Handbook for Sampling and Sample Preservaton of Water and Wastewater


United States	Environmental Monitoring and Support December 1980
Environmental Protection	Laboratory
Agency	Cincinnati OH 45268
Research and Development
vvEPA Handbook for Draft
Sampling and
Sample
Preservation of
Water and
Wastewater

-------
EMSL0220
HANDBOOK FOR
SAMPLING AND SAMPLE PRESERVATION
OF WATER AND WASTE WATER
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268

-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring and Support
Laboratory, U.S. Environmental Protection Agency, and approved for publication.
Mention of trade names or commercial products does not constitute endorsement
or recommendation for use.
ii

-------
FOREWORD
Environmental measurements are required to determine the quality of ambi-
ent waters and the character of waste effluents. The Environmental Monitoring
and Support Laboratory - Cincinnati engages in the following activities:
o Develops and evaluates techniques to measure the presence and concen-
tration of physical, chemical, and radiological pollutants in water,
waste water, bottom sediments, and solid waste.
o Investigates methods for the concentration, recovery and identifica-.
tion of viruses, bacteria, and other microbiological organisms in
water. Conducts studies determine the responses of aquatic organ-
isms in water.
o Conducts an Agency-wide quality assurance program to assure standard-
ization and quality control of systems for monitoring water and waste
water.
Standardized procedures for analyses of quality control become academic
if samples are not representative of their original environment or if changes
of constituent concentrations occur between time of sampling and analysis.
This publication presents techniques for sampling and sample preservation to
help alleviate these problems. Procedures have been standardized as much as
possible throughout this document. However, sampling techniques could not be
predetermined for all situations, so the use of statistical procedures to
establish location and frequency of sampling, number of samples, and parameters
to be analyzed is recommended when other guidelines do not exist. Sample
preservation methods and holding times are included for the 71 parameters
listed for the National Pollutant Discharge Elimination System program, prior-
ity pollutants, and selected biological species. Special handling or sampling
techniques are also included for the individual constituents. Personnel estab-
lishing a sampling program should find sufficient information to determine the
best techniques to apply.
Dwight G. Ballinger
Director
Environmental Monitoring & Support
Laboratory - Cincinnati
ill

-------
ABSTRACT
This research program was initiated with the overall objective of providing
guidelines for sampling and sample preservation of waters and wastewaters.
Information obtained from a review of the literature and the results of a
survey of field practices provides the basis for guidelines in general sampling
techniques, automatic samplers, flow measuring devices, a statistical approach
to sampling, preservation of physical, chemical, biological and radiological
parameters, and sampling procedures for waters emanating from municipal, indus-
trial, and agriculture sources. Sampling procedures for surface waters and
sludges are also included.
iv

-------
CONTENTS
Foreword ....... 		iii
Abstract		iv
Acknowledgement 		x
1.	Introduction 		1
2.	General Considerations for a Sampling Program 		3
2.1	Objectives of Sampling Programs 		3
2.2	Sampling Locations 		7
2.3	Sample Collection Methods 		10
2.4	Type of Sample		27
2.5	Planning a Sampling Program 		41
2.6	Field Procedures 		49
2.7	References		53
3.	Flow Measurements		55
3.1	Closed Conduit Flow Measurement		56
3.2	Flow from Pipes Discharging to the Atmosphere ....	67
3.3	Open Channel Flow Measurements		74
3.4	Miscellaneous Flow Measurement Methods 		104
3.5	Secondary Devices 		114
3.6	References	130
4.	Statistical Approach to Sampling 		133
4.1	Basic Statistics and Statistical Relationships ....	133
4.2	Determination of Number cf Samples 		176
4.3	Determining Sample Frequency 		179
4.4	Determination of Parameters to Monitor 		182
4.5	In--plant Sampling and Network Monitoring	191
4.6	References			211
5.	Sampling Municipal Wastewaters 		213
5.1	Background 	213
5.2	Objectives of Sampling Programs 		213
5.3	Frequency of Sampling		214
5.4	Location of Sampling Points 		215
5.5	Number of Samples			220
v

-------
CONTENTS (Continued)
5.6	Parameters to Measure	221
5.7	Type of Sample	221
5.8	Methods of Sampling	222
5.9	Volume of Sample and Container Type	223
5.10	Preservation and Handling the Samples 		223
5.11	Flow Measurements		224
5.12	References		225
6.	Sampling Industrial Wastewaters 		226
6.1	Background		226
6.2	Objectives of Sampling Programs 		226
6.3	Frequency of Sampling 			227
6.4	Location of Sampling Points 		228
6.5	Number of Samples		229
6.6	Parameters to Measure		233
6.7	Type of Sample		233
6.8	Method of Sampling		234
6.9	Volume of Sample and Container Type		235
6.10	Preservation and Handling of Samples 		235
6.11	Flow Measurement		235
6.12	References		236
7.	Sampling Agricultural Discharges 		237
7.1	Background		237
7.2	Objectives		 					237
7.3	Frequency of Sampling		237
7.4	Location of Sampling Points 		238
7.5	Number of Samples	
7.6	Parameters to measure 	
7.7	Type of Sample	
7.8	Method of Sampling		^40
7.9	Volume of Sample and Container Type		240
7.10	Flow Measurement	
7.11	References		^
8.	Sampling Surface Waters, Aquatic Organisms and Bottom
Sediments	246
8.1	Background		246
8.2	Objectives of the Study 		246
vi

-------
CONTENTS (Continued)
8.3	Parameters to Measure 	246
8.4	Location of Sampling Points 	 247
8.5	Number of Samples 	252
8.6	Frequency of Sampling ....... 	 256
8.7	Method of Sampling	256
8.8	Types of Samplers for Aquatic Organisms 	 256
8.9	Volume of Sample and Container Type	270
8.10	Preservation and Handling of Samples 	 270
8.11	Flow Measurement	270
8.12	References	270
9. Sampling of Ground Water	271
9.1	Background	271
9.2	Frequency of Sampling	274
9.3	Location of Sampling Points	276
9.4	Parameters to Measure	283
9.5	Type of Sample	289
9.6	References	300
10.	Sampling Sludges			302
10.1	Background	302
10.2	Objectives of Sampling Programs 	 302
10.3	Parameters to Analyze	303
10.4	Location of Sampling Points 	 303
10.5	Frequency of Sampling	304
10.6	Number of Samples	305
10.7	Type of Sample		 305
10.8	Method of Sampling	
10.9	Volume of Sample and Container Type	305
10.10	Preservation and Handling of Samples	307
10.11	Flow Measurement	
10.12	References 	
11.	Suspended Solids Sampling 	 .... 309
1.1	Representative Sampling Theory 	 309
1.2	Segregation Sampling Error 	 310
1.3	Field Sampling	317
1.4	Laboratory Subsampling 	 319
1.5	Guidelines of Sampling of Suspended Solids 	 320
vxi

-------
CONTENTS (continued)
11.6 References 			322
12.	Sampling, Preservation and Storage Considerations for
Trace Organic Materials 		323
12.1	Sample Collection Method 		331
12.2	Sediment Sampling 		342
12.3	Sampling Location 		343
12.4	Sample Container 		343
12.5	Sampling Procedure and Pretreatment of
Sample Equipment	346
12.6	Sample Preservation and Storage 		347
12.7	References 	353
13.	Sampling Radioactive Materials 		356
13.1	Background 	356
13.2	General Considerations 		358
13.3	Frequency of Sampling	361
13.4	Location of Sampling	362
13.5	Sample Volume	363
13.6	Sample Containers	363
13.7	Sample Filtration	364
13.8	Sample Preservation 		364
13.9	General Sampling Procedure - Water & Wastewater . .	366
13.10	Radiation Safety 		367
13.11	References 	368
14.	Collecting and Handling Microbiological Samples 		370
14.1	Background 	370
14.2	Analytical Methodology 			370
14.3	Sample Bottle Preparation	3'75
14.4	Sample Methods and Equipment 		377
14.5	Sample Frequency and Site Selection . 		381
14.6	Preservation and Transit of Samples 		389
14.7	References 	391
15.	Sample Identification and Chain of Custody Procedures . .	392
15.1	Sample Identification 		392
15.2	Chain of Custody			393
15.3	References 	402
viii

-------
CONTENTS (Continued)
16.	Quality Assurance 		403
16.1	Objectives	403
16.2	Elements of a Quality Assurance Plan	404
16.3	Personnel Training 		405
16.4	Quality Assurance in Sampling 		407
16.5	References	413
17.	Sample Preservation 		414
17.1	Methods of Preservation	414
17.2	Containers	422
17.3	Holding Time	426
17.4	Sample Volume	442
17.5	References	443
Appendix A	446
Appendix B		448
ix

-------
ACKNOWLEDGMENT
Much of the material in this manual is based on an Environmental Protec-
tion Agency (EPA) handbook prepared by Envirex Inc. Acknowledgment is given to
Envirex Inc. for use of the handbook. Special acknowledgment is given to the
staff of the Quality Assurance Branch, Environmental Monitoring and Support
Laboratory - Cincinnati, Ohio for their technical support. In particular, the
direction and support of Mr. Edward L, Berg, Project Officer, is appreciated.
Finally, thanks are extended to all the staff members of The Bionetics Corp-
oration, - technical, administrative and clerical - who participated in this
project and contributed to its success. Special thanks are given to the fol-
lowing technical staff members who contributed significantly to this report:
A.J. DiPuccio, T.V. Gala, V.L. Kowalski, G.Becus of the University of Cincin-
nati, Robert Graves and Harry Kolde of EPA.
x

-------
CHAPTER 1
INTRODUCTION
Obtaining representative samples and then maintaining the integrity of
the constituents is an integral part of any monitoring or enforcement program.
Standardization of the analytical techniques has been established to a high
degree, but the result of analysis is only as good as the sampling and the
sample preservation. The purpose of this handbook is to present the best
techniques currently available for sampling and sample preservation. The
recommendations were developed from an extensive research report which
included a literature review and survey of current laboratory and field prac-
tices. The handbook will allow personnel to determine the most effective
procedures for their specific applications.
In sampling, the objective is to remove a small portion of an environ-
ment that is representative of the entire body. It is then obvious that im-
proper sampling will give erroneous results. Once the sample is taken, the
constituents of the sample must stay in the same condition as when the sample
was collected. The length of time that these materials will remain stable is
related to the preservation method.
The sampling technique is determined by the type of water or wastewater
to be sampled. Salt waters are not included. Therefore., the following areas
are addressed in this handbook:
1.	Municipal wastewaters
2.	Industrial wastewaters
3.	Surface waters and sediments
4.	Agricultural runoff
5.	Wastewater sludges
6.	Groundwater
1

-------
General information on automatic samplers and flow monitoring is also
included.
Statistical methods have been presented in this handbook and will be
used to determine the following aspects of sampling programs:
1.	Number of samples	3. Location of sampling
2.	Frequency of sampling	4. Parameters of measure
Special consideration is given to sampling for suspended solids, trace
organics and radioactive substances.
Preservation methods are related to the parameters to be analyzed so, in
this handbook, these techniques are classified by parameter. The (71) para-
meters specified for the National Pollution and Discharge Elimination System
(NPDES) permit program in the Federal Register of October 16, 1973, priority
pollutants, and selected biological parameters are included.
2

-------
CHAPTER 2
GENERAL CONSIDERATIONS FOR A SAMPLING PROGRAM
Most definitions of water quality are use-related. Each water use pro-
duces wastewaters which are responsible for different types of pollution. For
example, thermal pollution is associated with power plant discharges; eutrophi-
cation of lakes is due to nutrient discharges; fish kills result from dis-
solved oxygen deficiency or toxic substances from industrial discharges. The
broad spectrum of waters (ground water, surface waters, lakes, estuaries,
coastal waters) and an equally broad spectrum of wastewaters (municipal wastes,
industrial wastewaters, surface run-offs) make monitoring of water quality a
formidable task. Sampling is one of the key elements in a monitoring program.
However, there is no unique sampling program that applies to all types of waters
and wastewaters. Nevertheless, features of a sampling program which are in-
cluded for all types of waters and wastewaters include:
1.	Objectives of Sampling Program
2.	Location of Sampling Points.
3.	Types of Samples.
4.	Sample Collection Methods.
5.	Flow Measurements.
6.	Field Procedures.
2.1 OBJECTIVES OF SAMPLING PROGRAMS
The objectives of a sampling program can be classified into four main
3

-------
categories, namely, planning, research or design, process control and regu-
lation. These objectives in an overall water quality program are interrela-
ted and cover different stages from planning to enforcement. In relation to
these objectives, the different sampling programs are compared in general
terms in Table 2.1. Since the objectives of a program directly affect all
aspects of sampling and laboratory analysis, determination of the objectives
is the first step in planning a sampling program.
2.1.1	Planning Objectives
Monitoring objectives of interest to an areawide or basin planner in-
clude :
1.	Establishment of baseline conditions.
2.	Determining assimilative capacities of streams.
3.	Follow effects of a particular project or activity.
4.	Identifying pollutant source.
5.	Assessing long-term trend.
6.	Allocating waste load.
7.	Project future water characteristics.
2.1.2	Research Objectives
Sampling is an essential part of most water/wastewater research projects
and is conducted to accomplish one or more of the following objectives:
1.	Determining the treatment efficiency for a unit process or overall
treatment system.
2.	Determining the effect of changes in process control variables.
3.	Characterizing influent and effluent streams and sludges.
4.	Optimizing chemical dosages, loadings for carbon adsorption columns,
4

-------
TABLE 2.1 COMPARISON OF SAMPLING PROGRAMS BASED ON OBJECTIVES

Objective
Planning
Research
Process Control
Regulatory


Design


Scope
General
Specific
Specific
Specific
Goals
Establish trends
New developments
Operation
Verification,

benchmarks
Modifications
quality control
compliance,

background levels.
Improvements

enforcement
Effort
Nonintensive and
Intensive
Nonintensive and
Nonintensive

unlimited
and limited
1imited
and limited

-------
for advance waste treatment processes or treatment of drinking
water.
5. Ascertaining health effects of effluents sludges, drinking waters
and ambient waters.
2.1.3	Process Control Objectives
Sampling to control water/wastewater treatment process and associated
systems is conducted primarily for internal use to accomplish one or more of
the following objectives to:
1.	Producing an effluent of the highest quality.
2.	Optimizing and maintaining physical, chemical, and biological process
control variables that affect treatment efficiency, i.e. mixed
liquor suspended solids, sludge withdrawal rate, dissolved oxygen,
chemical dosages, etc.
3.	Determining resource recovery from unit processes.
4.	Allocating the cost of treatment to a unit within a complex of unit
processes.
5.	Determining substances that are toxic or interfere with the treatment
system.
2.1.4	Regulatory Objectives
Most sampling and subsequent analyses are performed to meet the require-
ments of federal, state, or local regulatory agencies; or regulatory agencies
will sample and analyze to assure compliance. An example of regulatory moni-
toring is the National Pollutant Discharge Elimination System established in
accordance with the Federal Water Pollution Control Act Amendments of 1972
(P.L. 92-500). Specific objectives in collecting regulatory data vary con-
6

-------
siderably and often, overlap, but generally include the following:
1.	Verifying compliance with effluent limitations.
2.	Verifying self-monitoring data.
3.	Verifying compliance with NPDES permit.
4.	Supporting enforcement action.
5.	Supporting permit reissuance and/or revision.
6.	Supporting other program elements such as, water quality standards,
requiring wastewater data.
2.2 SAMPLING LOCATIONS
2.2.1	General Considerations
Usually, the sampling program objectives define either the approximate
or precise locations for sampling, e.g., Influent and effluent to a treatment
plant or water supply intake. Often, however, the sampling program objectives
give only a general indication, e.g.,effect of a surface runoff on a river
quality when assessing the quality of drinking water to a large community. For
programs of this type, careful selection of sampling locations is required.
Since water quality varies from place to place In most water systems,
locations appropriate to the information needs of a particular program must
be selected. The nature and extent of spatial heterogeneity may vary with
time, and can also differ markedly between systems of the same type, e.g.,
a typical case may be a zone of mixing of fresh and saline waters. Therefore,
no specific guidelines can be given on the exact locations for sampling; how-
ever, some general points are worth bearing in mind when considering sampling
locations.
2.2.2	Relevant Factors in Selecting Sampling Locations
The selected sampling locations must be representative sites. The term
7

-------
"representative point" is defined in 40 CFR, Part 35, Appendix A, p. 224, 1976
as:
1.	A location in surface waters or groundwaters at which specific
conditions or parameters may be measured in such a manner as to
characterize or approximate the quality or condition of the water
body; or
2.	A location in process waters or wastewaters where specific conditions
or parameters are measured that adequately reflect the actual condi-
tion of those waters or wastewaters.
A major consideration influencing the selection of the sampling locations
is the homogeneity of the water or wastewater. Turbulence and good mixing
enhance the homogeneity or the uniform distribution of constituents within the
body of water or wastewater, e.g., a stream just downstream of a hydraulic
jump or a lake during spring or fall turnovers. Non-homogeneity, on the other
hand results from:
1.	Poor mixing, e.g., thermal stratification in lakes or a river down-
stream of a waste discharge.
2.	Different densities of the constituents, e.g., floating oils or
settling suspended solids.
3.	Chemical or biological reactions, e.g., growth of algae in upper
layers of water body causing changes in pH.
Other considerations for the selection of sampling locations are:
establishment of general characteristics of a large body of water or waste-
water, pronounced degradation of water quality in specific areas, suitability
for flow measurements, convenience and accessibility.
8

-------
2.2.3 Selection of Sampling Locations (1)
Locations of the sampling points based on the considerations mentioned
in section 2.2.2 are:
1.	Homogeneity of Water or Wastewater
•	At significant outlets and inputs of lakes, impoundments, estuaries
or coastal areas that exhibit eutrophic characteristics.
•	At locations upstream and downstream of major population and/or
industrial centers which have significant discharges into flowing
stream.
•	Upstream and downstream of representative land use areas and morph-
ologic zones.
•	From several locations to obtain the required information.
2.	General Characteristics of Water or Wastewater (1):
•	At representative sites in mainstream of rivers, estuaries, coast-
al areas, lak.es or impoundments.
•	In major water use areas, such as public water supply intakes,
commercial fishing areas and recreational areas.
•	At representative sites in the individual waste streams.
•	At the mouths of major or significant tributaries to mainstreams,
estuaries or coastal areas*
3.	Pronoimced Water Quality Degradation:
•	At critical locations (which have the potential for displaying
the most pronounced water quality or biological problems) in water
quality limiting areas.
•	At critical locations within eutrophic or potentially eutrophic
lakes, impoundments, estuaries, or coastal areas.
9

-------
4.	Flow Measurement
• Locations where corresponding discharges are known or can be esti-
mated.
5.	Convenience, accessibility and practicability are certainly impor-
tant but they should be secondary to representativeness of sampling.
2.3 SAMPLE COLLECTION METHODS
Samples can be collected either manually or with automatic samplers.
Whichever technique is adopted, the success of the sampling program is direct-
ly related to the care exercised in the sample collection. Optimal perform-
ance will be obtained using trained personnel.
2.3.1	Manual Sampling
Manual sampling is the oldest method of sample collection. There is
minimal initial cost involved in manual sampling. The human element is the
key to the success or failure of manual sampling programs. It is well suited
to a small number of samples, but is costly and time consuming for routine
and large sampling programs. Table 2.2 lists some of the advantages and dis-
advantages of manual and automatic sampling.
2.3.2	Automatic Samplers
Automatic Samplers usage has increased because of cost savings, capabil-
ity of more frequent sampling, better reliability (2), and the NPDES permit
program.
Currently, there are many automatic samplers available with widely vary-
ing levels of sophistication, performance, mechanical reliability and cost.
Table 2.3 lists different automatic samplers and their characteristic features
(3). However, no single automatic sampling device is ideally suited for all
ro

-------
TABLE 2.2. THE ADVANTAGES AND DISADVANTAGES OF MANUAL AND AUTOMATIC SAMPLING
Type	Advantages	Disadvantages
Manual	Low capital cost
Compensate for various
situations
Note unusual conditions
No maintenance
Can collect extra
samples in short time
when necessary
Probability of increased
variability due to
sample handling
Inconsistency in collection
High cost of labor
Repetitious and monotonous
for personnel
Automatic
Consistent samples
Probability of decreased
variability caused by
sample handling
Minimal labor require-
ment for sampling
Has capability to
collect multiple
bottle samples for
visual estimate of
variability & analysis
of individual bottles
Considerable maintenance
for batteries & cleaning;
susceptible to plugging
by solids
Restricted in size to the
general specifications
Inflexibility
Sample contamination
potential
11

-------
TABLE 2.3 AUTOMATIC SAMPLERS AND THEIR CHARACTERISTIC FEATURES (3)
MANUFACTURER
MODEL NO.
o ~
£ S
A O
< o
DIMENSIONS
WO.» DPTH . * HT.
or DIA.x HT.
r —
° 5
SAMPLE
u
a. -/
MATERIALS EXPOSED
TO SAMPLES
II!
io
i
S"
9
x s
*-
z
TYPE OF
pump

CONTROLS
POWER
PORT ABLE I
or FIXED 1


EC O
D »¦
a. O
low
Top. f
'op S
olid
tale
u
«
«
a
•
£
f
s.
1/1

No.
Cap.
(mil
u
Bottles
Tubing
Other
BIF Sanitrol
41-4
670
27.3 k 25.4 xVAR
18.16
1
7570

Nalgene
Tygon
Fiberglass

762

Dipper


X

X



F
Brailsford
EVS-3B
672
30.5 * 22.9 « 48.3
8.72
1
3785

'olypfopylene
Tygon
Plexiglas
10.2
182
3.16
Vacuum


X

X
X


P
Brailsfoid
DC-F
296
30.5 x 24 * 48.3
8.72
1
7570

Polypropylene
Tygon
Teflon
23.2
213
3.16
Piston


X


X


P
Brailsford
DU-Z
373
30.5 * 22.9x 48.3
B.72
1
7570

Polypropylene
Tygon
Teflon
23.2
213
3.16
Piston


X


X


P
Brailsfoid
EP
373
Small
L
1
3785

Polypropylene
Tygon
Teflon
23.2
213
3.16
Piston


X


X


P
BVS
PP-100
700
31.8 * 25.4 x 4E
35
1
9463

Plastic
Tygon
PVC

6096
3.16
Pressure


X



X

P
BVS
PPR-100
900
43.2 x 49.5 x 45.1

1
5678
Ref.
Plastic
'Tygon
PVC

6096

Pressure


X


X
X

P
BVS
SE-400
2700
61 x 61 x 122
79.5
1
18.925
Ret.
Polyethylene
Plastic
PVC

975
12.7
Submersible


X
X
X



F
BVS
SE-600
2900
61 x 61 x 122
79.5
1
18.925
Ref.
Polyethy ten*
Plastic
PVC


50.8
Submersible


X
X
X



F
Bristol
M-4K7
941
7.6 x 30.4
3.2
1
3785

Polypropylene

Stainless



Pipeline


X
X
X



F
Chandler
SR-10
2245
27.2 x 59.7 x 108
45.4
1
8000
Ref.
Polyethylene
PVC
U
H
671

Vacuum
*

X
X
X



F
Collins
40-2R
1343
50.8 x 61 x 122
100
1
18,925
Ref.
Poiyethylene
Polyethylene
Polypropylene
H
610
9.5
Moyno


X

X



F
EMA
200 AC
239
20 x 83
9.1
1
U
Ice
U
Plastic
Aluminum

77
9.5
Solenoid
Plunge*


X

X
X


F
ETS
FS-4
1100
108 x 46 * 55
31.8
12
3785

Plastic

Noryl
L
B83

Peristaltic


X

X



P
Fluid Kinetics
Cuttom
Deii^n





Ret.








X
X
X
X



F
FIIC Corp.
Tiu-Test
2850
49.6 x 60.4 x 131
147.6
1
7500
Ref.
Polyethylene


93.3
457
50.8
Centrifugal


X
X
X



F
Horizon
7578
600
40.6 x 23.5 x 57.2
12.7
1
9463

Polyethylene
Tygon
Silicone

914
4.8
Peristaltic


X


X


P
Hydragard
FP
370
10.2 x 74
3.2
1
U

u
PlBStiC
Stainless


9.5
Pressure

X
X



X

P
Hydra-Numatic
HNS
1980
91.4 x 33.4 x 91.4
90.8
1
18,925

Polyethylene
Tygon
Bronze
75
457
12.7
Impeller

X
X

X



F
ISCO
1392
1200
49.5 x 53.3
18.2
28
500
Ice
Polyethylene
Tygon
Silicone
96.3
790
6.35
Peristaltic
*
X
X

X
i


P
ISCO
14B0
800
48.5 x 64.8
14.1
1
11.350
Ice
Polyethylene
Tygon
Silicone
24.1
790
6.35
Peristaltic
X
X
X

X
X


P
ISCO
1580
900
4B.5 x 64.8
14.1
1
11.350
Ice
Polyethylene
Tygon
Silicone
96.3
790
6.35
Peristaltic
X
X
X
X
X
X


P
Lakeside
T2
1855

25
1
U
Ref
u
Plastic
Plexiglas


12.7
Scoop

X


X



F
Manning
S-4000
1350
43.8 x 57.2
18.1
24
500
Ice
Polyethylene
Tygon
Plexiglas
H
670
9.5
Vacuum
X
X
X
X
X
X


P
Markland
1301
1150
43.2 x 30.5 x 71.1
27.2
1
7570

Polyethylene
Tygon
E.P.T.

914
6.35
Pressure


X
X

X
X

P
Markland
2104T-CLK
1250


1
7570

u
Tygon
E.P.T.

914
6.35
Pressure


X
X
X

X

F"
N-Con
Surveyot
275
Small
L
1
U

u
U
Buna-N
H
182
12.7
Impeller


X

X



P
N-Con
Seoul
520
35.6 x 15.3 x 43.2
10
1
3785

Polypropylen
Tygon
Silicone
12.1
457
6.35
Peristaltic


X

X
X


P





|

















X « HAS. U ¦ USER SUPPLIED, L ¦ LOW. H • HIGH
(continued)

-------
TABLE 2.3 (continued)
MANUFACTURE*
MODEL NO
»< m
o r
0. O
« o
DIMENSIONS
IID.i HT.
or DIA.* HT.
tens]
r _
ei
U4 —
£
SAMPLE
BOTTLES
c
ui?
MATERIALS EXPOSED
TO SAMPLES
si_
to
< E
X"
-T~
Sf
< E
~-
2
TYPE OF
PUMP
1 PURGE
: CYCLE
cc
NTBO
IS
p
OWER


PORTABLE
Of FIXED
No.
Cap.
{mt)
£ o
o
u
Botlfes
Tubing
Other
F low
0 'Op
r,m«
Prop
Solid
State
CJ
4
It
«
c
<£
e
c
l/S
N-Con
Sentry
1100
40.6 x 35.6 x 33
15.9
24
450

Glass
Tygon
Silicone
12.1
457
6.35
Peristaltic
X

X

X
X


p
N-Con
Trebler
1600


1
U
REF.
U

PVC

L

Scoop

X


X



F
N-Con
Sentinel

58.5 * 25.4 * 147.4
B4
1
7570
REF.
Polyethylene

PVC


50.8
U


X

X



F
NP Enterprises
NPE



1

REF.



H


Vacuum
X

X

X



F
Phips 6 Bird
8392-300
850


1
U

u

Stainless

305

Dipper


X

X
X


F
Pro-Tech
CG-125
800
33 ii 25.4 x 43.2
9.1
1
5678

TFE Resins
TFE Resins
PVC

911
3.16
Pressure
X

X



X

P
Pio-Tech
CG-150
300
33 x 25.4 x 43.2
9.1
1
5678

TFE Resins
TFE Resins
PVC

914
3.16
Pressure
X

X



X
X
P
Pro-Tech
CEL-300
1500
33 x 48.3 x 43.2
13.7
1
5678

TFE Resins
PVC
PVC
99 7
914
12.7
Submersible


X

X



P
Pro-Tech
DE1--240S
5700
76.2x81.2x182.9

24
100
REF.
TFE Resins
Stainless
PVC
99.7
914
12.7
Submersible


X

X



F
OCEC
CVE
570
38.1 x 38.1 x 60.9
24.9
1
1893
ICE
Glass
Tygon
Plexiglas
H
610
6.35
Vacuum
X

X

X



P
QCEC
E
1000
20.3 x 33 x VAR,
45.4
1
U



Stainless



Dipper


X

X



F
OCEC
CVE 11
950
38.1 x 43.2 x 38.1
15.9
1
3785
ICE
Glass
Plexiglas
Brass
H
610
12.7
Vacuum
X

X
X
X
X


P
QCEC
IF
960
39.4 x 7.7
10
1
U

U
U
Stainless



Plurtger into
oipel »ne


X

X



r
Sigmamoior
WD-1
650
34.3 x 25.4 x 36.9
14
1
9462

Plastic
Tygon

9.7
670
3.16
Mutating


X

X
X


P
Sigmamotor
m-s
1100
50 x 37 x £4
27
1
18,925

Plastic
Tygon

4.2
548
6.35
Finger


X

X
X


P
Sigmamoior
WW!-4-24
1100
50 x 37 x 64
25.4
24
450

Plastic
Tygon

9.7
670
3.16
Nutating
X

X

X
X


P
Sigmamotor
WM-6-24
1400
50 x 37 x 64
29
24
450

Plastic
Tygon

4.2
548
6.35
Finger
X

X

X
X


P
Sigmamotor
WAP-2
700
34.3 x 25.4 * 36.9
11.4
1
9462

Plastic
Tygon

9.7
670
3.16
Nutating

X


X



P
Sigmamotor
KAP-5
1050
50 x 37 x 64
19.1
1
18.925

Plastic
Tygon

4.2
548
6.35
Finger

X


X



P
Sigmamotoi
WW-1-24R
1525
53.4 x 55.9 x 88.4
56.8
24
450
REF.
Plastic
Tygon

9.7
670
3.16
Nutating
X

X

X



F
Sigmamotor
WAC-5R
1300
53.4 x 55.9 x 125
44.5
1
18.925
REF.
Plastic
Tygon

VAR.
670
3.16
Finger

X


X



F
SIRCO
B/ST-VS
1670-
2S50

127
24
473
REF.
Polyethylene
Plexiglas

H

9.53
Vacuum
K

X

X



F
SIRCO
B/IE-VS
IIOO-
27TB

123
1

REF.
Stainless
PVC
PVC

6096

Dipper


X

X



F
sinco
B/OP-VS
1375 '
2772

91
24

REF.
Polytttaylant
PVC,
Plexiglas



Pressurized
Source


X

X



F
SIRCO
MK-VS
675"
136*
40.7 x 40.7 x 55.3
17
i
24
1S.H0
soo

Plastic
PVC
Plexiglas
140
670
9.53
Vacuum
X

X
X
X
X


P
Soniord
NW-3
1000
39.4 X 39.4 X 68
23.2
24
473

Glass
Tygon
Stainless

396
6.35
Evacuated
bottles


X




X
P
Sonford
HG-4
500
33.B x 31.4 x 33.5

1
3785

Polyethylene

Stainless

53

Telescoping
tube


X

X
X


P
TMI
MARK 3B
845
36.8 x 66
14.5
12
570

Glass
Tygon
Stainless

300
6.35
Evacuated
boltlea


X




X
P '
THt
MARK 4B
950
38 x 38 x 47
20.2
24
570

Glass
Tygon
Stainless

300
6.35
Evacuated
bottles


X


X

X
P
Tri-Aid Sciences
CUSTOM
DESIGN





REF

Silicone


762
9.53
Peristaltic

X
X
X
X



F
Haste Watcher
CS/TP
1425
20 x 20 x 7
10.5
1
U

U
Tygon
Silicone
34
670
7.9
Peristaltic
X
X
X

X



F
X - HAS, U - USER SUPPLIED, L - LOW. H - HIGH

-------
situations. For each application the following variables should be consid-
ered in selecting an automatic sampler (4):
•	Variation of water or wastewater characteristics with time.
« Variation of flow rate with time.
•	Specific gravity of liquid and concentrations of suspended solids
•	Presence of floating materials.
Selection of a unit or a variety of units for sampling should be preceded
by a careful evaluation of such factors as:
•	The range of intended use.
•	The skill level required for installation of the automatic sampler.
•	The level of accuracy desired.
References 5,6,7,8 and 9, have useful information on the theoretical
design considerations and actual field performance data for automatic samplers.
2.3.2.1 Criteria for Evaluating Automatic Sampler Subsystems
There are usually five interrelated subsystems in the design of an auto-
matic sampler. The criteria for selecting subsystems are briefly described
below; more detailed information can be found in references 5,6, and 9.
2.3.2.1.1 Sample Intake Subsystem
The success of an automatic sampler in gathering a representative sample
is dependent upon conditions at a particular sampling site (4) and the
design of the sample intake subsystem. The reliability of a sample intake
subsystem is measured in terms of:
•	Freedom from plugging or clogging.
•	Non-vulnerability to physical damage.
•	Minimum obstruction to flow
•	Capability to draw a representative sample.
14

-------
•	Multiple intakes.
•	Rigid intake tubing or facility to secure or anchor the intake
tubing. Avoidance of sharp bends, twists, or kinks to prevent
clogging of intake line.
2.3.2.1.2	Sample Gathering Subsystem
Three basic sample gathering methods: mechanical, forced flow, and
suction lift are available in commercial samplers. Figures 2.1 and 2.2
illustrate forced flow and suction lift sample gathering subsystems, respec-
tively. Figures 2.3 and 2.4 illustrate a mechanical sample gathering subsys-
tem at weir and flume installations respectively. These subsystems are
compared in Table 2.4.
2.3.2.1.3	Sample Transport System
The majority of the commercially available composite samplers have
fairly small diameter tubing in the sample train. This tubing is vulnerable
to plugging, due to the buildup of fats, etc. Adequate flow rates must be
maintained throughout the sampling train in order to effectively transport
suspended solids.
To optimize sampler performance and reliability, the following features
and procedures are desirable:
•	The minimum size of the sample transport line should be
6 mm (1/4 in.) internal diameter.
•	For most applications, the sample should not contact metals
during transport.
•	The sample line should be transparent and flexible, and made of
an inert material such as Tygon^. If trace quantities are to be
15

-------
TABLE 2.4 COMPARISON OF SAMPLE GATHERING SUBSYSTEMS
Feature
Mechanical
Forced Flow
Suction Lift
Lift
High
High
Limited to 7.6 m(25
feet) or less
Sample integra-
tion over the
entire depth.
Possible
Possible with
pumps but not
with ejection
units.
Possible with
multiple intakes
Obstruction to
flow
Explosion-
proof
Fouling
Significant
Some
Dissolved gasses No problem
Sample Volume
Exposed parts
have a
tendency to
foul
Suitable for
wide range
Flexibility
Maintenance
Limited
Heavy
Less than
mechanical
subsystem
Pneumatic
ejection units
meet this re-
quirement
No problem
Not easily
fouled
Pump suitable
for wide range.
Very little
Some
Not suited but
if used, the
initial flow
should be dis-
carded.
Intake tubing of
less than 6ram(l/4")
I.D. is prone to
fouling
Should be inde-
pendent of
Pneumatic ejection vertical lift
units suitable for
sample volume
Moderate
Moderate but
costly
Maximum
Little
16

-------
PRESSURE OPERATION
Propellant under pressure from a source (A)
is metered by a control valve (B) for rate-
meter (C) into accumulator tank (D). On
reaching a preselected pressure, a pneu-
matic relay (E) releases the accumulated
propellant through inlet line (F) to the sample
intake chamber (G). Pressure »n the chamber
closes its chock valve (H) and propels the
samDle through outlet line (I) and into the
sample bottle (J). Excess propellant vents
through the sample line, thereby purging it of
liquid and incidentally providing protection
against line freezing in cold weather. The
resulting pressure drop rectoses the relay (E)
and the sampling cycle repeats at a repeti-
tion rate determined by adjusting the control
valve (B).
O
=

PRESSURE
OPERATION
Figure 2.1 Schematic of Forced Flow Type Sampler,

-------
FILL SENSOR
INLET HOSE
SAMPLE SIZE
ADJUSTMENT
MEASURING
CHAMBER
COMPRESSOR
]VALVE
STEPPING
MOTOR
ROTARY UNION
SPOUT
DISCRETE
COMPOSITE
SAMPLE BOTTLES
INTAKE
CONTROLS
Figure 2,2 Schematic of Suction Lift Type Sampler
18

-------
Parts List
No.	Description
1.	Motor-Reducer
2.	Drive Sprocket
3.	Driven Sprocket
4.	Roller Chain
5.	Scoop
6.	Scoop Counter Weight(not shown )
7.	Limit Switch
8.	Time Clock
9.	Alum. Sampler Casting
10.	Alum. Sampler Support
11.	Outlet Coupling
12.	Unilet Body & Cover
PLASTIC P'P!
V-NOTCH W6IR
(RECTANGULAR OR ClPOLlgTTI
WSIBSAR€ OPTIONAL!
GALV. ANCHOR
BOLTS OR ACKERMAN
Figure 2.3 Schematic of Mechanical Type Sampler (Weir Installation).
19

-------
Parts List
No.
Description
1.	Motor - Reducer
2.	Drive Sprocket
3.	Driven Sprocket
A.	Roller Chain
5.	Scoop
6.	Scoop Counter Weight(not shown)
7.	Limit Switch
8.	Time Clock
9.	Alum. Sampler Casting
10.	Alum. Sampler Support
11.	Outlet Coupling
12.	Unilet Body & Cover
5s
*tASTie*»P€

•OTTlI
GAUV .ANCHOR
•oi rs ort Adce^MAN,
Figure 2.4 Schematic of Mechanical Type Sampler (Flume Installation).
20

-------
measured, a method of testing for tube contamination is needed.
Avoid sharp bends, twists, or kinks to prevent clogging.
•	The sample line should be purged prior to and immediately after
each sample collection. A clean water purge is effective (4) but
not feasible in most instances. A complete air purge is suffi-
cient for non-permanent or winter operation.
•	The sampler should be capable of lifting a sample a vertical dis-
tance of 6.1 m (20 ft.) (7).
•	The sampler should be capable of maintaining line velocity of
0.6 to 3.0 m/sec. (2-10 ft/sec.) for vertical transport. (7) i
•	The importance of line velocity and isokinetic conditions (intake
velocity same as velocity of flow of water) depends on the con-
centration and density of the non-filterable suspended solids in
the water, the program requirements for accuracy of suspended
solids determinations, and any other parameters affected by sus-
pended solids concentrations. If a program requires maintaining
isokinetic conditions, dial adjustmant of intake velocity is a
desired feature.
2.3.2.1.4 Sample Storage Subsystem
Both discrete samples and composite samples are desirable for certain
applications. Discrete samples are subject to considerably more error intro-
duced through sample handling, but do provide opportunity for manual flow com-
positing and time history characterization of a waste stream during short per-
iod studies. The desired features of sample storage subsystems are:
•	Flexibility of discrete sample collection with provision for
single composite container.
21

-------
•	Minimum discrete sample container volume of 500 mil (0.13 gal.)
and a minimum composite container capacity of 7.57 SL (2.0 gal.).
•	Storage capacity of at least 24 discrete samples.
•	Containers of conventional polyethylene or borosilicate glass
and of wide mouth construction.
•	Capability for cooling samples by refrigeration or a space for
packing ice and maintaining samples at 4° to 6°C (39° to 43°F)
for a period of 24 hours at ambient temperature range between
-30° to 50°C. (-22° to 122°F)
•	Adequate insulation available for the sampler to be used in
either warm or freezing ambient conditions.
2.3.2.1.5 Controls and Power Subsystem
The following are desired power and controls features, many of which will
depend upon whether or not the sampler is to be portable or a permanent instal-
lation:
•	Capability for either AC or DC operation.
•	Battery life for 2 to 3 days of reliable hourly sampling without
recharging.
•	Battery weight of less than 9 kg (20 lb.) and sealed so no leak-
age occurs.
•	Solid state logic and printed circuit boards.
•	Timing and control systems contained in a waterproof compartment
and protected from humidity. Timer should use solid state logic
and a crystal controlled oscillator.
•	Controls to allow both flow-proportional sampling (directly
linked to a flow meter) and periodic sampling at an adjustable
22

-------
interval from 10 minutes to 4 hours.
•	Capability of multiplexing, i.e. drawing more than one sample
into a discrete sample bottle to allow a small composite over a
short interval. Also capability for filling more than one bottle
with the same aliquot for additon of different preservatives.
•	Capability of adjusting sample size and ease of doing so.
2.3.2.1.6 General Desirable Features
From the view of safety, maintenance, reliability and security in field
applications, the following general features are desired in an automatic
sampler:
•	Water tight casing to withstand total immersion and high humidity.
•	Vandal-proof casing with provisions for locking.
•	A secure harness or mounting device if sampler is placed in a
sewer.
•	Explosion-proof construction.
•	Sized to fit in a standard manhole without disassembly.
•	Compact and portable for one-man installation.
•	Overall construction, including casing, of materials resistant
to corrosion (plastics, fiberglass, stainless steel).
•	Sampler exterior surface painted a light color to reflect sun-
light.
•	Low cost, availability of spare parts, warranty, ease of main-
tenance, reliability and ruggedness of construction.
2.3.2.2 Installation and Use
2.3.2.2.1 General Consideration
Well designed equipment will yield good results only when properly
23

-------
installed and maintained. A few general guidelines follow:
•	When a sampler is installed in a manhole, secure it
either in the manhole (e.g. to a rung) above the high
water line or outside of the manhole (e.g. to an above
ground stake by means of a rope) .
•	Place the intake tubing vertically or at such a slope
to ensure gravity drainage of the tubing between
samples, avoiding loops or dips in the line.
•	Clean sample bottles, tubing and any portion of the
sampler which contacts the sample between setups.
Whatever methods of cleaning are used, all parts of
the sampler which come in contact with the sample
should be rinsed with tap water and then given a final
rinse with distilled water. A distilled water rinse
may not be necessary between setups on the same waste
stream.
•	Inspect the intake after each setup and clean, if necessary.
Exercise care when placing the intake(s) in a stream containing
suspended solids and run the first part of the sample to waste.
The velocity of flow should at all times be sufficient to pre-
vent deposition of solids. When a single intake is to be used
in a channel, place it at six-tenths depth (point of average
velocity) (10,11). For wide or deep channels where stratifica-
tion exists, set up a sampling grid.
•	Maintain electrical and mechanical parts according to the manu-
facturer's instructions. Replace the desiccant as needed. If
24

-------
a wet-cell lead-acid battery is used, neutralize and clean up
any spilled acid.
•	Position the intake in the stream facing upstream. In any case
limit the orientation of the intake with ±20 degrees on either
side of the head-on. Secure the intake by a rope at all times
with no drag placed on the inlet tubing.
•	After the installation is complete, collect a trial sample to
assure proper operation and sample collection. The sampler
must give replicate samples of equal volume throughout the
flow range. If the sampler imposes a reduced pressure on a
waste stream containing suspended solids, run the first part
of the sample to waste.
2.3.2.2.2 Winter Operation
For outdoor use in freezing temperatures special precautions should be
used to insure reliable sample collection and to prevent the collected
sample(s) from freezingt
•	Place the sampler below the freezing level or in an insulated
box.
•	When AC is available, use a light bulb or heating tape to
warm sampler. When installation below the freezing level is
not possible and line current is available wrap 1.2 to 1.8 m
(4 to 6 ft.) heat tapes [thermostatically protected 3°C (38°F)]
around the sample bottle and the intake lines. Loosely wrap
a large plastic bag (airline trash bags, 10 mil, GSA
#8105-808-9631) over the heat tape on the intake lines.
Place a large plastic bag over the sampler as loosely
25-

-------
as possible. (7)
•	Be certain to place the line vertically or at such a
slope to ensure gravity drainage back to the source.
Even with a back-purge system some liquid will
remain in the line unless gravity drainage is provided.
If an excess length of tubing exists, cut it off.
Keep all lines as short as possible.
•	Do not use catalytic burners to prevent freezing since
vapors can affect sample composition. When power is
unavailable, a well insulated box containing the
sampler, a battery and small light bulb are effective
in preventing freezing.
2.3.2.3 Selection of an Automatic Sampler
To choose an automatic sampler, list the desired features needed for
a particular sampling program and select the sampler that best fits the
requirements consistent with the sampling objectives.
The
following is a check list for selecting an automatic sampler:
1.
Vertical lift
2.
Submergence
3.
Explosion proof
4.
Intake tube: diameter/material
5.
Dissolved gases
6.
Suspended solids
7.
Oils and grease and floating material
8.
Organic priority pollutants
9.
Isokinetic sampling
26

-------
10.	Sample type: continuous, composite: time proportional, flow
proportional, etc.
11.	Multiple intakes
12.	Multiplexing
13.	Dependability
14.	Ease of operation
15.	Maintenance
16.	Availability
2.4 TYPE OF SAMPLE
Selection of the type of sample to be collected depends upon a number
of factors such as the variability of flow, variability of water or waste-
water quality, the accuracy required and the availability of funds for con-
ducting the sampling and analytical programs. All samples collected, either
manually or with automatic equipment, are either grab or composite samples.
2.4.1 Grab Samples
A grab sample is defined as an individual sample collected over a
period of time not exceeding 15 minutes. It can be taken manually, using
a pump, scoop, vacuum, or other suitable device. The collection of a grab
sample is appropriate when it is desired to:
1.	Characterize water quality at a particular time.
2.	Provide information about minimum and maximum concentrations.
3.	Allow collection of variable sample volume.
4.	Corroborate composite samples.
27

-------
2.4.2	Composite Samples
A composite sample is defined as a sample formed by mixing discrete
samples taken at periodic points in time or a continuous proportion of the
flow. The number of discrete samples which make up the composite depends
upon the variability of pollutant concentration and flow. A sequential
composite is defined as a series of short period grab samples each of which
is held in an individual container, then composited to cover a longer time
period. Six methods are used for compositing samples. Table 2.5 lists those
methods with their advantages and disadvantages. Choice of composite type
is dependent on the program and relative advantages and disadvantages of
each composite type.
2.4.3	Selection of Sample Type
Use grab samples when: (12,13,14)
1.	The stream does not flow continuously, e.g. batch dumps.
2.	The water or waste characteristics are relatively
constant.
3.	The parameters to be analyzed are likely to change
(i.e. dissolved gases, residual chlorine, soluble
sulfide, oil and grease, microbiological parameters, organics,
etc.).
4.	Information on maximum, minimum or variability is desired.
5.	The history of water quality is to be established based on
relatively short time intervals.
6.	The spatial parameter variability is to be determined e.g.
the parameter variability throughout the cross-section
and/or depth of a stream or large body of water.
28

-------
TABLE 2.5 COMPOSITING METHODS
Sample
mode
Compositing
priciple
Advantages
Disadvantages
Comments
1. Continuous
Constant
pumping rate
Minimal manual
effort, requires
no flow measure-
ment-
Requires large
sample capacity;
may lack represen-
tativeness for
highly variable
flows»
Practical but not
widely used.
2. Continuous
3. Periodic
Sample pumping
rate proportional
to stream flow-
Constant sample
volume, constant
time interval
between samples.
Most representa-
tive especially
for highly
variable flows;
minimal manual
effort.
Minimal
instrumentation
and manual effort;
requires no flow
measurement.
Requires accurate
flow measurement
equipment, large
sample volume,
variable pumping
capacity, and
power.
May lack
representativeness
especially for
highly variable
flows
Not widely used.
Widely used in both
automatic samplers
and manual sampling.
4. Periodic
Constant sample
volume, time
interval between
samples propor-
tional to stream
flow.
Minimum manual
effort.
Requires accurate
flow measurement/
reading equipment
Manual compositing
from flow chart.
Widely used in
automatic as well
as manual sampling,
(continued)

-------
TABLE 2.5 (continued)

Sample
Compositing
Advantages
Disadvantages Comments
mode
principle


5. Periodic
Constant time
Minimal
Manual compositing Not widely used in

interval between
instrumentation
from flow chart. automatic samplers

samples, sample

In abscence of but may be done

volume propor-

prior information manually.

tional to total

on the ratio of

stream flow

minimum to maximum

since last sample.

flow, there is a



chance of collecting



either too small or



too large individual



discrete samples for



a given composite



volume.
6. Periodic
Constant time
Minimal
Manual compositing Used in automatic

interval between
instrumentation.
from flow chart. samplers and widely

samples, sample

In abscence of used as manual

volume propor-

prior information method .

tional to total

on the ratio of

stream flow

minimum to maximum

at time of

flow, there is a

sampling„

chance of collecting



either too small or



too large individual



discrete samples for
a given composite
volume.

