WATER POLLUTION CONTROL RESEARCH SERIES • DAST-12
           OIL SAMPLING TECHNIQUES
S DEPARTMENT OF THE INTERIOR • FEDERAL WATER POLLUTION CONTROL ADMINISTRATION

-------
          WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Reports describe the
results and progress in the control and abatement of pollu-
tion of our Nation's waters.  They provide a central source
of information on the research, development and demonstra-
tion activities of the Federal Water Pollution Control
Administration, Department of the Interior, through inhouse
research and grants and contracts with Federal, State, and
local agencies, research institutions, and industrial organ-
izations.

Water Pollution Control Research Reports will be distributed
to requesters as supplies permit.  Requests should be sent to
the Planning and Resources Office, Office of Research and
Development, Federal Water Pollution Control Administration,
Departjnent of the Interior, Washington, D. C.  20242.

-------
PROGRESS REPORT:
                      OIL SAMPLING TECHNIQUES
                            Compiled By
             Oil & Hazardous Materials Research Section
         Northeast Region Research and Development Program
                  Edison Water Quality Laboratory
                        Edison, N. J.  08817
                              for the
           FEDERAL WATER POLLUTION CONTROL ADMINISTRATION
                  U. S. DEPARTMENT OF THE INTERIOR
                  Oil Pollution 	 15080QBJ 12/69
                           December 1969

-------
             FWPCA Review Notice
This report has been reviewed by the Federal
Water Pollution Control Administration and
approved for publication.  Approval does not
signify that the contents necessarily reflect
the views and policies of the Federal Water
Pollution Control Administration, nor does
mention of trade names or commercial products
constitute endorsement or recommendation for
use.

-------
                              CONTENTS

                                                              PAGE
Abstract ...........................
List of Figures  .......................   iv
List of Tables ........................    v
Introduction .................... .....    1

Section I - Preliminary Study on the Use of Solid Absorbents  .    3
            By:  Oil & Hazardous Materials Research.
                 Section, Edison, N. J.

       II - Oil Sampling Methods ...............   13
            By:  Fred K. Kawahara, Ph.D., FWPCA,
                 Cinn. , Ohio

      III - Oil Slick Collecting Device ............   17
            By:  Admiralty Materials Laboratory,
                 British Royal Navy, 1969

       IV - Sampling In Waters Polluted by Hydrocarbons,         21
            Part I  ......................
            By:  F. Edeline, et al . , Cebedeau Laboratory

        V - Sampling In Waters Polluted by Hydrocarbons,         35
            Part II  .....................
            By:  F. Edeline, et al . , Cebedeau Laboratory

       VI - Collection of  Slick Forming Materials  from the       49
            Sea Surface   ...................
            By:  W. D. Garrett, U.  S.  Naval Research
                 Laboratory

-------
                               FIGURES

                                                                 PAGE

Section I - Preliminary Study on the Use of Solid Absorbents

    Figure I - Absorbence of South Louisiana Crude Oil
               by Polyurethane Foam A	        10

          II - Absorbence of South Louisiana Crude Oil
               by Polyurethane Foam B	        10

         III - Absorbence of South Louisiana Crude Oil
               by Polyurethane Foam C	        10

          IV - VI  - Photographs of Field Tests	     11&12


Section IV - Sampling in Waters Polluted by Hydrocarbons, Part I

    Figure I - Sliding Cylinder	        27


Section  V - Sampling in Waters Polluted by Hydrocarbons, Part II

    Figure I - Movable Frame	        40

          II - Filter Aspiration Device	        40

         III - Semimicroextractor	        40

          IV - Chromatographic Tube	        40


Section VI - Collection of Slick-Forming Materials from the Sea Surface

    Figure I - Sea-Surface Collection Assembly
               in Draining Position 	        58

          II - Top View of Tank Experiment to Study
               Monolayer Retrieval  Method 	        58
                                  LV

-------
                                TABLES

                                                                PAGE

Section I - Preliminary Study on the Use of Solid Absorbents

    Table I - Oil Distribution on a Water Surface	    5

         II - Results of Oil Recovery Experiment
              Using Polyurethane Foams   	    8
Section IV - Sampling In Waters Polluted by Hydrocarbons, Part I

    Table I - Results of Camphor Test  .............   29

         II - Visual Description of Oil Slick Correlated          30
              with Thickness   .................

        Ill - Reproducibility Results Using Pail With Tap         31
              Method, Point A  .................

         IV - Reproducibility Results Using Pail With Tap         31
              Method, Point B  . .  . ..............

          V - Reproducibility Results Using Simple                32
              Pail Method  ...................

         VI - Reproducibility Results Using Sliding               32
              Cylinder Method  .................

        VII - Reproducibility Results Using Movable               33
              Probe Method  ..................

       VIII - Statistical Comparison of Various Methods  ....   34


Section V - Sampling in Waters Polluted by Hydrocarbons, Part  II

    Table I - Reproducibility Results Using Filter Paper          42
              Disk Method, Laboratory Tests  ..........
         II - Recovery of Oils From  Industrial  Plant Using        43
              Filter Paper Disk Method   ............

        Ill - Reproducibility Results Using Movable               44
              Frame Method   ..................

-------
                           TABLES CONT'D.

                                                                PAGE
   Table IV - Reproduclbility Results Using Filter
              Aspiration Device, Point A
          V - Reproducibility Results Using Filter               45
              Aspiration Device, Point B ..........

         VI - Cold Elution Recovery  ............     46

        VII - Hot Extraction Recovery  ...........     46

       VIII - Sampling Methods - Summary ..........     47
Section VI - Collection of Slick-Forming Materials From The
             Sea Surface

    Table I - Removal of Monolayers from Water Surfaces            56
             by Metal Screens  	

-------
                                 ABSTRACT









     Sampling of oil in the environment, depending upon the thickness




of the slick, can present certain operational problems, most paramount




of which is the collection of an adequate volume of sample required




for identification by chemical analyses.  Several basic "dip-stick"




techniques, which are primarily applicable for sampling slicks with




a thickness of greater than 2 mm, as well as suggested methods for




sampling thin oil slicks are discussed and illustrated.  Included in




this report are preliminary results on oil entrapment by solid absor-




bents obtained by the Oil & Hazardous Materials Research Section,




Edison, New Jersey.  Also reported are results of investigations per-




formed by foreign and U. S. scientists, using various types of




sampling equipment and materials.
KEY WORDS:  Oil, Sampling, Absorption-, Separation
                                V1L

-------
                               INTRODUCTION
      This report provides a description of selected methods known to
be available for obtaining representative samples of "thin" (less than
2 mm thick) oil slicks in the water environment.  The purpose of pre-
senting these techniques is not to recommend any one method of sampling;
but rather, to make available the latest technical information to per-
sonnel concerned with the problem of collecting representative samples
of oil slicks.  Needless to say, the methods outlined in this report
supplement the already known basic "dip-stick" methods for sampling
heavy or thick oil slicks.  Some of these techniques, which were dis-
cussed in the publication "Laboratory Guide for the Identification of
Petroleum Products", January, 1969, are repeated in this report for
purposes of continuity.

      Regardless of what method is used for sampling, it is important
that a chain of custody of the samples be properly maintained and record-
ed from the time the samples are taken until the litigation is completed.
In this regard, a record of time, place, and name and title of the person
taking the sample, and each person handling the sample thereafter must be
maintained and forwarded with the sample, using Form No. FWPCA 208 (7-68).

      Many precautions must also be observed when handling oil samples for
analyses since the character of the sample may be affected by a number of
common conditions including:

      a)  composition of the container 	 glass bottles should
          always be used since plastic containers, with the ex-
          ception of teflon, have been found under certain con-
          ditions to absorb organic materials from the sample.
          In some cases, the reverse is also true in that com-
          pounds have been dissolved from the plastic containers
          into the sample itself.  This problem also applies to
          the bottle cap liners; therefore, the portion of the
          cap that comes in contact with the sample  should be
          made of glass, teflon, or lined with aluminum foil;

      b)  cleanliness of the container 	 previously unused
          glass bottles are preferred.  If this is  impossible,
          bottles should be either acid cleaned or washed with
          a strong detergent and thoroughly rinsed and dried;

      c)  time lapse between sampling and analysis 	 since
          the chemical characteristics of most oils, especially
          the lighter fuel oils, change with time,  the time
          lapse between sampling and analysis  should be kept  to
          a minimum.  If analysis  cannot be completed within  2k
          hours,  samples can be preserved, depending upon  the
          volatility of the oil, by removal of  air  and exclusion

-------
          of light.  With heavier type oils, such as No. 4 and re-
          sidual oils, carbon dioxide may be used to displace the
          air.  If dry ice is available, (approximately 0.5 cu. in.) it
          may be added to the sample.  As soon as the effervescing
          has stopped, the jar should be sealed.  When carbon dioxide
          or another inert gas is not available, or in those instances
          where volatile components are present (No. 2 or lighter), the
          sample can be preserved by carefully filling the bottle to
          the top with water to displace the air.  All samples should
          be kept under refrigeration until analyses are completed;

      d)  collection of adequate volume of sample 	 it is desirable
          to obtain as much of a sample of the oil as possible.  It
          is suggested that 20 mis be considered as the minimum volume
          of oil needed to perform a series of "identification analyses"
          on light oils 	 No. 2 and below.  For heavier oils a minimum
          volume of 50 mis is required.
      As additional data on sampling oil in the environment are forthcoming,
material will be released in the form of addendums to this progress report.
After reading this report, any individual who has knowledge of, or who has
developed a sampling technique, is urged to make this information available
to the Oil & Hazardous Materials Research Section, Edison, New Jersey  08817.

