Make your own free website on

Conclusion of MMB Study

                      NEWS RELEASE   

November 25, 1997
Contact Public Relations: 
Mary C. Currie GGB, H&TD (415) 257-4548



    A report detailing the traffic safety study undertaken for a proposed one-foot-wide Movable Median Barrier developed by. Barrier Systems, Inc. (BSI) of Carson City, Nevada, for possible use on the Golden Gate Bridge has been released for public review. The report was prepared for the "Golden Gate Bridge, Highway and Transportation District (District) by the Traffic Institute of Northwestern University (Northwestern), Evanston, Illinois.

     The report is being made available to the general public through the Office of the District Secretary and local libraries. A list of report repositories is attached.  The report will be presented to the Building and Operating Committee of the District Board of Directors on Friday, December 5, 1997 at 10:00 am in the Board Room, Toll Plaza Administration Building, Golden Gate Bridge. 

    The results of the crash certification testing will be presented made by E-Tech Testing Services. Inc. Dr. Robert K. Seyfried, P.E., Director Transportation Engineering Division will present the Northwestern study findings. A 3-D visualization, created by AutoDesk, will also be presented in video format of a car crossing the Golden Gate Bridge in a lane adjacent to the barrier.

     In September 1996, the District retained services from the Northwestern Traffic Instruction to conduct an in-depth and detailed traffic engineering safety analysis of the "New One-Foot" BSI of Carson City, NV.  The purpose of the analysis was to evaluate whether the new one-foot barrier's installation is warranted by the anticipated trade-offs, namely, the barrier's potential benefit of preventing all cross-head-on collisions accidents, versus the anticipated potential detriments of increasing other types of accidents such as collision with the barrier, reduction in speed, reducing traffic capacity and the reality low costs of purchasing, operating, and maintaining the BSI barrier. Cost estimated at $ 6,800,000.00

     In March 1997, crash certification testing was held in Lincoln, Ca. in an open-field area, (abandon air strip) on the one-foot wide barrier system.  Certification was performed pursuant to National Cooperative Highway Research Program report No. 350. The Bridge District contributed $42,500 towards the cost of this testing, which was performed by E-Tech Testing Services, Inc. E-Tech submitted its report in May 1997. Cal-Trans reviewed the testing procedures in consultation with the Federal Highway Authority and approved the crash testing of the barrier. Northwestern Traffic Safety Study Report may be reviewed at the following repositories after December 1, 1997  


To obtain a copy of the certification testing an E-Tech contact the Office of the Bridge District,  Janet  S. Tarantino/Secretary to the Golden Gate Bridge: Toll Plaza Administration Building, Monday through Friday, 8:30 am to 4:30 p.m. Also, you may obtain a copy at your local libraries, listed below...

Belvedere-Tiburon Public Library, 6 Beach Road, Tiburon, CA 94929
Corte Madera Public Library, 707 Meadowsweet Drive, Corte Madera, CA 94925
Fairfax Public Library, 2097 Sir Frances Drake Boulevard, Fairfax, CA 94930
Inverness Public Library, 15 Park Avenue, Inverness, CA 94937
Larkspur Public Library, 400 Magnolia Avenue. Larkspur, CA 94939
Marin City Public Library, 164 Donahue Street. Marin City, CA 94965
Marin County Library, 3501 Civic Center Drive, San Rafael, CA 94903
Mill Valley Public Library, 375 Throckmorton Avenue, Mill Valley, CA 94941
Novato Public Library, 1720 Novato Blvd., Novato, CA 94947
Petaluma Public Library, 100 Fairgrounds Drive, Petaluma, CA 94952
Pt. Reyes Public Library, 4th & A. Point Reyes Sta., CA 94956
San Anselmo Public Library, I 10 Tunstead, San Anselmo, CA 94960
San Francisco Main Library Civic Center, Larkin & McAllister. San Francisco, CA 94102
San Geronimo Valley Public Library, PO Box 423, San Geronimo, CA 94963
San Rafael Public Library, I 100 E Street, San Rafael, CA 94901
Sausalito Public Library, 420 Litho St, Sausalito, CA 94965
Sebastopol Public Library, 7140 Bodega Hwy, Sebastopol, CA 95472
Sonoma County Library, Rohnert Park/Cotati, 6600 Hunter Rohnert Park, CA 94928
Sonoma County, Public Library, 3rd & E Street, Santa Rosa, CA 95401
Sonoma Public Library, 755 West Napa Street, Sonoma, Ca 95476
Stinson Beach, Public Library, 3470 Shoreline Hwy, Stinson Beach, CA 94970

Movable Median Traffic Barrier Safety Report/Study on the Movable Median Barrier  needed On the Golden Gate Bridge.

