EPA-670/2-75-042
June 1975
Environmental Protection Technology Series
METHODS TO TREAT, CONTROL AND MONITOR
SPILLED HAZARDOUS MATERIALS
National Environmental Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
EPA-670/2-75-042
June 1975
METHODS TO TREAT, CONTROL AND MONITOR
SPILLED HAZARDOUS MATERIALS
By
Roland J. Pilie, Robert E. Baler, Robert C. Ziegler,
Richard P. Leonard, John G. Michalovic,
Sharron L. Pek, Ditmar H. Bock
Calspan Corporation
Buffalo, New York 14219
Contract No. 68-01-0110
Program Element No. 1BB041
Project Officer
Joseph P. Lafornara
Industrial Waste Treatment Research Laboratory
Edison, New Jersey 08817
NATIONAL ENVIRONMENTAL RESEARCH CENTER
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
REVIEW NOTICE
The National Environmental Research Center
Cincinnati has reviewed this report and approved
its 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 com-
mercial products constitute endorsement or recom-
mendation for use.
-------
FOREWORD
Man and his environment must be protected from the adverse effects
of pesticides, radiation, noise and other forms of pollution, and the un-
wise management of solid waste. Efforts to protect the environment require
a focus that recognizes the interplay between the components of our physical
environmentair, water, and land. The National Environmental Research
Centers provide this multidisciplinary focus through programs engaged in
studies on the effects of environmental
contaminants on man and the biosphere, and
a search for ways to prevent contamination
and to recycle valuable resources.
Pollution from spills of hazardous materials is widely recognized as
very damaging to the water ecosystem and to the public health and welfare.
This report describes several promising new techniques for 1) preventing
chemical spills on land from reaching nearby surface or ground waterbodies
by gellation in place; 2) detecting and monitoring spills in a watercourse;
and 3) treating waters which have been polluted by spills of a wide variety
of organic and inorganic hazardous materials.
A. W. Breidenbach, Ph.D.
Director
National Environmental Research
Center, Cincinnati
iii
-------
ABSTRACT
A program was instituted to study the feasibility of treating, controlling
and monitoring spills of hazardous materials. Emphasis was placed on con-
sidering techniques and equipment which would be applicable to general
classes of chemicals rather than to specific hazardous polluting substances.
Several methods were investigated and found to be promising for removing or
detoxifying spills of hazardous chemicals "in situ". These included: the
use of sodium sulfide as a precipitating agent for spills of heavy metal-ion
solutions; the use of activated carbon packaged in porous fiber bags (carbon
"tea bags") for adsorbing a wide variety of soluble organic chemicals; and
the use of various acids or bases to neutralize spills.
Methods were studied to control spills on land and prevent their contaminat-
ing nearby surface or ground water. To this end, a four-component "universal
gelling agent" was developed to immobilize a spilled liquid.
A "cyclic colorimeter", a novel heavy metal-ion detector, was perfected and
laboratory tested, and a detection kit capable of sensing several chemicals
was developed.
A computer model was developed and refined to simulate the spread of a
spill when certain stream parameters and material characteristics are known.
Bioassay studies were conducted for several chemicals using at least three
species of biota. In addition, bioassays were conducted to estimate the
environmental effect of each of the various treatment methods developed.
-------
CONTENTS
Abstract
List of
List of
Sections
1
2
3
4
5
6
7
8
9
10
11
12
13
Figures .......
Tables
INTRODUCTION
SUMMARY
CONCLUSIONS
RECOMMENDATIONS
THE USE OF ACTIVATED CARBON FOR THE REMOVAL OF HAZARDOUS
CHEMICALS FROM WATER BODIES
TREATMENT OF WATER SPILLS WITH ION EXCHANGE RESINS ....
TREATMENT OF SPILLS OF ACIDS AND BASES BY DIRECT
NEUTRALIZATION
PRECIPITATION OF HEAVY METALS WITH SODIUM SULFIDE ....
IMMOBILIZATION OF HAZARDOUS CHEMICALS
DETECTION AND MONITORING
BIOASSAY STUDIES
MATHEMATICAL MODELING OF HAZARDOUS SPILLS AND SPILL
TREATMENT
HAZARDOUS CHEMICALS /COUNTERMEASURES
Page
iv
vi
vii
1
4
8
13
17
36
43
45
62
88
90
106
127
References 137
-------
LIST OF FIGURES
Figure No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Title
Residual Phenol Concentrations Resulting from Powdered
Phenol Removal by 8x30 Nuchar Under Ideal Conditions. . . .
Weight of Carbon Required per Unit Weight of Phenol Spilled
Bag Tests with Various Granular Carbons ..........
Tea Bag Granular Carbon Adsorption of Phenol from Distilled
Water, Tap Water, and Creek Water
Carbon Tea Bag Effectiveness Under a Variety of Conditions
Removal of Phenol by Activated Carbon (8x30 Mesh Nuchar) in
Pool Tests
Tea Bag Design, Channel Tests 2 and 3
Tea Bag Configuration for Pool Test Number 1
Comparison of Carbon Fibers with Powdered Activated Carbon
Toxicity of Na_S and NaOH (24 Hour Exposure)
Toxicity of Na S and NaOH as Functions of pH of Their
Solutions (24 Hour Exposure)
Lead Concentration as a Function of Titrant Added .....
Experimental Facility at Bethany Test Site
Immobilization of Hazardous Chemicals on Land with Gelling
Sketch of Pad for Vapor Suppression Tests ....
Plan View of "Lake" Cyclohexane Spill Treatment Test. . . .
Increasing the Effectiveness of Booms by Congealing the
Page
18
19
20
25
26
27
29
30
31
33
50
53
53
58
67
71
73
78
79
vi
-------
LIST OF FIGURES (Cont.)
Figure No. Title
20 Cross Section of the Ditch as Modified to Trap Heavier-
Than-Water Spills 80
21 Treatment of Sulfuric Acid Spill 83
22A Slow Reaction Which Occurs When Lime is Spread on Top of
Acid-Fly Ash Slurry 84
22B Cloud Formed by Intense Reaction that Occurs When Lime
Layer is Raked into Acid-Fly Ash Slurry Below 84
23 Algae Growth Curves for Acetone Cyanohydrin, Acrylonitrile
and Chlorine Tests at 10 ppm (60°F ±1°F) 93
24 Algae Growth Rate with Time after Exposure to 10 ppm
Methanol, Cadmium, and Chromate 94
25 Algae Growth Rate with Time after Exposure to 10 ppm
Phenol, Ammonium Hydroxide, and Mercury 95
26 Dissipation of Volatile Pollutants (ppm) with Time in an
Aerated System 96
27 Equilibrium Adsorption of Phenol by Nuchar (8x30) Ill
28 Concentration-Time Values Derived from Equation 8 (Curve)
Compared with Experimental Data (Points) 113
29 Concentration-Time Values Derived from Equation 4 (Curves)
Compared with Experimental Channel Data (Points) ..... 114
30 Dispersion of a 5000 Gallon Phenol Spill in a Non-Flowing
Water Body 118
31 Two Dimensional Stream (w/text) 119
G!
32 Mean, m = -^ (Zx,t) vs t 125
^)
33 Variance a Versus Time 126
VII
-------
LIST OF TABLES
Table No. Title Page
1 Adsorption Activated Carbon 21
2 Carbon Adsorption of Mn, Cu, Ni, Cr, and Hg 23
3 Removal of Arsenic by Ion Exchange Resins and
Activated Carbon in "Tea Bags" 37
4 Removal of Hexavalent Chromium by Ion Exchange Resins and
Activated Carbon 38
5 Removal of Trivalent Chromium by Ion Exchange Resins and
Activated Carbon 39
6 Removal of Phenol by Ion Exchange Resins and Activated
Carbon 41
7 Periodic Table of the Elements 46
8 Quantity of Standard Sulfide Treatment Solution Required
to Treat One Pound Spills of Indicated Compounds 47
9 Comparison of Heavy Metals and Na2S Hazards 52
10 Heavy Metal Analysis of Fish from Heavy Metal Sulfide
Exposure Tests (ppm of Wet Weight) 56
11 Sulfide Precipitation Experimental Results 57
12 Heavy Metal Spills 60
13 Compounds Tested with Universal Gelling Agent 60
14 List of Foam Concentrates for Flammable Liquid 65
15 Summary of Results on Combined Vapor Suppression and
Gelling Tests 75
16 The Chemical Analysis of Test Water Source (mg/1) 91
17 Mortality Data in Percent Deaths of Minnows with Time after
Exposure to 10 ppm of Nine Chemicals at 60°F 92
18 Safe and Lethal Concentrations of Pollutants Including the
LD for Fathead Minnows and for a 7-Day Period of
Observation. 97
Vlll
-------
LIST OF TABLES (Cont.)
Table No. Title
19 Length of Time before Physical Effects and/or Deaths
were Observed 98
20 Percent Survival of Fish with Time during Exposure to
a Phenol-Carbon Sediment (Aqua Nuchar A; Powdered Carbon) 99
21 Percent Survival of Fish Exposed to Phenol-Carbon
Sediment (Granular Carbon) 99
22 Percent Survival of Minnows Exposed to a Sediment of
Acrylonitrile-Granular Carbon and Sand 100
23 Estimates of Toxicity of Various Metals to Fish 102
24 Data on the Toxicity of Na~S to Minnows (Percent of Deaths
with Time) 103
25 Percent Deaths of Minnows with Time after Exposure to
NaCl, NaN03, and Na2S04 104
26 Summary of Data from Metal Precipitate Tests 105
27 Summary of Parameters Estimated in Treatment Model
(Phenol-Nuchar 8x30) 115
-------
SECTION 1
INTRODUCTION
1. OBJECTIVE OF THE PROGRAM. The objective of this program, sponsored under
EPA Contract No. 68-01-0110, was to investigate, develop and demonstrate methods
to treat, control and monitor spills of hazardous materials. The program was
broad-ranging and meeting its objective involved the performance of analytical,
laboratory, and field work with hazardous chemicals and consideration of the
many conditions of their manufacture, transport, storage, and use.
2. SCOPE OF THE PROGRAM. The program emphasized the development and identifi-
cation of methods, procedures, equipments, and techniques that will ameliorate
the spills of hazardous materials. Second, the scope of the program encompassed
all types of chemicals and heavy metal salts; it concentrated on those chemicals
and metals that are typical of classes of chemicals and that were deemed most
hazardous in a previous study performed by Battelle Institute for the EPA
(Ref. 1). (Instrumentation was developed to detect and monitor chemicals in
water.) Spills on land as well as in water were of concern; although the
feasibility of using some methods to prevent hazardous materials spilled on land
from reaching watercourses was shown, methods to treat spills in water were the
prime target of investigation. The effects of both the spilled chemicals and
the treatment methods on the aquatic environment were studied. Preliminary
information was gathered on the costs and logistics of applying treatment methods,
Specific tasks that were undertaken are best summarized in the list below.
(1) Assess existing and develop new methods for treating and controlling
spills in watercourses.
(2) Assess existing and develop new methods for treating and controlling
land spills.
(3) Develop methods for disseminating and/or applying treatment materials.
(4) Investigate means for reducing toxicity of hazardous materials
spilled in watercourses.
(5) Develop detection and monitoring techniques.
(6) Investigate flotation, containment, and skimming operations
(as appropriate).
(7) Perform bioassays to determine the effects of countermeasures on
aquatic organisms.
(8) Provide limited demonstration of developed techniques where
feasible and appropriate.
-------
(9) Evaluate logistics, cost, etc. for countermeasures developed.
(10) Prepare recommendations for programs to demonstrate the
effectiveness of the methods developed.
At the direction of the EPA, emphasis within the program in the implemen-
tation of these tasks was placed on the test and demonstration of control
techniques for near-term operational use and on instrumentation for detecting
and monitoring pollutants in watercourses. A second important emphasis was
the generalization of the technology developed for application to as many types
of chemicals as possible. This was accomplished by classifying chemicals into
groups with common physical and chemical characteristics, testing a variety of
chemicals from each class in small-scale laboratory experiments to assure common
response to a given treatment, and finally demonstrating the applicability of
the treatment concept on a selected member of the class in large-scale labora-
tory or field experiments. A major output of this program is a chart presented
in Section 13 that contains recommended procedures for treating and controlling
spills of some 250 chemicals.
Important program considerations with respect to methods, treatments, and
devices other than the necessity that they be effective were:
(1) The control or treatment method must be capable of being used for
a wide variety of chemicals and substances.
(2) The control procedure or techniques must be safe and simple to use.
(3) The equipment and material to be applied must be inexpensive and
portable.
(4) The recommended methods for treating the spill should be usable
throughout the country.
It is believed that this program has identified, and in many cases tested,
many of the best answers currently available for handling spills of hazardous
materials. Conclusions and recommendations of the program for improving and
demonstrating spill abatement technology are contained in Sections 3 and 4 of
this report.
This volume is a summary volume and presents only sufficient material to
enable the reader to have an understanding of the work performed, and of the
bases for the recommended treatments and demonstrations. Additional material is
available to describe in detail the work that was accomplished on:
(1) The treatment of spills of heavy metal compounds.
(2) The experimental work on carbon and resin adsorption.
(3) The design and development of special instrumentation.
-------
(4) The description of immobilization technology for treatment of spills.
(5) A description of the bioassay work accomplished.
(6) The modeling methodology developed in the program.
The reader is referred to this material for substantiative data and a description
of the detailed approach used throughout the program.
-------
SECTION 2
SUMMARY
The principal accomplishments of this program may be summarized as follows:
1. ACTIVATED CARBON TREATMENT OF CHEMICALS IN SOLUTION. Extensive experi-
mental and theoretical work was performed on the use of activated carbon for
removal of dissolved pollutants from surface waters. Bench scale experiments
were performed on twelve organic chemicals ane eight metallic ions, with ini-
tial pollutant concentration ranging from 1 to 1000 mg/1 and with carbon to
pollutant ratio ranging from 1 to 100. Five different types of powdered and
granulated carbon and several fibrous carbon products were used. Excellent
results were achieved with all organics tested except methanol and with some
metallic ions, particularly with Hg+2 (99% removal).
The principal problem with the use of activated carbon in natural waters
is the removal of the carbon from the water after adsorption is complete. To
solve this problem a concept was developed for treating spills with carbon
contained in well ventilated, porous bags that are floated in the waterway.
These "tea bags" containing the carbon are readily removable from surface waters.
Bench scale tests demonstrated that utility of this concept. Tests performed
in a simulated stream consisting of a 1 ft2 cross section, 28 ft long, 8 ft
wide race track channel with controllable flow velocity and adjustable bottom
roughness were responsible for evolving near optimum bag designs and estimates
of time required for treatment under different flow conditions. Tests per-
formed in 12 ft diameter swimming pools demonstrated the importance of a small
amount of wave action for providing adequate ventilation in non-flowing water-
ways. Full scale bags (10 ft long) were used to demonstrate that proper agi-
tation would be achieved with waves as small as 3 inches high. The ease of
recovering the carbon-filled bags from either type of waterway was demonstrated.
A short operational analysis performed on this concept indicated that a system
of nationwide distribution of activated carbon-filled bags for spill removal
is feasible and economically achievable.
In addition to work with free-powdered and granulated carbon products
and carbon filled "tea bags", bench scale tests were performed on newly
developed, experimentally available activated carbon fibers which resemble
steel wool. Such products, if ever available at prices comparable to granular
carbon, would offer significant advantages in speed of adsorption and therefore
pollutant removal efficiency without requiring packaging in porous bags to be
removable from a water course.
2. TREATMENT OF WATER SPILLS WITH ION EXCHANGE RESINS. Limited bench scale
experiments were performed to examine the potential for using ion exchange
resins for removal of dissolved chemicals that do not respond to activated
carbon treatment. Significant success was achieved. It appears that the
porous bag approach would be useful for dispersal and retrieval of these
products. No experiments work was performed with such packaging, however.
-------
3. PRECIPITATION OF HEAVY METALS WITH SODIUM SULFIDE. The treatment of
heavy metal compounds by precipitation of insoluble metal sulfides was evalu-
ated. The treatment is applied by introducing sulfide ions, derived from a
solution of sodium sulfide, into the spill. A wide range (literally thous-
dands) of heavy metal compounds can be treated by this method. Bench scale
tests were performed on 28 of these compounds on the program. Channel tests
were performed to test effectiveness of the treatment in flowing streams and
methods for detecting several spilled materials for treatment. Large scale
laboratory experiments were performed in 12 ft diameter swimming pools using
ferrous sulfate as the spill simulant. Small scale field experiments per-
formed under natural conditions also made use of that harmless stimulant.
These experiments demonstrated that small spills can be safely located by
injecting small amounts of sodium sulfide into the waterway and noting the
location of the visible precipitate and treated by dumping the appropriate
quantities of sodium sulfide into the traced spill from plastic bottles.
Large spills require carefully controlled treatment with sodium sulfide
to avoid potentially dangerous secondary pollution. An equipment concept
was developed in which metallic ion concentration measurements would be used
to automatically control the amount of sodium sulfide dispensed and feedback
would be used for compensating for measurement errors.
An experimental program was performed to train interested volunteer
firemen in the use of these treatment procedures to minimize hazards from
small heavy metal spills.
4. IMMOBILIZATION OF HAZARDOUS CHEMICALS. A gelling agent consisting of
four active ingredients and one inert powder was developed and tested success-
fully on 40 liquid chemicals in bench scale experiments. In each case the
chemicals were congealed into a thick, viscous mass that would not flow when
the beaker was turned on its side. The material was field tested in specially
constructed ditches with 1% and 2% grades. In all cases 55 gallon liquid
spills were totally immobilized in less than 80 ft by the application of 10
to 25% (by weight) of the agent. Similar tests against liquids floating on
water improved the efficiency of crude booms to the extent that 4 inch thick
layers of gelled material could be retained with no difficulty. In all cases
the gelled material was removed from the surface with shovels and returned to
the 55 gallon drums for later disposal.
Narrow slits in tanks were sealed to prevent further spillage by inject-
ing the gelling agent into the slit with sand blasters.
An operational analysis was performed which indicated the feasibility of
establishing caches of this material around the nation so that it could be
quickly transported to a spill site. The economics of such preparations are
reasonable.
Using concentrated sulfuric acid as the spilled material, a large scale
experiment was performed to demonstrate that highly reactive materials can be
immobilized by absorbing the liquids in fly ash. Neutralizing agents can be
-------
placed on the liquid-fly ash slurry for treatment. The reaction rate can then
be controlled by controlling the rate at which the neutralizing agent is mixed
into the slurry.
Spills of heavy metal compounds absorbed by the ground were immobilized
by treatment with sodium sulfide. Movement of the metal sulfide precipitates
formed under the surface by the procedure was eliminated except in conjunction
with soil in which they were formed.
Using ethylene dischloride as a test material, it was effectively demon-
strated that spills of liquids that sink in water can be trapped in excava-
tions in stream beds without interrupting stream flow.
5. DETECTION AND MONITORING. A cyclic colorimeter which uses modulation of
indicator injection to reduce sensitivity to natural turbidity and fouling was
designed, constructed and tests on this program. A spill detection kit con-
taining conductivity meter, pH indicator, odor samples, and colorimetric rea-
gents was developed and proven effective in tests involving volunteer firemen.
Tests were performed of the utility of conductivity meters, pH probes and
specific ion electrodes for detection and measurement of spilled acids, bases,
and metallic compounds. Catalytic combusters were tested for detection and
monitoring of spills of volatile organic solvents.
*> MODELING. Two types of models were developed on this program. In the
first type an expression was developed to describe the rate of removal of
a pollutant from solution by activated carbon adsorption in terms of three
constants that are readily determined by simple laboratory experiments and
a parameter which depends on the ventilation of the carbon by the solution
in the specific flow (or wave agitation) conditions of the spill.
The second type of modeling effort considered the dispersion of the
spilled chemical in a waterway. Two modeling concepts were developed. In
the simpler of the two models, the solution to primitive diffusion and flow
equations were programmed on a desk top computer to produce plots of the con-
tours of pollutant concentration at any selected time after a spill. The
model is intended for use in an on-site, adaptive prediction procedure that
makes use of measurements of pollutant concentration made at the spill site
to determine the dispersion coefficients to be used in predicting future con-
centration contours.
The second, a more sophisticated model, includes the possibility of incor-
porating a treatment effect in estimating the pollutant dispersion. This
model permits analysis not only of the concentration distribution of pollutant
but also of the mean, variance, skewness, and kurtosis of the distribution.
7. BIOASSAY STUDIES. Toxicity tests were performed with methanol, phenol,
acrylonitrile, ammonium hydroxide, acetone, cyanohydrine and chlorine using
fathead minnows as test subjects. Additional tests were performed to deter-
mine toxicity to the minnows of methods used to treat spilled materials and
of reaction products that remain after treatment.
-------
8. HAZARDOUS CHEMICAL/COUNTERMEASURES MATRIX. The final section of this
report is a compilation of recommended procedures for treating and controlling
some 250 recognized hazardous liquids having a high probability of being
involved in spills into or near water courses.
-------
SECTION 3
CONCLUSIONS
Specific conclusions based on work performed on this program are summarized
here. Specific recommendations are summarized in Section 4. A generalization
of conclusions and recommendations for treatment of spills of 250 of the most
hazardous chemicals in the form of a treatment matrix that is useful to the field
worker is presented in Section 13 of this report.
1. ACTIVATED CARBON TREATMENT OF CHEMICALS IN SOLUTION. Adsorption on acti-
vated carbon is an effective method for removal of many compounds from aqueous
solutions.
A dosage of 10 parts of carbon to one part pollutant is a reasonable com-
promise for spill treatment.
Carbon must be packaged in such a way as to permit removal after adsorption
is complete in order to prevent unacceptable esthetic damage and possible toxic
bleed off.
Packaging of granular carbon in porous bags,i.e-> carbon teabags, is an
effective procedure which permits dispersal and essentially complete retrieval
without causing a prohibitive decrease in adsorption rate.
A new product, aggregates of carbon fibers that resemble loosely-packed
steel wool, which is available in small experimental quantities only, appears
to offer similar dispersal and retrieval capability without auxiliary packaging
and does not decrease adsorption rate.
The cost of preparations on a national scale for treating spills with
activated carbon appears to be reasonable considering the potential benefits.
2. TREATMENT OF WATER SPILLS WITH ION EXCHANGE RESINS. Limited laboratory
experiments with ion exchange resins have demonstrated that these products are
effective for the removal from solution of some of the chemicals which do not
respond to activated carbon treatment. It appears that the teabag technique
would be suitable for dispersal and retrieval of these products.
3. PRECIPITATION OF HEAVY METALS WITH SODIUM SULFIDE. The use .of sodium
sulfide treatment to form heavy metal sulfide precipitates has been shown to be
effective in significantly reducing the concentration of many heavy metal ions
in solution and thus reducing the immediate hazard of heavy metal spills. But
since sodium sulfide is itself toxic, care must be taken in its application to
spills.
Since the metal sulfide precipitate formed upon contact with sodium sulfide
solutions is usually readily visible, it is appropriate to use the sodium sulfide
solution in small quantities to locate the plume of toxic heavy metal compound
8
-------
and define its boundaries in the water body. Appropriate quantities of the
treatment may then be added at the proper location to effect the precipitation
without creating a secondary hazard from the sodium sulfide. Analysis of
existing toxicity data indicates, however, that even if by error the heavy
metal spill were completely missed by the treatment, the sodium sulfide would
pose a problem that is less than one-tenth as severe as the initial spill.
Weighing the potential secondary hazard against the potential reduction in
toxicity of the initial spill recommends strongly that the sodium sulfide
treatment procedure be implemented for small spills of heavy metal compounds.
The secondary hazard that could result from errors in the treatment of
large spills is significant. For this reason, the treatment of large spills
should be based on local metal ion concentration measurements. Techniques for
such a treatment are more demanding in terms of equipment and operator skill
and require further development for effective field use.
No effective method has been achieved for removing the heavy metal pre-
cipitates from affected water. In most natural streams diffusion will reduce
the esthetic problem and residual toxicity associated with the suspended pre-
cipitate so that no significant problem is anticipated if treatment is limited
to small spills. More research is required on the eventual fate of the pre-
cipitate before conclusions can be drawn relative to the use of this technique
on large spills.
Local fire departments throughout the nation can be trained and supplied
with the materials and equipment required for applying the sodium sulfide
treatment to small heavy metal spills at a very small cost.
4. IMMOBILIZATION OF mZARDOUS CHEMICALS. Most hazardous chemicals (all tested)
spilled on land can be effectively immobilized by the application of the
gelling agent formulation developed on this program. This gelling agent con-
sists of equal quantities of four active ingredients, each selected for its
ability to congeal one class of chemicals, and one inert powder intended to
improve flow and dispersion characteristics. A weight of gelling agent equal
to 10 to 25% of that of spilled material is required for complete immobilization.
Surface flow can be stopped by creating dams or levees of congealed material
by the application of significantly smaller quantities of the gelling agent to
the leading edge of the spill, depending on the nature of the terrain on which
the spill occurs. Such a procedure would not inhibit percolation of the spilled
material into the soil, however.
Selective use of specific agents for congealing chemicals of the class for
which they are applicable could reduce the mass of agent used but would require
trained personnel for decision making as to which active ingredient to employ.
Once immobilized, the bulk of the congealed material can be removed from the
surface by simple mechanical means and placed in suitable containers for ship-
ment or further treatment. Because of the sticky nature of most congealed
chemicals, a very small residue always adheres to the surface after mechanical
removal and, depending on the nature of the problem, requires further cleanup.
-------
One of the ingredients used in the formulation has excellent lubricating
properties even when mixed with large quantities of water and could pose a
safety hazard. Care must, therefore, be taken to effect thorough cleanup of
the treated area in all regions where pedestrian or automotive traffic is expected.
The gelling agent formulation can be used Affectively for sealing long,
narrow (e.g., 1/4-inch wide) splits in containers to minimize the quantity of the
material that is spilled. In many cases, the specific gelling agents appear to
be more effective than the formulation for this purpose. More extensive testing
is required.
Gelling can be used very effectively to improve the efficiency of floating
booms commonly used to immobilize chemicals that float on water. In most cases
booms begin to leak when the thickness of the floating material increases to
about 1/16 inch. When the gelling agent is applied to spills of floating
chemicals, rafts as thick as 1/4 inch are formed even without booms. Simple
booms constructed of lengths of 4" x 4" lumber with a skirt of coarse screen
can retain congealed material in floating rafts as thick as four inches without
leakage. These thick rafts may be towed across the surface to beaches where
the congealed material can be removed mechanically to containers for further
shipment.
The cost of the present mix of the gelling agent is approximately $2.00/lb.
Discussions with chemical companies suggest that an equivalent mix could be
produced in large quantities at one-fourth that price. Preparations could be
made on a national baiss for immobilizing spills with the gelling agent at a cost
of under $10 million.
Many hazardous liquids can be immobilized by absorption in fine powdered
material such as fly ash which exists in great quantities in the United States.
The use of fly ash to immobilize concentrated sulfuric acid simultaneously
provides a mechanical method for controlling reaction rate of the acid with
neutralizing agents to prevent excessive bubbling and frothing. This procedure
in its crudest form could be used on real spills at its present state of
development.
Spills of most heavy metal compounds that have been absorbed by surface
soils can be effectively immobilized by treating the soil with sodium sulfide
and soaking with water to promote formation of heavy metal sulfide precipitates
beneath the surface. These particulates are immobile except in association
with land erosion.
Materials that are immiscible with water and are characterized by specific
gravity exceeding unit can be effectively trapped beneath flowing water in
excavations produced in stream bottoms. The spilled material can then be pumped
from the artificial sump into containers on the surface.
5. DETECTION AND MONITORING. Electrical conductivity measurements are effective
in detecting the presence and extent of ionic solute spills in water but are not
suitable for determining the effectiveness of treatment.
10
-------
Commercial pH probes and certain specific ion probes are highly effective
for detection and measurement of spills of acids, bases and metallic compounds
and in monitoring the effectiveness of treatment.
Catalytic combustors are useful for monitoring spills of volatile organic
solvents and indicating the potential fire hazard associated with them.
Less volatile organics are effectively observed with multicolor trans-
missometers.
A cyclic colorimeter which uses modulation of indicator injection to reduce
sensitivity to natural turbidity and fouling was designed and tested in the
laboratory. Colorimetric indicators are valuable for any spill detection kit.
