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.
8Database Description In the mid-1980s, the Construction Division of NYSDOT initiated a program to compile detailed information on traf- fic crashes and worker construction accidents that occur on their construction projects. Since the mid-1990s this proce- dure has generated reports for nearly all traffic crashes and worker construction accidents, with the reports initiated by department staff and contractors. These reports are supple- mented by standard police accident reports for many traffic crashes and some worker accidents. NYSDOT maintains a detailed database that is used to generate annual reports of work zone accidents, track overall safety trends, and prepare special reports addressing specific safety issues. The entire database is rather extensive. Researchers focused on the following variables for analysis: ⢠Time of accident occurrence (day or night); ⢠Time when work activities were typically performed (day, night, or both); ⢠Facility type; ⢠Work zone situation type at the time of the traffic crash (flagging, lane closure, mobile operation, etc.); ⢠Accident severity; ⢠Type of traffic crash occurring if applicable (rear-end, side- swipe, single vehicle, vehicle intrusion [a passenger vehicle traveling through the work zone and entering the work- space], vehicle impact with worker, etc.); ⢠Type of worker construction accident if applicable (falls, equipment accidents, trenching accidents, accidents be- tween work vehicles and workers, etc.); and ⢠Contributing factors to the traffic crashes (driver inatten- tion, poor driver judgment, etc.). Generally speaking, accidents occurring between 6 am and 6 pm were coded as daytime accidents, and those from 6 pm to 6 am were coded as nighttime accidents. Information about the work zone situation at the time of the crash, whether the vehicle intruded into the workspace, and whether the vehicle impacted a highway worker are all items not easily obtained (if available at all) from typical statewide traffic crash records systems. It should be noted that statewide statistics on the amount of day and night work operations each year in New York were not available to allow development of any type of crash rate measure. Data Reduction and Analysis Researchers obtained 6 years of NYSDOT work zone inci- dents (traffic crashes and construction accidents) for calendar years 2000 through 2005. The majority of projects that in- volved night work activity were performed on freeway and expressway facilities. Therefore, all freeway/expressway work zone incidents were first extracted from the full NYSDOT database for each year, using a combination of a âhighway typeâ data field and extensive knowledge of the New York roadway network by a member of the research team. The goal of this step was to develop a consistent basis for comparison of daytime versus nighttime traffic crashes and construction accidents. Researchers reviewed each incident report to verify the accuracy of the codes used in each of the data fields rela- tive to the brief narrative of the accident included in each record. In cases of obvious miscoding or where a code was not provided, researchers reviewed the descriptive narrative of the incident and manually inserted the correct code(s). Ulti- mately, more than 3,400 traffic crashes and construction accidents over a six-year time period (2000 through 2005) were available for analysis. Table 1 summarizes how these incidents were distributed across traffic crashes and highway worker construction accidents in both daytime and nighttime working conditions. Researchers then systematically segregated the data accord- ing to incident time period (day or night) and typical period C H A P T E R 2 NYSDOT Work Zone Accident Database Analysis
of work operation (day, night, or both). While most of the crashes and accidents included in the database occurred dur- ing active work periods, some did occur during periods when the work zone was inactive. Statistical distributions of the other data fields were then computed separately for daytime incidents at daytime work operations and for nighttime inci- dents at nighttime work operations. Further stratification was made to examine lane closure work operations as its own dis- tinct subset. As previously noted, temporary closure of one or more travel lanes on a freeway-type facility is often limited to night hours in order to avoid high traffic volume time periods. Direct comparison of incident characteristics of this particular work zone situation during the day to the same situation at night was thus of primary interest to the research team. Researchers then computed statistical distributions of the various data field elements and used chi-square statistical tests of independence to check whether the differences in the distributions between daytime and nighttime incidents were significant. As previously noted, exposure data were not available to allow traffic crash or construction accident rates to be calculated from these data. Findings Work Zone Traffic Crash Analysis