EPA-600/R-9 5-150b
vSeptember 1985
LOW OZONE-DEPLETING HALOCARBONS AS TOTAL-FLOOD AGENTS:
VOLUME 2—LABORATORY-SCALE FIRE SUPPRESSION AND
EXPLOSION PREVENTION TESTING
by
Stephanie R. Skaggs, Everett W. Heinonen, Ted A. Moore, and Jon A. Kirst
Center for Global Environmental Technologies
New Mexico Engineering Research Institute
The University of New Mexico
Albuquerque, New Mexico 87131-1376
Cooperative Agreement CR-817774-01-0
Project Officer:
N. Dean Smith
Air Pollution Prevention and Control Division
National Risk Management Research laboratory
U. S. Environmental Protection Agency-
Research Triangle Park, North Carolina 27711
Prepared for:
U. S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
WASHINGTON, DC 20460
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before comp
1. REPORT NO. 2.
EPA-600/R-95-l50b
4. TITLE AND SUBTITLE
Low Ozone-Depleting Halocarbons as Total-Flood
Agents: Volume 2-~Eaboratory-Scale Fire Suppres-
sion and Explosion Prevention Testing
5. REPORT DATE
September 1995
6. PERFORMING ORGANIZATION CODE
7. AUTHORISI
S. R. Skaggs, E. W. Heinonen, T.A.Moore, and
J. A. Kirst
8. PERFORMING ORGANIZATION REPORT NO.
NMERI OC 92/26
9. PERFORMING ORGANIZATION NAME AND ADDRESS
The University of New Mexico
Mew Mexico Engineering Research Institute
Albuquerque, New Mexico 87131-1376
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
CR 817774-01-0
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 10/90-3/93
14. SPONSORING AGENCY CODE
EPA/600/13
15.supplementary notes APPCD projec t officer is IS. Dean Smith, Mail Drop 62B 918'541-
2708. Volume 1 is "Candidate Survey."
16.abstract repQrt gives results from (l) flame suppression testing of potential
Halon-1301 (CF3Br) replacement chemicals in a laboratory cup burner using n-hep-
tane fuel and (2) explosion prevention (inertion) testing in a small-scale explosion
sphere using propane and methane as fuels. Test equipment and techniques are desc-
ribed. Agent performance is given in terms of the concentration required to achieve
flame extinguishment in the cup burner and as the concentration required to achieve
explosion inertion, defined as an explosive overpressure of 1 psi (6.9 kPa) or less.
Results are also expressed in terms of weight and storage volume equivalents, re-
flecting the weight and storage volume of a candidate agent required to achieve the
same performance effectiveness as Halon-1301. At the time of this project, the NFPA
(National Fire Protection Association) had selected a number of Halon-1301 replace-
ment candidates with potential for use in the near-term. All of these chemicals were
included in the laboratory-scale tests performed in this study. Results of this study
indicate that, of the NFPA candidates, only FC-3-1-10 (CF3CF2CF2CF3) and IIFC-23
(CF3H) would be acceptable based on NFPA standards for use in occupied areas de-
signed for protection against propane explosions and fires. HFC'-227ea (CF3CIJFCF3)
would also be acceptable for use in occupied areas for fire protection only.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. descriptors
b.identifiers/open ended terms
c. COSATl Field/Group
Pollution Rthers
Halohydrocarbons Alkene Compounds
Fire Protection
Explosion Pi^oofing
T ests
Czone
Pollution Prevention
Stationary Sources
Total-Flooding
13 B
07 C
13L,
14 B
07E
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
106
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
-------
FOREWORD
The U.S. Environmental Protection Agency is charged by Congress with pro-
tecting the Nation's land, air, and water resources. Under a mandate of national
environmental laws, the Agency strives to formulate and implement actions lead-
ing to a compatible balance between human activities and the ability of natural
systems to support and nurture life. To meet this mandate, EPA's research
program is providing data and technical support for solving environmental pro-
blems today and building a science knowledge base necessary to manage our eco-
logical resources wisely, understand how pollutants affect our health, and pre-
vent or reduce environmental risks in the future.
The National Risk Management Research Laboratory is the Agency's center for
investigation of technological and management approaches for reducing risks
from threats to human health and the environment. The focus of the Laboratory's
research program is on methods for the prevention and control of pollution to air,
land, water, and subsurface resources; protection of water quality in public water
systems; remediation of contaminated sites and groundwater; and prevention and
control of indoor air pollution. The goal of this research effort is to catalyze
development and implementation of innovative, cost-effective environmental
technologies; develop scientific and engineering information needed by EPA to
support regulatory and policy decisions; and provide technical support and infor-
mation transfer to ensure effective implementation of environmental regulations
and strategies.
This publication has been produced as part of the Laboratory's strategic long-
term research plan. It is published and made available by EPA's Office of Re-
search and Development to assist the user community and to link researchers
with their clients.
E. Timothy Oppelt, Director
National Risk Management Research Laboratory
EPA REVIEW NOTICE
This report has been peer and administratively reviewed by the U.S. Environmental
Protection Agency, and approved for publication. Mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Information
Service, Springfield, Virginia 22161.
i i
-------
ABSTRACT
Halon 1301 has been used to protect Alaskan North Slope oil and gas production facilities from fire and
explosion since the early 1970s. Since restrictions have been placed on future production of Halon 1301, an effort
has been undertaken to develop replacements. This report presents results from flame suppression testing of
potential replacement chemicals in a cup burner using n-heptane fuel and incrtion testing in a small-scale
explosion sphere, using propane and methane as the test fuels. Test equipment and techniques are described An
analysis of cup-burner results and the impact of several modifications to the inertion test techniques are discussed.
Fire suppression test results are presented for a wide range of halocarbon chemicals. For inertion testing,
differences between Halon 1301 inciting concentrations determined using the current test apparatus and previous
testing are briefly discussed. Agent performance is presented in terms of the concentration required to achieve
flame extinguishment or inertion (defined as an explosive overpressure of one psi or less) and weight and storage
volume equivalents, reflecting the weight and storage volume of a replacement agent required to achieve the same
effectiveness performance as Halon 1301.
Seventy-one chemicals, including Halons 1301 and 1211, were tested in the New Mexico Engineering
Research Institute (NMERI) 5/8-scale cup burner to evaluate their flame extinguishment capabilities. Thirty-six
chemicals, including Halon 1301, were screened for incrtion performance using propane as a fuel. A number of
highly effective flame suppression and inertion chemicals were identified as candidate replacements for Halon
1301. However, in some cases, questions about their toxicity, stability, corrosivitv. and global environmental
impacts still remain unanswered. Chemicals containing bromine or iodine performed best on a weight and volume
basis in both the extinguishment and inertion testing. Only chemicals containing bromine or iodine had a storage
volume equivalent (ratio of the storage volume of a candidate to that of Halon 1301) for fire suppression or inertion
of less than two, meaning that it would require less than twice the volume of chemicals to effect the same
performance as Halon 1301. Several iodine-containing candidates were identified that had flame suppression
capabilities superior to Halon 1301, but limited quantities prevented determination of the performance capabilities
for the iodinated compounds in the inertion testing. In addition to neat agent testing, limited testing was
performed on blends of chemicals to assess the potential of adding small amounts of brominated or iodinated
compounds to hydrofluorocarbons (HFCs). The addilion of two percent (by total gas volume) of brominated or
iodinated chemicals to some HFCs reduced the storage volume equivalent to less than two when propane was used
as a fuel.
iii
-------
CONTENTS
Abstract "• i i
Figures v i
Tables v i i i
Acknowledgments ix
1. Introduction 1
Physical Action Agents 1
Chemical Action Agents 2
2. Flame Suppression Testing 4
Cup Burner Apparatus 4
Cup Burner Testing Procedures 4
Gaseous Chemicals 10
Liquid Chemicals 10
Room Temperature Boiling Chemicals 10
Fire Suppression Data Reduction 10
Extinguishment Test Results 13
3. Explosion Prevention Testing 17
Explosion Sphere Apparatus 17
Ignition System 18
Data Acquisition 18
Explosion Prevention Test Procedures 20
Explosion Prevention Data Reduction 20
Explosion Prevention Test Matrix 22
Calibration and Baseline Testing 22
Peak Pressures Testing 22
Halon 1301 Testing 22
Replacement Chemical Screening 23
Blend Testing 23
Flammabilily Curve Testing 24
Explosion Prevention Test Results 25
Calibration and Baseline Testing 25
Peak Pressure Testing 25
Halon 1301 Testing 27
Replacement Candidate Screening 27
Blend Testing 32
Flanunability Curve Testing 34
Halon 1301/Propane 34
Halon 1301/Methane 36
HFC-23/Propane 37
HFC-227ca/Propanc 37
FC-3-l-10/Propane 39
4. Discussion 40
Fire Suppression Inter- And Intra-Laboralory Variability 40
Explosion Prevention Inter- And Intra-Laboratory Variability 41
Ignition Source 45
iv
-------
CONTENTS (concluded)
Purity Of Methane 51
Test Conditions 51
Cleaning 51
Mixing 51
5. Conclusions And Recommendations 52
References 57
Appendices
A Candidate Agents 59
B Halocarbon Naming And Numbering Rules 62
C Equipment Description 66
D Explosion Prevention Testi ng Procedures 70
E Partial Pressure Loading Procedures 73
F Overpressure Versus Chemical Concentration Curves 75
G Overpressure Versus Chemical Concentration Curves, Blends 93
V
-------
FIGURES
Number Page
1 NMERI 5/8-Scale Cup Burner Apparatus 5
2 Cup Burner Mixing Chamber 6
3 Cup Burner Configuration for Gaseous Agents 7
4 Cup Burner Configuration for Testing Liquids 8
5 Cup Burner Configuration for Testing Highly Volatile Liquids 9
6 Explosion Sphere Apparatus 19
7 Typical Explosion Overpressure versus Time Plot 21
8 Maximum Overpressure versus Fuel-to-Air Ratio, Methane 26
9 Maximum Overpressure versus Fuel-to-Air Ratio, Propane 27
10 Overpressure versus Chemical Concentration, Halon 1301 28
11 Peak Overpressure versus HFC-227ea Concentration, Propane 31
12 Inerting Concentrations of Iodinated Compounds 33
13 Overpressure versus Time, 134a and 134a/1311 Blend, Propane 35
14 Flammability Curve, Halon 1301/Propane 36
15 Flammability Curve, Halon 1301/Methane 37
16 Flammability Curve, HFC-23/Propanc 38
17 Flammability Curve, HFC-227ea/Propane 38
18 Flammability Curve, FC-3-l-10/Propane 39
19 Comparison of Cup Burner Extinguishment Concentrations as Measured
by Various Laboratories 42
20 Flammability Curves, NMERI and Fcnwal, Methane and Propane 43
21 Effect of Ignition Strength on Stoichiometric Methane-Air Explosions 46
22 Effective Spark Energies as a Function of Stored Electrical Energy 47
23 NMERI and Fenwal Ignition System Schematic 48
24 Halon 1301 Incrtion Concentration vs. Ignition Energy, Propane 49
25 Methane/Halon 1301 Inertion Concentrations. NMERI-Fenwal-Bureau of Mines 50
F-l Overpressure vs. Chemical Concentration, Halon 1301 76
F-2 Overpressure vs. Chemical Concentration, CFC-12 76
F-3 Overpressure vs. Chemical Concentration, CFC-114 77
F-4 Overpressure vs. Chemical Concentration, HCFC-22 77
F-5 Overpressure vs. Chemical Concentration, HCFC-124 78
F-6 Overpressure vs. Chemical Concentration, HCFC-142b 78
F-7 Overpressure vs. Chemical Concentration, FC-14 79
F-8 Overpressure vs. Chemical Concentration, FC-116 79
F-9 Overpressure vs. Chemical Concentration, FC-218 80
F-10 Overpressure vs. Chemical Concentration, FC-C318 80
F-l 1 Overpressure vs. Chemical Concentration, FC-3-1-10 81
F-12 Overpressure vs. Chemical Concentration, FC-4-1-12 81
F-l 3 Overpressure vs. Chemical Concentration, FC-5-1-14 82
F-14 Overpressure vs. Chemical Concentration, FC-6-1-16 82
F-l5 Overpressure vs. Chemical Concentration, HFC-23 83
F-l6 Overpressure vs. Chemical Concentration, HFC-32 83
F-17 Overpressure vs. Chemical Concentration, HFC-125 84
F-l 8 Overpressure vs. Chemical Concentration, HFC-134 84
vi
-------
FIGURES (concluded)
Number Page
F-19 Overpressure vs. Chemical Concentration, HFC-134a 85
F-20 Overpressure vs. Chemical Concentration, HFC-152a 85
F-21 Overpressure vs. Chemical Concentration, HFC-227ca 86
F-22 Overpressure vs. Chemical Concentration, HFC-227ea 86
F-23 Overpressure vs. Chemical Concentration, HFC-236ca 87
F-24 Overpressure vs. Chemical Concentration, HFC-236fa 87
F-25 Overpressure vs. Chemical Concentration, HFC-245cb 88
F-26 Overpressure vs. Chemical Concentration. HFC-254cb 88
F-27 Overpressure vs. Chemical Concentration, HFC-32.HFC125 Azeotrope 89
F-28 Overpressure vs. Chemical Concentration, HBFC-22B1 89
F-29 Overpressure vs. Chemical Concentration, HBFC-124B1 90
F-30 Overpressure vs. Chemical Concentration, F1C-13II 90
F-31 Overpressure vs. Chemical Concentration, FIC-11511 91
F-32 Overpressure vs. Chemical Concentration, 4-Bromo-3,3,4,4,-tetrafluoro-l-butene 91
F-33 Overpressure vs. Chemical Concentration, CO2 92
F-34 Overpressure vs. Chemical Concentration, N2 92
G-l Overpressure vs. FC-3-1-10 Blends Inerting Concentration, Methane 94
G-2 Overpressure vs. HFC-134a Blends Inerting Concentration, Methane 95
G-3 Overpressure vs. HFC-134a Blends Inerting Concentration, Propane. 96
G-4 Overpressure vs. HFC-32 Blends Inerting Concentration, Propane 97
vi i
-------
...3
14
15
16
22
23
24
29
30
31
34
35
39
40
41
44
54
55
60
61
63
74
TABLES
Second Generation Chemical Classes
Full Confidence Cup-Burner Extinguishment Values
Average Cup Burner Extinguishment Concentrations of Limited Confidence
Candidate Fire Suppression Effectiveness Ordered By Storage Volume Equivalent
Calibration and Baseline Test Scries
Chemical Screening Test Matrix
Blend Test Matrix
Chemical Inertion Concentrations
Explosion Prevention Performance Ordered By Storage Volume Equivalent, Propane.
Explosion Prevention Performance Ordered By Storage Volume Equivalent, Methane.
Blend Inertion Concentrations
Blend Explosion Prevention Performance Comparison
Flammability Curves Summaiy
Evaluation of Measurement Error in Cup Burner Experiments
Inter-laboratory Comparison of Cup Burner Extinguishment Concentrations
Mclliane/Halon 1301 Inertion Results
NFPA 2001 Halon 1301 (Halocarbon) Replacement Candidates
Cardiotoxicity Values and Inertion Design Concentrations For NFPA 2001 Agents....
PAA Candidate Group List Of Halon 1301 Replacements
CAA Candidate Group List Of Halon 1301 Replacements
Prefixes for Halocarbon Numbers
Partial Pressures Of Agent And Fuel
vi i i
-------
ACKNOWLEDGMENTS
This work was performed with sponsorship from the U. S. Environmental Protection Agency, Alaskan
North Slope oil and gas production companies, and the U. S. Coast Guard.
l'x
-------
SECTION 1
INTRODUCTION
Enclosed spaces containing flammable hydrocarbon fuels present challenging fire and explosion
protection problems. Of particular concern arc Alaskan North Slope petroleum handling facilities (1), where leaks
of flammable gaseous and liquid hydrocarbons can occur, presenting both fire and explosion hazards. At present,
the primary fire and explosion protection measure for such facilities is total-flood application of Halon 1301.
However, due to depletion of the ozone layer and resultant regulations, the availability of halon fire extinguishing
agents (including Halon 1301) will decrease substantially as a result of the Montreal Protocol restrictions. The
Center for Global Environmental Technologies (CGET) at the New Mexico Engineering Research Institute
(NMERI) has evaluated the potential use of low ozone-depleting halocarbons as Halon 1301 replacements for total-
flood application in enclosed facilities such as those found at the Alaskan North Slope.
Two methods have been established for producing chemicals with reduced ozone-depleting potentials
(ODP) (2). The first involves the use of chemicals that do not contain bromine atoms, which are known to be
several times more potent at causing stratospheric ozone depletion than chlorine. The second mechanism is to
incorporate structural features within the bromine-containing molecules such that their atmospheric lifetimes are
significantly reduced and the molecules will not reach the stratosphere in amounts that will deplete the ozone.
Both methods of reducing the ODP were the focus of this study. Candidates lacking bromine atoms were identified
and tested as part of this effort Unfortunately, the effectiveness performance for these chemicals was dramatically
inferior when compared to Halon 1301 for both fire suppression and explosion prevention (inertion). Candidates
containing bromine and iodine, many of which are believed to have lower atmospheric lifetimes, were also
screened and evaluated. While these chemicals showed promise in terms of fire suppression and inertion
effectiveness, a number of questions regarding chemical stability, toxicity, corrosivity, and global environmental
characteristics still remain unanswered.
PHYSICAL ACTION AGENTS
Halon replacements can be divided into two types depending on their mechanism of fire extinguishment:
physical action agents (PAAs) and chemical action agents (CAAs). The designation into two categories does not
imply that agents may not operate by both mechanisms, rather that one action appears to be the predominating
mode of extinguishment. Physical extinguishment occurs by a variety of mechanisms including: vapor-phase heat
absorption, liquid-phase heat absorption, evaporative cooling, thermal dissociation, dilution of fuel and oxygen,
and separation of fuel and oxygen; however, vapor-phase heat absorption appears to be the most important for the
1
-------
PAA halon replacements. The PAAs do not contain bromine or iodine, the two constituents that impart chemical
action (3). PAAs tend to have lower ODPs than CAAs because of the absence of bromine, but require two to four
times more agent (compared to Halon 1301) to effect extinguishment or prevent explosions. Nonetheless, these
replacements may oflcr much needed transitional protection during the period in which halons arc being phased
out of production and new, superior substitutes can be found and manufacturing facilities are established. Because
of the extensive time required to investigate new agents and technological approaches, plans were made early in
this effort to search for chemicals superior to PAAs to protect people and property on the North Slope from the
threat of fires and explosions.
CHEMICAL ACTION AGENTS
Chemical extinguishment occurs primarily by removal of combustion free radicals which inhibit chain
reaction mechanisms (e.g., hydrogen and oxygyn atoms, and hydroxyl free radicals). With halocarbon agents,
significant free radical removal requires the presence of bromine or iodine. This process occurs through the
hydrogen bromide (or iodine) flame suppression cycle as shown below for Halon 1301:
CBrF3 + H- -> CF3- + HBr
HBr + H* —> H2 + Bp
Br + Br + M -> Br2 + M*
Br2 + H- -> HBr + Br
fll
[2]
[3]
[4]
In general, while CAAs have superior explosion protection and fire suppression capabilities, they also
tend to have higher toxicities, and bromine-containing compounds tend to have high ODPs. A need now exists for
new chemically acting replacement (halocarbon) agents that have both low ODPs and high efficiencies.
Decreasing the tropospheric lifetime of halocarbons is one means of reducing the impact on the environment.
Several "new" mechanisms have been identified to reduce tropospheric lifetimes while keeping the substituents,
iodine or bromine, that allow chemical action for fire extinguishment. Among the mechanisms that allow
decreased tropospheric lifetimes are reaction of a carbon-carbon double bond with hydroxyl free radicals (alkenes),
photolysis (alkenes, iodides, and geminal dibromides), and rainout (polar molecules).
Some of the chemical classes that exhibit the desired tropospheric removal processes arc shown in
Table 1. Several of these classes have been shown to be highly effective in fire suppression testing. Also,
preliminary indications suggest that several of there compounds have near-zero ODPs, short atmospheric lifetimes,
and low global warming potentials (GWP)(4), making them attractive for use in fire suppression and explosion
prevention. These indications need to be verified, and fielding will require several years due to the requirements
2
-------
TABLE 1. SECOND-GENERATION CHEMICAL CLASSES
CHEMICAL FAMILY
TROPOSPHERIC DESTRUCTION PROCESS
Hydrogen-Containing Geminal
Dibromides
Reaction with Hydroxyl Free Radicals and
Tropospheric Photolysis
Fluoroiodocarbons
Tropospheric Photolysis
Polar-Substituent Bromocarbons
Rainout
Bromofluoroalkenes
Reaction of Double Bond with Hydroxyl Free Radicals
and Tropospheric Photolysis
for field and emissions testing, toxicity testing, shelf-life determination, assessment of compatibility with
engineering materials, and development of manufacturing capabilities.
