EPA-600/2-81-199
                                            September 1981
             OIL SLICK DISPERSAL MECHANICS
                          by
                  Chukwuka A. Osamor
                   Robert C. Ahlert
  Department of Chemical and Biochemical Engineering
      Rutgers, The State University of New Jersey
           New Brunswick, New Jersey   08903
                   Grant No. R805901
                    Project Officer
                 Leo T. McCarthy, Jr.
       Oil and Hazardous Materials Spills Branch
Municipal Environmental Research Laboratory--Cincinnati
     —         Edison, New Jersey  08837
      MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
          OFFICE OF RESEARCH AND DEVELOPMENT
         U.S. ENVIRONMENTAL PROTECTION AGENCY
               CINCINNATI, OHIO  45268

-------
                           DISCLAIMER


     This report has been reviewed by the Municipal  Environmental
Research Laboratory, U.  S.  Environmental  Protection  Agency, and
approved for publication.  Approval  does  not signify that the
contents necessarily reflect the views and policies  of the U. S.
Environmental  Protection Agency, nor does mention of trade names
or commercial  products constitute endorsement or recommendation
for use.

-------
                            FOREWORD
     The U.  S.  Environmental  Protection Agency was created
because of increasing public  and government concern about the
dangers of pollution to the health and welfare of the American
people.  Noxious air, foul  water, and spoiled land are tragic
testimonies  to  the deterioration of our natural  environment.
The complexity  of that environment and the interplay of its
components require a concentrated and integrated attack on the
problem.

     Research and development is that necessary first step in
problem solution; it involves defining the problem, measuring
its impact,  and searching for solutions.   The Municipal Environ-
mental  Research Laboratory develops new and improved technology
and systems  to  prevent, treat, and manage wastewater and solid
and hazardous waste pollutant discharges  from municipal and
community sources, to preserve and treat  public drinking water
supplies, and to minimize the adverse economic, social, health,
and aesthetic effects of pollution.  This publication is one  of
the products of that research and provides a most vital communi-
cations link between the researcher and the user community.

     Crude oils and petroleum-based products are extremely
complex systems and behave differently when discharged to marine
environments.  This study investigates the spreading and disso-
lution  behavior of small oil  slicks on water.  The use of
chemical surface-active agents to disperse oil slicks was also  -
evaluated in experiments conducted under  controlled conditions
in the  laboratory.
                                Francis T. Mayo, Director
                                Municipal Environmental Research
                                  Laboratory
                               1 i i

-------
                             PREFACE


     During the past five decades, petroleum has become the
world's major source of fuel  and power.  As a consequence of the
rapidly increasing demand for petroleum, the volume of crude
oils and refined products handled during the production, trans-
portation, and chemical manufacturing processes has increased
dramatically.  Anderson (1967) estimates that oil  accounts for a
third of world shipping tonnage.

     This dependence on oil  is causing increasing  oil  pollution
of marine environments by accidental discharges from tankers and
oil-ferrying vessels, from rupturing off-shore oil-wells, and as
a result of natural  seepage.   Intentional  discharges of bilge and
ballast water further aggravate oil  pollution.  Several key
sources of oily discharges into the  marine environment were
identified in a report by the National Academy of  Sciences (1975).
This report estimates the annual influx of oil into the oceans

at 6 x 10  metric tons.  A major fraction  of the influx was
attributed to spillage during tanker collisions, natural seepage,
and oil well mishaps.

     Oil pollution of marine  environments  has been occurring
during the past decades, but  public  awareness of the problem did
not intensify until  the Torrey Canyon Spill in 1967 and the 1969
oil well blow-out in Santa Barbara Channel.  Before these inci-
dents, autodepuration processes were considered adequate
cleansing mechanisms for oil  pollution of  all natural  water
bodies.  The increasing size  of oi1-transporting vessels with
reduced manueverability and  the shift of oil prospecting to
shallow waters of the outer  continental shelf have caused
considerable concern.  News  media coverage of oil  spills from
recent tanker collisions has  maintained public interest.  Pictures
of dead oil-soaked fowl and  fish and tarry oil lumps that wash
ashore to deface clean beaches make  emotional issues.

     Because of the  many problems caused by oil spills and
concern for a clean  marine environment by  the public and environ-
mentalist groups, legislation has been enacted at  both national
and international levels to  curb oil pollution.  During recent
years, research in oil pollution has increased as  many labora-
tories try to provide answers to numerous  questions concerning
the inputs, fates, and effects of petroleum in the marine
envi ronment.
                               i v

-------
     Initially, research studies focused on developing adequate
equipment and devices for containment, removal, and clean-up of
spilled oil.   Oil  spill  clean-up is a difficult problem for
several reasons:

     a)  Oil  spills occur at random, and the portion of the
         earth susceptible to oil  spills is large:  Approx-
         imately 70% of the earth  is .covered with water and
         60% of world oil production is transported across
         the oceans;

     b)  A majority of the world's coastlines are scattered
         and inaccessible in the event of an oil  spill;

     c)  In the presence of the enormous forces generated
         by winds, waves, and currents, oil slicks can cover
         large areas of the ocean  surface in a short time;
         thus the logistics of dealing with large oil  spills
         in remote areas can be overwhelming.

     More recent research efforts  have been focused on ecological
studies to understand the short- and long-term biological effects
of oil  pollution on marine organisms.  Success in these two
research areas has been limited because of lack of understanding
of the  physics and chemistry underlying the dispersal  processes
that act on oil after a spill occurs.  Research into the physical,
chemical, and mechanical processes that disperse  the components
of crude oils is necessary, and the rate of each  process has to
be quantified.  Knowledge of these rates will be  useful for eval-
uating  the biological effects of petroleum on marine biota and
future  decisions on development of more effective clean-up
methods.

     Since coastal waters are the  most productive part of the
sea, containment and physical removal of oil slicks are the
favored methods for oil  spill clean-up.  Present  technology in
booms and skimming devices limits  their use to calm waters.  Thus
in consideration of potential ecological damage to fish popula-
tions and coastal  marshes, as well as fire hazards from large oil
spills, use of chemical  oil dispersants appears to be  a promising
option, especially in view of the  nontoxic nature of the new
generation of dispersants.  The number of commercially available
dispersants is large, however, and manufacturer's claims for
these dispersants are difficult to evaluate and/or substantiate.
In the  absence of detailed laboratory studies, a  realistic
approach to dispersant application cannot be developed.  The role
of dispersants and the mechanistic steps in the breakup and
reformation of oil slicks following dispersant application are
poorly  understood.

     Recognition of these information gaps has provided the
stimulus for this research study on oil slick dispersal mechanics,

-------
particularly in the areas of spreading,  dissolution,  and  chemical
dispersion.
                              VI

-------
                            ABSTRACT
     This study investigates the spreading behavior of small  oil
slicks formed from spills of 12 oils.   The oils included crudes
of different origins as well as processed oils.  The increases
in area covered by the oils over time  were determined using
photographic techniques.  Spreading equations were used to
correlate experimental data.  These equations contain the physi-
cal properties of oil  and water phases, volume of oil spilled,
and time of spreading  as independent variables.  Derivation of
the equations parallels Fay's development.  Results show that
gravity, surface tension, and viscous  forces are important in
spreading.   Oil slick  configurations vary for different oils,
and profiles are neither rectangular,  circular nor spherical.

     The rates of dissolution of these oils in tap water at 25C
were investigated by equilibrating oil with water in open static
systems for 2 to 3 weeks.  Limits of solubilities have been
established for the oils.  Six oils were also equilibrated in
salt water at 25C.  The concentration-time profiles of the oils
vary in tap and salt water tests.  Concentrations of oil in
solution are lower in  tests with salt  water.  A segmented mathe-
matical model has been derived and was used to correlate experi-
mental data.  The model describes two  processes that occur when
oils are equilibrated  with water:  soluble and volatile compo-
nents of oil leach into solution initially and later evaporate K
from solution.  Experimental data confirm the nonequil i brium
nature of dissolution  and evaporation  processes.  The concentra-
tion of oil in the water column may be toxic to marine organisms
at certain  times during equilibration  processes.

     A large-scale laboratory system has been designed and was
used to evaluate the efficiencies of five commercial dispersants,
Parameters  investigated include oi 1 -to-di spersant ratios, oil
type, degree of agitation, and the effect of salt water.  The
mass transfer processes which cause chemically treated slicks to
break up into oil droplets and to disperse into the aqueous
phase upon  agitation,  have been identified.  The primary
mechanisms  have been quantified by analogy to homogeneous and
heterogeneous catalysis and detergency.  Phenomenological
theories have been developed for chemical di spersant action at
the oil/water interface and in the near-field water column.  The
use of chemicals to disperse oil slicks must be viewed as a


                               vi i

-------
management alternative based on knowledge of the toxicity of
dispersants to aquatic biota and interactions with oil/water
systems.

     This report was submitted in fulfillment of Contract No.
R805901 by Rutgers University under the sponsorship of the U.  S.
Environmental  Protection Agency.  This  report covers the  period
August 1, 1978 to January 31, 1981, and work was completed as  of
June 30,  1981.
                              VI 1 1

-------
                       CONTENTS

Foreword 	      i i i
Preface .	       iv
Abstract 	      vi i
Figures 	       xi
Tab! es 	      xi v
Abbreviations and Symbols 	      xvi
Acknowledgements 	       xx

     1.  Introduction 	        1
     2.  Conclusions 	        4
     3.  Recommendations 	        8
     4.  Research Objectives 	       10
     5.  Literature Review of Relevant Previous
              Work 	       12
              Spreading 	       13
              Oil Transport Due to Wind, Wave and
                   Current Forces 	       22
              Dissolution 	       26
     6.  Chemical Dispersion 	       36
              Oil/Water Systems 	       36
              Emulsion Formation  	       36
              Structure of Dispersants 	       40
              Colloid Chemistry Aspects of Chemical
                   Dispersion 	       42
              Viability of Chemical  Dispersion as
                   an Oil Spill Clean-up Method ...       44
     7.  Experimental 	       49
              Materials 	       49
              Data Aquisition Devices 	       54
              Setup 	       56
              Procedures 	       67
              Presentation of Experimental Data ...       80
     8.  Modeling of Oil Slick Dispersal Mechanics.       81
              The Mechanisms of Dissolution 	       81
              The Rate of Oil Spreading on Calm
                   Water 	       90
              Mechanisms of Chemical Dispersion of
                   Oil Slicks 	       99
              Verification of Dissolution and
                   Spreading Models with Experi-
                   mental Data 	      118

-------
     9.  Results and Discussion 	       120
              Dissolution 	       120
              Spreading 	       132
              Chemical  Dispersion 	       146

References 	       176
Appendices

     A.  Sample Calibration  Curves 	       189
     B.  Solubility Data in  Tap Water at 25C 	       192
     C.  Solubility Data for Selected Oils in  Salt
              Water at  25C 	       195
     D.  Experimental  Data for the Spreading of
              Oil  Slicks on  Calm Water at 20C  	       195
     E.  Experimental  Data for the Dispersion  of
              Oils with Oil  Dispersants 	       207
     F.  Computer Programs for Numerical Analysis
              of Experimental  Data 	       214

-------
                             FIGURES


Number                                                       Page

   1    Processes involved in the Fate of Spilled Oil  in
          the Marine Environment 	     2

   2    Surface Tension Forces Acting on an Oil  Lens 	    15

   3.   The Four Forces which Act on an Oil Film During
          Spreading 	    17

   4    Some Typical Surface-Active Agents 	    41

   5    Schematic Representation of a Spherical  Micelle 	    43

   6    Schematic Diagram of Experimental Setup  for Spreading
          Studies  	    58

   7    Schematic Diagram of Experimental Setup  for
          Dissolution Studies 	    60

   8    Schematic Diagram of Experimental Setup  for Chemical
          Dispersion Studies 	    61

   9    A Segment from the Spectra of the CC1A Extract of an
          Oil/Water Sample	    72

  10    Sequence Photographs of Oil  During Spreading 	    75

  11    The Three Phases of an Oil/Water System  	    82

  12    The Regions and Concentration Profiles of an Oil/
          Water System 	    85

  13    Schematic Diagram of an Oil  Slick on  Calm Water
          Showing Spreading and Retarding Forces 	    92

  14    Some Mass Transfer Processes During Chemical
          Dispersion of Oil Slicks 	   102
  15    Comparison of Profiles of Dissolution Model  and
          Experimental Data for #2 and #6 Fuel  Oils and
          Crudes from Nigeria,  La Rosa (Venezuela),  and
          Lagunillas - Tap Water 	   126

  16    Comparison of Profiles of Dissolution Model  and
          Experimental Data for Crudes from  Suniland,
          Sahara and Iran, and 8% Mixture of Crude Oil and
          Processed Oil - Tap Water 	   127

-------
Number                                                       Page
  17   Comparison of Profiles of Dissolution Model  and
          Experimental  Data for Crudes from Alaska, Arzew
          and Brass River - Tap Water ......................   128

  18   Comparison of Profiles of Dissolution Model  and
          Experimental  Data for #2 Fuel  Oil, Iranian and
          Brass River Crudes - Salt Water ..................   129

  19   Comparison of Profiles of Dissolution Model  and
          Experimental  Data for Nigerian, Alaskan,  and
          La Rosa Crudes - Salt Water ......................   130

  20   Comparison of Profiles of Dissolution Model  and
          Experimental  Data for Nigerian Crude .............   131

  21   Comparisons of Experimental Data  and Spreading
          Equations for #2 Fuel Oil (25  mis) ...............   147

  22   Comparisons of Experimental Data  and Spreading
          Equations for #2 Fuel Oil (50  mis) ...............   148

  23   Comparisons of Experimental Data  and Spreading
          Equations for #2 Fuel Oil (75  mis) ...............   149

  24   Comparisons of Experimental Data  and Spreading
          Equations for #2 Fuel Oil (100 mis) ..............   150

  25   Comparisons of Experimental Data  and Spreading
          Equations for Nigerian Crude (25 mis)  ............   151
  26   Comparisons of Experimental  Data and Spreading
          Equations for Nigerian Crude (50 mis) ............   152

  27   Comparisons of Experimental  Data and Spreading
          Equations for Nigerian Crude (75 mis) ............   153

  28   Comparisons of Experimental  Data and Spreading
          Equations for Nigerian Crude (100 mis) ...........   154

  29   Comparisons of Experimental  Data and Spreading
          Equations for Alaskan Crude (25 mis)  .............   155

  30   Comparisons of Experimental  Data and Spreading
          Equations for Alaskan Crude (50 mis)  .............   156

  31   Comparisons of Experimental  Data and Spreading
          Equations for Alaskan Crude (75 mis)  .............   157
  32   Comparisons of Experimental  Data and Spreading
          Equations for Alaskan Crude (100 mis) ............   158
  33   Dispersion of #2 Fuel  Oil with Products  A,  B,  C,  D
          and E at 1:1  Oil -to-Di spersant Ratio  .............   161

  34   Dispersion of #2 Fuel  Oil with Products  A,  B,  C,  D
          and E at 5:1  Oil -to-Di spersant Ratio  .............   162

  35   Dispersion of #2 Fuel  Oil with Products  A,  B,  C,  D
          and E at 10:1 Oi 1 -to-Di spersant Ratio ............   163


                              xii

-------
Number                                                       Page

  36   The Effect of Oil-to-Dispersant Ratio on the
          Dispersion of #2 Fuel  Oil  with Product A 	  165

  37   The Effect of Oi1-to-Dispersant Ratio on the
          Dispersion of #2 Fuel  Oil  with Product B 	  166

  38   The Effect of Oil-to-Dispersant Ratio on the
          Dispersion of #2 Fuel  Oil  with Product C 	  167

  39   The Effect of Oil-to-Dispersant Ratio on the
          Dispersion of #2 Fuel  Oil  with Product D 	  168

  40   The Effect of Oi1-to-Dispersant Ratio on the
          Dispersion of #2 Fuel  Oil  with Product E 	  169

  41   Dispersion of Iranian Crude with Products A, B, C, D
          and E at 5:1  Oi 1-to-Dispersant Ratio 	  170
  42   Dispersion of #6 Fuel Oil  with Products A, B,  C, D
          and E at 5:1  Oil-to-Dispersant Ratio 	  171
  43   Dispersion of #2 Fuel Oil  and Iranian Crude in Calm
          Water with Product B  (5:1  0/D Ratio) 	  172

  44   Dispersion of #2 Fuel Oil  and Iranian Crude with
          Product B (5:1  0/D Ratio)  under Continuous
          Agitation 	  174

  45   Dispersion of #2 Fuel Oil  and Iranian Crude in Salt
          Water with Product B  (5:1  0/D Ratio) 	  175
                             xiii

-------
                                    TABLES
Number
1
2
3
4
5
6

Spreading Laws for Oil Slicks 	
Solubility Values for Some Crude Oils and
Petroleum-Derived Products 	 ,
Analytical Data on Oil Samples as Performed by
Mason & Hanger-Silas Mason 	 ,
Results of Measurements of Physical Properties of
Oils 	
Summary of Dispersant Characteristics 	 ,
Maximum Solubilities of Oils in Tap Water and Times
Page
18
29
50
51
53

                 to Attainment  	  122

          7   Maximum Solubilities of Oils in Salt Water and Times
w               to Attainment	  123

          8   Oil Concentrations at "Saturation"  	  124

          9   Comparison of Final Values of Parameters from Fitting
                 Dissolution Model to Experimental Data and
                 Regression Statistics  	  125

         10   Summary of Coefficients of Spreading Equations for
                 Arzew Crude 	  134
         11   Summary of Coefficients of Spreading Equations for
                 Brass River Crude 	  135
         12   Summary of Coefficients of Spreading Equations for
                 Alaskan Crude  	  136

         13   Summary of Coefficients of Spreading Equations for
                 Iranian Crude  	  137
         14   Summary of Coefficients of Spreading Equations for
                 Sahara Crude  	  138
         15   Summary of Coefficients of Spreading Equations for
                 Nigerian Crude  	  139

         16   Summary of Coefficients of Spreading Equations for
                 #2 Fuel Oil 	  140

         17   Summary of Coefficients of Spreading Equations for
                 #6 Fuel Oil 	  141
%i»,.i

                                      xi v

-------
Number                                                       Page

  18   Summary of Coefficients  of  Spreading  Equations  for
          8% Crude 	   142
  19   Summary of Coefficients  of  Spreading  Equations  for
          Lagunillas  Crude  	   143
  20   Summary of Coefficients  of  Spreading  Equations  for
          La Rosa Crude 	   144

  21   Summary of Coefficients  of  Spreading  Equations  for
          Suniland Crude 	   145
                               XV

-------
                    ABBREVIATIONS AND SYMBOLS

                                                     2
a         constant or acceleration of the oil,  cm/sec
A         surface area, interfacial  area, or volume of oil  per
                                                              2
          unit length normal  to the direction of spreading,  cm
b         constant or width of the oil  slick, cm
c         constant
                               3                 3
C         concentration,  mol/cm  or mg/l or gm/cm
                                          3
AC        concentration difference, mol/cm
C"         average concentration, mg/£
C*        interfacial concentration of oil, mg/l
(C.S)     interfacial concentration of adsorbed dispersant  mole-
          cules, mg/l
(CO)     interfacial concentration of mixed micelles, mg/l
(C OS )    interfacial concentration of adsorbed mixed micelles,
  171  m    mg/l
C         solubility in water of pure hydrocarbon, mol/cm
C         concentration of hydrocarbon in bulk  water phase,
                3
          mol/cm
d         diameter of oil  droplet, cm,  or distance between
          charged micelles, cm
D         diffusion coefficient, cm/sec
D~         average diffusion coefficient, cm/sec
e         solubility enhancement factor
E         energy of adsorption or desorption, cal/gm-mol
f         dilution factor
                                                    2
F         flux of compounds through unit area,  mg/cm «day,  or
          force per unit  volume of oil, dynes/cm
Fo        spreading force due to net surface-tension of oil,
          dynes/cm
Fm        external spreading  force by a control monolayer,  dynes/
          cm
                               xv i

-------
                                             2
g         acceleration due to gravity, cm/sec
h         thickness of oil layer, cm
Ah        thickness of oil layer above water level,  cm
H         wave height, cm
k.        constants, or reaction/diffusion coefficients,  time"
 1        (i = -3,-2,-1,1,2,...)
kn        frequency factor, time"
                                     -1
K         diffusion coefficient, time
K, 9 ,    equilibrium reaction constants
 I , L. , 3
Kp        Blokker constant
£         length of a slick,  or characteristic length of  a slick
          in the direction of spreading, or wave length,  cm
L         wave length, cm
m         number of carbon atoms or adsorption sites, or
          aggregation number  of dispersant molecules
n         number of molecules per unit cross sectional area,
                      2
          molecules/cm
                                  2
N         flux of compound, mol/cm -sec
                                 23
N.        Avogadro's Number,  6x10   molecules/gm-mol
NB        Boltzmann's constant, dynes-cm/K
0         interfacial concentration of oil, mg/£
p         number of counter-ions not attached to the micelle
                   >e
                   .2
                                   2
P         rate  of permeation,  mg/cm -sec,  or  hydrostatic  pressure,
          gm/cm.sec
                                                    2           <
AP        hydrostatic pressure difference,  gm/cm«sec
                                  3
q         volumetric flow rate, cm /sec
Q         volumetric flow rate, cm /sec
r         radius of oil  slick, micelle or single dispersant mole-
          cule, cm
r, , o    rate of reaction for steps 1, 2 and 3, mg/£-sec
  I , C. , O
r» D R    rate of reaction for adsorption,  desorption,  and
  ' '      reaction controlling steps
R         gas constant,  cal/gm-mol. K, or combined rate of
          accommodation/solubi1ization/microemulsification,
          mg/£« sec
R(t)      radius of oil  slick as  a function of time of  spreading,
          cm
                             xvi

-------
S         solubility of oil in water, mg/£, or
          adsorption site
SQ        total concentration of adsorption sites
t         time, sec or day
trj        duration of spil 1 ,  sec
t         time to attainment  of maximum concentration, day
t         Saybolt universal seconds, sec
T         absolute temperature, K
u         velocity of flowing water, cm/sec
v         velocity, cm/sec or ft/sec or knots
                    • •
                    '3
v         volumetric flow rate of oil,  cm /sec
V         volume, cm'
V         rise velocity of an oil  droplet, cm/sec
V         volume of a water sample
w         width of an oil  slick, cm or van-der-Waals adsorption
          energy associated with -CHp- group,  cal/gm-mol
W         mechanical  energy or work input, ergs
x         distance in the horizontal  direction, or coordinate of
          the center  mass of the oil  slick, cm
X         mole fraction of compound
y         distance in the vertical direction, cm
z         distance normal  to the plane across which  mass trans-
          port occurs, cm
z,        valency of  surface-active ion
  \                                                              *>
Greek Symbols
a         surface-tension, dynes/cm
p         density,gm/cm
Ap        ratio of oil and water densities
                                 2
v         kinematic viscosity, cm /sec
6         thickness of interfacial layer, cm
6x        unit length along the direction of spreading of oil
          slick, cm
e         electron charge, or energy input per unit  mass and
          time, ergs/gni'Sec
tyQ        electrical  potential of the surface
y         activity coefficient

                              xvii i

-------
                                       2
r         modified concentration, mg/cm
6         fraction of the surface which is unoccupied by adsorbed
          molecu!es
9,  2      instantaneous contact angles formed by the oil phase at
  '        the  air  and water boundaries, degrees
<)>         weight fraction of the total volume of solution
          occupied by micelles
u         absolute viscosity,  gm/cm-sec
A         ratio of width to length of slick
Subscripts
b         boundary
c         current, counter-ions or calibration
crit      critical
d         diffusion or dispersant
E         evaporation
eq         equili brium
f         forward
g         gas  phase
H         refers to Henry's Law
i         ionic, single molecule or specie
L         liquidphase
m         aggregation number of micelles or maximum
n         non-ionic or nth compartment
o         o i 1
ow        between  oil and water
p         polarization
r         reverse
s         saturation
t         total
w         water phase, wave or wind
x         horizontal direction
y         vertical direction
0         zero
95        95  percent
00         equilibrium, bulk of air or when time is equal to
          i nfini ty
                               xix

-------
                        ACKNOWLEDGEMENTS
     The authors gratefully acknowledge the assistance and
coordination provided by Mr.  Leo McCarthy,  Jr.,  Project Officer,
of the EPA Environmental Research Laboratory in  Edison, New
Jersey.

     We  wish to thank Nick Bosko for technical  assistance.
Several  individuals and organizations contributed to the success
of this  study by donating samples of crude  oils  and dispersants.
Lynn Braun is thanked for typing this report.
                               XX

-------
                            SECTION 1


                          INTRODUCTION


     After an oil  spill  occurs, several  natural  processes begin
to operate on the oil  and cause it to  disperse.   The physical,
chemical, and mechanical  processes that disperse components of
crude oils have been identified (Pilpel, 1968; National  Academy
of Sciences, 1975; McAuliffe, 1977a),  but the rates of the indi-
vidual processes remain  to be quantified.  In the early stages
of an oil spill, the major dispersal  mechanisms  are natural
spreading, evaporation,  emu!sification,  dissolution, and
mechanical transport by  winds, waves,  and current forces.  During
the latter stages of an  oil  spill, other processes (e.g. microbial
decomposition, oxidation  [atmospheric, chemical, and photo-],
adsorption onto particulate matter, sedimentation, etc.) continue
to dissipate the oil.   These processes do not act singly but
compete with each other,  and the complex nature  of the inter-
actions often makes it impossible to  separate and quantify indi-
vidual rates.  Figure  1  shows the processes involved in the fate
of oil in the marine environment.

     The action of a specific dispersal  process  on oil components
and the effect on environmental damage caused by the oil spill
depends on a combination  of factors.   Straughan  (1972) has
discussed some of the  factors:  the type of oil, the dosage rate.
of oil, the physiography  at the site  of the spill, weather
conditions at the time of the spill,  the biota of the area, the
season of the spill, previous exposures  of the area to oil and
other pollutants, and  the method used  to treat the oil.

     To predict the fate  of oil spilled  in a marine environment,
it is necessary to quantify the rates  of the different dispersal
mechanisms.  Knowledge of how fast crude oils and petroleum
products spread on calm  water and in  the presence of external
forces (winds, waves and  currents) is  essential  before management
decisions can be made  to  contain the  oil by booming and/or to
initiate skimming operations.  Information on the rates of
dissolution and characteristics of emu!sification of specific
oils is useful for assessing hydrocarbon concentration levels in
water to which marine  organisms will  be  exposed.  Until  these
processes are understood  fully speculation will  continue on the
short- and long-term effects of oil pollution on marine environ-
ments.

                                1

-------
Figure 1.   Processes  Involved  in  the  Fate  of
           Spilled  Crude  Oil  in  the Marine
           Environment (Burwood  and Speers,
           1974)

-------
     Because the limitations of oil  containment and skimming
devices in rough water are becoming  evident,  other options to
deal  with oil  spills must be considered.   The addition of chemical
dispersants to oil  slicks to stimulate the formation of oil  drop-
lets  and enhance their dispersal  into the underlying water now
appears to be  an acceptable clean-up option in view of the
irreparable damage  to coastlines  and marine life that can result
from  large oil spills.  Dispersants  are easy  to use.  In remote
areas, adequate mixing between oil  slicks and dispersants can be
obtained from  the turbulence generated by wind, wave and current
forces at sea  without additional  mixing energy.  The mechanisms
of dispersant  action on oil slicks  have not been investigated,
and a rational approach to dispersant application has not evolved.

     The number of  crude oils and petroleum products that can be
spilled on water is large.  The behavior  of oils on water even
under similar  spill conditions will  vary  due  to different chemical
compositions.   Thus, there is a need for  research studies with
several oils in order to quantify (if possible), or to determine
the order-of-magnitude estimates  of  the rates of dispersal
processes for  specific oils.  Appropriate plans for contigencies
to mitigate the damage to marine  environments can be made more
accurately when the rates and mechanisms  of oil transport are
known.

-------
                            SECTION 2


                           CONCLUSIONS


     Oily discharges to aquatic systems are usually marked by
slick formation.  Surface oil  layers that form after spills have
many undesirable impacts on the environment, such as aesthetic
damage to beaches and shoreline, reduction of oxygen exchange at
the air-water interface, fouling of wild fowl, etc.  The
potential for damage by oil spills depends on the rates of the
dispersal mechanisms and other factors.  Spreading and dissolution
of oils in the underlying water are important dispersal
mechanisms, but these rates have not been quantified for many
petroleum-based systems.

     Crude and processed oils  are extremely complex systems and
behave differently when discharged to the environment; the origin
and composition of oil, processing history, and physical and
chemical  properties determine  the rates of spreading, dissolution
and other dispersal mechanisms.  In this study, the rates of
dissolution and spreading of 12 oils were investigated under
laboratory conditions.

     Rates of spreading were determined experimentally by measur-
ing the extent of area! covering with time of small oil spills on
calm water.  Four different volumes of oils were spilled and the-
areas covered by the slicks were determined photographically.  As
the oils  have different physical and chemical properties, the
variables investigated  include, density, viscosity, surface
tension and interfacial tension between oil and water.  Observa-
tions of  the configurations of the spreading slicks indicated
that the  oils did not spread preferentially as rectangular,
circular or elliptical  slicks.  In general, the shape of a slick
varies according to the type of oil spilled, the rate of
discharge and other factors.  It is influenced by thermal convec-
tion currents and molecular motions in the water column.

     Mathematical models were  derived for oils spreading on calm
water.  The derivation  of the  spreading equations follows Fay's
work.  The principal forces influencing spreading of oils on calm
water are gravitational, viscous, inertia!, and net surface
tension.   Gravity accelerates  spreading, causing oil  slick thick-
ness to decrease and the oil to spread laterally.  Viscous and

-------
inertia!  forces retard spreading, but the effect of the latter
appears to be small.  The value of net surface tension determines
whether spreading is accelerated or retarded.  These forces are
related to the physical  properties of the oil and water phases.

     By equating an accelerating force to a retarding force,
several spreading models containing only one empirical constant
can be derived (Fay, 1971).  The models fit the experimental data
with varying accuracies.  In general, the order of goodness of
fit, from best to worst, was gravity/viscous, surface-tension/
viscous,  gravity/inertia and surface-tension/inertia.  The
surface-tension/inertia  spreading model is independent of the
volume spilled; this equation is not valid for predicting the
aerial extent of slicks.  The effects of physical properties of
the oil and water phases can be determined from the spreading
equations.  The effect of temperature on rate of spreading was
not investigated, but it can be determined from indirect
influence on the physical and chemical properties of the oil and
water phases.

     In the open sea, calm conditions do not persist indefinitely.
Gross oil transport by natural  turbulence is superimposed on
spreading and the total  area covered by a slick must be determined
by summing the transport by the two mechanisms.

     The  rates of dissolution of the oils were determined by
measuring their solubilities in water during equilibration in
open static tests.  The  oils were equilibrated with tap water at
25C for 2 to 3 weeks.  Experimental data show increases in oil
concentrations initially and decreases later during the period of
equilibration.  Similar  trends  were exhibited by the experimental
data generated by equilibrating 6 oils with salt water solution,
however,  solubilities were lower in tests with salt water, and
the oils  attained maximum solubilities at slower rates.  Satura-
tion data were determined from  long-term, closed-system experi-^
ments.  Solubilities vary for different oils and depend on oil
composition.  Organic species in solution were not characterized,
but they  are believed to be low molecular weight hydrocarbon
compounds, for example aliphatics, aromatics, and substituted
organics.

     The  experimental data suggest that as oil slicks equilibrate
with water, volatile hydrocarbon species evaporate into the
atmosphere from the air/oil interface, and soluble species
dissolve  into the underlying column of water from the oil/water
interface.  These processes are not in equilibrium; hydrocarbons
continue  to evaporate from solution after the oil layer has been
depleted  of volatile hydrocarbons.  This process occurs even when
oil is present in solution at less than saturation concentra-
tions.  A segmented mathematical model was proposed to quantify
the rates of dissolution for the duration of the equilibration
period.  This model consists of equations for the solution and

-------
evaporation phases.  The model was used to fit experimental data
from the dissolution studies.  The results of the numerical
simulation show that the model fits experimental  data for a
majority of the oils fairly well.

     The conditions during the dissolution experiments correspond
to an unlikely worst case of an oil spill such that the oil
completely covers the water surface.   Under these conditions,
experimental  data suggest that low concentrations of oil  will
persist in the water phase after 2 weeks of equilibration.  In
the field, the concentration of oil in the water column below
surface slicks will be influenced by several factors, such as
water quality.  For example, dissolved organic matter is  present
at varying concentrations in aquatic systems.  Dissolved  organic
matter can solubilize organic compounds and increase oil  concen-
tration.  Water movements will have a dilution effect on  concen-
tration, but may cause oil droplets to be transported to  the
underlying column of water.  Other processes which disperse and
degrade petroleum operate simultaneously with dissolution.  The
concentration of oil in the aqueous phase will be influenced by
the rates of these mechanisms.

     Management decision to disperse oil spills with chemical
dispersants must be based on knowledge of the action of
commercial preparations and toxicity to marine organisms.  This
study has identified the mass transfer processes that lead to
formation and dispersion of droplets in chemically-treated oil
slicks.  Mathematical equations are proposed to quantify  the
rates of some of the principal mechanisms.  These equations were
not verified with experimental data,  since necessary input data
includes information that is deemed proprietary or cannot be
determined experimentally.

     The efficiencies of 5 commercial products to disperse 3 oi\s
of varying physical and chemical characteristics were evaluated "
in a large-scale laboratory system.  The design of the wave-tank
was based on current dispersion practices and the tank permits
spatially distributed sampling.  Variables investigated include
oi1-to-dispersant ratio, oil and dispersant types, and the
effects of agitation and sea salts.

     Experimental data show that efficiency increases with the
volume of dispersants added.  Oil concentrations decrease
gradually with time after dispersion:  the rate of decrease
varied for each dispersant and oil combination.  When the system
is mixed continuously, agitation causes unstable and stable
droplets to go initially into the aqueous phase.   When agitation
ceases, unstable droplets coalesce and migrate to the water
surface, oil  concentration in the aqueous phase decreases with
time and, finally, stabilizes.  The dispersants are classified
according to the efficiency of their action under test conditions
by measuring the quantity of extractable organic material in

-------
water samples.

     This study investigated the rates of spreading, dissolution
and chemical  dispersion of petroleum:   these processes are
important mechanisms for the dispersal of oil  spills.   Previous
research studies have identified the key processes and the rela-
tive importance of each process but research was limited to a few
crude oils.   Two crude oils from different sources may behave
differently when spilled, even when their physical properties
(density, viscosity, and surface tension) have similar values.
Models for predicting the rates of spreading must be correlated
with oil properties that can be measured easily using  a represen-
tative sample of the oil slick.  From  the point of view of those
charged with  the responsibility of containing oil slicks, models
based upon gross parameters (e.g.  oil  slick thickness) are of
limited value due to the difficulty and error of measuring this
parameter during the life of a slick.   Calm conditions do not
persist indefinitely in the field  and  ultimately, transport of
gross oil by  mechanical forces is  superimposed on natural
spreading.

     The interactions of wind, waves,  and tides in the presence
of oil slicks cannot be adequately simulated in the laboratory.
The mechanical transport of oil due to these forces is probably
more important than natural spreading  in the overall dispersion
of oil slicks, when the damage to  a coastline is considered.   But
the effects of these forces on oil slicks are known in general
terms; oil slicks become elongated and distorted.  Usually, the
slick breaks  into patches which drift  in the direction of the
wind at a speed proportional to the sum of the vector  velocities
created by transport forces.  The  influence of tides is minimal
due to the periodic and oscillatory nature of tidal movements.

     The dispersal of spilled oil  by applying chemical dispersants
appears to be a promising method of cleaning-up oil spills.
Proper use of dispersants could result in the efficient dispersal
of oil even in the absence of wave action.  A gap in knowledge
exists:  the  mechanism of action of dispersants must be better
understood,  and the rates of spreading and dissolution must be
established for a larger variety of crude oils and petroleum
products.

-------
                            SECTION 3


                         RECOMMENDATIONS


     Research on the rates of all  oil dispersal  mechanisms is
needed to improve mass balance calculations.   The variety of oils
that can be spilled is large and marine ecosystems vary in
susceptibility to the detrimental  effects of  spilled oils.

     Research in the following areas is recommended:

1.   Direct air sampling above slick surfaces  to  determine rates
    of evaporation of hydrocarbon  species;

2.   Rates of adsorption of oils on different  sediments should be
    determined to characterize adsorption isotherms;

3.   Decreases in rates of oxygen exchange across the air/sea
    interface, in the presence of  different oils and film thick-
    nesses ;

4.   Improvements in analytical determination  of  oil  concentra-
    tions can be realized by coating sampling devices and glass-
    ware used for extraction, storage,  etc.,  with hydrophilic
    films;
                                                                V.
5.   Experiments in coastal environments must  be  intensified to
    improve knowledge of the rates of physical  oil dispersal
    processes and effects of oil on aquatic biota;

6.   In the absence of information  on the composition of commercial
    dispersants, analytical  techniques  must be  developed to
    characterize water samples from dispersion  studies in terms
    of only the hydrocarbon  species contributed  by oil slicks;

7.   More critical comparisons of different dispersants for
    general detergent power  are needed;

8.   Further studies on the mechanisms of dispersant  action are
    required, with emphasis  on the effects of chemical structure
    of surfactants, solvent  types, and  other  active  components on
    dispersion efficiency;

-------
 9.   Studies  on  toxicity  of  dispersants  to  marine  life  will
     increase knowledge on environmental  impact  aspects  of
     dispersants;

10.   Rates  of penetration of  oils  and  chemically-treated  oils
     into  sediments  deserve  to  be  investigated;  and

11.   Further  development  of  the  phenomenological theories of
     dispersant  action  is required.

-------
                            SECTION 4


                       RESEARCH OBJECTIVES
     Oil  spills occur in the sea at random and without prior
warning.   A study of an actual  spill  in the field is expensive
and environmental conditions are changing constantly.   Response
to oil  spills is usually slow and measurement of important para-
meters  for correlation purposes is difficult.  Until recently,
public  opinion on field trials  with small "experimental" spills
has been  unfavorable.  Thus, inferences on the behavior of oil
spills  in the marine environment must be drawn from the results
of laboratory-scale experiments.

     Because of the complex nature of the interactions of disper-
sal processes, laboratory experiments simulating one dispersal
process and excluding all other processes are impossible to
design.  Monitoring changes in  the properties of the oil in the
upper layer without physically  perturbing the system is an
insurmountable problem.  In view of these limitations, the major
research  objectives are:

     1.  To investigate experimentally the spreading rates of
some crude oils and petroleum products on calm water,  during oil
slick initiation (approximately 30 minutes);

     2.  To develop rate equations from experimental data deter-
mined in  'I'-- important independent  variables to be correlated
in the  equations are oil volume, viscosity, density, and surface
tension.   These parameters are  important for determining the
spreading rates of oils.  They  are recognized to be most
desirable for correlating data;

     3.  To determine experimentally  the dissolution rates of
some crude oils and petroleum products in water - the  goal is to
quantitate the short-term distribution of total hydrocarbon in
water following an oil spill;

     4.  To evaluate the performance  of commercially available
dispersants in emulsifying and/or solubilizing crude oils in
water - the major criterion for this  evaluation is the additional
solubility, beyond natural  dissolution, due to treatment of oils
with dispersants.  A detailed mechanism to explain the action of

                               10

-------
       chemical dispersants on oil slicks is to be proposed; and

            5.  To study the surface slick environment with respect to
       surface disruption, droplet formation and motion to the extent
       possible with available instrumentation.

            The influence of factors such as temperature, water quality,
       environmental factors, etc. are considerable.  Attempts must be
       made to appraise these factors.

            Spreading, dissolution and chemical dispersion of oil spills
       are processes of significant interest in their own rights.  The
       major  reasons for exploring them lies in the belief that an
       increase in knowledge will assist efforts to:

            a)  assess the environmental impact of oil spills accurately,

            b)  design appropriate clean-up devices,

            c)  plan for contingency operations and execute plans
                successfully in the event of oil spillage,

            d)  develop sound rationales for dispersant selection and
                application, and

            e)  safeguard and protect the environment from damage caused
**"*              by oil spills.
                                      11

-------
                            SECTION 5


           LITERATURE REVIEW OF RELEVANT PREVIOUS  WORK


     The literature on oil  pollution of the marine environment
is of considerable bulk and still  growing.   A thorough review of
all  aspects of oil pollution, e.g.  oil  spill  magnitude,  legisla-
tion, containment, treatment and clean-up,  forecasting,  environ-
mental  effects, etc., is a  major task and outside  the scope of
this research.  Therefore,  this review will be limited to  the
literature on the specific  oil  dispersal processes to be investi-
gated.   The reader is directed  to  several  published reports
(Pilpel, 1968; Blumer and Sass, 1972; Boesch et al.,  1974;
National Academy of Sciences, 1973a,b,  1975;  Baker, 1976;  Smith,
1968, Hoult, 1969; Hepple,  1971; Wolfe, 1977; Nelson-Smith, 1972;
American Petroleum Institute, 1969, 1971,  1973a,b, 1975, 1977a,b,
1979)  for a general  overview of oil pollution of  the marine
environment.

     The environmental impacts  of  spilled oil are  numerous:  fire
hazard, fouling of beaches, damage  to shore property  and fishing
grounds, attenuation  of light and  gaseous oxygen transfer  between
the atmosphere and aquatic  biota,  etc.   The literature on  the
adverse biological effect of oil spills on  marine  organisms is
voluminous.  Because  of the proliferation of assay procedures
used to assess the toxicity of  oily pollutants to  marine life,
there is disagreement on the short- and long-term  effects  of oil
spills  on aquatic biota, in spite  of much published work in this
area.  In general, the long-term effects of oil  pollution  on
aquatic species have  neither been  fully investigated  nor under-
stood.   The extent to which oil is  retained,  metabolized or
excreted following ingestion by marine fauna  is not known.   The
National Academy of Sciences (1975) and the American  Petroleum
Institute (1973b, 1977b) have summarized,  and documented numerous
short-  and long-term  effects of oil spills  on marine  organisms.
Other pertinent references  on the  biological  effects  of  oil
spills  can be found in the  general  literature on oil  pollution
cited above.  Moore and Dwyer (1974) have reviewed the literature
on this subject.

