Turning flows at Motorway Service Areas

Transcription

1 TRANSPORT RESEARCH LABORATORY Turning flows at Motorway Service Areas Prepared for QS, Traffic Safety and Environment Division (TSE), Highways Agency S M Gray, M Al-Katib and G S Buckle TRL REPORT 441

2 First Published 2000 ISSN Copyright Transport Research Laboratory This report has been produced by the Transport Research Laboratory, under/as part of a Contract placed by the Highways Agency. Any views expressed are not necessarily those of the Agency. TRL is committed to optimising energy efficiency, reducing waste and promoting recycling and re-use. In support of these environmental goals, this report has been printed on recycled paper, comprising 100% post-consumer waste, manufactured using a TCF (totally chlorine free) process. Transport Research Foundation Group of Companies Transport Research Foundation (a company limited by guarantee) trading as Transport Research Laboratory. Registered in England, Number TRL Limited. Registered in England, Number Registered Offices: Old Wokingham Road, Crowthorne, Berkshire, RG45 6AU.

3 CONTENTS Page Executive Summary 1 1 Introduction 3 2 The 1994 data 3 3 The 1997 study 4 4 The 1998 extension to the study 4 5 Results The video surveys The questionnaire survey Journey details Reasons for stopping at MSA Driving experience Attitudes and comments 12 6 Discussion and conclusions 12 7 Acknowledgements 13 8 References 13 Abstract 15 Related publications 15 iii

4 iv

5 Executive Summary The Transport Research Laboratory (TRL) has undertaken a major research project for the Highways Agency entitled Motorway junction layout to increase capacity and safety at low cost. As part of this project TRL have been requested to look at the factors affecting the turning flows at Motorway Service Areas (MSA). A literature search has been carried out and details have been obtained of the facilities offered and gross floor areas (GFA) at a total of six MSAs included in video/questionnaire surveys in 1994 and One-day video surveys were carried out at four MSAs in December Both video surveys and questionnaire surveys were carried out at two different MSAs in July/ August Analysis of the video data for these six MSAs indicated a relationship between turn-in percentage and distance from previous services. To further investigate this relationship and to make the overall findings more robust, video surveys were carried out at a further six MSAs, early in In addition, an inventory of parking provision at all 12 sites studied was compiled in order to be able to test the hypothesis that a motorist s decision to turn into an MSA might be affected by his past experience of the availability of parking at that MSA. The main findings of these surveys are as follows: l Main reasons given for deciding to stop at the MSA were for rest, refreshment, toilets, and fuel. At both sites, the main reason for stopping at the particular MSA at which interviewing took place was convenience of time/distance. l Overall, 43% of respondents said they only use MSAs when absolutely necessary and a further 15% said they only use them when on business. General (volunteered) comments revealed that at least 45% of respondents considered MSAs to be expensive/overpriced. This study has confirmed the view that a motorist s decision to turn into a Motorway Service Area is influenced by his/her journey length, journey purpose, location of an MSA relative to origin and destination, the availability and proximity of other MSAs or alternative facilities off the motorway, and the amount of parking space available at an MSA, based on previous experience. Video surveys l Analysis of half-hourly turn-in flow and main flow recorded over a 12 hour period in both directions at each of the twelve MSAs shows a variable relationship between the two. Usually the relationship is positive but in a number of cases it was negative. l The relationship between turn-in percentage and main flow was less variable and almost always negative, i.e., the larger the main flow, the smaller the proportion of vehicles turning in to an MSA. l A stepwise regression analysis has indicated a diminution of turn-in percentage as main flow expands, an increase in turn-in percentage as distance to/from other MSAs increases, and a decrease in turn-in percentage as distance to/from junctions increases. Also, the greater the total number of parking spaces, the higher the turn-in percentage is likely to be. Questionnaire survey l A total of 2082 completed questionnaires were obtained at two Motorway Service Areas. Of the sample population, 64% were on a business trip at the time of interview, 33% were travelling for pleasure, and the remaining 3% were combining the two. Overall median journey length was 165 miles and median journey time was 3 hours 15 minutes. The mean time travelling since the last stop was 1 hour 34 minutes. 1

