National Academies Press: OpenBook
« Previous: 4. Performance Measures
Page 20
Suggested Citation:"5. Queue Spillback into Freeways." National Academies of Sciences, Engineering, and Medicine. 2020. Highway Capacity Manual Methodologies for Corridors Involving Freeways and Surface Streets. Washington, DC: The National Academies Press. doi: 10.17226/25963.
×
Page 20
Page 21
Suggested Citation:"5. Queue Spillback into Freeways." National Academies of Sciences, Engineering, and Medicine. 2020. Highway Capacity Manual Methodologies for Corridors Involving Freeways and Surface Streets. Washington, DC: The National Academies Press. doi: 10.17226/25963.
×
Page 21
Page 22
Suggested Citation:"5. Queue Spillback into Freeways." National Academies of Sciences, Engineering, and Medicine. 2020. Highway Capacity Manual Methodologies for Corridors Involving Freeways and Surface Streets. Washington, DC: The National Academies Press. doi: 10.17226/25963.
×
Page 22
Page 23
Suggested Citation:"5. Queue Spillback into Freeways." National Academies of Sciences, Engineering, and Medicine. 2020. Highway Capacity Manual Methodologies for Corridors Involving Freeways and Surface Streets. Washington, DC: The National Academies Press. doi: 10.17226/25963.
×
Page 23
Page 24
Suggested Citation:"5. Queue Spillback into Freeways." National Academies of Sciences, Engineering, and Medicine. 2020. Highway Capacity Manual Methodologies for Corridors Involving Freeways and Surface Streets. Washington, DC: The National Academies Press. doi: 10.17226/25963.
×
Page 24
Page 25
Suggested Citation:"5. Queue Spillback into Freeways." National Academies of Sciences, Engineering, and Medicine. 2020. Highway Capacity Manual Methodologies for Corridors Involving Freeways and Surface Streets. Washington, DC: The National Academies Press. doi: 10.17226/25963.
×
Page 25
Page 26
Suggested Citation:"5. Queue Spillback into Freeways." National Academies of Sciences, Engineering, and Medicine. 2020. Highway Capacity Manual Methodologies for Corridors Involving Freeways and Surface Streets. Washington, DC: The National Academies Press. doi: 10.17226/25963.
×
Page 26
Page 27
Suggested Citation:"5. Queue Spillback into Freeways." National Academies of Sciences, Engineering, and Medicine. 2020. Highway Capacity Manual Methodologies for Corridors Involving Freeways and Surface Streets. Washington, DC: The National Academies Press. doi: 10.17226/25963.
×
Page 27
Page 28
Suggested Citation:"5. Queue Spillback into Freeways." National Academies of Sciences, Engineering, and Medicine. 2020. Highway Capacity Manual Methodologies for Corridors Involving Freeways and Surface Streets. Washington, DC: The National Academies Press. doi: 10.17226/25963.
×
Page 28
Page 29
Suggested Citation:"5. Queue Spillback into Freeways." National Academies of Sciences, Engineering, and Medicine. 2020. Highway Capacity Manual Methodologies for Corridors Involving Freeways and Surface Streets. Washington, DC: The National Academies Press. doi: 10.17226/25963.
×
Page 29
Page 30
Suggested Citation:"5. Queue Spillback into Freeways." National Academies of Sciences, Engineering, and Medicine. 2020. Highway Capacity Manual Methodologies for Corridors Involving Freeways and Surface Streets. Washington, DC: The National Academies Press. doi: 10.17226/25963.
×
Page 30
Page 31
Suggested Citation:"5. Queue Spillback into Freeways." National Academies of Sciences, Engineering, and Medicine. 2020. Highway Capacity Manual Methodologies for Corridors Involving Freeways and Surface Streets. Washington, DC: The National Academies Press. doi: 10.17226/25963.
×
Page 31
Page 32
Suggested Citation:"5. Queue Spillback into Freeways." National Academies of Sciences, Engineering, and Medicine. 2020. Highway Capacity Manual Methodologies for Corridors Involving Freeways and Surface Streets. Washington, DC: The National Academies Press. doi: 10.17226/25963.
×
Page 32

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.