-------
Use composite samples when :
1.	Determining average concentrations.
2.	Calculating mass/unit time loading.
2.4.4	Method of Manual Compositing
When using a constant volume/ time proportional compositing method,
previous flow records should be used to determine an appropriate flow volume
increment so a representative sample is obtained without over-running the
bottle capacity or supply.
The preparation of the flow rated composite is performed in various
ways. Table 2.6 summarizes the techniques necessary for preparing composites
from time constant/variable volume samples.
2.4.5	Examples of Manual Compositing
Example 2.1 illustrates the method of manual compositing for time
constant/volume proportional to discharge since last sample, when records of
totalized flow are available.
Example 2.1A illustrates the method of manual compositing for time
constant/volume proportional to discharge since last sample, when records of
flow rates are available.
Example 2.2 illustrates the method of manual compositing for time
constant/volume proportional to instantaneous flow rate.
Example 2.3 illustrates the method of manual compositing for the
constant volume/time proportional to equal increment discharge passing the
sampling point, based on the past records of totalized flow.
Example 2.3A Illustrates the method of manual compositing for the
constant volume/time proportional to equal increment discharge passing the
sampling point, based on the past records of flow rates.
31

-------
TABLE 2.6 MANUAL PREPARATION OF VARIABLE VOLUME COMPOSITE
Type
Preparation
Equation
Time constant/propor-
tional
Determine volume since
last sample by
integration
- *Qh
iaq.
v.
¦	aliquot volume to be
extracted from ith
discrete sample
V£ « composite volume (known)
qi » flow rate when ith
discrete sample was
taken (from flow record)
¦	flow volume when ith
discrete sample was
taken
Q » flow volume when ith-1
discrete sample was
taken
• flow volume or rate
since last sample
(integration)
EAQj. ® total flow volume,
(estimated)
Time constant/volume
proportional to
instantaneous flow
rate
Note flow rate at each a^
time of discrete sample
collection
n
r*i
i"»i
Where, a^ = aliquot volume to be*
extracted from ith
discrete sample
*1
flow rate when ith
discrete sample was
taken (from flow record)
composite sample volume
desired
number of discrete samples
32

-------
Example 2.1: Manually Preparing a Composite Sample for:
Time Constant/Volume Proportional to Discharge Since Last Sample.
Given: A 500 discrete sample was taken at the end of each hour over
an 8 hour shift. A 3000 mS. composite is desired, A recording of
totalized flow is available.
Sample No.
CD
0
1
2
3
4
5
6
7
8
AQi
(liters)
0
858
3,462
8,254
12,347
17,950
21,225
24,600
25,750
(liters)
858
2,604
4,792
4,093
5,603
3,275
3,375
1,150
IA Q. = 25,750
Max a. = 653 m£
(m2.)
100
303
558
477
653
382
393
134
E3i = 3,000
a. (adjusted) =
a. (500/max a.)
l	i'
(mJ,)
77
232
427
365
500
292
301
103
2,297
Steps:
Enter Q. from record and calculate AO, = (X - Q. ,
1 i i-1
Calculate a^ « ^c (AQ^) , where Vc = 3000 mi.
ZAQi
3.	Check to see if maximum a^ exceeds discrete sample volume, i.e.
653 m£ > 500 mJl.
4.	If it does, adjust aliquot sizes using the relationship:
a. (adjusted)
'discrete sample volume'
max a 4
500
653
= 0-77
5. Determine the adjusted composite volume from (adjusted). This
example illustrates that although desired composite volume was
3,000 mil because of discrete sample volume size, only 2,297 mfc
of composite sample can be obtained.
33

-------
Example 2.1A: Manually Preparing a Composite Sample for;
Time Constant/Volume Proportional to Discharge Since Last Sample'
Given: A 500 m£ discrete sample was taken at the end of each hour over
an 8 hour shift. A 3,000 raS, composite is desired. A recording
of flow rate is available.
(adjusted) =
Sample No. 9 ^ AQ^ a^ a^ (500/max
(liters) (liters) (mil) (m£)
0 961
1 2,025 1,483 146 132
2 3,700 2,862 282 255
3 5,212 4,456 439 397
4 6,'004 5,608 553 500
5 5,018 5,511 543 491
6 4,002 4,510 444 401
7 3,089 3,546 349 316
8 1,847 2,468 244 221
IAQi=30,444 Iai=3,000 2,713
Max ai = 553 mil
Steps:
1. Enter q. from record and use trapezoidal rule to calculate
1
AQi - (
-------
5. Determine the adjusted composite volume from a^ (adjusted).
This example illustrates that although desired composite vol-
ume was 3,000 m£. (V ) because of discrete sample volume size,
only 2,713 m£ of composite samples can be obtained.
35
-------
Example 2.2: Manually Preparing a Composite Sample For:
Time Constant/Volume Proportional to Instantaneous Flow Rate.
Given: 500 discrete samples were taken at hourly intervals over
an 8 hour shift, A 2000 mX, composite is desired. A recording
of flow rate is available.
Sample No.
(i)
1
2
3
4
5
6
7
8
qi
(liters)
600
1,000
1,700
2,800
1,800
1,400
1,000
700
= 11,000
a.
x
(m£)
109
182
309
509
327
255
182
127
Za. = 2,000
x
(adjusted) =
a.^ x 500/ max ai
(m£)
107
179
304
500
321
250
279
125
1,965
max = 509 mJl
Steps:
1. Enter q^ from record and sum.
2. Calculate a^ = Vc/q^.
3. Check to see if maximum a^ exceeds discrete sample volume.
4. Adjusted aliquot volume = a. (500/509)= a. (adjusted). This
X
example illustrates that with an individual discrete sample capa-
city of 500 m£ only 1,965 mi volume of composite sample can be
obtained. If it is desired to collect a composite sample of 3,000
mz volume, obviously larger sized (750 tn£) capacity bottles or
greater sampling frequency will be required for collecting individ-
ual discrete samples.
36
-------
Example 2.3; Manually Preparing a Composite Sample For:
A Constant Volume/Time Proportional to Equal Increment Discharge.
Given: A 500 discrete sample was taken each time an average hourly
flow flowed past the sample point. Sampling period is 8 hours.
In addition, a 500 m£ sample was taken at the end of the sampling
period; A composite of 4,'OOOmJt is desired. A recording of total-
ized flow (from past record) is available.
Past Record
Period ^i(past) ^Qi (past)
ith hour (liters) (liters)
Actual
^i (actual)
(liters)
^i (actual)
(liters)
i Sample
(mil) No.
0
0
0
0
0

1
868
868
797
797

500
2
4,024
3,156
3,648
2,851

500
3
7,616
3,592
8,002
4,354

500
4
11,453
3,837
11,709
3,707
500

500
5
16,629
5,176
16,056
4,347

500
6
20,377
3,748
19,763
3,707

500
7
22,625
2,248
24,321
4,558

500
8
25,000
2,375
26,650
2.229
264

EAQ.
1
(past) " 25
,000
EAQ^ (actual) = 26
,650
Steps:

1
. Enter
from past
record and calculate AC^ = -
Qi-r
1
2
3
4
5
6
7
8
9
2. Determine the number of samples for the overall sampling period.
On the basis of the number of samples required for the overall
sampling period, P, determine the average flow from the past re-
cords for the time interval,T, between the successive discrete
samples. ^7
-------
In our case, the number of samples for the sampling period = 8
Overall sampling period, P = 8 hours.
_, , _ 8 hrs , ,
Time interval, T = —r—. = 1 hour
Average flow for the time interval between successive
samples from past = (past) = 25,000 = 3,125 liters.
p 8
3. Aliquot size a^ = 500 m£.
4. Collect each discrete sample every time 3,125 liters passes the samp-
ling point; and an additional one 500 sample aliquot at the end of
the sampling period.
5. Record the actual flow.
6. Note the total flow for the sampling period. In our case it is IAQ_^
(actual) = 26,650 liters.
7. Calculate the difference between IAQ (actual) and EAQ. (past) which
i i
is 26,650 - 25,000 = 1,650 liters. This is the flow which passes the
sampling point after taking the last sample for equal incremental dis-
charge, up to the end of sampling. This flow is sampled by the sample
taken at the end of the sampling period.
8. Compute the representative aliquot required for the unbalanced flow
in step 7 in proportion to the equal increment flow.
Required aliquot volume ~ £AQ^ (actual) - IAQ^ (past)
equal increment discharge volume
- 2Mi2mrAimof 4,264 m£.
38
-------
Example 2.3A Manually Preparing a Composite Sample For:
A Constant Volume/ Time Proportional to Equal Increment Discharge.
Given: A 500 mi, discrete sample was taken each time an average hourly
flow flowed pas-t the sample point. Sampling period is 8 hours.
In addition, a 500 mil sample was taken at the end of the sampling
period. A composite of 4,000 m£ is desired. A recording of in-
stantaneous flow..rate (from past records)is available.
Period
ith hour
q^Cpast)
(liters)
AQi(past)
(liters)
q.(actual)
tliters)
AQ.(actual)
tliters)
ai
(m£)
Sai
N
0
40
-
30
-

1
60

80
500
1
2
100

110

110
500
2
3
120

110

140

130
500
3
4.
160

150

500
4

160

165
500
5
5.
160

180

500
6

155

180

6.
150

180

500
7

130

145

7.
110

110

500
8

105

100

8.
100

86
9

£AQ^(past)
- 930

ZAQi(actual) = 950

Steps:
1. Enter q^ from past record and use trapezoidal rule to calculate AQ^
=(q^ + q£_2)/2 (another integration scheme could be used if warranted.
2. Determine the number of samples for the overall sampling period. On
the basis of number of samples required for the overall sampling period
39
-------
P, determine the average flow from the past records for the time inter-
val T, between the successive discrete samples.
In our case the number of samples for the sampling period = 8.
Overall sampling period, P = 8 hours.
Time interval, T 8 hours 1 hour.
8
Average flow for the time interval between successive samples from past
Eqt
records 930 116 liters.
= p = 8 =
3. Aliquot Size a^ = 500 m£.
4. Collect each discrete sample every time 116 liters passes the sampling
point and one additional aliquot of 500 at the end of the sampling
period.
5. Record the actual flows per unit of time interval selected, e.g. hours,
minutes, days.
6. Calculate the total actual flow for the sampling period. In our case it
is ZAQ^ (actual) = 950 liters.
7. Calculate the difference between £AQ. (actual) amd EAQ (past) which is
i
950-930 = 20 liters. This is the flow which passes the sampling point
after taking the last sample for equal incremental discharge, up to the
end of the sampling period.
8. Compute the representative aliquot required for the unbalanced flow de-
termined in step 7 in proportion to the equal increments.
Required aliquot volume TAQj(actual) - XAQ.(past)
44 = (a ) = (20g,) (500 mi)
equal increment discharge 1162-
volume
= 86 m&
40
-------
9. Composite volume = la^
= 8 aliquots of 500 rail + 86 miL from the aliquot
taken at the end of the sampling period.
- 4,086 m£.
2.5 PLANNING A SAMPLING PROGRAM
To achieve desired goals and performance, it is imperative that adequate
consideration is given to both the planning and execution of a sampling pro-
gram. While no ready made sampling program can be formulated which is appli-
cable to all situations, the following considerations are provided to help
plan an appropriate sampling program. The overall planning process can be
divided into four stages:
1. Preliminary Plan
; . Evaluation of Preliminary Plan
3. Final Plan
4. Program Evaluation
2.5. 1 Preliminary Plan
In this stage emphasis is upon collection of preliminary information on
the entity to be sampled, the sampling sites and the flow characteristics.
This information may be available from records of previous surveys. Where
such information is not available reconnaissance should be carried out to be-
come thoroughly familiar with actual site conditions. Table 2.7 shows the type
of information needed in most cases. The appropriate information for Table 2.7
should be collected with a minimum of effort and cost. Based on this informa-
tion, a preliminary sampling plan is drawn up. Preliminary sampling objectives
should be delineated and then details of plan such as anticipated parameters,
sample type, sample size, frequency, etc. specified. This information should
41
-------
TABLE 2.7 PRESURVEY INFORMATION
Entity:
Process details
Treatment Plant
C
)
l.
Industry
c
)
2.
River
(
)
3.
Estuary
(
)
4.
Sewer
(
)
5.
Water mains
/
V
\
/
6.
Plans:
Yes
No
Waste sources
flows
sewer maps
( )
( )
1.
P/C
water line network maps
C )
( )
2.
P/C
river and tributary maps
C )
( )
3.
P/C
treatment plant map3
( )
( )
4.
P/C
estuary zone maps
C )
C )
5.
P/C
Channel
Width
Depth
Pipe
Diameter_
Material
Flow Variability
Hourly max
Hourly min
Hourly average_
Daily max
Daily min
Daily average_
Manholes ( )
other
Diameter
or width_
Depth
P = pipe flow C =» open channel flow
(continued)
42
-------
TABLE 2.7 (Continued)
Topography
Level ( )
Slopes ( )
Vegetation ( )
Swamp
( )
Other ( )
Specify
Physical Charac—
teristics of Flow
Odor
Temperature
Oil and grease ( )
Clear
Turbid
( )
( )
Suspended solids
concentration
Safety
Steep banks ( )
Soft grounds ( )
Gases ( )
Specifiy
Stream Currents
Tubulent ( )
Sluggish ( )
Security
Fence ( )
Open ( )
Guarded ( )
Lighted ( )
Other ( )
Specify
Sampling Sites
Distance:
Near ( )
Remote ( )
Numbers:
Few ( )
Many ( )
Accessibility:
Road ( )
Bridge ( )
Other ( )
Specifiy
Convenience:
Sheltered ( )
Power available( )
Other ( )
Specify
Additional Information
43
-------
be recorded in a tabular form similar to Table 2.8.
An estimate of the resources (manpower and equipment) needed for the sampling
program should be made. Table 2,9 illustrates one form for keeping records of
available resources and estimated needs of a sampling program, A preliminary
sampling plan should include sample preservation and chain of custody proce-
dures .
2.5.2 Evaluation of Preliminary Plan
Circulate the preliminary sampling plan among other divisions (laboratory,
field personnel, quality assurance branch, etc.) connected with the sampling
program for their considerations and further deliberations before drawing up
a final sampling program.
2.5.3 Final Plan
The final sampling plan is based on the preliminary plan and subsequent
deliberations and coordination with the various personnel involved. The final
plan should spell out in detail and with clarity the various aspects of samp-
ling such as: objectives, sampling locations, number and frequency of samples,
sample types, preservation and chain of custody procedures, designation of
authorities, field procedures and other pertinent information so that the
sampling plan is executed in an efficient and well coordinated manner. Pre-
sampling briefing should be a key element in any sampling program.
2.5.4 Program Evaluation
The entire program should be evaluated after the samples are collected
and analyzed. This evaluation iis to determine the effectiveness of the final
plan and should serve to avoid future pitfalls and problems. The performance
evaluation should act as a tool to enhance the efficiency of the program and
quality df the data generated from a sampling program.
44
-------
TABLE 2.8 DETAILS OF SAMPLING

Paraaeters of
Interest
Saaple
Type
S&sple
Frequency
Number
of
Saoples
Field or
Lab
Analysis
Sasple
Volusse
Preservation
Hclding
Tirjes
Analytical
>fechods
Chain of
Custody
Procedure
Remarks
45

SHARP EDGED
ROUNDED
SHORT TUBE
BORDA

L
L
*

—*

fr
r

C
0.61 to
0.71
0.98
0.80
0.51
Figure 3.5 Coefficients of Several Types of Orifices (13)
100
THIN PLATE ORIFICE
90
AS ME
NOZZLE
80
60
CO U-l
to u-i
(U -H
>-1 Q
0)
O v
U M
-------
3.1.5 Elbow Meters
Flow acceleration induced in a fluid going around a bend (such as an
elbow) produces a differential pressure that can be used to indicate flow.
The pressure on the outside of an elbow is greater than m the inside, and
the pressure taps located midway around the bend (i.e, 45 degrees from
either flange) can be connected to a suitable secondary element for indicating
or recording.
For accurate flow measurement.straight pipe runs of at least 20 pipe
diameters should be provided both upstream and downstream of the elbow.
Accuracies of 3 to 10% are generally encountered although accuracies of 1 to
2% or better in some cases may be achieved if calibrated in place (5).
3.1.6 Pitot Tube
A schematic diagram of a simple pitot tube is shown in Figure 3.7. In
operation, the velocity of the flow is calculated from the difference in head
measured on the manometer. Pitot tubes measure the flow velocity at a point.
The basic formula is:
Vx = c/2gT
Vx = velocity at a point
C = coefficient of discharge obtained by calibration
Vc = velocity at the center
Vm = mean velocity = 0.83 Vc
H = measured pressure differential
Q = discharge volume
A = area of cross section of stream at the point of
measurement
Q = VmA
65
-------
m) n >) >))) > n n) un)n i) n n > ) n) i >) n n ) i t t >i))) > /1 rm
v,
\jl
T__
))))))))))* \)})> )})))))))
TT2

(r
>})))>)>))!))>?
P2"P1 = Pd
11 rt f f f > )T.

Figure 3.7 Pitot Tube Schematic
Annular
Orifice
If
.Float
Figure 3.8 Rotameter
66
-------
commercially available pitot tubes consist of a combined peizotneter and total
head meter. Pitot tube measurements should be made in a straight section up-
stream and free of valves, tees, elbows, and other fittings with a minimum dis-
tance of 15 to 50 times the pipe diameter. When a straight section is not
possible, a velocity profile should be obtained experimentally to deter-
mine the point of mean velocity. Pitot tubes are not practical for use with
liquids with large amounts of suspended solids because of the possibility of
plugging. In large pipes, the pitot tube is one of the most economical means
of measuring flows.
3.1.7 Rotameters
Rotameters (Figure 3.8) are tapered tubes in which the fluid flows
vertically upward. A metal float in the tube comes to equilibrium at a point
where the annular flow area is such that the velocity increase has produced
the necessary pressure difference. Rotameters are simple, inexpensive and
accurate devices for measuring relatively small rates of flow of clear, clean
liquids. For this reason they are often used to measure the water rate into
individual processing steps in manufacturing operations. To maintain accuracy
in a rotameter, it is absolutely essential that both the tube and float be
kept clean.
3.1.8 Electromagnetic Flowmeter
The electromagnetic flowmeter operates according to Faraday's Law of
Induction : the voltage induced by a conductor moving at right angles through
a magnetic field will be proportional to the velocity of the conductor through
the field. In the electromagnetic flowmeter, the conductor is the liquid
stream to be measured and the field is produced by a set of electromagnetic
coils. A typical electromagnetic flowmeter is shown in Figure 3.9. The
67
-------
induced voltage is subsequently transmitted to a converter for signal con-
ditioning.
Electromagnetic flowmeters have many advantages: accuracies of ± 1 per-
cent are achievable, a wide flow measurement range, a negligible pressure loss,
no moving parts, and rapid response time. However, they are expensive. Build-
up of grease deposits or pitting by abrasive wastewaters can cause error.
Regular checking and cleaning of the electrodes are necessary.
3.1.9 Acoustic Flowmeters
Acoustic flowmeters ,commonly used in water and wastewater flow measure-
ments, operate on the basis of travel time difference method. In the travel
time difference method, sound waves are transmitted diagonally across the pipe
or channel in opposite directions relative to the flow and the difference in
travel times upstream and downstream are measured (Figure 3.10)
Flowmeters must be installed according to manufacturer's instructions and
calibrated in place to eliminate errors due to uncertainties in nonlaminar
flow profile, error due to accoustic short circuit (where transducers are
mounted externally on the pipe) , and errors due to mechanical effects or to
variations in temperature, pressure or composition of water or wastewater.
According to the manufacturers, an accuracy of one percent of full scale is
achievable.(2,5)
3.2 FLOW FROM PIPES DISCHARGING TO THE ATMOSPHERE
The common techniques for measuring the flow from open ended pipes either
full or partly full are listed below. The orifice and flow nozzle techniques
which are not listed here are described in Sections 3.1.3 and 3.1.4 respec-
tively. Rotating element meters are described in Section 3.3.1.1.
3.2.1 Pipes Flowing Full
1. Vertical open end pipe (7)
68
-------
MAGNETIC
COIL
II
MAGNETIC
COIL
X
60 CYCLE A-C
Figure 3.9 Electromagnetic Flowmeter
/
TRANSDUCER A
'//SSSSSSSS^i'/SSSSSSSJ'SSJ'SS/SSSSSSSSS/SSSSJ'SSSSSffSSSSSSSS/-

/
e

•^v
V///SSS/S*SSSSSSSSSSSSSSSS/SSSSSS/S/SSfSrssrS*?<7fS**M*rr
TRANSDUCER B
Travel Time:
tAB = L/(c+Vcos0)
L^g » L/(c-Vcos0)
Figure 3.10 Principle of Acoustic Flowmeter (6)
69
-------
a. Weir flow: Q = o.249D 1*20^1*24 (Figure 3.11a)
? 025 0 53
b. Jet flow: Q = 0.171D H ' (Figure 3 .lib)
where, Q = flow m^/s(cfs = 35.34 m^/s)
D = internal pipe diameter, meters
H = distance from pipe outlet to top of crest,meters.
2. Horizontal or sloped open end
Q = 2.264 x 10~^ AX (Figure 3.11.C and e)
fy
where, Q = flow, m"^/s (cfs = 35.34 m"Vs)
2
A = cross sectional area of the pipe, m
X = distance from the end of the pipe to where Y is
measured, m.
Y = vertical distance measured at a distance X from
the pipe end, m.
3. Purdue Method (6)
It is similar to the trajectory method for the horizontal open ended
pipe, described in (2) above. To obtain the flow, the trajectory measure-
ments X and Y, Figure 3 .12 are used in conjunction with curves derived from
Purdue University experiments, on pipes 0.05 to 0.15m (2—6 inches) in diameter.
Figure 3.13 gives discharge data for Purdue trajectory method for X = 0, 6,
12, and 18 inches and inside pipe diameters of 2,4, and 6 inches.
3.2.2 Pipes Flowing Partially Full
1. Horizontal or sloped open end (7)
2.264 x 10-4 M (CF)
Q = (Figure 3. lid)
/ v
70
-------
H
It

\
a. Weir Flow b. Jet Flow
Vertical Open-End Pipe
D
_L

NS v\

c.
Horizontal Pipe Flowing
Full
\XVv°x\u\
\V>v\
XVVxV>\
\\^VXxHv
s \\\\
d. Horizontal Pipe
Partially Full
/A

e. Discharge From Open-End Pipe -y'.
H,
—at least 6d—
Open End
Hose * ¦ *'
6d
f. California Pipe Method
Figure 3.11 Techniques for Pipes Discharging to the Atmosphere (7)
-------
where,
Q = flow, m^/s (cfs = 35.34 m~Vs)
2
A = cross-sectional area of the pipe, m
X = distance from end of pipe to where Y is
measured, m
Y = vertical distance measured at a distance
X from the pipe end, m
CF = correction factors which are given in Table 3.3
2. Purdue Method (6)
This method can he used for partially full pipe discharging to
atmosphere using the curves (Figure 3.12) for X = 0, provided
the brink depth is less than 0.8 diameter.
3. California Pipe Method (6,7)
Q = TW,(Figure 3.1 If)
whe re,
Q = flow, m^/s (cfs = 35.34 m^/s)
T = 8.69 (1 - -J)1*88
W = d2'48
d = pipe diameter, m
a = distance from top of pipe to flow, m
The empirical equation is derived from experiments performed
on steel pipes from 3 to 10 inches in diameter and it is
imperative that a/d should be less than 0.5 and the straight
pipe length to the end of pipe should be at least 6d.
72
-------
TABLE 3.3 CORRECTION FACTORS FOR DISCHARGE
FROM PIPES PARTLY FULL (7)

Correction

Correction
R*
Factor
R*
Factor
R*
Factor
0.10
0.948
0.37
0.664
0.64
0.324
0.11
0.939
0.38
0.651
0.65
0.312
0.12
0.931
0.39
0.639
0.66
0.300
0.13
0.922
0.40
0.627
0.67
0.288
0.14
0.914
0.41
0.614
0.68
0.276
0.15
0.905
0.42
0.602
0.69
0.265
0.16
0.896
0.43
0.589
0.70
0.253
0.17
0.886
0.44
0.577
0.71
0.241
0.18
0.877
0.45
0.564
0.72
0.230
0.19
0.867
0.46
0.551
0.73
0.218
0.20
0.858
0.47
0.538
0.74
0.207
0.21
0.847
0.48
0.526
0.75
0.195
0.22
0.837
0.49
0.513
0.76
0.184
0.23
0.826
0.50
0.500
0.77
0.174
0.24
0.816
0.51
0.487
0.78
0.163
0.25
0.805
0.52
0.474
0.79
0.153
0.26
0.793
0.53
0.464
0.80
0.142
0.27
0.782
0.54
0.449
0.81
0.133
0.28
0.770
0.55
0.436
0.82
0.123
0.29
0.759
0.56
0.423
0.83
0.114
0.30
0.747
0.57
0.411
0.84
0.104
0.31
0.735
0.58
0.398
0.85
0.095
0.32
0.723
0.59
0.386
0.86
0.086
0.33
0.712
0.60
0.373
0.87
0.078
0.34
0.700
0.61
0.361
0.88
0.069
0.35
0.688
0.62
0.349
0.89
0.061
0.36
0.676
0.63
0.336
0.90
0.052
* R - F/D (Free board in pipe/inside pipe diameter)
73
-------
Figure 3.12 Trajectory Measurements, Purdue Method
. i Mr
i
i i

1
\ l. "V 1
\ \
i\» «v
¦ " 1 " -
-

\ \
V
\ \
\ \
\ \
\ \
*
-

\
\
\
\
\ \
\ \
\ X

\
\
\

Inside
iter " 2 Iff
\
\
V, \
4
\ 6 A
v
\
\
\
X » 18 Inch*#
... i mi! i
1 1
1 11 .1.
V
' •
\
\
\
V1111
\ \
\ %
\ *
' 1 _
5 10 20 50 100 200 500 1000 2000
. \ 1 llj 1
— —X " 6 iachag
1 1
" T 1 1 1
1
i <
1 (M
t
\
\

V
\

*
X - £
Inches
\
N
\

\ \
NA
\
\
<

-
-

"*—¦—\
\\

v\
\ \
\ X
\
\
>
\
-
m
1 t t ^
Inside diameter • 2 la«
. L i i
\ \
\ X
- \ N
\
1 1 "l 1
k 4 In*
vr
\ \
\ \
\ \
\ *
. \ ,
\
\
6 in\
\
I M i
»
5 10 20 50 100 200 500 1000 2000
Flow, gallons per minute
Figure 3.13 Discharge Data, Purdue Method
74
-------
3,3 OPEN CHANNEL FLOW MEASUREMENTS
Methods of flow measurements for open channels can be applied to flows
in non-pressure sewers since both have the same hydraulic characteristics.
Different methods in use can be grouped into the following broad classification:
1. Velocity methods
2. Head-Discharge Methods
3. Miscellaneous Techniques
3.3.1 Velocity Methods
Velocity of flow can be measured using different devices; namely, various
drag body current meters, eddy-shredding current meter, acoustic velocity meter,
doppler-shift velocity meter, electromagnetic current meter, rotating element
current meters. Various drag body current meters are compared in Table 3.4.
Pitot tubes are described in Section 3.1.6.
3.3.1.1 Rotating Element Current Meters
Of the rotating element current meters, Price and Pigmy meters are quite
commonly used. The principle of operation is based on the proportionality
between the velocity of water and resulting angular velocity of the meter
rotor. In conventional current meters there is a wheel which rotates when
immersed in flowing water and a device which determines the number of revo-
lutions of the wheel. The general relation between the velocity of the water
and number of revolutions of the wheel is given by: (1,2,4,5,6,16):
V = a+bN, where
V = velocity of water meters per second
a and b are constants
N » no. of revolutions per second
75
-------
TABLE 3 ,4 COMPARISON OF DRAGBODY CURRENT METERS (16)

Draghody Current Meter Typi
e

Factor
Vertical Axis
Deflection Vane
Horizontal Axis
Pendulum Type
Deflection Vane
Pendulum
Current
Meter
Inclinometer
Drag Sphere
Velocity Range
Wide range but
not suitable for
low velocities
Wide range
Wide range
Suitable for low
velocit ies-only
single velocity
range
Single velocity
range
Submerged
Installation
No
Possible
Pendulum ball
is submerged
Possible
Possible
Debris
Problem
Not a
Problem
Affects drag Not a problem
on line and
hence accuracy
of velocity
measurement
Not a problem
Output Recording
Mechanical
output
Electrical
output
No, manual
operation
No, data manually
processed
Electrical
output
Readout of
Deflection
Visual
No visual
readout

"""
Deflection can
be resolved-no
visual readout
Simplicity
Simple
Simple
Complex

-------
These current meters can be grouped into two broad classes: 1) vertical-
axis rotor with cups or vanes and 2) horizontal-axis with vanes. A number of
refrences give the details on different current meters. Figure 3,14 shows the
propeller current meter which is typical of a horizontal-axis current meter
with vanes. Figure 3.15 shows the Price current meter which is typical of a
vertical-axis rotor current meter with cups.
Practical considerations usually limit the ratings of these meters to
velocities ranging from 0.030 m/s (0.1 fps) to about 4.57 m/s (15 fps). The
comparative characteristics of these two types are summarized below (4):
1. Vertical-axis rotor with cups or vanes
a. Operates in lower velocities than do horizontal-axis meters.
b. Bearings are well protected from silty water.
c. Rotor is repairable in the field without adversely affecting the
rating.
d. Single rotor serves for the entire range of velocities.
2. Horizontal-axis rotor with vanes
a. Rotor disturbs flow less than do vertical-axis rotors because
of axial symmetry with flow direction.
b. Rotor is less likely to be entangled by debris than are
vertical-axis rotors.
c. Bearings friction is less than for vertical-axis rotors because
bending moments on the rotor are eliminated.
d. Vertical currents will not be indicated as positive velocities
as they are with vertical-axis current meters.
e. They have a higher frequency of mechanical problems.
77
-------
Counter
Propeller
o
Figure 3.14 Propeller Meter(17)
Electric
cable
Support cjb<«
Revolution counter
\
Figure 3.15 Price Meter (17)
78
-------
To determine the discharge (flow volume), in addition to velocity of flow
it is necessary to determine the area of flowing water or wastewaters. This
holds especially for large flows in rivers, lakes, and wide and deep channels.
A depth sounding is necessary at each vertical and width measurement of the
cross-section of flow to determine the area of flowing water or wastewater.
Sounding rods, sound weights and reels, handlines, and sonic sounders are
common equipment used for depth determinations. Marked cableways and bridges,
steel or metallic taps or tag lines are used for width determinations. For
details or procedures for depth and width determinations, see reference (4).
3.3.1.2 Measurement of Velocity
To determine the discharge at a particular cross-section, it is neces-
sary to determine the mean velocity of flow at that section. In drag body
current meters such as vertical-axis deflection vane, horizontal-axis pendu-
lum type deflection vane and pendulum current meters, it is possible to inte-
grate velocities at different depths in a particular section to obtain the
mean velocity of flow, whereas inclinometer, drag sphere, rotating element
current meters and pitot tubes measure velocity at a point. Therefore, to
obtain the mean velocity of flow at a particular vertical section, it is neces-
sary to take velocity measurements at different depths. The various methods
of obtaining mean velocities are:
1. Vertical-velocity curve
2. Two-point
3. Six-tenths-depth
4. Two-tenths-depth
5. Three point
6. Subsurface
79
-------
Table 3.5 compares these methods in relation to application, flow
depth, velocity measuring point(s), and accuracy.
3.3.1.3 Time of Travel-Velocity Methods
a. Salt Velocity Method (1,2,5,6)
The method is based on the principle that salt in solution increases
the conductivity of water. This method is suitable for open channels of
constant cross-section and for flow in pipes. Sodium chloride and lithium
chloride are commonly used. The basic procedure is as follows:
1. Install two pairs of conductivity electrodes down stream from the
salt injection point at known distances and sufficiently far apart
in the stretch of the channel.
2. Connect the recording galvanometer to the electrodes.
3. Inject the slug of salt solution.
4. The time for salt solution to pass from the upstream to the down-
stream electrodes, in seconds, is determined by the distance on
the graph between the centers of the gravity of the peak areas.
AL
5. Calculate the discharge, using the formula Q = , where,
Q = discharge in cubic meters (cubic feet).
A = cross-sectional area of flow, square meters (square feet).
L = distance between the electrodes, meters (feet).
T = recorded time for salt solution to travel the distance
between the electrodes, seconds.
B. Color Velocity Method
The color velocity method is used for measuring high velocit flows
in open channels. It consists of determining the velocity of a slug of
dye between two stations in the channel. This velocity, taken as the mean
80
-------
TABLE 3 .5 COMPARISON OF VARIOUS METHODS TO OBTAIN MEAN VELOCITY
Methods Vertical-Velocity Two-point Six-tenth depth Two-tenth Three point Subsurface
Considerations Curve Method Method Method Depth Method Method Method
Application
Not for routine
discharge and
measurements
Generally
used
Primarily used
for depths less
than 2.5 feet
During times of
high velocities
when measurements
at 0.6 and 0.8
depth aire not
possible
When velocities
in a vertical
are abnormally
distributed
When it is
impossible
to obtain
soundings
and the depth
cannot be es-
timated to an
approximate
0.2 depth
setting
To determine
coefficients
for application
to the results
obtained by
other methods
When more
weight to 0.2
and 0.8 depth
observations
is desired
00
Flow depth
requirement
Greater than
2.5 feet
Greater than
2.5 feet
0.3 foot to
2.5 feet
No depth
constraint
Greater than
2.5 feet
Greater than
2.5 feet
Velocity
measuring
point(s)
Mean velocity
At 0.1 depth
increments be-
tween 0.1 and
0.9 depth
From vertical-
velocity curve
0.2 and 0.8
depth below
the water surface
*0.2
+ V0.
0.6 depth
below the
water surface
Observed velocity
is the mean
velocity
0.2 depth
below the
water surface
'mean =Cxv0.2
C=Coef ficient
obtained from
vertical-velocity
curve. At that
vertical for the
particular depth
of flow
0.2, 0.6 and
0.8 depth below
the water sur-
face
mean =
V0.2+V0.&+V0.6
At least 2
feet below
the water
surface
V
mean =
C x V observed
C=Coefficient
obtained from
ver t ical-velo-
city curve at
that vertical
for the parti-
cular depth of
flow
Accuracy
Most Accurate
Gives consistent
and accurate re-
sults
Gives reliable
results
If C is accu-
rately known can
give fairly reli-
able results
Gives reliable
results
Gives rough
estimate as C
is difficult
to determine
accuratelv
Vp j = Velocity at 0.2 depth from water surface
Vjj j = Velocity at 0.6 depth from water surface
V = Velocity at 0.8 depth from water surface
0 • O
vmean = Mean velocity
-------
velocity, multiplied by the cross-sectional area of flow gives the
discharge. Commercial fluorescein or potassium permanganate may be used
as the coloring matter. The color velocity is computed from the obser-
vations of the time of travel of the center of the mass of colored liquid
from the instant the slug of dye is poured at the upstream station to the
instant it passes the downstream station, which is at a known distance
from the upstream station.
With fluorescent dyes, the use of fluorometer to detect the center of
the colored mass will enhance the accuracy of the results,
c. Floats
There are three types of float methods used for flow measurements,
namely, surface floats, subsurface floats and integrating floats. To de-
termine the flow velocity one or more floats are placed in the stream
and their time to travel a measured distance is determined. These methods
are simple, but from an accuracy standpoint, they should only be used for
estimating the discharge.
Various surface floats like corks, stoppered bottles etc. and sub-
merged floats like oranges measure the surface velocity. The mean velocity
of flow is obtained by multiplying with a coefficient which varies from
0.66 to O.bO (2). A more sophisticated version is the rod-floats, which
are usually round or square wooden rods. These rods have a weighted end
so that they float in vertical position with the immersed length extend-
ing about nine-tenth of the flow depth. Velocity measured by the time of
travel by these rods is taken as the mean velocity of flow. These floats
are used in open channels and sewers.
To obtain better results, the velocity measurements should be made on
82
-------
a calm (winds are very minimal) day in a sufficiently long and straight
stretch of channel or sewer of uniform cross-section and grade with a
minimum of surface waves. Choose a float which will submerge at least one-
fourth the flow depth.
A more accurate velocity measurement is obtained by using integrating
float measurements. The method is simple and consists of the release of
buoyant spheres (like ping pong balls) from the channel floor. As these
spheres rise they are carried downstream by the flow velocity. The time
from the moment of release to the moment when they surface, and the
distance traveled downstream are measured. The discharge is measured
using the following relationships:
Q = DV and V = -
t
where Q = discharge in cubic meters/sec. (cubic feet/sec.)
per unit width of channel
D = flow depth, meters (feet)
V = terminal velocity of the float, meters/sec. (ft./sec.;
L = distance traveled downstream by float, meters (.feet)
t = time of rise of float, sec.
In flows of large depth and velocity, integrating float methods with
two floats of different velocities of rise are used (18,19). The dis-
charge is calculated, using the relationship:
„ _ d
-------
float (1)> respectively; and t2 and are times of rise of float (2)
and float (1), respectively.
The integrating float method is simple and does not require any
laboratory calibration. It integrates the vertical velocity profile and
yields the mean velocity or discharge per unit width of the section. The
method is suited to low velocities and is especially useful for flows
having abnormal velocity profiles, and it has practically no lower velo-
city limit. To get better accuracy, the reach of the stream to be meas-
ured should be sufficiently long and straight and the bed fairly uniform.
Use a fast rising float so that distance travelled downstream is of short
length. The shape of the float should be spherical (18).
3.3.2 Head Discharge Methods
This technique takes advantage of the head discharge relationship that
exists when a liquid flows over an obstruction or through a specific (conver-
gent-straight-divergent) channel section.
3.3.2.1 Weirs
A weir may be defined as an overflow structure built across as open chan-
nel, usually to measure the rate of flow of liquid.
Depending upon the shape of the opening, weirs may be termed rectangular,
trapezoidal, triangular, etc. When the water level in the downstream channel
is sufficiently below the crest to allow free access of air to the area beneath
the nappe, the flow is said to be free. When the water level under the nappe
rises above the crest elevation the flow may be considered submerged; the degree
of submergence depends upon the ratio of upstream and downstream head (height of
water above crest elevation). The effect of submergence is to cause large inac-
curacies in the flow mesurements. Therefore, the use of submerged weirs as the
84
-------
flow measuring device is avoided.
In a sharp crested weir, flawing liquid does not contact the bulk head
but springs past it. If the bulk head is too thick for the liquid to spring
past, the weir is classed as broad crested.
Weirs may be contracted or suppressed. When the distances from the sides
of the weir notch to the sides of the channel (weir pool) are great enough (at
least two or three times the head on the crest) to allow the liquid a free, un-
constrained lateral approach to the crest, the liquid will flow uniformly and
relatively slowly toward the weir sides. As the flow nears the notch it ac-
celerates, and as it turns to pass through the opening, it springs free laterally
with a contraction that results in a jet narrower than the weir opening. If
a rectangular weir is placed in a channel whose sides also act as the sides of
the weir, there is no lateral contraction, and the weir is called a suppressed
weir. Various types of weirs are shown in Figure 3.16.
Most of the flow measurements are conducted on sharp crested weirs without
submergence and the subsequent discussion is limited to this type. For infor-
mation on sharp crested weirs with submergence and broad crested weirs, refer
to reference (2) and other books on hydraulics.
A typical sharp crested weir is shown in Figure 3.17. Figures 3.18 a,b,
and c, show the various dimensions required for fully contracted rectangular
Cipolletti and V-notch weirs.
The relationship between head and discharge for different weirs is given
in Table 3.6. For rectangular weirs, the Francis formula is widely used for
flow measurements. However, it should be born in mind that it is applicable
and accurate only for sharp crested fully contracted or suppressed weirs. On
the other hand Kindsvater-Carter formula is applicable to any type of sharp
crested rectangular weir. It gives accurate results and is being increasingly
85
-------
SHAPE OF THE WEIR CREST
SHARP CRESTED
BROAD CRESTED WEIR
Flow
Flqw ».
a.
Flo
SHAPE OF THE NOTCH
RECTANGULAR
V-NOTCH

TRAPEZOIDAL
Flow
Flow,
n
c.
INVERTED
TRAPEZOIDAL
COMPOUND

POEBING
£3"
APPROXIMATE LINEAR

PROPORTIONAL APPROXIMATE EXPONENTIAL
FLOW CONTRACTION

SUPPRESSED
CONTRACTED
Figure 3.16 Types of Weirs
86
-------
Point to measure Depth,H
I
Straight I. At Least A H
Inlat Run "
Appro*. 5.08 cm (2M)
t
m-T«.
/
;ir\i 1- --
- -^
Figure 3.17 Typical Sharp Crested Weir (3)
ChaaiMl
wall J
a. Contracted Rectangular Weir
b. Contracted V - Notch Weir
y Contractions ^
f
B
For Full Contraction
P>2H and
Contractions> 2H
c. Contracted Cipolletti Weir
Figure 3.I8 Various Dimensions for Fully Contracted
Rectangular, Cipolletti and V-Notch Weirs (6)
87
-------
TABLE 3.6 HEAD-DISCHARGE RELATIONSHIP FORMULAS
Weir Type
Contracted
Suppressed
Remarks
Rectangular
Francis Formulas
Kindsvater-Carter
formula
Cipolletti
V-Notch
Cone formula for
90° V-Notch only
Kindsvater-shen
f ormula
Q* = 3.33(L-0.2H3/2J
Q - 3.33((H+h)3/2-h3^2]
(L-0.2H)
Q = C L H
^ e e e
1.5
Q = 3.33LH3/2
Q = 3.33L((H+h) 3//2-h3^2)
Q " Ce Le He
1 • 5
Q = 3.367 LH3^2
Q = 3.367L(H+1.5h)3/2
Q = 2.49 H2"1*8
8
Q =15 Ce tan(9^ (2gHe5) 1/2 NA
NA
NA
NA
Approach velocity neglected
Approach velocity taken
into consideration
Approach velocity neglected
Approach velocity taken
into consideration
V-Notch weirs are not
appreciably affected by
approach velocity
Q = discharge in cubic feet per second
H = head in feet h = head in feet due
Cg= coefficient, "L = L+k. , where k, = ratio
He= H+0.003 NA = Not applicable
9 = Angle of the notch H^H+k^
L = crest length in feet
to the approach velocity (V), = 62/2g
of crest (L) to channel width (B), k^L/B
-------
used.
The rate of flow determines the type of weir to use. A rectangular weir
is preferable for flows greater than 3.4 cubic meters/min. (2 cubic feet/sec.)
V-notch weirs are used for flows of less than 0.17 cubic meters/min. (1.0 to
10 cubic feet/sec.) (2). The Cipolletti weir is also used in the same range
as the rectangular weir. The accuracy of measurements obtained by the use of
Cipolletti weirs, based on the formulas given in Table 3.6 is inherently not
as great as that obtained with suppressed rectangular and V-notch weirs (2).
With these ranges in mind, the minimum head should be at least 5 cm
(0.2 ft.) to prevent nappe from clinging to the crest, and because at smaller
depths it is difficult to get sufficiently accurate gauge readings. The crest
should be placed high enough so that the water flowing over will fall freely,
leaving an air space under and around the jets. Requirements for standard
weir installations are shown in Figures 3.18 a,b, and c for rectangular,
Cipolletti and V-notch weirs, respectively.
For shapes other than those mentioned above, head-discharge relationship
must be extablished through field calibration using the salt-dilution (Section
3.4.3) or other methods.
Flow rates for 60° and 90° V-notch weirs can be determined from the nomo-
graphs in Figure 3.19. Figures 3.20a and 3.20b should be used for flow rates
of V-notch weirs in conjuction with the Kindsvater-shen formula (6); the
cone formula should be used only with fully contracted V-notch weirs. Flow
rates for Cipolletti weirs can be obtained from Figure 3.21. Figure 3.22 is a
nomograph for flow rates for rectangular weirs using Francis formula; whereas
Figure 3.23a and 3.23b should be used in conjunction with Kindsvater-Carter
formula.
89
-------
24-q
20
18-
16-
14
12-
10:
9:
8:
7:
6 -
5-
4-
3-
2-
|J

7000
(-6000
5000
-4000
r 3000
*
:
2000
: »00
^000
;600
-400
300
r 200
r 100
p80
:60
40
30
b 20
10
8
4
u3
o
O
u>
r 4000
r 3000
1-2000
r 1000
1-800
;€00
-400
300
£-200
r WO
-80
-60
-40
r 30
r 20
r 10
r 8
-4
r3
i
I
r2
r 18
w
ui
o
z
o
<
Ui
*
r24
r 20
18
»
16
14
•12
10
9
8
7
-6
-5
-4
Ll
90 V-Notch
2.48
60°V-Notch
2.50
Q • 2.49H Q = 1.443h
Where Q = discharge in cubic feet per second H - head in feet
Figure 3.19 Nomograph for Capacity of 60° and 90°
V-Notch VJeirs (16)
90
-------
0.012
-u
" 0.008
0.004
0
0 20 40 60 80 100 120
Notch angle, degrees
Figure 3.20a Value of Kindsvater-shen Formula for
V-Notch Weir
0.60
-------
8 9 1000
I i J
7 8 9 100
I I I 1
Q, liter/s
Figure 3-21 Flow Rates for Cipolletti Weirs (17)
-------
1.0
1.5
2D j
Z5\
30 ¦;
-i
4J0|
50":
60
7.0-j
i
80 i
9.0
10.0;
H
W
kJ
A
PS
F^i
O
O
S5
W
•J
Note:
130
200¦
250*
Based on Francis weir formula
as follows:
o
LO
C^J
m
w
vD
vO
S •
o o
hJ w
Pn
a 11
83
Fn W
P „
w a
•4 U
PS
eg
w o
u
w >
pq ffJ
O
O Fk
Eh
W
Crf
W
PU
w
§
o
BOOO
5000
2000
1000
45
59
33
r27
:r 21
18
500-
15
+" 12
"ir
9
>- 6
2O0
100
50-3
20-
•O^ 3
5-
i- 2
05
-• b1
^4
0.2-
0.1 j-'4
J
0D5^
OJ02-
oa
o.ooo ¦;
-\a
Where:
10000-
9000
8000
TOOO-
6000
5000
4000^
25
20
--15
3000 -
10
9
5-8
7
2OO0--
w
k-4
w
in
33 1000
f*J 9O0
800
700
§ 600-
H
Sj 500-!
e>
Z 400
g 300C
h 2004-
1.0
OS
OB
0.7
Q6
:-0.5
0.4 »
o
. r 0.3
100
90
80-t
70-
60-
50-
45-
•0.2
¦Oil
or
Q = 3.33LH3/2 (for suppressed weir)
Q = 3.33(L-0.2H)H3/2
« 3.33LH3/2-0.66H5'2 (for con-
tracted weir with two end contractions)
Q = discharge, in cubic feet
per second
L * length of weir, in feet
H = head, in feet.
Figure 3.22 Nomograph for
Capacity of Rectangular Weirs (7)
93
-------
0.020
0.010
V ft
-0.010

-0.003 '
0
0.80
1.00
0.20 0.40 0.60
L/B
Figure 3.23a Value of for L/B Ratio Kindsvater-Carter
Formula for Rectangular Weirs (31)
4.2
4.0
3.8
3.6
3.4
3.2
3.0

L/B =
ly

c =3.2
e
2 + 0.40
H/Py

/>
0.9

"<0.8

\6
/

0.4 '
/
0.2 f

0.4 0.8
1.2
H/P
1.6 2.0 2.8
Figure 3.23b Value of Ce for H/P Ratio Kindsvater-Carter
Formula for Rectangular Weirs (21)
94
-------
3.3.2.1.1 Criteria for Installing Standard Weirs
To achieve the best accuracy in flow measurement the following criteria
should be met in installing standard weirs (2):
1. The upstream face of the bulkhead should be smooth and in
a vertical plane perpendicular to the axis of the channel.
2. The upstream face of the weir plate should be smooth,
straight, and flush with the upstream face of the bulkhead.
3. The entire crest should be a level, plane surface which forms
a sharp, right-angled edge where it intersects the upstream
face. The thickness of the crest, measured in the direction of
flow should be between 1 and 2 mm (about 0.03 to 0.08 in.)
Both side edges of rectangular weirs should be truly vertical
and of the same thickness as the crest.
4. The upstream corners of the notch must be sharp. They should
be machined or filed perpendicular to the upstream face, free
of burrs or scratches, and not smoothed off with abrasive cloth
or paper. Knife edges should be avoided because they are
difficult to maintain.
5. The downstream edges of the notch should be relieved by
chamfering if the plate is thicker than the prescribed
crest width. This chamfer should be at an angle of 45°
or more to the surface of the crest.
6. The distance of the crest from the bottom of the approach
channel (weir pool) should preferably be not less than twice
the depth of the water above the crest and in no case less
than 4.72 cm (1 foot).
95
-------
7. The distance from the sides of the weir to the sides of approach
channel should preferably be no less than twice the depth of water
above the crest and never less than 4.72 cm (1 foot). (Exception:
suppressed rectangular weir for which sides of the notch should be
coincident with the sides of the approach channel).
8. The overflow sheet (nappe) should touch only the upstream edges of
the crest and sides.
9. Air should circulate freely both under and on the sides of the nappe.
10. The measurement of head on the weir should be taken as the difference
in elevation between the crest and the water surface at a point up-
stream from the weir a distance of four times the maximum head on
the crest.
11. The cross-sectional area of the approach channel should be at least
8 times that of the overflow sheet at the crest for a distance up-
stream from 15 to 20 times the depth of the sheet.
12. If the weir pool is smaller than defined by the above criteria, the
velocity of approach may be too high and the staff gauge reading
too low, and the head discharge relationship given in Section 3.3.1.1
will not hold good.
3.3.2.2 Flumes
In contrast to weirs which have a tendency to settle the suspended par-
ticles near their upstream side, most of the flumes have a self cleansing
feature which makes them a preferred flow measuring device where sediment is a
factor in the stability of the stage (head) discharge relation.
Flumes are comprised of three sections: a converging upstream section,
a throat or contracted section, and a diverging downstream section. The size
96
-------
of flume is the width of the throat section.
The following factors must be considered in the location of flume (2):
1. Do not install flume too close to turbulent flow, surging or un-
balanced flow or poorly distributed velocity patcern.
2. Locate flume in a straight channel section having no bends
upstream of the flume.
3. For convenience install flume at a location which is readily
accessible, near the diversion point, and near the devices installed
to control the discharge.
Some of the flumes commonly used as flow measurement devices are des-
cribed below.
a. Parshall Flumes
Parshall flumes have been developed in various sizes (throat width)
from 2.50 mm (1 inch) to 15.24 m (50 feet). The configuration and standard
nomenclature for Parshall flumes is given in Figure 3.24. Strict adherence to
all dimensions is necessary to achieve accurate flow measurement.
Flow through a Parshall flume may be either free or submerged. The degree
of submergence is indicated by the ratio of the downstream head to the upstream
head (H^/Ha) - submergence ratio. The flow is submerged if the submergence
ratio is:
greater than 0.5 for flumes under 0.076 m (3 in.) size
greater than 0.6 for flumes 0.15 m -0.23 m (6 in. - 9 in.) size
greater than 0.7 for flumes 0.3 m - 2.44 m (1 to 8 ft.) size
greater than 0.8 for flumes bigger than 2.44 m (8 ft.) size
For a free flow in a Parshall flume of size (W), the upstream head (H^)
and discharge (Q) relationship is given by the general equation Q *= CWHn.
97
-------
a) Plan
Flow
Floor
b) Section
Figure 3.24 Parshall Flume Configuration and Nomenclature (17)
Table 3.7 gives the values of c, n, and Q, for different sizes (.W) of the
Parshall flume. Nomographs, curves or tables are readily available to deter-
mine the discharge from head observations. Flow curves are shown in Figure
3.25 to determine free flow through 0.07 m to 15.24 m (3 in. to 50 ft.) Par-
shall flumes (4).
For submerged conditions, correction factor should be applied to the
free flow determined using the relationship Q = CWHn. These correction factors
are given in Figure 3.26 for different sizes of the Parshall flume.
98
-------
GPM
3000
2000
1000
FLOW
60 80100
15 20
30/40
I 2 3 4
HEAD-INCHES
8 10
CFS
.1 2
HEAO-FEET
FIVE INCHES IS Ml Ml HUH FULL SCALE
HEAD WITH FOXBORO FLOAT ANO CABLE
METER
8 10
0.02
FLOW
0.01
THIRTY SIX INCHES IS MAX I HUM FULL
SCALE HEAD WITH FOXBORO FLOAT ANO
CABLE METER
Figure 3.25 Flow Curves for Parshall Flumes (3)
99
-------
TABLE 3.7 FREE FLOW VALUES OF C AND N FOR PARSHALL FLUME
BASED ON THE RELATIONSHIP Q = CWHn (7)
Flume Throat, W C n Max; Q». c^s
1
in
0.338
1.55
0.2
2
in
0.676
1,55
0.5
3
in
0,992
1.55
1.1
6
in
2,06
1.58
3.9
9
in
3.07
1.522W0,026
8.9
1
ft
4W(*)
16. 1
1.5
ft
If
II
24.6
2
ft
II
II
33.1
3
ft
II
II
50.4
4
ft
tl
II
67.9
5
ft
II
II
85.6
6
ft
It
II
103.5
7
ft
If
fl
121.4
8
ft
II
II
139.5
10
ft
39.38
1.6
200
12
ft
46.75
1.6
350
15
ft
57.81
1.6
600
20
ft
76.25
1.6
1000
25
f t
94.69
1.6
1200
30
ft
113.13
1.6
1500
40
ft
150.00
1.6
2000
50
ft
186.88
1.6
3000
(*)W in feet
100
-------
4 ft
2 in. 8 In.
8 ft
8 h to
30 ft
throat
widths
0.9
0.7
as
as
too
Ho
Submergence, j— , in percentage
Figure 3.26 Correction Factor for Flow Discharge Deter-
mination for Parshall Flumes(22)
b. Palmer Bowlus Flumes
Palmer Bowlus flumes are venturi flumes of a super-critical flow type de-
signed as a device to be inserted into an existing conduit with minimal site
requirements other than sufficient slope. Figure 3.27 shows various types
of Palmer Bowlus flumes. A laboratory study indicates that accuracies within
± 3% of the theoretical rating curve could be obtained at depths as great as
90% of the pipe diameter (23). The chief advantage of Palmer Bowlus flumes
101
-------
over Parshall flumes is their ease of installation in existing conduits,
sewers, etc. Standard Palmer Bowlus flumes are available to fit pipe sizes
15.2 cm (6 in.) to 2.4 meters (8 ft.). A disadvantage of Palmer Bowlus flumes
is that they have a small range of flow, about 20:1.
Diskin flumes (24), an unconventional type of Palmer Bowlus flume, are
portable devices but have limiting submergence, (H^/Hg), between 0.75 and
0.85, and are not suited to trashy or debris laden flows.
c• Cut-throat Flumes
These are in a way modified Parshall flumes without throat section and
flat bottom. (Figure 3.28). They are suitable for flat gradient channels;
level flow and every flume size having same wall lengths makes construction
easy and less costly. Analytical and experimental background on these flumes
can be found in reference (24).
d. Type IIS, H, HL Flumes
These flumes are primarily used in irrigation channels and small water
sheds. Figure 3.29 illustrates these flumes. Their main advantage is sim-
plicity of construction, and they have a wide range of flow. Details on dis-
charge ratings can be found in references (2,25). Their design incorporates
the sensitivity of a sharp crested weir and the self cleansing feature of a
Parshall flume.
e' Other Flumes
Trapezoidal flumes (Figure 3.30) have much larger capacities than rec-
tangular flumes of the same bottom width. Two common types of flumes are:
1) Tapezoidal flumes with bottom slope, 2) Trapezoidal critical depth flume.
Accuracy of ±2 percent is claimed for trapezoidal critical depth flumes.
The San Dimas flume (Figure 3.31) was developed specifically to pajafs
102
-------
End View
Longitudinal Mid Sections
Vertical^ Horizontal
(a) y.
nurizouta.
•crn
(b)
1GT dzp
Figure 3.27 Palmer Bowlus Flumes (3)
STILLING WELL
FLOW
STILLING WELL
FOR Hay
PLAN

