Twa 800

[cyrus15]

science 101

Hello everyone,

Attention: bsieker (Bernd)

Indeed, let’s call it Science 101

"Now let's see. Tanks have vents. Fluids (i.e. gases and liquids) stream from regions of higher pressure towards regions of lower pressure. What is the pressure inside and outside the tank?
If we go along with your premise that no air was in the tank to begin with (more on that later), the only thing inside the tank to create pressure is the fuel vapor pressure. The only value I have found is for JP-8 at 38 degrees Celsius: it is about 200Pa. Another document about Jet-A1 mentions "less than 10hPa", so, let's be generous and assume1000Pa.
Ambient air pressure at 13,000ft is around 50,000Pa, i. e. 50 times the inside pressure. To figure out which fluid will stream into which direction is left as an exercise to the reader."

I agree with your first paragraph.

The vapor pressure of JP-8 at 38 oC is correct but not in this case.

Here’s the deal: check out the MSDS link for JP-8
http://www.hess.com/ehs/msds/JP8_HOV_4088_clr.pdf

Under basic physical properties section 9 it says:

BOILING RANGE: 280 to 572 oF (140 to 300 oC)
VAPOR PRESSURE: 0.029 psia @ 100 oF (38 oC)
VAPOR DENSITY (air = 1): AP 4.5
SPECIFIC GRAVITY (H2O = 1): 0.75 - 0.80
PERCENT VOLATILES: 100 %
EVAPORATION RATE: Slow; varies with conditions
SOLUBILITY (H2O): Negligible

The vapor pressure is 0.029 psia, that is equivalent to 200 Pascal (Pa) you mentioned but that is in Absolute pressure reading (almost in vacuum condition). Meaning; you can’t mix absolute pressure and atmospheric pressure readings without the proper conversion.

To do so, we should get the vapor pressure at the corrected temperature added to the ambient pressure at also the corrected temperature and altitude to get the pressure inside the fuel tank at that altitude.

So, according to your figures (assuming it is temperature and altitude corrected) we should have 200Pa + 50000Pa to get the pressure inside the fuel tank at 13000 feet, that's why tank vent is needed to allow the 50000 Pa to get in the tank though in real life, the tank pressure is little higher than ambient pressure because of the vapor pressure, that’s the correct solution to the exercise.

In the issued AD (after the accident) one item calls for adding small fuel quantity in CWT and should be considered as part of zero weight of B747 (very much permanent additional weight) to prevent the in-tank pump from running dry and overheat.

more on absolute, gauge and ambient pressure (PSIA/PSIG) reading can be found in this link; http://www.iceweb.com.au/Technical/pressure_measurements_info_notes.htm

"The fuel pump as such was not found to be the ignition souce. Secondly, car fuel pumps are usually not submerged in the tank, but somewhere down the line, which their electrical parts away from the fuel line."

The first part of your paragraph is correct but not the second part, take a look at these websites:

See the picture of a typical automobile fuel pump housed in the tank at:
http://www.trustmymechanic.com/parts...nding_unit.htm

And here’s an article from popular mechanics December 1997 issue talking about replacing an in-tank pump.
http://www.popularmechanics.com/how_...e/1272236.html

Volkswagen Fuel pump (read the description)
http://replacement.vdubpartsdirect.com/parts/vdubpartsdirect/wizard.jsp?year=1998 &make=VW&model=PAS--004&category=E&part=Fuel% 20Pump&returnurl=null&dp=true

1997 Saturn fuel pump
http://replacement.vdubpartsdirect.com/parts/vdubpartsdirect/wizard.jsp?year=1997 &make=SN&model=SCSE--002&category=E&part=Fuel% 20Pump&returnurl=null&dp=true

Old cars (imperial 1981-1983) pump repair
http://www.imperialclub.com/Repair/Fuel/pump.htm

Go to"The Fuel Pump" paragraph at this link
http://www.picotech.com/auto/tutoria...-injection.htm

At the end it reads:
"Fuel passing across the fuel pump's armature is subjected to sparks and arcing; this sounds quite dangerous, but the absence of oxygen means that there will not be an explosion!

The majority of fuel pumps fitted to today's motor vehicles are fitted within the vehicle's petrol tank and are referred to as 'submerged' fuel pumps. The pump is invariably be located with the fuel sender unit and both units can sometimes be accessed through an inspection hole either in the boot floor or under the rear seat. Mounted vertically, the pump comprises an inner and outer gear assembly that is called the 'gerotor'. The combined assembly is secured in the tank using screws and sealed with a rubber gasket, or a bayonet-type locking ring. On some models, there are two fuel pumps, the submerged pump acting as a 'lift' pump to the external roller cell pump."

Also see the picture (Figure 1.2) in the same page.

