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.
4Problem Statement Most roadwork today involves reconstruction, rehabilita- tion, and maintenance of existing roadways and often occurs in or near moving traffic. More and more, agencies are doing roadway work on high-volume facilities at night to reduce adverse traffic impacts and complaints by the public that typically occur when the same work is done during the day. Minimizing traffic impacts of roadwork activities is a key emphasis of the Federal Highway Administration (FHWA) (1). Nighttime travel is commonly characterized by the following: ⢠Lower traffic volumes, ⢠A higher percentage of truck traffic, ⢠Higher operating speeds, ⢠Reduced visibility, and ⢠Higher concentrations of drowsy and impaired drivers. Traffic volumes at night on a roadway facility are typically much lower than during daylight hours, which is often the main reason that work is performed at night. Lower traffic volumes yield reduced vehicular exposure to the work zone, which is a key determinant associated with crash frequencies. Furthermore, working during nighttime hours reduces the likelihood and extent of traffic congestion that could result from performing that work. Traffic congestion has been associated with higher crash frequencies and rates by several researchers (2, 3). However, lower volumes provide greater maneuverability to drivers, and can allow higher operating speeds to occur into and through the work zone than would have been possi- ble had the work been done during the day. Obviously, lower light levels at night reduce visibility for drivers and workers relative to what would be available during the day. Based on crash data and other collected information, it is well recog- nized that greater concentrations of impaired drivers are on roadways at night than during the day (4, 5, 6). Driver ex- pectancy to encounter roadwork activity may also be less at night than during the day in some areas, depending on how extensively night work has been embraced by those regions in recent years and how well the night work is publicized to drivers. Each of these factors may have the potential to increase crash risk per vehicle and crash severity at night, compared to daytime conditions. The fact that arguments exist on both sides of the question as to whether working at night is more or less safe than work- ing during the day emphasizes the need for this research. Rather than continuing to rely on conjecture and subjective opinion, NCHRP initiated this project to objectively deter- mine and document how nighttime and daytime work zones affect traffic safety. Four specific objectives were identified: ⢠Determine the crash rates for nighttime and daytime work zones; ⢠Determine the nature of, and identify similarities and dif- ferences between, traffic-related crashes in nighttime and daytime work zones; ⢠Develop management practices that promote safety and mobility in nighttime and daytime work zones; and ⢠Develop work zone crash reporting recommendations to further improve the data collected on work zone crashes. Previous Research Work Zone Effects on Traffic Safety Over the past 30 years, numerous researchers have examined the influence of work zones on roadway crashes, primarily in terms of how normal crash rates or the likelihood of crashes changes when a work zone is installed at a particular location. In recent years, one finds that crashes typically increase ap- proximately 20 to 30 percent within work zones relative to the normal crash experience for those locations, although the amount of the increases varies from study to study (7, 8, 9, 10, C H A P T E R 1 Background
11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24). Differences in work zone designs, quality of traffic control device main- tenance, types of work performed, and other roadway and traffic characteristics probably contribute to the varying re- sults observed. In addition, recent studies have shown that the relationship between work zone crash likelihood and road- way (i.e., average daily traffic [ADT], lane and shoulder widths, etc.) and work zone characteristics (i.e., duration, length, etc.) are nonlinear (24, 25). As to whether such crash increases are more significant at night, the evidence is less clear. Some studies have found that nighttime crashes in work zones increase by a greater per- centage than daytime crashes (21, 26, 27), but other studies have found the increases in daytime and nighttime crashes to be similar (8, 19, 28). In essentially all studies described above, the changes in crashes were computed over the entire duration of a long- term roadway rehabilitation or reconstruction project, dur- ing times when work was occurring as well as when the work area was inactive. Projects involving major rehabilitation or reconstruction of the roadway often require temporary degradations in roadway geometry such as narrowed lanes, shortening of entrance or exit ramps, and the placement of temporary concrete barriers immediately adjacent to the travel lanes. These degraded geometric changes are left in place when work is occurring as well as when the work zone is inactive. Consequently, the changes in crashes reported in these studies actually represent the combined effect of both the degraded geometric conditions in the work zone and the influence of work activity itself. Although some consensus exists that overall crash rates increase during roadway repair and reconstruction activities (even if the amount of the increase is of some debate), the literature is less definitive with regards to whether or not crashes tend to be more severe, less severe, or as severe as under non-work zone situations. Several researchers have concluded that work zone crashes are no more severe than non-work zone crashes (8, 9, 13, 18, 19, 29, 30). However, others (21, 31, 32) have found that work zone crashes were more severe than non-work zone crashes or increased more signif- icantly than property-damage-only (PDO) crashes in the work zone in the databases they examined. Once again, differences in the type of analysis used and the characteristics of roadways and work zones examined may be at least partially responsible for the divergent findings. Despite the difficulties in properly attributing changes in crashes to the effects of the work zone design or to the actual presence of work activity itself, rear-end crashes are often found to be overrepresented in the crashes that do occur. The percentage of work zone crashes that involve