FORMATION/FLAMMABILITY OF J ET A AEROSOL IN AIRCRAFT F UEL TANKS
David Koert∗
Department of Mechanical Engineering, Wichita State University, Wichita, KS
EMAIL: David.Koert@wichita.edu
Kevin Loss and Nathaniel Reynolds
Boeing Commercial Airplanes, Wichita, KS
The ignition and explosion dynamics of a fuel vapor-droplet/air mixture in aircraft fuel tank ullage are
significantly different from that of vapor-only/air. In particular, the presence of fuel droplets can broaden the
flammability range as a result of the lowering of the lean flammability limit. Conditions in an empty aircraft fuel
tank may facilitate formation of droplets. A cloud of suspended fuel droplets may be formed when fuel residue in
an essentially empty tank is: 1) heated before take-off and then cooled while climbing after take -off, or 2) heated
from the bottom of the tank while the top surface of the tank is cooled. An experimental program to study the
formation of suspended fuel droplets in a large temperature-, pressure- regulated fuel tank segment has begun in
order to simulate fuel tank vapor dynamics leading to the formation of an explosive fuel vapor-droplet/air
mixture.
Fuel Tank Flammability - Flammability of Fuel-Vapor/Air Mixtures
When evaluating the risk of explosion in fuel tanks of commercial jet aircraft, discussion generally focuses
on the flammability envelope that lies between the lean flammability limit and the rich flammability limit for a
homogeneous, gas-phase mixture of fuel-vapor and air in the ullage. Since these limits are based on conditions of
thermodynamic equilibrium they are sometimes referred to as static limits of flammability. Dynamic (nonequilibrium) conditions can lead to wider flammability envelopes. Static flammability limits depend on the fuelvapor and oxygen concentrations. Since the fuel tanks are vented to the atmosphere surrounding the aircraft, the
concentration of oxygen depends on altitude. Because the fuel saturation pressure limits the equilibrium
concentration of the evaporated fuel vapor, the fuel concentration depends on ullage temperature. Thus, altitudetemperature flammability envelopes can be predicted from: 1) the flammability limits of jet fuel, 2) the saturation
pressure of jet fuel, and 3) the variation of the temperature and pressure of the standard atmosphere with altitude.
Equations for these properties are then used to develop altitude-temperature flammability envelopes, which are
often used in the discussion of fuel tank safety. Typical envelopes (Kuchta, 1973) are shown in Figure 1.
Figure 1. Altitude-Temperature Flammability Limits for Jet Fuels [Kuchta, 1973]
A reasonable way of using the information in these flammability envelopes in regards to flight safety is to
locate fuel tank ullage and venting temperatures at altitude. If they fall outside the flammability envelope for a
given fuel, then safe conditions exist.
∗
Corresponding author
Altitude-temperature flammability envelopes are misleading is due to the fact that flammable conditions may
exist well outside the flammability envelope (Scull, 1951; Barnett and Hibbard, 1956). This is because the
calculations used to determine the flammability envelopes are based on conditions of thermodynamic equilibrium,
whereas conditions within the fuel tank and venting system are often quite dynamic. For more volatile fuels such
as Avgas, where tank conditions are more likely to be above the rich limit, flammable mixtures can result if the
tank breathes in air. For low volatility fuels such as Jet A where tank conditions are more likely to be below the
lean limit, flammable mixtures can result if conditions exist under which mists and sprays are formed.
Fuel Tank Flammability - Flammability of Fuel–Droplet–Vapor/Air Mixtures
For the low volatility fuels such as Jet A, thermodynamic conditions within aircraft fuel systems are often
below the lower limit of the equilibrium-based flammability envelope. Within this region, the risk of fire and
explosion is still present when dynamic conditions leading to the formation of aerosols exist. Nester (1967) and
Ott (1970) have investigated the effects of mists and sprays on the lean flammability limit of jet fuels. Both
investigators demonstrated that the presence of suspended fuel droplets can significantly lower the lean limit. Ott
concluded that, “the application of static flammability limits to the dynamic environment found in aircraft fuel
systems can lead to gross errors in the assessment of fuel tank fire and explosion hazards.” Ott also noted that
two distinct dynamic conditions occurring in aircraft fuel systems can result in the formation of fuel mists: (1)
sloshing and vibration (agitation), and (2) pressure changes (homogeneous condensation).
