DOT/FAA/TC-19/27 Full-Scale Evaluation of

Federal Aviation Administration

William J. Hughes Technical Center Novec™ 1230
Aviation Research Division
Atlantic City International Airport
New Jersey 08405

August 2019

Final Report

This document is available to the U.S. public

through the National Technical Information
Services (NTIS), Springfield, Virginia 22161.

This document is also available from the Federal Aviation

Administration William J. Hughes Technical Center at
actlibrary.tc.faa.gov.

U.S. Department of Transportation

Federal Aviation Administration
 NOTICE

This document is disseminated under the sponsorship of the U.S.

Department of Transportation in the interest of information exchange. The
United States Government assumes no liability for the contents or use
thereof. The United States Government does not endorse products or
manufacturers. Trade or manufacturer’s names appear herein solely
because they are considered essential to the objective of this report. The
findings and conclusions in this report are those of the author(s) and do
not necessarily represent the views of the funding agency. This document
does not constitute FAA policy. Consult the FAA sponsoring organization
listed on the Technical Documentation page as to its use.

This report is available at the Federal Aviation Administration William J.

Hughes Technical Center’s Full-Text Technical Reports page:
actlibrary.tc.faa.gov in Adobe Acrobat portable document format (PDF).
 Technical Report Documentation Page
1. Report No. 2. Government Accession No. 3. Recipient's Catalog No.
DOT/FAA/TC-19/27
4. Title and Subtitle 5. Report Date
FULL-SCALE EVALUATION OF NOVEC™ 1230 August 2019
7. Author(s) 6. Performing Organization Code
Casey, Jeremy
8. Performing Organization Report No.

9. Performing Organization Name and Address 10. Work Unit No. (TRAIS)
General Dynamics Information Technology (GDIT)
SRA International Inc.
a CSRA Company and subsidiary of General Dynamics
200 Decadon Drive
Egg Harbor Township, NJ 08234
11. Contract or Grant No.
DTFACT-15-D-00007
12. Sponsoring Agency Name and Address 13. Type of Report and Period Covered
U.S. Department of Transportation
Federal Aviation Administration Final Report
Airport Safety and Operations Division (AAS-300)
800 Independence Avenue SW
Washington, DC 20591
14. Sponsoring Agency Code
AAS-300
15. Supplementary Notes
The FAA Airport Technology Research and Development Branch COR was Keith Bagot.
16. Abstract
In the early 2000s, 3M™ released Novec™ 1230 Fire Protection Fluid, which has no ozone-depleting potential and a minimal
global-warming potential. Prior to the Federal Aviation Administration’s (FAA’s) consideration for approval as a complimentary
agent in Aircraft Rescue and Firefighting (ARFF) operations, Novec 1230’s fire-extinguishing capabilities had to be evaluated in
full-scale fire tests similar to the ones described in DOT/FAA/AR-95/87. A research team at the William J. Hughes Technical
Center evaluated the extinguishing capabilities of Novec 1230 discharged from a flight line extinguisher in simulated wheel brake
fire involving hydraulic fluid tests, three-dimensional inclined-plane fire tests, and pan fire tests with 16- and 30-foot (ft)
diameters. While these tests were based on those performed in DOT/FAA/AR-95/87, some parameters were changed due to test
site restrictions and to enhance repeatability. Novec 1230’s performance on the simulated engine nacelle was previously evaluated
in AFCEC-CX-TY-TR-2014-0033, the results of which were accepted for this research effort.

Under this research effort, Halotron® I also was tested on three-dimensional inclined-plane fire tests, 16-ft pan fire tests, and 30-ft
pan fire tests. Halotron I’s performance on the simulated engine nacelle fire tests was previously evaluated in NAWCWD TM
8572, the results of which were used as data for this evaluation. Data from these tests were used to compare Novec 1230’s
performance with that of Halotron I. While Halon® 1211 is still an FAA-approved clean agent for use at airports, ARFF vehicle
manufacturers have not installed Halon 1211 extinguishing systems on their products in over a decade.

Throughout all experimental configurations, Novec 1230 required more agent by both weight and volume than Halotron I. The
difference between performance parameters by percentage were consistent between the three experimental configurations with the
exception of the simulated engine nacelle conducted by the United States Air Force and Navy. The results of the simulated engine
nacelle tests were comparable with the results of the other experimental configurations with the exception of extinguishment time.

17. Key Words 18. Distribution Statement

Clean agent, Halotron, Novec, Aircraft rescue, Firefighting, This document is available to the U.S. public through the
Wheel fire, Pan fire National Technical Information Service (NTIS), Springfield,
Virginia 22161. This document is also available from the
Federal Aviation Administration William J. Hughes Technical
Center at actlibrary.tc.faa.gov.
19. Security Classif. (of this report) 20. Security Classif. (of this page) 21. No. of Pages 22. Price
Unclassified Unclassified 55
Form DOT F 1700.7 (8-72) Reproduction of completed page authorize
 TABLE OF CONTENTS
Page

EXECUTIVE SUMMARY ix

1. INTRODUCTION 1

1.1 Background 1
1.2 Purpose 2

2. OBJECTIVES 2

3. EQUIPMENT AND MATERIALS 2

3.1 Fire-Extinguishing Agent and Equipment 2

3.1.1 Novec 1230 Properties 3

3.1.2 Amerex Model 776 Fire Extinguisher 4
3.1.3 Halotron I and Amerex Model 674 Fire Extinguisher 5
3.1.4 Modified Clean Agent Skid 6

3.2 Test Locations 7

3.3 Test Fuels 8
3.4 Data Collection 8

4. TEST SETUP AND PROCEDURES 11

4.1 Pretest Procedures 11

4.2 Test Configurations and Procedures 13

4.2.1 Simulated Wheel Brake Fire Involving Hydraulic Fluid Test 13

4.2.2 Three-Dimensional Inclined-Plane Fire Test 14
4.2.3 Pan Fire Tests 16
4.2.4 Simulated Engine Nacelle Running Fuel Fire Test 18

4.3 Posttest Procedures 21

4.4 Data Analysis 22

5. TEST RESULTS AND DISCUSSION 23

5.1 Simulated Wheel Brake Fire Involving Hydraulic Fluid Test Results 23
5.2 Three-Dimensional Inclined-Plane Fire Test Results 24
5.3 The 16-ft Pan Fire Test Results 27
5.4 The 30-ft Pan Fire Test Results 29
5.5 Simulated Engine Nacelle Running Fuel Fire Test Results 37
5.6 Drop-in Validation for Truck-Based Systems 39
5.7 Overall Agent Comparison 41

iii
6. CONCLUSIONS 42

7. REFERENCES 43

iv
 LIST OF FIGURES

Figure Page

1 Amerex Model 776 Fire Extinguisher 5

2 Amerex Model 674 Fire Extinguisher 6
3 Modified Fire Combat Clean Agent Skid 7
4 Mastech MS6300 Weather Meter 9
5 Kestrel 5000 Environmental Meter 9
6 Fill-Rite FR4211D Fuel Pump With a Tuthill 800C Flow Meter 10
7 Brecknell CS-2000 Hanging Scale 11
8 Fire Extinguisher Being Weighed 12
9 Simulated Wheel Brake Fire Involving Hydraulic Fluid Test Setup 13
10 Fully Involved Simulated Wheel Brake Involving Hydraulic Fluid Test Fire 14
11 Ramp With Catch Basin 15
12 Three-Dimensional Inclined-Plane Fire Test During Preburn 16
13 The 30-ft Pan With 16-ft Ring 17
14 Fully Involved Fire in 16-ft Ring 18
15 Fully Involved Fire in 30-ft Pan 18
16 Simulated Engine Nacelle Running Fuel Fire Test Apparatus 19
17 Simulated Engine Nacelle 20
18 Approach Path for Simulated Wheel Brake Fire Involving Hydraulic Fluid Test 24
19 Novec 1230 Three-Dimensional Inclined-Plane Fire Test 4 26
20 Halotron I Three-Dimensional Inclined-Plane Fire Test 4 27
21 Test 1: Novec 1230 30-ft Pan Fire 30
22 Test 2: Novec 1230 30-ft Pan Fire 31
23 Test 3: Novec 1230 30-ft Pan Fire 32
24 Test 1: Halotron I 30-ft Pan Fire 34
25 Test 2: Halotron I 30-ft Pan Fire 35
26 Test 3: Halotron I 30-ft Pan Fire 36
27 Average Flow Rate for Simulated Engine Nacelle Running Fuel Fire Tests 41

v
 LIST OF TABLES

Table Page

1 Physical Properties of Novec 1230, Halon 1211, and Halotron I at 77°F 3

2 Environmental Properties of Novec 1230, Halon 1211, and Halotron I 3
3 Specifications for Amerex Model 776 and Model 600 Fire Extinguishers 4
4 Specifications for Amerex Model 776 and Amerex Model 674 Fire Extinguishers 5
5 Properties of JP-4 and Jet-A Fuel 8
6 Dimensions of Tires Used in Wheel Brake Fires 13
7 Results for Novec 1230 Simulated Wheel Brake Fire Involving Hydraulic Fluid Test 23
8 Results for Novec 1230 Three-Dimensional Inclined-Plane Fire Test 25
9 Results for Halotron I Three-Dimensional Inclined-Plane Fire Test 25
10 Agent Performance Comparison for the Three-Dimensional Inclined-Plane Fire Tests 27
11 Results of the Novec 1230 16-ft Pan Fire Tests 28
12 Results of the Halotron I 16-ft Pan Fire Tests 28
13 Agent Performance Comparison for the 16-ft Pan Fire Tests 29
14 Results for Novec 1230 30-ft Pan Fire Test 29
15 Results for the Halotron I 30-ft Pan Fire Tests 33
16 Novec 1230 Tests Results From AFCEC-CX-TY-TR-2014-0033 37
17 The NAWCWD TM 8572 Test Results for Halotron I With Amerex Extinguisher 674 38
18 Summary of Results From Simulated Engine Nacelle Running Fuel Fire Tests 38
19 Results for Simulated Engine Nacelle Running Fuel Fire Test With Novec 1230 39
20 Summary of Results for Simulated Engine Running Fuel Fire Tests With Novec 1230 40
21 Summary of Novec 1230 Performance by Experimental Configuration 41

vi
 LIST OF ACRONYMS

ARFF Aircraft Rescue and Firefighting

atm Atmosphere
BTU British thermal unit
CFR Code of Federal Regulations
COF2 Carbonyl fluoride
DNE Did not extinguish
EPA Environmental Protection Agency
°F Degrees Fahrenheit
FAA Federal Aviation Administration
ft Foot
g Grams
gpm Gallons per minute
HCFC Hydrochlorofluorocarbon
HF Hydrogen fluoride
in. Inch
lb Pound
m Mass
mol Mole
mph Miles per hour
NFPA National Fire Protection Association®
psi Pounds per square inch
TAFB Tyndall Air Force Base
TDP Thermal decomposition product
UL Underwriters Laboratories®
U.S. United States
USAF United States Air Force
WJHTC William J. Hughes Technical Center

vii/viii
 EXECUTIVE SUMMARY

In accordance with the Montreal Protocol Clean Air Act, Halon® 1211 had been set to be phased
out due to its ozone-depleting qualities. Production was banned at the beginning of 1994, with
remaining stockpiles only permitted for use in emergency situations and critical firefighter
training. Since then, replacement clean gaseous firefighting agents have been developed and
tested. Some testing was detailed in DOT/FAA/AR-95/87, which compared the firefighting
capabilities of Halon 1211 with two potential replacements, Halotron® I and perflourohexane
(C6). The Federal Aviation Administration (FAA) approved Halotron I as the Halon 1211
replacement for use in Aircraft Rescue and Firefighting (ARFF) operations at certificated
Title 14 Code of Federal Regulations Part 139 airports.

In the early 2000s, 3M™ released Novec™ 1230, which has no ozone-depleting potential and a
minimal global-warming potential. Prior to the FAA’s consideration for approval as a
complimentary agent in ARFF operations, Novec 1230’s fire-extinguishing capabilities had to be
evaluated in full-scale fire tests similar to the ones described in DOT/FAA/AR-95/87. A research
team at the William J. Hughes Technical Center evaluated the extinguishing capabilities of
Novec 1230 discharged from a flight line extinguisher in simulated wheel brake fire involving
hydraulic fluid tests, three-dimensional inclined-plane fire tests, and pan fire tests with 16- and
30-foot (ft) diameters. While these tests were based on those performed in DOT/FAA/AR-95/87,
the research team changed some parameters due to test site restrictions and to enhance
repeatability. The United States Air Force (USAF) conducted previous evaluations, detailed in
AFCEC-CX-TY-TR-2014-0033, which provided data and results for Novec 1230 on the
simulated engine nacelle running fuel fire tests. In addition, a research team at Tyndall Air Force
Base evaluated Novec 1230 discharged from a truck-based system with a simulated engine
nacelle running fuel fire test at to evaluate Novec 1230’s performance as a drop-in agent and to
ensure the agent’s performance was not hindered when discharged from an unoptimized system.
The team conducted additional evaluations of Halotron I to provide comparable data on the
altered experimental configurations.

