cope with service perturbations, etc. this theoretical limit is reduced by 20% (i.e. the reduction factor) to provide a practical capacity for a section of line. Thus, a section of line that can theoretically handle one train every six minutes in each direction (i.e. 10 trains per hour each way) would have a practical capacity of eight trains per hour in each direction. Railtrack states that "the validity of this approach, while impossible to justify in strictly analytical terms, has been proven over a number of years" (Railtrack, 2001a).

5.1.4 It would be expected that a greater reduction factor (to achieve the practical capacity limit) may be necessary:

• in situations where either network capacity across a range of routes is highly utilised (since trains would be more likely to suffer delay and present themselves late at sections of route)
• where the mix of trains is more diverse than normal in terms of stopping patterns or maximum speeds, since theoretical train paths would be far less optimally utilised and/or
• where the infrastructure prevents trains with different characteristics from avoiding conflicts (e.g. through a lack of passing loops).

5.1.5 These considerations will be very specific to certain sections of route and are not discernible from the available data and literature. They require more detailed analysis than required to satisfy this study brief, but may be important factors when addressing specific capacity problems in the future. In this section, the specific infrastructure-related issues for routes around Glasgow, Edinburgh, Aberdeen and Dundee are analysed. Following the analysis of the infrastructure-related issues and train paths scheduled, this chapter turns to examine existing occupancy levels on routes into each of the cities, the main focus being on peak periods.

5.2 Rail Infrastructure Capacity Utilisation

5.2.1 This section discusses the four Railtrack studies that have been conducted on behalf of the SRA during 2001 and 2002, collectively referred to as the Railtrack Scotland Infrastructure Capacity Studies. First, it identifies the methodologies adopted, factors considered and general findings and, second, it considers the findings in the context of the routes serving each of the four cities. Measures identified in the Network Management Statement (Railtrack, 2002) are incorporated where relevant.

Study Methodologies and General Findings

5.2.2 Three of the four Railtrack Scotland studies were focused on the current utilisation of the network in terms of train paths, governed primarily by track layout and signalling (Railtrack, 2001a, 2002a, 2002b). The fourth study was concerned specifically with constraints resulting from existing platform lengths, which limit the maximum train lengths that can operate on particular routes.

Train Paths

5.2.3 Stage 1 of the Central Scotland Capacity Review analysed the capacity utilisation of each section of route within the study area, comparing current utilisation (or current utilisation plus committed/likely service enhancements) with the practical capacity of that section. The practical capacity was calculated as the number of trains per hour that could be handled by each section of the network (i.e. the theoretical capacity), scaled down by a reduction factor of 20% to account for junction conflicts, scope for recovery from service perturbations, etc. This is clearly a relatively crude measure of capacity utilisation, but is one that provides a consistent approach and is fairly standard within the rail industry.

5.2.4 This approach identified a total of 21 locations that were identified as "hot spots", where capacity utilisation currently (or with likely service enhancements below, referred to as projected peak utilisation) exceeds 80% of the practical capacity. A further 4 "hot spots" were identified in the follow-on review of the network through to the NESTRANS area. The 25 "hot spots" identified by Railtrack are shown in Table 5.1, listed in descending order of their current peak train path utilisation. This table also reveals which of the four cities have services that utilise the network at these locations, with those affected only slightly being shown in brackets.

5.2.5 The likely service enhancements that were included were as follows:

• Edinburgh Crossrail (implemented in June 2002)
• Stirling/Alloa/Kincardine
• Larkhall to Milngavie
• Virgin CrossCountry upgrade (implemented in September 2002)
• West Coast Route Modernisation
• East Coast Main Line Upgrade
• Glasgow to Kilmarnock half-hourly service (Incremental Output Statement (IOS))
• 6 trains per hour off-peak Ayrshire services (IOS)
• Glasgow Queen St to Dundee half-hourly off-peak services (IOS)
• Glasgow Central to Shotts Sunday service (IOS)
• Glasgow Queen St to Cumbernauld Sunday service (IOS)
• Glasgow Central to East Kilbride later Sunday service (IOS)
• Aberdeen CrossRail (IOS)
• Aberdeen to Inverness (IOS)

Table 5.1: Capacity "Hot Spots" Identified by Railtrack (Peak Period)

5.2.6 A number of separate initiatives were identified by Railtrack as being under development that, if implemented, should resolve certain of the "hot spot" capacity problems. They are identified in Table 5.1 with a , with the individual initiatives being as follows:

