[Federal Register Volume 70, Number 30 (Tuesday, February 15, 2005)]
[Rules and Regulations]
[Pages 7800-7827]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 05-2752]



[[Page 7799]]

-----------------------------------------------------------------------

Part II





Department of Transportation





-----------------------------------------------------------------------



Federal Aviation Administration



-----------------------------------------------------------------------



14 CFR Part 25



Special Conditions: Boeing Model 747-100/200B/200F/200C/SR/SP/100B/300/
100B SUD/400/400D/400F Airplanes; Flammability Reduction Means (Fuel 
Tank Inerting); Final Special Conditions; Rule

  Federal Register / Vol. 70, No. 30 / Tuesday, February 15, 2005 / 
Rules and Regulations  

[[Page 7800]]


-----------------------------------------------------------------------

DEPARTMENT OF TRANSPORTATION

Federal Aviation Administration

14 CFR Part 25

[Docket No. NM270; Special Conditions No. 25-285-SC]


Special Conditions: Boeing Model 747-100/200B/200F/200C/SR/SP/
100B/300/ 100B SUD/400/400D/400F Airplanes; Flammability Reduction 
Means (Fuel Tank Inerting)

AGENCY: Federal Aviation Administration (FAA), DOT.

ACTION: Final special conditions.

-----------------------------------------------------------------------

SUMMARY: These special conditions are issued for the Boeing Model 747-
100/200B/200F/200C/SR/SP/100B/300/100B SUD/400/400D/400F series 
airplanes. These airplanes, as modified by Boeing Commercial Airplanes, 
include a new flammability reduction means that uses a nitrogen 
generation system to reduce the oxygen content in the center wing fuel 
tank so that exposure to a combustible mixture of fuel and air is 
substantially minimized. This system is intended to reduce the average 
flammability exposure of the fleet of airplanes with the system 
installed to a level equivalent to 3 percent of the airplane operating 
time. The applicable airworthiness regulations do not contain adequate 
or appropriate safety standards for the design and installation of this 
system. These special conditions contain the additional safety 
standards the Administrator considers necessary to ensure an acceptable 
level of safety for the installation of the system and to define 
performance objectives the system must achieve to be considered an 
acceptable means for minimizing development of flammable vapors in the 
fuel tank installation.

DATES: The effective date of these special conditions is March 17, 
2005.

FOR FURTHER INFORMATION CONTACT: Mike Dostert, Propulsion and 
Mechanical Systems Branch, FAA, ANM-112, Transport Airplane 
Directorate, Aircraft Certification Service, 1601 Lind Avenue SW., 
Renton, Washington 98055-4056; telephone (425) 227-2132, facsimile 
(425) 227-1320, e-mail [email protected].

SUPPLEMENTARY INFORMATION:

Background

    Boeing Commercial Airplanes intends to modify Model 747 series 
airplanes to incorporate a new flammability reduction means (FRM) that 
will inert the center fuel tanks with nitrogen-enriched air (NEA). 
Though the provisions of Sec.  25.981, as amended by amendment 25-102, 
will apply to this design change, these special conditions address 
novel design features.
    Regulations used as the standard for certification of transport 
category airplanes prior to amendment 25-102, effective June 6, 2001, 
were intended to prevent fuel tank explosions by eliminating possible 
ignition sources from inside the fuel tanks. Service experience of 
airplanes certificated to the earlier standards shows that ignition 
source prevention alone has not been totally effective at preventing 
accidents. Commercial transport airplane fuel tank safety requirements 
have remained relatively unchanged throughout the evolution of piston-
powered airplanes and later into the jet age. The fundamental premise 
for precluding fuel tank explosions has involved establishing that the 
design does not result in a condition that would cause an ignition 
source within the fuel tank ullage (the space in the tank occupied by 
fuel vapor and air). A basic assumption in this approach has been that 
the fuel tank could contain flammable vapors under a wide range of 
airplane operating conditions, even though there were periods of time 
in which the vapor space would not support combustion.

Fuel Properties

    Jet fuel vapors are flammable in certain temperature and pressure 
ranges. The flammability temperature range of jet engine fuel vapors 
varies with the type and properties of the fuel, the ambient pressure 
in the tank, and the amount of dissolved oxygen released from the fuel 
into the tank. The amount of dissolved oxygen in a tank will also vary 
depending on the amount of vibration and sloshing of the fuel that 
occurs within the tank.
    Jet A fuel is the most commonly used commercial jet fuel in the 
United States. Jet A-1 fuel is commonly used in other parts of the 
world. At sea level and with no sloshing or vibration present, these 
fuels have flammability characteristics such that insufficient 
hydrocarbon molecules will be present in the fuel vapor-air mixture, to 
ignite when the temperature in the fuel tank is below approximately 100 
[deg]F. Too many hydrocarbon molecules will be present in the vapor to 
allow it to ignite when the fuel temperature is above approximately 175 
[deg]F. The temperature range where a flammable fuel vapor will form 
can vary with different batches of fuel, even for a specific fuel type. 
In between these temperatures the fuel vapor is flammable. This 
flammability temperature range decreases as the airplane gains altitude 
because of the corresponding decrease of internal tank air pressure. 
For example, at an altitude of 30,000 feet, the flammability 
temperature range is about 60 [deg]F to 120 [deg]F.
    Most transport category airplanes used in air carrier service are 
approved for operation at altitudes from sea level to 45,000 feet. 
Those airplanes operated in the United States and in most overseas 
locations use Jet A or Jet A-1 fuel, which typically limits exposure to 
operation in the flammability range to warmer days.
    We have always assumed that airplanes would sometimes be operated 
with flammable fuel vapors in their fuel tank ullage (the space in the 
tank occupied by fuel vapor and air).

Fire Triangle

    Three conditions must be present in a fuel tank to support 
combustion. These include the presence of a suitable amount of fuel 
vapor, the presence of sufficient oxygen, and the presence of an 
ignition source. This has been named the ``fire triangle.'' Each point 
of the triangle represents one of these conditions. Because of 
technological limitations in the past, the FAA philosophy regarding the 
prevention of fuel tank explosions to ensure airplane safety was to 
only preclude ignition sources within fuel tanks. This philosophy 
included application of fail-safe design requirements to fuel tank 
components (lightning design requirements, fuel tank wiring, fuel tank 
temperature limits, etc.) that are intended to preclude ignition 
sources from being present in fuel tanks even when component failures 
occur.

Need To Address Flammability

    Three accidents have occurred in the last 13 years as the result of 
unknown ignition sources within the fuel tank in spite of past efforts, 
highlighting the difficulty in continuously preventing ignition from 
occurring within fuel tanks. Between 1996 and 2000 the National 
Transportation Safety Board (NTSB) issued recommendations to improve 
fuel tank safety that included prevention of ignition sources and 
addressing fuel tank flammability (i.e., the other two points of the 
fire triangle).
    The FAA initiated safety reviews of all larger transport airplane 
type certificates to review the fail-safe features of previously 
approved designs and also initiated research into the feasibility of 
amending the regulations to address fuel tank flammability. Results 
from the safety reviews indicated a significant number of single

[[Page 7801]]

and combinations of failures that can result in ignition sources within 
the fuel tanks. The FAA has adopted rulemaking to require design and/or 
maintenance actions to address these issues; however, past experience 
indicates unforeseen design and maintenance errors can result in 
development of ignition sources. These findings show minimizing or 
preventing the formation of flammable vapors by addressing the 
flammability points of the fire triangle will enhance fuel tank safety.
    On April 3, 1997, the FAA published a notice in the Federal 
Register (62 FR 16014), Fuel Tank Ignition Prevention Measures, that 
requested comments concerning the 1996 NTSB recommendations regarding 
reduced flammability. That notice provided significant discussion of 
the service history, background, and issues related to reducing 
flammability in transport airplane fuel tanks. Comments submitted to 
that notice indicated additional information was needed before the FAA 
could initiate rulemaking action to address all of the recommendations.
    Past safety initiatives by the FAA and industry to reduce the 
likelihood of fuel tank explosions resulting from post crash ground 
fires have evaluated means to address other factors of the fire 
triangle. Previous attempts were made to develop commercially viable 
systems or features that would reduce or eliminate other aspects of the 
fire triangle (fuel or oxygen) such as fuel tank inerting or ullage 
space vapor ``scrubbing'' (ventilating the tank ullage with air to 
remove fuel vapor to prevent the accumulation of flammable 
concentrations of fuel vapor). Those initial attempts proved to be 
impractical for commercial transport airplanes due to the weight, 
complexity, and poor reliability of the systems, or undesirable 
secondary effects such as unacceptable atmospheric pollution.

Fuel Tank Harmonization Working Group

    On January 23, 1998, the FAA published a notice in the Federal 
Register that established an Aviation Rulemaking Advisory Committee 
(ARAC) working group, the Fuel Tank Harmonization Working Group 
(FTHWG). The FAA tasked the FTHWG with providing a report to the FAA 
recommending regulatory text to address limiting fuel tank flammability 
in both new type certificates and the fleet of in service airplanes. 
The ARAC consists of interested parties, including the public, and 
provides a public process to advise the FAA concerning development of 
new regulations. [Note: The FAA formally established ARAC in 1991 (56 
FR 2190, January 22, 1991), to provide advice and recommendations 
concerning the full range of the FAA's safety-related rulemaking 
activity.]
    The FTHWG evaluated numerous possible means of reducing or 
eliminating hazards associated with explosive vapors in fuel tanks. On 
July 23, 1998, the ARAC submitted its report to the FAA. The full 
report is in the docket created for this ARAC working group (Docket No. 
FAA-1998-4183). This docket can be reviewed on the U.S. Department of 
Transportation electronic Document Management System on the Internet at 
http://dms.dot.gov.
    The report provided a recommendation for the FAA to initiate 
rulemaking action to amend Sec.  25.981, applicable to new type design 
airplanes, to include a requirement to limit the time transport 
airplane fuel tanks could operate with flammable vapors in the vapor 
space of the tank. The recommended regulatory text proposed, ``Limiting 
the development of flammable conditions in the fuel tanks, based on the 
intended fuel types, to less than 7 percent of the expected fleet 
operational time (defined in this rule as flammability exposure 
evaluation time (FEET)), or providing means to mitigate the effects of 
an ignition of fuel vapors within the fuel tanks such that any damage 
caused by an ignition will not prevent continued safe flight and 
landing.'' The report included a discussion of various options for 
showing compliance with this proposal, including managing heat input to 
the fuel tanks, installation of inerting systems or polyurethane fire 
suppressing foam, and suppressing an explosion if one occurred.
    The level of flammability defined in the proposal was established 
based on a comparison of the safety record of center wing fuel tanks 
that, in certain airplanes, are heated by equipment located under the 
tank, and unheated fuel tanks located in the wing. The ARAC concluded 
that the safety record of fuel tanks located in the wings with a 
flammability exposure of 2 to 4 percent of the FEET was adequate and 
that if the same level could be achieved in center wing fuel tanks, the 
overall safety objective would be achieved. The thermal analyses 
documented in the report revealed that center wing fuel tanks that are 
heated by air conditioning equipment located beneath them contain 
flammable vapors, on a fleet average basis, in the range of 15 to 30 
percent of the fleet operating time.
    During the ARAC review, it was also determined that certain 
airplane types do not locate heat sources adjacent to the fuel tanks 
and have significant surface areas that allow cooling of the fuel tank 
by outside air. These airplanes provide significantly reduced 
flammability exposure, near the 2 to 4 percent value of the wing tanks. 
The group therefore determined that it would be feasible to design new 
airplanes such that airplane operation with fuel tanks that were 
flammable in the flammable range would be limited to nearly that of the 
wing fuel tanks. Findings from the ARAC report indicated that the 
primary method of compliance available at that time with the 
requirement proposed by the ARAC would likely be to control heat 
transfer into and out of fuel tanks. Design features such as locating 
the air conditioning equipment away from the fuel tanks, providing 
ventilation of the air conditioning bay to limit heating and to cool 
fuel tanks, and/or insulating the tanks from heat sources, would be 
practical means of complying with the regulation proposed by the ARAC.
    In addition to its recommendation to revise Sec.  25.981, the ARAC 
also recommended that the FAA continue to evaluate means for minimizing 
the development of flammable vapors within the fuel tanks to determine 
whether other alternatives, such as ground-based inerting of fuel 
tanks, could be shown to be cost effective.
    To address the ARAC recommendations, the FAA continued with 
research and development activity to determine the feasibility of 
requiring inerting for both new and existing designs.

FAA Rulemaking Activity

    Based in part on the ARAC recommendations to limit fuel tank 
flammability exposure on new type designs, the FAA developed and 
published amendment 25-102 in the Federal Register on May 7, 2001 (66 
FR 23085). The amendment included changes to Sec.  25.981 that require 
minimization of fuel tank flammability to address both reduction in the 
time fuel tanks contain flammable vapors, (Sec.  25.981(c)), and 
additional changes regarding prevention of ignition sources in fuel 
tanks. Section 25.981(c) was based on the FTHWG recommendation to 
achieve a safety level equivalent to that achieved by the fleet of 
transports with unheated aluminum wing tanks, between 2 to 4 percent 
flammability. The FAA stated in the preamble to Amendment 25-102 that 
the intent of the rule was to--

    * * * require that practical means, such as transferring heat 
from the fuel tank (e.g., use of ventilation or cooling air), be 
incorporated into the airplane design if heat sources were

[[Page 7802]]

placed in or near the fuel tanks that significantly increased the 
formation of flammable fuel vapors in the tank, or if the tank is 
located in an area of the airplane where little or no cooling 
occurs. The intent of the rule is to require that fuel tanks are not 
heated, and cool at a rate equivalent to that of a wing tank in the 
transport airplane being evaluated. This may require incorporating 
design features to reduce flammability, for example cooling and 
ventilation means or inerting for fuel tanks located in the center 
wing box, horizontal stabilizer, or auxiliary fuel tanks located in 
the cargo compartment.

    Advisory circulars associated with Amendment 25-102 include AC 
25.981-1B, ``Fuel Tank Ignition Source Prevention Guidelines,'' and AC 
25.981-2, ``Fuel Tank Flammability Minimization.'' Like all advisory 
material, these advisory circulars describe an acceptable means, but 
not the only means, for demonstrating compliance with the regulations.

FAA Research

    In addition to the notice published in the Federal Register on 
April 3, 1997, the FAA initiated research to provide a better 
understanding of the ignition process of commercial aviation fuel 
vapors and to explore new concepts for reducing or eliminating the 
presence of flammable fuel air mixtures within fuel tanks.

Fuel Tank Inerting

    In the public comments received in response to the 1997 notice, 
reference was made to hollow fiber membrane technology that had been 
developed and was in use in other applications, such as the medical 
community, to separate oxygen from nitrogen in air. Air is made up of 
about 78 percent nitrogen and 21 percent oxygen, and the hollow fiber 
membrane material uses the absorption difference between the nitrogen 
and oxygen molecules to separate the NEA from the oxygen. In airplane 
applications NEA is produced when pressurized air from an airplane 
source such as the engines is forced through the hollow fibers. The NEA 
is then directed, at appropriate nitrogen concentrations, into the 
ullage space of fuel tanks and displaces the normal fuel vapor/air 
mixture in the tank.
    Use of the hollow fiber technology allowed nitrogen to be separated 
from air, which eliminated the need to carry and store the nitrogen in 
the airplane. Researchers were aware of the earlier system's 
shortcomings in the areas of weight, reliability, cost, and 
performance. Recent advances in the technology have resolved those 
concerns and eliminated the need for storing nitrogen on board the 
airplane.

Criteria for Inerting

    Earlier fuel tank inerting designs produced for military 
applications were based on defining ``inert'' as a maximum oxygen 
concentration of 9 percent. This value was established by the military 
for protection of fuel tanks from battle damage. One major finding from 
the FAA's research and development efforts was the determination that 
the 9 percent maximum oxygen concentration level benchmark, established 
to protect military airplanes from high-energy ignition sources 
encountered in battle, was significantly lower than that needed to 
inert civilian transport airplane fuel tanks from ignition sources 
resulting from airplane system failures and malfunctions that have much 
lower energy. This FAA research established a maximum value of 12 
percent as being adequate at sea level. The test results are currently 
available on FAA Web site: http://www.fire.tc.faa.gov/pdf/tn02-79.pdf 
as FAA Technical Note ``Limiting Oxygen Concentrations Required to 
Inert Jet Fuel Vapors Existing at Reduced Fuel Tank Pressures,'' report 
number DOT/FAA/AR-TN02/79. As a result of this research, the quantity 
of NEA that is needed to inert commercial airplane fuel tanks was 
lessened so that an effective FRM can now be smaller and less complex 
than was originally assumed. The 12 percent value is based on the 
limited energy sources associated with an electrical arc that could be 
generated by airplane system failures on typical transport airplanes 
and does not include events such as explosives or hostile fire.
    As previously discussed, existing fuel tank system requirements 
(contained in earlier Civil Air Regulation (CAR) 4b and now in 14 Code 
of Federal Regulations (CFR) part 25) have focused solely on prevention 
of ignition sources. The FRM is intended to add an additional layer of 
safety by reducing the exposure to flammable vapors in the heated 
center wing tank, not necessarily eliminating them under all operating 
conditions. Consequently, ignition prevention measures will still be 
the principal layer of defense in fuel system safety, now augmented by 
substantially reducing the time that flammable vapors are present in 
higher flammability tanks. We expect that by combining these two 
approaches, particularly for tanks with high flammability exposure, 
such as the heated center wing tank or tanks with limited cooling, 
risks for future fuel tank explosions can be substantially reduced.

Boeing Application for Certification of a Fuel Tank Inerting System

    On November 15, 2002, Boeing Commercial Airplanes applied for a 
change to Type Certificate A20WE to modify Model 747-100/200B/200F/
200C/SR/SP/100B/300/100B SUD/400/400D/400F series airplanes to 
incorporate a new FRM that inerts the center fuel tanks with NEA. These 
airplanes, approved under Type Certificate No. A20WE, are four-engine 
transport airplanes with a passenger capacity up to 624, depending on 
the submodel. These airplanes have an approximate maximum gross weight 
of 910,000 lbs with an operating range up to 7,700 miles.

Type Certification Basis

    Under the provisions of Sec.  21.101, Boeing Commercial Airplanes 
must show that the Model 747-100/200B/200F/200C/SR/SP/100B/300/100B 
SUD/400/400D/400F series airplanes, as changed, continue to meet the 
applicable provisions of the regulations incorporated by reference in 
Type Certificate No. A20WE, or the applicable regulations in effect on 
the date of application for the change. The regulations incorporated by 
reference in the type certificate are commonly referred to as the 
``original type certification basis.'' The regulations incorporated by 
reference in Type Certificate A20WE include 14 CFR part 25, dated 
February 1, 1965, as amended by Amendments 25-1 through 25-70, except 
for special conditions and exceptions noted in Type Certificate Data 
Sheet A20WE.
    In addition, if the regulations incorporated by reference do not 
provide adequate standards with respect to the change, the applicant 
must comply with certain regulations in effect on the date of 
application for the change. The FAA has determined that the FRM 
installation on the Boeing Model 747-100/200B/200F/200C/SR/SP/100B/300/
100B SUD/400/400D/400F series airplanes must also be shown to comply 
with Sec.  25.981 at amendment 25-102.
    If the Administrator finds that the applicable airworthiness 
regulations (14 CFR part 25) do not contain adequate or appropriate 
safety standards for the Boeing Model 747-100/200B/200F/200C/SR/SP/
100B/300/100B SUD/400/400D/400F series airplanes because of a novel or 
unusual design feature, special conditions are prescribed under the 
provisions of Sec.  21.16.
    In addition to the applicable airworthiness regulations and special 
conditions, the Model 747-100/200B/

[[Page 7803]]

200F/200C/SR/SP/100B/300/100B SUD/400/400D/400F series airplanes must 
comply with the fuel vent and exhaust emission requirements of 14 CFR 
part 34 and the acoustical change requirements of Sec.  21.93(b).
    Special conditions, as defined in Sec.  11.19, are issued in 
accordance with Sec.  11.38 and become part of the type certification 
basis in accordance with Sec.  21.101.
    Special conditions are initially applicable to the model for which 
they are issued. Should the type certificate for that model be amended 
later to include any other model that incorporates the same or similar 
novel or unusual design feature, or should any other model already 
included on the same type certificate be modified to incorporate the 
same or similar novel or unusual design feature, the special conditions 
would also apply to the other model under the provisions of Sec.  
21.101.

