EPA-600/2-77-151
August 1977
Environmental Protection Technology Series
AGENT AND
APPLICATiflit TO
>BOTECTION
Office of Research and Development
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U S Environmental
Protection Agency, have been grouped into nine series These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields
The nine series are
1 Environmental Health Effects Research
2 Environmental Protection Technology
3 Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6 Scientific and Technical Assessment Reports (STAR)
7 Interagency Energy-Environment Research and Development
8 "Special" Reports
9 Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental 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.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-77-151
August 1977
MULTIPURPOSE GELLING AGENT AND ITS APPLICATION
TO SPILLED HAZARDOUS MATERIALS
by
J. G. Michalovic
C. K. Akers
R. E. Baler
R. J. Pilie
Calspan Corporation
Buffalo, New York 14221
Contract No. 68-03-2093
Project Officer
Joseph P. Lafornara
Oil and Hazardous Materials Spills Branch
Industrial Environmental Research Laboratory-Cincinnati
Edison, New Jersey 08817
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
LIBIU RY
U S EWO-.' -,.V, PROTECTION AGENCY
\J, O- L-i 4 '' i .\v ,,>»- ,..
-------
DISCLAIMER
This report has been reviewed by the Industrial Environmental Research
Laboratory-Cincinnati, 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 endorse-
ment or recommendation for use.
ii
-------
FOREWORD
When energy and material resources are extracted, processed, converted,
and used, the related pollutional impacts on our environment and even on our
health often require that new and increasingly more efficient pollution
control methods be used. The Industrial Environmental Research Laboratory-
Cincinnati (lERL-Gi) assists in developing and demonstrating new and improved
methodologies that will meet these needs both efficiently and economically.
This report is a product of the above efforts. It documents the labor-
atory and field studies conducted to determine the optimum formulation, and
dispersal means for a multipurpose gelling agent for immobilizing hazardous
material spills on land. As such it serves as a reference to those in
state, local and Federal Agencies and the transportation and chemical indus-
tries, and others who are interested in the control of spills of hazardous
materials. This project is part of a continuing program of the Oil and Ha-
zardous Materials Spills Branch, lERL-Ci to assess and mitigate the environ-
mental impact of pollution from hazardous material spills.
David G. Stephan
Director
Industrial Environmental Research Laboratory
Cincinnati
iii
-------
ABSTRACT
Under contract to the U. S. Environmental Protection Agency, a blend of
materials was formulated that would spontaneously gel a wide variety of
hazardous liquids. This blend, known as the Multipurpose Gelling Agent (MGA),
has been optimized to obtain a balanced formulation that will effectively gel
and immobilize most spilled hazardous liquids within minutes. The current
formulation, consisting of four powdered polymers and one inorganic powder,
has the ability to immobilize spilled liquids with the least amount of mate-
rial in the shortest period of time, but evaluations of alternate materials
are still in progress. In field testing of the blend, it was determined that
air conveyance modes of dispersal could be employed but with high losses due
to the effects of wind on the powdered form. Three compressed and granulated
forms of the gelling agent were then developed which are clearly superior to
the original powdery blend for delivery to liquid spill targets. Various
off-the-shelf dry solid dispersion devices were evaluated and the most prom-
ising systems field tested on simulated and actual spill targets, both in
pools and in linear ditches. The results show that MGA provides a potentially
cost-effective and efficient means to mitigate the damages from hazardous
liquid spills.
This report was submitted in fulfillment of Contract No. 68-03-2093 by
Calspan Corporation under the sponsorship of the U.S. Environmental Protec-
tion Agency. This report covers a period from 28 June 1974 to 28 August 1975,
and work with technical effort was completed as of 28 July 1975.
IV
-------
CONTENTS
Foreword iii
Abstract iv
Figures vi
Tables vii
I. Introduction 1
II. Summary, Conclusions and Recommendations 2
Summary and Conclusions 2
Recommendations 3
III. Laboratory Optimization of the Multipurpose Gelling Agent (MGA) . 4
Background 4
Gelling Agent Blend Optimization 6
Cost of MGA 11
Substitute Gelling Agent Materials 13
Storage Requirements 16
Preliminary Bioassay 17
Procurement and Physical Specifications 17
IV. Physical Form of Gelling Agent 20
Background 20
Optimization of MGA Physical Form 22
Forms Tested 22
Summary of Test Results 22
Detailed Test Results 24
Powder form 24
Agglomerate form 24
Tablet form 24
Graded form 27
Webbed form 27
Mat form 29
Roll compressed form 29
Spun polyolefin bag evaluation 30
V. Dispersal of the MGA 31
Background 31
Dissemination Systems Evaluations . . . . 33
Pressurized Tank. ........ .... <> . <, . . 33
Venturi/Compressed Air 34
Centrifugal Blower 36
Auger-Fed/Pneumatic Conveyor 36
Nozzle Configuration 40
Mass Distributions of Powdered MGA0 40
VI. Discussion . . . . 45
v
-------
FIGURES
Number Page
IV-1 Effect of wind on powdered form of MGA being dispersed .... 21
IV-2a "Graded" form of MGA 28
IV-2b "Graded" form of MGA (SEM, 20 KV, 40° tilt, SOX mag.) .... 28
V-la Venturi tube - compressed air nozzle 35
V-lb Diagram of Airplaco Econo-Vac nozzle 35
V-2 Venturi compressed air system blowing powdered form MGA ... 37
V-3 Long tube on Venturi-compressor 38
V-3a Dispersing agglomerated form 38
V-3b Gelled trichloroethane 38
V-4 Centrifugal blower (snowblower) 39
V-4a Experimental apparatus 39
V-4b Blowing powdered MGA 39
V-5 MGA dispersed from pneumatic-conveyor/auger-fed system
Powdered MGA being dispersed from 30-meter
(100-foot) hose . 41
V-6 Equal mass distribution contour map obtained from CC^-charged
dry chemical fire extinguisher dispersion of MGA 42
V-7 Mass distribution of MGA dispersed from CC^-charged dry
chemical fire extinguisher 44
VI
-------
TABLES
Number Page
III-l Compounds Tested with MGA 5
III-2 Minimum Amount of Gelling Agent Required to Form an
Immobile Gel (g/10 ml) 7
III-3 Effectiveness of Gelling Agent Blends 9
III-4 Formulation of MGA Blends 10
III-5 Amount of Blend Required for Gelling Alcohols 10
II1-6 Hydroxypropyl Cellulose [Klucel (Hercules)]
Gelling Qualities (g/10 ml) 11
III-7 Cost Per Pound (Per Kilogram) of Gelling Agent Blends C, D, E. 12
III-8 Potential Substitute Materials Considered 14
III-9 Physical Properties of MGA Components 18
111-10 Classification of MGA Components 19
IV-1 Test Liquids 23
IV-2 Efficiency of Physical Forms of MGA 25
IV-3 Gelling Efficiency of Powder MGA (Blends D § E) 26
V-l Dispersion Devices 32
VI1
-------
SECTION I
INTRODUCTION
The purpose of this investigation was to develop materials and equipment
which could be used to gel and, thereby, immobilize a wide variety of liquid
chemical compounds that may be spilled to the detriment of the environment.
The ideal countermeasure to any spill of a hazardous material is to
terminate the spill as soon as possible by sealing the material in its orig-
inal container. Once hazardous chemicals are on the ground, the most attrac-
tive secondary countermeasure consists of immobilization of the chemical to
minimize the affected land area, to prevent flow of hazardous liquids to sur-
face water, and to minimize percolation of liquids into the surface and sub-
soils. Immobilization of a spilled liquid with a gelling agent ideally leaves
the material in a form that can be easily removed by mechanical means.
Under an earlier EPA sponsored program on "Methods to Treat, Control and
Monitor Spilled Hazardous Materials" (Contract No. 68-01-0110)* a preliminary
blend of materials was formulated that was useful for gelling both organic
and inorganic liquids. The mixture was tested on 35 hazardous chemical com-
pounds that were successfully gelled. This original blend, because of its
ability to thicken and gel all hazardous liquids tested, was referred to as
the "Multipurpose Gelling Agent" (MGA).
The first objective of this study was to determine the gelling agent
formulation which would provide an optimum balance of gelling speed, ease of
application and product cost. Generic components and suitable generic sub-
stitutes were identified as far as possible. The requirements for storage
were identified and are reported herein. Consideration was also given to the
physical form of the agent to make it more amenable to effective delivery to
an interaction with a spilled liquid. Various physical forms of the MGA were
evaluated with respect to gelling efficiency and ease of dispersion.
The second objective was to evaluate various types of dry dissemination
devices that would be suitable for dispensing the gelling agent in a spill
situation. Consideration was given only to "off-the-shelf" equipment. The
types of devices showing the most promising principles of operation were
field tested on simulated spill targets to define dispersion, delivery capac-
ity, rate of delivery, portability, and unit cost.
*
This work was carried out under contract to the Environmental Protection
Agency, Edison, N.J. from 29 June 1971 to 27 November 1972.
-------
SECTION II
SUMMARY, CONCLUSIONS § RECOMMENDATIONS
SUMMARY $ CONCLUSIONS
The MGA developed under EPA Contract Number 68-01-0110, "Methods to
Treat, Control and Monitor Spilled Hazardous Materials," was optimized to
obtain a balanced formulation that would effectively gel a large variety of
spilled hazardous liquids.
A formulation, Blend D, consisting of a polyacrylamide, a polytertiary-
butylstyrene, a polyacrylonitrile rubber, a polycarboxymethyl cellulose, and
a fumed silica was determined to be the optimum gelling agent based on its
ability to immobilize the greatest variety of hazardous spill liquids with
the least amount of material.
Also evaluated were alternate materials that could possibly be substi-
tuted for components of the MGA. Commercial sources for each of the ingred-
ients were identified and preliminary generic descriptions obtained.
It was determined through early field testing that the powder form of
the MGA could be deployed using air conveyance modes of dispersal but only
with high losses due to effects of wind. Three compressed and granulated
forms of the MGA were developed which are clearly superior to the original
powdery blend for field dispersal onto liquid spill targets.
Various off-the-shelf dry solid dispersal devices were evaluated with
the most promising dispersal modes field tested. Auger-fed, pneumatic-
conveyor systems ("Rockduster") originally developed for the coal mining
industry were shown to be best suited for distribution of MGA at high rates
and acceptable accuracy. The portable "sandblaster" was also found to be an
acceptable delivery device.
Delivery nozzles bent at 30° to the horizontal plane of the dispersing
tube provide better accuracy and agent distribution than straight-pipe
nozzles.