Work Zone Conditions Where Crashes Occur Daytime and nighttime work zone crashes on NYSDOT freeways and expressways differed significantly in how they were distributed among the common work zone traffic control operations utilized on these facilities. As shown in Table 2, a substantially higher percentage of nighttime crashes occurred during lane closure operations than of daytime crashes (57.6 vs. 50.4 percent, respectively). On the other hand, a higher percentage of daytime crashes occurred where there was only minor traffic control present and no work was occur- ring (17.2 percent vs. 7.6 percent of nighttime crashes). As previously stated, the typical reason for working at night is that one or more travel lanes must be closed for several hours to perform the work, and doing so during the day would cause unacceptable traffic impacts. The greater relative frequency of temporary lane closures at night is also a likely explanation of the higher percentage of nighttime crashes during traffic control setup and takedown 9 Daytime Work Periods, Daytime Accidents Nighttime Work Periods, Daytime Accidents Nighttime Work Periods, Nighttime Accidents Daytime Work Periods, Nighttime Accidents Total Traffic Crashes 1762 9 316 102 2189 Construction Accidents 931 114 1045 Total 2693 9 430 102 3234 Table 1. Summary of NYSDOT work zone traffic crashes and construction accidents on freeways and expressways (2000-2005). Type of Work Zone Traffic Control in Use Daytime Work Operations, Daytime Crashes (n = 1757) Nighttime Work Operations, Nighttime Crashes (n = 316) Lane Closure Minor Traffic Control, Work Inactive Minor Traffic Control, Work Active Flagging Shoulder Closure Median Crossover Lane Shift During Traffic Control Setup/Takedown Full Road or Bridge Closure Other 50.4% 17.2% 8.3% 8.0% 5.4% 3.3% 3.2% 3.1% 0.7% 0.4% 57.6% 7.6% 9.5% 3.2% 0.9% 1.3% 0.9% 14.2% 3.8% 1.0% Chi-Square Test Results Daytime and nighttime distributions are significantly different from each other a a X2 = 94.510 > X2Crit (6, 0.05) = 12.592 Table 2. NYSDOT work zone crashes by traffic control conditions.
relative to daytime crashes (14.2 percent vs. 3.1 percent, re- spectively). Temporary lane closures require setup and take- down activities each night so that all travel lanes are returned to service for peak travel periods, which would imply a higher relative frequency of these activities occurring during night- time hours compared to daytime hours. Nevertheless, the percentage of such crashes during nighttime hours seems rather high given that the amount of time typically required for such setup and removal is typically much less than the du- ration of the actual temporary lane closure. It is also noted that crashes at flagger-controlled sites are more frequent in daytime operations, which most likely reflects the relatively lower usage of flaggers for traffic control at night. Severity of Work Zone Crashes The comparative effect of day versus night work on crash severity is illustrated in Table 3. Overall, work zone crash sever- ities trend slightly higher at night when consolidated across all types of work zone situations and for lane closure traffic crashes specifically. However, if only non-worker-involved crashes are considered, no significant differences in severity are detected between daytime and nighttime crashes. As can be seen in the table, worker-involved crashes tend to be more severe at night than during the day. Whereas 29.6 percent of worker-involved crashes during daytime lane closure opera- tions resulted in fatalities or injuries, 50.8 percent of such crashes at nighttime lane closure operations resulted in injuries or fatalities. A similar trend is evident when worker-involved crashes are examined across all work zone types. Over one- half of the worker-involved crashes at night resulted in in- juries or fatalities, and only about one-third of those types of crashes during the day involved fatalities and injuries. The designation of a crash as being worker involved does not automatically imply that the worker was actually hit by the vehicle and sustained the injury. In fact, for many of the worker-involved crashes, the driver or other occupants of the vehicle sustained the most serious injuries. In these instances, the worker was involved but somehow managed to avoid being struck. If the crash did involve a vehicle striking a worker, the result was usually quite severe. Overall, 93 per- cent of those workers who were struck by a vehicle during the 10 All Work Zone Traffic Control Types Lane Closure Traffic Control Work Zones Only Injury Severity Daytime Work Operations, Daytime Crashes Nighttime Work Operations, Nighttime Crashes Daytime Work Operations, Daytime Crashes Nighttime Work Operations, Nighttime Crashes All Traffic Crashes: Fatal Injury PDO (n = 1762) 1.4% 36.4% 62.2% (n = 304) 3.3% 43.4% 53.3% (n = 886) 1.1% 32.5% 66.4% (n = 182) 2.7% 41.2% 56.1% D aytime and nighttime distributions are significantly different a Daytime and nighttime distributions are not significantly different d Traffic Crashes: No Workers Involved