Both Nester (1967) and Ott (1970) investigated the flammability of mists and sprays created by mechanical
agitation of small amounts of fuel in the bottom of relatively small tanks (2 ½ gal. and 80 gal., respectively). The
mechanical agitation required to produce the mists and sprays reported by Ott were characterized by sloshing (1020 CPM at double amplitudes of 16º and 30º) and vibration (400-3200 CPM at amplitudes up to 0.05 in.).
Ignition limits were determined by measuring the pressure rise in the tank after an ignition attempt with a spark
igniter. Both researchers reported that the lean flammability limit temperatures could be lowered by as much as
60ºF.
No attempt was made by either Nester or Ott to characterize the mists/sprays even though the researchers
acknowledged that droplet number density and droplet size distribution were important parameters in assessing
the flammability limits of fuel vapor- droplet- air mixtures. Although not reported directly, one supposes that the
mist/spray formed by mechanical agitation could be rather short lived. Many of the droplets formed by
mechanical means could be too large to remain suspended, and thus fall due to the effects of gravity at varying
rates depending on size. Ott reports that during testing, sloshing and vibration continued during ignition attempts
– a necessity, perhaps, if the droplets formed by agitation were too large to remain suspended. In consideration
of these studies, one can only conclude that little is known about the stability of aerosols formed in aircraft fuel
systems.
When considering the stability of fuel mists/sprays formed by mechanical agitation, the chemical
composition of the droplets also comes into question. Since the droplets are presumably formed from the breakup
and dispersion of fuel in the liquid state, the chemical composition of droplets is likely to be essentially that of the
fuel, at least initially. Once formed, however, a droplet may be susceptible to a change in composition due to
evaporation depending on the amount of time it remains suspended and the degree of saturation in the fuelvapor/air mixture. If given sufficient time, lower molecular weight components would evaporate due to their
volatility thus shifting the composition of the suspended droplet toward the higher molecular weight components
of fuel. No attempt was made in the investigations of Nester or Ott to characterize the chemical composition of
fuel mist.
Fuel Cloud Formation in Fuel Tank Ullage by Homogeneous Condensation
As stated previously, Ott (1970) mentioned a second mechanism for the formation of fuel mists in aircraft
fuel tanks. In particular, Ott states that “pressure changes (reductions) can result in the formation of mists which
are relatively stable and may change the flammability characteristics of the ullage gases.” The mechanism for
mist formation which Ott calls “pressure changes (reductions)” undoubtedly refers to a process more accurately
described as homogeneous condensation. To date, no report of the results of a direct study of homogeneous
nucleation of jet fuel vapors is available.
Key Questions
The preceding discussions developed several important concepts, which provide the justification for the
investigation. These concepts are summarized and several key questions are highlighted.
The application of equilibrium-based flammability envelopes to the dynamic environment of aircraft fuel
systems can lead to serious errors in the assessment of the risk of fire and explosion. In particular, the risk of fire
and explosion is still present when dynamic (non-equilibrium) conditions such as those leading to the formation
of mists or sprays exist. In fact, when droplets of sufficient size and density distribution are present, fuel tank
ullage at any lean fuel-vapor/air ratio can be flammable, even in an atmosphere without any premixed fuel vapor.
