The Effects of Park and Ride Supply and Pricing on Public Transport Demand
4 Parking at railway stations
4.1 Rail passenger numbers continue to increase in Scotland as a result of various timetable changes including the introduction of new or faster services. Coupled with employment growth in the main urban centres11 and worsening road congestion, this has improved the competitiveness of rail versus other modes.
4.2 As a result, the availability of parking has reduced with many full before the end of the AM peak period. If parking is not available, there is a perception that some may choose to make their entire journey by car rather than use rail. Alternatively, the length of the access leg by car could be extended if there is a lack of parking at some stations. This relationship has not been examined satisfactorily by previous studies.
4.3 Rail based park and ride can help to support a number of objectives including:
- economic: the number of people able to access city centre employment will increase, given the speed and capacity characteristics of rail versus other modes
- environmental: rail transport has significant potential to lower CO2 emissions through modal shift from car
- social: by encouraging some motorists to switch modes for at least part of their journey, this will reduce congestion levels and deliver other qualitative benefits, including improved quality of life and amenity
4.4 The selection of case studies for primary research was carefully reviewed by the client steering group to ensure a representative sample was selected. Primary research was conducted at Bridge of Allan, Kirkcaldy, Perth and East Kilbride stations where the number of car parking spaces had been recently increased. This provided a reasonable cross-section covering geographic location, rural or urban characteristics and different charging structures. Two "control" stations (Stirling and Falkirk High) provided comparisons to understand the impacts if the number of spaces available was unchanged. The controls are located in the same geographic area as the primary case studies, so the specific impacts of car park expansions could be isolated. Table 4.1 summarises details of the specific car parks examined.
| Station | Number of Spaces | Date of Expansion | Usage | Operator | Rail Services | |
|---|---|---|---|---|---|---|
| Before the Extension | After the Extension | |||||
| Kirkcaldy | 274 | 594 | Nov 2006 | 80% | Local authority | Regular services to Edinburgh, plus trains to Aberdeen / Inverness |
| Bridge of Allan | 114 | 146 | Dec 2005 | 100% | First ScotRail | Regular trains to Edinburgh and Glasgow Queen Street |
| East Kilbride | 162 | 287 | May 2007 | 97% | First ScotRail | Regular trains to Glasgow Central |
| Perth | 54 | 160 | N/A | 92% | First ScotRail | Regular trains to Glasgow Queen Street, Edinburgh, Aberdeen and Stirling |
| Falkirk High | 215 | 215 | N/A | 98% | First ScotRail | Regular trains to Edinburgh and Glasgow Queen Street |
| Stirling | 276 | 276 | N/A | 100% | First ScotRail | Regular trains to Glasgow Queen Street, Edinburgh, Aberdeen and Perth |
Source: National Rail website, First ScotRail monitoring data illustrating car parking occupancies
4.5 In addition to the above stations, other examples benefiting from car park extensions including Markinch, Dunfermline Town, Rosyth, Musselburgh, Cupar, Carluke, Uddingston, Glengarnock, Johnstone and Prestonpans were included in the forecasting model. Ticket data for other 'control' stations, namely, Leuchars, Motherwell and Falkirk Grahamston, were also included.
Consultation with First ScotRail
4.6 The impact of car parking extensions on rail demand was discussed with First ScotRail (FSR). This explored the relationship between parking availability, the resulting impact on rail demand and the influence of other factors from the perspective of the operator.
Support for park and ride schemes
4.7 FSR is generally supportive of schemes to increase car parking availability, especially if they help to attract new rail revenue. However, the high capital costs and short duration of the remaining franchise (the current contract expires in November 2014) means the scope for FSR to lead proposals is limited. The fulfilment of a Committed Obligation in the Franchise Agreement would generally necessitate FSR leading the development of a scheme, rather than progressing schemes commercially.
