Skip to contentUnited States Department of Transportation - Federal Highway AdministrationSearch FHWAFeedback



Toll Facility Safety Study Report to Congress

Appendix B - Literature Review


There are many papers and reports that take an in-depth look at toll facilities and their operational characteristics, but few studies exist that focus on safety at toll facilities. Those studies that do exist and presented here according to the following four areas:

  • Safety Conscious Planning in Toll Facility Design.
  • Driver and Occupant Safety in the area of Toll Facilities.
  • Traffic Volume at Toll Plazas and the Impact on Safety.
  • Toll Collector Health and Safety.

Safety Conscious Planning in Toll Facility Design

Adequate toll facility design can have an impact on the safety of toll facilities once constructed and operational. Several papers and reports reviewed by the project team describe case studies and state-of-the-practice for toll facility design. In recent years, many of the studies deal with effective design for facilities with a combination of conventional and ETC lanes.

In National Cooperative Highway Research Program (NCHRP) Synthesis 240 Toll Plaza Design (Schaufler 1997), various practices are described related to the design, current practice, and use of ETC. One of the major considerations of safety design is the use of safety devices on the approach to toll islands. Devices include: concrete-filled steel bollards or I-beams, concrete ramparts, barrier shapes, concrete crash blocks, impact attenuators, and frangible devices (such as sand barrels). Toll agencies recognize that the purpose of the device is to protect the toll attendants and redirect vehicles back into the toll lane and away from the toll plaza structure. The synthesis report also describes the importance of appropriately designed traffic control devices since a lack of consistency can be confusing to new patrons and older drivers. Inadequate traffic control devices can also lead to slower processing of the tolls as well as abrupt maneuvers in a plaza. The author states that toll plazas should have consistent signing, marking, and signal practices to encourage consistent behavior.

A paper by McDonald and Stammer (2001) provides a contribution toward guidelines for toll facility design. The authors state that safe toll facility design is more critical than ever when ETC lanes are implemented. Safety considerations mentioned include the safety of drivers who stop to pay a toll using conventional lanes, drivers who proceed without stopping when using ETC lanes, and toll plaza workers. Various qualities of toll facilities change depending on the agency as well as the individual facility including lane configurations, reversible lanes, taper rates, transition lengths, lane width, and vertical geometrics.

With respect to lane configurations, toll facilities can contain a combination of toll payment options including full service booths, automatic coin machines, and ETC. These ETC lanes can be either express or dedicated depending on the facility. Care should be taken in the selection of speed limits for ETC lanes where speeds can range from 5 to 45 mph for dedicated ETC lanes and up to 60 mph or more for express ETC lanes. The use of reversible lanes is fairly common especially in areas where there is a directional effect such as morning versus evening rush hour.

McDonald and Stammer recommend that reversible lanes not be used with ETC express lanes unless the speed of the nonstop traffic is drastically reduced. A better option may be to place the high-speed lanes on the outside (or right side) of the toll facility, but the authors suggest that this would be inconsistent with the usual design of multi-lane roadways, in which faster traffic normally uses the left lanes.

McDonald and Stammer also found taper rates to be very different when comparing the transition between the roadway and the toll facility. There are two tapers, one on the approach and a second for the departure or merge. The authors suggest a modification to existing taper rate equations for the area where vehicles are anticipated to stop for toll payment. The authors suggest that at 40 mph or less, the approach taper equation should be:

Equation. Length equals the offset distance in feet times the posted speed limit squared divided by 105.

where L is the minimum length of taper in feet, S is the posted speed in mph, and W is the offset distance in feet. For 45 mph or more, the approach taper equation should be:

Equation. Length equals three-eighths the product of the offset distance times the posted speed in miles per hour.

Once vehicles have paid their toll, the authors suggest that the merge taper equation should be:

Equation. Length equals 1.5 times the sum of the offset distance in feet times the posted speed limit squared divided by 105 plus 5 times the offset distance in feet.

for exit speeds of 40 mph or less. When speeds leaving the plaza are more than 40 mph, as in the case for ETC lanes where traffic does not stop for the toll plaza, conventional taper equations are recommended.

