The Effects of Park and Ride Supply and Pricing on Public Transport Demand
5 Bus based park and ride
5.1 One of the main objectives for bus based park and ride is encouraging car users travelling to large urban centres to transfer onto public transport for part of their journey. Sites have been introduced at various locations across Scotland and the rest of the UK. Similar to the rail based schemes, these examples can also support wider objectives. As well as economic and social benefits, bus travel has significant potential to lower CO2 emissions through modal shift from car.
5.2 Similar to the overarching study methodology described in chapter 2, the bus case study uses a combination of existing secondary data with conclusions from new primary research to address the objectives. Existing literature was summarised, with two case studies carefully selected based on their current usage to ensure the overall sample sizes were sufficient. Outputs were incorporated into the forecasting models, with the combined analysis being used to determine the factors which contribute to a successful scheme.
Secondary research - overview of existing park and ride sites
5.3 In total, there are about 150 sites across the UK, with about 31m passenger journeys per annum in 2005/0615. About 70,000 parking spaces are available. Passenger usage has been used as a proxy to examine the success or otherwise of a scheme. The schemes attracting the highest number of passengers include York (almost 20% of the UK total), Norwich, Cambridge (both sites attract over 3m trips per annum) and Chester (about 2.5m trips). There have been relatively few publications since 2007 which examine the performance of park and ride on a national basis. Some local authorities monitor statistics for individual schemes, but much of the comparative analysis was collated before 2007.
5.4 The performance of other UK sites to identify the factors contributing to the success of park and ride has been examined to highlight any weaknesses that mean individual sites perform less well, so the lessons learned can be avoided when developing other proposals. Table A5.1 (in Appendix A5) collates scheme best practice from other sites with the indicators described below, supplemented by other measures. The main conclusions set out below form a subjective assessment of the contributory factors which have contributed to overall best practice:
- size of potential catchment: the most successful schemes serve an urban centre with a population of at least 100,000 people. The traffic levels diverting from the strategic road network approaching the urban centre are calculated in the 'Guidance to Scheme Promoters' section
- location: sites need to maximise their proximity to the potential catchment, with minimal delays from congestion. Effective signing is also required. For example, the bus based park and ride site at Falkirk is poorly located in relation to the strategic network and this reduces its attractiveness. Furthermore, one of the main factors contributing to the low usage of Hanley (Stoke-on-Trent) is its location, just 500m from the centre given the surrounding congestion
- service frequency: departures every 10 minutes mean passengers do not need a timetable. There are a small number of examples that operate at a lower frequency, but a service every 12-15 minutes is less attractive for short distance journeys. High Wycombe is served by just 4 buses per hour throughout the day and this is not frequent enough, particularly for short distance trips, given the site is located less than 1 mile from the town centre
- role of Demand Management: as noted earlier in the case studies, a successful park and ride is an integral component of urban transport policy. The most successful schemes are integrated into an overarching demand management strategy including the cost and availability of parking, especially in historic cities and towns. This is evident from the most successful English schemes which are predominantly located in historic cities
- costs: the ratio of bus fare using the park and ride to the typical peak and off-peak charges has been collated. Generally, a lower ratio indicates the cost of the park and ride is more competitive versus urban centre parking. Edinburgh and Norwich have the lowest ratios for charged spaces. The sites in Hanley and Ipswich have the least competitive pricing strategy for park and ride compared with central parking costs
5.5 In addition to the indicators presented in Table A5.1, the impact of bus priority measures to deliver reliable, competitive journey times, plus branding of vehicles are highlighted in the TAS report, although it is recognised these can be more difficult to quantify. The purpose of highlighting the other two items is to demonstrate there are some specific tasks which are supplementary to the issues flagged in Appendix A5.
5.6 There are a number of attributes which appear to have contributed to the operation of a successful park and ride. The historic characteristics of some cities, including the shortage of central parking, pricing strategy and reductions in road space, has contributed to their wider success. Park and ride sites can help to control congestion levels, particularly in historic centres where the generation of additional traffic could have a detrimental impact on the performance of the road network if it already operates close to capacity.
