National Academies Press: OpenBook

Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks (2017)

Chapter: Chapter 6 - Classifying and Evaluating Truck Freight Bottlenecks

« Previous: Chapter 5 - Identify and Quantify Truck Freight Bottlenecks
Page 57
Suggested Citation:"Chapter 6 - Classifying and Evaluating Truck Freight Bottlenecks." National Academies of Sciences, Engineering, and Medicine. 2017. Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks. Washington, DC: The National Academies Press. doi: 10.17226/24807.
×
Page 57
Page 58
Suggested Citation:"Chapter 6 - Classifying and Evaluating Truck Freight Bottlenecks." National Academies of Sciences, Engineering, and Medicine. 2017. Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks. Washington, DC: The National Academies Press. doi: 10.17226/24807.
×
Page 58
Page 59
Suggested Citation:"Chapter 6 - Classifying and Evaluating Truck Freight Bottlenecks." National Academies of Sciences, Engineering, and Medicine. 2017. Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks. Washington, DC: The National Academies Press. doi: 10.17226/24807.
×
Page 59
Page 60
Suggested Citation:"Chapter 6 - Classifying and Evaluating Truck Freight Bottlenecks." National Academies of Sciences, Engineering, and Medicine. 2017. Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks. Washington, DC: The National Academies Press. doi: 10.17226/24807.
×
Page 60
Page 61
Suggested Citation:"Chapter 6 - Classifying and Evaluating Truck Freight Bottlenecks." National Academies of Sciences, Engineering, and Medicine. 2017. Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks. Washington, DC: The National Academies Press. doi: 10.17226/24807.
×
Page 61
Page 62
Suggested Citation:"Chapter 6 - Classifying and Evaluating Truck Freight Bottlenecks." National Academies of Sciences, Engineering, and Medicine. 2017. Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks. Washington, DC: The National Academies Press. doi: 10.17226/24807.
×
Page 62
Page 63
Suggested Citation:"Chapter 6 - Classifying and Evaluating Truck Freight Bottlenecks." National Academies of Sciences, Engineering, and Medicine. 2017. Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks. Washington, DC: The National Academies Press. doi: 10.17226/24807.
×
Page 63
Page 64
Suggested Citation:"Chapter 6 - Classifying and Evaluating Truck Freight Bottlenecks." National Academies of Sciences, Engineering, and Medicine. 2017. Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks. Washington, DC: The National Academies Press. doi: 10.17226/24807.
×
Page 64
Page 65
Suggested Citation:"Chapter 6 - Classifying and Evaluating Truck Freight Bottlenecks." National Academies of Sciences, Engineering, and Medicine. 2017. Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks. Washington, DC: The National Academies Press. doi: 10.17226/24807.
×
Page 65
Page 66
Suggested Citation:"Chapter 6 - Classifying and Evaluating Truck Freight Bottlenecks." National Academies of Sciences, Engineering, and Medicine. 2017. Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks. Washington, DC: The National Academies Press. doi: 10.17226/24807.
×
Page 66
Page 67
Suggested Citation:"Chapter 6 - Classifying and Evaluating Truck Freight Bottlenecks." National Academies of Sciences, Engineering, and Medicine. 2017. Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks. Washington, DC: The National Academies Press. doi: 10.17226/24807.
×
Page 67
Page 68
Suggested Citation:"Chapter 6 - Classifying and Evaluating Truck Freight Bottlenecks." National Academies of Sciences, Engineering, and Medicine. 2017. Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks. Washington, DC: The National Academies Press. doi: 10.17226/24807.
×
Page 68
Page 69
Suggested Citation:"Chapter 6 - Classifying and Evaluating Truck Freight Bottlenecks." National Academies of Sciences, Engineering, and Medicine. 2017. Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks. Washington, DC: The National Academies Press. doi: 10.17226/24807.
×
Page 69
Page 70
Suggested Citation:"Chapter 6 - Classifying and Evaluating Truck Freight Bottlenecks." National Academies of Sciences, Engineering, and Medicine. 2017. Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks. Washington, DC: The National Academies Press. doi: 10.17226/24807.
×
Page 70

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.