• Edinburgh Waverley MasterPlan - additional platforms and remodelling of station throat, to relieve Waverley to Haymarket pressures
• Stirling/Alloa/Kincardine - reopening of route which will relieve some pressure on Haymarket to Dalmeny and Dalmeny to Inverkeithing sections
• Edinburgh Airport Rail Link - should provide more capacity for Haymarket to Newbridge
• Signalling and junction improvements - to increase capacity between Slateford and Midcalder and Midcalder and Carstairs
• East Coast Main Line Upgrade - this should deal with capacity problems between Monktonhall and Drem
• Glasgow Queen Street redevelopment - additional platforms to provide extra capacity on Queen Street to Cowlairs section
• Additional tracks between Greenhill and Cowlairs, to relieve pressure on this section
• Glasgow Airport Rail Link - provision of additional tracks between Shields Jn and Paisley
• Dynamic loop between Stewarton and Dunlop (IOS) - resolve capacity problems between Barrhead and Kilmarnock

5.2.7 It is clear that, even if all of these initiatives are implemented, there will still be many capacity "hot spots" that will be unable to cater for additional train paths during the peak periods. It is evident, though, that these capacity issues are primarily a peak problem, as can be seen from the same analysis of the off-peak (Table 5.2). Just four sections of route are identified as being problematic outwith the peak, three of which are covered by previously identified separate initiatives and the fourth by the Bathgate to Airdrie line reopening.

Table 5.2: Capacity "Hot Spots" Identified by Railtrack (Off-Peak Period)

5.2.8 Certain of the single track routes may also pose problems for increasing service provision in both directions, even where existing off-peak utilisation is well within acceptable limits. Particular problems would be the Kilmarnock line and the Largs, Balloch, Helensburgh, North Berwick and Newcraighall branches, where an additional train path per hour in each direction would be impractical with the existing infrastructure even at off-peak times.

Platform Lengths

5.2.9 In certain cases, the capacity of a rail route may be limited by the length of platform at one or more stations. Except where "grandfather rights" exist, trains are generally not allowed to stop at stations unless all doors are aligned with the platform. Table 5.3 reveals the existing platform length constraints on each route.

Table 5.3: Platform Length Constraints by Route

5.2.10 It is clear that a large number of the Glasgow and Edinburgh suburban routes can cater for trains of no more than 4 cars, although the inter-urban and Glasgow Electric routes are generally able to handle 6 car trains.

5.3 Summary of Infrastructure-Related Issues

5.3.1 This section provides a brief summary of the infrastructure-related issues and the implications that constraints identified will have for each of the four cities.

Implications for Rail Services to/from Glasgow

5.3.2 The key issues for Glasgow are as follows:

• peak period track capacity is almost fully-utilised into Glasgow Queen Street High Level, with platform capacity at the terminus being a particular constraint
• platform lengths on the Glasgow Queen Street to Maryhill, Dunblane and Cumbernauld/Falkirk Grahamston routes tend to limit train lengths to a maximum of 4 cars
• express services from Glasgow Queen Street to Edinburgh, Perth, Dundee and Aberdeen can generally operate at up to 6 car train lengths with existing station stopping patterns, but service frequencies are affected by several capacity "hot spots". Platform length data are not available for stations north of Dundee but, with the possible exceptions of Portlethen and Dyce, are likely to allow 6 car trains
• all Glasgow Electric routes can cater for train lengths of at least 6 cars, but very few stations can cater for anything longer than this
• all services on Glasgow Electric Routes that serve the low level lines through Queen Street or Central use sections of route defined as capacity "hot spots", potentially limiting additional services using the low level routes to traverse Glasgow
• all sections of route served by Glasgow Electric trains from Central High Level on the short-distance routes (i.e. to Cathcart Circle, Newton and Neilston) have peak capacity utilisation of less than 75%
• at Glasgow Central High Level, all platforms (with the exception of 11a) can cater for EMUs of at least 6 cars. No mention is made in the Railtrack reports of any platform availability and occupancy problems, but this will be clearly dependent upon train lengths and turnround times
• the Ayrshire and Inverclyde electric routes are constrained primarily by the section of route between Cardonald Junction and Wallneuk Junction (effectively Glasgow Central to Paisley Gilmour Street), which operates above capacity at peak times and at 75% of capacity off-peak. Even if the Glasgow Airport Rail Link is implemented, growth on the Ayrshire corridor will be limited by the "hot spot" between Paisley and Kilwinning
• platform lengths on the diesel lines from Glasgow Central to Whifflet, Edinburgh (via Shotts), East Kilbride, Barrhead, Kilmarnock and Paisley Canal limit train lengths to a maximum of 4 cars. A number of platforms at Central High Level are limited to 4 or 5 car diesel trains (of 23 metre car length). In addition, track capacity is fully-utilised at peak times on sections of the East Kilbride and Kilmarnock routes and is heavily utilised on sections of the Edinburgh (via Shotts) route
• Anglo-Scottish services to the WCML and ECML appear to be able to operate at up to 11 and 9 carriage trains respectively, with an additional carriage possible on each route (within Scotland at least) dependent upon station stopping patterns. Available track capacity is heavily utilised at peak times on much of the Glasgow to Edinburgh (via Carstairs) route, affecting services from Glasgow to both the WCML and ECML.