A spill detection kit containing a conductivity meter, pH indicator, odor
samples, and colorimetric reagents was developed and proved effective in tests
involving volunteer firemen.
6. MODELING. A mathematical model developed on this program accurately por-
trays the interaction of activated carbon with phenol in solution under a wide
variety of treatment conditions.
The model can be incorporated with a dispersion model to predict effects of
treatment and to permit an accurate evaluation of treatment. Simple laboratory
experiments can be performed with various carbon-pollutant systems to evaluate
constants of the model which will enable the investigator to make quantitative
comparisons of activated carbon products.
A small but multifaceted mathematical modeling effort was initiated late
in the program. One of the goals of this effort was to investigate the possi-
bility of using desktop computers to obtain useful though primitive solutions
to spill dispersion problems as an aid to decision making in the field. A
mathematical model was programmed for a Hewlett Packard (Model B-10) computer
which, using fundamental physical and chemical properties of the spilled material
and estimated dispersion constants, produces graphic printouts of expected con-
centration isopleths of the spilled material for selected times after the spill.
The value of such a simplified model stems from the ease with which an
interaction between the model and real time data acquired at a spill site can be
established. Quick comparisons of model predictions with measurements at the
spill can be made so that assumed model parameters, such as dispersion coefficients,
can be adjusted to be representative of the situation at hand. The model may
then be run to provide more realistic predictions of the location and config-
uration of the spill plume for any desired future times, and thereby permit more
realistic decisions to be made relative to the treatment of the spill. In view
of the serious uncertainties in values of dispersion coefficients and the wide
variations in these values with time and position in a given water body, we
believe that only through such an adaptive system can realistic predictions be
made.
11
-------
7- BIOASSAY STUDIES. Toxicity tests were performed with six frequently
spilled compounds using fathead minnows, Pimephales promelas, promelas, as
subjects. Arranged in order of increasing toxicity to the minnows, they are:
methanol, phenol, acrylonitrile, ammonium hydroxide, acetone, cyanohydrin, and
chlorine. The bioassay research extended also to studies of the toxicity of
the minnows to methods used to treat spilled chemicals, and of the reaction
products that may remain in the water after spill treatment.
12
-------
SECTION 4
RECOMMENDATIONS
1. ACTIVATED CARBON. Three additional investigations are required before
extensive preparations are made for implementing the activated carbon treatment
procedures developed on this program.
(1) A series of field experiments must be performed to learn how the
treatment can best be applied under several types of real con-
ditions, and to provide experience necessary for making realistic
decisions relative to implementing large-scale utilization of the
procedures. To avoid undue delay, we recommend that the carbon
teabag technique be used for these experiments. To properly
interpret the data from these experiments and to help in the
planning for implementation of the technique, the integrated
dispersal-treatment model should be made operational.
(2) A program should be initiated to develop and test a carbon fiber
product which has all of the characteristics necessary for
treatment of spills. We recommend that the program include the
joint efforts of a manufacturing concern such as Carborundum and
a corporation having environmental facilities. One goal should
be to develop a pilot plant with proper quality control for
producing an activated carbon fiber wool with the mechanical and
chemical properties of the best sample tested on this program.
Production capacity should be sufficient to permit assessment of
the product in the field and provide good estimates of large scale
production cost. A second goal should be to provide extensive
testing of the material, at first to aid in the generation of
specifications for the final product and subsequently to provide
field experience with its distribution, use and recovery. Small-
scale laboratory experiments should be performed with the final
product and a wide variety of potential pollutants to generate the
constants necessary for use in the treatment model.
(3) A detailed operational analysis should be performed to develop
the optimum plan for preparation for, and implementation of, the
carbon adsorption treatment procedures for real spills.
2. ION EXCHANGE RESINS. Ion exchange resins are more effective for removal of
some pollution from aqueous solutions than activated carbon. A more extensive
evaluation than was possible on this program is recommended before decisions are
made relative to incorporation of ion exchange resin treatment capability with
any activated carbon treatment capability that is implemented.
3. PRECIPITATION OF HEAVY METALS WITH SODIUM SULFIDE. On the basis of the
studies of heavy metal spill treatment which have been completed, a number of
13
-------
recommendations for further action are made. These include a pilot program to
verify the suitability of some techniques developed for general spill treatment
use, and further studies to refine other techniques and develop them to a state
of field utility. Studies are also recommended to improve the environmental
effects data base on which treatment decisions must be made.
The basic sulfide treatment process has been proved effective and reasonably
safe in laboratory studies and in field-scale experiments. A pilot program
should accordingly be undertaken including further field tests and the use of
the technique as a response to actual spills and culminating in the distribution
of information and materials for spill response to an adequate number of response
centers (volunteer fire companies for example are a possibility) to provide
rapid treatment of most spills. The program should include a demonstration of
the use of the sulfide treatment with a spill tracking kit like that developed in
this hazardous spills study which provides for the identification of a variety of
hazards.
Further study and development of advanced treatment techniques is recom-
mended to provide low-risk response to large spills. The use of specific-ion
electrodes for the metals, the cyclic colorimeter, or a titration and a sulfide
electrode to control localized sulfide additions should be studied further.
Problems to be solved in achieving effective treatment center primarily on
response rate limitations of the detector-human-treatment system and may best
be alleviated by development of a more highly automated approach.
Further study of the fate of the sulfide precipitates formed is recommended.
A better knowledge of the deposition pattern of the precipitates and of the
paths of possible reentry of the heavy metal into the biosphere could provide
greater confidence that the spill hazard has been eliminated, or lead to the
development of additional spill response procedures.
Finally, the development of a more comprehensive, consistent body of data
on the environmental effects of the heavy metals and of treatment materials is
considered highly desirable. Attempts to compare hazards and benefits for
treatment procedures brought to light a number of deficiencies in existing
toxicity data. They were found to be often inconsistent, incompletely reported
and reported for a wide variety of organisms (but only a few for any one substance)
necessitating cross-species comparisons. It is recommended that toxicity tests
be performed for all materials for a few selected test organisms with careful
attention to such variables as pH, temperature, and dissolved oxygen and with an
adequate control reproducing all factors except the presence of pollutant in the
test tank.
4. IMMOBILIZATION OF HAZARDOUS CHEMICALS. The immobilization procedures
developed on this program are now ready for experimental use on inadvertent
spills, even though significantly more development and testing is required to
optimize products, equipment and procedures.
(1) A few organizations with competent, imaginative field personnel
should be selected, trained and equipped to apply immobilization
procedures experimentally to real spills. Experience obtained
14
-------
should be documented to aid in the generation of instructions
for future use of the procedures. With the assistance of the
EPA, authorization should be obtained from local and state
authorities to apply these techniques in the field.
(2) The experimental investigation of methods for immobilization of
spilled material should be continued and expanded. Emphasis
should be placed on developing and testing equipment and techniques
for applying the procedures, and on reduction of the hazards, particu-
larly the fire hazards to personnel performing the operations.
Techniques should be tested on a larger scale simulating a wider
variety of terrain, stream and spill conditions.
(3) Significantly more attention should be given to methods for
interrupting a spill by sealing split or punctured containers.
(4) A development program is recommended for optimizing the gelling
agent which includes an investigation of compounds which can be
substituted one for the other to improve overall effectiveness
of the universal blend and provide special features that may be
most important for specific areas where special problems exist.
Tests should be performed to determine proper particle-size
distribution of each component in the mix and the degree of
fluidization required for ease of field deployment. The suscep-
tibility to degradation or deterioration under a variety of storage
conditions should be investigated. The end product of this effort
should be a list of specifications which the EPA could use in a
subsequent request for quotation for large quantities of the
material for nationwide distribution and stockpiling.
(5) The investigation of the use of ground sealants to prevent perco-
lation was extremely limited on this program because of the
inability to acquire proprietary materials. Further investigation
is recommended.
(6) A detailed operational analysis is required to determine optimum
methods for implementation of the techniques, including distri-
bution of equipment and agents throughout the country, trans-
portation means to spills, operational agencies to be used for
treatment, special problems associated with particular regions
and educational requirements for operational agencies.
5. PROTECTION OF PERSONNEL. There is a serious need for an investigation of
safety precautions to be used by spill cleanup teams. We recommend that such
an investigation be initiated immediately. Improved apparatus and procedures
should be developed where necessary and recommendations for safe handling of
spills should be widely disseminated among the appropriate agencies.
15
-------
6. DETECTION AND MONITORING. A kit of instruments and chemical indicators
similar to the Calspan spill kit should be made available to appropriate
federal, state, and local agencies that may be required to respond to spills
of hazardous materials.
A short educational movie on the proper use of the kit should be produced
and made available to these agencies on a loan basis.
The fact that pH electrodes have been provided with considerable resistance
to fouling and other interferences suggests that the same could be done for other
specific ion electrodes, e.g., sulfide, bivalent metal, and cyanide probes. Some
research and development in this area is in progress. This effort should be
encouraged.
Instruments to be incorporated in a device for treating spills of hazardous
substances were developed to a point where they could be tested in closed loop
control systems. Such tests should be conducted to establish stability and
accuracy of this treatment approach. The development of a suitable treatment
apparatus was recommended elsewhere.
A fieldable model of the cyclic colorimeter should be constructed and
tested under real life conditons at the outfall of a plant where spills may occur.
7. MODELING. The most difficult problem in predicting the location and geometry
of a spill plume is the selection of dispersion coefficients that are appropriate
to the situation at hand. Additional research is required to obtain sufficient
understanding for estimating these coefficients from hydrological information on
the different types of water bodies in which spills may occur. We recommend an
energetic pursuit of this goal. This effort will necessarily be prolonged and
expensive.
As an interim measure, we recommend that attempts be made to test the concept
of an adaptive modeling procedure for predicting plume characteristics in a
manner similar to that discussed in this report. This effort can be effectively
incorporated as part of a field research effort intended to test concepts for
cleanup of water spills.
8. - BIOASSAY STUDIES. Environmentalists require information on the effect of
pollution on complete ecological systems for a variety of purposes involving
chronic conditions. Under the practical conditions of an acute spill, which
presumably is a one-time event in a given water body, the utility of complete
information is mostly limited to the time period devoted to rejuvenation of the
water body. The immediate information required for decisions by the director of
a cleanup team is the dosage of the spilled material that will produce a signifi-
cant fish kill. To provide this information, it is recommended that bioassay
studies similar to those performed on this program be extended to provide base-
line toxicity data for a wide variety of chemicals that may be involved in spills.
A single species of moderately tolerant fish, such as the fathead minnow, should
be used as subjects in these experiments. Such a program would provide the data
necessary for making decisions in the field and simultaneously provide information
on the relative toxicities of the important compounds that are most likely to be
spilled.
16
-------
SECTION 5
THE USE OF ACTIVATED CARBON FOR THE REMOVAL OF
HAZARDOUS CHEMICALS FROM WATER BODIES
Activated carbon has been used for many years as an adsorbant for organic
and inorganic pollutants. It has been used in water treatment plants to remove
undesirable odors and tastes from drinking waters, and in recent years, has been
used to remove dissolved organics from municipal and industrial wastewater.
Because activated carbon is known to adsorb an array of organic substances as
well as some inorganic ionic species, it is a logical candidate as a broad
ppectrum treatment of hazardous organic liquid and heavy metal spills from water.
Early carbon adsorption experiments on this program were aimed at determin-
ing the effectiveness of powdered and granular carbons in removing six specific
chemical pollutants. The initial objective was to determine dosage requirements
for these chemicals, which included methanol, acrylonitrile, acetone cyanohydrin,
phenol, chlorine and ammonia. Tests were performed in the laboratory with
fixed volumes of solution and continuous, controlled agitation. Pollutant con-
centrations were measured at prescribed time intervals up to 24 hours. Typical
results pertaining to the ultimate adsoprtion capability are illustrated in
Figures 1 and 2 where the residual pollutant concentrations after 24 hours of
treatment with specific amounts of carbon are plotted.
Various carbon dosages from a ratio of 1:1 to 100:1 carbon-to-pollutant con-
centration were considered experimentally. Starting with an initial phenol con-
centration of 1000 ppm, in Figure 1, it can be seen that a 1:1 dosage of carbon
to phenol reduces the concentration of phenol about 15%, whereas a 10:1 ratio of
carbon to phenol reduces the concentration 95%.
Comparisons of Figures 1 and 2 show that the relative efficiency in terms
of total mass of pollutant removed per unit mass of carbon is almost independent
of granularity of the carbon used. From Figure 2, it is apparent that for all
carbon-to-phenol dosages exceeding approximately 10:1, the residual concentration
of pollutant in solution is independent of initial concentrations. The mass of
pollutant removed, however, is substantially different. This is best illustrated
by replotting the data in the form shown in Figure 3, from which the importance
of initiation of the treatment before substantial dilution occurs is readily
apparent.
The early experiments showed that generally the ultimate capacity of all
of the activated carbon products tested was approximately the same, and that a
carbon-to-pollutant ratio of 10:1 is near the optimum for treatment of most
organic spills. Table 1 lists examples of a number of different classes of
organic components that have been treated with lOx dosages of activated carbon
and the percent removal achieved at that dosage. Depending on the nature of the
pollutant treated, removal ranged from 48% to 99%. In general, the adsorptive
capacity of carbon for an organic pollutant can be related to the solubility,
density, polarity, and structure of the organic compound.
17
-------
10,000
1000
i
E
100
CARBON TO
PHENOL 10:1
10
CARBON TO
PHENOL 2:1
INITIAL PHENOL -
500 ppm
INITIAL PHENOL
50 ppm
CARBON TO
PHENOL 1:1
100
PHENOL CONCENTRATION AFTER TREATMENT, ppm
1000
Figure 1 RESIDUAL PHENOL CONCENTRATIONS RESULTING FROM
POWDERED CARBON DOSAGES (AQUA NUCHAR A)
18
-------
100,000
10,000 -
I
111
C/9
o
a
tc
1000
100
>- ; 4-'.^->'.
0.3
1.0
10
RESIDUAL PHENOL (ppm)
100
1000
Figure 2 PHENOL REMOVAL BY 8 x 30 NUCHAR UNDER IDEAL CONDITIONS
-------
INITIAL PHENOL
CONCENTRATIONS
00 ppm
250 ppm
500 ppm
1000 ppm
DATA FOR AQUA NUCHAR
ACTIVATED CARBON
10 20 30 40 50 60 70 80
PERCENT OF PHENOL REMOVED
90 100
Figure 3
WEIGHT OF CARBON REQUIRED PER UNIT WEIGHT OF PHENOL
SPILLED AS A FUNCTION OF PERCENT REMOVAL
20
-------
TABLE 1
ADSORPTION BY ACTIVATED CARBON
CHEMICAL
PHENOL
ACETONE CYANOHYDRIN
METHANOL
ACRYLONITRILE
CHLORINE
PYRIDINE
ISOPRENE
BUTANOL
BENZALDEHYDE
BENZENE
XYLENE
STYRENE
INITIAL
CONCENTRAnON
(ppm)
1000
500
100
1000
200
100
1000
200
15
1000
100
1000
500
200
100
1000
500
1000
500
1000
500
100
1000
500
100
500
250
50
200
100
200
100
20
CARBON
DOSAGE
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
10X
RESIDUAL
CONCENTRATION
(ppm)
3
2
1
400
110
70
830
132
10
490
72
0.05
0.05
0.15
0.05
145
71
110
110
249
163
52
9
6
2
27
23
20
29
32
6
7
9
1
PERCENT
REMOVAL
99
99
99
60
45
30
17
33
33
51
28
99
99
99
99
86
86
89
78
75
67
48
99
99
98
95
91
60
86
68
97
93
55
21
-------
The effectiveness of activated carbon for removal of several inorganic
salts from solution has also been investigated. Results are presented in
Table 2. It is evident that very high carbon-to-pollutant dosages are required
for substantial removal of some soluble metal salts. Required dosages for 50%
removal range from lOOOx for Cr+6 to 5x for tig when initial metal ion con-
centrations are 100 ppm. It is important to note that mercury, a highly toxic
heavy metal, may be very effectively removed from solution by activated carbon.
It is also worth noting that lead in the acetate form is also readily removed
by this means.
The data in the previously presented figures and tables illustrate the
effectiveness of 24-hour treatments with activated carbon. In most cases, 60%
to 80% of the reduction of pollutant concentration occurred within the first
three hours of treatment; frequently the removal occurred more rapidly. A review
of available data confirmed the hypothesis that any treatment procedure which
increased contact between the pollutant and the activated carbon increased
adsorption rates.
The early experiments demonstrated that the total capacity of activated
carbon for many classes of pollutants and the rate of removal of pollutants
from water were both consistent with anticipated needs for cleaning spill-
contaminated water. Our attention was, therefore, directed toward the develop-
ment of practical methods for the dispersal of activated carbon in water and the
removal of the carbon from the stream or lake after adsorption was complete.
1. POWDERED CARBON. In the laboratory studies, it was shown that the most
efficient adsorption is accomplished by powdered carbon in polluted water that
is sufficiently turbulent to keep the powder in suspension. However, 1000 ppm
of powdered carbon reduced the transparent depth of water to less than 2 milli-
meters. One percent of that concentration still produced a totally unacceptable
turbidity. Addition of flocculents created a thick, unacceptable bottom sludge,
and was never more than 99% efficient, leaving turbid water. Bioassay experi-
ments with sludge formed of carbon previously used to remove phenol from water
showed some toxicity to fish (50% fatal to fathead minnows in 30 days) .
Froth flotation techniques in which compressed air is bubbled through the
water in order to float carbon particles to the surface were found to be less
efficient than flocculation. Fifty percent recovery of carbon was achieved only
with the addition of surface active agents at concentrations which would pose a
secondary pollution problem, sometimes equal to the first.
It was concluded, therefore, that free powdered carbon was not suitable for
use in natural water, except as a last resort or where the pollutant laden
powder would settle to the bottom where it could be located and removed.
2. THE CARBON TEABAG. While granular carbon adsorbs pollutants less rapidly
than powdered carbon, its coarse grain size permits greater flexibility in the
design of dispersal and retrieval techniques. Among the first concepts con-
sid^red was a carbon filled porous cloth "teabag." The intial design involved
a vertically suspended bag with flotation at its top and well ventilated pockets
of activated carbon extending tor the desired depth in the water.
22
-------
Table 2
CARBON ADSORPTION OF Mn, Cu, Ni, Cr, Hg AND Pb
HEAVY METAL
TESTED
+2
Mn^
Mn , n
«*2
Mn , «
Mn ,
Mn
Cu+2
Cu ,0
fj
CuTo
T*2
Cu
4-2
Ni+2
Ni.9
Ni,,
+2
NiT~
+y
Ni
4-6
Cr
Cr+6
Cr+6
Cr+6
Cr+6
Hg+2
Hgj.o
Hg+2
Hgj.o
Hg
Cr+3
Ci+3
Cr+3
Cr+3
Cr+3
+2
Pb.f
Pb 9
Fb
+2
pblo
T«_ * ^
Pb j
Pb ..
Fb
ACTIVATED CARBON
DOSAGE
(ppm)
0
500
1,000
5,000
10,000
0
500
1,000
5,000
10,000
0
500
1,000
5,000
10,000
0
500
1,000
5,000
10,000
0
500
1,000
5,000
10,000
0
500
1,000
5,000
10,000
0
500
1,000
10,000
0
500
1,000
5,000
10,000
RESIDUAL METAL
CONCENTRATION
(ppm)
100
99
97
75
50
50
46
45
13.5
1.8
100
96
95
89.5
48
100
84
74
66
64
100
1
1
1
1
100
95
92.5
82.5
52.5
120
49
43
10.5
107
93
88
17
7.5
PERCENT REMOVAL
0
1
3
25
50
0
8
10
73
96.4
0
4
5
10.5
52
0
16
26
34
36
0
99
99
99
99
0
5
7.5
17.5
47.5
0
59.1
64.2
91.2
0
13
17.7
84
93
TEST CHEMICAL
MnCl2
CuSO,
4
NiCl2
K2Cr2°7
HgCl
£
CrCl,
3
Pb(C<,H_0_)o
2322
Pb(NO )
J *
23
-------
Considerable design work and testing in stirred beakers resulted in data
that are typified by the curves shown in Figure 4. Obviously, the coarse mesh
bag with 20 x 50 mesh carbon is most effective, but carbon losses through even
fine mesh cloth forced abandonment of the small grain carbon in favor of the
more coarse "Nuchar" 8 x 30 carbon, for which losses were less than 1%. This
material was used in most of the channel and pool tests described later and
appears to be a suitable compromise for use in the field.
Before proceeding to these larger scale tests, a short series of experi-
ments was performed to determine the effect of natural contaminants in water on
the removal of spilled material from water. Phenol was used as the test pollutant,
As indicated in Figure 5, the variations in water temperature over a range
expected in nature have more effect on the rate and extent of adsorption than
the type of water involved.
The small-scale experiments described above were followed by a series of
tests of the teabag concept under conditions that more nearly simulate natural
environments expected in the field. One set of experiments, channel tests, was
performed in a 1000-liter, race-track shaped channel 28 feet long, 8 feet wide
and having a 1 x 1 foot cross section. Stream velocity through the channel
could be controlled between zero and 1.5 ft/sec and bottom roughness could be
adjusted by adjusting the distribution of stones placed on the channel bottom.
Several teabag configurations were tested with three different types of carbon.
Pertinent illustrative data are presented in Figure 6 and compared with standard
beaker tests. In all cases, the carbon-to-pollutant ratio was 10:1. The bags
were permitted to move freely with the channel water. Ventilation of water
through the bags was produced only by the shear and turbulence in the channel.
Comparison of channel Tests 1 and 2 illustrates the importance of bag
design and packing. In Test 1, the bags used were constructed with l^inch wide
vertical pockets fully packed with carbon. Without changing other experimental
conditions, Test 2 was performed using bags with 1-inch wide horizontal pockets
that were half filled with the same carbon. The loose packing permitted less
restricted flow of polluted water through the charcoal, which was free to move
inside the pockets. The horizontal configuration prevented packing of the char-
coal by gravity. Obviously, the Test 2 configuration is superior.
Comparison of Tests 2 and 3 shows that there is no significant effect on
removal due to the type of bag fiber provided the mesh size is unchanged and
sufficiently coarse to permit flow-through of the contaminated water.
Comparison of Tests 3 and 4 demonstrates the importance of stream velocity
in ventilating the carbon in bags. The increased turbulence due to the 1 ft/sec
flow velocity in Test 4 caused a significantly increased pollutant removal rate
over the rate experienced in Test 3, with a flow velocity of 1/2 ft/sec. This
effect is further illustrated by comparison of these results with the "Beaker
No Stir" results that are included in the figure. Comparison of channel Test 4
with the "Beaker, Bag Stirred" results suggests that data acquired in our
standard beaker experiments may be compared directly with the channel tests per-
formed at 1 ft/sec with loosely packed, horizontally compartmentalized bags.
24
-------
1000
900
800
700
600
!
ILTRASORB 300
MESH BAG
NUCHAR 20 x 50
FINE MESH BAG
DARCO 8 x 35....
COARSE MESH BAG
FILTRASORB 400
MESH BAG
NUCHAR 8 x 30
MESH
NUCHAR
-20x50
MEDIUM
JBAG
NUCHAR 20 x 50 COARSE MESH BAG
80 100
MINUTES
180
Figure 4 BAG TESTS WITH VARIOUS GRANULAR CARBONS
25
-------
500
to
100
Ul
£
Q
5»
IU
K
o
D
A
DISTILLED WATER, ROOM TEMP.
TAP WATER, ROOM TEMP.
CATTARAUGUS CREEK WATER, ROOM TEMP.
DISTILLED WATER 36-40°F
I
3 4
TIME - HOURS
6
Figure 5 TEA BAG GRANULAR CARBON ADSORPTION OF PHENOL
FROM DISTILLED WATER, TAP WATER, AND CREEK WATER
-------
250 r
TIME IN HOURS
9 10 11
Figure 6 CARBON TEA BAG EFFECTIVENESS UNDER
A VARIETY OF CONDITIONS
27
-------
Channel Test 7 was performed with freely dispersed powdered carbon (Aqua
Nuchar A) to provide a baseline against which all other channel tests may be
compared.
With the importance of teabag ventilation thoroughly established in these
experiments, some concern developed as to the potential value of the concept for
treatment of spills that occur in still water such as ponds, lakes, or slow moving
streams. A final series of experiments was therefore performed in an outdoor
swimming pool, 12 feet in diameter and 3 feet deep filled with 8,000 liters of
water. In these tests, we sought to determine if motion associated with waves
could provide sufficient ventilation to produce acceptable pollution removal
rates. The data are presented in Figure 7 and compared with results of a
preliminary wave test performed in the channel.
It is apparent from the "still water" tests that even a small amount of
agitation is effective in increasing pollution removal rate. While totally
inadequate for effective treatment of spills, the adsorption rate during the
first eight hours of the experiment, a period of moderate breeze, was signifi-
cantly greater than that of the next eight hours during the nocturnal calm.
A truly significant increase in adsorption rate resulted from the artificial
generation of 2 cm high waves, 40 cm long in the second pool experiment.
Similar results were obtained with small artificial waves in the channel under
conditions of zero flow. These tests also demonstrated the need for mooring
the bags to prevent aggregation and drift due to wave action. A nylon fishing
line with fish hooks attached at 1-foot intervals was sufficient to maintain
proper separation and position.
This suggests that natural wave action in a lake should be adequate for
ventilation of carbon filled teabags. Under very calm water conditions, the
waves produced in the wake of a few small outboard motor boats could be used to
provide sufficient agitation to permit removal of a significant fraction of
spilled material from solution in a few hours.
Schematic drawings of the teabags used in the channel and pool tests are
presented in Figures 8 and 9, respectively. The bags were constructed of nylon
with thirty threads to the inch to provide a pore size of approximately 0.5 mm.
In both cases, each horizontal compartment was filled to one-half capacity with
Nuchar 8 x 30 activated carbon to produce a mean loading weight of 30g and SOOg
per bag, respectively. Either of these configurations could be useful for treat-
ment of spills in natural water, but neither configuration seems applicable for
a sufficiently broad class of water bodies.
A design of more general applicability would consist of an arbitrarily long
bag (20 to 100 feet) consisting only of carbon-containing compartments similar
to those illustrated in Figure 9. Separate flotation units which could be
readily attached to the bag at any desired location could be provided to permit
adjustment of the depth to which the bag would hang to accommodate the depth of
the water body and the nature of the spill. For example, in a 5-foot deep
stream, flotation units could be attached every eight feet to permit the bag to
hang to the 4-foot level, and just clear the bottom. If the spilled material were
28
-------
300
to
vo
250
200 hi-
s
O
UJ
a.
150
STILL WATER
t- -}- -t--
WAVES IN POOL
2 cm HIGH
40cm LONG
WAVES IN CHANNEL
2 TO 5 cm HIGH
04 8 12 16 20 24 28 32 36 40 44 43 52 56 60 64 68 72 76 80
TIME IN HOURS
100
Figure 7 REMOVAL OF PHENOL BY ACTIVATED CARBON (8 x 30 MESH NUCHAR)
IN POOL TESTS
-------
CO
o
7"
.
6"
r
1"
1
POLYURETHANE
' FOAM FLOTATION
GRANULAR
CARBON
30 g TOTAL
Figure 8 TEA BAG DESIGN, CHANNEL TESTS 2 AND 3
-------
1
2
1
i
4"
I
3" FLOTATION
i *
-1.5"
14
LOOSELY FILLED
CARBON
COMPARTMENTS
14
LOOSELY FILLED
CARBON
COMPARTMENTS
Figure 9 TEA BAG CONFIGURATION FOR POOL TEST NUMBER 1
31
-------
more dense than water and likely to be dissolving from a pool on the stream
bottom, the flotation units could be attached at, for example, 15-foot
intervals, so that most of the carbon would remain at lower levels where con-
centration is greatest. Attachment of the units at 2-foot intervals could
accommodate spills in shallow water or of materials that are less dense than
water and going into solution from floating pools of the concentrated pollutant.
While it was impossible to test the effectiveness of this concept in treating
a spill or simulated spilled hazardous material, two 10-foot long bags were
constructed to evaluate the mechanics of the idea. Tests in the Buffalo harbor
indicated that the desired flexibility can be achieved and that the amount of
agitation desired from wave action, either natural or artificially produced in
the boat wake, would produce the required ventilation.
