U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
PB-261 260
Chemical Analyses for
Water Pollutants
National Training & Operational Technology Or., Cincinnati, OH.
Nov 76
-------
NOTICE
THIS DOCUMENT HAS BEEN REPRODUCED FROM THE
BEST COPY FURNISHED US BY THE SPONSORING
AGENCY. ALTHOUGH IT IS RECOGNIZED THAT CER-
TAIN PORTIONS ARE ILLEGIBLE, IT IS BEING RE-
LEASED IN THE INTEREST OF MAKING AVAILABLE
AS MUCH INFORMATION AS POSSIBLE.
-------
•{ b b 0 8 4
PB 261 260
EPA 430/1 IS 013
CHEMICAL ANALYSES FOR WATER
POLLUTANTS
TRAINING MANUAL
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF WATER PROGRAM OPERATIONS
NAT|ON.V. "ECHNICAL
INFC *'.••••' :>. SERVICE
-------
BIBLIOGRAPHIC DATA
SHEET
1. Report No.
eport No.
EPA-4UO/1-75-013
3. Recipient's Accession No.
J!£L
'4. Title and "Subtitle
Chemical Analyses for Water Pollutants
5. Report Date
November 1976
'. Author(s) ^
Audrey E. Kroner
8- Performing Organization Kept. !
No. " 1
'. Performing Organization Name and Address
U.S. Environmental Protection Agency, OWPO
Municipal Operations and Training Division
.National Training & Operational Technology Center
Cincinnati, Ohio 45268
10. Project/'TasK, Work Unit No. ;
11. Contract/Grant No.
12. Sponsoring Organization Name and Address
13. Type of Report & Period
Covered
14.
15. Supplementary Notes
16. Abstracts
This training manual contains outlined information used by Lecturers when
presenting topics in EPA-NTOTC Course 100. 3, "Chemical Analyses for Water
Pollutants. " The contents concern individual constituents found in water (e. g.
dissolved oxygen) and information useful to analysts (e.g. statistics). A section
of laboratory procedures adapted for class groups is included.
17. Key Words and Document Analysis. 17a. Descriptors
Acidity, Alkalinity, Ammonia, Biochemical Oxygen Demand, Carbon,
Dissolved gases. Hardness, Inorganic nitrates, Nitrites, Nitrogen cycle,
Oxygen, Phosphorus, Quality Assurance, Statistical Analysis, Data Collection
and Evaluation.
17b. Identifiers/Open-Ended Terms
17e. COSATI Fie Id/Group
19. Security Class (This |21. "No. of Pages
Report) ' j
UNCLASSIFIED j
18. Availability Statement
Release to public
20. Security Class (This
Page
UNCLASSIFIED
1 NTIS-35 IRdV. 3-7ZI
USCOMM-OC r
-------
EPA-430/ 1-75-01 3
December 1975
CHEMICAL ANALYSES FOR WATER POLLUTANTS
This course is designed for chemists or technicians
who will perform chemical analyses for water pollutants.
The student will learn about analyses listed in the
guidelines developed pursuant to Section 304(g) of the
Federal Water Pollution Control Act Amendments of
1972 for NPDES permit applications and reports and
also for certifications issued by the States.
The student will learn theoretical concepts and perform
laboratory exercises pertaining to the following: Acidity
and Alkalinity Determinations, pH Measurement,
Biochemical Oxygen Demand, Dissolved Oxygen
(Winkler and Probe), Calcium and Magnesium Hard-
ness Determination, Nitrogen Analyses (Ammonia,
Kjeldahl, Nitrate, Nitrite), Phosphorus Determinations,
Total Organic Carbon Analysis, Total Solids Deter-
mination.
Discussion periods are scheduled after each laboratory
assignment is completed. Time is divided about equally
between classroom and laboratory work.
Persons attending should have fundamental knowledge
of inorganic chemistry and quantitative analysis. They
should also be able to use an analytical balance, volu-
metric glassware and titration assemblies.
U. S. ENVIRONMENTAL PROTECTION AGENCY
Office of Water Program Operations
TRAINING PROGRAM
-------
FOR KW OK l>
These mainuals are prepai-ed for reference use of students enrolled in
scheduled training courses of the Office of Water Program Operations.
Environmental Protection Agency.
Due to the limited availability of the manuals it is not
appropriate to cite them as technical references in
bibliographies or other forms of publication.
References to products and manufacturers are for
illustration only; such references do not imply product
endorsement by the Office of Water Program
Operations, Environmental Protection Agency.
The reference outlines in this manual have been selected and developed
with a goal of providing the student with a fund of the best available
current information pertinent to the subject matter of the course.
Individual instructors may provide additional material to cover special
aspects of their own presentations.
This manual will be useful to anyone who has need for information on
the subjects covered. However, it should be understood that the manual
will have its greatest value as an adjunct to classroom presentations.
The inherent advantages of classroom presentation is in the give-and-
take discussions and exchange of information between and among
students and the instructional staff.
Constructive suggestions for improvement in the coverage, content,
and format of the manual are solicited and will be given full
consideration.
-------
CONTENTS
Title or Description Outline Number
I DATA COLLECTION AND EVALUATION
Sample Handling - Field Through Laboratory 1
Statistics for Chemists 2
Accuracy-Precision-Error ^
Interlaboratory Quality Control Studies 4
II STANDARD SOLUTIONS, ACIDITY, ALKALINITY
Volumetric Analysis of Water Quality 5
Acidity. Alkalinity, pH and Buffers 6
Alkalinity and Relationships Among the Various Types of Alkalinities 7
III OXYGEN TESTS
Dissolved Oxygen - Factors Affecting DO Concentrations in Water 8
Dissolved Oxygen Determination (Winkler lodometric Titration and 8
Azide Modification)
Dissolved Oxygen Determination - Electronic Measurements 9
Biochemical Oxygen Demand Test Procedures 10
BOD Determination - Reaerated Bottle Probe Technique 10
Effect of Some Variables on the BOD Test 10
Mathematical Basis of the Biochemical Oxygen Demand (BOD) Test 10
IV SELECTED ANALYSES
Sources and Analysis of Organic Nitrogen 11
Ammonia, Nitrites and Nitrates 11
Total Carbon Analysis 12
Determination of Calcium and Magnesium Hardness 13
Solids Relations in Polluted Water 14
Phosphorus in the Aqueous Environment 15
100.3. 11.75
II
-------
2 C o n_t e n t s
Title or Description Outline Number
V LABORATORY PROCEDURES
Biochemical Oxygen Demand Test - Dilution Technique 16
Laboratory Procedure for Dissolved Oxygen (Winkler-Azide Procedure) 17
Determination of Kjeldahl Nitrogen 18
Determination of Nitrate-Nitrite Nitrogen 19
Laboratory Procedure for Total Solids 20
Laboratory Procedure for Total Hardness 21
Laboratory Procedure for Phosphorus 22
Acidity 23
Laboratory Procedure for Total Alkalinity 24
-------
SAMPI.K HANDLING - KII-: LP THROmi f I \HOK\TOKY
I PLANNING A SAMPLING PROGRAM
A Factors to Consider:
1 Locating sampling sites
2 Sampling equipment
3 Type of sample required
a grab
b composite
4 Amount of sample required
5 Frequency of collection
6 Preservation measures, if any
B Decisive Criteria
1 Nature of the sample source
2 Stability of constituent(s) to be measured
3 Ultimate use of data
II REPRESENTATIVE SAMPLES
If a sample is to provide meaningful and
valid data about the parent population, it
must be representative of the conditions
existing in that parent source at the
sampling location.
A The container should be rinsed two or
three times with the water to be collected.
B Compositing Samples
1 For some sources, a composite of
samples is made which will represent
the average situation for stable
constituents.
2 The nature of the constituent to be
determined may require a series of
separate samples.
C The equipment used to collect ;nc sample
is an important factor to consider.
ASTM* ' has a detailed section on various
sampling devices and techniques.
D Great care must be exercised when
collecting samples in sludge or mud areas
and near benthic deposits. No definite
procedure can be given, but careful
effort should be made to obtain a rep-
resentative sample.
Ill SAMP LE IDENTIFICATION
A Each sample must be unmistakably
identified, preferably with a tag or label.
The required information should be planned
in advance.
B An information form preprinted on the
tags or labels provides uniformity of
sample records, assists the sampler, and
helps ensure that vital information will
not be omitted.
C Useful Identification Information includes:
1 sample identity code
2 signature of sampler
3 signature of witness
4 description of sampling location de-
tailed enough to accommodate repro-
ducible sampling. (It may be more
convenient to record the details in the
field record book).
5 sampling equipment used
6 date of collection
7 time of collection
8 type of sample (grab or composite)
9 water temperature
10 sampling conditions such as weather,
water level, flow rate of source, etc.
11 any preservative additions or Techniques
12 record of any determinations done in
the field
13 type of analyses to be done in laboratory
WP.SUR.sg. 6. 3.74
1-1
-------
Sample Handling - Field Througji Laboratory
IV SAMPLE CONTAINERS
A Available Materials
1 glass
2 plastic
3 hard rubber
B Considerations
1 Nature of the sample - Organics
attack polyethylene.
2 Nature of constituents) to be determined
- Cations can adsorb readily on soine
plastics and on certain glassware.
Metal or aluminum foil cap liners can
interfere with metal analyses.
3 Preservatives to be used - Mineral
acids attack some plastics.
4 Mailing Requirements - Containers
should be large enough to allow extra
volume for effects of temperature
changes during transit. All caps
should be securely in place. Glass
containers must be protected against
breakage. Styrofoam linings are
useful for protecting glassware.
C Preliminary Check
Any question of possible interferences
related to the sample container should
be resolved before the study begins. A
preliminary check should be made using
corresponding sample materials, con-
tainers, preservatives and analysis.
D Cleaning
If new containers are to be used, prelim-
inary cleaning is usually not necessary.
If the sample containers have been used
previously, they should be carefully
cleaned before use.
There are several cleaning methods
available. Choosing the best method in-
volves careful consideration of the nature
of the sample and of the constituent(s) to
be determined.
i Phosphate detergents should r.oi bo
used to clean containers for phosphorus
samples.
'2 Traces of dichromatic cleaning solution
will interfere v, ith metal analyses.
L Storage
Sample containers should be stored and
transported in a manner to assure their
readiness for use.
V SAMPLE PRESERVATION
Every effort should be made to achieve
the shortest possible interval between
sample collection and analyses. If there
must be a delay and it is long enough to
produce significant changes in the sample,
preservation measures are required.
At best, however, preservation efforts
can only retard changes that inevitably
continue after the sample is removed
from the parent population.
A Functions
Methods of preservation are relatively
limited. The primary functions of those
employed are:
1 to retard biological action
2 to retard precipitation or the hydrolysis
of chemical compounds and complexes
3 to reduce volatility of constituents
B General Methods
1 pH control - This affects precipitation
of metals, salt formation and can
inhibit bacterial action.
2 Chemical Addition - The choice of
chemical depends on the change to be
controlled.
Mercuric chloride is commonly used
as a bacterial inhibitor. Disposal of
the mercury-containing samples is a
problem and efforts to find a substitute
for this toxicant are underway.
1-2
-------
Samp!<• Handling - I'lciii Tlinuigh I
To dispose of solutions of inorganic
mercury salts, a recommended
procedure is to capture and retain the
mercury salts as the sulfide at a high
pH. Several firms have tentatively
agreed to accept the mercury sulfide for
re-processing after preliminary con-
ditions are
3 Refrigeration and Freezing - This is
the best preservation technique avail-
able, but it is not applicable to all
types of samples. It is not always a
practical technique for field operations.
C Specific Methods
The EPA Methods Manual includes a
table summarizing the holding times and
preservation techniques for several
analytical procedures. This information
also can be found in the standard refer-
ences (!» 2» 3) as part of the presentation
of the individual procedures.
VI METHODS OF ANALYSIS
Standard reference books of analytical
procedures to determine the physical
and chemical characteristics of various
types of water samples are available.
A EPA Methods Manual
The Methods Development and Quality
Assurance Research Laboratory of the
Environmental Protection Agency, has
published a manual of analytical procedures
to provide methodology for monitoring the
quality of our Nation's Waters and to deter-
mine the impact of waste discharges. The
title of this manual is "Methods for Chem-
cal Analysis of Water and Wastes. "<2>
For some procedures, the analyst is
referred to Standard Methods and/or to
ASTM Standards.
B Standard Methods
The American Public Health Association,
the American Water Works Association
and the Water Pollution Control Federation
prepare and publish a volume describing
methods of water analysis. These include
physical and chemical proceaures. The
title of this book is "Standard Methods
for the Examination of Water and Waste -
water. "<3>
C ASTM Standards
The American Society for Testing and
Materials publishes an annual "book"
of specifications and methods for testing
materials. The "book" currently con-
sists of 33 parts. The part applicable
to water is a book titled, "Annual Book of
ASTM Standards, Part 23. Water;
Atmospheric Analysis". '
D Other References
Current literature and other books of
analytical procedures with related in-
formation are available to the analyst.
E NPDES Methodology
When gathering data for National Pollutant
Discharge Elimination System report
purposes, the analyst must consult the
Federal Register for a listing of approved
analytical methodology. There he will be
directed to pages in the above cited
reference books where acceptable pro-
cedures can be found. The Federal
Register also provides information con-
cerning the protocol for obtaining approval
to use analytical procedures other than
those listed.
VII ORDER OF ANALYSES
The ideal situation is to perform all
analyses shortly after sample collection.
In the practical order, this is rarely
possible. The allowable holding time
for preserved samples is the basis
for scheduling analyses.
-------
Sample Handling - Field Through Laboratory
A The allowable holding time for samples
depends on the nature of the sample, the
stability of the constituent(s) to be de-
termined and the conditions of storage.
1 For some constituents and physical
values, immediate determination is
required, e.g. dissolved oxygen, pH.
2 Using preservation techniques, the
holding times -for other determinations
range from 6 hours (BOD) to 7 days
(COD). Metals may be held up to 6
months.^2'
(2)
3 The EPA Methods Manual includes
a table summarizing holding times and
preservation techniques for several
analytical procedures. This information
can also be found in the standard
(123)
references ' ' as part of the
presentation of the individual
procedures.
4 If dissolved concentrations are
sought, filtration should be done in
the field if at all possible. Other-
wise, the sample is filtered as soon
as it is received in the laboratory.
A 0.45 micron membrane filter is
recommended for reproducible
filtration.
B The time interval between collection
and analysis is important and should be
recorded in the laboratory record book.
VIII RECORD KEEPING
The importance of maintaining a bound,
legible record of pertinent information
on samples cannot be over-emphasized.
A Field Operations
A bound notebook should be used. Informa-
tion that should be recorded includes:
1 Sample identification records (Sec
Part UI)
2 Any information requested by the
analyst as significant
3 Details of sample preservation
4 A complete record of data on any
determinations done in the field.
(See B, next)
5 Shipping details and records
B Laboratory Operations
Samples should be logged in as soon as
received and the analyses performed
as soon as possible.
A bound notebook should be used.
Preprinted data forms provide uniformity
of records and help ensure that required
information will be recorded. Such sheets
should be permanently bound.
Items in the laboratory notebook would
include:
1 sample identifying code
2 date and time of collection
3 date and time of analysis
4 the analytical method used
!i any deviations from the analytical
method used and why this was done
6 data obtained during analysis
7 results of quality control checks on the
analysis
8 any information useful to those who
interpret and use the data
9 signature of the analyst
1-4
-------
Sample Handling - Field Through Laboratory
IX SUMMARY
Valid data can be obtained only from a repre-
sentative sample, unmistakably identified,
carefully collected and stored. A skilled
analyst, using approved methods of analyses
and performing the determinations within
the prescribed time limits, can produce data
for the sample. This data will be of value
only if a written record exists to verify sample
history from the field through the laboratory.
REFERENCES
1 ASTM Standards, Part 23, Water; .
Atmospheric Analysis.
2 Methods for Chemical Analysis of Water
and Wastes, EPA-MDQARL,
Cincinnati, OH 45268. 1974.
3 Standard Methods for the Examination of
Water and Wastewater, 13th edition,
APHA-AWWA-WPCF, 1971.
Dean, R., Williams, R. and Wise, R.,
Disposal of Mercury Wastes from
Water Laboratories, Environmental
Science and Technology, October, 1971.
This outline was prepared by A. Donahue,
Chemist, National Training Center, MPOD,
OWPO. EPA, Cincinnati, Ohio 45268.
Descriptors: On-Site Data Collections,
On-Site Investigations, Planning, Handling,
Sample, Sampling, Water Sampling,
Surface Waters, Preservation, Wastewater
1-5
-------
STATISTICS FOR CHEMISTS
I INTRODUCTION
A Statistics may be defined, for our purpose,
as a collection of methods which have been
developed for handling numerical data
pertaining to samples or portions of entire
populations.
B The statistical methods with which we will
concern ourselves deal with the presentation
and analysis of numerical data from samples.
H FREQUENCY
A Definitions
1 Frequency - indicates how many times
a particular score occurs in a collection
of data
Frequency table - a tabular arrange-
ment of data, ranked in ascending or
descending order of magnitude,
together with the corresponding
frequencies
Frequency histogram - a set of
rectangles having bases on a horizontal
axis with centers at the given scores
and heights equal to the corresponding
frequencies (See Figure 1)
Frequency polygon - a line graph of
frequencies plotted against scores
(can be obtained by connecting mid-
points of tops of rectangles in the
frequency histogram) (See Figure 1)
Figure 1
Frequency Histogram & Frequency Polygon
Frequency
>-> CO GO 4^ O
1 1 1 1 1
X
X
X
X
j
X
^
V
s
X
\
s
\
\
\
1 1 1 1 1
98 99 100 101 102
Chloride ug/l
ST.25b.11.75
2-1
-------
Statistics For Chemists
B Application
Consider the application of the above
definitions to the following set of data,
obtained from twelve determinations for
chloride in water.
Results (tig/1)
100 101 99
101 100 100
99 102 100
98 101 102
numbcT of obscrvntions Hie iiiedian is
'-'! * I
tin1 :i v t' ra^r < >1 LIi
middle two scores.
4 Mean - arithmetic average of all the
values in the sample distribution, de-
noted by X. The formula for calcula-
ting the sample mean is
X= *!-
x
n
£ X,
Table 1
Frequency Table
Chloride (ng/1)
98
99
100
101
102
Frequency
1
2
4
3
2
II MEASURES OF CENTRAL TENDENCY
A Definitions
1 Central tendency - the tendency of
values to cluster about a particular
value in the distribution
2 Mode - that value which occurs most
frequently
3 Median - midpoint of an array of
scores. If there is an odd number of
observations, n, the median is
xn +1 where Xn +
2
n+ 1
2
distribution.
represents
the
value in the frequency
If there is an even
4
X = —l where there are n number
of values.
B Aids in calculation of the mean
Application of the following two statements
can reduce errors and amount of time
spent in calculating the mean of a
distribution.
1 Adding or subtracting a constant to or
from each score in a distribution is
equivalent to adding or subtracting the
same constant to or from the mean of
the distribution. Thus the following
formula:
X = X± C
c
where the Xj's are the
values in the distribution with mean X,
and the Xj ± C's are the_values in the
distribution with mean Xc.
Multiplying or dividing each score in
a distribution by a constant is equivalent
to multiplying or dividing the mean of
the distribution by the same constant.
Thus the following formulas:
(1) Xc = CX
or
(2) Xc = - where the X^s are the
values in the distribution with mean X,
2-2
-------
Statistics for Chemists
and the CXj's or the —±- "s are the
v_alues in the distribution with mean
X
C Application
Consider the application of the above
definitions to the previously mentioned
set of data, obtained from twelve deter-
minations for chloride in water, shown
in Table 1.
1 Mode = 100
100 + 100
2 2
2
= inn
_ X6 + X7
2
Denote the mean of the distribution in
Table 1 by X . If we add 100 to each
score in the distribution in Table 2, we
obtain the scores in the distribution in
Table 1; likewise if we add 100 to the
mean, X, of the distribution in Table 2,
we obtain the mean, X , of the distri-
bution in Table 1.
Thus X = X + 100
c
X = -=
c n
100
- = l(-2) + 2(-l) + 4(0) + 3(1) + 2(2) +
X
12
100
X = .25 + 100 = 100.25
c
3 Mean
n
_ 98 + 2 (99) + 4 (100) + 3 (101) + 2 (102)
12
= 100.25
4 Aid in Calculation
Consulting Table 1 and observing that
the values are in the neighborhood of
100 we might subtract 100 from each
score and obtain the following distribution:
Table 2
Frequency Table
Chloride (fig/1) Frequency
-2 1
-1 2
0 4
1 3
2 2
IV MEASURES OF DISPERSION
A Definitions
1 Dispersion - spread or variability of
observations in a distribution
2 Range - the difference between the
highest value and the lowest value
R = max - min
3 Average deviation - the sum of the
deviations of the values from their
mean, without regard to sign, divided
by the total number of data values (n)
The formula for calculating the average
deviation is:
2-3
-------
Statistics for Chemists
Average deviation of the mean (D) -
the average deviation of individual
data items from the mean (d) divided
by the square root of the number of
data items (n)
The. definition of the average deviation
of the mean can be expressed by the
formula:
d
D =
Variance - the sum of the squares of
the deviations of the values from their
mean divided by the total number of
data values (n) minus 1
The definition of the variance can be
expressed by the following formula:
- X)2
n - 1
Standard deviation - the square root
of the variance
The definition of the standard deviation
can be expressed by the following
formula:
- X)2
However, the formula commonly used
because of its adaptability to the hand
calculator is the following:
a -
n - 1
where there are
n number of values.
The dei'iuil.iim ui' the .stundur
deviation of the mean can be
expressed by the- formula:
S =
s
n
8 Relative standard deviation - the
standard deviation (s) expressed as
a fraction of the mean, s
X
The relative standard deviation is
often expressed as a percent. It is
then referred to as the coefficient
of variation (V):
V = ;— X 100 = %
X
The relative standard deviation is
particularly helpful when comparing
the precision of a number of deter-
minations on a given substance at
different levels of concentration.
B Aids in Calculation
Application of the following statements
can reduce errors and amount of time
spent in calculating the variance or
standard deviation of a distribution.
1 Adding or subtracting a constant to or
from each score in a distribution
doesn't affect the variance or standard
deviation of the distribution.
Thus the following formulas:
(D
2 2
s = s
c
Standard deviation of the mean (S) - the
standard deviation of individual data
items (s) divided by the square root of
the number of data items (n)
(2) s = s
c
where the X^'s are the values in
the distribution with variance s
and standard deviation s, and the
X.- + C's are the values in the
A — o
distribution with variance s
and standard deviation s .
2-4
-------
Statistics for Chemists
2 Multiplying or dividing each score in
a distribution by a constant is tMjuivnlent
to multiplying or dividing the variance
of that distribution by the square of the
same constant.
Thus the following formulas:
9 99
(1) B* * C S
C _
2 s2
(2) s 2 - £»
c £•*
2
n
1
2
4
3
2
-*
98
99
100
101
102
X = 100.
deviation - d
IXi- X|
2.25
1.25
. 25
. 75
1.75
25
nlX
2
2
1
2
3
11
-------
Statistics for Chemists
f
5 Standard deviation - s = / •
si
n Xj nXj X ,
1 98 98 9604
2 99 198 9801
4 100 400 10000
3 101 303 10201
2 102 204 10404
1203
/ 2
/*>•» -3? /£
a V n v
,„ (3)2
2 (£XiH 2 . 12 _ 16.25 .
£Xi - n £ u n !•«
n - 1
nX^i
9604 /EX2- (EX )2
s = / i i /I 48 ^ 1 ° '"*
19602 / -5^ y
40000 * n - 1
30603
20808 7 Standard deviation of the mean -
120617 s
vTir
ci* ioncm Using calculations from number 6,
91 1 - llUOUl
11
rTT » 1.21
6 Aid in calculation
Recalling that adding or subtracting a
constant to each score in the distri-
bution doesn't affect the variance or
the standard deviation of the distribu-
tion we can simplify the computations
by first subtracting 100 from each
score in the distribution, thus obtain-
ing the frequency distribution shown
in Table 2.
n
1
2
4
3
2
nCXj-C) (XrC)2 n(XrC>2
n - 1
8 Relative standard deviation expressed
as a percent (coefficient of variation)
v=-S- xioo
X
Using calculations from number 6 for
s = 1. 22 and from number 2 for
X = 100.25,
V =-
X
1.22
100.25
X 100
Figure 2
Normal Distribution Curve
Quantity Measured
2-6
-------
Statistics for Chemists
V INTRODUCTION TO NORMAL
DISTRIBUTION CURVE
A Statistics deals with theoretical curves
which are smoother than frequency
polygons, obtained from experiments In
real life. However, frequency distribu-
tions or frequency polygons of experimental
data often approximate a mathematical
function called the "normal" distribution
curve. (See Figure 2)
As shown in Figure 3, the frequency polygon
for the 12 determinations for chloride in
water is a fairly good approximation of the
normal curve. If, however, in the chloride
determinations we had obtained 103 instead
of 98 and 104 instead of 99 this distribution
would not have- been a good approximation of
the normal curve, as is shown in Figure 4.
Figure 3
Comparison of Normal Curve and Frequency Polygon
o
0)
g.
V
(-,
fc
97 98 99 100 101
Chloride ng/1
102
Figure 4
Comparison of Normal Curve and Frequency Polygon
o
0)
§.
0>
99
2-7
-------
Statistics for Chemists
If a frequency distribution is a good
approximation of the normal curve, we
can use some facts about the normal
curve to give us information about the
frequency distribution.
Figure 5 shows the normal distribution
in terms of the population mean a, and
the standard deviation of the population
a , and gives the percent of area under
the curve between certain points.
Figure 5
Normal Distribution Curve
+ l
-------
Statistics lot1 C
We may check the distribution of sample
data to see if it is a "normal" distribution
in the following manner. Substitute the
value of the sample mean (X) for the value
of the midline and substitute the value of
the sample standard deviation (s) for the
limits of the value spans where we might
expect certain percentages of the data
items to occur. Then we can check the
number of data items which actually do
occur within these value spans.
Figure 6 demonstrates this application
using the chloride data values from Table 1.
The data values are marked on the hori-
zontal line and the frequency of the
occurrence of each value is marked on the
vertical. The midline of the distribution
is marked at the value of the sample mean
(X = 100, See m C 3). The value of the
sample standard deviation (s = 1.21, See
IV C 5) is used to mark value areas under
the curve where different percentages of
data values will probably occur. Thus,
for the area X + Is, X - Is * 98. 79 and
X + 1 B - 101.21. Therefore, according
to the normal distribution curve shown in
Figure 5, we might expect about 68% of the
data items to have values between 99 and
101. (The values are rounded to whole
numbers since the data values are thus
recorded).
Consulting Table 1. we find that 75% or 9
of the 12 data items have values in this
range. This percentage is shown in
Figure 6 by the frequency polygon for the
data shown earlier in Figure 3.
Likewise assuming a normal distribution,
we would expect 95% of the observations
to lie within + 20 's from the population
mean. In fact, 100% of the observations
were within + 2 s's from the sample mean.
In both cases the observed percentages are
reasonably close to the expected percentages.
Other tests exist for determining whether
or not a frequency distribution might
reasonably be assumed to approximate
the normal distribution.
It would be good to become as familiar as
possible with the' normal distribution siru
an underlying normal distribution in
assumed for many statistical tcwta ol
hypothesis.
REFERENCES
1 Bennett, C.A. and Franklin, N. L.
Statistical Analysis in Chemistry and
the Chemical Industry. John Wiley
& Sons, Inc., New York. 1954.
2 Crow, E. L., Davis, F.A., and Maxfielri,
M.W. Statistics Manual. Dover
Publications, Inc., New York. 1960.
3 Dixon, W.J. and Massey, F. J.
Introduction to Statictical Analysis.
McGraw-Hill Book Co., Inc., New
York. 1957.
4 Ostle, B. Statistics in Research. The
Iowa State University Press, Iowa,
1963.
Youdon, W.J.
Chemists.
New York.
Statistical Methods for
John Wiley & Sons, Inc
1951.
This outline was prepared by L. A. Lederer,
Statistician, formerly with Analytical
Reference Service. Training Program,
NCUIH, SEC. Revised by Audrey E.
Donahue, Chemist, National Training Center,
MPOD, OWPO. USEPA. Cincinnati, OH 45268
Descriptors: Graphic Methods, Quality Control,
Statistical Methods, Statistics
9-0
-------
ACCURACY-PRECISION-ERROR
I INTRODUCTION
An analytical method is subject to errors.
These errors may affect the accuracy of the
method because they introduce bias into the
results. There are other types of errors
which affect the precision of the method
because they produce random fluctuations in
the data. The most desirable situation for
the analyst is shown in the diagram in
Figure 1 where the results are both precise
and accurate.
X
X
IMPRECISE AND INACCURATE
PRECISE BUT INACCURATE
ACCURATE BUT IMPRECISE
PRECISE AND ACCURATE
Figure 1. PRECISION AND ACCURACY
For results to be accurate, the analysis
used must give values close to the true
value.(See Figure 1)
B Precision
Precision is the degree of agreement
among results obtained by repeated
measurements on a single sample under
a given set of conditions. It is a measure
of the degree to which results "check. "
(See Figure 1)
C Note
It is possible to have precision without
accuracy. (See Figure 1)
II DETERMINATE ERROR AND ACCURACY
A determinate error is one which con-
tributes a constant error or bias to results,
causing them to be inaccurate. This
constant error makes it possible for results
to agree with each other (be precise) and
still be inaccurate.( See Figure 1)
Determinate errors have "assignable"
causes which can usually be identified and
either eliminated or controlled. (The terms
"determinate" error, "assignable" error,
and "systematic" error are synonymous).
A Sources of Determinate Error
1 Method error
Method errors are those that are
inherent in the procedure. These are
very serious and the hardest to detect
and correct. The most common
method error is the presence of inter-
ferences in the sample. Other
examples would be precipitation of
substances other than the desired
material, partial solubility of pre-
cipitates, and entrainment as in a
solvent extraction procedure.
CH. MET. con. lc. 11.75
v-1
-------
Accuracy-Precision-Error
2 Personal errors
Personal errors are attributable to
individual mistakes which are con-
sistently made by an analyst. These
errors are the result of consistent
carelessness, lack of knowledge or
personal bias. Examples are errors
in calculations, use of contaminated
reagents, non-representative sampling,
or poor calibration of standards and
instruments.
3 Instrumental errors
Instrumental errors are those which
are caused by an analytical instrument
or by the effects of the environment
acting on the instrument. Moisture in a
G.C. column, improper wavelength
markings on a spectrophotometer or
incorrect scoring on a buret would be
examples.
B Effects of Determinate Error
1 Additive
An additive determinate error is one
which has a constant value regardless of
the amount of analytically sought con-
stituent present in the sample.
(See Figure 2)
2 Proportional
A proportional determinate error
changes value according to the amount
of analytically sought constituent in the
sample. (See Figure 3)
IH DETECTION OF DETERMINATE ERROR
A Spiked Samples
1 Samples which can be determined before
and after the addition of a known con-
stituent ( in the concentration range of
interest) provide a way to detect
determinate errors. Spiked samples
should be representative and resemble
actual conditions as closely as possible.
The quantitation of bias can then be
obtained with the following measures.
a Mean error - the difference between
the mean of the data and the I nu-
n-suit.
b Uelative error - the mean error ol
a set of data expressed as a per-
centage of the true result.
EXAMPLE: An analyst determines
the nitrate content of the effluent
from his sewage treatment plant to be
0.5 mg/1. He then adds 1 mg/1 of
standard nitrate solution to the
sample. Table 1 shows the replicate
results obtained on the spiked
sample, and calculation of both
mean and relative errors.
Table 1
Sample: Effluent
Determination: Nitrate (Modified Brueine)
X.
i
1.55 mg/1
1.58
1.60
1.47
1.35
1 Calculation of mean error
7.55
X =
I. 51 mg/1
Mean error = +0. 01 mg/1
2 Calculation of relative error
0.01 X 100
Relative Error =
1.50
0.7'
2 Control charts
Trends and shifts on control charts
may also indicate determinant error.
Using spiked samples, the standard
deviation is calculated and control
limits (usually + 3 standard deviations)
for the analysis are set (see Figure 4.
For further discussion of control limits,
see reference 3, p. 62.
3-2
-------
4r-
1234
THEORETICAL VALUE
Fipure 2. ADDITIVE ERROR
1234
THEORETICAL VALUE
Figure 3. PROPORTIONAL ERROR
300
,23456 89
100
Figure 4. CONTROL CHART
-------
Accuracy-Precis ion-Error
3 In some cases, it is impossible to
spike a sample so that it resembles
actual conditions (e.g., BOD and
pesticide samples). Youden '**•'
provides excellent techniques for
detecting bias in this situation.
B Unknown Samples
1 Independent method
Analysis of a sample for a desired
constituent by two or more methods
that are entirely different in principle
(gravimetric and volumetric) may aid
in the estimation of determinate error.
However, another reliable method may
not be available or may be laborious to
perform.
2 Control charts
It is possible to plot a control chart
(Figure 4) even when it is not possible
to spike a sample. One can use as a
reference value an average of a series
of replicate determinations performed
on a composite check sample. Such a
sample must be preserved or stabilized
in such a way that the concentration
of the constituent being measured will
not change from day to day (see
reference 1).
3 Aliquoting
If the determinate error is additive,
the magnitude may be estimated by
plotting the measured quantity versus
a range of sample volumes or sample
weights. If the error has a constant
value regardless of the amount of
analytically sought constituent, then
a straight line fitted to the points will
not go through the origin. (See Figure 5)
C Youden's Graphical Technique(10, 11, 12)
Dr. W. J. Youden has devised an
approach to test for determinant errors
with a minimum of effort on the part
of the analyst and his laboratory.
Samples used may be of known (spiked)
or unknown composition.
1 Technique
Two different test samples (X and V)
are prepared and distributed for
analysis to as many individuals or
laboratories as possible. lOach
participant is asked to perform
only one determination on each
sample (NOTE: It is important
that the samples be relatively
similar in concentration of the
constituent being measured).
Table 2 shows the results on two
such samples analyzed for percent
potassium by 14 different laboratories.
The mean for each sample has been
calculated.
2 Interpretation - Figure 6
The vertical line drawn on the
graph represents the mean (X)
of all the results obtained on
Sample X; the horizontal line was
drawn through the mean (Y) of all
the results obtained on sample Y.
Each pair of laboratory results
can then be plotted as a point on
the graph (marked X. 1 etc.)
If the ratio of the bias (error) to
standard deviation is close to zero
for the determinations submitted
by the participants, then one would
expect the distribution of the paired
values (or points) to be close to
equal among the four quadrants.
The fact that the majority of the
points fall in the (+, +) and (-, -)
quadrants indicates that the results
have been influenced by some source
of bias or determinate error.
Furthermore, one can even learn
something about a participant's
precision from the graph. If all
participants had perfect precision
(no indeterminate error), then all
the paired points would fall on a
450 line passing through the origin.
Consequently the distance from such
a 450 line of each participant's
point provides an indication of that
participant's precision.
3-4
-------
Accuracy-Precision-Error
Tin; Yoiirk-n method lf> ali-io disi:»ssi%tl
in thin manual in llu- Hec-tion on
Inter laboratory Quality Control
Studies. A quantitative treatment
of this subject can be found in
references 11 and 12.
10i-
1 8
Z 6
O
2
1234
SAMPLE VOLUME (ml)
Figure 5
Laboratory
1
Sample X
9.74%
9.92
9.98
9.99
10.00
10.11
10.12
10.14
10.19
10.23
10.25
10.29
10.55
10.62
LI niplc N
8.50":,
8.28
8. 84
8.24
8.73
8.54
8.64
8.82
9.04
8.93
8.97
8.80
9.21
8.95
X = 10.15
Y = 8. 75
9.40
9.20
9.00
8 80 -
8.60
8.40
8.20
8.00
£
^
— QJ- -
a.
*
i/i
Y
X
1
Tt
3
X
8
X.
S
)(.
X
X
6
2 X Y
SAf
4PLE X (
11
X
10
%K)
13
X
4
9.00 9.20 9.40 9.60 9.80 10.00 10.20 10.40 10.60 10.80 11.00
Figure 6. YOUDEN'S GRAPHICAL TECHNIQUE
-------
Accuracy-Precision-Error
IV ELIMINATION OF DETERMINATE
ERROR
Various approaches can be used to eliminate
the source of determinate error. The
approach used depends upon whether the
source is personal, method, or instrumental.
A Personal
Great care must be used to avoid producing
an unconfldent attitude in a technician. It
is undesirable for the analyst to feel he is
being "policed. "
B Method and Instrumental
1 Blanks
Blanks can be used to correct for
interferences from reagents, sample
color, etc.
2 Correction factors
Examples of correction factors used
in environmental analyses to eliminate
determinate errors are the following:
a Recovery factors in organic
extractions
b Chemical yield values in gravimetric
analyses
c Counting efficiencies for radiation
counters
3 Standard addition
1 l|;un- 7. Mil!','',
AUDI ! IO\ (. li '• I''' 'C \ i \ll I 111 * ' ••
M- ni\i i \ ni\ i u .\
t 90
I 7°
g so
\n
,.<<
NOTK: The X value appearing along the
abscissa (i.e., \ + 20, X •> 40, etc.) re-
fers to the unknown added to the strontium
standard
4. Standard compensation
Another approach is to prepare the
standard so that its composition
resembles that of the sample as
closely as possible. The objective
of the approach is exactly the same
as that of standard addition - to
compensate for the presence of
interfering substances in the unknown.