Implications for Rail Services to/from Edinburgh

5.3.3 The key issues for Edinburgh are as follows:

• Waverley station and its approaches currently act as a capacity constraint, but this should be remedied by the proposed station redevelopment and remodelling of its track layout. In addition, three platforms are not capable of accommodating 6 car trains at present
• express services from Edinburgh to Glasgow Queen Street, Perth, Dundee and Aberdeen can generally operate at up to 6 car train lengths with existing station stopping patterns, but service frequencies are affected by several capacity "hot spots". Platform length data are not available for stations north of Dundee but, with the possible exception of Portlethen and Dyce, are likely to allow 6 car trains
• local services from Edinburgh to Bathgate, Dunblane, Glasgow Central (via Shotts) and Fife Circle/Dundee are limited to a maximum of 4 cars by platform lengths at a number of stations
• all services heading west from Waverley are affected by capacity constraints between Waverley and Haymarket and then by separate other sections of line on their respective routes
• stations on the Edinburgh to North Berwick and Newcraighall routes can accommodate trains up to 6 cars in length, although the current rolling stock and service patterns tend to preclude trains of more than 4 cars
• Anglo-Scottish services to the ECML and WCML appear to be able to operate at up to 11 and 9 carriage trains respectively, with an additional carriage possible on the WCML route (within Scotland at least) dependent upon station stopping patterns. Available track capacity is heavily utilised at peak times between Edinburgh and Carstairs, affecting services to the WCML. Track capacity between Monktonhall Jn and Drem Jn is becoming a significant issue for ECML services.

Implications for Rail Services to/from Aberdeen and Dundee

5.3.4 The key issues for Dundee and Aberdeen are as follows:

• many of the capacity "hot spots" affecting services to/from Aberdeen and Dundee are in the Central Belt, on the routes into Edinburgh and Glasgow
• further "hot spots" exist on single track sections of route, with the greatest problems north of Aberdeen. The single track section between Arbroath and Montrose is the only other capacity "hot spot" in the North East area.

5.4 Actual Versus Planned Capacity Provision

5.4.1 Much of the available data on capacity provision clearly relates to that timetabled (i.e. planned capacity). This may not reflect the actual situation; for example, if a peak period service is cancelled daily on a route with 4 trains per hour, this would result in actual capacity provided being just 87.5% of that planned over a two hour peak period. A similar reduction in actual capacity may also result from 2 or 3 car trains being provided where a 4 car one is scheduled (i.e. short-formed trains).

5.4.2 Some limited statistics are available from the Strategic Rail Authority (SRA, 2000; 2001a; 2001b; 2002) relating to problems such as service cancellations and short-formation trains. The following definitions are provided in the On Track statistical publication for the various incentive payments:

• Punctuality Incentive Payments (PIP) - this "measures lateness and cancellations on the day against the planned cancellations…If average lateness is better than the benchmark, the SRA pays the operator; if worse, the operator pays the SRA."
• Short Formation Incentive Payments (SFIP) - each operator must have a train plan showing how the required level of capacity will be provided. If they fail to meet this plan, an SFIP charge is applied.
• Timetable Change Incentive Payment (TCIP) - this "penalises operators who change the timetable from the printed version", but "with an incentive to handle disruption in a planned way, and to give passengers notice of amended services."

5.4.3 At the time of franchising, such incentive payments typically applied only to local and regional train operators. ScotRail's performance during the last three financial years can be seen in Table 5.4.

