Highway Capacity Manual Methodologies for Corridors Involving Freeways and Surface Streets (2020)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

- 8 - 3. Critical Review of the HCM This section presents a critical review of the methods in the Highway Capacity Manual 6th Edition, highlighting strengths and identifying weaknesses related to the evaluation of trips using freeways and surface streets. Overview of Performance Measures and Service Measures for HCM Analyses The HCM uses the service measures shown in Figure 1 for each type of facility. As shown, the HCM also provides for each facility a systems performance measure that can be converted to travel time. An HCM procedure that includes freeways, highways, and urban streets simultaneously needs to have a common set of performance measures, and ultimately, a common service measure. Other performance measures unique to each facility type may be useful for a variety of purposes, but in order to consider all facilities as an integrated system, a common and consistently used measure is essential. Source: Highway Capacity Manual 6th edition - Exhibit 2-2 Figure 1 – HCM service measures by system element and mode A detailed discussion on performance measures and the proposed trip-based evaluation framework is presented in Section 4 – Discussion on Performance Measures. Freeway Facilities This subsection presents and discusses specific issues related to HCM freeway facilities analysis that pertain to systems analysis. Performance measurement: Freeway analysis is based on tools that estimate the operational performance of each type of segment. The freeway systems analysis method evaluates operations in time and space considering, to some degree, interactions (queue effects) between consecutive segments. Operational performance is determined based on density and speed along each segment. In the case of

- 9 - ramps, performance measures are estimated separately for the influence area (consisting of the two rightmost lanes, and for a length of 1,500 ft.) and the remaining leftmost lanes. Based on these methods, one can derive travel time for each segment and stitch together the travel time for the entire facility. However, the quality of a particular trip from an origin to a destination cannot be obtained, because the method does not provide operational performance by lane. In the field, speed can vary widely depending on the lane used, and each origin-destination (O-D) is likely to use a specific lane or set of lanes along each segment. For example, travelers exiting at a congested off-ramp will experience a much different travel time than those using the left most lanes of the same segment. Currently, LOS for the facility is estimated as a function of the LOS for each segment, and does not consider the quality of the trip from a traveler’s perspective. However, travelers experience travel time over the entire system, as well as based on their O- D. A performance measurement scheme for the entire system should also consider the origins and destinations of the travelers within each segment. Furthermore, five of the six LOS ranges exist where speeds are relatively high (above approximately 50 mi/h) and only one LOS range is used to define the congested regime, which is of the highest interest in large urbanized areas. Many congestion management techniques will improve congestion (e.g., delay) but the facility will still be classified as LOS F under the current framework. Spillback from a downstream facility: The HCM addresses queue spillback only when it is fully contained within the boundaries of a freeway facility. Spillback onto the freeway may occur either due to inadequate capacity of the ramp proper, or due to inadequate capacity at the ramp terminal (typically the signal at the downstream interchange). The capacity of the ramp proper is defined as the off-ramp’s maximum allowable hourly flow rate based on its geometric characteristics (number of lanes, free-flow speed, etc.). The capacity of the ramp terminal is defined as the capacity of the signalized or unsignalized approach to the surface street. The current procedures in HCM provide guidance on estimating queue storage ratio on ramp terminals, but do not directly address situations when ramps have queue-to-storage ratio > 1.0 (LOS F at diverge ramps). Influence area for merge and diverge areas: Currently, the HCM methodology for merge and diverge areas focuses on predicting performance within a 1,500 ft. influence area and for the two rightmost lanes. Figure 2 provides a schematic of the diverge influence area, according to the HCM. When spillback occurs, it is likely that queue length extends upstream beyond 1,500 ft. Thus, the influence area of the junction may be significantly longer. Also, the influence area may vary by time period depending on the demand-to- capacity ratio and its variability. Lastly, the effects of spillback may affect additional lanes, as through vehicles attempt to avoid the spillback and they increasingly use the leftmost lanes to maintain their speed. The current freeway systems analysis framework, which uses a constant area (longitudinal distance and cross section) to define this influence area, must be revised to consider spillback effects. Source: Highway Capacity Manual 6th Ed. - Exhibit 14-7 Figure 2 – Schematic of a diverge influence area (HCM)

