EPA-600/2-76-234
September 1976
Environmental Protection Technology Series
ANALYTICAL VARIABILITY OF
FIVE WASTEWATER PARAMETERS
Petroleum Refining Industry
Robert S. Kerr Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Ada, Oklahoma 74820
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield. Virginia 22161.
-------
EPA-600/2-76-234
September 1976
ANALYTICAL VARIABILITY OF FIVE WASTEWATER
PARAMETERS--PETROLEUM REFINING INDUSTRY
by
Leon H. Myers
Thomas E. Short, Jr.
Billy L. DePrater
Fred M. Pfeffer
Treatment and Control Technology Branch
Robert S. Kerr Environmental Research Laboratory
Ada, Oklahoma, 74820
ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
ADA, OKLAHOMA 74820
-------
DISCLAIMER
This report has been reviewed by the Robert S. Kerr Environmental
Research Laboratory, U.S. Environmental Protection Agency, and approved
for publication. Mention of trade names or commercial products does
not constitute endorsement or recommendation for use.
ii
-------
ABSTRACT
Information is presented on a "Round Robin" laboratory analysis exercise
conducted by Environmental Protection Agency laboratories with coopera-
tive industrial and state agency laboratories. The initial sample
analyzed was obtained from a petroleum refinery activated sludge plant's
final clarifier effluent and represented a low contaminent level waste-
water. The second sample was taken from the discharge of an American
Petroleum Institute separator at a refinery and was characteristic of a
high contaminent level wastewater containing sulfide concentrations
which could cause analytical interference problems.
Samples were divided among 12 laboratories to be analyzed for chemical
oxygen demand, suspended solids, ammonia nitrogen, phenolics, and oil
and grease. The Robert S. Kerr Environmental Research Laboratory analyzed
six sample sets to determine intralaboratory deviation (repeatability),
while the other participating laboratories analyzed single samples to
provide data for interlaboratory deviation (reproducibility) determina-
tions. Study results are expressed in terms of averages, standard
deviation, and spike recoveries for intralaboratory, interlaboratory,
and combined evaluations.
A summary of the instruction seminar which was provided to the participants
between the low and high level sample runs to discuss analytical problems
and techniques used for the five parameters is presented.
This report was submitted by the Robert S. Kerr Environmental Research
Laboratory under the sponsorship of the Environmental Protection Agency.
Work was completed as of June, 1974.
iii
-------
CONTENTS
Sections Page
I Conclusions I
II Recommendations 3
III Introduction 4
IV Participant Selection 5
V Study Objective 9
VI Study Conditions 11
VII Sampling Procedures 13
Sample Sets 14
Location of Sample Points 17
Sample Delivery 17
VIII Participant Seminar 19
IX Statistical Evaluation 20
Preliminary Data Checking 20
Outlier Rejections 20
Data Presentation 21
Data Analyses 21
X References 35
XI Appendices 36
-------
FIGURES
No. Page
1 Project Organizational Structure 6
2 Equipment Used to Obtain Representative Sample 14
3 Sampling Pump Installation 15
4 Composite Sampling Container 15
5 Top View of Composite Sampling Container 16
6 Sample Sets for Cooperative Laboratories 16
7 Chemical Oxygen Demand Set-Up 39
8 Phenolics Set-Up 39
9 Oil and Grease (Hexane Extraction) Set-Up 40
10 Ammonia Nitrogen Set-Up 40
11 Suspended Solids Set-Up 41
vi
-------
TABLES
No. Page
1 Crude Capacity of Participants 7
2 Sample Volumes and Preservatives 17
3 TOC Analysis on Replicate Samples 19
4 COD Analytical Method Evaluation - Phase I 23
5 Suspended Solids Analytical Methods Evaluation - 24
Phase I
6 Ammonia Analytical Method Evaluation - Phase I 25
7 Phenolics Analytical Method Evaluation - Phase I 26
8 Oil and Grease Method Evaluation - Phase I 27
9 COD Analytical Method Evaluation - Phase II 28
10 Suspended Solids Analytical Method Evaluation - 29
Phase II
11 Ammonia Analytical Method Evaluation - Phase II 30
12 Phenolics Analytical Method Evaluation - Phase II 31
13 Oil and Grease Method Evaluation - Phase II 32
14 Statistical Comparison of Laboratory 01 with the 33
Other Laboratories (at the 95% confidence level)
15 Statistical Comparison of the Hexane and Freon 34
Methods for the Analysis of Oil and Grease
(at the 95% confidence level)
vii
-------
ACKNOWLEDGMENTS
The participation of member refineries of the Oklahoma Refiners Waste
Control Council (ORWCC); the Methods Development and Quality Assurance
Research Laboratory (MDQARL), Cincinnati, Ohio; Oklahoma State Health
Department; Oklahoma State University; and the assisting Programs of the
Robert S. Kerr Environmental Research Laboratory (RSKERL), Ada, Oklahoma,
are gratefully acknowledged.
The assistance of Messrs. Marion Buercklin, Malcolm Lider, Ed Sheets,
and John Skinner of the ORWCC; Dr. Sterling Burks, OSU; Bob Kroner and
Jim Lichtenberg, MDQARL; and William C. Galegar and Marvin L. Wood,
RSKERL, for project guidance are further acknowledged.
We are indebted to the personnel at the RSKERL who participated in the
analytical study and provided instruction for the seminar. Principal
RSKERL staff participants were Messrs. Jack H. Hale, Clarence Edmonson,
Kenneth Jackson, Mike Cook, Roger Cosby, Tommy Redman, John Matthews,
and Mrs. Dee Hutchings.
viii
-------
SECTION I
CONCLUSIONS
1. The chemical oxygen demand (COD) test had a repeatability (single
lab) expressed in terms of standard deviation of 9.5 milligrams per
liter (mg/1) for petroleum refinery wastewater which had a COD average
of 134 mg/1. Reproducibility (overall laboratories) for this same
refinery wastewater exhibited a standard deviation of 15.0 mg/1.
2. Suspended solids with an average concentration of 19 mg/1 had a
standard deviation for repeatability of 1.8 mg/1 and a standard devi-
ation for reproducibility of 5.2 mg/1.
3. Results obtained for the ammonia test with an average concentration
of 8.5 mg/1 exhibited a repeatability standard deviation of 0.1 mg/1 and
a reproducibility standard deviation of 0.9 mg/1.
4. The repeatability standard deviation for phenolics was 0.2 mg/1 for
a sample containing 5.5 mg/1; the standard deviation for reproducibility
of phenolics was 0.8 mg/1.
5. Oil and grease standard deviation for repeatability was 2.3 mg/1
and reproducibility was 2.9 mg/1 for a sample containing approximately
11 mg/1.
6. A t-test shows there to be no significant difference between the
quantitative results of the hexane and freon procedures.
7. The variance of the analysis for oil and grease is less for the
freon method than the hexane method.
8. Recovery of ammonia and COD spikes was greater than 95 percent
after instruction seminar.
-------
9. A comparison of results between Phases I and II indicate the
instruction seminar, which was held to achieve uniformity of analyti-
cal procedures accomplished:
a. A significant reduction in arithmetic and extreme outlier
value errors;
b. Enhancement of uniformity of laboratory technique;
c. Minimizing the COD mean values between intralaboratory
and interlaboratory results;
d. Improved spike recovery for COD and ammonia.
e. The standard deviations for COD, ammonia, and phenolics
were decreased.
10. Environmental Protection Agency (EPA) methodology for the parameters
studied appeared applicable for this petroleum refinery wastewater when
the analysts were properly instructed.
-------
SECTION II
RECOMMENDATIONS
1. Similar studies should be conducted with other industrial discharges
to evaluate variances for EPA procedures.
2. Instruction seminars need to be developed by EPA to properly instruct
analysts, whose results will be used by EPA for self-monitoring of
permit conditions and acquisition of data on demonstration grants.
3. Repeatability and reproducibility values should be established for
each EPA procedure on various industrial wastewaters.
4. An EPA Laboratory Service should be established to continue a
cross-match of methodology with wastewater discharge samples peculiar
to a specific industry. This established laboratory should be able to
accept a testing program semi-annually with participating industrial
and regulatory laboratories.
5. Phase I of this study should be repeated to determine the effect
of the training seminar on the low level parameters as opposed to the
high level parameters of Phase II.
-------
SECTION III
INTRODUCTION
Enactment of the Federal Water Pollution Control Act Amendments of 1972
provided this Nation with the legislation needed to clean up the domestic
waters. This law specified two goals for reducing and eliminating water
pollution: first, to provide water clean enough for swimming and other
recreational purposes by July 1, 1983, and second, to eliminate pollu-
tant discharges by July 1, 1985.
Within the framework of this legislation, industry discharging water
must obtain a discharge permit issued under the regulation of EPA. The
permit contains a description of the effluent limitations for the pollu-
tants of concern. Enforcement of the issued permit is administered by
EPA, either through monitoring stations or by site visits when samples
of water discharges may be obtained and analyzed for permit compliance.
The permit holder is required to provide analytical data on water dis-
charges to EPA on a routine basis. If a permit holder is in violation
of the permit conditions, court action may be pursued.
Since many states have legally provided a documented self-reporting
system of effluent chemical analysis for years, utilization of this
procedure to satisfy permit conditions is not a unique concept. Self-
reporting provides the company with control of the treatment facility
performance and the enforcement office with information on which dis-
charge point needs investigation for possible permit violations.
