-«jSt»»s Office of EPA-520/1-88-002
tit¥«nm«mal Protection Radiation Programs August 1988
Agency Washington, DC 20460
&EPA Sediment Monitoring at
Deep-Ocean Low-Level
Radioactive Waste
Disposal Sites
Methods Manual
-------
EPA 520/1-88-002
A METHODS MANUAL
for
SEDIMENT MONITORING AT DEEP-OCEAN
LOW-LEVEL RADIOACTIVE WASTE DISPOSAL SITES
edited by
James S. Booth
Atlantic Marine Geology Branch
U.S. Geological Survey
Woods Hole, Massachusetts
August 1988
Prepared in cooperation with the
U.S. Environmental Protection Agency
under Interagency Agreement DW14931699-01-0
Project Officer
James Neiheisel
Office of Radiation Programs
U.S. Environmental Protection Agency
Washington, DC 20460
-------
FOREWORD
In response to Public Law 92-532, the Marine Protection,
Research and Sanctuaries Act of 1972 as amended, the Environmental
Protection Agency (EPA) promulgated regulations in 1977 to control
disposals of waste materials in the oceans. The EPA is currently
developing revisions to the existing Ocean Disposal Regulations.
The EPA Office of Radiation Programs (ORP) is developing site
designation, waste package performance, and monitoring criteria
applicable to ocean disposal of low-level radioactive wastes
(LLW). The ORP is also preparing technical information reports to
support LLW disposal criteria.
This report is a methods manual for monitoring sediments in
the deep ocean. It is intended to be a frame of reference for
baseline monitoring to characterize LLW disposal sites, and for
site designation and trend assessment monitoring of sediments.
The Agency invites all readers of this report to send any
comments or suggestions to Mr. David E. Janes, Director, Analysis
and Support Division, Office of Radiation Programs (ANR-461),
Environmental ProtectioruAgency, Washington, DC 20460.
Richard J/ Guimond, Director
Office of Radiation Programs (ANR-458)
111
-------
PREFACE
This manual contains methodologies for monitoring specific
environmental parameters at LLW disposal sites. It includes
procedures for 15 types of analyses and specific recommendations
for shipboard operations related to the collection, handling, and
storage of sediment samples. These analyses and recommendations
are not necessarily permanent parts of the EPA regulations or
criteria for the disposal of LLW on the seafloor, nor is the
material included herein to be considered exclusionary: the manual
is a working document that represents a step toward the eventual
assurance that prudent site investigations will be conducted.
Moreover, the methods presented are not to be considered
definitive: analytical methods are constantly improving, new
methods can be developed, and technology is advancing. The
methods given do, however, provide a frame of reference and a
starting point.
In some cases, more than one method is given for a
particular analysis. This does not imply that the manual is
intended to be a methods encyclopedia. Rather, it simply
underscores the fact that alternative methods exist or, that for
some analyses, more than one method is in common use.
This manual is a product of the efforts of many
individuals. Contributors are: P. Colombo, M. Fuhrmann, and
R. Pietrzak of Brookhaven National Laboratories and Associated
Universities, Inc. in Upton, NY; R. Willingham and J. Nowland of
the U.S. Army Corps of Engineers in Marietta, GA; K. Fanning of
the University of South Florida in St. Petersburg, FLr and
J. Booth, M. Bothner, J. Hathaway, F. Manheim, C. Parmenter,
L. Poppe, and W. Winters of the Atlantic Marine Geology Branch,
U.S. Geological Survey (USGS) at Woods Hole, MA. The compilation
and editing of incorporated materials were accomplished by the
USGS. The draft final report was transferred from the USGS NBI
word processing system for editing on the EPA Office of Radiation
Programs (ORP) word processing system by P. Cuny and W. R. Curtis
of ORP. The report was then revised, as necessary, by Mr. Curtis
and prepared for publication.
v
-------
TABLE OF CONTENTS
Paqe
FOREWORD i i i
PREFACE V
LIST OF TABLES xi
LIST OF FIGURES xii
CHAPTER 1 SELECTION OF SEDIMENT MONITORING CRITERIA
Introduction 1
Radionuclides of Concern 1
Factors That Control Radionuclide Retention 3
Recommended Measurements 3
Geological
Sediment Texture 4
Mineralogy 4
Radiography 4
Geochemical
Distribution Coefficient 4
Oxidation-Reduction Potential 5
pH 6
Total Organic Carbon 6
Geotechnical
Index Properties 7
Vane Shear Strength 7
One-dimensional Consolidation Test 7
Static Triaxial Compression 7
References 8
CHAPTER 2 SAMPLING AND SHIPBOARD RECOMMENDATIONS
Introduction 9
Sampling 9
Handling 11
Storage 13
Shipping 14
Summary 15
References 16
GEOLOGICAL METHODS
CHAPTER 3 TEXTURE
Introduction 18
Procedure 18
Comments 36
Quality Assurance 39
Cost Analysis 39
References 41
vi i
-------
TABLE OF CONTENTS (Continued)
Page
CHAPTER 4 X-RAY DIFFRACTION MINERALOGY
Introduction 42
Procedure 42
Comments 44
Quality Assurance 47
Cost Analysis 48
References 49
CHAPTER 5 RADIOGRAPHY
Introduction 51
Procedure 52
Comments 53
Quality Assurance 53
Cost Analysis 53
References 54
GEOCHEMICAL METHODS
CHAPTER 6 DISTRIBUTION RATIOS (Rj)
Introduction 56
Procedure 58
Discussion 60
Quality Assurance 61
Cost Analysis 61
References 61
CHAPTER 7 IRON AND MANGANESE
Introduction 63
Procedure 63
Quality Assurance 64
Cost Analysis 65
References 65
CHAPTER 8 NITRATE
Introduction 66
Porewater Extraction 66
Nitrate Analysis 72
Nitrite 76
Nitrate + Nitrite 77
Quality Assurance 81
Cost Analysis 81
Discussion 81
References 82
vin
-------
TABLE OF CONTENTS (Continued)
Page
CHAPTER 9 Eh AND pH
Introduction 84
Eh 84
Procedure 85
Quality Assurance 86
Cost Analysis 87
pH 87
Procedure 88
Quality Assurance 89
Cost Analysis 90
References 91
CHAPTER 10 TOTAL ORGANIC CARBON
Introduction 92
Test Method A
Procedure 92
Quality Assurance 94
Cost Analysis 95
Test Method B
Procedure 95
Quality Assurance 96
Cost Analysis 96
Discussion of Methods 97
References 97
GEOTECHNICAL METHODS
CHAPTER 11 WATER CONTENT
Introduction 100
Procedure 100
Comments 103
Quality Assurance 104
Cost Analysis 104
References 104
CHAPTER 12 ATTERBERG (LIQUID AND PLASTIC) LIMITS
Introduction 105
Liquid Limit
Test Method A: Casagrande Drop Cup 107
Quality Assurance 110
Cost Analysis 110
Test Method B: Fall-Cone Penetrometer 113
Quality Assurance 116
Cost Analysis 116
Discussion 116
Plastic Limit
Procedure 116
Comments 119
Quality Assurance 119
Cost Analysis 119
References 120
ix
-------
TABLE OF CONTENTS (Continued)
CHAPTER 13 GRAIN SPECIFIC GRAVITY
Introduction 121
Test Method A: Water-Filled Pychnometer 121
Quality Assurance 122
Cost Analysis 122
Test Method B: Gas-Pressurized Pychnometer 122
Quality Assurance 124
Cost Analysis 127
Discussion 127
References 127
CHAPTER 14 LABORATORY VANE SHEAR STRENGTH
Introduction 128
Procedure 128
Comments 131
Quality Assurance 132
Cost Analysis 136
References 136
CHAPTER 15 ONE-DIMENSIONAL CONSOLIDATION
Introduction 137
Procedure 137
Comments 141
Quality Assurance 141
Cost Analysis 144
References 144
CHAPTER 16 STATIC CONSOLIDATED-UNORAINED COMPRESSIVE
STRENGTH
Introduction 145
Procedure 145
Comments 149
Quality Assurance 152
Cost Analysis 152
References 153
APPENDIX A Preparation of Randomly Oriented
Mounts for X-ray Diffraction 154
APPENDIX B Separation of Clay Fraction 160
APPENDIX C Silver Filter Preparation 171
APPENDIX D Ethylene Glycol Vapor Treatment 174
APPENDIX E Heat Treatment 176
APPENDIX F Flow Charts For Identifying Clay
Minerals With Various 001 Diffraction
Maxima. 178
APPENDIX G Preparation of Glass Slides for
Optical Microscopy 182
-------
LIST OF TABLES
Page
Table 1 Radionuclides Commonly Present in LLW. 2
Table 2 Correlation of Grain Size phi Classes with
Sieve Sizes and Numbers for ASTM and Tyler
Mesh Sieves. 23
Table 3 Comparison of Withdrawal Time and Depth Tables
for Temperatures of Suspension Maintained at
20 and 24 °C. 29
Table 4 Cost Per Sample Estimates for a Typical
Computerized Sedimentation Laboratory. 40
Table 5 Breakdown of Analysis Costs for Semiquantitative
Estimates of Mineralogical Composition Using
X-ray Powder Diffraction. 48
Table 6 Selected Radionuclides (Beta and Gamma
Emitters) Potential Associated With Nuclear
Reactor Wastes. 57
Table 7 Precision of Method A Using Dry Pretreatment
Method. 94
Table 8 Range, Sensitivity and Accuracy of LECO
Analyzer. 95
Table 9 Corrections to Natural Water Content Within
a Subsample to Account for a Salinity of 35 ppt
in the Pore Water. 103
Table 10 Corrections to Add to Raw Specific Gravity
to Account for a Salinity of 35 ppt in the
Pore Water. 127
XI
-------
LIST OF FIGURES
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10
Figure 11
Figure 12
Figure 13
Figure 14
Example of a REQUEST FOR ANALYSIS Form. 19
Example of a SAMPLE ID Form. 20
Differential Pressure Settling Tube for
Sand-Fraction Analysis. 25
Example of a Raw Data Record Produced by
an Automated Settling Tube. 27
A Coulter Counter Electroresistance
Multichannel Particle Size Analyzer with the
Sample Stand in a Faraday Cage. 31
Histogram of a Bimodal,, Poorly Sorted,
Silty Sand. 34
Arithmatic Ordinate, Cumulative Frequency
Curve of a Well Sorted, Unimodal Sand and a
Very Poorly Sorted, Bimodal Clayey Silt. 35
Moving Average Plot of a Core Penetrating
Interbedded Silty Sands and Clayey Silts. 37
Graphical Representation of Sand, Silt and
Clay Compositions in Silty Sand, Sandy Silt
and Silty Clay, by a Triangular Diagram. 38
A Metal Membrane Mount Used for X-ray
Powder Diffraction. 46
Gas-displacement Squeezer for Use with
Multisqueezer Manifold. 68
Schematic Diagram of an Inert-atmosphere
System for Extruding, Slicing and Squeezing
Sediment. 70
Schematic Diagram of the Configuration of a
Technicon Autoanalyzer II for the Determination
of Dissolved Nitrite. 73
Schematic Diagram of the Configuration of a
Technicon Autoanalyzer II for the Determination
of Nitrate + Nitrite. 74
Xll
-------
LIST OF FIGURES (Continued)
Page
Figure 15 Cadmium-Copper Reductor Column for the
Determination of Nitrate + Nitrite in
Marine Porewaters. 79
Figure 16 Typical Water Content Data Form. 101
Figure 17 Plasticity Chart Showing Location of
Some Types of Soil. 106
Figure 18 Mechanical Drop Cup Liquid Limit Device
and Grooving Tool. 108
Figure 19 Plot of Water Content versus Number of Drops
from a Drop Cup Liquid Limit Test. Ill
Figure 20 Typical Liquid Limit and Plastic Limit Test
Data Form. 112
Figure 21 Fall Cone Penetrometer and Electric Timer
Used for Determining the Liquid Limit of
Sediment. 114
Figure 22 Typical Fall-cone Atterberg Limits and
Summary Data Form. 115
Figure 23 Plot of Correlation Between Fall Cone and
Drop Cup Methods of Determining Liquid Limits. 117
Figure 24 Typical Data Form for Grain Specific Gravity
Determined by using the Water Filled Pycnometer
Method. 123
Figure 25 Gas Pressurized Pycnometer. 125
Figure 26 Form for Recording Data and Determining
Grain Specific Gravity by Gas Pressurized
Pycnometer. 126
Figure 27 Vane Shear Machine that Uses a Spring to
Apply Torque to the Vane. 129
Figure 28 Typical Vane Shear Data Form. 130
xni
-------
LIST OF FIGURES (Continued)
Page
Figure 29
Figure 30
Figure 31
Figure 32
Figure 33
Figure 34
Figure 35
Figure 36
Figure 37
Vane Shear Machine that Uses a Torque
Sensor to Rotate the Vane.
Torvane Shear Strength Devic, Soft Sediment
Sampler and Stiff Sediment Adapter.
Pocket Penetrometer and Soft Sediment Adapter,
Typical CRS Consolidation Test Data Form.
Typical Constant-rate-of-strain Consolidation
(CRSC) Sample and Test Chamber Configuration.
Constant-rate-of-strain Consolidation Test
Summary Form.
Typical Triaxial Test Data Form.
Typical Components of a Triaxial Test Device.
Consolidated-undrained Triaxial Compressive
Strength Test Summary Form.
133
134
135
138
139
142
143
147
148
150
151
xiv
-------
CHAPTER 1
SELECTION OF SEDIMENT MONITORING CRITERIA
Introduction
The scientific suitability of a deep-ocean site for
low-level radioactive waste (LLW) disposal depends on the type of
radioactive waste material; the packaging parameters; the natural
physicochemical characteristics of the site, including
oceanographic factors (i.e., collectively, the local marine
environment); and the waste's predicted impact on the biosphere,
including food chain relationships.
The monitoring criteria presented in this document are
those related to the physicochemical environment. The
measurements included will provide guidance within this category
for the site selection. These criteria, which include recommended
measurements, were established because they provide information on
the radionuclide retention capability of deep-sea sediment, which
is the ultimate host medium for the LLW. The measurements also
provide pertinent information on the geologic stability of the
site, the penetrability and subsequent settlement of the waste
packages, and on the corrosiveness of the environment with respect
to the packaging material. Finally, the measurements will help to
characterize the site for many other considerations which can be
applied to modeling or other pertinent evaluation activity.
The measurements included in this manual provide a
framework of scientific guidelines within which to evaluate a
candidate LLW site's ability to retain radionuclides of concern
over a specified time period and in a semiquantitatively
acceptable manner.
Radionuclides of Concern
LLW typically contains a variety of radionuclides, many of
which are present in small quantities and/or have very short
half-lives. These nuclides are generally not considered to pose a
serious health hazard per se. However, other nuclides may be
considered "radionuclides of concern." Such nuclides are defined
in terms of the following criteria: (1) half-life more than 5
years, (2) presence in relatively significant quantities, and (3)
biological toxicity (Wild et al 1981). Radionuclides that are
commonly present in LLW and that meet at least one of the
aforementioned criteria are listed in Table 1. Most of these
radionuclides will be eliminated or are present in negligible
amounts. However, three of the nuclides that will be present in
the waste package and meet all the criteria are cesium, cobalt,
and strontium (specifically the 137Cs, 60Co, and 90Sr
isotopes).
-------
Table 1. Radionuclides Commonly Present in LLW
Name Isotopic Symbol Half-life (Years)
Cobalt 60Co 5.3
Tritium 3H 12.3
Plutonium 241pu 14.4
Curium 244cm 17.6
Strontium 90Sr 28
Cesium !37Cs 30.2
Curium 243Qm 32
Plutonium 238pu 87
Nickel 63Ni 92
Americium 241Am 445
Carbon 14C 5,730
Plutonium 240Pu 6,580
Americium 243Am 7,650
Niobium 94Nb 20,000
Plutonium 239Pu 24,110
Nickel 59Ni 82,000
Technetium 99Tc 210,000
Cesium 135cs 2,000,000
Neptunium 237Np 2,100,000
Iodine 129I 16,000,000
Uranium 235U 710,000,000
Uranium 238U 4,510,000,000
-------
Factors That Control Radionuclide Retention
The primary barrier to radionuclide migration is the
waste package (the matrix and container) itself, but because the
package lies on or in the sea floor, an additional barrier is
available: the contiguous sediment. The relative effectiveness of
this natural barrier depends on the portion of the package that is
in contact with or embedded in the medium and the retention
potential of the sediment itself. The area of contact or
embedment may be roughly estimated a priori by using certain
geotechnical parameters and geologic information. The area of
conflict is also the implicit justification for the recommended
measurements, which serve as a means to estimate the effectiveness
of those sediments as a barrier.
Numerous factors control the inherent ability of the
sediment to retain radionuclides. Onishi et al (1981) have
examined selected radioisotopes and the factors that control their
uptake in natural environments. They show that the ability of
marine sediments to sorb these isotopes is governed by the type
and quantity of sorptive minerals, complexing ligands, amorphous
oxides, organic matter, and competing ions; pH; Eh; and the
potential for colloid formation.
Recommended Measurements
The measurements and associated methods in this manual
are intended to provide either direct or indirect information with
which to judge the radionuclide retention potential of a candidate
LLW site. In recognition of the multidisciplinary nature of these
assessments, the manual is divided into three parts: Geological,
Geochemical, and Geotechnical. Guidelines for proper sample
collection and treatment aboard ship are also included.
The methods presented herein are either generally
accepted as standard and widely used by laboratories concerned
with batch analysis, or they represent state-of-the-science
analytical procedures for those measurements for which there are
no standards or common methods. Although alternative methods are
available in many cases, use of these methods provides a basis for
comparison and evaluation and promotes uniform data reporting.
The types of measurements include those currently listed
in current ocean disposal regulations, those additonally
reommended because of the specific monitoring needs pertaining to
the disposal of packaged LLW at depths of 4,000 meters or more
(this depth is recommended by the International Atomic Energy
Agency pursuant to the London Dumping Convention agreement, of
which the United States is a signatory), and those that are
-------
presented as optional. A document entitled "Sediment Monitoring
Parameters and Rationale for Characterizing Deep Ocean Low-Level
Radioactive Waste Disposal Sites" (EPA 520/1-87-011) provides more
background information and detail on these measurements. The
following sections present each type of measurement and a summary
of that measurement's bearing on radionuclide retention in marine
sediments.
Geological Measurements
(1) Sediment Texture is a key variable for site
evaluation and is a required measurement under existing ocean
disposal regulations. Grain-size data may be used to estimate
basic radionuclide retention (i.e., sorptive characteristics), to
correlate data from other fundamental geotechnical measurements
(which in turn can be used to estimate package embedment), and to
obtain an indication of geologic stability (i.e., recent or active
dynamic processes on the seabed). Grain-size data often correlate
with organic content and other geochemical variables, may also be
used to calculate mineral percentages, and generally are the most
versatile of all the indirect monitoring methods.
(2) Mineralogy is important because of the
sorption properties of several of the radionuclides of concern,
including cesium and strontium. Sorption is primarily dependent
on cation exchange and large surface areas. Sediments rich in
zeolites, smectites, and other clay minerals have a high sorption
potential for the retention of the radionuclides relative to
sediments dominated by less sorptive minerals such as quartz,
feldspars, calcite, and nonclay phyllosilicates (i.e., muscovite
or biotite). Moreover, identification of minerals that may have
economic potential could disqualify a site for LLW disposal. An
accurate inventory of the mineralogical suite is therefore
required under the current ocean disposal regulations.
(3) Radiography of sediments provides information
that pertains to radionuclide retention, geologic stability, and
package embedment. Bioturbation, for example, which is detectable
in x-radiographs, can be evaluated as an indicator of vertical
mixing of the sediment (and vertical transport of radionuclides).
The presence of laminae, clasts or sand, and graded beds can also
be detected, all of which bear on the geologic stability of the
site.
Geochemical Measurements
(1) The distribution coefficient (Rd) reflects
the ability of a sediment to sorb a selected radionuclide under
specific (e.g., in-situ) conditions. It is therefore a direct
measure of the retention potential in a proposed site. It is
defined (Pietrzak et al 1981) as the ratio of specific
-------
activity in the sediment phase to that in the liquid phase for the
radionuclide of interest. That is,
R<3 = sediment activity/weight of sediment
liquid activity/volume of liquid
Thus, the greater the R^ value, the greater the nuclide
retention by the host sediment. The method of R^ measurement in
this manual provides the only direct measurement of radionuclide
retention potential in a candidate seafloor site. However,
because the quantitative effectiveness of scavenging (sorbing
nuclides from solution) depends on many factors such as
mineralogy, redox state, and the presence of competing ions, it is
essential that a profile is developed that includes the "fluff"
(the topmost surficial sediment and nepheloid layer particulates)
as well as sediment that may be at the limit of waste package
penetration.
Retention may also be examined in terms of diffusion.
The rate of radionuclide diffusion, from a source below the
sediment-water interface into the water column, may be an
important parameter to measure for many nuclides with low R^'s
or for those that are particularly abundant. Accordingly, the
measurement of diffusion coefficients is under consideration for
possible inclusion in this and related documents.
(2) Oxidation-Reduction Potential
Three separate methods are given in this
manual for evaluating the redox state of the sediment: (1) direct
measurement of Eh, (2) measurement of sediment Fe and Mn
concentrations in profile, and (3) measurement of nitrate
porewater concentration in profile, which is the most accurate,
complicated and expensive of the three. Retention of several
radionuclides of concern, including 6°Co, 99>pc an(j 237fjpf is
significantly affected by redox state, making this parameter one
of the key elements in a LLW site investigation. Redox state
changes from oxygenating (+) conditions at the sediment surface to
reducing (-) conditions at subbottom depths from several
centimeters to several meters in deep-ocean sediments (Schmidt
1979, Carpenter et al 1983, and Wilson et al 1983)). Thus, a
profile must be developed to determine the position of the
redox-state boundary in the sediment column and the relative
position of this boundary to waste package penetration depth.
In-situ redox potential of deep-ocean sediments is
extremely difficult to determine. Redox state is sensitive and
highly variable, and the removal of samples from their natural
geochemical environment may compromise the validity of subsequent
shipboard or laboratory measurements ashore, particularly if the
sample has been exposed to an oxygen-rich environment.
-------
A sulfide sulfur test to assess redox state may be
included in a subsequent version of this manual. Because the
reduction of sulfate to sulfide signals the presence of anaerobic
conditions, this test, which is simple and relatively inexpensive,
could provide, as do the other three redox methods, information on
the position of the boundary between oxygenating and reducing
conditions. Moreover, because sulfide formation requires the
presence of both iron and sulfate-reducing bacteria, which is the
case when a steel waste drum is in a reducing environment,
corrosion of the drum occurs through cathodic depolarization
(Uhlig 1971). This affects the efficiency of the primary
retention medium, the waste package. The sulfide sulfur test may
therefore provide predictive information on waste package
performance as well as redox state. Despite this additional
utility, however, the sulfide sulfur test is considered inferior
to the other three redox tests because it requires interpretation
of the sediment history, and consequently is generally a less
reliable test and specifically a less accurate test with regard to
the placement of the oxygenation-reduction boundary. Accordingly,
the sulfide sulfur test should only serve as a backup to one of
the other types of measurement.
(3) Hydrogen ion activity (pH) affects retention
of strontium, cobalt, plutonium and other important radionuclides
because it affects their solubility (Onishi et al 1981) . pH is a
standard geochemical descriptor and is necessary for understanding
or predicting reactions. To date, most measurements indicate that
pH may not vary significantly in deep-sea sediments, but recent
in-situ data suggest that it may be more variable than earlier
studies indicated.
As presented in this manual, pH is a simple measurement
utilizing a standard pH meter and probe. The measurement may be
done in conjunction with Eh measurements. As with Eh, however,
measurement results may not be representative of in-situ
conditions unless great care is taken. In the case of pH,
outgassing must be avoided, because the buffering capability of the
porewater at the probe site may be affected.
(4) Total organic carbon
The radionuclide retention capability of
a sediment can be adversely affected by organic matter. For
example, it has been demonstrated at LLW land sites that 60Co
can form complexes with organic constituents and that these
complexes are not easily sorbed (retained) by the host medium. In
effect, the geologic barrier is rendered inefficient by such
complexes. The isotopes 90Sr, 239Pu and others also form
complexes with organic constituents (Onishi et al 1981); although
a general lack of data on these nuclides precludes predictions
about retention, they too may affect the retention capability of
the host sediment. Measurement of organic carbon is required
under the current ocean dumping regulations.
-------
Geotechnical Measurements
(1) Index properties
Natural water content, liquid and
plastic limits, and grain specific gravity constitute the
geotechnical index property suite for this methods manual. These
mass physical properties are a basis for sediment classification;
they correlate with textural and compositional sediment attributes
and give insight into the stress history, stress behavior
(comprassion index), permeability and numerous other engineering
characteristics of sediments. More specifically, index property
data may be used to provide a first-order prediction of waste
package penetration and settlement into sediments. They will also
provida data on the geologic stability of a LLW site and a
first-order approximation of its retention potential.
(2) The vane shear test measures undrained shear
strength of a cohesive (fine-grained) sediment. Both "natural"
and remolded strengths are normally determined on a sample. The
former gives an indication of the maximum strength and the latter
of the minimum strength of the sediment. This test has direct
utility in evaluating waste package embedment, and also may be
used to assess geologic stability and the degree of compaction of
the sediment. The test is currently optional, but is being
considered as a recommended measurement for LLW site monitoring if
package burial becomes a disposal option.
(3) One-dimensional consolidation tests indicate
stress-strain relationships in a sediment and thus are directly
applicable to settlement analysis of an applied load such as a
waste canister. In addition, the test provides information on the
maximum past stress (overburden) on a site, which has implications
regarding erosion, mass movement, sedimentation rate, excess pore
pressure, and other factors that bear on geologic stability.
Consolidation testing is optional. The method presented in this
manual (constant-rate-of-strain) is the most rapid of the common
methods, and is relatively inexpensive.
(4) Static triaxial compression tests can
determine drained and undrained strength parameters as well as
stress-strain relationships. It is appropriate for analyzing
canister penetration and for estimating subsequent settlement
under the load conditions imposed by the waste package. Results
of this test also may be used in a variety of slope stability
analyses and other evaluations that bear on geologic stability.
-------
References
Carpenter, M.S.N. et al, 1983. Geochemistry of the near
surface sediments of the Nares Abyssal Plain, Institute
of Oceanographic Sciences Report 174, 66 pp.
Onishi, Y. et al, 1981. Critical review: radionuclide
transport, sediment transport, and water quality
mathematical modeling; and radionuclide absorption and
desorption mechanisms, Battelle Pacific Northwest
Laboratory, PNL-2901, 339 pp.
Pietrzak, R.F., Czyscinski, K.S., and Weiss, A.J., 1981.
Sorption measurements performed under site-specific
conditions - Maxey Flats, Kentucky and West Valley, New
York disposal sites, Nuclear and Chemical Waste
Management, v. 2, pp. 279-285.
Schmidt, R.L., 1979. The chemistry of water and sediment
from the benthic boundary layer at a site in the
Northwest Atlantic Ocean, Battelle Pacific Northwest
Laboratory, PNL-22842, UC11, 23 pp.
Uhlig, H.H., 1971. Corrosion and corrosion control,
Wiley Interscience, 96 pp.
Wild, R.E. et al, 1981. Data base for radioactive waste
management, U.S. Nuclear Regulatory Commission,
NUREG/CR01759, v. 2.
Wilson, T.R.S. et al, 1983. Status report on geochemical
field results from Atlantic study sites, Institute of
Oceanographic Sciences Report 175, 56 pp.
-------
CHAPTER 2
SAMPLING AND SHIPBOARD RECOMMENDATIONS
Introduction
The only general requirement for sampling is that the
upper 30 cm of sediment must be recovered. However, if the
recovered sample is to be truly representative of in-situ
conditions, it must be collected with tools of optimum design and
receive appropriate treatment during the pre-analyses phase.
These requisites are often compromised by cost and time
constraints, and by shipboard and laboratory pragmatisms. Thus, a
necessary preliminary to LLW disposal site investigations is to
define a level of sample quality that is both acceptable and
attainable within the practical constraints of field and
laboratory operations. This chapter provides guidelines and
rationales for sampling and sample treatment that support the
analytical methods found within this manual.
Sampling
Some disturbance or alteration of a sediment sample
during conventional sampling activities is inevitable. The simple
act of removing a parcel of sediment from its natural environment
and transporting it to the surface precludes the possibility of
ever obtaining a completely undisturbed sample because in-situ
stresses are released during recovery. The effect of this on
sample quality is especially obvious for a sediment that is
overprocessed in-situ or one that contains dissolved gases which
may ebullate as hydrostatic pressure is reduced from more than 400
atmospheres (at approximately 4,000-meters water depth) to 1 atm
on deck. However, if such alterations are limited to changes in
stress, the sampling activity was successful.
Although the dramatic change in stress cannot be
obviated, unless a pressurized sampling system is developed, it
may only be detrimental to overall sample quality in rare
circumstances. In most cases, other sources of disturbance have
the most affect on sample quality. These sources include varying
degrees of shock, sediment compression, skin friction and other
actions that accompany sampler penetration; as well as the
tension-release and "flow-in" changes in stress that may accompany
sampler withdrawal. These gross physical disturbances disrupt
sediment layering, which affects radiograph interpretation, alters
strength and consolidation properties and, potentially, affects
analyses that focus on the sediment profile (i.e., texture,
mineralogy and the suite of geochemical measurements. The
-------
fundamental goal of sampling -- to preserve in-situ integrity of
sediment mass to the highest degree possible -- can be largely
realized, however, by judiciously selecting the proper type of
sampling system.
A number of instruments have been developed for
collecting a sediment sample from the deep sea with a minimum of
disturbance. Practical considerations, however, limit selection
to box corers and large-diameter "gravity" corers, because other
systems are either too expensive (e.g., pressurized corers, and
bottom-sitting platform corers with fixed pistons), may cause too
much disturbance (e.g. grab samplers and piston corers), or have
some other problem. Box corers and gravity corers meet the
penetration and distrubance criteria, do not require an inordinate
amount of deckspace or special ship rigging, are readily available
and are relatively straightforward to operate and maintain. Box
corers that are particularly suited include the Soutar Box Corer
(Soutar et al, 1981), which features a slowed entry, and the
Hessler-Sandia MK-3. Though relatively expensive and cumbersome
compared to gravity corers, box corers offer the advantages of
large sample volume and reliability, and are unsurpassed with
regard to the quality of sample they recover. Moreover, they lend
themselves to subcoring, which is a noteworthy advantage in a
multidisciplinary activity such as LLW disposal site monitoring.
Gravity coring is quick, simple, and inexpensive. It
also offers the advantage of speed: the elapsed time between the
corer breaking the sea surface upon recovery and storage or the
start of processing of the core sample can be less than 5 minutes,
which may be an important factor for some types of geochemical
analyses. Some gravity corers even offer special features, such
as hydraulic damping during entry (Pamatmat 1971), which further
enhances the chances for high sample quality. To generally
achieve optimum sample quality, one should select a corer model
with an inside diameter of at least 5 cm. It should not require a
cutter (nose-cone), internal core catcher or liner. A Benthos
model 2171, hydroplastic corer (Richards and Keller 1961) with
external core catcher, or equivalent, is recommended. If a
gravity corer is used, it may be necessary to collect replicate
cores at stations. The actual time required for this extra coring
should not be considered a prohibitive disadvantage per se.
Further, such activity provides an opportunity to examine local
variability of the sediment because the replicates would be
recovered from different, though proximal, locations.
10
-------
Regardless of the coring system used, certain operational
caveats apply: (1) entry and withdrawal of the corer from the sea
bed must be done as slowly as possible to enhance recovery of a
high quality sample; (2) a pinger, attached to the corer, is
essential for estimating distance to the sea bed; (3) a reliable
means for stabilizing the corer when it breaks the sea surface
should be readily available to ensure that the sample is not
jerked around or banged into the ship or any of the deck gear; (4)
coring should not be attempted when the sea state is high enough
to cause difficulty in handling coring equipment or cores during
collection or recovery operations, to prematurely trigger a core
descending to the seafloor, to allow double entry (repeat
penetration) into the sea bed, or to anyway compromise core
quality; and, (5) the coring tube and liner/s/ should always be
kept clean and free from potential contaminants that could affect
the collected sample/s/. Corrosion products, such as rust, can
not only affect the results of geochemical analyses (e.g., Fe),
but they can also alter texture and mineralogy measurements.
A more complete discussion of coring and related factors,
as well as an extensive bibliography on the subject, are provided
by Lee (1985). Other pertinent references include Lee and
Clausner (1979), Rosfelder and Marshall (1967), and Hvorslev
(1949) .
