excerpt

"We know that these aircraft exploded, but there are many other factors other than the air conditioning packs that must be looked at to determine the exact ignition source in these accidents."

Precisely.  Read .... and then read on for

Fuel tank dangers: the fixable problem that may be here to stay

By Air Safety Online Staff

In most respects, July 17, 1996 was to be just another day for N93119, an ancient, 25-year-old Boeing 747-100 jetliner sitting on the tarmac at New York's John F. Kennedy International Airport, scheduled for an evening flight across the Atlantic.

On his last assignment with ABC Sports, producer John O'Hara hurriedly made his way through the airport to catch a flight to Paris to supervise coverage of the Tour de France. The 39-year-old five-time Emmy Award winning journalist was taking his wife Janet and daughter Caitlin along for the trip, while his sons Brian and Matthew stayed at home with grandparents.

O'Hara boarded the plane just moments before the scheduled takeoff time.

The packed 747, with its fading "Trans World" titles, was airborne by 8:20 PM.  O'Hara and 229 others aboard Flight 800 never made it to Paris.

"The crash of Flight 800 graphically demonstrates that, even in one of the safest transportation systems in the world, things can go horribly wrong," announced Jim Hall, a modest, former political aide with a slow Tennessee drawl, to a crowd of hundreds gathered at the National Transportation Safety Board headquarters in Washington, DC.

It was August 22, 2000 when the NTSB released their official 'probable cause' of the crash of TWA Flight 800, which occurred more than four years earlier.

The cause of the 1996 crash, the NTSB concluded, was an explosion in the aged aircraft's center fuel tank.

"The bottom line is that our investigation confirmed that the fuel-air vapor in the center wing tank was flammable at the time of the accident, and that a fuel-air explosion with Jet A fuel was more than capable of generating the pressure needed to break apart the center wing tank and destroy the airplane," Bernard Loeb, then Director of Aviation Safety at the NTSB, said at the hearing.

Six years prior to the TWA crash, the NTSB had investigated an accident involving a Philippine Airlines 737-300 jetliner in the South Pacific.  The plane, which had been sitting idle on a hot runway for hours, was packed with passengers.  Nine were killed when the plane's center fuel tank exploded.

Using data from the Philippine and TWA accidents, the NTSB theorized that if an aircraft's center fuel tank were hot enough -- NTSB tests proved that the tanks could reach temperatures in excess of 200 degrees Fahrenheit -- and if the tank was relatively empty, remnants of fuel could turn into an extremely explosive vapor.

But warm weather alone cannot create this type of explosive vapor in fuel tanks.  What other factors contribute to an overheated fuel tank?

"Air conditioning," said Christine Negroni, a former CNN aviation correspondent and author of Deadly Departure: Why the Experts Failed to Prevent the TWA Flight 800 Disaster and How It Could Happen Again.

Located directly beneath the center fuel tank of the Boeing 707, 737, 747, and 767 jetliners are large air conditioning units, which can heat up to hundreds of degrees, effectively heating the fuel tanks above as well.

The TWA 747 had been sitting on the tarmac at JFK Airport for more than two hours, and it was summertime; July.  Temperatures on the 17th of July in New York reached well into the 80s.

The Philippine plane had also been sitting for hours in hot weather as well, with the air conditioning packs running the entire time.

And there have more incidents involving Boeing planes.

Temperatures in the early afternoon of March 3, 2001 reached into the high 90s in Bangkok, Thailand.  At Don Muang International Airport, a Thai Airways Boeing 737, registered HS-TDC, was being prepared to fly Thailand's Prime Minister Thaksin Shinawatra and his son, as well as more than 140 other passengers to the northern city of Chiang Mai.

The 9-year-old jetliner's air conditioning packs had been running for 40 minutes on the ground at the airport, as well as during the entire previous flight.

Minutes before passengers were to be boarded, the center fuel tank exploded.  18 minutes later, the plane's right wing tank also exploded, and the resulting fire destroyed the aircraft.

One flight attendant was killed, and seven others were injured in the blast.  The National Transportation Safety Board and Boeing are assisting the Thai government in the investigation.