V INDETERMINATE ERROR AND
PRECISION
Sample interferences producing deter-
minate errors can be overcome by
adding equal amounts of unknown
sample to a series of standards. The
concentration of the unknown can then
be determined graphically from a plot
of the measured quantity (absorption,
emission, etc.) versus the standard
concentration. (See Figure 7)
Even when all determinate errors are
eliminated, every replicate analysis will
not give the same value. Such variation
in results is due to indeterminate error,
also known as random, chance or
uncontrollable error. Indeterminate
error affects the precision or agreement
among results.
-------
v-Precision- Error
A Sources of Indeterminate Error
Indeterminate errors are due to
unassignable or "chance causes."
Examples are Inadvertent contamination
of sample or glassware, variation in
reagent additions, or variations in
instrument response (see Figure 8).
B Effects of Indeterminate Error
Since the causes of these errors are
random, the effects are also random.
Fortunately, these random variations
conform to the "Laws of Chance" so
statistical measures of precision can be
used to quantitate indeterminate errors.
VI DETECTION OF INDETERMINATE
ERROR
A measure of the degree of agreement among
results can be obtained by analyzing a single
sample repeatedly under a given set of
conditions.
A Range
The range of the replicate results
(difference between the lowest and the
highest value) provides a measure of
indeterminate variations.
B Standard Deviation
An estimation of indeterminate error can
be obtained through a calculation of the
standard deviation. The following formulas
should be applied to random data which
follows a normal distribution. Normality
can be checked by ranking and plotting the
data on normal probability paper; it should
fall on a straight line (see Figure 9). Any
values which do not fall close to the straight
line (those values entire-led) ohould be
rejected in the calculation oi standard
deviation. Other statistical U:sts W
can be used to objectively evaluate the
rejection of outliers. The value obtained
for standard deviation for a particular
method may vary with the analyst, the
concentration range of the constituent,
and the composition of the sample
analyzed. The confidence of the
estimate is increased as the number of
results (n) used to compute the standard
deviation is increased.
1 Replicate results on the same sample:
(1)
s =
Xi
X =
f£(X. - X)2
n-1
value of single result
average (mean) of results on
same sample
n = number of results
Example : Table 3 contains a set of 5- day
BOD results obtained on a synthetic
sample containing 150 mg/1 of glucose
and 150 mg/1 glutamic acid. (Note: A 1%
dilution was used in the actual test). The
results are those submitted by laboratories
participating in a colaborative study .
Calculate the standard deviation of these
results.
X = 192 mg/1
36.0
= 58,200
-------
Accuracy-Precision-Error
2 3
BECKMAN B (650 m*)
METHYLENE BLUE SOLUTION
1cm CELL
25 REPLICATE READINGS
.528 .530 .532 .534 .536
ABSORBANCE
Figure 8. VARIATIONS IN SPECTROPIIOTOMETKR READINGS
FIGURES NORMAL PROBABILITY CURVE
»»,•» tMIM ft » « fO •> TO «0 90 40 )0 ZD
1*0
1M
340
uo
300
2*0
2ZO
ZOO
i.n
i.eo
140
I.ZO
1.00
•^
•
^
•
I •
^*
•
™
,-^x
4.M
0
•
i ,
L — '
oft aa
a 0460J ojasil 9 » to»«oso«i?oio we M» M*m* ntt
plclliiif mltnal - -^t * 2.0*3
-------
Accuracy-Freciaion-l^rror
SAMPLE;
Table 3
160 mg/1 gliu -use i 150ms /I ulutnmic ac'itl ( !"• dilution)
DETERMINATION. 5-day Biocheinu al Oxygen IVmun<1
X
i
100 mg/i
117
125
132
142
147
153
160
165
165
167
173
(78
18'J
190
196
19C
1U7
'••Data taken
Service, Tr
vx
-92 mg/1
-75
-67
-GO
-50
-45
-3J
-32
-27
-27
-25
-19
-14
- 3
- 2
4
4
5
(X| - X)" X.
8464 (mg/1)2 198 mg/1
5625 199
44H9 200
3600 200
2500 204
2025 210
1521 211
1024 212
729 215
729 223
625 224
361 227
196 229
9 238
4 247
16 250
16 ?59
U!j 274
Iron) ttnter. Oxygen Demand Report (Julj
unmg 1' run ram
H. A. Tuft Solitary l-'.np.n
Table 4 *
SAMPLE: Aqueous
DETERMINATION: Phosphate (Lucerna, Cet.du, and
Sample
A
B
C
D
E
F
G
H
I
J
K
Mg/1 P
51
53
33
39
53
54
47
47
50
51
48
47
50
50
47
47
42
42
50
51
59
60
2
d d Sample
2 4 L
0 0 M
1 1 N
0 0 O
1 1 P
1 1 Q
0 0 R
DOS
GOT
1 1 U
1 1 V
Vx
6 mg/1
7
3
8
12
18
19
20
23
31
32
35
37
46
55
58
67
82
, IPfiO), Analyi
leerinp CVuti'f.
(X. - X)"
36 (mg/1)2
49
64
64
144
326
361
400
529
961
1024
1225
1369
2116
3025
3::44
4489
67?.4
ii ,il Reference
Cuii'itmati, Ohio.
CONCENTRATION
RANGE:
Prat Method)
Mg/l P
54
56
38
37
52
52
58
58
54
54
54
54
43
48
52
51
53
53
46
47
42
42
30 - 70 mg/1
d d2
2 4
1 1
0 0
0 0
0 0
0 0
0 0
1 1
0 0
1 1
0 0
*Data obtained from Frank Scluckner. Proctor and Gamble Company.
3
-------
A ccuracy-Precision- Error
2 Duplicate results on different samples
In a laboratory where duplicates are
routinely run, it would be simple to
use the following formula for evaluation
of standard deviation.
/ £(d ) (See reference 5,
S = J 2k page 654) (2)
d = difference between duplicates
k = number of samples
Example: Table 4 contains a set of
phosphate results obtained on aqueous
samples in the range 100-200 mg/1.
Calculate the standard deviation of these
results.
:(d2)
16
22
8 = .61 mg/1
3 Duplicate and triplicate results on
different samples
•i
2(X. - X)2
n - k
(See reference 7,
page 73) (3)
Note: X = average of results on the
same sample
Example: Table 5 contains a set of %
nitrogen results obtained on unknown
organic compounds.
Calculate the standard deviation.
4 Use of range iO estimate standard
deviation
For a small number of replicates
(n < 10), the range can be used to
estimate the standard deviation (See
Table 6 ).
R
dN
= s
(4)
Example: Table 7 contains a set of
replicate nitrate results. Calculate
the standard deviation of these results.
a Use of formula (1)
X = 0.72 mg/1
n = 5
S(X. -X)2 = .0134
S =t
s = . 058 mg/1
b Use of formula (4)
R * . 14 mg/1
N
2.33
.14 mg/1
2.33
s = .060 mg/1
C Coefficient of Variation
An estimation of indeterminate error
can also be made by calculating the
coefficient of variation.
V = ;- X
X
100
By comparing the standard deviation
(dispersion) to the average or mean
value (central tendency) in a set of data
and expressing this relative standard
deviation as a percentage, the analyst
has a meaningful interpretation of the
degree of dispersion present. Indeter-
3-10
-------
Accuracy-Precision-Error
SAMPLE:
DETERMINATION:
Table 5*
Unknown Organic Compounds
•7. Nitrogc:. (Kjeldahl)
CONCENTRATION-
RANGE: lOfc - 20%
Sample
A
B
C
D
E
F
G
H
I
%N
16.48
16.45
16. 50
16.49
16.72
16.57
17.52
17.60
17.63
16.31
16j30
16.40
16.31
16.35
17.56
17.54
14.96
14.66
19. 15
18.89
*
16.47
16.50
16.65
17.58
16. 3J
16.35
17.55
14.81
19.02
x - !t
i
.01
.01
0
.01
.07
.08
.06
.02
.05
0
.0;
.05
.04
0
.01
.01
.15
. 15
.13
. 13
(X - XT
i
.0001
.0001
0
.0001
.0049
.0064
.0036
.0004
.0025
0
.0001
.0025
.OOlfi
0
.0001
.0001
.0325
.0225
.0169
.0169
Sample
J
K
L
M
N
O
P
Q
R
ToN
13.93
13.56
10.34
10. 19
17. 16
17. 13
15.01
15.05
12.44
12.70
12.73
14.37
14.36
11.85
11.85
14.79
14.70
14.70
17. 19
17. 14
X
13.75
10.27
17. 15
15.03
12. G2
14.37
11.85
14.73
17. 17
X - X
i
. 18
, Ib
.07
.08
.01
.02
.02
.02
. 18
.08
. 11
0
.01
0
0
.06
.03
.63
.C2
.03
(X. - X)2
. 0324
. 0361
.0049
.0064
.0001
.0004
.0004
.0004
.0324
.0064
.0121
0
. 000 1
0
0
.0036
.0009
.0009
.0004
.0009
*D«ta obtained from Frank Schickner, Proctor and Gamble Company.
FACTORS
Size
of Sample
2
3
4
5
6
7
8
9
10
*Natrella,
Table 6*
USED TO ESTIMATE THE STANDARD
DEVIATION FROM RANGE
(n) ^
1. 13
1.69
2.06
2.33
2. 53
2.70
2.85
2.97
3.08
(8)
Experimental Statistics, pp.
1
dN
.887
.591
.486
.430
.395
.370
.351
.337
.325
2-6.
Table 7
SAMPLE: Ohio River Water
DETERMINATION:
Xi
0.65 mg/lN
0.68
0.70
0.76
0.79
Nitrate (Modified
x.-x
-0.07
-0.04
-0.02
+0.04
+0.07
Brucine)
(X.-X)2
.0049
.0016
.0004
.0016
.0049
11
-------
AOOUIMO% -Precision-Error
minate error would be the cause of undue
dispersion.
Example: Table 3
s = 41 mg/1
X = 192 mg/1
V = ^ X 100
V • 21%
An indication of an individual's analytical
precision can be obtained by his partic-
ipation in an interlaboratory study. In
this case, the analyst performs only
one determination on each of two samples.
Using Youden's Graphical Technique as
an indication of an individual's precision
was discussed previously in this outline
in m C. Also see the outline on "Inter-
laboratory Quality Control Studies."
REFERENCES
1 Accuracy in Clinical Chemistry. Dade
Reagents, Inc., Miami, Florida.
2 Allan, Douglas H. Statistical Quality
Control. Reinhold Publishing Corp.,
New York. 1959.
3 American Society for Testing Materials
ASTM Manual on Quality Control of
Materials. Special Technical
Publication 15-C. 1951.
Bauer, E. L.
Chemists.
1960.
A Statistical Manual for
Academic Press, New York.
5 Bennett, Carl A. and Franklin, Norman L.
Statistical Analysis in Chemistry and
the Chemical Industry. John Wiley &
Sons, Inc., New York. 1954.
6 Chase and Rabinocvitz. Principles of
Radioisotope Methodology. Minneapolis
Burgess Publishing Company. 1964.
7 Mickley, Harold S., Sherwood, Thomas K.
and Reed, Charles E. Applied
Mathematics in Chemical Engineering.
McGraw-Hill Book Company, New York.
1957.
8 Natrella, M.G. Experimental Statistics,
National Bureau of Standards Handbook
91. U.S. Dept. of Commerce. 1963.
9 Schickner, Frank A. Personal Com-
munication. The Procter & Gamble
Company, Miami Valley Laboratories,
Research & Development Department,
P.O. Box39175, Cincinnati, OH 45239.
10 Youden, W.J. The Collaborative Test.
JAOAC 46:55-62. January 1963.
11 Youden, W.J. The Sample, The
Procedure, and the Laboratory.
Anal. Chem. 32:23-37A. Dec. 1960.
12 Youden, W.V. Statistical Techniques
for Collaborative Tests. Association
of Official Analytical Chemists, Box 540,
Benjamin Franklin Station, Washington,
DC . 1967.
This outline was prepared by Betty A.
Punghorst, Chemist, formerly with
National Training Center, and revised by
Audrey E. Donahue, Chemist, National
Training Center. MPOD. OWPO, USEPA,
Cincinnati, Ohio 45268
Descriptors: Accuracy, Data Collections,
Data Processing, Error Analysis, Errors,
Graphic Methods, Measurement, Monitoring,
Precision, Quality Control
3-12
-------
INTERLABOHATOKY QUALITY CONTUOL STUOIKS
I DEFINITION OF INTERLABORATORY
STUDY IN EPA
A It is a between-laboratory evaluation of
exact physical, chemical or biological
methods which yield a measurement of
desired properties.
B It is not an interlaboratory evaluation of
materials, reagents, or different test
conditions. It is not an initial study of a
method, nor a study to develop methods.
II PURPOSES OF INTERLABORATORY
STUDY
A Selection of an analytical method.
B Evaluation of an analytical method.
C Evaluation of laboratory and analyst
performance.
I PREPARATION FOR INTERLABORATORY
STUDY
A Laboratories are in control. Between-
replicate deviation will then be very
small and uniform in all laboratories.
B Ruggedness testing of a method has been
completed by the method developer, or
by a single qualified laboratory. Effects
of small changes in time, temperature,
pH, reagent and sample volume measure-
ments, etc., are known and corrected
for"7
C All laboratories and analysts are familiar
with the method to be tested. Any special
equipment or reagents required are
known and available., If all analysts are
not acquainted with the method, a simple
preliminary study is undertaken to
accomplish this. '
IV INSTRUCTIONS FOR THE STUDV
A Exact method write-ups are provided
to each analyst. The technical
description of an analytical method
is difficult. The language and organiza-
tion must be complete, yet simple and
unambiguous. These characteristics
are often conflicting.
Requirements in a method may seem
trivial. Yet if these are not recognized
and controlled, the entire study can be
useless.
B Explicit directions for sample preserva-
tion, sample make-up, time limitations,
sequence of analyses, etc. , are provided.
C Advance notice of tests is given so that
laboratories can integrate the test into
their program and realistic deadlines
can be established. Once established,
deadlines for agreement to participate,
for completion of analyses and for report-
ing are followed closely.
V THE TEST SAMPLES
A The sample is carefully designed to refleo:
the concentrations found in natural waters
and yet be within the best portion of the
concentration range for the method.
B Since precision of almost all methods
varies with concentration, a compre-
hensive study includes several levels of
concentration.
C Since accuracy of a method is very im-
portant, exactly known levels of con-
stituents are added both to distilled and
natural waters for testing. Stabilized
natural waters are not used as samples.
CH.MET. con.7a. 11. 75
4-1
-------
Interlaboratory Quality Control Stiali.-.s
VI SAMPLE FORM AND CONTAINER
A In the EPA Interlaboratory Program,
samples are prepared as concentrates
for final dilution at the testing laboratory.
Use of concentrates has several ad-
vantages. .This greatly reduces space
requirements and thereby greatly re-
duces cost of mailing. With use of sealed
glass ampules, preservation of con-
centrates is easily obtained through
steam sterilization. Chemical preser-
vatives can be used at fairly high levels
in the concentrate and can be removed as
interferences by the dilution of the con-
centrate to final sample volume. Theo-
retically, a well designed sample proper-
ly sealed in glass is stable indefinitely.
B Samples can be prepared as dilute sim-
ulated or natural water samples to be
analyzed, as is. These samples have an
advantage in that they reach the analyst
exactly as a routine sample. There is
no error in making a dilution and this
source of variance in the method is re-
moved.
Some disadvantages of such samples are:
the serious logistics problem from
storage if a large number of samples is
involved; a relatively high cost of mailing;
and a limited choice of preservation
methods because of the size of containers,
the probable use of plastic containers,
and only limited testing of water types.
VII INTERLABORATORY STUDY DESIGN
A Sample concentrates are prepared in sim-
ilar yet different concentrations at each
of several levels of analysis.
B True values are calculated from the
amount of a constituent added, and are
used to measure the accuracy of the study
method.
C Multiple constituents are prepared in a
sample by related groups of analyses.
For example, a mineral and physio..;]
.sample is analy/.rd for pll, alkalinity/
acidity, specific- conductance, total hard-
ness, sulfate, chloride, fluoride, solids,
calcium, magnesium, sodium and
potassium. A nutrient sample is analyzed
for ammonia nitrogen, nitrate nitrogen,
Kjeldahl nitrogen, orthophosphate and
total phosphorus.
D The analyst dilutes separate aliquots to
volume with distilled water and with
natural water of his choice.
E Single analyses are made for each
parameter. Replicates are of limited
value in evaluating a method for routine
analyses because of data handling
problems and failure to detect significant
differences by replication.
F Recoveries are compared for distilled
and for natural waters. Recovery from
distilled water should indicate method
bias. Differences in recoveries from
distilled and natural waters should
indicate interference by natural water
samples.
VIII DATA EVALUATION AND REPORTING
A Rejection of Outliers-Accuracy.
The T-Test is applied to all data using
the standard deviation of all data and
compared with a T value at the 99% con-
fidence interval. Accuracy is calculated
after rejection of this data.
B Precision Measures
A statistical summary is developed for
computerized treatment of data as
suggested by Larsen^3). Method pre-
cision is reported as standard deviation,
95% confidence interval, and coefficient
of variation. Method precision is reported
for each sample since it varies with
concentration.
4-2
-------
Intcrlaboratory Quality Control Studies
C Graphic Display
Methods are evaluated usinj> two-sample
charts described by Youden*^'. The
values reported by an analyst for a
sample pair became the x and y coordi-
nates for a single data point on the chart.
Data form ellipses around the true value
because of systematic error. Random
error is shown as the perpendicular
distance from the 45 slope. General
scatter indicates poor method pre-
cision. Bias is shown by movement of
the ellipse away from the true values.
Outliers are visible at some distance
from the major grouping of data. Ex-
amples shown on the following pages are:
Figure 1 - Outliers,
Figure 2 - Limited precision and
limited accuracy,
Figure 3 - Negative bias,
Figure 4 - Positive bias,
Figure 5 - Systematic error,
Figure 6 - Good precision and
accuracy.
IX INTER LABORATORY STUDY
REPORT (5' 6)
A Reports are slanted toward audience.
B All data are coded by laboratory and
analyst.
C Glossary of terms is necessary.
D Computer use increases statistical
capabilities tremendously. There is a
danger of "over-evaluation".
li Increased efficiency in reporting data is
achieved by using the computer print-out
directly in a report. See Figures 7
and 8.
REFERENCES
1 Youden, W. J. 1961. Experimental
Design and ASTM Committees. Materials
Research & Standards, 1(11):862-867.
2 Preliminary Study, Nutrient Analyses,
September 1969, Analytical Quality
Control Laboratory, FWPCA,
Cincinnati, Ohio.
3 Larson, Kenneth E., The Summarization
of Data, J. of Quality Technology, Vol.
1, No. 1, Jan. 1969.
4 Youden, W. J. 1967. Statistical Tech-
niques for Collaborative Tests. The
Association of Official Analytical
Chemists, Washington, D. C.
5 Winter, J. A. and Midgett, M. R. 1969.
FWPCA Method Study 1, Mineral and
Physical Analyses, Analytical Quality
Control Laboratory, FWPCA, Cincinnati,
Ohio.
6 Winter, J. A. and Midgett, M. R. 1970.
Method Study 2, Nutrient Analyses, Manual
Method, EMSL. USEPA, Cincinnati, Ohio.
This outline was prepared by J. A. Winter,
Chief, Method & Performance Evaluation,
EMSL, USEPA, Cincinnati, Ohio 45268.
Descriptors: Laboratory Tests, Quality
Control, Statistical Methods, Testing
Testing Procedures, Error Analysis, Graphic
Methods
4-3
-------
2.70
2.46-
2.22-
1.98
1.74
.50
FI6HE 1
AMMONIA NITR06EN, «t N/UTH, WUttBS
REJECTED BY T-TEST AT fl% WWAIIUTY.
FROM: METHOD STUDY 2, NUTRIENT ANALYSES,
MANUAL METHODS
SAMPLE 3
1.74
1.98
2.22 ' 2.46
2.70
1.9
6.0
5.0
4.0
3.0
2.0
i i i i
, HMKS
" KJEIIAHL NITtifiEN, i
6081 PRECISION WfTI
' ERRtR. FROM: NETNO
NITRIENT ANALYSES.
METIODS
-
Lkl
-a.
_«5
".
; •
2.0 3.
.
J
•
• •
' .1
i i i ^ i
UN/LITER. .
SYSTEMATIC
I STMT 2, • •
NANBAL .
* . •
»
-
-
-
SAMPLE 8
0 4.0 5.0 6.0 7.0
o
*!
f
CT
E.
O
O
V.
.Oil ' ' ' i
I F
- KJELDAHL NITROGEN,
.76
.52
.28
t04
0.?
- , - T 1 , , .
IGURE 2
HI N/LITER, POOR PRECISION. ~
FROM: METHOD STUDY 2, NUTRIENT ANALYSES,
MAN
CO
"UJ *
— J •
a.
. V*
^ * *
* •
1 1 1 1
UAL METHODS
t
t
_
•** *
• »*
*
SAMPLE 5
5
4
3
i
Fl(
i i
iURE 6
~ pH DATA, GOOD PRECISION AND GOOD ACCURACY. ~
FROM: METHOD STUDY 1, MINERAL AND PHYSICAL
AX
ALYSES
-
jrf
>fr
UJ
£ SAMPLE 3
-., . I
_
" i
•0.2
.28
.52
.76 1.0,
-------
. M-
.56
.42
.28
.14
AMMONIA NITROGEI
~ AS NEGATIVE BIAS.
ANALYS
.•
•*
• •
04 * *
•
0.
"^
i • i i
i i i i i i
FIGURE 3
1, mg N/LITER, LIMITED ACCURACY
FROM: METHOD STUDY 2, NUTRIENT ~
ES, MANUAL METHODS
.
-
-
SAMPLE 1
i i i i i i
0.0
.14
.28
.42
.56
.70
\
en
28
24
20
FIGURE 4 .
POTASSIUM ', mg K/LITER, POOR ACCURACY AS
A POSITIVE BIAS, SODIUM INTERFERENCE.
FROM: FWPCA METHOD STUDY I, MINERAL
AND PHYSICAL ANALYSES
a.
3E
«c
CO
16
SAMPLE 5
20
24
28
I.IOQOO CON*. LLK. L.171«« I9S «.!I
FIGURE 7
BOD DATA, STATISTICAL SUMMARY. PREPARED BY
COMPUTER FOR DIRECT USE IN REPORT.
FROM: METHOD RESEARCH STUDY 3, DEMAND ANALYSES
cr
o
£
-------
FIGURE 8
AMMONIA NITROGEN DATA, YOUDEN'S TWO-SAMPLE CHARTS.
PREPARED FOR REPORT BY COMPUTER PLOTTER
.6.
1A
OBTUBIUTO
SAMPLE 1
.6
.4.
.2
0
IB '
M
ill
*'' .
'. '.'
.
y|
^MPLE.l . .
J *2 »4 «o
1
10
NA1UWLIUTR
SAMFJ.E 3
-------
VOLUMETRIC ANALYSIS OF WATER QUALITY
I INTRODUCTION
A A standard solution is one whose com-
position and concentration are known
to a high degree of accuracy. In chem-
ical analyses, it is used to determine
the concentration of a particular com-
ponent in a sample.
B These chemical analyses very frequently
involve an acid-base reaction or an
oxidation-reduction reaction.
C Three important terms connected with
these two types of chemical reactions
are: mole, equivalent weight, and
normality. These terms will be de-
fined in following sections.
II ANALYTICAL CHEMICAL REACTIONS
A As stated above, in a chemical analysis
a volume of standard solution is brought
into contact with a volume of sample
in order to determine the concentration
of some particular component of the
sample.
B For a given volume of sample, it is
necessary to use a definite amount of
the standard solution--too much or too
little would give erroneous results.
C For example, one cannot simply mix
together random volumes of sodium
hydroxide and hydrochloric acid solu-
tions and expect the only two substances
produced to be sodium chloride and
water.
D Unless the concentrations of t h e two
reagents are known and the amounts
measured accurately, excess sodium
hydroxide or hydrochloric acid will
also remain at the end of the reaction.
E The reason for these limitations is
that when molecules react with
one another, they do so in definite ratio.
H
K
Unless the number of molecules of each
reactant is known, there will always be
an excess of one of the reactants re-
maining at the end of the chemical
reaction. As mentioned before, this
leads to erroneous analytical results.
Because of their size, it is not possible
to count out numbers of molecules.
However, the number of molecules in
a quantity of a chemical may be found
by determining its weight and consulting
a table which lists the weights of the
atoms making up the chemical.
For example, sodium hydroxide has the
formula NaOH. It can also be stated
that a molecule of sodium hydroxide
consists of one sodium atom, one hy-
drogen atom and one oxygen atom.
One sodium atom weighs 23 atomic
mass units (amu). An oxygen atom
weighs 16 amu; and a hydrogen atom
weighs 1 amu.
A molecule of sodium hydroxide, there-
fore, weighs 40 amu; all amu values
have been rounded off. Forty amu is
the molecular weight of sodium hydroxide.
A mole of any chemical is a number of
grams numerically equal to the molec-
ular weight of that chemical. One mole
of sodium hydroxide, therefore, con-
tains 40 grams (40 g).
Similarly, the atomic weight of a.
chlorine atom is 35 amu; that of a hy-
drogen atom is 1 amu; and the molec-
ular weight of the hydrogen chloride
molecule is 36 amu. One mole of hy-
drogen chloride weighs 36 g.
Synonyms for mole are: mol, gram
mol, gram mole, and gram molecular
weight.
Tables listing atomic weights of the
elements can be found in virtually all
PC. 17 b. 11.75
5-1
-------
Analytical Heaftions - Standard Solutions
of the texts used for high school and
first year college chemistry courses.
The remaining two terms mentioned in
1C (equivalent weight and normality)
willbe considered as they are related to
acid-base and oxidation-reduction reac-
tions .
Ill ACID-BASE REACTIONS
A Recall that an acid is a substance which
donates a hydrogen ion, or proton, (H+)
in a chemical reaction and a ba-se is a
substance which donates a hydroxide ion
(OH~) in a chemical reaction. Reactions
involving these two ions are termed
acid-base reactions.
B In the case of sodium hydroxide, the
molecular weight is 40 amu and one
mole of sodium hydroxide weighs 40 g.
One hydroxide ion is contained in the
sodium hydroxide molecule.
1 For a base, the number of grams
in a mole divided by the number of
hydroxide ions equals a quantity
called the equivalent weight. Ex-
amples are given below. All amu
values have been rounded off.
a Base - potassium hydroxide KOH
Atoms Number Wt/atom
K .1 39 amu
O 1 16
H 1 1
Total
39 amu
16
1
56 amu
One mole of KOH = 56 g
Number of hydroxide, ions = 1
Equivalent weight of KOH = 56 g
b Base - magnesium hydroxide
Mg(OH)2
Atoms Number Wt/atom Total
Mg 1 24 amu 24 amu
O 2 16 32
H 2 1 2
58 amu
One mole of Mg(OH)2 = 58 g
Number of hydroxide ions = 2
Equivalent weight of Mg(OH)2 = 29 g
For an acid, ih4_
98 amu
One mole of H2SO4 = 98 g
Number of hydrogen ions = 2
Equivalent weight of H2SO4 = 49 g
C Normality is a method of expressing
solution concentrations. If one equiv-
alent weight of a chemical is dissolved
in a solvent and the volume brought to
one liter (1), the concentration of the
solution is one normal (N).
1 The equivalent weight of KOH was
calculated to be 56 g. This amount
of the solid dissolved in water and
diluted to a liter would give a 1 N
solution.
2 The equivalent weight of HNO3 was
found to be 63 g. This quantity of
acid diluted to a liter would give a
1 N solution.
IV OXIDATION-REDUCTION REACTIONS
A The concepts of mole, equivalent weight
and normality, as described in previous
sections, apply also to oxidation -
reduction reactions.
5-2
-------
Analytical peai:iio.'i.s -_'^;a.'dar< 1 '-olu_: i_i
B
One definition of an oxidation is that it
involves an increase in the oxidation
state (charge) of an atom.
For example: FeCl2 — FeCls. In this
conversion the Fe has been oxidized
from +2 to +3.
A reduction is a decrease in the oxida-
tion state (charge) of an atom.
For example: KMnO4 -* MnO2- In this
conversion the Mn has been reduced
from +7 to +4.
The equivalent weight of an oxidizing
or reducing agent is calculated by
dividing the number of grams in a mole
of the reagent by the change in charge
involved.
For example: FeCl2 (used as a reducing
agent).
Atoms
Fe
Cl
Number
1
2
Wt/atom Total
56 anxu
70
56 amu
70
126 amu
126 g
One mole of
Change in charge » 1
Equivalent weight of FeCl2
126 g
The concentration of a liter of solution
which contains 126 g of FeCl2 is 1 N.
V PRIMARY STANDARDS
A A reagent of known purity is called a
primary standard. Primary standard
grade chemicals are available from
chemical supply houses and the National
Bureau of Standards. An accurately
measured quantity of a primary stand-
ard is used for the preparation of
standard solutions.
B Other requirements of a primary stand-
are are:
1 It must be stable at 105° C (the
temperature used for drying).
2 It should not be reactive with com-
ponents of the air, such as C>2 and
CO2-
It should have a high equivalent
weight so as to minimize any
errors in the analysis.
It should be readily available at a
reasonable cost.
VI STORAGE OF STANDARD SOLUTIONS
A Standard solutions should be prepared
using high quality distilled water.
B Care should be taken to insure the
cleanliness of the glass or plastic
bottle used for storage.
C Some solutions may decompose on ex-
posure to light and should be stored in
dark bottles.
D The stopper or cap should fit tightly.
VII CA LCULA TIONS
A The basic formula used in volume! ri.-
analysis is
(1 XN) of standard solution =
(1 XN) of sample
In a typical analysis, three of the four
quantities are known, or found, and
B
N of sample
(1 XN) of standard
1 of sample
This formula can be rearranged to give
g =1 XN Xequivalent weight
where g and equivalent weight (ew) re-
fer to the component being analyzed
and 1 and N to the standard solution.
For example: How many g of NaOH are
present in a sample if 100.0ml of 0.2N
HC1 are required for its titration?
g = 1 XN Xequivalent weight =
100.0 ml
1000.0 ml/I
= 0.8
-------
Analytical Reactions - Standard Solutions
REFERENCES 3 A.yros, li. II. yiiatiiimi ivc
Analysis, llnrpor atul Ufolhi-rs.
1 Hamilton, S. B. and Simpson. S. G. Now York. 1».
Quantitative Chemical Analysis
llth edition, The MacMillan Co.,
New York. 1958. This ourline was prepared by C. R. Feldniann,
Chemist, National Training Center, MPOD,
2 Blaedel, W.J. and Meloche, V. W. OWPO, USEPA. Cincinnati, Ohio 45268
Elementary Quantitative Analysis:
Theory and Practice. Row, Descriptors: Analytical Techniques, Chemical
Peterson and Co., N. Y. 1957. Analysis, Volumetric Analysis, Water Analysis
5-4
-------
ACIDITY, ALKALINITY, pH AND BUFFERS
I DEFINITIONS OF ACIDS AND BASES
A Arrhenius Theory of Acids and Bases
(Developed about 1887)
1 Acid: A substance which produces,
in aqueous solution, a hydrogen ion
(proton), IT*".
2 Base: A substance which produces,
in aqueous solution, a hydroxide
ion, OH".
3 The Arrhenius theory was confined
to the use of water as a solvent.
B Bronsted and Lowry Theory of Acids
and Bases (Developed about 1923)
1 Acid: A substance which donates,
in chemical reaction, a hydrogen
ion (proton).
2 Base: A substance which accepts,
in chemical reaction, a hydrogen
ion (proton).
3 Bronsted and Lowry had expanded
the acid-base concept into non-
aqueous media; i.e., the solvent
could, but did not have to be .water.
C There are other acid-base theories. The
two above are probably the most commonly
used ones when discus sing wastewater
topics however.
II DEFINITIONS OF ACIDITY,
ALKALINITY AND NEUTRALITY
A Acidity
A condition in which there is a prepon-
derance of acid materials present in
the water.
B Alkalinity
A condition in which there is a prepon-
derance of alkaline (or basic) materials
present in the water.
C Neutrality
m
It is possible to have present in
the water chemically equivalent
amounts of acids and bases. The
water would then be described as
being neutral; i.e., there is a
preponderance of neither acid nor
basic materials. The occurrence
of such a condition would be rare.
The term "neutralization" refers to
the combining of chemically equiv-
alent amounts of acids and bases.
The two products of neutralization
are a salt and water.
HC1
NaOH
NaCl
Hydro- Sodium
chloric Hydroxide
acid
Sodium
Chloride
(a salt)
The key word in the above definitions
is "preponderance." It is possible to
have a bas ion of acidity while there
are basic materials present in the
water, as well as conversely.
HOW ARE DEGREES OF ACIDITY AND
ALKALINITY EXPRESSED?
The pH scale is used to express various
degrees of acidity and alkalinity. Values
can range from Oto 14. These two ex-
tremes are of theoretical interest and would
never be encountered in a natural water or
in a wastewater. pH readings from 0 to
just under 7 indicate an acidic condition;
from just over 7 to 14, an alkaline condition.
Neutrality exists if the pH value is exactly
7. pH paper, or a pH meter, provides the
most convenient method of obtaining pH
readings. It should be noted, that under NPDES
Methodology, pH measurements are to be made
using a pH meter. Some common liquids andpH
values are listed in Table I.
CH.ALK. 3.3. 75
6-1
-------
Acidity. Alkalinity, pH and Buffers
TABLE 1. "pH Values of
Household lye
Bleach
Ammonia
Milk of magnesia
Borax
Baking soda
Sea water
Blood
Distilled water
Milk
Corn
Boric acid
Orange juice
Vinegar
Lemon juice
Battery acid
5.9 - 8.4 is the common
natural waters.
Common Liquids
13.7
12.7
11.3
10.2
9.2
8.3
8.0
7.3
7.0
6.8
6.2
5.0
4.2
2.8
2.2
0.2
pH range for most
IV HARD AND SOFT WATERS
In addition to being acidic or basic, water
can also be described as being hard or soft.
A Hard water contains large amounts of
calcium, magnesium, strontium, man-
ganese and iron ions, relative to the
amount of sodium and potassium ions
present. Hard water is objectionable
because it forms insoluble compounds
with ordinary soap.
B Soft water contains small amounts of
calcium, magnesium, strontium, man-
ganese and iron ions, relative to the
amount of sodium and potassium ions
present. Soft water does not form in-
soluble compounds with ordinary soap.
V TITRATIONS
A The conversion of pH readings into such
quantities as milligrams (mg) of acidity,
alkalinity, or hardness, is not easily
carried out. These values are more
easily obtained by means of a titration.
B In a titration, an accurately measured
volume of sample (of unknown strength)
is combined with an accurately measured
6-2
volume of standard solution (of known
strength) in the presence of a suitable
indicator.
C The strength (called normality) of the
sample is then found using the following
expression:
milliliters (ml) of sample X normality
(N) of sample * ml of standard solu-
tion X N of standard solution.
Three of the four quantities are known,
and
N of sample = ml of standard solution
X N of standard solution/ml of sample.
D In modified form, and a more specific
application of the above equation, alka -
linity is calculated in the following
manner (13th ed. Standard Methods).
mg of alkalinity as mg CaCO3/liter (1)
* ml of standard H~SO4 X N of standard
H,SOd X 50 X 1000/ml sample.
VI INDICATORS
The term "suitable indicator" was used
above. At the end of A titration, the pH of
the solution will not necessarily be 7. It
may be above or below 7. A suitable indi-
cator, therefore, is one which undergoes
its characteristic color change at the appro-
priate pH. Below are a few examples of
indicators and the pH range in which they
undergo their characteristic color changes.
In some cases, mixed indicators may be
used in order to obtain a sharper and more
definite color change. Again it should be noted
that under NPDES, pH meters are to be used
for the measurement of pH.
Indicator
Methyl Yellow
Methyl Orange
Methyl Red
Cresol Purple
Phenolphthalein
Alizarine Yellow
TABLE 2. pH
Operational
pH Range
2.8 - 4.0
3.1 - 4.4
4.4 - 6.2
7.4 - 9.0
8.0 - 9.6
10.0 - 12.0
Range of Indicators
-------
Acidity, Alkalinity , pH and Buffers
VII BUFFERS
A A buffer is a combination of substances
which, when dissolved in water, resists
a pH change in the water, as might be
caused by the addition of acid or alkali.
Listed below are a few chemicals
which, when combined in the proper
proportions, will tend to maintain the
pH in the indicated range.
Chemicals pH Range
A cetic A cid + Sodium Acetate 3.7- 5.6
Sodium Dihydrogen Phosphate +
Disodium Hydrogen Phosphate 5.8- 8.0
Boric Acid + Borax 6.8- 9.2
Borax + Sodium Hydroxide 9.2 - 11.0
TABLE 3. pH Range of Buffers
A buffer functions by supplying ions
which will react with hydrogen ions
(acid "spill"), or with hydroxide ions
(alkali "spill").
