600281216
May 1931
MODIFICATION OF SPILL FACTORS AFFECTING AIR POLLUTION
Vol. II - The Control of the Vapor Hazard From
Spills of Liquid Rocket Fuels
by
J . S . Greer
S.S. Gross
R . H . H i 11 z
M.J. McGoff
MSA Research Corporation
Division of Mine Safety Appliances Company
Evans City, Pennsylvania 16033
Contract No. 68-03-264J
Project Officer
John E. Brugger
Oil and Hazardous Materials Spills Branch
lunicipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental
Research Laboratory, U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the U.S.
Environmental Protection Agency, nor does mention of trade names
or commercial products constitute endorsement or recommendation
for use.
-------
FOREWORD
The U.S. Environmental Protection Agency was created be-
cause of increasing public and government concern about the
dangers of pollution to the health and welfare of the American
people. Noxious air, foul water, and spoiled land are tragic
testimonies to the deterioration of our natural environment.
The complexity of that environment and the interplay of its
components require a concentrated and integrated attack on the
problem.
Research and development is that necessary first step in
problem solution; it involves defining the problem, measuring
its impact, and searching for solutions. The Municipal En-
vironmental Research Laboratory develops new and improved tech-
nology and systems to prevent, treat, and manage wastewater and
solid and hazardous waste pollutant discharges from municipal
and community sources, to preserve and treat public drinking
water supplies, and to minimize the adverse economic, social,
health, and aesthetic effects of pollution. This publication
is one of the products of that research and provides a most
vital communications link between the researcher and the user
commun i ty.
This report addresses a preliminary evaluation of cooling
and foam blankets as mechanisms of reducing the release of
hazardous vapors from hypergolic rocket fuels. The results show
that foam has potential for controlling the vapor release from
hydrazine and nitrogen tetroxide based propellants.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
i i
-------
ABSTRACT
The hypergolic rocket fuels, hydrazine and nitrogen tetrox-
ide, are volatile hazardous materials of special interest to the
Air Force. Through monitoring of ongoing Environmental Pro-
tection Agency programs, the Air Force has maintained cognizance
of the developing state-of-the-art in spill control. This Air
Force supplement to the basic EPA program was a preliminary eval-
uation of the potential of cooling and foam covers to mitigate
the vapor hazard from hydrazine and nitrogen tetroxide.
Coolants exhibited some control over vapor release from the
hypergolic fuels. Liquid nitrogen was the most effective ma-
terial. Logistics were deemed a major disadvantage for the antic-
ipated spill scenarios.
Foams using commercial agents were beneficial with hydrazine
but were not effective against nitrogen tetroxide. Modified foam
systems incorporating acrylic resins were more effective. They
were able to maintain hydrazine concentrations at or below 0.5
ppm. Some control was also exhibited with nitrogen tetroxide
but there was intermittant vapor release through the foam.
Based upon the work of this program the acrylic foams offer
a promising approach to the control of the vapor hazard from
hydrazine and nitrogen tetroxide.
This report was submitted in partial fulfillment of Con-
tract No. 68-03-2648 by MSA Research Corporation under the spon-
sorship of the U.S. Environmental Protection Agency. This report
covers the period 22 January 1980 to 4 December 1980, and work
was completed as of 15 May 1981.
-------
CONTENTS
Foreword ill
Abstract iv
Tables v
Abbreviations and Symbols vi
Acknowledgment vii
1. Introduction 1
2. Summary and Conclusions 2
3. Recommendations 4
4. Program Plan 5
Phase 1 - Coolant Evaluation 5
Phase2-FoamStudies 5
Phase 3 - Demonstration 6
Final Report 6
5. CryogenEvaluation 7
Results with Nitrogen Tetroxide 7
Results with Hydrazine 8
Data Analysis 9
6. Foam Evaluation 10
Foam Agent Select ion 11
Testing of Commercial Agents 12
Testing Special Agents 15
Foam Modifications 17
Foam-Cryogen Combinations 17
7. Proposed Program Continuation 20
Hydrazine Vapor Control 20
Nitrogen Tetroxide Vapor Control 20
TABLES
Number Page
1 The Results of Screening Tests on Low Ex-
pansion Foams with Hydrazine 13
2 Screening of Potential Foam Modifiers 18
-------
AFFF
°C
cm
°F
g
gal
i n .
mi n
ml
oz
ppm
TLV
ABBREVIATIONS AND SYMBOLS
aqueous film forming foam
degrees Celsius
centimeter
degrees Fahrenheit
gram
gallon
inch
minute
m i 1 1 i 1 i t e r
ounce
parts per million
threshold limiting value
VI
-------
ACKNOWLEDGMENT
This work was performed under subcontract to the Environ-
mental Monitoring & Services Center of Rockwell International,
Newbury Park, California. The authors wish to thank George R.
Schneider, Project Officer, Environmental Monitoring & Services
Center of Rockwell International, and John E. Brugger, Project
Officer, U.S. Environmental Protection Agency, for their direc
tion and support.