Handling
Once the coring device is brought to the sea surface,
extreme care must be exercised. The core should be maneuvered
slowly to the deck or its cradle to avoid accidental bumping,
which causes physical disturbance, and should be set down as
gently as possible. Enough hands should be available to guide the
suspended corer and ensure that it does not get out of control.
Core length and penetration distance should be measured and
recorded as soon as the core is safely on board. The core should
be prepared quickly for storage (wrapped) or shipboard laboratory
work (moved to a "constant environment" location) to avoid adverse
effects (changes due to temperature, evaporation, drainage,
vibrations and other factors) from potential accidents due to
shipboard activities or the weather. Cores should always be
stored in an upright position if contained in liners. If the core
is too long to store vertically, it should be sectioned for
storage with each section capped and sealed.
11
-------
If a box corer was used, the cored sediment must be
processed immediately. Otherwise it may begin to drain. The
sediment sample should be subcored by smoothly inserting
large-diameter (16-7 cm I.D.), thin-walled (1-3 mm) tubes that
have been sharpened on the "cutting" end. It is recommended that
the tubes be made of plastic (e.g., polycarbonate) or some other
inert material to avoid chemical reactions that may alter
geochemical, or other, properties. Common core liner is usually
used. After the subcore tubes have been inserted, the outside
(box) corer may be taken apart and the rest of the original core
material removed. Once free, the subcore tubes should be cleaned
and inspected. They should always be kept vertical. If a
particular core is to be used for detailed geochemical profiling,
it may be desirable to maintain several centimeters of seawater
over the sediment to mitigate disturbance of the sediment-water
interface during handling. Alternatively, the supernatant seawater
can be siphoned off and stored for future analysis. For other
types of analyses, any space between the sediment surface and the
top of the liner should be plugged to prevent future movement of
the core in the liner or disruption of the sediment surface. The
plugs should be made of light-weight nonabsorbing material, such
as close-celled styrofoam or wax. Next, the tube should be capped
(commercial plastic or rubber end caps are suggested) and the caps
should be taped in place with several wraps of plastic electrical
tape. Care must always be taken not to invert the subcore during
handling, which can cause severe disturbance if the sample is not
"tight" in the tube or if the sediment is "soupy." The subcore
must not be tamped on the deck because tamping artificially
compacts the sediment and changes its strength, consolidation, and
water-content-related properties. In addition, its gross
structure and sensitive geochemical profiles may also be adversely
affected. All core tubes, caps, and utensils should be cleaned
carefully with appropriate acids and/or solvents before use. This
cleaning is not standardized; the reagents vary and depend on the
composition of the plastics.
Because the subcores are comparatively short, there is
less hydrostatic head and thus they are less apt to leak than
longer cores. However, surface samples are typically high in
water content and core ends should be waxed as further protection
against leakage. If the cores are not scheduled for immediate
analysis, the entire tube should be waxed to retard evaporation
through the tube walls, which, if plastic, tend to be pervious.
The tubes should be handled and stored in the in situ position and
labeled indelibly with identification information and an "up"
arrow.
The subcoring technique works well for box cores, but is
unnecessary with gravity cores. All statements made previously on
the handling of subcores apply to gravity cores if they are the
type that employ plastic pipe as a barrel (note that the two
recommended gravity coring systems are such types).
12
-------
If it is desirable to collect long cores (i.e., longer
than 11.5 meters), or for other reasons use of a core liner is
appropriate, care must be taken when the filled liners are
extruded from the core barrels. They should never be hammered or
jerked unless such action is absolutely necessary to free the
liner. If the core is to be sectioned, first cut the liner. This
should be accomplished without unnecessary vibration or pressure
on the liner because such disturbance is transmitted to the core.
This can be done by using tools especially manufactured for
cutting plastic pipe. It is suggested that such cuts be made
between clamps. Clamps hold the liner securely and prevent it
from flexing as pressure is applied. Without clamps, the portions
of the core near the cut will be disturbed unless a very rigid
liner is used. A hacksaw may also be used to cut the liner, but
it is not preferred because of the vibration caused by its use and
because it may affect texture analysis locally if plastic shards
get into the sediment. After the liner has been cut, the sediment
should be parted with a wire saw. The sections should be no more
than 1.5 m (approximately 5 ft) long, and should be cut much
shorter (approximately 0.5 m) if long-term storage (more than a
month) is anticipated. The sections should be capped, waxed, and
labeled as described above.
Whether the original cores were lined or unlined, all
core samples should be adequately protected from additional
disturbance or alteration at this stage. They may also be
x-radiographed on the ship to check for disturbance. This is
desirable because if problems are discovered, recoring is
possible. However, it is generally impractical.
Shipboard analytical work should always be seriously
considered during cruise planning. Samples intended for triaxial,
consolidation, or other complicated testing may need to be
prepared for safe transport because it is difficult to perform
these tests at sea, but there are numerous tests, such as nitrate
analysis, Eh/pH measurements, vane shear tests, and other
activities (i.e., subsampling) that should be done at sea to avoid
disturbances or alterations related to further handling and
storage of the samples. in addition, shipboard testing provides
site information while still in the field, and smaller subsamples
are more easily protected against evaporation or other changes
than whole cores or core sections.
Storage
Proper storage of the sealed sample tubes, whether at sea
or ashore, is an essential part of the pre-analysis activity. Of
paramount importance in this regard is storing the samples at the
proper temperature: all samples must be refrigerated. It may even
be desirable to freeze samples dedicated to certain geochemical
tests, particularly if shipping is required. Refrigeration
13
-------
temperature should be between 1-5 °C for general preservation,
and, because in situ temperatures at depths greater than 4000 m
are approximately 1 °C, storage at this temperature would match
in situ conditions. However, freezing of geotechnical samples
must be prevented, so care must be taken to avoid accidental
freezing if 1 °C is chosen as the storage temperature. Storage
in an air-conditioned environment is acceptable for a short period
of time, such as while at sea, if refrigeration is not available,
but is not recommended. Temperatures even at this level (15-22
°C) cause expansion and "softening" of the sediment (Mitchell
1976) , accelerate bacterial and other organic productivity, and
promote evaporation; all of which are detrimental to sample
quality. The storage area should ideally be kept at 100 percent
relative humidity as an additional safeguard against evaporation.
In fact, from core retrieval to storage, all practical safeguards
must be taken against evaporation because this process can
drastically alter many geochemical and geotechnical properties,
and even change the texture and mineralogy of authigenic minerals
(such as gypsum) which could begin to form. Sun, or other light,
that fosters organic growth should not access the area unless the
tubes or liners are opaque. For short periods of time (a few
weeks at most) strict humidity control may not be as critical.
The core tubes should be stored vertically upright with
the sediment surface at the top; the tubes should be well-padded
underneath, and tightly secured. The padding and secure fastening
are particularly important at sea because the samples need to be
isolated from high-frequency (motor-induced) and low-frequency
(motion-induced) vibrations which can cause dynamic loading and
settlement in the tubes. The core tubes should not be stacked.
Samples that are a product of shipboard subsampling and intended
for geotechnical index property testing, if properly sealed in
plastic bags and/or cans containing a wet tissue, need not be
protected so rigorously, but refrigeration is highly recommended.
Shipping
All samples should be sent to the laboratory as soon as
they reach port. Optimally, they should be hand-carried to their
destination in padded cases. If there are too many samples to
carry and long-distance shipping is unavoidable, air freight is
preferred to trucking. Air freight is faster and subjects the
samples to less vibration; storage temperature may be stipulated
within reason, but for short periods of time this is generally not
necessary. No matter what the conveyance, the tubes and other
samples should be shipped in their proper orientation and in
sturdy, well-padded boxes. The boxes should be marked "fragile"
and have "up" arrows clearly in view. Commercially available tip
indicators should be affixed to both the inside and outside of
each box so that if the tubes are inverted or on their side during
transport, it will be known. If possible, the boxes should be
constructed such that the preferred orientation is obvious.
14
-------
After they arrive at the laboratory, the samples should
be stored as discussed and the analyses should begin as soon as
possible. Storing cores for more than a few months should be
avoided because aging, which alters the properties, can occur
relatively quickly (Mitchell 1986).
Summary
The properties of marine sediments are extremely
vulnerable to alteration during sampling and subsequent activity.
Seemingly minor improprieties during field work may render part of
a derived data set worthless or, perhaps worse, leave the data
quality uncertain. Accordingly, all reasonable precautions should
be taken to avoid physical, chemical, or biological changes in the
samples, and every effort should be made to complete the analyses
of the samples as soon as possible.
In general, the following guidelines should be adhered
to: (1) samples should not be jarred, tamped, vibrated, or
pressure-loaded in any way; (2) storage temperatures should be
kept near freezing, or frozen in some cases; (3) humidity should
be as high as possible, and any circumstance that favors
evaporation or drainage should be avoided; (4) the storage area
should be free of light wavelengths that favor organic
productivity unless sample tubes are opaque; and, (5) cores should
be transported and stored as close to their in situ orientations
as possible. If these guidelines are followed, and if proper
sampling equipment has been used, the investigator may be
reasonably sure that the data are as representative of field
conditions as possible.
15
-------
References
Hvorslev, M.J., 1949. Subsurface exploration and
sampling of soils for civil engineering purposes, U.S.
Army Corps of Engineers Waterways Experiment Station,
Vicksburg, MS, 521 pp.
Lee, H.J. and Clausner, J.E., 1979. Seafloor sampling
and geotechnical parameters determination handbook, U.S.
Naval Civil Engineering Laboratory, Port Hueneme, CA,
Technical Report R-873, 128 pp.
Lee, H.J., 1985. State-of-the-art laboratory
determination of the strength of marine soils, in Chaney,
R.C., and Demars, K.R., eds. Strength testing of marine
sediments: laboratory and in situ measurements, American
Society for Testing and Materials, Special Technical
Publication 883, pp. 181-250.
Mitchell, J.K., 1976. Fundamentals of soil behavior,
John Wiley, New York, NY, 422 pp.
Mitchell, J.K., 1986. Practical problems from surprising
soil behavior, Journal of Geotechnical Engineering, vol.
112, pp. 255-289.
Pamatmat, M.M., 1971. Oxygen consumption of the seabed,
Shipboard and laboratory experiments, no. 5, Limnology
and Oceanography, vol. 16, pp. 536-550.
Richards, A.F. and Keller, G.H., 1961. A plastic-barrel
sediment corer, Deep-Sea Research, vol. 8, pp. 306-312.
Rosfelder, A.M. and Marshall, N.F., 1967. Obtaining
large, undisturbed, and oriented samples in deepwater, in
Richards, A.F., ed., Marine Geotechnique, University of
Illinois Press, Urbana, IL, pp. 243-262.
Soutar, A. et al, 1981. Sampling the sediment-water
interface; evidence of organic-rich surface layer, EOS,
vol. 62, p. 45.
16
-------
GEOLOGICAL METHODS
17
-------
CHAPTER 3
TEXTURE
Introduction
There are four main reasons for performing particle-size
analyses on sediments (Blatt et al. 1972): (1) the grain size is a
basic descriptive measure of the sediment; (2) grain size
distributions are characteristic of sediments deposited in certain
environments; (3) detailed study of the observed grain-size
distributions often yield information about the physical
mechanisms occuring during deposition or diagenesis; and, (4)
grain size can often be related to other properties (i.e.,
permeability or stability), and variations in these properties may
be predicted from variations in grain size. With this in mind,
the fundamental objectives of a grain-size analysis are to
accurately measure individual particle sizes, or hydraulic
equivalents, to determine their frequency distribution, and to
calculate a statistical description which adequately characterizes
the sample.
The accuracy of these measurements is limited by the
capability of the operator, the sampling (see Chapter 2) and
measuring techniques, and the equipment. Care and attention to
detail must be exercised to achieve the best possible results. As
with most types of analyses, there is no ultimate technique or
procedure that will produce the most desirable grain size data for
all cases. Several types of analyses have been developed over the
years to accommodate the types and sizes of samples and the
reasons for doing the analyses.
The recent development and commercial availability of
inexpensive microcomputers allows sedimentologists to construct
complete computerized particle-size analysis systems (Poppe and
others 1985). The major advantages of using these systems are the
time-, labor-, and cost-saying functions they afford. Most of the
laboratory equipment and procedures, and all of the data
processing described herein, may be adapted to these computerized
systems.
Procedure
1. Because the same analysis is not done for each
requestor and because the size data often become part of a data
base, there must be some formal system for recording what analyses
have been completed on each sample and for whom they were
performed. To this end, the use of "REQUEST FOR ANALYSIS" and
"SAMPLE ID" forms (Figs. 1 and 2) are recommended as
organizational aids. These completed forms should be included when
18
-------
SEDIMENTATION LAB
REQUEST FOR ANALYSIS
SUBMITTER:
Cruise ID Project ID
NO. OF SAMPLES:_
DATE SUBMITTED:
DATE RESULTS NEEDED BY:
WORK REQUESTED:
SPECIAL INSTRUCTIONS:
AFTER JOB IS COMPLETED:
DISCARD BULK:
RETURN TO SUBMITTER:
PURPOSE OF INVESTIGATION:
PAGE OF PAGES
Figure 1. Example of a REQUEST FOR ANALYSIS form.
19
-------
SUBMITTER
CRUISE ID_
PROJECT ID
PAGE
OF
PAGES
H-
2
0
Hi
(11
CO
w
H
O
S1
FIELD NO.
LAB NO.
LOG
DEGREES A
MINUTES (
LAT.
&TION
ND DECIMAL
N-S OR E-W)
LONG.
WATER
DEPTH
CM)
SAMPLING
DEVICE
AREA
TOP
DEPTH
(CM)
BOTTOM
DEPTH
(CM)
-------
samples are submitted for analysis. Appropriate identifers
include: requestor, cruise identification (ID), project ID, work
requested, purpose of investigation, sample ID, latitude and
longitude of collection site, water depth, top depth, bottom
depth, sample device, and sampling area.
After the forms are completed, a unique lab number should
be assigned to each sample. This number may consist of two
letters followed by three numbers (e.g., AA795), or of sequential
listing developed by the lab involved.
2. if the whole sample is not to be used, the sample is
split by one of three methods: a micro-splitter, cone and quarter,
or random bulk. The method selected for a given analysis will be
determined by the size of the sample, its homogeneity, and the
degree of accuracy that is required. As a general rule, the
larger the analyzed sample, the more accurate the grain-size
analysis. However, unless a sample contains abundant medium to
coarse gravel, a 50g split is usually sufficient. Larger samples
are more time consuming to analyze and do not provide
significantly more reliable results.
3. Place the split sample in a preweighed 100ml beaker.
These beakers should have previously been enscribed with a unique
number, therefore requiring no additional labels (which would
alter the total weight of the beaker). Dry the samples in a
convection oven at 100 °C. This temperature will drive off
unbound water without affecting the grain size. When the samples
are dry (overnight is usually sufficient), place them in a
desiccator to cool and then weigh them. Gross sample weight minus
the weight of the beaker gives the net sample weight.
4. Whole or fragmented calcite secreting micro- and
macro-organisms can bias the grain size distribution if they occur
in significantly high concentrations (greater than 5 percent).
Their presence alters the textural data and complicates
interpretation of sedimentary transport and deposition. Calcite
or aragonite may be dissolved selectively using cold, dilute (10
percent) HCl or a 5 pH, NaOAc buffer (Jackson 1956). After the
carbonate has been dissolved, the HCl or NaOAc buffer must be
removed by multiple decantations or centrifugations using a buffer
to remove the salts and distilled water to remove the buffer.
Care must be taken to use these dissolution techniques only when
detrital carbonate is absent. It is often easier to remove the
fragments of biogenic carbonate manually from the coarse fraction
(see below) than to treat the whole sample.
5. The removal of organic matter is necessary for
complete dispersion of the clay and, in sediments with high
organic content (greater than 5 percent), to prevent the organics
from being counted as part of the sample, which would bias the
21
-------
grain-size distribution. The sample is placed in a 600ml beaker
and a small volume (approximately 10 ml) of 30 percent hydrogen
peroxide is added. The sample is stirred and, if necessary, water
is added to slow the reaction and prevent bubbling over. More
hydrogen peroxide is added until the dark color of the organic
matter has largely disappeared; then the sample is washed three
times with a NaOAc buffer of pH 5 and once with methanol to remove
the remaining released cations.
6. To prevent flocculation of the sediment and to give
effective dispersion of the sediment during pipet analysis, the
soluble salts and exchangeable polyvalent cations are removed by
decantation or centrifugation with distilled water.
7. If the samples have been decalcified, treated with
hydrogen peroxide, or desalinated, repeat Step 3 and recalculate
the post-treatment net sample weight. Subtracting this value from
the original will determine the weight of the removed constituent.
8. Wet sieve the sample through a 63 micron, American
Society for Testing Materials (ASTM) #230 sieve to divide the
sample into coarse (sand and gravel) and fine (clay and silt)
fractions. A sodium hexametaphosphate solution (a 1,000ml beaker
of Calgon containing 80ml of formaldehyde for each 5 gallons of
distilled water) is recommended for wet sieving if an electronic
particle counter (e.g., Coulter Counter) is to be used in the
fine-fraction analysis. This solution acts as a dispersant and
electrolyte, which is necessary when electronic particle coulters
are used. A weaker electrolyte solution (0.5 percent) may be used
if the fine fraction is to be analyzed by pipet. A rubber spatula
and a squeeze bottle of the calgon solution provide the best
sieving results. The coarser sand and gravel fractions are
retained on the screen while the finer silt and clay fractions are
collected in the catch pan. Place the coarse fraction in a
preweighed 100ml beaker and the fine fraction in a Mason Jar. Wet
sieve, using only enough calgon solution to fit in a single 32-oz
Mason Jar to prevent any possible fractionational biasing of the
fine fraction.
9. Dry the coarse fraction in a convection oven at 100
°C. When the samples are dry (overnight is usually sufficient),
place them in a desiccator to cool before weighing them. Gross
coarse weight minus the weight of the beaker gives net coarse
weight.
10. Most of the biogenic calcite found in a sample is
commonly in the form of coarse pelecypod or gastropod shell
fragments. It is often easier and less time consuming to remove
this carbonate manually from the sample and reweigh the now
decalcified sample to determine the new net coarse weight.
Remember, the weight of the carbonate must be subtracted from the
net sample weight before entering the data into the computer.
22
-------
11. The coarse fraction is usually determined by use of a
settling tube (for a description of a typical settling tube, see
Schlee 1966). However, if this fraction weighs less than 5 grams
or if it contains greater than 5 percent foraminifera, which will
not settle properly in a sedimentation column, sieves must be
employed. If the coarse fraction grain-size distribution is to be
determined by sieving, assemble a bank of sieves (see Table 2
below). A bank of sieves generally consists of a cover, the -1,
0, 1, 2, 3, and 4 phi sieves, and a catch pan. The bank of sieves
is agitated in a shaker (such as a Ro-Tap) for at least 15
minutes. After sieving, weigh the sand and gravel fractions,
record the weights from each phi class, and calculate the relative
percents of the sand-fraction phi classes and the relative
percents of the gravel-fraction phi classes. These data, per se,
are normally entered directly into a computer or are manually
combined with the fine-fraction data to obtain an overall complete
grain-size distribution.
TABLE 2. Correlation of grain size phi classes with sieve sizes
and numbers for A.S.T.M. and Tyler mesh sieves."
PHI
CLASS
NOMINAL SIEVE
OPENING (mm)
SIEVE SIZES
ASTM # TYLER #
SIZE
FRACTION
4
3
2
1
0
.062
.125
.250
.500
1.000
230
120
60
35
18
250
115
60
32
16
SAND
_ 1
-2
-3
-4
-5
2
4
8
16
32
.000
.000
.000
.000
.000
10
5
0
0
1
.3125
.625
.25
9
5
2 . 5 GRAVEL
0.624
_
23
-------
12. If the sand-fraction grain-size distribution is to be
determined by a settling tube technique, the gravel-fraction
distribution (if present) must still be determined by sieve
analysis. Separate the dry sand and gravel fractions by using a
2.0mm (-1 phi) sieve, and weigh the gravel fraction. Net coarse
weight minus the gravel weight gives the sand weight. Assemble a
bank of gravel fraction sieves; then sieve to separate and weigh
each phi class; record the weight; and, calculate the
gravel-fraction relative percentiles.
13. The settling tube (e.g., Schlee 1966) provides a
means for rapid size analysis of sand-sized material by settling
the grains through a column of water (Fig. 3). One settling tube
design is based on using the pressure differential between two
columns of water that have a common head. The change caused by
the introduction of sediment within one of the columns is measured
by a transducer. Results are fed either to a strip chart recorder
or to an analogue-to-digital converter, and relayed to a
computer. As the sediment settles past the pressure transducer
port, the pressure differential decreases with time. Because the
sedimentation rate, in accordance with Stoke*s Law, is a function
of grain size, one can interpret the sand-fraction grain-size
distribution from the variation in pressure differential. Because
the design of this type of settling tube is common, a procedure
and comments for use of such a device are given below.
A warm-up time of about 5 minutes is required for
the transducer. During this time the operator should
check the system pressure lines for air bubbles and note
the settling column water level. Remove all air bubbles
from the lines and raise the water level to 3-5 mm above
the settling tube (see Fig. 3). The operator usually has
the option of introducing the sample by using the
automated flipper or by manually adding it with a
tablespoon (this is generally true for most automated
settling tubes). Both methods require practice to
achieve reproducibility, but the results are equally
accurate and comparable.
A micro-splitter is used to obtain a sample of
5-10 g. Optimal size is about 8 grams. Samples less
than 5 g generate too weak a signal from the pressure
transducer. A weak signal may be diluted by electrical
and mechanical noise, causing greater error in the
reproducibility of the data. Samples larger than 10 g are
more apt to form density currents down the sides of the
settling tube or, upon introduction, to form clumps of
finer sediment which act as much larger particles.
24
-------
Figure 3. Differential pressure settling tube for sand-fraction analysis.
25
-------
Whether the automated or manual technique is
employed, the sample must be dampened with just enough
water to: (a) afford enough intergrain cohesion to help
the sample adhere to the flipper; (b) insure simultaneous
introduction of the sample; and, (c) prevent air from
becoming trapped in the sample, which may slow the
settling rate of the particles.
If the automatic flipper technique is selected,
place the sample on the flipper using a tablespoon,
flatten the top of the sample to less than 1 cm in
height, and dampen the sample with an eyedropper. When
the switch that controls the flipper is shifted, the
sample will be introduced into the settling tube. When
the flipper is halfway between the vertical ?????? and
contact with the settling tube, move the switch back to
its original position. This will prevent the flipper
from crashing into the settling tube and introducing
noise into the system.
The manual spoon technique simply involves placing
the sample split into a tablespoon, dampening the sample
with an eyedropper, and dumping the sample in one smooth,
rapid motion (from a height of less than 3 cm) down the
center-of the settling tube.
The operator is encouraged to run a few standards
to check for equipment problems and practice sample
introduction. This exercise will help the operator to
produce more accurate and reproducible analytical results.
After the sample has settled to the bottom of the
tube and the pressure differential returns to
approximately zero, a hard copy of the sand fraction
grain-size distribution can be generated on a strip chart
recorder or the terminal (see Figure 4) .
14. The fine fraction (silt and clay), which has been
stored in Mason Jars of Calgon solution, may be analyzed by pipet
or Coulter Counter. The Coulter Counter determines particle
volume; the pipet method measures settling rates. The reason for
performing the analyses will determine which method will be used
(see Comments section, page 36).
The pipet method consists of preparing a dispersed (by
sonic probe), homogeneous (by vigorous stirring) suspension of the
fine fraction, diluting it with a 0.5 percent Calgon solution to
1,000 ml, and allowing the particles to settle in a graduated
cylinder (Polk 1974). The optimum sample is approximately 20
cm^ (about 15-g dry weight). With more sample, the grains
26
-------
Lab Number AC216
Operator:
Cruise I.D.
Project I.D.
Sample I.D.
Sample weight:
Coarse weight:
Sand weight:
53.8613
52.7069
50.1269
Gravel Fraction Percentages:
28.68
39.92
31.4
0
0
Wednesday 09:32:00 July 25, 1984
.5
.0
PHI
-1.
-1
-0.5
+0.0
+0.5
+ 1.0
+ 1.5
+2.0
+2.5
+3.0
+3.5
+4.0
VALUE
797
797
797
795
794
774
679
512
345
160
79
73
PERCENTAGE
0.000V.
0.0007.
0.000V.
. 2757.
1387.
. 7557.
, 0857.
23.0037.
23.0037.
25.4827.
11. 1577.
0.8267.
0.
0.
2a
13.
0.07.
0.37.
2.97.
36. 17.
48.57.
12.27.
Total percentage = 99.8 Remainder percentage = .199997
Corrected 0 Phi = .3 7. Corrected 4 Phi = 12.2 7.
********»******#***#*#**#*#****#* PLOT #*****#*###*************#*»*******»*»#**
Figure 4. Example of a raw data record produced by an automated settling tube.
27
-------
interfere with each other during settling; with less sample, the
experimental errors in weighing increase with respect to the
sample. Withdraw 20 mL samples of the suspension from levels at
5, 10, or 20 cm below the water surface at standard intervals of
time (see Table 3). Because all particles of a given equivalent
diameter (based on calculations using Stoke's Law) will have
settled below that level after a standard interval of time, the
samples should contain only finer particles. These aliquots are
either suction-mounted on preweighed filters and rinsed in
distilled water or placed in preweighed 100 ml beakers. If
preweighed beakers are used, the operator must remember to account
for the weight of the Calgon when calculating the phi fraction
weights. The pipet is rinsed with distilled water after each
extraction and the rinse water is also passed through the filter.
To begin the analysis, start the timer as soon as tha stirring rod
emerges for the last time. At the end of 20 seconds, insert the
pipet to a depth of 20 cm and withdraw exactly 20 ml. This is the
most important single step, as the subsequent analysis is based on
it, so be as accurate as possible. Continue the withdrawals at
the specified time intervals and depths. The filters are dried
and weighed and the size distribution is calculated from the
weight of sediment. The principle behind the computation is this:
if the fine sediment is uniformly distributed throughout the
entire 1,000 ml column by stirring, and exactly 20 ml is drawn at
each of the stated times, then the amount of mud in each
withdrawal is equal to 1/50 of the total amount of mud suspended
in the column at that given time and at that given depth (i.e.,
the amount of mud finer than the given diameter; all particles
coarser than the given diameter will have settled past the point
of withdrawal). The first withdrawal is made so quickly after
stirring and at such a depth that particles of all sizes are
present in suspension. Therefore, if we multiply the weight of
the first withdrawal by 50, we will obtain the weight of the
entire amount of mud in the cylinder. If we then withdraw a
sample at a settling time corresponding to a diameter of 6 phi,
and multiply by 50, we know that the product represents the number
of grams of mud still in suspension at this new time: the weight
of mud finer than 6 phi. Similiarly we can compute the weight
percent at any size and obtain an entire distribution.
Because of the length of time required to complete a
whole phi interval analysis (16 hours, 24 minutes to 10 phi at 20
°C) and because Brownian motion interferes with the settling of
less than 10-phi-sized particles, the pipet method is now used to
determine the silt/clay boundary in percent gravel-sand-silt-clay
analyses, and when a sample contains a significant amount of
material smaller than 0.4 microns in diameter. This latter
condition is important because a Coulter Counter determines a size
distribution based solely on the grain-size range it has been
calibrated to analyze (typically 0.4 to 62 microns). It is
28
-------
unfortunate that many low-energy and deep-water environments
contain significant amounts of very fine clay and colloidal-sized
material that would be ignored by the Coulter Counter.
To perform this silt/clay boundary analysis, collect
samples at 20 cm depth after 20 seconds elapsed time to determine
the concentration of sample in solution, and at a depth of 10 cm
after exactly 2 hours and 3 minutes to determine the concentration
of clay-sized (less than 4 microns) material in solution. From
the weight of sediment on these filters, the percent silt and clay
can be determined.
The time needed to perform a pipet analysis may be
reduced by placing the settling cylinders in a constant
temperature bath and lowering the viscosity of the settling medium
by increasing its temperature (see Part B of Table 3).
TABLE 3. Comparison of withdrawal time and depth tables for
temperatures of suspension maintained at 20 oC (Part A)
and at 24 oC (Part B)
PART A
PHI DIAMETER
t*i
4
5
6
7
8
9
10
WITHDRAWAL DEPTH (cm)
• * •*• •
" » « "
20
10
10
10
10
5
5
PART B
PHI DIAMETER
t*i
4
5
6
7
8
9
10
11
12
13
WITHDRAWAL DEPTH (cm)
:**:
20
10
10
10
10
10
5
5
5
5
TIME at 20° C
2 hr.,
4 hr. ,
16 hr., 24 min.
TIME at 24° C
20 sec.
1 min., 56 sec,
7 min., 44 sec,
31 min.
3 min.
6 min,
1 hr.
7 hr.
14 hr.
59 hr.
237 hr.
949 hr.
20 sec,
1 min., 45 sec,
6 min., 58 sec,
28 min.
51 min.
24 min.
50 min.
20 min.
20 min.
29
-------
15. The Coulter Counter is the most widely used particle
counter for fine-grained-sediment texture analysis (Fig. 5). It
permits easy, fast, and accurate analysis of whole phi grain-size
distributions, but, as with any precision instrument, care must be
exercised in its use. The instrument's optimum precision, as with
all the above analyses, is realized only if the operator is
conscientious in attention to detail and consistently follows
established procedure. The following procedure is provided
because of the general acceptance of the Coulter Counter. It is
by no means complete and is intended solely as a set of
guidelines. All operators are encouraged to familiarize
themselves with the complete Coulter Counter Manual to achieve the
best results.
The Coulter Counter is activated according to the
following steps: (1) the Coulter Counter and Population
Control Accessory are turned on and allowed to warm up
for 15 minutes; and (2) a noise check is made using the
dummy load.
Place a 200 micron tube on sample stand and turn
aperture slightly clockwise; set the Aperture Matching
Switch, Size Calibration Channels, and Size Calibration
Pot to the proper settings. Set the Gain switch to auto,
Active Channel switch to 14-3, Sampling switch to time,
Mode switch to volume, and Display Gain switch to XlO.
Fill a clean beaker with filtered (0.2 micron filter)
electrolyte; place the beaker on the sample stand
slightly away from the tube with the rubber spatula
stirrer centered in the bottom of the beaker; and set the
stirring motor to the proper RPM (a high constant speed
that does not cause surface turbulence). Check for
acceptable cumulative background count: the counts should
be less than 500 with the Faraday cage door closed.
Shake the sample in the Mason Jar vigorously and
use the whole-sample- mixer to completely suspend the
sample. While the sample is mixing, use an appropriate
pipet to transfer three equal subsamples from the top,
middle, and bottom of the Mason Jar to a clean beaker of
filtered electrolyte filled to a standard volume (200-300
ml). Never allow the tip of the pipet to touch the side
or bottom of the Mason Jar while subsampling. Place the
beaker on the sample stand, set stirring motor, and open
control stopcock above tube. Open auxiliary stopcock
just long enough to clear any bubbles from tube. Push
reset button. Bring concentration index meter to a
standard 0.03. If the concentration is above 0.03, add
more electrolyte to dilute to 0.03. If the concentration
is below this, add more subsamples to the beaker until
30
-------
Figure 5. A Coulter Counter electroresistance multichannel particle
size analyzer with the sample stand in a Faraday cage.
31
-------
0.03 is reached. Check that the concentration is less
than the maximum for 5 percent coincidence (less than
10,000 at manometer setting 2.0 CC, 14-3). After pushing
reset, wait 5 seconds, then push Accumulate. When the
calibration light comes on, push Stop and close the tube
stopcock. If the system is computer automated, record
the data from each channel to disk while the Scope
Display switch reads differential percent. This is
accomplished by pressing the Print/Plot button with the
Print/Plot switch set to interface. If the system is not
automated, the printer in the Population Control
Accessory (PCA) should be used to generate a hard-copy by
pressing the Print-Plot button with the Print/Plot switch
set to print. The operator should periodically check to
be sure that the PCA data match the LED readout. Remove
the beaker from the sample stand, cover it with
cellophane to avoid contamination, and reserve the sample
for the 30 micron tube analysis. Complete only enough
200 micron analyses to permit adequate time in which to
perform the 30 micron analyses on the same day.
Change to 30 micron tube; set Aperture Matching
switch, Size Calibration pot, Active Channel switch, Gain
Control, and Channel Selector switch to the propdr
settings. Place a clean beaker of electrolyte on the
sample stand and check for acceptable cumulative
background count. Do not use the stirrer motor during 30
micron tube analyses. Pour the sample saved from 200
micron analysis through a clean 20 micron micromesh sieve
into a clean electrolyte-rinsed beaker. Immediately
place beaker on the sample stand and open tube stopcock;
open auxiliary stopcock to clear bubbles from tube, and
then push reset. Push Accumulate button and stop when
the calibration light comes on. Record the data from
each channel to the computer disk. If a computer will
not be used to integrate the 200 and 30 micron data sets,
switch the gain control to Manual and match the
differential percent from channel 12 or 13 with the
differential percent from channel 5 of the 200 micron
tube analysis. Record the differential percentiles and
manually normalize the results.