57 C h a p t e r 6 6.1 Overview of Potential Causes of Truck Bottlenecks This chapter describes the process of identifying the causes of the bottlenecks that were identi- fied using the methodology described in Chapter 5. These causes can include recurring conges- tion, weather, crashes, construction, and a wide variety of other causative factors. In most cases, these causes can be identified based on a quantitative analysis conducted at a desktop using avail- able data. In other cases, this needs to be combined with field analysis to refine the understand- ing of the bottleneck. Similarly, a combined desktop and field analysis can be used to rank truck bottlenecks. Travel speed-based delay for all vehicles has been studied extensively by several research proj- ects. Figure 6-1 shows a distribution of the causes of travel speed-based delay for all vehicles on all types of roadways from previous research conducted by FHWA. Recurring congestion, traffic incidents, and weather were found to be responsible for 90 percent of all vehicle delay. Due to definitional differences, for this previous research the causal category “recurring congestion” was referred to simply as “bottlenecks.” The increased use of vehicle probe data has made the calculation of truck-specific travel speed-based delay more accurate and similar distributions of delay can now be developed for truck activity. This chapter is structured to examine this through the following sections: • Section 6.1. Overview of Potential Causes of Truck Bottlenecks, • Section 6.2. Identify Causes of Travel Speed-Based Truck Bottlenecks, • Section 6.3. Ranking Travel Speed-Based Bottlenecks, • Section 6.4. Identify Causes and Rank Process-Based Truck Bottlenecks, and • Section 6.5. Conduct Field Analysis to Refine Bottleneck Understanding. 6.2 Identify Causes of Travel Speed-Based Truck Bottlenecks Identifying the potential causes of truck bottlenecks is a process of overlaying the timing of bottlenecks with the timing of other activities that have the potential to cause the bottleneck. For example, if a bottleneck is identified between 11:00 a.m. and 11:30 a.m. on a Monday morning at a specific location, then information on truck and auto volumes, crashes, weather, and construc- tion should be examined during the same time period to determine which of these factors had the potential to have contributed to the bottleneck. Additional factors should also be considered depending on the specific type of location where the bottleneck occurred. For example, for locations near port terminals, additional factors can Classifying and Evaluating Truck Freight Bottlenecks

58 Guide for Identifying, Classifying, evaluating, and Mitigating truck Freight Bottlenecks include operating hours of gates. For locations on arterials, turning movement counts at inter- sections may need to be examined. Another consideration is often the need to maintain a corridor approach in identifying causes of bottlenecks. In some cases, relieving a bottleneck at one location shifts the bottleneck to a downstream location without providing broader system benefits. This is particularly possible when considering efforts to alleviate bottlenecks based on congestion. Alternatively, bottleneck relief may also result in higher speeds which exacerbate road geometry or safety issues along a corridor. To estimate systemwide impacts of bottleneck mitigation efforts, typically a travel demand model is needed. Additionally, outreach to roadway users can be used to determine how relieving specific point bottlenecks will impact other elements of the transportation network. 6.2.1 Example of Analysis to Identify Potential Causes This subsection provides a simplified example of how to calculate the causes of truck bottle- necks. Specifically, the example highlights how to determine the amount of truck delay caused by a vehicle crash. The example is conducted using three hypothetical segments (Segment 1, 2, and 3) over a 1-hour period that is divided into 6 10-minute time intervals. The three segments are con- tinuous segments along a single route in a single direction such that Segment 3 follows Segment 2 and such that Segment 2 follows Segment 1. All segments are assumed to be 1-mile long. Table 6-1 shows truck speeds in miles per hour by time interval for each of the three road segments and six time periods. The reference speed for each of the segments is assumed to be 60 mph. Any time intervals showing speeds that are recorded below 60 mph are assumed to be congested. The congested time intervals are highlighted in yellow for each time segment. Table 6-2 shows truck volumes for those same road segments and time periods. These vol- umes are typically available as estimates through state DOT vehicle classification counting pro- grams. Truck volumes can also be developed through special counts collected specifically for the purposes of bottleneck analysis. Source: FHWA Office of Operations, Traffic Congestion and Reliability: Trends and Advanced Strategies for Congestion Mitigation, December 2013. Non-Recurring (e.g., Special Events) 5% Bottlenecks 40% Traffic Incidents 25% Work Zones 10% Weather (Snow, Ice, Fog) 15% Poor Signal Timing 5% Recurring Congestion Figure 6-1. Causes of travel speed based bottlenecks for all vehicles.