In many instances, the buffer is composed
of a weak acid and a salt of the weak acid;
e.g., acetic acid and sodium acetate.
1 In water, acetic acid ionizes or
"breaks down" into hydrogen ions
and acetate ions.
HC2H3O2 - H+ + C2H3O2"
(acetic acid) (hydrogen ion) (acetate ion)
(proton)
This ionization occurs to only a
slight extent, however, most of the
acetic acid remains in the form of
HC2H3O2; only a small amount of
hydrogen and acetate ions is formed.
2 Thus, acetic acid is said to be a
weak acid.
3 In the case of other acids, ionization
into the component ions occurs to a
large degree, and the term strong acid
is applied; e.g., hydrochloric acid.
HC1
(hydrochloric
acid)
(hydrogen ion)
(proton)
cr
(chloride
ion)
D
4 The terms "strong" and "weak" are
also applied to bases. In water
solutions, those which break down
into their component ions to a large
extent are termed "strong", and
those which do notare "weak".
Sodium hydroxide is a relatively
strong base, while ammonium
hydroxide is realtively weak.
5 Sodium acetate (a salt of acetic acid)
dissociates or "breaks down" into
sodium ions and acetate ions when
placed in water.
NaC2H302 - Na+ + C2H302-
(sodium acetate) (sodium ion) (acetate ion)
This dissociation occurs to a large
extent, and practically all of the
sodium acetate is in the form of
sodium ions and acetate ions.
It would be difficult and expensive to
prepare large quantities of buffers for
use in a treatment plant. However,
certain naturally occurring buffers may
be available (carbon dioxide is an ex-
ample). It dissolves in water to form
the species indicated below.
C02 + H20
(carbon dioxide) (Water)
« H2C03
(carbonic acid)
H2C03
H
+ HC03-
(hydrogenion)(hydrogen car-
(proton) bonate ion)
(bicarbonate)
The hydrogen ions react with hydroxide
ions which might appear in the water
as the result of an alkali "spill".
H+ + OH" = H20
(in the (hydroxide ion
buffer) "spilled")
The hydrogen carbonate ions react with
hydrogen ions which might appear in
the water as the result of an acid "spill",
H+ + HCO3- = H2CO3
(hydrogen ion) (in the
(proton) buffer)
"spilled"
This buffering action will be in effect as
long as there is carbonic acid present.
6-*
-------
Acidity. Alkalinity, pH and Buffers
Buffering action is not identical with
a process in which acid wastes are This outline was prepared by cTR
"neutralized" with alkali wastes, or Chemist. National Training Center, MPOD,
conversely. The desired effect is OWPO, EPA, Cincinnati, OH 45268
achieved in both cases, however (i.e.,
the pH is maintained within a desired
range.) Descriptors; Acids, Acidity, Alkalis,
Alkalinity, Analytical Techniques, Buffers,
Buffering Capacity, Chemical Analysis,
Hydrogen Ion Concentration, Indicators,
Neutralization, Water Analysis.
6-4
-------
ALKALINITY AND RELATIONSHIPS AMONG THE
VARIOUS TYPES OF ALKALINITIES
I PRELIMINARY
The property of water referred to as alkalinity
is usually caused by the presence of hydroxyl,
carbonate and bicarbonate ions. To a lesser
extent, borates, phosphates and silicates
contribute but are generally present in
negligible amounts.
The concentration and ratio of the OH
C°3
and HCO, ions may be measured by titrating
a sample to certain specified pH's or end
points which are detected either by use of a
pH meter or by color indicators. Phenol-
phthalein is used for visual detection of the
first end point, (approximately pH 8) which
indicates the neutralization of NaOH and con-
version of CO ~ to HCO3 . A number of
indicators (methyl orange, methyl purple,
brom cresol green,etc.) are used for detection
of the second end point (pH 3-5) which indicates
the complete conversion of HCO to HO and
CO . The final end point is determined by
the amount of CO- " and HCO originally
present in the sample. If the end points are
determined electrometrically they are taken
as the mid-point of the greatest rate of pH
change per unit volume of titrant.
II RELATIONSHIPS BETWEEN HYDROXIDE,
CARBONATE, AND BICARBONATE
ALKALINITIES
The results obtained from phenolphthalein
and total alkalinity measurements offer a
means of classification of the principal
forms of alkalinity, if certain assumptions
are made. It must first be assumed that
interferences are absent and that bicar-
bonate and hydroxide do not exist in the
same solution. According to the system
presented in Standard Methods, 13th Edition:
A Hydroxide alkalinity is present if the
phenolphthalein alkalinity is more than
one-half the total alkalinity.
B Carbonate alkalinity is present if the
phenolphthalein alkalinity is not zero
but is less than the total alkalinity.
C Bicarbonate alkalinity is present if the
phenolphthalein alkalinity is less than
one-half the total alkalinity.
Table 1. Relationships Between Phenolphthalein Alkalinity, Total Alkalinity,
Carbonate Alkalinity, Bicarbonate Alkalinity and Hydroxide Alkalinity
Lecture
Notes
Case 1
Case 2
Case 3
Case 4
Case 5
Result of
Titration
P * T
P « IT
P= O
P > \1
p < IT
OH~ Alkalinity
as CaCO3
T
0
O
2P-T
O
CO. " "Alkalinity
as CaCO_
O
O
2P
O
2(T-P)
2P
HCO " Alkalinity
as CaCO
O
O
0
T
O
T-2P
P = Phenolphthalein Alkalinity
T = Total Alkalinity
CH.ALK.2d. 11.75
7-1
-------
Alkalinity and Relationships Among the Various Types oi AIkaUnities
Table 2. Stoichiometric Volumes of Solutions of Different Normalities
Standard Solution
Normality
Equivalent Volumes, ml
H0SO.
2 4
0.0200
10.0
9.4
10.0
6.3
NaOH
0.0189
10.6
10.0
10.5
6.6
Na,,CO,.
2 .1
O.Ol'.M)
9.9
9.5
10.0
6.3
Na H CO..
0.0125
16.0
15. 1
15.9
10.0
HI CASE EXAMPLES
The relationships involved in Table 1 may
best be explained by reference to the following
graphs. These were prepared by titrating
volumes of standard solutions of sodium
hydroxide, sodium carbonate, and sodium
bicarbonate with standard sulfuric acid.
The Stoichiometric volumes of the various
solutions are summarized in Table 2 for
convenience in the interpretation of the charts.
A CASE 1 - Where phenolphthalein alkalinity
» total alkalinity
B CASE 2 - Where phenolphthalein alkalinity
= one-half the total alkalinity
I 1 I I
Ml, fl.n^o N II(S(> ADOKO
6
pll
3
25 Mt ii.»'(M-i N N.UHI
II' •
The sharp break occurs at the point where
all of the NaOH has been exactly neutral-
ized by the acid. The pH and concentration
of the end products (Na SO and HO)
determine the pH at the equivalence point
between NaOH and H?SO.; in this case,
approximately 7.0.
The titration proceeds in 2 stages wherein
all of the CO- " is converted, first to
HCO3"and finally to H2CO . The first
end point occurs at approximately pH 8,
and at exactly half the volume of acid
used for the total titration. The end point
which occurs at approximately pH 4
represents the total alkalinity and requires
exactly twice the volume of acid used for
the first end point.
If either HCO ~ or OH' ions had been
present the titration volumes for the curves
would not have been of equal magnitude.
7-2
-------
Alkalinity and Relationships Among the Various Types of Alkalinities
C CASE 3 - Where phenolphthalein alkalinity
= 0
25 Ml. 0.0125 N Nnllc'O.j v» 11,020 N
(P • (1)
j
T The reaction proceeds in one stage with
the initial pH at approximately 8. 5 and
final pH at 4. 0. In this case the phenol-
phthalein alkalinity is zero and since no
conversion of CO- " to HCOg is noted the
total alkalinity can only be due to the
HCO3" ion.
D CASE 4 - Where phenolphthalein alkalinity
is greater than one-half the total alkalinity
The volume of acid required for the first
end point (phenolphthalein alkalinity) is
due to the OH" neutralization and con-
version of the CO "" to HCO, . The
second end point represents the complete
conversion of HC03 to H2CO . Referring
to Case 2 where the volume of acid was
similar for each end point, it is apparent
that a base responding to phenolphthalein
but which is not CO,, must be present.
Since it was originally assumed that OH
and HCO," do not exist in the same
solution we must conclude that the total
alkalinity is due to OH and COg .
E CASE 5 - Where phenolphthalein alkalinity
is less than one half of the total alkalinity
MI.lMlliu N H.,SO} AOOK1I
pO Ml. 0.020 N N.-..,r<>.( I Hi Ml. <>. IW N N..WO.Q vs "."20 N ",!><'.,
11> < .'. T)
If, in the reaction NaOH +
Na CO + HO, the NaOH exists in
excess quanuty, the final sample con-
tains NaOH and Na CO . (Case 4) in
which the volume of acid required for
the phenolphthalein end point is greater
than one half the total. In this case,
however, the situation is reversed,
wherein the volume of acid required
for the HCO " end point is greater than
one half the total. Referring again to
the reaction NaOH + NaHCO -* Na2C°3
HO, if NaHCOg is in excess the end
products must consist of Na^COg and
NaHCO and OH~ must be absent. The
end points consist, therefore, of Na^CO^
-* NaHCO (phenolphthalein end point)
and NaHCOg — H CO
7-3
-------
Alkalinity and Relationships Among the Various Types of Alkaiinitics
F CASE 6 - Where phenolphthalein alkalinity
is greater than one-half total alkalinity.
and
(10 + 12.6) _
= volume of acid required
The
second end point occurs at
l1P,.t
for conversion of CO to HCO«
5 Id II -ii -i'.
MI.O.IUU N llj""1, ADIIlil)
PlOMl. 0.01HO N NiiOM t H»MI. (I.OIiiri N
(r
IV
K'oH •
Following the original assumption that
OH" and HCO ~ are not compatible, with
— ml, the volume of
e, o
which is converted to H CO . This then
becomes the same as Case \.
COMPARISON OF ANALYTICAL
METHODS FOR ALKALINITY (According
to the Analytical Reference Service
Report JAWWA Vol. 55, No. 5, 1963)
DETERMINATION METHODS
TABLE 2—SlaliaiaU Summary (cnU.)
a
G C
i
12
3
6
PH
3
0
t
condition similar to Case 4 (P > IT).
1ASE 7 - Where phenolphthalein alkalinitj
s greater than one -half total alkalinity.
ill ii i
— ^^^"*-*^^ A (P) KN1> i'OIN 1 ~
-r ^^ p( r> I-:NJ> POINT __
III II 1
I in is :MI 2:1 :in r,
Ml.fl. H2U N ll..s<>, Al'l»i:i>
UttM
r
Vnr
No.o(
Vtton
R*fMnt
-------
Alkalinity and Relationships Among the Various Types of Alkalinities
SUMMARY
Amount Added
Avg. Deviation from
amount added
Standard Deviation
50% Range
17 mg/1 (as CaCO3)
3.1 mg/1
3.4 mg/1
2.0 mg/1
Method Most Commonly Electrometric
Used
Method Preferred
Total Number of
Observations
Electrometric
41
V PROCEDURE
The actual measurement of alkalinity is a
very simple procedure requiring only titration
of sample with a standardized acid and the
proper indicator. For phenolphthalein
alkalinity the end point and indicator are well
established. For the bicarbonate titration
the final end point is a function of the HCO3
concentration. With low amounts (< 50 mg/1
as CaCOJ the pH at end point may be
approximately 5. 0. With high concentrations
(> 250 mg/1 as CaCOJ the pH at end point may
be 4.5 to 3. 8. For all-purpose work, in
which the highest degree of accuracy is not
required, an end point at pH 4.5 using methyl
purple as the indicator is recommended.
The traditional methyl orange frequently
proves to be unsatisfactory because of the
indefinite color change at the end point and
also because of the low pH (3.8 - 3.9)
required to establish the change.
VI EPA ANALYTICAL METHODS.
A The Environmental Protection Agency,
Office of Water Program?. Analytical Quality
Control Laboratory has compiled a manual
of Analytical Methods which is to be used
in Federal laboratories for the chemical
analysis of water and waste samples.
The title of this manual is "'Methods for
Chemical Analysis of Water and Wastes,
1971."
B This manual lists two parameters which
are related to the subject matter in this
outline. They are total alkalinity and
total acidity.
C For the measurement of both of these
parameters, the recommended method
is volumetric, with the equivalence point
being determined electrometrically.
The use of a color indicator (methyl
orange) is recommended only for the
automated method.
D The procedure references cited by the
EPA Methods manual are the 13th
Edition of Standards Methods and ASTM
Standards, Part 23, 1970.
REFERENCES
1 Standards Methods for the Examination of
Water and Wastewater, 13th Ed.
APHA, Inc., New York. 1971.
2 ASTM Standards, Part 23. 1973.
This outline was prepared by R. C. Kroner, for-.r.
Chief, Physical and Chemical Methods,
Analytical Quality Control Laboratory,
National Environmental Research Center,
Cincinnati, OH 45268.
Descriptors: Alkalis, Alkalinity, Analytical
Techniques, Buffers, Buffering Capacity,
Chemical Analysis, Water Analysis
7-5
-------
OXVCKN
Factors Affecting DO Concentration In Water
I The Dissolved Oxygen determination is
a very important water quality criteria for
many reasons: \
A Oxygen is an essential nutrient for all
living organisms. Dissolved oxygen is
essential for survival of aerobic
organisms and permits facultative
organisms to metabolize more effectively.
Many desirable varieties of macro or
micro organisms cannot survive at
dissolved oxygen concentrations below
certain minimum values. These values
vary with the type of organisms, stage
in their life history, activity, and other
factors.
B Dissolved oxygen levels may be used as
an indicator of pollution by oxygen
demanding wastes. Low DO concen-
trations are likely to be associated with
low quality waters.
C The presence of dissolved oxygen
prevents or minimizes the onset of
putrefactive decomposition and the
production of objectionable amounts of
malodorous sulfides, mercaptans,
amines, etc.
D Dissolved oxygen is essential for
terminal stabilization wastewaters.
High DO concentrations are normally
associated with good quality water.
E Dissolved oxygen changes with respect
to time, depth or section of a water
mass are useful to indicate the degree
of stability or mixing characteristics
of that situation.
F The BOD or other respirometric test
methods for water quality are commonly
based upon the difference between an
initial and final DO determination for a
given sample time interval and con-
dition. These measurements are
useful to indicate:
II
1 The rate of biochemical activity in
terms of oxygen demand for a
given sample and conditions
2 The degree of acceptability
(a bioassay technique) for bio-
chemical stabilization of a given
microbiota in response to food,
inhibitory agents or test conditions
3 The degree of instability of a
water mass on the basis of test
sample DO changes over an
extended interval of time.
4 Permissible load variations in
surface water or treatment units
in terms of DO depletion versus
time, concentration, or ratio of
food to organism mass, solids, en-
volume ratios.
5 Oxygenation requirements
necessary to meet the oxygen
demand in treatment units or
surface water situations.
The DO test is the only chemical test
included in all Water Quality Criteria.
Federal. State. Regional or local
FACTORS AFFECTING THE DO
CONCENTRATION IN WATER
A Physical Factors:
DO solubility in water for an
air/water system is limited to
about 9 mg DO/liter of water at
20°C. This amounts to about
0. 0009% as compared to 21% by
weight of oxygen in air.
Transfer of oxygen from air to
water is limited by the interface
area, the oxygen deficit, partial
pressure, the conditions at the
WP.NAP.25. 11.75
8-1
-------
Dissolved Oxygen Determination
interface area, mixing phenomena
and other items.
Certain factors tend to confuse
reoxygenation mechanisms of
water aeration:
a The transfer of oxygen in air
to dissolved molecular oxygen
in water has two principal
variables:
1) Area of the air-water
interface.
2) Dispersion of the oxygen-
saturated water at the
interface into the body liquid.
The first depends upon the surface
area of the air bubbles in the water
or water drops in the air; the
second depends upon the mixing
energy in the liquid. If diffusors
are placed in a line along the wall,
dead spots may develop in the core.
Different diffusor placement or
mixing energy may improve oxygen
transfer to the liquid two or threefold.
b Other variables in oxygen transfer
include:
3) Oxygen deficit in the liquid.
4) Oxygen content of the gas phase.
5) Time.
If the first four variables are
favorable, the process of water
oxygenation is rapid until the liquid
approaches saturation. Much more
energy and time are required to
increase oxygen saturation from
about 95 to 100% then to increase
oxygen saturation from 0 to about
95%. For example: t\n oxygen-
depleted sample often will pick up
significant DO during DO testing;
changes are unlikely with a sample
containing equilibrium amounts
of DO.
c The limited solubility of oxygen
in water compared to the oxygen
content of air does not require
the interchange of a large mass of
oxygen per unit volume of water
to change DO saturation. DO
increases from zero to 50%
saturation are common in passage
over a weir.
d Aeration of dirty water is practiced
for cleanup. Aeration of clean
water results in washing the air and
transferring fine particulates and
gaseous contaminants to the liquid.
One liter of air at room temperature
contains about 230 mg of oxygen.
A 5 gal carboy of water with 2 liters
of gas space above the liquid has
ample oxygen supply for equilibration
of DO after storage for 2 or 3 days
or shaking for 30 sec.
Aeration tends toward evaporative
cooling. Oxygen content becomes
higher than saturation values at
the test temperature, thus
contributing to high blanks.
Oxygen solubility varies with the
temperature of the water.
Solubility at lOOC is about two
times that at 30° C. Temperature
often contributes to DO variations
much greater than anticipated by
8-2
-------
Dissolved Oxygen Determination
solubility. A cold water often has
much more DO than the solubility
limits at laboratory temperature.
Standing during warmup commonly
results in a lose of DO due to
oxygen diffusion from the super-
saturated sample. Samples
warmer than laboratory tempera-
ture may decrease in volume due
to the contraction of liquid as
temperature is lowered. The full
bottle at higher temperature will
be partially full after shrinkage
with air entrance around the stopper
to replace the void. Oxygen in the
air may be transferred to raise the
sample DO. For example, a
volumetric flask filled to the 1000 ml
mark at 30° C will show a water
level about 1/2 inch below the mark
when the water temperature is
reduced to 2QOC. BOD dilutions
should be adjusted to 2QOC + or -
1 1/2° before filling and testing.
Water density varies with tem-
perature with maximum water
density at 4QC. Colder or warmer
waters tend to promote stratification
of water that interferes with
distribution of DO because the
higher density waters tend to seek
the lower levels.
Oxygen diffusion in a water mass is
relatively slow, hence vertical and
lateral mixing are essential to
maintain relatively uniform oxygen
concentrations in a water mass.
Increasing salt concentration
decreases oxygen solubility
slightly but has a larger effect
upon density stratification in a
water mass.
The partial pressure of the oxygen
in the gas above the water interface
controls the oxygen solubility
limits in the water. For example,
the equilibrium concentration of
oxygen in water is about 9 mg DO/I
under one atmospheric pressure of
air, about 42 mg DO/liter in
contact with pure oxygen and 0 mj.;
1X">/ liter in contact with pure
nitrogen ((a 20" C').
B Biological or Bio-Chemical Factors
1 Aquatic life requires oxygen for
respiration to meet energ>
requirements for growth, repro-
duction, and motion. The net
effect is to deplete oxygen resources
in the water at a rate controlled
by the type, activity, and mass of
living materials present, the
availability of food and favor -
ability of conditions.
2 Algae, autotrophic bacteria, plants
or other organisms capable of
photosynthesis may use light
energy to synthesize cell materials
from mineralized nutrients with
oxygen released in process.
a Photosynthesis occurs only
under the influence of aacq-a?.:,
light intensity.
b Respiration of alga is
continuous.
c The dominant effect in terms
of oxygen assets or
liabilities of alga depends upon
algal activity, numbers and
light intensity. Gross algal
productivity contributes to
significant diurnal DO
variations.
3 High rate deoxygenation commonly
accompanies assimilation of
readily available nutrients and
conversion into cell mass or
storage products. Deoxygenation
due to cell mass respiration
commonly occurs at some lower
rate dependent upon the nature of
the organisms present, the stage
of decomposition and the degree
of predation, lysis, mixing and
regrowth. Relatively high
8-3
-------
Dissolved Oxygen Determination
deoxygenation rates commonly are
associated with significant growth
or regrowth of organisms.
Micro-organisms tend to flocculate
or agglomerate to form settleable
masses particularly at limiting
nutrient levels (after available
nutrients have been assimilated or
the number of organisms are large
in proportion to available food).
a Resulting benthic deposits
continue to respire as bed
loads.
b Oxygen availability is limited
because the deposit is physically
removed from the source of
surface oxygenation and algal
activity usually is more
favorable near the surface.
Stratification is likely to limit
oxygen transfer to the bed load
vicinity.
c The bed load commonly is
oxygen deficient and decomposes
by anaerobic action.
d Anaerobic action commonly is
characterized by a dominant
hydrolytic or solubilizing action
with relatively low rate growth
of organisms.
e The net effect is to produce low
molecular weight products
from cell mass with a. corre-
spondingly large fraction of
feedback of nutrients to the
overlaying waters. These
lysis products have the effect
of a high rate or immediate
oxygen demand upon mixture
with oxygen containing waters.
f Turbulence favoring mixing of
surface waters and benthic
sediments commonly are
associated with extremely
rapid depletion of DO.
Recurrent resuspension of
thin benthic deposits may
contribute to highly erratic
DO patterns.
g Long term deposition areas
commonly act like point
sources of new pollution as
a result of the feedback of
nutrients from the deposit.
Rate of reaction may be low
for old materials but a low
percentage of a large mass of
unstable material may produce
excessive oxygen demands.
C Tremendous DO variations are likely
in a polluted water in reference to
depth, cross section or time of day.
More stabilized waters tend to show
decreased DO variations although it is
likely that natural deposits such as leaf
mold will produce differences related
to depth in stratified deep waters,
A CKNOWLEDGMENTS
This outline contains significant materials
from previous outlines by J. W. Mandia
Review and comments by C. R. Hirth and
R. L. Booth are greatly appreciated
REFERENCE
1 Methods for Chemical Analysis of
Water and Wastes. EPA-MDQARL.
1974.
This outline was prepared by F. J. Ludzack.
Chemist, National Training Center, MPOD,
OWPO, EPA. Cincinnati, OH 45268 and
revised by Charles R. Feldmann, Chemist,
National Training Center
Descriptors : Aeration, Aerobic Conditions,
Air-Water Interfaces, Anaerobic Conditions,
Benthos, Biolcg ical Oxygen Demand, Dissolved
Oxygen, Water Pollution, Water Quality
8-4
-------
DISSOLVED OXYGEN
Determination By The Winkler lodometric Titration and Azide Modification
The basic Winkler procedure (1888) has been
modified many times to improve its work-
ability in polluted waters. None of these
modifications have been completely
successful. The most useful modification
was proposed by Alsterberg and consists of
the addition of sodium azide to control
nitrite interference during the iodometric
titration. The A zide modification of the
iodometric titration is recommended as
official by the EPA- WPO Quality Control
Committee for relatively clean waters.
A Reactions
1 The determination of DO involves
a complex series of interactions
that must be quantitative to provide
a valid DO result. The number of
sequential reactions also compli-
cates interference control. The
reactions will be presented first
followed by discussion of the
functional aspects.
MnSOx + 2 KOH -*Mn(OH)9 + K SO
2 Mn(OH)
MnCKOH)2 + 2
MnO(OH>
2 KI -» MnS0
(a)
(b)
(c)
(d)
(e)
Reaction sequence
The series of reactions involves
five different operational steps in
converting dissolved oxygen in the
water into a form in which it can
be measured.
a 02-»MnO(OH)2
Mn(SO4)2
I -> Thiosulfate (thio) or
phenylarsine oxide (PAO)
titration.
b All added reagents are in excess
to improve contact possibilities
and to force the reaction toward
completion.
The first conversion, O_ ~*
MnCKOH) (reactions a, 6) is an
oxygen transfer operation where
the dissolved oxygen in the water
combines with manganous
hydroxide to form an oxygenated
manganic hydroxide.
a The manganous salt can react
with oxygen only in a highly
alkaline media.
b The manganous salt and alkali
must be added separately with
addition below the surface of
the sample to minimize reaction
with atmospheric oxygen via
air bubbles or surface contact.
Reaction with sample dissolved
oxygen is intended to occur
upon mixing of the reagents arid
sample after stoppering the
full bottle (care should be used
to allow entrained air bubbles
to rise to the surface before
adding reagents to prevent
high results due to including
entrained oxygen).
c Transfer of oxygen from the
dissolved state to the pre-
cipitate form involves a two
phase system of solution and
precipitate requiring effective
mixing for quantitative
transfer. Normally a gross
excess of reagents is used
to limit mixing requirements.
Mixing by rapid inversion 25
to 35 times will accomplish
the purpose. Less energy is
required by inversion 5 or 6
CH,O.-do.31c. 11.75
8-5
-------
Dissolved Oxygen Determination
times, allowing the solids to
settle half way and repeat the
process. Reaction is rapid;
contact is the principal
problem in the two phase
system.
d If the alkaline floe is white,
no oxygen is present.
Acidification {reactions c and d)
changes the oxygenated manganic
hydroxide to manganic sulfate
which in turn reacts with
potassium iodide to form elemental
iodine. Under acid conditions,
oxygen cannot react directly with
the excess manganous sulfate
remaining in solution.
Iodine (reaction e) may be titrated
with sodium thiosulfate standard
solution to indicate the amount of
dissolved oxygen originally
present in the sample.
a The blue color of the starch-
iodine complex commonly is
used as an indicator. This
blue color disappears when
elemental iodine has been
reacted with an equivalent
amount of thiosulfate.
b Phenylarsine oxide solutions are
more expensive to obtain but
have better keeping qualities
than thiosulfate solutions.
Occasional use, field operations
and situations where it is not
feasible to calibrate thio
solutions regularly, usually
encourage use of purchased
PAO reagents.
For practical purposes the DO
determination scheme involves the
following operations.
Fill a 300 ml bottle* under
conditions minimizing DO
changes. This means that the
sample bottle must be flushed
with test solution to displace
the air in the bottle with water
characteristic of the tested
sample*.
*DO test bottle volumes should
be checked - discard those
outside of the limits of 300 ml
+ or - 3 ml.
To the filled bottle.
1) Add MnSO4 reagent (2 ml)
2) Add KOH, KI, NaN reagent
(2 ml)
Stopper, mix by inversion,
allow to settle half way and
repeat the operation.
Highly saline test waters
commonly settle very
slowly at this stage and
may not settle to the half
way point in the time
allotted.
To the alkaline mix (settled
about half way) add 2 ml of
sulfuric acid, stopper and mix
until the precipitate dissolves.
Transfer the contents of the
bottle to a 500 ml Erlenmeyer
flask and titrate with 0. 0375
Normal Thiosulfate*. Each
ml of reagent used represents
1 mg of DO/liter of sample.
S-6
-------
Dissolved Oxygen I ^.'termination
The same thing applies for
other sample volumes when
using an appropriate titrant
normality such as:
1) For a 200 ml sample, use
0. 025 N Thio
2) For a 100 ml sample, use
0.0125 N Thio
*EPA-WPO Method
The addition of the first two DO
reagents, (MnSO4 and the KOH, KI
and NaN~ solutions) displaces an
equal quantity of the sample. This
is not the case when acid is added
because the clear liquid above the
floe does not contain dissolved
oxygen as all of it should be con-
verted to the particulate MnCXOH^.
Some error is introduced by this
displacement of sample during
dosage of the first two reagents.
The error upon addition of 2 ml of
each reagent to a 300 ml sample
is -i- X 100 or 1. 33% loss in DO.
300
This may be corrected by an
appropriate factor or by adjust-
ment of reagent normality. It is
generally considered small in
relation to other errors in sampling,
manipulation and interference,
hence this error may be recognized
but not corrected.
Reagent preparation and pro-
cedural details can be found in
reference 1.
IV The sequential reactions for the
Chemical DO determination provides
several situations where significant inter-
ference may occur in application on
polluted water, such as:
A Sampling errors may not be strictly
designated as interference but have the
same effect of changing sample DO.
Inadequate flushing of the bottle con-
tents or exposure to air may raise the
DO of low oxygen samples or lower the
DO of supersaturated samples.
B Entrained air may be trapped in a DO
bottle by:
1
Rapid filling of vigorously mixed
samples without allowing the
entrained air to escape before
closing the bottle and adding DO
reagents.
Filling a bottle with low temperature
water holding more DO than that in
equilibrium after the samples warm
to working temperature.
Aeration is likely to cool the sampi-
permitting more DO to be introduced
than can be held at the room or
incubator temperatures.
Samples warmer than working
incubator temperatures will be
only partially full at equilibriurr
temperatures.
Addition of DO reagents results in
reaction with dissolved or entrained
oxygen. Results for DO are invalu
if there is any evidence of gas
bubbles in the sample bottle.
The DO reagents respond to any
or reductant in the sample capable 01
reacting within the time allotted. HOC1
orH2O2 may raise the DO titration
while H2S, & SH may react with sample
oxygen to lower the sample titration.
The items mentioned react rapidly ana
raise or lower the DO result promptly.
Other items such as Fe or SO, may
or may not react completely within the
time allotted for reaction. Many
organic materials or complexes from
benthic deposits may have an effect upon
DO results that are difficult to predict.
They may have one effect during the
alkaline stage to release iodine from
Kl while favoring irreversible
absorption of iodine during the acid
stage. Degree of effect may increase
with reaction time. It is generally
inadvisable to use the iodometric
titration on samples containing large
amounts of organic contaminants or
8-7
-------
Dissolved Oxygen Determination
benthic residues. It would be expected
that benthic residues would tend toward
low results because of the reduced iron
and sulfur content - they commonly
favor high results due to other factors
that react more rapidly, often giving
the same effect as in uncontrolled
nitrite interference during titration.
Nitrite is present to some extent in
natural waters or partially oxidized
treatment plant samples. Nitrite is
associated with a cyclic reaction during
the acid stage of the DO determination
that may lead to erroneous high results.
1 These reactions may be repre-
sented as follows:
2HN00 + 2 HI
2
2HN02*-
4H O + NO (a)
2^2
H0 + N0 (b)
These reactions are time, mixing
and concentration dependent and
can be minimized by rapid
processing.
2 Sodium azide (NaN,) reacts with
nitrite under acid conditions to
form a combination of N, + NgO
which effectively blocks the
cyclic reaction by converting the
HNO- to noninterfering compounds
of nitrogen.
3 Sodium azide added to fresh
alkaline KI reagent is adequate to
control interference up to about
20mgof NO." N/liter of sample.
The azide is unstable and grad-
ually decomposes. If resuspended
benthic sediments are not detectable
in a sample showing a returning
blue color, it is likely that the
azide has decomposed in the
alkaline KI azide reagent.
Surfactants, color and Fe+++ may
confuse endpoint detection if present
in significant quantities.
F Polluted water commonly contains
significant interferences such as C.
It is advisable to use a membrane
protected sensor of the electronic type
for DO determinations in the presence
of these types of interference.
G The order of reagent addition and prompt
completion of the DO determination is
critical. Stable waters may give valid
DO results after extended delay of
titration during the acidified stage. For
unstable water, undue delay at any stage
of processing accentuates interference
problems.
REFERENCE
1 Methods for Chemical Analysis of Water
and Wastes. EPA-MDQARL. 1974.
ACKNOWLEDGMENTS
This outline contains significant materials
from previous outlines by J. W. Mandia.
Review and comments by C. R. Hirth and
R. L. Booth are greatly appreciated.
This outline was prepared by F. J. Ludzack,
Chemist, National Training Center. MPOD,
OWPO, USEPA, Cincinnati, Ohio 45268. and
revised by C. R. Feldmann, Chemist, National
Training Center.
Descriptors: Chemical Analysis, Dissolved
Oxygen, Oxygen, Water Analysis
8-8
-------
DISSOLVED OXYGEN
DETERMINATION BY ELECTRONIC MEASUREMENT
I INTRODUCTION
A Electronic measurement of DO is attractive
for several reasons:
1 Electronic methods are more readily
adaptable for automated analysis, con-
tinuous recording, remote sensing or
portability.
2 Application of electronic methods with
membrane protection of sensors affords
a high degree of interference control.
3 Versatility of the electronic system
permits design for a particular measure-
ment, situation or use.
4 Many more determinations per man-
hour are possible with a minor expend-
iture of time for calibration.
B Electronic methods of analysis impose
certain restrictions upon the analyst to
insure that the response does, in fact,
indicate the item sought.
1 The ease of reading the indicator tends
to produce a false sense of security.
Frequent and careful calibrations are
essential to establish workability of the
apparatus and validity of its response.
2 The use of electronic devices requires
a greater degree of competence on the
part of the analyst. Understanding of
the behavior of oxygen must be supple-
mented by an understanding of the
particular instrument and its behavior
during use.
C Definitions
1 Electrochemistry - a branch of chemistry
dealing with relationships between
electrical and chemical changes.
Electronic measurements or electro-
metric procedures - procedures using
the measurement of potential difference^
as an indicator of reactions taking
place at an electrode or plate.
Reduction - any process in which one
or more electrons are added to an atom
or an ijn, such as O + 2e -* 2O~
The oxygen has been reduced.
Oxidation - any process in which one
or more electrons are removed from
an atom or an ion, such as ZnO - 2e
Zn
+2
The zinc has been oxidized.
Oxidation - reduction reactions - in a
strictly chemical reaction, reduction
cannot occur unless an equivalent
amount of some oxidizable substance
has been oxidized. For example:
- 4e 5±
°2+4e
2H20
4H hydrogen oxidize.
-2
2O oxygen reduced
Chemical reduction of oxygen may also
be accomplished by electrons supplied
to a noble metal electrode by a battery
or other energizer.
Anode - an electrode at which oxidation
of some reactable substance occurs.
Cathode - an electrode at which
reduction of some reactable substance
occurs. For example in I. C. 3, the
reduction of oxygen occurs at the
cathode.
Electrochemical reaction - a reaction
involving simultaneous conversion of
chemical energy into electrical energy
or the reverse. These conversions are
Note: Mention of Commercial Products and Manufacturers Oo-.s
Not Imply Endorsement by the Environmental Protection
Agency.
CH.O.do.32a. 11.75
9-1
-------
Dissolved Oxygen Determination
equivalent in terms of chemical and
electrical energy and generally are
reversible.
9 Electrolyte a solution, gel, or mixture
capable of conducting electrical energy
and serving as a reacting media for
chemical changes. The electrolyte
commonly contains an appropriate
concentration of selected mobile ions
to promote the desired reactions.
10 Electrochemical cell - a device con-
sisting of an electrolyte in which 2
electrodes are immersed and connected
via an external metallic conductor.
Tifc electrodes may be in separate
compartments connected by tube con-
taining electrolyte to complete the
internal circuit.
a Galvanic (or voltaic) cell - an
electrochemical cell operated in
such a way as to produce electrical
energy from a chemical change,
such as a battery (See Figure 1).
Folarographic (electrolytic) cell -
an electrochemical coll operated in
such a way as to produce a chemical
change from electrical energy
(See Figure 2).
Cithitt
POLAROGRAPHIC CELL
Filiri 2
Auto
Citkide
GALVANIC CELL
Flf HI 1
As indicated in I. C. 10 the sign of an
electrode may change as a result of the
operating mode. The conversion by the
reactant of primary interest at a given
electrode therefore designates terminology
for that electrode and operating mode.
In electronic oxygen analyzers, the
electrode at which oxygen reduction occurs
is designated the cathode.
Each cell type has characteristic advantages
and limitations. Both may be used
effectively.
1 The galvanic cell depends upon
measurement of electrical energy
produced as a result of oxygen
t
-------
Dissolved Oxygen Determination
reduction. If the oxygen content of the
sample is negligible, the measured
current is very low and indicator driving
force is negligible, therefore response
time is longer.
2 The polarographic cell uses a standing
current to provide energy for oxygen
reduction. The indicator response
depends upon a change in the standing
current as a result of electrons
released during oxygen reduction.
Indicator response time therefore is
not dependent upon oxygen concentration.
3 Choice may depend upon availability,
habit, accessories, or the situation.
In each case it is necessary to use
care and judgment both in selection
and use for the objectives desired.
H ELECTRONIC MEASUREMENT OF DO
A Reduction of oxygen takes place in two
steps as shown in the following equations:
H2°2
+ 2e
2e
2OH
20H
Both equations require electron input to
activate reduction of oxygen. The first
reaction is more important for electronic
DO measurement because it occurs at a
potential (voltage) which is below that
required to activate reduction of most
interfering components (0.3 to 0.8 volts
relative to the saturated calomel electrode -
SCE). Interferences that may be reduced
at or below that required for oxygen
usually are present at lower concentrations
in water or may be minimized by the use
of a selective membrane or other means.
When reduction occurs, a definite quantity
of electrical energy is produced that is
proportional to the quantity of reductant
entering the reaction. Resulting current
measurements thus are more specific for
oxygen reduction.
B Most electronic measurements of oxygen
are based upon one of two techniques for
evaluating oxygen reduction in line with
equation II. A. 1. Both require activating
energy, both produce a current propor-
tional to the quantity of reacting reductant.
