Overview
Travel time reliability reflects the distribution of travel time of trips using a facility over an extended period of time. This distribution arises from the interaction of a number of factors that influence travel times:
- Recurring variations in demand, by hour of day, day of week, and month of year;
- Severe weather (e.g., heavy rain, snow, poor visibility) that reduces capacity;
- Incidents (e.g., crashes, stalls, debris) that reduce capacity;
- Work zones that reduce capacity and (for longer-duration work) may also influence demand; and
- Special events (e.g., major sporting events, large festivals or concerts) that produce temporary, intense traffic demands which may be managed in part by changes to the facility’s geometry or traffic control.
There are two widely held ways that the same underlying distribution of travel times can be characterized. Each is valid and leads to a set of performance measures that capture the nature of travel time variability. They are:
1. Measures of the variability in travel times that occur on a facility or a trip over the course of time, as expressed through metrics such as a 50th, 80th, or 95th percentile travel time.
2. Measures of the reliability of facility travel times, such as the number of trips that fail or succeed in accordance with a pre-determined performance standard, as expressed through metrics such as on-time performance or percent failure based on a target minimum speed or travel time.
For convenience, the remainder of this chapter uses the single term reliability to characterize both the variability-based and reliability-based approaches to characterizing the same facility travel time distribution. A sufficiently long history of travel times is required to establish a facility’s travel time distribution—a year is generally long enough to capture most of the variability caused by the factors listed above.
The Highway Capacity Manual’s (HCM’s) freeway and urban street facility procedures (Chapters 10 and 16, respectively) describe average conditions along the facility during a user-defined analysis period, typically the peak 15 min of a peak hour, under typical conditions (e.g., good weather, no incidents). Because this value is an average, there will be times of the day or days during the year when conditions are better than the average, due to lower-than-average traffic demands. There will also be days when conditions are much worse, due to incidents, severe weather, unusually high demand levels, or a combination of these.
Chapter 36, Travel Time Reliability, presents methods that can be used to describe how often particular operational conditions occur and how bad conditions can get. This chapter’s variability and reliability performance measures can be used as the basis for quantifying the degree of severity of level of service (LOS) F (oversaturated) conditions, for developing agency performance standards for oversaturated facilities, and for quantifying the impacts of physical and operational measures designed to improve travel time reliability. Because travel time reliability is a new concept for the HCM, this chapter devotes a number of pages to describing the reliability concept, how reliability can be measured, and how reliability can be applied to analyses to better inform their results:
- The remainder of Section 1 presents definitions of reliability terms along with a high-level overview of the reliability methodology.
- Section 2 presents travel time variability and reliability concepts, including performance measures, illustrative reliability results from U.S. freeway and urban street facilities, potential data sources, and guidance on interpreting reliability results.
- Sections 3 and 4 describe at a high level the travel time distribution estimation methods for freeway and urban street facilities, respectively. These descriptions omit many of the computational details. Readers wishing a greater level of detail about the methods are referred to Chapter 37, Travel Time Reliability: Supplemental for the computational details. The cell formulas and Visual Basic macros in the FREEVAL-RL and STREETVAL computational engines, available in the Technical Reference Library in the online HCM Volume 4, provide the greatest level of detail.
- Section 5 presents default values for the methods, describes potential applications (use cases) for reliability analyses, and addresses the role of alternative tools (such as simulation) in evaluating travel time reliability.
- Section 6 provides seven example problems illustrating the application of the reliability methods to a freeway facility and an urban street facility.
- Section 7 lists this chapter’s references.
Chapter 37, Travel Time Reliability: Supplemental, provides the computational details of the reliability methodologies, presents variability statistics for a number of U.S. freeway and urban street facilities, and provides a method for measuring variability and reliability in the field.