NOTES;
* Always keep Faraday cage door closed during analyses.
* Make sure bubbles are cleared from tube aperture by
opening both stopcocks.
* If the tube clogs, brush the aperture opening. If the
tube is still clogged, clear tube in a distilled-water
ultrasonic bath. If this fails, replace the distilled
water with 50 percent nitric acid.
32
-------
* Always have aperture current in OFF position when not
running an analysis.
* Check AGC card weekly.
* After the analyses are completed, the sample stand,
Faraday cage, vacuum and electrolyte reservoir flasks,
and counter top should be rinsed in distilled water and
dried with paper towels.
* When the analyses are complete remove the 30 micron
tube, place the 200 micron tube on the sample stand, and
set all switches to the proper settings for a 200 micron
analysis.
16. There are innumerable ways to present the results of
textural analyses. The technician may compute the modified
frequency percentage; method of moment statistics; percentages of
gravel (greater than 2.0 mm), sand (greater than 63 microns and up
to 2 mm), silt (greater than 4 microns and up to 63 microns), and
clay (less than 4 microns); and the modified frequency percentages
for size distributions up to and including 11 phi to -5 phi (0.4
microns to 64 mm). The method of moments statistics generated
often include modal classes and frequencies; arithmetic mean,
median, and standard deviation; skewness; and kurtosis. A
histogram, a cumulative frequency plot, inclusive graphics
statistics (Folk 1974), and verbal equivalents for standard
deviation, skewness, and kurtosis may also be produced.
Because the most effective way to describe and summarize
a set of numbers is to visually inspect pictures of these numbers,
data graphics should do more than simply substitute for
statistical tables. They should convey both quantitative and
categorical information and allow the viewer to decode the
portrayed data with precision and efficiency. To do this, choose
graphical methods in textural analysis depending on whether you
are comparing changes (1) within a given sample, (2) within a
sedment core, or (3) between samples at one or more geographic
sites.
Histograms and cumulative frequency curves efficiently
show the size distribution with a given sample. A histogram is
essentially a bar graph in which the percentages for each size
grade (usually a whole phi interval) are plotted as columns (Fig.
6). The histogram is useful as a pictorial method to show sorting
and modality but cannot be used to readily determine any
statistical parameters such as the median or graphic mean. The
arithmatic ordinate, cumulative frequency curve is the most
commonly used graphical method for portraying the grain-size
distribution within a given sample (Fig. 7). The advantage of
this method is that all statistical parameters (median, mean,
standard deviation, skewness, etc.) may be calculated directly
from the curve allowing quantitative comparison between samples.
33
-------
30
LU
0 20
a: ^-u
LU
o_
g 10
LU
0 I 23456789 10 II 12
PHI SIZE
Figure 6. Histogram of a bimodal, poorly sorted, silty sand.
-------
100
90
80
E PERCENT
J\ O> -J
D 0 0
3 40
ID
1 30
o
20
10
9m ... ^
SAND^ / /
/ /
/ MEDIAN
j_ t
/
I •
/ r* CLAYEY SILT
• )~.-y
1 -A'S— "f 1 I 1 1 1 1 1 1 1 1
0 1 23456789 10 II l<
PHI SIZE
Figure 7. Arithmatic ordinate, cumulative frequency curve of a well sorted,
unimodal sand and a very poorly sorted, biraodal clayey silt.
35
-------
Moving-average plots are commonly used to show particle
size variation with depth in a sediment core (Fig. 8). The
advantages of these plots are that they are very pictorial,
combine the data from any number of samples, and distinctly show
gradual and abrupt changes in sediment texture; the disadvantage
with these plots is that the only quantitative information
conveyed is relative percentages among the gravel, sand, silt, and
clay fractions.
The triangular diagram (Shepard 1954) is recommended for
graphically comparing sample variations within and between sets of
textural data. In a triangular diagram, an equilateral triangle
is divided into smaller triangles as shown in Figure 9. It allows
expression of the texture fractions of a sediment in terms of
three components, sand at the apex, silt at the left-hand corner,
and clay at the right-hand corner. A three-dimensional diagram;
i.e., a pyramid, permits plotting four components, for instance a
gravel in addition to the other three size fractions. The
advantage of the triangular diagram is that more than one sample,
and for that matter more than one group of samples, may be readily
compared using one diagram. The disadvantage of the traingular
diagram, like that of the moving average plot, is that the only
quantitative information conveyed is relative percentages among
the sand, silt, clay, and perhaps gravel fractions.
Comments
The purpose of the analyses must be considered when
selecting which method will be used. For example, the Coulter
Counter determines particle volume, as opposed to the pipet
method, which measures settling rates. Therefore, if a researcher
is studying flow regimes and hydraulic equivalents, the pipet
method might produce more applicable data. Other
settling-velocity analysis methods used for fine-grain sizes are
the hydrometer (Buoyocoz 1928) and decantation methods. However,
these techniques are more difficult and less accurate (Folk 1974)
and thus are generally not recommended. Numerous journal articles
have compared the various techniques used for size analyses of
fine-grained suspended sediments. All operators contemplating the
use of any of these methods are strongly encouraged to familiarize
themselves with this literature. For example, Shideler (1976) and
Behrens (1978) compared the Coulter Counter with pipet
techniques. Hydrophotometers have been compared with the pipet
method by Jordan and others (1971) and with the Coulter Counter by
Swift and others (1972).
The fine-fraction considerations also apply to the coarse
fraction. The coarse fraction is usually determined by
settling-tube analysis. However, if this fraction weighs less
than 5 g or contains greater than 5 percent foraminifera, which
will not settle properly in a sedimentation column, the operator
must utilize sieves to measure nominal diameter. Earth (1984)
discusses these and other particle-size methods and considerations.
36
-------
Figure 8. Moving average plot of a core penetrating interbedded silty sands
and clayey silts.
70
25 50 75
WEIGHT PERCENT
GRAVEL
SAND
SILT
CLAY
37
-------
Figure 9. Graphical representation by a triangular diagram. Solid square
represents a silty sand composed of 65% sand, 25% silt, and 10%
clay; solid circle represents a sandy silt composed of 35% sand,
60% silt, and 5% clay; and a solid triangle showing a silty clay
composed of 15% sand, 40% silt, and 45% clay.
SAND
SILT
CLAY
38
-------
Quality Assurance
Extensive textural analysis performed on standards have
shown that the above methods, with the possible exception of the
hydrometer, will produce results with an accuracy of better than
plus or minus 10 percent (particle counter, 3 percent; sieves, 5
percent; pipet and sand-fraction settling tube, 5 percent).
However, optimum precision for these is realized only if the
operator is conscientious in attention to detail and consistently
follows established procedure. To monitor precision, the operator
should run replicates on at least every tenth sample.
As discussed earlier, the representativeness of any
grain-size distribution data set depends on the methods used to
obtain the data. These methods must be chosen in accordance with
the original purpose of the study. If it is more important to know
the actual size distribution than to know the hydraulic
equivalence, the sieve and the particle-counter analyses should be
performed rather than the settling-tube and pipet analyses.
The comparability problems between sets of data generated
by different devices are not as great as one might expect. For
example, the data produced by sieve and settling tube analyses
would be remarkably similar for a given sample because a settling
tube is usually calibrated using natural sediments sieved to known
fractions. On the other hand, pipet and hydrometer analyses will
usually produce data indicating a given sample is slightly finer
than will the data produced on a Coulter Counter or the
equivalent. This result occurs because fine-fraction particles
are generally plate shaped and do not settle as fast as the quartz
spheres upon which Stoke's Law, and therefore the pipet and
hydrometer settling analyses, is based. Also, a Coulter Counter
typically cannot be calibrated to analyze the very fine clay (less
than 0.4 microns) and colloidal portions of the size distribution.
Cost Analysis
Cost estimates are based on complete analyses (including
-5 to 11 phi size distribution; percents gravel, sand, silt, and
clay; statistics; histogram and cumulative frequency plots; and
verbal equivalents) performed on an automated system. Production
rate is 25 samples/week. The ranges in labor, and therefore,
total cost are dependent on the salary of the employee who
actually performs the analyses. The estimates shown in Table 4
assume typical salary levels for a technician, a supervisory
scientist, and the cost of using a Coulter Counter.
When other techniques are substituted, the cost/sample
changes. When sieves are used instead of the settling tube, the
cost/sample increases by $4/sample because of the increased time
39
-------
spent running the Ro-tap, weighing the phi fractions, and manually
entering the data into the computer. Similarly, utilizing a pipet
rather than the Coulter Counter would vastly increase the cost
because of the extended analysis time. Inasmuch as no more than
six or seven samples can be run concurrently because of the
frequency and resultant interference of sampling times, it would
take 16 to 17 hours to run six samples at room temperature (about
20° C) or 8-9 hours to run six samples using a constant
temperature bath (about 32° C). These settling analyses add
about three days, plus 24 hours of overtime in the case of the
room temperature scenario, to the time it takes to perform 25
analyses; in this case, the technician cost/sample total would
increase by $7 to $17. All cost estimates are based on the use of
an automated data-reduction system. If the calculations
necessary to normalize and combine the coarse and finefraction
distributions and generate the statistics are performed by hand,
the labor cost/sample would increase by about $21/sample.
All estimates are based on the analyses being performed
at a functioning sedimentation laboratory.
TABLE 4. Cost per sample estimates for a typical computerized
sedimentation laboratory.
Labor $ 20 to $35
Equipment depreciation, maintenance
and expendable materials $ 5
Overhead and Contingencies $ 20 to $30
Total 1. Coulter/settling tube $ 45
2. Coulter/sieve $ 50
3. Pipet/settling tube $ 65
4. Pipet/sieve $ 70
COST PER ANALYSIS AT BATCH RATE $ 70
40
-------
References
Earth, N.G., 1984. Modern methods of particle size
analysis, John Wiley and Sons, New York, NY, 309 pp.
Behrens, E.W., 1978. Further comparisons of grain size
distributions determined by electronic particle counting
and pipet techniques, Journal Sedimentary Petrology, vol,
48, no. 1, pp. 1213-1218.
Blatt, H., Middleton, G. and Murray, R., 1972. The
origin of sedimentary rocks, Prentice Hall Publishing
Co., Englewood Cliffs, NJ, 634 pp.
Buoyocoz G.J., 1928. The hydrometer method for studying
soils. Soil Science, vol. 25, pp. 365-369.
Folk, R.L., 1974. The petrology of sedimentary rocks,
Hemphill Publishing Co., Austin TX, 182 pp.
Jackson, M.L., 1956. Soil chemical analysis: advanced
course, Published by the author, Madison, Wis., 895 pp.
Jordan, C.F., Jr., Fryer, G.E. and Hemmen, E.H., 1971.
Size analysis of silt and clay by hydrophotometer,
Journal Sedimentary Petrology, vol. 41, no. 2, pp.
489-496.
Poppe, L.J., Eliason, A.H. and Fredericks, J.J., 1985.
APSAS: an automated particle-size analysis system, US
Geological Survey Circular 963, 77 pp.
Schlee, J., 1966. A modified Woods Hole Rapid Sediment
Analyzer, Journal Sedimentary Petrology, vol. 30, pp.
403-413.
Shepard, F.P-, 1954. Nomenclature based on
sand-silt-clay ratios, Journal Sedimentary Petrology,
vol. 24, pp. 151-158.
Shideler, G.L., 1976. A comparison of electronic
particle counting and pipet techniques in routine mud
analysis, Journal Sedimentary Petrology, vol. 46, no. 4,
pp. 1017-1025.
Swift, D.J.P., Schubel, J.R. and Sheldon, R.W., 1972.
Size analysis of fine-grained suspended sediments - a
review, Journal Sedimentary Petrology, vol. 42, no. 1,
pp. 122-134.
41
-------
CHAPTER 4
X-RAY DIFFRACTION MINERALOGY
Introduction
The petrographic examination of sediments by means of an
optical microscope is generally more accurate than X-ray powder
diffraction analysis for the sand-sized fraction, and gives
information on the genetic relationships and associations of the
minerals as well. However, quantitative analysis of silt-sized
particles by petrographic microscope is time consuming and
difficult, and is often almost impossible for clay-sized material.
Because of this, X-ray powder diffraction has become a common tool
for mineralogic investigations during the last four dacades.
Procedure
A split from each sediment sample is dried at 60° C,
mounted, and X-rayed as a randomly oriented powder (Appendix A).
The random orientation insures that the incident x-rays have an
equal chance of diffracting off the crystal lattice face of any
given mineral present in the sample. The use of a powder press to
make randomly oriented powder mounts is undesirable because
excessive force could cause preferred orientation of the
crystallite. The randomly oriented mounts are X-rayed between the
angles of 2° and 70° 20' using CuKo radiation and a scanning
rate of 2°/minute. Slower scanning rates substantially increase
analysis time and are usually not necessary if long-fine-focus or
fine-focus X-ray tubes are used.
The clay fraction from each sample split used in a
randomly oriented aggregate is separated by centrifuge (Appendix
B) and mounted as an oriented aggregate mount on a silver membrane
filter (Appendix C). These oriented aggregate mounts force the
clay minerals, usually plate-shaped phyllosilicates
(layer-silicates), to lie flat, allowing the operator to direct
the incident X-ray beam down- the c-axis of these minerals. It is
the c-axis that shows the extent of d-spacing expansion and/or
contraction indicative of certain clay minerals during the
subsequent treatments. Silver membrane filters can be used as
substrates for oriented clay samples because they are not affected
by heat or organic solvent treatments and because they can be
prepared rapidly and easily using readily available laboratory
equipment.
Each oriented aggregate mount is subjected to four
treatments to determine which clay minerals are present: air
drying, glycolation with ethylene glycol (Appendix D), heating to
400° C (Appendix E), and heating to 550° C. After each
42
-------
treatment the samples are X-rayed between the angles of 2° and
40° 20'. The individual clay mineral species are then
identified using the flow charts presented in Appendix F.
Mineral identification is useful only if some
quantitative distinction between the mineralogy of different
samples can be made. Truly quantitative evaluations are not yet
possible in complex mineral assemblages, but by utilizing the
areas and intensities of characteristic X-ray peaks, useful
semiquantitative approximations can be found. The intensity of a
mineral's characteristic X-ray diffraction peaks cannot be used as
a sole measure of its abundance because of the variations between
diffraction patterns that are caused by X-ray machine conditions
and by differences in sample thickness and the degree of preferred
orientation. in addition, different minerals, different atomic
planes within a mineral, and different samples of the same mineral
generally do not have the same capability to diffract X-rays.
Other complications arise from sample dilution by amorphous phases
invisible during X-ray diffraction analysis; a marine sample, for
example, may contain a significant siliceous microfossil component.
However, many semiquantitative methods exist and are
commonly used. The basic principle is that weight fractions can
be calculated from the peak areas and intensities, mass absorption
coefficients, densities, and constants. Some of the more common
methods are: spiking/dilution, internal standard (as described by
Klug and Alexander 1954), standardless (Zevin 1977), external
standard (Copeland and Bragg 1958, Geohner 1982 and Pawloski
1985) , and matrix flushing (Chung 1974). An adaptation of the
external standard method is usually preferred. in this method,
semiquantitative estimates of the minerals present in the randomly
oriented mounts are made by comparing at least three diffraction
maxima intensities and areas for each mineral within a sample,
with the areas and intensities recorded from a collection of
standards. Only peaks that do not represent over-lapping
diffraction maxima are chosen for the comparison. The collection
of standards should be extensive and consists of both
monomineralic and multiphase X-ray patterns. These patterns can
be generated from mineral samples purchased through any of the
scientific supply companies. A quartz standard is run daily to
monitor keV and mA drift, long term drift of the X-ray tube
(aging), and alignment problems, such as sample positioning in the
goniometer circle and slight goniometer alignment changes.
Changes in the quartz standard pattern are factored into the
collection of mineral standards. The relative percent of the
clay-mineral sample is estimated from the intensity and area of
the 4.5 angstrom peak, which is common to most clay minerals in
randomly oriented powder mounts.
During clay mineral analyses, useful comparisons may be
made from sample to sample by means of various ratios of peak
43
-------
areas. Relative percentages among the clay minerals are estimated
by a method adapted from Biscaye (1965). The method calculates
weighted peak area percentages for smectite, illite/mica,
kaolinite, chlorite, and mixed-layer smectites. The peaks and
weighting factors used are: (1) the area of the 17-angstrom
glycolated peak for smectite; (2) four times the 10-angstrom peak
area for illite/mica; and (3) twice the area of the 7-angstrom
peak for the combined total of kaolinite and chlorite. Individual
kaolinite and chlorite percents are assigned according to the
ratio of the 3.58-angstrom kaolinite and 3.54-angstrom chlorite
peaks. The weighting factor for smectite adjusted for the amount
of illite layers or chlorite layers present in the crystal
structure is used to calculate the semiquantitative estimates for
the mixed-layer smectites. All peak areas are generally computed
from the glycolated pattern.
The term illite/mica is used to refer to all 10-angstrom
minerals, except glauconite, which do not swell upon glycolation
and which are not destroyed when heated to 550° C. This term is
necessary because it is often difficult to distinguish between the
various minerals that show 10-angstrom reflections (such as
muscovite, biotite, sericite, lepidolite, illite, and phlogopite)
in normal diffraction analyses. The term smectite is used to
describe a group of swelling minerals that are represented
principally by montmorillonite, beidellite, nontronite, saponite,
hectorite, and sauconite. If necessary, the individual
illite/mica and smectite minerals may be differentiated by a
combination of diffraction and geochemical techniques. For
example, some of the illite/mica minerals and their polymorphs may
be distinguished by the location of the 060-hkl diffractions, and
montmorillonite and beidellite can be differentiated by lithium
saturation (Schultz 1969) .
A split is taken from each sample and mounted in
artificial Canada balsam (such as Piccolite or Caedex) as a smear
slide (Appendix G). These slides are used to check the
semiquantitative diffraction techniques, to detect amorphous
materials or material present in trace amounts, to detect
layer-silicate species concentrated in the sift fraction, and to
examine the biological debris. When these semiquantitative
estimates are made in conjunction with smear slides, the estimates
are generally considered to be accurate to within 10 percent of
their actual values; however, the smaller values may vary by
considerably more than this.
Comments
Silver membrane filters can be used as substrates for
oriented clay samples because they are not affected by heat or
44
-------
organic solvent treatments; they can be prepared rapidly and
easily using readily available laboratory equipment. The silver
of the substrates gives no maxima at diffraction angles less than
about 30° 20' using copper radiation. This contrasts with glass
(Croudace and Robinson 1983, Drevey 1973 and Pollastro 1982),
plastic-filter, or ceramic-tile substrates (Shaw 1965), which
produce significant amorphous or crystalline diffraction effects
between about 10° and 30° 20' and errors due to segragation
(Gibbs 1965). The high mass-absorption coefficient of silver to
copper radiation results in a lower overall background than these
other materials, but does not decrease diffraction intensitites
produced by the sample; the silver underlies, but is not mixed
with the sample.
Poppe and Hathaway (1979) have developed a method for
mounting silver substrates (Fig. 10) that permits easy
installation and removal of silver filters, presents a flat, even
sample surface in the correct position relative to the X-ray beam,
and fits automatic sample changers. Other silver filter mounting
methods include glued mounts, vacuum mounts, and top clamp
mounts. Each of these last techniques has significant
disadvantages. The glued mount is inconvenient and often leaves
the sample in a condition in which further treatments that are
important in the identification of clay minerals cannot be made.
If separate mounts are prepared, the differences in sample
thickness retained on each silver filter result in faulty surface
positions. Also, duplicate filters may show variations in
preferred orientation and thus give different diffraction
intensities. The vacuum mount is unacceptable because it can be
adapted to an automatic sample changer only with great
complexity. A mount in which the filter is held in place by
clamps on the top of the mount likewise will not consistently
reproduce an even sample surface, the vertical position of the
sample is highly variable, and the clamps tend to produce
diffraction and fluorescence effects at low goniometer angles.
An adaption of the external standard technique is
preferred. There are several important advantages to using this
peak area and intensity ratio technique as a semiquantitative
method: sample preparation time is minimal; the randomly oriented
aggregate mount need be X-rayed only once; and patterns are
generated in a reasonably short period of time (35 min.).
Internal standard methods, such as spiking/dilution,
generally require more time in which to prepare the samples and
calculate the results than do external standard methods. The
standardless technique involves complicated calculations and has a
greater inherent error than the internal or external standard
methods. Matrix flushing, which also involves complicated
45
-------
Figure 10. A metal membrane mount used for X-ray powder diffraction.
-------
calculations and increased time for sample preparation, is
applicable mainly for differentiating polymorphs and solid
solutions.
Satisfactory estimates of silica (opal) concentrations
may be made from this area of the 4.10-angstrom peak or from a
technique described by Goldberg (1958). It is sometimes desirable
to analyze the mineral suite of a more limited grain-size
fraction. When this is done, the sand, silt, and clay fractions
may be separated by sieve and settling techniques and each
fraction may be independently subjected to x-ray diffraction and
microscopic analyses.
Quality Assurance
The accuracy of any measurement is limited by the
capability of the operator, the sampling (see Chapter 2) and
measuring techniques, and the equipment. Semiquantitative X-ray
diffraction analyses of multiphase mineralic knowns have shown
that when the external standard method is used in conjunction with
an extensive mineral reference library, it will produce results
with an accuracy of better than +/- 10 percent of the amount
present. However, the accuracy decreases (owing to the counting
statistics, if nothing else) with the amount of a mineral
present. At a 1 percent true concentration, an estimate may be
off by a few hundred percent of its true concentration, resulting
in an estimate of 2 or 3 percent. The optimum precision is
realized only if the operator is conscientious in attention to
detail and consistently follows established procedure. To monitor
accuracy and precision, the operator must analyze a standard (such
as quartz) daily, occasionally rerun other monomineralogic and
multiphase standards, and run replicates on at least every tenth
sample.
Because semiquantitative results are only as good as the
strip chart trace of the X-ray diffraction pattern, care must be
taken when comparing semiquantitative X-ray analysis data sets
that have been generated using different sample preparation
techniques. For example, any method that requires gravitational
settling of the suspended clay fraction onto a glass slide or
ceramic tile, or one that does not instruct the technician to
grind the entire sample to less than 62 microns for the randomly
oriented mount, may cause partitioning within the sample and
prevent an accurate representation. For randomly oriented powder
sample preparation, a free-falling method such as the technique
described by National Bureau of Standards Monograph 25 (1971),
shown in Appendix A, is recommended to avoid the preferred
orientation often associated with other packed powder methods.
47
-------
Cost Analysis
Cost estimates are based on a complete analysis (randomly
oriented and oriented aggregate mounts and smear slides).
Production rate is typically 25 samples per week. Cost per sample
totals (Table 5) are given in ranges; exact costs would be
determined by the number of samples contracted. Labor costs, and
thus total costs, would vary according to the salary of the
employee who actually performs the analyses. Accordingly,
representative salaries for a technician and a supervisory
scientist have been factored into the estimate.
The use of glass slides as substrates for randomly
oriented mounts or as filter peel substrates for oriented
aggregate mounts would lower the expendable materials cost by $1
to $1.50 per sample. It would also, however, raise the labor cost
because the automatic sample changer could not be used. This
would require that a technician monitor the diffraction unit
instead of simultaneously working on a different phase of the
analysis. This could also limit operation of the diffractometer
to business hours and thus increase the cost per sample by $3.50
to $7.00.
The semiquantitative method chosen also affects the total
cost. The external-standard, spiking/dilution, and
internal-standard methods all require the use of a mineralogic
reference library, but the spiking/dilution and internal-standard
methods require more time for sample preparation and data
processing, resulting in a $3.50 to $8.00 cost increase per
sample. The standardless method would lower the cost per sample
slightly, despite the time needed for complicated calculations
that require extensive programming, because it eliminates the need
for an extensive reference library. Unfortunately, it is the
least accurate of the semiquantitative methods.
Numerous special techniques, such as the differentiation
of the individual illite/mica or smectite group minerals, are used
for specific research problems. These techniques invariably
increase the expendable materials and/or analysis time costs and,
therefore, the cost per sample. All estimates are based on the
assumption that analyses are performed at a functioning X-ray
diffraction laboratory.
TABLE 5. Breakdown of analysis costs for semiquantitative estimates
of mineralogical composition using X-ray powder
diffraction.
Labor $ 25
Equipment depreciation, maintenance, and
expendable materials $ 15
Cost of Overhead and Contingencies $ 35
COST PER SAMPLE AT BATCH RATE $75
48
-------
References
Biscaye, P.E., 1965. Mineralogy and sedimentation of
recent deep-sea clay in the Atlantic Ocean and adjacent
seas and oceans, Geol. Soc. Amer. Bull., vol. 76, no. 7,
pp. 803-832.
Chung, F.H., 1974. Quantitative interpretation of X-ray
diffraction patterns of mixtures - matrix-flushing method
for quantitative multicomponent analysis, Jour. Appl.
Crys., vol. 7, pp. 519-525.
Copeland, L.E. and Bragg, R.H., 1958. Quantitative
X-ray diffraction analysis: Ana. Chem., vol. 30, pp.
196-201.
Croudace, I.W. and Robinson, N.D., 1983. A simple,
rapid, and precise smear method for the preparation of
oriented clay mounts, Clay Min., pp. 337-340.
Drever, J.I., 1973. The preparation of oriented clay
mineral specimens for X-ray diffraction analyses by a
filter membrane peel technique, Amer. Min., vol. 58, pp.
553-554.
Gibbs, R.J., 1965. Error due to segragation in
quantitative clay mineral X-ray diffraction mounting
techniques: Amer. Min., vol. 50, pp. 741-751.
Goehner, R.P-, 1982. X-ray diffraction quantitative
analysis using intensity ratios and external standards,
Adv. X-ray Ana., vol. 25, pp. 309-313.
Goldberg, S.E., 1958. Determination of opal in marine
sediments: Journal of Marine Research, vol. 17, pp.
178-182.
Hathaway, J.C., 1956. Procedure for clay mineral
analyses used in the sedimentary petrology laboratory of
the US Geological Survey, Clay Min. Bull., vol. 3, pp.
8-13.
Klug, H.P. and Alexander, L.E., 1954. X-ray diffraction
procedures for polycrystalline and amorphous materials,
John Wiley and Sons, New York, NY
49
-------
References (Continued)
National Bureau of Standards Mongraph # 25, 1971.
Standard X-ray diffraction powder patterns, US Government
Printing Office, Washington, DC, Office, 3 pp.
Pollastro, R.M., 1982. A recommended procedure for the
preparation of oriented clay mineral specimens for X-ray
diffraction analyses, US Geological Survey Open-File
Report 82-71, 10 pp.
Poppe, L.J. and Hathaway, J.C., 1979. A metal membrane
mount for X-ray powder diffraction, Clays and Clay Min.,
vol. 27, no. 2, pp. 152-153.
Schultz, L., 1969. Lithium and potassium absorption,
dehydroxylation temperature, and structural water content
of aluminous smectites, Clays and Clay Min., vol.17, pp.
115-149.
Shaw, H.F., 1972. The preparation of oriented clay
mineral specimens for X-ray diffraction analysis by
suction onto a ceramic tile method, Clay Min., vol. 9,
pp. 349-350.
50
-------
CHAPTER 5
RADIOGRAPHY
Introduction
Radiographing cores of unconsolidated sediment allows the
nondestructive observation of certain textural and structural
features of the sediment by revealing density contrasts. This
"internal" view has proved invaluable in the study of marine
sediment because many of the relevant facts shown on the
radiograph are often obscured to the unaided eye. Features such
as laminae, clasts, changes in grain size, burrows, and shell
fragments are commonly observed in x-radiographs; and in some
cases, the degree of bioturbation can be estimated.
X-rays pass through limited masses of solid matter (e.g.,
sediment) and strike x-ray sensitive film, where the differential
absorption of X-rays -- which is caused by different densities
within the sediment — is recorded.
The energy of the X-ray produced by the X-ray tube is
expressed as units of kilovolts potential (kvp). X-rays that have
greater energy (harder X-rays) have a shorter wavelength and are
capable of greater penetration. A radiograph produced with hard
X-rays will have relatively little contrast compared to one made
with less energetic X-rays. Choice of energy is often a matter of
experimentation and depends on the thickness of the sample, its
density, and on its composition. For example, sediment with a
high calcium content will require greater energies for proper
exposures.
Resolution of detail and contrast in the radiograph are
controlled by the energy of the X-ray, the type of film used, the
exposure time, and the number of X-rays produced by the source
(Haublin 1971).
A variety of X-ray sensitive films is available, but for
sediment radiography, a fine-grained film that provides high
contrast is desirable. A typical industrial x-ray film that meets
this criterion is Dupont Cronex NDT55. Film should be used
without lead screen in pack, a common practice in industrial
radiography.
The intensity of the beam of X-rays governs the exposure
time. For example, if a good exposure was obtained at 4 mA fo^ 2
min. an exposure of equivalent quality would be produced at 1 mA
for 8 min. intensity is also a function of the distance of the
source to the plane of the film, as X-rays follow the inverse
square law.
51
-------
The equipment and facility must meet the various safety
requirements for users of X-rays. Proper shielding and remote
control of the X-ray unit are crucial.
Procedure
Most industrial or medical x-ray generating units can be
used for sediment radiography. Typically these units can (1)
generate energies of up to 150 kvp (which may be required for
thick samples), although 100 kvp is generally adequate; (2)
develop intensities greater than 8 mA; (3) accomodate samples
longer than 30 cm (the minimum length of cores for site
monitoring); and, (4) permit the samples in both vertical and
horizontal orientations.
Because sediment cores are typically cylindrical, a
radiograph that properly exposes the center of the core will
overexpose its edges. To eliminate this effect, use a plexiglass
box that is just large enough to hold the core. With the core in
the box, the remaining space can be filled with a very fine
grained powder that closely matches the composition of the core.
Powdered clay-mineral products such as a drilling mud or a
beneficiated bentonite product of Georgia Kaolin Company
(Elizabeth, NJ) or Englehart Industries (Carteret, NJ) have been
used successfully. This combination provides a consistent
thickness of material through which X-rays pass, giving a more
uniform exposure.
The core to be radiographed should not be frozen because
ice crystals may show in the radiograph. If possible, water
should be drained from the core to produce a sharper image (it may
be worthwhile to collect an additional core or subcore if drainage
seems necessary). Place the core in the plexiglass box, then fill
the box with clay, making sure there are no voids. Position the
box directly in front of the X-ray tube aperture. The sample
should be a distance of about 1 m from the X-ray source. The film
packet can be taped directly to the back of the box. Small lead
letters can be taped directly to the front of the film packet,
identifying the sample and providing orientation of the radiograph.
Considerable trial and error is often required to obtain
the optimum combination of voltage, amperage, and exposure time
needed to produce the best radiograph. If experimentation is
needed, begin by using the lowest energy (kvp) possible because
this enhances contrast relative to the "harder" X rays. As a
frame of reference, the following data are provided: the subject
is a 6.5-cm diameter core, containing overlying water, positioned
in a plexiglass box full of powdered clay.
Amperage: 6 mA
Voltage: 130 kvp
52
-------
Time: 3 min
Film: Dupont Cronex NDT-55
Distance: 42 in
Details on exposure are available from film
manufacturers. Film should be developed as the manufacturer
specifies. Make sure that the developer and other chemicals are
fresh or replenished. if an automatic developing system is used,
be sure that all rollers have been thoroughly cleaned before
beginning. Poor developing technique can cause mottling of the
radiograph.
Interpretation of radiographs is qualitative, but
radiography does allow observation of many sedimentological
features that are not visible normally. A useful overview of
radiograph interpretation is given by Bouma (1964). When
concerned with biogenic structures, interpretations from
paleontology often are helpful (Seilacher 1964). Interpretations
of primary sedimentary structures that can be observed in
radiographs are given in Middleton (1965).
Comments
If keeping a core specimen intact is unncessary and the
sediment is relatively firm, the core can be split lengthwise or
cut into tabular sections ("slabs") a few centimeters thick.
Preparing the sample in such a manner often results in a
radiograph that shows more detail and can eliminate the need for
clay fill around the core.
Longer radiographs can be produced if the x-ray head or
the core is mounted on a track. This configuration permits a
flow-type print to be produced, and may eliminate the need to
section cores.
Radiographs have been successfully taken at sea. This
capability can provide valuable site information while in the
field and help evaluate core quality as well.
Quality Assurance
Judging the quality of radiographs is subjective and,
because the procedure is nondestructive, the number of attempts
required to produce a suitable product is largely up to the
investigator. There are, therefore, no standards or established
criteria for evaluation of the product.
Cost Analysis
Time required for analysis 8 hours
COST PER RADIOGRAPH AT BATCH RATE $ 300
53
-------
References
Bouma, A. H., 1964. Notes on x-ray interpretation of
marine sediments, Marine Geology, vol. 2, pp. 278-309.