Novec 1230 extinguished the simulated wheel brake fire involving hydraulic fluid tests with an
average extinguishment time of 14.3 seconds while requiring an average of 78.2 lb of agent. On
the three-dimensional inclined-plane fire tests, Novec 1230 extinguished the fires in an average
of 13.7 seconds and required 85.9 lb of agent. Novec 1230 extinguished the 16-ft pan fires in an
average of 12.7 seconds with 78.2 lb of agent. On the 30-ft pan fire tests, Novec 1230 failed to
achieve complete extinguishment; however, in two of the three tests, the overwhelming majority
of fire was suppressed. On the simulated engine nacelle with running fuel fire tests, Novec 1230
extinguished the fires in an average of 20.1 seconds with 124.2 lb of agent. When discharged
from the truck-based system, Novec 1230 was discharged at a flow rate 38.7% greater than when
discharged from an extinguisher. At this increased flow rate, Novec 1230 extinguished the
engine nacelle fires in 40.3% less time and with 10.8% less agent as compared to the nacelle
extinguisher tests. This indicated that Novec 1230’s extinguishing capabilities are not hindered
when discharged from an unoptimized, truck-based system at a higher flow rate.

Under this research effort, the team also tested Halotron I in three-dimensional inclined-plane
fire tests, 16-ft pan fire tests, and 30-ft pan fire tests. Halotron I’s performance on the simulated
engine nacelle fire tests was previously evaluated in NAWCWD TM 8572, the results of which

ix
were used as data for this evaluation. Data from these tests were used to compare Novec 1230’s
performance with that of Halotron I. While Halon 1211 is still an FAA-approved clean agent for
use at airports, ARFF vehicle manufacturers have not installed Halon 1211 extinguishing
systems on their products in over a decade.

On the inclined-plane fire tests, Novec 1230 required 5.4% more time and 18.2% more agent by
weight when compared to Halotron I. Due to differences in each agent’s density, the research
team also performed a volumetric analysis. When comparing the volume of agent required to
extinguish the inclined-plane fires, Novec 1230 required 9.1% more agent than Halotron I. On
the 16-ft pan fire tests, Novec 1230 required 5.8% more time, 21.1% more agent by weight, and
11.8% more agent by volume. Neither agent was able to extinguish any of the 30-ft pan fires, but
an analysis of the amount of fire suppressed showed that both agents performed similarly. On the
engine nacelle fires, Novec 1230 extinguished the fires 30.7% faster than Halotron I, but
required 20.9% more agent by weight and 11.7% more agent by volume. The difference in
performance parameters between each agent was consistent throughout all equivalent tests, with
the exception of extinguishment time on the simulated engine nacelle fires conducted by the
USAF and U.S. Navy.

x
1. INTRODUCTION

Since Halon® 1211 ceased production in 1994 and the availability diminishes, stakeholders have
sought suitable replacement clean agents. 3M™ is currently seeking Federal Aviation
Administration (FAA) approval for a clean agent, Novec™ 1230, for use as a complementary
extinguishing agent at airports certificated by Title 14 Code of Federal Regulations (CFR) Part
139 [1]. Novec 1230 has no ozone-depleting potential, low global-warming potential, and an
atmospheric lifetime that is measured in days rather than years. Prior to the FAA’s consideration
for approval, the Novec 1230 extinguishing capabilities must be validated as outlined in the
technical report DOT/FAA/AR-95/87 [2], as required by 14 CFR Part 139.5.

1.1 BACKGROUND

A clean agent is defined as a gaseous extinguishing substance that leaves no residue and does not
conduct electricity. Halon 1211 was previously the primary clean agent used by Aircraft Rescue
and Firefighting (ARFF) departments. However, it was found to have ozone-depleting qualities,
and the United States (U.S.) Environmental Protection Agency (EPA) developed a plan to phase
out the use of Halon 1211 in accordance with the Montreal Protocol Clean Air Act [3].
Halon 1211 production has been banned since January 1 1994, with remaining stockpiles only
permitted to be used for emergency situations and critical firefighter training [3]. Since then,
several other EPA-compliant clean agents have been developed and tested.

In October 1995, the FAA released DOT/FAA/AR-95/87, which evaluated the fire-extinguishing
capabilities of two potential replacements for Halon 1211, Halotron® I and perflourohexane
(C6) [2]. The extinguishing capabilities of Halon 1211 and both potential replacement agents
were evaluated with a series of tests thought to represent real-world applications. These tests
consisted of (1) dry-pooled fuel fire tests, (2) three-dimensional inclined-plane flowing fuel fire
tests, (3) simulated engine nacelle running fuel fire tests, and (4) simulated wheel brake fire
involving hydraulic fluid tests. In addition, the throw ranges for each agent were compared using
hand line tests. It was recommended that any future testing of potential replacements for Halon
1211 follow those test setups as closely as possible [2]. Halotron I extinguished a sufficient
number of test fires while using an average of 1.5 times as much agent by weight than
Halon 1211 did in the same tests and subsequently received approval for ARFF use at FAA-
certificated airports. Agreements were later made to phase out production of the
hydrochlorofluorocarbon HCFC-123, the main component of Halotron I, by 2020 [4]. Halotron I
production is not thought to be impacted by this phase-out due to significant remaining
inventories and its ability to be recycled [4].

In the early 2000s, 3M released Novec 1230 for use as a clean extinguishing agent. Novec 1230
received approval as a suitable replacement streaming agent under the EPA’s Significant New
Alternatives Policy in January 2003 [5]. Based on the test setups described in
DOT/FAA/AR-95/87, a series of evaluations was developed for this research effort that would
provide a comparison of the extinguishment capabilities of Novec 1230 to those of the current
clean agent most widely used in the industry, Halotron I. Although Halon 1211 has been
considered the baseline agent in previous evaluations, it is no longer produced, and no domestic
ARFF vehicle manufacturers currently offer it as a complimentary agent system option.

1
1.2 PURPOSE

The purpose of this report is to describe and analyze the results from evaluations of the
extinguishing capabilities of Novec 1230 in (1) simulated wheel brake fire involving a hydraulic
fluid tests; (2) three-dimensional, inclined-plane fire tests; (3) 16- and 30-foot (ft) pan fire tests
(which were done in place of the dry-pool fire tests); and (4) simulated engine nacelle running
fuel fire tests. These test results will be compared to results from identical tests conducted with
Halotron I, based on those described in DOT/FAA/AR-95/87 [2]. Data from evaluations
conducted by the U.S. Air Force (USAF) and Navy were used as official test results and data for
the engine nacelle fires [6 and 7].

2. OBJECTIVES

The objectives of this series of experiments are as follows:

• Evaluate the extinguishing capabilities of Novec 1230 on a variety of full-scale fires.

• Quantify the amount of agent necessary to extinguish each type of fire.

• Compare applicable results to data from previous evaluations of clean extinguishing

agents.

• Obtain data for the currently accepted clean extinguishing agent performance on new
experimental configurations performed for this research effort.

• Quantify the performance of Novec 1230 in terms of the current clean extinguishing
agent.

3. EQUIPMENT AND MATERIALS

This section describes the materials and equipment used during each evaluation to discharge the
agent, prepare the experimental configurations, and collect all relevant data.

3.1 FIRE-EXTINGUISHING AGENT AND EQUIPMENT

To satisfy all the requirements outlined in DOT/FAA/AR-95/87, a research team evaluated

Novec 1230 when being discharged from a wheeled flight line fire extinguisher. The team also
conducted an additional series of experiments to validate the use of Novec 1230 as a drop-in
replacement for current truck-based, clean agent systems being discharged through a hand line,
without the need to optimize the existing system.

Due to the changes in the experimental configurations from DOT/FAA/AR-95/87, the team
repeated some tests using Halotron I to have accurate and comparable data. These tests included
both 16- and 30-ft pan fire tests and the three-dimensional inclined-plane tests. Previous tests
conducted by the U.S. Navy provided data on the engine nacelle fire tests using Halotron I [7].

2
3.1.1 Novec 1230 Properties

Developed by 3M, Novec 1230 is a fluorinated ketone that has the chemical symbol C6F12O and
is designated in clean agent standards as FK-5-1-12. It is a liquid under ambient conditions, but
can transition to a gaseous state upon discharge from an extinguisher. In both states, it is
electrically nonconductive. Novec 1230 extinguishes fires by absorbing energy from the
surrounding space, suppressing the conditions under which combustion is able to occur. Because
of its relatively low-vaporization heat and high-vapor pressure, it can evaporate over 50 times
faster than water. Physical properties for Novec 1230 compared with those of Halon 1211 and
Halotron I are given in table 1. [8]

Table 1. Physical Properties of Novec 1230, Halon 1211, and Halotron I at 77°F [8 and 9]

Property Novec 1230 Halon 1211 Halotron I

Molecular Weight
316.04 165.4 150.7
(grams per mole (g/mol))
Boiling Point at 1 atmosphere (atm)
120.6 25.0 80.6
(degrees Fahrenheit (°F))
Saturated Liquid Density
99.9 116.0 92.3
(pound mass per cubic foot (lbm/ft3))
Density of Gas at 1 atm
0.851 0.435 0.383
(lbm/ft3)
Vapor Pressure
20.5 38.7 95.0
(pounds per square inch (psi))

Novec 1230 is reported to have a minimal environmental impact [8]. This is considered to be a
desirable characteristic due to the previous phase out of Halon 1211 caused by its negative
effects on the environment. Although the environmental impacts are outside the scope of this
research effort, it may be an important consideration if Novec 1230 is considered for approval as
a clean extinguishing agent. Relevant environmental properties for Novec 1230, Halon 1211, and
Halotron I are given in table 2.

Table 2. Environmental Properties of Novec 1230, Halon 1211, and Halotron I [8 and 9]

Property Novec 1230 Halon 1211 Halotron I

Ozone-depleting potential 0.0 4.0 0.019
Global-warming potential 1.0 1890.0 341.0
Atmospheric lifetime (years) 0.014 16.0 3.5-11.0

As with other clean firefighting agents, there are still hazards associated with thermal
decomposition products (TDPs). Novec 1230, similar to many other clean firefighting agents,
decays into its TDPs when it reacts with a fire due to high temperatures, inducing a breakdown
of its component molecules. The TDPs of greatest concern for fluorine-based agents are
hydrogen fluoride (HF) and carbonyl fluoride (COF2). COF2 reacts with water and forms HF and
carbon dioxide. HF converts to hydrofluoric acid upon interaction with water, forming an acid
gas. Hydrofluoric acid is corrosive, toxic, and poses an immediate health risk, even in small

3
amounts. It has been determined that the highest concentration of HF that is tolerable for humans
in a 1-minute exposure is approximately 120 parts per million (ppm). [10]

Different fluorine-based firefighting agents have been found to produce HF concentrations that
are considerably higher than Halon firefighting agents, with Novec 1230 having the capability to
produce approximately 10 times the amount of HF than that of Halon 1311. The factors that have
been found to affect the concentration of HF and other TDPs produced the most are the ratio
between the fire size and the room volume, the concentration of the firefighting agent, and the
agent discharge time [11]. The resultant TDP concentrations can be vastly diminished in fixed
fire suppression systems by correctly designing the system for the application; however, when
used in a streaming application during an ARFF operation, the factors affecting the production of
TDPs are not easily controlled.

3.1.2 Amerex Model 776 Fire Extinguisher

Novec 1230 was discharged from an Amerex® Model 776 fire extinguisher for the simulated
wheel brake fire test, three-dimensional inclined-plane fire test, and the 16- and 30-ft pan fire
tests. For each test described in DOT/FAA/AR-95/87, agent was contained in an Amerex Model
600 extinguisher. Table 3 shows specifications for the Model 776 and Model 600 extinguishers,
and figure 1 is a photograph of the Model 776 extinguisher. It should be noted that tests with
Novec 1230 conducted in AFCEC-CX-TY-TR-2014-0033 used an Amerex Model 775
extinguisher [6]. All specifications of the 775 and 776 are identical with the exception of the
wheels on each unit.

Table 3. Specifications for Amerex Model 776 and Model 600 Fire Extinguishers [6]

Specifications Amerex Model 776 Amerex Model 600

Agent Novec 1230 Halon 1211
Underwriters Laboratories® (UL) 3A:80B:C 30A:240B:C
Rating
Capacity (lb) 150 150
Shipping Weight (lb) 422 (filled) 272 (empty) ±15 315 (filled) 165 (empty) ±15
Discharge Time (seconds (s)) 22 48
Cylinder—DOT 4BW240
Operating Pressure (psi) 125 200
Test Pressure (psi) 480 480
Burst Pressure (minimum psi) 960 960
Discharge Range (ft) 30 30–40
Operating Temperature Range (°F) -40 to +120 -65 to +120
Safety Disc Burst Range (psi) 400–500 400–500
Hose Length (ft) 40 50
Hose Diameter (in.) 1.0 0.75
Wheels (in.) 36 vulcanized rubber 16 × 4 semi-pneumatic
Height (in.) 62 59
Width (in.) 29 29
Depth (in.) 40 36

4
 Figure 1. Amerex Model 776 Fire Extinguisher

3.1.3 Halotron I and Amerex Model 674 Fire Extinguisher

Halotron I is currently the only produced, FAA-approved clean extinguishing agent for ARFF
applications. Table 4 shows the physical properties of Halotron I. The research team used
Halotron I in this research effort to provide comparable data on new or modified experimental
configurations.