3. CARBON FIBERS. Recently, several different types of activated carbon filters
have become available in small experimental quantities from the Carborundum
Company. Beaker tests similar to those for which data are presented in Figure 4
were performed with a variety of these materials to determine their potential for
removal of pollutants from water. One variety which resembles loosely packed,
fine grain steel wool in appearance shows excellent potential. Test data for
this material are compared with channel tests data obtained with freely dispersed
powdered carbon in Figure 10. In view of previously discussed results (Figure 6,
channel Test 4 compared to Beaker Test, Stirred), we do not consider the dif-
ference in test procedure used to be significant
It is apparent from Figure 10 that to within experimental accuracy the carbon
fibers are as effective as powdered carbon, both in removal capacity for the
test pollutant and in the rate of removal of the pollutant from water.
The wool-like fibers in the tested samples have a density very nearly equal
to the density of water. The structural strength of the strongest samples was
adequate to permit compression of the matrix for storage and shipment. When
placed in water, the matrix expanded almost to its initial configuration and
floated with the uppermost fibers at the water surface. Ventilation of the
loosely packed fibers, therefore, was excellent. Agitation of this sample did
not produce significant fiber fracture. Besides normal stirring, the wool was
repeatedly lifted from the beaker with a spatula and replaced without excessive
fragmentation. In the field, the material could be readily removed from water
using a coarse-mesh net or perhaps a grappling hook.
This steel-wool-like material, therefore, appears to have many of the
properties of activated carbon desirable for field use. Other types of carbon
fibers tested were either too frangible for consideration or less effective
chemically. Since the samples were produced by manual processing on an experi-
mental basis, the variability is quite understandable. Further experimentation
and development should be encouraged.
4. OTHER PACKAGING CONCEPTS FOR ACTIVATED CARBON. Two additional concepts
were investigated for packaging activated carbon in such a way that it might be
retrieved from natural water after treatment was completed. With the aid of
the Carborundum Company, activated carbon was bonded to waterproof paper using
32
-------
250
200
150 -
1
-i
O
z
ui
CL
100
01234
TIME IN HOURS
Figure 10 COMPARISON OF CARBON FIBERS WITH
POWDERED ACTIVATED CARBON
33
-------
techniques designed for sandpaper manufacture. Laboratory tests showed that
the product was in general less efficient than the carbon filled teabag,
primarily because the bonding agent occupied some of the available adsorption
sites and restricted access of the pollutant to approximately half of the
exterior surface of each carbon grain. Similar results were achieved in in-house
attempts to coat reticulated foam with granular carbon.
5. COST AND LOGISTICS. Since the large-scale laboratory and outdoor experi-
ments were not completed until late in this program, extensive analysis of cost
and logistic requirements for implementing the teabag concept for activated
carbon treatments was not possible. Furthermore, the potential advantages of
using the newly developed activated carbon fibers suggest that a vastly superior
treatment may become available in the near future. This analysis was, therefore,
intended only to determine whether the use of the concept of properly packaged
activated carbon showed enough promise of practicability to warrant further
development and testing.
The analysis was based on the assumption that 20,000 pounds of properly
packaged carbon would be stored at each of 200 locations throughout the United
States, and that 100,000 pounds of additional carbon would be stored at each
of five Air Force Bases or commercial airports, ready for shipment to any spot
in the nation. The purpose of this distribution would be to provide adequate
supplies of carbon at "local" warehouses to permit rapid treatment of small
spills; i.e., spills of up to 2,000 pounds of material, in a time frame which
is consistent with the rate at which the concentration of spilled material is
reduced to levels that are too low for effective treatment.
The local supplies would also be available for initiation of the treatment
of large spills at about the maximum rate which could be delivered in a very
rapid response situation. Even if the 200 caches of carbon were uniformly
distributed throughout the contiguous states, there would be 100,000 pounds of
material within 120 miles of any spill. The time required to make the decision
to treat with carbon, design a specific treatment procedure, and implement the
treatment procedure is estimated to be comparable to, or greater than, the time
required to transport the carbon 120 miles over land to the spill site
(assuming 1 hour for loading, 1 hour for transport of the closest carbon, and
3 hours for transport of the remaining 80,000 Ibs from nearby caches).
It was assumed that the first 50 tons of carbon to arrive at a large spill
would be dispersed from ten small boats supplied by local agencies (or by local
sportsmen to protect their waters). An 18-foot runabout could carry 1,000
pounds of material safely with a crew of three for dispensing the carbon.
Assuming 20 minutes for each trip, the 10 boats could disperse the carbon avail-
able through ground transportation within four hours after shipment is initiated.
The first air shipment of carbon could be available at most sites within that
period.
It seems reasonable from a logistics standpoint, therefore, to consider
stockpiling 5 million pounds of activated carbon. At the projected price of
$2.00/lb, for the carbon, the basic material cost would be $10 million.
Packaging might increase this number by 10%.
34
-------
We have found during the project that granulated activated carbon could
be procured and packaged in hand-made teabags for slightly under $1.50/lb.
Substantial savings could certainly be effected through automated procedures
that could be implemented to fill a large procurement order. The teabag
approach to carbon treatment could certainly be implemented, therefore, for
under $10 million.
35
-------
SECTION 6
TREATMENT OF WATER SPILLS WITH ION EXCHANGE RESINS
1. HEAVY METALS. It was shown previously that activated carbon is not
effective for the treatment of most heavy metal spills in water. Therefore,
laboratory investigations were conducted to assess the relative effectiveness
of ion exchange resins (I.E. resins) for the treatment of water spills as
compared to activated carbon. Ion exchange resins were investigated for
removal of As+^, Cr+^, and Cr . Tests were conducted in which ion exchange
resin was freely dispersed in water or contained in "tea bags" to facilitate
recovery of resin with exchanged pollutants. Parallel tests were run using
powdered carbon as the treatment agent to compare effectiveness of carbon
and I.E. resin.
The exchange resins used for the experiments were a mixture of cationic
and anionic resins normally used for demineralizing of water (Corning No.
3508A). "Tea bags" for holding the resins were made from dacron polyester
fabric having a mesh size of 0.4 mm x 0.5 mm.
Test solutions were prepared using sodium arsenate (Na2HAsO. . 7H_0),
chromium chloride (CrCl, . 6H_0), and potassium dichromate (K2Cr20_).
Five hundred milliliters tff eacn test solution and one bag containing I.E.
resin were placed in each of three beakers and mixed with a magnetic stirrer.
Periodically, samples of the solutions were withdrawn and analyzed for resi-
dual chromium or arsenic concentration. The results are summarized in Tables
3, 4 and 5.
By comparing removals of arsenic, trivalent and hexavalent chromium by
ion exchange resins to removal by activated carbons, it is seen that ion
exchange resins, as expected, are much more effective. This is so even when
powdered carbon is used and applied loosely, rather than in bags as shown
by the chromium data in Tables 4 and 5. For a contact period of one hour
using ion exchange resin in tea bags, As , Cr+3^ and Cr+6 removals were
85%, 91.5%, and 97%, respectively. By comparison, removal of As+5 by 8 x 30
mesh granular carbon in a tea bag was 5.6% in one hour and removals of Cr+3
and Cr in one hour using loose powdered carbons were 47.5% and 36%, respec-
tively. Dosages were the same for all treatments (i.e., 10 g/1).
Table 4 indicates the effect which placement of the resins in tea bags
has on rate of pollutant removal as compared to free dispersion of the resin
in water. Freely dispersed resin resulted in 90% removal of Cr+6 in 10 min-
utes, whereas resin contained in the bag required 40 minutes for an approxi-
mately comparable removal of 88%. Containment in bags is desirable, of course,
to facilitate removal of the pollutant from the water body after it has been
exchanged onto the resin.
36
-------
Table 3
REMOVAL OF ARSENIC BY ION EXCHANGE RESINS
AND ACTIVATED CARBON IN "TEA BAGS"
TARGET
POLLUTANT
ARSENIC
Na.HAsO.
2 4
TYPE, DOSAGE;
AND CONDITION
OF CHEMICAL
AGENT USED
ION EXCHANGE
RESIN (CORNING
NO. 350 8A) IN A
DACRON
POLYESTER BAG;
DOSAGE = 10 g/1
GRANULAR CARBON
(NUCHAR 8 X 30) IN
A DACRON
POLYESTER BAG;
DOSAGE = 10 g/1
ION EXCHANGE RESIN
(CORNING NO. 3508A)
AND GRANULAR CARBON
(NUCHAR 8 X 30) IN A
DACRON POLYESTER
BAG; RESIN DOSAGE =
5 g/1; CARBON
DOSAGE = 5 g/1
CONTACT
TIME
Minutes
0
5
30
60
240
1440
0
5
30
60
240
1440
0
5
30
60
RESIDUAL
CONCENTRATION
OF POLLUTANT
mg/1
AS ARSENIC
125
100
73
19
~ 0
~ 0
125
125
122
118
105
97
125
120
86
30
PERCENT
REMOVAL
0
20
42
85
~100
~ 100
0
0
2.4
5.6
16.0
22.4
0
4
31.2
76.0
37
-------
Table 4
REMOVAL OF HEXAVALENT CHROMIUM
BY ION EXCHANGE RESINS AND ACTIVATED CARBON
TARGET
POLLUTANT
HEXAVALENT
CHROMIUM
K0Cr.O_
9 97
<£* £ /
TYPE, DOSAGE;
AND CONDITION
OF CHEMICAL
AGENT USED
ION EXCHANGE
RESIN (CORNING
NO. 3508 A) IN A
DACRON POLY-
ESTER BAG;
DOSAGE - 10 g/1
ION EXCHANGE
RESIN (CORNING
NO. 3508 A);
LOOSE;
DOSAGE = 10 g/1
POWDERED CARBON
(NUCHAR C-190N);
LOOSE:
DOSAGE = 10 g/1
CONTACT
TIME
Minutes
0
6
10
40
60
120
0
5
10
15
20
30
45
60
0
60
w
RESIDUAL
CONCENTRATION
OF POLLUTANT
mg/1
AS CHROMIUM
100
68
48
12
3
0.3
100
27
10
4.5
1.5
0.8
0.4 p
0.3
100
64
\J^T
PERCENT
REMOVAL
0
32
52
88
97
99.7
0
73
90
95.5
98.5
99.2
99.6
99.7
0
36
38
-------
Table 5
REMOVAL OF TRIVALENT CHROMIUM
BY ION EXCHANGE RESINS AND ACTIVATED CARBON
TARGET
POLLUTANT
TRIVALENT
CHROMIUM
3
TYPE, DOSAGE;
AND CONDITION
OF CHEMICAL
AGENT USED
ION EXCHANGE
RESIN (CORNING
NO. 3508A) IN A
DACRON POLY-
ESTER BAG;
DOSAGE = 10 g/1
POWDERED CARBON
(NUCHAR C-190N);
LOOSE;
DOSAGE = 10 g/1
CONTACT
TIME
Minutes
0
6
10
40
60
120
0
60
RESIDUAL
CONCENTRATION
OF POLLUTANT
mg/1
AS CHROMIUM
100
65
48.5
16
8.5
0.5
100
52.5
PERCENT
REMOVAL
0
35
51.5
84
91.5
99.5
0
47.5
39
-------
Treatment of the metals as elemental forms with ion exchange resin required
resin to pollutant ratios of 100:1. However, the ion exchange resins used in
these experiments consisted of a mixture of both cation and anion I.E. resins.
The pollutants (As*15, Cr+3, Cr+6) are all cations. Thus, it is expected that
the same percentage removals of metal ions could be achieved with perhaps half
as much cation I.E. resin used alone, or a 50:1 treatment to pollutant ratio.
Because carbon treatment was the principal method of spill treatment
studied in this project, available resources did not allow for further evalua-
tion of ion exchange "tea bags" in channel or pool tests. In view of experi-
mentation which demonstrated greater removal of some metals with ion exchange
resins, further evaluation and development of spill treatment with exchange
resins is recommended. An important advantage of resins is that they are
easily regenerated for reuse.
2. PHENOL. Phenol removal by both cationic and anionic exchange resins
was evaluated, in laboratory beaker tests and compared to removal with acti-
vated carbon. Exchange resins used were AMBERLITE IR-120-H, a cation exchanger,
and AMBERLITE IRA 400-OH, an anionic exchanger. These resins are manufactured
by the Rohm and Haas Company.
All treatment agents were applied in loose form at a dosage of 10 g/1.
Initial phenol concentrations for all tests were 1000 mg/1. Agitation was
provided by magnetic stirring. From the results of these tests, tabulated in
Table 6, it is seen that the cation exchange resin does not remove phenol.
However, anionic exchange resin, applied loosely, removed a greater amount
of phenol than powdered carbon, but does not remove it quite as fast. After
five minutes contact, removals of phenol by anion exchanger and powdered
carbon were 76% and 97%, respectively. After ten minutes contact, the resin
removed almost as much as the powdered carbon. After ten minutes contact,
the powdered carbon approached its ultimate capacity for phenol whereas the
anion exchange resin continued to remove phenol to the conclusion of the
experiment (85 minutes). Granular carbon attained nearly the same removal
of phenol as powdered carbon but required nearly an hour more contact time.
The dosage of anion exchange resin as a ratio to initial pollutant con-
concentration in this test was 10:1. On a performance basis then, it would
be beneficial to further evaluate anion exchange resins for removal of phenol
and substituted phenols which behave similarly in water. Phenol in water
behaves as a weak acid and dissociates:
/
H+
40
-------
Table 6
REMOVAL OF PHENOL BY ION EXCHANGE
RESINS AND ACTIVATED CARBON
TYPE DOSAGE;
AND CONDITION
OF CHEMICAL
AGENT USED
CATION EXCHANGER
(AMBERLITE IR-120-H
TECH. ) ;
LOOSE;
DOSAGE = 10 g/1
ANION EXCHANGER
(AMBERLITE IRA-
400-OH) ;
LOOSE;
DOSAGE = 10 g/1
POWDERED CARBON
(NUCHAR C-190N);
LOOSE
DOSAGE = 10 g/1
GRANULAR CARBON
(NUCHAR 20 X 50)
LOOSE;
DOSAGE = 10 g/1
CONTACT
TIME
Minutes
0
900
0
1
5
10
60
85
0
5
10
30
60
120
180
0
5
10
30
60
120
180
RESIDUAL
CONCENTRATION
OF POLLUTANT
mg/1
1000
1000
1000
660
240
70
40
6
1000
57.5
56
54
52
50
48
1000
250
140
87.5
65
45
42
PERCENT
REMOVAL
0
0
0
34
76
93
96
99.4
0
94.3
94.4
94.6
94.8
95
95.2
0
75
86
91.3
93.5
95.5
95.8
41
-------
The negatively charged ring is exchanged on the anion exchange resin. In
a similar manner, a large number of organic amines which are weak bases could
be exchanged on cation exchange resins. The dissociation of two widely used
organic amine dyes in water- aniline and pyridine, are:
ANILINE
NH2+ H2O
Wl + H.
'
PYRIDINE
N-HT+OH"
The positively charged rings might be exchanged on cation exchange resins,
Available resources did not permit evaluation of ion exchange resins in "tea
bags" for organic pollutant removal either in laboratory tests or channel
tests. The carload or truckload costs for the anion and cation exchange
resins used were $.34/lb and $0.28/lb, respectively. Cost of the 8 x 30
granular carbons used in this research is $0.24/lb while powdered carbon
is $0.08/lb. Continued investigation of exchange resins for the treatment
of hazardous pollutant spills in water is recommended.
42
-------
SECTION 7
TREATMENT OF SPILLS OF ACIDS AND BASES BY
DIRECT NEUTRALIZATION
Among the hazardous materials singled out for special attention during
this project were ammonia, which upon immersion in water forms ammonium
hydroxide and chlorine, which reacts with water to form hypochlorous and hydro-
chloric acids. (The presence of activated carbon which might be added as
a countermeasure to such spills in water causes the above reactions to proceed
more rapidly and completely to the formation of their end products.)
In addition to these, many other compounds are either acidic or basic in
water solutions. These materials may be treated simply by applying stoichio-
metric amounts of neutralizing weak acids or bases, as appropriate. These
less completely ionized agents prevent risking the severe secondary danger of
creating adverse pH's by even minor over-treatment or lack of coincidence of
the deployed countermeasure with the original spill plume.
Neutralization of alkaline pollutants within two minutes by bubbling of
carbon dioxide through the water was demonstrated. Alternatively, neutraliza-
tion may be achieved immediately upon mixing with acetic acid, a weak acid.
It was concluded, however, that the use of carbon dioxide gas will not be cost
effective since secondary means of diffusing the gas into the water must be
developed to prevent the loss of large quantities of the neutralizing gas to
the atmosphere.
Similarly, acidic solutions (especially those containing hydrochloric
acid from the interaction of molecular chlorine and water) were readily adjusted
to neutral pH with treatments of weak basic substances such as sodium carbonate
(washing soda). Neutralizations of acid spills have also been achieved with
proper dosages of limewater and aluminum hydroxide.
It was important to determine if these simple neutralization schemes would
be applicable in the natural waters which generally are buffered by dissolved
substances and biological products at pH's of 7 and above. Using water supplies
from different sources (including well water - out-gassed distilled water, water
from Cattaraugus Creek, and strongly organic-polluted water from the Buffalo
River) which ranged in initial pH from pH 7 through pH 8.3, the above-stated
conclusions were confirmed experimentally. In no instance was it found that
significant deviations from simple stoichiometric neutralization occurred.
With weak acid and weak base countermeasures, a very large region of pH
"forgiveness" was found, without apparent influence from the initial pH of
the water into which the spill occurred or of the presence or absence of
adventitious buffering substances in the initial water supply. Thus, the
use of weak acid and weak base countermeasures to their appropriate base and
acid spills, respectively, can be recommended.
43
-------
As is true for any large spill, it will be difficult to estimate the
exact amount of spilled material that actually reaches the water or the
distribution of that material in the water. With large spills of both
acidic and basic materials, it seems particularly appropriate to utilize
an automatic treatment device of the type recommended on page 49 of this
report. The neutralization of acidic or basic materials spilled directly
onto the ground was studied during the large scale demonstration phase at
Calspan's Bethany, New York site. This experiment in which concentrated
sulfuric acid was neutralized with lime is described in Section 9.
44
-------
SECTION 8
PRECIPITATION OF HEAVY METALS WITH SODIUM SULFIDE
The treatment of spills of soluble heavy metal compounds, by precipitation
of the generally insoluble metal sulfides, was evaluated as part of the project,
The technique is potentially useful for many materials and was found to be
effective in treating solutions of most of the heavy metal compounds which
present a spill hazard. It was shown to be simple to apply and capable of
reducing the toxic hazard presented by heavy metal ions in solution.
The treatment is applied by introducing sulfide ions, derived from a
solution of sodium sulfide, into the spill. At concentrations of the heavy
metal ion that present a toxic hazard, formation of the sulfide precipitate
proceeds rapidly, and toxicity is reduced within seconds. The treatment
chemical is stabilized by the addition of some sodium hydroxide, in order
to prevent evolution of toxic hydrogen sulfide fumes.
In the case of small spills, that are typical of the majority of heavy
metal spills reported in the past two years, treatment can be applied on the
basis of a simple test: A small amount of sodium sulfide (Na2S) is injected
at the spill site. If a visible precipitate forms, additional treatment
material is applied until about one-half gallon (4 liters per kilogram) of
concentrated Na^S solution is used for every pound of spilled material. For
larger spills this process may result in inadequate treatment and more exact
procedures are advocated.
A wide range of heavy metal compounds can be treated by the sulfide method.
Since several thousand compounds should be considered, a detailed listing of
these compounds of the metals indicated by a slash mark in Table 7 are sub-
ject to sulfide treatment because the sulfide is their least soluble compound
whenever excess sulfide ions are available. To demonstrate the effectiveness
of sulfide treatment, soluble compounds of the commercially important metals
listed in Table 8 were investigated. Sulfide treatment was found to be un-
suitable for spills of two groups of heavy metals. (For the sake of brevity,
the term "metal" is used for the ions generated by solution of a compound in
the following. Bulk elemental heavy metals, if spilled, can usually be
recovered before they present a noticeable toxic hazard to waterways.)
In the first instance there are those which, like titanium and aluminum,
have greater affinity with oxygen than sulfur and whose reactive compounds
hydrolyze in the presence of sulfide ion to form hydroxides and oxides. These
often require no treatment, as they are quickly precipitated by mere contact
with natural water. A second problem arises in the treatment of high oxidation
states, chromium, molybdenum, manganese, vanadium, and tungsten, that cause
the metal to be presented in anionic species, e.g., as the chromate, molybdate,
or permanganate, etc. Sodium sulfide is often effective in reducing such
species to a lower valence from which precipitation, usually as a hydroxide,
45
-------
H
Li
Na
K
Rb
Cs
Fr
Be
Mg
Ca
Sr
1
Ra
Table 7
PERIODIC TABLE OF THE ELEMENTS
LEGEND
SUBJECT TO SULFIDE
TREATMENT
SUBJECT TO RELATED
TREATMENT
HEAVY METALS
Sc
La
Ac
Ti
Hf
104
Nb
105
Mo
|W
Tc
Ru
W^
Co
Rh
,
Ir
Ni
v'xx <
Au
Zn
Cd
M
In
TI
Si
Ge
Sn
N
Se '
Te
Br
At
RARE EARTH METALS
ACTINIDES
He
Ne
Ar
Kr
Xe
Rn
. Th 7
$$$
Pa
5|r
fr :l;
Pm
Np
Sm
Pu
Eu
Am
Gd
Cm
Tb
Bk
Dy
Cf
Ho
Et
Er
Fm
Tm
Md
Yb
No
Lu
Lr
-------
Table 8
QUANTITY OF STANDARD SULFIDE TREATMENT
SOLUTION REQUIRED TO TREAT ONE POUND SPILLS
OF INDICATED COMPOUNDS
COMPOUND
SILVER NITRATE
CADMIUM CHLORIDE
COBALT CHLORIDE
COBALT CHLORIDE
COPPER SULFATE
COPPER SULFATE
COPPER SULFATE
IRON SULFATE
IRON SULFATE
IRON SULFATE
IRON CHLORIDE
IRON CHLORIDE
MERCURY CHLORIDE
MANGANESE SULFATE
MANGANESE SULFATE
MANGANESE SULFATE
MANGANESE SULFATE
MANGANESE SULFATE
MANGANESE SULFATE
MANGANESE SULFATE
FORMULA
AgN03
CdC12
CoCl2
CoCl9 6H90
£ £
Cu2S04
CuS04
CuSO. 5H.O
4 2
FeSOj. R,0
H <6
FeS04 - 4H20
FeS04 - 7H20
FeCl3
FeCl3 6H20
HgCl2
MnSO.
4
MnS04 H20
MnSO. 2H00
4 2
MnS04 3H20
MnS04 4H20
MnSO. - 5H.O
4 2
MnS04 6H20
TREATMENT
0.270 gal
0.500 gal
0.706 gal
0.385 gal
0.411 gal
0.575 gal
0.367 gal
0.540 gal
0.409 gal
0.330 gal
0.848 gal
0.509 gal
0.338 gal
0.607 gal
0.543 gal
0.490 gal
0.447 gal
0.411 gal
0.380 gal
0.354 gal
0.851 Ib Na2S and 0
,040 Ib NaOH per gallon
47
-------
Table 8(Cont.)
COMPOUND
MANGANESE SULFATE
NICKEL SULFATE
NICKEL SULFATE
NICKEL SULFATE
LEAD ACETATE
LEAD ACETATE
ZINC CHLORIDE
ZINC SULFATE
FORMULA
MnS04
NiS04
NiSO,
4
NiS04
Pb(C2l
Pb(C2]
ZnCl2
ZnS04
7H.O
' 6H20
7H20
13°2)2 * 3H2°
MO, IOHOO
J f. & t,
7H2°
TREATMENT
0.331 gal
0.593 gal
0.349 gal
0.326 gal
0.242 gal
0.181 gal
0.673 gal
0.319 gal
48
-------
is possible. This process is often accompanied by the evolution of hydrogen
sulfide and is not recommended. Alternate methods for the treatment of chro-
mate and dichromate spills, which are quite common, have been investigated.
Barium compounds are produced in quantity, and although barium is not subject
to sulfide treatment, its fate in case of spills was considered. The carbon-
ate and particularly the sulfate of barium are sufficiently insoluble that
spills of moderate quantities of soluble barium compounds should quickly
precipitate in natural bodies of water containing their salts.
The correct treatment of a metal spill requires application of sulfide
treatment in a definite (stoichiometric) ratio that assures that each metal
ion combines with an appropriate number of sulfide ions to yield the insoluble
sulfide. Most of the commercial metals are bivalent, as is sulfur, and each
metal ion combines with one sulfide ion. It is also found that the atomic
weights of metals treatable by the sulfide method occur in groups centered
on 60, 105, and 200, and that the lowest group (ranging from 55 for Mn to
65 for Zn) contains the commercially important elements manganese, iron,
cobalt, nickel, copper, and zinc. Each pound of the ions of these metals
requires about 1.3 pounds of sodium sulfide for complete conversion to the
insoluble sulfide. Unfortunately, the metals are encountered as, compounds
with various anions and with water of hydration. Because of these, the ratio
of sodium sulfide treatment, in the form of a concentrated (at 0°C) solution
needed to treat a one-pound spill, varies from 0.181 gallons for lead acetate
to 0.848 gallons for ferric chloride.
Some method of estimating the quantity of treatment to be applied to a
spill is needed when the amount or kind of material spilled is not known.
The use of specific ion electrodes for the metals or for the sulfide ion
itself was investigated in a spill simulation facility, using iron as a
heavy metal. The flow conditions dictated a frequency of sulfide additions
of- at least one every ten seconds. The required operations (reading the
specific ion meter, determining sulfide addition, and making the addition)
proved too complex for all manual operation on this time scale, and the
treatment achieved was far from ideal. The iron concentration was nonetheless
substantially reduced. Residual iron in this case was estimated to be less
than 16 mg out of 520 mg added in the original spill. This is to be compared
to a residue of 7 mg out of 520 mg when the exactly stoichiometric treatment
was applied. The use of specific ion electrodes for the metals, the cyclic
colorimeter, or of a titration and a sulfide electrode, to control localized
sulfide additions should be studied further. Problems to be solved in achiev-
ing effective treatment center primarily on response rate limitations of the
detector-human-treatment system and may best be alleviated by the development
of a more highly automated approach.
In order to evaluate automated spill treatment procedures, an apparatus
whose general features are shown in Figure 11 could be used. The device could
be used for sulfide treatment of heavy metal spills as well as for acid-base
neutralization which follows similar rules. It utilizes direct control of
treatment and feedback to compensate for errors in measurement and would per-
mit design of a stable treatment control system.
49
-------
CONTROL COMPUTER
tn
O
TREATMENT
STORAGE
FEEDBACK
SENSOR
POLLUTANT
SENSOR
WATER INTAKE
(FOR DILUTION)
TREATMENT DISPENSER
Figure 11 LARGE SPILL TREATMENT BARGE
-------
Sulfide precipitation has been shown to be effective in significantly
reducing the levels of heavy metal ions in solution and thus reducing the
toxic hazard of a heavy metal spill. In the process, however, sodium sulfide,
a material which is itself toxic, is introduced into the environment. It is
important therefore to assess the hazards posed by this material and to com-
pare them to the benefits to be derived from its use. The preparation of a
completely satisfactory comparison between the hazards and benefits of the
sulfide precipitation treatment of heavy metal spills would require a signi-
ficantly greater body of data than that available at this time. Ideally the
spreading of plumes of heavy metals and sodium sulfide over a range of stream
conditions should be correlated with complete toxicity-concentration-exposure
time data for these materials to compare the environmental damage caused by
untreated spills, treated spills, and erroneous treatment where there was no
spill. In this study a much less rigorous approach has, by necessity, been
taken. One can make a rudimentary comparison between the consequences of not
treating a heavy metal spill and the consequences of applying sodium sulfide
treatment where no spill had occurred. The comparison is made in terms of
the volume of water contaminated at a level lethal to fish. Despite the crude
nature of these comparisons, it is felt that the differences between the metal
spill results and the sulfide "spill" results are sufficiently great to support
the conclusion that sulfide treatment should be recommended for heavy metal
spills, with some reservations, as outlined below.
