United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park, NC 27711
EMB Report 85-FPE-04
July 1986
Air
Hazardous Waste
Treatment, Storage, and
Disposal Facilities
Site-Specific Test Report
Texaco Refining
and Marketing Company
Delaware City, Delaware
-------
SITE-SPECIFIC TEST REPORT
TEXACO REFINING AND MARKETING, INC.
DELAWARE CITY, DELAWARE
ESED 85/12
EMB 85FPE04
Prepared by:
Entropy Environmentalists, Inc.
Post Office Box 12291
Research Triangle Park, North Carolina 27709
Contract No. 68-02-3852 and 68-02-4336
Work Assignment No. 3024 and No. 3501
EPA Task Manager
Clyde E. Riley
U. S. ENVIRONMENTAL PROTECTION AGENCY
EMISSION MEASUREMENT BRANCH
EMISSIONS STANDARDS AND ENGINEERING DIVISION
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
July 1986
-------
DISCLAIMER
This document has been reviewed by the Emission Standards and Engineering
Division, Office of Air Quality Planning and Standards, Office of Air, Noise
and Radiation, Environmental Protection Agency, and approved for publication.
Mention of company or product names does not constitute endorsement by EPA.
Copies are available free of charge to Federal employees, current contractors
and grantees, and nonprofit organizations - as supplies permit - from the
Library Services Office, MD-35. Environmental Protection Agency, Research
Triangle Park, NC 27711.
Order: EMB Report 85-FPE-04
11
-------
CONTENTS
Page
Figures v
Tables vi
1.0 INTRODUCTION 1-1
2.0 SUMMARY AND DISCUSSION OF RESULTS 2-1
2.1 Background Samples 2-4
2.2 Land Treatment Unit, Cell #4 (Process L) 2-10
2.3 Land Treatment Unit, Cell #8 (Process N) 2-12
2.4 Land Treatment Unit, Cell #3 (Process 0) 2-13
2.5 Unpaved Roads in Land Treatment Unit (Process M) 2-14
2.6 Repeatability and Reproducibility 2-16
2.7 Conclusions 2-18
3.0 PROCESS DESCRIPTION 3-1
3.1 Land Treatment Cells 3-1
3-2 Unpaved Roadways 3~3
4.0 SAMPLING AND ANALYSIS 4-1
4.1 Site Plot Plan 4-1
4.2 Land Treatment Unit, Cell #4 (Process L) 4-2
4.3 Land Treatment Unit, Cell #8 (Process N) 4-11
4.4 Land Treatment Unit, Cell #3 (Process 0) 4-15
4.5 Unpaved Roads in Land Treatment Unit (Process M) 4-18
4.6 Repeatability, Reproducibility, and Performance
Audit Samples 4-20
4.7 Background Samples 4-21
5.0 QUALITY ASSURANCE 5-1
111
-------
CONTENTS (continued)
APPENDICES Page
A RAW FIELD DATA AND SAMPLING LOGS A-l
Process Data Sheets and Sampling Grid Sketches A-3
Repeatability, Reproducibility, and Quality Assurance Samples A-9
Chain of Custody Forms A-10
B ANALYTICAL DATA B-l
EMB Split Sample Inventory B-3
Moisture Determination Data Sheets B-6
Screening Data Sheets B-43
Percent PM1(-, Determination Data Sheets 6-98
Metals Analysis Results B-110
Organic Cleanup Data Sheet B-118
Organics Analysis Results B-119
Quality Assurance Data B-135
Oil and Grease Analysis Results B-137
SAMPLING AND ANALYTICAL PROCEDURES C-l
Sampling Apparatus C-3
Sampling Location Selection and Documentation C-8
Sample Collection C-ll
Sample Handling and Transport C-14
Drying and Sieving Procedures C-16
Chemical Analyses C-19
Quality Assurance (QA) Procedures C-25
D SAMPLING PROGRAM PARTICIPANTS AND OBSERVERS D-l
PROCESS OPERATIONS DATA E-l
Summary of Processes Sampled During Site Survey E-3
Summary of Equipment for Processes Sampled During Site Survey E-4
IV
-------
FIGURES
Number Page
4.la Site plot plan of Texaco in Delaware City. 4-3
4.Ib Enlargement of site plot plan showing locations of land
treatment cells and sampling locations for background
and unpaved road samples. 4-5
4.2 Sampling grid, process dimensions, and sample numbers
for land treatment unit, Cell #4 at Texaco, Delaware City
(Process L). 4-6
i }
4.3 Sampling grid, process dimensions, and sample numbers
for land treatment unit, Cell #8 at Texaco, Delaware City
(Process N). 4-13
4.4 Sampling grid, process dimensions, and sample numbers for
land treatment unit, Cell #3 at Texaco, Delaware City
(Process 0). 4-16
4.5 Dimensions and sample numbers for the segments of unpaved
roads sampled in the land treatment unit at Texaco
(Process M). 4-19
C.I Example process grid. C-10
C.2 Label used for sample jars. C-12
-------
TABLES
Number Page
2.1 Sampling Plan for Texaco 2-3
2.2 Analytical Results of Silt Screening, Weight Loss on
Drying, and PMin Sieving, Fugitive Particulate from
TSDF (85/12), Texaco, Delaware City, DE 2-5
2.3 Analytical Results for Metals and Semivolatile Organics,
Fugitive Particulate from TSDF (85/12) 2-8
2.4 Summary of Oil and Grease Analysis, Fugitive Particulate
from TSDF (85/12) 2-11
2.5 Analytical Results for Repeatability and Reproducibility
Samples, Metals, Fugitive Particulate from TSDF (85/12) 2-17
4.1 Sample Drying Procedure Summary 4-9
4.2 Metals, Measurement Methods, and Detection Limits 4-10
4.3 Semivolatile Organics Analyzed For and Their Quantifiable
Detection Limits at Medium Concentration Levels 4-12
5.1 Quality Assurance Results for Metals Analysis 5~2
5-2 Quality Assurance Results for Semivolatile Organics Analysis 5~3
C.I Sampling Equipment Specifications C-5
C.2 Sampling Equipment Preparation and Clean-Up C-7
C.3 Metals and Measurement Methods C-20
C.4 Semivolatile Organic Compounds Measured C-22
C.5 Pesticides Analyzed For and Their Quantifiable
Dectection Limits C-24
C.6 Spiking Compounds: Acid Extractables II C-27
C.7 Spiking Compounds: Neutral Extractables V C-28
C.8 Spiking Compounds: Neutral Extractables VI C-29
C.9 Spiking Compounds: Pesticides II C-30
C.10 Spiking Compounds: Metals C-31
VI
-------
1.0 INTRODUCTION
On October 1? and 18, 1985, Entropy Environmentalists, Inc. collected
soil samples from four treatment, storage, and disposal related processes at
Texaco Refining and Marketing, Inc. (Texaco) located in Delaware City,
Delaware. The purpose of this sampling program was to provide preliminary
data on the magnitude of fugitive particulate emissions from various
processes at treatment, storage, and disposal facilities (TSDF's) and the
degree to which these emissions are contaminated. The U. S. Environmental
Protection Agency (EPA) anticipates utilizing the analytical data from this
program with emission models to estimate fugitive particulate emissions from
TSDF's. The information generated by this study may ultimately be used by
the Office of Air Quality Planning and Standards (OAQPS) of EPA to assess the
adequacy of regulations governing fugitive particulate emissions from TSDF's.
To accomplish the overall goals of this study, soil samples were
collected from representative processes at this plant and were submitted for
the appropriate analyses in order to determine the following:
The percent by weight of silt in the soil (i.e., material that
passes through a 200 mesh screen and has a nominal diameter
less than 75 urn) and the percent by weight of moisture in the
soil.
The degree of contamination in the soil silt fraction of
metals, semivolatile organics, and pesticides.
The percent by weight of soil silt that is less than 20 urn in
diameter based on a sonic sieving technique.
1-1
-------
o The particle size dependency of the degree of contamination
(i.e., greater or lesser degree of contamination in particles
with diameters not in excess of 20 urn) by conducting separate
analyses of different soil particle size fractions.
o The repeatability and reproducibility of the sampling and
analytical procedures for the entire sampling program (only
raw data are included in this report; a statistical summary
will be presented for all sampling sites in a later report).
At Texaco, the four processes sampled were three different cells within the
land treatment unit and unpaved road segments (3) within the land treatment
site. A pair of background samples along with samples to assess the
repeatability and reproducibility of the method were also taken.
Samples taken were analyzed for silt content, PMin content, metals,
cyanide, and semivolatile organics as described in Chapter 4. Research
Triangle Institute (RTI) conducted the analyses for metals and PEI and
Associates performed the analyses for the semivolatile organics. Additional
cleanup of the semivolatile organic extracts was performed by Triangle Labor-
atories, Inc. The outside laboratory that performed the metals analysis on the
reproducibility samples was PEI and Associates. For a cost benefit, all the
reproducibility samples for the entire study were analyzed at the same time.
EPA decided not to have the repeatability and reproducibility samples analyzed
for semivolatile organics because of their high oil and grease content.
Field sampling was performed by Mr. Steve Plaisance and Mr. Bernie von
Lehmden of Entropy Environmentalists. Mr. Phillip Englehart and Dr. Tom Lapp
of Midwest Research Institute (MRI) directed Entropy personnel regarding
specific processes to be sampled, the boundaries of the processes, and recorded
the pertinent process and operating characteristics. Mr. Gene Riley (EPA Task
Manager) of the Emission Measurement Branch (EMB) observed the sampling
<4
program. Mr. Bob Wojewodski, Senior Environmental Engineer, served as the
principal contact for Texaco.
1-2
-------
This report is organized into several chapters that address various aspects
of the sampling and analysis program. Immediately following this chapter is
the "Summary and Discussion of Results" chapter which presents table summaries
of data on silt and PMir) content and degree of contamination for each sample
fraction analyzed. Following the "Summary and Discussion of Results" Chapter
is the "Process Description" chapter (supplied by MRI) which includes
descriptions of each process sampled. The next chapter, "Sampling and
Analysis," presents the plot plan and sampling grid for each process. The
method of selecting the sampling grid and the sample collection procedures are
outlined, including any deviations and problems encountered. This chapter also
describes the sample preparation and analytical procedures used for each
sample; any deviations from the normal procedures are addressed. The
appendices present the Raw Field Data and Sampling Logs (Appendix A);
Analytical Data (Appendix B); detailed Sampling and Analytical Procedures
(Appendix C); Sampling Program Participants and Observers (Appendix D); and
Process Operations Data (Appendix E).
