EPA
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
'RTP, NC 27711
EMB Report 91 -IWW-07
MAY 1992
Air
INDUSTRIAL WASTEWATER
FIELD EVALUATION OF
WASTEWATER DRAIN SYSTEM
EMISSION TEST REPORT
ROHM AND HAAS COMPANY
BRISTOL FACILITY
BRISTOL, PENNSYLVANIA
-------
EMB Report #91-IWW-07
Industrial Wastewater Field Evaluation
of Wastewater Drain System
Emission Test Report
Rohm and Haas Company
Bristol Facility
Bristol, Pennsylvania
EPA Contract # 68D90054
Work Assignment #1-33
Prepared for:
Emission Measurement Branch
TSD/OAQPS
Research Triangle Park, NC 27711
Prepared by:
Radian Corporation
155 Corporate Woods, Suite 100
Rochester, NY 14623
May 1992
-------
TABLE OF CONTENTS
Section
1.0 INTRODUCTION 1-1
1.1 Project Description 1-1
1.2 Project Objectives 1-2
1.3 Project Organization and Responsibilities 1-2
1.4 Test Plan Changes 1-3
1.5 Report Organization 1-4
2.0 SUMMARY AND DISCUSSION OF RESULTS 2-1
2.1 Wastewater Characterization 2-1
2.2 Tracer Chemical Feed Rates 2-1
2.3 Wastewater Flow Measurements 2-14
2.4 Temperature Measurements 2-17
2.5 Tracer Concentrations in Wastewater Samples 2-17
2.6 Headspace Gas Samples 2-22
2.7 Mass Flow Rates at Wastewater Sampling Locations 2-24
2.7.1 Determination of Mass Flux at Locations A and D . . . . 2-24
2.7.2 Determination of Mass Flux at Locations B and C . . . .2-24
2.8 Collection System Air Emission Rates 2-26
2.9 Evaluation of Collection System Losses 2-29
3.0 FACILITY DESCRIPTION 3-1
4.0 SAMPLING LOCATIONS 4-1
4.1 Wastewater Sampling Locations 4-1
4.1.1 Location A 4-1
4.1.2 Location B 4-4
4.1.3 Location C 4-4
4.1.4 Location D 4-4
4.2 Headspace Gas Sampling Locations 4-8
4.2.1 Location A 4-8
4.2.2 Location B 4-8
4.2.3 Location C 4-8
4.2.4 Location D 4-8
ii
-------
TABLE OF CONTENTS (Continued)
Section
4.2.5 L/V-1 4-8
4.2.6 L/V-2 4-12
4.2.7 L/V-3 4-12
4.2.8 L/V-4 4-12
4.2.9 L/V-5 4-12
4.2.10 L/V-6 4-12
4.2.11 L/V-7 4-12
4.2.12 L/V-8 4-12
4.2.13 L/V-9 4-13
4.2.14 L/V-10 4-13
5.0 SAMPLING AND ANALYTICAL PROCEDURES 5-1
5.1 Inoculation of the Waste Stream 5-1
5.1.1 Reagents 5-1
5.1.2 Tracer Feed Systems 5-4
5.1.3 Tracer Solution Flow Determination 5-4
5.2 Sampling Procedures 5-5
5.2.1 Volatile Wastewater Samples 5-6
5.2.2 Semi-Volatile Wastewater Samples 5-7
5.2.3 Metals Wastewater Samples 5-7
5.2.4 Headspace Gas Samples 5-8
5.2.5 Leak/Vent Gas Samples 5-8
5.3 Field Measurements 5-8
5.4 Analytical Procedures 5-10
6.0 QUALITY ASSURANCE/QUALITY CONTROL RESULTS 6-1
6.1 Field Measurement Quality Control 6-1
6.2 Liquid Sample Quality Control Results 6-2
6.2.1 Blanks 6-3
6.2.2 MS/MSD Results 6-6
6.2.3 Field Duplicates 6-8
6.2.4 Add Spike Sample Results 6-8
6.2.5 Surrogate Spikes 6-10
6.2.6 Metals Quality Control and Standard Addition Samples . 6-12
iii
-------
TABLE OF CONTENTS (Continued)
Section
6.3 Canister Sample QC Results 6-12
6.4 Sample Chain-of-Custody 6-16
6.5 Field Records 6-16
APPENDICES (Bound Separately)
A: Project Participant List
B: Calculation Sheets
C: Analytical Results
D: Sample Collection Matrix
E: Analytical Matrix
F: Sample Chain-of-Custody Forms
G: Field Logbooks
iv
-------
LIST OF TABLES
Tables
2-1 Rohm and Haas Wastestream Characterization Results 2-2
2-2 Bucket and Stopwatch Measurements for Hexane/Acenaphthene Solution 2-3
2-3 Bucket and Stopwatch Method Measurements for Chlorobenzene/
1,1,1-Trichloroethane Solution 2-4
2-4 Bucket and Stopwatch Method Results for Metals Solution 2-5
2-5 Weight Change (AW) Versus Time for Feedstock Solutions 2-6
2-6 Calibrated Feed Jar Method for Hexane/Acenaphthene Solution .... 2-8
2-7 Calibrated Feed Jar Method For Chlorobenzene/l,l,l-Trichloroethane
Solution 2-9
2-8 Feed Rate Comparison Between 4 Methods for Hexane/
Acenaphthene Solution 2-10
2-9 Feed Rate Comparison for Chlorobenzene/l,l,l-Trichloroethane
Solution 2-11
2-10 Feed Rate Comparison for Two Methods Metals Solution 2-12
2-11 Total Mass In 48-Hours 2-13
2-12 Wastewater Flow Rate at Location A 2-15
2-13 Wastewater Flow Measurements at Location D 2-16
2-14 Calculated Wastewater Flows at Location D 2-18
2-15 Temperature Measurements 2-19
2-16 Tracer Concentrations In Wastewater Samples 2-20
2-17 Canister Field Sample Results 2-23
2-18 Mass Flow Rates for Tracer Compounds at Liquid
Sampling Locations 2-25
2-19 Air Emissions by Compound and Location 2-27
3-1 Compounds Present in the Wastewater, Rohm and Haas Bristol Plant . 3-2
5-1 Sampling Frequency and Physical Measurements at Each Location ... 5-2
5-2 Tracer Chemical Reagents 5-3
5-3 Analytical Procedures .5-12
6-1 Trip and Field Blank Results - Liquid Samples (/ig/L) 6-4
6-2 Method Blank Results - Liquid Samples (pg/L) 6-5
-------
LIST OF TABLES (Continued)
Tables
6-3 MS/MSD Results - Liquid Samples 6-7
6-4 Field Duplicate Results - Liquid Samples 6-9
6-5 "Add" Spike Sample Results 6-11
6-6 Method and Trip Blank Results - Canister Samples (ppbv) 6-14
6-7 Field Duplicate Results - Canister Samples 6-15
vi
-------
LIST OF FIGURES
Figure
3-1 Layout of the Chemical Sewer System Rohm and Haas Bristol Plant . . 3-3
4-1 Wastewater Sampling Locations 4-2
4-2 Sampling Configuration at Location A 4-3
4-3 Sampling Configuration at Location B 4-5
4-4 Sampling Configuration at Location C 4-6
4-5 Sampling Configuration at Location D 4-7
4-6 Configuration of the Gravity Flow Pumping Station Tank System . . . 4-9
4-7 Headspace Gas Sampling Locations 4-10
4-8 Configuration for Leak/Vent Sample Locations 4-11
vii
-------
1.0 INTRODUCTION
The United States Environmental Protection Agency (EPA) is gathering
information to improve the understanding of volatile organic compound air
emissions from industrial wastewater collection systems. The EPA Office of
Air Quality Planning and Standards, through its Chemicals and Petroleum Branch
(CPB) and Emission Measurement Branch (EMB), contracted Radian Corporation to
develop and implement a field sampling and analytical program to provide data
on volatile emissions from industrial wastewater collection systems. The
information from this study will be used to develop guidelines and regulations
for air emissions from wastewater for several industries under the Clean Air
Act.
1.1 Project Description
This work assignment involved support of the technical functions
carried out by the U.S. Environmental Protection Agency, Emission Measurement
Branch, Research Triangle Park, North Carolina. To improve the current
understanding of organic emissions from wastewater collection systems, the EPA
coordinated a field sampling and analytical program on the process sewer
system at the Rohm and Haas chemical manufacturing facility located in
Bristol, Pennsylvania. Radian was contracted to assist in the planning,
sampling, sample analysis and data evaluation for this program.
A pre-test survey was carried out by Radian and Research Triangle
Institute (RTI) engineers together with EPA staff to examine the facility.
The pre-test survey was performed on February 28, 1991, at which time
objectives were discussed and sample points were tentatively identified.
Samples from three locations along the wastewater collection system
were collected by plant personnel on April 4, 1991, as part of a wastewater
characterization/spiking study. The samples were analyzed at Radian
Corporation's Perimeter Park laboratory located in Morrisville, NC. The
results from this study were used to select appropriate tracer compounds for
the full-scale test at this facility. The sampling program was described in
the "Test Plan for Field Evaluation of Wastewater Drain System Emissions,"
1-1
-------
dated June 24, 1991. Preparation for sampling began in early June 1991 and
the full-scale field test was conducted from June 24 - 28, 1991, at the
facility.
1.2 Project Objectives
The purpose of the testing program was to provide data to corroborate
or improve existing emission models for air emissions of volatile materials in
wastewater collection systems. The specific objectives were:
• To conduct a tracer study at an industrial facility to acquire
emission data;
• To collect high quality data for wastewater concentrations,
volumetric flowrate, and temperature at several locations along
the collection system;
• To collect high quality data for equilibrium headspace
concentration and temperature at several locations along the
collection system; and
• To collect high quality data for gas concentration,
flowrate, and temperature at several system leaks
and/or vents along the collection system.
1.3 Project Organization and Responsibilities
This work assignment (No. 1-33) was performed under EPA contract No.
68D90054. The sampling and analysis portion of the project was performed by
Radian Corporation staff located in Rochester, New York, and Morrisville,
North Carolina, respectively. Site selection and sample location selection
was performed by EPA-EMB, EPA-CPB, Radian Corporation and RTI. Some on-site
physical measurements of flow direction at specific sewer system manholes were
performed by RTI. Valuable assistance throughout the whole testing program
was provided by the staff at the Rohm and Haas Bristol Facility.
1-2
-------
A list of the participants for the on-site portions of the program is
presented in Appendix A.
1.4 Test Plan Changes
The test plan was followed except for the following changes:
• A 3-hour inoculant equilibration period was allowed prior to the
start of sample collection;
• Sample location A was changed from the first manhole downstream
from the spiking location along the collection system, to the
inlet at the end of the open channel that was spiked;
• A weir was installed in the open channel upstream of the spike
location to measure the flow at location A, and the ISCO Flow
Pokes® were not used;
• The GFPS pump calibration was only performed once instead of three
repeat trials;
• Additional determinations of the tracer chemical feed rates were
performed;
• A Kurz Instruments Inc. Model 490 mini-anemometer was used to
measure air velocities at leak/vent locations instead of a TSI
Model 8350 velometer.
• Sample collection for each round began at location A and proceeded
to location D;
• Due to problems with the tracer feed system for acenaphthene, only
selected samples were analyzed by SW-846 Method 8270 'for semi-
volatile compounds; and
• Canister samples collected at locations A, B, C and D were
collected as grab samples instead of manual composites.
1-3
-------
All of the originally planned samples were collected. The test was
conducted for the complete 48-hour duration.
1.5 Report Organization
Results of the test are summarized and discussed in Section 2.0 of
this report. A detailed facility description is presented in Section 3.0, and
sampling locations are discussed in Section 4.0. Section 5.0 and 6.0 detail
sampling and analytical procedures, and quality assurance procedures and
results, respectively.
The following are presented in the Appendix volume of this report:
project participants, calculation worksheets, analytical results (laboratory
reports), sample collection matrix, analytical matrix, sample chain-of-custody
forms, and copies of the field logbooks.
1-4
-------
2.0 SUMMARY AND DISCUSSION OF RESULTS
The tracer study of drain system emissions at the Rohm and Haas
Bristol facility was performed to collect data to improve the current
understanding of air emissions from wastewater collection systems at an
operating industrial site. The field test sampling and analytical program met
all of the objectives of the program as stated in Section 1.2. In this
section the results are discussed and tabulated.
2.1 Wastewater Characterization
Samples from locations B, C, and D were collected by plant personnel
on April 4, 1991. These three samples were analyzed at Radian Corporation's
Perimeter Park laboratory for volatiles and semi-volatiles by EPA Methods 8240
and 8270, respectively. A sample from location D was also analyzed by EPA
Method 6010 for metals. These analyses were performed to characterize the
wastewater matrix at the facility along the drain line to be tested. Results
for the characterization samples are presented in Table 2-1. These results
were used .to select the appropriate tracer compounds for use in the full scale
test.
2.2 Tracer Chemical Feed Rates
The flowrates of the spiking compounds as well as a comparison of the
methods used to determine the flowrates are presented in this section.
Calculation worksheets are presented in Appendix B.
Volumetric, and mass flowrates for the hexane/acenaphthene,
chlorobenzene/1,1,1-trichloroethane, and the metals spiking solutions using
the "bucket and stopwatch" method are presented in Tables 2-2, 2-3 and 2-4,
respectively.
Each solution feed reservoir was placed on an ordinary bathroom scale.
