EPA/600/R-04/119
September 2004
Characteristics of Spilled Oils, Fuels, and Petroleum
Products:
2a. Dispersant Effectiveness Data for a Suite of
Environmental Conditions - The Effects of
Temperature, Volatilization, and Energy
by
George Serial, Subhashini Chandrasekar
Department of Civil and Environmental Engineering
University of Cincinnati
Cincinnati, Ohio
James W. Weaver
Ecosystems Research Division
National Exposure Research Laboratory
Athens, Georgia 30605
National Exposure Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
-------
Notice
The U.S. Environmental Protection Agency through its Office of Research and Development
funded and managed the research described here under contract (69-C-00-159) to the University
of Cincinnati. It has been subjected to the Agency's peer and administrative review and has been
approved for publication as an EPA document. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
11
-------
Foreword
The National Exposure Research Laboratory's Ecosystems Research Division (ERD) in Athens,
Georgia, conducts research on organic and inorganic chemicals, greenhouse gas biogeochemical
cycles, and land use perturbations that create direct and indirect, chemical and non-chemical
stresses, exposures, and potential risks to humans and ecosystems. ERD develops, tests, applies
and provides technical support for exposure and ecosystem response models used for assessing
and managing risks to humans and ecosystems, within a watershed / regional context.
The Regulatory Support Branch (RSB) conducts problem-driven and applied research, develops
technology tools, and provides technical support to customer Program and Regional Offices,
States, Municipalities, and Tribes. Models are distributed and supported via the EPA Center for
Exposure Assessment Modeling (CEAM) and through access to Internet tools
(www.epa.gov/athens/onsite).
At the request of the US EPA Oil Program Center, ERD is developing an oil spill model that
focuses on fate and transport of oil components under various response scenarios. Since crude
oils and petroleum products are composed of many chemicals that have varying physical
properties, data are required to characterize these fluids for use in models. The data and
regressions presented in this report illustrate the interaction between oils and dispersants under a
variety of environmental conditions. EPA expects these data to be useful both for modeling and
to provide a resource for the oil spill response community as a whole.
Rosemarie C. Russo, Ph.D.
Director
Ecosystems Research Division
Athens, Georgia
in
-------
Oil Spill Report Series
A series of research reports is planned to present data and models for oil spill planning and
response. To date, these include:
1. Oil Composition
Zhendi Wang, B.P. Hollebone, M. Fingas, B. Fieldhouse, L. Sigouin, M. Landriault, P. Smith, J.
Noonan, and G. Thouin, 2003, Characteristics of Spilled Oils, Fuels, and Petroleum
Products: 1. Composition and Properties of Selected Oils, United States Environmental
Protection Agency, National Exposure Research Laboratory, EPA/600/R-03/072.
2. Dispersants
George Serial, Subhashini Chandrasekar, James W. Weaver, 2004, Characteristics of Spilled
Oils, Fuels, and Petroleum Products: 2a. Dispersant Effectiveness Data for a Suite of
Environmental Conditions - The Effects of Temperature, Volatilization, and Energy,
United States Environmental Protection Agency, National Exposure Research Laboratory,
EPA/600/R-04/119.
3. Simulation Models
James W. Weaver, 2004, Characteristics of Spilled Oils, Fuels, and Petroleum Products: 3a.
Simulation of Oil Spills and Dispersants Under Conditions of Uncertainty, United States
Environmental Protection Agency, National Exposure Research Laboratory, EPA/600/R-
04/120.
As more reports are added to the series, they may be found on EPA's web site at:
http://www.epa.gov/athens/publications.
IV
-------
Table of Contents
Notice ii
Foreword iii
Oil Spill Report Series iv
List of Figures 3
List of Tables 4
Abbreviations and Acronyms 5
Abstract 6
Introduction 7
Background 8
Factors affecting dispersion 10
Objectives 11
Experimental Methods 12
Factorial Experimental Design 14
Methods 14
Weathering of Oil 14
Oil Standards 15
Dispersant Effectiveness Procedure 16
Sample analysis 16
QA/QC Checks 17
Precision 17
Accuracy 17
Method Detection Limit 17
Calculation Procedure 18
Results 18
Experimental Data Summary 18
Summary of Prudhoe Bay Crude Dispersal Characteristics 18
Summary of No. 2 Fuel Oil Dispersal Characteristics 27
Summary of South Louisiana Crude Dispersal Characteristics 33
Regression Equations 39
Conclusions 44
-------
References 46
Appendix 1: Experimental Data 50
Appendix 2: Replicate Study 60
Appendix 3: SOPs Appearing in the Quality Assurance Project Plan 71
-------
List of Figures
Figure 1 Baffled flask test apparatus 13
Figure 2 Comparison of regression equations (curves) against measured Prudhoe Bay Crude/no
dispersant efficiency 24
Figure 3 Comparison of regression equations (curves) against measured Prudhoe Bay
Crude/dispersant "A" efficiency 25
Figure 4 Comparison of regression equations (curves) against measured Prudhoe Bay
Crude/dispersant "B" efficiency 26
Figure 5 Comparison of regression equations (curves) against measured No. 2 Fuel Oil/No
dispersant efficiency 29
Figure 6 Comparison of regression equations (curves) against measured No. 2 Fuel
Oil/Dispersant "A" efficiency 31
Figure 7 Comparison of regression equations (curves) against measured No. 2 Fuel
Oil/Dispersant "B" efficiency 32
Figure 8 Comparison of regression equations (curves) against measured South Louisiana
Crude/No Dispersant efficiency 35
Figure 9 Comparison of regression equations (curves) against measured South Louisiana
Crude/Dispersant "A" efficiency 37
Figure 10 Comparison of regression equations (curves) against measured South Louisiana
Crude/Dispersant "B" efficiency 38
Figure 11 Estimated vs Measured % Dispersal of PBC with No Dispersant 42
Figure 12 Estimated vs Measured % Dispersal of 2FO with No Dispersant 42
Figure 13 Estimated vs Measured % Dispersal of SLC with No Dispersant 42
Figure 14 Estimated vs Measured % Dispersal of PBC with Dispersant "A" 42
Figure 15 Estimated vs Measured % Dispersal of 2FO with Dispersant "A" 42
Figure 16 Estimated vs Measured % Dispersal of SLC with Dispersant "A" 42
Figure 17 Estimated vs Measured % Dispersal of PBC with Dispersant "B" 42
Figure 18 Estimated vs Measured % Dispersal of 2FO with Dispersant "B" 42
Figure 19 Estimated vs Measured % Dispersal of SLC with Dispersant "B" 42
Figure 20 Average vs "First" single measured dispersal efficiency for PBC 66
Figure 21 Average vs "First" single measured dispersal efficiency for 2FO 66
Figure 22 Average vs "First single measured dispersal efficiency for SLC 66
-------
List of Tables
Table 1 % Dispersal Efficiency of Prudhoe Bay Crude with Dispersant "A" 19
Table 2 % Dispersal Efficiency of Prudhoe Bay Crude with Dispersant "B 20
Table 3 % Dispersal Efficiency of Prudhoe Bay Crude with No Dispersant 22
Table 4 % Dispersal Efficiency of No. 2 Fuel Oil with Dispersant "A" 27
Table 5 % Dispersal Efficiency of No. 2 Fuel Oil with Dispersant "B" 28
Table 6 % Dispersal Efficiency of No. 2 Fuel Oil with No Dispersant 29
Table 7 % Dispersal Efficiency of South Louisiana Crude with Dispersant "A" 33
Table 8 % Dispersal Efficiency of South Louisiana Crude with Dispersant "B" 34
Table 9 % Dispersal Efficiency of South Louisiana Crude with No Dispersant 35
Table 10 Coefficients of Regression Equations with Terms Determined by Step-Wise Linear
Regression 41
Table 11 Oil control experiments (Temperature = 5±1 °C) 50
Table 12 Oil + Dispersant 'A' experiments (Temperature = 5±1 °C) 50
Table 13 Oil + Dispersant 'B' experiments (Temperature = 5±1 °C) 50
Table 14 Oil control experiments (Temperature = 22 +1-1 °C.) 52
Table 15 Oil + Dispersant 'A' experiments ( Temperature = 22±1 °C ) 53
Table 16 Oil + Dispersant 'B' experiments ( Temperature = 22±1 °C ) 54
Table 17 Oil control experiments ( Temperature = 35 ±1 °C ) 55
Table 18 Oil + Dispersant 'A' experiments ( Temperature = 35 ±1 °C ) 56
Table 19 Oil + Dispersant 'B' experiments ( Temperature = 35 ±1 °C ) 57
Table 20 Results of oil-dispersant combinations at different temperatures (Dispersant 'A', 250
rpm, 34 ppt) 58
Table 21 Two way interactions 61
Table 22 Dispersant by temperature interaction ( Flask speed =250 rpm, Weathering =0%). . 61
Table 23 Dispersant by volatilization interaction (Flask speed =250 rpm, Temperature = 22±1
°C) 63
Table 24 Dispersant by speed interaction (Weathering = 0%, Temperature = 22 ± 1 °C ). ... 63
Table 25 Replicate determination: Oil + Dispersant 'A' experiments ( Temperature = 22±1°C ,
Speed = 200 rpm) 64
Table 26 Replicate determination: Oil + Dispersant 'B' experiments ( Temperature = 22±1 °C ,
Speed = 200 rpm ) 65
Table 27 Summary of Prudhoe Bay Crude replicates 67
Table 28 Summary of No. 2 Fuel Oil replicates 69
Table 29 Summary of South Louisiana Crude replicates 69
-------
Abbreviations and Acronyms
2FO - Number Two Fuel Oil
BFT - Baffled Flask Test
DCM - Dichloromethane
DOR - Dispersant-to-oil ratio
ERO3S - EPA Research Object-Oriented Oil Spill model
EXDET - Exxon Dispersant Effectiveness Test
NRC - National Research Council
PBC - Prudhoe Bay Crude Oil
QAPP - Quality Assurance Project Plan
RPD - Relative Percent Difference
RSD - Relative Standard Deviation
SFT - Swirling Flask Test
SLC - South Louisiana Crude Oil
SOP - Standard Operating Procedure
U.S.EPA - United States Environmental Protection Agency
-------
Abstract
Chemical dispersants are used in oil spill response operations to enhance the dispersion of
oil slicks at sea as small oil droplets in the water column. To assess the impacts of dispersant
usage on oil spills, US EPA is developing a simulation model called the EPA Research Object-
Oriented Oil Spill (ERO3S) model (http://www.epa.gov/athens/research/projects/eros/ and
(Weaver, 2004). Due to the complexity of chemical and physical interactions between spilled
oils, dispersants and the sea, an empirical approach to characterizing the interaction between the
dispersant and oil slick may provide a useful or practical approach for including dispersant action
in a model. The main objective of this research is to create a set of empirical data on three oils
and two dispersants that has the potential for use as an input to the ERO3S model. These data are
intended to give an indication of the amount of dispersal of these oils under certain conditions.
The US EPA is developing an improved dispersant testing protocol, called the baffled
flask test (BFT), which is a refinement of the swirling flask test (Venosa et a/., 2002). Use of
this protocol was the basis of the experiments conducted in this study. The variations in the
effectiveness of dispersants caused by changes in oil composition, dispersant type, and the
environmentally related variables of temperature, oil weathering, and rotational speed of the BFT
were studied. The three oils tested were South Louisiana Crude Oil (SLC), an Alaska North
Slope Crude (Prudhoe Bay Crude Oil, PBC), and Number 2 fuel oil (2FO). The two dispersants
with the highest effectiveness scores under certain test conditions reported earlier were selected
for this study. A factorial experimental design was conducted for each of the three oils for four
factors: volatilization, dispersant type, temperature and flask speed. Each of the four factors was
studied at three levels except for the dispersant factor where only two dispersants were
considered. Statistical analysis of the experimental data was performed separately for the three
oils. Empirical relationships between the amount of oil dispersed and the variables studied were
developed. The experiments showed that dispersal increased with mixing energy/flask speed for
each experiment performed, although there were cases with overlapping ranges of dispersal for
different flask speeds. In these cases, increases in dispersal due to lack of weathering or
-------
increased temperature evidently accounted for the overlap. In about half of the experiments there
was no significant relationship between weathering and dispersal. Where weathering was
significant, it was inversely related to dispersal. In either case, the weathering affect was small
compared to either flask speed or temperature. Dispersal did not show a consistent pattern with
temperature increase. For most of the experiments, either the maximum or the minimum
amount of dispersal occurred at the middle temperature of 22 °C.