1-3
-------
2.0 SUMMARY AND DISCUSSION OF RESULTS
This chapter presents a summary of the sampling and analysis results and a
brief discussion of significant deviations from the proposed sampling and anal-
ysis protocol for this program. Since the standard sampling and analytical
procedures are not addressed in this chapter, it is recommended that those
individuals who are not familiar with the sampling and analytical procedures
used in this study review Chapter 4, "Sampling and Analysis," prior to reading
this chapter.
Soil samples were collected from four processes at Texaco's refining and
marketing facility located in Delaware City, Delaware. The processes included:
(1) the land treatment unit, Cell #4, (2) the land treatment unit, Cell #3;
(3) the land treatment unit. Cell #8; and (4) three segments of unpaved roads
within the land treatment unit. Sampling and analysis were conducted using the
procedures described in the Sampling and Analysis Protocol which was written
specifically for this sampling program. The protocol was provided to the
facility prior to the sample collection. The procedures described in this
protocol are described again in detail in Chapter k and Appendix C of this
report.
As described in the Sampling and Analysis Protocol, this site-specific
report is intended to present the data relevant to the samples obtained at one
site in this study and the procedures used to obtain these samples. Some
statistical analyses will be performed on the data concerning this site;
however, the majority of statistical analyses will involve the data collected
over the entire study, including the repeatability and reproducibility and
quality assurance data, and will be included in the summary report to be
2-1
-------
completed at the conclusion of the program. With the exception of the data
from the screening conducted to determine silt contents, there is not
sufficient data to conduct meaningful statistical analyses on a site- or
process-specific basis.
The sampling plan for Texaco is shown in Table 2.1. The sampling
procedures were designed to obtain a representative sample of that portion of
the contaminated soil with the potential to become airborne. The analyses of
the collected samples were designed to measure the concentration of the most
likely compounds or elements that could be soil contaminates (metals, cyanide,
semivolatile organics, and pesticides). The sample collection techniques were
generally as follows: (1) for undisturbed hard surfaces a sweeping technique
was used to obtain surface samples only; (2) for moderately disturbed surfaces
a scooping technique was used to obtain near surface samples; and (3) for
surfaces that were mechanically disturbed to a specific depth, coring was used
to sample to the depth of the disturbance. The number of samples collected
within each process was a function of the variability expected in the degree of
contamination and/or the amount of sample that was needed for the analyses.
According to the Sampling and Analysis Protocol, the collected samples were
to be analyzed for metals, cyanide, and the semivolatile organics and
pesticides found on the hazardous substance lists (HSL) in the EPA Contract
Laboratory Program (CLP), Statement of Work. If significant quantities of
cyanide, semivolatile organics, or pesticides were not expected to be present
in a particular process, then analysis for those compounds was not performed.
MRI decided that at this particular site, cyanide and pesticides would not be
present in significant quantities in any process sampled, and therefore,
cyanide and pesticide analyses were deleted. All samples were analyzed for
metals and semivolatile organics, except the repeatability and reproducibility
2-2
-------
TABLE 2.1. SAMPLING PLAN FOR TEXACO
Process
Sampled
Land Treatment
Unit, Cell #4
Land Treatment
Unit, Cell #8
Land Treatment
Unit, Cell #3
Unpaved Roads
Within Land
Treatment Unit
Background
Samples
Within Land
Treatment Unit
Process
Designation
L
N
0
M
BCD
Number of
Samples
8*
8*
8
8
3
2
Collection
Method
Modified
coring**
(stainless
tube
Modified
coring**
(plastic
tube)
Scooping
Scooping
Sweeping
Scooping
Analyses
Loss on Drying
Silt and PM1Q Content
Semivolatile Organics
Oil and Grease content
Loss on Drying
Silt and PM Content
Metals
Loss on Drying
Silt and PM Content
Metals
Semivolatile Organics
Oil and Grease Content
Loss on Drying
Silt and PM Content
Metals
Semivolatile Organics
Oil and Grease Content
Loss on Drying
Silt and PM Content
Metals
Semivolatile Organics
Loss on Drying
Silt and PM Content
Metals U
Semivolatile Organics
* One to two cores for metals analysis (plastic core tube) and one to two cores for
organics analysis (stainless core tube) were taken from each of 8 single grid
cells.
**For each organic sample and each metal sample, 1 to 2 cores approximately
two to three inches deep were taken.
2-3
-------
samples. They were only analyzed for metals since the presence of oil and grease
was anticipated to interfere with the semivolatile organics analysis. Complete
lists of compounds or elements for which analyses were conducted and their
detection limits are presented in Chapter 4 (see Tables 4.2 and 4.3).
The analytical results are discussed in the following subsections. Complete
sampling data sheets are presented in Appendix A and all analytical data sheets
are presented in Appendix B.
2.1 BACKGROUND SAMPLES
Background samples were collected at a point not used for TSDF activities and
analyzed, because many compounds and elements are either naturally occuring in
the soil or may be present as a result of factors other than those which may be
attributed to Texaco's activities. The percent weight loss on drying (LOD)
measured for BGD-420 was 9-46 percent by weight. The samples were then
oven-dried at 105 C for 4 hours prior to being screened for silt content. The
silt content of the two background samples (sample identification numbers BGD-420
and BGD-421) averaged 13-7 percent by weight (see Table 2.2). The composite silt
material (sample identification number BGD-446) separated from the background
samples was sonic sieved. Material passing through a 20 urn sieve constituted the
PM n content. The PM1f, content averaged 30.82 percent by weight of the silt
material. The silt screening did not produce a sufficient amount of silt to
allow for the production of the "greater than PM10" (>PMlfJ and PM 0 fractions
for chemical analyses. "Greater than PM '' refers to the fraction of the silt
material that does not pass through the 20 urn sieve.
Results of the analyses for metals and semivolatile organics are shown in
Table 2.3- The analytical results for the metals and cyanide in the background
silt sample (sample ID BGD-455) are in terms of micrograms of the metal per
2-4
-------
TABLE 2.2.
ANALYTICAL RESULTS OF SILT SCREENING, WEIGHT LOSS ON DRYING, AND PM
FUGITIVE PARTICULATE FROM TSDF (85/12)
TEXACO, DELAWARE CITY, DELAWARE
10
SIEVING
Site and
Process
Texaco, Delaware City
Land Treatment, Cell #4
(Process L)
Texaco, Delaware City
Land Treatment Roads
(Process M)
Sample
ID
L-401-M
L-402-M
L-403-M
L-404-M
L-405-M
L-406-M
L-407-M
L-408-M
L-401-0
L-402-0
L-403-0
L-404-0
L-405-0
L-406-0
L-407-0
L-408-0
Average
Std. Dev.
M-409
M-410
M-411
Percent
Silt
11.8
6.9
8.2
3-5
6.8
11.3
5-7
6.3
9.1
5.6
8.0
3.0
7.4
7-4
4.2
8.9
7.1
2-5
8.2
19-7
10.4
Percent
Loss on
Drying
28.03
21.85
28.50
28.06
26.78
26.02
25-95
33.82
28.09
22.48
22.97
30.07
27.52
32.03
28.75
29.51
27-53
, 3-24
1.65
1.19
1.73
Sample Percent
ID
L-433
L-437
M-440
M-440
Average
Std. Dev
M-443
M-443
Average
Std. Dev
M-446
M-446
PM10
8.51
12.63
10.57
2.92
1.29
2.76
2.03
. 1.04
29.92
34.94
32.43
. 3.55
20.52
21.12
Average 20.82
Std. Dev. 0.42
(continued)
2-5
-------
TABLE 2.2. (continued)
Site and
Process
Texaco, Delaware City
Land Treatment, Cell #8
(Process N)
Texaco, Delaware City
Background Samples
Texaco, Delaware City
Land Treatment, Cell #3
(Process 0)
Texaco, Delaware City
Land Treatment, Cell #3
Repeatability and
Reproducibility
Sample
ID
N-412
N-413
N-414
N-415
N-416
N-417
N-418
N-419
Average
Std. Dev.
BGD-420
BGD-420
BGD-421
BGD-421
Average
Std. Dev.
0-422
0-423
0-424
0-425
0-426
0-427
0-428
0-429
Average
Std. Dev.
0-422rrl
0-422rr2
0-422rr3
0-422rr4
0-422rr5
Average
Std. Dev.
Percent
Silt
10.9
12.4
14.5
10.3
13.2
11.5
12.9
10.2
12.0
1-5
16.4
15.4
12.5
10.4
13-7
2.7
5.6
5-9
5-8
5-6
6.9
4.4
7-7
7-0
6.1
1.0
6.3
7.8
6.8
6.5
8.3
7-1
0.9
Percent
Loss on
Drying
21.89
9.46
30.49
28.96
32.71
29.47
30.78
27.91
29-97
1.85
Sample Percent
ID PM1Q
N-453 17.06
N-453 18.65
17.85
1.12
BGD-446 31.92
BGD-446 29.72
30.82
1.56
0-463 5.81
0-463 5.24
5-52
0.40
(continued)
2-6
-------
TABLE 2.2. (continued)
Site and
Process
Sample
ID
Percent
Silt
Percent
Loss on
Drying
Sample
ID
Percent
PMm
Texaco, Delaware City
Land Treatment, Cell #3
Repeatability and
Reproducibility
0-423rrl
0-423rr2
0-423rr3
0-423rr4
0-423rr5
7.1
8.3
4.5
6.1
11.7
31.70
29.23
35-53
28.56
32.48
Average
Std. Dev.
7-5
2.7
31.50
2.79
0-425rrl
0-425rr2
0-425rr3
0-425rr5
Average
Std. Dev.
6.7
1.9
3-7
2.7
7.8
4.6
2.6
27-75
30.41
33-21
31.49
27.98
30.17
2.33
2-7.
-------
TABLE 2.3. ANALYTICAL RESULTS FOR METALS AND SEMIVOLATILE ORGANICS
FUGITIVE PARTICULATE FROM TSDF (85/12)
Metals Analysis
Sample Identity
Eleient
AluninuB (AD
Antinony (Sb)
Arsenic (As)
Bar i us (Ba)
Beryllium (Be)
Cadaiua (Cd)
Chromium (Cr)
Cobalt (Co)
Copper (Cu)
Iron (Fe)
Lead (Pb)
Manganese (Mn)
Mercury (Hg)
Holybdenuit (Ho)
Nickel (Ni)
Osiiui (Os)
Seleniui (Se)
Silver (Ag)
Thallium (Tl)
Vanadiui (V)
Zinc (Zn)
Semi volatile Analysis
Saiple Identity
2-Hethylnapthalene
Phenanthrene
Pyrene
Land Treatnent Cell
Silt
N-448
(ug/g)
11,100
1.1
6.7
152
<1
<1
209
16.9
207
18,400
57.0
407
1.3
9.2
98.3
<1
2.3
<2
<2
200
248
Land Treatment
Silt
N-447
(ug/g)
N.D.