Transition lengths include both the queue area length prior to entering the toll plaza and the recovery zone after leaving the toll plaza. Appropriate distances need to be chosen to allow vehicles to decelerate and get back up to speed when leaving the toll plaza.

Transition lengths that are too short can cause conflicts and should be avoided. In general, the queue area was found to range from 75 feet to over 1000 feet and the recovery zone was found to range from 30 to 650 feet, depending on the facility. The authors state that the queue area will depend on demand, but the recovery zone should be a minimum of 75 feet in length to allow for slow deceleration of trucks.

McDonald and Stammer acknowledge that lane width at the toll plaza will depend on the lane type, but they recommend that wide load lanes meet State regulations, ETC express lanes be 12 feet wide with shoulders, ETC dedicated lanes be 12 feet wide, lanes with high truck traffic (≥ 30 percent) be 12 feet wide, lanes with substantial truck traffic (>10 percent but <30 percent) be 11 feet wide, and all other lanes be 10 feet wide, as a minimum.

McDonald and Stammer also discuss vertical geometrics in the areas of cross slopes and sight distance. The authors recommend that the cross slope for express ETC lanes be the same as that of the adjacent roadway to remove excess water and avoid icy conditions, and that a cross slope of 1.5 percent to 1.6 percent should be used on the approach, at the plaza, and on the departure. The authors also explain the importance of stopping and decision sight distance at toll plazas and thus urge the reader to consult the AASHTO Green Book for sight distance considerations and equations.

A recent report by FHWA, conducted by Wilbur Smith Associates (2006), State of the Practice and Recommendations on Traffic Control Strategies at Toll Plazas, also provides general information on toll plaza design and traffic control devices. The report describes several recommendations dealing with safety design at toll facilities. The first section addresses design in the approach zone and recommends the use of the transition zone taper values proposed by McDonald and Stammer (see above). Additionally, the authors suggest that if an upstream interchange is located in a way that drivers cannot change lanes quickly enough to reach the ETC lanes, then the maneuver should be physically prevented through barriers. A third recommendation dealing with the approach zone is that sensors should be placed at toll lanes to prevent oversized trucks from entering the lane.

The second section on safety design deals with the departure zone. The authors recommend that the transition zone taper values proposed by McDonald and Stammer be used and that recovery zones should be long enough to allow sufficient driver reorientation, acceleration, and initial merge. The authors again recommend that in departure zones, if a downstream interchange is located so close to the toll plaza that ETC users cannot safely change lanes to make the exit, then the movement should be restricted through physical barriers. The authors also suggest that when dedicated ETC lanes are used, a physical barrier should separate the dedicated lanes from other toll traffic until the other traffic has accelerated to two-thirds of the operating speed.

The report addresses various other safety design issues as well, suggesting, for example, that when express lanes are used, they should be located to the far left of the plaza for consistency and that they should be designed with shoulder and lane width characteristics that are similar to the mainline approach. The report also advises the use of lighting for safety in which the intensity levels and uniformity ratios should be based on the American National Standards Institute (ANSI) and Illuminating Engineering Society (IES) values.

A case study by Mohamed et al (2001) describes safety considerations related to the design of electronic toll plazas. The authors define five main lane types that are used at toll facilities (a classification similar to the one McDonald and Stammer describe) which are: manual, automatic, mixed, dedicated automatic vehicle identification (AVI), and express AVI. During a period of 3½ years from 1994 to 1997, data was analyzed for the Orlando-Orange County Expressway Authority in Florida. The data show that 32 percent of the crashes occurred at their 10 mainline toll plazas. The monthly crash rate also significantly increased when comparing the rate before ETC was added and after ETC was added (from 3.4 to 7.5 crashes per month). The authors identify several conflict points and dangerous behaviors including merging, queuing, and speeding vehicles. Some suggested reasons that ETC caused an increase in crashes could be driver unfamiliarity with the system, toll plaza configuration, and the possible speed variance between drivers that use ETC and drivers that do not. Solutions mentioned include increasing the width of the ETC lanes, arranging similar lane types within the plaza, providing more advance signing, adding variable message signs to show payment methods available and status of lanes, and an extensive use of pavement markings. Pavement markings could be used to reduce speeds, discourage weaving and lane changes, reduce driver confusion, and reduce conflict points.