4.8 As a result, Regional Transport Partnerships have taken the lead in delivering extensions, certainly in the SPT (Strathclyde Partnership for Transport) area, with operational and maintenance responsibilities reverting to Local Authorities upon completion. However, proposals need to be aligned with other considerations, for example, capacity constraints at existing sites. Carluke is an example of a Council promoted scheme where the proportion of spaces occupied prior to expansion was very low. With just an hourly service towards Glasgow, the scope to attract additional motorists to this station is limited. This demonstrates the importance of aligning objectives carefully.
Role of parking in influencing wider demand growth
4.9 There are a number of factors influencing rail growth, including the attractiveness of the rail service and the extent of crowding problems. Extending station car parks with a competitive rail service that operate parallel to congested roads can attract new users. This is a particular issue if the rail service is transformed. The route between Edinburgh and Glasgow via Shotts was transformed in December 2009, with the introduction of some services which do not stop at all stations on the line. This has reduced journey times between Edinburgh and Glasgow Central by up to 33 minutes from 96 minutes to 63 minutes. Passenger numbers at selected stations have increased significantly as a result of the improved service, creating capacity problems at a number of station car parks.
Options to expand parking provision
4.10 The potential benefits of decking (providing an additional, second level of car parking at existing car parks) were highlighted by FSR, particularly if the scope for surface level expansions was limited. Bridge of Allan, Johnston and Uddingston stations were suggested as examples that may require decking if additional spaces are required. Rising land costs, particularly for plots close to railway stations were highlighted as a risk, so decking could offer a quicker solution. This approach would minimise the extra land take, although the capital costs would be considerably higher compared with a surface car park. The cost per space for a surface car park is £5,000 - £10,000, whilst the costs for decked spaces are 2-3 times higher12.
4.11 The operator has responsibilities for maintaining the extra spaces as specified under the SQUIRE13 regime, whilst collecting revenue at the stations where charges are enforced. They also actively promote the expansion of station car parks through the local press and leaflets (for non-users) and advertisements at stations for existing users. Station car parks have full CCTV coverage, but FSR does not feel that 'Park Mark' or similar quality assurance deliver good value for money due to poor recognition by passengers.
4.12 Whilst FSR are supportive of proposals to expand car parking availability at railway stations, the opportunities are constrained by several factors. The short remaining duration of the operator's franchise term restricts the scope to develop new proposals. The wider interface with the timetable must be considered as part of the overall proposition. Whilst decking existing car parks can reduce implementation timescales, particularly if land needs to be purchased, the resulting capital costs can be prohibitively expensive.
4.13 A forecasting model was populated using LENNON rail ticket data (which represents single rail trips)14, plus other parameters, including change in city centre employment, GDP, housing and the availability of parking spaces at the stations. The modelling framework incorporated two main components:
- development of a modelling framework populated with LENNON data
- inclusion of selected parameters using the results from the attitudinal surveys
4.14 Demand models that link observed and stated behavioural responses were examined to understand the relationship between parking availability, quality and prices, along with the change in attitudinal behaviour. The change in rail demand is also dependent on the change in parking spaces, the change in parking charge and changes and the availability of other factors, such as the quality and security of the parking, for example, CCTV.
4.15 Time series LENNON rail ticket data was collated for all passengers (not just park and ride users) by period (4 weekly intervals) for the primary and control stations and incorporated into econometric modelling software. The control stations were selected since they had not benefited from additional parking provision as shown in Table 4.1. Data from these stations was incorporated into the modelling framework to isolate the specific impacts of the extra parking spaces. This would enable the impacts of the case study stations versus the controls to be isolated. The number of trips and revenue generated from the busiest flows were collated, since any changes in car parking availability would have the greatest impact on demand affecting these flows. For example, the highest number of trips originating from Bridge of Allan arrived at Edinburgh, Glasgow Queen Street, Stirling, Dunblane and Falkirk. Journeys and revenue were split by ticket type (season and others).