Novel or Unusual Design Features

    Boeing has applied for approval of an FRM to minimize the 
development of flammable vapors in the center fuel tanks of Model 747-
100/200B/200F/200C/SR/SP/100B/300/100B SUD/400/400D/400F series 
airplanes. Boeing also plans to seek approval of this system on Boeing 
Model 737, 757, 767, and 777 airplanes.
    Boeing has proposed to voluntarily comply with Sec.  25.981(c), 
amendment 25-102, which is normally only applicable to new type designs 
or type design changes affecting fuel tank flammability. The provisions 
of Sec.  21.101 require Boeing to also comply with Sec. Sec.  25.981(a) 
and (b), amendment 25-102, for the changed aspects of the airplane by 
showing that the FRM does not introduce any additional potential 
sources of ignition into the fuel tanks.
    The FRM uses a nitrogen generation system (NGS) that comprises a 
bleed-air shutoff valve, ozone converter, heat exchanger, air 
conditioning pack air cooling flow shutoff valve, filter, air 
separation module, temperature regulating valve controller and sensor, 
high-flow descent control valve, float valve, and system ducting. The 
system is located in the air conditioning pack bay below the center 
wing fuel tank. Engine bleed air from the existing engine pneumatic 
bleed source flows through a control valve into an ozone converter and 
then through a heat exchanger, where it is cooled using outside cooling 
air. The cooled air flows through a filter into an air separation 
module (ASM) that generates NEA, which is supplied to the center fuel 
tank, and also discharges oxygen-enriched air (OEA). The OEA from the 
ASM is mixed with cooling air from the heat exchanger to dilute the 
oxygen concentration and then exhausted overboard. The FRM also 
includes modifications to the fuel vent system to minimize dilution of 
the nitrogen-enriched ullage in the center tank due to cross-venting 
characteristics of the existing center wing fuel tank vent design.
    Boeing originally proposed that the system be operated only during 
flight and that the center tank would continue to be inert on landing 
and remain inert during normal ground procedures. Boeing has more 
recently stated that the FRM design may include the capability to be 
operated on the ground.
    Boeing has proposed that limited dispatch relief for operation with 
an inoperative NGS be allowed. Boeing has initially proposed a 10-day 
master minimum equipment list (MMEL) relief for the system. Boeing 
originally proposed that there be no cockpit or maintenance indication 
onboard for the NGS, and that periodic maintenance, using ground 
service equipment, be performed to verify system operation. More 
recently Boeing has stated that to meet operator needs and system 
reliability and availability objectives, built-in test functions would 
be included and system status indication of some kind would be 
provided. In addition, indications would be provided in the cockpit on 
certain airplane models that have engine indicating and crew alerting 
systems. The reliability of the system is expected to be designed to 
achieve a mean time between failure (MTBF) of 5000 hours or better.

Discussion

    The FAA policy for establishing the type design approval basis of 
the FRM design will result in application of Sec. Sec.  25.981(a) and 
(b), amendment 25-102, for the changes to the airplane that might 
increase the risk of ignition of fuel vapors. Boeing will therefore be 
required to substantiate that changes introduced by the FRM will meet 
the ignition prevention requirements of Sec. Sec.  25.981(a) and (b), 
amendment 25-102 and other applicable regulations.
    With respect to compliance with Sec.  25.981(c), AC 25.981-2 
provides guidance in addressing minimization of fuel tank flammability 
within a heated fuel tank, but there are no specific regulations that 
address the design and installation of an FRM that inerts the fuel 
tank. Since amendment 25-102 was adopted, significant advancements in 
inerting technology have reduced the size and complexity of inerting 
systems. Developments in inerting technology have made it practical to 
significantly reduce fuel tank flammability below the levels required 
within the rule. However, due to factors such as the limited 
availability of bleed air and electrical power, it is not considered 
practical at this time to develop systems for retrofit into existing 
airplane designs that can maintain a non-flammable tank ullage in all 
fuel tanks or during all operating conditions. These special conditions 
include additional requirements above that of amendment 25-102 to Sec.  
25.981(c) to minimize fuel tank flammability, such that the level of 
minimization in these special conditions would prevent a fuel tank with 
an FRM from being flammable during specific warm day operating 
conditions, such as those present when recent accidents occurred.

Definition of ``Inert''

    For the purpose of these special conditions, the tank is considered 
inert when the bulk average oxygen concentration within each 
compartment of the tank is 12 percent or less at sea level up to 10,000 
feet, then linearly increasing from 12 percent at 10,000 feet to 14.5 
percent at 40,000 feet and extrapolated linearly above that altitude. 
The reference to each section of the tank is necessary because fuel 
tanks that are compartmentalized may encounter localized oxygen 
concentrations in one or more compartments that exceed the 12 percent 
value. Currently there is not adequate data available to establish 
whether exceeding the 12 percent limit in one compartment of a fuel 
tank could create a hazard. For example, ignition of vapors in one 
compartment could result in a flame front within the compartment that 
travels to adjacent compartments and results in an ignition source that 
exceeds the ignition energy (the minimum amount of energy required to 
ignite fuel vapors) values used to establish the 12 percent limit. 
Therefore, ignition in other compartments of the tank may be possible. 
Technical discussions with the applicant indicate the pressure rise in 
a fuel tank that was at or near the 12 percent oxygen concentration 
level would likely be well below the value that would rupture a typical 
transport airplane fuel tank. While this may be possible to show, it is 
not within the scope of these special conditions. Therefore, the effect 
of the definition of ``inert'' within these special conditions is that 
the bulk average of each individual compartment or bay of the tank must 
be evaluated and shown to meet the oxygen concentration limits 
specified in the definitions section of these special conditions (12 
percent or less at sea level) to be considered inert.

[[Page 7804]]

Determining Flammability

    The methodology for determining fuel tank flammability defined for 
use in these special conditions is based on that used by ARAC to 
compare the flammability of unheated aluminum wing fuel tanks to that 
of tanks that are heated by adjacent equipment. The ARAC evaluated the 
relative flammability of airplane fuel tanks using a statistical 
analysis commonly referred to as a ``Monte Carlo'' analysis that 
considered a number of factors affecting formation of flammable vapors 
in the fuel tanks. The Monte Carlo analysis calculates values for the 
parameter of interest by randomly selecting values for each of the 
uncertain variables from distribution tables. This calculation is 
conducted over and over to simulate a process where the variables are 
randomly selected from defined distributions for each of the variables. 
The results of changing these variables for a large number of flights 
can then be used to approximate the results of the real world exposure 
of a large fleet of airplanes.
    Factors that are considered in the Monte Carlo analysis required by 
these special conditions include those affecting all airplane models in 
the transport airplane fleet such as: A statistical distribution of 
ground, overnight, and cruise air temperatures likely to be experienced 
worldwide, a statistical distribution of likely fuel types, and 
properties of those fuels, and a definition of the conditions when the 
tank in question will be considered flammable. The analysis also 
includes factors affecting specific airplane models such as climb and 
descent profiles, fuel management, heat transfer characteristics of the 
fuel tanks, statistical distribution of flight lengths (mission 
durations) expected for the airplane model worldwide, etc. To quantify 
the fleet exposure, the Monte Carlo analysis approach is applied to a 
statistically significant number (1,000,000) of flights where each of 
the factors described above is randomly selected. The flights are then 
selected to be representative of the fleet using the defined 
distributions of the factors described previously. For example, flight 
one may be a short mission on a cold day with an average flash point 
fuel, and flight two may be a long mission on an average day with a low 
flash point fuel, and on and on until 1,000,000 flights have been 
defined in this manner. For every one of the 1,000,000 flights, the 
time that the fuel temperature is above the flash point of the fuel, 
and the tank is not inert, is calculated and used to establish if the 
fuel tank is flammable. Averaging the results for all 1,000,000 flights 
provides an average percentage of the flight time that any particular 
flight is considered to be flammable. While these special conditions do 
not require that the analysis be conducted for 1,000,000 flights, the 
accuracy of the Monte Carlo analysis improves as the number of flights 
increases. Therefore, to account for this improved accuracy appendix 2 
of these special conditions defines lower flammability limits if the 
applicant chooses to use fewer than 1,000,000 flights.
    The determination of whether the fuel tank is flammable is based on 
the temperature of the fuel in the tank determined from the tank 
thermal model, the atmospheric pressure in the fuel tank, and 
properties of the fuel quantity loaded for a given flight, which is 
randomly selected from a database consisting of worldwide data. The 
criteria in the model are based on the assumption that as these 
variables change, the concentration of vapors in the tank 
instantaneously stabilizes and that the fuel tank is at a uniform 
temperature. This model does not include consideration of the time lag 
for the vapor concentration to reach equilibrium, the condensation of 
fuel vapors from differences in temperature that occur in the fuel 
tanks, or the effect of mass loading (times when the fuel tank is at 
the unusable fuel level and there is insufficient fuel at a given 
temperature to form flammable vapors). However, fresh air drawn into an 
otherwise inert tank during descent does not immediately saturate with 
fuel vapors so localized concentrations above the inert level during 
descent do not represent a hazardous condition. These special 
conditions allow the time during descent, where a localized amount of 
fresh air may enter a fuel tank, to be excluded from the determination 
of fuel tank flammability exposure.

Definition of Transport Effects

    The effects of low fuel conditions (mass loading) and the effects 
of fuel vaporization and condensation with time and temperature 
changes, referred to as ``transport effects'' in these special 
conditions, are excluded from consideration in the Monte Carlo model 
used for demonstrating compliance with these special conditions. These 
effects have been excluded because they were not considered in the 
original ARAC analysis, which was based on a relative measure of 
flammability. For example, the 3 percent flammability value established 
by the ARAC as the benchmark for fuel tank safety for wing fuel tanks 
did not include the effects of cooling of the wing tank surfaces and 
the associated condensation of vapors from the tank ullage. If this 
effect had been included in the wing tank flammability calculation, it 
would have resulted in a significantly lower wing tank flammability 
benchmark value. The ARAC analysis also did not consider the effects of 
mass loading which would significantly lower the calculated 
flammability value for fuel tanks that are routinely emptied (e.g., 
center wing tanks). The FAA and JAA have determined that using the ARAC 
methodology provides a suitable basis for determining the adequacy of 
an FRM system.
    The effect of condensation and vaporization in reducing the 
flammability exposure of wing tanks is comparable to the effect of the 
low fuel condition in reducing the flammability exposure of center 
tanks. We therefore consider these effects to be offsetting, so that by 
eliminating their consideration, the analysis will produce results for 
both types of tanks that are comparable. Using this approach, it is 
possible to follow the ARAC recommendation of using the unheated 
aluminum wing tank as the standard for evaluating the flammability 
exposure of all other tanks. For this reason, both factors have been 
excluded when establishing the flammability exposure limits. During 
development of these harmonized special conditions, the FAA and the 
European Joint Aviation Authorities (JAA) agreed that using the ARAC 
methodology provides a suitable basis for determining the flammability 
of a fuel tank and consideration of transport effects should not be 
permitted.

Flammability Limit

    The FAA, in conjunction with the Joint Airworthiness Authorities 
(JAA) and Transport Canada, has developed criteria within these special 
conditions that require overall fuel tank flammability to be limited to 
3 percent of the fleet average operating time. This overall average 
flammability limit consists of times when the system performance cannot 
maintain an inert tank ullage, primarily during descent when the change 
in ambient pressures draws air into the fuel tanks and those times when 
the FRM is inoperative due to failures of the system and the airplane 
is dispatched with the system inoperative.

Specific Risk Flammability Limit

    These special conditions also include a requirement to limit fuel 
tank flammability to 3 percent during ground operations, takeoff, and 
climb phases of

[[Page 7805]]

flight to address the specific risk associated with operation during 
warmer day conditions when accidents have occurred. The specific risk 
requirement is intended to establish minimum system performance levels 
and therefore the 3 percent flammability limit excludes reliability 
related contributions, which are addressed in the average flammability 
assessment. The specific risk requirement may be met by conducting a 
separate Monte Carlo analysis for each of the specific phases of flight 
during warmer day conditions defined in the special conditions, without 
including the times when the FRM is not available because of failures 
of the system or dispatch with the FRM inoperative.

Inerting System Indications

    Fleet average flammability exposure involves several elements, 
including--
     The time the FRM is working properly and inerts the tank 
or when the tank is not flammable;
     The time when the FRM is working properly but fails to 
inert the tank or part of the tank, because of mission variation or 
other effects;
     The time the FRM is not functioning properly and the 
operator is unaware of the failure; and
     The time the FRM is not functioning properly and the 
operator is aware of the failure and is operating the airplane for a 
limited time under MEL relief.
    The applicant may propose that MMEL relief is provided for aircraft 
operation with the FRM unavailable; however, it is considered a safety 
system that should be operational to the maximum extent practical. 
Therefore, these special conditions include reliability and reporting 
requirements to enhance system reliability so that dispatch of 
airplanes with the FRM inoperative would be very infrequent. Cockpit 
indication of the system function that is accessible to the flightcrew 
is not an explicit requirement, but may be required if the results of 
the Monte Carlo analysis show the system cannot otherwise meet the 
flammability and reliability requirements defined in these special 
conditions. Flight test demonstration and analysis will be required to 
demonstrate that the performance of the inerting system is effective in 
inerting the tank during those portions of ground and the flight 
operations where inerting is needed to meet the flammability 
requirements of these special conditions.
    Various means may be used to ensure system reliability and 
performance. These may include: System integrity monitoring and 
indication, redundancy of components, and maintenance actions. A 
combination of maintenance indication and/or maintenance check 
procedures will be required to limit exposure to latent failures within 
the system, or high inherent reliability is needed to assure the system 
will meet the fuel tank flammability requirements. The applicant's 
inerting system does not incorporate redundant features and includes a 
number of components essential for proper system operation. Past 
experience has shown inherent reliability of this type of system would 
be difficult to achieve. Therefore, if system maintenance indication is 
not provided for features of the system essential for proper system 
operation, system functional checks at appropriate intervals determined 
by the reliability analysis will be required for these features. At a 
minimum, proper function of essential features of the system should be 
validated once per day by maintenance review of indications or 
functional checks, possibly prior to the first flight of the day. The 
determination of a proper interval and procedure will follow completion 
of the certification testing and demonstration of the system's 
reliability and performance prior to certification.
    Any features or maintenance actions needed to achieve the minimum 
reliability of the FRM will result in fuel system airworthiness 
limitations similar to those defined in Sec.  25.981(b). Boeing will be 
required to include in the instructions for continued airworthiness 
(ICA) the replacement times, inspection intervals, inspection 
procedures, and the fuel system limitations required by Sec.  
25.981(b). Overall system performance and reliability must achieve a 
fleet average flammability that meets the requirements of these special 
conditions. If the system reliability falls to a point where the fleet 
average flammability exposure exceeds these requirements, Boeing will 
be required to define appropriate corrective actions, to be approved by 
the FAA, that will bring the exposure back down to the acceptable 
level.
    Boeing proposed that the FRM be eligible for a 10-day MMEL dispatch 
interval. The Flight Operations Evaluation Board (FOEB) will establish 
the approved interval based on data the applicant submits to the FAA. 
The MMEL dispatch interval is one of the factors affecting system 
reliability analyses that must be considered early in the design of the 
FRM, prior to FAA approval of the MMEL. Boeing requested that the 
authorities agree to use of an MMEL inoperative dispatch interval for 
design of the system. Boeing data indicates that certain systems on the 
airplane are routinely repaired prior to the maximum allowable 
interval. These special conditions require that Boeing use an MMEL 
inoperative dispatch interval of 60 hours in the analysis as 
representative of the mean time for which an inoperative condition may 
occur for the 10-day MMEL maximum interval requested. Boeing must also 
include actual dispatch inoperative interval data in the quarterly 
reports required by Special Condition III(c)(2). Boeing may request to 
use an alternative interval in the reliability analysis. Use of a value 
less than 60 hours would be a factor considered by the FOEB in 
establishing the maximum MMEL dispatch limit. The reporting requirement 
will provide data necessary to validate that the reliability of the FRM 
achieved in service meets the levels used in the analysis.
    Appropriate maintenance and operational limitations with the FRM 
inoperative may also be required and noted in the MMEL. The MMEL 
limitations and any operational procedures should be established based 
on results of the Monte Carlo assessment, including the results 
associated with operations in warmer climates where the fuel tanks are 
flammable a significant portion of the FEET when not inert. While the 
system reliability analysis may show that it is possible to achieve an 
overall average fleet exposure equal to or less than that of a typical 
unheated aluminum wing tank, even with an MMEL allowing very long 
inoperative intervals, the intent of the rule is to minimize 
flammability. Therefore, the shortest practical MMEL relief interval 
should be proposed. To ensure limited airplane operation with the 
system inoperative and to meet the reliability requirements of these 
special conditions, appropriate level messages that are needed to 
comply with any dispatch limitations of the MMEL must be provided.

Confined Space Hazard Markings

    Introduction of the FRM will result in NEA within the center wing 
fuel tank and the possibility of NEA in compartments adjacent to the 
fuel tank if leakage from the tank or NEA supply lines were to occur. 
Lack of oxygen in these areas could be hazardous to maintenance 
personnel, the passengers, or flightcrew. Existing certification 
requirements do not address all aspects of these hazards. Paragraph 
II(f) of the special conditions requires the applicant to provide 
markings to emphasize the potential hazards associated with confined 
spaces and areas where a hazardous atmosphere

[[Page 7806]]

could be present due to the addition of an FRM.
    For the purposes of these special conditions, a confined space is 
an enclosed or partially enclosed area that is big enough for a worker 
to enter and perform assigned work and has limited or restricted means 
for entry or exit. It is not designed for someone to work in regularly, 
but workers may need to enter the confined space for tasks such as 
inspection, cleaning, maintenance, and repair. (Reference U.S. 
Department of Labor Occupational Safety & Health Administration (OSHA), 
29 CFR 1910.146(b).) The requirement in the special conditions does not 
significantly change the procedures maintenance personnel use to enter 
fuel tanks and are not intended to conflict with existing government 
agency requirements (e.g., OSHA). Fuel tanks are classified as confined 
spaces and contain high concentrations of fuel vapors that must be 
exhausted from the fuel tank before entry. Other precautions such as 
measurement of the oxygen concentrations before entering a fuel tank 
are already required. Addition of the FRM that utilizes inerting may 
result in reduced oxygen concentrations due to leakage of the system in 
locations in the airplane where service personnel would not expect it. 
A worker is considered to have entered a confined space just by putting 
his or her head across the plane of the opening. If the confined space 
contains high concentrations of inert gases, workers who are simply 
working near the opening may be at risk. Any hazards associated with 
working in adjacent spaces near the opening should be identified in the 
marking of the opening to the confined space. A large percentage of the 
work involved in properly inspecting and modifying airplane fuel tanks 
and their associated systems must be done in the interior of the tanks. 
Performing the necessary tasks requires inspection and maintenance 
personnel to physically enter the tank, where many environmental 
hazards exist. These potential hazards that exist in any fuel tank, 
regardless of whether nitrogen inerting has been installed, include 
fire and explosion, toxic and irritating chemicals, oxygen deficiency, 
and the confined nature of the fuel tank itself. In order to prevent 
related injuries, operator and repair station maintenance organizations 
have developed specific procedures for identifying, controlling, or 
eliminating the hazards associated with fuel-tank entry. In addition 
government agencies have adopted safety requirements for use when 
entering fuel tanks and other confined spaces. These same procedures 
would be applied to the reduced oxygen environment likely to be present 
in an inerted fuel tank.
    The designs currently under consideration locate the FRM in the 
fairing below the center wing fuel tank. Access to these areas is 
obtained by opening doors or removing panels which could allow some 
ventilation of the spaces adjacent to the FRM. But this may not be 
enough to avoid creating a hazard. Therefore, we intend that marking be 
provided to warn service personnel of possible hazards associated with 
the reduced oxygen concentrations in the areas adjacent to the FRM.
    Appropriate markings would be required for all inerted fuel tanks, 
tanks adjacent to inerted fuel tanks and all fuel tanks communicating 
with the inerted tanks via plumbing. The plumbing includes, but is not 
limited to, plumbing for the vent system, fuel feed system, refuel 
system, transfer system and cross-feed system. NEA could enter adjacent 
fuel tanks via structural leaks. It could also enter other fuel tanks 
through plumbing if valves are operated or fail in the open position. 
The markings should also be stenciled on the external upper and lower 
surfaces of the inerted tank adjacent to any openings to ensure 
maintenance personnel understand the possible contents of the fuel 
tank. Advisory Circular 25.981-2 will provide additional guidance 
regarding markings and placards.

Affect of FRM on Auxiliary Fuel Tank System Supplemental Type 
Certificates

    Boeing plans to offer a service bulletin that will install the FRM 
on existing in-service airplanes. Some in-service airplanes have 
auxiliary fuel tank systems installed that interface with the center 
wing tank. The Boeing FRM design is intended to provide inerting of the 
fuel tank volume of the 747 and does not include consideration of the 
auxiliary tank installations. Installation of the FRM on existing 
airplanes with auxiliary fuel tank systems may therefore require 
additional modifications to the auxiliary fuel tank system to prevent 
development of a condition that may cause the tank to exceed the 12 
percent oxygen limit. The FAA will address these issues during 
development and approval of the service bulletin for the FRM.

Disposal of Oxygen-Enriched Air (OEA)

    The FRM produces both NEA and OEA. The OEA generated by the FRM 
could result in an increased fire hazard if not disposed of properly. 
The OEA produced in the proposed design is diluted with air from a heat 
exchanger, which is intended to reduce the OEA concentration to non-
hazardous levels. Special requirements are included in these special 
conditions to address potential leakage of OEA due to failures and safe 
disposal of the OEA during normal operation.
    To ensure that an acceptable level of safety is achieved for the 
modified airplanes using a system that inerts heated fuel tanks with 
NEA, special conditions (per Sec.  21.16) are needed to address the 
unusual design features of an FRM. These special conditions contain the 
additional safety standards that the Administrator considers necessary 
to establish a level of safety equivalent to that established by the 
existing airworthiness standards.

