Safety Evaluation of Corner Clearance at Signalized Intersections

CHAPTER 7. CROSS-SECTIONAL EVALUATION RESULTS

AGGREGATE ANALYSIS

Table 11 through table 16 present the estimated CMFs and related SE for each of the following target crash types.

• Total—all crashes within 250 ft of intersection (all types and severity levels combined).
  
  
• Fatal and injury—all injury crashes within 250 ft of intersection.
  
  
• Rear-end—all crashes coded as “rear-end” within 250 ft of intersection.
  
  
• Sideswipe—all crashes coded as “sideswipe” within 250 ft of intersection.
  
  
• Right-angle—all crashes coded as “right-angle” within 250 ft of intersection.
  
  
• Turning—all crashes coded as “right-turn” or “left-turn” within 250 ft of intersection.
  
  
• Nighttime—all crashes with lighting condition coded as “dark,” “dawn,” or “dusk” within 250 ft of intersection.

This study presents aggregate results by number of approach and receiving corners with driveways within 50 ft of the intersection. The study presents results separately for the number of approach corners (i.e., one or two) and number of receiving corners (i.e., one or two) compared to no driveways within 50 ft of the intersection on the approach or receiving corners, respectively.

For total crashes, the CMFs were 0.82 and 0.67 for corner clearance of 50 ft or less on one and two approach corners, respectively, compared to no driveways within 50 ft of both approach corners. The CMFs were 1.33 and 1.76 for corner clearance of 50 ft or less on one and two receiving corners, respectively, compared to no driveways within 50 ft of both receiving corners. All CMF estimates were statistically significant at the 95-percent confidence level.

Table 11. Results for total crashes.

For fatal and injury crashes, the CMFs were 0.79 and 0.62 for corner clearance of 50 ft or less on one and two approach corners, respectively, compared to no driveways within 50 ft of both approach corners. The CMFs were 1.29 and 1.68 for corner clearance of 50 ft or less on one and two receiving corners, respectively, compared to no driveways within 50 ft of both receiving corners. All CMF estimates were statistically significant at the 95-percent confidence level.

Table 12. Results for fatal and injury crashes.

For rear-end crashes, the CMFs were 0.79 and 0.63 for corner clearance of 50 ft or less on one and two approach corners, respectively, compared to no driveways within 50 ft of both approach corners. The CMFs were 1.36 and 1.86 for corner clearance of 50 ft or less on one and two receiving corners, respectively, compared to no driveways within 50 ft of both receiving corners. The CMF estimates were statistically significant at the 95-percent confidence level.

Table 13. Results for rear-end crashes.

For sideswipe crashes, the CMFs were 0.83 and 0.69 for corner clearance of 50 ft or less on one and two approach corners, respectively, compared to no driveways within 50 ft of both approach corners. These two CMF estimates were not statistically significant at the 90-percent confidence level. The CMFs were 1.31 and 1.71 for corner clearance of 50 ft or less on one and two receiving corners, respectively, compared to no driveways within 50 ft of both receiving corners. The CMF for one corner was statistically significant at the 95-percent confidence level, and the CMF for two corners was statistically significant at the 90-percent confidence level.

Table 14. Results for sideswipe crashes.

For right-angle crashes, the CMFs were 1.03 and 1.06 for corner clearance of 50 ft or less on one and two approach corners, respectively, compared to no driveways within 50 ft of both approach corners. Neither CMF estimates were statistically significant at the 90-percent confidence level. The CMFs were 1.42 and 2.02 for corner clearance of 50 ft or less on one and two receiving corners, respectively, compared to no driveways within 50 ft of both receiving corners. The CMF estimate for one corner was statistically significant at the 95-percent confidence level, and the CMF for two corners was statistically significant at the 90-percent confidence level.

Table 15. Results for right-angle crashes.

For turning (right- or left-turn) crashes, the CMFs were 1.00 and 1.01 for corner clearance of 50 ft or less on one and two approach corners, respectively, compared to no driveways within 50 ft of both approach corners. The CMFs were 1.22 and 1.49 for corner clearance of 50 ft or less on one and two receiving corners, respectively, compared to no driveways within 50 ft of both receiving corners. None of these CMF estimates were statistically significant at the 90-percent confidence level.

Table 16. Results for turning crashes.

For nighttime crashes, the CMFs were 0.94 and 0.87 for corner clearance of 50 ft or less on one and two approach corners, respectively, compared to no driveways within 50 ft of both approach corners. These two CMF estimates were not statistically significant at the 90-percent confidence level. The CMFs were 1.29 and 1.67 for corner clearance of 50 ft or less on one and two receiving corners, respectively, compared to no driveways within 50 ft of both receiving corners. The CMF estimate for one receiving corner was statistically significant at the 95-percent confidence level, and the CMF for two corners was statistically significant at the 90-percent confidence level.

Table 17. Results for nighttime crashes.

DISAGGREGATE ANALYSIS

The objective of the disaggregate analysis was to identify specific CMFs by crash type and different conditions. The analysis could also reveal those conditions under which the strategy was more effective. The research team considered several variables in the disaggregate analysis, including major and minor road traffic volume, number of lanes on the major and minor road, posted speed limit on the mainline, driveway density on the mainline, and presence of left- and right-turn lanes on the mainline. The multivariable regression models included interaction terms to investigate the potential differential effects of corner clearance with respect to the interacted variable. For example, the interaction term for major road traffic volume and number of major road approaches with driveways within 50 ft is the product of the two variables. A statistically significant interaction term would indicate an apparent differential effect of corner clearance across different traffic volumes or the other variables of interest.

The analysis results indicated that none of the interaction terms were statistically significant at even an 80-percent confidence level. While these results indicated no differential effect of corner clearance, the sample size may have been too small to detect differential effects at the desired level of confidence.