-------
         Preliminary Study on the Use

             of Solid Absorbents
             SECTION-I
Oil and Hazardous Material Research Section
     FWPCA, Edison, New Jersey, 1968

-------
             PRELIMINARY STUDY ON THE USE  OF SOLID ABSORBENTS

                         EDISON,  NEW JERSEY,   1968
I.   Background

     Before oil  portions  obtainable  from an  oil  slick  are  chemically
analyzed  in the  laboratory, the first and most  important requirement
is to secure a representative  sample of the  slick.   It is  therefore
apparent  that the  chemical nature  of the oil  should  not be altered
irretrievably by the  sampling  techniques and  furthermore there should
be no contamination of the sample  introduced  by  components of the
sampling  device.

     Sampling of oil  presents  many difficulties  not  immediately obvious.
An oil slick may vary in  thickness from several  inches down to a mono-
molecular layer measured  in microns  (10-4 cm.).  The quantity of sample
required  is therefore important since such will  determine  the area of
sweep.  For example,  5,000 gallons of oil, if assumed  to be evenly dis-
tributed  over one  square  mile  of water, will equate  to an  oil thickness
of 0.0071 mm.  If  200 ml  of sample is found necessary, then all the oil
must be recovered  from 28 square meters of open water.  If sampling re-
covery is 50 percent  rather than 100 percent, the sweep area must be
doubled.  Table I  below describes theoretical thickness and area, as-
suming an even distribution of oil for various magnitude oil spills.
                                TABLE I
                  Oil Distribution on a Water Surface
Spill Area
Gal/Sq.Mile
1,000,000
100,000
10,000
5,000
1,000
100
Spill Area
ml/Sq. Meter
1430
143
14.30
7.15
1.43
0.143
Area (Sq. Meters)
Req'd. to obtain
200 ml /sample
0.14
1.4
14
28
140
1400
Oil
Thickness
1.43
0.143
0.0143
0.0071
0.00143 = 1
0.000143 =
(mm)




-43XX
0.143/4

-------
      Prior  to initiating studies  at  Edison,  New Jersey,  it was determined
 that the ideal sampler  should have the following characteristics:

      a)   simplicity
      b)   ability  to  function under various conditions
      c)   few moving  parts
      d)   low cost
      e)   rapid recovery
      f)   no secondary treatment required  in  recovering the sample

      Because of these specifications,  mechanical devices using pumps and
 motors were eliminated  from consideration.   Likewise, the use of chemicals
 was  precluded due to possible change of oil  characteristics.  This led to
 emphasis and application of solid absorbent  materials for recovery of oils
 from the water environment.
 II.  Experimental

     A.  Screening of Absorbent Materials

         The following materials were evaluated to determine their effect-
     iveness in removing oil from water.

         a)  Teflon shavings
         b)  Polypropylene fiber
         c)  Rayon manufacture waste material
         d)  Glass-fiber insulation ("Rockwool")
         e)  Styrofoam
         f)  Polyurethane foam

     Whereas first-stage studies showed each of the above materials capable
of recovering oil from water, the polyurethane foam showed greatest promise
for absorbing larger volumes of oil.  Also, the polyurethane foam in sheet
form could be easily handled and did not disperse over the water surface as
did certain of the other absorbent materials.  Thus, further investigations
were undertaken using only this particular "sorbent".

     B.  Testing of Polyurethane Foam

         Polyurethane foam was employed in sheet form in varying thickness
  of 1/8-inch, 1/4-inch, and 1-inch.

-------
      C.   Polyurethane Foam Characteristics
                                        1  ft.   x 1/4"
                                        36 inch
                                        590 cm3
                                        15.9 g
                                        0.027  g/cm
                                        540 ml
                                        17 ml
                                        85  inch   x  7/8"
                                        7k   inch
                                        1210 cm3
                                        22.8 g
                                        0.011 g/cm
                                        900  ml
                                        11 ml
     a)   Foam A*

         Size
         Volume
         Volume
         Dry Weight
         Density
         Saturated water  volume
         Residual Oil  retained
           after squeezing

     b)   Foam B*

         Size
         Volume
         Volume
         Dry weight
         Density
         Saturated water  volume
         Residual oil  retained
           after squeezing

     c)   Foam C*

         Size
         Volume
         Volume
        Dry  weight
         Density
         Saturated water  volume
         Residual oil  retained
           after squeezing

D.  Experimental Results

    Table  II  shows the pertinent data obtained upon use and recovery
of South Louisiana Crude.  During these preliminary studies no attempt
was made to  control temperature, wind, degree of mixing slick area,
water makeup  (tap water used) or other variables that might affect
the inherent  characteristics of the crude oil on a water surface.
                                       44.6  inch  x  1"
                                       44.6  inch
                                       732 cm3
                                       15.5  g     3
                                       0.0212 g/cm
                                       620 ml
                                       7 ml
*Names and sources of Polyurethane Foams are retained in the files of the
 Oil & Hazardous Materials Research Section, Edison, New Jersey.

-------
                                 TABLE II
       Results of Oil  Recovery  Experiment Using  Polyurethane Foam;
     Oil  Available              Recovery  of Oil           Recovery  of Oil
         (mis)                       (mis)                     (%)

     ABC            ABC           ABC

     50    ^O     30           40     30      20          80     75     67
    100    140    145           60     100     105          60     71     72
    240    210    280          145     115     185          60     55     66
    475    555    665          225     305     415          48     55     62
    580   1090    895          130     645     450          22     59     50
   1345   1565  1495          380     600     530          28     38     35
   1365                      410                         30

     Oil  absorbency plots are defined in Figures 1, 2 and 3 respectively
for Foams A, B  and C.  Although data are not included in this report, the
foams demonstrated significantly higher recovery after  one use.  The oil-
laden foams after the initial "run" could absorb a much greater amount of
oil compared to the previously unused or "fresh" foam.

     To detect  possible changes in the character and composition of oil
via pickup, infrared spectra analysis was conducted on the oil before and
after pickup.  No differences were apparent by IR, although further veri-
fication of this reaction would be desired by means of chromatographic
separation techniques.  The oily mixture recovered from the foams after
passing through the wringer and gravity separation showed less than 0.1
percent water.

     E.   Conclusions

         Based upon limited experimentation it  was found that polyurethane
     foam is an effective material for absorbing oil from the surface of
     water.   Future work should be directed towards determining optimum
     conditions for oil  removal.   Additional  study should include,  but  not
     necessarily be limited to:

         a)   determining effect of foam thickness on oil pickup
         b)   determining effect of pore  size  (density of foam)
         c)   determine feasibility of  pre-treatment with oleophilic
             substances
         d)   testing with different grades  of oil
         e)   further field applications

-------
     On the basis of this preliminary study, it is suggested that poly-
urethane foam sheets could be used to sample "thin" oil slicks.  In actual
practice, several square feet of polyurethane foam can be attached with
staples to a wooden dowel or other rigid, floatable device.  The foam-
holding device in turn is attached to a line fixed to a pole.  The sampler
(in fisherman style) would steer or drag the sheet of polyurethane through
the oil slick and after a few minutes of contact time would remove the
foam from the water.  The foam is passed through a wringer, the oil col-
lected and the process repeated until a sufficient quantity of sample is
obtained.

     (Figures 4-6) are photographs taken of a unit field-tested and found
to be successful.  The system although simplified greatly, offers consider-
able promise and plans are underway to improve this sampling approach.