"An additional thought: Why do you think many modern jetairliners have tank interting systems, which feed nitrogen-enriched air into the tanks, to reduce the amount of oxygen, to reduce the chances of creating a flammable fuel/air mixture."

Fuel Tank Inerting System is nothing new (since the 50’s); it is associated mostly with military aircrafts particularly when flying in combat zone. Up to now most modern jetliner including B747 do not have this system.

The FAA tested the system you mentioned back in the 70’s and didn’t make it mandatory because of the advantages vs. cost are not favorable, but recently, (after the accident) new developments brought back the possibility to incorporate this system in commercial wide body jetliners particularly the ones with center fuel tank. Meanwhile, we just wait for the final word from the FAA.

[bsieker]

cyrus15, thanks for the reply.

I stand corrected on the car fuel pumps, I appreciate the links.

Now back to relevant matters.

Originally Posted by cyrus15

Here’s the deal: check out the MSDS link for JP-8
http://www.hess.com/ehs/msds/JP8_HOV_4088_clr.pdf

Yes, that is the document I was referring to.

The vapor pressure is 0.029 psia, that is equivalent to 200 Pascal (Pa) you mentioned but that is in Absolute pressure reading (almost in vacuum condition). Meaning; you can’t mix absolute pressure and atmospheric pressure readings without the proper conversion.

That is only partially correct. It is an absolute pressure if the gas is the only gas in the "atmosphere", but if there is another gas present, this value is also the partial pressure of that substance in the gas mixture. That's how gases work.

To do so, we should get the vapor pressure at the corrected temperature added to the ambient pressure at also the corrected temperature and altitude to get the pressure inside the fuel tank at that altitude.

What would create the ambient pressure inside the tank, to which you add the vapor pressure?

This only applies under very specific conditions, namely that there was atmopheric pressure in the vessel to begin with, created by some other gas, then you add some JP-8. In that case you are correct, we have to add both pressures, and some portiion of the vapor/air mixture would be vented to equalise inside and outside pressure.

So, according to your figures (assuming it is temperature and altitude corrected) we should have 200Pa + 50000Pa to get the pressure inside the fuel tank at 13000 feet, that's why tank vent is needed to allow the 50000 Pa to get in the tank though in real life, the tank pressure is little higher than ambient pressure because of the vapor pressure, that’s the correct solution to the exercise.

So, what is creating the 50,000Pa inside the fuel tank, if not air? (Hey, finally I also got to use that smiley! )

I guess it must be Phlogiston.



Bernd

[polzin]

This is a terrific discussion and I hate to admit maybe beyond my comprehension. But one thing still bothers me, the FBI didn't have time to do any of these tests or analyisis before they decided that it was a problem with the aircraft.
True? Or over the years have I forgotton the time line? Honestly don't know.
any comments....Thanks in advance.
kp

[PBL]

kp,

Originally Posted by kp

But one thing still bothers me, the FBI didn't have time to do any of these tests or analyisis before they decided that it was a problem with the aircraft.

Whether they had time or not, they didn't do it. The FBI were investigating a possible crime, not trying to work out the physics and chemistry of fuel tanks.

If you take a look at the docket, which is quite large, contained multiple-hundred-page reports by NASA and so on, you will find that most of that substantial amount of work by civil agencies (more than just the NTSB) on fuel tank flammability and ignition sources took place after the civilian agencies were allowed to start their investigation.

PBL

[cyrus15]

attention Bernd

Ok, let’s take a look at an empty tank on ground; let’s say it is 3 x 3 x3 meter 27 cubic m in volume with vent. As it is, it is full of air at the same atmospheric pressure say 101320 Pa (14.7 psig) so we have inside and out side pressure of that tank is 14.7psi. We are going to consider the major factors (mostly the physical properties listed in the MSDS).

Say we fill it up with 3cubic meter JP-8, we should have 24 cubic meters empty, but it is still occupied by air at the same pressure on ground (101320 Pa standard atmospheric pressure) and the remaining 3 cubic meter of air has been displaced by the fuel volume.

Let’s take that tank up to 13000 feet (flying fuel tank….!) now we have 50000Pa atmospheric pressure at that height in the tank acting on the surface of the fuel as well as all surfaces, so the rest of the atmospheric pressure (101320 Pa) has gone through the vent to the atmosphere leaving 24 cubic meters of less dense air inside the tank (less oxygen less other gases in the air) plus whatever fuel fume that would escape till the pressure equalize.

Now we have an enclosure with 3 layers of fuel in a liquid form, fuel vapor (heavier than air) and the rest is thinner air at that pressure / altitude / temperature combination. Say at that altitude you opened the filling cap of that tank and you dropped a burning wood stick (source of fire) in the tank, you will burn the air / fuel vapor mixture only in the area where it is combustible (the upper area of the vapor layer and the lower area of the air layer) and the fire well die very quickly leaving hot exhaust with insignificant amount of pressure escaping to the atmosphere.

