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.
25 This chapter summarizes some of the most relevant studies previously conducted on utility pole safety. In particular, it addresses research that attempted to quantify the effect of various traffic and roadway features on the frequency and severity of crashes into utility poles. It also reviews research on safety appurtenances and the literature on crash prediction. Factors Related to Crash Frequency Several studies attempted to analyze crash databases to identify some of the factors that drive the frequency of utility pole and other fixed-object crashes. Wright and Mak (1976) conducted a study to determine the relationships between single- vehicle fixed-object crashes and the roadway and other variables for urban two-lane streets in Georgia. The study showed that crash rates were most highly related to traffic volume, horizontal alignment, and number of intersections per mile. Perchonok et al. (1978) investigated the relationships between single-vehicle crashes and roadway and roadside features. Data were collected on more than 9,000 single-vehicle crashes on rural roads in six states. Horizontal alignment proved to be a major factor, with more than 40% of the crashes occurring at horizontal or vertical curves. Left curves and downgrades were overrepresented in crashes, but crashes were also overrepresented at the beginning of curves. Cleveland and Kitamura (1978) developed a macroscopic prediction model of roadside crashes on two-lane rural roads in Michigan. During the study, they collected and analyzed data for 270 2-mile (3.2-km) roadway sections with various geometric and traffic features. They developed crash-prediction models for different traffic volume groups. The most important variables for crash prediction were restriction on passing sight distance, frequency of roadside obstacles, and length of road with roadside obstacles within various distances from the road (i.e., pole offset). Fox, Good, and Joubert (1979) developed a crash predictive model to identify risk factors for nonintersection and intersection pole crashes. The variables resulting in the nonintersection crash model included ADT, lateral pole offset, pavement skid resistance, roadway width, hori- zontal curvature, pavement deficiencies, superelevation of the curve, and pole location. Wright and Robertson (1976) conducted a study of 300 fatal fixed-object crashes on rural Georgia roads for consideration in establishing priorities for removal or modification of road- side hazards. The roadway factors most closely associated with single-vehicle crashes were curvature (greater than 6 degrees) and downhill gradient (2% or steeper) before or at the curves. A great majority of fatal crashes also occurred on the outside of the horizontal curve. Jones and Baum (1980) reviewed more than 8,000 single-vehicle crashes from 20 urban areas in the United States, finding that the number of poles along a roadway segment (i.e., pole density C H A P T E R 4 Factors Associated with Utility Pole Crashes
26 Utility Pole Safety and Hazard Evaluation Approaches or pole spacing) was the most important variable in predicting the probability of utility pole crashes. Lateral pole offset from the road represented the next most important feature, followed by road grade, road path (curvature), and speed limit. Mak and Mason (1980) conducted a detailed study of crashes that involved utility poles, sign poles, and light poles in seven geographic areas in the United States. Pole crashes were found to be primarily an urban problem, with 85% of such crashes occurring in urban areas. The overall crash rate in terms of pole crashes per 100 million vehicle miles was 16 (i.e., 9.9 crashes per hundred million vehicle kilometers). Mak and Mason also concluded that the frequency of pole crashes was most highly associated with pole density, pole offset, and horizontal and vertical alignment. Factors Related to Pole Crash Severity Several researchers studied the effects of traffic and roadway variables on the severity of utility pole crashes. Fox, Good, and Joubert (1979) found crashes on horizontal curves to be slightly more severe than those on tangent sections because of the increased number of side impacts on curves. Utility pole crashes were more severe at nonintersections than at intersections, probably the result of lower vehicle speeds at intersections. The Jones and Baum (1980) study indicated that 49.7% of all utility pole crashes caused personal injuries. They observed that impact speeds and pole circumference were related to the severity of utility pole crashes, but the spacing and offset of utility poles did not affect utility pole crash severity. Mak and Mason (1980) reported a 50% chance that at least one