-------
600 T_ IP AM VOLUME^ 590 cm3	
                          Performed 11/22/68

                          Area = 144 inch2

                          1/4" Thickness
    0       400       800      1,200      1,600

              OIL AVAILABLE TO "A" (mis)
 FIGURE 1-Absorbence of South Louisiana Crude Oil
         by Polyurethane Foam A
                                                           2
                                                           at
                                                           UJ
                                                           O
                                                           u
                                 600

                                 500

                                 400

                                 300-

                                 200-

                                 100
                      o
      Performed 11/22/68

Foam Volume = 2,070 cm3
                    9
      Area = 144 inch

      7/8" Thickness
                                             400
  800
1,200
1,600
                                               OIL AVAILABLE TO "B" (mis)

                                 FIGURE 2-Absorbence of South Louisiana Crude Oil
                                          by Polyurethane Foam B
                                600
                                500
                             •=• 400 ^
                             o
                             UJ
                             2 300-
uj  200-
Of
_J
o  100-
                                                          Performed 11/22/68
                                                                         >)
                                                          Area = 44.6 inch

                                                          1" Thickness
                                   0        400       800      1,200     1,600

                                           OIL AVAILABLE TO "C" (mis)

                                FIGURE- 3-Absorbence of South Louisiana Crude Oil
                                         by Polyurethane Foam C

-------
                                  Figure 4:   Strips  of polyurethane,
                                  which can  vary in  thickness  from
                                  1/8 to 1-inch, are permanently
                                  fastened at one end to a wooden
                                  dowel which in turn is connected
                                  via a cord to  a "fishing pole".
Figure 5:  Foam sampler floats on
water surface.  Back and forth
motion of "fishing pole" moves
sampler through oil slick causing
absorption of oil into void areas
of polyurethane foam.
                                   11

-------
Figure 6:  Garage-type wringer is used to "squeeze-out" oil entrapped
within polyurethane sampler.  Oil-water mixture collected in tray is
transferred to separatory funnel and then to glass sample bottle.
                                12

-------
             Oil Sampling Methods

                      By

      Fred K. Kawahara, Ph.D., F.A.I.C.

       U. S. Department of the Interior
Federal Water Pollution Control Administration
           SECTION- II

-------
                          Oil Sampling Methods

                                   By

                   Fred K. Kawahara, Ph.D., F.A.I.C.

                    U. S. Department of the Interior
             Federal Water Pollution Control Administration
                   Division of Water Quality Research
                  Analytical Quality Control Laboratory
                             1014 Broadway
                         Cincinnati, Ohio  45202
     The sampling techniques outlined below were included in the publi-
cation "Laboratory Guide for the Identification of Petroleum Products",
January, 1969.

    "Oily materials may be collected from the surface of the water by
means of three devices.  The first is a glass, wide-mouth filtering funnel,
connected by teflon tubing to a two-way stopcock.  Volatile oil product
found upon the water surface is ladled with the aid of this device; the
lower water phase is discarded by opening the stopcock.  The upper petro-
leum phase is transferred to a large container.  Ladling and water-discard
operation should be repeated until a sufficient amount of oil (10 grams or
more) is collected.

     An alternative collector is a paint-free dustpan with a suitable stop-
cock attached to the handle.  Collection and concentration of several grams
of petroleum product can be achieved with this household device.  Heavy,
viscous material, such as asphalts, can be collected in a similar manner.
Transfer to the final collecting jar from the funnel, jar, or dustpan, is
possible with aid of a clean spoon or stick.

     The third device is a large household mop with a wringer attachment.
It is suggested that, before use, the sponge, whether derived from natural
or synthetic materials, be rinsed thoroughly with  a proper solvent, such as
chloroform.  This mop is passed through the oil pool; the absorbed materials
are squeezed out and transferred by means of a funnel into the collecting
jar.  The sampling operation is facilitated by attaching the bottle, scoop,
or mop to a long pole.  Where possible, reference  samples of oil  should also
be obtained from vessels or shore facilities."
                                     15

-------
 Oil Slick Collecting Device

              BY

Admiralty Materials Laboratory
   British Royal Navy, 1969
  SECTION-  III

-------
                      Oil Slick Collecting Device
                     Admiralty Materials Laboratory
                        British Royal Navy, 1969
     Information has been obtained from sources in the United Kingdom
concerning a highly-promising oil slick sampling device developed by the
Admiralty Materials Laboratory of the Royal Navy.  Following the Torrey
Canyon incident, the Laboratory was assigned responsibility for investi-
gating the distribution of oil over large water areas.  Part of this work
involved the sampling of thin oil slicks.  It has already been reported
that collection of such oil layers is far from simple because in many cases
visible slicks are exceedingly thin and, furthermore, it is difficult to
separate sufficient material for eventual analysis.

     The Laboratory has reportedly designed a device for collecting samples
of oil from thin slicks which currently is under Patent application.  The
device depends upon the principle whereby when water is stirred in a cylin-
drical vessel the surface assumes the shape of a parabola, and oil together
with other free-floating matter, will collect near the center of rotation.
The oil collector consists of an open-ended cylinder with a stirrer inside.
After the device is lowered into the water, the oil collecting at the center
of the cylinder is trapped in a central tube and retained when the cylinder
is withdrawn from the water.  It may be possible to extend this application
of recovering small oil portions to the larger task of scavenging oil slicks
spread over broad areas in harbors, estuaries, and possibly open waters.
Further information is being requested on this sampling technique.
                                    19

-------
         Sampling in Waters Polluted by Hydrocarbons.  Part I

                                  by

F. Edeline, Chem. Engr., A.I.Gx., and R. Heuze, Industrial Chemist,
                         Cebedeau Laboratory
                        SECTION- IV
                    English Translation Made From
           Tribune Cebedeau (Centre Beige Etude Doc. Eaux)
                    16, No. 234,  257-261  (1963)

-------
                             Table of Contents

                                                           Page

  I.   Introduction	    25

 II.   Selected Methods

       A.  Determination of Surface Film	    26

           1.  Camphor Test
           2.  Pail With Tap

       B.  Collection of a Representative Sample  	    26

           1.  Simple Pail
           2.  Sliding Cylinder
           3.  Movable Probe With Pumping

       C.  Measurement Results.  .	    28

           1.  Note on the Statistical Presentation of Results
           2.  Determination of the Surface Film

               a.  Camphor Test
               b.  Addendum Relative to Camphor Test
               c.  Pail With Tap

           3.  Collection of a Representative Sample

               a.  Simple Pail
               b.  Sliding Cylinder (Shell)
               c.  Movable Probe With Pumping

III.  Conclusions   	    34
                                    23

-------
              Sampling  in Waters  Polluted  by  Hydrocarbons.  Part I

 I.  Introduction

      The c[uantitative determination of hydrocarbons in waters is a
 difficult problem because the aim is to determine a whole range of
 products at the same time.

  ^. ^ Most of the time, current methods make use of hydrocarbon solu-
 bility in ether or carbon tetrachloride, which makes it necessary to
 determine hydrocarbons and lipids simultaneously as well as all the
 other compounds that may be present in the water and are soluble in
 ether.

      The problem of collecting samples is even harder to solve,
 especially if we consider the variety of forms that hydrocarbons can
 assume in water.  First of all, there are hydrocarbons present as a
 distinct phase, in the form of a surface film.  These surface films
 are obtained by spreading one liquid on another with a higher surface
 tension.  This spreading can proceed all the way to a monomolecular film.
 In the case of hydrocarbons (surface tension of the order of 30-35 dynes/
 cm),  the forces of cohesion oppose unlimited spreading, and we have so-
 called "coherent" films.

      This film often contains suspended lipophilic matter, which stabilizes
 the  film depriving it of some of .its mobility.  Hydrocarbon films present
 certain problems:   danger of ignition,  failure to show, obstruction of
 apparatus,  interference with gas exchange between water and atmosphere,
 etc.   It is therefore of  interest to determine this category of hydrocar-
 bons  separately.   We also find hydrocarbons and greases within the body of
 the water,  not  in  the dissolved state,  to be sure,  but in the form of
 emulsions,  sometimes very fine and very stable, or  incorporated in small
 aggregates  of  suspended matter.   In this form, they are especially trouble-
 some  for  waste  water purification plants and for plants treating  a water
 supply,  either  human or industrial.

      It  is  obvious  that hydrocarbon distribution in water is  anything but
 homogeneous, especially if  the aggregates  in which  they are incorporated
 are relatively  large.   It  should  be  noted  that the  two main forms  in which
 hydrocarbons are found  are  interconvertible to a certain extent.   The ob-
 servation of waste waters from rolling  mills,  which are highly polluted by
 oils, has shown that  aggregates  loaded  with drops of oil can  be maintained
 in suspension by agitation,  but as  soon as a  decrease  in turbulence occurs,
 such aggregates will  come to  the  surface,  wherein the  oil  immediately spreads
 in an iridescent film.

     We will not speak  of the  difficulties involved  in  recovering  the last
 traces of greases or  hydrocarbons  on the bottles used  for  sampling;  we  will
 assume this problem  has been  satisfactorily solved.  We  shall  content our-
 selves in this first  paper with comparing  various sampling methods  reported
 by specialized laboratories*.
*A second paper deals with specific methods proposed by Gebedeau.
                                    25

-------
     When we questioned the principal western laboratories studying
waters, we received as many different responses as there were labora-
tories consulted.  Many of these answers gave evidence that the problem
of sampling hydrocarbons had not been recognized as such.  Many others
could be considered variants of the same basic operating procedure.  For
each method described below, ten tests were made, either in the laboratory
or under "on-the-spot" operating conditions.  These series of quantitative
determinations allow us at best to evaluate the reproducibility of the
measurements for each of the methods, none of which seem worthy at this
time of being considered a reference method.  Until a fuller inquiry, we
consider the best method to be that which consistently gives the highest
results.  For the methods applicable to surface films and involving only
a small sampling surface, sampling was done from films made artificially
in the laboratory 	 which will give an idea of the absolute value of
the sampling carried out by these methods.
II.  Selected Methods

     A.  Determination of the Surface Film

         1.  Camphor Test

             (Origin:  Deutsche Einheitsverfahren H 17/18)

             This test consists in observing particles of camphor placed
         on the surface of the water.  If these particles are in a state
         of incessant agitation, there is no greasy film on the water.
         If, on the other hand, they remain motionless, there is a film.
         This test is a qualitative test.  This phenomenon, known as the
         "camphor dance", corresponds to the formation of a surface film
         from the solid product.

