Failures of Bridges and Highways
Infrastructure Considerations in Safety Board
Accident Investigations
Remarks by James Hall
Chairman, National Transportation Safety Board
to the American Automobile Association
Orlando, Florida, March 11, 1996

Thank you for inviting me to take part in this very interesting conference. As you are well aware, the Safety Board and the American Automobile Association have long shared a common interest in improving highway safety. We have worked together as partners on numerous projects to reduce accidents, injuries and fatalities. We are very pleased to have you as one of our partners on the Administrative License Coalition -- a group of more than 40 agencies, associations and private companies working to pass important State anti-drunk driving laws.

In December, I was pleased to attend a meeting of the coalition at your Washington offices and met with Jim Kolstad, our former Chairman and your current Vice President, NHTSA's Rick Martinez and other safety leaders to plan strategies for the current State legislative season.

Today I will talk to you about another topic on which we have a common concern. There is no question that the maintenance of our highway infrastructure is necessary to highway safety. It is also necessary that when we improve roads and build new facilities we follow ever-changing guidelines developed through safety research and operational experience gained over the years. Part of that research and experience is obtained through accident investigation.

The National Transportation Safety Board is the premier transportation accident investigation agency in this country, if not the world. Through our investigations we determine probable causes and formulate recommendations to improve transportation safety.

Most citizens know us best from the headlinegrabbing airline disasters. Many do not know of our work in railroad, marine, pipeline or highway investigations. Infrastructure concerns have been issues in all modes of transportation -- from antiquated air traffic control equipment, to aging airliners, to corroding pipelines, to the lack of collision avoidance systems in the railroad industry. Any of these would be good subjects for discussion, but today I want to concentrate on highway accidents that we have investigated over the years where either the crumbling highway infrastructure or the design set the stage for disaster.

But, no matter what we find in our investigations, no matter what improvements we recommend, no matter what regulators decree, this nation will not achieve its goals without commitment from those we entrust with our safety. As we will see, inspection and maintenance programs provide little protection for the public without that commitment.

Probably on the front line of infrastructure deterioration in the highway world are bridges; they are everywhere they cross rivers, swamps and bays, railroad tracks and yards, and other highways. There are more than 570,000 highway bridges in the United States. Of these, 33 percent are considered deficient. How many of these did you travel over to get here?

This does not mean that the deficient bridges are about to collapse or that they are unsafe. With proper load posting, roadway striping, signs, signals, roadside barriers, crash cushions, and other changes, our bridges can continue to serve traffic needs, albeit not to the desired level.

The Safety Board first got involved in the safety of bridges in our very first year of existence, 1967, when the Silver Bridge over the Ohio River at Point Pleasant, West Virginia collapsed, killing 46 persons. Our investigation revealed that better bridge inspection programs were needed.

In 1971, as a result of our recommendations following the Silver Bridge tragedy, the National Bridge Inspection Standards came into being. The Standards set policy regarding bridge inspection frequency, inspector qualifications, report formats, and inspection rating procedures. However, the future was not secured. Bridge repair and replacement needs far exceeded funding sources -- and that is still the major problem 25 years later.

In the 1980's, though some believed that the nation's bridge inspection program was maturing well, in actuality many states had poorly run programs with little oversight by the Federal Highway Administration. The consequences were inevitable, but no less shocking when they occurred.

On June 28, 1983, a 100foot, 3lane section of the I-95 Mianus River Bridge near Greenwich, Connecticut fell 70 feet into the river during the night. Luckily, the timing kept the death toll down to 3; had it collapsed during a rush hour, countless more would have certainly died.

The Safety Board found that the State's bridge inspection and maintenance programs, culminating in its inspectors' failure to detect displacement in the pinandhanger assemblies that connected the suspended span to the fixed structure of the bridge, had caused the collapse of the suspended span. The movement affecting the assemblies was the result of corrosion; that is, plain old rust, between the bridge girder and the 6-and-a-halffootlong hangers. The buildup of rust had pushed the hanger off its support pin.

The bridge had been inspected nine months before the accident but we found that the inspectors had not been given specific instructions for inspection of the pin and hanger assemblies, and the inspection itself was cursory. Does this sound like the level of commitment here the public deserved?