vehicle occupant will be injured in a utility pole crash, closely matching the Jones and Baum study results. Of the 1,000 utility pole crashes included in the study, 518 (51.8%) involved one or more injuries, and 16 (1.6%) resulted in one or more fatalities. Vehicle impact speed ranked as a major factor in crash severity; other factors included utility pole type (e.g., wood, metal), presence of yielding poles, vehicle characteristics (e.g., weight, size), and impact configuration (collision location and direction of impact). Griffin (1981) studied single-vehicle crashes in Texas, finding that 44.7% of utility pole crashes involved a personal injury. Furthermore, about 33.5% of such crashes resulted in a moderate injury (B-type injury) or worse, and 5.8% involved a serious injury (A-type injury) or a fatality. In their study of clear zones, Graham and Harwood (1982) uncovered no relationship between an agencyâs clear zone policy (i.e., 6:1 clear zone, 4:1 clear zone, no clear zone) and the severity of fixed-object crashes. Safety Appurtenances Several previous studies addressed the issue of the effectiveness of various crash-related counter- measures, such as placing utility lines underground (and removing the poles), increasing the lateral offset of poles, installing protective barriers (e.g., guardrails), reducing the number of poles, using yielding (breakaway) poles, and employing other options. Each countermeasure is discussed below. Most of the previous studies confirmed that burying utility lines reduces the overall severity of fixed-object crashes, based on the assumption that other less-rigid objects will be hit instead of the utility pole. The net effect is highly dependent on site-specific roadside characteristics, such as roadside slope and the number and type of other obstacles (e.g., trees, mailboxes, and other rigid objects). Some of the challenges encountered in running utility lines underground include
Factors Associated with Utility Pole Crashes 27 the high installation costs and the practice of using many utility poles to also carry attached streetlamps or other related features, along with other utilities. Hunter et al. (1978) suggested that moving poles further from the roadway will reduce fatal crashes, but that approach will not affect the overall crash frequency because vehicles will hit other obstacles after the poles are relocated. Increasing the lateral offset of utility poles is aimed at reducing the chance of a pole being struck by an errant vehicle. Studies by Mak and Mason (1980) and by Fox, Good, and Joubert (1979) identified an overrepresentation of crashes into poles sited within 10 feet of the roadway. In recognition of the possible increase in other fixed-object crashes as a result of pole reloca- tion, Rinde (1979) assumed no overall reduction in the frequency of fixed-object crashes but still concluded that a drop in crash severity likely would occur. Installing roadside hardwareâsuch as guardrails or other impact-attenuation devicesâin front of utility poles offers another technique for potentially mitigating the severity of a crash. Installing guardrails in front of poles will likely increase the frequency of fixed-object crashes because a guardrail would be a larger obstacle than the utility pole, and the guardrail must be placed closer to the roadway than the pole. Roadside Design Guide standards call for some separa- tion between the guardrail and the pole (AASHTO 2011b). Reducing utility pole density can decrease the frequency of utility pole crashes. Treatments that reduce the number of poles (i.e., pole density) include (1) shared use of multiple features on utility poles (i.e., poles that host multiple types of utilities, such as telephone lines, electric lines, and luminaries); (2) poles situated on only one side of the street instead of both sides; and (3) wider pole spacing along the roadway corridor. Jones and Baum (1980) concluded that pole density was the variable most strongly correlated with utility pole crash frequency although they did not quantify the precise impact of reducing the number of poles. One of the practical constraints when reducing the number of poles is the possible need for larger and more rigid poles to support the increased pole spacing and/or heavier utility lines. Thus, any countermeasure that decreases pole density can be costly, and the larger poles could produce an adverse effect on pole crash severity when poles are struck. The countermeasure of yielding poles is directed at reducing the severity of utility pole crashes; the expectation is that such yielding poles would not affect crash frequency. Several yielding pole designs have been developed and evaluated, such as (1) FHWA-approved (AD-IV) steel-reinforced safety poles or fiberglass