-4 *-
Li L,
SUBMERGED
¦ -FLOW

FLOW
"a
Hb

_—TRANSITION
FREE FLOW

. La r

B = VW-2Li/3=WL'2/3

La = 2iy9
ELEVATION ^
Figure 3.28 Rectangular Cut throat Flumes (5)
103
-------
HS
Figure 3.29 Type HS, Hand HL Flumes (5)
R=W
L=R+2h
u~x
PLAN
STILLING WELL
INTAKE
FLOOR ON
3% SLOPE
ZERO DATUM
FREE FALL
Figure 3.30 Trapezoidal Flume (5)
SIDE VIEW
Figure 3.31 San Dimas Flume (5)
-------
large amounts of sediment and debris. These flumes have the advantage that
neither approach conditions nor disturbances upstream or downstream have an
effect on their discharge ratings. Their rectangular cross-section makes them
less sensitive or accurate at low flows.
3.4 MISCELLANEOUS FLOW MEASUREMENT METHODS
3.4.1 Friction Formula
Measurements of channel or sewer bottom slope, depth of flow and flow
velocity can be used to only rough estimate the flow. The Manning formula is
commonly used for estimating a flow:
v = 0-453 R2/3 51/2
n
V = average velocity, m/s (fps ** 3.28 m/s)
n = coefficient of roughness
fcross sectional area of flowl
R - hydraulic radius, m [ wetted perimeter J
s = slope of energy grade line.
The Manning formula is widely used for the engineering design of sewers
and channels. However, for flow measurement,its usefulness is limited for a
number of reasons. It is difficult to assign an appropriate value to the rough-
ness coefficient which varies with the channel or sewer material (concrete,
brick, etc.), and the surface of the channel or sewer (new, old, etc). For
sewers, it varies also with the ratio of depth of flow to the depth when flow-
ing full. The other inaccuaracy that may enter into the flow measurement is
due to the slope of the energy grade line which is taken as the slope of the
channel or sewer. However, these two slopes may or may not be identical. For
various charts, tables and nomographs on the use of the Manning formula refer
to reference ( 26)
105
-------
3.4.2 Radioactive Tracer Techniques (7)
Radioactive tracer techniques measure the flow rate at the time of the
measurement. These techniques are simple and relatively inexpensive and the
equipment is portable. Successful use of these techniques requires a section
of the pipe or channel free of branch connections and requires turbulence at
the injection point for thorough mixing of the tracer. The tracer used must be
a gamma-ray emitter, must be compatible with the flowing liquid, and must have
a half-life longer than the duration of the test. Tracers generally used are
salts of cesium-134, iodine-131, sodium 24, or gold-198. There are two methods
of flow measurements by the radioactive tracers: 1) Two-Point Method and
2) Total-Count Method. Accuracies within 2% to 5% of the actual flow can be
achieved using these methods.
a. Two-Point Method
This method uses the time interval for the surge of tracer to pass
between two points separated by a determinable volume of the liquid. This
time interval is determined by peaks on the chronological chart of a
common amplifier-recorder connected to two G-M counters separated by a
known or determinable volume of a section of a pipe. The schematics of the
arrangement of the test is shown in Figure 3.32.
b. Total-Count Method
The basic principle of the total-count method is that a well mixed
finite quantity of radiotracer, A, passing through a measurement point will
produce a total number of N counts on a Scaler connected to a Geiger coun-
ter fixed in or near the stream some distance downstream. The value of N
is inversely proportional to the flow rate q and is directly proportional
to A, the quantity of the tracer mixed:
106
-------
A F
N = q , where F is a proportionality factor which is
characteristic of the isotope, the counter, and geometrical relationship
of the stream . Note that q is the flow rate at the tracer injection point.
The total-count method gains versatility through the divided-stream
principle: The same number of counts is obtained on the fraction or
split flow as is obtained on the total flow. This allows one to measure
a small fraction or bypass of the total flow.
To obtain accurate results, the numerical value of F must be deter-
mined in the laboratory by exposing the counter to a tracer solution in the
same geometrical arrangement as in the field test, to find the counting
rate that corresponds to a certain concentration of the tracer.
For example, if one desires to measure the flow of water/wastewater
through a 30.4 cm (12 in.) pipe, take a 60.8 cm (2 ft.) length of 30.4 cm
(12 in.) pipe closed at one end, and fill it up with water/wastewater con-
taining a known concentration of the radioactive tracer C to obtain milli-
curies per cubic meter (gallon). Strap the Geiger counter to the pipe and
connect it to a scaler. Determine' the number of counts per minute, n.
Then the factor, F, for cubic meters per minute (gallons per min.) is:
3, _ n Counts per minute
m /m n - c miHiCuries per cubic meter
Arrangement for the field measurement is schematically shown in
Figure 3.33, upper post. To place the measurement, inject a known amount
of tracer, A, either in a slug or gradually and record the total number of
counts, N. Calculate, the flow using the formula
A
Q " jj F substituting these values, and value of
107
-------
RATEMETE«
TRACER
INJECTOR
¦E_E
I
COUNTERS
n-
Figure 3.32 Schematic of Two Point Method (7)
TRACER
INJECTOR
A MILLICURIES
SCALER
I
cubIC MetEF
MINUTE
DQQ
N COUNTS
COUNTER
I
F CQUNTS / MILLICURIES =
i!
MINUTE/ CUBIC METER

quo
Figure 3.33 Schematic of Total Count Method (Upper Post) and
arrangement for the Determination of F-Factor (Lower Post) (7)
108
-------
o
F m /min. obtained above.
The divided-stream principle is used in a modified technique, the
sample-bucket technique, in which a fraction of total flow is passed
through a bucket containing the counter. The factor F is determined with
the actual bucket and the counter.
The procedure for measuring flow of a large open stream, such as a
river, is accomplished by floating the counter any place in the flow down-
stream from the injection point. The value of F, is predetermined by
submerging the counter at least 15.2 cm (6 in.) under the surface of liquid
in a tank at least 1.2 m (4 ft.) in diameter.
For better sensitivity a bundle of four counters connected in parallel
and enclosed in lucite pipe is used.
3.4.3 Chemical Dilution (2,6,7,27)
Chemical dilution technique often known as the salt dilution technique, is
applicable to both the open channel and pipe flow. This technique does not
require the stream dimensions or the measurements of fluid levels or pressures.
The flow is determined by measuring the concentration of the chemical at two
points downstream from the injection point. The following should be considered
when using this technique for flow measurements in waters and wastewaters:
Turbulence at the point of injection of the chemical should assure
thorough mixing (especially the lateral mixing) of the chemical
in the field.
Flow in the channel or pipe should be steady.
Chemical used should meet the following requirements:
Compatible with the fluid; no loss or deterioration of
the chemical in the fluid.
109
-------
. Nontoxic to plant and animal life.
. Easy and accurate quantitative detection at low concentration.
. Low cost of the chemical and the equipment.
Chemicals commonly used are lithium chloride (atomic adsorption analysis
of lithium) and fluorescent dyes (fluorometer measurement) such as sodium
fluorescein, Rhodamine B, Pontacyl Brilliant Pink B, and Rhodamine WT. However,
use of sodium fluorescein is not recommended as it is easily affected by light
and bacterial action. In waters/wastewaters with high suspended solids, there
will be pronounced loss of Rhodamine B dye. Pontacyl Brilliant Pink B and
Rhodamine WT dyes are compared in Table 3.8.
The chemical dilution technique is used in two ways: 1) continuous ad-
dition or 2) slug injection.
a. Continuous-Addition-of-Chemical Method
In this technique the chemical of known concentration is added at a uniform
rate to the stream and the dilution is determined after it has traveled
downstream at least a distance 100 times the width at the surface of the
fluid. The relationship to determine the flow is
A = q C1~C2
C2_C0 where,
A = stream discharge
Cq = natural (or background) concentration of the chemical
in the stream
Cj = concentration of the chemical injected
C2 = final concentration of the chemical at downstream
sampling point
q = rate of injection of the chemical
110
-------
TABLE 3.8 COMPARISON OF RHODAMINE B, REODAMINE WT AND
PONTACYL BRILLIANT PINK B DYES (27)
Factors
Rhodamine B
Rhodamine WT
Pontacyl Brilliant
(Pink B)
pH 5-10
Stable
Stable
Stable
Absorption
peak-visible
light range
550 mu
556 mu
560 mu
Fluorescence
peaks
570 mu
580 mu
578 mu
Suspended
Pronounced
Low
Low
solids absorption absorption absorption
-------
b. Slug Injection Method
In this method a known amount, S, of the chemical is added to the
stream. At a point sufficiently downstream (minimum 100 times the width
at the surface of the flow), the concentration, C, of the chemical during
its time of travel, At, is determined by continuously sampling from the
stream during the passage of the chemical wave and mixing this constant
continuous sample into a single container to obtain an "integrated sample"*
S
The flow is determined by the relationship 0 = —— where,
CAt
Q = stream discharge
S = amount of chemical injected
C = average concentration of chemical during its
passage over a downstream point during
time interval At.
3.4.4 Water Meters
An estimate of the flow can be obtained from water meter readings when
an instantaneous flow rate is not critical. This technique is used in a con-
fined area, such as the industrial plant. Water meters should be certified
periodically. When using the incoming and outgoing flow for an initial esti-
mate of the flow rate, all changes in the water quantity that occur in various
processes must not be overlooked. These changes may be due to water actually
consumed in the process, e.g., cement manufacturer, conversion of quick lime to
slaked lime, or the change of phase: water changing to steam, etc.
3.4.5 Measuring Level Change in Tank
In some instances the level change in a tank can be used to estimate flow.
To accomplish this, the volume of the tank related to depth must be established;
then the flow is allowed to enter and the level change with time is recorded.
Figure 3.34 gives the relationship of depth to volume for various shapes of the
112
-------
SPHERE
RIGHT CYLINDER
(C
c:
Total Volume
V = 1/6 ttD3 = 0.523A98D3
Partial Volume
V = 1/3 irdz (3/2 D-d)
Total Volume
V = 1/4 kiD^H
Partial Volume
V - 1/4 nD^h
ANY RECTANGULAR CONTAINER
y

>
H
I
TRIANGULAR CONTAINER
Case 1
Case 2
ELLIPTICAL
CONTAINER

H
¦fh
4
J*
\V
J'r
Total Volume
V = HLW
Partial Volume
V - hLVJ
Partial Volume (Case 1)
V - 1/2 hBL
Total Volume
V - 1/2 HBL
Partial Volume (Case 2)
V - 1/2 L (HB - hB)
Total Volume
V - *BDH
Partial Volume
V - uBDh
Figure 3.34 Equations for Container Volumes
113
-------
FRUSTUM OF A CONE
Case 1
Case 2
Total Volume
V = it/12 h(Dj2 + Dj D2 + D22)
Partial Volume
V = tt/12 h^2 + Dj d + d2)
CONE
Case 1
Case 2
(Case 1)
Partial Volume
V = 1/12 tt d2h
Total Volume
V = 1/12 71 D2H
Partial Volume (Case 2)
V = 1/12 tt(D2H - d2h)
PARABOLIC CONTAINER
Partial Volume
V = 2/3 hdL
Case 1
Total Volume
=2/3 HDL
Partial Volume
Case 2
Figure 3.34 (Continued)
-------
tank.
3.4.6 Pump Rates
When other methods are not available for flow measurement and a pump is
used in the system, the operating characteristics of the pump can be used to
estimate flow. One method Is to multiply the pumping time and the pump capacity
at the discharge pressure as obtained from manufacturer's head curves vs. flow
(28).
Another technique is to establish the pump's horsepower and determine the
capacity from the manufacturer's curves. However, these techniques should be
used only for estimates of flows.
3.4-. 7 Calibrated Vessels
Another technique useful for free falling water is to capture a known
volume of water over a recorded time interval. The flow rate is then estab-
lished for a specific time. More than one measurement is necessary to allow
accurate estimates; the volume chosen should allow time of collection to be
more than 10 sec. (29).
3.5 SECONDARY DEVICES
Secondary devices are the devices in the flow measurement system which
translate the interaction of primary devices in contact with the fluid into
the desired read-out or records.
These devices can be classified into two broad classes:
1. Non-recording type with
a. Direct read-out such as a staff gauge
b. Indirect read-out from fixed points as in a chain, wire
weight and float type.
2. Recording type, where the recorders may be graphic or digital.
Examples of recording type devices are: float in well, float in
flow, bubbler, electrical and acoustic.
115
-------
The advantages and disadvantages of the various secondary devices are
given in Table 3.9 and relative comparison of primary and secondary open
channel flow measurement devices is shown in Table 3.10. Table 3.11 compares
various recording type secondary devices.
3.5.1 Non-recording Type Secondary Devices
3.5.1.1 Staff Gauge
A staff gauge, shown in Figure 3.35a, is usually a graduated enameled
steel plate bolted to a staff. Care must be taken to install the gauges
solidly to prevent errors caused by change In elevation of the supporting
structure.
3.5.1.2 Hook Gauge
A hook gauge, shown in Figure 3.35b, is a modification to a staff gauge.
The gauge (hook) is manually brought to the water surface and the water eleva-
tion read.
3.5.1.3 Chain Gauge
The chain gauge, shown in Figure 3.35c, is a substitute for the staff
gauge and consists of a horizontal seal and a chain that passes over a pulley
to fasten a hanging weight. Water level is indicated by raising or lowering
the weight until it just touches the water surface. Sources of errors in the
measurement are,settling of supporting structure, temperature changes, changes
in length due to wear, and wind action.
3.5.1.4 Wire Weight Gauge
Wire weight gauge, shown in Figure 3.35d, is a modification of the chain
gauge and uses a wire or small cable wound on a reel. The reel is graduated
or a counter is used to give readings from a reference check bar of the water
elevations to the tenths and hundredths of a foot.
116
-------
TABLE 3.9 ADVANTAGES AND DISADVANTAGES OF SECONDARY DEVICES
Device
Advantages
Disadvantages
Hook gauge or
stage board
Common, accurate
Manual only, stilling
well may be needed
Differential Pressure Measurement
a. Pressure bulb
b. Bubbler tube
No compressed air
source, can be di-
rectly linked to
sampler
Self-cleaning, less
expensive; reliable
Can clog openeings,
expensive
Need compressed air or
other air source; can't
stand much abuse
Surface float
Inexpensive,
reliable
In-stream float catches
debris
Dipper
Quite reliable,
easy to operate
Oil and grease will
foul probe, expensive,
possible sensor loss
Ultrasonic
No electrical or
mechanical contact
Air bubbles may cause
echo rebounding
117
-------
TABLE 3.10 RELATIVE COMPARISON OF PRIMARY AND SECONDARY OPEN CHANNEL
FLOW MEASUREMENT DEVICES (a)
Primary devices
Secondary devices
Channel-char's
only (Manning
Hook gauge Differential Float
Ultra-
Characteristic
Suitable for continuous
measurement
Capability for sending
signal to sample (flow-
proportional sampling)
Need for stilling well
Low initial cost
Easy to install
High accuracy of measurement
Low maintenance (incl. cleaning)
Suitable for high solids
wastewater
Low susceptibility to fouling
(rags, debris, grease)
Wide flow range
Low headloss
Low auxiliary requirements
(manpower, compressed air,
AC power)
formula) Weir Flume stage board pressure Device Dipper sonic
na
na
2
2
2
1
na
na
1
1
3
3
+
3
3
2
3
3
+
+
1
2
+
+
+
+
3
1
3
2
1
+
+
1
+
+
3
+
+
(a)na = not applicable
- = no or not suitable
+ = yes or suitable
1 = fair frequently a problem
2 = good, sometimes a problem
3 = excellent, seldom or never a problem
-------
TABLE 3 .11 COMPARISON OF RECORDING TYPE SECONDARY DEVICES
Features
Floating in Well Float in Flow
Stilling Well
Sensing Flow
Level
Purge System
Moving Parts
Necessary
Indirectly
Not required
Presence of
moving parts
Not necessary
Directly
Not required
Presence of
moving parts
Bubbler
Electrical
Acoustic
Not necessary
Not necessary
Not necessary
Flow level
Flow level
Flow level
translated
translated
translated
into air
into electri-
into acoustic
back pressure
cal property
response
Maybe required
Not required
Not required
Absence of
Absence of
Absence of
moving part
moving part
moving parts

where sensing

element is

physically in

the flow.

Present where

probe is lowered

for flow sensing.

-------
4-s
9-i
PS
9-g
Iw"
9-=
"J "
9-1
a-§
5-1
5-Jj
3-1
2-g
*•—:r •
«B#1
: >
•
: ¦«
: r
.1—t.t.u
i **
---*f—1
a) Staff gauge
ra
V/
b) Hook gauge
d) Wire Weight gauge

Chain index monKi
/
~-Scale
• •
o o
1
fftofimawri)
— «y KJ «f tO 10 K O 91 O — AiV.y IO UN/
o o do'o'o obo r ' -1
^Weight
c) Chain gauge
Figure 3.35 Various Non-recording Type Secondary Devices
120
-------
3.5.2 Recording Type Secondary Devices
3.5.2.1 Float in Well
It essentially consists of a float (sensor weight) and a counter weight
connected via a cable to a wheel which rotates as the float rises or falls with
changes in the water level. The wheel is connected mechanically or electronic-
ally to the read-out, recorder, etc. The float is installed in a stilling well.
3.5.2.2 Bubbler
In a bubbler, Figure 3.36, a pressure transducer senses the back pressure
experienced by a gas which is bubbled at a constant flow rate through a tube
anchored at an approximate point with respect to a primary device. This back
pressure can be translated into water depth and subsequently related to dis-
charge .
Air Supply
2-4-2
Meter box and
Recorder
O
Pressure gauges and
reducing valve - normally
in meter box as part of
meter
Bubbler Pipe

1
i
or stilling well to
measure depth of flow

'W'J>W

Figure 3.36 Bubbler
121
-------
3.5.2.3 Electrical
These devices use the change in an electrical property (capacitance,
resistance, etc.) to sense liquid depth. The probe or sensor is a part of an
electrical circuit, and its behavior in a circuit is a function of its degree
of immersion.
3.5.2.4 Acoustic
With acoustic devices, continuous measurement of liquid depth is accom-
plished by measuring the time required for an acoustic pulse to travel to the
liquid-air interface and return. Of the two physical arrangements, liquid
path and air path measurement, the air path arrangement is commonly used since
installation is simplified, is independent of fluid velocity, and avoids any
contact with the fluid.
3.5.3 Errors in Flow Measurement (20,30)
The final measurement accuracy of a system (primary and secondary devices
included) depends on many factors.
3.5.3.1 Sources of Errors Related to the Primary Devices
Sources of errors described here are in relation to weirs and flumes, but
similar arguments can be developed for other devices. The following are the
main sources of errors:
• Basic errors in the discharge/head tables or formulas. In many in-
stances, the discharge tables, charts or formulas have been developed
empirically. They show experimental relationship. Therefore, extra-
polation beyond the range of observations from which they were developed
can lead to serious errors.
• Faulty fabrication or construction. Erroneous Length: An error of
0.1 foot in the length of a rectangular or Cipolletti weir will cause
122
-------
an error of 1% in the flow measurement of a one foot weir. A correspon-
ding error in 0.30 meter (one foot throat-width) flume will be 0.86% and
that in a four foot flume 0.23%.
Error due to transverse slope of weir crest. When the crest of the
rectangular or Cippoletti weir is sloped, the common practice is to
measure the head at the center of the crest. This leads to an error of
100S2L2
32H2
S = Slope of the weir crest
L = Length of the weir
H = Head at the center of the weir crest.
This error can be reduced to an insignificant amount if the discharge
is calculated as the difference of the discharges based on higher and
lower heads on the weir crest.
Stilling well not at a proper location. The head of the weir must be
measured beyond the effect of the drawdown. For standard weirs the
stilling well for the head measurement should be placed at a distance
upstream of four times the maximum head on the weir. For Parshall
flumes the locations of stilling wells for the head measurement bear a
definite relationship with the throat width. Substantial errors in the
field measurements have been traced to changes in the location or design
of the stilling well entrance.
Errors due to neglecting velocity of approach to weir. When the
velocity of approach is greater than 0.5 fps it should be considered
in the discharge formula. For a 0.2 ft head on the weir, this error
for approach velocities of 0.15 m/s, 0.30 m/s, and 0.46 m/s (0.5 fps,
1.0 fps, and 1.5 fps) is 2.7, 9.8, and 20.8 percent respectively. This
123
-------
error is less when the head on the weir is greater. For a 0.30 m (1 ft)
head,corresponding figures are 0.6, 2.2, and 4.7 percent. Use of
the Kindsvater-Carter formula will help alleviate this error.
• The error due to the reduction of depth of the weir pool. The height
of the weir, when less than twice the head on the weir, will introduce
an error of 5.6, 2.7, and 1.5 percent for ).06 meter (0.2 ft) head and
0.15, 0.30, and 0.61 meter (0.5, 1.0, 2.0 ft) height of the weir. A
corresponding error of a 0.5 foot head will be 13.1, 6.4, and 3.4 per-
cent respectively. This error can be corrected by using Rehbock's
formula, q = LH3/2(0.605 + —± + 0.08 S) or the
3 v 320H-3 P
Kindsvater-Carter formula. In a standard sized weir pool, this error
can be minimized or eliminated by proper maintenance and cleaning.
• Weir blade sloping upstream or downstream. The error introduced is
normally small. It becomes significant, however, if the face goes out
of plumb by a few degrees.
• Roughness of upstream face of weir or bulkhead. The roughness of the
upstream face of weir or bulkhead can cause an increase in the discharge.
The discharge is observed to increase by changing the roughness of the
upstream face of the weir bulkhead from that of a polished brass plate
to that of a coarse file for a distance of 30.48 cm (12 in) below the
crest. The increase ranges from 2 percent for 0.15 meters (0.50 ft)
head to about 1 percent for 0.412 meter (1.35 ft) head (30).
• Aeration of the nappe. Insufficient aeration of the nappe will increase
the discharge over the weir. It has been observed that for a drop in
pressure under the nappe by 20.32 mm (0.8 in) of water below atmospher-
124
-------
pressure, the discharge increased by 3.5 percent at 0.15 meter(0.5 ft)
head and about 2.0 percent at 0.30 meter(1.0 ft) head (30).
• Other errors may be due to submergence of the weir. Obstructions in
the measuring section, changes in the viscosity and surface tension,
unstable flow at very low heads, etc.
3.5.3.2 Errors in the Secondary Devices
• One of the most common errors is the incorrect zero setting of the
head gauge. This error is of the same magnitude as the error for mis-
reading the head.
• Error due to misreading the head is another common error. Common causes
of this error are incorrect location of the gauge, the dirty head gauge,
not using the stilling well, considerable fluctuations of the water
surface and carelessness on the part of the reader. For 30.48 cm -
12.19 m (12-48 in) Cipolletti and 90° V-notch weirs, a small error of
3.05 cm (0.01 ft) in reading will introduce an error approximately 7.5
percent in discharge results for the lower heads. For greater heads,
the error is less.
• The chart related errors are common to all the recording type devices.
These errors are the result of the variations in the chart due to humid-
ity, paper expansion and shrinkage.
• The error common to the totalizers is the variation in the speed of
totalizer drive motors.
• Other errors which are characteristics of particular secondary devices
are:
Float Devices (12)
The error due to a float lag which is similar to the "play" be-
125
-------
tween gears. Once the index is set to the true water level
while the water is rising, it will thereafter show the correct
water level. For a falling water level however, the index will
be above the true water level by the amount of the float lag
as shown in Figure 3.37a. If the index is set at true water
level at some intermediate point between rising and falling
water levels, the index will be proportionately low by the
amount of the float lag for rising water levels and high a sim-
ilar amount on the falling water level, as shown in Figure 3.37b.
For recorders and indicators,
F
float lag = 0.37 where F = force required to move the
mechanism, ounces. D = diameter of the float, inches, and
float lag in feet.
The error due to line shift. For every change in the water
level, there is a movement of float line from one side of
the float pulley to the other. This change of weight changes
the depth of floatation of the float, consequently the stylus
deviates from the true water height by a small amount. This
is dependent on the change in the water level since the last
correct setting, and the weight of the line used between the
float and the counter weight.
P
Error from live shift = 0.37 p2 AH where
p = weight per unit length of the line ounces
D = diameter of the float, inches
AH = change in water level, feet and error from line shift
shift in feet.
126
-------
Float Lag
True Water Level
Time
a) Showing float lag when index is set to
True Water Level while the water is rising
Float Lag
True^,Water Level
Time
b) Showing float lag when index is set at some
intermediate point between rising and falling
water levels.
Figure 3.37 Float Lag (12)
127
-------
If the error from line shift occurs when the counter weight
P
is submerged, the error = 0.34 —2 AH
The error for the submergence of the counter weight, is the
result of the reduced pull on the float which leads to the
increased depth of floatation. The error for the submergence
is given by AX.
Ay -* - - P(L-2&) r 0
AX SCWA I2 SjJ
where,
C = the counter weight
Sc = specific gravity of the counter weight
W = weight of the float
P = weight per unit length of the float line
L = total length of the float line from float to counter
weight
length of the float line, on the counter weight side
A = area of the float
Sj = specific gravity of the float line.
. The error due to fouling by trash or debris.
Bubbler
clogging of the exit and base of the bubble tube
aspiration effects due to velocity of flow
Errors due to temperature and aging
. Errors due to hysteresis (lag effect)
Electrical
Main error is due to foam, floating oil or grease in the
128
-------
liquid
Acoustic
The main errors are due to foam, highly turbulent-flow
and false echo in restricted sites like manholes, meter
vaults, etc.
3.5.3.3 Total Error in the Flow Measurement
• Often the total error in the flow measurement in a system is wrongfully
taken as the sum of the errors in the primary and the secondary devices. How-
ever, the total error in the flow measurement is the square root of the sum of
the squares of the individual errors (31). Illustrative example is given
below:
In the flow measurements through a 30.48 cm (12 in) Parshall flume, the
flow was 0.21 m-Vs (7.41 cfs) at 457.20 mm (18 in) of head. It was observed
that there was a 3 percent error in the flow measurement for the Parshall
flume. The error introduced by the use of a flow measurement formula was 1.5
percent. There was an error of 6.350 mm (1/4 in) in the measurement of the
throat. The error due to incorrect setting of zero was 3.175 mm (1/8 in) and
the error in the reading of the head was 3.18 mm (1/8 in).Calculate the total
percentage error.
Percentage error in the
head measurement (secondary
device)
(Head)1
Xn(e) = 100
Nl
(3.175)2 + (3.175)2
(457.20)2
» .982 - 1 percent approximately
Pecentage error in the
primary device dimensions
¦ 2% approximately
129
-------
Percent total error in the system
= X =
\
/Percent \
| error of1
Ithe flow/
(Percent \ 2
error of +
the formula/
(Percent\ 2
error of\ +
primary J
device /
/Percent ^2
I error of |
t secondary/
\ device J
= 32 + 1.52 + 22 + l2
= 4 percent approximately
130
-------
3.6 REFERENCES
1. The American Society of Mechanical Engineers, "Fluid Meters-Their
Theory and Application," Report of the ASME Research Committee on
Fluid Meters, 6th Edition, New York, N.Y., 1971.
2. U.S. Department of Interior, "Water Measurement Manual", Bureau of
Reclamation, 2nd Edition, revised, 1974.
3. Associated Water and Air Resource Engineers, Inc., "Handbook for In-
dustrial Wastewater Monitoring", U.S. Environmental Protection Agency,
Technology Transfer, August 1973.
4. Buchanan, T.J., and W.P. Somers. "Discharge Measurements at Gaging
Stations", U.S. Geological Survey, Techniques of Water Resource In-
vestigations, Book 3, Chapter A8, 1976.
5. Shelley, P.E., and G.A. ICirkpatrick. "Sewer Flow Measurement-A
State of the Art Assessment", U.S. EPA, EPA.^600/2-75-027 , November, 1975.
6. Kulin, Gerson and P.R. Compton. "A Guide to Methods and Standards for
the Measurement of Water Flow", U.S. Department of Commerce, National
Bureau of Standards, Special Publication 421, May 1975.
7. American Petroleum Institute. "Manual on Disposal of Refinery Wastes.
Chapter 4, 1969 p. 1-26.
8. King, H.W., "Handook of Hydraulics", 4th Edition, McGraw Hill, 1954.
9. Strater, V.L., "Fluid Mechanics", McGraw Hill, 1966.
10. Perry, R.H. and Chilton, C.H., "Chemical Engineers' Handbook", 5th
Edition, McGraw Hill, 1974.
11. American Society of Testing Materials. "Annual Book of ASTM Standards,
Part 31-Water", Philadelphia, Pennsylvania, 1976.
12. Leupold and Stevens Incorporated, "Stevens Water Resource Data Book", 2nd
Edition (revised), Beaverton, Oregon, 1975.
13. Kennard, J.K., "Elementary Fluid Mechanics", 4th Edition, John Wiley
& Sons, Inc., New York.
14. Blasso, L., "Flow Measurement Under any Conditions", Instruments and
Control Systems, 48., 2, pp. 45-50, February 1975.
15. Thorsen, T., and R. Oen. "How to Measure Industrial Wastewater Flow",
Chemical Engineering, 82, 4, pp. 95-100, February 1975.
16* Smoot, G.F., "A Review of Velocity-Measuring Devices", USDI, U.S.
Geological Survey, Open File Report, Reston, Virginia, 1974.
131
-------
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Simon, Andrew L., "Practical Hydraulics", John Wiley & Sons, Inc.
New York, 1976
Liu, H; "Analysis of Integrating-Flot Flow Measurement", Proceedings of
the American Society of Civil Engineers, HY5, September 1968. p. 1245-
1260.
Hajos, S. Neves, "Verfahren zur Messung Kleiner Wassergeschwindigkeitan",
Zentralhlav der Bauvewaltung 2h_ (44), 1904 pp. 281-283.
Bos, M.G., Editor, "Discharge Measurement Structures", working group on
Small Hydraulic Structures, International Instutute for Land Reclamation
and Improvement, Wageningen, The Netherlands. 1076
Kindsvater, C.E., and Carter, R.W. , "Discharge Characteristics of Rec-
tangular Thin-Plate Weirs", Paper No. 3001, Transactions, ASCE Vol. 124,
1959.
Robinson, A.R., "Simplified Flow Corrections for Parshall Flumes Under
Submerged Conditions", Civil Engineering, ASCE. Sept., 1965.
Wells, E.A. and H.B. Gotaas, "Design of Venturi Flumes in Circular
Conduits", American Society of Civil Engineering, 82^, p. 23, April 1956.
Skogerboc, G.V. , R.S. Benett, and W.R. Wallcer, "Generalized Dis-
charge Relations for Cut-Throat Flumes", Proc. American Society of Civil
Engineering, 9_8, IR4, pp. 569-583, December 1972.
Shelley, P.E., and G.A. Kirkpatrick. "An Assessment of Automatic
Sewer Flow Samplers", Office of Research and Monitoring, U.S. Environ-
mental Protection Agency, EPA-600/2-76-065, Washington, D.C., December
1975.
WPCF, ASCE, "Design and Construction of Sanitary and Storm Sewers", WPCF
Manual No. 9, ASCE Manual and Reports on Engineering Practice No. 37,
New York, 1969.
Regpolo, J.A., L.E. Myers, and K.J. Brust; "Flow Measurements with
Fluorescent Tracer", Proceedings of ASCE. HY5, 1966, p. 1-15.
Forester, R., and D. Overland, "Portable Device to Measure Industrial
Wastewater Flow", Jour. WPCF 46, pp. 777-778, April 1974.
Rabosky, J.G., and D.L. Koraido. "Gauging and Sampling Industrial
Wastewaters", Chemical Engineering, pp. 111-120, January 1973.
Thomas, C.W., "Common Errors in the Measurement of Irrigation Waters",
Proc. Paper 1362, ASCE IR2, 1957, p. 1-24.
Mougenot, G., "Weirs and Flumes", Water and Sewage Works, pp.79-81,
July, 1974.
132
-------
CHAPTER 4
STATISTICAL APPROACH TO SAMPLING
For every sampling program four factors must be established:
1. Number of samples
2. Sampling frequency
3. Parameters to be measured
4. Location(s) of sampling
These variables are usually established in varying degrees by the discharge
permit requirements which may or may not be scientifically sound. In those
cases where a new program Is being initiated or where the permit require-
ments need review, statistical methods and scientific judgment should be used
to establish the best procedures.
This chapter explains various statistical terms and techniques and their
applications to sampling. Each new concept is introduced with an example to
illustrate its use. After the basic terms are defined and illustrated,
statistical methods are introduced for analyzing data and determining the
above four factors* These methods are also illustrated with examples.
4.1 BASIC STATISTICS AND STATISTICAL RELATIONSHIPS
Data representing a physical phenomenon are broadly classified as
continuous (such as temperatures measured constantly and recorded as a con-
tinuous curve) or discrete (such as temperatures recorded hourly), and as
deterministic (able to be described by an explicit mathematical relationship
or formula) or nondeterministic (random). Due to the nature of water quality
133
-------
changes and the complexity of the processes affecting the water or wastewater
characteristics, there is no way to predict an exact value for a datum at a
future instant in time. Such data are random in character and are conven-
iently described in terms of probability statements and statistical averages
rather than by explicit equations. However, long-term changes in water qual-
ity tend to have a functional character with random fluctuation components.
Statistical evaluation techniques provide a tool with which to detect and quan-
tify both the deterministic and random components of a water or wastewater
quality record.
4.!•IStatistical Sample Parameters - Definitions and Examples ^
A wastewater stream was sampled once a week for a period of one year and
the concentration of a cerain parameter recorded. (See Table 4.1)
These data don't give much information as presented, so certain opera-
tions are performed on them to give them some meaning. Two things that give
useful information about a set of data are measures of location (such as
arithmetic mean and median) and measures of spread (such as range, variance
and standard deviation).
4.1.1.1 The Arithmet1c Mean
The arithmetic mean or simply the mean is one statistic used to lo-
cate the :'center" of a data set. It is defined to be the sum of all the ob-
servations divided by the number of observations (N) or, in symbols:
N
E xi
X = fail
N
134
-------
TABLE 4 .1 WASTEWATER PARAMETER DATA
Week
Concentration
Week
Concentration
1
35.8
27
31.1
2
33.0
28
3J.6
3
33.6
29
28.9
4
35.0
30
35.6
5
33.5
31
32.9
6
34.7
32
31.8
7
33.6
33
37.4
8
36.9
34
32.0
9
38.8
35
34.8
10
35.5
36
31.7
11
32.2
37
32.7
12
32.2
38
36.0
13
33.3
39
34.2
14
33.5
40
30.3
15
33.0
41
39.6
16
33.1
42
34.6
17
33.5
43
31.7
18
31.9
44
30.3
19
31.7
45
34.4
20
32.4
46
32.4
21
34.8
47
31.1
22
33.5
48
36.5
23
33.9
49
33.2
24
32.0
50
34.3
25
34.2
51
35.8
26
33.4
52
32.4
TABLE 4.2 WASTEWATER PARAMETER DATA IN NUMERICAL ORDER

Observation <1
Concentration
Observation #
Concentration
1
39.6
27
33.5
2
38.A
28
33.4
3
37.4
29
33.3
4
36.9
30
33.2
5
36-5
31
33.1
6
36.0
32
33.0
7
35.8
33
33.0
8
35.8
34
32.9
9
05.6
35
32.7
10
35.5
36
32.4
11
35.0
37
32.4
12
34.8
38
32.4
13
34.8
39
32.2
14
34.7
40
32-2
15
34.6
41
32.0
If.
34.4
42
32.0
17
34.3
43
31.9
18
34.2
44
31.8
19
34.2
45
31.7
20
33.9
46
31.7
21
33.6
47
31.7
22
33.6
48
31.1
23
33.6
49
31.1
24
33.5
50
30.3
25
33.5
51
30.3
26
33.5
52
28.9
135
-------
where: Xi are the observations, with i ranging from 1 to N
N is the number of observations
N
£ is the operator "sum" meaning to add together all the values
i=l
of the variable following it (in this case X^) as i covers the
integers from 1 to N.
N
I x± = xt + x2 + x3 + ...+xN
In the above example (from Table 4.1), X^ = 35.8, X2 = 33.0, ...,X^ = X^ =
32.4;
N
y X^ = 35.8 + 33.0 + 33.6 + ...+ 35.8 + 32.4 = 1748.3; and so the mean,
i=l _
which is denoted X (read "X-bar") is:
N
_ I \
X = i=l 1748.3 =33.6
N 52
The mean, unlike the median, uses all the data, and is therefpre more
representative of the whole set. Unfortunately, it is affected by extreme
values. If in Table 4.2 the first observation is replaced by 396.0
the median is still 33.5, but the mean becomes:
X = 396.0 + 38.8 + 37.4 + ... + 28.9 = 2104.7 » 40.5
52 52
which is considerably greater than the former value of 33.6.
The mean, since it makes use of all the data, is the most often used
measure of the "center" of a data set•
136
-------
4.1.1.2 The Median
The median of a set of data is the observation in the middle, that is,
the number that is located such that half of the observations are less than
it and half are greater. To find the median of a set of observations, the
data must first be arranged in numerical order as in Table 4.2.
If N is the number of observations in the ordered data set (in this case,
N N
N, is 52), then the median is defined to be the mean of the th and — + 1st
observations if N is even (between the 26th and 27th here, which would be 33.5)
or the ^^-th observation if N is odd (i.e. with 15 ordered observations, the
median is the 8th value).
The median is a good measure of the location of the center of a set of
data because it is unaffected by extreme values (i.e. if the largest observa-
tion were 396.0 instead of 39.6, the median would still be 33.5>. It's fault
is that it doesn't make use of all the information contained in the data, but
rather uses only the relative sizes of the observations.
4.1.1.3 The Range
In addition to knowing where the "center" of a data set is, it is useful
to know how spread out the data set is. One indicator of the spread of a data
set is the range, which is defined as the difference between the largest and
the smallest values in the set. For example, in Table 4.2, the largest is
39.6 (#1) and the smallest is 28.9 (//52) and so the range is R = 39.6 - 28.9
= 10.7.
Like the median, the range is simple to compute, once the data are
arranged in decreasing or increasing order, but doesn't use all the data, and
therefore does not carry much information.
137
-------
4.1.1.4 The Variance
The variance, which could be called the average of the squares of the
deviations of the data from their mean, is another indicator of how spread
out the observations are. To find the variance, you subtract the mean from
each observation, square each of these differences, sum the squared terms,
then divide the sum by one less than the number of observations, or, in
symbols;
N - ?
2 I (X± - X)2
St = i^l
N - 1
Table 4.3 shows how this is done; i is the week and X. is the correspond-
ing concentration.
52 26 _ ? 52 _ 2
I (X. -X)2 = I (X - X) +5; (X. - X) - 67.00 + 151.11 = 218.11
i-1 i=l i=27
N _ 52
, I (X. - X)2 I (X - 33.6)
Variance = S~ = i=l 1 = i=l = 218.11 = 4.28
X N - 1 51 51
o
There is another formula for computing Sjj which will be given herfe without
an example:
? , 2 ~2
I (X/) - N (X )
sx -
N - 1
This formula says to square each observation and sum the squares. Then multi-
ply the square of the mean (found earlier) by the number of observations (N),
subtract this from the sum of squares just computed, then divide by N-l. This
formula involves fewer steps since there is only one subtraction, as opposed
138
-------
TABLE 4.3 COMPUTATION OF THE VARIANCE
i
X.
l
(Xj-X)
(x.-x)2
X
i
X
l
(X.-X)
(X.-X)2
1
35.8
2.2
4.84
27
31.1
-2.5
6.25
2
33.0
-0.6
0.36
28
33.6
0.0
0.00
3
33.6
0.0
0.00
29
28.9
-4.7
22.09
4
35.0
1.4
1.96
30
35.6
2.0
4.00
5
33.5
-0.1
0.01
31
32.9
-0.7
0.49
6
34.7
1.1
1.21
32
31.8
-1.8
3.24
7
33.6
0.0
0.00
33
37.4
3.8
34.44
8
36.9
3.3
10.89
34
32.0
-1.6
2.56
9
38.8
5.2
27.04
35
34.8
1.2
1.44
10
35.5
1.9
3.61
36
31.7
-1.9
3.61
11
32.2
-1.4
1.96
37
32.7
-0.9
0.81
12
32.2
-1.4
1.96
38
36.0
2.4
5.76
13
33.3
-0.3
0.09
39
34.2
0.6
0.36
14
33.5
-0.1
0.01
40
30.3
-3.3
10.89
15
33.0
-0.6
0.36
41
39.6
6.0
36.00
16
33.1
-0.5
0.25
42
34.6
1.0
1.00
17
33.5
-0.1
0.01
43
31.7
-1.9
3.61
18
31.9
-1.7
2.89
44
30.3
-3.3
10.89
19
31.7
-1.9
3.61
45
34.4
0.8
0.64
20
32.4
-1.2
1.44
46
32.4
-1.2
1.44
21
34.8
1.2
1.44
47
31.1
-2.5
6.25
22
33.5
-0.1
0.01
48
36.5
2.9
8.41
23
33.9
0.3
0.09
49
33.2
-0.4
0.16
24
32.0
-1.6
2.56
50
34.3
0.7
0.49
25
34.2
0.6
0.36
51
35.8
2.2
4.84
26
33.4
-0.2
0.04
52
32.4
-1.2
1.44
26 _ 2
I (X. - X) = 67.00
i=I
I (X. - X)2= 151.11
i=27 1
-------
to N subtractions using the other method.
4.1.1.5 The Standard Deviation
The units of the variance are the square of the units of the mean (and
the original data), i.e. if the data are expressed in mg/l> the variance is
in mg^/l^. Because of this, the standard deviation, which is the square root
of the variance, is more commonly used as a measure of dispersion. In our ex-
ample the variance, Sx, is 4.28, and so the standard deviation is:
s = Jsl - ATI = 2.07.
X x
Since the data are expressed as mg/1, the standard deviation is also in mg/1.
The mean (X) and standard deviation (S ) are actually only estimates of
X
parameters known as the population mean (yx) and population standard deviation
(o ), which are discussed in Appendix A.
An interesting and useful fact about these two numbers is that in a nor-
mal population (which is discussed later and is a phenomenon which occurs
quite frequently), 68.3% of the observations will fall within u ^ a , 95.5%
X X
will be found within yx ~2ax, and 99.7% within ux * 3ax« Since X approxi-
mates vix and Sx approximates ox, these percentages will hold approximately
for X - Sx, X ± 2SX and X - 3SX>
4.1.1.6 Coefficient of Variation
This statistic provides a measure of the dispersion relative to the loca-
tion of the data set, so that the spread of the data in sets with different
means can be compared.
S
Coefficient of Variation - CV - ~
X
140
-------
4.1.1.7 The coefficient of Skewness
The coefficient of skewness is a measure of the degree of assymetry of
the data about its mean:
N
N I (X. - X)3
Coefficient of Skewness = k ~ i=l
(N-l) (N-2) Sx3
In our example, k = 52 (272.765) = .63(see Table 4.4)
51 (50) 8.870
A positive coefficient of skewness indicates high extreme values and,
as shown on page (4), leads to a mean greater than the median.
4.1.2 Harmonic Variations (2)
The use of the statistical concepts discussed so far depends on the
assumption that the data record is random. The identification and esti-
mation of the transient variations of a wastewater monitoring record is ex-
tremely important. It reduces the standard deviation, thereby making esti-
mators more reliable. The techniques used in identifying and evaluating
these components are trend removal and time series analysis.
4.1.2.1 Trend Removal
A trend in a wastewater monitoring record can usually be detected visu-
ally. Trends can be either linear (increasing or decreasing) or non-linear
(e.g. exponential or logarithmic). A trend may be defined as any harmonic
component whose period is longer than the record length. Trend removal is
an important step in data processing. If trends are not removed, large dis-
tortions can occur both in further data processing and in conclusions on the
probability distribution of the measured parameter. In many wastewater moni-
toring programs the evaluation or detection of the trend is a desired result
141
-------
TABLE 4.4 COMPUTATION 0? THE COEFFICIENT OF SKRWNESS
i
- X
(X - X)3
i
1
- X
-------
in itself.
The best method for evaluating a trend is the least-square procedure
which can be used if a random or harmonic component is superimposed on a linear
trend such that'
X(t) = x(t) + x'(t)
where X(t) is the data record expressed as a function of time. (in the
Table 4.1 data, t is expressed in weeks, and so X(l) = X-^ = 35.8,
X(2) = X2 = 33.0,..., X(52) = X52 = 32.4 ,
X(t) is the linear trend.
X'(t) is the random component.
In this case, the trend can be approximated by a straight line of the form
*
X(t) = a + bt.
The coefficients a and b are computed by regression analysis and can be
proven to be:

2 (2N + 1) I X(t)
-
6 I t X(t)
a =
t=l

t=l

N
(N - 1)

12 I t X(t)
-
6(N +1) I X(t)
b =
L t-1

t=l
(At) (N) (N-l)(N+l)
where N is the number of samples.
t is the sampling interval.
After removal of this linear trend, X(t), the new time series isi
X(t) = X'(t) - X(t) - (a +bt)
143
-------
Table 4.5 contains a data set with a linear trend. There follows an exam-
ple of identifying and removing this trend.
It can be seen in Figure 4.1 that the data contain an upward trend and
also a harmonic component. The trend is identified by finding X(t) = a + bt.
a = 2 ((2) (34) + D(lOA.l) - 6 (2133.1) =1.4
(34) (33)
b = 12(2133.1) - 6(35)(104.1) = 0.1
(1) (34) (33) (35)
Therefore the line X(t) = 1.4 + 0.1 t. Since a linear trend is removed by
subtraction, the new time series is:
X(t) = X(t) -(a + bt) = X(t) - (1.4 + O.lt)
Table 4.6 lists the adjusted data and Figure 4.2 shows the series after the
removal of the trend.
4.1.2.2 Time Series Analysis
Time series analysis is the most powerful method of analyzing a large
volume of data, such as continuous records with high frequency of data
acquisition. Since large amounts of data are required, time series analysis
should not be used for short surveys or low frequency monitoring when limited
amounts of data are available, or if part of the record is missing.
Auto^-Covariance and Auto-Correlation Analysis
These functions describe the dependence of the values of the data at
one time on the values at another time. An estimate of the auto-covariance
144
-------
TABLE 4.5 DATA SET WITH LINEAR TREND
Data
nomnut-atl nn

Data
computation
t
X(t)
tx(t)
t
X(t)
tx(t)
1
1.0
1.0
18
3.8
68.4
2
1.4
2.8
19
3.7
70.3
3
1.9
5.7
20
4,d
96.0
4
2.0
8.0
21
4.4
92.4
5
2.5
12,5
22
4.3
94.6
6
2.4
14.4
23
4.6
105.8
7
2.5
17.5
24
4.3
103.2
8
2.8
22.4
25
4.4
110.0
9
2.1
18.9
26
4.3
111.8
10
2.2
22.0
27
3.9
105.3
11
1.7
18.7
28
4.3
120.4
12
1.8
21.6
29
3.6
104.4
13
1.5
19.5
30
3.2
96.0
14
1.8
21.6
31
3.8
117.8
15
1.9
28.5
32
3.4
108.8
16
2.8
44.8
33
4.5
148.5
17
2.7
45.9
34
4.6
156.4

I t X(t) =
2139.5

I X(t) « 104
.90

t=l

TABLE
4.6 ADJUSTED DATA.SET
OF TABLE 4.5

Computation
Adjusted Data

Computation
Adjusted Data
t
x(t)
X(t)
t
x(t)
X(t)
1
1.5
-0.5
18
3.2
0.6
2
1.6
-0.2
19
3.3
0.4
3
1.7
0.2
20
3.4
1.4
4
1.8
0.2
21
3.5
0.9
5
1.9
0.6
22
3.6
0.7
6
2.0
0.4
23
3.7
0.9
7
2.1
0.4
24
3,8
0.5
8
2.2
-0.6
25
3.9
0.5
9
2.3
-0.2
26
4.0
0.3
10
2.4
-0.2
27
4.1
-0.2
11
2.5
-0.8
28
4.2
0.1
12
2.6
-0.8
29
4.3
-0.7
13
2.7
-1.2
30
4.4
-1.2
14
2.8
-1.0
31
4.5
-0.7
15
2.9
-1.0
32
4.6
-1.2
16
3.0
-0.2
33
4.7
-0.2
17
3.1
-0.4
34
4.8
-0.2
145
-------
X(t)
5.0
x(t)
4.0
X(t) 3.0
2.0
1.0
I 2 J 4 5 6 I S 9 10 II 12 13 14 15 Ift 17 II 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
t (weeks)
Figure 4.1 Series before removal of trend
-------
2.0
1.0
x(t) 0
-1.0
2.0
I 2 3 4 S t 1 8 9 10 II 12 I) I* IS 16 IT It 19 20 21 22 23 24 2) 26 27 28 29 30 31 32 33 34
t (weeks)
Figure 4.2 Series after removal of trend
-------
function (acvf) between two observations X(t) and X(t + u), separated by a lag
time, u, is given by;
N-~u _ _
cfu} = I I «x(t) - X)(X(t + u) - X)}
CW N t=i
where N is the number of observations in the record,
X is the mean of the N observationsc
c(u) is called the sample auto-covariance function of the time
series, and is a function of the lag time, u.
Using the data in Table 4.5, we find that
and so, for u = 4,
x = A21-U 3.1
34
c(4) = 34
1
- 34
. . .-I"
I (X(t) - 3.1] (X(t + 4) - 3.1)
(1.0 - 3.1) (2.5 -3.1) + (1.4 - 3.1)(2.4 -3.1)+
(3.2 - 3.1) " 3.1)
- ~ (22.19)= .65
34
Since the acvf is a measure of the dependence between values spaced a certain
distance apart, looking at c(u) for various values of u will give information
on this dependence. For example, in this set of data, c(4) = 0.65, c(l) =
1.06, and c(10) = 0.12. This shows that the auto-correlation decreases with
increased lag time and is quite small when u reaches 10.
Notice that, except for N rather than N-l in the denomenator, c(0) =
Sx , the sample variance. This says that the variance is just the auto-
148
-------
covariance of each observation with 'tself.
When the acvf is normalized by dividing by c(0), it becomes the sample
auto-correlation function (acf)
r(u) = cJiil
K } c(0)
which is an indicator of how much one observation is dependent on those around
it. It gives a visual picture (when plotted against the lag, u, between
points) of how the dependence damps out as the lag increases. This graph is
called the auto-correlogram. Figure 4.3 is the auto-correlogram for the data
in Table 4.5. The fact that the curve in Figure 4.1 Is somewhat like a sine
wave is reflected in the auto-correlation, which begins to show negative
correlation after u passess 11. For purely random data the acf would approach
zero as u increases. A periodic component in the record would result in a
periodic auto-correlogram with period similar to that of the original data.
The principal application of the acf is to establish the influence of values
at any time over values at a later time. It provides a tool for detecting
deterministic data which might be masked in a random background.
Variance Spectral Analysis
In the analysis of time series, the "variance spectrum" more commonly
known as "power spectrum" is a basic tool for determining the mechanism gen-
erating an observed series. The power spectrum is just the Fourier Transform
of the theoretical acvf, -y(u), and so is defined, as a function of frequency f,
by r (f) y(u) cos (2irfu) du
where y(u) ¦ E{ (X(t) -y) (X(t+u) -y )}.
(The expectation operator E is defined in Appendix A).
149
-------
2.0
1.0
c(u)
lag time, u (weeks)
Figure 4.3 Auto-correlogram
-------
By definition (cf. Section A.1.1) variance is a measure of the dispersion
of observations about their mean value. This dispersion may result from pure-
ly random fluctuations (noise) in the observed data as well as from determin-
istic (non random) fluctuations. These deterministic fluctuations may be the
result of trends (e.g. linear trends) as well as periodic components in the
record. Spectral analysis is a useful tool for the analysis of data records in
in which both random and deterministic fluctuations may be present as it
allows its user to separate these two types of fluctuations.
In spectral analysis of a data record (which ideally but not necesssarily
should be continuous) the power spectrum r(f) of the series is plotted against
frequency f. Figure 4.4 shows six hypothetical data records and their corres-
ponding power spectra.
Figure 4.4a shows a record on which all observed values are equal and
therefore equal to their mean. Their variance is zero and therefore the power
spectrum plot is zero at all frequencies.
Figure 4.4b shows a record with a linear trend. The variance in this
record is a result of the time dependent linear trend in the record. There
is no random or periodic dispersion about the mean, consequently all of
the variance (or power) spectrum is concentrated at the zero frequency.
Figure 4.4c shows a record exhibiting periodic harmonic fluctua-
tions with frequency £^. The variance in this record is a result of the
harmonic fluctuation of frequency ft about the mean. All the power spectrum
is concentrated at the fl frequency.
Figure 4.4d shows a record with purely random fluctuations (white noise)
about a constant mean value. The variance in this record is a result of these
purely random fluctuations. There is no trend or harmonic fluctuations. The
151
-------
power spectrum is uniformly distributed over all frequencies. Figure 4.4e
shows a record with purely random fluctuations superimposed on a linear trend.
Its power spectrum is the superposition of power spectra corresponding to
the linear trend record and the purely random record.
Figure 4.4f shows a record with purely random fluctuation superimposed
on harmonic variations of frequency fx • It's power spectrum is the super-
position of power spectra corresponding to the harmonic record and the purely
random record.
The power spectra depicted in Figure 4.4 are theoretical power spectra.
They are based on infinite continuous records. In practice records will be
of finite duration and discrete. When evaluating the power spectrum of a
finite duration record it is assumed that this finite record repeats itself
periodically at intervals of length equal to the duration of the given record.
When dealing with discrete records (or digital treatments of a continu-
ous record) the frequency of data acquistion is a frequency foreign to the
phenomenon under study which would appear in the power spectrum. These two
practical limitations on spectral analysis lead to distortion in the low
and high frequency regions of the spectrum known as "aliasing". The highest
frequency which can be resolved from a discrete record with sampling interval
At is the "Nyquist frequency"
f max 1
2 A t
Furthermore, the length of the record should be large enough to resolve
its periodic fluctuations. For example, spectral analysis of the portion AB
of the fecord in Figure 4.4f would lead to a power spectrum similar to that of
Figure 4.4e and not the actual power spectrum of Figure 4.4f.
152
-------
Besides, purely random fluctuations (white noise) are never met in prac-
tical applications where the theoretical power spectra depicted in Figures
4.4d-f would not be obtained. Rather, spectra similar to those of Figures
4.5 a-c would be encountered. In Figure 4.5a the absence of any significant
peak in the spectrum reflects the absence of any significant periodicity in
the record of 4.4d. In Figure 4.5b the presence of a significant peak at the
low frequency end of the spectrum is indicative of the linear trend in the
record of Figure 4.4e. The significant peak at frequency f\ on the spectrum
of Figure 4.5c reflects the presence of the harmonic component of frequency
fi in the record of Figure 4.4f.
The following rules of thumb should be followed when using spectral
analysis:
- The length of the record should be at least 10 times as long as the long-
est period of interest (e.g. 10 years of data if the annual period is the
longest period of interest).
- The sampling interval should be less than half the shortest period of inter-
est (which would then have the Nyquist frequency). A sampling interval of
one third or one fourth the length of the shortest period of interest is
recommended.
In view of the length of record and the high frequency of data acquisition
necessary for accurate spectral analysis, an overwhelming number of calcula-
tions will have to be carried out and treatment of the data on a digital com-
puter is necessary. In carrying out spectral analysis with the aid of a dig-
ital computer, the practitioner may wish to write his own program or take ad-
vantage of existing programs such as BMD02T, BMD03T, BMD04T, or SPECTRA which
are described in references (16,17),
153
-------
X(t)
t, time.
QJ
a
cn
u
0)
»
o
p-
frequency
(a) Constant record
X(t)
time
frequency
(b) Linear trend record
X(C)
wave length
corresponding toi
time
frequency
(c) Harmonic record
Figure 4.4 Typical theoretical power spectra for several records
154
-------
X(t)

time
(d) Purely random fluctuations
frequency
X(t)

