United States
                      Environmental Protection
                      Agency
 Hazardous Waste Engineering
 Research Laboratory
 Cincinnati OH 45268
                     Research and Development
EPA/600/S2-87/070  Jan. 1988
v>EPA          Project  Summary
                     Evaluation  of  Oil  Spill  Dispersant
                     Testing  Requirements
                       This research program was initiated
                     to evaluate the cost and effectiveness
                     of the procedures for testing oil spill
                     dispersants as specified in the National
                     Oil and Hazardous Substances Pollution
                     Contingency Plan, Annex X. The testing
                     procedure is described in detail in the
                     Standard EPA Dispersant Effectiveness
                     and Toxicity Tests (EPA-R2-73-201) and
                     in Annex X. These procedures were
                     examined using No. 2 and No. 6 fuel
                     oils and  six commercial oil spill dis-
                     persants. The methods were evaluated
                     in terms  of reliability, precision, cost-
                     effectiveness, and applicability.
                       Seven laboratory methods for testing
                     dispersant effectiveness using com-
                     mercial oil spill products and No. 2 and
                     No. 6 fuel oils were evaluated. The
                     tests  included the  EPA,  Mackay,
                     Russian,  French, Warren Spring, and
                     two interfacial tension  test methods
                     (one based on the du Nouy ring principle
                     and the other on drop-weight). These
                     tests were reviewed in terms of type,
                     scale, method  of applying mixing  en-
                     ergy, and the time required to conduct
                     a product evaluation. The experimental
                     results, compared in terms of the preci-
                     sion of the test data and how effective
                     the six nonionic  dispersants were,
                     demonstrate that the relative effective-
                     ness found for the dispersants varies
                     appreciably as a function of the testing
                     method. Reasons for the variation are
                     discussed, and recommendations are
                     presented on how to achieve dispersant
                     testing data that are more representative
                     of open-sea conditions.
                       On the basis of these findings, recom-
                     mendations for revision to the Standard
                     Dispersant Effectiveness  Test from
                     Annex X and the Standard Dispersant
                     Toxicity Test were made and have been
                     included as part of the full report.
  This Project Summary was developed
by EPA's Hazardous Watte Engineering
Research Laboratory, Cincinnati, OH, to
announce key findings of the research
report that Is fully documented In a
separate report of the same title (see
Project Report ordering Information at
back).
 Introduction
  In 1975 the National Oil and Hazardous
 Substances Pollution Contingency Plan
 (NCP) was developed in compliance with
 the Federal Water Pollution Control Act.
 This plan as currently amended (40 CFR
 Part 300) provides for a coordinated and
 integrated  response by  the Federal
 Government to protect the environment
 from the damaging effects  of pollution.
 One means of minimizing damage from
 oil discharges is to emulsify, disperse, or
 solubilize the  material  into the  water
 column, thereby  minimizing surface ef-
 fects.  Use of chemical  treating agents
 was specifically covered by Annex X of
 the NCP and is controlled under  its re-
 placement, Subpart H.
  Many chemicals have  been developed
 commercially to combat floating oil. When
 applied to oil spills, these chemicals (dis-
 persants) accelerate the  dispersal  of the
 oil into small globules and emulsions by
 reducing the mixing energy required for
dispersal. When  the oil  is dispersed  as
fine droplets, it  will not cling to solid
surfaces and the rate of biodegradation
and chemical transformation will  be in-
creased because  of the resulting greater
surface area. The effectiveness (in terms
of ability to move floating  oil  into the
water column) and toxicity  of  these
chemicals  and resulting oil:water:dis-
persant mixtures varies from oil to oil and
product to product.