Haublin, W. K., 1971. X-ray photography, iri Carver, R.
E., ed., Procedures in Sedimentary Petrology,
Wiley-Interscience, New York, NY.
Middleton, G. V., ed., 1965. Primary sedimentary
structures and their hydrodynamic interpretation, Society
of Economic Paleontologists and Mineralogists Special
Publication 12, 265 pp.
Seilacher, A., 1964. Bogenic sedimentary structures, in.
Imbrie, J. and Newell, N. D., eds., Approaches to
Paleoecology, John Wiley and Sons, New York, NY, pp.
296-315.
54
-------
GEOCHEMICAL METHODS
55
-------
CHAPTER 6
DISTRIBUTION RATIOS (Rd)
Introduction
The distribution coefficient is a measure of sorptive
behavior that is useful in characterizing transport of
radionuclides between bottom sediments and overlying fluids in
natural water bodies (Pietrzak and others 1981). It estimates the
partitioning coefficient between ions associated with solid
particles in a porous medium (sediments) and the same ions
associated with the liquid phase, based largely on surface
adsorption. It will yield useful comparative information
primarily for ions of type A in Table 6. These are ions that are
relatively insensitive to oxidation/reduction conditions in the
sediment, and are most likely to behave similarly under both field
and laboratory conditions.
The method is a short-term batch sorption technique based
on recommendations in Relyea (1980). Results are described in
terms of R^, the distribution ratio, or apparent distribution
coefficient, rather than the distribution coefficient (K^) . The
use of Kd implies presence of chemical equilibrium, which may
not be present, and/or may be difficult to document in the natural
systems in question. To the maximum extent possible, the test
should bracket certain parameters of the actual disposal
environment, such as solution concentrations, temperature, and pH.
The distribution ratio (R^) for a specific chemical
species is defined as:
Rd = cs/cl' where
Cs = mass of solute on the solid phase per unit mass of solid
phase, and
GI = mass of solute in solution per unit volume of liquid phase.
For radioactive species, Cs equals the activity of solute on the
solid phase per unit mass of solid phase, and C]_ equals the
activity of solute in solution per unit volume of liquid phase.
R<3 is often expressed in units of milliliters of solution per
gram of solid (sediment) under steady-state conditions. These
conditions are defined here as meaning that R^ values obtained
for three separate samples of sediment exposed to the contact
liquid for 3 to 14 days shall not differ by more than the expected
precision, other conditions remaining constant. Rd and Kd are
defined identically under equilibrium conditions; i.e., K^ =
Rd.
56
-------
Table 6. Selected radionuclides (beta and gamma emitters)
potentially associated with nuclear reactor waste (partly
from Bowen 1979 and Burraan 1986).
Isotope
Ag-110
Am-241
Co-60
Ce-Pr-144
Cs-134
Cs-137
Cm-243
Eu-155
Fe-55
H-3
1-129
1-131
Kr-85
Nb-95
P-32
Pm-147
Pu-239
Pu-240
Ru-106
Sb-125
Sm-151
Sr-Y-90
Tc-99
Zn-65
Zr-95
T 1/2 (yr)
0.69
433
5.1
0.78
2.1
30
32
4.9
2.7
12.3
16.0 x 106
0.022
10.3
0.1
0.04
2.6
24,100
6,570
1
2.7
90
28.8
2.1 x ID'
0.67
0.18
Category
D
A
C-D
C
A
A
A
C
C
A
B
B
G
C
A
A
C
C
A
C
A
A
C
D
A
Typical
Redox Reactions
RE analog
Co (II-III),
Ce (III-IV)
RE analog
Eu (II-III)
Fe (II-III) , S
I- - 104 -,
organics
I' - I04 -,
organics
Nb (IV-VI)
P04 -,
organics
RE analog
Pu (III-IV-V-VI)
Pu (III-IV-V-VI)
Sb (III-V), S, H
RE
Tc (II-IV-VII)
S
Key to Table
A = elements relatively unaffected by redox reactions
B = halogens (iodine)
C = transition and redox-active metals
D = heavy metals not directly subject to common reduction
processes, but subject to precipitation or complexation
by sulfide or other reduced phases
G = gas phase
T 1/2 = the half-life of each isotope listed (in years)
RE = rare earth
S = sulfide
H = an organic biomethyl group-former.
57
-------
Procedure
R<3 analysis should begin with samples that are fully
representative of the in-place sediments, including both organic
and inorganic components where the former are significant. Care
should be taken to preserve natural moisture content and, whenever
possible, the experiments should be conducted as soon as possible
after sediment recovery. This will minimize, but not eliminate,
phase changes that may occur during storage, such as reduction of
iron oxides, reduction of pore water sulfates to hydrogen sulfide
with ensuing precipitation of metal sulfides, or oxidation of
naturally occurring sulfides to sulfuric acid. Freezing in sealed
plastic containers that allow for expansion room is considered
optimal storage; if this cannot be done, sealed storage at 4 °C
is preferred to room temperature storage.
Effective diffusion coefficients for dissolved ions in
the sediments must be known as well as R^ values before
prediction of radionuclide transport can be attempted. Other
background information, including mineralogy, solid-phase
chemistry with special emphasis on Fe and Mn concentration, and
geochemical environment, possibly including pore water chemistry,
is recommended to be conducted. The specific parameters that are
most important will vary with each radionuclide.
To assess the effect of organic matter on the R^
values, you may have to perform parallel measurements of R
-------
greater forces or longer time. Allow the fourth wash to
remain in contact with the sediment for at least 24 h
with occasional stirring. Centrifuge and decant the
liquid as before. Save the fourth wash solution as an
"equilibrated contact solution."
3. Prepare a stock radioactive tracer solution by first
evaporating the "as received" radioactive tracer
solution. Wet the residue with 6M HC1 and allow the
solution to evaporate again. Add the contact solution
(filtered seawater should be used for marine samples) to
dissolve the tracer and allow this solution to stand for
1 week. Filter the solution using a 0.45 micron
polycarbonate membrane filter.
4. Determine pH in a sample of the contact solution plus
tracer.
5. Place 5 to 25 g of wet sediment into a centrifuge
tube or bottle. Add the tracer-spiked contact solution
and mix thoroughly. The volume of the solution should be
seven times the weight of the sediment. It is strongly
recommended that the procedure include evaluation of the
adsorption isotherm by making several runs with different
concentrations and ratios of solution to sediment.
Experiments have shown that R^ will vary with the
solution/sediment ratio (Traves 1978, Pietrzak and Dayal
1981). Analysis of R^ at several tracer concentrations
is recommended.
6. Run each sample set in triplicate. Shake the
contents of each tube on a laboratory shaker for at least
6 h at 3 day intervals. To demonstrate that a steady
state has been achieved, you shall have contact periods
of 3 to 14 days. Constancy will be observed during the
2nd to 5th shaking-observation cycles. If no steady
state is achieved in this time a longer period may be
employed.
7. Prepare blank samples by adding aliquots of the stock
tracer solution to the contact solution in quantities
identical to those used for the R^ determinations. The
blanks should stand in similar bottles as the the
sediments for the same period of time, and should be
filtered in the same manner as the samples. The blanks
are used to determine the original activity in the
tracer-spiked contact solutions as well as the plateout
on apparatus. Blanks should be prepared in triplicate.
8. Measure and report the pH of the mixtures.
59
-------
9. Centrifuge each sample for 20 min at a minimum of
1,400 g in a controlled temperature centrifuge. Separate
the liquid and solid phases.
10. If filtration is necessary or desirable, use a 0.45
micron polycarbonate membrane filter. Check for presence
of species of interest in the filters by the appropriate
analytical method (e.g. after ashing in a platinum
foil-lined, porcelain crucible). Check for the
possibility of tracer adsorption by filtering the
original tracer-spiked pre-equilibrated solution and
analyzing the filtrate.
11. For each shaking/observation cycle, take an aliquot
from the supernate and determine the concentration of the
species of interest by the appropriate analytical method.
The formula for calculating the distribution ratio, R^, for
radioactive species is given by:
AS x Vi
Ws x AI
where
As = activity of the sediment phase (found by subtracting AI
from the activity of the blank) ,
Ws = weight of the sediment phase corrected for moisture,
AI = activity of the solution equilibrated with the sediment, and
YI = volume of the solution equilibrated with the sediment.
Discussion
This test method applies to porous media in which
adsorption and ion exchange are the principal applicable processes
affecting radioactive ion distribution between solid and liquid
phases. Other processes affecting ion behavior are complex
formation (especially with organic matter), precipitation, and
co-precipitation. Biomethylation of specific phases (e.g., I, Sb)
may solubilize components (Craig 1980) .
Although the test is not intended for redox-active
species, reactions related to changes in organic-rich, reducing
sediments may indirectly affect uptake or release of specific
radionuclides through change in physicochemical conditions,
sorption, complexation and decomplexation. Special precautions
may be required for given tracer solutions. For example,
evaporation of stock tracers to dryness may volatilize iodine, and
render plutonium insoluble.
60
-------
The effective diffusion coefficient, governed by the
tortuosity of the fluid pathway through the sediment, can be
estimated for nonreactive ions by resistivity methods (Manheim and
Waterman 1974, Andrews and Bennett 1981), or by the tracer-implant
method (see Sayles 1984). An important additional factor
affecting ion transport in sediments is bioturbation. For a
treatment of this phenomenon and chemical models describing it,
see Aller (1982).
Measurements of pH are useful to characterize conditions
under which the laboratory measurements are carried out, but pH
values on sediments recovered from marine environments are usually
unreliable because of outgassing during sediment recovery. In
situ pH values can be better computed from interstitial water
measurements (Manheim and Schug 1978).
Quality Assurance
The estimated overall precision of measurement is 10 to
25 percent in terms of a coefficient of variation for R^ values
in the range of 1 to 400 mL/g. Error analysis may be treated by
standard statistical methods (ASTM 1976). More direct control on
ion transport measurements in sediments is through in situ
implanted tracer experiments (Sayles 1984). These and appropriate
porefluid studies also offer a means of evaluation of labile or
redox-active radionuclide transport in recent marine and nonmarine
sediments.
Cost Analysis
Time required for batch R^ determination 16 hours
COST PER ANALYSIS AT BATCH RATE $ 1,000
References
Aller, R.C., 1982. The effects of macrobenthos on
chemical properties of marine sediment and overlying
water, iri McCall, P.L. and Tevesz, M.J.S. (eds) ,
Animal-sediment relations, Plenum Publ. Corporation, pp.
53-102.
Andrews, D. and Bennett, A., 1981. Measurement of
diffusivity near the sediment-water interface with a
fine-scale resistivity probe, Geochimica et Cosmochimica
Acta, vol. 45, pp. 2169-2175.
ASTM Committee on Standard Statistical Methods, 1976.
Manual on presentation of data and control chart
analysis, ASTM Special Technical Publication, STD 15D,
04-015040-34.
61
-------
References (continued)
Bowen, H.J.M., 1979. Environmental chemistry of the
elements, Academic Press, London, NY, 333pp.
Burman, S., 1986. Ocean dumping of nuclear wastes, Man
and development, vol. 8, pp. 147-171.
Hathaway, J.C. et al, 1979. US Geological Survey core
drilling on the Atlantic shelf, Science, vol. 206, pp.
515-527.
Craig, P.J., 1980. Metal cycles and biological
methylation, iji Hutzinger, 0. (ed.), Handbook of
Environmental Chemistry, vol.1, Part A, Springer Verlag,
NY, pp. 169-227-
Manheim, F.T. and Waterman, L.S., 1974. Diffusimetry
(diffusion coefficient estimation) on sediment cores by
resistivity probe, Intitial Reports of the Deep-Sea
Drilling Project, vol. 22, pp. 663-670.
Manheim, F.T. and Schug, D.M., 1978. Interstitial waters
of Black Sea cores, Initial Reports of the Deep Sea
Drilling Project, vol. 42B, pp. 637-651.
Pietrzak, R.M. and Dayal, R., 1981. Radionuclide
sorption isotherms for materials from the Barnwell
Disposal Site, _in Evaluation of isotope migration-land
burial and water chemistry at commercially operated
low-level radioactive burial sites, Quarterly Progress
Report, NUREG-CR-2192, BNL-NUREG-51409, vol. 1, 35 pp.
Relyea, J.F., Serne, R.J. and Rai, D., 1980, Proposed
standard batch K^ procedure, in Appendix A - Method for
determining radionuclide retardation factors, Status
Report Battelle Pacific Northwest Laboratory Report
PNL-3349, 15 pp.
Sayles, F.L., 1984. Geochemical experiments on ISHTI, ir\
Brush, L.H. (ed), Subseabed disposal program, chemical
response studies, Annual Report October 1981 - September
1982, Sandia National Laboratories SSAND 82/2713, pp.
291-316.
Traves, C.C., 1978. Mathematical description of
adsorption and transport of reactive solutes in soil, ir\
A review of selected literature, Oak Ridge National
Laboratory Report ORNL - 5403, 67 pp.
62
-------
CHAPTER 7
IRON AND MANGANESE
Introduction
The following procedures for determining the
concentrations of Fe and Mn in marine sediments are taken from a
general analytical scheme which is appropriate for subsequent
analysis of other metals. The procedures presented were developed
for analysis of Al, Ba, Cd, Cr, Cu, Fe, Pb, Mn, Hg, Ni, V, and Zn
in marine sediments from the continental margin off the eastern
United States (Bothner and others 1984).
The methodology is appropriate for sediments having a
wide range of grain size. With a carbonate-rich sample, slowly
add the acid to avoid sample loss with the effervescence caused by
the generation of C02- The low drying temperature used in our
procedures minimizes the loss of volatile forms of mercury that
may be present.
Profiles of Fe and Mn can be used to locate the boundary
between oxidizing and reducing conditions. Chapters 8 and 9 also
present methods for locating this boundary.
Procedure
An aliquot from a well-homogenized original sample is
dried at 70 °C for 24-48 h until constant weight is achieved.
The dried sample is ground in an agate grinder to avoid the
contamination introduced by some other grinding materials (cobalt,
for example, from tungsten carbide grinders).
Two blanks containing all reagents are analyzed along
with the samples. All reagents are analyzed for contaminants
prior to use, as is always necessary. The Canadian marine
sediment standard MESS-1 (Marine Analytical Chemistry Standards
Program, National Research Council, Canada) and the US Geological
Survey marine sediment standard MAG-1 (Flanagan 1976) are analyzed
in each set of samples. A series of solution standards is
prepared which approximates the concentration levels expected in
the samples; this series is used as the standard in calibrating
the inductively coupled plasma (ICP) spectrometer.
Exactly 0.5 g of ground bulk marine sediment is added to
a covered teflon beaker and digested overnight (at least 8 h) with
5 ml of 72 percent HCI04, 5 ml of HN03, and 143 ml of HF at
approximately 140 °C. The next day, the covers are removed and
the temperature is increased to betweeen 180 and 190 °C, first
producing fumes of HCI04 and then evaporating the solution to
dryness. Care should be taken that easily oxidizable organic
matter is already consumed before the drying step. The residue is
63
-------
dissolved and diluted to exactly 25 ml with 8 N HCl. This
solution is referred to as stock solution A.
The concentration of Fe and Mn is determined by
inductively coupled plasma (ICP) spectrometry by using 1 ml of
stock solution A diluted to 10 ml with distilled H20. The
instrument used is a Jarrel Ash Plasma Direct Reader, model 1160.
The argon plasma torch is maintained by an induction field having
a frequency of 27.1 MHz. Aerosols aspirated into the plasma are
ionized and the characteristic radiation of the elements are
measured spectrophotometrically. Instrument conditions: argon
gas, 20 1/min; wavelength for Fe, 259.9 nm; wavelength for Mn,
257.6 nm; forward power, 1.1 kw; fixed cross flow nebulizer;
spectral band width, 0.036 nm; observation height, 16 mm; sample
flow rate, 0.8 ml/min.
For Fe, the determination limit for this procedure is 50
ug/g in the bulk sample. The average blank, as measured in
solution, is 0.02 ug/g.
The determination limit for Mn is 10 ug/g in the bulk
sample. The average blank, as measured in solution, is 0.006 ug/g.
Atomic absorption spectrophotometry is an alternative
method for the analysis of digested marine sediment (Buckley and
Cranston 1971).
Quality Assurance
Procedures for quality assurance typically include
replicate analyses of sediment standards and comparison of results
with certified values. The precision of Fe and Mn analyses, as
determined by the coefficient of variation for 12 replicates of
standard sediments, is approximately within 4 percent. The
accuracy of the methods is determined by comparing absolute values
determined with published "accepted values." For 12 replicates
of the standard MESS-1, measured Fe was equal to 2.92 +/- 0.04
percent. The published value is 3.0 +/- 0.2 percent^-. Mn was
measured at 476 +/- 18 ppm; the published value is 513 +/- 25
ppml. For five replicates of the USGS marine sediment standard
MAG-1, Fe was measured at 4.7 +/- 0.1 percent and Mn at 721 +/- 11
ppm. "Best" accepted values are 4.9 percent for Fe and 710 ppm
for Mn (Bothner and others 1984) .
1 Values reported by the Marine Analytical Chemistry Standards
Program, National Research Council, Canada.
64
-------
Cost Analysis
The following estimated cost analysis of Fe and Mn
presumes an adequately equipped laboratory and large (greater than
40) batches:
COST PER SAMPLE AT BATCH RATE $ 50*
Additional elements can be analyzed by TCP without further
separation at $ 7 per element.
Atomic absorption spectrophotometry (AAS) is another
method commonly used to determine the concentrations of metals in
digested sediment samples. The advantage of the ICP
instrumentation is that a number of elements can be determined in
the sample solutions simultaneously and, in some cases, with
greater precision at low element concentrations. If both types of
instrumentation are available, the use of ICP for determination of
Fe and Mn is more cost effective than the use of AAS.
References
Bothner, M. H. et al, 1984. The Georges Bank Monitoring
Program 1983: Analysis of trace metals in bottom
sediments, US Geological Survey Circular 915, 36 pp.
Buckley, D. F. and Cranston, R. E., 1971. Atomic
absorption analysis of 18 elements from a single
decomposition of aluminosilicate, Chemical Geology, vol.
7, pp. 273-284.
Flanagan, F. J., 1976. Description and analyses of eight
new USGS rock standards, US Geological Survey
Professional Paper 840, 188 pp.
65
-------
CHAPTER 8
NITRATE
Introduction
The objective of this method is to present techniques for
determining interstitial nitrate profiles in marine sediments.
This parameter permits assessment of the degree of oxidation in
marine sediments in proposed or existing low-level nuclear waste
disposal sites. The rationale is that the depth where
interstitial nitrate concentrations approach zero marks the level
below which redox-active radionuclides undergo major changes in
solubility. Some elements like Fe, Mn, Ce and Co may become more
mobile, whereas others like U, Pb and Cu may become less mobile as
a consequence of forming sulfides or other less soluble
compounds. Because nitrate is a highly labile component, careful
attention to sampling (see Chapter 2) and pore fluid extraction
technique is required. Two in situ techniques that require no
separate sediment sampling are in current use: (1) rn situ probes
(Sayles and others 1976), and (2) "peeper" techniques that allow
ions to migrate from the sediment across membranes into the
sampler (Kepkay and others 1981). However, the methods and
equipment are expensive, not commercially available, and do not
easily provide detailed interstitial gradients at intervals less
than a centimeter. Therefore, these approaches are not
recommended unless the laboratory doing the monitoring is already
familiar with them. The method proposed here involves fluid
extraction from a core sample and nitrate analysis. These two
steps will be treated in the remainder of this chapter.
Porewater Extraction
Once an acceptable core is at hand, there are two ways to
extract porewaters from slices of sediment: centrifugation and
squeezing. Ordinary contaqt between porewaters and the apparatus
used for these two techniques does not appear to alter
interstitial nitrate concentrations. There is, however, a general
source of contamination that must be avoided. Exposure of
reducing sediment to air can cause significant increases in
interstitial nitrate levels (Fanning and Maynard-Hensley 1980).
It can even make nitrate appear when none was present. This
limitation affects the choice of extraction apparatus and the
length of storage of samples, which should not be for more than a
few hours (for a more complete discussion of handling and storage,
see Chapter 2). Porewater extraction by centrifugation is
simple. Sediment is extruded and sealed in centrifuge tubes to
await centrifugation. Typical centrifuging conditions are 15,000
rpm for 3-4 min (Froelich and others 1983). Thus 15 to 20
sediment samples can be processed per hour. Fifty cm3 of
sediment can be put in each tube so reasonably large porewater
samples are obtained. Extruded pore waters will have some
66
-------
turbidity and should be membrane-filtered. This is done by
expelling pore fluid from a Luerlok syringe (disposable plastic
syringe whose tip locks onto a filter device) through a Swinney
filter attachment containing 1-cm diameter 0.4-u Nuclepore
membranes. The principal disadvantage of centrifugation is that a
suitable centrifuge such as a Sorvall SS3 is difficult to use at
sea; gimbaled platforms are necessary. The alternative is to
store the sediment in centrifuge tubes and spin them back on
land. However, sediment should not be stored in this fashion more
than a few hours, and logistics will rarely accomodate this
requirement. Transfer of core sections to containers flushed with
inert gas may be possible but still poses the risk that hard-won
samples may yield uncertain data. Therefore, for most monitoring
operations, sediment squeezing is the recommended method of
porewater extraction, even though squeezers appropriate to process
large numbers of samples are difficult to obtain commercially.
The recommended squeezer is plastic (Nylon or Delrin)
with a gas-driven membrane that presses whole sediment against a
filter to express porewater. For most unconsolidated marine
sediment, a high-pressure stainless-steel squeezer (Manheim and
Gieskes 1984) is not needed. There are two principal types of
plastic membrane squeezers. One is the Reeburgh (1967) squeezer
which is held together by a metal C-clamp. The second is the
threaded type (Fig. 11) in which the top screws into the bottom.
Although threaded squeezers are more complicatad in construction,
they have operational advantages over Raeburgh squeazers. Metal
clamps corrode, and Reeburgh squeezers are difficult to clean.
The threaded types, suited for multi-unit deployment on a gas
manifold, are depicted in Figure 11, and are the basis of this
method. The plastic top is screwed in and out of the base with a
question mark-shaped wrench. The stem of the question mark is the
handle, and a small rod protrudes from the internal curved portion
near the top. That rod fits into horizontal holes drilled into
the squeezer top. Ten threaded squeezers may be operated at once
inside a glove bag. Each has a Swagelok quick-connect fitting in
its top to deliver squeezing gas (N in Fig. 11) and a Clippard
3-way pneumatic valve to allow it to be operated independently of
the rest. Cost for building a threaded squeezer, including tubing
and fittings, is modest. The diameter of the filter in Figure 11
is 9 cm, equalling standard laboratory filter paper size. To be
absolutely certain that this source of error does not occur, one
should process all sediment inside a glove box or bag containing
an inert atmosphere (N2, Ar, or He). However, such an approach
is tedious and slow, and an alternative is to let sediment color
be the guide (Fanning and Maynard-Hensley 1980, Lyle 1983). If
all of a sediment core inside the liner appears to be brown or
tan, there will probably be a negligible air artifact, and the
sediment can be sliced and centrifuged or squeezed in air. If the
core seems to have zones which are green, gray, or black, then
67
-------
Figure 11. Gas-displacement squeezer for use with multisqueezer mani-
fold. A = threaded squeezer top; B = squeezer bottom with
inner taper (G); C = rubber membrane; D = sealing ring with
tapered edge (G); E = nylon screw; F = inner ring with two
tapered edges (G); G = sealing tapers; H = paper filter,
/ = Nucleopore membrane; J = exit hole; K = sediment
sample; L = plastic covering on sediment; M = locking notch
for tightening top.
68
-------
the sediment is probably reducing and a glove bag or box is
necessary. Note that sediment tends to streak along the liner
wall during penetration, which may mask the color of the bulk of
the sediment at a given level. If sediment from an area is being
sampled extensively, tests can be run on some initial cores to
determine whether or not an inert atmosphere is required.
Many varieties of glove boxes or bags are available,
ranging in price from tens (disposable plastic) to thousands of
dollars. All of them present common problems as may be noted in
Figure 12. First, the core liner or subcore tube (A) must be
transferred into the bag or box in such a way that air is kept
out. An effective seal is flexible rubber sheeting (B) held by a
Plastic ring (C). The next problem is sediment extrusion. After
the core liner is sealed and an inert atmosphere established
around it, the core cap may be removed and the supernatant water
sucked off. Then the bottom core cap may be removed, and a rubber
stopper (D) inserted to serve as a piston. The stopper and
sediment remain in place because of the rod (E) and floor mounting
(F), while the operator's surgically-gloved hands through the
ports (G) move the liner downward to extrude sediment for
slicing. The coupling (H) is necessary to allow changing to
different lengths of rods as the length of unprocessed sediment
shortens during the slicing. A third problem is determining when
atmospheric oxygen has been reduced to a safe level inside the bag
or box. Some bags require positive internal pressure to avoid
collapse. The inflation process (with inert gas) will lower the
oxygen concentration considerably, and opening a small exhaust
port and continuing to sweep the gases out of the bag for 20-30
min before opening the core will lower it even further. A
convenient test is to check whether or not a cigarette lighter can
be made to light. This test should be performed periodically as a
core is processed. A fourth problem is getting a rack of
squeezers (I) inside a glove bag or box. A rack that contains 10
squeezers such as those shown in Figure 12 is at least 2 ft high
and at least 3 ft wide. Many glove bags do not have entry ports
large enough to accommodate such a structure, and many glove boxes
are not tall enough.
1. Exit port. The exit port (J, Fig. 11) is designed to
admit the tip of a Luerlok syringe. Holes are drilled in
caps from 1-oz Nalgene polyethylene bottles. Hole size
matches the tips of disposable syringes. Disposable
syringes are cut off at the upper part of the barrel and
inserted into the exit port of each squeezer so that the
cap of the 1 oz bottles is held against the bottom of the
squeezer. Polyethylene bottles to receive the expressed
pore water can then be screwed to the caps beneath the
squeezers.
69
-------
Figure 12. Schematic diagram of an inert-atmosphere system for
extruding, slicing, and squeezing sediment. Labeled
components are discussed in -the text.
-------
2. Installation of the filters. Final filtering of
expressed porewaters is automatic because the central
well in the squeezer is machined to fit a 47 mm 0.4-u
Nuclepore membrane (I). in the well is placed a Whatman
42 (or equivalant) paper filter (H) which is wide enough
to fit between the lower tapered surface (G) of the inner
ring (F) and the tapered portion (G) of the squeezer
bottom. This arrangement makes a good seal to keep the
sediment from being extruded between the inner ring and
the squeezer bottom while the porewaters pass through the
two filters. (A problem with a Reeburgh squeezer is that
sediment extrusion into the 0-ring channel on the bottom
edge of the cylinder makes it hard to clean).
3. Loading of sample into the squeezer. First, a
sediment slice is extruded from the liner and the outer
contaminated layers scraped off. The sediment (K) is
placed on the filter paper and covered by a circle of
thin polyethylene film (L) wide enough to fit between the
inner ring (F) and the squeezer bottom. The sediment
thus becomes the "meat" in a "sandwich" of filter paper
on the bottom and polyethylene sheeting on the top.
Sediment flow outside the "sandwich" is minimal. The
compacted sample is easy to remove and put in plastic
bags for subsequent chemical or mineralogical analysis.
Cleanup between samples is quick and easy. All that is
necessary is to wipe the squeezer parts with a Kimwipe
tissue. It is almost never necessary to clean the rubber
membrane in the top. Much time is thereby saved when
processing large numbers of sediment samples to define
interstitial gradients. Sediment should always be spread
out in as thin a layer as possible, no more than 1 cm.
When thicker layers are squeezed, the sediment next to
the filter compacts during the first part of the
squeezing process and inhibits percolation from porewater
farther from the filter, reducing yield.
4. Completing the squeezing. The top of the squeezer is
screwed on and tightened with a wrench (note in Figure 11
that the bottom has notches to hold it during the
tightening). Inert gas is admitted slowly to expand the
membrane (C) against the polyethylene film (L). Most of
the porewater should flow into the receiving bottle in a
rush after a pressure increase of less than 5 psi. The
pressure is increased to 120 to 160 psi and held for 5 to
15 min. The receiving bottle with the porewater sample
is removed and capped, after which the squeezer is
disassembled, cleaned and reloaded for the next sample.
The Nuclepore membranes can even be dried with tissue and
71
-------
re-used; polycarbonate is tough. A normal porewater
yield is 10 to 15 cm3 from a 40 cm3 sediment sample,
which is about 50 percent of the porewater present in the
sample. Although optimum results are assured by on-board
analysis of samples within a few hours of extraction,
pore fluids may be frozen in well-sealed containers
flushed with inert gas, and analysis deferred for several
days.
Nitrate Analysis
Nitrate dissolved in pore waters cannot be easily
measured. Therefore it is reduced to nitrite in a Cd-Cu reductor,
and the nitrite is measurad colorimetrically. The reaction
between N02 ~ and sulfanilimide produces the diazonium ion
which is then reacted with N-(1-naphthyl)-ethylenediamine
dihydrochloride to form an intensely rose-colored azo-dye. That
color is measured at 520 to 540 nm in a spectrophotometer or
colorimeter. The procedure was first described for seawater by
Bendschneider and Robinson (1952) . Porewater may contain nitrite
as well as nitrate. Therefore the effluent from the Cd-Cu
reductor may contain more nitrite than that produced from
nitrate. To obtain interstitial nitrate levels, you must conduct
two analyses on each porewater sample. One analysis determines
the concentration of nitrite alone, and the other determines the
concentration of nitrite + nitrate in the reductor effluent. The
nitrate concentration in the porewater sample is then estimated by
difference.
Porewater is usually obtained in small volumes. A 10 to
15 cm3 of porewater is a large sample. This presents a problem
because the well-established manual methods for nitrate
determination are set up for 50 to 100 cm3 samples (Strickland
and Parsons 1972, Grasshoff 1976). A scaled-down manual method is
not available and, even if it were, would probably result in large
errors. Porewater samples could be diluted; however the
sedimentary regions of greatest interest are those with low
porewater nitrate profiles. Dilution of the porewater nitrate
concentrations would be counterproductive to a precise definition
of the profiles in those regions. The desired analytical method
will utilize from 1 to 3 cm^. Automated analysis such as
performed on an autoanalyzer can meet this need. Typical sampling
times are 0.5 min which, according to the flow charts for the
nitrite and nitrate + nitrite methods in Figures 13 and 14, will
consume only 0.37 cm3. The proper procedure is to place 0.5
cm3 of porewater in a 4-cm3 autoanalyzer sample cup for a
nitrate analysis. A duplicate determination thus consumes only
about 1 cm3, or 10-15 percent of a typical porewater sample.
This provides an inducement for using an autoanalyzer method,
which is faster than a manual method and more reproducible because
72
-------
TECHNICON AUTOANALYZER METHOD NITRITE
CONCENTRATION RANGE 0-3.00uM
TYPE SAMPLE SURFACE, SALINE AND PORE WATER
REFERENCE AAH
TO
SAMPLER
SAMPLE
TUBE SIZE
(inches)
.073
WASH
u>
COLORIMETER RECORDER
50mm TUBULAR F/C
520 m^i FILTERS
RATE'30 PER HOUR
(21)
SAMPLER IV
20-TURN 20-TURN
ooooooon 00000009
COIL COIL
AIR .030
| COLOR REAGENT .035
SAMPLE .035
WASTE .035
*" F/C PULL THRU
PROPORTIONING
PUMP
1
1,
"1 r1'
•IT*
T
COLOR FLOW RATE
CODE (ml /mi n)
GREEN -GREEN 2.00
BLACK-BLACK 0.32
ORANGE-ORANGE 0.42
ORANGE-ORANGE O.42
ORANGE-ORANGE 0.42
Figure 13. Schematic diagram of the configuration of a Technicon Auto-
analyzer II for the determination of dissolved nitrite.
The color codes listed are for the various sizes of tubing
used in the configuration.
-------
TECHNICON AUTOANALYZER METHOD NITRATE + NITRITE
CONCENTRATION RANGE Q-SOuM
TYPE SAMPLE SURFACE .SALINE AND PORE WATER
REFERENCE AAH
WASTE
c
REDUCTOR
RESAMPUE LINE
POLYETHYLENE
TUBING
COLUMN
20-TURN
OQOQQQQO
fe-
COIL
AIR
AMMONIUM
CHLORIDE
SAMPLE
AIR
COIL
COLOR
REAGENT
SAMPLE
WASH
F/C WASTE PULL THRU
COLUMN WASTE
TUBE SIZE
(inches)
030
.056
.030
RATE' 30 PER HOUR
(Z'l)
SAMPLER IV
.030
.030
.073
.051
.040
PULL THROUGH
COLOR
CODE
BLACK- BLACK
YELLOW-YELLOW
BLACK-BLACK
BLACK-BLACK
BLACK-BLACK
GREEN-GREEN
GRAY- GRAY
WHITE-WHITE
FLOW RATE
(ml/min)
0.32
I 20
0.32
0.32
0.32
200
1.00
0.60
PROPORTIONING
PUMP
COLORIMETER RECORDER
JSOmm TUBULAR F/C
520 my. FILTERS
Figure 14. Schematic diagram of the configuration of a Technicon Auto-
analyzer II for the determination of dissolved nitrate +
nitrite. The color codes listed are for the various sizes
of tubing used in the configuration.