As a result of this accident, the Safety Board made 28 bridge safety recommendations to the Connecticut and U.S. Department of Transportation, the FHWA, and industry groups. To the credit of these organizations, all of the recommendations were implemented.

Almost two years later, on April 24, 1985, two34footlong twin spans of the Chickasawbogue Bridge on U. S. 43 near Mobile, Alabama fell into water ranging from 10 to 30 feet deep after a steel Hpile support rusted through at the river bottom. Fortunately, there was no loss of life. The bridge had been inspected only three weeks before, but only those elements above water were looked at; for the structures in the murky creek, the corrosion remained undetected.

The Alabama Highway Department had previously discovered corrosion of underwater elements of bridges located in similar waters, yet had failed to inspect this bridge and other bridges underwater. Here again we saw a lack of commitment on the part of those on whom the public relied by failing to make the extra effort to look at underwater conditions and by failing to apply earlier findings to similar bridges around the State.

And Alabama was not alone. We found that thirtyfive states did not routinely inspect the underwater elements of bridges, and even where the states had active inspection programs, the detail and frequency of the inspections were of questionable adequacy.

As a result of our recommendations to the FHWA and the American Association of State Highway and Transportation Officials, effective underwater inspection programs were developed. But no program is effective if authorities don't follow it or ignore the findings of their own inspectors.

In April 1987, Connecticut's neighbor, New York State, suffered a major bridge failure on its Thruway system. A flood scoured out the footing of a 540foot-long bridge near Amsterdam, New York, dropping a 230foot section into the flooded Schoharie Creek. Four passenger cars and one tractorsemitrailer plunged into the creek, killing all ten occupants.

We determined that the probable cause of the collapse was the failure of the Thruway Authority to maintain adequate rip rap around the bridge piers. Had the rip rap, basically a rock barrier, been maintained to design levels, the bridge would have weathered the storm. Ironically, the missing rock was noted on 2 inspection reports and was to have been replaced, but that detail had been deleted from rehabilitation plans that had been prepared and let for contract a few years earlier.

Since the collapse, the New York State Thruway Authority and the New York State Department of Transportation have had a very aggressive bridge inspection program. This proactive program, the Bridge Safety Assurance Program, consists of a multistep process for identifying potential causes of bridge failure and for the subsequent rating of bridges as to the extent of their vulnerability to these failure modes.

The Safety Board highly commends New York State for this initiative. This tracks a major recommendation that came out of our investigation of the Mobile, Alabama Amtrak wreck that killed 47 people. We urged the U.S. Department of Transportation to determine the vulnerability of the nation's railroad and highway bridges to collisions from marine vessels.

The last three bridges discussed were all built in the 1950s, when the interstate program rushed us into the superhighway age. Perhaps the designers did not anticipate that future generations would neglect maintenance. With hindsight, the designs could have been more forgiving; yet the fact remains that many more such bridges exist. Bridges must be inspected thoroughly and maintained properly if we are to preserve our infrastructure. And that takes commitment of both scarce public funds and inspectors' time.

Sometimes, attempts to improve the infrastructure, when not well thought out, lead to their own problems.

On April 1, 1989, 3 spans of a bridge carrying northbound U.S. Route 51 over the Hatchie River in my home state of Tennessee collapsed during a flood, killing 8 persons in the 5 vehicles that plunged into the raging river with the spans.

The bridge had been built in 1931. In the mid-1970s, Route 51 was widened so a parallel bridge was built to carry southbound traffic, thus dedicating the old structure into carrying northbound traffic. What set up this disaster was that, while the original bridge was four fifths of a mile long, the new bridge was only 1,000 feet long. Earthen embankments constructed to carry the roadway to the new bridge in effect dammed up the flood plain just downstream of the old bridge.

The flood of March and April 1989 accelerated the pace at which the Hatchie River channel migrated, pushing the water under bridge piers not designed to be underwater. The poor design of the new bridge created the scenario that destroyed the old bridge.

Since this accident, the Federal Highway Administration has been encouraging States to evaluate the effects of new construction on existing structures during roadway and bridge improvements.

