xvEPA
United States
Environmental Protection
Agency
Office of Research and
Development
Washington DC 20460
EPA/540/R-95/520
August 1995
Site Characterization
Analysis Penetrometer
System (SCAPS)
Innovative Technology
Evaluation Report
SUPERFUND INNOVATIVE
TECHNOLOGY EVALUATION
-------
CONTACT
Laiy Jack and Steve Rock are the EPA contacts for this report. Lary Jack is presently with the new
Characterization Research Division (formerly the Environmental Monitoring Systems Laboratory) in Las
Vegas, NV, which is under the direction of the National Exposure Research Laboratory with headquarters in
Research Triangle Park, NC.
Steve Rock is presently with the new Land Remediation and Pollution Control Division (formerly the Risk
Reduction Engineering Laboratory) in the newly organized National Risk Management Research Laboratory
in Cincinnati, OH.
-------
EPA/540/R-95/520
August 1995
SITE CHARACTERIZATION ANALYSIS
PENETROMETER SYSTEM (SCAPS)
INNOVATIVE TECHNOLOGY EVALUATION REPORT
NATIONAL RISK MANAGEMENT RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
NATIONAL EXPOSURE RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
LAS VEGAS, NEVADA 89193
Printed on Recycled Paper
-------
Notice
The information in this document has been funded wholly or in part by the U.S. Environmental Protection Agency (EPA)
in partial fulfillment of Contract No. 68-CO-0047, Work Assignment NO. 0-40, to PRC Environmental Management, Inc.
It has been subject to the Agency's peer ad administrative reivew, and it has been approved for publication as an EPA
document. The opinions, findings, and conclusions expressed herein are those of the contractor and not necessarily those
of the EPA or other cooperating agencies. Mention of company or product names is not to be construed as an endorsement
by the agency.
-------
Foreword
The U.S. Environmental Protection Agency is charged by Congress with protecting the Nation's land, air, and water
resources. Under a mandate of national environmental laws, the Agency strives to formulate and implement actions leading
to a compatible balance between human activities and the ability of natural systems to support and nurture life. To meet this
mandate, EPA's research program is providing data and technical support for solving environmental problems today and
building a science knowledge base necessary to manage our ecological resources wisely, understand how pollutants affect
our health, and prevent or reduce environmental risks in the future.
The National Risk Management Research Laboratory is the Agency's center for investigation of technological and
management approaches for reducing risks from threats to human health and the environment. The focus of the Laboratory's
research program is on methods for the prevention and control of pollution to air, land, water and subsurface resources;
protection of water quality in public water systems ; remediation of contaminated sites and ground water, and prevention
and control of indoor air pollution. The goal of this research effort is to catalyze development and implementation of
innovative, cost-effective environmental technologies; develop scientific and engineering information needed by EPA to
support regulatory and policy decisions; and provide technical support and information transfer to ensure effective
implementation of environmental regulations and strategies.
This publication has been produced as part of the Laboratory's strategic long-term research plan. It is published and
made available by EPA's Office of Research and Development to assist the user community and to link researchers with their
clients.
E. Timothy Oppelt, Director
National Risk Management Research Laboratory
-------
Abstract
In August 1994, a demonstration of cone penetrometer-mounted sensor technologies took place to evaluate their
effectiveness in sampling and analyzing the physical and chemical characteristics of subsurface soil at hazardous waste sites.
The effectiveness of each technology was evaluated by comparing each technology's results to the results obtained using
conventional reference methods. The demonstration was developed under the Environmental Protection Agency's Superfund
Innovative Technology Evaluation Program.
Three technologies were evaluated: the Site Characterization and Analysis Penetrometer System (SCAPS) Laser Induced
Fluorescence (LIF) sensor developed by the Tri-Services (Army, Navy, and Air Force), the Rapid Optical Screening Tool
(ROST™) developed by Loral Corporation and Dakota Technologies, Inc., and the conductivity sensor developed by
Geoprobe® Systems. These technologies were designed to provide rapid sampling and real-time, relatively low cost analysis
of the physical and chemical characteristics of subsurface soil to quickly distinguish contaminated, areas from
noncontaminated areas. Results for the ROST™ and Geoprobe® technologies are presented in separate reports similar to
this one.