- 14 - 5. Queue Spillback into Freeways Queue spillback into freeway facilities occurs due to insufficient capacity in at least one element of the off-ramp, namely the ramp proper or the downstream ramp terminal. The impact of the spillback on the freeway mainline can be restricted to the exit area or can extend for miles, depending on a series of factors. This section summarizes the data collection and analysis process for developing the methodology to evaluate queue spillback into freeways. Field data and simulation were used to develop the procedure, which is presented in Appendix B - Off-ramp Queue Spillback Check and Appendix C - Off-ramp – Queue Spillback Analysis. Data Collection Data collection (Task 6) is a critical part in the development of analytical procedures to evaluate the interactions between freeway and urban street facilities. The research team used a combination of field data and simulation modeling in order to comprehensively evaluate a variety of designs and traffic conditions that can be found throughout the U.S. The use of simulated data to complement field data was made necessary for two important reasons: 1. The number of feasible traffic, control, and highway design characteristics available is extremely high, making a pure field data collection approach unrealistic. 2. Simulation allows us to isolate the effects of a specific variable (for example, the length of the acceleration lane) on traffic operations. Field data are subject to day-to-day noise in traffic attributes (because of fluctuations in demand, weather, incidents, special events, etc.) that make it difficult to isolate the effects of a single variable. Even though microsimulation cannot and should not replace field data collection, it can very effectively supplement field data. Properly calibrated simulation models can replicate driver behavior and aggressiveness based on simple observations of gap acceptance, headway distributions, number of lane changes per mile, etc. This section describes the data collection for each of the three major components, including: data requirements, dataset description, data reduction efforts, agencies contacted and application of microsimulation. Data Requirements The process of identifying potential study locations started by identifying sites that regularly experience spillback into freeways, and have instrumentation and data available that can be used by the research team. Another key requirement for suitable study locations was that the off-ramp is the primary and only freeway bottleneck to allow the observation of queue spillback effects in isolation. Any locations experiencing overlapping bottlenecks at the diverge region were discarded from the process. The data required for studying spillback into freeways is shown in Figure 4, and each of the components is briefly discussed next.

- 15 - Figure 4 – Data collection framework for queue spillback into freeway facilities 1. Video Cameras: video observations are the most critical data for the study of queue spillback, as they provide information that cannot be accurately captured by loop detectors. Video recordings can provide views of the off-ramp area so that the queues forming longitudinally and laterally due to spillback along the freeway can be observed and measured. It also provides important insight on driver behavior, lane changing activity, and forced merging. 2. Upstream Detectors: detectors are a critical source for speed and flow data. It is essential that detectors are able to provide data on individual lanes and in raw format (minimal aggregation). The available sources used for this project provide raw data aggregated in intervals between 20s to 60s. Locations with multiple detectors upstream of the off-ramp were also preferred, as they provide more comprehensive data to evaluate the changes in performance measures along the freeway longitudinally. 3. Downstream Detectors: downstream detectors are not strictly required for the analysis, but are useful to ensure that conditions downstream are undersaturated. Any sites where downstream detectors showed oversaturated conditions were discarded. 4. Ramp Detectors: ramp detectors along the off-ramp can provide data to estimate the off-ramp demand and capacity. 5. Geometry: geometric features such as number of lanes in the freeway and ramp, deceleration lane length, lane width and others can be obtained using satellite tools such as Google Maps. Agencies Contacted The strict requirements for data collection to observe off-ramp queue spillback, as previously mentioned, created a significant challenge. The research team contacted state Department of Transportation (DOT) staff in each of the 50 states, obtaining a positive response from 21 states (Figure 5). From this sample, we were able to identify study sites from five states that met all the required criteria: California, Georgia, Virginia, Florida and Minnesota. The remaining 16 states that responded were able to identify locations with queue spillback, but they did not meet the full data requirements previously stated, and therefore these locations could not be used.

- 16 - Figure 5 – Survey for off-ramp queue spillback study sites – outcome by state Dataset Description Several potential locations were selected for data collection, and only those that met all the criteria previously described were kept in the final dataset. For the selected locations, detector data were extracted from online repositories for each State DOT. Video recordings were extracted from online repositories or recordings provided by DOT staff. Table 1 summarizes the data sources by state. Table 1 – Data sources by State DOT – off-ramp queue spillback State Detector Data Source Video Recordings Source California Caltrans PeMS (1) Caltrans CCTV Map (2) Florida RITIS (3) Recording provided by agency staff Virginia VDOT PeMS (4) RITIS Georgia GDOT ATMS (5) GDOT ATMS Minnesota MN Data Tools (6) Recording provided by agency staff (1)http://pems.dot.ca.gov (2) cwwp2.dot.ca.gov (3)http://ritis.org (4)https://vdot.iteris-pems.com (5) navigator-atms.dot.ga.gov (6) http://data.dot.state.mn.us/datatools Table 2 summarizes the final list of ten study locations, with a description of their key characteristics and number of video observations. A video observation consists of one peak period recording, where the development and discharge of queues can be observed, which typically lasts between 2-4 hours.