Scale readings were taken at approximately 2-hour intervals. The weight
change during each time interval and the associated mass flow rate for each
solution is presented in Table 2-5.
2-1
-------
Table 2-1. Rohm and Haas Wastestream Characterization Results
Location B
Location C
Location D
Volatiles
(Method 8240)
Semi-Volatiles
(Method 8270)
I
N)
Metals
(Method 6010)
acetone (7540 ppb)
toluene (9.6 ppb)
total xylenes (5.4 ppb)
n-butyl ether (7.6 ppb-est)
ethyldimethylbenzene (5.0 ppb-est)
isopropyIbenzene (2.3 ppb)
phenol (328 ppb)
benzoic acid (259 ppb)
butyl benzyl phthalate (280 ppb)
isopropy I benzene (110 ppb)
C12H24 (68 ppb)
ethylhexyl acetate (30 ppb)
methylheptyl propenoate (63 ppb)
octyl propanoate (27 ppb)
phenyl ethanediol (330 ppb)
phenyl benzoate (250 ppb)
unknown oxygenated compound (91 ppb)
toluene ester of benzoic acid (970
Not analyzed
acetone (27,570 ppb)
benzene (67.1 ppb)
2-butanone (5100 ppb)
ethylbenzene (33.4 ppb)
styrene (5330 ppb)
toluene (2390 ppb)
total xylenes (204 ppb)
butyl 2-propenoate (940 ppb-est)
isobutyl propenoate (63.8 ppb-est)
methyl methacrylate (1960 ppb-est)
benzoic acid (121 ppb)
naphthalene (38 ppb)
oxygenated compound (2300 ppb)
tetramethyIbenzene (94 ppb)
Not analyzed
acetone (10,600 ppb)
benzene (45.5 ppb)
2-butanone (3800 ppb)
ethylbenzene (23.8 ppb)
styrene (4150 ppb)
toluene (2100 ppb)
total xylenes (143.4 ppb)
butyl 2-propenoate (600 ppb-est)
ethyl toluene (8.0 ppb-est)
isobutylpropenoate (43 ppb-est)
methyl methacrylate (1830 ppb-est)
benzoic acid (138 ppb)
naphthalene (39 ppb)
butoxyyethoxyethanol (630 ppb)
tetramethylbenzene (98 ppb)
aluminum (0.66 ppm)
barium (0.080 ppm)
boron (0.14 ppm)
calcium (5.3 ppm)
copper (0.028 ppm)
iron (3.0 ppm)
magnesium (1.6 ppm)
manganese (0.79 ppm)
nickel (0.011 ppm)
potassium (0.55 ppm)
silver (0.013 ppm)
sodium (12.0 ppm)
strontium (0.21 ppm)
vanadium (0.0040 ppm)
zinc (0.18 ppm)
est - estimated quantity
-------
Table 2-2. Bucket and Stopwatch Measurements for Hexane/Acenaphthene Solution
(Density calculated as 1 pound per liter or 0.454 g/ml)
Mass Flow Rate (ms/min)
Date
6/25
6/25
6/25
6/25
6/26
6/26
6/27
Time
1330
1530
1935
2300
0000
1245
1120
Flowrate (ml/min)
1.0
2.0
2.0
1.9
0*
2.0
2.0
Hexane
392
783
783
744
-
783
783
Acenaphthene
62
125
125
119
-
125
125
* Flow stopped due to crystallization in tube, restarted at 1245.
Note:
1. Sample collection began at 1608 on 6/25.
2-3
-------
Table 2-3. Bucket and Stopwatch Method Measurements for
Chlorobenzene/1,1,1-Trichloroethane Solution
Date
6/25
6/25
6/25
6/25
6/26
6/26
6/27
(Density
-
Time
1340
1530
1935
2300
0322
1136
1101
calculated as 2.9
Flowrate (ml/min)
1.0
2.0*
2.0
1.8
1.8
1.9
1.8
pounds per liter or 1.32g/ml)
Mass
Chlorobenzene
594
1188
1188
1069
1069
1128
1069
Flowrate (mg/min)
1.1. 1-Trichloroethane
726
1452
1452
1307
1307
1379
1307
* Flow reset from 1.4 at 1510.
Note:
1. Sample collection began at 1608 on 6/25.
2-4
-------
Table 2-4. Bucket and Stopwatch Method Results for Metals Solution
(Density calculated as 2 pounds per
liter or 0.908
g/ml)
Mass Flowrate (mg/min)
Date
6/25
6/25
6/25
6/25
6/26
6/26
6/27
Time
1308
1510
1935
2300
0322
1136
1101
Flowrate (ml/min)
7.7
6.2
5.7
5.4
5.0
5.5
5.4
Zinc
601
484
445
422
390
429
422
Copper
624
502
462
437
405
445
437
Note:
1. Sample collection began at 1608 on 6/25.
2-5
-------
Table 2-5. Weight Change (AW) Versus Time for Feedstock Solutions
NJ
Date
6/25
6/25
6/25
6/25
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/27
6/27
6/27
6/27
6/27
6/27
6/27
6/27
Time
Internal
1608-1807
1807-2004
2004-2221
2721-2359
2359-0202
0202-0358
0358-0555
0555-0758
0758-1013
1013-1203
1203-1401
1401-1600
1600-1800
1800-2000
2000-2200
2200-0000
0000-0200
0200-0355
0355-0600
0600-0807
0807-1000
1000-1200
1200-1407
1407-1555
Time (min)
119
117
136
98
123
116
117
123
135
110
118
119
120
120
120
120
120
115
125
127
113
120
127
108
Hexane/Acenaphthene Solution
Mass Flow Rate (mg/min)
AW (Ibs) Hexane Acenaph.
0.5 1,644 263
0.5 1,673 267
0.5 1,439 230
0.0 0 0
0.0* 0 0
0.0* 0 0
0.0* 0 0
NU**
NU
NW
NU
NU
NU
NU
NU
HU
NU
NU
NU
NU
NU
NU - -
NU
NU
Ch 1 orobenzene/ 1 , 1 , 1 - T r i ch I orethane
Solution
Mass Flow Rate (mg/min)
AU (Ibs)
0.5
0.5
0.0
0.5
0.5
0.0
1.0
0.5
1.5
0.5
0.5
0.5
0.5
0.5
0.5
0.0
1.0
0.5
1.0
1.5
1.0
2.0
0.5
0.5
Chloro-
benzene
858
873
0
1,042
830
0
1,746
830
2,270
929
866
859
851
851
851
0
1,702
888
1,634
2,413
1,808
3,401
804
946
1,1,1-
Trichlor-
oethane
1,049
1,067
0
1,274
1,015
0
2.134
1,015
2,774
1,135
1,058
1,049
1,041
1,041
1,041
0
2,081
1,086
1,998
2,949
2,210
4,157
983
1,156
Metals Solution
Mass Flow Rate (mg/min)
AU (Ibs)
2
2
2
2
2
1
2
0
2
3
2
1
2
2
2
2
1
1
2
2
2
2
2
1
Zinc
656
667
574
662
799
337
667
0
868
1,064
662
328
651
651
651
651
325
340
625
615
347
651
615
362
Copper
679
691
594
695
825
348
691
0
898
1,102
685
340
673
673
673
673
337
351
646
636
359
673
636
374
Flow stopped due to crystalization in delivery tube. ** Not weighed, delivery unit on hot plate, not able to weigh.
-------
The calibrated feed jar procedure was also used to measure changes in
feed volumes over 2-hour intervals. Tables 2-6 and 2-7 present the changes in
volume measured by the feed jar calibration technique plus the corresponding
calculated mass flow rate and changes in mass for the hexane/acenaphthene and
chlorobenzene/1,1,1-trichloroethane solutions, respectively.
The total mass of each compound used was determined by measuring the
volume of each solution before and after the test, and multiplying by the
solution's density to determine the mass. An average mass flow rate was then
calculated by dividing the total mass by the total time of the test. These
results are presented in Table 2-8.
Table 2-9 presents a comparison of four feed rate measurement methods
for the hexane/acenaphthene solution: bucket and stopwatch, weight change,
feed jar calibration, and total volume. The weight change was not effective
because the feed reservoir was removed from the scale during the test and
placed on a heating surface to keep the acenaphthene in solution. The overall
agreement of the other techniques was considered acceptable for this study.
Table 2-10 presents a comparison of the feed rate measurement
techniques for the chlorobenzene/1,1,1-trichloroethane solution. Overall, the
measurement techniques show acceptable agreement for this study.
Table 2-11 compares the feed rate measurement techniques for the
metals solution. Agreement between these techniques was also considered
acceptable for this study.
Overall, the bucket and stopwatch flow rate measurement technique was
the most sensitive for measuring small changes in volume, and was considered
the most accurate.
2-7
-------
Table 2-6. Calibrated Feed Jar Method for Hexane/Acenaphthene Solution
Date
6/25
6/25
6/25
6/25
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/27
6/27
6/27
6/27
6/27
6/27
6/27
6/27
Time Interval
1608-1807
1807-2004
2004-2221
2221-2359
2359-0202
0202-0358
0358-0555
0555-0758
0758-1013
1013-1203
1203-1401
1401-1600
1600-1800
1800-2000
2000-2200
2200-0000
0000-0200
0200-0355
0355-0600
0600-0807
0807-1000
1000-1200
1200-1407
1407-1555
Time
(min)
119
117
136
98
123
116
117
123
135
110
118
119
120
120
120
120
120
115
125
127
113
120
127
108
A Volume
(ml)
250
280
200
100
110
120
0
0
0
0
120
250
250
190
170
150
230
0
0
120
290
120
330
210
A Mass
(g)*
113.5
127
90.8
45.4
49.9
54,5
0
0
0
0
54.5
113.5
113.5
86.3
77.2
68.1
104.4
0
0
54.5
131.7
54.5
149.8
95.3
MFR":
Hexane
823
935
576
399
346
405
0
0
0
0
398
823
816
620
554
490
750
0
0
395
1005
392
1018
760
MFRb:
Acena
phthene
131
150
92
64
56
65
0
0
0
0
64
131
130
99
89
78
120
0
0
59
160
62
162
122
*: Mass flow rate for hexane expressed, in rag/min.
°: Mass flow rate for acenaphthene, expressed in mg/min.
•Density of solution calculated as 0.454 g/ml.
2-8
-------
Table 2-7. Calibrated Feed Jar Method For Chlorobenzene/1,1,1-Trichloroethane
Solution
Date
6/25
6/25
6/25
6/25
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/27
6/27
6/27
6/27
6/27
6/27
6/27
6/27
Time Interval
1608-1807
1807-2004
2004-2221
2221-2359
2359-0202
0202-0358
0358-0555
0555-0758
0758-1013
1013-1203
1203-1401
1401-1600
1600-1800
1800-2000
2000-2200
2200-0000
0000-0200
0200-0355
0355-0600
0600-0807
0807-1000
1000-1200
1200-1407
1407-1555
Time
(min)
119
117
136
98
123
116
117
123
135
110
118
119
120
120
120
120
120
115
125
127
113
120
127
108
AVolume (ml)
230
280
180
210
260
130
260
200
270
350
300
260
170
200
260
200
230
250
310
440
430
340
310
270
AMass*
(g)
304
370
238
277
343
172
343
264
356
462
396
343
224
264
343
264
304
330
409
581
568
449
409
356
MFRa:
Chloro-
benzene
1150
1423
787
1272
1255
667
1319
966
1187
1890
1620
1297
840
990
1286
990
1140
1291
1472
2058
2262
1684
1449
1483
MFRb:
TCA
1405
1739
963
1555
1533
816
1613
1180
1450
2310
1980
1585
1027
1210
1572
1210
1393
1579
1800
2516
2765
2058
1771
1813
' Mass flow rate for chlorobenzene, expressed in mg/min.
° Mass flow rate for 1,1,1-trichloroethane, expressed In mg/min.
* Density of solution calculated as 1.32 g/ml
2-9
-------
Table 2-8. Total Mass in 48 Hours
Compound Total Feed Mass/48 Hours Average Mass Flow Rate"
(mg/min)
Chlorobenzeneb 3778 1315
l,l,l-Trichloroethaneb 4617 1603
Hexanec 1272 442
Acenaphthene0 203 70
Zincd 1202 417
Copperd _ 1247 _ 433
* Assumes compound feed rate was constant and continuous during the entire 48-hour period.
° Based on a total feed volume of 6.36 liters of the chlorobenzene/1, 1, 1-trichloroethane solution at a
density of 1.32 g/mL.
c Based on a total feed volume of 3.25 liters of the hexane/acenaphthene solution at a density of 0.454
g/mL.
" Based on a total feed volume of 15.4 liters of the zinc sulf ate/ copper chloride solution with metals
concentrations of 78.07 mg of zinc/mL and 80.99 mg of copper /mL, and a solution density of 0.908 g/mL.
2-10
-------
Table 2-9. Feed Rate Comparison for Hexane/Acenaphthene Solution
HEXANE
K)
I
Time Interval
(hr)
0-8
8-16
16-24
24-32
32-40
40-48
Avg Flow Rate
Bucket & Stopwatch
(mg/min)
778
0
783
783
86
Weight Change
(mg/min)
1189
0
NW
NW
NW
NW
-
Jar Calibration
(mg/min)
683
188
305
620
286
794
479
Total Feed Volume
(mg/min) .
