TEMPORARY LOSSES OF HIGHWAY CAPACITY AND IMPACTS ON PERFORMANCE: PHASE 2

Title | Contents | Acknowledgements | Exec. Summary
1. Intro | 2. Approach | 3. Crashes | 4. Breakdowns | 5. Work Zones | 6. Weather | 7. Signal Timing
8. RR Crossings | 9. Toll Facilities | 10. PUD | 11. Results Summary | 12. Next Steps | 13. References

11. Results Summary

11.1 Results

Temporary capacity losses due to work zones, crashes, breakdowns, adverse weather, sub-optimal signal timing, railroad crossings, toll facilities, and urban PUD activities resulted in over three and a half billion vehicle-hours of delay on U.S. freeways and principal arterials in 1999 (Fig. 30, Table 37). Assuming an average vehicle occupancy of 1.6 persons, this translates into about six billion person-hours of delay. Assuming an average value of time of $15 per hour for each person impacted, temporary capacity losses produced about $55 billion in lost time alone in 1999. Because conservative assumptions have been used throughout this analysis, and because several significant sources of delay have not been included, these estimates are believed to be a lower bound on the actual impacts of TLC.

Figure 30. Non-fatal crashes and work zones account for over two-thirds of the delay from temporary losses of capacity.

Non-fatal crashes were estimated to be the source of most delay from temporary capacity reductions, accounting for 45.5 percent of estimated delay. Work zones accounted for about a quarter of delay from TLC. Breakdowns, adverse weather, and signal timing were next, causing 12, 9, and 8.1 percent, respectively. Delay from toll facilities, fatal crashes, railroad crossings, and urban PUD activities combined were responsible for just over one percent of estimated delay.

The sources of delay are tabulated by highway type, urban area size, peak period versus off peak, and congestion level in Tables 38 through 40. Surprisingly, over 85 percent of the delay estimated in TLC2 occurs in the off-peak period or on uncongested segments in the peak period. Whether or not in times and locations with recurring congestion, TLC2 indicates that Americans lose 2.5 hours for every 1,000 miles of travel due to delay from incidents, work zones, bad weather, poor signal timing, railroad grade crossings, double-parked urban delivery vehicles, and toll booths. Delay is over 4 hours per 1,000 miles of travel in very large urban areas, about 3 hours and 45 minutes in large urban areas, over 2 hours in small and medium areas, and 45 minutes in rural areas.

In terms of total delay from TLC events, urban areas were estimated to have experienced more delay than rural areas, and larger urban areas experienced more delay than smaller ones. Although Figure 31 shows that medium-sized urban areas experienced much less total delay than might have been expected (based on area type and size), when the amount of travel within each area type/size is considered, estimated delay for this urban area size follows the trend for other urban areas (Fig. 32).

Figure 31. Medium-sized urban areas experienced the least amount of delay from TLC events in 1999.

Figure 32. When the amount of travel in each area type/size is considered, the delay experienced in medium-sized urban areas is more in line with size (in terms of population).

In urban areas, more delay from TLC events was realized on freeways than on principal arterials, but in rural areas the opposite was true: other principal arterials experienced slightly more delay (Fig. 33). This also held true when the amount of vehicle travel on these highway types was considered.

Figure 33. Delay from TLC events was most prevalent on urban freeways.

Most of the delay from temporary capacity losses (60 percent) was experienced during off-peak periods (Fig. 34). About 27 percent of delay was experienced on uncongested highways during peak periods, and nearly 13 percent was experienced on congested highways during peak-periods.

Figure 34. Most delay from TLC events occurred during off-peak periods.

While total delay from temporary events is least on congested highway segments (Fig. 34), the picture is reversed when calculated per mile of travel (Fig. 35). This explains why unexpected delay seems worse to people traveling in the peak period. While the ratio of congested VMT to total VMT seems low in Tables 38 and 40 and figures 34 and 35, the TLC2 study overestimates the amount of VMT on congested highway segments because VMT on a segment congested in the peak hour is classified as congested for the entire peak period.

