Safety Evaluation of Continuous Green T Intersections

CHAPTER 5. DATA COLLECTION

Florida and South Carolina are two States with multiple CGT intersections that have existed for several years. Data from both of these States were used in the present study. The data collection procedure and summary of the data are provided in this chapter of the report.

FLORIDA

Florida began installing CGT intersections as early as 1972. CGT intersections considered in this study were constructed between 1972 and 2004, and the geometry remained unchanged during the safety evaluation period. The Florida Department of Transportation (FDOT) provided locations of CGT and comparable conventional signalized T intersections for analysis.

FDOT provided traffic volume data for the major and minor roads for each intersection where possible. When FDOT did not have traffic volume data, local jurisdictions were contacted to obtain traffic volume data. Traffic volumes for 2013 were available for all of the major roads, which included the high-speed, continuous through movement at the CGT intersections. When FDOT and the local jurisdictions did not have traffic volume data for the intersecting roadway, the Trip Generation Manual was used to predict the traffic volume based on the land uses along the properties adjacent to the minor intersecting roadway.⁽³⁴⁾ After these steps were taken, there were 4 CGT and 11 comparison intersections from Florida with missing traffic volumes on the minor street approach. Because most of the minor street approaches with missing traffic volume data were in residential areas, it was assumed that the volumes on these approach roadways would be approximately equal to 500 vehicles per day. One local jurisdiction (Melbourne, FL) performed a multiday traffic count for one of the missing minor street approaches and confirmed that a 500 vehicle per day volume was accurate.

FDOT also provided crash data in geographic information system files. These files were used to identify all crashes (total crashes), fatal and injury crashes, rear-end crashes, sideswipe crashes, and angle crashes within 250 ft of the intersections for 2008–2012 (inclusive).

Google Earth™ was used to collect other key variables, including variables related to the intersection design and traffic control features. These variables included the following:

• Posted speed limits on the major and minor street approaches.
  
  
• Lane and shoulder widths on all approaches.
  
  
• The number of through lanes on the major and minor approaches.
  
  
• The presence of right- and left-turn lanes on the through and intersecting roadways.
  
  
• The presence of a channelized right-turn lane from the major approach to the minor street.
  
  
• The presence of a channelized right-turn lane from the minor approach to the major street.
  
  
• Whether right-turn-on-red movements were permitted on either the major or minor street approaches.
  
  
• Whether the intersection is located on a horizontal curve.
  
  
• Whether a rail line crosses the intersecting roadway near the intersection.
  
  
• Whether there was a driveway (right-in, right-out only) where a fourth leg of the intersection would be.
  
  
• The skew angle for the intersection.

In total, there were 30 CGT intersections and 38 comparison intersections from Florida included in the analysis database. Variable names and the associated definitions are provided in table 1 for the Florida intersections. The variables include traffic volumes and posted speed limits on the through and intersecting roadways, lane-use controls, cross-section dimensions, geometric characteristics, and crash-related information.

Descriptive statistics for CGT intersection variables included in the Florida analysis data files are provided in table 2 and table 3.

Descriptive statistics for the comparison intersection variables included in the Florida analysis data files are provided in table 4 and table 5.

SOUTH CAROLINA

CGT intersections in South Carolina included in the database for this evaluation were installed between the period prior to 1990 through 2010. The South Carolina Department of Transportation (SCDOT) worked with the research team to identify the CGT and comparison intersections for use in the analysis.

Traffic volume data for the major and minor approach roads for each intersection were obtained from SCDOT where possible. Traffic volumes for 2013 were available for all of the major roads. When SCDOT did not have traffic volume data for the intersecting roadway, the Trip Generation Manual was used to predict the traffic volumes based on the land use of land adjacent to the minor street approach.⁽³⁴⁾ After these steps were taken, there were no intersections from South Carolina with missing traffic volumes.

SCDOT also provided crash data for 2009 through 2013. These files were used to identify all crashes (total crashes), fatal and injury crashes, rear-end crashes, sideswipe crashes, and angle crashes within 260 ft of the intersections for each of the years. The distance of 260 ft was selected due to the measurement values associated with the crash data. (It was not possible to use 250 ft.)

Google Earth™ was used to collect other geometric and traffic control data. The variables were the same as those collected for the Florida data.

In total, there were 16 CGT intersections and 21 comparison intersections from South Carolina included in the analysis database. One CGT intersection was constructed in 2009, and one was constructed in 2010. Thus, data for the years of construction and the year before (in the case of the CGT intersection constructed in 2010) were excluded from the analysis period. The descriptive statistics for CGT intersections from South Carolina are provided in table 6 and table 7.

The descriptive statistics for comparison intersections from South Carolina are provided in table 8 and table 9.