442
442
442
442
442
442
442
ACENAPHTHENE
Time Interval
(hr)
0-8
8-16
16-24
24-32
32-40
40-48
Avg Flow Rate
Bucket & Stopwatch
(mg/min)
124
0
125
125
94
Weight Change
(mg/min)
190
0
NW
NW
NW
NW
-
Jar Calibration
(mg/min)
109
30
49
99
45
126
76
Total Feed Volume
(mg/min)
70
70
70
70
70
70
70
NW - Not weighed; delivery unit on hot plate, not able to weigh
-------
Table 2-10. Feed Rate Comparison for Chlorobenzene/1,1,1-Trichloroethane Solution
CHLOROBENZENE
N)
K)
Time Interval
(hr)
0-8
8-16
16-24
24-32
32-40
40-48
Avg. Flow Rate
Time Interval
(hr)
0-8
8-16
16-24
24-32
32-40
40-48
Avg. Flow (g/hr)
Bucket & Stopwatch
(mg/min)
1173
1069
1128
1069
1110
1,1,1-
Bucket & Stopwatch
(mg/min)
1434
1307
1379
1307
1357
Weight Change
(mg/min)
693
852
1231
638
1659
1740
1136
TRICHLOROETHANE
Weight Change
(mg/min)
848
1041
1504
781
2028
2126
1388
Jar
Calibration
(mg/min)
1158
1052
1498
1026
1490
1720
1850
Jar
Calibration
(mg/min)
1415
1285
1831
1255
1822
2102
1618
Total Mass in
48 hours
(mg/min)
1315
1315
1315
1315
1315
1315
1315
Total Mass in
48 hours
(mg/min)
1603
1603
1603
1603
1603
1603
1603
-------
Table 2-11. Feed Rate Comparison for Metals Solution
Zinc
Time Interval
0-8
8-16
16-24
24-32
32-40
40-48
Avg. Flow (g/hr)
Bucket & Stopwatch
(mg/min)
459
390
429
422
425
Weight Change
(mg/min)
640
450
731
651
476
494
574
Total
Volume
417
417
417
417
417
417
417
Feed
(mg/min)
Copper
Time Interval
Bucket & Stopwatch
(mg/min)
Weight Change Total Feed
(mg/min) Volume (mg/min)
0-8
8-16
16-24
24-32
32-40
40-48
Avg. Flow (g/hr)
476
405
445
437
441
662
466
756
673
492
510
593
433
433
433
433
433
433
433
Total Flow in 48 hours (g)
14688
19150
2-13
-------
2.3 Wastewater Flow Measurements
Flow determinations were made at location A using a suppressed
rectangular weir. Measurements of the head at the weir were taken
approximately every 2 hours over a 48-hour period. The head measurements were
taken 3 inches behind the crest of the weir to the nearest 1/8-inch.
Results of the weir measurements and calculations are presented in
Table 2-12. The flows were calculated using the equation presented in Section
5.3 (Equation 5-1). Calculation worksheets are presented in Appendix B. The
calculated flows ranged from 23.4 gallons per minute (gpm) to 58.0 gpm. The
average flow during the 48-hour period was 37.6 gpm.
The flow rate of water at location A varied depending on the discharge
of "white water" from centrifuges in Building 34. Building 34 produced
suspension polymers, hard spherical polymethyImethacrylate-based particles
less than 1 mm in diameter. The batches were processed in a centrifuge to
remove the smaller diameter particles. The centrifuge wash water, which had a
milky white color, was batch discharged approximately once per hour. This
raised the base flow in the trench and turned the flow white. The elevated
flow generally lasted approximately 5 minutes, and cleared from the trench
within 5 minutes after the inflow cleared.
This "white water" discharge was of concern during the testing due to
the possibility of tracer adsorption by the polymer particles. Since the
water cleared rapidly from the collection system, all wastewater sampling was
conducted while the wastewater was clear. This was accomplished by
coordinating the sampling effort with Rohm and Haas personnel.
The "white water" flow rate was measured at the weir during three
events throughout the test period. The flows are presented in Appendix B.
The average flow was 74.6 gpm.
The flow at location D was determined by measuring the pump clock
readings and the level of water in the GFPS tanks at 2-hour intervals. These
measurements are presented in Table 2-13. Flowrates for each 8-hour sampling
interval were calculated using the equations presented in Section
2-14
-------
Table 2-12. Wastewater Flow Rate at Location A
Date Time
6-25 0 (1608)
2 (1807)
4 (2004)
6 (2201)
8 (2359)
6-26 10 (0202)
12 (0358)
14 (0555)
16 (0758)
18 (1013)
20 (1203)
22 (1401)
24 (1600)
26
28 (2000)
30 (2200)
6-27 32 (0000)
34 (0200)
36 (0355)
38
40 (0807)
42 (1000)
44
46 (1407)
48 (1558)
Head On Weir
(inches)
3/4
3/4
7/8
1-1/4
1-1/8
7/8
3/4
7/8
3/4
7/8
1-1/8
1-1/8
1-1/4
3/4
1-1/4
1-3/8
1-1/8
1-0
1-1/8
1-1/4
1-1/8
1-1/8
1-1/8
1-0
1-0
Flow Average Flow For 8-
(gpm) Hour Period (gpm)
23.4
23.4
29.4
50.3
42.9 33.9
29.4
23.4
29.4
23.4 29.7
29.4
42.9
42.9
50.3 37.8
23.4
50.3
58.0
42.9 45.0
36.0
42.9
50.3
42.9 43.0
42.9
42.9
36.0
36.0 40.1
2-15
-------
Table 2-13. Wastewater Flow Measurements at Location D
Date
6/25
6/25
6/25
6/25
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/26
6/27
6/27
6/27
6/27
6/27
6/27
6/27
6/27
6/27
Sample Interval
0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
34
36
38
40
42
44
46
48
Time of Reading
1600
1859
2037
2246
0051
0229
0420
0621
0838
1052
1235
1430
1655
1813
2019
2222
0027
0232
0415
0627
0854
1036
1243
1441
1637
Pump Clock Reading
(hours)
1039.6
1040.4
1041.0
1042.1
1043.4
1044.4
1045 . 6
1046.7
1047.7
1048.7
1049 . 6
1050.4
1051.4
1051.9
1052.9
1054.0
1055.0
1055.8
1056.6
1057.6
1058.9
1059.6
1060.6
1061.4
1062.2
Measured Water
Level In Tank
(inches)
45
51
62
30
70
70
30
70
70
70
30
30
70
70
30
70
30
70
30
30
70
30
70
70
70
2-16
-------
5.3. (Equations 5-2 and 5-3). The results of those calculations are
presented in Table 2-14. Calculation worksheets are presented in Appendix B.
The calculated flows ranged from 358 to 463 gpm. The average flow during the
48-hour period was 390 gpm.
The "white water" flow also affected the flow rate at location D. The
incremental flow at each "white water" event was 37 gpm above the base flow in
the collection system. Since each event lasted approximately 5 minutes and
occurred approximately once per hour, the total incremental flow at location D
in an 8-hour period was 1480 gallons. The Vp values in Table 2-14 were not
corrected for this volume since it would not change the calculated flow rates
significantly.
2.4 Temperature Measurements
Ambient temperature readings were taken at breathing zone height at
location B. Headspace and liquid temperature measurements were made at each
of the four sample locations. These measurements are presented in Table 2-15.
The average ambient temperature over the 48-hour sampling period was
24.2°C. The temperature remained fairly constant, but there were some
.noticeable downward trends of approximately 5-6°C during the night. Water
temperatures at each of the four sampling points remained stable over the
sampling period. However, there were increases of 2.3°C, 2.1°C, and 0.1°C in
water temperature from point A through D, for a total increase in average
temperature of 4.5°C. Average headspace temperatures at locations A and D were
within 1.2°C of the average ambient temperatures. At both of these locations
the headspace temperature was taken in an area open to the atmosphere. The
average of headspace temperatures taken at sites B and C were within 2.5°C of
the average water temperatures at the same site. At these locations the
headspace area above the water stream was covered by a manhole cover.
2.5 Tracer Concentrations in Wastewater Samples
Table 2-16 presents the results for the preliminary samples collected
before tracer inoculation and all of the field samples collected throughout
the test. Laboratory analytical reports are presented in Appendix C.
2-17
-------
Table 2-14. Calculated Wastewater Flows at Location D
Sample
Interval
0
8
16
24
32
40
Notes:
1. Rp - 838
2. r - 60
3. H, - 30
T8
(minutes)
228
486
708
924
1158
1356
gpra
inches
inches
TO
(minutes)
0
228
486
708
924
1158
H8
(inches)
70
70
70
30
70
70
H0
(inches)
45
70
70
70
30
70
VP
(gallons)
192288
216204
186036
179064
198036
165924
AT
(minutes)
531
467
497
452
507
463
Flowrate
(gpm)
362
463
374
396
391
358
The total flow into the tank over each 8-hour period can be calculated using the following equation:
Vp - [(T. - T0) * R,,] + [C • ft * r2 * (H, - H,)] -[C*»r*r1*(H0- H,)]
Where:
VD = Total volume into the tank (gallons)
T, • Pump clock reading at end of 8-hour interval (minutes)
T0 - Pump clock reading at beginning of 8-hour interval (minutes)
Rp • Pump rate (gpm)
C » Conversion factor for cubic inches to gallons (0.00433)
r - Gravity Flow Pumping Station tank radius (inches)
R, m Liquid height in tank at end of 8-hour interval (inches)
H, - Liquid height in tank at end of pumping cycle (inches)
H0 • Liquid height in tank at beginning of 8-hour interval (inches)
The flow rate in gpm over the 8-hour period can then be determined using the following equation:
F =
whsre:
AT » Actual elapsed time between the T, and T0 pump clock and liquid height readings (minutes).
2-18
-------
Table 2-15. Temperature Measurements
Location A
Time
(hrs.)
Pre
0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
34
36
38
40
42
44
46
48
Ambient
Temp.
(°C)
24.8
27.9
24.8
23.0
20.6
19.6
19.3
18.4
18.2
20.5
25.4
27.3
28.8
31.6
24.8
27.5
23.8
20.8
21.0
21.1
21.4
24.7
26.0
28.4
30.5
32.1
Head-
Space
Temp.
(°C)
22.8
30.1
24.8
26.3
21.6
21.6
19.7
18.4
18.1
22.2
26.8
32.5
32.3
29.2
29.3
28.1
23.5
21.6
21.0
20.6
21.1
23.9
27.7
34.1
33.5
30.3
Liquid
Temp.
(°C)
26.0
27.0
25.3
26.7
25.6
25.1
22.5
24.6
24.3
24.8
25.8
26.2
26.6
29.3
26.8
27.9
26.8
26.0
25.4
26.3
26.6
25.9
26.6
26.8
27.1
26.6
Location B
Head-
Space
Temp.
CO
30.8
33.9
28.0
26.4
27.2
27.2
24.5
22.9
22.6
25.5
32.0
32.9
34.7
32.7
31.0
29.8
27.2
24.6
22.9
25.5
27.0
26.7
32.0
34.1
33.5
37.0
Liquid
Temp.
(°C)
28.4
29.5
26.7
28.9
27.8
27.4
26.1
26.9
26.7
28.5
27.8
28.8
28.1
28.2
29.2
30.8
28.9
28.6
28.4
30.5
28.5
28.6
27.8
28.3
29.9
29.5
Location C
Head-
Space
Temp.
<°C)
27.0
31.5
27.8
28.4
23.2
23.4
23.8
23.0
24.9
26.8
27.9
31.1
27.7
30.0
28.9
28.4
28.4
28.0
28.0
28.6
26.8
28.1
28.2
29.4
30.3
31.1
Liquid
Temp.
CO
28.5
30.5
31.2
30.8
30.5
28.8
28.7
28.1
28.3
30.6
30.6
31.2
31.0
31.1
31.1
29.7
29.9
31.7
31.5
31.5
30.9
29.8
31.6
31.8
32.1
32.6
Location D
Head-
Space Liquid
Temp . Temp .
(«C) (°O
28.3
27.6
24.2
21.9
20.1
24.0
18.9
17.8
17.9
22.1
32.3
39.2
30.4
27.4
26.6
23.9
22.3
22.6
23.2
25.8
22.2
24.9
29.8
36.8
32.1
28.1
30.0
31.3
31.3
31.1
30.2
29.4
28.9
28.2
28.6
30.9
31.2
31.7
31.6
31.3
31.2
26.9
29.9
31.6
31.5
31.4
30.6
30.5
31.6
32.1
32.3
32.8
2-19
-------
Table 2-16. Tracer Concentrations in Wastewater Samples (pg/L)
Sample
A-Pre
A-0
A-8
A-16
A-24
A-32
A-40
B-Pre
B-0
B-8
B-16
B-24
B-32
B-40
C-Pre
C-0
C-8
C-16
C-24
C-32
C-40
D-Pre
D-0
D-8
D-16
D-24
D-32
D-40
Ch 1 o r ob enz ene
<5
23000
4100
34000
51000
41000
120000
<5
260
860
760
1200
1100
1000
<5
11
35
90
87
110
150
<5
<5
<5
<5
9.2
15
170
1,1,1-
Trichloroethane
<5
58000
4600
16000
45000
99000
120000
<5
660
2600
1500
1600
1900
2600
<5
190
99
310
140
180
340
<5
23
21
120
130
190
280
Hexane
<5
<1000
<1000
<1200
<2000
<500
<5
<5
<500
<500
<500
<2500
<500
<500
<5
<50
<50
<250
<50
<50
<50
<5
<5
<5
<5
<5
<5
<50
Acenaphthene
<10
<10
Not Analyzed
<10
Not Analyzed
Not Analyzed
Not Analyzed
<10
88
Not Analyzed
Not Analyzed
Not Analyzed
Not Analyzed
99
<10
49
Not Analyzed
Not Analyzed
Not Analyzed
Not Analyzed
18
<10
4.3
Not Analyzed
Not Analyzed
Not Analyzed
Not Analyzed
48
Copper
4.7
2300
2700
2300
1700
1400
1700
6.5
1800
1700
1600
1400
1300
1400
2.6
240
150
240
200
210
230
3.3
220
140
210
180
190
190
Zinc
9.0
2600
3000
2500
1900
1500
2000
20
2000
1900
1700
1500
1400
1500
14
290
180
280
240
240
270
15
280
170
250
220
240
250
2-20
-------
None of the organic tracer compounds were detected in the preliminary
samples, which indicates that the background level of these compounds was
essentially zero. The two metal tracers were detected in the preliminary
samples (copper at an average concentration of 4.3 /ig/L and zinc at an average
concentration of 14.5 /ig/L). These levels were one to two orders of magnitude
lower than the lowest concentration detected in the spiked samples. The
background contribution of these metals can be neglected in evaluation of the
data. For the analysis of tracer losses presented in subsequent sections of
this report, background subtraction was not performed for the metals data.
