Synthesis of Safety for Traffic Operations - Transports Canada
Synthesis of Safety for Traffic Operations - Transports Canada
Synthesis of Safety for Traffic Operations - Transports Canada
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
<strong>Synthesis</strong> <strong>of</strong> <strong>Safety</strong> <strong>for</strong> <strong>Traffic</strong> <strong>Operations</strong> March 2003<br />
models predict the annual number <strong>of</strong> crashes on an arterial approach (both directions<br />
included). Of the many variables that were found to influence crash occurrence, the type<br />
<strong>of</strong> signal control was included. Adaptive signal control was found to be safer than<br />
pretimed signal control, reducing crashes by 13% (CMF <strong>of</strong> 0.87).<br />
Night-time Flash<br />
Polanis (2002)<br />
Polanis (2002) reported on the removal <strong>of</strong> red/amber night-time flashing operation from<br />
19 intersections in Winston-Salem, North Carolina. Site selection in<strong>for</strong>mation is sketchy;<br />
the intersections are “not necessarily… high-crash locations”, but are described as<br />
locations where the crash pattern indicated a safety benefit from night-time flash<br />
removal. Target crashes are right-angle crashes that occur during the night-time flashing<br />
operation. The results <strong>of</strong> the study are shown in Table 3.56.<br />
Sixteen <strong>of</strong> the 19 intersections exhibited a statistically significant reduction in target<br />
crashes at the 95% level <strong>of</strong> confidence. Polanis aggregates the results <strong>for</strong> all intersections,<br />
which yields CMFs <strong>of</strong> 0.22 <strong>for</strong> target crashes, and 0.67 <strong>for</strong> all right-angle crashes.<br />
There is no reason to believe that removal <strong>of</strong> night-time flashing operation would have<br />
any measurable effect on right-angle crashes at other times <strong>of</strong> the day. Hence, rather than<br />
measuring the safety impacts on total right-angle crashes, it is more in<strong>for</strong>mative to use<br />
the non-target crashes (total right-angle crashes minus target crashes) as a control group.<br />
If this is the case the aggregated data produces the statistics shown in Table 3.57.<br />
These data indicate that the target crashes were reduced by 78%. However, the nontarget<br />
crashes were also reduced, by 19%. Since the night-time flashing operation should<br />
have had no effect on the non-target crashes, a 19% reduction in target crashes could<br />
reasonably be expected without removal <strong>of</strong> night-time flash. Taking into account this<br />
background crash reduction leads to a CMF <strong>for</strong> removal <strong>of</strong> night-time flashing operation<br />
<strong>of</strong> 0.27, or a 73% reduction in right-angle crashes during the flashing operation.<br />
It bears mentioning that the above study did not account <strong>for</strong> exposure. This is an<br />
important consideration because, although traffic volumes are likely increasing and<br />
omission from the analysis leads to under estimation <strong>of</strong> the safety benefits, the growth <strong>of</strong><br />
traffic during night-time flashing operation is likely significantly less than during the<br />
remainder <strong>of</strong> the day. If this is indeed the case, then not accounting <strong>for</strong> exposure leads to<br />
an over estimation <strong>of</strong> the safety benefits <strong>of</strong> night-time flash removal.<br />
Page 59