The research team used an Amerex Model 674 extinguisher to discharge Halotron I in all
respective tests conducted during this research effort. It should be noted that during the initial
series of tests conducted under DOT/FAA/AR-95/87 with Halotron I, the agent was discharged
from a Halon 1211 extinguisher with no performance modifications or alterations. Table 4 shows
the specifications of the Model 674 extinguisher compared to the Amerex Model 776
extinguisher used for Novec 1230 testing, and figure 2 is a photograph of the Amerex Model 674
extinguisher.

Table 4. Specifications for Amerex Model 776 and Amerex Model 674
Fire Extinguishers [6 and 7]

Specifications Amerex Model 776 Amerex Model 674

Agent Novec 1230 Halotron I
UL Rating 3A:80B:C 10A:120B:C
Capacity (lb) 150 150
Shipping Weight (lb) 422 (filled) 272 (empty) ±15 388 (filled) 238 (empty) ±15
Discharge Time (s) 22 38
Cylinder DOT 4BW240 DOT 4BW500
Operating Pressure (psi) 125 125
Test Pressure (psi) 480 480
Burst Pressure (minimum psi) 960 1200
Discharge Range (ft) 30 30–40

5
 Table 4. Specifications for Amerex Model 776 and Amerex Model 674
Fire Extinguishers [6 and 7] (Continued)

Specifications Amerex Model 776 Amerex Model 674

Operating Temperature Range (°F) -40 to +120 -40 to +120
Safety Disc Burst Range (psi) 400–500 400–500
Hose Length (ft) 40 50
Hose Diameter (in.) 1.0 0.75
Height (in.) 62 64
Width (in.) 29 30
Depth (in.) 40 42

Figure 2. Amerex Model 674 Fire Extinguisher

3.1.4 Modified Clean Agent Skid

Instead of using a wheeled flight line fire extinguisher for the simulated engine nacelle running
fuel fire tests, a truck-based system mounted to a skid was used for hand line application. This
was done for two reasons: (1) the USAF has performed identical experiments previously with a
reliable set of data, and (2) it allows for a comparison of extinguisher application to hand line
application. The use of this system provides validation to ensure Novec 1230 can be used as a
direct drop-in replacement agent without the need to modify existing systems. Figure 3 is a
photograph of the modified Fire Combat clean agent, skid-mounted system used in the test. The
research team modified the skid system to include instrumentation and data collection systems.
The team installed these modifications so they did not alter the performance of the system. The
agent storage tank was positioned on top of four load cells to monitor the weight of the tank for
determining the amount of agent discharged and flowrate across all the tests. Additionally, two

6
pressure transducers with snubbers were installed on this system, one on top of the agent tank
and the other immediately preceding the nozzle. These sensors were then connected to a data
acquisition system, which was connected to a laptop running a custom LabVIEW® program to
record and monitor all data. The specifications of this system were designed to meet the
standards of the National Fire Protection Association® (NFPA) 414 Standard for Aircraft Rescue
and Firefighting Vehicles for halogenated hand lines [12]. This Standard specifies a discharge
range of at least 25 ft, a minimum discharge rate of at least 5 lb/s, a hose with an inner diameter
of at least 1 in., and a hose length of at least 100 ft [12].

Figure 3. Modified Fire Combat Clean Agent Skid

3.2 TEST LOCATIONS

A research team at William J. Hughes Technical Center (WJHTC) conducted the simulated
wheel brake test, three-dimensional inclined-plane test, and the pan fire tests. The WJHTC team
performed these tests within a designated full-scale fire test area on a specialized concrete pad.
The pad is pitched towards a drain to contain and transfer the fuel, firefighting agent, and
contaminated water into a holding tank, thereby preventing contamination to the surrounding
environment. Additionally, a water hand line and a foam hand line were available for safety.

A research team at Tyndall Air Force Base (TAFB) conducted the simulated engine nacelle
running fuel fire tests using an F100 engine nacelle mockup. The team situated the nacelle on a
concrete catch basin with a sump pit, which was designed to collect and contain most of the fuel
and agent discharged during each test. After each test, the team pumped the sump into a holding
tank for disposal later. A high-pressure foam hand line was available for safety.

7
3.3 TEST FUELS

JP-4 fuel was used for the Halon replacement tests in DOT/FAA/AR-95/87. However, due to the
limited availability of JP-4 and the abundance of Jet-A fuel, the teams used Jet-A in the three-
dimensional inclined-plane fire tests, 16- and 30-ft pan fire tests, and the simulated engine
nacelle running fuel fire tests. Jet-A is a kerosene-based fuel with a regression rate lower than
JP-4 fuel. Table 5 shows the relevant properties of JP-4 and Jet-A fuels.

Table 5. Properties of JP-4 and Jet-A Fuel [13 through 17]

Properties JP-4 Jet-A

Flash Point (°F) 0.0 100.4
Auto-Ignition Temperature (°F) 474.8 410.0
Freezing Point (°F) -52.2 -42.8
Boiling Point Range (°F) 113-572 300-572
Vapor Pressure at 68°F (psi) 1.76 0.0077
Density at 71°F (lb/ft3) 47.45 50.57
Lower Flammability Limit (%) 1.3 0.7
Upper Flammability Limit (%) 8.0 5.0
Specific Energy 18451 18451
(British thermal unit per pound mass (BTU/lbm))
Regression Rate (in./min) 0.148 0.130
Specific Heat at 104°F (BTU/lbm°F) 0.514 0.490

The simulated wheel brake fire test was representative of hot brakes igniting hydraulic fluid and
the resultant oil and tire fire. The hydraulic fluid used for the testing in DOT/FAA/AR-95/87 [2]
was specified as Military Specification MIL-H-5606F [18]. The tests conducted with Novec
1230 used Mobil Aero™ HFA Aviation Hydraulic fluid, which met MIL-H-5606A [19]
specifications.

3.4 DATA COLLECTION

For all WJHTC evaluations, the team photographed the test area before, during, and after each
test and recorded all the tests with video cameras in at least two different positions to capture
alternate viewpoints. The team placed the cameras at a minimum offset of 90° from each other
around the test fixture, allowing for the most comprehensive coverage of each test and
uninterrupted documentation in the event that one camera’s view became obscured. The camera
positions for each test varied based on which test fixture was used, wind direction, extinguisher
discharge direction, and test area obstructions present.

The WJHTC team recorded data on wind speed, weight of the extinguisher before and after
testing, preburn time, discharge time, and time to extinguishment. The team members noted the
time and details of any peculiarities such as fire spreading outside the test area, re-ignition of fire
after the initial extinguishment, or accidental discharge of the extinguisher.

8
For all TAFB evaluations, the team took photographs of the test area before, during, and after
each test, which was also recorded by two video cameras from different positions. A laptop
computer running the custom LabVIEW program connected to the modified FireCombat skid
system recorded discharge time, agent discharged, and tank and nozzle pressures.

The teams measured wind speed and monitored it for at least 3 minutes prior to each test to
ensure there were no gusts that exceeded the respective test’s threshold. The teams used
handheld weather meters to measure and monitor the weather conditions at both test locations.
The WJHTC team used a Mastech® MS6300 (figure 4) and TAFB team used a Kestrel® 5000
Environmental Meter (figure 5). A research team member sampled multiple areas around the test
fixture prior to each test. If any measurement exceeded the threshold, testing was suspended until
weather conditions were more favorable. The average and peak wind speeds were both recorded
prior to each test.

Figure 4. Mastech MS6300 Weather Meter

Figure 5. Kestrel 5000 Environmental Meter [20]

9
For each WJHTC evaluation that used Jet-A fuel, the fuel was pumped through a Fill-Rite®
FR4211D with a Tuthill 800C flow meter, shown in figure 6. The team used this pump to
measure and distribute the fuel in each test. The team also used this pump in conjunction with a
stopwatch to determine the fuel flow rate when conducting experiments on the three-dimensional
inclined-plane test fixture.

Figure 6. Fill-Rite FR4211D Fuel Pump With a Tuthill 800C Flow Meter

In each experiment where an extinguisher was used, the extinguisher was weighed before and
after discharge. The team used a Brecknell® CS-2000 scale, shown in figure 7, to weigh the
extinguisher before each test, after each test, and during each refill procedure. The scale was
attached to each extinguisher and lifted by a forklift. Each measurement was taken a minimum of
two times and recorded. For the experiments that used the Fire Combat skid, the team
continuously monitored and recorded the weight using the LabVIEW program and integrated
load cells.

10
 Figure 7. Brecknell CS-2000 Hanging Scale

Two research team members monitored and recorded times using handheld stopwatches. The
stopwatches were used to determine when the preburn phase of each experiment concluded and
extinguishment efforts began. The discharge and extinguishment times were recorded by hand
then later verified and/or corrected using videos from each test.

4. TEST SETUP AND PROCEDURES

Sections 4.1 through 4.4 describe the pretest procedures, test configurations and procedures,
posttest procedures, and data analysis used during this research effort.

4.1 PRETEST PROCEDURES

Prior to conducting any tests, team members measured the wind speed with a handheld
anemometer at several locations around the test area to ensure the wind was consistently below
the maximum allowable speed for the respective test. During gusts, both the average and
maximum wind speeds were recorded. For the simulated wheel brake fire and three-dimensional
inclined-plane fire, if the wind speed surpassed the predetermined threshold by a minimal
amount, team members placed wind screens upwind of the test fixture (shielding the test area)
and rechecked the wind speeds in multiple areas around the test fixture. However, if wind speeds
in the immediate area could not be reduced below the test’s limit, testing was suspended until
conditions became favorable. In addition, testing was suspended if it began to rain to ensure
accurate results for the firefighting agent capabilities.

While weather conditions were being monitored, team members weighed the extinguisher a
minimum of two times and recorded the starting weight. Figure 8 is a photograph of the forklift
lifting the scale and extinguisher configuration. For the engine nacelle fires using the clean agent
skid unit, the data acquisition system measured and recorded the starting weight. This process
included connecting the laptop to the data acquisition system and confirming all sensors were
connected and functioning. Prior to the first test, the team calibrated the load cells by placing
calibration weights on the agent tank and confirming the accuracy of the measurements.

11
 Figure 8. Fire Extinguisher Being Weighed

The three-dimensional inclined-plane fire test and the 16- and 30-ft pan fires required the pans to
be filled with water. This was done to ensure a consistent and even fuel layer and to protect the
concrete test pad from spalling. For the 16- and 30-ft pan fires, the pan was filled until a
minimum ¾ in. of water completely covered the bottom of the pan. For the inclined ramp, the
team filled the pan approximately halfway with water (approximately 4 in.) prior to beginning
the fuel flow.

Additional preparations for each test included setting up the video cameras, wetting the test pad,
and setting up safety lines. For all WJHTC tests, prior to dispensing fuel, the team saturated the
concrete around the test fixture to prevent spalling from the heat of the fire. Also, during the pan
fires, team members stretched and charged an AFFF hose line. For TAFB tests, the team
stretched and readied an AFFF hose line to be used as a backup line.

After all other preparations were made and weather conditions were favorable, a team member
dispensed fuel into each test fixture. For the simulated wheel brake fire, the team member
dispensed hydraulic oil from graduated buckets to ensure proper fuel volume and then dispensed
the oil onto the test fixture. For the inclined ramp, a team member confirmed the flow rate by
measuring the time it took to dispense 4 gallons of fuel into a graduated bucket through an
alternate outlet. After confirmation, the team member shut down the fuel pump and configured
valves to allow fuel to flow over the ramp. For the 16- and 30-ft pan fires, a team member
dispensed fuel through a hose and nozzle that were connected to the fuel pump. The team
monitored the volume from the flow meter on the fuel pump, and shut it down once the desired
value was reached. For the engine nacelle, the team used the inline flow meter to confirm fuel
flow prior to the warming process. The team used a wheeled propane torch to ignite the fuel until

12
a significant portion of the fixture was on fire, after which the torch was removed from the test
area.

4.2 TEST CONFIGURATIONS AND PROCEDURES

Sections 4.2.1 through 4.2.4 describe the setup and execution of each test that was performed.
The research teams designed the tests to follow the ones outlined in DOT/FAA/AR-95/87 as
closely as possible, with aspects of those tests being altered if necessary. Additionally, the
simulated engine nacelle fire tests followed the procedures described in AFCEC-CX-TY-TR-
2014-0033 [6].