Discussion of Comments

    Notice of Proposed Special Conditions No. 25-03-08-SC for the 
Boeing Model 747-100/200B/200F/200C/SR/SP/100B/300/100B SUD/400/400D/
400F series airplanes was published in the Federal Register on December 
9, 2003 (68 FR 68563). Thirteen commenters responded to the notice.

General Comments

    Comment: One commenter supports the special conditions but states 
that ignition source prevention must still be provided. The commenter 
believes that the combination of flammability reduction and ignition 
source prevention is the most effective means to prevent fuel tank 
explosions.
    FAA Reply: The safety assessment required by Special Federal 
Aviation Regulation (SFAR) No. 88, Fuel Tank System Fault Tolerance 
Evaluation, identifies design and maintenance changes that are needed 
to prevent ignition sources in transport category airplanes. The FAA is 
developing a number of airworthiness directives (ADs) to address 
ignition sources resulting from single failures in all fuel tanks and 
combinations of failures in tanks that have been classified as high 
flammability. We will not issue ADs to address combinations of failures 
in high flammability tanks if the FRM is installed because of the 
significant improvement in fuel tank safety offered by the FRM required 
by this special condition. We are not considering a change to the 
current ignition prevention analysis requirements that include assuming 
a flammable ullage. No changes were made as a result of this comment.
    Comment: Two commenters believe the special conditions for the FRM 
are

[[Page 7807]]

not appropriate because the special conditions are written to fit the 
applicant's proposed design of an inerting system to reduce 
flammability of fuel tanks and are therefore considered ``prejudiced.'' 
One of these commenters adds that regulatory guidance should be 
unprejudiced and available before development of any design.
    FAA Reply: We do not concur. As stated earlier in this document, 
these special conditions are specific to certification of an FRM based 
on inerting technology. As discussed in AC 25.981-2, inerting, as well 
as other technologies such as cooling, is an acceptable means of 
compliance with Sec.  25.981(c). No changes were made as a result of 
this comment.
    Comment: Two commenters believe the limited FRM, as described in 
the special conditions, would not comply with the requirements of 
Sec. Sec.  25.981(c) and 25.1309 for new airplane designs (post 
amendment 25-102) with high flammability fuel tanks.
    FAA Reply: As stated earlier, these special conditions apply 
specifically to certification of an FRM for applicable Boeing Model 747 
series airplanes and do not apply to new airplane designs. However, we 
have determined that an FRM that complies with these special conditions 
would meet the intent of Sec.  25.981(c). No changes were made as a 
result of this comment.
    Comment: One commenter would support rulemaking to investigate 
amending Sec.  25.981 (and revising AC 25.981-2) to:
     Clarify that ``minimization of flammable vapors'' in 
accordance with Sec.  25.981(c) is to be accomplished through design 
features ensuring the tank will have inherent low flammability (e.g. 
venting, cooling, control of heat transfer, etc.); and
     Eliminate the possibility of compliance for future 
airplane designs through the installation of a limited FRM.
    FAA Reply: On February 17, 2004, the FAA Administrator announced 
plans to issue a notice of proposed rulemaking that will require 
approximately 3,800 Airbus and Boeing planes be fitted with systems 
that reduce the presence of flammable vapors in fuel tanks. This 
proposal could require airlines to install new systems to reduce fuel 
tank flammability on existing and newly produced larger passenger jets. 
We are also considering amending Sec.  25.981(c) and revising AC 
25.981-2 to further limit fuel tank flammability. No changes were made 
as a result of these comments.
    Comment: The commenter requests that before proceeding with any 
further regulatory activities, the FAA should provide additional 
detailed information on whether SFAR 88 changes are sufficient to cover 
the requirements of Sec.  25.981. The commenter believes that ``SFAR 88 
meets the requirement of Sec.  25.981(c)(2) and does not understand the 
need to also address Sec.  25.981(c)(1).'' This commenter also states 
that harmonization with the European Aviation Safety Agency (EASA) on 
these special conditions is essential for industry.
    FAA Reply: We do not concur with the commenter's first statement. A 
direct relationship between SFAR 88 and Sec.  25.981(c)(1), or Sec.  
25.981(c)(2), does not exist. SFAR 88 addresses ignition source 
prevention, while Sec.  25.981(c)(1) acknowledges an ignition source 
may be present under some remote circumstances. Section 25.981(c)(2) 
assumes that an ignition can occur--in essence that SFAR 88 was not 
successful and also flammable vapors are present--and requires that the 
resulting ignition of flammable vapor will not prevent continued safe 
flight and landing. The FAA has fully coordinated these special 
conditions with the JAA/EASA. No changes were made as a result of these 
comments.
    Comment: One commenter notes that although the special condition 
requirements for system reliability and performance are very specific, 
they do not address the qualification standards that the system will 
have to meet. Additional guidance on this subject would be appropriate. 
Another commenter expresses concern about use of the terms ``intended'' 
and ``expected'' in the special conditions when relating to an FRM. It 
is the commenter's opinion that the use of these terms indicates that 
the applicant is not confident that their design ``will'' or ``shall'' 
contribute to the overall safety of the airplanes.
    FAA Reply: We do not concur. In the preamble to the special 
conditions, we state that the applicant is required to show compliance 
with the applicable airworthiness regulations and special conditions. 
In part, the applicable regulations, Sec.  25.1301 and Sec.  25.1309, 
require the applicant to show that the equipment ``functions properly 
when installed'' and ``is designed to ensure that they perform their 
intended functions under any foreseeable operating condition.'' 
Irrespective of any wording in the preamble to the special conditions, 
the special conditions include requirements to address foreseeable 
specific safety issues that are not addressed by the current 
regulations. Any airplane that meets the requirements of the special 
conditions will maintain the level of safety intended by the applicable 
requirements of the Code of Federal Regulations (CFR). No changes were 
made as a result of these comments.
    Comment: One commenter states that there are various statements 
made throughout the special conditions that refer to reliability and 
maintenance of the system. It is the commenter's opinion that these 
statements are specific to implementation, and the actual approach 
should be derived using standard methodology used for certification of 
the airplane.
    FAA Reply: To achieve the desired safety level of the FRM, we 
believe the special condition requirements for determining reliability 
and maintainability of the FRM are necessary. This is to ensure that 
the FRM is an acceptable means by which the development of flammable 
vapors in the center wing tank is minimized as required by Sec.  
25.981. No changes were made as a result of this comment.
    Comment: One commenter notes that ``inert'' is not defined 
consistently throughout the special conditions. The commenter suggests 
the use of only one definition and proposes the definition used in 
special condition paragraph I. Definitions. The same commenter also 
requests clarification if linear extrapolation of oxygen concentration 
can be used for aircraft ceilings above 40,000 feet, and clarification 
of the difference between the terms ``bulk'' and ``bulk average.''
    FAA Reply: We concur that the definition of inert needs to be 
consistent throughout the special conditions and have therefore 
modified the definition of inert in the preamble to incorporate the 
definition of inert provided in paragraph I. Definitions of the special 
conditions. With respect to aircraft altitudes above 40,000 feet, we 
have added that linear extrapolation can continue for oxygen 
concentration from 14.5 percent at 40,000 feet to the required 
operating altitude. Concerning the use of bulk and bulk average in the 
special conditions, we have modified the preamble and special 
conditions to consistently use the term ``bulk average'' when referring 
to the fuel temperature or oxygen concentration within the fuel tank.
    Comment: The commenter requests that the FAA clarify if the FRM is 
a safety enhancement system or a safety system. The commenter notes 
that in the preamble discussion of the ``Inerting System Indication,'' 
the FAA states that the applicant may propose master minimum equipment 
list (MMEL) relief

[[Page 7808]]

be provided for airplane operation with the FRM unavailable. The 
system, however, is considered a safety system that should be 
operational to the maximum extent practical. If this system is 
considered a safety system, then a form of redundancy will have to be 
built in. At this time, the applicant's design does not show any 
redundancy.
    FAA Reply: The FRM is a safety system designed to provide an 
additional layer of protection to the ignition prevention means already 
in place. The system by itself is not intended to be fully redundant 
since it provides a second layer of protection. The FRM is intended to 
be a safety enhancement system that provides an additional layer of 
protection by reducing the exposure to flammable vapors in the heated 
center wing fuel tank. This protection, when added to ignition 
prevention measures, will substantially reduce the likelihood of future 
fuel tank explosions in the fleet. The applicant has proposed a 10-day 
MMEL relief period, but the Flight Operations Evaluation Board (FOEB) 
will determine and approve the appropriate MMEL intervals based on data 
the applicant submits to the FAA. The applicant must show that the 
fleet average flammability exposure of a tank with an FRM installed is 
equal to or less than 3 percent, including any time when the system is 
inoperative. No changes were made as a result of these comments.
    Comment: One commenter says the cost of the FRM is substantial and 
justification for it is debatable. The commenter believes the FRM will 
put a heavy economic burden on the slowly recovering airline industry 
and only supports the adoption of an FRM on new type designs and newly 
built airplanes as an improvement in fuel system safety. This commenter 
also says that considering the potential affects of this subject on the 
European airline industry, joint European position activity is critical 
to ensure that decisions are based on safety grounds and not on 
political motivations.
    FAA Reply: We do not concur with the commenter regarding the impact 
of cost associated with the issuance of the special conditions. These 
special conditions are unique to the applicant's certification of an 
FRM for the applicable Boeing Model 747 series airplanes and do not 
mandate that an FRM must be added to an operator's 747 fleet. They have 
been fully harmonized with EASA. The FAA announcement of issuance of a 
notice of proposed rulemaking that would propose retrofit and 
production incorporation of FRM into U.S-registered airplanes is a 
separate rulemaking effort that will require a cost benefit analysis 
and will be published for public comment. No changes were made as a 
result of this comment.
    Comment: One commenter notes that the applicant has planned a 3-
month, in-service evaluation (ISE) of the FRM. It is the opinion of two 
other commenters that a 4,000-hour (12 month) ISE should be specified 
before certification of the FRM because--
     It adds complexity,
     It has not yet been retrofitted in an in-service airplane,
     It has no proven track record for reliability, and
     Ground and flight tests are not sufficient to demonstrate 
overall reliability of the system.

    The commenters say that maintenance and performance features of the 
system were designed to support a 10-day relief under the MMEL program. 
If the demonstrated performance and reliability of the system meet 
design objectives, then the FAA should support the planned relief. 
Another commenter recommends a one-year in-service evaluation (ISE) 
program following the first installation of an FRM and prior to FRM 
installation on a production airplane. This commenter says that past 
experience has shown reliability and system degradation by oil 
contamination scenarios, with the engine and APU being the source, and 
carbon particle buildup on components similar to those required by the 
proposed FRM, due to airport and airplane turbine exhausts. This 
commenter believes that one year would be an adequate time for the 
manufacturer to develop and provide corrective actions for 
discrepancies or reliability issues with the FRM that are identified 
during the ISE program.
    FAA Reply: We do not concur with the commenters. The industry 
commonly conducts ISE through cooperative efforts between the type 
certificate holder and the airlines prior to fleetwide introduction of 
changes. While the FAA agrees an ISE might be appropriate, we 
traditionally do not mandate it. An ISE can be part of a manufacturer's 
incorporation strategy for optional equipment. FAA certification of a 
system is required before an ISE can be conducted on a U.S.-registered 
transport category airplane; therefore, an ISE is not related to 
certification requirements. The reliability reporting requirements in 
the special conditions will provide data to determine if actions are 
needed to correct discrepancies and improve system reliability after 
certification of the system. No changes were made as a result of these 
comments.
    Comment: Three commenters request that the FAA consider 9 percent 
as the maximum oxygen concentration at sea level. One commenter 
disagrees with the premise that the wing fuel tanks offer an acceptable 
minimum level of flammability exposure and is concerned about using 
this minimum level for development of inerting systems. The commenters 
believe that the maximum oxygen concentration of 12 percent at sea 
level should be considered as a level of reduced flammability rather 
than inert, and that 9 percent should be used as the long-term goal for 
defining a tank as inert. Another commenter states that 12 percent 
oxygen concentration will not protect the center or wing fuel tanks 
from external hazards and that 9 percent should be used to protect the 
tanks. The commenter requests clarification of why 12 percent oxygen 
concentration at sea level is specified in the special conditions 
instead of the maximum 9 percent.
    Three commenters want the minimum oxygen concentration percentage 
at sea level to be 10 percent. They refer to paragraph 7(a)(1) of AC 
25.981-2, which reads: ``An oxygen concentration of 10 percent or less 
by volume is acceptable for transport airplane fuel tanks inerted with 
nitrogen, without additional substantiation.'' One commenter believes 
this acceptable oxygen concentration establishes a minimum acceptable 
performance standard in terms of the threat (ignition source energy), 
and 10 percent or less should be the average design concentration for 
each fuel cell with no area at a concentration greater than 11.5 
percent. Another commenter says that 10 percent contradicts the 
definition of ``inert,'' as proposed, and would like the FAA to provide 
the acceptable oxygen concentration level (percentage by volume) and 
the fundamental justification for this level. Minimum performance 
inherent in the AC method must be guaranteed. The final commenter would 
like to know if AC 25.981-2 will be revised if the FAA believes that 12 
percent is adequate.
    Two commenters referenced applying an adequate safety factor to the 
maximum 12 percent oxygen concentration limit. One commenter referenced 
various reports they believe support the use of a 20 percent safety 
margin that should be applied to the FRM. The commenter states that the 
FAA uses safety factors in design of aircraft structure, components, 
and systems and to deviate from good design practice is not in the 
interest of public

[[Page 7809]]

safety. This commenter suggests that the FAA follow industry practice.
    FAA Reply: We do not concur with the commenters. The special 
condition requirement of 12 percent maximum oxygen concentration at sea 
level is based on FAA oxygen content testing and review of other test 
data, such as Navy gunfire tests. These data show that 12 percent 
oxygen concentration will prevent a fuel tank explosion for airplane 
system failure and malfunction-generated ignition sources. 
Additionally, data from the Navy testing provided in document NWC TP 
7129, ``The Effectiveness of Ullage Nitrogen-Inerting Systems Against 
30 mm High-Explosive Incendiary Projectiles,'' dated May 1991, shows 
that 12 percent oxygen concentrations are also very effective at 
mitigating the effects of a high-energy incendiary projectile 
puncturing the fuel tank ullage.
    We plan to revise AC 25.981-2 to include the definition of inert 
that is used in these special conditions.

Summary

    Comment: The commenter refers to the statement in the summary 
paragraph that the regulations do not contain adequate or appropriate 
safety standards. The commenter considers this statement invalid and 
fails to comprehend what is missing in the regulations to adequately 
address certification of an FRM and why special conditions would be 
required. The commenter agrees with the FAA that the FRM installation 
must comply with Sec.  25.981 at amendment 25-102, the fuel vent and 
exhaust emission requirements of part 34, and the acoustical 
requirements of Sec.  21.93(b). The commenter also believes that 
Sec. Sec.  25.831(b), 25.1301, 25.1307, 25.1309, 25.1316, 25.1321, 
25.1322, 25.1357, 25.1431, 25.1438, and 25.1461 might also apply.
    FAA Reply: Many of the regulations quoted by the commenter are 
applicable, and compliance with these requirements must be shown for 
certification of the FRM for the applicable Boeing Model 747 series 
airplanes. However, part 25 regulations do not contain adequate or 
appropriate safety standards for the performance of the FRM. The basis 
to issue special conditions is addressed in Sec.  21.16. No changes 
were made as a result of this comment.

Background

    Comment: This commenter believes ignition source prevention has 
failed. The commenter points to the 1997 notice, in which the FAA 
requested industry comments on the mitigation of hazards posed by 
flammable fuel tank vapors. In that notice, the FAA cites 13 fuel tank 
explosion/ignition events and three non-operational events, for a total 
of 16 during the 1959-1996 timeframe, before the Thailand B737 center 
wing tank explosion. The commenter says that since the ignition sources 
for the last three accidents are unknown, an FRM must safeguard against 
unknown ignition sources of unknown ignition energy. A significant 
number of single failures and combinations of failures can result in 
ignition sources within fuel tanks; therefore an acceptable system must 
safeguard against all (except extremely improbable) ignition sources 
within the fuel tank. The commenter also notes that approximately 550 
people lost their lives in these explosions.
    FAA Reply: The ignition prevention safety reviews conducted 
following the 1996 accident revealed many previously unknown single 
component failures that could result in ignition sources within the 
fuel tanks. We will issue additional ADs, where necessary, to require 
design or maintenance actions to address these newly discovered 
deficiencies. The safety reviews also identified combinations of 
failures that could result in an ignition source. Because service 
experience and analysis indicated that these combinations were less 
likely to occur, we determined that it was not practical to address 
them in existing airplanes. The safety reviews also confirmed that 
unforeseen design and maintenance errors exist and result in 
development of ignition sources. As discussed earlier in this document, 
the NTSB recommendations included not just preventing ignition sources, 
but also reducing fuel tank flammability. The NTSB concluded that ``a 
fuel tank design and certification philosophy that relies solely on the 
elimination of all ignition sources, while accepting the existence of 
fuel tank flammability, is fundamentally flawed because experience has 
demonstrated that all possible ignition sources cannot be determined 
and reliably eliminated.'' Therefore, the purpose of these special 
conditions is not to address additional rulemaking for prevention of 
ignition sources but to certificate a specific fuel tank FRM for Boeing 
Model 747 series airplanes. No changes were made as a result of this 
comment.
    Comment: The commenter states that service experience of airplanes 
certificated to the earlier standards shows that ignition source 
prevention alone has not been totally effective at preventing 
accidents. The commenter notes that after the TWA 800 accident, fuel 
tank system rulemaking activity started in such an excessive way that 
the FAA has mandated over 50 ADs and proposed changes to part 25. After 
other fuel tank explosion accidents prior to the flight TWA 800 
accident, the FAA did not change the design standards of fuel tank 
systems. SFAR 88 was the first real rulemaking activity where the FAA 
mandated ignition source reduction throughout the fleet. Those changes 
are not incorporated at this time. The commenter therefore believes the 
FAA cannot say that the past service experience for ignition source 
prevention alone has not been totally effective in preventing 
accidents. Currently, the results of ignition source prevention 
measures are unknown.
    This same commenter also believes that the addition of SFAR 88 and 
an FRM will not reduce the chance of maintenance induced errors and may 
have an opposite effect in that it could introduce the risk of further 
human factors errors.
    FAA Reply: We do not concur. Past experience shows that detailed 
design reviews, similar to those required by SFAR 88, have not been 
effective at eliminating ignition sources. Following an accident in 
1976, we conducted an exhaustive investigation and design review of the 
lightning protection features of the fuel tank system, including full 
scale testing of the wing. From this, we mandated design changes to 
improve lightning protection of the system. Subsequent review of the 
airplane design required by SFAR 88 revealed the need for additional 
bonding modifications that will be mandated. Failure of other 
components within the fuel tank system and components adjacent to the 
fuel tank could also cause ignition sources. These examples show that 
it is very difficult to identify all ignition sources during design. 
Additionally, past experience also indicates unforeseen design and 
maintenance errors can result in development of ignition sources.
    We have issued multiple ADs to address ignition source prevention 
and believe that implementation of design changes intended to prevent 
ignition sources identified by SFAR 88 will prevent about 50 percent of 
future fuel tank explosions. The more significant changes to fuel tank 
systems resulting from the SFAR 88 activity include:
     Features to prevent dry running of fuel pumps within the 
fuel tanks;
     Ground fault protection of fuel pump power supplies for 
pumps or wires exposed to the fuel tank ullage;
     Additional electrical bonds on some components;
     Electrical energy limiters on wiring entering fuel tanks 
that are normally

[[Page 7810]]

emptied and located within the fuselage contour;
     Electrical bond integrity checks; and
     Improved maintenance programs.
    While we believe these modifications and maintenance program 
changes will significantly improve safety, the results of the safety 
reviews conducted as part of SFAR 88 show there is uncertainty in the 
effectiveness of ignition source prevention alone. The addition of an 
FRM will significantly improve fuel tank safety by reducing or 
preventing flammable vapors in the fuel tank and will incorporate fail-
safe features into the fuel tank system that account for design and 
maintenance errors. No changes were made as a result of these comments.

Fuel Properties

    Comment: The commenter says that the new generation airplanes 
(B737NG, B757, B767, and B777) are not certified to use Jet B or JP-4 
wide-cut fuels. The commenter also points out that AD 85-11-52R1 
prohibits the use of Jet B and JP-4 on Boeing Model 737-300 series 
airplanes.
    FAA Reply: We do not concur. While wide-cut fuels are not commonly 
used in the world fleets, some of the airplanes mentioned do allow at 
least limited use. Other models are certified for unrestricted use. 
Significant use of lower flash-point fuels could affect the percentage 
of time the fuel tanks are flammable. Therefore, to achieve consistent 
flammability exposure, the flash point of the approved fuels must be 
considered in the analysis used for demonstrating compliance. No 
changes were made as a result of these comments.