Classifying and evaluating truck Freight Bottlenecks 59 Table 6-3 shows the calculation of delay along each of the three segments for the six time inter- vals. This is calculated by using the difference between the time taken to travel the segment using the reference speed and the time taken to travel the segment during the actual 1-hour period. This is calculated separately for each 10-minute interval. The travel times are calculated as the distance divided by the travel speed. Table 6-4 shows the timing of the crash that occurs on the roadway. Specifically, it shows that a crash occurred on Segment 2 between 11:10 a.m. and 11:20 a.m. This crash blocked a lane of traffic which was cleared between 11:30 a.m. and 11:40 a.m. with traffic returning to normal speeds by 11:50 a.m. Information on crashes is available in state crash databases. Information on incident clearance times is sometimes maintained by state DOTs. However, this data is stored at different levels of detail in different organizations. In some instances, it may need to be estimated based on the time it takes for speeds to return to the reference speed or clearance time of other similar incidents. This is discussed in greater detail in Chapter 3. Table 6-5 illustrates which of the delays shown in Table 6-3 has been “influenced” by the known crash. Not all of the delay that was calculated can be attributed to the crash. In particular, delay that occurred on Segment 1 which is upstream from the crash cannot be attributed to the Time Intervals Segment 1 Speeds (mph) Segment 2 Speeds (mph) Segment 3 Speeds (mph) 11:00 a.m. – 11:10 a.m. 60 60 40 11:10 a.m. – 11:20 a.m. 60 40 60 11:20 a.m. – 11:30 a.m. 60 20 60 11:30 a.m. – 11:40 a.m. 60 20 60 11:40 a.m. – 11:50 a.m. 60 40 40 11:50 a.m. – 12:00 p.m. 45 60 30 Table 6-1. Truck speeds on three road segments by time interval. Time Intervals Segment 1 Truck Volumes Segment 2 Truck Volumes Segment 3 Truck Volumes 11:00 a.m. – 11:10 a.m. 100 90 85 11:10 a.m. – 11:20 a.m. 110 100 95 11:20 a.m. – 11:30 a.m. 130 120 115 11:30 a.m. – 11:40 a.m. 125 105 95 11:40 a.m. – 11:50 a.m. 110 105 95 11:50 a.m. – 12:00 p.m. 90 85 80 Total Truck Volume 665 605 565 Table 6-2. Truck volumes on three road segments by time interval.

Time Intervals Segment 1 Delay Influenced by Crash Segment 2 Delay Influenced by Crash Segment 3 Delay Influenced by Crash Total Delay Influenced by Crash 11:00 a.m. – 11:10 a.m. 0.0 0.0 0.0 0.0 11:10 a.m. – 11:20 a.m. 0.0 0.8 0.0 0.8 11:20 a.m. – 11:30 a.m. 0.0 4.0 0.0 4.0 11:30 a.m. – 11:40 a.m. 0.0 3.5 0.0 3.5 11:40 a.m. – 11:50 a.m. 0.0 0.9 0.8 1.7 11:50 a.m. – 12:00 p.m. 0.0 0.0 1.3 1.3 Total Truck-Hours of Delay 0.0 9.2 2.1 11.3 Table 6-5. Truck-hours of delay “influenced” by the crash. Time Intervals Segment 1 Truck Delay Segment 2 Truck Delay Segment 3 Truck Delay Total Truck Delay 11:00 a.m. – 11:10 a.m. 0.0 0.0 0.7 0.7 11:10 a.m. – 11:20 a.m. 0.0 0.8 0.0 0.8 11:20 a.m. – 11:30 a.m. 0.0 4.0 0.0 4.0 11:30 a.m. – 11:40 a.m. 0.0 3.5 0.0 3.5 11:40 a.m. – 11:50 a.m. 0.0 0.9 0.8 1.7 11:50 a.m. – 12:00 p.m. 0.5 0.0 1.3 1.8 Total Truck-Hours of Delay 0.5 9.2 2.8 12.5 Table 6-3. Calculation of truck delay hours on three road segments by time interval. Time Intervals Segment 1 Crashes Segment 2 Crashes Segment 3 Crashes 11:00 a.m. – 11:10 a.m. – – – 11:10 a.m. – 11:20 a.m. – Crash Occurs – 11:20 a.m. – 11:30 a.m. – Lane Blocked – 11:30 a.m. – 11:40 a.m. – Crash Cleared – 11:40 a.m. – 11:50 a.m. – Scene Clear – 11:50 a.m. – 12:00 p.m. – – – Table 6-4. Timing of vehicle crash and incident clearance.