Many of the chemicals containing bromine or iodine identified herein have high molecular weights that
result in boiling points above 0 °C. Therefore, these higher boiling chemicals may likely require an alternative
method to gaseous dispersion for dispensing the agent. Consequently, misting technology has been investigated in
a separate effort as an alternative to deliver higher boiling point halocarbons in a three-dimensional, total-flooding
capacity.
A list of PAA candidates and CAAs from each chemical category above was developed in a previous effort
(2) (Appendix A). (Note: Appendix B provides information on naming and numbering of halocarbons.) The
candidates were separated into categories based on availability and toxicity information, with lite most promising
given the highest priority for testing. The chemicals were then tested, at laboratory scale in the NMERI 5/8-scale
cup burner, to determine their effectiveness in suppressing liquid hydrocarbon flames. The concentration required
to suppress the flame was determined as a gas volume percentage. Later, a laboralory-scalc explosion sphere
apparatus was designed and constructed to measure the inertion ability of selected candidates. Similar chambers
have been constructed by the Fenwal Corporation (5) and successfully used in graduate studies at Worcester
Polytechnic Institute (6,7,8,9). Fire suppression and inertion results were then used to determine a relative ranking
of performance among candidate chemicals, based on both weight and storage volume equivalents to Halon 1301.
3
-------
SECTION 2
FLAME SUPPRESSION TESTING
Chemicals were tested at the laboratory scale to determine their effectiveness in extinguishing
hydrocarbon flames. The recognized laboratory-scale fire suppression test method uses the cup-burner apparatus.
The original cup burner, developed by Hirst and Booth (10) at Imperial Chemical, Inc. (ICI) in the late 1970s,
required relatively large test quantities of chemicals. In an effort to reduce the amount of chemicals needed to test
the flame suppression capabilities, NMER1 developed and validated a 5/8-scale version of the ICI apparatus under
a separate effort (11). Flame suppression testing of all candidates was performed using the NMERI 5/8-scale cup-
burner apparatus. The equipment and test technique are fully described elsewhere (11); however, a brief
description is provided below.
CUP BURNER APPARATUS
The NMERI 5/8-scale cup burner (Figure 1) consists of a glass chimney with an inner diameter of 50 mm
and a height of 415 mm. These dimensions yield an approximate 37.5 percent reduction in linear dimensions
compared to a previous NMERI full-scale version, which was nearly identical to the ICI cup burner used by Hirst
and Booth. The flame cup, which passes through a port in the side of the chimney, is 13 mm in diameter, 165 mm
above a mixing chamber, and is placed in the center of the circular chimney. The flame cup is connected to a clear
plastic tube that passes through the chimney bottom to a side-arm flask filled with «-heptane. The flask sits on an
adjustable platform for fuel level control. The bottom of the chimney contains glass beads to help with agent/air
mixing. A small glass mixing chamber is connected to the bottom of the chimney through a stopper. The
extinguishing agent enters the mixing chamber through a side arm. Air is introduced into the mixing chamber via
another side arm with an interior attachment that channels the air to the bottom of the chamber (Figure 2), thus
ensuring the complete mixing of air and agent. Liquid agents are introduced into the mixing chamber by a syringe
pump. The agent side-arm is fitted with a septum for the syringe needle. Gaseous agents enter the mixing
chamber through flowmeters and tubing. Volumetric flow rates of air and agents are used to calculate the agent
percent molar concentration required to extinguish the flame.
CUP BURNER TESTING PROCEDURES
Different cup burner test methods are used depending on the boiling point of the material to be tested.
Chemicals that are gases at room temperature are tested by connecting the gas from a cylinder to the cup burner
through appropriate flowmeters. Chemicals with relatively high boiling points are metered with a syringe pump
4
-------
All Dimensions are in mm
unless otherwise noted.
Chimney
Side Port
(Plugged)
Flame Cup
415
Flame Cup
Stalk
82
Side Ports
(Plugged)
118
Glass Beads (5 Dia.)
Wire Mesh
AIR INLET
AGENT INLET
47
1/4 in OD
Figure 1. NMERI 5/8-Scate Cup Burner Apparatus.
5
-------
Beveled & Ground Glass
Circular Glass
Chamber
AGENT INLET ¦
47
X
Impinger (1dia.)
AGENT/AIR OUTLET
t
16.8
SjL
All Dimensions are in mm
unless otherwise noted.
• 1/4 in OD
Glass Tubing
AIR INLET
49
i
90
Figure 2. Cup Burner Mixing Chamber.
into a heated vaporization compartment just below the cup burner. Extinguishment concentrations of chemicals
that have boiling points near room temperature (approximately 25 ± 10 °C) are extremely difficult to measure,
since such materials do not vaporize well from cylinders, but have sufficiently high vapor pressures that a syringe
pump is difficult to use. A novel method using a capped burette was developed to obtain cup-burner
extinguishment concentrations for such materials. However, results obtained by this methods are still often
inconsistent. Future work is required to find suitable modifications of this method that will allow reliable
extinguishment concentrations to be obtained for chemicals with intermediate boiling points. Figures 3, 4, and 5
show the apparatus configuration for testing the various states of chemicals.
After the appropriate apparatus configuration was chosen, the fuel was placed in the flame cup by attaching a hose
from the cup to a flask resting on the vertically adjustable laboratory platform. The fuel level was maintained at
the rim of the cup by raising or lowering the fuel reservoir flask. Air flow was regulated at a constant pressure of
179 kPa (26 psi) and a flow between 5370 mL/min and 13,200 mL/min for gaseous agents. For high boiling
agents (liquids) the air flow was maintained at approximately 14,700 mL/min. The total flow rate of agent plus air
was kept within a pre-detcrmincd operating region (between 8,000 and 16,000 mL/min) for this cup burner (II).
6
-------
1000-mL Soap Film
Bubble Meter
Cup Burner Apparatus
j— Pressure Gage
Agent Rotameter
j— Pressure Gage
Calibrated Air Rotameter
Needle Control Valve
— Digital Thermometer
Needle Control Valve
AIR INLET
(Regulated to 26 lb/in2)
- AGENT INLET
(Regulated to 10 lb/in2)
Digital Thermometer
Figure 3. Cup Burner Configuration for Gaseous Agents.
-------
Cup Burner Apparatus
18 gauge 3-1/2 in Spinal Needle
Gas Tight Syringe
Syringe Clamp
Syringe Plunger
Plunger Pushing Mechanism
Variable Speed —v
Syringe Pump \
Pressure Gage
Septum
Heated Sand Bath
Calibrated Air Rotameter
Needle Control Valve
tAIR INLET
(Regulated to 26 lb/in2)
Digital Thermometer
Figure 4. Cup Burner Configuration for Testing Liquids.
-------
Stopper
10 mL Burette
Cup Burner Apparatus
Ring Stand
Pressure Gage
10
Calibrated Air Rotameter
18 gauge 3-1/2 in Spinal Needle
Needle Control Valve
Septum
Heated Sand Bath
— AIR INLET
(Regulated to 26 lb/in2)
Digital Thermometer
Figure 5. Cup Burner Configuration for Testing Highly Volatile Liquid.
-------
Gaseous Chemicals
Gaseous chemicals were stored in cylinders connected to the agent flowmeter, with agent delivery pressure
regulated at 6.894 kPa (10 psi). To provide a rough indication of the extinguishment concentration, an arbitrary
agent flowmeter selling was chosen, and the flow was increased by five units every 15 seconds until
extinguishment occurred. To determine more precisely the extinguishment concentration, a setting five units
below the crude "extinguishment concentration" was used as a starling point and the flow increased by one unit
every 15 seconds until extinguishment occurred. Since chemical quantities were often limited, the flow rate
corresponding to this flowmeter setting was measured using a bubble flowmeter. Five tests were performed for
each chemical in order to calculate the extinguishment concentration for that agent. The chimney temperatures,
air flowmeter settings, and agent flowmeter settings were measured.
Liquid Chemicals
For liquids, a syringe was used to deliver the agent into the cup burner mixing chamber. The syringe was
filled with agent, the needle was passed through a septum in the mixing chamber at the bottom of the cup burner,
and the syringe was placed on the syringe pump guides. The desired syringe pump setting was chosen, the syringe
pump and timer were activated, and the pump speed was adjusted upwards every 15 seconds until extinguishment
occurred. At extinguishment, volume versus time readings were taken to determine the flow rate of the agent
exiting the syringe. A total of five tests were performed for each agent to determine the minimum extinguishment
concentration.
Room Temperature Boiling Chemicals
For room temperature boiling chemicals, flow rates were controlled and determined using a 10 ml.
stoppered burette with a needle attachment. At extinguishment, volume versus time readings were taken as the
level of chemical in the burette dropped. Generally, 10 or more tests per chemical were required to determine the
average minimum extinguishment concentration.
FIRE SUPPRESSION DATA REDUCTION
Extinguishment concentrations of gaseous agents arc reported as volume percent concentration. The
molar flow rates for the air and gaseous chemicals were calculated using the ideal gas law and solving for n:
n = PV/RT (5)
where
n = molar flow rate (molcs/min)
P = atmospheric pressure (atm)
10
-------
V = volumetric flow rate (mL/min)
R = gas constant (82.06 mL-atm/mole-K)
T = gas temperature in the chimney (K).
Once molar flow rates were determined, the extinguishment concentration, in percent, was calculated
using the following equation:
EC nagcnt^(nagent+ nair) (6)
where
EC = extinguishment concentration (%)
"agent ~ ^ ^agent^ ^
nair =Pva,/RT <»>
and
Vagent = a8cnt volumetric flow rate (mL/min)
^air = air volumetric flow rate (mL/min)
Substituting Equations (7) and (8) into Equation (6) gives the molar percent extinguishment concentration
for gaseous agents as a function of the volumetric flow rates:
EC = vagcnt/(vagcnt+ ^air) ^
Based on analysis, the measurements of temperature and pressure within the cup burner apparatus are
unnecessary for determining gaseous agent extinguishment concentrations. These parameters fall out of the
equation when the substitutions are made into Equation (6). Accurate air and agent volumetric flow rate
measurements are the critical values in determining extinguishment concentrations.
The extinguishment concentration of liquid agents is calculated somewhat differently than that for
gaseous agents because the flow rates are monitored differently. While the flow rates of gaseous agents are
measured using calibrated rotameters, liquid agent flow rates are controlled and measured using a syringe pump.
Therefore, the molar flow rates are calculated for liquid agents using the following equation:
nliq agent ^'liq agent P^^liq agcnl 0®)
where
njjq ggent = molar flow rate of the liquid agent (moles/min)
11
-------
^liq agent
P
MWjjq agent
= volumetric flow rate of the liquid agent (mL/min)
= density of the liquid agent (g/inT.)
. = molecular weight of the liquid agent (g/mole)
The molar flow rate of air is determined using Equation (8). The extinguishment concentration in volume
percent is calculated according to Equation (6). When calculating liquid agent extinguishment concentrations,
pressure and temperature arc important since the values for njjq agCn( and najr are used in the calculation.
Cup-burner results are usually given as the minimum gas-phase concentration needed to suppress a flame.
One can then relate the result to a reference compound (usually Halon 1301 for total-flooding agents) to calculate a
Gas Volume Equivalent (GVEq). For example, if a candidate agent has an extinguishment concentration of ECc
and a reference compound has an extinguishment concentration of ECR, then the GVEq is given by Equation (11):
Note that the GVEq gives the increase in gaseous volume of a candidate agent as needed to provide an
extinguishment as measured by a cup burner equivalent to the reference agent. The reciprocal of the GVEq gives
the Gas Volume Effectiveness (GVEf, Equation 12), which is the effectiveness of a candidate material relative to
the reference compound. For GVEq, a higher number indicates a lower efficiency. The opposite is true for GVEf.
The gas-phase concentration of an agent required to extinguish a flame does not always provide an
accurate representation of an agent's efficiency. More important arc the weight and storage volume of an agent
required to give the same fire extinguishment capability as a reference compound. The Weight Equivalent (WEq)
is the ratio of the weight of the candidate agent to the weight of the reference agent. The equation used is
where "MW" denotes "molecular weight," the subscript "C" denotes "candidate," and the subscript "R" denotes
"reference agent." The Weight Effectiveness (WEf) is the reciprocal of the WEq and gives a measure of the
effectiveness of a candidate relative to that of the reference material as determined by the weight required to
suppress a cup-burner flame (Equation 14).
GVEq = ECc/ECr
(11)
GVEf= 1/GVEq = ECr/ECc
(12)
WEq = GVEq x (MWc/MWr) = (ECc/ECR) x (MWc/MWr)
03)
WEf= 1/WEq = 1/GVEq x (MWr/MWc) = (ECR/ECc) x (MWR/MWC)
(14)
12
-------
Note that as the molecular weight of the candidate agent increases, the WEq increases and the WEf
decreases. Thus, lower molecular weight materials appear to be more effective when effectiveness is measured by
weight. Like the numbers for gas volume, a higher efficiency is denoted by a higher WEf and a lower WEq.
The Storage Volume Equivalent (SVEq) is the amount of candidate agent as measured by storage volume
requirements relative to that required by a reference agent. The storage density of the agent is important in
determining the storage volume requirements. Obviously, a higher density means that less space is required.
Since most agents of interest here are stored as liquids (usually under pressure) because they require less volume,
liquid densities (LDs) are used to determine the SVEq (Equation 15).
SVEq = WEq x (LDR/LDC) = (ECc/ECr) x (MWc/MWr) x (LDr/LDc) (15)
Again, one can use a Storage Volume Effectiveness (SVEf) to measure the effectiveness relative to a
reference compound (Equation 16).
SVEf = 1/SVEq = (ECR/ECc) x (MWr/MWc) x (LDc/LDr) (16)
EXTINGUISHMENT TEST RESULTS
All of the cup-burner work reported herein employed //-heptane as the fuel. Average extinguishment
concentrations for the materials tested are presented in Tables 2 and 3. Some of the chemicals tested were not
halon replacement candidates and are included only as reference compounds. The values presented in these tables
have been rigidly scrutini/xd for possible experimental errors, suitability for testing wilh available methods,
flammability, and other factors which might affect the reported values. The values reported on Table 2 have met
all the criteria required for full confidence in their values subject lo the limitations presented above. The values
presented on Table 3 are for various reasons (e.g., flammability, limited quantities, boiling point, questionable
experimental conditions) deemed to be of limited reliability and are presented for completeness only. A discussion
of errors and sensitivities is presented in Section 4.
As noted above, not all the agents in Tables 2 and 3 were considered as Halon 1301 replacements. Many
of these chemicals were tested in order to gain understanding of the fire suppression mechanism and to develop a
database for predictive algorithm generation, not addressed in this project. Table 4 presents the GVEq, WEq, and
SVEq of each of the chemicals considered as Halon 1301 replacements, in order of increasing SVEq. As indicated
above, lower GVEq, WEq, and SVEq values denote more effective chemicals, when comparing to Halon 1301.
Values of unity suggest that the chemical is equally as effective as halon. Appendix B provides information on the
naming and numbering of halocarbons.
13
-------
TABLE 2. FULL CONFIDENCE CUP-BURNER EXTINGUISHMENT VALUES
Halocarbon No.
Halon No. Name
CAS No.
Exting.
conc.
(vol %)
CC-10 104
CFC-11 113
CFC-12 122
CFC-13 131
CFC-113 233
CFC-114 242
CFC-115 251
HCC-20 103
HCC-30 102
HCFC-22" 121
HCFC-122 223
HCFC-124** 241
HCFC-132b 222
HCFC-133a 231
HCFC-225ca/cb 352
FC-14" 14
FC-116** 26
FC-218** 38
FC-C-318** 48
FC-3-1-10** 4-10
FC-5-1-12 6-12
FC-5-1-14 6-14
FC-6-1-14 7-14
FC-6-1-16 7-16
FC-7-1-16 8-16
HFC-23** 13
HFC-32** 12
HFC-125** 25
HFC-134** 24
HFC-134a** 24
HFC-227ca" 37
HFC-227ea" 37
HFC-236ea** 36
HFC-236fa** 36
HFC-245cb** 35
BFC-13B1 1301
BFC-21B2** 1202
BFC-114B2 2402
BFC-7-1-17aB1 8-17-01
BCFC-12B1 1211
Tetrachloromethane (carbon tetrachloride)
Trichlorofluoromethane
Difluorodichloromethane
Chlorotrifluoromethane
1,2,2-T richloro-1,1,2-trifluoroethane
1,2-Dichloro-1,1,2,2-tetrafluoroethane
1-Chloro-1,1,2,2,2-pentafluoroethane
Trichloromethane
Dichloromethane
Chlorodifluoromethane
1.1-Difluoro-1,2,2-trichloroethane
2-Chloro-1,1,1,2-tetrafluoroethane
1.2-Dichloro-1,1-difluoroethane
2-Chloro-1,1,1-trifluoroethane
3.3-Dichloro-1,1,1,2,2-pentafluoropropane/
1,3-Dichloro-1,1,2,2,3-pentafluoropropane
(azeotrope)
Tetrafluoromethane
Hexafluoroethane
Octafluoropropane
Octafluorocyclobutane
Decafluorobutane (perfluorobutane)
Perfluoromethylcyclopentane
Tetradecafluorohexane (perfluorohexane)
Perfluoromethylcyclohexane
Hexadecafluoroheptane (perfluoroheptane)
Perfluoro-1,3-dimethylcyclohexane
Trifluoromethane
Difluoromethane
Pentafluoroethane
1,1,2,2-Tetrafluoroeiharie
1,1,1,2-Tetrafluoroeihane
1,1,1,2,2,3,3-Heptafluoropropane
1,1,1,2,3,3,3-Heptafluoropropane
1,1,1,2,3,3-Hexafluoropropane
1,1,1,3,3,3-Hexafluoropropane
1,1,1,2,2-Pentafluoropropane
Bromotrifluoromethane
Dibromofluoromethane
1,2-Dibromo-1,1,2,2-tetrafluoroethane
Bromo-heptadecafluoro-octane
Bromochlorodifluoromethane
56-23-5
75-69-4
75-71-8
75-72-9
76-13-1
76-14-2
76-15-3
67-66-3
75-09-2
75-45-6
354-21-2
2837-89-0
1649-08-7
75-88-7
127564-92-5
75-73-0
76-16-4
76-19-7
115-25-3
355-25-9
180-22-7
355-42-0
355-02-2
355-57-9
335-27-3
75-46-7
75-10-5
354-33-6
359-35-3
811-97-2
2252-84-8
431-89-0
431-63-0
690-39-1
1814-88-6
75-63-8
1858-53-7
124-73-2
423-55-2
353-59-3
7.6
7.8
7.6
7.3
6.2
6.4
6.3
10.5
14.1
11.6
6.3
8.2
7.9
7.6
6.5
13.8
7.8
6.1
7.2
5.0
3.7
4.4
3.5
4.0
3.2
12.6
8.8
9.4
11.2
10.5
6.5
6.3
6.6
5.6
8.2
2.9
1.8
2.1
2.4
3.2
(coiitinuecJ)
n-Heptane fuel, NMERI 5/8-scale cup burner.
Considered as candidate to replace Halon 1301 for total-flood fire extinguishment in present
work (Appendix A).
14
-------
TABLE 2. (CONCLUDED)*
Halocarbon No.
Halon No.
Name
CAS No.
Exting.
conc.
(vol %)
HBFC-22B1"
1201
Bromodifluoromethane
1511-62-2
4.4
HBFC-123aB1
2311
2-Bromo-2-chloro-1,1,1-trif!uoroethane
151-67-1
3.1
HBFC-123B2**
2302
2,2-Dibromo-1,1,1-trifluoroethane
354-30-3
1.9
HBFC-123aB2**
2302
1,2-Dibromo-1,1,2-trifluoroethane
354-04-1
2.0
HBFC-123bB1
2311
1 -Bromo-2-chloro-1,1,2-trifluoroethane
354-06-3
3.2
HBFC-124B1"
2401
2-Bromo-1,1,1,2-tetrafluoroethane
124-72-1
2.9
HBFC-142B1
2201
2-Bromo-1,1-difluoroethane
359-07-9
4.2
FIC-1311"
13001
Trifluoroiodomethane
2314-97-8
3.0
FIC-11511"
25001
Pentafluoroiodoethane
354-64-3
2.1
FIC-217bal1"
37001
1,1,1,2,3,3,3-heptafluoro-2-iodopropane
677-69-0
3.2
FIC-217cal1**
37001
1,1,2,2,3,3,3-heptafluoro-1-iodopropane
754-34-7
3.0
FIC-319al1**
49001
Nonafiuoro-1 -iodobutane
423-39-2
2.8
FIC-5-1-13al1 **
6-13-001
Tridecafluoro-1-iodohexane
355-43-1
2.5
N/A**
N/A
4-Bromo-3-chIoro-3,4,4-trifluoro-1-butene
374-25-4
4.5
N/A**
N/A
4-Bromo-3,3,4,4-tetrafIuoro-1-butene
18599-22-9
3.5
N/A
10-18
Perfluorodecalin
306-94-5
3.6
N/A
N/A
Chloropentafluorobenzene
344-07-0
5.4
N/A
N/A
1,3-Dichlorotetrafluorobenzene
1198-61-4
6.1
* n-Heptane fuel, NMERI 5/8-scale cup burner.
