Techniques of Water-Resources Investigations

of the United States Geological Survey

ChapterCl

LABORATORY THEORY
AND METHODS
FOr
SEDIMENTANALYSIS
By Harold P. Guy

Book 5
LABORATORY ANALYSIS

Click here to return to USGS Publications

UNITED STATES DEPARTMENT OF THE INTERIOR

CECIL D. ANDRUS, Secretary

GEOLOGICAL SURVEY

V. E. McKelvey, Director

First printing 1969

Second printing 1973

Third printing 1977

UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1969

For sale by the Branch of Distribution, U.S. Geological Survey,

1200 South Eads Street, Arlington, VA 22202
PREFACE
This series of manuals on techniques describes procedures for planning
and executing specialized work in water-resources investigations. The
material is grouped under major subject headings called books and further
subdivided into sections and chapters; Section C of Book 5 is on sediment
analysis.
The unit of publication, the chapter, is limited to a narrow-field of
subject matter. This format permits flexibility in revision and publication
as the need arises.
Provisional drafts of chapters are distributed to field offices of the
U.S. Geological Survey for their use. These drafts are subject to revision
because of experience in use or because of adva,ncement in knowledge, tech-
niques, or equipment. After the technique described in a chapter is suffi-
ciently developed, the chapter is published and is sold by the U.S. Geo-
logical Survey, 1200 South Eads Street, Arlington, VA 22202 (authorized
agent of Superintendent of Documents, Government Printing Office).
III
CONTENTS

Pege Page
. ..
Preface_----------------------------------- 111 Procedure for the VA tube-pipet method of
Abstract----------------------------------- 1 particle-size analysis- _ _ _ _ ___ ___ __ _ __ ___-_ _ 30
Introduction- _ _ __ _ _-_ _______ ___ -_ ___ __ __ __ _ I Equipment___-____-__----__---_________ 31
Acknowledgments _____ __ - _ __ - _ - _ __ - _ _ _ _ 1 Preparation of sample __________________ _ 32
Functions of a sediment laboratory- _ _ _ _ __ 1 Separation of sand from fines _________ 32
Units of measurement and definitions- _ _ __ 3 Splitting and dispersing _______ __ __ _ __ 33
Theory of particles falling in a fluid- _ _ _ __ _ 8 The VA tube analysis- _ _________________ 33
The settling medium ________________ 8 Tube selection _____ __ ___ _ __ __ _ __ __ _ _ 37
Stokes law for small particles _________ 9 Procedure_-________-_______________ 37
Drag-Reynolds number--- _ ___ __ __ _ __ 10 Chart interpretation- _ _ _ _ _ __ __ __ _ _- _ 38
Significant figures for reporting results--- __ 10 Procedure for the BW tube-VA tube method of
particle-size analysis- _ _ _ ___ __ __ __ __ ____ _ __ 38
Procedure for determining suspended-sediment
11 TheOden theory------ _________ -___-___ 38
concentration- - _ - ________ _______________ _
11 Equipment-- __ ___ __ __ __ __ __ __ __ __ __ __ __ 39
Evaporation method- _ __ ________________
12 Preparation of sample- _ _ _ __ __ __ _-_ __ ___ _ 40
Filtration method- _ - __ _ __________-_____
The BW tube analysis--- _ __ __ __ __ __ _ ___ _ 42
Laboratory forms- ______ ________________ 13
17 Dispersion-_______--_______________ 42
Procedure for concentration analysis- _ _ _ __
Withdrawals- __ _ __ _ _ _ ___ ___ __ _-_ ___ 42
Determination of particle-size distribution of Recording of data ______ __ __ _ __ __ _-_ _ 43
suspended sediments- _ _ ___ _ _ ____ __-_ __-_ _ _ 18 45
The Oden curve ____________________
Recommended frequency of analyses- _ _ __ _ 18 47
Limitations of BW tube for sand--- __ __ __ _
Method of analysis---- _ __ __ __ _ __ ___ __ _ __ 19
Determination of particle-size distribution of
Limitations- _ - _ _ __ __ _ __ ___ __ _ __ _ - _ _ 19 47
deposited sediment and soil samples_- - _ _ __ - _
Dispersed settling medium- _ __ _ __ __ - _ 19 Equipment and method of handling- _ __ __ _ 47
Native water settling medium ________ 20 48
In situ measurement- _ _ __ _ ___ _ __ __ __ _ __ _
Sample preparation for native water Nominal diameter by immersion- _ __ __ 48
analysis _______- __ __ __ _ __ __ __ __ __ __ _- _ 22 The Zeiss analyzer- _ __ __ _-_ _ - __ __ __ _ 48
Procedure for the sieve-pipet method of particle- Laboratory analysis _____________________ 49
sizeanalysis ____- ____ -__-___--__- ____ - ____ 23 Preparation and procedures- _ __ __ _ __ _ 49
Equipment ______ - ___ __ ._ _ __ _ ____ __ __ - __ 24 Calculation of results _________ ___ _ _ - _ 51
Preparation of sample-------------- _____ 26 Mechanical analyses of soil samples _ __ __ __ 51
Recording and decanting- _ _ _ ________ 26 Other determinations related to sediment anal-
Mechanical dispersion- - __- - __ __ __ __ _ 26 ysis-________-__________________________~ 51
Sieving ____ -__-___--__-.__--__-________ 28 Organic material ______ _ __ __ __ __ __ _ __ __ __ 51
Separation of sand from fines-__- - - - _ _ 28 Relationship to particle-size settling
Dry method-- _____-.._--_-- _._______ 28 media___--__-_____-_____________ 52
Wet method__---________-____---__ 28 Procedure of removal-.._ _ __ _ _- _ _ __ __ _ 52
Analysis of silt-clay fraction--- __--__---- 29 Aggregate destruction- _ ___ _ __ __ _ __ __ 52
Splitting ____--__--_..- ______________ 29 Dissolved solids- - _ _-_ _ ___ _ __ _ __ __ __ __ _ _ 53
Dispersing-------- _______ i _______ -- 29 Related water-quality analysis- _ _ - __ __ _ __ 53
Pipetting__-__--__-._- ____ -___- ____ 29 Specific gravity_-____--__-______________ 54
Calculation of results _____ _- __ _- __ __ __ __ _ 29 Specific weight.. _ - __ __ _ __ _ __ _ __ __ __ __ _ _ _ 54
Total weight of sediment in sample-. _ 29 Selected references- - __ __ __ ___ _ __ __ __ __ _ __ _ _ 55
Calculations ____- ___ __ __ __ ___ __ __ __ _ 30 Index_--___--___________________________-- 56
VI CONTENTS

ILLUSTRATIONS

P%Kl Page
FIoURx.9 l-3. Graphs showing: FIGWILES 8. Laboratory form for Particle-
1. Relation of nominal di- Size Analysis, Sieve-Pipet
ameter and fall veloc- Method- _ _ _ _ _ __ _ __ _ _ __ __ _ 27
ity for quartz par- 9. Photograph of VA tube and re-
ticles- _ _ __ _ __ _ _- - - - 6 cording apparatus- _ _ _ _ ___ _ 32
2. Relation between fall 16-12. Laboratory form for-
velocity and diameter 10. Particle-Size Analysis,
for quartz spheres- _- 10 VA Tube Method
3. Weight of dissolved sol- (12~cm tube)------ 34
ids for given size of 11. Use with the 186-cm
aliquot and given con- VA tube--- ________ 35
centration- - - _- _ __ _ _ 12 12. Recording steps in com-
puting particle size
4-6. Laboratory form for- when VA tube is used
4. Sediment Concentra- with sieve, pipet, and
tion Notes, Depth (or) other methods- 36
Integrated Samples 13. Sketch of the Oden curve----- 39
(Short form) _ _ _ __ __ 14 14. Photograph and diagram of
5. Sediment Concentra- bottom-withdrawal tube and
tion Notes, Depth stand- _ _ _ _ _ __ __ _ _ __ __ _ __ _ 40
Integrated Samples 15-17. Laboratory forms for-
(Comprehensive 15. Particle-Size Analysis,
form) _ - _ _ _ _ _____-- 15 BW Tube Method-- 41
16. Particle-Size Analysis,
6. Sediment Concentration
BW Tube Method 0
Notes, Point Inte- (experimental) _ _ _ _ _ 44
grated Samples..---- 16 17. Plotting Oden curve
7. Photograph and diagram of (experimental) _ - _ _ _ 46
apparatus for pipet particle- 18. Diagrammatic sketch of par-
size analysis--- __ __ __ _ __ _ __ 25 ticle analyzer--- _ _ __ _ _- _ __ _ 49