4.2.1 Simulated Wheel Brake Fire Involving Hydraulic Fluid Test

The purpose of this test was to represent an aircraft brake fire scenario involving hydraulic fluid.
The test setup consisted of an aircraft rim and tire mounted on a stand inside a 4-by-4-ft pan,
which is shown in figure 9.

Figure 9. Simulated Wheel Brake Fire Involving Hydraulic Fluid Test Setup

The aircraft tire used in the DOT/FAA/AR-95/87 test was a deflated McDonnell Douglas F-4C
aircraft tire, whose nominal dimensions are given in table 6 [21]. The deflated dimensions were
not provided in DOT/FAA/AR-95/87. The tire used in this research effort was a deflated Boeing
B-737-200 main gear tire, whose nominal dimensions are given in table 6 [21]. When deflated,
the tire had a 36-in. diameter and 14-in. width. The tire was mounted on a 16-in. rim with a
14-in. width. The allowable wind speed threshold for testing with this configuration was 3 miles
per hour (mph).

Table 6. Dimensions of Tires Used in Wheel Brake Fires [21]

Halotron I Evaluations Novec 1230 Evaluations

Research Effort (DOT/FAA/AR-95/87) (this report)
Tire Size (in.) 30X11.50-14.50 40X14-16
Rim Diameter/Width (in.) 14.50/9.75 16/14

13
To begin, the research team dispensed 2 gallons of hydraulic fluid in the bottom of the pan,
followed by 1 gallon on top of the tire itself, distributing the fluid as evenly as possible. The fluid
in the pan was ignited the first, followed by the fluid on the tire. After the fire became fully
involved, a 90-second preburn phase began. The fire was considered fully involved when all
areas of hydraulic fluid appeared to be ignited. An example of a fully involved fire is shown in
figure 10. After 90 seconds, the firefighter approached the tire and commenced extinguishing
efforts. When the research team confirmed the fire was fully extinguished, the team recorded the
time to extinguishment and discharge time and measured the extinguisher weight to determine
how much agent had been discharged.

Figure 10. Fully Involved Simulated Wheel Brake Involving Hydraulic Fluid Test Fire

Another difference between the simulated wheel brake fire testing conducted in this research
effort and DOT/FAA/AR-95/87 was the firefighter’s approach. In DOT/FAA/AR-95/87, the
firefighter took a direct approach parallel to the tire tread. For the simulated wheel brake fires
with Novec 1230 conducted during this research effort, the firefighter took an offset approach,
approximately 45° to the tire tread. The firefighter took this approach to provide consistency
with the currently taught ARFF-training tactics.

4.2.2 Three-Dimensional Inclined-Plane Fire Test

The purpose of this test setup was to simulate a fire propagated by fuel flowing over sloped
terrain into a pooled area. The test setup consisted of a concrete ramp, the top of which had a
pipe that dispensed fuel at a rate of 4 gallons per minute (gpm) through 27 evenly spaced, ¼-in.
holes. This rate was increased from the DOT/FAA/AR-95/87 tests to ensure the ramp was
adequately covered in fuel. The 3-gpm flow rate specified in DOT/FAA/AR-95/87 did not
completely cover the ramp so the research team gradually increased the fuel flow rate until the
ramp was adequately covered. The fuel flowed down the ramp into a metal catch basin, which
was partially filled with water to protect the concrete of the test pad. The ramp was 20 ft long,

14
61 in. wide, and had a slope of 4.7°. The catch basin was 4 ft long and 8 ft wide, with the wider
side oriented parallel to the edge of the ramp. This fixture is shown in figure 11.

Figure 11. Ramp With Catch Basin

Before each test began, the research team cleaned the ramp, drained and rinsed the catch basin to
remove any contaminants, and added new water. Then, the team pumped Jet-A fuel out of the
pipe at a rate of 4 gpm. Once the fuel had covered the ramp and began pouring into the catch
basin, the accumulation period began. This accumulation period lasted until 5 gallons of fuel had
been dispensed, ensuring adequate fuel accumulation on the ramp and in the catch basin. After
5 gallons of fuel had accumulated, the fuel was ignited. Ignition began at the top of the ramp,
below the fuel dispensing pipe, and continued down the ramp as necessary. Due to the
continuous flow of fuel, upper portions of the ramp had to be reignited during the ignition
process in some cases. It should be noted that the fuel flowing from the dispensing pipe did not
ignite until it splashed onto the ramp throughout the entirety of each test. When the fire became
fully involved, a 60-second preburn phase began, as shown in figure 12. Fully involved was
defined as when all of the catch basin and all of the fuel flowing down the ramp were on fire.
The firefighter then commenced extinguishing efforts, beginning at the catch basin and moving
up the ramp towards the fuel-dispensing pipe. When the fire was fully extinguished, the team
shut down the fuel supply to the ramp, recorded the time to extinguishment and discharge time,
and measured the weight of the extinguisher again to determine how much agent had been
discharged. The wind speed threshold for testing with this fixture was 5 mph.

15
 Figure 12. Three-Dimensional Inclined-Plane Fire Test During Preburn

4.2.3 Pan Fire Tests

In DOT/FAA/AR-95/87, the effectiveness of Halon 1211 and both potential replacement agents
were evaluated using a dry-pool fire test. This test was intended to simulate fuel spills during
flight line operations. The test was performed by pouring fuel onto a dry concrete surface over a
given area that varied between 200 and 800 sq ft, igniting the fuel, and attempting to extinguish
the fire in the shortest possible amount of time [2]. However, there were concerns that
performing such a test on the test pad at the WJHTC would induce spalling in the concrete and
that the low fuel quantity-to-area ratio would produce inconsistent fires, which could possibly
self-extinguish during extinguishing procedures. In light of these concerns, the research team
modified the testing parameters by increasing the fuel load for a given area and using water as
the surface medium for the pooled fuel. These modifications to the test procedures provided a
consistent fuel load and equivalent fire area for all same-sized pan tests regardless of application
method. Due to these alternative configurations, a direct comparison of results between this
research effort and that of DOT/FAA/AR-95/87 for representative spilled fuel fires does not
allow for an accurate comparison of extinguishing performance. For this reason, the research
team performed these tests with both Novec 1230 and Halotron I to allow for a comparison of
the data under identical test conditions.

The alternative configurations consisted of a shallow steel pan with a 30-ft diameter, and a steel
ring with a 16-ft diameter that could be placed in the pan and used to contain fuel in a smaller
area. Both the ring and the pan were 6 in. deep. The ring and the pan allowed for the testing of
fires with areas of approximately 200 and 700 sq ft, respectively. In both configurations, the pan
was partially filled with water prior to dispensing the fuel. Using water as a test bed protects the
concrete pad and test fixture from damage and produces a uniform fuel layer across the surface.
The uniform fuel layer ensured that the entire area would burn. Figure 13 shows a research team
member dispensing fuel into the 16-ft ring. Another research team member had previously filled
the 30-ft pan with water.

16
 Figure 13. The 30-ft Pan With 16-ft Ring

Prior to each test, the team cleaned the steel pan to ensure that there was no contaminates that
could influence each test. Cleaning consisted of debris removal, rinsing the pan with water, and
then removing any remaining puddles with squeegees. The pan was then filled with water until it
reached a depth of at least ¾ in. at the shallowest section.

After the team filled the pan with an adequate amount of water, they added the prescribed
amount of fuel to either the ring or the entire pan for each test. The volumes of Jet-A fuel
specified were 19.8 gallons for the 16-ft ring and 70 gallons for the 30-ft ring. These values were
determined based on a calculated fuel consumption of a steady-state, fully involved pooled fuel
fire with no wind, to ensure a burn time of at least 1 minute regardless of which fuel was used.
These amounts were determined to be sufficient to ensure the fuel had not been exhausted prior
to being considered extinguished for each configuration. The fuel was ignited after the firefighter
dispensed the proper amount of fuel into the ring or the pan. Ignition began at one point on the
edge of the fuel layer and continued around the edge until a significant portion of fuel was
ignited. The fire was then allowed to naturally spread across the pooled fuel. The fire was
considered fully involved once the entire pool was on fire, as shown in figures 14 and 15 for the
16-ft ring and 30-ft pan tests, respectively. Once the fire became fully involved, the team
conducted a 20-second preburn phase, after which extinguishing efforts began. When the
firefighter extinguished the fire or depleted the extinguisher, the team recorded the time to
extinguishment and total discharge time, and measured the extinguisher weight to determine how
much agent was discharged. The wind speed threshold for testing with these fixtures was 7 mph.

17
 Figure 14. Fully Involved Fire in 16-ft Ring

Figure 15. Fully Involved Fire in 30-ft Pan

4.2.4 Simulated Engine Nacelle Running Fuel Fire Test

This test utilized the USAF’s F100 nacelle mockup, pictured in figure 16, which is designed to
simulate an engine fire with fuel continuously leaking onto the ground. The TAFB research team
conducted these tests at the Silver Flag test site. All tests conducted during this research effort
for this experimental configuration utilized Novec 1230 discharged from a hand line supplied by
the Fire Combat clean agent skid unit. This alteration of discharge method was done for two
reasons: the USAF has previously conducted equivalent tests using an equivalent Novec 1230
extinguisher, and altering the discharge method would provide validation that hand line

18
application does not hinder the extinguishing capabilities of the agent. The previous tests
conducted by the USAF, as detailed in AFCEC-CX-TY-TR-2014-0033, produced a valid and
relative data set to be used in this research effort. In light of this available data set, it was decided
to use this experimental configuration as a verification of extinguishing capabilities when
discharged from a hand line [6]. The hand line application is representative of a typical ARFF
response under conditions matching this experimental configuration. The U.S. Navy also
performed clean agent tests on the simulated engine nacelle running fuel fire test fixture using
Halon 1211 and Halotron I. These tests were detailed in NAWCWD TM 8572, and the relevant
Halotron I tests were accepted as data to be used during this research effort [7].

The test apparatus consisted of two concentric metal cylinders that had a length of 16 ft and
internal volume of 189 ft3, and was raised 47 in. off the ground over a catch basin. The cylinder
was divided into four sections, with a 36-in.-long front section (fan section), the two 33-in.-long
middle sections (compressor and low pressure turbine sections), and the 90-in.-long rearmost
section. Between each section was a series of metal baffles that were 2 in. wide, 1/8 in. thick, and
spaced 2 in. apart. While modifications to this test fixture have been made since used in
DOT/FAA/AR-95/87, the dimensions have remained the same. Figure 16 is a photograph of the
nacelle, and figure 17 shows a cross-sectional diagram of the nacelle with relevant parts and
dimensions labeled. The nacelle can be moved around the catch basin to account for changes in
wind direction while keeping the afterburner section above the catch basin. [7]

Figure 16. Simulated Engine Nacelle Running Fuel Fire Test Apparatus

19
 Figure 17. Simulated Engine Nacelle [7]

The procedure for this test followed that of the rear-engine fire tests detailed in AFCEC-CX-TY-
TR-2014-0033, which was split into distinct pretest and test phases. To begin the pretest phase,
the team confirmed that all instrumentation and the data acquisition system was functioning and
accurate, the agent tank was pressurized, and the pressure and weight of the agent tank were
recorded. Next, a team member started the flow of Jet-A fuel at a rate of 2 gpm through the
afterburner nozzle. That fuel was then ignited, and the tail pipe was heated to a temperature of
within 25° of 550°F. A team member verified the temperature using a handheld infrared
thermometer. The fuel flow was then shut off, and the tail pipe was allowed to cool to a
temperature of within 25° of 475°F, again verified by a handheld infrared thermometer. A team
member then started fuel flows through the low-pressure turbine and afterburner nozzles at a rate
of 2 gpm each, providing a total flow rate of 4 gpm. The fuel was allowed to flow until 25
gallons had been dispensed. If spontaneous ignition occurred during this process, the team
member shut off the fuel flow, allowed the fixture to cool further, and restarted the fuel flow. [6]

Prior to beginning the test phase, team members charged the hand line and recorded
instrumentation measurements. Then, the team began the test phase by igniting the two fuel
flows from the low-pressure turbine and afterburner nozzles using torches inserted through the
corresponding ignition ports. Next, the firefighter ignited the fuel in the catch basin and allowed
it to burn for 15 seconds, after which extinguishing efforts commenced. Once the fire was
extinguished, a team member recorded the time to extinguishment and amount of agent
discharged as determined by load cell values on the clean agent skid. The wind speed threshold
for testing with this setup was 5 mph.

20
4.3 POSTTEST PROCEDURES

After a test was completed at the WJHTC, the research team weighed the extinguisher with any
remaining agent, hose, and nozzle. The extinguisher weight was used to determine the amount of
agent discharged during that test. A team member shut the extinguisher valve, discharged any
remaining agent in the hose, and then disconnected the hose from the extinguisher. The
extinguisher tank was positioned horizontally, and the discharge valve was opened to
depressurize the extinguisher. This was done to conserve as much agent as possible. The
Novec 1230 extinguisher was completely depressurized, while the Halotron I extinguisher was
bled until 5-10 psi remained. These differences are solely attributed to the refilling procedures
for each extinguisher.

