16-03-2012, 05:13 PM
Traffic Flow on Freeway Upgrades
TrafficFlowonFreewayUpgrades TRB.pdf (Size: 355.59 KB / Downloads: 43)
INTRODUCTION
Upgrades on freeways impose lower velocities on vehicles , which may lead to capacity losses.
While heavy vehicles are affected most severely, steep gradients have an impact on passenger
cars as well. As a consequence inclined sections constitute bottlenecks within the freeway
network. Therefore, the traffic flow on upgrades is of significant importance for the
determination of the required freeway dimensions and for traffic flow qualities within the
network. This is the reason why traffic flow on freeway upgrades plays an important role in the
current guidelines for highway capacities in many countries.
The American Highway Capacity Manual (HCM, 2000) traditionally describes the impact of
upgrades on traffic flow by passenger car equivalents for heavy vehicles, which strongly depend
on the grade and length of gradient for different types of vehicles.
The corresponding German guidelines RAS-Q (1) traditionally do not to use passenger car
equivalents (i.e. pcu values) for freeway traffic flow analysis. They describe traffic volume based
on “vehicles / hour” (sum of vehicles over all lanes; with the unit of “1” for motorized vehicles
of all types; traffic composition is described by the percentage of heavy vehicles ; = vehicles with
a maximum weight above 3.5 tons). The measure of effectiveness (MOE) for freeways is
represented by the travel velocity of passenger cars over longer sections of the freeway (e.g. 5
km). In the former RAS-Q (1) the impact of upgrades on freeways was described based on
Brannolte's investigations (2).However, it was found necessary to recalibrate these relationships
for the new guideline HBS (3), which is a German version of the Highway Capacity Manual.
These new investigations were embedded in a research project conducted by the authors at the
Ruhr University of Bochum on behalf of the German federal DOT (4, 5).
The objective of this research project was to develop speed-flow diagrams representing typical
traffic conditions for each combination of length and degree of gradient on German motorways.
These curves were to form the basis of the Freeway chapter in the HBS (3). Moreover, criteria for
the construction of additional lanes on upgrades (“crawler lanes”) had to be developed based on
Brilon, Bressler 3
3
traffic flow characteristics, traffic safety, and cost-benefit considerations. This paper mainly
focuses on the derivation of capacities and traffic flow characteristics for freeway upgrades.
The methods employed include empirical investigations based both on microscopic
measurements and continuous traffic monitoring as well as on microscopic simulation. Here the
simulation model was calibrated based on empirical data. The results have been analyzed by
statistical methods so as to develop models simple enough to be included into a guideline.
EMPIRICAL DATA ANALYSIS
First of all, a series of measurements was conducted at 8 sections of 2- and 3-lane freeway
upgrades in different parts of Germany. Travel time measurements were made by video
recordings of traffic situations on 5 different freeways, in each case at two points spaced between
2 and 3,6 km apart along the upgrade. Here the gradients varied between 2% and 6%, while the
total length of the uphill section varied between 2.5 and 5.6 km. At each section the travel speed
between two cross sections (distance between 1.6 and 2.6 km) was evaluated based on license
number identification. Each data set contained several 1000 vehicles at traffic flows between
1000 and 3000 veh/h, indicating that free flow conditions were generally observed during
measurements. Fig. 1 gives an impression of the range of observed traffic conditions regarding
both speed and volume. The highest speeds were observed on the A43 downgrade. This
downgrade section had been included for comparison. The low speeds on the A8 were influenced
both by the steep gradient (+5.6 %) and the traffic restrictions (speed limit: 100 km/h + truck
overtaking prohibition). The results show a clear impact of the degree of gradient on travel
speeds.
In addition, a local traffic flow analysis was performed at the upper end of each of the sections
under observation. The results are documented in (4). In this paper the emphasis is mainly on
travel velocities over extended freeway sections.
Another and much more comprehensive set of data was collected from 5 longer freeway upgrade
sections equipped with continuous counting facilities in connection with variable speed limit
control devices. In each of these instances, more than two successive measurement points along
the whole upgrade section could be analyzed continuously over more than one week (with raw
data at 1-minute intervals). As an example Fig. 2 shows a speed-flow diagram from the A43-
upgrade (s = 3.5 %; a section of the A43 different from that in Fig. 1) based on 5-minute
intervals, with an average truck percentage of 8 %. In this context, average speed always
corresponds to the space mean speed calculated from local measurements. We see free flow
traffic situations and congested intervals at both locations (200 m and 1000 m from the lowest
point). The maximum observed volumes, thus, give us an indication of the capacity, which at
around 3600 veh/h seems to be identical for both locations . This is regarded as typical for 2-lane
(per direction) freeways in Germany. Comparing data points from both measurements shows that
capacity is not affected by longer upgrades. At the downstream point (Q7), however, free-flow
speeds appear reduced due to the upgrade while on the other hand, congested speeds have
increased.
This effect was observed at all of the 5 test sections which all experienced traffic breakdowns
during the period of investigation. One rather important result came out clearly: traffic