Introduction
Transportation and consumption of petroleum products around the world has created the
potential for oil spills into the environment. Offshore drilling and production platforms are another
potential source of oil spills at sea. Other important sources of oil spills are facilities that store oil
and ships that clean or empty ballast water at sea. Oil spills at sea can affect the water surface, the
water column, sediments and shorelines. Oil initially forms a slick at the water surface due to its
immiscibility. Oil spills spread under the action of gravitational, viscous and surface tension forces
(Hoult, 1972). Wave and wind action disrupts the oil/water interface, resulting in the formation of
oil droplets that enter the water column. Usually only a small amount of spilled oil can be dispersed
in this fashion-without the addition of a chemical dispersing agent or dispersant (Delvigne, 1987a).
Depending upon the characteristics of the oil, entrainment of water can result in an emulsion of
varying stability (Wang etal., 2003).
Spilled oil will most likely spread over a large area if a quick response is not initiated. So
careful response planning and preparedness are essential for successful response to oil spills at sea.
In general, there are three major response operations for cleanup of oil spills at sea: mechanical
response, in-situ burning, and the use of chemical dispersants (NRC, 1989). Some of the recovery
options for oil on the seabed include manual removal by divers, dip nets, or seines; pump and
vacuum systems; nets and trawls; or dredging.
-------
Chemical dispersants are mixtures of surfactants and other substances that are usually
sprayed onto oil slicks to remove oil from the surface and disperse it into the water column at very
low concentrations (Lessard and Demarco, 2000). Surfactants are surface active agents that are
dissolved in one or more solvents. They have a chemical affinity for both oil (lipophilic) and water
(hydrophilic) molecules. When applied to an oil film, surfactants diffuse to the oil/water interface
and align themselves so that the lipophilic end of the molecule is attached to the oil phase and the
hydrophilic end extends into the water phase, thereby allowing some oil to mix into the top of the
water column in the form of tiny droplets. It is also believed that chemical dispersants help reduce
the droplet size when viscous shear is the dominant breakup mechanism (Ming and Chris, 1998).
Dispersant effectiveness is defined as the amount (quantity) of oil that the dispersant puts
into the water column compared to the amount of oil that remains on the surface. Many factors
influence dispersant effectiveness, including oil composition, sea energy, state of oil weathering, the
type of dispersant used and the amount applied, temperature, and salinity of the water. Certain
components of oil, such as resins, asphaltenes, and larger aromatics or waxes, are barely dispersible
(Fingas, 2000). Oils that are made primarily of these components will disperse poorly even when
dispersants are applied. On the other hand, oils that contain mostly saturates, will disperse both
naturally and when dispersants are added. The additional amount of oil dispersed when dispersants
are used compared to the amount that would disperse naturally depends strongly on the amount of
sea energy present (Mackay et a/., 1984). The amount of dispersant applied is very important to
effectiveness. It was found that the effectiveness falls to nearly zero for a light oil at a dispersant-to-
oil ratio (DOR) between 1:40 and l:60(Delvigne, 1987b; Fingas etal., 1993b; Fingas et a/., 1997).
Background
Many different types of dispersant test procedures and apparatus have been described in the
literature. In general, three approaches have been used for dispersant applications in these tests: (1)
premixing of a dispersant with the oil before the test begins (Fingas et al., 1989a; U. S Environmental
-------
Protection Agency, 1996); (2) premixing of the dispersant with water before oil is introduced to the
system.(Rewick et al., 1981 ; Rewick et al., 1984) and (3) mixing of the dispersant with the oil at
the oil-air interface as part of the testing procedure itself.
Clayton et al. (1993) evaluated the performance of the revised standard EPA method (U.S
Environmental Protection Agency, 1984), Environment Canada's Swirling Flask Test (SFT) (Fingas
etal, 1987; Fingas etal, 1989a; Fingas et al, 1989b; Blondinae^a/., 1997b) and the IFP-Dilution
test (Bardot et al., 1984; Desmarquest et al., 1985; Baling et al., 1990). They also evaluated three
versions of the SFT (premixed dispersant and oil, and 1-droplet and 2-droplet dispersant-to-oil-slick
addition). Based on their results, the SFT premixed procedure was recommended to EPA for testing
dispersants due to its relative simplicity and straightforwardness. Becker et al. (1991) studied the
Exxon Dispersant Effectiveness Test (EXDET) and found that the results obtained from this method
and the SFT were similar. The SFT was adopted in the final EPA regulation in September 1994.
The SFT was reexamined after its first year of use (IT Corporation, 1995). Due to discrepancies in
found in practice, a redesign of the test flask was considered by EPA. The design of the new test
which is referred to as the Baffled Flask Test (BFT) and the experiments leading to major changes
in the protocol reflecting a much more reproducible and stable effectiveness test were described by
(Serial et al., 200 la). The BFT has the promise of being able to overcome limitations of previous
test methodologies which included non-reproducibility and non-representativeness of field
conditions, as reported by earlier researchers. In this project, the baffled flask test was used to test
for variation in the effectiveness of dispersants caused by changes in temperature, oil composition,
oil weathering, dispersant type, and rotation speed.
Currently, US EPA is developing a simulation model called the EPA Research Object-
Oriented Oil Spill (ERO3S) model that is intended to assess the overall impacts (environmental,
health, safety) of chemical dispersant usage on oil spills. This model is designed for simulating a
portion of the oil slick behavior (Weaver, 2004). However, due to the complexity of chemical and
physical interactions between spilled oils and the sea, a portion of the behavior of the oil spill must
be based upon empirical data. The impacts of dispersants on oil slicks are best characterized by
-------
empirical data. So the aim of this project is to create a set of empirical data on three oils and two
dispersants that has the potential for use as an input to the ERO3S model. Depending on the nature
of the results, the data might be used to determine the amount of oil that is dispersed in a spill
scenario.
The three oils tested were South Louisiana Crude Oil (SLC), Prudhoe Bay Crude Oil
(PBC) and Number Two Fuel Oil (2FO). The two dispersants werre commercial products
denoted "A" and "B" for the purposes of this study. These two dispersants were those with the
highest effectiveness scores under certain test conditions reported earlier (Sorial etal., 2001b).
Of the three oils under investigation, PBC and SLC are medium weight EPA/API standard
reference oils and 2FO is a light refined oil.
Factors affecting dispersion
A number of factors influence the effectiveness of dispersants, including the properties of
the oil: viscosity, slick thickness, dispersant-to-oil ratio, surfactant loss at the water surface,
surface tension, emulsion formation, temperature, salinity, mixing energy, etc (Sorial etal.,
2001). Of these, the effect of temperature, mixing energy and weathering were studied. Of those
omitted: viscosity and surface tension changes with weathering and so are included implicitly,
and the oils chosen do not form appreciable emulsions (Wang et a/., 2003).
After dispersants have been added to the oil at sea and the small oil droplets formed,
mixing energy is further required to disperse these oil droplets in the water column. In order for
the dispersion process to be successful, oil droplets must stay submerged in the water column and
not return to the surface to re-coalesce into an oil slick. As reported by Clayton et al. (1993), the
application of dispersant reduces the interfacial tension between the oil and water, which results
in the formation of oil droplets. The initial droplet size distribution is controlled by mixing
energy, as well as the specific dispersant and oil, and dispersant-oil and oil-water ratios.
10
-------
Experimental studies performed by a number of scientists indicated that the sizes of the oil
droplets are inversely related to the amount of mixing energy input into test vessels. (Clayton et
a/., 1993) and (Fingas etal., 1993a) conducted experiments that indicated dispersants reduce the
size of the oil droplets, making re-coalescence unlikely.
The chemical composition and physical properties of a crude oil determine the behavior
of the oil and the way its properties will change when the oil is spilled at sea. Weathering
increases the viscosity of the oil due to evaporation of the lighter components. Oil viscosity has
been implicated as a major factor affecting the dispersibility of oil (Canevari etal., 2001). As oil
weathers and the viscosity increases, it has been demonstrated that the effectiveness of the
chemical dispersants decline (Baling, 1988). Low water temperatures increase the viscosities of
both the oil and the dispersant. As oil gets more viscous due to low water temperature or
weathering, the energy requirement for mixing the dispersant and oil also increases (Clayton et
a/., 1993). High water temperatures usually increase the solubility of dispersants in water and the
temperature of the spilled oil itself. So, an increase in temperature is expected to reduce oil
viscosity and hence improve dispersion. However, there have been conflicting results in the
trend of dispersant effectiveness with either increasing or decreasing water temperature (Mackay
and Szeto, 1981; Byford etal, 1983; Lentinen and Vesala, 1984; Fingas, 1991). The general
conclusion drawn from this body of work is that the ideal condition for dispersal is be a light oil
spilled on warm water with moderate wind and wave action (Paul et a/., 1999) that does not
impede response activities.
Objectives
The objective of this project was to use the baffled flask test data to generate empirical
dispersant effectiveness relationships to explore the behavior of oils under varying conditions and
to develop a data set for use in oil spill models. Three oils and two dispersants were tested for a suite
of environmental conditions including weathering, temperature and speed of the flask. Sub-
11
-------
objectives were:
To conduct a factorial experimental design for three oils, namely, SLC, PBC, and 2FO for
determining the significance of the above factors on the effectiveness of two dispersants
referred to as Dispersant 'A' and Dispersant 'B' in this study. Of the three oils chosen, PBC
and SLC are medium weight reference oils and 2FO is a light refined oil.
To determine if the baffled flask test has sufficient sensitivity for generating dispersant
effectiveness data over the specified range of environmental conditions.
To develop and evaluate quantitative, empirical relationships that relate dispersant
effectiveness to weathering, speed (mixing energy) and temperature for each oil/dispersant
pair.
Experimental Methods
The required glassware and analytical equipment that were used are listed below:
• Modified Trypsinizing Flask: 150 mL glass Trypsinizing flasks were modified to include
a stopcock near the bottom (Figure 1) to provide the BFT apparatus. These are
commercially available and can be obtained from Fischer Scientific.
Shaker Table: A shaker table with a variable speed control unit (40-400 rpm) and an
orbital diameter of approximately 0.75 inches (2 cm) was used to create turbulence in the
test flasks liquids.
• Micropipettor: A Brinkmann Eppendorf repeater plus pipettor with 100 |j,L and 5mL
syringe tip attachments capable of dispensing 4 |oL of dispersant and 100 |j,L of oil was
used, producing a delivery accuracy ±0.3% and precision ±0.25%.
Glassware: Glassware consisting of 25, 50, and 100 mL graduated cylinders, 125 mL
separately funnels with Teflon stopcocks, 1,2,5 and 10 mL pipettes, 50 mL crimp-style
amber glass vials and other lab glassware were used in the experiments.
12
-------
Syringes: 50, 100 and 1000 |j,L gas-tight syringes were used.
Spectrophotometer: A UV-visible spectrophotometer, capable of measuring absorbence at
340, 370 and 400 nm was used in the experiments.
Analytical balance: calibrated per standard laboratory procedures was used.
Synthetic Seawater: The synthetic sea water 'Instant Ocean', manufactured by Aquarium
Systems of Mentor OH, was used for the study. The synthetic sea water was prepared by
dissolving 68 g of the salt mixture in 2 liter of Milli-Q water to provide a salinity of 34 ppt.
Test Oils: Three types of oil samples provided by EPA-Prudhoe Bay Crude Oil (Alaska
North Slope Crude Oil - PBC), South Louisiana Crude Oil (SLC) and Number 2 Fuel Oil
(2FO) were used in the study.
Dispersants: Two dispersant samples were provided by EPA.
Methylene Chloride: Dichloromethane (DCM), pesticide quality, was used for extraction of
all sample water and oil-standard water samples.
i
Figure 1 Baffled flask test apparatus.
13
-------
Factorial Experimental Design
The response variable for the experiments conducted is the percent effectiveness of the
dispersant. The factors and levels of each of the factors are as follows: volatilization (0, 10, and
20% for SLC and PBC and 0, 3.8 and 7.6% for 2FO); dispersants 'A' and 'B'; temperature (4, 22 and
35 °C ); and flask speeds (150 , 200 and 250 rpm). With these levels for each of the factors, a
complete factorial experiment consisting of 54 runs was conducted for each oil. The total number
of samples prepared was 162: 3 oils, 3 temperatures, 3 flask speeds, 2 dispersants, 3 volatilizations.
In addition to these experiments, control experiments using the oils with no dispersant were also
conducted. For this, the number of flasks prepared were: 3 oils, 3 temperatures, 3 flask speeds, 3
volatilizations for a total of 81. Thus the experimental program consisted of 243 flasks for the
unreplicated portion of the experiments.
Methods
The complete set of standard operating procedures (SOPs) for the experiments appear in
Appendix 3. The following sections summarize the main points of each.
