An analysis was performed to quantity the accident effects of lane and shoulder widths on rural roads carrying fewer than 2,000 vehicles per day. The primary data base used in the research contained accident and roadway characteristic information for more than 6600 km (4,100 mi) of two-lane roadway sections in seven states. Independent data bases from three states (Minnesota, Illinois, and North Carolina) for roadways totaling more than 86,000 km (54,000 mi) were selected to validate the accident relationships found in the primary data base. Analysis of covariance was used to quantify accident relationships on these low-volume roads. Single-vehicle and opposite-direction accidents were classified as "related" accidents because the accident rates for these two types were found to be related to differences in lane and shoulder widths. The rate of related accidents was also affected by roadside hazard, roadway terrain, the number of driveways per mile, and state differences. No differences in accident rates were found between roadways with paved and unpaved shoulders. For lane widths of at least 3.0 m (10 ft), related accident rates were lower when wide shoulders were present than when narrow shoulders were present. For a given shoulder width, wider lanes were found to be associated with lower accident rates. Somewhat counterintuitively the accident rate was higher for 3.0-m (10-ft) lanes with narrow shoulders than for 2.7-m (9-ft) lanes with narrow or wide shoulders. For traffic volumes of 250 vehicles per day or less, accident rates did not differ significantly between paved and unpaved roads. For traffic volumes of greater than 250 vehicles per day paved roads have significantly lower accident rates than unpaved (dirt and gravel) roads. The research findings indicate that on low-volume roads lane widths as narrow as 2.7 m (9 ft) may be acceptable from a safety standpoint under certain conditions. The 1995 draft AASHTO policy chapter on local roads includes revised roadway width guidelines that reflect many of the research findings presented.
[1]
W W Hunter,et al.
PAVED HIGHWAY SHOULDERS AND ACCIDENT EXPERIENCE
,
1974
.
[2]
T J Foody,et al.
THE IDENTIFICATION OF RELATIONSHIPS BETWEEN SAFETY AND ROADWAY OBSTRUCTIONS REPORT & EXECUTIVE SUMMARY
,
1974
.
[3]
Patrick Shannon,et al.
PAVEMENT WIDTH STANDARDS FOR RURAL TWO-LANE HIGHWAYS. ABRIDGMENT
,
1976
.
[4]
E A Rinde,et al.
ACCIDENT RATES VS. SHOULDER WIDTHS. TWO LANE ROADS, TWO LANE ROADS WITH PASSING LANES (SEPT 1977)
,
1977
.
[5]
Charles Zegeer,et al.
COST-EFFECTIVENESS OF LANE AND SHOULDER WIDENING OF RURAL, TWO-LANE ROADS IN KENTUCKY
,
1979
.
[6]
Daniel S Turner,et al.
BEFORE-AFTER ACCIDENT ANALYSIS FOR TWO SHOULDER UPGRADING ALTERNATIVES
,
1982
.
[7]
W. Hunter,et al.
Safety effects of cross-section design for two-lane roads. Volume 1
,
1987
.
[8]
William W. Hunter,et al.
SAFETY EFFECTS OF CROSS-SECTION DESIGN FOR TWO-LANE ROADS - VOLUME I - FINAL REPORT
,
1987
.
[9]
C Zegeer,et al.
Roadway widths for low-traffic volume roads
,
1994
.
[10]
L King,et al.
THE NEW AASHTO METRIC POLICY ON GEOMETRIC DESIGN OF HIGHWAYS AND STREETS
,
1995
.