-------
  Most chemical oil dispersants are pro-
duced and marketed with little information
concerning their relative effectiveness and
toxicity. This makes it difficult to (1) select
products based on their effectiveness for
a particular  oil  spill, (2) estimate the
effect of  the dispersant on the environ-
ment,  and (3) determine the costs of
treatment.
  A reliable laboratory test procedure for
estimating dispersant  effectiveness  is
desirable for the selection of the  most
economically  and  environmentally ac-
ceptable  formulation. However, it is dif-
ficult to  take  into account all the test
parameters  that  adequately  represent
real-world conditions  on a  laboratory
scale. In  evaluating laboratory tests, it  is
necessary to identify parameters that are
important and those that may be disre-
garded without affecting the precision of
the test results.
  The  objectives of this study were to
evaluate  the Standard  EPA  Dispersant
Effectiveness and Toxicity Tests (EPA-R2-
73-201,  May 1973) when applied to six
commercial oil dispersant chemicals (A,
B, C, D, E, and F) in light of other existing
test protocols and to recommend modifi-
cations to the procedures that will en-
hance the sensitivity and reliability of the
method.
  Other methods examined (by literature
review) included the Mackay* test, the
Russian  test,  the  Wareen Spring test
(Warren  Spring Laboratory), and the
French test (French Ministry). Two  addi-
tional tests based on interfacial tension
were also investigated.
  Previous investigators have tried to
compare  the EPA standard test with other
reported  methods. The comparison is dif-
ficult because most laboratory methods
differ widely in design and procedure and
test  criteria are often unavailable. For
example, mixing energy is supplied by a
spray hose and circulation pump in the
EPA test, by a high-speed propeller in the
Russian  test, by  rotating  separatory
funnels in the Warren Spring and French
tests, and by a high-velocity airstream  in
the Mackay test. In the EPA and Russian
tests, dispersant is poured on the test oil
in a containment  ring; in the Warren
Spring and MacKay tests, the dispersant
is syringed on top of the test oil. In the
Russian,  Warren Spring, and French tests,
samples  are withdrawn for analysis after
mixing is stopped; in the EPA and Mackay
 Mention of trade names or commercial products
 does not constitute endorsement or recommenda-
 tion for use.
procedures sampling is conducted under
dynamic conditions.
  The present investigation focused on
(1) comparison of test data in terms of
dispersant ranking order, precision of test
results,  and time required to complete
the evaluation, (2) identification of key
experimental parameters that contribute
to poor test results, (3) recommendations
for improving current testing methods to
simulate the real world environment more
closely,  and (4) development of an im-
proved effectiveness test that is subject
to fewer interferences.
  Specifically,  the work consisted of the
following tasks:
  • Examine the physical and chemical
    parameters and test  prescribed by
    Annex X and evaluate their applic-
    ability.
  • Conduct laboratory effectiveness and
    toxicity tests of the six dispersants
    using the Annex X procedures.
  • Determine the cost and time required
    to establish a facility with the equip-
    ment necessary to perform the EPA
    tests routinely.
  • Evaluate other dispersant effective-
    ness test  protocols  (by  literature
    review) and compare the results with
    the EPA test.
  • Modify the EPA test and address its
    major deficiencies.
  • Suggest alternative  testing proce-
    dures, such as the drop-weight inter-
    facial tension test.
  As a  result  of this work, a  revised
standard dispersant effectiveness test, a
revised  standard toxicity test, a modified
EPA dispersant effectiveness test proce-
dure, and a drop-weight dispersant effec-
tiveness  test  procedure have been
prepared, and are presented in the project
report.

Conclusions
  Overall test characteristics, advantages
and disadvantages  are summarized in
Table 1. Also provided in this section are
conclusions regarding the dispersant ef-
fectiveness tests and the toxicity tests, as
well as the material and equipment
requirements.

Dispersant Effectiveness Tests

The Standard EPA Test
  • Test results varied with the type of
    oil  and the dispersant/oil ratio.
  • Data variance for the initial three
    products tested (C, E, and F) was
    greater than the value suggested in
    EPA-R2-73-201.
  •  Data variance after modifications foi
     tank cleaning  and oil/  dispersant
     addition (products A, B, and D) was
     less than  the  value  suggested  in
     EPA-R2-73-201.
  •  Considerable amounts of oil adhere
     to the  walls of the tank  and the
     plastic pumping system.
  •  Excessive time is required to conduct
     a dispersant evaluation.
  •  The test can be revised  to improve
     its performance (see Recommenda-
     tions).