In a press release dated April 11, 2001, the NTSB made note that the Thai accident was frighteningly similar to the Philippine Airlines accident eleven years earlier.

"The recording of the HS-TDC explosion has features that are similar to recorded features of a Philippine Airlines 737-300 center wing fuel tank explosion in May 1990.  Neither recording includes a precipitating sound of an initiating explosion that may have ignited the fuel tank," the release said.

"What Boeing has found was some problem with the air conditioning system but they are not sure. It may have been caused by explosives because of traces of C-4 explosives," Prime Minister Thaksin said.

FBI labs later found no traces of explosives, and a Boeing spokeswoman said Boeing never made any such comments relating to the air conditioning system on the Thai jet.

"It's painfully obvious that this really is a problem, and the NTSB realizes that," Christine Negroni, who has been studying the effects of overheated fuel tanks for years, told Air Safety Online.

"The thing that all of these aircraft have in common is that they have heat elements directly below the center fuel tanks," Negroni said, noting that the NTSB has been trying for years to make this a public issue.

Says Negroni, "This is like a tree falling in the forest, the reason that the NTSB is making this as poignant an issue as they are, is because it doesn't matter how much they say we have a problem, nobody wants to do anything about it."

Negroni may be right.  The Federal Aviation Administration has issued no directive in recent years, that Air Safety Online could immediately locate, regarding the use or position of air conditioning units on Boeing jetliners.

According to Negroni, Boeing has known of the dangers of heated fuel tanks and the placement of the air condition packs for a long time.

Boeing decided not to reposition the air conditioning units in later models of the Boeing 737 and 767.  (Boeing)

"The 707, 737, 747, and 767 center fuel tank design issue has been a known problem for 35 years, and nothing is being done about it.

"Many at Boeing throughout the 1980s thought that in retrospect, maybe the air conditioning pack under the center tank design wasn't such a good idea, but Boeing continued to make these bad decisions, and bamboozled the FAA to go along with it," Negroni said.

The National Transportation Safety Board, however, has presented this issue to the FAA at least three times in the past twenty years, and they are running into trouble making this a more public issue.

Perhaps if the Philippine and Thai events had resulted in larger accidents, the issue would have become more public.

"The only thing that makes these non events is that the planes were parked on the ramp.  Just because Thailand's Prime Minister wasn't killed in the accident, that doesn't mean this isn't a real alarm, but it still makes it hard for investigators to convey such a message," Negroni said.

"It's important to understand that when an airborne aircraft's center fuel tank explodes, that plane is going down.  It's only luck that these two aircraft weren't airborne."

In 1998 and 1999, the Federal Aviation Administration mandated that U.S. airlines install additional shielding on wiring in and around the center fuel tanks of older Boeing jets.  This, the FAA acknowledged, would reduce the risk of a spark causing an explosion in a tank with heated vapors.

Only now are authorities beginning to look for ways to prevent the heated vapors from being created.

In July 2000, the FAA launched a fuel tank inerting study, in which a Boeing 737, equipped with special sensors, would be used to test the effectiveness of a system that pumps nitrogen into the fuel tank while the plane is on the ground.

This is part of an FAA study that began in 1999 to determine how nitrogen could be pumped into planes at airports with a system that uses a membrane to separate nitrogen from air.

The nitrogen inerting system could prevent explosive fuel vapors from filling empty fuel tanks.  In fact, the United States Air Force has been using a similar system on their aircraft for years.

Unfortunately, it took the loss of more than 240 lives over 35 years to generate enough awareness for the FAA to begin an official study on the issue.

It's a clear-cut case of what aviation industry insiders call "Tombstone Technology."

"The thing that all these aircraft have in common is that they have
heat elements directly below the center fuel tanks.  We need to get rid of either the heat elements or the vapors, and we need to do this now," Negroni said.

The Boeing Commercial Aircraft Group did not change the location of the air conditioning packs in its newest 737NG ("new generation") or 767-400 jets, still directly underneath the center fuel tanks.

