TxDOT X-Ramps vs Diamonds: Safety and Operational Issues for Conversion Presented by: Karen Dixon, Ph.D., P.E. October 10, 2017
Outline Objective of this Safety Investigation Phase I Update Site Selection Data Collection & Analysis Findings to Date Current Phase II Activities 2
Objective of Safety Investigation Evaluate the safety implications of ramp configurations and adjacent driveway placement for frontage roads in Texas. Ramp configurations include traditional as well as X-configuration ramps. Pay particular attention to the position/influence of driveways. Note: Phase I focus is on frontage road safety performance. 3
Traditional (Diamond) Ramps Interchange #1 Interchange #2 Highway / Freeway Weave / Merge on Freeway 4
X-Ramp Configurations Interchange #1 Interchange #2 Highway / Freeway Weave / Merge on Frontage Road 5
Site Selection / Identification 6
Site Selection -- Developed GIS Intersection Points for Freeways and Arterials 7
Site Selection (continued) Refined Interchange Identification using GIS Tools a) Aerial View b) GIS Street Map View c) Applying 60 m Buffer d) Consolidated Intersection Point 8
Site Selection (continued) -- Developed a Stratified Random Sample 9
Selection of Study Sites Acquired full detailed data for 93 randomly selected locations (62 sites with X ramps and 31 sites with traditional) Acquired additional general data for approximately 30 traditional diamond sites (to be included in Phase II) Identified urban frontage road driveway and ramp configurations upstream and downstream of each randomly selected interchange 10
Data Collection and Documentation Confirmed Study Site Matches Study Objectives Identified Site Boundaries Acquired Data from RHiNo and GoogleEarth/Streetview Assembled Companion Crash Data for Each Study Location 11
Additional Site Data Aerial Photographs / Street View 1. Used GoogleEarth to tag key locations of interest 2. Developed Excel macro to measure distances between tag locations 3. Determined driveway density, proximity, and opportunity for influencing weaving at each study location 12
Example Site Boundaries 13
Identify Key Driveways and Gore Locations 14
Identify Locations CL Reference Point 2B (out of view to left) 0 ft DW B4 - Driveway #4 (Center) 874.9 ft ON B1 Painted Gore for On-Ramp B1 1160.9 ft NG B1 Physical Nose of Gore B1 1257.9 ft BG B1 Back of Gore for B1 1299.2 ft DW B5 Driveway #5 (beyond potential weave section) 1556.2 ft 15
Driveways that Influence Weave Traditional Ramps Interchange #1 Interchange #2 Highway / Freeway Driveways in Weaving Area Region Not Directly Affected Driveways in Weaving Area 16
Driveways that Influence Weave X-Ramps Interchange #1 Interchange #2 Highway / Freeway Region Not Directly Affected Driveways in Weaving Area Driveways not Directly in Weaving Area but may still have residual impacts 17
Summary of Data Elements Ramp Type Segment Length Acquired Distance between Gore Noses Gore Ratio (i.e. gore distance / segment length) Total segment driveways (and locations) Gore driveways (and locations) 18
Summary of Data Elements Acquired (continued) Access density along road (points per mile) Access density within gore regions (points per mile) Distance from gore nose to first upstream and downstream driveways for G1 and G2 Number of lanes ADT 19
Example Crash Data (Site 158) CRIS maps the crash to freeway centerline Direction of travel and crash type used for additional frontage road assignments (current activity) 20
Conduct a Safety Assessment Assemble Merged Database Suitable for Descriptive Statistics for 3 years, 5 years, and 7 years (last year is 2015) Use the Merged Database to Conduct Statistical Safety Analyses Assessed dates associated with each location Document and contrast X-ramps and traditional ramps at urban diamond interchange locations Determine overall safety impacts for total crashes as well as F+I crashes 21