** Considered as candidate to replace Halon 1301 for total-flood fire extinguishment in present
work (Appendix A).
TABLE 3. AVERAGE CUP BURNER EXTINGUISHMENT CONCENTRATIONS
OF LIMITED CONFIDENCE*
Halocarbon No. Halon No. Name
CAS No.
Exting.
conc.
Jvol%}_
20.OT
7.8*
7.1 §
8.3§
8.0t
12.5§
8.5*
4.9§
10.1#
55#
4 6#
5 6#
12.0§
HCFC-31
111
Chlorofluoromethane
593-70-4
HCFC-121
214
1,1,2,2-Tetrachloro-1-fluoroethane
354-14-3
HCFC-123
232
2,2-Dichloro-1,1,1-trifluoroethane
306-83-2
HCFC-123a
232
1,2-Dichloro-1,1,2-trifluoroethane
354-23-4
HCC-130a
204
1,1,1,2-Tetrachloroethane
630-20-6
HCFC-141b
212
1,1-Dichloro-1-fluoroethane
1717-00-6
N/A
N/A
1 -Bromo-3,3,3-trifluoro-1 -propene**
N/A
CFC-216ba
362
1,2-DichIoro-1,1,2,3,3,3-hexafluoropropane
661-97-2
HFC-254cb"
34
1,1,2,2-Tetrafluoropropane
40723-63-5
HCC-270fa
302
1,3-Dichloropropane
142-28-9
HCC-270da
302
1,2-DichIoropropane
78-87-5
HCC-272ea
32
1,2-DifIuoropropane
62126-90-3
FC-4-1-12
5-12
Dodecafluoropentane (Perfluoropentane)
678-26-2
* n-Heptane fuel, NMERI 5/8-scale cup burner,
t Questionable experimental conditions.
$ Insufficient quantity for accurate testing.
§ Flammable compound.
* Near room temperature boiling point.
" Considered as candidate to replace
Halon 1301 for total-flood fire
extinguishment in present work (Appendix A).
15
-------
TABLE 4. CANDIDATE FIRE SUPPRESSION EFFECTIVENESS ORDERED
BY STORAGE VOLUME EQUIVALENT
Halocarbori No.
MW
Liq. Dens.
Exting. conc.
GVEq
WEq
SVEq
(g/mole)
(g/mL)
(vol %)
BFC-21B2
191.81
2.42
1.8
0.62
0.80
0.50
HBFC-123B2
241.81
2.22
1.9
0.66
1.06
0.72
HBFC-123aB2
241.81
2.17
2.0
0.69
1.12
0.78
FIC-1311
195.91
2.36
3.0
1.03
1.36
0.87
FIC-115-11
245.91
2.07
2.1
0.72
1.20
0.87
HBFC-124B1
180.92
1.85
2.9
1.00
1.22
0.99
Halon 1301
148.91
1.50
2.9
1.00
1.00
1.00
HBFC-22B1
130.90
1.55
4.4
1.52
1.33
1.29
FIC-217cal1
295.92
2.06
3.0
1.03
2.06
1.50
FIC-217bal1
295.92
2.10
3.2
1.10
2.19
1.57
HFC-32
52.02
0.98
8.8
3.03
1.06
1.62
FIC-319al1
345.92
2.01
2.8
0.97
2.24
1.67
BTFB*
206.96
1.36
3.5
1.21
1.68
1.85
F1C-5-1 -13al1
445.94
2.05
2.5
0.86
2.58
1.89
HFC-236fa
152.04
1.37
5.6
1.93
1.97
2.16
BCTFBt
233.40
1.68
4.5
1.55
2.43
2.16
HCFC-133a
118.50
1.39
7.6
2.62
2.09
2.25
FC-116
138.01
1.59
7.8
2.69
2.49
2.35
HFC-236ea
152.04
1.42
6.6
2.28
2.32
2.45
HFC-227ea
170.03
1.42
6.3
2.17
2.48
2.62
FC-3-1-10
238.03
1.52
5.0
1.72
2.76
2.72
HFC-227ca
170.03
1.39
6.5
2.24
2.56
2.76
HCFC-124
136.48
1.38
8.2
2.83
2.59
2.82
FC-218
188.02
1.35
6.1
2.10
2.66
2.95
HCFC-22
86.47
1.17
11.6
4.00
2.32
2.98
FC-5-1-14
338.03
1.68
4.4
1.52
3.44
3.08
HFC-134a
102.03
1.20
10.5
3.62
2.48
3.10
FC-6-1-16
388.03
1.73
4.0
1.38
3.59
3.12
FC-14
88.00
1.33
13.8
4.76
2.81
3.17
HFC-245cb
134.05
1.20
8.2
2.83
2.55
3.18
HFC-125
120.02
1.23
9.4
3.24
2.61
3.19
HFC-134
102.03
1.20
11.2
3.86
2.65
3.31
FC-C-318
200.04
1.48
7.2
2.48
3.34
3.38
HFC-23
70.01
0.67
12.6
4.34
2.04
4.57
* 4-Bromo-3,3,4,4-tetrafIuoro-1 -butene
4-Bromo-3-chloro-3,4,4-trifluoro-1-butene
16
-------
SECTION 3
EXPLOSION PREVENTION TESTING
In addition to suppressing fires, halons are also used to prevent explosions; therefore, explosion
prevention testing was of key importance in this project A number of apparatuses have been used since the early
1970's to determine the explosion prevention concentration of particular chemicals. Early methods used visual
flame propagation inhibition as the criterion for determining whether a ftiel/agent mixture was considered inerted
by a chemical, while later methods relied on pressure measurements for more objective means of determining
inertion of a fuel environment. The current Halon 1301 inertion design concentration was determined by Fenwal
Safety Systems (5) in an apparatus where the suppression of an overpressure rise due to an explosion was used as
the criterion for inertion.
Fenwal Safety Systems (5), and later Das (6), considered a flammable mixture inert if the overpressure
due to the explosion of the hydrocarbon was suppressed to one psi or less. According to this convention, the
NMERI explosion sphere testing was designed to determine the concentration of each candidate, defined on a gas
volume basis, required to limit the overpressure to no more than one psi. This concentration was defined as the
inertion concentration (IC). Two separate ICs were determined. The IC of a stoichiometric fuel-to-air ratio (ICst)
was used in screening agents. The inerting concentration at the peak of the flammability curve (ICfl), was
determined for selected fuel and agent combinations.
Test equipment, methodology, and parameters were designed for this program to produce repeatable data
that could differentiate between the ICs for a large number of chemicals in a cost and time effective manner. No
attempt was made to exactly simulate conditions potentially seen in an actual explosive environment at the North
Slope. Rather, the emphasis was put on reliable equipment operation producing repeatable data. Results should be
used only in the context of this specific test apparatus operated under the standard conditions described herein, and
no extrapolation to other situations should be undertaken without consideration. In addition, comparisons with test
results from other explosion spheres must be understood in the context of the specific test equipment and
procedures used. This will be discussed in more detail in the section on inter- and intra-laboralory variability in
Section 4.
EXPLOSION SPHERE APPARATUS
The explosion sphere, modeled after the Fenwal Explosive Sphere (5), consisted of two 25-cm diameter
(9-3/4-inch) 304-stainless steel hemispheres welded on stainless steel flanges that could be fastened to form a
17
-------
sphere with a measured volume of 7930 cm (Figure 6). A stoichiometric mixture of fuel and air and the desired
concentration of chemical were introduced in the sphere using the partial pressure method to determine the correct
volumes of agent, fuel, and air (Appendix E). The partial pressure was measured by a 0- to 2.5-psi transducer.
A fan internal to the sphere provided mixing. The mixture was ignited by a variable power DC spark generated
between electrodes located in the center of the sphere. The resulting overpressure was detected by a 0- to 50-psi
pressure transducer and the pulse recorded on a Hewlett-Packard data acquisition system (HP DAS). Overpressure
relief was provided by a 3/4-inch safety vent disc (200 psi. Fike model number 3/4-inch PV-UT Nickel) installed in
a rupture disk holder on top of the sphere. Pipe nipples provided the inlets for the fuel, air, and chemical, pressure
transducer openings, and the vacuum and exhaust port, as well as the thermocouple and fan power penetrations.
A cryotrap was used to capture decomposition products and reduce the amount of explosion by-products released to
the atmosphere. A vacuum pump exhausted the gases through a U-shaped aluminum lube placed in a liquid
nitrogen bath. The tube and hoses were cleaned at the end of each test day. A more detailed description of the
equipment is contained in Appendix C.
Ignition System
The ignition system was designed to produce a range of spark energies. It consisted of a bank of
capacitors capable of being charged at low voltage using a power supply and discharged through a transformer
(Webster Standard Baseless 312) to boost voltage above 10,000 volts. The capacitors, which have a nominal rating
of 350 volts but were not charged above 200 volts for safety reasons, could be combined in series or parallel to
produce different values of total capacitance. For the range of spark energies used for these tests, three 2000
microfarad (mfd) capacitors were combined in series to give a total of 0.006 farad capacitance. By varying the
charging voltage, the ignition system produced a calculated value of 40,70, and 100 joules, using the formula:
E = 1/2 CVX where E equals the energy (in joules), C equals the value of the capacitance (in farads) and V equals
the charging voltage (in volts). The resulting energy was multiplied by an estimated factor of 0.85 to account for
transformer efficiency. The spark was formed across a 6-mra gap between electrodes of industrial ignitors (SL
Auburn model 1-6) threaded into the bottom flange of the sphere. The energy measured and reported herein was
the energy stored in the capacitors, which may not be the energy available in the spark (see Section 4 for a more in-
depth discussion).
Data Acquisition
Data acquisition and recording, as well as the charging of the capacitors, was automatically controlled through the
HP DAS (Appendix C). The system controller was an HP 86B computer, which monitored the system operation
and the partial pressures of the agent, fuel, and air during the loading process, and recorded the pressure-pulse
data. The HP 86B was connected to a printer and plotter to produce a printed record of the test. Test data were
stored on 3.5-inch floppy disks. An HP 3852 Data Acquisition/Control Unit and an HP 3488A Switch Control
18
-------
Thermocouple —
To Thermocouple
Reader —.
Burst Disk
To AC
SECTION - SIDE VIEW
— Agent Valve
(Typ)
Electrode
0
Mixing Fan
To Vacuum
To Transformer
Spare
Fuel In
Vacuum
Gage pressure
Gage
To
Loading and
Recording
Transducer
m
PLAN VIEW
Figure 6. Explosion Sphere Apparatus.
-------
Unit provided interface between the pressure transducers and the HP 86B and automatically charged the capacitors
lo the correct test voltage.
EXPLOSION PREVENTION TEST PROCEDURES
The test methodology which governed the conduct of the testing involved loading the correct amounts of
fuel, chemical, and air into the sphere, mixing the components to a homogenous mixture by the internal fan, and
discharging the DC spark in an attempt to ignite the mixture. The correct concentrations of each component were
determined using the partial pressure method (Appendix E), which provided the partial pressures of fuel and
chemicals. The valves which allowed the fuel and chemical to enter the sphere were operated manually. The
loading pressure transducer measured the internal pressure to the nearest .00689 kPa (0.001 psi), and when the
desired pressure of each component was reached, the valve was closed. After fuel and chemical were placed in the
sphere, air was added to raise the internal pressure to 101 kPa (14.7 psi) corresponding to sea level. The fan was
shut ofT, and after one or two minutes, depending on the fuel type, the capacitors were charged and then
immediately discharged through the electrodes inside the sphere. Any resulting explosion was detected through
the analog pressure gage and thermocouple reader and recorded by the computer. Appendix D provides a more
complete description of the test technique.
EXPLOSION PREVENTION DATA REDUCTION
After each test, a plot of overpressure versus time was generated (Figure 7). This plot automatically
accounted for the loading pressure by averaging the first 10 data points and subtracting this average from all
subsequent points, leaving only the pressure rise attributable to the explosion. The plot contained noise with a
magnitude of approximalcly ±2.07 kPa (0.3 psi). On plots where the maximum overpressure was over 68.9 kPa
(10 psi), the noise was ignored; however, where the maximum overpressure was less than 68.9 kPa (10 psi),
maximum overpressure was estimated by averaging the noise peaks.
Plots of the overpressure versus chemical concentration were generated, and tests were continued until at
least one test resulted in an overpressure of one psi or less. Statistical analyses (least squared linear regression)
were performed on the data and the ICst was taken as the concentration where the best-fit regression line passed
through one psi (Appendix F). The ICst was tabulated for each chemical and fuel. For selected chemicals, full
flainmability curves were determined in which a range of fuel and chemical concentrations were tested in addition
to the stoichiometric mixture. Chemical and fuel concentrations were plotted as points corresponding to explosion
or non-explosion. The flammability curve or envelope was drawn between the interface and the explosion and
non-explosion points.
20
-------
3.00
2.00 -
... f._ . , .
EPA SPHERE TEST SERIES
r? i.00 -
w
CL
LU
5
cn
cn
UJ
CO
a
cr
UJ
m
2:
<
I
o
...J
-i.00 —-
-2.00 -
- r
_t~ ' : '
-r —
-3.00
r !-
i i
! i
_j—h u
I ' :
i i
—-f- -
0 300 600 900 1200 1500 1800 2100 2400 2700
ELAPSED TIME (mS)
3000
Figure 7. Typical Explosion Overpressure versus Time Plot.
Gas volume, weight and storage volume equivalents and effectiveness were computed in a like manner to
cup-burner results (see pages 12 and 13). These formulas were used to calculate equivalency values for explosion
prevention by replacing EC with IC.
The SVEq of a blend of chemical candidates (see page 23) is calculated for each component separately
based on the ICst of the mixture and the percentage of each component. The following example illustrates the
calculation of the SVEq for the volumetric blend of 85 percent HFC-134a/15 percent FIC-13I1 for propane (ICst =
10.3 percent for the blend).
SVF^^SVtkj^SVEq
SVEqbknd = 0.85
IC,
blend
1311
V MWm,Y LDm,^
V ^'1301 J
mw,;
1301
+ 0.15
IC,
blend
V IC\yo\ J
mw.;
1311
MWninj
Ln
1301
ID,
1311 •
(17)
(18)
103 V 102.3
lw °'85' 4.3 A148
1.748 + 0301
^+0.15f^
.9/ V 4.3
10.3V 196 V 15
4.37 V 148.9/V 2.36.
SVEql
blend
2.05
(19)
(20)
(21)
21
-------
EXPLOSION PREVENTION TEST MATRIX
Before replacement candidates were tested, considerable testing was undertaken to optimize the test
equipment and technique and to develop baseline data on Halon 1301 for comparison to previous testing at other
laboratories.
Calibration And Baseline Testing
The primary purpose of die calibration and baseline test series was to determine optimum test conditions,
with regard to the strength of the ignition spark and ifiiel-to-air ratio and to develop baseline Halon HOI data for
comparison with results from other researchers to ensure that those results from this testing were comparable to
other efforts. The baseline test series was also designed to provide a reference for replacement chemical testing,
A matrix of the calibration test series is shown in Table 5.
TABLE 5. CALIBRATION AND BASELINE TEST SERIES
Agent Fuel Chemical concentration Ignition energy
(vol %) (joules)
PEAK PRESSURE
TESTING
None Propane None 40,70,90,100
None Methane None 40,70,90,100
HALON 1301
BASELINE TESTING
Halon 1301 Propane 2 to 6 40,70,100
Halon 1301 Methane 2 to 6 70
Peak Pressures Testing-
During peak pressure testing, the baseline explosive overpressure, without inciting chemicals, for propane
and methane was determined for a range of ignition energies. Tests were run at different ignition energy levels at
stoichiometric conditions and at conditions slightly richer or leaner than the stoichiometric to determine the effect
of fuel concentration and to ascertain which concentrations caused the highest overpressures.
Halon 1301 Testing—
The purpose of the Halon 1301 test series was to determine the ICS( of Halon 1301 for each fuel and to
verify the repeatability of the apparatus using at least two runs for each Halon 1301 concentration.
22
-------
Replacement Chemical Screening
After completion of the calibration and baseline Halon 1301 testing, the process of screening large
numbers of chemicals at stoichiometric fuel-to-air ratios was begun. The required TCst for each chemical was
estimated based on a ratio of its cup-burner test results to that of Halon 1301. An initial test was run at this
concentration and the chemical concentration was increased or decreased based on the resultant overpressure.
Sufficient tests were performed to draw a curve of overpressure versus chemical conccntralion such that the
concentration required to reduce the overpressure to one psi could be determined. Linear regression methods were
used to determine the ICsl of each chemical Table 6 is a listing of the chemicals tested.
TABLE 6. CHEMICAL SCREENING TEST MATRIX
Halons*
Halon 1301
Fluorocarbons
FC-14
FC-116
FC-218
FC-C318
FC-3-1-10
FC-4-1-12J
FC-5-1-14""
FC-6-1-16t
Bromofluoroalkenes
Bromotetrafluorobutene
Chlorofluorocarbons*>f
CFC-12
CFC-114
Hydrofluorocarbons
HFC-23
HFC-32
HFC-125
HFC-134
HFC-134a
HFC-152a
HFC-227ea
HFC-227ca
HFC-236fa
HFC-236ea
HFC-245cb
HFC-254cb
HFCs-32/125
Hyd roch lorof I uorocarbons
HCFC-22
HCFC-124
HCFC-142b
Hydrobromofluorocarbons
HBFC-22B1
HBFC-124B1
Fluoroiodocarbons
FIC-13I1
FIC-115Ilt
FIC-217caI1
FIC-31911
Others*
C02
N2
* Reference compounds.
T Not included in analysis.
Blend Testing
Because the iodinated and brominated chemicals performed significantly better than the HFCs, HCFCs,
and FCs, it was decided to blend minor percentages (up to 15 percent of the amount, or 2 percent of the total agent
concentration) with several chemicals lhat exhibited superior inertion performance and offered good cost,
availability, and toxicity tradeoffs. The goal was to determine whether a blend of chemicals could be found that
had acceptable inertion performance as well as adequate ODP and toxicity characteristics. Chemicals were added
to the sphere in the correct partial pressures to give the desired percentages of major and minor components.
23
-------
Three chemicals were used as the major components, FC-3-1-10, HFC-134a, and HFC-32, and three as
minor components, Halon 1301, FIC-13T1, and HBFC-124B1. In addition, equal amounts ofFIC-1311 and HBFC-
124B1 were substituted as the minor blend to determine the combined effect of brominated and iodinated agents.
The blend test matrix is presented as Table 7. Note that the total chemical percentage represents the total amount
of chemical, both major and minor components, used. The minor component percentage represents that
percentage of the previous amount, which is a minor component. For example, if the total agent percentage is
7 percent and the minor component percentage is 5 percent, the chemical concentration is 7 percent of which
95 percent is the major component and 5 percent is the minor component. Where two minor components were
used, the percentage of each is noted.
TABLE 7. BLEND TEST MATRIX
Fuel
Major
Component
Minor
Component
Total Agent^
Percentaqe*
Minor Comp
Percentage1
Methane
FC-3-1-10
FC-3-1-10
Halon 1301
HBFC-124B1
5-7
5-7
5-15
5-15
HFC-134a
HFC-134a
HFC-134a
HBFC-124B1
FIC-1311
FIC-1311 /
HBFC-124B1
5-7
5-7
5
5-15
5-15
10 (5 ea)
Propane
HFC-134a
HFC-134a
HFC-134a
FIC-1311
HBFC-124B1
FIC-1311/
HBFC-124B1
9.5-12
9.5-10
9.7-10
10-15
10-15
15 (7.5 ea)
HFC-32
FIC-1311
12-14
15
Total combined percentage of major and minor components.
t Expressed as a percent of total agent percentage.
Flammabilitv Curve Testing
All of the inertion tests of the screening lest series were conducted at a stoichiometric fuel-to-air ratio.
While it was known that a stoichiometric ratio may not identify the highest agent concentration required to inert
all fuel-to-air ratios, the number of agents to be screened limited the initial screening to a single fuel-air mixture,
and stoichiometric concentration was selected. The full range of fuel-chemical concentrations is described by a
flammability curve, or diagram. Three promising near-term candidates and Halon 1301 were tested to develop full
flammability diagrams using propane as a fuel and methane for Halon 1301.
Flammability testing involved additional testing, including the determination of the lower and upper
flammability limits of the fuel itself. These are the lowest and highest concentrations of fuel which result in an
explosion of one psi or more. Flammability limits on common fuels such as propane and methane arc available in
24
-------
the literature and serve as a comparison between this effort and previous testing. In order to generate the
flammability curve, tests were performed using selected fuel concentrations between the flammability limits, and
chemical concentrations were varied until one test resulting in an overpressure of one psi or greater (considered an
explosion) occurred. A new fuel concentration was then tested and the chemical inerting concentration was
determined at that fuel concentration. The ICfl fell between the maximum concentration resulting in an explosion
and the next highest concentration which did not cause an explosion. This represented the peak of the
flammability curve. The TCfj is reported as the concentration where an explosion did not occur.