It is important to note in reviewing the concentration data, that the
flow was increasing between each sample location due to the addition of other
unspiked wastewater entering at various points along the collection system; in
effect diluting the tracers as the wastewater flows downstream. Therefore,
decreases in sample concentration between sample points cannot be directly
attributed to losses to the air. This is particularly true between locations
B and C where a major step change in wastewater flow rate occurred.
Some qualitative observations made during the test will effect the
data analysis and interpretation. No problems were encountered with the
tracer feed system for the chlorobenzene/l,l,l-trichlorethane or the metals
tracers. The feed rates appeared to be reasonably constant throughout the
test period.
For the hexane/acenaphthene tracer solution, the feed rate was
intermittent and stopped completely at certain times during the test. This
was because the solubility of the acenaphthene in the hexane was very
temperature dependent. As the ambient air temperatures dropped throughout the
evening and night the acenaphthene precipitated in the tracer reservoir and in
the peristaltic pump tubing line, effectively blocking the flow. It was also
very difficult to get an accurate "bucket and stopwatch" flow measurement on
the hexane/acenaphthene feed because the acenaphthene would crystallize out
and block the flow when the feed line was lifted out of the wastewater stream.
Since the level of the solution in the feed reservoir did drop over specific
2-hour intervals it is known that some of the solution was being inoculated,
but the feed flow was not constant and very difficult to quantify. In an
effort to overcome the acenaphthene precipitation problem, the feed reservoir
2-21
-------
was placed on a hot plate and the solution was heated. This did achieve some
success in resolublizing the acenaphthene in the reservoir but was not
successful in unblocking the feed tubing. The tubing was changed twice during
the test period and all available tubing on-site was used, but tie line
blockage problems did persist. Due to the uncertainty of the tracer feed only
selected samples were analyzed for acenaphthene.
One important visual observation was made during the time period when
the hexane/acenaphthene was going in. The end of the feed tubing was
submerged below the water surface at the inoculation point but the hexane
appeared to rise immediately to the surface and appeared as a very thin light
refractive layer (droplet sheen) which could be observed to float over the
sample collection line at location A, which was submerged below the water
surface. By the end of the open channel, the circulation patterns in the
water had collected the droplet toward the center of the water surface. Since
the acenaphthene was dissolved in the hexane, this could explain why
acenaphthene was not detected at location A. Since the acenaphthene was
picked up at location B, it is possible that the acenaphthene was mixed better
by the time the water travelled through a longer section of the drain system.
It may not be possible to determine losses of hexane to the air, since the
hexane may have remained on the surface and floated over all of the submerged
sample lines at all four sampling locations. This might explain all of the
hexane non-detect values.
2.6 Headspace Gas Samples
Results for the volatile tracer compounds in the canister field
samples collected at each of the four wastewater sampling locations and the
seven leak/vent locations are presented in Table 2-17.
The average air velocity and the velocity range for flow of air out of
each manhole is also presented in the table. The air flow at location A was
inadvertently not measured during sample collection. There was no flow at
locations B and C during the test since these were covered with solid plywood
manhole covers and the sampling ports remained capped. No flow measurement
was made at location D since the outlet pipe was below grade and was not
accessible without confined space entry procedures.
2-22
-------
Table 2-17. Canister Field Sample Results3 (ppbv)
Sample Location
A
B
C
D
L/V-1
LA- 2
LA-4
LA-6
LA- 7
LA- 8
w LA-9
Average Velocity (fpm)
Not measured
Not measured
Not measured
Not measured
200
75
75
150
100
300
200
Velocity Range (fpm)
Not measured
Not measured
Not measured
Not measured
100-500
50-100
50-150
100-500
50-250
200-500
100-500
Chlorobenzene
3900b
1400
930
3900
17000
5500
410
3600
100
29000
<0.22
Hexane
48
6.4
91
52
8100
1400
11
220
8.5
650
<1
1,1, 1-Trichloroethane
6000"
1400
20000
17000
110000
13000
320
29000
110
86000
48
to
ppbv - parts per billion by volume
fpm - feet per minute
* - All canister samples were collected on 6/27. Samples at locations A, B, C an D were collected between 1215 and 1250. Samples at the leak/vent locations were
collected between 0935 and 1148.
" - The chromatographic peak for this analyte was saturated. A diluted run of this sample proved unusable. The actual concentration of this analyte may be higher
than the result presented here.
-------
2.7 Mass Flow Rates at Wastewater Sampling Locations
This section presents the results of calculations to determine the
mass flow rate of the tracer compounds at each of the four sampling locations.
2.7.1 Determination of Mass Flux at Locations A and D
Since wastewater flow rates and concentrations are available for
locations A and D the mass flux rate at these locations for all of the tracer
compounds can be determined by the following equations:
MFA-(CA)(FA) (2-D
MFD=(CD)(FD) (2-2)
where:
CA D - Tracer concentrations at locations A or D, respectively; and
FA D - Measured flow rate at locations A or D, respectively.
Results of these calculations are presented in Table 2-18.
2.7.2 Determination of Mass Flux at Locations B and C
Since wastewater flow rates could not be determined at points B and C,
the mass flux cannot be determined directly as described above. The
wastewater flow rates can be calculated indirectly by the following equations:
(C) (2-3)
Fc= (MF^)/(CC) (2-4)
where:
2-24
-------
Table 2-18. Mass Flow Rates for Tracer Compounds at Liquid Sampling Locations (mg/min)
Sample
A-0
A-8
A-16
A-24
A-32
A-40
B-0
B-8
B-16
B-24
B-32
B-40
C-0
C-8
C-16
C-24
C-32
C-40
D-0
D-8
D-16
D-24
D-32
D-40
Flow
(gpm)
34
30
38
45
43
40
46
44
54
55
54
53
330
480
340
370
320
310
360
460
370
400
390
360
Chlorobenzene
3000
470
4900
8700
6700
18000
45
140
160
250
230
200
14
63
120
120
140
170
3*
4*
4*
14
22
230
1,1,1-
Trichloroethane Acenaphthene
7500 NC
520
2300 NC
7700
16000
18000
120 15
430
310
330
390
520 20
240 61
180
400
190
220
390 21
32 6
37
170
200
280
380 65
Copper
300
310
330
290
230
260
310
280
330
290
270
280
300
270
310
280
260
270 '
300
250
300
270
280
260
Zinc
340
340
360
320
240
300
350
320
350
310
290
300
360
330
360
330
290
310
380
300
350
330
360
340
* Calculated using 1/2 of the detection limit for non-detect sample results.
NC - Not calculated
2-25
-------
MFA D = Mean mass flux for metal tracer compounds calculated using Equations 2-1 and 2-2 ,
CB c - Metal tracer concentrations at location B or C, respectively.
It is assumed that the MFAD values will be approximately equal for
the metals since no mass is expected to be lost to the air or to piping via
adsorption. That is, the mass flux rate will be constant at all of the
sampling locations.
Once the wastewater flow rates have been calculated, the mass flux for
the organic tracer compounds can be calculated using the following equations:
MFR = (C.) (F.) (2-5)
MFC= (Cc) (Fc) (2-6)
where:
CB c - Organic tracer concentrations at location B or C, respectively.
FB c - Flow rates calculated Equations 2-3 and 2-4 for location B or
C, respectively.
Results for these calculations are presented in Table 2-18.
2.8 Collection System Air Emission Rates
The air emissions by compound and location are presented in Table 2-
19. The emission rates for unmeasured points along the sewer system were
estimated. The estimates were based on the average of the measured flow rates
and the average of the compound concentration from the closest measured
location. Calculation worksheets are presented in Appendix B.
2-26
-------
Table 2-19. Air Emissions by Compound and Location (mg/min)
Manhole Location
A
L/V-1
L/V-2
Main Line points between L/V-2 & L/V-6
(3 locations)
Spur points between L/V-2 & L/V-6 (16
locations)
L/V-4
L/V-6
L/V-7
Spur point off L/V-7 (1 location)
L/V-8
L/V-9
Spur points off L/V-9 (5 locations)
Main Line point between L/V-8 & D
Spur points between L/V-8 & D (5
locations)
D
Total
1,1,1-
Trichloro-
e thane
689
265
25.9
70.9
4.48
.280
28.4
.036
.036
83.9
.062
.310
102
2.19
166
1440
Chloro-
benzene
77.9
34.9
9.00
12.8
4.80
.300
3.00
.028
.028
24.0
-
-
27.6
12.4
32.0
239
Hexane
31.7
12.8
1.80
1.77
.096
.006
.140
.002
.002
.410
-
-
.450
.169
.320
49.7
2-27
-------
Qualitatively, the air emissions are highest at location A and rapidly
decrease downstream of the tracer inoculation location. The air emissions for
1,1,1-trichloroethane at L/V-2 are less than 5% of the emissions at location
A. The emissions from manhole locations on the main line remain approximately
constant from L/V-2 to location D. Air emissions from manholes off the main
line were present, but very low. The emissions from manholes on branch lines
were at least an order of magnitude (10 times) lower than mainline emissions.
Taken overall, the air emissions follow a fairly logical pattern.
However, several anomalies in the data can be seen. The air emissions at L/V-
8 and points downstream are slightly higher than emissions at L/V-6 and L/V-2.
One explanation for this is the air flow pattern in the headspace of the sewer
system.
Air flow variations at all manholes were detected, and at several
locations, flow varied by a factor of five. Manholes located outside are
affected to some degree by the wind. Data taken on the first day of testing
by RTI personnel show a relationship between wind spe£d and air flow in the
sewer system. At low wind speeds, air flowed into the sewer system. At
moderate wind speeds (200 ft/min) flow was zero. At higher windspeeds (400 to
500 ft/min) air flowed out of the sewer system. Wind direction also had an
effect on sewer system air flow. Manholes located inside buildings will be
effected by the building ventilation system. Headspace areas of high air
velocities and stagnation may exist due to the layout of the sewer system.
Over time, air emissions may collect and then dissipate at manhole locations.
Air samples of the sewer headspace were taken only during venting of the
headspace.
Total air emissions from the sewer system do not equal the rate of
volatile tracer compound mass lost from the water. Air sampling of the sewer
headspace only during periods of venting should have given relatively high air
emissions. Total air emissions of 1,1,1-trichloroethane verify this, by
accounting for 120 to 125% of the mass lost from the water. However, air
emissions of chlorobenzene account for only 20 to 25% of the mass lost from
the water. Air emissions at point A were based on a straight line
extrapolation of emissions from L/V-1 and L/V-2. Chlorobenzene emissions may
not linearly decrease with the distance from the inoculation location, and
2-28
-------
could be much higher at location A than calculated. Also, the air volume of
the sewer system may be large enough to absorb and hold large concentrations
of volatile compounds. The water/air system within the sewer may not have
reached equilibrium when the air samples were taken, and air emissions may
continue after tracer inoculation is stopped.
2.9 Evaluation of Collection System Losses
The data collected during the study can be evaluated for each of the
tracer compounds. The observations, data, and calculated results differ for
each tracer compound and require separate analysis.
For hexane, there were problems with the feed system due to
crystallization of acenaphthene in the reservoir and feed lines. These
problems caused considerable uncertaintly regarding the quantitative feed rate
of hexane into the collection system. In addition, hexane was not detected in
any of the water samples.
However, qualitative visual observation, and the air sample results,
give some indications of the fate of hexane in the collection system.
Immediately after inoculation, the hexane/acenaphthene was observed to rise to
the surface, at which time a large fraction of the hexane probably volatilized
quickly. The resulting sheen droplet was observed to move to the center of
the water stream as the flow entered the pipe at location A. Since the hexane
was at the surface at location A, and the sample was collected below the water
surface, hexane would not be detected at point A. It is not known if the
hexane continued on the surface or completely volatilized by location B.
Either scenario would explain the water results at locations B, C, and D since
all of those samples were also collected from below the water surface. The air
sample results indicate that some of the hexane remained past location A since
hexane was detected in the headspace gas throughout the collection system.