-------
SECTION IV
PARTICIPANT SELECTION
An inquiry was received from the Oklahoma Refiners Waste Control Council
(ORWCC) expressing their concern of analytical variances experienced in
analyzing industrial wastewater samples at individual refineries. This
organization, composed of 11 refinery members, has employed the self-
reporting system for wastewater discharge analyses since 1955. Each
month, the individual refineries report analytical data to the Oklahoma
State Corporation Commission. Their selection as the participating
industrial organization for this study insures the presence of laboratory
experience in wastewater analyses, familiarity with analytical quality
control procedures, and presence of the necessary laboratory supplies
and equipment.
STEERING COMMITTEE
To attain the major and ancillary objectives of a study to define inter-
and intralaboratory repeatability and reproducibility, a steering com-
mittee including one EPA, one state, and three refinery representatives
was formed. Formation of the project plan was the responsibility of the
committee while the liaison and project direction were responsibilities
of Robert S. Kerr Environmental Research Laboratory (RSKERL) personnel.
To prevent fragmentation of information relevant to project completion
an organizational structure, such as the one used in this study shown in
Figure 1, is a necessity for a project such as this involving multiple
organizations.
The steering committee recommended four divisions for the analytical
study: (1) low-level contamination wastewater analysis, (2) participant
seminar to specify uniform procedures, (3) high-level contamination
wastewater analysis, and (4) spike recovery from an industrial waste-
water.
-------
R SKERL
RSKERL ANALYTICAL
LABORATORY
STEERING
COMMITTEE
PARTICIPANTS
NATIONAL PETROLEUM REFINERY-
ORGANIC CHEMICAL WASTES
RESEARCH
Figure 1. Project organizational structure
-------
Participant selection from the council was voluntary, with the agreement
that any refinery who volunteered would, of necessity, have to partici-
pate in the total program. Eight of the eleven member refineries agreed
to participate in the project. Those refiners who did not choose to
participate represent refineries who either contract their wastewater
samples for analysis or could not participate due to internal restric-
tions. Refinery size of the participants varied from 12,000 to 112,000
barrels of crude per calendar day (MBCD). The size distribution of the
participating refineries is shown in Table 1. Size variance of the
participants is an important factor since the population (industrial
participants) involved in the study should represent a spectrum from
small to large. Refinery size also reflects laboratory capabilities
for wastewater analyses since their analytical facilities are dependent
on refinery size.
Two Oklahoma state agencies, which are currently involved in analyzing
industrial wastewaters, requested participation in the program. Those
state agencies were the Oklahoma State Health Department, whose respon-
sibilities include analyses of petroleum refinery wastewater for the
Table 1. CRUDE CAPACITY OF PARTICIPANTS
Refinery
A
B
C
D
E
F
G
H
Thousand Barrels/Calendar Day
12.0
25.0
28.5
29.5
48.5
51.0
87.0
112.0
-------
Oklahoma Corporation Commission (Oklahoma's Enforcement Agency for
petroleum refineries), and the Zoology Department of Oklahoma State
University (OSU). The latter department has performed bioassays on
petroleum refinery wastewaters since 1956 and regularly analyzes refinery
wastewaters.
EPA laboratories participating in the study were the RSKERL at Ada,
Oklahoma, and the Methods Development and Quality Assurance Research
Laboratory (MDQARL) located in Cincinnati, Ohio. MDQARL participation
was primarily that of a referee laboratory while RSKERL duties included
project liaison, sampling, intralaboratory analyses, data analyses,
and report preparation. Related information and background including
study schedule, laboratory equipment needed, correspondence relating
to project liaison, report forms, and analytical instructions are located
in Appendix A.
-------
SECTION V
STUDY OBJECTIVE
The objective of this study was to observe the extent of analytical
variability incurred for a specific industrial wastewater under con-
trolled conditions for intralaboratory (repeatability), interlaboratory
(reproducibility), and spike recovery.
Enforcement of the Federal Water Pollution Control Act (PL 92-500),
which includes a self-reporting system, will be dependent on the pre-
cision and accuracy of the analytical results. Both parties in any
court action should be prepared to furnish sufficient evidence that
their analytical results are correct.
EPA's self-reporting system is based on two major variables: (1) analyti-
cal procedures, and (2) flow measurements. The basis of the reported
values is:
Discharge flow x parameter concentration ,..,.
units of production
This simple formula has two principal potential error sources: (1) ana-
lytical variability, and (2) flow measurement variability. For example,
a 30,000 barrel per day (30 MBD) refinery which has a discharge flow of
one million gallons per day (1 MGD) and a five-day biochemical oxygen
demand (BOD ) concentration of 20 mg/1 may have a discharge load of 5.56
pounds per unit product.
20 x 8.34 x 1 = 5>56 #/MBp (2)
If, because of analytical variability, the analyst reports 25 mg/1 BOD,.,
the discharge load would be 6.95 #/MBD which may constitute a permit
violation. If, in the event of a flow metering error, which is quite
-------
common, a value of 1.2 MGD was used instead of 1.0 MGD, the example
discharge load would be 6.67 #/MBD instead of 5.56 #/MBD. These sources
of error are always present because of the technology and methodology
involved in making the measurements.
Analytical variability, for the purposes of this study, is defined as
the concentration differences reported by participating laboratories for
a common industrial wastewater sample using a standardized testing pro-
cedure. Analytical variability is statistically measured by the standard
deviation test.
Repeatability (precision) is defined as the variability encountered by
a single analyst for the same sample analyzed with the same apparatus.
The statistical measurement for repeatability is also the standard
deviation test.
Reproducibility (precision) is defined as a measurement of the variability
encountered in results from analysts from two or more laboratories for
the same sample and is statistically evaluated using the standard devi-
ation test procedure.
Spike recovery (accuracy) is a percentage measurement of the retrieval
of a specified quantity of chemical added to the industrial wastewater.
10
-------
SECTION VI
STUDY CONDITIONS
A list of factors which affect analytical variability was prepared as a
guide to the study conditions. These analytical variation causes are
listed below:
1. Analysts' laboratory techniques
2. Procedure alteration
3. Poor quality control
4. Improper standards
5. Poor reagents
6. Arithmetic errors
7. Poor sampling techniques
8. Improper equipment, supplies, and glassware
9. Interferences from other contaminents
10. Insufficient time allotted for proper analyses
11. Lack of explicit method
These factors may be summarily reduced to three items of principal
concern: (1) sampling techniques, (2) interferences, and (3) laboratory
factors.
To provide uniformity for the study, the following conditions were
established prior to Phase I:
Sampling—A single industrial wastewater would be sampled and
replicates divided among the participants.
Interfering Compounds—One sample set would not contain sulfide,
while the other sample set would contain sulfide in sufficient concen-
tration to cause interferences.
11
-------
Analytical Procedure--The methods of choice to be followed would
be EPA's "Methods for Chemical Analysis of Water and Wastes," 1971.l
Quality Control—Each participant was requested to use the "Labora-
tory Quality Control Manual," 1972.2> 3
Arithmetic Errors—A self-reporting format was to be provided for
each analyst to submit necessary information to computer checking of
calculations.
Analytical Techniques—Although the Phase I sample set would be
analyzed in the stipulated manner, a participant seminar to provide
additional information on techniques would be provided prior to Phase II
sample set analyses.
Statistics—All data would be sent to the RSKERL, Ada, Oklahoma,
for statistical evaluation.
Parameters to be Determined—Parameters would include COD, phenolics,
total suspended solids (TSS), ammonia nitrogen, and oil and grease.
Those laboratories which had organic carbon instruments were requested
to report total organic carbon (TOC).
Participants--Various industrial and enforcement laboratories would
voluntarily participate. It was agreed to furnish each participant with
a code identification.
"Spike" Sample—Certified "spike" samples prepared by EPA's MDQARL,
formerly Analytical Quality Control (AQC), would be included in each
sample set for Phase I and II. Instructions for ammonia and COD spike
additions were to be included with each sample set.
Referee laboratory services would be provided by the MDQARL. Intra-
laboratory deviation would be determined at the RSKERL by analyzing six
sample sets for each phase of the study. Participants were requested to
begin analysis at specified times to insure uniformity of storage effect
on analyses.
12
-------
SECTION VII
SAMPLING PROCEDURES
Proper sampling of the industrial waste which was subdivided into 20
replicate samples was of utmost importance to the project. If any of
the 20 samples were not representative, the project would be a failure
from the beginning; unfortunately, there are not methods which can be
used to obtain one truly representative grab sample, much less divide
that one sample into 20 parts with the expectation that each of the
20 samples will be equal.
The sampling method used for the study involved centrifugally pumping
the water at a medium volume rate (10 gallons per minute) into a 35-
gallon drum which had an inert inner liner. Calculated amounts of
preservatives were added to the sample and an electric mixer was used
to mix the sample thoroughly. After five minutes of mixing, a replicate
sample was withdrawn through a valve located near the bottom of the
barrel into a previously numbered one-quart plastic or glass container.
The numbered one-quart sample containers were filled at random to mini-
mize the relation of the one-quart sample to the total volume in the
35-gallon drum. The samples were then placed into ice chests. A time
interval of 15-20 minutes was needed to obtain each sample set. Upon
completion of sample preparation for the five parameters, ice was placed
in each chest to assist preservation of samples. A schematic of the
sample equipment is depicted in Figure 2.