Three sites were selected for the demonstration, each contained varying concentrations of coal tar waste and petroleum fuels,
and wide ranges in soil texture.
This demonstration found that the SCAPS technology produced screening level data. Specifically, the qualitative assessment
showed that the stratigraphic and chemical cross sections from SCAPS technology were comparable to the reference
methods. The technology's identification of the relative magnitude of contamination generally matched the reference data.
The quantitative assessment found that the SCAPS data was most closely correlated to the total petroleum hydrocarbons and
volatile petroleum hydrocarbons data. Based on this study, the SCAPS technology appears to be capable of rapidly and
reliably mapping the relative magnitude of the vertical and horizontal extent of subsurface contamination, when that
contamination is fluorescent. This type of contamination includes petroleum fuels and polynuclear aromatic hydrocarbons.
IV
-------
Table of Contents
Section Page
Notice ii
Foreword Hi
Abstract iv
List of Figures vii
List of Tables vii
List of Abbreviations and Acronyms viii
Acknowledgements x
1 Executive Summary 1
2 Introduction 3
Demonstration Background, Purpose, and Objectives 3
Demonstration Design 4
Qualitative Evaluation 4
Quantitative Evaluation 6
Deviations from the Approved Demonstration Plan 7
Site Descriptions 7
3 Reference Method Results 9
Reference Laboratory Procedures 9
Sample Holding Times 9
Sample Preparation 9
Initial and Continuing Calibrations 10
Sample Analysis 10
Detection Limits 11
Quality Control Procedures 11
Confirmation of Analytical Results 12
Data Reporting 12
Quality Assessment of Reference Laboratory Data 12
Accuracy 12
Precision 12
Completeness 12
Use of Qualified Data for Statistical Analysis 12
Chemical Cross Sections 13
Atlantic Site 13
York Site 13
Fort Riiey Site 17
Quality Assessment of Geotechnical Laboratory Data 17
Geotechnical Laboratory 17
Borehole Logging 17
Sampling Depth Control 17
v
-------
Table of Contents (Continued)
Section
Page
Stratigraphic Cross Sections 17
Atlantic Site
-------
List of Figures
2-1 Typical Transect Sampling Line and Stratified Random Sampling Grid 6
3-1 TPH Reference Method Chemical Cross Section — Atlantic Site 14
3-2 PAH Reference Method Chemical Cross Section — Atlantic Site 14
3-3 TPH Reference Method Chemical Cross Section — York Site 15
3-4 PAH Reference Method Chemical Cross Section —York Site 15
3-5 TPH Reference Method Chemical Cross Section — Fort Riley Site 16
3-6 PAH Reference Method Chemical Cross Section — Fort Riley Site 16
3-7 Reference Method Stratigraphic Cross Section — Atlantic Site 18
3-8 Reference Method Stratigraphic Cross Section — Yorlc Site 19
3-9 Reference Method Stratigraphic Cross Section — Fort: Riley Site 19
4-1 Tri-Services SCAPS 24
4-2 SCAPS Panel Plot — Node 4 Atlantic Site 24
4-3 SCAPS Chemical Cross Section — Atlantic Site 29
4-4 SCAPS Chemical Cross Section —York Site 29
4-5 SCAPS Chemical Cross Section — Fort Riley Site 30
4-6 SCAPS Stratigraphic Cross Section — Atlantic Site 30
4-7 SCAPS Stratigraphic Cross Section — York Site 33
4-8 SCAPS Stratigraphic Cross Section — Fort Riley Site 33
4-9 Fluorescence Intensity vs. Wavelength — Node 4 Atlantic Site 34
5-1 Normalized LIF and Qualitative Reference Data — Atlantic Site 38
5-2 Normalized LIF and Qualitative Reference Data — Yoirk Site 38
5-3 Normalized LIF and Qualitative Reference Data — Fort Riley Site 39
List of Tables
Table
2-1
3-1
4-1
4-2
4-3
5-1
5-2
5-3
Page
Criteria for Data Quality Characterization Site
Comparison of Geologist Data and Geotechnical Laboratory Data — All Sites ...