Using 1975 quotations for raw materials, preparation and packaging, it
was estimated that large quantities of MGA may be prepared in granular form
for between $4.40 and $5.50 per kilogram ($2.00 and $2.50 per pound).
The MGA must be stored in moisture-proof containers at temperatures less
than 50°C (122°F).
-------
A ten-minute film was produced showing the various tested dispersion
devices, various physical forms of the agent, and the gelling of a test
chemical.
RECOMMENDATIONS
The MGA is now sufficiently well-defined to be manufactured and stored
in reasonably large batches. It is recommended that approximately 1350 kilo-
grams (3000 pounds) of this material be procured and utilized in large-scale
field experiments to more completely demonstrate its efficacy.
Smaller scale tests should be run with MGA forms produced by a variety
of manufacturing techniques so that the optimum physical characteristics for
both dispersal and gelling efficiency may be defined. Commercially available
substitute ingredients should be incorporated into small MGA batches, averag-
ing 45 kilograms (100 pounds) or so, to guarantee their suitability for free
substitution as market and/or material conditions dictate.
Auger-fed, pneumatic-conveyor systems should be obtained and field-
tested, followed by redesign, if necessary, for MGA distribution to large
spill targets.
Ultimate clean-up measures to be applied following successful use of the
MGA should be experimentally determined, and the prospects for economical
recovery of the spilled liquid from the containing gel mass should be in-
vestigated.
-------
SECTION III
LABORATORY OPTIMIZATION OF THE MGA
BACKGROUND
Under a completed EPA sponsored research program on "Methods to Treat,
Control, and Monitor Spilled Hazardous Materials" (Contract No. 68-01-0110),
Calspan developed and evaluated a mixture of several dry chemical materials
for immobilizing spills of hazardous liquid chemical substances. The hazard-
ous liquid chemicals considered in that program are listed in Table III-l.
They were selected for study on the basis of toxicity, volumes shipped, and
previous histories of spills.
There were a number of criteria used in the development of the MGA. One
was that the gelling agent should be able to immobilize a broad spectrum of
organic and inorganic liquids. The development of many specific agents would
create a perhaps insurmountable logistics problem. A second criterion was
that the presence of, or interaction of, the treatment agent should not create
a serious secondary hazard, such as increased danger of fire or explosion.
The MGA developed represented a new composition of matter, created by
the mechanical blending of equal weights of five specific components having
the following properties.
The first component was a material of the highly water-soluble poly-
electrolyte-type, typified by polyacrylamide. This material was selected
for its ability to thicken water and aqueous solutions.
The second component of the blend was a loosely crosslinked copolymer
of the class typified by polytertiary-butylstyrene copolymerized with
divinylbenzene. This material was selected to interact most strongly with
those liquids having almost no polarity and only poor solvent power such as
cyclohexane, gasoline fractions, and a variety of other inert spirits.
A third component was a material of the polyacrylonitrile-butadiene
copolymer class chosen to be especially effective against the very strongly
polar organic chemicals such as acrylonitrile, ethylene dichloride, and
other chlorinated or polar liquids.
The fourth required component of the blend was a material to cope with
the most difficult of all hazardous liquids to thicken, solidify, and immobi-
lize in place, as typified by methyl alcohol and other chemicals of the
alcoholic class. Materials suitable for this use included the polycarboxy-
methylcellulose polymers.
-------
TABLE 111-1. COMPOUNDS TESTED WITH THE MGA
ACETONE
ACETONE CYANOHYDRIN
ACRYLONITRILE
AMMONIUM HYDROXIDE
ANILINE
BENZALDEHYDE
BENZENE
BUTANOL
CARBON DISULFIDE
CARBON TETRACHLORIDE
CHLORINE WATER (SATURATED)
CHLOROFORM
CYCLOHEXANE
CYCLOHEXANONE
ETHANOL
ETHYL ACETATE
ETHYLENE DICHLORIDE
ETHYLENE GLYCOL
FORMALDEHYDE
GASOLINE
ISOPRENE
ISOPROPYL ALCOHOL
KEROSENE
METHANOL
METHYL ETHYL KETONE
OCTANE (2,2,4 TRIMETHYL PENTANE)
ORTHO-DICHLOROBENZENE
PETROLEUM ETHER
PHENOL (89%)
PYRIDINE
SULFURIC ACID
TETRAHYDROFURAN
TRICHLOROETHYLENE
WATER
XYLENE
-------
This four-component blend uses the commercial products with the follow-
ing trade names:
(1) Gelgard, Dow Chemical Corporation, to combat spills
of aqueous liquids;
(2) Imbiber Beads, Dow Chemical Corporation, to combat
spills of the inert spirits-type liquids;
(3) Hycar 1422, BF Goodrich Corporation, to combat the
polar organic chemical spills including the chlor-
inated hydrocarbons; and
(4) Carbopol 934, BF Goodrich Corporation, to combat
spills of alcoholic liquids.
For ease of delivery of this four-polymer blend, it required fluidization
to ensure rapid, smooth egress from commercial spray equipment. Addition of
fumed silica (e.g., Cabosil, Cabot Corporation) was used.
GELLING AGENT BLEND OPTIMIZATION
The four powdered polymers, each capable of congealing at least one
class of hazardous liquid into an immobile mass, were originally selected
and combined in equal parts by weight and fluidized with fumed silica to
form MGA.
To achieve an optimum blend, the proper proportioning of the above com-
ponents had to be pursued. A number of laboratory tests were performed to
obtain a formulation which would minimize material cost without sacrificing
the effectiveness against a broad spectrum of spills. Each test liquid was
treated with individual gelling agent components to ascertain the least
practical amount of material required to congeal specific fluids into immobile
gels. Also, a mixture of equal amounts (20% by weight) of each polymer and
fumed silica was added to each of the test liquids.
The amount of each polymer and of the mixture necessary to gel a liquid
was determined by adding small amounts of polymer to 10 ml of liquid in a 3-
inch aluminum pan. When the aluminum pan could be inverted without loss of
the gelled fluid the exact weight of polymer added was recorded. Table III-2
lists the experimental results.
It is apparent from Table III-2 that in all cases the mass of the blend
required to congeal the test liquids is less than five times the mass of the
specific active agent required to produce the same results. The ingredients
of the blend did not act independently. In all cases, a degree of synergism
exists which is associated, at least in part, with physical "sopping" of the
test liquid by nonactive components of the blend.
-------
TABLE 111-2. MINIMUM AMOUNT OF GELLING
AN IMMOBILE GEL (g/10ml)
10 ml
Acetone
Acetone Cyanohydrin
Acrylonitrile
Ammonium Hydroxide
Aniline
Benzaldehyde
Benzene
Butanol
Carbon Disulfide
Carbon Tetrachloride
Chlorine Water
Chloroform
Cyclohexane
Cyclohexanone
Ethanol
Ethyl Acetate
Ethylene Dichloride
Ethylene Glycol
Formaldehyde
Gasoline
Isoprene
Isopropyl Alcohol
Kerosene
Methanol
Methyl Ethyl Ketone
Octane
0-Dichlorobenzene
Petroleum Ether
Phenol
Pyridine
Sulfuric Acid
Tetrahydrofuran
Trichloroethylene
Water
Xylene
"M"
Gelgard
(g)
#
11. 6#
5.47
0.22
#
#
#
#
7.56
#
0.14
#
#
#
#
#
#
4.32
0.58
#
#
#
#
#
#
#
#
3.74
#
#
1.22
#
#
0.07
#
Imbiber Beads
(g)
#
#
4.21
#
#
2.95
0.58
#
0.79
0.42
#
0.47
0.95
2.52
#
0.89
1.10
#
#
0.74
0.42
#
2.37
#
0.84
4.05
0.74
0.58
#
1.58
#
0.68
0.63
#
0.74
AGENT REQUIRED TO FORM
1422 Hycar
(2% Cabosil)
(g)
0.58
1.23
0.58
#
0.72
0.54
0.54
#
2.50
#
#
0.33
#
0.54
#
0.72
0.43
#
#
3.62
1.52
i'
,/-
6.11
#
0.54
#
0.51
2.17
0.54
0.51
1.45
0.47
0.47
#
1.09
Carbopol
(g)
1.38
2.64
1.43
#
#
#
#
0.74
#
1.41
#
#
#
#
0.52
#
#
0.67
0.74
#
#
1.11
#
0.57
#
1.61
#
ff
1.80
#
#
0.94
#
#
#
Blend A*
(g)
1.15
2.11
1.25
0.32
1.47
1.57
0.59
2.47
1.23
1.20
0.47
0.76
1.13
1.35
0.98
1.05
1.05
0.78
0.52
1.27
0.83
2.38
2.15
1.03
1.13
1.89
1.08
1.03
1.20
1.03
0.98
0.76
0.81
0.12
1.18
* 20% of each of: Gelgard,
Carbopol, and Cabosil.
# No gel formed.
Imbiber Beads, Hycar
Cabosil),
-------
With these data available, the relative proportions of the five ingredi-
ents were modified in subsequent mixtures in an attempt to produce a blend
which could be used to treat all of the test chemicals with the same dosage.
An arbitrary goal of one gram or less of the gelling agent per 10 ml of liquid
was established.
Based on inspection of the data presented in Table III-2, Blends B and C
were prepared and tested with the same 35 representative chemical compounds
that had been used earlier. Table III-3 compares Blend B and C with the
original, equally proportioned blend (Blend A). Formulations of these blends
are listed in Table III-4.
From Table III-3 it is apparent that Blend B is inferior to Blend A, in
terms of mass agent required, for 19 of the 28 cases in which both were tested.
Blend C, however, is superior to Blend A in this respect for 25 of the 35
comparisons made. Furthermore, where C is not clearly superior to A, the
difference in mass requirements is, in each case, relatively minor.
If the amount of gelling agent from each blend tested is summed and
divided by the number of measurements, a numerical comparison can be made
among Blends A, B and C. The average amount of Blend A to gel 10 ml of liquid
is 1.15g. The amounts of Blends B and C are 1.48 and 0.94g, respectively,
This clearly shows the effectiveness of Blend C over Blends A and B.
Note that in most cases the arbitrarily established goal of l.Og of
agent per 10 ml of test liquid was achieved in Blend C. There are, however,
some chemical compounds, such as acetone cyanohydrin, isopropyl alcohol,
butanol, and octane, which required approximately twice that dosage rate for
gelation. These deviations from the goal suggested that further mixture
adjustments be made to enhance the gelling efficiency for alcohols. The
Carbopol concentration in Blend C was, therefore, increased, and the Hycar
concentration decreased to form Blend D. Results obtained with alcohols are
presented in Table III-5. Within the accuracy of the experiments, tests of
each of the other three gelling agent ingredients indicated no change in
dosage requirement relative to Blend C.