Classifying and evaluating truck Freight Bottlenecks 61 crash. Additionally, delay that occurs in time periods before the crash occurred cannot be attrib- uted to the crash. The total delay attributable to the crash is 11.3 truck-hours of delay which is lower than the total 12.5 hours of delay that was calculated in Table 6-3. The delay statistics computed in Table 6-5 can then be aggregated to estimate total delay in each segment, or total delay in specific time periods, or delay in some combination of segments (e.g., a defined urban corridor or urban area) for defined time periods (e.g., the a.m. peak period). The summary values shown in Table 6-5 also can be aggregated on the basis of whether spe- cific causation variables were present. For example, in Table 6-5, of 11.3 observed vehicle-hours of delay, 8.3 hours occurred when a crash was present in Segment 2 (Time Periods 11:10–11:20, 11:20–11;30, and 11:30–11:40. Consequently, just over 66 percent of the delay occurred when a crash was present. This does not mean that crashes “caused” 66 percent of all delay in this example, but it does suggest that crashes might be a significant contributor to freight delays observed at this location. Additional desktop analysis can be done to explore these relationships further. For example, data for these segments on other days at these same times could be analyzed to compare the amount of delay normally present without a crash. The number and duration of crashes occur- ring along this stretch of roadway also could be computed and reviewed. As mentioned earlier, more than one variable is often present when congestion occurs. For example, Table 6-6 shows when heavy rain was influencing the congestion measured in Table 6-2. Some of that rain occurred at the same time that a crash was present (Time Periods 11:20–11:30 and 11:30–11:40). Table 6-7 updates the “influence” characterization. Time periods and segments influenced only by rain are colored light blue. Time periods influenced only by the crash are shaded yellow. Time periods influenced by both factors are shaded a light orange. If the truck-hours of delay within each of these categories is aggregated and any delay associ- ated with a specific influencing factor is assigned to that factor, then the total delay is computed as follows: • Crash – 8.3 truck-hours (0.8 + 4.0 + 3.5) (influences up to 66.4 percent of all delay); • Rain – 10.5 truck-hours (4.0 + 3.5 + 0.9 + 0.8 + 1.3) (influences up to 83.7 percent of all delay); • No Cause – 12.1 truck-hours (5.0 + 7.1) (9.6 percent of all delay has no “other cause” identi- fied except volume); and • Total Delay – 12.5 truck-hours. If the individual delays associated with each factor are simply added, the total will exceed the actual total delay (20.0 truck-hours versus 12.5 truck-hours). However, the relative size of the Time Intervals Segment 1 Weather Segment 2 Weather Segment 3 Weather 11:00 a.m. – 11:10 a.m. Sunny Sunny Sunny 11:10 a.m. – 11:20 a.m. Sunny Sunny Sunny 11:20 a.m. – 11:30 a.m. Sunny Rain Rain 11:30 a.m. – 11:40 a.m. Sunny Rain Rain 11:40 a.m. – 11:50 a.m. Sunny Rain Rain 11:50 a.m. – 12:00 p.m. Sunny Rain Rain Table 6-6. Timing of weather incidents.

62 Guide for Identifying, Classifying, evaluating, and Mitigating truck Freight Bottlenecks delay numbers provides good insight into the types of conditions that are present when delay forms. “Shared” delay also can be evenly (or otherwise analytically) divided between delay influ- encing factors to provide insight into the relative significance of different congestion influencing factors. For example, if all “shared” delay is evenly divided between influencing factors, then delay is computed as follows: • Crash – 4.6 truck-hours (36.5 percent of delay here when “shared” delay is evenly divided, but it could influence up to 66.4 percent of all delay); • Rain – 6.8 truck-hours (53.8 percent of delay here when “shared” delay is evenly divided, but it could influence up to 83.7 percent of all delay); • No Cause Identified – 1.2 truck-hours (9.6 percent of all delay has no “other cause” identified except volume); and • Total Delay – 12.5 truck-hours. These same data can be presented in graphic formats that are easy for decision makers to understand, such as shown in Figure 6-2. The availability of the analytical NPMRDS (or other similar data- sets), in combination with other data sources, will allow significant investigation of the factors that cause or affect the size and timing of truck bottlenecks. Roadways that are not designed to a truck’s size and performance characteristics can result in truck delays, which can be analyzed in terms of a roadway’s attributes. Many attributes can be extracted from asset and spatial data catalogs (GeoData) maintained by DOTs and MPOs. The analysis of causation in this situation is where the roadway performs adequately for cars but poorly for large vehicles. Roadway attributes identified in the catalogs can include grades, horizontal align- ment, intersection type, and other geometrics. Chapter 7 expands on this approach. Segment 1 Truck-Hours of Delay Segment 2 Truck-Hours of Delay Segment 3 Truck-Hours of Delay Total Truck- Hours of Delay 11:00 a.m. – 11:10 a.m. 0.0 0.0 0.7 0.7 11:10 a.m. – 11:20 a.m. 0.0 0.8 0.0 0.8 11:20 a.m. – 11:30 a.m. 0.0 4.0 0.0 4.0 11:30 a.m. – 11:40 a.m. 0.0 3.5 0.0 3.5 11:40 a.m. – 11:50 a.m. 0.0 0.9 0.8 1.7 11:50 a.m. – 12:00 p.m. 0.5 0.0 1.3 1.8 Total Truck-Hours of Delay 0.5 9.2 2.8 12.5 Crash Rain Crash and Rain Table 6-7. Identification of multiple causes of truck bottlenecks (truck-hours). Case Study Highlight A study was performed by the I-95 Corridor Coalition to identify, classify, and evaluate all vehicle bottlenecks to establish baseline performance measures for a corridor spanning several Northeast states. INRIX speed data and FHWA HPMS volume data were used to conduct the analysis. Regarding nonrecurrent delay, incident and work zone data were not available for the corridor, but researchers considered weather conditions for the dates with the worst congestion days at each location. They determined that weather was likely a significance factor on those days. More details are in Appendix B.