         2.  Pail With Tap

             (Origin:  RIZA, Rijksinstituut voor Zuivering van Afvalwater)

             The sampling is done in a pail equipped with a tap at the
         bottom; wait five minutes after sampling, then draw off and
         collect the last liter with the greases and hydrocarbons of
         the surface film.  Rinse the pail with petroleum ether.  Refer
         the measurement to the cross section of the pail.  If necessary,
         repeat several times to collect sufficient material.

     B.  Collection of a Representative Sample

         1.  Simple Pail
                                              1
             (Origin:  EAWA6 - Eidg. Anstalt fur Wasserversorgung,
              Abwasserreinigung und Gewasserschutz)

             Collect about 5 liters from the surface layer using an
         ordinary pail.
                                      26

-------
2.  Sliding Cylinder

    (Origin:   Shell and American Petroleum Institute.  Method
     702-53)

    This ingenious apparatus permits rapid collection of a
2-liter representative sample of water, including the surface
film, to a depth  of about  50 cm.  The apparatus has (See Figure 1)
a sliding Plexiglas cylinder whose fall is controlled by a trigger.
In falling vertically, the cylinder cuts a cylinder in the water
similar to a core sample.   At the bottom of its travel, it meets
a stopper, which  hermetically seals it and permits withdrawing it
from the water.   In principle,  the apparatus is not suitable for
depths greater than 0.5 m.   We  found a less interesting version of
this method in The Analyst 81,  492 (1956), recommended by the Joint
ABCM-SAC Committee on Methods for the Analysis of Trade Effluents.
                                    trigger
                              handle
                        supporting rod
              fixed guide cylinder
          sliding Plexiglas cylinder
             (diameter 6.5 cm)

              fixed rubber stopper
                                               trigger
                                                bottom plate
                            FIGURE 1-Sliding cylinder.
                             27

-------
    3.  Movable Probe With Pumping
                                        n     H
       (Origin:  EAWAG - Bundesanstalt fur Gewasserkunde)

        The apparatus used consists of a lonj plastic tube 1/2
    inch in diameter, weighted at the end and provided with
    colored reference marks every 25 cm.  The tube runs in a
    peristaltic pump operated by a 6-v storage battery.  The
    probe is lowered to the bottom of the water to be sampled,
    and the probe is raised by a steady movement.  The cor-
    rections needed to take into account the volume of water
    contained in the. tube are made.

C.  Measurement Results

    1.  Note on the Statistical Presentation of Results

        Our tables show the n measurements of x made in each
    series.  For these measurements, the various statistical
    parameters are calculated below:

    Mean;   the arithmetic mean of all the measurements (x~).

    Standard deviation:  the square root of the variance where
                      a  _   / Z(x-x}z
                                  n- i
        There are about  2 chances out of 3 that a result picked
    at random will have  a deviation ^_ a  with respect to the
    mean .                            ~

    Standard deviation in %:    the standard deviation expressed
    in % of the mean 3C.

    Mean deviation;  calculated by the formula:
                                  q
        There are about 2 chances out of 3  that the exact value
    sought,  for which n measurements are made,  will fall  in the
    interval x + G m.  We give a m for 10 measurements.

    95% deviation;   equal to 1.96 Q .  There are 95 chances out
    of  100 that an isolated measurement will not differ from the
    value  sought by a deviation greater than this number.
                               28

-------
2.  Determination of the Surface Film

    a.  Camphor Test

        The tests were made in the laboratory on hydrocarbon films
    of known dimensions.  For this, we placed on a container of
    known surface area increasing amounts, measured by microburette,
    of used SAE 20 crankcase oil, filtered beforehand.  The films
    are characterized by the number of milligrams of hydrocarbon
    present per square meter of surface.  The same process was re-
    peated for all the laboratory tests.  The test is said to be:

    positive (+) when the camphor grains remain perfectly
                 motionless;

    doubtful (i) when they exhibit alternating motion and lack
                 of motion;

    negative (-) when they remain incessantly in motion.

        Table I below compiles the results obtained and shows that
    this test, for the hydrocarbon employed, actually permits de-
    termining that a film belongs to one or another of the three
    categories delimited by the following values:

                 from    0 to 175 mg/m^
                 from  175 to 325 mg/m
                 above 325 mg/m
                          TABLE  I

                                       Result
                            29

-------
             However,  when we  repeated  this  test with a more  fluid
         hydrocarbon such as petroleum  residuum, we  observed  a new
         phenomenon, which falsified  the results.  The camphor grains
         are  denser than the hydrocarbons, but  less  dense  than water,
         and  they  normally position themselves  at  the interface be-
         tween the two liquids.  At this point, however, they tend
         to form a surface film, too.   The camphor has no  difficulty
         in driving back a film of petroleum residuum, even a thick
         one.   This is why we  observe the camphor  dance on light and
         fluid hydrocarbon films.  In the case  of  a  heavier oil, this
         phenomenon does not occur, and the  method can be  used.

         b.   Addendum  Relative to the Camphor Test

             We believe it is  of interest to complete this analysis
         of the camphor test with Table II drawn from the  literature,
         where appearance of various  iridescent hydrocarbon films is
         described.  Since the amounts  of hydrocarbons were given in
         liters per  km ,  we converted them to mg/m ,  assuming a den-
         sity  of 0.92.
                               TABLE II
      Visual Description of Oil Slick Correlated With Thickness
            (Taken from Gas and Wasserfach 82, 182.  1939)

      Hydrocarbons	
                   Thickness
          mg/m        mm         	Appearance 	

            33.5    0.0365      Scarcely visible under the best conditions
            67      0.073       Visible - silver luster
           134      0.146       First traces  of color
           268      0.292       Visible bright-colored bands
           897      0.975       Duller colors
          1790      1.950       Dark colors
>40       >37     > 0.040       Continuous iridescent film
        c.  Pail With Tap

            Tables III and IV present the two series of results ob-
        tained at two points in the water lines of a plant (rolling
        mill).  The results are very poor, even in Series IV where
        we benefit from a surface of calm water.
                                  30

-------
                               TABLE III

      Reproducibility Results Using Pail With Tap Method, Point A
                         2
Measurement No.      mg/m

      1               205
      2               428
      3               578
      4              1120
      5               816
      6               133
      7               181
      8               398
      9              1590
     10               260
Mean:  601 mg/m
Standard deviation:  +  520 mg/m
Standard deviation, %:  ±  86.5%
Mean deviation of 10 measurements:
95% deviation: ±1019 mg/m
                       .
                    mg/m
                                TABLE IV

      Reproducibility Results Using Pail With Tap Method, Point B
                         2
Measurement No.      mg/m

      1              1040
      2               137
      3               193
      4               177
      5              8870
      6              1210
      7               319
      8               758
      9               870
     10               437
Mean:  1400 mg/m
Standard deviation:
Standard deviation,
Mean deviation of 10 measurements:
95% deviation:  - 16,425 mg/m2
 - 8380 mg/m"
&:  -  598%
                -  2650 mg/m2
     3.  Collection of a Representative Sample

         a.  Simple Pail

             The results obtained with this method, which has the
         advantage of extreme simplicity, are not very constant,  as
         may be judged from Table V.
                                     31

-------
                                TABLE V

            Reproducibility Results Using Simple Pail Method

                         2
Measurement No.      mg/m

      1               15.7
      2               15.0
      3               11.9      Mean:  12.35 rag/liter
      4                6.4      Standard deviation:  - 6.30 mg/liter
      5                3.5      Standard deviation, %:  ~ 51.0%
      6               15.6      Mean deviation of 10 measurements: - 2.0 nag/liter
      7               26.0      95% deviation:  - 12.35 mg/liter
      8                8.0
      9               12.0
     10                9.4
         b.  Sliding Cylinder (Shell)

             This method is the most convenient and reproducible.  It
         does not seem possible to modify it to permit sampling at depths
         greater than 50 cm.  Still, it is very likely that series of
         tests would show that the water layer between 0 and 50 cm depth
         is the most important from the viewpoint of the hydrocarbon con-
         tent.  These tests have not been made.  Table VI gives the measure-
         ment results.
                                TABLE VT

         Reproducibility Results Using Sliding Cylinder Method

                         2
Measurement No.      mg/m

      1                5.44
      2                3.64
      3                3.57     Mean:  4.22 mg/liter
      4                4.68     Standard deviation:  - 0.734 mg/liter
      5                3.91     Standard deviation, %:  17.4%
      6                3.64     Mean deviation of 10 measurements: - 0.232 mg/liter
      7                5.07     95% deviation:  - 1.44 mg/liter
      8                4.85
      9                3.31
     10                4.13
                                      32