TABLES

TABLE 1. Factors C for computation of sediment concentration in milligrams per

liter when used with parts per million or the ratio (times 1CP) of the
weight of sediment to the weight of the water-sediment mixture----- __ 4
2. Recommended scale of sizes by classes for sediment analysis- _ - _ _ _ _ _ __ _ _ 7
3. The subdivision of matter and the resulting properties----------------- 8
4. Fall velocity correction factors for use with figure 2 when temperatures
are different from 2O“C ________________ ------_- __________________- 10
5. GuidetoselectionofVAtubesize ______________ ----_--- _____________ 19
6. Time of pipet withdrawal for given temperature, depth of withdrawal,
and diameter of particles- _ __ __ _ __ _ __ __ _ _ __ _ _ - - __ - _ - _ _ _ _ __ _ __ _ _ _ __ 24
7. Bottom withdrawal tube sedimentation time table to be used with the
Odencurve----------------------------------------------------- 43
LABORATORY THEORY AND METHODS FOR SEDIMENT ANALYSIS

By Harold P. Guy

Abstract cians for sediment analysis. These procedures

are not individually acknowledged because they
The diverse character of fluvial sediments makes
have usually evolved as a result of the many
the choice of laboratory analysis somewhat arbitrary
and the pressing of sediment samples difficult. This sediment conditions encountered in the many
report presents some theories and methods used by kinds of streams in the United States and be-
the Water Resources Division for analysis of fluvial cause the scientists and technicians who have
sediments to determine the concentration of suspended- developed these highly efficient systems did so
sediment samples and the particle-size distribution of
in close association with fellow workers. It is
both suspended-sediment and bed-material samples.
Other analyses related to these determinations may regrettable that the writer cannot document all
include particle shape, mineral content, and specific these procedures in great detail ,&cause it is
gravity, the organic matter and dissolved solids of difficult to obtain complete information and be-
samples, and the specific weight of soils. cause there is need for standardization. Sedi-
The merits and techniques of both the evaporation mentation theories, however, are better ac-
and filtration methods for concentration analysis are
discussed. Methods used for particle-size analysis of knowledged ; but again, many contributions have
suspended-sediment samples may include the sieve evolved from individuals or groups and are not
pipet, the VA tubepipet, or the BW tube-VA tube de- fully documented.
pending on the equipment available, the concentration Among the many who have contributed to
and approximate size of sediment in the sample, and this chapter, as well as the earlier 1960 draft,
the settling medium used. The choice of method for
most bed-material samples is usually limited to pro-
special appreciation is extended to C. D. Allbert,
cedures suitabie for sand or to some type of visual F. 1C. Ames, B. C. Colby, B. R. Colby, C. R.
analysis for large sizes. Collier, R. K. Flint, J. C. Mundorff, G. Porter-
Several tested forms are presented to help insure a field, and R. B. Vice.
well-ordered system in the laboratory to handle the
samples, to help determine the kind of analysis re-
quired for each, to conduct the required processes, and Functions of a sediment laboratory
to assist in the required computations. Use of the A sediment laboratory in the Water Resources
manual should further “standardize” methods of fludal
Division of the U.S. Geological Survey has two
sediment analysis among the many laboratories and
thereby help to achieve uniformity and precision of principal functions: (1) the determination of
the data. suspended-sediment concentration of samples
collected from streams and (2) the determina-
Introduction tion of particle-size distribution of suspended
sediment, streambed material, and reservoir de-
Acknowledgments posits. Other kinds of analyses, usually associ-
The writer has attempted to document the ated with these two functions, may include quan-
“special” procedures developed and routinely titative determination and (or) removal of
used by many scientists and laboratory techni- organic matter and dissolved solids, the specific

1
2 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

gravity of sediment particles, the specific weight several challenges in addition to those encoun-
of soils and sediment deposits, particle shape tered in determining concentration. As discussed
determinations, and mineral analyses. later, the sieve-pipet method or the visual-
Data from the concentration and particle- accumulation tube-p@ method is recom-
size determinations are used to make computa- mended in order to obtain maximum informa-
tions of suspended-sediment discharge, com- tion regarding fall velocity. Since each kind of
putations of total sediment discharge, and analysis requires a rather narrow range of sedi-
computations of the probable volume and spe- ment quantity, the much larger range normally
cific weight of exposed and submerged reservoir found in sediment samples is likely to be a
deposits. These are but a few of the many uses problem. Compositing or grouping of the sedi-
made of concentration and particle-size data ment from several samples into one may be
of fluvial sediment. The reliability of these necessary when the stream concentration is low ;
computations and the utility of the data are or, a sample from a stream having a high con-
dependent on the accmacy and reliability of centration may need to be split so that an opti-
the laboratory analyses. This discussion of the mum smaller quantity of sediment may be
functions of a sediment laboratory is given to analyzed. For example, the quantity of sand to
provide insight into some of the problems en- be analyzed in the visual-accumulation tube can
countered in sediment analyses and to help guide range from 0.05 to 0.8 g (gram) for the smallest
the field technician in the collection of samples tube to about 5 to 15 g for the largest tube; or
that are suitable for the desired analysis. in the case of the pipet, the quantity of silt and
In the determination of suspended-sediment clay must be just sufficient to yield a concentra-
concentration, the quantity and characteristics tion between 2,000 and 5,000 ppm (parts per
of the sediment, and the chemical quality of million) in the sedimentation tube. Each kind
water from the stream (native water) must be of analysis has its own range of sediment
considered in the processing of the sample. Too quantity necessary to obtain optimum results.
small a quantity of sediment tends to magnify In addition to the problem of determining
errors inherent in determining the weight of the kind of analysis to use, it is necessary to
dry sediment, or in the transfer of material keep in mind the need to determine the concen-
from one container to another. On the other tration of sediment in the samples for which
hand, too large a quantity of sediment may cause particle-size distribution is determined. Provi-
problems with respect to splitting, drying, and sion should, therefore, be made to obtain the
weighing the sample. Samples collected with weights of all increments of sediment in the
certain types of mineralized water and (or) analysis as well as portions not used in the
containing colloidal clay result in difficult sep- analysis, The sample must be kept wet until
aration of sediment from the native water. A the particle-size analysis is complete to prevent
weight correction based on the quantity of the formation of aggregates that resist redis-
dissolved solids is often necessary when the persion. Division of a combined sample into
evaporation method is used to determine sue- separate portions for separate concentration and
pended-sediment concentration. These and other particle-size analysis should be avoided unless
aspects of suspended-sediment concentration the sample can be divided into two samples each
determinations are discussed later in this representing the streamflow at the given time.
chapter. Splitting may introduce quantitative and quali-
The determination of the sediment particle- tative errors that will affect the concentration
size distribution in a sample of streamflow offers and (or) particle-size results. Procedures for
LABORATORY THEORY AND METHODS FOR SEDIMENT ANALYSIS 3
determining particle-size distribution of fluvial ing on the geologic and hydrologic character-
sediment ,by sieve and sedimentation methods istics of any given stream and drainage basin.
are discussed in sections to follow. The fre- It may be useful to check the specific gravity
quency of sampling for analysis and the need and shape characteristics of sediment for a
for special analytical procedures for both native specific stream in order to convert the fall
and dispersed settling media are also discussed. diameter of sediment determined by the VA
The determination of particle-size distribu- tube to standard fall velocity or actual particle
tion of bed-material samples is different from size.
the determination of particle-size distribution
of suspended-sediment samples because the Units of measurement and definitions
quantity of material is usually large, the The methods used for collecting sediment
samples are usually sand with a considerably samples are described in a chapter of this series
larger median size and most samples will be in entitled “Field Methods for Fluvial Sedi-
a room-dry condition. TJnlike suspended-sedi- ment Measurement.” Though it is possible to
ment samples, which often have very limited determine many kinds of sediment concentra-
quantities of material available for analysis, the tion such as the temporal mean or the average
larger bed-material samples with proper split- along a line, over an area, and throughout a
ting allow the use of optimum quantities of specified volume, the sampling and compositing
material for all size ranges. procedures are usually designed to give a time-
Sieves are used to determine particle-size dis- integrated, velocity-weighted, cross-sectional
tribution of material z2.0 mm (millimeter), average value of concentration or size
whereas the VA tube (visual-accumulation distribution.
tube) is used most frequently to determine The common unit for expressing suspended-
particle-size distribution of material from 0.062 sediment concentration is milligrams per liter
to 2.0 mm. However, the sieve can be, and some- (mg/l) and is computed as one million times the
times is, used for the entire range of sand and ratio of the dry weight of sediment in grams
gravel sizes. Particle size determined by the VA to the volume of water-sediment mixture in
tube is referred to as sedimentation diameter cubic centimeters. Other units, such as percent
and the relationship-between sieve size and sedi- or parts per million ( ppm), have been used to
mentation diameter is well known for quartz express suspended-sediment concentration but
spheres. It is usually assumed that the particles are not recommended. In the laboratory it is
have a specific gravity of 2.65 unless the material more convenient to obtain the weight of the
in the sample appears to contain significant water-sediment mixture than to obtain its
quantities of minerals other than quartz. Both volume. Therefore, the following formula in-
specific gravity and particle shape vary depend- volving parts per million is used