Classifying and evaluating truck Freight Bottlenecks 63 This section provided a data analysis illustration with a simplified example. Appendix D pro- vides more detailed analysis procedures for the calculation of performance measures. Namely, Section D-3 of Appendix D provides segment and route calculation procedures. 6.3 Ranking Travel Speed-Based Bottlenecks The most straightforward way to rank the causes of travel speed-based truck bottlenecks is in decreasing order of total truck delay. For the example presented in Section 6.2, a simple ranking of causes would be as shown in Table 6-8. Travel speed-based bottlenecks can also be ranked across multiple locations in a similar fashion with locations that have the most delay having the highest ranking. There are several examples of bottleneck rankings that have been developed. The American Highway Users Alliance develops the annual report, Unclogging America’s Arteries (132), which includes a list of the top 50 worst truck bottlenecks in the U.S. The ranking is based on annual truck-hours of delay at freeway segment locations. The list also includes information on average queue length, annual lost value of time, and annual fuel wasted. Table 6-9 shows the top 17 truck bottlenecks from 2015 based on this report. Another example of the ranking of bottlenecks is shown the Texas 100 Most Congested Road­ ways List (23) analysis from 2014 (Table 6-10). This table shows the 10 most congested bottlenecks 9.6% 36.5%53.8% No Cause Crash Rain Figure 6-2. Delay by congestion influencing factor. Cause of Truck Bottleneck Hours of Truck Delay Ranking Rain 6.8 1 Crash 4.6 2 No Cause Identified 1.2 3 Total Truck-Hours of Delay 12.5 N/A Table 6-8. Ranking of causes of truck bottlenecks at single location.

64 Guide for Identifying, Classifying, evaluating, and Mitigating truck Freight Bottlenecks in Texas, ranked by the total number of hours of delay occurring annually, with those statistics normalized on a per mile basis to account for the fact that each reporting segment has a different length. The “worst” bottleneck in Texas under these criteria is the Interstate 610 road segment from Interstate 10 to Interstate 59. However, if Table 6-10 is sorted in terms of annual truck- hours of delay per mile, this road segment is only the fourth worst Texas roadway segment. The worst truck delay segment is Interstate 35 from U.S. Route 290 north to State Highway 71. Similarly, if segments were ranked on the basis of annual congestion cost, the ranking would again be different. No single ranking system is appropriate for all uses. Each performance measure (e.g., truck delay, total delay, expected travel rate or reliability, or the frequency with which congestion occurs) can be used to effectively rank locations. Each of those resulting rankings will likely be different. What these different rankings indicate is that the importance of any one bottleneck changes depending on which bottleneck attributes are most important to an individual decision Table 6-9. American Highway Users Alliance ranking of top truck bottlenecks.

Classifying and evaluating truck Freight Bottlenecks 65 maker. Rankings can even be created that are based on the relative (potential) causes of those delays—e.g., where are the largest freight bottlenecks where incidents have played a role in the size of that delay? Because this is an automated process, rankings can be developed for a variety of defined sub- sets of the highway system. Arterials can be ranked differently from freeways. Rankings can be computed by geographic portion of the state. They can be computed for roads exclusively within a given MPO’s jurisdiction. They can even be computed for specific categories of road, such as for priority truck routes. The outcome of these different ranking systems is better decision support. If the state legislature is interested in having congestion relief projects in different parts of the state, then rankings can be developed for those different geographic regions. If money is set aside for arterial improvements, congestion rankings can be developed for just those eli- gible roadways. Finally, an agency may wish to remove some types of truck delay from consideration in the ranking system. For example, delays caused by bad weather might be removed from a ranking intended to identify Rank Roadway From To County Annual Hours of Truck Delay per Mile Annual Truck Congestion Cost 1 I-35 U.S. 290N SH71 Travis 108,645 $72.33 2 I-610 I-10 U.S. 59/I-59 Harris 68,893 $20.99 3 U.S. 59 I-610 SH 288 Harris 51,604 $23.64 4 I-635 I-35E/U.S. 77 U.S. 75 Dallas 49,538 $33.59 5 I-10/U.S. 90 N. Elridge Pkwy Sam Houston Tollway W Harris 48,855 $13.43 6 I-345/US75/ I-45 Woodall Rodgers Freeway U.S. 175 Dallas 46,744 $9.36 7 U.S. 59 I-10/US90 SH 288 Harris 45,469 $11.60 8 I-10/U.S. 90 I-610 I-45 Harris 44,400 $21.17 9 I-45 Sam Houston Tollway N I-610 Harris 39,713 $31.08 10 I-10/U.S. 90 Sam Houston Tollway W I-610 Harris 38,295 $21.27 Source: Texas 100 Most Congested Roadways List. Texas Department of Transportation. Available: http://www.txdot.gov/inside-txdot/projects/100-congested-roadways.html. Last Accessed: April 10, 2015. Note that the full ranking of all segments throughout Texas beyond the top 100 are available here: http://mobility.tamu.edu/most- congested-texas/. (23) Table 6-10. Texas 100 most congested roadways. . . . delays caused by bad weather might be removed from a ranking intended to identify places to spend congestion relief money, whereas those same delays might be expressly highlighted to support the implementation of better road weather management activities. . . .