The hazard involved in treating a nonexistent spill or of mistreating a
spill clearly depends on the amount of material released to the environment.
Experience has shown that heavy metal spills typically involve small quantities,
of the order of a few kilograms. For comparison of effects, a spill of 4 kg
of each heavy metal compound for which toxicity data corresponds to data avail-
able for the sulfide is taken as an example. The volume of water which could
be contaminated to the reported lethal level by such a spill was calculated.
A comparison of these two volumes (Table 9) corresponding to no treatment
and to the worst possible treatment action provides a measure of the benefits
and hazards of sulfide treatment.
The ratios of the affected volumes vary widely, even with different data
for the same heavy metal, but in no case is the sodium sulfide toxic in more
than 5% of the volume calculated as lethal for the heavy metal. Further, it
may be seen that the affected volume is quite small for all the sodium sulfide
additions, even in the case of ZnSO treatment where the greatest amount is
added, and the sulfide toxicity corresponds only to 10% fish mortality in
seven days. The hazard occasioned by sodium sulfide is further alleviated by
the instability of the sulfide ions in the environment which was not accurately
reproduced in the laboratory toxicity tests where contact with water of lower
pH than that of the Na2S solution did not occur.
The acute toxicity of sodium sulfide solutions in various concentrations
was measured and compared with the toxicity of sodium hydroxide. Both sodium
sulfide and hydroxide raise the pH of their aqueous solutions and this alone
may occasion some toxicity. Figure 12 shows 24-hour toxicity versus concen-
tration of these compounds (in mg of compound per liter) while Figure 13 shows
51
-------
Table 9
COMPARISON OF HEAVY METAL AND Na S HAZARDS
HEAVY METAL
COMPOUND
cdci2
cdci2
cdci2
CuS04
HgCl2
HgCl2
Ni(N03)2
Pb(N03)2
ZnSO,
ZnSO^
AMOUNT Na2S
TO TREAT 4 kg
1.75 kg
1.75
1.75
1.25
1.20
1.20
1.15
0.9
2.0
2.0
WATER CONTAMINATED
BY HEAVY METAL
150,000 m3
400
240
3,300
300
30,000
800
12,000
13,000
8,000
WATER CONTAMINATED
BY Na2S
22 m3
18
12
30
9
10
30
25
50
50
52
-------
%OF
FISH
KILLED
100 r-
80
60
40
20
O
NaOH
(EMERALD
SHINERS)
(MINNOWS)
CONTROL
10 ~ 100
CONCENTRATION (ppm)
1000
Figure 12 TOXICITY OF
AND NaOH (24 HOUR EXPOSURE)
%OF
FISH
KILLED
10.0
11.0
12.0
Figure 13 TOXICITY OF Na^ AND NaOH AS FUNCTIONS OF pH
OF THEIR SOLUTIONS (24 HOUR EXPOSURE)
53
-------
the same toxicity as a function of pH. The pH is closely related to the molar
concentration of the compounds and Figure 16 indicates that there is relatively
little difference between the molar toxicity of sodium sulfide and hydroxide.
Many pollutants ultimately leave the environment by conversion to useful
or at least acceptable chemicals. For example, both chemical and biological
processes tend to oxidize most organics to water and carbon dioxide. Metals
differ in this respect; those metals whose ions are toxic will continue to
pose a threat indefinitely if they are not rendered insoluble and stored in
an environment which inhibits redissolution. The sulfide precipitation treat-
ment was shown to be effective in rendering many heavy metal ions insoluble.
Collection of the precipitate and return of the heavy metal content to commer-
cial channels proved to be impractical in some cases. When the sulfide pre-
cipitate is released into the environment, it may undergo slow redissolution
because its solubility product is finite. Ultimately the heavy metal is
converted to the least soluble species for which anions are available.
In well-aerated surface water the hydroxyde (OH) ion is most available
and hydroxides are commonly formed from available metal ions. These precipi-
tate to greater depth in the oceans, eventually encountering a reducing
environment where excess sulfide ion exists. In this environment, stable
deposits of heavy metal sulfides are formed in sediments.* It is believed
that the processes at the interface between the anaerobic interior of sedi-
ments and the aerobic water in shallow, well-agitated waterways are similar.
One may, therefore, assume that heavy metal sulfides are immobilized once
they are included in a stable sediment. The organic nature of bacteria may
lead to feedback, however, and this feature was investigated separately.
Several experiments were conducted to investigate the environmental effects
of sulfide precipitates remaining in the environment. A series of experiments
were run in which minnows were exposed to the metal sulfides individually for
periods of 60 days. No significant evidence of toxicity of the sulfides of
zinc, copper, lead, cadmium, or manganese was found in these tests. In the
tanks containing CoS and NiS, 20% and nearly 60%, respectively, of the fish
died within 60 days as compared with 5% mortality in the control tank. Maxi-
mum concentrations of dissolved ion in these tanks were 7.0 ppm Co and 8.0 ppm
Ni. Significantly more fish died in the tank containing silver sulfide than
in the control also, but no dissolved silver was ever detected so the reason
for this result is not clear.
One further experiment was conducted as a preliminary investigation of
the reentry of heavy metals into the biosphere following sulfide precipitation.
A simple ecosystem was established in a tank carpeted with a sediment made up
*
Mary Sears, Ed. Oceanography AAAS, Washington, D.C., 1961, Publ. No. 67; p.
555.
Fairbridge, R.W., Encyclopedia of Geochemistry, Van Nostrand Reinhold Co.,
New York, p. 51.
54
-------
of kaolin, and two grams each of the sulfides of silver, cadmium, cobalt,
copper, manganese, nickel, lead and zinc, and 5.3 grains of barium chromate.
The ecosystem contained bacteria, Euglena, approximately 300 Tubifex (annelid
worms), a few Hornwart plants, and 100 fathead minnows that were added after
one month was allowed for the establishment of the rest of the system. The
fish were observed to be quite healthy for a period of eight weeks. Samples
of fish were removed from the aquarium at 4 weeks after their addition and at
8 weeks when the experiment was concluded. Results of heavy metal analyses
on these fish are shown in Table 10.
The analytical results strongly suggest a continuing concentration of
the heavy metals in the tissue of the minnows, particularly in the case of
manganese and lead. In the light of the absence of adequate controls, such
as analyses of minnows aging in an environment not similarly enriched, the
results of this preliminary experiment cannot be considered definitive, but
further study in this area is clearly desirable.
The treatment methods considered for heavy metal spills have been studied
in an extensive series of experiments including beaker and channel experiments
in the laboratory, and ambitious field tests. Beaker experiments in the labora-
tory provided verification of the fundamental chemical process involved in sul-
fide precipitation of heavy metals.
In an early treatment experiment, one tenth molar solutions of CdCl2,
MnS04, CuS04, FeCl3, ZnCl2, CoCl2, Pb(C2H302)2, AgN03, and HgCl2 were treated
in 100 ml quantities with two tenth molar sodium sulfide solution. No deliber-
ate excess sulfide was added, and the sulfide additions were measured using
a 100 ml beaker with 10 ml graduations in accord with the rough control likely
to be achieved in the field. Final sulfide concentration was determined using
a specific ion electrode, and pH was measured with a wide range pH indicator
paper. Residual metal concentrations were determined by atomic absorption
spectrophotometry. Results are shown in Table 11.
Further experiments were conducted to explore the effects of pH on sulfide
precipitation effectiveness. Lead acetate was used as the test material in
these experiments, and varying amounts of 0.1 M sodium hydroxide solution were
added to the sodium sulfide treatment. Results are shown in Figure 14* It is
seen that at high sodium hydroxide additions, precipitation of lead hydroxide
becomes significant and less than the stoichiometric sulfide treatment is
required. Such additions present a real hazard to aquatic biota as large pH
changes may be readily induced. The addition of a small amount of sodium
hydroxide to the sodium sulfide solution, however, may be seen to enhance
treatment before use and prevents the evolution of hydrogen sulfide gas. A
solution containing 1 M sodium sulfide and 0.1 M sodium hydroxide was found
to be quite effective. (This solution was not used in the experiment illus-
trated in Figure 14).
A series of experiments conducted in a stream simulation test facility
developed for this program provided information on the transport of spills
55
-------
Table 10
HEAVY METAL ANALYSIS OF FISH FROM HEAVY METAL
SULFIDE EXPOSURE TESTS (ppm of wet weight)
ELEMENT
Co
Cu
Mn
Ni
Pb
Zn
4 WEEKS'
EXPOSURE
1.5
3.5
14.0
3.8
17.0
47.0
8 WEEKS'
EXPOSURE
1.9
4.3
21.0
3.8
21.0
51.0
56
-------
Table 11
SULFIDE PRECIPITATION EXPERIMENTAL RESULTS
ORIGINAL
COMPOUND
CdCl2
MuSO
A
CuSO,
FeCl3
ZnCl2
CoCl2
Pb(C2H302)2
AgN03
HgCl2
FINAL
pH
7
7
4
7
12
7
n.d.*
8
9
FREE S=
CONCENTRATION
(MOLAR)
-17
3x10
3xlO"8
-25
<10
12
1x10
IxlO"4
IxlO-19
IxlO"20
3x10" 7
IxlO'13
TOTAL
S= CONG.
(MOLAR)
-11
3x10 J"L
3xlO~2
*
n.d.
-6
1x10 °
IxlO"3
IxlO-13
n.d.*
IxlO"2
5xlO-10
RES.
METAL CONG.
(ppm)
2
n.d.* but veiry high
30
200
2
200
100
0
0
n.d.: Not Determined Exactly
57
-------
1000r-
500
8
30T
200
10
D
O
1000 ppm Na £-9 HjO/O.IM HCI (9:1)
1000 ppm NajS* H2O
1000 ppm Na^-Q H2O/0.1M NaOH (5:1)
1000 ppm NajS-B H2O/0.1M NaOH (2:1)
TITRANT SOLUTIONS
10
20 30 40 50
VOLUME OF TITRANT ADDED (milliliters)
60
Figure(14 LEAD CONCENTRATION AS A FUNCTION OF TITRANT ADDED
58
-------
and treatment under various flow conditions, and allowed evaluation of treat-
ment methods. The effectiveness of sulfide precipitation with even the very
simplest techniques for locating a spill and delivering the treatment was
demons trated.
Results of these channel tests indicated that under conditions typical
of most natural streams, addition of the appropriate amount of sodium sulfide
solution close to the plume of spilled material will still result in the
removal of the metal from solution. Mixing in these highly turbulent flows
rapidly eliminates any errors in the distribution of the treatment. Under
these conditions, spill location using small quantities of sodium sulfide
as a tracer is highly effective.
Under conditions of little or no flow encountered in ponds and lakes,
sodium sulfide delivered off target will not mix with the metal for a long
period of time. Accurate location of the spill, and proper distribution of
the treatment will then be important. Introduction of some artificial turbu-
lences (by outboard motor boats, for example) should enhance treatment effec-
tiveness. Experiments showed that under these quiescent conditions, concen-
trated heavy metal solutions collected at the bottom of the water body.
Measurements of ion concentration must then be made on the bottom waters
of the lake or pond to be treated in order to allow proper sodium sulfide
additions to be determined. Automatic spill treatment procedures as discussed
earlier (Figure 11) may be useful in treating spills under these conditions.
Some bodies of water such as canals and some rivers will have conditions
between the extremes discussed above, and may or may not require precise
treatment delivery and auxiliary mixing.
Field tests were conducted in a mountain brook using ferrous sulfate as
a harmless simulant of hazardous heavy metal spills. Sodium sulfide was found
to be highly effective for spill location and marking under these conditions.
Rapid and thorough spill treatment was achieved by dumping sodium sulfide
solution out of plastic bottles into the region which the sodium sulfide
indicator solution showed to contain heavy metal ions.
Sulfide precipitation of spills of heavy metal compounds appears to be
quite effective and ready for application in the field. Spills of heavy
metal compounds are often small as shown in the spill statistics of Table
12 covering 1971 and early 1972.* They should be treated quickly, before
the spill is diluted to a concentration less than the solubility of the metal
sulfide. Thus it is important that treatment material and personnel trained
in its use are available near the spill.
The recommended treatment material is a solution of sodium sulfide in
water. The solution should be fairly concentrated to conserve space and
reduce weight. On the other hand, crystallization of sodium sulfide at low
Private communication with Hal Bernard, EPA.
59
-------
Table 12
HEAVY METAL SPILLS
MATERIAL
CHROMIC ACID
CHROMIUM IN ACID
MERCURY
NICKEL SOL. & GROUND
LIMESTONE
POTASSIUM PERMANGANATE
TETRAETHYL LEAD
ZINC SULFATE
CONTAINER
STORAGE TANK
STORAGE TANK
TREATMENT PLANT
PILE
PLANT
RR TANK CAR
PLANT
CAUSE
RUPTURE
OVERFLOW
LEAK
AVALANCHE
EQUIPMENT FAILURE
DERAILMENT
EQUIPMENT FAILURE
QTY. SPILLED
140 Ibs
7
?
7
11 Ibs
?
7-1/2 Ibs
SPILLED INTO
CREEK, TENN. RIVER
MILL CREEK
DETROIT RIVER
ILLINOIS RIVER
OHIO RIVER
LAND
7
-------
temperature is to be avoided. A solution formulated at the freezing point,
containing 0.85 pounds of sodium sulfide per gallon of water, and 0.04 pounds
of sodium hydroxide, meets these requirements. It can be stored indefinitely
at normal temperatures in plastic containers. For manual application a two-
gallon bottle was found to be comfortable to carry and pour from. Rugged
bottles with a molded handle, either cylindrical or in the shape of a Jerry,
can be obtained for about three dollars in lots of ten thousand. As shown
earlier, two gallons of treatment are sufficient for spills of two to ten
pounds of common heavy metal compounds. Five gallon bottles which could
treat larger spills were found to be difficult to manage in a small boat
typical to a spill response situation.
Since a quick response to the spill is important, treatment should be
applied by personnel near the scene. In the course of this study it was
determined that firemen were interested in the spill problem and motivated
to learn spill detection and treatment techniques. A police agency was inter-
viewed and indicated that their first reaction to a hazardous spill would be
to protect the public from the hazard and to ask a fire department to assist
with its disposal. The agency felt that its personnel were rotated too often
to benefit from spill training. These observations support the view that
firemen can be trained in treating small spills. Many volunteer and paid
fire companies participate in weekly training sessions that include training
films on new or difficult techniques. A training film could be prepared to
illustrate detection of heavy metal spills and their treatment with sodium
sulfide. There are tens of thousands of fire departments in the USA. It is
estimated that about ten thousand of these in conjunction with the US Coast
Guard and environmental agencies of the separate states would be sufficient
to cover the waterways in which a small heavy metal spill may occur. The
cost of equipping all of these with sodium sulfide solution (four gallons
at $0.2 each or $2000) and containers (2.80 each or $56,000), and producing
and circulating training films (estimated $30,000) is estimated to be less
than $100,000.
In the immediate future it would be desirable to conduct a pilot program
involving the fire department(s) of an industrial city responding to real
spills that might occur, and a few simulated (iron) spills in order to perfect
training techniques and evaluate practical hazards of sulfide precipitation.
It is anticipated that the training and administrative cost of the program
will be greater than the costs of materials.
61
-------
SECTION 9
IMMOBILIZATION OF HAZARDOUS CHEMICALS
The extent of environmental damage produced by any spill is dependent on
the nature and quantity of spilled material and the distribution of the spill
within the environment. Obvious advantages can accrue whenever it is possible
to minimize the areal or volumetric extent of hazardous concentrations of spilled
material either by interrupting the spill before containers are empty or by
preventing the spread of material that has already spilled on land to a nearby
waterway. A significant fraction of this project was devoted to these ends, and
a wide variety of problems was considered.
Ideally, a spill should be terminated as soon as possible by sealing leaking
or split containers before they are empty. When hazardous chemicals do spill
onto the ground, immobilization of the chemical to prevent flow of hazardous
liquids to surface water and to minimize percolation of liquids to subterranian
aquifers should be the next line of defense. If spilled material reaches a
watercourse, it is desirable to minimize the spread of floating liquids across
the surface or trap water immiscible liquids in streams and ditches without
interrupting the flow of water. In all of the above cases, additional advantages
result if the immobilization procedure leaves the spilled material in a form
which can be safely and quickly removed and packaged for further treatment,
shipment or disposal.
The following methods were employed experimentally to accomplish these
goals:
(1) Four powdered polymers, each capable of congealing one class of
hazardous liquid into a viscous, sticky mass, were selected and
combined into what has been termed colloquially "the universal
gelling agent," and used to seal narrow splits in containers, to
completely immobilize liquids on land, to prevent percolation
into the soil, to reduce surface spreading and improve the effect-
iveness of booms on water, and to facilitate the trapping of liquids
floating on streams in small mesh nets or screens. The resulting
gel was easily removed in convenient form for subsequent treatment
or disposal.
(2) Fly ash was used to absorb and immobilize hazardous liquids. When
used to immobilize concentrated acids or bases, the fly ash served
an additional function by minimizing the spattering when neutralizing
agents were applied. Neutralization reaction rates were then
determined by the amount of mechanical or hydraulic mixing applied
as part of the treatment procedure. By controlling the spattering,
the treatment method was made safer to use and therefore more
likely to be used.
62
-------
(3) Sumps to trap dense, immiscible liquids beneath flowing water
and subsequent pumping to remove the spilled material from
the stream were demonstrated.
(4) Sodium sulfide solutions were used to precipitate spills of
heavy metal salts that had percolated into the surface. Such
precipitates were found to be immobile except in association
with erosion.
Results of our investigations of methods for immobilizing hazardous
chemicals are summarized in the following subsections.
1. "UNIVERSAL" GELLING AGENT FOR HAZARDOUS LIQUID SPILLS. One major subclass
of chemicals transported in huge volumes includes those precursor organic
chemicals which by polymerization, thickening, or other chemical techniques are
turned into solid end products, such as organic monomers which react to form
the polymers which are the common plastics of everyday commerce. An obvious
immobilization technique for this class of chemicals was to polymerize them in
place at the site of the spill either on the ground or in the leaking tank.
Most polymerization reactions, however, are exothermic, giving off substantial
heat as they proceed so that explosion dangers are inherent in the solidification
of material by this type of uncontrolled self-polymerization. In addition, a
further environmental danger would have derived from the use of polymerization
catalysts of the common type such as benzoyl peroxide or laurdyl peroxide,
since the catalysts, themselves, are often'poisonous and explosive. More
innocuous thickening, solidifying or immobilizing techniques were desired. Over
the course of the project, such a countermeasure was developed to encompass a
wide variety of hazardous spills, including monomers, organic solvents, feed
stock chemicals, and inorganic reactants.
The mechanism by which the "universal gelling agent" acts is as follows.
The individual gel components selectively interact with the appropriate
chemicals themselves to create an immobile semi-solid mass which is easily removed
by mechanical means. This is in contrast to other spill immobilization tech-
niques which have been described and demonstrated, which simply absorb
the spilled liquid into a finely powdered mass of inexpensive, easily accessible,
easily deployable materials (such as fly ash or Portland cement) with few, if
any, secondary environmental hazards.
The "universal gelling agent" represents a new formulation created by the
mechanical blending of at least four and preferably five or more specific
ingredients, each having a specific purpose. The first ingredient is a material
of the highly water-soluble polyelectrolyte-type, typified by polyacrylamide.
This material could be substituted by any of a number of other polymers including
proteinaceous materials such as gelatin and casein. It is important that this
powder, and all the other components of the blend, be particle-size controlled
within a precise range for speedy interaction with the target liquid, and for
ease of deployment. It should also be manufactured or admixed with a small
surface-active additive to assure rapid contact with aqueous liquids.
63
-------
The second component of the blend is a loosely cross-linked copolymer of
the ilk typified by polytertiary-butystyrene copolymerized with divinylbenzene.
This material is selected to ineract most strongly with those liquids having low
polarity (such as cyclohexane, gasoline fractions, and a variety of other inert
spirits).
A third component is a material of the polyacrylonitrile-butadience copolymer
class which is chosen to be especially effective against polar organic chemicals
such as acrylonitrile, ethylene dichloride, and other chlorinated liquids.
The fourth component of the gelling agent is a material to cope with the
most difficult of all hazardous liquids to thicken, solidify, and immobilize
in place, as typified by methyl alcohol and other chemicals of the alcoholic
class. Materials suitable for this use include the polycarboxylmethylcellulose
polymers or the polyethylene oxide materials. For these types of polymers,
one of the less expensive polysaccharide exudates produced by bacterial cultures
can be substituted.
The four component blend used experimentally on this program consisted of
equal parts of:
(1) Dow Chemical Corporation, Gelgard, to combat spills of aqueous
liquids,
(2) Dow Chemical Corporation, Imbiber Beads, for spills of the inert
spirits-type liquids (typified by cyclohexane),
(3) BF Goodrich Corporation, Hycar 1422, to combat the polar organic
chemical spills including the chlorinated hydrocarbons, and
(4) BF Goodrich Corporation, Carbopol (and in some cases, Union Carbide,
Polyox) to selectively thicken and control alcohol spills.
Ease of delivery of this four-polymer blend required fluidization to ensure
rapid, smooth egress from commercial spray equipment. A one-fifth by proportion
addition of fumed silica "Cabosil," produced by Cabot Corporation was used. It
is well-known in the paint and pigment industry that such finely powdered silicas
can be used as thickeners for most organic vehicles. While the kinetics of such
thickening action are far too slow to be useful in the manner envisaged here,
the addition of Cabosil to the "universal gelling agent" for the primary purpose
of fluidizing the blend for ease of field deployment had the secondary benefit
of providing a stiffer, hydrolysis-resistant gel of almost all hazardous liquids
tested.
Consultation with manufacturers of the polymers mentioned in the preceding
paragraphs and with others in competitive industries indicated that the final cost
for the "universal gelling agent" can be brought to 50$ per pound or less.
The chemicals against which the "universal gelling agent" has so far been tested,
with excellent results, are listed in Table 13. it was learned from laboratory
experiments that approximately 10 to 25% of the "universal gelling agent" by
64
-------
Table 13
COMPOUNDS TESTED WITH UNIVERSAL GELLING AGENT
ACETONE
ACETONE CYANOHYDRIN
ACRYLONITRILE
AMMONIUM HYDROXIDE
ANILINE
BENZALDEHYDE
BENZENE
BUTANOL
CARBON DISULFIDE
CARBON TETRACHLORIDE
CHLORINE WATER (SATURATED)
CHLOROFORM
CYCLOHEXANE
CYCLOHEXANONE
ETHANOL
ETHYLACETATE
ETHYLENE DICHLORIDE
ETHYLENE GLYCOL
FORMALDEHYDE
GASOLINE
ISOPRENE
ISOPROPYL ALCOHOL
KEROSENE
METHANOL
METHYL ETHYL KETONE
OCTANE (2,2,A TRIMETHYL PENTANE)
ORTHO-DICHLOROBENZENE
PETROLEUM ETHER
PHENOL (89%)
PYRIDINE
SULFURIC ACID
TETRAHYDROFURAN
TRICHLOROETHYLENE
WATER
XYLENE
65
-------
weight, based upon the weight of the original spilled liquid, is required for
complete immobilization. Field studies have shown, on the other hand, that it
is usually necessarybecause of inefficiencies inherent in field deployment
of the material and the poor mixing usually obtainedthat a minimum of double
this theoretical amount is regularly required.
The results obtained with the experimental blend used on this program
demonstrate the potential value of the gelling agent for combating liquid spills.
Little effort was devoted on the program to optimizing the blend either from
the utilitarian or cost-effectiveness viewpoints. It is recommended that
additional effort be devoted to optimization of the agent in preparation for
operational use.
2. FIELD EXPERIMENTS. Most of the field experiments performed on the project
were designed to investigate methods for immobilizing hazardous chemicals.
These experiments were performed at a half square mile tract of abandoned
farmland owned by Calspan near Bethany, New York. After a series of soil tests
showed that the possibility of significant percolation of purposely-spilled
chemicals into subterranean aquifers or surface drainage into nearby streams
was negligible, the experimental facilities depicted schematically in Figure 15
were constructed.
Two sets of ditches, 2 feet wide by 2 feet deep and 100 feet long, were
constructed with 1% and 2% grades, respectively, to determine the effectiveness
of immobilization procedures against surface flow of spilled materials. At the
foot of each set of ditches, a large excavation was constructed to house two,
12-foot diameter, 3-foot deep, plastic-lined swimming pools intended to capture
all spilled material before escaping to the natural environment. The basic ex-
perimental concept used in most experiments was to simulate two spills in each
experiment. One spill was treated experimentally in an effort to immobilize
the chemical and a second was used as a control: i.e., to establish a baseline
against which the behavior of the treated spill could be compared.
In accordance with the general theme of this project, the chemicals used
for experimental spills in the field were selected to represent classes of
materials that pose serious hazards in the real world. Pure water was used to
simulate weak aqueous solutions. Cyclohexane was selected as a serious real
threat which is representative of chemicals which are immiscible with water and
float on the surface, while ethylene dichloride was selected to represent
immiscible chemicals that are more dense than water.
Sulfuric acid, the hazardous chemical shipped (through the United States)
in greatest quantity, was chosen in an attempt to solve the special problem of
frothing that has produced serious equipment damage during treatment of land
spills of acids and bases with neutralizing agents. In attempts to immobilize
heavy metals, the eventual fate of the compound used in the experiments could not
be predicted. Therefore, a relatively harmless compound, manganese sulfate, was
selected for experimental use. The results of these experiments are summarized
below.
66
-------
PERFORATED
56 GAL
DRUM
EXCAVATION
ALUMINUM
OUTFALL
100 FT LONG DITCH
12 FT DIA
SWIMMING FOOL
GARDEN
HOSE
56 GAL DRUM,
AND SUMP
PUMPS
CONTROL DITCH
EXPERIMENTAL
DITCH
POOLS
POND
EXCAVATION
(NOT TO SCALE)
Figure 15 EXPERIMENTAL FACILITY AT BETHANY TEST SITE
67
-------
3. IMMOBILIZATION OF LAND SPILLS WITH GELLING AGENTS. To minimize cost and
clean-up problems after experiments, specific gelling components were used in
all large-scale field experiments. The "universal" agent was tested only in
the laboratory.
Initial immobilization experiments in the field were performed with pure
water spills to represent weak aqueous solutions and to develop safe and effec-
tive field test treatment procedures before using hazardous chemicals. In
these experiments, Gelgard was used exclusively as the gelling agent. Ten
such experiments were performed on 55-gallon water spills in the 1% and 2%
ditches.
The manufacturer recommends 0.4% by weight (1.8 lb/55 gallons) of Gelgard
to congeal water. In the field tests, dosages ranging from 2 to 6 pounds were
used to immobilize one to five-minute long 55-gallon spills in distances rang-
ing from 25 to 75 feet in the ditches. The overtreatment was required to com-
pensate for inefficient distribution of the agent on the spill. The poor dis-
tribution was caused mainly by the powder's falling on already congealed water
and by the need to create a stiff gel to break the momentum of water at the
head of the flow, creating a dam capable of retaining the pressure of as-yet-
untreated water. Since ditches were always saturated with water before experi-
ments, the control spills of water always resulted in 55 gallons flowing over
the outfall 100 feet downstream.
To minimize the area affected by the spill, it is necessary to inhibit
flow as soon as possible. It is therefore particularly important to treat
the head of the flow first. Not only does this create a dam to interrupt
the flow, but the head is a region of turbulence which produces excellent
mixing of the agent with the liquid, promoting efficient treatment. Typically,
the first dam of congealed material reached a depth of 1/2 to 1 inch before
overflow began. It was most effective then to move downslope to form a
second dam. With the flow inhibited by the first gel dam, the second was
more easily formed, and by continued treatment was built to depths sometimes
exceeding 2 inches before overflow began. By progressive treatment in this
way the final dam usually exceeded four inches in depth with a 55-gallon
spill before flow was terminated in the ditch. Once this was accomplished,
it was a simple matter to move upstream and treat the liquid trapped behind
the dams that had been formed earlier.
Several dispersal methods were tested with Gelgard in these experiments.
It was apparent that broadcasting the agent with shovels was a simple and
effective procedure for spills with minimum dimension of four to five feet.