Table 5.4: SRA Incentive Regimes (ScotRail)

5.4.4 The statistics appear to reveal a marked worsening of ScotRail's performance with regard to service performance and adherence to the timetable. In 1999/00, a bonus payment of over 5 million was made from the SRA to ScotRail; in 2001/02, however, a penalty payment of almost 10 million was made in the opposite direction. The main change has been in the PIP measure, which covers both late and cancelled trains. From the capacity point of view, it is the cancellation rate that is significant. In the statistics, a train is classed as cancelled if it runs less than half of its planned mileage. Table 5.5 displays ScotRail's cancellation rate, based upon the 'plan of the day' timetable, which may in itself be a reduction from the standard timetable.

5.4.5 Comparison between Tables 5.4 and 5.5 suggests that train lateness, rather than cancellation, has been the main cause of the increasing PIP penalty payment. There is no evidence from Table 5.5 that the cancellation rate is deteriorating, with the more recent summer period having a lower rate than the previous one, but the more recent winter cancellation rate was higher than the one before. Cancellations are greater during the winter than the summer. The statistics suggest that around 2% of all trains are cancelled. If these were operated rather than cancelled, there would be a marginal increase in capacity provided. It would perhaps be expected, however, that the cancellation rate is higher around the main urban areas than on inter-urban and rural routes, and at peak rather than non-peak times, so the impacts upon peak capacity provision in the main cities may actually be more severe than the aggregate statistics suggest.

Table 5.5: Train Cancellation Rate (ScotRail)

5.5 On-train Capacity Utilisation

5.5.1 It was identified earlier that there are very little published data relating peak patronage levels to capacity provided, thus creating difficulties in identifying the level of spare capacity existing at peak periods. This section analyses the little published data available, together with patronage data provided by SPT and the original occupancy surveys undertaken specifically for this study. As with the bus survey data, it is necessary to restate that all rail patronage analysis, both SPT and non-SPT, is based upon single day observations at each survey location/on each route and, as such, provides only a snapshot of rail capacity utilisation.

Published Train Occupancy Statistics

5.5.2 The only published data relate to the measure of passengers in excess of capacity (PIXC), but they refer solely to ScotRail services across the Forth Bridge into Edinburgh, this measures capacity utilisation on weekday trains arriving in Edinburgh between 0700 and 0959 and departing Edinburgh between 1600 and 1859. The acceptable level stipulated by the SRA is a PIXC value of 4.5% in one peak or 3.0% across both peaks. For journeys over 20 minutes, which includes the Edinburgh to Fife corridor, capacity is taken as the number of standard class seats on the train. These statistics have been published only for the last two years and are shown in Table 5.6. PIXC data are not published for any other rail routes in Scotland.

Table 5.6: Passengers in Excess of Capacity (PIXC) across Forth Bridge

5.5.3 This reveals that, in both peak periods and in both survey years, the supply of seats has been less than the number of passengers carried. Unless additional capacity is provided, it is clear from these data that any additional passengers could be accommodated only by increasing the number of standing passengers, since peak hour seats are fully occupied.

Train Occupancy for Glasgow Routes

5.5.4 The analysis of patronage on the routes to/from Glasgow is based upon data provided by SPT ( see Section 3.2). Table 5.7 reveals the peak period utilisation on each of the routes serving Glasgow City Centre. This is provided both in terms of seat utilisation and train utilisation, the latter incorporating SPT's assumed standing capacity for the particular train type.

5.5.5 For most routes, where services either terminate or commence at Glasgow Central or Queen Street, the peak loading occurs on the innermost section of route and it is a straightforward task to identify the utilisation of available capacity. For the two through routes (i.e. North Electric through Queen Street Low Level and Argyle through Central Low Level), there are several City Centre stations and peak loadings sometimes occur elsewhere along the route than at Queen Street or Central. These two routes are considered in more detail later in the report.

Table 5.7: Surveyed Capacity Utilisation of Peak Rail Services to/from Glasgow City Centre (2000/01)

5.5.6 There is considerable variation In occupancy levels between different routes, ranging from a low of 26% seat utilisation in the evening peak for the Maryhill route up to a maximum of 160% for the Shotts route in the evening peak. Four of the 10 routes for which complete data are available were found to be operating at or above the seated capacity during the morning peak, but only one is above seated capacity in the evening peak. The Shotts line is the only one currently operating with patronage above the total train capacity, this occurring in both the morning and evening peak periods. While there appears to be significant spare capacity across the Glasgow network, it is by no means clear that the target patronage growth can be catered for. This is analysed in depth in Section 7.2.