The Modified Test
  •  Test addresses  major deficiencies of
     standard  EPA  test: application  of
     constant mixing energy and elimina-
     tion of circulation pumps and plastic
     sampling system.
  •  Test results improve with increasing
     spray pressure.
  •  Optimized sampling position is 5.08
     cm (2 inches) below the  water sur-
     face and 7.62 cm (3 inches) from the
     tank wall.
  •  Effectiveness test  results  for  dis-
     persants B, D, and E are considerably
     lower than standard EPA test results.
     For  dispersants A  and C,  approxi-
     mately  the  same  results were ob-
     tained for both tests.
  •  Dispersant effectiveness  test results
     of dispersants A and B increase (by
     different factors)  with  increasing
     paddle speed.
  •  Oxygen  mass  transfer  coefficient
     measurements for the modified test
     suggest that the mixing energy is a
     factor of about 3.5 less than that for
     the standard test.

Du Nouy Ring Interfacial
Tension Test
  •  The test is  unsuited for  heavy oils
     such as No. 6 fuel oil because  of
     cleanliness problems and the curved
     meniscus of the oil.

The Drop-Weight Interfacial
Tension Test
  • The test is superior to other methods
     in terms of simplicity, required work-
     ing space, and  time.
  • The test measures the fundamental
     physical property that  applies to all
     surfactant/oil systems: the lowering
     of the  interfacial tension  between
     oil and water.
  • Test results are difficult  to compare
     with other methods  because  no
     mixing energy is required,  wall ef-
     fects are minimized, sampling prob-

-------
 able 1.   Review of Dispersant Effectiveness Tests
Test
EPA ISTD)
EPA (REV)
Mackay
Russian
French
Warren
Spring
Drop-
weight
Du Nouy
Ring
Type
Tank
Tank
Tank
Tank
Separatory
funnel
Separatory
funnel
Interfacial
tension
Interfacial
tension
Scale
(liter)
130
130
6
1
0.25
0.25
0.02
0.07
Mixing
Energy
Source
Spra y/ circulation
pump
Spray/circulation
pump
Air turbulence
Propeller
Rotating sep. funnel
Rotating sep. funnel
—
—
Time per
Test
(hr)
96
96
21
20
17
10
4
<4
Precision of
Test Results
(% S.D.)
32.3
6.4
10.1
8.8
13.3
14.5
6.0
N.C.
Major
Advantages
Large Scale
Large Scale
Simulates wave
action
Simple, bench scale
Simple, bench scale
Simple, bench scale
Very simple, rapid
Very simple, rapid
Major
Disadvantages
Time and cost; pump shear
effect; mixing energy
Time and cost; pump shear
effect; mixing energy
Expensive apparatus; wave
damping; high mixing energy
Propeller shear effect; bottom
sampling
Bottom sampling; wall effects
Bottom sampling; wall effects;
dispersant and oil premixed
Does not give value for amount
of oil dispersed
Does not give value for amount
of oil dispersed
     lems do not apply, and the results
     are not interpretable in terms of the
     amount of oil dispersed.
  • Three interfacial phenomena can be
     derived from a single drop-weight
     measurement. The critical micelle
     concentration (cmc) can be used to
     predict the dispersant concentration
     at  which monolayer  coverage is
     achieved at the oil water interface;
     the initial slope of the cmc curve can
     be  used to  predict the packing ef-
     ficiency of surfactant molecules at
     the oil water interface; and the drop-
     weight reduction is proportional to
     the interfacial  tension lowering
     between oil  and water.

The Mackey Test
  • Data variance is higher than in the
     EPA test.
  • The product ranking order is different
     than that based on the EPA results.
  • Apparatus  is not  readily portable,
     space  efficient, or time- and  cost-
     effective.
  • Magnitude of applied mixing energy
     appears to be too high.
  • Problem of wave damping could
     generate an unrealistic ranking trend
     in data for a series of dispersants.

The French, Russian, and
Warren Spring Tests
  • The amounts of oil dispersed in the
     French, Russian, and Warren Spring
     tests were generally lower than those
     in the EPA and Mackay tests.
  •  Bottom  sampling  discriminates
     against observation of larger particle
     size oil droplets and may skew test
     results.
  •  Wall effects may contribute to the
     higher data  variance than that ob-
     served for the EPA and Mackay test
     results.