6 2

7 1 Introduction The Transport Research Laboratory (TRL) has undertaken a major research project for the Highways Agency entitled Motorway junction layout to increase capacity and safety at low cost. As part of this project TRL was requested to look at the factors affecting the turning flows at Motorway Service Areas (MSA). Department of Transport Circular Roads 4/88 gave advice on the control of development on trunk roads, and DOT Roads Circular 1/94 supplemented that advice so far as the development of Motorway Service Areas was concerned. Further advice on the considerations which local planning authorities should take into account when considering planning applications for service areas on motorway and trunk roads was given in Planning Policy Guidance Note 13 Transport (PPG13). In August 1992 the Government announced that in future the private sector should take the initiative in identifying and acquiring MSA sites and seeking planning consent from local planning authorities, but the Department would continue to have interests in MSAs in relation to motorway safety and traffic management. MSAs were initially intended to be spaced at a frequency of one per 30 miles. This frequency is now nearer one per 15 miles with pressure from developers for an increased frequency. The Highways Agency requires information on the traffic generation effect, associated weaving flows and junction design aspects of MSAs. The Highways Agency, TSE Division, wishes to know how one estimates the flow into/out of a proposed MSA. There are many factors that might determine the traffic flow into/out of an MSA. These include: l traffic volumes on the main motorway upstream from that service area; l the distance (and subsequent journey time) between service areas; l the facilities available within the service area; l the gross floor area (GFA) of the service area; l motorway layout and weaving section length (e.g. the closeness of another junction) in the vicinity. TRL collected traffic flow information in December 1994 at four locations. Further studies were carried out in 1997 and 1998 with the aims of: i supplementing the 1994 data with additional information such as gross floor area, and type of facilities offered at each MSA; ii collecting further up-to-date information, by means of questionnaire and video surveys, either at the same MSAs as in 1994 or at selected alternatives; iii producing guidelines for estimating traffic flows into and out of an MSA. Because of the difficulty experienced in obtaining the approval of all the operators for the publication of all data collected, individual MSA s are not identified in this report. 2 The 1994 data A study of MSAs was made for the (then) Road Engineering and Environment Division (REED) in December The study measured traffic flows into and past four MSAs, referred to, in this report, as sites 94A, 94B, 94C and 94D. Additional information obtained on gross floor areas and type of facilities offered at these four MSAs is shown in Table 1. Table 1 Services offered at 1994 survey sites Floor area (square metres) 94A 94B 94C 94D North South East West East West Service offered -bound -bound -bound -bound -bound -bound Shop and shop store Toilets (incl. showers, baby change etc.) Restaurant (Back of House) Restaurant servery Restaurant seating t Amusements Circulation (Public) Circulation (Back of House) Staff areas t Plant/Boiler room Customer services/atm Roadside Diner (Back of House) Roadside Diner (Public areas) Fast Food (Back of House) Fast Food (Public areas) t Total

