27
The number of freeway construction and rehabilitation projects continues to increase because the highway infrastructure in the United
States is aging. The resulting work zones limit the number of vehicles operating within the affected freeway system. Each vehicle
traveling in a work zone is affected by features such as additional
signs, narrowed lanes, barrels or barriers close to the travelway,
trucks entering the construction area, reduced speed limits, and
workers near the open lanes. A lane closure generates the greatest
constriction to vehicles in a freeway work zone. Drivers accustomed
to the original configuration change their travel behavior to traverse
the area adjacent to the closed lane. Frequently, more vehicles want
to use the road than can be accommodated by the reduced lane configuration. The result is generally the formation of a queue of vehicles upstream of the lane closure.
Most state departments of transportation use the capacity values
in the Highway Capacity Manual (1) as a guide for analyzing work
zone lane closures. These values were primarily collected by the
Texas Transportation Institute (TTI) for urban freeways in Texas.
TTI updated these capacity values in the early 1990s (2). For purposes of comparison the updated values of 1,575 vehicles per hour
per lane (vphpl) for a two-lane to one-lane urban closure and 1,460
vphpl for a three-lane to two-lane urban closure will be used. The
major limitation to applying Texas data to the rest of the nation is
that the Texas freeway network has an extensive frontage road system, which permits vehicles to bypass congested segments of highways. Most other freeway systems in the United States do not have
K.K. Dixon, Georgia Institute of Technology, School of Civil and Environmental Engineering, Atlanta, Ga. 30332-0355. J.E. Hummer, North Carolina
State University, Civil Engineering Department, Box 7908, Raleigh, N.C.
27695-7908. A.R. Lorscheider, North Carolina Department of Transportation, P.O. Box 2501, Raleigh, N.C. 27611.
28
could be at various locations within the work zone for a few minutes
to several hours. The research described suggests two specific work
zone volumes that are appropriate for capacity analysis.
CAPACITY DEFINITIONS
Freeway Work Zone Capacity
Several previous studies attempted to identify the capacity of a freeway work zone. A late-1960s California study measured volumes
for 3-min intervals during a lane closure with congested conditions.
The California analysts averaged two consecutive 3-min counts separated by 1 min; they then multiplied the average value by 20 to
determine the 1-hr capacity values (4). The study did not adjust traffic volumes for the presence of heavy vehicles.
TTI initiated a series of work zone studies in the late 1970s and
early 1980s. Those research efforts identified capacity as full-hour
volume counts in lane closures with traffic queued upstream. The
researchers considered consecutive hours at the same location as
independent studies (5 ). Although the team monitored truck presence during data collection, it did not adjust volumes to passengercar equivalents.
TTI performed a short-term freeway work zone study (2) and
defined and measured capacity as the mean queue-discharge rate at
a freeway bottleneck. The team recorded measurements hourly;
however, the researchers averaged consecutive hour volumes at the
same site and recorded this average as one observation. TTI analysts
converted capacity to passenger-car equivalents by adjusting the
value by the heavy-vehicle factor of 1.7 presented for level terrain
in the 1985 Highway Capacity Manual (6 ). The TTI team assumed
that the standard adjustment factor for heavy vehicles during normal
traffic operations is directly applicable to freeway work zones. The
assumption of heavy-vehicle-factor transferability to work zones
permitted comparison of values between sites but did not address
changes in heavy-vehicle operation during queued conditions. Previous research, therefore, provides several definitions and measurements for work zone capacity. The exact physical freeway location
at which capacity occurs also seems to be elusive, but it is generally
assumed to occur at the end of the transition area.
Dixon et al.
29
TABLE 1
30
of the work zone (to verify free-flow status) and at consistent locations within the work zone.
The project team elected to use Vehicle Magnetic Imaging traffic
counters and classifiers developed by Nu-Metrics. The researchers
positioned these computer-driven devices in the center of a lane with
an adhesive tape coat or mounting straps nailed in place. Information about device type, operation, and placement can be obtained
from the project report Capacity and Delay for North Carolina
Freeway Work Zones (11).
The team primarily collected data in 5-min time bins and analyzed space-mean-speeds within these bins in a manner consistent
with previous freeway research. Select sequential data were collected for headway analysis and delay studies. The results of the
headway and delay studies will be available for publication in the
near future.