Weathering of Oil
The three oils (PBC, SLC and 2FO) were used in the study at three levels of volatilization
(weathering), each. The weathering of the oil was performed by bubbling air up through a one Liter
graduated cylinder filled with each oil sample. The volume of the oil remaining in the measuring
cylinder was recorded as a function of time. By using the one Liter volume and the volatilization
levels selected, the amount of oil that sticks to the cylinder is negligible compared to the amount
evaporated. The evaporative loss was then expressed as a volume percent.
n/ -r i <.-i- j Initial volume - Final volume 1AA
% oil volatilized - - x 100
Initial volume
14
-------
PBC and SLC were volatilized to 10% and 20%, whereas 2FO was volatilized to 3.8% and 7.6% .
Oil Standards
Standard solutions of oil for calibrating the UV-visible spectrophotometer were prepared
with each specific reference oil and dispersant combination used for a particular set of
experimental test runs. A 6-point calibration curve was generated for each set of experiments
(See Appendix 3).
For treatments with no dispersant, i.e, oil control experiments, only oil was used to make
the standard solution. First, the Oil Alone Stock Standard was prepared. The density of 2 mL of
each oil mixed with with 18 mL DCM added was measured by using a 1 mL gas-tight syringe,
and the concentration of the oil solution determined. Specific volumes of 11, 20, 50, 100, 125,
and 150 \iL of PBC-DCM stock or 100, 150, 200, 300 \iL of SLC-DCM stock or 150, 200, 400,
600, 800, 1000 |j,L of 2FO-DCM stock were each added to 30 mL of synthetic sea water in a
separately funnel and extracted thrice with 5 mL of DCM. The final DCM volume for each
extract was adjusted to 20 mL with DCM. The extracts were transferred to a 50 mL crimp-style
glass vial with a Teflon/aluminium seal. The contents of the sealed vial were mixed by inverting.
The vials were stored at 4 ± 2 °C until the time of analysis. Prior to any analysis, the
spectrophotometer ultraviolet lamp was turned on and allowed a 30-minute warm-up period.
For treatments with oil plus dispersant, Oil Plus Dispersant Stock Standards were first
prepared. The densities of 2 mL of each oil, 80 |j,L of the dispersant and 18 mL DCM was
measured using a 1 mL gas-tight syringe and the concentrations determined. These stock
solutions were used to prepare standard solutions in a similar way as described previously for the
Oil Alone Stock Standards.
15
-------
Dispersant Effectiveness Procedure
For each test, 120 mL of synthetic sea water, equilibrated to the desired temperature was
added to a modified trypsinizing flask (Baffled Flask). 100 |oL of oil was dispensed directly onto
the surface of the synthetic sea water using an Eppendorf repeator pipettor with a 5 mL syringe
tip attachment. The dispersant was then dispensed onto the center of the oil slick by using a 100
|j,L syringe tip attachment that was set to dispense 4 |j,L, giving a dispersant-to-oil ratio of 1:25.
The flask was placed on an orbital shaker and mixed for 10 minutes at the desired rotation speed,
at the end of which it was removed from the shaker and allowed to remain stationary on the
bench top for another 10 minutes. At the end of the settling time, 2 mL was drained from the
stopcock and discarded. Then a 30 mL sample was collected in a 50 mL measuring cylinder.
This 30 mL sample was then transferred to a 125 mL separately funnel and extracted three times
with fresh 5 mL DCM. The extract was then adjusted to a final volume of 20 mL and transferred
to a 50 mL crimp-style, glass vial with an aluminum/Teflon seal. These vials were stored at 4±2
°C until the time of analysis.
Sample analysis
The experimental sample extracts and the standard solutions were removed from the cold
room and allowed to equilibrate to the laboratory temperature. First, a blank solution (DCM)
was placed into the UV-visible spectrophotometer. Then the six calibration standards were
measured in the order of increasing concentration and the absorbence values recorded for
wavelengths of 340, 370 and 400 nm. After this, the experimental samples were measured.
Samples that exceeded the highest calibration standard point were diluted. A factor of 10
dilution was needed for the SLC.
16
-------
QA/QC Checks
Precision
Instrument precision obj ectives for the dispersant effectiveness tests were based on analyzing
measurements in replicate. The acceptance criterion was based upon agreement of the replicate
sample values within ±5 % of their mean value. The operator precision objectives were determined
by using the relative standard deviation (RSD) for percent dispersant effectiveness from four
replicate flasks. This acceptance criterion was established as an RSD less than 15% (Venosa, 2002)
D0r. Standard Deviation of Replicates 1AA
RSD = J-—£- x 100 (2)
Average Concentration
Accuracy
The accuracy was determined by using a mid-point standard calibration check after every 4
experimental samples analyzed. The acceptance criterion was based upon percent recovery of 90-
110%. The percent recovery was determined by using the following equation:
C
% R = — x 100 (3)
C
a
where %R is the percent recovery, Cm is the measured concentration of the check standard, and Ca
is actual concentration of the check standard.
Method Detection Limit
The concentration reporting limits (RLs) by UV-Spectrophotometer for PBC, SLC and 2FO
were 0.04,0.05 and 0.09 mg/L respectively. These RLs were at the low end of the calibration curves
for the analytes. The analyses for all these oils were measured within the calibration concentration
17
-------
range. For samples whose measured concentrations were above the range, the samples were diluted
and re-analyzed. Conversely, there was no sample whose measured concentration was below
detection limits.
Calculation Procedure
The dispersant performance (i.e, percent of oil dispersed, or Effectiveness) was determined
by:
J-MXO/ Total oil dispersed , „„
Eff% = i- X 100 (A\
D V \ '
roil v oil
where poil is the density of specific test oil, g/L, and Voi!is the Volume (L) of oil added to the test
flask. This calculation is a simple ratio of the oil dispersed into the water column to the amount of
oil added to the flask. The complete procedures for this calculation are given in Appendix 3.
Results
Experimental Data Summary
The percent dispersal effectiveness for each oil and dispersant combination are given in
Tables 11 to 20 of Appendix 1, along with replicate results given in Tables 22 to 26 of Appendix
2. Because the dispersant behavior differs by oil, each of the three oils are discussed separately
by using a set of histogram-like tables and plots of the data.
Summary of Prudhoe Bay Crude Dispersal Characteristics
Table 1 summarizes the 27 Prudhoe Bay Crude with dispersant "A" experimental data.
18
-------
Each experiment is represented by a "triple" that indicates its value of weathering, temperature
and speed, respectively. To compare relative effects, the values are indicated by H, M, or L
indicating that the high, medium or low value of the parameter was used. The data are grouped
in columns that represent a 10% increment in dispersal efficiency. Within each column, the
dispersal increases bottom to top. The highest dispersal, for example, occurred with the
combination M,M,H that appears in the 90 to 100% column in the topmost occupied row.
LLL LHL
MLL MHL LML
HLL HHL HML
0-10 10-20 20-30 30-40 40-50
LHH
LHH
HHM MHH
LLM MLH LHM
MLM HLH HHH
HLM MML MHM
50 - 60 60 - 70 70 - 80
HMM MMH
MMM LMH
LMM HMH
80-90 90-100
Weathering Temperature Speed*
*H = high, M = medium, L = low
Example M H L = medium weathering, high temperature, low flask speed
Table 1 % Dispersal Efficiency of Prudhoe Bay Crude with Dispersant "A"
Several features of the results are evident from this table. First, all of the experiments
produced at least 20% dispersal of the oil. These values are at least 14% higher than the control
experiments (Table 3, described below), indicating that dispersal above 20% does not occur in
these experiments without the dispersant. Second, dispersal of the oil is normally expected to
decrease with weathering and increase with speed and temperature. If this expectation was
19
-------
completely fulfilled, the least dispersal would occur at H L L and the most dispersal at L H H.
The cells containing these values are highlighted in each table, and the combination H L L,
indeed had the lowest dispersal. It is worth noting, however, that medium and low weathering
produced the next two lowest values of dispersal (2nd and 3rd entries in the 20 - 30% dispersal
column). This result indicates that any amount of weathering occurring with low speed and low
temperature results in only 20% to 30% dispersal. The highest dispersal, however, did not occur
as expected with lowest weathering, highest temperature and highest speed (L H H). Although
dispersal was high for this combination, it was not the highest.
The third point is indicated by the dotted line at 50% dispersal of the oil. All of the flask
speeds are low for all levels of dispersal less than 50% (left of the dotted line). These dispersal
levels occur with high, medium and low values of temperature and weathering. Thus, speed
appears to be the most significant factor that causes dispersal of 50% or less. The forth
observation is that all dispersal greater than 80% occurs with medium (not high) temperatures,
but all of the flask speeds are medium or high, again indicating the importance of energy input.
Weathering, again, seems to be relatively unimportant as high, medium and low values occur in
the 80% plus range of dispersal. In fact, all three values of weathering occur in almost all of the
10% dispersal increments (each column except 40% - 50%).
20
-------
MLL
HLL
MML LHM
LLM HMM
MLM MHL MHM LLH
HML HHL HHM LMM
LLL HLM LML LHL MMM
LMH
MMH
HMH
MLH
MHH
HLH
HHH LHH
0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100
Weathering Temperature Speed*
*H = high, M = medium, L = low
Example M H L = medium weathering, high temperature, low flask speed
Table 2 % Dispersal Efficiency of Prudhoe Bay Crude with Dispersant "B"
Table 2 displays data for Prudhoe Bay Crude with Dispersant "B." As for dispersant "A"
the dispersal is at least 20% in all cases. The minium and maximum dispersals occurred at the
expected end points (shaded boxes). Dispersal efficiency of greater than 80% occurs only with
high flask speed. At lower percent effectiveness, the pattern is less clear, than that for PBC with
dispersant "A," but high, medium or low weathering can occur over the entire range of dispersal
efficiencies. Most (two-thirds) of the low temperatures have dispersal efficiencies of less than
50%. Compared with dispersant "A", dispersant "B" gave fewer results that were less than 40%
or above 90%, which implies more assurance of dispersal between 40% and 90% despite the
varying conditions.
With no dispersant (Table 3), the maximum percent dispersal effectiveness was less than
21
-------
7%, with two-thirds of the efficiencies less than 4%. Dispersal above 4% mostly occurred with
high speed. These occurred in combination with each condition of weathering and temperature,
suggesting that speed was the most important factor.
LHL
HLL
MLL
LLL
0-1
LHM
HHM
MHM
LLM
HLM
MLM
HHL
MHL LMM
1-2 2-3
LML
LMH
HLH
MML
HML
3 -4
MMM
HMM
HHH
LLH HMH
MLH LHH MMH
4-5 5-6 6-7 7-8 8-9 9-10
Weathering Temperature Speed*
*H = high, M = medium, L = low
Example M H L = medium weathering, high temperature, low flask speed
Table 3 % Dispersal Efficiency of Prudhoe Bay Crude with No Dispersant
The responses of the Prudhoe Bay Crude in all experiments are plotted in Figures 2 to 4.
These, and subsequent figures for No. 2 Fuel Oil and South Louisiana Crude, oil show the
experimental data and regression lines described later. In some cases, there is only one
regression line per flask speed. This occurred when there was no statistically significant
relationship for weathering within the data set.
22
-------
The percent dispersal efficiency increased with flask speed in all cases. When
weathering was significant (Figures 6 and 7), dispersal increased inversely with weathering. In
each case, there was a tendency for the maximum dispersal to occur at the middle temperature.
This tendency was strongest for dispersant "A", but occurred for most cases of either dispersant
addition.
23
-------
ca
£
CD
100 —,
90 -
80 H
70 -
10
0
60 —
40 —|
30
20 H
Prudhoe Bay No Dispersant
o
o
A
A
150 RPM, 0% Weathering
150RPM, 10% Weathering
150 RPM, 20% Weathering
200 RPM, 0% Weathering
200 RPM, 10% Weathering
200 RPM, 20% Weathering
250 RPM, 0% Weathering
250 RPM, 10% Weathering
250 RPM, 20% Weathering
Estimate for 150 RPM
Estimate for 200 RPM
Estimate for 250 RPM
0
40
10 20 30
Temperature (C)
Figure 2 Comparison of regression equations (curves) against measured Prudhoe Bay Crude/no dispersant efficiency.
24
-------
CO
l_
CD
Q.
