EPA-R2-73-251
MAY 1973 Environmental Protection Technology Series
Feasibility of Plastic Foam Plugs
for Sealing Leaking
Chemical Containers
Office of Research and Monitoring
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and
Monitoring, Environmental. Protection Agency, have
been grouped into five series. These five broad
categories were established to facilitate furtber
development and application of environmental.
technology. Elimination of traditional grouping
was consciously planned to foster technology
transfer and a maximum interface in related
fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3, Ecological Research
e. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL
PROTECTION TECHNOLOGY series. - This series
describes research performed to develop and
demonstrate instrumentation, equipment and
methodoloqy to repair or prevent environmental
degradation from point and non-point sources of
pollution. This work provides the new or improved
technology required for the control and treatment
of pollution sources to meet environmental quality
standards.
-------
EPA-R2-73-251
May 1973
FEASIBILITY OF PLASTIC FOAM PLUGS FOR
SEALING LEAKING CHEMICAL CONTAINERS
By
R. C. Mitchell
C. L. Hamermesh
J. V. Lecce
Project #15090 HGW
Contract 68-01-0106
Project Officer
Ira Wilder
EPA
Edison Water Quality Research Laboratory, NERC
Edison, New Jersey 08817
Prepared for
OFFICE OF RESEARCH AND MONITORING
U. S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D. C. 20460
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402
Price 85 cents domestic postpaid or 60 cents OPO Bookstore
-------
EPA Review Notice
This report has been reviewed by the Environmental Protection
Agency and approved for publication. Approval does not sig-
nify that the contents necessarily reflect the views and poli-
cies of the Environmental Protection Agency, nor does mention
of trade names or commercial products constitute endorsement
or recommendation for use.
11
-------
ABSTRACT
A program was conducted to evaluate the feasibility of
methods for plugging leaks in damaged chemical containers
by application of suitable plastic barriers. Such a sys-
tem would be valuable in helping to prevent water pollu-
tion from spilled hazardous chemicals.
A large number of candidate sealants were evaluated in
laboratory screening tests, including various urethane
foams; polystyrene and polyvinyl acetate instant foams;
filled and unfilled epoxy systems; and polysulfide, butyl,
neoprene, and silicone rubber systems. The most promis-
ing results were obtained with the urethane foams. Addi-
tional evaluation and scaleup tests were made, including
sealing of leaks of many different hazardous chemicals,
application to leaks both under water and in air, and
sealing of leaks in 55-gallon containers.
The feasibility of this concept was demonstrated. As a
consequence of the success already realized, it is prob-
able that a practical and useful systems embodying this
approach, can be developed.
This report was submitted by the Rocketdyne Division of
Rockwell International in fulfillment of Project Number
15090 HGW, Contract 68-01-0106, under the sponsorship
of the Environmental Protection Agency. Work was corn—
pleted in August 1972.
This report has been given the Rocketdyne Internal Report
Number R-9054.
111
-------
CONTENTS
Abstract
Sect ions
Page
111
I Conclusions and Recommendations .
I
II Introduction . . . 3
III Screening and Evaluation Tests
Selection of Sealants
Testing
IV Scaleup Tests arid Feasibility Demonstration 11
V Preliminary Design and Development
Envelope of Practical Applications
Survey of Available Application Hardware
Preliminary Development Tests .
Chemicals Which can be Sealed .
Preliminary Design Concepts .
VI Acknowledgements
VII References
VIII Publications and Patents
IX Glossary
27
27
30
32
37
40
47
49
51
53
Appendix
Derivation of Equations for Envelope of Practical Applications
55
5
5
6
V
-------
12
• . . . 14
• . . . 15
• . 16
17
• • • . 18
• . 19
• . 20
• . 21
• • • • 22
23
24
25
• • 29
• . 35
• . 35
• 36
• 41
• 42
43
43
55
FIGURES
1. Laboratory Mixing/Delivery System for Potential Sealants
2. Simple Applicator With Foam
3. Sealing a 1-1/2-Inch Hole
4. Completed Seals
5. Improved Applicator
6. Preparing to Seal a 3-Inch Hole
7. Moving the Applicator Into Position
8. With Plunger in Place, Foam is Extruded Into the
3-InchHole
9. When Leak Has Been Stopped, Applicator is Extracted
10. 3-Inch Hole Securely Sealed With Foam
11. Foam Plug Formation Viewed From Inside Chemical Container
12. Ruggedness of Foam Plug
13. Benzene Leak Sealing Under Water
14. Envelope of Practical Applications for Sealing Leaks in
Nonsubmerged Containers
15. Pressurize Foam Cylinders With Rocketdyne Mixing Gun
16. Pressurizable Foam Cylinders, Mixing Gun, and
Polyethylene-Lined Plunger- In-Cylinder Applicator
17. Use of Mixing Gun to Supply Foam to Cylinder Applicator
18. Foam-Filled Balloon Concept
19. Hollow Rubber Cone Concept
20. Moving Applicator Concepts
21. Foam Cocoon Concept for Leaking Valves and Open Pipes
22. Sketch of Basic Container in Different Attitudes
vi
-------
TABLES
1.. Potential Sealants Screened . . 7
2. Hazardous Chemicals Tested and Small-Scale Sealing
Test Results 8
3. Probable Leak Sealing Success With 100 Hazardous
Chemicals . . . . 38
vii
-------
SECTION I
CONCLUSIONS AND RECOMMENDATIONS
CONCLUSIONS
1. There is a vital need for systems that can prevent dispersion of a
hazardous chemical. It is far better and easier to prevent a hazard-
ous chemical from entering a waterway than it is to attempt to remove
it after entry.
2. The feasibility has been demonstrated of a concept to help prevent
dispersion of leaking hazardous chemicals by application of a suit-
able plastic barrier (e.g., plastic foamed in place) to a leaking
chemical container to plug, seal, or enclose openings such as those
from cracks, ruptures, open seams, and damaged valves.
3. There are a number of potential sealants that might be considered for
use in a leak-plugging system (e.g., various types of polyurethane
foams, polystyrene foams, polyvinyl acetate foams, epoxy systems,
and rubber systems). Of those tested, the most promising are the
urethane foams.
4. Urethane foams can be used to seal leaks of a large variety of chem-
icals such as methanol, insecticides, benzene, and styrene.
5. The potential of this leak-sealing concept is great enough to justify
its development for operational use. It is probable that a practical
and useful system embodying this concept can be developed
RECOMMENDATIONS
1. It is recommended that a prototype of an operational system be devel-
oped and demonstrated embodying the use of urethane foams applied in
place to plug leaking chemical containers (considering a broad range
of leaks, e.g., cracks, ruptures, open seams, damaged valves, etc.).
2. It is recommended that a careful evaluation be made to assess whether
it is necessary and practical to attempt development of special seal-
ants and techniques for sealing leaks of particularly difficult chem-
icals that cannot be satisfactorily sealed with urethane foams (e.g.,
strong acids).
1
-------
SECTION II
INTRODUCTION
In a highly industrialized society, tremendous quantities of chemicals
are produced and shipped to other sites for various uses. Many of these
chemicals are quite hazardous when introduced into natural waters. The
likelihood of accidental release to watercourses is increased by the many
operations involved in the production, transfer, shipping, unloading, and
ultimate utilization of the chemicals. Thus, it is not at all surprising
that spills do occur.
Spills can occur in a variety of ways. For example, the most dramatic
and generally most serious type occurs when a container is violently
ruptured (sometimes accompanied by fire or explosion) and large quanti-
ties of the hazardous material are spilled almost instantaneously. A
less catastrophic spill results when the container maintains its integ-
rity, but suffers enough damage to cause leakage of the hazardous mate-
rial at a moderate rate. The leaking chemical can enter the watercourse
either directly (for example, as the result of a barge accident or a
land-based container falling into the water) or by flowing or being washed
into a drainage channel or percolating into the ground water supply.
Countermeasures that neutralize or treat hazardous chemicals that are
already mixed and in the waterway may require hours or days to be initi-
ated and involve the very difficult problem of handling large volumes of
water. Therefore, there is a vital need for a system that can prevent
further dispersion of the hazardous chemical by stopping the leak, regard-
less of whether the leaking container is on land or under water. The
program described herein was directed toward this need, with the under-
lying philosophy that it is far better and easier to prevent a hazardous
chemical from entering a waterway than it is to try to remove it after
entry.
The goal of the program was to demonstrate the feasibility of a concept
for plugging leaks of hazardous materials. The concept involves applica-
tion of suitable plastic barriers (e.g., plastic foamed in place) to a
leaking chemical container to plug, seal, or enclose cracks, ruptures,
open seams, damaged valves, etc. The major tasks involved: (1) screen-
ing tests to determine the suitability of various currently available
plastics (used in the broad sense to include rubbers, foamed systems,
etc., in addition to polymer systems more normally considered as “plastics”)
as sealants for selected hazardous chemicals, such as methanol, insecti-
cides, benzene, and styrene; (2) evaluation of parameters involved in
sealing leaks with selected sealant/hazardous chemicals; (3) demonstra-
tion at about the 10-gallon level of the feasibility of this concept; and
(4) preliminary work toward design and development of an operational
system.
3
-------
SECTION III
SCREENING AND EVALUATION TESTS
SELECTION OF SEALANTS
The objective of this phase of the program was to screen and evaluate,
on the laboratory scale, the utility of a wide variety of materials as
potential sealants for leaks from chemical containers. Several criteria
were selected for evaluating the sealants, including:
1. Sealants should develop sufficient strength and adhesion to a
painted metal container to completely seal a leak.