rear-end impacts increased 7 to 83 percent, depending on the study cited and location of the work zone examined in that study (8, 18, 21, 27, 33, 34). The creation of traffic and congestion during work activities is usually hypothesized as the major reason for the observed increase in rear-end crashes. Traffic and con- gestion can be created by temporary lane closures, other re- ductions in roadway capacity (i.e., narrowed lanes, lane shifts, etc.), drivers who are confused about their proper travel path and slow down, or the movement of construction equipment into and out of the work area. Nighttime versus Daytime Work Zone Crashes Decisions about whether to perform work in travel lanes during daylight hours or at night should be based, in part, on which approach is likely to yield the lowest crash costs over the duration of the project (35). Previously, only three small studies were identified that attempted to examine this question directly. In the first study, researchers in California examined eight construction projects where work was performed at night to minimize traffic disruptions (36). Researchers found that crash rates per million-vehicle-miles (mvm) of travel ex- posure were consistently and substantially higher at night when work activity was present. The magnitude of increases ranged from 67 to 156 percent, with an overall average crash rate increase of 87 percent. Researchers further stratified the data based on whether or not travel lanes were closed during the period of work activity each night. They found the crash rates at night during lane closures to be an additional 75 per- cent higher than during periods of work activity at night when travel lanes were not required to be closed. Researchers also examined crash rates on the basis of crashes per million hours (per mile per lane to normalize data between projects) and found the rates to increase an average of 122 percent dur- ing periods of night work activity (as compared to non-work hours). Unfortunately, the analysis could not answer the ul- timate question as to whether or not crash rates per mile of work zone and/or crash costs would have been higher or lower than this had the work been performed during the day instead. A second study used Illinois fatal work zone crashes and national estimates of work activity occurring during daytime and nighttime periods to assess the relative safety of day- time and nighttime work zones (37). The estimates of exposure came from a sampling process of scheduled work activity in- formation posted online by state and local transportation agencies (38). Based on their analyses, the researchers con- cluded that night work was five times more hazardous than daytime work activity. However, the lack of actual exposure data from Illinois work zones for use in the comparison was noted as a key limitation of the analysis. A third study examined five urban freeway reconstruction projects in Texas where all work activities that required lane 5
closures were done at night, and other types of work activities off the travel lanes occurred during the daytime (39). Two other smaller projects that involved nighttime lane clo- sures for pavement resurfacing activities were also examined. Researchers found that the major reconstruction projects ex- perienced some increase in crashes during both daytime and nighttime periods, indicating that temporary geometric de- gradations (closure of shoulders, reduced acceleration lane lengths at ramps, concrete barriers immediately adjacent to travel lanes, etc.) had an effect on crash likelihood. Crash in- creases were even larger during periods of work activity (day or night), on average, with the increase in crashes during night work activities somewhat higher than during daytime work activities. Given that the daytime work activities at these proj- ects did not involve lane closures whereas the nighttime work activities did, this finding was expected. Unfortunately, as with the previous study, these results do not provide any indication of how crashes might have been affected had the lane closures and work that was done at night actually been instead performed at each site during the day. Nighttime versus Daytime Worker Safety Although not all worker accidents involve traffic crashes, working at night does significantly impact the lives of high- way workers and so is of at least some relevance to the goals of this study. Overall, the safety impacts of performing road- work at night (relative to daytime operations) on highway workers have not previously been examined in detail. Work- ers generally perceive traffic speeds past the work site to be higher at night and so also perceive their level of safety to be diminished (40). The limited amount of accident data avail- able on highway workers has not necessarily confirmed this perception, however. A study by the National Institute of Occupational Safety and Health (NIOSH) examined fatal occupational injuries for highway construction workers be- tween the years 1992 and 2000 (41). Based on their assess- ments, the NIOSH researchers concluded that âworking at night is not responsible for the overall increase in highway worker deaths.â However, it should be noted that there were insufficient data to actually compare highway worker accident rates at night versus during the day. Implications for This Study There is a strong general consensus in the literature that work zones increase the likelihood of crashes that occur on a particular segment of roadway. Less agreement exists as to whether work zones result in more severe, less severe, or equally severe crashes as before work began. Certain types of crashes appear to be more affected by work zone presence than others, but again, the amount of influence is heavily site dependent. Such variability in studies to date is not surprising; work zones themselves are highly variable entities. The literature does imply that the amount of increased crash risk at a given project location is a combination of temporary changes in geometrics and influences due to work activity. Work activity influences can be drivers distracted by work operations and equipment, turbulence created by work vehicle or equipment access to and from the work area, and tempo- rary lane closures that increase traffic densities (possibly to the point of congestion) and require drivers to maneuver around the closure. However, efforts to better understand the relative contributions of work zone design and work activities to the increased crash risk are