N.D.
N.D.
PM10
N-450
(ug/g)
13,200
0.8
6.7
215
<1
<1
255
20.6
219
21,400
81.0
473
1.3
9.5
108
<1
2.5
<2
<2
227
287
Cell IB
PH10
N-449
(ug/g)
N.D.
N.D.
N.D.
18
>PH10
N-452
(ug/g)
940
0.6
6.2
272
<1
1.9
196
14.7
200
17,600
65.0
389
1.5
5.7
94.4
U
3.1
<2
<2
190
232
Cell 14
Silt
L-430
(ug/g)
11,000
<0.5
6.6
106
<1
<1
141
17.9
164
22,400
74.0
358
0.9
9.9
86.8
<1
4.2
<2
<2
267
225
Cell 14
Silt
L-434
(ug/g)
N.D.
N.D.
10.0 J
Cell «3
Silt
0-458
(ug/g)
11,900
0.8
7.4
190
<1
<1
142
17.7
198
21,400
92.0
508
1.6
<2
150
<1
3.2
<2
<2
352
296
Cell 13
Silt
0-457
(ug/g)
45.0 J
22.0 J
9.5 J
Roads
Silt
d-439
(ug/g)
11,400
<0.5
5.9
114
<1
1.4
96.7
16.5
110
19,700
49.0
392
0.9
<2
83.8
<1
2.2
<2
<2
207
225
Roads
Silt
H-438
(ug/g)
15.0 J
36.0 J
11.0 J
Background
Silt
BD6-455
(ug/g)
14,000
2.5
5.2
59.9
<1
1.5
21.3
9.5
31.9
19,400
15.0
206
0.1
<2
13.6
-------
gram of silt sample (dry basis). These results reflect the nominal concen-
trations of these materials present in the soil which are not a result of
Texaco's activities. The results for the background samples have not been
subtracted from the results for the other samples since the risk assessments use
the inclusive value of the degree of contamination. It should be understood,
however, that the actual outside contribution to the degree of contamination of
the soil is that portion of the contaminate concentration which exceeds the
nominal background level.
For the analysis of the semivolatile organic compounds, the background
samples were extracted as low-level samples following the U.S. EPA Contract
Laboratory Program, Statement of Work for Organic Analysis, 7/85 Revision
(refered to as the CLP in this report). The extract was concentrated and
subjected to an adsorption chromatography cleanup procedure instead of the CLP
gel permeation chromatography (GPC) cleanup procedure. The adsorption
chromatography procedure was developed to remove more aliphatic compounds from
the extract than possible with the GPC procedure. The presence of excessive
amounts of aliphatic material would require dilution of the extracts prior to the
gas chromatograph/mass spectrometer (GC/MS) analysis and a corresponding increase
in the sample detection limit. The dilution would be necessary to protect the
GC/MS from the large amounts of aliphatic compounds present in some samples
(particularly land treatment samples that contain considerable amounts of oil and
grease).
For the background sample a 10-fold dilution, determined by GC/flame
ionization detection (GC/FID), was required prior to the GC/MS analysis. Two of
the CLP hazardous substance list (HSL) semivolatile compounds were found, but at
concentrations below the quantifiable detection limit. The compounds detected
met the mass spectral criteria, but the values reported are estimates only.
2-9
-------
2.2 LAND TREATMENT UNIT, CELL #4 (PROCESS L)
Cell #4 in the land treatment unit (Process L) was sampled using a coring
technique to obtain samples to a depth of approximately two to three inches,
depending upon the soil conditions in each cell sampled. A sampling grid was
laid out and eight randomly selected cells were sampled. Because of the
potential for contamination from materials used in construction of the core
sampling equipment, the samples taken for metals analysis (sample
identification numbers L-401-M through L-408-M) were sampled with a PVC coring
tube, and the samples taken for semivolatile organics analysis (sample
identification numbers L-401-0 through L-408-0) were sampled with a stainless
steel coring tube. The oil and grease content was determined on a composite
sample made of aliquots taken from all sixteen samples from Cell #4. The oil
and grease content of the composite sample was 6.11 percent expressed on a dry
basis (see Table 2.4). The weight loss on drying averaged 27-53 percent by
weight (see Table 2.2). Following oven-drying at 105 C for 6 hours, the
sixteen samples were screened for silt content which averaged 7•1 percent by
weight.
The silt fractions separated from the samples taken for organics analysis
and those taken for metals analysis were each sonic sieved for PM-o-, content
which averaged 10.57 percent by weight of the silt. The silt screening did not
produce sufficient silt material to allow the production of >PM10 and PM-.n
material by sonic sieving for chemical analyses. Aliquots of the silt
fractions (one for metals and one for organics) were taken for semivolatile
organics and metals analysis. These were analyzed separately to determine the
degree of contamination. The analytical results for metals and semivolatile
organics are shown in Table 2.3- Like the background silt sample, the land
treatment Cell #4 silt sample was extracted by the low-level method. The
2-10
-------
TABLE 2.4. SUMMARY OF OIL AND GREASE ANALYSIS
FUGITIVE PARTICUALTE FROM TSDF (85/12)
Process Process
ID Description
L Land Treatment, Cell#4
N Land Treatment, Cell #8
0 Land Treatment, Cell#3
Site
Texaco, Del aware City
Texaco, Delaware City
Texaco , De 1 aware C i ty
Oil and
Grease
6. 11%
8.46%
8.92%
QUALITY ASSURANCE SUMMARY FOR OIL AND GREASE ANALYSIS
Process
ID
Sample
Description
Site
Oil and
Grease
Total Repeatability
0 Oil&Grease 0-rrl Comp
0 Oil&Grease 0-rrl Comp
Analytical Repeatability
0 Oil&Grease 0-rrl Comp
0 Oil&Grease 0-rrl Comp
Sampling Reproducibility
0 Oil&Grease 0-rr4 Comp
0 Mean of 0-rrl Comp
Performance Audit
BGD Spiked with 34 mg of paraffin oil
Texaco,Delaware City
Texaco,Delaware City
Mean
RPD
Texaco,Delaware City
Texaco,Delaware City
Texaco,Delaware City
RPD
Expected
0.39%
RPD
Found
0.36%
6.94%
7.91%
7.43%
0.48%
7.91%
7.30%
0.30%
8.12%
7.43%
0.35%
Recovery
92.3%
2-11
-------
extract was concentrated and cleaned by the adsorption chromatography
procedure. The sample extract required a 165-fold dilution for the GC/MS
analysis. One semivolatile organic compound was detected in the sample from
Cell #4. The contamination of the compound was below the quantifiable
detection limit of 54.4 ug/g.
With the exception of the use of the adsorption chromatography cleanup
procedure, the dilution of the semivolatile organic sample extracts prior to
analysis, and the increased quantifiable detection limit, all procedures
followed the Sampling and Analysis Protocol.
2.3 LAND TREATMENT UNIT, CELL #8 (PROCESS N)
Also in the land treatment unit, Cell #8 (Process N) was sampled using a
grid layout. Eight samples were collected within this grid in a random manner
as described in Chapter 4. The scoop sampling technique was employed. The
determination of oil and grease content and LOD was conducted on sample N-412.
The oil and grease content was 8.46% expressed on a dry weight basis (see
Table 2.4). The LOD for sample N-412 was 21.89 percent (see Table 2.2). All
eight samples were oven dried at 105 C for 6 hours and desiccated overnight
prior to silt screening. Each of the eight dried samples (sample identi-
fication numbers N-412 through N-419) was screened for silt content which
averaged 12.0 percent silt by weight (see Table 2.2). The two jars of silt
(sample identification number N-453)• resulting from screening samples N-412
through N-419, were then sonic sieved for PM-ir, content which averaged 17-85
percent by weight in the silt sample. Portions of the three fractions (silt,
>PMin, and PM1n) produced from the combined silt sample from Cell #8 were
analyzed for metals. All three fractions were analyzed for metals to determine
if the degree of contamination was less or greater in the PMin fraction
2-12
-------
(particle size dependent). The results for the metals are expressed in micro-
grams of the metal per gram of sample on a dry basis. The concentrations
measured for the background sample were not subtracted from the sample results.
As a cost saving measure, only the silt and PM-.0 fractions were analyzed
for semivolatile organics to determine the particle size dependency of the
degree of contamination. Like the background silt sample, the land treatment
Cell #8 silt and PM1(-. fractions were extracted by the low-level CLP procedure.
The extracts were concentrated and subjected to the adsorption chromatography
cleanup procedure. Prior to the GC/MS analysis, the silt extract was diluted
238-fold and the PM Q extract was diluted 150-fold. None of the semivolatile
HSL compounds were detected in the silt or PM1f) extracts at the quantifiable
detection limits of 78.6 ug/g and 49-5 ug/g, respectively.
With the exception of using the adsorption chromatography cleanup procedure,
diluting the extracts prior to the GC/MS analysis, and the increased quantifia-
able detection limit, all procedures followed the Sampling and Analysis Protocol,
2.4 LAND TREATMENT UNIT, CELL #3 (PROCESS 0)
Cell #3 of the land treatment unit (Process 0) was sampled using the
scooping technique. A sampling grid was laid out and eight randomly selected
cells were sampled. The determinations of the oil and grease content and LOD
were conducted on aliquots of sample 0-429 taken from Cell #3- The oil and
grease content of the sample was 8.92 percent expressed on a dry weight basis
(see Table 2.4). The LOD for sample 0-422 was 30.49 percent (see Table 2.2).
All samples (0-422 through 0-429) were oven-dried at 105 C for 6 hours and
desiccated overnight prior to screening. The resulting dried samples were
screened for silt content which averaged 6.1 percent by weight.
2-13
-------
The silt separated from the samples (sample ID number 0-463) was sonic
sieved from PMin content which averaged 5-52 percent by weight of the silt.
Since the amount of silt composite resulting from the silt screening was low
and the PMin content of the silt was also low, the decision was made not to
produce PMlf) or >PM1(-, for chemical analyses. The results for the analyses done
on the silt fraction for metals and semivolatile organics are shown in Table
2.3. Three semivolatile organic compounds were detected in the silt samples
from the Cell #3- The silt sample was prepared for organic analysis like the
background samples. The extract was diluted 259-fold prior to the GC/MS
analysis. All of the compounds were below the quantifiable detection limit of
85.6 ug/g, which means that the reported compounds were identified, but the
magnitude of the results are only an estimate. The analytical results for the
background sample were not subtracted from the sample results.