5.7 Schemes in Scotland generally attract a lower number of trips per annum compared with many English examples. Whilst the number of sites serving the major Scottish cities is generally lower compared with the most popular English examples, the number of trips per site is also less. The four schemes serving Edinburgh generate around 400,000 trips per annum, whereas five sites at York attract about 6m journeys. Although Edinburgh has a larger population versus York (about 450,000 people compared with 180,000 respectively), York receives double the number of tourists each year (7.1m visitors compared with 3.5m for Edinburgh). The higher number of tourists visiting York may have contributed to the increased park and ride use16.
5.8 We have reviewed two cities in England to understand why some English schemes attract a higher number of passengers compared with examples in Scotland. Norwich and York17 offer a good cross-section of evidence.
Norwich: A strategy was gradually implemented in the early 1990s to address the worsening congestion problems, reduce emissions and improve road safety. Six sites located close to the main strategic routes have been delivered. Although some sites are close to railway stations, buses offer a higher frequency so the scope for competition with other public transport is limited. Norfolk County Council has placed strong emphasis on quality. Sites have good lighting and security and are staffed. NCC provides financial support for the park and ride sites, choosing to prioritise the deployment of a modern, high quality fleet rather than adopt the measures to reduce costs. Bus fares are competitively priced compared with car parking in the centre. Park and ride fares are expressed as a charge per car to encourage family groups. Dedicated bus services operate every 10-12 minutes. Some bus priority measures have been introduced on some corridors. The network of sites has helped to reduce traffic levels crossing a city centre cordon. Cost, convenience and the limited availability of alternatives were identified as contributory factors.
York: The city attracts over 7 million tourists per year and is a prime retail location serving a large catchment in North Yorkshire and beyond. With inadequate space in the city centre to accommodate demand, park and ride sites have been introduced incrementally. As a result, the majority of main radial corridors are now served. The location of sites is adjacent to the outer ring road, minimising delays for car drivers. Dedicated buses depart every 10 minutes. First operates services commercially, with modern low floor vehicles. Articulated vehicles are used on some routes to support high commuting flows. Some priority measures have been introduced, although traffic congestion is becoming a constraint. Free parking is available, with competitively priced fares compared with the cost of city centre parking. Some car parks are full before 10am, despite 3,750 spaces being available. In parallel, the City Council has enforced a number of demand management measures. The City Council also has ambitious plans to further expand its park and ride offer. A funding proposal has been submitted to expand the number of spaces at an existing location and create two new sites.
5.9 These case studies have been compared with Ingliston and Bridge of Don18.
Ingliston, Edinburgh: Edinburgh is a major trip destination attracting commuters, shoppers and tourists. The city is served by several bus based park and ride sites, mainly located close to the A720 bypass. Monitoring surveys conducted by Edinburgh Council indicate users think services offer good value for money. However, there is limited evidence to demonstrate the impact of park and ride on wider traffic levels. Although Ingliston has over 1,000 parking spaces and served by frequent buses, the occupancy rate is just 50%.
There are several contributory factors including:
- its location on the A8 means some drivers (especially those using the M8 extension) will bypass the site
- lack of bus priority measures mean journey time reliability is poor
- there is limited opportunity to influence service patterns and market the service as buses are operated commercially (this applies to York too)
- there is significant competition with other public transport, especially rail
Bridge of Don, Aberdeen: Dedicated buses operate as a shuttle via Aberdeen city centre to Bridge of Don every 10 minutes during the peak periods (although this drops to every 15 minutes during the off-peak). The site is located close to the A90, about 3 miles from the city centre. Buses benefit from a range of priority measures to achieve reliable journey times.
In contrast with many other UK schemes, passenger numbers have declined sharply during the last five years. For example, the number of trips has declined from 300,000 per annum in 2000 to just 170,000 in 2006. Similar to the Edinburgh example, there are a number of factors that have contributed to this outcome including:
- slight decline in traffic levels using the A90 corridor towards the city centre that would be in-scope for park and ride
- competition with the Ellon park and ride located about 15 miles north of Aberdeen which operates as an inter-urban service. The Ellon site attracts about 80,000 trips per annum and partially overlaps with Bridge of Don
- lower parking charges enforced by the city council in the city has meant city parking is more attractive, based on price, compared to the park and ride site. This makes park and ride at Bridge of Don relatively less competitive
- staff presence at the site has been removed in response to funding cuts
5.10 The results from the secondary analysis highlight some of the main characteristics associated with successful park and ride sites. These conclusions have been used to identify some of the complementary factors and the financial metrics to guide scheme appraisal. We now turn to the primary research which was undertaken to help us understand the characteristics of demand for the Scottish sites and which will be interpreted in the context of the secondary analysis outlined above.