In 1990, Zilocchi described a success story on the Garden State Parkway dealing with the addition of branch toll lanes to improve safety and operations at toll plazas. This practice is used by several toll agencies when expanding toll plazas to create more toll lanes. Traditionally, toll plazas were expanded by adding lanes horizontally, which caused issues with increased right-of-way, construction time needed, disruption of traffic, opposition from the community, and cost. Branch lanes are used to add toll lanes either ahead of or after the existing lanes as a branch of the mainline plaza. Zilocchi stated that safety was a prime concern and thus the safety aspects were carefully monitored. One issue deals with motorists traveling too fast on the approach to the branch lanes, and thus signing was used to successfully cut down on approach speeds. Of additional concern was toll worker safety, and thus the New Jersey State Police were asked to carefully monitor the branch toll lanes due to an increased danger of robbery or other inappropriate action with tollbooth personnel in the branch lanes. Other concerns included protecting collectors from being injured by vehicles, and thus impact attenuators and sidewalks with steel guiderails were added to make the branch lanes as safe as the traditional lanes. Zilocchi reported that as of publication of the article, there were not any serious accidents in the branch lanes and that there did not seem to be any reason to think that the lanes were inherently unsafe as designed.

Driver and Occupant Safety in the Area of Toll Facilities

In 2003, Mohamed Abdel-Aty (2003) published a paper analyzing driver injury severity levels and what factors had the most influence. In the paper, models were developed at roadway sections, intersections, and toll plazas in Central Florida. The toll plazas analyzed were on the Central Florida expressway system on state roads 408, 417, and 528 for a total of 79 miles. Ten mainline toll plazas and 42 ramp tolls are located in the system. The ETC was added starting in 1994 and the installations were completed by 1998. Police reports were obtained in 1999 and 2000 to analyze the results.

Variables collected fell into the categories of crash, driver, vehicle, plaza, and roadway. Crash-related factors included age, gender, driver license type, alcohol involvement, driver violation, presence in an ETC lane, and ETC user. Vehicle factors included vehicle type, point of impact, number of impacts, and speed ratio (as compared with the posted limit). Plaza factors included mainline versus ramp, and roadway factors included weather condition, lighting condition, and time and day information. Of 1,932 total crashes on the system in 1999 and 2000, 447 crashes (23.1 percent) occurred near a toll plaza with 803 drivers involved. The type of crashes were: rear-end (40.1 percent), sideswipe (26.5 percent), fixed object (21.3 percent), backed-into (6.4 percent), and other (5.7 percent). As for data that indicated crash location relative to the toll plaza (725 vehicles), 43.6 percent of the accidents were when the vehicle was approaching, 43.7 percent were when the vehicle was at the plaza, and 12.7 percent were when the vehicle was leaving the plaza. When looking at injury severity, 61.0 percent had no injury, 21.0 percent had possible injury, 15.3 percent had evident injury, and 2.8 percent had a severe or fatal injury.

The results showed that older drivers, male drivers, and occupants not wearing a seat belt had a higher probability of severe injury. Additionally, drivers of passenger cars, vehicles struck on the driver's side, and vehicles that were speeding experienced higher severity injury. Additionally, it was found that vehicles with an ETC transponder had a higher injury rate to those without the transponders. This may occur when a vehicle without a transponder stops in the lane with fast-moving ETC transponder users.

Another study conducted in Florida by Abdelwahab and Abdel-Aty (2002) looked at the same data as the Abdel-Aty paper described previously and compared the use of logit models and artificial neural networks to analyze the safety at toll plazas. The authors state that an ETC transponder user is 11 percent more likely to be involved in a crash than nonusers. The results showed that a neural network approach was a better model to predict crash severities, but more important to this topic are the recommendations made after running the models. The first suggestion is that improvement is needed in lane markings at the approach to toll plazas. The second suggestion states that traffic signs should be used to deter motorists from stopping in ETC lanes. The authors also suggest that warning signs be placed downstream of plazas located at off-ramps to show that a toll plaza is located on the ramp. Additionally, the authors recommend that ETC lanes be made wide enough to accommodate heavy trucks that use the ETC system. Finally, it is suggested that the approach zone and toll plaza structure should be illuminated to enhance visibility.