Quantitative Evaluation Data for the Atlantic Site
Quantitative Evaluation Data for the York Site
Quantitative Evaluation Data for the Fort Riley Site
Regression Analysis Results for SCAPS and the Reference Methods — All Sites
Regression Analysis Results for the Average of Both SCAPS Pushes
and the Reference Methods — All Sites
Data for Mean SCAPS, TPH, and VPH — All Sites
. 5
20
28
28
28
42
44
45
VII
-------
List of Abbreviations and Acronyms
AEG Army Environmental Center
ASTM American Society for Testing Materials
bgs below ground surface
BTEX benzene, toluene, ethylbenzene, and xylene
CCAL continuing calibrations
cm centimeter
cm/s centimeters per second
CP cone penetrometer
DQO data quality objective
EPA Environmental Protection Agency
ERA Environmental Resource Associates
ETS Environmental Technical Services
FID flameionization detector
FMGP former manufactured gas plant
GC gas chromatograph
Geoprobe® Geoprobe®Systems
HPLC high performance liquid chromatography
Hz pulses per second (hertz)
ICAL initial calibrations
ITER innovative technology evaluation report
LCS laboratory control samples
LIF laser induced fluorescence
MDL method detection limit
Method OA-1 University of Iowa Hygienics Laboratory Method OA-1
micrograms per kilogram
micrograms per liter
micrometer
mg/kg milligrams per kilogram
mg/mL milligrams per milliliter
Mj megajoule
mL milliliter
mm millimeter
MMTP Monitoring and Measurement Technologies Program
MS matrix spike
MSD matrix spike duplicate
NRMRL National Risk Management Research Laboratory
NERL-CRD National Exposure Research Laboratory-Characterization Research Division
nm nanometer
OMA optical multichannel analyzer
%D percent difference
%RSD percent relative standard deviation
PAH polynuclear aromatic hydrocarbon
y^g/L
viii
-------
List of Abbreviations and Acronyms (Continued)
PARC Princeton Applied Research Company
PDA photodiode array
PE performance evaluation
PID photoionization detector
ppm parts per million
PRC PRC Environmental Management, Inc.
PRL PACE reporting limit
PTI Photon Technology, Inc.
QA/QC quality assurance/quality control
QAPP quality assurance project plan
ROST™ Rapid Optical Screening Tool
RPD relative percent difference
RSD relative standard deviation
SCAPS Site Characterization and Analysis Penetrometer System
SITE Superfund Innovative Technology Evaluation
TER technology evaluation record
TOC total organic carbon
TPH total petroleum hydrocarbon
USCS Unified Soil Classification System
USDA United States Department of Agriculture
VOC volatile organic compound
VPH volatile petroleum hydrocarbon
WES Waterways Experiment Station
IX
-------
Acknowledgements
We wish to acknowledge the support of all those who helped plan and conduct this demonstration, interpret data, and prepare
this evaluation report In particular, for demonstration site access and relevant background information, Dean Harger (Iowa
Electric Company), Ron Buhrman (Burlington Northern Railroad), Abdul Al-Assi (U.S. Army Directorate of Engineering
and Housing); for turn-key implementation of this demonstration, Eric Hess, Darrell Hamilton, Harry Ellis (PRC
Environmental Management, Inc.) (913) 281-2277); for editorial and publication support, Suzanne Ladish and Frank
Douglas (PRC); for technical and peer review Dr. T. Vo-Dinh (Oak Ridge National Laboratory), Robert Knowlton (Sandia
National Laboratories), and JeffKelley (Nebraska Department of Environmental Quality); and for EPA project management,
Lary Jack (National Exposure Research Laboratory-Characterization Research Division) (702)798-2373). In addition, we
gratefully acknowledge the participation of the developers, the Tri-Services Site Characterization Analysis and Penetrometer
System (SCAPS) group (410)612-6836).
-------
Section 1
Executive Summary
Recent changes hi environmental site character-
ization have resulted hi the application of cone
penetrometer (CP) technologies to site characterization.
With a variety of in situ physical and chemical sensors,
this technology is seeing an increased frequency of use
in environmental site characterization. CP technologies
employ a wide array of sampling tools and produce
limited investigation-derived waste.