0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0-08
0.09
0.0 0.5000 0.4960 0.4920 0.48R0 0.4840 0.4801 0.4761 0.4721 0.4681 0.4641
0.1 0.4602 0.4562 0.4522 0.44ft! 0.4441 0.4404 0.4164 0.4125 0.4268 0.4247
0.2 0,4207 0,41 fiH 0.4129 0./,»')() 0.4052 0.4011 (1.1974 0.1916 0.3897 0.1859
I). '! I). 1821 0. ')7H> 0.17*5 <>,3707 I). ¦Hilt* J), 16 )2 (l. 'l'.'M (l.ViV 0.3520 H.14«:i
0.4 0.3440 0.3409 0.3172 0.3130 O.'DOO 0.3264 0,1228 0.3192 0.31J6 0.3121
0.5
0.6
0.7
0.8
0.9
0.3085
0.2743
0.2420
0.2119
0.1841
0,1050
0.2709
0.2389
0.2090
0.1814
0.1011
0.2676
0.2358
0.2061
0.1788
0.29HI
0.2643
0.2327
0.2033
0.1762
0.2.946
0.2611
0.2296
0.2005
0.1736
0.2912
0.2578
0.2266
0.1977
0.1711
0.2B77
0.2546
0.2236
0.1949
0.1685
0.2841
0.2514
0.2206
0.1922
0.1660
0.2810
0.2483
0.2177
0.1894
0.1635
0.2776
0.2451
0.2148
0.1867
0.1611
1.0 0-1587 0.1562 0.1539 0.1515 0.1492 0.1469 0.1446 0.1423 0.1401 0.1379
1.1 0.1157 0.1335 0,1314 0.1292 0.1271 0.1251 0.1230 0.1210 0.1190 0.1170
1.2 0.1151 0.1131 0.1112 0.1093 0.1075 0.1056 0.1018 0.1020 0.1001 0.0985
1.3 0.0968 0.0951 0.0914 0.0918 0.0901 0.0H85 0.0869 0.0853 0.0838 0.0823
1.4 0.0808 0.0793 0.0778 0.0764 0.0749 0.0735 0.0721 0.0708 0.0694 0.0681
1.5
1.6
1.7
1.8
1.9
0.0668
0.0584
0.0446
0.0359
0.0287
0.0665
0.0537
0.0436
0.0351
0.0281
0,0643
0.0526
0.0427
0.0344
0.0274
0.0630
0.0516
0.0418
0.0336
0.0268
0.0618
0.0505
0.0409
0.0329
0.0262
0.0606
0.0495
0.0401
0.0322
0.0256
0.0594
0.0485
0.0392
0.0314
0.0250
0.0582
0.0475
0.0384
0.0307
0.0244
0.0571
0.0465
0.0375
0.0301
0.0239
0.0559
0.0455
0.0367
0.0294
0.0233
2.0 0.022H 0.0222 (1.0217 0.11212 0.0207 0.0202 0.0197
2.1 0.0179 0.0174 0.0170 0.0166 0.0162 0.0158 0.0154
2.2 0.0139 0.0116 0.0132 0.0129 0.0125 0.0122 0.0U9
2.3 0.0107 0.0104 0.0102 0.00990 0,00964 0.00939 0.00914
2.4 0.00820 0.00798 0.00776 0.00755 0.00734 0.00714 0.00695
2.5 0.00621 0.00604 0.00587 0.00570 0.00554 0.00539 0.00523
2.6 0.00466 0.00453 0.00440 0.00427 0.00415 0.00402 0.00391
2.7 0.00347 0.00336 0.00326 0.00317 0.00307 0.00298 0.00289
2.8 0.00256 0.00248 0.00240 0.00233 0.00226 0.00219 0.00212
2.9 0.00187 0.00181 0.00175 0.00169 0.00164 0.00159 0.00154
0.0192 0.0188 0.01H3
0.0150 0.0146 0.0143
0.0116 0.0U1 0.0110
0.00889 0.00866 0.00842
0.00676 0.00657 0.00639
0.00508 0.00494 0.00480
0.00379 0.00368 0.00357
0.00280 0.00272 0.00264
0.00205 0.00199 0.00193
0.00149 0.00144 0.00139
167
-------
Turning to a table of areas under the standard normal curve (Table 4.8), we
find that the area to the left of -2 ( which is the same as the area to the
right of 2) is 0.0228, which is, then, the probability that X is less than or
equal to 286 (written P (X £ 286) = 0.0228). This means that if a large number
of samples of size 100 are taken from this population, approximately 2.3% of
them will have sample means of 286 or less.
If the population parameters (yx and cr^) are unknown,we can use this
method to make inferences about them. Suppose we know that 70 and that
the mean of a random sample (X) with N=100 is 318. Can we reasonably assume
that the population mean, y , is 300?
x
We are testing to see if y = 300. We call this hypothesized value y
x o
and the hypothesis that u = p is called H ( the null hypothesis). We write
x o o
the null hypothesis:
H : y = y (in this case, H : y = 300)
o x o ox
Our alternative is that y^ ^ 300. This is called the alternative hypothesis
and is denoted H, : y ^ y
1 x o
z /i ® 2 ** i-96 and so the critical region for the rejection of H is
a/2 .025 0
{z:z<-1.96 or z>1.96}.
The test statistic we use is
318-300
z " Oj " W/T55 " 2'57'
In this case, z = 2.57>1.96 = z^/2» and s0 we reJect H0 and conclude that the
distribution from which the sample was taken has a mean other than 300.
168
-------
If both yx and ax are unknown, we can't use- the z-statistic as above
(since its calculation involves c?x), and so we use the statistic
X"Uo
SxMl
t =
which has a Student's t-distribution.
Example:
If, in the above example, the standard deviation is unknown, but we find
the sample standard deviation to be 70.5, then our test statistic is
318-300
t = , = 2.55.
70.5//100
Using Table 4.9, which gives values of tn.a (which is the number such that
p(tn>tn.a)= a, where tn has a Student's t-distribution with n degrees of free-
dom), we look under a=.025 (since we are using a two-tailed test at the .05
level of significance) and n=99. (The degrees of freedom, n, is just N-l) .
Since n=99 does not appear in the table, we take the number approximately 2/3
of the way between n=60 and n «= 120. Our test statistic, t=2.55, is greater
than that for n=60, and so we reject H : y a 300 in favor of H : y # 300.
J o x i x
Example:
If we take a different sample (of size 121 this time) from the same popu-
lation and compute a sample mean of 310 and a sample standard deviation 70.2,
we find
x-y0 310-300
t = ¦ =» 1.56.
sx/rti 70.2//T21
169
-------
TABLE 4.9 PERCENTAGE POINTS OF STUDENT t-DISTRIBUTION (18)
Value of t ;a such that p(t>t ;a)=a
n v n '
"Area == a

tn»
a

a
n
0.10
0.050
0.025
0.010
0.005
1
3.078
6.314
12.706
31.821
63.657
2
1.886
2.920
4.303
6.965
9.925
3
1.638
2.353
3.182
4.541
5.841
4
1.533
2.132
2.776
3.747
4.604
5
1.476
2.015
2.571
3.365
4.032
6
1.440
1.943
2.447
3.143
3.707
7
1.415
1.895
2.365
2.998
3.499
8
1.397
1.860
2.306
2.896
3.355
9
1.383
1.833
2.262
2.821
3.250
10
1.372
1.812
2.228
2.764
3.169
11
1.363
1.796
2.201
2.718
3.106
12
1.356
1.782
2.179
2.681
3.055
13
1.350
1.771
2.160
2.650
3.012
14
1.345
1.761
2.145
2.624
2.977
15
1.341
1.753
2.131
2.602
2.947
16
1.337
1.746
2.120
2.583
2.921
17
1.333
1.740
2.110
2.567
2.898
18
1.330
1.734
2.101
2.552
2.878
19
1.328
1.729
2.093
2.539
2.861
20
1.325
1.725
2.086
2.528
2.845
21
1.323
1.721
2.080
2.518
2.831
22
1.321
1.717
2.074
2.508
2.819
23
1.319
1.714
2.069
2.500
2.S07
24
1.318
1.711
2.064
2.492
2.797
25
1.316
1.708
2.060
2.485
2.787
26
1.315
1.706
2.056
2.479
2.779
27
1.314
1.703
2.052
2.473
2.771
28
1.313
1.701
2.048
2.467
2.763
29
1.311
1.699
2.045
2.462
2.756
30
1.310
1.697
2.042
2.457
2.750
40
1.303
1.684
2.021
2.423
2.704
60
1.296
1.671
2.000
2.390
2.660
120
1.289
1.658
1.980
2.358
2.617
a «= 0.995, 0.990, 0.975, 0.950, and 0.900 /oilow
from tn;l-0t = -tn»a
170
-------
Looking in Table 4.9 for a=.025 and n»N-l=120, we find that t
120;.025
= 1.980,
and so our test statistic does not fall in the critical region. Therefore,
we cannot reject HQ.
4.1.5 Confidence 1n te rva1s
4.1.5.1 Confidence Intervals for the Mean (1, 3)
In the example above, we tested a hypothesis about the population mean.
In a similar way we could construct an interval within which we would consider
a hypothesis for the mean tenable and outside of which such a hypothesis
would be untenable. We call this interval a confidence interval and its end-
points confidence limits.
In the previous example, a population mean of 300 was found to be con-
sistent with the computed statistics. Suppose we tested H0:MX=295 against
Hi:Hx^295. Then
Somewhere between 295 and 300 is a mean such that the computed t is equal to
tn;a,and this number is the lower confidence limit for the population mean.
Likewise, if we test H0:yx=322 against H-^,.)Jx:/322, t = -1.88, which is greater
than ~ti20;.025' an£* so ^o is acceptable. But a test of HQ:ux=323 yields
t=-2.03<-1.98 and so we reject Hc. Therefore, the upper confidence limit is
betwnen 322 and 323. The actual confidence limits for yx can be computed
from
3? -Vo
which is greater than t120;.025 = 1.980, and so we reject HQ in favor of H-^
,a/2 -
171
-------
which, in this example yield
yL —<^20. 025 * ~ x) --(1.98 x ZM -310) = 297.4
lai J , . U
-------
has a distribution (i.e. a chi-square distribution with n degrees of
freedom).
Using a chi-square table (Table 4.10), we can construct a 95% confidence
interval for a2 as follows:
Find X2N.a/2« which *s the number that
P (Y(j2>—- I (X.-y)2) - .95, and so
x N; .025 i»l x N; ,975 l56X
we have a 95% confidence interval for o2.
Example:
Suppose we have 10 observations from a normal distribution with mean 0
(i.e. N-10, yx=0). Then (Xi-vJx)2=(X^-0)2»X^2 and so
10 10
£ (X^-px)2 » £ ^i2, i*et this sum be equal to 113. A5.
i"i i»i
173
-------
Table 4.10 PERCENTAGE POINTS OF CHI-SQUARE DISTRIBUTION
(18)
Value of ;pj f such that Proh(^#l > tJ » a
*!
*
n
0.995
0990
0.975
0.950
O.'XK)
010
005
0.025
0010
0005
1
0.000039
000016
0 00098
0.0039
00158
2.71
3 84
5 02
6.63
7.88
2
00100
0 0201
0.0506
0.103
0.211
4.6)
5.99
7.38
9.21
10.60
3
0.0717
0.115
0 216
0.352
0 584
6.25
7.81
935
11.34
12 84
4
0.207
0.297
0 484
0711
1.06
7.78
9 49
11 14
13 28
14.86
5
0412
0 554
0 831
1.15
1 61
9.24
It 07
12 83
15.09
16.75
6
0676
0.872
1.24
1 64
2 20
1064
12 59
1445
16.81
18.55
7
0.989
1.24
1.69
2.17
2.83
12 02
14.07
16.01
1848
20.28
8
1.34
1.65
2.18
2.73
3 49
13.36
15.51
17 53
20.09
21.96
9
1.73
209
2.70
3.33
4.17
14.68
16 92
19 02
21.67
23.59
10
2.16
2.56
3.25
3.94
4.87
15 99
1831
20 48
23.21
25.19
11
2.60
305
3.82
4.57
5.58
17.28
19.68
21.92
24.73
26.76
12
3.07
3.57
4 40
5.23
6.30
18.55
21.03
23 34
26 22
28.30
13
3.57
4.H
SGI
5.89
104
19.M
22.36
24.74
27.69
2**2
14
4.07
4 66
563
6.57
7.79
21 06
23 68
26.12
29.14
31 32
15
4 60
5.23
626
7.26
8.55
22.31
25.00
27.49
30 58
32 80
16
5.14
5.81
6.91
7.96
9.31
23.54
26.30
28 85
32.00
34.27
17
5.70
6.41
7.56
8.67
10.08
24.77
27.59
30.19
33.41
35.72
18
6.26
701
8 23
9.39
10.86
25.99
28.87
31.53
34 81
37.16
19
6.84
7.63
8.91
10.12
11 65
27.20
30 14
32 85
36 19
38 58
20
7 43
8.26
9.59
10.85
12.44
28 41
31 41
34.17
37.57
40 00
21
8 03
8 90
10.28
11.59
13.24
29.62
32.67
35 48
3893
41.40
22
8 64
9.54
1098
12.34
14.04
3081
33.92
36 78
40 29
42 80
23
926
1020
11 69
1309
14.85
32.01
35.17
3808
41 64
44.18
24
9 89
10.86
1240
13.85
15.66
33 20
3642
39.36
42.98
45.56
23
10.52
11.52
13.12
14.61
16.47
34 38
37.65
40.65
44.31
46.93
26
U 16
12.20
13.84
15.38
17.29
35.S6
38 88
41.92
45.64
48.29
27
11.81
12.88
1457
16.15
18.11
36.74
40.11
43.19
46 96
49.64
24
12.46
13.56
15.31
16.93
18.94
37.92
41.34
44 46
48 28
50.99
29
13.12
14 26
1605
17.71
19.77
39.09
42.56
45 72
49.59
52.34
30
13.79
14.95
16.79
18.49
20.60
40.26
43,77
46.98
50 89
53.67
40
20.71
22.16
24.43
26.51
29.05
51 81
55.76
59.34
63 69
66.77
60
35.53
37.48
40 48
43.19
46.46
74 40
79 08
83 30
88.38
91.95
120
83.85
86 92
91.58
95.70
100.62
140.23
146.57
152.21
158 95
163.65
For 120, xj;i » «I I - r- + f, /5- I where «, li the desired percentage point for a standardised normal
distribution. L *" V J
174
-------
A 95% confidence interval (a/2 = .025) for a2 is then
("113.45, 113.45s
¦ (5.5, 34.9)
[ 20.48 3.25
4.1.5.2.2 Confidence Interval for if y is Unknown
It is also true (by the definition of S2) that
NS2
4
has a chi-square distribution with N-l degrees of freedom (X^_^) and so if yx is
unknown, we can find a confidence interval using S2, the sample variance.
Suppose in the above example, Sx = 3.6. We turn to Table 4.10 again and
find X2Q; .025 and X2Q; .975, which are 2.70 and 9.02, and so the interval is
NS2 NS2
. (9*12.96, 9x12.96) . 43.2)
X2^;.975 X29;.025 19.02 2.70
The confidence limits for the standard deviation are found by taking the square
root of those for the variance.
4.1.5.3 Relative Error of the Standard Deviation
n
Sx

where Q is the width of the confidence interval of the standard deviation
v2
N-l;l-ot/2 is defined above
(1-a) x 100% is the level of confidence of the interval.
175
-------
4.2 DETERMINATION OF NUMBER OF SAMPLES (6)
The number of samples necessary to reasonably characterize a water or
wastewater is determined after collecting some background data on the concen-
tration and variance of the concentration of the parameters under consider-
ation. These values can be estimated; however, estimation will decrease the
confidence in the results. Two techniques can be used to calculate the
number of samples, one based on the allowed sample variability, the other on
the accuracy of the sample mean. Each will give a desired value of N, the
number of samples needed, with the larger value to be chosen for application.
4.2.1 Determining Number of Samples from a Constraint on the Variability
To apply this method, the following information is needed:
1. Allowable error of the standard deviation (—-)
sx
2. Confidence level required (1-a)
Therefore, for this situation, one is estimating that the value of a
certain variable will occur within a specific interval. A normal distribution
of the data around the mean is assumed. The data should be checked for nor-
mality as in Section 4.1.3.1.
Example:
Determine the number of samples required from a wastewater monitoring
program such that the estimated standard deviation will be within 25% of its
true value (i.e. ± 12.5%) at a confidence level of 98%.
Here a = 1-.98-.02 and " 0.25. From Figure 4.9, the value of
bx bx
=0.25 is found on the vertical axis and a horizontal line is followed until
the curve for a » .02 is met. Then a vertical line is dropped to the hori-
zontal axis to find the number of observations needed (N"180 in this case).
176
-------
10.0
Graphical Solution to Equation
N-l; l-ct/2
0.25
0.05.
0.01
10
2
50
100
5
500 1000
N - 180
Sample Size, N
Figure 4.9 Determination of the number of samples based
on the required accuracy of extreme values
177
-------
4.2.2 Determining Number of Samples from a Constraint on the Mean Value
To apply this method, the following information is required:
1. Confidence level required (1-ot)
Sx
2. Coefficient of variation of the source to be sampled (CV= —)
X
3. The required accuracy of the sample mean.
A double iteration procedure is recommended, especially if the number
of samples is found to be small (N<30). For this calculation a normal dis-
tribution is assumed.
The first iteration uses the formula
CV x Z ,0 2
^ 0/100 ;
where CV =
X
0 is the allowed deviation of the sample mean from the true mean, ex-
pressed as a percent of the true mean.
zcx/2 is found in Table 4.8.
For the second iteration use
CV * ta/2;N-l,
N* <—iHoo—1
where tct/2;N-l is found in Table 4.9.
Example:
For a wastewater stream with an average daily concentration of 120
mg/1 BOD and a standard deviation of 32 mg/1, determine the number of daily
samples which would provide an accuracy of the daily averages within 5%.
178
-------
3 = 5
X = 120
Sx = 32
S 32
CV =-3=120 = 0-27
X
If we choose a = .05 (95% confidence level), then = ^.025 -*-s f°UI1^ i-n
Table 4.8 to be 1.96.
2
Step 1 N - (°-2^-96) = 109.3 = 110 samples
Step 2 Using N=110, find ta/2;N-l = t.025;109 in Table 4.9
to be approximately 1.983 (using linear interpolation),
so N -(°-275^983)2 = 114.6 - 115 samples.
If the accuracies of both the standard deviation and the mean are used
as criteria, choose the larger of the two values of N. In the example above,
Ng = 180 and N^" = 115, so 180 daily samples should be taken.
*•3 DETERMINING SAMPLE FREQUENCY
Although it requires the use of a digital computer, spectral analysis is
the method that should be used for determining sampling frequency because of
it's accuracy and the simplicity of the final interpretation.
4.3.1 Determination of the Sampling Frequency from Power Spectra (7,8,9,10)
It is imperative that a good set of historical data be available for anal-
ysis. Ideally, these data should be continuous. Practically, they should be
taken at a frequency that is higher than the highest expected frequency of
179
-------
harmonic variation components of the record. For example, if daily trends are
to be analyzed, hourly samples may be called for. At any rate, the length and
sampling interval of the record should satisfy the rules of thumb governing
spectral analysis (cf. Section 4.1.2,2). Ideally, in a discrete record, there
should be no missing points. Interpolation may be used if a few data points
are missing, when these are widely scattered on the record. Interpolated data
should account for no more than five percent of the total data.
The following examples illustrate the use of spectral analysis in the
determination of sampling frequency.
Example 1: The wastewater influent for the city of Racine, Wisconsin, was sam-
pled hourly in the summer of 1974 and TOC analyzed. The record is shown in
Figure 4.10. The average and variance were calculated to be 70.56 mg/£ and
1262.07 mg2/£2 respectively. Determine the optimal sampling frequency for
this plant.
The power spectrum corresponding to the record of Figure 4.10 ie obtain-
ed as depicted in Figure 4.11. This power spectrum exhibits a significant
peak at the 1/24 hr frequency and a less significant peak at 1/8 hr, Most
of the variability on the data occurs in the frequency band from 1/48 hr to
1/16 hri Since the last significant peak in the spectrum occurs at the
1/8 hr frequency, the sampling frequency which should be at least two times
the frequency of the last significant peak, corresponding to,the Nyquist
frequency* should be at least 1/4 hr. In order to clearly show the 1/8 hr
variability a sampling interval of 3 hrs. or even 2 hrs. is recommended in
accordance with the second rule of thumb. Note that this example, the first
rule of thumb stated in Section 4.1.2.2 is violated as the length of the record
in Figure 4.10 (7 days) Is less than 10 times the longest period of interest
180
-------
x = 70.56 mg/1
= 1262.07
Sun Mon Tues Wed Thur
Time
Fri
Sat
Figure 4.10 Time record of TOC of municipal wastewater at
Racine, Wisconsin
n
X!

at
4-1
¦H
4J
CO
w
§
1-1
o
0)
a.
«
n
a)
w w
h h ID 9) (I CO (0
X X ^ U U U
X X X X X
CM O
r-( r-< 03 I— vQ iTl
CO
M
X
to
co
H
,c
es
Figure 4.11
Frequency 1/hour
Power spectrum of TOC concentration of municipal waste-
water at Racine, Wisconsin
181
-------
(1 day). However the peak at the 1/24 hr frequency is so significant that it
cannot be explained by aliasing distortion alone.
Example 2: The power spectra of wastewater variation corresponding to two
typical types of industrial discharges are shown in Figures 4.12 and 4.13.
Determine the optimal sampling frequency.
The spectrum of Figure 4.12 exhibits two strong peaks in the frequency
band from 1/16 hr to 1/5 hr. This spectrum is typical for industrial plants
working 24 hours a day, seven days a week, with three shifts a day. Note the
absence of peaks on the low frequency region reflecting the absence of trend
in the record which would then appear to be stationary. Inasmuch as the last
significant peak occurs between the 1/6 hr and 1/5 hr frequency, a sampling
frequency of 1/2 hr is recommended (i.e. 2 x 1/4 hr).
The spectrum of Figure 4.13 displays a strong peak at the 1/24 hr fre-
quency and less significant peaks at the 1/12 hr and 1/6 hr frequencies.
This spectrum is typical for industrial plants working with one daily shift.
Here again, the absence of peaks in the low frequency region of the spectrum
is an indication of the stationariness of the record. In order to clearly
exhibit the 1/6 hr frequency component of the data a sampling interval of 2
hours is recommended in accordance with the second rule of thumb.
4.4 DETERMINATION OF PARAMETERS TO MONITOR
The decision as to which parameters to monitor is critical, since it is
not possible to monitor all parameters. There are two statistical methods to
help with this decision if prior regulations do not exist. The decision
variable for the first method is the probability of exceeding a standard and
the second is the correlation between parameters.
182
-------
(J 0.25 0.J0 0.15 0.20 0.25
Frequency (1/hour)
Figure 4.12 Power spectrum of industrial
plant discharge, Case 1
ro
U3
O
0.25
0.15
0.20
0.10
0.05
0
Frequency(1/hour)
Figure 4.13 Power spectrum of industrial
plant discharge, Case 2
183
-------
4.4.1 Probability of Exceeding a Standard
This method requires knowledge of:
1. The mean, u, or sample mean, X,
2. The standard deviation, o, or sample S.D., Sx,
3. The standard> X_, not to be exceeded for that parameter.
b
The probability of exceeding the standard is:
P (x > Xs) = P (z > Zj = a
Xs - X
where Za = —5— .
x
After computing Za, the probability, a, can be found in Table 4.8. Parameters
with the largest value of a have the highest sampling priority.
Example 1:
The effluent standard for an industry was determined to be 100 mg/1 of
Cl~. A wastewater quality survey has shown that the mean concentration of
chlorides was 75 mg/1 and the S.D. was 18 mg/1.
To determine the probability of the standard being exceeded:
Xc, - X 100 - 75 , ™
1. Determine Za = —2 = — = 1.39
Sx 18
2. Find a from Table 4.8 such that Za = 1.39. The value is
.0823, or 8.23%.
Often effluent standards will be specified for several parameters. Then
the parameters can be ranked in descending order of their probability of ex-
ceeding the standard. The priority of sampling will be in the same order.
Table 4.11 is an example of how this is done.
Example 2:
The standard for another parameter is 4 parts per million. The average
in the past was found to be 7ppm, with a S.D. of 2 ppm.
184
-------
Here XS = 4
X = 7
Sx = 2
Because of symmetry, p(z < -Za) = P'fz > Za) , and so, since Za = -1.5 in this
case, we look up +1.5 in the table, finding a = .0668. Since we want
P (z>-Za), we use the fact that p(z>-Za) = l-p(z<-Za) = 1-a. So the probabil-
ity of exceeding the standard is 1-a = 1-.0668 = .9332, or about 93.3%.
4.4.2 Correlation Between Measured Parameters (11)
Ideally, all important water quality parameters should be monitored,
but this is usually not economically feasible, so a method is needed for
deciding which parameters to omit. This is done by checking the closeness of
correlation among parameters of interest. It is known that a correlation
exists between many water quality parameters such as
B0D5 and TOC
COD and TOC
Chlorides and Conductivity
Total Dissolved Solids and Conductivity
Suspended Solids and Turbidity
Acidity, Alkalinity and pH
Hardness, Calcium and Magnesium
Hardness and Alkalinity
If a strong correlation exists between two or more parameters, the monitoring
of one parameter may be discontinued or monitored at a reduced frequency.
185
-------
In order to apply the technique, the following must be available:
1, A data record for all parameters of interest
2. A computer program for calculating correlation coefficients.
The relationship between two parameters X and Y can be linear or non-linear
(such as exponential, logarithmic, etc.)- If a non-linear relationship
exists, attempt to linearize the relationship, e.g. by using logarithms
of the values of X and Y, or some other functional approximation. Then
linear regression analysis provides a linear approximation of the form
A A
Y = a+bX. The coefficient of correlation, R^Y» will then be a measure of the
closeness of fit. The coefficient of correlation is determined from the
equation
N „ __
I (Xi-X) (Yi-Y)
i=l
R.
Ti
N N
I (Xi-X)2 I (Y.-Y)2
i=l i=l
Numerous computer package subroutines are available for the above analysis.
The hypothesis that a relationship exists between X and Y can be tested
at a given level of significance a (where 1-a is the confidence that the
hypothesis is true). If the obtained coefficient of correlation is such that
|RXy 1 51 Rc> where Rc is the minimal correlation coefficient, which can be
found in Table 4.12, the null hypothesis (that the correlation is zero) is
rejected.
If a pair of parameters has a correlation coefficient significantly
greater than the value from the table, one parameter in the pair is eligible
for elimination from or reduction of monitoring. The decision on which
186
-------
parameter should be eliminated will be based on the cost of data acquisition
and the priority of the parameter.
Example:
A wastewater system was surveyed for an extended period of time. As a
result of the survey, 25 sets of wastewater quality data were gathered. Each
set contained data on pH, TOC, COD, BOD, TKN, phosphorus, conductivity,
total dissolved solids, suspended solids, turbidity, lead, mercury, iron,
copper, alkalinity, acidity, hardness, calcium, magnesium, coliform bacteria,
fecal coliform and chlorides.
1. Determine the sampling priority of each parameter.
2. Determine which parameter measurements can be eliminated
or reduced.
First we find the probability that a parameter will exceed its standard.
This will determine the sampling priority of the standard.
The correlation analysis of the 22 parameters in Table 4.11 was per-
formed by a computer, using the formula given previously. From Table 4.12,
it was determined that
, 0.388 for a = .05
Rc. = {
0.496 for a = .01.
Table 4.13 shows the results of the analysis.
Sampling for total dissolved solids (TDS) has the highest priority, but,
because of the high correlation between TDS and conductivity, one of these
analyses can be eliminated. Total coliforms have the second highest priority,
but since the correlation between total and fecal coliforms is high, analyz-
ing for fecal coliforms is not necessary. The high correlation among BOD,
COD and TOC makes it possible to eliminate or reduce one or two of them.
187
-------
TABLE 4.11 SAMPLING PRIORITIES OF PARAMETERS FOR A TYPICAL WASTEWATER
Parameter
Water Quality
Mean,X
Standard
Z
P(X >xs)
Sampling

Standard, Xc

Deviation,S

Priority
pH
6.5 - 8.0
7.8
0.4
0.50
0.308
5
TOC
None
31
7.9
-
0
16 - 22
COD
70
60
11
0.91
0.181
7
BOD
30
20
8
1.25
0.125
9-10
TKN
5
3.5
1.5
1.00
0.158
8
Phosphates
1
0.5
0.2
2.50
0.006
15
Conductivity
None
320
80
-
0
16 - 22
Total dissolved

solids
500
491
125
0.072
0.472
1
Suspended Solids
30
28
5
0.40
0.34
4
Turbidity
20
19
3
0.33
0.37
3
Lead
5
3
1.0
2.0
0.0228
14
Mercury
5
2.5
1.5
1.67
0.047
13
Iron
10
7.8
1.9
1.16
0.123
11
Copper
7
0.8
0.15
1.33
0.0918
12
Alkalinity
None
-
-
-
0
16 - 22
Acidity
None
-
-
-
0
16 - 22
Calcium
None
-
-
-
0
16 - 22
Hardness
None
-
-
-
0
16 - 22
Magnesium
None
-
-
-
0
16 - 22
Total coliforms
100
81
65
0.29
0.386
2
Fecal coliforms
10
5
64
1.25
0.125
9-10
Chlorides
200
156
59
0.90
0.134
6
-------
TABLE 4.12 VALUES OF CORRELATION COEFFICIENT, p, FOR
TWO LEVELS OF SIGNIFICANCE (12)
Degrees of Freedom
Percent Level
of Significance, a
n = N - 1
Five
One
1
0.997
1.000
2
0.950
0.990
3
0.878
0.959
4
0.811
0.917
5
0.754
0.874
6
0.707
0.834
7
0.666
0.798
8
0.632
0.765
9
0.602
0.735
10
0.576
0.708
11
0.553
0.684
12
0.532
0.661
13
0.514
0.641
14
0.497
0.623
15
0.482
0.606
16
0.468
0.590
17
0.456
0.575
18
0.444
0.561
19
0.433
0.549
20
0.423
0.537
21
0.413
0.526
22
0.404
0.515
23
0.396
0.505
24
0.388
0.496
25
0.381
0.487
30
0.349
0.449
35
0.325
0.418
40
0.304
0.393
45
0.288
0.372
50
0.273
0.354
60
0.250
0.325
70
0.232
0.302
80
0.217
0.283
90
0.205
0.267
100
0.195
0.254
125
0.174
0.228
150
0.159
0.208
200
0.138
0.181
300
0.113
0.148
400
0.098
0.128
500
0.088
0.115
189
-------
TABLE 4.13 MATRIX OF CORRELATION COEFFICENTS
Parameter pH IPC COD """5 TEN P Ccci TPS SS T Pb Bg Fe Cm JLl't. Ac Ca Bard Hg Te
VO
O
pa
—

TOC
0
—

000
0
0.8
—

BOD5
0
0.68
0.63
—

IKK
0
0
0.15
0.18
—
Phosp
0
0
0.18
Q.2t
0.69
Conduct
0
0.30
0.41
0.35
0.33
TDS
0
0-25
0.35
0.48
0.41
SS
0
0.25
0.40
0.38
0.25
Turb
0
0.4
0.51
0.33
0.18
Pb
0.1a
0
0
0
0
Hg
0
0
0
0
0
Fe
0.1
0
0
0
0
Ca
0
0
O
0
0
Alk
0.6
0
0
0
0
Acid
0.6
0
0
0
0
Ca
0
0
0
0
0
Hard
0.1
0
0
0
0
Mg
0
0
0
0
0
T. Coll
0
0.31
0.3S
0.38
0
F. Coll
0
0.10
0.18
0.21
0
Chlor
0
0
0
0
0
0.17
. .
0.20
0-91
0.75
0.10
0.68
0.13
0
0.28
0
0.30
0
O.il
0
0.30
0
0.35
0
0.20
0
0.31
0
0.61
0
O.iO
0
0
0
0
0
0.53
0.18
—

0.59
0.89
—

0.31
0.18
0.15
—

0-23
0.25
0.31
0.70
—

0.39
0.58
0.61
0.18
0.23
—

0.23
0.31
0.25
0.69
0-59
0.41
—

0.41
0
0
0
0
0
0
—

0.15
0
0
0
0
0
0
0.-9
—

0.35
0
0
0
0
0
0
0.65
0
—

0.63
0
0
0
0
0
0
0.61
0.18
0.3S
—
0-31
0
0
0
0
0
0
0. Id
0
0.35
0.18 —
0
0.12
O.U
0
0
0
0
0
0
0
0 0
0
0.11
0.08
0
0
0
0
0
0
0
0 0
0.83
0
0
0
0
0
0
0
0
0
0 0
0.79 —
0 0 —
0 - no engineering relevance; assused no telazin.
-------
Testing for turbidity could also replace that for suspended solids. It is
also possible to eliminate at least one analysis from the group hardness,
coliform and alkalinity. Metals have relatively low priority and so at
least one of them can be reduced. Thus, the following streamlined program
is feasible:
Parameter Priority of Sampling
pH
high

TOC or COD
high

BOD
reduced

TKN
high

Phosphates
reduced

Conductivity
high

Suspended Solids or Turbidity
high

Lead
reduced
or
not
necessary
Mercury
reduced
or
not
necessary
Iron
reduced

Copper
reduced
or
not
necessary
Alkalinity
reduced

Hardness
reduced

Total Collforms
high

Fecal Coliforms
reduced
or
not
necessary
4.5 IN-PLANT SAMPLING AND NETWORK MONITORING
If the sampling locations have not been predetermined, there are sys-
tematic methods of determining the location of sampling points. However,
191
-------
these methods are only tools to aid sampling personnel and do not replace pro-
fessional judgment and experience.
4.5.1 Segmentation - Priority Technique
This technique can be applied to any large flowing network including
an industrial plant collection system, a municipal sewerage system, or even
a watershed network. To apply this technique the following information must
be known:
1. The mass flow rate of the parameter of interest, (Qwj Cwj).
2. The range of variation of the parameter input,
Pj =
-------
vO
U>
Plant 1
Storm
Water

Plant 2
e-
Plant 3
Water
Intake
€
13

Sanitary Waste
k8
14
Plant
15
16
Figure 4.14 Segmentation of a wastewater system
-------
such as the influent and/or effluent of a treatment plant. Using the corre-
lation analysis between the monitored segment and other upstream and down-
stream segments, it is possible to identify segments with low correlation to
the monitored segment. A second consideration should be the worth of the data
measured at the segment. For example, if the magnitude of a measured para-
meter and its variability are insignificant when related to other segments,
the segment will have a low priority for monitoring.
4.5.1.1 First Priority Sampling Points
The location of at least one sampling point is strictly determined iy the
basic objectives of a monitoring program, i.e. protection of the environment.
This objective requires that a sampling point be located just before a waste-
water discharge to a receiving water body. If the industry has several
wastewater outfalls, a sampling point should be located downstream from the
last outfall. In the case that the monitoring point is located in the re-
ceiving water body, an upstream station to monitor the upstream water quality
and quantity is necessary. This will allow the effect of the wastewater dis-
charge on the receiving water body to be clearly identified. If the water
intake for the industry is situated on the same water body, the upstream sam-
pling point can be conveniently located at the water intake.
4.5.1.2 Second Priority Sampling Points
Other important objectives of a sampling program can be to monitor the
quality of raw wastewater and to evaluate the efficiency of a treatment pro-
cess. Thus, a location for a second priority sampling point would normally
be at the influent to a treatment plant.
For small and middle-sized wastewater systems, sampling at the first
and second priority sampling points should be sufficient to meet most of the
194
-------
objectives and requirements established by regulatory agencies.
4.5.1.3 Third Priority Sampling Points
The location of additional sampling points may be necessary for large
wastewater systems with many inputs. Their purpose is to provide additional
information or warning. In this case, the method of segmenting the waste-
water system and determining sampling priorities for each segment can be of
use in establishing additional sampling points. Segmentation of a waste-
water system is accomplished by isolating the locations which substantially
modify the waste stream conditions. These locations include junctions of
wastewater streams, treatment units, wastewater overflow, flow dividers,
storm and cooling water inflows, and storage reservoirs. The following out-
lines a method of segmentation.
1. It is best to represent the wastewater system by a linear
graph technique. Such a graph consists of nodes or
junctions and branches or lines. All wastewater inputs
will enter the system through the nodes, and the nodes
also separate branches with different characteristics.
A branch is considered as a segment with uniform geometric,
hydraulic, and transform characteristics. The following
depicts the classification of some typical elements of
a wastewater system.
Nodes - manholes, changes of slope, changes in conduit
diameter, flow dividers, junctions of sewers and channels,
outfalls, influents and effluents to treatment steps, etc.
Branches - conduits, channels, treatment steps, bypasses,
adjacent receiving water bodies, storage reservoirs,
holding ponds, etc.
195
-------
For the industrial water/wastewater system of Figure 4.15, a
linear graph representation is shown in Figure 4.16.
In segmenting the system, each node should be uniquely numbered.
Wastewater input to each node should be characterized by the
range of variation
P.. = (QwjCwj) max - (QwjCwj) min,
which is, basically, the range of waste loads to the node j.
The units of will be g/sec if the flow Qw is expressed in
m3/sec and concentration in mg/1. It might be convenient
also to know the approximate frequency of fluctuations of
the input P^. A node table such as is shown in Table 4.14
should be prepared.
Each branch is identified by a double subscript AB, where A
is the number of the upstream node and B is the number of the
downstream node.
Coefficients of transformation and a._ should be assigned
Ad Ad
for each branch. The coefficient of transformation B^g describes
roughly how the variability of the wastewater is reduced in
this segment. In most cases can be determined approximately
from the geometry of the segment and treatment parameters. The
coefficient describes how the correlation is reduced in the
segment. The following values of the coefficients are recommended
* Short sewers and channels B. a „
AB AB
1.0 1.0
* Plug flow treatment steps,
long sewers and channels exp(-KT) 0.9 to 1.0
with decay
196
-------
* Completely mixed treatment steps
with short detention time
1-Etr/100
0.85 to 0.95
U«l/f)
* Completely mixed treatment steps
with long detention time t ,
(t»l/f) (2 (1+Kt) tf) (2tf)
* Storage and equalization i 1
reservoirs and holding ponds (2tf)~^ (21f) ^
with no decay
where
K = decay coefficients in the segment (in units of day~l)
t = detention time in the segment (in days)
f = frequency of fluctuations of waste inputs
Etr = treatment efficiency (in percent)
4. Determine and approximate ranges of wastewater quality variations
for each segment. This can be done by starting at the most up-
stream nodes containing wastewater inputs and moving downstream,
by the buffering capacity of segments and by new wastewater in-
puts (such as process discharges) in downstream nodes.
Figure A.17 Illustrates how this procedure is accomplished. JK is
the most upstream node containing a wastewater input and would there-
fore be the starting point. The range of wastewater variability will be
J
r ¦ P.
JK J
where r^ is the wastewater quality variation range in segment JK
downstream from J. Above, the downstream node K the variation range
is determined by
197
-------
INT^ K.E
l
X
PROCESS
4
PROCESS
5
FLOATATION
[AND TOXICITY
REMOVAL
T~®'
SQUALIZATIOS
AND
STORAGE
BIOLOGICAL
TREATMENT
PLANT
SLUDGE
HANDLING
SLUDGE
SOLIDS
EFFLUENT
MONITORING
<£>
00
SANITARY
WASTE
PROCESS

PROCESS

PROCESS
I

'
r

r
SEDIMENTATION

HEAVY
AND

METALS
NEUTRALIZATION

REMOVAL
SANITARY
WASTE
Figure 4.15 An industrial water/wastewater system
-------
00 Segment
I
0=0 Segment
[Tj—.Primary Sampling Segment J
Wastewater Input [S [-^-Secondary Sampling Segment
Node
- Sewer —j
or Channel
- Treatment!
Intake
Figure 4.16 Linear graph representation of the system
-------
At a node the variability range can be changed by wastewater inputs
to the node and by other upstream branches entering the node. For
a case where more than one input enters a node, the following re-
lationship (propogation of errors) can be used to compute the
variability range:
rAB ¦ q +1
-------
Variability Range
Monitoring ( N J
Point
JK
KL
LM
MN
JK
K
rJK
'J
J
rJK * &JK
K K
t * r
KL JK
L K
rKL = rKL * KL
" t
-------
0.23
S " Monitored segment
First, second, third
priority segments
for monitoring
0.26
P ¦ 0.63
100.72
P - 0.30
28.5
p - O.63
0.26
50.36
67.5
0.10
>3.3
0.52
p • 0.70
0.33
67.5
0.75
0.82
78.73
96.121
MonItored
segment
Figure *».18 Correlograph for segments
-------
in the correlation coefficient can be roughly estimated as follows:
In a Branch - multiply p by the coefficient a
B . B
In a Node - multiply p by the ratio *AB/rRC
where B is the node under consideration, AB is the branch located
farther away from the monitored segment, and BC is the branch lo-
cated closer to the monitored segment.
6. Additional sampling points should be located at the segment where,
theoretically, the correlation with the monitored point ends.
Since the correlation influence of both points extends both down-
stream and upstream, there will be an overlap such that each sam-
pling point will have an influence of r = /Rc, where Rc is the
critical point found in Table 4.12. If the number of samples is
not known, a value of Rc between 0.25 and 0.30 will give a good
estimate.
7. If there are several segments to be monitored, i.e. one or more
segments have a correlation level less than Rc, the priority can
be determined according to the magnitude of the variability range
r^j for the segment i j. The segment with the highest rjj will
have the highest priority.
8. Once a new sample location is established, the procedure is re-
peated to find the next sampling location.
9. The entire procedure should be repeated for each important
parameter.
Example:
Determine the locations of sampling points for the wastewater system
given in Figure 4.15. The analysis will be based on the COD information re-
presenting the organic load to the system.
203
-------
Step 1 - Divide the system into segments using the linear graph rep-
resentation, as in Figure 4.16.
Step 2 - Locate a first priority sampling point (P) at the effluent
channel (segment 1-2). Locate second priority sampling points
(S) at the influent to the treatment plant (segment 4-5)
and in the receiving water body, upstream and downstream from
the waste discharge.
Step 3 - Estimate the variability range of the inputs to the system
(Table 4.14).
Step 4 - Estimate 3 and a for each segment (Table 4.15).
Step 5 - Estimate the variation range in each segment. Proceed upstream
from the most downstream segment (Table 4.16).
Step 6 - Estimate the coefficient of correlation between wastewater
variations in each segment and the nearest monitored segment,
i.e. to segment 4-5. Proceed from the monitored segment (where
R =¦ 1.0) and work upstream (Table 4.16 right portion). Each
segment is correlated to the segment immediately downstream
toward the monitored point. Developing a correlograph (Figure
4.IS) at this stage will aid in the decision process in Step
7.
Step 7 - Once the correlation coefficients are estimated, find those
where R
-------
TABLE 4.14 WASTEWATER LOADS'TO NODES
CONSTITUENT: COD
Node
Maximal Loading
g/sec
Minimal Loading
g/sec
Pj
1
0
0
0
2
0
0
0
3
0
0
0
4

5
0
0
0
6
10
1.2
8.8
7
0
0
0
8
30.0
6.0
24.0
9
0
0
0
10
0
0
0
11
175
100
75
12
0
0
0
13
0
0
0
14
66.0
17.0
49.0
15
109
21.0
88.0
16
0
0
0
17
0
0
0
18
42
23
19
19
121.50
93.0
28.5
Fluctuations of maximum and minimum at most nodes - 1/8 hrs"^
205
-------
TABLE 4.15 COEFFICIENTS OF VARIATION IN BRANCHES
Branch Description £ a
1-2 Effluent Channel 1.0 1.0
2-3 Activated Sludge Plant 0.1 0.4
3-4 Equalization Basin 0.2 0.2
4-5 Sewer 1.0 1.0
5-6 Sewer 1.0 1.0
6-7 Sewer 1.0 1.0
7-8 Sewer 1.0 1.0
8-9 Sewer 1.0 1.0
9-10 Neutralization Plant 0.9 0.9
10-11 Sewer 1.0 1.0
5-12 Sewer 1.0 1.0
12-13 Flotation Unit 0.5 0.5
13-14 Sewer 1.0 1.0
14-15 Sewer 1.0 1.0
7_16 Sewer 1.0 1.0
16-17 Sewer 1.0 1.0
17-18 Chemical Coagulation 0.7 0.7
16-19 Sewer 1.0 1.0
206
-------
TABLE 4.16 DETERMINATION OF THE SAMPLING PRIORITIES OF SEGMENTS
Stfaent
Cpstresr variation range
r„ - (Ir2 + rp2)3-5
Downstream w:s:in range
r4 " ru * 3
Correlarlca c:c::
at the dov3S£re£=: noce
Pu " Pd fd
iciezt Is the branch
at the upstream aode
p - p9
Priority ::r
tertiary -zczltcr:
16-19
28.5

28.5
0.33
* 28-51/31.45 - 0-30
0.30
T2
17-18
19.0
19
« 0.7 - 13.3

0.14
0.14 0.7 - 0.10
T3
16-17
13.3

13.3
0.33
* 13.3/31-45 « 0.14
0.14

7-16
(28.52 + 13-32)5-5 - 31.45

31.45
0.81
* 31.45/78.24 - 0.33
0.33

10-11
75

0.63
0.63

9-10
75
75
* 0.9 - 67.5

0.70
0.7 * 0.9 « 0.63

8-9
67.5

67.5
0.75
* 67.5/71.64 - 0.70
0.70

7-8
(67.52 + 242)'-5 - 71.64

71.64
0.81
* 71.64/78.24 - 0.75
0.75

6-7
(71.Si2 + 31-452)0*5 - 78.24

78.2i
0.82
* 78.24/78.73 * 0.81
0.81

5-6
(78.2i2 ~ 8.82)5-5 - 78.73

78.73
1.0
* 78.73/96.12 * 0.32
0.82

14-15
88.0

88.0
0.26
* 88/100.72 = 0.23
0.23

13-14
(88- + i92)"-5 - 100.72

100.72

0.26
0.26
XI
12-13
100.72
100.
.72 * 0.5 - 5C.3*

0.52
0.52 * 0.5 * 0.26

5-12
50.36

50.36
1.0 * 50.36/46.12 - 0.32
0.52

4-5
(78.73* + 50.362)--5 « 96.12

96.12

1.0
1.0
Initial
aoai trrins
-------
low correlation levels. Both the values of R and of r^j
should be examined for these segments, the requirements and
objectives of the program should be considered, and then pro-
fessional judgment must be exercised.
In this example, segments 17-18, 16-17 and 16-19 are
neighboring segments with low correlation levels. Looking at
the variability values, we see that segment 16-19 has the
highest value, indicating the great fluctuations in wastewater
quality. Therefore, of these three, segment 16-19 might have
the highest priority. Segments 14-15, 13-14 and 12-13 are
also neighboring segments with low correlation levels. Seg-
ment 13-14 has the greatest variability and would therefore
be chosen. Since its variability is much higher than that of
segment 16-19, it would have the highest overall priority.
At this stage, correlation and variability values can be re-
calculated to see if monitoring at these points would satisfy
the program requirements. If not; the procedure should be
repeated.
4.5.2 Probability of Exceeding a Standard (13)
In locating sampling points in a receiving water body, the probability
of exceeding a receiving water standard should be considered. For all con-
servative substances and all nonconservative substances except oxygen and
possibly temperature and nitrates, the critical section would be located im-
mediately downstream from the outfall. The section with the highest probabil-
ity of violating the dissolved oxygen standard will be further downstream near
the "sag point." The location of the critical point can be approximately
208
-------
evaluated as follows:
The probability that the dissolved oxygen standard will be exceeded is
p(cds) - P(z>z£
Ds - P(x)
S(x)
)
which can be found in Table 4.9, where
C is the dissolved oxygen concentration
C3 is the dissolved oxygen standard
D is the oxygen deficit
Ds is the maximum allowable oxygen deficit
D(x)
KjLp
k2-*i
exp(=^) - expf^.)
+ D0 exp
which is the average oxygen deficit at distance x from the outfall
S(x) » x SLq x U is the standard deviation at distance x. (13, 14,
Lo is the average BOD discharge
S^0 is the S.D. of the BOD discharge
is the coefficient of deoxygenation
K2 is the coefficient of re-aeration
Dq is the initial oxygen deficit
U is the stream velocity
To find a maximal p(c
-------
that Zs = (ds-D(x)J/S(x) is a minimum. This can be accomplished by
finding the location x at which D(x)/S(x) is a maximum (and so P(D(x)>Dg)
is a maximum. The distance x can be found by plotting D(x)/S(x) against
x for given K-^, Dq, Lq and U, and then finding the x value corres-
ponding to the highest value of D(x)/S(x).
210
-------
4.6 REFERENCES
1. Haber, Audrey and Richard P. Runyon
General Statistics. Reading, Mass.; Addison-Wesley, 1969.
2. Bendat, J. S. and A. G. Piersol
Random Data: Analysis & Measurement Procedures
New York, Wiley-Interscience, 1971.
3. Hogg, Robert V. and Allen T. Craig
Introduction to Mathematical Statistics, 3rd Edition
London, The Macmillan Company, 1970.
4. Foster, H. A.
Theoretical Frequency Curves and Their Application to Engineering
Problems, Trans. ASCE Paper, 1532, p. 142-173, 1924,
5. Associated Water & Air Resources Engineers, Inc.
Handbook for Industrial Wastewater Monitoring
U.S. EPA Technology Transfer, 8-8 to 8-12, August, 1973.
6. Montgomery, H. A. C. and I. C. Hart
The Design of Sampling Programmes for Rivers and Effluents
Water Pollution Control (London, England), 73/. 77-98, 1974.
7. Drobny, N. L.
Monitoring for Effective Environmental Management
Proc. ASCE National Water Resources Engineering Meeting.
Atlanta, Georgia. January 24-28, 1972.
8. Gunnerson, C. G.
Optimizing Sampling Intervals
Proc. IBM Scientific Computing Symposium, Water and Air Resources
Management. White Plains, New York, 1968.
9. Sparr, T. M. and P. J. Schaezler
Spectral Analysis Techniques for Evaluating Historical Water Quality
Records (Presented at International Seminar and Exposition on Water
Resources Instrumentation, Chicago, June 4-6, 1974)
10. Wastler, T. A.
Application of Spectral Analysis to Stream and Estuary Field Studies
U.S. Department of HEW, Cincinnati, Ohio, p. 27, November, 1963.
11. Kaester, R. L., J. J. Cairns, and J. S. Crossman
Redundancy in Data from Stream Surveys
Water Research. 13: 637-642, August 1974.
211
-------
12. Fisher, R. A. and F. Yates
Statistical Tables for Biological, Agricultural and Medical Research
London, Oliver and Boyd, 1949.
13. Chamberlain, S. G., C. V. Beckers, G. P. Grimsrad, and R. D. Shull
Quantitative Methods for Preliminary Design of Water Quality Sur-
veillance Systems.
Water Resources Bulletin, 10: 199-219, April, 1974.
14. Thomann, R. V.
Variability of Waste Treatment Plant Performance
Journal ASCE Sanitary Division, 96: 819-837, January, 1970.
15. Eckenfelder, W. W.
Water Quality Engineering for Practicing Engineers
New York, Barnes and Noble, 1970.
16. Dixon, W. J.
BMD Biomedical Computer Programs
University of California Press, Berkley, CA 1973
17. Barr, A.J, j.h. Goodnight, J.P. Sail, J.T. Helwig
SPECTRA procedure, in A user's guide to SAS76
SAS Institue, Inc., Raleigh, N.C. 1976
18. Owen, Donald, B. Handbook of Statistical Tables,
Addison-Wesley Company, Reading, Mass.
212
-------
CHAPTER 5
SAMPLING MUNICIPAL WASTEWATERS
5.1 BACKGROUND
Municipal wastewater consisting of the spent waters from a community is
treated by chemical, physical, and/or biological means prior to discharge to
surface waters. Up to three stages of treatment are commonly used at munici-
pal treatment plants (1): primary (screening, sedimentation), secondary (acti-
vated sludge, trickling filter, etc.)» and tertiary (physical/chemical treat-
ment) . The wastewater characteristics vary with the size and habits of the
community, the type of collection system(combined or separate), the amount of
infiltration and the type of industrial discharges.
5.2 OBJECTIVES OF SAMPLING PROGRAMS
5.2.1 Regulatory
Sampling of municipal wastewaters is required by regulatory agencies for
the NPDES permit program. The location of sampling points, frequency, sample
type, etc. are specified in the NPDES permit. At the time of NPDES permit
modifications, incorporate the recommendations of Compliance Sampling
Inspection and use the statistical analysis of self monitoring data as a
rationale to specify the permit requirements.
5.2.2 Process Control
In addition, sampling is performed at municipal treatment plants for
process control purposes. This monitoring provides a check on the efficiency
of the process allowing the operator to make adjustments to optimize the
213
-------
Process efficiency.
5.2,3 Research and Development
The special needs of a research project will dictate the sampling pro-
gram. Hence each project must he considered individually and no general
guidelines can be given.
5.3 FREQUENCY OF SAMPLING
5.3.1 Established by Regulation
Follow the frequency requirements indicated in the permit issued by the
regulatory agencies.
5.3.2 Use of Statistics
Apply spectral analysis techniques (Section 4.3.2) to establish the op-
timum frequency. If the data required for this technique is not available:
1. Conduct a week-long survey collecting hourly samples.(For combined
municipal-industrial wastewaters choose a week of high industrial
production.)
2. Determine if any unusual industrial or community discharge occurred
during the sampling period (e.g. an extensive spill or extremely
heavy rainstorm) which may invalidate the data and necessitate a
repeat of the survey.
After data collection, the analysis of data should be performed as out-
lined in Section 4.3.2.
5.3.3 Surveillance Purposes
A poll of EPA Surveillance and Analysis Labs indicated a general con-
currence that for normally variable domestic wastewaters a minimum of 8 evenly
spaced grab samples collected over a 24 hour period, repeated for a minimum of
3 weekdays, will result in a fair estimate of water chemistry characteristics
(2).
214
-------
5.3.4 Other Considerations
Follow interim sampling frequencies prior to the generation of data for
statistical analysis. Frequencies appear in Tables 5.1 (3) and 5.2 (4).
5.4 LOCATION OF SAMPLING POINTS
Collect the sample at the location(s) specified in the permit. At these
locations collect the sample in the center of the channel at 0.4 to 0.6 depth
where the flow is turbulent, well mixed, and the settling of solids is minimal.
Sampling at 0.4 to 0.6 depth will avoid skimming of the water surface or
dragging the channel bottom.
For BOD analysis, it is recommended that samples be collected prior to
the disinfection step (5). For BOD and suspended solids, samples of plant
influent and effluent must be collected in order to calculate the removal of
these constituents. The sampling of wastewater for immiscible liquids, such
as oil and grease, requires special attention and no specific rule can be given
for selection of the most representative site for collection of an oil and
grease sample because of wide range of conditions encountered in the field.
In such cases, experience of the sampling team should be the guide in the
selection of the most representative site. (2)
5.4.1 Influent
Influent wastewaters are preferably sampled at points of highly turbulent
flow in order to insure good mixing; however, in many instances the desired
location is not accessible. Preferable raw waste sampling points are (6):
a. the upflow siphon following a comminutor (in absence of grit chamber);
b. the upflow distribution box following pumping from main plant wet well;
c. aerated grit chamber;
215
-------

TABLE 5.1 PROCESS TESTING GUIDE3
(3)
Process
Test
Frequency

PRETREATMENT

Grit
Volatile Solids
Daily
Removal
Total Solids
Daily

Moisture Content
Daily

PRIMARY TREATMENT

Primary
Settleable Solids
Daily
Sediment at ion
PH
Daily

Total Sulfides
Daily

Biochemical Oxygen Demand
Weekly

Suspended Solids
Weekly

Chemical Oxygen Demand
Weekly

Dissolved Oxygen
Weekly

Grease
Weekly

SECONDARY TREATMENT

Activated
Suspended Solids
Daily
Sludge
Dissolved Oxygen
Daily

Volatile Suspended Solids
Weekly

Turbidity
Daily
Trickling
Suspended Solids
Daily
Filter
Dissolved Oxygen
Daily
Oxidation
Dissolved Oxygen
Daily
Ponds
Total Sulfides
Daily

Total Organic Carbon
Weekly

Total Phosphorus
Weekly

Settleable Solids
Daily

pH
Daily

Total Sulfides
Daily
Final
Biochemical Oxygen Demand
Weekly
Sedimentation
Suspended Solids
Weekly

Chemical Oxygen Demand
Weekly

Dissolved Oxygen
Weekly

Turbidity
Daily

MB AS
Weekly
3
This is a minimum sampling guide, and is subject to change with plant site,
complexity of operation, and problems encountered.
(continued)
216
-------
TABLE 5.1 (continued)
Process Test Frequency

DISINFECTION

Chlorination
Chlorine Residual
Daily

MPN Coliform
Weekly

SOLIDS HANDLING

Thickening
Suspended Solids
Daily
Volatile Solids
Daily
Digestion
Total Solids
Weekly
Volatile Solids
Weekly

PH
Daily

Gas Analysis
Weekly

Alkalinity
Weekly

Volatile Acid
Weekly
Centrifuging
Suspended Solids
When in Operation
Volatile Solids
When in Operation
Vacuum Filters
Sludge Filterability
When in Operation

Suspended Solids
When in Operation

Volatile Solids
When in Operation
Incineration
Ash Analysis
When in Operation

ADVANCED TREATMENT

Chemical
Jar Test
Weekly
Coagulation &
Phosphorus Analysis
Weekly
Flocculation
Apparent Density
Weekly
Activated
COD
Weekly
Carbon
TOC
Weekly
Recarbonation
PH
Weekly
Ammonia
Ammonia Nitrogen
Weekly
Stripping
pR
Weekly
Filters
Suspended Solids
Daily

Turbidity
Daily
Microscreen
Suspended Solids
Daily

Chemical Oxygen Demand
Weekly
a This is a minimum sampling guide, and is subject to change with plant site,
complexity of operation, and problems encountered.
217
-------
TABLE 5.2 RECOMMENDED MINIMUM SAMPLING PROGRAMS FOR MUNICIPAL
WASTEWATER TREATMENT PROCESSES (4)
N5
00

u
•H

u
•

W
O
I
e

£
u

e
o

w
c:
0
o
c
o
u
V

4
u
«
i

*2
o

e
u
«
u
tft

0
«

Ow
"e
a.