With the exception of using the adsorption chromatography cleanup proce-
dure , diluting the extracts prior to the GC/MS analysis, and the increased
quantifiable detection limit, all procedures followed the Sampling and Analysis
Protocol.
2.5 UNPAVED ROADS IN LAND TREATMENT UNIT (PROCESS M)
Three segments of unpaved roads within the land treatment unit were
sampled. These segments were located (1) at the main gate to the unit (sample
M-409), (2) on the north-south access road at the midpoint of Cell #2 (sample
M-410), and (3) on the east road between Cells #3 and #4 (see Figure 4.1b).
Each segment was sampled using the sweeping technique.
A brush was used to sweep loose particulate from an 8-inch wide strip
across the width of each segment. Two sample jars were filled with each
sample. After drying, the samples were screened for silt content which
2-14
-------
averaged 8.2, 19-7. and 10.4 percent by weight, for sample numbers M-409.M-410,
and M-411, respectively (see Table 2.2). The weight losses on drying were
1.65, 1.19 i and 1-73 percent by weight, respectively. The silt samples
obtained from these samples were sieved for PMin content which averaged 2.03,
32.43, and 20.82 percent by weight for silt samples M-440, M-443, and M-446,
respectively. The low PMin content of road sample M-440 as compared with the
other two road samples (M-443 and M-446) correlated with the relatively low
silt content of the same sample. This was possibly due to the oil and grease
content of the road sample; however, this cannot be confirmed since the oil and
grease content of this sample was not measured. Since a sufficient quantity of
silt was not obtained from the silt screening, PMin and >PMin fractions were
not produced for the analysis of the metals and semivolatile organics. The
silt fractions from these three samples were combined.
The results of metal and semivolatile organic analyses for the composite
silt sample are presented in Table 2.3« The concentrations measured for the
background sample were not subtracted from the results for the unpaved road
silt sample. The silt sample was prepared for organic analysis like the
background sample and diluted 185-fold prior to GC/MS analysis. Three semi-
volatile organic compounds were found in the silt sample and were below the
quantifiable detection limit of 61.2 ug/g, (i.e., the mass spectral criteria
for these compounds were met for identifying the compounds, but the actual
magnitudes reported are only estimated values) .
With the exception of using the adsorption chromatography cleanup proce-
dure, diluting the extracts prior to GC/MS analysis, and the increased
quantifiable detection limit, all procedures followed the Sampling and Analysis
Protocol.
2-15
-------
2.6 REPEATABILITY AND REPRODUCIBILITY
As discussed in more detail in Chapter 4 and Appendix C, additional samples
were collected in three of the sampled grid cells in Cell #3 of the land
treatment unit (Process 0) for use in measuring the sampling and analysis
repeatability and reproducibility. The silt content values for these samples
are presented in Table 2.2. The results for the metals analysis for these
samples are presented in Table 2.5- Analyses for semivolatile organic
compounds were not conducted on these samples because the detection limits for
the samples were expected to be high due to the presence of oil and grease in
the samples. A summary report presenting the repeatability and reproducibility
results for the entire study will be completed at the end of the study.
The repeatability and reproducibility samples were also used for the oil
and grease analysis. For total repeatability, the composite sample (sample ID
0-rrl) made of aliquots from 0-422rrl, 0-423rrl, and 0-425rrl was analyzed in
duplicate (see Table 2.4). For analytical repeatability (defined here as the
gravimetric determination of the extracted residue), the gravimetric determina-
tion was conducted twice on the sample extract. Sampling reproducibility was
determined by comparing the oil and grease content of a composite sample
(sample ID 0-rr4) made of aliquots of 0-422rr4, 0-423rr4 and 0-425rr4 to the
oil and grease content of the 0-rrl composite. The rr4 samples were taken from
the same cells by a secondary sampler.
A performance audit was also conducted by spiking a background sample (from
another site) with paraffin oil dissolved in Freon TF. The spiked sample was
carried through the oil and grease analysis procedure. The recovery of the
paraffin oil was calculated to be 9^ percent (see Table 2.4).
2-16
-------
TABLE 2.5. ANALYTICAL RESULTS FOR REPEATABILITY AND REPRODUCIBILITY METALS SAMPLES
FUGITIVE PARTICUALTE FROM TSDF (85/12)
Sample Identity
Eleaent
AluBinuffl (AD
Arsenic (As)
Bariua (Ba)
Berylliua (Be)
Cadmiua (Cd)
Chromium (Cr)
Cobalt (Co)
Copper (Cu)
Iron (Fe)
Lead (Pb)
Manganese (Mn)
Hercury (Hg)
Nickel (Ni)
Selenium (Se)
Silver (Ag)
Vanadium (V)
Zinc (Zn)
Sample Identity
Element
Aluminum (AD
Arsenic (As)
Barium (Ba)
Beryllium (Be)
Bismuth (Bi)
Cadmium (Cd)
Chromium (Cr)
Cobalt (Co)
Copper (Cu)
Iron (Fe)
Lead (Pb)
Manganese (Hn)
Hercury (Hg)
Nickel (Ni)
Selenium (Se)
Silver (Ag)
Vanadium (V)
Zinc (Zn)
6rid No. 2
RTI PEI
0-465 0-467
(ug/g) (ug/g)
12,400 7,997
6.9 10.9
109 138
<1 <0.1
<1 68.3
137 181
14.5 20.2
174 346
21,000 32,500
69.0 2,620
425 2,100
l.l 1.2
141 34.4
2.5 2.2
<2 8.3
483 21.0
233 27,280
RTI
0-471
(ug/g)
13,700
7.7
121
_nni
percent
Recovery
-
72. n
12.11
101. 5Z
-
98.57.
-
-
129. OX
-
1 12.5Z
179. 4Z
-
-
72. 81/.
73. IX
-
-
,500 10,100
6.5 6.4
113 678
<1 <1
<1 <1
116 113
15.0 15.2
168 153
,600 18,300
60.5 66.0
400 401
0.8 0.9
489 485
2.7 2.9
<2 <2
,820 2,810
209 207
PEI RTI
0-483 0-485
(ug/g) (ug/g)
7,630 13,600
5.9 7.9
104 146
0.5 <1
0.8 <1
85.6 145
14.4 18.6
134 199
14,600 22,900
50.0 81.3
330 664
1.0 1.2
403 217
0.6 3.6
<0.7 <2
2,160 440
179 330
Unspiked Sample
0-458
(ug/g)
11,900
7.4
190
(1
<10
<1
142
17.7
198
21,400
92.0
508
1.6
150
3.2
<10
352
296
Oy IRC
Amount
(ug/g)
0
89.4
89.4
89.4
0
89.4
0
0
89.4
0
89.4
89.4
0
0
89.4
89.4
89.4
0
PEI
0-487
(ug/g)
10,050
8.6
155
<0.1
1.5
104
21.1
221
19,010
71.2
840
2.3
269
<0.3
<0.7
392
419
Found
0-497
6,577
74
211
79
0
80
95
13
224
13,069
126
422
0
96
57
60
199
294
Grid No. 16
RTI RTI PEI
0-491 0-492 0-493
(ug/g) (ug/g) (ug/g)
12,500 12,300 10
8.4 7.6
145 228
<1 <1
(1 <1
149 147
21.1 20.7
212 204
22,700 22,500 11
86.4 81.0
671 659
1.3 1.4
211 207
2.9 3.8
<2 <2
414 408
341 332
PEI
Percent
Recovery
-
74. 5X
23. 6X
88. 6Z
-
89. n
-
-
28. n
-
37.7Z
-96. OZ
-
-
60. 71
66. 81
-
-
,970
5.7
70
<0.1
1.1
114
15.0
178
,600
48.4
591
1.7
176
<0.3
<0.7
284
309
-------
2.7 CONCLUSIONS
No major problems were encountered during sample collection. It was felt
that the sampling program was successful in obtaining representative samples.
In the analyses of the samples, no problems were encountered in obtaining
silt content or determining PM1f) content. The results of the metals analyses
are also believed to be accurate.
The only significant problem encountered during the analyses was the fact
that the samples contained a significant amount of organics (oil and grease)
not found on the Hazardous Substances List. This prevented the semivolatile
organics analyses from being conducted at the level described in the analytical
protocol. Because of the high concentrations of organics, an alternative
sample cleanup procedure was used on the samples to remove these organics. The
cleanup procedure used on the semivolatile organic sample extracts appeared to
have little effect on the samples from this site. However, the detection
limits for these samples were believed to be the lowest levels practical for
the analysis of HSL semivolatile compounds by GC/MS.
2-18
-------
3.0 PROCESS DESCRIPTION
At this facility, sampling was undertaken for four processes, where
the term "process" refers to a likely source of potentially contaminated
fugitive particulate emissions. All of the process samples were taken
within the facility's land treatment unit; processes sampled include:
a. Three different cells within the land treatment unit; and
b. Unpaved road segments at three locations within the land treat-
ment unit.
The following process descriptions are based largely upon the information
provided by the facility and observations made during the course of the
survey/sampling effort.
3.1 LAND TREATMENT CELLS
The actual working surface of the land treatment unit is approximately
34 acres, and is configured into 12 discrete cells (see facility map, Fig-
ure 4.2). At the time of survey, 11 of the 12 cells were being used on a
strict, rotation basis. Average cell size is 2.85 acres; the cells range in
size from 2.36 to 3.40 acres. The unit has been in operation for about
3 years.
The land treatment unit is used exclusively to dispose of wastes
generated at the facility. Principal wastes and approximate quantities
treated during the past year are shown below.
EPA Hazardous
Waste No.
K050 and K051
K048 and K049
Quantity
(tons)
2,000
1,850
Non-RCRA waste 8,000-20,000
Description
API and CPI oil-water gravity separator
bottoms
Vacuum filter cake (VFC) from oil recovery
unit at facility wastewater treatment
plant (WWTP)
Dual cell gravity solids (DCG)--biosludge
from WWTP
In addition to these waste streams, the unit occasionally receives tank
bottoms associated with storage tank cleanup, as well as oil-contaminated soil
3-1
-------
The principal equipment types, functions, and approximate level of
activity for the treatment unit are summarized below.