5.5.7 As with the bus occupancy analysis earlier, specific services may well be operating at or above capacity due to the specific peaking of flows. In addition to the peak averages, therefore, it is important to consider particular trains that have existing capacity problems, or that would become problematic if patronage was to grow as desired by the targets.

5.5.8 Virtually all of the Glasgow local rail routes had one or more services experiencing overloading in terms of the seated capacity. In total, though, just 67 individual trains out of a weekday total of around 1,475 trains (i.e. 4.5%) were observed to have more passengers than seats at any particular time in the journey. 34 of these had passengers standing for 10 minutes of more, with one morning peak service each on the Argyle and Inverclyde lines having passengers standing in excess of 20 minutes. On certain services, primarily in the morning peak, significant numbers of people were observed having to stand for 10 minutes or more, as shown in Table 5.8. With just one exception, all these severely overcrowded services served Glasgow between 0800 and 0900 or 1700 and 1800.

Table 5.8: Surveyed Services with 50 or More Passengers Standing for 10 Minutes or More

5.5.9 Only 11 of the 67 overcrowded trains were longer than 3 cars, so the vast majority of overcrowding took place on trains that could be lengthened, subject to rolling stock being available. Of the overcrowded trains shown in Table 5.8, just 5 are currently operating at the maximum train length permitted by the platforms on the route.

Train Occupancy for Edinburgh Routes

5.5.10 The analysis of occupancy levels for the Edinburgh routes has been based upon the peak period surveys outlined in Section 3.2. Again, it should be borne in mind that these are snapshots of particular days and may not be fully representative of the general situation.

5.5.11 Table 5.9 reveals the survey analysis for Edinburgh-based routes in the peak periods. Capacity in the evening peak period appears to be slightly more of a premium than in the morning, although this is likely to be a result of service disruption that led to the cancellation of two evening peak trains on the Dunblane route (see below). Train capacity has been assumed to be 135% of seated capacity, based on data provided by ScotRail (which varies slightly to more specific train type assumptions used by SPT above).

Table 5.9: Surveyed Capacity Utilisation of Peak Rail Services to/from Edinburgh

5.5.12 In the morning peak period across all routes, seated capacity is almost fully utilised. The Fife express, Glasgow Queen Street and Bathgate routes were all observed to be carrying more than 20% more passengers than seats. Fife local services were 99% full (against seated capacity), with the other routes having spare seated capacity across the morning peak. In the evening peak, the number of passengers departing Edinburgh exceeded the provision of seats by 6%. All routes apart from Fife express, Newcraighall and Kirknewton were found to be operating at above the seated capacity, although there were virtually no spare seats on the Fife express route.

5.5.13 The Dunblane and Glasgow Queen Street routes were observed to be operating at around 40% above seated capacity. Care should be taken with the interpretation of the findings for these two routes, as there was considerable service disruption caused by a freight train derailment in Falkirk at the time of surveying. This particularly affected the evening peak, where both the 1703 and 1803 Edinburgh to Dunblane trains did not operate. This would be expected to lead to overloading on the Dunblane trains that did run, together with the Glasgow Queen Street services due to shared stops (i.e. Linlithgow, Polmont and Falkirk). Had the service operated to plan, seated capacity utilisation for the two routes combined would have been 91%, as opposed to the severe overloading observed. It is likely, however, that passenger numbers were lower than normal as some passengers would be likely to have made alternative travel arrangements or staggered their journeys. It would be reasonable to assume that these services would normally be operating at or just above seated capacity at present.

5.5.14 16 individual services to/from Edinburgh were observed to have 50 or more passengers standing (see Table 5.10). The majority of these services were observed to be only 2 or 3 cars in length. In three cases, shown in brackets in the "no. of cars" column, the services are booked to be longer trains but were not operating as such on the day surveyed. It is not known whether these observations of short-formed trains are a regular occurrence or a specific feature of the survey day. However, had the three trains operated as booked, one of them would have had sufficient seats for the total number of passengers and overloading on the other two would have been significantly lower. Further, as stated above, care should be taken with the Glasgow Queen Street and Dunblane services due to the service disruption and cancellation of trains on the survey day.