Various states of the sampling procedures were photographed for con-
venience of presentations.
Figure 3: Electric centrifugal pump installed on the clarifier
catwalk with the intake pipe submerged about eight inches below the
water surface. The discharge hose is directed to the sample barrel.
13
-------
Centrifigual
Pump
Unco
Foot Valve
Electric Stirrer
Inert Liner
^-Sampling Valve
Ice
Chest
Sampling
Container
Figure 2. Equipment used to obtain representative sample
Figure 4: The inert lined sampling barrel being filled with waste-
water to a pre-determined volume, as measured by depth. The electric
stirrer was started after the correct sample volume was obtained and the
preservative added.
Figure 5: Top view of the barrel shows the mixing pattern created
by the electric stirrer. Continuous stirring was necessary to prevent
settling of suspended solids and separation of oil which usually occur
with refinery wastewaters.
Figure 6: Ice chests were arranged numerically and pre-numbered
samples were placed in the box.
SAMPLE SETS
Sample volumes obtained and the preservative added for each sample set
are shown in Table 2. Each sample set included 12 one-quart sample
containers and the AQC samples for ammonia and COD spikes. Directions
on analytical procedures and quality control were mailed prior to the
sampling so the participants could correspond on problem areas. The
project instructions and format for reporting analyses are located in
Appendix A of this report.
14
-------
Figure 3. Sampling pump installation
Figure 4. Composite sampling container
-------
Figure 5. Top view of composite sampling container
Figure 6. Sample sets for
cooperative laboratories
-------
Table 2. SAMPLE VOLUMES AND PRESERVATIVES
Parameter
COD
TSS
Ammonia
Phenolics
Oil § Grease
Barrel Vol. Amount of Vol. of sample
(gal.) Preservative Preservative Drawn/set
17.0
17.0
22.0
22.0
17.0
Sulfuric Acid
(cone)
none
Mercuric Chloride
Phosphoric Acid
Copper Sulfate
Sulfuric Acid
(cone)
136.0 ml
none
3.52 gm
204,0 ml
68.0 gm
40.0 ml
2 qts.
2 qts.
3 qts.
3 qts.
2 qts.
LOCATION OF SAMPLE POINTS
Wastewater for the first sample phase was obtained from a final clarifier
at a petroleum refinery. The water had been biologically treated and
represented all the water used in the refinery with the exception of
sanitary sewage.
Wastewater for the second sampling phase was obtained from the discharge
of an API type separator. This water was not biologically treated and
contained sulfides which may cause interferences in analytical procedures.
SAMPLE DELIVERY
To insure uniformity of starting time, the participants were instructed
to begin analyses at 10:00 a.m. (CST). Since the Phase I samples were
obtained about 16 hours previous, the sample set could be air delivered
to MDQARL, Cincinnati, Ohio. Due to unforeseen circumstances, the sample
set was not delivered to MDQARL until the following day. A second sample
set was analyzed by RSKERL at the later date for comparative purposes
with the MDQARL samples.
17
-------
The second phase sample set was delivered to the seminar participants.
Initial analyses time was again set for 10:00 a.m. (GST). Correspondence
with the involved airline insured sample receipt of the MDQARL samples
on time.
18
-------
SECTION VIII
PARTICIPANT SEMINAR
An instruction seminar, designed to furnish information on laboratory
techniques, was presented to analysts participating in the study. A
summary of the discussions is presented by parameter in Appendix B.
SAMPLE SET UNIFORMITY
Prime consideration was afforded the problem of obtaining 20 sample sets
which contained virtually the same quantities of contaminants. .As a
preliminary screening device, TOC analysis was performed on suspended
solid samples from six sample sets selected at random. Results of these
analyses are listed in Table 3.
Results of this preliminary screening were sufficient to verify that
uniformity of the 20 sample sets did exist, and that they were accept-
able replicates.
Table 3. TOC ANALYSIS ON REPLICATE SAMPLES
Sample Set No. TOC (mg/1)
01 23.0
03 23.0
07 23.0
09 23.0
14 23.5
18 23.5
19
-------
SECTION IX
STATISTICAL EVALUATION
PRELIMINARY DATA CHECKING
Upon receipt of the completed data sheets from the various analysts and
laboratories, an IBM 1130 computer was used to check each analysis for
errors. When any error was detected, the mistake was either corrected
or the analyst was contacted for an explanation, depending upon the
nature and severity of the error. Whenever necessary the numerical
value of each result was adjusted so that all the results could be
represented with the identical number of decimal places.
OUTLIER REJECTIONS
All data were carefully reviewed and whenever appropriate, data points
were rejected for the following reasons:
1. EPA analytical methodology was not used;
2. The analytical method used did not apply to the range of
concentration in the sample;
3. The analyst reported the value as not reliable;
4. Statistical procedures revealed that the extreme values had
only a small chance of validity and would make a significant
change in the reported variability.
Outlier analyses proceeded in the following manner. Data which were
rejected for reasons 1, 2, or 3 above were first culled. The extreme
values were rejected by applying the two-tail "t" test to the remaining
values at a 95 percent probability level; that is, with a 95 to 5
assurance that the data rejected were invalid and should be rejected.
These values were probably caused by gross instrumental, .chemical, or
human error. In the Phase I study, 32 data points out of a total of
20
-------
221 were rejected. In the Phase II study only five out of a total of
194 were rejected. The reduction in the amount of data rejected from
the Phase I to Phase II indicates the effect of the instruction seminar.
DATA PRESENTATION
Tables 4 through 13 contain the results of the study with the data
grouped according to source. Presented are the analytical results for
each parameter for both phases of the study. The data analyzed at the
RSKERL, Laboratory 01, are in one group and the results are a measure
of the repeatability (intralaboratory variations). The data analyzed at
the other laboratories are presented in the second group, and the results
from this group are a measure of the reproducibility (interlaboratory
variations). Also presented are the average results and standard devi-
ations of the results from each group. Where applicable, the average
spike recovery is presented. Also presented are the average results,
standard deviation, and spike recovery for the data when both groups of
data are combined.
DATA ANALYSES
Inter- vs Intralaboratory Variation
In order to statistically compare the results from the RSKERL and the
other laboratories, the following analyses were carried out. The
variance (a2) of the analytical data from both groups was compared by
using the "F" test. The mean (x) of the data from both groups was
compared by use of the Student "t" test. Both of these tests were
carried out at the 95 percent confidence level. The results are pre-
sented in Table 14 for both Phase I and Phase II.
As expected, the repeatabilities were smaller than the reproducibilities
except for ammonia (Phase I), COD (Phase II), and oil and grease by the
Hexane method (Phase II). In these cases, there was no significant
difference in the variabilities. In the case of ammonia analyses, there
21
-------
were some instrumentation problems in Laboratory 01 which caused a
higher than normal deviation. This problem was eliminated in Phase II.
A comparison of oil and grease analysis for Phase I could not be made
because of the limited amount of data available from Laboratory 01 for
that test.
The normal expectation is that there should be no significant difference
in the means of the two groups of laboratories. In this study, the only
exceptions to this were for COD (Phase I), phenolics (Phase II), and oil
and grease by the Freon method (Phase II). The instruction seminar held
between Phase I and II may have improved the COD measurement. Phenolics
and the Hexane oil and grease tests also improved but not to the same
degree. The analytical methods may need revision for these two parameters.
Spike Recovery
The average spike recoveries in the Phase I study were 87.8 percent for
COD and 22.1 percent for ammonia. In the Phase II the average spike
recoveries were 95.1 percent for COD and 95.6 percent for ammonia.
There was an obvious improvement in spike recovery as a result of the
instruction seminar held between the two phases. This is particularly
true for the ammonia test.
Oil and Grease.Methods
A statistical evaluation was made to determine possible differences in
the two oil and grease analytical methods. These results are presented
in Table 15. A "t" test and "F" test were carried out at the 95 percent
confidence level. The results indicate that there is no significant
difference in the mean value of the two methods. However, it appears
that the Freon method yields a lower standard deviation than does the
Hexane method.
22
-------
Table 4. COD ANALYTICAL METHOD EVALUATION - PHASE I
(Before Instruction)
Lab. No.
Intralaboratory
01
01
01
01
01
01
COD
Results :
77.6
77.6
76.3
83.7
77.6
73.5
mg/1
Duplicate
-
77.6
79.6
75.5
75.5
77.6
73.5
Spiked
106.1
106.1
102.0
104.1
108.2
102.0
Spike
Recovery (%)
92.2
89.0
84.5
79.3
99.0
92.2
Average COD
Standard Deviation
Average Spike Recovery
Interlaboratory Results:
02
05
08
10
11
13
15
16
17
20
71.0
91.8
78.1
89.0
87.6
74.0
122.9
145.0
59.0
111.7
82.8
95.2
78.1
85.0
,7
,2
83.
82.
119.0
145.0
59.0
121.4
Average COD
Standard Deviation
Average Spike Recovery
Combined Results:
Average COD
Standard Deviation
Average Spike Recovery
77.1 mg/1
2.7 mg/1
89.4 percent
110.4
125.8
106.9
125.
115.
.5
.5
108.4
,5
.2
98.6
164.0
83.0
131.5
104.
93.