EXPLOSION PREVENTION TEST RESULTS
Calibration and Baseline Testing
Peak Pressure Testing--
Forty-seven tests using methane as a fuel without an inerting chemical were run to determine a baseline
explosive overpressure, to investigate the effects of rich or lean fuel-to-air ratios, and to study the influence of
spark energies. Figure 8 is a composite of all 47 tests showing maximum overpressure versus the amount (by
percent) that the mixture was rich or lean. This test series also clearly showed that variations in the maximum
overpressure produced could be expected even when input parameters were carefully controlled. The series also
indicated the importance of repeatability between tests.
Fuel-to-Air Ratio—Seven tests using a stoichiometric fuel-to-air ratio (9.5 percent for methane) produced a
mean overpressure of 651.5 kPa (94.5 psi). For the 24 tests in which the mixture was within ± 6 percent of the
stoichiometric, the mean maximum overpressure was 6460 kPa (93.7 psi). Excluding the one test in excess of
689 kPa (100 psi), the mean peak pressure of the stoichiometric tests was 6460.0 kPa (93.7 psi), and the mean of
tests with a fuel-to-air ratio between ± 6 percent was 645.9 kPa (93.4 psi). Statistical analysis was performed using
a one-way ANOVA showing that these values were not statistical different from each other (p=0.88). Therefore,
while the stoichiometric ratio produced the greatest overpressure, any fuel-to-air ratio between ± 6 percent of
stoichiometric produced statistically the same overpressure. Fuel-to-air ratios within ± 6 percent of stoichiometric
should result in consistent test results.
Stored Spark Energy—Of the 24 tests meeting the ± 6 percent of stoichiometric fuel-to-air ratio criteria, 9
were run at 100 joules, 3 at 90 joules, 8 at 70 joules and 3 at 40 joules. The mean maximum overpressures with
respect to spark energy were as follows:
25
-------
% Rich
% Lean
/ Stoichiometric
-20 -15 -10 -5 0 5 10
Fuel/Air Ratio, Vol %, Rich or Lean
15
20
25
Figure 8. Maximum Overpressure versus Fuel-to-Air Ratio, Methane.
100 joules - 654.9 kPa (95.0 psi) mean maximum overpressure
90 joules - 643 kPa (9.1.3 psi) mean maximum overpressure
70 joules - 642.5 kPa (93.2 psi) mean maximum overpressure
40 joules - 641.1 kPa (93.0 psi) mean maximum overpressure
Statistical analysis (one-way ANOVA) was performed and it was determined that no significant difference
existed in the means of the peak overpressure for the four energy levels (p = 0.21). Therefore, it was decided to
perform the remaining testing at 70 joules, which provided a balance between potential IC sensitivity to energy
level at 40 joules and potential electrical problems resulting from the higher voltages required for 90 or 100 joules.
It has been noted by several researchers (References 5 and 12, for example) that a non-stoichiometric fiiel-
to-air ratio may require a greater agent incrting concentration than stoichiometric mixtures. However, since this
project involved relative performance ranking, a stoichiometric mixture was used for all tests.
A brief series using propane as the fuel was conducted, which confirmed that the results seen with
methane could be extended to propane (Figure 9). Ten tests were run at 70 joules. Tests run at a stoichiometric
fucl-to-air ratio (4 percent for propane by volume) resulted in a mean maximum overpressure of 715.9 kPa
(103.85 psi), while those within ± 6 percent (of the stoichiometric fuel-to-air ratio) resulted in a mean maximum
overpressure of 718.1 kPa (J 04.17 psi). Fuel-to-air ratios outside that range resulted in lower peak overpressures.
26
-------
106 _% Lean_
105
'35
a.
a>
3
8
w-
a.
I
o
J*:
TO
«
a
: 104
% Rich
103
102
101 -
100
' Stoichiometric
)-
+-
-20 -15 -10 -5 0 5 10
Fuel/Air Ratio, Vol %, Rich or Lean
15
-—|
20
Figure 9. Maximum Overpressure versus Fuei-to-Air Ratio, Propane.
Halon 1301 Testing—
Fiflv-two tests were performed to determine the ICS[ of Halon 1301 with propane. Twenty tests were ran
to determine the Halon 1301 ICst for methane. Figure 10 shows the plot of overpressure versus Halon 1301
concentration for propane and methane. Additional data points above 5 percent concentration, which resulted in
0 kPa (0 psi) overpressure, were omitted from these graphs for clarity. The lCst for Halon 1301 using either
methane or propane for a fuel was approximately 4.3 percent. The ICfi will be discussed later in this section.
Replacement Candidate Screening
Thirty-five candidate chemicals were tested, the majority using propane or methane as a fuel. Although
not considered replacement candidates. CFC-12 and CFC-114 were included to provide a comparison between
CFCs and current candidates. Also, CO2 and N2 were tested for comparison purposes. Appendix F displays the
plots of overpressure versus chemical concentration, with a linear regression, best-fit curve. Data points that
resulted in high overpressure were omitted from selected plots to emphasize the concentrations approaching
inertion, as were data points above which no overpressure was seen. The ICst was defined as either the
concentration at which the regression line passed through one psi overpressure for those chemicals which the data
fit a linear regression or the first concentration resulting in 0 kPa (0 psi) above which no overpressure rise was
seen in subsequent higher agent concentrations. Table 8 reports the ICsj of the chemicals tested.
27
-------
t/>
Q.
4)
£
&
a
90.00
60.00
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00
¦ Propane
c Methane
-+-
4.00 4.05 4.10 4.15 4.20 4.25
Chemical Concentration, Vol. %
4.30
4.35
Figure 10. Overpressure versus Chemical Concentration, Halon 1301.
Tables 9 and 10 present the Gas Volume Equivalent (GVEq), the Weight Equivalent (WEq), and the Storage
Volume Equivalent (SVEq) of each chemical and fuel, ranking the agents by SVEq.
A dramatic difference in inertion characteristics was observed between CAAs (bromine- and iodine-
containing) and PAAs. PAAs, such as HFC-227ea, exhibit peak overpressure versus concentration curves similar
to Figure 11, where the overpressure increases relatively little as the chemical concentration decreases until the
concentration is much below that needed for inertion (about 5 percent in this case) and the overpressure rises more
substantially with relatively small agent concentration decreases. In other words, increasing the amount of inciting
agent has only a slight effect on overpressure until finally the overpressure is reduced to less than one psi and
complete inertion is achieved.
In contrast, those agents that inert through chemical mechanisms, such as Ilalon 1301 (Figure 10),
demonstrate a different peak overpressure versus concentration curve. As the ICst is approached, the peak
overpressure falls dramatically with a very small increase in agent concentration. In general, only those agents
that contain bromine or iodine exhibit this type of behavior. Oddly, both HCFC-142b and HFC-152a (Figures F-6
and F-20) show a similar shaped curve, which may be explained by their flammable nature acting as fuels and
causing a move from stoichiometric fuel-to-air ratios to fuel-rich ratios.
28
-------
TABLE 8. CHEMICAL INERTION CONCENTRATIONS*
Chemical
ICst (vol %)
ICst (vol %)
Propane as fuel
Methane as fuel
Halon 1301
4.3
4.3
CFC-12
10.8
15.0
CFC-114
9.8
10.8
HCFC-22
19.4
14.7
HCFC-124
12.6
9.1
HCFC-142b
19.5
5.8
FC-14
21.1
18.3
FC-116
15.9
10.4
FC-218
11.2
9.0
FC-C318
11.7
8.9
FC-3-1-10
9.6
7.8
FC-4-1-12
7.6
6.5
FC-5-1-14
7.2t
5.5
FC-6-1-16
6.5
5.1
HFC-23
19.5
14.0
HFC-32
17.8
not tested
HFC-125
14.7
9.7
HFC-134
14.5
not tested
HFC-134a
14.1
7.8
HFC-152a
10.1
6.9
HFC-227ca
12.9
not tested
HFC-227ea
11.9
8.1
HFC-236ea
11.7
not tested
HFC-236fa
10.5
not tested
HFC-245cb
10.7
not tested
HFC-254cb
9.7
not tested
HFCs-32/125
16.3
not tested
HBFC-22B1
8.8
6.1
HBFC-124B1
5.8
3.6
FIC-13I1
5.2
3.1
FIC-11511
11.0
not tested
FIC-21711
7.6 + 0.4t
not tested
FIC-31911
4.9 ± 0.6t
not tested
BTFB*
5.3
not tested
co2
29.7
not tested
No
38.0
not tested
* Inertirig concentration (1C) value at stoichiometric fuel concentration,
t Insufficient data to determine.
t 4-Bromo-3,3,4,4,-tetrafluoro-1 -butene.
29
-------
TABLE 9. EXPLOSION PREVENTION PERFORMANCE ORDERED
BY STORAGE VOLUME EQUIVALENT, PROPANE
Chemical
MW
LD
"Cst
GVEq
WEq
SVEq
(g/mole)
(g/mL)
(vol %)
Halon 1301
148.91
1.50(30)T
4.3
1.00
1.00
1.00
FIC-1311
195.91
2.36M2)*
5.2
1.21
1.59
1.01
HBFC-124B1
180.92
1.85®
5.8
1.35
1.64
1.33
HFC-152a#
66.05
0.91**
10.1
2.35
1.04
1.72
HBFC-22B1
130.92
1.55**
8.8
2.05
1.80
1.74
BTFB**
206.98
1.357*
5.3
1.23
1.71
1.89
FIC-31911
345.92
2.01(20)*
4.910.6s®
1.14§®
2.65s®
1.98s®
HFC-32#
52.02
0.98##
17.8
4.14
1.45
2.21
HFCs-32/125"*
67.3
1.07(21)##
16.3
3.79
1.71
2.40
FIC-217cal1
295.92
2.06(20)*
7.610.4s®
1.77s®
3.51s®
2.56®®
HFC-254cb
116.01
1 0s
97
2.26
1.76
2.64®
HFC-236fa
152.04
1.37***
10.5
2.44
2.49
2.73®
HFC-245cb#
134.05
1.20s
10.7
2.49
2.24
2.80s
HFC-134a
102.03
1.20**
14.1
3.28
2.25
2.81
HFC-134
102.03
1.20®
14.5
3.37
2.31
2.89
HCFC-124
136.48
1.38*
12.6
2.93
2.69
2.92
HFC-236ea
152.04
1.42®
11.7
2.72
2.78
2.93®
C02
44.0
1.03(-20)***
29.7
6.91
2.04
2.97
FIC-11511
245.91
2.07(28)*
11.0
2.56
4.22
3.06
N2
28.01
0.8
38.0
8.84
1.66
3.12
FC-4-1-12
288.03
1 63###
7.6
1.77
3.42
3.15
FC-116
138.01
1.59<-73)****
15.9
3.70
3.43
3.23
FC-14
88.01
1.33(-80)*
21.1
4.91
2.90
3.27
HFC-227ea
170.03
1.42****
11.9
2.77
3.16
3.34
HCFC-22
86.47
1.17*
19.4
4.51
2.62
3.36
HFC-125
120.02
1.23(20)®
14.7
3.42
2.76
3.36
FC-5-1-14
338.03
1 68###
7.2®
1.67®
3.80®
3.39®
FC-6-1-16
388.03
1 73###
6.5
1.51
3.94
3.42
FC-3-1-10
238.03
1 52###
9.6
2.23
3.57
3.52
FC-218
118.02
1.35*
11.2
2.60
3.29
3.65
HFC-227ca
170.03
1.39s
12.9
3.00
3.43
3.70
FC-C-318
200.04
1.48(11)*
11.7
2.72
3.66
3.70
HCFC-142b#
100.5
1.12**
19.5
4.53
3.06
4.10
HFC-23
70.01
0.67"
19.5
4.53
2.13
4.77
t
t
§
#
ft
**
§§
##
ttt
m
§§§
MMII
ftftfr
At 25 °C unless noted in parentheses.
NMERI Halocarbon Database.
PCR Research Chemical Catalog, 1990.
Estimate.
Flammable compound.
DuPont, "Physical Properties of
HFCs and HCFCs."
MSDS, Great Lakes Chemical Company.
4-Bromo-3,3,4,4-tetraf!uoro-1-butene.
Number of tests limited due to small sample amounts.
IC is midpoint of concentrations resulting in an explosion
(- value) and no explosion (+ value).
MSDS, Allied Signal.
60% HFC-32/40% HFC-125.
Measured.
Handbook of Chemistry & Physics (Ref. 13).
MSDS, Genium's Reference Collection.
3M Chemical Company.
MSDS, SCM Specialty Chemicals,
tttt Great Lakes Chemical Company.
30
-------
100
80
60
\
\
\
1311
« 40
20
31911
217call
11511
6 8
Agent, Vol %
12
14
Figure 12. Inerting Concentrations of lodinated Compounds.
-------
TABLE 11. BLEND INERTION CONCENTRATIONS
Major component
Minor components)*
1CS|, methane
(vol % of mixture)
1CS{, propane
(vol % of mixture)
FC-3-1-10
None
7.8
N.T.t
Halon 1301
5.7
N.T.
HBFC-124B1
5.1
N.T.
HFC-134a
None
7.8
13.5
HBFC-124B1
N.T.
9.9
F1C-13I1
5.1*
10.3
HBFC-124B1/FIC-13I15
N.T.
9.7
HFC-32
None
N.T.
17.5
HBFC-1311
N.T.
13.0
* Minor components are 15 percent of total agent amount except where noted.
t N.T.=Not Tested
$ 10 percent concentration of minor agent.
§ 7.5 percent HBFC-124B1 - 7.5 percent F1C-13I1 for propane.
The addition of up to 15 percent of a brominated or iodinated chemical decreases the ICst of HFC-134a by
up to 30 percent for propane and up to 36 percent for methane. This corresponds to a SVEq decrease from 2.64 to
2.02, a reduction of 23 percent. If the decrease was based solely on an arithmetic average of the percentages of the
respective agents, the new SVEq would be 2.39, a decrease of only 9.5 percent. When 7.5 percent FIC-13I1 and
7.5 percent HBFC-124B1 are added, the volume efficiency is 1.92, a decrease of 27 percent. The SVEq blends
with two minor components were similarly calculated (Table 12). The addition of the minor component in the
blends does indeed decrease the ICst and the SVEq far more than a simple arithmetic average of the two agents,
and the combination of iodine and bromine decreases the SVEq even more than the addition of either component
alone.
Figure 13 is a plot of two tests, one using only HFC-134a at a concentration of 10 percent as the inerting
agent and one using a mixture of 15 percent F1C-13I1 and 85 percent HFC-134a at 10 percent, illustrating the
reduction of overpressure due to the addition of FIC-13I1 to HFC-134a.
Flammabilitv Curve Testing
Halon 1301/Propane-
Figure 14 illustrates the flammability curve for propane and Halon 1301. The ICsl was 4.3 percent
(Figure 10). The chemical concentration at the peak of the flammability curve (ICfj) was between 6.05 and
6.21 percent at a fuel concentration of 5 percent. The corresponding point for the Fenwal testing (5) was
34
-------
TABLE 12. BLEND EXPLOSION PREVENTION PERFORMANCE COMPARISON
Major
Minor
Methane
Propane
IC$t of
mixture
(vol %)
SVEq of
mixture
%change in
SVEq
ICst of
mixture
(vol %)
SVEq of
mixture
%change
in SVEq
FC-3-1-10
None
7.8
2.86
N.T.'
—
Halon 1301
5.7
1.97
-31
N.T.
—
HBFC-124B1
5.1
1.77
-38
N.T.
—
HFC-134a
None
7.8
1.56
13.5
2.68
HBFC-124B1
N.T.
—
9.9
2.02
-25
FIC-1311
5.11
1.01
-35
10.3
2.05
-23
HBFC-124B1/
N.T.
—
9.7
1.96
-27
FiC-13115
HFC-32
None
N.T.
17.5
2.18
FIC-1311
N.T.
—
13.0
1.72
-21
* Minor blends are 15 percent of total agent amount except where noted,
t Not tested
$ 10 percent concentration of minor agent.
§ 7.5 percent HBFC-124B1 - 7.5 percent FIC-1311 for propane.
EPA SPHERE TEST SERIES
CO
Q.
UJ
cr
3
cn
in
UJ
cr
CL
cr
UJ
CD
X
c
X
o
134a
I34a/i31i
600
900 iZOO 1500 1800 3100 2400 2700 3000
ELAPSED TIME (raS)
Figure 13. Overpressure versus Time, HFC-134a and HFC-134a/FIC-13l1 Blend (85/15),
10 Percent Total Concentration, Propane.
35
-------
TABLE 10. EXPLOSION PREVENTION PERFORMANCE ORDERED
BY STORAGE VOLUME EQUIVALENT, METHANE
Chemical
MW
(g/mole)
LD*
(g/mL)
'cst
(vol %)
GVEq
WEq
SVEq
FIC-1311
195.91
2.36f-42)t
1.85+
3.1
0.72
0.95
0.60
HBFC-124B1
180.92
3.6
0.84
1.02
0.82
Halon 1301
148.91
1.5(30)§
4.3
1.00
1.00
1.00
HFC-152a
66.05
0.91#
6.9
1.60
0.71
1.17
HBFC-22B1
130.92
1.55**
6.1
1.42
1.25
1.21
HCFC-142b
100.5
1.12t
5.8
1.35
0.91
1.22
HFC-134a
102.03
1.20#
7.8
1.81
1.24
1.55
HCFC-124
136.48
1.38t
9.1
2.12
1.94
2.11
FC-116
138.01
1,59(-73)tt
10.4
2.42
2.24
2.11
HFC-125
120.02
1.23(20)#
1.42W
9.7
2.26
1.82
2.22
HFC-227ea
170.03
8.1
1.88
2.15
2.27
FC-C-318
200.04
1.48(-11)t
7.8
1.81
2.44
2.47
HCFC-22
86.47
1.17#
14.7
3.42
1.99
2.55
FC-5-1-14
338.03
1.68§§
5.5t
1.28t
2.90t
2.59t
FC-6-1-16
388.03
1.73§§
5.1
1.19
3.09
2.68
FC-4-1-12
288.03
1.63§§
6.5
1.51
2.92
2.69
FC-14
88
1,33(-80)t
18.3
4.26
2.52
2.84
FC-3-1-10
238.03
1.52S§
7.8
1.81
2.90
2.86
FC-218
188.02
1.35tt
9.0
2.09
2.64
2.94
HFC-23
70.01
0.67#
14.0
3.26
1.53
3.43
* At 25 °C unless noted in parentheses. ** MSDS, Great Lakes Chemical Company,
t PCR Research Chemical Catalog, 1990. tt MSDS, SCM Specialty Chemicals.
t Estimate. tt Great Lakes Chemical Company.
§ NMERI Halocarbon Database . §§ 3M Chemical Company.
# DuPont, "Physical Properties of HFCs and HCFCs."
110.0
100.0
90.D
80.0
S. 700
tt
5 60.0
!
S 40.0
30.0
20.0
10.0
0.0
T
4.0
A
5.0
7.0 8.0 90
Chemical Concentration, Vol. %
10.0
11.0
12.0
Figure 11. Peak Overpressure versus HFC-227ea Concentration, Propane.
31
-------
Limited quantities of selected iodine-conlaining agents precluded adequate testing in order to determine
the ICgf as was done for the other agents. In some cases (FICs-21711 and -31911) only three or four tests could be
performed. The ICst reported in Tables 8 and 9 are the midpoints of those concentrations which resulted in an
explosion, designated by the minus values, and those which did not result in an explosion, designated by the plus
values. The actual incrting concentration falls between those limits, and the midpoint was chosen as an average
value. Like Halon 1301 and brominated chemicals, the overpressure did not gradually decline as the agent
concentration decreased; rather, there was either a large explosion (generally 30 psi or more) or no explosion.
Explosions of low overpressure were not seen at all, indicating that a very narrow range in concentrations is
required to inert a mixture. Figure 12 shows the inertion performance of the four iodinated compounds.
Compounds containing more than four caibons generally have boiling points too high to be tested by current
sphere methods and were not included in this testing.
(Note: After each test with an iodinated chemical that resulted in an explosion, the inside of the sphere
was covered with a "purplish" coating (likely free iodine), which permanently discolored the cpoxy scaling
penetrations and the electrodes. The day after completion of the testing of iodinated chemicals, the house vacuum
supplying the fume hood in which the sphere operated was extremely weak, although the vacuum was full strength
at other locations throughout the laboratory. Tracing the vacuum line back to the input, it was found that the
plastic stem of the shut-off valve connecting the fume hood vacuum line to the main line had disintegrated,
clogging the valve, the inside of which was covered with a substance the same color as that found inside the
sphere. By-products were either not being trapped by the liquid nitrogen trap or they passed through the trap too
rapidly to be cooled by the liquid nitrogen. Since over 300 tests had been run up to that time, it is difficult to
attribute the disintegration of the valve to the iodinated chemical alone. It is believed that use of inertant
containing iodine would not cause such damage due to its one-time use.)