VI 1
-------
SECTION 1
INTRODUCTION
Liquid rocket propellents present one class of volatile
hazardous chemicals. The toxic hazard has been well recognized,
and the Air Force in its facilities provides for the contingency
of a spill of these materials. As the state-of-the-art in spill
control has advanced, the Air Force has monitored the progress
with interest for its own applications. The success of aqueous
foams in controlling vapor hazards (which has been demonstrated
in earlier EPA-sponsored programs) and the potential for cryo-
genic cooling (which was being demonstrated in this program) was
justification for the Air Force to undertake an investigation of
these techniques as mechanisms to control the vapor hazard from
spilled rocket fuels.
The existing program was supplemented by Air Force funding
for the specific purpose of investigating cooling and aqueous
foam blankets to mitigate the vapor hazard of hydrazine and
nitrogen tetroxide. This study was primarily a survey and was
not intended to develop a total system. State-of-the-art tech-
nology was used to the fullest extent.
-------
SECTION 2
SUMMARY AND CONCLUSIONS
The investigation assessed the ability of two techniques --
cooling and foam covers -- to mitigate the vapor hazard of hydra-
zine and nitrogen tetroxide. Each technique was pursued in-
dependently for each of the propellant materials. The program
used a series of laboratory tests to evaluate the various methods
and materials for vapor suppression. Vapor concentrations above
the two fuel materials were measured using detector tubes. Tests
in which the concentrations exceeded the maximum limit of de-
tector tubes, about 35 ppm, were considered failures.
The coolant portion of the study investigated wet ice, dry
ice, liquid CO?) and liquid nitrogen. For nitrogen tetroxide,
liquid nitrogen was able to reduce the vapor concentration below
the detector tube limitation. Wet ice had an immediate but small
effect that persisted only as long as the ice was solid. As
liquid formed, it reacted to form nitric acid. The heat of for-
mation more than offset the cooling effects. To maintain the low
vapor level, fairly large quantities of nitrogen were required,
with fairly frequent make up indicated.
The adding of coolants to hydrazine reduced vapor levels be-
low the limits of the detector tubes, but they were not able to
approach the TLV. Wet ice produced the slowest reduction of
vapor concentration with hydrazine. Because of the accompanying
dilution, as the ice melted, the lowest vapor concentration was
ultimately achieved.
The evaluation of foam systems initially considered com-
mercially available foam agents. Representative materials of
protein, fluoroportein, AFFF, and surfactant foams were selected
based upon characteristics identified in a previous EPA program.
A second set of materials, which covered the so-called "alcohol"
or "polar solvent" foam agents, was added to these.
None of the materials selected were able to exercise any
control over the vapor release rate of nitrogen tetroxide. In
contrast, almost all were able to reduce substantially the vapor
concentration above hydrazine. These levels varied from 2 to 20
ppm depending on the foam. The length of control also varied as
a function of foam type.
-------
A third type of foam system was also tested. These agents
were based upon acrylic modified foams previously developed by
MSA which gel when used on ammonia or amine materials. The gelled
foam forms a stable barrier with a low permeation of hydrazine
vapor. There is some drainage which mixes with the hydrazine and
gels the top liquid layer. The two layers were able to reduce the
vapor concentration into the range of 0.5 ppm.
These foam formulations were also beneficial in reducing
vapor release from nitrogen tetroxide. They were not able to
give continuous control, however. The vapor build up beneath
the foam layer resulted in intermittent vapor pulses (breathing).
Although these foams were better than any other foam evaluated,
they were less than desired. They did provide a basis for fur-
ther development.
The acrylic modified .foams tended to form a skin in con-
tact with nitrogen tetroxide. A series of polymeric materials
were tested alone and in combination with the acrylic modifier
in an effort to improve the control over nitrogen tetroxide.
Within the time limit of this program no major improvements over
the original MSA formulations were found.
The use of coolants as a method of controlling the vapor
hazard from spilled liquid hydrazine or nitrogen tetroxide is
rejected. The tests showed some reduction in vapor release, but
when this was equated with the volume of coolant material required
and the logistics involved for projected scenarios, cooling was
not deemed a viable approach.
Aqueous foam systems offered significant benefits for hydra-
zine. Several commercially available agents were shown to be
able to reduce vapor levels to the 2 to 5 ppm range. None of the
agents were effective against nitrogen tetroxide.
Acrylic modified aqueous foams developed previously for
ammonia and like materials were shown to be effective for hydra-
zine. Foam layers reduced vapor levels to the 0.5 ppm range.
These agents also had some positive effects on nitrogen tetroxide.
At the conclusion of this program, these foams offered the best
mechanism to control the vapor hazard from the two types of rocket
propellant materials.