8 3 The 1997 study Following discussions with two of the operators of MSAs, it was decided that, the sites used by TRL in 1994 being rather old, any intended data collection exercise should include more modern sites which would have different operational characteristics. Two new sites, operated by two different companies, were selected and both questionnaire and video surveys were carried out at each site. The target for the questionnaire surveys was to obtain approximately 1000 interviews at each site, between the hours of 7am and 7pm. Respondents were sought in all parts of the MSA, car parks, coach and lorry parks, fuel station, outside the main concourse and indoors, to include a cross-section of drivers of both sexes, of any age, with and without passengers. The aim of the questionnaire was to determine why, and at what stage in a journey, a motorist decides to turn into an MSA. The video surveys sought to establish turn-in flows at each of the two MSAs as percentages of the upstream main line flows on a half-hourly basis throughout the day between 7am and 7pm. The method of data collection was SVHS video recording, analysed through the Camera and Computer Aided Traffic Sensor system (CCATS) video processor package. Details of gross floor areas and services offered at site 97A are shown in Table 2. The questionnaire survey at this site took place between Monday 14 July and Friday 25 July 1997 inclusive, but excluding Saturday and Sunday. Table 2 Services offered at site 97A Floor area (square metres) Service offered Eastbound Westbound Fast Food (Public and Back of House areas) Shop and shop store Toilets (including showers, baby change etc.) Restaurant (Back of House) Restaurant servery Restaurant seating Amusements ATM - 7 Circulation (Public) Circulation (Back of House) Staff areas Phones - 13 Roadside Diner (Back of House) Roadside Diner (Public areas) Total Table 3 shows gross floor areas and services offered at site 97B. The questionnaire survey at this site was conducted on weekdays between Monday 4 August and Friday 15 August inclusive. Table 3 Services offered at site 97B 4 The 1998 extension to the study One-day video surveys were carried out at a further six MSAs in January/February 1998, sites 98A to 98F. Survey dates were as follows: Site 98A - 27/1/98 (repeat* at one camera position on 17/2/98) Site 98B - 29/1/98 Site 98C - 3/2/98 (repeat* at one camera position on 17/2/98) Site 98D - 4/2/98 Site 98E - 10/2/98 Site 98F - 11/2/98 * Repeats were necessary because of camera failure. In addition, an inventory of parking provision at each of the twelve video survey sites studied between 1994 and 1998 was compiled in March/April 1998, because it was felt that there could be a relationship between turn-in flows and the amount of parking space available. Regular drivers along a motorway might avoid a particular MSA because they had found it difficult to find a parking space on a previous occasion. 5 Results Floor area (square metres) Services offered Northbound Southbound Shop and shop store Public toilets Restaurant (Back of House) Restaurant (Front of House) Amusements Customer services/atm 9 39 Public circulation General staff areas Roadside Diner (Back of House) 93 - Roadside Diner (Front of House) 92 - Total The video surveys As was the case for the 1994 data, the 1997 and 1998 main carriageway and turn-in flow data were analysed, wherever possible, through the Camera and Computer Aided Traffic Sensor system (CCATS) video processor package. In a few cases, because of camera angle or poor lighting conditions, it was necessary to analyse the data manually. Flow data obtained from all the video surveys were tabulated and analysed. Main flow was calculated as the sum of Turn-in flow and Straight-on flow, and percentage Turn-in as the calculated percentage of main flow traffic turning into a Motorway Service Area. 4

9 The average turn-in flows and main flows per half hour for all sites are shown in Table 4. Graphical comparisons of average main flow, average turn-in flow and average turn-in percentage for the grand total of twenty-four sites (twelve MSAs, two directions each), are given in Figures 1, 2 and 3, together with some basic statistics. Table 4 Average turn-in flow and main flow (vehicles per half hour) Main Number Turn-in flow Turn-in of cases 7:00am 7:00am Percen Exam Direction -7:00pm -7:00pm -tage 1 - ined 94A Southbound N=24 Northbound N=24 94B Eastbound N=24 Westbound N=24 94C Eastbound N=24 Westbound N=24 94D Eastbound N=24 Westbound N=20 97A Eastbound N=24 Westbound N=24 97B Southbound N=24 Northbound N=24 98A Eastbound N=24 Westbound N=24 98B Southbound N=24 Northbound N=24 98C Southbound N=24 Northbound N=24 98D Eastbound N=21 Westbound N=24 98E Southbound N=24 Northbound N=24 98F Southbound N=24 Northbound N=24 Averages N=569 Averages N=528 1 This is calculated by averaging the individual half hour percentages rather than by dividing the sum of turn-in flows by the sum of main flows. 2 Due to missing observations during the period 7:00am to 9:00am the averages for both half hour main flow and turn-in flow are calculated for the period 9:00am to 7:00pm. 3 Again, due to missing observations, the averages for both the main flow and turn-in flow are calculated for the period 7:00am to 7:00pm but excluding the periods 10:30am to 11:00am and 2:00pm to 3:00pm. 4 This excludes the two sites 94D Westbound and 98D Eastbound where data was incomplete as explained above. Simple regression analysis has been used to investigate the relationship between: l turn-in flow and main flow; l turn-in percentage and main flow; l turn-in flow and half hour time intervals. In each case there were 24 models, one for each direction at each MSA. These suggested a range of potential relationships and were often not statistically significant. No doubt this was largely a result of having, at most, only twenty-four observations per site subject to the vagaries that a single day may impose rain, bright sunshine, congestion following an accident, road works, fog, congestion due to a shock wave progressing down the motorway etc. Due to this inherent variability within sites the analysis focuses on linear models with data from all MSAs considered together, thus attempting to identify the general model. The findings were as follows: i Though some of the 24 site models suggested that the relationship between turn-in flow and main flow was negative and others that it was positive the overall model was statistically significant and indicated a positive relationship: Turn-in flow = x Main flow This model is visually depicted in Figure 4. The small squares correspond to known observations and the thick dark line reflects the expected mean value of turn-in flow (with regard to all sites surveyed) given a value for main flow (and assuming a linear relationship). The dashed lines above and below the regression line represent the region within which 95% of individual observations for turn-in flow will reside, i.e., given an observation for main flow (from the sites surveyed) we would expect on nineteen occasions out of twenty that the turn-in flow will reside within the two dashed lines. This linear model is not ideal, and though significant (p<0.01) only explains 16.4% of turn-in variability. A slight improvement can be obtained through application of a power model where: Turn-in flow = 2.841xMain Flow This explains 17.1% of variability. ii The relationship between turn-in percentage and main flow (Figure 5) was more consistent over the 24 individual site models, all except one showing a negative relationship, i.e., the larger the main flow, the smaller the proportion of vehicles turning in. The equation for the overall model for all sites combined, which was statistically significant (p<0.01), was: Turn-in percentage = x Main flow The performance of this linear model is substantially better than the previous relationship between turn-in flow and main flow with 19.6% of variability in turn-in percentage explained by this model. iii The regression analysis of turn-in flow with regard to half hour time interval proved inconclusive. This may be due to the variation in peak times for turn-in flows as shown in Table 5. 5