The techniques differ in the means of
supplying the activating potential; one
employs a source of outside energy, the
other uses spontaneous energy produced
by the electrode pair.
1 The polarographic oxygen sensor
relies upon an outside source of
potential to activate oxygen reduction.
Electron gain by oxygen changes the
reference voltage.
a Traditionally, the dropping mercury
electrode (DME) has been used for
polarographic measurements. Good
results have been obtained for DO
using the DME but the difficulty of
maintaining a constant mercury drop
rate, temperature control, and
freedom from turbulence makes it
impractical for field use.
b Solid electrodes are attractive
because greater surface area
improves sensitivity. Poisoning
of the solid surface electrodes is
a recurrent problem. The use of
selective membranes over noble
metal electrodes has minimized
but not eliminated electrode con-
tamination. Feasibility has been
improved sufficiently to make this
type popular for regular use.
2 Galvanic oxygen electrodes consist of
a decomposable anode and a noble
metal cathode in a suitable electrolyte
to produce activating energy for oxygen
reduction (an air cell or battery). Lee. .-.
is commonly used as the anode because
its decomposition potential favors
spontaneous reduction of oxygen. The
process is continuous as long as lead
and oxygen are in contact in the electrolyte
and the electrical energy released at
the cathode may be dissipated by an
outside circuit. The anode may be
conserved by limiting oxygen availability.
Interrupting the outside circuit may
produce erratic behavior for a time
after reconnection. The resulting
9-3
-------
Dissolved Oxygen Determination
nr
current produced by oxygen reduction
may be converted to oxygen concen-
tration by use of a sensitivity coefficient
obtained during calibration. Provision
of a pulsed or interrupted signal makes
it possible to amplify or control the
signal and adjust it for direct reading
in terms of oxygen concentration or to
compensate for temperature effects.
ELECTRONIC DO ANALYZER
APPLICATION FACTORS
A Polarographic or galvanic DO instruments
operate as a result of oxygen partial
pressure at the sensor surface to produce
a signal characteristic of oxygen reduced
at the cathode of some electrode pair.
This signal is conveyed to an indicating
device with or without modification for
sensitivity and temperature or other
influences depending upon the instrument
capabilities and intended use.
1 Many approaches and refinements have
been used to improve workability,
applicability, validity, stability and
control of variables. Developments
are continuing. It is possible to produce
a device capable of meeting any reasonable
situation, but situations differ.
2 Most commercial DO instruments are
designed for use under specified con-
ditions. Some are more versatile than
others. Benefits are commonly reflected
in the price. It is essential to deter-
mine the requirements of the measure-
ment situation and objectives for use.
Evaluation of a given instrument in
term* of sensitivity, response time,
portability, stability, service
characteristics, degree of automation,
and consistency are used for judgment
oa a cost/benefit basis to select the
most acceptable unit.
B Variables Affecting Electronic DO
Measurement
1 Temperature affects the solubility of
oxygen, the magnitude of the resulting
signal and the permeability of the
protective membrane. A curve of
oxygen solubility in water versus
increasing temperature may be concave
downward while a similar curve of
sensor response versus temperature
is concave upward. Increasing
temperature decreases oxygen solubility
and increases probe sensitivity and
membrane permeability. Thermistor
actuated compensation of probe
response based upon a linear relation-
ship or average of oxygen solubility
and electrode sensitivity is not precisely
correct as the maximum spread in
curvature occurs at about 17° C with
lower deviations from linearity above
or below that temperature. If the
instrument is calibrated at a temperature
within + or - 5° C of working temperature,
the compensated readout is likely to be
within 2% of the real value. Depending
upon probe geometry, the laboratory
sensor may require 4 to 6% correction
of signal per ° C change in liquid
temperature.
2 Increasing pressure tends to increase
electrode response by compression
and contact effects upon the electrolyte,
dissolved gases and electrode surfaces.
As long as entrained gases are not
contained in the electrolyte or under
the membrane, these effects are
negligible.
Inclusion of entrained gases results in
erratic response that increases with
depth of immersion.
3 Electrode sensitivity changes occur as
a result of the nature and concentration
of contaminants at the electrode sur-
faces and possible physical chemical or
electronic side reactions produced.
These may take the form of a physical
barrier, internal short, high residual
current, or chemical changes in the
metal surface. The membrane is
intended to allow dissolved gas pene-
tration but to exclude passage of ions
or particulates. Apparently some ions
or materials producing extraneous ions
within the electrode vicinity are able
to pass in limited amounts which
9-4
-------
Dissolved Oxygen Determination
become significant in time. Dissolved
gases include 1) oxygen, 2) nitrogen,
3) carbon dioxide, 4) hydrogen sulfide,
and certain others. Item 4 is likely to
be a major problem. Item 3 may pro-
duce deposits in alkaline media; most
electrolytes are alkaline or tend to
become so in line with reaction n.A. 1.
The usable life of the sensor varies
with the type of electrode system,
surface area, amount of electrolyte
and type, membrane characteristics,
nature of the samples to which the
system is exposed and the length of
exposure. For example, galvanic
electrodes used in activated sludge
units showed that the time between
cleanup was 4 to 6 months for electrodes
used for intermittent daily checks of
effluent DO; continuous use in the mixed
liquor required electrode cleanup in 2
to 4 weeks. Each electrometric cell
configuration and operating mode has
its own response characteristics.
Some are more stable than others.
It is necessary to check calibration
frequency required under conditions
of use as none of them will maintain
uniform response indefinitely. Cali-
bration before and after daily use is
advisable.
4 Electrolytes may consist of solutions
or gels of ionizable materials such as
acids, alkalies or salts. Bicarbonates,
KC1 and KI are frequently used. The
electrolyte is the transfer and reaction
media, hence, it necessarily becomes
contaminated before damage to the
electrode surface may occur. Electro-
lyte concentration, nature, amount and
quality affect response time, sensitivity,
stability, and specificity of the sensor
system. Generally a small quantity of
electrolyte gives a shorter response
time and higher sensitivity but also may
be affected to a greater extent by a
given quantity of contaminating sub-
stances.
5 Membranes may consist of teflon,
polyethylene, rubber, and certain
other polymeric films. Thickness
may vary from 0.5 to 3 mils (inches X
1/1000). A thinner membrane will
decrease response time and increase
sensitivity but is less selective and
may be ruptured more easily. The
choice of material and its uniformity
affects response time, selectivity and
durability. The area of the membrane
and its permeability are directly
related to the quantity of transported
materials that may produce a signal.
The permeability of the membrane
material is related to temperature and
to residues accummulated on the
membrane surface or interior. A
cloudy membrane usually indicates
deposition and more or less loss of
signal.
6 Test media characteristics control the
interval of usable life between cleaning
and rejuvenation for any type of
electrode. More frequent cleanup is
essential in low quality waters than for
high quality waters. Reduced sulfur
compounds are among the more
troublesome contaminants. Salinity
affects the partial pressure of oxygen
at any given temperature. This effect
is small compared to most other
variables but is significant if salinity
changes by more than 500 mg/1.
7 Agitation of the sample in the vicinity
of the electrode is important because
DO is reduced at the cathode. Under
quiescent conditions a gradient in
dissolved oxygen content would be
established on the sample side of the
membrane as well as on the electrode
side, resulting in atypical response.
The sample should be agitated
sufficiently to deliver a representative
portion of the main body of the liquid
to the outer face of the membrane.
It is commonly observed that no
agitation will result in a very low or
negigible response after a short period
of time. Increasing agitation will cause
the response to rise gradually until
some minimum liquid velocity is reached
that will not cause a further increase
in response with increased mixing
energy. It is important to check
mixing velocity to reach a stable high
signal that is independent of a reasonable
change in sample mixing. Excessive
9-5
-------
Dissolved Oxygen Determination
may create a vortex and expose
the sensing surface to air rather than
sample liquid.. This should be avoided.
A linear liquid velocity of about 1 ft/sec
at the sensing surface is usually
adequate.
8 DO sensor response represents a
potential or current signal in the
mill!-volt or milli-amp range in a
high resistance system. A high quality
electronic instrument is essential to
maintain a usable signal-to-noise ratio.
Some of the more common difficulties
include:
a Variable line voltage or low batteries
in amplifier power circuits.
b Substandard or unsteady amplifier
or resistor components.
c Undependsble contacts or junctions
in the sensor, connecting cables, or
Instrument control circuits.
d Inadequately shielded electronic
components.
e Excessive exposure to moisture,
fumes or chemicals in the wrong
places lead to stray currents,
internal shorts or other malfunction.
C Desirable features in a Portable DO
Analyzer
1 The unit should include steady state
performance electronic and indicating
components In a convenient but sturdy
package that is small enough to carry.
2 There should be provisions for addition
of special accessories such as bottle
or field sensors, agitators, recorders,
line extensions, if needed for specific
requirements. Such additions should
be readily attachable and detachable
and my!1***1" good working characteristics.
3 The instrument should include a
sensitivity adjustment which upon
calibration will provide for direct
reading in terms of mg of DO/liter.
4 Temperature compensation and temp-
erature readout should be incorporated.
5 Plug in contacts should be positive,
sturdy, readily cleanable and situated
to minimize contamination. Water
seals should be provided where
necessary.
6 The sensor should be suitably designed
for the purpose intended in terms of
sensitivity, response, stability, and
protection during use. It should be
easy to clean, and reassemble for use
with a minimum loss of service time.
7 Switches, connecting plugs, and con-
tacts preferably should be located on
or in the instrument box rather than
at the "wet" end of the line near the
sensor. Connecting cables should be
multiple strand to minimize separate
lines. Calibration controls should be
convenient but designed so that it is
not likely that they will be inadvertently
shifted during use.
8 Agitator accessories for bottle use
impose special problems because they
should be small, self contained, and
readily detachable but sturdy enough
to give positive agitation and electrical
continuity in a wet zone.
9 Major load batteries should be
rechargeable or readily replaceable.
Line operation should be feasible
wherever possible.
10 Service and replacement parts avail-
ability are a primary consideration.
Drawings, parts identification and
trouble shooting memos should be
incorporated with applicable operating
instructions in the instrument manual
in an informative organized form.
D Sensor and Instrument Calibration
The instrument box is likely to have some
form of check to verify electronics,
battery or other power supply conditions
for use. The sensor commonly is not
included in this check. A known reference
9-6
-------
Dissolved Oxygen Determination
sample used with the instrument in an
operating mode is the best available
method to compensate for sensor variables
under use conditions. It is advisable to
calibrate before and after daily use under
test conditions. Severe conditions.changes
in conditions, or possible damage call for
calibrations during the use period. The
readout scale is likely to be labeled -
calibration is the basis for this label.
The following procedure is recommended:
1 Turn the instrument on and allow it to
reach a stable condition. Perform the
recommended instrument check as
outlined in the operating manual.
2 The instrument check usually includes
an electronic zero correction. Check
each instrument against the readout
scale with the sensor immersed in an
agitated solution of sodium sulfite
containing sufficient cobalt chloride to
catalyze the reaction of sulfite and
oxygen. The indicator should stabilize
on the zero reading. If it does not, it
may be the result of residual or stray
currents, internal shorting in the
electrode, or membrane rupture.
Minor adjustments may be made using
the indicator rather than the electronic
controls. Serious imbalance requires
electrode reconditioning if the electronic
check is O.K. Sulfite must be carefully
rinsed from the sensor until the readout
stabilizes to prevent carry over to the
next sample.
3 Fill two DO bottles with replicate
samples of clarified water similar to
that to be tested. This water should
not contain significant test interferences.
4 Determine the DO in one by the azide
modification of the iodometric titration.
5 Insert a magnetic stirrer in the other
bottle or use a probe agitator. Start
agitation after insertion of the sensor
assembly and note the point of
stabilization.
Adjust the instrument calibration
control if necessary to compare
with the titrated DO.
If sensitivity adjustment is not
possible, note the instrument
stabilization point and designate
it as ua. A sensitivity coefficient,
is equal to
where DO is the
titrated value for the sample on
which ua was obtained. An unknown
IIS
DO then becomes DO « — . This
factor is applicable as long as the
sensitivity does not change.
Objectives of the test program and the
type of instrument influence calibration
requirements. Precise work may
require calibration at 3 points in the
DO range of interest instead of at zero
and high range DO. One calibration
point frequently may be adequate.
Calibration of a DO sensor in air is a
quick test for possible changes in
sensor response. The difference in
oxygen content of air and of water is
too large for air calibration to be
satisfactory for precise calibration
for use in water.
IV This section reviews characteristics of
several sample laboratory instruments.
Mention of a soecific instrument does not
imply EPA endorsement or recommendati.
No attempt has been made to include all the
available instruments, those described are
used to indicate the approach used at one
stage of development which may or may not
represent the current available model.
A The electrode described by Carrit and
Kanwisher (1) is illustrated in Figure 3.
This electrode was an early example of
those using a membrane. The anode was
a silver - silver oxide reference cell with
a platinum disc cathode (1-3 cm diameter).
The salt bridge consisted of N/2 KC1 and
9-7
-------
Dissolved Oxygen Determination
KOH. The polyethylene membrane was
held in place by a retaining ring. An
applied current was used in a polarographic
mode. Temperature effects were relatively
large. Thermistor correction was studied
but not integrated with early models.
Figure 3
The Beckman oxygen electrode is another
illustration of a polarographic DO sensor
(Figure 4). It consists of a gold cathode,
a silver anode, an electrolytic gel con-
taining KC1, covered by a teflon membrane.
The instrument has a temperature readout
and compensating thermistor, a source
polarizing current, amplifier with signal
adjustment and a readout DO scale with
recorder contacts.
SENSOR ELECTRONICS
IX.-""1"
E
, 2
1
\N^
y
M
s
i3
—t t —
k«'
! M *f I..... r,.k.
A|lt«t*r Art.. I.
Afil.l.r ling
-------
Dissolved Oxygen Determination
The Galvanic Cell Oxygen Analyzer (7,8)
employs an indicator for proportional DO
signal but does not include thermistor
compensation or signal adjustment.
Temperature readout is provided. The
sensor includes a lead anode ring, and
a silver cathode with KOH electrolyte
(4 molar) covered by a membrane film
(Figure 7).
•rociiion Galvanic Call Oxyfon »robe
Connecter Leodi
Thermlttor Cable
Retainer
. Tapered Section
to Pit iOD •ottlei
Plaitlc Membrane
Retainer Ring
lead Anode Ring
Silver Cathode •
Polyethylene Membrane
The Weston and Stack Model 300 DO
Analyzer (8) has a galvanic type sensor
with a pulsed current amplifier adjustment
to provide for signal and temperature
compensation. DO and temperature
readout is provided. The main power
supply is a rechargeable battery. The
sensor (Figure 8) consists of a lead anode
coil recessed in the electrolyte cavity
(50% KI) with a platinum cathode in the tip.
The sensor is covered with a teflon mem-
brane. Membrane retention by rubber
band or by a plastic retention ring may be
used for the bottle agitator or depth
sampler respectively. The thermistor
and agitator are mounted in a sleeve that
also provides protection for the membrane.
G The EIL Model 15 A sensor is illustrated
in Figure 9. This is a galvanic cell with
thermistor activated temperature com-
psnsation and readout. Signal adjustment
is provided. The illustration shows an
expanded scheme of the electrode which
when assembled compresses into a sensor
approximating 5/8 inch diameter and 4 inci
length exclusive of the enlargement at the
upper end. The anode consists of com-
pressed lead shot in a replaceable capsule
(later models used fine lead wire coils),
a perforated silver cathode sleeve around
the lead is covered by a membrane film.
The electrolyte is saturated potassium
bicarbonate. The large area of lead
surface, silver and membrane provides
a current response of 200 to 300 micro-
amperes in oxygen saturated water at
2DOC for periods of up to 100 days use (8).
The larger electrode displacement favors
a scheme described by Eden (9) for
successive DO readings for BOD purposes.
V Table 1 summarizes major characteristic
of the sample DO analyzers described in
Section IV. It must be noted that an ir.._-.; .
analyst may adapt any one of these for .-;••. .<
purposes on a do-it-yourself program. The
sample instruments are mainly designer L>-
laboratory or portable field use. Those
designed for field monitoring purposes may
include similar designs or alternate designs
generally employing larger anode, cathoce,
and electrolyte capacity to approach better
response stability with some sacrifice in
response time and sensitivity. The electrciv
controls, recording, telemetering, and
accessory apparatus generally are semi-
permanent installations of a complex nature.
A CKNOW LEDGMENTS:
This outline contains certain materials from
previous outlines by D. G. Ballinger,
N. C. Malof, andJ.W. Mandia. Additional
information was provided by C.R. Hirth,
C.N. Shadix, D.F. Krawezyk, J. V/oo.h-i.
and other t;.
-------
Dissolved Oxygen Determination
WESTON & STACK
DO PROBE
CORD
CORD RESTRAINER
SERVICE CAP
PROBE SERVICE CAP
ELECTROLYTE FILL SCREW
PROBE BODY
PLATINUM CATHODE
CONNECTOR PINS
PIN HOUSING
LEAD ANODE
REMOVABLE PROBE SHIELD
AND THERMISTOR HOUSING
Figure 8
9-10
-------
Made! A1SA ELECTRODE COMPONENT PARTS
Cable Sealing
Nut
A 15017
Cable
Connection
Cover
A15016A
'O' Ring
R524
'O' Ring
R389
R385
Membrane— Securing
'O' Ring
R317
Anode
Contact
A 4 5 0140
(With Sleeve S24)
Anode
Contact
Holder
A15015A
Lead Anode
Co*f
(A150
ilete
24A)
„ •
1
S
'O' Ring
R612
Membrane Securing
'O' Ring
R317
Silver Cathode
A15013A
Filler Screw
Z47I
. — ^ [III!
1 111 II ILL. —
— ' lllllll Illlllll'1 —
O' Ring
R622
Note: Red wire of cable connects to Anode Contact Holder
Black wire of cable connects to Anode Contact
Membrane not shown E. I. L. part number T22
'O' Ring
R622
End Cap
A15011A
a
en
co
O
a
m
n>
O
a>
CD
I
O
a
FHI.K
-------
TAB1.K \
CHARACTERISTICS OF VAHIOHS LABORATUHY UO INSTRUMKNTS
DO Temp.
Carrit &
Kanwisher
Beckman
Yellow Springs
51
Yellow Springs
54
Precision
Sci
Weston tk
Stack
300
Delta
T5
Delta
85 _
Anode
silver -
silver ox.
ring
Aq*
ring
Ag
coil
"
Pb
ring
Pb
coil
Pb
Lead
Lead
Cathode
Pt
disc
Au
disc
Au
ring
silver
disc
Pt
disc
Ag
Silver
disc
Silver
disc
Elec
KC1
KOH
N/2
KC1
gel
KC1
so In
sat.
KOH
4N
ja
40%
KHCO,
KOH
IN
KOH
IN
Type
pol*
pol
pol
galv*
galv
galv
galv
galv
Membr
polyeth
teflon
teflon
*polyeth
teflon
teflon
teflon
teflon
Sig.
Adj.
no
yes
yes
yes
no
yes
yes
yes
yes
Comp.
Temp. Rdg.
no
ves
yes
no*
yes
yes
yes
no
yes
yes
ves_
yes
yes
no
yes
Accessories for
which designed
Recording temp.
& signal adj. self
assembled
recording
field and bottle
probe
recording field
bottle & agitator
probes
agit. probe
depth sampler
recording
field bottle &
agitator probe
field bottle &
agitator probe
*Pdl - Polarographic (or amperometric)
**G*lv - Galvanic (or voltametric)
REFERENCES
1 Carrit, D.E. and Kanwisher, J.W.
Anal. Chem. 31:5. 1959.
2 Beckman Instrument Company. Bulletin
7015, A Dissolved Oxygen Primer,
Fuller-ton. CA. 1962.
3 Instructions for the YSI Model 51 Oxygen
Meter, Yellow Springs Instrument
Company. Yellow Springs. OH 45387.
4 Instructions for the YSI Model 54 Oxygen
Meter, Yellow Springs Instrument
Company. Yellow Springs. OH 45387.
5 Technical Bulletin TS-68850 Precision
Scientific Company, Chicago, IL 60647.
6 Mancy, K.H.. Okun, D.A. and Reilley,
C.N. J. Electroanal. Chem. 4:65.
1962.
7 Instruction Bulletin, Weston and Stack
Model 300 Oxygen Analyzer. Roy F.
Weston. West Chester, PA 19380.
8 Briggs. R. and Viney, M. Design and
Performance of Temperature Com-
pensated Electrodes for Oxygen
Measurements. Jour, of Sci.
Instruments 41:78-83. 1964.
9-12
-------
Dissolved Oxygen Determination
9 Eden, R. E. BOD Determination Using 11 Methods for Chemical Analysis of Water
a Dissolved Oxygen Meter. Water and Wastes, EPA-MDQARL. 1974.
Pollution Control, pp. 537-539. 1967.
10. Skoog, D.A. and West, D. M. Fundamentals
of Analytical Chemistry. Holt,
Rlnehart & Winston, Inc. 1966.
This outline was prepared by F. J. Ludzack
Chemist, National Training Center, MPOD, OWPO,
USEPA, Cincinnati, OH 45268 and Nate
Malof, Chemist, EPA.OWPO, National
Field Investigations Center, Cincinnati, OH
Descriptors : Chemical Analysis, Dissolved
Oxygen, Dissolved Oxygen Analyzers,
Instrumentation, On-Site Tests, Water Analysis,
Analysis, Wastewater, Oxygen
9-13
-------
BIOCHEMICAL OXYGEN DEMAND TEST PROCEDURES
I OXYGEN DEMAND OF POLLUTED
WATERS
Established practice includes common use
of the BOD test as a tool for estimation of
the bio-oxidizable fraction of surface waters
or wastewaters discharged to them. Any
index including a quantity per unit time such
as th« BOD is a rate expression. The
ultimate demand is more important than any
one point on the progression. The results
of a bottle test with minimum seeding and
quiescent storage are not likely to be as
high as those on the same influent in a mix-
ing situation and abundant seed of secondary
treatment or receiving waters. The BOD
is "a" fraction of total oxygen requirements.
A The particular technique used for BOD
commonly is specified by State agencies
and/or supervisors. They are required
to interpret the results as obtained by
laboratory testing. It is essential that
the tester and the interpreters have a
common understanding of what was done
and how. It is highly advisable to main-
tain a given routine until all concerned
agree upon a change.
1 Each particular routine has many un-
deflnable factors. The particular
routine is not as important as the con-
sistency and capability with which the
result was obtained.
2 This outline and Standard Methods
discusses several valid approaches for
obtaining BOD results. Selection of
"method" is not intended in this outline
or in the EPA Methods Manual*2'.
B The common 5-day incubation period for
BOD testing is a result of tradition and
cost. Initial lags are likely to be over
and some unknown fraction of the total
oxidizable mass has been satisfied after
5 days.
C A series of observations over a period of
time makes it possible to estimate the
total oxidizable mass and the fraction
oxidized or remaining to be oxidized at
any given time. The problem is to define
the shape of the deoxygenation pattern and
its limits. A fair estimate of the shape of
the deoxygenation pattern is available by
observations at 1, 2, or 3 days, 7 days
and 14 days.. Increased observations are
desirable for more valid estimates of
curve shape, rate of oxidation and total
oxidizable mass or ultimate BOD.
D Increasing impoundment of surface waters
and concurrent increases in complexity
and stability of wastewater components
emphasize the importance of long-term
observation of BOD. The 5-day observation
includes most of the readily oxidizable
materials but a very small fraction of the
stable components that are the main factors
in impoundment behavior.
II DIRECT METHOD
A With relatively clean surface waters, the
BOD may be determined by incubation of
the undiluted sample for the prescribed
time interval. This method is applicable
only to those waters whose BOD is less
than 8 mg/1 and assumes the sample
contains suitable organisms and accessory
nutrients for optimum biological
stabilization.
B Treated effluents, polluted surface waters,
household and industrial wastewaters
commonly require dilution to provide the
excess oxygen required for the oxygen
demand determination. General guidelines
for dilution requirements for a given BOD
range in terms of the percent of sample in
BOD dilution water are:
For a 5-day BOD of
5-20 mg/1, use 25 to 100% sample
For a BOD of
20-100 mg/1, use 5 to 25% sample
For a BOD of
100-500 mg/ 1, use 1 to 5% sample
For a BOD of
500-5000 mg/1, use 0. 1 to 1. Q°/c sample
III PROCEDURES
A Cylinder Dilution Technique
CH.O.bod.57g. 11. 75
10-1
-------
Biochemical Oxygen Demand Test Procedures
1 Using an assumed or estimated BOU
value as a guide, calculate the factors
for a range of dilutions to cover the
desired depletions. Those dilutions
ranging from a depletion of 2 mg/l
and a residual of 1 mg/1 are most
reliable. At least three dilutions in
duplicate should be used for an
unknown sample.
2 Into a one-liter graduate cylinder (or
larger container if necessary) measure
accurately the required amount of mixed
sample to give one liter of diluted waste.
Fill to the one liter mark with dilution
water. Carefully mix. The initial DO by
calculation includes IDOD (VIII) a
determined initial does not. Both are
essential to estimate significance of
IDOD. Entrapment of air bubbles during
manipulation must be avoided.
3 Siphon the mixture from the cylinder into
three 300 ml glass stoppered bottles,
filling the bottles to overflowing.
4 Determine the DO concentration on one
of the bottles by the appropriate
Winkler modification and record as
"Initial DO".
5 Incubate the two remaining bottles at
20°C in complete darkness. The
incubated bottles should be water-sealed
by immersion in a tray or by using a
special water-seal bottle.
6 After 5 days of incubation, or other
desired interval, determine the DO on
the bottles. Average the DO concentration
of the duplicates and report as "Final DO".
B Direct Dilution Technique
1 It may be more convenient to make the
dilution directly in sample bottles of
known capacity. A measured volume of
sample may be added (as indicated in
A-l) above', and the bottle filled with dilution
water to make the desired sample
concentration for incubation. In this
case, the sample must be precisely
measured, the bottle carefully filled,
but not overfilled, and the bottle volumes
comparable and known. Precision is
likely to be poorer than for cylinder
dilution.
2 Continue the procedure as in A-4, 5,
and 6 above.
C Seeded Cylinder Dilution. Technique
1 Many wastewaters may be panially or
completely sterile as a result of
chlorination, effects of other toxic
chemicals, he&t, unfavorable pH or
other factors detrimental to biological
activity. Validity of the BOD result
depends upon the presence of organisms
capable of prompt and effective bio-
degradation and favorable conditions
during the particular test. Correction
of the cause resulting in sterilization
must be corrected by adjustment,
dilution, etc. , prior to reinoculation to
achieve meaningful BOD data. Receiving
water, biologically treated effluents, and
soil suspensions are a good source of
organisms likely to be adapted for
stabilization of wastewaters. Untreated
wastewaters provide numerous organisms
but are likely to contain nutrients
contributing to excessive seed corrections
and may require appreciable time for
adaptation before test waste oxidation
becomes significant.
2 The amount of added inoculant must be
determined by trial. The concentration
added should initiate biochemical
activity promptly but should not exert
enough oxygen demand to unduly reduce
the oxygen available for sample
requirements.
3 Estimate the sample concentration
desired in accordance with A-l and
C-2 above and add the sample aliquot to the
dilution cylinder.
4 Add approximately half of the required
amount of dilution water to the sample
and mix. This is necessary to assure
that the concentrated waste does not
exert a toxic effect on the seed organisms.
5 Measure a suitable aliquot of seed into
the bottle or cylinder and fill with
dilution water. Mix the combined sample,
seed and dilution water without excessive
air entrainment.
6 Continue as in III-A steps 4, 5, and 6 above.
IV INTERPRETATION OF RESULTS
Standard Methods includes a calculation
section that is valid and concise. Preceding
it are details of reagent preparation and
10-2
-------
Biochemical Oxygen Demand Test Procedures
procedures for the test. These will not be
reprinted here. This section considers
certain items that may cause concern about
the validity of results unless they are care-
fully considered and controlled.
A The initial DO of the BOD test obviously
should be high. The method of attaining
a high DO can trap the analyst.
1 Aeration of dilution water is the most
commonly considered treatment.
This technique does produce a high DO
but it is a treacherous ally.
a Dirty air passing through clean
dilution water can produce clean
air and dirty water. This is a
simple air-washing operation.
Filtering the entering air stream
may remove brickbats and 2 x 4's,
but filters tend to pass organic
gases, fine aerosols, and partic-
ulates.
b A stream of air passing through
water tends to cool the water by
evaporation 1 to 3 C below ambient
temperature. The cooled liquid
picks up more DO than it can hold
at ambient temperature. The
physical loss of oxygen may produce
an erroneously high depletion value
for a determined initial DO, or a
low depletion on a calculated
initial DO. Erroneous blanks are
a particular concern. The dilution
water temperature /DO shift is
critical.
2 Raising DO by allowing the sample to
equilibrate in a cotton-plugged bottle
for 2 or 3 days permits oxygenation
with minimum air volume contact.
3 Shaking a partially filled bottle for a
few seconds also oxygenates with min-
imum opportunity of gas washing con-
tamination, supersaturation, or
temperature changes.
B Seeding always is a precarious procedure
but a very necessary one at times. Often
the application of seed corrections is a
if you do,
if you don't"
situation. Hopefully, seed corrections
are small because each individual
biological situation is a "universe" of
its own.
1 Unstable seeding materials such as
fresh wastewater have "seed11 organisms
characteristic of their origin and
history. Saprophytes resulting in
surface water stabilization may be a
small fraction of the population. Re-
actable oxygen-demanding components
produce excessive demands upon test
oxygen resources.
2 A seed containing viable organisms at
a lower energy state because of limited
nutritional availability theoretically is
the best available seed source. An
organism population grown under
similar conditions should be most
effective for initiating biochemical
activity as soon as the nutrient situation
favors more activity. The population
should not be stored too long because
organism redistribution and die-out
become limiting. This type of seed
would most likely be found in a surface
water or a treatment plant effluent
with a history of receiving the particular
material under consideration.
3 Seed sources and amounts can only be
evaluated by trial. Different seed
sources and locations require checkout
to determine the best available material
from a standpoint of rapid initiation of
activity, low correction, and predictable
high oxygen depletion under test.
C Chlorination and BOD results fundamentally
are incompatible. Chlorination objectives
include disinfection as the number one
goal. Chlorine is notoriously non-specific
in organism effects. Chlorine acts like
an oxidant in the DO determination. Test
organisms are less suitable for activity
than they were before chlorination.
Nutrients may be less available after
chlorination. Certainly the conditions are
less suitable for biological response after
chlorination. Dechlorination is feasible
with respect to the oxidizing power of
10-3
-------
Biochemical Oxygen Demand Test Procedures
chlorine, but many organic chlorine com-
pounds that do not show strong oxidizing
action still have toxic effects on biologic
response.
Numbers are obtainable after dechlorination
and reseeding. The meaning of these
numbers is obscure. At least two states
(New York and New Jerssy) specify BOD's
before chlorination only.
V PRECISION OF THE BOD TEST
A The DO test precision often has been used
to suggest precision of the BOD result.
DO precision is a relatively minor and
controllable factor contributing to BOD
results. Other factors such as organism
suitability, members, adaptation and
conditional variables are much more
difficult to control or to evaluate.
B The Analytical Reference Service report
on Water Oxygen Demand, July, 1960
(Sample type VII) included the results of
seeded samples of glucose-glutamic acid
BOD results from 34 agencies on 2, 3, 5
and 7 day incubations.
The relative geometric standard deviation
(average) was 19% on 2% sample and 24%
on 1% sample concentrations. Rate
coefficients ranged from 0. 10 to 0. 27 with
a median of 0. 16 from 21 different
laboratories that participated in rate studies.
VI ALTERNATE BOD TECHNIQUES
Reaeration methods are becoming increasingly
popular in order to approach more nearly the
actual waste concentration in the receiving
water. It is common to obtain "sliding" BOD
results related to the concentration of waste
in a series of dilutions of the same sample.
This may result from greater possibilities
for toxic effects at higher concentrations, or
to a different selection of organisms and change
in oxidation characteristics at low concentrations
of sample. The most reliable estimate of stream
behavior is likely to be from that dilution closest
to the wastewater dilution in the receiving water.
Reaeration can be accomplished by the
usual series techniques by dumping all of
the remaining sealed bottles into a common
container when the residual DO reacfies about
1. 0 mg/1. After reoxygenation, the remaining
bottles are refilled and a new initial DO
determined. Subsequent dissolved oxygen
depletions are added incrementally as a
summation of the total oxygen depletion
from the start of the test. If necessary,
the reaeration technique may be performed
several times but at a sacrifice of double
DO determinations for each day on which
reaeration occurs.
B Special methods of reaeration have evolved
to minimize the extra manipulation for
reaeration of individual sample dilutions.
1 Elm ore Method
A relatively large volume of the sample
is stored in an unsealed bottle. Small
bottles are withdrawn in sets of 5 or
more, sealed, incubated, and the DO
determined at appropriate intervals.
When the DO concentration in the smaller
bottles reaches 1. 0 mg'l, a new set is
withdrawn from the large unsealed
bottle, after reoxygenation if necessary.
2 Orford Method
The deoxygenation is carried out in a
large sealed jug from which samples
for DO are withdrawn at appropriate
intervals. To maintain the waste level
and a sufficient DO in the jug, additional
waste is added from a second open
container. See diagram.
C Excess oxygen may be provided by
oxygenation with commercial oxygen
instead of with air to increase the initial
oxygen content for incubation while limiting
the number of dilutions or reaeration steps
When oxygen is used in place of air the
oxygen saturation in water at 20°C is about
40 mg/1 instead of 9 mg/1. Limited results
are available hence the analyst must verify
his technique. The DO tends to decrease
as soon as the bottle is opened hence, about
35 mg/1 of oxygen content is the top of the
practical working concentration. There has
been no evidence that the biota is inhibited
by the higher oxygen content with respect
to BOD progression.
D Reaeration or Oxygenation Advantages and
Limitations.
1 Reaeration expands the range of BOD
results obtainable directly at field
concentrations, but is not advisable for
applications when the sample BOD
exceeds 50 mg' 1.
10-4
-------
Biochemical O.xygen Demand Test ['rot edures
2 Dilution water problems are eliminated,
to the extent that the stream sample may
be tested without dilution.
3 Incubator storage space becomes a real
problem i'or multiple sample routine.
VII Dissolved oxygen electrodes, polarographic
and others, are feasible for use in BOD
determinations, often making it possible to
make an estimate of DO or BOD when sample
interference prevents a valid Winkler DO
determination.
Electronic probe DO makes it possible to
determine many successive DO's at different
time intervals on the same bottle with
negligible sample loss. Reaeration or
extended time series, therefore, are more
feasible.
Another outline in this series describes
response of reaerated BOD,, with electronic
DO probes.
It is the rebponsibility of the analyst to
evaluate:
1 Applicability of the specified technique
and sample.
2 To determine requirements for mixing
and possible thermal effects while
mixing in terms of instrument response
and biochemical reaction.
3 To evaluate long-term calibration or
standardization and their effects upon
precision and accuracy of the BOD
result.
VIH IMMEDIATE DISSOLVED OXYGEN
DEMAND (IDOD)
Immediate dissolved oxygen demand include;
dissolved oxygen utilization requirements of
substances such as ferrous iron, sulfite
and sulfide which are susceptible to high
rate chemical oxidation.
REAERATION METHODS FOR i.O.D. DETERMINATION
reservoir
AA
AAAA
sealed bottles
ELMORE METHOD
d.o. samples
OtFOiD METHOD
10-5
-------
Biochemical Oxygen Demand Test Procedures
A The IDOD is an apparent response as
indicated by a specified technique. Since
DO titration is based upon iodine titration,
any factor that causes Io response different
from that produced by the reaction of KI
and molecular oxygen confuses the IDOD
determination,
B IDOD Determination
1 The IDOD determination includes the
determination of DO on a sample and
dilution water separately. A waste
likely to have a significant IDOD is
unlikely to show a DO.
2 According to mixing theory, it should
be possible to calculate the DO of any
definite mixture of the sample and
dilution water from the DO of component
parts and their proportion.
The same relative proportions of sample
and dilution water should be mixed
without air entrainment and the DO
*6 arbitrarilv elected
4 Any difference between the calculated
initial DO as obtained in 2 above, and
the DO determined in 3 above, may be
designated as IDOD.
5 Sample aeration. DO interference, and
other factors affect results for IDOD.
C Sample Calculation of IDOD
1 Sample DO checked and shown to
be 0. 0 mg/ 1
Dilution water DO found to be 8. 2 mg! \
Assume a mixture of 9 parts of dilution
water and 1 part (V/V) of sample.
Calculated DO -
1X0=0
9 X 8.2 = 73.8
10 parts of the mixture contain 73. 8/10
or 7. 4 mg DO/1. Note that mixing has
reduced the DO concentration because
the original amount is present in a
larger package.
The mixture described above was held
for 15 minutes and the DO determined
was 4. 3 mg/1.
100
7c sample"
used
IDOD = DO . - DO . . ->
calc detm
= 7.4 - 4. 3 x 10
= 31 mgIDOD/1
REFERENCES
1 Standard Methods, APHA-AWWA-WPCF
13th fid. 1971.
2 Methods for Chemical Analysis of Water
and Wastes, EPA-MDQARL. 1974.