-------
         c.  Movable Probe With Pumping

             Contrary to what we thought, this method proved highly
         unsatisfactory.  We think one of the reasons for this lies
         in the use of a very long tube for sampling (10 m), which
         calls for pumping for some time before collecting the sample,
         if we wish to be certain that the sample comes from the de-
         sired depth.  But in the case of abundant suspended matter,
         it may well not be moved along the tube at the same velocity
         as the liquid.  Furthermore, the weighted tube touches the
         bottom of the water before being drawn up; consequently, it
         produces an eddy that may resuspend the deposits.  Finally,
         the suction of the pump itself contributes to the formation
         of turbulence.  Perhaps it is possible to remedy these draw-
         backs, at least in part.  We have not tried to do so, for
         this method also has the disadvantage of calling for a heavy
         piece of equipment that is hard to handle.  Table VII shows
         the results of our measurements.
                               TABLE VII
           Reproducibility Results Using Movable Probe Method

                         2
Measurement No.      mg/m

      1              2.30
      2              0.90
      3              5.30        Mean:  4.06 rag/liter
      4              6.10        Standard deviation:  - 2.65 mg/liter
      5              2.10        Standard deviation, %:  - 65.3%
      6              7.60        Mean deviation of 10 measurements: - 0.84 mg/liter
      7              1.90        95% deviation:  5.19 mg/liter
      8              1.00
      9              6.80
     10              6.60
                                   33

-------
III.  Conclusions

     The first conclusion we can draw from this comparative study (a
conclusion that an analytical chemist cannot fail to draw) is that all
the methods in the collection are deplorably poor.  The best would not
satisfy even the least exacting person.  It is, however, a fact with
which we shall have to reckon and for which we should be forewarned by
multiplying the number of samples taken.

     For the purpose of tentatively classifying the different methods
tested, we have drawn up Table VIII which contains the standard deviations
in per cent and also gives the number of samples that must be taken in
order to be certain (threshold of 0.95) that the mean of the measurements
on the sample will not deviate more than 25% (or 50%) from the actual
value.  This table contains only the most favorable series for the methods,
which have been the object of several series of measurements.
                               TABLE VIII
               Statistical Comparison of Various Methods
                                        Standard      No. of measurements
                                      deviation, %    necessary  (see text)

                                                         50%       25%

Determination  of  the  surface film
   Pail with tap	         86.5           12        47

Collection  of  a representative  sample
   Sliding  cylinder	         17.4            1         2
   Simple Pail	         51.0            4        16
   Movable  probe  with pumping  .  .         65.3            7        26
     According  to the table,  it  does  not  seem at  all  possible  to  retain  the
method  of  the "pail  with tap" for  the determination of  the surface  film.
This is why Cebedeau is  particularly  eager  to find another method.   This
research will be the subject  of  a  forthcoming paper*.   The semiquantitative
camphor test is to be retained only for heavy hydrocarbon films.
 *This  paper is given immediately following in this compilation of reports
  on sampling.
                                     34

-------
         Sampling in-Waters Polluted by Hydrocarbons.   Part II

                                   by

F. Edeline, Chem. Engr., A.I.Gx., and R. Heuze, Industrial Chemist,
                           Cebedeau Laboratory
                         SECTION-V
                      English Translation Made From
             Tribune Cebedeau (Centre Beige Etude Doc.  Eaux)
                        18, No. 255, 75-79 (1965)

-------
                           Table of Contents

                                                       Page

  I.  Introduction	      39

 II.  Methods Designed at Cebedeau 	      39

      A.  Determination of the Surface Film

          1.  Filter Paper Disks
          2.  Movable Frame
          3.  Metal Cylinder Lined With Filter Paper

III.  Measurement Results	      41

      A.  Note on the Statistical Presentation

      B.  Determination of the Surface Film

          1.  Filter Paper Disks
          2.  Moveable Frame
          3.  Metal Cylinder Lined With Filter Paper and
              Repulsion by a Surface-Active Solution

 IV.  Conclusions	      47
                                   37

-------
           Sampling  in Waters  Polluted  by  Hydrocarbons.  Part II
I.  Introduction

     Several months ago, when making a study of stream pollution by hydro-
carbons, Cebedeau had occasion to work on the problem of collecting samples,
certainly a practical problem, and of relevance to the interests of this
meeting.  (Topic No. 23, "Practical Measures to be Taken for the Prevention
of Pollution by Mineral Oils in Ports".)

     Questions asked of the principal western laboratories brought us as
many different answers as there were laboratories consulted.  We have al-
ready reported the results of a special study of these methods (1).  Be-
cause these methods do not satisfy us we have designed still other methods
in the Cebedeau Laboratories to be described here with a brief critique of
their application.

II.  Methods Designed at Cebedeau

     A.  Determination of the Surface Film

         1.  Filter Paper Disks

             Sampling is accomplished by throwing flat on the water a
         loose-textured filter paper disk of the usual dimensions avail-
         able on the market (diameter 9 or  11 cm).   Such a filter can be
         wetted either by water or by hydrocarbons,  but once wetted by
         one, it no longer retains the other liquid.  It is therefore
         important to place this paper flat on  the water (this confines
         the use of the method to calm waters)  and to remove it with
         caution with the aid of tweezers,  also limiting the running of
         water over the filter, which would cause an error by loss of
         materials.

         2.  Movable Frame

             For sampling, we use a  square  wooden  frame exactly  1 m on
         a  side.  The fourth  side of this frame is movable  (See  Figure  1)
         so that after  aim2 film has been intercepted on  the  surface  of
         the water, we  can bring this  surface  to 1 m x  0.2 m, and finally
         collect the film  in  a wide-necked  bottle  opening just  below the
         surface.

         3. Metal  Cylinder Lined with Filter  Paper  and Repulsion by a
             Surface-Active Solution

             Use  is  made  of a copper cylinder  7 cm tall, with a circum-
         ference  slightly smaller  than the  side of a sheet  of chromato-
         graphic  paper  (i.e.,  a  diameter  of about  17.5  cm)  (See Figure  2).
         This  cylinder  is  provided with  a handle and a  5 mm side hole.   It
         is lined with  a  strip of  chromatographic  paper (loose-textured
         paper) which  is  carefully applied  to  the  metal and  attached by
         small  clamps.
                                       39

-------
       stop
                            1-m sliding side
       0.2-m sliding side
               handle
                                                                                copper cylinder
 strip of
                                                                               side hole
                                                            1 ' r* u '            / _| •
                                                    chromatographic paper    diame»«r 0.5 mm)
             FIGURE 1-Moveable frame
                                                                FIGURE 2-Filter  aspiration device
filter-support tube
  tared receiver tube
                              bend
                               filter strip or paper
                               carrying the hydrocarbons
                                                                    ether-
                                                                                  ground stopper
tared receiver tube
                                                                                    strip rolled into a tube and
                                                                                    carrying the hydrocarbons
                                                                                     plastic rings
           FIGURE 3-Semimicroextractor
                                                                FIGURE 4-Chromatographic tube

-------
             Before  sampling,  the  strip  is completely impregnated with
         distilled water.  Excess water  is applied and runs over the
         paper.  From this point on, the clamps become useless and may
         be removed.

             In sampling, the  apparatus  is plunged into the water and
         immersed to within a  centimeter of the upper edge of the strip
         of paper.  At this time,  one or two drops of a concentrated
         solution of commercial detergent are rapidly applied by means
         of a dropper to the surface of  the water at the center of the
         intercepted circle.   This produces a surface-active film that
         tends to drive the hydrocarbon  film toward the periphery of the
         apparatus (which is easily discernible).

             When the hydrocarbon  film is confined to the peripheral ring
         1 or 2 cm wide, the apparatus is taken out of the water by a
         steady and very slow  movement, which has the effect of "aspirating"
         the hydrocarbon film  onto the strip of filter paper.  The apparatus
         is then set vertically, with the side opening on the bottom, and
         so as to avoid any running of water outside the strip.  The water
         is absorbed by a wad  of cotton  placed outside the hole.  The strip
         may then be detached, dried, and hydrocarbons removed from the
         paper.  To recover the hydrocarbons, we can extract them hot with
         petroleum ether in a  semi-microextractor (See Fig. 3) of the
         Quickfit type, or elute them cold in a chromatographic tube
         (See Fig. 4).

III.  Measurement Results

     A.  Note on the Statistical Presentation

         Our tables present n measurements of x made in each series.  For
     these measurements, we calculated the various statistical parameters
     below:

     Mean:   the arithmetic mean of all the measurements (x);

     Standard deviation;  the square root of the variance where
                                   n — i
         There are about 2 chances out of 3 that one result taken at
     random will have a deviation  5./a/   with respect to the mean.

     Standard deviation in %:  the standard deviation expressed in %
     of the mean x".