Acenaphthene was similarly affected by the problems with feed rate
quantitation. Since it also appeared to remain on the surface at location A,
it was not detected in the water samples. However, the acenaphthene did
appear to mix into the water by location B, and was also detected at locations
C and D. However, the sample results and the calculated mass flow rates
2-29
-------
showed high variability and inconsistent trends. Along with the potential for
low analytical quantitation for acenaphthene (discussed in detail in Section
6), it is difficult to reach any conclusion on the fate of acenaphthene in the
collection system.
Feed rate calculations for chlorobenzene show good agreement between
the four measurement techniques. The "bucket and stopwatch" method is
considered to be the most reliable. The feed rate calculations confirm the
visual observation that the feed of the chlorobenzene/1,1,1-trichloroethane
solution was reasonably constant throughout the test.
The mass flow rate calculations for location A indicated more mass
passed there than was put in with the tracer feed. A likely explanation is
that the poor mixing observed in the channel for hexane was also a factor for
the chlorobenzene. This is possible since the concentration of chlorobenzene
in the channel was expected to be near the solubility in water. Since
chlorobenzene, unlike hexane, is more dense than water, the chlorobenzene
would have tended to sink below the surface. As the flow moved down the
channel, the chlorobenzene would also have centered in the stream, as did the
hexane. This would have created a zone of artificially high sample
concentration below the surface of the water at the pipe inlet, resulting in
observed sample concentrations which were biased high.
Disregarding the data from location A, it is still possible to
calculate the chlorobenzene losses from the collection system between the
spiking location and locations B, C, and D. Unfortunately, the losses in the
initial segment of the channel cannot be quantitatively distinguished from the
overall loss throughout the system.
Based on an average feed rate of 1119 mg/minute for chlorobenzene,
there appears to be approximately 80 to 90% lost by location B and 98 to 99%
lost over the whole system by location D. Calculations of air emission rates
did not give very good closure of the mass balance as discussed in Section
2.8.
In addition, 1,1,1-trichloroethane also exhibited poor agreement
between mass feed rate at the spike location and location A. This compound is
2-30
-------
also more dense than water, and it is likely that the same mechanism for this
sample bias at location A occurred due to insufficient mixing in the channel.
As with chlorobenzene, the values for location A should be disregarded in
evaluation of the data.
Based on an average feed rate of 1367 mg/min for 1,1,1-trichloro-
ethane, there appears to be approximately 60 to 70% lost by location B, and 80
to 97% lost over the whole system by location D. Mass balance closure was
much better for 1,1,1-trichloroethane than for chlorobenzene based on the air
emission rate calculations presented in Section 2.8.
For the metal tracer compounds, the feed rate determinations showed
good agreement between the four methods, which confirms the qualitative
observation that the feed rate was reasonably constant throughout the test.
The mass flow rate of the metal tracers at location A showed good
agreement with the feed rate calculations. This indicates that the mixing
problem observed for the organic tracers did not affect the metal tracers.
This is likely because the metals were already dissolved in a water phase and
were well below their solubility limits at the flow rate in the channel.
There was also good agreement between the mass flow rates at locations
A and D, which confirmed that there were no losses of the metal tracers to
adsorption. This validated the use of these compounds for the purpose of
calculating wastewater flow rates at locations B and C.
There is a consistent trend to the data at all four locations that
also warrants discussion. There appears to be a consistent increase in the
measured concentrations over time throughout the test. That is, more of the
volatile tracers seem to remain in the water and less volatilization occurs.
One possible explanation is that as time passes and the headspace gas
concentration increases over the water, the concentration gradient decreases,
which results in a lower driving force for volatilization. It appears that 48
hours may have been insufficient time for that concentration gradient to reach
steady state, which would have leveled out the sample concentrations over
time.
2-31
-------
3.0 FACILITY DESCRIPTION
The field test program was conducted from June 24 - 28, 1991, at the
Bristol Plant of Rohm and Haas Delaware Valley Inc., located in Bristol
Township, Bucks County, Pennsylvania. The Bristol Plant supplies specialty
chemicals for automotive, pharmaceutical, medical, paint, lighting, packaging,
food, ink, adhesives, and other applications, both in the United States and
for world-wide export. The plant has five main process lines, producing
Rhoplex® Emulsion Polymers, Paraloid® Polymers, Acrybase® Suspension
Polymers, Acryloid® Solution Polymers, and Acryloid® Molding Pellets. In
addition, the plant serves as a monomer distribution center for the Northeast
and Canada, and runs two processes on a toll basis for other companies. There
are over 100 buildings and approximately 1,400 employees on the 800-acre site,
which includes a 120-acre manufacturing and research park. Two 34-acre fresh
water supply lakes are also on site, along with 54 acres used for tertiary
wastewater treatment lagoons, which receive flow from the facility's
biological treatment plant.
Wastewater from open-ended pipes discharges into open conduits, and
then into an underground chemical sewer system. An accumulation tank, known
as the Gravity Flow Pumping Station (GFPS), collects the gravity flow sewer
discharges, from which the water is pumped to the on-site wastewater treatment
system. All testing occurred in the gravity flow section of the collection
system, upstream from the GFPS outlet. Constituents of the process wastewater
stream, as provided by Rohm and Haas, are presented in Table 3-1. Figure 3-1
provides an overview of the chemical sewer system layout in the section of the
facility selected for this test program.
The gravity feed section of the wastewater collection system is
composed of 8- and 12-inch cast iron piping. The majority of the constituents
in the waste stream originate from air pollution abatement devices. Equipment
cleaning during process change-over activities and maintenance activities also
contribute to the wastewater.
3-1
-------
TABLE 3-1. Compounds Present in the Wastewater,
Rohm and Haas Bristol Plant
Material Boiling Point (°C) Vapor Pressure (mm Hg)
Acrylamide
GAA
Acrylonitrile
Butyl Acrylate
Butyl Cellosolve
Butyl Methacrylate
Desmodur W
Ethyl Acrylate
Ethyl Benzene
Fascat 4202
Formaldehyde
GMAA
Isobutyl Methacrylate
Isopropanol
Methanol
Methyl Acrylate
Methylamyl Ketone
Methyl Methacrylate
Toluol
Vinyl Acetate
Xylene
88
141
77.4
148
336
160
100
108
204
98
160
135
82.3
64
80
149
101
110
73
137
0.21 @50°C
4 @20°C
84 @20°C
3.3 @20°C
0.67 @20°C
4.9 @20°C
0.001 @25°C
31 @20°C
10
-------
N
Figure 3-1. Layout of the Chemical Sewer System
Rohm and Haas Bristol Plant
3-3
-------
4.0 SAMPLING LOCATIONS
The spiking location and the locations of the four wastewater sampling
points were determined based on the pre-test survey visit information and the
wastewater screening/spiking study results. The locations were chosen based
on the objectives of the program to track the fate of the tracer compounds
through the drain system. Canister samples of headspace gas were also
collected at each of the wastewater sampling locations. In addition, several
system leak/vent locations were identified by EPA-CPB and RTI personnel during
the test. Canister samples were also collected from these locations.
A description of each sample point location is provided in the
following subsections.
4.1 Wastewater Sampling Locations
Wastewater samples were collected at four sampling locations in the
gravity flow section of the collection system. These locations are shown in
Figure 4-1.
4.1.1 Location A
Sample location A was at the east end of the open channel located
south of Building 34. This sample location is shown in detail in Figure 4-2.
The sample was collected at the inlet to the pipe at approximately half the
depth of the water. This location was approximately 16 feet downstream of the
spiking point.
The flow at location A was measured by a wooden weir placed into the
channel approximately 4.5 feet upstream of the spiking point. To ensure that
all of the flow was measured by the weir, a wooden dam was placed on the
concrete pad to direct flow into the channel upstream of the weir.
Rohm.Drn
EPAr 4-1
-------
Figure 4-1. Wastewater Sampling Locations
4-2
-------
*>.
I
10
DRAIN
11'-0'
CURB
OPEN TRENCH
7" - 9" DEEP
12" WIDE
BUILDING 34
2X4 PLACED TO
CHANNEL FLOW
WEIR
CONCRETE
4'-8''2"
CHEMICAL RESERVOIRS
15'-11"
TO INNOCULATION POINT
CHEMICAL FEED PUMPS
STREET
u
T
L 6.5"
INLET
PIPE - 3*
WATER DEPTH
SITE A - TOP VIEW
SAMPLE INLET AT FACE OF PIPE
SCALE: 1" =5'
Figure 4-2. Sampling Configuration at Location A
-------
4.1.2 Location B
Sample location B was at the manhole located at the intersection of
the facility roads near Buildings 30 and 36. This sample location is shown in
detail in Figure 4-3. A sample had previously been collected at this location
during the screening/spiking study.
At this location, the steel manhole cover was replaced with a plywood
cover for the duration of the test. The plywood cover was used to maintain
the vapor-water equilibrium within the manhole. Holes were drilled in the
cover and sample ports were installed for water sampling, headspace gas
sampling and temperature measurements.
4.1.3 Location C
Sample location C was at the manhole located at the intersection of
the facility roads near Buildings 46, 51 and 67. This sample location is
shown in detail in Figure 4-4. A sample had previously been collected at this
location during the screening/spiking study.
At this location, the steel manhole cover was replaced with a plywood
cover for the duration of the test. The plywood cover was used to maintain
the vapor-water equilibrium within the manhole. Holes were drilled in the
cover and sample ports were installed for water sampling, headspace gas
sampling and temperature measurements.
4.1.4 Location D
Sample location D was at the junction box located upstream of the GFPS
tanks. The sample location is shown in detail in Figure 4-5. A sample had
previously been collected at this location during the screening/spiking study.
During the test, the GFPS system was configured by plant personnel so
that the entire inflow from the junction box would go into the east tank (the
west tank inlet was closed). All of the water pumped out of the system was
also from the east tank (the west tank pump was shut down); however, the
interconnecting valve between the two tanks was open so that the liquid level
4-4
-------
WATER SAMPLING PORT
AIR SAMPLING
PORT
2" DIA. TEMPERATURE
GAUGE PORT
TOP VIEW - COVER
WATER SAMPLE
PROBE (2'-5''2">
AIR SAMPLE PROBE
WATER LEVEL
8" INLET PIPE
12" OUTLET PIPE
SIDE VIEW - MANHOLE
SITE B
SCALE: 1 " = 1'
Figure 4-3. Sampling Configuration at Location B
4-5
-------
WATER SAMPLING PORT
AIR SAMPLING
PORT
2* 01A. TEMPERATURE
GAUGE PORT
TOP VIEW - COVER
WATER SAMPLE
PROBE <5'-9">
12* INLET PIPE
AIR SAMPLE PROBE (3'-7s'4">
J
s
1
PE -.
^
} '
V
(
\
npj
1
^
*
«
i[
i ' •
a/ '
t
*
1
in
i
12* PI
r"
. /r
t
WATER LEVEL
12* OUTLET PIPE
SIDE VIEW - MANHOLE
SITE C
SCALE: 1" • 2'
Figure 4-4. Sampling Configuration at Location C
4-6
-------
SAMPLE
LOCATION
12" INLET
PIPE
2'
5'-8'
16" OUTLET
PIPE
DEPTH TO WATER - 5'
TOTAL WATER DEPTH - 1'-4"
TOP VIEW
SITE D
SCALE: 1" = 2
Figure 4-5. Sampling Configuration at Location D
4-7
-------
would be equalized in the two tanks. This was necessary to maintain adequate
storage capacity in case of rain. A detailed diagram of the GFPS tank system
is presented in Figure 4-6. Therefore, the flow at the GFPS tanks was.
measured to provide the flow estimates for sample location D.
4.2 Headspace Gas Sampling Locations
Headspace gas samples were collected at each of the wastewater sample
locations and at seven leaks/vents along the collection system. These
locations are shown in Figure 4-7. The leak/vent sample locations are shown
in detail in Figure 4-8.
4.2.1 Location A
The grab sample was collected 1 to 2 inches above the surface of the
water, at a point 3 to 4 inches inside the pipe at the east end of the open
channel.
4.2.2 Location B
The grab sample was collected from the air sampling port on the wooden
manhole cover.
4.2.3 Location C
The grab sample was collected from the air sampling port on the wooden
manhole cover.
4.2.4 Location D
The grab sample was collected 1 to 2 inches above the surface of the
water, at a point 3-4 inches inside the inlet pipe at the junction box.
4.2.5 L/V-1
The grab sample was collected 1 to 2 inches below the manhole cover in
the hole around the pipe.
4-8
-------
16" INLET
PIPE
16" INLET
PIPE
P1B PUMP
LEVEL
CONTROLLER
(TYP.)
TOP VIEW
WALKWAY
EQUALIZATION
VALVE
A
v
1
5
o
\
\ 1
S
U3
oo
\
I
\ ¥ ¥
/
(
:>
TT TT
(
)
/ nMiNunMju.
/ /— WALKWAY
/ /
/
TT/
SIDE VIEW
SCALE: 1* - 6
Figure 4-6. Configuration of The Gravity Flow Pumping Station Tank System
4-9
-------
L/V-4
N
L/V-9
Figure 4-7. Headspace Gas Sampling Locations
4-10
-------
6" PIPE
MANHOLE COVER
LIFT HOLE (TYP.)
L/V-1
L/V-2
L/V-4
1" x 3"
OPENING
1" x 3"
OPENING
1" x 3"
OPENING
1 " x 3 "
OPENING
L/V-6
L/V-7
L/V-8
L/V-9
Figure 4-8. Configuration for Leak/Vent Sample Locations (Manhole Top Views)
-------
4.2.6 L/V-2
The grab sample was collected 1 to 2 inches below the manhole cover in
the large hole in the cover.