The EPA's Monitoring and Measurement
Technologies Program (MMTP) at the National
Exposure Research Laboratory, Las Vegas, Nevada,
selected CP sensors as a technology class to be evaluated
under the Superfund Innovative Technology Evaluation
(SITE) Program. In August 1994, a demonstration of
CP-mounted sensor technologies took place to evaluate
how effective they were in analyzing the physical and
chemical characteristics of subsurface soil at hazardous
waste sites. Prior to this demonstration, two separate
predemonstration sampling efforts were conducted to
provide the developers with site-specific samples. These
samples were intended to provide data for site-specific
calibration of the technologies and matrix interferences.
The main objective of this demonstration was to
examine technology performance by comparing each
technology's results relative to physical and chemical
characterization techniques obtained using conventional
reference methods. The primary focus of the
demonstration was to evaluate the ability of the
technologies to detect the relative magnitude of
fluorescing subsurface contaminants. This evaluation is
described in this report as the qualitative evaluation. A
subordinate focus was to evaluate the possible
correlations or comparability of the technologies
chemical data with reference method data. This
evaluation is described hi this report as the quantitative
evaluation. All of the technologies were designed and
marketed to produce only qualitative screening data.
The reference methods for evaluating the physical
characterization capabilities were stratigraphic logs
created, by a geologist from soil samples collected by a
drill rig equipped with hollow stem augers, and soil
samples analyzed by a geotechnical laboratory. The
reference methods for evaluating the chemical
characterization capabilities were EPA Method
418.1 and SW-846 Methods 8310 and 8020, and
University of Iowa Hygienics Laboratory Method OA-1.
In addition, the effect of total organic carbon (TOC) on
technology performance was evaluated.
Three technologies were evaluated: the Site
Characterization and Analysis Penetrometer System
(SCAPS) laser induced fluorescence (LIF) and CP
sensors developed by the Tri-Services (Army, Navy, and
Air Force), the Rapid Optical Screening Tool (ROST™)
developed by Loral Corporation and Dakota
Technologies, Inc., and the conductivity sensor
developed by Geoprobe® Systems. These technologies
were designed to provide real-tune, relatively low cost
analysis of the physical and chemical characteristics
(primarily petroleum fuels and coal tars) of subsurface
soil to quickly distinguish contaminated areas from
noncontaminated areas. The SCAPS technology is
designed and operated to produce screening level data.
Results of the demonstration are summarized by
technology and by data type (chemical or physical) hi
individual innovative technology evaluation reports
(ITER). In addition to the three technology-specific
ITER's, a general ITER that examines cone
penetrometry, geoprobes, and hollow stem auger drilling
hi greater detail has been prepared.
The purpose of this ITER is to chronicle the
development of the SCAPS technology, its capabilities,
associated equipment, and accessories. The document
concludes with an evaluation of how closely the results
obtained using the technology compare to the results
obtained using conventional reference methods.
One hazardous waste site each was selected hi Iowa,
Nebraska, and Kansas to demonstrate the technologies.
-------
The sites were selected because of thek varying
concentrations of coal tar waste and petroleum fuels, and
because of thek ranges hi soil textures.
This demonstration found that the SCAPS
technology produces screening level data. Specifically,
the qualitative assessment showed that the stratigraphic
and the chemical cross sections were comparable to the
reference methods. The SCAPS sensors did not requke
sample collection, and thus, avoided the sampling
difficulties encountered by the reference methods during
this demonstration. The relatively continuous data
output from the LIF sensor eliminated the data
interpolation requked by the reference method. This
also increased the apparent resolution of the sensor's
data.
The SCAPS LIF operator also qualitatively
identified changes in contaminant type by detecting
significant changes hi peak emission wavelength. The
gross soil classifications identified by the technology
generally matched the reference method classifications.