Much of the material was lost in our narrow ditches with this procedure.
When attempts were made to sprinkle the material from shovels, the distribu-
tion of the agent across the surface was usually uneven. As water flowed
around the thicker regions, an impenetrable gel formed at the surface, pro-
ducing a large clump with dry, unused agent at its center. For more uniform
distribution, sprinklers resembling salt shakers were constructed from three-
pound coffee cans to which six-foot long handles were attached. These
68
-------
sprinklers proved to be very effective for treating in narrow confinements
and were used extensively in subsequent experiments.
Hand pump dusters normally used for insecticide application in home gar-
dens could not deliver the agent fast enough to stop a 1/2-gallon per second
flow in the 2-foot wide ditches. Larger hand-powered dusters intended for
agricultural use delivered material at an adequate rate but were so tiring
that continuous operation could not be maintained. This type of duster could
be equipped with battery-powered explosion-proof motors to produce effective
portable equipment for treatment of spills that have reached remote areas
(e.g., where the spilled material flows into heavily wooded areas through
gulleys or small streams).
Dry chemical fire extinguishers produced an airborne plume of agent that
was too wide to treat spills in confined areas. Since fire extinguishers
cannot hold a sufficient mass of agent for treating large spills, they do
not seem practical for use against material that is already on the ground.
However, with a modified nozzle design, they could be made into very effec-
tive portable devices for sealing narrow splits in containers.
High pressure dispersal devices appear to be suitable for immobilizing
large spills that require large amounts of gelling agent where the use of
compressor-operated equipment is appropriate. Although paint sprayers proved
effective, they could not deliver the agent at a sufficient rate to treat
large spills. In these experiments, sand blasters, deliverying 5 to 10 pounds
of agent per minute, provided an appropriate distribution for large spills
even though some of the agent was always blown out of the simulated spill
area by the wind.
The bulk of gelled material was easily removed mechanically by shoveling
the material into 55-gallon drums. Heavy earth-moving equipment could be use-
ful for large spills of nonflammable materials, but should be avoided where a
fire hazard exists. The sticky consistency of most gelled chemicals makes
pumping inappropriate at least for land spills.
Typically, 75 to 85% of the spilled material was recovered in gelled form
during our experiments. Our experience was that some material was always
inaccessible to shovels after gelling chemicals on land. But the amount was
a small fraction of the mass lost in control experiments by a combination of
evaporation and percolation into the soil when the ditches had not been pre-
saturated with water before the experiment. In some cases in which Gelgard
is used on highways or city streets, thorough washdown should be required to
eliminate the lubricating effect of dilute mixtures of this material and water.
Like most polyelectrolytes, Gelgard produces a nearly friction-free environment
that could be extremely hazardous to pedestrian and automotive traffic. Manual
scrubbing followed by thorough rinsing would be adequate for small spills.
Vacuum scrubbers commonly used on airports for hangar floor and ramp washing
would be ideal for clean-up after treatment of large spills. Before leaving
the area of a spill treated with Gelgard, all affected surfaces should be
tested while wet, since friction is not affected by the dry form.
69
-------
In addition to the experiments with water, experiments were performed in
the ditches to test immobilization procedures against cyclohexane and ethylene
dichloride. In both cases, the 2% ditches were saturated with water but no
standing water was present. Seventy-five pounds of Imbiber beads were used
to completely immobilize the 55-gallon cyclohexane spill within 80 feet of
the spill point. Forty-one and one-half pounds of Hycar arrested the 55-
gallons of ethylene dichloride in slightly less than 70 feet. Aftej: the
treatment was completed in each case, water was pumped through four garden
hoses into the spill zone to determine if the gelled material could be dis-
lodged. As water seeped under the cyclohexane, buoyancy dislodged approxi-
mately 50% of the material in 30 minutes, but none flowed down the ditch.
The ethylene dichloride was unaffected. A surge flow of water, produced by
releasing 55 gallons at the head of the ditch in 2 minutes, washed approxi-
mately 75% of the gelled cyclohexane into the pool. The consistency of the
floating material was such that it could easily have been trapped with a
small mesh net or screen. Again, the ethylene dichloride was unaffected.
In both cases, a total of 75% of the spilled material was recovered in
gelled form by shoveling and 75% of the respective control spills was flowed
into drums placed at the outfall 100 feet downstream. A sequence of four
photographs are included in Figure 16 to illustrate the results of these
experiments.
Thus, it was demonstrated that significant benefit can be achieved by
using gelling agents to immobilize land spills of these two classes of
materials.
4. PROTECTION OF PERSONNEL AND PROPERTY. The experience acquired in pre-
paringforandexecutingtheseexperiments constituted an extremely valuable
portion of the overall results. Since the chemicals used in these demonstra-
tions were toxic by skin contact and inhalation, and represented significant
fire hazards, a variety of safety precautions had to be followed. Similar
and improved preparations must be made in advance for the protection of mem-
bers of clean-up teams for real spills.
Gas masks and air packs were provided to team members working in areas
of potential hazards from vapors of the spilled material. Oxygen was avail-
able in case of inadvertent inhalation of vapors in spite of these precautions.
In addition, two vapor sensors were used to warn against potential hazards
either from inhalation or fire.
Protective clothing, including boots, trousers with elastic ankle bands,
slip-over jacket with hood and elastic waistbands, and gloves, were provided
to all workers. When worn with face masks, the clothing seemed quite adequate
as temporary protection against skin contact. Garden hoses with running water
were available continuously to rinse these garments in case of minor splashing
or immersion. A small plastic swimming pool filled with water was maintained
on site for total immersion of any team member suffering extensive contact
with the spilled chemical.
70
-------
(A) TYPICAL SPILL CONFIGURATION
(C) COMPLETE IMMOBILIZATION
IN 79 FEET IN THIS CASE
(B) TREATMENT WITH GELLING AGENT
REMOVAL OF GELLED MATERIAL WITH
ALUMINUM SHOVEL. MAXIMUM THICKNESS
OF CONGEALED CYCLOHEXANE WAS ABOUT
4 INCHES IN THIS EXPERIMENT
Figure 16 IMMOBILIZATION OF HAZARDOUS CHEMICALS ON LAND WITH GELLING AGENTS
-------
Two types of protective suits* were used in these experiments. Disposable
plastic garments available from Protective Clothing Supply** were light and
fairly comfortable but were more susceptible to tears and abrasions than the
heavier reusable garments available. Both types of protective gear were
extremely warm when worn in sunshine and neither type offered significant
protection against fire.
Significant improvements could be made in protective clothing. For
example, incorporating a metallized finish with the fabric or plastic would
seem appropriate to reflect heat from the sun for increased comfort and,
more importantly, to offer some protection against the initial flash and
subsequent radiation from fires.
With the materials used in our experiments (and many of the spills that
occur in the real world), the fire hazard is by far the most serious. We
protected ourselves at the Bethany site by limiting experiments to conditions
with minimum windspeed of 5 mph, carefully monitoring the vapor concentration
in the working area and simply leaving the area once when the vapor concentra-
tion in the ditch exceeded the lower explosion limit. Our plan for fire was
to let it burn and use fire extinguishers only to'assist in rescue operations
if necessary. Such behavior is not acceptable in the real world where the
spill site cannot be selected and carefully prepared with fire breaks in
advance. The potential fire hazards must be considered in each spill and
suitable tactical measures must be implemented for protection of personnel
and property. Additional research is required to develop countermeasures
which are specifically applicable to the spill situation.
Only two concepts for minimizing the fire hazard were test on this pro-
gram. Attempts were made to mix a variety of dry chemical fire-fighting
agent with the universal gelling agent in the hope that if a fire started,
the C02 generated by combustion of the gel would extinguish the fire or at
least reduce its intensity. But laboratory experiments showed that even
with 50% dry chemical fire-fighting agents, there was no significant reduc-
tion of hazard. Dry chemicals and foam were effective, however, for extin-
guishing fires fueled by the gelled material.
The second approach, that of reducing evaporation of spilled material
to maintain incombustible vapor-air mixtures above the spill, was more success-
ful. To accomplish this end, the spilled material was treated with a high
density fire-fighting foam before application of the gelling agent. These
vapor suppression experiments were performed on a specially-equipped test
pad used for a variety of fire-fighting experiments at Calspan. The test
pad (Figure 17) consists of a concrete floor with three concentric steel
rings which are held in the concrete and rise 3 inches above the concrete
surface. The liquid level within the concentric rings is equalized
X
'Safety First" Disposable two-piece suits with hood.
1243 Military Road, Buffalo, New York.
72
-------
STEEL RINGS
3 IN. HIGH
Figure 17 SKETCH OF PAD FOR VAPOR SUPPRESSION TESTS
Table 14
LIST OF FOAM CONCENTRATES FOR FLAMMABLE LIQUID
NATIONAL'S FOAM LIQUID
FLAMMABLE LIQUID
AER-0-FOAM XL-3
AER-0-FOAM "991 -6%
ETHYLENE BICHLORIDE
CYCLOHEXANE
METHANOL
73
-------
through holes in the rings. Fifty-five gallons of flammable liquid on the
test pad forms a pool approximately 1-1/2 inches deep.
Two instruments were used to measure the reduction in vapor concentration
above the spills: a Johnson-Williams Model CD 800 W with recorder output,
and a model G vapor sniffer. These instruments measure the percent by volume
of the vapor in air and give a meter reading calibrated in the percent lower
explosion limit (%LEL) for the vapor. An explosion could occur at 100% LEL.
The devices used were calibrated for methanol, which has the lowest LEL of
materials used in the experiments. Vapor measurements were made at known
distances from the liquid surface of the spill before and after application
of the foam.
High density foam was applied with a National Foam System, Inc. "Hand-
held Nozzle" Model RP-6B with the HL metering-check valve combination con-
nected directly to the nozzle. Table 14 lists the foams and the flammable
liquids on which they were used.
The foam was applied by splashing it in front of the outside steel ring
so that it would gently flow onto the flammable liquid. Approximately 2.5
gallons of foam concentrate were applied in each of the 3 experiments which
took approximately 50 seconds each time. The resultant foam depths over the
flammable liquids ranged from 1/2 to 1 inch.
The results of these experiments are summarized in Table 15 . Note that
in each case, the vapor concentration was reduced by a factor of at least
three. Best results, reduction by factors of 6 to 10, were achieved with
the cyclohexane and ethylene dichloride which are water insoluble. It was
postulated that as the foam rolled onto the methanol, some of the methanol
was dissolved by the water in the foam and transported to the top of the
foam layer.
All the experiments were performed in fresh-to-moderate breezes, which
limited the initial vapor concentration (measured 1 to 3 inches above the
surface of the spill) to low values. Significantly better improvement ratios
were observed under the quiescent conditions of the laboratory, where high
initial concentrations existed at the same distance above the surface.
In each of these three experiments, specific gelling agents were used:
Imbiber beads for cyclohexane, Hycar for ethylene dichloride> and Carbopol
for methyl alcohol. In all cases, the presence of the foam appeared to
inhibit contact between the spilled liquid and the powdered treatment.
Vigorous mixing was required to promote gelling.
In subsequent experiments performed on a laboratory scale, each of the
specific gelling ingredients was mixed with Gelgard and applied to foam-
treated surfaces of the same flammable test chemicals. In each case, the
Gelgard reacted with the water in the-foam and caused it to disperse. The
remaining ingredients apparently remained free to react with the flammable
liquids since gelling of these materials proceeded at a more nearly normal
74
-------
Table 15
SUMMARY OF RESULTS ON COMBINED
VAPOR SUPPRESSION AND GELLING TESTS
FLAMMABLE LIQUID
VAPOR READING IN L.E.L.
NO FOAM
WITH FOAM
GELLING MATERIAL
AND AMOUNT
COMMENTS
ETHYLENE DICHLORIDE
6-8
0.8
-J
en
CYCLOHEXANE
24
METHANOL
HYCAR 1422 60
IMBIBER 55
BEADS
CARBOPOL 50
FOAM APPEARS TO SLOW GELLING
ACTION. MAY BE DUE TO
BOTH FLUOROCARBON AND WATER
USED TO GENERATE FOAM.
LARGE BEADS WERE USED; THESE
REACT SLOWLY. HERE AGAIN
INTERFERENCE WITH FLUOROCARBON
AND WATER IS SUSPECTED.
CARBOPOL SEALS ITSELF OFF WHEN
APPLIED BY HAND LEAVING UNREACTED
MATERIAL INSIDE LARGE CLUMPS;
REQUIRED MIXING TO GET GELLED
PRODUCT.
-------
rate. But the effectiveness of the foam in minimizing the fire hazard was
substantially reduced.
These experiments suggest that in a large spill the fire hazard due to
spilled liquids might first be minimized by coating the entire spill with
foam. The universal gelling agent might then be used to treat relatively
small portions of the spill-affected area. When gelling is complete in one
area, additional foam could be applied while the gelling treatment is applied
elsewhere. By such a procedure, the overall fire hazard might be kept to
manageable proportions. Before firm conclusions can be drawn, however, test-
ing would be required.
In addition, only high-density foams were tested on the program. Con-
sidering the variety of fire-fighting foams and application techniques avail-
able, substantially more testing is warranted before making specific recom-
mendations .
5. THE USE OF GELLING AGENTS ON WATER SPILLS. Limited small-scale experi-
ments using gelling agents to immobilize hazardous liquids floating on water
showed that the fraction of liquid that had not gone into solution could be
readily congealed. In tests with benzene, cyclohexane, and gasoline, the
congealed material continued to float indefinitely. Hycar-treated acryloni-
trile, on the other hand, floated for approximately 48 hours in the beaker
experiments and then gradually sank to the bottom. (Bioassay tests with both
benzene and acrylonitrile showed that fathead minnows mistook small floating
particles of gelled material for food and died within a few hours. These
tests may not be valid, however, since the aquarium fish were trained to eat
food sprinkled on the surface.) The ease and speed with which the gelled
material could be removed from the water surface suggest, in any case, that
the procedure could be quite useful and that larger scale tests are warranted.
Several large-scale laboratory experiments were performed in 12-foot
diameter swimming pools to determine the effectiveness of gelling procedures
for immobilizing spills of materials which are immiscible with water and less
dense than water. Cyclohexane was used as the test material and Imbiber beads
were used for treatment.
Cyclohexane spilled on clean water initially forms thin beads that per-
sist for several minutes. After minor agitation, presumably as a result of
a small amount of cyclohexane going into solution locally, the bead spreads
spontaneously into an extremely thin film. In a variety of experiments in
which cyclohexane was both spilled from beneath the water surface and poured
onto the surface, the spontaneous spreading was prevented by treatment with
Imbiber beads. As spillage and treatment continued, the gelled material
reached thicknesses up to approximately a half centimeter as it gradually
drifted outward from the spill center. Light breezes caused noticeable
drift of rafts of the congealed material, and no tendencies to sink were
observed.
76
-------
The gelled cyclohexane could be readily confined and immobilization com-
pleted with unsophisticated booms constructed of short lengths of 4 x 4 lumber
to which screen wire had been attached. These booms completely confined the
gelled material and permitted it to be compressed to thicknesses of about
four inches.
Field experiments were performed with this hazardous material at our
Bethany site. The 15-by-30 foot excavation at the foot of the 1 percent
ditches was filled with water to simulate a lake. A 24-foot boom constructed
of 4.x 4fs, screen, and an 8-inch plastic skirt was placed around the outfall
in such a way as to confine the spilled material to about one quarter of the
excavation surface. Fifty-five gallons of cyclohexane were spilled 15 feet
upstream from the outfall of a ditch in which a continuous water flow was
maintained with two garden hoses. The experimental configuration is shown
in Figure 18 . Sixty pounds of Imbiber beads were broadcast onto the spill
from shovels as the cyclohexane entered the excavation. As in the pool experi-
ments, the treatment was highly effective. The cyclohexane was congealed
into a floating mass with an average thickness of approximately 1/4 inch that
was completely contained within the boom. The boom was used to corral the
treated material to the ramp area of the pond (See Figure 18) and to compress
the gell to thicknesses up to 2 inches without loss. Aluminum shovels were
then used to skim the l-to-2-inch layer of stacked gelled material from the
water into 55-gallon drums. Approximately 75% of the original spill volume
was collected from the ramp area. The losses can be attributed to evaporation
and to adhesion to the walls of the excavation. These results are illustrated
in the four photographs presented as Figure 19.
After removal of the bulk of the gel, the surface of the pond was swept
with a piston film of Sorbitan Monooleate to complete treatment. Bulk water
contained 3.8 ppm of cyclohexane.
There was no control test performed in conjunction with this experiment.
There is little doubt, however, from tests performed in swimming pools that
untreated cyclohexane would have escaped through the crude boom and covered
the entire excavation. A skimming operation involving special equipment
would have been required to recover the spilled material.
77
-------
EXCAVATION PIT
PLUGGED
DRAINAGE DITCH
FLOATING BOOM
RAMP LEADING
DOWN INTO PIT
#1 TREATMENT STATION
#2 TREATMENT STATION
55 GALLON DRUM CYCLOHEXANE
2-100' - 1% TRENCHES
Figure 18 PLAN VIEW OF "LAKE" CYCLOHEXANE SPILL TREATMENT TEST
78
-------
]
3
(A) CYCLOHEXANE SPILL IN PROGRESS
AND TREATMENT WITH IMBIBER BEADS
(B) THE INITIAL BOOM CONFIGURATION
(C) BOOM TOWED TO RAMP AREA FOR
REMOVAL OF GELLED CYCLOHEXANE
(D) NOTE ABOUT 2-INCH THICKNESS
OF GEL IN BOOM AND SHOVEL
Figure 19 INCREASING THE EFFECTIVENESS OF BOOMS BY CONGEALING THE TRAPPED LIQUID
-------
6. GRAVITY COLLECTION OF HEAVIER-THAN-WATER CHEMICALS IN STREAMS. An experi-
ment was performed to determine whether or not liquids that are immiscible with
water and have a specific gravity exceeding unity could be effectively trapped
in an artificial depression in a stream bed. Ethylene dichloride (specific
gravity: 1.1) was used as the test material in one of the ditches with a 2%
gradient. For the demonstration, an excavation was made in the trench approxi-
mately 75 feet from the spill drum.
To simulate a very slow-moving stream, a sandbag dam was constructed below
the excavation and the ditch was filled with water until the dam overflowed.
Water was allowed to back up into the trench to a position about 40 feet above
the excavation. A 250 ml beaker was placed in a small depression between the
excavation and the dam. The experimental configuration is illustrated sche-
matically in Figure 20. The 55-gallon spill was accomplished in two minutes
from the position indicated in the figure. Three hoses maintained a continuous
flow (about 15 gpm) of water from the spill zone over the dam.
ETHYLENE DICHLORIDE
SPILLED UPSTREAM
GROUND LEVEL
ETHYLENE PI CHLORIDE jj
Figure 20 CROSS SECTION OF THE DITCH AS MODIFIED TO TRAP
HEAvTER-f HAN-WATER SPILLS
80
-------
Immediately after spilling at the end of the trench, the solvent formed large
streaming beads that moved with the shallow flowing water. Small pockets of
the material were trapped in depressions along the bottom of the trench and in
turbulent eddys behind a few large clumps of clay that had fallen into the
trench. Upon reaching deeper water (approximately 1/2 inch), the material
settled completely beneath the surface of the water and continued to flow under
the influence of gravity into the excavation. When the spill had been completed
and flow of ethylene dichloride could no longer be detected in the ditch, the
beaker was removed and found to contain no visible ethylene dichloride. It was
then replaced. The "sniffer" did not detect ethylene dichloride vapors above
the water surface, but small concentrations were detected above trapped beads
of solvent. These beads, incidentally, could be easily dislodged by agitating
the water in the vicinity of the obstruction or depression.
After a variety of peripheral experiments of this kind, the dam was re-
moved to produce rapid flow of water over the excavation and thereby simulate
the effect of a fast stream. When the flow had reached equilibrium, the beaker
was again removed. Several additional surge flows were produced by spilling
55 gallons of water into the ditch and additional beaker samples withdrawn.
In no case was ethylene dichloride visible in the beaker.
Collected material was then pumped with a submersible pump from the de-
pression in the trench to another drum. Forty-seven gallons were recovered.
Samples of water which overflowed the dam were analyzed for ethylene
dichloride using a H/P 5750 gas chromatograph with a flame ionization detector.
The samples collected from the beakers (mounted between the depression and the
dam) were found to have ethylene dichloride concentrations ranging from 300 ppm
before the surge flow to 4400 ppm after the surge flow, which may be compared
with the nearly pure ethylene dichloride in the bottom of the depression. When
the dam was removed to initiate surge flow, ethylene dichloride concentration
was measured at 2000 ppm in a sample collected at the outfall. A sample of
water collected from the overflow of the dam just before the surge flow was
produced contained only 4.4 ppm ethylene dichloride. It is not known whether
the higher concentrations that accompanied the surge flows were due to entrain-
ment of ethylene dichloride from the excavation or due to entrainment of water
that had rested above the ethylene dichloride in. the excavation for long periods
of time and thereby contained the solvent in solution, but in either case it
is concluded that removal of sunken hazardous material should be undertaken
whenever possible to minimize contamination of the water column.
Based on our small-scale field tests, the separation of substances with
specific gravity exceeding one from water by gravity appears to be a viable
way to recover spills in waterways. Conceivably, advance notice of an upstream
spill would provide time for an excavation to be produced in a downstream loca-
tion for the entrapment of the denser material traveling along the stream bottom.
81
-------
Benefits of this technique include:
1. High percentage recovery of original spill volume.
2. Spill material in many cases may be recovered in nearly
unadulterated, easily recoverable form.
3. Minimum investment in equipment required.
4. Materials and equipment used are easily obtained.
In consideration of the use of this technique, it must be recognized that
artificial excavations must be made in streams. The natural depressions in
streams are present because the hydrodynamics in the stream is such that heavier-
than-water bottom sediments are naturally removed from those locations. While
such locations should be searched for pockets of trapped material, it is not
reasonable to expect that a given natural depression would trap significant
quantities for long periods of time.
7. TREATMENT OF SULFURIC ACID LAND SPILLS. One of the principal problems
that has occurred in attempting to treat concentrated acids and bases spilled
on land is caused by the severe bubbling and frothing that occurs when large
quantities of neutralizing agents are bulldozed into the pool of material.
Particularly difficult situations have been reported when treating sulfuric acid
spills with lime. In consideration of this problem, it was reasoned that a
process that would permit the control of the reaction rate could be of value.
One concept that evolved was the use of fly ash to absorb and immobilize the
acid. The limestone could then be added after immobilization was complete.
Under such conditions, neutralization would occur only at the surface of con-
tact between the two materials. After the frothing that accompanied the initial
contact had ceased, the two substances could be periodically mixed at a control-
led rate by some remote means to cause the reaction to occur after equipment
and personnel had been removed to safer locations.
After laboratory tests indicated that significant control could be achieved
by this means, a small field demonstration was performed. A 15-gallon neoprene
drum of sulfuric acid was placed in the bottom of a 100-foot, 1% grade trench,
approximately 10 feet from the upstream end. One hundred pounds of fly ash was
plac'ed 10 feet downslope from the acid drum, while four piles of lime were
strategically located at intervals directly downstream from the fly ash dam, as
illustrated in Figure 21.
82
-------
15 gal
SULFURICACID
FLY ASH
DAM
LIME DAMS
100 Ib
120 Ib
120
120
160
[1% GRADE TRENCH (100' x 2' x 2')]
Figure 21 TREATMENT OF SULFURIC ACID SPILL
The measured pH's of pond water, water in the flooded excavation at the end of
the trench, and residual water puddled in the bottom of the trench, were 8.0,
7.9, and 8.0, respectively, before spilling the acid. Three control lines were
attached to the container of sulfuric acid. Personnel outfitted with complete
sets of protective clothing, including PVC suits, rubber boots, neoprene gloves,
and protective goggles, adjusted tension on the lines to trip the acid drum and
produce a controlled spill.
The initial fly ash dam effectively prevented flow of sulfuric acid in the
trench, but seepage into the fly ash was slow. In the interest of time, an
additional 50 pounds of fly ash were shoveled on top of the liquid acid to form
a stiff slurry, which included all of the fly ash in the initial dam. Two
hundred twenty pounds of crushed limestone were then shoveled onto the slurry.
There was an initial strong reaction of lime with the acid with frothing caused
by C(>2 evolution (H2S04 + CaCO^ = Ca 804 + C02 + H20). This reaction subsided
after five minutes and was reiterated upon mixing with a garden rake. Reaction
rate was found to be controlled by degree of mixing.
The degree of control of the reaction rate which can be achieved by this
procedure is vividly illustrated with the two photographs presented in Figure 22.
The froth formed by the initial layer of lime on the acid fly ash slurry is
visible in Figure 25A just above the shovel handle. The second lime layer is
visible on top of this subsided froth layer in the lower right corner of the
photograph. No evidence of a chemical reaction is indicated where the lime and
froth layer meet.
83
-------
Figure 22A SLOW REACTION WHICH OCCURS WHEN LIME IS SPREAD
ON TOP OF ACID-FLY ASH SLURRY
Figure 22B CLOUD FORMED BY INTENSE REACTION THAT OCCURS
WHEN LIME LAYER IS RAKED INTO ACID-FLY ASH
SLURRY BELOW
-------
As the lime was raked into the acid fly ash slurry (Figure 22B), the
accelerated reaction was so intense that a dense cloud formed in the ditch
above the newly formed froth. Just prior to this experiment, a thunderstorm
had saturated the Bethany site. The cloud probably consisted of water condensed
out of the air on condensation nuclei that were generated by the reaction.
Tiny sulfuric acid droplets may have been ejected into the air by breaking of
bubbles in the heavy froth that was formed. The bubble breaking mechanism is
a well known source of sea salt nuclei that predominate in maritime atmospheres.
Regardless of the source of the cloud, it is evidence of the energetic reaction
that occurred wherever fresh limestone was mixed into the acid slurry. When
the reaction subsided, the cloud disappeared. When the slurry was raked with-
out the addition of fresh lime, the cloud did not form.
This experiment was completed by gradually mixing 950 pounds of lime into
the acid slurry over a 3.5 hour period. In the latter portion of the experiment
water was mixed in with the lime and acid slurry to help drive the reaction to
completion. In the final stage, slaked lime was mixed into the slurry to raise
the pH to 8. The slurry was then shoveled into 55 gallon drums and the residue
washed into the excavation below.
8. IMMOBILIZATION OF HEAVY METAL SPILLS ON LAND. Solutions of heavy metals
can be immobilized on the surface by using any of the techniques described
earlier; i.e., by congealing the solution with the universal gelling agents or
absorption with fly ash or portland cement. Whatever fraction of the solution
>that has percolated into the soil is unaffected by these procedures, however.
It was postulated that this material could be immobilized by application of
sodium sulfide to the contaminated soil in order to convert the soluble salt
to an insoluble sulfide. Heavy metal sulfide particles formed underground
should be completely immobile except by erosion of the land surface.
To test this concept, 55 gallons of 5% MnS04 solution were spilled in each
of the two 1% grade ditches at such a slow rate that all of the liquid soaked
into the ditch bottom. One of the ditches was then treated with sodium sulfide
using a 3.5 gallon garden sprayer. Fifty-five-galIon drums of water were then
spilled rapidly at the head of each ditch and samples collected at the outfall
about 10 seconds after passage of the leading edge of the surge flow. A steady
flow was then maintained in the two ditches between periodic surge flows which
were continued for the next 24 hours.
At the conclusion of the experiment, the water collected in the 12-foot-
diameter pools contained 72 ppm and 6 ppm of dissolved Mn in the control and
treated experiments, respectively. These numbers correspond to 10% and 1% of
the total spilled material in the two ditches. The remainder of the metal was
presumably trapped in the soil.
The experiment demonstrated conclusively that the treatment can signifi-
cantly decrease the transport of spilled heavy metals to water-courses in sub-
sequent surface runoff.
85
-------
If the sodium sulfide treatment is used for treatment of land spills, it
must be recognized that significant overtreatment may be necessary due to
formation of sulfides of other competing metals already in the soil.