-------
SECTION 3
RECOMMENDATIONS
The prime recommendation to be made is to continue to in-
vestigate improvement in the acrylic modified foams. Although a
number of modifications were tried without measurable improve-
ment, these reflected materials which were available within the
time constraints of this program.
he acrylic materials which have been used are stable pri-
marily o n the basic side. Certain other materials that exhibit
better stability have been recommended by the polymer manufac-
turers. In addition, there is the potential for providing foams
that gel on the acid side. These foams are based upon silicic
acid. There has been some prior work in this area but it was
for a very different application. Thus the evaluation of silicic
acid gels will have to start with basics.
Certain of the potential modifications which were attempted
in the test program were incompatible with the surfactant ma-
terial employed in the acrylic modified formulation. This pre-
vented their full evaluation.
Beyond the evaluation of additional materials for a nitrogen
tetroxide system, there is a need to conduct some larger scale
tests to fully verify the suitability of whatever foam system is
selected as the best available technology. This should address
the types of spill scenarios which might be encountered. Thus,
at the conclusion of the test program, if results are positive,
spill control systems can be designed for each scenario.
The test program should also address the clean up procedures
which can or should be used after the vapor control system has
been utilized.
-------
SECTION 4
PROGRAM PLAN
The program had as its objective the evaluation of two
techniques, foam covers and surface cooling, as mechanisms to
control the vapor hazard from spilled liquid rocket propel Iants.
The planned program had three scheduled phases. Phase 1 was con-
cerned with the evaluation of coolant technology. Phase 2 eval-
uated foam technology. With positive results from the first two
phases a demonstration of capability was conducted as Phase 3.
The total program covered a twelve week period. The first
week was used for planning and organizing the program and allowed
for materials acquisition. The final two weeks were used to co-
ordinate all information and prepare a final report. The re-
mainder of the schedule provided for a demonstration.
PHASE 1 - COOLANT EVALUATION
This phase was scheduled for two weeks. It utilized, to
the extent possible, the results of the main program on coolants.
This included definition of candidates, form of application, the
ratio of coolant to spilled material and the techniques for
measuring vapor concentrations.
The total two weeks were spent on laboratory studies. One
additional week of the program was set aside to evaluate test
data and formulate conclusions and decisions.
PHASE 2 - FOAM STUDIES
This phase covered four weeks. Two weeks were devoted to
evaluations of commercial agents with both hydrazine and nitro-
gen tetroxide. One week was given to the evaluation of the
special agents proposed for test. The final week was for testing
of modifications which were indicated by the test results.
The laboratory study drew upon prior work to establish can-
didate foam agents both commercial and special. Test procedures
paralleled those successfully employed in other similar studies.
The data derived in this phase was analyzed and conclusions
formulated in the same week as for the coolant data analysis.
-------
PHASE 3 - DEMONSTRATION
The fate of this phase depended upon the results of the
first two phases. It was scheduled to cover two weeks; one week
for preparation and one week for conducting the field work.
The details of the demonstration would be evolved during
the course of the program through discussions with the Air Force.
This would be a small scale field test encompassing a spill sur-
face area no larger than 25 square feet. The actual test parame-
ters depended on the mechanism selected for vapor control, foam
or coolant. In either case the demonstration would parallel
those used either in the main study on coolants or prior EPA work
with foams. The uncertainties concerned restrictions the Air
Force might want to impose to simulate real time spill scenarios.
FINAL REPORT
A final report covering all aspects of the program was sub-
mitted in draft form during the 12th week of the program.
-------
SECTION 5
CRYOGEN EVALUATION
The use of refrigerants to cool a liquid spill and thereby
reduce the equilibrium vapor pressure has been the subject of the
first and main part of this program. Although the test fluids
had changed for the supplemental part of the program, the basis
for selecting acceptable cooling candidates was still the same.
It was possible to argue for limiting the coolant to dry ice but
there were circumstances which compromised that position. First,
both propellant materials were soluble in or reactive with water,
and second, TLVs were so low that only liquid nitrogen might be
effective. It was decided initially to evaluate all four forms,
but since liquid C02 is essentially applied as solid snow, dry
ice was the only form of C02 evaluated.
The tests conducted were similar in all cases. The volume
of propellant was 50 ml (0.01 gal) for N£04 and 100 ml (0.03 gal)
for hydrazine. The volume of coolant was 200 grams (7.05 oz)
for both wet and dry ice and 100 ml (0.03 gal) for liquid nitro-
gen.
RESULTS WITH NITROGEN TETROXIDE
Wet Ice
Approximately 200 g (7.05 oz) of ice was required to pro-
vide an excess when added to 50 ml (0.01 gal) of liquid dinitro-
gen tetroxide. An equilibrium temperature of -10°C (14°F) was
measured after one minute. The vapor concentration decreased to
between 30 and 50 ppm over the liquid, after this temperature
was attained.
Although some 60 minutes was required to melt all of the
ice, the control of the vaporization rate was lost as soon as
measurable melting of the ice had occurred. This took approxi-
mately 15 minutes. After this 15 minutes the cooling by the ice
was in competition with the heat being generated by the water-
N204 interaction.