CJ
o
M
3

u
3
00
3
o

O
U

oo
fl.
>»

>»

«
o

•o
00

1
9
o
9
Oi
u

6
J
X

a
c
«8
s

«
u

u
SM

o
u
s

e
hi
•H
O

OQ
b

Si
3

•5
u

*r4

O
«
o

o
o
B
b

>
O

u
O
u
p-

u
H
o
<
«
as

9
CO
3
6
3C

£
ft.

3
-------
TABLE 5.2 (continued)
O
sl r2
O
S F
u
fr
a
hi
flu
S F
3
S F
H
S F
<
S
S F
S F
S F
6
2
S F
S F
-S
s
S F
S F
m -2
S F
Micro Analysis
Ortho-F
Sludge Vol.
Dl9. S
IB AS
Metals
Plant Flow
G 2/W
C 3/W
C 2/M
C 2/K
R
G 3/0
0 l/D
Chlor. UsU.
C 1/D
B
Collfora
c t/w
Fecal Coliforv -
c i/v
Alk. C 2/W
C l/W
Jar Test
c l/W
Hardness
C l/W
G 3/D
C 3/W
C 2/W
C 1/W
G 3/U
C 1/D
C l/V
C 2/W G l/t>
C 1/D
1* S • type of saaple
2. F ¦ frequency
G - Grab
C * 24 hour composite
D * Day
W - Week
n - Mooch
R ¦ Record continuously
Mo " Mooltor continuously
-------
d. flume throat; and
e. pump wet well.
In all cases, samples should be collected upstream from recirculated
plant supernatant and sludges.
5.4.2 Ef fluent
Collect effluent samples at the most representative site downstream from
all entering waste streams. When manually compositing effluent samples
according to flow where no flow measuring device exists, use the influent
flow measurement without any correction for time lag. The error in influent
and effluent flow measurement is insignificant except in those cases where
extremely large volumes of water are impounded (such as in reservoirs) as a
result of influent surges coupled with highly restrictive effluent discharge
(7).
5.4.3 Pond Sampling
Composite samples should be employed even for ponds with long detention
times because of the tendency of lagoons to short circuit. If dye studies or
past experience indicate a homogeneous discharge, a grab sample may be taken as
representative of the waste stream.
5.4.4 In-Plant Location
Apply the statistical technique outlined in Section 4.5 to determine
in-plant sampling locations. In addition to these locations, sample all other
unit processes periodically or when the variability of a parameter adversely
affects the efficiency of a unit process,
5,5 NUMBER OF SAMPLES
Use one or more of the following methods to determine the number of
samples!
220
-------
1. Follow permit requirements by regulatory agencies.
2. Apply statistical methods in Section 4.2 to the data from the
preliminary survey.
3. Use the frequency data to establish number of samples (e.g. 1 sample
every 6 hours will establish 4 samples per day),
5.6 PARAMETERS TO MEASURE
The NPDES permit for each -municipal treatment plant spells out in detail
the effluent limitations and monitoring requirements for that particular plant.
For evaluating the plant performance, regardless of the size, these parameters:
Biochemical Oxygen Demand (5 day), solids, pH and flow should be monitored
routinely (8).
Secondary analyses include:
1. Fecal Coliform and Chlorine Residual
2. Temperature 8. Dissolved Solids
3. Dissolved Oxygen 9. Alkalinity
4. Total Solids 10. Metals
5.. Total Volatile Solids 11. COD
6. Nitrogen Series 12. Oil and Grease
7. Phosphorus 13. Organic Priority Pollutants
as required
Table 5.2 indicates the parameters to analyze the efficiency or the
effectiveness of the various unit processes. Changes are allowed to compen-
sate for specific plant conditions.
5.7 TYPE OF SAMPLE
Use composite samples for all overall monitoring (.6) and grab samples for
checking individual unit processes. Use one of the following types of com-
posite samples to properly estimate mass loading;
221
-------
1. Periodic, time constant, sample volume proportional to stream flow.
2. Periodic, sample volume constant, time proportional to stream flow
since the last sample.
Other composite types may be used if comparable results can be demon-
strated.
5.8 METHODS OF SAMPLING
Choose manual or automatic sampling depending on how the advantages and
disadvantages of the methods apply to the specific program. (Refer to Chap-
ter 2). Adequate care should be exercised in sample collection. Only trained
personnel should be entrusted the task of sample collection. Much of the un-
certainty regarding the collection of suspended solids can be minimized if
samples are collected at isokinetic conditions or at higher intake velocities.
5.8.1 Automatic Sampler
Automatic samplers for municipal wastewaters must be capable of collec-
ting representative suspended solids samples throughout the collection and
treatment system. While sampler selection will depend on site conditions,
the following guidelines are suggested:
1. For sampling raw wastewater and primary effluent, use a sampler
having an intake velocity greater than 0.76 ra/sec. (2.5 ft./sec.).
For sampling a final effluent with no visible solids, a sampler
having a lower intake velocity may be acceptable (2).
2. To determine the effectiveness of an automatic sampler to collect
suspended solids, statistically compare the suspended solids value
of the composite sample from the automatic sampler with the mean
value of the manual grab samples. The minimum compositing period
222
-------
should be six hours with a maximum individual sample frequency of
one hour (7). The ratio of the automatic sampler suspended solids
value to the manual grab suspended solids value varies throughout
the plant. For influent and primary effluent the acceptable ratio
is 1.6 - 2.0 and for the final effluent it is 0.9 - 1.3.(9)
5.9 VOLUME OF SAMPLE AND CONTAINER TYPE
The volume of sample obtained should he sufficient to perform all the
required analyses plus an additional amount to provide for any split samples
or repeat examinations. Although the volume of sample required depends on
the analyses to be performed, the amount required for a fairly complete analy-
sis is normally 7.57 liters (2 gallons) for each laboratory receiving a
sample. The laboratory receiving the sample should be consulted for any
specific volume requirements. Individual aliquot portions of a composite
sample should be at least 100 milliliters (0.21 pints) in order to minimize
sampler solids bias. Depending on the sampling frequency and sample volume,
the total composited sample should be at least 7.57 liters (2 gallons) (6).
Use a separate sterilized container for coliform analysis. See chapter 12
for organic collection methods. Collect chlorine residual or oil and grease
samples in a glass container. Plastic is acceptable for the other recommen-
ded analyses. Specific information for water quality or organic parameter
types is given in chapter 17.
5.10 PRESERVATION AND HANDLING THE SAMPLES
Follow the guidelines establishing test procedures for the Analysis of
Pollutants (AO CFR 136) as amended in Federal Register Vol. 41, No. 232,
December 1, 1976, and for organics, vol. 44, No. 233, December 3, 1979.
223
-------
Techniques indicated in Chapter 17 to collect and preserve the sample can
be used as a general guide.
5.11 FLOW MEASUREMENTS
The flow measurement technique selected should be in relation to the
sampling location, type of flow, etc. Follow the guidelines enumerated in
Chapter 3 on Flow Measurements. Primary and secondary flow measurement
devices should be calibrated prior to taking flow measurements.
224
-------
5.12 REFERENCES
1. Metcalf and Eddy, Inc. Wastewater Engineering. McGraw Hill, New York,
1972.
2. Harris, D. J. and W. J. Keefer, Wastewater Sampling Methodologies and Flow
Measurement Techniques. EPQ 907/9-74-005, U.S. Envirotvmental Protection
Agency, Region VII, 1974. 117 pp.
3. URS Research Co., Procedures for Evaluating Performance of Wastewater
Treatment Plants. PB 228 849/6, National Technical Information Service,
Springfield, Virginia.
4. Anon. Estimating Laboratory Needs for Municipal Wastewater Treatment
Facilities. PB 227 321/7. National Technical Information Service,
Springfield, Virginia.
5. Henderson, F. M. Open Channel Flow. MacMillan Co., New York, 1966.
6. Anon. NPDES Compliance Sampling Manual. U.S. Environmental Prdtection
Agency, Washington, D. C., June 1977.
7. Barth, E. F. U.S. EPA Inter-Office Memo dated August 22, 1975.
8. Water Pollution Control Federation Highlights. Vol. 12 H-I, April 1975.
9. Anon. Comparison of Manual (Grab) and Vacuum Type Automatic Sampling
Techniques on an Individual and Composite Sample Basis. EPA-330/1-74-001,
U.S. Environmental Protection Agency, Denver, Colorado, 1974. 29 pp.
225
-------
CHAPTER 6
SAMPLING INDUSTRIAL WASTEWATERS
6.1 BACKGROUND
Industrial wastewaters vary significantly in pollution characteristics.
This chapter presents general guidelines and considerations so that effective
sampling programs can be established for varied situations.
6.2. OBJECTIVES OF SAMPLING PROGRAMS
6.2.1 Regulatory
Sampling of industrial wastewaters is required by regulatory agencies
for the NPDES permit program. The location or sampling points, frequency,
sample type, etc. are specified in the NPDES permit. At the time of NPDES
permit modifications, incorporate the recommendations of Compliance Sampling
Inspection and use the statistical analysis of self monitoring data as a
rationale to specify the permit requirements.
6.2.2 Process Control
In addition, sampling is performed within the plant to monitor individ-
ual waste streams, as an indirect check on the process efficiencies, to
compute material balances, etc.
6.2.3 Research and Development
The special needs of a research and development project will dictate the
sampling program; such projects are summarized below (1);
1. To explore potential recovery from a given department or unit
226
-------
process, considering process modifications and studying the
economics thereof.
2. To define factors influencing character of wastes from a given
department or unit process,
3. To investigate and demonstrate variations in the character and
concentration of combined waste.
4. To establish a sound basis for the treatment of residual wastes.
Each such project must be considered individually and no general guide-
lines can be given.
6.3 FREQUENCY OF SAMPLING
6.3.1 Established by Regulation
Use permit requirements when compliance monitoring is the objective. If
the sampling frequency is not specified by regulation, sampling interval
should be less than one hour (3), and if data is available use the statistical
methods as a tool to determine the frequency of sampling.
6.3.2 Use of Statistics
Apply the statistics outlined in Section 4.3., to obtain frequency of
sampling whenever possible. Background data must be collected to determine
mean and variance. One of the following procedures can be used to obtain this
information (listed in order of preference) if it has not been previously
collected:
1. Conduct a week long preliminary survey consisting of the hourly
samples to characterize the system.
2. Conduct one 24 hour survey taking hourly samples (as outlined in
chapter 2). Analyze individual samples if batch dumps are suspected.
Any weekly pattern must be considered and samples taken on the day
227
-------
of the greatest variation of the parameters of interest.
3. Obtain data from a plant with the same type of industrial operation.
However, where processes differ, take samples to quantify the
variation.
After data collection, use production figures to determine extreme values,
assuming a linear operating relationship (which is not always the case).
6.3.3 Other Considerations
Consider variable plant operations when determining frequency:
1. Seasonal operation
2. Less than 24 hour per day operation
3. Special times during the day, week or month set aside for cleanup
4. Any combination of the above
When monitoring these types of operations, it is necesssary to sample
during normal working shifts in the season of productive operation. Figure
6.1 gives procedures for the various situations.
6.4 LOCATION OF SAMPLING POINTS
6.4.1 Effluent Monitoring
Regulatory permits establish effluent monitoring points within a plant.
The permit may specify only the total plant discharge or a specific discharge
from a certain operation or operations. Consult permits for th^se locations,
or use those recommendations for obtaining representative samples given in
chapter 2.
6.4.2 In-Plant Locations
To achieve process control or to design and implement in-plant pollution
control programs, in-plant sample locations are necessary, Use the following
procedures to determine the sampling locations:
228
-------
1« Become familiar with the plant processes and sources of wastes
from unit operations.
2. Ascertain the sewer layout in the plant. If a sewer plan exists
thoroughly review the sewer plan and examine each sewer to deter-
mine its course and destination. Where a sewer plan is not available,
the only practical way to determine the sewer layout is by dye-
tracing.
3. Determine the exact source and the point at which each waste stream
enters the sewer.
4. Sample each waste stream and the plant outfall. By doing so, each
waste stream is characterized and the outfall characterizes the total
plant effluent.
5. Sample each batch discharge.
6. If a point of upset exists within the plant, establishment of a samp-
ling station or monitoring equipment at that point will allow early
detection.
7. If data on different waste streams is available from past records,
use statistical techniques outlined in Section 4.5.1 as an aid to
establish the critical sampling locations within the plant.
NUMBER OF SAMPLES
Determine the number of samples from the following:
1. Follow NPDES permit requirements
2. Where NPDES permit is not applicable;
Apply statistical methods (Section 4.2) to data from a
preliminary survey.
. To effectively determine the concentration and types of pollu^
229
-------
Variable
Constant
Seasonal
Operation
Plant
Operation
Specific
cleanup
time
No Individ-
ual cleanup
discharges
Sampling
during
operation
season
2k hour per
day work
shift
Les* than
2k hour day
Separate com-
posite over
cleanup
period
Year round
operation
Sample
during
working
shlfts
Sample at all
times with
special empha-
sis on worse
than average
Figure 6.1 Factors of plant operation to be considered In
the design of the sampling program-(2)
-------
TABLE 6,1 NPDES EFFULENT LIMITATION PARAMETERS BY INDUSTRY
u.
Temperature Discharges XX X
¦OO-SOay XXXXXX X X XX X X XX XXXXXXX
Suspended So) Ids XXXXXXX XXXXXXXXXXXXX XXXXXXX X
Oils, Fats I Grease XX XXX XXXX X X
Amonia X X X X X X
K>
CO MI trite-Nitrogen X
I™-
Nitrate-Nitrogen XX X
Nitrogen (Kjeldahl) X X
Phosphorus X X XX
Sulfite X
Sulfide X XXXX
Sulfate
Chloride
Chlorine
Fecal Col Ifora (act. XX X
Fluoride X X X X X X X
Arsenic
larliae
Boron
Cadalue
Chroalu* X XXX X XXX
Cobalt
Capper X X
-------
TABLE 6.1 (Continued)
to
1
t
-J Tt
»
«
o»
01
«*-
>-

**
»
u

flC
•

—
u
0

1
v
**

0
u
c
V

—

*J
c
9)
w
1
_
0

s
9
W
0
<9
g
i_
4-1
0
Ifc
u1
0
—
a.
tn
u.
«L
—

X
X
X
X
X
X
X
X
X
X
X
c
%¦
1
>. «
X W 4. « *> O V C V — u u * CO
TS- -O « *-WOL*.q-««'--i40£ O
V3««u— c— *-» c o» « — o u. a._t.
- • C (A « w 9*9 U <0W «rt 0. ^ i_ C t * *r«F
A. MA.
to Le#d X XX
"" xxxxxxxxxxxxxxxxxxx X
Manganese
X X
Mercury * x
Nickel X X
*'"e XX XX
Phenol s X XX XX
PCB5
Aldrln
Oleldrln
Heptachlor
Color X
COO X X X X X X X
Cvanlde XX XX
Iron X
Surfactants X
AlifnlniM X
Arsenic X
Settleable Solids
-------
tants discharged, collect no less than three operating day
composite samples (3)
6.6 PARAMETERS TO MEASURE
6.6.1 NPDES Requirements
Parameters required for measurement in NPDES permits are listed by in-
dustry in Table 6.1 (2). These are the parameters commonly required and are
to be used as a minimum guideline where exact permit specifications do not
exist.
6.6.2 Other Parameters
Application of the techniques from Section 4.4 is a rational method of
establishing parameters to measure. However, if process control is desired,
measure the critical constituent. For example, if a distillation tower is
to be controlled, monitoring the organic carbon content of the discharge
stream may provide early information of leaks in the system.
6.7 TYPE OF SAMPLE
In any program, the type of sample, either composite or grab must be
established. Permit restrictions will determine the type for effluent moni-
toring, but for in-plant surveys both types should be considered and the most
appropriate chosen. For in-plant surveys where data does not exist carry out
a preliminary survey to determine the variability of individual streams to
arrive at a decision on the type of sample to be collected. Collect pro-
portional composite samples to determine the average waste water quality of
each waste stream or total plant effluent, depending upon the sampling ob-
jective,
Collect grab samples in the following situations:
1. If a batch discharge is to be characterized.
233
-------
2. If the flow is homogeneous and continuous with relatively constant
waste characteristics so a grab sample is representative of the
stream,
3. When the extremes of flow and quality characteristics are needed
(e.g., for design purposes).
4. When one is sampling for a parameter which requires that the entire
sample be used for analysis with no interior transfers of containers
(e.g., oil and grease).
5. When sampling for parameters which change character rapidly such as
dissolved gases or those which cannot be held for a long length of
time before analyses (e.g., bacteria counts, chlorine, dissolved
oxygen and sulfide).
6.8 METHOD OF SAMPLING
Choose manual or automatic sampling depending upon how the advantages
and disavantages of the methods apply to the specific sampling program. (Refer
to chapter 2). Adequate care should be exercised in sample collection, and
only trained personnel should be entrusted the task of sample collection.
6.8.1 Automatic Samplers
If an automatic sampler is to be used, the actual type of sampler is
determined by the constituents in the wastewater. A list of samplers and
their features are given in Table 2.3. The features and techniques for use
of automatic samplers are discussed in Section 2.3.2. To choose a sampler,
list the features needed for sampling the type of industrial wastewater, as
outlined in Section 2.3.2.3. If the variablity of the wastewater is not
known or expected to be high, a multiplex feature which takes more than one
sample into a single bottle is desirable. This would allow samples to be
234
-------
collected at short time increments such as once every 10-15 minutes.
Another possible feature would be to fill more than one sample bottle at a
time interval. This multiple bottle technique would allow use of one bottle
for the composite and the other for possible discrete analysis. Once the
needed features have been established, the sampler which best matches these
features can be selected. Available samplers may need adaptation. It is
imperative that the stream be well mixed at the sampling point to avoid prob-
lems when using automatic samplers in streams with a high solids content.
6.9 VOLUME OF SAMPLE AND CONTAINER TYPE
The volume of sample to be taken is determined by the number of analyses
to be performed on the sample. If this has not been determined, a grab sample
volume, a minimum of 1.51% (2 gal.) and an individual composite volume of 0.4S,
(0.11 gal.) should be taken. The container type is also contingent upon the
analysis to be run.
6.10 PRESERVATION AND HANDLING OF SAMPLES
This procedure is contingent upon the types of parameters to be anal-
yzed. Follow "Guideline Establishing Test Procedures For the Analysis of
Pollutants" (40 CFR 136) as amended in Federal Register, Vol. 41, No 232,
December 1,' 1976. Specific techniques are indicated by the parameter in
US EPA's "Methods of Chemical Analyses of Water and Wastewater, 1979", (4)
and Table 3.1.
6.11 FLOW MEASUREMENT
Flow measurement techniques adopted should be in relation to the samp-
ling location, type of flow, etc. Follow the guidelines enumerated in Chap-
ter 3 on Flow Measurements. Primary and secondary devices should be cali-
brated prior to taking flow measurements.
235
-------
6.12 REFERENCES
1. Black, H.H. Procedure for Sampling and Measuring Industrial Waters.
Sewage Industrial Wastes. 24:45, January 1952.
2. N.F.I. C-Denver. Effluent Limitations Guidelines for Existing
Sources and Standards of Performance for New Sources for 28 Point
Source Categories. Denver, p. 122, August 1974
3. Rabosky, J. G. and D. L. Koraido. Gaging and Sampling Industrial
Wastewaters. Chemical Engineering £50 p. 111-120, January 8, 1973.
4. U.S. Environmental Protection Agency. Methods for Chemical Analyses
of Water and Wastewater, 1974. ESML, Cincinnati, p. 298, 197 9.
236
-------
CHAPTER 7
SAMPLING AGRICULTURAL DISCHARGES
7.1 BACKGROUND
Agricultural discharges can be separated into two broad wastewater cate-
gories: 1. concentrated animal waste or manure from a confined feedlot;
2. runoff from an agricultural watershed. These two types of Wastewater
differ mainly in the concentration of pollutants. Runoff from fields, associ-
ated almost entirely with rainfall and snowmelt events, is characteristically
much less polluted, while feedlot runoff is a highly concentrated point source.
The values for constituents of field runoff depend on the amount and intensity
of rainfall or snowmelt, land use, topography, soil type, use of manure or
fertilizer, etc.
7.2 OBJECTIVES
There are two main objectives in sampling agricultural discharges:
1. Research - to study both field and feedlot runoff.
2. Regulatory - to monitor field or feedlot runoff or effluent from
feedlot runoff treatment.
7.3 FREQUENCY OF SAMPLING
7.3.1 Feedlot Discharge
7.3.1.1 Regulatory
Follow the sampling frequency given in the discharge permit. Daily
sampling is the maximum requirement in most permits.
237
-------
7.3.1.2 Other
Apply the spectral analysis techniques as outlined in Section 4. . Col-
lect preliminary data if not available by conducting one of the following (in
order of preference)
a. A one week survey collecting hourly grab samples where the discharge
is continuous.
b. A 24-hour survey collecting hourly grab samples.
Calculate the mean and variances as Indicated in Section 4. and apply
a computer program for spectral analysis (Section 4. ).
7.3.2 Field Runoff
Apply the statistical methods outlined in Section 4. if possible. Col-
lect preliminary data by sampling every 5 minutes for the duration of several
runoff events (1). Collect and analyze samples individually or composite
them proportional to flow, depending on the objectives of the study. Since
most of the variability in the runoff occurs during the initial part of the
runoff hydrograph on the rising side of flow crests, sampling is the most
critical at this time.
7.4 LOCATION OF SAMPLING POINTS
7.4.1 Feedlot Discharge
Channel feedlot runoff to a central point by sloping or trenching if no
treatment is provided. If treatment is provided, sample effluent from the
treatment system.
7.4.2 Field Runoff
Select a site downstream of the runoff area at a point where runoff col-
lects into a channelized flow. Use the topography of the area to locate this
238
-------
point. Choose a location with sufficient depth to cover the sampler intake
without excavation.
7.5 NUMBER OF SAMPLES
The number of samples for both feedlot discharge and field runoff are
calculated in the following manner:
1. Follow regulatory requirements.
2. Apply the statistics in Section 4. after the mean and variance are
determined through a preliminary survey (see Section 7.3).
7.6 PARAMETERS TO ANALYZE
7.6.1 Established by Regulation
Analyze all parameters required by discharge permits.
7.6.2 No Requirements
Analyze the following parameters (2,3,4):
1. Nutrients (total phosphate and nitrogen series)
2. Demand
3. Physical/Mineral (total and suspended solids)
4. Microbiological (fecal coliform and fecal streptococci)
Other analyses such as metals or pesticides may be necessary depending on
the nature of the study.
7.7 TYPE OF SAMPLE
Do not collect a single grab sample due to the high variability of run-
off. Collect a series of samples for analysis, or form a composite sample
according to flow using one of three methods:
1. Constant sample volume, time between sampling periods proportional
to stream flow.
239
-------
2. Sample volume proportional to total stream flow since last sampling
period; constant time between sampling periods.
3. Sample volume proportional to instantaneous stream flow rate; con-
stant time between sampling periods.
Use method 1 whenever possible, since this technique will allow a large
number of samples to be taken at high flows. Choose a flow volume increment
that will not exceed the bottle supply. An automatic sampler and integrated
flow measurement device is necessary for this type of sampling. Both methods
2 and 3 are acceptable also, but not preferred.
7.8 METHOD OF SAMPLING
Collect samples either automatically or manually; analyze the discrete
samples separately or composite them proportional to flow. For sampling field
runoff, use an automatic system activated by runoff through the flume. Typical
sampling/flow measurement stations are shown in Figures 7.1 and 7.2. If feed-
lot runoff contains large particulate matter (e.g., corn cobs), manual sam-
pling will be necessary.
7.9 VOLUME OF SAMPLE AND CONTAINER TYPE
Use multiple sample containers to provide the best preservation for
specific parameters. For example, if the parameters given in Section 7.6.2
(nutrients, demand, physical/mineral, microbiological) are to be analyzed,
three containers and three preservation techniques would be required for each
sample.
240
-------
Self Starting
Stage Recorder
Stilling Well
H Flume
Automatic
Sampler
Figure 7.1 View of field
installation (from 5)
-------
\
H Flume
Figure 7.2 View of field installation (7)
-------
ro
OJ
STRIP CHART
-RECORDING PEN
FLOW
HYDROGRAPH
12 v.ie:
SOLENOID

SAMPLING
CONTACTS
SAMPLE
SAMPLE
BOTTLE
A
X

CLAMP
r
•FLOAT
RUNOFF
Figure 7.3« Schematic of water level recorder
and sampler arrangement (from 5)
-------
Container
Parameter Group
Technique
1
Nutrients
Add H2S04 to pH 2 or 40-400
mg/1 HgCl2 and refrigerate
at 4°C
2
Demand
(Physical/Mineral)
Ice as soon as possible
after collection
3
Microbiological Collect in sterile container
and ice as soon as possible
7.10 FLOW MEASUREMENT
Select the flow measurement device based on the specific application and
the need for accuracy. A type H flume is advantageous because of its wide
range of accuracy (3,6). The measurement instrumentation should include a
continuously recording flow chart, with a pressure-sensitive record preferred
to ink. A schematic of a typical installation is shown in Figure 7.3. More
detailed information on flow measurement is given in Chapter 3.
7.11 REFERENCES
1. Miner, J.R., L.R. Bernard, L.R. Fina, G.H. Larson, and R.I. Lipper.
Cattle Feedlot Runoff Nature and Behavior. Journal WPCF. 38: 834-847,
October 1966.
2. Humenik, F.J. Swine Waste Characterization and Evaluation of Animal
Waste Treatment Alternatives. Water Resources Research Inst., Univ. of
North Carolina, Raleigh, NC, June 1972. 152p.
3. Harms, L.L., J.N. Dornbush, and J.R. Andersen. Physical and Chemical
Quality of Agricultural Runoff. Journal WPCF. 4_6: 2460-2470,
November 1974.
4. Robbins, J.W.D., D.W. Howells, and G.J. Kriz. Stream Pollution from
Animal Production Units. Journal WPCF. 4hi 1536-1544, August, 1972.
5. Harms, L.L. South Dakota School of Mines and Technology. Rapid City,
South Dakota. Personal Communication to Environmental Sciences Division.
December 20, 1974.
244
-------
6. Madden, J.M. and J.N. Dornbush. Measurement of Runoff and Runoff
Carried Waste from Commercial Feedlots. Proc. Int. Symposium on
Livestock Wastes. Ohio State Univ., Columbus, Ohio. April 19-22,
1971. 44-47
7. Leonard, R.A., C.N. Smith, G.W. Lar?dale, and G.W. Bailey, Transport
of Agricultural Chemicals From Small Upland Piedmont Watersheds.
Environmental Research Laboratory, Office of Research and Development,
Athens, GA, EPA-600/3-78-056, May 1978.
245
-------
CHAPTER 8
SAMPLING SURFACE WATERS, AQUATIC ORGANISMS, AND BOTTOM SEDIMENTS
8.1 BACKGROUND
The sampling of rivers and streams, estuaries, lakes and oceans, biolog-
ical organisms, and their associated bottom sediments are considered in this
chapter. Methods of sampling are directly affected by the objectives of the
study and parameters which are to be analyzed. Therefore, the decisions re-
garding parameters must be made at the beginning of the study in order to
develop a rational sampling program.
8.2 OBJECTIVES OF THE STUDY
The main objectives of sampling surface waters, aquatic organisms, and
sediments are:
1. Evaluation of the standing crop, community structure, diversity,
productivity and stability of indigenous aquatic organisms.
2. Evaluation of the quality and trophic state of a water system.
3. Determination of the effect of a specific discharge on a certain
water body.
8.3 PARAMETERS TO ANALYZE
Selection of parameters is dependent on the objectives and extent of the
program or study and must be performed prior to the development of the samp-
ling plan. Surface waters and sediments are commonly analyzed for the chem-
ical and biological parameters listed in Table 8.1.
246
-------
TABLE 8.1 COMMON ANALYSES FOR SURFACE WATER, AQUATIC ORGANISMS
AND SEDIMENT SAMPLING
Chemical
Biological
Dissolved Oxygen
Phosphate
Nitrogen Series
Alkalinity-
Silica
pH
Specific Conductance
Solids (TDS, TS, TSS)
Organic Matter and Demand
Color
Turbidity
Pesticides
Heavy Metals
Fish
Benthic Macroinvertebrates
Periphyton
Phytoplankton
Zooplankton
Macrophytes
Macroalgae
8.4 LOCATION OF SAMPLING POINTS
Select the study site based on the program objectives, the parameters of
interest, and the sampling units. For example, the following guidelines are
suggested in the EPA Model State Water Monitoring Program (1) for selecting
long-term biological trend monitoring stations:
1. At key locations in water bodies which are of critical value for
sensitive uses such as domestic water supply, recreation, propa-
gation, and maintenance of fish and wildlife.
2. In major impoundments near the mouths of major tributaries.
3. Near the mouths of major rivers where they enter an estuary.
4. At locations in major water bodies potentially subject to inputs
of contaminants from areas of concentrated urban, industrial, or
agricultural use.
5. At key locations in water bodies largely unaffected by man's
activities.
In order to avoid bias, use one of the following random sampling plans to
determine sampling points within the study site. Random sample selection is
discussed in more detail in the EPA Biological Field and Laboratory Methods
Manual (2).
247
-------
8.4.1 Simple Random Sampling
Use a simple random sampling plan when there is no reason to subdivide
the population from which the sample is drawn. Then the sample is drawn such
that every unit of the population has an equal chance of being selected. First,
number the universe or entire set of sampling units from which the sample will
be selected. This number is N. Then from a table of random numbers select as
many random numbers, n, as there will be sampling units selected for the sample.
Select a starting point in the table and read the numbers consecutively in any
direction (across, diagonal, down, up). The number of observations, n (sample
size), must be determined prior to sampling. For example, if n is a two-digit
number, select two-digit numbers ignoring any number greater than n or any
number that has already been selected. These numbers will be the numbers of
the sampling units to be selected.
8.4.2 Stratified Random Sampling
Use a stratified random sampling plan if any knowledge of the expected
size or variation of the observations is available. To maximize precision,
construct the strata such that the observations are most alike within strata
and most different among strata, i.e., minimum variance within strata and max-
imum variance among strata. Perhaps the most profitable means of obtaining
information for stratification is through a prestudy reconnaissance (a pilot
study). For information on conducting a pilot study, consult the EPA Biological
Methods Manual (2). Stratification is often based upon depth, bottom type,
isotherms, or other variables suspected of being correlated with the para-
meter of interest. Select as many strata as can be handled in the study. In
practice, however, gains in efficiency due to stratification usually become
negligible after only a few divisions unless the characteristic used as the
248
-------
basis o£ stratification is very highly correlated with the parameter of
interest (2).
8.4.3 Systematic Random Sampling
Use a systematic random sampling plan to assure an adequate cross section
while maintaining relative ease of sampling. A common method of systematic
sampling involves the use of a transect or grid. However, choose a random
starting point along the transect or grid to introduce the randomness needed
to guarantee freedom from bias and allow statistical inference.
8.4.4 Nonrandom Sampling
Use a nonrandom sampling plan if justified by the study site, or parameters
of interest, or the type of study being undertaken. For example, the following
sample locations might satisfy the program objectives:
_____ Parameter Sampling Location
Fish Shoreline sampling
Benthic macroinvertebrates Right, left bank, midstream or
transect
Periphyton Shoreline sampling
Phytoplankton Transect or grid
Zooplankton Transect or grid
Macrophytes Shoreline sampling or transect
Chemical Transect or grid
8.4.4.1 Impact of Point Discharges
A transect sampling scheme may be used to determine the impact of a point
discharge.
1. Place lines transecting the receiving water at various angles from
the discharge point.
2. Choose sampling intervals randomly or uniformly or by the methods
described in Section 8.4.4.2.
3. Choose two remote control points to use as background.
4. See Figure 8.1 for example.
249
-------
Poi nt
Source
o Control
Point
o Control
Point
Shoreline
Figure 8.1 Example of transect sampling scheme in reser-
voirs, lakes and coastal waters
A grid sampling scheme may also be used but is not applicable to all
biological parameters. Grid placement must be contained in a similar environ-
ment (e.g. all ripples or all pools) for a valid comparison.
1. Set up grids across and through the area to be sampled (i.e., in both
width and depth directions versus length) as required by the program.
2. The grid size is dependent upon the degree of lateral and vertical
mixing. If the amount of mixing is unknown, then take a larger
number of samples across and through the stream than would be other-
wise desirable.
3. Choose the number of samples randomly, uniformly or using the pro-
cedure in Section 8.4.4.2.
4. Choose a control point upstream of the grid system.
5. See Figure 8.2 for an illustration of the grid method.
250
-------
Poim
Sourt
Control
Point
Figure 8.2 Example of grid sampling scheme in rivers
8.A.4.2 Spatial Gradient Technique
This technique may be used for the rational selection of sampling sta-
tion locations (3,4). It presupposes the existence of historical data or
some reasonable estimate of the expected variability of the parameters to be
monitored over the region of interest, say, along the length of the river.
This technique has greater applicability for chemical than biological para-
meters .
1. Collect historical or comparable data to estimate the mean and
variance of the parameter of interest, Y.
2. Plot the maximum and minimum values of the parameter concentration
versus distance along the river (Figure 8.3).
251
-------
sample S :ation A
Distance Along River
Figure 8.3 Use of spatial gradient technique for
maximum spacing of sampling stations
3. Calculate a slope for both lines(Gmax and Gm^n).
4. Determine the difference between the slopes, i.e., GmiW - G .
r max mir
5. Determine the maximum allowable error in the estimates of the
parameter value at Point B.
AY
6. d « max
max r r
«mov _
""max - umin
7. Use this d to determine distance between points on a transect or
grid in a grid pattern.
8.5 NUMBER OF SAMPLES
The following information is summarized from the EPA Biological Methods
Manual (2).
252
-------
8.5.1 Simple Random Sampling
Use one of the following two methods depending on the decision variable.
1. Estimation of a Binomial Proportion - An estimate of the proportion
of occurrence of the two categories must be available. If the cate-
gories are presence and absence, let the probability of observing a
presence of P (0 < P < 1) and the probability of observing an absence
byQ (0 30, use t =* 2. This nQ ensures with a 0.95 probability
that P is within d of its true value.
b. For nQ > 30, use a second calculation where t is obtained from a
table of "Student's t" with nQ - 1 degrees of freedom. If the
calculation results in an n0, where
no
<0.05
no further calculation is warranted. Use nQ as the sample size.
no
If -— > 0.05, make the following computation:
N
2. Estimation of a Population Mean for Measurement Data - In this case
an estimate of the variance, S^, must be obtained from some source,
and a statement of the margin of error, d, must be expressed in the
253
-------
same units as are the sample observations.
t2s2
a. For n_ > 30, use n = 75
0 o cK
no
b. For nQ < 30, recalculate using t from the tables, and if —jj— >
n
n = 2
n
, , o
After a sample of size, n, is obtained from the population, the
basic sample statistics may be calculated. If the sample size, n,
is greater than 5 percent of the population > 0.05), a correc-
tion factor is used so that the calculation for the sample variance
is:
? N-n a n
S = N n-1
8.5.2 Stratified Random Sampling
Conduct a pilot study or obtain from other sources reliable estimates of
the variance within strata. If historical data has been .collected, use
optimal allocation to determine the total number of samples.
254
-------
tZ(ENksk)'
N2 d2
1 +
'Ht2
N^d2
Where t = Student's t value (use 2 for estimate)
Nk - number of sampling units in stratum k
2
sk ® variance of stratum k
sk = ® standard deviation of stratum k
N = total number of sampling units in all strata
d = acceptable parameter error
If no data is available, use proportional allocation to determine the
total number of samples:
t2*Vk2
Nd2
n
ENksk
1 0 2
N d
Use the following equations to determine the number of samples to be
collected in each stratum, n^:
nN s
Optimal allocation: n^ » k k
nN
Proportional allocation: n, « £_
2Nk6k
k = ~T
8.5.3 Systematic Random Sampling
Determine the number of samples to be taken on the grid or transect using
the methods given in Section 8.A.A.2 or 8.5.1.
255
-------
8.6 FREQUENCY OF SAMPLING
While the frequency of sampling will often be determined by the program,
use the Model State Water Monitoring Program (1) guidelines for guidance in
trend monitoring (Table 8.2).
8.7 METHOD OF SAMPLING
While compositing of individual grab samples is permitted for most
chemical parameters, as a rule do not composite biological samples. For
biological parameters collect single grab samples in replicate.
8.8 TYPES OF SAMPLERS FOR AQUATIC ORGANISMS
Choose the type of sampler that meets the needs of the sampling program
by considering the advantages and disadvantages of the sampler type. In gen-
eral, equipment of simple construction is preferred due to ease of operation
and maintenance plus lower expense. Advantages and disadvantages of various
water bottles are shown in Table 8.3 and illustrated in Figure 8.4. This
equipment is useful for chemical, phytoplankton and zooplankton sampling.
Corers and bottom grabs (Tables 8.4 and 8.5 and Figures 8.5 and 8.6) are use-
ful for sediment sampling. Nets and substrate samplers are covered in
Tables 8.6 and 8.7 and Figures 8.7 and 8.8.
There are inherent advantages of using a diver for sediment sampling. The
diver can ascertain what is a representative sample in addition to taking pic-
tures and determining qualitatively the current velocity.
256
-------
TABLE 8.2 MODEL STATE WATER MONITORING PROGRAM GUIDELINES FOR BIOLOGICAL MONITORING (1)
Community
Parameter Priority3
Collection &
analysis method*5
Sampling frequency0
Plankton
Counts and identification
1
Grab samples
Once each; in spring, summer

Chlorophyll a;

and fall

Biomass as ash-free weight

Periphyton
Counts and identification;
1
Artificial
Minimally once annually

Chlorophyll a;
2
substrate
during periods of peak

Biomass as ash-free weight
2

periphyton population

density and/or diversity
Macrophyton
Areal coverage;
2

Minimally once annually

Identification;
2
As circumstances
during periods of peak

Biomass as ash-free weight
2
prescribe
macrophyton population

density and/or diversity
Macroinver-
Counts and identification
1
Artificial and
Once annually during periods
tebrate
Biomass as ash-free weight
2
natural
of peak macroinvertebrate

Flesh tainting;
2
substrates
population density and/or

Toxic substances in tissue^

diversity
Fish
Toxic substances in tissue**
1

Once annually during

Counts and identification;
2

spawning runs or other

Biomass as wet weight;
2
Electrof ishing
times of peak fish

Condition factor;

or netting
population density

Flesh tainting
2

and/or diversity

Age and growth
2

a Priority: 1) Minimum program; 2) Add as soon as capability can be developed,
b See EPA Biological Methods Manual. c. Keyed to dynamics of community.
d See Analysis of Pesticide Residues in Human and Environmental Samples, "USEPA, Perrine Primate
Research Lab, Perrine, FL 32157 (1970)," &"Pesticide Analytical Manual," USDHEW, FHA, Wash,D.C.
-------
TABLE 8.3 COMPARISON OF WATER SAMPLERS

Device
Application
Container Type
Advantages
Disadvantages
Nanseti Bottle
Phytoplankton
Teflon lined
Able to use in series
Small volume
Kemmerer Bottle
Chemical
Bacteriological
Zooplankton
PVC
Brass
Acrylic plastic
No cross contamination
Fixed capacity
from 0.4-16 L
Van Dorn Bottle
Chemical
Bacteriological
Zooplankton
Phy topiankt on
PVC
No cross contamination
Fixed capacity
from 2-30 L
Simple Bottle
Chemical
Bacteriological
Glass
Easy to make
Cross contamina-
tion
Pumps
Chemical
Zooplankton
Phytoplankton
Vanes
Large volume,
samples a vertical
water column, contin-

uous sample
-------
TABLE 8-4 COMPARISON OF BOTTOM GRABS
Device
Advantages
Disadvantages
Ponar
Ekman
Tall Ekman
Peterson
Smith-Mcintyre
Diver
Safe, easy to use, prevents escape of
material with end plates, reduces shock
wave, combines advantages of others,
preferred grab in most cases
Use in soft sediments and calm waters,
collects standard size sample
(quantitative), reduces shock wave
Does not lose sediment over top; use
in soft sediments and calm water,
standard sample size, reduces shock wave
Quantitative samples in fine sediments,
good for hard bottoms and sturdy and
simple construction
Useful in bad weather, reduces premature
tripping, use in depths up to 1500 m
(3500 ft), flange on jaws reduces
material loss, screen reduces shock waves,
good in all sediment types
Can determine most representative
sampling point and current velocity
Can become buried in soft sediments
Not useful in rough water; not useful
if vegetation on bottom
Not useful in rough waters, others as
for Ekman
May lose sampled material,premature
tripping, not easy to close; does not
sample constant areas; limited sampling
capacity
Large, complicated and heavy, hazardous
for samples to 7 cm depth only, shock
wave created
Requires costly equipment and
special training
-------
TABLE 8.5 COMPARISON OF CORING DEVICES
Device
Advantages
Disadvantages
Kajak or
K.B. Corer
Moore (Pfleger)
0'Conner
Elgmork's
Jenkins
Enequist
Does not impede free flow of
water, no pressure wave, easily
applied to large area
Valve allows sample to be held
Can sample water with hard bottoms
Sample easily removed, good in soft
muds, easy to collect, easy to
remove sample
Good in soft sediments and for
collecting an undisturbed
sediment-water interface sample.
Visual examination of benthic
algal growth and rough estimates
of mixing near the interface after
storms can be made
Good in soft/medium sediments,
closing mechanism
Careful handling necessary to avoid
sediment rejection, not in soft
sediments
Not in deep water
Not in hard sediments
Complicated
Does not penetrate hard
bottom
Kirpicenko
Soft and hard bottoms, various
sizes, closes automatically
Not for stony bottoms
-------
TABLE 8.6 COMPARISON OF NET SAMPLING DEVICES
Devices
Application
Advant ages-
Disadvantages
Wisconsin Net
Zooplankton
Efficient shape
concentrates
sample
qualitative
Clarke-Bumpus
Zooplankton
Quantitative
No point sampling,
Difficult to mea-
sure accurately
depth of sample
261
-------
TABLE 8.7 COMPARISON OF SUBSTRATE SAMPLERS

Type of Substrate
Advantages
Disadvantages
1. Artificial
Reduces compounding effects of
substrate differences

Modified
Hester-Dendy
Reduces compounding effects of
substrate differences, multiplate
sampler
Long exposure time, difficult
to anchor, easily vandalized
Fullner
Wider variety of organisms
Same as modified Hester-Dendy
EPA Basket Type
Comparable data, limited extra
material for quick lab processing
No measure of pollution on
strata, only community formed
in sampling period, long exposure
time, difficult to anchor, easily
vandalized
EPA Periphyton
Sampler
Floats on surface, easily anchored,
glass slides exposed just below surface
May be damaged by craf t;
easily vandalized
2. Natural

May be difficult to Quantitate
Any bottom or
sunken material
Indicate effects of pollution, gives
indication of long term pollution
Possible lack of growth
-------
Nansen Water Bottle

Van Dorn Sampler
Figure 8.4 Water Bottles
(Courtesy of Wildlife Supply Co.)
263
-------
Ekman Grab
Ponar Sampler (two sizes)
Figure 8.5 Bottom Grab Samplers
(Courtesy of Wildlife Supply Co.)
264
-------
Smith-Mclntyre (Aberdeen) Grab
Figure 8.5 (continued) Bottom Grabs
265
-------
D

Valve
Cylf nder
ro
on
on
w
Clamp

Nose of Sampler
Side View-Vertical Core Sampler
Figure 8.6.
~
rnT
A
Elgmork's Core Sampler
Core samplers
-------
Clark-Bumpus Sampler
Wisconsin Net
Figure 8.7 Nets and Related Samplers
267
-------
Side View
1
1
1
1
1
1 p
0 0 :
»
»
i
<
i
i
i
i

T
•
0
LI

-T

f
5

¦
C-€j 0
i

-L
Top View
EPA Periphyton sampler. Plexiglass frame supported by
two styrofoam floats. Rack holds eight glass microscope
slides .
Figure 8.8 Periphyton samplers
268
-------
I
t*
Limestone filled basket sampler
Modified Hester-Dendy type multiple-plate artificial
substrate
Figure 8.8 (Continued)
269
-------
8.9 VOLUME OF SAMPLE AND CONTAINER TYPE
The size of sample is dependent on the expected amount of the chemical
parameter to be analyzed. The container type is also dependent on parameter
type. Refer to Section 17 for specific information relative to the chemical
parameters which are to be analyzed. Refer to the Biological Methods Manual
(2) for container type and sample volumes, where applicable.
8.10 PRESERVATION AND HANDLING OF SAMPLES
Refer to Section 17,1 for specific information regarding preservation
and handling of samples relative to the chemical parameters to be analyzed,
and to the EPA Biological Methods Manual (2) for aquatic organism preserva-
tives .
8.11 FLOW MEASUREMENT
Flow measurement in rivers is accomplished by the combined use of a cur-
rent meter to measure the stream velocity and a stage recorder to measure
the surface elevation of the river. Consult USGS gaging stations for addi-
tional or historic information. See Section 3 for more details.
8.12 REFERENCES
1. National Water Monitoring Panel. Model State Water Monitoring Program.
U.S. EPA Report No. EPA-440/9/74-002. U.S. EPA Office of Water and
Hazardous Materials. June 1975.
2. Weber, C.I., ed. 1973. Biological Field and Laboratory Methods for
Measuring the Quality of Surface Waters and Effluents. National
Environmental Research Center, Office of Research and Development, U.S.
EPA, Cincinnati, Ohio, EPA 670/4-73-001.
3. Hill, R.F. Planning and Design of a Narragansett Bay Synoptic Water
Quality Monitoring System. NEREUS Corp., 1970
4. Drobny, N.L. Monitoring for Effective Environmental Management.
Proc. ASCE National Water Resources Engineering Meeting. Atlanta,
Georgia. January 24-28, 1972.
270
-------
CHAPTER 9
SAMPLING OF GROUND WATER
9.1 BACKGROUND
Ground water is an important source of water, particularly in regions of
scarce surface water supplies and polluted surface waters. The quality of
ground water can be altered by various events: intrusion of sea water,
seepage from waste injection wells, leachate movement from landfill, waste-
water lagoon seepage, etc. Increased use of ground waters for drinking,
industrial and other purposes necessitates monitoring the quality of ground-
water.
The hydrologic cycle (1) shown in Figure 9.1 illustrates the various com-
ponents involved in ground waters, their interrelation and necessity of
considering all components in a monitoring program. It is evident that the
quality and quantity of waters entering and leaving the subsurface have an
immense influence on the quality and quantity of ground water.
A comprehensive ground water monitoring program goes beyond sampling
subsurface waters and includes sampling appropriate waters at the land surface
(springs, ponds, lakes, rivers, lagoons, etc.), surface soils and sub-surface
soils. Therefore, knowledge is required of hydrogeology, well hydraulics,
geochemistry, and physical chemistry.
271
-------
Condensation
Snow
S '
Precipitation
\• * Surface
\ . runofl
Transpiration
Spring
Evaporation
Water table
Lake
Ocean
Figure 9.1 The hydrologic cycle
immmMmmmmsmmmim
VADOSE ZONE
aturaiion'.zone
AQUIFER
.E FFECTlVE. BAS e o F AQUIFER
LAND SURFACE
J
WASTE DISCHARGE
WELL DISCHARGE
272
-------
The subsurface environment is exceedingly complex with contrasting con-
ditions existing at short vertical and horizontal- distances from one another
(2). Therefore, subsurface soil sampling is needed to assess any interactions
between subsurface soil and water. These interactions c?n be classified into
three general processes which may affect the quality of ground water:
1. Physical processes 2. Geochemical processes
. filtration . complex ion pair formation
. dispersion . acid-base reaction
. dilution . inorganic redox reactions
. adsorption-desorption . ion exchange
. precipitation-solution
3. Biochemical processes
. solute uptake in biosynthesis
. solubilization
. mineralization of organics
. catalysis of inorganic redox reaction
These interactions can act on substances to slow migration, increase con-
centrations, decrease concentrations, or alter the original substance. With-
out sampling and analysis of subsurface solids, the data generated by a ground
water sampling program can be misinterpreted. For example, the reduction of a
constituent in water percolating through a soil may be by filtration, ion ex-
change, biological uptake or any combination of these processes. An increase
in the concentration of the constituent in the filtrate can be attributed to a
change in the percolating water, a failure of the soil retention mechanism, or
another water source. Without the proper solids analysis many questions arise.
What was the removal mechanism? Has the capacity of the mechanism been
273
-------
exceeded or exhausted? Is the mechanism renewable? Can the increase be
attributed to a leaching or solution process?
This chapter addresses the sampling of ground water in the subsurface
environment with the realization that it constitutes one part of a complete
ground water monitoring program.
9.2 FREQUENCY OF SAMPLING
The frequency of sampling is selected by considering the factors listed
in this section.
9.2.1 Program Objectives
A program's objective can be classified as planning, regulatory, process
control or research and development. In general the frequency of sampling
required increases as one moves from planning to research and development.
9.2.2 Statistics
Application of statistical techniques to establish the optimum sampling
frequency is dependent upon the availability of sufficient and reliable data.
General guidelines are set forth in Section 4.3.1 on sampling frequency and
data required for spectral analysis applications.
9.2.3 Budgetary Limitations
A sampling program is required to operate within monetary limits. These
limits are reflected in three forms: (1) Manpower, (2) Sampling facilities,
and (3) analytical facilities. However, because water quality changes are the
result of the slow movement of water in the aquifer,frequency can be related to
the direction of movement and the travel time to minimize cost.
9.2.4 Ground Water Basin Characteristics
Basin characteristics may require irregular sampling frequency. Such
characteristics include:
274
-------
hydrogeology . flooding
climatology . tides
seasonal variations in vegetation
9.2.5 Other Considerations
Additional influences that affect sampling frequency include:
. seasonal utilization of surface recharge, injection or
extraction sites
. sampling location in relation to elevation ( e.g. topsoil,
vadose zone or zone of saturation)
. non-point sources of contamination
. removal or addition of overlaying sediments by natural
actions or man's activities
. geochemical properties
9.2.6 General Guidelines for Sampling Frequency
Under natural conditions, the quality of ground water will change with
time. Rates of change are related to rate of flow which in turn is governed
by the hydrogeologic situation. Some ground water basins unaffected by man
show annual fluctuations in quality produced by seasonal variations in re-
charge, level changes, and discharge. Some general guidelines for sampling
frequency are:
Sanitary landfill sites, 50-100 days (4)
Trends in water quality, annually (5)
Salt water intrusion, frequently (6)
Rapid changes - continuous, daily or weekly
Sampling period to define periodic changes, every two years
Nfcar or down stream from a known pollution source, semi-monthly,
275
-------
monthly, or bimonthly
Ground water flowing toward wells or being pumped from wells,
semi-annually
9.3 LOCATION OF SAMPLING POINTS
9.3.1 General
Sampling location guidelines applicable to all ground water sampling
situations cannot be devised. Each sampling program has unique characteristics
based upon its geological setting as shown in Figure 9.2. Therefore, a three
dimensional approach, i.e. the latitude, longitude and mean sea elevation of
the sampling point must be specified.
Most sampling programs assess the potential of ground water pollution.
Major sources and causes of ground water pollution, methods of waste disposal
and categories of pollutant discharge are shown in Table 9.1.
9.3.2 Preliminary Investigation
A systematic investigation of an area by highly trained personnel is
essential in determining the optimum location of sampling points. The type and
number of professional services essential to the program of investigation will
vary with the magnitude and complexity of the sampling program. As a minimum,
at least one geologist is required, while those of a complex nature may require
more than one geologist and other professionals such as environmental engineers,
geotechnical engineers, and chemists.
A preliminary investigation should consist of:
. Determining pertinent features of geological framework.
. Characteristics of contaminants
. Man-made contingencies causing the movement of water, whether
from pumping from wells, disposal of waste, or accidents that
276
-------
WASTE SITE
WATER TABLE
i ZONE OF_^-r-
SATURATIQN
' Sf^A
' • ' D
>»• .••/ ..•
j MONITOR WELLS
/ ^ -WATER WELL