Primary research - overview of findings
5.11 Sites at Bridge of Don near Aberdeen and Ingliston near Edinburgh were chosen to undertake primary research to try and understand the differences in usage. This selection also offered a reasonable geographic coverage, a mixture of catchment sizes and different city centre parking strategies. Both park and ride sites selected for primary research serve eastern Scotland. However, there is a comprehensive rail network serving Glasgow, so the opportunities for complementary bus based park and ride schemes are reduced. Whilst there are several examples serving the Central Belt, their level of usage varies and ensuring a robust sample size formed one of the considerations when assessing the suitability of potential sites.
Journey times to the park and ride
5.12 Figure 5.1 indicates almost 60% of respondents interviewed at Bridge of Don and Ingliston spent less than 20 minutes travelling to the park and ride, whilst a further 15% travel less than 30 minutes. Interestingly, over 20% of respondents travel for more than 40 minutes, including 10% for more than hour. Bridge of Don has a higher number of respondents with a journey time less than 10 minutes, and a shorter travel time to the park and ride (about 20 minutes compared with 25 minutes to Ingliston), implying the catchment to the Bridge of Don is smaller.
Figure 5.1: Journey Times to the Park and Ride
Source: Arup analysis of Accent data, sample size shown
Catchment analysis - trips to the destination
5.13 A significant proportion of trips using the Ingliston park and ride have a destination in Edinburgh city centre with 'The Gyle' also a popular location, as shown in Figure 5.2. Other destinations in Edinburgh are less convenient by park and ride, explaining why less than 15% have a destination in these locations. The concentration of trips to central Aberdeen is even more pronounced compared with Edinburgh, with over 95% to the city centre.
Figure 5.2: Distribution of trip destinations
Source: Arup analysis of Accent data. Sample size shown
5.14 Figure 5.3 indicates the most popular journey purposes are commuting and shopping. For both Ingliston and Bridge of Don, journeys to work account for almost 50% of total trips, with shopping accounting for a further 25%. The 'other' journey purposes account for a relatively small proportion of the total.
Figure 5.3: Journey purpose
Source: Arup analysis of Accent data, sample size shown
Availability of parking spaces
5.15 In Figure 5.4, over 95% of users had no problems finding a space at both sites. Of the 5% of users that did report problems, these occurred about once a week so the incidence of parking problems is very small. Monitoring data collected by Aberdeen and Edinburgh Councils which illustrates the usage of sites endorse this conclusion19 .
Figure 5.4: Availability of car parking
Source: Arup analysis of Accent data, sample size shown
5.16 The type of ticket purchased is shown in Figure 5.5. In total, about 80% of passengers use single or return tickets. Usage of alternative tickets is relatively small, with less than 15% using day tickets or other (multi-modal) products. Passengers using the Bridge of Don park and ride generally used return tickets, with most people from Ingliston using single tickets.
Figure 5.5: Type of ticket used
Source: Arup analysis of Accent data, sample size shown
5.17 The maximum bus fare people are willing to pay is presented in Figure 5.6 and is based on the current fare plus a series of potential fare increases from £0.50 up to £3.50. Across both sites, almost 25% would not be willing to pay anything extra, suggesting some users have available free or low cost parking. About 30% could pay up to £1.00 extra to use the park and ride, implying the cost to park in Aberdeen is relatively cheap. Interestingly, about 12% of users in Edinburgh would be willing to pay over £3.00, indicating the alternative parking choices are very expensive.
5.18 It is useful to consider these alternative costs in relation to the park and ride fares which can range from £2.10 to £2.30 for regular users (infrequent users will pay more). Furthermore, it is important to acknowledge there are other contributory factors influencing travel behaviour including the convenience of using park and ride relative to driving into the city centre and the scope to avoid traffic congestion.