Table 5.10: Surveyed Edinburgh Services with 50 or More Standing Passengers

Train occupancy for Aberdeen routes

5.5.15 The analysis of occupancy levels for Aberdeen trains has been based upon the peak period surveys outlined in Section 3.2. Table 5.11 reveals the analysis of the rail survey at Aberdeen. It should be noted, however, that two pre-peak trains (one from the south and one from the north) were cancelled on the survey day, so that the observed passenger loadings on the following peak trains are likely to be far higher than normal.

Table 5.11: Surveyed Capacity Utilisation of Peak Rail Services to/from Aberdeen

5.5.16 In addition, one morning and three evening trains to/from the south were observed to be 2 cars in length instead of the booked 3 cars, thus inflating the occupancy levels from the normal, assuming they do normally operate as booked. Two of the three evening trains that were short-formed departed between 1700 and 1800 - had they operated as planned, the seat utilisation to/from the south in the evening peak would have been approximately 70% rather than 98% and the evening peak Aberdeen average would have been reduced to 64%.

5.5.17 In both three hour periods, no train outwith the main peak hour was more than around 80% full in terms of seats, with most being less than one-third full. Over the full three hour survey period, seat occupancy was found to be 56% in the morning and 47% in the evening. Had the two cancelled trains operated as planned, seat occupancy in the morning period would also have been 47%, on the assumption that total passenger numbers would have been the same (i.e. prospective passengers did not abandon their rail journey).

Train Occupancy for Dundee Routes

5.5.18 The analysis of occupancy levels for the Dundee routes has been based upon the peak period surveys outlined in Section 3.2. Table 5.12 reveals the analysis of the rail survey at Dundee. This reveals considerable spare capacity on all routes in both peak periods. In the morning peak, no train was more than 80% full against seated capacity, while in the evening peak the highest occupancy was 64% of seated capacity.

Table 5.12: Surveyed Capacity Utilisation of Peak Rail Services to/from Dundee

5.6 Known Rolling Stock Developments

5.6.1 There are two particularly significant developments affecting the amount and type of rolling stock. The first is currently in the final stages of implementation, and involves the introduction to service of a fleet of 40 new Class 334 electric multiple units (EMUs), effectively replacing the elderly Class 303s. This has not been a direct replacement of old with new, but has involved a cascade of different types of EMUs across the Ayrshire Coast, Inverclyde, South Electric and Argyle lines. Table 5.13 reveals the capacities of the different types of EMUs, with all but the Class 320 being involved in this cascade process.

Table 5.13: Capacities of Different Types of EMUs (SPT area)

5.6.2 Of the other classes of train, it can be seen that the new Class 334s have fewer seats than the older rolling stock, with 15% fewer seats than the Class 318s. The main reason for this reduced seating capacity is that seats are arranged four abreast in two of the three carriages, rather than the five abreast in older rolling stock and one of the Class 334 carriages. Dependent upon the internal train layouts, it would perhaps be expected that actual overall train capacities will be more similar between the Class 334s and earlier designs than is shown in Table 5.13, due to greater standing space in the new trains. If the aim is to seat all passengers (including patronage increases), then the differences in seated capacities may be significant. In general terms, it is also not clear how much daily variation there is in terms of which unit types actually operate each service; if this varies from the booked type then there may be capacity implications. It is not known whether this is a particular problem on the Scottish rail network.

5.6.3 The Class 334 trains have been introduced primarily on the Ayrshire Coast, Inverclyde and Argyle lines. On the latter, they appear mainly to have replaced Class 303s, with only minimal capacity implications. On the Ayrshire Coast and Inverclyde routes, however, services were previously operated mainly by Class 318s and Class 314s respectively, so it is on these routes that the changes in seated capacity may be significant. It is not known whether the new trains have taken over services on a like-for-like basis, in which case the capability to seat patronage growth will be reduced, or whether there has been a replacement of 3 car trains with 6 car ones with a corresponding increase in capacity provided. From the passenger count data, which pre-dated the introduction of the new trains, it would appear that 6 car Class 334s are necessary on further services than those previously booked for 6 cars in order to ensure that standing passengers are kept to a minimum.

