Conditions and Performance Chapter Listing

Conditions and Performance Home Page

Introduction

Summary

Highway Operational Performance

Transit Operational Performance

Measuring Traffic Congestion

While congestion is conceptually easy to understand, it has no widely accepted definition. This is because the perception of what constitutes congestion varies from place to place. What may be considered congestion in a city of 300,000 may be greatly different than perceived traffic conditions in a city with 3 million people, based on varying history and expectations. Because of this, transportation professionals examine congestion from several perspectives.

Three key aspects of congestion are its severity, extent, and duration. The severity of congestion refers to the magnitude of the problem, as measured by the average overall travel speed, travel time delay, or the length of queues behind bottlenecks. The extent of congestion is defined by the geographic area (the portion of the population or portion of total travel affected). The duration of congestion is the length of time that the traffic flow is congested, often referred to as the "peak period" of traffic flow.

Daily vehicle-miles of travel (DVMT) per lane-mile is the most basic measure of how much travel is being accommodated on our highway systems since it is a count-based metric. It is based on actual counts of traffic, not on calculations which are in turn based on actual data. The traditional congestion measure in this report has been volume service flow (V/SF), the ratio of the volume of traffic using a road in the peak travel hour to the capacity or service flow of that road. V/SF is limited because it only addresses peak-hour and disregards the duration of congestion. As travel volume grows on a given highway section, after a certain point peak-hour congestion tends to stabilize even as total hours of congestion continue to increase. Focusing only on the V/SF measure alone can lead to erroneous conclusions about highway operating performance. This report adds a new indicator of congestion, delay. Delay incorporates the effects of congestion throughout the day, not only during the peak hour of travel.

DVMT per Lane-Mile

The volume of travel per lane-mile has increased over the past 10 years on every functional highway system for which data are collected. For urban Interstate the rate of increase from 1987 to 1997 is 2.0 percent, and for rural Interstate the rate of increase is 3.4 percent. DVMT per Lane-Mile for each system is shown in Exhibit 4-3. Whatever other measure is used to estimate congestion or its effects, there is no doubt that the density of traffic is increasing, especially on the higher functional systems.

Exhibit 4-3. DVMT per Lane-Mile, 1987-1997

V/SF Ratio

Volume/service flow (also known as the volume/capacity ratio) is a measure of the severity of congestion. The V/SF is the ratio between the volume of traffic actually using a highway during the peak hour and the theoretical capacity of the highway to accommodate traffic. The higher the ratio, the more congested the facility.

Congestion reported in this chapter is based on a threshold value of 0.80. This typically represents Level of Service (LOS) D, as described in Exhibit 4-4. This volume of traffic is 80 percent of the maximum that can be accommodated on a highway, but freedom to maneuver is noticeably limited and incidents result in substantial delays. Higher V/SF ratios represent more severe congestion, escalating into a breakdown in traffic flow at LOS F. Procedures for calculating the V/SF ratio are described in the Transportation Research Board's Highway Capacity Manual (HCM). It should be noted that this measure of congestion is still a subjective issue, even with engineering standards.

Exhibit 4-4. Description of Levels of Service

Exhibit 4-5 describes the percentage of peak-hour urban traffic that operates at a V/SF threshold of 0.80 or higher. The severity of congestion was somewhat greater on urban Interstates in 1997 than in 1993, increasing from 52.6 to 53.3 percent of all peak-hour traffic operating under congested conditions. For the same period peak-hour congestion was declining on other freeways and expressways until 1997, when it increased to 45.7 percent. Meanwhile, congestion severity decreased on other urban principal arterials between 1993 and 1997. Further years of estimating congestion may provide a clearer picture of the long-term trends in congestion.

Exhibit 4-5. Percent of Congested Travel on Urban Principal Arterial Highways, 1993-1997
Peak-hour travel with V/SF >= 0.80 based on 1994 Highway Capacity Manual

Delay

The Federal Highway Administration 1998 National Strategic Plan established a target of reducing delays on Federal-aid highways by 20 percent in 10 years, in terms of hours of delay per 1000 VMT. The delay values used in this report are modeled rather than measured. Currently we have no efficient way to measure delay directly. (See "Future Research," on page 4-12.) Delay is calculated as the difference between estimated actual travel speed and free-flow travel speed. Note that the delay calculations are in terms of vehicle-hours of delay, so that one hour of delay affects the same number of vehicles in one location as another. To the extent that vehicle occupancy differs from place to place, the number of people affected by one vehicle hour of delay may differ.

Delay is a new measure relative to the other two measures used in this report. How well it tracks perceived congestion remains to be seen. Several more years of use will be needed to determine the validity of the procedures used to calculate the value and the credibility of the results.

Exhibit 4-6 shows trends in delay since 1993. For each of the four types of areas shown, delay in 1997 was greater than in 1993. Delay increased from 8.27 to 9.35 hours between 1993 and 1995, but declined to 8.97 hours in 1997. Most urban highways have experienced less delay since 1995. Delay on Urban Interstates has fallen below 1993 levels. As shown in Exhibit 4-7, there is far more delay on Urban Interstates in the areas with more than 200,000 population than in the smaller urban areas or in rural areas.

Exhibit 4-6. Daily Delay, 1993-1997

Exhibit 4-7

Daily Delay on Interstate Highways in 1997 (Hours per Thousand Vehicle Miles Traveled)

Exhibit 4-8

Daily Delay on Other Principal Arterials in 1997 (Hours per Thousand Vehicle Miles Traveled)

The greatest delay occurs on urban other principal arterials, in urbanized areas with more than 200,000 residents. These are higher-level roads that are accommodating metropolitan growth. As shown in Exhibit 4-8, delay on these routes was 50 percent greater than delay on the same functional system in small urban areas under 50,000 population.

Despite the overall decline in delay observed since 1995, rural delay continues to increase. Every rural functional system had higher average delay in 1997 than in 1995.