Fatal Injuries PDO (n = 1423) 1.5% 37.3% 61.2% (n = 195) 3.6% 40.0% 56.4% (n = 771) 1.3% 32.9% 65.8% (n = 123) 2.4% 38.2% 59.3% Chi-Square Test Results Chi-Square Test Results Daytime and nighttime distributions are not significantly different b Daytime and nighttime distributions are not significantly different e Traffic Crashes: Workers Involved Fatal Injuries PDO (n = 339) 0.9% 32.7% 66.4% (n = 107) 2.8% 49.5% 47.7% (n = 115) 0.0% 29.6% 70.4% (n = 59) 3.3% 47.5% 49.2% Chi-Square Test Results Daytime and nighttime distributions are significantly different c Daytime and nighttime distributions are significantly different f a X2 = 12.609 > X2Crit (2, 0.025) = 7.378 b X2 = 4.757 < X2Crit (2, 0.025) = 7.378 c X2 = 12.068 > X2Crit (1, 0.025) = 5.024 d X2 = 8.671 > X2Crit (2, 0.025) = 5.024 e X2 = 1.912 < X2Crit (1, 0.025) = 5.024 f X2 = 7.596 > X2Crit (1, 0.025) = 5.024 Table 3. NYSDOT work zone crash severity.
day sustained injury, as did 100 percent of workers who were struck at night. Types of Traffic Crashes Occurring Significant differences were detected in the types of collisions that occurred on freeways and expressways during daytime versus nighttime work zone operations. Table 4 illustrates the distribution of daytime and nighttime collision types for all types of work zone situations and also for lane closure work zone operations only. The percentage of the crashes that in- volve rear-end collisions is substantially lower at night for all work zone traffic control types combined and for work zone lane closure operations in particular. Presumably, the lower traffic volumes present at night allow work activities to be accomplished with fewer disruptions in traffic flow and fewer abrupt speed changes by vehicles. The decrease in rear-end crashes at night is offset by small increases in the percentage of intrusion crashes, impacts with truck-mounted attenua- tors (TMAs), and impacts with work equipment, materials, and/or debris outside of the work area. One possible explanation for these increases is that in day- time conditions, the traffic congestion that is created migrates the majority of the crashes (most of them rear-end collisions) upstream away from the work area to where drivers first have to make significant adjustments in their speed. At night, this crash impetus does not exist upstream, and so the conse- quences of driver inattention congregate in and around the area of work activities themselves. Add to this the fact that drivers at night are more likely to be impaired than during the day, and it is fairly easy to understand why intrusions and im- pacts with TMAs would make up a greater proportion of nighttime work zone crashes. Statistically, the differences be- tween daytime and nighttime periods shown in Table 4 are highly significant when all work zone traffic control types are considered together. Limiting the analysis to lane closure traf- fic control work zone crashes, differences between daytime and nighttime conditions are still statistically significant but not by as large of an amount. Table 4 also illustrates that the percentage of intrusion crashes involving workers is substantially higher at night work operations than during daytime work operations. The amount of the increase is the same for lane closure crashes (0.7 percent daytime vs. 3.8 percent nighttime) and for all work zone crashes (0.7 percent daytime vs. 3.8 percent night- time). However, the absolute frequency of these crashes was very low; totally only 12 crashes during daytime periods and 11 crashes during nighttime periods over the 6-year period of analysis. Intrusion crashes involving construction equipment and those involving construction debris or materials are also higher at night but to a lesser degree. Intrusion crashes still make up only a small proportion of all types of traffic crashes in either daytime or nighttime work zones. The percentages shown in Table 4 are consistent with previously reported trends in work zone intrusion crashes (47, 48). As stated previously, it 11 All Work Zone Traffic Control Types Lane Closure Traffic Control Work Zones Only Key Crash Types Daytime Work Operations, Daytime Crashes (n = 1762) Nighttime Work Operations, Nighttime Crashes (n = 315) Daytime Work Operations, Daytime Crashes (n = 886) Nighttime Work Operations, Nighttime Crashes (n = 182) Rear End Other Multi-Vehicle Single Vehicle Run-Off- Road Intrusion Impacts: with Workers with Equipment with Debris/Other Non-intrusion Impacts: with Workers with Equipment with Debris/Other Impact with TMA Other Miscellaneous Types 49.0% 16.8% 9.4% 0.7% 3.7% 3.1% 0.4% 3.6% 6.8% 2.7% 3.8% 35.6% 14.3% 9.8% 3.8% 4.8% 3.8% 0.3% 6.7% 8.6% 9.8% 2.5% 59.1% 14.8% 5.3% 0.7% 4.7% 4.3% 0.2% 2.5% 4.9% 2.5% 1.0% 45.6% 13.2% 7.1% 3.8% 6.0% 4.4% 0.5% 4.4% 7.7% 4.9% 2.4% Chi-Square Test Results Daytime and nighttime distributions are significantly differenta Daytime and nighttime distributions are significantly differentb a X2 = 118.360 > X2Crit (7, 0.025) = 16.013 b X2 = 21.828 > X2Crit (7, 0.025) = 16.013 Table 4. NYSDOT traffic crash types.