Fire Triangle

    Comment: The commenter points to the FAA statement, ``Because of 
technological limitations in the past, the FAA philosophy regarding the 
prevention of fuel tank explosions to ensure airplane safety was to 
only preclude ignition sources within fuel tanks.'' It is the 
commenter's opinion that there never was a technological limitation. 
The commenter refers to a test the FAA conducted in the 1970s of a 
nitrogen fuel tank inerting system on a DC-9 airplane, and that system 
maintained oxygen concentration less than 8 percent under all normal 
and emergency flight conditions. The commenter also listed other 
airplanes that use NEA, liquid nitrogen, and explosion suppressant 
systems to minimize fuel tank flammability. The commenter further 
points out that in March 2002, the Aviation Rulemaking Advisory 
Committee (ARAC) concluded that fuel tank inerting may provide safety 
benefits and warrants continued industry and government research. Then, 
in December 2002, an on-board nitrogen generator intended to pump the 
inert gas into an emptying fuel tank was unveiled. The commenter states 
that all of this demonstrates the capabilities of industry.
    FAA Reply: While we agree with the commenter that the earlier 
systems were available, we do not agree that they were practical for 
commercial transport airplanes because of the cost, complexity, weight, 
and poor reliability of the systems. The FRM that will be certified for 
installation on Boeing Model 747 series airplanes reduces fuel tank 
flammability by inerting the tanks with nitrogen using hollow fiber 
membrane technology that does not require installation of an air 
compressor to produce NEA, thereby reducing cost, complexity, and 
weight. As previously discussed, more recent research has found that a 
simpler inerting system that reduces the oxygen concentration of the 
fuel tank to 12 percent or less at sea level is sufficient in achieving 
the desired safety level. No changes were made as a result of these 
comments.

Fuel Tank Harmonization Working Group

    Comment: The commenter points to several references throughout the 
preamble discussion to a flammability exposure of 2 to 4 percent and 
requests that this be changed to 5 percent. The commenter says that the 
ARAC, in their 1998 report, estimated wing fuel tank exposure as 5 
percent. The commenter also points to the reference to 3 percent 
flammability value for the wing fuel tanks in the preamble discussion 
of ``Definition of Transport Effects'' and requests that this also be 
changed to 5 percent.
    FAA Reply: We concur in part. Although the ARAC report did identify 
a flammability exposure of 2 to 6 percent in the Task Group 8 section, 
in other locations of the report a generalized value of 5 percent was 
used. In the original discussion in the proposed special conditions, we 
incorrectly referenced a range of 2 to 4 percent instead of the actual 
value of 2 to 6 percent. We consider the estimated range that was based 
on a flammability analysis of a number of different airplane models to 
be more representative of the wing fuel tank flammability range across 
various airplane models. No changes were made as a result of these 
comments.
    Comment: The commenter says that the data presented in the 
discussion of the Fuel Tank Harmonization Working Group should be for 
historical reasons, and the criteria used for determining the need for 
an FRM should be AC 25.981-2.
    FAA Reply: We do not concur. The purpose of AC 25.981-2 is to 
provide guidance for demonstrating compliance with Sec.  25.981(c) to:
     Minimize fuel tank flammability; and
     Mitigate the hazards if ignition of the fuel vapors 
occurs.

The AC does not provide criteria to determine if a system is required 
to reduce flammability in fuel tanks.
    We infer from the commenter's remarks that they believe these 
special conditions will mandate the installation of an FRM, which is 
not the case. These special conditions do not represent rulemaking to 
mandate the reduction of a fuel tank flammability system. Instead, they 
are required to support certification of novel features of the FRM not 
addressed by existing regulations, and include additional requirements 
to address warm day operations during ground, takeoff, and climb 
portions of the flight where previous accidents have occurred. No 
changes were made as a result of these comments.
    Comment: One commenter considers the flammability range of l5 to 30 
percent of fleet operating time for fuel tanks containing flammable 
vapors, as documented in the ARAC report, a large range. This range 
indicates that the actual percent depends on assumptions. This 
commenter believes that a Monte Carlo analysis should not be a part of 
the certification process as it is an analysis that is based on flawed 
assumptions. The commenter considers use of statistical methods more 
consistent with the FAA philosophy for fail-safe designs. The commenter 
believes that aviation safety would be undesirably low if a Monte Carlo 
analysis was used for the design and certification of navigation and 
guidance systems, ground proximity warning systems, weather radar, wind 
shear avoidance, engine fire protection, etc. Another commenter also 
contends that the assumptions used in the Monte Carlo analysis are not 
supported by historical data.
    FAA Reply: We do not concur with the first comment. The 15-30 
percent addresses the range of average flammability exposures across 
the airplane models in the fleet. Specific airplane models will have a 
fixed average flammability exposure. We do agree that variations in 
assumptions for the analysis could result in large differences in the 
results of the

[[Page 7811]]

flammability analysis. For this reason, the special conditions 
incorporate specific parameters that must be used when determining fuel 
tank flammability. The Monte Carlo methodology has been used in a wide 
range of industries to address safety concerns. Previous ARAC 
activities recommended use of the Monte Carlo method for calculating 
average fuel tank flammability exposure. This methodology has recently 
been used by industry to evaluate the flammability exposure of fuel 
tanks as part of the SFAR 88 activities. We therefore expect the 
applicant as well as industry already have a good understanding of how 
to use the model. No changes were made as a result of these comments.

FAA Rulemaking Activity

    Comment: The commenter notes that the ARAC recommendations 
referenced in this discussion did not use the word ``reduction.'' The 
commenter believes that the word ``reduction'' in Sec.  25.981(c) needs 
further study. The commenter also says that the 2 to 4 percent 
flammability of unheated aluminum wing fuel tanks should not be used as 
a criterion in the special conditions, and notes that AC 25.981-2 does 
not specifically address the center wing fuel tank like the special 
conditions but includes all tanks (including wing tanks).
    FAA Reply: We do not concur with the comment concerning the use of 
unheated aluminum wing fuel tanks as the criterion for an acceptable 
level of fuel tank flammability. AC 25.981-2 does provide clarification 
under section 5, paragraph (d)(3), that the intent of Sec.  25.981 is 
``to require that the exposure to formation or presence of flammable 
vapors is equivalent to that of an unheated wing tank in the transport 
airplane being evaluated.'' The special conditions incorporate the 
intent of Sec.  25.981(c) and also include additional requirements for 
warm day conditions where previous accidents have occurred. The special 
conditions also include requirements to address novel design features 
that are not covered under the applicable airworthiness standards of 
part 25. No changes were made as a result of these comments.

Fuel Tank Inerting

    Comment: Two commenters say the applicant's proposed design does 
not include an essential verification system (NEA sensors and 
indication) to ensure that the appropriate nitrogen concentrations will 
be directed into the fuel tank to displace the fuel vapors in the 
ullage space. One commenter compares this to the statement in the 
discussion of ``Criteria for Inerting'' that the combination of 
ignition prevention and reduction of flammable vapors in the tank will 
substantially reduce the number of future fuel tank explosions.
    FAA Reply: We do not concur. To comply with the special conditions, 
the applicant must demonstrate that the FRM meets the specific 
performance and reliability requirements. An indication system would be 
required if it is shown that the FRM cannot meet these requirements 
unless one is installed. No changes were made as a result of these 
comments.
    Comment: The commenter requests that the reference to ``using the 
size difference'' in the first paragraph be changed to ``using the 
absorption difference,'' as this would more accurately reflect how 
hollow fiber membranes function.
    FAA Reply: We concur with the commenter and revised the sentence to 
read: ``* * * the hollow fiber membrane material uses the absorption 
difference between the nitrogen and oxygen molecules to separate the 
NEA from the oxygen.''
    Comment: The commenter says that it does not have to be pressurized 
air from the airplane engines that is used to produce NEA; compressed 
air from any source can be used.
    FAA Reply: We agree, however these special conditions address a 
specific system design for the applicable Boeing Model 747 series 
airplanes using bleed air from the airplane engines to generate NEA. We 
recognize there may be other means to achieve the same goal. No changes 
were made as a result of this comment.
    Comment: The commenter contends that technology has not kept up 
with the need to eliminate the need for stored nitrogen because hollow 
fiber technology does not produce enough NEA to inert the center tank 
during all phases of flight, including descent. Hollow fiber 
technology, as described in the special conditions, will not inert the 
wing tanks.
    FAA Reply: We do not concur. The applicant has selected hollow 
fiber technology as a means to produce NEA to inert the center wing 
tank on Model 747 series airplanes. The applicant must show that the 
FRM will inert the center tank. Hollow fiber technology could be used 
to inert wing fuel tanks; however, there is no requirement in the 
special conditions to do so. No changes were made as a result of this 
comment.

Criteria for Inerting

    Comment: The commenter requests that this discussion be revised as 
shown below. The commenter says the FAA proposed wording implies that 
the 9 percent military and 12 percent commercial oxygen concentration 
values are intended to be equivalent. The 9 percent is a military limit 
for zero exposure. The 12 percent is a benchmark for evaluating 
minimization of flammability exposure, equivalent to wing tanks.

Criteria for Inerting

    Earlier fuel tank inerting designs produced for military 
applications were based on defining ``inert'' as a maximum oxygen 
concentration of 9 percent. One major finding from the research and 
development efforts conducted by the FAA was the determination that 
the 9 percent maximum oxygen concentration limit established to 
protect military airplanes was significantly lower than necessary to 
prevent significant pressure rise for the majority of ullage 
conditions. This FAA research supports a value of 12 percent as a 
benchmark at sea level for determining when the likelihood of 
significant pressure rise is low. The test results are currently 
available on FAA Web site: www.fire.tc.faa.gov, and will be 
published in FAA Technical Note `Limiting Oxygen Concentrations 
Required to Inert Jet Fuel Vapors Existing at Reduced Fuel Tank 
Pressures,' report number DOT/FAA/AR-TN02/79.
    It should be noted that the 12% benchmark is not intended to 
claim that ignition is impossible below 12%. 14 CFR 25.981 (c) 
requires minimization of flammability, not elimination. ARAC 
evaluations concluded complete elimination of flammability was 
impractical and unnecessary. 14 CFR 25.981(c) was based on reducing 
flammability exposure to equal or less than wing tanks, which have 
an acceptable safety history. The 12% benchmark is used to divide 
exposure time when significant pressure rise is unlikely, from 
exposure time when significant pressure rise is more likely. Testing 
indicates there is also significant ability to inhibit ignition for 
many fuel vapor conditions when oxygen content is above 12%, but no 
credit is taken for these conditions.
    As a result of this research and the 12 percent benchmark, the 
quantity of nitrogen-enriched air that is needed to inert commercial 
airplane fuel tanks was reduced. This reduction in nitrogen-enriched 
air, coupled with advancements in design technology, facilitates the 
development of an effective flammability reduction system that 
approaches simple and practical.''

    FAA Reply: We do not concur. The 12 percent requirement in the 
special conditions is based on testing of flammability using electrical 
ignition sources caused by airplane system failures. It is not intended 
to address combat threats. However, data from the Navy tests concludes 
that inerting to 9 percent oxygen has little benefit over 12 percent 
for protection of fuel tanks from overpressure caused by ignition from 
30 millimeter Hi energy incendiary rounds.

[[Page 7812]]

No changes were made as a result of this comment.

Type Certification Basis

    Comment: The commenter points to two statements concerning 
compliance with Sec.  25.981, which appear to be confusing regarding 
applicability to the FRM. First, the commenter asks for clarification 
as to the extent to which Sec.  25.981 is applied to the system. The 
commenter assumes it is only those areas exposed to fuel vapor under 
normal operation. The commenter also points to paragraph two of the 
``Novel or Unusual Design Features,'' which states that compliance is 
required for the changed aspects of the airplane by showing that the 
FRM does not introduce any additional potential ignition risk into the 
fuel tanks.
    FAA Reply: There are two aspects of the FRM concept. First, it is 
the means chosen to achieve the requirements of Sec.  25.981(c) to 
minimize fuel tank flammability for the applicable 747 series 
airplanes. In this case, the applicant chose to introduce NEA into the 
center wing tank and assure that it is dispersed throughout. Having 
made that choice, the applicant is required to ensure that the changes 
introduced by the system (i.e., FRM) do not introduce any potential 
ignition sources into the tank. No changes were made as a result of 
this comment.
    Comment: The commenter says that compliance with Sec.  25.981 
applies to certification of fuel tanks and not to the installation of 
an inerting system, although fuel tank inerting may be one way to show 
compliance with Sec.  25.981(c)(1).
    FAA Reply: We do not concur. The applicant has proposed to 
voluntarily comply with Sec.  25.981(c), amendment 25-102, for 
certification of the performance of an FRM to reduce flammability in 
the center wing fuel tanks of Model 747 series airplanes. Additionally, 
as stated in the preamble to these special conditions, the applicant 
must also ensure that installation of an FRM will meet the ignition 
source prevention requirements of Sec.  25.981(a) and (b), as well as 
all the other applicable part 25 regulations. No changes were made as a 
result of this comment.
    Comment: The commenter requests that the 747-Classics effectivity 
be removed from the special conditions. The commenter says that few 
747-Classics remaining in service may fall within the total 3 percent 
exposure criteria, and failing that should pose a far lower risk for 
the following reasons:
     The majority of ignition reduction modifications (IRM), 
including the improved maintenance procedures, will be implemented 
prior to any reasonable FRM compliance date;
     AD 98-20-40 fuel quantity indicating system protection 
upgrade has been fully incorporated on all 747-Classics; and
     With the two 737 accidents, it appeared that the center 
wing tank (CWT) fuel pumps were inadvertently left running with an 
empty CWT, and although it could not be confirmed that the pumps were 
at fault, the IRM requirement to automatically (or otherwise) shut 
pumps off at low pressure will eliminate this possible ignition source.

There may be an argument that the older airplanes are at a greater risk 
and therefore should be FRM protected, but the historical events and 
sample in-tank inspections tend to rebuff this proposition.
    FAA Reply: We disagree with the commenter that the center wing fuel 
tank on 747 Classic airplanes falls within the 3 percent fleet average 
flammability exposure criteria because initial flammability exposure 
analyses of these airplane models has shown the flammability to be well 
above 3 percent. We estimate there are currently about 95 747-100, -
200, and -300 airplanes in service today in the United States. Though 
ignition source prevention ADs have been incorporated on these 
airplanes and additional ADs will be incorporated as a result of SFAR 
88 rulemaking, as we said earlier in this document experience 
demonstrates that all possible ignition sources cannot be determined 
and reliably eliminated. Reducing or preventing flammable vapors from 
forming in high flammability fuel tanks will significantly improve fuel 
tank safety. These special conditions support certification of the 
applicant's FRM design for possible installation on Boeing Model 747 
series airplanes. These special conditions do not mandate any changes 
to current airplanes. No changes were made as a result of these 
comments.

Novel or Unusual Design Features

    Comment: The commenter requests that the phrase ``by showing that 
fuel tanks'' in the second paragraph of this discussion be deleted 
because the beginning of the sentence establishes the requirement to 
comply with Sec.  25.981(a), and (b). The method of compliance is the 
applicant's responsibility.
    FAA Reply: We do not concur with the commenter. This last phrase 
provides a condensed explanation to the reader of what is required for 
compliance with Sec.  25.981(a) and (b). No changes were made as a 
result of this comment.
    Comment: This comment concerns the discussion of how the applicant 
proposes to operate the FRM. The commenter says the applicant must be 
allowed the freedom to design the system and must ensure that all 
features of the FRM are addressed properly so that hazardous conditions 
do not occur and the system complies with Sec. Sec.  25.1301 and 
25.1309 and other applicable requirements.
    Another commenter requests that the system description be replaced 
by the following to focus on requirements and not prescribe design:

    The proposed FRM uses a nitrogen generation system (NGS). Engine 
bleed air will flow through an air separation module (ASM) that will 
separate the air stream into nitrogen-enriched air (NEA), which will 
be supplied to the center fuel tank, and oxygen-enriched air (OEA), 
which will be exhausted overboard. The FRM will also include 
modifications to the fuel vent system. Certain features of the FRM 
may introduce a hazard to the airplane if not properly addressed.

    FAA Reply: We do not concur with the commenters. This section of 
the special conditions preamble appropriately defines what the novel or 
unusual design features of the FRM are that require special conditions 
under Sec.  21.16. No changes were made as a result of these comments.
    Comment: This commenter says the special conditions do not 
adequately address the descent control valve function as it relates to 
the high flow versus low flow mode. The Monte Carlo analysis is not 
based on test data or historical data to predict the effectiveness of 
the NGS on descent.
    FAA Reply: We do not concur. The special conditions require that 
the applicant validate the inputs to the Monte Carlo analysis by ground 
and flight tests and substantiate that distribution of NEA is effective 
at inerting the fuel tank for the performance conditions required. No 
changes were made as a result of these comments.
    Comment: It is the commenter's opinion that the proposed 10-day 
MMEL relief for the system is unjustified. The commenter says all 
components are Line Replaceable Units (LRU) that can be replaced within 
``typical'' turn around time. A long relief time defeats the purpose of 
the system. If limited dispatch relief is granted, then it should be 
restricted to conditions (cold temperature) in which development of 
flammable vapors in the fuel tank is of low probability. The commenter 
points to AC 25.981-2,

[[Page 7813]]

paragraph 4(h), which addresses limited operations based on outside air 
temperature.
    FAA Reply: The special conditions do not approve an MMEL dispatch 
interval. As stated previously, even though the applicant has proposed 
a 10-day MMEL dispatch interval, the Flight Operations Evaluation Board 
(FOEB) will determine and approve the appropriate MMEL relief intervals 
based on data submitted by the applicant. The applicant must show that 
the fleet average flammability exposure of a tank with an FRM installed 
is equal to or less than 3 percent, including operating time with an 
FRM. No changes were made as a result of these comments.
    Comment: This commenter says the MMEL procedure is a result of 
system design (safety system or not, redundancy, etc.) and reliability 
of the system. It is up to the applicant to design their system to 
satisfy both the regulations and their customers.
    FAA Reply: We concur. The special conditions require the applicant 
to submit data that show compliance with the special conditions for 
their proposed MMEL dispatch interval. The FOEB will assess the data in 
determining if the interval is appropriate. No changes were made as a 
result of this comment.
    Comment: The commenter contends that the existing technology for 
hollow fiber technology presently has a mean time between failure 
(MTBF) of less than 2,000 hours, which is different than the 5,000 
hours identified in this section.
    FAA Reply: To comply with the specific reliability requirements, 
the applicant will have to consider the MTBF or life limit of the 
hollow fiber technology in their FRM design. The design and compliance 
with the special conditions will dictate what the MTBF will be. No 
changes were made as a result of this comment.

Discussion

    Comment: Three commenters contend that the statement ``* * * due to 
factors such as the limited availability of bleed air and electrical 
power, it is not considered practical at this time to develop systems 
for retrofit * * *'' is not appropriate and is incorrect. One commenter 
says this issue would be better addressed in documentation and 
discussion rather than this section of the special conditions. The 
discussion should be limited to the issues considered and the data 
presented in the proposed special conditions. The second commenter says 
that on all commercial airplanes during normal operation (all engines 
operating and all generators operating), excess bleed-air and 
electrical power is available. The last commenter requests removal of 
the words ``Since amendment 25-102 was adopted,* * * it is not 
considered practical at this time to develop systems for retrofit into 
existing airplane designs that can maintain a non-flammable tank ullage 
in all fuel tanks or during all operating conditions.'' The commenter 
says the wording suggests that a more stringent requirement than that 
established by amendment 25-102 has been demonstrated to be practical. 
The FAA has not proposed, substantiated, or adopted rulemaking to 
support this statement. Changes to the requirements of Sec.  25.981(c) 
are not the subject of these special conditions.
    FAA Reply: We do not concur with the commenters but believe 
clarification is needed to fully understand the context of the 
statement that is at issue. As stated earlier, the FAA Administrator 
has made public statements concerning our intention to propose 
rulemaking that would amend Sec.  25.981(c). During the public process 
following issuance of any proposal, comments will be welcome. The 
purpose of this statement in the special conditions is to provide 
justification for the level of performance required within the 
proposal. Although the complexity and sizing of inerting technology has 
been reduced such that it is a viable method for reduction of 
flammability in fuel tanks, there are still restrictions in existing 
airplanes today that would limit an inerting system from being 100 
percent effective at inerting the fuel tank during all operating 
conditions. No changes were made as a result of these comments.
    Comment: One commenter expresses concern that an FRM that complies 
with Sec.  25.981(c), amendment 25-102, may not preclude fuel tanks 
from routinely being flammable under the specific operating conditions 
present when recent accidents occurred. The commenter says that if the 
FAA believes the above statement is true, then it has not specified the 
right regulations. The commenter believes a repeat of the Philippine, 
TWA, or Thai incidents would be prevented by compliance with Sec.  
25.981(c).
    FAA Reply: The FRM is intended to add an additional layer of safety 
for high flammability fuel tanks by reducing the existence of flammable 
vapors in the center wing tank. It is important to recognize that this 
system does not totally eliminate flammable vapors in the tank during 
all operating conditions. The special conditions include requirements 
that will address specific risk elements for warm day ground and climb 
profiles where accidents have occurred which is a more stringent 
requirement than Sec.  25.981(c). The FRM will augment the ignition 
source prevention measures in substantially reducing the risk for 
future fuel tank explosions. No changes were made as a result of these 
comments.