Equipment (commercial
designation if available)
Farm tractor (Ford 7700)
with implements:
1. Chisel plow
2. Disc harrow
3. Lime bucket
4. Backdrag
Bulldozer (John Deere 350c)
Two vacuum
14 wheels
trucks--4 axles,
Dump truck--3 axles,
10 wheels
Function
Chisel plow and disc harrow
used for initial incorpora-
tion and subsequent soil
cultivation, respectively.
Lime bucket used to add
lime to soil for pH con-
trol. Backdrag used to
smooth surface prior to
waste application.
Activity Units
Chisel plow used minimum 4 to
6 times/yr. Disc harrow used
minimum of 8 to 12 times/yr.
Lime bucket on as-needed
basis. Backdrag minimum 4 to
6 times/yr.
Functions include spreading Based on survey observations
of solid material on treat- activity—4 to 6 hr/day.
ment plots, and pulling of
backdrag over surface.
Delivery of high liquid
content wastes to treat-
ment plots.
Delivery of solid wastes
to treatment plots
3,000-gal. capacity; avg.
3 loads/day, when operating,
Plant considers "full load"
~ 14,500 Ib; 2 loads/day.
The basic sequence of operations for the unit are as follows:
a. Waste application;
b. Initial incorporation (chisel plow);
c. Addition of lime as needed (lime bucket);
d. Soil cultivation (disc harrow);
e. Smooth surface (backdrag); and
f. Repeat steps d and e.
separator sludges)
is the principal
is maintained
High liquid content wastes (predominantly API and CPI
are applied from a vacuum tank with a 4-in. main. Gravity
mechanism for spreading; to facilitate spreading tillage
perpendicular to the surface contour. To prevent undesirable runoff into
the drainage swale (see facility map), (along) contour plowing is used at
the extreme lower end of each treatment cell. Wastes with higher solids
content are delivered by dump truck. Waste spreading is accomplished by a
bulldozer fitted with a blade. Observations during the survey suggest that
3-2
-------
the solids are spread quite uniformly over the cell surface to a depth of
2 to 3 in.; gravity spreading of the higher liquid content waste produces
considerably less uniform conditions. In the case of the liquid wastes it
is quite possible that areas closest to the point of application experience
far heavier loadings than areas at the opposite end of the treatment cell
(i.e., closest to the drainage swale).
The targeted application rate for the treatment unit is set at 80 bar-
rels of oil per acre per application. This loading allows reapplications
on the order of 4 to 6 times per year. The treatment unit foreman indicated
that application frequency shows some sensitivity to seasonal climate varia-
tions. For example, during the winter months, the trafficability of the
surface typically decreases making it more difficult to incorporate wastes.
As a result, frequency of reapplication may decrease during these periods.
The facility uses lime to immobilize the metals.
This practice creates a favorable environment for precipitation of metals
in the zone of incorporation (nominal 8- to 12-in. depth). As a "typical"
value, the cells are limed once per week.
As noted earlier, sampling was conducted for three different cells
within the treatment unit. The samples represent three different points in
time after application. Cell 3 samples were taken immediately (< 4 hr)
after initial incorporation of the waste streams (combination of VFC and
DCG—biosludges). Cell 4 samples were taken approximately 1 week after ap-
plication of waste material (predominantly DCG). Initial incorporation as
well as subsequent cultivation had already been completed. 'Cell 8 samples
represent conditions approximately 40 days after application.
3.2 UNPAVED ROADWAYS
In addition to samples from the actual treatment surface, samples were
also collected from the unpaved roadways located within the treatment unit.
The treatment unit is effectively isolated from the rest of the facility,
and as a result traffic on the roadways is restricted to activity directly
associated with the land treatment operation. Estimated traffic volume on
the roadways is < 25 vehicle passes per day.
3-3
-------
4.0 SAMPLING AND ANALYSIS
This chapter outlines the procedures used for (1) the sampling conducted at
Texaco Refining and Marketing, Inc. (Texaco) and (2) the analysis of the
samples collected. Included are descriptions of the location of each process
sampled and the sampling grid for sample collection. Sample handling,
preparation, and/or analysis specific to this facility or any process therein
are described in detail. Any deviations from the standard sampling and
analysis procedures (see Appendix C) are discussed.
Four processes were sampled at Texaco: (1) the land treatment unit,
Cell #4; (2) the land treatment unit, Cell #8; (3) the land treatment unit,
Cell #3; and (4) three segments of unpaved roads within the land treatment
unit. The samples from each of these processes were analyzed for weight loss
on drying (LOD), silt and PMin content, metals, and semivolatile organics. The
samples from the land treatment cells were also analyzed for oil and grease
content. A tabular presentation of the sampling plan for Texaco which
specifies the number and types of samples and the locations at which they were
collected can be found in Chapter 2 (see Table 2.1). The subsections that
follow further describe the sampling locations, sampling grid schemes, and
applicable sampling and analytical procedures.
4.1 SITE PLOT PLAN
Figures 4.la and 4.1b show the site plot plans for the landfill and land
treatment facilities. The scale of Figure 4.la is approximately 1 inch equals
250 feet, the scale of 4.1b is approximately 1 inch equals 230 feet. The
4-1
-------
locations of the land treatment cells and unpaved roads sampled are shown on
these site plot plans. Pertinent topographical features, both natural and
man-made, are also shown.
4.2 LAND TREATMENT UNIT, CELL #4 (PROCESS L)
Cell #4 (Process L) is in the center on the western side of the land
treatment unit at the Texaco facility (see Figure 4.1b). The shape of Cell #4
approximated a rectangle with dimensions of 224, 430, 220, and 434 feet.
Based on these dimensions, MRI designated that the sampling grid for
Process 0 be a 200 x 400 foot rectangle centered within the cell having 50-foot
square grid cells. The sampling grid was then laid out using surveyor's stakes
and tape. The grid cells were numbered starting in the northeast corner of
Cell #4 as shown in Figure 4.2.
MRI directed that eight grid cells be sampled; a random number table was
used to select the specific grid cells for sampling (see Appendix C). One of
the grid cells selected (#1) was rejected by MRI because it was on two of the
process boundaries and adjacent to another cell already selected. Grid cell #2
was substituted in its place.
MRI determined that for the sample collection, the coring technique would
be used at this process. Within each cell, a sampling template was randomly
tossed four times. . The cored sample aliquots were taken from inside the areas
defined by the template. The application of the basic coring technique (see
Appendix C) proved to be difficult and a modified coring technique was devised
based on discussions between MRI and Entropy personnel. The modified technique
involved taking one or two 2- to 3~inch cored aliquots from each of the four
template areas using each type of core tube (stainless steel or plastic).
Because of the potential for sample contamination by the coring tube
4-2
-------
I/>^T:">
M/jssr j
1 < I.?,'"-—,'?v ^'
ETHYL CORPORATION !
(FORMERLY OIAUOND SHAMROCK CORPORATION)!
GETTY REFINLNO AND MARKETING COMPANY
\ UW1 W 01« (Ml .SM)-
.
'
(Unul.? Hail •"
J_
'-Ttr.:
3n —!
Tlf
^LTt4
1^
Vl
^•^i'/.W/^
*s
L
/
y
'/
/ ' ^li,\v <«i"/' " P \x -\^N \\ \—i— /
/ / AR£A9\ \2 \ \ l,'/•'//' / U t\\ \ \\ \\\\ \ ' AREA; ^
/ I *Vl3fcl|/l|J/J'|l\V^ \ V\*\\V\ \ I
s;^iiSii%Sil&A
/
--AA
T~~~\~^T7~r~r)rr'y-rr'? j \ ri--y
\ } i */*''SM/'9' \/TJ
} J ! !'/////I///' /, / y
;v/"/r/ /"//// ,//"/ /
/
/ X
Li--/z.^
•2 •
iiz
*'
;;///////
/jjjL/j;/
7T-\
^V^/v
\J-2^
* i ii i
1
i »
"? ^
i»
rrr
Trf
= :i
•f
MI.UJ.IO
I. . JS.7*S
i| .^'wJ1!" :t
lf\...-^f yima ' \
y^- .i If"* ronai ,'• -
£-v-V-:-v - -.
DREDGED MATERIAL
STORAGE AREA NO.1
l DISPKN. «» .2*-—^^
v\ •
\\
I'JWSALMKE* *4 '
OlSfOSAL ABU »4 \
DISKJSti AJtU •*
S~}.~-' •"«-•»•!
,'>- X-.VA
IBU ;» V\
i STCUflU / . ,\
\\ \ - ;""« \\ '
%: ;'"( .
• • ^'\ ^ \
>« v\ • \
> • \-x, ; i
^.•\V:-\v
•3- ^
^f^K-J
,'/•: /I '01
i • ; .''41 .
\:rS,
MIL
trocwiU
\ "
t> x
^
I !» SLIOG£ Ml
!'^
'.ft
% m
/
.&
^-13
on:»rrt •cMu^taT
l^>
V\X
//i
%L
JlJ
i
A-
S
Tf*
s
V \V
j._'—^.' -_.
,-''/ » )TA
-V(^
-*;7j
ir///"
r
s-ij——
^^Vx
LAND TREATMENT FACIUTY !
ra ('
:\\V\ \
l>
\\
^
• SBC.W1.70 .
f ^.J.flii >0
ILEH. • }i.«40
— 18' P(C FOaCtfAIl
TO E*< hVTT •
Ukvtl BCU
FACILITY
^'in
,'( ! LAKH -i tin .: K;-.
011 TH£»t«£JIT . • • I -
>..<*_
V/\ M0,')7»
f*^-'-:-i
i .^:"--l I
- - ' KEY MAP
LAND TREATMENT AND LANDFILL FACILITIES
l. XU UIUTitm UTEI 70 TH£ lEHTICAl OlSTAftd !» fEET ICAUnED FKH
is KU suuttt.
L «U OOCIMUS IK «TT tt«T» TO THE STXTf OF DCk^Kf OXTOIuns ]
sjnm x:« »nfo»ii
«. C8XA3 SW4U CanTX
9iUb UATCt HM17DBIIC •
'iniKTca
gum iota**? •
EUTH ::a
GKJW StlfXU CQMTDUX I
.*'
.->•*•
- .. - -—" * •"' Ln*M< tii»» rtflt^
\^l:i.-.--w«»i~««—•
.'.**: --".".
Figure 4.la. Site plot plan of Texaco
in Delaware City, DE.
& Well • Water Data Table
CUEVTOPrf
CASING
EL£V. «l HjO.