Blend Testing
The outstanding performance of FIC-1311 and HBFC-124B1 led to a short series of tests to determine
whether minor blends of these agents could improve the performance of other agents. Appendix G shows results
superimposed on the plots of maximum overpressure versus concentration for the major components. Except for
the blend of HFC-134a and FIC-13I1, only blends containing 15 percent by volume of the minor blending agent
(7.5 percent of each when there were two minor components) were plotted. For FC-3-1-10, where the number of
data points was limited, no lines were drawn (Figure G-l, FC-3-1-10/1301 blend). When adequate data points
were available to draw a curve, the ICsl was determined as the intersection of the curve and the 1-psi line (Figure
G-2, HFC-134a/FIC-1311 blend). Table 11 reports the lCst results on blends in tabular form.
32
-------
10.00
9.00
8.X
>?
3
> 7.00
ft 600
o
"5 4.00
3.00
*
2.00 *¦
1.00
0 00
¦ N0ft-&Ypl03K>ff
~ Sxplction
:>•: IfWted Rtgicn
3 ¦'
3 00 4.00
Chemical Concentration, Vol. %
o-:
u
700
Figure 14. Flammability Curve, Halon 1301/Propane.
6.1 percent at 5 percent fuel concentration. The lower flammability limit was determined to lie between
2.35 percent and 2.5 percent (as compared to 2.3 percent for the Fenwal results), while the upper limit was between
9.5 and 9.8 percent (as compared to 10.2 percent for the Fenwal results). In general, the maximum inerting
concentrations and shapes of the curv es were similar for both the NMERI and Fenwal sets of test data.
Halon 1301/McLhane—
Figure 15 presents the flammability curve for methane and Halon 1301. The ICst was 4.3 percent (Figure 10). The
maximum concentration at the peak of the flammability curve (ICfl) was determined to lie between 4.75 and
4.9 percent at a fuel concentration of 8.5 percent. The corresponding value for the Fenwal testing was 7 percent at
an 8.7 percent fuel concentration. The lower and upper flammability limits were found to be greater than 5 percent
and between 13 and 14 percent, respectively, in contrast to 5 and 15 percent for the Fenwal results. The ICfj was
42 percent lower for the NMERI tests than for the Fenwal tests, although the peaks occurred at approximately the
same fuel concentration and the shapes of the curves were similar.
36
-------
— 6.00 +
I)
I1
4.X
2.00
200 3.00 4.00
Chemical Concentration, Vol. %
5.00
6.00
Figure 15. Flammability Curve, Halon 1301 /Methane.
HFC-23/Propane--
Figuxc 16 illustrates the flammability curve for propane and HFC-23. The flammability- limits are
described in the propane/Halon 1301 section. The ICj] was between 20.2 and 20.4 percent at a fuel concentration
of between 3 and 3 .2 percent. The ICS{ was 19.8 percent. No flammability curve test results from other
researchers using HFC-23 to inert propane were available for comparison.
HFC-227ea/Propanc—
For these tests, chemically pure propane was used, in comparison with standard bulk propane used for fuel
in prior testing. The baseline stoichiometric overpressure was 696 kPa (101 psi) as compared to an average of
710 kPa (103 psi) from the bulk propane. The lower flammability limit was between 2.1 and 2.2 percent,
compared to a value of 2.12 percent reported in Reference 13. The upper flammabilily limit was between 9.47 and
9.8 percent, compared to 9.35 percent reported in Reference 13. Thirty-two tests were run. including 8 used to
determine the lower and upper flammability limits. The ICfl was between 11.5 and 11.7 percent at a fuel
concentration of 4 percent (Figure 17).
37
-------
• Nem-Explosion
Q Explosion
tnerted Rtgfon
;.svy>.vA|>.v:-
(•v
-------
FC-3-1-10/Propane—
Figure 18 illustrates the flammability curve for propane and FC-3-1-10. The ICfj was between 9.75 and
9.85 percent at a fuel concentration of 4.36 percent. The ICst was 9.5 percent, which corresponds to the point on
the plot where the stoichiometric fucl-to-air ratio intersects the flammability curve. Nole that the shape of the
curve is more rounded than the propane/I IFC-23 curve (Figure 16), but similar to the HFC-227ea propane curve
(Figure 17).
All of NMERI's flammability curve results are reported in Table 13.
1000
... f!:
8 00
7.00
9 600
1
"E
e
500
§
o 4 00
3
300
200
1 00
0.00
0.00
2.00
¦ Non-Explosion
u Explosion
- Inerted Region
4.00 6.00 8.00
Chemical Concentration, Vol. %
c
M.
(Mm
0
10.00
12.00
Figure 18. Flammability Curve, FC-3-1-10/Propane.
Agent
TABLE 13. FLAMMABILITY CURVES SUMMARY
Fuel
Flammability Limits
JCfL
Lower
Upper
Agent,
Fuel,
(vol %)
(vol %)
(vol %)
(vol %)
Halon 1301
Propane
2.4-2.5
9.5-9.8
6.2
5.0
Halon 1301
Methane
5.0
13.0-14.0
4.9
8.5
HFC-23
Propane
2.4-2.5
9.5-9.8
20.4
3.0-3.2
HFC-227ea
Propane*
2.1-2.2
9.5-9.8
11.7
4.0
FC-3-1-10
Propane
2.4-2.5
9.5-9.8
9.9
3.4
* High purity propane.
39
-------
SECTION 4
DISCUSSION
FIRE SUPPRESSION INTER- AND INTRA-LABORATORY VARIABILITY
Extensive studies of the experimental variables that may affect the accuracy and precision of cup-burner
results were performed to validate the extinguishment concentrations obtained by these procedures. This work
included determination of flow measurement errors and the sensitivities of extinguishment concentrations to these
errors. Analyses of measurement and calculation techniques indicated that inherent errors in the measurement of
air and agent flow were most critical in determining the precision of the extinguishment concentration, and an
extensive series of measurements were made to determine the magnitude of these errors. Results of these
measurements are presented on Table 14.
TABLE 14. EVALUATION OF MEASUREMENT ERROR IN CUP BURNER EXPERIMENTS
Measurement
Number of
samples
Mean value
(mL/min)
95 Percent confidence limit
(mL/min)
Air Flow
43
7322
±655 (8.9 percent)
Agent Flow (gas) High Rate
12
1494
±35 (2.3 percent)
Agent Flow (gas) Intermediate Rate
12
1001
±14 (1.4 percent)
Agent Flow (gas) Low Rate
12
496
±11 (2.2 percent)
Agent Flow (liquid) High Rate
10
3.73
±0.22 (5.9 percent)
Agent Flow (liquid) Low Rate
10
2.44
±0.18 (7.4 percent)
These inherent measurement errors result from the difficulties in reading gas burettes or syringes and
measuring time with stopwatches. When these errors are propagated through the extinguishment concentration
calculation, they result in 95 percent confidence limits of 10.1 percent (gases) and 17.9 percent (liquids) of the
extinguishment concentration reported. These values correspond to standard deviations of 5.0 percent and
8.8 percent, respectively.
A survey of the literature and halon industry contacts indicates that at least five other organizations utilize
cup burners to measure fire extinguishing capabilities. Table 15 provides a comparison of extinguishment results
from these other laboratories.
40
-------
TABLE 15. INTER-LABORATORY COMPARISON OF CUP BURNER EXTINGUISHMENT
CONCENTRATIONS, VOLUME %•
Chemical
NMERI
Naval Res.
Lab.
Great
Lakes
ICI
Univ. of
Tenn.
Fenwal
Mean
Std. dev.
percent1
HFC-23
12.4
12.0
12.7
N/A
12.6
12.0
12.3
2.2
HFC-125
9.4
8.8
9.3
N/A
N/A
8.1
8.9
5.2
FC-3-1-10
5.5
5.2
4.1
N/A
5.7
5.5
5.2
10.1
Halon-1211
3.2
3.6
3.3
4.0
3.5
3.8
3.6
7.1
Halon-1301
2.9
3.1
3.1
4.0
2.7
3.0
3.1
12.1
HFC-227ea
6.3
6.6
6.0
N/A
N/A
5.8
6.2
4.4
HBFC-22B1
4.4
4.1
3.9
4.6
N/A
3.9
4.2
6.1
n2
31
30
N/A
30
N/A
N/A
30.3
1.3
co2
20.4
21
N/A
21
N/A
28
22.6
12.4
* NMERI 5/8-scale cup burner, Naval Research Laboratory (14), Great Lakes Chemical
Company (15), Imperial Chemical Industries (16), University of Tennessee, Knoxville (17),
Fenwal (18).
t Standard deviation (Std. dev.) as a percentage of the mean value. Note that single outlier
values for FC-3-1-10, Halon-1301, and CO2 account for large deviations.
Analysis of the available data indicates that, despite the differences in cup burner design and variations in
test techniques, the extinguishment values for compounds agree well between laboratories, generally within + 5 to
10 percent, which is approximately the same variability as predicted based on error analysis (Figure 19).
EXPLOSION PREVENTION INTER- AND 1NTRA-LABORATORY VARIABILITY
Figure 20 illustrates a comparison of the flammability curves of Halon 1301 with propane or methane
determined in the NMERI laboratory with those determined in the Fenwal laboratory (5). These Fenwal data were
used to derive the Halon 1301 incrtion concentration in the National Fire Protection Association (NFPA) 12A
standard. While the curves for propane arc nearly identical, the methane test results vaiy significantly. For
propane, the ICfl reported by each organization is in the range of 6 percent. However, for methane, NMERI
reports a Halon 1301 ICf| of 4.75 to 4.9 percent, while Fenwal reported an TCpj of 7 percent. In an attempt to
explain this difference, a review of past inertion methods and results was undertaken (Table 16). As seen in the
table, a variety of methods have been used to determine the Halon 1301 "inertion concentration."
41
-------
M
6
u_
X
....
o
u.
X
o
;gt*
• -Shrift. •
-h
H-
d
©
m
o
LL.
± ±
Agent
'.'o ¦
;"i£:
A'lO
° NMER1
* NRt
¦ GLCC
ICI
A U of Term
"" Feriwal
~
10 percent-
variation'^
5 percent
variation
Figure 19. Comparison of Cup Burner Extinguishment Concentrations as
Measured by Various Laboratories.
Early methods used visual (lame propagation determinants to specify whether a mixture was considered
flammable. Later methods used pressure recordings as the criterion for flammability. Spark type and energy were
also varied considerably between methods. The humidity of the air was not reported for any test. Nonetheless, a
wide range of so-called inertion concentration for Halon 1301 have been determined for methane, ranging from
4.25 to 40 percent. Differences in measurement techniques, chamber sizes, ignition sources and energies, and
means of preparing and measuring the various components have led to these significant differences between the
inerting concentrations (Table 16). Potential factors that could impact inertion results include ignition source
lenergy level, type of spark (inductive vs. strictly capacitive component)], method of generation (gap for NMERI
vs. corona discharge for Fenwal), duration of spark, location of electrodes within sphere, spark gap, mixing
techniques (loading technique, fan mixing vs. natural mixing, amount of mixing time, amount of time after fan is
shut off), purity of methane fuel, test conditions (test temperature, humidity inside sphere, contaminants inside
sphere), size and construction of sphere, and method of cleaning sphere between tests.
42
-------
NMERI
FENWAL
NMERI
FENWAL
0 12345678
Halon 1301, Vol %
0 1 2 3 4 5 6 7
Halon 1301, Vol %
Figure 20. Fiammability Curves, NMERI and Fenwal, Methane and Propane.
43
-------
TABLE 16. METHANE/HALON 1301 INERTION RESULTS
Reference
Apparatus type
Ignition type
Ignition
Inertion
Inertion
energy
criterion
conc.,
% by vol.
Du Pont (19)
Mason jar
Kitchen match head
176 J*
Visual zero
9.0t
(1971)
(0.965 L)
w/ spark plug
flame
propagation
Du Pont (19)
55 gal. drum
Kitchen match head
176 J*
Visual zero
9.0t
Intermediate
(233 L)
w/ spark plug
flame
scale tests
propagation
(1971)
Du Pont(19)
Mason jar
Kitchen
176 J*
Visual zero
9.0t
(1972)
(0.965 L)
match head
flame
propagation
AC spark
27 J/sec*
N
4.3
AC spark
130GJ/sec*
(1
40.0
Explosion
Kitchen match head
176 J*
N
6.75t
Burette
(10.2 cm
AC spark
27 J/sec*
«
4.25
diameter x
121.9 cm)
(9.2 L)
AC spark
1300J/sec*
It
10.0
Fenwal (5)
Explosion
DC; 2-series 525
~11 J*
1 psig
7.0t
(1976)
Sphere (5.6 L)
microfarad
capacitors charged
to290 V; electrodes
attached to graphite
rods
Bureau of Mines
Flammability
DC Spark Match
< 1 J*
1 psig
-5.0*
(20) (1979)
Chamber (8 L)
35 J*
-9.0*
Bureau of Mines
Flammability
Pyrotechnic Igniior
1000 J*
Unspecified
> 17.0
(21) (1991)
Chamber (20 L)
NMERI (this
Explosion
DC; 3-2000 mfd
**
70 J
1 psig
4.9t
work) (1992)
Sphere (7.9 L)
capacitors charged
to 165 V
* Effective energy.
** Stored energy.
t lcf|.
* lcst
44
-------
Ignition Source
Results from previous Halon 1301/methane inertion testing (Table 16) indicate that the reported Halon
1301 concentrations required to inert methane were highest using the more energetic ignition energies. While the
criterion for inertion differed (zero versus 25 percent tube length flame propagation), several trends can be
observed. The 1972 Du Pont mason jar and explosive burette tests, igniting the mixture by a 27 J/sec AC spark ,
resulted in inertion concentrations of 4.3 and 4.2 percent respectively, which correspond more closely to the
NMERI results than those of Femval. The 1971 Du Pont mason jar and intermediate scale tests and the 1972
mason jar and explosive burette tests, both of which used the same apparatus but 176 J/sec kitchen matches for
ignition, resulted in inerting concentrations of 6.75 and 9 percent. Tests using a 1300 J/sec AC spark resulted in
maximum inerting concentrations of 40 and 10 percent, for the mason jar and explosion burette, respectively.
While the results of these early tests may be questioned due to lack of uniformity of test procedures and analyses,
trends indicate that lower concentrations of Halon 1301 are required to inert mixtures ignited by low energy sparks
than those ignited by higher energy sparks or non-spark ignition sources. The Bureau of Mines measured an
inerting concentration of approximately 5 percent for a spark energy of less than 1 J and a concentration of
approximately 9 percent for a 35 J match (Figure 21) (20). Using a 1000 J pyrotechnic ignitor resulted in an
inertion concentration in excess of 17 percent (potentially not the maximum because of limited data availability)
(21). Figure 21 also shows the results of the NMERI testing.
Unfortunately, comparable results with propane as a fiiel are not available. The only other data available
besides the NMERI and Fenwal results are from a 1971 Du Pont mason jar study where the inertion concentration
(ICfj) for Halon 1301 was found to be 9 percent for propane using a 176 J kitchen matchhead (see Table 13).
Most research has focused on the magnitude of the ignition energy and has not explored the effects that
the ignition type (AC spark, DC spark, pyrotechnic, match head) may have on inerting concentrations. Nearly all
inertion results, including those generated by NMERI, are based on the premise that the energy available in the
spark is the energy stored in the capacitors which is then directly transformed into an effective spark energy.
Hertzberg, Conti, and CashdoIIar (22) calculated the effective energy contained in a spark at various stored energy
levels, charging voltages, and chamber volumes (Figure 22). By measuring the pressure rise due to the spark in
very small chambers, they calculated the effective spark energy and found that it could be only a fraction of the
stored energy. In a parallel effort, Hertzberg et al. (20), indicated that the conventional Halon 1301 inerting value
of between 4 and 5 percent is obtained only with a weak ignition source, and is not a true flammability limit but
rather the concentration required to prevent ignition by a spark source. The true limit, according to Reference 20,
appears to be in the 8 percent range. Bartknccht (23) indicates that the maximum flammability range for
explosions occurs at 10,000 J. but does not describe how this value is determined.
45
-------
8
Fuel ; Methane
7
1 Match (35 J Effective)
6
5
4
3
NMERI Spark
(70 J Stored)
Spark (<1 J Effective)
2
1
Halon Concentration, Vol %
Figure 21. Effect of Ignition Strength on Stoichiometric Methane-Air Explosions (Ref. 20).
The ignition configuration used by Hertzberg and colleagues (20) is similar to that used by NMERI
(Figures 22 and 23). The effective energy, £eff is a function of the volume of the test chamber, VD, and the pressure
rise generated by the spark, AP. The data set in Figure 22 most representative of the NMERT sphere is the labeled
curve. Although data are not reported above 15 J stored electrical energy, the curve of effective versus stored
energy begins to approach horizontal at that value, indicating that beyond 10 J stored energy, the effective spark
energy remains at about 1 J. This is the region (5 to 15 J stored energy) where Rangasamy (7), Silva Filho (8), and
Duartc (9), using propane as a fuel, reported that Halon 1301 inertion concentration became insensitive to higher
levels of ignition energy in their Master's Degree theses at Worcester Polytechnic Institute (Figure 24). This
suggests the possibility that the reported IC limit may have been a result of a limitation of the spark ignition
system, rather than the inciting capability of Halon 1301.
46
-------
Transfer Efficiencies: 71.4 %-
Maximum
2%
10%
25 %,
Circuit
Spark Gap
(6 mm)
o
Voltage, E
300 - 400
60-150
V0 = Chamber Volume
10-3
Iff1
10'
1
Stored Electrical Energy, 1/2 CE2, J
Figure 22. Effective Spark Energies as a Function of Stored Electrical Energy (adapted from
Fig. 2 of Ref. 22). Original © by The Combustion Institute, 1984. Reproduced with
permission.
47
-------
10 kv
Transformer
6 mm Gap
_L_ _i_ 2C0Q mfd
| Capacitors
165 v
Sphere
(7.9 I)
(a) NMERi (70 Joules Stored)
raohite Red
290 v —
525 mfd
Capacitors
Sphere
[5-6 I)
(b) FtNWAL (11 Jcuies Stored)
Figure 23. NMERI and Fenwal Ignition System Schematic.
48
-------
8
Fuel: Propane
6
4
~ Silva Filho (DC Spark)
O Varadharajalu (DC Spark)
A DAS (Pyrotechnic Match)
V Duarte (AC Spark)
2
0
0
40
20
60
80
100
120
140
Ignition Energy (JouIqs)
Figure 24. Halon 1301 Inertion Concentration vs. Ignition Energy, Propane (Ref. 9).
This point is farther demonstrated by the results of Zlochower and Hertzberg (21) of the U. S. Bureau of
Mines. A 1000 J pyrotechnic igniter was used as a source to ignite a Halon 1301/metliane mixture. The reported
concentration for inertion was 17 percent Halon 1301 at 10 percent mclhane concentration, and this was
potentially not the highest concentration required because sufficient data points to construct a full flammability
curve were not provided. Figure 25 is a comparison of the NMERI, Fenwal, and Bureau of Mines data for the
Halon 1301 inertion of methane. It indicates that, since both the flammability region and maximum inerting
concentration for the Fenwal tests were greater than those for NMERI, the possibility exists that more effective
energy was available in the Fenwal testing than the NMF.RI testing. In addition, Bartknecht (23) stated
"The higher the energy transferred from the ignition source to the surrounding gas mixture, the
wider the range of concentrations permitting autonomous flame propagation. Especially the
upper explosion limit will be moved toward higher gas concentrations."
This is illustrated in Figure 25 where the upper flammability limit is between 13 and 14 percent for the NMERI
tests, 15 percent for the Fenwal tests (and also the generally accepted value for methane), and 17.5 percent using
the Bureau of Mines higher energy source.
49
-------
-a Fenwal (11 J Stored)
-o- — NMERI ( 70 J Stored )
Bureau of Mines (f 000 J Effective)
20
15
10
5
Stoichiometric
Fuel/Air Concenlralion
0
10
5
15
20
0
Halon 1301, Vol %
Figure 25. Methane/Halon 1301 Inertion Concentrations, NMERI-Fenwal-Bureau of Mines.
Figure 23 represents a schematic of the Fenwal and NMERI ignition circuits. Differences between the
two methods include a graphite rod located between the electrodes in the Fenwal test set up, a higher charging
voltage for Fenwal (290 volts vs. 165 volts for NMERI), and tltc absence of a transformer in the Fenwal circuitry.
The Fenwal ignition method could potentially produce a more energetic spark for two reasons. The charging
voltage for the Fenwal testing was not boosted through a transformer as was the NMERI voltage, and was 290 volts
compared lo the NMERl's 165 volts. Accordingly, this would potentially raise the effective spark energy to a
higher level than achieved in the NMERI testing. Also, because of the corona discharge method used by Fenwal,
less energy was potentially expended to initiate the spark channel between the electrodes, and more energy was
available for ignition of the flammable mixture.
50
-------
Purilv of Methane
The chemical grade of methane, or other chemicals, is nol discussed in reports of past work; therefore, it
is difficult to determine if this has played a role in producing a discrepancy in test results.
Test Conditions
Test conditions for previously reported work indicate that all the studies were performed under similar
conditions, i.e., 25 °C and 1 atm. Therefore, test conditions appear to be an unlikely source of discrepancy.