2. The setting time of the sealants should be short.
3. Sealants should fill a hole completely, regardless of the irreg-
ularity of the opening.
4. Sealants should have satisfactory chemical r&sistance to import-
ant hazardous chemicals such as benzene, phenol, methanol, and
insecticides.
The sealants tested can be divided into two groups: expanding and non-
expanding types.
The expanding sealants included various types of foam systems such as In-
stant Foams (Monsanto Corporation, St. Louis, Missouri) and urethane
foams [ Nagle Corporation, chicago, Illinois; MSA Corporation, Pittsburgh,
Pennsylvania; Expanded Rubber and Plastics Corporation, Torrance, Calif-
ornia (Stafoam); and Olin Corporation, New Haven, Connecticut (Auto-
froth)]. These materials have desirable attributes for this application
in that the foams will conform to an irregularly shaped hole and the ex-
pansion of the foam should result in a tighter seal. This approach
assumes that the foam ingredients are applied either before expansion
occurs or while it is occurring. Thus, the normal extent of shrinkage
which accompanies curing is not likely to result in the foam pulling
away from the holes. Should shrinkage be so large that some leaks remain,
a second application of foam to these few sites could correct the problem.
Foaming occurs as the result of expansion of trapped gas. The gas may be
generated by a chemical reaction such as that of an isocyanate and water
to produce carbon dioxide or by the decomposition of a blowing agent,
such as N,N’-dimethyl N,Nt_dinitrosoterephthalamide (Nitrosan, DuPont
Corporation, Wilmington, Delaware). In some foams, volume increase is
due to a physical process in which a low-boiling chemical which was dis-
solved in the formulation at higher pressures vaporizes when the formu-
lation is released from a container at ambient pressures. Some foams
employ both mechanisms for volume increase. The foam systems screened
in this program included ones employing each mechanism arid others with
the combination.
5
-------
An equally important consideration is whether a foam system undergoes a
chemical reaction during the expansion, resulting in a crosslinked pro-
duct. Crosslinking is an advantage for such foams in that a crosslinked
material is considerably more resistant to chemical attack than the cor-
responding uncrosslinked polymer. A shortcoming of crosslinked systems
is that the chemical leaking from the container may inhibit or retard
the crosslinking reaction, and thus decrease or destroy the utility of
the foam.
The nonexpanding sealants examined included materials that reacted to
achieve crosslinked structures as well as systems that were thermoplstic
and served merely as mechanical seals. Various epoxy systems (both filled
and unfilled), quick-curing silicone rubbers, and catalyzed polysulfide
rubbers are examples of the former. Swollen neoprene, butyl rubber, and
uncatalyzed polysulfide rubbers are all of the thermoplastic nonexpanding
type.
ThSTING
Initial screening tests were made using 8-ounce cylindrical paint cans as
the containers for the leaking liquids. The cylindrical cans were painted
with baked alkyl resin enamel (yellow drum paint, Techniform Laboratories,
Venice, California). This enamel is used to paint the outside surface of
chemical drums. Adhesion to an unpainted metal surface is generally
greater than to a low-surface-energy painted surface. The worst case of
adhesion is that of a greasy or dirty surface. However, this condition
is difficult to reproduce; therefore,painted metal surfaces were used as
a standard low-energy test surface. Sealants that were effective in
plugging leaks because of their adherence to the outside of the containers
could,therefore,be screened on these small cans with assurance that their
adhesion behavior on large-scale containers would be similar. Holes of
various sizes from about 1/8- to 3/4-inch diameter were punched near the
bottom of the can, and the test hazardous chemical was poured into the
can, while covering the hole temporarily, to prevent leakage while filling
the can. At the start of the test, the cover over the hole was removed
and the sealant candidate was applied using a spatula or spraying foam
directly from its dispenser as rapidly as possible while the chemical was
flowing through the hole. This procedure was employed for tests both in
air or under water.
Table 1 lists the materials screened as potential sealants and their per-
formance in plugging small-diameter, low-head leaks of either benzene or
methanol, primarily in tests as described in the previous paragraph. As
a result of these early successes, testing was expanded to include most
of the hazardous chemicals that are ranked among the top 20 in the pri-
ority list (Ref. 1 and 2). In Table 2, these chemicals are listed as
well as the degree of success obtained using various sealants (primarily
polyurethane foams or epoxy resins) to seal holes in painted containers.
Most of these tests were made with 1/4- to 3/4-inch holes in 8-ounce con-
tainers. It should be noted that even with these off-the-shelf unopti-
mized sealants, successful plugging of leaks of most of these chemicals
6
-------
TABLE 1. POTENTIAL SEALANTS SCREENED
Urethane Foam
(St afoan)
Urethane Foam
(Autofroth)
Polyurethane
Putty (tJrabond)
Epoxy Putty
(Po lypoxy)
Epoxy Putty
(Epoxylite 3351)
Epoxy Putty (Sea
Coin ‘-Poxy Putty)
Steel-filled Epoxy
Putty (Devcon SF)
Lead-filled Epoxy
Putty (Devcon L)
Polysulfide Rubber
Putty (EC612)
Butyl Rubber Sealer
(EC12O2-T)
Swelled Uncompounded
Neoprene Rubber
Rapid-curing
Silicone Rubbers
*setting tines of less than about
Very rapid foaming; no adhesion to metal; may
be suitable for mechanical seal
Very rapid foaming; no adhesion to metal; may
be suitable for mechanical seal
Very rapid foaming; no adhesion to metal; may
be suitable for mechanical seal
Aerosol-type cans give inadequate mixing and
delivery; setting tine probably too long to
be useful*
Components may be promising, if formulation
is modified for shorter setting time*
Type of Material
Source
Indication of Results
Polystyrene Instant
Foam
Modified (rubbery)
Polystyrene Instant
Foam
Polyvinyl Acetate
Instant Foam
Urethane Foam
Urethane Foam
Monsanto
Monsantao
Monsanto
Nalge
MSA
Expanded
Rubber and
Plastic
Olin
Poly Resins
Pettit Paint
Epoxylite Corp
Permal it e
Plastics
Devcon Corp.
Devcon Corp.
3M
3M
Rocketdyne
Preparation
Polysil
Setting time of basic components much too long;
very promising when modified for shorter
setting time*
Several formulations available; promising
characteristics
Sets up slowly on exposure to moisture or
liquid water
Setting time too long*
Setting time too long*
Promising (shorter setting time still needed)*
Sets up fairly rapidly to form high-strength
solid
Sets up fairly rapidly to form high-strength
solid
Low strength requires backing for permanent
seal
Too weak in tape form for permanent seal
High flexibility helps plug irregular holes
but is extruded slowly by hydrostatic head
Several formulations tested; insufficient
rigidity to seal large holes
10 seconds are desired
7
-------
TABLE 2. HAZARDOUS CHEMICALS TESTED AND SMALL-SCALE SEALING TEST RESULTS
1
Sealed leaks of aqueous solutions (50 to 90
“A” and “W” indicate tests performed in air
0Q
Hazard Ranking
(Ref. 2)
Hazardous Material
Results of Sealing Tests
Stafoam Urethane
Foam
Sea-Goin’
Epoxy Putty
Other Sealants
1
Phenol
Sealed (A)t
Failed (A)
Butyl Rubl r-Sealed
(A)
2
Methyl Alcohol
Sealed (A,W)
Sealed (A,W)
Epoxy Putty-Sealed
MSA Urethane-Sealed
(A)
(A)
3,8,
14, 18
Insecticides, Rodenticides:
DDT 95 o Solu./Water
Dieldrin
Sealed (A)
Sealed (W)
Sealed (A)
4
Acrylonitrile
Failed (A)
Failed (A)
Polysulfide Rubber-
Sealed (A)
5
Chiorosulfonic Acid
Failed (A)
6
Benzene
Sealed (A,W)
Sealed (A,W)
MSA Urethane-Sealed
(A,W)
9
Phosphorus Pentasulfide
Failed (A)
10
Styrene
Sealed (A)
Sealed (A)
11
Acetone Cyanohydrin
Sealed (A)
13
Nrrnyl Phenol
Sealed (A,W)
Sealed (A,W)
15
Isoprene
Sealed (A)
Sealed (A)
16
Xylene
Sealed (A)
17
Nitrophenol
Sealed (W)
Sealed (W)
percent) for 18 hours, then failed.
and submerged in water, respectively.
-------
was accomplished. Both chlorosulfonic acid and phosphorus pentasulfide
could not be sealed. These chemicals probably require a very special
sealant because they may be too reactive with water to permit practical
sealing. Most of the other top 20 hazardous chemicals can probably be
sealed with any sealant that is effective with either benzene or phenol.
For example, sealants useful for benzene leaks would also likely be effec-
tive against the other hydrocarbons, styrene, isoprene, and xylenes.
Sealants useful for phenol would probably also seal acrylonitrile and
acetone cyanohydrin.
As indicated previously, some of the systems evaluated, both the expand-
ing and nonexpanding types, react to yield crosslinked products. For the
epoxies and urethanes, these crosslinking reactions are base catalyzed.