     Mean deviation;  calculated by the formula:
                           am
                                  /IT

-------
         There  are about  2 chances  out of  3 that the exact value
     that we want to determine and  for which we make n measure
     ments will  fall in the  interval x - a m.  We give a m for 10
     measurements.

         95% deviation:   equal to 1.96 a .  There are 95 chances
     out of 100  that an isolated measurement will not differ from
     the value  sought by  a deviation greater than this number.

     B.  Determination of the Surface Film

         1.  Filter Paper Disks
                                                       «
             For these tests we used Schleicher and Schull SS 5893
         filters 12.5 cm. in diameter with a surface area of 123 cm2
         for the tests in the laboratory, and filters 11 cm. in
         diameter with a  surface area of 95 cm2 for the industrial
         plant tests.  The results of the laboratory tests are given
         in Table I.  The tests performed  in the plant were made on
         a settling basin for rolling-mill waste water, at a point
         chosen for its mild surface agitation.  These results are
         given in Table II.
                                TABLE I

Reproducibility Results Using Filter Paper Disk Method, Laboratory Tests

        Measurement,
No.       mg/m	

 1          642
 2          527
 3          263         Mean:  562 mg/m2
 4          663         Standard deviation:  - 138 mg/m
 5          505         Standard deviation, %:  - 24.6%
 6          537         Mean deviation for 10 measurements:  - 43.7 mg/m2
 1          453         95% deviation:  - 270 mg/liter
 8          758
 9          684
10          589
                                  42

-------
                                             TABLE II
              Recovery of Oils From Industrial Plant Using Filter Paper Disk Method
No_.

 1
 2
 3
 4
 5
 6
 7
 8
 9
  mg/m
Introduced

   2190
   1815
   1580
   1329
   1117
    9914
    872
    763
    671
  mg/m
Recovered

  2570
  1700
  1650
   841
   837
   748
   643
   403
Deviation,

  + 17.4
  -  6.4
  +  4.4
  + 11.7
  - 24.7
  - 15.8
  - 14.2
  - 15.7
  - 39.9
Mean:  90.8%
Standard deviation: + 19.15%
Standard deviation,%: i 21.1%
Mean deviation for 10 measurements: + 6.06%
95% deviation: + 37.53%
                                                 43

-------
         In Table II we have made a statistical analysis of the
     percentages.  We see that in the range of films tested (from
     650 to 2200 rag/in  we recover on the average less hydrocarbon
     than introduced:  90.8% vs 100%.  This suggests a systematic
     error of the order of -10%, but statistical theory does not
     allow us to call this deviation significant.  The precision
     or reproducibility is somewhat better in the laboratory,
     which is entirely normal.

         The difficulty of this method is obviously in the removal
     of the filter after impregnation.  In spite of the use of
     tweezers and the speed, a certain amount of water running
     over the filter cannot be avoided, causing material loss.

     2.  Movable Frame

         Test results using the Movable Frame Method shown in
     Table III were quite unsatisfactory, due especially to the
     difficulty of recovering the hydrocarbons adhering to the
     walls of the frame.
                               TABLE III
           Reproducibility Results Using Movable Frame Method
         Measurements,
No.          mg/m2	

 1           259
 2           487
 3            26.3      Mean:  172 mg/m2
 4            22.3      Standard deviation:  + 238 mg/m
 5            50.3      Standard deviation, %T  + 138%
 6           390        Mean deviation for 10 measurements: ±75.2
 7           616        95% deviation:  466 mg/m2
 8            20.1*
 9            28.4

     3.  Metal Cylinder Lined with Filter Paper and Repulsion by a
         Surface-Active Solution

         Two series of measurements were made at different points.*
     The first series given in Table IV was carried out by a more ex
     perienced operator than the second series given in Table V.  It
     is to be hoped that the scatter of the results will be reduced
     by repeated practice with the method.
*Further description of the nature of samples and type of oils was not
provided in the translation of the report.
                                   44

-------
                                TABLE IV
    Reproducibility Results Using Filter Aspiration Device,  Point  A


           Measurements,
No.            mg/m2

 1             236
 2             134          Mean:  249 mg/m                 2
 3             220          Standard deviation:  + 69.4 mg/m
 4             285          Standard deviation, %:  ± 27.9%
 5             315          Mean deviation for 10 measurements: ±22.0 mg/m
 6             338          95% deviation:  ± 136 mg/m2
 7             216
                                TABLE V
    Reproducibility Results Using Filter Aspiration Device, Point B


           Measurements,
No.            mg/m	

 1             314
 2             212
 3             342          Mean:   422 mg/m
 4             358          Standard deviation:   +205 mg/m
 5             322          Standard deviation,  %:  ± 41.8%
 6             298          Mean deviation  for  10 measurements:  ±64.9 mg/m
 7             318          95% deviation:   ± 402 mg/m2
 8             607
 9             897
 10             554


          The hydrocarbons can be  recovered  by cold  elution or  by
      hot extraction.   Cold elution gives unstable results, for its
      duration varies  greatly  from sample to sample.   Table VI  shows
      unsatisfactory results obtained from two different  series of
      measurements.   Here we are dealing with an intermediate stage
      in the treatment of the  sample, and we can only  assume  that
      it introduces  single or  double errors.
                                      45

-------
    No.
                                 TABLE VI
                          Cold  Elution Recovery


                                                  Measurements,
                                                                     B
 Theoretical                                 53.8
      1                                      44.8                    13.6
      2                                      49.3                    11.2
      3                                      50.7                     8.1
      *                                      50.2                    13.8
      5                                      55.0                    10.2
      6                                      51.6                     6.3
      7                                      44.9                     7.6
      8                                        -                      8.5
      9                                        -                      8.2
         In recovery by hot extraction, we determined the minimum
     duration that would assure complete extraction of a sample and
     concluded that a two-hour extraction was sufficient.

     Sample deposited on the strip              38.5 mg
     Recovered after 1/2 hour                   27
     Recovered after 1 hour                     31.6 mg
     Recovered after 1-1/2 hours                35.25 mg
     Recovered after 2 hours                    35.25 mg

         It remained to check the sampling methods in the hot extraction
     technique on films of known density.  The results of 5 measurements
     are presented below (Table VII).  Compared to the other methods,
     these results are highly satisfactory.  We believe that the varia-
     bility of the result should be attributed not to the sampling itself
     (which is virtually quantitative) but to the difficulty of preparing
     exactly similar films.
                               TABLE VII
                        Hot Extraction Recovery
         Measurements,
No.      	mg	
 1          25.90      Theoretical amount to be recovered:  26.15 mg
 2          21.05      Average amount recovered:  24.35 mg
 3          26.00      Standard deviation: - 2.06 mg
 4          23.70      Standard deviation, %: ± 8.46%
 5          25.10      Mean deviation for 10 measurements: + 0.65 mg
                       95% deviation:  + 4.04 mg
                                   46

-------
 IV.   Conclusions

      For  the purpose of classifying the different methods  tested, we
 compiled  Table  VIII,which lists  the standard deviations in per  cent
 and  also  gives  the  number of  samples that must be taken in order to be
 certain  (threshold  of 0.95) that the mean of the measurements on the
 sample does  not deviate more  than 25% (or 50%) from the actual  value.
 This table contains only the  most favorable series for  the methods,
 which were the  object of several series of measurements.   It was con-
 cluded that  the first two methods could be retained.
                                TABLE  VIII
                      Sampling Methods 	  Summary

                                      Standard       No.  of Measurements Needed
Determination of the  Surface Film   Deviation, %        25%         50%

Metal cylinder lined  with  filter
  Paper	         8.46              1            i
Filter paper disks	        24.6               1            4
Wooden frame	       138                30          H8
                             Literature Cited
(1) Edeline, F., and Heuze, R., "Sampling in Waters Polluted by Hydrocarbons",
Tribune Cebedeau (Centre Beige Etude Doc. Eaux) 16, No. 234, 257-261 (1963).

-------
      Collection of Slick-Forming Materials from the Sea Surface

                                  by

                            W. D. Garrett

                  U. S. Naval Research Laboratory
                         Washington, D. C.
                        SECTION- VI
The following materials are taken from two separate reports prepared by
Dr  W  D. Garrett on the same subject.  The first paper was in the form
of a manuscript submitted February 14, 1962 as Project SF 001-06-01,
Task 1500 in response to NRL Problem 003-17.  The second paper appeared
in Report of NRL Progress, March 1965 issue.