Prepared for the Golden Gate Bridge, Highway and Transportation District
By the Northwestern University Traffic Institute CONCLUSIONS
PO Box 1409   Evanston, IL 60204  October 29, 1997


The following is a summary of conclusions regarding the anticipated impacts of a movable median barrier system on safety and traffic operations on the Golden Gate Bridge and its approaches:

  • 1.     The current accident rate up to October 29, 1997,on the Golden Gate Bridge is 0.64 accidents per million vehicle-miles of travel (123,000 per day) which is significantly less than would normally be expected for this type of roadway facility. This current accident rate is also approximately one-half of the accident rate on the Bridge prior to the improvement project which widened the roadway width to 62 feet-accident rates on weekends are significantly higher than the overall accident rate. The presence of stopped or slow vehicles is a significant contributing factor in accident occurrence under normal driving conditions. Forty-five (45) percent of the accidents on the Bridge involve injuries and fatalities. In a five-year period, 24 accidents occurred in which one or more vehicles crossed the dividing line into oncoming traffic, which produced a head-on collision. These crossover accidents had a higher average severity, with 69 percent involving injuries or fatalities.

    2.      Accident experience on several comparable facilities suggest that there remains a significant concern that overall accidents as well as injury accidents could increase on the Golden Gate Bridge if a movable median barrier system is not  installed. However, the most directly applicable accident experience (Auckland Harbor Bridge and San Diego- Coronado Bridge) suggest that overall accidents as well as injury and fatal accidents would be reduced if a movable median barrier system were installed (which have installed barriers now). The observed total accident frequency reductions after installation of movable median barrier systems on these bridges actually exceeded the number of crossover accidents that were eliminated by the barrier systems, but there are no more head-on collisions.

    3.      No increases in non-crossover injury and fatal accident frequency were observed on either bridge. in fact a decrease in non-crossover injury and fatal accidents occurred on the Auckland Harbor Bridge. It is not intuitively obvious why such decreases in non-cross over accidents would occur after installation of a movable median barrier system. Nevertheless, the current data from the Auckland Harbor Bridge and the San Diego-Coronado Bridge indicate the likelihood of secondary safety benefits from the movable median barrier system in reducing non-crossover accidents. It must be recognized that this contradictory data increases the uncertainty in estimating the safety implications of installing a movable median barrier system on the Golden Gate Bridge. Nevertheless, it appears that the most likely outcome of installation of a movable median barrier on the Golden Gate Bridge would be a small reduction in property damage accidents (0 to 30%), an elimination of most fatal accidents, and a 20% to 40% reduction in injury accidents.

    4.      The Narrow Quickchange Movable Barrier system appears to satisfy all of the desired performance criteria for application on the Golden Gate Bridge except for maximum lateral deflection which is somewhat more than the desired criterion of 30 inches. Because relatively few barrier impacts would be expected to exceed this 30-inch criterion in actual practice, the Narrow Quickchange Movable Barrier system may be considered marginally acceptable in this regard. If a movable median barrier system were to be installed on the Golden Gate Bridge, the Narrow Quickchange Movable Barrier system would be preferred over the standard Quickchange Movable Barrier system. No other movable barrier systems are known to meet desired criteria. However, if any other systems become available which satisfy appropriate performance criteria, they also should be considered for potential application.

  • 5.     The south end of a movable median barrier system should be terminated in the toll plaza area, north of the toll booths, anchored, and treated with an acceptable crash cushion. The north end of a movable median barrier system should be extended beyond the end of the existing freeway median barrier and terminated within the northbound freeway roadway.

  • 6.     Through the horizontal curves on the San Francisco and Marin approaches, a movable median barrier system would restrict the ability of drivers in the adjacent lanes to see obstacles in their path. This sight restriction would limit the design speed on these curves to 30 to 35 mph based on current highway design criteria. As a practical matter, it is unlikely that drivers would reduce their speeds to 30 to 35 mph on these curves in response to warning or regulatory speed signs.

  • 7.     For lanes adjacent to a movable median barrier system that are maintained as 10 or more feet in width, no significant changes in lateral positions of vehicles within the lanes are likely, and no decreases in traffic carrying capacity of the roadway are likely. The installation of a movable median barrier system would result in the loss of one or more feet of overall roadway width. If implemented, the barrier can and should be positioned such that the minimum width of any pair of lanes is at least 20 feet Desirably, it should be positioned such that the existing 11-foot curb lane widths are also retained. None of the barrier positioning alternatives examined in this study were entirely satisfactory; all required the use of one or more lanes which were less than desirable widths.

  • 8.     A movable median barrier system would result in reduced speed, efficiency, and flexibility in responding to and removing traffic accidents or stalled vehicles on the Bridge. This would result in longer periods of congestion and increased chances for accidents related to slow or stopped traffic.