Therefore, acid interferes with the reaction and attacks the cross linked
product. Thus, while these cured sealants may plug leaks of nonpolar
chemicals, they are unsuitable with phenol or acrylonitrile. The fail-
ure to seal may be due either to the phenol, which is acidic and inter-
feres with the cure, or the lack of resistance even of the cured sealants
to these potent solvents. The temporary (18 hours) sealing of phenol bya
urethane foam reported in Table 2 indicates that for this specific sys-
tem, lack of resistance to the solvent is the mode of failure, rather
than inadequate curing.
The four different types of rubber tested are inert to phenol and,there-
fore. could seal phenol leaks. However, they swell or dissolve in aro-
matic hydrocarbons. Uncompounded Neoprene W (Elastomer Chemicals, DuPont
Corporation, Wilmington, Delaware) was allowed to swell in either carbon
tetrachioride or benzene, forming a relatively soft, highly spongy mass
before testing as a sealant. In this condition, it successfully sealed
a 2-inch-diameter hole near the bottom of a 5-gallon drum full of water.
However, the material still had a finite viscosity and so was slowly
extruded through the hole; a permanent seal was required to prevent the
leak from redeveloping.
One. technique developed and tested to improve the ability of the nonexpand-
ing materials to “bridge” a hole was to use a supporting or backing mate-
rial. In its simplest form, this concept involved use of a “patch” in
which the sealant was coated on the supporting fabric (e.g., nylon cloth)
and the patch was then applied to the leak. Another 2-inch-diameter hole
was permanently sealed by first applying a plug of the poly-sulfide rubber
putty (3M Sealer EC-612, Minnesota Mining and Manufacturing, Minneapolis,
Minnesota), followed by a patch of Sea Goin’ Poxy Putty (Catalog Number
1324, Perinalite Plastics, Costa Mesa, California) on a strip of nylon
cloth that encircled the container. Several silicone rubber formula-
tions were investigated using the nylon cloth technique. They had excel-
lent chemical resistance, and it was possible to shorten the curing time
to the order of 10 seconds; however, the mechanical properties (primarily
rigidity and adherence) were not satisfactory for plugging leaks.
9
-------
The “instant foams” consist of a polymer dissolved in a very low-boiling
solvent and contained in pressurized cylinders. Upon opening a valve, a
solid foam is formed very quickly. These foams were not promising for
leak sealing, primarily for two reasons: the extremely fast solidifica-
tion of the foam and the lack of any significant degree of adhesion to
metal surfaces. Attempts to use these materials to seal leaks by forming
mechanical plugs were not successful. While it is conceivable that other
higher-boiling solvents could be used which would vaporize at a slower
rate and thus delay setting of the foam, the current high cost of the
foams and their limited availability led to the decision not to pursue
this approach further.
A number of the potential sealants had some desirable characteristics,
e.g., they either filled irregular holes, adhered well to metal, cured
rapidly, or gave a strong cured product. However, for most of the sol-
vents, at least one or more of these properties was seriously unsatis-
factory. As a result of these tests, it was concluded that the urethane
foams had the most potential for use in the larger-scale tests. This
decision was based not only on the considerations enumerated above, but
also on such factors as the ability to dispense large volumes of foam.
This is of vital importance in large-scale tests.
10
-------
SECTION IV
SCALEUP TESTS AND FEASIBILITY DE?VL NSTRATION
The primary purpose of this phase of the program was to evaluate the abil-
ity of a foamed system to plug leaks from larger holes and under conditions
where the hydrostatic heads were greater than those encountered in the
laboratory tests. Other factors that were examined included: barrier
type (e.g., patch, outside plug, plug expanded outside and inside), thick-
ness of barrier, speed of application, type of applicator, technique of
application, and temporary support during the curing stage.
Much of the preparation of foams in this phase employed hand mixing in
beakers. This technique was used to permit rapid evaluation and accurate
control of composition of the foams in these tests. By weighing out the
ingredients and then hand-mixing, the ratio of components could be care-
fully controlled with little effort. However, it was obvious that to
employ this concept practically, a more flexible mixing/delivery system
was required. Since no available mixing system was found that could per-
form the necessary functions, a system was designed and constructed
(Fig. 1). It uses a 1/4-inch diameter in-line static mixer (Kenics Cor-
poration, Danvers, Massachusetts) and can continuously supply a mixed,
metered, two-component product over a range of small flowrates (up to about
1 lb/mm). There are three chemical supply cylinders, fabricated of 2-
inch-diameter, stainless-steel tubing approximately 12 inches long; two
were used for the foam constituents and the third was used for a solvent
with which the system could be flushed (methylene chloride was used as
the solvent with urethane foams). An inert gas pressurization system was
provided (nitrogen was used) to transfer chemicals from the supply cylin-
ders. This device was used for tests with urethane foams, and could have
been modified to permit its use with more viscous systems, such as the
epoxy formulations. This mixing system was used only for a limited num-
ber of tests because it quickly became desirable and possible to progress
to a much larger scale.
Early tests in this phase indicated that a key to successful plugging of
large holes against appreciable hydrostatic heads is the expansion of the
foam both inside and outside the hole to form a structural bridge. To
achieve this, it is necessary to apply the foam at a point in its curing
cycle where the foam is sufficiently mobile so that it can flow through
the hole into the container and then expand to provide a bridge across
the opening. However, during this stage, the foam must also have suffi-
cient internal cohesion so that the chemical pouring from the container
does not wash it away by mechanical action. It is desirable for the foam
to be still capable of adhering to the container surface so that an inti-
mate and strong bond can be achieved. These requirements led to the con-
cept of either a special applicator design (large cross section, long
residence time) or a two-stage application technique. In the latter con-
cept, the foam is mixed externally and allowed to cure for a few seconds
until it reaches the right stage of tackiness. It is then applied to a
leak with some type of device that provides temporary support for an addi-
tional few seconds until sufficient rigidity has developed.
11
-------
5DZ11-l/28/72-S1J
Figure 1.
Laboratory Mixing/Delivery System for Potential Sealants
I ,
PRESSURIZATION
LI NE
MICROMETER
METERING
VALVES
F L L L
THREE-WAY
BALL VALVES
I
._ UPPLY 1
1 CYLINDERS: TWO
(IN FRONT) FOR
FOAM COMPO-
NENTS, ONE FOR
FLUSHING
SOLVENT
IN-LINE STATIC
MIXER
SAMPLE OF
ED FOAM
F
f
12
-------
This two-stage application technique was tested in two simple forms
(described in the following paragraphs), with considerable success.
Figures 2 through 10 illustrate some of the large-scale tests that were
made with polyurethane foams. The simplest of the two applicator forms
(Fig. 2 through 4) involved the use of a large suction cup (“plumber’s
helper”), which was filled with foam that was allowed to set for a few
seconds until the proper tackiness was reached. The cup was then placed
against the leaking stream and some of the foam was forced through the
hole with the plunger. The sides of the cup provided enough restraint
so that the necessary compression of the foam cells could be achieved,
resulting in expansion on the inside of the hole. It was not necessary
for the cup to conform exactly to the surface of the container. This
technique was used successfully to seal a 3-inch hole in a 55-gallon drum
with 3 feet of head. It was also used successfully to seal 3-inch under-
water leaks. Figure 2 shows a rubber plunger filled with foam prior to
use. In Fig. 3, the plunger is being used to plug a 1-1/2-inch-diameter
hole in a 55-gallon chemical drum. Two completed seals are seen in Fig. 4.
In some cases, the rubber plunger cap adhered to the seal and was left in
place but,in other cases, it could be removed after the foam set.
The second of the two applicator forms tested was a simple version of a
piston in a cylinder, shown in Fig. 5. A charge of premixed but uncured
foam was placed in the cylinder, either from hand mixing or from an in-
line foam-mixing gun. Then the open end of the applicator was placed
over the leak. Although Fig. 6 and 7 show the operator moving the appli-
cator directly into the liquid jet, it is preferrable in some cases to
move the applicator into position from the side of the jet. Then, with
the applicator in place, the piston is moved to force the foam against
and through the leak opening (Fig. 8). Figures 9 and 10 show the com-
pleted seal of this 3-inch-diameter leak using the Stafoam polyurethane
foam.
Using the technique described above, a silent color motion picture (Ref. 3)
showing several tests involving the sealing of leaks of benzene, methanol,
and water from 55-gallon drums both in and out of water was prepared and
submitted to the Environmental Protection Agency. This film graphically
demonstrates the feasibility of the concept for plugging leaking chemical
containers.
Figure 11 contains four frames from this motion picture, showing the for-
mation of a foam plug in air, viewed from inside the container. The first
frame shows the hole, looking from the inside of the container, with the
liquid flowing outward. In the second and third frames, the foam plug
and gas bubbles can be seen leaving the foam surface. In the fourth
frame, the plug is completely formed. The plug formation was very rapid--
approximately 1/2 second for the entire process.
To illustrate the ruggedness of a foam plug, Fig. 12 shows additional
frames taken from the motion picture. After the outside portion of the
plug is sawed off and removed, the remaining portion of the plug inside
13
-------
. 4
¶ — —
a
a
S
I
0
Figure 2. Simple Applicator With Foam
SDZ 1 1-l/28/72-S 1E
-------
5DZ 11-1/28/72-S1F
U i
i,i
1.
Figure 3. Sealing a 1-1/2-Inch Hole
-------
414
0•
5DZ11-l/28/72-S1A
Figure 4.