| WATER TABLE
fZONE OF
SATURATION
WASTE SITE
MONITOR WELLS
WATER WELL
' G/?0/ '
Figure 9.2 lypes of Dispersion of Contaminated Wastewater
277
-------
TABLE 9.1 SOURCES OF POLLUTION, DISCHARGE CATEGORIES AND METHOD OF DISPOSAL ( 3)
Category
Common Method oF Disposal
N5
a>
Municipal
Sewer leakage
Sewage effluent
Sewage sludge
Urban runoff
Solid wastes
Lawn fertilizers
Agricultural
Evapotranspiration
and leaching
(return flow)
Fertilizers
Soil adsoadnents
Pesticides and
hi»rbic ides
AniAal wastes
(feedlots and
daires)
Stockpiles
Industrial
Cooling water
Process waters
Storm runoff
Boiler blowdown
Stockpiles
Water treatment
plant affluent
Hydrocarbons
Tank and pipeline
leakage
Oilfield wastes
Briucs
Hydrocarbons
Mining wastes
Miscellaneous
Polluted precipitation
aiid surface water
Septic tanks and
cespools
Highway deicing
Seauater intrusion
Point
Line
Diffuse
Percolation
Pond
X
X
X
X
na
X
na
na
Surface Spreading
and irrigation
Seepage Pits
and Trenches
Dry Stream
Beds
Landtills
na
X
X
X
X
X
X
na
na
X
na
X
X
na
na
X
na
na
na
na
na
na
Disposal Injection
Wells Wells
na
X
X
X
X
X
na
X
X
na
na
na
X
na
na
na
-------
cause pollution.
Sources of geologic information are the United States Geological Survey
(USGS), United States Soil Conservation Service, state geological agencies,
local health departments, universities, commercial geotechnical firms, and
professional societies.
The location and characteristics of past, present and future developments
can be determined from various maps and studies produced by government agencies
and private concerns. Land use plans, areawide assessments, transportation
studies, open space plans, and zoning plats are a few examples. Such items as
sewers, lagoons, storage facilities, mining activities, agriculture activities,
and flood protection facilities should be examined.
When necessary geophysical methods should be utilized to confirm and expand
existing information. These methods can be classified under two general cate-
gories: (1) surface and (2) subsurface investigations.
9.3.2.1 Surface Investigation
Surface investigation of subsurface conditions by geophysical methods in
this category is at best an estimate of the extent of contamination underground.
These methods however are useful as a complement to existing information and
are economically attractive when compared to subsurface investigative methods.
Geophysical methods are based on the measurement or detection of physical
properties and consist of seven basic types of geophysical measurements:
Electrical - resistivity measurement which relates soil
characteristics to extent of mineral contamination
Seismic - change in the propagation of sound from small
explosion to determine degree of porosity of soil
Radioactive - gamma and beta measurements to determine the
degree of natural radiation or from other radioactive sources.
279
-------
Thermal - temperature measurement to indicate dilution from a
pollutant source
Gravity - change in density of material
Magnetic - detection of magnetic pollutants
Electro-magnetic - remote sensing of electromagnetic energy from
earths surface.
These measurements are interpreted in terms of porosity, water content,
density, mineral content, water quality and geologic formation. Correct inter-
pretation usually requires auxiliary subsurface investigation. The electrical
resistivity and seismic refractory methods are most commonly employed.
After assessing available information and the data collected, and taking
into account the sampling program objectives, the need for subsurface investi-
gation should be determined.
9.3.2.2 Subsurface Investigation
Regardless of the methods utilized in subsurface investigations, a hole or
shaft in the earth is required. The exploration is conducted at the surface by
placement of equipment in the hole or shaft. Both existing and new-made wells
and test holes are commonly chosen as access points for the subsurface equip-
ment. Wells can be constructed by the following principles: (1) dug, (2) augered
(3) driven, (4) jetted, and (5) drilled.
A prudent approach is utilization of existing wells when possible and
correlation of the data and measurements gathered through them with preexisting
information to determine the necessity and location of new holes or wells.
Geophysical measurements obtainable from a test hole or well include:
potential logging . radioactive logging
resistivity logging . caliper logging
temperature logging
280
-------
The preexisting information, surface investigation data and subsurface
investigation data should be correlated to determine the necessity, extent and
location of ground water sampling points required to meet the program objec-
tives.
9.3.3 Vertical and Horizontal Sampling Points
Consider the following factors for location of sampling points:
Ground water movement - movement of water from a high to low gradient
stagnant areas, movement induced by sampling and production wells, and
inflow-outflow from surface water. Tracers in ground water have limited
value; chlorides and tritium move with water, dyes and cations sorb on
earth materials.
. Horizontal placement - Govern the number and placement of wells by the
geohyrologic conditions and the disposal operation of the pollution
source. The routine practice of placing wells in a circle around the
site may not be acceptable. Use a minimum of 2 or 3 wells when the
direction of movement is known,; For refuse disposal sites (8), use:
(a) up-land sites - place wells within the site and below the
refuse cells
(b) valley floor sites - place wells outside the site and on the
down flowside.
Vertical placement - locate withdrawal point at a representative depth;
to average concentration gradients or at a depth commensurate with the
objective, e.g. early detection of salt water intrusion. Well depth
should be at a depth to collect samples regafdless of seasonal fluctu-
ation in the water table. Contaminated water from waste site tends to
be more concentrated in upper part of zone of saturation; salty water
lower part.
281
-------
9.3.4 Representative Samples
Sample Collection - Pump from well until temperature, pH, and
specific conductance are constant; standby, observation, new or
little used wells may require one day of pumping. At a minimum
pump until pH is constant.
Cased Wells - Backfill and plug with cement the annular space
between the zone of saturation and the surface to avoid contami-
nation from drilling muds and additives.
Pump Type - Do not use airlift pumps when analyzing for such
parameters as pH, carbonate, bicarbonate, temperature, and purgeables.
Preservation - Preserve samples for organic analyses according
to EPA procedures (4°C)-do not acidify since this procedure en-
hances volatilization of undissociated fatty acids, precipitates
humic-like organics and facilitates hydrolysis and oxidation
of complex organics.
Trace Metal Samples - Use plastic construction for well casing
material. Most wells already contain metallic casing, screens,
and pump column, therefore minimize contact time with walls.
Contamination - Avoid contamination of the sample from soil
bacteria, particulate matter, and atmospheric oxygen.
Small Springs - Collect samples from small springs in uncon-
solidated deposits by driving a well point or slotted pipe
to a depth of 1 meter or less into ground adjacent to spring.
» Large Springs - Sample from large springs in consolidated
rock rather than other types of springs. Sample spring in
upswelling water by forcing a bottle held by hand or attached
282
-------
to a rod for oxygen contaminated sample. Use a thief
sampler for manual metal sampling or collect samples using
an all plastic submersible electrical pump and garden hose.
Dissolved Parameters - Filter water through a 0.45 filter;
some metals, i.e. iron and manganese, form colloidal particles
that pass through this filter. Use a 0.1 filter to correct
this problem.
9.4 PARAMETERS TO MEASURE
9.4.1 General
The analyses performed on samples should be a function of the sampling
program objective. Some analyses must be performed in situ or on site and
others in the laboratory.
Table 9.2 shows the classification of pollutants and pollution indi-
cators in ground water. Table 9.3 summarizes major inorganic pollutants and
pollution indicators in the ground water. Table 9.4 shows the relation
between pollution sources and pollution types.
9.4.2 Parameters for Planning Objectives
The data required for planning should be of a general nature. The para-
meters utilized for collecting data should generally not indicate specific
pollution problems but overall water quality and quantities of an area. Some
commonly employed measurements are:
PH
nitrate
oxidation reduction potential chloride
total coliforms
water level
total dissolved solids
TOC
283
-------
TABLE- 9.2 CLASSIFICATION OF POLLUTANTS AND POLLUTION INDICATORS IN GROUNDWATER
Physical
Organic
Chemical
Inorganic
Chemical
Bacterio-
logical
Radio-
logical
Temperature
Density
Odor
Turbidity
Carbon
Chlorophylls
Extractable organic
matter
Methylene
blue active
substances
Nitrogen
Chemical oxygen
demand
Phenolic material
Pesticides
(insecticides
and herbidides)
Major
constituents
Other
constituents
Trace elements
Gases
Coliform group
Pathogenic
micro-
organisms
Enteric viruses
Gross alpha
activity
Gross beta
activity
Strontium
Radium
Tritium
-------
TABLE 9.3 INORGANIC CHEMICAL POLLUTANTS AND POLLUTION INDICATORS
Major
Others
Pollutant or Pollution Indicator
Drinking Water Trace Other Trace
Gases
calcium
magnesium
sodium
potassium
carbonate
bicarbonate
sulfate
chloride
nitrate
total dissolved solids
pH
electrical conductivity
oxidation potential
silica
boron
fluoride
nirogen forms
phosphorus forms
hardness
iron
manganese
arsenic
barium
cadmium
hexavalent chromium
copper
cyanide
lead
selenium
silver
zinc
mercury
vanadium
molybdenum
bromide
iodide
nickel
aluminum
cobalt
lithium
sulfide
beryllum
methane
hydrogen sulfide
carbon dioxide
dissolved oxygen
residual chlorine
organic purgeables
-------
TABLE 9.4 RELATION BETWEEN POLLUTION SOURCES AND POLLUTION TYPES

Inorganic
Trace
Organic
Bacterio-
Radio-
Source
Physical
Chemical
Elements
Chemical
logical
logical
Municipal

Sewer leakage
minor
primary
secondary
primary
primary
minor
Sewage effluent
minor
primary
secondary
primary
primary
minor
Sewage sludge
minor
primary
primary
primary
primary
minor
Urban runoff
minor
secondary
variable
primary
minor
minor
Solid wastes
minor
primary
primary
primary
secondary
minor
Lawn fertilizers
minor
primary
minor
minor
minor
minor
Agricultural

Evapo transpiration

and leaching
minor
primary
minor
minor
minor
minor
Fertilizers
minor
primary
secondary
secondary
minor
minor
Soil amendments
minor
primary
minor
minor
minor
minor
Pesticides
minor
minor
minor
primary
minor
minor
Animal wastes

(feedlots & dairies)
minor
primary
minor
secondary
primary
minor
Stockpiles
minor
primary
minor
variable
variable
minor
Industrial

Cooling water
primary
minor
primary
minor
minor
minor
Process waters
variable
primary
primary
variable
minor
variable
Storm runoff
minor
secondary
variable
primary
minor
minor
Boiler blowdown
primary
secondary
primary
minor
minor
minor
Stockpiles
minor
primary
variable
variable
minor
variable
Water treatment

plant effluent
minor
primary
secondary
minor
minor
minor
Hyd ro carb ons
secondary
secondary
secondary
primary
minor
minor
Tank & pipeline

leakage
variable
variable
variable
variable
minor
variable
(continued)
-------
TABLE 9.4 (continued)

Inorganic
Trace
Organic
Bacterio-
Radio-
Source
Physical
Chemical
Elements
Chemical
logical
logical
Oilfield wastes

Brines
primary
primary
primary
minor
minor
minor
Hydrocarbons
secondary
secondary
secondary
primary
minor
minor
Mining wastes
minor
primary
primary
cariable
minor
variable
Miscellaneous

Polluted precipitation

and surface water
variable
variable
variable
variable
variable
variable
Septic tanks and

cespools
minor
primary
minor
secondary
primary
minor
Highway deicing
minor
primary
minor
secondary
minor
minor
Seawater intrusion
primary
primary
primary
minor
minor
minor
-------
9.4.3 Parameters for Regulatory Objectives
Parameters employed for regulatory purposes are used in the areas of
permit compliance, surveillance, pollution detection and water quality
assessment. The parameters selected or specified for measurement should be
a direct function of the process or discharge under consideration. In addition
to parameters that may be listed in a permit, minimum standards are set under
legislation such as the Federal Water Pollution Control Act, the Safe Drinking
Water Act and the Resource Conservation and Recovery Act.
9.4.4 Parameters for Process Control Objectives
Process control parameters may be viewed by the relations the process has
to ground water. The ground water may function as a water source or disposal
method or both. For example: a water to be used for non-potable purposes
should not require the same analyses as when potable purposes are intended;
parameters appropriate for assessing the effects of aquifer recharge with
municipal waste water would not be appropriate for assessing the effects of a
sanitary landfill on aquifer quality.
9.4.5 Parameters for Research and Development Objectives
Parameter selections for research and development needs are usually more
specific and involve fewer numbers than the previous categories. This is the
result of the nature of the research and development work, i.e. an area of
limited extent.
288
-------
9.5 TYPE OF SAMPLE
9.5.1 General
Collect ground water samples by manual or composite methods as discussed
in Chapter 2 or by accumulation columns in Chapter 12. Mmual grab samples are
most commonly collected since movement of water is slow and water quality does
not normally exhibit sudden drastic changes.
9.5.2 Grab Samplers
Collect grab samples by one of the following methods: (1) transport the
water to the surface in a container, (2) transport the water through a closed
conduit and discharge on the surface, and (3) construct tile or ditch lines for
relatively high water tables.
9.5.2.1 Container Transport
Collect ground water samples with a depth integrated or point sampler.
A depth integrating sampler shown in Figure 9.3 (9) is simply a container
equipped with a holding and submerging mechanism which collects water through-
out the vertical profile. Other depth integrating samplers known as bailers
(10,11) Figures 9.4aand 9.4b, are lowered through the water and are filled
through the bottom inlet which contains a check valve for retaining water when
retrieved.
Use point samplers to collect a sample at a specific depth. Two types
are the flow-through and water-tight samplers. Flow-through samplers are
shown in Figures 9.5.a(12), 9.5.b(13), 9.6.a, and 9.6.b(14). Water tight
samplers are shown in Figure 9.7 (9).
9.5.2.2 Closed Conduit Transport
Use a pump, compressed gas or a vacuum to transport the water sample to
the surface. Do not use vacuum systems for organic purgeablea.
289
-------
~
Figure 9.3 Depth-integrating samplers
290
-------
.D. Teflon
Rope to surface
.Nickel Wire
Cable
_ Eyebolt
^"TTp Cap
Airholes

Tubing
3/4" diameter
Glass Marble
— Reducing
coupling
¦1" diameter Teflon
Extruded Rod
Rubber ball
5/16" diameter
Hole
Nuts and washers
for weight
a. Depth-integrating
bailer
Figure 9.4
291
b. EPA bailer
-------
.Water-tight plug
Suspension lug
Solenoid
Latch
Plunger
84 cm
Connecting rod
Plastic body
-Phosphor-bronze spring
Plunger
Adjustment nut
7 cm
a. Tate Sampler
Suspension
cone
upper valve
4-1
Lower
valve
b. Frost Sampler
Figure 9.5 Point Flow—through Samplers
-------
Standard *
pipe *pl&
N>
vo
U»
Drilled 009
tafitcaMe
Leather
L6«»*nr>g
cable
Messenger
we
-------
Figure 9.7
-------
Wells permanently equipped and continuously flowing may be sampled from a tap
in the conduit; submersible electrical pumps and pitcher pumps (up to 23 feet
head) may be used where applicable to pump water to the surface.
Use pressure or vacuum (not purgeables) lysimeters to collect samples
from the vadose zone, i.e., the zone above the water table. (15, 16, 17, 18)
This device consists of a porous ceramic cup equipped with a small diameter
PVC pipe (not for organics) for sample accumulation as shown in Figures 9.8.a,
9.8.b, and 9.8.c. Operation consists of applying a slight vacuum to the cup
during filling, releasing the vacuum, and applying pressure to the cup to
transport the water to the surface. Excessive pressure will cause air bubbl-
ing with loss of dissolved gases in the sample.
Use a piezometer installed in a boring, Figure 9.9 (19), to also collect
water from the vadose zone utilizing either an air lift, Figure 9.10.a (20) or
pressure system, Figure 9.10.b, to transport water to the surface.
295
-------
WUBOin TUBING
COPPER
TUBING
Figure 9.8.a Wallahan Lysimeter
•e-WAY PUMP
°5f°
PRESSURE-
VACUUM IN
BENTON ITE
3/16" COPPER
TUBE
PLASTIC PIPE
(24'XONG
608 mm.
POROUS CUP
BCNTONITE
SAMPLE
r bottle
DISCHARGE
TUBE
132 mm.
BOREHOLE
SANO
BACKFILL
SUPER-SIL
Figure 9.8.b Parizek and Lane Lysimeter
296
-------
TUBING TO SURFACE
CONNECTORS
PIPE-THREAD SEALANT
PVC PIPE CAP
PIPE
PVC CEMENT
POLYETHYLENE TUBING
BRANCH
FEMALE ELBOW
13
POPPET CHECK VALVE
Ei
•Of-
® C\J
CONNECTORS
EPOXY CEMENT
POLYETHYLENE TUBING
POROUS CU»
Figure 9.8.c Wood Lysimeter
297
-------
PRESSURE-VACUUM
LINE
LOW PERMEABILITY
MATERIAL
BOREHOLOE

V/.
:v^.v\sv.v

mm
1:®
liliSS

l'» .V
•lil

DISCHARGE LINE
LANO SURFACE
POROUS OR SLOTTED
PVC PIPE
CHECK VALVE
SAND BACKFILL
POLYETHYLENE TUBING
END CAP
"T"AN0 ELBOW FITTINGS
SAMPLE COLLECTION
CHAMBER
END CAP
Figure 9.9 Details of a Low-cost Piezometer Modified
for Collection of Water Samples
298
-------
FO
VO
«o

PUMPING
LIFT
-Z-
-O
o.-
o•
o
o.
•— —- —»
h»=-
T_ h.
hm* Mont mum height to which the
oir-woter mixture will rise
hw* Submerged length of the oir line
"V- Density of the oir-woter mixture
7m* Density of woler
* Potentiometric surfoce
Figure 9.10.a Air-lift System
PUMPING
LIFT
t
~ _T
h
hmoi *Moximum tfteight to which woler in
the nylon >hose will rise relotive to
the woter level between the hose
ond piezometer cosing
Ah " Differenc* between hmol ond the
hydraulic heod in Ihe tormotion
2- * Potentiometric surfoce
Figure 9.10.b Pressure System
-------
9.6 REFERENCES
1. Todd, D.K. , Ground Water Hydrology, John Wiley & Son, New York, 336 pp,
1959.
2. Todd, D.K., Tinlin, R.M., Schmidt, K.D., and Everett, L.G., "Monitoring
Ground Water Quality: monitoring Methodology", EPA Report No. 600/4-76-026,
169 p., 1976.
3. Warner, D.W., "Rationale and Methodology for Monitoring Groundwater Polluted
by Mining Activities", EPA 680/4-74-003.
4. "Sanitary Landfill", ASCE Manual No. 39, American Society of Civil Engineers,
New York, 1976.
5. Tinlin, R.M., "Monitoring Ground Water Quality: Illustrative Examples",
EPA Report N. 600/4-76-036, 1976.
6. "Ground Water Management", ASCE Manual No. 40, American Society of Civil
Engineers, New York, 1972.
7.* Johnson, A.I., "Portable Equipment for Borehole Geophysical Exploration",
United States Dept. of the Interior Geological Survey, open-file report,
1963.
8. Palmquist, R., Sendlein, L.L., "The configuration of Contamination Enclaves
from Refuse Disposal Sites on Floodplains", Groundwater, Vol. 13, No. 2,
March-April 1975.
9. Brown, E., Skougstad, M.W., and Fishman, M.T., "Methods for Collection and
Analysis of Water Samples for Dissolved Minerals and Gases". Chap. A-l,
Techniques of Water Resources Investigation of the United States Geological
Survey, Book 5 Laboratory Analysis, 1971
10. Dunlap, W.J., McNabb, J.F., Scalf, M.R., and Cosby, R.L., "Sampling for
Organic Chemicals and Microorganisms in the Subsurface", EPA Report
No. 600/2-77-176, 1977.
11. Wood, W.W., "Guidelines for Collection and Field Analysis of Ground Water
Samples for Selected Unstable Constituents". Chapter D-2, Techniques of
Water Resources Investigations of the United States Geological Survey,
Book 1, Collection of Water Data by Direct Measurements. U.S. Geological
Survey, 24 pp-, 1976.
12. Tate, T.K., "Variations in the Design of Depth Samplers for Use in Ground
Water Studies", Water and Water Engineering, June, 1973, p. 223.
13. Frost, R.C., Bernascone, T.F., and Cairney, T., "A Light-weight and Cheap
Depth Sample", Journal of Hydrology, Vol. 33, No. 1/2, pp. 173-178, 1977.
* Delete this reference in final copy.
300
-------
14. Johnson, A.I., "Portable Equipment for Borehole Geophysical Exploration",
United States Dept. of the Interior Geological Survey, open-file report,
1963.
15. Wood, W.W., "A Technique Using Porous Cups for Water Sampling at any Depth
in the Unsaturated Zone", Water Resources Research, Vol. 9, No. 2,
pp. 486-488, 1973.
16. Wagner, G.H., "Use of Porous Ceramic Cups to Sample Soil Water Within the
Profile", Soil Science, Vol. 94, p. 377-386, 1962.
17. Parizek, R.R. and Lane, B.E., "Soil-Water Sampling Using Pan and Deep
Pressure-Vacuum Lysimeters", Journal of Hydrology, Vol. 11, No. 1,
pp. 1-21, 1970.
18. Wengel, R.W. and Griffin, G. F., "Remote Soil-Water Sampling Technique",
Soil Science Society of America Proceedings, Vol. 35, No. 4, pp. 661-664,
1971
19. U.S. EPA " Procedures Manual for Ground Water Monitoring at Solid Waste
Disposal Facilities", EPA /530/SW-611. Aug. 1977.
20. Trescott, P. C., and Pinder, G. F., "Air Pump for Small-Diameter Piezo-
meters", Ground Water, Vol. 8, No. 3, p. 10-15, 1970.
301
-------
CHAPTER 10
SAMPLING SLUDGES
10.1 BACKGROUND
The quantity and composition of sludge varies with the characteristics
of the wastewater from which it is concentrated and with the concentration
process used. Some common types of sludge are:
1. Coarse screenings from bar racks
2. Grit
3. Scum from primary settling tanks
4. Primary settling tank sludge
5. Return and waste activated sludge
6. Flotation or gravity thickened sludge
7. Aerobic or anaerobic digester sludge
8. Drying bed sludge
9. Vacuum filter cake
10. Sludge press cake
11. Centrifuge sludge
12. Fine screening backwash water
13. Sand filter backwash water
14. Sludges from special treatment processes such as the treatment of
industrial wastes or combined sewer overflows.
Sludge sampling methods are usually confined to water and wastewater
plants, either municipal or industrial. The sampling programs employed are
concerned mainly with the following sludges: primary settling tank sludge,
return and waste activated sludge, thickened sludge, digester sludge, and the
resulting cakes produced by sludge drying methods.
10.2 OBJECTIVES OF SAMPLING PROGRAMS
10.2.1 Process Control
Most sludges are measured for various process control reasons including
the following:
302
-------
1. Optimization of sludge drawoff procedure
2. Determination of the efficiency of a concentration process
3. Determination of the loadings to the process
4. Evaluation of feed material for subsequent sludge conditioning
techniques which may vary with changing feed characteristics
5. Control of the activated sludge process, i.e., the mixed liquor
suspended solids (MLSS) concentration
6. Control of blanket depths in clarifiers
7. Determination of sludge characteristics that may be detrimental
to digester processes
10.2.2 Research
Research projects require specific sampling techniques which are
determined by the program.
10.3 PARAMETERS TO ANALYZE
The parameters to analyze will depend on the objective of the process.
For example, analysis of total and suspended solids content of the sludge is
necessary to determine the efficiency of a sludge thickening process. A guide
for parameters to analyze is shown in Figure 10.1. Additional parameters to
analyze include: heavy metals, pesticides, and nutrients.
10.4 LOCATION OF SAMPLING POINTS
10.4.1 Flowing Sludges
10.4.1.1 Piping
Collect samples directly from the piping through a sampling cock having
a minimum I.D. of 3.8 cm (1.5 in) (1).
10.4.1.2 Channels
Collect samples at the measuring weirs, or at another point where the
sludge is well mixed.
303
-------
10.4.2 Batch Sludges
10.4.2.1 Digesters
Collect samples from a mixed sink which is fed through lines attached at
different levels in the digester. Be certain to waste sludge accumulated in
the lines prior to sampling (1).
10.4.2.2 Tanks
Mix tank thoroughly and collect samples. Or collect samples at various
depths and locations in the tank. Mix samples together prior do analysis.
10.4.3 Specific "In Plant" Locations
The following locations are recommended for sludge sampling at waste-
water treatment plants:
1. Primary Sludge - Draw sludge from the settling tank hoppers into a
well or pit before pumping, mix well and then collect a representative
sample directly from this well. Alternately, collect samples from
openings in pipes near the sludge pumps or from the pump itself (4).
2. Activated Sludge - Collect samples at:
a. the pump suction well
b. the pump or adjacent piping
c. the point of discharge of the return sludge to
the primary effluent.
The sample point should be located in a region of good agitation to insure
suspension of solids (4).
3. Digested Sludge - Collect samples at the point of the discharge of
the digester drawoff pipe to the drying beds or the drying equip-
ment (3) .
4. Bed Dried Sludge - Collect equal sized samples at several points
within the bed without including sand. Mix thoroughly (4).
5. Filtered Sludge - Collect equal size portions (possibly by using a
cookie cutter) at the filter discharge (4).
10.5 FREQUENCY OF SAMPLING
The extreme variability of sludges creates a need for frequent sampling
304
-------
to achieve accurate results. Each composite sample should be composed of at
least 3 individually obtained samples (4). Sample batch operations at the
beginning, middle and end of a discharge, or more frequently if high vari-
ability is suspected (4). Tapped lines should also be sampled in three sepa-
rate intervals because of variations in the sludge at the drawoff source (i.e.,
clarifler, digester, etc.) Minimum frequencies for various sludge processes
are included in Figure 10.1.
10.6 NUMBER OF SAMPLES
The number of samples is determined from the frequency and the number to
include in the composite. Refer to Figure 10.1 for minimum guidelines.
10.7 TYPE OF SAMPLE
Collect grab samples when analyzing an unstable sludge for a parameter
which is affected by the instability, or when analysis is required as soon as
possible (e.g., sludge volume index test for activated sludge samples).
Analysis of composite samples is recommmended in all other situations
to reduce the effects of sludge variability. Use at least three individual
samples to form the composite. Wherever possible, collect frequent discrete
samples and composite according to flow rate (5).
10.8 METHOD OF SAMPLING
Automatic samplers are not commonly available for sludge sampling due to
the high fouling potential and solids content of the wastewater. Use manual
sampling techniques in most situations unless special adaptations can be made.
10.9 VOLUME OF SAMPLE AND CONTAINER TYPE
Use a wide mouth container to sample sludges. The size and material of
container depends on the parameters to be analyzed. In general, a clean
305
-------
3
I
rt 00

-H

H 0)

U
u

3
L
F
U H
CJ »rl
> K
¦H ^
(Q
o -a
Temperature
PH
BOD
SS
IS
TVS
Alkalinity
Volatile Acids
Settleable Solids
2/W
1/D
1/D
1. F - frequency
2. L » location
Su
2/W
1/D
1/D
Su
1/W
1/D
1/D
1/H
2/W
1/D
1/D
1/D
1/D
AD
AD
2/W
2/W
3/W
Mn
1/D
1/W
1/W
1/D
2/W
Mn
I/W
1/W
AD
AD
Mn
1/D
2/0
1/D
1/D
Where

Mn
•
monitor

H
«
hour

D
-
day

week

AD
•
at drawoff

Su
*
subnacanc

1
•
influent

P
-
product aludge
or cake
C
-
centrate

F
-
filtrate

Is
•
In situ

S
*
supernatant or
decant
U
¦
underflow

Figure 10.1
Recommended minimum sampling programs for
municipal wastewater sludge treatment
processes (2).
306
-------
borosilicate glass container is preferable to reduce the possibility of adsorp-
tion of organics to the container wall; however, polyethylene can be used for
non-organic type analyses. See Chapter 17 for more details.
10.10 PRESERVATION AND HANDLING OF SAMPLES
Preservation methods are discussed in Chapter 17. Be certain to complete-
ly mix the sample after a preservative is added to disperse the chemical and
allow adequate preservation. Considerable mixing or homogenization is required
prior to aliquot removal to insure representative portions are obtained.
Further studies on the preservation of sludges appear warranted.
10.11 FLOW MEASUREMENT
For flowing lines do not use flow measuring devices which will be easily
fouled by solids (e.g., orifice, venturi meter). Use a permanently installed
self-cleaning or non-obstructive device such as a magnetic flow meter.
Batch sludge discharges are not easily quantified in terms of volume
discharged. Make estimates from pump capacity, the change in depth in a tank
or well and time of pumping or other appropriate methods.
10.12 REFERENCES
1. Joint Committee of American Society of Civil Engineers and Water
Pollution Control Federation. Sewage Treatment Plant Design -
WPCF Manual of Practice, No 8, 1967.
2. Estimating Laboratory Needs for Municipal Wastewater Treatment Plants.
U.S. EPA, Office of Water Program Operations, Washington, D.C., Report
No. EPA-430/9-74-002» Operation and Maintenance Program. June 1973,
pp. A-l to A-29.
3. Technical Practice Committee - Subcommitte on Operation of Wastewater
Treatment Plants. Operation of.Wastewater Treatment Plant - WPCF
Manual of Practice No. 11, 1970.
307
-------
4. New York State Department of Health. Manual of Instruction for
Sewage Treatment Plant Operators, New York, NY, Health Education
Service, 308 p.
5. Technical Practifce Committee - Subcommittee on Sludge Dewatering
Sludge Dewatering - WPCF Manual of Practice No 20, 1969.
308
-------
CHAPTER 11
SUSPENDED SOLIDS SAMPLING
A key water quality parameter is suspended solids since it impacts upon
such activities as the design of Wastewater Treatment plants, turbidity removal
in drinking water, sediment control in streams, and disinfection. Also, the
concentration of other key water quality parameters is related to suspended
solids since the solid structure may contain biochemical and chemical oxygen
demand materials, trace metals, nutrients, pesticides and toxic-hazardous mate-
rials absorbed on the surface. Therefore, it is imperative that representative
samples be collected for suspended solids.
11.1 REPRESENTATIVE SAMPLING THEORY
For solids distributed uniformly within a given system and containing the
same chemical and physical properties, any sample taken shall be representative.
However, most systems in practice contain suspended solids varying in physical
and/or chemical properties; the degree of non-uniformity in practice ranges from
slight to large and subsequently causes errors in obtaining a representative
sample.
11.1.1 Sampling Error
The error in sampling suspended solids in the field or subsampling from a
previously collected sample is attributed to two factors: (1) solid segregation
effects, and (2) random distribution of solids:
(a) Segregation Effects - Error in sampling due to significant differences
309
-------
between solid particles in specific gravity, size, and shape,
(b) Random Solid Distribution - Error due to imperfect sampling or homo-
genization procedures. For example, a mixture of 1,000 green beads
and 5,000 yellow beads, color being the only difference, is homogen-
ized as completely as possible, however, a sample of 24 beads will
not always contain four green beads but may vary from zero to eight.
The magnitude of this type of error depends on the size of the
sample being withdrawn.
Segregation effects are more pronounced in field sampling since solids are
difficult to mix thoroughly or process through devices that eliminate solid
segregation. Random effects are more pronounced in the laboratory since segre-
gation effects can be minimized by homogenization of the wastewater sample.
11.2 SEGREGATION SAMPLING ERROR
Typical waters/wastewaters contain solid particles which vary in size,
shape, and specific gravity. All of these properties influence the particle
settling rate which must be exceeded to keep the solid suspended and prevent
segregation of solids within the water/wastewater system being sampled. The
theoretical settling rate of a spherical solid in a quiescent aqueous medium
is given by Stokes law:
\ . "2 (y sv>s
18 v
Where: V = settling velocity
s
D = sphere diameter
S = specific gravity of solid
s
Sw = specific gravity of water
v = kinamatic viscosity of water
310
-------
11.2.1 Particle Size
Stokes law indicates that the settling velocity increases with increasing
particle diameter. The size of solids found in water/wastewater vary as shown
in Figure 11.1. Approximately 90% of all solids are less than 1 mm in size.
11.2.2 Specific Gravity of Solids
Stokes law also indicates that the settling rate increases with increasing
specific gravity of the solid. The specific gravity of suspended solids found
in waters/wastewaters vary from 0.8 to 3.5, some of which are shown below:
Specific
Material Gravity
Oils, other organics 0.95
Flocculated mud particles with 95% water 1.03
Municipal
(a) Effluents 1.15
(b) Influent 0.8 - 1.6
(c) Grit 1.2 - 1.7
Aluminum Floe 1.18
Iron Floe 1.34
Sand 2.65
Calcium Carbonate Precipitate 2.70
11.2.3 Shape of Solids
The settling velocity formula of Stokes applies to spherical particles,
however, most waters/wastewaters contain solids of non-spherical shape. In
general solids with irregular shapes tend to settle at lower rates than spherical
particles of the same specific gravity (1). Shapes encountered in waters/waste-
waters include:
311
-------
CODE
T-T Secondary Effluent
® Surface Runoff
O Municipal Grit Chamber Effluent
0 Areated Grit Chamber Effluent
^ Digested Sludge
9 Ohio River Water
0 Combined Storm Sewers
15 20
80 85
30 40 50 60 70
Percentage Less Than By Weight
Figure 11.1 Suspended solid particle sizes in various waters/waste waters (2).
312
-------
Type
Shape
(a) Microbiological and paper scraps
Placoid
(b) Sand grains
Angular
(c) Plastic monomers
Spherical
(d) Fibers - wood, rayon, nylon
Cylindrical-stringy
11.2.4 Settling Velocities
Experimentally determined settling velocities (2) for various
solid types are:
(a) Erosion soil run-off - Ranges from .015 -r 10.1 cm/sec
(.0005 - 0.33 ft/sec).
(b) Grit chamber effluent - Mean of 0,54 cm/sec (.0017 ft/sec).
(c) Primary clarifier design for settable solids removal -
11.2.5 Scouring Velocity
Sampling of horizontal flowing open channels and pipes for suspended solids
must be conducted at velocities which assures adequate mixing. Stratifica-
tion or segregation of solids are classified as follows:
(a) Bed load - Solids that move by saltation, rolling, or sliding along
or near the bottom surface.
(b) Suspended solids or suspended load - solids that are supported by the
upward components of turbulent currents and that stay in suspension for
appreciable amounts of time. The equation for estimating the velocity
(3) to transport solids is:
.028 - .043 cm/sec (.0009 - .0014 ft/sec).
V = (g) (S - 1) Dg » 1,486 R1/6 B (S - 1) Dg
s t n
313
-------
Where:
Vs = Scouring velocity
S = Specific gravity of the particle
Dg = Diameter of particle
B = 0.04 to start scouring and 0.8 for scouring
f = Friction factor - .03 for concrete
n = Manning roughness factor - See Table 11.1
R = Hydraulic Radius - See Table 11.2
g = 32.2 ft/sec^.
TABLE 11.1 VALUES OF MANNING'S ROUGHNESS COEFFICIENT n
Glass, plastic, machined metal 0.010
Dressed timber, joints flush 0.011
Sawn timber, joints uneven 0.01A
Cement plaster 0.011
Concrete, steel troweled 0.012
Concrete, timber forms, unfinished 0.014
Untreated gunite 0.015-0.017
Brick work or dressed masonry 0.014
Rubble set in cement 0.017
Earth, smooth, no weeds 0.020
Earth, some stones and weeds 0.025
Natural river channels:
Clean and straight 0.025-0.030
Winding, with pools and shoals 0.033-0.040
Very weedy, winding and overgrown 0.074-0.150
Clean straight alluvial channels 0.031d*/6
d D-75 size in ft,
314
-------
TABLE 11.2 VALUES OF HYDRAULIC RADIUS FOR
VARIOUS CROSS SECTIONS
% = area of stream cross section; "equivalent diameter" = 4R^
wetted perimeter
Shape of Cross Section RH
Pipes and ducts, running full:
Circle, diam. = D D
4
Annulus, inner diam. = d. outer diam. " D
Square, side = D
Rectangle, sides a,b
Ellipse, major axis = 2a, minor axis = 2b
(D ~ d)
4
D
"5"
ab
2(a + b)
ab
K(a + b)*
Open channels or partly filled ducts:
Rectangle, depth = y, width ¦ b by
b + 2y
D
Semicircle, free surface on a diam. D ^
Wide shallow stream on flat plate, depth »
o
Triangular trough, /, ® 90 , bisector <1 m Y
vertical, depth * y, slant depth = d ^ 2
Trapezoid (depth = y, bottom width * bj: yb + y/ /3~
Side Slope 60° from horizontal b + 4y/
Side slope 45° Y^ +
b + 2
(continued)
315
-------
120
Sampling rates
100
— 0,25 normal
— 0.50 normal
~ — 3.0 normal
Normal sampling rate; ratio
of intake velocity to stream
velocity equals 1.0.
Range of Stokes' Law
o 20
-20
-40 L
0.01
0.02 0.03 0.04 0.06
0.1
0.2
0.3 0.4 0.5
Sediment Size in Millimeters
Figure 11.2 Relation of sediment size to errors in sediment concentration.
.i I,
Standard nozzle
0,15—mm. sediment
Stream.velocity
6 ft,/sec. (Fig. 11)
O — 4 ft./sec.
D — 3 ft./sec.
rO-
=D—
I I n 4 I I , J.
-1 I I L.
0.15 0.2 0.3 0.4 0.6
Relative Sampling Rate —
1.0 1.5 2.0
Intake Velocity
3.0 4.0 5.0
Stream Velocity
Figure 11.3 Effect of stream velocity on errors in sediment concentration.
316
-------
TABLE 11.2 (continued)
Film (thickness = t) on wail of vertical t -t2/D = t (approx.)
wetted wall tower of diameter - D
_ Values of K. If S = (a - b)/(a + b),
S = 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
K = 1.010 1.023 1.040 1.064 1.092 1.127 1.168 1.216 1.273
11.3 FIELD SAMPLING
Collection of suspended solids in the field can be performed manually or
automatically, however significant differences in results can be expected when
sampling non-homogeneous systems such as raw municipal wastewaters as shown
in Table 11.3 (4). Also, automatic samplers with high intake velocities, i.e.
2-10 ft/sec., will capture about 1.5 - 2.0 times more solids than manual flow
proportional or manual grab sampling methods. However, as the system becomes
more homogeneous with respect to solids, i.e., final effluent values in Table
11.3, intake velocities or method of sampling becomes less important in
obtaining comparable results.
Intake velocities above or below stream velocities for suspended sediment
solids (S.g. 2.65) within Stokes law, i.e., Reynolds number less than 1.0, do
not result in any significant error as shown in Figure 11.2 (5). However, as
the particle size increases, significant error occurs when the intake/stream
velocity ratio varies from 1.0. This relationship, Figure 11.3, between the
Relative Sampling Rate Ratio as error in concentration has a negative slope,
i.e., when the intake velocity is less than the stream velocity, more solids
will be collected and when the intake velocity exceeds the stream velocity,
leas solids shall be collected.
The rationale for this inverse relationship is illustrated in Figure 11.4.
Therefore, in order to insure representative sampling, the intake/stream velo-
city ratio should be unity (isokinetic flow).
317
-------

stream lines

sediment path
CFZZZZZZZZZZZZj
intake
6
-------
TABLE 11.3 RICHARDS-GEBAUR SEWAGE TREATMENT PLANT NFS COMPARISON
RATIO OF SAMPLING METHOD VALUE TO MANUAL FLOW VALUE

Sample

Date

Intake
Velocity
Station
Method
May 21
May 2 2
May 23
Average
ft/sec.

QCEC
2.099
1.155
1.755
1.669
2-5

1SC0
0.991
0.431
1.406
0.942
2
Influent
Manual Flow
1.0
1.0
2.0
1.0
—

Manual Grab
1.223
0.697
0.820
0.907
—

Hants
3.141
1.537
1.449
2.042
2.5
Primary
Effluent
Sigmamotor
Manual Flow
0.783
1.0
0.700
1.0
0.968
1.0
0.817
1.0
0.25

Manual Grab
0.981
0.975
1.170
1.042
—

Hants
1.354
0.743
1.387
1.161
2.5
Final
Effluent
Braiisford
Manual Flow
0.822
1.0
0.769
1.0
1.225
1.0
0.939
1.0
.02

Manual Grab
0.951
0.794
1.209
0.985
—
11.4 LABORATORY SUBSAMPLING
Subsampling from previously collected field samples may be subject to
error resulting from segregation effects, i.e., particle size and specific
gravity. As shown in Figure 11.5, the shake and pour technique achieves
93% recovery of solids with specific gravities in the range of 2.2-2.6 and
319
-------
particle sizes less than 50 microns; magnetic stirring improves percent re-
coveries .
Subsampling recoveries of 100 percent for solids having specific gravities
ranging from 1.05-1.14 can be expected up to 500 microns. Therefore, to
insure representative subsamplitig, the entire sample should be thoroughly
blended and as large an aliquot used as possible.
11.5 GUIDELINES FOR SAMPLING OF SUSPENDED SOLIDS
Minimize sampling errors caused by segregation effects by sampling in a
well mixed or turbulent zone.
Minimize random sampling errors in the laboratory by homogenizing the
sample and using as large a sample aliquot as possible.
Maintain the flow rate in the sample lines to effectively transport sus-
pended solids. For horizontal runs, the velocity must exceed the scouring
velocity and in vertical runs, the velocity must exceed the settling velocity
of the particle.
For solids falling within the range of Stokes law, consistant represent-
ative samples can be obtained at intake/stream ratio either greater or less
than 1.0. For solids falling outside Stokes law, an intake/stream ratio of
1.0 is recommended.
The geometry of the intake has little effect upon the representativeness
of the sample, however, the intake should face into the stream at no more than
20 degrees from the direction of stream flow.
320
-------
90
80
70
60
50
40
30
20
10
0
Magnetic Stirring
Shako—Pour
Shake—Pour
Legend
@ and ^ e.g. 2.2 — 2.6 <
[T) s.g. 1.0S - 1.14
500
200
1000
70 100
Particle Slate — Microns
.5 Percent recovery vs particle size during subsampling with different mixing techniques
321
-------
11.6 REFERENCES
1. EPA 600/2-76-006, Design, and Testing of a Prototype Automatic Sewer
Sampling System, March 1976.
2. EPA 670/2-75-011, Physical and Settling Characteristics of Particulates
in Storm and Sanitary Wastewaters, April, 1975.
3. WPCF Manual of Practice, No. 9, ASCE, 1970. p. 88.
4. Harris, D.J., Keffer, W.J., EPA 907/9-74-005, Wastewater Sampling Method-
ologies and Flow Measurement Techniques. 1974.
5. Interagency Committee on Water Resources, 1940. A Study of Methods Used
in the Measurement of Analysis of Sediment Loads in Streams: Laboratory
Investigation of Suspended Sediment Samplers, Report No. 5.
322
-------
CHAPTER 12
SAMPLING, PRESERVATION AND STORAGE CONSIDERATIONS
FOR TRACE ORGANIC MATERIALS
The presence of organic compounds in water and wastewater is regulated
by the Safe Drinking Water Act (SDWA) and the Clean Water Act (CWA).
The SDWA has established maximum contaminant levels (1,2) for the
following organic chemicals:
(a) Chlorinated hydrocarbons:
Endrin Methoxychlor
Lindane Toxaphene
(b) Chlorophenoxys:
2,4-D 2,4,5-TP (sllvex)
(c) Trihalomethanes:
Trichloromethane Bromodichloromethane
Dib romochlo romethane Trib romomethane
Some chemical indicators of industrial contamination are listed in Table
12.1. This list contains chemicals which have been detected in drinking
water supplies and for which the possibility of adverse health effects exists.
The presence of these chemicals is indicative of chemical pollution; this list
is not exhaustive, but serves merely as a guide (3).
A court settlement agreement involving the Natural Resources Defense
Council, et al. and the Environmental Protection Agency (EPA Consent Decree)
resulted in EPA publishing a list of 65 compounds and classes of compounds
(Table 12.2). The Consent Decree required that EPA regulate these compounds
via the Federal Water Pollution Control Act (subsequently amended by the Clean
323
-------
Water Act). EPA's list of 129 priority pollutants (Table 12.3) is an out-
growth of the Consent Decree's list of 65.
Specific toxic pollutant effluent standards will be promulgated for the
129 priority pollutants, thus far they have been promulgated (4,5,6) for the
following:
Analytical procedures for the identification of organic compounds can be
found in a number of publications (7 through 22). However, analytical results
are only meaningful if the sample analyzed is truly a representative sample
of the media you are testing. Chemical analysis for organics present at trace
levels places high demands on sampling techniques.
Aldrin/Dieldrin
Benzidine
DDT (DDD, DDE)
Endrin
Toxaphene
PCB's
324
-------
TABLE 12.1 CHEMICAL INDICATORS OF INDUSTRIAL CONTAMINATION (23)
I. Aliphatic halogenated hydrocarbons:
Methane derivatives:
D? chloromethane
Trichlorofluoromethane
Ethane derivative's;
1.1-dichloroethane
1.2-dichloroethane
hexachloroethane
Unsaturated hydrocarbons:
Trichloroethylene
Tetrachloroethylene
Vinyl chloride
1.1-dichloroethene
Other halogenated compounds:
1.2-dichloropropane
bis(2-chloroisopropyl) ether
II. Cyclic aliphatic compounds:
Chlorinated hydrocarbons:
Lindane
BHC
Cyclodienes:
Chlordane
Aidrin
Dieldrin
Dichlorodifluororoethane
Carbon tetrachloride
1.1.1-trichloroethane
1.1.2-trichloroethane
1,1,2,2-tetrachloroethane
1.2-dichloroethene
1.3-dichloropropene
Hexachlorobutadiene
2-chlorovinyl ether
Bis(2-chloroethyI) ether
Kepone
Toxaphene
Heptachlor
Heptachlor epoxide
Endrin
Hexachlorocyclopentadiene
325
-------
TABLE 12.1 (continued)
III. Aromatic hydrocarbons:
3,4-benzo flu oranthene
benzo(k)fluoranthene
1,12-benzo perylene
Benzenes:
Benzene
Toluene
Xylenes
Halogenated aromatics:
Chlorinated naphthalenes
Chlorobenzene
Dichlorobenzenes
Polychlorinated biphenyls
Pentachlorophenol
Bromobenzene
DDT
Other aromatic hydrocarbons:
Nitrobenzene
Dinitrotoluene
fluoranthene
indeno(1,2,3, c, d)pyrene
benzo(a)pyrene
Ethylbenzene
Propylbenzene
Styrene
DDE
DDD
Chlorophenols
Trichlorobenzenes
4-bromophenylphenyl ether
4-chlorphenylphenyl ether
Hexachlorobenzene
Phthalate esters
Atrazine
326
-------
TABLE 12.2 65 TOXIC POLLUTANTS OR CLASSES OF TOXIC POLLUTANTS (21)
Acenaphthene
Acrolein
Acrylonitrile
Aldrin/Dieldrin
Antimony and compounds
Arsenic and compounds
Asbestos
Benzene
Benzidine
Beryllium and compounds
Cadmium and compounds
Carbon tetrachloride
Chlordane (technical mixture and metabolites)
Chlorinated benzenes (other than dichlorobenzenes)
Chlorinated ethanes (including 1,2 dichloroethane
1,1,1—trichloroethane, and hexachloroethane)
Chloroalkyl ethers (chloromethyl, chloroethyl,
and mixed ethers)
Chlorinated naphthalene
Chlorinated phenols
Chloroform
2-chlorophenol
Chromium and compounds
Copper and compounds
Cyanides
DDT and metabolites
Dichlorobenzenes (1,2-,1,3- and 1,4-dichlorobenzenes)
Dichlorobenzidine
Dichloroethylenes (1,1- and 1,2-dichloroethylenes)
2,4-dichlorophenol
Dichioropropane and dichloropropene
2,4 Dimethylphenol
Dinitrotoluene
Diphenylhydrazine
Endosulfan and metabolites
Endrin and metabolites
Ethylbenzene
Fluoranthene
Haloethers
Halomethanes
Heptachlor and metabolites
Hexachlorobutadiene
Hexachlorocyclohexane ( all isomers)
Hexachlorocyclopentadiene
Isophorone
Lead and compounds
Mercury and compounds
Naphthalene
Nickel and compounds
Nitrobenzene
Nitrophenols (including 2,4-dinitrophenol,
dinitrocresol)
Nitrosamines
Pentachlorophenol
Phenol
Phthalate esters
Polychlorinated biphenyls (PCB's)
Polynuclear aromatic hydrocarbons (including
benzanthracenes, benzopyrenes, benzofluoran-r
thene, chrysenes, dibenzanthracenes and
indenopyrenes)
Selenium and compounds
Silver and compounds
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
Tetrachloroethylene
Thallium and compounds
Toluene
Toxaphene
Tri chloroethylene
Vinyl Chloride
Zinc and compounds
-------
TABLE 12.3 PRIORITY POLLUTANTS
I. Phthalate esters:
IV.
Dimethyl phthalate
Diethyl phthalate
Di-n-butyl phthalate
II. Haloethers
Bis(2-chloroethyl)ether
Bis(2-chloroisopropyl)ether
2-chloroethylvinyl ether
Di-n-octyl phthalate
Bis(2-ethylhexyl)phthalate
Butylbenzyl phthalate
Bis(2-chloroethoxy)methane
4-chlorophcnylphenyl ether
4-bromophenylphenyl ether
Bis(chloromethyl)ether
III. Chlorinated hydrocarbons:
Hexachloroethane 1,3-dichlorobenzene
Hexachlorobutadiene 1,4-dichlorobenzene
Hexachlorocyclopentadiene 1,2,4-trichlorobenzene
1,2-dichlorobenzene Hexachlorobenzene
2-chloronaphthalene
Nitroaromatics and Isophorone:
Nitrobenzene
2,6-dinitrotoluene
Nitrosoaraines:
2,4-dinitrotoluene
Isophorone
N-nit rosodimethylamine N-nitrosodipropylamine
N-nitrosodiphenylamine
VI. Dioxin:
2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
VII. Benzidines:
Benzidine
VIII.Phenols:
3,3-dichlorobenzidine
1,2-diphenylhydrazine
Phenol
2,4-dimethylphenol
2-chlorophenol
2,4-dichlorophenol
2,4,6-trichlorophenol
Pent achloropheno1
4-chloro-3-methylphenol
2-nitrophenol
4-nitrophenol
2,4-dinitrophenol
4,6-dinitro-2-methylphenol
328
-------
TABLE 12.3 (continued)
IX. Polynuclear aromatics:
Acenaphthene
Pluoranthene
Naphthalene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Chrysene
X. Pesticides & PCB's:
Acenaphthylene
Anthracene
Benzo(g,h,i)perylene
Fluorene
Phenanthrene
Dibenzo(a.h)anthracene
Indeno(1,2,3-cd)pyrene
Pyrene
Aldrin
Dieldrin
Chlordane
DDD
DDE
DDT
A-endosulfan
B-endosulfan
Endosulfan
Endrin
Endrin aldehyde
Heptachlor
Toxaphene
Heptachlor epoxide
Alpha-BHC
Beta-BHC
Delta-BHC
Gamma-BHC
Toxaphene
Aroclor 1242
Aroclor 1254
Aroclor 1221
Aroclor 1232
Aroclor 1248
Aroclor 1260
Aroclor 1016
XI. Purgeables:
Benzene
Chlorobenzene
Toluene
Ethylbenzene
Carbon tetrachloride
1,2-dichloroethane
1,1,1-trichloroethane
1,1-dichloroethane
1f1,2-trichloroethane
1,1,2,2-tetrachloroethane
Chloroethane
Chlorodibromomethane
Tetrachloroethylene
Chloroform
1.1-dichloroethylene
1.2-transdichloroethylene
1,2-dichloropropane
1,1-dichloropropylene
Methylchloride
Methylenechloride
Methylbromide
Bromoform
Di chlo rob romomethane
Trichlorofluoromethane
Trichloroethylene
Vinyl chloride
Dichlorodifluoromethane
XII. Acrolein & Acrylonitrile:
Acrolein Acrylonitrile
329
-------
TABLE 12.3 (continued)
XIII. Inorganics:
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Cyanide
Mercury
Nickel
Selenium
Silver
Thallium
Zinc
Asbestos
330
-------
12.1 SAMPLE COLLECTION METHOD
The method of sampling shall be either manual or automatic. Sampling
practices, as specified in Chapter 2, should be followed, except as indicated
in this chapter.
12.1.1 Manual Sampling
The considerations outlined in Chapter 2 are applicable. However, the
sample collector and container material should be borosilicate glass. Collec-
tors and containers constructed of borosilicate glass will minimize sample
contamination. Grab samples obtained for analysis involving purgeable organics
shall be sealed in such a manner as to eliminate entrapped alr(7). This type of
sample,collected without headspace, is illustrated in Figure 12.1.
12.1.2 Automatic Sampling
Although continuous automatic sampling is probably the best method for
collecting truly representative samples, certain precautions must be taken.
Automatic sampling equipment must be free of Tygon and other potential sources
of contamination such as plastic, or rubber components (23). Tygon tubing is
a potential source of phthalate ester contamination. Teflon is acceptable
and may be used in other parts of the sampling system as required.
Automatic samplers used to obtain samples for trace organics analysis
may need special design features. An experimental sampler has been developed
which is capable of collecting grab samples for purgeable organics analysis
and collecting samples on accumulator columns for non-purgeable organics
analyses (24). All system components in contact with the sample are either
constructed of Teflon or glass; this includes a specially designed Teflon-
bellows pump. Various illustrations of this system are shown in Figures 12.2
through 12.6.
331
-------
Sampling systems utilizing carbon or macroreticular resin in columns
have been employed for sampling organic® in ground water (25,26,27). The
accumulator column in these systems is located between the water to be
sampled and the pump, therefore, special Teflon type pumps are not needed.
These type systems are illustrated in Figures 12.7 through 12.10.
Automatic samplers can be used to collect grab samples or composited
grab samples. EPA's 600 series methods for analyzing non-volatile organic
priority pollutants make reference to these types of automatic samplers.
Screw cap
Teflon/Silicon Septum
(Pierce #12722 or equiva—

Convex Meniscus(Sample)
I :
40 mL borosilicate glass
vial (Pierce #13075 or
equivalent)
. i
Figure 12.1 Collection Bottle (21,22)
332
-------
Figure 12.2 Automatic sampler opened to show the 26 purgeable
sample bottles in position.
333
-------
Figure 12.3 A 140 mL purgeable sample bottle for the
automatic sampler.
*
334
-------
Figure 12.4 Automatic sampler opened to show 7 or the 14
accumulator columns. Another bank of 7 is
located behind the visible bank.
335
-------
Figure 12.5 A 1.8 x 27 cm empty accumulator column for
the automatic sampler.
336
-------
Figure 12.6 Automatic sampler pump with container removed.
Teflon bellows are at the bottom.
337
-------
TYOON
TEPLON
TUBE
CARBON
COLUMN
(3•>»")
LAND SURFACE
TUBE
PERISTALTIC
O
PUMP
s
¦Ml
WELL CASING
rTEFLON TUBE
TO CALIBRATED
RECEIVER
WATER TABLE
GROUND WATER
Figure 12.7 Ground water Sampling System (26)
338
-------
GLASS WOOL
PLUG
30 MM
230
MM
$ 50/30 JOINT
Figure 12.8 Carbon adsorption column (27)
339
-------
STOPCOCK
3-WAY TEFLON
-s
GLASS WOOL
PLUG
MM
TEFLON TUBING
6 MM t.D.
CM
TEFLON
CONNECTOR
12 MM 1.0.