9. LOGISTICS AND COST OF IMMOBILIZATION PROCEDURES. Considering the wide
variety of potential applications, treatment procedures and materials for
immobilization of hazardous chemicals, it was not surprising that numerous pos-
sibilities exist for preparations to implement these techniques. At the state
of development of the immobilization concept up to the end of this program, it
was not appropriate to perform an exhaustive operational analysis of possible
use concepts and associated preparation requirements and costs. Investigations,
therefore, were limited, and intended only to provide sufficient information
to generate an opinion as to the eventual practicability of the concept. For
this study, use of the universal gelling agent was considered and the following
assumptions made.
1. To be effective., attempts to seal leaking containers should be made
within half an hour of the occurrence of the spill-producing accident
so that small-scale local supplies and equipment are required.
2. To be effective, immobilization procedures for materials that are al-
ready on the surface should be initiated within a few hours after the
occurrence of the spill.
The only practical concept visualized for maintaining local supplies of
the universal gelling agent available for sealing leaky containers is to pro-
vide such supplies and equipment to either local fire departments or local law
enforcement agencies. It is anticipated that equipment similar to dry chemical
fire extinguishers with specially designed nozzles suitable for ejecting a
narrow stream of gelling agent five to ten feet from the operator can be devel-
oped for this purpose. Judging from the cost of existing fire extinguishers,
such equipment could be procured in quantity for under $100/unit. Ten thousand
local organizations could then be equipped for a cost of $1,000,000.
The requirements for distribution of sufficient equipment and supplies for
treatment of material on the surface is substantially more complex and must be
considered on a regional basis. The distribution of caches located in Nevada,
for example, need not be as dense as in the industrial regions of the northeast.
Delivery of equipment and supplies from the cache to a spill in metropolitan
areas must almost necessarily be by truck. In rural areas, delivery and direct
dispersal from helicopters would be advantageous in many cases. Packaging pro-
blems are significantly different in the two cases.
Without attempting to solve these problems, it was assumed that as many
as 200 sites throughout the country might be equipped and supplied for immobi-
lizing spilled chemicals with the universal gelling agent and that delivery
would be accomplished with ground transportation.
86
-------
The following additional assumptions were made. Each site is supplied
with 10,000 pounds of universal gelling agent (to provide a capability for
treating 20,000-pound to 50,000-pound spills, depending on the chemical) pack-
aged in 200 piastic-lined cardboard drums that are permanently stored on 20
warehouse pallets for ready loading on trucks. The 50-pound drums were selected
to permit handling by one man in the field. Delivery equipment would include
two modified sand blasters, one air compressor, and ten aluminum shovels. Pro-
tective clothing and gas masks for ten workers were also assumed at each site.
Assuming a price of $2.00/lb. as an upper limit (slightly more than paid
for our mix procured in experimental quantities), the outlay for the gelling
agent would be $20,000 per site. At the projected price of $.50/lb., this cost
could be as low as $5,000. The following unit prices are appropriate for other
equipment and supplies: drums, $1.50; pallets, $4.00; sand blasters, $100.00;
compressors, $5,500.00; shovels, $5.00; complete set of protective clothing,
$17.00; and gas masks, $45.00. Cost per site for these items, plus $300.00
for miscellaneous items not considered, would be $7,000.00 Assuming $2.00/lb.
for the 200 sites throughout the country, this amounts to $1.5 million for
the gelling agent; cost'per site would be $27,000. Total cost for the 200 sites
would be $5.5 million. To this should be added the $1 million for equipping
10,000 fire departments for sealing split containers.
In our opinion, these costs for preparation for immobilization of spilled
hazardous chemicals are not excessive. The cost of training personnel, storing
materials, and administering the program has not been considered. Certainly,
further development of the concepts for immobilization of spilled materials is
warranted.
87
-------
Section 10
DETECTION AND MONITORING
For the effective treatment of spills of hazardous material, it is imper-
ative that the spill be detected and identified rapidly and treated while it
is concentrated. Rapid detection makes it possible to discover a spill before
it is dilute and to reduce damage from the spill by means of localized treat-
ment, impoundment and closing of municipal water intakes. A concentrated spill
is easier to treat than a dilute spill and fewer biotae (including humans) are
exposed to the spilled and treatment chemicals. If and when a spill occurs,
the geographical extent and concentration of the spilled material must be
monitored during the course of the treatment. Delineation of the amount and
extent of untreated material permits cost-effective treatment and limits the
exposure of the environment to excesses of the treatment chemicals.
For many spills, anticipation is impossible and early detection depends
on the report of the person responsible for the spill or on environmental signs
which are obvious to untrained observers (gross changes in color, odor, fish
behavior, etc.). On the other hand, in certain probable spill locations, such
as harbors and industrial rivers, arrays of automatic detectors could be cost-
effective. To maximize the ratio of spill damage prevented to instrumentation
cost, such detectors should react to a wide spectrum of possible pollutants,
as do pH indicators, which respond to spills of acids, bases, and chemicals
(such as liquified chlorine and ammonia) which form acids or bases when mixed
with water. Furthermore, to be suitable for such applications, instrumentation
must require little maintenance and be resistant to deterioration in the hostile
environment characteristic of sewers and industrial rivers.
The requirements on equipment and methods for monitoring the extent of a
spill and the progress in its treatment are somewhat different. Such equipment
does not require the same degree of environmental immunity as the detection
instruments since it need be deployed only during the treatment of a spill and
need survive only a few hours of environmental exposure. Temporary deployment
also obviates the need for extended periods of maintenance-free operation and
permits rather rapid expenditure of consumables, e.g., indicator chemicals.
Since an operator can be deployed with the equipment, automatic operation is
not necessary and an operator is available for interpreting the equipment out-
puts. Finally, a few highly portable instruments can cover a rather wide area
os possible pollution sites; so cost is not quite so critical as for the detec-
tion devices. On the other hand, monitoring equipment must give semiquantitative
results at the very minimum, an indication of the endpoint of the spill-
treatment titration and should provide a fairly specific indication of the
nature of the spilled material. Furthermore, although an operator is available,
he, in general, will not be skilled in chemistry or electronics; so the instru-
ment should require only the normal observational skills of the average person,
as does a colorimetric indicator.
88
-------
During 1971-72 a study was undertaken to select methods and instrumentation
that are commercially available or could be quickly developed to meet the afore-
mentioned detection and/or monitoring equipment criteria, with major emphasis
on applications to the detection and monitoring of spills into watercourses.
A wide range of chemical and physical phenomena were explored in this effort,
as described in detail in a separate volume of this report. Electrical con-
ductivity was found to be highly effective in detecting the presence and measur-
ing the extent of untreated ionic solute spills but was essentially worthless
in indicating the effectiveness of the treatment. Commercial pH probes and
certain other specific ion probes proved to be highly effective in detecting
and measuring spills of acids, bases, and metallic salts and in monitoring the
effectiveness of their treatment. Volatile organics triggered indications from
commercial catalytic combustion sensors while less volatile organics were detect-
ed with a multicolor transmissometer. Colorimetric indicators proved to be
extremely helpful and were used both in a spill detection kit designed for use
by firemen and in an automatic "cyclic colorimeter," which uses modulation of
indicator injection to reduce its sensitivity to turbidity and fouling. Colori-
metric indicators were not reliable when added to a water body (e.g., as a stripe
sprayed from a low flying airplane) to delineate the extent of the spilled
material. Other experiments indicated the effectiveness of dyes as spill tracers
and of the sense of smell as a spill detector. One of the most significant
outcomes of the effort was the development of a spill detection kit containing
a conductivity meter, pH indicator, odor samples, and colorimetric reagents
geared to the prescribed spill treatments. This kit has been proven effective
in tests involving volunteer firemen as operators.
Details of the instrumentation portion of this project are presented in
a separate report.
89
-------
Section 11
BIOASSAY STUDIES
Bioassay experiments followed the procedures recommended in Standard
Methods , 13th Edition. The water pH, temperature, level of dissolved oxygen,
exposure to ambient light, and concentrations of toxicant were all controlled
and recorded, as well as additional parameters of the water quality including
its alkalinity and hardness. Water quality data are listed in Table 16. The
test temperature for all of our experiments was 60°F ±1°F and the lighting
cycle was adjusted to produce 12 hours of simulated daylight and 12 hours of
darkness.
The LDso values for the six major compounds of interest in this study,
over a seven day exposure period, were determined. In addition, observations
were made of the minimum concentration of each substance which produced any
observable untoward effects and/or deaths of the test organisms. Numerous
well-characterized controls were included in all bioassay tests.
The selected test organisms were the fathead minnow, Pimephales promelas
promelas. (acclimated to the test water for 24 to 48 hours prior to each experi-
ment), algae, Cfr \ arnyc 1 omonas reinhardtii. and plants, Ceratophvllum demersum.
During the experiments, daily records were kept of fish behavior and fish deaths.
Upon the death of any fish, it was immediately removed from the assay tank.
A brief summary of the large amount of data generated in these bioassay
tests is provided in the following tables and graphs. Table 17 and Figures 23
through 25 report the findings after challenging both fish and algae with 10 ppm
concentrations of nine chemicals of major interest in this program. Recogniz-
ing that there was a natural decay rate of each of these chemicals during the
assay procedure, especially due to continuous aeration of the system, data were
also collected on the dissipation rates which were encountered. Figure 26 plots
these data.
Tables 18 and 19 provide information on the time elapsed before observable
detrimental effects were encountered with lower than 10 ppm concentrations of
each nominally hazardous chemical. The approximate LDso values are also reported.
One of the major tasks of the bioassay program was to determine the poten-
tial secondary environmental consequences of application of our recommended
countermeasures . Chief among the countermeasures for soluble hazardous chemicals
in water has been treatment with activated carbon. Thus, both powdered and
granular activated carbon with adsorbed test pollutant (phenol and acrylonitrile)
were used as a "secondary pollutant" challenge in our bioassay system. Tables 20
21 and 22 document the findings that adverse secondary effects were indeed
observed after exposure of fish to the pollutant-loaded carbon sediments, whether
in powdery or granular form. The bulk water concentrations of pollutants were
well below the levels found, in the preliminary series of tests, to cause adverse
90
-------
Table 16
THE CHEMICAL ANALYSIS OF TEST WATER SOURCE (tng/1)
PARAMETER
TOTAL HARDNESS
CALCIUM
MAGNESIUM
POTASSIUM
IRON
COPPER
ZINC
CADMIUM
CHROMIUM
NICKEL
LEAD
ALKALINITY
METHOD*
122B
110C
127B
14 7A
124D
119A
16 5 A
109A
117B
129A
125B
102
TOTAL SUSPENDED SOLIDS 148B
TOTAL SOLIDS
SULFATE
NITRATE
NITRITE
AMMONIA
PHENOL
CHLORINE
CHLORIDE
FLOURIDE
CYANIDE
pH
DISSOLVED OXYGEN
148B
156C
133A
134
132B
U.V. Absorption
114A
Specific Ion Electrode
me
Specific Ion Electrode
-
218B
TEST WATER
132
95
37
1.2
0.220
0.025
0.015
nil < 0.01
nil < 0.01
nil< 0.01
nil< 0.05
90
14
174
25
0.13
0.0230
0.07
nil < 0.25
nil< 0.01
58
1.09
nil< 0.5
7.34
9.85
Section as refers to Standard Methods, 13th ed.
91
-------
Table 17
MORTALITY DATA IN PERCENT DEATHS OF MINNOWS WITH TIME
AFTER EXPOSURE TO 10 ppm OF NINE CHEMICALS AT 60°F
1 HOUR
2 HOURS
3 HOURS
6 HOURS
1 DAY
2 DAYS
7 DAYS
14 DAYS
CONTROL
0%
0%
0%
0%
0%
0%
0%
0%
CHLORINE
WATER
86.7%
100%
-
-
-
-
-
-
10 ppm He
(HgCl2)g
1.7%
96.7%
100%
-
-
-
-
-
ACETONE
CYANOHYDRIN
0%
13.3%
23.3%
43.3%
93.3%
100%
-
-
10 ppra Cd4"4"
(Cd C12)
0%
0%
0%
3.3%
76.6%
93.3%
93.3%
93.3%
AMMONIUM
HYDROXIDE
0%
1.7%
5.0%,
3.3%
45%
45%
46.7%
50%
PHENOL
0%
0%
0%
0%
0%
0%
3.3%
6.7%
10 ppm
Cr+6
(Na2Cr2 0?)
0%
0%
0%
0%
0%
0%
0%
0%
ACRYL-
ONITRILE
0%
0%
0%
0%
0%
0%
0%
0%
METHANOL
0%
0%
0%
0%
0%
0%
0%
0%
VO
10
-------
vo
u>
1000
800
m
tc.
\u
a.
iu
3
600
400
200
6 8
TIME (days)
10
12
14
Figure 23 ALGAE GROWTH CURVES FOR ACETONE CYANOHYDRIN
ACRYLONITRILE AND CHLORINE TESTS AT 10 ppm (60°F ±1°F)
-------
1000 r-
900
800
700
600
I 500
ui
400
300
200 <
CONTROL
O METHANOL
CADMIUM
a CHROMATE
2 POINTS ( A & a )
6 8
TIME - DAYS
10
12
14
Figure 24 ALGAE GROWTH RATE WITH TIME AFTER EXPOSURE TO
10 ppm METHANOL, CADMIUM, AND CHROMATE
94
-------
iooo r
900 -
800 -
700 -
600 -
500 -
400 -
300 -
200
CONTROL
O PHENOL
AMMONIA
O MERCURY
TIME - DAYS
Figure 25 ALGAE GROWTH RATE WITH TIME AFTER EXPOSURE TO
10 ppm PHENOL, AMMONIUM HYDROXIDE, AND MERCURY
95
-------
10
8
£ 6
CHLORINE (10 ppm INITIAL)
I
I
~ 10
i
H 8
i .
2
Q
UJ
«r
4
2
0
1 2
TIME (hours)
PHENOL (10 ppm INITIAL)
J I
_L
I I 1
0 1
10 r
234567
TIME (days)
ACETONE CYANOHYDRIN
(10 ppm INITIAL)
ACRYLONITRILE (10 ppm INITIAL)
23456
TIME (days)
8
01234567
TIME (days)
10
8
6
4
2
0
, AMMONIUM HYDROXIDE
Ny (10 ppm INITIAL)
1 1 1 1 1 1 1
012345
TIME (days)
Figure 26 DISSIPATION OF VOLATILE POLLUTANTS (ppm) WITH TIME
IN AN AERATED SYSTEM
96
-------
Table 18
SAFE AND LETHAL CONCENTRATIONS OF POLLUTANTS
INCLUDING THE LD^O FOR FATHEAD MINNOWS
AND FOR A 7-DAY PERIOD OF OBSERVATION (ppm)
POLLUTANT
ACETONE CYANOHYDRIN
ACRYLONITRILE
AMMONIUMHYDROXIDE
CHLORINE
METHANOL
PHENOL
MAXIMUM
SAFE LEVEL
0.5
20
5
0.05
10,000
10
MINIMUM
LETHAL DOSAGE
1.0
(40)
7.5
0.1
15,000
40
LD50
1.42
30
10
0.165 (4 day)
17,800
56.5
97
-------
Table 19
LENGTH OF TIME BEFORE PHYSICAL EFFECTS AND/OR
DEATHS WERE OBSERVED
POLLUTANT
EFFECTS OBSERVED AT:
DEATHS OCCURRED AT:
METHANOL
10,000
15,000
20,000
25,000
30,000
50,000
ACRYLONITRILE
10
20
40
80
100
ACETONE CYANOHYDRIN
0,
0.
1.
1.
.0
,25
2.5
5.0
10.0
20.0
CHLORINE
0.05
0.
0.
1.
1.
1
5
0
25
2.5
5.0
AMMONIUM HYDROXIDE
1.25
2.5
5.0
7.5
10
20
40
PHENOL
10
15
20
40
80
100
150
200
1 hour - 24 hours
1 hour - 24 hours
15 min. - 24 hours
15 min. - 24 hours
Immediately
Immediately
None
None
24 hours and after
6-24 hours
6-24 hours
None
3-6 hours
30 min. - 24 hours
1 hour - 24 hours
30 min. - 24 hours
30 min. and after
15 min. and after
Immediately
None
3 hours - 30 hours
2 hours - 4 days
30 min. and after
15 min. and after
5 min. and after
Immediately
None
None
None
6 hours - 5 days
None observed
None observed
None observed
None
1 hour - 24 hours
15 min. - 48 hours
15 min. and after
10 min. and after
5 min. and after
Immediately
Immediately
None
30 hours only
24 hours - 48 hours
2 hours - 24 hours
3 hours - all dead
30 min. - all dead
None
None
5 days and after
12 - 24 hours
12 - 24 hours
None
None
24 hours (20%)
None
3-24 hours (80%)
1-1/2% - 24 hours (100%)
2-1/2% - 24 hours (100%)
1-1/2 hours (100%)
None
24 hours - 48 hours (60%)
24 hours - 4 days (70%)
3 hours - 8 hours (100%)
2 hours - 6 hours (100%)
1 hour - 4 hours (100%)
30 min. - 1-1/2 hours (100%)
None
None
None
24 hours - 5 days (60%)
3 hours - 24 hours (50%)
45 min. - 75 min. (100%)
30 min. (100%)
None
None
None
48 hours
24 hours
24 hours
1 hour -
1 hour -
- 6 days (30%)
- 6 days (90%)
- 6 days (80%)
4 days (100%)
3 days (100%)
98
-------
Table 20
PERCENT SURVIVAL OF FISH WITH TIME DURING EXPOSURE
TO A PHENOL-CARBON SEDIMENT
(AQUA NUCHAR A; POWDERED CARBON)
TIME
INITIAL - 5 DAYS
6-12 DAYS
13 - 21 DAYS
22 DAYS
23 DAYS
24 DAYS
25 DAYS
28 DAYS
29 AND 30 DAYS
PHENOL TEST
100%
96%
92%
88%
80%
72%
68%
68%
68%
CONTROL
100%
100%
1002
100%
100%
100%
100%
98%
98%
Table 21
PERCENT SURVIVAL OF FISH EXPOSED TO PHENOL-CARBON SEDIMENT
(GRANULAR CARBON)
TIME
INITIAL
7 DAYS
15 DAYS
30 DAYS
AMOUNT OF PHENOL
IN WATER
0.1 - 0.3 ppm
niKO.10 ppm
nil< 0.10 ppm
0.9 ppm
% SURVIVAL
IN PHENOL
100%
95%
78%
45%
% SURVIVAL
IN CONTROL
100%
97%
86%
65%
99
-------
Table 22
PERCENT SURVIVAL OF MINNOWS EXPOSED TO A SEDIMENT OF
ACRYLONITRILE-GRANULAR CARBON AND SAND
TIME
0-3 HOURS
1 DAY
2 DAYS
3 DAYS
4 DAYS
% OF FISH
SURVIVAL
100%
80%
48%
30%
0%
DUPLICATE TEST
100%
100%
100%
-
0%
CONTROL
100%
100%
100%
100%
100%
100
-------
effects by themselves. Therefore, it was concluded that activated carbon with
absorbed pollutants cannot be allowed to persist at the stream bottom in natural
waterways supporting mixed biological communities.
Another countermeasure developed during the HASP program was the immobili-
zation of hazardous liquid spills with a powdered gelling agent. Bioassay tests
in which living organisms were challenged with gel-immobilized benzene and
acrylonitrile showed that here, again, rapid adverse effects were noted (cul-
minating in fish death). Thus, it is necessary that even gelled hazardous
liquids be removed from biologically productive watercourses as rapidly as
possible. On the other hand, it was noted that exposure of fish to high solids
concentrations (ranging up to thick slurries of carbon in water with concen-
trations of 100,000 ppm) could, over short contact times, be tolerated by fish
in the absence of a hazardous organic pollutant.
Another countermeasure developed under this program was the addition of
sodium sulfide solutions to pollutant plumes of dissolved heavy metal salts on
the theory that the precipitated metallic sulfide ores would be innocuous.
Bioassay tests on this system indicated that the sodium sulfide itself, probably
because of its very high natural pH, was also a toxic to varying degrees over
long water contact times (60 days) and that the dissolved secondary sodium salts
produced in the water column could also, on occasion, have adverse biological
effects. These long term studies with sodium sulfide and the heavy metal sulfide
precipitates also included exposure of a plant, the hornwart, in order to assess
the potential deleterious effects on water plants. In some instances, severe
deterioration of this plant was noted upon exposure to the supposedly inert
heavy metal ores residing at the bottom of the test tanks. All of these find-
ings are documented in Tables 23, 24, 25 and 26.
Recommendations for future study include providing the substantial amount
of missing information on the relative toxicities of a greater variety of haz-
ardous chemicals (as determined through static bioassay tests such as those
reported here). It is also to be recommended that determinations be made of
the tests organisms' (fathead minnows) ability to recover after a brief initial
exposure to a high concentration plume of toxicant and to exposure of various
toxicant levels in a continous flow system.
101
-------
Table 23
ESTIMATES OF TOXICITY TO FISH OF VARIOUS METALS
METAL COMPOUND
BARIUM Cl
N03
CADMIUM Cl
N0_
S04
CHROMIUM (+3)
(Na) CHROMATE
(Na) DI CHROMATE
(K) CHROMATE
COPPER Cl
N03
804
COBALT Cl
N03
MANGANESE Cl
C03
S04
NICKEL Cl
N03
so4
Pb Cl
N03
S04
SILVER N03
ZINC Cl
N03
504
TOXIC
CONCENTRATIONS
(ppm)
50
200
0.9 (SOFT H20)
5.0 (HARD H20)
6
0.3-7
1000
5.2
1.2
40
300-400
60
75
3.3
0.18
0.15
16
10-15
15
300
40
500
5
1
10-16
0.5
10
25
0.04
15
5.7
3
FISH USED
SALMON
"FISH"
FATHEAD MINNOWS
FATHEAD MINNOWS
FUNDULUS
STICKLEBACK
FISH
TROUT
STICKLEBACK
MINNOWS
FISH
FISH
FISH
FISH
FISH
TROUT
FUNDULUS
STICKLEBACK
FISH
FISH
FISH
FISH
FISH
FISH
GOLDFISH
FISH
FISH
FISH
FISH
FISH
FISH
FISH
TYPE OF TEST
72 HOUR
96 HOUR
96 HOUR
96 HOUR
__
3 HOUR
24 HOUR
INCIP- TOX. LEVEL
5 DAY
__
200 HOUR
102
-------
Table 24
DATA ON THE TOXICITY OF Na2S TO MINNOWS
(% DEATHS WITH TIME)
TIME
INITIAL
0.25 HOURS
0.5 HOURS
1.0 HOURS
1.25 HOURS
3 HOURS
6 HOURS
24 HOURS
48 HOURS
72 HOURS
7 DAYS
INITIAL pH (AVE)
CONCENTRATION (ppm)
0
0
0
0
0
0
0
0
0
0
0
0
8.33
10
0
0
0
0
0
0
0
0
0
0
0
8.93
20
0
0
0
0
0
0
0
0
0
0
0
9.14
40
0
0
0
0
0
0
0
0
10%
10%
10%
9.50
80
0
0
0
0
0
0
0
10%
20%
20%
20%
9.95
100
0
0
0
0
0
0
0
20%
30%
30%
30%
-
200
0
0
0
80%
100%
-
300
0
10%
50%
100%
11.42
500
0
60%
100%
11.70
1000
0
100%
11.15
-------
Table 25
PERCENT DEATHS OF MINNOWS WITH TIME AFTER EXPOSURE TO
NaCl, NaN03, & Na2s°4
DAYS
1
2
3
4
5
6
7
CONTROL
0%
0%
0%
0%
0%
0%
0%
1000 ppm
Na2SO,
0%
0%
0%
0%
0%
0%
0%
1000 ppm
NaNO
20%
20%
20%
20%
20%
20%
20%
1000 ppm
NaCl
30%
30%
30%
30%
30%
40%
50%
104
-------
Table 26
SUMMARY OF DATA FROM METAL PRECIPITATE TESTS
METAL
PRECIPITATE
TESTED
CONTROL
ZINC SULFIDE
CADMIUM SULFIDE
LEAD SULFIDE
COOPER SULFIDE
COBALT SULFIDE
SILVER SULFIDE
NICKEL SULFIDE
BARIUM CHROMATE
CONTROL
MANGANESE SULFIDE
CONTROL
RANGE OF METAL
CONCENTRATION IN
TEST H20
THROUGHOUT
TEST (ppm)
-
0.01-1.20
0.005-0.70
0.35-0.67
0.05-0.30
2.4-7.0
niKO.Ol
2.9-8.15
3.0-4.5 (Ba+2)
0.40-0.80 (CrO^j)
3.0-18.25
-
EQUILIBRIUM
CONCENTRATION OF
METAL (ppm) FROM
THE SOLUBILITY
OF THE SULFIDE
-
5.0
t
1.0
0.8
0.2
3.0
0.1
2.4
1.0
3.0
-
TO XI CITY
RANGE
TO FISH
(FROM
LITERATURE)
-
0.3-15
0.01-7
0.1-25
0.01-3.3
7-16
0.04
0.08-16
50-200 (Ba+2)
75-400 (Cr04)
40-500
-
% SURVIVAL
OF FISH
AFTER 60 DAYS
95.4
100.0
98.0
94.0
92.0
80.0
74.5
41.0
53.0
49.0
63.0
61.0
CONDITION OF THE
CERATOPHYLLIUM
AT TERMINATION
HEALTHY AND WITH
GREEN FOLIAGE
HEALTHY AND WITH
GREEN FOLIAGE
NO LEAVES, ONLY
A GREEN STALK
HEALTHY AND WITH
GREEN FOLIAGE
20-30% WITH LEAVES
HEALTHY AND WITH
GREEN FOLIAGE
20-30% WITH LEAVES
DISINTEGRATED TO
A PILE OF DEBRIS
HEALTHY
HEALTHY
HEALTHY
HEALTHY
O
01
-------
SECTION 12
MATHEMATICAL MODELING OF HAZARDOUS SPILLS
AND SPILL TREATMENT
In the course of deploying treatment material for a hazardous chemical
spill in a water body, it is desirable to understand and characterize two
important phenomena: (1) the time rate of consumption of the pollutant as
a result of treatment; (2) the spreading or dispersion of the pollutant
throughout the fluid. Since these are extremely complex phenomena to examine
in real life, it becomes necessary to resort to mathematical models to suit-
ably reproduce the above phenomena. The primary objective of such models is
to simulate a variety of spill and treatment phenomena and to thereby facili-
tate the prediction of whether treatment in certain cases is feasible or not.
Both types of modeling were considered on this program. In the subsequent
subsections, we will describe first a mathematical model of activated carbon
treatment for removal of pollutants from solution. This model, based on para-
meters which are readily determined in the laboratory for any carbon-pollutant
system, is useful for describing the rate of decrease of pollutant concentra-
tion as a result of carbon treatment. The dispersion models developed on the
program are described thereafter.
1. MATHEMATICAL MODEL OF ACTIVATED CARBON TREATMENT
For the purpose of developing the rate model for the adsorption of pollu-
tant, it was assumed that at time zero, V liters of solution containing the
contaminant A are contacted with M grams of activated carbon in tea bags.
Furthermore, it is assumed that the overall rate of uptake of A by the carbon
depends on the following three physical steps.
1. Diffusion of A, from the bulk solution, through a very thin film of
relatively stagnant solution surrounding the carbon particles.
2. Diffusion of A through the internal pores of the activated carbon
particles.
3. Physical adsorption of A on an active site within the carbon particles.
In treating Steps 1 and 2, very simple rate equations are used, and tempera-
ture and concentration gradients within the external film and within the carbon
particles are not taken into account. Thus, film and pore diffusion mechanisms
(Steps 1 and 2) are treated as a single diffusional resistance. With these
assumptions, the appropriate rate of depletion of A via film and pore diffusion
is given by:
106
-------
dt
, (C - C )
fa as
(1)
where
TQ2 A.