4.2.7 LA-3
This sample location was a floor sump covered with a grate located
inside Building 34. No sample was collected since air was flowing into the
system at this location at the time of sampling.
4.2.8 L/V-4
The grab sample was collected 1 to 2 inches below the manhole cover in
one of the lift holes.
4.2.9 L/V-5
No specific L/V-5 location was identified.
4.2.10 L/V-6
The grab sample was collected 1 to 2 inches below the manhole cover in
one of the holes in the cover.
4.2.11 L/V-7
The grab sample was collected 1 to 2 inches below the manhole cover in
the hole in the cover.
4.2.12 L/V-8
The grab sample was collected 1 to 2 inches below the manhole cover in
the hole in the cover.
4-12
-------
4.2.13 L/V-9
The grab sample was collected 1 to 2 inches below the manhole cover in
one of the holes in the cover.
4.2.14 L/V-10
This sample location was an outdoor sump covered with a grate located
south of Building 120. No sample was collected at this location.
4-13
-------
5.0 SAMPLING AND ANALYTICAL PROCEDURES
This section describes the sampling procedures used during the tracer
study at the Rohm and Haas plant. Samples were collected for analysis of
volatile, semi-volatile, and metal tracer materials. Table 5-1 provides a
summary of the types and numbers of samples taken at the four air/water
sampling locations, as well as the physical measurements taken.
5.1 Inoculation of the Waste Stream
The wastewater collection system was inoculated with six tracers:
n-hexane, chlorobenzene, 1,1,1-trichloroethane, acenaphthene, zinc sulfate,
and copper (II) chloride. These were selected based on assessment of the low
background levels present in the wastewater, the spike recoveries (good
recovery indicated that the wastewater had little matrix effect on the
compound), the solubility at the inoculation point, the facility's National
Pollutant Discharge Elimination System (NPDES) permit limits, and the relative
toxicity/reactivity of the compound.
The tracers were introduced into the wastewater collection system in
the open channel located outside of Building 34 (Figure 3-1). The inoculation
of the system began three hours prior to sampling, to ensure that equilibrium
had been established.
5.1.1 Reagents
Table 5-2 lists the six chemical reagents that were used for the test.
The table includes the chemical manufacturer, formula weight, purity (in
percent), and the total quantity of each reagent that was available for the
test. The total quantities were estimated based on assumed flows at sample
locations A and D, assumed percent losses through the system, and the required
target concentrations for analytical detection at location D. The formula
weights and purities were provided by the manufacturer or chemical distributor
of the reagent.
5-1
-------
Table 5-1. Sampling Frequency and Physical Measurements at Each Location
Sample
Composite
Start
Time
(hrs)
0
8
16
24
ui 32
1
to
40
Location A - Pipe
Sample Type
8240/8270
6010
8240/8270
6010
8240/8270
6010
8240/8270
6010
8240/8270
6010
8240/8270
6010
Canister
Sample Totals:
Field
Measurements
Liquid Flowrate
Liquid Temperature
Air Temperature
Liquid Flowrate
Liquid Temperature
Air Temperature
Liquid Flowrate
Liquid Temperature
Air Temperature
Liquid Flourate
Liquid Temperature
Air Temperature
Liquid Flowrate
Liquid Temperature
Air Temperature
Liquid Flowrate
Liquid Temperature
Air Temperature
8240 - 6
8270 - 6
6010 - 6
Canister - 1
Location B - Manhole
Sample Type
8240/8270
6010
8240/8270
6010
8240/8270
6010
8240/8270
6010
8240/8270
6010
8240/8270
6010
Canister
Field
Measurements
Liquid Temperature
Air Temperature
Liquid Temperature
Air Temperature
Liquid Temperature
Air Temperature
Liquid Temperature
Air Temperature
Liquid Temperature
Air Temperature
Liquid Temperature
Air Temperature
8240-6
8270 -6
6010 -6
Canister -1
Location C - Manhole
Sample Type
8240/8270
6010
8240/8270
6010
8240/8270
6010
8240/8270
6010
8240/8270
6010
8240/8270
6010
Canister
Field
. Measurements
Liquid Flourate
Liquid Temperature
Air Temperature
Liquid Flowrate
Liquid Temperature
Air Temperature
Liquid Flowrate
Liquid Temperature
Air Temperature
Liquid Flowrate
Liquid Temperature
Air Temperature
Liquid Flowrate
Liquid Temperature
Air Temperature
Liquid Flowrate
Liquid Temperature
Air Temperature
8240 -6
8270 -6
6010 -6
Canister -1
Location D - Junction Box
Sample Type
8240/8270
6010
8240/8270
6010
8240/8270
6010
8240/8270
6010
8240/8270
6010
8240/8270
6010
Canister
Field
Measurements
Liquid
Temperature
Air Temperature
Liquid
Temperature
Air Temperature
Liquid
Temperature
Air Temperature
Liquid
Temperature
Air Temperature
Liquid
Temperature
Air Temperature
Liquid
Temperature
Air Temperature
8240 -6
8270 -6
6010 -6
Canister -1
Notes: 1. Composite aliquots for 8240 were collected at 2-hour intervals for an 8-hour period and composited in the laboratory.
2. Composite aliquots for 8270 and 6010 were collected at 2-hour intervals for an 8-hour period and composited in the field.
3. Discrete field measurements were taken at 2-hour intervals over the 48-hour test period.
4. In addition, grab samples for 8240, 8270 and 6010 were collected prior to tracer inoculation at all four locations.
-------
Table 5-2 Tracer Chemical Reagents
t_n
u>
Chemical Name
Chlorobenzene
1,1,1-Trichloro-
e thane
n-Hexane
Acenaphthene
Cupric Chloride,
Dihydrate
Zinc Sulfate,
Septahydrate
Manufacturer
Fisher Scientific
Aldrich Chemical Co.
Fisher Scientific
Aldrich Chemical Co.
Mallinckrodt
Mallinckrodt
Formula Weight
112.56
133.41
86.18
154.21
170.48
287.54
Purity (%)
99.96
97
87.1
99
99
100.7
Total Quantity
4 liters
4 liters
8 liters
500 grams
12 kilograms
19.5 kilograms
-------
5.1.2 Tracer Feed Systems
The six tracer compounds were inoculated using three separate feed
systems. Each system consisted of a tracer solution reservoir, a Masterflex®
peristaltic pump controller, a pump head for the appropriate flow rate, and
Viton® rubber tubing. The outlet lines from the three pumps were submerged
below the water surface in the inoculation channel.
The hexane/acenaphthene solution was prepared by dissolving 500 grams
of acenaphthene, which is a solid, in 8 liters of n-hexane. This feed
solution had a concentration of 62.5 grams of acenaphthene per liter of hexane
and was pumped out of a 3-gallon glass tracer reservoir.
The chlorobenzene/1,1,1-trichloroethane solution was prepared by
mixing 4 liters of each reagent. The chlorobenzene and the 1,1,1-
trichloroethane were completely miscible and were pumped out of a 3-gallon
glass tracer reservoir.
The metals solution was prepared by dissolving 19.5 kg of zinc sulfate
and 12 kg of cupric chloride in 12 gallons (53L) of distilled water. This
feed solution had a zinc sulfate concentration of 368 g/L and a cupric
chloride concentration of 226 g/L. This solution was pumped out of a 15-
gallon plastic tracer reservoir using Teflon® tubing instead of Viton® rubber
tubing.
5.1.3 Tracer Solution Flow Determination
The feed rate of the tracer solutions was determined throughout the
test using four different techniques. In addition, the density of each feed
solution was determined by weighing one liter on a scale.
Bucket and Stopwatch Method (all three tracer solutions) - At periodic
intervals the flow rate was measured by placing the inoculant line into a 5 ml
graduated cylinder. The accumulated volume was measured for a time period
between 30 seconds and 2 minutes. Three repeat trials were made for each
determination.
5-4
-------
Weight Change Method (all three tracer solutions) - Each tracer
solution reservoir was placed on a scale and tared prior to preparation of the
feed solutions. The reservoirs were kept on the scales throughout the test
and the weights were recorded at each two-hour sample collection interval.
Calibrated Feed Jar Method (hexane/acenaphthene and chloro-
benzene/1,1,1-trichloroethane solutions only) - The level of solution in each
glass reservoir was marked with an indelible marker at 2-hour intervals
throughout the test. At the end of the test the remaining solution was
measured and removed, and each jar was refilled with distilled water using a
graduated cylinder. The incremental volume between each mark on the jar was
recorded.
Total Feed Volume Method (all three tracer solutions) - The remaining
volume of each tracer solution at the end of the 48-hour test period was
measured. This volume was subtracted from the initial volume to calculate the
total feed volume.
5.2 Sampling Procedures
Before the inoculation feed pumps were turned on, grab samples of the
wastewater stream were collected for volatiles, semi-volatiles and metals at
each of the four locations for use as matrix characterization samples. These
were labelled as "Pre" samples.
After 3 hours, tracer inoculation was assumed to have reached
equilibrium. Composite samples were collected over six 8-hour intervals
during a 48-hour period. At each of the four locations the following samples
were collected and measurements made:
• VOA wastewater samples for SW-846 Method 8240 analysis, with each
sample consisting of five VOA subsamples collected after 0, 2, 4, 6,
and 8 hours which were composited in the laboratory;
• Semi-volatile and metals wastewater samples for EPA SW-846 Methods
8270 and 6010 analysis, with each sample consisting of five subsamples
collected every 2 hours, which were composited in the field;
5-5
-------
• Discrete liquid temperature readings at 2-hour intervals;
• Discrete liquid flowrate readings at 2-hour intervals at locations A
and D;
• Discrete headspace temperature readings at 2-hour intervals; and
• One SUMMA®-polished canister headspace gas grab sample for GC/MS
analysis, collected during the last 8-hour sampling interval.
Composited samples were composed of sample fractions collected every 2
hours during each 8-hour sampling interval. Because the fifth 2-hour interval
for one 8-hour sampling period also represented the first 2-hour interval of
the next 8-hour sampling period, the volumes from these intervals added to
composited samples had to be half that of the second, third, or fourth 2-hour
intervals. Samples for volatiles analyses were composited at the laboratory;
semi-volatiles and metals samples were composited on site. A sample
collection matrix is presented in Appendix D.
Isco® peristaltic sampling pumps and the tubing used to collect water
samples were dedicated to each sampling location to avoid cross-contamination
between sampling points. The pump tubing was purged with the wastewater
before each sample.collection. Teflon® tubing was used to collect air and
water samples to avoid interference and contamination from sampling equipment.
All water samples were packed with ice in coolers immediately after
collection, and shipped via Federal Express overnight delivery to Radian's
Perimeter Park analytical laboratory, in Morrisville, North Carolina.
In addition, one-time canister grab samples for GC/MS analysis were
collected at seven system leaks and/or vents along the collection system. The
velocity of the discharge was measured when the sample was collected.
5.2.1 Volatile Wastewater Samples
The volatile organic tracer samples were collected over six
5-6
-------
consecutive 8-hour sampling intervals. Three 40-mL VOA vials were filled with
water every two hours (at 0, 2, 4, 6, and 8 hours), for a total of five sample
subsets per 8-hour interval. Samples were composited at the laboratory, using
the injection syringe to pull an aliquot from one VOA vial for each 2-hour
interval. The aliquot volumes pulled from each selected 2-hour interval vial
were as follows, for each 8-hour composited sample:
0 hour 2 hour 4 hour 6 hour 8 hour
4 mL 1 mL 1 mL 1 mL 4 mL
The remaining VOA vials from each 2-hour interval served as extra samples in
case of breakage. All sample bottles used were certified by the manufacturer
to contain less than 5 /Jg/L of any of 38 different volatile organic compounds.
5.2.2 Semi-Volatile Wastewater Samples
Every 2 hours, a grab sample was collected using a 1-liter glass
graduated cylinder. The cylinder was first rinsed with a small amount of
sample, which was collected in a bucket and discarded into the Gravity Flow
Pumping Station tank after all sampling for that 2-hour sampling episode was
complete. Each subsample was poured into a 2-L glass bottle, which was kept
on ice. After all five 2-hour interval samples were collected, the composite
volume was mixed in the 2-L bottle, and the composited sample poured into two
1-L wide-mouth, glass bottles, which were then labelled, taped with
transparent tape, and stored on ice. The second 1-L bottle served as an extra
sample in case of breakage. The volumes collected at each 2-hour interval
were as follows:
0 hour 2 hour 4 hour 6 hour 8 hour
250 mL 500 mL 500 mL 500 mL 250 mL
5.2.3 Metals Wastewater Samples
Samples for metals analysis were collected similarly to the semi-
volatile wastewater samples, except that 1) 1-L Nalgene plastic bottles and 1-
L Nalgene plastic graduated cylinders were used instead of glass bottles, and
2) samples were preserved with nitric acid to a pH of 2 or less.