-------
both samples and standards are processed under carefully
controlled timing and operational conditions. Autoanalyzers are
common to many water quality laboratories, and, although the
initial cost of the instrument is high, operating expenses are
reasonable.
The procedure for a specific system, the Technicon
Autoanalyzer II, is as follows:
Samples are placed in the 4-cm3 plastic sampling cups
which are arranged in a circle around a tray. The
mechanical sampler inserts a stainless steel tube in each
cup and sucks out sample for a carefully controlled time
period. The tube is withdrawn, placed in a deionized
water wash for another carefully controlled time period,
and then inserted in the next sample cup along the tray.
During the wash step, deionized water is sucked into the
system such that samples, standards, and blanks are
separated from each other by pulses of wash water. Air
bubbles are inserted at regular intervals along the
stream of samples, standards, blanks, and wash water to
prevent those entities from blending with each other due
to backward streaming of liquids along the tubing walls.
Color is developed downstream of the sampler after
reagents are added to the stream and the mixture passes
through mixing coils (Figs. 13 and 14). A peristaltic
pump both sucks and moves samples and injects reagents.
The air bubbles are removed just before the stream in
which color is developing passes through a flow cell in a
colorimeter. There the intensity of the color is sensed
and appears as a series of peaks on a moving strip
chart. The heights of the peaks are measures of the
concentrations of the substance of interest in the
various samples, standards, and blanks being analyzed.
Further details may be found in the standard references
of Strickland and Parsons (1972), Grasshoff (1976) and in
the Technicon literature (Technicon Industrial Methods
161-71W/B, 1976, and 158-71W, 1972). Since nitrite and
nitrate + nitrite will be analyzed on the same porewater
sample, time can be saved by sampling for both methods
from the same 4-cm3 cup and using a splitter to divide
the stream out of the stainless steel sampling tube.
Tubing on the output nipples of the splitter can be
selected to provide the input flow rates recommended for
each method: 0.42 ml/min for nitrite (Fig. 13) and 0.32
ml/min for nitrate + nitrate (Fig. 14).
75
-------
Nitrite
Refer to Figure 13 during this discussion. Only one
reagent is required, the DIAZO color reagent which is labelled
COLOR REAGENT in Figure 13. It is prepared as follows:
Sulfanilamide (CgHs^C^S) 5.0 g
Concentrated Sulfuric acid 50.0 ml
N-(1-naphthyl)-ethylenediamine dihydrochloride
(C12H14N2°22HC1) °-25 9
Brij-35 0.5 ml
Deionized water 500.0 ml
Add 50 ml of concentrated sulfuric acid and 5.0 g of
sulfanilamide to 350 ml of deionized water. Some heating
may be necessary to dissolve the sulfanilamide although
the heat of dilution of the concentrated acid is usually
sufficient. Dissolve 0.25 g of
N-(1-naphthyl)-ethylenediamine dihydrochloride and 0.5 ml
of Brij-35 in the solution. Dilute to 500 mL with
deionized water.
The DIAZO color reagent should be kept in a dark bottle
and stored in a refrigerator (4 °C); it should last 2 to 3
months if so preserved. Brij-35, a reagent supplied by Technicon,
controls air-bubble surface tension and wall friction inside the
tubing carrying the sample stream. The control is necessary to
guarantee that the air bubbles are of uniform size and spacing.
Brij-35 is very important; without it or with too little of it,
peaks on the strip chart become erratic and very difficult to
read. If the DIAZO reagent has a pinkish caste after preparation,
discard it and try again with fresh reagents. The old reagents
are probably contaminated.
Two types of standards are required. The first is the
Stock standard which is very concentrated and is used for a long
time. The second is the Daily standard which is prepared by
diluting the Stock standard every time samples are run.
Preparation protocol is as follows: Stock standard -- 5 mM
N02- Dry reagent grade NaN02 and weigh out 0.3450 g to +/-
0.1 mg. Dissolve in a little deionized water in a 1-liter Class A
volumetric flask. Add a few drops of chloroform (CHC13) and
dilute to the mark. Daily (or working) standards -- prepare four
or five working standards by diluting aliquots of the Stock
standard with seawater. Use Class A volumetric flasks. The range
of the working standards should be 0 - 5 uM.
Chloroform is added to the Stock standard to retard
bacterial activity, and, if the Stock standard is kept in a dark
bottle, it can last for up to a year. The Daily standards are
76
-------
made up in seawater because of refractive index changes that occur
in the colorimeter flow cell when the boundary between a porewater
sample and the deionized water wash solution crosses the light
path. if the working standards are made up in seawater, then the
refractive index changes are the same for both samples and
standards. Almost any seawater will do, as long as it is not
anoxic and has a salinity of at least 33. Open-ocean surface
seawater is preferable, if available.
A blank solution is required because almost no seawater
can be guaranteed to be "nitrite-free." A 0.7-M NaCl solution is
an excellent blank solution. There is a possibility that some
reagent-grade NaCl is slightly contaminated with nitrite.
Therefore, the analyst should first prepare two blank solutions:
one made with ULTREX NaCl and the other with ordinary
reagent-grade NaCl. If the reagent grade gives a peak with a
height greater than that of the ULTREX NaCl, then the
reagent-grade NaCl should be re-crystallized once or twice to
remove the nitrite contamination.
The following protocol should be followed in performing
nitrite analyses on porewaters. Configure the Autoanalyzer
according to Figure 13. Run all samples in duplicate, and run a
set of standards and blanks for every 20 samples. Adjust the
baseline to between zero and 3 percent. Adjust the gain until the
peak for the most concentrated working nitrite standard is nearly
full scale on the recorder (90 - 100 percent). The cam in the
sampler should be selected so that the sampling tube is in each
cup for 0.5 min. Wash time should be selected to give suitable
separation between sample peaks. Then the run of samples,
standards, and blanks should be completed. Measure the heights of
the peaks (PH) above the baseline on the strip chart (a low-level
nitrite peak often resembles two mountains connected by an
elevated plain). The peak height is the distance from the
baseline to the elevated plain. Nitrite concentrations should be
calculated according to the following equation:
[N02l = (PHgample - PHb;Lank) x F
"F" is the least-squares slope of the regression of
concentration vs. peak height for the set of working standards for
the run. A sample of the diluting seawater for those standards
should be given a "zero" concentration, and the concentrations of
the other standards calculated on the basis of that assumption and
the nitrite added in the aliquots of the Stock standard. Use of
the 0.7-M NaCl blank solution corrects for the fact that the
diluting seawater may contain some nitrite.
Nitrate + Nitrite
The protocol for the Autoanalyzer processing of nitrate +
nitrite has two main parts (Fig. 14). In the first, the sample
77
-------
stream comes from the sampler (or one outlet of the sample
splitter), is injected with air bubbles and ammonium chloride
solution, is de-bubbled, and flows into the Cu-Cd reductor column
where the nitrate is reduced to nitrite. Then, in the second
part, the effluent from the reductor column is re-injected with
air bubbles, is injected with color reagent which reacts with the
nitrite, and is de-bubbled and passed through the colorimeter flow
cell as in the nitrite method.
The first preparation the analyst must make is the Cu-Cd
reductor column (Fig. 15). Two solutions are required to make and
maintain the reductor columns: (1) a solution of 10 g NH4C1 per
liter to which 3-5 drops of concentrated NH4
-------
1/3 x 1/4
SLEEVE
Figure 15. Cadmium-copper re due tor column for the
determination of nitrate + nitrite in.
marine porewaters. The column shown is
of Technicon design with fittings that
attach to a Technicon Autoanalyzer II.
79
-------
If kept in a tightly capped dark bottle, the Stock
standard should last for months since the CHC13 retards
bacterial activity. Working standards are made with seawater to
guarantee that refractive index changes are the same for both
samples and standards. This approach is similar to that used in
the nitrite method. However, in the case of the nitrite + nitrate
method, the seawater should be surface seawater reasonably far
from land. Deeper seawater or seawater too close to land can have
as much nitrate as porewaters. Salinity should be at least 33. A
0.7-M NaCl blank solution is used for the same reason and with the
same precautions for the nitrate + nitrite method as for the
nitrite method. The Autoanalyzer should be configured according
to Figure 14. All samples are run in duplicate, and a set of
standards and blanks as run for every 20 samples. An important
beginning step is to "activate" the Cu-Cd reductor column by
passing a seawater solution rich in nitrate through it. A 2-mM
solution is recommended, and 2-3 cups in the sampler tray should
be filled with this solution. This "activation" process is
necessary to make certain that the first sample or standard to
pass through the reductor is reduced with the same efficiency as
all the other samples and standards. Without "activation" the
reductor may not reach maximal efficiency until several standards
or samples have passed through. One should avoid connecting the
color reagent line to the reductor effluent until the large
nitrate "spike" has passed. Otherwise, an intense purple color
forms that has to be washed out before other samples can be
processed. The nitrate + nitrite run proceeds similarly to a
nitrite run. Sampling time is 0.5 min, wash time is selected to
optimize peaks; baseline and gain are adjusted by the same
criteria; and peak heights are determined the same way.
Concentrations of nitrate + nitrite are calculated by the equation:
[N0§ + N02l = (PHsampie - PHblank)
"F" is the least-squares slope of the regression of concentration
vs. peak height for the set of working standards for the run.
Because the Stock standard is so high in nitrate, only small
aliquots will be necessary to prepare the working standards, and
all of them can be considered to have the same low nitrite
concentrations. For the purpose of obtaining the factor "F," it
is reasonable to assume that nitrite concentration is zero.
Therefore, the concentrations used in the standardization
regression are taken to be the same as the increases produced by
the standard additions of nitrate. As with the nitrite method, use
of the 0.7-M NaCl blank solution corrects for the fact that the
diluting seawater may contain some nitrite or nitrate. When a run
is over the NH4C1 solution should be pumped through the Cd-Cu
reductor column. The lifetime and functioning of the column are
greatly enhanced by storage in Nl^Cl. The reductor column has a
finite lifetime, and it is critical to test the column
periodically for reduction efficiency. The test is to add spikes
80
-------
of NaNC>3 and NaN02 to the same sample. The additions should
be such that both the nitrate and nitrite concentrations in the
sample are to be increased 10 M higher. Then the sample is passed
through the nitrite and the nitrate + nitrite protocols
simultaneously in the normal way. The increases in peak heights,
for the altered sample will show whether or not the reducing
efficiency of the reductor column is suitably close to 100
percent. Air bubbles are a serious problem in the reductor
column. The column should be installed with no bubbles entering
at the connectors, and columns separated from the Autoanalyzer
should be stored with NH4Cl-solution-filled Tygon tubing between
their ends.
Quality Assurance
Because of the labile nature of nitrate and the fact that
relatively few sets of analyses have been performed with this
method, determination of the accuracy of the test has not been
feasible. Precision has been examined, however, using the
percent recovery test (EPA 1979), in which a controlled nitrate
spike is run on the autoanalyzer, an 85 percent recovery was
accomplished at a concentration of 2 uM and a value of 98 percent
was achieved at a 10-uM concentration. Replicate analyses, in the
range from 0 to 30 uM, yielded a maximum standard deviation of
0.3. In addition, 58 pairs of duplicate analyses on stream
samples showed an average range of difference of 0.8 uM.
As stated previously, data quality is the most vulnerable
when an orignally "reduced" sample is exposed to oxygen.
Furthermore, nitrate can be lost from an oxidized sample if the
sample is sealed off from oxygen. For this reason, samples should
be processed within a few hours of core recovery.
Cost Analysis
COST PER ANALYSIS AT BATCH RATE $ 450
Discussion; Alternative Methods To Measure Oxidation State
The simplest alternative to nitrate measurement as a
monitor of oxidation state in sediments is color. The boundary
between oxygenated and anoxic sediments is often marked by a color
change from gray-brown or brown to green. However, this is
obviously an imprecise measurement. Measurement of
oxidation-reduction potential using a platinum probe and reference
electrode in sediments often yields qualitatively useful results,
but cannot be relied upon as a quantitative tool (Whitfield 1974,
Stumm and Morgan 1981, pp. 490-497). Interstitial oxygen has
been measured by recent investigators (Revsbech and others 1980,
Revsbech 1983, and Van der Loeff and Lavaleye 1986). The elegant
81
-------
method of Revsbech (1983) provides detailed interstitial gradients
of dissolved oxygen using microelectrodes inserted into sediment.
With additional development these techniques may become usable for
routine analysis (Helder and Bakker 1985). However, even with
improved techniques for oxygen determination, nitrate measurements
retain value. This is because nitrate is not present as a
potential contaminant in air, and more important, because
porewater oxygen frequently becomes depleted well above the
sediment zones in which metals can be reduced (Froelich and others
1979).
References
Bendschneider, K. and Robinson, R.J., 1982. A new method
for the determination of nitrite in seawater, Jour. Mar.
Res., vol. 11, pp. 87-96.
EPA, 1979. Handbook for analytical quality control in
water and wastewater laboratories, EPA-600/4-79-019.
Fanning, K.A. and Maynard-Hensley, V., 1980. Oxidative
changes to nitrate and boron in marine pore waters,
Nature, vol. 287, pp. 38-41.
Froelich, P.M. et al, 1979. Early oxidation of organic
matter in pelagic sediments of the eastern equatorial
Atlantic: suboxic diagenesis, Geochim. Cosmochim. Acta,
vol. 43, pp. 1075-1090.
Froelich, P.M. et al, 1983. Pore water fluoride in Peru
continental margin sediments: uptake from seawater,
Geochim. Cosmochim. Acta, vol. 47, pp. 1605-1612.
Helder, W. and Bakker, J.F., 1985. Shipboard comparison
of micro- and minielectrodes for measuring oxygen
distribution in marine sediments, Limnology and
Oceanography, vol. 30, pp. 1106-1109.
Grasshoff, K., 1976. Methods of Sea Water Analysis,
Verlag Chemie, New York, 317 pp.
Kepkay, P.E., Cooke, R.C. and Bowen, A.J., 1981.
Molecular diffusion and the sedimentary environment:
results from the ijn situ determination of whole sediment
diffusion coefficients, Geochim. Cosmochim. Acta, vol.
45, pp. 1401-1409.
Lyle, M. , 1983. The brown-green color transition in
marine sediments: a marker of the Fe(III)-Fe(II) redox
boundary, Limnology and Oceanography, vol. 28, pp.
1026-1033.
82
-------
References (continued)
Reeburgh, W.S., 1967. An improved interstitial water
sampler, Limnology and Oceanography, vol. 12, pp. 163-165
Revsbech, N.P., Sorensen, J. and Blackburn, T.H., 1980,
Distribution of oxygen in marine sediments: Limnology
and Oceanography, vol. 25, pp. 403-411.
Rutgers van der Loeff, M.M. and Lavaleye, M.S.S., 1986.
Sediments, fauna and the dispersal of radionuclides at
the N.E. Atlatnic dumpsite for low-level radioactive
waste, Netherlands Institute for Sea Research, Texel,
134 pp.
Sayles, F.L. et al, 1976. A sampler for the in situ
collection of marine sedimentary pore waters, Deep-Sea
Res., vol. 23, pp. 259-264.
Soutar, A. et al, 1981. Sampling the sediment-water
interface -- evidence for an organic-rich surface layer,
EOS, vol. 62, 45 pp.
Strickland, J.D.H. and Parsons, T.R., 1972. A Practical
Handbook of Sea Water Analysis, Bull. 167, 2nd. ed.,
Fish. Res. Bd. Canada, Ottawa, 310 pp.
Stumm, W., and Morgan, J.J., 1981. Aquatic chemistry,
2nd Ed. John Wiley Sons, New York, NY, 780 pp.
Whitfield, M., 1974. Thermodynamic limitations on the
use of platinum electrodes in Eh measurements, Limnology
and Oceanography, vol. 19, pp. . 857-865.
83
-------
CHAPTER 9
Eh and pH
Introduction
E^ (oxidation-reduction potential) and pH (hydrogen ion
activity) are key variables in defining the geochemical
environment of any seafloor area. Where inferrable (accurate
determinations are difficult), they are indicators of the types of
reactions that are taking place, and also allow prediction of the
rate of these reactions and the products. Eh and pH also
influence a sediment's capability to retain certain radionuclides.
The methods described herein may be used to determine the
Eh and pH of a moist sediment sample. They are both
nondestructive tests, and can be used on small sediment samples.
These two tests are often linked because they employ similar
instrumentation and are performed in a similar manner. Also, it
may be convenient to conduct Eh and pH tests in conjunction with
the test for nitrate (see Chapter 6).
Eh
The oxidation-reduction potential (Eh)/ as measured by
this method, is defined in ASTM definition D-1129 (ASTM 1982) as
the electromotive force developed by a noble metal electrode
immersed in the sample and referred to a hydrogen electrode.
In a simple, reversible reaction at equilibrium, Eh is
given by the Nernst equation:
RT [A]a [B]b
Eh = E° = + In
NF [C]C [D]d
where R is the gas constant, T is the absolute temperature in
degrees Kelvin, N is the number of electrons exchanged in the
reaction equation, F is the Faraday constant, and E° is the
standard potential of the reaction in volts. A and B are the
oxidized species and C and D are the reduced species. E° is
related to the Gibbs free energy of the reaction.
In practice, the Eh is measured as the potential
developed between a reference electrode and a noble metal
electrode, typically platinum. This oxidation-reduction potential
is related to the standard hydrogen cell which is the hydrogen
gas-hydrogen ion couple (Langmuir 1971).
84
-------
Procedure
Most laboratory pH meters are suitable for Eft
measurements when using the millivolt scale and the appropriate
electrode. Typically, a combination electrode containing both the
platinum and reference electrode is desirable. It is convenient
to have a probe that has a plastic body and is easily cleaned for
shipboard measurements. A battery-operated meter is preferred.
Several solutions can be used as standards to calibrate
Eft measurements, many of which use the Fe++/Fe+++ couple.
The redox standard solution recommended in ASTM D-1498 is prepared
as follows:
Dissolve 39.21 g of ferrous ammonium sulfate, 48.22 g of
ferric ammonium sulfate and 56.2 mL of sulfuric acid in
water and dilute to 1 liter. Using a platinum probe
normalized to the hydrogen electrode (see section on
calculations), the Eft should read +675 mV.
Three other reference solutions are also available. Two
are recommended by the manufacturers of the Orion redox electrodes
(Orion Research, Inc. 1978). Another, and perhaps the most
commonly used, is Zobell's solution (Langmuir 1971).
Make sure that the probe is filled with the fresh
solution recommended by the manufacturer. Rinse the probe with
flowing distilled water from a squeeze bottle and blot the water
on the probe with laboratory wipes. Do not rub the probe because
this may generate a static electric charge.
Place the probe into the standard solution, agitate the
solution until the Eft reading begins to stabilize, and then let
it rest. Adjust the meter to the millivoltage of the standard.
Rinse the probe as before and insert it about 1 cm into the moist
sediment sample. Allow the meter to stabilize and record the
value, including the sign, to the nearest 10 mV- If the reading
is unstable, slight movement of the probe may help. Remove the
probe and rinse. When storing the probe it should either be kept
in distilled water or the filling solution should be drained.
Condition of the sediment sample is of extreme importance
to this measurement. The sample must be absolutely fresh and must
not have been exposed to air. The test has no thermodynamic
validity in the presence of oxygen. However, it may still have
qualitative value, so analysis should be performed immediately
upon retrieval. If practical, use a core liner in which you have
drilled holes large enough for the probe to pass through. Seal
the holes with duct tape prior to sampling. These can be slit
upon retrieval and the probe inserted for the measurement.
85
-------
The meter-read value for Eh is a redox potential
specific to the probe, reference electrode, filling solution and
the sample. By convention, this value is normalized to the
hydrogen electrode for each probe and filling solution by the
probe manufacturer. The calculation is:
ENHE = Eobs + Eref
where ENHE is tne oxidation-reduction potential of the sample
relative to the normal hydrogen electrode; EODs is tne potential
developed between the platinum electrode and the reference
electrode, and can be observed at the meter; and Eref is the
potential developed by the reference electrode relative to the
normal hydrogen electrode. Eref is a constant for the probe and
filling solution and is provided by the probe manufacturer.
Quality assurance
Electrochemical analysis of Eh in natural systems has
within it inherent sources of error that limit its usefulness to
that of a semiquantitative method. It can be used to determine
general redox regimes in sediment, and in many cases this is
probably adequate.
In most natural water systems, there are generally
multiple redox reactions occurring simultaneously. The measured
potential is the sum of these reacti9ns. The redox potential
observed is therefore a mixed potential. Moreover, concentrations
of reactants participating in redox reactions are often low and
not at equilibrium. This causes drifting of measurements as well
as uncertainities in identifying the reactions that dominate the
system (Lindberg and Runnelis 1984) .
Redox reactions in sediment are somewhat less
problematical because concentrations of elements involved in redox
reactions are typically higher in sediment than in water. Also,
biological action in the sediment often produces a relatively
broad range of E^ values that make E^ changes more obvious.
For example, the upper 'few centimaters of sediment may well be
ventilated by bioturbation while just below this upper layer,
bacterial action causes a zone of sharply decreasing Eft.
A major source of potential error in electrochemical
measurement is fouling of the platinum electrode, particularly
over a series of many analyses. Oxygen can poison the reactive
surface of the electrode by forming oxides or hydroxides of
platinum. H2S, H2, CH4, CO, and films of organic material
and ferric oxhydroxides are all reported to interfere with Eft
measurements (Langmuir 1971).
86
-------
Electrodes can be cleaned in several ways. ASTM D-1498
recommends immersion in warm aqua regia for one min (ASTM 1982).
Others recommend first soaking the electrode in detergent
solutions, then cleaning the platinum with scouring powder or a
very fine abrasive such as jeweler's rouge (Langmuir 1971).
Temperature differences between the sediment being
analyzed, the probe, and the reference standard may cause drifting
of measured values. This effect is typically small, perhaps 15 mV
between 20 °C and 5 °C.
Cost analysis
COST PER ANALYSIS AT BATCH RATE $ 10
####f###lt##tt###tt##tt####tt##ttt#tt##tt####t#####t##tt##t###tt###########
pH
The pH of a solution is a measure of the hydrogen ion
activity therein and is expressed as:
pH = -log10[H+]
In practice, however, pH is not a direct measure of
[H+]; rather, it is determined electronically by reference to
one or two (preferable) standard buffer solutions that ideally are
near or bracket the pH of the unknown solution. This is necessary
because, unless the solutions are diluted and composed of simple
electrolytes, the liquid junction potential and individual ion
activities cannot be evaluated easily and, therfore, a rigorous
determination of [H+] cannot be made (Bates 1973) . Seawater
and, thus, the interstitial water of nearly all surficial marine
sediments, is not a dilute solution. Such waters typically have
ionic strengths well above that of a "dilute" solution (I = 0.7
vs. I equal to or less than 0.1). Accordingly, comparative
measurements based on buffers and reference potentials are the
operational norm for seawater and seawater-related solutions as
well as most other natural aqueous solutions.
pH is commonly determined by observing the potential
difference between a glass electrode and a reference electrode in
a test solution. In effect, a pH meter is a specially calibrated
voltmeter that is connected to paired probes or a combination
probe. Thus, in operation, pH is defined as a voltage measurement
made on cells of the type
glass electrode solution (X) KCl(aq) reference electrode
in which solution (X) is either a buffer solution of known pH (for
meter calibration) or the test solution of unknown pH. Skirrow
(1975) and Stumm and Morgan (1970) give details of test
conventions and pH theoretical considerations.
87
-------
Procedure
This method requires a pH meter, preferably portable. A
glass combination electrode should replace the noble-metal
combination electrode used for En measurements. Standard buffer
solutions are required for calibration and should generally
encompass a range from pH 7 to 10. Proceed as follows:
1. Sediment samples, ideally, should be tested as soon
as possible after retrieval. To facilitate this, prepare
core liner in which holes have been drilled at suitable
intervals before coring operations begin. Make sure the
holes are large enough to accomodate the probe. Tape
should be used to cover these holes during sampling. The
tape can be slit or removed to allow measurement.
Subcores of box cores can also be taken in this specially
prepared liner.
2. Fill the pH probe with fresh filling solution as
recommended by the manufacturer (Note: permanently filled
probes are available, but are generally less stable and
less accurate [at +/- 0.03 pH units] than a refillable
type [accuracy +/- .02]).
3. Connect the probe to the meter and ascertain that all
functions are normal.
4. Let the probe equilibrate in fresh seawater for at
least 1 h before performing the analysis. The seawater
should be kept at a temperature that is close to that of
the samples, which, because the samples are retrieved
from water depths greater than 4000 m, will be
approximately 1-2 °C. The standard buffers should be
at a similar temperature.
5. Calibrate the pH metar by first setting the
temperature compensation adjustment. This will probably
be in the range as stated above. Assuming buffer
solutions of approximately pH 7 and 10 are selected,
measure the pH of the "pH 7" buffer and adjust the meter
to read the buffer value (make sure that the actual
buffer value used is corrected for the cold
temperature). Similarly, measure the pH of the "pH 10"
buffer. If necessary, adjust the "slope" so the meter
reads the proper, temperature-corrected value for this
buffer. Repeat both measurements and record the
results. Always rinse the probe carefully between
measurements and dab it dry with soft paper. Do not wipe
the probe as this may cause a static electric buildup on
its surface.
88
-------
6. Slit or remove the tape covering the predrilled hole
and insert the probe about 2 cm into the sediment. Allow
the reading to stabilize and record the results.
7. Rinse the probe with clean seawater before proceeding
with other measurements.
8. At the end of the sequence of measurements, repeat
measurements of the two standard buffer solutions and
record the results.
9. Correct, if necessary, the sample pH values in
accordance with the change of buffer values.
Quality assurance
Measurement of pH in marine waters can have a high degree
of precision. in fact, according to Strickland and Parsons
(1968), a precision of +/- 0.03 units should be obtainable and may
actually be surpassed if proper technique and precautions are
exercised. Of particular note in this regard are temperature
effects, use of reliable buffer solutions, and probe maintenance.
Temperature affects pH in two ways. It can change the
electrode potential and it can change the ion activity within the
test solution. Both of these problems may be avoided if the
sediment pore water (i.e., the test solution), the probe, and the
standards are at the same temperature. It is also important to
calibrate the pH meter by setting the temperature compensation
control. It is desirable and common to measure pH immediately
upon sample retrieval; temperature should be measured at that time
on at least one sample from a given area and the compensation set
accordingly (probably at 1-2 °C, again assuming sediment samples
are recovered from greater than 4000 m water depth). The probes
and standard solutions should be kept under refrigeration at a
temperature that matches the measured temperature. Although
equations are available to correct a pH value to its in situ
equivalent, room temperature equilibration of all components is
not recommended because of the potentially adverse effects on
other measurements.
Using fresh buffer solutions of appropriate values is
essential for quality pH measurements. Also, the presence of
C02 or other contaminants may alter the buffer pH value
unacceptably. The National Bureau of Standards has certified nine
standard buffers and provides information on the preparation and
characteristics of buffer solutions. Commercially available
buffers are generally accurate to +/- 0.01 pH units. Numerous
publications are available that provide practical information on
buffers (e.g., Beckman 1983).
89
-------
Appropriate probe maintenance is also essential.
Although there are many possible reasons for a slow response, a
long stabilization time or inordinate drift, most of the time the
problem can be traced to an electrode defect. The chance of
developing a probe problem is particularly high during shipboard
work because the probes are inserted directly and repeatedly into
the sediment, which can lead to clogging of the fibrous reference
electrode, or if H2 is present, can lead to poisoning of the
electrode. Accordingly, probes should be constantly tested and
rejuvenated, unclogged, or replaced as necessary. The shipboard
environment is generally harsh on probes and electronic systems in
general, and care should be taken to ensure that all equipment is
maintained and stored properly and that the pH meter is properly
protected from salt air or physical abuse.
Because cores undergo a substantial reduction in ambient
pressure as they are brought to the surface, questions have arisen
concerning the possible effect of outgassing on pH; that is, as
pressure is released, gases usually dominated by C02 are lost,
affecting the pore fluids of core, especially at the insert hole
for the probe. Loss of C02 causing increase in pH can occur
even when C02 is not saturated in situ, if methane is lost,
sweeping other gases with it. The likelihood of spurious results
increases as the pressure differential between sea floor and
shipboard increases. When shipboard pH measurements are made,
investigators should be aware of this possible outgassing and its
effect on pH values.
Cost analysis
COST PER ANALYSIS AT BATCH RATE $ 10
###############################################§####################
90
-------
References
American Society for Testing and Material (ASTM), 1982.
Annual book of standards, vol. 31, Philadelphia, PA
Bates, R. G., 1973. Determination of pH: Theory and
Practice: John Wiley, New York, NY, 435 pp.
Beckman Instruments, Inc., 1983. The Beckman Handbook of
Applied Electrochemistry, Beckman Instruments, Inc.,
Bulletin No. 7739, 68 pp.
Langmuir, D., 1971. E^-pH determination, d^n Carver, R.
E., ed., Procedures in Sedimentary Petrology,
Wiley-Interscience, New York, NY, 653 pp.
Lindberg, R. E. and Runnelis, D. D. , 1984. Groundwater
redox reactions: an analysis of equilibrium state applied
to En measurements and geochemical modeling, Science,
vol. 225, pp. 925-927.
Orion Research, Inc., 1978. Instruction manual, Orion
Platinum Redox Electrode, Model 96-78, Cambridge, MA,
Skirrow, G., 1975. The dissolved gases -- carbon
dioxide, i_n Riley, G.P. and Skirrow, G. , eds., Chemical
oceanography, vol. 2, Academic Press, New York, NY, pp.
1-192.
Strickland, J. D. H. and Parsons, T. R., 1968. A
practical handbook of seawater analysis, Fisheries
Research Board of Canada Bulletin 167, 311 pp.
Stumm, W. and Morgan, J. J., 1970. Aquatic chemistry,
John Wiley, New York, NY, 583 pp.
91
-------
CHAPTER 10
TOTAL ORGANIC CARBON
Introduction
The organic content of marine sediments reflects the
chemical and physical conditions of the depositional and
postdepositional environment. Many geochemical studies routinely
measure organic carbon as an indicator of the amount of biogenic
matter in sediment samples. Because inorganic phases of carbon
are present in varying amounts, usually in the form of carbonates,
analytical techniques must be used to differentiate between the
organic and inorganic phases. Two methods for the analysis of
organic carbon in marine sediments are presented here, and the
analytical procedures, accuracy, and applicability of the methods
are discussed.
Test Method A
Procedure
The first method of organic carbon analysis involves high
temperature combustion using a LECO W12, model 1761-100, carbon
analyzer on samples from which the inorganic carbon (carbonate
carbon) has been removed. Removal of carbonate (pretreatment) is
accomplished by using either wet or dry leach techniques.
Wet leach pretreatment using HCL (10 percent by volume)
1. Samples to be analyzed should be dried thoroughly in an oven
at approximately 90 °C and then ground to homogeneity (fine
powder) using an agate mortar and pestle. There should be no
discernable differences in texture or color after grinding.
Samples can be transferred to appropriately sized glass vials for
drying and storage.
2. Redry samples for about 1 h at 90 °C, then transfer to
dessicator and allow to cool before weighing.
3. Weigh between 0.1 and 2.5 g of sediment into acid-cleaned
50-ml beakers (which should be purified in a muffle furnace at 500
°C to ensure against contamination). Samples that contain
relatively high C content (3 to 7 percent) require less material;
samples with high CaC03 content require the most. Samples
should be weighed to four decimal places and the weights
recorded. Beakers should be labeled carefully before weighing.
4. Prepare a 10 percent (by volume) HC1 solution (1.2 N), and
place samples on a tray in a frame hood. Carefully add 10 ml of
acid to each beaker. Care must be taken with carbonate-rich
samples to avoid fizzing. The samples are allowed to remain in a
fume hood overnight or until no further reaction is visible.
92
-------
5. Rinse sample residues carefully onto pre-ashed glass-fiber
filters (47-mm diameter, 0.4-micron opening) and heat them to 500
°C for 1 h in a muffle furnace. Use an all-glass Millipore
filtering assembly connected by a trap to a vacuum pump. Use only
distilled water, which has been prefiltered, to rinse, and be sure
to rinse any sediment from sides of glass funnel.
6. After rinsing acid from sample, fold filter in half twice
using forceps. Transfer to a porous LEGO combustion crucible.