Meanwhile, Boeing told all 737 operators they should not use fuel boost pumps in the center fuel tank when the tank is empty, the Seattle Post-Intelligencer reported.  Boost pumps transfer fuel from one tank to another.

Boeing also issued a service letter in May of 2000, as a result of the TWA crash investigation, asking operators to use ground source air conditioning, which does not use the aircraft's air conditioning packs.

The Federal Aviation Administration is expected to issue an airworthiness directive soon making the recommendation mandatory, the newspaper said, citing an unnamed industry source.

"The air conditioning packs are a known heat source," said Liz Verdier, a spokeswoman for the Boeing Company.  "We know that these aircraft exploded, but there are many other factors other than the air conditioning packs that must be looked at to determine the exact ignition source in these accidents."

"We still haven't found any one cause for all of these fuel tank explosion accidents.  As we continue these investigations, there may be more causes found, and we will work to fix those," Verdier said.

"Boeing is actively working with the FAA to make fuel tanks as safe as possible," Verdier said, adding, "These aircraft have been flying safely for years and years.  If we can make them even safer, we will."

Before the fuel tank inerting program can be implemented, the FAA must complete a "Cost of Implementing Ground-Based Fuel Tank Inerting in the Commercial Fleet" study to determine whether the additional level of safety is economically feasible for U.S. airlines.

  The FAA on fuel tank inerting

TWO INPUTS

-----Original Message-----
From: IASA Safety [mailto:safety@iasa-intl.com]
Sent: Saturday, 11 August 2001 12:26 AM
To: gammontech@gammontech.com
Subject: That Vexing Fuel-Tank Inerting Problem

Dear Sirs re: your site at http://www.gammontech.com/index.htm

FAA's Fuel Tank Harmonisation Working Group Recommendations:

For fuel vapor reduction, five of the options considered reduce the exposure to flammable fuel vapor. These are:

· Insulate the heat source adjacent to fuel tanks;

· Ventilate the space between fuel tanks and adjacent heat sources;

· Redistribute mission fuel into fuel tanks adjacent to heat sources;

· Locate significant heat sources away from fuel tanks;

· Sweep the ullage of empty fuel tanks.

  Perhaps you might have some professional opinions about this additional idea (below). Practical, impractical? Indeterminable? Not worth pursuing?   regards IASA Safety www.iasa.com.au and www.iasa.com.au   The Proposition   If the FAA won't play ball with nitrogen inerting of airliner fuel-tanks, maybe there's another simpler way to stop the ullage filling up with flammable vapours:
 