The chemical cross sections for the LIF sensor showed
close agreement to the reference method cross sections
in identifying low, medium, and high zones of
contamination. Generally, the relative LIF intensity was
positively related to the concentration of total petroleum
hydrocarbons and total polynuclear aromatic
hydrocarbons. In only one case during this
demonstration did the SCAPS LIF sensor not identify
fluorescence above background for zones sampled that
indicated contamination. Reference method sampling
indicated contamination hi the 100 's of the parts per
million (ppm) range at Node 5 at the York site. The
failure of the SCAPS LIF sensor to identify this zone
may have been a result of the horizontal separation
between the SCAPS and reference method sampling
points, and inherent matrix heterogeneity. The
quantitative assessment found that the SCAPS LIF data
was most closely correlated to the TPH and volatile
petroleum hydrocarbons (VPH) data. Due to matrix
heterogeneity up to 50 percent of die original data set
used in the quantitative evaluation was eliminated as
outliers. This greatly reduced the predictive value of the
regression models, however, the remaining data was still
used to identify trends. The quantitative data assessment
also produced a first approximation of a detection
threshold for the SCAPS LIF sensor. For TPH and
VPH, based on thek regression models, the fluorescence
intensity (background corrected) at 0 milligram per
kilogram was 157 and 336, respectively. In addition, the
lowest concentrations of TPH and VPH detected during
the quantitative assessment were 60 and 19 mg/kg,
respectively. Both of these low concentrations had
fluorescence intensity readings near the thresholds
(157 and 336) discussed above.
Based on the continuous data output for both the
chemical and physical properties of soil, the SCAPS
sensors (physical and chemical) appear to be valuable
tools for qualitative site characterization. The lack of
better correlation for the quantitative evaluation cannot
be solely attributed to the technology. It may also be
due to the combined effect of matrix heterogeneity, lack
of instrument calibration, uncertainties regarding the
exact contaminants being measured, and the age and
constituents hi the waste. Based on the data from this
demonstration, it is not possible to conclude that the
technology can or cannot be quantitative hi the
configuration used during this demonstration. Based on
the effects listed above, potential users should not expect
the SCAPS LIF sensor to produce data which shows a
high degree of correlation when comparisons with
conventional data are made on a point-by-point basis.
Verification of this technology's performance should be
done only on a qualitative level. Even though it cannot
quantify actual levels of contamination or identify
individual compounds, it can produce relative
contaminant distribution data very similar to
corresponding data produced by conventional methods,
such as drilling and laboratory sample analysis, and it
can monitor changes hi emission wavelength to identify
possible changes hi contaminant constituent. The
general magnitude of the LIF sensor data directly
correlated to the general magnitude of contamination
detected by the reference method. The SCAPS
performance during this demonstration showed that it
could generate this data faster than the reference
methods and with little to no waste generation relative to
the reference methods. The cost associated with using
this technology to produce the qualitative data was
approximately $42,000, as compared to the $55,000 used
to produce the reference cross sections, hi this case, the
SCAPS LIF and CP sensors cost less than reference
methods, produced almost 1,200 more data points
(continuously) than the conventional approach, and
provided data in a real-tune fashion. It should be noted
that the technology's data is screening level, while the
reference method approach produced definitive data.
The question that this demonstration cannot answer is
whether or not it is better to have few data points at the
highest data quality level or many more at a lower data
quality level. Issues such as matrix heterogeneity may
greatly reduce the need for definitive level data hi an
initial site characterization. Critical samples will always
requke definitive analysis.
-------
Section 2
Introduction
The purpose of this ITER is to present information on
the demonstration of the SGAPS LIF and CP sensors, a
system designed to provide screening type data on the
physical and chemical characteristics of subsurface soil.
This system uses laser light to cause fluorescing
contaminants hi soils to fluoresce and measures the
resulting fluorescence. Currently, this technology is being
used most commonly to detect PAH compounds associated
with petroleum fuel.
More detailed information regarding aspects of this
report can be found hi the January 1995 technology
evaluation record (TER) for this demonstration. The TER
contains all of the raw data and is not intended for general
circulation, however, portions of the TER can be accessed
by contacting the EPA technical project manager.
The SCAPS sensors were demonstrated in conjunction
with two other sensor technologies: (1) the ROST™
developed by Loral Corporation and Dakota Technologies,
Inc., and (2) the conductivity • sensor developed by
Geoprobe®. The results of the demonstration of these
other two technologies are presented in individual ITERs
similar to this document. An additional general ITER was
prepared which discusses the history, sampling, and other
capabilities of cone penetrometry, Geoprobe®, and hollow
stem auger drilling. Complete details of the
demonstration, descriptions of the sites, and the
experimental design are provided hi the August 1994 final
demonstration plan for geoprobe- and CP-mounted
sensors. This information is briefly summarized hi this
document.