4.16 The model was segmented to reflect the type of ticket being purchased (season and other) and the length of trip (up to 20 miles or longer). LENNON data was available for about 6 years, and 170 station-to-station flows were included in the model. The inclusion of almost 15,500 records meant the sample size was robust. The model incorporated other variables, including the capacity of the railway station car park, or the availability of spaces. A number of variables were also incorporated into the model using results from the primary research. A more detailed description of the modelling methodology is presented in Appendix A1.
Results from the primary data collection
4.17 To supplement the analysis presented in Chapter 3, further analysis of the primary research has been completed. The most popular destination from Kirkcaldy is Edinburgh Waverley (57 respondents out of 101). With the exception of Haymarket, there were fewer than 10 respondents travelling to other destinations. This reflects the high frequency rail service to Edinburgh from Kirkcaldy with 5-6 services per hour. Over 45 respondents from Bridge of Allan travelled to Edinburgh Waverley, with a further 33 to Glasgow Queen Street. The distribution of passenger journeys is consistent with the train service pattern, with 2 trains per hour to Edinburgh and an hourly service to Queen Street. Fewer than 10 people made journeys to other stations including Falkirk Grahamston. Almost all respondents surveyed at East Kilbride station were travelling to Glasgow Central, with fewer than 10 trips to any other station. The lack of direct services to other major destinations appears consistent with these results, as shown in Figure 4.1.
Figure 4.1: Number of single trips from Kirkcaldy, Bridge of Allan, East Kilbride and Perth
Source: Arup analysis of Accent data
4.18 Commuting and shopping were the most popular trip purposes as shown in Figure 4.2. Commuting trips accounted for 40-50% of the total, depending on the individual station. The importance of shopping trips differed, ranging from around 10% (Perth) to 35-40% (Falkirk High and Stirling). The relatively low percentage of shopping trips from Perth was offset by the high proportion visiting friends and relatives. The timings of the surveys (afternoon and evening peak) may have influenced the journey purpose.
Figure 4.2: Journey purpose
Source: Arup analysis of Accent data, sample size shown
4.19 Figure 4.3 illustrates the type of ticket purchased by the respondents at the six stations. Overall, 'season' tickets are the most popular type, especially at Bridge of Allan, Falkirk High and Stirling. In contrast, the percentage from Kirkcaldy using season tickets was relatively small (about 15%). The use of off peak tickets is higher from Kirkcaldy, Falkirk High and Perth, highlighting the role of shopping and other leisure trips from these stations.
4.20 These results illustrate some interesting trends, particularly when examined alongside the journey purpose analysis. In particular, the percentage of respondents purchasing season tickets from Kirkcaldy, Falkirk High and Perth is lower than the proportion of commuters, implying some may be travelling less than 5 days a week and therefore choosing to buy alternative products. The percentage buying off-peak tickets at Bridge of Allan and Stirling is low (less than 20%), highlighting the importance of peak flows.
Figure 4.3: Ticket type based on results of passenger interviews
Source: Arup analysis of Accent data, sample size shown
Parking costs and the availability of spaces
4.21 The stations with extended car parks are generally free, with all respondents at Kirkcaldy and Bridge of Allan benefitting. In Perth, about half of the travellers parked for free. Charges are supposed to be levied, but some travellers do not pay since they perceive the likelihood of checks to be minimal. Over 90% of traveller's park free of charge at East Kilbride. The selected stations without extensions have a higher proportion of chargeable spaces. At Falkirk High and Stirling, only 1% parked for free. Other respondents had paid £2.50 per day (Falkirk High) and £3.00 per day (Stirling). Table 4.2 presents the results.