weight of sedimentx l,OOO,OOO

mgP=C (PPm)=c[ weight of water-sediment mixture 1
where C is given in table 1. Total sediment con- generally not reliable because variation due to
centration has, in some cases, been expressed on a particle characteristics and compaction makes
volume basis, especially for sands and other the assumption of a specific weight for the con-
coarse material. Such volume measurements are version of volume into weight difficult.
4 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

Table 1 .-Factors C for computation of sediment concentration in milligmms per liter when used with parts per million or the
mtio (times 10s) of the weight of sediment to the weight of the water-sediment mixture
[The factors are based on the assumption that the density of water is 1.000, plus or minus 0.005, the range of temperature is O=2Q°C, the specillc
gravity of sediment is 2.65, and the dissolved solids concentration is less than 10.000 parts per million]

Ratio c Ratio c Ratlo C

o- 15,900 1. 00 234,006-256,000 1. 18 417,000-434,000 1. 36

16,000- 47, 000 1. 02 257,000-279, 000 1. 20 435,000-451,000 1. 38
47,000- 76,000 1. 04 280,000-300,000 1. 22 452.006-467.000 1. 40
77,006-105,000 1. 06 301,00&321,000 1. 24 468; 000-483; 000 1. 42
106,006-132,000 1. 08 322,006-341,000 1. 26 484,000-498,000 1. 44
133,006-159,000 1. 10 342,006-361,000 1. 28 499,006-513,000 1. 46
160,000-184,000 1. 12 362,000-380,000 1. 30 514,000-528,000 1. 48
185,000-209,000 1. 14 381,000-398,000 1. 32 529,000-542,000 1. 50
210,000-233,000 1. 16 399,000-416,000 1. 34

Several measures of particle size should be specific grayity of 2.65 and has the same standard fall
defined as the basis for simple, accurate, and velocity as the particle.
The sedimentation diameter of a particle is the
practical determination and expression of sedi-
diameter of a sphere that has the same specific gravity
ment sizes. Some measures of particle size are and terminal uniform settling velocity as the given
difficult to evaluate and are seldom determined ; particle in the same sedimentation fluid.
others are determined only approximately. The standard sedimentation diameter of a particle is
Before the basic concepts of these measurements the diameter of a sphere that has the same specific
gravity and has the same standard fall velocity as the
are discussed, several definitions relating to par- given particle.
ticle size l are quoted essentially as reported in 8&e distribution, or simply distribution, when
(U.S. Inter-Agency Committee on Water Re- applied in relation to any of the size concepts, refers
sources, Subcommittee on Sedimentation, 1957b, to distribution of material by percentages or propor-
p. 11) : tions by weight.
Fall velocity and settling velocity are generally terms
The nom&al diameter of a particle is the diameter which may apply to any rate of fall or settling as dis-
of a sphere that has the same volume as the particle. tinguished from standard fall velocity.
The sieve diameter of a particle is the diameter of
a sphere equal to the length of the side of a square sieve The basic concept of “size” of a sediment
opening through which the given particle will just pass. particle is expressed in terms of volume (the
The standard fall velocity of a particle is the average nominal diameter as defined above). When con-
rate of fall that the particle would attain if falling sideration of the density or mass of the particle
alone in quiescent, distilled water of infinite extent
and at a temperature of 24°C.
is necessary, specific gravity is used. It is recog-
The standard fall diameter, or simply fall diameter, nized that sieves, frequently used for the separa-
of a particle is the diameter of a sphere that has a tion of sediments into “size” grades, do not grade
particles entirely by size but by shape as well.
IThe meaning of the term “particle” is often rather am- Ideally, sieving should show relative quantities
biguous. Consider any given sample of soil which contains a
considerable number of relatively large aggregates; these of the sample in terms of square openings; but
aggregates are “particles,” and many properties of soil de- in reality, many irregular particles that have
pend on their size and stability. The aggregates in turn
consist of grains which are also “particles” and which are nominal diameters much greater than the stated
oommonly referred to as the “ultimate particles” of soil. The sieve openings will pass through a given sieve.
size of the ultimate particles also determines many soil
properties. These so-called ultimate particles may also be The above definition of sieve diameter means
considered as not truly ultimate because the grains may be that the sieve diameter of the particle is the
composed of two or more crystals of the same or different
minerals. length of the side of the smallest square open-
Obviously, it is important to be aware of the kin& of ing through which the particle could pass. Thus,
particles to be measured. In sediment investigations, the
ultimate particle size should be used unless otherwise specified. it is assumed that the nominal and sieve diam-
Ultimate particle size in fluvial sediment can be de5ned for eters of a sphere are always equal. It should be
practical purposes as the particle size obtained by standard
methods of preparation and dispersion of sediment samples. mentioned further that the nominal diameter
Stnndard methods of preparation and analysis in native water of a sieved fraction may be larger than the in-
usually results in floocules of partioles, especially in the clay
sizes. dicated sieve diameter if the sieve is badly worn
LABORATORY THEORY AND METHODS FOR SEDIMENT ANALYSIS 5
or if some openings have been stretched out of The physical size of particles, especially for
their “square” shape. gravel and larger particles, may be expressed
One of the dynamic properties necessary for on the basis of the length of three diameters
sediment transport studies is the velocity of fall of the particle (a, b, and c) which are mutually
of the individual particles in quiet water. Terms, perpendicular (Inter-Agency Report 12, p. 12).
such as “settling velocity,” “fall velocity,” “ve- The axes are differentiated as a=the longest
locity of settling,” or “velocity of fall,” have axis of the particle, b= the intermediate axis of
been applied to this fundamental characteristic. the particle, and c=the shortest axis of the
Standard fall velocity allows various investiga- particle. The harmonic mean diameter (Krum-
tors to compare their data directly. This fall- bein and Pettijohn, 1938, p. 128) is defined as
velocity concept is simple and logical ; however, dn=3 abc/(ab+bc+ac) and is considered to be
a size concept is so thoroughly embedded in the closely related to diameters calculated on the
concepts relative to the field of sedimentation basis of Stokes law. In addition to the harmonic
that an expression or measure of diameter is mean, the arithmetic mean d,= (a+b+c)/3 and
needed. Hence, fall diameter is related to fall the geometric mean diameters d,=3J& have
velocity in that a given particle has only one fall been used by various investigators to express
diameter which is independent of the type of size. The important value of these physical
material, the concentration of analysis, or the measurements is in the expression of a particle
method of analysis. shape factor.
Studies have shown that a particle settling as The sha e factor sf =c/@ is useful in relating
nominal dP iameters to fall velocity or fall di-
one of several surrounding particles generally ameter. Figure 1 shows the relation of naturally
falls with a velocity different from that for the worn quartz particles to their fall velocity for
particle settling alone. The fall-diameter distri- shape factors of 0.5, 0.7, and 0.9 and at tem-
bution of a sample may be described as the dis- peratures of O“, lo:, 20°, 30°, and 40’ C. The
tribution determined as though each particle shape factor of 0.7 1s about average for natural
sediments.
were dropped separately and a summation made The usefulness of shape factors is limited in
of the fall diameters of the particles. This dis-
routine sedimentation studies, for even if the
tribution is based on weight and fall diameter
sieve, nominal, and axis length diameters are
and is believed to be the most basic and desirable
known, the fall diameter may be determined
expression for the sedimentation analysis of a more easily and accurately through fall velocity
sample.
data than by the shape factor. In summary, the
The definition of sedimentation diameter does factor of shape may be kept in proper perspec-
not restrict the conditions under which the tive by remembering that for smooth quartz
settling velocity may be determined. The spheres all four diameters, nominal, sieve, axis
relation of diameter to settling velocity depends length, and fall are the same ; that the relative
only on the specific gravity of the particle. Any effect of shape varies with size; and that the
of the various settling velocities, depending on effect of shape is more significant with sieve
the nature of the sample, the concentration, the than with the sedimentation type of analyses,
fluid, and the apparatus, must be used with a at least for the smaller sizes.
determination or estimate of specific gravity to With respect to met’hods of particle-size anal-
establish sedimentation diameter. It is apparent yses, the following distinction is made between
from the definition previously given that the a dispersed system and a stratified system as re-
standard sedimentation diameter of a particle ported in (U.S. Inter-Agency Committee on
depends only on the volume and shape of the Water Resources, Subcommittee on Sedimenta-
particle, and that the relationship of nominal tion, 1957b, p. 13) :
diameter to standard sedimentation diameter
A diapereed system is one in which part&l- begin
becomes a measure of the effect of shape and to settle from an initial uniform dispersion and in
roughness on the settling velocity of the particle which particles of sedimentation sizes settle together.
in water at 24’C. Size distribution may be determined by measuring the
6 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