Once the discharge valve of the Novec 1230 extinguisher was removed, the extinguisher was
hung from the scale. New agent was pumped into the extinguisher from a 55-gallon drum while
monitoring the weight. Once it was filled to the desired weight, the extinguisher was slowly
pressurized with dry nitrogen while being agitated to promote the dissolution of nitrogen into the
agent and achieve the required pressure. To ensure consistency between tests, the extinguisher
was refilled and repressurized to the same amounts each time, a total weight of 422 lb and an
internal pressure of 125 psi. After refilling and repressurizing the extinguisher, the team member
weighed it again and recorded that as the starting weight for the next test, if the test was expected
to occur the same day. An on-site 3M representative verified this process.

To refill the Halotron I extinguisher after being mostly depressurized, a team member connected
a set of valves and gauges to the extinguisher’s discharge port. These valves and gauges were
connected to a bulk tank, which was pressurized with a cylinder of argon. The extinguisher was
then hung from the scale to monitor the weight during the refill process. The liquid side of the
bulk tank was opened to allow Halotron I to flow into the extinguisher until a total weight of
468 lb had been reached. Then, the liquid side of the Halotron I tank was closed and the gas side
was opened, allowing argon and gaseous agent to flow into the extinguisher and pressurize it.
The extinguisher was slowly pressurized to 150 psi, as verified by the extinguishers pressure
gauge. The valve and associated hoses were then removed from the extinguisher, and the
extinguisher was weighed again. This weight was used as the starting weight if another test was
expected to be conducted. An on-site representative from American Pacific verified this process.

Refilling procedures differed slightly for the engine nacelle tests, which used the clean agent skid
unit. First, a team member isolated the agent tank from the rest of the system using the discharge
valve and pressurization valves and then depressurized the tank. The clean agent skid unit is
equipped with an agent fill port, which was then connected to the pump and used to refill the
agent tank. The skid’s load cells were used to monitor the weight during the refilling process.
Agent was pumped into the tank until the desired weight was achieved. The fill hose was then
disconnected, and the tank pressurization valve was opened very slowly to promote the
dissolution of nitrogen into the agent. After reaching equilibrium, the pressurization valve was
fully opened.

Following the completion of each test, the team serviced and prepared each test fixture for the
following test. This involved removing any water, fuel, or agent that remained; rinsing off the

21
fixtures; and removing any standing water. After the team serviced the extinguishers and test
fixtures and recorded all data, the test was considered complete.

4.4 DATA ANALYSIS

To determine the discharge time and time to extinguishment for tests that used an extinguisher,
the values logged on the data sheets were cross-referenced with video footage from the
corresponding tests. Time to extinguishment was calculated as the time between the initial
discharge and the earliest time when no fire could be seen in any of the footage. Discharge time
was calculated as the total time the firefighter had the hose nozzle open and was dispensing
agent. The amount of agent used was divided by the total discharge time to determine the
discharge rate.

From each video, a team member measured the time when the firefighter was directing discharge
at the test area, as well as the time that discharge took to extinguish only the fire within the
confines of each test fixture. These measurements represented the required discharge and
extinguishment times, respectively. The team member multiplied the required extinguishment
time by the calculated discharge rate to provide an estimate for just the amount of agent that was
necessary to extinguish the test fire. In some instances, the firefighter directed the discharge to
fire that had spread outside the test fixture. When this occurred for a minimum of 1 second, the
team member used the discharge rate to estimate the weight of agent not directed at the test
fixture and accounted for this when calculating the required discharge and time to
extinguishment values.

For each test configuration, the team made a table listing the test number, date, average wind
speed, maximum wind speed during gusts, test fixture extinguishment time, total discharge time,
total agent used, estimated agent used for test fixture extinguishment, and flow rate for each test
in that setup. Any tests that were not considered to be official were not included in the tables and
not used in any of the corresponding calculations.

The team performed further analysis to calculate the volume of agent discharged during each
evaluation. This was done for both Novec 1230 and Halotron I by using the saturated liquid
densities at 1 atm for each agent. The liquid density was assumed to remain constant, and the
vapor weight for each agent was assumed to be negligible in this analysis. These assumptions are
corroborated by the manufacturer recommended use of a calibrated liquid level gauge to
determine the amount of agent present in the skid unit during refilling procedures. [22] This
method provided another basis of comparison of the results in an attempt to account for the
difference in agent densities. The value for estimated volume of agent used is derived from the
weight of each agent divided by the saturated liquid density of the respective agent.

Next, the team compared the results of each evaluation with Novec 1230 to results from
evaluations of Halotron I, using Halotron I as the baseline. This was done due to Halotron I
being the only currently produced, FAA-approved clean agent at indexed airports. The research
team felt this would be a more accurate representation of agent performance due to the
availability of equivalent testing data and the real world ubiquitous use of Halotron I.

22
5. TEST RESULTS AND DISCUSSION

The WHJTC research team tested Novec 1230 in March and April of 2017 and tested Halotron I
in August through October of 2017. The TAFB research team tested Novec 1230 at the end of
May 2017. This section details the results of those tests, while comparing them to the results of
similar tests with Halotron I, the only currently produced, FAA-approved, clean-extinguishing
agent.

5.1 SIMULATED WHEEL BRAKE FIRE INVOLVING HYDRAULIC FLUID TEST

RESULTS

The research team conducted four simulated wheel brake fire tests. The results of three tests were
considered to be official and used in the analysis. The results of Test 3 were discarded due to
reignition of the hydraulic fluid caused by a small pocket of fire not seen by the firefighter.
Table 7 shows the results for the extinguishment time, discharge time, and agent used for each
test.

Table 7. Results for Novec 1230 Simulated Wheel Brake Fire Involving Hydraulic Fluid Test

Average Maximum Extinguishment Discharge Agent Agent Discharge

Wind Wind Time Time Used Required Rate
Test Date (mph) (mph) (s) (s) (lb) (lb) (lb/s)
1 3/7/2017 2.5 2.5 17 17 91 91.0 5.35
2 3/7/2017 0.0 2.5 15 16 85 79.7 5.31
4 3/7/2017 1.1 2.2 11 15 87 63.8 5.80
Average 14.3 78.2

The Test 1 fire was extinguished in 17 seconds, with a total discharge time also of 17 seconds
and 91 lb of agent used, all of which was necessary to extinguish the fire. The Test 2 fire was
extinguished in 15 seconds, with a discharge time of 16 seconds and 85 lb of agent discharged,
an estimated 79.7 lb of which was necessary to extinguish the fire. The Test 4 fire was
extinguished in 11 seconds with 15 seconds of discharge time. A total of 87 lb of agent was used
during Test 4, 63.8 lb of which was estimated to be necessary to extinguish the fire. On average,
each of the three official wheel brake fire tests took 14.3 seconds to extinguish and required 78.2
lb of agent to extinguish the fixture.

The Novec 1230 tests with this setup differed from the evaluations described in
DOT/FAA/AR-95/87 in the direction that the firefighter took to approach the tire. During tests
with Halon 1211 and Halotron I, the firefighter took an approach path that was parallel to the tire
tread, whereas for the Novec 1230 testing, the firefighter made a diagonal approach instead, as
shown in figure 18. This diagonal approach path was taken to more closely align with currently
taught ARFF techniques. Because of this, a larger section of the tire surface and pan was
obstructed, which could have adversely affected the firefighter’s ability to extinguish the fire as
compared to previous testing. This may have resulted in additional repositioning around the test
fixture to completely extinguish the fire. This obstructed view was the reason Test 3 was
removed from the official results. A small pocket of fire remained on the opposite side of the tire
from where the firefighter was positioned after closing the bail, and subsequently the fire
reignited the test fixture.

23
 Figure 18. Approach Path for Simulated Wheel Brake Fire Involving Hydraulic Fluid Test

Each Novec 1230 simulated wheel brake fire test had shorter extinguishment times than the test
that preceded it. The same firefighter was used for each test, which could be an indicator of the
firefighter becoming familiar with the test fixture and the extinguishing process, resulting in
more efficient agent application. For each test, the portion of the fire that was directly in front of
the firefighter was extinguished quickly, taking between 4 to 6 seconds. The firefighter spent the
remainder of the extinguishment time repositioning around the fixture while applying agent.

The results for the Novec 1230 simulated wheel brake fire tests will not be compared to the
results of Halotron I from DOT/FAA/AR-95/87. While the test configurations were very similar,
the research team concluded that the differences between the tests conducted in both research
efforts skewed the results significantly. It should be noted that only one test for Halotron I was
conducted during the evaluations described in DOT/FAA/AR-95/87. After thorough review of
video footage from both series of tests, the research team found that the offset approach in
conjunction with the larger tire caused a significantly greater portion of the test fixture to be
obscured compared to previous testing. This affected the firefighter’s view of the fire and
hindered the direct application of agent. As previously described, the majority of the discharge
times during the Novec 1230 tests consisted of the firefighter repositioning around the test
fixture to ensure all fire was extinguished. In light of the differences between test parameters and
the associated effects, a comparison of the results from both tests would not be representative of
either agent’s extinguishing capabilities.

5.2 THREE-DIMENSIONAL INCLINED-PLANE FIRE TEST RESULTS

A total of four Novec 1230 three-dimensional inclined-plane fire tests were conducted, three of
which were considered to be official and used in this analysis. The Test 2 results were discarded
due to an improperly filled extinguisher. The results for all official Novec 1230 tests are
provided in table 8.

24
 Table 8. Results for Novec 1230 Three-Dimensional Inclined-Plane Fire Test

Average Maximum Extinguishment Discharge Agent Agent Flow

Wind Wind Time Time Used Required Rate
Test Date (mph) (mph) (s) (s) (lb) (lb) (lb/s)
1 3/9/2017 0.9 2.2 10 11 74 67.3 6.73
3 3/21/2017 0.0 1.8 11 16 94 64.7 5.88
4 3/28/2017 2.9 4.0 20 21 132 125.8 6.29
Average 13.7 85.9

The Test 1 fire was extinguished the in 10 seconds, with a discharge time of 11 seconds. A total
of 74 lb of agent was discharged, 67.3 lb of which was necessary to extinguish the fire. The
Test 3 fire had a discharge time of 16 seconds, 5 seconds of which was not directed at the test
fixture but at fires around the fixture. A total of 94 lb of agent was discharged while requiring
64.7 lb to extinguish the fixture. The Test 4 fire was extinguished in 20 seconds with a discharge
time of 21 seconds. A total of 132 lb of agent was discharged, requiring 125.8 lb to extinguish
the text fixture. On average, a discharge of 13.7 seconds and 85.9 lb of Novec 1230 was required
to extinguish the fires.

The research team conducted additional tests on this fixture using Halotron I due to the
differences in experimental configurations from the tests conducted in DOT/FAA/AR-95/87. A
total of four tests were conducted overall, with three tests being considered official. The results
for the official tests are listed in table 9. Test 1 was not considered an official test due to a
clogging issue with the fuel-dispensing pipe on the test fixture.

Table 9. Results for Halotron I Three-Dimensional Inclined-Plane Fire Test

Average Maximum Extinguishment Discharge Agent Agent Flow

Wind Wind Time Time Used Required Rate
Test Date (mph) (mph) (s) (s) (lb) (lb) (lb/s)
2 10/23/2017 0 1.6 11 13 72 60.9 5.54
3 10/23/2017 0 0.7 10 12 68 56.7 5.67
4 10/23/2017 2.5 3.8 18 19 106 100.4 5.58
Average 13 72.7

The Test 2 fire was extinguished in 11 seconds with a discharge time of 13 seconds. A total of 72
lb of agent was discharged with 60.9 lb required to extinguish the fire. The Test 3 fire was
extinguished in 10 seconds with a discharge time of 12 seconds. A total of 68 lb of agent was
discharged, of which 56.7 lb were necessary to extinguish the fire. The Test 4 fire was
extinguished in 18 seconds with a discharge time of 19 seconds. A total of 106 lb of agent was
discharged, of which 100.6 lb was necessary to extinguish the fire. For the three official inclined-
plane tests conducted at the WJHTC, Halotron I took 13 seconds and required 72.7 lb of agent on
average to extinguish the fires.

As shown in tables 8 and 9, Test 4 required a relatively drastic increase in agent required by
weight. Novec 1230 required 1.9 times the amount of agent as the average of Tests 1 and 3.
Halotron I required 1.7 times the agent by weight as compared to the average from Tests 2 and 3.