100
90 -
80 —\
70 -
60
40
30
20
10
0
0
Prudhoe Bay Crude DispersanFA"
A
A
A
150RPM,0%w
150 RPM, 10% w
150 RPM, 20% w
200RPM,0%w
200 RPM, 10% w
200 RPM, 20% w
250RPM,0%w
250 RPM, 10% w
250 RPM, 20% w
— Estimate 150 RPM 0%w
•-- Estimate 150 RPM 10%w
•-- Estimate 150 RPM 20% w
Estimate 200 RPM 0% w
Estimate 200 RPM 10%w
Estimate 200 RPM 20% w
— Estimate 250 RPM 0% w
•— Estimate 250 RPM 10% w
Estimate 250 RPM 20% w
10
20
30
40
Temperature (C)
Figure 3 Comparison of regression equations (curves) against measured Prudhoe Bay Crude/dispersant "A" efficiency.
25
-------
CO
w
CD
a
100
90 -
80 —\
70 -
60
50
40 -
30 —\
20
10
0
0
Prudhoe Bay Crude DispersanFB"
A
A
A
150RPM,0%w
150 RPM, 10% w
150 RPM, 20% w
200RPM,0%w
200 RPM, 10% w
200 RPM, 20% w
250RPM,0%w
250 RPM, 10% w
250 RPM, 20% w
— Estimate 150 RPM 0%w
•-- Estimate 150 RPM 10%w
•-- Estimate 150 RPM 20% w
Estimate 200 RPM 0% w
Estimate 200 RPM 10%w
Estimate 200 RPM 20% w
— Estimate 250 RPM 0% w
•— Estimate 250 RPM 10% w
Estimate 250 RPM 20% w
10
20
30
40
Temperature (C)
Figure 4 Comparison of regression equations (curves) against measured Prudhoe Bay Crude/dispersant "B" efficiency.
26
-------
Summary of No. 2 Fuel Oil Dispersal Characteristics
Dispersal of the No. 2 Fuel Oil with dispersant "A" ranged from about 10% to nearly
100% (Table 4). Most dispersal above 70% occurred at high speed, but even 80% - 90%
dispersal was achieved at medium speed. These very high dispersals occurred with low, medium
and high values of both weathering and temperature. The lowest dispersals - 10% to 40% -
occurred with low speed, but there was also dispersal of 50% to 70% at low speed (Table 4).
Dispersal of 40% to 50% occurred with medium speed.
Dispersal of 2FO with dispersant "B" ranged from 20% to nearly 100% (Table 5).
Dispersal above 70% occurred with medium and high speeds. All of the highest dispersals (90%
- 100%) occurred at the highest speed with medium or high temperatures. Dispersals of 70% -
80% occurred at low to medium temperatures. With four exceptions, dispersal increased with
increasing speed, while any of the weathering and temperature levels occurred across the range of
dispersals.
With no dispersant the maximum dispersal was nearly 11% (Table 6). Dispersal above
this value required use of the dispersant. Low values of dispersal tended to be associated with
low temperature and speed. Conversely both temperature and speed tended to be higher where
dispersal was high.
27
-------
MLL
HLL
LML
LHM
MHM
LLM
LHL HHM
MHL MLM
LLL HHL HLM MML HML
LLH
MLH LHH
MHH HMM HMH
HHH MMM MMH
HLH LMM LMH
0-10 10-20 20 - 30 30 - 40 40 - 50 50 - 60 60 - 70 70 - 80 80-90 90-100
Weathering Temperature Speed*
*H = high, M = medium, L = low
Example M H L = medium weathering, high temperature, low flask speed
Table 4 % Dispersal Efficiency of No. 2 Fuel Oil with Dispersant "A"
LML
LLM
MLM
HLM
MLL LHL
HLL HHL
LLL MML MHL
0-10 10-20 20 - 30 30 - 40 40 - 50
HML
LHM
MHM
HHM
50 - 60 60 - 70
LMM
MMM
HMM
LLH
MLH
HLH
70-80
LHH
LMH
MHH
HHH
HMH
MMH
80-90 90-100
Weathering Temperature Speed*
*H = high, M = medium, L = low
Example M H L = medium weathering, high temperature, low flask speed
Table 5 % Dispersal Efficiency of No. 2 Fuel Oil with Dispersant "B"
28
-------
MMM
0-1 1-2
MMH
HML
HLL
LLL
MLL
2-3
LLM
MLM
HLM
LML
3-4
MLH
HLH
LMM
HMM
4-5 5-6
LHL
MML
LLH
MHL
6-7
LHM
MHM
HHL
HHM
LMH
HMH
7-8 8-9
LHH
MHH
HHH
9-10 10-11
Weathering Temperature Speed*
*H = high, M = medium, L = low
Example M H L = medium weathering, high temperature, low flask speed
Table 6 % Dispersal Efficiency of No. 2 Fuel Oil with No Dispersant
Figures 5 to 7 show dispersal of No. 2 Fuel Oil over the suite of experimental conditions.
With the exception of an anomolous point (no dispersant, 22 °C, 200 rpm and 10% weathering),
dispersal increased with flask speed. For the dispersant treatments, weathering did not result in
significant patterns, as each flask speed is represented by only one curve. The data show highest
dispersal at the highest temperature for the control at all speeds, dispersant "B" at 250 rpm and
dispersant "B" at 150 rpm and 10% weathering. Otherwise, the data follow the pattern for
Prudhoe Bay Crude, where maximum dispersal occurred at the middle temperature of 22 °C.
29
-------
iuu —
on
yu
80 —
~?r\
i\j
60 —
03
CD -
o ^0
Q ~
£
40 —
on
OU
20 —
1 n
I U
_
n
u
^^
s^&^=r tt — "^-ssiij
s^*9§,<«B^ —
^x j ^ " * ™ B^n
11
No. 2 Fuel 0\\ No Dispersant
<)> 150RPM,0%w
^ 150 RPM, 10% w
^ 150 RPM, 20% w
| | 200RPM,0%w
Q 200 RPM, 10% w
| 200 RPM, 20% w
A 250RPM,0%w
A 250 RPM, 10% w
A 250 RPM, 20% w
Estimate 150 RPM 0%w
— - — • — • — •— Estimate 150 RPM 10% w
Estimate 150 RPM 20% w
Estimate 200 RPM 20% w
Estimate 250 RPM 1 0% w
Estimate 250 RPM 20% w
0
30
40
10 20
Temperature (C)
Figure 5 Comparison of regression equations (curves) against measured No. 2 Fuel Oil/No dispersant efficiency.
30
-------
cc
0)
CD
Q_
Q
v^
100 -^
on
yu
80 —
vn
/ u
60 —
_
50
—
40 —
30
20
10
0
T
No. 2 Fuel Oil Dispersant "A"
O
o
A
A
150 RPM,0% Weathering
150RPM, 10% Weathering
150 RPM, 20% Weathering
200 RPM, 0% Weathering
200 RPM, 10% Weathering
200 RPM, 20% Weathering
250 RPM, 0% Weathering
250 RPM, 10% Weathering
250 RPM, 20% Weathering
Estimate for 150 RPM
Estimate for 200 RPM
Estimate for 250 RPM
0
40
10 20 30
Temperature (C)
Figure 6 Comparison of regression equations (curves) against measured No. 2 Fuel Oil/Dispersant "A" efficiency.
31
-------
100 —,
90 -
80 —|
70
60 H
03
0)
w 50
Q
40 —|
30
20 -
10
0
No. 2 Fuel Oil Dispersant "B"
o
o
A
A
150 RPM, 0% Weathering
150 RPM, 10% Weathering
150 RPM, 20% Weathering
200 RPM, 0% Weathering
200 RPM, 10% Weathering
200 RPM, 20% Weathering
250 RPM, 0% Weathering
250 RPM, 10% Weathering
250 RPM, 20% Weathering
Estimate for 150 RPM
Estimate for 200 RPM
Estimate for 250 RPM
0
40
10 20 30
Temperature (C)
Figure 7 Comparison of regression equations (curves) against measured No. 2 Fuel Oil/Dispersant "B" efficiency.
32
-------
Summary of South Louisiana Crude Dispersal Characteristics
The South Louisiana Crude showed generally high dispersal with dispersant "A" as most
of the values were above 50% (Table 7). Values above 70% occurred at both medium and high
speeds, while those lower than 70% occurred with only medium or low speeds. Various values
of weathering occurred throughout the range of dispersal values, indicating a minor influence of
weathering on dispersal. Experiments with the highest temperatures had dispersals between
60% and 80%.
MMH
LHH LLH
MMM LHM LMH
LMM HLM HHH MLH
LLL LHL LLM HLH LML
HML MLL HHL HHM MHH HMH
MML HLL MHL MLM MHM HMM
0-10 10-20 20 - 30 30 - 40 40 - 50 50 - 60 60 - 70 70 - 80 80-90 90-100
Weathering Temperature Speed*
*H = high, M = medium, L = low
Example M H L = medium weathering, high temperature, low flask speed
Table 7 % Dispersal Efficiency of South Louisiana Crude with Dispersant "A"
33
-------
Dispersal of SLC with dispersant "B" showed a cluster of values above 70% (Table 8).
These high values were associated with medium or high speeds and various combinations of
temperature and weathering. Low dispersal occurred mostly with low speeds. The control
(Table 9) showed most of the relatively high dispersals (7% and above) occurred with medium or
high speeds, and uniformily high temperatures. The lowest dispersal values (<2%) occurred
mostly with low speeds and temperatures, but with each weathering condition.
HLL
LMM LHL
MML MHL LLL
HML HHL MLL
HLH
MHM
LHM
LMH
HHM
MMM HMH
HMM HHH
LLM MMH
MLM LLH
HLM MLH
LHH
MHH
LML
0-10 10-20 20 - 30 30 - 40 40 - 50 50 - 60 60 - 70 70 - 80 80-90 90-100
Weathering Temperature Speed*
*H = high, M = medium, L = low
Example M H L = medium weathering, high temperature, low flask speed
Table 8 % Dispersal Efficiency of South Louisiana Crude with Dispersant "B"
34
-------
MLH LMH
HLH MMH
HLL HML HLM LLH HHL
MLL LMM MLH MMM HMM
LLL MML LLM MHL LHL LML
HMH
LHM HHM
MHM MHH LHH HHH
0-1
1-2
2-3
3-4
4-5
5-6
6-7
7-!
8-9
9-10
10-11
Weathering Temperature Speed*
*H = high, M = medium, L = low
Example M H L = medium weathering, high temperature, low flask speed
Table 9 % Dispersal Efficiency of South Louisiana Crude with No Dispersant
Figures 8 to 10 show the data and regression curves for South Louisiana Crude.
Dispersal increased with speed for all experimental conditions. Significant effects of weathering
were indicated for dispersant "A" and the control experiments. These showed the inverse
relationship between weathering and dispersal for dispersant "A" and some of the control.
Although the scatter in the data at 150 rpm and 200 rpm makes the relationships less clear for
dispersant "A", results for both dispersants "A" and "B" show minimum dispersal at the middle
temperature (22 °C) at 150 and 200 rpm but maximum dispersal at this temperature for 250 rpm.
The shapes of the regression lines indicate a relationship that is concave up at lower speeds and
transitions into one that is concave down at higher speeds.
35
-------
100 -
90
80
70 -
60 —\
03
CD
Q.
W
Q
5° ~
40
30
20
10
0
South Louisiana Crude No Dispersant
A
A
A
150RPM, 0%w
150 RPM, 10% w
150 RPM, 20% w
200 RPM, 0% w
200 RPM, 10% w
200 RPM, 20% w
250 RPM, 0% w
250 RPM, 10% w
250 RPM, 20% w
— Estimate 150 RPM 0%w
- - - Estimate 150 RPM 10% w
• - - Estimate 150 RPM 20% w
Estimate 200 RPM 0% w
Estimate 200 RPM 10%w
— Estimate 200 RPM 20% w
— Estimate 250 RPM 0% w
• - - Estimate 250 RPM 10% w
— - Estimate 250 RPM 20% w
0
40
10 20 30
Temperature (C)
Figure 8 Comparison of regression equations (curves) against measured South Louisiana Crude/No Dispersant efficiency.
36
-------
03
2
CD
o.
100
90
80
70
60
40
30
10 —
0
South Louisiana Crude Dispersanf'A"
0> 150RPM, 0%w
^ 150 RPM, 10% w
^ 150 RPM, 20% w
Q 200 RPM, 0% w
n 200 RPM, 10% w
| 200 RPM, 20% w
A 250 RPM, 0% w
A 250 RPM, 10% w
A 250 RPM, 20% w
Estimate 150 RPM 0%w
Estimate 150 RPM 10% w
Estimate 150 RPM 20% w
Estimate 200 RPM 0% w
Estimate 200 RPM 10% w
Estimate 200 RPM 20% w
Estimate 250 RPM 0% w
Estimate 250 RPM 10% w
Estimate 250 RPM 20% w
0 10 20 30 40
Temperature (C)
Figure 9 Comparison of regression equations (curves) against measured South Louisiana Crude/Dispersant "A" efficiency.