d
I I I I
--=I

, -
, -

) -
, -

v -

) -

4 -

Skl313Wllllv\I NI ‘kl3131”\ltfl~ WNIWON

LABORATORY THEORY AND METHODS FOR SEDIMENT ANALYSIS

concentration of sediment at given intervals of depth America. This scale has a fixed geometric inter-
and settling time, as in the pipet method, or the distri- val with each grade one-half as large as the pre-
bution may be obtained from the quantity of sediment
ceding, as suggested by Udden in 1898. Grade
remaining in the suspension after various settling
times, as in the bottom-withdrawal-tube method. scales are, without exception, based on unequal
A atratified system is one in which the particles start class intervals as a result of the large range of
falling from a common source and become stratified sizes for most sediment. If equal intervals were
according to settling velocities, as in the visual-accumu- chosen to give full significance to the smaller
lation-tube method. At any given instant, the particles
sizes, then an excessive number of classes would
coming to rest at the bottom of the tube are of one
sedimentation size only and are finer than the particles be necessary with the result of virtually no sig-
that have previously settled out and are coarser than nificance between classes for the larger sizes.
those remaining in suspension * * *. The unequal class intervals may make the ap-
The size-gradation scale, or the distribution plication of certain statistical analyses to the
of the size classes, is a descriptive function data more difficult. Table 2 gives the number
which serves to place nomenclature and termi- for Tyler and U.S. Standard sieves and the size
nology on a uniform basis. A grade scale has limits for each class name of the recommended
been defined by Krumbein and I’ettijohn (1938, gradation scale.
The phi scale shown in column 6, derived by
P- 76) “* * * as an arbit,rary division of a con-
tinuous scale of sizes, such that each scale unit Krumbein (1934)) is a logarithmic transforma-
or grade may serve as a convenient class in- tion that has integers for the class limits that
terval for conducting the analysis * * *.” In increase with decreasing grain size. This. scale
was developed to make it easier to directly ap-
such a scale of sizes, it is recognized that a
ply conventional statistical practices to sedi-
natural sediment has a continuous size distribu- ment-size data.
tion without implication of natural grouping. Specific gravity is the ratio of the weight of
The Wentworth grade scale is favored in a substance to the weight of an equal volume
0
Table I.-Recommended scale of sizes by classes for sediment analysis
[After Lane and others (1@47)]

U.S. Metric units Phi

zl% standsrd Class n8me value ’ English units (feet)
NO. sieve No.1 (millimeters) (micrometers) (4)
(0 (2) (3) (4) (5) (6) (7)

Boulders---- _-_ __ __ _ __ >256 ._____ >O. 840

Large cobbles __________ 256 -128 -8 0.840 -0.420
Small cobbles--- - _ __ __ 128 - 64 -7 0.420 -0. 210
Very coarse gravel- - _ _ _ 64 - 32 -6 0. 210 -0. 105
Coarse gravel __________ 32 - 16 -5 0.105 -0. 0525
Medium gravel-------- 16 - 8.0 __--__---------_ -4 0. 0525 -0. 0262
Fine gravel ______-_-___ 8. 0 - 4.0 __--_--___-----_ -3 0. 0262 -0. 0131
Very fine gravel-------- 4. 0 - 2.0 -2 0. 0131 -0. 00656
Very coarse sand-- ____ 2. 0 - 1.0 2000 -1000 -1 0. 00656 -0. 00328
Coarse sand _______ --__ 1000 - 500 0 0. 00328 -0. 00164
Medium sand __________ ; 80 I ; ;; 500 - 250 +l 0. 00164 -0. 000820
Fine sand---_---_----.. 0: 25 - 0: 125 250 - 125 +2 0. 000820 -0. 000410
Very fine sand _________ 0. 125 - 0. 062 125 - 62 +3 0. 000410 -0. 000205
Coarsesilt___-__-___-- 0. 062 - 0. 031 62 - 31 +4 0. 000205 -0. 000103
Mediumsilt___-___--_- 0. 031 - 0. 016 31 - 16 +5 0. 000103 -0. 0000512
Fine silt- _________ -_-- 0. 016 - 0. 008 16 - 8 +6 0. 0000512-O. 0000256
Very fine silt.. - _ _ ____ - _ 0. 008 - 0. 004 8- 4 +7 0. 0000256-O. 0000128
Coarse clay- _ _ _ _ _ _ __-_ 0. 004 - 0. 0020 2 +8 _____ ________---_-__
Medium clay ________ -- 0. 0020- 0. 0010 ; I +9 --------____________
Fine clay- _ _ _ _ ____ _ _- _ 0. OOlO- 0. 0005 l- A5 +10 --------____________
Very fine clay--------- 0. 0005- 0. 00024 0. 5- 6. 24 +11 --------___________-

1 For marimum size of the iven class.

* Sieve openings are marke di in inches and millimeters.