Cap with aluminum foil and label. Record the number next to the
appropriate sample.
7. Place crucibles in the oven to dry and desiccate until you
begin to run the analysis using the LEGO Carbon Analyzer
previously described. Standards should always be run whenever the
combustion chamber is cleaned or traps in the sample lines are
changed.
Dry leach pretreatment
1. Sample preparation, grinding, and drying procedures duplicate
those described in steps 1 through 3, of the wet leach
pretreatment. However, samples may be transferred directly into
LECO crucibles after weighing, or weighed directly into the
crucibles themselves. Number the crucibles by scratching the
surface with a metal stylus.
2. Using an empty desiccator equipped with vacuum sleeve and
perforated ceramic plate, which you have first removed, add enough
concentrated HC1 to cover the bottom to a depth of about 1/2
inch. Replace the ceramic plate and arrange the crucibles on the
bottom in a systematic fashion. Crucibles should not be capped.
Close the desiccator and leave it in the fume hood for a minimum
of 3 days. Always follow appropriate safety precautions. Samples
may be placed on glass petri dishes for ease in transferring;
otherwise, handle with forceps. In either case, wear gloves, lab
coat, and safety glasses.
3. After a minimum of 3 days, remove samples from desiccator and
allow to vent for about 4 h in fume hood. Then, transfer samples
to a second dessicator which has been prepared in a similar way
using concentrated NH40H. Place samples in the desiccator,
cover, and allow to stand half an hour to neutralize the absorbed
HC1. Dry the samples in an oven until processing, or transfer to
a desiccator until LECO-WR 12 analysis. Always desiccate samples
in between analyses to guard against H20 absorption.
Total organic carbon determination
This method of analysis utilizes the difference in
thermal conductivity between two gases, C02 and 02/ to measure
93
-------
the amount of carbon released from a sample during combustion at
1,500 °C for 60 sec. The sample is combusted in an induction
furnace through which purified 02 flows at a controlled rate.
The carbon is oxidized to C02 and the sulphur to S02; residual
phases remain behind as oxides. A thermal conductivity cell,
consisting of matched thermistors connected by a Wheatstone
bridge, registers a current imbalance when C02 is allowed to
enter the measuring cell. This current is amplified, producing a
positive reading on the digital voltmeter that is proportionate to
the amount of carbon in the sample. S02f C02» and combustion
solids are collected by a series of traps during various phases of
the operation. A detailed description of the operation of the
LEGO analyzer is provided in the user's manual supplied by the
manufacturer.
Quality Assurance
Dry leach method of sample preparation
The dry leach method of sample preparation for organic
carbon analysis has been used in a number of different sediment
types. The following data (Table 7) were obtained for samples
analyzed in triplicate using the dry leach method. The range of
carbon values found for each suite of analyses is listed along
with standard deviation. The relative standard deviation (rsd)
obtained for each sample was averaged, and the mean value for each
set determined.
Table 7. Precision of Method A using dry pretreatment method.
Sample No. of Range, in percent, of Mean rsd
Site Analyses Organic Carbon [+/- Std. Dev.] (percent)
A * 13 0.76 +/- 0.04 - 1 15 +/- 0.33 +/- 0.76
B * 6 3.12 +/- 0.01 - 6.82 +/- 0.01 +/- 0.87
C * 6 1.68 +/- 0.18 - 6.59 +/- 0.06 +/- 1.70
D ** 3 3.50 +/- 0.21 - 4.62 +/- 0.42 +/- 0.62
E ** 7 0.50 +/- 0.004 - 0.72 +/- 0.014 +/- 0.84
Notes to Table
rsd = (standard deviation/percent organic carbon) x 100.
A = COST well No. 1 (Gulf of Mexico)
B = Cariaco Trench C = New York Bight
D = Lydonia Canyon axis E = Georges Bank shelf
* = Whelan, J. and Pratt, M. Unpublished data, Woods Hole
Oceanographic Institution, Woods Hole, MA
** = sediment trap samples, US Geological Survey, Woods Hole, MA
94
-------
Wet leach method of sample preparation
Ten sediment trap samples from the axis of Lydonia Canyon
on Georges Bank were analyzed in triplicate after carbonate had
been removed by wet leaching. Organic carbon percentiles for
these samples ranged from 2.78 +/- 0.01 to 4.00 +/- 0.14; mean rsd
was +/- 1.05 percent.
The range, sensitivity, and accuracy of the LEGO analyzer
on 1-g samples, as reported by the LEGO Corporation 1986 brochure,
is listed in Table 8. The regular range is satisfactory for most
carbon determinations
Table 8. Range, sensitivity, and accuracy of LEGO analyzer.
Range Sensitivity Accuracy
Regular 0.001 % C +/- 0.002 % C or 50%
(0.010 - 5.000 % C) of carbon present
Low 0.0001 % C +/- 0.0005 % C at
(0.005 - 0.200 % C) lower levels to
+/- 0.002 % C at
upper levels
Cost Analysis
Test Method A is based on use of a LECO-WR12 carbon
analyzer. Unit cost for batch processing depends on pretreatment
procedures. Time for preparation and analysis is 0.5 - 1 h per
sample.
COST PER ANALYSIS AT BATCH RATE $ 30 (wet leach)
$ 20 (dry leach)
####tt######ttttt########t#f###tt#####t###t###############tttt#########tt#
Test Method B
Procedure
The amount of organic carbon in sediments and sedimentary
rocks is determined by calculating the difference between total
carbon (TC) and carbonate carbon (1C) contained in a sample.
Total carbon is determined by using a Perkin-Elmer High
Temperature Combustion Elemental (CHN) Analyzer; C02 is
determined using a Coulometrics Inc. model 5011 coulometric
titrator, (Engleman 1985, in press).
95
-------
Total carbon
Samples to be analyzed are dried, thoroughly ground, and
redried before weighing. A 1- to 2-mg sample is weighed on a
microbalance and transferred to a quartz combustion tube for
analysis by the CHN elemental analyzer. A complete description of
instrument operation is available in the user's manual, and will
not be given here. However, the analyzer works on the same
principles as the LECO-W12 described previously- Acetanalide
standards (71 percent TC) are used to calibrate the instrument;
the relative standard deviation (rsd) is +/- 0.3 percent C (Brown,
F., personal communication 1986).
C02 analysis
This method of analysis is based on the exact measurement
of the quantity of electricity that passes through a solution
during an electrochemical reaction. Between 10 and 100 mg of
sample (more if carbonate content is high) are weighed into
porcelain crucibles and transferred to the reaction tube. The
tube is attached to the coulometer and 5 ml of 2N perchloric acid
is added; the tube is heated. The carrier gas (air), which is
scrubbed with KOH to remove C02, passes through the tube and
carries the evolved C02 through a Ag2S04 scrubber to remove
chlorine and sulphur. The C02 is bubbled into the coulometric
cell, which contains a partially aqueous medium containing
ethanolamine and a colorimetric indicator. The C02 is
quantitatively absorbed and converted to dioxyethylcarbamic acid
by the ethanolamine. The coulometer electrically generates enough
base to return the indicator solution to the starting point. The
starting and end points are determined photometrically. Because
the coulometer is electronically calibratad, empirical calibration
on the basis of carbon dioxide standards is not necessary.
However, it is recommended that sediment standards be prepared and
routinely analyzed at the beginning and end of a run.
Quality Assurance
Method accuracy is +/- 0.4 percent C02 (Zoanne Brown,
USGS Analytical Laboratories, personal communication 1985) based
on replicate analyses. Engleman and others (1985) report a
difference between within-day and day-to-day precision of
approximately 5 percent rsd (relative standard deviation) or less
over a concentration range of 0.01 - 36 percent C02 using the
coulometric titrator.
Cost Analysis
Test method B is based on use of a coulometric titrator.
Time for preparation and analysis is 0.5 - 1 h per sample.
COST PER ANALYSIS AT BATCH RATE $ 30
####tt###########t###################################################
96
-------
Discussion of Methods
Various techniques for the analysis of organic carbon in
soils, sediments, and sedimentary rocks have been described in the
literature and the merits and disadvantages discussed. Some of
these are included in the references listed below. The major
difficulty encountered in any of these analyses involves the
accurate separation of the inorganic (carbonate) and organic
phases of carbon. Wet leach methods, followed by filtration, may
result in loss of the more labile organic compounds, particularly
from recent marine samples that are high in carbonate, certain
soil types, and suspended-matter samples (see, Gross and others
1972, Roberts 1973, Leventhal and Shaw 1979, Froelich 1980, Weliky
and others 1983). Roberts (1973) estimated that from 9 to 44
percent organic carbon was lost from recent carbonate samples when
treated with various concentrations of sulfuric acid.
Preliminary comparison of the dry vs. wet leach method
was made using a standard sample which contained about 40 percent
carbonate and which was run in triplicate using both preparation
techniques. Mean organic carbon values obtained for the dry vs.
wet leach methods were 0.615 +/- 0.012 with rsd at 2.0 percent,
and 0.510 +/- 0.006 with rsd of 1.2 percent, respectively (David
Brewster, USGS, unpublished data 1981). The rsd of replicate
analyses using different samples (sediment-trap and piston-core
samples from Georges Bank) is 1.05 percent for the dry leach
method and 0.84 percent for the wet leach method. However, more
detailed comparisons need to be made. The dry leach method is not
recommended for samples with a high concentration of
organic-diatomaceous rich oozes or carbonate contents greater than
60 percent (Jean Whelan, WHOI, personal communication 1985). For
this level of carbonate, method II might be preferable.
In summary, the size and composition of the sample are
important when selecting an appropriate method of analysis. If
sample size is small, the CHN analyzer or Coulometric method may
be more suitable. If the organic content is extremely low and the
carbonate content high, methods that require more time-consuming
and sophisticated techniques, such as discussed by Froelich
(1980), may be required.
References
Boyce, R. E. and Bode, W. G.. DSDP, vol. IX, chapt. 13,
pp. 797-816.
Byers, S. C., Mills, E. L. and Stewart, P. L. , 1978. A
comparison of methods of determining organic carbon in marine
sediment with suggestions for a standard method, Hydrobiologic,
vol. 58, no. I, pp. 43-47.
97
-------
Dean, W. E., Jr., 1974. Determination of carbonate and
organic matter in calcareous sediment and sedimentary rocks by
loss on ignition: comparison with other methods, Journal of
Sedimentary Petrology, vol. 44, no. 1, pp 242-248.
DSDP Appendix III, Shore-based laboratory procedures,
DSDP, vol. IV, pp. 746-748.
Engleman, E. E., Jackson, L. L. and Norton D. R., 1985.
Determination of carbonate carbon in geological materials by
coulometric titration, Chemical Geology, vol. 53, pp. 125-128, in
press.
Froelich, p. N., 1980. Analysis of organic carbon in
marine sediments, Limnology and Oceanography, vol. 25, no. 3, pp.
564-572.
Gibbs, R. J., 1977. Effect of combustion temperature and
time, and the oxidation agent used in organic carbon and nitrogen
analyses of sediments and dissolved organic material, Journal of
Sedimentary Petrology, vol. 37, no. 2, pp. 547-550.
Gross, G. M. , 1971. Carbon analysis, ir\ Carver, R. E.,
ed., Procedures in Sedimentary Petrology, Wiley- Inter science, New
York, NY, pp. 573-580.
Gross, G. M. et al, 1972. Distribution of organic carbon
in surface sediment - Northeast Pacific Ocean, in Pruter, A.T. and
Alverrons, D.L., eds., Columbia River Estuary and Adjacent Ocean
Water, University of Washington Press, Seattle, WA, pp. 254-264.
Kolpack, R. L. and Bell, S. A., 1968. Gasometric
determination of carbon in sediments by hydroxide adsorption,
Journal of Sedimentary Petrology, vol. 38, pp. 617-620.
Leventhal, J. S. and Shaw, V- E., 1970. Organic matter
in Appalachian Devonian black shale: I. Comparison of techniques
to measure organic carbon and II. Short range organic carbon
content variations, Journal of Sedimentary Petrology, vol. 50, no.
1, pp. 77-81.
Mills, G. L., and Quinn, J. G., 1979. Determination of
organic carbon in marine sediments by persulfate oxidation,
Chemical Geology, vol. 25, pp. 155-162.
Roberts, A. A., Palacas, J. G. and Frost, I. C., 1973.
Determination of organic carbon in modern carbonate sediments,
Journal of Sedimentary Petrology, vol. 43, no. 4, pp. 1157-1159.
Weliky, K. S. et al, 1983. Problems with accurate carbon
measurements in marine sediments and water column particulates: A
new approach, Limnology and Oceanography, vol. 28, pp. 1252-1259.
98
-------
GEOTECHNICAL METHODS
99
-------
CHAPTER 11
WATER CONTENT
Introduction
The water content, defined as the mass of water
(including dissolved components such as salt) divided by the mass
of soil solids, is one of the most important and fundamental soil
parameters. it is also one of the simplest to determine. The
water content can indicate possible grain sizes of a sample
because clay particles tend to adsorb water to their surfaces.
For instance, a high water content typically indicates that a
sediment has a high clay content. Some clay minerals, such as
montmorillonite (smectite), have a greater tendency than others to
attract water particles (Lambe and Whitman 1969, p. 44). A low
water content, on the other hand, may mean that a coarser grain
size is present or that a clay has been heavily loaded, which
caused the adsorbed water to be squeezed out. When compared to
other measured properties such as Atterberg limits, water content
can be used to predict certain engineering behavior or may be
evidence that particular geologic processes have occurred. Water
content is used in many phase relation equations and is related to
the shear strength of a saturated clay (Lambe 1951, p. 8).
Procedure
Applicable ASTM standard: D2216-80, Standard method for
laboratory determination of water (moisture) content of soil,
rock, and soil-aggregate mixtures (ASTM 1987, pp. 355-358).
1. Select a representative specimen that has a mass of
at least 25 g. If a discontinuity or change in sediment type is
encountered at a particular level, representative specimens should
be obtained from each material. Spacing between specimen
subsampling depends on the overall test-program objectives and
should be specified as an appropriate interval; e.g. 10 cm. If a
sample had previously been bagged, thoroughly remold the sediment
with a spatula before obtaining a water content specimen.
2. Record the cruise and core identifiers, sample
interval, water content jar identifier, and mass of a clean dry
water content container.
3. Place the moist specimen in the container and
determine the mass of the container plus moist material by using a
balance that has a precision of +/- 0.01 g. Record the combined
mass on a data form (e.g., Fig. 16).
100
-------
WATER CONTENT
Cruise/Project-
Oven:
. Date:_
Initials:.
Oven Temperature:.
Salinity Correction (ppt):_
. Drying Time:.
Comments:.
Dish
No.
Mass
Dish
+
Wet
Sed.
(g)
Mass
Dish
+
Dry
Sed.
(g)
Mass
Dish
(g)
Mass
Water
(g)
Mass
Dry
Sed.
(g)
Water
Content
(%)
Corn
Water
Content
(??)
Figure 16. Typical water content data form.
101
-------
4. Place the sample and container into a drying oven
that can maintain its drying temperature within +/- 5 °C.
Depending on the type of sediment being dried, an oven temperature
between 60 and 110 °C should be used. Geotechnical testing
laboratories often use a temperature of 105 °C (Lambe 1951, p.
10), however, an oven temperature of 60 +/- 5 °C may be more
appropriate for materials containing significant amounts of
hydrated water or organic material (Liu and Evett 1984, p. 7).
5. After the material has dried to a constant mass,
remove the container from the oven (typically 8 to 24 h).
6. Immediately place the hot container into a desiccator
to cool.
7. When the sample has cooled to a temperature at which
it can be handled easily, place the container and sample on the
balance to determine the combined mass of the container and dry
sediment. Record the results. The dried water content sample can
be saved for future grain specific gravity testing. The dried
sample should not be used for grain-size analysis or x-ray
diffraction studies.
8. Subtract the mass of the container and dried sediment
from the mass of the container and wet sediment to obtain the mass
of water without dissolved salt.
9. Determine the mass of dried sediment and salt
precipitate by subtracting the container mass from the combined
mass of the container and dried material.
10. The water content of the specimen (w) uncorrected for
salt content in the pore fluid, can be determined from the
following equation:
w = mass of the water
mass of the dry sediment and salt
11. The water content (wc) corrected for salt in the
pore fluid, can be determined from Table 9 (assuming a salinity of
35 ppt). Find the appropriate percentage in the right column, and
add it to the uncorrected water content. For example, if the
uncorrected water content value was 95 percent, 7 percent should
be added, producing a corrected water content of 102 percent. The
corrected water content can also be determined from the following
equation using any salinity value:
/I + S
wc = \ 1000-S / x 100
(l
MS
+ S
1000-S
- / S
( 1000-S
)MW
x Mw\
102
-------
where: wc = water content (in percent) corrected for a
particular salinity value,
S = salinity (in ppt),
Mw = mass of water without salt, and
Ms = mass of sediment including salt.
Table 9. Corrections to natural water content within a subsample
to account for a salinity of 35 ppt in the pore water.
Uncorrected Add
natural water content (percent) (percent)
less than 12 0
13 to 31 1
32 to 46 2
47 to 59 3
60 to 70 4
71 to 81 5
82 to 90 6
91 to 99 7
100 to 108 8
109 to 116 9
117 to 123 10
124 to 130 11
131 to 137 12
138 to 144 13
145 to 150 14
151 to 156 15
157 to 162 16
163 to 168 17
169 to 174 18
175 to 179 19
180 to 185 20
186 to 190 21
191 to 195 22
196 to 200 23
12. The report (data sheet) should include the following:
water content of the specimen to the nearest 0.1 or 1 percent,
depending on the purpose and required precision of the test;
indication of test specimen having a low mass (below 25 g);
indication of test specimen containing more than one soil type
(layered, etc.); indication of any material (size and amount)
removed from the test specimen; and, the method of drying if
different from oven drying at 110 +/- 5 °C
Comments
An accurate determination of water content depends on
adequate sampling, handling, shipping, and storage of core
sections or water content samples (see Chapter 2). Leakage of
103
-------
pore water must be prevented, as must compaction of unsampled
sediment cores still within liners. Core sections or sample bags
should be well sealed.
Water contents must be corrected for the salt that
precipitates out of the pore water during drying because
corrections can exceed 10 percent of the actual water content
value.
The equation used to determine water content must always
be stated because some investigators define it as mass of water
(including salt) divided by the total sample mass. Using that
definition, the water content must be less than 100 percent,
whereas the definition in this section allows the water content to
exceed 100 percent.
Quality Assurance
Although ASTM (1987) has not yet developed requirements
for the precision and accuracy of this test method, Bennett and
others (1970) estimated that the precision was +/- 1 percent. The
method is suitable for all marine sediments, although great care
must be taken in some sediment types (e.g., sands) and in surface
sediments to ensure that the data represent the in situ conditions.
Cost Analysis
Time required for each water content analysis (not
including drying or cooling times): 10 minutes
COST PER SAMPLE AT BATCH RATE $ 4
######################################################f#############
References
American Society for Testing and Materials, 1987. Annual
book of standards: soil and rock, building stones, and
geotextiles, ASTM , Philadelphia, PA, vol. 04.08, 1189 pp.
Bennett, R. H., Keller, G. H. and Busby, R. F., 1970.
Mass property variability in three closely spaced
deep-sea sediment cores, Journal of Sedimentary
Petrology, vol. 40, no. 3, pp. 1038-1043.
Lambe, T. W., 1951. Soil testing for engineers, John
Wiley, New York, NY, 165 pp.
Lambe, T. W. and Whitman, R. V., 1969. Soil mechanics,
John Wilwy, New York, NY, 553 pp.
Liu, C. and Evett, J.B., 1984. Soil properties testing,
measurement, and evaluation, Prentice Hall, Englewood
Cliffs, N.J., 315 pp.
104
-------
CHAPTER 12
ATTERBERG LIMITS (LIQUID AND PLASTIC LIMITS)
Introduction
The liquid limit and plastic limit (called Atterberg
limits) are the two parameters used most often to distinguish the
boundaries between the consistency states of fine-grained soils.
The liquid limit is the water content that separates liquid- from
plastic-behaving remolded sediment, and the plastic limit
separates plastic from semisolid behavior. Therefore, if the
water content is above the liquid limit, the remolded sediment
will behave like a liquid; if the water content is below the
liquid limit, but above the plastic limit, the remolded sample
will exhibit plastic behavior.
The Atterberg limits are very useful parameters because
they indicate the water contents over which sediment behaves
plastically. Liquid and plastic limits are related to the amount
of water that is attracted to the surfaces of the individual
sediment particles. Nonplastic behavior is typically exhibited by
predominantly coarse-grained material. Typically, the higher the
clay mineral content of a sediment, the greater will be the amount
of adsorbed water on the clay particles and, hence, the higher the
Atterberg limits.
Other sample parameters that can be determined from
Atterberg limits include the plasticity index (Ip) which is the
difference between the liquid limit and plastic limit and the
liquidity index (IL) [(natural water content minus plastic
limit) + Ip], which relates the in situ water content to the
Atterberg limits. The latter is useful for estimating approximate
sediment stress histories. The plasticity index is often plotted
versus liquid limit on a plasticity chart (Fig. 17); the location
of the data indicates what type of sediment is present and also
the amount of compressibility that can be expected to follow
engineering-type loading (Peck and others 1974, p. 22).
Two methods of determining the liquid limit are currently
in use. Within the United States, the ASTM method of using a
brass drop cup is more popular. Elsewhere, however, the fall-cone
method prevails. Some data exist that indicate better precision
can be obtained with the fall-cone method (Head 1980).
105
-------
400
300
200
100
0
50
40
30
20
10
0
Vol
"(Ms
/
-
Benton
(Wyom
came clt
xi co Cit
te
ng)
"\
•i*
Z^ i^Orgor
(Flush
Re
lot
200
40
L 1 qu i
S*
Vori
0
ic sil
ing n
F-Lme"A"
ous
f pe
tar
nea(
i
types
at
d cla
Jows
y —
L.I.I
0 600 80
d limit
Expansive soils
(Ref. 14o)^^
Glacial clays^
d tro
erite
7
4
pical
clay
/
CL
X
-ML
/
A
V
\
IX
10 20
\
/
C
\'j&'
0
^-»^
/
v
^
^
H
^
/
*n
m
•3$
/
/
,'i
fy
w
*A
' — - —
/
/
^.
ffi
<*
m
w*
-Loe
ML and OL
30
/
C
/
^
im
S^i
^
ss
/
H
/*
A
fc >-y
i
2L_
/
;
A
z
r
/
/
^
,x7y
ft
p
Kaol
^~-
Mica
/
Sodi
Mon1
jm
mor
LL — iUl
PI =25C
\
y^
/{
n an
ceou
MH and OH
d all
ssilt
*/
i-
/
jvial
s
llonite
D to 600
D to 550
1 ,
y
/
'"&" line
Diatc
earth
clay:
mac
sous
40 50 60 7C 80 90
60
50
Liquid limit
Figure 17. Plasticity chart showing location of some types of soil (Hunt,
1984, p. 148). The following abbreviations are used; C: clay;
H: high plasticity; L: low plasticity; LL: liquid limit; M:
silt; 0: organic; and PI: plasticity index.
106
-------
Liquid Limit
Test Method A: Casagrande Drop Cup
In the following method, the liquid limit is defined as
the water content at which both sides of a remolded pat of soil,
placed in a standard cup and cut by a groove of standard
dimensions, will flow together at the base of the groove for a
distance of 13 mm (0.5 in) when the soil is subjected to 25 shocks
from the cup being dropped 10 mm in a standard liquid-limit
apparatus operated at a rate of 2 shocks per second.
Procedure
Applicable ASTM Standard: D4318-84, Standard test method
for liquid limit, plastic limit, and plasticity index of soils
(ASTM 1987, pp. 763-778).
1. Obtain a representative sediment sample. Note:
although ASTM recommends removing material retained on a
No. 40 sieve (425 micron), that practice artificially
biases the test results to indicate that a more plastic
material is present.
2. After calibrating the apparatus, place a portion of
the remolded soil in the cup of the liquid limit device
(Fig. 18) at the point where the cup rests on the base.
Squeeze it down, and spread it into the cup to a depth of
about 10 mm at its deepest point, tapering it to form an
approximately horizontal surface. Take care to eliminate
air bubbles from the soil.
3. Form a groove in the soil by drawing the tool,
beveled edge forward, through the soil. When cutting the
groove, hold the grooving tool against the surface of the
cup and draw it in an arc, maintaining the tool
perpendicular to the surface of the cup throughout its
movement. In soils where a groove cannot be made in one
stroke without tearing the soil, cut the groove with
several strokes of the grooving tool. Alternatively, cut
the groove to slightly less than required dimensions with
a spatula and use the grooving tool to bring the groove
to final dimensions. Exercise extreme care to prevent
sliding the soil pat relative to the surface of the cup.
4. Verify that no crumbs of soil are present on the base
or the underside of the cup. Lift and drop the cup by
turning the crank at a rate of 1.9 to 2.1 drops per
second until the two halves of the soil pat come in
contact at the bottom of the groove along a distance of
13 mm (0.5 in).
107
-------
DIMENSIONS
LETTER
MM
LETTER
MM
Az
54
± 0.5
N
24
B "
2
± O.I
P
28
C^
27
± 0.5
R
24
£ "
56
± 2.0
T
45
F
32
t/ 4
47
± 1.0
G
10
V
3.6
H
16
liV
13
•J
60
± 1.0
/
6.5
K*
50
± 2.0
L *
150
± 2.0
M"
125
± 2.0
ESSENTIAL DIMENSIONS
CAM
ANGLE
DEGREES
V DIAMETER
CRS OR BRASS PIN
-HARD RUBBER BASE CONFORMING SOFT RUBBER CONFORMING TO
TO SPECIFICATION IN 6.1.1 SPECIFICATION IN 6.1.2
Hand-Openled Liquid Limit Device
30
60
90
120
150
180
210
240
270
I 300
330
360
CAM
RADIUS
0.742 R
0.753 R
0.764 R
0.773 R
0.784 R
0.796 R
0.81 B R
0.854 R
0.901 R
0.945R !
-t
0.974 R .
0.995R
1.000 R
DIMENSIONS
LETTER
MM
"LETTER
MM
A4
2
l± O.li
~G
10
MINIMUM
B4
II
±0.2
~H
IJ
C4
40
± 0.5
J
6O
0"
B
± O.I
1 K4
IO
±0.05
E 4
50
± 0.5
- LA
60 DEG
± 1 DEG
"F"" "
2
±0.1
N
20
ESSENTIAL D/UFHSIONS
°BACK tT LEAST 15 MM FROM TIP
NOTC : DIHCNSION n SHOULD BC 1.9-2.0 AHD DIMENSION D
SHOULD BE S.O-a.l WHEN NEW TO ALLOW FOR
ADEQUATE SERVICE LIFE
Grooving Tool (Optional Height-of-Drop Cage Allached)
Figure 18. Mechanical drop cup liquid limit device and grooving tool (ASTM,
1987, p. 774-775).
108
-------
5. Verify that an air bubble has not caused premature
closure by observing that both sides of the groove have
flowed together with approximately the same shape. If a
bubble has caused premature closing of the groove, reform
the soil in the cup, adding a small amount of soil to
make up for any lost in the grooving operation, and
repeat steps 2 to 4. if the soil slides on the surface
of the cup, repeat steps 2 through 4 at a different water
content. If, after several trials at successively higher
and lower water contents, the soil pat continues to slide
in the cup or if the number of blows required to close
the groove is always less than 25, record that the liquid
limit could not be determined, and report the soil as
nonplastic without performing the plastic limit test.
6. Remix the soil sample on the glass plate without
adding to, or removing pore water from the sediment and
return a pat of soil to the cup, performing steps two
through five. When the operator has recorded at least
two trials within one count from each other, record the
number of drops, N, required to close the groove. Remove
a slice of soil approximately 20 mm wide, extending from
edge to edge of the soil cake at right angles to the
groove and including that portion of the groove in which
the soil flowed together. Place in a weighed container
and cover.
7. Return the soil remaining in the cup to the glass
plate. Wash and dry the cup and grooving tool prior to
the next trial.
8. Remix the soil specimen on the glass plate, adding or
reducing water to increase or decrease the water content
of the soil and change the number of blows required to
close the groove. Repeat steps two through seven for at
least two additional sets of trials. One set of the
trials should be for a closure requiring 25 to 35 blows;
one for closure between 20 and 30 blows; and one trial
for a closure requiring 15 to 25 blows.
NOTE Some investigators add saline water to the sediment
or use absorbent material to remove pore water so that
the salinity of the sediment is not changed.
9. Determine the water content, wc, of the soil
specimen from each trial in accordance with the prior
section, making sure that the water contents are
corrected, at least approximately, for salt content.
Make all weighings on the same balance. Initial
weighings should be performed immediately after
completion of the test.
109
-------
10. Plot the relationship between the water content,
wc, and the corresponding number of drops, N, of the
cup on a semilogarithmic graph with the water content as
the ordinate on the arithmetical scale and the number of
drops as the abscissa on the logarithmic scale (Fig.
19). Draw the best-fit straight line through the three
or more plotted points.
11. The test information can be recorded on a form
similar to Figure 20. On the graph, the point at which
the best-fit line intersects the 25-drop abscissa line
corresponds with the liquid limit of the soil.
Computational methods may be substituted for the graphic
method. The liquid limit should be rounded to the
nearest whole number.
Quality Assurance
ASTM states that no interlaboratory testing program has
yet been performed to determine field-wide precision. However,
they presented the precision of the test results as performed by
different individuals in one laboratory.
Within Laboratory Precision for Liquid Limit
Average Value Standard Deviation
Soil A: WL 27.9 1.07
Soil B: WL 32.6 0.98
Cost Analysis
Time required for each drop cup liquid limit analysis
(not including drying or cooling times): 1 h
COST PER SAMPLE AT BATCH RATE $ 20
#########################################t######tt###################
110
-------
50
49
48
c
o
o
g 46
45
44
Liquid limit = 46.4
10
15 20 25 30
Number of drops
35 40
50
60
Figure 19. Water content versus number of drops from a drop cup liquid limit
test (Liu and Evett, 1984, p. 34).
Ill
-------
Pro
\ri-t
Boring »
LIQUID AND PLASTIC LIMIT TESTS
Date
Io. Sample No.
LIQUID LIMIT
Run No.
Tare No.
it
is
Tare plus wet soil
Tare plus dry soil
Water
W
w
Tare
Dry soil
Water content
W
8
V
Number of blows
Water content, w
1
2
3
5 10 20 30
Number of blows
1* 5
TT
• •
PL
6
" " '.'. '. '. Symbol
140
PLASTIC LIMIT
Run No.
4"!
3 a
Tare No.
Tare plus wet soil
Tare plus dry soil
Water
U
w
Tare
Dry soil
Water content
W
w
Plastic limit
Ret
larks
1
2
3
"* 5
rroni
Lty chart
Natural
Water
Content
Technical. rnnpiit.»ri hy i~h"-lc£d by
Figure 20. Typical liquid limit and plastic limit test data form (Dept. of
the Army, 1980, p. 111-19).
112
-------
Test Method B: Fall-Cone Penetrometer
In the following method, the liquid limit is defined as
the water content at which a cone with specific dimensions and
mass penetrates the flat surface of remolded sediment to a
prescribed distance.
Procedure
Applicable standard: BS 1377, 1975, Test 2 (A); British
standard test for liquid limit - cone penetrometer method, British
Standards Institution, London, England.
1. Obtain a representative sediment sample.
2. Thoroughly remold the material in a large evaporating
dish with a spatula.
3. Fill the metal penetrometer cup with sediment, being
careful not to trap air bubbles within the sediment
during the placement procedure.
4. Evenly scrape off any material above the top of the
cup with a spatula or wire saw, leaving a flat sediment
surface.
5. Place the sediment onto the liquid limit device
making sure that the cone tip barely comes into contact
with the sediment surface (Fig. 21).
6. Release the cone and allow it to penetrate the
sediment for exactly 5 sec; record the penetration depth.
7- Remove the sediment cup from the device, clean the
cone, and remold the sediment within the cup. Repeat
steps two through six until the second penetration is
within 0.2 mm of the previous depth.
8. With a small spatula, take a water content subsample
from the zone adjacent to the penetration void.
9. Return the remaining sediment from the penetrometer
cup to the evaporating dish. Change the water content so
that three equally spaced penetrations between 10 and 30
mm are obtained.
10. Plot the water content versus penetration depth for
the three trials on a data form (Fig. 22) and determine
the liquid limit (corrected for salt content) related to
the standard penetration depth for the particular cone in
use. For example, a cone with a mass of 80.00 +/- 0.05 g
113
-------
Figure 21. Fall cone penetrometer and electric timer used for determining
the liquid limit of sediment (the timer automatically stops
penetration after five seconds have elapsed).
114
-------
Date
ATTERBERG LIMITS AND SUMMARY DATA
Sample I.D.