Blend D was, therefore, selected as the principal mixture to be used in
experiments pertaining to control of physical characteristics of the gelling
agent and testing of dissemination equipment. Orders were placed for suffi-
cient quantities to perform these experiments. Subsequently experiments were
performed on substitute gelling agent materials which led to Blend E described
below.
A substitute material which has potential for the immobilization of
alcohols, the most difficult of all hazardous liquids to gel, was evaluated.
Klucel is a hydroxypropyl cellulose, nonionic, water-soluble cellulose ether.
Klucels M and H both have the same structure, with M having a higher molecu-
lar weight. Table III-6 lists data which show these materials to be superior
to Carbopol 934 in gelling the low-molecular-weight alcohols. In addition,
the gels formed with the Klucel were more elastic and possessed greater
coherence.
-------
TABLE 111-3. EFFECTIVENESS OF GELLING AGENT
(g/10ml OF CHEMICAL REQUIRED FOR
Chemical
(10 ml)
Acetone
Acetone Cyanohydrin
Acrylonitrile
Ammonium Hydroxide
Aniline
Benzaldehyde
Benzene
Butanol
Carbon Disulfide
Carbon Tetrachloride
Chlorine Water
Chloroform
Cyclohexane
Cyclohexanone
Ethanol
Ethyl Acetate
Ethylene Bichloride
Ethylene Glycol
Formaldehyde
Gasoline
Isoprene
Isopropyl Alcohol
Kerosene
Methanol
Methyl Ethyl Ketone
Octane
0-Dichlorobenzene
Petroleum Ether
Phenol
Pyridine
Sulfuric Acid
Tetrahydrofuran
Trichloroethylene
Water
Xylene
BLENDS A, B,
GELATION)
ANDC
Blend
A
(g)
1.15
2.11
1.25
0.32
1.47
1.57
0.59
2.47
1.23
1.20
0.47
0.76
1.13
1.35
0.98
1.05
1.05
0.78
0.52
1.27
0.83
2.38
2.15
1.03
1.13
1.89
1.08
1.03
1.20
1.03
0.98
0.76
0.81
0.12
1.18
B
(g)
1.26
3.05
1.75
0.35
1.44
1.23
1.12
4.59
1.33
1.30
0.35
0.60
1.26
1.19
1.75
1.23
0.95
1.82
0.77
1.40
1.30
4.24
1.86
1.65
0.88
*
*
*
1.30
*
*
*
*
0.25
1.12
C
(8)
0.95
2.23
1.14
0.50
0.95
0.82
0.56
2.60
0.58
0.74
0.61
0.37
0.93
0.74
1.33
0.66
0.53
1.03
0.69
0.90
0.74
2.15
1.59
1.14
0.72
1.99
0.69
0.69
0.80
0.61
0.98
0.37
0.50
0.48
0.69
*Not tested.
-------
TABLE
Blend Gelgard
A 20
B 15
C 5
D 5
E 5
111-4. FORMULATION OF MGA
Imbiber Beads
20
30
30
30
35
Hycar
20
35
35
30
20
BLENDS (%
Carbopol
20
15
20
25
*
BY WEIGHT)
Cabosil Klucel H
20 *
5 *
10 *
10 *
10 30
*Not used.
TABLE MI-5. AMOUNT OF BLEND REQUIRED TO GEL ALCOHOLS
Liquid (10 ml)
Methanol
Ethanol
Isopropanol
1 - Butanol
BLEND
(g)
C
D
1.14
1.33
2.15
2.60
0.92
1.17
1.47
2.21
10
-------
TABLE 111-6. GELLING QUALITIES OF HYDROXYPROPYL CELLULOSE (KLUCEL M
AND KLUCEL H) AND CARBOPOL 934 (g/10ml)
Acetone
Acetone cyanohydrin
Acrylonitrile
Butanol
Ethanol
Ethylene glycol
Isopropyl alcohol
Methanol
Phenol (89%)
Klucel M
0.81
1.39
1.15
1.02
0.95
2.27
1.02
0.88
1.29
Klucel H
0.66
1.25
1.22
0.99
0.79
2.27
1.02
0.79
1.18
Carbopol 934
1.38
2.64
1.43
0.74
0.52
0.67
1.11
0.57
1.80
Klucel H was incorporated into Blend D as a substitute for Carbopol
later in this study (Blend E). This blend was tested for dosage requirements
and gelling speed. As indicated in Table IV-3, Blend E is slightly superior
to Blend D. The overall advantages, however, were not considered to be so
important as to warrant repetition of packaging and dissemination experiments
performed with Blend D.
COST OF MGA
A preliminary cost analysis of the MGA is based upon the costs of order-
ing and producing 450-kg (1000-lb) batches of agent. The cost per pound will
decrease as the quantity of material ordered increases. This reduction in
cost, however, is not expected to exceed 15 to 20%. The cost of Blends C, D
(standard MGA formulation), and E are given in Table III-7. Costs given re-
flect only the price per pound of the base materials. The cost for blending,
forming, and packaging is not included in Table III-7. For a 450-kg (1000-lb)
batch, the cost for blending is $0.55/kg ($0.25/lb) and packaging $0.22/kg
($0.10/lb). The cost for preparation of a special form (i.e., pressing,
rolling, chipping) is not included because no such processing has yet been
done with large lots. These costs, however, would not be expected to increase
the price above $0.55/kg ($0.25/lb). Final costs of the gelling agent would
be expected to be in the $4.40 to $5.50/kg ($2.00 to $2.50/lb) range. These
costs reflect 1975 economic conditions and are substantially higher than our
1973 estimates of $1.10/kg ($0.50/lb). If current economic trends continue,
the cost per pound of gelling agent is expected to increase further.
11
-------
LLJ
Q"
C/3
O
LU
CQ
1-
LLJ
^i
0
3
LU
C3
LL
0
S
oc
a
0
5
DC
LU
Q.
Q
Z
D
O
a.
cc
UJ
Q.
CO
o
o
^
UJ
CQ
^
H
w
13
0>
CQ
Q
13
g
CQ
CJ
13
c
a>
CQ
-&*5-
, *
,0 bO
i-H Ai
\ /
o\°
**
/ \
**tB
0\°
<&r
f ^
^ o bO
r-H ^
V31 fc?
0\°
LOLO LOO OO OO OO
CM LO Tj- tO CN CN LO .-H vOI^
o CN .-i 01 -st T-H i OLO r~-oo
i i CM ooo ^nto i r^Lo ^Hio
OO r-Hr-H OO Oi-H OO
/~^ I N f ^ f V f N
LOrH r^i-H r-H\O LO^ vQf-s
OLO ooto I-^LO i to i i r--oo
(N'st (NVD O.-H tNJLO rH tO
LO O O * O O
tO (N tO t 1
LOLO OO OO OO OO
CNLO i (to IOOO LOLO MDt^
orsi \OCT) ^Hvo CNt^ i r~ oo
r-H(M OOOO CslTf \OtO 1 1 tO
OO OT-H OO O.-H OO
\^__J \^_J \^__J *~_J *-_^
/ \ / \ / s t * / \
LOr-i r^-i t i < \o oo or^
OLO oo^o r^Lo LOLO i r^oo
CN-st (N1\D OrH CN1LO i 1 tO
>. / V 1 V 1 \ 1 V '
LO O O LO * O
tO tO CN i-H
LOLO OO LOO OO OO
(NLO i-HtO OOvD OO \Or-~
OCNJ vOOl ^t'st OO 1 t-^OO
i-H(N] OOOO CNJLO LO r- 1 1 r-HtO
OC5 Or-H C3O O i 1 OO
V. ' V 1 \ / \ / \ /
t \ t \ / \ / \ 1 N
LO r i r^r-i i i vo oo \Dr~»
OLO ooto t^LO LOLO i r--oo
CN^t CN^O Oi-H CN1LO i ItO
LO O LO O -it O
in
13 ^r
a to
m CN
<* r-H K
^1 0) PH r-H -H
03 rQ rH O fl> f)
DJ) -H rt X) O O
I-H ^3 o ?H 3 ,a
i) £ X rt !-H rt
f n . . >-i t r T "^s F i
U HH ,J_ t^J H^H V >
O O
O 00
^ O
"~1 ^
CN -st
^ /
LO LO
h^ O
r-» LO
01 to
r-H "st
o ^o
00 «O
00 LO
00 r-H
r-H ^T
v '
MH
0
/~~>
M
,^
^
^ 1
13
r> pj
r-H 0)
\ i 1
be- CQ
"8
M
O
12
-------
It should be noted that since the start of this work, the manufacture of
Gelgard M has been discontinued by the Dow Chemical Corporation. Small
supplies of Gelgard are still available, and Dow will resume manufacture if
an order is sufficiently large. Substitute material has also been found that
works similarly to Gelgard. This material (Kelzan) is discussed with other
potential substitute materials in the next section.
SUBSTITUTE GELLING AGENT MATERIALS
In order to broaden the scope and universality of the gelling agent, a
search for substitute gelling agent materials was performed. Manufacturers
of gelling and thickening compounds were contacted and products obtained for
laboratory evaluation (Table III-8).
The materials were evaluated against the following criteria:
(1) - Material must be a dry powder with low moisture content.
(2) - Material must form viscous gels without extreme stirring
or heating.
(3) - Material gelling efficiency must compare directly with
current MGA components.
Each of the potential substitute materials was screened using' these
representative liquids from the four major chemical classes:
Class I - liquids (aqueous solutions) Water
Class II - liquids (nonpolar organics) Cyclohexane/Isooctane
Class III - liquids (polar organics) Trichloroethane
Class IV - liquids (alcohols) Methanol
Specific Product Evaluations
Pluronic F-127 (BASF Wyandotte Corporation)--The literature on Pluronic
F-127(polyoxyethylene and polyoxypropylene) states that two techniques have
been developed for preparing Pluronic gels -- one based on cooling a hot
solution; the other on warming a cool solution. Neither of these techniques
is applicable to our specific problem.
When Pluronic F-127 was added to various liquids (at ambient tempera-
ture) no thickening occurred. This product was rejected as a candidate
material.
Alcogum L-ll (Alco Chemical Corporation)--This thickener is supplied in
liquid form.It was rejected as a substitute material.