66 Guide for Identifying, Classifying, evaluating, and Mitigating truck Freight Bottlenecks places to spend congestion relief money, whereas those same delays might be expressly high- lighted to support the implementation of better road weather management activities, even if those activities are not applied exclusively to those road segments. The size, scope, and ranking of bottlenecks also change depending on exactly how the road- way segment encompassing the “bottleneck” is defined. Where does the bottleneck start? And where does it end? Given detailed data, it is easy to follow the formation, growth, and eventual dissolution of a given bottleneck on any particular day. On the basis of that specific observa- tion, the analysis can determine the exact length and duration of the congestion. However, the congestion that forms today (e.g., 1.5 miles long, lasting for 90 minutes) is different from the congestion that forms tomorrow (e.g., 0.3 miles long, lasting 20 minutes) and from what forms next Friday afternoon (e.g., 12 miles long, lasting 6 hours, thanks to a crash involving a rolled truck hauling fuel). Different bottleneck definitions for a specific location will result in different analytical outcomes. For example, if the bottleneck segment described above is defined as being 12 miles long, the estimated total delay for the segment will be larger than if the bottleneck is defined as being only 0.5 or 1 miles long. But the total delay per mile computed for the longer bottleneck location will be much lower than if the bottleneck is defined as one of the shorter distances, because much of the longer segment is not as congested as the shorter road segments that are closer to where congestion typically starts—the actual “bottleneck” itself. Complicating the definition of the road segments for which bottle- necks will be computed is that numerical analyses can only be per- formed for roadway segments for which data are available. This means that it is not always possible—from existing data—to accurately mea- sure the actual length of a queue associated with a truck bottleneck. This is a limitation of the NPMRDS and other probe speed datasets. These probe datasets typically describe the average travel time for the entire segment for which data are reported. For example, a truck may travel at 70 mph over the first 4.5 miles of a 5-mile-long segment, but then fight through stop-and-go traffic over the last half-mile, averag- ing 15 mph. The result is a reported travel time (~5.86 minutes) that accurately reflects the travel time over the entire segment and that can be converted to an average speed estimate of ~51 mph. However, while the delay measurements based on that travel time and speed are cor- rect, the data limit the ability to directly identify the very slow speeds and queue that formed over the last half-mile of that segment. One common way of reporting roadway segments—especially when delay or travel time is used as the bottleneck ranking statistic—is to group smaller segments into modestly long road seg- ments that stretch from one major interchange to another. The contiguous small segments that make up these larger reporting segments typically have similar numbers of lanes and operating characteristics. They generally do not contain known geometric bottlenecks (for example, caused by major merging movements or lane drops) in the middle of the defined segment. They can range from 4 to 15 miles and constitute a length of roadway that might logically be turned into a major construction or improvement project. Additional details about segmentation are covered in Section 4.4 of this Guidebook. Table 6-11 taken from a recent FHWA webinar about the use of the NPMRDS (26), gives an example of these longer reporting segments provided by Wisconsin DOT. In Table 6-11, road segments range from 5 to 15 miles. They are defined as occurring from one major interchange to Case Study Highlight Since 2002, the American Transportation Research Institute (ATRI) has partnered with FHWA on the Freight Performance Measures (FPM) initiative. The FPM monitors the performance of selected truck-based freight facilities. The report provides rankings and performance on 100 of the most congested locations in the United States. Locations are not selected by specific criteria for inclusion in the study, but rather are identified as freight-significant based on multiple years of analysis, past research, surveys of private- and public- sector stakeholders and based on speed and volume datasets. More details are included in Appendix B.

Classifying and evaluating truck Freight Bottlenecks 67 another. The information in this table can be used to rank bottlenecks, based on subtracting the normal travel time from the worst peak-travel time. Each location can then be ranked based on this differential. In this example, the interchanges do not need to be freeway-to-freeway movements but can simply be locations at which major changes in traffic volumes occur. Some local insight is typically needed to create these longer segments, but insight gained by reviewing the performance of the shorter segments also can help guide the definitions of these longer reporting segments. Longer segments are particularly useful in the basic bottleneck identification and ranking process—that is, the “desktop” analysis. The use of longer segments limits the size of the output tables, which reduces the time needed for staff to review them. Moderately long segments also help ensure that total delay is effectively captured. Once completed, the desktop analysis results support a fairly quick and effective ranking process. On the basis of those results, agencies can then perform more detailed analyses that look at roadway performance within these longer segments. These “field analyses” are performed only for the highest-priority roadway segments. Each agency decides how many and which of these identified bottleneck sections it will study in more detail. In this way, the desktop analysis helps agencies manage their work load and helps ensure that the resources they apply to more detailed studies are efficiently allocated. 6.3.1 Desktop Analysis of Bottleneck Impacts on Travel Times One other way to examine the importance of identified bottlenecks is to examine their impacts on truck trip travel times. This can be accomplished by first using knowledge of the key freight movements in the state to develop a list of important freight O/D. For example, these movements could be from one of the major manufacturing centers in a state to a major port, or to the state border on an Interstate that leads to a major shipping destination for the commodities in question. It is then possible to compute paths or “trips” from the origin to the destination of each key freight movement. By using the cube analysis structure that describes the potential causation Table 6-11. Performance measures reported for longer segment lengths.