10 The Sites 94A Southbound 94A Northbound 94B Eastbound 94B Westbound 94C Eastbound 94C Westbound 94D Eastbound 94D Westbound 97A Eastbound 97A Westbound 97B Southbound 97B Northbound 98A Eastbound 98A Westbound 98B Southbound 98B Northbound 98C Southbound 98C Northbound 98D Eastbound 98D Westbound 98E Southbound 98E Northbound 98F Southbound 98F Northbound Average main flows (vehicles per half hour) Cases examined 2 : 569 half hour flows (528) Overall average half hour main flow: 1450 vehicles (1482) Overall standard deviation in half hour main flow: 526 vehicles (522) Standard error of average half hour main flow: 22 vehicles (23) Largest registered half hour main flow: 2968 vehicles, 98A Westbound Smallest registered half hour main flow: 116 vehicles, 98B Eastbound Greatest variation in half hour main flow: 98E Southbound Least variation in half hour main flow: 94D Westbound (94B Eastbound) 1 The two black bars correspond to sites where data was incomplete i.e. sites 94D Westbound and 98D Eastbound 2 Half hour periods for which both main flow and turn-in flow were recorded. Note: Figures in brackets are for the reduced data set where 94D Westbound and 98D Eastbound are excluded. Figure 1 Comparison of main flows 1 6

11 The Sites 94B Eastbound 94B Westbound 94A Southbound 94A Northbound 94C Eastbound 94C Westbound 94D Eastbound 94D Westbound 97A Eastbound 97A Westbound 97B Southbound 97B Northbound 98A Eastbound 98B Southbound 98B Northbound 98C Southbound 98C Northbound 98D Eastbound 98D Westbound 98E Northbound 98F Southbound 98F Northbound 98A Westbound 98E Southbound Average turn-in flows (vehicles per half hour) Cases examined 2 : 569 half hour turn-in flows (528) Overall average half hour turn-in flow: 169 vehicles (169) Overall standard deviation in half hour turn-in flow: 84 vehicles (87) Standard error of average half hour turn-in flow: 4 vehicles (4) Largest registered half hour turn-in flow: 935 vehicles, 98E Southbound Smallest registered half hour turn-in flow: 15 vehicles, 94B Eastbound Greatest variation in half hour turn-in flow: 98E Southbound Least variation in half hour turn-in flow: 94B Westbound 1 The two black bars correspond to sites where data was incomplete i.e. sites 94D Westbound and 98D Eastbound. 2 Half hour periods for which both main flow and turn-in flow were recorded. Note: Figures in brackets are for the reduced data set where 94D Westbound and 98D Eastbound are excluded. Figure 2 Comparison of turn-in flows 1 7