Dispersant Toxlclty Testa
  •  Post-test  mortality  and  sublethal
     effects in fish exposed to some dis-
     persant products indicate that an
     LC60 estimate based on the test may
     not be adequate for predicting the
     environmental impact of a dispersant.
  •  Because  of differences in test
     methodologies, comparisons of toxi-
    city rankings obtained from bioassays
    performed on brine  shrimp  and
     mummichogs are meaningless.
  • The higher LC50 values reported for
    the toxicity of dodecyl sodium sulfate
    to brine shrimp were approximately
    five times greater than  the  lower
    LC50 values. A portion of this variation
    is probably due to differences in the
    age of  the dodecyl sodium  sulfate
    stock solution.
  • The standard seawater formulations
    given  in  EPA-R2-73-201 are  not
    stable at pH 8.  For  example, pH
    levels were as  low a 6.9 in control
    jars immediately after shaking, even
    though the pH of the seawater stock
    had been  8.0  ±  0.1  only 30-60
     minutes  earlier.  It appears  that
     several salts were not included in
     the formulations.
  • Grass shrimp (Crongon franciscorum)
     appear satisfactory as a test organism
     using the same experimental proce-
     dures and diluent  water specified
     for the mummichogs.

 Time, Material, and Equipment
 Requirements
  Equipment setup  time,  performance
 time, and cost can  have  a significant
 effect on  the  general  acceptance  of  a
 testing program. As a part of our effort,
 we maintained an inventory of the equip-
 ment and  materials used and a record of
 the time required to perform each phase
 of the work. The objective of this  task
 was to identify the cost and time required
 to establish a facility with the equipment
 necessary to perform the Annex X tests
 routinely.
  Table 2  summarizes the major equip-
 ment required to conduct  the Annex X
 testing procedures. The  estimated costs
 do  not  reflect the  purchase of major
 equipment (Table 2), which may be  sub-
 ject to wide price variations.

 Recommendations

Dispersant Effectiveness Tests

General
  Nome of the laboratory test procedures
adequately tests the effectiveness of

-------
Table 2.    Major Equipment Required to Conduct Annex X Tests

     Test                                    Equipment
Dispersant effectiveness
Dispersant toxicity
Standard EPA test tank and spraying apparatus
Beckman Model Zeromatic II pH meter
Turner Model 350 Spectrometer

Constant temperature bath (SRI manufactured}
Shaker table
Assorted gas-tight syringes (Hamilton)
Light box (SRI manufactured)
Blender
Miscellaneous organism maintenance equipment and glassware
Physical properties

  Viscosity


  Specific gravity

  Miscibility


  Flash point

  pH

  Pour point

  Ionic activity

  Metal content

Chlorinated hydrocarbons
Brook field Model LVF Viscometer
Master line Model 2095 constant temperature bath

Miscellaneous hydrometer tubes. VWR Scientific Co.

Constant temperature bath
Miscellaneous glassware and chemicals

Setaflash Model OJSF

Beckman Model Zeromatic II

Pour point jars, VWR Scientific Co.

Miscellaneous glassware and chemicals

Varian Model AA6 Atomic Absorption Spectrometer

Microtek Model 220 gas chromatograph
Coulson Electrolytic Conductivity Detector
two-dimensional dispersants. The surface
and interfacial  tension  measurements
discussed  later  are the only evaluation
methods currently available, and these
have not been systematically studied.
  A series of standard test oils should be
established. One or two crude oils should
be  added to the Annex X testing; No. 2
fuel oil could be eliminated because it is
unlikely that dispersants would be  used
on  it in the field.
  The manufacturers of several products
recommend application rates [that is, ratio
of dispersant to oil 
-------
Table 3.    Recommendations for Improving the EPA Dispersant Effectiveness Test"
Problem
Effect on
test numbers Recommendations
Effect of Na2S04 drying         ~5% high
  material

Effect of dispersant on oil       ~5% high
  measurement

Dispersant blank correction not  Variable
  linear with concentration

Water and oil blanks not        High
  required

Oil hfs on pump and tubing     Variable
  parts

Oil loss to tank walls and ring    Low
  during spraying and foaming

Recirculation pumping rate     Variable
  not constant

Impossible to supply repeated    Variable
  mixing energy

Eddy currents around water     Variable
  return tube

Tank and tubing cleaning       Variable
Dispersant application may be    Variable
  different from field condition?
            Eliminate step because H2O does not interfere
              with method

            Make blank correction
            Measure dispersant blank correction at
              different concentration levels

            Measure blanks and make appropriate
              correction

            Use Teflon or PVC (see compatibility
              test results)

            Coat tank with lipophobic material such as
              Teflon

            Use constant voltage power supply or
              constant speed pump

            Use fine spray over ring or use constant speed
              propeller mixer

            Extend tube deeper into tank
            Coat tank with lipophobic material; use
              demountable pump and tubing connections

            Follow manufacturer's recommended
              application method and fate
6 EPA-R2-73-201.