"Rf = rate of diffusional depletion of A, yrf r~~
C = concentration of A in bulk solution, j°
C = concentration of A in pore solution within carbon
particle,
= diffusional transfer coefficient, hr
-1
t - time, hr
The third step, namely, physical adsorption of A on an active site within
the carbon particle, can be well represented by the following reversible second-
order reaction:
A in pore solution with-
in carbon particle
Available active
sites within
carbon
Adsorbed complex
on active sites
i.e.,
[A] + [Sorbent] ^ [A-Sorbent]
(2)
Based on Equation (2) the net rate of uptake of A by adsorption on active sites
within the carbon particles may be written
k c
a as
j
d
(3)
where
mg A
a
net rate of adsorption uptake of A, r3~
107
-------
q = concentration of A adsorbed on carbon surface,
1 ^VfcA^^-L&w J. UI..J-VS1.1 W J. fi C&U0W J. 1*T ^U WLl \~a.J. LS WLl OUJ.JLCLV~C« ~
a g carbon
liters
k = adsorption rate constant, . ,
a r ' mg A-hr
Q = ultimate adsorption capacity of carbon, °r
A
(Q-q ) = available adsorption capacity of carbon, °-r
k, = desorption rate constant, hr"~
In writing Equation (3), it is assumed that the local adsorption at an active
site is proportional to the liquid phase concentration within the particle
C and the available adsorptive capacity (Q-q ). The desorption is assumed
to be proportional only to the quantity of A adsorbed, q . Equation (3) may
be put in an alternate form:
dqa
R = : = k [C (Q-q ) - K q ] (A)
a dt a as a a
where
kd m A
K = equilibrium constant -, -rf
^ k ' liter
a
We have thus far identified two rate expressions: the rate of film diffu-
sion with pore diffusion of A from bulk solution to an active site as given by
Equation (1) and the net rate of adsorption of A onto an active site as given
by Equation (4). It is now necessary to combine these two rate expressions so
that the unmeasurable quantity Cas is eliminated and an overall rate expression
is obtained in terms of Ca, the measurable quantity. This is done by assuming
that there is no continual accumulation of A in the film and the carbon interior
but rather that all A transported by film and pore diffusion is taken up by
adsorption within the carbon particle. This can only happen when the rates in
Equations (1) and (4) are equal, and therefore
R = VR- = MR (5)
I cl
where
A
R = overall rate of uptake of A,
108
-------
(Note that the quantities of V and M in Equation (5) are used in order to satisfy
the dimensional units on both sides of the equalities in Equation (5) .)
Therefore, substituting Equations (1) and (4) into Equation (5) and elimi-
nating Cas, the following overall rate expression is obtained:
-VdC C (Q-q ) - Kq
R -- * = a V - ^a
dt ^ CQ-qaJ
Mka * VKf
If it is assumed that the quantity of A in the pore fluid and in the stag-
nant film is negligible, then qa can be expressed as
q = (C - C )
Ma ao a M
= (C - C ) e
ao a
where
mg A
(7)
C = initial concentration of A, -..
ao liter
£ = V/M
Equation (7) says that the amount of A adsorbed on the carbon interior is equal
to the amount of A depleted in the bulk solution. Substituting Equation (7)
into Equation (6) finally yields
-dC C 2 + AC + B
a a a /«%
dt * 1 + 1 (C + C)
k k a
a f
where
A = Qf - C + K
x ao
B = -KC
ao
C = Q' - C
ao
Q' = Q/e
109
-------
Equation (8) is .the desired rate expression which describes the depletion
of A in the bulk solution in terms of the bulk phase concentration Ca> There
are also four physical parameters of interest in the above rate expression;
namely, Q, K, ka and kf. The first two parameters, Q and K, are equilibrium
parameters and may be determined as follows. At equilibrium, ,a = 0 and
hence Equation (4) becomes
k [c (Q-q ) - Kq~|= 0
a |_as x na MaJ
i.e. ,
c c
as K
qfl Q Q
d
A known quantity of A can be stirred in a beaker with a known quantity
of carbon for an extended period of time at the end of which equilibrium may
be assumed to have been attained. At equilibrium, the bulk concentration Ca
is the same as the pore concentration Cas; also, qa is obtained by Equation
(7) . This is repeated for different concentrations of C^o and carbon quanti-
Cas
ties M so that enough points are obtained for a plot of - versus C .
Q cLS
na
According to Equation (9) , this plot should yield a straight line with slope
1 K
^ and intercept ^-. Such a plot is shown in Figure 27 and the values of Q and
K are determined as 154.288 and 14.387 mR phenol respectively0
g carbon r J
The third parameter, k , would be expected to depend on the type of carbon
used and also the solute A \as also would the first two parameters, Q and K) .
The fourth parameter, k-, would largely depend on the physical environment such
as the type and degree of agitation and local turbulence in the vicinity of
the carbon particles. Parameters ka and kf would have to be determined by
fitting the rate expression in Equation (8) to actual concentration vs time
data obtained from controlled laboratory tests under different flow conditions.
The parameter ka is best determined under conditions where resistance due
to film transfer is expected to be the least, whereby all resistance to the
transfer of pollutant from solution to carbon can be assumed to be due to
adsorption only. Such a condition can be realized in a beaker in which phenol
solution and loose granular carbon is vigorously- agitated or stirred. In this
manner of contacting, film transfer resistance (*- ) is negligible and Equation
(8) reduces to f
(10)
110
-------
4000
3000
2000
1000
C0 - 100 mg/
NITIAL PHENOL CONC.
g CARBON
K - 14.387 mg PHENOL
100
200
300
400
500
600
700
Figure 27 EQUILIBRIUM ADSORPTION OF PHENOL BY NUCHAR (8 x 30)
111
-------
The parameter k in Equation (10) was therefore manipulated such that
the concentration-time history predicted by Equation (10) was in close agree-
ment with experimental results according to a least squares error criterion.
liters
The value of k determined by such a fit was found to be 0.002167 phenol-hr "
Figure 28 shows the close fit obtained between model and data.
Since the adsorption rate constant ka depends on a phenomenon occurring
largely within the carbon, it will always be constant for a given carbon and
pollutant regardless of how the carbon is deployed (tea bags or otherwise) or
the external physical environment in the vicinity of the carbon particles.
In direct contrast, the film transfer coefficient, kf, depends almost entirely
on the local turbulence, mixing and agitation in the vicinity of the carbon
particles. Therefore, in evaluating kf, attention must be paid to the kind
of flow situation at hand.
In the estimation of kf, channel tests were used to generate concentration
vs time data. These data were obtained at different flow velocities, and k-
was determined for each flow velocity such that Ca vs t as predicted by Equa-
tion (8) was in close agreement with experimental channel data.
Figure 28 shows concentration-time values derived from Equation (8) as
compared with experimental data. There are three sets of channel data pre-
sented in Figure 29; these are at flow velocities of 0, 0.5, and 1.0 ft/sec.
In each case, the carbon dosage is 250 mg/liter. The 0 ft/sec data were
actually generated in an unstirred beaker of 250 mg/liter phenol solution
treated with a lOx dose of carbon in a tea bag. Examination of Figure 29
shows that the experimental data for each run (0, 0.5, and 1.0 ft/sec) may
be well represented by the rate expression in Equation (8). The calculated
curves generally fit the -data in a satisfactory manner. Also, the film trans-
fer coefficient increases with increase in flow velocity, which is to be
expected.
The results of the parameter estimation of the treatment model are sum-
marized in Table 27.
2. CONCLUDING REMARKS
Mathematical modeling is of great interest in that it furthers the under-
standing of spill dispersion and treatment, and displays in considerable detail
the interactions between various parameters of interest. More importantly, per-
haps, models are potentially important tools in aiding practical decisions.
In the next subsection, we discuss the modeling of the spill dispersion and
the treatment of spills. The actual coupling of these two models occurs via
Equation (11) when the R(C) term is added. However, from the complex nonlinear
form of R(C) as can be seen in Equation (8), no simple analytical solution (as
in Equation 12)) is possible. The only way to obtain solutions is by numeri-
cal techniques implemented on a computer. Furthermore, sufficiently fine resolu-
tion may require thousands of points of storage capacity and this figure is then
to be multiplied by the number of elements (flow components, stream depth, and
so forth) pertinent to the model. The finer the spatial resolution, the smaller
112
-------
250
002167 LITERS
mg PHENOL-hr
8 9 10
Figure 28 CONCENTRATION-TIME VALUES DERIVED FROM EQUATION 8
(CURVE) COMPARED WITH EXPERIMENTAL DATA (POINTS)
113
-------
250
ft/sec, kf-0.0151 hr'1
5 ft/iec, kf - 0.252 hr'1
0 ft/sac, kf - 0.567 hr'1
9 10
Figure 29 CONCENTRATION TIME VALUES DERIVED FROM EQUATION 4
(CURVES) COMPARED WITH EXPERIMENTAL CHANNEL DATA
(POINTS)
114
-------
Table 27
SUMMARY OF PARAMETERS ESTIMATED
IN TREATMENT MODEL (PHENOL-NUCHAR 8 x 30)
PARAMETER
Q
K
k
a
kf at 0.0 ft/sec
kf at 0.5 ft/sec
kf at 1.0 ft/sec
ESTIMATED VALUE
154.288
14.387
0.002167
0.0151
0.252
0.567
UNITS
mg Phenol
gm Carbon
mg Phenol
liter
liters
mg Phenol-hr
1/hr
1/hr
1/hr
115
-------
the time step permissible without violating computational stability. All this
obviously calls for computers with vast storage capacities and speedy computa-
tional abilities. However, this certainly does not present any obstacle since
rapid advances in computer design have made such computation possible.
The major limitation is physical in nature. As discussed later,
the incomplete knowledge of dispersion coefficients hampers realistic dynamic
simulation of the dispersion model. Also, in the case of the treatment model,
various carbon-pollutant combinations must be analyzed to determine constants
such as Q, K and ka. These physical coefficients must be established in
order that the simulation of hazardous spills and their treatment is not left
to guesswork and speculation. These various constants and coefficients could
be set up in some form of unified information retrieval system such that for
a given spill in a specific water body all the necessary parameters are made
available to the individual (s) who are faced with the decision of where and
when to treat a given spill. It is in this capacity that the greatest benefit
from mathematical models will be realized.
3. MATHEMATICAL MODEL OF POLLUTANT DISPERSION IN FLOWING STREAMS
The equation which governs the distribution of a pollutant in a continuum
is the classical mass conservation equation
,1 *c M-n ^ / ) ** - K f2t ... k ^c *. k 7zc Rfc.)
u -^ "K "R ;
where
C = concentration of the pollutant
u,v,w = velocities of the fluid in x, y, and z directions
K .K ,K = turbulent or eddy dispersion coefficients,
X V Z
3 respectively
x,y,z = longitudinal, lateral and vertical coordinates,
respectively
R (C) = rate at which pollutant is being consumed
t = time
At 'present, we shall confine ourselves to Equation (11) without the
term R(C) . We shall discuss this term in more detail.
The solution to Equation (11) (without R(C)) is
116
-------
. (12)
K?
where it is assumed that u is the only non-zero stream flow velocity and K is
given by
K =
where M is the total mass of the spill.
Equation (12) represents the dispersion of material spilled at a specific
point. An isoconcentration line would appear as an expanding ellipse whose
center of symmetry is moving with a velocity u in the x direction. After a
while the ellipse would begin to contract, representing the exhaustion of
material within the boundary line. Figure 30 shows such elliptical contours
at different time instants in a nonflowing water body. The contours represent
the motion of the 10-ppm concentration boundary resulting from a 5000-gallon
phenol spill, 100 meters from the shoreline. Note that after about 24 hours,
the 10-ppm boundary has reached a maximum and begins to recede toward the
shoreline. This example serves to illustrate the natural dilution process
involved and also the order of magnitude of time and distance necessary for
dilution.
The necessary computations were performed on a Model 9810 Hewlett-Packard
desktop calculator, while the results were plotted via an associated Hewlett-
Packard plotter. The Model 9810 can be card programmed for up to 2000 program
steps. Equation (12), along with a portable calculator equivalent to the Model
9810, allows the user a great deal of flexibility in that "on-site" calculations
can be made while the spill is still spreading downstream. Rapid checks can
be made on the spread and location of the flowing spill since computation time
is of the order of a few seconds.
The values of Kx, Ky, and Kz used in this example are typical or "ballpark"
values for nonflowing fluids. In the case of moving fluids, the convective and
turbulent eddy effects have a marked effect on the dispersion coefficients. The
major drawback to a realistic simulation of the dynamic process is the incomplete
knowledge of the turbulent transport processes. The dispersion coefficients,
in general, vary both in space and time in a manner that is not well understood.
The question of how these coefficients vary with flow and/or channel geometry
needs to be investigated both experimentally and analytically to aid in the
application of models to the prediction of spill behavior.
Until more complete information becomes available on representative values
of dispersion coefficients for different types of water bodies under different
conditions, it appears that some on-site adaptive prediction procedure will be
required to provide the field worker with information needed for decision-making.
Recognizing that the field Corker needs information quickly but has no requirement
117
-------
oo
250
r
200 -
150 -
ui
UJ
100 -
PLUME BOUNDARY = 10 ppm CONTOUR
K
0.1 m2/sec
0.05 m2/sec
Kz = 0.0005 m2/sec
t = 24 HOURS
t = 6 HOURS
t = 1 HOUR
200 250
METERS
300
350
400
450
Figure 30 DISPERSION OF A 5000 GALLON PHENOL SPILL IN A NONFLOWING WATER BODY
-------
for extreme accuracy in predicted data, it appears that a simple model such as
described above could serve adequately for the calculations. Initial calcula-
tions could be made with assumed values of dispersion coefficients that are
based solely on experience of the operator. Simultaneously, field data could
be gathered on, for example, concentration as a function of position and rough
isopleths could be plotted. Comparison could then be made of these rough iso-
pleths and model-predicted contours to determine the nature of the error in
the assumed dispersion coefficients. Adjustments could be made until a reason-
able fit is achieved and the values of the coefficients thus selected could be
used to make predictions for longer times in the future. Repetition of the
procedure could provide upgraded predictions and account for different circum-
stances as they change with time or position in the water body.
During the course of the above modeling study, a parallel analytical effort
was directed toward developing a flow model which would permit the analysis of
not only the concentration distribution of the pollutant in the water body but
also the analysis of the mean, variance, skewness and kurtosis of the concen-
tration distribution. In addition, the model would require very simple numeri-
cal techniques to obtain solutions.
The mathematical model generated in this parallel effort was derived from
the basic diffusion equation for a passive contaminant in a moving environment.
In addition to considering diffusion, the possibility of incorporating a chemical
reaction effect was included. This was considered to be of importance for the
present study since the effect of a treatment on the dispersing pollutant was
of interest. Since the scope of this analytical effort was to be limited, a
two-dimensional model was considered realistic.
Consider a two-dimensional stream, as shown in Figure 31, with a velocity
profile U(z). Let it be assumed that the turbulent transport of material
Figure 31 TWO-DIMENSIONAL STREAM
119
-------
parallel to the axis can be described by eddy diffusivity coefficients Kx and
K . Let Kx and K- be independent of x and t and be at most functions of z.
If the concentration field of the passive contaminant is written as C(x,z,t),
the governing differential equation may be written as (Ref . 1) :
-------
Although it is not necessary to do so in employing the moment method,
the assumption was made to let Kx and Kz be constants for the cases to be
considered in this program. Therefore combining Equations (14) and (15),
it can readily be shown that the resulting series of equations are formed:
« 2e
a ^-°- - K y ° = 0
" -v *. " y o u
o ar t z o 2
z
e 2e
a . * ! _ y & 1 = IT fzl 6
V 7 t Kz 7 2 U Cz) 9o
z
V
e 2e
T-- o- O
93! FT - Kz TT ' 6Kx ei + 3 u
Z
e V
0 . UL _ v ? 4 = i2K e + 4 u
9. T7 K- <> o LZ v °O H
4 fft Z ? £ TS. f.
Z
- d ^ &
or in general _£_ZL - k *" . p.. (3- ,*;
The equations in (16) are solved sequentially since for the nth
equation Fn (z,t) involves only elements 00,..., 9n_j. The above equations
are solved subject to the boundary conditions that
*ei
% = 0 at z = 0, 1: i.e.. at the bottom of the shear
a z
and at the free surface. It is also assumed that the contaminant is deposited
in the stream instantaneously att=0atx=z=0. CRe£ used to nondimen-
sionalize C(x,z,t) in Equation (13) is chosen so that
1
e (z,t) dz = 1
o
o
121
-------
The left-hand side of (16) has the form
T e - A_K J^i i- o 1 ...
L 9i ~ ?t z 3- 2 1 0»-L'-"
z
If one obtains the Greens function for this operator L, then the moments 6^
can be expressed by
r t 1
9 - G(z,z, t-T) F. (z. T) dzMT (18)
i \J J i 1
o o
where G(z,z, t-T) is the Greens function. Thus it can be seen that the
original problem of solving Equation (14) has been reduced to a problem of
quadrature rather than finite differences and should in theory be simple to
perform. The function G(z,z] t-T) has been developed for an analogous heat
transfer problem in Reference 3 and is therefore readily available. The first
four moments 0O, 0]_, ©2, 83 have been programmed on a computer.
Before passing on to a discussion of the calculated moments of the concen-
tration distribution subject to dispersion alone, let us consider a possible
means of incorporating a sink or consumption rate term. Since the method of
moments is inherently limited to a linear problem, the form of rate term must
be linear. Therefore for the present work this rate was taken to be propor-
tional to the concentration itself. That is
R = k C(x,z,t) (19)
where RREF = ^REF* since laboratory measurements of k were available for certain
reactions. Physically, this reaction can be interpreted as follows: at some
time T after the initial spill has occurred, the treatment is dispersed uni-
formly over the zone in which the pollutant is found.
The concentration at each point decreases (as a result of treatment) at
a rate proportional to the concentration. The inclusion of (19) into Equation
(14) and the subsequent development of the moment equations proceeds in a
straightforward manner. Equations similar to (16) result except the operator
defined in (17) is replaced with the operator
(20)
such that
t
T Q = 1
i
T' fl - F <21>
L e - F
122
-------
where the right-hand side F^ has the same form as the right-hand side of (16) .
If a transformation _ -ktrr is defined, the system (21) may be reduced to
the set: i ~ e i
a-t
F. (z,t)
(22)
which has the identical form as the set (16). Thus one may proceed as follows
when analyzing a dispersing reacting system where R is given by (19) : (i)
Compute the solution 0^ to (16) subject to the boundary conditions for the
nonreacting system for O^t^-T; i.e., if a treatment is added at a time
the solution takes the form r
e± (z,t)
T< t<: T
(z,t) = e (z,t)
e-k(t- )
where 0^ is given by the output of the computer program. This scheme permits
the analysis of a number of treatment situations to be applied for effect on
a .single computer run. Noting that the variation of dimensionalizing param-
eters also permits consideration of a number of physical cases for a single
value of a dimensionless constant, it is clear that the present formulation
can yield a tremendous amount of physical insight for a single computer run.
In order to be able to translate the meaning of the variables 00> QI,...,
03 into a physical interpretation on the quantity of interest, the concentra-
tion distribution, the following parameters were introduced.
Mean m1 = e,/0 ,
1 o
Variance cr
1/2
Skewness 2T- =
1 3/2
123
-------
Kurtosis
2 =
o
e2/9o -
-3
The values of the parameters m, cr , If , and If _ characterize the general fea-
tures of (C(x,z,t). For example, the mean, m, can be viewed as describing
the movement of the centroid of the concentration distribution at a particular
value of z. The variance,cr, provides a measure of how the pollutant in each
layer is dispersing relative to its centroid.
The skewness and kurtosis of a distribution relate the symmetry of the
distribution about the centroid and the behavior of the concentration distri-
bution far from the centroid, respectively.
An example of the mean and the variance behavior for two sets of physical
parameters are shown in Figures 32 and 33, respectively. The physical para-
meters for the cases considered are
Case 1.
=5.15
Case 2.
K
x
5.15
= 5.1 x 10
U(z) = z1/7
K = 5.15 x 10
z
U(z)
1/7
= 0.408
= 0.408
Recall in interpreting these cases the dimensionless variables Kx, Kz, U(z) and
z have the definitions:
^S *-U*
K K _
K = -5, K = ^ U(z) - U(z)/U, z = "zVD
DU DU
where quantities with -v are dimensional. As a result, the above two cases
can be considered appropriate for a number of physically realistic situations
with different values of the parameters "K , !c , Tj(z) , D, and II- The only
restriction is that the combinations musl^yie^ valv-s of the parameters used
in Case 1 or 2.
Detailed discussion of the mean, variance, skewness, kurtosis, examples
of the effect of treatment for realistic values of k (in Equation (19)), as
well as a comparison of the dilution effect versus the neutralization effect
on the concentration, will be included in the detailed report on modeling.
124
-------
to
l/l
N
N
II
E
<
in
I
U(Z) = z177 i
r i i
1000
2000
t: TIME (DIMENSIONLESS)
3000
e
1
Figure 32 MEAN, m = -^ (Zx, t) vs t
00
-------
to
10
1000
2000
t: TIME (DIMENSIONLESS)
3000
figure 33 VARIANCE
-------
SECTION 13
HAZARDOUS CHEMICALS/COUNTERMEASURES
The matrix presented in this section is a compilation of recommended pro-
cedures for treating and controlling some 250 recognized hazardous liquids
having a high probability of being involved in spills into or near watercourses.
The chemicals are listed in alphabetical order. After each compound name is an
"instant recognition" symbol designed to trigger in the minds of an environ-
mental control team the selection of specific countermeasures to combat spills
of that compound, that selection depending for the most part on the physical
behavior of the compound in water (e.g., floats vs sinks and soluble vs
insoluble) as well as its chemical nature.
The symbols shown are in the form of arrows whose thickness and orientation
visually suggest the recommended treatment. A bold upright arrow, for example,
identifies those liquids which are lighter than water and essentially water
insoluble. The hazardous spill control team should therefore immediately consider
the application of oil booming and skimming as the most effective technology to
control such a spill in a watercourse. Conversely, a dense arrow pointing down-
ward indicates that the material is heavier than water and essentially water
insoluble (ten parts per million or less solubility) in which case lagooning,
trenching, and sump or vacuum pumping should be considered to localize and
remove that spilled substance.
A less dense arrow curving upward (i.e., toward the surface) indicates to
the investigator that liquids so marked in the matrix are lighter than water and
soluble in the water. He will thus recognize immediately that massive spills of
such a substance will be mostly localized at, or immediately below, the water
surface, and that the greatest plume dimension (as the spill spreads) will remain
near the surface. Oil booming and skimming technology will be effective in early
stages of such spills. Carbon adsorption, neutralization or precipitation would
be required for the portion of the spill that is in solution. The most difficult-
to-treat, hazardous chemical is one identified by an arrow which curves first
downward and then upward to indicate that the substance is heavier than water
and highly water soluble (greater than 10 ppm, when known toxic levels are less
than or equal to this concentration). The maximum plume dimensions of this
spreading pollutant mass will occur at the stream bottom and the most useful
countermeasures will be vacuum pumping of the concentrated pools in the stream
bottom soon after spill occurrence followed by activated carbon adsorption,
neutralization, and/or precipitation.
The first column to the right of the instant recognition symbols shows the
hazard rating of each substance as taken from previously reported work by the
Environmental Protection Agency. These numbers range from 1 to 250, with the
lower numbers indicating the greater hazards. The ratings were deduced from a
computer program and series of algorithms which took into account such factors
as the abundance of each material manufactured and transported, and its relative
127
-------
toxicity. The subscripts f'a" and "b" appended to some of the ratings
in the column denote which of the essentially water-soluble liquids are lighter
than water (a) and heavier than water (b).
The matrix also shows the solubility of each liquid in water at normally
encountered temperatures, the density of the liquid, and (when known either
from our own studies or from other sources) the approximate lower concentrations
of the chemical in water that kill fish of some specie.
The remainder of the countermeasure matrix specifies the individual coun-
termeasures which are effective for controlling that chemical if spilled on
the ground or into water. The eventual goal of any spill treatment is either
the complete neutralization or the removal of the spilled material to eliminate
hazardous and esthetic consequences. The immediate goals of treatment will
frequently be to provide for the safety of personnel and property. In many
spills the major hazards will involve toxic vapors and potential fire. These
hazards must be considered first. The spilled material should be identified
and appropriate medical and fire fighting personnel consulted immediately.
Citizens may have to be evacuated from the affected area and proper equipment
made available to protect the spill treatment teams. The application of fluro-
carbon-water foams will be useful to minimize the evaporation of spilled material,
and thereby reduce both the toxic and fire hazard for those materials so indi-
cated in the matrix.
Another objective of the treatment will often be to prevent the spread
of the spilled material. A variety of common-sense mechanical countermeasures
may be obvious to even the untrained worker. Solids that have been spilled on
the ground may be covered with plastic sheets, for example, to prevent dissolu-
tion in rainwater. Trenches may be dug for the same purpose or to arrest the
surface flow of spilled liquids. Such countermeasures are applicable to spills
of virtually any material and frequently may be applied before the arrival of
special facilities.
The use of mechanical barriers is lumped in the matrix with chemical
barriers, such as antiwetting agents that minimize ground percolation and arrest
surface flow. More testing is required before specific chemical barriers can
be recommended, however.
Other specific treatments of land spills included in the matrix are those
which eliminate hazards without removing the material (neutralization of acids
and bases and precipitation of heavy metals) and those which immobilize the
spill and alter its form to facilitate mechanical removal (use of gelling agents
and powdered materials to absorb the liquid bulk).
Countermeasures which are effective against hazardous materials spilled
into water are also shown. Countermeasures for floating liquids include the
use of natural barriers or oil spill control booms to limit spill motion for
that class of materials that float on water; the use of surface active agents to
128
-------
compress floating materials into smaller areas or move them from otherwise in-
accessible areas; and the injection of the "universal gelling agent" to solidify
liquids that are either free floating, compressed by surface active agents, or
trapped in booms. As with land spills, the use of mechanical means to remove
the immobilized masses of pollutants is required after any of these treatments.
Removal should be accomplished as soon as possible to minimize aquatic damage.
For liquids which are more dense than water, the effective countermeasures
include the use of natural deep water pockets, excavated lagoons, or sandbag
barriers to trap the dense material at the stream bottom before removing it
with suction hoses while it is still in its liquid state. Alternatively, the
"universal gelling agent" can be injected to solidify the trapped mass in place
for subsequent removal by standard dredging methods.
For those materials of an acidic or basic nature, the countermeasures
indicated on the matrix include the neutralization with weak acids and bases,
as appropriate, as the safest possible technique. Highly water-soluble materials,
especially those which are more dense than water, disperse rapidly throughout
the water volume. For most of these materials the countermeasures are limited
to the application of activated carbon or ion exchange resins. To avoid secon-
dary pollution problems in either case, the treatment materials must be pack-
aged in such a way as to permit their complete removal by mechanical means.
In the case of spills of soluble heavy metals salts, which are also more
dense than water, the application of stoichiometric amounts of sodium sulfide
solutions is recommended to precipitate the heavy metals as their sulfide ores.
Because of the potential secondary pollution, this treatment is recommended
only for small spills except where special equipment is available for measuring
the local pollutant concentration and metering the treatment solution applied
at that location.
129
-------
H
U)
O
COOf
A iioATt MTIII IIIIOLUILI
V BMl 0*1111 IMOLMLI
f rioAn KATIII KH.UIII
I^J omul. <*'! KU.UUI
MAT! HIAL NAMC
ALmAMTICALi
ACITAUMHVM f
ACITIC ACID ^J
ACITIC ANNVMtOI ^J
ACItOHl f
ACITOM CTANOMVMIN f
ACITOdlTlllll f
ActTontmoM I^J
ADM. ACTLMALIOMAANHVDIMOII fcl
>M>AOD f
>.w> ADO imni . uin f
MAT ACID I^J
IJI>T ACIO nrin IALTI I^J
ACflYUCACIO I^J
ACHvioMTmii ^
AMMCACID ^J
^O-OU-O-OHTO-K-^.., f
ALCOHOL* MONONVOmC «TI*t f
ALCOHOL*. tlTlltS MXVMVOWC ^J
ALDIKVDIt AND KITONIt f
ALMIN TOXAVHIM OMOUT J>}k
ALUMWUM FLUOftlOf I^J
ALUMINUM tUL'ATI lALUMi fcf
AMiMtS TOTAL f
AMMONIA ^
MMrwtit rr^MVMH if
Bianri'ur MM>Att Ik
AMMONIUM ACfTATI fc
AMWONHMOHOfllH \
AMMONIUM MTHATI 1^
AMMOHIUM PtHCHLOHATI ^1
AMVL ALCOHOL f
AftiLINt \J
AMINIC COMPOUUM ^
AMIUM CAHHMATI ^J
MhZAiOCHVM ^J
if
isa
\\l
m
H
3
107
too
11
M
IJO
fl
X)
ID
143
in
m
4
no
u
Mi
M
It
n
20
ra
i
n
»
1*1
i
V*
rt
tM
. H
M
«
!
i
rs
WMO
MMO
MO MO
HMO
logo
MMO
IMO
MJBO
100
mm
110
IMO
MMO
IMO
IOMO
100
IOMO
MMO
IOMO
001
MMO
HOMO
MMO
MMO
MMO
HOMO
S»MO
JMMO
MMO
MMO
ITMO
»"»
1
10
IOW
i
3
i|
S3
e§
*?tu
1»
tair
ow
0*»
en
tM
1 1
01
OJ
1 1
1 1
IM
OJ
1 1
0*
ot
1>
ot
04
XOf
»71
0>
OOOJ
0-001
IT?
t M
1A1
1.0
>
001
ion
»V1
-4]
104
p!