     Following an oil spill, several physical, chemical, mechani-
cal  and biological processes begin  to act on  components  of  the
oil, as shown in Figure 1.   Pilpel  (1968)  classified  the

                               12

-------
processes into two types:  dispersal  and destructive.  Dispersal
processes include aerosol formation,  evaporation, spreading,
emulsification,  dissolution, solubi1ization ,  sedimentation and
mechanical transport by wind, wave, and current forces.
Processes that dissipate oil spills include oxidation and micro-
bial degradation.  The destructive processes  and some of the
dispersal processes are selective, i.e. they  act on certain
components of petroleum.  Several  factors influence the  extent
of dispersion and destruction of oil  spills:   physical and
chemical  characteristics of the oil and environmental conditions.
Important characteristics of the oil  are:  total volume, specific
gravity,  viscosity, chemical composition, type and extent of
refinery  treatment.  Wind and current  velocities, wave climate,
ambient temperature, salinity, location of spill and time of the
year are  important environmental  factors.

     Some processes operate on some oil components more  rapidly
than others and  there are numerous interactions among the
processes.  The  interactions are so complex and poorly understood
that a recent attempt to develop a mass balance model of oil
spills was marginally successful  (American Petroleum Institute,
1977c).  An oil  cycle similar to the  nitrogen cycle may  be
considered to exist, but speculations  will continue on the sinks
for petroleum.

     The  discussions to follow focus  on previous works on the
spreading, and dissolution of oil  slicks.  Section 6 discusses
chemical  dispersion and emulsification.


SPREADING

     Spreading of oil on water can be  viewed  as consisting of
two parts:  the  natural tendency to spread on calm water in the<
absence of external forces and spreading as a result of  mechani-
cal transport due to wind, wave,  and  current  forces.  Both phases
of spreading are important, but this  section  focuses primarily on
the former.

     As the first few drops of oil are released onto water, the
oil spreads horizontally and spontaneously.  This phenomenon has
been observed by several investigators (Blokker, 1964; Berridge
et al . , 1968a, O'Brien, 1970) and  is  called flashing or  flash
spreading.  Spreading occurs faster at the edges than in the bulk
of the oil causing a thin film to  be  formed very rapidly at the
water surface.  The rate of spreading  is rapid in the initial
stages but decreases with time.  Flash spreading is a function of
the quantity of  surface active compounds in the oil and  the
underlying water.  All crude oils  and  petroleum products do not
flash.  For example, pure hydrocarbons usually form lenses on
water  (Langmuir, 1933).


                               13

-------
     One of the earliest studies of the spreading of oil is
reported by Langmuir.   He studied the formation of oil lenses on
water by pure hydrocarbons and derived the following equation
for calculating the equilibrium thickness of an oil lens:
                     .2
                      00
where

     h           = the equilibrium thickness of the oil lens,  cm
      oo
     a ,  a ,  a   = surface tensions of oil  and water, and inter-
                   facial tension between oil  and water,
                   dynes/cm, respectively

     p ,  p       = density of oil and water,  gm/cm ,
                   respectively

     g           = gravitational  constant

Langmuir's equation was derived for pure hydrocarbons and its
ability to predict the spreading  behavior of crude oils and
petroleum products is severely limited.

     Some pure hydrocarbons spread on water.  Pomerantz et al.
(1967) studied the spreading behavior of low-molecular weight
hydrocarbons.   They found that a  balance on the surface tension
forces determined whether the oils spread or formed lenses.  The
spreading force of the oil is defined as




where Fo  is the spreading force (dynes/cm).  Surface tension
forces acting  on an oil slick are shown in  Figure 2.  The oil
will spread if Fo is positive.  If Fo is negative, the oil will
form a lens.   Aging processes cause changes in a , a  and a  ,
                                                WO      O W
and Fo is not  constant with time.  Thus, the spreading force can
be positive during the initial stages of spreading and become
negative  during the final stages.  Since Equation (1) can be
written as

                        2     -2F°(PW)
                       h  ™                                  f "3 \


the equilibrium thickness (h^) is undefined when Fo is positive.

     The  presence of indigenous surface active components in oil
   water  may  exert a counter force against  the spreading force
or

                              14

-------
of oils.   Therefore, use of monomolecu!ar surface films to
confine spreading oil  slicks has been studied by several  investi-
gators (Garrett, 1969,  Garrett and Barger,  1970; Cochran  and
Scott, 1971; and Milz  and Fraser,  1972).   In the presence of an
external  spreading force, a modified Langmuir's equation  is used
to calculate the thickness of oil  lenses:
                             -2(Fo-Fm)p
                                       w
                              gpo(pw-po)
                                                             (4)
where
     h = thickness of the oil  lens, cm

    Fm = external  spreading force applied by the control  mono-
         layer, dynes/cm

     Some spreading oils form  thin continuous layers that are a
few molecular diameters thick.   On the sea,  such oil films are
responsible for the irridescent colors that  are often the first
tell-tale signs of oil  pollution.  A majority of oils do  not
spread into films  that  are continuous.  Spreading oils usually
break up into smaller patches  and/or pools that are separated by
streaks of oil-free water.  Because spreading patterns are random
and irregular, the final configuration for an oil slick is
unpredictable.  Generally, spreading causes  an increase in the
area covered by oil slicks beyond that expected if the oil
remained stationary.
                                                w
                           ow
                            wate
           Figure 2.
Surface Tension
on an Oil  Lens
Forces  Acting
                               15

-------
     Blokker (1964) studied the spreading of a few oils in rec-
tangular and circular geometries.   He found that after a slick
attained a mean thickness of 2 cm, the spreading rate became
independent of the spreading history, but was approximately
proportional to the instantaneous  mean layer thickness.  The
following equations were derived for the rate of spreading of a
homogeneous oil slick:
     (a)  Circular slick
— v
Pw
                                                             (5)
     (b)  Rectangular slick
                                                             (6)
where
     dt(&t) = diameter (length)  of the slick at time t,  cm

     d (H ) = initial  diameter (length)  of the slick, cm
          U,
          W
            = specific gravity of the oil  and water,  respectively
         V  = volume of the oil,  cm

          b = width of the oil  slick or channel,  cm

          t = time, min
            = the Blokker coefficient,  min
                                          -1
Blokker observed pronounced flashing with some oils;  this phenome-
non was attributed to the presence of surface active  components
in the oils.

     Berridge et al.  (1968a) and Jeffrey (1971) used  Blokker's
equations to  correlate their experimental data.  Results showed
wide scatter  and they reported wide variations in the values of
the empirical coefficient (the Blokker coefficient)  determined
for the oil  types tested.

     Fay (1969,1971)  developed a theory for oil spreading on calm
water.  Equations that give order-of-magnitude estimates of the
increase in  area of an oil  slick with time were developed from
consideration of the  forces that act on an oil slick.  According
to Fay (1969) there are four major forces acting on  oil  during
spreading:   gravity,  viscous, inertia, and surface tension.  On
the assumption that oil is  a homogeneous liquid with  constant

                              16

-------
properties, Fay identified three distinct stages during  the
spreading process that depend on the major forces acting at  the
time.  Spreading forces are balanced by retarding forces in  each
stage of the spreading process.  The three stages and the
sequence in which they occur are gravity-inertia, gravity-viscous,
and surface tension-viscous.  Figure 3 is a representation of  the
forces.  Gravity and surface-tension forces tend to spread the
oil, while inertia and viscous forces impede spreading.

     The force of gravity causes a horizontal spreading  motion of
an oil  slick by creating a pressure gradient between the pool  of
oil and surrounding water.  An alternate explanation is  that
floating oil has an elevated center of gravity and greater poten-
tial energy.  As a consequence, the lighter oil seeks a  constant
level and spreads sidewise, losing excess potential energy.  As
the oil spreads, pool thickness diminishes and the gravity force
decreases.  Surface tension forces act at the leading edges  of
the oil slick as shown in Figure 2.  Table 1 summarizes  the
spreading equations derived by Fay for each spreading regime.
                                         «•• *-.-«*•*»«.)
                                      SLFFACE  TENSION
                                        FRICTION
               Figure 3.  The Four  Forces which Act on
                          an Oil Film during  Spreading
                          (from Fay, 1971 )
                               17

-------
       Table 1.  Spreading laws for oil  slicks (Fay, 1971)
        Regime                 One-dimensional




   Gravity-Inertia         i = 1.5(AgAt2)]/3                 (7)


                                      •) -3/9  "1/9 1/4

   Gravity-Viscous         I = 1.5(AgA^tJ/Vv1'  )             (8)
                                             w


                                     ? -3  o   1/4

Surface-tension-Viscous    £ = 1.33(a t /p'v )               (9)
                                          W W



                                Axi symmetric




   Gravity-Inertia         r = 1.14(AgVt2)1/4               (10)




   Gravity-Viscous         r = 1.45(AgV2t3/2/vyV/6       (11)
                                              W



Surface-tension-Inertia    r = 2.3(a2t3/p2,vll)1/4            (12)
                                         W  W
where
       A = 1-P0/PW
     £,r = length and radius of the oil  slick, cm, respectively „

                                            3
   p ,p  = densities of oil  and water, gm/cm , respectively
    o  w
                                         2
       g = gravitational  constant, cm/sec



       A = volume per unit length normal  to the direction of
                        2
           spreading, cm



       V = total  volume of oil, cm

                                           2
      v  = kinematic viscosity of water,  cm /sec
       W


a ,a ,a   = surface tension of oil, water  and interfacial  tension

           between oil  and water, dynes/cm, respectively



Fay (1971) contends that, as time progresses, a large spill will

pass through the three  stages of spreading in succession.  How-




                               18

-------
ever, a small  spill  follows surface tension-viscous spreading
from the start.   Fay derived an equation to estimate the maximum
area of an oil  slick
     A = K
              .2,,6
                      1/8
(13)
where
     a  = surface tension of oil, dynes/cm
      S = solubility of the oil  in water, gm/cm
                                             2
      D = diffusivity of the oil  in water, cm /sec
      K = a constant of order unity
      V = volume of oil, cm
                                                2
     vw = y/p  = kinematic viscosity of water, cm /sec
     p  = density of water, gm/cm
      w
Fay (1969) extended these spreading models to account for the
spread of a slick from a steady source in a moving stream.   For
the case of a  one-dimensional slick, the following equations
were derived:
                                  2   1/3
     Gravity-inertia     a = (Agvx /u)                      (14)
     Gravity-viscous     a = (Agv2x3/2/v,1/2u7/2)1/4         (15)
                                        W
     Surface-tension-          232   3 1/4
          viscous        £ = (a x /p v u )                  (16)
                                    W W
where
     v = rate  of discharge of oil, cm /sec
     x = distance from the source in the direction of flow,  cm
     u = water velocity, cm/sec
There is disagreement on the numerical values of the coefficients
in Equations (7) through (12) (Hoult and Suchon, 1970;  Fannelop
and Waldman, 1971).  These equations have been used to  correlate
experimental data for actual spills.  The results  appear satis-
factory.
                              19

-------
     O'Brien (1970) conducted experiments  on the spreading  of
oil  on water from a stationary leaking source and found constant
spreading rates after an initial  time interval.   Also,  surface
tension had a greater influence on spreading than did the gravity
force.

     Hoult (1972) showed that there are similarity solutions  to
the  boundary-layer equations that predict  the same results  as
Fay's equations.

     Buckmaster (1973) studied gravity-viscous spreading and
derived an expression for oil slick size as a function  of time:


               R(t) = 1.76(gA)1/4V1/V1/8t3/8              (17)
                                      W

where R(t) is the half width of the oil  slick and g,  A, V,  v, t
have been defined previously.

     By solving the diffusion equation, Murray (1972) derived
expressions for the maximum width and length of  a slick, as a
function of the concentration of oil  in the water, i.e. the
boundary concentration:


                       w = - *4~ -                     (18)
                           (2Tre)1/2vcCb



                       a = —3 — ,                          (19)
where
                                                               <
     w = width of the slick,  cm

     I - length of the slick, cm

    Cu = apparent boundary concentration contour,  gm/cm

     Q = rate of emission of  the source, gm/sec
                                         2
     K = diffusion coefficient of oil,  cm /sec

    v  = ambient current speed, cm/sec
     t*

These equations were derived  based on the assumption  that  slick
geometry is approximately elliptical.  Therefore,  the ratio
(X) of slick width to length  and the area (A)  of the  slick are:
                              20

-------
                           4(2Tr)1/2KC.
                           - r— ^                      (20)
                       A = _ Q _                   (21
                                                            (
Murray (1972) claimed his analysis is in good agreement with
observations made during an oil  spill.

     Other investigators have studied the spreading of oil  on
surfaces other than water (Glaeser and  Vance, 1971; McMinn, 1972;
Chen et al., 1974; Chen and Charles, 1976; Weiskopf and Uzuner,
1979).  The spreading history of oil on ice was found to be
similar to that of oil on water, except for an additional  effect
due to the roughness of ice surfaces.

     Chen et al .  (1974) investigated the gravity-viscous spread-
ing of crude oils on artificially prepared ice surfaces and
correlated experimental data with


                -^j - 0.24(tpogV1/3/y0)1/5 + c             (22)


where

     R = radius of the oil  slick, cm

     V = oil volume, cm

     t = time, sec

    p  = density of the oil , gm/cm

    yQ = viscosity of the oil, gm/cm-sec
                                       2
     g = gravitational constant, cm/sec

     c = constant

Chen and Charles (1976) studied  the surface tension-viscous
spreading of oil  on ice and fitted experimental data with
KV
                           (c-1)/6(tg1/2/V1/6)c             (23)
where V,c,t,g have been defined,  and K is a  function  of oil  type
and temperature.

                              21

-------
     The efforts of many investigators have increased present
knowledge of the spreading of oil  on water, however,  all  of the
approaches and equations proposed  to describe spreading rates of
oils rely on one or more of the following assumptions:

     1.   the thickness of the oil  is considered uniform during
         spreading;

     2.   cessation of spreading occurs when a minimal thickness
         has been achieved;

     3.   slick configuration is either rectangular,  circular or
         elliptical throughout spreading;

     4.   no consideration is given to changes, with  time, of
         physical and chemical properties of the oil  and
         variations in composition;  and

     5.   there is no termination of  spreading.

     Crude oils and petroleum products are complex mixtures of a
very large number of hydrocarbons  and other compounds (Rossini,
1960).   Each oil component has specific properties,  but the bulk
oil  properties are not simple functions of component  properties.
Oil  properties vary with time from the moment the oil is  spilled.
The most important causes of this  variation in calm  water during
long exposure times are the actions  of weathering processes:
evaporation of the lighter components of the oil  mixture  into
the air, dissolution of the soluble  components of the oil
(surface active compounds, low-molecular weight hydrocarbons and
polar compounds) and, possibly, emu!sification of the oil
components attached to highly energized surface active  components,
These weathering processes cause changes in oil density,
viscosity and surface tension.  During the latter stages  of
spreading, other weathering processes, e.g. bacterial degradation
contribute to varying oil properties and composition  (Zobell,
1969).   These changes, in addition to the thinning of the oil
slick,  may retard and eventually stop the spreading  process.

     Spreading is one of the most  important oil dispersal
mechanisms.  The extent of the surface area of the spreading oil
at any  time influences the rate of other dispersal processes.
Thus, a  thorough mathematical treatment of spreading  is desirable,
Because  all dispersal processes are  interrelated in  a complex
way, experimental verification of  existing theories  on  spreading
will continue pending the derivation of better models.


OIL TRANSPORT DUE TO WIND, WAVE AND  CURRENT FORCES

     When considering the vulnerability of a coastline  and
hazards  to fish populations from large oil spills, the  spreading

                               22

-------
characteristics of oil, particularly mechanical  transport by
wind, wave, and current and tidal  forces are very important.
The fact that oil  has a calming effect on waves  at sea had been
known since the days of Benjamin Franklin (Giles and Forrester,
1970).  But, oil  pollution and, in particular,  the action of
these external  forces did not attract research  efforts until
recently.   Since  ignorance of the  interaction of these forces
and oil  slicks  exists, a brief review of some of the significant
contributions to  present knowledge follows.   Also, Fallah and
Stark (1976a) have provided an overview and  a comprehensive
bibliography of many scattered publications  on  the movement of
spilled  oil at  sea.

     The literature  is replete with studies  on  the effects of
wind, waves and currents on transport processes  at air-sea inter-
faces.  It is clear  that despite numerous observations of these
phenomena, during  the past two centuries, understanding has not
improved significantly.  Wind, waves and currents interact in a
complex  way and cause gross motion of water  on  the sea surface.
In the presence of oil slicks, the influence of  these forces
become difficult  to  uncouple and to quantify.

     Several investigators (Teeson et al, 1970;  Schwartzberg,
1971; Warner et al., 1972; Smith,  1977) have studied the effects
of wind  velocities on oil slicks;  it is difficult to uncouple
transport due to  wind from that due to waves and currents.
Winds blowing over oil slicks induce several movements:  oil is
moved relative  to  the surface waters by wind-induced drift
(leeway),  by wind-induced surface  currents (wind shear), and by
wind-induced water transport.  The principal motion depends on
the wind velocity.

     Teeson et  al. (1970) experimented with  circular polyethylene
sheets and found  drift speeds that varied from  2 to 4% of the
wind speed.  This  range of values  is in agreement with reported"
values of wind-induced surface drift (Wu, 1968).  Other investi-
gators have observed oil drift velocities that  range from 3 to
4% of the wind  speed (Smith, 1968; Nelson-Smith, 1972).

     Smith (1977)  determined the leeway of oil  slicks and found
a linear relationship between oil  leeway and wind velocity:
                    v  = 0.0179v  + 0.0196
                                w
(24)
where
     VQ = oil slick leeway, knots

     v,, = wind velocity 10 meters above the water level,  knots
      W
                              23

-------
The oils that were tested
wind speed.   In the range
had no significant effect
drifted at approximately 3.6% of the
of wind velocities studied, oil  type
on leeway.
     The influence of variable wind velocities on surface drift
was investigated by Warner et al.  (1972).   They derived model
equations for the horizontal  components of drift velocity for an
oil patch and claimed good agreement between predictions and the
observed trajectory of an actual  spill.

     However, there is disagreement on the direction of wind-
induced oil  drift.  Teeson et al .  (1970) measured the angles of
deflection of circular polyethylene sheets drifting on water
under the action of wind.  Their  results showed angles of deflec-
tion that varied from 0 to 20 degrees from the direction of the
wind.  But Murray (1975) observed  an actual  spill and found the
oil slick to be oriented 10 to 40  degrees  to the right of the
local wind direction.

     Since the behavior of winds  can be characterized as random,
Tayfun and Wang (1973) used Monte  Carlo methods to simulate the
movement of oil slicks by winds and currents.   Two techniques
were used in the simulation:   random walk  and  time series.   The
models were applied to simulation  of the movement and distribu-
tion of an instantaneous oil  spill, with minimal success.

     The combined effect of wind,  wave, and  current forces  was
investigated by Teeson et al . (1970).  They  concluded that  oil
lenses drift at a rate equal  to the vector sum of the individual
drift velocities.  On the basis of experimental data, Schwartzberg
(1971) concluded that high winds  cause waves and the presence of
waves reduces wind drift.  He correlated the combined effect of
winds and currents with
                            = v.. + 0.56v.
                               w
                                   (25)
where
     VQ = combined drift of the oil  slick,  cm/sec

     v  = wind drift,  cm/sec
      W

     v  = current drift, cm/sec


Schwartzberg (1971) found that waves produced negligible small
drift in the absence of wind.   In the presence of wind,  waves
reduced the drift to about 2.7%.
     In the presence of both wind-driven and
Hoult (1972) proposed the following equation
drift of the center of mass of an oil  slick:
                   tidal  currents,
                   for the combined
                               24

-------
                         at • »c + °-035««                   <26>

where x is the coordinate of the center of mass of the oil  slick,
cm.   Hoult accounted for the effect of wind-driven waves in the
term  0.035v  and suggested that additional  terms be included in
            W
the  equation if waves are present.

     Laboratory studies (Alofs and Reisbig,  1972; Reisbig and
Pottinger, 1973) have shown that the transport of oil  slicks by
wave-induced drift will be higher than that  predicted  by Stokes1
theory for gravity waves.

     Reisbig and Pottinger (1973) proposed two equations for
wave-induced drift of oil slicks.  The equations for small  and
large oil  slicks are
                     vo - (^)2vw+0.445({f)vw                 (27)
and
         vn = (T^V  +  1 .3sin[1638(r - 0.005)].(l-e L )       (28)
          0     L    W              L.
where

     v  = oil  velocity, cm/sec


     v  = wave vecocity, cm/sec
      W

      H = wave height, cm

      L = wave length, cm

      d = slick diameter, cm

     Theoretical  analysis of the motion of oil  slicks under the
influence of currents and wind has been presented by Shukla and
Stark (1974).   They considered random wind characteristics and
derived a probabilistic model  to describe the displacement of the
center of gravity of an oil  patch using a Markovian time series
approach.  This model was developed further by Fallah and Stark
(1976b) who included wind shift at random intervals and transi-
tional drift due  to the sudden shifts in wind direction.

     The subject  of oil spill  forecasting is growing as numerous
models claim ability to accurately predict the trajectory of oil
spills (Hess and  Kerr, 1979).   Because of diversity of assump-
tions in these models and the  absence of accurate input data, a
majority of the models are impotent.   The mechanics of oil-sea

                                25

-------
interactions are poorly understood.   Thus, deficiencies exist in
all  models and their predictive capabilities are questionable
(Stolzenbach et al., 1977; Lissauer  and Murphy, 1979).

     In summary, waves, wind, and tidal currents cause  gross
motion of oil  slicks and increase their areal  extent over that
expected for oil spreading naturally on calm water.   Drifts due
to tidal currents and waves are not  additive,  but the effect of
tidal  currents may be negligible because of their oscillatory
nature.  The effect of wind is to cause the oil to drift along
the direction  of the wind and to induce currents in  the water
column.  When  both wind-driven waves and tidal  currents are
present, complex interactions occur  and quantification  of indivi-
dual  drift rates is difficult.  A simple approach involves
addition of the individual drifts.   Waves, wind, and tidal
currents elongate and distort surface slicks.   Furthermore, these
forces at sea  cause oil slicks to break up into patches of differ-
ent sizes.  The motion of each patch may become independent of
the parent slick; large patches may  reform and  continue to be
influenced by  the velocity and direction of wind, waves and
currents.  The ultimate goal  of any  predictive  model for the
motion of oil  slicks on water is a  combination  of the two types
of oil transport processes, i.e., spreading on  calm  water and
gross motion produced by wind, wave, and current forces.


DISSOLUTION

     When oil  is spilled on water,  evaporation  and dissolution
compete for the low molecular weight hydrocarbons.  Since the
process of evaporation is more important during the  initial
stages of an oil spill, it has received more attention  than
dissolution.  Rates of dissolution  of crude oils and petroleum
products are important because they  influence  the hydrocarbon
burden that aquatic biota must bear.

     Detailed  studies of the dissolution rates  of crude oils and
petroleum products in water are scarce.  Most  of the work on the
dissolution rates of oils in water  have involved determination
of aqueous solubilities of single hydrocarbons  and simple mixtures
of hydrocarbons.  Since knowledge of the solubility  of  hydro-
carbons in water is vital to understanding the  toxicity of crude
oils  to marine organisms, a brief review of hydrocarbon solubili-
ties  follows.

     Several investigators have studied the aqueous  solubility
of single hydrocarbon compounds and  mixtures of hydrocarbons
(McAuliffe, 1963,1966,1969a,b; Leinonen et al. , 1971; Leinonen
and Mackay, 1973; Sutton and Calder, 1974,1975; Eganhouse and
Calder, 1976;  and references cited  inter alia).
                               26

-------
     A few investigators have tried to relate hydrocarbon solu-
bility to some measurable property (properties)  of the hydro-
carbon.   Bohon and Claussen (1951) found an inverse relation
between  solubility and molar volume for aromatic hydrocarbons.
McAuliffe (1966) experimentally determined a linear relation
between  the logarithm of solubility and molar volume for hydro-
carbons  of the same homologous group.   Hermann (1972)  showed that,
for a large number of hydrocarbons, the logarithm of solubility
varies linearly with the calculated size of the  solvent cavity
just large enough to fit a solute molecule.

     Leinonen et al . (1971) and Leinonen and Mackay (1973)
proposed a correlation for predicting  the solubility of hydro-
carbons  in water.  Eganhouse and Calder (1976) studied the  solu-
bility of medium-molecular weight aromatic hydrocarbons singly
and in combination.  Experimental data they presented  show  the
effects  of mutual solubilities, i.e.,  the effect the presence of
one compound has on another.  Their data showed  mutual decreases,
mutual increases or no change in solubility.

     Because solubilities reported for certain hydrocarbons are
higher than values calculated based on vapor pressures and  ideal
solution assumptions (i.e. Henry's Law), Peake and Hodgson  (1966,
1967) suggested that some hydrocarbon  compounds  can be
"accommodated" in water.  Through the  process of accommodation,
hydrocarbon molecules "dissolve" in water beyond usual thermo-
dynamic  solubility limits.

     Factors that influence hydrocarbon solubility include  pH,
temperature, the concentration of dissolved salts and  the
presence of surface-active agents in the water column.  Electro-
lytes in solution increase or decrease solubility by causing
salting-in or salting-out of hydrocarbon molecules (McDevit and
Long, 1952).  Gordon and Thome (1967a) and Eganhouse  and Calder
(1976) have shown that salting-out occurs in natural  and artifi-1*
cial seawater, but the effect is fairly small.  Salting-out of
hydrocarbons is directly proportional  to salinity.  Klevens
(1950) and Gordon and Thorne (1967b) found that  the presence of
surfactants increases the solubility of hydrocarbons,  if the
surfactant concentration in water exceeds the critical micelle
concentration.

     The molecular basis of hydrocarbon-water interactions  from
the perspective of the degree of aggregation of  hydrocarbon
molecules in water is not understood.   This subject has been
reviewed by Shaw (1977).  According to Shaw there are  three
degrees  of aggregation:

     1.   true solution in the thermodynamic sense, i.e. molecularly
dissolved hydrocarbon;

     2.   colloid and molecular aggregations less than  1 ym  in
size; and,
                               27

-------
     3.   particles larger than 1  ym in size.

These categories are not rigidly separated and hydrocarbon
species  exist in a continuum from molecularly dissolved to
visible  particulates.

     Despite the variation among the values of aqueous solubili-
ties of  hydrocarbon compounds reported by different investigators,
it is widely accepted  that aromatic hydrocarbons are more soluble
than aliphatic hydrocarbons.  The solubility of a given oil  in
water will  depend largely on the individual solubilities of the
hydrocarbon components,  mutual interactions between the solubili-
ties, the degree of accommodation, salting-in and salting-out
phenomena,  and the effects of natural  solubilizers (surfactants)
present  in  the water or  in the oil.  Low molecular-weight hydro-
carbons, particularly  the aromatic types, will dissolve in water
rapidly; higher molecular weight hydrocarbons will be leached
gradually into the underlying water, at a reduced rate.  Dilution
in the water column is important in determining the overall
dissolution rate.  The degree to which dissolution is complemented
by the formation of emulsions is an important factor, also.

     Several investigators have reported solubilities for crude
oils and petroleum products in seawater (Boylan and Tripp, 1971;
Parker et al., 1971; Smith and Maclntyre, 1971; Frankenfeld,
1973; Gordon et alI. , 1973; Anderson et al., 1974; Boehm and
Quinn, 1974; Lysyj and Russell, 1974).  Different methods were
used to  a)   introduce  the oil to water, b)   equilibrate the  oil/
water systems, and c)   measure the concentration of oil in water.
The solubility values  are summarized in Table 2.

     The results of experiments conducted under closed static
conditions  will  be different from those of open systems.  Also,
the duration of the equilibrium process has an effect on the
quantity of oil  dissolved in water.  For example, Lysyj and
Russell  (1974) found an  accelerated dissolution of oil into  the
aqueous  phase after a  stabilization period  that lasted up to  8
days.  These salient points must be considered when comparing
solubility  values from different studies.  Unfortunately,
pertinent information  is not always provided.

     Measurements of the concentration of oils in water under oil
slicks are  sparse because of the limitations  of sampling.  It is
not an easy task to sample spatially under  an oil slick without
disturbing  the floating  oil.  McAuliffe et  al. (1975) measured
the concentrations of  dissolved hydrocarbons  under controlled oil
slicks in open water.   The highest concentration of oil reported
was 60 yg/£ in a sample  collected 20 minutes  after the spill.
Spatial  variation of oil concentraiions in  the water could not be
reported because of the  sampling  scheme used.  Also,  their
samples  were probably  contaminated because  some samples were
retrieved by repeatedly  dipping a steel  bucket into the slick.


                               28

-------
                         TABLE 2

        Solubility  Values for Some  Crude Oils
               Petroleum-Derived Products
         and
011 (s)
Kuwait crude

Kuwait crude

Three fuel oils

Venezuelan crude 1
Venezuelan crude 2
Louisiana
North African crude
#2 fuel oil
Bunker fuel
Venezuelan crude
and two fuel oils
#2 fuel oil
12 fuel oil
Bunker oil
Crude oil
South Louisiana crude
Bunker C residual
#2 fuel oil
Kuwait crude
#2 fuel oil

JP-5 fuel
DF-2 fuel
Navy distillate
Detergent lube oil
Non-detergent lube
Regul ar gasol ine
Premi urn gasol ine
Duration
Temp. of
(°C) Experiment
(days)
23 0.5

36
65
25



20-25 5



19-21 7

23 16
23 16
25 10
25 10


20 0.5

25 0.5




25 42



Oil Concen-
tration in
Seawater Reference
(ppm)
1453a

400
500
860b

6.2
6.1
5.9
8.3
6.5
1 .9
267b

9.7C
7.0C
2.4C
44. Oc
19.8
6.3
8.7
10.4
380

110
363
325
78
33
612
623
Boylan and Tripp
(1971)
Parker et al .
(1971)
Smith and
Maclntyre (1971 )
Frankenf el d
(1973)




Gordon et al .
(1973)
Lu and Polak
(1973)


Anderson et al .
(1974)


Boehm and Quinn
(1974)
Lysyj and Russell
(1974)





Aromatic species only,    Average value;
Distilled water
                           29

-------
An oil  slick is in a nonequilibrium condition with respect to
dissolution and evaporation; because the latter process dominates,
McAuliffe (1977) suggested that the highest concentration of oil
in water should occur within the first few hours of the initiation
of an oil spill.

     There has been little research on the kinetics of solutions
of oils into water.  For open  water and under static conditions,
Lyons and Rideal (1929) have proposed an equation for the rate of
dissolution of two immiscible  liquids when one liquid (solute) is
layered on the other (solvent).  They assumed that the rate of
dissolution is proportional  to the concentration gradient and
that the interface between the two liquids is saturated with the
solute.  The equation for the  rate of dissolution is

                                  C -C
where

     HP                                    "3
     j£- = rate of dissolution of oil, gm/cm *sec


      D = diffusion coefficient, cm/sec

     C  - oil concentration at saturation, gm/cm


      C = oil concentration, gm/cm

      6 = thickness of the interfacial  layer between the liquids,
          cm (0.2 to .06 cm)

     Lu and Polak (1973) studied the rates of dissolution of
three oils:  #2 fuel, medium bunker, and crude.  They correlated^
experimental data with the following equation:


                              - keat                         (30)


The integrated form of the equation is
where
                          Coo-Ct = kaeat                      (31)
       ..C^ = concentration and maximum concentration of oil  in
             the water per square meter of the interface,
             mg/1iter

         t = dissolution time, days

       k,a = constants

                               30

-------
The values of k and a determined by least-squares analysis are
presented below:


                                  k(mg/m2)
          #2 fuel  oil                1043        -0.423

          Crude oil                  8915        -2.380

          Bunker oil                  459        -0.503


These experiments  were conducted in closed systems and stirring
was provided for the  tank contents.  Since oil  spills occur
generally in open  water, the results of dissolution studies in
closed tank systems  cannot be considered representative of
natural  dissolution  rates for the oils tested.   Oil in open water
is not in an equilibrium condition as a result  of evaporation and
dissolution.  When dissolution studies are conducted in closed
systems, evaporation  of the volatile fractions  of oils is reduced,
These components are  the soluble portion of crude oils and petro-
leum products.   Thus,  the rate of dissolution is enhanced due to
the solubility of  these components of petroleum that would
ordinarily partition  between the liquid and gas phases.

     A process that  is similar to the dissolution of crude oils
is the rate of dissolving of monolayers.  This  subject is
discussed in many  textbooks on colloid chemistry.  According to
Adamson  (1967), the  rate of dissolution of monolayers may be
considered as a single diffusion process into an infinite medium.
Permeation rate (P)  is given by
where

     C = the bulk concentration,  mg/cm
                                      2
     D = the diffusion coefficient,  cm /sec

     t = time,  sec
                                      2
     r = modified concentration,  mg/cm

The bulk concentration is assumed to be established  immediately
below the film  in equilibrium with it.  By assuming  r  is  propor
tional to C, the equation can be  integrated to  yield
                                                             (33)
                               31

-------
where K is a proportionality constant.

     Harrison et al.  (1975) tried to quantify the relative rates
of evaporation and dissolution within a few hours of slick initia^
tion.  They assumed the water layer immediately in contact with
the oil  slick is in equilibrium with the oil.  Therefore,  the
dissolution rate can  be expressed by a  liquid phase dissolution
mass transfer coefficient multiplied by an interfacial  aqueous
concentration of the  oil.  The rate of  dissolution (R.)  was
defined as
                          Rd = KdS.X.Yi                       (34)
where
                                    p
     R. = rate of dissolution, mol/m -sec

     K. = mass transfer coefficient, m/sec

     S. = solubility of oil  in water, mol/m

     X. = mole fraction of oil

     Y.: = activity coefficient of oil


     A value of 5.5 x 10"  m/sec was used for the liquid phase
mass transfer coefficient.  This value had been suggested by Liss
and Slater (1974) as a typical liquid phase mass transfer coeffi-
cient on the basis of a study of gas exchange at the air-sea
interface.  The analytical method used by Harrison et al. (1975)
was not sensitive enough, while the sampling technique perturbed
the dissolution process and  the compositions of the oil  and water
phases.

     Neely et al. (1976) proposed a mathematical model to predict
concentration-time profiles  for spills of chloroform onto a river
They considered the river as a series of continuously-stirred
flow reactors and assumed that chloroform holdup occurs  in the
first reactor.  Three distinct time phases were considered to
account for the rate of infusion of chloroform during the
accident.  By writing material balance equations for each phase
and for the flow of contaminant through an arbitrary compartment,
a series of differential equations were derived:

                       dC

                    Vn

where
                               32

-------
     C  = concentration of oil, mg/cm

     Vn = volume of a compartment, cm
                                          3
      q = volumetric flow of the river, cm /sec

     k  = rate constant for evaporation, cm/sec
                                           2
      A = surface area of a compartment, cm

      n = the nth compartment (n > 2)

The parameters in the equations were adjusted during numerical
integration to fit the equations to experimental  data.   They
claimed the model was adequate for generating concentration-time
profiles corresponding to data from an actual spill.  This model
cannot be used to predict the rates of dissolution of oil  slicks
in water because chloroform is heavier than water and does not
form a surface slick.  Similarly, the model proposed by
Thibodeaux (1977) for spills of chemicals heavier than  water
cannot be applied to determining the rates of dissolution  of oils
that float on the water surface.

     Rates of dissolution of oils are generally slow and should
permit accurate determination.  Appropriate sampling procedures
must be adopted to avoid perturbing the surface slick and  the
underlying aqueous phase.  The concentration of oil  in  water
samples must be measured accurately, also.  Duration of the equi-
libration period is of utmost importance.  Several factors that
influence the solubility of hydrocarbons in water are also impor-
tant for oils.  For example, the solubility of any crude oil  or
petroleum product will be enhanced by the presence of surface-
active compounds in the oil or water (Boehm and Quinn,  1973).
Emulsification and solubi1ization by micelle formation  (Elworthy^
et al. , 1968) may become important when seawater  contains
sufficient quantities of dissolved organic matter.  Dissolution
rates are valuable in the interpretation of laboratory  bioassay
experiments.   They are necessary to establish whether organisms
in water in close proximity to a spill  will be exposed  to  toxic
levels of oil species.

     Literature on the spreading and dissolution  of oil spills
has been reviewed in this section.  Other weathering processes
act on oil and cause it to dissipate or disperse  from the  moment
it is spilled, but this review is limited to spreading  and disso-
lution.  To conclude this section it is necessary to comment
briefly on a few mathematical  models that have been proposed to
predict the behavior of oil spills on water.

     A mathematical model of the behavior of an oil  spill  on
water was presented by Mackay and Leinonen (1977).  The model
uses a simulated oil composition.  Thus, the physical  and  chemical

                               33

-------
characteristics of the components in the model  "oil" are known.
Oil  weathering processes quantified in the model  are evaporation,
spreading, dissolution, horizontal  and vertical  diffusion, and
natural  and chemical  dispersion.  Oil  composition, sea state,
wind speed, temperature, etc., can  be varied in  the model.  Infor-
mation generated by the model  includes size and  thickness of oil
slick, "oil" properties and composition and the  amounts of "oil"
evaporated, dissolved and disoersed as a function of time after
the spill.  The model uses Blokker's equation (Blokker, 1964) for
the rate of spreading, while the rate of dissolution of each
component in the oil  is given  by

                         N = k(eXCD-Cw)                      (36)


where

     k = dissolution  mass transfer  coefficient,  cm/sec

     e = solubility enhancement factor

     X = mole fraction of oil  component

    C  = solubility in water of pure component,  mol/cm
                                               •3
    Cl( = bulk water phase concentration, mol/cm
     w

The predictive ability of this model in actual  spills will be
limited because the derivation of the model assumes simple mix-
tures of hydrocarbons.

     Another report (API, 1977c) presents a computer simulation
model for the mass balance of  oil spills.  Rate  equations
available in the literature were used to quantify several dis-
persal processes in the model.  Some dispersal  processes were not
included in the model because  of lack of knowledge.  Despite the
complexity of this model, it cannot predict accurately the
behavior of oil spills on water.  Stolzenbach et  al . (1977)
reviewed a few models and found the predictive  capability of
existing models to be limited.

     The fate of oil  spilled on water is extremely complex to
model because of the  numerous  processes acting  to dissipate and/
or disperse the oil.   The marine environment is  a complex system,
also.  Assuming that  the individual  rates of all  weathering
processes can be quantified, combination of the  individual rates
may not adequately describe the fate of spilled  oil because of
the complex interactions between the mechanisms.   The behavior of
crude oils and petroleum products on water will  be different than
simple mixtures of several  pure hydrocarbons.  Until accurate
data from studies of  individual  oil  weathering  processes become
available, mathematical models of the fate of oil spills should


                                34

-------
       be viewed with caution.  As more research is directed towards
**""      quantifying the rates of individual dispersive processes, the gap
       in knowledge will disappear.
                                       35

-------
                            SECTION 6


                       CHEMICAL DISPERSION
     The dispersal  of oil  by treating slicks with chemical  disper-
sants cannot be understood fully without discussing briefly other
relevant aspects of this subject.


OIL/WATER SYSTEMS

     When oil  is mixed with water, an appreciable degree of
mixing does not necessarily occur.  The oil  and water portions
usually separate into distinct phases when the mixture is allowed
to remain quiescent for a  few minutes.  Almost complete phase
separation can occur depending on the type of oil, the volume
ratio of oil-to-water, and the intensity of agitation.  Complete
separation of oil  and water phases will not occur when the  system
is mixed vigorously.  Because crude oils and petroleum products
are slightly soluble in water, a small portion of the oil will
dissolve in the water phase.  Also, colloidal dispersions and oil
droplets (>1 ym) will form.  The formation of colloids and  drop-
lets by oils when they are mixed with water is an important
mechanism for the dispersion of oil slicks.
                                                                ht

EMULSION FORMATION

     Although emu! si f ication is treated in many textbooks on
colloid and interfacial chemistry (e.g. Davies and Rideal,  1963),
it is poorly understood.  Only those aspects of emu! sif ication
important to the understanding of the dispersal of crude oils and
petroleum products  will be discussed in this section.

     Crude oils and petroleum products can interact with water to
form two types of emulsions:  water-in-oil (w/o)  and  oil-in-water
(o/w).  A majority  of crude oils have a natural tendency to form
w/o emulsions (Berridge et al . , 1968a,b).   Stability  and droplet
size distributions  are the most important  characteristics of
emulsions.   Emulsions of oil-in-water are  generally not stable,
except when surface-active compounds are added to the system.  On
the other hand, w/o emulsions can be stable  for long  periods* of
time.  The formation of stable w/o emulsions have been linked to

                               36

-------
the presence of indigeneous surface-active components in crude
oils.   Berridge et al .  (19685) suggested that asphaltenes and
metal  complexes in crude oils are responsible for the stability
of w/o emulsions.   Canevari (1969a) identified porphyrin com-
pounds as promoting the formation and stability of w/o emulsions,
also.   Since these compounds (metal complexes, surface-active
agents and porphyrins)  are not present in most distillate
products, these refined petroleum products do not form stable w/o
emulsions.  Emulsions  which are stable, such as "chocolate mouse"
can form tar lumos due  to the action of weathering processes.

     Turbulent flow conditions favor small droplet size distribu-
tions.  As the oil is  broken into small droplets, they are dis-
persed by dilution into the larger volume of seawater.  The
smaller the droplets,  the faster they are disoersed by water
movements.  So emu!sification can cause the concentration of oil
in the water below surface slicks to exceed tolerable levels in
localized areas, and cause lethal effects to aquatic biota.
Water-in-oil emulsions  do not mix readily with seawater.  When
droplets are large (>75 ym) , they are no longer neutrally buoyant
In this case, the droplets may floculate and/or coalesce and rise
to the surface.  Oil that has coalesced may merge with the parent
slick or form a separate pool of oil at the water surface.  A few
research studies that  have increased the knowledge of emulsifica-
tion are reviewed below.

     The breakup of buoyant droplets in a turbulent fluid was
studied by Kolmogoroy  (1949).  The maximum stable drop size was
related to flow conditions and two dimension!ess parameters, the
Weber and Reynolds Numbers.

     Hu and Kintner (1955) experimentally investigated the
splitting of oil droplets due to the density difference between
oil and water phases.   The following equation was used to corre-
late their experimental data:


                  dcrit = K452 x 10"2(aow/Ap)               (37)

where

     d   -t = critical  size of oil droplet, cm

        AP = (PW-P0)/PW

     p  ,p  = density of oil and water, gm/cm , respectively
      0  W
       a ., = interfacial tension between oil and water, dynes/cm
        0 W

     Hinze  (1955) derived an empirical correlation for turbulent
emu!sification, in a couette apparatus with low oil concentra-
tions

                               37

-------
d95 - °-725^}
                                2       K     -2/5
                              wow
                    PCT

                                      w ow

where
                                                             (38)
     dgr = diameter of the oil  droolets such that 95% of the
           emulsions are made up of droplets smaller than this
           value, cm

      pl( = density of water, gm/cm
       W

      y  = viscosity of water,  gm/cm»sec

     a   = interfacial tension  between oil  and water, dynes/cm
      0 W

       e = energy input per unit mass and time, ergs/gm-sec

     The size distribution of oil  particles in seawater that
result from the emu!sification  of oil slicks has been measured
under laboratory and field conditions.  Gordon et al. (1973) used
a Coulter Counter to determine  oil particle sizes in a hydro-
carbon/water system mixed under laboratory  conditions.  The
greatest number of particles had sizes ranging from 1 to 30 ym.