12 The Sites 94A Southbound 94A Northbound 94B Eastbound 94B Westbound 94C Eastbound 94C Westbound 94D Eastbound 97A Eastbound 97A Westbound 97B Southbound 97B Northbound 98A Eastbound 98A Westbound 98B Southbound 98B Northbound 98C Southbound 98C Northbound 98D Eastbound 98D Westbound 98E Northbound 98F Southbound 98F Northbound 98E Southbound 94D Westbound Average turn-in percentage Cases examined 2 : 569 half hour turn-in percentages (528) Overall average half hour turn-in percentage: 12.6 % (12.1%) Overall standard deviation in half hour turn-in percentage: 5.7 % (5.3%) Standard error of average half hour turn-in percentage: 0.2 % (0.2%) Largest registered half hour turn-in percentage: 38.0%, 94D Eastbound Smallest registered half hour turn-in percentage: 3.5%, 98F Northbound Greatest variation in half hour turn-in percentage: 98E Southbound Least variation in half hour turn-in percentage: 98B Northbound 1 The two black bars correspond to sites where data was incomplete i.e. sites 94D Westbound and 98D Eastbound. 2 Half hour periods for which both main flow and turn-in flow were recorded. Note: Figures in brackets are for the reduced data set where 94D Westbound and 98D Eastbound are excluded. Figure 3 Comparison of turn-in percentages 1 8

13 Turn-in (vehicles per half hour) Main flow (vehicles per half hour) Figure 4 Relationship between turn-in flow and main flow Turn-in percentage Main flow (vehicles per half hour) Figure 5 Relationship between turn-in percentage and main flow 9

14 Table 5 Turn-in flow peak times Site Direction Turn-in flow peak 94A Southbound 9:30am 10:00am Northbound 1:30pm 2:00pm 94B Eastbound 6:00pm 6:30pm Westbound 5:30pm 6:00pm 94C Eastbound 9:30am 10:00am Westbound 2:30pm 3:00pm 94D Eastbound 8:30am 9:00am Westbound 1 9:30am 10:00am 97A Eastbound 12noon 12:30pm Westbound 1:00pm 1:30pm 97B Southbound 1:00pm 1:30pm Northbound 8:00am 8:30am 98A Eastbound 9:30am 10:00am Westbound 2:00pm 2:30pm 98B Southbound 8:00am 8:30am Northbound 8:30am 9:00am 98C Southbound 8:30am 9:00am Northbound 9:30am 10:00am 98D Eastbound 2 3:30pm 4:00pm Westbound 8:00am 8:30am 98E Southbound 5:00pm 5:30pm Northbound 9:00am 9:30am 98F Southbound 12:30pm 1:00pm Northbound 12:30pm 1:00pm 1 Missing observations during the period 7:00am to 9:00am. 2 Missing observations for 10:30am to 11:00am and 2:00pm to 3:00pm. Table 6 shows mean percentage turn-in rates for the twelve MSAs surveyed, together with distances to next services and from previous services, distances from last junction and to next junction, and numbers of parking spaces, all factors which might have a bearing on turn-in flows. A stepwise regression technique applied to the total data set of all half hour flows covering all sites with: dependent variable: Turn-in percentage (the relationship between turn-in percentage and main flow was more robust than the relationship between turn-in flow and main flow). prospective independent variables * : Distance from previous junction Distance from previous MSA Distance to next junction Distance to next MSA Number of car parking spaces Number of other parking spaces Size of main flow Number of time interval in the day (7:00am to 7:30am = 1, 7:30am to 8:00am = 2, 6:30pm to 7:00pm = 24) produced the following model: Turn-in percentage = x Size of main flow x Distance from previous junction x Distance from previous MSA x Distance to next junction x Distance to next MSA x (Number of car parking spaces + Number of other parking spaces) Table 6 Comparison of results for all surveyed MSAs Average Distance Distance Coach turn-in from from Distance and percentage previous 1 previous to next Car lorry 7:00am MSA junction junction parking parking 7:00pm (miles) 2 (miles) (miles) spaces spaces 94A Sbnd Nbnd B Ebnd Wbnd C Ebnd Wbnd D Ebnd Wbnd A Ebnd Wbnd B Sbnd Nbnd A Ebnd Wbnd B Sbnd Nbnd C Sbnd Nbnd D Ebnd Wbnd E Sbnd Nbnd F Sbnd Nbnd Distance to next services in any direction is the same as distance from previous services in the opposite direction. 2 Where several previous or next service stations are present, depending on the motorway route taken, only the nearest of them is included. At the time of the survey, there was no continuous motorway route from site 94B to alternative MSAs. 3 Due to missing observations during the period 7:00am to 9:00am averages are calculated for the period 9:00am to 7:00pm. 4 Again, due to missing observations, averages are calculated for the period 7:00am to 7:00pm excluding the periods 10:30am to 11:00am and 2:00pm to 3:00pm. This model explains 35% of turn-in percentage variability and tends to indicate that: l the further alternative MSAs are from the site the higher the turn-in percentage; l the nearer junctions are to the site the higher the turn-in percentage; l the more parking spaces there are the higher the turn-in percentage; l the greater the main flow the lower the turn-in percentage. It is not immediately clear why the distance to/from the next/previous junction should matter and, in fact, alternative models suggest that the relationship between turn-in percentage and distance to junctions is not robust. Due to there being no continuous motorway route from site 94B to alternative MSAs, this site was excluded from this analysis. * Gross floor area was available for only half the MSAs and so was excluded from the regression analysis. 10