D/0  ratios is incorrect.  Such  blanks
should  be incorporated  in any revised
procedure.

Modified EPA Dispersant
Effectiveness
  The modified EPA Dispersant Effective-
ness Test addresses the three major areas
that needed improvement in the Standard
EPA Test: (1) a repeatable way to apply
the mixing energy; (2) a sampling device
to eliminate  the  circulation pump and
connecting plastic  tubing; and  (3) an
energy  source  for simulating wave and
tidal action after the application of mixing
energy. The modified EPA Dispersant Ef-
fectiveness Test apparatus  consists es-
sentially of the Standard EPA Test tank
and containment  ring with the following
modifications:
  • Mixing energy is supplied by a full
    cone synthetic seawater spray de-
     livered by a  standard spray  nozzle.
    The spray covers the entire  tank
    surface, including the oil/dispersant
    containment ring.
  • Spraying is actuated by a solenoid
    valve.
               • Spray nozzle is adjustable in height
                 to maintain full surface coverage at
                 different  spraying  energies (pres-
                 sures).
               • The energy  source for simulating
                 wave and tidal action is a low-shear
                 paddle stirrer operated at a calibrated
                 speed.
             Because oxygen mass transfer coefficient
             (k K) measurements for the modified test
             showed that the applied mixing energy is
             about 3.5 times  less than that for  the
             standard test, additional dispersant effec-
             tiveness tests using the modified proce-
             dure should  be conducted  at a  higher
             mixing energy to increase k £ to at least
             20 cm h °, which is characteristic of a
             medium sea state condition.
             The du Nouy Ring Interfacial
             Tension Test
               This should be dropped from further
             consideration because the ring method
             gives unreliable results for  heavy oils
             such as No. 6 fuel oil.
Drop-Weight Interfacial
Tension Test
  The Drop-Weight Interfacial  Tension
Test measures the fundamental  physical
property that  applies to all  surfactants:
the  lowering  of the  interfacial  tension
between oil and water. This test is also
superior to other  methods  in terms of
simplicity,  required working space, and
test time.
  Several features of the Drop-Weight
Test that  need  additional development
include:
  • Develop the test so it can  be used
     with crude oils; this will involve using
     a finer bore capillary to  increase the
     oil-drop detachment time.
  • Evaluate  a broader range of dis-
     persant products to obtain statistics
     concerning the  precision  of test
     results.
  • Develop the drop-time test to elimi-
     nate the need to weigh the oil drop
     and simplify the  apparatus  for field
     application.
  • Modify the test to obtain information
     concerning relative diffusion rates
     of dispersant through an oil  column.
  • Calculate the drop-weight reduction,
     which is  proportional to the  inter-
     facial tension lowering, from the test
     results, and use this value to predict
     dispersant efficiency.

The Mackay Test
  The major  advantage of  the Mackay
Test is that it attempts to simulate the
was/e action  of the  open-sea  environ-
ment. The problem of wave damping, that
is,  the reduction or elimination  of circu-
lating waves for  some oil/dispersant
combination during testing,  needs addi-
tional study to relate the magnitude  of
the effect to the environment

The French,  Russian, and
Warren Spring Tests
  Because the  French,  Russian,  and
Warren Spring tests suffer from unrealis-
tic mixing energy applications, large wall
effects, and bottom sampling, which dis-
criminates  against larger particle size oil
droplets, these tests should not  be con-
sidered as reliable indicators of dispersant
performance.

Alternative Dispersant
Effectiveness Methods
  It  may be desirable to develop new
testing methods to supplement or replace
procedures that address the question of
mixing  energy  variations  and those

-------
specific areas not covered by the drop-
weight method.
  A simple test based on participate size
counting or light scattering (described in
the project report) could be developed at
reasonable cost to  provide information
not derived from other current testing
methods. An  effective surfactant will
generate a finer oil droplet distribution
than a less effective product given the
same mixing  energy and temperature
conditions. It would appear that a testing
method based  on  this approach  could
provide information that is fundamentally
related to dispersant effectiveness.