     Witmer and Gollan (1973) used optical  techniques to measure
the size distribution of oil droplets in discharged tanker
ballast waters.  The most frequent droplet  size was about 10 ym
and the smallest particle observed was 3 ym.

     Lien and Phillips (1974) used a Coulter Counter to measure
droplet sizes in an emulsion.  The most abundant particle size
was found to be slightly larger than the minimum size detectable^
by the instrument.

     Forrester (1971) counted and measured  oil  particles in the
water column after an oil spill  from a tanker.   The sizes varied
from 5 to 85 ym.  The number of particles with smaller sizes was
greater than those with larger  sizes.  Forrester oroposed a model
to predict the probability of oil  drops breaking up in a turbu-
lent field.  The model predicts a high probability for large
droplets and low probability for small droplets.

     There are no standard emulsion preparation methods, and
different methods give different droplet size distributions.  The
state-of-the-art for measuring  size distributions of emulsions is
the Coulter Counter.  This instrument is expensive and cannot be
afforded by many laboratories.   A major disadvantage of this
equipment is the coalescence of emulsion during measurement.
Thus,the final size distribution may be different from that'of
the original oil/water mixture.   It is not  surprising that few

                               38

-------
reported measurements of the size distribution of emulsions show
different results.

     Emulsions can  be carried to subsurface depths after they are
formed.  When conditions are calm,  the motion of small  drop-
lets (< 1 ym) is expected to be Brownian.  Larger oil droplets
will rise and return to the surface slick.  The rise velocity of
droplets can be  predicted by a modified form of Stokes1 Law
proposed by Bond (1927)

                             kgd2(p -p )
                        Vr -    IS:                         (39)
                         k = 2/3 +y /y                       (40>

where

     V  = rise velocity of a droplet, cm/sec

      d = diameter of oil  droplet, cm

  p ,p  = density of oil  and water, gm/cm ,  respectively
   0  W

  y ,y  = viscosity of oil and water, gm/cm*sec, respectively
   u  w
                                  2
      g = gravity constant, cm/sec

     Under turbulent conditions, large oil  droplets may be broken
into smaller droplets and  transported away from the center of
mass of the surface slick  by sea currents.   Forrester (1971)
found oil particles distributed as far as 250 Km from the source.
of an oil spill.

     Fallah and Stark (1976c) developed a model to describe the
diffusion of oil  droplets  in the underlying  water using a random
walk approach.  The model  was applied to the problem of disper-
sive oil  losses from oil  booms and a hypothetical slick was used
to illustrate the model.   Similarly, Leibovich (1975) developed
a model to describe the turbulent dispersion of oil droplets into
the water column.

     Emu!sification is an  important process  for the dispersal of
oil spills.  The size distribution of oil droplets is important,
since the solution of hydrocarbon by molecular diffusion occurs
across droplet surfaces.   Emulsification is  an effective disper-
sal mechanism for oil spills for several reasons:
                               39

-------
     1.  oil  droplets can be transported faster than bulk oil;

     2.  dissolution from droplets occurs at an accelerated pace
         in water masses;

     3.  attachment by adsorption onto suspended particles can
         cause sedimentation of the oil; and

     4.  oil  droplets are small enough to be assimilated and
         gradually decomposed by marine organisms and bacteria.


STRUCTURE OF DISPERSANTS

     Chemical  dispersants are blends of surface-active compounds
(or surfactants) in suitable solvent bases.   The variety of
chemical  compounds that possess surface-active properties are so
numerous  that  several names have been used to describe disper-
sants (e.g. concentrated soap solutions, detergents, solubilizers,
emulsifiers,  etc.).  In general, surfactants are linear organic
molecules containing both polar and nonpolar groups.  These groups
impart to the  molecules their amphiphilic properties.  Thus, the
presence  of water- and oil-soluble groups at the ends of a
generally linear molecule is an important characteristic of
surface-active compounds.

     Poliakoff (1969) reviewed the state-of-the-art for chemicals
used for  dispersing spilled oil.  He found that a considerable
number of chemical compounds have been developed for emulsifying
oil.  A typical  dispersant consists of a blend of surface-active
compounds in aqueous, alcoholic or hydrocarbon solvent base.
Usually,  dispersant formulations include additives (e.g.
dispersant aids, water softeners, corrosion  inhibitors, etc.) to
enhance efficiency.  Dispersants can be divided into two groups:
ionic and non-ionic.  The former can be divided into two sub-
groups:  an ionic and cat ionic.

     A surfactant molecule is often represented as a tadpole-like
structure in which the "head" is water-loving (hydrophilic) while
the "tail" is  oil-loving (hydrophobic).  Figure 4 shows some
typical surfactant types:  anionic, non-ionic and cationic.
Anionic surfactants have structures that most closely resemble
soaps (Nelson-Smith, 1972).

     The  chemical structure of a dispersant, particularly the
ratio of  oil-  and water-soluble groups, is an important property.
Dispersants can  be classified according to the hydrophilic/
lipophilic balance, also (Davies and Rideal, 1963).

     There have  been interesting improvements in the technology
of dispersants,  particularly concerning their toxicity and self-
mixing properties.  Dispersant technology is so advanced that

                               40

-------
 SOAP      CH3.CH2.CH2...      ...CH2.COONa
 ANIONIC  <
                            ...CH2.O.SO3Na
          CH3.CH2»*   ...CH2
                                   SO3Na
 NON-IONIC   CH3.CH2...(CH2O.CH2)nO.CH2.CH2OH
 CATIONIC   CH3.CH2.CH2...
 CH3
 IXCH3
.Nv
 I SCH3
 Br
Figure 4.  Some Typical  Surface-Active Agents
           with the  Lipophilic "Tail" to the
           Left and  the  Hydrophilic "Head" to
           the Right  (Nelson-Smith, 1972)
                    41

-------
commercial  formulations may not be composed entirely of disper-
sants with  the same chemical  structure.   Great variations are
possible in the solvent base  used to dissolve the surfactant
molecules,  and the chemical structure and characteristics of the
dispersant.

     Surface-active compounds have been  found in crude oils
(Siefert, 1969; Siefert and HowellI,  1969).   The concentration of
surfactants in petroleum may  be significant depending on the
source and  extent of refining of the crude.  Corrosion inhibitors,
de-icing agents and combustion promoters are blended with some
petroleum products, thereby replenishing indigenous surfactants
removed during refining.  Seawater also  contains organic compounds
of both natural and anthropogenic origin that are surface-active.
Sieburth et al. (1976) have shown that these materials are
abundant in the surface microlayer of the sea.


COLLOID CHEMISTRY ASPECTS OF  CHEMICAL DISPERSION

     With turbulent mixing such as might be provided by water
movements due to wind, wave,  tidal,  and  current forces at sea,
oil  slicks  can break up into  oil droplets and disperse into the
underlying  water.  Also, the  dispersion  of one liquid into
another can be enhanced by modifying the interfacial tension
between the two immiscible liquids.

     A reduction of the interfacial  tension between two immiscible
liquids can cause emulsions to form.  This reduction can be caused
by a)  adding dispersants to  the oil/water system,  b)  pre-mixing
dispersant  and oil before introducing the oil into  water, and
c)  surface-active compounds  that may be present in the oil or
water phase before the liquids are mixed.  Surface-active
compounds present in seawater (e.g.  dissolved organic matter),
crude oils  and petroleum products enhance the emulsion-forming
tendency of oil slicks.  The  type of emulsion dispersants form
depends somewhat on whether the surfactants are preferentially
soluble in  the oil or water phase.   Water-soluble surfactants
form o/w emulsions, while oil-soluble surfactants form w/o
emulsions in accordance with  an empirical relation  proposed by
Bancroft (1915).

     A surface-active compound forms aggregates called micelles
in aqueous  solution when the  concentration of the surfactant
reaches a critical value called the  critical  micelle concentra-
tion.  The  formation and properties  of micelles have been
discussed by Elworthy et al.  (1968)  and  Tanford (1973).  Figure
5 is a representation of a spherical micelle.  Micelles have two
distinct regions of dissimilar properties.   In aqueous solutions,
the central micro-region will dissolve hydrocarbon  molecules,
while the outer region is soluble in water.  Thus micelles can


                               42

-------
incorporate hydrocarbon molecules into the micro-region and swell
(see Figure 5).   In hydrocarbon solutions, the structure of
micelles is reversed, that is the hydrophilic portions of the
micelles are shielded from the hydrocarbon solvent.

     Colloidal  particles containing amphiphilic and  hydrocarbon
molecules are called mixed micelles (Shaw, 1970).   This phenome-
non is well known as the chemical basis of detergency.  The
interaction of oil/water/dispersant molecules leads  to a micellar
type of accommodation.   The  resulting clusters of  mixed micelles
(oil droplets enclosed  in thin sheets of surfactant  molecules)
are stable because the  size  distribution is in the colloidal
range.  Additional  stability of the droplets is created by the
presence of a film of surfactant molecules which protects the oil
droplets and prevents coalescence by repelling droplets from  one
another.
              Figure 5.   Schematic  Representation  of
                         a Spherical  Micelle (Shaw,  1970)
                               43

-------
VIABILITY OF CHEMICAL DISPERSION AS AN OIL SPILL CLEAN-UP METHOD

     Decisions associated with public safety and ecosystem
protection must be made after oil  spills occur.  These decisions
are aimed at mitigating environmental damage to fish populations,
spawning and nesting areas, shorelines, beach property, etc.  The
major thrust of any oil spill clean-up and control  activity
should be recovery of the oil.  Therefore, diversion, physical
containment and total recovery of  spilled oil by mechanical
devices are methods favored by individuals and groups concerned
with preserving the environment.

     Mechanical skimming devices require that oil  be concentrated
at the water surface for recovery  to be efficient.   Anchoring
booms at sea poses serious problems.  The length of booming
equipment necessary to confine a major oil spillage presents
complicated handling problems at sea.  The efficient performance
of booms and skimming devices is limited to sea states with waves
of less than 2 feet in height and  current speeds of less than 1
knot.  Furthermore, logistic considerations make deployment of
these devices difficult in remote  areas that cannot be reached by
over-land means.

     Limitations  on the use of containment equipment and mechani-
cal skimmers are  now widely recognized and other counter-measures
have been proposed.  Opinion is growing that there  is a need for
other spill control options, until  there are improvements in the
state-of-the-art  for containment and removal of oil spills
(Canevari, 1969b; McCarthy, 1977;  McCarthy et al.,  1978).
Chemical dispersion of oil spills  is emerging as an alternative
counter-measure for large oil spills, and the proper use of
dispersants for controlling oil spills has become  a subject of
increasing world-wide interest.  The option to use  dispersants  to
treat oil spills  is attractive for  several reasons.

     There are circumstances when  the use of booms  and skimming
devices may be hampered or impractical, such as with oil  spills
under ice, piers, jetties and docks.  These areas  are not easily
accessible for booming, and it is  impossible to use conventional
recovery techniques.

     In other situations, when there is a threat to life, due to
fire hazard, or when beach properties and commercial  fishing
grounds are in danger of being impacted, dispersants may be the
only means of protecting valuable  areas.

     Skimming equipment cannot completely recover  oil slicks even
when containment  devices are used.   It may be desirable to use
chemicals to disperse the remaining oil.

     The logistics of aerial application of dispersants in remote
areas, not accessible by land, appear favorable (Ross et al ., 1978)

                               44

-------
     Crude oils form w/o emulsions naturally.   A w/o emulsion is
not desirable because it coalesces readily and it is tougher than
an o/w emulsion to disperse.   Addition of surfactants to oil
slicks enhances formation of o/w emulsions that are easily
dispersable.

     Finally, the volume of the oceans is tremendous and the
capacity to assimilate oil  droplets is infinite.  These reasons
support chemical  dispersion as a candidate for controlling oil
spills.

     Damage to a  coastline  and marine biota from a large oil
spill  can be  devastating.  The situation can be made worse if the
oil spill is  localized.   Mitigating the environmental damage due
to oil spills is  the objective of any oil spill clean-up opera-
tion.   This objective can be met by careful and by controlled use
of dispersants in coastal and offshore waters  and in well  flushed
estuaries.  The USEPA is aware of the need to  use dispersants,
under certain conditions, to protect vulnerable shoreline.

     Previously,  the use of dispersants posed  two major problems.
First, the toxicity of the  first-generation dispersants was
greater than  that of the oils to be dispersed.  The disastrous
consequences  to aquatic  life due to toxicity of dispersant appli-
cation during the Torrey Canyon spill are well documented (Smith,
1968).  Since, a  drastic reduction in toxicity has been achieved.
A wide variety of chemical  dispersants available commercially,
with different surface-active components and solvent bases, are
biodegradable and less toxic than their predecessors.

     Another  problem associated with the use of dispersants has
been the energy input required to break a floating oil  slick into
droplets.  The equation  governing droplet formation is:
                           Wk
where
      W.  = amount of mechanical  (mixing) energy or work input,
           ergs
                               p
       A = interfacial  area, cm

     a   = interfacial  tension between oil  and water,  dynes/cm
      0 W

     A reduction in interfacial  tension occurs when dispersants
are added to oil/water systems,  therefore the energy input
required for emu! sif ication is lower for oil /water/di spersant
systems than oil/water systems.   Actually, the amount of mechani
cal energy required to disperse oil  slicks  will  be greater than
that predicted by Equation (41).  The mixing  energy that is


                               45

-------
V..
required to disperse a large oil  spill  may not be readily availa-
ble.  The turbulence created by the propeller action of a
flotilla of ships has been used in some situations.   Some of the
newer dispersants have self-mixing capabilities (Caneyari, 1973,
1975).  The mechanism of action of these dispersants is similar
to the diffusion and stranding mechanism that is characteristic
of surfactants capable of spontaneous emulsification (Davies and
Rideal, 1963).  With self-mixing  dispersants, it is  not necessary
to provide mixing energy.  In this case the dispersion of oil
droplets can derive energy from the motions of the underlying
water column.

     The application of chemical  dispersants onto oil  slicks to
stimulate the formation and dispersion  of oil droplets is a key
alternative.  The method is not new; dispersants were  used to
disperse oil during a major oil spill incident, i.e. the Torrey
Canyon (Smith, 1968).  Oil droplet formation and emulsification
occur naturally, but to a limited extent because of  the presence
of surface active components in oil and in seawater.  Chemical
dispersion increases the natural  rate of dispersion  of spilled
oil, and the use of dispersants is a choice between  reliance on
natural processes to disperse the oil and acceleration of the
dispersive process by chemicals.   Dispersants must be  selected  as
the spill management alternative  with as much knowledge and
understanding of the mechanism of action of dispersants as
possible.


LABORATORY TESTS OF THE EFFECTIVENESS OF DISPERSANTS

     The literature on laboratory and field studies  of dispersants
is extensive.   References on the  subject can be found  in API
publications (1969, 1971, 1973, 1975, 1977a and 1979)  and a
special publication by the ASTM (McCarthy et al. , 1978).  The  bujk
of the published work has been on evaluation of the  toxicity of
dispersants on selected marine biota (API,  1973b; Renzoni , 1973;
and literature cited inter alia).  Tests in immediate  response  to
fortuitous spills are possible, but often expensive  and incon-
clusive.

     The effectiveness of dispersants varies.  Several  factors
influence effective oil dispersal:  oil  properties,  characteris-
tics of the water phase (salinity and concentration  of dissolved
organic matter), pH, ratio of oi1-to-dispersant, method of appli-
cation of the dispersant, spray drop size,  contact time between
oil and dispersant, temperature,  dispersant structure,  intensity
of agitation or mixing, and environmental factors (wind and wave
climates and tidal conditions).  In the absence of detailed field
tests, comparison tests can be performed in the laboratory under
controlled conditions.  Laboratory tests are necessary to assemble
prior information to support selection  of the best dispersant,


                               46

-------
specific to oil  type and particular spill  circumstances.

     Since decisions to use or not to use  dispersants in  the
field will be based in part on results of  laboratory tests, it is
important that laboratory tank systems and test conditions
simulate actual  field situations to the maximum possible  extent.
Because of the proliferation of methodologies, a realistic
approach to chemical dispersion has been slow to evolve and
claims by manufacturers have not been substantiated.

     Methods used to evaluate the effectiveness of dispersants
involve the application of the chemical dispersant on oil  float-
ing on water and agitation of the oil/water/dispersant mixture by
means of a mechanical device.  The absence of a universally
accepted laboratory system and standardized testing procedures
has led several  laboratories to develop tests to assess the
effectiveness and efficiency of chemical dispersants.  Fitzgerald
(1977) has listed some of the more common  procedures:  the
Swirling Beaker, the Warren Springs Separatory Funnel, the Mobil
Static Dispersion Test and the USEPA Simulated Environmental Tank
Test (SET).  Recently, Mackay et al. (1978) proposed a new
laboratory device for performing similar tests.  Because  of
inadequate correlation of testing procedures and variations in
designs of experimental apparatus, test results may not be
comparable and the performance of dispersants in the field may
not be consistent with laboratory results.

     In order to develop a standard laboratory system to  evaluate
the effectiveness of dispersants, several  factors must be
considered in the design of the laboratory system.  These  factors
have been discussed by Osamor and Ahlert (1981).  Also, this
paper critiqued  two available procedures:   the SET and the Mackay
Apparatus.

     The kinetics of dispersant action, as well as the details o"f
processes occurring at oil/water interfaces, are unknown.
Canevari (1969a,b, 1978) has described some dispersion mechanisms.
The basic function of dispersants is to break up oil slicks into
tiny droplets, to help to disperse or solubilize oil into  the
underlying water column.  A good dispersant produces an oil-in-
water emulsion that is stable and has no tendency to flocculate
or coalesce.

     The ultimate goals of laboratory experiments are assessment
of the efficiency of dispersants and an ability to extrapolate
test results to  predict performance during actual  field use.
Exact prediction of field performance is too complex, but  labora-
tory results will give preliminary indications of how a particular
crude oil or petroleum product will respond to treatment  with a
specific dispersant.  Difficulty lies mainly in the inability to
simulate in the  laboratory conditions that occur in the field.
Marine environments are very complex and conditions are changing

                               47

-------
constantly.   Experience in the use of dispersants in  the field is
growing.   It is important to draw from field experience for
proper design of laboratory systems to evaluate the effectiveness
of chemical  dispersants.
                               48

-------
                            SECTION 7

                          EXPERIMENTAL
MATERIALS

Crude Oils and Petroleum Products

     A total  of 12 crude oils and petroleum products were studied
The oils had  various origins and some of the oils had vague
histories.  Iranian crude oil was provided by the USEPA (Edison,
N.J.) for this investigation.  The fuel  oils, i.e.  #2 and #6, as
well  as an unidentified oil  consisting of 8% crude  and 92% oil
derivatives,  were supplied by Sun Oil (Marcus Hook, Penna.).   The
oil consisting of 8% crude oil  was designated as 8% crude for
identification purposes.  The remaining  eight crude oils were
obtained from Mason & Hanger-Silas Mason Co., Inc., OHMSETT
(Leonardo, N.J.).  These crude oils are  Nigerian light,
Lagunillas, La Rosa (Venezuela), North Slope (Alaska), Brass
River (Algeria),  Suniland, Sahara and Arzew (Algeria).

     The oil  samples are identified and  listed in Table 3.  Also,
Table 3 gives analytical test data on the oil samples, as
performed by  Mason & Hanger-Silas Mason.  Measurements of some
key properties (specific gravity, viscosity, surface and inter-
facial tensions)  were made at the time the oils were received.
The results of these measurements are reported in Table 4.  The
data  of Tables 3  and 4 show good agreement; the slight discrepan-
cies  could be due to the effects of aging during storage and  to
differences in analytical methodology.

     Sufficient quantities of the oils were procured so that  all
tests could be performed on samples from the same batch.  All of
the oils were stored in the original  containers with tight caps,
to prevent evaporation.  Samples were withdrawn as  quickly as
possible and  unused portions were discarded.

     As there are variations in the physical and chemical  proper-
ties  of the oils, it is assumed that  there are substantial
variations in compositional  characteristics because of the
different origins.  They represent a  large fraction of oils that
are transported on the oceans.   Although these oils have a range
of physical and chemical properties that bracket the properties

                               49

-------
                 TABLE 3

Analytical Data on Oil Samples as Performed
       by Mason & Hanger-Silas Mason

Oil
#2 Fuel

#6 Fuel

8% Crude

Iranian


Nigerian

Laguni 1 las

La Rosa Venezuela

North Slope,
Al aska
Brass River,
Al geria
S u n i 1 a n d


Sahara


Arzew, Algeria


a = 26 C
b = 83.3 C

Specific
Gravity
0.864

0.951

0.902

0.851


0.832

0.947

0.905

0.892

0.813

0.894


0.797


0.805




Visco
(m2/
x 10
4
6
2876
6
27
9
10

8
7
8
266
17
28
2
20
5
5
8
19

7
5

8
5

5
c =
d =
sity Surface
Tension
s) (N/m)
-6 x TO'3
.96a 32.0
.81b
.97a 38.0
.16b
.05a 34.0
.46b
.58a 29.2
h
.23
.Oc 26.0
.5d
.Oc 33.3
.od
.1° 31.2
.9d
.Oc 28.0
.7d
.8C 26.9
.2d
.Oc 29.0
d
.4
.8C 24.7
d
.2
.3C 26.2
H
.4°
100 F
210 F
Interfacial
Tension
(N/m)
x 10"3
28.8

16.2

16.0

28.4


24.2

21 .7

26.7

20.3 *

25.6

23.4


25.2


20.7




                     50

-------
       o

        o
        S-
        (U
        a.
        o
        s_
       Q.
        10
        U
CO
,
4J
I/I
0
*8
;» co
c
•^
^
CJ
in
co
>>
+j
>
IO
« £
0 ~
•r*
H-
eu
D.
"> 0
CM

•»—
O
10 LO
ID CM
CM CM
r-» i—
CO CO
CM CM
LO «J- LO 00 00
i— co to o r»»
r^ r^ r». co co
JD
cn O
CM O
• o
co i—
10
•a- o
IO O
• o
CO CM
O LO O
en LO LO
O) CO CTl
o o o
LO O O
cri (O in
Cn CO Cn
O O O
in CM
vo in
o CD en
r- 00
•o
Ol
1- 4->
•4J Q. •— •— •—
.
•
co
CM
in
cr>
CO
O
CM
0
o»
o
in
o
en
o
0)
T3
3
s-
C_)
»«
CO
r>>
r^.
CM

cn
CM
in
en
CM
co
CO
CO
CM
O
O
«9-
CO
o
LO
LO
CO
0
r-~
in
CO
o
Iranian
«»•
r^
co
«s-
CM
CM
en
CM
CM
co
«»
CM
CM
IO
VO
CM
CO
O
O
CO
CO
o
CM
CO
CO
o
Nigerian
cn
vo
CO
in
CM
o
^
co
CO
(O
o

CM
CM
CTl
O
co
03
*»•
CO
co
«9-
VO
in
10
CO
CO
«c
r^
cn
CO
O
CM
0
cn
o
in
O
o»
o
La Rosa, Venezuela
en
cn
CM
CM
o
CO
CO
CM
co
co
•
CM
CM
VO
r-..
CO
CM
CO
CO
O
o
cn
CO
o
o
en
CO
0

S_
0
•z.
*t
*t
^
CM
CM
^~
cn
CM
en
CM
CO
en
CM
CO
o
CO
o
CM
CO
0
in
CO
0
10
i.

<
!_
O)
>
•F—
at
in
in
10
s-
CQ
r~
10
CM
en
vo
CM
CO
CO
«o
U3
o
CM
10
CO
CM
O
en
CO
o
in
cn
CO
0
in
cn
CO
0
•o
c
ro
c
3
GO
CO
r^
CM
CM
CM
en
rs.
CM
CM
CO
.
0
Sahara
«•
cn
CM
!••.
en
CM
o
CO
CM
CM
CM
CM
CM
O
cn
r-.
o
in
en
r-
o
r^
en
p«.
0

                                                                                                                      eu
                                                            51

-------
of numerous crude oils and petroleum products, they may not
necessarily reflect oils having, historically, the highest
frequency of discharges into marine environments.

Chemical Dispersants

     The variety of commercially available dispersants has grown
rapidly with the success of research by the detergent industry.
Imaginative trade names for dispersants have been encountered in
the literature.   Several dispersants are claimed to be able to
disperse all oil slicks under different environmental conditions.

     It is impossible to test all dispersants.  Some dispersants
were selected from the list of dispersants "approved" by USEPA.
The list was compiled on the basis of commercial availability and
suitability of compositional data supplied by the manufacturers.
Other considerations by the USEPA, for inclusion of a dispersant
in the list, are stability during storage, non-toxic effects on
marine organisms, biodegradabi1ity of key constituents, and ease
of application.   The dispersants that were considered for testing
are shown in Table 5.  The characteristics of the dispersants
were summarized  from the manufacturers' literature.  Only five of
the dispersants  were tested; they are referenced in subsequent
discussions as Products A, B, C, D and E.  These dispersants are
classified below to reflect the  characteristics considered during
selection.

        Classification              Dispersant         Product

1.  Self-mix                       Corexit-9527           A

2.  Non-ionic
      Hydrocarbon-Solvent Base     BP 11OOX               B
      Aqueous-Solvent Base         Corexit 7664           C

3.  Ionic
      Hydrocarbon-Solvent Base     Congo                  D
      Aqueous-Solvent Base         Proform TM             E


     All of the  dispersants were received directly from the manu-
facturers.  Special  storage was  not required for any of the
dispersants.

Chemicals and Glassware

     All chemicals were used as  purchased from the manufacturers.

     Spectral grade Carbon T.etrachl ori de was purchased from
Fisher Scientific (Cat. #C-199)  and used to extract oil  from
water samples.
                               52

-------
                         CO
                         o
                         4J
                         CO
                         (U
                         4->
                         O
                         (O
                         S-
                         >a
                   LO

                   UJ
                   -J
                   CQ
                         co
                         s_
                         (U
                         Q.
                         CO
                         S-
                         (O
                         3
                        CO
'V,

 U.
O CO O
(J U VO

»r»
*"
S- • _
3 +•>•:?•
o a-ib
a.




.c
10 -t-> U.

Ll.


^
>r™ O)LU
~2 "-^* O

a *~


il»-^
o u.
•0
Q.VO
OO- — •
*•> II
1 %
pH ^3
O

°2


-u
c
ia
U)
s.
0)
Q.
Vt
O
Lu
r*. CM

en CM
u u
z a.
a. ca
o
CM



CO
1/1 iZ"

«^- ^?

CM £3.

r~

^_
CM




.C
 <4_
E -i-
ia u
S °"




i

O
O CM
re 3:



O i—
O CM
I ^•^^v^

O f—
in CM


i °O
1 in
i



CJ CJ
LO 00
CO £ S
. •— a. o.
i
O vo i~-
VO CM CO


^_
in
i •




in ^ ^
CM U.
CO 0

o *— •
v>
O  P^ r—
CO r— r— r—  vi

>
c
r™
o
_a

IO
oo
u
t/?
CO









1
s-

-------
     Nitro Fast Red B, an oil-soluble red dye was obtained from
Sandoz Colors and Chemicals (East Hanover, N.J.).

     Instant Ocean-Sea Salt (Aquarium Systems Co., Ohio) was
used, according to manufacturer's instructions, to prepare the
salt water solutions.

     Granular anhydrous Sodium Sulfate (Cat.  #S-421), separatory
funnels,  pipets, flasks, graduated cylinders, beakers, glass
containers and ofner glasswares were purchased from Fisher
Scientific.

     Trimethoxysilane  (3,3,3-Trif1uoropropyl) was obtained from
Petrarch  Systems (Levittown, Pa.), and used to render glasswares
and sampling tubes hydrophilic.


DATA AQUISITION DEVICES

Weight and Temperature

     All  weight determinations were made on a Mettler Type H6T
analytical balance.  The balance has a capacity of 160 grams and
was manufactured by Mettler Instrument Corp.  (Hightstown, N.J.,
Ser. #184453).  Measurement can be read to the fourth decimal
pi ace.

     Liquid temperatures were measured with mercury thermometers.
Thermometers were calibrated in degrees Centigrade and could be
read to a precision of ± 0.1 C.  Thermometers were purchased from
Fisher Scientific (Cat.  #15-166A).

Specific  Gravity

     Glass float hydrometers were used to measure specific gravi-
ty.  Two  Fisher Scientific hydrometers (Cat.  #11-5100 and
#11-520A) with different scale ranges were used.   The first
hydrometer was calibrated from 0.650 to 1.000 with 0.005
divisions, while the second had a range of 1.000  to 1.220 and
0.002 divisions.

Kinematic Viscosity

     Kinematic viscosity measurements were performed using a
Saybolt viscometer.  The viscometer was manufactured by Precision
Scientific (Chicago, 111., Ser. #14-V-2).  The device uses the
capillary-tube principle for measuring liquid viscosities.  The
liquid is allowed to drain from the bottom of a cylinder through
a short capillary tube.   The time required to drain 60 ml of the
liquid is recorded, in seconds using a stopwatch.   Time is taken
as the kinematic viscosity of the liquid.  The viscosity is in
Saybolt Universal  seconds and can be expressed in the units,

                               54

-------
 2
m /sec, by using the following equation


             v = ± = (0.00237ts - l^-)x9.29xlO~5            (42)


where t  = Saybolt Universal  seconds.  The accuracy of kinematic

viscosity measurements depends on the accuracy of the timing.
Reported viscosity values are averages of triplicate measurements.

Surface and Interfacial  Tension

     All surface and interfacial  tension measurements were made
with a Fisher Scientific Model 21 Surface Tensiomat (Cat.  #14-814,
Ser. #1910).   The manufacturer's  specifications for the platinum-
iridium ring  (Cat. #14-812-5, #1240) are:

                  Mean Circumference = 5.985 cm

                        R/2          = 53.6135350

The Tensiomat is a torsion balance which measures the force
required to pull a platinum-iridium ring free from a liquid
surface (surface tension) or across the interface of two immiscible
liquids (interfacial tension).  To measure surface tension, the
ring is submerged in the liquid and slowly raised out of the
medium.  As the ring is  pulled from the liquid, the force
necessary to  separate the ring from the liquid is measured and
read directly on a dial  to a precision of ± 0.05 dyne/cm.   For
interfacial tension measurements, the ring is first submerged in
the heavier liquid.  Then, the lighter liquid is layered on the
heavier liquid and the ring is pulled from the heavier to  the
lighter liquid.

     The device was used in both  manual and automatic mode; the
results from  both measurements were comparable.  Triplicate
measurements  were made for each liquid and the results averaged.
The manufacturer's manual gives details of all operational
procedures and necessary corrections to the dial  readings.   Prior
to use, the device was calibrated according to manufacturer's
instructions.  Usually,  the surface tension of laboratory  distill-
ed water is determined first and  checked for close agreement with
the known value prior to sample measurements.

Infrared Spectrophotometer

     To quantify the concentration of oil in water samples,
infrared spectra of the  samples were obtained with a Perkin-Elmer
Model 599 Infrared Spectrophotometer (Ser. #113277).  This equip-
ment utilizes the double beam technique and measures the differ-
ential absorbance/transmission between two samples, i.e. a


                               55

-------
reference "cell" and a sample "cell".  The conditions used in
obtaining the spectra of all  samples were wide spectral  slit,
scan time of 6, and chart expansion of 1.

     The spectra of CC1.-extracts of water samples were  measured

using matched pairs of IR grade quartz rectangular (1 cm path
length) or cylindrical (5 cm  and 10 cm path lengths)  cuvets
supplied by International Crystal Laboratories (Elizabeth, N.J.).


SETUP

     This study was conducted in three phases:  spreading, disso-
lution and chemical dispersion.   A different experimental  setup
was used in each phase.  The  setups for each phase are presented
separately.

Spreading Experiments

     The equipment used to conduct spreading experiments is
described below:

Test Tank--
     The test tank is a rectangular, transparent aquarium made
from 1/4-inch thick Plexiglass to facilitate visual  observations.
The dimensions of the tank are 5 feet long by 3 feet  wide by 2
feet high.  The tank was sealed against leakage with  methyl
chloride and a silicone rubber compound.   Plexiglass  bars were
used to reinforce the tank at five locations at the tank walls,
so it could sustain the outward hydrostatic pressure  of  about 200
gallons of water.  Tank contents can be drained through  an outlet
port connected via a garden hose to a centrifugal  pump.

Oil Feeding System--
     There are three methods  by which contaminant  oil could be
introduced into a receiving body of water:  1)  instantaneous,
2)  leakage at a constant rate, over a fixed time  interval, and
3)  a combination of both methods.  Since method 2)  is the most
common method for oily discharges, a flow system was  designed to
feed a steady quantity of oil to the center of the test  tank.

     The oil  feeding system consists of a variable speed Master-
flex peristaltic pump, with controller, manufactured  by  Cole-
Parmer Instrument Co. (Chicago, 111.).  The pump,  which  is capable
of delivering 167 mls/min with no head, is used to deliver oils
from a holding tank.  Oil flow from the pump is directed to a
specially designed feeder made out of 1/8-inch diameter  stainless
steel tubing.  The tubing is  positioned such that  oil is
discharged from one end of the tube located approximately 4
inches below the water surface.  Thus, the oil feeding system was
used to introduce the oil to  be tested, beneath the  water surface,
at a constant rate over a fixed time.

                               56

-------
Photographic Documentation of the Area!  Extent of Spreading Oils--
       The measurement of the area!  extent of spreading oils is
basic and essential  to studies of oil  transport.  Conventional
sequential still  photography was used  to document the area
covered by the spreading oils.  Black-and-white photographs were
obtained with a Canon F-l, 35 mm single-lens reflex camera
equipped with a motor drive unit.  A 28 mm f/2.8, wide-angle lens
was used.  The camera was equipped with an interval timer and
other peripherals to provide remote control  capability and to
sequence photographs at a selected interval.  The interval timer
can be preset so  that the camera photographs continuously (3.5
frames per second) or varies the exposure time from 0.5 seconds
to 3 minutes per  frame.  The camera was equipped with a film
chamber which enhances the film-holding capacity of the camera  to
250 exposures.  Ordinary and high contrast 35 mm black-and-white
films exposed through multiple filters (yellow and polarizing)
were used to document the spreading boundaries of the oils inves-
tigated.  The filters were used to decrease the reflectivity of
the oil films and to improve contrast.

     The camera was  mounted directly above the test tank by
attaching it to the  extension arm of a Technol TC-1 copy stand.
Even with this arrangement, it was not possible to locate and
delineate accurately the area! extent  of #2 fuel oil.  In this
case, the oil was dyed with an oil-soluble red dye, i.e. Nitro
Fast Red B,   When mixed with oils, this dye does not modify the
surface tensions  of  the oils (Hollinger and Mennella, 1973).  The
quantity of the dye  used was very small, but it enabled the
boundaries of the oil to be identified.   Surface-tension measure-
ments of the oil  and dye mixture confirmed the findings of
Hollinger and Mennella (1973).

     Estes and Golomb (1970) have claimed that oil  on the sea
surface does not  image well in the spectral  bands recorded by
conventional black-and-white or color  photography.   Photographic^
imagery is always affected by the contrast between the object and
background.   To improve contrast between the spreading oils and
background,  the bottom of the tank was painted white.  Further-
more, several flood  lights were positioned at various locations
such that contrast was enhanced by the angle of the light.  At
sea, such an arrangement would be difficult, if not impossible.

     A schematic  diagram of the experimental setup used to study
the spreading rates  of oils, is shown  in Figure 6.
                               57

-------
                                             OJ
                                             CT
                                             C     CU
                                            •r-     C
                                            •O    1-
                                             CU    i—
                                             O)
                                            U_  C  S-
                                             i  •,-  oj
                                            i—  re -M
                                            i-  i-  re
                                            in vo
                                                           cu
                                                          •I—
                                                          •o






-r
1
1


I





I
II

i \ •
\\






I




1


\







i


r

i
l
i






                                         to
                                         i.
                                         
                                         E -M--
                                         tO CO r—
                                         s-    -o
                                         CT>, O
                                         o  a. o
                                         s-  o •—
                                        Q- O U-
                                        r— CXJ CO
                                                           cu
                                                           i.
                                                           a.
                                                          co
                                                           o
                                                          M-

                                                           o.
                                                           0)
                                                          co
                                                          re
                                                          
.c
 o
CO
VO
 eu
 s_
 C7>
58

-------
Dissolution Experiments

     Figure 7 is a schematic diagram of the apparatus used to
study the rates of dissolution of oils.  It consists of a cylin-
drical  glass tank 15 inches in internal diameter and 15 inches
high.

     A copper coil immersed in the tank facilitates temperature
control  of the water in the tank.  A constant temperature thermo-
static bath is used to circulate cooling water through the coil
to maintain the contents of the tank at a nearly constant temper-
ature of 25 ± 0.2 C.

     Water samples are withdrawn from the tank by gravity flow
through  a glass tube of 1/4 inch internal diameter connected to a
short piece of tygon tubing.  The glass tube is located along the
side of  the tank such that the inlet is positioned at the center
of the tank.  Prior to each experiment, the sampling tube is
treated  to make the surfaces hydrophilic to eliminate adsorption
of oil  onto the tube wall.  Sample withdrawals occurred with the
tube positioned 6 inches from the water surface.

Chemical Dispersion Studies

     The apparatus used for chemical dispersion studies is shown
in Figure 8.  The major components in the apparatus are the test
tank, the device for applying dispersants, and the equipment for
mixing the tank contents.  The apparatus is discussed below.  The
discussion on each component also focuses on some factors that
were considered during the design of the apparatus.  Two setups
that have gained some recognition in chemical dispersion studies
are critiqued.

Test Tank--
     The size of the test tank is important because it limits the
volumes  of water, oil and dispersant that can be used.  The tank
should be large enough to minimize wall surface interactions and
the vertical dimension must be related to the rate of slick
breakup  and droplet migration.  In other words, the experiment
loses most of its value when droplets commence to impact the
bottom of the tank and the concentration gradient reverses.  If
the tank size is too small, the quantity of water becomes another
variable.  In this case, the water/oil/dispersant ratio becomes
important and the relevance of results to field situations is
questionable because the quantity of water in the field is
essentially infinite.  The larger the tank size, the wider the
variations in the quantities of oil  and dispersants that can be
used and the larger the volume of test water that can be main-
tained.

     The configuration of the tank is important because of  •
influences on the wave form.  A cylindrical tank provides

                               59

-------
cooling
 water
                                     Sampling
                                      tube
Figure 7.  Schematic  Diagram of Experimental
           Setup  for  Dissolution Studies
                       60

-------
 O  co
4-> T-
 00
X
    U
 U T-
•i-  S.
 CO
 O)
J3
 O
 S-
D_
                                       a>       •*-
                                (j 0) "O H Q. Q.
                                0) O"O C E  E
                               i— O  
                                   oo  o r— CM
                                                           (d
                                                           u
                                                          •r-
                                                           E
                                                           o>
                                                          J=
                                                          o

                                                           s_
                                                           o
                            a.
                            3
                            +->
                            QJ
                                                           s.
                                                            3
                                                           (O *->
                                                           u  o
                                                           •r- •!"-
                                                           +J  V)
                                                           its  S-
                                                           E  
-------
symmetry and limits variations to two spatial  coordinates,
however, optical  measurements are difficult.   Thus, the choice of
a tank configuration must be predicated on measurement system
selection and the device chosen to impart mixing energy.

     Since hydrocarbon concentrations can be  evaluated as
functions of space and time, sample volume and sampling frequency
are important factors determining the size of the tank.  Sampling
with time reduces the liquid volume in the tank but must  not
influence significantly the concentration gradients in the hori-
zontal and vertical dimensions.  The volume of the laboratory
tank is critical  for test results to be meaningful and relevant
to those in the field.

     Several tank systems have been used in previous investiga-
tions of the effectiveness of dispersants.  For example,  Murphy
and McCarthy (1969) and McCarthy et al. (1973) have described the
Simulated Environmental Tank (SET) system.  It is a cylindrical
tank, 24 inches in diameter and 28 inches high.  The Mackay
apparatus consists of a cylindrical glass tank 29 cm in diameter
and 29 cm high (Mackay et al., 1978).  Wall surface interactions
will be significant due to the small volumes  of these tanks.

     The tank used for dispersion studies is  the same as  that
used for spreading experiments.  This tank has been described
previously, but it was modified slightly by inclusion of  3/8 inch
Plexiglass tubes  at different locations in the tank.  Sampling is
accomplished by gravity flow through tubes positioned at  15
locations in the  tank.

     The arrangement of the sampling tubes in the tank is as
follows:  two adjacent vertical sides have three sampling loca-
tions each at various heights (2, 8, 14 inches) above the bottom
of the tank, while a third side has the remaining tubes.   The
fourth side is used for visual  observation.  Sampling tubes are
located such that water samples can be withdrawn at three depths
and 6 inches from the walls of the tank.  Three additional  samples
were generally withdrawn at three different depths but at the
center of the tank.

Dispersant Application--
     The effective use of dispersants requires not only an effi-
cient material  but an efficient application technique.  Improper
and inefficient application techniques result in unsatisfactory
performance of the dispersant.   For dispersants to perform at
maximum efficiency, they must be applied with proper equipment.
The proper application method is to spread the chemical evenly on
the oil slick.

     In the field, the most common methods of applying dispersants
onto oil slicks are hand-held spraying equipment, fire-hoses and
spray booms mounted on work boats or vessels  of opportunity, e.g.


                               62

-------
helicopters and planes.  Use of hand-held sprayers is limited to
small  spills.   For treatment of moderate and large oil  spills,
fire-hose systems and spray booms are necessary because of the
large  oil slick areas to be treated.   Aerial spraying is attrac-
tive because of the possibility of application in remote areas,
high dispersant dosage rates and fast response.  In general,
spraying equipment is designed to spray the dispersant  neat
(particularly for hydrocarbon-solvent-based dispersants) or to
spray  a dilute dispersant solution by using an eductor  that feeds
the chemical into a stream of seawater.  Several  spraying systems,
including spray booms, are available  commercially.  Spray booms
feature different nozzle arrangements; spraying height, spray
size,  swath area, etc. can be controlled.

     At the laboratory scale, several methods have been used to
introduce dispersants to treat oil:

     a)  mixing chemical with water  prior to introducing the oil
         to be dispersed;
     b)  pouring dispersant from a container;
     c)  syringe injection or pipetting of dispersant into the
         middle of the patch of oil  on water;
     d)  mixing dispersant with oil  before pouring the  mixture
         onto water;  and,
     e)  spraying with hand-held spray cans.