Completed Seals
C’
-------
Figure 5. Improved Applicator
5DZ11-1/28/72-S1D
-------
I
5DZ11-1/31/72-S1A
/
.
Preparing to Seal a 3-Inch Hole
Co
: •
I
Figure 6.
-------
‘p
5DZ 1 1-l/3 1/72-S1D
‘ 0
4 *
I
Figure 7. Moving the Applicator Into Position
-------
L
5DZ11-l/31/ 72-SiC
Figure 8. With Plunger in Place, Foam is Extruded Into the 3-Inch Hole
0
-------
SDZ1 1-1/31/72-SiB
Figure 9. When Leak Has Been Stopped, Applicator is Retracted
p- I
-------
Figure 10. 3-Inch Hole Securely Sealed With Foam
t
SDZ11-1/31/72-S1E
-------
(a) Chemical is flowing outward b) Foam is being applied from
through hole in container wafl outside of container and
starts to extrude through
hole
(c) Foam extrusion is completed;
gas bubbles escape from curing
foam
Figure 11. Foam Plug Formation Viewed From
Inside Chemical Container
(d) Plugging is complete
23
-------
(c) Remaining portion of plug inside
container is sufficient to func-
tion as a seal
Figure 12. Ruggedness of Foam Plug
I
F4 ___
(a) Completed foam plug is being
sawed flush with drum outer
surface
-.. . ______
(b) Outer portion of foam plug
is removed
24
-------
the container is sufficient to continue to function as a seal, as shown
in the bottom frame in Fig. 12.
A benzene leak-sealing test in which the seal is applied underwater is
shown in the sequence of photographs given in Fig. 13. The benzene can
be seen gushing from the hole in the first frame. The application of
urethane foam to seal the leak is shown in the second and third frames.
The completed plug, lifted above the water surface, is shown in the
fourth frame.
Figure 13. Benzene Leak Sealing Under Water
25
-------
SECTION V
PRELIMINARY DESIGN AND DEVELOPMENT
Based upon the successful demonstration of the feasibility of the concept,
preliminary investigative work was started toward eventual development of
a system for operational use. This work is summarized in the following
subsections, describing: the results of a study to assess the range of
applicability of this type of leak-sealing system, a continual survey of
available foam application hardware, preliminary development tests that
were made, predictions of the hazardous chemicals which can be sealed,
and some preliminary design concepts that were generated.
ENVELOPE OF PRACTICAL APPLICATIONS
A question of considerable importance is: What types of hazardous spills
can be effectively treated by use of the sealing techniques being devel-
oped? A preliminary study was made to define an envelope of applicability
in terms of characteristic parameters of a leaking container (size of hole,
size of container, and liquid head in the container) within which this type
of technique should be considered.
It is obvious that a very large hole in a small container will result in
a complete loss of the contents before corrective action can be taken.
Therefore, there is an upper limit on hole size (as a function of con-
tainer size, liquid head, and response time) which it is practical to
consider. Efflux times can be calculated for any given container geom-
etry and orientation, hole size and location, and type of liquid. An
analysis was made, using the following assumptions and definitions:
1. Leak sizes may be specified by a single characteristic diameter, D.
2. Containers have a cross-sectional area that does not vary with
liquid height and is specified by a characteristic diameter, Dc•
3. Leaks occur at the bottom of the vessel.
4. The effective container height, H, divided by the characteristic
diameter is a known value, e (generally less than the actual
container height divided by the actual diameter).
The time required for the liquid level in the container to drop to some
fraction (f) of H is (the derivation is given in the Appendix):
r 21 V 5 ’ 6
T = 8.21 [ 1 - (f)l/ j 1 (1)
where T is time in seconds, V is the initial container volume in gallons,
C is the discharge coefficient of the hole, and D is given in inches.
27
-------
Using values of C = 0.61, T = 120 seconds, and f = 0.70 (i.e., losing 30
percent of the contents in 2 minutes), the resulting relationship between
the leak size and the size of a container is:
0.135 V 5 ’ 12
(2)
e
The results are plotted as curves Al (for e = 1.5) and A2 (for e = 0.5)
in Fig. 14.
Another boundary condition can be derived by considering the limits of
human strength. It is assumed that the leak-sealing technique to be
developed must be suitable for application by one man. Any of the sys-
tems envisioned at present would require that essentially all of the flow
be stopped at once. This means that the operator must be able at least
to exert a force sufficient to hold a plug in place against the hydraulic
pressure of the leak for a few seconds. Alternatively, the force required
to overcome the momentum of a liquid stream could be as much as 20 percent
greater (see the Appendix for details); however, this can probably be
reduced substantially by design of an applicator to deflect at least part
of the momentum of the fluid stream. Consequently, the limiting con-
straint, based on human strength, was taken as the force required to with-
stand the static pressure of the leak. The maximum usable force which
can be exerted in a horizontal direction by one man (F) is estimated to be
50 lbf. Using the same assumptions as before, a relationship can be de-
rived (see the Appendix) for the maximum hole size that can be plugged in
tanks of various sizes. For V in cu ft, D in feet, F in lbf, and the
fluid density, p, in lb/cu ft , the relationship is:
= . 2 e 6( 3 *)3 (3)
This equation (with p = 62.4 lb/cu ft and F = 50 lbf) is plotted as curves
81 (for e = 1.5) and B2 (for e = 0.5) in Fig. 14.
The area to the left of both curves Al and Bi (for e = 1.5) or A2 and B2
(for e = 0.5) is then the applicable area for development of leak-sealing
techniques. It should be noted that the assumptions used are “worst case”
(e.g., the hole was assumed to be at the bottom of the container, whereas
it actually could be anywhere). Also, a circular hole was assumed, pro-
viding the greatest distance to be bridged by the sealant for a given leak
size. The total region for practical applications of this concept is much
broader than the limiting boundaries shown in Fig. 14.
The final consideration is to assess the capabilities of the actual leak-
plugging techniques which will be developed (i.e., the maximum hole size/
hydraulic head combinations that can actually be handled by the developed
technique). It already appears promising, from the work of this project,
that the method can be developed to the point of sealing leaks in most or
all of the applicable region to the left of curves A and B in Fig. 14.
28
-------
100,000
80,000
6o,ooo —
140,000 —
20,000 —
10,000
8000
6000
14000 —
U)
z
0
-j
-J
>-
I-
L)
0.
z
z
0
2000 —
1000
800
600
400 —
200
100
80
60
0
CURVE Bi
(e — 1.5)
AREA WHERE HYDRAULIC
PRESSURE IS TOO GREAT
FOR LEAK TO BE STOPPED
BY ONE MAN
CURVE B2
APPLICATIONS
I AREA FOR PRACTICAL
2
CHARACTER I ST IC
3
DIAMETER OF LEAK, INCHES
5
Figure 14. Envelope of Practical Applications for
Sealing Leaks in Nonsubmerged Containers
U)
I-
U)
I-
a
0
-J
\
\
U,
I d
1-
z
0
‘I
-a
a
a
z
AREA WHERE LEAKAGE
RATE IS TOO RAPID
TO ALLOW CORRECTIVE
ACTION BEFORE CONTENTS
OF CONTAINER ARE LOST
40
I
1
4
29
-------
The study outlined in this section indicates that the region of practical
applicability is broad enough and the potential capability of the leak-
sealing system is great enough to justify its development for operational
use.
SURVEY OF AVAILABLE APPLICATION HARDWARE
A preliminary survey was made to investigate commercially available hard-
ware systems that can mix and/or deliver foams, nonfoamed polymer systems,
caulking materials, etc. This survey was limited in scope, but should be
continued and expanded in connection with any subsequent development pro-
ject. This type of information is important to make maximum effective
use of existing hardware in any development work:
A wide variety of mixers and delivery equipment is commercially available
for handling polymer systems of diverse types. In general, this equip-
ment is designed to be used under rather different circumstances than
those encountered in sealing leaking containers of hazardous materials.
To prevent hazardous materials from contaminating a waterway, the appli-
cation of a sealant to the leaking container should meet a number of
requirements, including the following:
1. The sealant should be sufficiently flexible to conform to
irregular holes in the leaking container.
2. The sealant should be (or become) sufficiently rigid to with-
stand appreciably hydrostatic heads.
3. The sealant should be chemically inert both to the hazardous
material and with water to permit underwater sealing.
4. The sealant should have satisfactory temperature limits.
5. Application should be possible to a variety of container surface
geometries and coatings (including dirty container walls).
6. Application should be made shortly after discovery of the leak.
7. Application should be performed without the operator exposing
himself unduly to the hazardous material or its fumes.
S. The applicator should be self-contained and portable.
These requirements might be fulfilled with a number of sealants. The
requirements that the sealant be flexible yet have a high yield stress
can probably best be satisfied by a two-component sealant such as urethane
systems. The unreacted components have a fairly low viscosity, but react
rapidly after mixing to form high-modulus crosslinked polymers. Rapid,
complete mixing of these systems after discovery of a leak will be essen-
tial for successful application. In the normal commercial applications
of polymers, usually not for sealing leaks, emphasis is more often placed
on long pot life (of the order of hours) so that the polymer can be mixed
in larger batches and applied with ample time before it sets up. In
30
-------
sealing leaking containers of hazardous materials, on the other hand, pri-
mary emphasis must be on plugging the leak rapidly to prevent pollution
of the waterway. Both the choice of sealant system and application hard-
ware must fulfill different functions than in usual commercial polymer
applications. To provide background relative to the kinds of hardware
development necessary for prototype design, some typical commercial poly-
mer application equipment is described below.