From an earlier email:
"c.  There seems to have been little discussion throughout the TWA800 saga about the virtues or benefits of fuel additives to raise auto-ignition flash-points (page 41). I would have thought that any improvements in this area could be achieved without loss of calorific value per unit mass of fuel. And every step in the right direction helps disproportionately. An alternative to adulterating the whole fuel load in this way might be (for "at risk" a/c) simply to add a physically immiscible fraction of distillate (or ?) to the CWT that would still be less dense than the Jet A, float on top of residual fuel and suppress Jet A vapours (much as a thin layer of oil floating on water stops evaporation). The quite small quantities involved would mix well via downstream filtration (ie. emulsify), when pumped into the wing-tanks, HP pump-sprayed and then atomised into the combustion chamber - and so should not unduly affect the normal power-plant or APU combustion process. As you know, jet engines will burn just about anything and a small adulteration of another fluid, whether combustible or not, would not have any great impact upon fuel system components. Adding a relatively small amount of fluid to the CWT (only) would be a far simpler proposition than setting up nitrogen-inerting facilities worldwide. As you know, all sorts of lubricious and anti-icing additives are already added to the basic fuel.  Some aircraft, particularly military ones, actually require it."  
The basic idea is that the additive would be required (to be added by the refuellers) once the temperature exceeded a given figure, +30 deg celsius (say) (unless the CWT was to be at least half-filled - whereby the cooling heat-sink effect would then stop the underlying aircon packs from heating the fuel sufficiently to create a build-up of vapour). I would imagine that 25 US gallons would be enough to give a one inch top-layer of a "damping" membrane overlay on the residual fuel (i.e. in an essentially empty 747 CWT).
It's worth mentioning the concept - because some brains-trust somewhere might pick up that ball and run with it.   ___________________________________________________________________________________
From Mariner 9   Surely a cheap & simple system would be to continually circulate the fuel from the centre tank to the wing tanks. As long as the fuel is maintained below the flash point (minimum 38C for Jet A-1) flammable vapours should not be generated.
Heavier distillates will invariably have a heavier density, and thus will not "float on top". Additionally, there would be problems with CFPP, freeze point, & MSEP of these materials.   ______________________________________________________________________________________
Dear Mariner 9   Being just a dumb pilot I haven't a good handle on all your terms. (CFPP, freeze point, & MSEP ).
The immiscible fluid (for the atop-fuel inhibitor membrane) that I had in mind wouldn't have nasty surfactant or interfacial tension properties and would form a long-lasting fine emulsion on pump-out (even if the exit from the CWT tank required a high-speed swirl-pump to properly emulsify (rather than centrifugal or displacement-type pumps) . In fact in an ideal world it would be hygroscopic. Micro-filter separators or clay adsorption are options for halting most of the introduced fluid's passage onwards into the APU or engines, however I've not heard of too many jet engines stopping (as recip engines are wont to) due to water in the fuel. Anything short of large particulate contaminants (in the micron range) will simply get burned up.
Admittedly fuel system components are not happy about ice and loss of lubricity, corrosive fluids or particulate contam. It might be a challenge to find a fluid with the desired properties (including a low freeze-point) that wouldn't clog the coalescing media. I also wouldn't have thought that CFPP would be a big consideration because of the low quantities of this introduced fluid that would be involved. But I also doubt that any form of filtration would keep it permanently in the CWT (which would be desirable) - not even if it was a perma-gel.   The trick would be to find a fluid with all the right properties, that would be cheap, stable and available in suitable quantities - and I guess you're saying that that would be unlikely. Anyway happy to hear any more thoughts (or options) on this, particularly anything "learned" about the desirable properties (SG etc).   regards
IASA Safety http://www.fire.tc.faa.gov/pdf/EXECSUM.pdf   Fuel Tank Harmonization Working Group