15 5.2 The questionnaire survey A total of 2082 completed questionnaires were obtained, 1064 at site 97A and 1018 at site 97B. Respondents were predominantly male at both locations (overall 83%) but those interviewed at site A were generally younger. Proportions in the various age bands are shown in Table 7. Table 7 Age distribution Up to Table 9 Place of interview Place of interview Percentage In concourse 2% Outside concourse 15% At fuel station 18% In car park 48% In lorry park 12% In coach park 4% Other e.g. rest area, caravan park 1% 1997 Site A 14% 41% 38% 7% 1997 Site B 10% 38% 46% 6% A detailed comparison of the separate results for sites A and B was undertaken to determine whether there were any significant differences between the two. The conclusion drawn was that the two samples were not dissimilar, and that results would therefore be reported for the two locations combined. Where significant but non-sensitive differences were detected between the two sites, mention will be made in the relevant section below. Approximately equal numbers of interviews were obtained on each side of the motorway, eastbound and westbound at site A, and northbound and southbound at site B. Interviews were to be spread throughout the day, but in the event, more were obtained in the morning between 7am and 12 noon (63%) than in the afternoon up to 7pm. A similar number of interviews were completed on each of the five days of the week, Monday to Friday. Only the driver of the vehicle arriving at the MSA was interviewed; the vehicles driven by respondents are shown in Table 8. Table 8 Respondents vehicles Vehicle driven Car LGV HGV Motorcycle Bus/coach 70% 9% 14% 3% 4% The place of interview is shown in Table 9. (N.B. Interviewers were instructed to interview inside the concourse only if weather conditions dictated - the weather was recorded as dry for 94% of interviews.) At the time of interview, 64% of respondents were on a business trip, 33% were travelling for pleasure, and the remaining 3% were combining the two. The number of passengers carried are shown in Table 10. The maximum number of adult passengers recorded was 51; the maximum number of children carried was 53. Sixty-three per cent of respondents were travelling completely alone, with neither adult nor child passengers. Table 10 Number of passengers Number of passengers Age of passengers or more Adults 66% 24% 5% 2% 1% 3% Children 88% 5% 4% 1% <1% 1% Journey details Respondents were asked a number of questions about the journey they were engaged in, to determine their origin and destination and hence the total length of the journey, the time the journey would take in total, and the time the respondent had been travelling since the last stop, or since setting out on the journey if there had been no previous stop. The distribution of journey length is shown in Table 11. Journey length ranged from as little as four miles to a staggering 860 miles, with a median length of 165 miles (mean 180 miles). These were claimed to be one-way journeys and with a total estimated journey time of up to 20 hours, it is likely that in some cases the journey would be spread over 2 days (eg long-distance lorry drivers and delivery drivers/sales representatives on round trips). Median total journey time was 3 hours 15 minutes. (The median is a measure of central tendency. It is the middle value of a variable when the values are placed in order from the smallest to the largest. The median is less sensitive to extreme values than the more commonly quoted mean.) The median journey length has also been calculated separately for the two sites, in each direction and for each of the two categories of journey purpose, business and pleasure. Results are in Table 12. Median journey lengths at site B are higher than those at site A. This can be explained by their relative positions in the motorway network. In each case, the median journey length for pleasure trips is greater than that for business trips. The length of time travelling since the last stop varied, according to the questionnaires, from a few minutes to 16 hours 30 minutes. Anything over five hours was considered unlikely and therefore excluded from the calculation of mean time travelling since last stop, which was 1 hour 34 minutes. Table 11 Length of journey Length of journey (miles) > 500 7% 15% 23% 22% 15% 9% 4% 2% 1% 1% 1% 11