3.5
10
y-spect (Nal)
89Sr
50.5d
1.0
0.5
CS and LBBC
90Sr
28.5y
1.0
0.2
CS and LBBC
nij
8.04d
2.0
0.2
CS and LBBC

10.0
0.4
IOR, y-spect

0.4
10
y-spect (Ge)
137Cs
30. Oy
0.4
10
y-spect (Ge)

1.0
0.3
CS and LBBC

3.5
10
y-spect (Nal)
226Ra
1600y
1.0
0.02
RE
228Ra
5.75y
2.0
0.1
CS and LBBC
Ra (total)
—
2.0
0.06
CS and IPC
Gross alpha
—
0.1
0.5
IPC

0.5
0.1
IPC
Gross beta

0.1
2.0
LBBC

0.5
0.5
LBBC
* Calculated at the 99.7 percent (three-sigma) confidence level, based on
1000-minute counting intervals and typical counting efficiencies and instrument
background levels.
Methods:
CS Chemical separation technique (10)
IOR Ion-exchange resin
IPC Internal proportional counter
LBBC Low background beta counter
LSC Liquid scintillation counter
RE Radon emanation and counting by alpha scintillation cell (10)
y-spect Gamma-ray spectroscopy, "Nal" denotes a 10cm X 10cm Nal (Tl)
detector and "Ge" an 85cm3 Ge (Li) detector
360
-------
13.3 FREQUENCY OF SAMPLING
13.3.1 Regulatory
As specified in : 1) license or regulations issued by the NRC or NRC
Agreement State, 2) EPA drinking water standards, or 3) permits from other
governmental agencies.
13.3.2 Surveillance
Frequency of sampling must be based on an evaluation of:
1) types, amounts and potential hazards of radionuclides discharged,
2) their behavior in the environment,
3) waste discharge practices,
4) nature of use of local environment, and
5) the distribution and habits of potentially affected populations (5).
A minimum grab sampling program for surveillance of nuclear power reactors
(4) that may be applicable to other types of facilities recommends the following
minimum frequencies:
1. Surface water - monthly.
2. Ground water, from sources likely to be affected - quarterly.
3. Drinking water supplies - sample at the water intake with a con-
tinous flow proportional sampler. If impracticable, obtain a
monthly grab sample at the reservoir when its holding time exceeds
one month; if less, make sampling frequency equal to reservoir
holding time.
4. Sediment - semiannually.
13.3.3 Other (e.g. testing effectiveness of waste treatment or control methods)
Frequency determined by objectives of investigation.
361
-------
13.4 LOCATION OF SAMPLING
Unless specified in regulatory licenses, requirements or permits, selec-
tion of proper sampling locations is based on judgment (see Section 13.3.2).
As a guide, the EPA recommends for surveillance of light-water reactor sites (4):
1. Surface water - At streams receiving liquid waste, collect one sample
both upstream and downstream of the discharge point. Obtain downstream
sample outside of the restricted area at a location no closer than 10
times the stream width to allow for mixing and dilution. At facility
sites on lakes or large bodies of water, sample near but beyond the
turbulent area caused by discharge. The upstream sample provides data
on background radioactivity. Collect the background sample just above
but beyond any influence by the discharge. Record the discharge flow
rate at the time of sampling.
2. Drinking water - sample all water supplies with intakes downstream and
within 10 miles of a nuclear facility. If none exists, sample the
first water supply within 100 miles.
3. Ground water - necessary when a facility discharges radioactive waste
to pits or trenches. When local ground water is used for drinking or
irrigation, at a minimum, sample the nearest affected well. Subsurface
movement of most radionuclides is retarded by the filtering and ion-
exchange capacity of soil; tritium, however, moves more rapidly with
seepage.
4. Sediment - samples to detect accumulation of undissolved or adsorbed
radionuclides in beds of streams or other bodies of water receiving
liquid effluents from nuclear facilities are collected: 1) downstream
near the discharge outfall but beyond the turbulent area, 2) down-
362
-------
stream of the discharge at locations where sediment is observed to
accumulate, e.g. at bends of streams or dam impoundments, and 3) up-
stream near the discharge outfall but beyond its influence, to deter-
mine background radionuclides.
See also Section 8.4 of this manual for additional guidance in selecting
proper sample locations.
13.5 SAMPLE VOLUME
Determining necessary sample volume depends on the types and number of
analyses to be performed and the sensitivity of available analytical instruments.
For surveillance purposes, obtain the following minimum volumes:
Measurement Volume, liters
Gamma-ray spectroscopy (Nal detector) 3.5*
Gamma-ray spectroscopy (GeLi detector) 0.4
Gross alpha or beta only 0.1
Liquid scintillation - tritium only 0.01
*Water can be subsequently used for analyses requiring chemical
separations (e.g.,89gr 90gr).
Sediment analyses usually require 1 kg. of sample (5).
Obtain larger volumes or weights when sample splitting or replicate
analyses is required for quality control purposes.
13.6 SAMPLE CONTAINERS
Use sample containers constructed of material that minimizes radionuclide
losses by adsorption or other processes during collection and storage. Con-
tainers made of fluorinated hydrocarbon material (e.g. Teflon) are preferred
because of their resistivity to adsorption. Polyethylene and polyvinyl chlo-
363
-------
ride are also recommended (11). Glass and metal containers tend to retain
radionuclides (12). Glass bottles also are more subject to breakage during
handling.
When adsorption problems persist, try: 1) washing container and sampling
apparatus with HC1 or HNO^ before sampling or 2) flushing the container and
apparatus with the liquid to be collected before final sampling (13). Test for
adsorption by analyzing used containers by gamma-ray spectroscopy when this
type of radionuclide emission is present. For other emitters, use successive
acid leachings with hot aqua regia and analyze the leachate (12).
Use caps or container covers that seal tightly to prevent handling losses
and maintain preservation.
Discard container after use to eliminate possibility of cross-contamina-
tion through re-usage. If for economic reasons the more expensive containers
are to be used again, test for adsorbed contamination as described above.
13.7 SAMPLE FILTRATION
Filter water and wastewater sample when the radionuclide contents in
either or both the suspended solids and dissolved matter fractions are to be
determined. Filter as soon as practicable after collection to assure that no
redistribution occurs during storage before analysis (12). Use membrane or
glass fiber filters since these types resist adsorption effects (11). Filter
before adding preservative or other substances to the sample since they can
effect changes in distribution (14).
13.8 SAMPLE PRESERVATION
Radionuclides at very low concentrations (parts per billion, or less),
typical of most environmental water samples, are subject to many little
understood chemical and physical processes (11, 12, 15). Variations in
364
-------
original sample concentration or homogeneity can result from: 1) adsorption
on sampling apparatus, container walls or solid material in the sample (5),
2) co-precipitation of radionuclides due to precipitation of Fe and Mn in
ground water samples exposed to air (15), 3) ionic exchange with components of
glass containers (12), 4) uptake by bacteria, algae or other biological matter
in the sample (13), and 5) formation of colloids (12). Many of these problems
are thought to occur because the amounts of stable isotopes are insufficient
for carrying the radioactive nuclides of the same element (11).
The standard preservation technique for radionuclides in water and waste-
water samples consists of adding concentrated HC1 or HNO3 to obtain a pH of <2
(14,15). Several exceptions exist:
1. Tritium - add no acid; begin analysis immediately upon return to
the laboratory (10).
2. Carbon 14 - see tritium
3. Radiocesiums - use HCl only
4. Radioiodines - see tritium: acid oxidizes iodides to iodines
which are rapidly lost through volatilization (12). For samples
containing 3jj, 14^ or 131-j- along with radionuclides requiring
preservatives, obtain duplicate samples and add acid to only one.
Add acid preservative after sample collection (but not before filtration-
see Section 13.7) or as soon as practicable but do not delay beyond 5 days (14).
When acid preservation is not desirable: 1) add iaotopic carriers of the
same elements as the radionuclides (12), 2) refrigerate samples at or near
their freezing temperature to retard chemical reaction rates and to inhibit
bacterial growth (16).
Samples of bottom sediments require no preservation additive (13).
365
-------
13.9 GENERAL SAMPLING PROCEDURE - WATER AND WASTEWATER
The following procedure summarizes the elements of good practice for
collecting and preserving samples of water and wastewater for radionuclide
measurements. These guidances apply to the situation where no unusual circum-
stances exist:
1. Flush sample lines, equipment or other apparatus and sample con-
tainer with sample medium to minimize adsorption effects. Use
type of containers recommended in Section 13.6.
2. Avoid floating debris and bottom sediments when sampling surface
waters. When aliquoting large samples containing significant
amounts of suspended solids, vigorously shake or mix to assure
representative subsamples.
3. Wash sampling apparatus with distilled water to minimize con-
tamination of subsequent samples.
A. Filter sample as soon as practicable after collection when radio-
nuclide distribution in soluble and/or insoluble phases is to be
determined (See Section 13.7). Use membrane or glass fiber
filters.
5. Add preservative of the required type to liquid samples (see
Section 13.8). When concentrated HC1 or HNO3 is the indicated
type, add to obtain a pH of <2. In cases of mixtures of radio
nuclides, for some (3jj, 14^, 131j) of which acid preservation is
not recommended, collect replicate samples and treat only one
with acid.
6. Seal sample container tightly. Complete sample data label includ-
ing time of collection for decay corrections.
366
-------
7. Analyze samples containing short-lived radionuclides as soon as
possible.
8. Discard sample containers after use or test for contamination if
expensive types of containers are to be used again.
13.10 RADIATION SAFETY
Storage of large numbers or volumes of samples containing radioactivity is
a potential source of exposure to workers occupying the area. However, this is
unlikely with environmental samples due to low radionuclide content. If in
doubt, survey the area periodically with a beta-gamma survey instrument, e.g.,
a Geiger-Mueller (GM) meter. Note that sample containers reduce all alpha-
particle and much beta-particle radiation. If levels above instrument back-
ground occur at work stations, consult a radiation safety specialist for
procedures to reduce levels and for proper disposal techniques when samples are
no longer needed.
367
-------
13.11 REFERENCES
1. U.S. Nuclear Regulatory Commission, "Standards for Protection Against
Radiation"* Title 10, Code of Federal Regulations, Part 20, Federal
Register, U.S. Government Printing Office, Washington, D.C. (1975).
2. Office of Water Supply, U.S. Environmental Protection Agency, "National
Interim Primary Drinking Water Regulation", Report No. EPA-570/9-76-003,
U.S. Government Printing Office, Washington, D.C. (1977)
3. Report by Committee 4 of the International Commission on Radiological
Protection, "Principles of Environmental Monitoring related to the Handling
of Radioactive Materials", ICRP Publication 7, Pergamon Press, Oxford (1965).
4. Office of Radiation Programs, U.S. Environmental Protection Agency,
"Environmental Radioactivity Surveillance Guide", EPA Report No.
ORP/SID 72-2 (1972).
5. National Council on Radiation Protection and Measurements, "Environmental
Radiation Measurements", NCRP Report No. 50 (1976).
6. DeVoe, J. R., " Radioactive Contamination of Materials Used in Scientific
Research", National Academy of Sciences - National Research Council, Nuclear
Science Series Report No. 34 (1961)
7. Martin, M.J. (ed.), "Nuclear Decay Data for Selected Radionuclides", Oak
Ridge National Laboratory Report. ORNL-5114 (March 1976).
8. Martin, M.J. and P.H. Blichert-Toft, "Radioactive Atoms", Nuclear Data
Tables A8, Nos. 1-2 (1970).
9. Lederer, C.M., Hollander, J.M., and Perlman, I., Table of Isotopes, John
Wiley, New York (1967).
10. Krieger, H.L., "Interim Radiochemical Methodology for Drinking Water", U.S.
Environmental Protection Agency, Report No. EPA-600/4-75-008 (Revised)
368
-------
CHAPTER 14
COLLECTING AND HANDLING MICROBIOLOGICAL SAMPLES
14.1 BACKGROUND
Fecal contamination from warm-blooded animals and man is present in
certain industrial effluents and in urban and rural runoff municipal waste-
waters, and can cause serious diseases and other health problems in drinking
water supplies, in waters used for recreation, agriculture, or in the food,
dairy and beverage industries. Consequently, the Federal Water Pollution Con-
trol Act Amendments (Clean Water Act), the Marine Protection, Research, and
Sanctuaries Act (Ocean Dumping), and the Safe Drinking Water Act require moni-
toring of water supplies, ambient waters and wastewater effluents for compli-
ance with bacterial limits (1, 2 and 3).
In order to control pathogens discharged into these waters, selected
groups of microorganisms are monitored as indicators of the sanitary quality
of a stream or water supply. These include "total" bacteria (standard plate
count), total coliform bacteria, fecal coliform bacteria, and fecal strepto-
cocci, as well as the pathogens themselves: Salmonella, Shigella, Giardia,
Pseudomonas, Klebsiella, Pneumoniae, Clostridium spp, Viruses, etc.
14.2 ANALYTICAL METHODOLOGY
The bacterial parameters: Standard Plate Count, Total Coliform, Fecal
Coliform, Fecal Streptococci and Salmonella will be discussed.
For a more detailed description of the methodologies see Standard
Methods and the EPA Manual (4,5).
370
-------
14.2.1 Standard Plate Count
The Standard Plate Count (SPC) Method is a direct quantitative measure-
ment of the viable aerobic and facultative anaerobic bacteria in a water en-
vironment that are capable of growth on plating medium. This test is usually
performed by suspension and growth of the sample in agar (pour plate) but may
be done as surface growth on a spread plate or membrane filter procedure.
Although no one set of plate count conditions can enumerate all organisms
present, the Standard Plate Count Method provides the uniform technique re-
quired for comparative testing and for monitoring water quality in selected
applications.
This simple technique is a useful tool for determining the bacterial
density of potable waters for quality control studies of water treatment
processes. The Standard Plate Count provides a method for monitoring changes
in the bacteriological quality of finished water throughout a distribution
system to indicate the effectiveness of chlorine in the system as well as the
possible existence of cross-connections, sediment accumulations and other prob-
lems within the distribution lines. The procedure may also be used to monitor
quality changes in bottled water or emergency water supplies.
14.2.2 Coliforms
The coliform or total coliform group includes all of the aerobic and
facultative anaerobic, gram-negative, nonspore-forming, rod-shaped bacteria
that ferment lactose in 24-48 hours at 35°C in a multiple-tube most probable
number (MPN) procedure or that produce a golden-green metallic sheen within
24 hours at 35°C in the membrane filter (MF) procedure. The definition in-
cludes the genera: Escherichia, Citrobacter, Enterobacter, and Klebsiella.
The coliform group may be subdivided into the two following categories:
371
-------
1. Coliforms normally of fecal origin (primarily Escherichia coli
types).
2. Coliforms usually associated with vegetation and soils (Citrobacter„
Enterobacter, Klebsiella, and Escherichia spp), which may occur in
fecal matter but in smaller numbers than _E. coli.
The two analytical techniques recommended by EPA and Standard Methods
for enumeration of coliforms are the MPN and the Single-Step, Two-Step and
Delayed Incubation MF Methods (4,5)
Although microbiological standard for public water supplies and drink-
ing waters are based on total coliform numbers which include coliform from
sources other than human and animal feces, the trend in recent years is to
provide a more accurate estimate of the sanitary quality of the water tested
by conducting fecal coliform analyses.
14.2.3 Fecal Coliform Methods
The fecal coliforms are part of the total coliform group. They are
defined as gram-negative nonspore-forming rods that ferment lactose in 24 ±
2 hours at 44.5 ± 0.2°C with the production of gas in the multiple-tube
procedure or produce acidity with blue colonies in the membrane filter pro-
cedure.
The major species in the fecal coliform group is Escherichia coli, a
species indicative of fecal pollution and the possible presence of enteric
pathogens. No method is presently available which distinguishes human fecal
coliforms from those of other warm-blooded animals.
The analytical techniques for identifying fecal coliforms in water are
the direct MF, the delayed-incubation MF and the multiple-tube, MPN methods.
The test is applicable to the examination of lakes and reservoirs, wells
372
-------
and springs, public water supplies, natural bathing waters, secondary non-
chlorinated effluents from sewage treatment plants, farm ponds, storm-
water runoff, raw municipal sewage, and feedlot runoff. The MF test has been
used with varied success in marine waters.
14.2.4 Fecal Streptococci
The term, fecal streptococci, is used to describe the streptococci
which indicate the sanitary quality of water and wastewater. The fecal
streptococci group includes the serological groups D and Q: Streptococcus
faecalis, S. faecalis subsp. liquifaclens, S. faecalis subsp. zymogenes,
S. faecium, S. bovis, S. equinus, and S. avium.
The MF, MPN and direct pour plate procedures can be used to enumerate
and identify fecal streptococci in water and wastewater.
Fecal streptococci data verify fecal pollution and may provide addi-
tional information concerning the recency and probable origin of pollution.
In combination with data on coliform bacteria, fecal streptococci are used
as a supplement to fecal coliform analyses when a more precise determination
of sources of contamination is necessary. The occurrence of fecal strepto-
cocci in water indicates fecal contamination by warm-blooded animals. They
are not known to multiply in the environment.
Further identification of streptococcal types present in the sample
may be obtained by biochemical characterization. Such information is useful
for source investigations. For example, S.bovis and S. equinus are host
specific and are associated with the fecal excrement of non-human warm-
blooded animals. High numbers of these organisms are associated with pol-
lution from meat processing plants, dairy wastes, and run-off from feedlots
and farmlands. Because of limited survival time outside the animal intesti-
373
-------
nal tract, their presence indicates very recent contamination from farm
animals.
14.2.5 Salmonella
The genus Salmonella is comprised of a large number of serologically
related, gram-negative, nonspore-forming bacilli that are pathogenic for warm-
blooded animals including man, and which are found in reptiles, amphibians and
mammals. They cause enteritis and enteric fevers via contaminated water, food
or food products. Because Salmonella are responsible for many outbreaks of
waterborne disease, increased efforts have been made to identify and enumer-
ate them.
Generally the numbers of Salmonella present are small, so that a larger
sample volume (> a liter) is required to isolate this pathogen than for coli-
form and fecal coliform analyses. Since negative result does not assure ab-
sence of Salmonella or other pathogens, analyses for indicator organisms are
usually run concurrently, to measure the potential health risk..
Recommended methods for recovery of Salmonella from water and waste-
water and their subsequent identification are presented in Standard Methods
and the EPA Manual (4,5). The methods are particularly useful for recrea-
tional and shellfish-harvesting waters. No single method of recovery and
identification of these organisms from waters and wastewaters is appropri-
ate for all sampling situations. Rather the method is selected based on
the characteristics of the sample and the microbiologist's experience with
the procedures. Multiple option techniques are described for sample con-
centration, enrichment, isolation and identification.
14.2.6 Enteric Viruses (4)
Viruses excreted by animal and man also pollute waters. These
374
-------
viruses are present in domestic sewage even after waste treatment and enter
streams and lakes that serve as the source of water for many communities.
Viruses are excreted much lower in numbers than coliform bacteria, and do not
multiply outside of the animal or man host. Dilution in ambient waters, sew-
age treatment, and water treatment further reduce viral numbers in the envi-
ronment. However it has been demonstrated that infection can be produced by
a few viral units.
Sample concentration is needed to demonstrate and quantitate viruses
in clean or potable waters because the numbers are quite low. For clean
waters, 400 liters or more of water must be sampled to detect viruses. The
most promising method for concentrating small quantities of viruses from
those waters is adsorption onto a microporous filter. Viruses are removed
from the filter with a protein eluant or glycine buffer at a controlled pH.
Viruses may be reconcentrated a second time.
Measuring viruses in wastewaters and natural waters is even more
difficult because of suspended solids. For such samples, the aqueous poly-
mer two-phase separation technic may be used directly for virus recovery but
the sample size is limited to 2 to 4 liters.
After concentration of viruses and elution, the eluate is analyzed by
cell culture or whole animal assay.
At this time, the routine examination of the waters and wastewaters
for enteric viruses is not recommended. However, for special needs such as
wastewater reuse, disease control, or special studies, virus testing can be
done but only by competent virologists with proper facilities.
14.3 SAMPLE BOTTLE PREPARATION £4,5)
Sample bottles must be resistant to sterilizing conditions and the
375
-------
solvent action of the water. Wide-mouth, screw-cap or ground-glass stop-
pered glass bottles, or heat-resistant plastic bottles (preferably polypropy-
lene) may be used if they can be sterilized without producing toxic materials.
Screw capped bottles must be equipped with neoprene rubber liners or other
materials that do not produce bacteriostatic or nutritive compounds upon
sterilization.
14.3.1 Selection and Cleansing of Bottles
Select bottles of sufficient capacity to provide a volume necessary
for all analyses anticipated. Use at least a 125 mL bottle for a minimum
sample volume of 100 mL and to provide adequate mixing space. Discard bottles
which have chips, cracks, and etched surfaces. Bottle closure must be cap-
able of creating a water-tight seal. Before use, thoroughly clean bottles
and closures with detergent and hot water and rinse with hot water to remove
all traces of detergent. Then rinse three times with a good quality labora-
tory pure water. A test for bacteriostatic or inhibitory residues on glass-
ware is described in Standard Methods and in EPA's Manual (A,5).
14.3.2 Use of Dechlorinating and Chelating Agents
Use a dechlorinating agent in the sample bottle when water and waste-
water samples containing residual chlorine are anticipated. Add 0.1 mL of
a 10 percent solution of sodium thiosulfate to each 125 mL (4 oz) sample
bottle prior to sterilization.
Use a chelating agent when waters are suspected of containing more
than 0.01 mg/L concentration of heavy metals such as copper, nickel, zinc,
etc. Add 0.3 mL of a 15 percent solution ethylene diamine tetra-acetic acid,
tetra sodium salt (EDTA), to each 125 mL (4 oz.) sample bottle prior to
sterilization (6,7).
376
-------
14.3.3 Wrapping of Bottles
Protect the tops and necks of glass stopper bottles from contamination
by covering them with aluminum foil or kraft paper before sterilization.
Screw-cap closures do not require a cover.
14.3.4 Sterilization of Bottles
Autoclave glass or heat-resistant polypropylene plastic bottles at
121°C for 15 minutes. Glassware may be sterilized in a hot air oven at 170°C
for two hours. Ethylene oxide gas sterilization is acceptable for plastic
containers that are not heat resistant. Before use, store sample bottles
sterilized by gas overnight to allow the last traces of gas to dissipate.
14.4 SAMPLING METHODS AND EQUIPMENT (5)
These methods are applicable for sampling potable water, streams and
rivers, recreational waters such as bathing beaches and swimming pools, lakes
and reservoirs, marine and estuarine waters, shellfish harvesting waters,
and domestic and industrial waste discharges.
In no case should a composite sample be collected for microbiological
examination. Data from individual samples show a range of values which
composite samples will not display. Individual results give information
about industrial process variations. Also, one or more portions that make
up a composite sample may contain toxic or nutritive material and cause
erroneous results.
Collect samples by hand if possible* If depth samples are required
or if the sampling sites are difficult to access such as bridges or banks
adjacent to surface waters use a sampling device.
377
-------
Do not rinse bottle with sample, but fill it directly to within 2.5-5 cm
(1-2 in.) from the top for proper mixing of the sample before analysis. Use
caution to avoid contaminating the sample with fingers, gloves or other mater-
ials.
Locate and then carefully identify the sampling site on a field log
sheet and on a chain of custody tag, if this is required, and on a label.
(See Chapter 15).
14.4.1 Tap Sampling
Do not collect samples from spigots that leak or that contain aeration
devices or screens. In sampling direct connections to water main, flush the
spigot for five minutes to clear the service line. For wells equipped with
hand or mechanical pumps, run the water to waste for five minutes before the
sample is collected. Remove the cap aseptically from the sample bottle.
Hold the sample bottle upright near the base while it is being filled. Avoid
splashing. Replace bottle closure and hood covering.
14.4.2 Surface Sampling by Hand
Collect a grab sample directly into a sample bottle prepared as des-
cribed in Section 14.3. Remove the bottle top cover and closure and protect
them from contamination. Avoid touching the inside of the closure. Grasp
the bottle securely at the base with one hand and plunge it mouth down into
the water, avoiding surface scum. Position the bottle towards the current
flow and away from the hand of the collector, the shore, the side of the
sampling platform, or boat. The sampling depth should be 15 to 30 cm (6-12
in.) below the water surface. If the water body is static, an artificial
current can be created by moving the bottle horizontally in the direction it
is pointed and away from the sampler. Tip the bottle slightly upwards to
378
-------
allow air to exit and the bottle to fill. After removal of the bottle
from the stream, tightly stopper and label the bottle.
14.4.3 Surface and Well Sampling by Weighted Bottle Frame
When sampling from a bridge or other structure above a body of water.
Place the bottle in a weighted frame that holds the bottle securely. Remove
the cover and lower the device to the water. It is preferable to use nylon
rope which does not absorb water and will not rot. Aim the bottle mouth up-
stream by swinging the sampling device first downstream, and then allow it
to drop into the water, without slack in the rope. Pull the sample device
rapidly upstream and out of the water, simulating the scooping motion of
grab sampling. Take care not to dislodge dirt or other material from the
sampling platform.
If sampling a well that does not have pumping machinery, use a weighted
sterilized sample bottle. Avoid contaminating the sample with surface scum
or dislodged material from the sides of the well.
14.4.4 Depth Sampling
Several additional devices are needed for collection of depth samples
from lakes, reservoirs, estuaries and the oceans. These depth samplers re-
quire lowering the sample device and/or container to the desired depth, then
opening, filling, and closing the container and returning the device to the
surface. Although depth measurements are best made with a pre-marked steel
cable, the sample depths can be determined by premeasuring and marking a
nylon rope at intervals with non-smearing ink, paint, or fingernail polish.
The following list of depth samplers is not inclusive but can serve as a
guide: The ZoBell J-Z, the Niskin, the New York Dept. of Health, and the
379
-------
Kemmerer samplers.
14.4.5 Sediments and Sludge Sampling
Microorganisms attach to particles and artifacts in water and are
found in large numbers at the bottom sediment/interfaces in any body of water.
Sewage solids in treated domestic wastewaters and sludges contain very large
numbers of microorganisms which pass into receiving streams, lakes and oceans
and then settle into the bottom sediments. This is a particular concern in
the ocean dumping program because of the concentrated disposal of very large
amounts of sludge in selected ocean dump sites. Microorganisms in these mater-
ials are periodically released into the overlying waters as the bottoms are
disturbed.
Sediments and bottom materials are difficlt to sample because of the
variable composition, size, density and shape of particles and the lack of
homogeneity. They vary from light, fluffy particles to compacted high den-
sity, solid layers.
Grab samples are not usually satisfactory for quantitative bottom
sampling because they may contain material which is not representative. How-
ever, they give an indication of the processes that occur.
Corers are used in quantitative work though none is entirely satis-
factory. The Eckman corer is used when sampling from small boats. The Wild-
life Co. (Saginaw, Michigan) coring device is used in shallow water (15 meters
or more). In extremely shallow water a lucite tube can be inserted into the
sediment by hand, and capped by a stopper. The Van Donsel-Geldreich sampler
can be used to collect soft sediments or muds in relatively deep waters. It
uses a sterile plastic bag in a weighted frame to collect the sample and
then closes the bag with a wire loop.
380
-------
14.5 SAMPLE FREQUENCY AND SITE SELECTION (5)
14.5.1 Frequency of Sampling
The frequency of sampling depends upon the type of pollution that is
to be measured. Cyclic pollution and its duration are measured as frequently
as practical immediately downstream from the source. Uniform pollution loads
are measured at greater distances downstream from the source and at less fre-
quent time intervals than cyclic pollution. A common approach for short-term
studies is to collect samples from each site daily and advance the sampling
intervals one hour during each 24-hour period to obtain data for a 7-10 day
study.
Often the numbers of samples to be collected are specified by NPDES
permits, drinking water regulations, or by State requirements. Some stand-
ards require a minimum number of samples to be collected each month. Other
standards are less explicit and simply indicate that the geometric mean coli-
form density shall not exceed a certain level each month, with no more than
10%, 20%, etc. of samples exceeding a certain value. Where the number of
samples required is undetermined, a sufficient number should be collected to
measure the variations in conditions.
14.5.2 Raw Water Supplies
Reservoirs, and lakes used as water supplies, are sampled at inlets,
other possible sources of pollution, the draw-off point, the quarter point
intervals around the draw-off point at about the same depth, and the reser-
voir outlet.
14.5.3 Potable Water Supplies
Coliform standards for potable water supplies established by Public
381
-------
Health Service Act of 1962 were amended by the Safe Drinking Water Act of
1974 (SWDA) and its amendments (8). The levels for the 1962 PHS Standards
were retained in the SDWA but were redefined as Maximum Contaminant Levels
(MCLs). As with the previous standards, the MCLs emphasize the importance
of collecting samples at regular intervals, in numbers proportionate to the
population served, and at points representative of conditions in the distri-
bution system. A set protocol was established for repeat sampling when posi-
tive coliform results occur. For application of the MCLs, the frequency of
sampling and the location of sampling points is established jointly by the
utility, the Reporting Agency, and the Certifying Authority.
The SDWA also specifies that any laboratory generating data for public
water supplies, as required under the Act, must be certified according to
the procedures and criteria in the Laboratory Certification Manual (9). The
laboratory facility, personnel, equipment and instrumentation, sampling meth-
odology, quality control, data reporting and necessary action responses are
specified.
14.5.4 Distribution Systems
Sample locations should be representative of the distribution system
and include sites such as municipal buildings, public schools, airports and
parks, hydrants, restaurants, theaters, gas stations, industrial plants and
private residences. A systematic coverage of such points in the distribution
system should detect contamination from breaks in water lines, loss of pres-
sure, or cross-connections. The sampling program should also include special
sampling locations such as dead-end distribution lines that are sources of
bacterial contamination, and far reaches of the distribution lines where
chlorine residual may have dissipated.
382
-------
The minimum number of samples which must be collected and examined each
month is based upon the population density served by the distribution system.
Samples should be collected at evenly spaced time intervals throughout the
month. In the event of an unsatisfactory sample, repetitive samples must be
collected until two consecutive samples yield satisfactory quality water.
Check samples from any single point or special purpose samples must not be
counted in the overall total of monthly samples.
Standard Sample: The standards for microbiological quality are based upon the
number of organisms allowable in a standard sample. A standard sample for the
membrane filter technique is at least 100 mL. For the MPN test, a standard
sample consists of five standard portions of either 10 mL or 100 mL.
14.5.5 Lakes and Impoundments
Sampling points in a recreational impoundment or lake should include
inlets, sources of pollution, grids or transects across the long axis of the
water body, bathing areas and outlets.
14.5.6 Stream Sampling
The objectives of the initial survey dictate the location, frequency and
number of samples to be collected.
A. Selection of Sampling Sites: A typical stream sampling program in-
cludes sampling locations upstream of the area of concern, upstream
and downstream of waste discharges, upstream and downstream from
tributary. Downstream sites should be located far enough below entry
of discharge or tributary to allow thorough mixing. For more com-
plex situations, where several waste discharges are involved, sam-
pling includes sites upstream and downstream from the combined dis-
charge area and samples taken directly from each industrial or muni-
383
-------
cipal waste discharge. Using available bacteriological, chemical and dis-
charge rate data, the contribution of each pollution source can be determined.
B. Small Streams: Small streams should be sampled at background sta-
tions upstream of the pollution sources and at stations downstream
from pollution sources. Additional sampling sites should be located
downstream to delineate the zones of pollution. Avoid sampling
areas where stagnation may occur (backwater of a tributary) and
areas located near the inside bank of a curve in the stream which
may not be representative of the main channel.
C. Large Streams and Rivers: Large streams are usually not well mixed
laterally for long distances downstream from the pollution sources.
Sampling sites below point source pollution should be established
to provide desired downstream travel time and dispersal as deter-
mined by flow rate measurements. Particular care must be taken to
establish the proper sampling points at: the upper reach control
station, non-point sources of pollution, waste discharges as they
enter the stream, quarter-point samples below the pollution sources
to detect channeling, tributaries, and downstream from tributaries
after mixing. Occasionally, depth samples are necessary to deter-
mine vertical mixing patterns.
14.5.7 Recreational Waters
A. Selection of Sampling Sites: Select sampling sites which reflect
the quality of water throughout the recreational area. Boat mar-
inas, waste drainage from dry well restrooms and other public build-
ings, any upstream flows from impounded rivers or drainages into
lakes, reservoirs or impounded streams, as well as the lake or body
384
-------
of water itself should be sampled.
Sampling sites at bathing beaches or other recreational areas
should include upstream or peripheral areas and locations adjacent
to natural drains that would discharge stormwater, or run-off areas
draining septic wastes from restaurants, marinas, or garbage collec-
tion areas.
Swimming pool water should be monitored at least daily during maxi-
mum use periods, preferably at the overflow. It is important to
test swimming pool samples for neutralization of residual chlorine
at pool side to assure that the dechlorinating agent was effective.
Depths: Sampling in bathing areas should be standardized at 1 foot
for shallow depths and at 3 feet for swimming depths.
G. Frequency and Time: Collect samples daily during high-use seasons.
Select high use days (Fridays, weekends and holidays) and sample
during peak period of the day, generally in the afternoons. Sample
estuarine waters at high tide, low tide and ebb tide to obtain a
measure of the cyclic changes in water quality.
14.5.8 Domestic and Industrial Waste Discharges
When it is often necessary to sample secondary and tertiary wastes from
municipal waste treatment plants and various industrial waste treatment opera-
tions, sampling must be adjusted to meet the specific situation. If plant
treatment efficiency varies considerably, collect grab samples around the
clock at selected intervals for a three to five day period. If it is known
that the process displays little variation, fewer samples are needed. The
NPDES has established treatment plant effluent limits for wastewater discharg-
ers. These are often based on maximum and mean values. A sufficient number
385
-------
of samples must be collected to satisfy the permit and/or provide statistic-
ally sound data and give a fair representation of the bacteriological quality
of the discharge (10).
14.5.9 Marine and Estuarine Sampling
Sampling marine and estuarine waters requires the consideration of other
factors in addition to those usually recognized in fresh water sampling. They
include tidal cycles, current patterns, bottom currents and counter-currents,
stratification.climatic conditions, seasonal fluctuations, dispersion of dis-
charges and multi-depth samplings.
The frequency of sampling varies with the objectives. When a sampling
program is started, it may be necessary to sample every hour around the clock
to establish pollutional loads and dispersion patterns. The sewage discharges
may occur continuously or intermittently.
When the sampling strategy for a survey is planned, data may be avail-
able from previous hydrological studies done by Coast Guard, Corps of Engin-
eers, National Oceanic and Atmospheric Administration (NOAA), U.S. Geological
Survey, or university and private research investigations. In a survey, float
studies and dye studies are often used to determine surface and undercurrents.
Initially depth samples are taken on the bottom and at five feet increments
between surface and bottom. A random grid pattern for selecting sampling
sites is established statistically.
A. Marine Sampling: In ocean studies, the environmental conditions
are most diverse along the coast where shore, atmosphere and the
surf are strong influences. The shallow coastal waters are particu-
larly susceptible to daily fluctuations in temperature and seasonal
changes. Sampling during the entire tidal cycle or during a half
386
-------
cycle may be required. Many ocean studies such as sampling over the
continental shelf involve huge areas where no two areas are the same.
Selection of sampling sites and depths are most critical in marine
waters. In winter, cooling of coastal waters can result in water
layers which approach 0°C. In summer, the shallow waters warm much
faster than the deeper waters. Despite the higher temperature, oxy-
gen concentrations are higher in shallow than in deeper waters due
to greater water movement, surf action and photosynthetic activity
from macrophytes and the plankton.
Moving from the shallow waters to the intermediate depths, one ob-
serves a moderation of these shallow water characteristics. In the
deeper waters, there is a marked stabilization of conditions. Water
temperatures are lower and more stable. Deep waters have limited
turbulence, little penetration of light, sparse vegetation, and a
layer of silt and sediment covering the ocean floor.
B. Estuarine Sampling: When a survey is made on an estuary, samples
are often taken from a boat, ususally making an end to end traverse
of the estuary. Another method involves taking samples throughout
a tidal cycle, every hour or two hours from a bridge, or from a boat
anchored at a number of fixed points.
In a large bay or estuary where many square miles of area are involved,
a grid or series of stations may be necessary. Two sets of samples
are usually taken from an area on a given day, one at ebb or flood
slack water, and the other three hours earlier, or later, at the
half tide interval. Sampling is scheduled so that the mid-sampling
time of each run coincides with the calculated occurrence of the
tidal condition.
387
-------
In locating sampling sitas, one must consider points at which tribu-
tary waters enter the main stream or estuary, location of shellfish
beds, and bathing beaches. The sampling stations can be adjusted as
data accumulate. For example, if a series of stations one-half mile
apart consistently show similar values, some stations may be dropped
and others added in areas where data shows more variability.
Considerable stratification can occur between the ocean's salt water
and fresh river water. It is essential when starting a survey of an
unknown estuary to find out whether there is any marked stratifica-
tion. This can be done by chloride determinations at different lo-
cations and depths. It is possible for stratification to occur in
one part of an estuary and not in another.
On a flood-tide, the more dense salt water pushes up into the less
dense fresh river water causing an overlapping, with th.e fresh
water flowing on top and forming the phenomenon called a salt water
wedge. As a result, stratification occurs. If the discharge of pol-
lution is in the salt water layer, the contamination will be concen-
trated near the bottom at the flood tide. The flow or velocity of
the fresh water will influence the degree of stratification which
occurs. If one is sampling only at the surface, it is possible that
the data will not show the polluted underflowing water which was con-
taminated at a point below the fresh water river. Therefore, where
stratification is suspected, samples at different depths will be
needed to measure vertical distribution.
C. Shellfish-Harvesting Waters: Water overlying shellfish-harvesting
areas should be sampled during periods of most unfavorable hydro-
388
-------
graphic conditions, usually at low tide after heavy precipitation.
However, shellfish beds are sometimes exposed during low tide and
must be sampled during other tidal conditions. Procedures for samp-
ling of shellfish and water in shellfish growing areas are governed
by the National Shellfish Sanitation Program's Manual of Operations
(11).
14.6 PRESERVATION AND TRANSIT OF SAMPLES (4,5)
The adherence to sample preservation and holding time limits is critical
to the production of valid data. Samples exceeding these limits should not
be analyzed. The following rules must be observed.
14.6.1 Storage Temperature and Handling Conditions
Bacteriological samples should be iced or refrigerated at a temperature
of 1-4°C during transit to the laboratory. Insulated containers are prefer-
able to assure proper maintenance of storage temperature. Care should be
taken that sample bottles are not totally immersed in water during transit
or storage.
14.6.2 Holding Time Limitations
Although samples should be examined as soon as possible after collec-
tion, they should not be held longer than six hours between collection and
initiation of analyses (12). This limit is applied to fresh waters, seawaters
and shellfish-bed waters. The exception is water supply samples mailed in
from water treament systems. Current regulations permit these samples to he
held up to 30 hours.
Although a holding time of six hours is permitted sewage samples, or-
ganically-rich wastes and marine waters are particularly susceptible to rapid
389
-------
increases or die-away and should be held for the shortest time possible, to
min imiz e change.
If the specified holding time limits cannot be observed the following
alternatives should be considered:
A. Temporary Field Laboratories; In situations where it is impossible
to meet the six hour maximum holding time between collection and
processing of samples, consider the use of temporary field labora-
tories located near the collection site.
B. Delayed Incubation Procedure: If sampling and transit conditions
require more than six hours, and the use of field laboratories is
impossible,consider the delayed incubation procedures for total and
fecal coliforms and fecal streptococci.
C. Public Transportation: Occasionally, commercial forms of transit
such as airlines, buslines or couriers are used to transport samples
contained in ice chests to the laboratory. These should be consid-
ered only when storage timetemperature requirements and the proper
disposition of the samples can be assured.
390
-------
14.7 REFERENCES
1. Federal Water Pollution Control Act Amendments of 1972, Public Law
92-500, October 18, 1972, 86 Stat. 8 16, 33 United States Code (USC)
Sec. 1151.
2. Marine Protection, Research and Sanctuaries Act of 1972, Public Law
92-532, October 23, 1972, 86 Stat. 1052.
3. Safe Drinking Water Act, Public Law 93-523, December 16, 1974, 88
Stat. 1660, 42 United States Code (USC) 300f.
4. Standard Methods for the Examination of Water and Wastewater, 14th
Edition, Washington, D.C., 1975, 1193 pp.
5. Bordner, R.H., J.A. Winter, and P.V. Scarpino, ed., Microbiological
Methods for Monitoring the Environment. U.S. EPA, Environmental
Monitoring and Support Laboratory, Cincinnati, EPA 600/8-78-017,
December, 1978.
6. Shipe, E.L. and A. Fields. Comparison of the Molecular Filter Tech-
niques with Agar Plate Counts for the Enumeration of E. Coli in Vari-
ous Aqueous Concentrations of Zinc and Copper Sulfate. Appl. Micro.
2:382, 1954.
7. Shipe, E.L. and A. Fields. Chelation and a Method for Maintaining the
Coliform Index in Water Supplies. Public Health Reports. 71:794,
1956.
8. 40 CRF 141, National Interim Primary Drinking Water Regulations,
December 24, 1975, pp. 59566-59585.
9. Manual for the Interim Certification of Laboratories Involved in
Analyzing Public Drinking Water Supplies, Criteria and Procedures,
EPA 600/8-78-008, U.S. EPA, EMSL-Cincinnati, May, 1978.
10. 40 CFR 136, Guidelines Establishing Test Procedures for Analysis of
Pollutants, October 16, 1973, pp. 28758-28760, December 1, 1976, pp.
52780-52786, and further amendments.
11. Hauser, L.S. (ed.), 1965, National Shellfish Sanitation Program. Manual
of Operations, Part 1: Sanitation of shellfish growing areas. U.S.
Public Health Service, Washington, D.C.
12. Public Health Laboratory Service Water Subcommittee, 1953. The effect
of storage on the coliform and Baoterium Coli counts of water samples.
Storage for six hours at room and refrigerator temperatures. J. Hyg.
51:559.
391
-------
CHAPTER 15
SAMPLE IDENTIFICATION AND CHAIN OF CUSTODY PROCEDURES
The successful implementation of a monitoring program depends on the
capability to produce valid data and to demonstrate such validity (1). In
addition to proper sample collection, preservation, storage and handling
appropriate sample identification and chain of custody procedures are neces-
sary to help insure the validity of the data.
15.1 SAMPLE IDENTIFICATION
15.1.1 Sample Number
Assign each sample container a unique number for identification in the
field and laboratory. The identification number should have as few digits as
possible to discourage abbreviation. The following guidelines should facili-
tate proper identification:
1. Use preprinted rolls of peel back labels assigned from the laboratory
to a sampling crew.
2. For relatively small numbers of samples use sequential numbering and
affix a label to each bottle. When the sample is placed in two or
more containers, assign two or more numbers to that sample. For
large numbers of samples such as encountered in river, lake, or
estuary sampling, use a five digit number, the first two numbers in-
dicating the week of the year. When a sample is split into two or
more parts, use one sample number and apply a color coded label to
each sample which indicates the type of preservative added.
-------
Therefore, once the type of preservative has been indicated, the
general group of parameters to be analyzed on that sample is estab-
lished. For example, a blue label indicates that nitric acid has
been added, therefore, the analyst could obtain an aliquot from this
sample for metal analysis.
3. Note the date and preservative on the label.
4. Note additional information in the field notebook.
15.2 CHAIN OF CUSTODY (2)
15.2.1 General
The regulatory body must be able to demonstrate the reliability of its
evidence in pollution cases by proving the chain of possession and custody of
any samples which are offered for evidence or which form the basis of analyt-
ical results introduced into evidence in any water pollution case. Therefore,
it is imperative that each regulatory body and its laboratory prepare written
procedures to be followed whenever evidence samples are collected, transferred,
stored, analyzed, or destroyed. The primary objective of these procedures is
to create an accurate written record which can be used to trace the possession
of the sample from the moment of its collection through its introduction into
evidence. The following guidelines on the chain of custody procedures are
provided. However, in those cases where state chain of custody procedures
apply, follow the procedures which satisfy state rules or laws for the intro-
duction of evidence into enforcement or judicial proceedings.
A sample is in custody if it is:
1. In actual physical possession, or
2. In view, after being in physical possession, or
393
-------
3. In physical possession and locked up so that no one could tamper
with it.
15.2.2 Sample Collection
1. Limit handling the sample to as few people as possible.
2. Obtain samples using the guidelines in this handbook.
3. Attach sample tags, (Figure 15.1), securely to the sample container
at the time the sample is collected. The tag should contain as a
minimum: station number and location, date, time taken, type of
sample, sequence number (first sample of the day-sequence No. 1,
second sample-sequence No. 2 etc.), analyses required and the name
of the person taking the sample. The tag must be filled out
legibly in waterproof ink.
4. Record field measurements and other pertinent information in a
bound field notebook to refresh the memory of the sampling personnel
in the event that a witness is required at an enforcement proceeding.
A separate set of field notebooks should be maintained for each survey
and stored in a safe place where they can be protected and accounted
for at all times. Establish a field data record format (Figure 15.2)
to minimize field entries or possible omissions. The following
information should be included:
date field measurements such as:
time temperature
Survey name conductivity
type of samples taken DO
volume of each sample pH
type of analyses flow
sample numbers other pertinent information
sample location or observation
394
-------
(FRONT SIDE)
o

CHAIN
OF CUSTODY RECORD

Sample No.
Time Taken (hrsi
Date
Taken
Source of Sample
Sreservatlve
Sample Collector
W!tness (es)
Remarks: (Analyses Requires,
Sample Type, etc.)
(BACK SIDE)
l hartby certify thai 1 received thl» i*mpte and dlepoted of It at noted below:
ffc
Received from
Date Received
Time Received
4) k
OCT
O
£
Disposition of Sample
Signature
1 haraby ctrtlfy that 1 racalvad chit sampl* and dlspoaad of ft *i notad below:
°*
oo
-e
4)«0
OOO
4)
*
deceived from
Date Received
Time Received
Jisposition of Sample
Signature
1 haraby ctrtlfy that 1 obtained this »««plt and dWpltchad It as ihown balowi
±-
0
fZ ft
u—
+* c
? *
o«te Obtained Time Obtained
Source
Oate Dispatched Tim* Dispatched Method of Shipment
D
Sent to
Signature
Figure 15.1 Chain of Custody Record Tag
395
-------
Samplers;
FIELD DATA RECORD
Statioa
Sanple Dace of \
Number1 Collection,
t 1
Time
pH

Other Parameters
Sample; Sample
Taken 'Received
temperature

I I 1

1 7

i
!
|

Figure 15.2 Sample - Field Data Record
396
-------
The entries should be signed by the person taking the sample. Assign
a survey coordinator or designated representative the responsibility
for preparing and retaining field notebooks during and after the sur-
vey .
5. The sample collector is responsible for the collected samples until
they are properly dispatched to a receiving laboratory or turned
over to an assigned custodian. He must assure that each container
is in his possession or view at all times or is stored in a locked
place where no one can tamper with it.
6. Take color slides or photographs of the outfall sample location and
any visible water pollution. Document in writing on the back of the
photo the following information: signature of the photographer,
time, date, and site location. Photographs of this nature should be
handled according to the established Chain of Custody procedures to
prevent alteration.
15.2.3 Transfer of Custody and Shipment
In transfer of custody, each custodian of samples must sign, record and
date the transfer. Regulatory agencies may develop their chain of custody
procedures tailored to their needs. These procedures may vary in format and
language but should contain the same essential elements regarding sample
identification and chain of custody procedures. Historically, sample transfer
under chain of custody has been on a sample by sample basis which is awkward
and time-consuming. However, EPA's National Enforcement Investigation Center
(NEIC), Denver, has set a precedent with its bulk transfer of samples. Bulk
transfer is speedier, reduces paperwork and the number of sample custodians.
The following description of chain of custody is similar to that of NEIC -
Denver (3).
397
-------
Samples must be accompanied by a Chain of Custody Record which,
includes the name of the survey, collector's signature, station
number, station location, date, time, type of sample, sequence
number, number of containers and analyses required (Figure 15.3) When
turning over the possession of samples, the transferor and trans-
feree must sign, date, and record time on the sheet. This record
sheet allows transfer of a group of samples in the field to the
mobile laboratory or to other designated laboratories. When a
custodian transfers a portion of the samples identified on the
sheet to the field mobile laboratory, the individual samples must be
noted in the column with the signature of the person relinquishing
the samples. The field laboratory person receiving the samples
should acknowledge the receipt by signing in the appropriate column.
If a custodian has not been assigned, the field custodian or the
sample collector has the responsibility for packaging and dis-
patching the samples to the laboratory for analysis. The "Dispatch"
portion of the Chain of Custody Record must be filled out, dated,
and signed.
Samples must be carefully packed in shipment containers such as ice
chests, to avoid breakage. The shipping containers must be locked
for shipment to the receiving laboratory.
Packages must be accompanied by the Chain of Custody Record showing
identification of the contents. The original must accompany the
shipment. A copy is retained by the survey coordinator.
If samples are delivered to the laboratory when appropriate personnel
are not there to receive them, the samples must be locked in a
398
-------
designated area within the laboratory in a manner so that no one can
tamper with them. The same person must then return to the laboratory
and unlock the samples and deliver custody to the appropriate
custodian.
15.2.4 Laboratory Custody Procedures
1. The laboratory must designate a "sample custodian" and an alternate
to act in his absence. In addition, the laboratory must set aside
as a "sample storage security area" an isolated room with sufficient
refrigerator space, which can be secured locked from the outside.
2. Samples should be handled by the minimum number of people.
3. The custodian should receive the incoming samples and indicate
receipt by signing the Chain of Custody Record Sheet accompanying
the samples and retaining the sheet as a permanent record. Couriers
picking up samples at the airport, post office, etc. must sign
jointly with the laboratory custodian.
4. Immediately upon receipt, the custodian must place samples in the
sample room which should be locked at all times except when samples
are removed or replaced by the custodian. To the maximum extent
possible, only the custodian should be permitted in the sample room.
5. The custodian shall maintain the intergity of the sample by appro-
priate storage.
6. The custodian must distribute samples to the personnel who are to
perform tests.
7. The analyst must record information in his laboratory notebook or
analytical work sheet, that describes the samples, the procedures
performed and the results of the tests. The notes must be retained
399
-------
CHAIN OF CUSTODY RECORD
Survey Samplers; (Signature)
Station
Number
Station Location
Date
Time
Sample Type
Seq,
No.
No. of
Containers
Analysis
Required
Water
Air
Comp,
Crab