Cleaning
Fenwal indicated that in all tests resulting in a pressure increase, the vessel was opened and cleaned by
wiping with a clean cloth and blowing away residue with air. After all other tests, the sphere was evacuated and
air introduced to clean the sphere. This test program determined that the sphere had to be opened and cleaned
with acetone between ever)' test. It was also opened and blown with compressed air if the test was aborted during
the loading process.
Mixing
Fenwal loaded the sphere with agent and fuel from a vacuum, and mixed the components as the air
entered the chamber, returning the mixture to atmospheric pressure. As shown in Appendix D, this test program
determined that the fan had to run for either one or two minutes to achieve acceptable repeatability. The mixing
techniques (methane is lighter than air, propane heavier) and fuel purity could conceivably be contributing factors
Test conditions, while not reported by Fenwal, would most likely not vary between the two fuels, nor would
physical properties of the sphere or cleaning techniques. This leaves the spark ignition as a probable cause of the
difference between the NMERI and Fenwal results.
51
-------
SECTION 5
CONCLUSIONS AND RECOMMENDATIONS
The results of the fire suppression and explosion prevention testing clearly demonstrated that those agents
containing either bromine or iodine exhibited superior effectiveness. Most of the bromine- and iodine-containing
candidates required slightly less or slightly more amounts of chemical on a storage volume basis (SVEq = 0.50 to
1.57) for fire suppression effectiveness, and between 1.0 and 1.9 times more agent compared to Halon 1301 on a
storage volume basis for inertion against propane explosions. When comparing the chemicals on a weight
equivalency basis, the other brominated and iodinated compounds required slightly more agent compared to Halon
1301 due to most of the chemicals having higher molecular weights than Halon 1301.
Results showed that small amounts of brominated or iodinated agents added to major components of
physically acting agents can result in a mixture which has a SVEq in the vicinity of 2.0, in some cases significantly
lower than the PAAs would have alone. Blends may offer some advantages in lowering the toxicity and cost of
some of the CAAs. However, questions still remain concerning the use and efficacy of blends because it is
speculated that the blend components may "separate" when dispersed in a realistic total flood application and not
provided the expected fire suppression or inerlion effectiveness. In addition, blends introduce logistics and
handling problems, unless an azeotropic mixture can be identified. As a result of these issues, single component
agents have been the major focus of this program. Nonetheless, the concept of blending chemicals has been
validated and may offer advantages to certain tradeoffs.
The results obtained from this effort represent laboratory-scale testing within controlled parameters.
However, care must be taken not to extrapolate results to larger scale without verification. The fire suppression
results reported herein highly agree with other research laboratories. The explosion prevention results for methane
and Halon 1301 do not correspond precisely with other testing; however, equipment and procedural differences
have been linked to the variations in test results. Nonetheless, the screening methods used here proved to be
reliable and repeatable, which allowed the relative ranking of a number of halocarbon agents. In addition, for
explosion prevention testing, since it has been shown that a stoichiometric fuel-to-air ratio may result in an
inerting concentration less than required to inert the mixture, testing at other fuel-to-air ratios should be
undertaken before setting design recommendations.
Although the laboratory-scale fire suppression and explosion prevention testing results identified several
highly effective candidates, several considerations were taken into account when recommending which chemicals
should be investigated further as candidates for replacing Halon 1301. Since this effort emphasized North Slope
52
-------
oil and gas facilities, the nature of the threat unique to these applications was of key importance. For the most
part, the North Slope facilities are considered occupied areas and are generally accessible to human inhabitants.
Although these areas are labeled as occupied areas, in fact they may actually be infrequently inhabited.
Nonetheless, because human occupancy is a major concern, the toxicity of the candidates must be sufficiently low
to allow for human exposure at safe levels. These issues were addressed in Reference 2 when the candidates were
prioritized into groups.
With the phase out of Halon 1301 at the end of 1993, the North Slope facilities will need to continue their
fire and explosion protection measures using other chemicals or technologies when current stocks of Halon 1301
are depleted. As a consequence, near-term replacements are needed now. This need for a near-lerm replacement
limits the number of chemicals which can be considered as candidates to only those which are available now or
will be available in the near future (Group 1, Table A-l, Appendix A). Also, a significant amount of toxicity
information must be known on the chemical to allow for its use in occupied areas. In addition, the candidate must
be approved by the Environmental Protection Agency under the Significant New Alternatives Policy (SNAP)
Program if it is to replace an ozone-depleting substance such as Halon 1301. This approval further limits the
number of candidates. Finally, under Alaskan codes, as in many states, the State Fire Marshal governs the fire and
explosion technologies that arc implemented within Alaska. The Fire Marshal utilizes the Uniform Building
Codes (UBC) and state fire codes which are based on National Fire Protection Association (NFPA) standards to
determine if fire and explosion protection methods are safe. NFPA, which develops standards for fire and
explosion protection, has formulated a committee to establish a standard for near-term Halon 1301 replacement
agents (NFPA 2001). This standard will serve essentially the same function as NFPA 12A in specifying safe
methods for utilizing near-term replacement agents. Since a fire marshal may use the NFPA standards to
determine safe fire protection methods and since NFPA is preparing a standard on selected ncar-lcrm chemicals,
the candidates recommended for further investigation as near-term Halon 1301 replacements must be selected from
those chemicals being considered in the NFPA 2001 standard. Table 17 lists the chemicals included in NFPA
2001.
AH of the chemicals presented in Table 17 were evaluated under the present effort. IICFC Blend A was
also tested, although the results for the blend were not reported in Section 3 because no difference in effectiveness
was noted compared to the major component (HCFC-22). The extinguishment concentration for HCFC Blend A
was measured to be 11.2 percent in the NMERi 5/8-scale cup burner and the ICst's for propane and methane were
18.0 and 13.3 percent, respectively. The candidates being included in the NFPA 2001 standard did not exhibit the
best performance for fire suppression or explosion prevention effectiveness in the present test evaluation. However,
these agents are or will be available in bulk and toxicity information exists which allows one to make a
determination of their safe use. NFPA has established guidelines within the standard to specify which agents can
be used in occupied areas. Accordingly, the fire suppression or inertion design concentration must be below the
53
-------
Table 17. NFPA 2001 HALON 1301 (HALOCARBON) REPLACEMENT CANDIDATES*
Number designation
Chemical name
Chemical formula
FC-3-1-10
Perfluorobutane
C4F10
HBFC-22B1
Bromodifluoromethane
CHF2Br
HCFC Blend A
Dichlorotrifluoroethane (HCFC-123, 4.75%)
chci2cf3
Chlorodifluoromethane (HCFC-22, 82%)
chcif2
Chlorotetrafluoroethane (HCFC-124, 9.5%)
CHCIFCF3
lsopropenyl-1 -methylcyclohexene (3.75%)
HCFC-124
Chlorotetrafluoroethane
CHCIFCF3
HFC-125
Pentafluoroethane
chf2cf3
HFC-227ea
Heptafluoropropane
CF3CHFCF3
HFC-23
Trifluoromethane
CHF<*
* Reference 24.
cardiac sensitization no observable adverse effect level (NOAEL) for agents to be used in occupied areas. The
NOAEL is determined in dogs where the chemical is administered by inhalation for a period of minutes and a dose
of epinephrine (adrenaline) is then given to "challenge" the dogs' cardiac sy stems. The NOAEL level for the
NFPA 2001 agents are listed in Table 18 along with the design concentrations for inertion and fire suppression.
Agents may only be used if their design concentration is below the NOAEL. As seen from Table 18, only
FC-3-1-10 and HFC-23 would be acceptable for use in occupied areas being designed for protection against
propane explosions and fires. HFC-227ea would also acceptable for use in occupied areas for fire protection only.
HCFC Blend A would be acceptable for use in occupied areas for fire protection if the manufacturer's design
concentration, which is significantly lower than the traditional design concentration, is used rather than that
determined under the normal condition. Since the North Slope facilities utilize Halon 1301 for fire suppression in
some cases independent of explosion protection, it is recommended that the near-term agents FC-3-1-10, HFC-23,
HFC-227ea, and HCFC Blend A be further investigated at a larger scale to determine their fire suppression and
explosion prevention effectiveness in more realistic scenarios.
Of the other agents that are not commercially available in the near-term, it is recommended that other
PAAs be investigated only at a low level of effort and that the main emphasis for future agents focus on CAAs.
These agents offer the greatest potential for replacing Halon 1301 since they have near or better effectiveness. Of
the CAAs, the fluoroiodocaibons (in particular FIC-13I1, the iodinated analog of Halon 1301) appear to be the
most promising from an effectiveness perspective. The haloalkenes such as BTFB also exhibited superior
effectiveness. Unfortunately, for both the fluoroiodocarbons and haloalkenes, as the molecular weight of the
54
-------
TABLE 18. CARDIOTOXICITY VALUES AND INERTION DESIGN CONCENTRATIONS FOR NFPA
2001 AGENTS*
Number designation
Design concentrationt
NOAELt, (vol %)
LOAEL§, (vol %)
Inertion, propane Fire suppression
(vol %) (vol %)
FC-3-1-10
10.6
6.0
40
>40
HBFC-22B1
9.7
5.3
2
3.9
HCFC Blend A
19.8
7.2 (13.4)#
10
>10
HCFC-124
13.9
9.8
1.0
2.5
HFC-125
16.2
11.3
7.5
10
HFC-227ea
13.1
7.6
9.0
10.5
HFC-23
21.5
15.1
50
>50
* Reference 24.
t Minimum design concentration is determined by the laboratory test plus 10 percent safety factor for
inertion and 20 percent safety factor for fire suppression. Values from this work.
* NOAEL = Highest tested concentration at which no adverse cardiac effects were observed. Adverse
cardiac effects generally refer to multiple irregular heart beats, arrhythmias, or death.
§ LOAEL = Lowest tested concentration at which adverse cardiac effects were observed.
* Manufacturer's design concentration determined by UL Canada as provided in NFPA 2001.
Design concentration based on this work given in parentheses.
chemical increases, the boiling point also increases and the volatility decreases. Therefore, chemicals with boiling
points above 20-30 °C may require misting to disperse them in a total flood manner. Of the CAAs investigated in
this effort, only F1C-13I1 is available in sufficient quantities to allow for larger scale testing. However, the price of
the material makes field-scale testing cost prohibitive under the present project. Future efforts should investigate
FIC-13I1 as a Halon 1301 replacement agent for North Slope and related applications.
In summary, it is recommended that:
1. FC-3-1-10, HFC-23, IIFC-227ea, and HCFC Blend A should be evaluated at the field scale as
near-term Halon 1301 replacement agents to determine their fire suppression and explosion prevention
capabilities.
2. Fluoroiodocarbons and bromofluoroalkcncs should be further investigated as far-term Halon 1301
replacement agents emphasizing global environmental concerns (short atmospheric lifetimes, zero OOP, and low
GWP), stability and shelf-life determinations, fire suppression and inertion effectiveness at large-scale, toxicity,
and manufacturability issues.
3. FIC-13I1 should undergo an immediate and rigorous evaluation to determine its near-term
potential as a Halon 1301 replacement focusing on stability, global environmental characteristics (atmospheric
55
-------
lifetimes, ODP, and GWP), acute toxicity and cardiac sensitization, large-scale fire and explosion prevention
capabilities, material compatibility, as well as recruiting manufacturing sources that will produce the material in
bulk at reasonable prices.
4. Other chemical classes besides halocarbons should be investigated to determine their possible use
as total flood fire and explosion prevention agents.
56
-------
REFERENCES
1. Pope, P. R., "North Slope Oilfield Fire Protection: The Halon Issue," Proceedings of the International
Conference on CFC and Halon Alternatives, Washington, DC, October 1989.
2. Skaggs, S. R., Tapscott, R. E., Nimitz, J. S., and Moore, T. A., Low Ozone-Depleting Halocarbons as
Total-Flood Agents: Volume 1-Candidate Survey, EPA-600/R-95-150a, Air Pollution and Control
Division, National Risk Management Research Laboratory, Research Triangle Park, NC, September 1995.
3. Barat, R. B., Sarafim, A F., Longwell, J. P., and Bozzelli, J. W., "Inhibition of a Fuel Lean Ethylene/Air
Flame in a Jet Stirred Combustor by Methyl Chloride: Experimental and Mechanistic Analysis,"
Combustion Science and Technology, Vol. 74, pp. 361-378, 1990.
4. Walters, E. A., and Molina, M., "Global Environmental Characteristics of Second-Generation Halon
Replacements," Proceedings of the Halon Alternative Technical Working Conference 1993, pp. 31-36,
Albuquerque, NM, May 11-13, 1993.
5. Dalzell, Warner G., A Determination of the Flammability Envelope of Four Ternary Fucl-Air-Halon
1301 Systems, Fenwal Report No. PSR-624, Fenwal Incorporated, Ashland, MA, January 1977.
6. Das, A., Relationship Between Ignition Source Strength and Design Jnerting Concentration of Halon
1301, M. S. Thesis, Worcester Polytechnic Institute, Worcester, MA, 1986.
7. Rangasamy, V„ Relationship Between Strength of DC Spark Ignition Source and Design Inerting
Concentration of Halon 1301 for Stoichiometric Propane/Air Mixtures, M. S. Thesis, Worcester
Polytechnic Institute, Worcester, MA, 1987.
8. Silva Filho, J., Relationship Between Strength of DC Spark Ignition Source and Design Inerting
Concentration of Halon 1301 for Stoichiometric Propane-Isohutane-Air Mixtures, M.S. Thesis, Worcester
Polytechnic Institute, Worcester, MA, 1988.
9. Duarte, L., Relationship Between Strength of AC Spark Ignition Source and Design Inerting
Concentration of Halon 1301 for Stoichiometric Propane-lsobutane-Air Mixtures, M.S. Thesis,
Worcester Polytechnic Institute, Worcester, MA, 1989.
10. Hirst, B., and Booth, K., "Measurement of Flame-Extinguishing Concentrations," Fire Technology,
Vol. 13, No. 4,1977.
11. Moore, T. A., Mcxire, J. P., Nimit/,, J. S., Lee, M. E., Beeson, H. D., and Tapscott, R. E., Alternative
Training Agents, Phase II - Laboratory-Scale Experimental Work, ESL-TR-90-39, Vol. 2 of 4, Tyndall
Air Force Base, FL, August 1990.
12. Zabetakis, M. E., "Flammability Characteristics of Combustible Gases and Vapors," Bulletin 627, Bureau
of Mines, Pittsburgh, PA, 1964.
13. Weast, R. C., Editor, Handbook of Chemistry and Physics, 67th Edition, CRC Press, Boca Ralon, FL,
1986.
14. Sheinson, R. S., and Driscoll, D. C., "Fire Suppression Mec hanisms: Agent Testing hy Cup
Burner," Proceedings of the International CFC and Halon Alternatives, Vol. II, Washington, DC,
October 10-11,1989.
57
-------
15. Robin, M. L., "Evaluation of Halon Alternatives," Proceedings of the Halon Alternatives Technical
Working Conference, pp. 16-38, Albuquerque, NM, 1991.
16. Fletcher, N., Wild, J. D., and Winterton, N., "Clean Agent Fire Extinguishants: A Low Ozone Depleting
Potential Transitional Substitute'," 1991 International Conference on CFC and Halon Alternatives,
Baltimore, MD, 1991.
17. Adcock, J. L., Mathnr, S. B., Huang, H-Q, Mukhopadhyay. P., and Wang, B-H, "Fluorinated Ethers:
A New Family of Halons?," Proceedings of the Halon Alternatives Technical Working Conference,
pp. 83-96, Albuquerque, NM, 1991.
18. Senccal, J. A., "Halon Replacement Chemicals: Perspectives on the Alternatives," Proceedings of the
International Telecommunications Fire Protection Sytnposium, New Orleans, LA, 1992.
19. Ford, C. L., "An Overview of Ilalon 1301 Systems," HalogenatedFire Suppressants, ACS Symposium
Series Number 16, Gann, R. G., Editor, American Chemical Society, Washington, DC, 1975.
20. Hertzberg, M., Cashdollar, K. L., Lazzara, C. P., and Smith, A. C., "Inhibition and Extinction of Coal
Dust and Methane Explosions," Bureau of Mines Report of Investigation, Bulletin 8708, Pittsburgh, PA,
1982.
21. Zlochower, I. A., and Hertzberg, M., "The Inerting of Methane-Air Mixtures by Halon 1301 (CF3Br) and
Halon Substitutes," Proceedings of the Ilalon Alternative Technical Working Conference 1991, pp. 122-
130, Albuquerque, NM, 1991.
22. Hertzberg, M., Conti, R. S., and Cashdollar, K. L., "Spark Ignition Energies for Dust-Air Mixtures:
Temperature and Concentration Dependencies," Twentieth Symposium (International) on Combustion,
pp. 1681-1690, The Combustion Institute, Pittsburgh, PA, 1984.
23. Bartknecht, W., Explosion, Springer-Verlag, New York, NY, 1981.
24. NFPA 2001 Standard on Clean Agent Fire Extinguishing Systems. National Fire Protection Association,
Quincy, MA, 1994 Edition.
58
-------
APPENDIX A
CANDIDATE AGENTS
Tables A-l and A-2 contains the listing of Halon 1301 replacement candidates by NMERI-defined groups.
Table A-l is for PAAs and Table A-2 is for CAAs. Group 1 contains those agents currently in production with low
toxicities. Group 2 contains agents with incomplete toxicity data, but which are believed to be relatively non-toxic,
and are available at least in small quantities. Group 3 contains agents with unknown or suspected high toxicities
or for which no data are available. Group 4 contains chemicals that are deemed unacceptable for further
consideration based on either toxicological or environmental properties.
59
-------
TABLE A-1. PAA CANDIDATE GROUP LIST OF HALON 1301 REPLACEMENTS
Group
Halocarbon
No.
Formula
Comments
1
FC-14
cf4
unknown tox. but expected to be low
Recommended
HCFC-22
chcif2
low toxicity
for
HFC-23
chf3
long atmospheric lifetime
Lab-testing
HFC-32
ch2f2
flammable
FC-116
CF3CF3
low toxicity
HCFC-124
CHCIFCF3
low toxicity
HFC-125
CHF2CF3
limited availability
HFC-134a
ch2fcf3
low toxicity
HCFC-142b
ccif2ch3
flammable
HFC-143a
ch3cf3
flammable
HFC-152a
chf2ch3
very low toxicity
FC-218
cf3cf2cf3
low toxicity
HFC-227ea
cf3chfcf3
unknown tox. but expected to be low
FC-3-1-10
c4f10
toxicity expected to be low
FC-C318
c4f8
decomposition products may be toxic
2
HCFC-124a
chf2ccif2
unknown tox. but expected to be low
Recommended
HFC-134
chf2chf2
unknown tox. but expected to be low
for
HFC-236ea
cf3chfchf2
unknown tox. but expected to be low
Lab-testing
HFC-236fa
cf3ch2cf3
acute toxicity known
HFC-245cb
cf3cf2ch3
unknown tox. but expected to be low
3
HCFC-142a
chcifch2f
no data
Unavailable
FC-C216
cf2cf2cf2
no data
Commercially
HCFC-226ba
cf3ccifchf2
no data
HFC-227ca
cf3cf2cf2h
acute toxicity known
HFC-C234
cf2cf2ch2
no data
HCFC-235cc
ch2fcf2ccif2
acute toxicity known
HCFC-235da
cf3chcichf2
anesthetic at low concentrations
HCFC-235ca
chf2cf2chcif
anesthetic at low concentrations
HFC-236cb
ch2fcf2cf3
no data
HCFC-244db
cf3chcich2f
no data
HCFC-244bb
cf3ccifch3
anesthetic at low concentrations
HCFC-244ca
chf2cf2ch2ci
no data
HCFC-244fb
ccif2ch2chf2
acute toxicity known
HCFC-244cb
ch2fcf2chcif
no data
HCFC-244da
chf2chcichf2
anesthetic at low concentrations
HFC-245eb
cf3chfch2f
no data
HFC-245ea
chf2chfchf2
no data
HFC-245ca
chf2chf2ch2f
no data
HFC-254eb
cf3chfch3
probably flammable, no data
HFC-254ea
chf2chfch2f
probably flammable, no data
HFC-254fa
chf2ch2chf2
probably flammable, no data
HFC-254ca
ch2cf2ch2f
probably flammable, no data
HFC-254cb
chf2cf2ch3
probably flammable, no data
4
HCFC-31
ch2cif
cancer-causing in lab animals
Unacceptable
60
-------
TABLE A-2. CAA CANDIDATE GROUP LIST OF HALON 1301 REPLACEMENTS
Group
Halocarbon
No.
Formula
Comment
1
FIC-1311
CF3I
unknown tox. but expected to be low
Recommended
FIC-11511
cf3cf2i
unknown tox. but expected to be low
for
FIC-217cal1
cf3cf2cf2i
unknown tox. but expected to be low
Lab-testing
FIC-31911
CF3CF2CF2CF2l
unknown tox. but expected to be low
FIC-5-1-13al1
CF3(CF2)4CF2I
unknown tox. but expected to be low
BTFB*
CH2=CHCF2CF2Br
unknown tox. but expected to be low
2
FIC-7-1-17al1
CF3
-------
APPENDIX B
HALOCARBON NAMING AND NUMBERING RULES
Many refrigerants, all halon fire extinguishing agents, and many substitutes for these materials are
halocarbons, gaseous or liquid chemicals that contain carbon and one or more of the elements fluorine, chlorine,
bromine, and iodine. Some of these chemicals also contain hydrogen. This appendix provides information on
naming halocarbons used as refrigerants, halons, and replacements.