- 17 - Table 2 – Summary of study locations – data collection – off-ramp queue spillback LOCATION D ir ec tio na l la ne s Se gm en t T yp e D ow ns tr ea m R am p Te rm in al O ff- R am p La ne s V id eo O bs er va tio ns Pe ak P er io d Sp ill ba ck R eg im e Miami/FL - I-95 SW 25th Rd. 3 Diverge Signalized Intersection 2 13 PM 4 Tampa/FL - I-275 NB to W Kennedy Blvd 4 Diverge Freeway Merge 2 6 PM 4 Norfolk/VA - I-64 WB to Northampton Blvd 4 Diverge Signalized Intersection 2 10 PM 3 Centreville/VA - I-66 WB to SR-28 (Sully Rd) 4 Diverge Signalized Intersection 1 11 AM 3 Centreville/VA - I-66 EB to SR-28 (Sully Rd) 4 Diverge Signalized Intersection 1 11 AM 3 Minneapolis/MN - I-35W SB to 35th St. 4 Weaving Signalized Intersection 1 5 PM 3 Atlanta/GA - I-285 NB to I-20 5 Diverge Freeway Merge 1 7 PM 3 Miami/FL - I-75 SB to SR 826 5 Diverge Freeway Merge 2 13 AM 4 Atlanta/GA - I-285 NB to GA-141 5 Diverge Freeway Merge 1 12 PM 3 Sacramento/CA - SR-99 NB to SR-50 5 Weaving Freeway Merge 2 5 AM 4 TOTAL 93 Schematics and sample screenshots of the study locations are presented in Figure 6 through Figure 15. Source: Photo provided by the Florida Department of Transportation Figure 6 – Miami, FL – I-95 SB to SW 25th Rd. Source: Photo provided by the Florida Department of Transportation Figure 7 – Tampa, FL – I-275 NB to SR-60

- 18 - Source: Photo provided by CATT Lab (Ritis.org) Figure 8 – Norfolk, VA – I-64 WB to Northampton Blvd Source: Photo provided by CATT Lab (Ritis.org) Figure 9 – Centreville, VA – I-66 WB to SR-28 (Sully Rd) Source: Photo provided by CATT Lab (Ritis.org) Figure 10 – Centreville, VA – I-66 EB to SR-28 (Sully Rd) Source: Photo provided by the Minnesota Department of Transportation Figure 11 – Minneapolis, MN – I-35W SB to 35th St.

- 19 - Source: Photo provided by the Georgia Department of Transportation Figure 12 – Atlanta, GA – I-285 NB to I-20 Source: Photo provided by the Florida Department of Transportation Figure 13 – Miami, FL – I-75 SB to SR 826 Source: Photo provided by the Georgia Department of Transportation Figure 14 – Atlanta, GA – I-285 NB to GA-141 Source: Photo provided by the California Department of Transportation Figure 15 – Sacramento, CA – SR-99 NB to SR-50

- 20 - Data Reduction Video observations were the most challenging part of the data reduction, as the process of obtaining the recordings was very time-consuming. After the recordings were obtained from the respective agencies, the research team performed a first screening at the videos for an initial validation, which included the following conditions: • Camera is able to show the queues in the diverge area as required for project purposes; • No occurrence of incidents, disabled vehicles, lane closures or other factors that may affect the daily dynamics of vehicular traffic; • No overlapping bottlenecks – the off-ramp is the only source of congestion. After this first screening, the next step was downloading detector raw data (speed and flow) from online repositories for the periods captured in the video files. Then, any observations with the following issues were also excluded from the dataset: • Poor detector data health, as evaluated by the data repositories; • Missing or null readings for long periods, making the data analysis unfeasible. The valid observations remaining after the initial screenings make up the study dataset, previously presented in Table 2. The next step in the data reduction was to measure the queue lengths from video observations. Queue lengths were measured visually from video starting from the exit point by lane, as illustrated in Figure 16. Vehicles queued on the adjacent lane were counted as part of the deceleration lane queue, even though they are in a mainline lane. We chose this approach because the downstream end of this queue is not always at the same location. Also, this approach makes the L2 queue measurements consistent with the L1 queue, which starts from the end of the deceleration. Measurements were taken in time intervals consistent with the detector aggregation times so queue lengths and flow/speed data can be matched for further analyses. For example, if raw detector data from a given location are aggregated in 20- second intervals, then queue lengths were measured every 20 seconds. Figure 16 – Back-of-queue length measurement – off-ramp queue spillback After the data reduction, queue lengths were matched with detector data to provide insights on the effects of queue spillback in the freeway performance. Figure 17 provides a sample of data illustrating queue spillback, where (a) shows the development of queues over time, while (b) shows the speed drops consistent with the development of mainline queues. At this site, queues along lane 2 were longer than those in lane 1 due to demands at the downstream intersection. Speeds on lane 1 (S1) are lower than lane 4 (S4), consistent with the occurrence of queue spillback on the right side of the freeway.