Stearates (Witco Chemical Organics Division)--Stearate gels are formed
by heating a solution of stearates plus the solvent to be thickened and then
13
-------
LLJ
CC
LU
Q
CO
z
0
O
CO
<
E
LU
r"
<^
IE
LU
(
1-
fe
CO
D
CO
_J
<
z
LU
|_
O
Q-
00
^~
WM
LU
_J
CO
^
1-
hH
1 1
1 I
t/1 h- 1
»-/ ^"~*
1 1
t I
CD
E
CTJ
Z
o
H
CD
i
CD
g
TO
1
(Tj
f-l
H
X
ff
rt
PH
g
0
u
X X X X X X
x x x x * x
x x x x x
XXX*XX XX XXX
0
H
-d
X
,c .
CTJ *~*
o
o &*
rH C
E ^
t, CD
f1 CD
Cll "
Jii f_i CD (D W
73 O 00 ^H O
aJ . e o x -^
0 ^ X^ PH , 3
^ f- ^H 3 O O i
g 2 ^ ""* F^1 4-inJO CD
'^ "2 CD O X fn -H nj
H .H i CD»_/(D CD'-'-l-' 0(D X
"TCxX t-'XPi E 4->OJO 4J4-1 P^
XP^°'C' rtr-(CD 3 (/)£L,O nJrt O
|-)rH 'rvl "x0*^ °° cee'C'i n
CD x g x ^ ^ c 333 ^ ^ x
^E.S^ ° -HO ClJOO r-H^H^H OO X
£ (g m a: ^ s "-^ xzz <<< o.o< ac
^
-i I-H rt O
--H O *
^ 1 (x t/1 CTi
< CD "-J
NO O i 1
N .H E .H -H O
CDXW^c: C^3 i < PL,--I
JH OO MO n*'H(D'~H OOOCD
4-1 ^rH 8 ^ r^'Sc.u t-^'Sa
l> IS < s: ^-- £ ^5 3
CD _,
13 -S
_o " E tr^
IH S <" ±1
rrt O ^fj *~O I/)
U J5 U 0 CD
U O >-i
.-, O t-n O O 3
0 0 tu 000 0
Bs .H OC/3 1-1 O +J PL, fn
<3 rH< CDX'H CD
^5 o
f^ h (/)
to n3 o ii
3 1-1 O
O O fc X
CD & nj o
3 C r-t O
a' O o ii
< z cx <:
I
I
i i
14
-------
cooling the mixture in a stepwise manner. This product was rejected as a
candidate material.
Gantrez-AN (GAP, Dyestuff and Chemical Division)--Four Gantrez-AN samples
were received for evaluation (Gantrez-AN, 119, 139, 149, 169). The literature
stated that a slurry of Gantrez-AN should first be prepared at 95-99°C. After
10 to 20 minutes of high speed stirring, the slurry would become viscous.
When Gantrez-AN was evaluated using our laboratory procedure, no thickening
occurred. Therefore, the product was rejected as a substitute material.
PVP (GAP, Dyestuff and Chemical Division)--Polyvinylpyrolidone did not
perform well as a thickening agent under our evaluation procedure. It was
rejected as a substitute material.
Soloid (Kelco Company)--The commercial development data sheet states that
"Soloid is a new solvent thickener that is soluble in most aromatic, aliphatic
and chlorinated hydrocarbon solvents. Soloid is not soluble in methylene
chloride."
Soloid was found not to wet with Class I liquids even when vigorously
stirred. The material floated on the surface. Class II and Class III
liquids gelled extremely quickly. Soloid did not thicken Class IV liquids
but tended to swell approximately three times in size.
Soloid was compared with Imbiber beads in gelling isooctane (a Class II
liquid). Although the Soloid compound formed a gel very fast, the Soloid
powder tended to float on the surface. Mixing was needed to get the powder
into the liquid mass, with a noticeable adverse "flour-bag" effect (gelling
the exterior surface of masses of agent and rendering the inner material
unavailable for reaction). It was found that approximately four times the
amount of Imbiber beads compared to Soloid was required to gel isooctane.
Soloid is, therefore, a candidate substitute material for the MGA com-
ponents for both Class II and Class III liquids.
Kelzan (Kelco Company)Kelzan is a xanthan gum-polysaccharide which is
a thickening agent for Class I liquids. The data sheet obtained shows 1%
aqueous solutions to have high viscosities. Kelzan was compared with Gelgard
M. When Kelzan was added to water, the powder spread immediately over the
surface and formed a surface-gel of sufficient strength to float additional
Kelzan on top of the surface. Upon stirring, the powder would agglomerate
around the stirrer, producing a large "flour-bag" effect. Approximately
twice the volume of Kelzan compared to Gelgard, was needed to gel equal
volumes of water. However, the "flour-bag" effect was so great that less
than this amount of Kelzan was used in the gelling effort. This product is a
potential candidate substitute component in the MGA.
Klucel H and M (Hercules)--Klucel was evaluated as a substitute material
for Class IV liquids. A direct comparison was made between Klucel and
Carbopol 934. It took approximately twice the amount of Carbopol 934 compared
to Klucel to get methanol. Thus, Klucel is a potential substitute component
for the MGA.
IS
-------
STORAGE REQUIREMENTS
Samples of the powdered form of the MGA, Blend D, were tested to find
the limitations that are imposed upon storage. Containers of MGA were stored
from 0 - 100% humidity and from -20 to 103°C (4 to 217°F). These samples were
examined weekly for two months and less frequently thereafter.
The MGA stored at room temperature in an open container (relative humid-
ity 50 - 80%) formed a lumpy mass while MGA stored at 100% relative humidity
formed a solid spongy mass.
The powdered MGA when stored at high temperature [103°C (217°F)], changed
color from white to deep tan in one month, indicating a possible degradation
of the organic polymers. Tests on the partially degraded material showed the
gelling effectiveness was decreased but not eliminated.
Samples were also stored for six months outdoors, exposed to the elements
in containers to evaluate the type of container and the stability of the agent
over wide temperature ranges. Fiber drums (Fibre-Pak) manufactured by the
Continental Can Company were filled with MGA and placed outdoors in an open
unprotected area where they were subjected to the natural environment. The
weather over the test period, February through July 1975, was typical of
Western New York State. A summary of the general weather over the test period
is listed below:
general weather variable (sunny, rainy, snowy)
temperature -13.9°C (7°F) to 31.7°C (89°F)
relative humidity 20 to 100%
winds calm to 35.8 m/sec (80 mph)
The MGA powder stored showed no indication of any deterioration and there
was no change in quality when compared to laboratory stored samples. The
storage container was not damaged by the weather although there was super-
ficial rusting of the metal parts.
Requirements for storage based on our laboratory evaluation are that the
MGA must be stored at temperatures less than 50°C (122°F) and be free from
moisture.
16
-------
PRELIMINARY BIOASSAY
A bioassay screening test was performed to determine the effect of
various concentrations of MGA on creek minnows. This particular test was
performed to evaluate the consequences of large amounts of MGA washing into
a stream or lake. Fish were exposed for 72 hours to 0, 0.01, 0.05, 0.1, 0.5
and 1% MGA (by weight) in water. A 1% solution results in a viscous solution
in which the fish could barely swim. The maximum safe concentration was found
to be 0.01%. The acute effects of gelling agent were due mainly to the
thickening solution which caused suffocation and gelled the fish into the
surface layer. After 24-48 hours, MGA precipitated out of solution and re-
mained at the bottom with no effect on surviving fish. Fish exposed to 0.05%
MGA solution lost equilibrium from 30 minutes to 24 hours but recovered. The
results presented, however, are only preliminary indicators of the toxicity
of the MGA upon the water quality when it is used in spill control.
PROCUREMENT AND PHYSICAL SPECIFICATIONS
The physical specifications of the components of the MGA as they were
received from their respective manufacturers are listed in Table III-9. Some
manufacturers stated that certain information, such as molecular weight dis-
tribution, was not available or was proprietary information. The manufactur-
ing specifications, as to type of processing or the type quality control,
were classed as proprietary. The generic name of each material is given in
Table 111-10, but this merely serves as a guide to chemical class since these
are polymeric materials and actual manufacturing specifications would have to
be used to reproduce the material. Cabosil was the only material for which
the generic name and manufacturing specifications would be obtained. This
material, a product of the Cabot Corporation, is a fumed silicon dioxide.
17
-------
V)
z
LU
z
o
a.
O
O
^f
O
^>
LU
0
CO
1 1 1
CC
LU
QL
O
CC
Q.
-1
^t
O
CO
^--
I
a.
ai
_!_
^~
«>
LU
00
^l
h~
£_f
0
d
o
CD
N
H
CO
CJ
C
cti
rH
Cti
CD
PH
Q-4
^
X co
cti cti
CD -P
X
P
O -H
PH 03
CO rH
CD
rH
P / ^
CO 0\°
H '
O
S
0
H
CD
CO
m1
i/>
co
cd
rH
CJ
S
0
U
CD
g
cd
C
CD
'"O
cd
rH
CU
(5
0
fi
E
3
0)
V
^
CD
-a
Z
0
PH
*
rf"I
S
X
X
CD
C
CO -H
3 fH
0 0
CD rH
3 to x
cr-d cj
cd .H
£3 bO f -,
rH CT O
CD -H CD CM
CD rH ^ I
CD
13
H
g
Cti
rH
X
J_J
o
cti
X
i i
0
ex
"O
cd
bo
rH
CD
CD
CD
C
0
C
^>
O
o
LO
CD
CD
r^
PH
CO
r^
*
X
K5
O
i l
*
1
O
P rH
rH CJ
CD X/ ,
C -P O CD
rH &C Cti
rH .H X
CD PH CD CD
CD CO > ' X
13 CD
CD p!
^ CD
C H
H X
rH H->
1 CO
tO rH
co X
O r^
rH i ^
CJ cd
co
cd
0)
03
rH
CD
Pi
rH
X>
6
rH
^
rC
bO
H
rH
co
E
,j
LO
4_)
CO
^3
"0
1^
s
o e
rH p_,
v &
i i
^j-
i i
CM
.
bO
CD
CO rH
i. O
0 C
rC Cd
o x:
rH CJ +J
(1) r I CD
CD Cti S
1
X
rH
O CD
PH £-
?^ CD
X rH
0 X
Q ^f^
rH P
Cti CD
CJ S
S
cn
1-1 %.