Dispersant Toxicity Tests
  There are fundamental differences in
the procedures  given  in Annex  X for
shrimp and fish toxicity tests. It  is not
meaningful to compare the sensitivities
of these organisms because any observed
differences could be due to differences in
the test procedures and  not to physiology.
Implementation of the following recom-
mendations should eliminate  the  major
differences between the test protocols as
well as  reduce variability  related  to
acclimation stress.
  • The same  seawater formulation
    should be used for  brine shrimp and
    fish  tests, unless  tests show that
    physiological differences  exist that
    mandate otherwise.
  • The  technique used to mix toxicant
    solutions should be the  same for
    both bioassays. The blender used for
    the brine shrimp tests provides vio-
    lent mixing that  is probably less
    consistent with  the mixing energy
    available in natural systems than is
    the energy from the shaker method
    used for the fish tests.
  • Since  many manufacturers suggest
    application rates that are less than
    the 1 part dispersant to 10 parts oil
    recommended in the EPA procedures,
    the toxicity associated with the re-
    commended application rate should
    also be evaluated.


Fish and Brine Shrimp Bioassays
  The test protocol does not require oil-
only bioassays. Such bioassays should
be incorporated into the protocol.
  The response of test organisms to the
reference toxicant, dodecyl sodium sulfate
(DSS), appears to vary with the age of the
toxicant.  Another compound should be
used  as  a  reference  toxicant. Phenol
appears to  have some promise, but tests
with grass shrimp indicate that toxicity
can vary, depending on the manufacturer.
  As stated earlier, the standard seawater
formulations do not appear to be stable at
pH 8. The pH of the fish formulation was
adjusted with NaOH, but the brine shrimp
solution could not be adjusted with either
NaOH or NaHC03, because a precipitate
formed almost  immediately. The brine
shrimp formulation exhibits a pH of 8 for
a short time after mixing, but becomes
less basic very quickly. If 8 is the desired
pH, a suitable buffer should be incorpo-
rated into the formulation. The seawater
formulation in  EPA-R2-73-201  differs
from those given by Tarzwell  in  that
NaHC03 was omitted from both formula-
tions and Na2S04 was omitted from one.
Addition of NaHC03 would undoubtedly
improve the pH stability.
  At least  one of the products tested
appeared to leave the  dispersed phase
fairly quickly after mixing. To reduce vari-
ability in the response due to exposure to
different  degrees of  dispersion, all test
organisms  should be introduced to the
test containers at the same time relative
to the  end of  mixing. That  is, all  con-
centrations in one replicate or all repli-
cates at  one concentration should be
mixed simultaneously and the organisms
added  at approximately the  same  time
(within 5 minutes) after mixing.
  It is unclear from EPA-R2-73-201 what
constitutes a desirable test program. For
instance, should all tests be performed
simultaneously or can they be performed
at different times? The optimum experi-
mental design would be to initiate all
bioassays (oil, dispersant, dispersant/oil,
reference toxicant) simultaneously for any
given organism  and dispersant.  This
would ensure a uniform test population
and simplify comparisons of dispersant
and dispersant/oil toxicity.
  Some dispersant  and/or oil volumes
and viscosities are too great to be handled
conveniently with syringes. The use of
pipettes or graduated cylinders should be
permitted when required for measuring
larger volumes.
  The  method promulgated  in EPA-R2-
73-201 for glassware cleaning appears
to be somewhat  more  rigorous than is
required, at least with No. 2 oil. Immersion
of the  test containers in hexane for 10
minutes requires a considerable amount
of hexane if several containers are done
at once, or a considerable amount of time
if all jars are done separately. As hexane
is quite flammable and expensive, use of
two  hexane rinses followed by a  hot
water  rinse, two detergent  scrubbings,
and tap and distilled water rinses would
be safer, more economical, and less time-
consuming. In  addition, it is difficult to
see why the results of multiple scrubbings
and rinses cannot be accomplished with
the  automatic  dishwashing facilities
present in many laboratories.