Definition of Inert

    Comment: One commenter believes that 12 percent oxygen 
concentration at sea level cannot be assured unless the oxygen 
percentage within the ullage of the fuel tank is monitored and 
measured. The commenter says oxygen monitoring by percentage is needed 
to verify if the center wing fuel tank is inert per the definition 
supplied in the special conditions, and to determine if the inerting 
system is inoperative. The commenter says there is a need to know the 
oxygen concentration in the center tank for airplanes operated in 
warmer climates. If NEA is lost, the risk factor needs to be accounted 
for in the analysis. If it is lost because of a leak surrounding the 
NGS, there will be a higher than normal oxygen level in that 
compartment. The commenter would encourage further investigation, 
testing, and analysis of existing data to support the definition of 
inert in all locations and all fuel tanks for the Model 747 series 
airplanes and eventually on the Model 737, 757, 767, and 777 airplanes, 
as referenced in the ``Novel or Unusual Design Features'' discussion.
    Two commenters believe that the level of oxygen concentration 
should be monitored at the most critical location in the fuel tank to 
verify adequate system operation. One of the commenters believes that 
an indication should be generated if the oxygen concentration in the 
fuel tank rises above the maximum allowable concentration for greater 
than a specified time. This would prevent transient conditions from 
generating nuisance indications. The other commenter says that the 
system indications should monitor adequate system performance 
throughout the flight profile, which is something a periodic ground 
check cannot ensure. Besides the obvious safety and reliability 
benefits, it is not understood how else the reporting requirements of 
special condition III(c) could be met. Although AC 25.981-2 does not 
require cockpit indications for an inerting system, this commenter 
would support rulemaking intended to revise AC 25.981-2.
    Two commenters believe that an indication system that displays the 
inerting system functionality should be available to the flightcrew. 
Relying solely on preflight or ground crew checks leaves out a valuable 
resource for

[[Page 7814]]

monitoring the system status. The flightcrew should be aware if the 
system is functioning. If it is not, changes in the flight profile 
should be made to ensure the airplane is out of the regime where the 
center fuel tank is in the most danger.
    FAA Reply: We do not concur with the commenters. There are no 
requirements in the special conditions for oxygen concentration 
monitoring, but there is nothing that precludes a monitoring system and 
associated crew indications from being developed. While monitoring of 
oxygen concentrations is one means of determining system performance, 
other indications such as pressure measurements, flow measurements, 
valve positions etc., as well as periodic functional checks may be used 
to provide assurance that the system is functional. The concerns listed 
by the commenters are included in the analysis and testing the 
applicant must perform to show that the FRM meets the special condition 
flammability and reliability requirements. No changes were made as a 
result of these comments.
    Comment: The commenter requests the word ``localized'' in the 
second sentence of the first paragraph in this section be deleted. The 
commenter also requests that the rest of the paragraph after the second 
sentence (i.e., ``Currently there is * * * be considered inert'') be 
deleted. The commenter believes the addition of a requirement to 
individually address all tank compartments is not in accordance with 
the principles used to date to develop a practical and commercially 
viable system that will minimize the average fleet flammability 
exposure. It is already conservative to estimate flammability based on 
average fuel temperature because the average fuel temperature is 
typically higher than the majority of the tank surfaces. This approach 
represents the theoretical flammability of a tank where all the tank 
surfaces are at this uniform temperature. In reality, when the fuel 
temperature is high enough to result in evolution of sufficient vapors 
to cause a flammable ullage near the fuel surface, the temperatures of 
the sides and top of the fuel tank are cooler, resulting in 
condensation that significantly reduces the actual flammability of the 
tank ullage.
    FAA Reply: We concur, in part, with the commenter. We have revised 
the definition of ``flammable'' in the special conditions to read, 
``With respect to a fluid or gas, flammable means susceptible to 
igniting readily or to exploding (14 CFR part 1, Definitions). A non-
flammable ullage is one where the gas mixture is too lean or too rich 
to burn and/or is inert per the definition below.''
    We do not concur with the comment that the bulk average fuel 
temperature should be used to determine flammability. The ARAC used a 
bulk average fuel temperature to provide a comparative flammability 
level for various fuel tanks on different airplane models. The ARAC 
used a simplified methodology that assumed the fuel tank was one large 
volume and that the liquid fuel and fuel vapor in the tank would mix, 
forming a uniform mixture. In this case, using the bulk average fuel 
temperature would provide a realistic representation of the actual fuel 
tank flammability.
    This simplified approach, however, does not reflect the actual 
design of some fuel tanks. In reality, some fuel tanks have 
significantly different flammability exposures within different 
compartments of the fuel tank due to barriers installed in the tank, to 
prevent sloshing of fuel. These barriers do not allow significant 
mixing of the fuel and vapors. For example, some center fuel tanks 
extend from the center wing box out into the wing. Other tanks located 
in the center wing box have barriers that create separate compartments 
within the tank. In these cases, the portion of the fuel tank in the 
wing or that exposed to a cold air source may be much cooler and little 
mixing within the different portions of the fuel tank would occur. If 
the fuel temperature in the part of the tank located in the wing or 
other colder section were used in the analysis, the results would not 
represent the actual flammability of those portions of the tank where 
cooling did not occur. We have therefore modified the special 
conditions to revise the discussion in appendix 2 to address those 
airplanes that have significantly different flammability exposures 
within different compartments of the fuel tank due to the design of the 
tank, such as a center fuel tank that extends from the center wing box 
out into the wing. For these fuel tanks, the appendix requires 
evaluation of the compartment with the highest flammability for each 
flight phase. We do not expect that determining which compartment to 
evaluate will require a detailed analysis of each compartment. In most 
cases, a qualitative assessment, considering ambient temperatures and 
other relevant factors will be sufficient.

Determining Flammability

    Comment: This commenter says the Monte Carlo analysis should also 
consider the center tank theoretically in an unheated condition, not 
heated by adjacent equipment.
    FAA Reply: We do not concur. The Monte Carlo analysis as used in 
these special conditions is specific for determining fuel tank 
flammability exposure and certifying an FRM that reduces the 
flammability of a specific center wing tank. No changes were made as a 
result of this comment.
    Comment: This commenter points out that in the second paragraph of 
the ``Flammability'' discussion the FAA says ``to quantify the fleet 
exposure, the Monte Carlo analysis approach is applied to a 
statistically significant number (1,000,000) of flights where each of 
the factors described above is randomly selected.'' Table 6 in appendix 
2 of the special conditions defines lower flammability limits if the 
applicant chooses to use fewer than 1,000,000 flights. The commenter 
says the number of runs should be defined as ``when the average results 
become stable,'' and the criteria for assessing these results should 
then be 3 percent.
    FAA Reply: We do not concur. Monte Carlo analyses in general 
require the applicant to run a large number of cases for the results to 
be accurate. The special conditions contain a method for an applicant 
to run fewer cases if they are able to show that they meet the required 
3 percent fleet average and 3 percent warm day flammability exposure 
limits for the fuel tank under evaluation. No changes were made as a 
result of this comment.
    Comment: The commenter requests that the following sentence be 
added to the end of the last paragraph of the ``Flammability'' 
discussion: ``However, fresh air drawn into an otherwise inert tank 
during descent does not immediately saturate with fuel vapors, and 
hence localized concentrations above the inert level during descent do 
not represent a hazardous condition.'' This is because fresh air drawn 
into the fuel tank through the vent during descent is not flammable, 
and will not cause the tank to become flammable during descent. Fresh 
air near the vent has not had the time necessary to mix with the bulk 
tank ullage, and thus will not be inert. However, the same lack of 
mixing time also precludes the presence of a flammable vapor level in 
this same region. Counting these non-hazardous periods as ``flammable'' 
would increase system size, weight, and associated costs with no 
benefit.
    FAA Reply: We concur and have modified the preamble discussion of 
``Determining Flammability'' to add the following sentence: ``However, 
fresh air drawn into an otherwise inert tank during descent does not 
immediately saturate with fuel vapors; hence, localized concentrations 
above the inert

[[Page 7815]]

level during descent do not represent a hazardous condition.''

Definition of Transport Effects

    Comment: One commenter says the FAA statement that the effects of 
mass loading and the effects of fuel vaporization and condensation with 
time and temperature changes have been excluded is flawed, because FAA 
documents clearly indicate that ``transport effects'' are important. 
Another commenter also believes that the analysis model should include 
``transport effects'' as well as flammability effects on heated 
unusable (empty, 0 quantity indication) fuel in the center wing tank. 
This second commenter says the fuel temperature within a specific 
compartment of the tank could be within the flammable range for the 
fuel type being used if the tank was empty and heat sources were next 
to the compartment.
    FAA Reply: We do not concur with the commenters. As stated in the 
definition of ``transport effects'' in the special conditions and the 
earlier discussion, this term includes two physical phenomena that 
affect the concentration of fuel vapor in the fuel tank ullage. The 
first is referred to as low fuel conditions or ``mass loading.'' At low 
fuel quantities there may be insufficient fuel in the fuel tank at a 
given pressure and temperature for the concentration of fuel vapor to 
reach the equilibrium level that would form if fuel were added to the 
tank.
    The second is the change in fuel vapor concentration in the fuel 
tank ullage caused by fuel condensation and vaporization. This change 
in fuel vapor concentration is caused by temperature variations on the 
fuel tank surfaces that result in a vapor concentration different from 
the concentration calculated using the bulk average fuel temperature.
    We excluded both of these effects because they were not considered 
in the original methodology ARAC used to establish the proposed 
flammability requirements. If this effect had been included in the wing 
tank flammability exposure calculation, it would have resulted in a 
significantly lower wing tank flammability exposure benchmark value.
    The ARAC analysis also did not consider the effects of the low fuel 
condition (or ``mass loading''), which would lower the calculated 
flammability exposure value for fuel tanks that are routinely emptied, 
such as center wing tanks. As explained earlier, when the amount of 
fuel is reduced to very low quantities within a fuel tank, there may be 
insufficient fuel in the tank to allow vaporization of fuel to the 
concentration that would be predicted for any particular temperature 
and pressure.
    No changes were made as a result of these comments.

Flammability Limit

    Comment: The commenter requests that the reference to ``during 
descent'' be changed to ``after high rate descent'' to more accurately 
reflect conditions.
    FAA Reply: We do not concur. The commenter provided no 
substantiation to clarify why they believe the tank would be able to 
maintain an inert ullage during descent mode that is not classified as 
a high rate of descent. Both the performance of the FRM and the rate of 
descent may impact the oxygen concentration level in the fuel tank and 
both need to be considered. No changes were made as the result of this 
comment.
    Comment: The commenter says that the 3 percent exposure criteria, 
referenced in this discussion, appears to be premised on the good 
service history of main and non-heated reserve fuel tanks. However, 
heated center wing tanks (CWTs) make up only a small percentage of the 
total number of tanks in use. If the exposure times for non-heated 
tanks are summed, it is likely to be close to the total overall 
exposure period for heated CWTs. If exposure period were the only 
criterion, then one would expect to see non-heated tank incidents. It 
is probable that the operating requirements (fuel remaining in tanks) 
have as much to do with the good service history as the exposure level. 
SFAR 88 Ignition Reduction Modifications will significantly reduce the 
ignition risk of the heated CWT to a level where perhaps they are not 
quite as safe as the main tanks but on a false premise. If the non-
heated tanks had an average 6 percent exposure, it is unlikely that the 
service history would differ. Setting the exposure design criteria to 3 
percent or lower may not be as relevant as indicated in these special 
conditions, and even a small shift upward could significantly influence 
the cost of installation and maintenance. A more important criterion 
could be the fact that many CWT components remain uncovered for the 
majority of time, with the possibility of an intermittent latent 
ignition type defect coming into play when inerting is unavailable. 
Therefore, the commenter states it may be more appropriate to consider 
additional MMEL limitations to help mitigate whatever is the remaining 
exposure risk. This may include ensuring that if CWT components fail, 
power is removed and not reapplied until the component is replaced and/
or some fuel is left in the CWT under certain defect conditions. It 
should also be noted that it is important to ensure that inerting does 
not become a substitute over time for the quick and effective clearance 
of CWT defects.
    FAA Reply: We agree with the commenter concerning the limitations 
of ignition source prevention. Minimization of ignition sources, such 
as component failure, removal of power, etc., was the goal of SFAR 88 
but it is recognized that absolute elimination of ignition sources is 
not possible. Flammability reduction provides a significant improvement 
in fuel tank safety in conjunction with ignition source prevention but, 
as such, it is important to recognize that this system will not 
necessarily eliminate all flammable vapors at all operating conditions. 
However, the warm day flammability exposure requirements in these 
special conditions would prevent fuel tank flammability during 
conditions where the past three fuel tank explosions occurred. By 
combining the two approaches, the risks for fuel tank explosions can be 
substantially reduced. Compliance with the special conditions will also 
ensure that neither the performance nor the reliability of the FRM will 
be greater than 1.8 percent of the fleet average flammability exposure, 
thereby further minimizing the exposure risk. The MMEL for each 
airplane model was reviewed as part of SFAR 88 and limitations on 
operations. We do not believe that additional MMEL requirements would 
be needed unless the FRM is unable to meet the performance, 
reliability, or warm day requirements in the special conditions. No 
changes were made as a result of these comments.

Specific Risk Flammability Limit

    Comment: The commenter says that because the issue of fuel tank 
flammability is primarily one of specific risk, they do not understand 
why the Monte Carlo analysis does not include MMEL relief and dispatch 
with the FRM inoperative in the evaluation of specific risk against the 
requirement of special condition paragraph II (b).
    FAA Reply: We did not include the effect of MMEL in special 
condition paragraph II (b) because the intent is to address the 
performance of the FRM under warm day conditions on the ground, in 
takeoff, and in climb, which are high risk. The fleet average 
flammability exposure includes the affects of reliability and including 
this in the warm day (that is, specific risk) is redundant. No changes 
were made as a result of this comment.
    Comment: The commenter requests that reference to ``conducting a 
separate

[[Page 7816]]

Monte Carlo'' be changed to ``analyzing a subset of the fleet average 
Monte Carlo'' to more accurately reflect how the analysis has been 
developed.
    FAA Reply: We do not agree. The applicant can analyze either a 
subset of an overall analysis or conduct a separate Monte Carlo for the 
warm day ground, takeoff, and climb cases. The applicant is still 
required to run the analysis to meet the fuel tank flammability 
exposure limit for the number of simulated flights as shown in Table 6 
of appendix 2. No changes were made to the special conditions because 
the method has not been limited.

Inerting System Indications

    Comment: The commenter says the four elements (when the FRM is 
operating and inerts the tank, when the FRM is operating but does not 
inert the tank, when the FRM is not operating properly and the operator 
is unaware of the failure, and when the FRM is not operating and is on 
the MMEL) mentioned in the first paragraph of this discussion should be 
included for fleet average flammability exposure. Paragraph II (e) of 
the special condition states that ``sufficient accessibility for 
maintenance personnel, or the flightcrew, must be provided to FRM 
status indications that are necessary to meet the reliability 
requirements of paragraph II (a) of these special conditions.'' The way 
this special condition is written is unclear and leaves it to the 
applicant's opinion of what the ``status indication'' should be. The 
commenter would therefore like to see this special condition explicitly 
address the four elements mentioned above.
    FAA Reply: We do not concur. The special conditions require the 
overall FRM reliability to meet a minimum standard and allow the 
applicant to optimize the design. The type of indications that would be 
required to meet the reliability requirements is design dependent; 
therefore, the special conditions do not require specific indications. 
No changes were made as a result of this comment.
    Comment: This commenter believes it would be cost beneficial and 
easier for operators if the look and feel of the FRM indication system 
is the same across all fleets. Operators already deal with different 
indication design philosophy across different fleets, so the argument 
of consistency is not appropriate. Where possible and depending on 
cost, a strong consideration should be made to align the FRM indication 
with existing indication philosophy. In the case of the 747-400, this 
should be by way of an Engine Indication and Crew Alert System (EICAS) 
status message.
    FAA Reply: We do not concur. As stated earlier, the special 
conditions do not dictate a specific design but rather state that 
indication and/or maintenance checks will be required to ensure that 
the performance and reliability of the FRM meets the special condition 
requirements. The look and feel of an indication system is beyond the 
scope of these special conditions. No changes were made as a result of 
these comments.
    Comment: The commenter believes that an FRM requires a redundant 
system to address any future foreseeable events and/or conditions. 
Consideration should be given to apply the FRM on newly certificated 
airplanes, and only where it is feasible to existing airplanes.
    FAA Reply: We do not concur. As stated earlier, the FRM is intended 
to be a system that provides an additional layer of protection by 
reducing the exposure to flammable vapors in the heated center wing 
fuel tank. This protection, when added to ignition prevention measures, 
will substantially reduce the likelihood of future fuel tank explosions 
in the fleet. These special conditions are only applicable to 
certification of an FRM for the affected 747 series airplanes for which 
an application was received. No changes were made as a result of these 
comments.
    Comment: The special conditions state that, ``at a minimum, proper 
function of essential features of the system should be validated once 
per day by maintenance review of indications or functional checks, 
possibly prior to the first flight of the day.'' This is a specific 
implementation and is taken to be for 747 series airplanes only. If the 
special condition material is intended to be used for other projects, 
the sentence should be ``proper function of essential features of the 
system should be monitored.''
    FAA Reply: The special conditions require that the FRM for the 
applicable 747 airplanes meet specific performance and reliability 
requirements. Various design methods to ensure this may include a 
combination of system integrity monitoring and indication, redundancy 
of components, and maintenance actions. The initial 747 FRM design 
features, as presented to the FAA, would require daily monitoring of 
system performance to meet the reliability requirements. Daily checks 
may not be needed on all FRM and are only one way of monitoring proper 
function of essential system features. Continuous system monitoring by 
maintenance computers with associated maintenance messages may also be 
used. A combination of maintenance indication or maintenance check 
procedures could be used to limit exposure to latent failures within 
the system, or high inherent reliability may be used to make sure the 
system will meet the fuel tank flammability exposure requirements.
    The type of FRM indications and the frequency of checking system 
performance (maintenance intervals) must be determined as part of the 
FRM fuel tank flammability exposure analysis. These special conditions 
will be used as the starting point for developing special conditions 
for other airplane models, listed in the preamble, for which the 
applicant is considering certification of an FRM. No changes were made 
to these special conditions as a result of these comments because they 
are applicable to the 747.
    Comment: Two commenters question the same discussion in the 
preamble, specifically the sentence that reads, ``if system maintenance 
indication is not provided for features of the system essential for 
proper system operation, system functional checks will be required for 
these features. They believe that, at a minimum, proper function of 
essential features of the system should be validated once per day by 
maintenance review of indications or functional checks, possibly prior 
to the first flight of the day.'' The comments indicate the commenter 
interpreted the statement to mean that daily checks are required. One 
commenter says that accomplishing the functional checks prior to the 
first flight of the day is not practical, because maintenance personnel 
are not available at all destinations. It could be 2 to 3 days before 
the affected airplanes would be at an appropriate location where 
maintenance is available. The validation check would better align with 
the operators' maintenance programs if the interval were based on 
flight hours. The applicant and airplane operators have discussed this 
topic at length, and believe that an interval of 75 flight hours would 
provide a conservative validation of the system's functionality and 
allow the check to be accomplished by qualified maintenance personnel. 
The commenters also say there is no historical data to support FRM 
validation only once per day. They recommend continuous monitoring.
    FAA Reply: As discussed earlier, we concur with the commenters that 
the need for daily checks will depend on the FRM design. The preamble 
discussion was not intended to mandate daily checks by maintenance 
personnel. As noted earlier, the need for system