This outline was prepared by F. J. Ludzack,
Chemist, National Training Center, MPOD,
OWPO, USEPA, Cincinnati, Ohio 45268.
Descriptors: Biochemical Oxygen Demand,
Chemical Analysis, Dissolved Oxygen,
Water Analysis, Analysis, Wastewater
10-6
-------
BOD DETERMINATION - REAERATED BOTTLE PROBE TECHNIQUE
I INTRODUCTION
A Customary dilution technique has certain
limitations with respect to the BOD
determination such as a prior commit-
ment on sample dilutions to be used,
number of bottles to be included to permit
Wlnkler DO determinations to be made
throughout a predetermined test interval,
(with sample destruction for each DO test),
possibility for anomalous effects due to
dilution water, dilution, or inconsistent
response between test bottles.
B DO probe technique offers the advantages
of nondestructive DO testing, possibilities
of adjusting routine according to sample
behavior during test, and retaining the
sample with minor losses for long term
observation (if desired) and for supplemen-
tary tests at the end of the BOD test
interval.
Note: Use of the probe results in a loss
of about 1/2 ml of each 300 ml sample.
Use dilution water with a glass rod or
bead to bring the sample volume back to
300 ml for further testing.
C This test was performed in triplicate on
the same sample at 100% concentration
with reaeration as needed to illustrate
results obtainable.
H PROCEDURE
A Sample
Final effluent (catch sample at 1100 hours)
of the Advanced Waste Treatment Research
Activated Sludge (Unit B) Sanitary Engineering
Center Experimental Wing (September 3, 1970).
One hundred percent concentration.
B DO Probe Calibration
The probe was calibrated daily in air
saturated tap water versus the azide
modification of the iodometric titration.
A replicate test bottle of the titrated
sample was retained for initial probe
adjustment, if necessary, and to recheck
sensor response at the end of the use
period. Zero sensor response in cobalt
treated sulfite solution was checked at
three day intervals. Electronic response
and battery condition were checked daily.
C Reaeration
1 When the DO test indicated 1. 0 mg
DO/liter or appeared likely to reach
that point prior to the next test reading
the test bottles were reaerated and a
new initial DO was obtained.
2 An adapter was inserted into the neck
of the sample bottle indicating low DO
and an empty bottle inserted on the
opposite end. The combination was
shaken vigorously, while in an inverted
position to allow the transfer of sample
to the empty bottle; immediately after
transfer the combination was reinverted,
shaken as before and sample returned
to the original bottle.
3 The sample was allowed to stand for
at least five minutes to allow entrained
air to rise (generally formed a froth
ring). The froth ring was raised into
the neck of the bottle by displacement
with about 1/2 ml of dilution water.
It usually was necessary to tilt the
bottle slightly and roll it to sweep
fine air bubbles off the shoulder of the
bottle and into the neck. After careful
reinsertion of the DO sensor these
bubbles were displaced by tipping the
assembly and discarding contents of
the BOD bottle lip containing the frothy
residue.
4 It is recognized that surface active
materials and possibly other components
would concentrate in the discarded froth.
It would be possible to limit this effect
CH.O.bod.lab.2a. .11.75
10-7
-------
BOD Determination - Reaerated Bottle Probe Technique _ __
by allowing a longer time for the froth
to break--say 15 to 30 minutes. This
was not attempted for this demonstration.
D No attempt was made to determine DO at
any regular time interval. Observation
time ie recorded for each test on a
24-hour clock basis with zero time at
midnight.
E Results are recorded in tabular form
including the date (Column 1), time in
hours (2), sample temperature (3), hours
of incubation (4), and the triplicate sample
data. For each of the samples the
observed DO, the change since the last
observation (ADO) and the summation of
all the DO depletions observed during
the incubated interval are recorded as
L ADO. A bold face line in the tabulated
data in the DO column indicates DO before
reaeration; the number below the line
indicates DO after reaeration to be used
as a new initial for the next observation.
F Near the end of the fifth day of incubation
(116 hours) one of the triplicates (no. 3)
was not reaerated to check the effect of
complete depletion of oxygen. It was
reaerated on the following day and thereafter.
G The results plotted in graphic form are
presented following the BOD tabular data.
H The results of a similar respiration test
on mixed liquor from the same activated
sludge units are tabulated to show the
effects of different types of feed on DO
depletion (respiratory activity). In this
case, time is in minutes, DO is in mg/1
and ADO/minute also is in mg/1. The
first test in Table 2 represented 100 ml
of return sludge plus plant effluent to fill
the bottle. Approximately one percent
nutrient agar solution was added by
syringe (1.0 ml) to produce the response
noted in the second series; one ml of five
percent mercuric chloride produced the
effects in series 3. A new replicate
sample (100 ml sludge) and plant influent
to fill the bottle produced the results in
series 4.
Ill SUMMARY
A Table 1 shows that it is possible to obtain
consistent results using a DO probe plus
reaeration for BOD technique. The
major requirement is to carefully calibrate
the DO instrument and sensor on a daily
basis. It is not recommended to reaerate
as many times as found necessary here.
It could have been diluted at any time
since the sample was available. Had
the BOD response at five days been
substantially complete only two
reaerations would have been necessary.
However, since the sample was working
into second stage BOD it proceeded to
react accordingly. The oxygen demand
of 15 mg/liter at five days increased to
30 mg/liter in seven days and reached
45 mg/liter in ten days. The sample that
was allowed to deplete on the fifth day
slowly recovered but by the twelfth day
or thereafter equaled or exceeded the
BOD of the samples with residual DO at
all times. This technique requires much
less incubator space, time and manip-
ulation with an added advantage of
retaining the sample for adjustment or
extended observation. The results do
not depend upon a preconceived guestimate
which may or may not fit the situation.
B Table 2 indicates the possibility for
estimating feed acceptability in an
activated sludge treatment plant. On
many occasions an inquiry from an
industrial plant, a different type of
inflow or perhaps a new plant requires
a decision regarding effects at the treat-
ment plant. A respiration test by probe
technique offers an opportunity to rate
the situation within ten minutes after
probe calibration. A test of sludge and
effluent oxygen demand gives an estimate
sludge respiratory activity now. When the
same sludge is subjected to a new feed of
an acceptable nature respiratory activity
rises as in series 2. If the feed happened
to be highly toxic as in series 3 (mercuric
chloride) respiratory activity stops--
10-8
-------
BOD Determination - Reaerated Bottle Probe Technique
obviously a dead or severely shocked
condition. A replicate sludge (Series 4)
with the regular plant influent shows a
prompt increase in activity. Oxygen
requirements, permissible load ratio,
acceptable feed to sludge ratios (return
sludge adjustment) all may be estimated
on the basis of DO tests related to overall
plant performance.
The graphic of BOD versus time includes
the COD--an estimate of first stage
oxygen demand and the TKN for the
composite effluent sample. If the total
Kjeldahl nitrogen (TKN) is multiplied by
its oxygen equivalent (4. 57) then the COD
plus TKN times 4,57 gives an estimate
of ultimate oxygen demand (138 mg/1).
The twenty day BOD was about f>f> \
or forty percent of this value. The effluent
sample obtained at 1100 hours is unlikely
to contain a significant portion of the high
load period considering the primary,
aerator and secondary clarifier detention.
Also the COD and TKN may have included
materials that were chemically oxidized
but gave a poor response to biological
oxidation. The BOD gives the rapid
respiratory oxygen demand. COD plus
the oxygen equivalent of N (on the same
sample) gives a quick estimate of potential
ultimate demand.
This outline was prepared by F. J. Ludzack,
Chemist, National Training Center, MPOD,
OWPO, USEPA, Cincinnati, Ohio 45268.
Descriptors: Analysis, Biochemical Oxygen
Demand, Chemical Analysis, Dissolved Oxvs
Wastewater, Water Analysis
10-9
-------
BOD Determination - Reaerated Bottle Probe Technique
TABLE 1
BOD Tabulated Results - Reaerated Sample; Probe DO
9/3/70 100% Sample
Date
9/3/70
9/3/70
9/3/70
9/4/70
9/4/70
9/8/70
9/8/70
9/9/70
9/10/70
9/11/70
8/11/70
9/12/70
9/13/70
9/14/70
9/15/70
9/16/70
* 9/20/70
Time
1340
1355
1620
1010
1600
1340
0940
1235
1200
0800
1400
2035
1130
1200
1340
1200
1240
Toc
24.0
24.0
22. C
20.2
20.6
19.5
19.8
20.0
20.0
20.0
20.0
20.0
20.0
20.0
19.6
19.3
21.6
Hrs.
Incub
0
0.25
2.6
20.5
26.3
72
116
143
166
186
192
222
238
262
288
310
480
Sample 1 ,
DO
6.7
6.7
6.5
4.9
8.2
7.4
1.0
8. 1
0.6
8.6
1.1
8.2
1.0
8.2
1.2
8.4
5.0
8.1
0.5
8.2
6.3
8.3
7.0
5.7
5. 1
3.8
ADO
-
0.2
1.6
0.8
6.4
7.5
7.5
7.2
7.0
3.4
7.6
1.9
1.3
1.3
0.6
1.3
EADO
0.2
1.8
2.6
9.0
16.5
24.0
31.2
38.2
41.6
49.2
51. 1
52.4
53.7
54.3
55.6
Sample 2
DO
6.7
6.7
6.5
4.7
8.3
7.5
1.5
8.1
0.5
7.6
1. 1
8.2
1.2
8.6
1.1
8.3
5.3
8.4
0.6
8.2
3.8
8.4
6.9
6.2
5.8
4.6
ADO
-
0.2
1.8
0.8
6.0
7.6
6.5
7.0
7.5
3.0
EADO
0.2
2.0
2.8
8.8
16.4
22.9
29.9
37.4
40.4
I
7.8
4.4
1.5
48.2
52.6
53. 1
0.7 | 53.8
0.4 j 54.2
1.2 j 55.4
Sample 3
DO
6.8
6.8
6.6
4.8
7.9
6.9
1. 1
7.4
0.4
0.0
7.6
2.0
8.0
2.0
8.5
5.7
8.6
0.4
8.2
4.6
8.3
0.9
8.3
4.5
8.2
7.2
4.9
ADO
-
0.2
1.8
1.0
5.8
7.0
0.4
5.6
6.0
2.8
EADO
0.2
2.0
3.0
8. 8
15. 8
16.2
21.8
27. 8
30.6
I
8.2 ' 38.8
3.6
42.4
7.4 49 . 8
3.8
1.0
2.3
53.6
54.6
56.9
* Incubator power off
10-10
-------
BOD Determination - Reaerated Bottle Probe Technique
ACTIVATED SLUDGK HKSPIKATION DATA 1>O fliOHK
Time
Min
0
1
2
3
4
0
1
2
3
0
1
2
3
0
1
2
3
Temp
240
24.5
24.8
24.2
DO
7.2
6.6
6.2
5.6
5. 1
4.7
3.8
2.7
1.7
4.7
4.8
4.7
4.7
5.9
4.6
3.3
2.0
ADO
Series 1
0.6
0.4
0.6
0.5
Series 2
0.9
1.1
1.0
Series 3
Nil
Series 4
1.3
1.3
1.3
Sample
100 ml Return Sludge & Effluent
+ 1 ml Nutrient agar
+ 1 ml HgCl2 Soln
100 ml Sludge + Influent
10-11
-------
o
I
GRAPHIC OF REAERATED PROBE BOD RESULTS
STARTING 9/3/70 SAMPLED 1100 HOURS
COMPOSITE SAMPLE RESULTS FOR 9/3/70
COD-69 m/LITER
TKN-15 Hf/LITER
ULTIMATE DEMAND 138 mg/LITER
23 DO DETERMINATIONS
1/2 ml DISPLACED PER
DETERMINATION
LOSS
ml * 100
300
= 3.8%
SAMPLE ALLOWED TO DEPLETE
INCUBATION DAYS jhrs)
1 1 1
8(216) 10(268) 12(312)
I
14
I
16
I
18
I
20
-------
EFFECT OF SOME VARIABLES ON THE BOD TEST
I TIME
,-kt
A The common equation y^=L(l-10 "") for
BOD relationship indicates time as a
variable. The rate coefficient (k^lndicates
that a specific percentage of material
initially present (oxygen) will be used
during a given time unit. Each successive
unit of time has less reactant present
initially than the preceding interval, hence
a definite precentage decrease results in
successively smaller amounts of reactant
use per unit of lapsed time. Increasing
kj results in a larger percentage oxygen
use per unit of time and also increases the
change in reactant mass among successive
time intervals.
B Adney's work for the British Royal Com-
mission cited 5 days passage time from
source to the ocean as maximum for
English streams. The 8th Report (1909)
largely established BOD philosophy in-
cluding the 5-day interval. At 5 days,
initial lags generally have terminated and
a substantial fraction of the long-term
oxygen demand has been exerted. If only
one time interval can be used, 7 days
permits better scheduling. Any one time
Interval is "a" fraction of the total oxygen
requirement; this is a poor reference
point if we do not know how it arrived.
For example, the percentage of
oxidizable material stabilized in terms
of oxygen use at various rate factors
are:
% oxidized
(logio>
k., (loe,,,) in 5 days KI (log )
0.1)5
0. 10
0. 15
0. 25
0.50
This range (K =2.3 k^ is commonly en-
countered in wasteWater stabilization with
the higher rates characteristic of fresh oxi-
dizable material that is readily converted.
in 5 days
42%
67%
84%
94%
99+%
K! (L
0. 11
0.23
0.34
0.57
1. 15
The lower coefficients are characteristic
of cell mass at later stages of oxidation
and of low-rate reactants in general.
The oxygen utilization at specified inter-
vals of time are required to estimate k.,..
and L, the estimate of oxygen use at
infinite time. It is common to observe
results at equal intervals of time but
this is not essential as long as
the time intervals are accurately known.
The initial time periods are critical as
an error of a few hours in time represent-
a relatively large change in reactant mpss
in a system at maximum instability. Un-
equal time periods can be plotted to define
the curve from which any given interval
can be selected as desired.
D Increasing impoundment of surface wa-r:
provides more time for stabilization or
relatively inert soluble or suspended
pollutants and for organism adaptation
to the situation or pollutants. Long te- ;.
BOD's are essential to indicate changes
in the pattern of oxygen demand vs. time.
It may be expected that one or more
plateaus will be evident in the BOD cur ,
followed by a temporary rise in rate
during second stage oxidation or thereaftr-
Anaerobiosls may cause a rise in rate
coefficient after aerobic conditions are
re-established. Eventually k stabilize L.
at very low values.
1 Rate coefficients tend to be difficu.".:
interpret during long term BOD's
because of progressive changes and
other factors.
a The relative error of the DO test may
be a large fraction of the incremental
DO change during low rate periods.
b Cell mass may agglomerate under
quiescent test conditions and decrease
nutrient availability.
CH.O.bod.56d. 11.75
10-13
-------
Effect of Some Variables on the BOD Test
c U it not likely t)iat recycled nutrients
under aerobic test conditions will
Have as much effect as recycle from
anaerobic benthic deposits in a
stream.
2 The BOD result tends to underestimate
d«oxygenation relative to surface water
behavior because of interchanges,
turbvdence, biota, and boundary effects.
Reeceokig doeo not occur in a sealed
bottle but rcseeding IB inevitable in a
atrwuu or Treatment unit.
U TEMPERATURE
A Effect on Oxidation Hate
Temperature is one of the important con-
trolling labors in any biological system.
In the BOD reaction, changes in tempera-
ture prodjce acceleration or depression
of the rate of oxidation. Figure 1 shows
the changes in the value of k at tempera-
tures from 0 - 25°C on a common
wastewater.
B Teat Temperature
In th« BOD test procedure an arbitrary
tempera U: re ie usually selected for
convenience even though a wide temperature
range exists under natural conditions.
Incubctioa of the teat containers at 20°C
for tha whole period is now accepted
practice in the U.S.; 18. 5°C is preferred
in England. Camp (ASCE, SA591.1, Oct.
65) recommended light and dark bottle
immersion in the stream.
C Temperature Cor rection
When it is necessary to calculate the rate
of oxidation at a temperature other than
20°, the following relationship may be
used:
^ . *
0, 3
0, 2
0. 02
J
L——Ji—.
where:
IJJ
j = rate coefficient at temperature T
kr> -rait? at coefficient al te rape ran; re T,,
6 - tempera lure t oa/f-;-lu;,;, :'or which
Streeter z»A Phe'.p,-- obtained w.e value
i, 047, e changes *ith temperature; it
appears to be higher in the range of
5-15°C thar. ir. iht range of 30 to 40°C.
The value given refers tc 15-30°.
The cited temperature coefficient appears
reasonable for household wastes. It may
not apply lor other wastes where developing
or seed organisms may not tolerate tem-
perature changes as readily. A given
temperature coefficient should be checked
for applicability under specified conditions.
A The organisms involved in biochemical
conversions apparently have an optimum
response near a pH of 7. 0 providing other
environmental factors are favorable; a pH
range of about 5. 5 to 8. 3 apparently is
acceptable (Figure 2). Reactivity is likely
to be significantly lower on both sides of the
acceptable pH range but microbial adapta-
tion may extend the limits appreciably.
For example, trickling filters have operated
with better than 50% treatment efficiency
at pH 3 and 10 after adaptation.
10-14
-------
Effect of Some Variables on the BOD Test
2 C
& >.
Fvgur* 2
4 6 pH 8
B Adjustment of Concentrated Samples
When wastes are more acid than pH 6. 5 or
more alkaline than pH 8. 3, adjustment to
pH 7. 2 is advisable before reliable BOD
values can be obtained.
C Dilution Samples
Standard dilution water is buffered at pH
7.2. Sample-dilution water mixtures should
be checked to make sure that the sample
buffer capacity does not exceed the capacity
of the dilution water for pH adjustment.
IV ESSENTIA L MINERA L NUTRIENTS
A Importance
In 1932 Butterfield reported on the role of
certain minerals in the biochemical oxidation
of sewage and concluded that deficient
minerals often upset metabolic response.
In addition, he found that inadequate nitrogen
and/or phosphorus was a common cause of
low BOD results in industrial wastewaters.
(Figure 3)
J S 4 5 « T
Time In D»yi
Effect of Minerml Nutrients on BOD
H Standard Methods Dilution Water
The dilution water specified for the BOD
test approximates USGS estimates for an
average U.S. mineral content of surface
water except for added phosphate buffer.
It is assumed to provide essential mineral
nutrients for most wastewaters but cannot
be expected to meet requirements for
grossly deficient wastewater nutrients both
mineral and organic. Ruchhoft (S.W. J.
13:669, 1941) summarized committee action
leading to the present dilution water.
C Other Dilution Considerations
There is a trend toward the use of receiving
water, storage-stabilized if necessary, to
evaluate waste behavior. It is advisable
to minimize dilution and consider the
nutrient level likely in the receiving water
as most valid. Any change in the environ-
ment, such as dilution, upsets the
microbial balance and requires adaptive
changes.
V MICROBIOLOGICA L POPU LA TION
A Need for Complex Flora and Fauna
Butterfield, Purdy, and Theriault (Pub.
Health Rep. 393, 1931) demonstrated that
an isolated species of organisms was not
as effective in biological stabilization as
a variety of species. Figure 4 summarires
some of their data. Bhatta and Gaudy
(ASCE, SA3, 91:63, June 1965) reinvestigated
this factor. Many studies have emphasized
the need for a mixed biota in the BOD test.
It appears that bacteria are capable of
varied activities, but all species are not
capable of synthesizing all required nutrients.
Certain bacterial species may be capable
of producing enzymes, amino acids, or
growth factors needed for their use and by
other species for optimum performance.
It has been shown that oxygen demand
becomes minimal when some limit of
bacterial population has been reached.
Predation prevents such an approach to
maximum numbers and maintains a con-
tinuing bacterial growth and recycle of
nutrients among a mixed population. The
net effect is a symbiotic relation among
mixed organisms tending to enhance the
rate of stabilization or utilization of
oxygen as in the BOD test.
10-15
-------
Effect of Some Variables on the BOD Test
B Organism Adaptation
1 Early invest!gallons in "elation to the
BOD test considered domestic H a stew:-re i-;:-:
primarily. The saprophytic organisms
involved in stabilization eithf-r were
prevent in adequate numbers or quickly
multiplied to attain effective populations.
<
2 ThJperiod of adjustment required to
shift enzyme production needed to utilise
an'energy source differ ant from that
previously utilized or to shift population
variety from that favored by one food tt>
that favored by another food i,-- con-
sidered an adaptation period. Dilution.
temperature, oxygen tension, pH,
nutrient type, inhibitory substances,
light and other changes all are common
inducements for microbial adaptation.
Mutation of organisms nisy be encountered
during adaptation but usual1,;- i? not a
factor.
3 The developments in industry and
technology have resulted in discharge
of new and more varied waatewatc-r
constituents. Microorganisms may
adapt themselves to the use of a new
substance as an energy source providing
the energy and environment are favor-
able. The receiving stream usually shows
development of an adapted microbiota
for a new or different discharge con-
stituent within hours, days or weeks
after fairly regular discharge. The
time for adaptation depends on the na ure
of the constituent, available energy..
tolerance of the organisms, and environ-
mental conditions.
C Seedtnl
The amount of seed and its selection must
be determined experimentally. The most
effective inoculant would be that which
WOttW produce the maximum BOD response
with minimum lag period and negligible
•eed demand. This would mean some
nfMExfrnum population adapted to feed and
conditions at a minimum equilibrium energy
nutrient supply.
pro.'irrssiixi on s. synthetic feed with
•'vet s--a*?r am'- :-ttU^ sewage inoculants
ai several cor?ce^trations. The river
\v?t*?r resulted in higher BOD with
>;eg?ieible I?.?? and set-d correction. The
deed correction at 20% concentration
of inoculant was less than 0.3 mg.
DO/lat 5 day a. i» would be possible to
use this river v/atsr as a diluent without
excessive oxygen loss to produce more
valid BOD progression for that receiving
water. The lower waetev/ater inoculant
concentration resulted in a definite BOD
lag. Higher wastewater concentrations
produced comparable BOD progression
earlier but resulted in high seed
corrections and lowered availability of
rissolved oxygen for the sample.
2 A good secondary treated effluent
,j.roduc«d results similar tc river wai^r
inoculation with higher seoo corrections
per increment of applied tncculant.
Soil suspensions also are v^ry effective
sou; ces of seed organisms with minor
s^ed corrections if chey are reasonablj
stabilized surface soils.,
3 It appears that the BOD progression
meat nearly indicating receiving water
oxidation would be one based upon
receiving water dilution or inoculated
with organisms from it.
•1 A new or unusual wastewater may
require adapted organisms not present
in sufficient numbers in the receiving
watei-. Development of an adapted seed
from soil suspensions, plant effluents
or receiving water may be necessary to
evaluate oxidation potential in a plant
or receiving water at some future tims.
Enrichment culture technique is bene-
ficial whore small concentrations of the
test wastewater are applied regularly
v.'ith increases in w&stewater concen-
trations as BOD or respiration activity
indicates increasing tolerance and
oxidation of the test waste. Both time
and concentration limits are useful to
characterize the waste-water and its
acceptability for biological stabilization.
10-16
-------
Effect of Some Variables on the BOD Test
All tormi In river w»t»r
Mixed occur!* t> plankton
Pure culture B. Aerogene* at plankton
Mlied culturt ixrlerU
Pure culturt 8. A*rogtntm
4 3 6 7 «
Time in Daya
l of Biological Form* on Oxygen Depletion
Figure 4
DO DEPLETION VS SEED
6 CONCENTRATION & TYPE (ax oiucost-otuiAMtc AC.)
RIVER WATER
STALE W W
0.5% STALE SEW -
0.2%
0.1%
1-20% RIVER WATER
DAYS
Figui
It must be recognized that BOD
progressions are most likely to err
on the low side. A meaningful BOD
test should seek the highest consistent
oxygen demand feasible for sample and
conditions.
D Algae
When large numbers of algae are present
in surface waters, they produce significant
changes in the oxygen content. Under the
influence of sunlight excess oxygen is
produced while a net deficit occurs in the
dark. The result is a wide variation in
surface water DO depending on sample
time.
When stream samples containing algae are
incubated in the laboratory the algae
survive for a time, then die because of the
lack of light. Short-term BOD determina-
tions may show the influence of oxygen
production by the algae. When the algae
die, they release the stored organic load
for recycle and increase the BOD. There-
fore, samples incubated in the dark may
not be representative of the deoxygenation
process in the stream, since the benefits
of photosynthesis are lacking. Conversely,
samples incubated in the light, under
conditions of continual photosynthesis,
will yield low BOD values.
The influence of algae on BOD is one of
the most difficult variables to evaluate.
More research is needed to develop
satisfactory methods for the accurate
determination of BOD in the presence o;
large numbers of algae. Light and dark
bottle incubations suggest the magnitude
of effects.
IV TOXICITY
A Effect
Since satisfaction of the DOD is accom-
plished through the action of microorgan-
isms, the presence of toxic substances
will result in depression of the oxidation
rate. In many cases, toxicity will
produce a lag period, until tolerant
organism activity becomes significant.
Figure 6 shows the effect of cyanide on the
BOD curve. A prominent lag period is
exhibited in the 2 ppm curve, while at
10 ppm the lag extends beyonu the fifth day.
An activated sludge may be adapted to work
effectively in degradation of 60 mg C'N/1.
-------
Effect of Some Variables on the BOD Test
Table I
i i
TUB* ill D*jm
Ctttct al Cyuld* on BOO at Domestic
(M Scwaf* IB Formula C OUatiaa Wiur)
Figure 6
Heavy metals have similar effects depending
on history and environment. The effects of
copper and chromium are illustrated in
Figure 7.
Waste
cone.
10%
5%
2%
1%
0.5%
Depletion
3.51
4.53
2.80
1.52
0.74
5 day BOD
35
91
140
152
148
EFFECT OF HEAW \IET\LS O\ BOD
Figure 7
B Detection
In laboratory determinations of BOD the
absence of toxic substances including
chlorine must be established before the
results can be accepted as valid.
Comparison of BOD values for several
dilutions of the waste will indicate the
presence or absence of toxicity. In Table 1
the calculated BOD for the dilutions show
higher values in the more dilute concen-
trations. It is apparent that toxicity was
present and that the toxic effect was diluted
out at a waste concentration of 2% or less.
VII NITRIFICATION
A Mechanism
The oxidation process, as exemplified by
the equation:
y = L (l-10~kt)
presumably involves the oxidation of
carbonaceous matter or 1st stage oxygen
demand.
C H O -2 C02 + H20
The rate coefficient is normally high, giving
nearly complete oxidation in a few days.
When nitrogenous material is present its
oxidation can be shown as:
NH3
_ O _
N°2 N°3
Nitrogen oxidation may be delayed for
several days during BOD tests unless
suitable micro-biota are initially available.
Under some circumstances these two
oxidations can proceed simultaneously and
the resultant BOD curve will be a com-
posite of the two reactions.
k t
v = „ (1-10 c) + L (1-10" n
't L c n
where yt «= the simultaneous BOD of the car-
bonaceous and nitrogenous phases or 1st and
2nd stage demands.
kc and kn = the rate coefficients appli-
cable to the carbonaceous and nitrogenous
materials respectively.
16-18
-------
Effect of Some Variables on the BOD Test
B
L and L « the ultimate oxygen demands
characteristic of the two phases respectively.
This is the general formula for any system
characterized by two simultaneous reactions.
Principal conditions governing simultaneous
carbon and nitrogen oxidation:
1 Presence of an effective nitrifying
culture at the beginning of the test
interval (nitrifiers grow relatively
slowly).
2 Maintenance of adequate DO, believed
to be a minimum of 0. 5 to 1.0 mg/1,
for nitrifier activity.
3 Available nitrogen - in excess of that
required for synthesis. This is believed
to require a minimum of about 7 mg/1
to support active nitrification on a
continuous basis.
4 Nitrifiers appear to be more sensitive
to toxicity than most saprophytic
organisms, hence are likely to be
inhibited more readily. This is
particularly evident during nitrite to
nitrate'conversion.
It may require 5 to 10 days to establish
nitrification if the population was not
nitrifying initially. This is the basis for
the sequential carbonaceous and nitrogenous
oxidation of sewage oxidation.
1 Effects on the BOD curve indicate a
typical pattern such as in Figure 8.
The influence of nitrification in the
production of a secondary rise in the
BOD curve is so well known that any
secondary rise may be erroneously
attributed to nitrification whether or
not nitrification was involved. Actually,
a secondary rise in the curve may be
due to any oxidation system assuming
dominance after the initial oxidation
system has been completed.
2 The nitrification phenomena occurs
simultaneously in many streams,
treated effluents or partially stabilized
samples. The designation of a secondary
BOD rise to nitrification should be
based on analysis, not curve shape.
C The extent of nitrification is conclusively
shown only by periodic analysis of
ammonia, organic, nitrite and nitrate
nitrogen. The conversion of ammonia
and organic nitrogen to oxidized nitrogen
is a definite indication of nitrification.
D Nitrification Inhibition
Plant efficiencies from a BOD standpoint
can be erroneous because nitrification
generally is not established during the
usual incubation of influent samples but
may be a major factor in effluent
incubations. It requires about 2 times
the oxygen to convert NEL -N to NO_ -N
as to convert C to CO hence this is a
major fraction of stream oxygen use.
Most secondary treated effluents are
characterized by a larger fraction of
carbon than nitrogen removal which
accentuates the problem.
Pasteurization of samples, methylene
blue, chromium, and acid treatment
followed by neutralization have been used
to inhibit nitrification for estimation of
carbonaceous BOD only. Any inhibition
of nitrification also produces a change in
the sample or its behavior and may
partially inhibit carbonaceous oxidation.
Nitrification is a factor in stream self-
purification and treatment. It does not
appear realistic to alter it for convenience.
The most realistic approach to carbon-
aceous oxidation is the measurement of
CO2 or COD.
Figure 8
10-19
-------
Effect of Some Variables on the BOD Test
VIII EFFECT OF DILUTION
When a series of dilutions are made on a
BOO sample usually the result a vary to the
extent that only an approximate BOD value
is obtained.
Table 2
INTERPRETATION OF BOD DATA
Sample cone. DO
i
, Depletion
Initial ; 8. 2
Final:
1* , 5-5 2-7
2* 3. 3 ; 4.9
4% 0.'
) Complete
BOD
"
2TO
245
A For example, in Table 2, 1%, 2% and 4%
concentrations of sample were used. The
4% concentration became anaerobic before
the end of 5 days. The 5-day BOD of the
1% concentration was 270 and that of the
2% concentration was 245.
B Statistically one value is more reliable
than the other.
Dilation
1%
2%
Difference
DO depletion
5.5 mg/1
3. 3 mg/1
2. 2 mg/1
The difference in depletion between 1 and
2% dilations is 2.2 mg/1. This difference
may be attributed to an additional 1% of
sample added to the original 1%. If the
difference is multiplied by the dilution
factor of 100 to obtain the BOD, the result
is 220 mg/L.
1 We now have three estimates of the
BOD on a one percent concentration
basis from the two dilutions:
a the actual 1% depletion gives 270
b 2%/2 depletion gives 245
c (2%- 1%) depletion gives 220
Statistically the probabilities of being
nearer the actual value goes with the
nearest two of three. The 4% value
of 8.2 dupletion/4 as a minimum
possible BOD 1% concentration gives
a BOD of at least 200.
There is the possibility that higher
concentrations may reflect significant
toxicity while lower concentrations
tend to reflect a greater proportion of
dilution water. The toxicity problem
does not appear to be significant since
the 4% sample concentration indicated
a BOD of at least 200. The higher
BOD at 1% sample concentration may
be due to a contaminated dilution water
or to the fact that a similar number of
seed organisms had less food and
utilized certain fractions that they had
passed by when they had more choice
with the 2% sample concentration.
Data is insufficient to resolve this one.
Incubations having a depletion of at
least 2 mg DO/liter and a residual of
at least 1 mg DO/liter are indicated
to be most valid**). Both the 1 and
2% concentrations fit this requirement
in Table 2. An average error of
+ or -0.1 ml on the DO titration would
have a smaller relative error upon
the 2% depletion.
We have a reasonable presumption
that the sample BOD of about 230 was
a good estimate. We do not have an
unequivocal basis for so stating.
Possible variations in results with
different dilutions of a given sample
are subject to many uncertainties in
the test routine.
If some cause is known - such as a
titration eror, the inclusion of ex-
traneous substances producing high
or low response, or a definite procedural
error that rules out a valid estimate of
the sample BOD- that result should be
labeled as a lost cause or unreported.
Otherwise, report what was obtained
to the best of your ability with the
provision of uncertainty for uncon-
trollables.
10-20
-------
Effect of Some Variables on the BOD Test
A CKNOW LEDGM ENT:
Certain portions of this outline contain
training material from prior outlines by
D. G. Ballinger and J. W. Mandia.
REFERENCE:
Standard Methods, APHA-AWWA-WPCF,
13th edition, 1971.
This outline was prepared by F. J. Ludzack,
Chemist, National Training Center, MPOD,
OWPO, USEPA, Cincinnati, Ohio 45268.
Descriptors: Algae, Ammonia, Bacteria,
Biochemical Oxygen Demand, Essential
Nutrients, Microorganisms, Nitrates, Nitri-
fication, Nitrites, Nutrients, Oxygen Demand,
Rates, Time, Toxicity, Waste Dilution,
Water Analysis, Chemical Analysis
10-21
-------
MATHEMATICAL BASIS OF THE BIOCHEMICAL OXYGEN DEMAND (BOD) TEST
Part 1
I FUNDAMENTAL CONCEPTS
For a number of years, the oxidation of
organic waste substances in surface waters
haa been under investigation. The first re-
ported observation of oxygen depletion was
on the Seine River below Paris (1877). Other
studies followed in Germany, England and
the U.S. Certain fundamentals are now
universally accepted. These are:
A Dissolved oxygen in the water is reduced
during stabilization of the organic material.
B As long as dissolved oxygen is present,
the rate of oxidation is independent of the
actual amount of oxygen available.
C The type and numbers of biological forms
present is an important factor.
D Measurement of changes in oxygen content
can be related to quantity and character
of oxidizable organic matter.
II FIRST ORDER RATE CONCEPT
In order to illustrate the mathematical
relationships in the BOO reaction, assume
the following laboratory observations:
A A set of replicate bottles is filled with
river water and sealed so that outside
air is excluded.
B Each day, one of the bottles is analyzed
for DO content, and the results tabulated.
(See Table 1.)
It will be noted that on each successive
day, the DO concentration is less than
the day before. That is, oxygen is being
consumed by biological or chemical
action in the water. The oxygen residual
when plotted against time forms the
deoxygenation curve (Figure 1).
The oxygen demand or BOD curve is re-
presented by the summation of observed
oxygen depletions ve time (Figure 2). This
Table 1.
Days
DO
Oxygen depletion
Per day Cumulative
0
1
2
3
' 4
5
6
7
8
9
10
9.2
7.4
5.8
4.6
3. 7
2.9
2.3
1.8
1.5
1.2
0.9
0
1.8
1.6
1.2
0. 9
0. 8
0.6
0.5
0.3
0.3
0.3
-
1.8
3.4
4.6
5.5
6.3
6.9
7.4
7. 7
8.0
8.3
CH. O.bod. 58c.ll.75
Preceding page Hank
10-23
-------
Mathematical BatuB of the BOD Teat
Figure 1
is an inverted representation of the de-
oxygenation curve. Inclusion of the daily
equilavents of oxygen demand in Figure 2
shows that increasingly smaller amounts
of oxygen normally are used during each
successive day.
If the reaction follows a Ist-order
rate system, a constant percentage of the
oxygen present at the beginning of each
day will be used during that day. There-
fore, the rate coefficient (K in nat. log.
or k in log,0) is constant from day to day.
Kj =^.3 k
:
Similarly, the organic matter present in
the sample is being oxidized, so there is
progressively less oxidizable material
present each successive day. The rela-
tionship between oxygen demand and the
amount of oxidizable material present
can be stated as follows: The oxygen
demand per unit time is proportional to
the amount of unoxidized material present.
Streeter and Phelps^ *• ^stated the concept
as - the rate of biochemical oxidation of
organic matter is proportional to the con-
centration of unoxidized substance,
measured in terms of oxidizability. Such
a rate ia termed a First Order Reaction
Rate. The same authors also stated that
there is no logical reason to expect one
raie coefficient ["but iTlnay"appear so
I of crude measurement and the effects of
many individual oxidation systems.
TIMI IN MfTS
Figure 2. INTEGRAL BOD CURVE
The curve as shown in Figure 2 is a
typical BOD curve. When interpreted in
light of the fundamentals previously dis-
cussed, it is apparent that:
1 With organic materials which are
easily oxidized, the reaction proceeds
rapidly, nearing stabilization in a
few days. The curve is therefore
steep, rapidly approaching a maximum.
2 As oxidation nears completion the BOD
curve will approach a limit or maximuir
value that can be represented by a line
above the time axis and parallel to it.
This line represents the ultimate oxyger
demand (L) for the sample and condi-
tions of the test. The curvature of the
BOD line as it approaches the maximum
is a function of the rate of oxidation.
i The BOD test commonly is applied for
samples containing mixed components
each of which may differ in availability
as a nutrient. The population of organ-
isms generally contains many varieties.