-------
                          Table of Contents
  I.   Introduction 	



 II.   Preliminary Survey of Techniques  	     53



III.   Screen Collection Method  	     55



 IV.   Field Collection Equipment  	     57



  V.   Verification of Screen Collection Method  	     57


                                                                   59
 VI.   Summary 	



      Addendum on Isolation and Analysis of  Collected              6

      Organic Material  	



      References  	





  Figure^



  1.  Sea-Surface Collection Assembly in Draining Position  . .      58



  2.  Top View of Tank  Experiment to Study Monolayer Retrieval      58

      Method    	
                                     51

-------
        Collection  of  Slick-Forming Materials  from the Sea  Surface
 I.   Introduction

     Aggregations  of  surface-active  materials  on large  natural bodies
 of water  such as rivers,  bays,  and the  oceans  are often responsible  for
 visible alterations in  the  appearance of  the surface.   These  surface-
 active substances  can originate from man-made  sources,  but they may  also
 arise from products of  the  marine biosphere.   Whatever  the source, these
 polar-nonpolar substances accumulate at the surface  and are compressed by
 the  convergent action of wind and water into areas in which wind-induced
 capillary waves are damped.     The calmed area or "sea  slick" stands out
 in contrast to the background water  body  since skylight reflectance  from
 the  slick is altered  by its action in damping  the capillary waves.

     The physical  properties of slicks  and their responses to environmental
 changes have been  studied by several investigators.  Adam (1) developed a
 spreading drop method for the measurement of ocean surface film pressures
 in situ.  Lumby and Folkard (2) applied such spreading  oil measurements to
 slicks in order to study the effect  of  air movement  and currents on  the
 film pressure of the  slick.  Some of the  physical  properties of sea  slicks
have been studied by  Ewing  (3)  while Dietz and LaFond (4) noted the pre-
dominance of calmed areas of the sea in the vicinity of water possessing
high biological activity.

     The objective of this research was to develop a technique to collect
and recover the constituents of  natural oceanic  slicks  so that the chemi-
cal composition of the  responsible film-forming  materials could be es-
tablished and be used to develop a physical and  chemical picture of the
ocean surface.

 II.  Preliminary Survey of Techniques

     Laboratory research was conducted  toward  the development of an ef-
fective field method  for the collection of the  surface-active chemicals
responsible for sea slicks.  Film-pressure measurements have indicated
 that areas of the ocean in which capillary waves  are damped often consist
of only slightly compressed monomolecular layers  of  surface-active sub-
 stances.  Film pressures as low as 2 to 4 dynes/cm, were found in clearly
visible slicks, and values rarely exceeded 10  dynes/cm.  To obtain a suf-
ficient quantity of the surface film material  for identification and study
 in the laboratory, it was necessary  to  collect  the film from a large area
 of the water surface.   In addition,  it  was desirable for the method chosen
 to be as independent  as possible of  film  pressure so that large areas of
 film could be removed efficiently even  when the  surface pressure was low.
                                     53

-------
     A chemical characterization of the oily material extracted by
Timmons and Jarvis (5) from plankton taken from the near-surface water
of the Bay of Maine showed that they consisted of considerable propor-
tions of Gi6) Cl8» C20> and C22 fatty acids and their triglycerides.
Since plankton organisms are ubiquitous in the world's oceans, it was
assumed that chemicals of this type would in part constitute many
natural slicks.

     For these reasons, methods for the removal of monolayers of oleic
acid, stearic acid, and plankton oil from the surface of both distilled
and synthetic sea water were examined.  The plankton oil used in this
work was extracted from a near-surface plankton collection made in May
of 1960 in the North Atlantic Ocean (40°N, 71°W).  The plankton were
primarily Calanus.  The oil extracted from them has been kept under re-
frigeration and the method of extraction was the same as that used by
Timmons and Jarvis (5).

     The monolayers were spread at various degrees of film compression
on water contained in a Pyrex glass hydrophil tray filled with a Cenco
float-type torsion head.  A monolayer was spread and a waxed glass
barrier was moved toward the floating torsion head barrier compressing
the film to the desired film pressure.  Film retrieval was attempted
and a barrier was again advanced in the direction of the torsion head
to restore the original pressure.  The resulting decrease in area of
the monolayer at constant pressure was a measure of the quantity of
material which had been removed.

     Preliminary studies with mineral oil as a floating absorbent for
the slick-forming materials proved unsuccessful because no significant
amount of organic monolayer was found to dissolve into the oil.  Further-
more, an approach of this nature was not considered promising since any
water-insoluble material which would dissolve a surface-active chemical
from a water surface would have its own spreading pressure increased.
It would no longer exist as a lens but would spread out over  the water
and make recovery difficult.

     A series of solid materials in sheet form were contacted horizontally
with various monolayers.  The solid sheets were held firmly  in a metal
frame which was immersed parallel  to the water surface on which  the mono-
layer was spread.  The candidate adsorbers  included clean glass, aluminum
foil, nylon, polyethylene, Plexiglas, polyvinylfluoride, glass paper  (one
micron fiber diameter), Teflon, and bondable Teflon  (one side treated with
sodium in liquid ammonia). The Teflon and polyvinylfluoride  polymers ex-
hibited no capacity for adsorption of the monolayers.  The glass, nylon,
Plexiglas, aluminum foil, and glass paper removed an area of  monolayer
equal to the area of the solid in  contact with the liquid surface.

-------
     Two adsorbers, polyethylene and activated polytetrafluorothylene
(treated with sodium  in liquid ammonia) appeared most promising.  How-
ever, this technique  was also discarded because of the inability to
desorb the collected  organic material by solvent washing.  Other dis-
advantages inherent in this method were limited adsorptive capacity,
possibility of desorption onto areas of water containing no surface
film, and selective adsorption.

III.  Screen Collection Method

     Experiments with 16 mesh (16 wires per inch) window screen indi-
cated that segments of the surface of a liquid were removed intact
when a screen was placed either in horizontal contact with or drawn
vertically through the liquid surface.  In this manner many small
discrete areas of the surface film between the wires could be trapped
and lifted from the bulk water.  Any materials contained in the en-
trapped surface film  were also carried along.  When the screen was
drawn through the liquid with its plane parallel to the surface, com-
plete independence from film pressure variations was achieved.  Although
some adsorption occurred on the screen wires, the main mechanism for
film removal was the  entrapment between the wires.

     Film-balance experiments showed that a monolayer could be transferred
from a water surface  by draining the screen entrapped liquid into a col-
lection vessel.  The  results of the effect of mesh size on this procedure
are summarized in Table I.  As the number of wires per unit area increased,
the operation of the  screen approached that of a solid adsorber.  In these
measurements, metal screens of various mesh sizes were contacted with a
synthetic sea water surface covered  with an oleic acid monolayer at a
film pressure of 5 dynes/centimeter.  The screen was removed and drained
into an open vessel and the operation was repeated.  After each contact
the film balance barrier was advanced to restore the original 5 dynes/
centimeter surface pressure.  The area swept by the barrier was then equal
to that removed by the screen.  In the case of the 16-mesh screen, the
apparent screen efficiency dropped to a value of approximately 75 percent
after the first surface contact and remained constant thereafter.  In the
case of the tighter screens, the efficiency decreased to a lower value
because a larger wire area was available for adsorption while less open
space existed for water entrapment.  It is also seen that the first contact
with the surface removed a greater area of monolayer since both adsorption
and water-entrapment  mechanisms were operating.  After the first contact,
little further adsorption occurred.  When the distance between the wires
became large, as in the case of the 4-mesh screen, the trapped-water film
became .unstable and ruptured almost immediately before transfer could be
effected.
                                    55

-------
                                                    TABLE I
                          Removal  oŁ Monolayers from Water Surfaces by Metal  Screens*
                                  Monolayer-Oleic acid on synthetic sea water
                                          Film pressure- 5 dynes/cm
 Screen
 (Mesh)
Surface Contact
    Number
Area of Monolayer
     Removed
      (cm2)
Area of
 Screen
   (cm2)
  Screen**
Efficiency
                                                                                           Open Space
                                                                                            of Screen
   4

   16
   30
   80
Water film between screen wires ruptures almost immediately

                                         100
       1
       2
       3
       5
      10

       1
       2
       3
       5

       1
       2
       3
       5
      99.6
      74.4
      73.1
      74.4
      75.0

     114.0
      57.8
      49.2
      46.5

      74.0
      32.2
      21.4
      19.9
  120
   85
    99.6
    74.4
    73.1
    74.4
    75.0

    95.0
    48.2
    41.0
    38.7

    87.1
    37.9
    25.2
    23.4
                                       60.2
                                                                              40.8
                                                                              19.4
 *Screens drained into separate container after each surface contact.
**Area of monolayer removed divided by area of screen.
                                                      56

-------
     By using this technique with a 16-mesh screen, oleic acid,
stearic acid, and plankton oil monolayers were retrieved from a
sea water surface at pressures ranging from 1 to 10 dynes/cm.
Since the screen efficiency was approximately 75 percent for all
of the film pressures studied, there was no apparent dependence
of the method on the degree of compaction of the monolayers within
these ranges.  It was determined experimentally that a 0.15 mm-thick
layer of water was collected by the 16-mesh screen made from 0.14-mm
diameter wire.

IV.  Field Collection Equipment

     A 16-mesh Monel screen of 0.014 inch diameter wire containing
60.2 percent open space was incorporated into a 30 by 24 inch aluminum
frame.  Two 30-inch aluminum handles of 3/4 inch rod were bolted onto
the frame.  In practice the screen was placed through the water surface.
It is then withdrawn horizontally (parallel to water) through the water
surface and drained immediately into a sample bottle as illustrated in
Figure 1.  This procedure is repeated until the sample bottle is filled.
Approximately 200 to 250 surface contacts are required to collect a
5-gallon sample, which corresponds to a sea surface area of from 1000 to
1250 square feet per sample.