This section summarizes general information about
the demonstration such as the purpose, objectives, and
design. Section 3 presents and discusses the validity of
the data produced by the reference methods used hi the
evaluation of two SCAPS sensors: the LIF sensor and the
CP sensor. Section 4 discusses the SCAPS sensors, their
capabilities, and equipment and accessories. Section 5
evaluates how closely the results obtained using the
SCAPS sensors compare to the results obtained using the
reference methods. Section 6 discusses the potential
applications of the technology. Section
7 presents developer comments, EPA response to
developer comments, and developer update on the
technology.
Demonstration Background, Purpose,
and Objectives
The demonstration was developed under the
Measuring and Monitoring Technologies Program
(MMTP), a component of the EPA's SITE Program. The
goal of the MMTP is to identify and demonstrate new,
viable technologies that can identify, quantify, or monitor
changes hi contaminants at hazardous waste sites or that
can be used to characterize a site cheaper, better, faster,
and safer than conventional technologies.
The SCAPS LIF sensor uses LIF to detect the
subsurface presence or absence of fluorescing compounds,
such as petroleum fuels and coal tar wastes. This
technology is attached to and advanced into the soil with
a conventional CP sensor. The SCAPS LIF and CP
sensors, were designed to provide rapid, continuous, hi situ
real-time, relatively low cost analysis of the physical and
chemical characteristics of subsurface soil. The
identification of subsurface chemical characteristics
involves quickly identifying the presence or absence of
contamination, and relative concentrations. These
capabilities would allow investigation and remediation
decisions to be made more efficiently and quickly,
reducing overall project costs such as the number of
samples that need to be submitted for costly and time
consuming confirmatory analyses, and costs associated
with multiple mobilizations.
The primary focus of the demonstration was to
evaluate the ability of the technologies to detect the
relative magnitude of fluorescing subsurface contaminants.
This evaluation is described in this report as the qualitative
evaluation. A subordinate focus was to evaluate the
possible correlations or comparability of the technologies
chemicid data with reference method data. This evaluation
is described hi this report as the quantitative evaluation.
-------
evaluate the possible correlations or comparability of the
technologies chemical data with reference method data.
This evaluation is described in this report as the
quantitative evaluation. All of the technologies were
designed and marketed to produce only qualitative
screening data.
There were three objectives for the qualitative
evaluations, and one objective for the quantitative
evaluation conducted during this demonstration. The
first qualitative objective evaluated the SCAPS LIF
sensor for its ability to vertically delineate subsurface
soil contamination. Cross sections of subsurface
contaminant plumes produced by the technology were
visually compared to corresponding cross sections
produced by the reference methods. The second
qualitative objective evaluated the SCAPS CP sensor for
its ability to characterize physical properties of
subsurface soils. The third qualitative objective was to
evaluate the SCAPS sensors for their reliability,
ruggedness, cost, and range of application. The SCAPS
LIF sensor was quantitatively evaluated on how its data
compared to the data from the reference methods, and an
attempt was made to identify the technology's threshold
detection limits.
Demonstration Design
The experimental design of this demonstration was
created to meet the specific qualitative and quantitative
objectives described in Section 3. The experimental
design was approved by all demonstration participants
prior to the start of the demonstration. This experi-
mental design is detailed in the final demonstration plan
(PRC 1994).
Sample results from the SCAPS sensors were
compared to results from the reference methods. For
this demonstration, the reference methods included
standard SW-846 methods for measuring petroleum
hydrocarbons and PAHs, and borehole logging and
sampling by a geologist using hollow stem auger
drilling. These comparisons are called intramethod
comparisons. These comparisons were used to
determine the quality of data produced by the
technology. Two data quality levels were considered
during this evaluation: definitive and screening data.
These data quality levels are described in Data Quality
Objectives Process for Superfund - Interim Final
Guidance (EPA 1993;.
Definitive data are generated using rigorous
analytical methods, such as approved EPA reference
methods. Data are analyte-specific with confirmation of
analyte identity and concentration. Methods produce
tangible raw data (e.g., chromatograms, spectra, digital
values) in the form of paper printouts or
computer-generated electronic files. Data may be
generated at the site or at an off-site location, as long as
the quality assurance/quality control (QA/QC)
requirements are satisfied. For the data to be definitive,
either analytical or total measurement error must be
determined.