http://www.fire.tc.faa.gov/pdf/TG3.pdf    Fuel-tank inerting (Task Group 3 Report) http://www.gammontech.com/gg28.htm
http://www.gammontech.com/menu1.htm
CFPP = Cold Filter Plug Point? Paraffin waxing clog-up of filters?
Solution = add PriFlow antigel at a later stage?
http://www.priproducts.com/priflow.htm   MSEP = EM-SEP  (WSIM)?
_________________________________________________________________________   http://www.gammontech.com/menu1.htm   http://www.gammontech.com/gg28.htm   NO. 28                              WSIM, MSEP and Swift Kit          
APRIL, 1984  
Comment: This is a totally revised issue of GamGram No.28. Rewriting it became essential because during the 12 years since the original issue was published many changes have occurred. The basic problem has been that the industry requires more exacting information, better test reproducibility and data that relate more closely to the filter separators that are being used in aviation today. This has resulted in the development of better test procedures and refined testing apparatus. The original title was "How to Measure WSIM".   REVISED OCT.1996   Most jet fuel supply systems include a piece of equipment known as a filter separator. Monitor filters are also used but they are not the subject of this GamGram. Unfortunately, very few people know the conditions that must exist for the filter separator to do its water removal job properly. If the water could always be expected to lie in the bottom of a tank with the fuel on top, it would be a simple job to drain it away, and most operators do that regularly. But fuel with water in it that goes through a centrifugal pump becomes an emulsion of literally millions of tiny water drops that do not settle to the bottom of the tank for long periods of time. It is this emulsion that the coalescer elements in the filter separator must deal with.
They must gather the tiny water drops together so that they will become large drops (coalesce) that can rapidly settle to the bottom and be drained away.   The enemy of a coalescer element is "surfactant" or surface active agents that prevent water drops from gathering together into large drops. They are chemical molecules that seek and influence a surface. The particular surface they "like" is the surface of a water drop in the fuel. The reason they like a water surface is because they have 2 "heads". One head likes fuel; the other head likes water. So, if the fuel contains surfactants and if water is present, those 2 - headed molecules "zoom" to the surface of the water drop just like bees go for honey. The fuel "heads" orient themselves to stay in the fuel and the water heads are captured by the water. Ultimately, the entire water drop is surrounded by a surfactant film making it impossible for 2 water drops to coalesce together because they cannot come into contact.   In the early days of jet fuel handling, it became obvious that a test was necessary to find out if a batch of fuel was contaminated with surfactants to an extent that coalescer performance would be jeopardized. The Water Separometer Index (WSI) test was developed and after modification it became the WSIM test (pronounced "wiz-um"). A reading of 100 was excellent, meaning that coalescers would perform very well. If the reading was as low as 70 the fuel was considered very poor. Extremely contaminated fuel could be "zero".   The modern instrument that measures the surfactant contamination of the fuel is currently called the Micro-Separometer ®. It is a highly refined version of the original equipment. The reading is still 100 for the best fuel but instead of referring to it as the WSIM rating, it is called MSEP (pronounced Em-sep). ® Registered trademarks by Emcee Electronics Corp.   Both the WSIM and the MSEP equipment are based on the same idea; an emulsion of water and the fuel sample is forced through a pad of fiberglass coalescing media. An optical device measures the haze in the effluent. The less haze detected, the higher the rating and vice versa. While precision (repeatability) has never been very good for either test, MSEP has proven to be superior to WSIM. Another big problem has been that the test over-reacts
to Stadis 450, the additive that improves fuel conductivity. In other words, a low MSEP fuel may perform quite adequately in a real coalescing performance test. Considerable pressure from users has influenced great effort to overcome these problems.   Possibly the most important variable that has been investigated has been the replacement of fiberglass with the same coalescing media that is used in
manufacturing modern coalescer elements that have passed the tests that are specified in API 1581, Revision 3. The new material looks somewhat like heavy paper; it contains very, very fine glass fibers. Fiberglass insulation (used in older coalescer designs) is such an inconsistent material that coalescer manufacturers were forced to find better media several years ago. The device that holds the fiberglass pads in the current version of the Micro Separometer is an aluminum capsule called the Alumicel ®. So what we are saying is that in the future new Alumicels are expected to contain a paper-like coalescing material instead of fiberglass. Meanwhile, the currently available Alumicels must be considered valid. As of June 1996, encouraging test results show that the instrument itself will probably not have to be revised. This is very good news for owners of the model known as Mark V Deluxe.   This review of tests that attempt to determine the effect of surfactants on jet fuel would not be complete without a comment on the technical property that is involved. That property is interfacial tension", and in our business it means "strength of the interface between the fuel and water. If the film of molecules at the interface is strong, large water drops can exist. As the interfacial film decreases in strength, the smaller the water drops will be
until the mixture of water and fuel becomes an emulsion. The measurement of interface strength is performed in the laboratory by a delicate instrument called a "tensiometer". It is definitely not a field instrument but a kit has recently entered the market that performs this measurement in the field.
It is called "SWIFT KIT" and is marketed by Velcon Filters, Inc. This kit is particularly useful in checking the performance of clay treaters that are used in our industry to capture and remove surfactants that cause the interfacial tension to decrease; clay adsorbs the surfactant molecules as described in GamGram No.14. Therefore, by checking the interfacial tension (IFT) before and after the fuel has passed through the clay, the operator can quickly assess the performance of the clay. This can also be determined with a Micro-Separometer but that is a more time consuming and expensive test.   In conclusion, the Micro-Separometer has proven to be the most reliable instrument for evaluating the ability of a fuel sample to have its water removed by a filter separator. A program is currently underway to improve repeatability and we will further revise this GamGram to reflect the results of that investigation when it has been completed.  

NOTE: Here in December 1997, Emcee now has a new Alumicel under test and up for approval. The idea is to make it more resistant to error caused by Stadis 450. Tests have been very positive, and soon this new cell should be approved.