[[Page 7817]]

functional checks and the interval between the checks will be 
established based on the level of ``system maintenance indication 
provided for features of the system essential for proper system 
operation'' and the reliability of the system. If continual system 
monitoring is provided or features of the system have high inherent 
reliability, daily checks would not be needed to meet the reliability 
requirements in these special conditions. As we stated in the preamble, 
the determination of a proper interval and procedure will follow 
completion of the certification testing and demonstration of the 
system's reliability and performance prior to certification. The time 
interval between system health checks and maintenance will be 
established by the reliability analysis, any airworthiness limitations, 
and the FOEB. We agree with the commenter that providing a design with 
continuous system monitoring is desirable; however, we do not agree 
that this feature should be required by the special conditions because 
it would mandate specific design features and not allow design freedom. 
No change was made as a result of these comments.
    Comment: Concerning accomplishment of a daily check for proper 
function of the FRM, the commenter says past experience has shown that 
extended ground time and maintenance induced errors can happen. The 
commenter also contends this is contradictory to the statement that, 
``determination of a proper interval and procedure will follow 
completion of the certification testing * * *.'' The commenter 
recommends that the maintenance review board (MRB) procedure, outlined 
in AC 121-22, be used to develop the Instructions for Continued 
Airworthiness.
    FAA Reply: Instructions for Continued Airworthiness are established 
as part of certification of the FRM to the performance and reliability 
requirements in these special conditions. The MRB procedure, as 
outlined in AC 121-22, will be used to define how an MRB will be 
conducted. No changes were made as a result of these comments.
    Comment: Concerning the MMEL dispatch inoperative interval, four 
commenters believe the proposed MMEL interval of 10 days should be 
shortened and the FRM be operational to the maximum extent practical. 
One commenter says 10 days represents approximately 2.74 percent of a 
year, and contends that the FRM components (bleed-air control valve, 
ozone converter, heat exchanger, filter, and ASM) can be readily 
removed and replaced by a line mechanic during a typical turnaround. 
The commenter believes that several of the FRM components can cause 
system malfunction (produce low quality NEA) without any indication. 
These malfunctions cannot be predicted by analysis or by test. A second 
commenter notes that the FAA and industry have adopted a 3-day MMEL 
relief interval for other inoperative safety systems, such as flight 
data recorders, while another commenter states that catastrophic events 
brought about the development of an FRM; therefore, the importance of 
such a system is easily seen.
    FAA Reply: We do not concur with the commenters regarding setting a 
specific MMEL interval in the special conditions. The FOEB process, as 
previously discussed, will determine the appropriate MMEL dispatch 
interval. No changes were made as a result of these comments.
    Comment: One commenter believes that if the reliability analysis 
shows that a 10-day MMEL will allow the overall fleet flammability 
exposure limit to meet the requirements listed in the special 
conditions, then the 10-day MMEL should be acceptable. A second 
commenter requests clarification that the MMEL relief will be 
determined using standard methods, and that the reference to warm 
climates in the last paragraph of this section refers to inclusion in 
the Monte Carlo analysis and not to a limitation in the MMEL specific 
to warm ambient temperatures.
    FAA Reply: The standard processes (FOEB review), as discussed 
above, will be used to determine the appropriate MMEL dispatch 
interval. These same processes may also determine if a limitation is 
needed in the MMEL for warm day operation based on the results of the 
analysis. No changes were made as a result of these comments.
    Comment: The commenter says that if the FRM is inoperative, there 
might be some conditions in which the percentage of oxygen 
concentration is as high as 30 percent while the airplane is in the 
climb flight profile. An operational consideration might be to transfer 
fuel into the center tank or to carry extra fuel in that tank until 
level cruise is attained. This procedure addresses the internal energy 
sources discussed in current advisory circulars. The commenter contends 
that whether or not the FRM is in low or high flow mode, it cannot keep 
up with the need due to pressure and temperature changes and out-
gassing of the fuel.
    FAA Reply: We do not concur. The special conditions require that 
the flammability analysis take into account any periods where the FRM 
is inoperative or does not have the capacity to maintain a non-
flammable fuel tank ullage. We agree with the commenter that out-
gassing of dissolved air in the fuel may affect the oxygen 
concentration in the fuel tank during certain flights. These special 
conditions require that this factor be considered when determining the 
portion of the flammability exposure evaluation time (FEET) when the 
FRM cannot maintain a non-flammable ullage. This portion of the fleet 
average flammability exposure is limited to 1.8 percent. The special 
condition requirements are intended to provide an additional layer of 
protection to the existing certification standards that require designs 
to preclude fuel tank ignition sources. This balanced risk management 
approach of precluding ignition sources and reducing flammability 
exposure in certain fuel tanks provides two independent layers for 
preventing fuel tank explosions in those tanks. No changes were made as 
a result of these comments.
    Comment: The commenter requests that the entire discussion of 
``Inerting System Indications'' be reworded. It is the commenter's 
position that the special conditions should establish the certification 
requirements not already established by existing part 25 requirements. 
The commenter says that the reliability requirement for the FRM is 
clearly established in paragraph II (a)(2) of the special conditions as 
to not contribute more than 1.8% overall fleet flammability exposure. 
The commenter believes the required inspections and associated 
inspection intervals should be developed by the applicant in support of 
complying with the 1.8% limit. The applicant should have the 
flexibility to design a system that has high reliability (at higher 
equipment cost) with fewer inspections required, or lower reliability 
and higher frequency of inspection with less time allowed for MMEL 
dispatch. The commenter also believes that this is consistent with 
Sec.  25.981(c), amendment 25-102, where it specifically states that 
``minimize'' means to incorporate practicable design methods to reduce 
the likelihood of flammable vapors.
    FAA Reply: We do not concur. The special conditions do provide the 
applicant with flexibility to design the FRM either to higher 
reliability and longer inspection intervals or lower reliability with 
more frequent inspections, as long as the contributions for either 
performance of the system or its reliability are not greater than 1.8 
percent of the total 3 percent fleet average flammability exposure. The 
approved maintenance procedures and

[[Page 7818]]

intervals established by the FOEB will be based on the applicant's 
fleet average flammability exposure data submitted to the FAA. No 
changes were made as a result of these comments.

Affect of FRM on Auxiliary Fuel Tank System Supplemental Type 
Certificates

    Comment: The commenter believes the applicant should validate, as 
part of the certification effort, that the performance and reliability 
requirements for the FRM are met for any approved combination of 
auxiliary fuel tank installations. The commenter does not understand 
how installation of an FRM on an airplane with auxiliary fuel tanks can 
be adequately assessed ``during development and approval of the service 
bulletin for the FRM.''
    FAA Reply: We concur and have added a requirement in special 
condition II (a)(3) for the applicant to ``identify critical features 
of the fuel tank system to prevent an auxiliary fuel tank installation 
from increasing the flammability exposure of the center wing tank above 
that permitted under paragraph II (a)(1) and (2) and to prevent 
degradation of the performance and reliability of the FRM.'' We have 
also added a requirement under paragraph III (a)(3) to establish 
airworthiness limitations to address these features.

Disposal of Oxygen-Enriched Air

    Comment: One commenter refers to the statement, ``the OEA produced 
in the proposed design is diluted with air from a heat exchanger, which 
is intended to reduce the OEA concentration to non-hazardous levels.'' 
The commenter says that although this is a particular solution to the 
hazard, it should not be seen as the only solution. The term 
``hazardous'' is open to interpretation; thus, this discussion is 
considered as too design specific.
    FAA Reply: We agree with the commenter that there are a number of 
different means of addressing any hazards associated with the OEA. 
These special conditions are applicable to the applicant's proposal for 
certification of their FRM design. The description of the particular 
design feature noted by the commenter was not intended to limit other 
means of compliance should another applicant propose an FRM. We will 
evaluate each FRM based on the proposed design. No changes were made as 
a result of these comments.
    Comment: The commenter requests that the first paragraph of this 
discussion be replaced with the following: ``The FRM produces both 
nitrogen-enriched air (NEA) and oxygen-enriched air (OEA). The OEA 
generated by the FRM could result in a fire hazard if not disposed of 
properly. Compliance with existing requirements of Sec.  25.863 are 
sufficient to address potential leakage of OEA due to failures and safe 
disposal of the OEA during normal operation.'' The commenter requests 
this change to make OEA leakage compliance requirements consistent with 
those applicable for other flammable leakage zone items.
    FAA Reply: We concur with the commenter that certification of the 
FRM will require the applicant to evaluate installation of equipment in 
a flammable fluid leakage zone for compliance with Sec.  25.863. 
However, compliance with Sec.  25.901 is required to ensure that no 
single failure or malfunction, or probable combination of failures, 
will jeopardize the safe operation of the airplane. Depending on where 
the OEA is discharged, other part 25 regulations might apply. No 
changes were made as a result of these comments.

Applicability

    Comment: The commenter notes that the airplane applicability is not 
consistent. Furthermore, the commenter says Sec.  25.981(c), amendment 
25-102, is only applicable to new type designs, and therefore these 
special conditions should apply to new type designs and may extend to 
newly built airplanes. If the special conditions were proposed for 
other Boeing Model airplanes (737, 777, etc.), the commenter believes 
the standards established for the 747 airplanes should also be 
applicable for these models.
    FAA Reply: We concur with the commenter that the airplane 
applicability was inconsistent in certain sections of the proposed 
special conditions in that these sections excluded the 747-100B and 
747-300 series airplanes. We have corrected the applicable sections of 
the final special conditions to show the applicability as Boeing Model 
747-100/200B/200F/200C/SR/SP/100B/300/100B SUD/400/400D/400F series 
airplanes. The applicant has voluntarily proposed to show compliance 
with amendment 25-102 plus the additional requirements of the special 
conditions for an inerting system for the affected Boeing Model 747 
series airplanes. As stated earlier, these special conditions will be 
the baseline for the other airplane models for which the applicant 
plans to seek approval of an FRM. No changes were made as a result of 
this comment.

Special Conditions

I. Definitions
    Comment: The commenter requests the definition for flammable be 
revised to read as follows:

    Flammable. With respect to a fluid or gas, flammable means 
susceptible to igniting readily or to exploding (14 CFR Part 1, 
Definitions). A non-flammable ullage is one where the gas mixture is 
too lean or too rich to burn and/or is inert per the definition of 
inert below. For the purposes of these special conditions, a fuel 
tank is considered flammable when the ullage is not inert and the 
fuel vapor concentration is within the flammable range for the fuel 
type being used. The fuel vapor concentration of the ullage in a 
fuel tank shall be determined based on the average fuel temperature 
within the tank. This vapor concentration shall be assumed to exist 
throughout all bays of the tank. An exception to this shall be 
utilized when one or more major portion of the tank is exposed to 
grossly dissimilar heating conditions. In this situation, the vapor 
concentration of this major portion shall be determined 
independently based upon the fuel temperature of this portion.

    The commenter requests this change because the wording, as proposed 
in the notice, is inconsistent with the modeling methods required in 
appendix 2 of the special conditions. The development of the concept of 
assessing average fleet flammability exposure using a Monte Carlo 
analysis was based on the use of an average bulk fuel temperature of 
the entire center wing fuel tank. This is the parameter that was 
defined in conjunction with the conclusion that achieving a 3 percent 
average fleet flammability exposure criteria would be considered 
equivalent to providing similar characteristics to the type 
certificated model's unheated aluminum wing tanks when the same fuel is 
used in the calculation, as required by Sec.  25.981(c). None of the 
Monte Carlo analytical modeling to date by the FAA, the two ARAC 
studies, or the Boeing Company have been based on individual tank 
compartment fuel temperatures. Each of these analyses has been based on 
the average temperature of the fuel and applying the flammability 
exposure based on that fuel temperature to all bays. The commenter 
references FAA Report DOT/FAA/AR-TN99/65 for supporting test data.
    FAA Reply: We concur, in part, with the commenter. As stated 
earlier, we have modified the definition of flammable to ``With respect 
to a fluid or gas, flammable means susceptible to igniting readily or 
to exploding (14 CFR part 1, Definitions). A non-flammable ullage is 
one where the gas mixture is

[[Page 7819]]

too lean or too rich to burn and/or is inert per the definition of 
inert below.''
    To ensure that flammability of individual bays is accounted for in 
the Monte Carlo analysis, we have added clarification in appendix 2 
that reads:

    For the purposes of these special conditions, a fuel tank is 
considered flammable when the ullage is not inert and the fuel vapor 
concentration is within the flammable range for the fuel type being 
used. The fuel vapor concentration of the ullage in a fuel tank 
shall be determined based on the bulk average fuel temperature 
within the tank. This vapor concentration must be assumed to exist 
throughout all bays of the tank. For those airplanes with fuel tanks 
having different flammability exposure within different compartments 
of the tank, the flammability of the compartments must be analyzed 
individually in the Monte Carlo analysis. The highest flammability 
exposure must be used in the analysis. For example, the center wing 
fuel tank in some designs extends into the wing and has portions of 
the tank that are cooled by outside air, and other portions of the 
tank that are insulated from outside air. Therefore, the fuel 
temperature is different than the portion of the fuel tank in the 
wing.

    Comment: One commenter says use of the term ``employee'' in the 
definition for ``hazardous atmosphere'' is questionable. The commenter 
considers it more appropriate to extend the definition to cover the 
risk to maintenance personnel, passengers, flightcrew, etc.
    FAA Reply: We concur with the commenter and have revised the 
definition of ``hazardous atmosphere'' to address any person(s).
    Comment: A commenter requests clarification of the definition of 
inert (what is the percentage at sea level to meet the 12 percent or 
less oxygen limit at 10,000 feet?). The commenter also asks if the NEA 
supply can keep up with demand through 10,000 feet. The commenter says 
the altitude should be 15,000 feet because TWA 800 exploded at 13,500 
feet. The commenter also says there is conjecture that the oxygen 
concentration in the fuel tank ullage will have to be less than 10 
percent at sea level to keep the oxygen level below 12 percent at 
10,000 feet.
    FAA Reply: We do not concur. The definition of inert is based on 
FAA testing as explained previously. No changes were made as a result 
of these comments.
    Comment: In reference to the definition of a Monte Carlo analysis, 
the commenter notes that the FAA used the ARAC analysis in the model as 
the means of compliance with the special conditions. The commenter says 
this analysis did not include transport effects, which they believe 
should be included, as well as flammability effects on center wing tank 
heated unusable (empty, 0 quantity indication) fuel. They say the fuel 
temperature within a specific compartment of the tank could be within 
the flammable range for the fuel type being used if the tank was empty 
and heat sources were next to the compartment.
    FAA Reply: We do not concur. As explained earlier, we excluded both 
of the phenomena (mass loading and fuel vaporization and condensation) 
that are part of the definition of transport effects, because they were 
not considered by ARAC when they established the flammability 
requirements. If they had included these effects in the wing tank 
flammability exposure calculation, the wing tank flammability exposure 
benchmark value would have been significantly lower, which could result 
in more restrictive requirements for center wing tank flammability 
exposure. No changes were made as a result of these comments.
    Comment: Two commenters request clarification of the definition of 
operational time. One commenter proposes the definition be revised to 
read as follows for consistency with AC 25.981-2 and the Monte Carlo 
analysis: This commenter says the current definition would not result 
in a clearly defined number of flights per day for use in the Monte 
Carlo analysis and would basically define the daily operational time as 
one continuous period of time.

    ``Operational Time. For the purpose of these special conditions, 
the time from the start of preparing the airplane for flight (that 
is, starting and connecting the auxiliary or ground power unit to 
the aircraft electrical system) through the actual flight and 
landing, and through the time to disembark any payload, passengers 
and crew.''

    FAA Reply: We concur in part. Because the definition of operational 
time in these special conditions is not consistent with the definition 
in 14 CFR part 1, Definitions, we have replaced ``operational time'' 
with the term ``flammability exposure evaluation time (FEET).'' We have 
revised the definition to read as follows:

    Flammability Exposure Evaluation Time (FEET). For the purpose of 
these special conditions, the time from the start of preparing the 
airplane for flight, through the flight and landing, until all 
payload is unloaded and all passengers and crew have disembarked. In 
the Monte Carlo program, the flight time is randomly selected from 
the Mission Range Distribution (Table 3), the pre-flight times are 
provided as a function of the flight time, and the post-flight time 
is a constant 30 minutes.

    Comment: This commenter believes additional definitions need to be 
added such as operational time, fleet average, etc., for clarification.
    FAA Reply: We concur in part. The definition of operational time is 
already addressed in Special Condition I. Definitions, and we have 
added additional definitions for clarification as needed.
II. System Performance and Reliability
    Comment: Several commenters request clarification of paragraph II 
(a)(2). One commenter assumes that the FRM can be non-operational for 
1.8 percent of the airplane operational life. This commenter says 
elsewhere in the special conditions more stringent requirements are 
implied (for example ``shortest practical MMEL relief''), which is 
inconsistent. The commenter considers the 1.8 percent requirement to be 
sufficient. Another commenter requests explanation of the percentage 
figures quoted in paragraphs II (a), (b), and (c).
    FAA Reply: The 1.8 percent maximum contribution requirement for an 
inoperative FRM is for an airplane fleet, not an individual airplane. 
The special conditions limit the maximum fleet average flammability 
exposure to 3 percent. The performance or reliability contributions can 
be up to 1.8 percent, as long as the overall fleet average flammability 
exposure does not exceed a total of 3 percent. The contribution for FRM 
performance would be limited to 1.2 percent if the reliability 
contribution were 1.8 percent. The 3 percent warm day requirement is a 
separate performance requirement that must be met for warm day ground, 
takeoff, and climb flight profiles and therefore does not include the 
contribution for reliability of the system. All of these requirements 
establish the minimum safety standards. No changes were made as a 
result of these comments.
    Comment: The commenter refers to the statement in paragraph II (c) 
that ``the applicant must provide data from ground testing and flight 
testing'' to show compliance with paragraphs II (a), (b), and (c)(2). 
The commenter believes that the means of compliance should be left to 
the applicant. The paragraph should therefore read, ``The applicant 
must provide appropriate data * * *''
    Comment: Another commenter also requests a change to paragraph 
II(c). This commenter suggests the following: ``The applicant must 
provide data from analysis and/or testing.'' The commenter says use of 
analysis and/or testing is consistent with normal processes used to 
demonstrate compliance with part 25 requirements.

[[Page 7820]]

    FAA Reply: We do not concur with the commenters. The wording of the 
special condition is consistent with other regulations where test data 
is needed to demonstrate compliance. Analysis alone is not considered 
adequate for demonstrating compliance with the special condition 
requirements because with this new technology there is not a sufficient 
experience base from which to derive a reliable analysis. No changes 
were made as a result of these comments.
    Comment: One commenter requests clarification why paragraph II (c) 
has been included in the requirements listed under paragraphs II 
(c)(1), II (d), and III (a).
    FAA Reply: We infer from the comment that the reference to 
paragraph II (c) should be removed from paragraphs II (c)(1), II (d), 
and III (a) and we concur. We have therefore revised the special 
conditions to change the reference in the noted paragraphs to paragraph 
II (c)(2).
    Comment: The commenter requests that the four elements involved 
with the fleet average flammability exposure, as referenced in 
``Inerting System Indications,'' be included in paragraph II (e).
    FAA Reply: We do not concur. The special conditions do not dictate 
a specific design, but rather state that indication and/or maintenance 
checks will be required to ensure that the performance and reliability 
of the FRM meets the special conditions requirements. No changes were 
made as a result of this comment.
    Comment: The commenter recommends that paragraph II (f) be expanded 
to state that appropriate markings are required for all inerted fuel 
tanks, tanks adjacent to inerted fuel tanks, and all fuel tanks 
communicating with the inerted tanks via plumbing. The plumbing 
includes, but is not limited to, vent system, fuel feed system, refuel 
system, transfer system and cross-feed system plumbing. NEA could enter 
adjacent fuel tanks via structural leaks. It could also enter other 
fuel tanks through plumbing, if valves are operated or fail in the open 
position. The hazardous markings should also be stenciled on the 
external upper and lower surfaces of the inerted tank to ensure 
maintenance personnel are aware of the possible contents of the fuel 
tank.
    FAA Reply: We concur in part. We revised paragraph II (f) to 
clarify that any fuel tank with an FRM must be marked as required, as 
well as any confined spaces or enclosed areas that could contain NEA 
under normal conditions or failure conditions. The special condition 
already requires the applicant to mark access doors and panels to any 
fuel tank that communicates with an inerted tank.
    Comment: Two commenters say that in paragraph II (g) it is not 
clear which ``normal'' operating conditions the FAA is referring to, 
and if this requirement is intended to address any FRM failures, or 
only hazards related to the oxygen-enriched air. Both consider the 
criteria specified in this paragraph to be inadequate. One commenter 
says the FRM installation must be shown to comply with the safety 
requirements of Sec.  25.1309 (demonstrate that an inverse relationship 
exists between the probability of an event, failure condition, and its 
severity). The second commenter requests that paragraph II (g) be 
revised to read: ``Oxygen-enriched air produced by the nitrogen 
generation system must not create a hazard during all FRS operating 
conditions and it must be established that no single failure or 
malfunction or probable combination of failures will jeopardize the 
safe operation of the airplane.''
    Comment: Another commenter requests paragraph II (g) be revised to 
read: ``Oxygen-enriched air produced by the nitrogen generation system 
must not create a hazard during normal operating conditions (refer to 
14 CFR 25.863).'' The commenter requests this change to make OEA 
leakage compliance requirements consistent with those applicable for 
other flammable leakage zone items.
    FAA Reply: We concur, in part, with the commenters. The intent of 
this requirement is to address any hazards associated with both normal 
operating and failure conditions and not just when the FRM is 
operating. This intent was not clear in the original proposal. We have 
revised paragraph II (g) to state that, ``Any FRM failures, or failures 
that could affect the FRM, with potential catastrophic consequences 
must not result from a single failure or a combination of failures not 
shown to be extremely improbable.'' Note that approval of the FRM 
design will require the applicant to evaluate installation of equipment 
in a flammable fluid leakage zone for compliance with Sec.  25.863. 
However, compliance with the existing general requirements of Sec.  
25.901 is required to ensure that no single failure or malfunction or 
probable combination of failures will jeopardize the safe operation of 
the airplane.
III. Maintenance
    Comment: The commenter requests paragraph III (a) be changed to: 
``Maintenance and/or inspection tasks needed to identify items without 
failure indication, so that FRM reliability does not fall below the 
values assumed in the Monte-Carlo analysis, must be identified as 
Airworthiness Limitations.'' The requirement to identify Airworthiness 
Limitations for all maintenance and/or inspection tasks is 
unprecedented in part 25 certification and would impose an unjustified 
burden on operators. The application of this special condition wording 
to other parts of the fuel system would, in essence, require an 
Airworthiness Limitation to inspect the flight deck lights for basic 
indications such as pump low pressure lights and status messages. It is 
the commenter's position that identifying Airworthiness Limitations 
only for items without failure indication will ensure that the desired 
inspections to identify latent failures are accomplished, without an 
impractical burden on the operators.
    FAA Reply: We concur, in part, with the commenter. Paragraph III 
(a) is not intended to apply to all maintenance and/or inspection 
tasks, just those necessary to identify failures related to FRM 
performance and reliability requirements. No changes were made as a 
result of these comments.
    Comment: The commenter requests that paragraph III(c)(1) be changed 
to: ``Develop and introduce an event monitoring and reporting system 
acceptable to the primary certification authority.'' The commenter 
requests this change because the proposed requirement to track 
inoperative time would result in the introduction of new recordkeeping 
processes, which, in turn, will result in a significant increase in the 
maintenance and operational burden on the operators. The commenter 
accepts that the FRM system reliability should be initially monitored, 
but the requirement should allow the flexibility for existing operator 
and reliability reporting systems to be used to evaluate actual in-
service system reliability, at practical costs.
    FAA Reply: We do not concur. We believe the applicant will be able 
to gather the required data from operators using existing reporting 
systems that are currently in use for airplane maintenance, 
reliability, and warranty claims. We anticipate the operators would 
provide this information to the applicant through existing business 
arrangements. No changes were made as a result of these comments.
    Comment: One commenter believes initiation of component and/or 
system modification should also be included in paragraph III (c)(4) for 
correcting failures of the FRM that increase the fleet flammability 
exposure. Another commenter says paragraph III (c)(4) is not clear as 
to whether this statement

[[Page 7821]]

refers to the 3 percent flammability requirement of paragraph II (a) or 
II (b), or both. This commenter believes paragraph III (c)(4) should 
specifically address the requirements of both paragraphs II (a) and II 
(b) of the special conditions.
    FAA Reply: We concur with the commenters that paragraph III (c)(4) 
needs clarification. We have revised this paragraph to read: ``Develop 
service instructions or revise the applicable airplane manual, per a 
schedule agreed to by the FAA, to correct any failures of the FRM that 
occur in service that could increase the fleet average or warm day 
flammability exposure of the tank to more than the exposure 
requirements of paragraphs II (a) and II (b) of these special 
conditions.''
    Comment: The commenter requests that an additional requirement be 
added that would instruct an applicant to provide training material to 
the industry to incorporate any new design system. This would include 
any specific dangers and safety factors. The amendment of all technical 
documentation, including Airplane Maintenance Manual (AMM), Airplane 
Flight Manual (AFM), etc., is not enough.
    FAA Reply: We do not concur with the commenter. The applicant must 
provide service bulletins that will instruct the operators how to 
properly install an FRM, which should include any specific dangers or 
safety factors that need to be considered during installation. The 
applicant is also responsible for providing any materials necessary to 
ensure an operator knows how to properly operate and maintain the 
system. Training is outside the scope of these special conditions. No 
changes were made as a result of this comment.