- 10 - Impacts of spillback on weaving segments: Similar considerations to those described above are also relevant for weaving segments. Along weaves, the on-ramp that is part of the weave may be affected in different ways based on whether the spillback reaches its gore point or not. The area available to the two weaving traffic streams would be reduced as a function of spillback. When longer queues prevail, the upstream on-ramp may be entirely blocked, and it is possible that in this case the on-ramp and off-ramp function independently, as there is no space available for weaving maneuvers. Capacity constraints and speed reduction due to spillback: Spillback along the mainline results in blockage of one or more lanes. Thus, capacity is reduced during such conditions, and the capacity reduction depends on the frequency and duration of lane blockage. Also, lanes adjacent to those blocked are likely to have reduced operating speeds. Those effects and their impact on the overall capacity and quality of service must be addressed. Ramp throughput during congested conditions: In order to consider the effects of an on-ramp queue to upstream surface streets, it is necessary to estimate the queue length as a function of demand and capacity. However, when the freeway mainline is congested, it is not clear what the capacity of the ramp is. In other words, the HCM does not provide the actual relative contributions of the ramp demand vs. the mainline demand at the merge during congested conditions; it assumes that the merge operates as a “zipper”, with equal contributions from the shoulder lane and the on-ramp. However, the resulting ramp flow may be a function of geometry, the extent and duration of congestion on the mainline, or prevailing driver behavior in the region. The ramp capacity may be higher for weaving segments with lower freeway to off-ramp demands and a longer weaving length. Estimating this discharge rate is necessary in order to estimate the resulting queue length along the on-ramp. In cases where ramp metering is present, the discharge rate can be determined and used to obtain the mainline input flow as defined by the HCM for oversaturated freeway segments in Chapter 25. Spillback effects along the mainline: Revised analyses should also consider the effects of spillback to upstream segments, including basic freeway segments. For example, current procedures determine whether the subject diverge ramp operates as an isolated ramp, considering the respective freeway and ramp demands. These procedures should be revised to consider that the area of influence may vary as a function of queue length and other factors related to spillback (for example the number of lanes blocked and their respective frequency). Additionally, the lanes of the upstream segments may be differently affected, depending on the spillback queue length. Urban Streets This subsection presents and discusses specific issues related to HCM urban streets analysis that pertain to systems analysis. Performance measurement. The urban streets methodology considers a variety of different facilities, including street segments, signalized intersections, interchange ramps terminals (IRT), two-way stop- controlled intersections (TWSC), all-way stop-controlled intersections (AWSC) and roundabouts. Each one of these facilities have an associated MOE; intersections use control delay, while segments use travel speed. Thus, development of a performance measurement framework for this group of facilities is easier, as these existing measures can be converted to travel time. Spillback from a downstream facility onto a signalized IRT. The connection between freeways and urban streets is typically at an IRT. The HCM does not currently address spillback from freeways into IRTs, however, the IRT methodology does address spillback from one signalized intersection to the other.

- 11 - The procedure includes an adjustment to consider spillback from the downstream intersection to the upstream in the form of additional lost time. This lost time is estimated for each upstream movement as a function of the downstream queue length and storage availability. Similarly, the HCM Urban Streets methodology provides an adjustment to the saturation flow rate to account for spillback effects. It is not clear whether one approach may be preferable than the other. However, we recommend that the HCM provides a consistent approach for addressing spillback onto signalized intersections and interchanges, whether the queue is originating from a downstream intersection or from a freeway on-ramp. Spillback from a downstream facility onto a roundabout IRT. Roundabouts are especially sensitive to queue spillback, since it can result in complete gridlock for all movements. The current roundabouts procedure does not evaluate roundabouts considering spillback. Evaluation of actuated control for an intersection/interchange with spillback. The current signalized methodology analysis framework considers both pre-timed and actuated control. In the case of pre-timed control, signal control is an input, and the effective green can be adjusted to account for spillback. However, in the case of actuated control, the signal phase duration is variable. This creates the following issues when spillback occurs: a) the phase duration estimation would be impacted and thus the methodology needs to be adjusted accordingly; and b) actuated control results in variable effective greens and thus variable arrivals to the downstream ramp. These variable arrivals would result in variable queues at the ramp, which will in turn affect the phase duration at the signal. This creates an iterative process which an HCM-type analysis cannot address. This research project uses suitable assumptions and simplifications to provide a reasonable framework for systems analysis, and uses equivalent pre-timed control to estimate phase durations. Queue length at the ramp receiving traffic from an intersection/interchange. Any type of queueing analysis depends on the arrival patterns and the service patterns. In the case of a ramp junction receiving traffic from an upstream intersection, the queue is calculated as a function of: a) the arrival patterns upstream of the ramp - these are a function of the type of control, as well as the arrivals from each incoming traffic stream; b) the departure rate at the ramp into the mainline - this is equal to the arrivals when conditions are undersaturated. However, for oversaturated conditions, it is not clear what the discharge rate is, and how it varies within an analysis period.