yielding poles and (2) the steel slip base. Studies confirmed that yielding poles (i.e., steel-reinforced safety poles) can be effective in reducing pole crash severity. Other countermeasures also have been employed to directly or indirectly reduce the frequency or severity of utility pole crashes. For example, Jones and Baum (1980) suggested that the use of occupant restraints (lap belts and shoulder harnesses) likely constitutes the most cost-effective countermeasure for reducing utility pole crash severity. Other proposed indirect methods for minimizing vehicle encroachments beyond the roadway include (1) improved roadway delinea- tion, (2) warning signs in advance of high-risk locations, (3) skid-resistant pavement overlays, (4) widening of lanes and shoulders, (5) rumble strips, (6) enhanced highway lighting, and (7) improved roadway alignment through reconstruction. Research on Utility Pole Crash Prediction In an unprecedented study for FHWA, Zegeer and Parker (1983) sought to determine the factors associated with utility pole crashes. The first phase of the study emphasized assessing the effect of various traffic and roadway variables on the frequency and severity of utility pole crashes. Data for crash and roadway characteristics were collected for more than 1,500 roadway sections
28 Utility Pole Safety and Hazard Evaluation Approaches covering approximately 2,500 miles of rural and urban roads in four states: North Carolina, Washington State, Michigan, and Colorado. Roadway sections in the data sample exhibited various road widths (two-lane to six-lane), terrain conditions, and curbed and uncurbed designs. The AADTs ranged from 1,000 to about 60,000 vehicles per day, with pole densities between 10 to 90 poles per mile and pole offset distances between 2 feet and 30 feet from the roadway. Multiple area types (urban, urban fringe, and rural areas) were included as well as various roadside conditions. The data were analyzed by using several statistical techniques (e.g., correlation, analysis of variance and covariance, and contingency-table analysis) to identify key factors associated with crashes. The following roadway features correlate most strongly with the frequency of utility pole crashes: â¢ AADT â¢ Pole offset â¢ Pole density. The relationship between utility pole crashesâcalled âaccidentsâ in Zegeer and Parker (1983)â and the utility pole offset and pole density is illustrated in Figure 9. The relationship between traffic volume and utility pole crashes is shown in Figure 10. In terms of crash severity when poles were within 10 feet of the roadway, wooden poles exhibited a significantly higher crash severity than metal poles. However, most of the metal poles in the study carried luminaires and featured slip or frangible bases, which are designed to break away on impact. Crash severity also rose higher on roads with greater curvature and in some speed limit categories. Figure 9. Relationship between utility pole crash frequency and the pole offset and pole density.
Factors Associated with Utility Pole Crashes 29 Figure 10. Relationship between utility pole crash frequency and traffic volume (AADT). Linear and nonlinear regression models were developed for utility pole crashes as a function of key roadway features. The best-fit crash-prediction model was based on the following relationship: Acc Mi Yr 9.84 10 ADT 0.0354 Density Offset 0.04 5 0.6 [ ][ ] ( ) [ ]( )( ) ( ) ( ) ( ) = Ã Ã + Ã â â where Acc/Mi/Yr = accidents (utility pole crashes) per mile per year ADT = average daily traffic (average number of vehicles per day) Offset = average distance from the road to the poles (in feet), for all poles in the section Density = number of poles per mile within the section, with poles on both sides of the road counted. The model was validated and displayed satisfactory predictive abilities, specifically an R-squared value of 0.63, a low constant (â0.04), and a low standard error (0.572). The model was verified in several ways, using sections from states that covered a wide range of traffic and roadway conditions. A nomograph (Figure 11), developed based on the utility pole crash-prediction model, enables a simple graphical determination of the expected number of utility pole crashes for various road- way conditions. For example, the estimated number of utility pole crashes per mile per year can be determined for a roadway with an AADT of 10,000 vehicles, a density of 60 poles per mile, and an average pole offset of 5 feet: (1) enter the nomograph at the 10,000 ADT point at the bottom left; (2) proceed up to the curve labeled as 60 poles per mile and then to the right to the 5-foot offset line; and (3) go directly down to the value on the X-axis, which shows 1.15 utility pole crashes per mile per year.