| No charge | Less than 80p | £1.50 | £2.50 | £3.00 | £3.50 | No. Of Responses | |
|---|---|---|---|---|---|---|---|
| Kirkcaldy | 100% | 101 | |||||
| Bridge of Allan | 100% | 115 | |||||
| East Kilbride | 92% | 8% | 107 | ||||
| Perth | 50% | 3% | 47% | 36 | |||
| Falkirk High | 100% | 64 | |||||
| Stirling | 1% | 99% | 64 |
Source: Arup analysis of Accent data
4.22 Respondents were interviewed to explore their perceptions of parking availability. There are noticeable differences between the results, with at least 70% of respondents always finding a space at Kirkcaldy and East Kilbride. This suggests the car park extensions at East Kilbride and Kirkcaldy have alleviated the previous constraints regarding the availability of spaces. Furthermore, nearly 60% of respondents at Perth always found a space. In contrast, only 20% of users at Bridge of Allan, Falkirk High and Stirling could always find a parking space. This highlights the popularity of the car park at Bridge of Allan. The introduction of parking charges at Dunblane may have increased the demand for station parking at Bridge of Allan. About 40% of users at Bridge of Allan and Stirling had problems finding a space on at least 3 occasions every 10 visits. Figure 4.4 presents the results.
Figure 4.4: Likelihood of difficulty in parking (every ten visits)
Source: Arup analysis of Accent data, sample size shown
4.23 The occurrence of parking problems affecting peak versus off peak passengers did not highlight any specific problems. Exploring the results in terms of commuting versus leisure travel as a proxy showed no difference in perceived availability of parking spaces. However, anecdotal evidence did highlight some car parks filling up early and this coincided with commuting patterns in the morning peak. Interviews with passengers arriving in the morning peak and who could not park were partially examined as part of the non-user surveys. Larger sample sizes would, however, be required to explore this issue more fully.
Awareness of car park extensions and attitudinal responses
4.24 The awareness of the improvements and the subsequent passenger behaviour in this section was only examined at Kirkcaldy, Bridge of Allan and East Kilbride.
4.25 The majority of respondents surveyed at Kirkcaldy (84%) and East Kilbride (66%) were aware of the improvements to these station car parks. The results for Bridge of Allan indicate an entirely different conclusion, with 61% of respondents not aware of the improvements. At least 80% of respondents stated they would have used their current station, even if additional parking spaces were not available. The results are shown in Figure 4.5.
Figure 4.5: Awareness of improvements and changes in user behaviour
Source: Arup analysis of Accent data, sample size shown
4.26 Figure 4.6 illustrates the change in travel behaviour for respondents who used the station prior to the improvements, and their mode of access. Only one-third of respondents interviewed at Kirkcaldy previously used the car park, with about 45% parking elsewhere. The results suggest there are fewer alternative parking choices available for passengers using Bridge of Allan and East Kilbride. With the exception of Kirkcaldy, the number of 'other' responses, including walking to the station or catching a bus or taxi was very small.
Figure 4.6: Behaviour before car park was extended: existing users
Source: Arup analysis of Accent data, sample size shown
4.27 Figure 4.7 presents the travel behaviour for respondents who are only using rail in response to the car park improvements. In contrast with the results presented in Figure 3.8, it is worth highlighting the relatively small sample size. Around 20-30% of users at Bridge of Allan and Kirkcaldy used to drive, indicating the expanded car parks have removed some trips from the network. Furthermore, 30-40% of users previously drove to a different station, so the extension may have reduced the car distance to reach the station. This result may also have been influenced by the introduction of both on-street parking restrictions and charging at the car park adjacent to Dunblane Station by Stirling Council. (The sample size for East Kilbride is too small to draw any meaningful conclusions.)
Figure 4.7: Travel behaviour before car park was extended - new users
Source: Arup analysis of Accent data, sample size shown
4.28 About 50% of respondents at Kirkcaldy, Bridge of Allan and East Kilbride stated they would park elsewhere if the number of spaces was reduced. Over 40% of users at East Kilbride would travel from a different station, and this reflects the close proximity of alternatives. Passengers interviewed at Falkirk High and Stirling indicated they would use a different mode of access to these stations. This reflects the improved journey opportunities available (higher frequencies, faster journey times) compared with adjacent stations. The results are shown in Figure 4.8.