16 Table 12 Median journey length for respondents at MSAs by journey purpose Median journey length (miles) Direction Business Pleasure 1997 Site A Eastbound Westbound Site B Southbound Northbound Reasons for stopping at MSA Eighty-five per cent of respondents had planned, before setting out, to stop somewhere on their journey; slightly less than half had planned to stop at the particular service station at which they were interviewed. For 24% of interviewees, this was the first time they had stopped at this service area. Respondents were asked for their main reasons for deciding to stop on this occasion. The results are shown in Table 13, separately for the two sites as there were some differences between them. (N.B. Some respondents gave more than one reason; reasons given by less than 5% of respondents at both sites are not included). Table 13 Main reasons for stopping at MSA Table 14 Miles travelled in UK Miles travelled All roads Motorways up to 10,000 21% 42% 10,000 to 20,000 26% 20% 20,000 to 30,000 13% 14% 30,000 to 40,000 11% 10% 40,000 to 50,000 10% 5% 50,000 to 60,000 5% 3% 60,000 to 70,000 3% 2% 70,000 to 80,000 3% 2% 80,000 to 90,000 1% 1% 90,000 to 100,000 4% 1% 100,000 to 150,000 3% 1% more than 150,000 1% <1% Attitudes and comments Respondents were asked which of four possible responses best described their attitude towards using Motorway Service Areas. Results are shown in Figure 6. The other category in the figure includes those respondents who said only that they considered MSAs convenient or handy. Only use when absolutely necessary 43% Like MSAs 33% Site A Site B To have break/stretch legs 35% 47% Use toilets 31% 46% Coffee/drink 23% 27% Buy petrol/diesel 21% 18% Buy book/newspaper/cassette/cd/ sweets/cigarettes 11% 16% Have something to eat inside MSA 9% 10% Get takeaway 7% 11% Consume food/drink from home 2% 6% More interviews at site A than at site B were conducted at the fuel station. More respondents at site B were travelling for pleasure, and those stopping at site B had longer total journeys, which may explain the higher percentages quoting rest and refreshment as reasons for stopping on this occasion. Reasons given for stopping at this particular service station were similar for sites A and B, the most often quoted reason being convenience of time/distance (55% of respondents) Driving experience Of the 1997 TRL survey population, 40% said they travelled on motorways daily, either as a passenger or driver, a further 33% at least once a week, and another 14% at least once a month. The distribution of miles travelled in the United Kingdom each year is shown in Table 14. Seventeen per cent of respondents said they stopped at a Motorway Service Area daily, a further 36% at least once a week, and another 21% at least once a month. Only use when on business 15% Figure 6 Attitude to MSAs Other 4% Prefer to leave motorway & find facilities elsewhere 5% A specific question about the adequacy of signs on the motorways for the service areas revealed that 96% of respondents found the signs adequate. Observations made by those who thought the signs inadequate were that they were too late, they were often obscured by large vehicles travelling in the nearside lane, there were not enough of them, and they should specify particularly the type and price of fuel available. The principal responses to a final question asking for general comments about Motorway Service Areas (i.e. volunteered opinions) are shown in Table Discussion and conclusions A motorist s decision to turn into a Motorway Service Area depends on a number of factors. The questionnaire survey carried out as part of this study has shown that the main reasons for stopping at an MSA are for rest, 12