LEVEL
9.14
« 4Z
T.4T
0 100 20)
suu ii ?tn
A
A
/s
A
A
A
A
AI
tSoIO »O« r iST
UNIT 10 - • . SITE
LAND TREATMENT AND
LANDFILL FACILITIES
ICAU ^
AS NOTED
1201-110-CE-91
4-3
4-4
-------
LAND TREATMENT FACBJTY
Figure 4.1b. Enlargement of Site Plot Plan Showing Locations of Land Treatment
Cells and Sampling Locations for Background and Unpaved Road Samples,
4-5
-------
SO-
SO1
224
^ fe.
i
434
1
1
©
L-402
9
13
©
L-405
©
L-4^6
L-308
29
L-401
6
10
14
18
22
26
30
3
7
(77)
L-403
(is)
L-404
19
23
27
31
4
8
12
16
20
L-407
28
32
*_
* 220'
TYPICAL
CELL
j
4301
1
J
N^ ,.,
SCALE: 0.5" - SO1
FIGURE 4.2. SAMPLING GRID. PROCESS DIMENSIONS. AND SAMPLE NUMBERS FOR LAND
TREATMENT UNIT. CELL *4 AT TEXACO. DELAWARE CITY (PROCESS L).
-------
materials of construction, two samples were taken from each grid cell, one for
metals analysis (using the plastic coring tube) and one for organics analysis
(using the metal coring tube). The sixteen samples taken from the eight grid
cells sampled in this process were numbered using the following scheme:
L-401-0, L-401-M, L-402-0, M-402-M L-408-0, L-408-M.
Oil and grease analysis was performed on the land treatment, Cell #4 soil
samples. An aliquot was taken from each sample and mixed to form a soil sample
composite. The oil and grease analysis was performed following Method 503 D,
Extraction Method for Sludge Samples, found in Standard Methods for Examination
of Water and Waste Water, 1985, l6th Ed. The method involved acidifying the
sample with concentrated HC1 followed by drying the acidified sample by mixing
the sample with an equal weight of anhydrous magnesium sulfate. The soil
mixture was extracted with Freon TF, 1,1,2-Trichloro- 1,2,2-trifluoroethane, in
a Soxhlet extraction apparatus. The weight of the extracted residue was
determined after distilling the solvent and drying the residue. The loss-on
drying (LOD) was determined on the soil composites to calculate the oil and
grease as a percent of dry solids. The following formula was used for the
calculation:
„, n,, , _ Weight of Residue x 100 %
% Oil and Grease =
Corrected Dry Weight of Soil Sample
The quality assurance (QA) for the oil and grease analysis used the repeat-
ability and reproducibility (R&R) samples. A soil composite was made from three
R&R samples collected by the primary sampler from the three R&R cells sampled at
the Texaco site. Duplicate aliquots of the primary soil composite were separately
analyzed for oil and grease. By comparison of the results from the duplicate oil
4-7
-------
and grease analysis, the degree of repeatability for the total system was
determined.
A second soil composite was made from three of the R&R samples collected by
the secondary sampler from the three R&R cells at the Texaco site. An aliquot of
the secondary soil composite was analyzed for oil and grease. By comparing the
average oil and grease value for the primary soil composite with the oil and
grease value for the secondary soil composite, a measure of the sampling
reproducibility was made.
A QA spike was done using the background sample from another site. A USEPA
quality control sample for oil and grease analysis consisting of paraffin, oil
dissolved in Freon TF was used. The spike provided a concentration of 0.34$ of
oil and grease to the background sample. The spiked sample was then extracted and
analyzed for oil and grease. The percent recovery of the spiked oil and grease
mixture was calculated by the following formula:
Percent Recovery = Spiked Sample % - Background Sample %
Calculated Spiked Sample %
Portions of each of the samples from this process were first analyzed for
weight loss on drying (LOD) by drying for 12 to 16 hours in a 105 C oven. All
samples were then oven dried at 105°C for 6 hours (see Table 4.1). Following
drying, each sample was screened to determine percent silt content and then sonic
sieved to determine percent PMin content (see Appendix C for a complete
explanation of sample handling during these analyses). Material passing through
the 20 urn sonic sieve constituted the PMin fraction. The portion of the silt
fraction that did not pass through this sieve was referred to as the "greater than
PM " (>PM ) fraction. With the low yield of silt from the eight samples and
4-8
-------
TABLE 4.1. SAMPLE DRYING PROCEDURE SUMMARY
Sample Process
ID Description Drying Procedure
L Land Treatment, Cell #4 Oven Dried at 105°C for 6 hours
M Land Treatment Roads Desiccated for 24 hours
N Land Treatment, Cell #8 Oven Dried at 105°C for 6 hours
0 Land Treatment, Cell #3 Oven Dried at 105°C for 6 hours
O-R&R Land Treatment, Cell #3 Oven Dried at 105°C for 6 hours
BCD Background Sample Oven Dried at 105°C for 4 hours
the low PMin content of the silt, the decision was made not to produce PMin and
fractions for the chemical analyses.
The set of samples collected for metals analysis and the set of samples
collected for organics analysis were each utilized to make separate composite
samples of the silt. Portions of the silt composite for metals analysis were
sent to RTI for the analysis . The procedures used for analysis of the metals
followed the methods outlined in the EPA publication, "Testing Methods for
Evaluating Solid Waste," SW-846. The metals measured and the detection limits
for the analytical methods used are shown in Table 4.2. Samples for analysis
of all metals except mercury (Hg) were prepared by acid digestion using EPA
Method 3050 (SW-846) . Mercury samples were prepared and analyzed by the cold
vapor atomic absorption procedure following EPA Method 7471 (SW-846) . Two
modifications were used in the final dilutions,.,of . the digestates. The samples
for inductively-coupled argon plasmography (ICAP) determination by EPA Method
6010 and furnace atomic absorption determination of antimony (Sb) by EPA Method
704l were diluted to achieve a final concentration of 5# HC1. The sample
digestates for arsenic (As) determination by EPA Method 7060, for selenium (Se)
determination by EPA Method 7740, and for thallium (Tl) determination were
diluted to achieve a final concentration of 0.5% nitric acid.
4-9
-------
TABLE 4.2. METALS, MEASUREMENT METHODS, AND DETECTION LIMITS*
Element
Aluminum (Al)
Antimony (Sb)
Arsenic** (As)
Barium** (Ba)
Beryllium (Be)
Cadmium** (Cd)
Chromium** (Cr)
Cobalt (Co)
Copper (Cu)
Iron (Fe)
Lead** (Pb)
Manganese (Mn)
Mercury** (Hg)
Molybdenum (Mo)
Nickel (Ni)
Osmium (Os)
Selenium** (Se)
Silver** (Ag)
Thallium (Tl)
Vanadium (V)
Zinc (Zn)
Detection Limits (ug/g)*
ICAP*** GFAA*** Cold Vapor .
75-0
0.05
01
. X
0.3
0.1
0.1
0.3
0.5
2.0
75-0
0.3
O.l
Om
. Ul
0.2
1 9 _____
1 . £. — — — — —
01 _____
. J- — — — __
0.05
0.2
0.2
n R
U . O — — — — —
0.1
Detection limits were calculated as three times the standard deviation of
the values measured for compounds at or near the suspected detection limit
in the background sample. For compounds not detected in the background
sample, the detection limits were calculated as three times the standard
deviation of the background noise. Fe, Mg, and Al detection limits were
determined using low level standards as three times the standard deviation
of the values measured.
**
Eight RCRA metals
*•**
ICAP = Inductively-Coupled Argon Plasmography
GFAA = Graphite Furnace Atomic Absorption
AA = Atomic Absorption
4-10
-------
A portion of the silt composite for organics analysis was sent to PEI for
the analysis for compounds listed in Table 4.3- The sample was prepared by
sonication extraction (EPA Method 3550, SW-846) using the procedure specified
in the EPA Contract Laboratory Program (CLP), Statement of Work for Organic
Analysis, 7/85 Revision. The extract was prepared at the low concentration
level and screened by gas chromatography with a flame ionization detector
(GC/FID). It was found to contain over 20 ug/g of organic compounds. The
extract was then transferred to Triangle Laboratories for cleanup by adsorption
chromatography. The extract was concentrated and a 200 rag portion was
removed. The 200 mg portion was redissolved in methanol/methylene chloride
(1:1) and chromatographed on Sephadex LH-20. The cleanup procedure used only
6.2% of the original sample which represents a 16-fold dilution.
The cleaned extract was returned to PEI and screened again by GC/FID.
Based on the results of the screening, the sample was diluted another 10.3 fold
to protect the gas chromatograph/mass spectrometer (GC/MS). The cumulative
dilution of 165-fold raised the sample's quantifiable detection limit to
54.4 ug/g.
4.3 LAND TREATMENT UNIT, CELL #8 (PROCESS N)
Cell #8 (Process N) was located toward the southwest corner of the land
treatment unit at Texaco (see Figure 4.1b). The shape of the process
approximated a trapezoid with two right angles; the dimensions were 490, 254,
430, and 252 feet (see Figure 4.3).
Based on these dimensions, MRI directed that the sampling grid be laid out
toward the center of the process as a 200 x 400 foot rectangle with a typical
grid cell being a 50 foot square. The grid cells were numbered from left to
right starting in the northeast corner of the sampling grid (see Figure 4.3).
4-11
-------
TABLE 4.3. SEMIVOLATILE ORGANIC COMPOUNDS FOR ANALYSIS
ACENAPHTHENE
ACENAPHTHYLENE
ANTHRACENE
BENZO (a) ANTHRACENE
BENZOIC ACID
BENZO (a) PYRENE
BENZO (ghi) PERYLENE
BENZO (b) FLUORANTHENE
BENZO (k) FLUORANTHENE
BENZYL ALCOHOL
BIS (2-CHLOROETHOXY) METHANE
BIS (2-CHLOROETHYL) ETHER
BIS (2-CHLOROISOPROPYL) ETHER
BIS (2-ETHYHEXYL) PHTHALATE
4-BROMOPHENYL PHENYL ETHER
BUTYL BENZYL PHTHALATE
4-CHLOROANILINE
4-CHLORO-3-METHYLPHENOL
2-CHLORONAPHTHALENE
2-CHLOROPHENOL
4-CHLOROPHENYL PHENYL ETHER
CHRYSENE
DIBENZO (a,h) ANTHRACENE
DIBENZOFURAN
1,2 DICHLOROBENZENE
1,3 DICHLOROBENZENE
1,4 DICHLOROBENZENE
3,3'-DICHLOROBENZIDINE
2,4-DICHLOROPHENOL
DIETHYLPHTHALATE
2,4-DIMETHYLPHENOL
DIMETHYL PHTHALATE
DI-N-BUTYLPHTHALATE
2,4-DINITROPHENOL
2,4-DINITROTOLUENE
2,6-DINITROTOLUENE
DI-N-OCTYL PHTHALATE
FLUORANTHENE
FLUORENE
HEXACHLOROBENZENE
HEXACHLOROBUTADIENE
HEXACHLOROCYCLOPENTADIENE .