- 21 - Figure 17 – Sample of data illustrating queue spillback: (a) queue length and (b) speeds by lane Qualitative Observations The videos obtained were used to observe operations during off-ramp queue spillback, in order to inform the development of analytical models. These observations are discussed in the following paragraphs. Unbalanced lane usage at off-ramp For off-ramps with two lanes, it was observed that queues may not distribute evenly along the two lanes. Drivers typically choose a position at the off-ramp based on their next movement at the downstream ramp terminal. Therefore, if the downstream ramp terminal has one particular movement with excessive demand- to-capacity ratio, queues are likely to develop along the lane connected to that specific movement, while the other off-ramp lane remains underutilized. Figure 18 shows examples of unbalanced demand at the off- ramp on (a) the right lane and (b) the left lane of a two-lane off-ramp.

- 22 - Source: Photos provided by (a)CATT Lab (Ritis.org) and (b)Florida Department of Transportation Figure 18 – Unbalanced lane usage at the off-ramp: (a) right lane and (b) left lane Increasing operational impacts across lanes when queues are longer When multiple detector stations are present upstream of the off-ramp, it is possible to compare the effects of queue spillback along multiple locations during the exact same time. Field data show that in the area close to the ramp exit, the effects of queue spillback are restricted to the blocked lanes, while the through vehicles in the leftmost lanes are not affected. However, for locations further upstream, speed drops are more homogenous and evenly spread across all lanes of the freeway. Figure 19 illustrates the impacts of queue spillback at three locations upstream of an off-ramp bottleneck: (a) at the ramp influence area, (b) 1,800 ft. upstream of the exit and (c) 5,500 ft. upstream of the exit. Source: Photos provided by the Florida Department of Transportation Figure 19 – Effects of queue spillback across freeway lanes at three locations upstream of the exit experiencing spillback (I-275 EB – Tampa, FL) Number of blocked lanes is inherent to each location experiencing spillback When there is queue spillback from an off-ramp, queues may block one or two lanes in the freeway mainline (Figure 20). Video observations of queues show that at each site experiencing recurrent queue spillback that extends beyond the deceleration lane, the number of blocked lanes does not change. In other words, queue length does not affect whether the site experiences the conditions of Figure 20(a) or (b). The condition shown in Figure 20(b) occurs more frequently in locations with more aggressive driver behavior.

- 23 - When the segment is a lane-drop (rather than a diverge) the exiting traffic can access the off-ramp with a single lane change. Therefore, drivers are more likely to wait until they are closer to the exit to change lanes, blocking the adjacent through lane. Figure 20 – Blockage of one (a) or two (b) mainline lanes due to off-ramp queue spillback Simulated Sites Given the complexity of obtaining field data to evaluate queue spillback, simulation was used to complement the data collection. The selection of locations for simulation took into consideration the use of pre-calibrated locations that were available to the research team. The simulated sites were used with increasing input demands so that queue spillback into freeways would occur. Table 3 presents the list of simulated locations for this study. Table 3 – Summary of simulated locations – off-ramp queue spillback Location Number of freeway lanes Segment type Downstream ramp terminal Number of off-ramp lanes I-105 SB to Bellflower Blvd. 3 Diverge Signalized int. 1 I-105 NB to Bellflower Blvd. 4 Diverge Signalized int. 1 I-710 SB to Martin Luther King Blvd. 4 Diverge Signalized int. 2 I-105 WB to Garfield Ave. 4 Major Diverge Signalized int. 2 I-710 SB to I-105 5 Diverge Freeway merge 2 I-710 SB to Martin Luther King Blvd. 5 Weaving Signalized int. 2 Simulation models developed and calibrated with field data in AIMSUN were used to obtain simulated data. The following assumptions were made for the simulated data: Queue measurement: Queued vehicles in a freeway off-ramp are not completely stationary. During queue spillback, exiting vehicles are typically moving at slow speeds. Additionally, vehicles are closely spaced but not as close as vehicles stopped in an intersection. Queue lengths were measured visually from video observations. However, quantitative criteria must be set to define freeway queue lengths in microsimulation. Vehicles were considered in a queue if they meet the following criteria: • Vehicle speeds are not greater than 5 mi/h • Distance between vehicles is shorter than 40 ft. Detectors: The greatest benefit of microsimulation is the ability to obtain a variety of performance measures at any desired location of the simulated network. For field data, speed and flow readings were

