Testimony of Jim Hall, Chairman
National Transportation Safety Board
before the Subcommittee on Aviation
Committee on Transportation and Infrastructure
regarding Accident Involving TWA Flight 800

Good morning Mr. Chairman and Members of the Subcommittee, and thank you for inviting me to appear before you to discuss the investigation of the TWA flight 800 accident. With me today at the table are Dr. Bernard Loeb, the Director of our Office of Aviation Safety, and Dr. Vernon Ellingstad, Director of our Office of Research and Engineering. Also, many of our investigators, who have worked virtually full time on this investigation for the past year, are with us here today.

Mr. Chairman, TWA 800 has been the most extensive investigative effort in the Safety Board's 30-year history. We have been on scene on Long Island for a full year, by far a record. The costs of recovering the victims and the wreckage from this tragedy have been high. Testing and research have been extensive, but we believe the money is well spent.

And, the American people can be proud of the selfless determination of hundreds of investigators from dozens of organizations, who have worked so diligently to find the cause of this tragedy.

I do not plan to dwell on too much historical information about the investigation today. The effort has been monumental -- for example, the Systems Group has had over 40 experts covering the broad responsibilities of that group. That number is a fraction of the behind-the-scenes support being provided by the Safety Board, parties to the investigation, and outside specialists.

I would now like to describe to you where we are today with the investigation and where we are going in the near future.

Although much of the work in many of our investigative areas has been essentially completed -- including operations, powerplants, maintenance records, structures, cabin reconstruction, medical factors, flight data recorder, cockpit voice recorder, trajectory analysis, and data base management - we still have months of tests and research ahead of us. Depending on future developments, we may reopen certain investigative areas. Areas that have continuing work include aircraft systems and fire and explosion. I will talk more about those areas shortly.

Before we could get to this point in the investigation, a massive underwater search and recovery effort was necessary. Since diving operations began on July 18, 1996, there were 677 surface-supplied dives, 3,667 scuba dives, and 209 remotely operated vehicle dives. After the diving operations were completed on November 2, 1996, we employed contract trawlers with specially rigged nets to drag the ocean bottom. Trawling continued throughout the winter and early spring. Trawling ceased on April 30, 1997, and 85 randomly selected sites on the ocean bottom were videotaped to ensure that it was clear of wreckage. The activity ended on May 18.

The diving and trawling operations covered about 40 square miles of ocean floor. Literally thousands of items were recovered from the bottom of the ocean and brought to the hangar for study. We believe we have recovered between 95 and 98 percent of the airplane.

In January 1997, we began to reconstruct the center section of the airplane in order to be able to better demonstrate the relationship of the various pieces of structure and systems and the sequence of the breakup of the airplane. The full- scale reconstruction, which is about 94 feet long, is the largest ever completed in the world. It consists of almost 900 pieces of wreckage. It has been extremely beneficial to the investigation.

Mr. Chairman, I would now like to discuss our progress to date.

There is no evidence of a bomb or a missile impact in the wreckage.

Based on evaluation of the recovered wreckage and a detailed evaluation of the sequence of events, we have determined that the fuel/air vapor in the center fuel tank exploded and that the explosion of the tank initiated the breakup of the airplane. We have not yet determined what ignited the fuel vapor in the center tank.

The determination of the sequence of events was reached with the participation and agreement of the parties to the investigation, as well as outside specialists from the United States and overseas.

Our investigation continues to concentrate on two main areas. First, we are attempting to determine the ignition source of the fuel/air vapor in the center tank. Second, we are attempting to understand the composition and characteristics of the fuel/air vapor in the fuel tank. To accomplish these two tasks, extensive testing and research has either been completed or is under way. We have used a host of independent laboratories and facilities, and have not cut corners in seeking the best available resources. Let me detail some of the still ongoing work, starting with the problem of ignition.

Determination of what ignited the fuel/air vapor

There are generally six primary ignition scenarios or theories currently being pursued - all of which have been known to us for many months. We are examining each theory carefully and conducting laboratory experiments and other scientific tests that will help us decide which ignition scenarios might be ruled out.

I will now discuss each theory and describe how we are studying it.

Center tank scavenge pump-This scenario involves the potential for overheating or other ignition energy from a failure mode in the scavenge pump that has not been recovered. As to other pumps in the center fuel tank, we have examined the jettison pumps and found no evidence that they were involved in the ignition of the fuel/air vapor. These tests were conducted at the NASA facility in Huntsville, Alabama.

Although we have not recovered the scavenge pump from the accident airplane, we have conducted several tests on exemplar scavenge pumps. We also have researched the service history of the pump on the accident aircraft, specifically, and scavenge pumps in general, to determine a possible failure scenario that could explain the accident.

Static electricity-This scenario involves the potential for generation of static electricity on an ungrounded component in the center fuel tank -- Wiggins couplings or Adel clamps -- that could lead to a spark and ignition of the fuel/air vapor. We have been conducting extensive laboratory tests at the Wright Laboratory at Wright-Patterson Air Force Base in Ohio, and at the Naval Research Laboratory in Maryland, to determine whether static electricity can be generated within the center tank sufficient to provide a spark that will ignite the fuel/air vapor. Additional static electricity tests are planned for the next few weeks. We also have in progress additional laboratory tests at Wright-Patterson involving fuel system components from TWA flight 800.