Figure 5.6: Willingness to pay higher bus fares for park and ride
Source: Arup analysis of Accent data, sample size shown
5.19 The impact on travel behaviour if car parking spaces were removed is explored. The results demonstrate the availability of parking is a fundamental factor influencing the attractiveness of park and ride. There is a small number of current park and ride users who could switch to a local bus service for the whole of the journey, but the majority of existing users would choose an alternative mode in the absence of sufficient parking. These results are presented in Figure 5.7
Figure 5.7: Change in travel behaviour if parking was not available
Source: Arup analysis of Accent data, sample size shown
Passenger safety considerations
5.20 The attitudinal responses of measures to improve safety or quality at the park and ride site were also considered, with the percentages relating to the total figures. The first topic considers user responses if CCTV was not available. Nearly 80% of respondents at both sites would continue to use the park and ride if CCTV was not available. The availability of lighting appears a more important issue. Less than 50% of respondents at both sites would use the park and ride if the site was poorly lit, although some users would only be willing to use the site during the summer. The combination of no tarmac road to the site, no lighting and no CCTV has a detrimental impact, with just 30% of respondents willing to use the park and ride. A further 40% at both sites would only use the site in the summer, in response to the security issues. Figures 5.8 - 5.10 illustrate the results.
Figure 5.8: No CCTV available - would you continue to use the site?
Source: Arup analysis of Accent data, sample size shown
Figure 5.9: No lighting available - would you continue to use the site?
Source: Arup analysis of Accent data, sample size shown
Figure 5.10: No CCTV, lighting and no tarmac road - would you continue to use the site?
Source: Arup analysis of Accent data, sample size shown
Primary research - overview of non user behaviour
5.21 A total of 120 interviews with non-users were completed. These discussions were conducted in Aberdeen and Edinburgh city centres to understand the factors that influenced their travel behaviour. Respondents were carefully screened to ensure the characteristics of their journey meant they could switch to park and ride.
Existing car journey times to the park and ride site
5.22 Respondents were interviewed to understand journey times to central Aberdeen or Edinburgh. The interviewees made relatively long access trips to both sites compared with drivers already using the park and ride, with over 40% travelling over 40 minutes. The percentage of respondents making short trips is small, with no-one interviewed having a travel time to the park and ride of less than 10 minutes to either site.
5.23 One of the main factors affecting the choice between driving or park and ride is the cost and availability of car parking in the city centre. About 55% of respondents have free parking available, predominantly in central Edinburgh and this is a major factor influencing their travel choice. Charges for public car parks are set by the individual Councils and these comparisons should help to inform the pricing strategy for park and ride.
5.24 The most popular journey purposes are commuting and shopping accounting for over 50% of the total trips.
Analysis of the modelling outputs
5.25 The results from the Stated Intention surveys were analysed to understand the proportionate change in demand in response to specific changes in bus fares, frequencies or the likelihood of not getting a parking space. A more detailed overview of the modelling methodology is presented in Appendix A4.
Likelihood of getting a space and impact on demand
5.26 The elasticity relating to the chance of not obtaining a parking space is reported. The impact of not finding a space has been modelled. Therefore, moving from 0-10% chance of not finding a space would lead to a 19% reduction in bus demand, whilst a 20% chance of not finding a space would result in a 34% reduction in bus demand at the park and ride sites. These are considerably higher than the figures produced for rail and probably reflect that the park and ride sites at Ingliston and Bridge of Don currently have a large amount of spare parking capacity at all times, whilst there is a higher number accessing these park and ride sites by car. Consequently, users will be more sensitive to any changes in parking availability.
5.27 The pricing elasticity was tested to see if revenue maximising fares could be identified, since the bus fare elasticity of -0.306 reported in Appendix A2 would indicate that in the short term at least, there is scope to increase fares (a 10% increase in fares would only reduce demand by 3%, resulting in increased revenue). A number of further fare elasticities were calculated to see whether the data allowed for the identification of a revenue maximising price. A range of price increases to the base fare of £3 (return) were considered, ranging from 50p to £5 (i.e. larger than an operator would consider). The corresponding bus fare elasticities were found to range from -0.55 to -0.75 respectively.
5.28 This would suggest a revenue maximising fare cannot be estimated from the data due to the limitations of the survey results. Results were inconclusive from the survey data. In the short term, bus demand is very inelastic, that is to say current passengers find it very difficult to find an alternative mode of transport to make the same journey they currently make in the short term. In the medium to long term we would expect the bus elasticity to become less inelastic as people respond to the bus fare increases by using other modes or changing their destination, so the scope for making large changes to the bus fare is lessened.