- 24 - available according to the position of detectors and were different for each study location. For the simulated sites, detector stations were placed at three locations upstream of the exit: 1,500 ft, 4,000 ft and 8,000 ft. Speeds and flows on a lane-by-lane basis were extracted in 20-second intervals, similarly to the lowest resolution obtained from field data. Forced merging: As queues extend further upstream of the ramp exit, drivers wishing to exit are more likely to change lanes aggressively to join the back of the queue. Some vehicles may fail to join the back of the queue and then attempt to find a gap to merge into the queue (forced merging). A microsimulation model where drivers can perfectly anticipate a lane blockage ahead and always join the back of queue would fail to replicate the field conditions of an off-ramp queue spillback. Therefore, for the simulation models, a vehicle with an O-D that includes the off-ramp as a destination adjusts its position stochastically, which allows a possibility that it may miss the back of the queue. If that happens, it attempts to find gaps in the queue to perform a lane change, as illustrated in Figure 21. Figure 21 – Example of a forced merging maneuver for a vehicle attempting to join the off-ramp queue Figure 22 provides an example of a simulated location where queue spillback occurs (I-105@Bellflower Interchange). As shown, at the vicinity of the off-ramp the blockage occurs in the rightmost lane, with negligible effects on the adjacent lanes. At sections further upstream, additional turbulence starts affecting the performance of the other lanes in the freeway mainline. As shown, the simulation replicates the field observations as expected. Figure 22 – Micro-simulated off-ramp queue spillback (I-105@Bellflower Interchange – Los Angeles, CA)

- 25 - Queue Spillback Check The first step to evaluate queue spillback into freeways is to assess whether it is expected to occur. The detailed methodology for off-ramp spillback check is provided in Appendix B - Off-ramp Queue Spillback Check. The methodology evaluates two potential capacity bottlenecks in the off-ramp (Figure 23): • Ramp proper: If the off-ramp demand is greater than the capacity of the ramp proper, spillback is expected to occur. • Downstream ramp terminal (intersection or a merge segment, in the case of a freeway-to-freeway connector): if there is insufficient capacity at the ramp terminal, queues will develop along the off- ramp. The procedure then compares the expected queue length and the available queue storage. If the expected queue length is greater than the queue storage, spillback is expected to occur. Figure 23 – Procedure for identifying spillback occurrence at an off-ramp/weaving segment If queue spillback is not expected to occur, no adjustments are required for the current HCM methods. If queue spillback is expected to occur, the procedure for evaluating queue spillback must be applied.

- 26 - Evaluation of Queue Spillback Impacts If queue spillback is expected to occur, the analyst must refer to the procedure for evaluation of impacts of queue spillback, described under Appendix C - Off-Ramp Queue Spillback Analysis. The procedure takes an approach similar to the Oversaturated Segment Evaluation (HCM Chapter 25), where the freeway facility is analyzed in 15-second time steps instead of 15-minute time periods. The facility structure also changes from segments to link-nodes for the analysis, and the proposed methodology expands the structure to include the off-ramp (Figure 24). Figure 24 – Expanded link-node structure to evaluate the off-ramp segment

Next: 6. Queue Spillback into Urban Streets »
Highway Capacity Manual Methodologies for Corridors Involving Freeways and Surface Streets Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

The procedures detailed in the 6th Edition of the Highway Capacity Manual (HCM) estimate capacity and several operational measures, including those determining Level of Service, for freeway facilities as well as surface streets.

The TRB National Cooperative Highway Research Program's NCHRP Web-Only Document 290: Highway Capacity Manual Methodologies for Corridors Involving Freeways and Surface Streets introduces materials to help modify the freeway analysis methods and the urban street methods so that the effects of operations from one facility to the other can be evaluated.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!