Fuel quantity indicating system-This scenario involves the potential for an electrical short circuit in an airplane wire bundle outside the tank that leads to a spark or overheating and ignition from a fuel quantity indication probe or compensator in the center fuel tank. We have examined the recovered portions of fuel probes from the center tank, the fuel pump cockpit switches, and other fuel system components in our laboratories and in the Wright Laboratory. We have conducted tests of exemplar fuel quantity probes at the Lear-Siegler factory in Seattle, Washington, to determine whether an electrical short circuit could be passed into the tank as an ignition source. We have also examined the wires, wire bundles, and wire conduits recovered from the TWA flight 800 wreckage.

No. 3 fuel tank electrical conduit-This scenario is related to a known history of deterioration of wires in an aluminum conduit that passes through the No. 3 fuel tank. The scenario involves the potential for a spark leading to ignition of vapor in the fuel tank vent tubes and the flame propagating to the center fuel tank. Examination of the wreckage has so far proved inconclusive, but this work continues. I will discuss shortly a flight test that will contain instrumentation to examine this theory.

Small explosive charge-This scenario involves the possibility that a small explosive charge detonated near the center fuel tank could lead to ignition of the fuel/air vapors. In August 1996, we learned about the availability at Brunthingthorpe in the United Kingdom of a retired Boeing 747 that was to be used for baggage container explosive hardening tests being conducted by the FAA and the Defense Evaluation and Research Administration of the U.K. This test was part of the research engendered by the bombing of Pan American flight 103 in 1988. We joined this effort to record and identify the sound spectral signatures of explosives when recorded on the cockpit voice recorder system.

In late July and early August, we will conduct additional tests on the Brunthingthorpe airplane. These will involve setting off small explosive charges in selected locations around the center wing tank to determine the damage that results and to make comparisons with the wreckage of TWA flight 800. If the center tank is not damaged significantly during those tests, we plan to conduct a full scale fuel/air explosion test.

High speed particle penetration--This scenario involves the possibility that a high speed fragment from a meteorite, space debris, or missile warhead could penetrate the center fuel tank and cause ignition.

In cooperation with the FBI, we have conducted tests and examinations of the wreckage to determine if a high speed fragment may have penetrated the center fuel tank and provided an ignition source. Test plates of aluminum were subjected to high speed penetration of various size particles and these specimens were compared to more than 150 holes found in the structures of TWA flight 800. Experts from Brookhaven Laboratories on Long Island assisted in this work, as did experts from the Naval Weapons Center at China Lake, California. To date, we have found no evidence of high speed particle penetrations; however, that work continues.

Determination of Conditions under which fuel/air vapors in fuel tanks are explosive, and the minimum energy needed to ignite the vapors.

Besides the work to determine possible ignition scenarios, we have been conducting numerous tests, and more are planned, to better understand the flammability and explosive potential of Jet A fuel. I need to point out that very little is known about the composition and characteristics of Jet A fuel, despite its use for many years. We need to understand the characteristics of the fuel to evaluate its susceptibility to ignition and to understand the propogation of the explosion that caused the accident. For example, we would like to determine where the ignition took place in the center tank, how it propogated, and how the environmental conditions affected the event. If we can learn that, we might be closer to determining what the ignition source was, and we can develop more definitive corrective actions, both mechanical and environmental. We have met with specialists from throughout the world to assist us in this effort.

We recently leased a Boeing 747 for flight tests. The airplane has been instrumented with temperature and pressure sensors, and vapor sampling equipment to provide a detailed characterization of the environment in the center tank and the rest of the fuel system. It will be flown from JFK International Airport in the next day or two to determine the temperature profile and chemical composition of the fuel/air mixture in the center tank under conditions similar to those of TWA 800. We are being assisted by the Desert Research Institute of the University of Nevada at Reno, and we trust that the data gathered from these flight tests will bring us closer to our goals of determining the cause of the accident and developing accident prevention measures.

We have been conducting tests at CalTech and the University of Nevada at Reno to determine the chemical characteristics of Jet A fuel under a variety of conditions. These tests include measurements of explosive temperatures, pressures, minimum ignition energy, and fuel vapor composition. We recently obtained fuel samples from Athens, Greece to compare with samples taken from an airplane that has flown from Athens to JFK. Those samples will also be examined to determine if the characteristics of the fuel change during flight.

Of course, we had analyzed fuel samples from both JFK and Athens immediately after the accident.

Once we have determined the chemical composition of the fuel/air vapor, we plan to conduct scale model tank explosion tests assisted by experts from CalTech and other laboratories. We have already conducted small-scale explosion tests using a single chamber test vessel. Because the Boeing 747 center fuel tank is a more complex structure, we need to evaluate the effects of its multiple interconnected compartments on the ignition and explosion physics. Depending on the results of these scale tests, we may conduct full-scale tests in which we will blow up one or more center tanks salvaged from retired Boeing 747s.