25
The research team attributed this increase to the increased wind velocity present in both tests.
Unlike the other official tests in both test series, the average wind speed for Test 4 of each agent
was greater than the measured peak winds in previous tests, but it still was below the threshold
for this experimental configuration. Additionally, the wind direction in both tests was partially
directed towards the firefighter, as shown from the smoke in figures 19 and 20. Due to the wind
direction and the inability to adjust the test fixture, the firefighter was required to approach the
test fixture and execute extinguishing procedures from a partially downwind direction in both
tests. This means that smoke, gaseous agent, and extinguishment byproducts were being
continually blown at the firefighter, possibly causing visual obstructions. This may have resulted
in inefficient agent application, agent being carried away by the wind instead of toward the fire,
and likely increased oxygen supply to allow the fuel to burn more rapidly. This combination of
effects is thought to be the cause of the increased agent required in both tests. Due to the
presence of comparable test conditions, peak winds below the specified threshold, and a similar
increase of agent required, the research team deemed Test 4 of each agent to be representative of
agent performance and considered them both to be official tests.

Figure 19. Novec 1230 Three-Dimensional Inclined-Plane Fire Test 4

26
 Figure 20. Halotron I Three-Dimensional Inclined-Plane Fire Test 4

Table 10 provides a summary of the results for both Novec 1230 and Halotron I on the three-
dimensional inclined-plane flowing fuel fires. The values provided in the table are the average
results of both agents, and all difference values listed reference the results of the Halotron I
testing as the baseline. On average, Halotron I performed marginally better than Novec 1230 in
this experimental configuration. Overall, Novec 1230 took 0.7 seconds longer (a 5.4% increase)
and required 13.2 lb more agent by weight (an 18.2% increase) than Halotron I on average to
extinguish the test fixture. When considering agent required by volume, Novec 1230 required
0.072 ft3 more agent than Halotron I, a difference of 9.1%. As previously discussed, in tests with
sustained winds over 1 mph, both agents required significantly more agent to extinguish the fire.

Table 10. Agent Performance Comparison for the Three-Dimensional Inclined-Plane Fire Tests

Extinguishment Agent Estimated Time Weight Volume

Time Required Volume Difference Difference Difference
Agent (s) (lb) (ft3) (%) (%) (%)
Halotron I 13.0 72.7 0.788 0 0 0
Novec 1230 13.7 85.9 0.86 5.4 18.2 9.1

5.3 THE 16-FT PAN FIRE TEST RESULTS

A total of three 16-ft pan fire tests were conducted using Novec 1230; all tests were considered
to be official. Table 11 provides these test results.

27
 Table 11. Results of the Novec 1230 16-ft Pan Fire Tests

Average Maximum Extinguishment Discharge Agent Agent Flow

Wind Wind Time Time Used Required Rate
Test Date (mph) (mph) (s) (s) (lb) (lb) (lb/s)
1 3/21/2017 2.0 3.6 12 16 106.0 79.5 6.63
2 3/21/2017 2.0 2.9 10 14 91.0 65.0 6.50
3 3/28/2017 3.6 6.7 16 19 107.0 90.1 5.63
Average 12.7 78.2

Test 1 had a total discharge time of 16 seconds, and the fire was extinguished in 12 seconds. A
total of 106 lb was discharged, of which 79.5 lb was required to extinguish the fire. Test 2 had a
total of 14 seconds of agent discharge, and the fire was extinguished within 10 seconds. A total
of 91 lb of agent was discharged, of which 65 lb was required to extinguish the fire. Test 3 had a
total discharge time of 19 seconds, and the fire was extinguished in 16 seconds. During Test 3, a
total of 107 lb of agent was discharged, of which 90.1 lb was required to extinguish the fire. On
average, Novec 1230 extinguished the 16-ft pan fires in 12.7 seconds and required 78.2 lb of
agent.

Due to the differences between experimental configurations between the 16-ft pan fire test and
the dry-pool fires of DOT/FAA/AR-95/87, the research team conducted additional tests using
Halotron I. The most notable differences between testing configurations was the amount of fuel
and substrate on which the fuel was distributed. A total of three 16-ft pan fire tests were
conducted using Halotron I; all tests were considered official. Table 12 shows the results of these
tests.

Table 12. Results of the Halotron I 16-ft Pan Fire Tests

Average Maximum Extinguishment Discharge Agent Agent Flow

Wind Wind Time Time Used Required Rate
Test Date (mph) (mph) (s) (s) (lb) (lb) (lb/sec)
1 8/14/2017 0.0 2.2 11 12 67 61.4 5.58
2 8/14/2017 3.4 6.7 13 14 75 69.6 5.36
3 8/16/2017 2.2 3.8 12 13 68 62.8 5.23
Average 12 64.6

Test 1 had a total discharge time of 12 seconds, taking 11 seconds to extinguish the fire. A total
of 67 lb of agent was discharged, requiring 61.4 lb to extinguish the fire. Test 2 had a total
discharge time of 14 seconds and an extinguishment time of 13 seconds. A total of 75 lb of agent
was discharged, of which 69.6 lb was required to extinguish the fire. Test 3 had a total discharge
time of 13 seconds and an extinguishment time of 12 seconds. A total of 68 lb of agent was
discharged, of which 62.8 lb was required to extinguish the fire. On average Halotron I
extinguished the 16-ft pan fires in 12 seconds and required 64.6 lb of agent.

A summation and comparison of the 16-ft pan fire tests results of Novec 1230 and Halotron I are
provided in table 13. The values provided in the table are the average results of both agents, and
all difference (i.e., time, weight, and volume) values listed are based on the results of the
Halotron I tests as the baseline. A comparison of results from the 16-ft pan fire tests show

28
Halotron I performed marginally better than Novec 1230. Novec took an average of 0.7 second
longer to extinguisher, an increase of 5.8%. Novec required an additional 13.6 lb of agent on
average to extinguish the fire, an increase of 21.1% as compared to Halotron I. When compared
to the volume of agent discharged, Novec 1230 required approximately 0.083 ft3 more than
Halotron I, an increase of 11.8%.

Table 13. Agent Performance Comparison for the 16-ft Pan Fire Tests

Extinguishment Agent Estimated Time Weight Volume

Time Required Volume Difference Difference Difference
Agent (s) (lb) (ft3) (%) (%) (%)
Halotron I 12.0 64.6 0.7 0 0 0
Novec 1230 12.7 78.2 0.783 5.8 21.1 11.8

5.4 THE 30-FT PAN FIRE TEST RESULTS

The research team conducted a total of three 30-ft pan fire tests using Novec 1230; all tests were
considered official. Table 14 shows the total discharge times and amount of agent discharged for
each 30-ft pan fire test. Since none of the 30-ft pan fires were successfully extinguished with
Novec 1230, values for the amount of time and agent required to extinguish the fires could not be
determined. Additionally, values for the times when the discharge began to sputter and reached
the gas point are provided in the table. The time to sputter is defined as the time when the agent
discharge stream begins to change from a continuous flow of liquid agent to bursts of liquid
agent separated by gaseous discharge. These gaseous bursts start slow and become more rapid
until the gas point is reached. The gas point is defined as the time when the discharge is only
propellant gas and gaseous agent. These values are also listed in table 14 in addition to total
discharge time. The flow rate values for all tests conducted in this experimental configuration are
based on the gas point time instead of the total discharge time. Although the flow rate is likely
reduced during the time the discharge is sputtering, the effects are assumed to be minimal, and
therefore negligible. This assumption is corroborated by the discharge stream consisting of
primarily liquid agent and no significant reduction in extinguishing capabilities was observed
during this period of time. Liquid agent may still be able to be discharged after the gas point has
been reached; however, it is not discharged in a continuous stream, and extinguishing
capabilities are vastly diminished. The time to sputter has been provided as both a reference
point in each test and a characterization of the discharge stream at the point the fire was most
diminished.

Table 14. Results for Novec 1230 30-ft Pan Fire Test

Average Maximum Discharge Time to Time to Agent Flow

Wind Wind Time Sputter Gas Point Used Rate
Test Date (mph) (mph) (s) (s) (s) (lb) (lb/s)
1 3/28/2017 2.9 5.8 33 17 23 150 6.52
2 4/05/2017 2.2 6.9 27 20 25 145 5.8
3 4/13/2017 2.7 2.7 23 18 22 140 6.36
Average 23.3 145

29
During Test 1 of the Novec 1230 30-ft pan fire, the agent discharge began to sputter 17 seconds
after discharge began, and the extinguisher reached its gas point after 23 seconds. In total, the
discharge time was 33 seconds and 150 lb of agent was discharged. Figure 21(a) shows the point
when the least amount of fire remained in the test fixture. It should be noted that this was the
only 30-ft pan fire test that the firefighter did not step into the pan prior to the gas point. The
firefighter did advance into the pan and closer to the remaining fire immediately following the
gas point; however, no further progress was made, and the fire continued to grow during this
time. Figure 21(b) shows the point immediately following the end of discharge.

(a)

(b)

Figure 21. Test 1: Novec 1230 30-ft Pan Fire

Test 2 of the Novec 1230 30-ft pan fire had similar results to Test 1, with the fire becoming
greatly diminished but failing to be extinguished. Test 2 had a total discharge time of 27 seconds
and a gas point of 25 seconds. Test 2 had the longest discharge time prior to beginning to sputter.
During Test 2, the firefighter knocked down a greater portion of the fire and made entry into the
pan significantly quicker than during Test 1. The firefighter stepped into the pan just as the
discharge began to sputter. The firefighter experienced difficulty while attempting to extinguish
the fire within inches of the back edge of the pan. The firefighter directed the agent stream in
sweeping motions from one side of the remaining fire to the other, closely following the
perimeter of the pan. As the firefighter directed the agent stream back and forth across the fire,
the fuel reignited prior to completing a single sweep. An example of this is shown in figure
22(a), where the firefighter has just completed a left-to-right sweep but the fuel on the left side
has already reignited. This is also the point when the least amount of fire remained in the pan
during Test 2. This continued until the firefighter ceased discharging the extinguisher. The
firefighter continued discharging the extinguisher for 2 seconds after reaching the gas point,

30
during which the fire grew rapidly. Figure 22(b) shows the remaining fire immediately following
the end of discharge. A total of 145 lb of agent was discharged during Test 2.

(a)

(b)

Figure 22. Test 2: Novec 1230 30-ft Pan Fire

Novec 1230 failed to extinguish the 30-ft pan fire in Test 3 as well. The discharge from the
extinguisher began to sputter 18 seconds after discharge began, and reached the gas point
4 seconds later, a total of 22 seconds after discharge began. The firefighter ended the discharge
1 second after the extinguisher reached its gas point for a total discharge time of 23 seconds. In
total, 140 lb of agent was discharged during this test. Although the team did not perform a
quantitative analysis, a qualitative analysis of video footage reveals a significantly lower
percentage of the pan was extinguished in Test 3. The point when the fire was diminished
furthest is shown in figure 23(a). The firefighter had the greatest difficulty extinguishing the
leading edge of the fire during Test 3, causing rapid fuel reignition from all directions and
preventing advancement into the pan. The firefighter was able to extinguish a portion of the fire
eventually, allowing for advancement into the pan 1 second after the discharge had begun to
sputter and also allowing a greater portion of the fire to be knocked down. Unfortunately, by the
time any significant progress was made, the extinguisher had begun sputtering and, shortly after,
reached the gas point. The firefighter ended the discharge within 1 second of the extinguisher’s

31
gas point and exited the pan. Figure 23(b) shows the fire remaining in the pan immediately after
the discharge ended.

(a)

(b)

Figure 23. Test 3: Novec 1230 30-ft Pan Fire

Although all attempts were made to keep all variables constant between tests (such as the use of
the same firefighter using the same agent application technique from an upwind direction), the
team later found minor discrepancies through in-depth review of the video footage from each
evaluation. One such discrepancy the team found during the footage review from Test 3 was the
distance from the 30-ft pan at which the firefighter began discharging. The firefighter began
discharging agent closer to the pan than in the previous two tests, limiting the amount of agent
discharged on approach to the pan and concentrating the agent that was discharged to a smaller
area. The team considers this discharge distance to be the main varying factor in results from
Tests 1 and 2. While this distance was also observed in other tests, it appeared to have less of an
effect than what was observed upon review of Test 3. Thorough review of the recorded footage
from both current and previous tests showed there was an optimal range from the firefighter
where Novec 1230 exhibited the greatest extinguishing capability. When the agent stream was

32
directed in close proximity to the firefighter, the discharge was too concentrated to effectively
extinguish any meaningful area, causing rapid reignition after the discharge was directed
somewhere else. At greater distances, the agent stream appeared to become too dispersed to
effectively suppress the fire. During Test 1, the firefighter initially attempted to extinguish the
fire without entering the pan. With the exception of the leading edge, the majority of the fire was
quickly knocked down, but the agent stream became ineffective when attempting to extinguish
the fire remaining on the opposite side of the pan. This deviation from discharging agent within
the optimal range is not considered to be the only reason extinguishment was not achieved, but is
considered a large factor. Furthermore, these fire tests were conducted after all the tests
described in previous sections were completed, with the same firefighter used in all tests. This
means that the firefighter had discharged the extinguisher and extinguished many fires preceding
these tests. This shows that even with experience, maintaining the optimal range between the
nozzle and base of the fire is challenging.