124.6
96.8
66.3
61.5
77.7
48.5
94.1 mg/1
25.3 mg/1
86.8 percent
87.7 mg/1
21.5 mg/1
87.8 percent
23
-------
Table 5. SUSPENDED SOLIDS ANALYTICAL METHODS
EVALUATION - PHASE I
(Before Instruction)
Lab . No .
Intralaboratory Results :
01
01
01
01
01
01
mg/1
Suspended Solids
17.5
14.0
15.5
13.5
11.0
14.5
Duplicate
14.0
13.5
13.5
12.0
13.5
12.5
Average Suspended Solids = 13.8 rag/1
Standard Deviation = 1.7 mg/1
Interlaboratory Results:
02 22.6 14.0
05 25.5 21.3
06 10.0 9.0
08 10.0 13.0
10 17.0 18.0
11 12.0 9.0
12 18.0 9.0
13 7.0 9.0
15 10.5 8.5
16 11.0 11.0
17 15.0 12.0
20 12.0 12.0
Average Suspended Solids = 13.2 mg/1
Standard Deviation = 4.9 mg/1
Combined Results:
Average Suspended Solids = 13.4 mg/1
Standard Deviation = 4.1 mg/1
24
-------
Table 6. AMMONIA ANALYTICAL METHOD EVALUATION - PHASE I
(Before Instruction)
Lab . No .
Ammonia
mg/1
Duplicate
Spike
Spiked Recovery (%)
Intralaboratory Results :
01
01
01
01
01
01
01
10.4
11.6
10.2
12.6
12.1
12.4
12.6
10.1
11.4
10.4
12.6
12.4
12.2
12.4
Average Ammonia
Standard Deviation
Average Spike Recovery
10.8
12.2
8.7
13.8
13.7
13.5
13.8
11.7 mg/1
1.0 mg/1
= 50.4 percent
40.4
51.5
-117.6
88.2
106.6
88.2
95.6
Interlaboratory Results :
02
05
08
10
11
13
15
16
17
20
12.3
10.0
8.4
12.4
12.1
14.0
12.0
11.4
12.1
11.1
12.4
10.0
8.4
12.4
11.9
11.2
11.9
11.5
12.0
11.0
Average Ammonia
Standard Deviation
Average Spike Recovery
13.6
10.6
7.0
13.6
13.2
5.6
-
12.3
13.4
13.0
11.4 mg/1
1.4 mg/1
= 0.0 percent
91.9
44.1
-102.9
88.2
88.2
-514.7
-
62.5
99.3
143.4
Combined Results:
Average Ammonia
Standard Deviation
Average Spike Recovery
11.5 mg/1
1.2 mg/1
22.1 percent
25
-------
Table 7. PHENOLICS ANALYTICAL METHOD EVALUATION - PHASE I
(Before Instruction)
Lab. No.
Intralaboratory Results :
01
01
01
01
01
01
01
ug/1
Phenolics
15.0
14.4
9.7
14.0
13.9
13.9
13.5
Duplicate
11.3
11.3
10.2
11.2
11.7
9.8
10.8
Average Phenolics = 12.2 pg/1
Standard Deviation = 1.8 yg/1
Interlaboratory Results:
02 14.0 15.0
05 24.0 21.9
08 3.3 4.0
10 20.0 17.0
11 29.0 29.0
13 4.0 4.0
16 15.6 20.7
17 12.5 3.8
20 12.0 13.0
Average Phenolics = 15.2 ug/1
Standard Deviation = 8.1 ng/1
Combined Results:
Average Phenolics = 13.9 ug/1
Standard Deviation = 6.3 yg/1
26
-------
Table 8. OIL AND GREASE METHOD EVALUATION - PHASE I
(Before Instruction)
mg/1
Lab. No. Hexane Duplicate Freon Duplicate
Combined Results:
01 4.7
02 4.5
06 -
08 6.8 8.7
10 6.4
11-
lS 4.5 5.5
15 9.9 8.5
16 4.9
17 -
20 -
Average Hexane
Standard Deviation Hexane
Average Freon
Standard Deviation Freon
Average Oil and Grease
Standard Deviation Oil and Grease
5.1
2.6
6.4
-
6.3
6.1
-
-
6.3
3.3
3.6
=
=
_
-
5.4
-
-
5.4
-
-
-
3.8
4.0
6.5 mg/1
2.0 mg/1
4.9 mg/1
1.3 mg/1
5.6 mg/1
1.8 mg/1
27
-------
Table 9. COD ANALYTICAL METHOD EVALUATION - PHASE II
(After Instruction)
Lab. No. COD
Intralaboratory Results :
01 125.5
01 138.2
01 142.2
01 146.3
01 125.5
01 125.5
mg/1
Duplicate
125.5
138.2
146.3
146.3
125.5
125.5
Spiked
485.8
503.9
503.9
495.8
485.8
485.8
Spike
Recovery (%)
97.4
98.8
106.9
94.5
97.4
97.4
Average COD
Standard Deviation
Average Spike Recovery
Interlaboratory Results:
02
05
06
08
10
11
15
16
17
20
127.5
130.3
116.9
150.2
119.5
144.0
143.9
172.7
139.4
142.6
,3
.5
135.5
130.3
121.0
142.
119.
140.0
143.9
168.7
151.4
142.6
Average COD
Standard Deviation
Average Spike Recovery
Combined Results:
Average COD
Standard Deviation
Average Spike Recovery
134.2 mg/1
9.5 mg/1
98.7 percent
,1
.1
474.
456.
423.4
505.9
478.1
508.0
503.7
494.0
498.0
487.8
92.6
88.1
82.3
97.2
96.9
98.9
97.2
87.4
95.3
93.3
139.1 mg/1
15.0 mg/1
92.9 percent
137.3 mg/1
13.3 mg/1
95.1 percent
28
-------
Table 10. SUSPENDED SOLIDS ANALYTICAL METHOD
EVALUATION - PHASE II
(After Instruction)
mg/1
Lab. No. Suspended Solids Duplicate
Intralaboratory Results :
01
01
01
01
01
01
17.5
21.5
19.0
20.5
20.0
20.5
Average Suspended Solids
Standard Deviation
Interlaboratory Results :
02
05
06
08
10
11
12
15
16
17
20
Average Suspended
Standard Deviation
Combined Results:
Average Suspended
Standard Deviation
18.2
22.3
16.0
17.0
20.0
19.0
9.0
19.5
19.0
20.0
15.0
Solids
Solids
14.5
20.5
19.0
18.5
19.5
20.0
19.3 mg/1
1.8 mg/1
19.1
33.2
13.0
18.0
19.5
21.0
5.5
19.3
18.0
22.0
17.0
18.2 mg/1
5.2 mg/1
18.6 mg/1
4.3 mg/1
29
-------
Table 11. AMMONIA ANALYTICAL METHOD EVALUATION - PHASE II
(After Instruction)
Lab. No.
Ammonia
mg/1
Duplicate
Spike
Spiked Recovery (%)
Intralaboratory Results :
01
01
01
01
01
01
8.7
8.7
8.8
8.7
8.8
8.7
Average
8.8
8.7
8.7
8.9
8.8
8.6
Ammonia
Standard Deviation
Average
Spike Recovery
15.3
15.4
15.4
15.4
15.4
15.3
8.7 mg/1
0.1 mg/1
= 97.4 percent
96.3
98.5
97.8
97.1
97.1
97.8
Interlaboratory Results :
02
05
08
10
11
12
15
16
17
20
8.5
7.6
7.0
8.7
8.5
8.4
9.2
9.3
9.1
9.4
Average
8.4
7.6
5.6
8.7
8.5
8.5
9.1
9.2
9.2
9.1
Ammonia
Standard Deviation
Average
Spike Recovery
14.0
14.1
11.2
15.1
14.8
16.5
15.3
14.3
16.4
15.4
= 8.5 mg/1
0.9 mg/1
= 94.6 percent
96.3
95.6
72.1
94.1
92.6
118.4
90.4
89.0
106.6
90.4
Combined Results:
Average Ammonia
Standard Deviation
Average Spike Recovery
8.6 mg/1
0.7 mg/1
95.6 percent
30
-------
Table 12. PHENOLICS ANALYTICAL METHOD EVALUATION - PHASE II
(After Instruction)
yg/1
Lab. No. Phenolics Duplicate
Intralaboratory Results:
01 5480 5367
01 5470 5500
01 5480 5150
01 5320 5367
01 5320 5283
01 5070 4800
Average Phenolics = 5300.6 yg/1
Standard Deviation = 206.5 yg/1
Interlaboratory Results:
02 6338 6088
05 6600 6720
08 6600 5000
10 4250 4400
11 6150 6200
16 5400 5500
17 5100 5200
20 6080 6550
Average Phenolics = 5761.0 yg/1
Standard Deviation = 795.8 yg/1
Combined Results:
Average Phenolics = 5563.7 yg/1
Standard Deviation = 650.4 yg/1
31
-------
Table 13. OIL AND GREASE METHOD EVALUATION - PHASE II
(After Instruction)
mg/1
Lab. No. Hexane Duplicate Freon Duplicate
Intralaboratory Results:
01 11.6 - 12.0
01 9.3 -
01 10.8 12.4
01 7.1 - 11.7
01 4.8 - 10.5
01 10.0 - 11.4
Average Hexane = 8.9 mg/1
Standard Deviation Hexane = 2.5 mg/1
Average Freon = 11.6 mg/1
Standard Deviation Freon = 0.7 mg/1
Average Oil and Grease = 10.1 mg/1
(both methods)
Standard Deviation Oil and Grease = 2.3 mg/1
(both methods)
Interlaboratory Results:
02 10.4 9.7 10.9 11.6
06 8.5 7.4
08 9.9 10.6
10 10.4 - - 10.3
11 12.3 14.4
12 11.3 12.4
15 10.4 - - -
16 15.1 - - -
17 17.9 19.2
20 10.8 11.6
Average Hexane = 12.5 mg/1
Standard Deviation Hexane = 3.4 mg/1
Average Freon = 10.9 mg/1
Standard Deviation Freon = 2.0 mg/1
Average Oil and Grease = 11.8 mg/1
(both methods)
Standard Deviation Oil and Grease = 2.9 mg/1
(both methods)
Combined Results:
Average Hexane = H-2 mg/1
Standard Deviation Hexane = 3.5 mg/1
Average Freon = H-1 mg/l
Standard Deviation Freon = 1.7 mg/1
Average Oil and Grease = H-2 mg/1
(both methods)
Standard Deviation Oil and Grease = 2.8 mg/1
(both methods)
32
-------
Table 14. STATISTICAL COMPARISON OF LABORATORY 01
WITH THE OTHER LABORATORIES (at the 95% confidence level)
Variance, a2
Analysis
Phase I:
COD
Solids
Ammonia
Phenolics
Phase II:
COD
Solids
Ammonia
Phenolics
F Value
84.80
8.60
1.90
19.08
2.48
7.99
137.00
14.80
Statistically21
Significant
Yes
Yes
No
Yes
No
Yes
Yes
Yes
t Value
2.97
0.52
0.60
1.51
1.13
0.85
1.25
2.22
Mean, x
Statistically
Significant
Yes
No
No
No
No
No
No
Yes
Oil and Grease -
Hexane 1.76
Oil and Grease -
Freon 8.14
No
Yes
2.40
0.92
Yes
No
If the "F" value is statistically significant (Yes), this indicates
that the interlaboratory deviation is larger than the intralaboratory
deviation. Normally it is expected that the "F" will be significant.