Dry Ice
The addition of crushed dry ice caused agitation of the
N204- The dry ice particles sank into the liquid, where they
7
-------
sublimed. A dense vapor resulte. that was composed of COg,
and condensed moisture. The concentration of ^04 never dropped
below 50 ppm even though the addition of 200 grams (7.05 oz) of
dry ice was sufficient to create a continuous layer of dry ice
over the Ng04 surface.
Liquid Nitrogen
The addition of liquid nitrogen to nitrogen tetroxide
created a similar response to solid C02- A dense vapor cloud
was released which contained red streaks indicating the carry off
of N204- The addition of 100 ml (0.03 gal) of liquid nitrogen
resulted in the formation of floating islands of LN2- Solid
particles which sank into the nitrogen tetroxide were observed
to form at the liquid-liquid interface.
As long as excess IN? persisted on the surface, N204 vapor
levels were suppressed into the 10-15 ppm range. As soon as the
liquid nitrogen boiled off, 5-10 minutes, the N 2 0 4 vapor concen-
tration increased rapidly beyond the detection limits of detector
tubes.
RESULTS WITH HYDRAZINE
Wet Ice
The addition of 200 grams (7.05 oz) of ice to hydrazine had
no immediate effect. The vapor concentration did decrease slowly
reaching 50 ppm, the limit of the detector tube, in about 2
minutes, and decreased to 20 ppm in ten minutes. At that time
the temperature of the hydrazine-ice mixture had decreased to
-3°C (27°F). These conditions persisted for about 20 minutes,
which was the time necessary for melting of the ice.
Dry Ice
The application of dry ice particles to hydrazine created
the same effect as for N204. There was a heavy cloud which
sampling showed entrained considerable quantities of hydrazine
vapor. Hydrazine levels were always above the maximum limit of
the detector tubes.
Liquid Nitrogen
The liquid nitrogen-hydrazine tests also paralleled those
with N 2 0 4. The development of visible LN2 on the spill surface
caused a solid material to form. This material had a tendency to
sink into the liquid hydrazine rather than form a floating layer.
Vapor concentrations in the range of 5 to 10 ppm were
achieved as long as liquid nitrogen was maintained on the surface.
When it disappeared, the vapor concentration rapidly rose to
-------
levels beyond the limits of the cetector tubes.
DATA ANALYSIS
The data showed that some reduction in the vapor hazard
could be achieved with coolants. Liquid nitrogen gave the best
results. In evaluating the data, consideration was given to
tentative scenarios. In all cases, fast if not immediate re-
sponse was necessary. The logistics of moving coolant to a spill
site made that approach unattractive. On-site storage or in-
place material formation was possible but both equipment costs
and maintenance requirements would be very high.
It was decided on the basis of this analysis to hold up
further work on this approach, pending results from the foam
portion of the program. As will be seen, the foam approach was
sufficiently successful that no further work was done or recom-
mended on the coolant system.
-------
SECTION 6
FOAM EVALUATION
Aqueous foams have seen wide use in the control of the vapor
hazard from spilled hydrocarbon liquids. In the past few years
there has been a number of investigations of the ability of foam
to restrict the vapor hazard from a variety of spilled volatile
chemicals not necessarily falling in the hydrocarbon category.
It is clear from this work that foams are effective barriers
to vapor release for those chemicals which are not soluble in
water. Water solubility or reactivity can affect the normal
types of foam in use by the fire services, with changes in pH
providing the most significant effect.
Aqueous foams are more tolerant of changes to the basic side
than to the acid side. Limited work has shown that the aliphatic
amines such as diethylamine can be benefited from a foam blanket.
The effect is to reduce the vapor concentration above the spill,
for some duration of time depending on the type of foam, the
depth of the blanket and the expansion of the foam. The vapor
does ultimately penetrate the blanket and the concentration above
the foam will tend to reestablish an equilibrium condition.
Hydrazine is expected to behave in a similar manner. The
rate of permeation through foam blankets is unknown, however,
and whether standard foam systems can provide the necessary de-
gree of control must be determined.
Efforts to use foam to control ammonia spills resulted in
the development of foam systems which would gel on interaction
with the ammonia. A similar action was observed with other chemi-
cals which shifted the pH in the basic direction. Once gelled,
the foam provides a greater restriction to the permeation of the
gas phase. Thus these foams provided enhanced vapor mitigation.
The use of foam on spills of materials which provide an acid
response has been limited to date to those materials which have
a severe toxicity problem. In none of these cases is the foam
able to form a blanket due to degradation by the acid action.
The main purpose for foam use is a mechanism to apply water gently
to reactive materials to convert the vapor to a less hazardous
form: S03 to H2S04, SiCl4 to HC1, etc., and release the material
as an aerosol rather than a vapor since aerosols tend to settle
10
-------
out rather than disperse into t h . air.
One would expect N204 to behave in a similar manner. Some
possibilities exist beyond normal foams. Foam solutions con-
taining silicates can be gelled by shiftingpH in the acid di-
rection forming silicic acid. These are usually formed by slight
acidulation and whether they would persist in contact with N204
is not known.