Review of Findings: X-Ramp F+I Crashes (3 yr) Critical variables include: Gore Ratio (i.e. length of gore divided by length of segment) Number of Driveways in Gore Region (within 400 or gore physical nose) Upstream distance from Nose of G1 to closest driveway Prevailing Number of lanes between G1 and G2 22
Review of Findings: X-Ramp Total Crashes (3 yr) Critical variables include: Gore Ratio Number of Driveways in Gore Region (within 400 or gore physical nose) Upstream distance from Nose of G1 to closest driveway Downstream distance from nose of G2 to closes downstream driveway Prevailing Number of lanes between G1 and G2 23
F+I Equation for X-Ramps Parameter Estimates Parameter Estimate Standard Error t Value Approx Pr > t Intercept 0.046317 0.358278 0.13 0.8971 GoreRatio 2.089092 0.464500 4.50 <.0001 GoreDWs 0.039265 0.011883 3.30 0.0010 DupG1-0.000323 0.000153-2.11 0.0352 LnsG1toG2 0.343152 0.086948 3.95 <.0001 24
Total Crash Equation for X-Ramps Parameter Estimates Parameter Estimate Standard Error t Value Approx Pr > t Intercept 1.041589 0.219128 4.75 <.0001 GoreRatio 2.580698 0.298749 8.64 <.0001 GoreDWs 0.023180 0.008454 2.74 0.0061 DupG1-0.000426 0.000096911-4.40 <.0001 DdownG2-0.000454 0.000115-3.94 <.0001 LnsG1toG2 0.394799 0.051814 7.62 <.0001 25
Review of Findings: Traditional (Diamond) Ramp F+I & Total Crashes (3 yr) Critical variables include: Number of Driveways in Gore Region Prevailing Number of lanes between G1 and G2 Note: Crashes within 250 of intersection excluded as they were considered crashes due to the intersection (per HSM recommended procedure) 26
3 Year F+I for Traditional Ramps Parameter Estimates Parameter Estimate Standard Error t Value Approx Pr > t Intercept -1.535518 0.555519-2.76 0.0057 GoreDWs 0.066647 0.018193 3.66 0.0002 LnsG1toG2 0.878466 0.204930 4.29 <.0001 27
3 Yr Total for Traditional Ramps Parameter Estimates Parameter Estimate Standard Error t Value Approx Pr > t Intercept -0.920540 0.289482-3.18 0.0015 GoreDWs 0.065629 0.009286 7.07 <.0001 LnsG1toG2 1.175524 0.104178 11.28 <.0001 28
Example Calculation X-Ramp Configuration with: Number of Driveways in Gore Region = 5 Gore Ratio = 0.3 (so 30% of segment is in gore region defined as 400 ft upstream and downstream of gore nose) Distance to closest upstream driveway from G1 = 385 ft Distance to closest downstream driveway from G2 = 360 ft [only critical for total crashes Number of lanes between gores (at weave location) = 3 Predicted number of F+I crashes is 6 (so 2 per year) Predicted number of Total crashes is 24 (so 8 per year) 29
Example Calculation Traditional Ramp Configuration with: Number of Driveways in Gore Region = 5 Gore Ratio = 0.3 (so 30% of segment is in gore region) Distance to closest upstream driveway from G1 = 385 ft Distance to closest downstream driveway from G2 = 360 ft [only critical for total crashes Number of lanes between gores (at weave location) = 3 Predicted number of F+I crashes is 4 (so just over 1.3 per year) Predicted number of Total crashes is 19 (so just over 6.3 per year) 30
Comparison of Example Values for Frontage Road Segment Crashes F+I predicted crashes per year (for segment) Total predicted crashes per year (for segment) X-Ramps 2 8 Traditional Ramps 1.3 (round to 2) 6.3 (round to 7) 31
Additional Observations Number of driveways in gore region critical for both X-ramps and Diamond ramps X-ramp configurations, given the same characteristics, can be expected to result in slightly more frontage road crashes than for diamond applications Number of frontage road lanes is always influential when predicting frontage road segment crashes Phase II (currently underway) -- Evaluating the safety effects of the adjacent freeway segments 32
Phase II Activities Assess the safety effects of the adjacent freeway segments (both at X-ramp and Diamond ramp locations) Develop self-calculating spreadsheet for equations Assemble charts to demonstrate site characteristic sensitivity 33
Questions? Karen Dixon, Ph.D., P.E. Phone: 979-845-9906 Email: k-dixon@tti.tamu.edu 34