Traffic Volume at Toll Plazas and the Influence on Safety

There is a scarcity of published research dealing with the effect of traffic volume on safety at toll plazas; however, a paper by Chang et al. (2000) investigated how volume-to-capacity (v/c) ratios affected safety on a freeway section with two mainline toll booths. The study section included a four-lane freeway with four interchanges, two toll gates, and two tunnels. It was found that the accident rate of the toll facility section was higher than the other sections. The overall relationship of crashes represented a U-shaped curve with v/c ratios from 0.1 to 0.8 and an accident rate from 90 to 2,714 crashes per million kilometers traveled. When analyzing the curve, the fewest crashes occurred at the toll facility section when the v/c ratio was 0.57. The authors do state, however, that the model did not fully explain the relationship and that additional explanatory variables should be considered.

Toll Collector Health and Safety

A comprehensive review of toll collector health and safety was performed by Szeinuk et al. (2006) for the International Brotherhood of Teamsters. As identified in other reports, the authors state that ETC systems may contribute to the health and safety risks of toll collectors. Overall, the report identified several concerns to toll collectors. Outside the tollbooth, these concerns include:

  • Crossing the toll plaza due to poor design, lack of signs, lack of tunnels, and danger of trips and falls.
  • Poor maintenance of plaza facilities.
  • Motorist issues including disregard for signs, speeding, failure to stop, and dangers of tractor-trailers.
  • Lack of enforcement resulting in a vulnerability to security.

Inside the tollbooth, risks to toll collectors include:

  • Lack of protection
  • Uncomfortable uniforms
  • Ergonomic concerns (i.e., uneven floors, uncomfortable seats, fixed window height, door jamming, poor lighting, and repetitive motion)
  • Inadequate air conditioning and heat
  • Respiratory issues
  • Noise-induced issues
  • Motorist behavior from rudeness to road rage
  • Poor sanitation
  • Sense of insecurity
  • Understaffing
  • Respiratory and cardiovascular diseases
  • Musculoskeletal injury
  • Cancer
  • Noise-induced hearing loss
  • Chemical exposure
  • Vehicle-pedestrian accidents
  • Homicide, assault, and violence
  • Slips and falls
  • Mental health, job strain, and stress

Several remedies are suggested by the author, including minimizing exposure of tollbooth collectors to vehicular exhaust and noise; improving ergonomics of toll collection facilities; protecting collectors from injury, violence, and slips; and providing an adequate work frame to minimize the risks for strain and shift-work-related disease.

A paper by Feist et al. (2001) detailed the results of a survey of tollbooth agencies regarding traffic noise and the performance of a headset to increase noise control. The surveys were conducted at the Portage Barrier Toll Collection Plaza located west of Portage, Indiana. Sources of noise included emissions from trucks and motorcycles, idling engines, and accelerating vehicles, which are all low frequency sounds. The employees identified that the presence of noise caused discomfort, made communication with patrons difficult, and caused fatigue. The survey showed that a properly configured and comfortable active noise control headset would be acceptable to employees. One caution noted by the authors, though, is that the device may be confused by motorists as a source of entertainment rather than a noise reduction device. The authors add that this may be overcome with a professional-looking design.

The Wilbur Smith Associates report on State of the Practice and Recommendations on Traffic Control Strategies at Toll Plazas emphasizes that safe access be given to toll collectors at toll facilities. Specifically, the report recommends that toll collectors should never have to cross an ETC lane to reach their plazas and that, preferably, tunnel or overhead access should be given for all lanes. The authors recognize that tunnel or overhead access can not be provided for every toll booth, but it should be designed that toll collectors would not have to cross more than one toll lane to get to their assigned booth.