Screening data are generated by rapid, less precise
methods of analysis with less rigorous sample
preparation. Sample preparation steps may be restricted
to simple procedures, such as dilution with a solvent
instead of elaborate extraction/digestion and cleanup.
Screening data provide analyte identification and
quantification, although the quantification may be
relatively imprecise. At least 10 percent of the
screening data are confirmed using analytical methods
and QA/QC procedures and criteria associated with
definitive data. Screening data without associated
confirmation data are not considered to be of known
quality.
Since this technology is new and Innovative,
approved EPA methods for in situ laser induced
fluorescence do not exist. For the purpose of diis
demonstration, the lack of approved EPA methods did
not preclude the technology from being considered
definitive. The evaluation of this technology as to its
quantitative capabilities was included to provide potential
users a complete picture of die technology's capabilities.
However, the developer never claimed that the
technology was quantitative. The main criteria for data
quality level assignment was based on the comparability
of the technology's data to the data produced by the
reference methods. Table 2-1 defines the statistical
parameters used to define die data quality levels
produced by SCAPS. These criteria were defined in the
approved demonstration plan, and accepted by the
developers. These are based on past SITE demon-
strations of monitoring technologies.
The sampling and analysis methods used to collect
the baseline data for this demonstration are currently
accepted by EPA for providing legally defensible data.
This data is defined as definitive level data by Superfund
guidance. Therefore, for the purpose of this
demonstration, these technologies and analytical metiiods
were considered reference metiiods.
Qualitative Evaluation
Qualitative evaluations were made through
observations and by comparing stratigraphic and
chemical cross sections from the technology to cross
sections produced from the reference methods. The
reference method for the stratigraphic cross sections was
-------
TABLE 2-1. CRITERIA FOR DATA QUALITY CHARACTERIZATION SITE
Data
Quality
Level
Definitive
Screening
Statistical Parameter
r2 = 0.80 to 1.0, and the slope3 and y-intercept are statistically similar to 1.0 and 0.0, respectively, the
precision (RSD) is less than or equal to 20 percent and inferential statistics indicate the two data sets
are statistically similar.
r2 < 0.80, the precision (RSD) is greater than 20 percent, and the technology meets its developer's
performance specifications, inferential statistics indicate the two data sets are statistically not similar;
or in the case where the regression analysis indicates the data is of definitive quality, but the
inferential statistics indicate the data sets are statistically different.
Notes:
r2
RSD
Since the SCAPS technology did not produce data in equivalent units to the reference method, the slope cannot
be used to assess accuracy, however, comparability can still be evaluated.
Coefficient of determination.
Relative standard deviation.
continuous sampling with a hollow stem auger advanced
by a drill rig and the corresponding borehole logs
created by a geologist. In addition, the technology's
ability to determine subsurface soil texture at discrete
intervals was further compared to data produced by an
off-site geotechnical laboratory. Soil samples were
analyzed for total organic carbon (TOC) using the
90-3 Walkley-Black Method; and soil texture analysis
was performed by American Society of Testing
Materials (ASTM) Method D-422.
The reference methods for the chemical cross
sections were subsurface sampling using a drill rig and
off-site sample analysis by EPA Method 418.1 and
SW-846 Method 8310. EPA Method 418.1 produces
data on TPH concentration. EPA Method 8310 produces
data on PAH concentrations. These reference methods
were selected for the qualitative evaluation based on
recommendations made by the developer, consideration
of the types of fluorescing target compounds, and the
project objectives.
To qualitatively assess the ability of the SCAPS LIF
sensor to identify the presence or absence of
contamination and produce contaminant distribution
cross sections, the technology was required to
continuously sample at five points located along a
transect line at each of the demonstration sites (Figure
2-1). These points were called sample nodes. The
transect line was placed across an area of known
subsurface contamination identified during
predemonstration sampling activities and previous
investi-gative sampling at these sites. A 6-foot by 6-foot
area was marked around each sample node. This area
was subdivided into nine sections of equal, size,
identified as Sections A through I. At least one sample
node per site was placed outside the area of contami-
nation.