5.29 The analysis presented in Chapter 5 highlights the main issues affecting the viability of bus based park and ride. The detailed review of the case studies, plus the other secondary data collated and the conclusions emerging from the primary research, has highlighted some important issues to be cognisant of when developing future bus based park and ride policies:
- objective 1 (changes to parking supply and pricing affecting public transport usage): since there is sufficient parking space at the existing car parks, this objective is evaluated in terms of the impact on demand if spaces were removed. There are two datasets collated as part of this study that explores the relationship between these variables. The outputs from the modelling work suggested a 10% chance of not finding a space would lead to a 19% reduction in bus demand, whilst a 20% chance of not finding a space would lead to a 34% reduction. Furthermore, the results from the primary research indicated over 60% of respondents would make their entire journey by car if there was insufficient parking available. Both datasets clearly demonstrate the importance of ensuring adequate parking is available throughout the day. Car users will continue their journey if sufficient parking facilities are not available
- objective 2 (changes to parking supply and pricing influencing modal shift): the monitoring data from City of Edinburgh Council indicates there are about 1,200 park and ride trips per day using all park and ride sites (Hermiston, Ingliston and Straiton) and this equates to less than 1% of the trips crossing the A720 for journeys towards the city centre. As a result, the impact of park and ride on wider modal shift is relatively small. The scale of congestion relief benefits is dependent on wider network performance, since any traffic diverting from roads that are operating close to capacity will have a greater impact compared with uncongested routes. Based on the current usage of most park and ride sites in Scotland, the impact on congestion and emissions will be relatively small
- objective 3 (relative importance of complementary factors): there are a range of criteria that determine the level of use associated with individual sites. Using the results from the primary research, combined with the case study analysis, these attributes include accessibility to the strategic road network, availability of parking spaces, opening times and the frequency of bus services, plus the relative competitiveness of park and ride in terms of journey time and cost versus the alternative journey made by car
- objective 4 (identification of undesirable outcomes): compared with the rail market, there is very limited evidence of undesirable outcomes generated from bus based park and ride. The results from the primary research indicate some respondents chose to drive to Ingliston to catch the park and ride rather than using their local bus service. A combination of higher bus frequencies and cheaper fares compared with their local service contributed to this decision making. However, these disbenefits are offset by the number of car kilometres removed from the network. The number of car trips removed from the network is 2.5 to 3 times higher than the additional mileage generated from respondents driving to the park and ride to take advantage of the bus services available there
- objective 5 (financial metrics to guide scheme appraisal): using outputs from a number of existing schemes, the operating costs for services departing every 10 minutes range from £800,000 to £1m per annum for each site depending on the operating period. Site operating costs vary depending on the specification offered and could range from £75,000 to £100,000 per annum. Depending on the overall commercial position, these costs would be attributable to the operator and / or Council. It is assumed the capital costs will be funded through Local Transport Plan investment. The value of removing car trips from the network in terms of decongestion benefits, accident savings and emissions should be quantified to make the case for the investment
- for a site to break even in financial terms, about 1,200 passengers per weekday would be required based on the operating costs described above. This is a high number of trips, and highlights the importance of locating park and ride in the optimum position to achieve a robust financial case. Based on previous consultancy studies20 , the percentage of drivers passing the site with a city centre destination ranges from 20-40%. It is estimated park and ride attracts between 10-40% of motorists depending on the characteristics of the schemes and the parking structure enforced in the urban centre. Car occupancies of about 1.45 persons could be assumed based on STAG guidance for AM peak trips21 . This equates to about 20,00022 vehicles per day using roads adjacent to the site (total cars * in-scope trips * mode share / car occupancies
- objective 6 (optimum pricing policy): analysis of the price elasticity to identify the optimum price structure was inconclusive. Incremental bus fare increases were considered, and the corresponding elasticities were found to range
- from -0.55 to -0.75. The revenue maximising fare cannot be estimated from the data since the results from the survey data are inconclusive. Whilst passengers may find it difficult in the short term to find an alternative mode, bus elasticity would become less inelastic in the longer term as people respond to the bus fare increases by using other modes or changing their destination. Consequently the scope to make changes to fares is lessened, though not in the short term