V.  Verification of Screen Collection Method

     The screen collection procedure was verified in a 17-foot tank by
the following experiment.  A compressed oleic acid monolayer was spread
onto the surface water on one side of a paraffin wax-coated cord.  The
experimental situation depicted in Figure 2 shows the positions of the
floating cord before the commencement of monolayer retrieval.  When the
surface of the tank was almost completely covered by monolayer, as indi-
cated by the position of the cord (area I), a drop of oleic acid was
placed on the remaining uncovered surface (area II).  This lens acted
as a monomolecular piston to drive the barrier cord back across the tank
as the oleic acid monolayer in area I was removed by the screen entrap-
ment method.  As the monolayer was retrieved the barrier moved across
the tank decreasing the area of I.  From the number of screen contacts
and an estimation of the area of monolayer removed, it was determined
that the screen method was about 70 percent area efficient in continued
use.  This figure is in good agreement with the film-balance data recorded
in Table I for the 16-mesh screen.

     The fact that 100 percent efficiency is not attained is due to an
initial adsorption of a portion of the monolayer onto the screen wire.
This material remains adsorbed onto the screen wires, thereby deactiva-
ting the metal so that additional adsorption does not occur during the
remainder of the sampling operation.  A number of these experiments were
performed until a 5-gallon container was filled with the water-oleic
acid mixture.  By calculation, 55 milligrams of oleic acid had been re-
moved.  Concentration and recovery of the oil yielded 38 milligrams of material,
                                   57

-------
           fill mark
                                     collection screen
                                       handles bolted to  frame
                                    plastic sample bottle
      FIGURE 1-Sea-surface collection assembly in draining position
          oleic acid piston
         monolayer & lens
floating barrier  cord
                         Oleic Acid Monolayer
                         Spread From Benzene
                                Solution
                                                 17' tank
            FIGURE 2-Top view of tank experiment to study
                     monolayer retrieval method

-------
     Field verification was obtained by retrieving a condensed monolayer
of oleic acid which had been spread on the Chesapeake Bay.  No natural
slick was apparent in the area to give a background film as interference.
Processing the 4-gallon sample collected yielded 19 milligrams of oleic
acid.  The yield was somewhat reduced by large quantities of silt present
in the samplewaich when filtered off removed oleic acid by adsorption.

     The method was also applied to a natural slick of unknown origin on
Chesapeake Bay.  Film-pressure measurements using Adam's spreading drop
method varied from 3 to 12 dynes/centimeter and averaged 5 dynes/centimeter
in the slick.  The water beneath the surface was unusually clear for this
body of water and little silt appeared in the sample.  Approximately 27
milligrams of surface-active material was isolated from the 5-gallon sample.
The multi-component mixture was separated by selective solubility in various
solvents.  The chemical and physical properties of the collected slick and
its constituents will be included in a future report on the constitution of
natural slicks sampled at various stations in the Atlantic and Pacific
Oceans, the Bay of Panama, and the Gulf of Mexico.

     These experiments demonstrated that it was possible to collect signi-
ficant quantities of materials which are spread out upon the sea surface
in monomolecular thicknesses, making possible the retrieval of sea slick
material in concentrated form.

VI.  Summary

     A method for the removal and isolation of the substances present in
natural and man-made oceanic oil slicks had been developed.  The technique
utilizes a planar metal wire screen which when drawn through a water sur-
face removes discrete segments of the surface layer between the wires.  In
this manner, a 0.15-mm film of surface water containing all constituents
of the surface film may be collected.  This approach is found to be effective
regardless of the film pressure; hence, it may be applied to monolayers under
low film compression.
                                     59

-------
Addendum on Isolation and Analysis of Collected Organic Material
     In sampling, the screen enclosed in a metal frame with handles is
placed into the water parallel to the surface, withdrawn, and drained
into a collection vessel.  The resulting concentrate of polar material
and sea water is further concentrated by the subsequent precipitation
of a ferric hydroxide sol within the sample bottle.  The organic con-
stituents are occluded and adsorbed by the resulting gelatinous preci-
pitate which after settling is separated from the sea water.  The con-
centrate is acidified to dissolve the ferric hydroxide and extracted
with various solvents.  The solvents are pumped off under vacuum leaving
behind the nonvolatile surface-active materials of which the slick was
composed.

     A procedure has been developed for the concentration and separation
of the collected slick constituents from the background sea water.  The
technique was preliminarily studied by Jeffrey and Hood (6) who found
that 95% of a complex mixture of all organic material derived from an
algae culture was coprecipitated with a ferric hydroxide gel.  This
approach has also been recently reported by Williams (7) as a method
for the concentration of organic acids in sea water samples.  The fol-
lowing is the procedure used in this Laboratory for the isolation of
slick material.

     The pH of the sample is adjusted to about 8.0 with sodium carbonate
solution.  Then 50 mg of ferric chloride per liter of sample is added
with stirring from a 1% solution.  To obviate cumbersome filtration,
the resulting sol is allowed to settle for several hours and the super-
natant water is siphoned off.  The sol is then acidified with 1:1 hydro-
chloric acid to a pH of 2.0, with stirring until the ferric hydroxide
dissolves.  A deep yellow but generally clear solution results.  The
mixture is then extracted with several 100-ml portions of organic sol-
vent .  The solvent is pumped off under vacuum at ambient temperatures
to avoid thermal modification of the solute.  The solvents used should
be freshly distilled and contain no appreciable residue upon evaporation.
Solvents used in this work were petroleum ether, benzene, ether, and
chloroform.

     Experimentally, the above procedure was found to be about 70% ef-
ficient when benzene or petroleum ether was used in the solvent extraction
step.  Hundred-mg portions of oleic acid, dodecyl alcohol, and plankton
oil were dissolved separately in 5 ml of acetone and dispersed into
separate 5-gallon containers of synthetic sea water.  Recovered from the
mixture by the foregoing procedure were 75 mg of oleic acid, 65 mg of
dodecyl alcohol, and 68 mg of plankton oil.
                                   60

-------
     Centrifugation was used as an alternate method for the isolation of
the slick material.  It was found that most of the organic materials are
adsorbed onto any particulate matter present in the sea water sample.
These solids and biological entities are thrown out onto the wall of the
centrifuge cylinder from which the organic matter may be removed by sol-
vent extraction.  The centrifugate is then treated by the ferric chloride
precipitation method to retrieve any substances not removed by centrifuga-
tion.

     Ideally it is desirable to collect only those chemicals contributing
to the slick phenomena.  To remove the undesired biological entities, silt,
and flotsam from the collected surface chemicals, the Chesapeake Bay samples
were passed through a motor-driven Sharpies super centrifuge.  A continuous-
feed cylinder was driven at 23,000 rpm developing 13,200 g at the cylinder
wall.  When the clear centrifugate was extracted with organic solvents less
than 1 mg of organic material was isolated.  A duplicate sample was collected
from the same bay slick.  This sample was not centrifuged and yielded 12.8 mg
of organic materials by extraction.  Solvent treatment of the solids collected
on the centrifuge cylinder from the centrifuged sample resulted in a 13.7 mg
collection of chemicals whose infrared spectra approximately matched the
spectra obtained for the materials from the uncentrifuged sample.

     The centrifuge removed not only the particulate matter but also most
of the surface-active matter by adsorption onto the discontinuous phase.
Thus, the centrifuge appeared to be a tool for the isolation of the slick
constituents from the background sample water.  A liner of clean aluminum
foil was inserted into the centrifuge bowl to facilitate the removal of
the collected solids.  After centrifugation of the sample, the liner is
removed and washed with solvent to dissolve and desorb the adherent surface-
active materials.

     In the event that the collected sample does not contain sufficient
particulate matter to adsorb the surface-active constituents, it is recom-
mended that all centrifugates be treated with ferric chloride as detailed
in the preceding section.
                                     61

-------
                               REFERENCES
1.  N. K. Adam, Proc. Roy. Soc., London (B) 122:134-272 (1937).

2.  J. R. Lumby and A. R. Folkard, Variation in the Surface Tension
    of Sea Water in situ, Bull. Inst.  Oceanographique, No. 1060:1-19
    (1956).

3.  C. Ewing, Slicks, "Surface Films and Internal Waves", J. Mar. Res.
    9:161-187.

4.  R. S. Dietz and E. C. LaFond, "Natural Slicks on the Ocean", J. Mar.
    Res. 9:69-79 (1950).

5.  C. 0. Timmons and N. L. Jarvis, A Chemical Characterization of Plank-
    ton Oil, NRL Memorandum Report 1033, March 1960.

6.  L. M. Jeffrey and D. W. Hood, Journal of Marine Research 7:247 271.

7.  P. M. Williams, Nature 189:219 (1961).
                                                       GPO 953-741
                                   62

-------