5-7
-------
5.2.4 Headspace Gas Samples
At all four of the water sampling sites, air grab samples were taken
during the last 8 hours of the 48-hour sample period. Air samples were
collected using evacuated SUMMA®-polished canisters, and analyzed using GC/MS
at Radian's Perimeter Park analytical laboratory. Canister sample collection
and analysis procedures were based on the EPA Compendium Method TO-14. Air
was drawn into the canister through Teflon® tubing from above the waste
stream. At location A the tubing section was very short (2 to 3 inches) and
was not purged. Prior to sample collection at locations B and C, the tubing
was purged with headspace air using an SKC Inc. Model 224-PCXZR3 Airchek
Sampling Pump at a flow rate of 2.5 L/min for a period of 30 seconds. Purge
time was kept to a minimum to avoid creating a negative pressure within the
manhole locations which might pull in ambient air. Prior to sample collection
at location D, the tubing was purged with headspace air using the same pump at
the same flow rate for a period of 60 seconds because of the extended length
of tubing necessary at that location.
5.2.5 Leak/Vent Gas Samples
In addition to air samples at the manhole locations, grab air samples
were taken at seven locations determined on-site. These sample locations
represented avenues of vapor leakage from the sewer system. Air samples were
collected in evacuated SUMMA®-polished canisters using Teflon® tubing.
Tubing was inserted approximately 1 to 2 inches into each manhole cover. All
samples were collected only when the air flow direction was out of the
manhole. Simultaneous air velocity measurements were taken at each location,
using a Kurz Instruments Inc. Model 490 mini-anemometer. In addition, flow
direction at the time of sampling was determined by RTI personnel using an
Alnor Velometer Jr. Directional Indicator.
5.3 Field Measurements
Liquid and air temperatures were taken at 2-hour intervals over the
48-hour test period at all four locations, using a YSI Model 3000 T-L-C meter.
This portable temperature read-out was carried from one sample location to the
5-8
-------
next, and the probe lowered into the wastewater. This was performed at the
manholes at locations B and C through a pre-drilled sample port that was
capped after temperature measurements were taken. For each manhole, the probe
was lowered to obtain the headspace temperature, then lowered further to
obtain the liquid temperature, and then removed from the manhole. At location
A, the probe was held a few inches above the sample stream for the headspace
temperature, and in the stream for the water temperature. At sample location
D, the probe was lowered to obtain the headspace temperature at the pipe drop-
off point, then removed. The liquid temperature was measured by dropping the
probe into the water just in front of the outlet pipe leading into the
junction box. Air and liquid temperatures were taken after sampling to avoid
disrupting the air-water equilibrium at each location. Ambient air
temperature measurements were made at location B.
The flow in the sewer system at the spiking point and location A was
measured at 2-hour intervals over the 48-hour test period by measuring the
height of the water stream over the rectangular weir placed in the open
channel. The flow rate was measured at the same time as sample collection.
Using the head measurements, along with the crest length of the weir
(12 inches), it was possible to calculate the flow over the weir. The
following equation, taken from Isco Open Channel Flow Measurement Handbook
(page 34), was used to calculate the flow:
Q = K * L * H1'5 (for a suppressed rectangular weir without end
contractions) (5-1)
Where, Q - Flow rate (gallons per minute)
H - Head on the weir (inches)
L - Crest length of weir (12 inches)
K - Constant (1495 for gpm)
The GFPS at sample location D was equipped with two new Gould Model
3171 water pumps. Both pumps had been installed within one year prior to the
test, and had a manufacturer's specified flow rate of 900 gallons per minute
(gpm). Flow was regulated by water level controllers which turned the pumps
on and off at specified levels. Pump clocks were in-place for these pumps
5-9
-------
which monitored totalized pump run time. Flow rates were measured by
recording the total pumping time from the pump clocks and the water level in
the tank at 2-hour intervals over the 48-hour test period.
The total flow into the tank over each 8-hour period can then be
calculated using the following equation:
Vp - [
-------
All preliminary ("Pre") samples collected just prior to spiking were
analyzed for the full list of Method 8240, 8270,and 6010 compounds. Samples
collected after sewer spiking began were analyzed for the spiking target
compounds only: chlorobenzene, hexane, and 1,1,1-trichloroethane by Method
8240, acenaphthene by Method 8270, and copper and zinc by Method 6010.
Acenaphthene proved difficult to spike into the sewer stream at a
constant rate. It was found that as temperatures cooled during the evening
hours, acenaphthene crystallized out of the acenaphthene/hexane
solution,clogging up the delivery lines. For this reason, only samples which
were not affected by this problem were analyzed for acenaphthene. These
included the Pre, 0, and 40-hour samples.
Due to the elevated levels of methyl methacrylate present in the
wastewater at location D during the characterization study, which caused
analytical difficulties with Method 8240, an extended bake-off period for the
Tekmar analytical trap in the instrument system was allowed between sample
analyses. During the analysis of the field test samples, which were not
diluted to reduce the concentration of methyl methacrylate, a hexane matrix
spike was performed after every fifth sample analysis. Hexane recovery
remained acceptable (100% ± 20%), so it was not necessary to replace the
Tekmar trap and reanalyze field samples to determine recoveries for hexane,
chlorobenzene, and 1,1,1-trichloroethane. The extended bake-off period
appeared to restore the efficacy of the Tekmar analytical trap sufficiently so
that the adsorbing surfaces were not saturated and acceptable surrogate
recoveries and matrix spike recoveries could be obtained.
Because of the focus on hexane, chlorobenzene, and 1,1,1-trichloro-
ethane with concentration adjustments to keep these three field-spiked
compounds within the instrument calibration range, it was not possible to
simultaneously accurately quantify all of the 8240 analytes which might have
been present in the field samples. Samples collected prior to tracer
inoculation were analyzed undiluted for the complete list of 8240 compounds.
As the concentration of the semi-volatile tracer compound,
acenaphthene, decreased along the wastewater collection system line due to
losses and dilution, it was expected that the concentration would drop below
5-11
-------
Table 5-3. Analytical Procedures
Sample Type
Analysis
Canisters
Volatiles by GC/MS (EPA Compendium Method
T014)
Liquids
Volatiles (EPA SW-846 Method 8240)
Semi-Volatiles (EPA SW-846 Method 8270)
Metals (EPA SW-846 Method 6010)
5-12
-------
analytical detection. For this test, insufficient acenaphthene was inoculated
to achieve detection at location D due to the facility's NPDES permit limits.
Therefore, a minor modification to Method 8270 was made. This modification
involved sample extract concentration to 300 microliters instead of 1
milliliter. This effectively achieved approximately a three-fold dilution of
the samples from location D. This procedure was checked and surrogate
recoveries were checked prior to performing the procedure on field samples.
No adverse effect on quantitation was observed.
5-13
-------
6.0 QUALITY ASSURANCE/QUALITY CONTROL RESULTS
The overall quality, or Quality Assurance (QA), of data for this
project is defined as the combination of all Quality Control (QC) samples and
procedures. The quality control regimen for this project included field
measurement QC, analytical QC, and field sample QC.
Overall the data met the quality assurance objectives with a few
exceptions. Blank sample results indicated that there was no systematic
sample contamination in the field, during sample handling and transport, and
during analysis. Recovery factors were consistent, indicating there was no
matrix effect on tracer recovery between sample locations or at different
sample concentrations. The following problems were noted which are discussed
in detail below:
• Acenaphthene exhibited poor recovery (35-60%) in matrix spike,
matrix spike duplicate, and add spike samples, which suggests that
field sample recovery was also low; and
• Canister sample field duplicate results indicated a possible low
bias for chlorobenzene in diluted samples.
The QC procedures and results are presented in the sections below.
Field sample and analytical QC discussions are provided for each of the two
sample types: liquid samples and canister samples.
Although the test plan mentioned the possibility of external
performance audit samples being provided by the EPA Emissions Measurement
Branch for analysis by Radian, no audit samples were received or analyzed.
6.1 Field Measurement Quality Control
The test plan discussed the potential use of Isco Flow Pokes® for
measuring the wastewater stream flow rate at sampling location A. Since the
original sampling location A at the manhole was not sampled, the Flow Pokes®
were not used. Instead, flow rates were measured at the new sample location A
6-1
-------
by use of a weir which was shaped to conform to the channel dimensions.
Stream height over the weir was measured every 2 hours, concurrent with
sampling at location A.
The GFPS pumping rate was determined in a single trial, before testing
began, by measuring the change in tank level over time with the inflow blocked
off. Three of these initial measurements were called for in the test plan.
However, since taking this measurement required plant personnel to stop the
flow into the GFPS tank, only one initial flow measurement was made to avoid
backing up the plant's sewer and overflowing the manholes.
Tracer feed rates and total volume of delivered solution were measured
in four different ways, as a cross-check. The metals and organics were stored
in containers resting on scales, from which readings were recorded every 2
hours. Feed rates were also confirmed using a graduated cylinder and
stopwatch periodically. The level in each container was marked every 2 hours.
Lastly, the volumes remaining in each container were measured at the end of
the test to verify the total amount delivered to the sewer system.
Acenaphthene crystallized out of solution at lowered (nighttime) temperatures
and therefore constant flow of the acenaphthene/hexane solution could not be
maintained.
6.2 Liquid Sample Quality Control Results
Both field sample quality control and analytical quality control
measures were instituted for the liquid samples collected for this project.
Analytical QC included the analysis of method blanks to assess laboratory-
introduced contamination, and matrix spike/matrix spike duplicates to assess
both analytical accuracy and analytical precision. Field sample QC included
the collection and analysis of field blanks and trip blanks to help assess
possible modes of sample contamination, and field duplicates to assess overall
reproducibility of sample results. Results for these QC measures are
presented below. Other analytical QC procedures described in the test plan,
and in SW-846 for each respective method, were followed as well; these results
are kept with the laboratory data at Radian's Perimeter Park laboratory.
6-2
-------
6.2.1 Blanks
Trip and field blanks were collected for analysis of the target
volatiles "and semi-volatiles. Table 6-1 presents the results. One of the two
trip blanks was analyzed for volatiles; the other for semi-volatiles. One of
the two field blanks was analyzed for volatiles; the other for volatiles and
semi-volatiles. Although all analyses were not performed on all the blank
samples , enough information is available to make an assessment as to sources
of possible sample contamination.
Method blanks were analyzed by each method daily, for each day of
sample analysis. Table 6-2 provides the method blank results. As seen in
this table, no target analytes were detected in any of the method blanks other
than a one-time occurrence of acenaphthene (0.77
Small amounts of chlorobenzene and 1,1, 1-trichloroethane were detected
in one trip blank and in both of the field blanks . The concentrations of
these compounds were similar for both types of blank samples. Since these
compounds appear at equivalent levels in the trip and field blanks , but do not
appear in the method blanks, it would appear that the sampling and handling
procedures may have contributed small levels of these compounds to the field
samples. However, the concentrations of chlorobenzene and 1,1,1-
trichloroethane in the field samples were generally orders of magnitude above
the levels found in the trip and field blanks; it is unlikely that the
contributions described here had notable impacts on sample results.
Neither hexane nor acenaphthene were detected in either the trip
blanks or the field blanks .
6-3
-------
Table 6-1. Trip and Field Blank Results - Liquid Samples (pg/L)
Sample
Trip
Blank 1
Field
Blank 1
Trip
Blank 2a
Field
Blank 2
Sample
Date
6/26/91
6/26/91
6/27/91
6/27/91
Chlorobenzene
0.94
1.1
N/A
1.1
1,1,1-
Trichloroethane
0.67
1.3
N/A
1.3
Hexane
<5
<5
N/A
<5
Acenaphthene
N/A
N/A
<10
<10
N/A - Not Analyzed
* - Trip Blank 2 was inadvertently omitted from Method 8240 analysis on the chain-of-custody.
6-4
-------
Table 6-2. Method Blank Results - Liquid Samples (/ig/L)
Volatiles (Method 8240)
Analysis Date
6/28/91
7/1/91
7/2/91
7/25/91
8/12/91
Semivolatiles (Method 8270)
Analysis Date
Chlorobenzene
<5
<5
<5
<5
<5
Acenaphthene
1.1. 1-Trichloroethane
<5
<5
<5
<5
<5
Hexane
<5
<5
<5
<5
<5
7/19/91 <10
8/5/91 0.77
8/5/91 <10
8/6/91 <10
6-5
-------
6.2.2 MS/MSD Results
Matrix spike (MS) samples are samples to which a known quantity of
analyte(s) is added. Evaluation of the percent recovery of these spiked
analytes allows for assessment of the method's accuracy. Each sample selected
for spiking was spiked in duplicate. The analysis of matrix spike duplicates
(MSDs) allows assessment of the analytical variability.
Table 6-3 gives results for the MS/MSD samples. Each of the six
tracer compounds was used to spike the MS/MSD samples. As seen from the
table, nearly all spiked samples showed acceptable analytical recoveries.
Hexane showed one recovery that was unexpectedly high (188%); however, the
other three recoveries were nearer 100% and no trends appear that suggest
consistent problems in quantitating hexane. Acenaphthene showed the potential
for low recovery with recoveries of 60 and 34% for the one sample that was
spiked in duplicate. Since the other MS/MSD sample was not analyzed for
acenaphthene, an assessment of systematic low recovery for this chemical could
not be confirmed or disproved.
Analytical variability was generally within the Radian QA/QC goals of
Relative Percent Difference (RPD) values < 50% for volatiles and semi-
volatiles, and < 20% for metals. In two instances the RPD value fell outside
these specifications; once for hexane (at 86%) and once for acenaphthene (at
55%). As the other RPD for hexane was only 4%, no consistent precision
difficulties appear to be present for the analysis of hexane. It is not
possible to assess whether there is a consistent problem in analytical
precision for acenaphthene, as MS/MSD analyses were run on only one of the two
samples. However, the RPD value available for acenaphthene is not far from
the expected level of precision. Furthermore, replicate analyses of the Add-
samples for acenaphthene at sample locations B and C give RPD values less than
50% (see Appendix C). Therefore, it is unlikely that analytical imprecision
is unduly large for acenaphthene.