The NYS Thruway Authority contracted with Vollmer Associates in 2000 to perform a Toll Lane Safety Study Report. This study was undertaken following a fatal pedestrian accident in 1999, and the purpose of the study was to analyze the issue of toll collector safety operations in light of the accident and to determine whether any additional measures could be effective and practical in improving the safety of toll collectors crossing active toll lanes. This study included an industry review, a summary of practices at other facilities, and a review of measures noted in the citation issued by the New York State Department of Labor. It also included information about how 8 different toll authorities deal with 16 different safety practices. Six of the eight had safety programs in place or under development. Seven of the eight had formal lane crossing procedures with requirements ranging from wearing safety vests, to holding up a hand or stop sign paddle, to making eye contact with the driver, to verbally asking a vehicle to stop while paying a toll and notifying them to wait until the toll collector crosses in front of them. Various traffic control devices are also used to show crosswalks and close toll lanes. These devices include traffic lights, stop signs, striping, gates, and traffic cones. Agencies primarily use gates for safety as well as violation deterrence. Five agencies had data available on crashes. Four of the five did not have any injuries dealing with toll collector/vehicle crashes. One agency had one instance of a toll collector/vehicle injury crash in which a vehicle with a transponder was at an ETC lane and then changed lanes to a staffed lane and failed to stop in the staffed lane even though a stop sign was present.

The NCHRP Synthesis 240 also describes security issues that occur at toll facilities. Access cards are increasingly being used at toll administration buildings to monitor the individuals who are entering and leaving the facility as well as to deny access to unapproved individuals. Various devices have also been used at toll booths, including surveillance cameras, silent alarms, bulletproof toll booth glass, and bill alarms.


Abdel-Aty, M.A. "Analysis of driver injury severity levels at multiple locations using ordered probit models." Journal of Safety Research, vol. 34, pp. 597-603, 2003.

Abdel-Aty, M.A. and H.T. Abdelwahab. "Artificial Neural Networks and Logit Models for Traffic Safety Analysis of Toll Plazas." Transportation Research Record, No. 1784, pp. 115-125, 2002.

Chang, J., C. Oh, and M. Chang. "Artificial Neural Networks and Logit Models for Traffic Safety Analysis of Toll Plazas." Fourth International Symposium on Highway Capacity. Transportation Research Circular E-C018, Transportation Research Board, 2000.

Feist, J.P., L. Mongeau, and R.J. Bernhard. "Tollbooth Operators' Response to Traffic Noise and the Performance of an Active Noise Control Headset: Survey Results." Transportation Research Record, No. 1756, pp. 68-75, 2001.

McDonald, Jr., D.R. and R.E. Stammer, Jr. Contribution to the Development of Guidelines for Toll Plaza Design. Journal of Transportation Engineering, pp. 215-221, May-June 2001.

Mohamed, A.A., M.A. Abdel-Aty, and J.G. Klodzinski. "Safety Considerations in Designing Electronic Toll Plazas: Case Study." ITE Journal, pp. 20-24, March 2001.

Schaufler, A.E. Synthesis of Highway Practice 240: Toll Plaza Design. Transportation Research Board, 1997.

State of the Practice and Recommendations on Traffic Control Strategies at Toll Plazas. Unpublished report by Wilbur Smith Associates for Federal Highway Administration, 2006.

Szeinuk, J., K. Krauskopf, R. Herbert, J. Dropkin, and M. Goldberg. "The Health and Safety of Toll Collectors." Mount Sinai-Irving J. Selikoff Center for Occupational and Environmental Medicine, New York, January 2006.

Toll Lane Safety Study Report. Unpublished report by Vollmer Associates LLP for the New York State Thruway Authority, July 2000.

Zilocchi, G.P. "Branch Toll Lanes: A Success Story at the Garden State Parkway." ITE Journal, pp. 21-24, September 1990.

More Information


Bryan Cawley
Office of Asset Management, Pavements, and Construction
E-mail Bryan

Updated: 04/07/2011

United States Department of Transportation - Federal Highway Administration