Appendix 1: Monte Carlo Analysis

    Comment: The commenter requests the following note be added to 
paragraph (b)(3): ``Note: localized concentrations above the inert 
level are allowed provided the volume of the non-inert region would not 
produce a hazardous condition.'' The commenter says the fresh air drawn 
into the fuel tank through the vent during descent will not be 
flammable and will not cause the tank to become flammable during 
descent. The commenter believes that counting these non-hazardous 
periods as ``flammable'' would increase the system size, weight, and 
associated costs with no benefit.
    FAA Reply: We agree that a note paragraph would be appropriate and 
have added the following to paragraph (b)(3): ``Note: localized 
concentrations above the inert level as a result of fresh air that is 
drawn into the fuel tank through vents during descent would not be 
considered as flammable.''
    Comment: The commenter requests the following change to paragraph 
(b)(5): ``Proposed MMEL/MEL dispatch periods including action to be 
taken when dispatching with the FRM inoperative.'' The commenter says 
the MMEL process is outside the scope of the special conditions. The 
specific MMEL time should be based on fleet data for similar systems, 
not a prescriptive mandate of 60 hours. The actual inoperative MMEL 
interval and corresponding fleet exposure used in the Monte Carlo 
analysis is one of a number of items whose inoperative interval would 
be substantiated as part of achieving part 25 certification. During any 
part 25 certification project, providing acceptable substantiating data 
to the FAA for assumptions and analytical processes is the 
responsibility of the applicant.
    FAA Reply: The establishment of an MMEL dispatch interval will be 
achieved through the certification process, whereby the Flight 
Operations Evaluation Board (FOEB) will review the applicable data 
submitted by the applicant to determine if the proposed dispatch 
interval is appropriate. However, the special conditions include the 
requirement in appendix 1, paragraph (b)(5), to allow the applicant to 
use an inoperative FRM interval that is shorter than the maximum 
proposed interval of ten days, if they can substantiate that the 3 
percent flammability requirement can be met when operating with an 
inoperative FRM. Otherwise, 60 flight hours must be used in the 
analysis for a proposed 10-day MMEL dispatch interval. No changes were 
made as a result of these comments.
    Comment: The commenter contends that in paragraph (b)(5) it should 
be noted that the assumed 60 flight hours for a 10-day MMEL is the 
``average'' MMEL/MEL dispatch inoperative period.
    FAA Reply: We recognize that not all MMEL inoperative periods will 
typically occupy the full allowed MMEL dispatch interval. To account 
for this, the special conditions require an average 60 flight hours to 
be used in the Monte Carlo analysis for a 10-day MMEL dispatch 
interval. This is based on using an average airplane utilization of 12 
hours per day, and an average of one-half the proposed 10-day MMEL 
dispatch interval. No changes were made as a result of this comment.

Appendix 2: Atmosphere

    Comment: The commenter says that oxygen monitoring would eliminate 
the need to compute the transitional temperature, as required in this 
section of appendix 2. This is because the oxygen monitoring system 
measures the temperature in the tanks and uses that temperature in the 
calculations to determine the oxygen percentage present.
    FAA Reply: From the comment, we infer that the commenter is 
questioning why a temperature needs to be calculated for the Monte 
Carlo analysis when an oxygen sensor can be used to measure temperature 
in the fuel tank. Modeling the atmosphere during climb and descent 
using the tables in appendix 2 is required to determine the 
flammability exposure for use in the Monte Carlo analysis. It is not 
related to possible design features such as an oxygen sensor. No 
changes were made as a result of this comment.
    Comment: The commenter would like to know who would make the 
decision regarding the use of lower flash point fuels for more than 
1percent of the fleet operating time. The commenter asks how this 
determination will be made to apply to a particular airplane flown with 
a particular defined flight profile. Another commenter believes there 
should be allowance for factoring in a higher flash point for fuels if 
used for more than 1 percent of the fleet operating time.
    Comment: A third commenter requests that the 3rd and 4th sentences 
in paragraph three of the ``Atmosphere'' discussion be changed to:

    Table 2 is based on typical use of Jet A type fuel, with limited 
TS-1 use. If an airplane fleet is expected to operate with low flash 
point fuels (such as JP-4) more than 1 percent of its operating 
time, or intermediate flash point fuels (such as TS-1) more than 10 
percent of the fleet operating time, then the Monte Carlo analysis 
must include fuel property variation acceptable to the FAA for these 
approved fuels.

    The commenter believes this change clarifies that some TS-1 fuel is 
already included in the Table 2 distribution, and adds a separate usage 
limit for low and intermediate flash point fuel that would require 
development of new worldwide fuel type studies only if exceeded. 
Currently, there are no data available to use for a statistical 
distribution of non Jet-A type fuels and it is unreasonable to expect 
an applicant to provide a Monte Carlo analysis incorporating a 
flammability exposure dataset for these other fuels where the 
appropriate data is not available. The impact on the flammability 
analysis of

[[Page 7822]]

up to 10 percent use of intermediate flash point fuels would be small; 
therefore, the study is not justified unless it is expected that the 
use of these fuels would exceed 10 percent.
    FAA Reply: We agree, in part, with the commenters. The fuel 
properties tables in appendix 2 of the special conditions include a 
distribution of flash points reflecting an FAA survey of jet fuels used 
in both U.S. domestic and international routes. The tables therefore 
include an allowance for use of lower flash points fuels. The intent of 
the Monte Carlo analysis method is to provide a standardized analysis 
method to compare the flammability of the fuel tank under evaluation to 
the established flammability limits. The flammability limits were 
established based on a Monte Carlo analysis using the flash point table 
in these special conditions. To simplify the standardized analysis, we 
have deleted the need to consider other fuel flash point distributions 
from these special conditions.

Appendix 2: Oxygen Evolution

    Comment: The commenter asks, if 12 percent or less oxygen 
percentage is tolerable at 10,000 feet (as opposed to 20.9 at sea level 
before NEA is available to the fuel tank), what oxygen concentration is 
needed on the ground at departure if the FRM is not fully effective 
immediately after engine start? Can the available NEA high flow rate 
keep up with the possible out gassing of the 30 percent oxygen level in 
the fuel in order to be at an oxygen level of 12 percent or less at 
10,000 feet?
    FAA Reply: The flammability requirements in the special conditions 
will limit the maximum oxygen concentration. We expect that if the FRM 
were not designed so that the oxygen concentration of the center wing 
fuel tank ullage is below 12 percent at sea level, it would not meet 
these requirements. It is also not possible to meet the specific risk 
requirements in the special conditions for warm day operations if the 
FRM does not reduce the oxygen concentration level below 12 percent 
during ground operations. The affects of oxygen evolution during climb 
must be accounted for in the analysis required by these special 
conditions. These special conditions do not preclude exceeding the 12 
percent oxygen concentrations during transient conditions. For example, 
the tank may no longer be inert during a high descent rate or during a 
rapid climb where the tank could be above the 12 percent oxygen level 
for short periods of time. As previously discussed, we do not believe 
it is practical to require an FRM that would inert the fuel tank during 
all operational conditions within the airplane operating envelope. No 
changes were made as a result of these comments.
    Comment: The commenter says the last sentence of this discussion 
should read, ``The applicant must provide the assumptions relating to 
air evolution rate'' because provision of substantiated data would not 
be possible due to the uncertain manner in which air evolves from the 
fuel during climb.
    FAA Reply: We agree with the commenter that air evolution rates are 
uncertain and can vary from flight to flight depending on the fuel load 
and the conditions under which the fuel was loaded. However, we do not 
agree that it will not be possible to provide data to substantiate the 
air evolution rate for the center wing fuel tank. The FAA has not seen 
large transients related to air evolution during airplane model testing 
(FAA Report No. DOT/FAA/AR-01/63, ``Ground and Flight Testing of a 
Boeing 737 Center Wing Fuel Tank Inerted With Nitrogen-Enriched Air.'' 
We would expect air evolution rates determined by flight testing with 
typical fuel loading to be representative of those anticipated in 
service, so this data should be sufficient to address the effects of 
air evolution on oxygen concentrations. No changes were made as a 
result of this comment.

Other

    In addition to the changes to the special conditions in response to 
comments, we made some changes to provide additional clarification in 
certain areas. Because those changes do not change the intent of the 
special conditions, they are not included in the discussion of 
comments.

Applicability

    As discussed above, these special conditions are applicable to the 
Boeing Model 747-100/200B/200F/200C/SR/SP/100B/300/100B SUD/400/400D/
400F series airplanes. Should the type certificate be amended later to 
include any other model that incorporates the same or similar novel or 
unusual design feature, or should any other model already included on 
the same type certificate be modified to incorporate the same or 
similar novel or unusual design feature, the special conditions would 
also apply to the other model under the provisions of Sec.  21.101.

Conclusion

    This action affects only certain novel or unusual design features 
on Boeing Model 747-100/200B/200F/200C/SR/SP/100B/300/100B SUD/400/
400D/400F series airplanes. It is not a rule of general applicability 
and affects only the applicant who applied to the FAA for approval of 
these features on the airplane.

List of Subjects in 14 CFR Part 25

    Aircraft, Aviation safety, Reporting and recordkeeping 
requirements.


0
The authority citation for these special conditions is as follows:

    Authority: 49 U.S.C. 106(g), 40113, 44701, 44702, 44704.

The Special Conditions

0
Accordingly, pursuant to the authority delegated to me by the 
Administrator, the following special conditions are issued as part of 
the type certification basis for Boeing Model 747-100/200B/200F/200C/ 
SR/SP/100B/300/100B SUD/400/400D/400F series airplanes, modified by 
Boeing Commercial Airplanes to include a flammability reduction means 
(FRM) that uses a nitrogen generation system to inert the center wing 
tank with nitrogen-enriched air (NEA).
    Compliance with these special conditions does not relieve the 
applicant from compliance with the existing certification requirements.
    I. Definitions. (a) Bulk Average Fuel Temperature. The average fuel 
temperature within the fuel tank, or different sections of the tank if 
the tank is subdivided by baffles or compartments.
    (b) Flammability Exposure Evaluation Time (FEET). For the purpose 
of these special conditions, the time from the start of preparing the 
airplane for flight, through the flight and landing, until all payload 
is unloaded and all passengers and crew have disembarked. In the Monte 
Carlo program, the flight time is randomly selected from the Mission 
Range Distribution (Table 3), the pre-flight times are provided as a 
function of the flight time, and the post-flight time is a constant 30 
minutes.
    (c) Flammable. With respect to a fluid or gas, flammable means 
susceptible to igniting readily or to exploding (14 CFR part 1, 
Definitions). A non-flammable ullage is one where the gas mixture is 
too lean or too rich to burn and/or is inert per the definition below.
    (d) Flash Point. The flash point of a flammable fluid is the lowest 
temperature at which the application of a flame to a heated sample 
causes the vapor to ignite momentarily, or ``flash.'' The test for jet 
fuel is defined in ASTM Specification D56, ``Standard Test Method for 
Flash Point by Tag Close Cup Tester.''
    (e) Hazardous Atmosphere. An atmosphere that may expose any 
person(s) to the risk of death,

[[Page 7823]]

incapacitation, impairment of ability to self-rescue (escape unaided 
from a space), injury, or acute illness.
    (f) Inert. For the purpose of these special conditions, the tank is 
considered inert when the bulk average oxygen concentration within each 
compartment of the tank is 12 percent or less at sea level up to 10,000 
feet, then linearly increasing from 12 percent at 10,000 feet to 14.5 
percent at 40,000 feet and extrapolated linearly above that altitude.
    (g) Inerting. A process where a noncombustible gas is introduced 
into the ullage of a fuel tank to displace sufficient oxygen so that 
the ullage becomes inert.
    (h) Monte Carlo Analysis. An analytical tool that provides a means 
to assess the degree of fleet average and warm day flammability 
exposure time for a fuel tank. See appendices 1 and 2 of these special 
conditions for specific requirements for conducting the Monte Carlo 
analysis.
    (i) Transport Effects. Transport effects are the effects on fuel 
vapor concentration caused by low fuel conditions, fuel condensation, 
and vaporization.
    (j) Ullage, or Ullage Space. The volume within the fuel tank not 
occupied by liquid fuel at the time interval under evaluation.
    II. System Performance and Reliability. The FRM, for the airplane 
model under evaluation, must comply with the following performance and 
reliability requirements:
    (a) The applicant must submit a Monte Carlo analysis, as defined in 
appendices 1 and 2 of these special conditions, that--
    (1) Demonstrates that the overall fleet average flammability 
exposure of each fuel tank with an FRM installed is equal to or less 
than 3 percent of the FEET; and
    (2) Demonstrates that neither the performance (when the FRM is 
operational) nor reliability (including all periods when the FRM is 
inoperative) contributions to the overall fleet average flammability 
exposure of a tank with an FRM installed is more than 1.8 percent (this 
will establish appropriate maintenance inspection procedures and 
intervals as required in paragraph III (a) of these special 
conditions).
    (3) Identifies critical features of the fuel tank system to prevent 
an auxiliary fuel tank installation from increasing the flammability 
exposure of the center wing tank above that permitted under paragraphs 
II (a)(1) and (2) of these special conditions and to prevent 
degradation of the performance and reliability of the FRM.
    (b) The applicant must submit a Monte Carlo analysis that 
demonstrates that the FRM, when functional, reduces the overall 
flammability exposure of each fuel tank with an FRM installed for warm 
day ground, takeoff, and climb phases to a level equal to or less than 
3 percent of the FEET in each of these phases for the following 
conditions--
    (1) The analysis must use the subset of 80 [deg]F and warmer days 
from the Monte Carlo analyses done for overall performance; and
    (2) The flammability exposure must be calculated by comparing the 
time during ground, takeoff, and climb phases for which the tank was 
flammable and not inert, with the total time for the ground, takeoff, 
and climb phases.
    (c) The applicant must provide data from ground testing and flight 
testing that--
    (1) Validate the inputs to the Monte Carlo analysis needed to show 
compliance with (or meet the requirements of) paragraphs II (a), (b), 
and (c)(2) of these special conditions; and
    (2) Substantiate that the NEA distribution is effective at inerting 
all portions of the tank where the inerting system is needed to show 
compliance with these paragraphs.
    (d) The applicant must validate that the FRM meets the requirements 
of paragraphs II (a), (b), and (c)(2) of these special conditions, with 
any combination of engine model, engine thrust rating, fuel type, and 
relevant pneumatic system configuration approved for the airplane.
    (e) Sufficient accessibility for maintenance personnel, or the 
flightcrew, must be provided to FRM status indications necessary to 
meet the reliability requirements of paragraph II (a) of these special 
conditions.
    (f) The access doors and panels to the fuel tanks with an FRM 
(including any tanks that communicate with an inerted tank via a vent 
system), and to any other confined spaces or enclosed areas that could 
contain NEA under normal conditions or failure conditions, must be 
permanently stenciled, marked, or placarded as appropriate to warn 
maintenance crews of the possible presence of a potentially hazardous 
atmosphere. The proposal for markings does not alter the existing 
requirements that must be addressed when entering airplane fuel tanks.
    (g) Any FRM failures, or failures that could affect the FRM, with 
potential catastrophic consequences must not result from a single 
failure or a combination of failures not shown to be extremely 
improbable.
    III. Maintenance. (a) Airworthiness Limitations must be identified 
for all critical features identified under paragraph II (a)(3) and for 
all maintenance and/or inspection tasks required to identify failures 
of components within the FRM that are needed to meet paragraphs II (a), 
(b), and (c)(2) of these special conditions.
    (b) The applicant must provide the maintenance procedures that will 
be necessary and present a design review that identifies any hazardous 
aspects to be considered during maintenance of the FRM that will be 
included in the instructions for continued airworthiness (ICA) or 
appropriate maintenance documents.
    (c) To ensure that the effects of component failures on FRM 
reliability are adequately assessed on an on-going basis, the applicant 
must--
    (1) Demonstrate effective means to ensure collection of FRM 
reliability data. The means must provide data affecting FRM 
availablity, such as component failures, and the FRM inoperative 
intervals due to dispatch under the MMEL;
    (2) Provide a report to the FAA on a quarterly basis for the first 
five years after service introduction. After that period, continued 
quarterly reporting may be replaced with other reliability tracking 
methods found acceptable to the FAA or eliminated if it is established 
that the reliability of the FRM meets, and will continue to meet, the 
exposure requirements of paragraphs II (a) and (b) of these special 
conditions;
    (3) Provide a report to the validating authorities for a period of 
at least two years following introduction to service; and
    (4) Develop service instructions or revise the applicable airplane 
manual, per a schedule agreed on by the FAA, to correct any failures of 
the FRM that occur in service that could increase the fleet average or 
warm day flammability exposure of the tank to more than the exposure 
requirements of paragraphs II (a) and (b) of these special conditions.

Appendix 1

Monte Carlo Analysis

    (a) A Monte Carlo analysis must be conducted for the fuel tank 
under evaluation to determine fleet average and warm day 
flammability exposure for the airplane and fuel type under 
evaluation. The analysis must include the parameters defined in 
appendices 1 and 2 of these special conditions. The airplane 
specific parameters and assumptions used in the Monte Carlo analysis 
must include:
    (1) FRM Performance--as defined by system performance.

[[Page 7824]]

    (2) Cruise Altitude--as defined by airplane performance.
    (3) Cruise Ambient Temperature--as defined in appendix 2 of 
these special conditions.
    (4) Overnight Temperature Drop--as defined in appendix 2 of 
these special conditions.
    (5) Fuel Flash Point and Upper and Lower Flammability Limits--as 
defined in appendix 2 of these special conditions.
    (6) Fuel Burn--as defined by airplane performance.
    (7) Fuel Quantity--as defined by airplane performance.
    (8) Fuel Transfer--as defined by airplane performance.
    (9) Fueling Duration--as defined by airplane performance.
    (10) Ground Temperature--as defined in appendix 2 of these 
special conditions.
    (11) Mach Number--as defined by airplane performance.
    (12) Mission Distribution--the applicant must use the mission 
distribution defined in appendix 2 of these special conditions or 
may request FAA approval of alternate data from the service history 
of the Model 747.
    (13) Oxygen Evolution--as defined by airplane performance and as 
discussed in appendix 2 of these special conditions.
    (14) Maximum Airplane Range--as defined by airplane performance.
    (15) Tank Thermal Characteristics--as defined by airplane 
performance.
    (16) Descent Profile Distribution--the applicant must use either 
a fixed 2500 feet per minute descent rate or provide alternate data 
from the service history of the Model 747.
    (b) The assumptions for the analysis must include--
    (1) FRM performance throughout the flammability exposure 
evaluation time;
    (2) Vent losses due to crosswind effects and airplane 
performance;
    (3) Any time periods when the system is operating properly but 
fails to inert the tank;

Note: localized concentrations above the inert level as a result of 
fresh air that is drawn into the fuel tank through vents during 
descent would not be considered as flammable.