68 Guide for Identifying, Classifying, evaluating, and Mitigating truck Freight Bottlenecks factors for truck bottlenecks (as illustrated earlier in Figure 4-5), agencies can compute travel times with these paths. Then, by computing travel times over these paths for multiple days and start times, it is possible to compute the travel-time reliability of these key freight movements. It also is possible to determine which bottlenecks each of the trips passes through and the amount of time lost to those bottlenecks for each of the trips. Examining the delays in each bottleneck versus the total trip time and total trip reliability allows the analyst to understand the relative importance of each bottleneck in relation to the travel-time reliability of the key freight trips in the state or region. 6.4 Identify Causes and Rank Process-Based Truck Bottlenecks The ranking and cause analysis for process bottlenecks is somewhat different than the straightforward ranking analysis for congestion bottle- necks. First, to examine process bottlenecks, the analyst starts with an understanding of the cause of the truck delay. The analysis process is based on the specific type of trucking restriction (e.g., low-height bridge) required by a known problem (e.g., a given bridge does not meet standards—which is known though agency records and is likely an item that trucking firms complain about to the agency). The ranking process involves examining the relative size of the various deficiencies. Different rankings could be computed on the basis of the different performance statistics mentioned above: • Total cost imposed on the trucking community or • Number of trucks inconvenienced by a given restriction. More likely, however, process bottlenecks will be ranked on the benefit-to-cost ratio of the required mitigation, and that requires an understanding of the appropriate mitigation for each process bottleneck. 6.5 Conduct Field Analysis to Refine Bottleneck Understanding The desktop analysis provides the ability to quickly describe, scope, and rank truck bottle- necks across an entire region or state. It also allows a state or region to quickly grasp the overall delay trend (i.e., are hours of delay increasing or decreasing over time?). However, the limitations imposed by the need to have widely available, consistent data sources precludes the desktop analysis process from incorporating all of the local detail that is needed to perform the effective planning and engineering required to cost-effectively mitigate bottlenecks. In addition, understanding the overall trend always begs the questions, “Why is that trend occurring?” and “How does that trend apply to this particular location of interest?” Answers to those questions typically require more site-specific analysis. Consequently, the desktop analysis process is designed to be only the start of the bottleneck analysis effort. It provides enough information for the agency to effectively select the locations on which to perform more detailed analyses. The next step in the bottleneck identification and evaluation process is conducting those detailed field analyses. The field analysis starts with the results from the desktop analysis. In many cases, it relies on the same tools and reports that are available to the desktop analysis, but it involves a deeper examination of a limited . . . to examine [noncongestion-related] bottlenecks, the analyst starts with an under standing of the cause of the truck delay. . . . [the field analysis] relies on the same tools and reports that are available to the desktop analysis, but it involved a deeper examination of a limited number of (usually contiguous) roadway segments.

Classifying and evaluating truck Freight Bottlenecks 69 number of (usually contiguous) roadway segments. The field analysis also typically incorporates additional data into the bottleneck analy- sis that are not available statewide. In some cases these data already exist at the field study location but are not available at other parts of the state. This commonly occurs when the field study is performed on a major urban corridor, where large amounts of data already exist because of existing traffic management systems or because other stud- ies performed in the area have collected those data. In other cases these additional data must be collected specifically for the field analysis. In still other cases, agency staff that work in the area can describe in detail some of the contributing causes of local bottlenecks. Taking advantage of this local knowledge is an important part of the field analysis process. In the end, these additional data sources are developed to provide more depth to the analysis about why observed travel patterns are occurring and how those delays might best be mitigated. As a starting place for the field analysis, the results from the desk- top analysis describe when and where bottlenecks are occurring and provide insight into the factors that influence the formation and size of the resulting truck delays. Starting at this point allows the analyst to progress from a simplistic understanding of the factors that influence bottlenecks to a more detailed understanding of exactly what is causing bottlenecks on the priority corridors/location they are studying. For example, in the field analysis, the analysts might look at not just the overall amount and general timing (e.g., a.m. versus p.m. peak delay) of the delay reported for the large roadway segment, but they might examine the exact timing and formation of that delay on specific days, examining details such as the following: • Where within the larger reporting segment does a bottleneck form, and how does it propagate from that initial bottleneck location? • Is congestion routinely forming at one or more specific points within the study corridor, or is it forming throughout the corridor because of simply too much volume? • Is the delay occurring at specific points in the corridor because of known geometric attributes (e.g., high ramp volumes, or major weaving movements)? • Does congestion form randomly in time and space as a result of vehicle crashes? • Are crashes within the corridor randomly distributed or are they concentrated in specific locations, and if they are in specific locations, what are the attributes of those crashes and the locations where they are occurring? It is common to specifically collect data for field analysis. For exam- ple, the agency might collect a new vehicle classification-based traffic count to obtain better truck volume data. Truck volume data available at the statewide level might be weak in a location selected for more detailed analysis, and improving the estimate of truck delays might make collecting those data important. Similarly, the agency might obtain data on factors such as transportation system management and operations (TSM&O) strategies being conducted within that corridor. These data would be used to inform the analyst whether specific bottleneck mitigation strategies already were being implemented in the study corridor. The availability of those services would then set in motion additional analyses, such as the response time of the existing incident management program, the nature of the crashes that resulted in the largest delays, and the size and scope of those incident management efforts. Case Study Highlight A recent study by the Hampton Roads Transportation Planning Organization (HRTPO) identified freight bottlenecks for highways that are expected to be part of the National Freight Network and forecast likely future truck bottleneck locations. In this field analysis, researchers considered many aspects that could cause bottlenecks, including defining deficient bridge structures, identifying height and lane width restrictions, pavement condition, and truck delay on the highway network. More details are in Appendix B. It is common to specifically collect data for field analysis.