One of the most directly applicable lines of mixers and delivery equip-
ment that have been identified is supplied by Semco, a division of Pro-
ducts Research E Chemical Corporation (Glendale, California). Their
equipment line includes air-powered and mechanical caulking and sealant
guns that use disposable sealant cartridges containing from 2-1/2 to 12
ounces of material. These guns are portable, especially when the air-
powered ones are driven with a self-contained portable cylinder of gas,
which is also marketed by Semco. Another of their important services
consists of prepackaging custom-designed two- and three-part sealants in
disposable cartridges that fit their guns. The various components are
mixed in the cartridge, just before application, using equipment also
supplied by Semco (but which requires an electrical power source). The
applicators are all simple nozzles; however, openings are available cover-
ing a range of sizes and shapes from a 27-gage needle to one with a rec-
tangular opening 1-3/4-inch wide.
Equipment similar in function and size range to that supplied by Semco
is also marketed by Pyles Industries (IVixom, Michigan). Another line
of urethane foam delivery equipment is available from the Olin Corpora-
tion Plastics Division (Brookpark, Ohio). These have many desirable
features. This kind of delivery hardware sold by Semco, Pyles, and Olin
more nearly approximates the needs for sealing leaks in containers of
hazardous chemica].s than any other equipment about which information has
been obtained; however, their characteristics are not unique.
Much other equipment is available for mixing and applying various cross-
linking polymer systems such as epoxies and polyurethanes. Most of this
hardware is apparently designed for much larger-scale delivery than is
desirable or necessary for the one-shot production of a seal for a leaking
container. For example, Glas-Craft of California supplies a “portable
spray-up system” for the application of glass-reinforced polyester resin
at the rate of 20 lb/mm of laminate. This apparatus is hardly “portable”;
the shipping weight is approximately 400 pounds. A somewhat smaller gel-
coat system marketed by the same supplier is also mounted on a cart, and
both systems are electrically driven. Similar comments about portability
and electrical power requirements can be made about equipment supplied
by the Sealzit Division of Flintkote Company (Riverside, California),
Gusmer Corporation (J-Ioboken, New Jersey), and I-Iardman Incorporated
(Belleville, New Jersey). Other companies contacted whose equipment was
unsuitable for sealing leaking containers rapidly included Hunt Process
Company (Santa Fe Springs, California), Grover Pump Company (Montebello,
California), and Zepco Manufacturing Company (Burbank, California).
31
-------
Although the Semco, Pyles, and Olin equipment are perhaps the best of
these investigated (i.e., the most directly applicable to plugging leak-
ing chemical containers) and include equipment with many useful features,
none of them satisfy all of the requirements needed for the applications
that are the subject of this program. The most serious deficiency is
that only simple nozzles are available for use as the applicator. This
type of nozzle is perfectly satisfactory for the intended purposes of the
equipment, but is completely unsuitable for use in plugging difficult
leaks of hazardous chemicals. A second deficiency (in the case of multi-
component sealants) is that the available mixing arrangements are neither
portable or self-contained (electric power is necessary to drive a mixing
motor). A third deficiency, for adaptation to plugging leaks, is that
the available systems all have a very short distance between the control
point for sealant flow (typically a trigger arrangement) and the delivery
point. The operator of the sealant applicator would, therefore, neces-
sarily be too close to the leaking hazardous material. Greater separa-
tion would be necessary for safety.
The conclusion that suitable equipment for the present application is not
“off the shelf” was further confirmed by a number of individuals thor-
oughly familiar with the industry including personnel at Hardinan Incor-
porated, Permalite Plastics (Costa Mesa, California), and NASA-Ames.
None of the hardware systems thus investigated is completely suitable for
the requirements of plugging leaks of hazardous chemicals. However, some
of the components and features will be useful in subsequent development
work, and should be utilized wherever this is practicable.
PRELIMINARY DEVELOPMENT TESTS
During the scaleup tests, the large suctiOn cup (plumber’s helper) was
used as a foam applicator with considerable success in sealing a 3-inch
hole in a 55-gallon drum with 3 feet of head. The basic plumber’s helper
technique was then improved and the plunger-in cylinder technique was
developed. In this improved technique (which was shown in Fig. 5 through
10 and in the movie film), a length of Teflon pipe was used as a cylinder.
The portion of the cylinder above the plunger was filled with foam, the
open end of the cylinder was placed over the leak and the plunger was used
to force foam against and through the hole in the leaking container. This
application technique, even in its simplest form, was quite successful in
eliminating some of the previous problems associated with deflection of
the applicator and foam.
Not all tests were successful with the plunger-in cylinder technique.
Experience gained during the testing identified several problems associated
with the technique of application, one of which was control of the pres-
sure exerted during application. The appearance of that portion of the
foam extruded through the hole in some unsuccessful leak sealing tests
showed that the seal was broken by forcing more foam through the hole
after the foam had set up. No consistent relationship could be observed
32
-------
between the force exerted on the applicator plunger and the appearance
of the foam after the test, since most of the force was consumed in over-
coming internal friction in the applicator. Although the applicator bar-
rel was made of Teflon, adhesion of foam to the inside surface produced
enough friction to make it impossible to empty the applicator after a
certain degree of cure had been reached. This problem was eliminated by
a modification of the application technique in which the applicator bar-
rel was lined with a thin, disposable polyethylene bag. The foam was
placed into the bag, and the plunger then forced the foam out of the bag
with only negligible friction between the polyethylene arid Teflon sur-
faces. Using the modified technique, it was possible to exert a uniformly
reproducible pressure on the foam i ’t the hole. An additional advantage
achieved by this change in applicator design was that the rubber base of
a plunger was no longer in contact with the foam and therefore was not
bonded to it. The plunger then served as a piston to eject the foam
from the applicator, and the plunger cound be removed entirely after
application of the foam to the leak.
The need for controlling and adjusting the foam composition to obtain
better seals with a specific chemical was demonstrated in many of the
scaleup and preliminary development tests in which the Stafoam polyure-
thane system was used. This foam system consists of: (1) a hydroxyl-
containing polymer (“polyol”), (2) an isocyanate (toluene diisocyanate)
or a mixture of toluene diisocyanate and other polyisocyanates, and (3)
a low-boiling fluorocarbon. In addition, an amine (e.g., triethylamine)
is dissolved in the polyol and serves as a catalyst for the crosslinking
reaction between the polyol and the isocyanate component.
Structurally, the polyol is an aliphatic polyester with hydroxyl end
groups and,therefore,is susceptible to solution by polar chemicals (e.g.,
methanol). By contrast, the isocyanate is aromatic in nature and more
likely to be dissolved by benzene, toluene, or xylene. If a successful
seal is to be made against a specific chemical, such factors must be taken
into account.
The Stafoam components as provided by the supplier are recommended for
use at 85:100 isocyanate to polyol weight ratio. At this level, the seals
obtained against benzene leaks were fairly good. However, superior seals
were produced when the isocyanate component ratio was increased to 95:100.
This may seem somewhat surprising in that, with the aromatic nature and
the increased content of the isocyanate component, this composition should
be more susceptible to attack by benzene. However, this is more than com-
pensated for by the larger degree of crosslinking obtained at the high
isocyanate level. With this fonnulation, even better seals could be
obtained by speeding up the crosslinking reaction by adding 4 milli-
liters of triethylene to 195 milliliters of the combination. This is
advantageous in that the action of benzene on the composition is only
that of a solvent for the isocyanate component. By speeding up the cross-
linking reaction, the time available for the benzene to dissolve unreacted
isocyanate is markedly reduced. This results in a better foam.
33
-------
The sealing of a methanol leak is complicated by its chemical reactivity
with the isocyanate. To compensate for this, additional isocyanate was
added to bring the isocyanate:polyol ratio to 100:100. This provides
another advantage in that the increased quantity of isocyanate reduces
the solubility of the polyol in methanol. However, for successful seal-
ing of methanol leaks, it was found necessary to speed up the curing
reaction substantially. Thus, 7 to 8 milliliters of triethylamine were
added to every 200 milliliters of the foam combination.
Seals of the modified Stafoam combination held up in benzene for well over
24 hours with no evidence of deterioration. The modification developed
for methanol exhibited signs of attack by methanol after 8 to 12 hours
of exposure. Small-scale seals with the hazardous chemicals tested were
generally intact after at least several days.
To progress to the next steps in developing practical leak-sealing sys-
tems, it was necessary to obtain the foam components in pressurizable
containers. The polyurethane foam system that was used in most of the
scaleup tests was the Stafoain product of Expanded Rubber and Plastics
Company (ERP). This was the most suitable urethane of the foam systems
that had been screened. However, the components were available only in
bulk; therefore, it had been necessary to mix the components by hand in
an open container. Attempts were made to obtain these components in
pressurizable cylinders; however, this was not possible within the time
requirements of the program. After examination of the alternatives, it
was decided to obtain a froth polyurethane foam system (Autofroth I) from
Olin which was very similar to the improved foam which was recommended
by ERP.
With the components in pressurizable cylinders and with the use of a mix-
ing gun, it was possible to supply the foam directly into the leak-seal
applicator previously used. It was found that the foam properties and
leak-sealing performance of a given system were much better when used in
this mode than when mixed by hand. Figures 15 through 17 show urethane
components in pressurizable cylinders and a mixing gun built by Rocketdyne.