Two distinct dynamic conditions occurring in aircraft fuel systems can result in the formation of fuel mists: (1)
sloshing and vibration (agitation), and (2) homogeneous condensation. Although homogenous condensation is a
common mode of condensation that can lead to the formation of a fuel fog, no report on the results of a study of
homogeneous nucleation of jet fuel vapors is available. Two situations conducive to droplet formation by
homogenous nucleation may occur often enough in fuel tanks that they warrant investigation: 1) diabatic
decompression of the ullage space, typically encountered whenever an aircraft climbs over an extended period
such as at take-off; and 2) situations where the fuel in the bottom of a tank is at a higher temperature than the top
surface of the tank, typically encountered when the aircraft sits on the ground on a hot day with the air
conditioning running in the passenger compartment. Little is known about tank vapor dynamics related to the
stability of aerosols formed in aircraft fuel systems. In view of the issues highlighted in this summary, the
following questions arise:
1) What are the conditions that arise in aircraft fuel tanks that lead to the formation of fuel mists due to
homogeneous nucleation of jet fuel vapor?
2) What is the droplet number density and size distribution of a fuel mist formed by homogeneous nucleation
of jet fuel vapor in an aircraft fuel tank?
3) What is the chemical composition of the droplets in a fuel mist formed by homogeneous nucleation of jet
fuel vapor in an aircraft fuel tank?
Fuel Tank Test Facility
A facility for the study of the formation of suspended fuel droplets (a fuel cloud) in fuel tank Ullage has been
built and is described in the following. Reference should be made to the Figure 2 for clarification of the details of
design. A 7'x3'x3' tank, constructed of steel, was built to model bay five of a Boeing 747 center wing tank at onehalf scale. A steel skin is welded over a steel framework to assure sufficient rigidity for optical ports during
operation, where tank pressure may be reduced to as-low-as 10 psi vac. Optical ports, located on opposite sides
of the tank, provide access for the droplet sizing/velocimetry instrumentation. A water jet exhauster is used to pull
a vacuum on the tank. Process heater/chillers using circulating water/glycol provide heating and cooling of
Figure 2. Fuel Tank Test Facility for Studies of Homogenous Nucleation of Jet Fuel Vapors.
various surfaces via cooling jackets in and on the tank. The jet ejector and heater/chillers are sized to allow the
tank to match temperature-pressure profiles at climb rates up to 1000 ft/min. An open reservoir for jet fuel is
located in the bottom of the tank.
Experimental Program
Experimental studies of homogeneous condensation during tank decompression will be conducted with the
goal of: 1) matching the decompression rate over a range of typical rates of assent, and 2) simulating tank wall
temperatures and heat transfer rates to promote realistic tank hydrodynamics. In order to conduct studies of this
type, spatially resolved measurements of droplet size and number density will be made, as well as measurements
of the velocity field. Several series of experiments are planned. A series of experiments with hydrodynamically
stable conditions (uniform wall temperature) to promote fog formation will be conducted in order to characterize
droplet size distribution and number density of fuel fog. Another series of experiments with realistic wall
temperatures and heat transfer rates will be conducted to determine likely scenarios for fog formation during
preflight- and flight- operations. Lastly, a series of experiments will be conducted to identify remediation
methods that remove or prevent fuel fog by adjustment of wall temperature and heat transfer rates. To date,
preliminary studies have begun to identify tank conditions where rapid formation of fuel fog occurs.
References
Barnett, H. C. and Hibbard, R. R. (1956), “Properties of Aircraft Fuels,” NACA TN 3276.
Kuchta, J. M. (1973), “Fire and Explosion Manual for Aircraft Explosion Investigators,” Technical Report
AFAPL-TR-73-74, Air Force Aero Propulsion Laboratory, Air Force Systems Command, Wright-Patterson
Air Force Base, Ohio.
Nester, L. (1967), “Investigation of Turbine Fuel Flammability within Aircraft Fuel Tanks,” Final Report DS-677, Naval Air Propulsion Test Center, Philadelphia.
Ott E. (1970), “Effects of Fuel Slosh and Vibration on the Flammability Hazards of Hydrocarbon Turbine Fuels
within Aircraft Fuel Tanks,” Technical Report AFAPL-TR-70-65, Fire Protection Branch of the Fuels and
Lubrication Division, Wright-Patterson Air Force Base, Ohio.
Scull, W. E. (1951), “Relation between Inflammables and Ignition Sources in Aircraft Environments,” Report
1019, Lewis Flight Propulsion Laboratory, Cleveland, Ohio.