  • 9.     The outcome of a benefit-cost analysis of the potential installation of a movable median barrier system on the Golden Gate Bridge is highly sensitive to the value of accidents avoided and to the annual operating and maintenance costs for the system. As a result, the benefit-cost analysis does not provide a clear-cut answer concerning economics of investments in such a movable median barrier system. Rather, the benefit-cost analysis should be considered one factor, among many, that must be evaluated by the Golden Gate Bridge, Highway and Transportation District in deciding on a course of action.

* A number of important changes have occurred since the Traffic Safety Study for a Proposed Movable Barrier on the Golden Gate Bridge 
(Ref. 1) was prepared by the Northwestern University Traffic Institute in 1985.  These include:

1.     Movable barrier systems have been installed on a number of permanent and temporary applications throughout the world, and useful accident data from some of these installations is available.

2.     New technology and refinements in previous designs have eliminated several characteristics of candidate movable barrier systems and end treatments that were considered objectionable in the previous study.

    We continue to have reservations about several negative impacts of a movable median barrier system on the Golden Gate Bridge. Among these concerns are the affect of a barrier system on driver sight distance, the potential for secondary impacts due to barrier deflection and vehicle rebound, the loss of potential refuge for disabled vehicles in the buffer lanes and consequent potential for rear-end collisions, and reduced speed and efficiency of emergency vehicle response. The benefit-cost analysis performed does not yield a clear-cut indication of the economic desirability of such a movable median barrier system. 

    Nevertheless, the movable median barrier systems in use in several permanent installations around the world have generally been successful in virtually eliminating head-on collisions and have not generally increased other accidents. Although the Golden Gate Bridge continues to enjoy an enviable traffic safety record, increasing severity of accidents in recent years is a concern. If the Golden Gate Bridge, Highway and Transportation District elects to install such a movable median barrier system, the Narrow Quickchange Barrier system appears to best satisfy desired performance criteria for application on the Golden Gate Bridge.


IV. Issues Requiring Further Study: 
(Committee of the whole item No. 2 for January 16, 1998/ page 10)
Several important issues remain to be resolved, if the Golden Gate Bridge, Highway and Transportation District elects to move forward with the installation of a movable median barrier system on the Golden Gate Bridge. These include:

1.  Lateral positioning of the barrier system: Several alternative barrier positioning schemes were evaluated in the Northwestern study. None were fully satisfactory in terms of desired lane widths. Staff will review these alternatives based on the relative importance of the various lane width criteria to determine that scheme on balance might best meet the needs of the District if a movable median barrier were installed.  The feasibility and advisability of replacing raised ceramic lane markers along the lane lines with flush or recessed markings will be evaluated in the evaluation of alternative barrier configurations.

2.  Anchorage for the San Francisco end of the barrier system: As discussed in this report, satisfactory functioning of the barrier system and crash cushion at the San Francisco end requires the development of an anchorage system. Although such an anchorage appears at least at this point to be technically feasible, it must be designed and tested before a movable median barrier system can be installed.

3.  Guidance system for the barrier transfer vehicle: Because of the relatively narrow lanes and the possible need to locate a barrier system with its base adjacent to or straddling the raised pavement markers on the Bridge, precise placement of the barrier system is important. A guidance system which assures consistent, accurate placement of the barrier system as it is moved from one position to another must be designed and tested.

4.  Procedures for emergency vehicle response:  In conjunction with emergency vehicle operating agencies, strategies must be developed for responding quickly and effectively to accidents on the Bridge, depending on lane configurations. Of particular concern is the development strategies for accessing the accident site, removing stalled or damage vehicles, and relocating the barrier system if it has been displaced by the accident. Specialized equipment such as double-ended tow trucks may need to be acquired.

5. Potential Implementation Strategy:  The ultimate decision of whether or not to move forward with the installation of a movable median barrier system must be made by the Golden Gate Bridge, Highway and Transportation District. If the District determines that the installation of a movable median barrier system is desirable, it should consider the feasibility and desirability of initially leasing such a system for a period of two to three years. As previously discussed, there are a number of technical issues relative to such a movable median barrier system that have not been satisfactorily resolved. A trial installation would permit an evaluation of such a system on the Golden Gate Bridge, and may lead to a better understanding of such issues before a final commitment to purchasing the system is made. During the trial installation period, accident frequency and severity should be closely monitored. Frequency and magnitude of barrier impacts and displacements should also be carefully recorded. Finally, alternative emergency vehicle response strategies can be tested and evaluated.


The information and/or report on this page is not the complete report supplied here. If you would like to see the whole report, you may obtain it through your local library, or contacting the Golden Gate Bridge, Highway & Transportation District.  Library name's and address's are supplied above.  
This page  was updated May, 2000

"Return Home Safe"

 Copyright 1996-2000. "Citizens for a Safe Golden Gate Bridge."  All rights reserved