The research team also conducted additional 30-ft pan fire tests using Halotron I. As previously
stated, this was due to the changes in experimental configurations between this research effort
and the dry-pool fires described in DOT/FAA/AR-95/87. The team conducted a total of three
30-ft pan fire tests with Halotron I; all tests were considered official. Halotron I also failed to
extinguish any of the 30-ft pan fires. The test results are provided in table 15, which presents the
equivalent data from the Halotron I tests that was provided with the Novec 1230 tests,
maintaining the same definitions and methods of calculations.

Table 15. Results for the Halotron I 30-ft Pan Fire Tests

Time
Average Maximum Discharge to Time to Flow
Wind Wind Time Sputter Gas Point Agent Used Rate
Test Date (mph) (mph) (s) (s) (s) (lb) (lb/s)
1 8/16/2017 3.1 4.9 36 22 31 153 4.94
2 10/3/2017 0 0.9 36 25 34 149 4.38
3 10/3/2017 0 3.4 33 23 30 152 5.07
Average 31.6 151.3

During Test 1 of the 30-ft pan fires with Halotron I, the extinguisher began to sputter after
22 seconds, reached the gas point after 31 seconds, and was discharged for a total of 36 seconds.
A total of 153 lb of agent was discharged. Test 1 also had the highest winds of the three tests,
and was the only test with a sustained wind. The firefighter extinguished the bulk of the fire from
outside of the pan, advancing into the pan approximately 4 seconds after the extinguisher began
to sputter. The time when the firefighter began advancing into the pan was also the same time
when the fire had been diminished the most, as shown in figure 24(a). As shown in the previous
Novec 1230 tests, the fire persisted around the perimeter of the opposite edge of the pan.
Ultimately, the fire was unable to be extinguished, with the firefighter ending discharge
5 seconds after the gas point. The fire continued to grow after the gas point had been reached;
however, the firefighter continued to direct the miniscule pockets of liquid agent being
discharged at the fire. The size of the fire immediately following the end of discharge is shown in
figure 24(b).

33
 (a)

(b)

Figure 24. Test 1: Halotron I 30-ft Pan Fire

Test 2 of the 30-ft pan fire tests with Halotron I failed to extinguish the fire, with fire persisting
around the perimeter of the pan opposite the firefighter. The extinguisher began to sputter after
25 seconds, reached the gas point after 34 seconds, and the discharge was ended after
36 seconds. A total of 149 lb of agent was discharged during this test. Unlike the Test 1 with
Halotron I, the firefighter was able to extinguish the fire around the perimeter of the pan furthest
away by advancing into the pan slightly more rapidly, approximately 3 seconds before sputtering
began. Even though fire remained on the top lip of the pan in this area, it did not reignite the fuel
in the pan during the discharge period. As the firefighter was extinguishing the fire most distant
from the initial discharge point, the fuel on the right side of the pan reignited. The firefighter
then redirected the discharge towards the reignited fuel. At this point in the test, the fire had been
diminished furthest, as shown in figure 25(a). As the firefighter attempted to extinguish the
remaining fire, the extinguisher quickly reached its gas point. The firefighter then ended the
discharge and exited the pan, as shown in figure 25(b).

34
 (a)

(b)

Figure 25. Test 2: Halotron I 30-ft Pan Fire

Test 3 of the 30-ft pan fire tests with Halotron I had similar results to the previous two tests.
During Test 3, the extinguisher began to sputter 23 seconds after discharge began, reached its gas
point after 30 seconds, and had a total discharge period of 33 seconds. A total of 152 lb of agent
was discharged during this test. The firefighter was able to extinguish a significant portion of the
pan, although it took slightly longer than the other two tests. This was due to fire remaining
around the entire perimeter of the pan as the fire in the middle of the pan began to be
extinguished. This required the firefighter to redirect the agent around the edge of the pan closest
to his position prior to advancing into the pan. The extinguisher began sputtering 1 second after
the firefighter advanced into the pan. The fire in the pan was diminished furthest as the
extinguisher reached its gas point, as shown in figure 26(a). Although there appears to be a large
amount of fire remaining, the bulk of it is located outside of the test fixture. The firefighter

35
continued discharging the extinguisher for 3 seconds after reaching the gas point. The firefighter
then ended the discharge and exited the pan, as shown in figure 26(b).

(a)

(b)

Figure 26. Test 3: Halotron I 30-ft Pan Fire

Overall, neither Novec 1230 nor Halotron I successfully extinguished the fires in the 30-ft pan
fire tests, which means a quantitative analysis of performance could not be performed on these
tests. The team made the following observations regarding test data:

• With the exception of Test 3 with Novec 1230, each fire was able to be diminished to a
small area around the perimeter of the pan regardless of agent used.

• Neither agent provided any protection against reignition.

• Both agents exhibited an optimal range in which the discharge was most effective.

36
• Neither agent was able to extinguish fire on the opposite side of the pan without the
firefighter advancing into the pan.

5.5 SIMULATED ENGINE NACELLE RUNNING FUEL FIRE TEST RESULTS

Detailed in AFCEC-CX-TY-TR-2014-0033, previous tests evaluated the extinguishing

capabilities of Novec 1230 discharged from an Amerex Model 775 extinguisher on a simulated
engine nacelle running fuel fire. The research team considered these evaluations to be
satisfactory representations of this research effort’s requirements and deemed it unnecessary to
repeat them due to the availability of an accurate data set. The team accepted a portion of this
data as official results for this research effort and used the data for the following analysis. Table
16 shows the test results that were accepted and considered official results. Of the 11
evaluations, the team accepted 7 as official tests. The team excluded Tests 1 and 8 due to the lack
of a reported discharge time, excluded Test 3 due to a structural defect with the test fixture, and
excluded Test 4 due to a failure to extinguish the fire. Only the results in table 16 will be used for
all analyses. [6]

Table 16. Novec 1230 Tests Results From AFCEC-CX-TY-TR-2014-0033 [6]

Wind Extinguishment Discharge Agent Agent Discharge

Speed Time Time Used Required Rate
Test (mph) (s) (s) (lb) (lb) (lb/s)
2 0.0 20 25 132 105.6 5.28
5 6.2 22 21 122 122* 5.81
6 6.7 22 22 139 139.0 6.32
7 2.6 20 20 138 138.0 6.90
9 5.0 21 22 139 132.7 6.32
10 2.9 18 19 121 114.6 6.37
11 1.1 18 20 124 111.6 6.20
Average 20.1 124.2
*Value based only on total agent discharge

The fires accepted as official results required an average of 124.2 lb of agent and 20.1 seconds to
be extinguished. The value listed for agent required in Test 5 was not calculated based on the
extinguishment time and discharge rate; instead, the overall agent used is listed. This is due to
Test 5 having a longer extinguishment time than discharge time.

Halotron I’s extinguishing capabilities also have been evaluated previously on simulated engine
nacelle running fuel fire tests using the F100 nacelle test fixture. These evaluations are detailed
in NAWCWD TM 8572. [7] The report evaluated two types of Halotron I extinguishers of
equivalent capacities, the Buckeye W-150 and the Amerex Model 674. To maintain consistency
through any comparisons, the research team used only the Amerex Model 674 test results as
official results and in the following analysis. These tests used JP-8 fuel for the fires, using the
same engine nacelle mockup and procedures as previously described tests. A total of 11 Amerex
extinguisher tests were conducted, of which 3 fires were successfully extinguished. Only the data

37
from the successful tests were accepted as official results and used in any analysis or
comparison. The accepted test results are given in table 17. The agent required values listed in
table 17 were determined in NAWCWD TM 8572, unlike the previous agent required values.
The values in table 17 are likely to be more accurate due to being based on a dynamic flow rate
instead of averaged across the whole discharge period. It should be noted that the largest
difference between values obtained between the two methods for this set of test results was
approximately 3% with a 0.5% average difference. Additionally, the DOT/FAA/AR-95/87
results for the simulated engine nacelle running fuel fires with Halotron I were not used for any
analysis or comparisons. This is due to the use of an updated F100 nacelle fixture and the agent
being discharged from an optimized extinguisher. The results from NAWCWD TM 8572 were
deemed to be more representative of the agent’s performance using the same test fixture and
procedures that Novec 1230 was evaluated with. [7]

Table 17. The NAWCWD TM 8572 Test Results for Halotron I With Amerex
Extinguisher 674 [7]

Wind Extinguishment Discharge Agent Agent

Speed Time Time Used Required Flow Rate
Test (mph) (s) (s) (lb) (lb) (lb/s)
1 0-3 44 45 136 135 3.02
2 0-6 17 20 89 73 4.45
11 5-8 26 34 131 100 3.85
Average 29 102.7

For the accepted NAWCWD TM 8572 tests, the average extinguishment time was 29 seconds
and required an average of 102.7 lb of Halotron I to extinguish the fires [7]. Compared to Novec
1230, Halotron I required an additional 8.9 seconds (44.3%) to extinguish the fires, but did so
using 21.5 lb (17.3%) less agent.

A summary of the average results for both Halotron I and Novec 1230 on the simulated engine
nacelle flowing fuel fire tests is presented in table 18. The estimated volume of agent discharged
was based on the saturated liquid densities of each agent. All values listed for the differences
between the results are referenced to the Halotron I values. While not part of the following
analysis, it is important to note that Halotron I extinguished 3 out of the 11 nacelle fires, and
Novec 1230 extinguished 9 out of 11 fires.

Table 18. Summary of Results From Simulated Engine Nacelle Running Fuel Fire Tests

Extinguishment Agent Estimated Time Weight Volume

Time Required Volume Difference Difference Difference
Agent (s) (lb) (ft3) (%) (%) (%)
Halotron I 29 102.7 1.113 0 0 0
Novec 1230 20.1 124.2 1.243 -30.7 20.9 11.7

As shown in table 18, Novec 1230 extinguished the fires 30.7% faster than Halotron I. However,
Novec 1230 required more agent by weight and volume than Halotron I. By weight, Novec 1230

38
required 21.5 lb more agent than Halotron I, an increase of 20.9%. By volume, Novec 1230
required 0.13 ft3 more agent on average than Halotron I, an increase of 11.7%.

5.6 DROP-IN VALIDATION FOR TRUCK-BASED SYSTEMS

During this research effort, the team conducted further performance evaluations of the
Novec 1230 on the simulated engine nacelle running fuel fire tests. In each of the previously
described evaluations, the Flight Line extinguisher had been optimized to maximize the agent
performance. However, these evaluations were conducted to validate that the agent’s
performance was not affected when being discharged from a truck-based system, which had not
been optimized for the agent. This means that no modifications were made to the system in
preparation of discharging a new agent. The only change made to the system’s configuration was
the propellant gas. Novec 1230 requires nitrogen to be used as the propellant gas instead of
argon, which is used for Halotron I. The regulator was not changed, and it was set to the
system’s factory specification.

In total, the research team performed three official simulated engine nacelle running fuel fire
tests, discharging Novec 1230 from a hand line supplied by a truck-based system. The test results
and average results of all tests are given in table 19.

Table 19. Results for Simulated Engine Nacelle Running Fuel Fire Test With Novec 1230

Average Maximum Extinguishment Discharge Agent Agent Flow

Wind Wind Time Time Used Required Rate
Test Date (mph) (mph) (s) (s) (lb) (lb) (lb/s)
1 5/26/2017 0.0 1.0 15 20 161.6 121.2 8.08
2 5/26/2017 1.5 1.5 13 16 140.6 114.3 8.79
3 5/26/2017 2.0 4.0 11 14 123.3 96.9 8.81
Average 13 110.8

All three attempts resulted in successful extinguishment of the fire. Although this was an
expected result due to the increased flow rate and agent tank capacity of the truck-based system,
the team performed these tests to ensure system optimization was not required to maintain the
agent’s extinguishing capabilities. Test 1 had an extinguishment time of 15 seconds, with a
discharge time of 20 seconds. A total of 161.6 lb of agent was used, of which 121.2 lb was
required for extinguishment. For Test 2, the extinguishment time was 13 seconds, with a
discharge time of 16 seconds. A total of 140.6 lb of agent was used, of which 114.3 lb was
necessary to extinguish the fire. Test 3 was extinguished in 11 seconds with a total discharge
time of 14 seconds. A total of 123.3 lb of agent was used, of which 96.9 lb was necessary to
extinguish the fire. On average, the fires were extinguished in 13 seconds and required 110.8 lb
of agent to extinguish the fire.

It should be noted that the same firefighter was used for all three tests, which were conducted in
relatively rapid succession. As shown by the discharge and extinguishment times between tests,
the firefighter’s proficiency increased between each test. The firefighter who performed these
tests has fought many fires using the engine nacelle test fixture and has previous experience
discharging Novec 1230 from an extinguisher. Even with this prior experience, significant
reduction in extinguishment times and the amount of agent discharged was observed between

39
tests. While further testing was not performed, the extinguishment times and agent required
values were not expected to be further reduced significantly.