Number Two:        Match the facts to the 737 incidents and TWA800:

  Read the quoted story below and you might begin to appreciate that they have known all along what caused TWA800 to explode. When you have read all that was known (below) and then (right-click "save as..")  downloaded and viewed the colour photos from the parent pdf (acrobat) file at http://www.iasa.com.au/PDF/FUELTANKTERMINALSTRIP.pdf, you will appreciate that:   a.  silver-soldered terminal connections interact with the sulfur in JetA fuel to form conductive silver-sulfidation deposits. The accumulation of these electrically-conductive deposits on fuel tank FQIS terminal blocks means that there is a breakdown in that terminal block's insulation and then as little as a 9 volt transistor radio battery can cause an arc across that FQIS terminal block. So we might conclude that not a lot of misdirected electricity was required to ignite that vaporous centre wing tank of TWA800 (or the 737's in Manila and Bangkok). The aging wire and a minor amount of induced current was all that was required to arc across the FQIS terminal block.   b.  Smiths Industries, BFG and many other responsible manufacturers of intank electrics have been very quietly changing over to nickel-plated terminations on wiring, hoping that this nasty problem will just "go away".   c. The FAA has recently decided that a quiet program with the cooperation of the relevant sections of industry is in everybody's best interests. It would, after all, be a huge program to swap out all immersed silver-soldered wiring in fuel-tanks, wouldn't it? So in due course such "anomalies" (see below) as exploding airplanes should decrease.  

Boeing Analytical Engineering Report 2-5323-WP-91-97, dated March 10, 1991. that described Boeing's examination of a sulfide deposit, on a harness from a fuel quantity indicator from the right wing tank of a 757, The fuel quantity indicator had a documented history of irregularities after only 750 hours in service.  The report noted "it is readily observed that the contaminant has migrated up into the conductor even after only 750 hours in service".   Boeing Laboratory Report 9-5576-P+CA-025P, dated March 30, 1993, that described sulfide deposits on electrical hardware from the fuel tanks of 737, 747, and DC-10 airplanes.
Boeing Laboratory Report 95576-P+CA-025P1, dated April 29, 1993, that also described sulfide deposits on electrical hardware from the fuel tanks of 737, 747 and DC-1O airplanes and included two additional reports obtained through a literature search): Silver Corrosion by aviation Turbine Fuel , dated September 1970. written at the Journal of the Institute of Petroleum; and Copper and Silver Corrosion by Aviation Turbine Fuels, dated April 22, 1973 written at the Indian Institute of Petroleum.
Boeing Analytical Engineering Report 9-5576-WP-97-272, dated August 5th 1997, that described electrical tests of FQIS parts containing sulfide deposits.   During FAA-sponsored tests conducted in response to Safety Recommendation A-98-37 (issued as a result of this accident) researchers at the University of Arizona also used the method developed by UDRI to create silver-sulfide deposits in a laboratory. During subsequent tests, these deposits served as an ignition mechanism for Jet A fuel vapor. The results of these tests were discussed at a November  9, 1999 meeting involving FAA, Safety Board. and industry personnel. According to University of Arizona personnel, "the formation of sulfur-containing conductive deposits from Jet A fuel on silver wire occurred with both a.c. and d.c. current.  Additional tests in this area are ongoing".   In response to AFRL's findings from the 1990 vapor ignition incident. BFGoodrich eliminated the use of silver-plated components in the FQIS and began using nickel-plated wire, gold-plated ring connectors, and sealant Those improved components have been used in military airplanes since about 1993. According to BFGoodrich, since that time there has been a large reduction in the FQIS anomalies that had associated with silver sulfate deposits (such as FQIS inaccuracies). Although a Boeing 1991 engineering report indicated that silver should never contact sulfur-containing liquids because of the susceptibility to sulfidation, and Boeing uses nickel-plated (instead of silver-plated) wiring in its newly manufactured 777 and 737-NG airplanes, Boeing indicated in a December 7, 1999, letter to the Safety Board that it does not recommend replacing silver-plated FQIS components in existing airplanes.