0
8 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

of water at 4°C. The measurement is then action, size, and concentration of particles in
dimensionless. It should not be confused with fluids. The milling, ceramic, cement, abrasive,
specific weight which is defined as the weight and paint industries usually accomplish the
per unit volume ; for example, water has a spe- separation into size classes in a rising column
cific weight of 62.4 pounds per cubic foot and of air. In the fields of geology, soils, and river
a specific gravity of 1.0. hydraulics, quiet water is the most common
settling medium. Whatever the medium, the
Theory of particles Falling in a Fluid principle involved measures the tendency for
particles of similar specific gravity to fall at a
Table 3 illustrates the relative position of the rate that is related to the size and shape of
finer particles commonly presem in sediment the particles. Several laws or mathematical ex-
samples with respect to the general subdivisions pressions have been developed for expressing
of particulate matter. The table also contains a the relation of particle size to settling velocity.
list of the properties for a given size range as These laws, especially Stokes law, serve as a
well as related information such as the limit basis for many size-separation techniques for
of the microscope and the diameter of pores of particles in the silt and clay classes. Particularly
various types of filtering materials. Ordinarily, important for the discussion to follow are the
sediment-size analyses are made only for mate- pipet and BW tube (bottom-withdrawal-
rial down to about 0.002 mm in the coarse- tube) methods. For most particle-size analysis,
suspensions category. Further separations either sieves or the VA tube are used for 0.062
would require excessive time and lack practical mm and larger sizes. The latter has the advan-
value in hydraulic engineering work ; however, tage of being a hydraulic method and helps to
in many scientific fields a knowledge of the size insure continuity among the size classes repre-
distribution of material finer than 0.002 mm senting the range from the fine to the coarse
can be useful. particles. In the following discussion, the set-
tling medium is assumed tobe water because it is
The settling medium
the universal environment of fluvial sediment
Most methods of determining particle-size and has most of the desirable physical charac-
distribution gradation are based upon the teristics for analytical use.

Table 3.-The subdivision of matter and the resulting properties

[From table 1, U.S. Inter-Agency Committee on Water Resources, Subcommittee on Sedimentation, 18411

SIZECB*B*cTERISTICs
Millimeters...............--mm.. l-o. 001 0. ooGu1-0. OKKHJl
Micrometers................s----~-- 1, ooo-1 O. “E: Y”’ O. ??p:z: E
Nanometers _____...._________ urn.. 1, ooo, oKl-1,c0o 1, GOO-100 loo-l O. “Et Y”’
Relative number of particles-. . __. 10 e10 ‘* 10 1110 ‘8 10 lb10 2’
Relative surface of particles _.__.__ :I:; : 10 s-10 1 10 LlO 8 10 LlO ’

QUALITIES STUDIED COARSE COLLOIDAL COLLOIDAL MOLECULAR

s”SPENsIONs S”SPENSION8 SOLUTIONS eOL”TIONs

Appearance in water __... ..____._. Very cloudy ___._________ Turbid ___.__ . _ . _ __ __ __ _ Clear _.__________._.___._
Particles observed- _._________... With naked eye ._____... With microscope... .__.. With ultra-microscope..
Rate of settllng _.______..__._..__._ Quickly or overnight.. __ Slowly or not at all...... Do not settle ._..____..._
Particles separated from water-. _ With filter p8per.. .__ With clay fllters . . . . . ..__ With ultra-5lter . .._ _.._.
Determination of particle sizes- __ _ Microscope, centrlfugsl Ultra-microscope, ultra-
settling, absorption or centri uge.
dispersion of light.
Form after evaporation... _ __ __ __. Powders and gels _..._ ___ Gels __ _. _ _ __ _ __ _
Soil separate.7 .._________.....__.__ Sand, silt, and (or) clay. _ Suspended clay ..___..... Ultra clay . . .._______..._

NOTE.-Limits of size measurements of microscopes and filtering materials:

Limit of microscope (ultra violet light) . . . .._._.___........._..._____...______._....~.~.~~~..~......~..~.~.~.~~~~~.~~.... loOr
Limit of ultramicroscope . ..__.____._..._.___.~.~~~....._._____..._._._______....._.~..~....______________..._____...~~~~ 10~
Limitofultraftlter.......------..----.-.-.-.-.-.....-----.-.....-----.-.-...-.--.-.-..-.....---...-.-..-..-----.--.---..- llr
Brownian movement starts at about __..___ _______.___.. _____.___..._______.....-.-.-.-----.-. .______._._..____.._.---- C
Diameter pores of hardened filter paper-. _. .._... _____. _....__________..._.-...------.-.-..-.-.---.......---.-...-.-...- 1.5r-2.2~1
Diameter pores ofCh8mberland fllter......................---.-----------------------..-.----..---.-------.-.---.-.-. O.ZvO.+
Diameterofbacteris.....-......----.-.--------.....-------~-~..--..-----.--------------.~.-.------~--.-----.-.-.~--.--- 0.5r-1.2~
Diameter of 2CGmesh particles __._...___________...-.-------.....-.-.---.-----.-.---.-.-.-...-.-----.-.....-..-.-...---- 74~
LABORATORY THEORY AND METHODS FOR SEDIMENT ANALYSIS 9

Stokes law for small particles g=acceleration due to gravity (980 centi-
The nature of particle movement in a settling meters per second,
medium depends on a balance of the accelerat- pJ=density of sphere in grams per centi-
ing and retarding forces which, in turn, depend meter,
on the size, shape, and density of the particle p,=density of fluid in grams per centi-
and the viscosity and (or) turbulence induced meter.
in the fluid. The resistance to motion can be The fall velocity as defined by Stokes law is
any one of three types of drag (Rouse, 1938;
U.S. Inter-Agency Committee on Water Re-
sources, Subcommittee on Sedimentation, 1941,
p. 24):(l) deformation, (2) surface, or (3) form, Assuming that the specific gravity of the parti-
depending on the interrelation of the character- cles is 2.65 and that water is the settling medium,
istics of the particle and fluid. The deformation the fall velocity is
drag is the resistance due entirely to the vis-
cosity of the fluid as the relative movement of 0.8983 d2 Jz
v=
the fluid extends a considerable distance away P Or d=0.9487
from the particle. Stokes law expresses the where &= diameter of particle in millimeters.
viscous fall of particles under such conditions, Figure 2 gives a graphical solution to the Stokes
usually for- quartz sediment particles finer than equation for a water temperature of 20°C. Vis-
0.062 mm. If the fall velocity of a particle is cosity correction factors for use with the graph
somewhat faster than that normal for the Stokes when temperatures other than 20% are used are
law range, then the inertial effects of the fluid tabulated in table 4.
reduce the viscous influence to a thin layer Basic assumptions in the application of Stokes
around the particle which produces the effect law are (1) that the particles be of sufficient size
of surface drag. For the larger sizes of particles, to overcome Brownian movement, or be suffi-
the inertial effects of the fluid become in-
ciently large for the fluid to be considered homo-
creasingly important in the flow around a par- geneous in relation to the particle size, (2) that
ticle, and therefore the form type of resistance
the particle be smooth and rigid and of near
dominates. Such flow causes a wake or low
spherical shape, and (3) that the particle be
pressure area behind the particle, and eddies
falling at a uniform velocity as it would in a
are formed; the net result is a force opposing
fluid of unlimited extent with the resistance of
motion.
the fall due only to the viscosity of the fluid.
Stokes law, as developed in 1851, states that
These assumptions are in general not difficult
the viscous resistance to fall of a particle in a
to satisfy (U.S. Inter-Agency Committee on
fluid is equal to the net downward force of the
Water Resources, Subcommittee on Sedimenta-
particles. Remembering t,hat the net downward
tion, 1941, p. 33-36).
force is the force of gravity on the spherical
Greatest consideration must be given to the
particle minus the buoyant force of the liquid,
effect of the proximity of particles to each
or
other whereby the deformation effects extend
413 (r$ Psd -4/3 CT? P/d, into the path of nearby particles. It is also con-
then these forces can be stated as ceivable that a group of particles creating a
region of high concentration may cause a local-
ized area of higher density than the settling
where medium and thus settle as a unit at an abnor-
r=radius of sphere in centimeters, mally high rate. The sphericity of the particle is
cc=dynamic viscosity of the fluid in dyne- of no great concern since the particle velocity is
seconds per centimeter, compared with sedimentation or nominal diam-
v=velocity of fall in centimeters per eter for most investigations. From a practical
second, viewpoint, a consideration of the shape of these
10 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

