Office of Operations
21st Century Operations Using 21st Century Technologies
Table of Contents

Traffic Analysis Toolbox Volume X:
Localized Bottleneck Congestion Analysis
Focusing on What Analysis Tools Are Available, Necessary and
Productive for Localized Congestion Remediation

2.0 Background and Discussion

The term "bottlenecks," in the context of this guidance, is hereafter confined to the genre of "recurring" traffic bottlenecks, as opposed to "nonrecurring" ones. Recurring bottlenecks are predictable in cause, location, time of day, and approximate duration; e.g., the ones that we encounter in our everyday commutes. Nonrecurring bottlenecks are random (in the colloquial sense) as to location and severity. Examples include crashes, weather events, and even "planned" events, such as work zones and special events, all of which are irregular in occurrence and location.

Let's dispense with nonrecurring bottlenecks for a moment. Nonrecurring bottlenecks are more prone to empirical study; i.e., based on or characterized by observation and experiment instead of theory. Said bottlenecks trigger traffic control plans (TCP) that are either premeditated or reactionary to the event. Tweaking the plan can improve it either in real-time or "for the next time" it is needed. "Dynamic Lane Merges" (DLM) are increasingly being tested, empirically, to increase the safety and efficiency of nonrecurring bottlenecks. DLMs essentially are active traffic management plans that "kick in" when excessive queues are detected, say in work zones. Messages are enacted that display proactive information on how and when to merge. The messages shut off when the queues begin to dissipate. Recurring bottlenecks, however, have historically been studied by the academic community using non-empirical means, like microsimulation.

A "localized" recurring bottleneck may be considered to be a defined event (i.e., cause) in a defined location; e.g., a lane drop, a weave, an intersection, or an off- or on-ramp. For example, repeat congestion at one movement of an interchange over a couple of hours each day would be "local," whereas a "mega" bottleneck or systemic congestion would be considered to be an undersized interchange, and is not the focus of this guidance.

It hardly needs to be said that in the "mega" case, micro simulation is always warranted due to the complexity of the facility or facilities. However, it is recognized herewith that lesser problems typically require comparatively lesser study and solutions. The key is finding the cutoff point at which project execution meets project need, decision-justification, and budget, both in terms of project analysis and project implementation. For example, the insufficiency of a left turn phase at a ramp terminal (causing queuing back down the ramp to the mainline) would not warrant a full-blown study of the mainline, as much as it would warrant an adjustment of the signal timing.

2.1 Common Causes of Bottlenecks

Recurring, localized bottlenecks occur any time the rate of approaching traffic is greater than the rate of departing traffic. The causal effect can usually be attributed to the existence of at least one of two factors:

  • Decision points, such as on and off-ramps, merge areas, weave areas, lane drops, tollbooth areas, and traffic signals; or
  • Physical constraints, such as curves, underpasses, narrow structures, or absence of shoulders.

Recurring bottlenecks usually disperse from the rear of the queue, as the volume crush dissipates and the confluence regains its ability to process the traffic more or less under free flow conditions. Nonrecurring bottlenecks, as a point of differentiation, can disperse from the front or rear, depending on whether the cause is incident-related (e.g., crash or work zone) or volume-related (e.g., special event crush load), respectively.

One can even imagine a compounded situation, where a decision point (off-ramp) is preceded by a physical constraint (sharp curve). This type of bottleneck congestion is more complex to mitigate because both the decision point and physical constraint must be addressed to deal with the bottleneck. Further, it is difficult to predict the largest contributor if there are multiple causes.

Each bottleneck cause has its own mitigation strategies. To select the appropriate strategy, planners must understand the bottleneck's causes before attempting to prescribe solutions, which will be discussed in Section 3.0.

2.2 What Role does Analysis Play?

As transportation agencies continue to seek innovative, cost-efficient solutions to reduce and eliminate bottlenecks, analysis of alternatives has become a necessary decision-support process.

By definition, the planner models the study area and measures the performance of several preselected criteria. If no improvements were made ("no-build" scenario), how would the corridor operate in the future? Conversely, what effects would be incurred if the alternatives were implemented? Alternatives analysis can be developed to compare operational forecasts under different scenarios.

Because analysis is useful for so many stages of the decision-making process, a variety of methods exist. It is important to note that the methods vary greatly; no one tool can model all scenarios or proposed improvements. Thus, selecting the appropriate tool based on the goals and objectives of the project is critical, and is the focus of this guidance.

2.3 Bottleneck Mitigation

In some cases, the most cost-effective ways to relieve bottlenecks are through the simplest geometric or operational improvements. Many of these solutions can be executed as safety projects, Federal-aid non-exempt projects, or even maintenance activities. When applied properly, these strategies can produce very high benefit-cost ratios because of the smaller footprint solution, the lower-cost design solution, and the lower life-cycle cost, including planning, design, construction, operations, and maintenance. Some of the more common low-cost mitigation strategies include the following:

  • Signal retiming. Many congested corridors can achieve bottleneck reductions by simply optimizing the timing of traffic signals or their timing offsets between intersections.
  • Restriping. Remarking traffic lanes to add auxiliary lanes or acceleration/deceleration lanes can increase capacity or redirect volume more efficiently.  Some of the common restriping techniques include preventing weaves or sharp turns that cause slowdowns; restriping lanes to provide more, although slightly narrower lanes; or converting short sections of shoulders into travel lanes.
  • Signage and Signals. Signs and signals can be designed to purposely restrict specific movements to the benefit of others (i.e., STOP sign, YIELD sign at the minor approaches) or prohibit inefficient movements (i.e., restrict U-Turns or left turns crossing heavy opposing traffic). On the flipside, these strategies also can be used to prioritize heavy movements to prevent bottlenecks from forming (i.e., right turn on red, and providing exclusive left-turn signals).
  • Installing Loop Detectors. Installing loop detectors ahead of traffic lights can help reduce queuing by dynamically prioritizing the busiest approaches on-demand.
  • Ramp or Driveway Removal (or Modification). Closing, relocating, metering, or combining ramps, especially low-volume ones, can unclog some traffic streams. In the case of ramp modifications, temporary closures can test a hypothesis. The ramp could always reopen if the cure is worse than the symptom.

In other cases, more costly solutions might be necessary, including rebuilding or redesigning the area in the vicinity of the bottleneck location. These strategies also can produce significant benefit-cost ratios, but the cost will always be higher than the strategies listed previously. These strategies may have high life-cycle financial costs (planning, design, construction, operations, and maintenance) or social costs (such as forcing drivers to relearn lane directions or turns). Regardless of the nature of the cost, these improvements must be planned well before their implementation, because they are costly and difficult to undo. The following project examples include:

  • Washington State DOT Integrated Operations/Construction Programs in Puget Sound Region and Seattle. A new exit ramp was constructed along I-405/SR 67 to minimize weaving.
  • Post Street Restriping Project. In San Francisco, California, Post Street between Kearny Street and Montgomery Street was a two-way street that was converted to a one-way street to increase its capacity during the p.m. peak-period.

There are many successful case studies of transportation agencies implementing low-cost, high-cost, or a combination of solutions to relieve recurring, localized congestion. The Federal Highway Administration's (FHWA) web site https://ops.fhwa.dot.gov/bn/index.htm has many other brief examples of localized bottleneck solutions.