Figure 2 shows typical placement of capacity analysis devices for
a standard 2-to-1. The team placed devices that monitor speed and
volumes at an advanced warning location that was before or adjacent to any signs warning motorists of impending lane closures.
Motorists in the traffic stream could not see the lane closure from
the advanced warning location. These advanced warning devices
verified normal free-flowing conditions and tested the assumption
that capacity limitations within the work zone were a result of
the construction and were not normally present due to recurrent
congestion.
A classifier was typically located at the end of the transition area.
Previous work zone research (2) identified the end of the transition
location as the critical capacity location, so a speed-flow relationship was evaluated at that location whenever feasible. An additional
classifier was positioned adjacent to the activity area (located
approximately at the middle of the construction zone) to make possible a speed comparison between vehicles located adjacent to the
activity area and vehicles entering the work area. The activity-area
device location was static (i.e., once put in place the device did not
move until data collection terminated at the site), but the construction activity typically moved forward (in direction of travel) over
time. As a result the activity-area device monitored speed adjacent
to active work for only a portion of the collection period. Counters
were typically positioned at the beginning of the transition to provide insight into vehicle lane changing within the transition region.
The team used similar device configurations at 3-to-2 and 3-to-1
lane closures. Because the number of devices was limited placement
at crossover configurations did not include the advance-warning
device location.
Dixon et al.
31
FIGURE 3
32
and night capacities (1,637 versus 1,644 vphpl) implies queued vehicles (in Segment 3) behave similarly during day and night. Behavior at the end of transition in Segment 1 does not appear to exhibit
the same characteristic. The average speed during the day was essentially constant at 85 kph (53 mph) for volumes below 1,100 vph. The
constant pattern is consistent with the uninterrupted facility speedflow models provided in the Highway Capacity Manual (1).
A time sequence speed-flow plot for the night data at Site 7
showed a reduction in speed as volumes increased in Segment 1.
This observed speed-flow behavior might be explained by the presence of light towers, night glare, or reduced peripheral visibility.
The speed of vehicles at the end-of-transition location in Segment 1
appears to be influenced by light conditions, whereas the queued
behavior appears to remain consistent during day and night. The
queues occurred primarily when vehicles operated at speeds less
than 48 kph (30 mph).
Dixon et al.
33
adjacent to the activity area. A greater variability in capacity observations occurred adjacent to active work than at the end of transition, likely due to the dynamic work activity.
A smaller capacity occurred at the activity-area device than at the
end of transition for all the rural observations. Generally, Table 3
shows a reduction in rural capacity of approximately 100 vph. The
lower capacity value indicates that closure capacity should be evaluated at the physical location where vehicles are subject to the greatest restriction. Sites with heavy work activity are subject to the
greatest limitation adjacent to construction instead of at the end of
transition. The transition region functions as an initial bottleneck,
whereas the activity area represents a second, more constricted
bottleneck. Essentially, the lane transition meters vehicles before
their encounter with the work activity. The presence of two queues
(one at the transition and one within the work zone) is likely during
the early stages of congestion, whereas later the work area queue
may grow backwards to envelop the transition queue.
Table 3 also compares activity area and end-of-transition capacities for the 2-to-1 urban sites. The distinct reduction of capacity
observed at the rural sites did not clearly occur at the urban sites.
The data indicate reduced capacity at Sites 9 and 14 in the vicinity
of active work, whereas Sites 7 (daytime) and 16 exhibited relatively
higher capacity values. Active work data were not available for the
night construction at Site 7. The active work capacities generally
appear to conform to type of work activity. Urban sites with moderate and heavy work activity exhibited capacities ranging from
1,458 to 1,560 vph. Site 16 did not exhibit reduced capacity values
at the activity area; however, as previously discussed, the end-oftransition capacity observed was considerably lower than it was at
other urban 2-to-1s.
Activity area data at Site 10 (a 3-to-1) exhibited a queued condition for 15 min. The short queue duration resulted in an activity area
capacity of 1,440 vph; however, the limited duration of capacity
conditions restricts the usefulness of this short-term value. A capac-
ity reduction appeared to occur for the 3-to-1 adjacent to the activity area in a fashion similar to that observed for the 2-to-1 closures.
Comparison of Capacities
Table 4 summarizes general work zone capacities observed on
North Carolina freeways. The North Carolina values are compared
with the updated Texas queue-discharge values given in Table 1.