37
-------
CO
CO
CD
Q.
CO
Q
£
IUU
90 —
80 —
70
60
_
50 —
40 —
South Louisiana Crude Dispersanf'B"
O 150 RPM, 0% Weathering
^ 150 RPM, 10% Weathering
^ 150 RPM, 20% Weathering
Q 200 RPM, 0% Weathering
Q 200 RPM, 10% Weathering
| 200 RPM, 20% Weathering
A 250 RPM, 0% Weathering
A 250 RPM, 10% Weathering
A 250 RPM, 20% Weathering
Estimate for 150 RPM
Estimate for 200 RPM
Estimate for 250 RPM
30
20
10
0
0 10 20 30 40
Temperature (C)
Figure 10 Comparison of regression equations (curves) against measured South Louisiana Crude/Dispersant "B" efficiency.
38
-------
Regression Equations
A linear regression empirical model was fit to the experimental data for each of the
oil/dispersant combinations. The model takes the following form:
% Efficiencyoil:dispersant = P0 + Pww + P,f + P,5
where w represents weathering in %, t is the temperature (water) in °C, and s is the flask speed in
RPM. The terms were chosen to include linear1 and parabolic2 effects of each variable and
possible two- and three-factor3 interactions. If all variables and interactions were statistically
significant, the model would include 15 terms. Because for each oil/dispersant combination there
are no more than 27 data points, no additional interaction or non-linear terms were included in
the model. Data from the replicate study (Appendix 2) were used to enhance the regressions:
each replicated point at the speed of 200 rpm and dispersants "A" and "B" was replaced by the
average result from the replicate study. As seen in the results (Table 10), only a few terms were
significant for a given oil/dispersant combination as determined by step-wise multiple regression
with an acceptance/rejectance level of 0.05. Between 4 and 9 terms represented all the data for
these experiments on the three oils tested. Notably, the step-wise regression showed that adding
more of the 15 possible terms did not improve the fits.
The various parameters of Equation 5 for the various oil - dispersant combinations are
1 Coefficients appearing in the first line of Equation 5.
Coefficients appearing in the third line of Equation 5.
Coefficients appearing in the second and forth lines of Equation 5.
39
-------
given in Table 10 together with R2 values, which indicates the linearity of the model. Generally,
R2 values above 90% indicate good linear fits. With the exception of 2FO with no dispersant
(86.9%), all the R2 values were above 90%. Regression equation terms that include weathering
as a variable are highlighted in Table 10 with gray shading. Note from the table that none of the
regressions include weathering alone as a term. This indicates the secondary nature of
weathering as a variable as described previously for each oil. Figures 11 to 19 show comparisons
of estimated and measured values of dispersal efficiency. Each of the plots show that the data
cluster along the 1:1 line, indicating, obviously, a close match. Prudhoe Bay Crude with either
dispersant (Figures 14 and 17) and the South Louisiana Crude with dispersant B (Figure 19)
show particularly tight clustering along this line.
40
-------
Table 10 Coefficients of Regression Equations with Terms Determined by Step-Wise Linear Regression
Factor(1)
constant
w
t
s
wt
ws
ts
wts
w2
t2
s2
w2t2
W2S2
t2s2
W2t2S2
R2
Prudhoe Bay Crude
No Dispersant Dispersant A Dispersant B
-5.9325 -264.6 -15.16
1.2090 4.222 3.506
2.609
-8.386e-3
-4.120e-3 -8.386e-3
-4.845e-5
-1.038e-2
-1.979e-2 -9.697e-2 -2.817e-2
1.468e-4 -5.409e-3 1.433e-3
2.6e-7
91.1% 97.5% 98.2%
No. 2 Fuel Oil
No Dispersant Dispersant A Dispersant B
1.490 -112.0 -17.65
10.67 3.032
0.6617
-2.452e-3
-1.4089e-3 -2.435e-2
6.996e-3 -0.2000 -6.313e-2
9.871e-5 1.256e-3
1.39e-6
9e-8 1.30e-6
86.9% 96.7% 94.8%
South Louisiana Crude
No Dispersant Dispersant A Dispersant B
-17.25 41.39
-0.1381 -8.873
0.1680 0.1762
-6.391e-3
-1.631e-3
7.656e-4 4.092e-2
4.382e-3 0.1516
-3.3750e-4
9.99e-6
5e-8
-2.87e-6
98.2% 90.8%
-69.24
-9.149
1.322
4.132e-2
0.1178
-2.970e-3
-2.26e-6
98.6%
(1) ™, =
w = weathering, t = temperature, s = speed
41
-------
I
LU
15 —i
10 —
15 —i
10 —
5 —
15 —i
10 —
5 —
^ I ' I ' I
0 5 10 15 0 5 10 15 0 5 10 15
Measured Measured Measured
Figure 11 Estimated vs Measured % Figure 12 Estimated vs Measured % Figure 13 Estimated vs Measured %
Dispersal of PB C with No Dispersal of 2FO with No Dispersal of SLC with No
Dispersant. Dispersant. Dispersant.
80 100
I ' I ' I
0 20 40 60 80 100 0 20 40 60
Measured Measured
Figure 14 Estimated vs Measured % Figure 15 Estimated vs Measured %
Dispersal of PBC with Dispersant
"A".
I ' I ' I ' I
20 40 60 80 100
Measured
Figure 16 Estimated vs Measured %
Dispersal of 2FO with Dispersant "A" Dispersal of SLC with Dispersant
"A".
E
'•Si
LU
100 —i
80 —
60 —
40 —
20 —
100 —i
80 —
60 —
1
20
100 —,
80 —
60 —
40 —
20 —
80 100
I ' I ' I ' I I ' I ' I ' I '
40 60 80 100 0 20 40 60 80 100 0 20 40 60
Measured Measured Measured
Figure 17 Estimated vs Measured % Figure 18 Estimated vs Measured % Figure 19 Estimated vs Measured %
Dispersal of PBC with Dispersant Dispersal of 2FO with Dispersant "B". Dispersal of SLC with Dispersant
"B". "B".
42
-------
Figure 2 (described previously) shows a comparison of the regression equations and
measured values plotted for the Prudhoe Bay Crude with no dispersant. The squares, for
example, should cluster about the 200 rpm dashed line. The measured values, however, span
almost the entire range of dispersal for speeds of 150 rpm to 250 rpm. This result indicates that
the measured variation in dispersal at 200 rpm is as great as the fitting error in the regression
equations. The coefficients for these regressions (Table 10) contain no terms that involve
weathering. Thus, the amount of volatilization weathering that occurs does not affect the
dispersal efficiency. Therefore, the three curves for the different speeds represent all possibilities
for dispersal of the oil.
Similarly, Figure 3 shows a comparison of the regression equations and measured values
plotted for the Prudhoe Bay Crude with dispersant A. The regression equations for this pair
contain no terms involving weathering (Table 10), so that the regression equations only need to
be plotted for speed and temperature. The graph shows the inverted parabolic shape of the curves
(i.e., highest dispersal at the mid-temperature), and the experimental data for each speed and
percent weathering. That weathering is unimportant for this oil and dispersant is shown by the
data points falling generally near each other regardless of the amount of weathering.
43
-------
Conclusions
A factorial experimental design for determining the impact of temperature, oil type, oil
weathering, and rotation speed on the effectiveness of two dispersants was studied and then
implemented. Regression formulas were developed that provide a smoothed representation of
the data, using a minimum number of significant parameters.
The experimental results obtained in this study reveal the following:
1) Under each set of experimental conditions dispersal increased with increasing flask
speed. This reflects increased mixing energy supplied to the flask for each increase in speed.
When the other variables - temperature and weathering - were varied, however, some lower
speed experiments yielded relatively high dispersal (e.g., PBC-A, SLC-none, SLC-A), that
overlapped the range of dispersals produced at the highest flask speed.
2) Where weathering was significant, increased weathering reduced dispersal (PBC-A,
PBC-B, 2FO-none, SLC-none, SLC-A). For other oil/dispersant combinations, the variation of
dispersal with weathering did not follow a consistent pattern. The magnitude of this effect was
generally small in either case.
3) The changes in dispersal with temperature did not show consistent behavior. In some
cases, the peak dispersal occurred at the middle temperature (PBC-none, PBC-A, 2FO-A, 2FO-B,
SLC-A-250 rpm, SLC-B-250 rpm), contrary to expectation. Also contrary to expectation, in
many of the other experimental combinations, the minimum dispersal also occurred at the middle
temperature (2FO-none, SLC-A-150 rpm, SPC-B-150 rpm, SLC-B-200 rpm).
4) The regression equations provided a reasonable fit to the experimental data. The
equations reproduced, in a smoothed fashion, the main behavior indicated by the data. As
expected, however, the regression curves did not go through every data point. The deviation was
44
-------
partly due to scatter in the data (see SLC, 150 rpm, 22 °C) and partly the nature of regression. In
some cases (e.g., SLC, 250 rpm, 35 °C) the regression curves lay somewhat above all the
measured data points. Improved approaches to fitting, and adding more experimental data might
overcome these difficulties.
In order to establish further the behavior of dispersants with these oils, additional work
could be performed:
1) to establish more strongly the temperature effect,
2) to determine the effect of salinity variation,
3) to determine the amount that viscosity of the test oils increased during weathering,
4) to determine the tendency for dispersed oil to refloat, and
5) to predict dispersal in the flask experiments using the empirical regression equations.
45
-------
References
Bardot, C., C. Bocard, G. Castaing and C. Gatellier, 1984, The Importance of a Dilution Process
to Evaluate Effectiveness and Toxicity of Chemical Dispersants. Proc. Seventh Annual
Arctic Marine Oilspill Program Technical Seminar, 1984.
Blondina, G. I, M. L. Sowby, M. T. Ouano, M. M. Singer and R. S. Tjeerdema, 1997b, "A
Modified Swirling Flask Efficacy Test for Oil Spill Dispersants," Spill Science and
Technology Bulletin 4177-185.
Byford, D. C., P. J. Green and A. Lewis, 1983, Factors Influencing the Performance and
Selection of Low-temperature Dispersants. Proc. Sixth Arctic Marine Oil Spill
Program, 1983, Edmonton, Canada.
Canevari, G. P., P. Calcavecchio, K. W. Becker, R. R. Lessard and R. J. Fiocco, 2001, Key
Parameters Affecting the Dispersion of Viscous Oil. Proc. 2001 International Oil Spill
Conference, Tampa, Florida.
Clayton, J. R., J. R. Payne and J. S. Farlow, 1993, "Oil Spill Dispersants, Mechanical Action and
Laboratory Tests," 5, 55-87.
Dating, P. S., 1988, A Study of the Chemical Dispersibility of Fresh and Weathered Crudes.
Eleventh arctic and marine oilspill program, Vancouver, British Columbia, Canada.
Dating, P. S., D. Mackay, N. Mackay and P. J. Brandvik, 1990, "Dropsize Distributions in
Chemical Dispersion of Oil Spills:Towards a Mathematical Model," Oil & Chemical
Pollution 7173-198.
Delvigne, G. A. L., 1987a, Droplet Size Distribution of Naturally Dispersed Oil. Fate and Effects
of Oil in Marine Ecosystems. Dordrecht, The Netherlands: pp 29-40.
Delvigne, G. A. L., 1987b, Experiments on Natural and Chemical Dispersion of Oil in
Laboratory and Field Circumstances. 1985 Oil Spill Conference, Los Angeles, CA.,
American Petroleum Institute, Washington, D.C.
Desmarquest, J. P., J. Croquette, F. Merlin, C. Bocard, G. Castaing and C. Gatellier, 1985,
Recent Advances in Dispersant Effectiveness Evaluation:Experimental and Field
Aspects. 1985 Oil Spill Conference, Los Angeles, CA, American Petroleum Institute.
46
-------
Fingas, M. F., 1991, Dispersants:A Review of Effectiveness Measures and Laboratory Physical
Studies. Environmental Emergencies Technology Division, Environmenta Canada,
Ottawa, Ontario, Canada.
Fingas, M. F., 2000, Use of Surfactants for Environmental Applications. Surfactants:
Fundamentals and Applications to the Petroleum Industry. Chapter 12.. Cambridge
University Press.: 461-539.
Fingas, M. F., V. M. Dufort, K. A. Hughes, M. A. Bobra and L. V. Duggan, 1989a, Laboratory
Studies on Oil Spill Dispersants. Oil Spill Dispersants:New Ecological Approaches.
ASTMSTP1018. American Society for Testing and Materials, Philadelphia,?A: 207-219.
Fingas, M. F., E. Huang, B. Fieldhouse, L. Wang and J. V. Mullin, 1997, The Effect of Energy.