30 Utility Pole Safety and Hazard Evaluation Approaches To illustrate the results of applying the predictive model to a countermeasure such as pole relocation, consider the values of the crash-reduction factors shown in Table 4. The first column lists the utility pole distance in the current situation; the proposed new pole offset distances are shown at the top of each of the remaining columns. To use the chart, find the current offset (first column), and move to the right along the row to locate the proposed offset column. The cell at that intersection notes the expected percent reduction in crashes. For example, if a pole is currently 5 feet from the road, moving it to 10 feet should result in approximately a 56% reduction in crashes. By using the nomograph with various ADT, pole offset, and pole density values, the sensitivity of the model to such factors is easily seen. In the previous example, changing the pole offset from 5 feet to 15 feet, for instance, would reduce the number of crashes from about 1.15 down to 0.55, approximately a 50% reduction in predicted pole crashes. The safety effects of changing combinations of pole offset and pole density can be seen as well. This crash-prediction model was used in the Zegeer and Parker (1983) study to compute the expected crash reductions for various countermeasures related to relocating poles and/or reducing the number of poles exposed to motorists within a roadway section. A series of tables was generated for this report on ARFs for pole relocation and pole density reduction, as given in the utility pole userâs manual by Zegeer and Cynecki (1986). For example, Table 5 in this synthesis report shows the expected ARFs for reducing pole density at a site with an ADT of 25,000 vehicles. Table 5 shows separate calculations for different pole offsets (3, 7, 15, and 25 feet from the road), using different pole densities before and after improvement (10 to 70 poles per mile, in increments of 10 poles per mile). Similarly, Table 6 corresponds to the expected ARFs for increasing pole offset from the roadway edge-line for an ADT of 25,000 vehicles. Table 6 gives separate calculations for densities of 20, 40, and 75 poles per mile, using poll offsets of 2 to 15 feet from the road before improvements and 6 to 30 feet after improvements. The full userâs manual (Zegeer and Cynecki 1986) includes more tables than this synthesis report. Figure 11. Utility pole crash-prediction nomograph (Zegeer and Parker 1983).
Factors Associated with Utility Pole Crashes 31 Table 4. Accident (crash) reduction factors (Zegeer and Parker 1983). Table 5. Accident (crash) reduction factors associated with reducing pole density (Zegeer and Cynecki 1984).
32 Utility Pole Safety and Hazard Evaluation Approaches Table 6. Accident (crash) reduction factors associated with increasing lateral pole offsets (Zegeer and Cynecki 1984). The Zegeer and Parker (1983) study also identified the following several factors associated with the likelihood of serious injuries and deaths for the 9,583 utility pole crashes in the research database: â¢ Pole type. For roadway sections where pole offsets from the road were 10 feet or less, wooden poles (compared to metal poles) were associated with a significantly greater severity of inju- ries and deaths. This outcome is likely because many of the metal poles in the database were luminaire poles with frangible bases that break away when struck. â¢ Horizontal curvature. Utility poles on roadway sections with increasing curvature experi- enced more severe utility pole crashes (when compared to tangent sections) for certain speed limit categories (i.e., speed limits under 35 mph and over 50 mph). â¢ Speed limit. No significant effect was documented between roadway speed limit and crash severity. This outcome possibly resulted from fewer categories of injury severity (PDO, injury, and fatality) compared to previous studies, such as that of Jones and Baum (1980), which analyzed more detailed data on crash severity.
Factors Associated with Utility Pole Crashes 33 In summary, crash frequency was clearly related to ADT, pole offset, and pole density, with lesser factors including type and size of pole, roadway curvature, and roadway type (divided or undivided). Crash severity was most related to pole type and roadway curvature. Table 7 summarizes the relationships between (1) utility pole crash frequency and severity and (2) various roadway features, based on a literature review by Zegeer and Parker (1983). Table 7. Summary of relationships between utility pole crash frequency and severity versus roadway factors (Zegeer and Parker 1983).