Exceptions noted are: Ammonia (Phase I), COD (Phase II), Oil and
Grease - Hexane (Phase II).
If the "t" value is statistically significant (Yes), this indicates
that the average results from Laboratory 01 are different from
those of other laboratories. Cases where the means are different
are: COD (Phase I), Phenolics (Phase II), Oil and Grease - Hexane
(Phase II).
33
-------
Table 15. STATISTICAL COMPARISON OF THE HEXANE AND FREON
METHODS FOR THE ANALYSIS OF OIL AND GREASE
Cat the 95% confidence level)
Variance, a2 Mean x
Statistically Statistically
Phase F Value Significant t Value Significant
I
II
2.19
4.24
No
Yes
2.15
0.11
No
No
34
-------
SECTION X
REFERENCES
1. Analytical Quality Control Laboratory. Methods for Chemical
Analysis of Water and Wastes, 1971. U.S. Environmental Protection
Agency, Cincinnati, Ohio. EPA-16010 07/71. July 1971.
2. Analytical Quality Control Program. Laboratory Quality Control
Manual, 2nd Edition. U.S. Environmental Protection Agency, Ada,
Oklahoma. 1972.
3. Analytical Quality Control Laboratory. Analytical Quality Control
in Water and Wastewater Laboratories. U.S. Environmental Protection
Agency, Cincinnati, Ohio. June 1972.
35
-------
SECTION XI
APPENDICES
Appendix Page
A Related Information and Background 37
Study Schedule 37
Laboratory Equipment Needed 43
Correspondence Relating to Project Liaison 46
Report Forms 52
Analytical Instructions 59
B Seminar Summary 61
Chemical Oxygen Demand 61
Total Suspended Solids 62
Distillation of Ammonia Procedure 62
Phenolics 63
Oil and Grease 63
36
-------
APPENDIX A
RELATED INFORMATION AND BACKGROUND
STUDY SCHEDULE
Purpose of Study
To scientifically determine the repeatability and reproducibility of
quantitatively measuring amounts of impurities in petroleum refining
wastewaters.
Participants in Study
Through the years an excellent working relationship has been established
between the RSKERL and ORWCC on refinery waste treatment research
studies. The participants will be the member refineries, the RSKERL,
and the MDQARL, which will provide referee laboratory service during
the entire study.
Study Program
Split samples will be preserved and sent to the participants for their
analyses. There will be three samples to be analyzed: (1) Low Level
Sample (2) Low Level Sample, and (3) EPA Analytical Quality Control
Sample. On receipt of the first sample set, the analyst will add the
AQC sample (3) to the (2) sample. This sample will be referred to as
the spike sample. Specific instructions will be sent with the sample
set.
Analytical procedures must be run according to the EPA methodology.
If additional procedures are used for analyses, a reference to that
procedure should be recorded.
To facilitate analytical expediency, a range of concentrations will be
supplied with each sample similar to the illustration below.
37
-------
SAMPLE #1
Range (mg/1) Parameter Procedure Page
40-150 Chemical Oxygen Demand High Level 24-28
.01-.3 Phenolics 4-AAP 232-234
5-25 Oil and Grease Hexane Ext. 217-220
20-50 Ammonia Nitrogen Distillation Proc. 134-140
15-40 Total Suspended Solids Non-Filterable 278-279
Analyses will be performed for chemical oxygen demand, phenolics,
oil and grease, ammonia nitrogen, and total suspended solids by the
participating laboratories.
Each participant's analysis will be returned to the RSKERL where
compilation and statistical analyses will be performed on the data.
Training Symposium
To provide analytical uniformity for the second set of samples, a
seminar will be held at the RSKERL in February. At this seminar,
the chemists and technicians, who will be analyzing the second set
of samples, will receive specific instruction on procedure and
analytical techniques. The seminar will be presented on the first
day, and a set of samples will be analyzed by the participants on
the second day. The second set of samples will be delivered the
following week.
It would be desirable if the EPA quality control procedures were
followed during both testing exercises. If, for some reason, this
cannot be done on the first sample set, it should be so noted. To
maintain uniformity, quality control has to be done on the second
sample set.
For your information, Figures 7-11 are photographs typical of each
set-up. Laboratory equipment needed with cost data of the set-up
is also provided. If you do not have the equipment, you may consider
ordering it from a supplier.
38
-------
-
OJ
»y • «n
r -
Figure 7. Chemical oxygen demand set-up
Figure 8. Phenolics set-up
-------
-U
o
Figure 9. Oil and grease (hexane
extraction) set-up
Figure 10. Ammonia nitrogen set-up
-------
DANGEP
Figure 11. Suspended solids set-up
Sampling Dates
It is anticipated the first sampling will be done on January 15;
the training symposium on February 5; and the second sampling on
February 13. The results of the survey will be available by
April 1.
41
-------
Study Outline
Date
December 1973
5
6
13
18
January 1974
15
15
15
18
23
30
February 1974
5
13
13
15
20
27
March 1974
15
25
April 1974
Event
EPA ORWCC Committee Meeting
Presentation of Study to Members
Concurrence of Participants
Study Profile Mailed to Participants
Sample Collected
Sample Transferred to Participants
Analysis Performed
Sample Data Sent to RSKERL
Data Summary Sent to Participants
Committee Meeting for Second Phase
Training Seminar
Second Set of Sample
Analysis Performed
Data Sent to RSKERL
Data Summary Sent to Participants
Committee Meeting
First Draft Report
Rewrite
Present to ORWCC at OSU
42
-------
LABORATORY EQUIPMENT NEEDED
Equipment Cost
Chemical Oxygen Demand:
1. Reflux apparatus
500 ml Erlenmeyer flask with 24/40 standard $ 3.50
taper joint
or 300 ml boiling flask with 24/40 STJ 3.34
2. Allihn condenser with 24/40 STJ at bottom, 12" or 13.00
equivalent
or Freidrichs condenser with 24/40 STJ at 16.33
bottom
3. 1 - 25 ml burets (teflon plug) 12.00
4. Assorted pipets
5. Glass beads
Oil and Greases:
1. Soxlet extractor (Corning 3740 or equivalent) 21.00
medium
2. Soxlet thimbles to fit extractor (33x80) fat free 6.00
(box of 25)
3. Flask 125 ml Corning No. 4100 or equivalent 3.75
4. Allihn condenser (bulb type) to fit extractor with
45/50 STJ bottom
5. Source of vacuum
6. Buchner funnel 12 cm or on which will accept 13.00
7. Filter paper - Whatman No. 40 or equivalent 2.11
11 cm (pkg. 100)
8. Muslin cloth discs 11 cm (cut from muslin bought
at drygoods store)
9. Filter flask to fit buchner
Funnel 1000-2000 ml 8.00
10. Stopper to adapt funnel to filter flask .50
11. Filter aide (Hyflo Super-eel) - John Manville 2.25
Corp. or equivalent - 1 Ib.
12. Oven set for 103° C
43
-------
13. Water bath (may be made)
14. Steam bath
15. Glass beads
16. Source of heat for extraction steam bath or
standard method advocates a heating mantle
for flask, etc.
17. Balance (analytical)
18. Dessicator for drying + plate (price varies)
Ammonia Nitrogen:
All glass distillation apparatus with 800-1000
ml flask or Kjeldahl distillation rack. The
phenol distillation equipment may be used.