THE HALOCARBON NUMBERING SYSTEM
Since these materials have complicated chemical names, it has become a general practice to give
commercial halocarbons a systematically derived number. The Halocarbon Numbering System was developed by
DuPont for Freon chemicals in the late 1930s. The system was later expanded and formalized in a standard
(ANSI/ASHRAE Standard 34) by the American Society of Heating, Refrigerating, and Air-Conditioning Engineers
(ASHRAE) and the American National Standards Institute (ANSI). Note that this standard covers much more
than just halocarbons, since many non-halocarbon refrigerants exist. Halocarbon Numbers normally contain 2 or 3
digits and, for pure halocarbons, the number is related to the chemical formula. In some cases, lower case letters
are added after the number to distinguish between different compounds that have the same atom counts (and,
therefore, the same number) but different atom arrangements. Increasingly, halocarbon numbers are not being
applied just to refrigerants, but also to halons, halocarbon solvents, and replacements for these materials.
The refrigeration industry usually places a letter "R" in front of the Halocarbon Number; however,
regulatory statutes are increasingly using a series of letters denoting the type of compound. For example,
compounds containing only chlorine and fluorine (in addition to carbon) have numbers preceded by "CFC," which
stands for "chlorofluorocarbon." Though not universally accepted or standardized, other prefixes arc being
increasingly used. Below (Table B-l) are listed some of the prefixes used. Note that two prefixes are used for
perfluorocarbons: "FC" and "PFC."
62
-------
TABLE B-1. PREFIXES FOR HALOCARBON NUMBERS
Prefix
Elements in chemical
Chemical family
BC
Br, C
Bromocarbon
BCC
Br, CI, C
Bromochlorocarbon
BCFC
Br, CI, F, C
Bromochlorofluorocarbon
BFC
Br, F, C
Bromofluorocarbon
CC
CI, C
Chlorocarbon
CFC
CI, F, C
Chlorofluorocarbon
HBC
H, Br, C
Hydrobromocarbon
HBCC
H, Br, CI, C
Hydrobromochlorocarbon
HBCFC
H, Br, CI, F, C
Hydrobromochlorofluorocarbon
HBFC
H, Br, F, C
Hydrobromofluorocarbon
HC
H, C
Hydrocarbon
HCC
H, CI, C
Hydrochlorocarbon
HCFC
H, CI, F, C
Hyd roch lorof I uorocarbon
HFC
H, F, C
Hydrofluorocarbon
HFE
H, F, C, 0
Hydrofluoroether
FC
F,C
(Per)fluorocarbon
FE
F, C, O
(Per)fluoroether
PFC
F,C
Perfluorocarbon
PFE
F, C, O
Perfluoroether
In the Halocarbon Numbering System, the first number gives the number of carbon atoms minus one,
followed by the number of hydrogen atoms plus one and the number of fluorine atoms (e.g., the number can be
represented by C-l, H+l, F). The number of chlorine atoms is not directly given. When the first digit is zero (for
the case of one-carbon compounds), it is omitted. Note that the halocarbon number for all one-carbon compounds
has only two digits: CFC-11, CFC-12, HCFC-22, etc.
For example, the halocarbon CC12F2 has one carbon atom (the first digit is equal to 1 -1 = 0, and it is
dropped), no hydrogen atoms (the second digit [now the first] is equal to 0 + I = l), and two fluorine atoms (the
third digit is 2). Therefore, this compound is CFC-12. The compound CHC1FCF3 has two carbon atoms (the first
63
-------
digit is 2 -1 = 1), one hydrogen atom (the second digit is 1 + 1 = 2), and four fluorine atoms (the third digit is 4).
This compound is HCFC-I24.
It is also possible to work backwards from the number to the formula. For example, the halocarbon
number HCFC-123 has a first digit of 1. Therefore, the number of carbon atoms is two since 2-1 = 1. The second
digit is 2, and there is only one hydrogen atom since 1 + 1=2, The third digit is 3 and this gives the number of
fluorine atoms directly. The number of chlorine atoms was not directly given, but it is indirectly given. This is
because the total number of atoms is known. The total number of atoms other than carbon in the compounds of
interest here is given by 2n + 2, where n is the number of carbon atoms. Thus, the number of atoms other than
carbon is 4 for one-carbon compounds, 6 for two-carbon compounds, and 8 for three-carbon compounds. Since
HCFC-123 has two carbon atoms, there must be six other atoms. Only four atoms have been accounted for: one
hydrogen and three fluorine atoms. There must be two chlorine atoms, and working backwards the formula
C2HC12F3 is obtained.
The formula is usually written to show which atoms are attached to which carbon atoms. There are three
ways to attach these atoms together: CHC12CF3, CHC1FCC1F,, and CHF2CCI2F. No other way exists. Molecules
having different arrangements of atoms are called "isomers." When writing a formula this way, the atoms
following each symbol "C" are the atoms attached to that carbon atom. For example. CHC12CF3 has one carbon
with a hydrogen and two chlorine atoms attached to it and a second carbon atom with three fluorine atoms attached
to it. (Note that this is the same compound as CCl2HCF3 and CF3CHC12, where the atoms have just been
rearranged and the molecule switched end-for-end.)
Three different arrangements of atoms arc distinguished by the assignment of halocarbon numbers:
The lower case letter added is based on the symmetry of the molecule. For two-carbon compounds,
absence of a letter indicates the most symmetrical isomer, while an "a" indicates the next most symmetrical isomer,
"b" the next, and so on. The symmetry is determined by adding the atomic masses of the atoms attached to each
carbon atom. The isomer with the smallest difference in the sum of the masses on the two carbon atoms receives
no letter, the next smallest difference receives an "a", the next a "b," etc. The approximate masses of the atoms are
1 for hydrogen, 19 for fluorine, and 35.5 for chlorine.
CHC12CF3
CHC1FCC1F2
CHF2CC12F
HCFC-123
HCFC-123a
HCFC-123b
64
-------
It is not essential that one know all of this, but there arc a couple of items of interest. First, one knows
immediately that the replacement (or proposed replacement) refrigerants HCFC-142b, HFC-134a, and HFC-152a
must exist as two or more isomers. Otherwise, there would be no reason to add a small letter to their halocarbon
number. Second, there are no geometric isomers possible for one-carbon compounds. Thus, no halocarbon number
having two digits (meaning only one caibon atom) has a small letter attached to it.
HALON NUMBERS
Halon fire extinguishing agents, particularly those containing bromine, are often named using an alternate
numbering system, the Halon Numbering System. This designation system is also sometimes used for materials
other than fire extinguishants. The Halon Numbering System designation lists, in order, the number of carbon,
fluorine, chlorine, and bromine atoms in a molecule. NMER1 has extended this convention to add a 5th digit to
designate the number of iodine atoms, if needed. Trailing zeros are dropped. Thus, Halon 1211 is CBrClF2 and
Halon 1301 is CBrF3. The Halon Numbering System cannot specify isomers. Thus, both CBrF2CBrF2 and
CF3CBr;F are designated Halon 2402. The Halon Numbering System is not used for cyclic or unsaturated
compounds.
TRADENAMES
Some trade names are based on either the halocarbon number or the halon number of the compound;
others are just "invented" by the company. In many cases, a pure halocarbon refrigerant is named using the
halocarbon number with a trade name prefix. For example, Elf Alochem uses "Forane®" (e.g., Forane® 134a) and
ICI uses "Klea®" (e.g., Klea® 32).
The companies commercializing halon replacements tend to use different systems. Great Lakes uses
9) ® ®
"FM" (Fire Master ) numbers that have no relationship to halocarbon or halon numbers. FM 100 and FM 200
are HBFC-22B1 and HFC-227ea, respectively. Similarly, the North American Fire Guardian "NAF®"
designations (e.g., NAF S-III ) for their blends have no systematic relationship to composition. Du Pont uses the
halon number with the prefix "FE" (for 'Tire Extinguishant"). FE-13 is Halon 13 (HFC-23), FE-241 is Halon
241 (HCFC-124), and FE-232® is Halon 232 (HCFC-I23). 3M Company uses a modified version of the halon
system with the prefix "PFC" (for "pcrfluorocarbon"). PFC-410 is FC-3-1-10 and PFC-614 is FC-5-1-14.
65
-------
APPENDIX C
EQUIPMENT DESCRIPTION
Test objectives were achieved using the following Hewlett-Packard electronic components:
(1) HP 86 computer - runs customized BASIC software and serves as the master controller for the listed
devices as well as for a dot matrix printer and for a multipen plotter. All communications between the HP-86 and
the other devices arc via an IEEE 488 interface.
(2) HP 3852A data acquisition and control (D/C) card cage system with the high speed voltmeter option
44702A cards installed - used to digitize data and monitor the signal that starts a test. This unit has its own
microprocessor and runs subroutines that have been downloaded from the HP 86 computer. In addition, this unit
has several built-in timing circuits.
(3) HP 3488 switch and control (S/C) unit with the 44471A single-pole switch option and the 44470A
double-pole switch option installed—the single-pole switches provide control functions and the double-pole
switches act as multiplexers by connecting the outputs from different devices to the voltmeter input of the 3852A to
monitor voltages.
(4) HP 82906A dot matrix printer - used to print program listings, test documentation, and pressure data.
(5) HP 7475 multipen plotter - used to make graphs of data.
(6) HP 9122 dual disk drive unit - stores data and programs on 3.5-inch floppy disks.
Additional equipment consisted of a capacitor charge/discharge system built at NMERI and normally used
to power Mutual Inductance Particle Velocity (MIPV) transducers. The MIPV box had space for up to six
2000-microfarad capacitors connected in parallel and could be reconfigured to generate sparks of various energy.
It also had a relay for charging these capacitors and a silicon controlled rectifier (SCR) for capacitor discharge.
A manual/auto control box (MACB) was specifically constructed for this test series. This box contained
miniature 5-volt printed circuit board relays, push buttons for manual control functions, light-emitting device
(LED) indicators to monitor control functions, a terminal strip to provide for the connection of batteries and
transducers, and a 10-volt regulated power supply. The purpose of the MACB box was to provide manual control
capabilities, such as charging and discharging the spark, and to isolate all functions so that the discharging of the
capacitors and the resulting high voltage could not feed back into any other control or data circuits.
66
-------
Batteries and power supplies were used to provide power for operating relays and charging capacitors.
Kulite™ and Druck transducers were used to measure the loading and explosion pressures respectively. Amplifiers
were used to drive long lines and improve the signal-to-noise ratios for all transducers; the Druck amplifier was
contained in the device and an external amplifier was used for the Kulite™.
OVERVIEW
Test objectives were accomplished by running customized BASIC software on the HP 86 computer and by
making a few simple measurements to verify proper operation of the data acquisition and control system. Software
was written at NMERI in the BASIC computer language and consisted of a program specifically for testing plus
other software written to provide for auxiliary operations.
SPECIFIC OPERATION
Monitoring the Loading of Agents. Fuel, and Air
The loading of agent, fuel, and compressed air into the sphere is monitored by a Druck 2.5-psi pressure
transducer, placed in the tubing matrix. A valve has been placed ahead of the transducer. This valve, when
closed, will protect the transducer from possible damage resulting from the pressure in the sphere caused by the
explosion. The output of this gage is amplified by the gage itself and fed to the input of the system voltmeter
through a switch in the HP 3488 S/C unit operating under software control.
After software has set up the proper control and monitoring circuit path, a loop structure is entered
causing the following sequence:
1. System voltmeter takes a single reading.
2. Reading is transferred to HP 86 computer.
3. Computer converts millivolts to psi using a third-order polynomial equation whose constants
were determined by laboratory calibration values.
4. This value is displayed on the operator's monitor and on a second monitor visible to the person
loading the sphere.
5. Program waits user-specified amount of time.
6. Cycle repeats.
The technician has been provided with target psi levels for fuel, agent, and compressed air, which are also
shown on the screen.
67
-------
Charge Capacitors
The user enters the desired energy, in joules, and the value of the capacitors in the MIPV box prior to test.
The HP 86 computer calculates the equivalent voltage for this value of energy and corrects it for a 15 percent
energy loss in the high voltage transformer. This value and the voltage monitoring routine are downloaded into
the 3852A immediately after the control and monitoring switches have been set in the 3488A.
Under software control, the computer closes two switches in the 3488A S/C unit. One switch activates an
isolation relay in the MACB box causing a 24-volt relay in the MIPV box to close. The 24-volt relay connects the
output of the high voltage power supply across the capacitors causing them to be charged. A series resistor
provides isolation between the charging capacitors and the high voltage power supply and controls charging time.
The other switch connects the voltmeter in the 3852A unit across the same input terminals of the MIPV box
through a voltage divider. The ratio of the voltage divider is 21.19 to 1. This is necessary since the voltmeter can
only measure up to 10.24 volts, and charge voltages are typically greater than 150 volts.
After the switches have been closed and the data and subroutine downloaded to the 3852A, the HP 86
computer waits two seconds to allow the charging process to begin before calling the charging subroutine in the
HP 3852A. The HP 86 computer then waits for a data entry from the HP 3852A. After the calculated charge
voltage has been attained, the HP 3852A D/C unit sends the last measured value back to the HP 86 computer,
which converts it to equivalent joules and displays it while the computer system is waiting for a signal to start the
test.
Start Test Signal
To start a test sequence, software causes the HP 86 computer to download a command to the 3852A D/C
unit. The 3852A then waits for a dry closure (short circuit) input to its 'Event In' terminal, which occurs when the
start button is pushed. When this signal is received, the 3852A signals the HP 86 computer that this command has
been received. This starts the completion of the test sequence.
Immediately upon acknowledgment that the test sequence has started, the HP 86 computer initiates the
data acquisition subroutine in the 3852A. This subroutine takes 1000 data points (limited by the internal memory
of the HP 3852A D/C unit) at the interval specified by the user in the initial setup portion of testing software.
Nominally these intervals arc 1 to 3 ms, which allows the recording of 1 to 3 sec of data.
68
-------
Since it lakes the HP 3852A approximately 400 to 450 ms to configure the voltmeter before running the
subroutine, the HP 86 computer must wait this amount of time before continuing with additional control functions
to ensure that the voltmeter is acquiring data from the transducer when test phenomena are occurring.
The transducer used to monitor explosion data is a Kulite™ model XTM-190, 0 to 50 psi. This transducer
is a sealed gage type. Although the overpressures expected can rise to 105 psi, this gage has been calibrated to
100 psi and has provided data consistent with higher pressure transducers used in preliminary testing. The
transducer is located in the air intake/transducer manifold adjacent to the loading transducer.
Discharge Capacitors
While the HP 3856 D/C is acquiring data, the HP 86 computer sends a command to the HP 3488A S/C
unit to close a switch. The closing of this switch causes an isolation relay in the MACB box to close, which places
12 volts across the gate of a silicon-controlled rectifier (SCR) in the MIPV box. This allows the capacitors in the
MlPVbox to discharge into the primary winding of a high voltage transformer. The secondary of this transformer
is connected to two electrodes that have been placed inside the sphere at a 6-mm separation. As the capacitors
discharge, a single arc forms across the electrodes. The voltage generated will depend on the amount of energy
specified and value of charging circuit capacitance.
Data Conversion
At the completion of the test, a file for header information and raw voltage data is created and the data
stored. An option is given to convert the voltage data to pressure data using the conversion factor. After
conversion, all data are stored in a separate pressure data file.
Printing/Plotting
Data can now be printed on the HP printer. Also, an option to plot the data is presented. A routine to
determine the maximum and minimum pressures is employed, and these pressures arc displayed on the screen.
Plot parameters such as the maximum and minimum values of the axes and the time of plot can be determined, and
the plot is produced on the screen or in hard copy using the inullipen plotter.
69
-------
APPENDIX D
EXPLOSION PREVENTION TESTING PROCEDURES
TEST METHOD DEVELOPMENT AND VALIDATION
Prior to testing of replacement agents, adequate preliminary testing and validation was accomplished to
ensure reliable results. Only after the method was established and found to be repeatable and reliable, Halon 1301
was tested to obtain baseline data to which the replacement agents were compared.
Ignition Source Strength
The energy content in the ignition spark as well as the type of source can significantly influence the
minimum concentration of agent required to inert the mixture. Figure 24 in Section 4 illustrates results from three
researchers using different ignition techniques to inert stoichiometric mixtures of propane with Halon 1301. As
shown, the ICS{ is proportional to the ignition energy level below a minimum energy, while above that level, the
ICst is energy independent. This phenomenon was investigated by testing Halon 1301 at increasing ignition spark
strength (40, 70, and 100 J). It was found that for the system used in the present testing, the ICS( was apparently
independent of spark energy above 40 J. This was because of the effective energy phenomenon. Therefore, to
allow a margin of safety, 70 J was selected as the ignition spark energy for the remaining testing.
Mixing
During preliminary testing, a discrepancy was noted in the explosive overpressure results depending on
the order in which the agent and fuel were added to the sphere. For example, a 5 percent concentration of Halon
1301 inerted a stoichiometric fuel-to-air mixture when the agent was loaded first, but not when the fuel was loaded
first. It was determined that a stratification of the three gaseous components existed within the sphere and that a
homogeneous mixture was required for repeatability. A 4-inch square box fan was, therefore, installed in the
bottom of the sphere to mix the components. The fan was activated for one minute when propane was the fuel and
for two minutes when methane was the fuel, improving repeatability between tests. The increased time for
methane was instituted due to inconsistent overpressures resulting from the 1-minute mixing period and may have
70
-------
resulted from the lower density of methane. All agent results were obtained from tests employing the noted mixing
procedures.
Cleaning
During preliminary testing, the sphere was not physically disassembled and cleaned. The by-products of
combustion were evacuated and the sphere flushed with compressed air. This procedure was followed throughout
the peak overpressure testing of the fuels alone. During baseline Halon 1301 testing, however, inconsistencies
among results from successive tests were noted. It was surmised that all materials were not completely removed by
flushing the sphere with air; therefore, the sphere was opened after each test and wiped out inside with acetone.
Although this increased the time to conduct each test, data repeatability was increased, and this procedure was
followed for all succeeding tests.
Test Methods And Procedures
Tests were conducted inside a laboratory fume hood, with exhaust gases passing through a ctyotrap before
being released to the environment. The following test procedures were used (specific equipment and software
operations are described in Appendix C):
1. The HP DAS was turned on and the computer program loaded. Test information, including the
test name; fuel type and partial pressure; agent type; concentration; and partial pressure; and test date and number,
was entered. The transducer power supplies were turned on, and the transducers were allowed to warm up for
1/2 hour prior to testing.
2. All testing was performed at a pressure corresponding to standard sea level (101 kPa [14.7 psi]
about 17.9 kPa [2.6 psi] above the normal ambient pressure in Albuquerque). To ensure repeatability and simplify
calculations, a correction between local atmospheric pressure and sea level was determined at least twice per test
day. A barometer located adjacent to the sphere provided the local atmospheric pressure.
3. After the transducers were powered for 1/2 hour, the amplifier gain and excitation voltage were
measured and recorded. The agent to be tested was connected to the upper pipe penetration. Fuel and agent were
bled as required to ensure that no air was trapped in the lines.
4. Prior to the first test of the day, the capacitors were charged to the required voltage and
discharged to check the operation of the spark system. The chamber was assembled and evacuated to
approximately -15 inches Hg„ and allowed to stand for several seconds to check for leaks.
71
-------
5. The correct amounts of agent, fuel, and air were calculated for the required percentage of agent
according to the partial pressure method (Appendix E).
6. The valve connecting the lower range loading transducer to the chamber was opened and the
transducer zeroed to atmospheric. A monitor next to the sphere displayed internal pressure in the sphere and the
desired partial pressures of agent, fuel, and air. The operator added the components to the required pressures by
controlling the input valves on the sphere.
7. After all components were loaded into the sphere, the valve connecting the loading transducer
was closed. The mixing fan was turned on for one minute (propane) or two minutes (methane) to ensure that the
components were completely mixed. After the fan was turned off, a delay of 30 seconds ensured that the mixture
returned to quiescent conditions.
8. The internal sphere temperature was recorded, and the desired spark energy was entered into the
computer. The capacitors were charged.
9. When the capacitors were charged to the voltage required to produce the desired energy, the
computer indicated that the test was ready to begin. A push button discharged the capacitors across the electrodes,
the explosion (if any) occurred, and the pressure pulse was recorded.
10. The computer calculated pressure data from the voltage data. Both voltage and pressure results
were stored on a 3.5-inch disk. The results could then be plotted or printed as desired.
11. The products of combustion were exhausted through the cryotrap by the house vacuum system.
12. The chamber was opened, and the inside was cleaned with paper towels moistened with acetone,
and the chamber was allowed to dry by evaporation of the acetone. The electrodes were sanded lightly with emery
paper to remove any contaminants.