fairly limited in the literature. An even more critical factor that has not been previously evaluated in the literature with any degree of success is the actual difference in safety between performing a particular work activity or project at night versus doing that same activity or project at the same location during the day. Rela- tive safety is one of the key recommended considerations that practitioners face when assessing whether or not to do a par- ticular project or project task at night (35). A few studies have provided some insight into the amount by which the normal nighttime crash rate increases if work is performed at night. However, the increase that would have occurred in the day- time crash rate at those locations if the work had been done during the day instead has not been quantified. Unfortunately, opportunities to evaluate this question directly at individual project sites are almost nonexistent. In most cases, the pri- mary reason that an agency and highway contractor work at night at a location is that there is a need to close one or more travel lanes for a temporary period to complete the work, and doing so during daytime hours would generate unacceptable traffic delays and queues as well as severely limit the contrac- torâs ability to move work vehicles and materials into and out of the workspace. Unacceptable impacts on adjoining prop- erties may also occur from temporary lane closures during daytime hours. Related to this comparison of nighttime-daytime crash risk increase is the choice of appropriate measures of risk to use, i.e., should it be crash rate per mile of work zone or crash rate per mvm? Traditional crash rates normalize crashes on the basis of vehicular-miles of travel or a similar measure of exposure. This rate reflects a level of risk to an individual driver traversing that particular roadway segment. Percent- age changes in this rate or similar indicators, such as the percentage change between actual and expected number of crashes in a given time period, thus indicate how individual driver risk is affected by the presence of the work zone. Cer- tainly, this indicator of motorist risk is an important consid- eration. However, from the practitionersâ perspective, the decision whether to work at night or during the day must also consider the consequences of increased crash risk to the 6
driving population as a whole. Whereas the increase in crash risk to an individual motorist may, in theory, be greater at night than during the day, the much lower traffic volumes (and thus vehicle exposure) that typically exist at a given location at night may more than offset this incrementally higher increased risk. Higher traffic volumes during the day mean that the same number of crashes will produce much lower crash rates per mvm. If the day versus night decision is for a given work zone, then it would appear that the practi- tioner will want to minimize the number of crashes (assuming equal severity). Thus, crash rate per mile of work zone would appear to be a better comparison metric. Certainly, differ- ences in the severity of the increased crashes may also exist between daytime and nighttime work operations, which also must be considered in the analysis. Together, this implies that the use of additional crash costs, normalized on the basis of the amount of work activity required at a given project loca- tion, will most closely reflect the information that highway agencies must weigh in their decisions of whether or not to work at night. Finally, although emphasis is traditionally placed on un- derstanding and measuring the safety impacts of highway work zones to the motoring public, when considering the differences between daytime and nighttime work operations, the consequences to highway workers also need to be taken into consideration. Unfortunately, very little data on this issue exist. The highway contracting community perceives work- ing at night to be a significantly greater risk to workers than working during the day. National databases (such as the Bureau of Labor Statistics) do not allow for a thorough com- parison of nighttime and daytime highway work condition safety. Data sources that provide at least some insight into dif- ferences in highway worker risks during these work periods are sorely needed. Study Overview Researchers ultimately adopted a two-pronged investiga- tion for this research project, based on data sources available to the research team. The first prong utilized the NYSDOT Work Zone Accident database. This database is a one-of-a- kind resource developed in the 1980s and expanded over the years, specifically for use in tracking all types of work-zone- related traffic crashes and worker accidents on NYSDOT con- struction projects statewide. The database has been a valuable asset to both NYSDOT personnel and other researchers in examining various questions about work zone features and crash characteristics that cannot be examined through tradi- tional state traffic crash records and databases (42, 43, 44, 45, 46). For this study, researchers explicitly examined the rel- ative differences in the types and severities of traffic crashes and worker construction accidents during both daytime and night- time work operations on New York freeway and expressway facilities. Even more importantly, the database included spe- cific information on various types of worker-involved traffic crashes and construction accidents sustained during both day and night work activities; this is something that is not available at the present time nationally in any other database. The second prong of the research effort was the collection and analysis of crash experiences of a four-state sample of work zones. For some of the projects, work activities were done predominantly during the daytime; for other projects, work activities were done mainly at night. For still other proj- ects, some work activities were done during the day (mainly those activities that did not require the temporary closure of travel lanes), and other activities that required travel lanes to be closed for several hours were performed at night. Four states were included in the analysis: ⢠California, ⢠North Carolina, ⢠Ohio, and ⢠Washington. These states were selected because they provide access to their statewide traffic crash and roadway inventory databases through FHWAâs Highway Safety Information System (HSIS) and because they reportedly had sufficient numbers of night work projects ongoing during the time period of interest in this study. 7