70 Guide for Identifying, Classifying, evaluating, and Mitigating truck Freight Bottlenecks The analyst could then compare the observed congestion patterns and statistics with the existing traffic management efforts on those roadways, as well as compare those outcomes with the state-of-the-art or state-of-the-practice for mitigating the types of congestion identi- fied in the study area. For example, if the field analysis indicated that a significant portion of “worst” travel days occurred when truck-involved crashes occurred, and the review of the incident management system did not include heavy-duty tow trucks, then one obvious mitigation approach would be to offer ways to speed access for those larger response vehicles. A good field analysis also includes agency staff who work in the geo- graphic region containing the bottleneck along with private-sector freight stakeholders that operate trucks or ship goods on the roadways of concern. Agency staff familiar with the adopted local plans and the local political and organizational working relationships must contrib- ute their knowledge of these plans and relationships to the field study. Understanding the local organizational relationships is often a key to successful implementation of bottleneck mitigation efforts. Leveraging existing plans and local interests can greatly speed the implementation effort and frequently decrease the cost of bottleneck mitigation. There- fore, partnering with local agencies, reaching out to local stake holders, and working across silos can help with the field analysis. Private-sector freight stakeholders can provide many pieces of valu- able information in the truck bottleneck evaluation process. Most nota- bly, they can provide information on the causes of why trucks slow down at a certain location, including road curvature, grades, lane width, or other safety concerns. For process-based delays, they are critical for understanding how truck patterns are altered due to regulations, includ- ing weight restrictions, truck bans, time-of-day restrictions, and other causes of truck delays. At this point in the analysis, it is generally a good practice to allow the private sector to comment on the accuracy of the analysis and provide input on some of the causes of what has been identified in the data. The outcome of these more detailed analyses is insight necessary to determine the types of improvements that are required to reduce the observed congestion. This mitigation is discussed in the next chapter of this Guidebook. Case Study Highlight The Oregon DOT recently identified all vehicle bottlenecks and recommended mitigation strategies for five corridors in Oregon in response to FHWA’s Localized Bottleneck Reduction (LBR) Program. The first tier of the two-tier analysis used loop detector and historical crash data to identify bottlenecks for a typical commute during the morning and afternoon peak periods. The second tier validates this analysis by reviewing existing documentation, available video footage, and field observation. The research team identified typical causes of the localized bottlenecks and suggested improvement strategies. More details are in Appendix B. Staff familiar with the adopted local plans and local political and organizational working relationships must contribute their knowledge of these plans and relationships to the field study.

Next: Chapter 7 - Options for Mitigating Truck Bottlenecks »
Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB's National Cooperative Highway Research Program (NCHRP) Research Report 854: Guide for Identifying, Classifying, Evaluating, and Mitigating Truck Freight Bottlenecks provides transportation agencies state-of-the-practice information on truck freight bottlenecks using truck probe data rather than traditional travel demand models. The report embraces a broad definition of truck freight bottlenecks as any condition that acts as an impediment to efficient truck travel, whether the bottleneck is caused by infrastructure shortcomings, regulations, weather, or special events. The comprehensive classification of truck freight bottleneck types described in this report provides a standard approach for state departments of transportation, metropolitan planning organizations, and other practitioners to define truck freight bottlenecks and quantify their impacts.

This project produced the following appendices available online:

  • Appendix A: Selected Details of State-of-the-Practice Review
  • Appendix B: Short Summaries of Selected Case Studies
  • Appendix C: Data Quality Control Examples
  • Appendix D: Additional Performance Measure Discussion and Analysis Procedures
  • Appendix E: Truck Bottlenecks and Geometrics
  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!