HEXACHLOROETHANE
INDENO(l,2,3-cd) PYRENE
ISOPHORONE
2-METHYL-4,6-DINITROPHENOL
2-METHYLNAPHTHALENE
2-METHYLPHENOL
4-METHYLPHENOL
NAPHTHALENE
(Continued)
4-12
-------
TABLE 4.3. (continued)
4-NITROANILINE
NITROBENZENE
2-NITROANILINE
3-NITROANILINE
2-NITROPHENOL
4-NITROPHENOL
N-NITROSO-DI-N-PROPYLAMINE
N-NITROSODIPHENYLAMINE
PENTACHLOROPHENOL
PHENANTHRENE
PHENOL
PYRENE
1,2,4-TRICHLOROBENZENE
2,4,5-TRICHLOROPHENOL
2,4,6-TRICHLOROPHENOL
MRI determined that eight grid cells would be sampled. A random number
table was used to select the grid cells for sampling (Appendix C). At MRI's
direction, grid cell #17 was substituted for grid cell #28, because #28 was
adjacent to two cells which had already been selected.
Because this process involved a disturbed surface, MRI decided that it
would be sampled using the scooping technique (see Appendix C). As for
Process L, the sampling template was randomly tossed four times within each
cell sampled. The sample from each cell consisted of four soil aliquots taken
inside the areas defined by the template. The eight samples taken were
numbered N-412 through N-419- Figure 4.3 shows the grid layout and the cell
from which each sample was taken.
An aliquot of sample number N-412 from this process was analyzed for oil
and grease content by the procedure described for Process L. A portion of
sample N-412 was also analyzed for weight loss on drying (LOD) by drying for 12
to 16 hours in an oven at 105 C. Later, all samples were dried in an oven at
105 C for 6 hours. Following drying, the samples were analyzed for
4-13
-------
50'
50'
TYPICAL
CELL
490'
254'
1
1
5
9
13
©
N-416
21
25
29
2
©
N-412
10
H
18
22
26
®
N-418
3
7
(TT)
N-413
15
19
23
(2?)
N-417
31
4
8
(l2)
N-414
(Te)
N-415
20
24
28
(32)
N-419
252'
430'
,
N
SCALE: 0.5" = 50'
FIGURE 4.3. SAMPLING GRID, PRXESS DIMENSIONS, AND SAMPLE NUMBERS FOR LAND
TREATMENT UNIT, CELL "8 AT TEXACO, DELAWARE CITY (PROCESS N).
-------
percent silt content and percent PM..,-. content (see Appendix C for specifics of
sample handling during each of these analyses).
Portions of the silt, PMin> and >PMin fractions from this process were
submitted to RTI for metals analysis and portions of the silt and PM10 fraction
only, were submitted to PEI for analysis of semivolatile organics. As a cost
saving measure, the >PM1f, fraction was not analyzed for semivolatile organics
since the particle size dependency of the degree of contamination will
be determined using only the concentration values for silt and PM-,,-, fractions.
All fractions were analyzed for metals and semivolatile organic compounds as
described previously for the samples from Process L. As for the Process L sample
extracts, the Process N extracts required dilution prior to GC/MS analysis which
resulted in the higher quantifiable detection limits of ?8.6 ug/g for the silt
extract (diluted 238-fold) and 49.5 ug/g for the PM _ extract (diluted 150-fold).
4.4 LAND TREATMENT UNIT, CELL #3 (PROCESS 0)
Cell #3. designated Process 0, is located next to Cell #4 in the center of
the east side of the land treatment unit at Texaco (see Figure 4.1b). The
process boundaries approximated a trapezoid with sides of 364, 224, 410, and
218 feet. Based on the process shape and dimensions, the sampling grid was
designated and laid out to occupy a rectangular area 200 by 360 feet; the
typical grid cell was a 40-foot square (see Figure 4.4). The grid cells were
numbered from left to right starting in the northeast corner of the sampling grid.
MRI determined that eight grid cells would be sampled. A random number
table was used to select the grid cells for sampling (Appendix C) and no
selected cells were eliminated.
4-15
-------
224'
40'
40'
TYPICAL
CELL
364'
1 — — '
6
It
©
0-425
21
26
31
36
41
k. , . ..
@
0-422
©
0-423
12
17
22
27
32
37
42
3
8
13
18
23
28
33
©
0-428
©
0-429
4
®
0-424
14
19
@
0-426
29
34
39
44
. 3
10
15
20
(*V
0-427
30
35
40
43
1
MATERIAL
410'
218'
SCALE: 0.0125"" 1'
FIOURE 4.4. SAMPLING GRID, PROCESS DIMENSIONS, AND SAMPLE NUMBERS FOR LAND
TREATMENT UNIT, CELL »3 AT TEXACO, DELAWARE CITY (PROCESS 0).
-------
MRI determined that for the sample collection, the scooping technique would
be used at this process (see Appendix C). Within each cell, the sampling
template was randomly tossed four times. As for Process N, the sample from each
cell consisted of the four soil aliquots (two scoops each) taken from inside the
areas defined by the template. The eight samples were numbered 0-422 through
0-429- Figure 4.4 shows each sample and the corresponding grid cell from which
it was taken.
An aliquot of sample 0-429 was taken for analysis of oil and grease content
by the procedure described previously for Process L samples. Sample 0-429 was
also analyzed for weight loss on drying (LOD) by drying for 12 to 16 hours in a
105°C oven. Later, all the samples from this process were oven-dried at 105°C
for 6 hours. They were then analyzed for percent silt content and percent PM1f.
content (see Appendix C for a complete explanation of sample handling during
these analyses).
Using the screening and sieving techniques described in Appendix C, all the
samples from this process were used to make a composite sample of the silt. As
for Process L, the decision was made not to produce PM..... and >PMin material for
chemical analysis. Portions of the composite samples of the silt were sent to
RTI and PEI for metals and semivolatile organics analysis, respectively. All
samples were analyzed for metals and semivolatile organic compounds as described
previously for the samples from Process L. As for the Process L sample
extracts, the Process 0 extracts required dilution prior to the GC/MS analysis
which resulted in the higher quantifiable detection limit of 85.6 ug/g after the
260-fold dilution.
4-1?
-------
4.5 UNPAVED ROADS IN LAND TREATMENT UNIT (PROCESS M)
Three separate segments of unpaved roads (Process M) within the land
treatment unit were sampled. As may be seen in Figure 4.1b, the areas sampled
were located:
(1) at the main gate to the land treatment unit (sample number M-409),
(2) on the north-south access road across from the midpoint of Cell #2
(sample number M-410), and
(3) on the east road of the unit between Cells #3 and #4 (sample number
M-411).
Each of the samples taken (see sample numbers above) was from a rectangular
area and spanned the road (centered) for 18 feet and was 8 inches wide (see
Figure 4.5).
Because unpaved roads consist of hard-crusted, undisturbed surfaces, MRI
recommended sampling this process using the sweeping technique. A disposable
brush was used to sweep the loose particulate from the surface of each road
area into a disposable scoop, which was then used to deposit the particulate
into the appropriate sample jars.
Ten-gram portions of the sample from each road segment were first analyzed
for weight loss on drying (LOD) by drying for 12 to 16 hours in a 105°C oven
and then all the samples were dried for 1 day in a desiccator. They were
analyzed for percent silt content and percent PM1(-. content (see Appendix C) .
Since a sufficient quantity of silt could not be obtained from the silt
screening, PM1f, and >pMin material was not produced for chemical analysis.
Portions of the silt fraction of the samples were submitted to RTI and PEI
for metals and semivolatile organics analysis, respectively. They were
analyzed for metals and semivolatile organic compounds as described previously
for the samples from Process L. As for the Process L sample extracts, the
4-18
-------
M-409
18'
M-410
8" '
18'
M-411
18'
FIGURE 4.5. DIMENSIONS AND SAMPLE NUMBERS FOR THE SEGMENTS OF UNPAVED
ROADS SAMPLED IN THE LAND TREATMENT UNIT AT TEXACO (PROCESS M).
-------
Process M extracts required dilution that resulted in the higher quantifiable
detection limit of 62.5 ug/g after a 185-fold dilution.
4.6 REPEATABILITY, REPRODUCABILITY, AND PERFORMANCE AUDIT SAMPLES
As part of the sampling conducted within the Process 0 boundries at Texaco,
samples were taken for measurement of repeatability (within laboratory
precision) and reproducibility (between laboratory precision) and for
performance audits. Three of the grid cells (numbers 2, 7, and 16) previously
sampled were sampled for these purposes.
Within each of these cells, the primary sampler (in this case, Mr. Bernie
von Lehmden) took three samples (only two needed) and the secondary sampler (Mr.
Steve Plaisance) took two samples (only one needed), all from the same template
#
area. Samples taken by the primary sampler were used to measure both total and
analytical repeatability and analytical reproducibility. They were also spiked
for the performance audits (see Chapter 5 and Appendix C). Samples taken by the
secondary sampler were used to measure total reproducibility (see Appendix C).
Sampling was conducted using the scooping technique.
Weight loss on drying determinations, drying, and silt content determinations
for these samples were done as described for the samples from Process L. The
analyses for metals were done using the same methods previously discussed.
Research Triangle Institute (RTI) (for within laboratory precision) and PEI (for
between laboratory precision) conducted the metals analysis. The decision was
made not to have the repeatability and reproducibility samples analyzed for
semivolatile organics because of the higher detection limits anticipated due to
the presence of oil and grease in the samples.
*Sampling and analytical.
4-20
-------
4.7 BACKGROUND SAMPLES
Two background samples were taken at Texaco outside the western boundary of
the land treatment unit. One sample (BGD-420) was taken approximately 75 feet
west and 25 feet north of ground water sampling well No. 26, and the second
sample (BGD-421) was taken approximately 25 feet north of the first sample (see
Figure 4.1b). The scooping technique was used for sample collection..
The background samples were analyzed for weight loss on drying (LOD) and
then dried in an oven at 105 C for 4 hours. They were also analyzed for
percent silt and percent PM-,,-, content (see Appendix C) .