This gun was a working tool (not a development item) and continued to
evolve during its use, as described in the following three paragraphs.
The sequence in Fig. 17 illustrates the initial increase in foam volume
as it begins to cure.
A series of tests was made using Olin urethane systems and the Rocketdyne
mixing gun. In the first few tests, it was obvious that poor mixing of
the poiyol and isocyanate was obtained. An increase in cylinder pressure
(to increase the velocity of the impinging streams) plus a change in the
size of the gun’s “mixing element” did not noticeably improve the results.
In addition, partial plugging of the mixer element was occurring. Several
tests were conducted and one successful sealing test was made on a 2-inch-
diameter hole in a 5-gallon can filled with water. This was followed by
a nunber of unsuccessful attempts at sealing benzene. The first Olin
foam formulation tested (designated C-2) has a relatively long cure time
34
-------
Rocketdyne Mixing Gun
I .
Figure 16. Pressurizable Foam Cylinders, Mixing Gun,
and Polyethylene-Lined Plunger-In-
Cylinder Applicator
Figure 15. Pressurizable Foam Cylinders With
-
35
-------
(on the order of 20 to 30 seconds) before it can be used for sealing.
To shorten the cure time, a small quantity of amine was added and hand
mixed into the foam components after they had been placed into the
plunger-in-cylinder applicator. This reduced the cure time, but still
no completely successful seals were obtained. However, it became appar-
ent in these tests that the weight ratio of isocyanate to polyoi (which
should be nominally 1:1) was high and thus adversely affected the prop-
erties of the foam.
An attempt was made to control the weight ratio of the two foam components
(isocyanate and polyol) by installing a throttling valve between the iso-
cyanate cylinder valve and the mixer gun valve. Short-duration(2 to 3
seconds) flow checks were made to determine the ratio of the two foam
components. The flow checks were simple and consisted of simultaneously
flowing the two individual streams into two containers and weighing the
material collected. The technique of using a throttling valve to control
the isocyanate flowrate was not entirely satisfactory. Flow checks had
to be made every few runs and, in a number of cases, the ratio had changed.
A new pi ocedure and/or technique must be developed to maintain the desired
ratio of the foam components. In the leak-sealing tests with the throt-
tling valve installed in the isocyanate line, two successful seals were
obtained with each of the two test fluids (water and benzene). Seals
were made on a 2-inch-diameter hole in a 5-gallon container. In these
tests, no mixing element was used on the mixing gun. The foam components
were introduced directly into the plunger-in-cylinder applicator and mixed
with a spatula.
Tests were next made using a Kenics static mixer in the gun. This com-
ponent basically consists of a straight section of tubing containing a
series of metal ribbons, twisted about an axis coincident with the tube
centerline, and designed to repetitively subdivide and comingle segments
Figure 17. Use of Mixing Gun to Supply Foam
to Cylinder Applicator
36
-------
of the flow, thereby causing mixing as the fluids flow through the tube.
Also, a large in-line mixer (similar to the Kenics static mixer) was con-
structed, incorporated into the mixing gun, and used in two benzene seal-
ing tests. These tests were successful with respect to achieving adequate
mixing; however, additional work will be necessary to improve the control
of component mixture ratio.
The preliminary development test efforts just described comprised initial
steps in the development of an operational system. This work can provide
a foundation for a subsequent development program. The goal of a devel-
opment effort should be to produce a system that operates satisfactorily
anywhere within the envelope of practical application discussed earlier,
and embodies the following characteristics:
1. Ability to completely plug leaks
2. Permanent or long-term plugging
3. Rapid response
4. Moderate weight and portability (use by one man)
5. Safe and easy to use by untrained personnel
6. Reasonable cost
7. Usable with liquid heads as high as those encountered in tank
cars
8. Flexibility in hole size, shape, location, etc.
9. Usable with dry, dirty, and wet surfaces
10. Compatible with wide range of hazardous chemicals
11. Long shelf life
12. Wide temperature tolerance (both in storage and application)
13. No external power requirements (except possible use of com-
pressed gas)
14. No secondary pollution problems
15. Allows salvage of hazardous chemical remaining in tank
CHEMICALS WHICH CAN BE SEALED
Many of the other high-hazard chemicals are chemically similar to those
that were used in leak-sealing tests during this program. Tentative pre-
dictions were made of the expected behavior of the 100 most hazardous
soluble chemicals (Ref. 1 and 2) toward polymeric sealants (particularly
urethane foams). These predictions are summarized in Table 3. Where it
is indicated that successful sealing is not probable, the decision was
based on one or more of the following:
1. The compounds have excessively high vapor pressures at ambient
temperature.
37
-------
TABLE 3. PROBABLE LEAK SEALING SUCCESS WITH
100 HAZARDOUS CHEMICALS
Hazard Ranking. Ref. 2
(1 is aost hazardous)
Hazardous Chemical
Can Probably
be Sealed
Can Probably
Not be Sealed
Prediction Not
Yet Possible
1 Phenol X
2 Hethyl Alcohol X
3 Insectides and Rodenticides Cyclic X
4 Acrylonitrile X
S Chiorosulfonic Acid X
6 Benzene X
7 A onia X
8 Misc. Cyclic Insecticides X
9 Phosphorus Pentasulfide X
10 Styrene X
ii Acetone Cyanohydrin X
12 chlorine X
13 Nomyl Phenol X
14 DOT X
15 Isoprene X
16 Xylenes X
17 Nitrophenol X
18 Aidrin-Toxaphene Group X
19 A oniu Nitrate X
20 Aluminum Sulfate X
21 Nitric Acid X
22 Herbicides and Plant Horisones, X
Cyclic
23 Dyes, Total X
24 Tetraethyl Lead
25 A onia Sulfate X
26 Fungicides, Total, Cyclic X
27 Sulfuric Acid X
2 na ogenal.ec tlyarocarovns
29 Pncspnc.rus, ked A
29 Phosphorus, White X
30 2. 4-D Acid Esters and Salts X
31 Benzoic Acid X
32 Foraalde’ yde X X See text)
33 2, 4-0 Acid X
34 Sodiun Dichroir.ate and Chroeate X
35 Pesticides and Insecticides. Acyclic X
36 Tetranetnyl Lead X
37 Ethers, Total X
38 Ferrous Sulfate X
39 Sodium Sulfide X
40 Hydrochloric Acid X
41 Nickel Compounds X
42 Bentaldehyde X
43 Hydrogen Cyanide X
44 &ityl Alcohol. N- and Iso- X
45 chlorinated Isocyanurates X
46 Calcium Fluoride X
47 Hexaisethylenedianine I
48 Fatty Acids I
49 Pyridine I
50 Lead Coiiipounds I
51 Naphthalene I
52 Carbon Disulfide X
53 Hypochlnites X
54 Calcium Hypocalorite
55 Copper Sulfate I
56 Sodiua Hydroxide
57 Acids, Acylbalides and Anhydrides
58 Dodecyl Mercaptan I
59 Phosphoric Acid
60 Nitrobenzene X
I
38
-------
TABLE 3. (Concluded)
I
Hazard
(1 is
Ranking. Ref. 2
most hazardous)
Hazardous Chemical
Can
be
Probably
Sealed
Can Probably
Not be Sealed
Prediction Not
Yet Possible
61
Methyl Parathion
X
62
Alcohols, tonohydrie. IMsubstituted
N
63
Chromic Acid
X
64
Aldehydes and Ketone
X
65
Misc. Cyclic Herbicides
X
66
Fluorine Hydrofluoric Acid
X
67
Misc. Acyclic Insecticides
A
68
Potassium Iodide
X
69
Sodium Carbonate
X
70
Mines, Total
X
71
Aniline
X
72
Aluminum Fluoride
A
73
Munonia Compounds
X
74
Carbon Tetrachloride
A
75
Furfural
A
76
Lindane
X
77
Sulfur Dioxide
A
78
Asusonium Perchlorate
A
79
Mercury Cumpounds
A
80
Acetonitrile
A
81
Toluene
X
82
Trimethy lamine
x
83
Mercury Fungicide
A
84
Lead As-senate
X
85
Ethyl aenzene
86
Perchioric , cid
87
Methy1 s, ,i,,p
“
88
89
Pentachlorophenol
Sodium Hydrosuifite
A
A
90
Acetaldehyce
A
91
Ajumoniun Chloride
A
92
Ethylenediamine
A
93
Acetic Acid
A
94
Calcium Carbide
X
95
Barium Carbonate
X
96
Cycloheay lamirte
A
97
Silver Nitrate
A
98
Arsenic Compounds
A
99
Ethyl Alcohol
A
100
Acetone
X
39
-------
2. The chemicals react rapidly with liquid water or moisture or
other components in air.
3. The compounds are strong acids or bases, or strong oxidizing
reducing agents and may react with the components of the
sealants.
The predictions for phosphorus and formaldehyde depend on the state of
these chemicals. Red phosphorus probably can be sealed, while white
phosphorus probably cannot be sealed. Formaldehyde as a monomeric gas
has a normal boiling of -21 C, and would probably not be sealable because
of its vapor pressure. In aqueous solution (37 percent) as formalin or
in the polymeric forms of meta- or para-formaldehyde, it probably could
be sealed by present methods.