Name
Plasticity Index (Ip) = WL~Wp
Liquidity Index (IL)
Bulk Density (yt)
Void Ratio (e)
Porosity (n)
Dish f
Wt. Dish + Wet Soil
Wt. Dish + Dry Soil
Wt. Dish
Wt. Water
Wt. Dry Soil
Liquid Limit %
Penetration mn
Avg. Penetration
Dish #
Wt. Dish + Wet Soil
Wt. Dish + Dry Soil
Wt. Dish
Wt. Water
Wt. Dry Soil
Plastic Limit 7.
Gs(l+Wc/100)
1+e
- We x Gs
100
1+e
o
LJ
I
<
10 20 30
PENETRATION (mm)
LIQUID LIMIT
PLASTIC LIMIT
SUMMARY OF RAW RESULTS
Liquid Limit
Plastic Limit
Liquidity Index
Plasticity Index
Nat. Water Content
Specific Gravity
SUMMARY OF SALT CORRECTED RESULTS
**********************************************************************************************
Liquid Limit (WL)
Plastic Limit (Wp)
Liquidity Index (IL)
Plasticity Index (Ip)
Nat. Water Content (Wc)
Specific Gravity (G )
Bulk Density (Yt)
Void Ratio (e)
Porosity (n)
*********************************************************************************************
Figure 22. Typical fall-cone Atterberg limits and summary data form.
115
-------
and an apex angle of 30 degrees requires a penetration of
20 mm to define the liquid limit. Round off the liquid
limit value to the nearest whole number.
Quality Assurance
The precision of the fall-cone penetrometer liquid limit
test is not known at the present time, however, it may be more
precise than the drop cup method.
Cost Analysis
Time required for each fall cone liquid limit analysis
(not including drying and cooling times): 1 h
COST PER SAMPLE AT BATCH RATE $20
####################################################################
Discussion
Head (1980) states that although the fall-cone test is no
quicker to perform, it is more dependable because the test
mechanics are based on the remolded static shear strength alone,
without dynamic factors entering the analysis. Head also notes
that the cone method produces more consistent results than the
Casagrande method. Some researchers note that up to liquid limits
of 100 percent, results between the two methods show little
difference (Fig. 23) [Head 1980, p. 67, Wasti and Bezirci 1986] .
Plastic Limit
Introduction
The plastic limit test is typically performed immediately
after the liquid limit test and provides the lowest water-content
value at which a soil behaves plastically in a remolded state.
The plastic limit is determined by first pressing a small portion
of plastic soil together, rolling it into a 3.2-mm (1/8-in)
diameter thread (which gradually removes the water), and repeating
the process until the thread crumbles and can no longer be pressed
together and rerolled. The water content of the soil at this
stage is reported as the plastic limit.
Procedure
Applicable ASTM standard: D4318-84, Standard test method
for liquid limit, plastic limit, and plasticity index of
soils (ASTM 1987, pp. 763-778).
1. Select approximately a 20-g representative portion of
soil from the material prepared for the liquid limit
test. Thoroughly remold the sample.
116
-------
200
100
°c
LIQUID
LIMIT
(B.S.cone
test)
X1
LLB= LLC-
/
y ./
y^X"^
)
/
^ /
s^ ^-Ten
/
otive correlat
curve
s
on
) 100 2OO , 300 400 5C
LIQUID LIMIT (Casagrande test) LLc
Figure 23. Correlation between fall cone and drop cup methods of determining
liquid limits (Head, 1980, p. 67). The following abbreviations
were used; B.S.: British standard; LLgt liquid limit determined
using the British standard; and LLC: liquid limit determined
using the Casagrande drop cup method.
117
-------
2. Change the water content of the soil to a consistency
at which it can be rolled by spreading and remolding
continuously on a glass plate to encourage evaporation.
The drying process may be accelerated by exposing the
soil to the air current from an electric fan or by
blotting with hard surface paper towels or
high-wet-strength filter paper (to avoid adding fiber to
the soil) .
3. From the 20-g mass, select a 1.5- to 2.0-g portion.
Form the test specimen into an ellipsoidal mass. Roll
this mass between the palm or fingers and the
ground-glass plate with just enough pressure to roll the
mass into a thread of uniform diameter along its entire
length. The thread should be further deformed on each
stroke so that its diameter is continuously reduced and
its length extended until the diameter reaches 3.2 +/-
0.5 mm (0.125 +/- .020 in). This should take no more
than 2 min. The amount of hand or finger pressure
required will vary greatly according to the soil. A
normal rate of rolling for most soils should be 80 to 90
strokes per minute, counting a stroke as one complete
motion of the hand forward and back to the starting
position. This rate of rolling may have to be decreased
for very fragile soils.
4. When the diameter of the thread is approximately 3.2
mm, break the thread into several pieces. Squeeze the
pieces together, knead together, reform into an
ellipsoidal mass, and reroll to 3.2 mm. Repeat this
gathering, kneading and rerolling, until the thread
crumbles under the pressure required for rolling and the
soil can no longer be rolled into a 3.2-mm diameter
thread. If crumbling occurs when the thread has a
diameter greater than 3.2 mm, this shall be considered a
satisfactory end point, provided the soil has been
previously rolled into a thread 3.2 mm in diameter.
5. Gather the portions of the crumbled thread together
and place them in a preweighed container. Immediately
cover the container.
6. Select another 1.5- to 2.0-g portion of soil from the
original 20-g specimen and repeat steps three to five
until the container holds at least 6 g of soil.
7. Repeat the full process until a second container,
holding at least 6 g of soil, is prepared.
8. Determine the salt-corrected water content, in
percent, of the soil contained in the containers (refer
118
-------
to that procedural section) and enter all data on a test
form such as Figure 22. Plastic limit results should be
rounded off to the nearest whole number. If either the
liquid limit or the plastic limit cannot be determined,
or if the plastic limit is greater than the liquid limit,
the sediment is nonplastic (NP).
Comments
The test method for determining the plastic limit is
straightforward; however, ASTM standard D4318-84 should be
consulted to ensure that the rate of rolling the thread, time
allowed to perform the test, etc. are properly performed. If the
ASTM standard is not followed exactly (except for the removal of
the 425-micron fraction), the results can be misleading. For
example, a nonplastic soil can appear to exhibit slight plasticity.
Numerous investigators have found that liquid limit and
plastic limit values are significantly affected by the amount of
organic matter that is present in the sediment (Booth and Dahl
1986). Typically, an increase in organic content increases both
the liquid and plastic limits. Therefore, organic content should
be measured and reported for those sediments suspected of
containing significant amounts of organic matter.
Quality Assurance
ASTM states that no interlaboratory testing program has
as yet been performed to determine field-wide precision. However,
the precision within one laboratory of the test method performed
by different individuals is as follows
Within Laboratory Precision for Plastic Limit
Average Value Standard Deviation
Soil A: Wp
Soil B: Wp
21.9
20.1
1.07
1.21
Cost Analysis
Time required for each plastic limit determination (not
including drying or cooling times): 30 min
COST PER SAMPLE AT BATCH RATE $10
#####################t#t##ttt#tt#t#t#####ttt#tt###f###f#t#ttt#ttttt##tt#tt##
119
-------
References
American Society for Testing and Materials, 1987. Annual
book of standards: soil and rock, building stones and
geotextiles, ASTM, Philadelphia, PA, vol. 04:08, 1189 pp.
British Standards Institution, 1975. Standard test for
liquid limit-cone penetrometer method, London, England.
Booth, J. S. and Dahl, A. G., 1986. A note on the
relationships between organic matter and some
geotechnical properties of a marine sediment, Marine
Geotechnology, vol. 6, no. 3, pp. 281-297.
Head, K. H., 1980. Manual of soil laboratory testing,
volume 1: soil classification and compaction tests,
Pentech Press, London, England, 339 pp.
Hunt, R. E., 1984. Geotechnical engineering
investigation manual, McGraw Hill, New York, NY, 983 pp.
Liu, C. and Evett, J.B., 1984. Soil properties testing,
measurement, and evaluation, Prentice-Hall, Englewood
Cliffs, NJ, 315 pp.
Peck, R. B., Hanson, W. E. and Thornburn, T. H., 1974.
Foundation engineering, John Wiley, New York, NY, 514 pp.
US Deptartment of the Army, 1980. Engineer Manual EM
1110-2-1906: laboratory soils testing, Washington, DC,
388 pp.
Wasti, Y. and Bezirici, M. H., 1986. Determination of
the consistency limits of soils by the fall cone test,
Canadian Geotechnical Journal, vol. 23, no. 2, pp.
241-246.
120
-------
CHAPTER 13
GRAIN SPECIFIC GRAVITY
Introduction
The specific gravity of soil is defined as the ratio of
the mass of a unit volume of a material at a stated temperature to
the mass in air of the same volume of gas-free distilled water at
a stated temperature (ASTM 1987, p. 210). The grain specific
gravity can be used, in conjunction with the salt-corrected water
content, to estimate in situ overburden stresses as well as the
possible presence of certain minerals.
Two methods are presently used to determine the grain
specific gravity. The traditional method (see A below), involves
filling a glass pycnometer with sediment and distilled water;
removing the entrapped air (possibly by boiling); cooling to room
temperature; measuring the water's temperature; weighing the
device; and finally, determining the grain specific gravity after
correcting the density of the water for temperature. A newer
method (see B) relies on a self-contained apparatus that simply
involves weighing the sample and placing it into the pycnometer to
measure volume (some devices are completely automated and measure
volumes of five samples simultaneously), and then making a simple
calculation.
Test Method A: Water-filled pychnometer
Procedure
Applicable ASTM Standard: D854-83, Standard test method
for specific gravity of soils (ASTM 1987, p. 210-213).
1. Calibrate pycnometer.
2. Place a sediment sample in the pycnometer. If a
volumetric flask is used, the sample should have a mass
of at least 25 g; if a stoppered bottle is used, the
sample should have a mass of at least 10 g.
3. Add sufficient distilled water to fill the volumetric
flask about three-fourths full or the stoppered bottle
about half full.
4. Remove entrapped air by subjecting the contents to a
partial vacuum or boiling gently for at least 10 min,
occasionally rolling the pycnometer to assist in the
removal of the air. Subject the contents to reduced air
pressure either by connecting the pycnometer directly to
an aspirator or vacuum pump, or by using a bell jar.
Note: some soils boil violently when subjected to reduced
121
-------
air pressure. If that happens reduce the air pressure at
a slower rate or use a larger flask. Allow heated
samples to come to ambient temperature before proceeding
with the analysis.
5. Fill the pycnometer with distilled water, clean the
outside, and dry with a clean, dry cloth. Determine the
weight of the pycnometer and contents, and the
temperature of the contents. Calculate the specific
gravity of the soil according to the ASTM standard
D854-83 and record the information on a form similar to
Figure 24. Grain specific gravity should be reported to
two decimal places.
Quality Assurance
ASTM has determined the precision of the test for
cohesive soils as:
Standard Acceptable
Deviation Difference *
Single-operator precision 0.021 0.06
Multilaboratory precision 0.056 0.16
*The difference between the results of two properly
conducted tests should not exceed the acceptable
difference.
Cost Analysis
Time required for each test: 2 h
COST PER ANALYSIS AT BATCH RATE $40
################################################################tt##<
Test Method B: Gas-pressurized pychnometer
Procedure
1. Grind an oven dried sample to a fine sand-sized
powder using a mortar and pestle.
2. Place the sample in a small evaporating dish or water
content tin and leave it in an oven at a temperature
between 60 and 110 °C for a minimum of 8 h.
3. Remove the sample from the oven and let it cool to
room temperature in a desiccator so that moisture in the
air won't be adsorbed by any clay minerals.
122
-------
Soils Testing Laboratory
Specific Gravity Determination
Sample No..
Boring No. .
Depth
Description of Sample
Tested by
Project No.
Location
Date.
[A] Calibration of Pycnometer
(1) Weight of dry, clean pycnometer, Wp
(2) Weight of pycnometer + water, Wpw _
(3) Observed temperature of water, 7~/
[B] Specific Gravity Determination
_g
°c
Determination No.:
Weight of pycnometer + soil + water,
Wpws (g)
Temperature, Tx (°C)
Weight of pycnometer + water at 7"x,
Wpw(a*Tx)(g)
Evaporating dish no.
Weight of evaporating dish, W0 (g)
Weight of evaporating dish + oven-dried
soil, Wds (g)
Weight of solids, Ws (g)
Conversion factor, K
Specific gravity of soil
KWS
"* Ws + Wpw(*\Tx)-Wpws
1
2
3
Figure 24. Typical data form for grain specific gravity determined using the
water filled pycnometer method (Liu and Evett, 1984, p. 23).
123
-------
4. Place the sample into a pycnometer cup of known mass
and place it into the pycnometer (Fig. 25).
5. Determine the volume of the soil grains according to
the manufacturer's instructions.
6. Remove the sample and cup from the pycnometer and
determine the mass of the sample and cup.
7. Specific gravity is determined by dividing the mass
of soil grains by the volume of soil grains and is
reported to two decimal places.
8. Correct the specific gravity for salt content by
using Table 10. The calculations can be made on a form
similar to Figure 26. The corrected grain specific
gravity can also be determined from the following
equation, using any salinity or salt density values:
Jsc
Ms
("
- I s
1 1000 - S
S
1000 - S
X W X Mg )
x w x MS \
) P
Ps /
where: Gsc = grain specific gravity corrected for a particular
salinity and salt density value,
Ms = mass of sediment (including salt) of pycnometer
sample,
Vs = volume of sediment (including salt) of pycnometer
sample,
S = pore-water salinity (in ppt),
w = water content (in decimal form) not corrected for
salt content,
Ps = sea-salt density (typically 2.18 g/cm^), and
pw = distilled water density at a temperature of 4 °C
(1 g/cm3).
Quality Assurance
Manufacturer's listing of the accuracy of volume
determination for particular pycnometers range from +/- 0.1 to 0.2
percent or +/- 0.05 cc. Check the manufacturer's information to
determine the precision of a particular pycnometer.
124
-------
Figure 25. Gas pressurized pycnoraeter.
125
-------
PYCNOMETER
SAMPLE ID
WEIGHT
MEASURED
VOLUME
TARE
CORRECTED
VOLUME
S.G.
RAW SALT COR
SALT COR.
W.C.
Figure 26. Form for recording data and determining grain specific gravity by
gas pressurized pycnometer.
126
-------
Table 10. Corrections to add to raw specific gravity to account
for a salinity of 35 ppt in the pore water. ~~
Salt-corrected Add to raw
water content (percent) specific gravity
18 to 67 .01
68 to 116 .02
117 to 165 .03
166 to 214 .04
Cost Analysis
Time required for each test (not including drying and
cooling times): 5 to 20 min.
COST PER ANALYSIS AT BATCH RATE $10
t#ttttt#######f############tt#ttt######tt####tt##########tt################
Discussion
The use of a pressurized gas pycnometer (Method B) is
much preferred because the precision is greater and the time
required to perform each test is less than that required by the
ASTM method.
If using Method B, extreme care must be exercised to
minimize the sample's exposure time to air after it has been
removed from the oven. Some clay minerals quickly attract
moisture in the air to their surfaces, thereby decreasing the
measured specific gravity values.
References
American Society for Testing and Materials, 1987. Annual
book of standards: soil and rock, building stones and
geotextiles, ASTM, Philadelphia, PA, vol. 04:08, 1189 pp.
Liu, C. and Evett, J.B., 1984. Soil properties testing,
measurement, and evaluation, Prentice-Hall, Englewood
Cliffs, NJ, 315 pp.
127
-------
CHAPTER 14
LABORATORY VANE SHEAR STRENGTH
Introduction
The miniature vane shear test is performed to determine
an approximate value of the undrained shear strength in
fine-grained soil. The sensitivity (that is the ratio of natural
undrained shear strength divided by the remolded undrained shear
strength) can also be calculated. The test consists of inserting
a four-bladed vane into the sediment, rotating the shaft connected
to the blades, and measuring the torque required to shear the
sediment. By assuming a particular failure surface within the
soil, the undrained shear strength (su) can be calculated.
Although ASTM (1987) has a standard for vane shear
testing in the field, it does not yet have one for laboratory
testing although such a standard is presently under review. The
related method D2573-72, standard test method for field varre shear
test in cohesive soil, is currently being revised. Some
information is pertinent to both types of tests.
Procedure
1. Make sure that the core section and vane shear
machine (Fig. 27) are securely positioned so that neither
will move during testing.
2. Take an initial reading on the rotation dial or set
the initial reading to 0°.
3. Insert the vane into the sediment so that the top of
the vane is about one vane height below the sediment
surface. The center of the vane should be at least 1.5
vane diameters away from any liner surface or wall.
4. Rotate the vane, or the spring top, at a rate of
90° per minute until a peak torque is reached. Record
the peak value on a data form similar to Figure 28.
5. Remove the drive belt and remold the sediment by
rotating the vane by hand rapidly through at least one
revolution. Another, more time consuming, method is to
remove the sediment and physically remold it within a
plastic bag to avoid entrapping air. Carefully replace
the sediment into a container and re-insert the vane.
6. Reattach the belt (if necessary), rotate the vane,
and read the peak torque. Record the peak value as the
remolded peak on the data form.
128
-------
Figure 27. Vane shear machine that uses a spring to apply torque to the vane,
129
-------
CRUISE .
CORE ID.
LAT.
VANE SHEAR DATA
LONG.
GEN. LOCATION
RECOVERED LENGTH
OPERATOR'S NAME
PENETRATION
SPRING SERIES
AND NUMBER
NATURAL SH
DEGREES
EAR STRENGTH
SuCkPa)
REMOLDED £
DEGREES
5HEAR STRENGTH
Sr-(kPo)
StCSu/Sr)
DESCRIPTION AND EVIDENCE
OF DISTURBANCE
Figure 28. Typical vane shear data form.
130
-------
7. Extract the vane.
8. Remove a 10- to 20-g water-content sample from the
zone where the vane test was run.
9. Insert a nonwater-absorbing plug in the resulting
hole.
10. The undrained shear strength, su, can be determined
from the following equation:
4T
su =
d2h + 0.667*rd3
where;
T = measured torque (determined from equipment
calibration),
d = diameter of vane, and
h = height of vane.
The factor, 0.667, represents uniform end shear
resistance at the top and bottom of the vane. Other
assumptions regarding end shear resistance changes the
factor from 0.5 for a triangular distribution to 0.6 for
a parabolic distribution (Bowles 1979, p. 381).
11. Determine the water content of the sediment.
12. Record the results on a form similar to Figure 28.
Shear strength values should be recorded to the nearest
0.1 kPa.
Comments
The vane shear test is relatively simple to perform.
However, any of a number of operational errors can result in
seriously skewed results. To keep the accuracy as high as
possible, consider the following points:
(1) To ensure that undrained conditions prevail, the vane
shear test should only be performed in fine-grained sediment,
using a rotation rate of about 90° per min. Coarse-grained
sediment should not be tested because they can lose much of their
confining stress prior to testing, thereby causing laboratory
strength determinations to be low.
(2) To ensure valid test results a soil must exhibit
plastic behavior, contain less than 15 percent very fine sand, and
must not display drainage or tension cracks during shear.
131
-------
(3) Lee (1985) summarizes the current uses and
limitations of the laboratory vane shear test, and suggests that
plastic soils with a liquid limit above 30 percent can be tested.
(4) Be sure vane is placed in an area of the core where
it will not contact any gross unconformities, i.e. clasts, shells,
gravel or sand lenses. Sharp climbs in peak torque values may be
an indication that the vane has contacted the liner or a clast
(e.g., shell or rock fragment) within the matrix.
(5) Sediment with shear strengths greater than 100 kPa
should not be tested with the vane shear machine because failure
conditions deviate significantly from the assumed mode (Noorany
1985) .
(6) Sample disturbance and improper storage can severely
affect measured strength values. Therefore, methods that impart
little disturbance must be used to obtain, transport, and store
sediment cores (see Chapter 2).
Vane blade size is not very important although a blade of
equal height and diameter (e.g. 12.7 mm x 12.7 mm) or one with
height equal to twice the diameter is often used. In very soft
sediment, a larger (e.g. 25.4 mm) blade should be used to increase
the measured torque and, thereby, the accuracy of the measurement.
Most vane shear machines use a spring to apply torque to
the vane. Recently torque sensors have been used to apply and
measure the torque (Fig. 29). The torque sensor method possesses
some advantages over the traditional method: it is able to provide
a hard copy of stress versus time plot; it measures post-peak
behavior; and, it turns the vane at a constant rate.
Spring-mounted systems tend to turn the vane at an extremely slow
rate at first, then rapidly increases the rate as failure is
approached. However, few data exist to indicate that the added
expense of a torque-sensor system is justified by increased
accuracy.
Two other methods of rapidly determining undrained shear
strength are occasionally used: the torvane (Fig. 30) and pocket
penetrometer (Fig. 31). Although the torvane is typically more
accurate than the pocket penetrometer, the laboratory vane shear
test is superior to both and should be performed whenever possible.
Quality Assurance
Although methods to determine accuracy of the laboratory
vane shear test have not been formulated, a precision of 0.5
degree of spring rotation can be obtained. However, the test must
be conducted carefully because some vane shear machines have up to
a 2 degree "play" that must be eliminated.
132
-------
Figure 29. Vane shear machine that uses a torque sensor to rotate the vane
(middle: torque sensor signal conditioner, right: strip chart
recorder).
133
-------
-
JOT
Figure 3d. Torvane shear strength device (center), soft sediment adapter
(left), and stiff sediment adapter (right).
134
-------
Figure 31. Pocket penetrometer and soft sediment adapter.
135
-------
Cost Analysis
Time required for each analysis (not including drying or
cooling times): 15 min.
COST PER SAMPLE AT BATCH RATE $20
########t#t#t###tt##########t###################################t####
References
American Society for Testing and Materials, 1987. Annual
book of standards: soil and rock, building stones and
geotextiles, ASTM, Philadelphia, PA, vol. 04:08, 1189 pp.
Bowles, J. E., 1979. Physical and geotechnical
properties of soils, McGraw-Hill, New York, NY, 478 pp.
Lee, H. J., 1985. State of tne art: laboratory
determination of the strength of marine soils, in Chaney,
R. C. and Demars, K. R. , eds., Strength testing"~of marine
sediments: laboratory and in situ measurements, American
Society for Testing and Materials Special Technical
Publication 883, Philadelphia, PA, pp. 181-250.
Noorany, I., 1985. Laboratory soil properties, ir\
Rocker, Karl, Jr., ed., Handbook for marine geotechnical
engineering, US Naval Civil Engineering Laboratory, Port
Hueneme, CA, 19 pp.
136
-------
CHAPTER 15
ONE-DIMENSIONAL CONSOLIDATION
Introduction
The constant-rate-of-strain consolidation (CRSC) test is
performed to evaluate the laterally confined one-dimemsional
stress-strain properties of a cylindrical wafer of sediment. Test
results can be used to determine the stress history (maximum past
stress) and the rate at which consolidation occurs. Typically, a
test is performed in four parts: saturation, load, rebound and
reload. Results from this test are often used to predict the
amount and rate of settlement of a proposed engineering structure.
Procedure
Applicable ASTM standard: D4186-82, Standard test method
for one-dimensional consolidation properties of soils using
controlled-strain loading (ASTM 1987, pp. 709-715).
1. Flush the equipment lines with de-aired water to
de-air the porous stones.
2. Place the test specimen into the CRSC system's
confining ring, either by carefully trimming a sediment
sample to the correct dimensions or by pushing a cutting
ring into the sediment. Softer sediment typically
requires the latter technique.
3. Trim the sample to the correct height with a wire saw
and fill any irregularities in the sample with material
from the trimmings. Obtain water content samples from
the top, middle, and bottom trimmings.
4. Determine the weight and dimensions of the sample and
record the data on a form similar to Figure 32.
5. Place the sample on the machine pedestal and assemble
the confining apparatus, including filter papers and
porous stones (Fig. 33).
6. Place the top cap on the sample. The top cap should
be made of a strong, light-weight material. This is
extremely important when testing very soft marine
sediment.
7. Assemble the chamber and fili it with de-aired water.
137
-------
CONSTANT RATE OF STRAIN CONSOLIDATION TEST
Test I.D./File Name:
Date/Start:
Tested By:
Project Title:_
(Cruise No.)
Core 1.0.-.
Subsectioned Interval (m):_
Test Sample Interval (m):
System Number:
t Readings:
Raw Disc:
End:
Rate of Feed (mm/mln):
Gearbox Lever (A-E):
Cell Press. (kPa):
Back Press. (kPa):
Reduced Disc:
PRE-CONSOLIDATION SPECIMEN DATA
Diameter (mm)
Height (mm)
Wt. cutting ring +
filter papers + sample (g)_
Wt..cutting ring +
filter papers (g)
Wt. wet sample (g)
SAMPLE DESCRIPTION AND COMMENTS
WATER CONTENT FROM TRIMMINGS
top side bottom
Container ID
Wt. wet soil +
container (g)
Wt. dry soil +
container (g) _____
Wt. water (g)
Wt. container (g)
Wt. dry soil (g)
Water content (%)
Average wc (%)
POST-CONSOLIDATION SPECIMEN DATA
Height (mm)
Container I.D.
Wt. dry sample + container (g)
Wt. container (g)
Wt. Dry sample (g)
Wt. water (g)
Water content (%)
Pre-consol wc (%)
Wt. wet sample + container (g)
Wt. dry sample + container (g)
Wt. water (g)
Wt. dry sample (g)
Water content (%)
Post-consol wc (%)
Figure 32. Typical CRS consolidation test data form.
138
-------
load
loading piston
\jn t ^
oilfiller plug
load
transducer
perforated
load platen
porous disc
cutting ring
seals
pore pressure
measurement
point
to back
pressure system
back
pressure
transducer
load frame
platen
i
differential
pressure
transducer
to pore »
pressure panel
Figure 33. Typical constant-rate-of-strain consolidation (CRSC) sample and
test chamber configuration (Head, 1986, p. 1208).
139
-------
8. Connect the load, deformation, pore-pressure, and
cell pressure (or differential transducer) measuring
devices. Check that systems are operating correctly and
that measurements are within bounds.
9. Make sure the sample does not swell. This is done
either by limiting potential vertical deformation or by
applying just enough seating stress to counteract the
swelling tendencies of the sediment. It is extremely
important to not overload the sample at this point.
10. Fully saturate the sediment interstitial pore spaces
and equipment lines with water by dissolving any
remaining bubble-phase gas. This is accomplished by
elevating the cell pressure (often to 300 kPa) without
allowing any pore fluid drainage. Determine a
pseudo-B-coefficient if possible. The B-coefficient
(change in pore pressure divided by change in chamber
pressure) indicates complete saturation if the value is
1.00- Lower values typically represent partially
saturated soils that may require higher stresses to
ensure adequate saturation. An attempt should be made to
reach full saturation, but, if this cannot be achieved,
make sure that the sample is at least nearly saturated (B
is less than 0.95) before continuing.
11. Let pore pressures within all systems and components
equilibrate, often overnight.
12. Vertically strain the sample at a rate that will
produce a change in pore pressure that is between 3 and
20 percent of the applied vertical stress at any time
during the test. The strain rate may be adjusted during
the test if it appears that the excess porewater pressure
will not fall within those limits.
13. At discrete intervals, record load, deformation, cell
pressure, and pore pressure response during the test.
14. If required, put a rebound curve into the test to
determine unloading and recompression characteristics.
15. Continue to apply load until the capacity of the
system is reached or until no further information is
required.
16. Remove the sample from the testing device; determine
dimensions, mass, and water content. Record information
on a form similar to Figure 32.
140
-------
17. Determine compression characteristics from the ASTM
standard or other related articles dealing with this test
(e.g., Lambe 1951, Lambe and Whitman 1969). The
following plots should be generated for each CRSC test
with the vertical effective stress (kPa, logarithm scale)
as the abscissa: (1) void ratio, (2) excess pore pressure
(kPa), (3) excess pore pressure divided by the total
vertical stress (percent), (4) coefficient of
consolidation (cm^ per sec), and (5) coefficient of
permeability (cm per sec). Test information can be
summarized on a data sheet similar to Figure 34.
Comments
Although the CRS consolidation test is not often used for
coarse-grained material, it is applicable to all fine-grained
sediment. Very soft marine sediment, however, typically presents
additional problems. When handling those samples, care must be
exercised to avoid sediment deformation during the trimming
process. Also, because normally consolidated and
under-consolidated shallow-subbottom samples have typically
experienced very low maximum past stresses, the sample must not be
overloaded during the initial stages of the test. The piston
bushings in the chamber must possess minimal friction or severe
overestimations of the maximum past stress could result. To
ensure the best possible test results, the sediment should be
sampled, handled, transported, and stored in a manner that will
minimally disturb the samples (see Chapter 2).
Note A less desirable method, using an oedometer
apparatus, is sometimes used to determine consolidation properties
of material. Because it utilizes an incremental loading schedule
that requires complete dissipation of excess pore pressure between
loadings, it is a very time-consuming test and often requires
weeks to perform. The incremental loading test only gives one
data point for each applied load (a load increment takes 12 to.24
h to complete), whereas the CRS consolidation test presents an
almost continuous void ratio-stress curve. Within the above
limitations, the incremental loading system is still adequate for
testing most soils and it is, indeed, the more traditional
approach. The ASTM standard for the incremental test is D2435-80,
standard test method for one-dimensional consolidation properties
of soils (ASTM 1987, pp. 388-394).
Quality Assurance
ASTM states that undisturbed soil samples from
homogeneous soil deposits at the same location often exhibit
significantly different consolidation properties. Because of
sample variability, no method exists to evaluate the comparative
precision of various consolidation tests on undisturbed samples.
141
-------
CONSTANT RATE OF STRAIN CONSOLIDATION TEST RESULTS
Test I.D./File Name:
Project/Cruise: Date Core Obtained:
Location of Core: Lat.; Long.;_
Core Retrieval:
Shape & Dimensions:_
Method of Shipping & Handling:_
Storage:
Temperature:
Problems in Handling/Storage:_
Boring/Core ID:
Extruded Sample Increment (m):
Type of Material:
Tested Sample Increment (m):_
Problems/Comments of Test:
LVDT:
Validity/Discrepancies of Test:
Frame: Load Cell: u Trans.:
Test Performed By:
Data Reduction By:
Checked By:
Raw Disc: Reduced Disc:
Date of Consol. :_
Date:
Date:
Saturation Pressure (kPa):
Strain Rate (mm/min):
Time for Consol:
B Coefficients; Initial:
Final:
Bulk Density Before Consolidation (kNm):
Heights (cm); Initial: Final:
(day/hour/min):
Area (cm^):
Water Content (%); Trimgs:_
Calc. Init.:
Final:
Ave. w Above Test Sample (%):
Ave. G Above Test Sample:
; Meas:
; Assumed
; Meas:
; Assumed:
Average Effective Unit Wt (kN/m3):_
o'VQ (kPa): ; o'^ (kPa) :
_; Casagrande:
; Other:
o'e (kPa):_
; OCR:
Figure 34. Constant-rate-of-strain consolidation test summary form.
142
-------
Test No:
cc
Cr
cv
k
JD
(lab); max: ; ave: ; min:
(lab):
(cm2/sec) @ o'vo: ; @ o'^:
(cm/sec) @ <;'„„: : @ a'^:
= el - e2 _
el - e3
e.: ; 0.42 e^:
Disturbance Index. ID Degree of Disturbance
.15-. 30 Sm»ll amount of disturbance
.30-. 50 Moderate disturbance
.50-, 70 Much disturbance
.70 Extreme disturbance (remolded)
; Cr (field):
; ave. virgin:
; ave. virgin:
t
£
*
5
^ T_t
rrr.t=.
!•( t)Eltl 1
Likwiivt — ^>
CMvntiHv LM*
6i
CHDCIM
C/M""
r-+
i \ *••
041.. -^*x
Curve Type: normal , sensitive , remolded , continuous curve_
rebound changes Cc slope: no , yes
Test Suite Results: averaged , weighted
o'yo (kPa):
o' (kPa): OCR:
°'vm :-
Cc (lab): max:_
Cr (lab):
, ave:
, min:
_; Cc (field):_
cy (cm2/sec) @ o'vo:_
k (cm/sec) @ o' 0:
, ave. virgin:
@
_> *•
_, ave. virgin:
Comments/Notes:
Figure 34. (cont). Constant-rate-of-strain consolidation test summary form.
143
-------
A suitable test material and method of sample preparation
have not been developed for determining laboratory variances due
to the difficulty in producing identical cohesive soil samples.
Therefore, no estimates of precision for this test method are
available.
Cost Analysis
Time required for each consolidation test (not including
drying or cooling times): 2 to 4 days
COST PER CONSOLIDATION TEST $400
####################################################################
References
American Society for Testing and Materials, 1987. Annual
book of standards, soil and rock, building stones and
geotextile, ASTM, Philadelphia, PA, vol. 04:08, 1189 pp.
