This report summarises the results from the fourth sub-trial of a larger track trial investigating the reactions of road users to Low Level Cycle Signals (LLCS) under different junction configurations. The trials were conducted at a specially constructed typical “urban” four-arm junction built at TRL’s test track. In this trial the LLCS were accompanied by three different cycle reservoir depths of 5m, 7.5m and 10m. The LLCS were positioned on a separate pole to the standard traffic signals, with the LLCS being at the second stop line and the main signals being at the first stop line. This junction layout was trialled both with and without an ‘early release’ for cyclists ahead of the vehicle traffic. Trials were conducted for two different road user groups over fifteen days, with a total of 1,290 participants: 1,117 cyclists (9 days) and 173 car drivers (6 days). In the cycle trial two group sizes were tested: a ‘small group’ of eight cyclists and a ‘large group’ of 16 cyclists. In the car trial, the data from a previous study was re-used as the baseline scenario of a 5m cycle reservoir, which involved 88 participant car drivers (3 days). The main study objective was to gather evaluation evidence on different sizes of cycle reservoir for groups of cyclists and individual car drivers, specifically when combined with LLCS being mounted on separate poles to the main signals. The standard depth of ASLs is currently 4 to 5 metres, although orders have been granted for a small number of sites to have ASLs that are 7.5m deep and DfT’s consultation update to the TSRGD (May 2014) includes 7.5m ASLs. Key findings are listed at the end of each sub-section and are referenced here in square brackets. The average occupancy of the different reservoirs was found to be as follows: - The average occupancy of the 5m reservoir was 8.0 cyclists when trialled with large groups (16 cyclists) and 6.5 cyclists when trialled with small groups (eight cyclists) [F6.a, F6.b]. - The average occupancy of the 7.5m reservoir was 13.0 cyclists when trialled with large groups (16 cyclists) [F6.b]. - The 10m reservoir was sufficiently large enough to hold at least 16 cyclists in almost all instances when trialled with large groups (16 cyclists) [F6.b]. These findings suggest that the following reservoir depths may be considered for a junction with a one-lane approach: - A cycle reservoir between 5m and 7.5m deep when the required storage space is 8 to 13 cyclists. - A cycle reservoir greater than 7.5m deep when the required storage space is 13 or more cyclists. - A rule of thumb seems to be 1.7 cyclists per metre of reservoir depth for a one-lane approach. Around half of the cyclists said that there were times when it was difficult or impossible to see the LLCS when waiting at the junction [F4.b]. This was greater with the larger groups (40%). The most common reason being that the LLCS was obscured by other cyclists [F4.c]. Of the cyclists who said it was difficult to see the LLCS and it affected how they went through the junction, over 40% said that they followed the cyclists in front, whereas a quarter said that they tried to reposition themselves so that they could see the LLCS [F4.d]. When asked specifically about the height of the LLCS, the most common response was “about right” (58%). Again the larger the group the more cyclists thought that the LLCS should be mounted higher (43-49% in the largest group) [F3.b]. These findings suggest that the height of the LLCS is about right, although could be mounted higher where large groups of cyclists are likely to be present. Similar to the previous trials, the longer early releases encouraged a higher proportion of cyclists to turn right ahead of the oncoming car. For the small group of eight cyclists, the average number of cyclists who turned right ahead of the oncoming car ranged from 1.3 with the 2-second early release up to 5.6 with the 5-second early release. For the large group of 16 cyclists, this ranged from 1.6 with the 2-second early release up to 8.6 with the 5-second early release [F8.b]. It was also found that in the scenarios with the deeper cycle reservoirs there was a higher average number of cyclists who turned right ahead of the oncoming car [F8.c]. This might be explained by the car on the opposing approach being set back further from the junction, resulting in a larger gap in which more cyclists could turn. The findings indicate that most of the cyclists who undertook the right turn movement in front of the oncoming car made a judgement to undertake this movement safely based on the junction layout, amount of early release and time required to clear the junction. However, there is evidence to suggest that some cyclists made this judgement based on the behaviour of the other cyclists in front, whereas a minority expected the car to wait for them [F8.e, F8.f]. Car drivers understood the different sized reservoirs equally well [F10.a] and the majority of car drivers thought that the size of the cycle reservoir they experienced was ‘about right’ [F11.a]. Common comments were that the size of the reservoir should be based on the location and volume of cyclists using the junction and that there is a need to strike a balance between space for cyclists and motor vehicles [F11.b]. The trials with the deeper cycle reservoirs were associated with a small decrease in compliance [F14.a], although the majority of encroachment was only up to 1.25m past the first stop line [F14.b]. Over half of the car drivers stopped more than 2.5m before the first stop line for the 7.5m and 10m reservoir depths, suggesting that a substantial proportion of car drivers stopped quite far back from the stop line, possibly in order to see the main signals that were located on the separate poles at the first stop line [F14.c]. When the reservoirs were 7.5m or 10m, the car drivers were more likely to start moving on the LLCS early release, compared to the 5m reservoir [F15.a]. Thus, the average Reaction Times for the car drivers were typically around half a second faster in the trials with the deeper reservoirs, compared to the trial with the 5m reservoir [F15.b]. The average Entry Time was highest in the scenario with no early release and a cyclist in front, suggesting that the car driver often had to wait for the cyclist before entering the junction. For the scenarios with no early release and no cyclist present, the average Entry Time increased by about 0.2 seconds for each additional 2.5m of reservoir [F15.d]. For the trials with an early release, the 7.5m reservoir resulted in a small increase of about 0.1 seconds to the average Entry Time, whereas the 10m reservoir resulted in an increase of about 0.6 seconds, compared against the 5m reservoir [F15.e]. (Author/publisher)
Samenvatting