Figure 35. On a per-mile-of-travel basis, delay from TLC events was more likely to occur during peak periods on congested roadways.

11.2 Comparing TLC2 Results to Others

This study has broken new ground in developing the first "bottom-up," nationwide estimates of temporary losses of highway capacity and resulting delay. In the course of the research, much has been learned about both data sources and methodologies that can be applied to improve and expand information about TLC impacts.

The TLC2 estimates were compared to two sets of estimates by Texas Transportation Institute (TTI), including estimates of incident delay for 85 urban areas in 1999 in the 2004 Urban Mobility Study (Schrank and Lomax 2004) and unpublished estimates of incident delay in 1999 for all FHWA-recognized urban areas in support of FHWA's annual Performance and Accountability Report. FHWA-recognized urban areas are urbanized areas defined by the U.S. Bureau of the Census with populations of at least 50,000—local transportation officials may adjust Census-defined borders. There were just over 400 urban areas in 1999. The TTI estimates are based on the most recent TTI procedures applied to 1999 data, with more extensive quality checks for the published estimates for the 85 areas than for the unpublished estimates for all urban areas. Both published and unpublished TTI estimates of delay per person are converted to vehicle delay with TTI's vehicle occupancy factor of 1.25. The comparable TLC2 estimates are for vehicle delay from crashes and breakdowns, but cover all urban and rural areas. Neither TLC2 nor TTI estimates cover roads other than freeways and other major arteries.

The delay estimates for crashes and breakdowns in TLC2 are slightly higher than both sets of incident delay estimates from TTI (Fig. 36). The differences between TLC2 and TTI for all urban areas vary slightly by urban area size (Fig. 37). A complete assessment of the causes of differences between TTI and TLC2 estimates is beyond the scope of this study.

Figure 36. A comparison of three studies that estimate delay: TLC2, TTI's Urban Mobility Study (85 urbanized areas), and TTI's estimate for the FHWA Office of Operations (all urbanized areas).

Figure. 37 TLC2 delay estimates are somewhat higher than TTI's (all-urban-area study) for most urban area sizes.

11.3 A Composite Picture of Delay

No single empirical or modeling study provides a comprehensive estimate of all sources of delay. However, a comprehensive picture can be assembled by combining elements of TLC2 and TTI's unpublished estimates for FHWA's annual Performance and Accountability Report. The resulting picture is a very approximate composite because very different methods are used in TLC2 and the TTI studies. TTI estimates recurring delay from weekday commuting peaks, while TLC2 estimates recurring delay from two elements (suboptimal signal timing and tollbooths) not covered by TTI. TTI estimates nonrecurring delay from relationships between incident delay and recurring delay in urban areas, while TLC2 uses a bottom-up approach to estimating nonrecurring delay from a variety of sources in both urban and rural areas.

The composite picture uses TTI's estimate of recurring delay for all urban areas, the TLC2 estimate of recurring delay for suboptimal signal timing and tollbooths, and the more comprehensive TLC2 estimates of nonrecurring delay. The resulting 5.1 billion hours of delay is 35 percent recurring and 65 percent nonrecurring. Recurring delay is a higher percentage in the larger cities (Fig. 38).

Figure 38. Very large urban areas had a somewhat greater share of recurring delay than other area types.

The combined TLC2-TTI estimate suggests a slightly higher contribution of nonrecurring delay than the composite picture compiled for the Federal Highway Administration from a variety of studies and professional judgments. The relative delay shares by source in these composite pictures are within 10 percentage points in every category except incidents (Table 41).

Neither composite picture is complete and adequately supported with empirical observations. Recurring congestion from weekend and holiday travel in all areas and recurring weekday congestion in rural areas are poorly captured, if at all. The full effects of bottlenecks, the extent and intensity of most forms of temporary capacity reductions, and the consequences of dramatic increases in trucking are not adequately understood or based on robust empirical studies. Delay on roads smaller than freeways and other major arteries is another unexplored part of the picture. Substantial data collection and analysis are necessary before a complete picture of delay can be framed.

Previous | Table of Contents | Next