I
to
too
too
10.000
1
IMO
e
r
»
xoo
00»
MO
t
*
0*
100
100
too
100
too
t
10000
ID
If IT tmit ON TMI OHOUNO TH* COUNT 1 MM AtUfflt
DOTtlOWOILOMAHl IMICTIVI
h
HI
i\h
M U1 flAlUATtON
MTHVIMIOAHOH
DILUTI ACID
P^
P
A\
nli
::!§
tilt
0
J
is!
:«»
ip
13!
ill
o»
,!l
1
$iu
;!!
P!
I!!:
*
IP IT imLSiNTOVIATfN TMt COUNT jHM«»iU!
m
*Ms
§'!
if
!!if
Hi
i
Ml
m
i
ij
i!
I
In
0)
i*
\!
i
m
»
*
«
*
*
ATHUAI HAM
IA1M4AMT1CAII
ACfTALMMTM
ACfTKACie
AM TIC AMnDWM
AOTOM
ACtTOM CTAMKrOOUM
ACCTOMTIHLI
AtlTOmOKOM
ACIOi. ACVLMAUOU * AMHVOOUOtl
lAOAOO
IAOAOO miM>wi.n
U>TACtO
M J-T ACtO OTIIIO IALTI
ACttVLK AOO
lamomma
ADWCACIO
>LCONO1I.»O«J«»1>«IC UO»UO»TI
AtcgMMi. ann. mvMvaac
ALMHTOM AND U room
AUMMIUOI nuomot
AuaoMMi muAniALwoi
AIOMI*. TOTAL
«»<
A«M»IUil»Ul»n
ANMOMM ACfTATI
AMOOMUHCHUIIMM
AAMOMIWMTIUTI
IUIC1IIIIHI flUCHlOKATI
AMTL ALCOHOL
AMIL>M
AUXIIIC COIVOIMDS
OAlHMCAMniUTt
MIUALOINVOI
-------
H
U)
H
cooc
^ PLOATt, WAT!* INSOLUtU
^ MMt. MATIN IHVOttMLl
f ru>*Ti wArm KM.UW.I
fel OMMLWATlftKM.U.N.1
HATIIIUL KAMI
IA1PMAMTICAU
HNZtMt f
M WOC ACID ^|
MNZOVi PtftOKlOf ^J
MhrvL cmwuot fr
MflVLllUHD"" 4*
ONK ACID tel
HOMO* ^
MITAOIIM MMoMTtD f
»MTVl ACOATI ^J
MJTVL ACHVIATI A
UTVL ALCOMOt N AMD ISO W
UTVLAMMU ^
UTVL MvoftortKOxiM r
UTVHiC ACID
CALCKMAMUMATt %J
CALCIUM CAKIIM \J
CALCIUM CMLOMIM HJ
CALCIUM » L UOftlM ^J
CALCIUM MVOMOIIDt ^
CALCIUM MVfOCMtOftlTI W
CALCIUM WTHATI ^J
CALCIUM moVMArt WttASlC %J
DinaOtl DtSULf IM \J
CAMON TITHACHLOKIM \|
CHiomns aiQutf noi ^J
CMLOflOUPinNI ^J
CMtCWOKMIM \J
PliflflB""""** *
ON.OMOMl'OMC ACID ^J
CHMMKACIO \J
CITHlC ACID ^j
COfMHWiPATI ^J
CUMCMC 4
CVCiOHIKAHONI '
i
pi
ill
f
It
1H
M
14C
!»
ttl
H>
b
M
in
an
lit
m
\i>
M
ft*
t*7
U
M
1)
141
Itl
IB
ft
U
ite
«»
X*
i«
i
i
1
_
100
IOOQO
001
001
If MO
41.TOB
>
.0000
01
IIUW
H00>
1MO
taatt
>«
V400H
II
IM
IJO«
10JOO
uoo
HO
MMO
m
(000
l»
IMO
IWOjOOO
IOJ009
M
10
10
1 M
IM
>J
1 1
ift
AH»
!
2V
It*
10
00*
Oft*
!i!
i
X
IDS
re
100
root
10
k
in
1000
10
<0
100
>
too
uoo
no
100
000
m
10
M
00
oft
1
1
If IT VltlSON fM MOUND TNI COUOTIMOIMUM1
OOTTIO MIOTI « IHICTtVI
hi
iiii
1
iii
ill
Hi
3;£§
liSi
m
I
|s|
'?»
H
G.*"
is;
iff
iP
li
.
_
i
i
*
*
^
%
*
*
If IT VU.UMTONAT1II. TMC COU«ni»«.U.*JMI DOTTU .MLOM AAf IF WTIVf
li!,
iHi
:si
i
fi
iSSi
SIS
*H
9 !
iiii
_
h
ii!
Hi
gE.I
li
t-^ U
fill
9)
0)
^
i[.'i
>i*l
All
A
0)
0)
_
i j
i:!
jij
!i I
I
r
si
is
i
III
i
i
,
o
MtHIM.«JMOI
iimi«ooiiu>u
MUM
ommcMio
omOVlMMlUM
mm.oa.amot
onnuuHOUiT
omwaao
MM
UVA0IOMO. MM«ITfO
»«UTVL ACtTATt
UTTt ACUVtAn
UTVIUOOM1 HMO KO
UTVLAHHII
OUTYI manopioaaoi
OUTTHCACtO
CAtau»M»m*n
UUMiCMMM
OUOUHKUOMM
CALCHP^ MviMONmo
CAUMH MYTCCMUMTt
CALCIUM MTMTt
CUCHM MOOMATf . OMMK
CAi»maouin«
OMom TiTucMiaom
CHtomi iuaw«uoi
CNCOMOOUOIM
CMUMOKmi
cnm^-nvM
cmomouiroHM; *ao
CMoaoKKie
OTMCAOO
camnoutMTi
CUWM
CTO.OMIUM
CTCtOMUNOm
-------
CO
to
CVCLOHtXVL ALCOHOL j
CVCLOHtKVLAMNt j
DOT 4
MCVL ALCOHOL ^
WIMMOCHLOflOrHOPANI \
DMMTVL PHTMAIATI %
DtCHLOMOIIftZINL *\
OKHLOHQOmLXMIOMt THAMI ^
MITHAMHAMMI ^
OIITHVLAMNI T
WITHVLIHft dVVCOL \
DINlTftOAMliINf <|
OIOCTVL PH THAI AT IS \
DlHtOmt«( 01 VCOL ^
DIVIHVI « NZINC «1
a. r
OOMCVL UNIINI j
DOOfCVl MIKr APT AN J
DV» TOTAl \
mCHLOflOHVOMIM ^
ITMVl ALCOHOL FTHAMOL /
CYMAMOLAMMIS J
THINS TOTAL J
ITHVL ACITATC jV
ITHVL ACHVLATt f
ITHVLAMWH |
ITHVL UN«Mt 7*
ITHVL CHLOftlOf ^
f THVL UTIHS ALL CHAOIS f
ITHVL POUMATI /
tTHVLINf f
ITHVLfNIOIAMM j
CTHVLIttC DtCHLMIDI ^
ITHVLINICLVCOL \
THVLINIONIM f
fTHVLlWMNt f
f THVLHf *VL ACNVLATI f
MftBAM If tMUTtt nSTICIDt ]
»IHftOU«CULPATE ^
» LUO«INffHVDItO*lUOMIC ACJO J
rOMMALOfHVDC j
POHMIC ACID j
PUMAMIC ACID ^
PLMGICIOIft ACVCLK \
flMGICtOd H -. TLlAHl JUt j
PUMGICIDIi TOM. CVCLtC \
\m
w
14
b
MB
1M
111
t«
IH
tM
»b
Ik
Mft
tJ
tM
IM
U
I)
h
«
IM
IT
in
?
114
m
1IT
Ml
l»
t7
Ik
17?
in
in
in
1H
3i
M
a
21>
»1
111
110
M
4,700
1000
001
0 1
100
M
im
001
wjno
11.000
M.OOO
001
10
WOOD
001
100
10000
1000
W 000
oot
SO 000
MAM
MOOO
MAOO
too
tcjeeo
140
57*0
Wt MO
OAOO
100
MJOOO
001
OjOOO
WJOOO
MJOO
MMO
100
1JBOO
HAM
UMO
ojne
Tjooa
100
100
001
DO*
001
oaj
on
1 1
i «
t j
1 4
100
0 M
1 t)
1*
1 ?
IO1
0<3
Of
ot
no
10
1 1
on
04
OJ
Of
on
01
am
on
OJ
on
OJOOt
OM
1 1
1 1
oaa
9M3
OJ»
OJ
It
0«1
OJ1S
04 IS
IM
Off
01
OJ
1
oot
IOMD
JO
1*00
t
100
1000
nooo
1000
IO
*
.
f 000
ft
1000
1*00
too
xo
10
m
JO
IM
HUM
IM
1
tM
*
MO
4*
W
IM
MO
OQM
oooot
'
9
»
*
*
*
*
*
»
*
*
*
*
»
*
CVCLOH(KVL ALCOHOL
CVCLOHIXVLAMIM
OOT
MCVL ALCOHOL
DHMOUOCHLOAOMtOPAMf
MMSTYL PHTMA1ATV
OtCMlMOUNaNt
OKHL OHOOIP LUOMHM THAMf
DJfTHAMOUUWM
OIITHYLAMUtC
DIITHVUMI OLVCOt
CHMtTMO AMILtNf
MOCTVl PHTHALATIS
DIMIOrVLf « Ot VCOL
DfVINVL MM7CM
OOOf CVL HM2f Ml
DOMCVL Mff ACAPTAM
DVIS. TOTAL
(nCHLOftOHVMIIN
f THVL ALCOHOL IITHANOLI
ITHAMOLAMINCK
I THINS TOTAL
f THVL ACf TATI
CTHVL ACNVLATI
ITHVLAMMCS
ITHVL MNXf Ml
ITHVL CHLOMDfl
CTHVL HTIM ALLOHAMI
ITHVLPOdMATC
KTHVLIM
ITHVLfNIOIAMrm
ITHVLf Ml DtCHLONIOC
ITHVLINC OLVCOt
ITHVIIW OXIOf
THVLIMMMC
ITHVLHIMVL ACHVLAT1
IMAM IFIMUATII KtTIOM
IflftOUC WLtATI
FLUoniMmvonof LUOHIC ACIO
OMtALMHVDf
OmHCACIO
UMAJWCACIO
UMOICIMt. ACYCLIC
UMGIOM1 MSCtLLAIWOUS
LMGIODU. TOTAL CYCLIC
-------
to
U)
CODE
^ FLOATS. WATER INSOLUBLE
^ SINKS. WATER INSOLUBLE
f FLOATS. WATER SOLUBLE
J^ SINKS. WATER SOLUBLE
MATERIAL NAME
t ALPHABETICAL!
PHJMPURAL P\J
FURFURVL ALCOHOL \|
flLYCIRINE ^J
GLVOMAL \J
HIPTANE. MIKED f
HIMICIDU PLANT HORMONES. ACYCUC f
Mf R41C1DES PLANT MOftMONES. CYCLIC f
HtffBtCTOES MISCELLANEOUS CYCLIC f
HBXAMETHVLENEDIAMINE ' f
HEXANE 4f
HEXVL ALCOHOL f
HVOAOCARBONS-HALOGENATID *«J
HVDffOCHLORIC ACID P^J
HYDROGEN CYANIDE f
HYDROGEN PEROXIDE ^J
HYDHOGENSULFIDE f
HVPOCHLMITIS |^|
INSECTICIDES. MISCELLANEOUS ACYCLIC f
INSECTICIDES. MISCELLANEOUS CYCLIC £,
ISOBUTANE jft
ISOCVANUMATES CHLORINATED (fcl
1SOOC1VL ALCOHOL >jf
ISOPINTANI A
ISOPRENI A
ISOPROPVL ACITAfl jf
ISOPHOPVL ACITONI *
ISOPROPVL ALCOHOL ^
ISOPROPVL ETHIR ^
LACTIC ACIDS "^
LAUROVL PEROXIDI ^f
LEAD ARKENATI "^
LI AD COMPOUNDS \J
LtNDAMt *
MAGHtSIUM COMPOUNDS \J
MACNISIUM tUt ' ATI k^
MAI 1 1C ANHVDRIOt P^i
S
n
IM
2B4
210
1*
IM
22
.
u
IM
M
40
43
1U
U
B
4ft
1S4
IS
»14
JtO
t»
117
711
IJb
4
U
U
71*
iM
144
ITSSOLUBILItVIN
NATCH IN IPAHTS PER Mil LIQNI
S3 MO
MttQ
OjOOO
MAO
W
100
wo
1.000
1.000
**
SJOO
MAM
MjOOO
OjOOO
SOMO
SOMO
1.000
1000
001
001
10000
10 000
001
oos
30400
14000
WOOD
7000
10000
Ol
1000
MOOO
MOM
I 000
ISO 000
WOOD
£
5
I.IB
112
,.,
1.M
04i
OB
04
Ot
04
OM
041
1 1
141
048?
t44
047
1 t
04
04
0*
17
00
OS
on
047*
01
on
017
1 7
70
»J
to
09
10
IM
OtI
CONCENTRATIONS OMATER
THAN THIS ARE KNOWN TO
KILL FISH IPAHTS MR MILLION)
jj
t
BjBBB
IjBM
tJM
M*
100
1MB
200
IM
04S
to
t
OOOJ
0001
1
100
"">
wo
too
MIO
M
OS
0 1
400
10000
100
If It 1
MUTRAIIIATION
wiMiancm
WOIIM
CARMMATI
,
.
.
.
.
.
.
.
'
PILL* OM TW GROUND THt COUMTIMHtAft
oomoojiMLoiiAiii cmciivi
MUTMUUTIOH
tTHVINIOAIIOII
DILUTI ACID
Pi
llii
liii
.
.
,
.
[pOmMCMANICAl AMD/Oft
CMMJCAl «AMII US TO
MIVINT SMIAO W UAUIIIAl
,
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
'
i
.
.
.
«
.
«
*
.
.
.
«
«
.
.
«
«
«
UMIS
APPIICATIOMOF 'LV
AIM 0* COUNT nmof II
TO AMMIna LIQUID
ULK
m
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
i
.
.
.
.
.
IF IT SPILLS INTO NATIH THI COUNTS BMIAKjms DOTTID m MLON AMI IMf CTIVI
ifif
*ci*
.
ust nmrAci ACTIVI
AOIHTSTOCOvmSS
AMD TMGKtN STILLIO
MATImAl
.
.
.
.
.
.
.
.
.
.
.
a
.
.
.
.
.
.
.
.
.
.
.
riMMCT UNWIMAL OILLINO
AOINT TOSOLIOI'V TMAmD
MAM AMD IKMAU
tMCCTIVf MS* OP MAMS
.
.
.
.
.
.
'
.
lih
i
1
,
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
i :s
iii
III!
.
.
.
MUTMIIK KITH IIM
OH BCAMONATI
.
.
.
.
.
.
NIUTftALIZt WITH MLUTI
AdO OM WITH HIMOVAILI
Tram ADD
.
[UIC MCCHANICAl OMCOOIS OH
LIFTS TO M MOVE IMMOBUI2ID
MASS! ft OP POL LUTANTS AND
MCCIP1TATIS
,
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
,
.
.
MATIPIIU KAMI
lALFMAMTICALI
pUflPIJHAL
PUNPUNVL ALCOHOL
GLVCfPJINi
CLVOKAL
HfPTAfM MIXtD
HHWODIS PIAMT HOKMOMS. ACTCLIC
HCASJCIDIS > PIAMT HONMOMIS. CTCLK
HHUHODIft MMCCiLANIOUS CYCLIC
HCXAMfTHVLIIrtCNAMMt
HtMANt
HCXVL ALCOHOL
MVDKOCAIWOIrX HALOfllMATID
HVMOCHLOHIC ACID
HVDHOCIN CTAMIDI
HYOKOCIX fOO«IOI
HVONOGI* SULFIDI
HVPOCHLOKITIS
INSfCTICIDIft MISCflLANtOUt ACYCLIC
INUCTICIDEft. MISCILLANIOUSCYCLIC
INMCTICIOIS . KOOCNllCIOft CYCLIC
ISOtUTAMC
ISOCVANUHATIS CHLOMINATCD
ISOOCYTl ALCOHOL
ISOPINTANI
.IO..INI
ISOPMOPVt ACI'Ari
ISOPIIOP-YL ACETONC
ISOPflOPVL ALCOHOL
ItOPHOPVL ITHCfl
LACTIC ACIOt
LAUMOVL Pf ftnxlfw
Lf AD AHSINAT1
If AD COMPOUNDS
LINDANt
MAGNESIUM COMPOUMDC
MAONISIUM SUl F ATI
MALIIC ANHVOHIM
-------
U)
Mflcudv fr
MIMCUflV COMPOUND! ^
HHKUKV VUMOlCIMI A
WTNVl ACITATI Jf
M1TMVL ALCOHOL IMITMAMOL' 4f
MinmAMMM W
MITHVl ITHVl HITOHI Jf
MITHVL WTHACIIVLATI «
MITHVL PAMATHION 4f
HOMO I THAMOLAMINI fcj
MOWHOUNI 4f
MAJAM Jf
NAPWTHALIM \J
MCKIl COMPOUNDS P^J
MICKIL SULfATI P^J
MT1IIC ACIO P^J
WTNOANILIMB P^J
MTftOMNft*PI ^J
MTMPNCMOl P^J
MTMOUBOIIIDI 4|
NOMM A.
NONVL PHI MOL A
OttAUC ACIO fe|
PMATHION 41
MMTAfMLOfUWMIMOt Ik 1
^7
PIMTAMI 4f
Pf flCHLO
IJ
144
1 If
t 4
0001
or
0*4
ii
01
019
IJ
10
OJ
107
OJ
IJ
IJ
00
103
Ik
1 »
10
104
113
2X3
IU
OJU
OJI7
OJ
a«
tftJOQO
JO
§MO
HO
01
w
100
10
1
n
i
a
20
10
tooo
1
J.OW
0 1
100
1
'
01
1*
1
0 1
too
uo
U
10
MO
100
100
no
»
4)
4>
*
4)
4)
»
4)
4)
4)
4>
4)
4)
4)
4)
»
*
4)
*
*
4)
4>
4)
4)
4>
4)
MNCUIIV
omcunvcowouwi
MNCUNV NMOICIMS
MTHVL ACf TAT1
MTKYL ALCOHOi. PMTMAMOU
MTHVUUMMM
l
HfTHVL ITNVL CTOM
HTTNVL HIYMAamAn
MTHVL PAAATMOH
MONO ITHANO1AMMC
MQMIKM.HH
MAMM
UVHTALIM
MCKIL COMPOUNDS
MKKCL SULPATI
MfTMCAOD
CTMOAIttUMI
MTMOMimM
MTHOTMIMOt
wmouooxnc
MOMtM
MOMVLPMMOL
OXMKAOO
PAHATMMM
*MM»M«Mni
rm^
PUTT AM
MMMLOmCAOO
PfHCMUMOfTHVUM
PUIIUUtl MJMCnOCM. ACVCMC
PHfMOi
PMlim MmCLWIC ACITATt
PMOVMOmC AOO
PMOjmOMU*
PHOIPMOIIM OKTCHUMUC4
PHOtPMOIlUt Pf NTAJUtPlM
PlttOJIiUBui TIUCMLOIMOC
PHTHALK AMMVOMOff
POTASRUM COMPOLWOB
POTASSIUM MVDNOXIOC
PUIAUIUW iQOtyt
*OI AkbUM PV KOPHOSPMA 1 1
POTA^MUM HILf ATI
«tO*»C*MC AOO
PHOFVi ACIfATt
NP«mrvt ALCOHOI >NPIIOPAMOt>
-------
Ul
COM
^ FLOATS. MATIN INSOLUtLl
Y SINKS. WATIM INBOLIMLI
f FLOATS. WATIR SOmaLI
F^J SUMS. WATIH SOL ML I
MATINIAL KAMI
IAL9HAUTICAU
MOPVLAMINIS f
FflOTVLtMt f
FftOFVLIM DfCHLONIM F^|
FftOFVLINI OLVCOL F^l
FflOFVLINt OXIM f
FVNIOmi f
PLVIM CVAMOI ^
UVIft NITHATI 1*^
SODIUM IM TALI f
ODIUM ACITATI Ik
SODIUM KMATt It
SODIUM COMPOUNDS It
SODIUM CAHOONATI It
SODIUM CMLOHATI F^
SODIUM OKHNOMATI ANOCHHOMATI 1^
SODIUM FLIXMIDf \
SODIUM HVDNOSULF IT t 1^
ODIUM HVDHOIIDt *\
SODIUM MfTHVLATI l\
ODIUM PMO*PMAT| 1%
SODIUM SILICATE l\
ODIUM SULFATf 1^
SODIUM SULFIDI ^
SODIUM SULFITE 1^
SOHBITOl 1^1
STVfllNI f
SU.LFUN Jr
SVLFUFJ DIOXIDt FlJ
CULFUNK ACID I^J
TITHAITHVLIM OLVCOL 1^
TCTNAL1N f
TOLUfNI f
tNICHLONOCTHANI ^
,»,e,.:nno..H,l,« \|
1
1W
in
i»
m
n
1ST
IN
IM
rat
IN
I4O
H
m
IM
in
,,
m
14*
ni
to
2ft
n
»
!!
tM
»
1
«
117
i
|
HMO
MO
uro
tojtoo
MJOB
OJ3
t»JOO
anaat
HMO
10. HO
MMO
lOjOOO
mjooo
73MO
4JJOO
IHjOOB
4JOJOOO
04»
KOK
tea*
««o
IMjOlO
110JOOO
OAOO
IMO
401
m iwcnvt
il
!ps
IM5
yll
illli
-
vgz
.
0
fii
.
ih
ii!
iiii
.
ill
t!
jjjp
.
11
P
1*1'
hit
il
if
i
.
.
0
1
.
.
.
.
0
,
.
.
.
MAT* HULL MUM
1A1FMAMTICA1.I
MIOPVLAMUNIS
MorviiNt
MOTTLtMt OWMLOFttOf
FNOFTLmi QLTCOL *
FNOFVLINt OHIDt
rvtuomt
STLvm CVAMUM
CtLVIN WTNATI
SODIUM -Ml TALI
OOMJM ACITATI
ODIUM KMATI
tOOTUH COMFOUNOS
SOOnVCAffMNATI
KXMUMCM.OAATC
OOHMI7«MraMAT1 ANOCMNOMATt
OO*UM FLUOHIOC
OOHJM KVONOCU1FITI
SODIUM HTONOKKM
ODIUM MCTHVLATI
KMMMFHOWMATI
OOMJM SILCATI
SODMMSULFATt
SODIUM SUL »IM
OOJUMCUlFITI
OMMTOL
STYNtM
ULnM
ULFUH DKWIOI
U1FUMICACIO
1IT1LAITHVLIMI OL*COL
T«T MAIM
TITNAITHVL LIAO
TOLUfNf
TNlCHLONOfTHAJM
T NiCHLONOI TH VL 1 NI
-------
TmCHlOIIOFlUOmXMTHAKt ^
Hi TmcHUXmHIIMH. AM) «*LTI ^
TmfTHAMCH.AM»» Jf
TflllTimlNt OLVCOt Ifcl
rfllMITHVLAMNI f
TUMCN11NI f
U«« l^|
VINVLACITATI |f
VINVL tTHlH ^
XVLINtS ^
imc MI»II I^J
ONC CHKWIH HJ
lINCCOMroUNM |^|
ItNCWlMn kl
Ib
u»
1«>
ItT
U
111
171
112
-
-
10>
-
001
041
1*MO
OMO
1.000
1.M»
nojooo
WjOOO
1
to
I.OOO
4J30.000
4-DO.OOO
10 JW
14
1 1
on
1 1
O.M
0«S
1}
Ot»4
on
OJt
30
211
211
194
200
0
JO
1000
no
i
14.000
n
10
i
«
»
'
*
*
*
»
tftKHI.O«O«LUO«IM
-------
REFERENCES
1. Dawson, G.W.; Shuckrow, A.J.; and Swift, W.H.; "Control of
Spillage of Hazardous Polluting Substances;" Report prepared
by Battelle Memorial Institute, Pacific Northwest Laboratories
for the Federal Water Quality Administration, Department of
the Interior, Program No. 1508, Contract No. 14-12-866,
November, 1970.
2. Crank, J., The Mathematics of Diffusion, Oxford University
Press, 1956.
3. Saffman, P.G., "The Effect of Wind Shear on Horizontal Spread
from an Instaneous Ground Source," Q. Journal of The Royal
Meterological Society, V88. 1962.
4. Carslaw, H.S., and Jaeger, J.C., Conduction of Heat in Solidj,
Oxford University Press, 1947.
137
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-670/2-75-042
3. RECIPIENT'S ACCESSION«NO.
4. TITLE AND SUBTITLE
METHODS TO TREAT, CONTROL AND MONITOR SPILLED
HAZARDOUS MATERIALS
5. REPORT DATE
June 1975; Issuing
Date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S) Roland J. Pilie, Robert E. Baier, Robert C.
Ziegler, Richard P. Leonard, John G. Michalovic,
Sharron L. Pek, and Ditmar H. Bock
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
1BB041; ROAP 21AVN; Task 033
Calspan Corporation
Buffalo, New York 14219
11. CONTRACT/««A»* NO.
68-01-0110
12. SPONSORING AGENCY NAME AND ADDRESS
National Environmental Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A program was instituted to study the feasibility of treating, controlling and
monitoring spills of hazardous materials. Emphasis was placed on considering tech-
niques and equipment which would be applicable to general classes of chemicals rather
than to specific hazardous polluting substances. Several methods were investigated
and found to be promising for removing or detoxifying spills of hazardous chemicals
"in situ". These included: the use of sodium sulfide as a precipitating agent for
spills of heavy metal ion solutions; the use of activated carbon packaged in porous
fiber bags (carbon "tea bags") for adsorbing a wide variety of soluble organic chemi-
cals; and the use of various acids or bases to neutralize spills. Methods were
studied to control spills on land and prevent their contaminating nearby surface or
ground water. To this end, a four-component "universal gelling agent" was developed
to immobilize a spilled liquid. A "cyclic colorimeter", a novel heavy metal ion
detector, was perfected and laboratory tested, and a detection kit capable of sensing
several chemicals was developed. A computer model was developed and refined to simu-
late the spread of a spill when certain stream parameters and material characteristics
are known. Bioassay studies were conducted for several chemicals using at least three
species of biota. In addition, bioassays were conducted to estimate the environmental
effect of each of the various treatment methods developed.
17.
KEY WORDS AND DOCUMENT ANALYSIS
a.
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
*Water treatment, *Decontami-
nation, *Hazardous materials,
*Activated carbon treatment,
*Gelation, Detection, Chemical
removal (water treatment),
Monitors, Biology, Neutraliza-
tion, Precipitation (chemistry)
Hazardous materials spill cleanup,
Hazardous materials spill control,
Hazardous materials spill monitor-
ing, Hazardous chemical spills,
Activated carbon "tea bags",
Cyclic colorimeter, "Universal
gelling agent", Bioassay
13B
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)'
UNCLASSIFIED
21. NO. OF PAGES
148
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
138
. S. GOVERNMENT PRINTING OFFICE: 1975-657-594/5*03 Region No. 5-11
------- |