0 \
PH ^
O
J3
cti
CJ
P
C
CD
bO
C
3
6
H
oX0 O
o o
00 LO
V
CD
rd
3
0
PH
C
cti
H
00
C7}
O
uo
V
f
bO r-,
CO
CD T3 C
^-^ CD O
rH ^-> XI
Cti CO cti rH
O -r-l .H O
PH C rH 0
Cti O rH
rH bO rH 13
co x;
io i co
O 13 S ft rH
rH CD 3 -H CD
O r^! -H rH X>
CD C 13 P XI
rH -H CD -H 3
0. rH B C rH
CN
CM
i (
rH
cd
CJ
X
K
CD
r^
O
C
g
a
^^>
i i
t^^
co
CD
^
cti
( i
H-t
,-fj
?
co
O rH
^ e p o3
x/ PH CO P
V PH O CD
E E
CD
^ cti
rH rH
3 ^
XI
rj
CM N1
CM O
O O
t-O
CD
P
d
cti CD
CD C4_|
N X
-------
tf\
\L
^~
z
in
o
Q.
^£
o
Q
h"
LLJ
^n
0
Z
_J
LU
O
1
^_
^y
^_
^J
II
O
z
o
1
O
TT"
^^
CO
CO
_J
o
CJ
"
_L
IV
LU
^J
CO
^c
1-
S_i
£-H
t-H
rJ
I-H
f 1
^D
W
W
3
u
I-J
u
rH
s
w
u
Oi
tu
os
r^
t-*
U
tL,
3
Z
3
W
s
^
z
w
Q
H
t/1
r-H / s
03 S
43 0
3 4-1
O «
Ul rH
4)
rH P
0) o!
4-1 5
03
? 4)
C
X: -H
4J rH
H O
U rH
r-;
4-1 O
O
oS -
rH .
41 M
4)
4-1
i 1
O
£-1
o
4)
i 1
4)
X
i i
O
A.
4> 13
i 1 -H
"3 oj
rH rH
0 X
1 O
rH 03
4) X
OS O
3: PL,
C
rt
rH W)
4) -H
1) _J-*
rH C U
03 4) -H
o u S
H
E 2 >
O O 13
43 Q C
CJ nS O
O rH Tf
S r\i 13 vo
SO -H OO
CM S ^t"
13
iH
Pj
00
T )
o
a
r <
* 4-*
M
bfl ^
. C
d>
O 4) t/>
03 43 P
rH 0 -H
4) rH rH
P O -H
(3 X OH
rH O (/)
1
X
rH
O
P.
O 4>
0 (3
4)
0) N
C (3
4) 4)
X rH
4-1 X
w q
X-H
4-J >
3 -H
43 13
I
X43
rH 4J
OS -H
H S
4-1
4) 4)
4-> Is!
X-H
rH rH
O 4)
f\ , C
q
ni
?H bO
4) -H
4-1 43
r-l C O
03 4) -H
u u S
H
e s >
4) O T3
43 Q C
U nS O
0 rH Tt
S CM 13 \O
O O -H OO
Q CM S rt
rH
4)
43 t/1
H 13
43 oS
E 4)
rH CQ
r\i
*
U 1)
H 13
C -H
03 rH
fciO O
rH rH
O 43
O
rH -H
03 TJ
rS
O 4)
a. q
Xr^
rH X
M43
C 4-J
O 4)
4-1 *
in
bfl ' *
43 . 4)
4-1 0 rH
H v -* «i 1
^ rH
t/1 4-1
+-) f 1 -f-l
0 o! C
03 0 O
SH -H rH
a> 6 X
4-1 4) rH
C 43 O
rH O 03
4)
q
4)
H
13
03
4-1
3
CO
1
4)
rH rH
H 4)
rH 43
4J 43
H 3
q rH
0
rH rH
X 4)
S X
a) «-H
X 0
rH p !
o o
a, o
13
XrH
c ca o
43 0) -H
U P, 4) 43
H 6 4) O
rH 0 fH
13 U H "
0 13
o rH ^ q
U 03 OS o3
O O rH
-H 0) rH
tt, e o > to
4) O 4) rH
43 rH rH Tj-
CQ U \O U TH-
rH CM
OS CM
O Tt
XrH
W
"
O
c
03
43
4J
4)
e
*.
fl
&0
0)
>,
I/I
I 1
0
43
0
O
1 1
OS
43
P
H
4-1 <v
U rH
OS O
rH q
4) o!
4-1 4-1
q 3
rH 43
rH
4)
X
1 1
o
p.
4)
q
4)
1 1
X
43
4->
4)
X
rH
O
I 1
X
X
o
43
rH
03
U
-o
XrH
q ca o
43 03 -H
O PL, 0) 43
H E 4) O
rH O rH
13 U H -
O 13
o rH ^ q
t_2 OS nJ oS
O O rH
-H 4) rH
UH S O > tO
4) O 4) rH
« 43 , 1 t-H ^~
ca u \£) u *
, i
o
P.
o
43 TJ~
JH to
OS CTl
U
^;
c
0)
51
H
t/1
13
q
oS
4)
j_^
3
4-1
X
H
E
4)
£>
0
43
03
0) t/1
43 ^
4-1 Oj
E
4)
N 13
H OJ
13 rH
H rH
3 aj
it bfi
<4-l
O 43
H 4-1
03
O
H
i 1
H
in
*O
4)
tL,
q
o
H 4-1
P 4)
03 0)
rH rH 1/1
O 4-J [/)
rH ^
O 43
t_J M -
H C
4-j as o
O 4J
43 LO (/I
Oi CM O
U rH CO
rH
H
l/l
O
43
03
U
LO'
o
rH
rH
CM
0
19
-------
SECTION IV
PHYSICAL FORM OF GELLING AGENT
BACKGROUND
In the early stages of development the MGA consisted of a fine powdery
mixture of components described previously. That consistency was selected
to provide maximum surface area and thereby achieve maximum speed of reaction
with spilled liquids. Early field experiments revealed some disadvantages of
the powder form. The most serious was the drift of the fine powder from the
target area even under conditions of light to moderate winds (Figure IV-1).
In addition, the first particulates to contact the spilled liquid reacted so
quickly that a thin surface film of gel was formed in upper layers which
prevented penetration of additional agent into lower layers that had not been
affected. Unless sufficient turbulence existed in the stream to provide the
necessary mixing of agent with the spilled liquid, flow continued at the
lower levels.
It was apparent from these experiments that the optimum consistency of
the gelling agent would be one that preserved the large amount of surface
area to promote rapid and complete reaction, and yet contained particles of
sufficient size to be relatively unaffected by wind and sink into the volume
of the spilled liquid rather than react only with the surface layers.
Various techniques for pelletizing and forming loose aggregates of the powder
were investigated in an attempt to achieve these results.
It was recognized early in this investigation, which was performed con-
currently with investigations of dispersal techniques and equipment, that the
utility of a given gelling agent form was in some cases dependent on the dis-
persal technique used. Since different dispersal equipment may be used for
different spills, depending on the size, nature and location of the spill,
the objective of this investigation was to select a material form which was
usable with the widest variety of potential equipment and which was effective
under static conditions without artificial mixing.
20
-------
Figure IV-1. Effect of wind on powdered form of MGA being dispersed
21
-------
OPTIMIZATION OF MGA PHYSICAL FORM
Forms Tested
The material forms included:
(1) The original fine powder;
(2) The powder agglomerated by a water spray;
(3) A variety of tablets;
(4) Crushed tablets that were sieve graded into two mesh ranges;
(5) Webbed, dustless powder;
(6) A mat formed from the webbed powder; and
(7) A roll compressed form chipped into proper sizes.
The standard test procedure used consisted of gently pouring l.Og of
MGA into a vial containing 10 ml of test liquid (the 10:1 gelling ratio) and
determining the time required for complete gelation. The criterion for
gelation was that no flow should be observed when the vial was gently tipped.
When gelation was not observed in this test the experiment was repeated with
a doubling of MGA until the liquid was gelled (i.e., 5:1, 2.5:1, 1.25:1
gelling ratio).
Each of the MGA forms was tested against three test liquids from each of
the four chemical classes (Table IV-1).
Summary of Test Results
The results of these experiments may be summarized as follows. The
"roll compressed" form, produced by chipping the solid layer resulting from
roller-mill compression of the powder, and the two "graded" samples, produced
by grinding MGA tablets and subsequent sieving for size distribution control,
appear to represent the near optimum compromise in material form. These do
not react with spilled liquids as quickly as the original powder. However,
even under light wind conditions the quantity of MGA reaching the target area
may be increased by factors as great as five. These three forms behave well
in wind and in some liquids sink slowly through liquid rather than reacting
with surface layers only and appear to be useful with all dispersal equipment
with which they were tested. However, sufficiently large samples of these
forms were not available for testing with all types of equipment.
From the standpoint of ease of manufacture, it appears that the "roll
compressed" form has advantages over the other two. On the other hand, the
gelling efficiency of the graded forms is slightly superior. The "roll com-
pressed" form and the "graded" forms should be retained as candidate
materials for large-scale testing.
22
-------
TABLE IV-1. TEST LIQUIDS
CLASS I
SATURATED CHLORINE WATER
CONCENTRATED AMMONIUM HYDROXIDE
WATER
CLASS H
KEROSENE
BENZENE
CYCLOHEXANE
CLASS HI
O-DICHLOROBENZEIME
ETHYLENE DICHLORIDE
CARBON TETRACHLORIDE
CLASS 12
METHANOL
ETHYLENE GLYCOL
ISOPROPANOL
23
-------
The agglomerated form of material performed admirably in the laboratory
tests, but was not significantly better than powder in windy conditions.
Additional manufacturing costs (which are substantial for this form) are not
justified.
The only other acceptable MGA form tested in these experiments was the
original powder. Because of problems discussed earlier, however, it must be
considered a poor fourth choice.
Results obtained with the materials tested are summarized in Tables IV-2
and IV-3 and discussed briefly in the following paragraphs.
Detailed Test Results
Powder Form--The powder form was obtained by mixing the individual com-
ponents of the MGA according to the Blend D formulation (5% Gelgard M, 30%
Imbiber Beads, 30% Hycar 1422, 25% Carbopol 934, 10% Cabosil). This mixture
was most effective in laboratory tests and was used as the control formula-
tion for comparison of gelling efficiency of various forms of MGA.
In most cases, the powdered form gelled the test liquids in less than
10 minutes with a 10:1 gelling ratio (10 ml of liquid to 1.0 g of MGA).
However, the alcohols were not gelled satisfactorily. Kerosene, ethylene
glycol and isopropanol did not gel even after 60 minutes with the 10:1 gel-
ling ratio. Ratios of 5:1 and 2.5:1 were required to gel these materials.
From data presented previously, which were obtained with mixing, it should
be expected that these materials should have gelled with the 5:1 gelling
ratio. As previously, a large part of the problem in the static ex-
periment was that the initial portion of the powder to contact the test
liquid would interact with the liquid to form a protective shell. A "glob"
of agent would thus be formed, with the agent in the inner portion of the
aggregate remaining dry and not participating in the reaction.