5.6.4 The second of the rolling stock developments relates to the recent announcement of an order for 28 new 3 car diesel multiple units (DMUs), most likely additional Class 170 Turbostars. Some of these are planned to strengthen existing inter-urban express services, but others will allow a cascade of rolling stock across a number of diesel lines in central Scotland. This planned new rolling stock is proposed to be used as follows:

• 11 Standard/First Class trains for the Edinburgh/Glasgow/Aberdeen/Inverness quadrangle
• 11 Standard Class only trains for the Edinburgh to Fife Circle, Bathgate and Dunblane lines
• 6 Standard Class only trains for SPT, for the Glasgow to Dunblane (and Alloa) service.

5.6.5 The planned impacts of the introduction of these trains on peak service capacity provision can be seen in Table 5.14 based on information provided by ScotRail. On certain routes, the proposed increases in provision resulting from longer trains will be quite considerable, particularly on the Edinburgh suburban routes and the Edinburgh to Glasgow express route. Increases on the Bathgate, Fife Circle and Dunblane routes are contingent upon platform lengthening at a number of stations to allow 6 car trains to operate. However, it is believed that completion of this infrastructure work is an integral part of the funding agreement for the new trains. Platform lengthening on the East Kilbride line is also expected to take place, allowing 6 car operation on this line too.

5.6.6 This analysis focuses solely on the high peak periods. On many of the routes, valuable increases in seated provision are proposed on the shoulders of the peak. On the Central Belt to Aberdeen/Inverness routes, information from ScotRail shows strengthening of peak trains resulting in 5% more seats between Glasgow and Aberdeen, 7% more seats between Edinburgh and Aberdeen and 21% more seats on the Glasgow/Edinburgh to Inverness route.

5.6.7 Further analysis of the rail targets and future capacity provision within Scotland are contained in Chapters 8 and 9 in the light of the discussions with the relevant bodies.

Table 5.14: Central Belt Planned Capacity Increases Resulting from New DMUs

5.7 Anglo-Scottish Infrastructure and Train Capacity Issues

5.7.1 Little data are published relating specifically to capacity provided or passenger numbers on Anglo-Scottish express trains. While these services are not a major focus of this study, they are important in that the Scottish Strategic Rail Study growth projections generally assume greater growth on Anglo-Scottish routes than on internal Scottish ones. Table 5.15 summarises the changing service provision on the Anglo-Scottish routes between Summer 1996 and Summer 2002. This compares the last year of British Rail operation and the most recent period, when services were provided by GNER (on the East Coast) and Virgin Trains (West Coast and CrossCountry, the latter services using both East and West Coast Main Line to serve Scotland).

Table 5.15: Anglo-Scottish Express Train Services (Number per Day, 1996 & 2002)

5.7.2 The number of Anglo-Scottish express trains has increased from 572 per week in Summer 1996 to 616 per week in Summer 2002, an increase of just under 8%. Across all three routes, half of the increase in the number of trains has been at weekends. GNER services on the East Coast accounted for 60% of the growth in the number of trains, with the Virgin West Coast and CrossCountry routes making up the remainder. During the 1996 to 2002 period, there was no significant change in the composition of the train fleet, so capacity provided in terms of number of seats is likely to have increased at a broadly similar rate to the number of trains operated.

5.7.3 Virgin CrossCountry introduced its "Operation Princess" revised network in September 2002, which saw a dramatic timetable change linked to the squadron service introduction of the new Voyager train fleet. The basis of this change was the introduction of a regular-interval timetable, generally comprising more frequent but shorter trains. In most cases, 7 car trains with around 430 seats have been replaced by 4 or 5 car trains with 188 or 250 seats respectively. Operation Princess led to a doubling in trains operated on Anglo-Scottish routes, but with only a 22% increase in seating provision, based on the Monday to Friday southbound timetable. This increase has subsequently been scaled back slightly, due to the withdrawal of a number of services resulting from problems with the new timetable, and it would appear that the current Anglo-Scottish CrossCountry seated provision is only around 15% higher than in the old timetable.

5.7.4 Quite clearly, in aggregate, Operation Princess has provided greater capacity. However, CrossCountry services using the West Coast Main Line have lost approximately 1,000 seats (again based on Monday to Friday southbound), more than offset by an increase of around 2,200 on CrossCountry services on the East Coast Main Line. Glasgow has lost out on seats provided, as has Edinburgh to West Coast Main Line destinations, leading to severe overcrowding on many of these services. On the East Coast Main Line, the number of daily services has increased from 3 to 16, but these largely duplicate the GNER services between Edinburgh and York. It has been reported, in any case, that some of these new services north of York/Newcastle are to be withdrawn in Summer 2003.