fine particles for the purpose of relating settling reduce the viscous influence. The Reynolds
velocity to physical size would be nearly an in- number, R, indicates the ratio of inertial forces
surmountable task, hence, the use of the sedi- to viscous forces and, therefore, the character
mentation diameter concept. of the resistance to fall. For spherical particles
falling in a fluid, R is a dimensionless quantity
0.5000 defined as vdpJp. The resistance to fall of these
larger particles is CDA8p,/2. where CD is the
coefficient of resistance and A is the projected
0.2000 cross-section area of the particle. This resist-
ance is equated to the net downward force as
0s 0.0100 0 1000 previously shown in deriving Stokes law.
%i
k-5
“, 0.0050 0.0500 % Cdtip,= %+dp,- P,).
z
Li The relationship between CD and R has been
5+
2 0.0020 0.0200 established empirically for spheres and for
5 some other shapes (Inter-Agency Report No. 4,
z- 0.0010 0.0100 p. 39, and Inter-Agency Report No. 12, p. 21
” and 51). The data indicate that Stokes law is
s
9
y 0.0005 applicable up to R=O. 1 and sometimes up to
0.0050
2 0.4. The relationship between CD and R is of
academic interest only because methods of
0 0002 analysis using sedimentation methods must be
calibrated in terms \pf some definite, easily
0 0001 understood, and readily reproducible unit of
0.001 0.002 0.005 0.010 0.020 0.050 0.100 sediment size. Standard fall velocity and fall
DIAMETER OF SPHERE, IN MILLIMETERS diameter provide a simple foundation for the
expression of the size distribution of samples.
Figure I.-Relation between fall velocity ond diameter
(Stokes law) for quartz spheres in water at 20°C. Particle size for these larger sizes is based on a
comparison with actual samples from which
Table 4.-Fall velocity correction factors for use with figure the individual particle fall velocities have been
2 when temperatures are different from 2O’C.
determined.
correctionfactor
Correction fwtm Correctionfactor
TemFcture 0 PT
- Temy;ture 0 BT
- Significant Figures For reporting result;
cm PlD
The number of significant figures to be used
18 1. 051 I 29 0. 811
19 1. 027 30 . 794 in recording data from analysis in the sediment
E 1. 027
000 30 . 794
776 laboratory is the product of compromises be-
;:21 1.. 976
976
000 ZB
ZB . 760
22
22 . 953
953 33 . 746 tween the need for uniformity in the final tables
E: . 930
908
930 34
34 ..714
729 of data, precision of measurement, and also the
E: . 908 I 35 .714
. 700 degree of precision necessary for the use of the
f l . 867
887 i5
38
36 . 700
673 data.
27 . 848 40 . 648
28 .829
For concentration of sediment in a sample, it
is practical and desirable to report to the nearest
1 mg/l up to 999 mg/l and then for higher values
Drag-Reynolds number
to use three significant figures. Often the third
significant figure in the series 500 to 999,501O to
Stokes law is not applicable for quartz 9990, and so forth, has very weak significance,
particles larger than about 0.060 mm falling in but is justified on the basis of uniformity. Again,
water because the inertial effects of the fluid in some instances it may be desirable to usz
LABORATORY THEORY AND METHODS FOR SEDIMENT ANALYSIS 11
the nearest 0.1 mg/l up to 9.9 mg/l, especially ment concentration. Thus, the filtration method
between 0.1 and 0.9 mg/l. These recommenda- may best be used on samples ranging up to about
tions are based on the assumption t-hat the net 10,000 mgJ1 of sediment that is mostly sand
sediment can be weighed to the nearest 0.0001 g and about 200 mg/l of sediment that is mostly
and the water-sediment mixture can be weighed clay. The evaporation method may be best used
to the nearest 1 g. when the concentration is more than these
With respect to size gradation, the goal limits.
should be to publish the percentage of material
in each fraction to the nearest whole percent. In Evaporation method
some analyses, where fractions may contain 2 The evaporation method consists of allowing
percent or less, it may be useful to report to the the sediment to settle to the bottom of the sam-
nearest 0.1 percent. Whenever practicable, the
ple bottle, decanting the supernatant liquid,
net quantity of sediment in a given fraction
washing the sediment into an evaporating dish,
should be weighed to the nearest three signifi-
and drying it in an oven. The main advantage
cant figures. With balances weighing to the near-
of the method over the filtration method is the
est 0.0001 g, the significant figures must be re-
simplicity of equipment and technique. The
duced to two when the quantity is less than
method works well if the sediment readily set-
0.0100 g.
tles to the bottom of the sample container be-
cause of its coarseness or because of the natural
Procedure for Determining flocculation of the suspended clay. Conversely,
with suspensions of naturally dispersed clay,
Suspended-Sediment Concentration the settling time may make the method imprac-
tical unless special procedures are used.
‘The best methods for making determinations
of concentrations and particle-size distributions NOTE A.-Special procedures to handle unsettled sam-
from suspended-sediment samples must be se- ples involve the use of (1) a filter tube, in which case,
it may be just as easy to adopt the complete filtration
lected from numerous possible procedures. The
method, or (2) a flocculating additive to reduce the
samples may contain only minute quantities of settling time. In the Erst procedure, the sample is al-
sediment of very small particle size or may con- lowed to #settle as much as practical and then, by use
tain large quantities of sediment with a wide of a Berkefeld Elter tube (or a similar device) and a
range of particle size. Samples with such ex- vacuum apparatus, most of the native water is removed
from each bottle. The sediment in suspension adheres
tremes usually require the use of special pro-
to the outside of the filter tube and the water passes
cedures. Samples for particle-size analysis al- through the walls. The materia1 on the outside of the
ways require special processing and computa- tube is removed by back flushing with native water
tional procedures to insure the determination of from the sample and can be treated in the same man-
the concentration as well as the particle-size dis- ner as the dissolved solids content.
A second special procedure is to use a Eocculating
tribution with only one drying of the sediment.
agent to reduce the settling time of ‘the naturally
Each of the several methods for determining dispersed clays. This method will require the use of
concentration has advantages and disadvan- a correction factor to ,be applied to the weight of dried
tages. The two most commonly used methods are sediment. Such a factor is likely to ,be more of an esti-
evaporation and filtration. The filtration mate ‘than an exact value because of the difficulty in
determining the amount of flocculant incorporated in
method may be somewha,t faster than the evap-
the Enal weight and the amount of dissolved ma-
oration method for samples of low concentra- terial precipitated. Hence, a flocculant should not be
tion. However, larger quantities of sediment used unless necessary and then only a minimum
tend to clog the filters and inhibit the use of this amount. As the samples decrease in concentration, the
method for the higher concentrations. The evap- opportunity for significant error increases because of
the estimated correction. Dilute HCl or C?aC% are the
oration method requires an adjustment for dis-
flocculants commonly used. The one most effective will
solved solids if the dissolved solids content is depend on the nature of the sediment and the dissolved
high, especially for samples having low sedi- minerals in the water. Extreme care is necessary in
12 TECIUXIQUES OF WATER-RESOURCES INVEX3TIGATIONS