Agglomerate Form--A small batch of MGA (Blend D) was sent to Valentine
Laboratories (Clifton, New Jersey) for agglomeration. Particles, 60 mesh in
size, were formed by agglomeration with water mist using a GLATT Agglomerator.
The bulk density of the agglomerated material was approximately twice that of
the powder.
In all but one case, the agglomerated form of the MGA was equal to or
better than the powdered form. All liquids gelled within the 10 minutes
design parameter, except isopropanol which did not gel. While this form has
several advantages over powder, its performance in wind was not a significant
improvement. The additional manufacturing expense is, therefore, not justi-
fied.
Tablet Form--A variety of tablets of different sizes and physical character-
istics was produced from the powder using a disk-making device and a hydrau-
lic press operated in several stops between 5000 and 20,000 psi. Tablet
configurations ranging from thin wafers (1-mm thick) to pellets 6 mm
(1/4 inch) in diameter and 6 mm (1/4 inch) thick were formed. Most were
24
-------
<
u
s
LL.
O
C/3
S
CC
0
LL
_l
<
O
C/3
>
I
a.
LL
O
>
o
Z
UJ
o
LU
UJ
csi
>
UJ
_l
CO
<
Q |
lu E
S«S
dSri
o| x
***
E
l- E
5"
m *
< UJ
H- X
E
CN
il
X
o
WEBBED
0 5 mm
RADE E
- 2 0 mm
w E
£°
<
K in
0 S
OMERATE
- 0 75 mm
CJ ^
0 0
<
* i
UJ UD
0 0
?U
°- §
o
TEST LIQUID
CLASS
. s g s
LL
; o o o
Z (N
* o r- r*
* O O O
1 SAT CHLORINE WATER
2. CONC, AMMONIUM HYDROXIDE
3 WATER
I WATER
SOLUBLE
(M
in § iS
LL
O
Z CM CM
^ O O
U CJ O
z z z
LO un in
LO LO
T CM O
r 6 r
LTJ r- ID
LO
0 N
r- O -
-~ o
LD <- in
LO
0 CM in
«- o <-
^ o o
LD in
CM CM
r- o o
o o o
LO
CM
CO O *J
^ c5 o
LO < »-
1 KEROSENE
2 BENZENE
3 CYCLOHEXANE
H. NON-POLAR
ORGANICS
LO
o o ^Z
CM CO uO
0 O 0
O O O
Z Z Z
in LO LO
r- ^ CM
o o o
LO ro
CM M
O 0 --
00-,
t- «- LO
in LO M
CM CM n
o o o
o o o
LO lO LO
fM CM CM
O 0 O
o o o
in LO
CM CM Ln
o o <-
o o o
1 O OICHLORO BENZENE
2 ETHYLENE DICHLORIDE
3 CARBON TETRACHLORIDE
m. POLAR
ORGANICS
C C in
in in
-------
TABLE IV-3. GELLING EFFICIENCY OF POWDER MGA (BLEND D & E)
CLASS
I. WATER
SOLUBLE
H. NON-POLAR
ORGANICS
m. POLAR
ORGANICS
EZ. ALCOHOLS
TEST LIQUID
1. SAT. CHLORINE WATER
2. CONC. AMMONIUM HYDROXIDE
3. WATER
1. KEROSENE
2. BENZENE
3. CYCLOHEXANE
1. O-DICHLORO BENZENE
2. ETHYLENE DICHLORIDE
3. CARBON TETRACHLORIDE
1. METHANOL
2. ETHYLENE GLYCOL
3. ISOPROPANOL
POWDER
BLEND D
* **
10/1 0.5
10/1 7
10/1 7
5/1 3
10/1 0.25
10/1 4
10/1 0.25
10/1 0.25
10/1 1.5
10/1 7
2.5/1 15
5/1 10
POWDER
BLENDE
10/1 1.5
10/1 1
10/1 .5
5/1 1.5
10/1 0.25
10/1 2.5
10/1 0.33
10/1 0.25
10/1 0.75
10/1 6
5/1 7
10/1 3
* GELLING RATIO
** MINUTES TO GEL
26
-------
mechanically stable, i.e., they could be bent or dropped into a hard
surface without breaking. However, in all cases the reaction rate of
tablets with test liquids was so slow that the concept of using a tablet
form of the MGA was abandoned. The experiments with tablet forming tech-
niques were important in that these methods were used as the first step
in manufacture of the graded forms of MGA which are very promising.
Graded FormLarge tablets [76 mm (3 inches) in diameter x 3 mm (1/8 inch)
thick] of Blend D formed at 6804.6 atm (10,000 psi) were ground into small
pieces using a laboratory blender. The ground tablets were then sieved into
two fractions (Grade I and Grade II). A photograph of the graded form is
shown in Figure IV-2.
Grade I contained a size distribution from 0.25 mm to 1.0 mm. Grade II
contained a size distribution of 1.0 mm to 2.0 mm. The particles larger than
2.0 mm were returned to the blender while the finer particles were recycled
to the press. The two sieved fractions (Grades I and II) were evaluated
separately.
Grade I of MGA was extremely effective in the static tests with 10:1
gelling ratio against all test liquids except kerosene, and isopropanol, for
which a 5:1 and 2.5:1 ratio was required respectively. It was also very
effective in subsequent field tests against simulated spills. In the latter
experiments there were some material losses due to wind but the effect was
minor.
Grade II was somewhat disappointing in the static tests, requiring at
least a 5:1 gelling ratio for immobilizing seven of the twelve test liquids.
In the field tests, however, it appeared to be as effective as Grade I
material against the test liquids. Furthermore, there were virtually no wind
losses.
On the basis of these experiments we recommend that both Grade I and
Grade II material be retained as candidate materials for large-scale field
testing.
Webbed Form--The Harshaw process (Harshaw Chemical Company) forms a web of
fibers within the powder mixture in order to form a "dustless" powder. The
process is as follows: 1% (by weight) Teflon K is added to the MGA powder.
The material is then mixed thoroughly with the blender and heated for 30
minutes at 103°C (217°F). Fiber formation is initiated by grinding the heated
mix with a mortar and pestle. This forms a dough-like material which is then
placed in a blender to break it into pieces less than 1 mm in size.
The webbed form of agent is about as effective as the graded forms of
the MGA in static tests, but was not very effective in windy conditions, so
its further investigation was abandoned.
27
-------
Figure IV-2a. Graded form of MGA.
Figure IV-2b. Graded form of MGA (SEM, 20KV, 45° tilt, SOX mag.)
28
-------
Mat Form-- The mat form was produced by rolling the dough-like mass from the
webbed procedure (above) into a thin sheet approximately 1.6 mm (1/16 inch
thick which was then cut into 6-mm (1/4-inch) squares.
Data showed that the mat form did not gel any of the test liquids in
less than 45 minutes and its investigation was, therefore, terminated.
Roll Compressed Form--An experimental "production run" of the rolled com-
pressed form of MGA was made by the Valentine Laboratory. A Fitzpatrick com-
pactor was used to produce a continuous sheet of compressed material approxi-
mately 1-mm thick which was then chipped into small pieces, 5 mm x 5 mm, with
a Fitzpatrick comminutor mill.
As shown in Table IV-2, static tests of this material produced gels in
five of the 12 test liquids using 10:1 gelling ratio, one with a 5:1 ratio,
five with a 2.5:1 ratio and one with a 1.25:1 ratio. These results are sub-
stantially inferior to those obtained with Grades I and II. Three reasons
for this decrease in performance can be given. First, the "rolled" form is
much more compressed than the tablets from which the graded forms were made.
(They are very rubbery in texture and cannot be broken easily or cut except
with a sharp knife.) Consequently, the porosity of the "roll compressed"
form is low and contact with the liquid is poor. Second, in manufacture of
both the "roll compressed" form and the tablets, the outer surface of the
MGA is very smooth and has few, if any, pores. In grinding the thick tablets
to produce graded forms a large amount of porous surface originally contained
within the tablet is produced. However, in chipping the "rolled" form into
5 x 5 x 1 mm chips only a small amount of inner surface is exposed, and
approximately half of the total exterior surface of the particles consists of
nonporous material that was in contact with the rollers. Again, a substan-
tial reduction in contact of MGA with liquid results. Third, as evidenced by
the differences in static performances of Grades I and II MGA ultimate
particle size distribution is important. The "roll compressed" material
tested consisted of substantially larger particles than either of the graded
forms.
Since there is no apparent fundamental difference in the processes used
for producing the rolled form and the tablets, it should be possible to manu-
facture "roll compressed" material which is as efficient as the"graded" forms.
This would require 1) reduction of the pressure applied in the roller mill;
2) increasing roller separation to produce a thicker sheet of MGA; and 3)
adjusting the comminutor mill to produce smaller particles. These changes in
process can be achieved. Unfortunately time restraints prevented another
production run during this program.
Because of potential economic advantages of producing the "roll com-
pressed" form of MGA compared with producing the "graded" forms, it should be
modified as suggested above and retained as a candidate material for large-
scale testing.
29
-------
Spun Polyolefin Bag Evaluation
Several advantages could be achieved if the MGA could be disseminated in
porous bags. Wind drift would be eliminated; retrieval of the MGA-spilled
liquid gel would be greatly simplified; manual dissemination for special
purposes, such as formation of a dike-perimeter to prevent flow of a spill,
could also be simplified.
To test the utility of porous bag packaging, spun polyolefin bags were
obtained from Gedcor Environmental Protection Corporation (Westland, Michigan)
and filled to half capacity with several of the MGA forms discussed in the
previous subsection. The bags were submerged for 10 minutes in one chemical
from each class of test liquids and removed for measurement of weight change
due to absorbed material. In all cases the quantity of liquid collected was
substantially smaller than that gelled in previous experiments with the same
agents freely disseminated. Upon examining the bag contents, it was found
that reaction of the test liquids with the outer layer of gelling agent had
formed a gel that protected the inner portions of the agent from contact with
the liquid.
In an attempt to eliminate this "flour-bag" effect, numerous wicking
agents (including paper pulp fibers, straw, vermiculite, sawdust and IBM
card chips) were mixed with the MGA before packaging, but no significant
improvement was observed. Only the "roll compressed" form could be completely
wetted and here the reaction rate, apparently limited by a partial "flour-bag"
effect, was greatly retarded.
As a result of these simple experiments, the concept of packaging MGA in
porous bags for dissemination was abandoned.
30
-------
SECTION V
DISPERSAL OF THE MGA
BACKGROUND
The Thomas Registry and the on-line computer search facilities at Calspan
were used to identify manufacturers of "off-the-shelf" dispersal devices for
powders. The search primarily produced agricultural equipment manufacturers.