A second possibility is controlled neutralization. With
chlorine bearing materials such as SiCl4 with application of a
foam containing NH3(NH40H soln), the chlorine is reacted to
NH4C1. This reaction occurs rapidly at the foam-chemical inter-
face and results in a floating layer of NH4C1 . This forms a
barrier to the transfer of water in and vapor out. The reaction
between the water and the chemical slows down and vapor release
is sharply reduced.
There are some potential reactions with N 2 0 4 which might
behave in this fashion using certain hydroxides or alkali car-
bonates. Not all neutralization reactions behave the same way,
however. NaOH solution foams are not effective against chloride
materials; in fact, the reactivity becomes quite violent.
There are currently five basic foam agent types commercially
available. These are proteinaceous materials derived from natu-
ral protein, aqueous film forming foam (AFFF) which employ a
fluorocarbon surfactant, fluoroprote ins which are combinations of
AFFF and protein, synthetic materials which are hydrocarbon sur-
factant base, and "alcohol" or "polar solvent" agents which tend
to be proprietary.
Within foam technology there are two expansion ranges in
use. There are theoretical design limits but in practice the
limits are 5 to 20:1 for low expansion and 250 to 750:1 for high
expansion foams. Only synthetic foaming materials encompass
both expansion ranges. The others are restricted to the low
range. AFFF can make high expansion foam but its breakdown rate
is extremely fast.
The evaluation program included representative foam agents
from each category. Fluorocarbon materials perform quite poorly
with water reactive materials. They are in wide use by the Air
Force for fire protection. For that reason they were included.
FOAM AGENT SELECTION
Commerc i a! Agents
In EPA-sponsored work it has been clearly shown that regard-
less of the type of foam, expansion or the volatile material in-
volved, the better the water retention capability of the foam the
1 1
-------
better is its performance. This tact was used to minimize the spe-
cific foam agents to be evaluated in the prior study. Further con-
trol of the number of test materials can be exercised by selecting
a constant foam agent addition of 6 percent for low expansion. The
EPA wor'k also shows that there is little difference between 3 and
6 percent forms of the same agent from the same manufacturer.
The agents selected during prior EPA programs were selected
for this program also. They are listed below. They represent
the best drainage characteristics of their class.
Protein - National AeroFoam 6 % Regular
Fluoroprotein - National XL 6 %
Fluorocarbon - National Aero-0-Water 6%
Surfactant - MSA Ultrafoam V
The choice of the MSA agent for the surfactant candidate was made
on the basis that it has the best drainage of all surfactant ma-
terials in both the low and high expansion mode and is the only
surfactant agent which meets the specifications of Underwriters
Laboratories for both low and high expansion foam.
In addition to these four agents, three additional agents
were included. These are the Universal agent of National Foam,
which reportedly is effective against ammonia; a fluorocarbon
base polar solvent foam, 3M ACT; and a protein base alcohol foam,
Rockwood All Purpose.
Special Agents
Three categories of special agents were included in the
selected lists. Two gelling systems previously investigated by
MSA for amine and ammonia were added; one using acrylic modifi-
cation and the second a carboxy vinyl additive. The third cate-
gory was the silicic acid system. This system is not well defined
and its inclusion was as a general category.
It should be noted that as the program progressed, further
modifications were made in these special agents to accommodate
observed deficiencies.
TESTING OF COMMERCIAL AGENTS
Each of the selected commercial agents was carried through
a screening to assess compatibility with both hydrazine and nit-
rogen tetroxide. All of the agents were tested in the low ex-
pansion mode. The AFFF, synthetic, fluoroprotein and the fluoro-
carbon polar solvent foam were also tested in the high expansion
mode .
For those foams and expansion which passed the screening
12
-------
test, additional tests were run ^ o establish a basis for assess-
ing the efficiency of vapor control. This series of tests used
detector tubes to determine breakthrough times. With vapor perm-
eation a sharp discontinuity is usually observed in vapor concen-
trations above the foam. This occurs when either the foam collap-
ses completely or it becomes saturated with the vapor phase. The
time to reach this point is a relative measure of the control
capability of the foam blanket.
Screening Tests for Compatibility
The compatibility for each agent was determined by measuring
the time necessary for the foam generated to collapse to one-half
its original height. In the case of low expansion tests about
7.6 cm (3 in.) of 8 to 1 expansion foam was placed over 50 ml
(0.01 gal) of the propellant material in a 500 ml (0.15 gal)
beaker.
The tests with nitrogen tetroxide were easy to evaluate; all
foams collapsed rapidly with only minimal control of the vapori-
zation rate. Of the seven agents tested, the best results were
obtained using MSA Type V, AFFF and 3M ACT polar solvent agent.
These survived because there was some foam regeneration by the
vapors being released. This action appeared to be primarily en-
trainment. When the foam did collapse all of the entrained vapor
was released .
The tests with hydrazine had more positive results. The
times for reduction to half the foam volume using low expansion
foam are given in Table 1. These data do not indicate any cap-
ability in vapor suppression, only foam compatibility.