The Texas volumes appear to conform to values observed in North
Carolina moderate-construction activity areas. The North Carolina
end-of-transition observations exhibited, on average, a volume 10
percent greater than the Texas end-of-transition values. Rural work
zone values for Texas were not available. The higher capacity finding supports the theory that two volumes are critical to work zone
capacity analysis. The collapse volume represented by Segment 2 in
Figure 1 (identified as work zone capacity) and a lower queuedischarge value occurring following queue formation both must be
considered. A future comparison of activity-area queue-discharge
values to activity-area work zone capacity appears warranted. The
lower rural capacity values (compared with urban) emphasize the
dramatic influence driver familiarity has on traffic operations within
work zones. The lack of a parallel frontage road system and the
possible East Coast driver influence may have a minor influence on
values obtained for the study described.
34
Recommendations
The small number of work zone sites with 3-to-2, 3-to-1, and
crossover configurations prohibited the determination of conclusive
capacity values for these scenarios. Additional capacity analysis for
these lane closure configurations should be pursued. Analysis of
work zones located in more than one state may be necessary to
gather a large enough data base.
Additional analysis of truck influence in work zones is also suggested. Finally, the development of a table representing volumes for
queue initiation (collapse or capacity value) and queue discharge for
each lane closure configuration is warranted.
ACKNOWLEDGMENTS
FHWA and NCDOT provided support for the research described
herein. The authors thank Troy Peoples, Terry Hopkins, and Mike
Stanley of NCDOT for their guidance; Laura Cove of FHWA for her
insight; and Marcus Cole of NCDOT for his project coordination
efforts.
REFERENCES
1. Special Report 209. Highway Capacity Manual. TRB, National
Research Council, Washington, D.C., 1994.
2. Krammes, R. A., and G. O. Lopez. Updated Short-Term Freeway Work
Zone Lane Closure Capacity Values. Report FHWA/ TX-92/1108-5.
FHWA, U.S. Department of Transportation, and Texas Department of
Transportation, Austin, 1992.
3. Part VI Standards and Guides for Traffic Controls for Street and Highway Construction, Maintenance, Utility and Incident Management
Operations. Manual on Uniform Traffic Control Devices, 1988 edition,
Revision 3. FHWA, U.S. Department of Transportation, 1993.
4. Kermode, R. H., and W. A. Myyra. Freeway Lane Closures. Traffic
Engineering, Vol. 40, No. 5, 1970, pp. 1418.
5. Dudek, C. L., and S. H. Richards. Traffic Capacity Through Work Zones
on Urban Freeways. Report FHWA/TX-81/281228-6. Texas Department of Transportation, Austin, 1981.
6. Special Report 209. Highway Capacity Manual. TRB, National
Research Council, Washington, D.C., 1985.
7. May, A. D. Traffic Flow Fundamentals. Prentice-Hall, Inc., N.J., 1990.
8. Hall, F. L., V. F. Hurdle, and J. H. Banks. Synthesis of Recent Work on
the Nature of Speed-Flow and Flow-Occupancy (or Density) Relationships on Freeways. In Transportation Research Record 1365, TRB,
National Research Council, Washington, D.C., 1992, pp. 1218.
9. North Carolina Department of Transportation Roadway Design
Standards. North Carolina Department of Transportation, Raleigh, 1992.
10. Manual on Uniform Traffic Control Devices. FHWA, U.S. Department
of Transportation, 1988.
11. Dixon, K. K., and J. E. Hummer. Capacity and Delay for North
Carolina Freeway Work Zones. Final report. North Carolina Department of Transportation, Raleigh, 1995.
12. Traffic Control in Saturated Conditions. Organization for Economic
Co-Operation and Development Road Research Group, 1981.
13. Rouphail, N. M., and G. Tiwari. Flow Characteristics at Freeway Lane
Closures. In Transportation Research Record 1035, TRB, National
Research Council, Washington, D.C., 1985, pp. 5058.
The views and opinions expressed in this paper are those of the authors and
do not necessarily reflect the views and opinions of FHWA, NCDOT, or
North Carolina State University. The authors assume full responsibility for
the accuracy of the data and conclusions presented in this paper.
Publication of this paper sponsored by Committee on Traffic Safety in Maintenance and Construction Operations.