Some of these methods, such as a) and d) are not  practical in the
field.   These two methods will also  provide maximum contact
between the dispersant and oil and cause efficiencies in labora-
tory tests to be higher than those in the field.

     The method for dispersant application in laboratory studies
must reflect techniques used in the  field.  The method  must be
reproducible and characterizable.  Therefore, use of hand-held
spray  cans or an atomizer/nebulizer  is suggested.  Oda  (1969)
described a method for applying dispersants in a  fine spray on
oil slicks.  Application of dispersant was achieved by  means of  a
spraying device fitted on an aerosol  bomb containing a  pressurized
propellant.  Spray cans equipped with triggers may be suitable
for applying dispersants.  The difficulty with using hand-held
sprayers in comparative tests is the  variability  in hand motions
and the applied pressure on the spray trigger during dispersant
application.  The droplet size and swath of the spray are diffi-
cult to control, also.  Therefore, an atomizing system  is
preferred.  Regardless of the method  used, the dispersant must be
applied uniformly and directly to the floating oil, in  the form
of small droplets, and not as a fog  or mist.

     In SET and Mackay systems, a ring is used to contain the oil
slick  and the dispersant is poured into the ring.  In general,
dispersants are used  in field situations where oil slicks cannot
be contained by booms due to spreading forces. Thus, the tech-

                               63

-------
niques for applying dispersants in both SET and Mackay systems
are not valid.

     Chemical  dispersants were applied by an ad hoc atomizing
unit.   This  system for applying dispersants consists of an adjus-
table  atomizing nozzle manufactured by D. B. Smith & Co.  (Utica,
N.Y.,  Model  #147).  The nozzle has a general purpose setting that
controls the drop size of the spray and the swath width.   This
type of nozzle  produces a uniform flat v-shaped spray pattern
instead of a hollow cone and the setting varies the size  of the
nozzle orifice  such that "fine-to-coarse" sprays can be produced.
The nozzle is  connected to a short piece of metal  tubing  equipped
with two inlets.   Laboratory compressed air flows through Tygon
tubing to one  of  the inlets.  The air flow rate is monitored by a
flow meter (Brooks Instrument, Hartfield, Pa.,  Model #1555-
04C1AZZ).  Dispersant solution is delivered to  the second inlet
by a variable-speed Masterflex peristaltic pump (Cole-Parmer,
Chicago, 111.,  Ser. #51526).  The flow rate of  dispersant solu-
tion is regulated by means of a Masterflex controller connected
to the pump.

     To apply  a sample dispersant, compressed air and the disper-
sant solution  are forced through the nozzle.  By using different
combinations of flow rates of air and dispersant, the atomizing
system can produce different spray sizes and swaths.

     A carriage resembling a railroad car was mounted on  top of
the tank to  continuously reposition the atomizing nozzle.  The
car is positioned on a track consisting of two  aluminum rods.
The rods are connected to two laboratory stands so that the track
crosses the  tank  lengthwise.  The car is pulled by a fishing line
attached to  the rotating shaft of a low-speed gear motor  (Merkle-
Korff  Gear Co.).   This setup permits the nozzle to travel from
one end of the  tank to the other at a constant  speed of 0.075 ft/
sec to simulate the transport of a sprayer attached to a  boat or"
plane  during dispersant application in the field.

     The design of the dispersant application system allows
several methods of dispersant application to be evaluated.   For
example, the swath and droplet size of the dispersant spray can
be varied by adjusting the orifice of the nozzle and varying the
flow rates of  air and dispersant solution.  The height of the
nozzle above the  water level in the tank can be varied by changing
the height of  the track above the tank.  The effects of variations
in the impact  velocity of dispersant sprays on  oil slicks,  spray
angle, single  and multiple passes, and spraying time per  pass can
be investigated with this system.
                               64

-------
Application of Mixing Energy--
     Some form of mixing or agitation must be provided to
chemically treated oil  slicks for complete dispersion.  There is
a direct relationship between mixing energy and the performance
of chemical dispersants.  In the field, natural wave action may
provide the agitation required to disperse the treated oil.  But,
during calm conditions  at sea, mixing energy has to be provided
to disperse oil  slicks  even when dispersants have been applied.

     Smith and McCracken (1977) and Smith (1978) have described
the major methods of supplying mechanical energy to treated oil
during field conditions:  agitation by high pressure fire hoses,
specially constructed wooden breaker-boards in tow by vessels,
and turbulence produced by the propeller action of ship wakes.
These methods were investigated at the OHMSETT test facility.
The results of these investigations show that the efficiency of
the dispersion varies according to the technique.  The depth of
droplet migration, and  the rates of coalescence of oil droplets
and slick reformation will vary for different methods.  This
underscores the importance of investigating different methods of
supplying energy to disperse oil slicks.

     When oil spills occur in remote regions, dispersants can be
applied from the air but there are no means of providing mixing
energy.  In such cases, natural wave motions are relied upon to
provide effective dispersal of the oil.

     In laboratory tests, the contents of the test tank are
agitated by mechanical  devices, such as shakers, pumps, vortex
blowers, impellers, etc.  For example, the 0111/water/dispersant
mixture in the SET test is mixed by the shearing action of a
pump.  In the Mackay system, air is bubbled through the tank.
Mechanical devices can  create zones of dissimilar intensities of
mixing, which influence the local droplet size distribution and k
the depth of oil droplet penetration.  The stability of emulsions
is a function of mechanical energy input; unstable phases will
tend to remain dispersed in the presence of turbulence.

     The intensity of mixing provided by laboratory devices cannot
be compared with field  methods for dispersing chemically-treated
oil spills.  The energy of agitation per unit volume of liquid  in
laboratory tests is likely to be much greater than under actual
field conditions.  Shackleton et al . (1960) studied the emulsify-
ing characteristics of  several pumps used for deballasting opera-
tions and found that stable emulsions were formed as a result of
the shearing action of  these pumps.  It is questionable whether
the mixing caused by wave motion is as intense as that provided
by laboratory equipment.  In the oceans, wave motions vary in a
random manner, and variations occur in both space and time.
Forrester (1971) measured the sizes of oil globules following a
spill by the tanker Arrow and found that oil globules formed in
natural wave motions were relatively large in size.

                               65

-------
            The experimental  tank must be provided with means to impart
       mixing energy.   Two devices that are well  known and have been
       used extensively in wave studies are submersible ultrasonic
       transducers and wave generators of the paddle type.  It is
       preferred to control both frequency and amplitude of mixing
       energy, while not physically disrupting the slick or the upper
       water column.  An ultrasonic transducer and a wave generator meet
       these requirements and both have a minimum physical presence
       inside the tank.  Also,  they permit close  regulation of the
       turbulent structure in the tank, with direct influence on inci-
       pient slick breakup and  droplet motions.   With these devices, the
       turbulence in the tank can be varied, characterized and reproduced
       By using a large tank, interference from  waves produced by reflec-
       tion at the walls can  be minimized and/or  artificial wave
       dampeners may be used.

            To impart  mixing  energy to the tank  contents, a wave genera-
       tor was designed.  It  consists of a paddle, made from 18 mm thick
       galvanized steel, that is 12 inches high  and 35 inches wide.   The
       paddle is hinged to two  flexible aluminum  plates, such that it is
       suspended vertically and dips 3 inches into the water in the tank.
       The paddle is capable  of generating surface waves when driven by
       a cyclic mechanical drive.  The paddle is  driven by a 3-inch
       diameter Plexiglass disk mounted on the shaft of a high-torque,
       brush-type electric motor (Bodine Electric Co., Chicago, 111.,
       Ser. # 3424955).  The  disk is mounted off-center on the motor
       shaft to produce an eccentric sweep each  half-turn.

            Since the  paddle  was installed at one end of the tank, the
       system generates surface waves as the rotational force from the
       motor shaft is  transmitted to the eccentric disk.  As the disc
       turns, it displaces the  paddle.  The forward and backward move-
       ments of the paddle displace the water surface and generate waves.
       Contact between the paddle and the disk is maintained by two    "
       springs connected to the paddle and the tanks, one on each side
       of the disk.

            Torque transfer from the motor to the paddle is controlled
       by varing the voltage  to the motor using  a Variac.  Wave amplitude
       and frequency are controlled by varying the speed of the motor
       and the depth of paddle  immersion in the  water.  The maximum
       displacement of the paddle is 1.5 inches.   By allowing the system
       to operate for  about 10  minutes, a constant wave pattern can be
       established in  the tank.  The wave pattern is reproducible and
       artificial devices were  not used at the tank walls to dampen the
       waves.
W

                                      66

-------
PROCEDURES

Cleaning of Glassware and Test Tanks

     The capacity of oils and hydrocarbons to adsorb on surfaces
is well known.  In the absence of adequate cleaning procedures,
the results of laboratory experiments may be biased due to arti-
facts from contamination of experimental  apparatus.  Separatory
funnels for extracting hydrocarbons from thief samples, collec-
tion bottles, etc., are critical  to realistic determination of
oil concentrations.  Thus, glassware must be cleaned thoroughly
to minimize contamination of samples.  Cleaning of glassware and
laboratory equipment is an energy-intensive and time-consuming
process in experimental studies of oil/water systems.

     All glassware was washed in  soap solution (Sparkleen) and
rinsed in running warm tap water.  Then,  the glassware was
extracted with acetone and carbon tetrachloride and dried in an
oven at a temperature exceeding  150C.   Cleaned glass  containers
were kept capped until used.

     The plexiglass tank was cleaned by scrubbing the  walls with
sponge and Sparkleen soap.  The tank was then rinsed with hot
water from a hand-held hose.  Test water was always examined for
visible oil sheen before each spreading experiment.  If an oil
sheen was present, the cleaning procedure was repeated.  For the
chemical dispersion studies, the  sampling tubes were disengaged
and cleaned independently.  At the beginning of each experiment,
test water was sampled and extracted with carbon tetrachloride.
If the infrared spectra of the sample indicated the presence of
residual oil, the cleaning process was repeated.

Hydrophilic Treatment of Glass Surface

     The spurious development of  multiple phases during sampling^
of hydrocarbon/water systems poses great difficulty and biases
determinations of oil concentrations in water samples.  This
phenomenon may be due to surface/oil interactions during sampling
As glass surfaces are wetted preferentially by oil, adsorption
onto surfaces of sampling probes  constitutes a sink for hydro-
carbons when test solutions containing low concentrations of oil
are sampled.

     In order to minimize interactions between glass surfaces and
oil, and experimental artifacts via sampling devices,  it was
decided to precondition the surfaces of collection bottles,
sampling tubes, and separatory funnels to make them hydrophilic.

     Glass surfaces were treated  with trimethoxysi1ane to render
them hydrophilic.  The treatment  procedure described in the manu-
facturer's brochure was used.  It consists of a)  washing the
glassware with a suitable detergent, b)  dipping glassware in a


                               67

-------
solution of trimethoxysi 1 ane for a few minutes,  and c)  air- or
oven-drying the glassware.

     The manufacturer claims that the wetting agent forms a
protective film on the glassware.  This film is  stable to clean-
ing with organic solvents  and solutions that have a broad pH
range, for over three months.

Preparation of Seawater Solutions

     Synthetic seawater solutions used in this study were prepared
in batches, as required.  Instant ocean sea salt was dissolved in
50 gallons of tap water in  a 55 gallon drum.  Complete dissolution
of the salt in water was achieved by a high speed mixer.   A 50 ml
sample was taken from each  batch and evaporated  to dryness in an
oven .

     The salt content of these samples showed that the percent
salt varied slightly for different batches; the  average was
approximately 3.51%.  This  value is in agreement with the salt
content of natural seawater.

Sample Collection

     Water samples were collected for analysis during experimental
investigations of the dissolution and chemical dispersion of oils.
Sampling can be complicated by emulsion formation or contamina-
tion.   The procedure for sample collection must  assure that a
representative portion of  the test system liquid is withdrawn.

     For the dissolution studies, the first sample in each experi-
ment was collected by means of vacuum suction.  Thereafter,
sampling occurred by gravity flow.  The first few mis of  solution
that were retrieved during  sampling were always  discarded.  This
corresponds to the liquid  holdup in the probe.  Approximately
250 mis of sample were collected daily for analysis.

     During chemical dispersion studies, samples were collected
by gravity flow through the probes.  Preliminary investigations
showed small gradients in  oil concentrations when samples from
different locations in the  tank were analyzed.  So, a three-
dimensional matrix of water samples was always collected.  50-ml
samples were collected from each of the 15 locations.  The
samples were combined to give a sample that is representative of
the tank contents.

     All water samples were collected directly into separatory
funnels.  The funnels were  stoppered immediately to prevent
volatilization and changes  in sample composition.  Usually,
samples were extracted and  analyzed within 10 hours so that
preservation of samples was not necessary.
                               68

-------
•v-     Determination of Oil  Concentrations in Water

            Measurement of the concentrations of oil  in water samples is
       a necessary part of most investigations of oil/water systems.
       The determination of hydrocarbon concentrations in water poses
       several analytic problems because of the very low concentrations
       expected for a wide range of hydrocarbons.   A review of the major
       analytical  techniques shows great diversity, but as yet no single
       method is a panacea for all types of problems associated with the
       determination of total  and hydrocarbon species concentrations in
       water.  The primary methods now available for measuring the
       concentrations of oil in water are spectroscopic, gravimetric,
       and chromatographic techniques.

            Infrared spectrophotometry is the most commonly used method.
       The use of infrared (IR) spectroscopy to quantify the oil content
       of water samples is an  established procedure.   The method is
       sensitive,  accurate and efficient.  This technique has become
       popular, also, because  of the short time required for analysis.
       But, this method is primarily for the analyses of alkanes and to
       a lesser extent of aromatic compounds with side chains.  In this
       method, the oil/water sample is extracted with carbon tetra-
       chloride (CCl^) and the total oil  content of the extract is

       quantified by measuring the maximum absorption attributable to

       methylene (-CH-) stretching frequencies at 2930 cm   in 1, 5 or
•«*,,,                   J
       10 cm path-length quartz crystal cells.  Routine determinations  of
       the total concentration of oil  in water samples were made with
       the IR method.

            Determination of the concentration of oil in water samples
       using the IR method has been described in detail by Gruenfeld
       (1977).
                                                                       w
            The procedure involves:

            1.  extracting oil from water samples  with an organic
                sol vent;

            2.  analyzing extracts with an IR spectrophotometer; and,

            3.  referring the  maximum absorbance of the extract to a
                calibration curve to determine the oil  content of the
                extract.

       Gruenfeld (1977) discussed the influence of salt and acid
       addition to water samples, prior to extraction with the solvent,
       and the detection limits of oil by IR.  Salt and acid were not
       used in the present assays.
                                      69

-------
Extraction--
     Oils are soluble in
carbon tetrachloride was
oil  from water
in high purity,
is recommended
15 mis of CC1,
                    most organic  solvents.   Spectral  grade
                    selected as  the solvent  for extracting
          samples for several  reasons:
           with little or no  spectral
          by the EPA.  To extract  oil
          are added to the separatory

          The funnel  is stoppered  and
                                          :   it can be obtained
                                           interference,  and it
                                           from a water sample,
                                           funnel containing the

water sample.  The funnel  is stoppered and shaken vigorously by
wrist action for a period  of 3 minutes in an inverted position.
The stoppered funnel  is placed on a stand and allowed to  remain
undisturbed for about 1 hour.   By this time, the content  of the
funnel has separated into  two  distinct phases.   The bottom phase
containing CC1, + oil is drained into a clean,  solvent-rinsed

bottle.   The extraction is repeated on the water phase with
another 15-ml portion of solvent.  The extracts are combined and
shaken with 2 gms of anhydrous, granular sodium sulfate (Na^SC,)

to absorb moisture and water droplets that may be entrained in
the extract.  The volume of the oil-free water sample is  measured
using a graduated cylinder.

Infrared Analysis--
     Two matched IR cells  (1,  5 or 10 cm path-length) are care-
fully rinsed with CCl* and one cell, i.e. the reference cell is

filled with CC1,.  The sample  cell is filled with extract.  The

cells are placed in the IR spectrophotometer and the differential

spectrum is scanned from 3200  to 2700 cm"  wave numbers.   The
maximum absorbance of the  extract is measured:   it occurs at

about 2930 cm" .  The concentration of oil in the sample  can be
determined by referencing  the  maximum absorbance to a calibration
curve of absorbance versus concentration for the cell size used.
If the maximum absorbance  is greater than 1.0,  a smaller  cell  is
used or the extract is diluted with a known  amount of solvent.
The concentration of oil-in-water is determined from the  follow-
ing equation :
                              Cc-VCCl
                         Co -
                                                        (43)
where
     C  = concentration of oil-in-water,  yg/ml
C  =
          concentration of oil  in the extract determined from
          the calibration curve,  yg/ml  CC1,

          volume of CC14 used to  extract the sample,  ml
                               70

-------
      f = dilution factor (if used,  otherwise f = 1)

     V  = volume of water sample, ml

      The instrument was checked for baseline drift regularly.
CC1, from the same bottle was used for the extractions and for

filling the reference cell.   A segment from the spectra of an
oil/water sample is shown in Figure  9.

Calibration Curves--
     A concentration vs. absorbance  plot derived from IR scans  of
several known concentrations of oil  in CC1. (solvent) is used for

quantitating oil in water samples.  A standard solution is
prepared by dissolving a weighted amount of oil into  100 ml  of
CCK.  Aliquots of the standard solution are diluted  with CC1.  to

obtain solutions at different concentrations.  Infrared analysis
of at least four solutions provides  absorbance versus concentra-
tion data which is used to prepare a calibration curve for the
specific oil and cell size.

     The absorbance and concentration data were correlated by
linear regression analysis.   The correlation coefficients range
from 0.98 to 1.0, indicating significant linearity.  The inverse
equations of the regression  expressions were used to  determine
oil  concentrations in water  samples.   Appendix A shows sample
calibration curves for one of the oils.

Efficiency of Extraction--
     The efficiency of the method that was used to extract oil
from water samples depends on several factors, such as the quan-
tity of oil in the sample, the volume of CC1., oil type, and the
distribution of oil between  CCl/and water phases.

     Two water samples containing 0.5 and 1.0 gms of  oil were
prepared as follows.  The oils were  dissolved in 5 mis of acetone
and the acetone solutions were spiked into 1 liter of water.  For
a control, 5 mis of acetone  were spiked into 1 liter  of water.
The three water samples were then extracted with CCl^, and the
absorbance of the extracts determined.  By subtracting the
absorbance for the control from those for the samples, the quan-
tity of oil extracted in each sample was determined.   Four oils
were used in this experiment:  #2 fuel, #6 fuel, Nigerian crude,
and Iranian crude.

     In all experiments, the percent recovery was greater than
90%.  The good results may be attributable to the quantity of CC1.

used in all extractions and  the hydrophilic treatment of glassware
surfaces which minimizes the adsorption of oil.


                               71

-------
CO
CC
o
«/>
03
      3500
      3000
WAVENUMBERCCiN/T1)
2500
        Figure 9.  A Segment from the Spectra of the  CC1.
                   Extract of an Oil/Water Sample
                              72

-------
Protocol for Studying Spreading Rates of Oils on Calm Water

     Experimental studies to determine the spreading rates of the
12 oils were conducted in the open, rectangular Plexiglass tank.
Because of the constraint imposed by the two horizontal dimensions
of the tank (5'x3'), the duration of experimental  runs was limited
to the first 35 minutes after slick initiation.  A run was dis-
continued after 35 minutes or when the spreading oil contacted the
sides of the tank.  This was necessary to minimize wall-effects on
the spreading slick.

      The  temperature  of  the  tap  water  flowing  into  the  tank  was
 controlled  by  adjusting  the  flow rates  of  cold  and  hot  water
 such  that  water  temperature  was  approximately  20±1C.   It  was not
 necessary  to  control  the temperature of  the  water during  a run
 because of  the short  duration  of the experiments.


     For each oil, four different volumes (25,  50, 75, 100 mis)
were spilled.  Although the oils were spilled at different rates,
they were discharged completely within  10 minutes.

     The spreading experiments were performed according to the
following procedural sequence.

     1.  Tank was cleaned and rinsed as described  in the cleaning
procedure section.

     2.  Tank was filled with tap water at 20 ± 1C to a depth of
12 inches.   Water was checked for surface films or irridescent
color.  Cleaning  procedure was repeated if a surface film was
observed.

     3.   Flood lights were turned on.

     4.  Camera was mounted on the Techno!  copy stand such that "
the whole tank could be viewed through  the view finder.

     5.  A circular piece of carbon paper (3.625 inches in
diameter)  was floated on the water surface.   The camera was
focused on the paper.   Exposure level  and shutter  speed were
selected to secure good contrast between paper  and background.
All required adjustments on camera, motor-drive, flood lights,
etc.  were completed.

     6.  The carbon paper was photographed.   This  served as the
basis for determination of the exact magnification of photographic
images of the spreading oil.

     7.  The oil-feeding system was installed and  the volume  of
test oil  plus an  allowance for clingage was  measured into  the oil
holding flask.   The flask was connected to  the  oil-feeding system.
                                73

-------
     8.  The peristaltic pump was turned on.

     9.  The stopwatch was started and the interval timer acti-
vated as soon as oil was discharged from the oil-delivery tube
and rose to the water surface.  The interval timer was adjusted
to sequence photograph at 1  minute interval  .

    10.  The stopwatch was stopped when the  oil  has been
discharged completely and pertinent data recorded.

     In each run, ample time was allowed for the water in the tank
to become quiescent and the  procedure above  was  standard for all
the oils tested except #2 fuel oil.  Because of  the poor contrast
of #2 fuel, a small quantity of red dye was  dissolved into the oil
before it was measured into  the holding tank.

Determination of Oil Slick Spreading Areas--
     All photographic films  were developed and printed into
8"xlO" prints in the dark-room facilities of the Rutgers Depart-
ment of Mechanics and Material Sciences (New Brunswick, N.J.).
Figure 10 shows sequence photographs of one  of the spreading oils.
The areas of the photographic images in the  photographs were
measured using a Compensating Polar Planimeter (Model #620010,
Ser. #85-203), manufactured  by Keuffel and Esser Co.  (Morristown,
N.J.).

     The instrument differs  slightly from a  needle compass because
of the presence of a tracer  point and a measuring wheel connected
to a vernier scale.  To measure an area, the tracer point is run
around the periphery of the  profile from a starting (and finish-
ing) point.  As the figure is traced, the measuring wheel rotates.
After a complete circuit of  the figure, the  distance  which the
wheel has revolved is determined from the difference  of the
initial and final readings on the vernier scale.

     For each series of photographs, the area  of the  image of the
reference carbon paper was first determined.  Since the actual
area of the carbon paper is  known, the areas of  the images of the
spreading oils in the same sequence photographs  could be deter-
mined.  Each photograph was  traced twice and the readings
averaged.

     The planimeter was calibrated by measuring  several areas of
rectangular, cylindrical and spherical geometries.  From compari-
sons of the measured and actual areas, the precision  of the
instrument was deemed greater than 99%.  Thus, uncertainties in
measurement of actual spill  areas are due largely to  identifica-
tion of spill boundaries as  a result of poor imagery.

Accuracy of the Spreading Areas--
     It was observed during  the spreading experiments that the oil
slicks did not spread uniformly.  Regions with different thick-

                               74

-------
Figure 10.   Sequence  Photographs  of Oil  During Spreading
                          75

-------
nesses of oil  were visible and appeared as different intensities
in the dark color of the images when photographs were printed.
Usually, darker regions were surrounded by lighter ones.  Most of
the oil  was present in the dark regions.

     The spreading patterns of the oils showed great diversity,
also.  Different configurations could be seen for different oil
types; however, spreading patterns were more round than rectangu-
lar.  A few oils spread in a circular fashion.  It was not known
whether the different configurations were caused by the molecular
motions of water and small advection currents or by the unequal
surface tension forces at the free surfaces of the oil slicks.

     Situations were encountered in which the spreading oil
covered the water surface within a few minutes and impacted the
tank walls.  Experiments were repeated in such cases.  In other
cases, the determination of the area! extent of spreading oil  was
impossible with a planimeter because spreading was not continuous.
These oils usually formed fingers that were disjointed and
separated by water streaks.

     It was difficult to identify regions with similar contrast
in all the photographs because of different spreading behaviors.
Clearly, the accuracy of the spreading areas depends on the
configuration  and spreading behavior of the slick.  For two
identical experiments, the areas covered by the same oil could
vary by as much as 20%.  This underscores the need for a more
accurate technique for determining the spreading areas of oil
slicks.   A color densitometer may be better suited for measuring
spreading areas directly from photographic negatives.

Protocol for Studying the Dissolution Rates of Oils

     The experiments to study dissolution rates of oils were
conducted in three phases.  In the first phase experiments, the v
dissolution of oils in tap water was studied.  All twelve crude
oils and derivatives were tested during this experimental phase.
In the initial stages of the experiment, two oils (Nigerian and
Iranian crudes) were equilibrated with water for three weeks.
The results of these experiments showed that equilibrium was
achieved in about two weeks.  Therefore, the duration of subse-
quent dissolution experiments was reduced to two weeks.

     During the second phase of experiments, the dissolution
rates of only  six oils in salt water were studied.  Five crude
oils and #2 fuel oil were tested.  The crude oils are:  Nigerian,
La Rosa (Venezuela), Brass River (Algeria), Iranian and North
Slope (Alaska).  All oi1/salt-water systems were allowed to
equilibrate for two weeks.

     Experiments using tap water and salt water were conducted in
the dissolution test tank described earlier.  The procedure

                               76

-------
outlined below was used for the experiments during both phases.

     1.   Test tank was cleaned according to the established
         cleaning procedure.

     2.   Cooling coils attached to the constant temperature
         bath were installed  in the tank.   The sampling tube
         was installed, also.

     3.   30 liters of water (tap or 3.5% salt water) was
         metered into the tank.

     4.   Water was sampled by vacuum suction and analyzed for
         oil contamination.  If there was  contamination, the
         procedural  sequence  1) to 4) was  repeated.

     5.   300 mis of oil were  layered on the water by gently
         pouring the oil  via  the side of the tank.

     6.   The constant temperature bath was turned on.

     7.   The water was sampled daily for the duration  of the
         experiment.

     Prior to each sample collection, the  position of  the sampling
probe was adjusted gently so  that all samples are collected 6
inches below the oil/water interface, in the center  of the
cylindrical tank.  The volumes of water samples collected varied
from 250 to 500 mis.  During  sampling, precaution was  taken so
that the surface slick was not disturbed.

     In  the third phase of the dissolution experiments, concen-
trations at saturation were determined for all twelve  oils.  The
procedure followed was different from that used in the first two
phases.   In brief, 200 mis of each oil were poured into 1-liter
separatory funnels containing 500 mis of tap water.   The funnels
were stoppered tightly and stored in the dark for three months.
After this time, 400 mis  of the aqueous phase was withdrawn from
each separatory funnel.  Samples were extracted with CC1. and

analyzed to determine their oil content.

     It  was assumed that  the  oil concentrations determined by
this procedure would be close to the "actual" saturation values
of the oils tested.   The  rationale in adopting this  procedure
stems from the fact that  oil/water systems can become  mutually
saturated when the ratio  of oil to water is close.  Since oils
equilibrate slowly with water, prolonged equilibration and
"closed" experiments are  necessary conditions for accurate deter-
mination of concentrations at saturation.
                               77

-------
Protocol  for Chemical  Dispersion Studies

     Although several  procedures have been used to determine the
effectiveness of oil  spill  dispersants,  there is no meaningful
common ground regarding acceptable procedures for evaluating
dispersants.

     The  efficiencies  of 5 dispersants were evaluated in this
study.  The dispersants are referred to  as products A, B, C, D,
and E.  Three oils were treated with the dispersants.  They are
#2 fuel  oil, Iranian  crude, and #6 fuel  oil.   The degree of diffi-
culty with which these oils can be dispersed  varies from "easy"
to "tough".

     Several variables influence the effectiveness of oil spill
dispersants, and it is cumbersome to evaluate all the factors,
even for  one oil/dispersant pair.  The effects of three oil-to-
dispersant dosage rates (1:1, 5:1 and 10:1 vol/vol) were studied
using #2  fuel oil.  The chemical dispersion of the other two oils
was studied at a 5:1  oi1-to-dispersant ratio  only.  Thus, one
emphasis  of this comparative study concerns the influence of
oil/dispersant ratio  on the effectiveness of  chemical dispersants,
These tests above were conducted with similar levels of agitation,

     The  effect of turbulence on the effectiveness of dispersants
is well  known.  The efficiency of oil spill dispersants increases
with the  intensity of  agitation.  As some manufacturers have
claimed  that it is not always necessary  to provide mixing energy
for certain dispersants, tests were conducted with product B to
determine the efficiency of the dispersion with continuous agita-
tion and  without agitation.

     The  effect of salt water on the dispersion of oil slicks
with dispersants was  also investigated with product B.  Tests to
determine the effects  of salt water and  turbulence on chemical
dispersion were conducted at an oil-to-dispersant ratio of 5:1.
Iranian  crude and #2  fuel  oil were used.

     Before actual comparative tests were begun, standard proce-
dures for dispersant  application and wave generation were
developed in preliminary tests.

     All  dispersants  were atomized by pumping the dispersant
solution  at 2cc/sec and flowing compressed air at 375 cc/min to
the nozzle.  The nozzle tip was adjusted to produce a fine spray
with a 9  inch swath at the water line.  The railroad car (and
nozzle)  travels at a  constant height of  18 inches above the water
level.  The dispersant solution is sprayed vertically downwards
at a constant angle onto the oil layer and hits the water surface
with a constant impact velocity.  The direction of travel of the
buggy is  always from  the downstream end  of the tank (where wave
breaking  occurs) to the upstream end (where waves are generated).

                                78

-------
       Usually  dispersants  were  applied  in  1  pass  as  the  nozzle  travelled
       across  the  tank  at a  constant  speed  of 0.075 ft/sec.   The  buggy
       was  stopped,  at  selected  points along  the tank,  during  application
       of  300  mis  of dispersant  in  order  to discharge  the  chemical  in 1
       pass.

            The standard waves  in all  tests were generated  by  adjusting
       Variac  output to  65  volts.   At  this  setting, the eccentric  disc
       of  the  wave generator makes  50  revolutions  per  minute.  At  the
       beginning of  wave generation,  waves  can  be  seen  breaking  at  the
       downstream  end of the tank.  After about 10 minutes  of  continuous
       wave generation,  a quasi-equi1ibrium is  established.   Surface
       waves are random  and  have a  characteristic  amplitude of approxi-
       mately  3 inches  and  a wave length  of about  15  inches.

            The procedural  sequence for  chemical dispersion studies
       follows.

            1.   The  tank was cleaned  according  to  the  established
                procedure and filled  with tap water to  a  depth of
                18 inches.   The  temperature of  the tap  water was
                maintained  at 25 ±  1C  by  controlling  the  flow
                rates of hot and cold  water.

            2.   The  water was sampled  and the sample  analyzed  to
                determine whether residual  oil  was present.   If
                necessary,  the  tank was  cleaned again.

            3.   The  wave generator  was turned on and  the  setting
                on the  Variac checked.   Waves were generated  for
                up to 10 minutes to establish a steady-state wave
                and  current  pattern in the  tank.

            4.   300  mis  of  oil  were poured  on the  water in  the
                tank.   The  oil  was  allowed  to spread  for  approximately
                5  minutes.

            5.   The  volume  of dispersant  to be  tested  was  measured
                into the dispersant holding vessel.

            6.   Application  of  the  dispersant onto the  oil  was commenced.

            7.   Wave generation  was stopped after  15 minutes.

            8.   As soon  as  the  wave generator was  stopped,  approximately
                50 ml samples were  collected  from  each  of  the  15
                sampling probes, using gravity  flow.   Samples  were  trans-
                ferred  into  1-liter separatory  funnels  for  extraction
                and  analysis.
W

                                      79

-------
     9.  Samples were collected every fifteen minutes for the
         first hour, and at thirty minute intervals during the
         next hour.  Later, samples were collected twice at
         hourly i ntervals.


PRESENTATION OF EXPERIMENTAL DATA

     Experimental data are  presented in Appendices B to E.
Appendices B and C include  data for oil solubility studies in tap
and salt water, respectively.   All of the oils were equilibrated
in tap water; only six oils were equilibrated in salt water.   The
experimental data for runs  in  tap and salt water show similar
trends.  The concentration  of  oil in water rises during the first
few days; after it dropped  gradually.  Generally, concentration
for runs in salt water were approximately an order-of-magnitude
(or more) lower than runs in tap water.

     Appendix D contains data  for areas covered by spreading  oils.
Four different volumes were spilled for each oil.  The experi-
mental data shows a steady  increase in area covered by any oil
during the first few minutes;  later, the area increased at a
slower rate.

     Appendix E shows the data for the chemical dispersion studies
The dispersion data for #2  fuel oil  and the 5 dispersants
(Products A to E) at 1:1, 5:1, and 10:1 oil-to-dispersant (0/D)
ratios are given in Tables  El, E2 •, and E3, respectively.  It  can
be seen that oil  concentration declined throughout the sampling
period, that is after the wave generator had been turned off.
Also, oil concentrations are higher for high dispersant dosage
rates.  Tables E4 and E5 present the data for the dispersion  of
Iranian crude oil and #6 fuel  oil, respectively.  Table E6
summarizes the data for the dispersion of #2 fuel oil and Iranian
crude in salt water.  When  this data is compared with those in
Tables E2 and E4, for the same oil/dispersant pair, two conclu-
sions can be made concerning the effect of salts.  First, there
were no significant differences between the data for the disper-
sion of #2 fuel oil in tap  and salt water.  In contrast, Iranian
crude behaved differently in dispersion studies in tap and salt
water.  Table E7 compares the  dispersion behavior of #2 fuel  oil
and Iranian crude under calm conditions and continuous agitation.
The low oil concentrations  in  tests without agitation suggest the
necessity for additional mixing energy for efficient dispersion
of chemically-treated oil spills.  When mixing energy is supplied
continuously to disperse oil slicks, almost complete dispersion
can be achieved.
                               80

-------
                           SECTION 8

            MODELING OF OIL SLICK DISPERSAL MECHANICS
     Mathematical  models are important for advance quantitative
and qualitative assessment of contamination of marine environ-
ments by oil spills.   It is impossible to derive models that are
valid for all  oil  spill  situations because the rates of oil
dispersal processes are  site dependent.   Also, the hydrodynamics
of the water column under an oil  slick and numerous environmental
factors control the rates of oil  slick dispersal processes.
Since a majority of factors affecting oil dispersion cannot  be
quantified accurately, there is little merit in sophisticated
models when accurate  input data are unavailable.

     In this section, simplistic  models  for the rates of
(i)  dissolution and  (ii)  spreading on  calm water of crude  oils
and petroleum products are developed.  A major feature of each
model is the ability  to  fit experimental data closely.  These
models should provide reasonable  estimates of these dispersal
processes under the environmental  conditions for which the models
are valid.  Also,  the models will  require as input only physical
properties of the  oil and water phases and other properties  of
the system which can  be  determined easily.

     Finally,  the  mechanisms of chemical dispersion and rate
expressions for some  of  the mechanisms will be presented.
Because of the complexity of chemical dispersion processes,  an
overall rate expression  is difficult to  derive.


THE MECHANISMS OF  DISSOLUTION

     Discharges of oil on water masses can occur instantaneously,
continuously or in a  combination  of both rates.  In the models to
be derived, it is  assumed that the oil has been discharged
completely before  commencement of the dissolution process.  Thus,
the oil/water system  considered here is  an unlikely worst case of
an oil spill,  such as an oil pool  which  covers the entire surface
of a lake or pond.  This situation is different from other oil/
water systems, e.g. those formed  from natural seeps, as infusion
and dissolution of oil occur simultaneously.
                               81

-------
     Figure 11  shows an oil/water system in dynamic equilibrium.
The system consists of three phases:  air, oil and water.  The
thickness of the oil layer is much less than the other two phases.
Regions of major interest are the air/oil  and oil/water inter-
faces.   The system may be visualized as the worst possible oil
spill  in a marine environment, as the oil  is assumed to cover the
entire  water surface.

     The following mass transfer processes occur when the system
is allowed to equilibrate:

     a)  volatile hydrocarbon evaporate at the air/oil interface;
     b)  hydrocarbon species diffuse from the bulk oil layer to
         both the air/oil and oil/water interfaces; and
     c)  hydrocarbon species diffuse into and dissolve in the
         bulk water phase.

     These processes occur simultaneously and the individual rates
are dependent on each other.  The rates of evaporation and disso-
lution  depend on the supply of volatile and soluble hydrocarbon
species from the bulk oil.  The relative rates of both processes
depend  on the extent to which the species partition into gas and
water phases.

     Before deriving the equation for the rate of dissolution of
crude oils and petroleum products in water, it is useful to
consider the trend of the experimental  data presented in Appendix
B.  The data for all experimental runs show the following trends.
                            AIR
                                      	air/oil  interface
                            OIL
                            UATCD            oil/water interface
       Figure 11.  The Three Phases of an Oil/Water System
                                82

-------
At the beginning of the equilibration processes,  oil  concentra-
tions in water are low but increase gradually to  maximum values,
within a few days.  All oils did not reach maximum concentrations
in water on the same day.   In general,  oils reached maximum solu-
bility levels in water after approximately eight  days.   Oil
concentrations declined gradually,  during the later stages of the
equilibration period.

     The trend in the  experimental  data suggests  that diffusion
of hydrocarbon from the saturated oil/water interface into the
bulk liquid column occurs  during the early stages of the equili-
bration process, until the concentration of oil  in water reaches
a maximum.   During this period,  the driving force for the diffu-
sion process is the concentration gradient, i.e.  the differential
concentration between  the  oil/water interface which is  assumed to
be saturated with oil  and  the concentration of oil in the bulk of
the water column.  The water column is  assumed to have  a uniform
concentration of oil.

     Movement of hydrocarbons to the oil/water interface occurs
by molecular diffusion from the  bulk oil layer.   This transfer of
soluble/volatile hydrocarbon species is necessary to "saturate"
the interface but causes depletion  of these materials in the bulk
oil layer.   Molecular  diffusion  of  similar materials occurs from
the bulk oil layer to  the  air/oil interface where they  are lost
to the atmosphere by evaporation.  Thus, evaporation and dissolu-
tion processes are occurring simultaneously, but  these  processes
may not necessarily be in  equilibrium.

     After  maximum oil concentrations have been  reached, the
steady decline in concentration  of  oil  in the water phase is
caused by evaporation.  Volatile hydrocarbons escape and deplete
the top layers of the  oil.  If evaporation of hydrocarbons is to
proceed, the transfer  of volatile hydrocarbons from the bulk oi'K
layer must  be maintained.   When  the bulk oil layer is depleted of
volatile hydrocarbons, diffusion of hydrocarbon  to the  air/oil
interface continues with transfer of oil already  dissolved in the
water column.  This is due to the greater partition coefficients
of hydrocarbons with air than water.  Thus, some  of the hydro-
carbon that evaporates at  the air/oil interface  is derived from
the oil in  aqueous solution, even though the oil  may be present
at a concentration which is less than saturation.  The  final
concentration of oil during the  later stages of  equilibration
depends on  the hydrocarbon species  present in each oil.  Some
soluble hydrocarbon species are  not volatile and  are retained in
solution.

Formulation of the Kinetics of Dissolution

     Crude  oils and petroleum-based products contain organic
materials that are soluble in water.  Typically,  these  water-
soluble species include hydrocarbon compounds that contain from 1

                                83

-------
to 20 carbon atoms (C-i-Cog).  as we^  as hydrocarbon  compounds
with nitrogen, sulfur,  and oxygen molecules,  and organo-metal1ic
compounds.   The complete range of soluble and volatile compounds
cannot be identified for any  oil  type.

     In order to derive an expression for the rate of dissolution
of oils in  water,  it is important to  comment  briefly on the
transport of materials  through systems  consisting of multiple
phases.  Several processes which involve transport of materials
between layers of  different phases are  important in  nature.  For
example, the exchanges  of gas across  air/water interfaces  (Liss,
1973; Liss  and Slater,  1974;  Broecker and Peng,  1974), and  evap-
oration of  hydrocarbons (Mackay and Matsugu,  1973; Mackay  and
Wolkoff, 1973; Butler,  1975;  and Cohen  et al ., 1978) have  been
investigated.

     The rate of transfer of  materials  across interfaces can  be
calculated  by several  methods depending on the physical  system
and the given set  of conditions.   Danckwerts  (1951)  described
several models which have been proposed to explain transport
across multiple phases.  The  concept  of stagnant films at  inter-
faces is well entrenched in engineering literature.   Also,  the
film theory is useful  for visualizing processes  at interfaces.
This concept is applied here.

     The air/oil and oil/water interfaces in  Figure  11  are  each
assumed to  consist of two films.   For example, for the air/oil
interface,  it is assumed that there is  a thin film of gas  on  the
air side and a thin oil film  on the oil side.   Similarly,  oil and
water films exist  on either side of the oil/water interface.
Since the thickness of  the oil layer  formed by a majority  of
spills is negligible,  it is convenient  to visualize  an oil/water
system as consisting of only  two films:  gas  and liquid films on
the air and water  sides, respectively.   Thus,  the oil  layer is
considered  to be an extended  interface  separating air and  water
phases.  The regions of an oil/water  system and  theoretical
profiles of the oil concentrations are  shown  in  Figure 12.

     The two-film  theory assumes that transport  of a material
from one phase to  the  other occurs by molecular  diffusion  through
both films.  If the transfer  is a steady-state process,  there is
no concentration build-up in  the interface and resistance  to
transport is due to the gas and liquid  films.   Gas and liquid
film resistances may be considered to be in series and the  rate
of transfer of material is controlled by the  film which offers
the greatest resistance to diffusion.  This theory has been used
with great  success to  explain rates of  gas transfer  between air
and water and evaporation of  liquids.  In some situations,  only
one of the  two resistances is significant.  For  instance,  there
is no liquid film  resistance  during the evaporation  of pure
liquids because of the  absence of a concentration gradient  in pure
1 iquids .