TABLE 1. THE RESULTS OF SCREENING TESTS OF
LOW EXPANSION FOAMS WITH HYDRAZINE
Time Required to Obtain 1/2
Agent Original Foam Volume (ml n )
National "Aero-Foam Regular" >60
National "XL-3%" >60
Rockwood "All Purpose" ^ 1
NationalUniversa! >60
3M "Light Water" 0,4 7
3M "Alcohol Solvent Concen-
trate" ^60
MSA Ultrafoam V >60
13
-------
Similar tests were done for the foaming agents capable of
forming stable foams at higher expansion ratios. These foam
samples were prepared by blowing the foaming agent solutions to
form foams having expansion ratios between 100 and 150 to 1. The
results were poor in all cases indicating unsuitability for the
high expansion mode.
Vapor Control by Selected Commercial Agents
The five commercial agents which appeared to be compatible
with hydrazine in the low expansion form were carried forward to
assess their ability to block vapor release from hydrazine. It
was decided there was no justification to look further at high
expansion nor to do an evaluation of any commercial agent with
nitrogen tetroxide.
The same test set up as was used for the compatibility tests
was employed. Detector tubes were used to measure vapor concen-
trations above the level of the foam blanket. The best results
were obtained with National Universal and MSA Type V.
Regular Protein Foam (National) --
Protein foam was compatible with hydrazine. Foam bubbles
at the interface expanded as vapor was absorbed, forming a layer,
which was gradually assimilated into the foam mass. The foam
drained normally but the collapse rate was reduced because of re-
generation by the vapors. A vapor concentration of less than 10
ppm was maintained for approximately 60 minutes by a 7 cm (^3 in.'
layer of foam. Beyond this time the vapor concentration in-
creased at a fairly rapid rate.
Fluoroprotein Foam (National) --
This foam behaved in much the same way as regular protein.
Interfacial bubbles formed as vapor was assimilated and some
volume expansion occurred. The level of hydrazine vapors above
the foam was low, about 1 ppm, but the breakthrough time was
short covering only 30 minutes. A high drainage rate caused the
foam to become friable and lose its barrier effect.
Polar Solvent Foam (3M) --
The fluorocarbon based polar solvent foam was compatible
with hydrazine, whereas the protein base material was not. The
fluorocarbon foam behaved much like protein and f1uoroprotein but
there was a tendency for large bubbles to form and percolate i n -
termittantly through the foam layer. An average vapor concen-
tration of 20 ppm was measured for about 20 minutes after appli-
cation. At that point the vapor level began a measurable in-
crease.
14
-------
Universal Foam (National) --
This agent, suitable for both normal and polar materials,
followed the same pattern as all previous agents. The average
vapor level above the foam blanket was 5 ppm. Breakthrough time
wasabout60minutes.
Synthetic (MSA) --
Over the first 15 minutes this agent gave the best per-
formance of all of those tested. The low expansion cover appear-
ed to achieve complete blockage of hydrazine vapor for that period
of time. After 30 minutes the vapor level had increased to 5 ppm
and then maintained that level for an additional 30 minutes.
Data Clarification --
These data should be reviewed with care since there is no
guarantee that the performance of these representative materials
will be characteristic of all agents in that class. In fact, the
opposite may be true. The selection process attempted to identify
the best in each category. National Universal, 3M ACT and f^SA
Type V, although commercially available, are proprietary formu-
lations not matched by other manufacturers. The data presented
for these three foam agents should be treated as being unique.
The data for protein, fluoroprotein and the two materials
which were eliminated during screening, AFFF and alcohol foams,
can probably be taken as representative. Differences between
foam agents from different manufacturers is not great in terms
of drainage and chemical compatibility.
TESTING SPECIAL AGENTS
Prior work had indicated that three special foam agents
might be effective against the propellant materials. Two systems
have been fairly well documented in the prior work. These were
systems which gelled on contact with alkaline materials. One
system used an acrylic polymer and the second employed a carboxy
vinyl polymer. Both systems had evolved during a program aimed
at developing an effective foam system to control the vapor
hazard from ammonia. The difference between the two systems con-
cerns the rate of gellation and the stability of the gel which
fo rms.
Both foams were very effective in controlling the vapor re-
lease from ammonia, being able to keep the concentration below
the odor threshold for a measurable period of time. The simi-
larity between ammonia and hydrazine would suggest that these
foams would be effective against that material also.
The third system proposed for consideration utilized silicic
1 5
-------
acid gels. In prior work foam systems had been defined incorpo-
rating sodium silicate (water glass) solutions. Foams formed
from such solutions could be gelled by acidulation, converting
the silicate to silicic acid. Several applications had been con-
sidered for this type of foam but no significant effort developed
to exploit this type of system. It thus is in a more rudimentary
state than the foams which gel on the alkaline side.
Testing with Hydrazine
The two gelling agents identified as MSA60 and 95 were eval-
uated against hydrazine in the same fashion as the commercial
agents. Both materials were compatible with hydrazine. The gel-
lation provided a stable foam mass with little collapse. Each gel
eventually broke down but the effective life of the 7 cm (3 in.)
layer approached four hours.