Fish Bioassay
  At present, predominantly wild-caught
mummichogs are used in the dispersant
bioassay. Inherent in the use of such fish
are genetic differences, seasonal varia-
tions, disease, differences in nutritional
state, and variation in availability at dif-
ferent times of the year. The feasibility of
rearing the mummichog in the laboratory
in a manner similar to that now being
used with the fathead minnow should be
investigated.  If a successful method was
developed, it would ensure the availability
of healthy  fish,  in any age class, at all
times of the year. A somewhat different
approach would be to use  a different,
easily reared fish species for all ranking
tests. An example of such a species would
be the sheepshead minnow (Cyprinodon
variegatus), which is already quite popular
for use in embryo-eawal and chronic
studies.
  Some fish died after being transferred
to clean water  following the 96-hour
exposure; these delayed mortalites con-
stitute a loss of information potentially
important to decision-making. A recovery
period  of one week in clean seawater
following exposure to the oil/water dis-
persion should  be added  to the fish
bioassay.
  The reference bioassay should be per-
formed in the same manner as the dis-
persant bioassay.


Brine Shrimp Bioassay
  In preliminary tests conducted with
brine shrimp, the range of effect for some
toxicants encompassed two orders of
magnitude. No guidance is given in EPA-
R2-73-201  for the concentrations to be
used in definitive tests under such cir-
cumstances.  The appropriate concentra-
tion spread to be used for two orders of
magnitude  should  be  determined and
incorporated  into the protocol.
  Filtering  the  diluent seawater  is not
required for  the fish  bioassays  but is
required for the brine shrimp assay. Why
the seawater must  be filtered for the
brine shrimp is uncertain. This step should
be eliminated or a larger pore size filter
(0.7 mm) used when filtering the brine
shrimp  water.  This will considerably
reduce  the time required  for filtration
with probably no effect on toxicity.

-------
Physical Properties
  Pass-fail criteria for the physical pro-
perties  listed  in  Annex X  should be
developed.
  The Weatherburn  Ionic Activity test
and the measurement of cyanide should
be eliminated from Annex X.  The manu-
facturer should be required to specify
whether the product is cationic, anionic,
or nonionic.
  More specific test procedures for heavy
metals  and chlorinated hydrocarbons
should  be specified  in  the Annex X
methods.  For example, the sensitivity of
the analyses for the heavy metals  could
be improved if different techniques were
used. The  simplest would be the graphite
furnace method (the estimated sensitivi-
ties are given in Section 4 of the project
report).  However, this method requires
specialized equipment that may not be
available in all laboratories. Alternatively,
it should be possible to develop a sample
digestion  method to concentrate  the
elements  for analysis by the aspiration
method.  Separate methods would be
required for dispersants  based on water
and hydrocarbon solvents and for mercury
and arsenic.
  Similarly, it  should be possible  to
develop a simple procedure for estimating
the concentration of chlorinated hydro-
carbons in the dispersant products. The
dispersant (water- or solvent-based) could
be partitioned between water and hydro-
carbon  solvent and  the organic  layer
analyzed by a gas chromatograph equip-
ped with a halogen-specific detector.
This Project Summary was prepared by staff of Woodward-Clyde Consultants,
  Walnut Creek, CA 94596 and SRI International, Menlo Park, CA 94025.
Leo T. McCarthy (deceased) was the EPA Project Officer (see below for present
  contact).
The  complete report,  entitled "Evaluation  of  Oil Spill Dispersant Testing
  Requirements." (Order No. PB  87-232  633/AS; Cost: $18.95. subject to
  change) will be available only from:
        National Technical Information Service
        5285 Port Royal Road
        Springfield, VA 22161
        Telephone: 703-487-4650
John S. Farlow can be contacted at:
        Releases Control Branch
        Hazardous Waste Engineering Research Laboratory—Cincinnati
        U.S. Environmental Protection Agency
        Edison, NJ 08837

-------
                                                                                                               * I •«-.
                                                                                                           METER
                                                                                                         62501091
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
     BULK RATE
POSTAGE & FEES PAID
        EPA
   PERMIT No, G-35
Official Business
Penalty for Private Use $300

EPA/600/S2-87/070
           QOQQ329
                          PS

-------