A progression of varieties occurs from
those favored by an excess of initial
nutrients to those favored by cell mass
and eventually to those tolerating a low
available energy level.
-------
Mathematical Basis of the BOD Test
4 This type of system may have apparent
characteristics of 1st order reaction
mechanisms for limited periods of
time. It is likely that two or more 1st
order mechanisms may be apparent if
observation time, nutrients and condi-
tions are appropriate. The data also
may be interpreted as a zero order, 2nd
order or multiple order reaction
mechanism. It is not always possible
to prove reaction order nor is it essen-
tial to do so. The main objectives are
to be able to estimate the ultimate BOD,
the amount of BOD remaining after a
given time interval and to obtain rea-
sonable agreement among observed and
calculated data. The first order
reaction concept provides "a" means
to meet the need without implying that
the reactions are necessarily 1st order
or that it is the only means of approach.
t
L
L1
y
then:
time in days
ultimate demand
demand remaining at time t
demand satisfied at time t
fraction of L remaining at time t
= fraction of L oxidized at time t
The relationships may be expressed as a
differential equation with respect to "t"
dL
dt
-KL
(1)
UI BOD EQUATION
Using the typical BOD reaction curve, it is
possible to develop equations expressing the
various relationships. Labeling the coordi-
nates as in Figure 3.
where K is a rate coefficient for deoxygena-
tion as a loge.
changing to
-Kt
(2)
Figure 3
Log
(L-y) _
10 L
kt
and:
Table 2
Equivalents in the BOD Equation
BOD remaining at
time t
Fraction of BOD
remaining
Conversion of loge
to log 10
Fraction of BOD
satisfied
L-y/L L
K
t/.
L
2.3k
(3)
10
-kt
10-25
-------
Mathematical Basis of the BOD Teat
from Table 2
1 - 10
-kt
(4)
multiplying both sides of equation (4) by
L:
= Ui - icfkt)
rearranging and cancelling gives:
y s L(l - 10
~kt)
This is the usual form of the BOI? equation.
To avoid confusion with other "k" values,
the deoxygenation rate is usually written
as ki and the equation becomes.
y = L (1 -
(6)
The components in Figure 5 are as in HI
where L is the ultimate oxygen demand at
infinite time and y refers to the demand
satisfied at time t. The term kj refers to
a "velocity constant" in physical chemistry
terminology. Section n B3 indicates
factors that result in system changes
during progression of the BOD test that
may affect the value of kj. For this
reason the term "rate coefficient" is used
for ki to recognize that kj may not be a
"constant" except for limited time periods.
As given, kj is an expression of the slope
dL / as a log and is related to the
rate of oxidation.
The L value is a laboratory number re-
garded as a theoretical rather than an
actual limit. The limit is useful for an
estimation of the effects of a waste upon
the receiving water where the total demand
is more valid than a part of it as for
example a 5 day BOD.
IV SIGNIFICANCE OF BOD CONSTANTS
A The effect of kj upon y is shown in
Figure 4.
The upper curve represents the BOD of a
domestic sewage. The kj is 0. 15.
At this rate, 50% of the demand will be
exerted in the first two days, indicating
that the organic matter is readily avail-
able as food for the organisms. In five
days, 83% of the demand has been satis-
fied and in 15 days, the oxidation is
essentially complete.
The lower curve represents the BOD of
a relatively stable sample. The kj
in this case is 0. 05. In five days
only 44% of the demand is satisfied and
at 15 days, there is still 17% of the
material unoxidized. Note that at the
lower rate, it takes 15 days to accom-
plish 83% oxidation, while at the higher
rate, the same percentage can be accom •
plished in 5 days.
MVS
EFFECT OF R RATE ON SHORT TERM MID.
Figure 4
Laboratory tests on the BOD of waste
materials are generally 5-day tests.
In the case of these two samples, the
standard 5-day interval would represent
vastly different degrees of oxidation.
V LOG OF % BOD REMAINING
Using the mathematical relationships estab-
lished above, a second BOD equation can be
developed.
10-26
-------
Mathematical Basis of the BOD Test
From Table 2:
L
-k t
= 10 1 = fraction remaining
-k t
10 1 X 100 " % remaining
-k t
log (10 1 X 100) = login% remaining
= 10'
further:
,-k.t
10g1Q(10 1 X 100) « 2 - kjt
then; ^
log of % BOD remaining = 2 - k^ (7)
VI SUMMARY
The relationships presented herein form the
basis for development of engineering esti-
mates from laboratory data. Validity of the
estimate depends upon sampling, prompt and
correct BOD technique, judicious data
analysis, and realization that the laboratory
BOD test does not include many factors
controlling stream behavior.
Two examples of the estimation of k and L
values are presented in Part 2, which begins
on the next page.
ACKNOWLEDGEMENT:
Certain portions of this outline contain
materials made available by D. G.
Ballinger, F. P. Nixon and D. E.
Baumgartner.
REFERENCES
1 Streeter, H. W., and Phelps, E. B.
Public Health Bulletin 46, USPHS.
1925.
2 Phelps, E. B. Stream Sanitation.
John Wiley & Sons, New York. 1944.
10-27
-------
MATHEMATICAL BASIS OF THE BIOCHEMICAL OXYGEN DEMAND (BOD) TEST
ESTIMATION OF k and L
Part 2
I Several methods are available that may
give a fair estimate of k and L providing
that the observed data plots in a form that
suggests a first order reaction rate fit. The
fact that stream samples frequently consist
of partially stabilized river water mixed
with varying amounts of more recent waste -
water additions leads to numerous situations
where observed ciata calculated as "a" first
order reaction does not give a good fit of
observed and calculated data. The low rate
oxidation of the aged material and high rate
oxidation of the new material simply cannot
be explained in terms of one k^ and L.
The BOD test is not precise enough to dis-
tinguish the infinite number of individual
rate systems included in deoxygenation of
a mixed wastewater by the mixed organisms
involved. The persistence of the one first
order reaction rate concept (incorrectly
called a monomolecular rate) partially is due
to the practical limitations of the test. High
assimilative oxidation rates generally have
been partially completed in the sewer, during
compositing or sample storage before BOD
analysis, or in the receiving water. Remain-
ing deoxygenation may lake the form charac-
teristic of cell mass and storage products,
(endogenous oxidation) which for a few days
shows a ^ of about 0. 1. If the observation
period was extended beyond the usual 7 to
10 days for rate estimation a. progressive
change of k-, would become apparent as the
more oxidizable components disappear and
relatively inert biological or other residues
became a larger and larger fraction of the
total oxidizable mass. The increased im-
poundment of surface water makes long-term
deoxygenation more and more significant.
The computed k-, and L are a result of the
selected measurement routine, past history
of the sample, and the manner of processing
the result. The validity of the estimate
depends upon engineering judgment and a
realization of the variables and effects
involved. No two individuals will derive
precisely the same values from given data
but common sense approaches using mathe-
matics as a tool rather than as a deity, will
produce useful estimates for resolution of
the problem.
II Useful methods of deriving k..
include the least squares method of Reed and
Theriault (Public Health Bulletin 127, 1927);
the Thomas Slope Method (Sew. Works J. 9,
No. 3, 425. 1937), the Moments Method of
Moore, Thomas, and Snow (Sew. and Ind.
Wastes 22, No. 10, 1343, 1950) and the
Rapid Ratio Method of Sheehy (JWPCF 32,
No. 6, 646, 1960). The first three of
these presume a singe rate coefficient
and do not give a good fit of data when de -
oxygenation fails to follow the pattern. Cer
tain other factors enter the picture even
when the curve apparently follows first
order characteristics (Ludzack et al, Sew.
and Ind. Wastes 25, No. 8, 875, 1953).
Sheehy's procedure elucidates the dis-
appearance of high rate components during
early stages and can be very useful for
rapid calculations.
The daily difference method outlined by
Tsivoglou (Oxygen Relations in Streams,
SEC Tech. Report W-58-2. p. 151, 1958)
is an adaptation of Fair and of Velz methods
and gives a graphic picture of observed data
and predominant rate changes with time.
The method is rapid, versatile, and gives a
great deal of information in a simple package
that is readily assembled. This technique
was used to illustrate its applicability as
"one" means of estimation.
A Using the data given in Table 1, Part 1.
Plot y vs. t on cartisian coordinate
paper. Draw a curve of best fit through
the observed points, including lags,
plateaus, if any, to get a picture of
the deoxygenation curve. If a majority
of observed values describe a smooth
Preceding page blank
10-29
-------
Estimation of k and L
progression, those that fail to fit may
be considered erratic and curve values
-may be used for subsequent operations.
Lags should be eliminated by curve
fitting and taking the observed points
after fhe lag termination. (Figure 2)
Plot the daily differences, corrected
if necessary, on semilog paper with
time on the linear scale and the daily
difference on the log scale. The
differences are conventionally plotted
as 1/2, 1-1/2, 2-1/2 days. etc. to
illustrate intervals rather than points.
(Figure 5)
Examination of Figure 2 shows identi-
cal daily differences for the last 3 days
or k^ - 0 for that interval in Figure 5.
That la unlikely considering the short
period of observation and the results
of the first 6 days.
It may be assumed that deoxygenation
follows a smooth curve after consider-
ing possible lag periods or multicom-
ponent rates operating in combination
or sequentially. Daily differences,
therefore, should be calculated from a
line of best fit such as Figure 2 rather
than from the individual points included
in Table 1 which may be affected by
experimental error.
Figure 5 is a plot of the daily differences
from Table 1 on the Iog10 scale vs
time on the arithmetic scale. The
corrected curve of best fit was not used
here to show the effects of the last 3
points on the logarithmic pattern. The
slope of the line was well defined on
the basis of the first 7 points. Note
that the daily differences are plotted
at the middle of the time interval, i. e.,
the deoxygenation occurring from 0 to
1 day appears at the 1/2 day position
and that occurring during the second
day appears at the 1-1/2 day position.
It is advisable to check arithmetic on
the work sheet by comparing the sum
of the daily differences in Table 1 with
that resulting from the line of best fit
in Figure 5. In Table 1: 1.8+1.6 +
1. 2 + . . . . + 0. 3 = 8. 3; in Figure 5;
1. 88 + 1. 50 + 1. 20 + . . . . + 0. 26 =
8. 4. Since the difference in sums is
1 part in 84 it is apparent that serious
errors did not occur in arithmetic or
plotting.
From Figure 5, the number at the zero
intercept of the daily difference slope
is a function of L. Other intercepts
are a function of L^ using equivalents
in TabLe 2. The number representing
L at 0 time = 2. 1; the number repre-
senting Lt when t = 10 is 0. 23. The
percent BOD remaining when t = 10 is
L+/L X 100 or
0. 23 X 100
2. 1
= 10.9 or !
8 From equation 7:
L°glO °^ *^e ft BOD remaining = 2 ~kjt
The Log 10 of 11% « 1. 04 and t - 10
then 1.04 • 2 - k1 X 10
or ki » 2 - 1. 04/10 = -—• = 0. 096 or
10 0.10
9 From equation 6:
yt = L (1 - 10-klt) = L (1 -±*- )
From Table 1, y = 8. 3, and
(1 - ^ ) =0.89 (Step 7)
The estimate of L becomes
y1Q = L (1 - i*) or 8. 3 = L (0. 89)
8.3
L =
0.89
= 9. 34 or 9. 3 mg BOD/1.
10-30
-------
Estimation of k and L
10 The laboratory test gives estimates
of the rate coefficient or k) of 0. 10 and
of L or the Ultimate Oxygen Demand
of 9. 3 mg of BOD per liter of sample.
The data approximates a system con-
sisting primarily of the oxidation of
partially stabilized cell mass and
storage products and plots as a one
rate system during the observation
period.
Equations 6 and 7 provide the funda-
mental relationships and the equivalents
in Table 2 may be read from Figure 5.
The data in. Table 3 present a different
situation involving more recent
contamination.
days
1
2
3
4
5
6
7
8
9
10
Table 3
BOD
2. 24
3. iS
3. 81
4. 22
4.56
4. 88
5. 12
5. 32
5. 49
5. 61
per day
2. 24
0.94
0. 63
0. 41
0. 34
0. 32
0. 24
0. 20
0. 17
0. 12
The differences among the first 3 ob-
served daily differences and the line
ab are 1. 55, 0. 36 and 0. 14 from
Figure 7. If the differences are plott-
ed on Figure 7, (coded as x) a new
rate cd can be estimated that gives a
reasonable fit with 2 of the 3 intervals.
The first daily difference is above the
slope cd suggesting a 3rd rate group
higher than ab or cd. Since this would
be based upon a single point in 1 series,
in which some discrepancy is apparent
in Figure 7 with respect to line fitting,
it is advisable to regard it as a
possibility as shown by the graphic
but withhold judgement on the 3rd rate
group for further evidence. The curve
in Figure 6 may be approximated by a
two rate system operating simultan-
eously with one of the two nearly
completed in 3 days.
3 Mathematically this situation may be
expressed as:
Yt - L .(1 -
cd
+ L . (1 -
ab
or
= L , (1 - •) + L . { 1 - -
cd L , ab L ,
cd ab
where -~ may be obtained from
L
Figure 7 at the designated time intervals
for each slope.
4 From Figure 7 the zero and 10 day
intercepts of line ab are 0. 75 and 0. 13.
The line of best fit of the observed
BODs in Table 3 are shown as a
smooth curve in Figure 6.
Figure 7 shows the daily differences
on the serai log scale vs the time
interval on the arithmetic scale. The
line ab is a reasonable fit for the slope
described by 5 of the 7 daily differences
after the 3rd day. It is apparent that
the curve in Figure 6 cannot be
described as one first order reaction.
_ 0.13 X 100
L 0.75
remaining when t = 10.
= 17.4% BOD
The log1Q of 17. 4 is 1.24.
If the log of the % BOD remaining =
2 - kt
then, 1. 24 = 2 - kab X 10 and kab =
2 - 1.24 0.76
10
10
= 0.076
10-31
-------
5 From Figure 7 the zero and 2 day
intercepts of line cd are 1.40 and 0. 22.
'cd
0.22 X 100
1.40
15. 7% BOD
remaining when t * 2. The logio of
15. 7 is 1. 20 as in Step 3:
2 - 1. 20 0.80
kcd * 2 "2
= 0.40
6 From the relation in B2, Lcd may be
approximated by substituting the ob-
served value for the 2 day BOD in
Table 2 or 3.18. The fraction ofLj
L
from *ero to 2 days is
0.53
ab
or 0. 71
(i . ±3. ) . o. 29. therefore the
Lab
use of the 2 day observed value for
line cd is incorrect because 0. 29 Lfib
is included in the observed value.
Since L . is unknown, the first estimate
of L wul be incorrect by the value
of 0^28 L h but may be approximated
later.
and
-it =
Lcd
then a high approxi-
mation of L-cd becomes.
3. 18 = Lcd X 0. 84
Lcd-05 "3.8mgBODA.
7 A similar approximation of LaD may be
set up for t « 10 where the observed
Y 10 from Table 2 of 5. 6 is corrected
by substracting Led which should be
almost complete by 10 days.
becomes 5. 6 - 3. 8 = 1. 8 and
Substituting the appropriate numbers
as in step 5:
T 1'8
Lab r 0. 826
- 2.18 or 2.2 mg/1.
8 A better approximation of Lcd is now
possible assuming that the 2 day ob-
served value included the following:
= 3. 18 - 0. 29 X 2. 2 or 3. 18 - 0. 84 = 2. 34
and
= 2. 9 mg BOD/1 instead of 3. 8
A second approximation of Lab would
raise the estimate of Lab because
Y10&b = 5. 6 - 2.9 - 2. 7 and
t - 2'7
^ab " 0. 826
and an estimate of the L values of the
sample plotted in Figure 6 would be
2. 9 + 3. 3 or 6.2 mg BOD/ liter.
= 3- 28 or 3- 3
BOD/1
C The substitution of the derived values for
the appropriate and ultimate oxygen
demand into the expression given in B2
should give values comparable to the ob-
served BODs in Table 3 or Figure 6.
Further, it should be possible to approxi-
mate curves for the individual rate systems
to obtain daily vectors, the sum of which
would compare with the observed points.
For example:
Where L , = 2. 9
cd
L
.-k!
3.3
and k , * 0. 40
cd
and k = 0. 076
ab
and 10" may be approximated by -=*•
from Figure 7. (See Table 2)
» 2.9X 0.60+3.3 X 0. 16
= 1. 74 + 0. 53 = 2. 27. observed = 2. 24
» 2. 4 -t- 0. 9 « 3. 3 observed « 3.18
I.. I. -
* 1.3
= 2. 9 + 1. 3 = 4. 2 observed = 3. 81
2.9+ S.S
-------
Estimation of k and L
= 2. 9 + 2. 7 = 5. 6 observed 5. 61
These data show a reasonable approxi-
mation of the observed BODs by a
reverse calculation from the derived
rate and ultimate BOD obtained from
them. The agreement indicates that
approximations made were reasonably
correct and that the composite BOD
curve in Figure 6 may be described as
a 2 rate system as indicated by the
BODs in Figure 6. The largest error
occurred in calculation of the com-
ponents for ¥3 at the point in time where
the higher rate system was being phased
out. Curve fitting, approximations of
correction factors for estimation of L,
plotting, slide rule and other errors
contribute to estimates of kj and L in
addition to the errors related to the
BOD test.
An approximation of kj is possible
from 2 BOD values obtained at different
intervals of time by using the data
presented in the Theriault Tables or
that given nomographically by the chart
on the Rates of BOD Satisfaction.
This outline was prepared by F. J. Ludzack,
Chemist, National Training Center, MPOD,
OWPO». .USEPA, Cincinnati, OH 45268.
Descriptors: Analysis, Analytical Techniques,
Biochemical Oxygen Demand. Data Handling,
Graphical Analysis, Graphical Methods,
Mathematical Models, Mathematical Studies,
Mathematics
10-33
-------
Estimation of k and L
Figure 5. l.og.~ Slope of the Dailj Differences from Figurr 2
l« o ...... __ -- •
iw.w
90
.w
a o
7 0
60
5 A
4M
•V
w 3 O
W
u
2
£
w
£ 2.0
H-l
D
>>
d
s
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
X
X
1
X
2
v
N
3
^-JB
*N
MMMWW^M
D.
5
'X
•B^^BiaB
5
fcYS
N,
S
i««*M«B*«*«
N
^s
— o —
•••^•^••M
7
— ^N
•—
8 9 K
-------
Figure 6. £ BOD of the Data from Table 2.
\
CA
6
COMPOSITE BODs FROM TABLE 2
BODs AS DESCRIBED BY LINE cd FIGURE 7
BODs AS DESCRIBED BY LINE ab FIGURE 7
DAYS
-------
Estimation of k and L
10.0
Figure 7. Log10 Slopes of the Daily Differences From Figure 6 BODs
0.1
-------
Estimation of k and
Thi-riauli Table
NUMERICAL VALUES OF THE FUNCTION (1 - 10" S
FOR THE RANGE k = 0.040 to k = 0. 250
Period of
incubation
(days)
7 _
8- _
i n - — - - - -
1 9_____
j.0--------
1 R-.----
1 7_»__ .
Ofl--
99_
00 _. „
Period of
incubation
(days)
A
0.04
0.088
.168
.241
. 308
.369
.425
.475
. 563
. 602
.637
.669
.698
.725
.749
.771
.791
.809
.826
.842
.855
.868
.880
.890
.900
0.11
0.224
.397
.532
fi37
0.05
0. 109
0.206
0.292
.369
.438
.499
.553
.602
.645
.(384
.718
.749
. 776
.800
.822
.842
.859
.874
.888
.900
.911
.921
.929
.937
.944
0. 12
0.241
.425
.563
.669
0.06
0. 129
.241
.339
.425
.499
.563
.620
.669
.712
.749
.781
.809
.834
.855
.874
.890
.905
.917
.928
.937
.945
.952
.958
.964
.968
0. 13
0.259
.450
.593
.699
Value of k
0.07
0.149
.276
. 383
.475
.553
.620
.676
.725
.766
.800
.830
.855
.877
.895
.911
.924
.935
.945
.953
.960
.966
.971
.975
.979
.982
Value 01 k
0.14
0.276
.475
.620
.725
0.08
0.168
.308
.425
.521
.602
.669
.725
.771
.809
.842
.868
.890
.909
.924
.937
.943
.956
.964
.970
.975
.979
.983
.986
.988
.990
0.15
0. 292
.499
.645
.749
0.09
0.187
.339
.463
.563
.645
.712
.766
.809
.845
.874
.898
.917
.932
.945
.955
.964
.970
. 976
.981
.984
.987
.990
.991
.993
.994
0. 16
0.308
.521
.669
.771
0. 10
0. 206
.369
.499
.602
.684
.749
. 800
.842
.874
.900
.921
. 937
.950
.960
.968
.975
.980
.984
.987
.990
.992
.994
.996
.997
0. 17
0.324
.543
.691
.791
/C -57
-------
Estimation of k and L
Theriault Table
Period of Value of k
incubation
,. . 0.11 0.12 0.13 0.14 0.15 0.16
(days)
5 .718 .749 .776 .800 .822
67 01 fiflQ QQ__L fl R (^ P^ A
........ ^ IO1 • OUo • OOrk • ODD * OI *±
7— .830 .855 .877 .895 .911 .924 .935
8- .868 .890 .909 .924 .937 .948 .956
9 .898 .917 .932 .945 .955 .964 .970
10-------- .921 .937 .950 .980 .968 .975 .980
U .938 .952 .963 .971 .978 .983 .987
12 — .952 .964 .972 .979 .984 .988 .991
13 .963 .972 .980 .985 .989 .992 .994
14 .971 .979 .985 .989 .992 .994 .996
15- .978 .984 .989 .992 .994 .996 .997
16 — .983 .988 .992 .994
17 .987 .991 .994 .996
18 .990 .993 .995 .997
19 .992 .995 .997 .998
2Q .994 .996 .997 .998 —-
Period of" ~~~~~ Value of k
ln°"bat*on 67T8oTTi6726 oT2i Ol 57235724 oTlT
(days)
1 0.339 0.354 0.369 0.383 0.397 0.411 0.425 0.438
2 .563 .583 .602 .620 .637 .653 .669 .684
3 .712 .731 .749 .766 .781 .796 .809 .822
4 .809 .826 .842 .855 .868 .880 .890 .900
5 .874 .888 .900 .911 .921 .929 .937 .944
6 .917 .928 .937 .945 .952 .958 .934 .968
7 .945 .953 .960 .966 .971 .975 .979 .982
8- .964 .970 .975 .979 .983 .986 .988 .990
9— .976 .981 .984 .987 .990 .991 .993 .994
10 .984 .987 .990 .992 .994 .995 .996 .997
U .990 .992 .994 .995
12 .993 .995 .996 .997
13 .995 .997 .997 .998
14__ .997 .998 .998 .999
15 .998 .999 .999 .999
-------
Estimation of k arid L
RATES OF B.O.D. SATISFACTION
FOR VALUES OF K. FROM .05 TO.60
100
80
60
40
8 =
'\JL \ V-\ T~T\"~V \~ I "\"" ""'"Xl ~' - ^-L ' J t ^i j - {
8 10 12 I4 '6 i8 20
-------
SOURCES AND ANALYSIS OK ORGANIC' NITROGEN
I INTRODUCTION
A Organic nitrogen refers to the nitrogen in
combination with any organic radical.
For sanitary engineering the main interest
is the nitrogen contained in proteins,
peptides, amines, amino acids, amides
and related compounds of animal or
vegetable origin.
II SOURCES OF ORGANIC NITROGEN
A Natural Origin
1 Dead animal and plant residues
2 Animal wastes - urea, feces
3 Autotrophic organisms - algae, s.bact.
4 Heterotrophic organisms
B Industrial Origin
1 Food processing wastewater - meat,
milk, vegetables
2 Pharmaceutical wastes, antibiologicals
3 Plastics - polyamides, nitriles
4 Chemical intermediates or products
5 Dye industry - azo, nitro
III TOTAL KJELDAHL NITROGEN
PROCEDURE
A Organic nitrogen is determined using the
Total Kjeldahl Nitrogen (TKN) method.
This determination includes both organic
nitrogen and free ammonia. By dis-
tilling the free ammonia off the sample
before the determination, organic nitrogen
can be determined directly.
B Scope
The procedure converts nitrogen compon-
ents of biological origin such as amino
CH.N.Sa. 11.75
acids, proteins and peptides to ammonia
but it may not convert the nitrogenous
compounds of some industrial wastes
such as amines, nitro compounds,
hydrazones, oximes, semi-carbazones
and some refractory tertiary amines.
C Method Summary
The sample is heated in the presence of
a concentrated sulfuric acid-potassium
sulfate-mercuric sulfate mixture and
evaporated until sulfur trioxide fumes are
obtained and the solution becomes color-
less or pale yellow. The residue is
cooled and diluted, then treated and made
alkaline with a hydroxide-thiosulfate
solution. The ammonia is distilled off
and then determined either by nessler-
ization or by titration.
1 Nesslerization (colorimetric) is the
method used when the ammonia-
nitrogen concentration is less than 1. 0
mg N/liter.
2 For ammonia-nitrogen concentration?
above 1. 0 mg N /liter, the ammonia i=?
determined by titration with 0. 02 N
H SO in the presence of a mixed
indicator.
For a detailed description of the
procedure and reagent preparation,
consult the EPA methods manual.
D Precision and Accuracy
1 Thirty-one analysts in 20 laboratories
used the Total Kjeldahl Nitrogen
procedure to analyze natural water
samples containing the following increments
of organic nitrogen: 0. 20, 0. 31, 4. 10
and 4.61 mg N/liter.
a Precision results for a standard
deviation were 0.197, 0.247, 1.056
and 1. 191 mg N/liter, respectively.
b Accuracy as bias was +0. 03, +0. 02,
+0. 04 and -0. 08 mg N/liter, respect-
ively.
11-1
-------
Sources and Analysis uf Organic Nitrogen
2 The nature and composition of ex-
traneous materials affect analytical
recovery. High salt concentrations
may raise digestion temperature.
High concentrations of organic sample
components may require excessive acid
during digestion tending to low nitrogen
yield.
3 The digestion temperature is critical.
380 to 390°C usually gives high
analytical recovery on the more re-
fractory nitrogen compounds of natural
origin. Nitrogen losses occur above
42
-------
AMMONIA, NITRITES AND NITRATES
I SOURCES AND SIGNIFICANCE OF
AMMONIA, NITRITES AND NITRATES
IN WATER
The natural occurrence of nitrogen com-
pounds is best demonstrated by the nitrogen
cycle (Figure 1).
A Ammonia
1 Occurrence
Ammonia is a product of the micro-
biological decay of animal and plant
protein. In turn, it can be used
directly to produce plant protein.
Many fertilizers contain ammonia.
2 Significance
The presence of ammonia nitrogen
in raw sura surface waters might indicate
domestic pollution. Its presence in
waters used for drinking purposes
may require the addition of large
amounts of chlorine in order to
produce a free chlorine residual.
The chlorine will first react with
ammonia to form chloramines be-
fore it exerts its full bactericidal
effects (free chlorine residual).
B Nitrites
1 Occurrence
Nitrite nitrogen occurs in water
as an intermediate stage in the
biological decomposition of organic
nitrogen. Nitrite formers (nitro-
somonas) convert ammonia under
aerobic conditions to nitrites. The
bacterial reduction of nitrates can
also produce nitrites under anaero-
bic conditions. Nitrite is used as
a corrosion inhibitor in industrial
process water.
2 Significance
Nitrites are usually not found in
surface water to a great extent.
Tiie presence of large quantities
indicates a source of wastewater
pollution.
Nitrates
Occurrence
Nitrate formers convert nitrites
under aerobic conditions to nitrates
(nitrobacter). During electrical
storms, large amounts of nitrogen
(No) are oxidized to form nitrates.
Finally, nitrates can be found in
fertilizers.
Significance
Nitrates in water usually indicate
the final stages of biological sta-
bilization. Nitrate rich effluents
discharging into receiving waters
can, under proper environmental
conditions, cause stress to stream
quality by producing algal bloomi-.
Drinking water containing exces-
sive amounts of nitrates can cause
infant methemoglobinemia.
II PRESERVATION OF AMMONIA, NITRATE
AND NITRITE SAMPLES<8)
A If the sample is to be analyzed for Ammom;.
or Nitrate, it may be preserved with 2rr,I
concentrated sulfuric acid per liter and
cooled to 4°C.
B For nitrite, cool to 4CC and analyze as soon
as possible.
C Mercuric chloride is effective as a preserv-
ative but its use is discouraged because:
1 The Hg ion interferes with some of the
nitrogen tests.
2 The Hg presents a disposal problem.
D Even when "preserved", conversion from one
nitrogen form to another may occur. Samples
should be analyzed as soon as possible.
CH.N.6e. 11.75
11-3
-------
Ammonia, Nitrites and .Nitrates
PLANT
PROTEIN
ANICN
FOOOSTUFF
THE NITROGEN CYCLE
Figure 1
11-4
-------
Ammonia, Nitrites and Nitrates
111 l)ET?:i
1 Reaction
The sample is distilled in
the presence of a borate
buffer at pH 9. 5
NH
H+
H+ + Na2 B4O7
Buffer
pH 9. 5 maintained
The ammonia in the dis-
tillate is then measured by
either of two techniques.
1)
2)
Nesslerization is used
for samples containing
less than 1 mg/1 of
ammonia nitrogen.
Absorption of NH3 by
boric acid and back ti-
trationwith a standard
strong acid is more
suitable for samples
containing more than
1 mg NH3 N/l.
NH3 +HBO2-
BO2"
H
HB0
Methylene Blue
pH 7.8 - 8.0
Green
pH 6. 8 - 7.0
Purple
11-5
-------
Ammonia, Nitrites and Xitrates
2 Interferences
a Cynate may hydrolyze, even at
pH 9. 5.
b Volatile organics may come over in
the distillate, causing an off-color
for Nesslerization. Aliphatic and
aromatic amines cause positive
interference by reacting in the acid
titration. Some of these can be
boiled off at pH 2 to 3 prior to
distillation.
c Residual chlorine must be removed
by pretreatment with sodium
thiosulfate.
d If a mercury salt was used for
preservation, the mercury ion
must be complexed with sodium
thiosulfate (0. 2 g) prior to distil-
lation.
3 Precision and Accuracy*")
Twenty-four analysts in sixteen
laboratories analyzed natural water
samples containing the following
amounts of ammonia nitrogen:
0.21, 0.26, 1.71, andl.92mg
NH3-N/liter.
a Precision
The standard deviation was: 0. 122,
0.070, 0.244, and 0. 279 mg
NH3-N/liter, respectively.
b Accuracy
The bias was: -0.01, -0,05.
+ 0.01, and -0.04 mg NH3-N/liter,
respectively.
C Selective Ion Electrode(8)
1 Principle
A hydrophobic, gas-permeable mem-
brane is used to separate the sample
solution from an ammonium chloride
internal solution. The ammonia in
the sample diffuses through the
membrane and alters the pll of tin-
internal solution, which is sensed by
a pH electrode. The constant level ot'
chloride in the internal solution is
sensed by a chloride selective ion
electrode which acts as the reference
electrode.
2 Interferences
a Volatile amines are a positive
interference.
b Mercury forms a complex with
ammonia so it should not be used
as a preservative.
3 Precision and Accuracy
In a single laboratory (EPA) four surface
water samples were analyzed containing
the following amounts of ammonia
nitrogen: 1.00, 0.77, 0.19, and0.13mg
NH3-N/liter.
a Precision
The standard deviations were 0.038,
0.017, 0.007, and 0.003 nig
NHg-N/liter, respectively.
b Accuracy
The % recovery on the 0. 19 and 0. 13
concentrations was 96% and 91%
respectively.
IV DETERMINATION OF NITRITE
A Diazotization' '
1 Reaction
a Under acid conditions, nitrite
ions react with sulfanilic acid to
form a diazo compound.
b The diazo compound then couples
with o-naphthylamine to form an
intense red azo dye which exhibits
maximum absorption at 540 nm.
11-6
-------
Ammonia, Nitrites and! Nitrates
NH,,
N^= N
OH
+ NO -i- 2H ••
pH1.4
HSO
•*• 2H^
'3 ""3*
SULFANILICACID DIAZO COMPOUND
N— — N
OH
H+ + NOl
H,,0
2
SO H
b In alkaline solution the nitro
derivative rearranges to
form & yellow -colored com -
pound which exhibits maxi-
mum absorption at 410
+ 3KOH
OK
2 Interferences
COLORLESS
YELLOW
a There are very few known
interferences at concentrations
less than 1000 times that of
nitrite.
b High alkalinity (greater than
600 mg/liter) will give low results
due to a shift in pH.
3 Precision
On a synthetic sample containing
0. 25 mg nitrite nitrogen/1, the ARS
Water Minerals Study (1961) re-
ported 125 results with a standard
deviation of ±0. 029 mg/1.
V DETERMINATION OF NITRATE
/C \
A Phenoldisulfonic AcidVD'
1 Reaction
a Phenoldisulfonic a c i d re-
acts with nitrate to produce
a nitro derivative.
2 Interferences
Chloride ion under the acid
conditions of the test intro-
duces negative interference.
6C1' + 2NO3~ + 8H+
3C12 f + 2NO f + 4H20
Silver sulfate can be used
to precipitate Cl", but due
to incomplete precipitation
of Ag+, an offcolor or
turbidity is produced when
the final color is developed.
(Note: This difficulty can
be overcome by using
NH4OHas the alkali.)
Nitrites in concentrations
greater than 0.2 mg N/l
introduce positive interfer-
ence. However, in most
waters, the concentration
H-7
-------
Ammonia, Nitrites and Nitrates
of nitrite is insignificant
compared to nitrate.
c Color and turbidity may be
removed by using A1(OH>3
suspension or by floccula-
tion with ZnSO4 and alkali.
3 Precision and accuracy
On a synthetic sample con-
taining 1. 0 mg nitrate N/l,
and 200 mg Cl'/l. the ARS
Water Nutrients Study (1966)
reported 46 results with a
standard deviation of 10.399
mg/1 and a mean error of
-0.31 mg/1.
B Brucine Sulfate(6> 7'8)
1 Reaction
Brucine, a strychnine compound
reacts with nitrate to form a
yellow compound which exhibits
maximum absorption at 410 nm.
The reaction according to t h P
procedure as outlined in
Standard Methods does not
follow Beer's Law. However,
a recent modification by Jenkins
andMedsker (2) has been devel-
oped. Conditions are controlled
in the reaction so that Beer's Law
is followed and concentrations
below 1 mg nitrate N/l can be
determined.
2 Interferences
a Nitrite may react the same
as nitrate but can be elim-
inated by the addition of sul-
fanilic acid to the brucine
reagent.
b Organic nitrogen compounds
may hydrolyze and give
positive interference at low
(less than 1 mg/1) nitrate
nitrogen concentrations.
c Residual chlorine may be
eliminated by the addition of
sodium arsenite.
3 Preci.sii'ii umt A' rut-Be
(8)
a Twenty seven analysts in 15 lab-
oratories analyzed in natural
water samples containing the
following increments of inorganic
nitrate: 0.16, 0.19, 1.08 and
1.24 mgN/liter.
b Precision result s as standard
deviation were 0. 092, 0. 083,
0.245, and 0.214 mg N/liter
respectively.
c Accuracy expressed as bias was
-0.01, +0.02, +0.04 and+0.04
mg N/liter, respectively.
C Cadmium Reduction*6- 8>
1 Reaction
A filtered sample is passed through
a column containing granulated
copper-cadmium to reduce nitrate to
nitrite. The nitrite (that originally
present plus reduced nitrate) is deter-
mined by diazotizing with sulfanilamide
and coupling with N-(l-naphthyl)-
ethylenediamine dihydrochloride to
form an intensely red azo dye which
is measured spectrophotometrically.
To obtain the value for only nitrate,
more of the filtered sample is tested
using the same colorimetric reaction
but without passing it through the
reduction column. The resulting value
represents the nitrite originally present
in the sample. Subtracting this nitrite
value for the non-reduced sample from
the nitrate + nitrite value for the re-
duced sample gives the value for nitrate
originally present in the sample.
2 Interferences
a Build-up of suspended matter in the
reduction column will restrict
sample flow. Filtration or floc-
culation with zinc sulfate should
remove turbidity.
11-8
-------
Ammonia, Niu lies and Nitrates
b High concentrations of iron,
copper or other metals may
interfere. EDTA may be used
to complex these.
c Large concentrations of oil and
grease in a sample can coat the
surface of the cadmium. Pre-
extracting the sample with an
organic solvent removes oil and
grease.
3 Precision and accuracy^ •
In 11 laboratories, three samples
were analyzed containing the follow-
ing amounts of nitrate nitrogen-
0.05, 0. 5, and 5 mg
NO3-N/liter.
a Precision
The relative standard deviation
was 96.4%, 25.6%, and 9.2%,
respectively.
b Accuracy
The relative error was 47. 3%,
6.4%, and 1.0%, respectively.
4 Automated cadmium reduction
The EPA Methods Manual^8) contains
details for the automated procedure.
D Hydrazine Reduction
A method using hydrazine to reduce
nitrate to nitrite followed by subsequent
measurement of nitrite by diazotizatioii
was reported by Fishman, et al. W
The means to determine nitrate is the
same as above in the Cadmium Reduction
Method. Subtraction of nitrite (deter-
mined from non-reduced sample) from
the total nitrite (reduced nitrate +
original nitrite) will give the original
nitrate nitrogen concentration.