PRELIMINARY DESIGN CONCEPTS
A nuz er of preliminary design concepts were generated for consideration
in application systems. These concepts are concerned with the broad
range of leak situations likely to be encountered in practice. These
include leaks in air and underwater, and different configurations such
as round and elongated holes, cracks, open seams, leaking valves, and
open pipes. When leaks occur underwater, the hydrostatic external pres-
sure acts to reduce the net pressure behind the leak and makes leak seal-
ing easier, compared to air.
Several concepts and ideas are described briefly in the remainder of this
section. It should be noted that these are intended only to illustrate
possibilities for using foams to seal various types of leaks. The empha-
sis is primarily on the applicator element; each concept would make use
of similar systems (also to be developed in the future) for storing,
feeding, and mixing the foam. Two concepts are illustrated for use in
plugging irregularly shaped holes: the foam-filled balloon (Fig. 18) and
the hollow rubber cone (Fig. 19). An approach to using a moving appli-
cator head for plugging split seams and elongated openings is illustrated
in Fig. 20; a variation of the hollow rubber cone idea might also be con-
sidered. The idea of using a foam cocoon or foam dome for sealing leaks
from damaged valves or piping is illustrated in Fig. 21.
In the foam-filled balloon applicator (Fig. 18), an expandable rubber
tube (or balloon) is attached securely to the end of a foam supply sys-
tem. The end of the delivery tube with the balloon is pushed through
the opening in the leaking container. The premixed foam components are
injected into the balloon through the tube (both through the end and
through selected circumferential holes). As the foam expands, the foam-
filled balloon expands on both sides of the hole to form a mechanical
bridge, and this securely holds the plug in place. The final foam den-
sity probably is not at all critical, i.e., a broad range of densities
would be substantially equivalent in performance. Some advantages of
this system are its simplicity and that the foam during its curing
period is protected by the rubber tube, therefore eliminating any
40
-------
PRESSUR I ZED
FOAM CYLINDERS
FOAM
SUPPLY
STEP I. TUBE WITH BALLOON INSERTED
STEP 3. FULLY EXPANDED AND CURED
BALLOON HELD SECURELY Iti
FOAM-Ft LIED
HOLE
Figure 18. Foam-Filled Balloon Concept
CHEMI CAL
TANK WALL
QUICK-
D I SC ONNE CT
I I QU I D
CHEMICAL
STATIC
CONTROL MIXER
DEVICE
EXPENDABLE
TUBE SECTION
THROUGH HOLE
RUBBER
TUBE
— — -I
-—-
STEP 2.. BALLOON, FILLED WITH EXPANDING FOAM,
CENTERED IN HOLE
41
-------
QUICK-
DISCONNECT
CHEMICAL
TANK WALL
tIQUID
CIIEMI CAL
PRESSURI ZED
FOAM CYLINDERS
FOAM
SUPPLY
STEP 1. RUBBER CONE, .JUST FILLED WITH FOAM,
IS BEING INSERTED INTO HOLE
FOAM
STEP 3. LEAK STOPPED; CURED FOAM AND SLIGHTLY
EXPANDED CONE HOLDS CONE IN POSITION
Figure 19. Hollow Rubber Cone Concept
CONTROL
DEVICE
STAT I C
MIXER
HOLLOW RUBBER
CONE WITH SLITS
L J
STEP 2. WHILE FOAM IS EXPANDING, CONE IS
HELD IN PLACE, STOPPING BULK FLOW
42
-------
SPL$T SEAM
ELONGATED OPENINGS
Figure 20.
FOAM
FOAM
— __ 4 LEAK
Moving Applicator Concepts
LLAA UEt VAL.V
Th
LEAK -
Figure 21. Foam Cocoon Concept for Leaking Valves and Open Pipes
PLASTIC REINFORCING
BAG (TIGHT FIT
NOT NECESSARY)
ELEVAT ION
END VIEW
FOAM FOAM
43
-------
possible reactions with the leaking chemical, and much of the erosive force
of the fluid stream. A disadvantage is that the balloon prevents the foam
from adhering to leaking container walls. Therefore, although bulk flow
will be stopped, some seepage may exist and this will require additional
application of foam for a complete seal. Although the balloon might be
punctured by a jagged edge, the foam probably will have developed enou b
strength to avoid any serious leakage. The foam components could be in-
cluded in the tube and pressurized with air or nitrogen to drive them
through a static mixer and then into the balloon.
In the hollow rubber cone concept (Fig. 19) a semirigid rubber cone, with
slits in it, is connected by its plastic or metal base to a tube or some
other foam supply arrangement. It is filled with the mixed foam compon-
ents, and then it is positioned or jammed into the hole in the chemical
container. As in the preceeding concept, the rubber cone acts as a pro-
tective barrier to the foam as it is expanding and curing. The slight
flexibility of rubber cone allows it to conform to some extent to the
shape of an irregular opening in the chemical container. As the foam
expands, the excess material comes out of the slits and fills in the
remaining gaps between the cone and container opening. As opposed to
the balloon concept, the foam can adhere to the container walls, thus
forming a good leak-proof seal. With a cone of the proper rigidity, as
the foam expands, the portion of the cone inside the container will ex-
pand slightly, helping to lock it in place. The foam would probably
adhere to the rubber plug, if allowed to cure while in contact. This is
desirable in the case of sealing a hole, but undesirable if this type of
device were used to seal an elongated opening or seam. Each operational
application system would probably need an assortment of different size
cones; the choice of sizes would be based on a study of the tank failure
modes most prevalent.
In both of the above concepts, the balloon or cone plug can be positioned
on a long tube; this will afford the operator some protection from the
leaking chemical. The tube can be used to transfer the foam components
from sn l1 pressurized cylinders to the plugging device or applicator.
The transfer tube and mixing tube can be discarded after use. The cost
of a static mixing tube when produced in large quantities would be
reasonable.
When the opening is long and narrow, as in split seams and elongated holes,
the basic approach would be to use a moving applicator, as illustrated in
Fig. 20. The applicator is rigid and would have various application heads
contoured to follow a corner, a curved surface, or a plane surface. Be-
cause the openings are narrow, it is expected to be possible to seal by
spraying foam directly into an opening; the application starts at one end
of an opening and progresses to the other end.
A foam cocoon concept (Fig. 21) is envisaged for leaking valves or open
(broken) pipes. A dome or reinforcing bag, with cutouts for piping where
necessary, restricts the leak path and makes it possible for the foam to
surround the leak completely (i.e., by forming a cocoon).
44
-------
The preliminary design concepts presented illustrate some possible modes
of using foams to seal various types of leaks in chemical containers.
In developing a prototype system, the concept being considered would be
broken down into its component parts (e.g., such as the applicator, foam
supply system, etc.); then, each component part would require testing and
development as necessary until it approaches or meets the desired
requirements.
45
-------
SECTION VI
ACKNOWLE DGEMENTS
The support of the project by the Environmental Protection Agency and the
interest and involvement exhibited by the Project Officer, Ira Wilder,
are acknowledged with sincere thanks.
This project was conducted in the Advanced Programs Department at Rocket-
dyne with Dr. B. L. Tuffly, as Program Manager, responsible for overall
administration, and Dr. R. C Mitchell, as Project Engineer, responsible
for the technical content and conduct of the program. The other members
of the project team were Drs. C. L. Hamerinesh, M. Kirsch, J. E. Sinor,
and Messrs. J. V. Lecce and J. J. Vrolyk. (This report has been assigned
a Rocketdyne control number of R-9054.)
47
-------
SECTION VII
REFERENCES
1. Wilder, Ira and J. Lafornara, “Control of Hazardous Materials Spills
in the Water Environment: An Overview,” presented before the Division
of Water, Air and Waste Chemistry, American Chemical Society, Washing-
ton, D. C., September 1971.
2. Dawson, G. W., A. J. Shuckrow, and W. H. Swift, Control of Spillage
of Hazardous Polluting Substances , Battelle Memorial Institute, Rich-
land, Washington, FWQA Contract No. 14-12-866, November 1970.
3. R-8897, Methods to Control Hazardous Materials pi11s , 16 mm color!
silent film, Rocketdyne Division, Rockwell International, Canoga Park,
California, March 1972.
49
-------
SECTION VIII
PUBLICATIONS AND PATENTS
Mitchell, R. C., M. Kirsch, C. L. Hamermesh, and J. E. Sinor, “Methods
for Plugging Leaking Chemical Containers,” Proceedings of the 1972
National Conference on Control of Hazardous Material Spills , Houston,
Texas, 21-23 March 1972.
Several disclosures are being evaluated by the Rocketdyne Patents Depart-
ment to decide whether to file patent applications.
Si
-------
SECTION IX
GLOSSARY
A cross-sectional area of fluid stream
= cross-sectional area of container (in horizontal plane)
= cross-sectional area of rupture, for flow
C = discharge coefficient = C C
cv
C = contraction coefficient = area vena contracta
C area hole
C = velocity coefficient =
D diameter of cylindrical container
e = geometry ratio of container =
f = fractional liquid height after time T, i.e.,
£ = height after time T
original height
g = local gravitational constant
g = conversion factor necessary to make Newton’s Second Law
dimensionally correct when using ibf and ibm 32.174 lbin-ft
lb f-sec
H = length of cylindrical container
h = effective liquid head above rupture opening
p = local pressure
Q = volumetric flowrate
T = time required for liquid level to drop to some fraction (f) of H
u = local velocity (any effects of nonuniform velocity distributions
neglected for these purposes)
u local velocity for ideal fluid
2
V = volume of container = r D H/4
C
Z = elevation of specified station (referenced to any consistent datum)
p = fluid density
53
-------
APPENDIX
DERIVATION OF EQUATIONS FOR ENVELOPE OF PRACTICAL APPLICATIONS
Derivations are given in this section for the equations used in Section V
to define an approximate envelope of practical applicability of methods
for stopping leaks in chemical containers. The symbols are defined in
the Glossary.