                               84

-------
                  BULK AIR
                            GAS FILM
                               LIQUID FILM
                                                  AIR/OIL
                                                   INTERFACE
                                                  OIL/WATER
                                                   INTERFACE
         Figure 12.  The  Regions and  Concentration  Profiles
                     of an Oil/Water  System
     For the most general case of non-steady-state  transfer  of a
compound between two phases, the basic  equation  for the  one-
dimensional case has the form:
dC   n
                                  d2C
                                                              (44}
where
                             3
     C = concentration, mg/cm
                        2
     D = diffusivity, cm /day

     z = direction normal to the plane across  which  transfer
         occurs, cm

     t = time, days

If the transfer process occurs by molecular  diffusion  and  steady-
state conditions apply, Pick's first law  is  used  to  determine  the
diffusional flux in the direction of transport:
                                85

-------
                            F = D                            (45)


where
                                                               2
     F = the flux of the compounds through the layer, mg/daycm
                        2
     D = diffusivity, cm /day

     C = concentration, mg/cm

     z = distance in the direction of transport, cm

Equation (45) is usually written as

                             F = kAC                         (46)

where

     k = mass transfer coefficient or velocity of the transfer
         process, cm/day

    AC = the concentration difference between the boundary
         surface and the average bulk concentration, mg/cm3

     Thus,  the flux of a material  across a layer is proportional
to the concentration driving force.  The magnitude of k is deter-
mined by the geometry and the flow characteristics of the system.

     The experimental data suggests that the dissolution process
could be divided into two parts with separate rate expressions
for the solution and evaporation phases.  The combined equations
yield the rate of dissolution for the duration of the equilibra-
tion period.
                                                                K.
     During the early stages of equilibration when oil is
dissolving  in the aqueous phase, a material  balance on oil in the
bulk water  phase is

     Accumulation = Input by diffusion from the liquid film
                         i = k   .(r _r"\                      (47 \
                        dt    L V^ L  '                      (   '
where
     j P
     -nr = rate of change of oil  concentration in the bulk liquid
                      3
          phase, mg/cm -day

     k.  = diffusivity of soluble species in the liquid, cm/day
                               86

-------
                              2
      A = interfacial  area, cm
                              o
      V = volume of liquid, cm
                                                        3
     C,  = concentration of oil in the liquid film, mg/cm

      C = average concentration of oil  in the liquid phase,

          mg/cm

The interfacial area and the initial  volume of liquid are known
from the experimental  system, but V is  not constant because of
loss of liquid from sample withdrawals.   If it is assumed that
the liquid film is saturated with oil,  the rate of accumulation
of oil  in solution at  any time is given  by


                    =MCs-C)     t  <  tm                    (48)
               TT(0) = 0       (Initial  Condition)
where

     K.  = the mass transfer coefficient (k, A/V) ,day~

     C  = oil concentration at "saturation" at the liquid film,
      5   mg/1

      Cf = average oil concentration in  the liquid, mg/1

Equation (48) is valid for the solution phase, that is up to some
time (tm) at which the concentration of oil in the aqueous phase
is maximum.   Also, it is assumed that the system is well  mixed.
Actually, oil concentration will vary with distance from  the oil/
water interface.  In this study, preliminary experimental runs
showed that the concentration of the sample taken 6-inches below^
the interface was the closest to the average value of the oil
concentrations from all the locations.

     During the latter part of the equilibration period,  the
concentration of oil in the aqueous phase decreases as a  result
of evaporation of hydrocarbons from solution in which the oil
concentration is less than saturation.   A material balance on  oil
in the water phase gives the rate of oil  loss:

             Depletion = Diffusion across the liquid film


                     - -KL(C-C*)     t  >  tm                  (49)
where C* is the oil  concentration at the oil/water interface and
K^ is the average mass transfer coefficient of the volatile

fraction of the dissolved species.  The derivation of the equa-

                               87

-------
tion relating C* to C follows.

     The rate of evaporation of oil  from solution is controlled
by three resistances due to the liquid film,  the oil layer,  and a
gas film.   The resistance to diffusion due to the oil  layer  is
negligible because of the small distance the  volatile  species
travel.   If there is no accumulation in the oil  layer,  the rate
of diffusion of volatile species from solution is in equilibrium
with the rate of evaporation at the  air/oil interface.   Under
steady-state conditions, Pick's first law is  applied to the  liquid
and gas  phases to determine the diffusional flux in the direction
of transport
                       KL(C-C*)  - Kg(cg-cj
where
                                                      (50)


                                                      "
K,,K  = liquid and gas mass transfer coefficients,  day"  ,
    9   resectively
C\C*,C
       ,
      9
                  = concentrations of volatile/soluble species
                    in liquid bulk, in the liquid at the inter-
                    face,  in the gas at the interface, and in
                    bulk air, mg/1 , respectively
The interfacial  concentrations can be related by an empirical
equation  of the form

                            Cg = KHC*                        (51 )

Equation (51 )  is similar to Henry's Law,  where KH is the Henry's

Law constant or  the partition coefficient of the volatile species
in gas and liquid phases.

     The concentration of these materials in air, C^,  is negligi-
ble.   Substituting for C  in Equation (50)  and eliminating C^
gives                   9
                        K[(C-C*)  =
                                                      (52)
Solving for C* in Equation (52)  and substituting  the  result  into
Equation (49), gives
                    —
                    dt
                                   t > t
                                        m
                                                      (53)

where
                                   (Initial  Condition)
                               88

-------
                         -L = J_ + _1_                      (54)
                         v     "   v \f                       *3 '
                         KE   KL   Kg H

and   C  = maximum concentration attained at time equal  t ,  mg/1.

     The duration of the equilibration period determines which  of
the two processes, i.e.  solution or evaporation,  is operative.
Equations (48) and (53)  jointly give the rate of  dissolution.
The lag period between the transfer of oil  into water and the
beginning of evaporation from solution is accounted for  by using
different integration limits for the two equations.

     Actually, these equations are for individual hydrocarbon
species that are volatile and soluble.  Thus, the final
expressions should be summations over all volatile and soluble
species.  Crude oils and petroleum products contain numerous
compounds with varying solubilities and volatilities.  It is an
impossible task to identify completely all  the compounds in  any
given oil that dissolve  into water and later evaporate from  solu-
tion.  The analytical method used in this study measures only  the
total concentration of extractable organics.

     If the mass transfer coefficients are  assumed constant,
integration of Equations (48) and (53) and  substitution  of the
initial conditions gives


                C = Cs(l-e~ L )    (t < tm)                  (55)

and

                C = CmexPC-KE(t-tm)]   (t > tm)              (56)

where
                                  -K, t
                       Cm - Cs(l-e  L m)                     (57)

The dissolution equations can be considered as a  segmented model
with a joint point at t  .  If the equilibration period is less

than t , Equation (55) gives the variation  of oil concentration
as a function of time of equilibration.  If the oil/water system
is allowed to equilibrate beyond tm, the concentration of oil  in

the liquid phase varies  with time according to Equation  (56).
                               89

-------
THE RATE OF OIL SPREADING ON CALM WATER

     It is impossible to predict accurately the rate of spreading
of an oil  slick.   Spreading rates are site dependent and are
influenced by the hydrodynamics of the underlying water and
surface air columns.

     A few theoretical  studies which have improved the knowledge
of oil slick transport  have been reviewed.  More recent experi-
mental investigations have involved actual field studies in which
large volumes of oil  have been spilled intentionally at sea.  The
costs associated with such studies are exorbitant and the infor-
mation derived from actual field tests may be relevant only to
specific spill situations.

     Usually, the spreading of oils on water is considered to
consist of two independent mechanisms (Hoult, 1972).  The first
mechanism is the tendency of the oil to spread as a gravity wave
on calm water; the second mechanism comprises the gross transport
of oil masses in the  presence of external forces, i.e. the
convective forces of  winds, currents, tides and waves.  The total
area covered by an oil  spill is a combination of the areas
covered by both spreading mechanisms.  Knowledge of oil transport
by convective forces  at sea is improving because of increasing
observations of actual  spills.

     Two approaches have emerged from previous work for dealing
with the spreading of oils on calm water.  A most significant
contribution was made by Fay (1969, 1971), who proposed three
stages during the spreading history of oil slicks on calm water
for oils discharged instantaneously.  Each stage is a balance
between a retarding and a spreading force.

     Using a different  approach, Murray (1972) derived an
expression for the area covered by an oil slick from the solution
of the one-dimensional  Fickian diffusion equation.  Fannelop and
Waldman (1971), Hoult (1972) and Buckmaster (1973) showed that
expressions similar to  Fay's can be derived by solving the basic
equations for describing movement and mixing of a contaminant,
i.e., continuity and  momentum equations, with specific boundary
conditions.

     The approach used  by Fay is better from the standpoint of
ease of estimation of the areal extents of oil slicks on calm
water.  This method is  adopted here.  The goal is to develop
expression(s) that permit prediction of areal increase with time,
for an oil slick spreading from a stationary source, using as
input data only the limited information available at a spill
site.  This information includes an estimate of the total volume
of oil spilled, duration or rate of spill, and the physical
properties of the oil and water, i.e. viscosity, density, surface
and interfacial tension.

                                90

-------
     Assumptions are that:

     a)  physical  properties of oil  and water phases do not vary
         with time;
     b)  the thickness of the oil  layer is always smaller than
         the horizontal  dimensions of the slick;  and
     c)  oil is discharged  onto water at a relatively constant
         rate.

     The first  assumption is only  valid during the initial  stages
of spreading.  Since the experimental spreading studies lasted
only 35 minutes, changes in the physical and chemical properties
of oil  and water phases  were not considered significant.   In
general, weathering processes operate on oil slicks and cause
changes in oil  composition.  Evaporation causes volatile  compo-
nents of the oil to be lost to the atmosphere and dissolution
causes  soluble  components to leach into water.  The effects of
other weathering processes, e.g. microbial degradation, photo-
oxidation, etc., may be  significant also, depending on the
duration of oil exposure.  Density and viscosity  of oil increase
with time but the net value of the surface tension balance  may be
positive or negative at  any instant.   The second  assumption is
necessary if the oil is  to  be in hydrostatic equilibrium  in the
vertical dimension.

Formulation of  Spreading Equations

     Figure 13  is a schematic diagram showing an  oil slick  on
water.   The thickness (h) of the slick varies with time,  t.  The
oil floats a height Ah above the mean water surface.  If  the
flowrate of oil onto water, Q, is  assumed to be fairly constant
then

                     V = Qt         t < tD                   (5B)


                     V = Vt         t > tD                   (59)

where

     V  = volume of oil,  cm

    V.  = total  volume spilled, cm

     Q  = volumetric oil  flowrate,  cm  /sec

     t  = time,  sec

    tp  = duration of oil spill, sec

If exchanges of oil at the  free surface boundaries of the slick
due to  evaporation and dissolution are negligible, conservation


                                91

-------
                                                          w
                     WATER
            Figure 13.
Schematic Diagram of an Oil  Slick
on Calm Water Showing Spreading
and Retarding Forces
of mass requires that
where
                             V = Ah
                                     (60)
     A = areal  extent of the slick,  cm
                                                                w

     h = mean thickness of slick, cm

     The forces which cause oil  slicks to spread and/or shrink
are identified  and considered next.

Surface Tension--
     Surface tension forces at the leading edges of oil  slicks
influence the spreading rate of  oils on calm water.  Figure 13
shows the direction of surface tension forces acting on a slick.
The net value of the balance of  the  surface tension  forces, at
any time, determines the contribution of surface tension to the
overall spreading rate of the oil.  The resultant surface tension
(a) is
                   a =
•cos  eraow
               • cos a,
(61)
                               92

-------
where

   a ,a ,a   = surface tensions of water,  oil  and interfacial
               tension between oil and water,  dynes/cm,
               respectively

       9,,92 = instantaneous contact angles formed by the oil
               phase at the air and water  boundaries, degrees,
               respectively

Oil  slick thickness is smallest along the  periphery of slicks;
the effect of surface tension is significant here because the
contact angles are nearly zero.  Actually, the contact angles
decrease with time as the oil spreads outwards.   Since the value
of cos 9 approaches 1 as 9 approaches zero degrees, Equation (61)
can be written as
Depending on the value of a, surface tension forces may accelerate
or retard spreading.

     A positive value for a speeds spreading,  while a negative
value has the opposite effect.   The net surface tension divided
by the area of the slick gives  the surface tension force per unit
volume of oil (F )
                           Fa = k^/A                        (63)
where
     a = net surface tension,  dynes/cm
                              2
     A = area of the slick,  cm

    k, = proportionality factor

The factor k-j was introduced to account for the variations  of the

contact angles and the surface and interfacial  tensions  with  time.

     The spreading force due to surface tension may not  be  uniform
along the free surfaces of the oil slick.   Variation of  surface
tension along the periphery  of oil slicks  are partially  responsi-
ble for unsymmetric spreading  patterns.

Gravity--
     The effect of gravity on  an oil  slick can  be evaluated by
considering the two components of the gravity vector,  g   (in  the
                                                       X
direction of spreading) and  g   (normal  to  the spreading),  as



                                93

-------
shown in Figure 13.   The component g  of the gravity force acts
vertically downward  and produces only hydrostatic pressure.   This
pressure increases with distance measured downward from the free
surface of the oil slick.   Since the oil slick is in hydrostatic
equilibrium,  a hydrostatic balance on the oil  is
                           37 • -Vo                        (64)
where
                                                       2
P = the hydrostatic pressure on the oil,  gm/cm»sec
    vertic
    cm/sec
    g  = vertical  component of acceleration due to gravity
     j
    p  = density of oil, gm/cm
     y = vertical  coordinate axis, cm
Integrating Equation (64) over the thickness of the oil  layer
above the water surface yields the magnitude of the hydrostatic
pressure difference (AP).
                       P = 9yP0Ah = k2gpoAh                  (65)
where
     k2 = constant of proportionality
     Ah = thickness of the oil layer above the mean water surface,
          cm
To eliminate Ah from Equation (65), a buoyancy equation  is used.v
Since the oil  floats on water, the equation of buoyancy  is
                        gpw(h-Ah) = gP()h                     (66)
where
     Ah = thickness of the oil above mean water level,  cm
      h = mean oil slick thickness, cm
  p ,p  = densities of water and oil, gm/cm ,  respectively
   W  0
                                            2
      g = gravitational acceleration, cm/sec
Solving for Ah gives
                               94

-------
                         Ah = (1  -   )h                      (67)


Substituting Equation (67) in Equation (65)  yields

                        P = k29P00-P0/Pw)h                   (68)

The gravity force per unit volume of oil  which corresponds to the
pressure force is


                      Fg ' "Z9P0<1-P0/P,,>T-        •          <69'

Eliminating h in Equation (69),  using Equation (60),  gives


                      % • "2gp0


This pressure force is caused by  the downward-acting  force of
gravity, tending to decrease the  height of the oil.

     The component g  of the acceleration due to  gravity generates
                    A
flow in the x-direction, also.   The magnitude of  this force is
                        Fg  = gxPQh(x)6x                     (71)
                          /\
where
      g  = the x-direction component of the acceleration due to
       x                  2
           gravity,  cm/sec

    h(x) = variation of oil slick thickness along the x-direction,
           cm

      <5x = unit width along the x-axis, cm

Since g  is much smaller than g  the influence of F   on the
                                                    /\
spreading oil is negligible.

Viscous--
     At the water surface, the dynamics of the spreading process
depends on the magnitude of the viscous forces at the air and
water interfaces.  The air at the top and the water layer in
close proximity to the oil, i.e.  the water-boundary layer,  exert
viscous drags on the oil.   In the absence of turbulent wind, the
                               95

-------
viscous drag at the air interface is insignificant.  The shear
stress acting on the oil  slick at the oil/water interface is the
viscous force which impedes oil  spreading.   The magnitude of this
viscous force per unit volume of oil, F ,  is

                            Fv = TA/V                        (72)


where

     T = shear stress exerted by the water  column on the oil
                         2
         slick, gm/cm»sec

For Newtonian fluids, there is a linear relation between shear
stress and strain rate or gradient in velocity


                            T -  y0 &                        (73)
where

     y  = viscosity of oil, gm/cm«sec


     -rr- = velocity gradient in the oil  in the direction of oil
     an                 _,
          thickness, sec

If the velocity profile in  the oil is assumed to be linear in the
vertical  direction, then

                          *J»'   \/      \i                      ,
                             'l-^l                      <">
v - 4                          (75)
and


                              W — 	
where

     k3 = proportionality constant

      v = spreading velocity, cm/sec

      I - characteristic length in the direction of spreading

          (A = t2), cm

      t = time of spreading, sec

      h = mean oil  slick thickness, cm
                               96

-------
Equation (74) is substituted into (73) to give


                              k3P0A"2
Substituting Equation (76) into (72) yields


                              W1/2  A
                         Fv = ^Ih -- T                    <77>

When h is eliminated in Equation (77) by using Equation (60), the
viscous force becomes
                                                             (78)
Inertia--
     During the early stages, i.e.  the initiation of spreading,
the force of inertia is important.   As a slick accelerates from
rest, the motion is retarded by force of inertia.  The magnitude
of this force, F.. ,  can be approximately determined by a momentum

balance in the direction of motion  of the slick

                           FT = K4P0a                        (79)

where

     PQ = mass density of the oil,  gm/cm
                                         o
      a = acceleration of the oil,  cm/sec

     k^ = proportionality constant

The acceleration of the oil can be  written as


                             a = -4                          (80)
                                  *
where
     I = characteristic length of the oil  slick,  equals  (A)1/2,
         cm

     t = time of spreading,  sec
                               97

-------
The inertia force becomes


                        Fi = k4P0Al/2/t2                     (81)

Equations of Spreading

     The equation of spreading is derived by equating the forces
which cause and oppose spreading.  These equations are (63), (70),
(78), and (81).  Thus, the spreading equation is

                          F +F  = F +F.                      (82)
                           a  g    vi                      v   '

                                      R/?        1 17
                             2   k u A      k p A
          •   (7   i     /    . \ V     3 o        4 o             / n *\ \
           1 A + k29Po^" p)~3 = 	2	 + 	2	         ^   '


where

     Ap = Pn/P » ratio of oil to water densities
           0  W

Algebraic manipulation of Equation (83) to solve for A as a func-
tion of the other terms  is not practical.  Following the method
of Fay (1969,1971), each of the spreading forces can be equated
to the retarding forces  and the expression is solved for A to
give several spreading regimes.

     For F  = F  (surface-tension/viscous spreading)

kl
a 3o
*" )
A = k5
/2t
yo
                                       2/7                   (85,
            kl  2/7
where k  =
Fo
  r F  = F. (surface-tension/inertia spreading)
                          1
                            g .                              ,86)
                              -- - -                     (86)
                               98

-------
                         A = k,
                                at'
        2/3
                                 (87)
where k,. = (-r-i-)
       b    k4

For F  = FV (gravity/viscous spreading)
                                         5/2
                    kgp
                                 (88)
                    A = k.
                           gp
            2/11
                                 (89)
where k? = (^-)
               2/11
For F  = F.. (gravity/inertia spreading)
                                2   k.p A
                                         1/2
                                                             (90)
                     A = k<
g(l-Ap)V2t2
                                        2/7
(91)
where k
               2/7
       8
     Equations (85), (87), (89) and (91) are functions of only
the physical  properties of oil  and water, volume of oil and
elapsed time.   These equations  relate the area! extent of oil
slicks spreading on calm water  to elapsed time.  It is interesting
to find that  the equation for the surface-tension/inertia stage
is independent of the volume of oil spilled.  As such, this equa-
tion is not expected to be a good predictor of the area of a
slick.  The volume of oil spilled is related to the duration of
the spill  and  the time of spreading via Equations (58) and (59).
MECHANISMS OF CHEMICAL DISPERSION OF OIL SLICKS

     The importance of dealing with oil spills by diversion^
containment and collection of floating oil  cannot be underempha-
sized.   These methods are not practical in  all spill situations,
                               99

-------
for several reasons cited earlier.  For example, there is a sea-
state threshold beyond which containment and recovery of oil
slicks are impossible.  Dispersion of oil  slicks by chemical
treatment may be the only option available to reduce adverse
impacts on the environment.   The economics of this oil spill
cleanup method are favorable when aerial spraying can be used and
when mixing energy can be provided by winds, waves, tides and
currents.

     Previous investigations of chemical dispersion of oil  slicks
were limited to laboratory tests of the effectiveness of disper-
sants and toxicity to marine life forms.  Thus, no phenomenologi-
cal theories are available to explain the  mechanisms of dispersant
action on oil slicks during  cleanup operations.

     In this section, the mass transfer processes that lead to
dispersion of oil  slicks in  water are identified.  Primary
mechanistic steps  are distinguished from those that are ancillary.
Mathematical equations are presented to quantify the key
mechanisms.  The effect of introducing mixing energy into oil/
water dispersant systems by  using mechanical devices or the action
of turbulence generated during high sea states will be discussed
in qualitative terms.  The limitations of  the model will be
apparent as consequences of  the assumptions made as the model is
developed.

Mechanistic Steps

     The mechanisms of chemical dispersion are poorly understood.
An attempt to explain dispersion of oil slicks with chemical
agents was made by Canevari  (1969a,b).  He proposed three mechan-
istic steps:

     a)  diffusion of dispersant molecules through the oil  layer;
     b)  incorporation of oil globules into micelles; and
     c)  diffusion of micelles into the underlying column of
         water where they become stranded.

Although these mechanisms are useful in providing a general
description of dispersion processes, they  do not include all the
mass transfer processes that contribute to chemical dispersion.
Also, it may be that none of the mechanisms cited is the rate-
determining step.   Canevari  (1969b) described the mode of action
of the self-mix dispersants  by an analogy  to the "diffusion and
stranding" mechanism that occurs in spontaneous emulsification
processes (Davies  and Rideal, 1963).

     Chan et al. (1976) used detergency theory to explain the
mechanism of solubilization  in a detergent-saturated solution.
Similarly, Prudich and Henry (1978) discussed their experimental
data for the transfer of hydrophilic-coated-mineral-matter  parti-
cles from a hydrocarbon phase to an aqueous phase using concepts of

                               100

-------
       partial detergency.  Drawing from the efforts of these investiga-
       tors, concepts in colloid chemistry and homogeneous catalysis, an
       attempt will be made to create a general description of oil slick
       dispersion using chemical dispersants.

            Dispersion practice varies and depends on several factors,
       e.g. size of oil slick, accessibility of the oil spill site, etc.
       The case considered here is as follows.  Dispersant is applied
       aerially in a fine mist, such that the droplets settle gently on
       the oil layer.  The impact velocity of the spray is negligible
       and the dispersant does not penetrate the oil; it forms a uniform
       film at the surface of the oil slick.

            In the absence of body forces in the water column (i.e.,
       prior to mixing of the dispersant/oiI/water system) and with
       sufficient contact between dispersant and oil, the processes that
       cause transfer of oily material from the oil layer to the aqueous
       phase are:

            a)  diffusion of dispersant molecules (and micelles) from
                dispersant solution to dispersant/oil (d/o) interfaces;
            b)  distribution of molecules at d/o interfaces;
            c)  saturation of d/o interfaces;
            d)  diffusion of molecules through the oil layer to oil/water
                (o/w) interfaces;
            e)  distribution of molecules at o/w interfaces;
            f)  adsorption of micelles at o/w interfaces;
            g)  formation of mixed micelles and interfacial complexes at
                o/w interfaces;
            h)  desorption of mixed micelles from o/w interfaces;
            i)  diffusion of mixed micelles into the bulk aqueous phase;
            j)  dissolution/accommodation/solubilization of oil particles
                in aqueous phase; and
            k)  diffusion of free molecules back to o/w interfaces or
                away from the zone of contamination.

       These mechanisms give a comprehensive picture of chemical disper-
       sion processes.  The fundamental dispersion processes are (f),
       (g), (h), (i) and (j); these steps are important for anionic,
       cationic and nonionic dispersants.  The other steps are
       auxilliary.  Steps (a) to (e) can be lumped together and consider-
       ed as a diffusional step across the oil layer.  Figure 14 shows
       the mass transfer processes that are important in dispersion
       processes.

            The relative importance of each of these steps varies,
       depending on the situation.  In some situations, some of the
       secondary processes can become important.  The importance of
       individual steps or processes depends on the structure of the
       surface-active agent, the oil, the solvent base used to formulate
       the dispersant, and the concentration and dosage of the surfac-
"**"""     tant.  For instance, hydrocarbons are generally miscible with
%**>

                                      101

-------
                               a>
102

-------
other hydrocarbons, because of the near zero Gibbs free energy of
mixing.   Thus, dispersants formulated in hydrocarbon solvents
will  diffuse through oil  slicks faster than those that are
aqueous-based.

Formulation of Model Equations

     Before mathematical  equations are derived for the fundamen-
tal processes, a brief discussion of micellization is necessary
to an understanding of chemical dispersion of oil  slicks.

     The behavior of a surface-active agent depends on its state
of solution.  Generally,  dispersion processes are favored  if the
surfactant is present in  the solvent solution as micelles  rather
than  as  single ions or molecules.  When sufficiently concentrated,
dispersant solutions contain micelles.  These aggregates are
considered to be thermodynamically stable and contain from 10 to
more  than 100 surfactant  molecules.  For example, sodium lauryl
sulfate  (CH3(CH2)1-|-NaS04) forms micelles consisting of 62 mole-

cules (Elworthy et al.,  1968).
     The size of a micelle is determined by the structure;  micelle
                               o
diameters range from 40 to 100 A.   The concentration of surfactant
at which micelles become significant,  i.e.  the critical micelle
concentration (CMC), is related to the number of carbon atoms  in
a straight chain dispersant molecule by
where
                         log CMC = a-bm
      m = the number of carbon atoms
(92)
    a,b = constants

Nonionic surfactants form micelles at lower concentrations  than
ionic types, at the same hydrocarbon chain length.   Since
commercially available dispersant formulations may  consist  of
more than one type of surfactant molecules, the CMC of a  mixture
will lie between the values of the components.

     The law of mass action is usually applied to  the association
of single molecules to form aggregates and complex  molecules.   A
basic assumption is that equilibrium exists between single  mole-
cules and micellar aggregates.  For non-ionic surfactants,  the
reversible process is
                           mC.
                                r

                              103
                                    m
(93)

-------
If the activities are equated to concentrations,  the equilibrium
constant, K ,  for the micellization of non- ionic  dispersants is
                               k*   Cm
                          Kn = r = :£                       (»«)
                                r   C.
where
    k,r,k  = rate constants for the forward and reverse reactions,
               -1
            sec  ,  respectively
                                                               3
       C  = mass concentration of aggregates or micelles,  gm/cm
       C. = mass concentration of single surfactant molecules,
                 3
            gm/cm
        m = the association number for the surfactant
For ionic surfactants, the reversible process is
                                    kf
                       mC.+(m-p)Cc — b Cm                    (95)
and
                            k*      Cm
                               -     m
                            _       .
                        i    k    rmr(m-p)
                             r   LiLc
where
     C  = mass concentration of counter-ions,  gm/cm
      p = number of counter-ions not attached  to the micelle
The degree of ionization of the ionic  surfactant is p divided  by
m.
     If the structure of a dispersant  is known,  it is possible
to determine m.   For dispersants with  a  linear hydrocarbon
compound, m equals the number of -CH^- groups.   Unfortunately,
information on the structure and types of  surface active agents
in commercial dispersant formulations  is proprietary, and m is
unknown for commercial dispersants.
     Since micelles in solutions of  ionic  surface-active agents
are charged, they repel  each other.   According to Hartley (1976),
                               104

-------
the distance between charged micelles is

                         d3 = _8l_ (j)r3                        (97)
                              3/2

where

     d  = distance between the centers of the micelles, A°

     <(>  = weight fraction of the total volume occupied by
          micelles

     r  = micelle radius (assumed spherical), A°

The formation of micelles in dispersant solutions influences the
rate of chemical dispersion.  The diffusion of surface-active
materials to various interfaces and transport of oil  droplets
into bulk aqueous phase are affected by the degree of micelliza-
tion.  The total quantity of surfactant available for dispersion
is increased when surface-active agents are present in crude
oils, petroleum products and surface layers of seawater prior to
the application of dispersants to oil slicks.

Diffusion of Micelles to Interfaces--
     Micelles must diffuse through the dispersant solution and
the oil layer to reach oil/water interfaces.   A high  concentra-
tion of surface-active material in the dispersant solution is
favorable to high diffusive flux.  Dispersants formulated in
hydrocarbon solvents approach molecular solution and  diffuse
rapidly through the oil layer.  Dispersants formulated in water
must undergo two diffusion steps; through both aqueous solvent
and oil layers.  Because of these two diffusion steps, the former
dispersants may be slightly more efficient than those of the
latter type.

     Since micelles and unaggregated molecules are in equilibrium
in dispersant solutions, both species diffuse but the flux of
micelles will be less than for single molecules.  The rate of a
diffusion process may be expressed by Fick's  First Law.  For
simple diffusion, Fick's Law can be written for micelles and
sing!e molecu!es .


                            m     m  3x



                           F  -  D  ^1


where
                              105

-------
                                                           2
F ,F. = mass fluxes of micelles and single molecules, gm/cm -sec,
        respectively

Dm»D,- = diffusion coefficients of micelles and dispersant mole-
 IT]  1            e*
        cules, cm /sec, respectively

ar   a r
  m    i
-T— , -r— = local concentration gradients in the direction of
 a X    a X                    «
           diffusion, gm/cm

The combined diffusive flux (F.) for both species is


                           Fd = Fm+Fi                       (100)

The actual Fd will be smaller than the sum of the individual

fluxes because of interactions between the diffusing species and
the effect of other intermolecular forces.

     Diffusion coefficients of micelles and single molecules can
be estimated from the simple Stokes-Einstein Equation

                                 NDT



where
     —                            2
     D = diffusion coefficient, cm /sec

    Ng = Boltzmann's constant, dynes cm/°K

     T = absolute temperature, °K
                                           2
     y = viscosity of solvent, dynes-sec/cm

     r = mean hydrodynamic radius of the micelle or molecule,  cm

The diffusion coefficient for single molecules will  be greater
than that for a micelle.

     Ward and Tordai (1946) derived an expression for the rate of
diffusion of molecules to interfaces.  They applied the penetra-
tion theory of mass transfer across a liquid-liquid interface  and
derived the following equation for the rate of diffusion


                      dji _ /_D_J/2 /  A > o                 ,., nn\
                      dt  " Wt'

where
                               106

-------
     n = the number of molecules arriving at a unit cross section-
                                               2
         al  area of the interface, molecules/cm
     D = diffusion coefficient of the molecule through the medium,
           2
         cm /sec

     t = time, sec
                                23
    N, = Avogadro's number, 6x10   molecu!es/gm-moles

     C = concentration of surface-active molecules diffusing from
         the surface, gm-moles/£

Equation (102) assumes

     a)  molecules diffuse at a uniform velocity;
     b)  there is no back-diffusion; and
     c)  there is no energy barrier, mechanical  mixing or thermal
         convention in the medium

When an energy barrier opposes the process,  a lower  rate of
diffusion will result.  This equation is applicable  to the diffu-
sion of single molecules, micelles, and mixed micelles.

     Equation (102) can be integrated to give

                               i/o  N
                          g/Dt J/2/ _A  N r                   Mm^


The diffusion of surface-active agents  from  the  bulk oil layer to
o/w interfaces will be rapid because it involves molecular diffu-
sion and small oil slick thicknesses.

     At the interface, micelles dissociate  into  single molecules
prior to adsorption and when sufficient molecules have been
adsorbed, some molecules "react" with oil  particles  to form mixed
micelles that desorb.  The proposed reaction between m molecules
and oil to form a mixed micelle is
                         mC.j+0 	 CmO

     The assumed sequence of reversible steps is

                              k.
     (1)                Cn.+S = (CjS)     (Adsorption)
                               107

-------
                              k2
     (2)            m(C.S)+0 = (CmOSm)    (Reaction)
                       I      |x     m  Ml

                             k
                              3
     (3)            (Cm°Sm) == cm°+mS     (Desorption)
                            k-3
where
       C. = concentration of single molecules
      CmO = concentration of mixed micelles
        S = an adsorption site
        m = number of adsorption sites
    (C.jS) = adsorbed single molecule
  (C OS ) = adsorbed mixed micelles
        0 = concentration of oil
The individual steps are  next considered  in greater detail.
Adsorption of Dispersant  Molecules--
     The rate of adsorption of surface-active molecules at o/w
interfaces is governed by the rate of diffusion  from the oil
layer.   Since single molecules and micelles diffuse together,
they arrive jointly at o/w interfaces where they adsorb as single
molecules.
     From step 1, the rate of adsorption  is
                       r*   =   C.S-kCC.S)                   (104)
where
       C. = concentration of surfactant molecules
        S = an adsorption site
    (C.S) = concentration of adsorbed molecules
   k-, , k -, = adsorption coefficients for the forward and reverse
            reactions
                               108

-------
       All  concentration  terms  refer  to  concentrations  at  the  interface.
"*****     The  adsorption  coefficients  follow  Arrhenius'  Law

                               k  =  kQ exp(-E/RT)                    (105)

       where

            k   =  frequency  of adsorption,  sec"

             E  =  energy  of  adsorption,  cal/gm-mole

             R  =  gas  constant,  cal/gm-mole»K

             T  =  absolute temperature,  K

       At  equilibrium, the  rates  of the  forward  and  reverse  reactions
       are  equal

                               SklC1  =  k.^C.S)                     (106)

       and

                                     k,    (C.S)
                               Kl  = TT7=-rc—                     (107)


       where  K-j  is  the equilibrium  adsorption constant.

            Surfactant molecules  are  held  at  the  interface  by  molecular
       forces.   They  are  oriented  with  hydrophobic groups  anchored  to
       the  oil  phase  and  hydrophilic  groups  immersed  in the  aqueous
       phase.

       Formation  of Mixed Micelles  and  Interfacial Complexes--
            When  a  sufficient number  of  dispersant molecules  have been
       adsorbed at  the interface,  a "reaction" occurs  between  a  number
       "m"  of absorbed molecules  which  are close  neighbors  and micro oil
       particles  to form  mixed  micelles  and/or interfacial  complexes.
       The  stoichiometry  of this  "reaction"  is not known.   The reaction
       may  be a simple coexistence  of oil  and dispersant molecules  as
       the  oil  particles  are merely incorporated  into  the  hydrophobic
       core of  the  micelles.  The  formation  of mixed  micelles  is an
       important  mechanism  in the  chemical  dispersion  of oil.  If the
       reaction is  assumed  to be  bimolecular, then the  rate  of the
       reaction can be written  as


                           r2 = k20(CiS)m-k_2(CmOSm)                (108)


       where
                                      1 09

-------
      0 = amount of oil  particle incorporated into the micellar
          core

(C OS ) = surface concentration of adsorbed mixed micelles
 k2'k-2 = coefficients of the forward and reverse reactions

Free oil  is always in excess at the interface and Equation
(108) can be simplified to


                     r2 = k2^CiS^m"k-2^Cm°Sm^               009)
       *
where k~  = k20

The equilibrium constant for the reaction is


                       K  = £L_ . ^m^mj                   (11Q)
                        2   k-2    (C.S)m


This reaction will not occur if the number of molecules adsorbed
on adjacent sites is less than the association number of the
surface-active compound.

Desorption of Mixed Micelles--
     The  desorption of mixed micelles from the interface exposes
additional surfaces for adsorption of molecules arriving at the
interface by diffusion from the oil  layer.  In this  process, m
adsorption sites become available.  The rate of desorption
according to step 3 is

                     v  = \t (r rK \-\(  <;mr n                M11 ^
                     r3   k3(Cm°V  k-3S Cm°                U'V

where

  (CmOSm) = concentration of adsorbed mixed micelles

      CmO = concentration of mixed micelles at the interface

       Sm = "m" number of adsorption sites which is  equal  to mS

   k3,k_3 = desorption coefficients


At equilibrium, the desorption constant K, is
                             k      SmCO
                               no
                                                             -, i 7 \
                                                                '

-------
As the mixed micelles desorb from the interface,  they take with
them oil  particles that are occluded in their core regions.

     The coefficients of the reaction and desorption steps follow
Arrhenius1 Law, also.  Davies and Rideal  (1963) present equations
for estimating the desorption energies (E)  of some nonionic and
ionic surfactants for a few interfaces in which the bottom phase
is water.  For nonionic and ionic surfactants, the equations are

                            E = A +mw                       (113)

and

                         E = Ap+mw-z^^                     (114)


where

    A  = energy of polarization of the unionized  polar group,
     p   cal/g-mole
m
       = the number of -CH2- groups
     w = van der Waals adsorption energy associated with each
         -CHg- group, cal/gm-mole

    z-| = valency of the surface-active ion

     e = electron charge

    ^  = electrical  potential  of the surface

Comparison of Equations (113)  and (114)  shows that the desorption
energies of nonionic surfactants are usually greater than those  *•
of ionic agents of the same chain length.   Davies and Rideal
(1963) give values for ImM of  sodium laurly sulfate for a para-
ffinic oil/water interface, with no salt added,  as

                  Ap = +160 (20C),  +1800 (50C)

                   w = +810

               z-|£^0 = -5000


In the presence of salt, the value  of z-|e^0 increases.

Diffusion of Mixed Micelles into the Aqueous Phase--
     As mixed micelles desorb  from  the interfaces, the layer
directly below the interface becomes saturated  with mixed
micelles.  The concentration gradient that builds up between  this
layer and the bulk aqueous phase provides  the driving force  that


                              111

-------
causes  mixed micelles to diffuse away from the
This diffusion step is similar to the diffusion
sant molecules and micelles.   Equation (102) is
this mass transfer.
                                         surface layer.
                                         of single disper-
                                         applicable to
Dissolution/Accommodation/Solubilization of Oil  in the Aqueous
Phase--
     The transfer of oil  into  the underlying column of water is
the ultimate goal of chemical  dispersion.   During the initial
stages  of the dispersion  process,  the concentration of oil  in  the
aqueous phase is nearly zero.   Then, as the mixed micelles  diffuse
into water, dissolution of the oil  in water occurs.  The first
micelles release their oil particles into  water,  where the  oil
goes into molecular solution.   The mixed micelles become unaggre-
gate molecules and, if the oil/water system is static, these
molecules diffuse back to the  interface.  In this case,  there  is
no net  loss of dispersant; but,  in  flow systems,  the dispersant
molecules may be flushed  away  from the zone of contamination.

     As the diffusion of  mixed micelles into the  water layer
continues, the concentration of oil  increases.  In time, oil in
water will be present as  molecularly dissolved,  accommodated and
solubilized species.  These forms  coexist  in water until the
system  reaches "supersaturation."   At this point, microemulsions
are formed.  These coalesce into  larger particles -that can  become
buoyant, rise and recoalesce with  the surface slick.  These
processes occur only in chemically treated and static oil/water
systems when the dosage of dispersant is large enough
accommodation/solubilization/microemu 1 si on phenomena.
heavily contaminated waters, these phenomena will not
in flow systems as dilution of oil  in solution occurs
by inflowing fresh water.
                                               to  lead to
                                                Except in
                                               take place
                                               continuously
     The combined rate of these processes is derived from a
material balance on the oil  in solution.   The rate of change of
oil concentration in the aqueous phase is equal  to the rate  of
molecular diffusion of mixed micelles from the interface which is
assumed to be saturated.  It follows that
                          R =
                       KL(CmO-C)
(115)
where
     K,  = overall  liquid mass transfer coefficient
    V
   concentration of mixed micelles at the o/w interfaces
   on the water side

= average concentration of oil  in  the aqueous phase
                               112

-------
The concentration of oil  in water consists of oily material  from
the slick as well as hydrophobic portions of mixed micelles  in
solution.  Under field conditions, some oil  may sediment after
adsorbing onto solid particles in suspension.

The Rate-Determining Step

     The concept of a rate-limiting step is  useful when consider-
ing a sequence of several steps.  If a step  is assumed to be the
slowest step,then its rate controls the rate of the overall
reaction.  All steps except the limiting step are assumed to be
at steady-state or in equilibrium and the overall rate is derived
in terms of the slowest step and relevant parameters.

     The diffusion steps  are probably not rate-limiting as some
mixing is usually available to chemically treated oil/water
systems.  The effect of mixing on oil/water/dispersant systems
will  be discussed later.   Thus, the rate-determining steps are
most likely adsorption, reaction or desorption.

Adsorption Control Model--
     Adsorption is a surface phenomenon and  is important in  homo-
geneous and heterogeneous catalysis.  Several models have been
proposed in the literature for the kinetics  of adsorption.  The
model used here is based  on the treatment by Langmuir-Hinshelwood-
Hougen-Watson (Carberry,  1976).  This treatment assumes
i)  monolayer coverage, ii)  no interactions between adsorbed
molecules, and iii)  homogeneous surfaces and uniformly energetic
adsorption sites.  By mass-action, the rate  of adsorption is
proportional to the concentration of the adsorbent and the
fraction of the surface that is unoccupied (0)

                          rA = k^.d-6)                    (116)


where k, is the adsorption coefficient.

     The total concentration of adsorption sites includes those
either vacant or containing adsorbed species.  It follows that

                      SQ  = s+(CiS)+(cm°Sm)                  (117)

where  S  = total concentration of adsorption sites

        S = adsorption sites that are vacant.

The equilibrium constants are used to eliminate the adsorbed
species in Equation (117).  From Equation (107), (C.S) is equal
to                                                 ^

                          (C.S) = SK]Ci                     (118)
                               113

-------
Since all sites are identical and if all surfactant molecules are
similar, Equation (118) can be written as


                         (C.S)m = mK.C^.S                    (119)


and


                                (C.S)m

                          mKl =   C.S                        (120)


The product of Equations (110) and (118) is solved for (C OS ) to
give                                                     m  m


                       (CmOSm) = mSK1K2Ci                   (121)


If Equations (118) and (121) are substituted into (117), the
result is
                                      ,K2Ci                 (122)

Equation (122) can be written as


                      ^~ = TT	r~~^	T                 (123)


Since Equation (123) is equal to (1-e),  i.e. the fraction of the
interface that is unoccupied, the expression for the adsorption-
control  model, Equation (116), becomes

                                 k,C.

                      rA = (i+K C.+mK K  C.)                 (124)
The interfacial  concentration term C. cannot be measured directly.
The procedure to relate C..  to CmO is as follows.  First, Equations

(110), (112) and (120) are  multiplied; the result is
                                    SmCmO                   (125)
Using the relation mS = Sm and solving for C.  gives
                                  Cm°

                           CT  = K K K                       (126)
                            1    KKK
                               114

-------
Substituting Equation (126) into (124) and simplifying the result
shows that

                                K.C 0
                                                            (127)
                       A   d+KBCmO+KcCmO)
where  KA = k1/K]K2K3
       Kg = 1/K2K3
The interfacial concentration term (CO) cannot be measured

directly, also; it must be related to the concentration of oil in
the bulk water phase through Equation (115).