    (4) Expected system reliability;
    (5) The MMEL/MEL dispatch inoperative period assumed in the 
reliability analysis (60 flight hours must be used for a 10-day MMEL 
dispatch limit unless an alternative period has been approved by the 
FAA), including action to be taken when dispatching with the FRM 
inoperative (Note: The actual MMEL dispatch inoperative period data 
must be included in the engineering reporting requirement of 
paragraph III(c)(1) of these special conditions.);
    (6) Possible time periods of system inoperability due to latent 
or known failures, including airplane system shut-downs and failures 
that could cause the FRM to shut down or become inoperative; and
    (7) Affects of failures of the FRM that could increase the 
flammability of the fuel tank.
    (c) The Monte Carlo analysis, including a description of any 
variation assumed in the parameters (as identified under paragraph 
(a) of this appendix) that affect flammability exposure, and 
substantiating data must be submitted to the FAA for approval.

Appendix 2

    I. Monte Carlo Model. (a) The FAA has developed a Monte Carlo 
model that can be used to develop a specific analysis model for the 
Boeing 747 to calculate fleet average and warm day flammability 
exposure for a fuel tank in an airplane. Use of the program requires 
the user to enter the airplane performance data specific to the 
airplane model being evaluated, such as maximum range, cruise mach 
number, typical step climb altitudes, tank thermal characteristics 
specified as exponential heating/cooling time constants, and 
equilibrium temperatures for various fuel tank conditions. The 
general methodology for conducting a Monte Carlo model is described 
in AC 25.981-2.
    (b) The FAA model, or one with modifications approved by the 
FAA, must be used as the means of compliance with these special 
conditions. The accepted model can be downloaded from the Web site 
http://qps.airweb.faa.gov/sfar88flamex. On this Web site, the model 
is located under the page ``Flam Ex Resources,'' and is titled 
``Monte Carlo Model Version 6a.'' The ``6a'' represents Version 6A. 
Only version 6A or later of this model can be used. The following 
procedures, input variables, and data tables must be used in the 
analysis if the applicant develops a unique model to determine fleet 
average flammability exposure for a specific airplane type.
    II. Monte Carlo Variables and Data Tables. (a) Fleet average 
flammability exposure is the percent of the mission time the fuel 
tank ullage is flammable for a fleet of an airplane type operating 
over the range of actual or expected missions and in a world-wide 
range of environmental conditions and fuel properties. Variables 
used to calculate fleet average flammability exposure must include 
atmosphere, mission length (as defined in Special Condition I. 
Definitions, as FEET), fuel flash point, thermal characteristics of 
the fuel tank, overnight temperature drop, and oxygen evolution from 
the fuel into the ullage. Transport effects, including mass loading, 
flammability lag time, and condensation of vapors due to cold 
surfaces, are not to be allowed as parameters in the analysis.
    (b) For the purposes of these special conditions, a fuel tank is 
considered flammable when the ullage is not inert and the fuel vapor 
concentration is within the flammable range for the fuel type being 
used. The fuel vapor concentration of the ullage in a fuel tank must 
be determined based on the bulk average fuel temperature within the 
tank. This vapor concentration must be assumed to exist throughout 
all bays of the tank. For those airplanes with fuel tanks having 
different flammability exposure within different compartments of the 
tank, where mixing of the vapor or NEA does not occur, the Monte 
Carlo analysis must be conducted for the compartment of the tank 
with the highest flammability. The compartment with the highest 
flammability exposure for each flight phase must be used in the 
analysis to establish the fleet average flammability exposure. For 
example, the center wing fuel tank in some designs extends into the 
wing and has compartments of the tank that are cooled by outside 
air, and other compartments of the tank that are insulated from 
outside air. Therefore, the fuel temperature and flammability is 
significantly different between these compartments of the fuel tank.
    (c) Atmosphere. (1) To predict flammability exposure during a 
given flight, the variation of ground ambient temperatures, cruise 
ambient temperatures, and a method to compute the transition from 
ground to cruise and back again must be used. The variation of the 
ground and cruise ambient temperatures and the flash point of the 
fuel is defined by a Gaussian curve, given by the 50 percent value 
and a  1 standard deviation value.
    (2) The ground and cruise temperatures are linked by a set of 
assumptions on the atmosphere. The temperature varies with altitude 
following the International Standard Atmosphere (ISA) rate of change 
from the ground temperature until the cruise temperature for the 
flight is reached. Above this altitude, the ambient temperature is 
fixed at the cruise ambient temperature. This results in a variation 
in the upper atmospheric (tropopause) temperature. For cold days, an 
inversion is applied up to 10,000 feet, and then the ISA rate of 
change is used.
    (3) The analysis must include a minimum number of flights, and 
for each flight a separate random number must be generated for each 
of the three parameters (that is, ground ambient temperature, cruise 
ambient temperature, and fuel flash point) using the Gaussian 
distribution defined in Table 1. The applicant can verify the output 
values from the Gaussian distribution using Table 2.
    (d) Fuel Properties. (1) Flash point variation. The variation of 
the flash point of the fuel is defined by a Gaussian curve, given by 
the 50 percent value and a  1-standard deviation value.
    (2) Upper and Lower Flammability Limits. The flammability 
envelope of the fuel that must be used for the flammability exposure 
analysis is a function of the flash point of the fuel selected by 
the Monte Carlo for a given flight. The flammability envelope for 
the fuel is defined by the upper flammability limit (UFL) and lower 
flammability limit (LFL) as follows:
    (i) LFL at sea level = flash point temperature of the fuel at 
sea level minus 10 degrees F. LFL decreases from sea level value 
with increasing altitude at a rate of 1 degree F per 808 ft.
    (ii) UFL at sea level = flash point temperature of the fuel at 
sea level plus 63.5 degrees F. UFL decreases from the sea level 
value with increasing altitude at a rate of 1 degree F per 512 ft.
    Note: Table 1 includes the Gaussian distribution for fuel flash 
point. Table 2 also includes information to verify output values for 
fuel properties. Table 2 is based on typical use of Jet A type fuel, 
with limited TS-1 type fuel use.

[[Page 7825]]



Table 1.--Gaussian Distribution for Ground Ambient Temperature, Cruise Ambient Temperature, and Fuel Flash Point
----------------------------------------------------------------------------------------------------------------
                                              Temperature in Deg F
-----------------------------------------------------------------------------------------------------------------
                                                             Ground ambient    Cruise ambient
                         Parameter                             temperature       temperature    Flash point (FP)
----------------------------------------------------------------------------------------------------------------
Mean Temp.................................................             59.95               -70               120
Neg 1 std dev.............................................             20.14                 8                 8
Pos 1 std dev.............................................             17.28                 8                 8
----------------------------------------------------------------------------------------------------------------


                                                            Table 2.--Verification of Table 1
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                             Ground          Cruise                           Ground          Cruise
  % Probability of temps & flash point being below the       ambient         ambient       Flash point        ambient         ambient       Flash point
                     listed values                         temperature     temperature        Deg F         temperature     temperature     (FP)  Deg C
                                                              Deg F           Deg F                            Deg C           Deg C
--------------------------------------------------------------------------------------------------------------------------------------------------------
1......................................................            13.1           -88.6           101.4            -10.5           -67.0            38.5
5......................................................            26.8           -83.2           106.8             -2.9           -64.0            41.6
10.....................................................            34.1           -80.3           109.7              1.2           -62.4            43.2
15.....................................................            39.1           -78.3           111.7              3.9           -61.3            44.3
20.....................................................            43.0           -76.7           113.3              6.1           -60.4            45.1
25.....................................................            46.4           -75.4           114.6              8.0           -59.7            45.9
30.....................................................            49.4           -74.2           115.8              9.7           -59.0            46.6
35.....................................................            52.2           -73.1           116.9             11.2           -58.4            47.2
40.....................................................            54.8           -72.0           118.0             12.7           -57.8            47.8
45.....................................................            57.4           -71.0           119.0             14.1           -57.2            48.3
50.....................................................            59.9           -70.0           120.0             15.5           -56.7            48.9
55.....................................................            62.1           -69.0           121.0             16.7           -56.1            49.4
60.....................................................            64.3           -68.0           122.0             18.0           -55.5            50.0
65.....................................................            66.6           -66.9           123.1             19.2           -55.0            50.6
70.....................................................            69.0           -65.8           124.2             20.6           -54.3            51.2
75.....................................................            71.6           -64.6           125.4             22.0           -53.7            51.9
80.....................................................            74.5           -63.3           126.7             23.6           -52.9            52.6
85.....................................................            77.9           -61.7           128.3             25.5           -52.1            53.5
90.....................................................            82.1           -59.7           130.3             27.8           -51.0            54.6
95.....................................................            88.4           -56.8           133.2             31.3           -49.4            56.2
99.....................................................           100.1           -51.4           138.6             37.9           -46.3            59.2
--------------------------------------------------------------------------------------------------------------------------------------------------------

    (e) Flight Mission Distribution. (1) The mission length for each 
flight is determined from an equation that takes the maximum mission 
length for the airplane and randomly selects multiple flight lengths 
based on typical airline use.
    (2) The mission length selected for a given flight is used by 
the Monte Carlo model to select a 30-, 60-, or 90-minute time on the 
ground prior to takeoff, and the type of flight profile to be 
followed. Table 3 must be used to define the mission distribution. A 
linear interpolation between the values in the table must be 
assumed.

                                    Table 3.--Mission Length Distribution Airplane Maximum Range--Nautical Miles (NM)
--------------------------------------------------------------------------------------------------------------------------------------------------------
                           Flight length (NM)                                                       Airplane maximum range (NM)
--------------------------------------------------------------------------------------------------------------------------------------------------------
                   From                                  To                1000    2000    3000    4000    5000    6000    7000    8000    9000    10000
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                                                Distribution of mission lengths (%)
                                                                         ---------
0.........................................  200.........................    11.7     7.5     6.2     5.5     4.7     4.0     3.4     3.0     2.6     2.3
200.......................................  400.........................    27.3    19.9    17.0    15.2    13.2    11.4     9.7     8.5     7.5     6.7
400.......................................  600.........................    46.3    40.0    35.7    32.6    28.5    24.9    21.2    18.7    16.4    14.8
600.......................................  800.........................    10.3    11.6    11.0    10.2     9.1     8.0     6.9     6.1     5.4     4.8
800.......................................  1000........................     4.4     8.5     8.6     8.2     7.4     6.6     5.7     5.0     4.5     4.0
1000......................................  1200........................     0.0     4.8     5.3     5.3     4.8     4.3     3.8     3.3     3.0     2.7
1200......................................  1400........................     0.0     3.6     4.4     4.5     4.2     3.8     3.3     3.0     2.7     2.4
1400......................................  1600........................     0.0     2.2     3.3     3.5     3.3     3.1     2.7     2.4     2.2     2.0
1600......................................  1800........................     0.0     1.2     2.3     2.6     2.5     2.4     2.1     1.9     1.7     1.6
1800......................................  2000........................     0.0     0.7     2.2     2.6     2.6     2.5     2.2     2.0     1.8     1.7
2000......................................  2200........................     0.0     0.0     1.6     2.1     2.2     2.1     1.9     1.7     1.6     1.4
2200......................................  2400........................     0.0     0.0     1.1     1.6     1.7     1.7     1.6     1.4     1.3     1.2
2400......................................  2600........................     0.0     0.0     0.7     1.2     1.4     1.4     1.3     1.2     1.1     1.0
2600......................................  2800........................     0.0     0.0     0.4     0.9     1.0     1.1     1.0     0.9     0.9     0.8
2800......................................  3000........................     0.0     0.0     0.2     0.6     0.7     0.8     0.7     0.7     0.6     0.6
3000......................................  3200........................     0.0     0.0     0.0     0.6     0.8     0.8     0.8     0.8     0.7     0.7
3200......................................  3400........................     0.0     0.0     0.0     0.7     1.1     1.2     1.2     1.1     1.1     1.0
3400......................................  3600........................     0.0     0.0     0.0     0.7     1.3     1.6     1.6     1.5     1.5     1.4
3600......................................  3800........................     0.0     0.0     0.0     0.9     2.2     2.7     2.8     2.7     2.6     2.5

[[Page 7826]]

 
3800......................................  4000........................     0.0     0.0     0.0     0.5     2.0     2.6     2.8     2.8     2.7     2.6
4000......................................  4200........................     0.0     0.0     0.0     0.0     2.1     3.0     3.2     3.3     3.2     3.1
4200......................................  4400........................     0.0     0.0     0.0     0.0     1.4     2.2     2.5     2.6     2.6     2.5
4400......................................  4600........................     0.0     0.0     0.0     0.0     1.0     2.0     2.3     2.5     2.5     2.4
4600......................................  4800........................     0.0     0.0     0.0     0.0     0.6     1.5     1.8     2.0     2.0     2.0
4800......................................  5000........................     0.0     0.0     0.0     0.0     0.2     1.0     1.4     1.5     1.6     1.5
5000......................................  5200........................     0.0     0.0     0.0     0.0     0.0     0.8     1.1     1.3     1.3     1.3
5200......................................  5400........................     0.0     0.0     0.0     0.0     0.0     0.8     1.2     1.5     1.6     1.6
5400......................................  5600........................     0.0     0.0     0.0     0.0     0.0     0.9     1.7     2.1     2.2     2.3
5600......................................  5800........................     0.0     0.0     0.0     0.0     0.0     0.6     1.6     2.2     2.4     2.5
5800......................................  6000........................     0.0     0.0     0.0     0.0     0.0     0.2     1.8     2.4     2.8     2.9
6000......................................  6200........................     0.0     0.0     0.0     0.0     0.0     0.0     1.7     2.6     3.1     3.3
6200......................................  6400........................     0.0     0.0     0.0     0.0     0.0     0.0     1.4     2.4     2.9     3.1
6400......................................  6600........................     0.0     0.0     0.0     0.0     0.0     0.0     0.9     1.8     2.2     2.5
6600......................................  6800........................     0.0     0.0     0.0     0.0     0.0     0.0     0.5     1.2     1.6     1.9
6800......................................  7000........................     0.0     0.0     0.0     0.0     0.0     0.0     0.2     0.8     1.1     1.3
7000......................................  7200........................     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.4     0.7     0.8
7200......................................  7400........................     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.3     0.5     0.7
7400......................................  7600........................     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.2     0.5     0.6
7600......................................  7800........................     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.1     0.5     0.7
7800......................................  8000........................     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.1     0.6     0.8
8000......................................  8200........................     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.5     0.8
8200......................................  8400........................     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.5     1.0
8400......................................  8600........................     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.6     1.3
8600......................................  8800........................     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.4     1.1
8800......................................  9000........................     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.2     0.8
9000......................................  9200........................     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.5
9200......................................  9400........................     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.2
9400......................................  9600........................     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.1
9600......................................  9800........................     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.1
9800......................................  10000.......................     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.1
--------------------------------------------------------------------------------------------------------------------------------------------------------

    (f) Fuel Tank Thermal Characteristics. (1) The applicant must 
account for the thermal conditions of the fuel tank both on the 
ground and in flight. The Monte Carlo model, available on the 
website listed above, defines the ground condition using an 
equilibrium delta temperature (relative to the ambient temperature) 
the tank will reach given a long enough time, with any heat inputs 
from airplane sources. Values are also input to define two 
exponential time constants (one for a near empty tank and one for a 
near full tank) for the ground condition. These time constants 
define the time for the fuel in the fuel tank to heat or cool in 
response to heat input. The fuel is assumed to heat or cool 
according to a normal exponential transition, governed by the 
temperature difference between the current temperature and the 
equilibrium temperature, given by ambient temperature plus delta 
temperature. Input values for this data can be obtained from 
validated thermal models of the tank based on ground and flight test 
data. The inputs for the inflight condition are similar but are used 
for inflight analysis.
    (2) Fuel management techniques are unique to each manufacturer's 
design. Variations in fuel quantity within the tank for given points 
in the flight, including fuel transfer for any purpose, must be 
accounted for in the model. The model uses a ``tank full'' time, 
specified in minutes, that defines the time before touchdown when 
the fuel tank is still full. For a center wing tank used first, this 
number would be the maximum flight time, and the tank would start to 
empty at takeoff. For a main tank used last, the tank will remain 
full for a shorter time before touchdown and would be ``empty'' at 
touchdown (that is, tank empty at 0 minutes before touchdown). For a 
main tank with reserves, the term empty means at reserve level 
rather than totally empty. The thermal data for tank empty would 
also be for reserve level.
    (3) The model also uses a ``tank empty'' time to define the time 
when the tank is emptying, and the program uses a linear 
interpolation between the exponential time constants for full and 
empty during the time the tank is emptying. For a tank that is only 
used for longrange flights, the tank would be full only on longer-
range flights and would be empty a long time before touchdown. For 
short flights, it would be empty for the whole flight. For a main 
tank that carried reserve fuel, it would be full for a long time and 
would only be down to empty at touchdown. In this case, empty would 
really be at reserve level, and the thermal constants at empty 
should be those for the reserve level.
    (4) The applicant, whether using the available model or using 
another analysis tool, must propose means to validate thermal time 
constants and equilibrium temperatures to be used in the analysis. 
The applicant may propose using a more detailed thermal definition, 
such as changing time constants as a function of fuel quantity, 
provided the details and substantiating information are acceptable 
and the Monte Carlo model program changes are validated.
    (g) Overnight Temperature Drop. (1) An overnight temperature 
drop must be considered in the Monte Carlo analysis as it may affect 
the oxygen concentration level in the fuel tank. The overnight 
temperature drop for these special conditions will be defined using:
     A temperature at the beginning of the overnight period 
based on the landing temperature that is a random value based on a 
Gaussian distribution; and
     An overnight temperature drop that is a random value 
based on a Gaussian distribution.
    (2) For any flight that will end with an overnight ground period 
(one flight per day out of an average of ``x'' number of flights per 
day, (depending on use of the particular airplane model being 
evaluated), the landing outside air temperature (OAT) is to be 
chosen as a random value from the following Gaussian curve:

                          Table 4.--Landing OAT
------------------------------------------------------------------------
                                                             Landing
                       Parameter                           temperature
                                                             [deg]F
------------------------------------------------------------------------
Mean Temp.............................................             58.68

[[Page 7827]]

 
neg 1 std dev.........................................             20.55
pos 1 std dev.........................................             13.21
------------------------------------------------------------------------

    (3) The outside air temperature (OAT) drop for that night is to 
be chosen as a random value from the following Gaussian curve:

                           Table 5.--OAT Drop
------------------------------------------------------------------------
                                                            OAT Drop
                       Parameter                           temperature
                                                             [deg]F
------------------------------------------------------------------------
Mean Temp.............................................              12.0
1 std dev.............................................               6.0
------------------------------------------------------------------------

    (h) Oxygen Evolution. The oxygen evolution rate must be 
considered in the Monte Carlo analysis if it can affect the 
flammability of the fuel tank or compartment. Fuel contains 
dissolved gases, and in the case of oxygen and nitrogen absorbed 
from the air, the oxygen level in the fuel can exceed 30 percent, 
instead of the normal 21 percent oxygen in air. Some of these gases 
will be released from the fuel during the reduction of ambient 
pressure experienced in the climb and cruise phases of flight. The 
applicant must consider the effects of air evolution from the fuel 
on the level of oxygen in the tank ullage during ground and flight 
operations and address these effects on the overall performance of 
the FRM. The applicant must provide the air evolution rate for the 
fuel tank under evaluation, along with substantiation data.
    (i) Number of Simulated Flights Required in Analysis. For the 
Monte Carlo analysis to be valid for showing compliance with the 
fleet average and warm day flammability exposure requirements of 
these special conditions, the applicant must run the analysis for an 
appropriate number of flights to ensure that the fleet average and 
warm day flammability exposure for the fuel tank under evaluation 
meets the flammability limits defined in Table 6.

                      Table 6.--Flammability Limit
------------------------------------------------------------------------
                                                               Maximum
                                                             acceptable
        Number of flights  in Monte Carlo  analysis           fuel tank
                                                            flammability
                                                                 (%)
------------------------------------------------------------------------
 1,000....................................................          2.73
5,000.....................................................          2.88
10,000....................................................          2.91
100,000...................................................          2.98
1,000,000.................................................          3.00
------------------------------------------------------------------------


    Issued in Renton, Washington, on January 24, 2005.
Ali Bahrami,
Manager, Transport Airplane Directorate, Aircraft Certification 
Service.

[FR Doc. 05-2752 Filed 2-14-05; 8:45 am]
BILLING CODE 4910-13-P