Settling Time and Shaking Time on the Swirling Flask Dispersant Apparatus.
Proceedings of the Twentieth Arctic Marine Oilspill Program Technical Seminar.
Fingas, M. F., K. A. Hughes and M. A. Schweitzer, 1987, Dispersant Testing at the
Environmental Emergencies Technology Division. Tenth Arctic Marine Oilspill Program
Technical Seminar, Alberta, Canada, Conservation and Protection, Environment Canada.
Fingas, M. F., D. A. Kyle, J. B. Holmes and E. J. Tennyson, 1993a, The Effectiveness of
Dispersants: Variation with Energy. Proceedings of the 1993 International Oil Spill
Conference, American Petroleum Institute, Washington, D.C.
Fingas, M. F., D. A. Kyle and E. J. Tennyson, 1993b, Physical and Chemical Studies on
Dispersants: The Effect of Dispersant Amount and Energy. Proceedings of the Sixteenth
Arctic and Marine Oil Spill Program Technical Seminar.
Fingas, M. F., D. L. Munn, B. White, R. G. Stoodley and I. D. Crerar, 1989b, Laboratory Testing
of Dispersant Effectiveness:The Importance of Oil-to-Water Ratio and Settling Time.
1989 Oil Spill Conference, San Antonio,TX, American Petroleum Institute,Washington,
D.C.
Hoult, D. P., 1972, "Oil Spreading on the Sea," Annu. Rev. Fluid Mech. 4341-368.
IT Corporation, 1995, Investigation of Reported Problems with Swirling Flask Dispersant
Effectiveness Test-Final Report, Contract No. 68-C2-0108, Work Assignment No. 3-53,
JTN 816453, U.S.Environment Protection Agency, Edison, NJ 08837,
47
-------
Lentinen, C. M. and A. M. Vesala, 1984, Effectiveness of Oil Dispersants at Low Salinities and
Low Water Temperatures. Symposium of Oil Spill Chemical Dispersants -Research
Experience and Recommendations, West Palm Beach, Florida, American Society for
Testing and Materials, Philadelphia, PA.
Lessard, R. R. and G. Demarco, 2000, "The Significance of Oil Spill Dispersants," Spill Science
and Technology Bulletin 6(1) 59-68.
Mackay, D., A. Chau, K. Hossain and M. Bobra, 1984, Measurement and Prediction of the
Effectiveness of Oil Spill Chemical Dispersants. Oil Spill Chemical Dispersants.
Research. Experience and Recommendations. ASTM STP 840. American Society for
Testing and Materials, Philadelphia, PA: 38-54.
Mackay, D. and F. Szeto, 1981, The Laboratory Determination of Dispersant Effectiveness-
Method Development and Results. Proc.1981 Oil Spill Conference, Atlanta, GA,
American Petroleum Institute, Washington, D.C.
Ming, L. and G. Chris, 1998, "The Relationship Between Oil Droplet Size and Upper Ocean
Turbulence," Marine Pollution Bulletin 36(12) 961-970.
NRC, 1989, Using Oil Spill Dispersants on the Sea, Report of the Committee on Effectiveness of
Oil Spill Dispersants ,National Academy Press, Washington, D.C,
Paul, G., H. B. Charles and G. P. Stephan, 1999, Proving Dispersants Work. Proc.1999
International Oil Spill Conference, Seattle, Washington, D.C.
Rewick, R. T., K. A. Sabo, J. Gates, J. H. Smith and L. T. McCarthy, 1981, An Evaluation of Oil
Spill Dispersant Testing Requirements. Proc.1981 Oil Spill Conference, Atlanta,GA,
Americam Petroleum Institue, Washington, D.C.
Rewick, R. T., K. A. Sabo and J. H. Smith, 1984, The Drop-weight Interfacial Tension Method
for Predicting Dispersant Performance. Oil Spill Chemical Dispersants. Research
Experience and Recommendations.ASTM STP 840. American Society for Testing
Materials, Philadelphia, PA: 94-107.
Sorial, G. A., K. M. Karen, H. Edith, A. D. Venosa and D. W. King, 2001, Development of a
Rational Oil Spill Dispersant Effectiveness Protocol. Proc. 2001 International Oil Spill
Conference.
48
-------
Serial, G. A., A. D. Venosa, K. M. Koran, E. Holder and D. King, 200la, "Oil Spill Dispersant
Effectiveness Protocol-Part I Impact of Operational Variables," accepted for publication
in ASCE Journal of Environmental Engineering.
Sorial, G. A., A. D. Venosa, K. M. Koran, E. Holder and D. King, 200Ib, "Oil Spill Dispersant
Effectiveness Protocol-Part II Performance of the Revised protocol," accepted for
publication in ASCE Journal of Environmental Engineering.
U.S Environmental Protection Agency, 1984, National Oil and Hazardous Substances Pollution
Contingency Plan:Final Rule, 40 CFR Part 300. Federal Report Register. Vol. 49(29192-
29207).
U.S Environmental Protection Agency, 1996, Swirling Flask Dispersant Effectiveness Test, Title
40 Code of Federal Regulations. Pt.300. appendix 0(245-2501
Venosa, A. D., D. W. King and G. A. Sorial, 2002, "The Baffled Flask Test for Dispersant
Effectiveness: A Round Robin Evaluation of Reproducibility and Repeatability," Spill
Science and Technology Bulletin 7(Nos.5-6) 299-308.
Wang, Z., B. P. Hollebone, M. Fingas, B. Fieldhouse, L. Sigouin, M. Landriault, P. Smith, J.
Noonan and G. Thouin, 2003, Characteristics of Spilled Oils, Fuels, and Petroleum
Products: 1. Composition and Properties of Selected Oils, United States Environmental
Protection Agency, National Exposure Research Laboratory, . Washington, DC.,
EPA/600/R-03/072.
Weaver, J. W., 2004, Characteristics of Spilled Oils, Fuels, and Petroleum Products: 3a.
Simulation of Oil Spills and Dispersants Under Conditions of Uncertainty, United States
Environmental Protection Agency, National Exposure Research Laboratory, Washington,
D.C., EPA 600/R-04/120.
49
-------
Appendix 1: Experimental Data
The following tables contain the dispersant effectiveness data from the set of 243
experiments described previously in the text.
Oil
SLC
SLC
SLC
PBC
PBC
PBC
2FO
2FO
2FO
Weathering
Condition
0%
10%
20%
0%
10%
20%
0%
3.8%
7.6%
Percent
Effectiveness
at 150 rpm
0.51
0.51
0.52
0.00
0.00
0.00
2.24
2.18
2.31
Percent
Effectiveness at
200 rpm
2.73
2.66
2.65
1.58
1.52
1.52
3.73
3.63
3.54
Percent
Effectiveness at
250 rpm
4.42
4.40
4.24
4.45
4.06
3.80
6.44
5.86
5.80
Table 11 Oil control experiments (Temperature = 5±1 °C)
50
-------
Oil
SLC
SLC
SLC
PBC
PBC
PBC
2FO
2FO
2FO
Weathering
Condition
0%
10%
20%
0%
10%
20%
0%
3.8%
7.6%
Percent
Effectiveness
at 150 rpm
54.50
52.51
50.91
21.10
20.96
20.36
21.09
19.21
18.65
Percent
Effectiveness at
200 rpm
77.32
71.60
69.65
58.91
57.31
55.34
45.67
42.67
41.92
Percent
Effectiveness at
250 rpm
89.65
85.90
84.09
69.71
65.75
64.04
79.29
77.90
73.77
Table 12 Oil + Dispersant 'A' experiments (Temperature = 5±1 °C)
51
-------
Oil
SLC
SLC
SLC
PBC
PBC
PBC
2FO
2FO
2FO
Weathering
Condition
0%
10%
20%
0%
10%
20%
0%
3.8%
7.6%
Percent
Effectiveness
at 150 rpm
51.80
50.99
46.52
30.02
28.23
26.67
24.18
24.42
24.41
Percent
Effectiveness at
200 rpm
72.59
71.75
70.71
48.85
48.76
46.44
48.14
47.12
47.01
Percent
Effectiveness at
250 rpm
83.13
81.25
77.73
84.11
83.48
81.21
73.06
72.75
72.25
Table 13 Oil + Dispersant 'B' experiments (Temperature = 5±1 °C)
52
-------
Oil
SLC
SLC
SLC
PBC
PBC
PBC
2FO
2FO
2FO
Weathering
Condition
0%
10%
20%
0%
10%
20%
0%
3.8%
7.6%
Natural Dispersal
150 rpm
1.26
1.10
1.61
3.86
3.44
3.25
3.24
0.00
2.63
Natural Dispersal
at 200 rpm
5.77
4.19
5.19
2.43
4.89
4.63
5.70
0.60
5.57
Natural Dispersal
at 250 rpm
7.07
5.44
8.45
3.81
6.29
5.58
7.12
2.78
7.02
Table 14 Oil control experiments (Temperature = 22 +1-1 °C.)
53
-------
Oil
SLC
SLC
SLC
PBC
PBC
PBC
2FO
2FO
2FO
Weathering
Condition
0%
10%
20%
0%
10%
20%
0%
3.8%
7.6%
Percent
Effectiveness
at 150 rpm
57.19
34.21
39.79
40.82
60.55
43.20
48.99
58.25
69.28
Percent
Effectiveness at
200 rpm
87.64
69.76
91.97
80.12
84.33
86.00
82.93
84.07
87.38
Percent
Effectiveness at
250 rpm
97.31
89.80
92.88
95.95
97.94
90.61
95.52
97.85
98.76
Table 15 Oil + Dispersant 'A' experiments ( Temperature = 22±1 °C )
54
-------
Oil
SLC
SLC
SLC
PBC
PBC
PBC
2FO
2FO
2FO
Weathering
Condition
0%
10%
20%
0%
10%
20%
0%
3.8%
7.6%
Percent
Effectiveness
at 150 rpm
28.64
26.99
26.42
52.32
49.24
47.69
48.94
33.38
56.39
Percent
Effectiveness at
200 rpm
76.67
75.45
73.54
69.18
63.47
62.34
78.00
76.49
75.89
Percent
Effectiveness at
250 rpm
90.68
86.03
88.99
87.96
85.15
83.53
98.24
92.79
94.75
Table 16 Oil + Dispersant 'B' experiments ( Temperature = 22±1 °C )
55
-------
Oil
SLC
SLC
SLC
PBC
PBC
PBC
2FO
2FO
2FO
Weathering
Condition
0%
10%
20%
0%
10%
20%
0%
3.8%
7.6%
Natural Dispersal
at 150 rpm
4.92
4.12
5.39
0.53
1.04
1.06
6.89
6.23
7.44
Natural Dispersal
at 200 rpm
8.90
8.25
10.39
1.77
1.63
1.74
7.84
7.50
7.33
Natural Dispersal
at 250 rpm
11.09
9.68
11.94
5.47
4.76
4.46
10.81
10.73
10.22
Table 17 Oil control experiments ( Temperature = 35 ±1 °C )
56
-------
Oil
SLC
SLC
SLC
PBC
PBC
PBC
2FO
2FO
2FO
Weathering
Condition
0%
10%
20%
0%
10%
20%
0%
3.8%
7.6%
Natural Dispersal
at 150 rpm
81.64
74.12
73.81
33.62
33.06
31.93
34.60
34.19
32.19
Natural Dispersal
at 200 rpm
96.28
94.20
87.64
72.21
71.47
68.46
47.92
46.55
45.11
Natural Dispersal
at 250 rpm
98.17
97.41
90.69
75.92
72.79
71.68
89.79
76.87
75.62
Table 18 Oil + Dispersant 'A' experiments ( Temperature = 35 ±1 °C )
57
-------
Oil
SLC
SLC
SLC
PBC
PBC
PBC
2FO
2FO
2FO
Weathering
Condition
0%
10%
20%
0%
10%
20%
0%
3.8%
7.6%
Percent
Effectiveness
at 150 rpm
42.88
42.68
41.60
63.59
58.23
56.29
41.74
41.53
41.71
Percent
Effectiveness at
200 rpm
76.69
77.57
76.11
67.69
67.02
66.87
53.13
52.67
52.64
Percent
Effectiveness at
250 rpm
92.06
91.39
87.76
89.23
82.34
80.79
98.92
97.98
96.06
Table 19 Oil + Dispersant 'B' experiments ( Temperature = 35 ±1 °C )
58
-------
Weathering
Oil Condition
SLC
SLC
SLC
PBC
PBC
PBC
2FO
2FO
2FO
0%
10%
20%
0%
10%
20%
0%
3.8%
7.6%
%Effectiveness at 5
°C
89.7
85.9
84.1
69.7
65.8
64.0
79.3
77.9
73.8
%Effectiveness at 22
°C
97.3
89.8
92.9
96.0
97.9
90.6
95.5
97.9
98.8
%Effectiveness at 35 °C
98.2
97.4
90.7
75.9
72.8
71.7
89.8
76.9
75.6
Table 20 Results of oil-dispersant combinations at different temperatures (Dispersant 'A', 250
rpm, 34 ppt)
59
-------
Appendix 2: Replicate Study
Based on a statistical analysis (ANOVA) of the factorial experimental design, and
assumed linear interactions between factors, significant two-way interactions were identified.