Interfacial Reaction Control Model--
     The rate of reaction is


                    rR - k2(C.S)m-k_2(CniOSm)                d28)

Substituting for the adsorbed species using Equations (120) and
(121) gives

                    rR = mk2K1SCi-mk_2SCmO/K3               (129)


where Sm has been replaced with mS.  Equation (129) is simplified
to give
Substituting for S using Equation (123), results in

                       [mk2K1C.-mk_2CmO/K33S0



If, however, C. is replaced in Equation (131) by using Equation

(126), the result simplifies to

                       (VKeqK3-k.2/K3)mS0CmO
                   i  ~    Aiii/ /•*"" rT_i_ \/ r> A A                   \ I *5 c


where
                              11 5

-------
            KB  -  l/KeqK3


            Kc  =  m/K3

           Ke   =  overall  equilibrium  constant

               =  KO/  - 2

            Equation  (132)  is  the  overall  rate  of  dispersion  when  the
       formation  of micelles  is  the controlling step.

       Desorption-Control  Model --
            In  Langmurian  terms,  the  rate  of desorption  is  assumed to  be
       proportional to  the  concentration of the desorbing  specie(s)  or
       the  fraction of  the  surface occupied, 9.  Thus,  the  rate  of
       desorption is


                             rD  '  MCm°Sm> -  k39                   <133>

       Substituting for  (CmOSm)  using  Equation  (121)  gives

                                rD =  mk3K]K2SCi                     (134)

       If S  is  replaced  in  Equation (134)  with  Equation  (123), the
       result  it
                              D
       Substituting  for  C.  using  Equation  (126)  and  simplifying  gives

                                       KnCmO
                                        D  m
                              D  "  (1+KB

       where

           KD  =  mk_3SQ


             B  =     2  3

           Kp  =  m/ Ko


       This equation  is  similar  to  that  derived  for  the  case  when  the
       rate of  adsorption  is  the  limiting  step.  Also,  Equation  (136)
       is  related  to  the  concentration of oil  in  the  bulk  fluid through
       Equation (115).

**»»

                                     116

-------
     When mixed micelles desorb from the interface,  surfactant
molecules are lost from the surface areas.   Desorption of mixed
micelles exposes additional areas at the interface.   Thus, space
is available for adsorption of surface-active molecules reaching
the interface by diffusion from the bulk oil  or water column.

     The overall rate of dispersion of chemically treated oil
spills will  vary according to the mechanism(s) which is(are)
rate-limiting.   Field conditions will  determine the  mechanism
which is the slowest step.

The Effect of Mixing

     Mixing is  usually supplied to oil slicks after  they are
treated by dispersants by mechanical  means  and/or water motions
induced by the  action of winds, waves, tides, and currents.   The
hydrodynamics of the system will vary according to the origin  and
intensity of mixing.

     The amount of work required to remove  completely a unit of
oil from the oil layer is called the work of  detergency.  The
magnitude of this work for complete detergency is
                          W .  = a +a -o"
                           d     w  o  ow
                                                    (137)
where
       W
cl = work of detergency,  dynes/cm
aw'CTo'aow
    surface tensions of water and oil,  and interfacial
    tension between oil and water, dynes/cm,  respectively
The presence of surfactants at oil/water interfaces causes reduc-
tion in interfacial  tension between oil  and water and leads to
lower values of Wd<   Also, a high energy barrier must be overcome

for mixed micelles to form and desorb from oil/water interfaces;
therefore, the system must be supplied with additional  energy for
complete detergency.   When the work of detergency is equal to or
less than the work input into the system,  complete dispersion may
occur.  If the
be complete.
       energy input is less than W,,  dispersion  will  not
     It is important to be able to estimate the amount of mixing
energy required to achieve a specific degree of emulsification.
Unfortunately, one of the difficulties in effective dispersant
treatment of oil  spills is the inability to estimate  the fraction
of energy input available to aid detergency.  In the  field,  the
largest part of the energy introduced into the system is dissipa-
ted into the water column.  Inadequate mixing has adverse effects
on chemical  dispersion as dispersant can be transported through
the water phase before mixed micelles are formed.  Therefore,  it
                              117

-------
is not surprising that greater efficiencies are usually reported
for small  scale laboratory tests of the effectiveness of disper-
sants, in  which the energy input per unit volume of fluid is
large.  It is necessary to allow the oil  and dispersant to make
sufficient contact prior to agitating the system.

     The rate of input of mechanical energy is important as it
influences the total  area of oil/water interfaces  available for
the mass transfer processes discussed earlier.  Additional oil/
water interfaces are created as the oil layer is physically
sheared into droplets according to the following equation

                           A = Vow                       "38)

where

     W.  =  amount of energy required to create the  interfaces


Stability  of the oil  droplets is a function of mechanical  energy
input and  other factors.  If the intensity of agitation is low,
oil droplets that are formed will  be coarse and unstable.
Unstable phases will  tend to remain dispersed only in the
presence of turbulence.  In the absence of turbulence,  or after
these phases have migrated to regions of low shear, coalescence
is possible; these droplets may coalesce with the  parent slick or
form separate patches.

     Fine  oil droplets will form if a system is agitated suffi-
ciently; these droplets are stable and resist coalescence.
Weathering processes, turbulent diffusion, waves,  currents and
tides, and Brownian diffusion transport oil droplets through the
water column to points far from the source of the  spill.  The
movement of dispersed oil droplets in water depends on  hydro-
static and hydrodynamic forces and the sizes of the droplets.


NUMERICAL  ANALYSIS OF DISSOLUTION  AND SPREADING EQUATIONS USING
EXPERIMENTAL DATA

     The dissolution model was tested with experimental data.
The values of four parameters in the segmented model are unknown,
and must be estimated for any oil.  These parameters are GS, tm>
K., and K£.  For each oil tested,  a value for the  joint point,

t  , was estimated from the experimental data.  Since t   corres-

ponds to the time when oil concentration is maximum, the data in
columns 2  (Tables 6 and 7) served  as input values  for t  during

the regression analysis.  Similarly, C$ corresponds to  oil concen-

trations at saturation.  The input values for this parameter are
listed in  Table 8.  The initial estimates for K.  and Kr were unity,

                              118

-------
     The equations of spreading are functions of the volume of
oil, the duration of the spill, physical  properties of oil  and
water phases, and time of spreading.   In  each equation, only the
value of one empirical constant must  be determined from experi-
mental  data.  Iteration was started with  unity as the initial
estimate of the constants in the four spreading equations.

     Least-squares regression analysis of the dissolution func-
tions and spreading equations was performed on an IBM 360/370
computer using the SAS (Statistical Analysis System).  SAS  is a
commercial  statistical package that is widely used and highly
documented  (SAS Institute,  1979).  Minimum programming effort is
required and the package has been tested  adequately.  The
Marquardt method was used to estimate unknown parameters.  This
method is available as an option in the nonlinear parameter esti-
mation program (NLIN) of the SAS package.  NLIN regresses the
residuals on the partial derivatives  of the functions with  respect
to the parameters until the iterations converge.  The Marquardt
procedure is an extension of the Gauss-Newton and Steepest  Descent
methods and can converge with relatively  poor starting guesses
for the unknown parameters  (Marquardt, 1963).  The least squares
objective function which is minimized is  the sum of the squares
of the residuals between the predicted values and experimental
data.  The  built-in convergence criterion in the NLIN program has
              Q
a value of  10  .

     The listings of sample computer  programs, used to run  the
regression  analyses and estimate unknown  parameters in the  disso-
lution and  spreading models are given in  Tables Fl and F2 of
Appendix F.
                              119

-------
                            SECTION 9

                      RESULTS AND DISCUSSION
DISSOLUTION
     The rates of dissolution in water of crude oils and
processed petroleum have been investigated in open static tests.
The tests correspond to the worst condition of an oil  spill,  e.g.
oil completely covering the surface of a small  lake.  Oils were
equilibrated with water at 25C for periods of 2 to 3 weeks and
oil concentrations in solution were measured routinely by infra-
red spectrophotometry.

Tap Water Studies

     In studies using tap water as the test fluid, 9 crude oils,
2 processed oils and an oil mixture containing 8% crude and the
remaining processed oil were investigated.  Experimental data
for oil concentrations  and equilibration time have been presented
in Appendix B.

     The results of solubility measurements suggest the old adage
"that oil does not mix  with water" is false.  The concentration
profiles for all oils increased gradually during the first few
days of equilibrium, reached maximum levels within 9 days and
decreased afterwards.  Maximum concentration levels were not
achieved after equivalent periods of equilibration, but all oil
concentrations stabilized after approximately 12 days  of equili-
bration.  Despite variations in maximum solubility levels,
experimental data for all oils showed similar trends,  except
Suniland crude which reached maximum solubility in water in the
first day.   Since samples were taken daily, it is not  known
whether this crude actually achieved maximum solubility earlier.

     The range of oil concentrations varied from zero  to a little
less than 200 ppm.  Generally, the processed oils, i.e., #2 and
#6 fuel oils, were less soluble than the crude oils.  The rate
and extent  of solution  of an oil depends on the chemical composi-
tion of the oil.  Crude oils from Brass River and Arzew exhibited
similar solubility behavior that differed slightly from the other
crude oils.  The rates  of dissolution of these two crudes were
stable for  the first few days, then accelerated, until maximum


                               120

-------
solubility levels were reached.   Ironically, these two crudes are
from the same geographical  area  (Algeria).   Table 6 shows the
maximum concentrations in solution and the  time these maximum
concentrations were reached.

Salt Water Studies

     The rates of dissolution of 5 crude oils and one refinery
product were investigated in  aqueous solutions containing 3.5%
sea salt.   The experimental  data are presented in Appendix C.

     In general, solubility levels in salt  solutions are lower
than those in tap water.   Thus,  the rates of dissolution are
lower.   Longer equilibration  periods were also required to
achieve maximum concentrations in salt solution.   The only excep-
tion is Brass River crude.   Table 7 shows the maximum concentra-
tion levels and required  equilibration periods.  The last column
shows the percent difference  when maximum concentrations in tap
and salt water are compared.   The solubilities of #2 fuel oil in
tap and salt water are comparable.  Decreases of  up to 99%
occurred for Alaska and Brass River crudes.   These results indi-
cate that hydrocarbon levels  in  seawater will be  much lower than
in fresh water systems.

     Oil concentration levels in salt water  represent the levels
that would exist in marine  waters with equivalent salt content.
The effect of sea salts is  to decrease the  solubility of oils by
"salting-out" hydrocarbons.   In  marine environments, this reduc-
tion in solubility may be offset by the presence  of dissolved
organic matter in sea water.   These materials can solubilize oils
and increase the solubilities of oils in salt water beyond the
limits  of solubilities in tap water.

Saturation Studies

     The concentration of oil in water at saturation was deter-
mined for each oil.  These  tests were conducted in closed vessels
and the oil/water systems were allowed to equilibrate for 3
months.

     Oil concentrations at  "saturation" are  given in Table 8.
For any oil, the concentration at saturation is significantly
greater than the maximum  solubility in open  water (Table 6).   It
is necessary to caution that  these values may not be true satura-
tion concentrations; however, they should be close due to the
prolonged equilibration of  the oil and water systems in closed
vessels.
                              121

-------
                             TABLE 6

            Maximum Solubilities of Oils in Tap Water
                     and Times to Attainment
Oil
Nigerian Crude
Lagunillas Crude
La Rosa Crude
#2 Fuel
#6 Fuel
Sum' land Crude
Alaska Crude
8% Crude
Iranian Crude
Sahara Crude
Brass River Crude
Arzew Crude
Time (t )
(days)
6
9
5
7
2
1
3
2
4
2
8
8
Maximum Solubility
(ppm)
4.5
2.5
3.4
1 .7
0.9
61 .2
152.1
78.6
15.0
24.0
176.0
196.0
Comparison of Predictions by the Dissolution Model  and Experimen-
tal Data

     During regression, the values of the parameters in the
dissolution equations  are adjusted in the iterative process to
give final values that best fit the equations to experimental
data.   Convergence was achieved in all numerical analyses.  Table
9 summarizes the final values of the four parameters for all the
oils.   Statistical tests of the significance of the values of K,

and K,. showed that these parameters are significantly different
                                           2
from zero.  The correlation coefficients, R , are close to unity,
indicating that there is good correlation between the model and
experimental data.  Correlation coefficients were less than 0.9
in only two cases:  Brass River and Arzew crudes.

     The final  values of the two mass transfer coefficients, K.

and Kr, vary for different oils but all  values are  smaller than

unity.  Variations in the final values of the mass  transfer
coefficients are caused by the different characteristics of the

                               122

-------
                            TABLE 7

           Maximum Solubilities of Oils in Salt Water
                    and Times to Attainment
Oil
Nigerian
La Rosa, Venezuela
North Slope, Alaska
Brass River, Algeria
Iranian
#2 Fuel
Time (tj
(days)
14*
12-13
4
4
7
5-6
Maximum
Solubility
(ppm)
1 .5
2.5
1 .1
1 .4
1 .5
1 .5
Decrease "*"
(%}
67
26
99
99
90
12
   * End of Equilibration
   + Based on Maximum Solubility in Tap Water
soluble and volatile components in the oils.   In tests using tap
water, K£ was always larger than K.;  but the  reverse is true in

tests using salt water.   Several factors that were not investiga-
ted influence KE and K,.   They are the temperature of the bulk

oil and water phase, salinity, and pH.  In the field, KE and K,

are functions of several  other factors,  such  as the wind velocity
profile above the oil  layer, turbulence  in the water phase,  and
water quality.  The final  values of C ,  in six out of twelve oils,

converged to the experimental  "saturation" concentrations
tabulated in Table 8.   The differences were small  in those cases
where the values did not  agree.  The  values of C$  in tests
conducted in salt water  are smaller than those in  tap water  and
reflect the lower solubilities of oils in salt water.  The final
values of tm correspond  to the times  required for  the oils to
reach maximum solubility  in the specific medium.  These values
are comparable to those  given  in Tables  6 and 7.

     Comparison of concentration-time profiles, using the disso-
lution model and experimental  data, can  be made by considering
Figures 15 to 20.  Experimental data  and model profiles are  shown
in Figure 15 for #2 and  #6 fuel oils, and crudes from Nigeria,
Lagunillas and La Rosa.   The dissolution model predicts experi-
mental data fairly accurately  in all  cases.  The profiles for
Suniland, Sahara and Iranian crudes,  and the  oil mixture contain-

                               123

-------
                            TABLE 8

               Oil  Concentrations at "Saturation"
                Oil                   Concentration  (CV)
         	(ppm)	

          Nigerian Crude                    29.9
          LagunillasCrude                  13.7
          La  Rosa  Crude                     17.4

          #2  Fuel                             9'3
          #6  Fuel                            21'12
          Suniland Crude                   102.6
          Alaska Crude                     235.8

          8%  Crude                         120'9
          Iranian  Crude                     23.3
          Sahara Crude                      35.6
          Brass River Crude                270.7
          Arzew Crude                      306.0
ing 8% crude are displayed in Figure 16.   The results show that
the predictions by the model  are comparable with experimental
data for Sahara and 8% crudes,  but the dissolution  model  does  not
fit the data as well  for crudes from Iran and Suniland.   Dispari-
ties occur during the later stages of equilibration.   Figure 17
compares the results  of the model  and data for Alaska,  Brass
River, and Arzew crudes.  The fit  between the data  and  the model
is good for Alaska crude;  it is poor for  Brass River  and  Arzew
crudes, because of the significantly different behavior  of these
two crudes.

     Profiles of the  dissolution model and experimental  data,
from solubility studies of #2 fuel oil and Iranian  and  Brass
River crudes in salt  water, are displayed in Figure 18.   Similar
graphs are shown in Figure 19 for  Nigerian, Alaskan and  La Rosa
crudes.  The model predicts experimental  data accurately  in all
cases.  Figure 20 shows the data and model for Nigerian  crude
using the results of  tests in tap  and salt water.   Despite
slightly different behaviors of the crude,in the two  test liquids,
the data and models indicate that  the final concentrations are
similar.


                               124

-------
                      TABLE 9

      Comparison of Final  Values of Parameters
from Fitting Dissolution Model  to Experimental Data
                             Parameters
Oil
A. Tap Water
Nigerian Crude
La Rosa Crude
Lagunil las Crude
#2 Fuel
#6 Fuel
Iranian Crude
Sahara Crude
Suniland Crude
8% Crude
Alaska Crude
Brass River Crude
Arzew Crude
B. Salt Water
#2 Fuel
Iranian Crude
Brass River Crude
La Rosa Crude
Alaska Crude
Nigerian Crude
KL
(day"1)

0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.

0.
0.
0.
0.
0.
0.

055
040
025
06
012
42
60
90
71
20
07
07

330
220
299
054
060
120
KE
(day'1)

0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.

0.
0.
0.
0.
0.
0.

0801
090
080
no
098
300
260
951
163
28
601
540

060
01
025
044
06
05
*m
(day)

3
5
8
7
4
4
2
1
1 .5
3
9
8

5
6
3.8
11
3
12
Cs
(ppm)

29.
17.
13.
5.
15.
17.
35.
100.
144.
290.
270.
306.

2.
2.
2.
6.
5.
2.

9
4
7
0
0
0
6
0
6
0
0
0

0
0
0
0
0
0
R2

0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.

0.
0.
0.
0.
0.
0.

993
989
994
980
948
960
990
922
988
968
860
860

992
996
997
980
962
978
                        125

-------
   1C r— i—•
f-  I/) -r-  QJ
i-  O C  3 3
     3 U_
O)
i-  (B IB CM VO
  _J _J =«==«=
C/5
>-
                                                               LU
                                                               CD
                                                               o
                                                               LU
                                                                           (O
                                                                        r— W>
                                                                         (O O
                                                                         O) ro
                                                                         E-l
                                                                        •i—
                                                                         S-  •
                                                                         <_>
3
r--0
O C
tf> (C
(O
•r- to
Oi—
  'I-
<*- O
O

W 0)

-------
o
                                                       en
                                                       CD
                                                       ID
                                                       O
                                                       LU
                                                                 (O +•>
                                                                +J X
                                                                s- co
                                                                OJ
                                                                Q.-O
                                                                X C
                                                                TJ  «
                                                                C C
                                                                ia to
                                                                   S-
                                                                T3 -O S.
                                                                O C O)
         C  -3!
         O <0
         •r- i. Q.
         +* 10 10
         3 .C t—
         r— (O
         O t/0 I
         CO
         «/)  »^-
         •r- T3 t-
         Q C O
           (O
         u_ i— T3
           C CO
           3 CO
         10 OO O)
         O)    O
         r- E O
         •r- O S-
         M- 5- Q.
         O <4-
         S-   T3
         Q. CO C
           Ol to
         <4- -O
         O 3 r—
           S- -r-
         C O O
         O
         CO i. 
-------
                                                              ca
        (O   -r-
       ^   i-
        v>   a>
        
       i— S- -i-
       oo o a:
       •(-> a»  to
       S- IM  (O
       o s-  J.
       z < ca
       f— c\j
                               B
  m         •         •         •
ca       ca        ca       ca
ca       in        ca       in
en       CM
            C\J       —t       T-1
                                                              CO
                                                                     UJ
                                                                     QQ
                                                                     C3
                                                                     UJ
rt) fO
+J •!-
c s-
O) OJ
E 01
•r— r—
s. «c
OJ	
a.
x s.
LlJ O
   >
-a -I-
c: o:
 re
•a s.
o co
                                                                        4->  2
                                                                        3  
                                                                        i- M- <0
                                                                        (B   2
                                                                        o. 
-------
            (O
            •r-
            i.
            
       O CJ >
            •^
       r-  c o:
       0)  (0
       3 •!- CO
       u_  c 10
          IB — i CO
       •— CM
                          CsJ
in
  •
CM
CS

CM
in
                                                             CJ
                                                             C\J
                                                 CO
                                                             ca
                                                               •
                                                             CM
                                                             IS
                                                              •
                                                             00
CO


o

LU

»—i
H-


O
i—i
H-

az
                                                        o
                                                        LU
                                                             IS)
                                                                  10 -W
                                                                  •M i—
                                                                  c to
                                                                  
                                OJ T-
                                T3 OS
                                o
                                s: co
                                  CO
                                c to
                                o s.
                                •r- CD
                                                                  O 10
                                                                  CO
                                                                  CO C
                                                                  •r- to
                                                                  Q -r-
                                                                     c
                                                                  M- 10
                                                                  O S-
                                                                    i— c
                                                                  CO
                                                                  o»  •
                                                                  «t- O
                                                                  O
                                                                  s_ ,—
                                                                  Q-  Ol
                                                                                cvi
                                                                              c=»=
                                                                              o
                                                                              CO S_
                                                                             •i- O
                                                                              i- <4-
                                                                              tO   t-
                                                                              O. to 0>
                                                                              E 4-> -M
                                                                              o 
-------
                                                   >-

                                                   Q
                                                   cn
                                                   ID
                                                   C3
                                                   UJ
                                                            (O
                                                            4J i-
                                                            1= O)
                                                            O) +->
                                                            E re
                                                            •«- 3
                                                            s_
                                                            Ol -l->
                                                            Q.f—
                                                            X (O
                                                            uj oo
re to
   01
r— -O
01 3
T3 S-
00
z:
   re
c 10
O O
•i- o:
•M
3 re
r— _l
O
10 T3
CO C
•r- re
O
   c
>*- re
o •*
   (O
to re
0> r—
i— 
                                                            o re
                                                            o a
                                                            CTl
                                                            O)
                                                            s_
(Wdd)   NOIlVdiN33N03  110
                          130

-------
        s-
      i- 
      Q.r—
      to 
                                                                       O 
-------
     The rate of dissolution in water of some crude oils and
petroleum-based products has been investigated in open static
tests.  The tests correspond to an unlikely worse case of an oil
spill, such that the oil completely covers the water surface.
The experimental data suggests that the dissolution process can
be divided into two phases.   Initially, soluble and volatile
organics diffuse into the underlying water; later, some volatile
materials evaporate from solution.  Thus,  the dissolution process
is influenced by evaporation in the final  stages of equilibration.
The amounts of these hydrocarbon species vary according to the
composition of the oil, hence, variations  exist in maximum solu-
bilities, times to attainment, and dissolution rates.   A segmen-
ted dissolution model has been applied successfully to correlate
the experimental data of the oils.  Although the goodness-of-fit
of the dissolution model to  the experimental data varies, a model
based on the two processes has been shown  to be capable of
reasonable predictions of the solubilities of different oils, as
a function of time, in water below a surface slick.

     In general, the results of solubility measurements indicate
that these oils exhibit low  solubilities in water.  The processed
oils, #2 and #6 fuels, are less soluble than the crude oils.  The
solubility behavior of oils  depends on their chemical  composition.
For fuel oils, the extent of refinery processing and additives
are important factors affecting solubility.  The data  show that
maximum concentrations were  not achieved by all the oils after
equivalent periods of contact.  In spite of variations in maximum
solubility levels and times  to attainment, the experimental data
show similar trends.

     The results of the dissolution studies give probable oil
concentration levels in the  water immediately under a  surface
slick, when these oils are spilled on water.  Oil concentrations
are low and may not be hazardous to marine organisms when exposure
is limited.  Concentration levels may be dangerous when oils
attain maximum solubilities  in water.  A class of marine organisms
which has adapted to the sea surface environment is neuston.  The
solubility limits established for these oils are important to
bioassay studies of the toxicity of oils to neuston.  Oil concen-
trations will be lower in sea water.  Also, water movements at sea
will continuously decrease oil concentration by dilution.  In
contrast, the dissolved organic matter in  the surface  microlayers
of the sea will increase oil concentration by solubilization.


SPREADING

     The rates of spreading  of 12 oils were investigated.  The
experimental runs consisted  of spilling specific volumes of oil
onto calm tap water.  Four different volumes of each oil were
spilled at different flow rates.  The effects on spreading of
several parameters, such as  the volume of  oil spilled, duration

                               132

-------
of spill  and physical  properties of the oils,  i.e.  viscosity,
density,  surface tension and interfacial  tension between oil  and
water, were investigated.   These parameters are important in
determining the increase with time of the area of an oil slick
spreading on calm water.  Water temperature is a factor in the
rates of  spreading of oils but it was not investigated in this
study because the effect of temperature should be determined  by
its influence on the properties of the oil  and water phases.   The
effect of temperature will be proportional  to  changes in the
values of the physical  properties of the  two phases.  The data
generated in each experimental run include  areal extents of the
oil as a  function of the time of spreading.  Photographic tech-
niques were used and the areas covered by the  oils  were measured
from photographic images.

     All  the oils did not  spread uniformly  and several slicks
developed regions with varying thicknesses  of  oil.   In some
cases, subjective judgment was used to determine the areal extent
of a slick if its profile  was not properly  defined  in the photo-
graphic prints.  Spreading patterns vary  for different oils and
there is  no preference for elliptic or circular geometries.  The
initial configurations of  a slick during  the initiation of
spreading and the final  profile at the end  of  an experiment are
influenced by the rate of  oil discharge,  the duration of the
spill, thermal convection  currents in the water column, proper-
ties of the oil and spreading forces.  The  configurations and
areas of  slicks may vary even when the slicks  are formed by the
same oil  and under the same conditions.

     Spreading equations were derived to  correlate  experimental
data.  The development of  these equations parallels Fay's work
and yields four spreading  equations:  surface-tension/viscous
(STV), surface-tension/inertia (STI), gravity/viscous (6V), and
gravity/inertia (GI).   Each spreading equation defines a spread-
ing regime in which only two opposing forces dominate the spreadr
ing behavior of a slick.  The equation for  each regime is then
derived by equating two  forces, one of which accelerates
spreading and the other  retarding spreading.  The dependent
variable  in each equation  is the area covered  by the spreading
slick and the independent  variables include the properties of the
oil and water phases.   This information is  usually  available  at
sites of  spills or can be  determined experimentally.  Each
equation  contains an empirical constant.

     Experimental data from spreading studies  are presented in
Appendix  D.  The spreading equations were fitted to experimental
data to estimate the values of the empirical coefficients.  The
final values of the coefficients and the  regression statistics
are summarized in Tables 10 to 21.  The correlation coefficients
indicate  the models are  capable of fitting  experimental data  with
varying accuracies.  In  general, the order  of  the goodness-of-fit
of the spreading equations from the best  to the worst is

                               133

-------
GV > STV > 61 > STI.   As expected the fit with the surface-
tension/inertia model  was the worst;  this model  predicts  that the
area of spreading is  independent of the volume of oil  spilled.
Thus, this equation is not acceptable as a model  for predicting
variation of the area  covered by an oil slick with time.
                            TABLE 10

         Summary of Coefficients  of Spreading  Equations
                         for Arzew Crude
Equation
1. Surface-Tension/Viscous




2. Surface-Tension/Inertia




3. Gravity/ Viscous




4. Gravity/Inertia




Vol .

25
50
75
100

25
50
75
100

25
50
75
100

25
50
75
100
K

15.695
12.356
11 .400
10.289

0.0392
0.0466
0.0549
0.0603

25.922
18.274
15.908
13.800

3.208
2.541
2.342
2.133
R2

0.966
0.9745
0.9769
0.9580

0.5550
0.5776
0.6040
0.579

0.9895
0.9937
0.9935
0.9720

0.8514
0.8695
0.8804
0.8585
                              134

-------
                   TABLE  11

Summary of Coefficients of  Spreading  Equations
             for Brass  River  Crude
Equation
1. Surface-Tension/Viscous




2. Surface-Tension/Inertia




3. Gravity-Viscous




4. Gravi ty/ Inertia




Vol.

25
50
75
100

25
50
75
100

25
50
75
100

25
50
75
100
K

7.869
4.462
4.533
3.747

0.0183
0.0132
0.0169
0.0189

13.559
7.136
6.901
5.289

1 .705
0.905
0.909
0.799
R2

0.9757
0.9636
0.9479
0.9860

0.7792
0.5845
0.6096
0.7843

0.9528
0.9847
0.9656
0.9739

0.9718
0.8589
0.8547
0.9716
                     135

-------
                   TABLE  12

Summary of Coefficients of  Spreading  Equations
               for Alaskan  Crude
Equation
1. Surface-Tension/Viscous




2. Surface-Tension/Inertia




3. Gravity/Viscous




4. Gravity/ Inertia




Vol .

25
50
75
100

25
50
75
100

25
50
75
100

25
50
75
100
K

59.989
49.778
28.064
28.304

0.0772
0.0908
0.0664
0.0775

68.238
52.281
27.060
26.332

5.175
4.154
2.402
2.396
R2

0.9953
0.9785
0.9921
0.9847

0.6791
0.6563
0.6582
0.6426

0.9986
0.9901
0.9987
0.9955

0.9278
0.8961
0.9176
0.9024
                     136

-------
                   TABLE  13

Summary of Coefficients of  Spreading  Equations
               for Iranian  Crude
Equation
1. Surface-Tension/Viscous




2. Surface-Tension/Inertia




3. Gravity/Viscous




4. Gravity/Inertia




Vol .

25
50
75
TOO

25
50
75
100

25
50
75
100

25
50
75
100
K

34.650
24.920
20.314
18.990

0.0573
0.0594
0.0609
0.0670

47.472
30.818
23.619
21 .139

4.399
3.126
2.544
2.373
R2

0.9853
0.7905
0.9676
0.9621

0.6355
0.5918
0.5906
0.5871

0.9966
0.9898
0.9873
0.9817

0.9003
0.8684
0.8648
0.8584
                     137

-------
                   TABLE  14

Summary of Coefficients of  Spreading  Equations
               for Sahara Crude
Equation
1. Surface-Tension/Viscous




2. Surface-Tension/Inertia




3. Gravity/Viscous




4. Gravity/Inertia




Vol .

25
50
75
TOO

25
50
75
100

25
50
75
100

25
50
75
100
K

41 .544
28.351
17.955
20.108

0.0616
0.0618
0.0511
0.0642

67.098
41 .554
24.490
26.951

6.555
4.415
2.845
3.056
R2

0.9665
0.9415
0.9739
0.9044

0.5564
0.5437
0.5973
0.5526

0.9905
0.9666
0.9916
0.9226

0.8525
0.8242
0.8746
0.7904
                     138

-------
                                   TABLE  15

                Summary of Coefficients of  Spreading  Equations
                              for Nigerian  Crude
Equation
1. Surface-Tension/Viscous




2. Surface-Tension/Inertia




3. Gravity/Viscous




4. Gravity/Inertia


•

Vol .

25
50
75
100

25
50
75
100

25
50
75
100

25
50
75
100
K

10.300
8.342
6.459
7.486

0.0182
0.0193
0.0198
0.0280

15.232
11 .492
8.287
9.429

1 .584
1 .207
0.954
1 .094
R2

0.9674
0.9920
0.9974
0.9849

0.8198
0.6820
0.7542
0.9005

0.9324
0.9987
0.9907
0.9727

0.9868
0.9242
0.9649
0.9944
c
                                     139

-------
                   TABLE 17

Summary of Coefficients  of Spreading  Equations
                for #6 Fuel  Oil
Equation
1. Surface-Tension/Viscous




2. Surface-Tension/Inertia




3. Gravity/ Vis co us




4. Gravity/ Inertia




Vol .

25
50
75
100

25
50
75
100

25
50
75
100

25
50
75
100
K

5.552
6.440
5.219
6.329

0.0021
0.0041
0.0045
0.0059

4.709
4.809
3.607
4.288

0.182
0.220
0.182
0.213
R2

0.9846
0.9425
0.8590
0.9448

0.7443
0.9012
0.9658
0.8994

0.9680
0.8960
0.7998
0.9090

0.9598
0.9977
0.9595
0.9867
                     141

-------
                   TABLE  18

Summary of Coefficients of Spreading  Equations
                 for  8% Crude
Equation
1. Surface-Tension/Viscous




2. Surface-Tension/Inertia




3. Gravity/ Viscous




4. Gravity/ Inertia




Vol .

25
50
75
100

25
50
75
100

25
50
75
100

25
50
75
100
K

30.991
27.293
25.342
25.264

0.0488
0.0567
0.0644
0.0730

39.925
32.973
28.636
27.895

3.657
3.024
2.791
2.700
R2

0.9141
0.9917
0.9987
0.9907

0.9111
0.8410
0.7748
0.7715

0.8634
0.9780
0.9918
0.9905

0.9878
0.9887
0.9700
0.9433
                     142

-------
                   TABLE  19

Summary of Coefficients of Spreading  Equations
             for Lagunillas  Crude
Equation
1. Surface-Tension/Viscous




2. Surface-Tension/Inertia




3. Gravity/Viscous




4. Gravity/ Inerti a




Vol .

25
50
75
100

. 25
50
75
TOO

25
50
75
100

25
50
75
100
K

21 .614
13.338
35.032
35.032

0.0112
0.0102
0.0351
0.0351

19.840
11 .085
26.719
26.719

0.953
0.587
1 .462
1 .462
R2

0.8623
0.8745
0.9761
0.9761

0.9555
0.9549
0.8423
0.8423

0.7976
0.8149
0.9470
0.9470

0.9696
0.9756
0.9949
0.9949
                     143

-------
                   TABLE  20

Summary of Coefficients of Spreading  Equations
               for La  Rosa Crude
Equation
1. Surface-Tension/Viscous




2. Surface-Tension/Inertia




3. Gravi ty/Vi scous




4. Gravity/Inertia




Vol .

25
50
75
100

25
50
75
100

25
50
75
100

25
50
75
100
K

61 .194
56.427
57.586
53.202

0.0798
0.1001
0.1327
0.1378

60.062
51 .190
48.649
44.097

4.261
3.739
3.869
3.442
R2

0.9835
0.9754
0.9839
0.9465

0.7458
0.6488
0.6840
0.6597

0.9662
0.9864
0.9884
0.9540

0.9636
0.8920
0.9163
0.8620
                     144

-------
                    TABLE 21

Summary of Coefficients of Spreading  Equations
                for Suniland  Crude
Equation
1. Surface-Tension/Viscous



2. Surface-Tension/Inertia




3. Gravity/Viscous




4. Gravity/Inertia




Vol .
25
50
75
100

25
50
75
100

25
50
75
100

25
50
75
100
K
52.768
30.937
27.396
24.414

0.0701
0.0673
0.0717
0.0751

73.803
38.545
32.233
27.386

6.179
3.742
3.254
2.908
R2
0.9516
0.9958
0.9818
0.9852

0.5720
0.7414
0.6645
0.6569

0.9726
0.9918
0.9896
0.9999

0.8437
0.9527
0.9060
0.9083
                       145

-------
     The values of the coefficients differ for each oil  and
spreading regime.   Within each spreading regime slight to large
differences can be seen in the values of the coefficients for the
four oil volumes spilled.  These differences are caused  by experi-
mental  error due to variations in thermal  convection currents in
the water and, perhaps, air motions which are unavoidable even in
the laboratory environment.  It is probable that for each spread-
ing regime, the mean of the four coefficients is a more  accurate
value for the empirical constant.

     The experimental  data and predictions by the four spreading
equations are given in Figures 21 to 32 for #2 fuel oil, Nigerian
and Alaskan crudes.  The data and model profiles show that oil
slicks  spread faster initially and spread more slowly as the
slicks  age.

     The effects of oil properties, i.e. density, viscosity,
surface tension and interfacial tension between oil and  water can
be judged from the spreading equations.  Increases in density and
viscosity have negative effects on the spreading behavior of oils
The net effect of surface tension varies according to whether it
is positive (accelerates spreading) or negative (retards spread-
ing).  Data for #6 fuel oil and Lagunillas crude show that
viscous oils spread more slowly than less viscous oils.   As the
ratio of the density of oil-to-water decreases, spreading rate
increases.   Since spreading of oils on water persists longer than
the duration of the spill, the initial effect of the rate of oil
discharge is cancelled by the total volume of the oil spilled.
Therefore,  the rate of discharge of oil is important in  the early
stages  of a spill  and  in situations where the oil is discharged
continuously, e.g., oil seeps.

     The mathematical  models used here are valid only for spread-
ing on  calm seas.   The results of the experiments have shown
that, even  for this simple situation,oi1s behave differently.
Information on the behavior of a greater variety of oil  types is
needed.  In the open sea, gross oil transport commences  almost
immediately after oil  is spilled.  The transport processes due to
turbulence  created by  wind, current, wave and tidal forces will
be superimposed on natural spreading.  The rates of these
processes are at present difficult to quantify, however, these
modes of transport are of importance in evaluating the potential
damage  to marine ecosystems by oil slicks.


CHEMICAL DISPERSION

     When oil slicks are agitated by turbulent forces at sea,
they break  up into small droplets that disperse into the water.
The formation and dispersal of oil droplets can be aided by
application of chemical dispersants to the slick.  Thus, infor-
mation  on the mechanisms leading to dispersion of droplets is

                               146

-------
                                   "O
                                   c:
                                   (U
                                   01
                                   co
SI

in
                                                   to  re

                                                   O +•>
                                                   O  L-
                                                   C/>  CO
                                                   •r-  C
 c: c  «/> re
 o o  3 i-

 (/)(/)(_>£-.
 c c  to  H-1

 co co  >> >>
 U o •+•> 4J
 ro ro T- T-

 ^ ^  ro ro
 3 3  S- S-
CO t/> CD CD
                                                  i— CM CO <
                                                      csa

                                                      in
                                                      CO
                                                                  CO
                                                                  LO
                                                                  CSJ
                                                      Q

                                                      CD
                                                      OJ
                                                                         CO
                                                                         LU
ET

OJ
                                                                                   s-
                                                                                   o
                                                                                   I/)
                                                                                   c
                                                                                   O
                                                                                   rC
                                                                                   3
                                                                                   CT
                                                                       re
                                                                       CD
                                                                       i.
                                                                       o.
                                                                      oo
                                                                                   re

                                                                                   re
                                                                                   re
                                                                                  a
                                                                                   re
\\
\

\\ \
\
e \
rr> \
\
\
\ '\
iA \
e \\ \
e \ \

1..
tSJ
•
CD
V
— -i:
i

*
CO
A
\ -
\
\ -
X \
•*• ^ ^. \
^==r^J
S> SI
^ g
(S
.-f
in
w-H
CSI
Q

1S>
m
in
ts
si
CO CS>
LU +J
CO =
O «u
< E
CJ E^
Q-E
X
LU LT>
CVJ
«l ^^
O
o o
to
S- (U
(0 3
Q.U-
O CNJ
O =t*=
00
< CO
3
                                                                                   cn
'id
                                        JO  V3HV
                                        147

-------
                                                 CO
                                                 LU
                                                 LU
                                                 Q
                                                 LU
                                                 CO
                                                 Q.
                                                         J_
                                                         o
                                                         C
                                                         o
                                                         ro
                                                         3
 re
 Oi
                                                         Q.
                                                         to
                                                         -o
                                                         C
                                                         (O
                                                         CO
O)
E


at v>
Q-r-
X E
LU
  O
4- in
o —

to i —
c -i-
o o
                                                         i- OJ
                                                         «O 3
                                                         o cv
                                                         <_> =«*=
                                                        CM
                                                        CM
                                                         o;
                                                         3
                                                         CD
('id  'OS)  QV3ddS JO V3&V
                       148

-------
                                     ~0ne~'
                                      3 -r-

                                      8t
                                      tO  O)
 c c to ro
 O O 3 •!-
•r- v- O •»->
 to to o fc.
 C C tO QJ
 OJ , >,
 o o •»->•*->
 (O fl3 *^~ "r1"
M- 4- > >
 S- S- (O «O
 3 3 S- i-
CO CO C3 CS
                                                   J
  LO
  CO
is
 ICO
                                                    in
                                                   !OJ
                                                           CO
                                                           LU
                                                           LU
                                                           C3
                                                           LiJ
                                                           CO
                                                           Q.
                                                           -C
                                                           _J
                                                           LU
                                                                      f-
                                                                      o
                                                                      to
                                                                      c
                                                                      O
                                                                      ce
                                                                      3
                                                                      cr
                                                                      UJ

                                                                      en
                                 to
                                 O)
                                 S-
                                 CL
                                                                      c
                                                                      «8
                                                                      (O
                                                                      fO
                                                                      o
                                 O) to
                                 Q-r-
                                 X E
                                LU
                                   un
                                                                      o o
                                                                      CO
                                                                      •r- r^
                                                                      S- 0}
                                                                      re =
                                                                      O.U-
                                                                      E
                                                                      O OsJ
                                                                      (_> =«»=
                                                                      O)
                                                                      s_
                                                                      3
                                                                      O)
('Id   'OS)  QV3da'S  dO  V3HV
                            149

-------
-o

CU
en
CD
                                     (/> to

                                     o •*->
                                     o s~
                                     in o
 C E CO (13
 O O =5 v
•I— •!— O +J
 to to o s-
 c= c t/> cu
 CO CD -r- C
I— h- 2» t-i

 CD CO >> >>
 O O -M -4-*
 
to
o

r— •
to
[ ^
c
CD
E
S- to
CD i—
0. E
X
LU CD
O
l|— f__
0 	
                                                                 VI r—
                                                                 O O
                                                                 10
                                                                 S- (U
                                                                 ro 3
           IS
                                                                 O CM
                                                                 CJ =tfc
                                                                 CM


                                                                 O)
                                                                 S-
                                                                 3
                                                                 CD
rid  'OS)   avBdds  JQ
                           150  .