Relinquished by; (Signature)
Received by: (Signature)
Date/Time
1
Relinquished by: (Signature)
"Received by: (Signature)
Date/Time
I
Relinquished by; (Signature)
Received by: (Signature)
Date/Time
Relinquished by: (Signature)
Received by: (Signature)
DaU'/Time
1
Dispatched by: (Signature) Date/Time Received for Laboratory by:
i
Date/Time
_L_
Method of Shipment:
Distribution: Original - Accompany Shipment
1 copy - Survey Coordinator Field Files
Figure 15.3 Sample - Chain of Custody Record
A 00
-------
as a permanent record in the laboratory and should include any
abnormalities which occurred during the testing procedure. In the
event that the person who performed the tests is not available as a
witness at the time of the trial, the regulator/ agency may be able
to introduce the notes in evidence under the Federal Business Records
Act.
8. Standard methods of laboratory analyses must be used as described in
the "Guidelines Establishing Test Procedures for Analysis of
Pollutants," 38 F. R. 28758, October 16, 1973. If laboratory
personnel deviate from standard procedures, they should be prepared
to justify their decision during cross-examination.
9. Laboratory personnel must be responsible for the care and custody of
a sample once it is handed over to them and should be prepared to
testify that the sample was in their possession and viewed or secured
in the laboratory at all times from the moment it was received from
the custodian until the tests were run.
10. Once the sample testing is completed, the unused portion of the
sample together with all identifying tags, the laboratory records,
and other documentation of work must be returned to the custodian.
11. Samples, tags and laboratory records of tests may be destroyed only
upon the order of the Laboratory Director, who will first confer
with the Chief, Enforcement Specialist Office, to make certain that
the information is no longer required.
401
-------
15.3 REFERENCES
1. Crim, R. L., Editor. Model State Water Monitoring Program. EPA-440/9-
74-002, U.S. Environmental Protection Agency, Washington, D.C. 1974,
pp.
2. In Press. Microbiological Methods for Monitoring the Environment/Water
and Wastewater, U.S. Environmental Protection Agency, Cincinnati, Ohio.
1976.
3. Anon. Compliance Monitoring Procedures. U.S. Environmental Protection
Agency, Denver, Colorado, 1975. 61 pp.
402
-------
CHAPTER 16
QUALITY ASSURANCE
Quality assurance is an integral part of all sampling programs. The ob-
jectives of quality asssurance are to assure that the data generated is:
1. Meaningful 4. Precise
2. Representative 5. Accurate
3. Complete 6. Comparable
Data must be representative of the condition being monitored. To enable
comparison with different data and with stated program objectives, data must
be presented in standard units. Quality assurance for a sampling program
should address all elements from sample collection to data reporting while
permitting operational flexibility. A quality assurance plan should include,
as an essential part, a continuing education and training program for the
personnel involved in the monitoring program. This will enhance quality assur-
ance capabilities and aid in keeping pace with the scientific advancement
occuring in the field.
16.1 OBJECTIVES
For the implementation of an effective and meaningful quality assurance
program it is imperative that its objectives are well defined, documented and
cover all activities that affect the quality of the data. Such written ob-
jectives are needed to assure:
1. Effective participation in the quality assurance program by various
403
-------
personnel in different organizations involved in a sampling program.
2. Uniform thinking and rationale among the personnel participating in
a sampling program.
3. Appropriate action at all levels among participating organizations.
4. Integrated and planned course of action.
5. Performance evaluation against stated objectives.
To meet the above objectives, one individual within the organization
should be designated the Quality Assurance (QA) Coordinator. The QA Coordin-
ator should undertake activities such as quality planning, auditing, and re-
liability. The QA Coordinator should also have the responsibility for coord-
inating all quality assurance activity so that complete integration of the
quality assurance plan is achieved.
16.2 ELEMENTS OF A QUALITY ASSURANCE PLAN (1)
The quality assurance plan will contain the following elements:
1. A policy to establish parameter analytical criteria (accuracy, pre-
cision, detection limit) for monitoring activities. Field, sample
handling, and test procedures are best established only after estab-
lishment of criteria.
2. A systematic policy for selection and use of measurement and sampling
methodology. Where available, approved methodology must be used.
Where alternate methodology is necessary or where approved methodolo-
gy does not exist, the quality assurance plan should state how the
alternate or new methodology will be documented, justified, and ap-
proved for agency use.
3. Documentation of operating procedures. The QA Coordinator should es-
tablish the format for the procedures and see that the documentation
404
-------
is done.
4. Intra-office quality assurance audits or acceptance criteria.
The QA Coordinator as part of the documented methodology or operating
procedures will approve or specify the intra-office audits. Detailed
quality assurance procedures are necessary for:
Personnel selection.
Sample site selection.
Sample collection, handling and preservation.
Calibration and maintenance of instruments and equipment (field
and laboratory)»
Intra-office audits (field and laboratory) for data acceptance
with documentation for agency data credibility.
Review and approval of data before they are released.
Scheduled intra-office audits (field and laboratory) through the
QA Coordinator to assess the accuracy of field and laboratory
methodology•
An audit by the QA Coordinator on a systematic basis to see that
all the above activities are being done.
16.3 PERSONNEL TRAINING (1)
Successful implementation of a quality assurance plan ultimately depends
upon the competence of the monitoring personnel. All personnel involved in
any function affecting data quality (sample collection, analysis, data re-
duction and quality assurance) should have sufficient training in their ap-
pointed jobs to contribute to the reporting of complete and high quality data.
The quality assurance plan should therefore provide for periodic assessment of
training needs and should describe the manner in which training is to be ac-
complished, This will include both in-house and external training and educa-
tion.
405
-------
Several methods of training are available to promote achievement of the
desired level of knowledge and skill required. The following are the train-
ing methods most commonly used in the pollution control field:
16.3.1 On the Job Training (OJT)
An effective OJT program could consist of the following:
Observe experienced professionals perform the different tasks in
the measurement process.
Perform tasks under direct supervision of an experienced profession-
al .
Perform tasks independently but with adequate quality assurance
checks.
16.3.2 Short-term Course Training
A number of short-term courses (normally two weeks or less) are available
that provide knowledge and skills to more effectively implement the NPDES mon-
itoring program. Course schedules can be obtained from the EPA Training Cen-
ters.
16.3.3 Long-term Course Training
Numerous universities, colleges, and technical schools provide long-term
(quarters or semester length) academic courses in wastewater treatment, ana-
lytical chemistry, environmental engineering, and other disciplines.
16.3.4 Training Evaluation
The quality assurance plan needs to address training evaluation.
Training should be evaluated in terms of (1) the level of knowledge and
skill achieved by the operator from the training, and (2), the overall effec-
tiveness of the training (including determination of training areas that need
improvement).
A good means of measuring skill improvement is to assign the trainee a
work task. Accuracy and/or completeness are commonly used indicators to asses
406
-------
the trainee's proficiency. The tasks should be similar to the following
forms:
1. Sample Collection. Trainee would be asked to list or preferably
perform all steps in a sample collection for a hypothetical or real
case. This would include selection of sample site, duration and
frequency of sampling, type of samples collected(grab or composite),
sampling and flow measuring equipment that would provide the highest
quality data. In addition, the trainee would be asked to perform
selected calculations. Proficiency would be judged in terms of com-
pleteness and accuracy.
2. Analysis. Trainee would be provided unknown samples for analysis
normally measured in the field. As defined here, an unknown is a
sample whose concentrations are known to the work supervisor (OJT)
or the training instructor (short-term course training) but unknown
to the trainee. Proficiency would be judged in terms of accuracy.
16-4 QUALITY ASSURANCE IN SAMPLING
As a first step for quality assurance in sample collection the sampling
program should delineate the details on sampling locations, sample type, sample
frequency, number of samples, duration of sampling, sample volume, sample col-
lection methods, equipment to be used for the sample collection, sample con-
tainers, pretreatment of containers, type and amount of preservative to be used
blanks, duplicates/triplicates, spiked samples, replicates, chain of custody
procedures, and any other pertinent matter which will have a bearing on the
quality assurance in sample collection and handling. Guidelines, on the above
can be found In this manual.
Despite a well defined sampling program, appropriate sampling and field
407
-------
TABLE 16.1 QUALITY ASSURANCE PROCEDURES FOR FIELD ANALYSIS AND EQUIPMENT (1)
Parameter
General
Daily
Quarterly
o
00
1. Dissolved
Oxygen
Membrane
Electrode
Winkler-
Azide
Method
Enter the make, model, 1. Calibrate meter using
serial and/or ID num-^
ber for each meter in
a log book.
Record data to near-
est 0.1 mg/1.
manufacturere1s instruc-
tions or Winkler—Azide
method.
2. Check membrane for air
bubbles and holes.
Change membrane and
KC1 if necessary.
3. Check leads, switch con-
tacts etc. for corrosion
and shorts if meter
pointer remains off scale«
Duplicate analysis should
be run to check the pre-
cision of the analyst.
Duplicate values should
agree within ±0.2 mg/1.
Check instrument calibration
and linearity using a series
of at least three dissolved
oxygen standards.
2. pH
Electrode
Method
Enter the make, model
serial and/or ID num-
ber for each meter in
a log book.
1. Calibrate the system
against standard buffer
solutions of known pH
value at the start of a
sampling run.
Take all meters to the lab-
oratory for maintenance, cal-
ibration and quality control
checks.
-------
TABLE 16.1 (continued)
Parameter General Daily Quarterly
2. pH (continued)
2. Periodically check the
buffers during the sample
run and record the data
in the log sheet or book.
3. Be on the alert for erratic
meter response arising from
weak batteries, cracked
electrode, fouling etc.
4. Check response and linear-
ity following highly acidic
or alkaline samples. Allow
additional time for
equilibration.
4. Check against the closest
reference solution each
time a violation is found.
5. Rinse electrodes thorough-
ly between samples and after
calibration.
3. Conductivity
Enter the make, model,
serial and/or ID num-
ber for each meter in
a log book.
1. Standardize with KC1 stand-
ards having similar specific
conductance values to those
anticipated in the samples.
Calculate the cell constant
using two different standards
1. Take all meters to lab
for maintenance, cali-
bration and quality
control checks.
2. Check temperature
compensat ion.
3. Check date of last plat-
inizing and replatinize
if necessary.
(continued)
-------
TABLE 16.1 (continued)
Parameter
General
Daily
Quarterly
3. Conductivity
(continued)
(continued)
Cell Constant =
Standard Value
Actual Value
Specific Conductance =
.Reading multiplied by
Cell Constant
Rinse cell after each
sample to prevent carry-
over.
4. Analyze NBS or EPA
reference standard
and record actual vs.
observed readings in
the log.
4. Residual
Chlorine
Amperometric
Titration
Enter the make, model,
ID and/or serial num-
ber of each titration
apparatus in a log
book. Report results
to nearest 0.01 mg/1
Refer to instrument
manufacturer's instruc-
tions for proper opera-
tion and calibration
procedures.
Return instrument to
lab for maintenance
and addition of fresh,
standardized reagentst
t Biweekly
-------
TABLE 16,1 (continued)
Parameter
General
Daily
Quarterly
5. Temperature
Manual
Thermistors:
thermographs
etc.
Enter the make, model,
serial number and/or
ID number and tempera-
ture range for each
thermometer. All standard-
ization shall be against an
NBS or NBS calibrated ther-
mometer. Readings should
agree within 1°C. If en-
forcement action is anti-
cipated, calibrate the
thermometer before and after
analysis. All data shall be
read to the nearest 1°,
Report data between 0°-9°C
(32°-48°F) to one signifi-
cant figure; between 10°-
99°C (50°-210°F) to two
significant figures.
1. Check for air spaces
or bubbles in the
column, cracks, etc.
Compare with a known
source if available.
Enter the make, model,
serial and/or ID number of
the instrument in a log
book. All standardization
shall be against an NBS or
NBS calibrated thermometer.
Reading should agree with-
in 1°C.
Check thermistor or
sensing device for
response and operation
according to the manu-
facturer's instructions.
Check at two temperatures
against an NBS or equiva-
lent thermometer. Enter
data in a log book. Temp-
erature readings shall
agree within 1°C or the
thermometer shall be re-
placed or recalibrated.
Accuracy shall be deter-
mined throughout the ex-
pected working range 0°
to 50°C (32°to 120 F) A
minimum of three tempera-
tures within the range
should be used to verify
accuracy. Preferable
ranges are; 5°-10°, 15°-
25°, 35°—45°C.*(41o-50°,
59°-77°, 95°-113°F)
Accuracy shall be deter-
mined throughout the ex-
pected working range of 0°
to 50°C (32° to 120°F) A
minimum of three tempera-
tures within the range
should be used to verify
the accuracy.
* Initially and Bi-annually
(continued)
-------
TABLE 16.1 (continued)
Parameter
General
Daily
Quarterly
5. Temperature (continued)
Thermistors;
The rmog raphs
etc. (cont.)
If enforcement action is Record actual vs. standard
anticipated refer to the temperature in log book,
procedure listed in Manual
above.
Preferable ranges are:
5°-10°, 15°-25°, 35°-45°C.
(41°t-50° , 59°- 77°, 45°-
113°F)*
6. Flow Measure-
ment
7. Automatic
Samplers
Enter the make, model,
serial and/or ID number
of each flow measurement
instrument in a log
book.
Enter the make, model
serial and/or ID num-
ber of each sampler in
a log book.
Install the device in
accordance with the manu-
facturer's instructions
and with the procedures
given in this manual.
Affix record of calibra-
tion by NBS, manufacturer
or other, to the instru-
ment log.§
Check intake velocity vs.
head (minimum of three
samples) and clock time
setting vs. actual time
interval.
* Initially and Bi-annually
§ Annually
-------
testing procedures, errors crop up due to equipment malfunction which
adversely affects the quality. Therefore, as a second step for quality
assurance, procedures should be developed for routine testing, maintenance
and calibration of the equipment. Manufacturer's instructions are appropriate
guides on these procedures. These procedures should establish routine main-
tenance, testing and calibration intervals, set up written procedures for
maintenance, testing and calibration, list the required calibration standards,
determine the environmental conditions requiring calibration, and generate a
documentation record system. Equipment should be labeled to indicate the
calibration data and when the calibration or maintenance expires. Table 16.1
contains a listing of quality assurance guidelines for field analysis, equip-
ment calibration and documentation. (1)
As a third step in quality assurance, random control checks should be
performed to make sure that appropriate sampling guidelines on sample col-
lection, handling and chain of custody are followed by the field personnel;
and deviations, if any, are rectified. Analytical quality control as an aid
to quality assurance can be performed through duplicate, split, and spiked
samples; sample preservative blanks, precision, accuracy and control charts.
For more details on analytical quality control refer to EPA's "Handbook for
Analytical Quality Control in Water and Wastewater" (2).
16.5 REFERENCES
1. U.S. EPA, NPDES Compliance Sampling Manual, Compliance Branch, (Draft),
May,4, 1977.
2. U.S. EPA Handbook for Analytical Quality Control in Water and Waste-
water Laboratories. Technology Transfer, June 1972.
413
-------
CHAPTER 17
SAMPLE PRESERVATION
Immediate analysis at the sampling site will preclude the need for
sample preservation, however this procedure is not practical in most
situations. Therefore, sample preservation and other related aspects of
sample handling should be established to maintain the representativeness of
the sample until analysis.
Complete and unequivocal preservation of samples, either domestic
sewage, industrial wastes, or natural waters, is a practical impossibility.
Regardless of the nature of the sample, complete stability for every
constituent can never be achieved. At best, preservation techniques can
only retard the chemical and biological changes that take place in a sample
after the sample is removed from the parent source. To maintain the
integrity of the sample.appropriate selection of containers, pretreatment
of containers if necessary and the holding times form the integral part
of the sample preservation program.
17. 1 METHODS OF PRESERVATION
Methods of preservation are relatively limited and are intended
generally to:
1. Retard biological action.
2. Retard hydrolysis of chemical compounds and complexes.
3. Reduce volatility of constituents.
Preservation methods are generally limited to chemical addition, pH
control, refrigeration, and freezing. Combinations of these methods are of-
414
-------
ten used for the preservation of the sample.
17.1.1 Chemical Addition
The most convenient preservative is a chemical which can be added to a
sample bottle prior to sampling. When the sample is added, the preservative
disperses immediately, stabilizing the parameter(s) of concern for long
periods of time. When a preservative added to preserve some of the para-
meters interferes with other parameters, collect additional samples for those
parameters. For example, concentrated nitric acid added for the preservation
of some of the metals would interfere with BOD, so an additional sample must
be collected for BOD.
17.1.1.1 pH Control
pH control to preserve the sample is dependent upon chemical addition.
As an example, to keep metal ions in a dissolved state concentrated nitric
acid is added to lower the pH to less than 2.
17.1.2 Freezing
Freezing has been the subject of many preservation studies (1-16). It is
felt by some that freezing would be a method for increasing the holding
time and allowing collection of a single sample for all analysis. However,
the residue solids components (filterable and nonfilterable) of the sample
change with freezing and thawing (8). Therefore, return to equilibrium
and then high speed homogenization is necessary before any analysis can be
run. This method may be acceptable for certain analysis but not as a
general preservation method.
17.1.3 Refrigeration
Refrigeration (or icing) has also been studied with various results (10-12,
17-21). This is a common method used in field work and has no detrimental
415
-------
effect on sample composition. Although it does not maintain integrity for
all parameters, it does not interfere with any analytical methods.
17.1.4 Preservation Guidelines
For NPDES samples, the permit holder must use sample preservation and
holding times for different parameters (organic and inorganic) as per the
proposed guidelines published in the Federal Register(22) and shown in Table
17.1. Use of alternative preservation methods is permissible if the preser-
vation effectiveness is demonstrated by supporting data.
Table 17.2 provides additional references and furnishes data on preser-
vation methods, storage and holding times for different parameters found in
various literature sources. However, for a specific application of the data,
reference to the original publication should be made.
17.1.5 Alternative Preservation Methods
Alternative preservation methods with specific preservatives can be used
if the preservation effectiveness of the parameters can be demonstrated by
supporting data. One way of obtaining the supporting data is through preser-
vation studies. Such preservation studies must include details on the
following:
1. Type of water/wastewater used as a sample in the experiment
2. Type of containers used.
3. Pretreatment of the container and the glassware used.
4. Preservation methods used.
5. Specific temperatures or temperature range used.
6. Duration of storage.
7. Stored in light or darkness.
416
-------
TABLE 17.1 CONTAINERS, PRESERVATION, AND HOLDING TIMES
Measurement3
1 Acidity
2 Alkalinity
3 Ammonia
Container
Preservative0
Maximum
Holding Timec
BACTERIA
4-7 Coliform, fecal
8 Fecal streptococci
9 Biochemical oxygen demand
10 Biochemical oxygen
demand Carbonaceous
11 Bromide
12 Chemical oxygen demand
13 Chloride
14 Chlorinated organic
compounds
P,G
P,G
P,G
P,G
P,G
P,G
P.G
P,G
f,G
P,G
G,teflon-lined
cap
Cool 4°C
Cool 4°C
Cool 4°C
R2SO4 to pH<2
Cool 4°C
0.008% Na2S203 8
Cool 4°C
0.008% Na2S203 8
Cool 4°C
Cool 4°C
None Required
Cool 4°C
H2S04 to pH<2
None Required
Cool 4°C
0.008% Na2S203
g
14 days
14 days
20 days
6 hours
6 hours
48 hours
48 hours
28 days
28 days
28 days
48 hours(until
extraction)
30 days(after
extraction)
(Continued)
-------
TABLE 17.1 (cont'd)
Measurement
Container
Preservative
Maximum
Holding Time^
15 Chlorine, total residual
16 Color
17-18 Cyanide, total and
amendable to chlorination
19 Dissolved oxygen
Probe
Winkler
P,G
P,G
P,G
P,G
G bottle & top
G bottle & top
Determined on site
Cool, 4°C
Cool, 4°C
NaCR to pR>12
0.008% Na2S203
Determine on site
Fix on site
2 hours
48 hours
14 days
1 hour
8 hour
-p-
i—
00
20 Fluoride
21 Hardness
22 Hydrogen ion (pH)
23 & 92 Kjeldahl and organic
nitrogen
METALS e
40-41 Chromium VI
58-59 Mercury
24-87 Metals
P
P,G
P,G
P,G
P,G
P»G
P,G
None Required
HNO^ to pH<2
Determine on site
Cool, 4°C
H2SO4 to pH<2
Cool, 4 C
HNO3 to pH<2
0.05% K2Cr207
HNO3 to pH<2
28 days
6 months
2 hours
28 days
88 Nitrate
P,G
Cool, 4°C
48 hours
28 days
6 months
48 hours
(Continued)
-------
TABLE 17.1 (Cont'd)
Measurements'
Container
Preservative0
Maximum
Holding Time^
•is
f—
vO
88(a)1
Nitrate-nitrate
89 Nitrite
90 Oil and Grease
91 Organic Carbon
P,G
P »G
G
P,G
93-206
ORGANIC COMPOUNDS]
Extractables (including
phthalates, nitrosamines
organochlorine pesticides,
PCB's, nitroaromatics,
isopho rone, polynuclear
aromatic hydrocarbons, halo-
ethers, chlorinated hydro-
carbons and TCDD)
Extractables (phenols)
G, teflon-lined
cap
Purgeables (Halocarbons
and Aromatics)
Purgeables (Acrolein and
Acrylonitrite)
G, teflon-lined
cap
G, teflon-lined
septum
G, teflon-lined
septum
Cool, 4°C
H2S04 to pH<2
Cool, 4°C
Cool, 4°C
Cool, 4°C
H2SO4 to pH<;2
Cool, 4°C
0.008% Na.S?0 8
^ L 3
Cool, 4°C
H2S04 to pH<2
0.008% Na2S203g
Cool, 4°C
0.008% Na2S203
Cool, 4°C
0.008% Na2S203B
28 days
48 hours
28 days
28 days
7 days(until
extraction)
30 days(after
extraction)
7 days(until
extraction)
30 days(after
extraction)
14 days
3 days
(Continued)
-------
TABLE 17,1 (Cont'd)
Measurement*
Container
Preservative0
Maximum
Holding Timec
.p-
NJ
O
207 Orthophosphate
208 Pesticides
209 Phenols
210 Phosphorus (elemental)
211 Phosphorus
RADIOLOGICAL
212-216 Alpha, Beta and Radium
217 Residue, total
218 Residue, Filterable
219 Residue, nonfilterable
220 Residue, settleable
221 Residue, volatile
73 Silica
222 Specific conductance
P,G
G, teflon-lined
cap
P,G
G
P,G
P,G
P,G
P,G
P,G
P,G
P,G
P
P,G
Filter on site
Cool, 4°C
Cool, 4°C
0.008% Na2S203^
Cool, 4 C
H2S04 to pH<2
Cool, 4°C
Cool, 4°C
E2SO4 to PH<2
HNO3 to pH<2
Cool, 4°C
Cool, 4°C
cool, 4°C
Cool, 4°C
Cool, 4°C
Cool, 4°C
Cool, 4°C
48 hours
7 days(until
extraction)
30 days(after
extraction)
28 days
48 hours
28 days
6 months
14 days
14 days
7 days
7 days
7 days
28 days
28 days
(Continued)
-------
TABLE 17.1 (Cont'd)
Maximum
Measurement2
Container'3
Preservative0
Holding Times^
223
Sulfate
P,G
Cool, 4°C
28 days
224
Sulfide
P,G
Cool, 4°C
Zinc Acetate
28 days
225
Sulfite
P,G
Cool, 4°C
48 hours
226
Surfactants
P,G
Cool, 4°C
48 hours
227
Temperature
P,G
Determine on site
immediately
228
Turbidity
P,G
Cool, 4°C
48 hours
a Parameter numbers refer to Table I
b Polyethylene (P) or Glass (G)
c Sample preservation should be performed immediately upon sample collection. For composite samples
each aliquot should be preserved at the time of collection. When use of an automatic sampler makes
it impossible to preserve each aliquot, then samples may be preserved by maintaining at 4°C until
compositing and sample splitting is completed,
d Samples should be analyzed as soon as possible after collection. The times listed are the maximum
times that samples may be held before analysis and still considered valid. Samples may be held
for longer periods only if the permittee, or monitoring laboratory, has data on file to show that
the specific types of samples under study are stable for the longer time.
Some samples may not be stable for the maximum time period given in the table. A permittee, or moni-
toring laboratory, is obligated to hold the sample for a shorter time if knowledge exists to show
this is necessary to maintain sample stability.
e Samples should be filtered immediately on-site before adding preservative for dissolved raetals.
f Guidance applies to samples to be analyzed by GC, LC, or GC/MS for specific organic compounds.
g Should only be used in the presence of residual chlorine.
h Not available in Table I
-------
8. Quality Control Samples: Spikes, duplicates.
9. Blanks; controls.
10. Number of samples analyzed, and results.
11. Statistical analysis, precision and accuracy.
17.2 CONTAINERS
A variety of factors affect the choice of containers and cap material.
These include resistance to breakage, size, weight, interference with
constituents, cost and availability. There are also various procedures
for cleaning and preparing bottles depending upon the analyses to be
performed on the sample.
17.2.1 Container Material
The two major types of container materials are plastic and glass (23).
Glass: Plastic:
r
1. Kimax o% P^rex brand
(borosilicate)
2. Vycor
3. Corning
4. Ray-Sorbor Low-Actinic
5. Corex
All these materials have various advantages and disadvantages. Kimax or
Pyrex brand borosilicate glass is inert to most materials and is recommended
where glass containers are used. Conventional polyethylene is to be used
when plastic is acceptable because of reasonable cost and less absorption
of metal ions . The specific situation will determine the use of
glass or plastic. However, use glass containers for pesticides, oil and
grease, and other organics. Table 17.3 summarizes the advantages and
422
1. Conventional polyethylene
2. Linear polyethylene
3. Polypropylene
4. Polycarbonate
5. Rigid polyvinyl chloride
6. Teflon
-------
TABLE 17.2 INFORMATION ON PRESERVATION AND STORAGE OF PARAMETERS IN
VARIOUS WATERS AND WASTEWATERS
Parameters
Sample type
Preservation
method
Container
Material
Temperature Holding Time
DEMAND
PARAMETERS
Biochemical
Oxygen
Demand
(BOD)
Raw sewage
Raw sewage
Raw semi-treated
or fully treated
domestic sewage
Raw waste water
N.S.
N.S.
Glass
N.S.
37°C
10°-24°C
1°C
4°C
Frozen in a mixture Polyethylene Approximate-
of acetone and dry
ice or finely
ground dry ice
Freezing
ly -5°C
Polyethylene -15°C
coated milk
cartons
1:4 settled sewage 60-80 mg/1 Hgdl^
to water from a
natural stream
Raw sewage
890 mg/1 HgCl2
Plastic
Plastic
Room temper-
ature
Room temper-
ature
6-12 hours (21)
12-24 hours (21)
6 days (21)
Up to 1 day in
composite sampling
systems (10)
6 months; on thawing
either with warm water
or at room temperature,
analyze using seeded
technique (5)
236 days, analyze
using seeded technique
(8)
18 days (23)
43 days (25)
N.S.-Not Stated.
(continued)
-------
TABLE 17.2 (continued)

Parameters
Sample type
Preservation
method
Container
Material
Temperature
Holding Time
DEMAND
PARAMETERS
-

Chemical
Oxygen
Demand
(COD)
1:4 settled
sewage to water
from a natural
stream
60-80 mg/1 HgCl^
Plastic
Room temper-
ature
18 days (23)

Raw sewage
890 mg/1 HgCl2
Plastic
Room temper-
ature
43 days (25)

Raw sewage
N.S.
Glass
37°C
10e_24°C
1°C
6-12 hours (21)
12-24 hours (21)
6 days (21)

Raw sewage
N.S.
N.S.
4°C
Several days (10)
Dissolved
oxygen (DO)
Sea water
0.5 % chlor-
oform + 0.5%
phenol
Glass
22°C
20 days (24 )

Sea water
Acidulating water
to pH 1.5 with
2.5 ml H2SO4 ;
5 ml HC1 per liter
of sample
Glass
22°C
22 days
-------
TABLE 17.2 (continued)
Parameters
Sample Type
Preservation
Method
Container
Material
Temperature
Holding Time
DEMAND
PARAMETER

Total
Organic
Carbon (TOC)
Settled sewage,
biological
filter effluent
1 ml saturated
Ag^SO^ solution
(i.e. 4 mg of Ag+)
to a liter of sample
Glass
Refrigerate
at 4°C
3 days (26}
METALS:

Aluminum
Waters in the
zone of mixing
of river and
sea waters in
estuaries
Samples frozen
rather than
acidified
Polyethylene
-20°C
In dark, 14
days <27 )

Natural fresh
water
1 ml of 4M H2SO,
per 100 ml sample
and filtered through
glass-fiber filters
Polyethylene
Room temper-
ature
4 weeks (28)
Cadmium
Stock aqueous
solutions pre-
pared in labor-
atory
Acidification to pH 2 Polyethylene
with HNOj and borosili-
cate glass
N.S.
32 days (29)
Lead
Stock aqueous
solutions pre-
pared in labor-
atory
Acidification to
pH 2 with HN03
Borosilicate
glass
N.S.
24 days (29)
(continued)
-------
TABLE 17.2 (continued)
Parameters
Sample Type
Preservation
Method
Container
Material
Temperature
Holding Time
METALS: cont.

Mercury
Distilled water
solutions con-
taining 0.1-10.0
Acidified with 5%
(v/v) HN03 + .05%
Cr2072"
Polyethylene
N.S.
10 days
(30)

Distilled water
solutions con-
taining 0.1-10.0
Acidified with 5%
(v/v) HNO, + .01%
Cr2072- 3
Glass
N.S.
5 months
(30)
Potassium
1:4 settled
sewage and nat-
ural stream water
Approx. 1.5 ml
saturated HgCH^ per
liter of sample (60-
80 mg/1 HgCl2)
Plastic
Room Temper-
ature
18 days
(23)
Silver
Stock aqueous
solutions pre-
pared in labor-
atory
Acidification to
pH 2 with HNO3
Polyethylene
Room temper-
ature
36 days
(29)
Sodium
1:4 settled sew-
age and natural
stream water
Approx. 1.5 ml
saturated HgC^ per
liter of sample (60-
80 mg/1 HgCl2)
Plastic
Room temper-
ature
18 days
(23)
Zinc
Stock aqueous
solutions pre-
pared in labor-
atory
Acidification to
pH 2 with HNO3
Polyethylene N.S.
preferred over
borosilicate
glass
60 days
(29)
(continued)
-------
TABLE 17.2 (continued)

Parameters
Sample Type
Preservation
Method
Container
Material
Temperature
Holding Time
METALS: cont.

Cadmium
Natural
lake water
Acidified to
pH 1
Pyrex glass
and
polyethylene
-15°C
184 days (31)
Copper
Natural
lake water
Acidified to
pH 1
Pyrex glass
and
polyethylene
-15°C
184 days (31)

.25 ml 3.5 N nitric
acid after arrival
at the laboratory
25 ml glass
vials with
polyethylene
snap-caps
Room temper-
ature
1 year (32)
Manganese
Natural
lake water
Acidified to
pH 1
Pyrex glass
and
polyethylene
-15°C
184 days (31)
Zinc
Natural
lake water
Acidified to
pH 1
Pyrex glass
and
polyethylene
-15°C
184 days (31)

.25 ml 3.5 N nitric
acid after arrival
at the laboratory
25 ml glass
vials with
polyethylene
snap-caps
Room temper-
ature
1 year (32)
(continued)
-------
TABLE 17.2 (continued)

Parameters
Sample type
Preservation
method
Container
Material
Temperature
Holding Time
NUTRIENTS:

Ammonia
Nitrogen
Relatively
unpolluted
bay waters
40 mg Hg+2 per
liter of sample
Plastic
4°C
30 days (12)

Sea waters
(off shore)
0.4 g phenol per
100 ml of sample
Glass
N.S.
2 weeks (14)

Slow freezing
Polyethylene
Frozen
20 days (14)

Near shore
and estuarine
waters (filtered
Freezing
Glass tnbes
polyseal
caps
-23°C
3 months (7)
and fortified
samples)
Synthetic fresh Unpreserved Polyethylene 4°C 1-3 days (33)
water, unpollut-
ed fresh water,
(filtered)
chemically treated
domes t ic sewage,
polluted sea water
(filtered)
(continued)
-------
TABLE 17.2 (continued)
Parameters
Sample Type
Preservation
Method
Container
Material
Temperature Holding Time
NUTRIENTS:
Ammonia
Nitrogen
Ammonia
(soluble)
Strongly pollu-
ted water
Strongly pollu-
ted water
Raw Sewage
Surface runoff
Amended and un-
amended river
water
Approx. 1.5 ml
saturated HgCl« per
liter (75 mg/lj
Approx. 3.0 ml of
40% formalin solu-
tion per liter of
sample (890 mg/1)
890 mg/1 HgCl2
Freezing
Refrigeration
Phenylmercuric
acetate (PMA):
20 mg PMA per
liter of sample
40 mg HgC^ per
liter of sample
Freezing
Phenylmercuric
acetate (PMA):
20 mg PMA per
liter of sample
Plastic
Plastic
N.S.
Plastic
Plastic
Plastic
Plastic
Plastic
Plastic
Room Temper-
ature
Room Temper-
ature
N.S.
18 days (23)
18 days (23)
-20°C
4°C
4°C
4°C
-20°C
43 days (25)
In dark, 12 wks.
In dark, 12 wks.
(34)
In dark, 12 wks.
(34)
In dark, 12 wks.
(34)
In dark, 12 wks.
o (34)
4 C or 23 C in dark, 12 wks.
(34)
(continued)
-------
TABLE 17 .2 (continued)
Parameters
Sample Type
Preservation
Method
Container
Material
Temperature
Holding Time
NUTRIENTS:
cont.

Ammonia

(soluble)
Tile drainage water
Freezing
Plastic
-20°C
In dark, 12 wks.
cont.

Phenylmercuric

(34)

acetate (PMA):
Plastic
4°C
In dark, 12 wks.

(20 mg PMA per

(34)

liter of sample)

40 mg HgCl2 per
plastic
4°C
In dark, 12 wks.

liter of sample

(34)
Kjeldahl
Relatively unpol-
40 mg Hg+2 per
Plastic
4°C
7 days (12)
nitrogen
luted bay waters
liter of sample

Synthetic fresh
Unpreserved
Polyethylene
4°C
Up to 3 days (35)

water, unpolluted

fresh water, chem-
1 ml of 0.02%
Polyethylene
4 C
Up to 3 days (35)

ically treated do-
mercury (II) chlor-

mestic sewage and
ide per 100 ml of

polluted sea water
sample

Strongly polluted
Approx. 1.5 ml of
Plastic
Room Temp.
18 days (23)

water
saturated HgCl2 per

liter (75 mg/lf

Raw manure slur-
Freezing and fast
Whirl pack
N.S.
5 weeks (36)

ries, oxidation
thawing or slow
bags

ditch mixed liquor
thawing

Refrigeration
Whirl pack
6-10°C
5 weeks (36)

bags

(continued)
-------
TABLE 17 .2 (continued)

Parameters
Sample Type
Preservation
Container
Temperature
Holding Time

Method
Material

NUTRIENTS: cont.

Kjeldahl

Acidification with
Whirl pack
6-10°C
5 weeks
(36)
nitrogen cont.

cone HjSO/ to a
bags
6-10°C
5 weeks
(36)

pH of 2

Nitrate
Relatively unpol-
40 mg Hg+^
Plastic
4°C
Up to 3
days (35)
Nitrogen
luted fresh vater

(filtered), chem-

ically treated do-
1 ml of 0.02% mer-
Polyethylene
4 C
28 days
(35)

mestic sewage, pol-
cury (II) chloride

luted sea water
per liter of sample

(filtered)

4 to 1 mixture of
22 or 66 mg of mer-
Glass
22±2°C
3 weeks
(37)

surface water and
cury (II) chloride

settled sewage
per liter of sample

Cone. B^SO^ 0.8 ml
Glass
22+2°C
3 weeks
(37)

per liter of sample

Strongly polluted
Approx. 1.5 ml of
Plastic
Room Temp.
18 days
(23)
water sample saturated mercury (II)
chloride solution
per liter of sample
(i.e. 60-80 mg/1 of
mercury (II) chloride)
(continued)
-------
TABLE 17 .2 (continued)
Parameters
Sample Type
Preservation
Method
Container
Material
Temperature Holding Time
NUTRIENTS: cont.
Nitrate
Nitrogen Cont.
Nitrite
Nitrogen
Surface runoff,
tile drainage
water, river water
Surface runoff
Strongly polluted
water
Approx. 3.0 ml of Plastic
40% formalin solu-
tion per liter of
sample (890 mg/1)
Freezing Plastic
20 mg PMA per liter Plastic
of sample
40 mg HgCl2 per Plastic
liter of sample
Approx. 1.5 ml of Plastic
saturated mercuric
chloride solution
per liter of sample
(i.e. 60-80 mg/1 of
mercuric chloride)
Room Temp. 18 days (23)
-20 C In dark, 12 wks.
< 34)
4 C or 23 C In dark, 12 wks.
4°C
(34)
In dark, 3 wks.
(34)
Room Temp. 18 days (23)
Sea water (fil-
tered) and nitrate
enriched
Approx. 3.0 ml of Plastic
40% formalin solution
per liter of sample
(890 mg/1)
Freezing
Pyrex glass
Room Temp. 18 days (23)
-18°C
220 days (4)
(continued)
-------
TABLE 17.2 (continued)
Parameters
Sample Type
Preservation
Method
Container
Material
Temperature
Holding Time
NUTRIENTS:
cont.

Nitrite
Nitrogen
coat.
Lake water
(unenriched)
1 ml saturated mer-
curic chloride per
liter of sample
Glass
Refrigerated
at 6°C
11 days
(20)

Lake water
(enriched with
nitrite)
1 ml saturated mer-
curic chloride sol-
ution per 300 ml of
sample
Glass
Refrigerated
at 6°C
6 days
6 days
(20)
(20)

Relatively un-
polluted bay
waters
+2
40 mg Hg per liter
of sample
Plastic
4°C
7 days
(12)

4 to 1 mixture of
surface water and
settled sewage
66 mg of mercury
(II) chloride per
liter of sample
Glass
22±2°C
45 days
(37)
(continued)
-------
TABLE 17 .2 (continued)
Parameters
Sample Type
Preservat ion
Method
Container
Material
Temperature Holding Time
NUTRIENTS: cont.
Orthphosphate
or total phos-
phate
Waters contain'
ing algae
Refrigeration
N.S.
3-5°C
Overnight (11)
Polluted fresh
water, polluted
sea water, strong-
ly polluted sea
water, biologi-
cally treated
sewage
Estuarine waters
1 ml of 8N sulfuric
acid per 100 ml of
filtered sample
Polyethylene N.S.
2+
40 mg Hg per
liter of sample
Oj.
40 mg Hg per
liter of sample
Glass
Glass
-10°C
4°C
For samples that
cannot be analy-
zed within 8 hrs.
(38)
One month (12)
Few days (12)
Soluble Inor-
ganic Phos-
phorus (SIP)
Strongly polluted
waters
Surface runoff
Slow freezing
and sediment re-
moved by centri-
fugation
Approx. 1.5 ml Plastic
saturated HgCl per
liter (75 mg/lj
N.S. N.S.
N.S. N.S.
Room Temp. 18 days (23)
2°C
-20°C
3 days (38)
3 days (38)
(continued)
-------
TABLE 17.2 (continued)
Parameters
Sample Type
Preservation
Method
Container
Material
Temperature
Holding Time
NUTRIENTS:
cont.

Soluble
Surface runoff,
Freezing
Plastic
-20°C
In dark, 12 wks.
Inorganic
tile drainage

(34)
Phosphate

Phenylmercuric
Plastic
4°C
In dark, 6 wks.

acetate (PMA)

(34)

20 mg PMA per

liter of sample

40 mg HgCl? per
Plastic
4°C
In dark, 6 wks.

liter of sample

(34)

Ammended river
Freezing
Plastic
-20°C
In dark, 12 wks.

water

(34)

(45 ml river

water + 5 ml of

solution contain-
40 mg HgCl per
Plastic
4 C
In dark, 12 wks.

ing 100 ppm NH.-N,
liter of sample

(34)

100 ppm of NO^-N

and 5 ppm of ortho

phosphate), and

natural rainwater

Seawater
Addition of Chlor-
Polyethylene
-5 to -10°C
Stored until

oform (0.6-0.8% v/v)

thawed for anal-

before freezing

ysis (39)
(continued)
-------
TABLE 17.2 (continued)
Parameters
Sample Type
Preservation
Method
Container
Material
Temperature

Holding Time
PHYSICAL/MINERAL

Alkalinity
1:4 settled sewage
Approx. 1.5 ml of
Plastic
Room temp.
18
days
(23)

to natural stream
saturated mercuric

(Not in dark)

water
chloride solution

per liter of sample

(60-80 mg/1 HgCl^)

Chloride
1:4 settled sewage
Approx. 1.5 ml of
Plastic
Room temp.
18
days
(23)

and natural stream
saturated mercuric

(Not in dark)

water
chloride solution

per liter of sample

(60-80 mg/1 HgCl2)

Raw sewage
890 mg/1 HgCl2
N.S.
N.S.
43
days
(25)
Conductivity
1:4 settled sewage
Approx. 1.5 ml of
Plastic
Room temp.
18
days
(23)

and natural stream
saturated mercuric

(Not in dark)

water
chloride solution

per liter of sample

(60-80 mg/1 HgCl^)

Raw sewage
890 mg/1 HgCl2
N.S.
N.S.
43
days
(25)
Total hardness
1:4 settled sewage
Approx. 1.5 ml of
Plastic
Room temp.
18
days
(23)

and natural stream
saturated mercuric

(Not in dark)

water
chloride solution

per liter of sample

(60-80 mg/1 HgCl2)

(continued)
-------
TABLE 17.2 (continued)
Parameters
Sample Type
Preservation
Method
Container
Material
Temperature Holding Time
PHYSICAL/MINERAL
cant.
Magnesium
hardness
Phenols
Sulfate
raw sewage
890 mg/1 HgCl,
All types of water 1.5 ml of IN NaOH
and waste waters per liter
All types of water
and waste waters
N.S.
3 ml 10% CuSO, sol-
ution per liter of
sample
1:4 settled sewage Approx. 1.5 ml
and natural stream saturated mercuric
water chloride solution
per liter of sample
(60-80 mg/1 HgCl2)
Raw sewage
890 mg/1 HgCl2
N.S,
N.S.
Stoppered
glass bot-
tles
Stoppered
glass
bottles
Plastic
N.S.
N.S.
N.S.
N.S.
Refrigeration
43 days (25)
(40)
preferably to
analyze shortly
after collection
(19)
Analyze within 2
days (19)
Room Temp. 18 days (23)
(Not in dark)
N.S.
43 days (25 )
-------
disadvantages of these materials.
TABLE 17.3 COMPARISON OF GLASS AND
PLASTIC CONTAINERS
Borosilicate Glass
Conventional Polyethylene
Interference
with sample
Weight
Resistance to
breakage
Cleaning
Sterilizable -
Space
Inert to all con-
stituents except
strong alkali
Heavy
Very fragile
Easy to clean
Yes
Takes up considerable
space
Good for all constituents
except pesticides and oil
and grease
Light
Durable
Some difficulty in
removing adsorbed
components
In some instances
Substantial space savings
during extended field
studies.
17.2.2 Container Caps
There are two main types of plastic container caps: polyethylene and
bakelite with liners. Use polyethylene caps (ease of cleaning) except if
these caps do not fit tightly to the container or if pesticides or oil and
grease analyses are to be performed. Teflon liners should be used for
pesticides and oil and grease samples. There are three liner types avail-
able and the advantages/disadvantages are listed in Table 17.4.
17.2.3 Container Structure
Use a wide mouth container in most instances. This structure will permit
438
-------
easy filling and sample removal. It is also easily cleaned, quickly dried,
and can be stored inverted. Use a narrow neck bottle when interaction with
the cap liner or outside environment is to be minimized. Use a cleaned sol-
vent container for pesticide sample collection (24).
TABLE 17.4 COMPARISON OF CAP LINERS
Liner Type
Advantages
Disadvantages
Wax coated paper
Generally applicable
to most samples
Must be inspected prior
to each use because of
deterioration

Inexpensive
Cannot use with organics
Neoprene
Same as wax coated
paper
Same as wax coated paper
Teflon
Applicable for all
analyses
Minimizes container/
sample interaction
High cost
17.2.4 Disposable Containers
Use disposable containers when the cost of cleaning is high. These con-
tainers should be precleaned and sterile. The most commonly used disposable
container of this type is the molded polyethylene cubitainer shipped nested
and sterile to the buyer. However since their cubic shape and flexible sides
make them almost impossible to clean thoroughly, use these containers only
once.
17.2.5 Container Washing
The following procedure should be followed to wash containers and caps
for inorganic and general parameters;
1. Wash containers and caps with a non-phosphate detergent and
439
-------
scrub strongly with a brush (if possible wash liners and
caps separately).
2. Rinse with tap water, then distilled water.
3. Invert to drain and dry.
4. Visually inspect for any contamination prior to storage.
5. If the container requires additional cleaning, rinse with
a chromic acid solution (35 mL of saturated sodium dichromate
solution in 1 liter of sulfuric acid - this solution can be
reused). Then rinse with tap water and distilled water and
dry as indicated above.
17.2.6 Container Preparation
For certain parameters, a special cleaning procedure is needed to avoid
adsorption or- contamination due to interaction with container walls. These
procedures are outlined below.
1. Metals: If metals are to be analyzed, rinse the container
with a solution of one part nitric acid to four parts water,
then with distilled water. If phosphorus is to be analyzed,
rinse the container with a solution of one part hydrochloric
acid to one part water,followed by distilled water.
2. Organics: If Oil and Grease or Pesticides are to be analyzed,
rinse the sample container with methylene chloride, followed
by acetone. For Pesticide analysis, use pesticide grade hexane
or acetone. The container should have been previously cleaned
with chromic acid solution as described in Section 17.2.5.
Treat the container caps similarly.
440
-------
3. Sterilization: For microbiological analyses, sterilize
the container and its stopper/cap by autoclaving at
121°C for 15 minutes or by dry heat at 180°C for two hours.
Heat-sensitive plastic bottles may be sterilized with
ethylene oxide at low temperatures. Wrap bottles in kraft
paper or cover with aluminum foil before sterilization to
protect against contamination. An acceptable alternative
for emergency or field use is sterilization of containers
by boiling in water for 15 minutes.
17.3 HOLDING TIME
Holding time is the time interval between collection and analysis.
In general, the shorter the time that elapses between collection of a
sample and its analysis, the more reliable will be the analytical results.
It is impossible to state exactly how much time may be allowed to
elapse between collection of a sample and its analysis; this depends on
the character of the sample, particular analysis to be made, and the
conditions of the storage
For NPDES purposes, in accordance with Federal Register, part 136
guidelines follow the recommendations given in Table 1, pp. xvi - xix,
Methods for Chemical Analysis of Water and Wastes. U.S. Environmental
Protection Agency, 1979.
For information purposes, however, data relating to holding times for
general and inorganic parameters was collected from various literature
sources and is tabulated in Table 17.2.
441
-------
17.4 SAMPLE VOLUME
The volume of sample collected should be sufficient to perform all the
required analyses plus an additional amount to provide for any quality con-
trol needs, split samples or repeat examination. Although the volume of
sample required depends on the analyses to be performed, the amount required
for a fairly complete analysis is normally about 8 liters, (about 2 gallons).
The laboratory receiving the sample should be consulted for any specific
volume requirements. Individual portions of a composite sample should be at
least 100 milliliters in order to minimize sampler solids bias. Depending
on the sampling frequency and sample volume, the total composited sample
should be a minimum of 8 liters (about 2 gallons). Refer to EPA's "Methods
for Chemical Analysis of Water and Wastes 1979" for the sample volumes
required for specific types of pollutant analyses.
442
-------
17.5 REFERENCES
1. Collier A. W., and K. T. Marvin. Stabilization of the Phosphate Ratio
of Sea Water by Freezing. U.S. Government Printing Office, Washington,
71-76, 1953.
2. May, B. Z. Stabilization of the Carbohydrate Content of Sea Water
Samples. Limnology and Oceanography, 5: 342-343, 1960.
3. Heron, J. Determination of Phosphate in Water after Storage in Polyeth-
ylene. Limnology and Oceanography, 5: 316-321, 1960.
4. Procter, R. R. Stabilization of the Nitrite Content of Sea Water By
Freezing. Limnology and Oceanography, 7: 479-480, 1962.
5. Fogarty, W. J., and M. E., Reeder. BOD Data Retrieval Through Frozen
Storage. Public Works, 88-90, March 1964.
6. Morgan, F., PE, and E. F. Clarke. Preserving Domestic Waste Samples by
Freezing. Public Works, 73-75, November 1964.
7. Marvin, K. T. and R. R. Proctor. Stabilizing the Ammonia-Nitrogen
content of Estuarine and Coastal Waters by Freezing. Limnology and
Oceanography, 10: 288-289. 1965.
8. Zanoni, A. E. Use of Frozen Wastewater As A Test Subtrate. Public
Works, 72-75, November 1965.
9. Tyler, L. P. and E. C. Hargrave. Preserving Sewage Seed for BOD
Analysis, Water and Sewage Works, 12: 181-184, May, 1965.
10. Agardy, F. J. and M. L. Kiado. Effects of Refrigerated Storage on the
Characteristics of Wastes, Industrial Waste Conference (21st) Purdue
University, 1966.
11. Fitzgerald, G. P., and S. L. Faust. Effect on Water Sample Preservation
Methods on Release of Phosphorous From Algae. Limnology and Oceanography,
12: 332-334, 1967.
12. Jenkins, D. The Differentiation, Analysis and Preservation of Nitrogen
and Phosphorous Forms in Natural Waters, Advances in Chemistry Series
73. American Chemical Society, Washington, D. C., 265-279, 1968.
443
-------
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Philbert, F. J. The Effect of Sample Preservation by Freezing Prior to
Chemical Analysis of Great Lakes Water, Proc 16th Conference. Great
Lakes Res. 282-293, 1973.
Degobbis, D. On the Storage of Sea Water Samples for Ammonia Determin-
ation, Limnology and Oceanography, 15: 146-150. January, 1970.
Burton, J. D. Problems in the Analysis of Phosphorus Compounds, Water
Research, Great Britain, 7: 291-307, 1973.
Harms, L. L., and J. N. Dornbush, and J. R. Anderson. Physical and
Chemical Quality of Agricultural Land Runoff. Journal T-TPCF, 46: 2460-
2470, November 1974.
Phillips, G. E., and W. D. Hatfield. Preservation of Sewage Samples.
Water Works and Sewage Journal, 285-288, June 1941.
Moore, E. W. Long Time Biochemical Oxygen Demands at Low Temperature,
Sewage Works Journal, 13 (3): 561-577, May, 1941.
Ettinger, M. B., and S. Schott, and C. C. Ruchott. Preservation of Phenol
Content in Polluted River Water Samples Previous to Analysis. Journal-
AWWA, 35: . 299-302, March 1943.
Brezonik, P. L., and G. F. Lee. Preservation of Water Samples for In-
organic Nitrogen Analysis with Mercuric Chloride. Air and Water
Pollution (Great Britain), 10: 549-553, 1966.
Loehr, R. C., and B. Bergeron. Preservation of Wastewater Samples Prior
to Analysis. Water Research (Great Britain) 1: 557-586, 1967.
Federal Register, 40CFR, Vol. 44 , No. 244 , 75029-75052, December 18, 1979.
Hellwig, D.H.R. Preservation of Water Samples. International Journal
of Air Water Pollution, 8: 215-228, 1964.
Zobell, C.E., and B.F. Brown. Studies on the Chemical Preservation of
Water Samples. Journal of Marine Research, 5 (3): 178-182, 1976.
Hellwig, D.H.R, Preservation of Waste Water Samples. Water Research,
1: 79-91, 1976.
Van Steendeven, R.A. Parameters Which Influences the Organic Carbon
Determination in Water. Water South Africa, 2(4): 156-159, 1976
444
-------
27. Hydes, D. J. and P. S. Liss. Fluorimetric Method for the Determination
of Low Concentrations of Dissolved Aluminum'in Natural Waters. Analyst,
101: 922-931, December, 1976.
28. Dale, T., and A. Henricksen. Intercalibration Methods for Chemical
Analysis of Water. Vatten, (1): 91-93, 1975.
29. Struempler, A. W. Adsorption Characteristics of Silver, Lead, Cadmium,
Zinc, and Nickel on Borosilicate Glass, Polyethylene and Polypropylene
Container Surfaces. Analytical Chemistry, 45 (13): 2251-2254, 1972.
30. Feldman, C. Preservation of Dilute Mercury Solutions. Analytical
Chemistry. 31: 99-102, July 1974.
31. Clement, J. L. Preservation and Storage of Water Samples for Trace
Element Determination. Department of Civil Engineering, University of
Illinois, Urbana, Illinois, 1972. 40 pp.
32. Henricksen, A., and Balmer. Sampling, Preservation and Storage of
Water Samples for Analysis of Metals. Vatten: (1): 33-38, 1977.
33. Dahl, I. Intercalibration of Methods for Chemical Analysis of Water.
Vatten. (4): 336-340, 1973.
34. Klingaman, E. D. M. and D. V. Nelson. Evaluation of Methods for
Preserving the Levels of soluble Inorganic Phosphorus and Nitrogen in
Unfiltered Water Samples. Journal Environmental Quality, 5 (1): 42-
46, 1976.
35. Dahl, I. Intercalibration of Methods for Chemical Analysis of Water.
Vatten (2): 180-186, 1974.
36. Prakasam, T. B. S. Effects of Various Preservation Techniques on the
Nitrogen Profile of Treated and Raw Poultry Waste. Draft Copy, 1975.
37. Howe, L. H., and C. W. Holley. Comparison of Mercury (II) chloride
and Sulfuric Acid as Preservatives for Nitrogen Forms in Water Samples.
Environmental Science and Technology, 3 (5): 478-481, May 1969.
38- Nelson, D. W., and M. J. M. Roomkens. Suitability of Freezing as a
Method of Preserving Run Off Samples for Analysis of soluble Phos-
phate. Journal Environmental Quality. 1 (3): 323-324, 1972.
39. Gilmartin, M. Changes in Inorganic Phosphate Concentration Occurring
During Seawaste Sample Storage. Limnology and Oceanography, 12:
325-328, 1967.
40. Baylis, J. E. Procedure for Making Quantitative Phenol Determinations.
Water Wastes and Sewage. 79: 341, 1932.
445
-------
Appendix A - Population Parameters
I. Populations and Samples (1,3)
Most sampling is done on a non-continuous basis, and so the data gathered
give an incomplete picture of the true condition of a water or wastewater. If
monitoring were done continuously, the data would be presented as a curve (f(t),
where f is the function which gives the value of the parameter at time t) rather
than as a discrete set of points (numbers). Therefore, the definitions of mean
and variance given in Section 4.1.1 could not be applied. This continuous
function defines a "population" from which the samples are taken. This popula-
tion has a mean and a variance of which the the sample mean and sample variance
(which are the mean and variance defined in Section 4.1.1) are only estimators.
This is why it is best to take as many samples as possible — more data reveals
more information about the population.
The Population Mean
The population mean, yv , is defined by
X
CO
yx ¦ E(X) = /_oo xfx(x) dx
where E(X) is another expression for the mean and is read "the expected value
(or expectation) of X."
f (x) is the density function of X, which is a function defining the
A
distribution of X.
446
-------
The Population Variance
2
The variance, ox, of the population is defined by
o£ = Var (X) = e((X~ux)2) « /_w (x-px)2fX
-------
Append ix B
Areas under the Normal Curve (1,3)
The graph of the probability density function of the standard normal
distribution,
=¦ __ exp(-x2/2),
/2tt .
is shown in Figure 4.6. It is the familiar bell-shaped curve. For any point
z, the area under the curve to the left of z is determined by
OxCOdt
which we have seen to be P(Zz). The normal distributions is symmetric about its
mean, and so P(Z>y +c) = P(Zc) = P(Z<-c).
There is a property of probabilities which says that, under certain
conditions which are not discussed here, P(Z>c or Z<-c) = P(Z>c +
P(Z<-c) = 2P(Z>c) and so if P(Z>c or Z<-c) = a then P(Z>c) = a/2 (which is
the area of the shaded region in Figure A.8).
448
* U.S. GOVERNMENT PRINTING OFFICE: 1M0-757-064/0192
-------