Foam 60 underwent what appeared to be a slight interaction
as it was applied. The drainage formed small gel globules in the
hydrazine. After ten minutes the vapor concentration above the
gelled foam decreased to about 1 ppm and decreased further with
time. After 30 minutes the concentration was less than 1 ppm.
After several hours the gel broke down and vapor control was lost.
Foam 95 formed a gel layer more readily than Foam 60. The
foam expanded slightly in the first few minutes after application
rather than interacting. The drainage formed a gel layer on the
surface of the hydrazine. When the gel formed, the hydrazine
level above the foam was in the range of 0.5 ppm. It remained
at this level for the duration of foam life. The foam did de-
grade but at a slower rate than Type 60 and it did not become as
friable as it aged.
Testing with Nitrogen Tetroxide
As with the commercial agents, all agents selected were
evaluated against both propellant materials. Both gelling foams
were evaluated with ^04 for compatibility. Neither material
could survive against the tetroxide but they were superior to all
other agents so far tested.
Type 60 reacted with nitrogen tetroxide to form a film. This
provided a base upon which a foam layer can be built. This foam
absorbed the vapor and underwent some expansion. The 7.6 cm (3
in.) layer was permeated in about 15 minutes. Once the foam has
been fully permeated by the nitrogen tetroxide it collapses slow-
ly taking about five minutes for full degradation. By making up
the foam level intermittently or by initially using a thicker
layer, some control of the vapor release can be achieved. There
was not continuous release, rather the foam releases vapor in
pulses. The drainage from the foam converted the N 2 0 4 to nitric
acid. This foarn thus provided the best control of all materials
1 6
-------
tested to date.
Foam 95 reacted slowly with N204. There was some film for-
mation but it was discontinuous. In general, it behaved simi-
larly to Type 60 but did not provide the same degree of control.
Some tests were run with silicate foams. None of the
systems tried were successful. The sodium silicate exaggerated
the reaction between the foam and the nitrogen tetroxide which
more than offset any benefit the foam might have. The formation
of a silicic acid gel was never achieved. Some variations in
the materials ratios of the silicates are possible, but there
was not enough time to evaluate all the possible permutations of
N a 2 0 • S i 0 2 ratios.
FOAM MODIFICATIONS
The Type 60 and 95 foams were shown to be quite effective
against hydrazine with Type 95 found to be the superior foam.
Against nitrogen tetroxide no foam truly was effective. Type 60
gave the best response. The fact that it tended to form a film
over the surface of the tetroxide was deemed to be the basis for
its control. As time allowed in the program, a variety of water
soluble polymeric materials were screened to assess their cap-
ability to form a film in contact with nitrogen tetroxide. The
time available in the program only permitted a cursory exami-
nation based upon visual observations of foam-nitrogen tetroxide
interactions. These observations for the materials tested are
tabulated in Table 2 .
In the course of the screening study there were some in-
compatibilities between additives and surfactant materials. The
major problem was a gelling reaction. For some of the promising
additives to be fully evaluated it will probably be necessary to
totally formulate the system optimizing the additive-surfactant
combination. A beginning was made in this direction at the end
of this program. It showed that surfactants based upon olefins
or betaines had the best stability to nitrogen tetroxide.
FOAM-CRYOGEN COMBINATIONS
Tests were conducted to determine the effectiveness of com-
bining cooling agents with foam systems. 3M "Alcohol Solvent
Concentrate" foaming agent was chosen for these tests from among
the four commercial foams that were partially compatible with
N204. The foam had a life of about 15 to 20 minutes (in low ex-
pansion form), when placed directly over the liquid. When the
foam was placed over tetroxide cooled with wet ice, its useful
life was increased to about one hour.
Dry ice and liquid nitrogen used in combination with the
"Alcohol Solvent Concentrate" foaming agent did not perform as
17
-------
TABLE 2. SCREENING OF POTENTIAL FOAM MODIFIERS
Observations of Foam-Nitrogen
Material _ Tetroxide Interaction
Rhoplex Forms globular layer between foam and N204
Acrylic Latex
Acrysol Forms strands of elastomerlc polymer over
Acrylic Latex N204; in some cases forms an interlayer of
semi-solids between the foam and
Polyvinyl Forms an interlayer of semi-solids between
Ether foam and
Polyvinyl Reacts with N204, but does not form an
Alcohol interlayer
Guar Reacts with N204 to form granular ppt but
Derivatives does not form interlayer over N204 with
foams
Polyvinyl Slow dissolution; does not retard reactions
Acetate when included in foam formulations
18
-------
anticipated. The large volume o gas released by the boiling or
sublimation of the coolant expanded the foam layer very rapidly.
Rather than replacing the foam lost to normal collapse, this
rapid expansion of the foam cover made it more susceptible to
collapse from the N02 also being evolved. The working life of
the foam remained almost unchanged when combined with dry ice
and was increased very slightly when used in combination with
liquid nitrogen, only when foam placement was delayed until the
liquid nitrogen evaporated.