BASIC CONSIDERATIONS
The region of application is considered to be limited by two factors:
1. Time for response (If the rupture is too large, there will not
be enough time to respond.)
2. Limits of human strength (The limiting constraint is taken as
the hydraulic force of the fluid stream, which will be trans-
mitted to the leak-sealing device, assumed to be portable and
operated by one man.)
THE SYSTEM
A cylindrical tank of diameter D , height H, and volume V (= ir .D H/4) is
considered. An opening of characteristic diameter D is located at depth
h below the free surface of the liquid in the container, as shown in Fig.
22 for containers at different attitudes.
(a)
(b)
(c)
Figure 22..
LIQUID LEVEL AT ANY TIME
EFFECTIVE CENTER
OF RUPTURE
Sketch of Basic Container in Different Attitudes
55
-------
TIME FOR RESPONSE
If we apply the Mechanical Energy form of Beimoulli’s Theorem to a rup-
tured container between station I (free liquid surface) and station 2
(at the vena contracta of the stream gushing from the rupture) (see Fig.
22):
*2 *2
U 1 -U,,
(Z—Z) —&-+ + (4)
1 2 2 g p
where
Z 1 -Z 2 = h by definition
<< u and can be neglected
= (We are not considering pressurized tanks in this case.)
Equation 1 reduces to
= v’ j (5)
Equation 2 gives a simple expression for the velocity of an ideal fluid
discharging from a container (neglecting friction effects). In reality,
frictional effects will keep the actual velocity of the discharge stream
below the value given by Eq. 5. Introducing a velocity coefficient, C,
u 2 (6)
Applying the Equation of Continuity and introducing a contraction coeffi-
cient, C , (to account for the vena contracta effect), converts Eq. 6 into:
Q=u 2 A 2 =u 2 AhC=CCA/ i (7)
or
Q=CA \/ ji (8)
Now, we want to use this result to develop an expression for the time
required for the liquid level in the container to drop by a specific
amount
dh_ Q 9
c
56
-------
Q is a function of h (Eq. 8) and, in the general case, A will also be
a function of h (specific for a given container geometry and attitude)
tb hb A(h)
5 dt=- (10)
t h
a a
To proceed further, it is necessary to specify Ac(h). Taking the simple
case of a cylindrical tank with its axis vertical (Fig. 22(a)):
A = rr D 2 /4
Using this and Eq. 8 in Eq. 10, plus specific limits which assume the worst
case (shortest time available) of starting with the container full of
liquid and the rupture at the bottom (h = H),
T ff1 irD 2
f dt = - 4CAhV’
T (D) 2 (; )l/2 (1 - f 112 ) (11)
Introducing e = H/DC and V = m D H/4, and thereby eliminating H and Dc
yields
5/6 1/2 ________________
T = 2(i) (1 £ ) c D 2 e (2g) 2 (12)
T — 0.3049 (1 - f 1 ” 2 ) V 5 ’ 6 13
— C D 2 e ( a)
with T in seconds, V in ft 3 , and D in feet (C, e, and f are dimensionless).
If it is desired to express Eq. 13a with V in gallons and D in inches,
the constant is changed:
1/2 5/6
_ 8..209(l f )V
CDe
which is given as Eq. 1 in Section V of this report. Typical values for
a sharp-edged (not countoured) hole are C 0.62, C 0.98, C 0.61.
57
-------
If we take as a criterion of maximum allowable flow (minimum allowable
response time),
f = 0.70 in I = 120 seconds
Then Eq. lOb becomes
D = 0.1353 vS/1 2 /e1 /6 (V in gallons, D in inches) (14)
which is given as Eq. 2 in Section V of this report.
Numerical results for several values of e are:
D/V 5 ” 12
0.25 0.170
0.5 0.152
1.0 0.135
1.5 0.1265
2.0 0.121
3.0 0.113
4.0 0.107
This range of values of e covers essentially any practical situation,
e.g., e 2 to 4 would be typical of a truck or railway tanker upended
to sit vertically; e 1.5 to 2 is typical of chemical drums; and e. .0.25
to 0.5 would represent truck or railway tankers in a normal (horizontal
axis) orientation. Strictly speaking, one would need to change the A,,(h)
function in’Eq. lOand derive a correspondingly changed Eq. 14 for any tank
orientation other than a vertical cylinder (Fig. 22(a)), with slightly
changed numerical constants in the tabulation above.
Because of the 1/6 power of e in Eq. 14, D is relatively insensitive to
changes in e. For present purposes (to provide a perspective for think-
ing about sizes of ruptures to be plugged), one may use a single value
of e, e.g., e = 1.5, so that Eq. 14 becomes
D = 0.1265 V 5112 (15)
LIMIT OF HII4AN STRENGTH
The-force required to counteract the hydraulic pressure is
F=khp 1 — (16)
n gc
58
-------
If we take, for a representative case, h = H (i.e., a full tank), with
Ah = It D 2 /4, and V = it D H14 = it H 3 /4 e 2 , Eq. 16 becomes:
/ 2 l/3
F=!(4eV
4\ir/ gc
Cubing both sides and rearranging:
3
VD 6 = 16 F (17)
2 2 ( 1_\ 3
iT e \ gcj
with V in cu ft. D in feet, p in ibm/cu ft. and F in lbf. This equation
is given in Section V as Eq. 3.
COMPARISON OF FORCES REQUIRED TO OVERCOME HYDRAULIC
PRESSURE AND FLUID MOMENTUM
The force required to overcome the static pressure from a fluid height
h with a hole of area Ah is
F =p.LhA. (18)
p gc n
The maximum force required to overcome the momentum of a gushing fluid
stream escaping from a hole is obtained by assuming that: (1) the stream
is being deflected at the vena contracta (i.e., point of maximum velocity),
(2) the total momentum of the stream is imparted to the applicator. For
these assumptions, the impulse—momentum principle gives the following
equation for the force required to overcome the stream momentum:
Q p u 2
F = (19)
m gc
where u 2 is the mean velocity at the vena contracta. Eliminating u 2 and
Q by the use of Eq. 8 and 10 gives:
F =2C C 2 p -& -hA. (20)
m C v gc n
Comparison of Eq. 18 and 20 gives:
F
= 2C C 2 (21)
F cv
p
59
-------
Using typical values for a sharp-edged circular hole (C = 0.62, C = 0.98)
produces the result
F
l.20
p
which means that the force required to overcome the momentum of the liq-
uid stream could be as much as 20 percent greater than that necessary
to overcome the static pressure developed by blocking the same opening.
It may be possible to reduce m substantially by clever design of the
applicator head to cause appropriate deflection of the fluid stream as
the applicator is moved into position. For this reason, it was decided
to use the definable value of F as the requirement on human strength.
60
-------
SELECTED WATER 1. Rep fl No.1 2. sxoxi No
RESOURCES ABSTRACTS k
INPUT TRANSACTION FORM I
4 .. Feasibility of Plastic Foam Plugs for Sealing
Leaking Chemical Containers
. ;r ) Mitchell, R. C., Hamermesh, C. L, Lecce, J. V.,
Kirsch, M. and Sinor, J. E. Project No.
EPA 15090 HGW
Rocketdyne Division of Rockwell International Corporation j1 Contract/Grant N
Canoga Park, California EPA 68-01-0106
13 Type of R*poz a 4
Period Covered .
12 Sponsoring Orgawzat ion
Environmental Protection Agency report number,
EPA—R2--73—251, May 1973.
A program was conducted to evaluate the feasibility of methods for plugging leaks
in damaged chemical containers by application of suitable plastic barriers. Such a
system would be valuable in helping to prevent water pollution from spilled hazardous
chemicals.
A large number of candidate sealants were evaluated in laboratory screening tests,
including various urethane foams; polystyrene and polyvinyl acetate instant foams;
filled and unfilled epoxy systems; and polysulfide, butyl, neoprene, and silicone
rubber systems. The most promising results were obtained with the urethane foams.
Additional evaluation and scaleup tests were made, including sealing of leaks of many
different hazardous chemicals, application to leaks both under water and in air, and
sealing of leaks in 55-gallon containers.
The feasibility of this concept was demonstrated. As a consequence -of the success
already realized, it is probable that a practical and useful system, en4odying this
approach, can be developed. (Mitchell - Rocketdyne)
17a. Descriptors
*Water Pollution Control, *Sealants, *Chemicals, *Leakage, *Accidents, *mansportation
l7b. Identif Jots
*plugging Chemical Leaks,
Pollution
*Hazardous Chemicals Spills, *Prevention of Water
2 ç. CO VVRR F :; & Gr .nip
05G
79. Si urityC ass. 21. .No.of Send To
(Report) Pages
20. Security ClassY. 22.. Price WATER RESOURCES SCIENTIFIC INFORMATION CENTER
(Page) WASHINGTON 0 C 20240
R. C. Mitchell I Rocketdyne Division of Rockwell
Aiii.. .t.-iia .iOna1 (. .L%JL1
------- |