using either dilute HCl or CaCL as a flocculant. Carbon- 0.1000

a,t,e minerals are readily decomposed in acid 80htiOUS

0.0500
and many other minerals are slightly soluble in weak
acid solutions; therefore, the use of hydrochloric acid
is not recommended. ‘Calcium chloride is highly hy- 0.0200

groscopic; therefore, if the sample has been treated

O.OloLl
with calcium chloride, the use of closed weighing b&-
tles is recommended instead of open evaporating dishes 0.0050
for the determination of concentration.
0.0020
The supernatant liquid is carefully decanted
from the sample so that the sediment in the 0.0010
bottom of the container will not be disturbed
0.0005
nor removed. A small quantity of the super-
natant liquid is therefore left with the sedi-
O.WO.2
ment. A specified amount of such liquid should 10 20 50 loo 200 5w lwa 2ow 5wo 10,ooo
be left with each sample, usually about 20 to DISSOLVED SOLIDS CONCENTRATION.
IN MILLIGRAMS PER LITER
50-~2 ml (milliliters). This specified amount is
especially significant if a correction for dis- Figure 3.-Weight of dissolved solids for Qiven size of aliquot
and given concentration.
solved solids is necessary. Distilled water is
used to wash the sediment and remaining liquid
into the evaporating dish. The contents of the Filtration method
dish are then dried in an oven at a temperature
that is slightly below the boiling point. If the The filtration method for the determination
boiling point is reached, then sediment may be of suspended-sediment concentration has ob-
lost from the dish. After all visible moisture has vious advantages over the evaporation method
evaporated, the temperature should be raised and should be used whenever it is practical to
to llO°C for at least 1 hour. do so. The method usually utilizes a Gooch
The weight of natural dissolved solids in the crucible in conjunction with various types of
supernatant water is included in the weight filter material. The Gooch crucible is a small
of dry material obtained by the evaporation porcelain cup of approximately 25 ml capacity
method and must be subtracted to determine with a perforated bottom. The crucible is easily
the weight of sediment. The following may be adapted to an aspirator system and vacuum
used as a guide to determine if a correction for filtration ; it is lighter in weight and consumes
the weight of the dissolved solids is necessary. less oven and desiccator space than the eva-
A 5 percent or more improvement in the results porating dishes; its tare weight is less likely
of the sediment determination will be realized to change during the weighing because less sur-
when the dissolved solids concentration is equal face is exposed to sorption by moisture ; and
to or greater than the sediment concentration, dissolved material passes through the crucibles
assuming that an aliquot of 20 ml of native and thus eliminates the need for a dissolved-
water is used in the evaporation, and that 400 g sohds correction.
of water-sediment mixture is contained in the In most laboratories the Corning # 32960
original sample. For example, a dissolved-solids Pyrex fritted glass crucible and glass-fiber filter
correction should be made when the dissolved disks, # 934-AH or similar specifications, have
solids concentration is 200 mg/l or higher and proven satisfactory for filtration of most types
t,he sediment content is 200 mg/I or less. of sediment samples. These filter disks show no
Similarly, a 2 percent improvement will result weight loss during filtration, save considerable
if the dissolved solids concentration is 40 per- preparation time, result in more uniform filtra-
cent or more of the sediment concentration. The tion, and make cleaning of the crucible easier
weight of the dissolved-solids correction for dif- than do other types of filters. However, the
ferent aliquots and a range of concentrations filter disks become clogged rather rapidly when
can be obtained from figure 3. some types1- of fine sediment are filtered.
0
LABORATORY THEORY AND METHODS FOR SEDIMXNT ANALYSIS 13
Better filtration results have been obtained the adsorbing agent should reduce the settling
for these fine-grained sediment samples by time and prevent the rapid clogging of the
using a glass-fiber filter disk in conjunction with filter. The main advantage of the adsorbing
an asbestos mat. The crucible with this extra agent over the use of the flocculating agent is
asbestos mat is prepared by placing the glass- that a precise correction, instead of an estimated
fiber filter disk in the crucible while vacuum is correction, can be made for the amount retained
applied and then pouring an asbestos slurry on on the filter. Acids or salts are not usually
top of the disk also while vacuum is applied. recommended because undesirable reactions
(See the next paragraph ‘concerning the prep- with the sediment are likely to occur. As already
aration of asbestos mats.) During the filtration mentioned, the use of a fiber-glass filter disk in
process, the somewhat coarse asbestos mat re- conjunction with an asbestos mat will eliminate
tains much of the sediment that would ordi- most filter problems.
narily clog the glass-fiber disk. The glass-fiber Sometimes it is possible to utilize another
disk prevents loss of asbestos or loss of very advantage of the filtration over the evaporation
fine sediment that would pass through the as- method when the concentration or the amount
bestos mat. This filter is recommended if inac- of sediment is small. Specifically, preparation
curacy or inconvenience results from the use of time can be saved by using the same mat two or
a single type of filter because little additional more times. When the crucible and mat are re-
time is required for the preparation of the dou- used, the gross weight of the prior use becomes
ble glass-fiber asbestos-type filter. the tare weight for the new use.
If glass-fiber filters are not available, asbestos
is considered satisfactory because of its inert- Laboratory forms
ness, resistance to heat, and low hygroscopicity.
The asbestos filter mats can be prepared in Figures 4,5, and 6 show forms used in record-
two ways. The simplest and most convenient ing data and making computations for deter-
way is to prepare a slurry of shredded asbestos mination of suspended-sediment concentration.
and distilled water, and pour a small volume of These forms (SedimRnt Concentration Notes,
this slurry into the crucible while vacuum is Depth Integrated Smples [Short form]; Sedi-
applied. A uniform mat of asbestos is thus ment Ctientration Notes, Depth Integrated
formed in the bottom of the crucible. The mat Samples [ Comp+e?wn&ve f mm]; Sediment
should be rinsed with distilled water while Concentration Notes, Point Integrated Sm-
vacuum is still applied ; the crucible is then pies) provide an orderly means of recording the
oven dried, cooled in a desiccator, and the tare sample information, the necessary weights, and
weight determined. Another method for prep- the results of computations for the concentra-
aration of the asbestos mat is to place a small ,tion of the sample. On the Short form for depth
portion of loose, uniform asbestos on the bottom integrated samples (fig. 4)) space is provided
of the crucible, and wet the asbestos with dis- for 20 bottles or samples; the other two forms
tilled water while vacuum is applied. provide for only 10. The forms serve as a record
If the sediment sample contains extremely of samples collected for each sampling site. Each
fine material, an actual loss of fine material bottle should be recorded in the chronological
through the filter may result in a turbid filtrate. order of sampling at ,the time of sample weigh-
This problem is treated in one of three ways: ing. Appropriate notes can be made on the
(1) add a known quantity of an adsorbing form if one or more samples are later with-
agent, such as activated carbon, which adsorbs drawn or particle-size analysis or wmposited
the colloids as a result of the difference in elec- for concentration. If more space is needed for
tric charges, (2) impregnate the filter with remarks, reference by number should be made
sediment and then re-filter the filtrate, or (3) to the back of the sheet ; or, expanded notes can
use a flocculating agent of the acid or salt type. be retained on the front by using space
In addition to preventing loss through the filter, ordinarily used for other bottles.
TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