In the last several decades there has been a vast change in highway design concepts due to increased awareness of the safety impact of various highway features. The Federal Highway Administration has been active in this area. We believe in the constant need to take into account the total roadwaydrivervehicle system. One of our most recent accident investigations illustrates this:

On July 27, 1994 a tractor cargotank semitrailer loaded with 9,200 gallons of propane, traveling east on I287 in White Plains, New York, drifted across the left lane onto the left shoulder, struck a 27inch-high Wbeam guardrail and overturned onto its left side. The resulting fireball rose 200 to 300 feet in the air as the tank was propelled a distance of 300 feet. Luckily the accident happened at 12:30 a.m. when the highway was relatively empty of vehicles. Although the driver died and several houses burned, there were no other fatalities.

We determined that the driver probably fell asleep. Our investigators documented only 11.5 hours of sleep in the previous three days.

We also felt that the design of the highway contributed to the accident. While the highway had been designed to the 1954 American Association of State Highway Officials Standards, these standards did not, and still do not significantly consider large vehicles. In the White Plains accident, since the bridge had four columns it did not collapse.

In two previous accidents we investigated, the bridges did collapse from collisions with tractor semitrailers. On May 19, 1993, on I-65 near Evergreen, Alabama, a cement carrier left the paved road on the right side, overran the W-beam guardrail and sheared the near column of a two-column bent. The bridge fell to the roadway and an automobile and another truck struck the fallen spans, killing the two drivers.

On August 4, 1994, on I-30 near Hooks, Texas, a tractor semitrailer went into the median and sheared on of the two bridge columns. That bridge also fell onto the roadway but no other vehicle collided with it. The three occupants in the tractor were killed -- the driver, his passenger and their dog.

In both of these accidents, the structures were many decades old and were designed without redundancy or column protection. The guardrail was designed to protect errant automobile drivers from impacting the columns -- but what of today's large, heavy trucks?

When bridges can no longer meet the transportation demands placed upon them due to clearances, loads, or safety features, they pose a hazard to the motoring public. Recently the Safety Board investigated two accidents involving insufficient clearances for railroad cars in which bridge spans that carried highway traffic collapsed. New York just completed a study of bridge failures and determined that overloaded vehicles comprised the second most common mode of failure. During the next fiscal year, the Safety Board will place its investigative emphasis on functionally obsolete bridges and substandard barrier systems on bridges.

Some infrastructure improvements can be retrofit fairly easily. For example, those of you from the northeast who have driven on the New Jersey Turnpike should take solace in the knowledge that a 42inch-high concrete barrier separates opposing traffic. The Safety Board had investigated several accidents before the installation of this barrier, but has investigated none since.

We have asked the FHWA to evaluate the performance of roadside barriers and highway geometrics to be compatible with large trucks and change the highway design standards to meet the safety needs of large trucks. Additionally we have asked them to require that the approximate 160,000-mile National Highway System, which will carry 40 percent of the nation's highway traffic and 70 percent of the nation's heavy truck traffic, be based on heavy truck operating characteristics. We realize that this isn't going to happen overnight - or even as long as I am chairman of the Safety Board - but it is a goal we should aim for as we reconstruct our highway system.

The Highway Safety Act of 1973 established specific highway safety programs intended to reduce the number and severity of highway traffic accidents through engineering improvements to hazardous highway locations and elements. It has been a successful program; it can be made better and we will continue to recommend safety improvements where needed.

Maintaining our highways in good condition and designing safety into the highway system not only reduces injuries and fatalities but also makes economic sense. A recent report by the U.S. Department of Transportation estimated that investments of $68.2 billion to improve the highway system ($24.6 billion for capacity expansion and $43.6 billion for system preservation) would yield a return of $2.60 for every dollar spent. Such a return is well worth the expense. I hope organizations like yours will advocate the investment of necessary funds to keep our country's highway transportation system the most efficient and safest in the world.

Finally, looking into the future, Intelligent Vehicle Highway Systems will open a new world of transportation possibilities for us, and a new world of infrastructure needs that must be addressed.

But, whatever infrastructure improvement program we embark upon, you need to make sure that highway authorities in your communities are as committed as you are to improving the safety of your roadways and bridges. This conference is an admirable step in that direction, but don't leave your commitment here in the Florida sun.

Keep the pressure on back home, and let us know if we can help.

I want to congratulate Triple A on this conference, and thank you all for inviting me today.

 

Jim Hall's Speeches