17 Table 15 General comments Comment % of respondents Expensive/overpriced 45% Convenient/handy/useful 16% They have improved/are improving 8% Comments on food quality/variety 8% Tidy/clean 8% OK/fine/adequate/good 7% Good facilities 7% Poor facilities/lacking something 7% Not enough MSAs/poor distribution 6% Comments on parking areas/layout etc. 6% refreshment, toilets, and fuel. However, with 45% of respondents volunteering the opinion that, in general, MSAs are expensive, and 43% claiming that they only use them when absolutely necessary, there must be other factors which determine the choice of one MSA in preference to another, and the use of an MSA in preference to leaving the motorway and finding facilities elsewhere. All of the MSAs in the sample surveyed offered similar facilities; gross floor areas were only available for six of the MSAs studied and this variable was therefore excluded from a regression analysis. Total journey length will certainly have an effect on a motorist s decision to turn into a Motorway Service Area. Long trips may necessitate two or more stops. The timing of a stop may be influenced by journey purpose, e.g., a businessman may find it convenient to stop for light refreshment half an hour prior to a meeting; a family on a long journey may stop at lunchtime or for an evening meal. The model developed in this study relating turn-in percentage to main carriageway flow past an MSA could be refined if the characteristics of the main flow, at a particular time of day, in terms of journey purpose, origin and destination (and hence journey length and location of the MSA relative to origin/destination), were known. Such information about main carriageway flow could be derived from traffic models such as the Motorway Users Model (for SE England), developed for the Department of Transport in The main findings of TRL s studies (1994, 1997 and 1998) of Motorway Service Areas are as follows: Video surveys l Analysis of half-hourly turn-in flow and main flow recorded over a 12 hour period in both directions at each of the twelve MSAs shows a variable relationship between the two. Usually the relationship is positive but in a number of cases it was negative. l The relationship between turn-in percentage and main flow was less variable and almost always negative, i.e., the larger the main flow, the smaller the proportion of vehicles turning in to an MSA. A base level of 19.5% is reduced by 0.5% per 100 vehicles in the half-hourly main flow. Thus for an average half hour main flow of 2000 vehicles, the percentage of vehicles turning in to an MSA would be approximately 9.5%. l A stepwise regression analysis has indicated a diminution of turn-in percentage as main flow expands, an increase in turn-in percentage as distance to/from other MSAs increases, and a decrease in turn-in percentage as distance to/from next junction increases. Also, the greater the total number of parking spaces, the higher the turn-in percentage is likely to be. Questionnaire survey 1 A total of 2082 completed questionnaires were obtained at two Service Area sites. Of the sample population, 64% were on a business trip at the time of interview and 33% were travelling for pleasure, while the remaining 3% were combining the two. Overall median journey length was 165 miles and median journey time was 3 hours 15 minutes. The mean time travelling since the last stop was 1 hour 34 minutes. 2 Main reasons given for deciding to stop at the MSA were for rest, refreshment, toilets, and fuel. At both sites, the main reason for stopping at the particular MSA at which interviewing took place was convenience of time/distance. 3 Overall, 43% of respondents said they only use MSAs when absolutely necessary and a further 15% said they only use them when on business. General (volunteered) comments revealed that at least 45% of respondents considered MSAs to be expensive/overpriced. This study has confirmed the view that a motorist s decision to turn into a Motorway Service Area is influenced by his/her journey length, journey purpose, location of an MSA relative to origin and destination, the availability and proximity of other MSAs, and the amount of parking space available at an MSA, based on previous experience. 7 Acknowledgements The 1997 study described in this report, and the 1998 extension, form part of the research programme undertaken as part of a Highways Agency Research Project, Motorway junction layout to increase capacity and safety at low cost, Project Officer Mr John Smart, Traffic, TSE Division. The interview survey was carried out by Mrs Joan Franklin, TRL, and her team of interviewers. The video surveys were carried out by TRL s Visual Media Centre. The authors gratefully acknowledge the assistance given by Directors of the Operating Companies involved and staff at individual MSAs. 8 References Department of Transport/Welsh Office (1988). The control of development on trunk roads. Circular 4/88 (Department of Transport), Circular 42/88 (Welsh Office). Department of Transport/Welsh Office (1994). Motorway Service Areas. Roads Circular 1/94 (Department of Transport), Circular 34/94 (Welsh Office). 13

18 Department of the Environment/Department of Transport (1994). Planning Policy Guidance: Transport (PPG13). Stationery Office, London. Department of Transport (1994). Motorway Users Model. WS Atkins Planning Consultants and Colquhoun Transportation Planning. 14

19 Abstract The Transport Research Laboratory was requested to look at the factors affecting the turning flows at Motorway Service Areas, as part of a Highways Agency research project entitled Motorway junction layout to increase capacity and safety at low cost. To this end, TRL carried out a questionnaire survey of motorists at two service stations, and also one-day video surveys at these two MSAs and six others, to supplement and compare with data collected by means of video surveys at four MSAs in December This report presents the results of this study. Related publications TRL168 Falling asleep at the wheel by J A Horne and L A Reyner (price 35, code J) CT79.1 Driver fatigue update ( ). Current Topics in Transport: selected abstracts from TRL Library's database (price 20) RR61 The transport implications of company financed motoring by J M Hopkin (price 25, code G) SR83 Layout of lorry parks: dimensions of stalls and aisles by M Brannam and J D Longmore (price 20) Prices current at April 2000 For further details of these and all other TRL publications, telephone Publication Sales on or , or visit TRL on the Internet at 15

20 16