Table 21 lists these and the specific conditions selected for replication. Two replicates were
included for each two-way interaction and four replicates of two of these experiments were
conducted.
The results of the three two-way interactions namely: temperature by dispersant,
volatilization by dispersant and speed by dispersant are shown in Tables 22 to 24. Factors that
were not involved in the two-way interactions were fixed at certain values: speed-250 rpm,
volatilazation-0%, and temp-22 °C. These tables show a comparison between the replicate
results and the results reported earlier. This comparison was based on determining the Relative
Percent Difference (RPD) which is given by:
Avg. eff of current results - Eff of previous results 1ftft
— x iuu (6)
Avg. eff of current results
The acceptance criterion was an RPD of less than 15%.
A four-replicate study was also conducted for all the three oils, the three levels of
volatilization and the two dispersants at 22 °C and a flask speed of 200 rpm. The total number of
experiments conducted was thus: 3 oils * 3 volatilization * 2 dispersants * 4 replicates =72. The
precision objectives were determined by using the relative standard deviation (RSD) for percent
effectiveness based on four replicate flasks. The results of these experimental runs are shown in
Tables 25 and 26. The RSD is given by:
60
-------
RSD =
Standard Deivation
Average Effectiveness
x 100
(7)
The acceptance criterion an RSD of less than 15%. The replicate data were used in developing
the regression formulas described in the main body of the report.
OIL
2FO
SLC
SLC
PBC
PBC
PBC
INTERACTION
Temperature *
Dispersant
Temperature *
Dispersant
Speed * Dispersant
Temperature *
Dispersant
Volatilization *
Dispersant
Speed * Dispersant
CONDITIONS NO: OF
EXPERIMENTAL
RUNS
0% volatilization * 2 dispersants * 3
temperatures *250 rpm * 2 replicates
0% volatilization * 2 dispersants * 3
temperatures *250 rpm * 2 replicates
0% volatilization * 2 dispersants * 3
speeds * 22 °C * 4 replicates
0% volatilization * 2 dispersants * 3
temperatures *250 rpm * 2 replicates
22 °C * 250 rpm * 3 volatilizations * 2
dispersants * 2 replicates
0% volatilization * 2 dispersants * 3
speeds * 22 °C * 4 replicates
12
12
24
12
12
24
Table 21 Two way interactions.
61
-------
Oil
PBC
PBC
PBC
PBC
PBC
PBC
2FO
2FO
2FO
2FO
2FO
2FO
SLC
SLC
SLC
SLC
SLC
SLC
Temp(°C)
22
35
6
22
35
6
22
35
6
22
35
6
22
35
6
22
35
6
Dispersant %
Effectiveness
of replicates
A
A
A
B
B
B
A
A
A
B
B
B
A
A
A
B
B
B
Rl
96.20
80.32
71.41
88.66
95.62
82.41
95.97
74.28
75.11
90.64
90.79
74.26
96.25
98.69
90.18
92.82
91.29
85.58
R2
94.05
74.20
87.01
84.76
94.38
70.04
95.87
81.60
64.06
92.13
96.84
75.08
98.96
98.33
90.11
88.60
83.73
85.74
Avg eff.
95.12
77.26
79.21
86.71
94.99
76.22
95.92
77.94
69.58
91.39
88.82
74.66
97.60
98.51
90.15
90.71
87.51
85.66
RSD
1.59
5.59
13.92
3.18
0.92
11.47
0.07
6.65
11.22
1.15
3.14
0.77
1.96
0.25
0.05
3.28
6.11
0.13
%Eff. of
previous
samples
95.95
75.92
69.71
87.96
89.23
84.11
95.63
89.79
79.29
98.24
98.92
73.06
97.35
98.17
89.65
90.68
92.06
83.13
RPD
0.87
1.73
11.99
1.44
6.07
10.34
0.30
15.20
13.95
7.50
11.37
2.15
0.26
0.35
0.55
0.03
5.19
2.96
Table 22 Dispersant by temperature interaction (Flask speed =250 rpm, Weathering =0%).
62
-------
Oil
Weathering
Dispersant
% Effectiveness
replicates
of
Avg
eff.
RSD
Rl R2
PBC
PBC
PBC
PBC
PBC
PBC
0%
10%
20%
0%
10%
20%
A
A
A
B
B
B
89.96
99.82
90.08
89.89
87.21
83.91
93.76
96.88
94.72
86.19
86.63
88.84
91.86
98.35
92.40
88.04
86.92
86.37
2.92
2.11
3.55
2.96
0.47
4.04
%Eff. of
previous
samples
95.95
97.94
90.61
87.96
85.15
83.53
RPD
4.45
0.42
1.93
0.09
2.03
3.29
Table 23 Dispersant by volatilization interaction (Flask speed =250 rpm, Temperature = 22±1 °C).
63
-------
Oil
SLC
SLC
SLC
PBC
PBC
PBC
SLC
SLC
SLC
PBC
PBC
PBC
Speed
150
200
250
150
200
250
150
200
250
150
200
250
Disp
A
A
A
A
A
A
B
B
B
B
B
B
% Effectiveness of replicates
Rl
54.69
88.70
99.52
51.25
79.38
99.05
30.48
71.38
91.03
49.73
76.09
92.46
R2
55.7
92.51
98.40
46.37
82.43
95.80
30.18
72.29
93.74
49.87
76.15
88.49
R3
55.18
87.71
98.05
48.00
81.28
96.05
31.47
71.79
89.38
49.70
72.15
93.77
R4
69.97
87.88
96.90
46.38
78.99
98.10
28.79
70.98
99.74
48.76
72.21
97.78
Avg eff.
58.
89.
98.
48.
80.
97.
30.
71.
93.
49.
74.
93.
.88
.20
.20
.00
.52
.20
.23
.61
.47
.51
.15
.12
RSD
12.56
2.52
1.09
4.78
2.01
1.60
3.67
0.78
4.86
1.03
3.07
4.11
%Eff. of
previous
samples
57.21
87.45
97.35
40.82
80.12
95.95
28.64
76.67
90.68
52.32
69.18
87.96
RPD
2.87
1.98
0.88
14.96
0.49
1.33
5.40
7.06
3.03
5.51
6.70
5.54
Table 24 Dispersant by speed interaction ( Weathering = 0%, Temperature = 22 ± 1 °C).
64
-------
Oil
SLC 0%
SLC 10%
SLC 20%
PBC 0%
PBC 10%
PBC 20%
2FO 0%
2FO 3.8%
2FO 7.6%
% Effectiveness of replicate samples
Rl
88.99
70.54
87.80
80.11
86.24
88.43
81.26
92.54
89.06
R2
88.30
73.89
87.80
78.63
88.14
89.07
80.68
88.66
84.79
R3
88.28
67.44
89.75
82.43
88.65
83.79
76.91
91.94
82.89
R4
87.79
73.48
91.13
78.92
90.09
65.78
78.87
88.66
86.08
Average
effectiveness
88.34
71.33
89.12
80.02
88.28
81.76
79.43
90.45
85.71
RSD
0.55
4.20
1.82
2.16
1.80
13.35
2.47
2.30
3.02
%Eff. of
previous
samples
87.45
69.94
91.97
80.12
84.33
86.00
82.97
84.03
87.88
RPD
1.01
1.96
3.14
0.12
4.57
5.17
4.45
7.09
2.54
Table 25 Replicate determination: Oil + Dispersant 'A' experiments ( Temperature = 22±1°C , Speed = 200 rpm).
65
-------
Oil
SLC 0%
SLC 10%
SLC 20%
PBC 0%
PBC 10%
PBC 20%
2FO 0%
2FO 3.8%
2FO 7.6%
% Effectiveness of replicate
Rl
66.16
76.38
73.96
75.18
69.18
68.32
71.70
73.34
66.63
R2
72.35
75.76
74.22
75.24
72.82
69.01
70.22
75.70
63.89
R3
70.13
75.33
71.90
71.28
79.88
69.66
74.00
71.84
75.05
samples
R4
65.29
75.40
72.75
71.34
73.34
68.77
76.59
73.20
68.80
Average
effectiveness
68.48
75.71
73.20
73.26
73.81
68.94
73.12
73.52
68.60
RSD
4.86
0.64
1.48
3.07
6.03
0.81
3.81
2.18
6.92
%Eff. of
previous
samples
76.67
75.45
73.54
69.18
63.47
62.34
78.00
76.49
75.89
RPD
11.28
0.34
0.46
5.57
14.00
9.57
6.66
4.03
10.64
Table 26 Replicate determination: Oil + Dispersant 'B' experiments ( Temperature = 22±1 °C , Speed = 200 rpm).
The entire suite of replicates are summarized in Tables 27, 28, and 29 and Figures 20, 21,
and 22. Figures 20, 21, and 22 show plots of the average of all measurements versus the first
single measured value. The latter are the original data measured before the replicate study.
These plots show generally that the averages and the first single values show a one-to-one
relationship.
66
-------
Figure 20 Average vs "First" single Figure 21 Average vs "First" single Figure 22 Average vs "First single
measured dispersal efficiency for measured dispersal efficiency for measured dispersal efficiency for
PBC. 2FO. SLC.
The tables list the values of the experimental conditions (weathering, speed, and
temperature), the number of replicates, the average and standard deviation of the replicates and
the first single value measured. The last column gives the difference between the original
measurement (first single value) and the average of all measurements in units of standard
deviation.
NSD =
% Efficiencyoriginal
measurement
- Average % Efficiency
Standard Deviation
(8)
This measure indicates how the first measurement made of dispersal for a given set of conditions
differed from the average and indicates how well a single measurement predicts the average.
67
-------
Weathering
(%)
Speed
(rpm)
Temperature
(°C)
Number of
Replicates
Average of all
Experiments
Standard
Deviation
First
Single
Value
NSD
PrudhoeBay Crude with Dispersant "A"
0
0
0
0
0
10
10
20
20
150
200
250
250
250
200
250
200
200
22
22
5
22
35
22
22
22
22
5
9
o
J
9
o
J
5
o
J
4
o
J
46.56
80.25
76.04
95.44
76.81
87.49
98.21
81.77
91.80
3.778
1.472
9.535
2.651
3.156
2.230
1.489
9.640
2.540
40.82
80.12
69.71
95.95
75.92
84.33
97.94
88.43
90.61
-1.519
-0.09
-0.664
0.1924
-0.282
-1.417
-0.181
0.6909
-0.469
Prudhoe Bay Crude with Dispersant "B"
0
0
0
0
0
10
10
20
20
150
200
250
250
250
200
250
200
250
22
22
5
22
35
22
22
22
22
5
9
o
J
10
o
J
5
o
J
5
o
J
50.08
73.20
78.85
89.80
93.08
71.74
86.33
67.62
85.43
1.330
2.513
7.680
3.865
3.389
6.019
1.062
2.991
2.962
52.32
69.18
84.11
87.96
89.23
63.47
85.15
62.34
83.53
1.6842
-1.6
0.6849
-0.476
-1.136
-1.374
-1.111
-1.765
-0.641
Table 27 Summary of Prudhoe Bay Crude replicates.
68
-------
Weathering
(%)
Speed
(rpm)
Temperature
(°C)
Number of
Replicates
Average of all
Experiments
Standard
Deviation
First
Single
Value
NSD
No. 2 Fuel Oil with Dispersant "A"
0
0
0
0
3.8
7.6
200
250
250
250
200
200
22
5
22
35
22
22
5
4
4
4
5
5
80.14
72.01
95.85
80.90
89.17
86.14
2.324
2.555
0.1506
6.636
3.390
2.447
82.97
79.29
95.63
89.79
84.03
87.88
1.2177
2.8493
-1.461
1.3397
-1.516
0.7111
No. 2 Fuel Oil with Dispersant "B"
0
0
0
0
3.8
7.6
200
250
250
250
200
200
22
5
22
35
22
22
5
4
4
4
5
5
74.10
74.26
93.10
93.84
74.11
70.05
3.239
0.8703
3.480
4.807
1.921
5.251
78.00
73.06
98.24
98.92
76.49
75.89
1.2041
-1.379
1.477
1.0568
1.2389
1.1122
Table 28 Summary of No. 2 Fuel Oil replicates.