is not possible to determine from the analysis whether the ac- tual risk of such crashes increases at night, or whether the shift in the relative frequencies between daytime and night- time work conditions is instead the result of a lower relative frequency of rear-end crashes associated with nighttime op- erations. Still, the fact that intrusion crashes involving work- ers are a greater proportion of nighttime work zone traffic crashes in general may partially explain why the highway worker perceives night work to be more hazardous. Although night work is associated with a lower percentage of rear-end crashes relative to daytime work activities, the question still exists as to whether the rear-end crashes that do occur at night are more severe because of generally higher traffic speeds. The comparison of day and night operations in Table 5 indicates that a slightly greater percentage of rear-end crashes in night work operations involves fatalities and in- juries, both for all work zones and for those at lane closures. However, the differences are not statistically significant. A sim- ilar comparison of the severity of workspace intrusion crashes between daytime and nighttime conditions is shown in Table 6. Similar trends are evident when all work zone types combined are considered, specifically for lane closures. The percentage of fatalities and injuries associated with intrusion crashes is greater at night than during the day. However, whereas the dif- ference in percentages between nighttime and daytime is sta- tistically significant for all work zone types combined, it is not when only lane closure intrusion crashes are considered. Work Zone Construction Accident Analysis Types of Construction Worker Accidents Table 7 presents the relative frequency of different types of construction worker accidents reported at NYSDOT work zones on freeway and expressway facilities during daytime and nighttime work activities. Overall, the percentages are similar between daytime and nighttime conditions, with no statistically significant differences detected. Accidents involv- ing tools or construction materials are the largest category, followed by strains caused by slipping or tripping and acci- dental contacts with utilities. The percentage of accidents that involve workers being hit by construction vehicles or equip- ment in the work zone is nearly identical for daytime and 12 All Work Zone Traffic Control Types Lane Closure Traffic Control Work Zones Only Injury Severity Daytime Work Operations, Daytime Crashes (n = 863) Nighttime Work Operations, Nighttime Crashes (n = 112) Daytime Work Operations, Daytime Crashes (n = 524) Nighttime Work Operations, Nighttime Crashes (n = 83) Fatal Injuries PDO 0.3% 35.2% 61.9% 2.7% 42.9% 54.4% 0.6% 31.1% 68.5% 2.4% 34.9% 62.7% Chi-Square Test Results Daytime and nighttime distributions are not significantly differenta Daytime and nighttime distributions are not significantly differentb a X2 = 3.307 < X2Crit (1, 0.025) = 5.024 b X2 = 1.622 < X2Crit (1, 0.025) = 5.024 All Work Zone Traffic Control Types Lane Closure Traffic Control Work Zones Only Injury Severity Daytime Work Operations, Daytime Crashes (n = 133) Nighttime Work Operations, Nighttime Crashes (n = 39) Daytime Work Operations, Daytime Crashes (n = 86) Nighttime Work Operations, Nighttime Crashes (n = 26) Fatal Injuries PDO 2.3% 36.8% 60.9% 7.7% 53.8% 38.5%a 2.2% 36.3% 61.5% 0.0% 56.4% 43.6%b Chi-Square Test Results Daytime and nighttime distributions are significantly differenta Daytime and nighttime distributions are not significantly differentb a X2 = 7.419 > X2Crit (1, 0.025) = 5.024 b X2 = 2.728 < X2Crit (1, 0.025) = 5.024 Table 5. Severity of NYSDOT work zone rear-end crashes. Table 6. Severity of NYSDOT workspace intrusion crashes.