The procedure was adapted to the Auto
Analyzer by Kamphake, et al. (3) and
can be found in detail in the 1971 edition
of the EPA Methods Manual.(8)
REFERENCES
1 Fishman, Marvin,]., Skougstad, Marvin
W , and Scarbio, George, .1 r
Diazotization Method for Nitrate and
Nitrite. JAWWA 56:633-638 May, I960
2 Jenkins, David and Medsker, Lloyd L
Brucine Method for Determination of
Nitrate in Ocean, Estuarine and Fresh
Waters. Anal Chem 36:610-612
March, 1964.
3 Kamphake, L. J., Hannah, S. and Cohen, J.
Automated Analysis for Nitrate by
Hydrazine Reduction. Water Research,
1, 205. 1967.
4 Lishka, R. J. , Lederer, L. A., and
McFarren, E. F. Water Nutrients
No. 1, Analytical Reference Service
1966.
5 Sawyer, Clair N, Chemistry for Sanitary
Engineers. McGraw-Hill Book Co. ,
New York. 1960.
6 Standard Methods for the Examination of
Water and Waste Water. APHA. AWWA.
WPCF. 13th Ed. 1971.
7 ASTM Book of Standards, Part 23, 1970.
8 Methods for Chemical Analysis of Water &
Wastes, EPA-MDQARL, Cincinnati,
Ohio 45268. 1974.
This outline was prepared by B. A. Punghorst,
former Chemist, and C. R. Feldmann,
Chemist, and revised by A. E. Donahue,
Chemist. National Training Center, MPOD,
OWPO, USEPA, Cincinnati, Ohio 45268.
Descriptors: Ammonia, Chemical Analysis,
Nitrates, Nitrites, Nitrogen, Nitrogen
Compounds, Nitrogen Cycle, Nutrients, Water
Analysis, Water Pollution Sources
11-9
-------
TOTA L CARBON ANALYSIS
I INTRODUCTION
A History of Carbon Analyses
In the wake of a rapid population, growth,
and the increasing heavy use of our
natural waterways, the nation, and indeed
the world, is presented with the acute
problem of increased pollutional loads on
streams, rivers and other receiving
bodies. This has resulted in a growing
awareness of the need to prevent the
pollution of streams, rivers, lakes and
even the oceans. Along with this aware-
ness has developed a desire for a more
rapid and precise method of detecting and
measuring pollution due to organic
materials.
B The Methods
In the past, two general approaches have
been used in evaluating the degree of
organic water pollution.
1 The determination of the amount of
oxygen or other oxidants required to
react with organic impurities.
2 The determination of the amount of
total carbon present in these impurities.
C Oxygen Demand Analyses
The first approach is represented by
conventional laboratory tests for deter-
mining Chemical Oxygen Demand (COD)
and Biochemical Oxygen Demand (BOD).
One of the principal disadvantages of these
tests is that they are limited primarily
to historical significance, that is, they
tell what a treatment plant had been doing,
since they require anywhere from two
hours to five days to complete. Since up
to now no faster method has been
available, traditional BOD and COD
determinations have become accepted
standards of measure in water pollution
control work even though they are
essentially ineffective for process
control purposes.
Until the introduction of the Carbonaceous
Analyzer, all methods taking the second
approach, the total carbon method of
evaluating water quality, also proved
too slow.
II THE ANALYSIS OF CARBON
A Pollution Indicator
Now the carbonaceous analyzer provides
a means to determine the total carbon
content of a dilute water sample in
approximately two minutes. With proper
sample preparation to remove inorganic
carbonates, the instrument determines
the total organic carbon content in the
sample.
B Relationship of Carbon Analysis to BOD
and COD
This quantity varies with the structure
from 27 percent for oxalic acid through
40 percent for glucose to 75 percent for
methane. The ratio of COD to mg carbon
also varies widely from 0.67 for oxalic
acid through 2.67 for glucose to 5. 33 for
methane. Representative secondary
sewage effluents have given a ratio of
COD to carbon content of between 2. 5 and
3. 5 with the general average being 3.0.
The BOD, COD and carbon contents of
these and some other representative
compounds are summarized in the follow-
ing table.
OH. MET. 24a. 11.75
12-1
-------
Total Carbon Analysis
Sample
Sioaric Acid - C -H^O™
Glucose - C-H. 9Or
G 12 6
Oxalic Acid - C2H2O4
Benzoic Acid - C,HCO0
I D i
Phenol - CgHgO
Potassium Acid Phthalate
KHC.H O4
Salicylic Acid - C7HgO3
Secondary Effluent, Clarified
IT (1
It ft ft
5 -Day
BOD-mg/mg
.786
.73
. 14
1.38
. 05 to 2. 1 de-
pending upon
concentration
.95
1. 25
13*
23*
4*
COD-
mg/mg
2.91
1.07
.18
1.97
2.36
1. 15
1.60
75*
67*
36*
ro Carbon
76
40
27
69
77
47
61
21*
12*
7*
* In units of mg/1
III THE CARBON ANALYZER
A Principle of Operation
Basically the carbonaceous analyzer con-
sists of three sections - a sampling and
oxidizing system, a Beckman Model 315
Infrared Analyzer, and a strip-chart
recorder.
C*rbon*ceou> Anal7*er Schematic
a carrier from cylinder
A micro sample (20-40 fjtl) of the water to be
analyzed is injected into a catalytic com-
bustion tube which is enclosed by an electric
furnace thermostated at 950°C. The water
is vaporized and the carbonaceous material
is oxidized to carbon dioxide (CO2) and
steam in a carrier stream of pure oxygen.
The oxygen flow carries the steam and
CO2 out of the furnace where the steam is
condensed and the condensate removed.
The CO2, oxygen and remaining water vapor
enter an infrared analyzer sensitized to
provide a measure of CO2- The output of
the infrared analyzer is recorded on a strip
chart, after which, the curve produced can
be evaluated by comparing peak height with
a calibration curve based upon standard
solutions. Results are obtained directly in
milligrams of carbon per liter.
Application
(D
Results show that the method is applicable
for most, if not all, water-soluble organic
compounds -- including those that contain
sulfur, nitrogen, and volatiles.
Nonvolatile organic substances can be
differentiated from volatiles, such as
carbon dioxide or light hydrocarbons by
12-2
-------
Total Carbon Analysis
determination of carbon both before and
after the sample solution has been blown
with an inert gas.
C Sample Preparation
The Carbonaceous Analyzer is often
referred to as a total carbon analyzer
because it provides a measure of all the
carbonaceous material in a sample, both
organic and inorganic. However, if a
measure of organic carbon alone is de-
sired, the inorganic carbon content of
the sample can be removed during sample
preparation.
1 Removal of inorganic carbon
The simplest procedure for removing
inorganic carbon from the sample is
one of acidifying and blowing. A few
drops of HC1 per 100 ml of sample
will normally reduce pH to 2 or less,
releasing all the inorganic carbon as
CC>2. Five minutes of blowing with
a gas free of CC>2 sweeps out the CC>2
formed by the inorganic carbon. Only
the organic carbon remains in the
sample and may be analyzed without
the inorganic interference.
2 Volatile carbonaceous material
Volatile carbonaceous material that
may be lost by blowing is accounted
for by using a dual channel carbon
analyzer. Beckman's new analyzer
has the previously detailed high
temperature (950°C) furnace plus a
low temperature (150°C) one. Using
quartz chips wetted with phosphoric
acid, the low temperature channel
senses only the CO2 (freed by the
acid) in the original sample. The
remaining organics and water are
retained in the condenser connected
to this low temperature furnace.
None of the organics are oxidized by
the 150°C furnace.
By injecting a sample into the low
temperature furnace, a peak repre-
senting the inorganic carbon is ob-
tained on the strip chart. Injecting
a nonacidified sample into the high
temperature furnace yields a peak
representing the total carbon. The
difference between the values deter-
mined for the two peaks is the total
organic carbon.
3 Dilute samples
If the sample is dilute (less than 100
mg/liter carbon) and is a true solution
(no suspended particles) no further
preparation is required.
4 Samples containing solids
If the sample contains solids and/or
fibers which are to be included in the
determination, these must be reduced
in size so that they will be able to pass
through the needle which has an opening
of 170 microns (needles having larger
openings may be obtained if necessary).
In most cases, mixing the sample in a
Waring Blender will reduce the particle
size sufficiently for sampling.
IV PROCEDURE FOR ANALYSIS
A Interferences
Water vapor resulting from vaporization of
the sample, causes a slight interference in
the method. Most of the water is trapped
out by the air condenser positioned immed-
iately after the combustion furnace. How-
ever, a portion of the water vapor passes
through the system into the infrared de-
tector and appears on the strip chart as
carbon. The water blank also appears on
the standard calibration curve, and is
therefore removed frorn the final calcu-
lation. In tests of solutions containing the
following anions: NO3, Cl", SO'2, PO^,
no interference was encountered with con-
centrations up to one percent.
B Precision and Accuracy
The recovery of carbon from standard
solutions is 98.5 - 100.0 percent. The
minimum detectable concentration using
the prescribed operating instructions is 1
mg/1 carbon. Generally, the data are
reproducible to + 1 mg/1 with a standard
deviation of 0.7 mg/1 at the 100 mg/1 level.
12-3
-------
Total Carbon Analysis
V APPLICATIONS
Several of the many research and industrial
applications of .ho Carbonaceous Analyzer
are listed below:
A Determine the efficiency of various waste -
water renovation processes, both in the
laboratory and in the field.
B Compare a plant's waste outlet with its
water inlet to determine the degree of
contamination contributed.
C Monitoring a waste stream to che.ck for
product lose.
D Follow the rate of utilization of organic
nutrients by micro-organisms,
E To detect organic impurities in inorganic
compounds.
VI ADVANTAGES OF CARBON ANALYZER
A Speed
The Carbonaceous Analyzer's most
important advantage is its speed of
analysis. One analysis can be performed
in 2-3 minutes. This speed of analysis
brings about another advantage, economy
of operation. Working with dilute samples,
one man can run ten to twenty carbon
determinations per hour. This is probably
more than the number of COD or BOD
tests that can even be started, much less
completed, in the same period of time.
B Total Carbon
Another advantage is that the measure of
carbon is a total one. The oxidizing
system of the analyzer brings about com-
plete oxidation of any form of carbon. No
compound has been found to which the
method is inapplicable.
VII CONCLUSIONS
The Carbonaceous Analyzer provides a
rapid and precise measurement of organic
carbon in both liquid and air samples. It
should be found useful for many research
and industrial applications, a few of which
have been mentioned.
Because of its rapidity it may be found more
useful than the more time-consuming BOD
and COD measurements for monitoring
industrial waste streams or waste treatment
processes.
REFERENCES
1 Van Hall, C. E. , Safranko. John and
Stenger, V. A. Anal. Chem. 35,
315-9. 1963.
2 Van Hall, C. E., and Stenger, V. A.
Draft of Final Report - Phase I - Con-
tract PH 86-63-94, Analytical Research
Toward Application of the Dow Total
Carbon Determination Apparatus to the
Measurement of Water Pollution.
3 Van Hall, C. E. . Stenger, V. A.
Beckman Reprint - R6215. Taken from
Paper Presented at the Symposium on
Water Renovation, Sponsored by the
Division of Water and Waste Chemistry,
ACS in Cincinnati. Jan. 14-16, 1963.
This outline was prepared by Robert T.
Williams, Chief, and revised by Charles J.
Moench, Jr., Waste Identification and
Analysis Section, MERL, USE PA, Cincinnati,
Ohio 45268.
Descriptors: Biochemical Oxygen Demand,
Carbon, Chemical Analysis, Chemical Oxygen
Demand, Organic Matter, Organic Wastes,
Water Analysis, Instrumentation, Nutrients
12-4
-------
DETERMINATION OF CALCIUM AND MAGNESIUM HARDNESS
I INTRODUCTION
A Definition of Hardness
USPHS - "in natural waters, hardness
is a characteristic of water which re-
presents the total concentration of just
the calcium and magnesium ions .ex-
pressed as calcium carbonate. If
present in significant amounts, other
hardness-producing metallic ions should
be included. "
B Other Definitions in Use
1 Some confusion exists in understanding
the concept of hardness as a result of
several definitions presently used.
2 Soap hardness definition includes hy-
drogen ion because it has the capacity
to precipitate soap. Present definition
excludes hydrogen ion because it is not
considered metallic.
3 Other agencies define hardness as
"the property attributable to presence
of alkaline-earths".
4 USPHS definition is beat in relation
to objections of hardness in water.
II CAUSES OF HARDNESS IN WATERS OF
VARIOUS REGIONS OF THE U. S.
A Hardness will vary throughout the country
depending on:
1 Leaching action of water traversing
over and through various types of
geological formations.
2 Discharge of industrial and domestic
wastes to water courses.
3 Uses of water which result in change
in hardness, such as irrigation and
water softening process.
B Objections to Hardness
1 Soap-destroying properties
2 Scale formation
C Removal and Control
Hardness may be removed and controlled
through the use of various softening oper-
ations such as zeolite, lime-soda, and
hot phosphate processes. It can also be
removed by simple distillation or com-
plex formation with surface active agents
(detergents).
Ill DETERMINATION OF HARDNESS
/c \
A Two Methods in Use
1 Compleximetric method (EDTA)
2 Calculation - by the use of appro-
priate factors, the hardness due
to ions other than calcium and
magnesium may be converted to
an equivalent calcium carbonate
hardness.
B Compleximetric Method
1 Principle of determination
Ethylenediaminetetra acetic acid
(EDTA) is a sparingly soluble ammo
polycarboxylic acid which forms
slightly ionized and very stable color-
less complexes with the alkaline-
earth metals.
2 Interferences: iron, manganese,
nickel, and zinc.
3 Procedure: Time and pH considera-
tions.
4 Calculations of total hardness
assuming a known volume of titrant
of EDTA.
5 Precision and accuracy.
CH.HAR.3c. 11.75
13-1
-------
Determination of Calcium ind Magnesium Hardness
C Determination of Calcium Hardness
1 Principle of determinations
Murexide indicator forms a salmon-
colored complex with calcium whose
ionization constant is of a higher
value than that of the Ca EDTA complex.
2 Interferences: Heavy metals and Sr.
3 Procedure: Time of tritation and
proper lighting conditions are critical
factors.
4 Calculation of Ca hardness.
5 Precision and accuracy.
D Determination of Magnesium Hardness
1 Calculation by difference method
most commonly used.
2 Equivalent of Ca hardness is sub-
tracted from total hardness equiva-
lent!, the difference attributable to
magnesium equivalents.
3 Other methods such as pyrophosphate
method, where calculation by differ-
ence method cannot be used.
IV ANALYTICAL QUALITY CONTROL
METHODS
A The Environmental Protection Agency,
Analytical Quality Control Laboratory
has published a manual titled Methods
for Chemical Analysis of Water and
Wastes. 1974.
B
The procedure for the determination
of hardness, as discussed in this
manual, was excerpted from the
13th ed. of Standard Methods*4'
and part 23 of ASTM. <5>
REFERENCES
1 Barnard, A. J. Jr., Broad, W. C. and
Flaschka, H. The EDTA Titration.
J. T. Baker Company. 1957.
2 U. S. Public Health Service. Drinking
Water Standards. USPHS Report,
Volume 61. No. 11, 1946.
3 Rainwater, R. H. and Thatcher, L. L.
Methods for the Collection and Anal-
ysis of Water Samples. U. S. Geo-
logical Survey Water Supply, Paper
1454. 1960.
Standard Methods for the Examination
of Water and Wastewater, 13th ed.
APHA, AWWA, WPCF, Method
122B, 1971.
ASTM Standards, Part 23, Method
D1126-67, 1973.
Methods for Chemical Analysis of
Water and Wastes, EPA-MDQARL.
Cincinnati, OH, 1974.
This outline was prepared by B. V. Salotto,
Research Chemist, Waste Identification and
Analysis Section, MERL. EPA, and revised
by C. R. Feldmann, Chemist, National
Training Center, MPOD, OWPO, USEPA,
Cincinnati, OH 45268.
Descriptors: Calcium, Chemical Analysis,
Hardness, Magnesium, Water Analysis
13-2
-------
SOLIDS RELATIONS IN POLLUTED WATER
I MPN, oxygen demand, and solids have
been major water pollution control criteria
for many years. This discussion ie con-
cerned primarily with solids and their inter-
relations with oxygen demand.
A Engineered treatment or surface water
self-purification depends upon:
1 The conversion of soluble or colloidal
contaminants into agglomerated masses
that may be separated from the-water.
2 Oxidation of putrescible components
into stable degradation products.
3 Item 1 is the major concern in most
treatment systems because item 2
requires a greater investment In time,
manpower and capital costs.
B Stress on oxygen demand removal fre-
quently results in an unduly small amount
of attention to the contribution of solids
in the oxygen demand picture.
1 Oxygen demand formulations generally
are specified to be applicable in the
absence of significant deposition.
2 Increasing impoundment, tidal estuaries,
and Incomplete solids removal, gener-
ally ensure that solids deposition will be
significant.
3 The BOD test stresses the fraction of
oxygen demand that is exerted relatively
rapidly. It includes only the fraction of
unstable material that is exerted under
test conditions - usually short term un-
der aerobic conditions.
4 Contributions of a bed load of solids to
oxygen demand frequently are incom-
pletely recognized because they have a
more local effect, are difficult to measure,
tend to move, and are incompletely un-
derstood. The onset of an aerobic action
in deposited solids increases hydrolysis
rate. High molecular weight cell mass
splits into small molecules that are
soluble and more available for high
rate oxidation upon feedback into oxy-
gen containing water.
II A given wastewater may have several
forms of solids in changeable proportions
with time and conditions.
A Solids may be classified among the charted
forms according to biological chemical
or physical properties. It is not possible
to precisely classify a given material into
any one form because they usually are
mixtures that may include or be converted
into other forms.
B Interrelationships are indicated by diagram
in Figure 1. Some changes occur more
readily than others.
1 Settleable solids generally consist of
a mixture of organic, inorganic, en-
trained dissolved or colloidal solids
with living and dead organisms.
a They may be hydrolyzed into smaller
sizes to aquire colloidal or
dissolved characteristics.
b They may be converted into a larger
fraction of living material or vice
versa.
2 Colloidal solids also are likely to be a
mixture of organic and inorganic
materials, living or dead containing
associated dissolved materials.
a They are likely to agglomerate to
form settleable masses with time
or changing conditions.
b Chemical reactions may solubilize
the colloidal masses for recombina-
tion into other forms.
3 Dissolved solids are most readily avail-
able of all forms for biological, chemi-
cal or physical conversion.
PC.6b. 11.75
14-1
-------
Solids Relations in Polluted \Vaier
Figure 1. SOLIDS INTERRELATIONSHIPS IN WATER
a They may be assimilated into cell
mass to become colloidal during a
state of rapid growth or form settle -
able masses during slow growth.
Surface phenomena are likely to en-
courage inclusion of other forms of
solids with cell mass.
Stabilization occurs with each conversion
because energy is expended with each
change.
1 Biological changes involve oxidation to
obtain enough energy to synthesize cells.
Products consist of cell mass and
degradation products.
Oxidation tends toward production
of CO2, H2O, NOg, SO^^etc. CO2
has limited water solubility and
partially leaves the aqueous environ-
ment. More soluble constituents
tend to remain with it for recycle.
Biological residues tend to recycle.
Eventually the mass consists of
relatively inert and largely insolu-
ble residues. These make up the
14-2
-------
Solids Relations in Polluted Water
bed load that may decompose at a
rate of less than 1% per day, but
accumulates in mass until it be-
comes dominant in water pollution
control activities.
Ill Treatment is an engineered operation
designed to utilize events in surface water
self-purification in a smaller package in
terms of space and time. Effective treat-
ment presupposes oxidation either in the
plant or in facilities to minimize feedback
of solids components to the aqueous
environment.
A Primary treatment consists of a separation
of floatable or settleable solids and re-
moval from the used water.
1 Prompt removal is essential to minimize
return of solubilized or leached materi-
als from the sludge mass.
2 Sludge and scum are highly putrescible
and difficult to drain.
3 Subsequent treatment prior to disposal
serves to enhance drainability, reduce
solids or volume for burial or burning.
B Secondary treatment generally involves
some form of aerobic biological activity
to oxidize part of the colloidal and dis-
solved contaminants and convert most of
the remainder into a settleable sludge.
1 Aerobic systems favor assimilation of
nutrients into cell mass at the expense
of part of the available energy repre-
sented by oxidation to products such as
CC>2 and water.
2 Cell mass and intermediate degradation
products are only partially stabilized
and tend to recycle with death and decay.
3 Rapid growth of cells tends to produce
sludges that are highly hydrated and
difficult to concentrate.
4 A compromise must be reached to pro-
duce a favorable balance among separa-
tion of solids and feedback of lysed
materials.
a Cell growth continues as long as
energy and nutrients permit.
b Under limiting nutrient conditions
the population may show a relatively
low rate overall dieoff but variety
is changing.
c Species most favored by the new con-
ditions tend to grow while others
die, lyse and release part of their
stored nutrients for subsequent use.
Anaerobic digestion of solids separated
during treatment is one process used to
increase solids stability and drainability
while reducing total volume or mass for
disposal.
1 Growth of cell mass is relatively slow
under anaerobic conditions while hy-
drolytic cleavage is relatively large
in comparison to that during aerobic
metabolism,
a Feedback to the aqueous environment
represents a significant fraction of
the input in the form of ammonia,
colloidal solids, low molecular
weight acids, and other products.
b Mass of the sludge is reduced by the
fraction of methane, carbon dioxide,
and other gases produced in process.
c Remaining solids tend to be more
concentrated, are lower in putresci-
bility and give up their water more
readily.
2 The liquid fraction of the products re-
maining after anaerobic digestion con-
tain nutrients, oxygen demand, solids,
and malodorous constituents that are
objectionable in surface waters if re-
leased without further aerobic
stabilization.
3 Digester liquids are much more con-
centrated than raw sewage and nutrition-
ally unfavorable, hence they are difficult
to treat and tend to shock aerobic
treatment processes.
14-3
-------
Solids Relations in Polluted Water
D Suitable disposal of solids resulting from
treatment operations takes various forms
of which the most desirable is to oxidize
it completely to gaseous products of oxi-
dation or inert ash. The objective is to
limit feedback into the aqueous environ-
ment and delay the charted interchanges.
1 Deposition of sludge as a cover on crop
land is effective providing surface
drainage is properly designed to limit
washoff.
2 Burial in areas above the flood plain
provides a suitable disposal,
3 Incineration under controlled con-
ditions to produce complete burning
of organics is effective but requires
close control.
a Incineration of digested sludge is
feasible but the digestion process
results in removal of part of the
heat content of the sludge as met-
hane carbon dioxide and water. Aux-
Dary heat may be required for wa-
ter evaporation if solids concentra-
tion is below 25%.
b Raw sludge incineration relieves
the auxilary heat problem but stor-
age may lead to odor problems. The
raw sludge may require chemical
treatment to enhance dew ate ring,
secondary sludges are more diffi-
cult to dewater and may require
centrifugation, floatation, or other
treatment to concentrate sludge for
heat balance purposes.
c Incineration requires close control
of complex equipment to maintain
acceptable environmental control.
Usually, after burners and off gas
scrubbers are necessary to avoid
air pollution.
IV ANALYTICAL PROCEDURES
A The Analytical Quality Control
Laboratory, Office of Water Programs,
Environmental Protection Agency has
published a manual titled. Methods for
Chemical Analysis of Water and Wastes,
1974.
B In this manual solids are classified into
four groups:
1 Residue, Total Filterable
2 Residue, Total Non-Filterable
3 Residue. Total
4 Residue, Volatile
C The procedure for the determination of
1,2, and 3 above is given in the EPA
Manual.
For the Volatile Residue procedure.
Standard Methods for the Examination
of Water and Wastewater, 13th Ed.,
page 536, Method 2243(1971) is the
reference cited by the EPA Manual.
This outline was prepared by F. J. Ludzack,
Chemist, National Training Center, MPOD,
OWPO, USEPA, Cincinnati, Ohio 45268.
Descriptors: Anaerobic Digestion, Oxygen,
Oxygen Demand, Sludge, Sludge Disposal,
Sludge Treatment, Solids, Waste Water
Treatment, Water Pollution Sources
14-4
-------
PHOSPHORUS IN THE AQUEOUS ENVIRONMENT
I Phosphorus is closely associated with
water quality because of (a) its role in
aquatic productivity such as algai blooms,
(b) its sequestering action, which causes
interference in coagulation, (c) the difficulty
of removing phosphorus from water to some
desirable low concentration, and (d) its
characteristic of converting from one to
another of many possible forms.
A Phosphorus is one of the primary nutrients
such as hydrogen (H), carbon (C)f
nitrogen (N), sulfur (S) and phosphorus (P).
1 Phosphorus is unique among nutrients
in that its oxidation does not contribute
significant energy because it commonly
exists in oxidized form.
2 Phosphorus is intimately involved in
oxidative energy release from and
synthesis of other nutrients into cell
mass via:
a Transport of nutrients across
membranes into cell protoplasm is
likely to include phosphoryiation.
b The release of energy for meta-
bolic purposes is likely to
include a triphosphate exchange
mechanism.
B Most natural waters contain relatively low
levels of P (0.01 to 0. 05 mg/1) in the
soluble state during periods of significant
productivity.
1 Metabolic activity tends to convert
soluble P into cell mass (organic P) as
a part of the protoplasm, intermediate
products, or sorbed material.
2 Degradation of cell mass and incidental
P compounds results in a feedback of
lysed P to the water at rates governed
by the type of P and the environment.
Aquatic metabolic kinetics show marked
influences of this feedback.
The concentrations of P in hydrosoils,
sludges, treatment plant samples and
soils may range from 102 to 106 times
that in stabilized surface water. Both
concentration and interfering compo-
nents affect applicability of analytical
techniques.
II The primary source of phosphorus in the
aqueous system is of geological origin.
Indirect sources are the processed mineral
products for use in agriculture, household,
industry or other activities,
A Agricultural fertilizer run-off is related
to chemicals applied, farming practice
and soil exchange capacity.
B Wastewaters primarily of domestic
origin contain major amounts of P from:
1 Human, animal and plant residues
2 Surfactants (cleaning agent) discharge
3 Microbial and other cell masses
C Waste*waters primarily of industrial
origin contain P related to:
1 Corrosion control
2 Scale control additives
3 Surfactants or dispersants
4 Chemical processing of materials
including P
5 Liquors from clean-up operations of
dusts, fumes, stack gases, or other
discharges
III Phosphorus terminology is commonly
confused because of the interrelations among
biological, chemical, engineering, physical,
and analytical factors.
CH.PHOS.4c. 11.75
15-1
-------
Phosphorua injthc Aqueous Environnu nt
Biologically, phosphorus may be available
as a nutrient, synthesized into living mass,
stored in living or dead cells, agglomerates,
or mineral complexes, or converted to
degraded materials.
Chemically, P exists in several mineral
and organic forms that may be converted
from one to another under favorable
conditions. Analytical estimates are
based upon physical or chemical techniques
necessary to convert various forms of P
into ortho phosphates which alone can be
quantitftted In terms of the molybdenum
blue colorimetric teat.
Engineering interest in phosphorus is
related to the prediction, treatment, or
control of aqueous systems to favor
acceptable water quality objectives.
Phosphorus removal is associated with
solids removal.
Solubility and temperature are major
physical factors in phosphorus behavior.
Soluble P is much more available than
insoluble P for chemical or biological
transformations and the rate of conversion
from one to another is strongly influenced
by temperature.
Table 1 includes a classification of the
four main types of chemical P and some
of the relationships controlling solubility
of each group. It is apparent that no
clear-cut separation can be made on a
solubility basis as molecular weight,
substituent and other factors affect
solubility.
Table 2 includes a scheme of analytical
differentiation of various forms of P
based upon:
1 The technique required to convert an
unknown variety of phosphorus into
ortho P which is the only one quanti-
tated by the colorimetric test.
Solubility characteristics of the sample
P or more precisely the means required
to clarify the sample.
a Any clarification method is subject
to incomplete separation. Therefore,
it is essential to specify the method
used to interpret the yield factor of
the separation technique. The
degree of separation of solubles
and insolubles will he significantly
different for:
1 Membrane filter separation
(0.5 micron pore size)
2 Centrifugation (at some specified
rpm and time)
3 Paper filtration (specify paper
identification)
4 Subsidence (specify time and
conditions)
Analytical separations (Table 2) like those
in Table 1, do not give a precise separa-
tion of the various forms of P which may
be included quantitatively with ortho or
poly P. Conversely some of the poly and
organic P will be included with ortho P if
they have been partially hydrolyzed
during storage or analysis. Insolubles
may likewise be included as a result of
poor separation and analytical conditions.
The separation methods provide an
operational type of definition adequate
in most situations if the "operation"
is known. Table 2 indicates the nature
of incidental P that may appear along
with the type sought.
15-2
-------
Phosphorus in the Aqueous Environment
Table 1
PHOSPHORUS COMPOUNDS CLASSIFIED BY
CHEMICAL AND SOLUBILITY RELATIONS
Form
Water Soluble
(1)
Insoluble
(1)
1. Ortho phosphates
(PO/3
Combined with monovalent
cations such as H,+ Na
Combined with multi
valent cations such
„ +2 .,
as Ca Al
^
Fe
+3
2. Poly phosphates
' "4"5(P3°9
and others depending upon
the degree of dehydration.
as in 1 above
Increasing dehydration
decreases solubility
(a)as in 1 above
(b) multi P polyphosphates
(high mol. wt.) in-
cluding the "glassy"
phosphates
3. Organic phosphorus
R-P, R-P-R (2)
(unusually varied nature)
(a) certain chemicals
(b) degradation products
(c) enzyme P
(d) phosphorylated nutrients
(a) certain chemicals
(b) cell mass, may be
colloidal or agglom-
erated
(c) soluble P sorbed by
insoluble residues
4. Mineral phosphorus
(a) as in 1 above
(a) as in 1 above
(b) as in 2 above
(c) geological P such as
phosphosilicates
(d) certain mineral com-
plexes.
(1) Used in reference to predominance under common conditions.
-3
(2) R represents an organic radical, P represents P, PC>4> or its derivatives.
Total P in Table 2 includes liquid and
separated residue P that may exist in
the whole sample including silt, organic
sludge, or hydrosoils. This recognizes
that the feedback of soluble P from
deposited or suspended material has a
real effect upon the kinetics of the
aqueous environment.
15-3
-------
Phosphorus in the Aqueous Environment
Table 2
PHOSPHORUS COMPOUNDS CLASSIFIED BY
ANALYTICAL METHODOLOGY
Desired P Components
Technique
(1)
Incidental P Included
(2)
1. Ortho phosphates
No treatment on clear
samples
Easily hydrolyzed
(a) poly phosphates -
(b) organic -P, -
(c) Mineral -P, + or
2. Polyphosphates
(2)-
-------
Phosphorus in the Aqueous Environment
IV I'olyphosphates are oi major interest in
cleaning agent formulation, an dispfrmuit::,
and for corrosion control.
A They are prepared by dehydration of ortho
phosphates to form products having two or
more phosphate derivatives per molecule.
1 The simplest polyphosphate may be
prepared as follows:
NaO NaO
;/• =0
O
O
HO
NaO'
\p
mono sodium ortho
phosphate (2)
disodium dihydrogen
polyphosphate
2 The general form for producing
polyphosphates from mono substituted
orthophosphates is:
n (NaH2PO4)
3 Di-substituted ortho phosphates or
mixtures of substituted ortho phosphates
lead to other polyphosphates:
insoluble polyphosphate than the
cation in the form of insoluble ortho
phowphute. InsolubiJiiy incf-anc;-: with
the number of I1 atoinw in thi-
polyphosphate. The "glassy" poly
phosphates are a special group with
limited solubility that are used to aid
corrosion resistance in pipe distribu-
tion systems and similar uses.
B Polyphosphates tend to hydrolyze or
"revert" to the ortho P form by addition
of water. This occurs whenever
polyphosphates are found in the aqueous
environment.
1 The major factors affecting the rate of
reversion of poly to orthophosphates
include:
a) Temperature, increased T increases
rate
b) pH, lower pH increases rate
c) Enzymes, hydrolase enzymes
increase rate
d) Colloidal gels, increase rate
e) Complexing cations and ionic
concentration increase rate
f) Concentration of the polyp'nosphau
increases rate
disodium hydrogen
ortho phosphate
Nail |PO<
mono SO!
The polyphosphate series usually
consist of the polyphosphate anion
with a negative charge of 2 to 5.
Hydrogen or metals commonly occupy
these sites. The polyphosphate can be
further dehydrated by heat as long as
hydrogen remains. Di or trivalent
cations generally produce a more
2 Items a, b and c have a large eiiect
upon reversion rate compared with
other factors listed. The actual
reversion rate is a combination of
listed items and other conditions or
characteristics.
3 The differences among ortho and ortho
+ polyphosphates commonly are close to
experimental error of the colorimetric
test in stabilized surface water samples.
A significant difference generally
indicates that the sample was obtained
relatively close to a source of poly-
phosphates and was promptly analyzed.
This implies that the reversion rate of
polyphosphates is much higher than
generally believed.
-------
Phosphorus in the Aqueous Environment
V SAMPLING AND PRESERVATION
TECHNIQUES
A Sampling
1 Great care should be exercised to
exclude any benthic deposits from
water samples.
2 Glass containers should be acid rinsed
before use.
3 Certain plastic containers may be
used. Possible adsorption of low con-
centrations of phosphorus should be
checked.
4 If a differentiation of phosphorus forms
is to be made, filtration should be
carried out immediately upon sample
collection. A membrane filter of
0.450 pore size is recommended for
reproducible separations.
B Preservation
1 If at all possible, samples should be
analyzed on the day of collection. At
beet, preservation measures only
retard possible changes in the sample.
a Possible physical changes include
solubilization, precipitation,
absorption on or desorption from
suspended matter.
b Possible chemical changes include
reversion of poly to ortho P and
decomposition of organic or min- .
eralP.
c Possible biological changes
include microbial decomposition
of organic P and algal or
bacterial growth forming organic
P.
2 Refrigeration at 4 C is recommended
if samples are to be stored up to
24 hours. This decreases hydrolysis
and reaction rates and also losses due
to volatility.
3 Addition of 2 ml concentrated
H2SO4 or 40 mg HgCl2/liter is
recommended for longer storage
periods. This chemical limits
biological changes.
a HgC !„ is an interference in the
analyucal procedure if the
chloride level is low (less than
50 mg. Cl/1). See Part VI
B 3 below.
VI THE EPA ANALYTICAL PROCEDURE*6}
A This is a colorimetric determination,
specific for orthophosphate. Depending
on the nature of the sample and on the
type of data sought, the procedure in-
volves two general operations:
1 Conversion of phosphorus forms to
soluble orthophosphate (See Fig. 1);
a sulfuric acid-hydrolysis for
polyphosphates, and some
organic P compounds,
b persulfate digestion for organic
P compounds.
2 The color determination involves
reacting dilute solutions of phosphorus
with ammonium molybdate and
potassium antimonyl tartrate in an
acid medium to form an antimony-
phosphomolybdate complex. This
complex is reduced to an intensely
blue-colored complex by ascorbic
acid. The color is proportional to
the orthophosphate concentration.
Color absorbance is measured at
880 nm or 650 nm and a concentration
value obtained us ing a standard curve.
Reagent preparation and the detailed
procedure can be found in the EPA
manual.
The methods described there are
usable in the 0.01 to 0.5 mg/liter
phosphorus range.
15-6
-------
Phosphorus in the Aqueous Environment
Tot»l KydtolyiabU
A Orthophoiphttta
total
?hoiphorui
Filter 0.45 Micron Membrane Filter
Coloriwtry
V \
p
I Soluble Soluble Uydrol7»«bU
[ Ortbcphosphatt 6 Orthopho«ph»c«
ltior«ir.ic polyphosphat*
phosphorus + some orpuaic
c ocpound* phosphor u«
I ColorlMtry
!To:«l Soluble
inorganic +•
organic
phCMpburu*
coapouo4«
FIGURE 1
ANALYTICAL SCHEME FOR DIFFEREKHATION OF PHOSPHORUS FORMS
B Interferences
1 Erroneous results from, contaminated
glassware is avoided by cleaning it
with hot 1:1 HC1, treating it with
procedure reagents and rinsings
with distilled water. Preferably
this glassware should be used only
for the determination of phosphorus
and protected from dust during
storage. Commercial detergents
should never be used.
2 High iron concentrations in samples
can precipitate phosphorus.
3 If HgCl- is used as a preservative,
it interferes if the chloride level of
the sample is less than 50 mg
Cl/liter. Spiking with NaCl is then
recommended.
4 Others have reported interference
from chlorine, chromium, sulfides,
nitrite, tannins, lignin and other
minerals and organics at high con-
centrations.
/ e \
C Precision and Accuracy
1 Organic phosphate - 33 analysts in
19 laboratories analyzed natural
water samples containing exact in-
crements of organic phosphate of
0.110, 0.132, 0.772, and 0.882 mg
PI liter.
Standard deviations obtained were
0.033, 0.051, 0.130 and 0.128
respectively.