Concurrently with the explosion tests, we plan to conduct computer modeling of the fuel/air explosions to better understand the propogation of an explosion and the consequent pressures produced throughout the Boeing 747 center tank.

Mr. Chairman, I need to point out that all of these tests that I have briefly described are extremely complex and nothing of this magnitude has ever been conducted before by the Safety Board. Because of the highly technical nature of the tests, and the potential danger posed to those conducting them, each phase of each test is very time-consuming.

I believe that the flight tests that are ongoing this week, the tests in the United Kingdom, and other tests at various universities and laboratories will bring us closer to our goal of preventing similar accidents in the future.

Lastly, as we do in all major airline accidents, the Safety Board is developing plans to hold a public hearing on this accident in December in Baltimore, Maryland. Excellent facilities are available, and Baltimore is a convenient location for many family members and other participants. At that public hearing, we plan to take sworn testimony on all of the relevant issues related to this tragic accident. However, we cannot open a public docket and conduct a public hearing concerning the evidence we have gathered until the FBI declares that it is no longer conducting a criminal investigation into the loss of TWA flight 800. We anticipate that they may do so in the near future.

Though this investigation is still ongoing, the Safety Board issued four safety recommendations to the FAA that urged both short-term and long-term actions to reduce the potential for a fuel/air vapor explosion in the center fuel tanks of Boeing 747s, as well as in fuel tanks of other aircraft. We suggested possible means to reduce the explosive potential of the fuel vapor, such as adding cold fuel to the center tank before takeoff, providing insulation or other methods to reduce the transfer of heat from the air conditioning units beneath the center tank, or inerting the tank by replacing the explosive vapor with a harmless gas.

FAA responded with a request for public comments in the Federal Register, posing questions that it wanted answered by the aviation industry and the scientific community before it acted on those recommendations. The comment period closes August 1. The FAA stated that it was concerned that the safety recommendations proposed major changes in requirements for fuel tank design and fuel management in transport category airplanes because the current airworthiness standards of the Federal regulations assume that the fuel vapor (ullage) in the fuel tanks is flammable. Current design and certification requirements concentrate on the elimination of ignition sources. However, we are asking for an additional safeguard - control or elimination of flammable vapors.

NTSB agrees with FAA that there are questions in need of answers before agreement on long-term prevention can be reached. We anticipate working closely and cooperatively with FAA to develop long-term solutions. But, we also believe that more could be done in the interim to reduce the possibility of another fuel tank explosion in the meantime. The probability is already very low, but if it might be made lower, without significant cost, we believe that effort should be made. Consequently, on July 1, 1997, the Safety Board classified the FAA's response to the short term recommendations as "unacceptable."

As you know, Mr. Chairman, our issuance of recommendations before completion of an investigation is not unusual; in fact, it occurs quite often. We issued recommendations 7 days after the Roselawn, Indiana ATR crash in 1994, and 20 days after the ValuJet crash into the Everglades last year. We issued recommendations following the Sioux City, Iowa DC-10 crash on 4 separate occasions before our final report was adopted, the first less than a month after the accident.

In addition, it is our regular practice to classify the responses to our recommendations. There are currently 358 open recommendations to the FAA, 31 of them - less than 9 percent - are currently classified as unacceptable response or action.

Mr. Chairman, the Safety Board is fully aware that the safety record of the Boeing 747 and many other airplanes over the past few decades has been excellent, and fuel tank explosions have been extremely rare events. However, the evidence gathered during the investigation of TWA flight 800 and from other previous accidents indicates that they do occur and that extraordinary steps may need to be taken to prevent similar accidents.

Our senior staff and investigators have been meeting regularly with the FAA and Boeing engineers, as well as outside specialists, to discuss the complex questions that have been raised by this tragic accident and to develop appropriate solutions. We all remain committed to determining the ignition source of the fuel/air vapor in the center tank of TWA flight 800. However, we also believe it is imperative to initiate steps toward the reduction of explosive vapor in fuel tanks. We will continue to work closely with the FAA and Boeing to devise corrective measures in a timely manner.

Mr. Chairman, let me make something very clear about these recommendations. We are not saying that our short-term recommendations would prevent every accident in the future, but we do believe that they would have prevented the TWA flight 800 accident and some of the previous accidents involving explosive fuel/air vapors.

Before I close, I would like to mention that, as part of the Safety Board's new role related to families of victims of airline accidents, we will be assisting the families in memorializing the first anniversary of the TWA flight 800 accident next week. Several days of activities have been planned by the family organizations, and they are being supported by units of local, state and the Federal government. Many other organizations from Long Island that were part of the search and recovery efforts are also assisting the families. At the families' request, we will provide them access to view the reconstructed wreckage at Calverton and to answer questions about the progress of our investigation. We expect about 750 family members to participate.

I now would be pleased to answer any questions that you may have.