5.7.5 For the future, within Scotland, the Railtrack capacity studies reveal some scope for additional train paths on both Anglo-Scottish main lines. However, there is likely to be more of an issue of mix of flows (e.g. high-speed passenger, local passenger, freight) on the long two-track sections of route towards the English border, which may limit the practical number of additional paths. Much growth in long distance traffic would be expected to occur at what would traditionally be regarded as non-peak track utilisation times. It is likely, though, that bottlenecks south of the border, rather than in Scotland, would be more of a problem in increasing train frequencies.

5.7.6 The West Coast Main Line infrastructure is currently being upgraded, with completion due by 2005/06. This should provide capacity for additional trains on the southern stretches and 13 West Coast trains per day are planned from 2004, rather than the current 7 to 9 trains per day. However, this increased frequency will be of Pendolino trains, which have around 20% fewer seats in total per 9 car train than the trains they are replacing (and around 30% in terms of Standard Class). Therefore, while the number of trains will increase by around 60% per weekday, the number of seats provided will be around 30% higher than at present. It is likely that only around 12% more Standard Class seats will be provided on West Coast Anglo-Scottish trains.

5.7.7 A comprehensive upgrade of the East Coast Main Line, to allow more frequent long distance services and greater freight capacity, had been proposed by the SRA. However, the 2003 Strategic Plan sees this project de-scoped, with some elements to be implemented by 2005 and others deferred indefinitely (SRA, 2003). The current capacity utilisation study is examining the potential for better utilisation of the existing infrastructure, together with the identification of certain enhancements that would increase capacity and improve performance. In the longer term, by 2010, a daytime core service of 2 trains per hour between London and Edinburgh is envisaged. For much of the day, however, this frequency already operates between the two cities, so the additional number of trains may not be particularly significant. The other key initiative planned for East Coast services is train lengthening. GNER is currently adding an extra coach to each train in its diesel fleet, creating 9 coach trains and increasing capacity of these trains by around 15%. In the Scottish context, these trains operate all GNER services north of Edinburgh (to /from Aberdeen/Inverness), plus a limited number between Edinburgh and London only. In the longer term, lengthening of the entire fleet to 10 coaches is an option identified in the Strategic Plan, which would further increase capacity by up to a further 15%.

5.7.8 For all Anglo-Scottish routes, train occupancy levels are not currently known. However, for long distance travel, particularly leisure, there may be considerable scope to increase passenger numbers through yield management, whereby existing capacity on quieter trains is filled.

5.8 Summary

5.8.1 This chapter has presented a detailed analysis of the existing situation with regard to infrastructure utilisation and service provision on the rail network within Scotland, together with a brief assessment of Anglo-Scottish issues. It has also analysed train occupancy data, both that provided by SPT and collected elsewhere for this study, albeit with the caveat that all such data provide only a snapshot of train occupancy. Future developments known of at the time of the study have been incorporated; however, as stipulated by the project brief, the impacts of potential large-scale developments (e.g. Airdrie to Bathgate reopening; Airport rail links and the Waverley line reopening) have not been explicitly analysed.

5.8.2 In summary, a number of infrastructure "hot spots" have been identified which, without enhancement work, will not be able to cater for increased train service frequencies at peak times. In many cases, separate initiatives have been developed in order to provide additional capacity, but there still remain a number of locations where additional peak train paths will not be available. In terms of on-train capacity utilisation, considerable peak period variations between routes (and, indeed, specific services) have been identified. Overcrowding has been found to be a feature of a significant minority of services at present, with Glasgow experiencing greater problems in the morning peak and Edinburgh routes being more overcrowded in the evening peak. Morning peak overcrowding was identified in the Aberdeen area, though this may have been solely due to service cancellations on the survey day.

5.8.3 Many of the existing on-train capacity problems for local services, particularly on the Edinburgh routes, should be resolved by the planned introduction of additional rolling stock and the associated station platform extension programme, which will allow service strengthening to take place. The analysis of Anglo-Scottish service provision has been more limited, though service frequency enhancements and train lengthening should provide additional capacity - there are still concerns as to whether this will be sufficient to cater for the SSRS planned patronage growth.

5.8.4 The key issues and implications for the patronage targets arising from this rail analysis are presented in Chapter 7.