(For price see Phenol) If phenol distilling
units are used, ignore 3, 4, and 5.
1. Titration setup utilizing a buret or the
colorimetric test utilizing nessler tubes
and a spectraphotometer. Colorimetric
should be utilized for low levels.
2. Nessler tubes (fall form, 50 ml)
3. Kjeldahl distilling rack (2 units)
4. Kjeldahl flask 800-1000 ml
5. Kjeldahl bulb, Iowa Seate type or equivalent
6. Erlenmeyer flasks (500 ml) glass stoppered
7. Various standard lab glassware
Phenol:
1. Distilling apparatus with Graham condenser,
Corning No. 3360 or equivalent 1000 ml
2. 1000 ml beakers
3. pH meter
4. 500 ml graduated cylinders or
500 ml volumetric flasks
5. Photometric equipment for work @ 460 mu
6. Filter paper 11 cm
7. Separatory funnels 1000 ml Squibb form with
ground glass stoppers and teflon stopcocks
130.00-250.00
150.00-200.00
Price will
depend on
heat source.
250.00-1000.00
approx. 19.00
each 2.50
300.00-400.00
6 for 15.00
each 5.00
each 5.00
each 29.80
approx. 2.00
9.00
5.00
various prices
approx. 2.00
13.00
44
-------
8. Nessler tubes 50 ml tall form (matched) 2.50
9. Buret 10 ml 20.00
10. Balance (see Oil and Greases)
11. Various common lab glassware, such as pipets, etc.
Solids, Non Filterable - Suspended:
1. Glass fiber filter discs, 4.7 cm or 2.2 cm without
organic binder. Reeve Angel type 934-H or
984-H, Gelmon type A, or equivalent.
2. Filter holder, membrane filter funnel or Gooch
crucible adaptor.
3. Suction flask 500 ml approx. 7.00
4. Gooch crucibles 25 ml if 2.2 cm filter used 1.40
5. Drying oven 103°-105° C price varies 80.00-500.00
6. Desiccator depends on size, approx. 20.00
7. Analytical balance 200 gm capacity price varies 800.00-1000.00
capable of weighing to 0.1 mg. (An
automatic, such as Mettler or equivalent)
45
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
^ ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
P. O. BOX 1198
ADA, OKLAHOMA 74820
January 18, 1974
Dear Participant:
On January 29 the first samples for the Round Robin Study
will be distributed to the various participating laboratories.
All samples should be delivered in time to begin the analyses
by 11:00 a.m. at the latest.
The samples will consist of:
1. A petroleum refinery effluent which has been sampled in
a manner which should produce comparable analytical results
on all parameters.
2. A duplicate of No. 1 which should be run to satisfy data
requirement concerning the reproducibility of results within
a given laboratory.
3. A duplicate of No. 1 which will receive a spike or addition
of a standard substance to test for interferences in the
refinery effluent.
The samples are to be run for: (a) suspended solids or non-
filterable solids, (b) phenols, (c) ammonia-N, (d) chemical oxygen
demand, and (e) oil and grease.
Each sample will be appropriately preserved and will be tagged
to indicate the proper analysis.
At the present time, it is feasible to spike the samples for
only ammonia-N and chemical oxygen demand, thus these, will be the
only two analyses which will be required on the spiked samples.
46
-------
The samples and number of analyses to be performed are:
Set No. 1
No. S (Lab No.) L refinery effluent for suspended solids = 1
No. P (Lab No.) L refinery effluent for phenol = 1
No. N (Lab No.) L refinery effluent for ammonia-N = 1
No. C (Lab No.) L refinery effluent for COD = 1
No. 0 (Lab No.) L refinery effluent for oil and grease = 1
Total Analyses 5
Set No. 2
No. S (Lab No.) D duplicate refinery effluent for suspended solids = 1
No. P (Lab No.) D duplicate refinery effluent for phenol = 1
No. N (Lab No.) D duplicate refinery effluent for ammonia-N = 1
No. C (Lab No.) D duplicate refinery effluent for COD = 1
No. 0 (Lab No.) D duplicate refinery effluent for oil and grease = 1
Total Analyses 5
Set No. 3
No. N (Lab No.) S duplicate refinery effluent and spike for
ammonia-N = 1
No. C (Lab No.) S duplicate refinery effluent and spike for COD = 1
Total Analyses 2
The methods to be used will be found in the EPA manual "Methods
for Chemical Analysis of Water and Wastes," 1971, which you should have
already received. As a suggestion, the parameters and applicable tests
would be:
Phenols - EPA Manual, page 232 - use procedure of Standard Method
13th Edition beginning on page 502 with distillation and use chloroform
extraction, pages 504 through 506.
Suspended Solids - EPA Manual, pages 278 through 279.
Ammonia-N - EPA Manual, pages 134 through 140 beginning distilla-
tion and using either the titration or the colorimetric procedures.
(The RSKERL plans to use the titration procedure.)
Oil and Grease - The Freon (trichlorotrifluoroethane) extraction
method of Standard Methods 13th Edition, page 254 ^is recommened unless
reagents and glassware are not available. Hexane extraction should be
the next choice, EPA Manual, pages 217 through 220. If possible, do
onfe oil and grease sample by each method.
47
-------
Chemical Oxygen Demand - EPA Manual, page 19, gives the low
level procedure. It is advised that the test prescribed is Standard
Methods, pages 495 through 499, using the higher normality dichromate
and ferrous ammonium sulfate be the method of choice for the samples
which we will be analyzing.
Instructions for Chemical Oxygen Demand Spike of Refinery Effluent
A vial of concentrate will be received with the set of samples
to be run for the Round Robin Study.
The demand concentrate is prepared by dissolving known amounts
of analytical reagent-grade glucose and glutamic acid in distilled
water. Each concentrate can be analyzed for biochemical oxygen
demand (BOD), chemical oxygen demand (COD), and total organic carbon
(TOC). COD is the only analysis we are requesting at this time,
however.
When diluted according to instructions, the COD of the sample
will be increased in the range of 0 to 50 mg/1.
The concentrate has been preserved by autoclaving the sealed
ampul, after which repeated analyses over a period of weeks by the
Analytical Quality Control Laboratory of EPA located at Cincinnati,
Ohio, were conducted to insure stability of the concentrate. However,
the concentrate must be diluted and analyzed immediately after the
ampul is opened to avoid degradation of the compounds.
When you are ready to begin the analysis, open the ampul by
snapping the top off at the break area on the neck and dilute with
the No. C (Lab No.) S refinery effluent as follows: Dilute 15 ml
of concentrate 1 to volume in a 500 ml volumetric flask with the
refinery effluent.
The analysis for COD may now be carried out as described in
the EPA Manual and Standard Methods 13th Edition.
Instructions for Ammonia-N Spike of the Refinery Effluent
A vial of concentrate will be received with the set of samples
to be run for the Round Robin Study.
The concentrate is prepared by dissolving known amounts of
analytical grade chemicals in distilled water for exact and pre-
planned concentrations. When diluted according to instructions,
the concentration of ammonia-N added to the sample will fall within
the range of less than 3 mg/1 as NH3-N.
48
-------
Each sample will be analyzed at least six times at the RSKERL to
determine variation.
The Analytical Quality Control Laboratory of EPA, located at
Cincinnati, Ohio, who prepares these standards, has repeatedly analyzed
the standards over a period of months to assure stability.
The first vial you recieve will be designated Concentrate 1 Nutrients,
and will contain inorganic nitrogen and phosphorus forms. Laboratories
interested in their accuracy in analyzing for nitrate-N and orthophosphate
can utilize the concentrate to investigate these values as well. Analysis
for ammonia-N is the only analysis we are specifying at this time, however.
The range for nitrate-N will fall below 2.0 mg/1 and orthophosphate as
P will be below 1.0 mg/1 for those who are interested.
All constituents are present in a soluble form. Do not filter the
concentrates. They have been preserved so that no changes will occur.
However, the preservative treatment is not effective after dilution.
Therefore, the samples must be analyzed immediately after opening and
diluting.
To begin the analysis, open the ampul by snapping the top off the
scored break area on the neck and dilute a 10 ml aliquot of the concen-
trate to 500 ml with the refinery effluent specified for ammonia-N
analysis spike (No. N (Lab No.) S). Approximately 25 ml of the concen-
trate is supplied. The 500 samples may now be analyzed for ammonia
nitrogen as required.
Yours truly,
49
-------
*
\ l/ g UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
^^^vP
•V pROit0 ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
P. O. BOX 1198
ADA. OKLAHOMA 74820
January 28, 1974
Dear Participant:
You will find attached to this letter several blank forms for
reporting the following analyses:
1. Chemical oxygen demand - 3 copies
2. Suspended solids - 2 copies
3. Oil and grease (Hexane) - 2 copies
4. Oil and grease (Freon) - 2 copies
5. Ammonia (Spectrophotometric) - 3 copies
6. Ammonia (titrimetric) - 3 copies
7. Phenolics - 2 copies
We would like to request that you use these forms in reporting
the results of your laborator's analyses. Your cooperation will
greatly facilitate the statistical analyses of the results. One
form should be completed for each sample analyzed. (There will be
an excess of oil and grease and ammonia forms as there are two methods
for running both of these analyses.)
Each form will request the following information:
Storet Number - Already filled in on the form.
Laboratory Code - Already filled in on the form.