6-6
-------
Table 6-3. MS/MSD Results - Liquid Samples
Compound
A-16
MS MSD
(% recov.) (% recov.) RPD
D-32
MS MSD
(% recov.) (% recov.) RPD
Chlorobenzene
1,1,1-
Trichloroe thane
Hexane
Acenaphthene
Copper
Zinc
79
92
188
60
100
92
88
84
75
34
102
92
11
9
86
55
2
0
113
153
115
N/A
88
81
116
126
120
N/A
89
84
3
19
4
NC
1
4
RPD - Relative Percent Difference
N/A - Not Analyzed
NC - Not Calculated
Underlined values fall outside the QA/QC goal of RPD <50I for volatiles and semi-volatiles, or <20Z for metals.
6-7
-------
6.2.3 Field Duplicates
Field duplicate samples are samples collected in the same manner and
location as the original sample. They provide insight as to the total
variability of the sample data, including sampling and analytical variability
and sample homogeneity. Four field duplicate samples were collected over the
48-hour sampling period, one at each of the four sample locations.
Table 6-4 provides the field duplicate results, reporting both the
concentrations found for the original field sample and its duplicate.
Relative Percent Differences (RPDs) are also calculated for each duplicate
set. RPDs are calculated as the difference between the two sample recoveries,
divided by the average and expressed in percent.
As seen from this table, good reproducibility in metals concentrations
was seen for all four sets of duplicates. RPD values for copper and zinc
analyses ranged from zero to seven percent. Higher variability was found for
chlorobenzene and 1,1,1-trichloroethane (variability could not be determined
for hexane, as it was not detected in any of the field samples).
Variability was relatively low for 1,1,1-trichloroethane in the field
duplicates; RPD values ranged from 10 to 98%, with three of the four values
falling at or below 40 percent. The RPD values that could be calculated for
chlorobenzene ranged from 54 to 128 percent. This higher variability appears
to be independent of concentration. Since the analytical variability for
chlorobenzene was low (see Section 6.2.2) it would appear that the major
contributor to the variability seen here for chlorobenzene was due to either
the sampling and handling processes, or to sample nonhomogeneity. Field
sample results for chlorobenzene should be reviewed with the potential for
increased variability in mind.
6.2.4 Add Spike Sample Results
In order to determine whether the wastewater stream matrix would
impact the recovery of the tracer compounds in the field samples, extra field
samples were collected at varying times for each sample location. These
samples, the "Add" samples, were later spiked with known concentrations of the
6-8
-------
Table 6-4. Field Duplicate Results - Liquid Samples
Sample ID Duplicate ID Analyte
A-40 ' G-40 Chlorobenzene
1 , 1 , 1-Trichloroethane
Hexane
Copper
Zinc
B-40 H-40 Chlorobenzene
1 , 1 , 1-Trichloroethane
Hexane
Copper
Zinc
Zinc
C-8 E-8 Chlorobenzene
1 , 1 , 1-Trichloroethane
Hexane
Copper
Zinc
D-8 F-8 Chlorobenzene
1 , 1 , 1-Trichloroethane
Hexane
Copper
Zinc
Concentrations
(Mg/L)
120000
120000
<5
1700
2000
1000
2600
<5
1400
1500
170
35
99
<5
150
180
ND
21
<5
140
1500
57000
98000
<5
1800
1900
220
3900
<5
1400
1600
180
20
34
<5
150
180
9.0
19
<5
140
1600
RPD
71
20
NC
6
5
128
40
NC
0
7
6
54
98
NC
0
0
NC
10
NC
0
7
RPD - Relative Percent Difference.
NC - Not Calculated; one or both concentrations for this analyte were Not Detected.
Note: Field duplicate samples were not analyzed for acenaphthene.
6-9
-------
tracer compounds, at roughly half of the original sample concentration for
each compound. (Pre samples, whose original tracer concentrations were very
low, or non-detectable, were arbitrarily spiked with 50 ppb of organic tracers
and varying amounts of the metal tracers.)
Additionally, selected samples were spiked again with concentrations
roughly equal to the original sample concentration. This was done in order to
determine whether the variability in surrogate recoveries would increase with
varying sample spike concentrations. These samples included the 0-samples
from location D, and the Pre samples. Pre samples were spiked with
concentrations double the original spike concentration.
Table 6-5 presents the spike recovery results for the Add samples.
Generally, there appears to be no notable increase in variability from
the Pre samples to the samples collected after tracer inoculation began. If
anything, the variability appears to decrease after inoculation began. This
may be due to the increased difficulty in recovering .small amounts of
chemicals that could be expected with the low initial spike compound
concentrations in the Pre samples. While some variability is noted over time
and from location to location for chlorobenzene, no upward or downward trends
are apparent in spike recovery for any of the target compounds. No recovery
factor adjustment appears necessary for sample results, with the possible
exception of acenaphthene. Acenaphthene showed poor recovery for both the
Pre- and 0-samples. This matches the relatively low recoveries seen for
acenaphthene in the MS/MSD samples.
Further, no trend of notable differences in recovery is noted from the
two levels of spike concentrations. It would not appear that spiking levels
contributed a marked impact on spike recoveries.
6.2.5 Surrogate Spikes
Surrogate spikes are compounds added to all samples, blanks, and
calibration standards undergoing organic compound analysis. These compounds
are not normally found in environmental samples; they indicate when an
analytical problem exists with the instrumentation, or when the instrument is
6-10
-------
Table 6-5. "Add" Spike Sample Results
CTi
I
Pre
Sample
Location
Chlorobenzene A
B
C
D
Hexane A
B
C
D
1,1,1-Trichloroethane A
B
C
D
Acenaphthene A
B
C
D
Copper A
B
C
D
Zinc A
B
C
D
* Value to Che average recovery from repeated
b Two separate spike samples were prepared and
Amount
Spiked
(Hg/L)
50/100
50/100
50/100
50/100
50/100
50/100
50/100
50/100
50/100
50/100
50/100
50/100
200/400
200/400
200/400
400
28/56b
8/16b
4/8b
4/12"
100/200b
100/200"
40/80b
40/80"
% Recov.
85a/82
85/90
95/82
81/99
97a/94
86/88
118/90
94/95
96a/90
97/91
102/98
100/104
50a/40a
58/42
70/40
41
102/90
100/88
85/82
120/125
111/91
84/97
99/102
99/89
analytical determinations for
analyzed. The
0
Amount
Spiked
57
130
5.7
N/S
N/S
N/A
N/A
N/S
15
330
94
12/20b
200
400
200
200
4000
4000
4000
400/1200"
8000
4000
4000
40/1200"
this sample.
first spike amount is rouahlv
% Recov.
85
118
89
-
-
98
91
-
120
100
100
100/110
54
57
60
21
98
98
93
97/90
97
98
97
101/89
SOX of the or it
Amount
Spiked
68
3.8
45
N/S
N/S
N/S
N/A
N/S
32
7.7
150
58
NA
NA
NA
NA
4000
4000
400
400
4000
4000
400
4000
Einal sample
16
% Recov.
86
77
58
-
-
-
88
-
96
92
78
96
-
-
-
-
92
101
95
94
92
102
97
98
concentration (
Amount
Spiked
64 '
6.1
4.3
4.6
N/A
N/A
N/A
N/A
57
8.1
7.0
65
NA
NA
NA
NA
3200
2800
400
400
4000
3600
400
400
24/32c
% Recov.
81
100
80
103
88
88
115
88
76
102
103
100
-
-
-
-
95
102
101
102
97
101
100
101
IK for dilution) :
the second is roughly 100X).
c Organic samples from the 24-hour composite period were spiked; metals samples from the 32-hour composite period were spiked.
NA - Hot available.
N/S - Not spiked.
-------
not responding to a particular class of compounds. Metals samples are not
spiked with surrogates.
Surrogate recoveries were acceptable for the volatile organic
analyses. Of the 51 samples and blanks that were reported, only one surrogate
fell below the acceptance criteria, and by only one percentage point.
Surrogate recoveries fell outside the expected recovery range more
frequently for semi-volatiles. Of the 28 samples and blanks which were
analyzed, more than half had at least one surrogate recovery outside the
expected range. The most frequently missed surrogate recoveries were for 2-
fluorobiphenyl (9 recoveries missed), 2-fluorophenol (13 recoveries missed),
and nitrobenzene-d5 (9 recoveries missed). In general, recoveries were lower
than expected more often than they were high.
The most meaningful surrogate to relate to acenaphthene is 2-
fluorobiphenyl. Chemically it is the most similar and the retention times are
reasonably close. The low recovery observed for this surrogate generally
supports the low recoveries observed for MS/MSD and Add Spike samples.
6.2.6 Metals Quality Control and Standard Addition Samples
Two types of QC samples were analyzed concurrent with the metals
analyses: quality control samples, and standard addition samples. Recoveries
for the quality control samples, which were analyzed several times along the
sample queue, ranged from 99 to 103 percent. Recoveries for the standard
addition samples ranged from 102 to 111 percent. Recoveries were lower for
the MS/MSD standard addition samples, ranging from 81 to 92 percent. Results
may be found with the sample results in Appendix C.
6.3 Canister Sample QC Results
As with the liquid samples, quality control samples were collected and
analyzed with the 11 canister samples. These quality controls included two
sets of field duplicates and one trip blank. Method blanks were also analyzed
daily for each day of field sample analysis. In general, no systematic
problems with sample contamination were found. Field duplicate results
6-12
-------
indicate the possibility that the canister dilution process had an impact on
analytical accuracy.
Method Blanks
Two method blanks were analyzed concurrently with the field samples;
one on 7/25/91, and the second on 8/12/91. Results are presented in Table 6-
6. None of the three target volatiles, or any of the other TO-14 volatiles
were reported in either of the method blanks.
Trip Blanks
One canister trip blank was carried through all the sampling and
handling processes with the field samples. This trip blank was then analyzed
with the field samples; results are listed with the method blanks in Table 6-
6. Neither chlorobenzene nor hexane was detected in the trip blank. A small
quantity of 1,1,1-trichloroethane was detected, at 3.2 ppbv. This amount
represents only a small fraction of the levels of 1,1,1-trichloroethane found
in most of the canister field samples. For samples containing less than 100
ppbv of this compound, the possibility of minor contributions from the
sampling and handling procedures should be considered.
Field Duplicates
Two sets of canister field duplicates were collected. One set was
collected at one of the liquid sampling points (location B), and the other at
leak/vent location L/V-6. Table 6-7 presents the field duplicate results,
providing both sample concentrations and the RPD values for each duplicate
set.
The RPD results for the L/V duplicates indicates good reproducibility
between samples. The results for the location B duplicates show lowered
reproducibility. The relatively high RPD values (and concurrent imprecision)
for hexane may be explained by the low concentrations being compared. Higher
variability may be expected as concentrations approach the detection limit,
and in this case, the hexane concentrations are roughly six and three times
6-13
-------
Table 6-6. Method and Trip Blank Results - Canister Samples (ppbv)
Blank ID Chlorobenzene Hexane 1,1,1-
Trichloroethane
Method Blank (7/25/91) ND ND ND
Method Blank (8/12/91) ND ND ND
Trip Blank ND ND 3.2
ND - Not Detected.
6-14
-------
Table 6-7. Field Duplicate Results - Canister Samples
Sample ID Duplicate
B-42 ' H-42
L/V-6 L/V-11
ID Analyte
Chlorobenzene
Hexane
1,1, 1 -Trichloroethane
Chlorobenzene
Hexane
1 , 1 , 1 -Trichloroethane
Concentrations
(ppbv)
1400
6.4
1400d
3600
220
29000d
470d
2.6
1000d
4200
240
31000d
RPD
100
84
33
15
9
7
d - Sample was diluted for this analyte.
6-15
-------
the detection limit for the location B canister sample and its duplicate,
respectively.
Chlorobenzene also exhibited low reproducibility in the location B
field duplicates, with an RPD value of 100 percent. The chlorobenzene
concentration for the location B canister sample (1400 ppbv) is several orders
of magnitude above the detection limit. The duplicate canister for location
B, however, was substantially diluted, and the quantitated concentration (470
ppbv) is much nearer the detection limit for chlorobenzene following dilution.
Canister dilution involves the addition of a known amount of ultra-high purity
nitrogen into the canister. It is possible that the dilution process may have
exerted a negative bias on the volatile results; results from diluted
canisters should be viewed with this possible bias in mind.
6.4 Sample Chain-of-Custody
Chain-of-custody forms accompanied the samples from the field, through
analysis, and to sample storage. Chain-of-custody forms documented sample ID
numbers, composite fraction (if applicable), sample dates, types of analyses
to be performed, preservation techniques, shipment dates, and the identity of
all personnel who handled the samples. Copies of the chain-of-custody forms
are presented in Appendix F. The forms were stored in plastic bags for
shipment with the samples. Sample coolers were sealed with chain-of-custody
seals to indicate, whether the sample containers had been compromised during
shipping. Samples were shipped from the field to the laboratory by next-day
express mail service.
6.5 Field Records
All field notes and raw data were recorded in bound laboratory
notebooks which were kept at each wastewater sampling location. Copies of the
notebook pages are presented in Appendix G.
6-16
------- |