13. At the end of each test day, the cryotrap and tubing were cleaned with acetone. When the agent
was changed, the exhaust and agent valves were also cleaned with acetone.
72
-------
APPENDIX E
PARTIAL PRESSURE LOADING PROCEDURES
The required mixture of fuel, agent, and air was achieved by measuring the partial pressures of the three
components. According to the idea] gas law, the volumetric fraction x 100 of each component in a mixture of
gases equals the ratio of the partial pressures of each component. Using Dalton's Law, the sum of the partial
pressures of all constituents equals the total pressure, according to Equation E-l:
patm =ph + pf + pa (E-i)
where
PATM = Standard atmospheric pressure (760 torr) (at sea level).
PH = Partial pressure of halocarbon = 760 torr times the desired volumetric fraction of agent.
Pp = Partial pressure of fuel = (ambient pressure - Ph) X volumetric fraction of fuel in the
stoichiometric fuel to air ratio.
PA = Partial pressure of air - 760 torr - Ph - PF-
Components may be added in any order, as long as the partial pressure of each agent is maintained. For
this testing, the agent is the first component added to the evacuated chamber. The valve in the agent supply line is
opened until the pressure transducer reads the desired partial pressure of agent. Likewise, the fuel valve is opened
until the transducer reads the desired pressure increase equal to the partial pressure of the fuel. Finally, the air
valve is opened and the mixture raised to 760 torr.
The stoichiometric fuel to air ratio for each of the three fuels was calculated by balancing the combustion
equation of each fuel with oxygen and dividing by the volumetric fraction of oxygen in air (0.2095) to get the
following fuel to air and fuel to total mixture ratios:
methane: 1:9.55 or 1 part fuel to 10.55 parts mixture
propane: 1:23.86 or 1 part fuel to 24.86 parts mixture
73
-------
Based on these ratios, Table E-l was generated to provide the partial pressures of agent and fiicl for agent
concentrations from 1 to 20 percent and each of the fuels, based on 760 torr total pressure in the chamber.
TABLE E-1. PARTIAL PRESSURES OF AGENT AND FUEL
Percent
Partial pressure
Partial pressure
Partial pressure
Partial pressure
agent
agent
fuel + air
propane
methane
(psi)
(psi)
(psi)
(psi)
0.01
0.147
14.553
0.585
1.381
0.02
0.294
14.406
0.579
1.367
0.03
0.441
14.259
0.574
1.353
0.04
0.588
14.112
0.568
1.339
0.05
0.735
13.965
0.562
1.325
0.06
0.882
13.818
0.556
1.311
0.07
1.029
13.671
0.550
1.297
0.08
1.176
13.524
0.544
1.283
0.09
1.323
13.377
0.538
1.269
0.10
1.470
13.230
0.532
1.255
0.11
1.617
13.083
0.526
1.241
0.12
1.764
12.936
0.520
1.227
0.13
1.911
12.789
0.514
1.213
0.14
2.058
12.642
0.509
1.199
0.15
2.205
12.495
0.503
1.185
0.16
2.352
12.348
0.497
1.172
0.17
2.499
12.201
0.491
1.158
0.18
2.646
12.054
0.485
1.144
0.19
2.793
11.907
0.479
1.130
0.20
2.940
11.760
0.473
1.116
74
-------
APPENDIX F
OVERPRESSURE VERSUS CHEMICAL CONCENTRATION CURVES
Appendix F presents plots of the explosion overpressure vs. chemical concentration for stoichiometric
fucl-to-air mixtures. A straight line fit using the least squares method has been drawn using the data points
generated during testing. For that line, a best-fit equation and resulting R-squared value have been determined.
For some chemicals, including Halon 1301, nitrogen, and carbon dioxide, where the transition from explosion to
non-explosion occurs within a narrow agent concentration range, the line connects the point of explosion and non-
explosion. A similar line was drawn for FC-5-1-14, which exhibited large data scatter, possibly due to a boiling
point close to the sphere's temperature. In these cases, the line is indicated by dashes, and no equation has been
derived. The stoichiometric inerting concentration (IC^) lias been determined as the chemical concentration at
which the straight line fit crosses the 1 psi line. To convert psi to kPa, multiply the psi value by 6.984.
75
-------
'55
Q.
09
>
o
90.00 T
80 00
70.00
60.00 ^
50.00
40.00
30.00
20.00
10.00 "
0.00
4.00
" Propane j
u Methane i
o
o
4.05 4.10 4.15 4.20 4.25
Chemical Concentration, Vol. %
4.30
4.35
Figure F-1. Overpressure vs. Chemical Concentration, Halon 1301.
I
ft)
S.
k_
0)
>
o
11.00
10.00
9.00
8.00
7.00
6.00
5.00
4.00
3.00
2.00
1.00
0.00
Overpressure = -66.78 (Cone.) ~ 941.98
R squared = 0 83
\ Overpressure « -2 66 (Cone.) * 41.95
\ R squared = 0.76
¦ Propane
3 Methane
9.00 10.00 11.00 12.00 13.00 14.00
Chemical Concentration, Vol %
\ o
\
X
15.00
16.00
Figure F-2. Overpressure vs. Chemical Concentration, CFC-12.
76
-------
\
¦ Propane
c Methane
\c
Overpressure = -18 95 (COnc ) ~ 205 82
Rsquared* 0.87
Q/erpressure» -8.00 (Cone.) ~ 79.49
R squared = 0 76
-+-
\
I
\
8.00 8.50 9.00 9.50 10.00 10.50 11.00
Chemical Concentration, Vol. %
11.50
12.00
Figure F-3. Overpressure vs. Chemical Concentration, CFC-114.
5.00 --
Overpressure = -7.19 (Cone.) + 106.42
R squared = 0 88
o \
¦ Propane
u Methane
e-
0)
\
\
i
¦
Overpressure = -0.43 (Cone.) + 9.33
Rsquared= 0 85
1
¦
, r
i ;
, ^ , ""'.v. ,
13.00 14.00 15.00
16.00 17.00 18.00 19.00
Chemical Concentration, Vol. %
20.00 21.00 22.00
Figure F-4. Overpressure vs. Chemical Concentration, HCFC-22.
77
-------
8.00
Overpressure = -2 62 (Cone.) + 33 86
R squared = 0.79
7.00
Overpressure = -3.32 (Cone)+ 3108
R squared = 0.74
6.00
5.00
Propane
4.00
Methane
® 3.00
2.00
1.00
0.00
8.00
7.00
9.00
10.00
11.00
12.00
13.00
Chemical Concentration, Vol %
Figure F-5. Overpressure vs. Chemical Concentration, HCFC-124.
¦ Propane
c Meihane
Overpressure = -16.72 (Cone ) + 97.26
R squared = 0.75
-4-
CXerpressure = -183 (Cone ) * 36.59
R squared = 0.98
\
\
\
\
\
I ( J )
9 10 11 12 13 14 15 16
Chemical Concentration, Vol. %
:\
H
17 18 19 20 21
Figure F-6. Overpressure vs. Chemical Concentration, HCFC-142b.
78
-------
8 00
Overpressure = -9.22 (Cone.) + 195.60
R squared - 0.88
7.00
Propane !
6.00
Methane
O- 5.00
4.00
3.00
Overpressure = -112 (Cone ) ~ 21.54
R squared = 0 52
2.00
1.00
0.00
16.00
17.00
18.00
19.00
20.00
21.00
22.00
Chemical Concentration, Vol. %
Figure F-7. Overpressure vs. Chemical Concentration, FC-14.
8.00
7.00
6.00 "f
•05
a- 5.00
S
3
$ 4.00
e-
® 3.00 +
o
2.00
Overpressure » -3 49 (Core.) * 36.39
R squared = 0.98
\
1.00
0.00
6.00
8.00
\
o \
\
¦ Propane
J Methane
t
Overpressure = -1 36 (Cone ) ~ 22 73
R squared = 0.91
\
f\
10.00 12.00 14.00
Chemical Concentration, Vol %
16.00
18.00
Figure F-8. Overpressure vs. Chemical Concentration, FC-116.
79
-------
10.00 -
Propane
Methane
8.00
7.00
» ~
8.
6.00 ~
0>
h»
5.00 ^
£
r
Overpressure = -2.17 (Cone)+ 25 24
R squared = 0.94
4.00 '' Overpressure = -4.41 (Cone)+\)3 47
R squared = 0 78 s-
3.00 ¦¦ °\
>
o
2.00
1.00
0.00
7.00
9.00
10.00
11.00
12.00
Chemical Concentration, Vol. %
Figure F-9. Overpressure vs. Chemical Concentration, FC-218.
5.00 T
4.50
Propane
Overpressure = -2.32 (Cone )~ 28.1
R squared = 0.98
4.00 "
Methane
3.50 --
Q.
,r 3.00
2.50 -¦ Overpressure = -2.18 (Cone.) ~ 20 42
R scared = 0 88
2.00 --
1.50
1.00
0.50
0.00
9.00
10.00
8.00
11.00
12.00
13.00
Chemical Concentration, Vol. %
Figure F-10. Overpressure vs. Chemical Concentration, FC-C318.
80
-------
4.50
Propane
(Xrerpressure * -2.86 (Core ) ~ 23.24
R squared = 0.93
4.00 --
Methane
3.50
•55 3.00
Overpressure = -2.16 (Cone.) + 21 82
R squared = 0.98
£ 2.50
£ 2 00
1.50 "
1.00
0.50
0.00
7.00
6.00
8.00
9.00
10.00
11.00
Chemical Concentration, Vol. %
Figure F-11. Overpressure vs. Chemical Concentration, FC-3-1-10.
2.
e-
%
o
10.00
9.00
8.00
7.00
6.00
5.00
4.00
3.00
2.00
1.00
0.00
\
¦ Propane
n Methane
CVerpressure = -3.OS (Core ) ~ 20 90
Rsquared » 0.51
\
N Overpressure" -4.24 (Cone)* 33 18
Rsquared = 0 87
\
o
¦ ~"
5.00
-I—
5.50
-h
•V
6.00 6.50
Chemical Concentration, Vol. %
V \
- ^ 1 n • • * » «
7.00 7.50 8.00
8.50
Figure F-12. Overpressure vs. Chemical Concentration, FC-4-1-12.
81
-------
4.00
3.50
3.00
& 2.50 t
S
i200
w
1.50
O
1.00 --
0.50
0.00
" Propane
D Methane
-+-
X
5.00 5.50 6.00 6.50 7.00 7.50
Chemical Concentration, Vol. %
8.00
8.50
Figure F-13. Overpressure vs. Chemical Concentration, FC-5-1-14.
6.00
5.00
"55 4.00
a.
4)
tw
e-
V
3.00
1.00
0.00
¦ Propane
u Methane
__ Oerpressure« -1.79 (Cone.) ~ 10.16
2-00 - Rsquared = 0 52
\
t).
V
4.00
4.50
Overpressure = -3 12 (Cone ) ~ 21.38
R squared = 0.84
\
\
1
$. 'V
5.00 5.50 6.00
Chemical Concentration, Vol. %
6.50
7.00
Figure F-14. Overpressure vs. Chemical Concentration, FC-6-1-16.
82
-------
14.00 t
12.00
10.00 -¦
"55
Q.
£ 8.00
.v
e-
0)
>
O
6.00 -¦
4.00
2.00
Overpressure = -4141 (Cone.) + 581.19
R squared = 0 94
¦ Propane
0 Methane
o.oo -*-¦
-f-
-f
\ O/erpressure = -15 3 (Cone ) + 30 76
\ R squared1 0 67
¦ \ .
H
12.00 13.00 14.00 15.00 16.00 17.00 18.00 19.00 20.00 21.00 22.00
Chemical Concentration, Vol. %
Figure F-15. Overpressure vs. Chemical Concentration, HFC-23.
Overpressure = -1.74
-------
O/erpcessure = -0 62 (Cone ) ~ 12 94
R squared = 0.49
1.80 T
Propane I
1.60
Methane !
1.40 --
Q.
1.00
* 0.80
Overpressure = -072 (Cone ) + 0 04
\ R squared = 0.64
o 0.60 --
0.40 -
0.20
0.00
14.00
8.00
9.00
10.00
11.00
12.00
13.00
15.00
16.00
Chemical Concentration, Vol %
Figure F-17. Overpressure vs. Chemical Concentration, HFC-125.
1.80
1.60
1.40
w 1 20
a.
I 1.00
8
I 0.80
£ 0.60
0.40
0.20
0.00
12.
00
Overpressure = -0.43 (Cone ) + 7.75
R squared = 0.92
\
¦ Propane
13.00 14.00 15.00
Chemical Concentration, Vol. %
16.00
17.00
Figure F-18. Overpressure vs. Chemical Concentration HFC-134.
84
-------
4.00
3.50
3.00
2.50
2.00
« 1.50 4
S.
1.00 f
Propane
° Methane
I Overpressure - -8.60 (Cone ) + 68 30
I R squared = 0 96
Overpressure = -0.65 (Cone.) ~ 10.18
R squared = 0 76
7.00 8.00 9.00 10.00 11.00 12.00 13.00 14.00 15.00 16.00
Chemical Concentration, Vol. %
Figure F-19. Overpressure vs. Chemical Concentration, HFC-134a.
25 oo
CVerpressure « -22.68 (Cone ) + 153.97
R squared = 0 57
20.00 T
35
Cl
15.00
I 1
0>
>
O
0.00
5.00
\
0.00 -f— I
4.00 5.00
;v+
-) m tn
6.00 7.00
* Propane
Methane
Overpressure- -2.57 (Cone J + 27.08
R squared = 0 69
-t-
8.00 9.00 10.00 11.00
Chemical Concentration, Vol. %
12.00
Figure F-20. Overpressure vs. Chemical Concentration, HFC-152a.
85
-------
2.50
2 00
OwprasMre= ¦% 67 (Cone ) ~ 22.64
R*qu**d ¦ 0 94
2
3
I
I
0.50
000
12.00 12.20 12.40 12 60 12.80 13.00 13 20 13 40 13 60
Chemical Concentration, Vol. %
Figure F-21. Overpressure vs. Chemical Concentration, HFC 227ca.
8.00
7.00
Propane
Methane
6.00
Q- 5.00
Overpressure = -1 17 (Core ) ~ 14 94
Rsquared" 0 92
4.00
Overpressure = -1.45 (Cone.) + 12.81
R squared = 0.89
$ 3 00
2 00
>K}_
0.00
6.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00
Chemical Concentration, Vol. %
Figure F-22. Overpressure vs. Chemical Concentration, HFC-227ea.
86
-------
2.50 T
\
2.00 -
\
\
\
X
\
\
Overpressure = -1.23 (Cone) + 15 35
Rsquared = 0 77
-+-
\
-+-
_Ak_
Propane
10.00 10.50 11.00 11.50 12.00
Chemical Concentration, Vol. %
12.50
13.00
Figure F-23. Overpressure vs. Chemical Concentration, HFC-236ea.
2.50
2.00
CL
® 150
5
a>
>
O
1.00
0.50
0.00
¦ Propane
\
\
Overpressure = -0.766 (Cone.) + 9.03
Rsquared * 0 91
\
I H
¦$-
-t-
—I— »
-t— 1
8.00 8.50 9.00 9.50 10.00 10.50 11.00 11.50 12.00 12.50 13.00
Chemical Concentration, Vol. %
Figure F-24. Overpressure vs. Chemical Concentration, HFC-236fa.
87
-------
3.00
Propane
2.50 -
2.00
1
Cverpressure = -1.75 (Conc.J ~ 19.68
R squared = 0.74
k.
e-
S
o
0.50
0.00
9.50
10.00
10.50
11.00
Chemical Concentration, Vol %
Figure F-25. Overpressure vs. Chemical Concentration, HFC-245cb.
Propane
Overpressure = -2.38 (Cone)* 24.03
R squared « 0.93
\
-+
\
4' \
t
-i
.50 8.00 8.50 9.00 9.50 10.00 10.50 11.00 11.50 12.00 12.50
Chemical Concentration, Vol %
Figure F-26. Overpressure vs. Chemical Concentration, HFC-254cb.
88
-------
¦ Propane
Overpressure = -1.30 (Cone.) ~ 22.23
R squared = 0.91
\
o.oo : 1 —i— t — 1- ¦
1450 15.00 15.50 16.00 16.50
Chemical Concentration, Vol %
. • _
17.00
17.50
Figure F-27. Overpressure vs. Chemical Concentration, HFC-32:HFC125 Azeotrope.
<»
9)
k-
I
5
60.00 j
55.00 ^
50.00 -
45.00
40.00
35.00
30.00
25.00
20 00
15.00
10.00
5.00
0.00
4.00
Overpressure = -362.97 (Oonc ) ~ 2041 87
R squared = 0.77
CVerpfessufe » -143.70 (Cone.) ~ 1175.08
R squared = 0 50
Propane
D Methane
T •" i"rn r-
4.50 5.00 5.50 6.00
T
6.50 7.00
Chemical Concentration, Vol. %
7.50
8 00
8.50
Figure F-28. Overpressure vs. Chemical Concentration, HBFC-22B1.
89
-------
50
45
Propane
40
\ Overpressure = -45 06 (Cone.) ~ 264.32
\ Rsquared = 0.71
Methane
35
30 -
20
Overpressure = -204.75 (Cone ) * 740.56
R squared = 0 96
a o-
- n
0
3
3.5
4
4.5
5
5 5
6
6.5
Chemical Concentration, Vol. %
Figure F-29. Overpressure vs. Chemical Concentration, HBFC-124B1.
70
60 t
50
40 f
30
20
10
¦ Propane
D Methane
Overpressure = -193.19 (Cone) + 1005
Rsquared = 0 21
Overpressure = -403.66 (Cone.) ~ 1257
Rsquared = 0.92
ULH—U [ i
H— 1 -I- h
2 2.2 2.4 2.6 2.6 3 3.2 3.4 36 3.8 4 42 44 4.6 48 5 5.2 5.4 5.6
Chemical Concentration, Vol %
Figure F-30. Overpressure vs. Chemical Concentration, FIC-1311.
90
-------
70.00
65.00
60.00
55.00
50.00
I 45.00
£ 40.00
I 35.00
Propane
Overpressure = -81.93 (Cone.) ~ 910 56
R squared « 0 71
8.00
-C
.00
—T ;-—
10.00 11.00 12.00 13.00
Chemical Concentration, Vol %
14.00
15.00
16.00
Figure F-31. Overpressure vs. Chemical Concentration, FIC-11511.
1.40 --
1.20 -
1.00 -
'55
Q.
£ 0.80
|
o. 0.60 -
<5
a
0.40 +
Propane
O/erpressure = -0.62 (Cone.) * 4 25
R squared * 0.94
0.20
0.00
4.00
t—
4.50
_..v
t-
5.00 5.50 6.00
Chemical Concentration, Vol. %
6.50
7.00
Figure F-32. Overpressure vs. Chemical Concentration, ^Bromo-S.SA^-tetrafluoro-l-butene.
91
-------
CO
Cl 50 +
40 -¦
30 -
20
10
0
16
17
18
) -
19
- I
20
t—
22
¦ t
23
¦ Propane
21 22 23 24 25
Chemical Concentration, Vol. %
i—
26
- I
27
t
28
29
30
Figure F-33. Overpressure vs. Chemical Concentration, CC>2-
50 T
45
40
35
30
25
I" 20 t
O
X
* Propane
1 * 1 *—
35 36 37 38 39 40 41 42
Chemical Concentration, Vol. %
43
44
45
Figure F-34. Overpressure vs. Chemical Concentration, N2.
92
-------
APPENDIX G
OVERPRESSURE VERSUS CHEMICAL CONCENTRATION CURVES, BLENDS
93
-------
5
C4
C4 - 4.5% 1301
4
C4-3.2% 1301
-© C4-13.2% 1301
C4 - 5.3% 124B1
3
C4 - 9.2% 124B1
C4-14.6% 124B1
--<~ C4-15% 152a
2
1
0
4.0
5.0
5.5
6.0
7.0
8.0
9.0
10.0
11.0
3.0
4.5
Total agent concentration, percent
Figure G-1. Overpressure vs. FC-3-1-10 Blends Inerting Concentration, Methane.
-------
~ OA
(5.1,41.6) (6.1,36.4) (7.0,12.8)
134a
134a-15% 124B1
-A 134a-15%- 152a
134a-15%142b
134a-10% 1311
134a -5% 1311 -5% 124B1
5
6
7
10
4
8
9
Total agent concentration, percent
Figure G-2. Overpressure vs. HFC-134a Blends Inerting Concentration, Methane.
-------
A
(10,8.1)
5
134a
134a-15% 1311
4
134a- 15%124B1
134a - 7.5% 1311 - 7.5% 124B1
-•A 134a-15% 152a
3
134a- 15%142b
2
1
0
8
9
10
11
13
12
14
15
16
Total agent concentration, percent
Figure G-3. Overpressure vs. HFC-134a Blends Inerting Concentration, Propane.
-------
20
H 32
B 32 - 15% 1311
15
10
5
0
Total agent concentration, percent
Figure G-4. Overpressure vs. HFC-32 Blends Inerting Concentration, Propane.
------- |