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.
NCHRP Project 17-61 3 CHAPTER 1 Background Introduction Work zone safety has been an area of concern to highway agencies, the construction industry, and the traveling public for nearly 40 years. A number of studies have been performed during that time to try and understand how the presence, design, and operation of work zones affect crash frequency and severity, manner of collision, and other crash characteristics. The earliest efforts to understand and predict the effects of work zones on crashes examined changes in crash rates (see Graham et al. 1977, Wang and Abrams 1980, as examples). Although researchers typically found that crashes did increase in work zones relative to pre-work zone conditions, the magnitude of the increases varied widely. Over time, odds ratio analyses using comparison sites and checks for comparability (Ullman and Krammes 1991), poisson/negative binomial regression models (such as Venugopal and Tarko 2000, Khattak et al. 2002, and Chen and Tarko 2012), and other analytical approaches (such as a quantitative risk assessment modeling approach proposed by Meng et al. 2010) were developed and applied. Even with more sophisticated analysis techniques, considerable variation has been found with regards to work zone effects upon crashes. Differences in work zone policies and procedures between agencies and regions likely explain some of the reasons for the inconsistent findings. Low sample sizes and regression-to-the-mean effects, and the inherent randomness in crash occurrence and outcome undoubtedly also contributed to some of the inconsistencies in the findings. Sample size effects are particularly prevalent in work zone safety analyses given the relatively short duration of many work zones, and the frequency at which the configuration of the work zone may change over the course of a project. Taken cumulatively, the studies typically indicate that work zones do result in more crashes relative to pre-work zone conditions. However, the magnitude of that increase has been difficult to predict. Several studies have also examined the effect of work zones on crash severity, utilizing a range of analytical techniques. In that most studies have shown total crashes to increase in work zones, a common research question has been whether the crashes that occur tend to be more severe or less severe than in pre-work zone conditions. Most studies have found that severe (fatal + injury) crashes have increased by a lower amount than have property- damage-only (PDO) crashes (examples include Hargroves and Martin 1980, Richards and Faulkner 1981, and Garber and Zhou 2002) or increased by a similar amount as PDO crashes (see Ha and Nemeth 1995, Lindly et al. 2000, and Qin et al. 2007). Even so, examples also exist where crash severity has increased by a greater percentage than has PDO crashes (such as Burns et al. 1989, Ullman and Krammes 1991, and See et al. 2009). Meanwhile, other studies have strived to identify and evaluate the importance of various contributing factors or other causal indicators to better understand how such crashes can be mitigated. For example, several researchers have examined the spatial distribution of crashes within a work zone. Study results tend to indicate that the activity area is the predominant location for work zone crashes (as indicated by Garber and Zhou 2002, Khattak and Targa 2004, Salem et al. 2006, and Akepati and Dissanayake 2011). However, given that the activity area can often be the longest portion of the work zone relative to the advance warning, transition, and termination segments, such a finding is not unexpected. As might also be expected, a greater proportion of crashes occurring in the advance warning area tend to involve rear-end collisions, whereas sideswipe collisions tend to occur most frequently in the transition area. Overall, the literature indicates that rear-end collisions typically increase the most and are often the predominant type of work zone crash (as illustrated by Rouphail et al. 1988, Hall and Lorenz 1989, Wang et al. 1995,
NCHRP Project 17-61 4 Daniel et al. 2000, Mohan and Gautam 2002). More recently, these increases in rear-end collisions have been shown to vary by type of work activity, time-of-day, and traffic volumes (Ullman et al. 2008). However, despite the amount of previous research performed, it is difficult for practitioners to objectively incorporate the safety effects of various work zone design alternatives, operating strategies, and countermeasures into work zone impacts assessment and transportation management plan (TMP) development. For instance, the AASHTO Guide for Reducing Work Zone Collisions lists nearly 40 treatments and strategies that are believed to have the potential to reduce work zone collisions (Antonucci et al. 2005), but quantitative guidance as to the level of effectiveness of any of these treatments/strategies in a work zone environment is extremely limited. This makes it difficult for practitioners to effectively assess the cost versus safety tradeoffs of work zone design element choices, maintenance of traffic alternative analyses, or the provision of specific crash mitigation strategies. Project Objectives The objective of NCHRP 17-61 was to develop new information and comprehensive guidance on the characteristics of work zone crashes and the effectiveness of certain engineering, enforcement, education, emergency services, and public policy countermeasures intended to reduce work zone crash frequency and severity. To accomplish this, the following major efforts were undertaken: ï· Assessment of data and findings from previous work zone safety study results to quantify, where possible, the effect of various work zone features, operational strategies, and potential safety countermeasures upon work zone crashes in terms of crash modification factors (CMFs); ï· Assessment of existing non-work zone CMFs already developed for the AASHTO Highway Safety Manual (HSM) or included in the Crash Modification Factor Clearinghouse for their potential suitability for applying them to work zone situations; ï· Mining of detailed crash databases for additional insights into work zone crash characteristics and work zone-related contributing factors; ï· Analysis of the effects of multiple work zone design features upon crashes using a multi-state dataset of work zone projects on interstate, freeway, and multi-lane highways developed under this research effort; and ï· Development of implementation guidance on how to utilize the information developed through this research effort to perform work zone safety impact assessments of alternative design decisions, operating strategies, and countermeasure deployments. The results of the efforts regarding the first, second, and last bullet above have been packaged into a stand-alone implementation guide separate from this document. This report specifically addresses the results of the third and fourth bullet above. Contents of this Report This technical report documents the methodology and results of the various research activities performed on this project. Following this introductory chapter, Chapter 2 documents the results of analyses performed on detailed work zone crash report diagrams and narratives extracted from three existing crash databases: the Virginia Department of Transportation (VDOT) Roadway Network System (RNS) Crash database, the National Highway Traffic Safety Administration (NHTSA) National Motor Vehicle Crash Causation Survey (NMVCCS), and the combined NHTSA/Federal Motor Carrier Safety
NCHRP Project 17-61 5 Administration (FMCSA) Large Truck Crash Causation Study (LTCCS). Work zone crashes identified in each of these datasets were examined in detail to better define and understand the work zone-related contributing factors and sequences of events that resulted in those crashes. Next, Chapter 3 presents the data, methodology, and results of an analysis of work zone crashes correlated to traffic operating conditions (specifically, periods of queuing and non-queuing) during nighttime temporary lane closures in Texas. Also included in this chapter is an evaluation of the crash- reducing effect of two types of work zone crash countermeasures, work zone ITS-based end-of-queue warning systems, and portable rumble strips (PRS) deployed transversely upstream of the temporary lane closure. Chapter 4 provides documentation of an effort to develop predictive statistical models of work zone crashes as a function of several key design variables (lane width, shoulder width, barrier presence, and lane shifts) using a database of work zone project plans, traffic volume, and crash data from a sample of work zones across the U.S. The intent of the analysis was to establish useful work zone crash modification factors (CMFs) describing the proportional change in crashes attributable to the various factors. These CMFs could then be used when performing trade-off analyses of alternative work zone traffic-handling options and strategies for a particular project on a particular roadway segment. Finally, Chapter 5 presents a summary of the key findings from these efforts and provides recommendations for additional research. A separate implementation guide describing the concept of work zone safety impact assessment, available CMFs, and estimation procedures, has also been produced as part of this research effort.