A summary of the results from the simulated engine nacelle with running fuel fire tests with
Novec 1230 is provided in table 20. These results are the overall averages from both series of
tests that used Novec 1230. The values listed for differences between results references the tests
that were conducted with an extinguisher.

Table 20. Summary of Results for Simulated Engine Running Fuel Fire Tests With Novec 1230

Extinguishment Agent Flow Time Weight Flow Rate

Time Required Rate Difference Difference Difference
Discharge Type (s) (lb) (lb/s) (%) (%) (%)
Extinguisher 20.1 124.2 6.17 0 0 0
Hand Line 13 110.8 8.56 -40.3% -10.8% 38.7%

As shown in table 20, an increase in flow rate of 38.7% reduced extinguishment times by an
average of 40.3%. The amount of agent required to extinguish the fires was also reduced by
10.8% when discharged through a hand line. Although there was an average reduction in agent
required, the overall amounts are comparable. Overall, there is no indication that the Novec 1230
extinguishing capabilities are hindered when discharged from an unoptimized truck-based
system. In fact, all comparable results indicate the performance of the agent may be enhanced
when discharged at a higher rate.

The team performed additional analysis on the flow rate of the agent discharged through the
hand line. The team calculated the flow rate from the weight data obtained from the load cells on
the Fire Combat skid unit and recorded through a custom LabVIEW program. This was done to
ensure that the agent met the NFPA 414-required minimum-specified flow rate of 5 lb/s [12].
The flow rates in each test exceeded the minimum acceptable flow rate and remained above that
level for the entire duration of each discharge. Between the three tests, the average discharge rate
was 8.56 lb/s. A graph of the flow rate over the entire discharge period for each test is shown in
figure 27. A rolling average filter was applied to the data in an attempt to reduce the variance
between measurements due to signal noise.

40
 10
9
8
7
Flow Rate (lbs/s) 6
5 Test 1
4 Test 2
3 Test 3
2
1
0
0 10 20 30 40 50
Time (s)

Figure 27. Average Flow Rate for Simulated Engine Nacelle Running Fuel Fire Tests

5.7 OVERALL AGENT COMPARISON

Table 21 provides a summary of Novec 1230’s performance compared to that of Halotron I

across all equivalent tests. Throughout all experimental configurations, Novec 1230 required
more agent by both weight and volume than Halotron I. Novec 1230 required an average of
20.1% by weight and 10.9% by volume more agent than Halotron I in equivalent tests. The
difference between performance parameters by percentage were consistent between the three
experimental configurations, with the exception of the simulated engine nacelle running fuel fire
tests conducted by the USAF and U.S. Navy. The agent required results of the simulated engine
nacelle tests were consistent with the results of the other experimental configurations with
Novec 1230 requiring more agent by weight and volume than Halotron I. However, Novec 1230
extinguished the simulated engine nacelle fire tests notably faster than Halotron I. Including the
simulated engine nacelle fires, Novec 1230 extinguished the test fires 19.5% faster than
Halotron I. When disregarding the results of the simulated engine nacelle tests, Novec 1230 took
an average of 5.6% longer than Halotron I to extinguish the test fires.

Table 21. Summary of Novec 1230 Performance by Experimental Configuration

Time Weight Volume

Difference Difference Difference
Test Type (%) (%) (%)
Inclined Plane 5.4 18.2 9.1
16-ft Pan 5.8 21.1 11.8
Engine Nacelle -30.7 20.9 11.7

41
6. CONCLUSIONS

For the simulated wheel brake fire involving hydraulic fluid tests, Novec™ 1230 required an
average of 14.3 seconds and 78.2 pounds (lb) of agent to extinguish the fires. Since the research
team used a larger tire and offset approach path, a significant portion of the test fixture was
obstructed, which required the firefighter to reposition around the test fixture. In each test, the
fire around the test fixture that was not obstructed on the firefighter’s initial approach was
extinguished in 4 to 6 seconds. In each test, the firefighter spent the remainder of the time
repositioning around the fixture and applying agent to the rest of the fire. Therefore, comparing
the results from DOT/FAA/AR-95/87 for the simulated wheel brake fire with Halotron® I would
not produce a valid conclusion of either agent’s extinguishing capabilities. Overall, Novec 1230
extinguished all three official simulated wheel brake fire tests. The results show the
extinguishment time and agent required were reduced in each successive test. The simulated
wheel brake fire test in which Novec 1230 performed best extinguished the fire 23% faster with
18% less agent than the average of all three tests.

In the three-dimensional inclined-plane tests, Novec 1230 took an average of 13.7 seconds and
required 85.9 lb of agent to extinguish the fires. Halotron I took 13 seconds and required 72.7 lb
of agent to extinguish the fires. Comparing these results, Novec 1230 took 0.7 seconds longer
and required 13.2 lb more agent to extinguish the fires, increases of 5.4% and 18%, respectively.
When considering the volume of agent required, Novec 1230 required 0.072 cubic feet (ft3) more
than Halotron I, an increase of 9.1%. Of note for this test series were the effects of a sustained
wind on the performance and application efficiency for both agents. In tests with comparable
sustained winds, Novec 1230 required 90% more agent and Halotron I required 70% more agent
to extinguish the fires than other equivalent tests for each respective agent.

For the 16-ft pan fire tests, Novec 1230 had an average extinguishment time of 12.7 seconds and
required 78.2 lb of agent to extinguish the fires. Halotron I had an average extinguishment time
of 12 seconds and required 64.6 lb of agent to extinguish the fires. Compared to Halotron I,
Novec 1230 took 5.8% more time and required 21.1% more agent by weight to extinguish the
fires. When considering the volume of agent required to extinguish the fires, Novec 1230
required 11.8% more agent to extinguish the fires. Both agents successfully extinguished all of
the 16-ft pan fire tests.

Both Novec 1230 and Halotron I were unable to extinguish any 30-ft pan fires. Although the
extinguishment time and agent quantity required values could not be determined, the team made
a qualitative performance comparison based on the area of fire remaining. In two out of three
Novec 1230 tests, and in all Halotron I tests, an overwhelming majority of the fire was
suppressed with only small areas of fire remaining around portions of the pan perimeter. The
cause for the test result in which Novec 1230 did not suppress the majority of fire in the pan was
attributed to the distance at which the firefighter began discharging. Based on the area of fire that
was suppressed in each test, both agents demonstrated nearly equivalent extinguishing
capabilities. Additionally, these test observations showed that both agents’ discharge streams
have optimal ranges where extinguishment is most effective.

For the simulated engine nacelle running fuel fire tests, data from
AFCEC-CX-TY-TR_2014-0033 were accepted as official tests for Novec 1230, and data from

42
NAWCWD TM 8572 were used as official tests for Halotron I. On average, Novec 1230
required 124.2 lb of agent and 20.1 seconds to achieve extinguishment and Halotron I, required
an average of 102.7 lb of agent and 29 seconds. This means that although Novec 1230
extinguished the fires 30.7% faster than Halotron I, it required 20.9% more agent by weight to do
so. By volume, Novec 1230 required 0.13 ft3 more agent than Halotron I, an increase of 11.7%.
Of note was the success rate of both agents, with Novec 1230 extinguishing 9 of 11 fires and
Halotron I extinguishing 3 of 11 fires.

The research team also used simulated engine nacelle with running fuel fire tests to evaluate the
performance of Novec 1230 when used as a drop-in replacement in a truck-based system. On
average, the agent flow rate increased 38.7% when discharged through the hand line of the Fire
Combat Skid system compared to the extinguisher application. Increasing the flow rate reduced
extinguishment times by 40.3%, and also reduced the amount of agent required to extinguish the
fires by 10.8%. These results indicate that Novec 1230’s extinguishing capabilities are not
hindered when being discharged from a truck-based system.

7. REFERENCES

1. Certification of Airports, 14 Code of Federal Regulations (CFR) § 139. (2012).

Washington, DC: Government Printing Office.

2. Federal Aviation Administration (FAA). (1995, October). Full-scale evaluations of

Halon 1211 replacement agents for airport fire fighting (DOT/FAA/AR-95/87).
Washington, DC: FAA.

3. U.S. Environmental Protection Agency. (2018, July 31). The accelerated phaseout of
Class I ozone-depleting substances. Retrieved from https://www.epa.gov/ods-
phaseout/accelerated-phaseout-class-i-ozone-depleting-substances (date last visited
06/7/19).

4. AMPAC™ HALOTRON. (2012, October). Halotron® I clean agent HCFC

phase-out questions, just the facts. Retrieved from
http://halotron.com/pdf/Halotron_JUST_THE_FACTS.pdf (date last visited 06/21/19).

5. U.S. Environmental Protection Agency. (2017, November 7). Substitutes in streaming

agents. Retrieved from https://www.epa.gov/snap/substitutes-streaming-agents (date last
visited 06/7/19).

6. Air Force Civil Engineering Center (AFCEC). (2014, November). Evaluation of the
Amerex model 775 wheeled extinguisher with Novec 1230, (AFCEC-CX-TY-TR-2014-
0033). Retrieved from https://apps.dtic.mil/dtic/tr/fulltext/u2/a613087.pdf (date last
visited 06/21/19).

7. Naval Air Warfare Center Weapons Division (NAWCWD). (2018, September).

Qualification of an acceptable alternative to Halon 1211 DOD flightline extinguishers,
(NAWCWD TM 8572). Retrieved from https://apps.dtic.mil/dtic/tr/fulltext/u2/
a605516.pdf (date last visited 06/21/19).

43
8. 3M. (2018, January). 3M™ Novec™ 1230 fire protection fluid [Technical data].
Retrieved from http://multimedia.3m.com/mws/media/124688O/3mtm-novectm-1230-
fire-protection-fluid.pdf?fn=prodinfo_novec1230.pdf (date last visited 04/15/19).

9. NFPA, Inc. (2017, April 3). Physical properties comparison, Halotron I and Halon 1211.
Retrieved from http://www.nfpa.it/halotron_referen1.htm (date last visited 04/15/19).

10. 3M Specialty Materials Division. (n.d.). A continuation of thermal decomposition product

testing with C6 F-ketone, increased concentrations. Retrieved from
https://www.nist.gov/sites/default/files/documents/el/fire_research/R0201302.pdf (date
last visited 06/21/19).

11. Ditch, B.D. (2002). Thermal decomposition products testing with 1,1,1,2,2,4,5,5,5
nonafluoro-4-trifluoromethyl pentan-3-one (C6 F-ketone) during fire extinguishing
(Master’s thesis). Retrieved from https://web.wpi.edu/Pubs/ETD/Available/etd-0106103-
152708/unrestricted/bdditch.pdf (date last visited 06/21/19).

12. National Fire Protection Association (NFPA). (2017, January). Standard for aircraft
rescue and firefighting vehicles. NFPA 414. Quincy, MA: NFPA.

13. Coordinating Research Council, Inc. (CRC). (1983). Handbook of aviation fuel
properties, (CRC report No. 530). Retrieved from https://apps.dtic.mil/dtic/tr/
fulltext/u2/a132106.pdf (date last visited 06/21/19).

14. Chevron Phillips Chemical Company LP. (2018, June 7). Jet-A aviation fuel, (Safety
Data Sheet Number 100000014588, version 2.3).

15. U.S. Department of Health and Human Services (USDHHS), Public Health Service,
Agency for Toxic Substances and Disease Registry. (1995, June). Toxicological profile
for jet fuels JP-4 and JP-7. Atlanta, GA: USDHHS.

16. DiNenno, P.J. et al. (Eds.). (2002). SFPE Handbook of Fire Protection Engineering (3rd
ed.). Gaithersburg, MD: Society of Fire Protection Engineers.

17. FAA. (1968, December). Foam and dry chemical application experiments,
(Interim Report NA-68-34 [RD-68-55]). Retrieved from
https://apps.dtic.mil/dtic/tr/fulltext/u2/680068.pdf (date last visited 06/21/19).

18. Department of Defense (DoD). (1980, August 29). Military Specification (MIL)-H-
5606F, Hydraulic fluid, petroleum base; aircraft, missile, and ordnance
[SUPERSEDING MIL-H-5606E] [S/S by MIL-PRF-5606H]. Washington, DC: DoD.

19. DoD. (1957, February 27). MIL-H-5606A, Oil, hydraulic; aircraft, petroleum base
[SUPERSEDING MIL-O-5606] [S/S BY MIL-PRF-5606H]. Washington DC: DoD.

20. Kestrel Meters. [Photograph]. Retrieved from https://kestrelmeters.com/products/kestrel-

5000-environmental-meter (date last visited 06/21/19).

44
21. The Goodyear Tire and Rubber Company. (2018, June). Global aviation tires. . Retrieved
from https://www.goodyearaviation.com/resources/pdf/databook-6-2018.pdf (date last
visited 06/7/19).

22. Fire Combat. (2014, February 3). Fire Suppression System Model 13624 460 lb. Halotron
I Skid System. Marinette, WI: Fire Combat.

45/46