Figure 4.8: Alternative travel behaviour if parking availability is constrained
Source: Arup analysis of Accent data, sample size shown
4.29 The importance of CCTV varies at the six stations. In Kirkcaldy, Bridge of Allan, East Kilbride and Perth, the majority of respondents would still use the car park even if CCTV was not available. The availability of lighting is of greater concern when compared with CCTV, particularly at Kirkcaldy and East Kilbride. Whilst the majority of respondents using Bridge of Allan and Perth would continue their current behaviour even if the car park was not lit, they are located close to other pedestrian links, helping to remove some safety concerns. Only one-third of users from Kirkcaldy and East Kilbride would continue to use the station if lighting was removed, although a higher proportion of respondents would use East Kilbride during the summer / daylight conditions. At two of the control stations (Falkirk High and Stirling), over 90% of respondents stated they would not use the car park if there was no lighting.
4.30 The impact on user behaviour was examined if CCTV and lighting were not available, along with no tarmac on the road surface. At Kirkcaldy, Bridge of Allan and East Kilbride about 50% of respondents would not use the car park, although 15-30% would use Kirkcaldy and East Kilbride during daylight hours. Almost 97% stated they would not use Falkirk High and Stirling if these facilities were removed. Figure 4.9 illustrates the results.
Figure 4.9: Willingness to use the car park without CCTV, lighting and an untarmaced road
Source: Arup analysis of Accent data, sample size shown
Results from the modelling outputs
Link between parking availability and demand
4.31 The outputs from the econometric analysis illustrates that a 10% increase in parking spaces would lead to a 0.43% increase in season ticket trips and a 0.35% increase in non-season ticket trips based on the sample of data analysed. The impacts are smaller for local journeys less than 20 miles. The impact for the case study stations is shown in Table 4.3 and is presented in terms of the number of additional rail trips per day when the elasticities are applied. This equates to 12 wholly new trips per day from Kirkcaldy and 3 extra trips per day from Bridge of Allan. The extra demand from Perth and East Kilbride is 11 daily trips. The results from the case studies could be applied to other stations, although scheme promoters need to ensure the characteristics of the stations to be expanded are similar to the above examples, only with due attention to local circumstances, given that the demand impact will vary according to the specific characteristics of each station.
4.32 The modelling methodology estimates the change in demand based on percentage based adjustments. Stakeholders have also requested this impact be expressed in terms of the change in demand which is specifically related to the number of parking spaces provided. The relationship between the number of parking spaces before and after the expansion and the total number of new trips, suggests each additional 100 spaces generates between 4 and 10 extra journeys per day.
| Station | Estimated Number of Daily Trips |
Change in season ticket trips / day |
Change in non-season tickets /day |
Total | Number of additional parking spaces |
Extra daily trips per 100 parking spaces |
|---|---|---|---|---|---|---|
| Kirkcaldy | 3,444 | 4 | 8 | 12 | 320 | 3.75 |
| Bridge of Allan | 754 | 1 | 2 | 3 | 32 | 9.38 |
| East Kilbride | 2,743 | 4 | 7 | 11 | 125 | 8.80 |
| Perth | 2,848 | 4 | 7 | 11 | 106 | 10.38 |
Source: Arup analysis of ITS and Office of Rail Regulation (2009/10 Station Counts). Trip generation based on 30% season tickets and 70% non-season tickets
4.33 The econometric modelling estimated the change in demand if the likelihood of finding a space was reduced as shown in Appendix A1. If the likelihood of not finding a space increased from 0% to 20%, this would lead to a 4.3% reduction in rail demand. If the likelihood of not getting a space increased to 10%, demand would be reduced by 2.2%. The model outputs indicate there would be virtually no increase in non-season ticket demand if parking availability was increased and there were ample free alternatives to the station car park.
4.34 A range of scenarios were tested to examine the impact of changes to parking charges. These tests include the introduction of a parking charge to £1 for locations where parking was previously free, or increasing the parking charge to a higher value, for example, from £1 to £2. Responses have been used to assess the percentage using rail which would not alter their travel behaviour, and the proportion parking elsewhere or switching to another mode depending on absolute changes to the parking charges. The relationship shown is non-linear, since each respondent was not presented with every price increase. If the parking charge is increased by £1, park and ride demand would be reduced by 4.9%. A 3.0% reduction in park and ride usage would occur if there is ample free parking available in an alternative location nearby.