That combustion is extremely inefficient producing inherently low energy yield because of improper fuel air ratio and also equally important the amount of air available in the tank that can support a significant fire, let alone explosion.

If you take a cup half filled with gasoline and you light it up with a lighter you will have a “poof” like fire and it will die quickly (please don’t try this at home) why the fire will not continue burning? Because the combustible layer is crudely atomized with air in unsustainable fashion.

fuel pump (booster, transfer or scavenge pump) are submerged under the fuel, even when that tank is almost empty, it is still well under the fuel vapor layer where no oxygen reaching that part of the tank even on ground because the vapor density of the fuel is higher than air, it will stay at the lower area of the tank.

The vapor pressure plus the evaporating rate of the fuel in the tank will significantly increase in low atmospheric pressure condition at that altitude, in effect, it will increase the pressure inside the tank and push the air out in proportion to the volume it create. Add to this the fuel temperature effect where further more increases the pressure as it rise. So at this point we have even less air volume in the tank and more fuel vapor layer because of the high vapor density.

Another major factor to consider is fuel atomization. When we talk about combustion we are referring to mixing fuel and air in a ratio that can produce a desired amount of heat out of burning, in real life fuel tanks are not designed for this. That's why Nick Charles is disputing the documentary he have seen and asking why those guys having air circulation fans in the tank that help atomizing fuel by creating air fuel mixture that can produce the desired results the producer want in his “hollywoodish” documentary, and in the process he insult the intelligence of the viewer, real life fuel tank is not like this.

It is very much the norm to conceder 1 part gasoline mixed (in a carburetor) with 14.7 part of air then subjected to compression (using piston) in hope to reach a critical mass where oxygen and fuel are well combined to produce a useful combustion in piston engine after ignition at sea level. This is just to give an idea of fuel/air ratio for something we know. For anybody who fly a Cessna (none fuel injected) he would be very much familiar with lean and rich fuel / air mixture, and when to lean out the mixture and when not. Basically it is adjusting the proper air/fuel ratio to optimize the combustion to produce useful power out of the engine when the altitude changes to accommodate the changing oxygen amount (actually oxygen density). Fuel tank doesn’t have pistons or atomization aid and it doesn’t change its volume.

Speaking of volume; so if we have 24 cubic meter of air at that altitude (thinner, low in oxygen content in comparison to sea level) and 3 cubic meter of JP-8 knowing that they are in layers with insignificant atomization, what would be the amount of energy that it can be produced to blow up the tank? Are fuel tank that frail?.

Fuel tank is very damp enclosure (water vapor mixed with fuel vapor) particularly in coastal area where the humidity is high. Aircrafts are known for having water collected in the tank due to the condensation of water vapor in the air that occupies the empty space in the tank. The accumulated water creates vapor (low pressure effect) making fire starting and propagation harder. In smaller aircraft, water level inspection in the fuel tank is part of the pre-flight procedure by actually taking sample.

In real life we are not anywhere near to have a significant fire in a fuel tank at that altitude neither the condition is conducive.

At this point, I will let you in a little secret:
B747 CWT vent is equipped with a float valve that blocks the vent in case of overfilling condition. Also, it will not allow fuel to come out when the aircraft is banking. The venting channels are interlinked with the other tanks (in the wing) and connected to a device called “surge box“at the wing tip and then to the atmosphere. This box regulates the pressure fluctuation and also keeps a little positive pressure inside the tank to reduce foaming. Further more, this will facilitate more fuel fume to escape to the wing tank and from there to the atmosphere during pressure changing, also allowing less air coming back to the center tank to equalize the pressure.

This design is approved by the FAA and Boeing engineers. As I said earlier about the inerting system, they didn’t see much of advantages for such a system, and the condition of the fuel tank designs are not notoriously life threatening knowing that jetliners are different breed of birds than military and aerobatics ones. There is so much DOT regulation on fuel and flammable vessels designs and safety that also didn’t change much the principle fuel tank design.

Reality dictate, hence the term “theory and practice”. You will be surprise of how many aircraft lost because of “fuel tank explosion” in mid air, search it, check out the FAA website (I don’t have the link right now) for the numbers. Safety shouldn’t be compromise for any reason and in my book any preventable loses of human life is not excusable.

The thing that troubled me the most is the elaborate effort using esoteric science to prove some thing scientifically not sound, tested over in real life for decades and proven it is no issue. At the end, people believe what they want.

[fueldrinker]

CWT and other fuel tank explosions?

Is this just a Boeing problem or have other aircraft suffered?

[IGh]

Ullage Deflagration: Other tanks, other a/c makers

Fueldrinker asked about other tank deflagrations suffered by non-TBC airliners (above).