Date - The date this analysis was started. Please use the
following format: 01/29/74 for January 29, 1974.
Sample Number - As printed on the sample tag and container. An
example of a sample number is 0-25-L. The first letter of the sample
number signifies the analysis. 0 is for oil and grease, N is for ammonia,
P is for phenol, C is for COD, and S is for solids. The second set of
digits is the laboratory code. The last letter will be either L, D,
or S. Sample numbers ending with "S" are to be spiked.
50
-------
Time at Start of Analysis - Please use military time (1300 for
1:00 p.m.)
Time at Completion of Analysis - Military time. If the analysis
is completed on a day other than the starting date, please make note
of this in the comments section.
Is Sample Spiked - Check yes or no.
Method Number - Already filled in on the form.
Analytical Results - Detailed information is requested for each
analysis. This will help verify the results. Completion of this
section will eliminate possible questions about simple arithmetic
errors and will provide a complete record of the test.
Comments - Any additional information that the analyst desires
to report.
Analyst's Initials
The samples in this study will be taken from a refinery waste
stream. We expect the analyses of this stream to fall within the
following ranges:
Ammonia 10 - 30 mg/1
COD 75 - 125 mg/1
Phenolics 0.05 - 0.3 mg/1
Suspended Solids 20 - 40 mg/1
Oil and Grease 5-25 mg/1
As this waste stream varies from day to day, we cannot be certain
that the results will always fall within this range.
Please use the enclosed self-addressed envelope to return your
results. We would appreciate it very much if you could mail out the
results by February 1, 1974.
Yours very truly,
Enclosures
51
-------
CHEMICAL OXYGEN DEMAND
ANALYSIS SHEET
Storet Number: 00335
Laboratory Code: 16
Date:
Sample Number:
Time at Start of Analysis: _
Time at Completion of Analysis: _
Is the Sample Spiked: Yes _ No _
Method Number: _ 1 _
Analytical Results:
Normality of K2Cr_0_ _ N
mis K^Cr-O- for normality titrations _ mis
mis Fe (NH.) (S0.)2 used for normality titrations _ mis
Normality Fe (NH4)2 (S04)2 =
mis K2Cr207 x (Normality K2Cr207) N
mis Fe (NH4)2 (S04)2 Used
mis K_CrJD_ used in test _ mis
mis sample used _ mis
mis Fe (NHJ2 (SOJ2 used in blank titration _ mis
ml Fe (NHJ2 (S04)2 used in sample titration _ mis
mg/1 COD = [(mis Fe (NH4)2 (S04)2 for blank -
mis Fe (NH4)2 (SO^) x (N of Fe (NH4)S04)2
x 8000)] * mis of sample used = _ mg/1
Comments :
Analyst's Initials _
52
-------
NON-FILTERABLE (SUSPENDED) SOLIDS
ANALYSIS SHEET
Storet Number: 00550
Laboratory Code: 16
Date:
Sample Number:
Time at Start of Analysis:
Time at Completion of Analysis:
Is the Sample Spiked: Yes No
Method Number: 1
Analytical Results:
Volume filtered ml
Dried weight of crucible, mat, and residue gms
Dried weight of crucibles and mat gms
Weight of residue gms
mg/1 suspended solids =
weight residue x 1,000,000 _ mg/1
volume filtered
Comments:
Analyst's Initials
53
-------
OIL AND GREASE BY HEXANE EXTRACTION
ANALYSIS SHEET
Storet Number:
Laboratory Code:
Date :
Sample Number:
Time at Start of Analysis:
Time at Completion of Analysis:
Is the Sample Spiked:
Method Number:
Analytical Results:
Volume of sample used
Weight of flask and oil
Tare weight of flask
Weight of oil
•*
Comments :
QQ55Q
16
Yes
No
mis
grams
grams
grams
Analyst's Initials
54
-------
OIL AND GREASE BY FREON EXTRACTION
ANALYSIS SHEET
Storet Number: 00550
Laboratory Code: 16
Date:
Sample Number:
Time at Start of Analysis:
Time at Completion of Analysis:
Is the Sample Spiked: Yes No
Method Number: 2
Analytical Results:
Volume of sample used mis
Weight of flask + residue gms
Weight of flask gms
Weight of material found in blank run on solvent.
(The same amount of solvent used as sample is
placed in a tared flask and evaporated. The
increase in wt. constitutes a blank.) gms
mg/1 oil or grease = [(flask + residue) gms -
(wt. flask) gms - wt. Blank, gms] x 1,000,000 mg/1
ml sample
Comments:
Analyst's Initials
55
-------
AMMONIA NITROGEN
(Spectrophotometric Procedure)
ANALYSIS SHEET
Storet Number: 00610
Laboratory Code: 16
Date:
Sample Number:
Time at Start of Analysis:
Time at Completion of Analysis:
Is the Sample Spiked: Yes No
Method Number: 2
Analytical Results:
*
• Volume of the 500 mis of distillate which is
Nesslerized. mis
Standard Curve: mg NHLN/1 Optical Density (Adsorbance)
Optical density (adsorbance) for sample
mg NH,H from standard curve. mg/1
mg/1 ammonia-N = (mg NHgN/1) x 1000 _ -
.8 x volume Nesslerized
Comments:
Analyst's Initials
Use 400 ml of sample, distill and dilute to 500 ml. This is the
distillate.
56
-------
AMMONIA NITROGEN
(Titrimetric Procedure)
ANALYSIS SHEET
Storet Number: 00610
Laboratory Code: 16
Date:
Sample Number:
Time at Start of Analysis:
Time at Completion of Analysis:
Is the Sample Spiked: Yes No
Method Number: 1
Analytical Results:
Volume of sample distilled mis
Volume of 0.01 N H2S04 used mis
,, . (mis of H-SOJ x .14 x 1000 ...
mg/1 ammonia = _ 2 4^ = mg/1
(mis of sample)
Comments:
Analyst's Initials
57
-------
PHENOLICS
ANALYSIS SHEET
Storet Number: 32730
Laboratory Code: 16
Date:
Sample Number:
Time at Start of Analysis:
Time at Completion of Analysis:
Is the Sample Spiked: Yes No
Method Number: 1
Analytical Results:
Volume sample distilled ml
Volume extracted ml
Standard Chart: yg phenol Absorbance
Absorbance of sample
yg phenol from calibration curve yg
,, , , (yg phenol) x 1000 ,,
yg/1 phenol = ml of original sample extracted = pg/1
Comments:
Analyst's Initials
58
-------
ANALYTICAL INSTRUCTIONS
Phase II - Round Robin Study
The samples to be run in the second part of the Round Robin Study have
been fixed with the EPA approved preservatives for the particular
analyses indicated on the container.
You will spike only the samples for chemical oxygen demand and ammonia
nitrogen as was done on the previous round of the study.
This set of samples represents a waste which should produce greater
quantities of the test substances than the sample taken for Round I.
The waste should also contain greater numbers of interfering substances.
The contaminants in this waste have been previously analyzed over a
period of time and have been found to fall within the following ranges:
Phenol 1-10 mg/1*
Ammonia-N 10-40 mg/1
Suspended Solids 20-60 mg/1
Chemical Oxygen Demand 200-500 mg/1
Oils 20-100 mg/1
* Phenols are in the milligram not microgram range.
Note: Sulfides are present in this sample in the range of 25 mg/1 at
the time of sampling.
The samples are to be analyzed using the EPA methodology specified in
Round I.
All samples will be identified as specified in the information you
received for Round I. The sample information sheets will be identical
to the Round I study.
Since this sample batch represents higher levels of contaminants to
be found in refinery wastewaters, the spikes are accordingly higher.
-------
By following the outlined spiking procedure, values in this specified
range should be produced in addition to the background values.
Chemical Oxygen Demand 200-500 mg/1
Ammonia Nitrogen 3-10 mg/1
Procedure for Ammonia-N Spikes
To begin the analysis, open the concentrate 2 ampul by snapping the
top off at the break area on the neck. Dilute 10 ml of the concen-
trate to 500 ml in a volumetric flask using the wastewater sample
indicated for the spike. Carry out the analysis on this spiked sample
in the same manner you are conducting the rest of the analytical
testing for ammonia-N.
Chemical Oxygen Demand Spike
Begin the analysis by snapping off the neck of the ampul for chemical
oxygen demand which has been supplied with your sample set. Dilute
15 ml of the ampul's contents to volume in a 500 ml volumetric flask
using the contents of the sample indicated for COD spike. The COD
analyses may now be performed in the prescribed manner.
If you have any questions, please call Billy DePrater or Bob Benefield
at the Robert S. Kerr Environmental Research Laboratory, 405/332-8800,
extension 335 and 235, respectively.
You may also wish to speak to the Resident Specialist about a particular
parameter. They are as follows:
Fred Pfeffer Oil and Grease Ext. 305
Kenneth Jackson COD Ext. 212
Roger Cosby Phenols Ext. 210
Mike Cook Suspended Solids and Ext. 300
Ammonia Nitrogen
Clarence Edmonson Ammonia Nitrogen Ext. 238
60
-------
APPENDIX B
SEMINAR SUMMARY
An invitation was extended to the analysts to participate in a seminar
designed to furnish information on laboratory techniques for each
parameter. The premise of the seminar was to minimize inherent errors
in laboratory technique by reviewing stepwise procedural steps. Besides
the obvious technique errors for each test procedure, analysts devised
shortcuts which may magnify the obvious technique errors. The seminar
was held at RSKERL with 24 in attendance. Instructors were the RSKERL
analysts who participated in Phase I of the study. A tape recording
was made of each lecture and a synopsis of laboratory technique problem
areas by parameter follows:
CHEMICAL OXYGEN DEMAND
1. All glassware needs to be chemically clean. Steam out con-
densers and flasks with 50 percent sulfuric for one and one-half hours,
cool, wash down condensers with distilled water and cap the condensers
with aluminum foil. Wash the flasks with distilled water and cap.