The following companies were contacted for descriptive literature of their
equipment:
Buffalo Turbine Agriculture
Equipment Company
John Deere
FMC-Agriculture Machinery
Division
Vandermolen Corporation
Spraying Systems Company
Root-Lowell Corporation
R. E. Chapin Manufacturing
Works Incorporated
H. D. Hudson Manufacturing
Company
Standard Container Company
International Harvester
Gehl Company
Mine Safely Appliances
Company (MSA)
U.S. Government
Gowanda, New York
Syracuse, New York
Jonesboro, Arkansas
Livingston, New Jersey
Wheaton, Illinois
Lowell, Michigan
Batavia, New York
Chicago, Illinois
Montclair, New Jersey
Chicago, Illinois
West Bend, Wisconsin
Pittsburgh, Pennsylvania
Washington, D.C.
Table V-l lists the information available on the dispersal devices.
31
-------
CO
LU
O
>
Ul
Q
"Z.
O
CO
oc
UJ
o.
CO
Q
r^
UJ
_J
CQ
«
H
Z
UJ
U
1-
o
o
SuJ-
u. cc y
cc t
"Jet a.
°0~
is
Ul
UJ
rsj
(/)
_i
Ul
O
O
1
a.
|
UJ
~ yj (/) w
... 1- Z T CD O
i =3 3 O O 0
E a H to w t/)
UJ
Z
oa
< tc
o
i
D
CD
u, u.
0 0
1- t-
a. a.
0 0
UJ UJ
_l _j
U) IU
II
cc cc
a. a.
tfl CO
111 UJ
> 2 Z
aooooooo uj uj
JJJCCCCCCCCDCCCGC IZ
zooooooozzis
NU,U-U.U.U.U,U. _(£3(C|
i-'f-ai<-uit-mjnSS i i
1 '^^|5§§^§*
pooo
isll
in in
a> to
SR2£?SS^SRI1
1 1 1 1 1 i?
QOQOQ gg x
inr^rninlnbbV^^bu)
Q
0
Q
z
g
*
TY OVER
CAPACI
THING WITH DUST
O
Z
HUDSON
CO
Z 5
LB CAPACITY
o
§
V
<
o:
u
s
in
CHAPIN
WORKS
O
UJ u.
cc S
ILOWER
tLOWER
ADDLE FORAGE E
ADDLE FORAGE E
a. a.
V 10
3 S
ss
RNATIONA
VESTER
II
oc
i
o
3
Q
t_
§
i
100 FT WI1
A
z
c
n
N THROW SHELL C
<
U
8
,_'
§
S
o
u
UJ
0
1C
UJ
fc
&
Q
Z
15 LB/MIN.. 200 PO
o
o
S
LO
V
N
,_
0
X
S
'
M BEAN DIV
si
U. =1
UJ
z t
= ?
u. >
o t oc
» O UJ
CAPACIT
10LBCAI
NT DISPEI1
' ui
-1 00 y
20LBIN 19 SEC, 1
^ITANT GAS DISPE
NGE RIOTCONTH
252
_J
w S
S 0
H
o
o
c/l
3
32
-------
In general, liquid-carrier systems have replaced dry-chemical dispersal
systems. The agricultural blowers which direct powder via a nozzle or exit
tube are no longer in production. They have been replaced with blower
systems which disperse powder into the air forming a large dust cloud. The
equipment is used for agricultural orchard dusting.
There are two systems used in the mining industry that disperse lime
dust as a dry powder over the walls and floor of the mine shaft in order to
eliminate underground explosions. One system consists of a hopper loaded,
auger fed, pneumatic conveyor. This unit transports the powder, at approxi-
mately 3.6 to 4.5 kg (8 to 10 pounds) per minute, through a 51-mm (2-inch)
diameter hose for distances up to 50 meters (200 feet). The other system used
is a pressurized tank which holds up to 2.7 long tons (3 short tons) [2.5
cubic meters (88 cubic feet)] of lime dust. The pressurized tank delivers
0.19 cubic meter (6.6 cubic feet) of lime per minute through a 30-meter
(100-foot), 51-mm (2-inch) diameter hose. The system is pressurized with a
3.82 cu m/min (135 cfm)/3.58 atm (38 psi) air compressor.
In addition to the investigation of devices for delivering bulk material
to spills, a brief evaluation was made of devices for dispersing canisters
and containers. These devices were considered for air-dropping MGA onto
spills that were otherwise inaccessible. In principle, the canisters would
spread the MGA over the target.
DISSEMINATION SYSTEM EVALUATIONS
Four dispersion systems were evaluated: (1) a pressurized tank, (2) a
Venturi/compressed air, (3) a centrifugal blower, and (4) an auger-fed/
pneumatic conveyor.
System evaluations were performed using powder agglomerate, Grade I,
Grade II and Rolled forms of MGA. The effect of wind, amount of MGA arriving
at target, dispersal pattern, and ease of dispersion were measured.
Pressurized Tank
A rechargable C0,,-charged dry chemical fire extinguisher was used to
evaluate the pressurized tank concept. An Ansul Company (Marinette, Wiscon-
sin) Model A-l-C fire extinguisher was loaded with approximately 0.22 kg
(1/2 Ib) of MGA, sealed and charged with a (XL cartridge.
The nozzle configuration of the test equipment used was designed for
discharging a fine powder for extinguishing fires and the larger MGA forms
(Grade II and Rolled) plugged the nozzle. However, with a properly designed
nozzle size, the larger size particles of MGA could be used.
A stainless steel tray [61 x 76 x 5 cm (24 x 30 x 2 inches)] was used to
collect the MGA that settled on the simulated spill target. The distance
between the nozzle and the simulated target was selected so that the maximum
amount of MGA would remain at the target site. Results were as follows:
33
-------
Powdered:--Less than 10% of the powdered MGA was collected in the
simulated target. A detailed study of the mass distribution of
MGA components settled from the plume under zero wind conditions
is presented in a later subsection. The plume was wide and long.
Even in light wind [107 to 214 m/min (4 to 8 mph)] most of the powder
blew away from the target area.
Agglomerate:--Less than 20% of the MGA hit the simulated liquid
spill target. The plume was not as large as the powdered MGA;
however, there was still considerable dispersal due to wind.
Grade I:--Approximately 50% of the MGA was collected on the simu-
lated liquid spill target. The plume was much smaller than the
other plumes and covered the target area.
Grade II:--The material was too large to permit expulsion from
the nozzle.
Roll Compressed:--The material was too large to permit expulsion
from the nozzle.
These initial experiments demonstrated very vividly the importance of
maximizing the mass to aerodynamic drag ratio of the MGA. In addition, they
demonstrated that the pressurized tank method of dispersion can be considered
a candidate system if a large reservoir is used with an air compressor. The
reservoir could be, for example, a tank approximately 2 meters (6 feet) in
diameter and 5 meters (15 feet) long, located on a truck or trailer bed that
is also equipped with a gasoline-powered air compressor. In this configura-
tion, assuming that the tank was 3/4 full of MGA, the tank would hold approxi-
mately 2268 kg (5000 Ib) of powdered MGA. This type delivery system is
commercially available from Mine Safety Appliance (Pittsburgh, Pennsylvania).
A possible disadvantage of this system is that when operated with an
ambient air compressor, condensation produced by expansion between the tank
and nozzle, or perhaps within the reservoir could react with the agent and
prevent proper dispersal. Additional tests are required with a large size
unit.
Venturi/Compressed Air
To evaluate this type of system, an Econo-Vac sandblasting unit
(Airplaco, Clipper Manufacturing Company, Kansas City, Mo.-Model EV-14M) was
used with an Ingersoll-Rand G85 gasoline-driven air compressor (Figure V-l).
A 5.6-liter (1-1/2-gallon) spill of trichloroethane was used as a test
spill. The exit nozzle was held approximately 60 cm (2 ft) from the liquid
surface. Various forms of MGA were evaluated.
34
-------
Figure V-1a. Venturi tube compressed air nozzle.
AIR CONTROL VALVE
NOZZLE
ATTACHMENT
ADAPTOR
MGA INPUT
i-AIR INPUT
Figure V-1b. Diagram of Airplaco Econo-vac nozzle.
35
-------
Powder--Wind, as well as the volume of air from the sandblasting
unit, caused considerable amounts of MGA to be blown away from the
target. It took approximately 2 minutes of dispersion to gel the
liquid into a gelatinous but not solidified mass. The mixture
would slowly flow when the tray was tipped (Figure V-2).
The Econo-Vac sandblaster delivered approximately 1.5 kg (3-1/2 Ib) of
powdered MGA per minute with the Venturi feed tube placed directly into
a fiber drum of MGA.
Agglomerate--The air from the sandblaster caused considerable mix-
ing action of the MGA with the liquid spill when the straight non-
flexible exit nozzle was used. This form of MGA gelled the tri-
chloroethane into a solid mass in 30 seconds, which did not flow even
when the tray was inverted (Figure V-3).
Grade I § II Both could be effectively dispersed with this type
of delivery system. The larger particles (Grade II) were least
affected by the wind but both performed well. Both grades pro-
duced an excellent gel with approximately the same amount of
material.
Rolled Compressed-- This was the optimum physical form for this
dispersion system. The material could be dispersed approximately
6 m (20 feet) without adverse wind effects. During the evaluation, the
MGA mixed well with the 5.6-liter (1-1/2 gallon) test liquid spill (tri-
chloroethane) and gelled it in minimal time.
This system of dispersion is flexible and commercially available. It is
a definite mode of dispersion for all the forms of multipurpose gelling agent
tested.
Centrifugal Blower
To evaluate this mode of dispersion, a snowblower (Ariens, 5 hp) was
used. The snowblower was tipped so that the screw-fed hopper was facing up
allowing the MGA to be poured directly into the blower. The centrifugal
blower blew the MGA into the air over the liquid spill target. The wind
effect on this system is severe using both powder and agglomerate forms of
MGA (Figure V-4). Less than 10% of the powder and 20% of the agglomerate
landed on a 6 x 6 m (20 x 20-foot) target even though the wind conditions
were near calm [maximum gusts were 134 m/min (5 mph)]. This dispersion
mode has, therefore, been eliminated as a possible delivery system.
Auger-Fed/Pneumatic Conveyor
A Bantam 400 Rockduster (Mine Safety Appliances Company, Pittsburgh,
Pennsylvania) was used to evaluate this mode of dispersion. The Rockduster
consisted of a hopper which auger-feeds the MGA to a moving air stream. The
volume of air is varied depending on the density of the material being dis-
persed. This unit can transport the powder through a 5-cm (2-inch) ID
hose up to 60 m (200 feet) long.