-------
c c  «/) te
O O
f-v-  O •»->
CO CO  O S-
C C  «/>
0) OJ ••- C
I  I
0) OJ >> >,
  O +•> +•>
(O fO
                                                          CO
                                                          LU
                                                          LU
                                                         CD
                                                         LU
                                                         CO
                                                          LU
C
O
to
3
CT
-o
CO
O)

Q.
co

-o

CO

CO

CO
O


CO

c

E
 cu
 Q-LO
 x oo
                                                                 14-  0)
                                                                 o "a
                                                                    3
                                                                 in  s-
                                                                 e o
                                                                 o
                                                                 co  c
                                                                 •r"  CO
                                                                 J_ -t-
                                                                 co  S_
                                                                 a. 
-------
c c in to
O O 3 v
•r- -i- O 4->
trt IO O S-
      
OJ
   I
0) O>
O O -4-> +-»
to to •>— -I—

   1- to to
3 3 I- S-
OO GO CD CJ
                                                           UJ
                                                           LU
                                                          UJ
                                                          on
                                                          Q_
                                                          -C
                                                          _J
                                                          UJ
                                                                    c
                                                                    o
                                                                    (O
                                                                    3
                                                                    cr
                                                                    en
                                                                    c
                                                                    O)
                                                                    S-
                                                                    Q.
                                                                   co

                                                                   -o
                                                                    c
                                                                    to

                                                                    to
                                                                   4->
                                                                    (O
                                                                   Q



                                                                    to
                                                                    O)
                                                                    Q-O
                                                                    x in
                                                                      
                                                                    O T-
                                                                   o z
                                                                   VD
                                                                   C\J
                                                                    OJ
  ('Id  'OS)   QV3ddS  JO  V3dV
                               152

-------
c c in re
o o 3 T-
•i- -r- O •*->
   in o s-
   QJ
   O •(-» 4->
   ro •>— •>—
»*-<*->>
 S- S- (O fO
 3 =5 S- S-
   CO CD CD
                                                             CO
                                                             Lul
                                                             CD
                                                             UJ
                                                             CO
                                                             a.
LU  '
                                                                      i.
                                                                      o
         in
         c
         o
                                                                      to
                                                                      •3
                                                                      CT
         CT>
         •O
         re
         O)
         i.
         a.
         C/)
                                                                      rO

                                                                      re
                                                                      to
                                                                      o
                                                                       re
                                                                       at
                                                                       Q-LO
                                                                       x r-»
                                                                      n_ a>
                                                                       o -a
                                                                         3
                                                                       t/> S-
                                                                       c. o
                                                                       o
                                                                       tO C
                                                                      •r- re
                                                                       i- •!-
                                                                       re 5-
                                                                       CL O)
                                                                       E 01
                                                                       O ••-
                                                                      o z:
                                                                      OJ
                                                                       OJ
                                                                       s-
  Cld   'OS)   QV3ddS  JO  V3dV
                                 153

-------







O)
01
0)








to
3
O
O
C/)
c
o
(/)
c:
O)
H-
1
Q)
O
ro
i-
3
^_

CO
•r—
4^
i-
0)
c
C (/>
O 3
•r- O
«/> (J
C CA
0> -f-
1— >
1 «^.
Q) >>
O -t->
S- «0
3 S-
tsi cr>
CM ro






fC _
-M
d)
c

•*
>^
£
to
i.
^
IS
^srj~




SJ
in
CO





IS
cS
CO

                                                        in
                                                        CM
                                                       OJ
                                                             05
                                                             LU
                                                             LU
                                                             Q
                                                             LU
                                                             CO
                                                             Q.
                                                             LU
                                                                     i.
                                                                     o
CO
 to
 3
 o-
LU

 cn
 c

•o
 «o
 Ol
 i-
 o.
oo

•o

 CO

 ro
+J
 

CD 10
E i—
•r- E
i.
O) O
O.O
X r-
111 v_^-
                                                                     S-
                                                                    c o
                                                                    o
                                                                    tO E
                                                                    •r- to
                                                                    i- T-
                                                                    (O f-
                                                                    CL 01
                                                                    E 0>
                                                                    O '1-
                                                                    oo
                                                                    CM
                                                                    o>
                                                                    s_
                                                                    en
        ('Id  'OS)  QV3&JS  dO  V3tfV
                                   154

-------
               »e
-u



 O -£>
 O 5-
 to O)

> HI

 C C CO (TJ
 O O 3 i-
•r- T- O +>
 «/> to o s-
 c c 10 o>
 Q» Q) -r- cr
                                                 (S5
                                                 •^~
C3

LO
CO
                 SJ
                                                       CO
                                                       UJ
                                                       UJ
                                                       Q
                                                       UJ
                                                       CO
                                                       Q_
                                                       UJ
                                                                 s-
                                                                 o
                                                                «*-
                                                                to
                                                                3
                                                                cr
                                                                LU
                                T3
                                 (O
                                 
                                                                10
                                                                •r- C
                                                                1- f8
                                                                (O -^
                                                                O. CO
                                                                E ro
                                                                O i—
                                                                cr>
                                                                CVJ
                                              ISl
Cld   "OS)  QV3ddS  JO  V3HV
                            155

-------
                            0)
                               t/i to
                               3 •!-
                               O •»->
                               O 5-
                               O O)
                               •r- C
                               c c «/> re
                               O O 3 T-
                               •r- -1- O •!->
                               i/> 10 o s-
                               c c= to o>
                                O O •!-> 4->
                                re re *|"1" >r~
                               M- <4- > >
                                s- i- re re
                                3 3 S- S-
                               OO CO CD tD
in
CO
Si
  •
S)
CO
                                                   Lf>
                                                   OJ
                                                         CO
                                                         LjJ
                                                         LLl
                                                         a
                                                         LU
                                                         CO
                                                         Q_
                                                         LU
                 s-
                 o
SI
  •
Si
re
3
CT
LU

Ol
C
*^-
-o
re
o>

CL
CO

-o
c
re

re
4J
re
o


re
                 
                                                                    10 3
                                                                    C S-
                                                                    O O
                                                                    (/>
                                                                    •r- C
                                                                    s- re
                                                                    «a -^
                                                                    CL en
                                                                    E re
                                                                    O r—
                                                                    o <:
                                                                    o
                                                                    ro
                                                                    0)
                                                                    3
                                                                    TO
rid  'OS)  av3Hds  JO
                              156

-------
                                               co 03
                                               3 T-
                                               o -u
                                               u s-
                                               to a)
                                               c c: to 
                                               to to o s_
                                               c c  o>
                                               Q)  H-I

                                               > >>
                                               O O +-> 4->
                                               «O ~
                                               M- «*- > >
                                               S- S- (O ro
                                               3 3 i- J-
                                               OO CO CD CD
                                               r— CM CO
LO
OJ
03
CS1
  •
CD
                                                                     s-
                                                                     o
                                                                in
                                                    Gi
                                                    CO
                                                                in
                                                                CM
                                                           CO
                                                           UJ
                                        C\J



                                        esi

                                        in
                                                                       UJ
                                                                       CO
                                                                       CL.
                                                                in
                                                                csi
                                                                     to
                                                                     CT
                                                                     UJ
                                                        •a
                                                         as
                                                         Q)
                                                         s-
                                                         CL
                                                        to
                                                                                 to

                                                                                 to
                                                                                 
                                                                     £
                                                                     0)
                                                                     £ ^™^>
                                                                     •I- 
                                                                     •r- C
                                                                     s- re
                                                                     03 -^
                                                                     a. to
                                                                     E re
                                                                     o r-
                                                                     o <:
00
                                                                                 O)
          CiJ   'OS)  QV3cJdS  dO  V3cJV
                                          157

-------
                            S_ S- to fO
                            3 3 i- S-
                            00 00 CD
                                                          s_
                                                          o
                                                  CO
                                                  UJ
                                                 LU
                                                 CD
                                                 UJ
                                                 CO
                                                 Q_
                                                 LU
 tO

 cr
LU

 CD
 c

-o
 to
 
                                                         E (O
                                                         O i—
                                                         CM
                                                         CO
                                                         o>
(UJ  'OS)  QV3ddS  dO  V3dV
                          158

-------
essential  to planning appropriate clean-up operations, using
commercially available dispersants.

     This  study has identified the principal  and secondary mass
transfer processes that are important in chemical  dispersion.   A
theoretical  formulation has been proposed to  explain these
mechanisms of chemical dispersion.  This theory is based on
detergency and catalysis concepts.  Mathematical models were
derived to quantify the principal mechanisms  according to the
steps that are rate limiting.   The models must be judged on their
merits as  they were not verified by experimental data, due to
lack of accurate input data.   Because of the  variety of commer-
cially available dispersants,  knowledge of the mechanisms of
chemical dispersion is essential to establishing a rationale for
dispersant selection.  This work has identified these mechanisms.

     As this study progressed, the limitations of modeling became
apparent.   For example, the chemical structure and the degree  of
association  or micel1ization  of surface-active compounds in
commercial dispersant preparations must be known.   This informa-
tion is proprietary.   The properties of the surfactants are useful
for determining the rate controlling steps.  This study under-
scores the importance of basic data utilizing known surface
active compounds in free form  and mixtures of surfactants.  The
results of the experimental data from dispersion tests using 5
commercial dispersants and 3  oils are discussed next.

     Current practices in chemical dispersion tests report the
effectiveness of chemical dispersants in terms of the percent
dispersion.   The percent dispersion is based  on the concentration
of oil in  solution, following  violent agitation of an oil/water/
dispersant system, as a fraction of the concentration of oil that
would result if all the oil were dispersed in the water.  This
practice is  not followed here  because it has  little merit and  it
does not reflect accurately the efficiency of dispersants.  A
brief explanation follows.

     Usually, dispersants are  formulated in hydrocarbon or
aqueous solvents.  Hydrocarbon-base dispersants have some advan-
tages over aqueous-base dispersants.  For instance, hydrocarbon
solvents are easily miscible  with oil slicks.  This may lead to
faster dispersion of oil slicks.

     Furthermore, the sources  of hydrocarbon  in an oil/water/
dispersant system are the slick material and  the dispersant.
Thus, a hydrocarbon-base dispersant will contribute a larger
amount of  hydrocarbon to the  system than an aqueous-base disper-
sant containing the same surface-active compounds.  Since
commercial dispersants contain a variety of surface-active com-
pounds and additives  with hydrocarbon molecules, the fraction  of
hydrocarbons in different dispersants will vary with the structure
of the surface-active compounds (the hydrophobic group), and the


                               1 59

-------
concentration of surfactant, hydrocarbon solvent and additives.

     A water sample from a dispersion test will  contain some
hydrocarbon from all  the sources.   Analytical  techniques do not
discriminate between  the hydrocarbon fractions contributed by the
hydrocarbon solvent,  the hydrophobic group of the dispersant, and
the oil  slick.   Surely, a hydrocarbon-base dispersant will show
higher oil  levels in  aqueous solutions even though the efficiency
may be identical to the aqueous-base dispersant containing the
same surface-active compounds and  additives.   The effect of the
type of solvent on the efficiency  of dispersants could be
partially resolved through use of  calibration curves based on
oil and dispersant mixtures.

     Even when the structure of the surfactant and its concentra-
tion in the dispersant solution are known, it is not easy to
determine the fraction of oil introduced into solution by treating
the surface slick.  The possible sinks for surfactants in the
system are a)  the bulk oil  phase  and dispersant solution,
b)  adsorption at the oil/water interface, as free or mixed
micelles, and c)  the bulk water phase, as unaggregated molecules
or in mixed micelles.  It is impossible to determine these con-
centrations separately.  Therefore, development of a meaningful
criterion to rank commercial dispersants without information on
the composition of the formulations remains a major task.

     The trends in experimental data gathered from the dispersion
of 3 oils with 5 dispersants under different  test conditions will
be discussed.  Slicks formed by spilling 300  mis of oil were
dispersed with dispersants in 1:1, 5:1 and 10:1  oi1-to-dispersant
ratios.   The experimental data have been presented in Appendix E.
Figures 33 to 35 are  concentration-time profiles for the disper-
sion of #2 fuel oil with the five  products at 1:1, 5:1, and 10:1
oil-to-dispersant ratios, respectively.  The  maximum oil concen-
tration represents the average oil and dispersant concentration  K
in the bulk fluid immediately after agitation was stopped.  An
analogy to current practice  of using percent  efficiency can be
made by normalizing the maximum oil concentration by the oil
concentration that would result if all the oil was completely
dispersed in the volume of the aqueous phase.   Therefore, 100%
dispersion corresponds to 425 ppm.  The profiles show decreases
in oil concentration  with sampling time for all  products.  The
greatest decrease occurred with product B.  Oil  concentrations
stabilized after about 2 hours.  If all the products are assumed
to contain equal concentrations of the same solvent, the order of
the relative ease of  dispersion of #2 fuel oil is given below

             0/D Ratio         Dispersion of  #2 Fuel Oil

                 1:1                D>B>A>C>E

                 5:1                D>A>B>C>E

                10:1                D>B=A>C>E

                               160

-------
                                                      ro
                                                      to

                                                      Q
                                                      CO

                                                       *
                                                      «c

                                                      to

                                                      u
                                                      3
                                                      -o
                                                      o
                                                      S-
                                                      Q.
                                                      r— -M
                                                      t- 1C
                                                      o a:
                                                      a> c
                                                      3 (O
                                                      CM O)
                                                      =tt= a.
                                                      OQ

                                                      c o
                                                      O •!->
                                                      •I— I
                                                      «/) r—
                                                      S- -I-
                                                      CO
                                                      ro
                                                      (U
(Wdd)  NOIiVdlN33N03 110
                      161

-------
                                              CO
                                              ce
                                              ID
                                              CD
                                              1C
                                                      res
                                                      •o
                                                      c
                                                      ro
                                                      O



                                                      CO
4J
u
3
-o
o
S-
o.
                                                      •^ 03
                                                      o a:

                                                      I— 4J
                                                      0) C
                                                      3 re
                                                      
-------
CO
ID
CD
•^x

UJ
                                                            ro
                                                            UJ

                                                            -O
                                                            c
                                                            ro
                                                            a
                                                           CO
                                                            10
                                                            •!->
                                                            u
                                                            3
                                                            T3
                                                            O
                                                            S_
                                                            Q-
                                                           •r- O
                                                           r- re
                                                           •r- CC
                                                           o
                                                            (!) ro
                                                            3 10
                                                           U- S_
                                                              0)
                                                           CM Q.
                                                           =te 10
                                                             •r-
                                                           <4- Q
                                                            O I
                                                              O
                                                            C •»->
                                                            O I

                                                            t/l -r-
                                                            t- O
                                                            ai
                                                            Q-r—
                                                            10 ..
                                                           •r- O
                                                           O r—
                                                           ID
                                                           CO
                                                            a>
                                                            i-
                                                            en
(Wdd)  NOI1VH1N33N03  110
                          163

-------
This sequence was determined on the basis of stabilized oil
concentration levels in the liquid phase.  Product B indicates
that initial  dispersion is efficient but, because stable emulsions
are not formed,  coalescence occurs readily.   Migration of
coalesced droplets to the surface reduces the concentration  of oil
in the liquid phase.  On the other hand, product D forms fairly
stable emulsions.  The rates of dispersion for products C and E
are lower than for product A.

     The effect  of varying the ratio of oi1-to-dispersant can be
seen in Figures  36 to 40.  In  general,  higher rates of dispersion
occurred as the  volume of dispersant added was increased.  The
rate of dispersion increased significantly when the volume of
Product A added  was increased.  Other dispersants were less  effec-
tive.   Products  B and E did not significantly increase rates of
dispersion when  the oil-to-dispersant ratio  was increased from
10:1 to 5:1.   These results seem contrary to some manufacturers
claims that oil  slicks can be  completely dispersed at oil-to-
dispersant ratios as low as 100:1.  Number 2 fuel oil is one of
the easiest oils to disperse,  but complete dispersion was not
achieved with the dispersants  tested, even when oil-to-dispersant
ratio  of 1:1  was used.

     The concentration-time profiles displayed in Figures 41 and
42 represent  the dispersion of Iranian  crude and #6 fuel  oil,
respectively.  Both oils are more difficult  to disperse than #2
fuel oil.  The degree of difficulty is  reflected in the levels of
oil concentrations that result from the dispersion of the oils
with the 5 products.  The concentration levels for each product
are least when #6 fuel oil is  dispersed.  The behavior of the
products in these tests was similar to  their behavior in the
dispersion of #2 fuel  oil.  Oil  concentration decreased with
sampling time.  This series of tests was performed at a 5:1  oil-
to-dispersant ratio and the following order  indicates relative
dispersion efficiency of the products:
                                                                <
             I rani an Crude       D>B>A>C>E
             #6  Fuel Oil         D > B  > C > A > E

When the relative efficiencies of the products are compared, on
the basis of  these dispersion  studies with the 3 oils, an impor-
tant conclusion  can be drawn:   the relative  efficiencies of  the
products do not  depend on the  type of oil.  This conclusion  is
significant,  also, as manufacturers tend to  classify their
products according to specific oils.  It follows that a good
dispersant will  be effective in dispersing both "easy" and "tough"
oils.

     Product  B was investigated in a further series of tests.
Figure 43 shows  the experimental  data when #2 fuel  oil and
Iranian crude were dispersed in calm water.   Low levels of oil


                              164

-------
                                                CO
                                                o:
                                                UJ
                                                         o
                                                         c
                                                         o
                                                         $_
                                                         
                                                         •r- O
                                                         O 3
                                                          i -a
                                                         o o
                                                         4J f~
                                                          i a.
                                                         o •»->
                                                           •r-

                                                         O

                                                         •(-> -r-
                                                         CJ O
                                                         q_ a>
                                                         LU 3
                                                            u.
                                                         a;
                                                         SZ CO
                                                         K- =«=
                                                         VO
                                                         ro
                                                          QJ
                                                          s_
                                                          CD

                                                         U-
(Wdd)  NOIlVdlN30N03  110
                       165

-------
                                           in
                                                        c
                                                        o
                                                        S-
                                                        O)
                                                        Q.
                                                        Q

                                                        a>
                                                        C
                                                        O
                                          CO
CO
tt:

V—'

LjJ
                                          CSJ
        ro
        
        s.
        d) CO
        Q.
        (O -t->
        •i- o
        Q -3
         I T3
        O O
        •(-» S-
         I Q.

        •r- JC
        O -t->
                                                        o o
                                                        O)
                                                          QJ
                                                          3
                                                        -E 00
                                                        I— =«=
                                                        CO

                                                        O)
                                                        s_
                                                        3
                                                        O)
(Wdd)  NOIlVyiN33NOO 110
                        166

-------
                                                CO
                                                ce
                                                LU
                                                        c
                                                        o
                                                        •r—
                                                        in
                                                        s-
                                                        O)
                                                        ca-
                                                        t/I
                                                        •r-
                                                        O

                                                        
-------
                                                CO
                                                cc
                                                UJ
                                                        M-
                                                        O
                                                         o
                                                        •r—
                                                         10

                                                         0)
                                                         Q.
                                                         V)
                                                        •r—
                                                        a

                                                         01
                                                         o
                                                         o
-(->
c
ro
                                                         0) O
                                                         a.
                                                         VI 4->
                                                         •i- O
                                                         O 3
                                                         I  73
                                                         O O
                                                         •M S_
                                                         I  Q-
                                                         O 4J

                                                         M- S
                                                         O
                                                         O O
                                                         O)
                                                         «*- r-
                                                         l)_ Q}
                                                         UJ 3
                                                           U_
                                                         
                                                         a>
(Wdd)  NOIlVdlN33N03  110
                         168

-------
                                                CO
                                                ct:
                                                LU
                                                          c
                                                          o
                                                          •r—
                                                          in
                                                          s.
                                                          cu
                                                          Q.
                                                          0)
                                                          c
                                                          o
                                                           LU
                                                          a.
                                                          V> 4->
                                                          •i- O
                                                          Q 3
                                                           I T3
                                                          O O
                                                          +» i.
                                                           I D_
                                                          o -u

                                                          «*- 2
                                                          O


                                                          O O

                                                          «*-1—

                                                          LU =1
                                                            u_

                                                          J= CSJ
                                        ISl
                                                          a>
                                                          s_
                                                          CT)
                                                          •i—
                                                          U_
(Hdd)  NOIlVdlN33N03  110
                         169

-------
                                                  CO
                                                  a:
-o
c
«o

a

 *
o

 M
co

 *
<:

i/>
4->
O
3
•o
O
S-
n.

x:
•M
•i—
2

a) o
•o -F-
3 +J
s- re
o a:
                                                          (O C
                                                          •r- ro
                                                          C to
                                                          -< Q.
                                                          00


                                                          C O
                                                          O 4J
                                                          •r-  I
                                                          WJ r—
                                                          S- -r-
                                                          O) O
                                                          a.
                                                          to i —
                                                          (U
                                                          s-
(Wdd)  NOIlVdlN33N03 110
                         170

-------
                                                        c
                                                        to
                                                        o

                                                         *l
                                                        CO
                                                        to
                                                        -o
                                                        O
                                                        i.
                                                        Q_
                                                        •r- (O
                                                        o o:
                                                        (U C
                                                        3 10
                                                        U- (A
                                                          s_
                                                        VO (U
                                                        =*te Q.
                                                          to
                                                        M- v-
                                                        O Q
                                                          I
                                                        c o
                                                        o +-»
                                                        •r- I
                                                        10 i—
                                                        S- f-
                                                        O) O
                                                        Q.
                                                        IO i—

                                                        o ur>
                                                        evi
                                                        
-------
                                        2


                                        O)
                              LU
                                        c
                                       •r-

                                        
                                        s- o
                                        O) 3
                                        Q.-0
                                        l/> O
                                        •r- i-
                                        Q Q.
                                        co



                                        S-
                                        3
                                        cn

                                        U.

-------
 concentrations  are  achieved.   In contrast, oil  concentrations
 shown  in  Figure 44  are  high  because agitation was provided
 throughout  the  sampling  period.  Figures 43 and 44 suggest the
 importance  of providing  mixing to oil/water/dispersant systems,
-to  increase rates of  dispersion.  Higher rates  of dispersion
 result if the system  is  agitated continuously.  This appears to
 be  the ideal method to  disperse oil slicks.  Because of high
 costs  associated with providing mixing energy continuously to
 chemically  treated  oils,  this method of dispersion is impractical
 During high sea states,  mixing energy can come  from gross water
 movements.

      Finally, Figure  45  describes the dispersion of #2 fuel oil
 and Iranian crude in  salt water.  The behavior  in salt water was
 similar to  that in  tap  water.  Except at high dilution sea salts
 may not have a  significant effect on rates of chemical dispersion
 ofoilslicks.
                               173

-------
                                                  CO
                                                  LU
                                                          CO
                                                          •o
                                                           o
                                                           s-
                                                          o.
O)
T3 C
3 O
$- f-
O 4J
   rtJ
C 4^
H3 -i-
•r— U>
C «C

t-i 3
   O
•O 3
C C
CO i-
   4->
r- C
•«- O
O O

r- $-

10 c a: o •r- a to -^ s- o OJ £X>— «/) •• •r- IT) «*


-------
                                               CO
                                               LU
                                                       •u
                                                       •T—

                                                        f


                                                        i.
                                                        V
                                                       +J
                                                        re
                                                        (O
                                                        CO
-o
3
S-
o

c
fO
                                                        C
                                                        03
                                                        •i- O
                                                        O •«-
                                                          +J
                                                        t— ro
                                                        a) o:
                                                        3
                                                        u_ a

                                                        co o
                                                        =*=
                                                        o ca
                                                        •r-
                                                        tr> •(->
                                                        s_ o
                                                        
-------
                           REFERENCES
Adamson, A. W. (1967).  Physical Chemistry of Surfaces.
     edition, Interscience Publishers, New York, 747 p
                                                  2nd
Alofs, D.
     tion
     2(4)
    J. and  R. L.
    of Oil  Slick
      439-443.
Reisbig (1972).
Movement Caused
 An Experimental  Evalua-
by Waves, J.  Phys.  Oceangr.
American Petroleum Institute (1969).
     Conference on Prevention and Control
                                 Proceedings
                                       Oil
                           of the Joint
                   of
     Washington, D. C.

American Petroleum Institute (1971).  __^__
     Conference on Prevention and Control of
          Spills.API
                                 Proceedings
                                       Oil
                           of the Joint
                          Spills.  API
     Washington, D. C.
American Petroleum Institute (1973a).
     Conference on Prevention and Controlof Oil
                                  Proceedings of  the Joint
                                           Spills.  API
     Washington, D. C.

American Petroleum Institute (19735).  Effects of Oil  and
     Chemically Dispersed Oil on Selected Marine Biota - A
     Laboratory Study.  Report by Battele, Pacific Northwest
     Laboratories, Washington, Publication No. 4191, API,
     Washington, D. C.
American Petroleum Institute (1975).  	
     Conference on Prevention and Control of
                                 Proceedings
                                       Oil
                           of the Joint
                          Pol 1ution
                      API
     Washington, D. C
American Petroleum Institute (1977a).  Proceedings of the Oil
     Spi 11  Conference (Prevention, Behavior, Control, Cleanup)
ipil
IPI,
          Washington, D. C
                                 Oil Sill Studies
American Petroleum Institute (1977b)
     Strategies and Techniques.  Proceedings of a Workshop at
     Gurney's Inn, Montauk, L.I
     Publication No. 4286, API
                          _ from  Feb. 24-26,
                          "Washington,  D. C.
                           1976
American Petroleum Institute
     Environment - Phase II.
                        (1977c).  Fate of Oil  in a Water
                        A Dynamic Model of the Mass Balance
     for Released Oil .   Publication No.  4313, API, Washington,
     D. C.
                               176

-------
American Petroleum Institute (1979).  Proceedings of the Oil
     Spill  Conference (Prevention, Behavior, Control, Cleanup).
     API, Washington, D. C.

Anderson, J.R.L. (1967).  Guardian Survey, p. 17, April  25.

Anderson, J.  W., J. M. Neff, B. A. Cox, E. H. Tatem, and G. M.
     Hightower (1974).  Characteristics of Dispersions and Water-
     Soluble  Extracts of Crude and Refined Oils and Their
     Toxicity to Estuarine Crustaceans and Fish.  Mar. Biol., 27:
     75-88.

Baker, J. M.  (ed.) (1976).  Marine Ecology and Oil  Pollution,
     Wiley,  New York.

Bancroft, W.  D. (1915).   J. Phys.  Chem., 19, p. 363.

Berridge, S.  A., R. A. Dean, R. G. Fallows, and A.  Fish  (1968a).
     The Properties of Persistent  Oils at Sea.  J.  Inst. Pet.,
     54: 300-309.

Berridge, S.  A., M. T. Thew, and A.  G. Lorriston-Clarke  (1968b).
     The Formation and Stability of Emulsions of Water in Crude
     Petroleum and Similar Stocks.  J. Inst. Pet.,  54: 333-357.

Blacklaw, J.  R., J. A. Strand, and P.  C. Walkup (1971).   Assess-
     ment of  Oil Spill Treating Agent  Test Methods,  Proceedings
     of Joint Conference on Prevention and Control  of Oil  Spills.
     API, Washington, D. C. 253-261  p.

Blokker, P.  C.  (1964).  Spreading  and  Evaporation of Petroleum
     Products on Water.   Presented at  the 4th International
     Harbour  Conference, June 22-27, Antwerp, Belgium.

Blumer, M.  and J.  Sass (1972).  Oil  Pollution:  Persistence and
     Degradation of Spilled Fuel Oil.   Science, 176: 1120-1122.

Boehm, P. D.  and J. G. Quinn (1973).  Sol ubi 1 ization of  Hydro-
     carbons  by Dissolved Organic  Matter in Seawater.  Geochim.
     Cosmochin. Acta, 37: 2459-2477.

Boehm, P. D.  and J. G. Quinn (1974).  The Solubility Behavior of
     No. 2  Fuel Oil in Seawater.  Mar. Poll. Bull., 5(7):  101-105,

Boesch, D.  F.,  C.  H.  Hershner, and J.  H. Milgram (1974).  Oil
     Spills  and the Marine Environment.  Ballinger  Publishing
     Co., Cambridge,  Mass.

Bohon, R. L.  and W. F. Claussen (1951).  The Solubility  of
     Aromatic Hydrocarbons in Water.  J. Amer. Chem. Soc., 73:
     1571-1578.
                              177

-------
Bond, W.  N.  (1927).   Bubbles and Drops and Stokes1 Law.  Phil.
     Magazine, (7),  4, p. 889.

Boylan, D. B.  and B. W. Tripp (1971).  Determination of Hydro-
     carbons in Seawater Extracts of Crude Oil and Crude Oil
     Fractions.  Nature, 230: 44-47.

Broecker, W. S. and  T. H. Peng  (1974).  Gas Exchange Rates
     Between Air and Sea.  Tellus, 26, 1-2, 21-35.

Buckmaster,  J. (1973).  Viscous-gravity Spreading of an Oil
     Slick.   J. Fluid Mech., 59 (3), 481-491.

Burwood,  R.  and 6.  C. Speers (1974).  Photo-oxidation as a Factor
     in the  Environmental Dispersal  of Crude Oil.  Est. and
     Coast.  Mar. Sci., 2, 117-135.

Butler, J. N.  (1975).  Evaporative Weathering  of Petroleum
     Residues:  The  Age of Pelagic Tar. Mar. Chem., 3, 9-21.

Canevari, G. P. (1969a).  The Role of Chemical Dispersants in Oil
     Clean-up.  In:   Oil on the Sea.  (D. P. Hoult, ed.) Plenum
     Press,  New York, pp. 29-51.

Canevari, G. P. (1969b).  General  Dispersant Theory.  Proceedings-
     Joint Conference on Prevention  and Control  of Oil Spills,
     API, Washington, D. C., pp.  171-177.

Canevari, G. P. (1973).  Development of the Next Generation
     Chemical  Dispersant.  Proceedings of the  Joint Conference on
     Prevention and  Control of  Oil Spills.  API, Washington, D. C,
     pp.  305-308.

Canevari, G. P. (1975).  A Review of the Utility of Self-Mixing »
     D.ispersants in  Recent Years.   Proceedings - Conference on
     Prevention and  Control of  Oil Pollution!API, Washington,
     D. C.,  pp. 337-342.

Canevari, G. P. (1978).  Some Observations on  the Mechanism and
     Chemistry Aspects of Chemical Dispersion, Chemical Disper-
     sants for the  Control  of Oil  Spills (L. T.  McCarthy, G. P~7
     Lindblom, and  H. F. Walter,  e'ds.), ASTM STP 659, American
     Society for Testing and Materials, Philadelphia, Pa. pp.
     5-17.

Carberry, J. J. (1976).  Chemical  and Catalytic  Reaction
     Engineering.   McGraw-Hill  Book. Co., New  York, p. 642.

Chan, A.  F., D. F.  Evans and E.  L. Cussler (1976).  Explaining
     Solubilization  Kinetics, AIChE  J., 22(6), 1006-1012.
                               178

-------
,       Chen, E. C., J. C. K. Overall and C. R. Phil 1ips(1974).  Spread-
            ing of Crude Oil on an Ice Surface.  Can. J.  Chem. Eng., 52,
            pp. 71-74.

       Chen, E. C. and D. K. Charles (1976).  Surface Tension Spreading
            of Crude Oil on Ice.  Mar. Sci. Comm. 2, 1, 1-11.

       Cochran, R. A. and P. R. Scott (1971).  The Growth of Oil Slicks
            and Their Control  by Surface Chemical Agent.   J. Pet.
            Technol., 23, 781-787.

       Cohen, Y., W. Cocchio and D. Mackay  (1978).  Laboratory Study of
            Liquid-phase Controlled Volatilization Rates in Presence of
            Wind Waves, Environ. Sci. Technol. 12(5), pp. 553-558.

       Danckwerts, P. V. (1951). Significance of Liquid-Film Coefficients
            in Gas Absorption.  Ind. Eng. Chem., 43(6), pp. 1460-1467.

       Davies, J. T. and E. K. Rideal (1963).  Interfacial Phenomena.
            Academic Press, New York, pp. 480.

       Eqanhouse, R. P. and J. A. Calder (1976).  The Solubility of
            Medium Molecular Weight Aromatic Hydrocarbons and the
            Effects of Hydrocarbon Co-solutes and Salinity.  Geochim.
            Cosmochim. Acta, 40: 555-561.

       Elworthy, P. H., A.  T.  Florence, and C. B. Macfarlane (1968).
            Solubilization by Surface Active Agents.  Chapman and Hall,
            Ltd., London, pp.  11-116.

       Estes, J. E. and B.  Golomb (1970).  Oil Spills:  Methods for
            Measuring their Extent on the Sea Surface.  Science, 169,
            676-678.

       Fallah;, M. H. and R. M. Stark (1976a).  Movement of Spilled Oil "
            at Sea.  J. Mar. Technol. Soc., 10, 3-18.

       Fallah, M. H. and R. M. Stark (1976b).  Random Drift of an
            Idealized Oil Patch.  Ocean Engng., 3: 83-97.

       Fallah,  M. H. and R. M. Stark (1976c).  A Probabilistic Model
            for Dispersive Oil Losses Beneath Oil Booms.   Ocean Engng.,
            3: 383-390.

       Fannelop, T. K. and G.  D. Waldman (1971).  The Dynamics of Oil
            Slicks on Creeping Crude.  Paper No. 71-14, American Inst.
            of Aeronautics and Astronautics, New York.

       Fay, J. A. (1969).  The Spread of Oil Slicks on a  Calm Sea.  In:
            Oil on the Sea (D. P. Hoult, ed.), Plenum Press, New York,
            p p. 5 3 - 6 3 .
W

                                      179

-------
Fay, J.  A. (1971).  Physical  Processes in the Spread of Oil on a
     Water Surface.   Proceedings of the Joint Conference on
     Prevention and Control  of Oil  Spills"API, Washington,
     0.  C., pp. 463-467.

Fitzgerald, D. E. (1977).  Utilization of Dispersants in Offshore
     Areas.  Proceedings - 1977 Oil Spill Conference (Prevention,
     Behavior, Control, Cleanup).   API, Washington,  D".C.,  pp.
     395-398.

Forrester, W.  D.  (1971).  Distribution of Suspended  Oil Particles
     Following the Grounding  of the Tanker Arrow, J. Mar.  Res.
     29(2), 151-170.

Frankenfeld,  J. W. (1973).  Factors Governing the Fate  of  Oil  at
     Sea:   Variations in the  Amounts of Dissolved or Dispersed
     Materials During the Weathering Process.  Proceedings  of the
     Joint Conference on Prevention and Contro1~bf Oil  Spills.
     API,  Washington, D. C.,  pp. 485-495.

Garrett, W. D. (1969).   Confinement and Control  of Oil  Pollution
     on  Water  with Monomolecular Surface Films.   Proceedings of
     the Joint Conference on  the Prevention and  Control of  Oil
     Spil1s.API, Washington, D.  C., pp. 257-261.

Garrett, W. D. and W. R. Barger (1970).  Factors Affecting  the
     Use of Monomolecular Surface  Films to Control Oil  Pollution
     on  Water, Env.  Sci. Techno!.,  4(2), pp.  123-127.

Giles,  C.  H.  and  S.  D.  Forrester (1970).  Wave Damping:  The
     Scottish  Contribution,  Chem.  and Ind., pp.  80-87.

Glaeser, J. L. and G. P. Vance (1971).  A Study  of the  Behavior
     of  Oil Spills an the Arctic.   U. S. Coast Guard Report No.
     7.14108/A/001-002,  NTIS,  Springfield, Va.

Gordon,  J. E.  and R.  L. Thome (1967a).  Salt Effects on the
     Activity  Coefficients of Naphthalene in  Mixed Aqueous
     Electrolyte  Solution.  I. Mixtures of Two Salts, J. Phys.
     Chem., 71(31),  pp. 4390-4399.

Gordon,  J. E.  and R.  L. Thome (1976b).  Salt Effects on Non-
     electrolyte  Activity Coefficients in Mixed  Aqueous Electro-
     lyte  Solutions.   II.  Artificial and Natural Seawaters.
     Geochim.  Cosmochim. Acta. 31,  pp. 2433-2443.

Gordon,  D. C., P. D.  Keizer  and N.  J. Prouse  (1973). Laboratory
     Studies  of the Accommodation  of Some Crude  and  Residual Fuel
     Oils  in  Seawater.   J. Fish Res. Board Can., 30, pp. 1611-
     1618.
                               180

-------
     '•" Gruenfel d,  M.  (1973).   Extraction  of Dispersed Oils from Water
*****          for Quantitative  Analysis  by  Infrared Spectrophotometry.
            Env.  Sci.  Techno!.,  7(7),  pp.  636-639.

       Harrison,  W.,  M.  A.  Winnik,  P.T.Y.  Kwong,  and D.  Mackay (1975).
            Crude  Oil  Spills:   Disappearance of Aromatic and Aliphatic
            Components from Small  Sea-Surface Slicks.  Env.  Sci.  Techno!.,
            9(.30),  pp. 231-234.

       Hartley,  G.  S.  (1976).   Micelles-Retrospect  and Prospect.
            Mice!1ization,  Solubilization,  and Microemulsions (K.  L.
            Mittal,  ed.)  Plenum Press,  New  York,  pp. 23-43.

       Hepple,  P.,  ed. (1971).   Mater  Pollution by  Oil,  Inst. of
            Petroleum, London.

       Hermann,  R.  B.  (1972).   Theory  of  Hydrophobic Bonding, II.   The
            Correlation  of  Hydrocarbon  Solubility in Water with Solvent
            Cavity  Surface  Area.   J.  Phys.  Chem., 76(19),  pp. 2754-2758.

       Hess,  K.  W.  and C.  L.  Kerr  (1979).   A Model  to Forecast the
            Motion  of  Oil  on  the  Sea,  Proceedings of the Joint Confer-
            ence  on  Oil  Spill  (Prevention,  Behavior, Control  and Clean-
            up) .API, Washington,  D.  C.,  pp.  653-663.

       Hinze,  J.  0.  (1955).   Fundamentals  of the  Hydrodynamic Mechanism
            of  Splitting  in Dispersion  Processes.  AIChE J.,  1, 289-295.

       Hollinger,  J.  P.  and R.  A.  Mennella  (1973).   Oil  Spills:
            Measurements  of their  Distributions and Volumes  by Multi-
            frequency  Microwave Radiometry.   Science, 181: 54-56.

       Hoult,  D.  P.,  ed.  (1969).   Oil  on  the Sea.  Plenum  Press,  New
            York.
                                                                       k.
       Hoult,  D.  P.  (1972).   Oil  Spreading  on the Sea.  Ann.  Rev.  Fluid
            Mech.,  4,  pp.  341-368.

       Hoult,  D.  P.  and  W.  Suchon  (1970).   The Spread of Oil  in a  Channel.
            Fluid  Mech.  Lab.,  Dept.  of  Mech. Eng.,  Mass. Inst. Techno!.
            25  pp.

       Hu,  S.  and  R.  C.  Kintner  (1955).   The Fall  of Single Liquid  Drops
            Through  Water.   AIChE  J.,  1,  p.  42.

       Jeffrey,  P.  F.  (197!).   Large  Scale  Experiments on  the Spreading
            of  Oil  at  Sea  and  its  Disappearance by  Natural Factors.
            Proceedings  of  the  Joint  Conference on  Prevention and
            Control  of Oil  Spi!1s.API,  Washington, D.  C.,  pp. 469-474.
                                      181

-------
       ,  H.  P.  (1950).   Effect of Electrolytes Upon the Solubili-
     zation  of  Hydrocarbons and Polar Compounds.   J. Amer. Chem.
     Soc.,  72:  3780-3785.

Kolmogorov,  A.  N.  (1949).   On the Disintegration  of Drops in
     Turbulent  Flow.   Doklady Akad.  Nauk S.S.S.R., 66, p. 825.

Langmuir, I.  (1933).   Oil  Lenses on  Water and the Nature of Mono-
     molecular  Expanded Films.  J. Chem. Phys. 1, pp. 756-776.

Leibovich,  S.  (1975).   A Natural Limit to the Containment and
     Removal  of Oil  Spills at Sea.  Ocean Engng., 3, pp. 29-36.

Leinonen, P.  J.,  D.  Mackay and C. R.  Phillips (1971).  A Correla-
     tion for the  Solubility of Hydrocarbons in Water.  Can. J.
     Chem.  Eng.,  49,  pp. 288-290.

Leinonen, P.  J.  and  D.  Mackay (1973).  The Multicomponent Solu-
     bility  of  Hydrocarbons in Water.  Can.  J. Chem. Eng., 51,
     pp.  230-233.

Lien, T.  R.  and C.  R.  Phillips (1974).  Determination of Particle
     Size Distribution  of  Oil-in-Water Emulsions  by Electronic
     Counting.   Env.  Sci.  Techno!. 8(7), pp. 558-561.

Liss, P.  S.  (1973).   Processes of Gas Exchange Across an Air/
     Water  Interface.   Deep-Sea Res., 20, pp. 221-238.

Liss, P.  S.  and P.  6.  Slater (1974).   Flux of Gases Across the
     Air-Sea  Interface.  Nature, 247: 181-184.

Lissauer, I.  M.  and  D.  L.  Murphy (1979).  Oil Spill Forecasting -
     Where  is it  Going.  Proceedings  of the Joint Conference on
     Oil  Spill  (Prevention", Behavior, Control, and Clean-up).
     API, Washington,  D. C., pp. 649-652.

Lu, B.C'.Y.  and  J.  Polak (1973).  A Study of the Solubility of Oil
     in Water.   Technology Development Report #EPS-4-EC-76-l,
     EPS, Dept.  of the  Environment,  Ontario, Canada, p. 25.

Lyons,  C. G.  and  E.  K.  Rideal (1929).  On the Stability of
     Unimolecular  Films, III.  Dissolution in Alkane Solutions.
     Proc.  Royal  Soc.,  Section A, CXXIV, p.  334.

Lysyj,  I. and E.  C.  Russell (1974).   Dissolution  of Petroleum
     Derived  Products  in Water.  Water Res., 8: 863-868.

Mackay, D.  and  P.  J.  Leinonen (1977).  Mathematical Model of the
     Behaviors  of  Oil  Spills on Water with Natural and Chemical
     Dispersion.   Report #EPS-3-EC-77-l9, EPS, Dept. of the
     Environment,  Ontario, Canada, p. 84.
                              182

-------
Mackay, D.  and R.  S.  Matsugu (1973)
     Hydrocarbon Spills on Land and
     51,  pp.  434-439.
                                      Evaporation Rates of Liquid
                                    Water.   Can.  J.  Chem. Eng.,
Mackay, D.  and A.  W.  Wolkoff (1973).   Rate of Evaporation of Low-
     Solubility Contaminants from Water Bodies to Atmosphere,
     Environ.  Sci. Techno!., 7, pp.  611-614.

Mackay, D.,  J. S.  Nadeau and C. Ng (1978).  A Small-Scale Labora-
     tory Effectiveness Test.   Chemical Dispersants for the
     Control  of Oil  Spills (L.  T. McCarthy, G. P. Lindblom, and
     H. F.  Walter, eds.) ,  ASTM  STP 659, American Society for
     Testing  and Materials,  Philadelphia,  Pa., pp.  35-49.
Marquardt, D. W. (1963).   An Algorithm for Least
     tion of Nonlinear Parameters.   J. Soc. Ind.
     11(2), pp.  431-441 .
                                                 Squares Estima-
                                                 Appl.  Math. ,
McAuliffe, C
     Nature,
              (1963).   Solubility in Water of
             200 (4911 ): 1092-1093.
CTC9
Hydrocarbons
McAuliffe, C.  (1966).
     paraffin, Olefin,
     Hydrocarbons.   J.
                       Solubility in Water of Paraffin,  Cyclo-
                       Acetylene, Cyclo-olefin, and Aromatic
                       Phys.  Chem.,  70(4): 1267-1275.
McAuliffe, C.  (1969a).   Determination of Dissolved Hydrocarbons
     in Sub-surface Brines.   Chem.  Geol. ,  4(1969): 478-479.
McAuliffe, C.  (1969b).   Solubility in Water of Normal
     Alkane Hydrocarbons, Science, 163:  478-479.

McAuliffe, C.  (1977).   Evaporation and Solution of C,
                                                         and C
                                                              10
        and C-

         Fate
                                                            and
     Hydrocarbons from Crude Oils on the Sea Surface.
     Effects of Petroleum Hydrocarbons in Marine Organisms and  •*
     Ect) systems .   (D.  A.  Wol fe, ed. ) .  Pergamon Press,  New York,
     pp. 33-372.