2. Fix the samples with sulfuric acid.
3. Use correct volume for analysis.
4. Do not use graduated cylinders to measure the sample volume.
Use large base pipettes.
5. Blend sample.
6. Use automatic pipette for potassium dichromate addition.
7. When adding sulfuric acid, keep the flask cool to the touch.
8. Wash down side of flasks.
9. Boil sample for two hours only.
10. Allow refluxed sample to cool to room temperature.
11. Results may be unreliable if it takes less than 10 ml of
ferrous ammonia sulfate to titrate the dichromate to the end point.
61
-------
12. Analyze reagents daily for normality.
13. Set up a blank, ferrous ammonia sulfate standard unknown and
a standard, one spike per eight samples and one duplicate per eight
samples for quality control.
14. Check arithmetic errors.
15. Reflux at 145° C.
16. There should not be more than 0.2 ml of potassium dichromate
difference when titrating the blank and the ferrous ammonia sulfate
standard.
TOTAL SUSPENDED SOLIDS
1. Thoroughly shake the sample before decanting.
2. Obtain constant weight on crucibles.
3. Always use tongs when transferring crucibles.
4. Correct oven temperature.
5. Rinse graduate with distilled water.
6. Crucibles should be chemically clean, preheated to correct
temperature and cooled in a desiccant.
DISTILLATION OF AMMONIA PROCEDURE (1.0-25.0 mg/1 of ammonia)
1. Ammonia free distilled water.
2. Correct normalities for reagents.
3. Fresh indicator is needed for good end point.
4. Check reagents prior to analysis for purity.
5. Boil out glass system prior to analysis.
6. Use correct aliquot of sample.
7. Check arithmetic calculations.
8. Precision and accuracy is based on sample size.
9. Titration method is better than nesserlization for industrial
waste because of interfering ions.
62
-------
PHENOLICS
1. Standardize stock standard by titration w/Na«S000.
L. £. O
2. Stock standard should be refrigerated and standardized every
two weeks.
3. Working phenol standard should be prepared only on the day
it is used. Degradation occurs!!
4. Samples should be analyzed as quickly as possible. Within
four hours without preservation. Within 24 hours with preservation.
5. Always run standards for calibration curve through the phenol
still the same way samples are (preservation included).
6. Guard against thermal degradation of distillate (i.e., be
sure distillate is cool).
7. Buffer sample pH to 10.0 for sample prior to color development.
8. Be sure CHC1, extract is dry (free of H_0) before reading on
spectrophotometer.
OIL AND GREASE
Freon Method
This is approved method of EPA in Standard Methods for oil and grease
of concentrations of <1000 mg/1.
Sample--
1. Should be collected in glass bottle (greater than 1 L) with
screw cap with teflon liner.
2. 5 ml of H7So. (1.1) should be added at time of collection.
3. pH should be <3, lower pH does not have any adverse effect.
4. When sampling any floating oil, film should be excluded.
5. Sample should be refrigerated from sample time through
analysis.
6. Analysis should be started within 24 hours.
7. Glass sample bottle should have a one liter mark so as to
obtain correct volume.
8. Suggestion made to use pH paper to adjust pH of sample.
63
-------
Procedure--
1." Technique from Standard Methods.
2. Extract oil and grease with 40 ml of Freon; do this two times
and combine, also wash sample bottle with 15-20 ml of Freon and combine
with extract.
3. Distill over H20 bath (80°-85° C).
4. Dry outside of flask, dessicate and weigh.
5. Use controlled HLO bath, then dry air, then illuminating gas
to displace the gaseous content of the flask.
6. Use tongs to handle distilling flask.
7. Obtain constant weight of flask by repeating HO bath, dessi-
cation, weight of flask three times.
8. Blank should be obtained by carrying 100 ml of Freon through
the same procedure as the sample; it should also be dried, dessicated,
and weighed three times to obtain a constant weight. The blank is then
substracted from the sample weight.
9. Volume of sample is obtained by putting sample in graduated
cylinder after extraction has been done. This is more accurate than
relying on marked line on sample bottle.
Advantages —
1. Freon advantages over petroleum ether or Hexane method:
a. Freon is non-flammable,
b. Method is faster, and
c. Freon is heavier thus being the bottom phase in separatory
funnel.
Disadvantages and Problems —
1. Prone to form emulsion with oil and grease of concentration
>1000 mg/1.
2. > oil and grease concentration = > emulsion problem.
3. Gravimetric technique—thus have problems with analytical
balance; large flask used to distill has weight of 100-120 gm.
64
-------
4. Oil and grease concentration of <5 mg/1 is unreliable.
5. Constant weight of 200 ml distilling flask is +_ 3-4 mg/1
(limitation to test).
6. Freon boiling point = 47° C--one will have problem with the
concentration of contaminates in the solvent every time the container
is opened and freon vapors escape.
7. Always siphon freon with glass tube, not tygon tubing, etc.
8. Emulsion problems can be improved by pH adjustment, salting,
technique, or by filtration.
9. From references in Standard Methods (Taras and Blom) part
of oil content is finely dispersed in sample. If sample is milky after
extraction try to use Nacl to salt out oil.
10. 300 gm/1 Nacl to acidified sample is mentioned in references.
(5 gm/1 Nacl is said to be just as effective—from EPA Lab, Edison,
this can be verified by data from the RSKERL, Ada.)
11. If the salting out method is employed, use 3 to 4 aliquots
of freon instead of only 2. Be sure to include this additional amount
of freon in the blank calculations.
12. Or use a long-stemmed funnel with Na_SO.(l" layer) on glass
plug. Wash with freon.
Miscellaneous--
1. Due to solvent specificity, this calculation of oil and grease
should be reported as Freon extractable oil and grease content, not
oil and grease.
2. Not recommended to reuse freon by condensing it.
Discussion from Group—
1. One man made mention of picking up weight which he believed
came from acid. He said H-SO. residue would add weight to sample.
(Instructor questioned.)
2. Another man had problems with algae contamination of his
samples. He tried to eliminate problem with use of gooch and glass
wool, washed with freon.
65
-------
3. Instructor stated work was being done on possible use of
CC1. in same technique.
Soxhlet Method
As given in EPA Manual for oil and grease 1000 mg/1:
1. Is a general physical filtration method using muslin and
filter aid suspension.
2. Muslin + filter aid is extracted in soxhlet with hexane.
3. Procedure calls for 20 cycles/hr. Instructor said is was
only possible to obtain 14 cycles/hr.
4. Instructor said is was possible a better method would be
to add one mg filter aid to sample and shake before filtration, thus
adding additional contact with filter aid.
5. Use hand covering to keep oil on hands from being added to
sample. (Filter paper used to clean sample bottle and should be
handled by something other than hands.)
6. If one uses plastic golves, one should run a blank to see
if hexane is extracting any plastic.
7. Also, at end of procedure when drying flask, one should
use dry air when solvent is being removed.
66
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-76-234
3. RECIPIENT'S ACCESSION>NO.
4. TITLE AND SUBTITLE
Analytical Variability of Five Wastewater Parameters-
Petroleum Refining Industry
5. REPORT DATE
September 1976
(I ssuing date*)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Leon H. Myers, Thomas E. Short, Jr., Billy L. DePrater
and Fred M. Pfeffer
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Robert S. Kerr Environmental Research Lab.
Office of Research and Development
U.S. Environmental Protection Agency
Ada, Oklahoma 74820
- Ada, OK
10. PROGRAM ELEMENT NO.
1BB036
11. CONTRACT/GRANT NO.
NA
12. SPONSORING AGENCY NAME AND ADDRESS
Same as above
13. TYPE OF REPORT AND PERIOD COVERED
Final report
14. SPONSORING AGENCY CODE
EPA/600/15
15. SUPPLEMENTARY NOTES
Prepared in cooperation with the Oklahoma Petroleum Refiners Waste Control Council.
16. ABSTRACT
Samples were divided among 12 laboratories to be analyzed for chemical
oxygen demand, suspended solids, ammonia nitrogen, phenolics, and oil and
grease. The Robert S. Kerr Environmental Research Laboratory analyzed six
sample sets to determine intralaboratory deviation (repeatability), while
the other participating laboratories analyzed single samples to provide data
for interlaboratory deviation (reproducibility) determinations. Study
results are expressed in terms of averages, standard deviation, and spike
recoveries for intralaboratory, interlaboratory, and combined evaluations.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. cos AT I Field/Group
Chemical analyses, Data processing,
Chemical Oxygen Demand
Ammonia Nitrogen
Phenolics
Oil and Grease
Suspended Solids
09B
07A
18. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (ThisReport)
21. NO. OF PAGES
75
20. SECURITY CLASS (Thispage)
22. PRICE
EPA Form 2220-1 (9-73)
67
GOVERNMENT PRINTING OFFICE: 1977-757-056/5533 Region No. 5-11
------- |