nighttime work activities and account for 6 to 7 percent of all construction worker accidents. Severity of Construction Worker Accidents The severity of construction accidents occurring during day- time and nighttime work operations is compared in Table 8. During night work, a slightly greater percentage of construc- tion accidents resulted in injuries than during daytime work, although this small difference is not statistically significant. It is particularly noteworthy that only three worker fatalities resulted from construction accidents over a 6-year period (0.3 percent of 931 reported accidents), and they all occurred in daytime operations. Summary of Findings The following key findings are drawn from the NYSDOT work zone accident database regarding work zones on free- way and expressway facilities, which are most often targeted for night work operations. These findings pertain to both traffic crashes and construction accidents. ⢠About half of daytime work zone traffic crashes and 60 per- cent of nighttime work zone crashes on NYSDOT freeways and expressways occur during traffic lane closures. This statistically significant increase in lane closure crashes at night is probably the result of the higher relative frequency of lane closure operations at night, rather than a higher crash risk. ⢠Overall, there appears to be little difference in traffic crash severity between daytime and nighttime work operations on freeway and expressway facilities. However, worker- involved traffic crashes at nighttime work zones were sig- nificantly more severe than in daytime. It should be noted that worker âinvolvementâ did not necessarily imply that the worker was struck by the vehicle; the higher percentage of severe crashes at night was often drivers or passengers in the vehicle rather than the worker. If a worker was struck, it usually resulted in an injury regardless of whether it oc- curred during the day or at night. ⢠Rear-end collisions comprise a smaller proportion of work zone traffic crashes at night work zone operations than dur- ing daytime operations. This result is consistent with ex- pectations that moving work activities to nighttime hours reduces congestion and queuing that can lead to higher rear-end crash frequencies. Also, working at night does not appear to result in more serious rear-end crashes (when they do occur) than during daytime work operations. ⢠Crashes involving workers, construction vehicles or equip- ment, and construction materials and debris (both intrusion and non-intrusion crashes) comprise a greater percentage of crashes at night than during the day. Intrusion crashes involving workers (those of most concern to agencies and highway contractors) are a higher percentage of crashes at night than during daytime hours. However, they are only a small proportion of the total work zone crash experience in either time period. ⢠Intrusion crashes at nighttime work operations for all types of traffic control combined are significantly more severe than at daytime work operations. A similar trend is evident for lane closure crashes, although the differences between night- time and daytime conditions are not statistically significant. ⢠The types of construction worker accidents occurring at NYSDOT freeway and expressway work zones do not differ significantly between daytime and nighttime operations. ⢠The severity of construction worker accidents does not dif- fer significantly between daytime and nighttime operations. 13 Accident Type Daytime Accidents (n = 931) Nighttime Accidents (n = 114) Falls â Elevated Slip/Trip Strain Accident with Tool or Material Worker Hit by Equipment or Construction Vehicle Other Equipment or Vehicle Accident Utility Contact Other Accident 4.1% 17.6% 30.4% 6.3% 8.5% 26.4% 6.7% 6.2% 18.6% 27.4% 7.1% 13.3% 17.7% 9.7% Chi-Square Test Results Daytime and nighttime distributions are not significantly differenta a X2 = 4.523 < X2Crit (6, 0.025) = 12.592 Table 7. NYSDOT construction worker accident types. Injury Severity Daytime Accidents (n = 931) Nighttime Accidents (n = 114) Fatal Injury None/PDO 0.3% 65.2% 34.5% 0.0% 73.5% 26.5% Chi-Square Test Results Daytime and nighttime distributions are not significantly different a a X2 = 2.846 < X2Crit (1, 0.025) = 3.841 Table 8. NYSDOT construction worker accident severity.