69
-------
Weathering
(%)
Speed
(rpm)
Temperatur
e(°C)
Number of
Replicates
Average of
all
Experiments
Standard
Deviation
First
Single
Value
NSD
South Louisiana Crude with Dispersant "A"
0
0
0
0
0
10
20
150
200
250
250
250
200
200
22
22
5
22
35
22
22
5
9
o
J
7
o
J
5
5
58.55
88.62
89.98
97.92
98.40
71.06
89.69
6.454
1.538
0.2879
1.158
1.363
2.669
1.899
57.21
87.45
89.65
97.35
98.17
69.94
91.97
-0.208
-0.761
-1.146
-0.492
-0.169
-0.42
1.2006
South Louisiana Crude with Dispersant "B"
0
0
0
0
0
10
20
150
200
250
250
250
200
200
22
22
5
22
35
22
22
5
8
o
J
7
o
J
5
5
29.91
71.36
84.82
92.28
89.03
75.66
73.27
1.194
3.133
1.463
3.745
4.603
0.4327
0.9485
28.64
76.67
83.13
90.68
92.06
75.45
73.54
-1.064
1.6949
-1.155
-0.427
0.6583
-0.485
0.2847
Table 29 Summary of South Louisiana Grade replicates.
70
-------
Appendix 3: SOPs Appearing in the Quality Assurance Project Plan
These procedures were taken from the Quality Assurance Project Plan (QAPP) developed
for this project (Suidan et a/., 2001).
Quality Assurance Project Plan Appendix 1: SOP for Oil Alone Stock Standard Preparation
a. Weigh a clean vial with a loose Teflon and aluminum cap (x grams).
b. Add 2 ml of the specific reference oil to the vial and re-weigh vial with the cap (y grams).
c. Add 18 ml DCM to the vial and re-weigh the vial with the cap (z grams).
d. Crimp the aluminum cap and mix the vial contents by hand shaking.
e. Measure the density of the specific reference oil + DCM by either using a density bottle
or a 1 ml gas tight syringe (first weigh the syringe empty and then when full up to the 1
ml mark with the solution) (poil+DCM> §/L)-
f. Determine the concentration of the oil solution
(y-x)
Concentration, g/L = (1)
71
-------
Quality Assurance Project Plan Appendix 2: SOP for Oil Plus Dispersant Stock Standard
Preparation
a. Weigh a clean vial with a loose Teflon and aluminum cap (x' grams).
b. Add 2 ml of the specific reference oil to the vial followed by 80 |j,L of the
dispersant (to make a ratio of dispersantoil = 1:25) and re-weigh vial with the cap
(y' grams).
c. Add 18 ml DCM to the vial and re-weigh the vial with the cap (z' grams).
d. Crimp the aluminum cap and mix the vial contents by hand shaking.
e. Measure the density of the specific reference oil + dispersant + DCM by either
using a density bottle or a 1 ml gas tight syringe (first weigh the syringe empty
and then when full up to the 1 ml mark with the solution) (p'oii+dispersant+ocM^ §/L)-
f Determine the concentration of the oil solution
(y'-x')
Concentration, g / L = ———— (2)
(Z - X ) / P0ii+diSpersant+DCM
72
-------
Quality Assurance Project Plan Appendix 3: SOP for Preparation of Standard Solutions from
the Stock Standard Solutions
a. Add a specific volume (see Step f for the required volumes) of the stock oil-DCM
solution prepared in Quality Assurance Project Plan Appendix 1, or of the stock
oil-dispersant-DCM solution prepared in Quality Assurance Project Plan
Appendix 2, to 30 ml of synthetic sea water in a separately funnel.
b. Extract the oil-water mixture with a 5-ml volume of DCM (15 seconds of
vigorous shaking followed by 2-minute stationary period to allow for phase
separation). For this first extraction, drain only about 3 ml from the separately
funnel due to a web-like emulsion formation at the solvent/water interface. Collect
the extract in a 25-mL graduated cylinder.
c. Repeat the extraction two more times (3 times total), each time using a 5-ml
portion of DCM, and drain the solvent to the solvent water interface (no web like
emulsions are formed after the first extraction). Do not allow any water to be
drained. Collect the extracts in a 25-mL graduated cylinder. Adjust the final
DCM volume for the combined extracts to 20 ml with DCM.
d. Transfer the DCM extract to a 50-ml crimp style glass vial with an
aluminum/Teflon seal. Mix contents of the sealed vial by inverting at least 10
times.
e. Store the vials at 4 ± 2 °C until the time of analysis.
f. The quantities of oil (or oil/dispersant) used to achieve the desired six
concentrations for developing a 6-point calibration curve are as follows: (a) for
Alaska North Slope Crude and Number 2 Fuel Oil, the volumes of the stock oil
(or oil/dispersant) standard added to the 30 ml seawater are: 1 l|iL, 20 |iL, 50 joL,
73
-------
100 |iL, 125 |iL, and 150 |iL; (b) for South Louisiana Crude the volumes are: 20
ML, 50 |iL, 100 |iL, 150 |iL, 200 |iL, and 300 |iL. For Alaska North Slope Crude
and Number 2 Fuel Oil, the maximum volume that can be added is 150 |iL
because absorbence saturation in the spectrophotometer occurs above this
concentration value. For South Louisiana Crude, volumes below 20 |iL are not
considered because the absorbence is close to the detection limit of the
spectrophotometer.
74
-------
Quality Assurance Project Plan Appendix 4: SOP for Spectrophotometer Calibration
a. Remove the standard vials from the cold room or the refrigerator and allow to
equilibrate at the laboratory temperature.
b. Scan a pure DCM solvent at the three analytical wavelengths, namely, 340, 370,
and 400 nm, and set the calibration at a zero absorbence reading.
c. Determine the absorbence of the six standards at each of the three analytical
wavelengths (i.e., 340, 370, and 400 nm). The calibration standards should be
introduced in increasing order of concentration.
d. Determine the area under the absorbence vs. wavelength curve between
wavelengths 340 and 400 nm by using the trapezoidal rule according to the
following equation:
(Abs340 + Abs370)*30 (Abs370 + Abs400)*30
Area= + (3)
e. Before DCM-extracts of dispersed oil-water samples can be analyzed for their oil
content, the UV-visible Spectrophotometer must meet an instrument-stability
criterion. This criterion is determined with the six oil standards analyzed in step c
and involves determination of response factors (RFs) for the oil at each
concentration level. The response factor at each concentration level is determined
by using the following equation:
Theoretical Concentration
Area
75
-------
where: theoretical concentration = oil concentration in g/L of DCM in standard
solution; Area = the area under the absorbence vs. wavelength curve between
wavelengths 340 and 400 nm determined by equation (3).
f. Instrument stability for the initial calibration is acceptable when the RFs of the six
standard extracts are <10% different from the overall mean value for the six
standards.
IRF-AV.RE] M
%dlerence=! [xlOO (5)
g. If one or more of the standard oil extracts does not meet this linear-stability
criterion, then the "offending" standard(s) must be prepared a second time (i.e.,
extraction of the specified amount of oil from 30-ml of seawater for the
"offending" standard according to the procedure in Quality Assurance Project
Plan Appendix 3). If replacement of the re-analyzed standard solution(s) in the
standard curve meets linear stability (i.e., no RF > 10% different from the overall
mean), then the analysis of the sample extracts can begin.
h. If the initial-stability criterion is still not satisfied, analysis of the sample extract
cannot begin, and the source problem (e.g., preparation protocol for the oil
standards, spectrophotometer stability, etc.) must be determined and corrected.
i. Determine the slope of the calibration points by using linear regression with zero
intercept.
y--mx (6)
where:
76
-------
y = area under absorbence curve; x = concentration of oil, g/L; m = slope
j. The initial six-point calibration of the UV-visible spectrophotometer is required at
least once per day.
77
-------
Quality Assurance Project Plan Appendix 5. SOP for SAMPLE ANALYSIS
a. Remove the experimental water sample extract vials from the cold room or
refrigerator and allow to equilibrate at the laboratory temperature.
b. Once a successful initial calibration curve has been established and verified,
introduce the experimental method blank sample (i.e., sea water alone run in the
experimental flask and then a 30 ml sample withdrawn and extracted with DCM
three times), followed by the experimental samples. Analyze the samples at the
three wavelengths (i.e., 340, 370, and 400 nm) and record the values.
c. If any sample result exceeds the highest calibration standard (which was
sometimes the case for Prudhoe Bay Crude Oil), the sample needs to be diluted
with DCM and re-analyzed to fall within the calibration range of the instrument.
If any sample result is less than the lowest calibration standard, report the result
with a footnote indicating that the concentration value was lower than the lowest
calibration standard.
d. Introduce a standard check after the analysis of the four experimental samples.
e. The sequence of analyses is thus: (1) solvent blank; (2) six calibration standards
for the specific test oil plus dispersant; (3) experimental sample method blank; (4)
four experimental samples (same test oil and dispersant); (5) solvent blank; (6)
mid-point standard calibration check; (7) six calibration standards for the specific
test oil alone; and (8) experimental sample of oil control blank (i.e., oil with no
dispersant). It is worthwhile to note that for the full test only 4 replicates of oil
control blanks of each test oil and 4 replicates of method blanks are conducted.
Therefore, the above sequence is applicable only when a method blank and oil
control blank are analyzed. Otherwise, skip sequence # 3, 7, and 8 if no blanks
78
-------
are analyzed.
f. The acceptance criterion for the four or more experimental sample replicates is
based upon the relative standard deviation (RSD) as determined by Equation 7 of
Appendix 2 to be less than 15%. If the RSD is greater than 15% for one of the
replicates, this replicate should be flagged and one more replicate run for this
specific oil/dispersant test. However, if the replicates seemed clustered into two
groups and the RSD is greater than 15%, all the replicates for this specific
oil/dispersant test need to be rerun and the original data flagged. If a situation
occurs where the acceptance criterion is not met for the third trial, report all
results, and indicate that the acceptance criterion was never met.
g. At least 5% of all UV-visible spectrophotometric measurements will be performed
in duplicate as a QC check on the analytical method. The absorbence values for
the duplicates should agree within ± 5% of their mean values.
h. Review QC results as early as possible to determine if all acceptance criteria are
met. Notify data user if any QC check falls outside of acceptance limit.
79
-------
Quality Assurance Project Plan Appendix 6. Calculation Procedure for Experimental
Samples
a. The calculation procedure for determining the calibration curve is presented under
Quality Assurance Project Plan Appendix 4.
b. Determine the area for the absorbence values obtained for the experimental
samples by using Equation (3) in Quality Assurance Project Plan Appendix 4.
c. Determine the concentration of oil of the experimental samples by using the
following equation:
[Area as determined by Equation (3)) , .
Concentration of oil, g / L = —— — — : I')
v Slope or the calibration curve )
d. Determine the mass of oil dispersed in the 30 ml of extracted experimental sample
by using the following equation:
Mass of oil, g = Concentration of oil * VDCM (8)
where VDCM = the final volume of the DCM-extract of the water sample (0.020 L).
e. Determine the mass of the total oil dispersed by the following equation:
Vtw
Total oil dispersed, g = Mass of oil * —— (9)
V * f*m /
ew-
where:
Vtw = total water volume in the testing flask (120 ml),
Vew = volume of water extracted for dispersed oil content (30 ml).
80
-------
Determine the dispersant performance (i.e., percent of oil dispersed, or Eff) based
on the ratio of oil dispersed in the test system to the total oil added to the system
as follows:
Total Oil Dispersed xlOQ (1Q)
Ail * Voi!
where:
poil = density of the specific test oil, g/L
Voil = Volume (L) of oil added to the test flask (100 |iL = 10'4 L)
g. Calculate Effusing Equation 10 for coupled experiments with and without
dispersant (Effd and Effc, respectively). Effc is the effectiveness of the control
and represents natural dispersion of oil in the test apparatus. It is determined
by using the average of the four replicates of the specific oil control runs.
Effd is the measured, uncorrected value.
h. Calculate the final dispersant performance of the chemical dispersant agent
after correcting for natural dispersion using the following equation
EffD = Effd - Effc (ll)
where:
EffD = percent dispersed oil due to dispersant only,
Effd = percent oil dispersed with dispersant added,
Effc = percent oil dispersed with no dispersant added (average of the four
replicate oil controls).
i. Report all the calculations in the form of a spread sheet.
81
-------
j. Report number of times test acceptance criteria (when the replicates for the
experimental samples are conducted) were violated and how the problem was rectified.
82
------- |