Portions of the silt fraction generated by screening were sent to RTI and
PEI for metals and semivolatile organics analysis, respectively. They were
analyzed for metals and semivolatile organic compounds as described previously
for the samples from Process L. The low-level extraction and adsorption
chromatography cleanup procedure used resulted in a quantifiable detection
limit of 3-3 ug/g after a 10-fold dilution prior to the GC/MS analysis.
4-21
-------
5.0 QUALITY ASSURANCE
The quality assurance (QA) measures for the chemical analyses were
conducted internally by each laboratory. For the metals analysis, RTI used
National Bureau of Standards (NBS) water (1643 B) as check samples for the
accuracy of the instrumentation. A marine sediment reference material (MESS-1)
acquired from the Marine Analytical Chemistry Standard Program of the National
Research Council of Canada and an NBS fly ash sample (1633 A) were used as QA
samples to check the overall accuracy of the digestion and analysis
procedures. One process sample was spiked with eight elements and their
percent recoveries calculated to assess matrix effects. Another sample (N-452)
was analyzed as duplicates to demonstrate analytical precision. Results of
these checks are presented in Table 5-1-
For the QA on the analysis of the semivolatile organics and pesticides, PEI
used a sample (0-457) for a matrix spike (MS) and a matrix spike duplicate
(MSD). The percent recoveries were determined and the relative percent
difference (RPD) for the duplicates calculated (see Table 5.2). The percent
recoveries for all compounds in both the MS and MSD samples, except pyrene in
the MS sample, were outside the QA limits. Acenaphthene and 2,4-dinitrotoluene
were recovered at levels above the QA limits for both MS and MSD samples. In
the MSD sample only, 4-nitrophenol and pyrene were recovered at levels above
the QA limit. The remaining compounds were not detected at all. The spike
concentrations were all below the quantifiable detection limit for the MS and
MSD samples because of the dilutions required for the GC/MS analysis.
5-1
-------
TABLE 5.1. QUALITY ASSURANCE RESULTS FOR METALS ANALYSIS
Sasple Identity
Elements (ug/g)
Aluainua (AD
Antinony (Sb)
Arsenic (As)
Bariua (Ba)
BerylliuB (Be)
Cadffliua (Cd)
Chromium (Cr)
Cobalt (Co)
Copper (Cu)
Iron (Fe)
Lead (Pb)
Manganese (Hn)
Mercury (Hg)
Molybdenua (Ho)
Nickel (Ni)
DsniuB (Os)
Selenium (5e)
Silver (Ag)
Thai liuu (Tl)
Vanadiua (V)
Zinc (Zn)
cyanide
EPA Check
Expected
(ug/g)
-
8.2
43.0
-
29.0
9.1
7.1
43.0
8.9
.
43.0
13.0
-
-
-
•
7.6
-
25.2
130
10.0
-
Sanple
Found
(ug/g)
-
9.0
43.6
-
30.5
7.7
6.8
40.1
12.3
-
43.0
12.9
-
-
-
-
6.9
-
26.7
123
10.0
-
NBS Fly Ash 1633 A
Expected
(ug/g)
140,000
7.0
145
1500
12.0
1.0
196
46.0
118
94,000
72.4
190
-
29
127
-
10.3
-
5.7
300
200
-
Found
(ug/g)
17,000
2.6
129
700
4.2
5.5
35.4
25.0
38.5
22,200
31.8
27.9
-
-
53.3
-
7.7
-
3.3
121
69.2
-
NRC Seditent MESS-1
Expected
(ug/g)
58,000
-
10.6
-
1.9
0.6
71.0
10.8
25.1
36,500
34.0
513
-
-
29.5
-
0.4
-
0.7
72.4
191
-
Found
(ug/g)
18,000
<0.5
7.9
87.3
1.4
0.4
40.1
10.2
22.3
25,000
23. B
344
-
-
33.5
-
<0.5
-
<2
54.0
171
-
Matrix Spike
Added
(ug)
-
-
10.0
100.0
100.0
100.0
100.0
-
100.0
-
10.0
100.0
0.40
100.0
100.0
-
11.0
100.0
10.0
100.0
100.0
-
Recovered
(ug)
-
-
8.9
. 239.0
94.2
90.2
97.1
-
96.5
-
8.6
102.0
0.36
91.5
93.7
-
10.0
81.8
11.0
32.8
94.4
-
Percent
-
-
89. OX
239*
94. 21
90.27.
97. IX
-
96. 5X
-
86. OX
102.0X
90. OX
91. 5X
93. 7X
-
90. 9X
91.81
110. OX
32. 8X
94. 4X
-
Duplicates
N-452
(ug/g)
-
0.6
6.2
272
<1
1.9
196
14.7
200
-
65.0
389
-
5.7
94.4
<1
3.1
-
<2
190.0
232
-
N-452
(ug/g)
-
1.0
7.7
147
<1
<1
189
11.8
196
-
55.0
379
-
<2
89.0
<1
2.9
-
<2
178.0
229
-
5-2
-------
TABLE 5.2. QUALITY ASSURANCE RESULTS FOR SEHIVOLATILE OR6ANICS ANALYSIS
SOIL SURROGATE PERCENT RECOVERY SUMMARY
Sanple Identity
Surrogate Coopounds
Nitrobenzene-d5
2-Fluorobiphenyl
Terphenyl-dl4
Phenol-d5
2-Fluorophenol
2,4,6-Tribronophenol
Silt
L-434
01
667,
164X
01
ox
ox
Silt
11-438
01
01
148X
OX
OX
OX
Silt
N-447
OX
OX
119X
OX
OX
OX
PH-10
N-449
OX
OX
105X
OX
OX
OX
Silt
0-457
OX
207X
259X
OX
OX
OX
Silt
BGD-454
OX
171
B3X
OX
OX
661
Sasple
Blank
OX
OX
120X
OX
OX
54X
Matrix
Spike
OX
821
110X
OX
OX
OX
Matrix Spike
Duplicate
OX
OX
192X
OX
OX
OX
SOIL MATRIX SPIKE/MATRIX SPIKE DUPLICATE RECOVERY SUMMARY
Sanple Identity
0-457
Compound
1,2,4-Trichlorobenzene
Acenaphthene
2,4-Dinitrotoluene
Pyrene
N-Nitrosodi-n-Propylaaine
1,4-Dichlorobenzene
Pentachlorophenol
Phenol
2-Chlorophenol
4-Chl or o-3-nethyl phenol
4-Nitrophenol
2-Hethylnapthalene
Phenanthracene
Napthalene *
Fluorene *
N-Nitrosodiphenylaffline *
3,3'-Dichlorobenzidine *
Benzo(a) anthracene *
Benzo(g,h,i)perylene *
Spike Unspiked
Cone. Saaple
(ug/g)
3.25
3.25
3.25
3.25
3.25
3.25
6.50
6.50
6.50
6.50
6.50
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
(ug/g)
0.0
0.0
0.0
9.5 J
0.0
0.0
0.0
0.0
0.0
0.0
0.0
45.0 J
22.0 J
0.0
0.0
0.0
0.0
0.0
0.0
Matrix Percent
Spike Recovery
(ug/g)
0.0
B.2 J
7.3 J
8.0 J
0.0
0.0
0.0
0.0
0.0
0.0
0.0
64.0 J
58.0 J
B.2 J
9.1 J
22.0 J
24.0 J
14.0 J
14.0 J
OX
253X
225X
-4BX
OX -
OX
OX
07.
OX
OX
OX
142X
264X
-
-
-
-
-
-
Matrix Spike
Duplicate
(ug/g)
0.0
8.5 J
5.3 J
17.0 J
0.0
0.0
0.0
0.0
0.0
0.0
8.5 J
65.0 J
47.0 J
7.40 J
12.00 J
28.00 J
51.00 J
0.00
0.00
Percent
Recovery
OX
262X
164X
232X
OX
OX
OX
OX
OX
OX
131X
144X
214X
-
-
-
-
-
-
RPD
OX
-0.4X
3. 97.
-37. 9X
OX
OX
OX
OX
OX
OX
07.
-0.27.
2.67.
-
-
-
-
-
-
Sasiple Detection Liait (ug/g)
85.6
90.6
105.4
= Compound Mas not detected in the unspiked sanple and Has not spiked, but was detected in the aatrix spike sample
and/or natrix spike duplicate saaple.
= Estieated value where the compound meets the nass spectral or chroaatographic criteria
but is below the quantifiable liait
METHOD BLANK SUMMARY FOR SEMIVOLATILE ORBANICS ANALYSIS
Blank ID
Compound Identity
None Detected
Concentration
None Detected
Saople Blank for
Seoivolatile Organics
5-3
-------
Six compounds were detected below the quantifiable limit in the MS and/or
MSD that were not detected in the unspiked sample. The cause of the compounds
not being found in the unspiked sample was probably the result of the dilution
of the samples.
All samples received, including the laboratory blanks, were spiked with
surrogate compounds and the percent recoveries of these compounds were
determined (see Table 5-2). Nitrobenzene-d... phenol-d.., and 2-fluorophenol
5 5
were not detected in any sample. Terphenyl-d..^. was detected in all the
samples; the recovery of terphenyl-d^ for samples L-434, M-438, 0-457. and the
0-457 MSD was above the QA limit. For 2,4,.6-tribromophenol, recoveries for the
method blank and the background sample were within the QA limits; the compound
was not detected in the rest of the samples. For 2-fluorobiphenyl, sample
L-434 and the 0-457 MS showed recoveries within the QA limits; sample 0-456
showed a recovery above the QA limit and the background sample showed a
recovery below the QA limit. The surrogate compound was not detected in the
other samples. Again, the dilution of the sample prior to the GC/MS analysis
was thought to be the cause of not detecting the surrogate compounds.
Analysis was conducted on a blank sample consisting of a purified solid
matrix spiked with surrogate compounds and carried through extraction and
concentration (see Table 5-2). The CLP specifies limits for the blanks on the
levels of common phthalate esters and Hazardous Substances List (HSL)
compounds. In the blank sample, no phthalate esters or HSL compounds were
detected at a quantifiable limit of 0.33 ug/g.
Entropy conducted a independent performance audit by spiking a silt sample
from the repeatability and reproducibility sample set. Two aliquots of a silt
composite made from sample 0-458 were used for the metals spikes (samples 0-495
and 0-497). The elements and their concentractions in the spiking solution
5-4
-------
used for the metals spike are listed in Table C.10 of Appendix C. The metals
spike was added to achieve approximately 100 ug/g concentration with the exact
concentration depending on the actual sample weight. The exact concentration
of the metals spike, the analysis of the unspiked silt sample and the spiked
sample, and the percent recoveries for each element are presented in Table 5-2.
5-5
------- |