Accuracy results as bias, mg P/liter
were: +0.003, +0.016, +0.023 and
- 0.008, respectively.
2 Orthophosphate was determined by
26 analysts in 16 laboratories using
samples containing orthophosphate
in amounts of 0. 029, 0.038, 0.335
and 0.383 mg P/liter.
Standard deviations obtained were
0.010, 0.008, 0.018 and 0.023
respectively.
'7
-------
Phosphorus in the Aqueous Environment
Accuracy results as bias, mg P/liter
were -0.001, -0.002. -0.009 and
-0. 007 respectively.
O Automated Methods
The EPA Manual also contains a
procedure for an automated colorimetry
method using the ascorbic acid reduction
method.
VII VARIABLES IN THE COLORIMETRIC
PROCEDURE
Several important variables affect
formation of the yellow heteropoly
acid and its reduced form, molybdenum
blue, in the colorimetric test for P.
A Acid Concentration during color develop-
ment is critical. Figure 2 shows that
color will appear in a sample containing
no phosphate if the acid concentration
is low. Interfering color is negligible
when the normality with respect to
approaches 0.4.
1 Acid normality during color develop-
ment of 0. 3 to slightly more than
0.4 is feasible for use. It is prefer-
able to control acidity carefully and
to seek a normality closer to the
higher limits of the acceptable range.
2 It is essential to add the acid and
molybdate as one solution.
3 The aliquot of sample must be
neutralized prior to adding the
color reagent.
U
iu
0.05 mg P
/\
\ / STD-BIK
BLANK
O.I 0.2 0.3 0.4 OJ 0.6
NO*MAl(TY
Figure 2
0-PHOSPHATE COLOR
VS ACIDITY
Choice of Reductant - Reagent stability,
effective reduction and freedom from
deleterious side effects are the bases
for reductant selection. Several re-
ductants have been used effectively.
Ascorbic acid reduction is highly
effective in both marine and fresh water.
It is the reductant specified in the
EPA method.
Temperature affects the rate of color
formation. Blank, standards, and
samples must be adjusted to the same
temperature (* 1°C). (preferably room
temperature), before addition of the
acid molybdate reagent.
Time for Color Development must be
specified and consistent. After addition
of reductant, the blue color develops
rapidly for 10 minutes then fades grad-
ually after 12 minutes.
15-8
-------
Phosphorus in the Aqueous Environment
VIH DETERMINATION OF TOTAL
PHOSPHORUS
A Determination of total phosphorus
content involves omission of any
filtration procedure and using the acid-
hydrolysis and persulfate treatments
to convert all phosphorus forms to the
test-sensitive orthophosphate form.
B Determining total phosphorus content
yields the most meaningful data since
the various forms of phosphorus may
change from one form to another in a
short period of time. (See part V, Bl)
K DEVELOPMENT OF A STANDARD
PROCEDURE
Phosphorus analysis received intensive
investigation; coordination and validation of
methods is more difficult than changing
technique.
A Part of the problem in methods arose
because of changes in analytical objectives
such as:
1 Methods suitable to gather "survey"
information may not be adequate for
"standards".
2 Methods acceptable for water are not
necessarily effective in the presence
of significant mineral and organic
interference characteristic of hydro-
soils, marine samples, organic
sludges and benthic deposits.
3 Interest has been centered on "fresh"
water. It was essential to extend them
for marine waters.
4 Instrumentation and automation have
required adaptation of methodology.
B Analysts have tended to work on their own
special problems. If the method
apparently served their situations, it was
used. ,Each has a "favorite" scheme that
may be quite effective but progress
toward widespread application of "one
method has been slow. Consequently,
many methods are available. Reagent
'acidity, Mo content, reductant and
separation techniques are the major
variables.
C At the present time there is not sufficient
data to warrant EPA endorsement of the
P procedure for sediment-type samples,
sludges, algal blooms, etc. Following
is a procedure (not included in the EPA
manual) which is useful when solids
are present in samples:
1 If sample contains large particles,
grind and emulsify solids in a blender.
2 Transfer 50 ml sample, or aliquot
diluted to 50 ml, into a 250 ml
Erie run ever flask.
3 Add 6. 0 ml of 18N H2SO4, 5 ml
concentrated HNO"o, 2 berl saddles
and digest on hot plate.
4 Digest until the disappearance of nitric
acid fumes and the appearance of white
503 fumes. Continue digestion for
approximately 5 minutes. Cool before
proceeding with Step 5.
5 Add 2 ml of HNO3-HCK>4 mixture
and 5 ml concentrated HNO3. Continue
digestion until all of the nitric acid is
driven off and dense fumes of perchloric
evolve. Perchloric acid requires
dilution with sulfuric acid and prior
destruction of most organics for safety.
15-9
-------
Phosphorus in the Aqueous Environment
Cool. Add approximately 40 ml
distilled water and transfer to 100 ml
volumetric flask.
Add 2-3 drops of phenolphthalein and
concentrated ammonium hydroxide
until a pink color is seen. Then
discharge the pink color with the
strong sulfuric acid. It is advisable
to add an equivalent amount of salt
formed during neutralization of digested
samples to the calibration standards to
equalize salt content during color
development.
Determine orthophosphate according
to the usual color procedure.
ACKNOWLEDGMENT:
Materials in this outline include significant
portions of previous outlines by J. M. Cohen,
L. J. Kamphake, andR. J. Lishka. Important
contributions and assistance were made by
B. C. Kroner, E. F. Barth, W. Allen Moore,
Lloyd Kahn, Clifford Risley, Lee Scarce.
John Winter, and Charles Feldmann.
REFERENCES
2 Gales, Morris E. . Jr., Julian, Elmo C. ,
and Kroner, Robert C. , Method for
Quantitative Determination of Total
Phosphorus in Water. JAWWA 58:
(10) 1363. October 1966.
3 Lee, G. Fred, Clesceri, Nicholas L. and
Fitzgerald, George P., Studies on the
Analysis of Phosphates in Algal Cultures.
Int. J. Air & Water Poll. 9:715. 1965.
4 Barth, E. F. and Salotto, V. V.,
Procedure for Total .Phosphorus
in Sewage and Sludge, Unpublished
Memo, Cincinnati Water Research
Laboratory, FWQA. April 1966.
5 Moss, H. V., (Chairman, AASGP
Committee) Determination of Ortho
Phosphate, Hydrolyzable Phosphate
and Total Phosphate in Surface Water,
JAWWA 56:1563. December 1958.
6 Methods for Chemical Analysis of Water
& Wastes, EPA, MDQARL, Cincinnati.Ohio
45268, 1974.
This outline was prepared by Ferdinand J.
Ludzack, Chemist, Audrey E. Donahue,
Chemist, both with National Training Center,
MPOD, OWPO, USEPA, Cincinnati, Ohio
45268.
Jenkins, David, A Study of Methods
Suitable for the Analysis and
Preservation of Phosphorus Forms
in the Estuarine Environment. DHEW,
Central Pacific River Basins Project.
SERL Report No. 65-18. University
of California. Berkeley, Calif.
November 1965.
Descriptors :
Chemical Analysis, Nutrients, Phosphates,
Phosphorus, Phosphorus Compounds,
Pollutant Identification, Sampling, Water
Analysis, Water Pollution Sources.
15-10
-------
BIOCHEMICAL OXYGEN DEMAND TEST
DILUTION TECHNIQUE
I OBJECTIVES: to prepare a sample dilution for series observation of oxygen depletion versus
time for the estimation of BOD , rate anc
Winkler iodometric titration of dissolved
will be used to check the titrimetric DO.
time for the estimation of BOD , rate and ultimate demand. The azide modification of the
Winkler iodometric titration of dissolved oxygen will be used. An electronic DO analyzer
II PROCEDURE:
A Each group will prepare 1 dilution of either effluent or influent at an assigned concentration
to enable filling 7 standard BOD bottles. One of these will be used for electronic DO mon-
itoring by the lab instructor, the other Bix are for daily student determinations. Identify
each bottle with group number, I for influent or E for effluent and the assigned concentration
in %.
B Calculate the volume of sample to be used to make up a total diluted volume of 3500 ml at
the assigned concentration. Then dilute the sample as follows:
1 Add dilution water to the dilution bottle to make up about 1/2 of the total amount of
distilled water to be used. Add 1 ml each of the four mineral supplements. (These
solutions are prepared at 4 times the concentration of Standard Methods mineral
supplement reagents so the added volume is lower). Mix the supplemented dilution
water thoroughly.
2 Carefully mix the bottle of sample to be used to a uniform suspension. If large chunks
are evident, homogenize. Rapidly pour out, with mixing, into a graduate, the
amount of sample required and return to lab station.
(Mixing, time, and precision are important. Your sample may be non-representative to
the extent of carelessness in mixing, allowing settling while adjusting measurement or
measurement errors.)
3 Add the measured quantity of sample to the dilution bottle.
4 Carefully add dilution water to make up 3500 ml of diluted sample. Use some of the dilu-
tion water to rinse the graduate into the bottle.
5 Stopper and shake the dilution bottle to mix the sample and to ensure that you have an
acceptable initial DO. This step will add DO or reduce supersaturation. About 15
seconds of mixing is usually adequate.
C Filling bottles is simplified if you place the dilution bottle on the center pedestal, insert
the siphon and line up your identified bottles with stoppers removed. Have a container
larger than the bottles available to accommodate the overflow while filling them. Otherwise,
move the apparatus near a sink. Fill the siphon tube and waste some of the diluted sample
to rinse this tube. Be sure that the siphon tube is used in a manner so that it is always
full of the diluted sample. Agitate the diluted sample regularly to avoid undue collection of
solids on the bottom.
16-1
CH, O. bod. lab. 3. 11.75
-------
Biochemical Oxygen Demand Test Dilution Technique
1 Place first bottle to be filled in the overflow container or hold over sink. Insert the
siphon delivery tip all of the way down into the bottom of the test bottle. Tilt the
bottle to cover the delivery tip with water promptly. Start the siphon flow slowly to
decrease turbulence until the delivery tip is well-covered. Fill to overflowing. Slowly
raise the delivery tip while overflowing the sample. Stop the flow with the bottle com-
pletely full. Insert the bottle stopper and turn it slightly to seat it. Fill the remaining
bottles in the same manner.
2 Each bottle will have the same DO and sample mix as the next if you have followed a
consistent procedure, mixed carefully, and have not changed technique enroute.
3 One of the seven bottles la to be retained for initial tit ration of DO by your group. Another
bottle is to be turned in to the lab instructor for use in performance checks on all class
dilution*. The remainder are to be placed into trays for 20° incubation. Be certain
that all bottles are properly identified.
D Dissolved Oxygen Determination
1 You are encouraged to check the DO of your sample on the electronic analyzer. A
calibrated unit will be available with assistance. The DO analyzer will not result in
more than 1/2 ml loss of sample on probe insertion. You will be able to add DO reagents
and continue the Winkler titrimetric DO on the same bottle.
2 An Initial DO value (DO ) using the Winkler method must be determined today. On succeed
ing days, the DO. will b% determined on the incubated samples for estimation of BOD
progression. (If reaeration is required, the technique will be demonstrated). All data
is to be recorded on the sheet provided in this manual.
The Winkler-Azide procedure as presented in outline 17 is to be used for these DO
determinations.
3 A blank value for each stock bottle of distilled water will be determined by the staff.
REFERENCE
Standard Methods for the Examination of
Water and Wastewater, 13th edition.
APHA, AWWA, WPCF, New York.
p. 489. 1971.
This outline was prepared by F. J. Ludzack.
Chemist, National Training Center, MPOD,
OWPO, USEPA, Cincinnati. Ohio 45268.
Descriptors: Analytical Techniques,
Biochemical Oxygen Demand, Chemical
Analysis. Laboratory Tests, Water Analysis
16-2
-------
Biochemical Oxygen Demand Test Dilution Technique
HOD DATA SHEKT
Group No.
Initial Date
Buret
Readings
DO
Buret
Readings
DO,
ADO
Sf/ID
BOD
Buret readings - initial and final
DO - initial DO in mg/liter
o
DO = DO at indicated incubation time t
£DO» DO change (DO - DOt); a depletion
Sf = the sample content as a decimal fraction in the dilution
ID = sample identification (influent or effluent)
BOD= DO - DO (according to Standard Methods, 13th edition, 1971)
Sf
16-3
-------
LABORATORY PROCEDURE FOR DISSOLVED OXYGEN
Winkler-Azide Procedure
I INTRODUCTION
The azide modification is used for most
sewage, effluents, and streams which con-
tain nitrate nitrogen and not more than 1 mg
of ferrous iron/1. If 1 ml 40% KF solution
is added before acidifying the sample and
there is no delay in titration, the method is
also applicable in the presence of 100-200
mg/1 ferric iron.
The "Methods for Chemical Analysis of
Water & Wastes, 1974, " published by the
Environmental Protection Agency,
recommends the Winkler-Azide
jmethod using the full bottle technique.
REAGENTS
A Manganous Sulfate Solution
Dissolve 480 g MnSO
water and dilute to
- - 4H O in distilled
1 liter.2
B Alkali-Iodide-Azide Reagent
Dissolve 500 g sodium hydroxide and 150 g
potassium iodide in distilled water and
dilute to 1 liter. To this solution add 10 g
sodium azide. NaNg dissolved in 40 ml
distilled water.
C Sulfuric Acid, Cone.
. The strength of this acid is 36 N.
D Starch Solution
Prepare an emulsion of 10 g of soluble
starch in a mortar or beaker with a small
quantity of distilled water. Pour this
emulsion into 1 liter of boiling water,
allow to boil a few minutes, and let settle
overnight. Use the clear supernate.
This solution may be preserved by the
addition of 5 ml per liter of chloroform
and storage in a 10° C refrigerator.
E Sodium Thiosulfate Stock Solution 0.75 N
Dissolve 186. 15 g Na^Og- 5 HO in boiled
and cooled distilled water and dilute to
1 liter. Preserve by adding 5 ml chloroform.
F Sodium Thiosulfate Standard Titrant 0. 0375N
Dilute 50. 0 ml of stock solution to 1 liter.
Preserve by adding 5 ml of chloroform.
Standardize with potassium biiodate.
HI PROCEDURE
A Addition of Reagents
1 Manganous sulfate and alkali- iodide-
azide
To a full sample bottle (300 ml + 3 ml
BOD incubation), add 2 ml manganous
sulfate solution and 2 ml alkaline-
iodide reagent with the tip of each
pipette below the surface of the sample.
2 Replace the stopper.
3 Rinse under running water.
4 Mix well by inverting 4-5 times.
5 Allow the precipitate to settle until at
least 100 ml of clear supernate have
been produced.
6 Repeat steps 4 and 5.
7 Add 2 ml cone, sulfuric acid with the
tip of the pipette above the surface of
the sample.
8 Stopper the bottle.
9 Rinse under running water.
] 0 Mix by inverting several times to dissolve
the precipitate.
1 1 Pour contents of bottle into a wide mouth
500 ml Erlenmeyer flask.
B Titration
1 Titrate with 0. 0375 N thiosulfate to a
pale straw color.
CH.O.do.lab. 3b. 11.75
17-1
-------
Laboratory Procedure for Dissolved Oxygen
2 Add 2 ml starch solution indicator.
3 Continue titrating to the disappearance
of the blue color.
C Calculation
mla tttrant X 0.0375 X8000
300
fl
*
mla titrantX 0.0125 X80 » mg/1 DO
mis tltrant X 1 « mg/1 DO
mis titrant • mg/1 DO
REFERENCE
Methods for Chemical Analysis of Water and
Wastes, EFA-MDQARL,. Cincinnati, Ohio.
1974.
This outline was prepared by J. W. Mandia,
Chemist, formerly with National Training
Center, and revised by C. R. Feldmann,
Chemist, National Training Center, MPOD.
OWPO, EPA, Cincinnati, OH 45268.
Descriptors: Analytical Techniques, Chemical
Analysis, Dissolved Oxygen, Laboratory
Tests, Oxygen. Water Analysis
-------
DETERMINATION OF KJELDAHL NITROGEN
(MICRO APPARATUS-NESSLERIZATION)
I REAGENTS
A Distilled Water
This should be ammonia-free. Pass
distilled water through an ion exchange
column with strongly acidic cation resin
mixed with a strongly basic anion -resin.
B Sulfuric Acid (20%)
20 ml acid/ 100 ml distilled water
C Mercuric Sulfate Solution
Dissolve 8g mercuric oxide in 50 ml of
20% sulfuric acid. Dilute to 100 ml with
distilled water.
D Digestion Reagent
Dissolve 134g potassium sulfate in about
650 ml distilled water. Add 200 ml
concentrated sulfuric acid. Add 25 ml of
mercuric sulfate solution (C above) and
dilute to 1 liter.
E Sodium Hydroxide - Sodium Thiosulfate
Solution
Dissolve 500g sodium hydroxide and 25g
sodium thiosulfate pentahydrate in distilled
water and dilute to 1 liter.
F Mixed Indicator
Mix 100 ml of methyl red indicator
(200 mg in 100 ml 95% ethyl alcohol) with
50 ml of methylene blue indicator solution
(200 mg in 100 ml 95% ethyl alcohol).
G Boric Acid Solution, 2%
Dissolve 20g boric acid in distilled water
and dilute to 1 liter.
H Ammonium Chloride Stock Solution
Dissolve 3. 819g NH4C1 in distilled water
and dilute to 1 liter. 1.0 ml = 1.0 mg
NH3-N.
CH.N.lab.8. 11.75
I Ammonium Chloride Standard Solution
Dilute 10. 0 ml of the stock solution with
distilled water to 1 liter.
1.0 ml = 0.01 mg NH3-N
J Sodium Hydroxide Solution
Dissolve 160 g sodium hydroxide in 500 ml
distilled water.
K Nessler Reagent
Dissolve lOOg mercuric iodide and 70g
potassium iodide in a small volume of
water. Add this mixture slowly to the
sodium hydroxide solution (J above), then
dilute to 1 liter.
II EQUIPMENT PREPARATION
A This procedure should be used if the
apparatus has been out of service for 4
hours or more.
1 Add about 50 ml of a 1:1 mixture of
ammonia-free distilled water and
sodium hydroxide-sodium thiosulfate
solution to each of the micro Kjeldahl
flasks to be used.
2 Add glass beads to each flask.
3 Attach a flask to the steam distillation
apparatus and distill about half the
mixture.
4 Add 1 ml of the Nessler reagent to the
distillate to check for ammonia.
a If the distillate is colorless, the
equipment is ammonia-free and the
procedure can be repeated with the •
next flask to be used.
b If the distillate is yellow, discard it,
distill another half of the mixture and
check this distillate with 1 ml Nessler
reagent. Repeat the process until the
distillate is colorless.
18-1
-------
Determination of Kjeldahl Nitrogen
HI DIGESTION OF SAMPLE
A Preparation of Digestion Mixture
1 Shake the sample.
2 Measure 50.0 ml sample into a 100 ml
Kjeldahl flask.
3 Add 10 glass beads.
4 Add 10 ml of the digestion reagent.
B Digestion
1 Place the flask to a properly ventilated
Kjeldahl digestion apparatus.
2 Turn on the heat source.
3 Evaporate the mixture until sulfur
trioxide (SO3) fumes are given off.
fumes are white. Also, the solution
will be pale yellow. )
4 Continue heating for an additional 30
minutes.
5 Turn off the heat source.
6 Cool the residue in the flask.
IV STEAM DISTILLATION
A Preparation of the Digestion Residue
1 Add 30 ml of ammonia-free distilled
water to the digested residue in the
Kjeldahl flask.
2 Connect the flask to the ground glass
joint of the micro steam distillation
apparatus.
3 Add 5 ml of 2% boric acid to a 50 ml
Erlenmeyer receiving flask.
4 I'osition the Erlenmeyer flask so that
the tip of the condenser is below the
level of the boric acid solution in the
receiving flask.
5 Carefully add 10 ml oJ the flodlum hyHroxidr-
sodium thiosull'ate solution from the
dropping funncL.
B Distillation
1 Turn on the heat source.
2 Distill at a rate of 6-10 ml/minute up to
the 30 ml mark on the Erlenmeyer
receiving flask.
3 Remove the receiving flask.
4 Put a small beaker under the condenser
tip to receive any additional distillate.
5 Turn off the heat source if there are no
more digestion residues to distill.
V COLORIMETRY FOR AMMONIA-
NESSLERIZATION
If the ammonia content is found to be greater
than 1 mg/liter, a titration procedure should
be used^ rather than Nesslerization.
A Preparation of Standards and Sample
1 Label nine 50. 0 ml Nessler tubes with
the following: 0, 0.5, 1. 2, 4, 5, 8,
10, and S.
2 Pipet the following volumes of ammonium
chloride standard solution into the corre-
spondingly labeled tubes: 0. 5 ml, 1. 0 ml,
2.0ml, 4.0ml, 5.0ml. 8.0ml, and
10.0 ml.
3 Pour the 30 ml of distillate collected
from the sample into the Nessler tube
marked "S".
4 Pour ammonia-free distilled water into
the tube labeled "0", bringing the volume
to the 50.0 ml line.
5 Add ammonia-free distilled water to each
of the remaining 8 tubes, bringing the
volume of each to the 50. 0 ml line.
18-2
-------
Determination of Kjeldahl Nitrogen
6 Add 1.0 ml of Nessler Reagent to each
of the 9 tubes.
7 Mix the solution in the "0" tube using a
cap or a rubber stopper on the top, then
inverting the tube three times.
8 Remove, rinse, and dry the cap or
stopper.
9 Repeat steps 7 and 8 to mix the contents
of the other 8 tubes.
10 After mixing, let the tubes stand in a
rack for 20 minutes before getting
absorbance readings.
B Spectrophotometric Readings
1 Turn the instrument on.
2 Set the wavelength at 425 nm.
3 After the 20 minute time span, use the
contents of the "0" tube to adjust to zero
absorbance on the spectrophotometer.
4 Using the contents of the tube labeled
0. 5, rinse, then fill an instrument cell.
5 Place the cell in the holder and record
the absorbance value from the instrument.
6 Discard the contents from the cell.
7 Use ammonia-free distilled water to
rinse the cell three times.
8 Repeat steps 4 through 7 to obtain
absorbance values for the rest of the
standards and for the sample.
9 Turn off the instrument.
VI CALIBRATION CURVE
A Constructing the Curve
1 Calculate the concentration of each
standard by multiplying the ml of
working standard used times 0.01 mg/ml,
which is the concentration of the
standard solution. This was diluted to
50. 0 ml in the Nessler tube, so the
result is mg NH3-N/50.0 ml. For
example, if 0. 5 ml of standard was used,
the concentration is
(0.5)(0.01) = 0.005 mg NH3-N/50.0 ml.
2 Plot the absorbance values for the
standards against these calculated
concentrations,
3 Draw the best straight line from zero
through all the points.
B Using the Curve
1 To find the NHj-N concentration in the
sample, locate its absorbance value on
the curve.
2 Find the corresponding mg NH3-N/50.0 ml
by dropping a vertical line to the
concentration axis.
3 Record this result.
VII FINAL CALCULATIONS
A Use this formula to calculate Total Kjeldahl
Nitrogen:
TKN. mg/1 =
Where:
X
B
ml sample C
A = mg NH,,-N/50. 0 ml from curve
B = ml total distillate including boric acid
C = ml distillate taken for Nesslerization
ml sample = ml of original sample taken
B An example calculation using the value from
the calibration curve would be:
TKN, mg/1 = AX 1000 B_
ml sample C
If:
A = 0.045
B = 30 ml (25 ml distillate + 5 ml boric acid)
C = 30 ml
ml sample - 50 ml
18-3
-------
Determination of Kjeldahl Nitrogen
Then:
TKN, mg/1 =
20
0.045X J000
I?
1
= 0.045X20X 1
= 0.045 X 20
= 0.90
TKN « 0. 90 mg/1
This outline was prepared by A. E. Donahue,
Chemist, National Training Center, MPOD,
OWPO. USEPA. Cincinnati. Ohio 45268.
REFERENCE
1 Methods for Chemical Analysis of Water
and Wastes, EPA-MDQARL. 1974.
Descriptors: Ammonia, Analytical Techniques,
Chemical Analysis, Laboratory Tests. Nitrogen
Nitrogen Compounds, Nutrients. Water Analysis
18-4
-------
LABORATORY PROCEDURE FOR TOTAL SOLIDS
I INTRODUCTION
A This procedure was excerpted from methods
for Chemical Analysis of Water and Wastes,
1971, Environmental Protection Agency,
Office of Water Programs, Analytical Quality
Control Laboratory.
B The procedure is applicable to surface and
saline waters, domestic and industrial wastes.
C The practical range of the determination is
10-30000 mg/1.
D Nonhomogenous materials (large floating
particles or submerged agglomerates) should
be excluded from the sample. Floating grease
and oil should be included in the sample and
dispersed in a blender before measuring the
aliquot.
E Samples should be analyzed as soon as
possible.
H EQUIPMENT
A Porcelain, Vycor, or platinum evaporating
dishes, 100 ml capacity; smaller sizes may
be used as required.
B Muffle furnace, 550- 50°C
C Drying oven, 103 - 105 C
D Desiccator
E Analytical balance
F Steam bath or Oven at 98°C
C Store the dish in the desiccator and weigh
just before use.
D Weigh the dish on an analytical balance.
E Shake the sample container vigorously.
F Measure 100 ml of the well mixed sample
in a graduated cylinder. (At least 25 mg
of residue should be obtained; less volume
of sample may be used if the sample appears
to be high in solids content. If it is low in
solids content, more sample may be added
to the dish after drying).
G Rapidly, but without spilling, pour the
sample into the evaporating dish.
H Dry the sample on a steam bath, or at 98 C
(to prevent boiling and splattering) in the
oven.
I Dry the evaporated sample in the oven at
103 - 105 C for at least one hour.
J Cool the dish in the desiccator and then
weigh it.
K Repeat the heating at 103 - 105°C, cooling
and weighing until the weight loss is less
than 4% of the previous weight, or 0. 5 mg,
whichever is less.
IV
CALCULATIONS
mg total solids/1 = (wt dish + residue)* -
(wt dish)*x 1000 x 1000
ml of sample
* in grams
III PROCEDURE
A Heat the clean evaporating dish in a muffle
furnace at 550- 50°C for 1 hour.
H Cool the dish in a desiccator.
This outline was prepared by C. R. Feldmann,
Chemist, National Training Center, MPOD,
OWPO, USEPA, Cincinnati, Ohio 45268.
Descriptors: Analytical Techniques, Chemical
Analysis, Laboratory Tests, Solids, Water
Analysis
PC. lab. 16a. 11.75
20-1
-------
I,AROHATOHY PHOCKIMiKi: !•'()]( TOTAL
JNICSS
I REAGENTS
A Buffer Solution:
1 Dissolve 16. 9 g of NH Cl in 143 ml of
cone. NH4OH. 4
2 Dissolve 1. 179 g of analytical reagent
grade disodium ethylenediamine tetra-
acetic acid dihydrate(Na9EDTA-2H9O)
and 0. 644 g of MgCl •6K9O in 50 ml of
distilled water. z i
3 Add the solution from (2) to the solution
from (1) with mixing, and dilute to 250
ml with distilled water. Addition of
small amounts of Na_EDTA-2H2O or
MgCl-- 6H2O may be necessary to attain
exact equivalence.
4 The buffer should be stored in a plastic
or resistant glass container tightly
stoppered to prevent CO absorption
and NH, loss. Discard me buffer when
1 or 2 ml added to the sample fails to
produce a pH of 10. 0 + 0.1 at the end
point of the titration.
B Inhibitor:
1 Most water samples do not require the
use of an inhibitor. In the presence of
certain interfering ions, however, an
inhibitor may be needed to sharpen the
endpolnt color change. Several types
of Inhibitors may be prepared or pur-
chased. Sodium cyanide, NaCN, is one
of the simpler inhibitors to use.
2 Add 0. 25 g of powdered NaCN to the
sample and adjust the pH to 9. 9-10. 1.
Caution: NaCN is poisonous. Use
Targe amounts of water when flushing
solutions containing NaCN down the
drain. Do not acidify solutions contain-
ing NaCN; volatile, poisonous hydrogen
cyanide, HCN, would be liberated.
C Indicator:
1 Eriochrome Black-T dye (EBT) is use-
ful for the determination. Other commer-
cial grades or laboratory formulations
of the dye are also satisfactory.
Prepare the indicator in dry powder
form by grinding together 0.5 g of the
dye and 100 g of NaCl.
D Standard Calcium Carbonate,
Weigh 1. 000 g of anhydrous, primary
standard grade CaCO,, and transfer it to
a 500 ml Erlenmeyer'rlask. Add 1:1 HCI
(equal volumes of cone. HCI and water)
dropwise and with swirling of the flask
until the CaCO has dissolved. Bring
the volume of liquid to about 200 ml with
water, boil a few seconds to dispel CO.,
cool, and add a few drops of methyl rea
indicator. Adjust the color of the solution
to an intermediate orange by the dropwise
addition of 1:1 HCI or 1:4 NH OH (1 volume
of cone. NH OH + 4 volumes of water).
Transfer the solution quantitatively to a
one liter volumetric flask and dilute to the
mark with water. (1.0 ml = 1.0 mgCaCOJ
O
E Na EDTA-2H9O (0. 01 M):
£i «£
Dissolve 3. 723 g of the dry reagent grade
Na EDTA-2H»O in distilled water and
dilate to 1 liter. 1. 0 ml of the 0. 01 M
solution = 1. 0 mg of CaCO,. Check the
concentration of this solution by titration
against the standard calcium carbonate
solution as described in n below.
n STANDARDIZATION OF THE i\a_EDTA'
2H20:
A Dilute 25. 0 ml of CaCO standard to about
50 ml with distilled water in a 125 ml
Erlenmeyer flask against a white background.
B Add 1-2 ml of the buffer solution and check
the pH to ensure that it is 9. 9 to 10. 1.
C Add approximately 0. 2 g of the indicator.
D Add the Na EDTA^HoO slowly and with
stirring until the color changes from a
rose to a blue color. If the color change
is not sharp, repeat the determination
using the inhibitor. If the endpoint is still
not sharp, prepare a fresh supply of
indicator.
E The titration should take less than 5 minutes,
measured from the time of buffer addition.
CH. HAR.lab. 3b. 11.75
21-1
-------
Laboratory Procedure for Total hardness
In an analysis of this type it is advant-
ageous to carry out a preliminary, rapid
titratton in order to determine approxi-
mately how much titrant will be required.
This is accomplished by adding the
NftalpDTA-ZHoO at a fast dropwise rate
unurehe color change is observed.
PROCEDURE
Repeat steps n A through n F using sample
in place of CaCOg standard. The amount of
sample taken should require less than 15 ml
of NaBtfA'aO titrant.
IV CALCULATION
A Standardisation of the Na2EDTA-2H2O:
REFERENCES
1 Standard Methods for the Examination
of Water and Wastewater. 13th Edition,
A PHA,-AWWA-WPCF (1971)
p. 179, Method 122B.
2 ASTM Standards. Part 23, Water;
Atmospheric Analysis 1972
p. 170, Method Dl 128-67.
3 Methods for Chemical Analysis of
Water and Wastes. 1974,
Environmental Protection Agency,
MDQARL, Cincinnati. OH.
ml of CaCO3 equal to 1. 0 ml of the Na2EDTA-2H2O
(symbol B>
ml of CaCO,
ml oma-EDTA'ZHgO required tor tttration
B Total Hardness
^ r*n /1 - A x B x 1000
as mg CaC03/1 - ml of sample
This outline was prepared by C. R. Feldmann,
Chemist. National Training Center, MPOD,
OWPO, USEPA. Cincinnati. Ohio 45268.
Descriptors: Analytical Techniques, Calcium,
A » ml of Na2BDTA-2H2O for tttration of sample Chemical Analysis, Hardness. Laboratory
Tests, Magnesium, Water Analysis
21-2
-------
ACIDITY
1. Scope and Application
1. 1 This method is applicable to surface waters, sewages and industrial wastes,
particularly mine drainage and receiving streams, and other waters containing
ferrous iron or other polyvalent cations in a reduced state.
1. 2 The method covers the range from approximately 10 mg/1 acidity to approxi-
mately 1000 mg/1 as CaCOs, usmS a *>° nil sample.
2. Summary of Method
2. 1 The pH of the sample is determined and a measured amount of standard acid is
added, as needed, to lower the pH to 4 or less. Hydrogen peroxide is added, the
solution boiled for several minutes, cooled, and titrated electrometrically with
standard alkali to pH 8.2.
3. Definitions
3.1 This method measures the mineral acidity of a sample plus the acidity resulting
from oxidation and hydrolysis of polyvalent cations, including salts of iron and
aluminum.
4. Interferences
4.1 Suspended matter present in the sample, or precipitates formed during the
titration may cause a sluggish electrode response. This may be offset by allowing a
15-20 second pause between additions of titrant or by slow dropwise addition of
titrant as the endpoint pH is approached.
5. Apparatus
5. 1 pH meter, suitable for electrometric titrations.
6. Reagents
6. 1 Hydrogen peroxide (H2O2- 30% solution).
6. 2 Standard sodium hydroxide, 0.02 N.
6. 3 Standard sulfuric acid, 0. 02 N.
7. Procedure
7.1 Pipet 50 ml of the sample into a 250 ml beaker.
7. 2 Measure the pH of the sample. If the pH is above 4. 0 add standard sulfuric acid
in 5. 0 ml increments to lower the pH to 4.0 or less. If the initial pH of the sample
is less than 4. 0, the incremental addition of sulfuric acid is not required.
7. 3 Add 5 drops of hydrogen peroxide.
CH.ALK.lab.5. 11.75 23"1
-------
7. 4 Heat the sample to boiling and continue boiling for 2 to 4 minutes. In some
instances, the concentration of ferrous iron in a sample is such that an additional
amount of hydrogen peroxide and a slightly longer boiling time may be required.
7. 5 Cool the sample to room temperature and titrate electrometrically with standard
alkali to pH 8. 2.
8. Calculations
8.,
where:
A = vol. of standard alkali used in titration
B * normality of standard alkali
C = volume of standard acid used to reduce pH to 4 or less
D = normality of standard acid
8. 2 If it is desired to report acidity in millequivalents per liter, the reported values
as CaCC*3 are divided by 50. as follows:
Acidity as meq/1 = "^ CaC°3
50
9. Precision
9. 1 On a round robin conducted by ASTM on 4 acid mine waters, including
concentrations up to 2000 mg/1. the precision was found to be 1 10 mg/1.
10. References
10.1 The procedure to be used for this determination can be found in:
ASTM Standards, Part 23. Water; Atmospheric Analysis, p 124, D-1067, Method
E (1973).
Standard Methods for the Examination of Water and Wastewater, 13th Edition.
p 370. Method 201 (Acidity and Alkalinity) (1971).
(The preceding is a copy of pp 1 and 2 from
Methods for Chemical Analysis of Water
and Wastes. EPA-MDQARL. 1974.)
Descriptors: Acids, Acidity, Analytical
Techniques, Chemical Analysis, Laboratory
Tests, Neutralization, Water Analysis
23-2
-------
LABORATORY PROCEDURE FOR TOTAL ALKALINITY
I REAGENTS
For detailed discussion of reagent preparation,
consult method reference 3.
A Carbon Dioxide - Free Distilled Water
B Standard Sodium Carbonate Solution
C Hydrochloric Acid Titrant (0. 02 N)
H STANDARDIZATION OF THE HYDRO-
CHLORIC ACID TITRANT
A Set the temperature reading on the pH
meter dial to match the temperature of
the buffer and sample solutions.
B Standardize the pH meter against a
reference buffer solution. Check against
a second buffer solution.
C Weigh accurately 0. 088 + 0. 001 g of the
dried sodium carbonate and transfer it to
a 500 ml conical flask.
D Add 50 ml of water and swirl to dissolve
the carbonate.
E While stirring the solution (magnetic bar
and sttrrer), add the hydrochloric acid
titrant from a 100 ml buret until a pH of
4. 5 is attained.
F Calculate the normality of the hydrochloric
acid solution as follows:
III PROCEDURE
A Pipette 50 ml of the sample into a 150 ml
beaker.
B Titrate with the hydrochloric acid to
pH4.5.
C Calculation
Total alkalinity as mg of CaCO3/l =
A X N X 50000
A =
0.053 X C
B
A = ml of standard HC1 titrant
N = N of standard HC1 titrant
B = ml of sample
50 * equivalent weight of CaCOg
1000 - converts ml to liters
REFERENCES
1 Methods for Chemical Analysis of Water
and Wastes, EPA-MDQARL, Cincinnati,
Ohio 45268, 1974. p. 3.
2 Standard Methods for the Examination of
Water and Wastewater, 13th ed., 1971,
p.370.
3 Book of ASTM Standards Part 23, 1972,
p. 143.
A = normality of the hydrochloric acid
B = g of sodium carbonate used
C = ml of hydrochloric acid consumed
This outline was prepared by C. R. Feldmann,
Chemist, National Training Center, MPOD,
OWPO, USEPA. Cincinnati, Ohio 45268.
Descriptors: Alkalis, Alkalinity, Analytical
Techniques, Chemical Analysis, Laboratory
0.053 = milliequivalent weight of Na2C03 Tests, Neutralization, Water Analysis
CH.ALK.]ab.2a.ll.75
24-1
------- |