4.35 The analysis presented above highlights the main issues emerging from the user surveys and the modelling outputs. The purpose of this section is to consider how these conclusions help to address the overall study objectives and these are set out below.
- objective 1 (changes to parking supply and pricing affecting public transport usage): Outputs from the model indicated a 10% increase in parking spaces would lead to a 0.43% increase in season ticket trips and a 0.35% increase in non-season ticket trips. Table 4.3 illustrates the impact on rail demand for the case study stations. Furthermore, if parking charges were increased by £1 (either from free to £1 or £1 to £2), rail demand would be reduced by 4.9%. A 3.0% reduction would occur if there is ample free parking available in an alternative location nearby. In the absence of sufficient parking at a station, the majority of passengers will use alternative solutions to access it
- objective 2 (scope of parking supply and cost to influence modal shift): The scope to influence overall mode choice if station parking was increased is relatively small. If additional parking was available, about 1.5% new trips would be generated. However, the wider positive impacts in terms of congestion and emissions will be smaller since the increased availability of parking has encouraged some existing users to drive to the station rather than opting to walk, cycle or catch the bus
- Based on the survey results using Kirkcaldy as a case study (the sample size is larger compared with Bridge of Allan or East Kilbride), the additional car trips travelling to the station as result of the extended car parks rather than using alternative modes is broadly equal to the reduction in car distance. This estimate of car distance removed from the network is based on the level of new rail trips generated from the parking improvements, the proportion switching from car based on the survey results and the assumed trip length from Kirkcaldy. This is similar to the journeys previously travelling to Kirkcaldy by bus, cycle or on foot and an assumed average distance to the station, again based on survey results. As a result, the change in car kilometres removed from the network resulting from the expanded car parks is negligible. However, this conclusion may be different depending on the characteristics of each station
- objective 3 (relative importance of complementary factors): Conclusions from stakeholder feedback, along with the survey results from park and ride travellers using the station regarding the impact of 'softer' measures, help to illustrate the relative importance of CCTV, lighting and a paved road. Some of these factors were assessed individually or in groups. For example, discussions with FSR highlighted the wider role of employment and other factors generating demand, the impact of timetable changes and the availability of spare capacity. The benefits of CCTV, lighting and other measures to create a safe waiting environment were also highlighted based on results from the primary research. The absence of such measures would reduce the attractiveness of park and ride
- objective 4 (identification of undesirable outcomes): The review of evidence to address Objective 2 highlighted the negligible change in car kilometres removed from the network if railway station car parks were expanded using Kirkcaldy as a case study. The mode of access before and after the changes to parking availability indicated around 20% of the sample at Kirkcaldy previously walked, cycled, took the bus, or car shared
- objective 5 (financial metrics to guide appraisal): The capital cost to extend existing car parks is about £5,000-£10,000 per space, although this could be higher if decking is required. Based on the estimated levels of trip generation using Kirkcaldy as a case study (about 26 new rail trips per day, based on 260 days per year) and the cost of an average return rail fare (based on MOIRA data for all journey purposes £5.57, 2009/10 data). The revenue stream that could be generated from additional rail passengers would generally be insufficient to provide a financial pay-back in less than 10 years, though there may be instances when this is possible.
- objective 6 (optimum pricing policy): the analysis highlighted the change in demand if prices were increased by £1 (for example, either from free to £1, or £1 to £2). Rail demand would be reduced by 4.9% if prices were increased by £1 or 3.0% if there is ample free local parking. Furthermore, about 55% of the remaining rail passengers would park elsewhere. Regardless of whether the car park is free or already charged, the increased parking charges would mean the revenue loss from rail passengers switching to other modes would exceed the income from the newly introduced parking charges. The revenue impacts would be even greater if existing parking charges were raised