Head, K. H., 1986. Manual of soil laboratory testing,
volume 1: soil classification and compaction tests,
Pentech Press, London, 339 pp.
Lambe, T. W., 1951. Soil testing for engineers, John
Wiley, New York, NY, 165 pp.
Lambe, T. W. and Whitman, R. V-, 1969. Soil mechanics,
John Wiley, New York, NY, 553 pp.
144
-------
CHAPTER 16
STATIC CONSOLIDATED-UNORAINED TRIAXIAL COMPRESSIVE STRENGTH
Introduction
The triaxial test measures the drained and undrained
stress-strain properties of soil. A right-circular cylinder of
sediment is enclosed in a watertight membrane within a
fluid-filled test chamber. After saturation of entrapped
bubble-phase air into the pore water is completed, radial and
vertical stresses on the sample are elevated and consolidation is
allowed by permitting drainage. After consolidation is finished,
the sample is vertically loaded at a constant strain rate until
failure (typically 15 percent strain) is reached. This
predetermined strain level will, in most cases, allow the sample
to reach its peak strength. While sample compression is
progressing, either the operator or an automatic data acquisition
system is recording axial load, axial deformation, pore pressure,
and cell pressure.
Although three main types of triaxial tests are performed
[unconsolidated-undrained (UU), consolidated-drained (CD), and
consolidated-undrained (CU)], the test method discussed here
pertains specifically to the CU test. However, with only slight
modifications in the testing procedure, the other two types of
tests can be run.
The shear strength of sediment in triaxial compression
depends on the stresses applied, the time allowed for
consolidation, the strain rate, and the stress history of the
soil. In this test, strength is measured under undrained
conditions, and the test is applicable to field conditions where
soils that have fully consolidated under one set of stresses are
subjected to a rapid stress change without time for drainage to
occur. Data from the test can be used to determine soil
characteristics in terms of total or effective stresses.
Procedure
ASTM does not have a standard for the
consolidated-undrained test, although one is presently in review.
However, it does give a test standard for the
unconsolidated-undrained test (D2850-82, standard test method for
unconsolidated, undrained compressive strength of cohesive soils
in triaxial compression (ASTM 1987, pp. 451-456).
1. Apply silicone grease to the triaxial-chamber bottom
pedestal and to the sample's top cap.
145
-------
2. Flush all system lines with de-aired water. Some
investigators use salt water in the pore pressure lines;
however, that fluid has a severe corrosive effect on most
metal fittings. De-air the porous stones by boiling.
3. Trim the sediment sample to the appropriate
dimensions: its height should be approximately twice its
diameter. For most sediment, a standard soil lathe can
be used for trimming. However, some extremely soft
marine sediments will deform under their own weight if
left standing. For those sediments, a miniature
thin-walled piston sampler can be used to obtain a
relatively undisturbed sample (Winters 1987) . In
operation, the piston is held fixed at the sediment
surface while the thin-walled tube, having an inside
diameter equal to the outside diameter of the test
sample, is pushed into the sediment. Although this
sampling technique disturbs the sediment somewhat, the
procedure does allow otherwise unsuitable sediment to be
tested. Record applicable information on a data form
similar to Figure 35.
4. Quickly place the sample on the triaxial machine
pedestal, making sure that the top and bottom porous
stones and filter papers are in the correct position.
5. Place the top cap, made of a strong, light-weight
material, on the sediment.
6. Place radial filter paper drains on the sample.
7. Place a thin membrane over the sample and seal both
the bottom pedestal and top cap with two 0-rings.
8. Assemble the chamber and fill it with de-aired water
(Fig. 36) .
9. Connect all measuring devices (load cell, strain
gauge, pore- and cell-pressure measuring devices or
differential transducer) to the chamber after ensuring
that they are operating correctly-
10. Slowly saturate the sample by simultaneously or
alternately increasing the cell pressure and back
pressure in less than 50-kPa increments. Do not saturate
in increments greater than the final consolidation
stress. Final back pressure should be at least 300 kPa.
11. After saturation is complete, usually overnight, as
indicated by a B coefficient (change in pore pressure
divided by the change in cell pressure) greater than
0.95, allow all stresses to equilibrate.
146
-------
Test ID/
Consol. File Name:
Project Title:
(Cruise No.)
Core ID:
Subsectioned Interval (m):
Test Sample Interval (m) :
System Number:
Raw Disc:
Reduced Disc:
Consol . Rdgs . : -
TRIAXIAL DATA SHEET
Test ID/
Shear File Name:
Date/Start: End:
Tested By:
Rate of Feed (mm/min):
Gearbox Lever (A-E):
Cell Press. (kPa):
Back Press. (kPa) :
Consol. Stress (kPa):
Shear Rdgs. ; -
Filter papers:
Membrane:
"B" value from printout:
Final Dvol rdg. (cc):
Initial Dvol rdg. (cc); ~
Total water expelled (cc):
Trimmed Diameter (mm):
Trimmed Ht. (mm):
Piston Factor (mm):
Piston Ht. (mm):
INIT RDGS
PP kPa:
DL mm:
AX kN:
CP kPa:
SHEARED SAMPLE
Init. Ht. (PH-PF) (mm):
Piston Ht. (mm): ~
Calc. Post-Consol Ht. (mm):
Piston Ht. (mm):
Calc. Post-Shear Ht. (mm):
Post-Consol Ht. (LVDT) (mm):
Meas. Post-Shear Ht. (mm):
POST SHEAR
CP -
PP -
C-P -
WATER CONTENT FROM TRIMMINGS
Top Side Bottom
Container ID
Wt. wet soil +
container (g)
Wt. dry soil +
container (g)
Wt. water (g)
Wt. container (g)
Wt. dry soil (g)
Water content (g)
wc salt corrected(%)
Average wc (%)
SAMPLE DESCRIPTION
PRE-SHEAR SPECIMEN DATA
Wt. 2 f.p. + wet sample (g)
Wt. 2 f. papers (g)
Wt. wet sample (g)
Wt. dry sample (below) (g)
Wt. water (g)
Water content (%)
POST-SHEAR SPECIMEN DATA
Container ID
Wt. wet sample +
membrane + cont. (g)
dry sample + cont.
membrane (g)
container (g)
wet sample (g)
dry sample (g)
water (g)
Wt.
Wt.
Wt,
Wt.
Wt,
Wt.
Water content (%)
(8)
COMMENTS AND OBSERVATIONS
Figure 35. Typical triaxial test data form.
147
-------
post and bracket for
strain dial gauge
stem
piston bushing
air bleed plug
PISTON
oil filler valve
or plug
membrane
BASE PEDESTAL
\ \gland
0-ring seal
CELL BASE
Figure 36. Typical components of a triaxial test device (Head, 1986, p. 801).
148
-------
12. Consolidate the sample to the required stress by
elevating the cell pressure above the back pressure and
permitting drainage. Sometimes, especially if large
consolidation stresses are to be applied to very soft
sediment, the final consolidation state is reached by
alternately increasing the cell pressure and allowing
drainage between increments. Plot the volume change of
the sample according to the log-time or square-root-time
method (Bishop and Henkel 1962, and Department of the
Army 1980). If the log-time method is used, allow
consolidation to continue for at least one log cycle of
time (or overnight) after primary consolidation has
ceased. If the square-root-time method is used,
consolidation should continue for at least 2 h after
primary consolidation has ceased.
13. Close the drainage valve and shear the sample at a
constant rate such that pore-pressure equalization occurs
throughout the sample. For fine-grained sediment, an
appropriate strain rate typically will cause 15 percent
strain to occur after several hours. If a test is
performed at too fast a rate, severe pore-pressure
measurement inaccuracies could result.
14. Measure and record load, deformation, and pore- and
cell-pressure (or differential transducer) readings
throughout the test.
15. Continue loading until 15 percent axial strain occurs.
16. Remove the sample from the chamber. Record the
dimensions and mass on a data form similar to Figure 35.
17. Perform required calculations and plot data as
specified (Bishop and Henkel 1962, Head 1986). As a
minimum, the following plots should be produced: (1)
volume change during consolidation (cc) versus the square
root or logarithm of time (min); (2) q (o^ minus 03)
divided by 2; (3) q (kPa) versus strain (percent); and
(4) change in pore pressure (kPa) versus strain
(percent). The results can be summarized on a form
similar to Figure 37.
Comments
The consolidated-undrained triaxial compressive strength
with pore pressure measurement test is a valuable tool. In
addition to determining static strength characteristics that could
be used for total stress analyses such as waste package seabed
penetration, the test can also be used to measure drained
parameters that are useful for analyzing slower in situ shear
mechanisms where sufficient time is available for complete
pore-pressure dissipation.
149
-------
TRIAXIAL SAMPLE AND TEST DATA
Test No.:
Project/Cruise:
Location of Core:_
Core Retrieval:
Date Core Obtained:
La t.: Long.:
Shape & Dimensions:
Method of Shipping & Handling:_
Storage:
Temperature:
Problems in Handling/Storage:
Boring/Core ID:
Extruded Sample Increment (cm):_
Type of Material:
Type of Test:
Tested Sample Increment (cm):_
Problems/Comments, of Test:_
Validity/Discrepancies of Test:
Frame: Load Cell: u Trans.:_
Test Performed By:
Data Reduction By:
Checked By:
Raw Disc:__
Type of Consolidation:
LVDT:
DVOL:
Date of Shearing:_
Date:
Date:
Reduced Disc:
Maximum Past Vertical Stress (kPa):
Back Pressure(kPa}:
Radial Consolidation Pressure (kPa):
Vertical Consolidation Pressure (kPa):_
Induced OCR: OCR Based On:_
Strain Rate (mm/min):
Type of Membrane:
Determined By:
Time of Shearing (min):
Type 5, Material of Drain(s):
Thickness (cm):__
Thickness (cm):
Figure 37- Consolidated-undrained triaxial compressive strength test summary
form.
150
-------
Test No:_
Height/Diameter; Trimmed: Tested:
Membrane Correction Applied: Filter Drain Correction Applied:
Bulk Density Before Consolidation (kNm3):
Heights (cm); Initial: Consolidation: Final:
Water Content (%); Initial: Consolidation: Final:_
Volume (cm3); Initial: After Consolidation:
Area (cm2); Inital: ; Consol. Area (cm2)=Ac=Consol. volume"
Consol. Height
B Coefficients; Before Consolidation:
After Consolidation: ,
Before Shearing:
At Failure; q (kPa): p1 (kPa):_
A Coefficient:
Change in Pore Water Pressure (kPa):
Axial Strain (%):
Type of Failure:
$' (maximum) (degrees): Su/p1 - c/p':
<(>' (at maximum q) (degrees):
$' (at peak-max. obi.) (organlcs only) (degrees):
Comments/Notes:
Figure 37. (cont). Consolidated-undrained triaxial compressive strength test
summary form.
151
-------
Almost all deep-sea sediments can be tested using the
Procedures described above. However, particles that are greater
than 1/6 the diameter of the test sample must not be present prior
to testing. To insure that the best possible test results are
obtained, the sediment should be sampled, handled, transported,
and stored in a manner that will minimally disturb the samples
(see Chapter 2).
Testing soft sediment presents special problems. Extreme
care must be exercised in the handling and trimming of the
samples. Friction in the top-cell piston bushing should be
minimal or severe strength overestimations can result. Thin
membranes, e.g. prophylactics, should be used so that inordinate
amounts of measured load won't be due to membrane stiffness.
Opinions vary on how to obtain strength measurements in
certain circumstances. Some investigators would suggest
consolidating to the rn situ overburden stress; some wouldn't
consolidate the sample at all; and others would first consolidate
to stresses much higher than the rn situ values, then, knowing the
stress history, would back calculate what the undrained shear
strength could be. A combination of all three test types is
possible. More sophisticated tests, e.g. anisotropically
consolidated triaxial strength tests, can also be performed. Lee
(1985) presents a summary of current methodologies used for
performing triaxial testing on marine sediment.
When evaluating the strength characteristics of marine
sediments, a laboratory that has had previous experience in
determining and interpreting offshore strength characteristics
should be consulted.
Quality Assurance
Methods for determining accuracy and precision have not
been formulated for the CU triaxial test by ASTM. Lee and
Clausner (1979) state that even when using special techniques to
minimize disturbance, the best accuracy attainable is +/- 20
percent. Typically, accuracy is much worse.
Cost Analysis
Time required for each triaxial test (not including
drying and cooling times): 2 to 4 days
COST PER TRIAXIAL TEST $400
####################################################################
152
-------
References
American Society for Testing and Materials, 1987. Annual
book of standards, soil and rock, building stones and
geotextiles, ASTM, Philadelphia, PA, vol. 04:08, 1189 pp.
Bishop, A. W. and Henkel, D. J., 1962. The measurement
of soil properties in the triaxial test: Edward Arnold,
London, 227 pp.
Head, K.H., 1986. Manual of soil laboratory testing,
volume 3: effective stress tests, John Wiley, New York,
NY, pp. 743-1238.
Lee, H. J., 1985. State of the art: laboratory
determination of the strength of marine soils, in Chaney,
R. C., and Demars, K. R., eds., Strength testing of
marine sediments: laboratory and ijn situ measurements,
American Society for Testing and Materials Special
Technical Publication 883, Philadelphia, PA, pp. 181-250.
Lee, H. J. and Clausner, J. E., 1979. Seafloor soil
sampling and geotechnical parameter determination
handbook, US Naval Civil Engineering Laboratory Technical
Report R873, Port Hueneme, CA, 128 pp.
US Department of the Army, 1980. Laboratory soils
testing iin Engineer Manual EM 1110-2-1906, Washington,
DC, 388 pp.
Winters, W.J., 1987. Guidelines for handling, storing,
and preparing soft marine sediments for geotechnical
testing, US Geological Survey Open-File Report 87-278, 11
pp.
153
-------
APPENDIX A
Preparation of Randomly Oriented Mounts
For X-ray Diffraction
154
-------
APPENDIX A
Preparation Of Randomly Oriented Mounts
For X-ray Diffraction
Materials Required:
plastic tape
weighing paper
pestle and mortar
stiff brush
.063 mm sieve
spatula
pencil
sample holder
2 pieces of glass
slide cut to cover
the sample holder
opening
2. Tape one cover glass
over the opening of the
sample holder. Leave a tab
for later removal. Fold
the tape against itself
so that it will not stick
to other objects. Place the
glass slide down on a piece
of weighing paper.
Grind the sample throughly
so that all of it can be
brushed easily through a
230 mesh (;063 mm) sieve.
The particles must be much
finer than .063 mm to avoid
size fractionation of the
minerals. The sieve is used
only to achieve even
distribution.
»
-------
(optional) Place a
second holder over the
first as a mask. This
enables the buildup of
a thick enough layer
for later packing down,
while maintaining a
clean metal surface on
the first holder.
k
!/ I
Place the sieve over the
sample holder and brush
the sample from the mortar.
(Note: the sample may
become contaminated with
trace amounts of elements
from the sieve and utensils.
These will not be detectable
by X-ray diffraction but may
affect the results of
emmision spectroscopy. Do
not plan to reuse this cut
for sensitive chemical
analyses.
6. Use the spatula to loosen the
sample that has stuck to the
mortar and brush the material
into the sieve.
-------
7. Brush the sample through
the sieve into the cavity
of the sample holder. The
purpose of the brushing
is to obtain even
distribution and minimize
perferred orientation of
the particles.
8. Remove the sieve and
mask.
9. Place the sample holder
on a clean piece of
weighing paper, tap the
powder remaining on the
mask onto the first
weighing paper and
replace the mask over
the sample holder.
r
-------
10. Pour the excess powder
on the first weighing
paper into the sample
holder. Distribute the
powder evenly.
11. Use a glass slide to
pack the sample into
the holder firmly
enough so that it will
not fall out, deform,
or slide, but not so
firmly that perferred
orientation will be
produced on the
opposite surface (that
will later become the
top surface).
12. Add sample from the vial
as a filler if necessary
to creat a firm pack.
This part of the sample
need not be as firmly
ground, as it will become
the bottom surface and
will not be exposed to
the X-ray beam.
-------
13. Pack as in step 11 above,
then tape a glass cover
on the surface of the
holder or use a metal
backing clip (not shown).
The tape will not stick
if powder remains on the
metal surface. First hold
the cover glass in place
and wipe the metal surface
with your finger. Fold a
little tab of the tape as
before.
14. Turn the sample holder over
so that the bottom side
faces up. Carefully lift
the first tab of tape and
remove the cover slide. This
will be the surface exposed
to X-rays and should be
smooth, uniform, and flush
with the metal surface. If
not remake the mount.
15. Label the sample holder
with a pencil.
-------
APPENDIX B
Separation of Clay Fraction
160
-------
APPENDIX B
Separation Of Clay Fraction
1. Materials required:
50 ml centrifuge tubes
centrifuge tube caps
250 ml beakers
evaporating dishes
test tube rack
sodium metaphosphate
spatula
marking pencil
lab tissues
ultra "sonic probe
Vortex mixer
stop watch
thermometer
2. Label centrifuge tubes
with the marking pencil.
3. Label 250 ml beakers
with the marking pencil.
-------
4. Label evaporating dishes
with the marking pencil.
5. Pour enough sample to
fill the rounded part
of the bottom of the
centrifuge tubes.
6. Add an amount of sodium
metaphosphate to cover
the end of a narrow
spatula as shown.
-------
7. Add distilled water to
fill the tube slightly
more than half full.
8. Place cap on the
centrifuge tube.
9. Place tube on the
Vortex mixer, press
down and tip tube to
the side to achieve
vortex. Be carefull
that your fingers do
not smear the labeling
on the tube. Mix until
sediment in bottom of
tube is aa in suspen-
sion.
-------
lOa. Rinse and wipe off
tip of ultrasonic
probe.
b. Turn on ultrasonic
probe power supply.
c. Adjust for maximum
power.
11. Disperse sample for
15-20 seconds with
ultrasonic probe.
12. Add distilled water
to fill tube to with-
in 1 cm of the top of
the tube, place cap
on tube and mix the
contents using the
Vortex mixer.
-------
13. Place tubes in
centrifuge. Be sure
centrifuge is bal-
anced by having the
opposite tubes filled
equally. 2, 4, 6, or
8 tubes may be used.
Close cover.
14. Check water temperature
and find centrifuging
time in minutes and
seconds (Table B-l).
15. Turn power switch to
"not timed". Turn
speed controller first
to zero and then to 40.
-------
t
I6ec
17
18
19
20
21
22
23
2k
25
?6
07
c I
28
29
30
31
32
33
3k
35
ba — B--
P - PC
x 1000
6.723
6.551*
6.396
6.235
6.081*
5-938
5.800
5.666
5.529
5.1*08
"5.?R7
5.172
5.05^
U.9i*6
I*.8k3
1*.737
If. 61*0
1*.5M*
k U?o
If .366
c=12.22b
82.2
80.2
78.1
76.0
71*. 3
72.5
70.8
69.2
67.5
66.2
6k ,6
63.2
61.8
6o.k
•59.1
57.7
56.6
55.5
e;k.k
53.3
T=c+l*0
122.2
120.2
118.1
116.0
111*. 3
112.5
110.8
109.2
107.5
106.2
ink, 6
103.2
101.8
100.1*
QQ.l
97.7
96.6
95.5
Oli.li
93.3
T - td
92
90
88
86
8U
83
81
79
78
76
7*
73
72
70
6q
68
67
66
6k .
63
mln.
1
1
L
1
1
1
1
1
1
1
l
1
1
1
1
1
1
1
1
1
sec.
32
30
28
26
2lf
23
21
19
18
16
i«;
13
12
10
9
8
7
6
k
3
T +30
1:62
1:60
1:50
1:56
l:5k
1:53
1:51
1:1*9
1:1*8
1:1*6
1«k5
1:1*3
1:1*2
1:1*0
1:39
1:38
1:37
1:36
l:3k
1:33
Table B-l. Table constructed from the formula given below and in Hathaway
(1956) to show the relationship between the temperature of the sediment
suspension and the centrifuging time necessary to separate the silt and
clay fractions at 2 micrometers. T+30 is the total centrifuging time
-------
counting acceleration and deceleration. T-td is the time at which decelera-
tion is started. This table assumes a ratio of starting radius of the
particle from the axis of the centrifuge to the ending radius (bottom of the
centrifuge tube) of 1.787, a centrifuge speed after 30 seconds of accelera-
tion of 1400 rpm, a deceleration time of 30 seconds, and a particle density
of 2.65.
The formula may be represented by the equation
.
T = Yl loglO R, + 2 (ta+td)
3.81r*N*(p-p0) 3
where: T total time in seconds
ta time of acceleration
td time of deceleration
/? viscosity in poises
RI initial distance from the axis of rotation
Rj final distance from the axis of rotation
r radius of the particle in centimeters
N angular velocity in revolutions/second
p density of the particle in grams/cm3
pe density of the medium in grams/cm3
For ta and td the acceleration and deceleration are assumed to be
constant.
-------
16. When needle of the
tachometer starts to
move (tap it to make
sure it starts) start
stop watch and move
speed controller to
52.
17. When the needle
reaches 700 rpm turn
speed controller to
65.
18. When the needle
reaches 1400 rpm
(this should take
about 30 seconds
from the start of
the stop watch) move
speed controller back
to 52. Adjust speed
if necessary to keep
needle on 1400 rpm.
-------
19. When the stop watch
reads the time deter-
mined from the table,
turn the power switch
off and return speed
controller to zero.
20. Allow machine to coast
down to 1000 rpm. Then
push brake button
momentarily several
times to maintain a
constant rate of decel-
eration.
21. Speed should read near
zero by the time the stop
watch reads T+30 seconds.
\
-------
22. Remove tubes from the
centrifuge, pry off
caps, and pour super-
natant liquid into the
appropriate beakers.
Be careful that the
sediment in the bottom
of the tube is not
poured off. Repeat
steps 7 through 22
until the supernatant
liquid is clear (about
4 or 5 times) .
Silt Fraction Preparation
1. Use squeeze bottle with
distilled water to wash
residue in bottom of
centrifuge tube into
evaporating dish.
// 3''
Place evaporating dish
into circulating oven
at 60-70°C.
Remove when dry and
scrape silt into a
sample vial.
-------
APPENDIX C
Silver Filter Preparation
171
-------
APPENDIX C
Silver Filter Preparation
This method shows the use of one whole silver filter per sample. A
more economical practice would be to cut each filter in half, thereby
getting two samples X-rayed per filter. This is possible because only
an area 2 cm by 1 cm is actually subjected to the X-ray beam and a mask
may be constructed so that only this area is exposed during filtration.
This technique has the added advantage in that a smaller sample would
be concentrated into a smaller area and thus give better diffraction
max ima.
1. New silver filters should
be soaked in alcohol over-
night, then rinsed with
alcohol on the filtering
apparatus. Label the filter
around the edge with a
pencil; write lightly, the
filters are delicate.
2. Center filter on vacuum
filter apparatus.
3. Place filter funnel on top
of the filter and clamp
it down.
/
-------
Pour a small amount of the
sample suspension into the
filter funnel (enough to
cover the filter). Repeat
after the water is pulled
through the filter.
5. Remove clamp and filter
funnel and pry up an edge
of the filter with a
pointed spatula or razor
blade.
6. Lift filter with filter
forceps and place it in
a plastic capsule to dry
and store till X-ray
analysis.
-------
APPENDIX D
Ethylene Glycol Vapor Treatment
175
-------
APPENDIX D
Ethylene Glycol Vapor Treatment
1. Pour ethylene glycol
to about 1 cm depth
in base of desiccator.
2. Place silver filter
with oriented clay
aggregate on shelf
of desiccator.
Additional shelves
may be stacked if
necessary.
Place desiccator in
oven at 60°-70°C for
about four hours.
Longer times will not
hurt the samples. Do
not remove filters
until they are run on
the X-ray diffractometer.
-------
APPENDIX E
Heat Treatment
177
-------
APPENDIX E
Heat Treatment
1. Set temperature
control to 400°C.
2. Place filter in
oven using tongs.
Leave sample in
oven for 1/2 hour.
3. Remove filter by
pulling it forward
with a wire hook
until the front of
the filter can be
grasped by the tongs.
4. X-ray the sample and
repeat the procedure
at 550°C.
-------
APPENDIX F
Flow Charts For Identifying Clay Minerals
With Various 001 Diffraction Maxima
179
-------
DlMggregatloa, fractlonation
i __ _
X-rays of ori
1
Clay <2*- ] [ Silt 2-62M | 1 Sand >fa2* |
anted aggregates
X-raya of randomly oriemted powders
nh r'n rVi ri~i ri
Treated vith Heated to kXicC
ethylene glycol at least £ hr
at least J hr, li hre. by
Heated to 550*C Klectron micrograph
at least i kr if required
| vapor pressur
[^A>-
expands to 51-J-U
vlth rational se-
quence of higher orden
Rxpands but glTt»
irrational sequence
Peats sttgr be broad
»
W«y disappear or
^lv* i?*A *i»**;lii4[ ^
Increase In intensity
of 2«A spacing
Nsy give Irrational
sequence "*"
Msy give irrational
sequence — mmy shov
Increase In intensity
I-
i
«-| Increase In intensity I » j
060 near 1.5
OoO near 1.50A
060 near 1-5
OoO near l.JOA
060 near 1.50A
U60 near l.?2A |
060 near 1.5*A I
APPENDIX F
If required
Non-clay minerals determined by cosfArlson of patterns 6, 7, 3, 9
and 10 vltb standard patterns
Regularly Interetratlfled chlorlte-anotsnrllloDlte
or chlorlta-Termleullte "correnslte" l^/
a-ml™iy lutentratlfled culorite-sDOtaarillonlte lV/
Oaorite 5/
Dioctaoedral chlorite TJ
Teraleullte r[/
Dloctahedral rermicullte 12/
MctBOrlllonlte 16/
Trloctahedral MiDtsurlLlanlte
or rermlculit* l6/ rj/ l4/
Cheak vlth (lycarol treatment
•octmorlllonlte glres 17.7A
Termlcullte (Ires 1*-15A
Flow chart for the identification of clay minerals with 001 diffraction maxima at 29 or
14 angstroms on the air dried pattern. Chart shows the effects of ethylene glycol and
heat treatments on the location of these peaks.
-------
I Sample (-
Diaaggregatlon, fractlonatlon
[Clay
X-rays of oriented aggregates
[ Silt 2-621*1
20-621"
| Sand >62cl
X-raye of randomly orlamted powders
Air dried
Treated with Heated to fcOO°C
ethylene glycol at lea.it £ hr
at lea.it J hr, U brs. by
vapor pressure
leatad to 550'C Electron micrograph
at leeat i ar If
Collapses to ~ 10A H Collapses to
go change
-»J ffUght collapse I——1»4 Further collapae
06O near l.J
Destroyed — mmy
•urrlTc »ll^bt
lover tcflpen.tur«
If requlrad
loo-clay mluerala determined by comparison of patterns 6, 7, 8, 9
and 10 with standard patterns
Bagularly Interstratlfled Tarmlculita-mlca VJ/
nafularly Interstratlfled chiorlte-«ic« 1J_/
Randomly Interatratlflad rermiculite-mlca 6/ Ik/
j»-~t«— iy Intentratlflad chlorlte-miea 6/ I*/
Interatratlflad auntaortllonlta-aUea 1X/ J/ Ifl/
Interatratlflad trloctahedral •antmorUlonlte-miea
IV J/ !§/
*- Interatratlfied dloctahedral rermlcullte-mlea
U/ 12/ IV
- ^ Beplollte §/
-»- Attapulglte or palygoratite 9/
Flow chart for the identification of clay minerals with. 001 diffraction maxima at 10-14,
12-12.5, and 10.5 angstroms on the air dried pattern. Chart shows the effect of
ethylene glycol and heat treatments on the location of these peaks.
-------
[ Saayle \ —
Diaesfgregation, fractiooatian
i
Air
c
10,
~U
X-rays of oriented aggregates
[Clay <£v
1 1
dried Treated vita Heated to »00;C Beated to 550*C Klectron •IcrQfraj
1 ethylene Klyool at least } hr at least i ar If required
at least } hr, 1> hrs .by i
1 | vapor presaure | [ [
D ra -ttod m m CD
n— ^.n.^ — H*""-"! — HI.— •• —
Doayveetrical or lo change or
1 »- slight ahift *- alight sharpening
either vmj
1 • | Expand* to ~ llA )• — *-| Collapaes to ~ 7 A |-»-| Destroy**! ) »-] »itoul»r •arpboO.ac
^ Increa»e. in Bpacing I -_,,....„, ^ 7i 1 .J DMtroy
or T>P «h-nfi« 1 CQjJJ***i u 'A | *| "••tr°T
— 1»-
— *-
-H
—
| Silt 2-62CJ Sand XiS* |
X-raye of randomly orteated powders
I 6 | I 7 | 1 a | | * \ \ ™ \
06O ne«x 1.JOA
Oieek hkl'a
efalnst stej^ards
for pal^aarphs
060 betvevn
1-51A - 1.52*
5A peek Tary »ee*
060 near 1-5*A |—
060 rarlable I—
y LJ ObO near 1-^9 A 1—
K* I >-| Flbrooa .wrpbolocy [*•
—
rt[ J Tubular •arffcaloor U
1 - U- ! ^ •- -S TW t-awinr
* 1 ** '""•ng" * *^ "d* r ue •/
»4 »—J Plmty •orpholacy
1 1 * * ^^
i« • -mt*—M.i nshal
P He Ba«sjitl •• 1 afc>Jf|Ui(ll
»-) Platy awrphelotj
-
Check akl'a
afain.it •tandJLr4
Check hfcl'a
a»p.J_a«t •taadard
OoO oear 1-^9 A
hi baodj preaant
ObO oee>r 1.49A
hk baikU pra«eiit
•odarataly broad
ocr|«4 hHl ' • reaolrad
-
L
oto.oMr i.5»A|-
0*0 new 1.50AJ—
if required
Han-clay minerals determined by cojqpariaon of patterns 6, 7, 8, 9
and 10 vlth standard pattern!
— •- Htca, muscorlUc, rarlous polyBOrphs ll/ 2O/ Ik/
— »- Olauconite or celadonlte 3J./ lV
(Boecoellte If Y beartnj) l^/
»~ Xlotite, Tmrious polyaorpas ll/ 19/
^— Interstratifled Mlra scotsrtrlllnnitf , or •ica-Termlcullte
"Ullta" ijl/
>- Otryaotllt \J I/
>- Antl«arlte k/ i/
Disordered kaallnlte ("fireclay") %/
001 'i
<~ Kaallnlte
_. Check against W
-«*=*"• standard patterns ^
^ OHortte P1"* lD<:rM«l ^
viBJ-°rl** Intensity of 1*A »pacln«
_. Dloctahedral cheek IncreaM In
ealorlte Intensity of 14A spacinc
Flow chart for the identification of clay minerals with- 001 diffraction maxima at Id or
7 angstroms on the air dried pattern. Chart shows the effect of ethylene glycol and
heat treatments on the location of these peaks.
-------
APPENDIX G
Preparation of Glass Slides for Optical Microscopy
183
-------
APPENDIX G
Preparation Of Glass Slides For Optical Microscopy
1. Materials required:
mounting medium
plastic dropper bottle
felt marker pen with
water soluble ink
spatula
distilled water in a
dropper bottle
hot plate
glass cover slips
glass microscope slides
glass scribe
Pour a small amount of the
mounting medium into the
plastic dropper bottle.
Any real or artificial
(i.e. Piccolite or Caedex)
Canadian balsam may be
used.
2. Mark the slide temporarily
with the felt marking pen.
A pen with water soluble ink
is used because the mark is
not soluble in xylene which
may be necessary later to
clean the excess mounting
medium from the slide. Set
the hot plate at about 300°
F. The exact setting must be
determined by tests; the
cured mounting medium must
no longer be tacky, but not
be yellow or brittle.
3. Place a small amount of
sample in the center of the
slide.
-------
4. Add a drop of distilled
water.
t«
*
m r *-^\
x T i\ f
5. Mix and spread the sample
with the spatula tip.
a
6. Place the slide on the hot
plate to dry.
-------
7. Place one or two drops of
mounting medium on the
center of the sample.
8. Place a square glass
cover slip over the sample
after the mounting medium
has been allowed to cook'
for about one minute.
9. Press the cover slip
down with a pair of
forceps. The mounting
medium should spread
and fill the space
beneath the cover glass.
-------
10. If the mounting
medium does not fill
all of the space,
add a drop right at
the edge of the cover
slip. Capillary
action should draw the
mounting medium under
the cover slip.
11. Allow the slide to cook
for another minute or
two, then remove it from
the hot plate and allow it
to cool. If the slide has
too much mounting medium,
it can be cleaned with a
razor and washed with
xylene.
12. The slide should then be
permanently marked with a
glass scriber. This is not
done earlier because glass
chips would often end up in
the sample and may be con-
fused with volcanic glass
shards.
------- |