McAuliffe, C.,  A. E.  Smalley, R.  D.  Groover, W.  M.  Welsh, W.  S.
     Pickle and G.  E.  Jones (1975).   The Chevron Main  Pass Block
     41  Oil Spill:   Chemical and  Biological  Investigations.
     Proceedings  of the Joint Conference on  Prevention and
     Control of Oil  Spil 1 s.  API, Washington,  D. C.,  pp.  521-529,

McCarthy, Jr.,  L. T.  (1977).  Considerations for Field Use of
     Dispersants.  Proceedings Oil  Spill Conference (Prevention.
     Behavior,  Control, Cleanup).  API,  Washington, D. C. , pp .
     399-401 .

McCarthy, Jr.,  L. T.,  I.  Wilder and  J. S. Dorrler (1973).
     Standard  Dispersant  Effectiveness and Toxicity Tests. .
     Environmental  Protection Technology Series  Report #EPA-R2-
     73-201 , pp.  57.

                               183

-------
 McCarthy,  Jr.  L.  T.,  6.  P.  Lindblom  and  H.  F.  Walter  (1978).
      Chemical  Dispersants  for  the  Control of  Oil  Spills.   ASTM
      Special Technical  Publication No. 659, American  Soc i e ty  for
-"'     Testing and  Materials,  Philadelphia, Pa.,  p.  318.

 McDevit, W.  F. and  F. A. Long  (1952).  The  Activity Coefficient
      of Benzene in  Aqueous  Salt  Solutions.  J.  Amer.  Chem.  Soc.,
      74: 1773-1777.

 McMinn, T.  J.  (1972).   Crude Oil Behavior on  Arctic Winter  Ice.
      U. S.  Coast  Guard  Report  No.  73418, NTIS,  Springfield, Va .

 Milz,  E. A.  and J.  P. Fraser (1972).  A  Surface-Active  Chemical
      System  for Controlling  and  Recovering  Spilled Oil  from the
      Ocean.  J. Pet.  Technol., 24, pp. 255-262.

 Moore, S.  F. and  R. L.  Dwyer (1974).  Effects  of  Oil  on  Marine
      Organisms:   A  Critical  Assessment of Published Data.   Water
      Res.  8, pp.  819-827.

 Murphy, T.  A.  and L.  T.  McCarthy Jr.  (1969).   Evaluation  of the
      Effectiveness  of Oil  Dispersing  Chemicals, Proceedings of
      the Joint Conference  on Prevention  and Control of  Oil  Spills.
      API,  Washington, D. C., pp. 199-207.

 Murray, S.  P.  (1972).   Turbulent Diffusion  of  Oil  in  the  Ocean.
      Limnol . Oceanogr.,  17(5): 651-660.

 Murray, S.  P.  (1975).   Wind  and  Current  Effects on Large-Scale
      Oil Slicks.  Proceedings  of the  Offshore  Technology
      Conference.  Dallas, Texas.  Paper #OTC  2389,  pp.  523-533.

 National Academy  of Sciences (1973a).  Background  Papers  for  a
      Workshop  on  Inputs. Fates,  and  Effects of  Petroleum  in the ^
      Marine  Environment. Vo~  lj Report  prepared  for  the  Ocean  "
      Affairs Board, Office  of  Naval  Research,  Washington,  D.  C.

 National Academy  of Sciences (1973b).  Background  Papers  for  a
      Workshop  on  Inputs. Fates and Effects  of  Petroleum  in  the
      Marine  Environment. Vol.  11,  Report prepared  for the Ocean
      Affairs Board, Office  of  Naval  Research,  Washington,  D.  C.

 National Academy  of Sciences (1975).  Petroleum in the  Marine
      Environment.   Workshop  on Inputs. Fates,  and  the Effects of
      Petroleum in the Marine Environment.   NAS, Washington, D. C. ,
      p. 107.

 Neely, W.  B.,  G.  E. Blau and T. Alfrey,  Jr. (1976).   Mathematical
      Models  Predict Concentration  -  Time Profiles  Resulting from
      Chemical  Spill in  a River.  Env. Sci.  Techno!.,  10(1): 72-76.
                               184

-------
Nelson-Smith, A. (1972).  Oil  Pollution and Marine Ecology.  Elek
     Science, London.

O'Brien, J. A. (1970).  Oil Spreading on Water from a Stationary
     Leaking Source.  The Chem. Engr., CE, pp. 407-409.

Oda, A. (1969).  Laboratory Evaluation of Chemical Oil Disper-
     sants, Proceedings - Joint Conference on Prevention and
     Control of Oil Spil1s.API, Washington, D. C., pp. 199-207.

Osamor, C. A. and R. C. Ahlert (1981).  A Laboratory System for
     Evaluating the Effectiveness of Dispersants and Criteria for
     Ranking Chemical Dispersants.  Submitted to Environ. Sci.
     Technol .

Parker, C. A., M. Freegarde and C. G. Hatchard (1971).  The Effect
     of Some Chemical and Biological Factors on the Degradation
     of Crude Oil at Sea.  In:  Water Pollution by Oil (P.  Hepple,
     ed.) Inst. of Petroleum,  London, pp.237-244.

Peake,  E. and 6. W. Hodgson (1966).   Alkanes in Aqueous Systems.
     I. Exploratory Investigations on the Accommodation of
     C20~C'33 n-alkanes in Distilled  Water and Occurrence in
     Natural Water Systems.  J. Amer. Oil Chem. Soc., 43(4):
     215-222.

Peake,  E. and G. W. Hodgson (1967).   Alkanes in Aqueous Systems.
     II. The Accommodation of C-ip'^se ""Alkanes in Distilled
     Water.  J. Amer. Oil Chem. Soc., 44(12): 696-712.

Pilpel, N. (1968).  The Natural Fate of Oil in the Sea.
     Endeavour, 100(27): 11-13.

Poliakof.f, M. Z. (1969).  Oil  Dispersing Chemicals:  A Study of w
     the Composition. Properties and Use of Chemicals for
     Dispersing Oil Spills.  Prepared for FWPCA on Contract No.
     14-12-549.  U. S. Dept.  of Interior, Washington, D. C.,
     p. 27.

Pomerantz, P., W. C. Clinton  and W.  A. Zisman (1967).  J. Colloid
     Interface Sci., 24: 16-28.

Prudich, M. E. and J. D. Henry, Jr.  (1978).  The Mechanisms of
     Transfer of Hydrophobic  Coated  Mineral Matter Particles from
     a  Hydrocarbon to an Aqueous Phase, AIChE J., 24(5): 788-795.

Reisbig, R. L. and J. M. Pottinger (1973).  The Simulation  of
     Deep Water Waves for Oil  Drift  Studies.  Proceedings of the
     Missouri Academy of Science.
                               185

-------
Renzoni, A.  (1973).   Influence of Crude Oil,  Derivatives and
     Dispersants on Larvae.  Mar. Poll. Bull., 4, pp.  9-13.

-Ross, C. W., P. B. Hildebrand and A. A. Allen  (1978).  Logistic
     Requirements  for Aerial Application of Oil Spill  Dispersants
     Chemical  Dispersants for the Control of  Oil Spills. ASTM
     STP 659,  (L.  T.  McCarthy, Jr., G. P. Lindblom and H.  F.
     Walter, eds.) American Society for Testing and Materials,
     Philadelphia, Pa., pp. 66-80.

Rossini, F.  D.  (1960).  Hydrocarbons in Petroleum, J.  Chem.
     Educ. 37(11) : 554-561.

SAS  Institute  (1979).  A  User's  Guide  to SAS  Programming.  SAS
     Institute, Raleigh,  N. C.

Schwartzberg,  H. 6. (1971).  The Movement of  Oil Spills.
     Proceedings of the Joint Conference on Prevention and
     Control of Oil Spills.API, Washington,  D. C., pp. 489-494.

Seifert, W.  K.  (1969).  Effect of Phenols on  the Interfacial
     Activity  of Crude Oil  (California) Carboxylic Acids and the
     Identification of Carbazoles and  Indoles, Anal. Chem., 41,
     pp. 562-568.

Seifer, W. K.  and  W.  6. Howells  (1969).  Interfacially Active
     Acids in  a California  Crude Oil.  Isolation of Carboxylic
     Acids and  Phenols, Anal. Chem., 41, pp.  554-562.

Shackleton,  L.R.B., E. Douglas and T.  Walsh (1960).  Pollution of
     the Sea by Oil,  Trans. Inst. Mar. Engrs.  (U.K.),  72(11),
     pp. 409-439.

Shaw, D. G.  (1977).   Hydrocarbons in the Water Column.  Fate and
     Effects of Petroleum Hydrocarbons in Marine Organisms and  "
     Ecosystems (D. A. Wolfe, ed.), Pergamon  Press. New York,
     pp. 8-18.

Shaw, D. J.  (1970).   Introduction to Colloid  and Surface Chemis-
     try, 2nd  Edition, Butterworths, London.

Shukla, D. K.  and  R.  M. Stark (1974).  Random  Movement of  Oil
     Patches.   Sci. of the  Total Environ., 3,  pp. 117-125.

Sieburth, J. M., et al.   (1976).  Dissolved Organic Matter and
     Heterotrophic Microneuston  in the Surface Microlayers of
     the North  Atlantic.  Science, 194, pp. 1415-1418.

Smith,  C. L. (1977).  Determination of the Leeway of Oil Slicks.
     In:  Fate  and Effects  of Petroleum Hydrocarbons in Marine
     Organisms  and Ecosystems (D~ A. Wolfe, ed.), Pergamon Press,
     New York,  pp. 351-362.

                               186

-------
Smith, C. L. and W. 6. Maclntyre (1971).  Initial Aging of Fuel
     Oil  Films on Seawater.  In:  Proceedings of the Joint
     Conference on Prevention and Control of Oil Spills.  API,
""""   Washington, D. C., pp. 457-461.

Smith, G. F. (1978).   Techniques for Mixing Dispersants with
     Spilled Oil, Environmental Protection Technology Series,
     Report IERA-600/2-78/128, NTIS, Springfield, Va., p. 38.

Smith, G. F. and W. E. McCracken (1977).  Techniques for Mixing
     Dispersant Treated Oil Slicks into the Water.  In:
     Proceedings of Oil Spill Conference (Prevention, Behavior,
     Control ,  CleanupTAPI, Washington, D. C., pp. 403-406.

Smith, J. E. (1968).   Torrey Canyon, Pollution and Marine Life.
     Cambridge University Press, Cambridge,England,p~.196.

Stolzenbach, K. D., 0. S. Madsen, E. E. Adams, A. M. Pollack and
     C. K.  Cooper (1977).  A Review and Evaluation of Basic
     Techniques for Predicting the Behavior of Surface Oil
     Slicks.  Report #222.  School  of Engineering, Massachusetts
     Institute of Technology, Cambridge, Mass., p. 321.

Straughan,  D.  (1972).   Factors Causing Environmental Changes
     after an  Oil Spill.  J. Pet. Techno!., 24, pp. 250-254.

Sutton, C.  and J. A.  Calder (1974).  Solubility of Higher Molecu-
     lar Weight n-Paraffins in Distilled Water and Seawater.
     Env. Sci. Techno!., 8(7), pp.  654-657.

Sutton, C.  and J. A.  Calder (1975).  Solubility of Alkyl Benzenes
     in Distilled Water and Seawater at 25°C.  J. Chem. Eng.
     Data,  20(3), pp.  320-322.

Tanford,  C. (1973).  The Hydrophobic Effect. Wiley, New York,
     p. 200.

Tayfun, M.  A.  and H.  Wang (1973).  Monte Carlo Simulation of Oil
     Slick Movements.   ASCE J., Waterways, Harbors and Coastal
     Eng. Div., 99, pp. 309-324.

Teeson, D., F. M. White and H. Schenck, Jr. (1970).  Studies of
     the Simulation of Drifting Oil by Polyethylene Sheets.
     Ocean Engng., 2,  pp. 1-11.

Thibodeaux, L. J. (1977).  Mechanisms and Idealized Dissolution
     Models for High Density (p > 1), Immiscible Chemicals
     Spilled in Flowing Aqueous Environments.  AIChE J., 23 (4):
     544-553.
                               187

-------
Ward, A.F.H.  and
     Tensions of
     Effects.  J
      L.  Tordai  (1946).   Time-Dependence of Boundary
      Solutions.   1.   The Role of Diffusion in Time-
       Chem.  Phys.,  14(7),  pp. 453-461.
Warner, J.  L., J.
     the Movement
     No. OTC-155,
       W.  Graham and R.  G.  Dean (1972).   Prediction of
       of  an Oil Spill  on the Surface of Water, Paper
       Offshore Techno!.  Conf., Dallas,  Texas.
Weiskopf, F.  B.  and M.  S.  Uzuner (1977).   Oil  Slick Spreading
     Beneath  a Uniform Ice Cover in Presence of a Current
     Proceedings of the Joint Conference  on Oil Spills (Preven-
     tion, Behavior, Control  and Clean-upTTAPI, Washington,
     D. C.,  pp.  297-300.

Witner, F. E.  and A. Gollan (1973).  Determination of Crude Oil
     Concentration and Size Distribution  in Ship Ballast Water,
     Env. Sci. Technol,,  7, pp.  945-948.

Wolfe,  D. A.,  ed. (1977).   Fates and Effects of Petroleum in
     Marine  Organisms and  Ecosystems, Pergamon Press, New York.

Wu, J.  (1968).  Laboratory Studies of Wind-Wave Interactions,
     J. Fluid  Mech., 34,  pp.  91-112.
Zobell, C.
     Sea.
E.  (1969).
Proceedings
Microbial  Modification of Crude Oil  in the
of the Joint Conference on Prevention and
     Control  of Oil  Spills.API, Washington, D.  C., pp.  317-326.
                               188

-------
    100.0
LU
o
m
cc
o
CO
m
                                        l.R cell: 1 cm


                                 Corr. Coefficient:  .9999
               50.0   100.0  150.0  200.0  250.0 300.0
                   OIL CONCENTRATION  (PPM)
     Figure Al .   Calibration Curve for Nigerian Crude

                 using  1  cm cuvets.
                          189

-------
    100.0
        Ifl
       • L
     60.0.
LU
O


DO
cn
o
CO
GO
40.0
     20.0
      0.21
                                 . 1.R cell: 5 cm

                           Corr. Coefficient: .9998

                          _	1	i	
        0.0     20.0    40.0    60.0     80.0    100.0


                   -OIL CONCENTRATION (PPM)
     Figure A2.  Calibration Curve for Nigerian Crude
                using  5  cm cuvets.
                          190

-------
O

<
00
a:
o

CD
    100.0
       .0
     60.0
40.0
     20.0
      0.3
                                        l.R cell: 10 cm

                                 Corr. Coefficient: .9995
        0.0     10.0    20.0    30.0     40.0    50.0


                   OIL CONCENTRATION  (PPM)
     Figure A3.   Calibration Curve for  Nigerian Crude

                 using 10 cm cuvets.
                          191

-------
            APPENDIX B



SOLUBILITY DATA IN TAP  WATER  AT  25C
Time
(Days)
0
1
2
3
4
5
6
7
8
9
10
n
12
13 -
14
15
16
17
18
19
20
21
Oil Concentration (PPM)
Nigerian
0.0
1 .5
3.5
4.3
4.5
3.5
3.2
3.0
2.8
2.4
2.4
2.4
2.2
- 2 . 1
1 .8
1 .6
1 .5
-
1 .5
1 .3
1 .3
1 .2
Iranian
0.0
6.0
9.8
11.6
15.0
10.0
6.8
3.5
2.7
2.7
2.6
2.5
2.3
2.3
2.1
2.1
2.0
-
1 .8
1 .7
1 .5
1 .5
#2 Fuel
0.0
0.7
0.9
1 .0
1.2
1 .3
1 .4
1 .7
1 .5
1 .3
1 .2
1 .0
1 .0
1 .0
0.9







#6 Fuel
0.0
0.3
0.4
0.5
0.6
0.9
0.5
0.4
0.4
0.4
0.4
0.4
0.3
0.3
0.3







La Rosa
0.0
0.8
1 .1
1 .5
2.3
3.4
2.8
2.3
2.4
2.1
2.0
1.9
1 .9
1 .6
1 .5
1 .2






Sahara
0.0
14.0
24.9
20.7
13.4
11 .9
10.8
7.0
3.7
3.1
2.6
2.2
0.9^
0.9^
0.8







               192

-------
V..-
                            APPENDIX B  (Continued)



                      SOLUBILITY DATA IN TAP WATER AT 25C
Time
(Days)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Oil Concentration (PPM)
Laguni 1 1 as
0.0
0.5
0.7
1 .1
1 .2
-
1 .9
1 .9
2.3
2.5
2.2
2.1
1 .9
1 .7
1 .6
1 .5
Suniland
0.0
61 .2
15.2
12.4
8.9
6.9
6.5
6.3
5.4
4.7
4.8
4.2
4.1
4.1
4.2

8% Crude
0.0
73.4
78.6
73.9
47.6
37.3
36.0
32.9
28.4
26.8
26.3
20.6
17.9
13.0
11 .8

Alaska
0.0
59.1
81 .6
152.1
75.4
70.3
60.9
52.6
28.5
20.1
19. 8
18.2
15.4
15.2
13.9
14.6
Brass
0.0
8.2
8.5
11.0
13.3
37.6
91 .5
108.9
176.0
159.1
77.9
52.6
11 .3
5.8
5.3

Arzew
0.0
12.5
13.0
16.0
28.2
58.2
83.4
131 .8
196.0
89.5
30.5
23.5
13.5
13.4
11 .0

                                      193

-------
                      APPENDIX  C
-SOLUBILITY  DATA  FOR  SELECTED OILS  IN SALT WATER AT 25C
Time
(Days)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Oil Concentration (PPM)
Nigerian
0.0
0.4
0.6
0.9
0.9
0.9
1 .0
1 .1
1 .2
1.2
-
-
1 .3
1 .3
1 .5
Iranian
0.0
0.6
0.7
1 .0
1 .1
1 .2
1.4
1 .5
1 .4
1 .4
1 .4
1 .4
1 .4
1 .4
1 .3
#2 Fuel
0.0
0.9
1 .0
1.2
1 .4
1 .5
1 .5
1 .4
1 .4
1 .3
1 .2
1 .0
1 .0
1 .0
1 .0
La Rosa
0.0
0.7
1.2
1 .3
1 .4
1 .4
1 .6
1 .7
2.0
2.3
-
-
2.5
2.5
2.4
Alaska
0.0
0.4
0.7
0.8
1 .1
0.6
0.5
0.6
0.5
0.5
0.5
0.5
0.5
0.5
0.5
Brass
0.0
0.7
0.7
1 .2
1 .4
1 .3
1 .3
1 .2
1 .2
1 .2
-
-
1 .1
1 .1
1 .0
                         194

-------
                    APPENDIX D

EXPERIMENTAL DATA FOR THE SPREADING OF OIL  SLICKS
               ON CALM WATER AT 20C
TABLE 01.   AREA OF OIL SLICKS FORMED BY  ARZEW CRUDE
Time
(mins)
0
2
3
4
5
6.
8
10
15
20
25
30
35
Area (ft2)
25+
1.75*
0
3.89
4.39
4.94
4.96
5.08
5.20
5.22
5.25
5.29
5.32
5.35
5.44
50+
2.00*
0
4.47
4.81
5.14
5.61
5.94
6.08
6.20
6.27
6.33
6.38
6.40
6.46
75+*
3.27
0
4.79
5.27
5.82
6.30
6.80
6.97
7.05
7.20
7.35
7.50
7.65
7.79
100+
6.08
0
4.95
5.38
5.75
6.14
6.61
6.89
7.23
7.48
7.67
7.82
-
7.99
+ Volume of oil  spilled,  cm
* Duration of spill,  min
                         o




                        195

-------
TABLE D2.   AREA OF OIL SLICKS FORMED BY  BRASS  RIVER  CRUDE
Time
(min)
0
2
3
4
5
6
8
10
15
20
25
30
35
Area (ft2)
25+
2.5*
0
0.16
1 .09
1 .67
1 .92
2.07
2.17
2.31
2.63
2.93
3.27
3.38
3.45
50+
4.55*
0
1 .04
1 .65
2.00
2.05
2.08
2.11
2.13
2.15
2.16
2.22
2.24
2.28
75+
7.00*
0
0.12
1 .85
2.21
2.44
2.74
2.80
2.81
2.82
2.83
2.86
2.87
2.89
100+
4.53*
0
0.07
1 .20
1 .56
1 .76
2.12
2.35
2.56
2.86
2.88
3.08
3.49
3.75
                               ^
    + Volume of oil  spilled,  cm
    * Duration of spill,  min
                           196

-------
TABLE D3.   AREA OF OIL SLICKS FORMED  BY  ALASKA  CRUDE
Time
(min)
0
2
3
4
5
6
8
10
15
20 	
25
30
35

25+
1 .12*
0
4.64
5.22
5.46
5.63
5.73
5.83
6.07
6.69
7.36
7.80
8.20
8.56
Area (ft2)
50+
6.23*
0
4.13
5.20
6.08
7.11
7.30
8.01
8.20
8.77
8.99
9.17
9.33
9.49
75+
2.92*
0
3.74
4.18
4.73
4.92
5.20
5.41
5.90
6.23
6.46
6.69
6.89
7.08
100+
4.15*
0
4.26
4.76
5.39
6.10
6.47
6.64
7.12
7.43
7.62
7.80
7.96
8.10
    Volume of oil  spilled,  cm'
  * Duration of spill,  min
                        197

-------
TABLE D4.  AREA OF OIL SLICKS FORMED BY IRANIAN CRUDE
Time
(min)
0
2
3
4
5
6
8
10
15
20 ..."...
25
30
35
Area (ft2)
25+
3.00*
0
4.47
4.97
5.25
5.41
5.45
5.60
6.23
6.52
6.71
6.97
7.18
7.36
5°+*
3.20*
0
5.13
5.60
5.97
6.23
6.34
6.46
6.56
6.71
6.91
7.20
7.35
7.49
75 +
3.40*
0
5.17
6.16
6.29
6.34
6.39
6.45
6.56
6.93
7.19
7.43
7.53
7.66
100+
3.67*
0
6.22
6.64
6.70
6.99
7.08
7.13
7.33
7.67
7.84
8.05
8.28
8.48
+ Volume of oil  spilled, cnT
* Duration of spill, min
                         198

-------
 TABLE D5.   AREA OF OIL SLICKS FORMED BY SAHARA CRUDE
Time
(min)
0
2
3
4
5
6
8
10
15
20
25 	
30
35
Area (ft2)
25+
1 .55*
0
8.58
9.03
9.45
9.94
10.32
10.56
10.77
10.84
10.86
10.90
10.97
11 .06
50+
4.00*
0
9.33
9.87
9.99
10.11
10.19
10.36
10.58
10.80
10.92
10.97
11 .01
11 .03
75+
3.12*
0
6.49
6.70
7.49
7.61
7.98
8.22
8.37
8.77
8.94
9.12
9.28
9.45
100+
7.52*
0
9.14
9.43
10.21
10.48
10.66
11 .12
11.15
11 .20
11 .29
11 .35
11 .46
11 .54
4.
  Volume of oil  spilled,  cm'
* Duration of spill,  min
                          199

-------
      TABLE D6.   AREA OF OIL  SLICKS  FORMED BY  #2  FUEL
Time
( m i n )
0
2
3
4
5
6
8
10
15
. 20
25
30
35
Area (ft2)
25+
1.83*
0
4.08
5.23
5.92
6.24
6.69
6.88
6.96
7.09
7.21
7.32
7.45
7.57
50+
3.70*
0
2.87
3.86
4.64
5.34
6.02
6.44
6.59
7.12
7.40
7.85
8.32
8.74
75+
6.73*
0
3.69
5.32
6.20
6.57
7.10
7.40
8.03
8.58
8.97
9.28
9.52
9.75
100*
7.22*
0
4.81
7.34
8.05
8.78
9.32
9.54
9.82
10.58
10.62
10.90
11 .14
11 .33
  +                          3
    Volume  of  oil  spilled,  cm

.  * Duration of  spill,  min
                           200

-------
 TABLE D7.   AREA OF OIL SLICKS FORMED BY NIGERIAN CRUDE
Time
(min)
0
2
3
4
5
6
8
10
15
20
25 ' ~"
30
35
Area (ft2)
25+
1 .88*
0
0.47
0.73
1 .10
1 .53
1 .68
1 .83
2.07
2.48
2.77
2.99
3.12
3.23
50+
4.88*
0
1 .19
1 .64
1 .99
2.16
2.37
2.57
2.66
2.77
2.87
3.11
3.17
3.19
75+
4.63*
0
0.70
1.18
1 .68
1 .80
2.02
2.44
2.55
2.80
3.04
3.19
3.30
3.39
100+
9.88*
0
0.45
1.05
1 .31
1 .64
1 .97
2.26
2.64
3.34
4.00
4.57
5.03
5.40
  Volume of oil  spilled,  cm
* Duration of spill,  min
                          201

-------
   TABLE D8.   AREA OF OIL SLICKS FORMED BY #6 FUEL
Time
( m i n )
0
2
3
4
5
6
8
10
15
20
2:5
30
35
Area (ft2)
25+
1 .50*
0
0.04
0.10
0.13
0.15
0.16
0.17
0.18
0.19
0.20
0.22
0.23
0.26
50+
1 .88*
0
0.07
0.14
0.15
0.17
0.18
0.20
0.26
0.32
0.38
0.43
0.49
0.55
75+
2.00*
0
0.08
0.09
0.12
0.14
0.16
0.19
0.20
0.27
0.34
0.43
0.55
0.72
100+
3.23*
0
0.13
0.22
0.24
0.27
0.30
0.32
0.35
0.40
0.47
0.57
0.70
0.88
+                          3
  Volume of oil  spilled,  cm

* Duration of spill,  min
                          202

-------
TABLE D9.   AREA OF OIL SLICKS FORMED BY LAGUNILLAS CRUDE
Time)
( mi n )
0
2
3
4
5
6
8
10
15
20 :
25
30
35
Area (ft2)
25 +
1 .57*
0
0.14
0.19
0.25
0.31
0.35
0.42
0.57
0.94
1 .17
1 .36
1 .51
1 .63
1
50+
2.30*
0
0.13
0.19
0.25
0.29
0.37
0.40
0.53
0.83
1 .05
1 .23
1 .37
1 .50
75+
3.47*
0
0.35
0.65
0.91
1 .13
1 .34
1 .48
1 .79
2.51
2.97
3.33
3.60
3.86
100+
3.30+
0
0.68
1 .24
1 .51
1 .88
2.14
2.50
2.85
3.24
3.74
4.11
4.42
4.72
   Volume of oil  spilled,  cm
 * Duration of spill,  min
                          203

-------
   TABLE DID.   AREA OF OIL  SLICKS  FORMED BY  8%  CRUDE
Time
(min )
0
2
3
4
5
6
8
10
15
20
25 ''"~:~K:
30
35
Area (ft2)
25+
3.0*
0
0.65
1.07
1 .16
1 .67
2.33
2.75
3.49
4.87
5.57
6.31
6.75
7.08
50+
6.57*
0
1 .15
2.08
2.76
3.11
3.81
4.05
4.53
5.76
6.76
7.13
7.46
7.98
75 +
5.52*
0
1 .95
2.95
3.77
4.34
4.84
5.50
6.09
7.14
7.61
7.99
8.30
8.59
100+
8.7*
0
2.58
3.76
4.70
5.42
5.65
5.98
6.86
7.57
8.26
8.90
9.49
10.06
+ Volume of oil  spilled,  cm
* Duration of spill,  min
                          204

-------
TABLE Dll.   AREA OF OIL SLICKS FORMED BY  LA ROSA  CRUDE
Time
( m i n )
0
2
3
4
5
6
8
10
15
20
25
30
35
Area (ft2)
25 +
2.7*
0
1 .47
2.15
2.91
3.59
4.15
4.88
5.69
5.95
6.35
6.53
6.56
6.70
50+
5.33*
0
4.52
5.17
5.68
6.03
6.31
6.51
6.90
7.47
7.50
7.82
8.11
8.43
75+
4.57*
0
6.13
6.77
7.11
7.24
7.49
7.84
7.98
9.84
10.09
10.53
11 .09
11 .25
100+
8.7*
0
6.51
7.17
7.80
8.00
8.34
8.39
9.09
10.21
10.52
10.90
11 .42
11 .48
  Volume of oil  spilled,  cm
* Duration of spill,  min
                          205

-------
TABLE D12.   AREA OF OIL SLICKS FORMED BY  SUNILAND CRUDE
Time
( m i n )
0
2
3
4
5
6
8
10
15
20
.... _;..•
25
30
35
Area (ft2)
25+*
4.07*
0
7.13
7.29
7.39
7.48
7.69
7.95
8.11
8.31
8.51
8.70
8.79
8.97
s°:
5.0
0
3.19
4.54
4.66
5.07
5.51
5.87
6.57
7.44
8.44
8.90
8.93
9.05
75+
5.0*
0
5.12
5.69
6.09
6.37
6.75
7.26
7.90
8.28
8.72
9.06
9.31
9.46
100+
4.4*
0
5.26
5.91
6.43
6.77
7.15
7.68
8.35
9.00
9.42
9.58
9.63
9.70
+                          3
  Volume of oil  spilled,  cm

* Duration of spill,  min
                         206

-------
                                   APPENDIX  E

              DATA  FOR THE  DISPERSION  OF  OILS  WITH  OIL  DISPERSANTS
•St.-''
                  TABLE  El.   DISPERSION  OF  #2  FUEL  WITH
                              SELECTION OIL  DISPERSANTS  (0/D 1:1)
Sampl ing
Time
(Hours)
0.0
0.25
0.50
0.75
1 .0
1 .5
2.0
3.0
4.0
Concentration of Organics (PPM)
A
0.0
74.8
54.1
39.5
31 .9
30.8
24.6
24.1
19.8
B
0.0
377.4
225.1
138.1
108.2
68.1
57.6
57.2
50.3
C
0.0
20.1
19.5
19.1
18.6
17.3
16.5
16.1
15.3
D
0.0
165.1
163.3
160.4
158.1
153.4
145.2
138.8
131 .6
E
0.0
18.8
18.7
14.8
9.7
9.4
6.4
6.2
6.1
                                      207

-------
TABLE E2.   DISPERSION OF #2 FUEL OIL WITH SELECTED
             OIL DISPERSANTS (0/D 5:1)
Sampl ing
Time
(Hours)
0.0
0.25
0.50
0.75
1 .0
1 .5
2.0
3 . Q _
4.0
Concentration of Organics (PPM)
A
0.0
47.4
34.0
22.5
20.4
17.7
15.7
14.7
12.8
B
0.0
82.1
44.5
28.2
21 .7
20.1
14.5
10.7
6.9
C
0.0
16.2
10.3
9.1
8.5
7.7
7.3
7.3
7.0
D
0.0
75.7
69.4
63.1
57.5
46.5
45.9
45.1
43.0
E
0.0
5.9
3.9
2.5
1 .6
1 .6
0.7
0.6
0.5
                       208

-------
TABLE E3.   DISPERSION OF #2 FUEL OIL WITH SELECTED
               DISPERSANTS (0/D 10:1)
Sampl ing
Time
(Hours)
0.0
0.25
0.50
0.75
1 .0
1 .5
..2.0;
3.0 :;::
4.0
Concentration of Organics (PPM)
A
0.0
25.0
14.4
9.9
8.5
6.0
6.0
6.0
6.0
B
0.0
36.7
24.5
16.4
10.2
9.0
6.8
6.2
5.8
C
0.0
6.7
5.6
4.8
4.5
4.1
3.8
3.8
2.9
D
0.0
89.4
60.1
43.3
32.7
32.0
31 .8
26.6
26.3
E
0.0
2.5
2.3
1 .4
1 .3
0.8
0.8
0.8
0.8
                         209

-------
TABLE E4.   DISPERSION OF IRANIAN CRUDE OIL WITH
           SELECTED DISPERSANTS (0/D 5:1)
Sam pi ing
Time
(Hours)
0.0
0.25
0.50
0.75
1.0
1 .5
2.0
,3.0 _
4.0 '----
A
0.0
12.8
9.1
7.6
7.3
7.2
6.4
6.3
5.9
Concentration
B
0.0
71 .9
40.1
26.0
21 .9
20.2
19.6
9.3
8.5
of Organi
C
0.0
6.5
6.2
5.4
5.4
5.2
4.5
4.3
3.4
cs (PPM)
D
0.0
37.3
34.0
33.8
33.1
32.3
32.1
29.6
29.9
E
0.0
0.5
0.3
0.3
0.2
0.2
0.2
0.2 .
0.1
                          210

-------
TABLE E5.  DISPERSION OF #6 FUEL OIL WITH SELECTED
           DISPERSANTS (0/D 5:1)
Sampl ing
Time
(Hours)
0
0
0
0
1
1
2
3
4
.0
.25
.50
.75
.0
.5
.0
.0^:
.0
0
1
1
1
1
1
1
1
1
A
.0
.7
.6
.5
.5
.4
.4
.3
.2
Concentrate
B
0.
35.
32.
21 .
16.
16.
12.
9.
7.
0
0
7
7
9
7
5
6
9
ion of Organics (PPM]
C D
0.
5.
3.
3.
3.
3.
2.
2.
1 .
0
8
6
6
1
1
4
0
7
0.
24.
18.
17.
16.
14.
12.
11 .
10.
0
9
4
8
5
3
2
8
6
I
E
0.
0.
0.
0.
0.
0.
0.
0.
0.

0
4
4
4
3
3
2
2
2
                          211

-------
TABLE E6.   DISPERSION OF #2 FUEL  AND IRANIAN  CRUDE
           OIL WITH PRODUCT B (SALT WATER,  0/D 5:1)
Sampling Time
(Hours)
0.0
0.25
0.50
0.75
1 .0
1 .5
2.0
~3.0^
4.0
Concentration of Organics (PPM)
#2 Fuel Oil
0.0
28.6
23.0
15.7
12.9
11 .8
8.0
7.5
5.6
Iranian Crude Oil
0.0
73.6
52.8
38.4
30.6
30.4
22.9
20.0
13.6
                        212

-------
TABLE E7.   DISPERSION OF #2 FUEL AND IRANIAN CRUDE
           OIL WITH PRODUCT B (0/D 5:1)  WITH AND
           WITHOUT AGITATION
Sampl ing
Time
(Hours)
0.0
0.25
0.50
0.75
1 .0
1 .5
2.0
3.0
4.0 ~:
Concentration of Organics (PPM)
#2 Fuel Oil
Without
Agitation
0.0
2.6
2.3
1 .5
1 .3
1 .3
1 .3
1.0
1 .0
Continuous
Agitation
0.0
158.6
170.4
170.8
175.4
241 .2
248.4
285.0
327.4
Iranian Crude Oil
Without
Agitation
0.0
1 .8
1 .8
1 .6
1 .6
1 .5
1 .5
1 .4
1 .4
Continuous
Agitation
0.0
36.5
41 .9
43.1
69.0
72.2
73.4
84.1 fc
84.7
                       213

-------
                      APPENDIX F


        COMPUTER PROGRAMS FOR NUMERICAL ANALYSIS
                 OF EXPERIMENTAL DATA
                       TABLE Fl

   Computer  Program for Fitting Experimental  Data
               to  Dissolution Equations
00100   DPT IDNS  LS=SOJ
00200   DRTR;
00300   INPUT T  C;
00400   PPDC NLIN  BEST=10 METHOD=r-1RRQUl=iRDT;
00500   PRRRMETERS  KL=.01
00600               KE=.04
00700               TM=3
f i n :-: fi n               f: :•":=£9.95
00900   EDUHDS CS>0~TM>OJ
01000   IF TOTM THEN  DO?
01100   MODEL C=CS+a. 0-EXP C-KL*T» j
01eoo   END;
01300   ELSE DD;
01400   MODEL C=CS* <1. 0-EXP C-KL*TM> > -*-EXP <-KE* -;T-TM» J
01500   END;
01600   OUTPUT DUT=NEl..l PREDICTED=CHRT;
01700   PRDC PRINT;
•pisoo   vflR T c  CHRT;
01-913CO   PRDC PLOT;
02 0013   PLOT C*T CHRT*T=''*'VDVERLflY;
                         214

-------
                            TABLE F2

         Computer Program for Fitting Experimental  Data
                     to Spreading Equations
 0 0 1 d 0
 00 £00
 00300
 00400
 00500
 no Ann
 00700
 00800
 nn900
 01000
 01100
 01200
 01300
 01400
 01500.
~0160ti
 01700
 0 1 '3 0 0
 01900
 nsnnn
 02100
 02200
 02300
 02400
 02500
 02600
 02700
 02800-
 02900
 03000
 03100
 03200
 03300
 03400
 03500
DPT I QMS LS=80;
DflTfiJ
IHPUT T fl;
IMFILE Ti;
PRDC ML IN METHOD=MflRQIJflRIiT;
PflRflMETERS K=i;
VT=255
TD= 105.0?
Q=VTXTD;
STW=73.4;
STD=£8. 0?
STDW=29.7;
S I GMfl=STI...I-STD-STOI..J 5
VISD=. 01594?
IF T
=TD THEN V=VTJ MODEL fl=K+ OUTPUT DUT=NEW PREDICTED=ftl 5 PRDC PRINT; VflR T fl Hi? PRDC PLDT; PLOT H+T fll*T='*xxDVERLflY; PRDC NLIN METHDD=MflRQUflRDT; PflRflMETERS K=1J VT=25; Tn=io5. o; Q=VT-'TD; STW=73.4? 28. 05 -T2. 0x7. 0> 03700 03800 03900 04000 SI GMfl=ST W-STD-STDW ? SPD=.797; IF T=TH THEN V=VTJ MODEL fl=K> < *•'•» C2. 0x3. 0> > OUTPUT OUT=NEW PREDICTED=ft2; PRDC PRINT; VflP T fl fl2; PRDC PLOT; PLOT fl+T fl£+T=-" + '-xDVERLflY; 215
-------
TABLE F2  (Contd.)
 04100    PRDC MLIM METHOri=MRRQURRnT?
 04200    PflRflMETERS K=l?
 04300    VT=25?
 04400    TD=105.0?
 04500    Q=VTxTD?
 04600    Gp=9ftn;
 04700    SPD=.797;
 04800    SPW=1.0?
 049 0 0    nEL=SPDxSpl.,.l;
 05 0 0 0    VISD=.01594 ?
 05100    IF  T
=TD THEM V=VT? 053 0 0 MODEL R=K> < CGR+SPD*- < 1. 0-DEL> 05400 OUTPUT OUT=NEW PREBICTEIi=fl3? 05500 PRDC PRIMT? 05600 VfiR T fl R3? 05700 PRDC PLDT? 05800 PLDT fl+T R3*T=-'*"'xDVERLRY; 059HO PRDC MLIM METHDH=MflRQUflRDTJ -136000" "PflRflMETERS K=l? - - - =96100^ VT=25? 06200 TD=105.0? 06300 Q=VTxTD? 06400 GR=980? 06500 SPO=.797? 06600 SPW=1.0? 06700 DEL=SPOxSPW? 06800 IF T=TD THEM V=VT? 07 00 0 MDDEL R=K> < *• V*-* 07100 OUTPUT DUT=NEW PREDICTEIi=fl4; 07200 .PRDC PRIMT? 07300 ;MfiR T R R4? 07400 PROC PLDT? 07500 PLOT R+T R4*T=-'* -xDVERLRY? C2. l. 0» 0x 216
-------
                                   TECHNICAL REPORT DATA
                             iu.vr read /«ii/i/(V/>»'V :in l/ic reverse bcjorc fo
  ,i • , T>. 1 rvO
  _ EPA-600/2-,
    . L A\U-S-UBTI I Lt
                     DISPERSAL MECHANICS
                        6. PERFORMING ORGANIZATION CODE
  AUTHORS)	"

     Chukwuka A. Osaiaor    and
Robert C. Ahlert
  PERFORMING.OHGANIZATION NAME AND ADDRESS
     Dept. of Chemical  & Biochemical Engineering
     Rutgers, The  State University of New Jersey
     New Brunswick,  New Jersey  08903
 12. SPONSORING AGENCY NAML AND ADDRESS
     Municipal Environmental  Research Laboratory-Cin.,OH
     Office of Research  & Development
     U.S. Environmental  Protection Agency
     Cincinnati,  Ohio  45268
                                                           3. RECIPIENT'S ACCLSSION-NO.
                                                           5. REPORT DATE
                                                             September 1981	
                           six
                           1H
                                                           8. PERFORMING ORGANIZATION REPORT NO
                        10. PROGRAM ELEMENT NO.
                                 AUN1K
                        11. CONTRACT/GRANT NO.

                             Grant No.  R-805901
                        13. TYPE OF REPORT AND PERIOD COVERED
                        Final
                        14. SPONSORING AGENCY CODE
                                EPA/600/14
15. SUPPLEMENTARY NOTES
                Project  Officer -  Leo T. McCarthy, Jr.     (201)  321-6630
16. ABSTRACT
     This study  investigates  the spreading and dissolution behavior  of small  oil slicks
 brmed from spills  of  12  oils.   The increases in area covered  by  the oils  during spread-
ing experiments  were determined using photographic techniques.  Spreading  equations were
derived and used to correlate experimental data.  Derivation of the  equations parallels
Fay's development.

     The rate of dissolution  of the oils in tap water at 25°C  were investigated by equi-
librating oils with water in  open static tests.  Limits of solubilities  have  been estab-
lished for the oils from  results of long-term equilibration in closed vessels.   Six oils
were also equilibrated with salt water.  A segmented mathematical model  has  been arrived
ind used to correlate  experimental  data.  The model describes  two processes  that occur
during equilibration:   soluble  and  volatile components of oil  leach  into solution
initially, and later evaporate  from solution.
           y^ detailed  description of the mass transfer process occurring  during chem-
 cal dispersion of oil  spills  has  been made.  The primary mechanisms  have  been quanti-
 ied by analogy to homogeneous and heterogeneous catalysis and detergency.   To evaluate
the effectiveness of five commercial dispersants, a large-scale laboratory system has
Deen designed.  Parameters investigated include oil and dispersant types,  oil-to-
iispersant ratios, degree of agitation, and the effect of salt water. __
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
          Dispersal
          Spreading
          Dissolution
          Miceles
          Chemical Dispersion
           b.IDENTIFIERS/OPEN ENDED TERMS


           Mechanisms
           Dispersal Processes
           Petroleum
                                                                           COSATi Held/Group
 3. DISTRIBUTION STATEMENT


          Release to Public
           19. SECURITY CLASS (Tlus Report)
                Unclassified
21. NO. OF PAGES
      237
           20. SECURITY CLASS (This page)

                Unclassified
                                     22. PRICE
EPA Form 2220-1 (9-73)
                                            217

-------