36
-------
Figure V-2. Venturi compressed air system blowing powdered form MGA
37
-------
a. DISPERSING AGGLOMERATED FORM
b. GELLED TRICHLOROETHANE
Figure V-3. Long tube on Venturi-compressor.
38
-------
a. EXPERIMENTAL APPARATUS
b. BLOWING POWDERED MGA
Figure V-4. Centrifugal blower (snowblower).
39
-------
The hopper was filled with approximately 0.05 cubic meter (2 cubic feet)
of powdered MGA and used a 30 m (100-foot) section of exit hose. This con-
figuration delivered 3.6 kg (8 Ib) of MGA per minute with less than 20% loss
due to wind (Figure V-5).
The unit as manufactured is driven either by a 3.7-kW (5-hp) electrical
unit (440 or 220 volts DC) or by a hydraulic unit [0.75 I/sec at 69 atm
(12 gpm at 1000 psi)]. Neither of these modes of power is ideal for portable
field application, but the unit can be modified to use a 120 volt AC motor
or a gas-driven motor. This unit is the most promising off-the-shelf equip-
ment tested.
Nozzle Configuration
Several experiments were also performed to evaluate possible nozzle con-
figurations that might be used with these systems. The Venturi tube system
equipped with a 4.4-cm (1-3/4-inch), 3-m (10-foot)-long exit tube was used.
Best results were achieved using a 4.4-cm (1-3/4-inch) diameter tube that
had been bent at an angle of 30° to the exit tube. This simple device per-
mitted the operator to accurately direct the plume of the MGA at the region
of the spill he wished to treat. Directing the plume down caused the plume to
strike the spilled liquid at relatively high velocities, thereby minimizing
wind drift of powder and, more importantly, causing substantial agitation of
the liquid as the agent entered it. This agitation promoted rapid and com-
plete reaction of the agent with the liquid.
Other configurations tested included a deflection plate, a nozzle with
an expanded opening and, of course, the straight pipe. These configurations
were substantially less effective in permitting accurate placement of the
plume.
Mass Distribution of Powdered MGA
The mass dispersion of the powdered form of MGA was evaluated indoors to
eliminate the adverse effect of wind.
MGA was placed in a CC>2-charged dry chemical fire extinguish rr whose
nozzle was collimated with a 1.2 cm (1/2-inch) diameter tube, 50 cm
(20 inches) long. The nozzle was pointed at a target 2.4 m (8 feet) away
and the entire amount of MGA was discharged in several short blasts.
A collection grid system consisting of petri dishes was used. After
expulsion, the amount of gelling agent collected in each dish was determined.
The distribution profile is shown in Figure V-6. The outer line in Figure
V-6 shows to what extent the finer materials of the blend drifted and con-
tour lines illustrate equal mass distributions. Seventy-five percent of the
material dispersed was collected within the target area (broken line).
In order to determine the distribution of each MGA component another
discharge of universal gelling agent was performed. In this experiment, the
collection grid consisted of the petri dishes set at one foot intervals in a
40
-------
Figure V-5. MGA dispersed from pneumatic conveyor/auger-fed system. Powdered
MGA being dispersed from 30-meter (100-foot) hose.
41
-------
X = COLLECTOR POSITION
75% OF MGA DISPERSED
WAS COLLECTED WITHIN
BROKEN LINE AREA
NOZZLE POSITION
meters
Figure V-6. Equal mass distribution contour map obtained from CO2-charged dry
chemical fire extinguisher dispersion of MGA.
42
-------
linear array. The nozzle was held in a horizontal position approximately
three feet above ground. The MGA was again discharged from a CCL dry
chemical fire extinguisher.
The dispersion distribution characteristics were determined by viewing
each collection dish with a light microscope at low magnification (40x).
Each MGA component showed a dispersion pattern that was predictable using its
density and particle size. The dispersion characteristic of each component
is as follows:
Gelgard--The Gelgard is of an irregular shape and has the appear-
ance of gravel. The distribution peaked around 2 meters from
the nozzle with no material dispersed beyond 5 meters.
Imbiber BeadsThe distribution of these spherical-shaped beads
peaked at 0.75 to 1 meter and beads were only seen sparingly beyond
2 meters.
Hycar 1422--This material is also spherical but only 1/3 the dia-
meter of the Imbiber Beads. Its distribution also peaked at 0.75
to 1 meter, declining rapidly at 3 meters. Unlike Imbiber Beads,
it was readily seen over the entire collection grid.
Carbopol 934--This component has the appearance of a very fine
powder and is adsorbed to the Gelgard, Imbiber Beads and the Hycar
1422; it was found to be dispersed over the entire pattern.
Cabosil--The Cabosil consists of very fine, irregular, transparent
pieces. Their distribution was observed over the entire collec-
tion grid but was mainly found beyond the 2 meter mark.
The mass distribution of the total mixture peaked at 5 meters with a
half width of 1 meter (Figure V-7).
This experiment revealed a particularly important disadvantage of the
powdered form of MGA, i.e., separation of components with distance from the
nozzle. (This separation cannot occur with any form tested except powder.)
It is apparent that, if the powdered form of MGA is used on real spills, a
nozzle configuration which permits direct impaction of the plume on the
liquid is essential.
43
-------
456
METERS FROM NOZZLE
Figure V-7. Mass distribution of MGA dispersed from CO2-charged dry chemical fire extinguisher.
44
-------
SECTION VI
DISCUSSION
Efforts to optimize the gelling agent blend produced clear cut results.
Blend D, consisting of Gelgard M, Imbiber Beads, Hycar 1422, Carbopol 934
and Cabosil, is substantially superior to other blends of the original mater-
ials tested. Blend E, in which Klucel H was substituted for Carbopol 934 in
Blend D, showed slightly superior performance on the alcohols. However,
since no experience was obtained in the manufacture of optimum forms of Blend
E, we believe that Blend D should be retained as the prime candidate for
large-scale experiments. Tests also showed that "Kelzan" can be substituted
for Gelgard and Soloid for Imbiber Beads without seriously affecting perform-
ance. Since manufacturers of the four components of Blend D occasionally
remove and substitute items in their product lines it would be wise to per-
form simple experiments by the manufacturing of selected MGA forms using
substitute materials in anticipation of the possible elimination of one of
the components from the market.
The experiments with different physical forms of MGA produced clear
evidence that the "agglomerate" form and "graded" forms were superior to all
others in static tests. However, the "agglomerate" form can be eliminated on
the basis of its poor performance in wind and its high manufacturing costs.
Grade I was superior to Grade II in the static tests but inferior to Grade II
in wind. The smaller particles in Grade I, i.e., 250 Am to ^SOO/xm, were
subject to significant wind drift, while the larger particles in Grade II
gelled slowly. The latter effect is not necessarily a disadvantage in treat-
ing real spills, since the larger particles sink more rapidly into a spill to
react with the volume of liquid rather than the upper surface.
We believe, therefore, that the optimum "graded" form should consist of
particles between 0.5 and 2 mm. This material can be produced by the same
manufacturing techniques used during this program (exercising proper selec-
tion of sieve mesh).
The "roll compressed" form used on this program was inferior to the
"graded" forms in static tests, but performed well in the field. As dis-
cussed in the text, the reasons for the reduction in performance are under-
stood and can very probably be corrected by slight adjustments in manu-
facturing procedures. Because of potential economic advantages of these
manufacturing procedures, we recommend that the "roll compressed" form be
retained for testing in large-scale experiments.
45
-------
Three kinds of off-the-shelf equipment proved to be useful in the small-
scale field experiments performed on this program. These include:
(1) Pressurized tanks similar to equipment manufactured by
Mine Safety Appliances, Company of Pittsburgh, Pennsylvania;
(2) Venturi/compressed-air types manufactured for sandblast-
ing purposes; and
(3) Auger-fed, pneumatic-conveyor equipment similar to the
"Rockduster," also manufactured by Mine Safety
Appliances.
The "Rockduster" appears to be most suitable for treatment of large
spills. Because of costs and the difficulty in transporting this equipment,
it is probably suitable for distribution only with large stockpiles of
gelling agent. Less expensive, easily transportable type equipment, such
as sandblasters, would be more appropriate for distribution with smaller
local supplies of the agent, such as at firehouses.
All three forms should be retained as candidates for large-scale testing.
Any of these units can be used with straight pipe nozzles or nozzles in
which the exit tube is bent at approximately 30°. Large-scale tests should be
performed with straight pipes equipped with detachable extensions bent at 30°.
46
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1 REPORT NO.
EPA-600/2-77-151
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Multipurpose Gelling Agent and Its Application to
Spilled Hazardous Materials
5. REPORT DATE
August 1977 issuing date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
J.G. Michalovic, C.K. Akers, R.E. Baier, R.J. Pilie
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORG ^NIZATION NAME AND ADDRESS
Calspan Corporation
Buffalo, New York 14221
10. PROGRAM ELEMENT NO.
1BB610
11. CONTRACT/GRANT NO.
68-03-2093
12. SPONSORING AGENCY NAME AND ADDRESS
Industrial Environmental Research Lab.
Office of Research & Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
- Gin., OH
13. TYPE OF REPORT AND PERIOD COVERED
Final Report
14. SPONSORING AGENCY CODE
EPA/600/12
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Previously, a blend of materials was formulated that would spontaneously
gel a wide variety of hazardous liquids. This blend, known as the Multipurpose
Gelling Agent (MGA), has been optimized to obtain a balanced formulation that
will effectively gel and immobilize most spilled hazardous liquids within minutes,
The current formulation, consisting of four powdered polymers and one inorganic
powder, has the ability to immobilize spilled liquids with the least amount of
material in the shortest period of time. In field testing of the powdered blend,
it was observed that when air conveyance modes of dispersal were employed high
losses occurred due to the effects of wind. Three compressed and granulated
forms of the gelling agent were developed which are clearly superior to the
original powdered blend for delivery to liquid spill targets. Various off-the-
shelf dry solid dispersion devices were evaluated and the most promising systems
field tested on simulated and actual spill targets, both in pools and in linear
ditches. The results show that MGA provides an efficient means to mitigate the
damages from hazardous liquid spills.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Hazardous Materials, Gelling Agents,
Gelation, Dispensers, Decontamination
Hazardous Materials Spill
Clean-up Hazardous
Material Spill Control
Multipurpose Gelling Agen
13B
3. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
55
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
47
*US. GOVERNMENT PRINTING OfFICE: 1977-757-0 56/6501
------- |