The AAR for TWA800 has an Appendix G that list some earlier cases, that cites Douglas cases (omitted many other examples).

The first modern investigation of an airliner ullage deflagration was in 1959, done by the Italian authority. That was Lockheed:


TWA 891 / 26Jun59, TWA Super Constellation L-1649-A N7313C flying near electrical storm. Broke-up in flight. Near Milan Italy, Vicinity of Olgiato Olona, Varese Province Italy....

... At disintegration altitude was between 11000' and 12000'; speed about 170 KIAS. Experiments concluded that critical variable existed for gasoline vapor ignition: tests in presence of non-static electrical discharge showed vapors ignite only if plane were CLIMBING. Uncertainty, vapor ignition tests only showed that "hazard cannot be excluded and vapors would actually ignite if the electrical discharge were non-static and sufficiently intense." Witnesses near site observed a lightning stroke, sound of explosion, then saw the fall of the plane's burning wreckage....

P.C. = from seven hypotheses , process of elimination discarded six; the remaining hypothesis deemed probable: Breakup due to explosion of fuel vapors in Fuel Tank #7, followed by excess pressure or further explosion in Tank #6; #7 Tank fuel vapors set-off by ignition of gasoline vapors from static electricity discharges (streamer corona) that developed at vent outlet.

Recommendation: More research and tests of fuel tank explosion, and instructions to pilots to avoid flight through such hazardous weather....

... Rpt of Board of Inquiry of Ministry of Defense, Republic of Italy, Rpt date Nov '60 (USA's CAA AAR File # 1-0045).]

= = = = / / / = = =

Here's another non-Boeing example:

British Eagle Int'l Airlines (BEIA) / 30Aug66 Britannia G-ARKA London, during pressure refueling at 0015 hrs an explosion occurred within starboard wing causing tank rupture and limited structural damage.

A/c had been on ground nine hours since flight, heavy rain at time of incident; crew onboard doing pre-flight checks. Few obvious ignition sources in region of tank, electrostatic discharge suspected within tank. First known incident of "ESD" in a civil a/c [numerous Canadian military a/c had suffered such tank explosions].

Conclusions from RAE Tech Rpt 67080, Apr'67, "Investigation into a Fuel Tank Explosion, Britannia G-ARKA" by M.E. Rogers & H.W.G. Wyeth reveal lack of evidence:

"... mist or foam generated by refuelling must have been ignited by an electrostatic discharge in the tank ullage. There is no positive evidence to support this choice of ignition source and, in fact, conditions seemed unsuitable for dangerous charge accumulation at the time of this incident. However all other ignition sources are discounted."

CRC Nov'75 "Aviation Fuel Safety" states this ullage as JetA vapors, low order tank explosion, moderate damage.

[The AAIB had several observers there at Calverton (wreckage collection at the old F-14 factory), Dave King managed to get co-workers to do a search, AAIB faxed that RAE rpt within a few hours. AAIB guys told us they were surprised that AAIB had passed that case to the RAE for investigation.]

[clearedtocross]

Half and half

Bristol Britannia did run on jet fuel (Proteus turboprops), but the Super Constellations ran and still run on AVGAS (gasoline).

[skiingman]

This is old, but I'm a pedant:

Originally Posted by bsieker

Secondly, car fuel pumps are usually not submerged in the tank, but somewhere down the line, which their electrical parts away from the fuel line.

This has been largely false for twenty five years or so. Even early EFI cars with downstream pumps usually had an in tank pump as well. Most modern fuel injected gasoline cars have a sole pump in the tank, submerged in fuel when full. Direct injection systems go back to the old days of engine driven pumps, but still have in tank pumps AFAIK.

Also, you speak of venting. In the context of gasoline automobiles, evaporative emissions are controlled by simple and robust systems. The moisture and fuel contamination problems that kill people in aircraft are extremely uncommon in modern automobiles. I have a vehicle I drive twice a month or so that runs on the same tank of fuel for months with zero issue. (It also does other things that appear to be magic compared to most light aircraft, such as always starting on the first try, never fouling plugs, consuming practically no oil at 15 years old, etc.)

Jet-A, of course, works much differently. I would be very interested in learning how much HC leaves a large airliner in typical use. I would assume the evaporative part is minuscule compared to the poorly combusted part, but I'm curious nonetheless. I expect refueling is the biggest offender.

[35hPA28]

Discovery Channel documentaries

On a side note, DC documentaries can be quite pasteurized. I recall watching one ( Mayday/Crash Scene Investigation) about AA965 and no mention at all was made about any discussion between pilots and cabin crew union/ERC representative on board regarding delays at Miami and possible too long on-the-clock time leading to the need for the crew to take an extra day´s rest before the return flight, as per regulations. IIRC, that was pointed as a contributing factor to the accident.
Added to clarify: ...pointed as a contributing factor by investigators.