Any advantages gained by combining the effects of coolants
and foam agents were only temporary since the coolant-foam
system, when once applied to a spill, could not be replenished
without breaking the foam cover. The combined treatment pro-
cedure thus had only a limited benefit.
19
-------
SECTION 7
PROPOSED PROGRAM CONTINUATION
HYDRAZINE VAPOR CONTROL
aca
abilities of the foam, will allow the design of a foam vapor
hazard protection system for those areas where spill potential
t- A 1 3 U 5 ,
NITROGEN TETROXIDE VAPOR CONTROL
The data developed in the program so far would indicate that
foam containment lies with systems which can develop a film bar-
rier to the nitrogen tetroxide. The acrylic systems seem to have
the best potential. Certain foams appear to have greater possi-
bilities than those tested so far. They need to be mated to a
compatible surfactant material for a fuller evaluation.
Some further effort in silicate systems can be justified.
This should address low Na20 materials to slow down the reaction
with the tetroxide which destroyed the foams in the early tests
with silicate solutions. It is possible that the silicate could
be combined with the acrylic system to compliment each other.
Those suggested efforts must be first conducted within the
laboratory. If a system is found which is effective, it should
be carried forward to a field scale test. These tests would be
the s m as for hydrazine. These should include those restraints
wh c would exist in the field and address subsequent clean up
nroc-dures The final step, as with hydrazi ne woul d be the de
sign of a spill control system for the projected hazard.
The foregoing would appear to suggest separate sys terns for
e
20
-------
IP
1. REPORT NO.
TECHNIC L REPORT DATA AWBERQ UBRAf^y J/S
'L'SS? read Instructin" :-i tilt! reverse before completing) '
2. 3. RECIPIENT'S ACCESSION NO.
i
4. TITLE AND SUBTI , _ -
MODIFICATION OF SPILL FACTORS AFFECTING AIR POLLUTION
Vol. II - The Control of the Vapor Hazard From Spills
Of Liquid Re :ket Fuels
7. AUTHOfl(S)
J.S. Greer, S.S. Gross, R.H. Hiltz, M.J. McGoff
9. PERFORMING ORGANIZATION NAME AND ADDRESS
MSA RESEARCH CORPORATION
Division of Mine Safety Appliances Company
Evans City, Pennsylvania 16033
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati , Ohio 45268
5. REPORT DATE
May 1981
6. PERFORMING ORGANIZATION CODE
s. PERFORMING ORGANIZATION REPORT r
80-197
EPA
10. PROGRAM ELEMENT NO.
1 1. CONTRACT/GRANT NO.
68-03-2648
i
13. TYPE OF REPORT AND PERIOD COVERED ]
Final - ,1an 1 QSO - Ror 1 PRO
14. SPONSORING AGENCY CODE
!
15. SUPPLEMENTARY NOTES !
Monitoring Agency: Rockwell International, Newbury Park, California 91320
15. ABSTRACT
The hypergolic rocket fuels, hydrazine and nitrogen tetroxide, are volatile hazardous
materials of special interest to the Air Force. Through monitoring of ongoing Envi-
ronmental Protection Agency programs, the Air Force has maintained cognizance of the
developing state of the art in spill control. This Air Force supplement to the basic
EPA program was a preliminary evaluation of the potential of cooling and foam covers
to mitigate the vapor hazard from hydrazine and nitrogen tetroxide.
Coolants exhibited some control over vapor release from the hypergolic fuels. Liquid
nitrogen was the most effective material. Logistics were deemed a major disadvantage
for the anticipated spill scenarios.
Foams using commercial agents were beneficial with hydrazine but were not effective
against nitrogen tetroxide. Modified foam systems incorporating acrylic resins were
more effective. They were able to maintain hydrazine concentrations at or below 0.5
ppm. Some control was also exhibited with nitrogen tetroxide, but there was inter-
mittent vapor release through the foam.
Based upon the work of this program, the acrylic foams offer a promising approach to
the control of the vapor hazard from hydrazine and nitrogen tetroxide.
17. KEY WORDS AND DOCUMENT ANALYSIS '
a. DESCRIPTORS
Evaporation Control, Hydrazine,
Nitrogen Tetroxide, Hazardous Materials,
Chemical Spills
is. DISTRIBUTION STATEMENT
Release to pub! ic
b. IDENTIFIERS/OPEN ENDED TERMS
Spill Control, Hazard-
ous Materials Spil Is,
Hydrazine, Nitrogen
Tetroxide, Vaporization
Control , Foam Systems
19. SECURITY CLASS (This Report/
INCLASSIFIED
20. SECURITY CLASS iThil page)
UNCLASSIFIED
c. COSATl Field/Group
i
21/09
07/04
06/20
i
!
i
21. NO. OF PAGcS
26
22. PRICE
EPA Form 2220-1 (Rev. 4-77! PREVIOUS EDITION 13 OBSOLETE
22
------- |