1
-
:P
L1
LL

r IW IN3
JO

-II4
JO
1 -H31: lH313M
16 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

1
- -

- - 1

- - -

- - -

- - - -
k
t s
5z0
i
P
i
tid
M
.g
p ;
.y
0 : 5
a
fi 3
lN3W tax
d 3ldWt’
JO JO
lH913M
I-
lH913A - -
LABORATORY THEORY AND METHODS FOR SEDIMENT ANALYSIS 17

The entrv for D.S. Cow. (dissolved solids) 4. At the sample weighing station, record the sample
information from the sample labels on the
Y

is used when the evaporation method is used appropriate concentration form. (See figs. 4, 5,
and then only when the dissolved-solids con- and 6.)
tent is relatively high, when an adsorbing or 5. Record the gross and tare weights to the nearest
flocculating agent has been added to the sample, gram for each sample, assuming the net sample
or when an adsorbing agent is used with the to weigh in exe- of 200 g. It is often desirable
to use a balance with a tare removing attach-
fikration method. Allthough details concer- ment and a procedure that will directly yield the
ing the principles of dissolved-solids deter- net weight of the sample. This avoids copying
minations are discussed in the last section of and computation errors, but puts greater
this chapter, “Related water-quality analysis,” responsibility on the operator.
it is appropriate to outline the essentials of the 6. Store the weighed bottles on a convenient rack or
table for several hours, or overnight, to assure
determinations here. The dissolved solids con- ,that the sediment settles from suspension. In
tent of a stream usually does not change ap- the event that the sediment does not settle, fol-
preciably from day to day for normal flow, but low the procedure outlined in “Note A (p. ll).”
usually diminishes rapidly with increasing Samples to be analyzed for particle-size distribu-
storm runoff. Such periods may be determined tion should be separated from those to be
analyzed for concentration only.
by a study of the gage heights recorded on the 7. Some laboratories at this point obtain and record
samples and (or) by the increase of sediment the conductance of the supernatant water for
content associated with storm runoff. A single each sample set. This is not published but is used
dissolved-solids determination is usually ade- as a control on the person taking the samples
quate for several daily samples when the and as an aid in finding and correcting errors
made in recording the field and lab data. In
streamflow rate is relatively uniform. The com- this case the comprehensive form (ilg. 5) would
posite determination is made from a sample be used.
prepared by withdrawing nearly equal volumes 8. Decant most of the sediment-free water using
,of the sediment-free water from each daily care not to disturb or remove sediment. If a dis-
sediment sample. The sample, whether for a solved-solids correction is necessary when using
the evaporation method, then the amount of
single day or a composite for several days, is
supernatant liquid left with the s&ment should
evaporated to dryness, heated at 110” C for 1 be approximately equal to one of the aliquots
hour, cooled in a desiccator, and weighed to ob- indicated in ilgure 3.
tain a unit weight. The dissolved solids correc- 9. Use distilled water to wash the remainder into a
tion to be applied to the dry weight of sediment previously weighed evaporating dish or illtra-
tion crucible, depending on the method used.
solids in each evaporating dish is equal to the
Check to be certain that the container number
product of the unit volume correction and the on the form Sediment Gmmntration Notes
volume of water evaporated from the dish. Also corresponds with the giveu sample or samples.
see page 12 and figure 4. 10. The sample bottles must be thoroughly cleaned, air
dried, recapped and packed in carrying cases for
reuse.
Procedure for concentration analysis 11. Dry the evaporating dishes or crucibles and, after
loss of all visible water, heat at about 110” C
With the above information it is possible to for 1 hour. The evaporating dishes must be
compile the following step-by-step procedure dried at a temperature low enough to prevent
for concentration analysis : “spattering” of sediment from the dishes by
boiling action.
1. Inspect the general condition of the samples as 12. Cool the containers in a desiccator at room
they are received at the laboratory. Loose caps temperature.
should be replaced and bottles with excessively 13. Weigh the containers to the nearest 0.6661 g on an
dirty exteriors should be cleaned. analytical balance. The tare weight of the evap-
2. Store samples in a cool dark room to prevent ex- orating dish should be obtained before each use
cessive evaporation and growth of organisms. because of the loss of weight caused by small
3. Arrange the samples for a given location in chips, frequent washing, and slight differences
chronological order while assembling them for in washing or cleaning methods. The tare weight
transport to the sample weighing station. of the crucible must be obtained before each use
18 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

because of differences in weights of the tlltering size analyses should be adequate to descrisbe the
mats, especially when the asbestos mats are used. pertinent characteristics of sediment particles
If sediment concentrations are greater than 1,660
so that satisfactory comparisons can be made be-
mg/l, then it may not be necessary to obtain a
new tare weight each time a crucible with a tween sediments collected from different places
glass-fiber filter mat is used. or from the same place at different times.
14. Compute the net weight of sediment in the container
and deduct, if necessary, the dissolved-solid cor-
rection.
Recommended Frequency OF analyses
15. Compute the concentration of the sample in parts The following recommendations are for the
per million determined on the basis of one mil-
determinations of ultimat,e sizes by use of a dis-
Ron times the ratio of the net weight of sedi-
ment to the net weight of the water-sediment mix- tilled-water settling medium and a chemical dis-
ture. The computed parts per million can then be persant for the silt and clay fractions. The rec-
expressed in the common units of milligrams per ommended frequency of native-water analysis is
liter by application of the conversion factor given discussed in the section “Native water settling
in table 1. It is convenient to make the conversion
medium” in respect to the theories, problems,
from parts per million to milligrams per liter
when the data are transferred from the form and methods of such analysis. Each sample
Sediment Conce~~tration Notes to other uses. Usu- should be depth integrated and represent the
ally concentrations are less than 16,660 parts per true size distribution in the cross section at the
million in which case parts per million equals time of sampling. Thus, for most streams, sam-
milligrams per liter. However, when the concen-
ples must be collected by the ETR (equal-tran-
tration does exceed 16,066 ppm the appropriate
conversion factor should be penciled plainly below sit-rate) method or by sampling at an adequate
the concentration on the form as a reminder to number of the centroids of equal discharge. If
the data user that milligrams per liter is dlf- the “daily” samples collected at one vertical near
ferent from parts per million. midstream are representative of the sediment
characteristics in the entire cross section, the
Determination of Particle-Size selection should normally ‘be adequate to cover
a wide range of flow conditions.
Distribution of Suspended Sediments Judgment concerning the selection of samples
Complete definition of the variations of par- will be enhanced by a study of the conditions
ticle-size distrilbution for a stream throughout for which particle-size samples of past records
the ranges of discharge and with time is pres- for the given station were selected, if available;
ently not economically feasible. It is feasible, and (or) by a study of the sample information
however, to analyze a sufficient number of sam- recorded on the form Sediment Con.cen.tratbn
ples to provide representative data for a range Notes (fig. 4) for the current water year. For
of the many conditions occurring throughout many streams, information on sediment char-
the year. acteristics is usually lacking during the first year
Particle-size data have many uses. One use is of sampling on a given stream ; therefore, some
to compute the probable specific weight of sedi- samples, particularly from high-stage flows,
ment as it will deposit in reservoirs. This com- should be set aside to be analyzed at a later time
putation requires plotting the median particle either for particle size or concentration at the
size against sediment load in tons per day. In- end of the water year when the facts are known
formation concerning the behavior of sediment concerning the range and other conditions of the
particles in different environments may be year’s flow. These extra samples should, of
studied and, therefore, requires analyses for course, be tightly sealed, tared and stored in a
particle-size distribution of fine particles in dark place.
both native and dispersed settling media. The Size analyses of suspended sediment for the
size distribution of both suspended and bed ma- so-called daily station should be made on a mini-
terial has an important use in the computations mum of five or six samples per year if selected to
of total (measured plus unmeasured) sediment represent various flow conditions and range of
discharge. The frequency and kind of particle- concentrations. Additional samples, up to a