Once each 6-foot by 6-foot area was marked,
sampling points for each technology and the reference
methods were assigned randomly at each node. This
produced a stratified random sampling design. Sections
for sampling at each node were only used once.
The potential effect of organic matter was evaluated
qualitatively by TOC analysis of soil samples. This
evaluation was intended to examine potential trends
between TOC content, and how data analyzed using the
technology compared to that obtained by the reference
methods.
The chemical and geotechnical data generated by the
technology was used to produce qualitative data
regarding contaminant and stratigraphic cross sections
along each transect line. These cross sections were
compared to cross sections generated by the reference
method results from soil samples collected with a drill
rig. The comparison of contaminant cross sections
involved visual comparisons with overlays, and
minimum and maximum depths of contamination at each
location along a transect line.
Other factors that underwent qualitative evaluation
were technology costs, ease of operation, ruggedness,
instrument reliability, environmental sampling capability,
and production rates. PRC assigned a person to work
with the developer to become knowledgeable in the use
and application of the SCAPS sensors. Through this
work, PRC was able to assess the operational factors for
the technology.
-------
FIGURE 2-1. TYPICAL TRANSECT SAMPLING LINE AND STRATIFIED RANDOM SAMPLING GRID
STRATIFIED RANDOM SAUPLINR GRID
TYPICAL TRANSECT SAUPI INC I IMF
LEGEND
1 NUU8ERED SAMPLING NOOC
«£,. CONTINUOUS VERTICAL MEASURING POINT FOR
QUALITATIVE ASSESSMENT
^> TARGETED 6 INCH DEPTH INTERVAL FOR
^^ OUANTITAIIVE ASSESSMENT
During the demonstration, a total of 78 soil samples
were collected and analyzed by the reference methods,
and used in the qualitative data evaluations. These
samples were distributed as follows: 28 from the
Atlantic site, 26 from the York site, and 24 from the
Fort Riley site. Only sample data reported as positive
values were used in the evaluation. As described hi the
approved demonstration plan, sample data reported as
"not detected" were not used. As stated hi the approved
demonstration plan, the elimination of these points are
not expected to have a lesser or similar effect as
assigning arbitrary values to non-detects.
Quantitative Evaluation
The SCAPS LJF sensor was evaluated quantitatively
on its ability to chemically characterize subsurface soil
contamination relative to classes of contaminants and
specific contaminants. This evaluation consisted of
comparing data generated using the technology to data
obtained using the reference methods over a wide range
of concentrations. The reference method for the
chemical cross sections soil sampling was hollow stem
auger drilling. University of Iowa Hygienics Laboratory
Method OA-1 (VPH), SW-846 Method 8020 benzene,
toluene, ethylbenzene, and total xylenes (BTEX),
SW-846 Method 8310 (PAH), and EPA Method
418.1 (TPH) were used as the reference analytical
methods. This demonstration attempted to determine if
the results from the SCAPS LIF sensor could be
correlated to results from the reference methods, and if
the technology was able to differentiate between different
types of contamination, such as PAHs, BTEX, coal tars,
and petroleum fuels. In addition, PRC attempted to
determine the detection thresholds for these classes of
contaminants.
To quantitatively assess the comparability of the data
produced by the SCAPS LIF sensor to the reference
methods" data, the demonstration plan required the
technology to conduct its most accurate and precise
measurements at discrete depths at each sampling node.
These depths represented zones of initial contaminant
detection, medium, and high fluorescence. However, at
the start of the demonstration, the developer of the
SCAPS sensors informed PRC that the data produced
during standard dynamic push mode was the most
accurate data that could be produced. Therefore, the
SCAPS LIF data for quantitative evaluation was the
same as that used hi the qualitative evaluations.
The locations for the reference method sampling for
the quantitative evaluation were selected after reviewing
the SCAPS and ROST* data for a site. Sample intervals
that showed similar data from both technologies were
selected as reference method sampling intervals.
Reference method sampling intervals represented zones
of initial contaminant detection, medium, and high
6
-------
fluorescence. The data produced at these intervals
was used to quantify contamination, identify con-
taminants, establish precision control limits, and
establish contamination detection thresholds. Reference
method data was used to assess the comparability of the
data produced by the SCAPS LIP sensor to reference
method chemical analysis.
For the quantitative evaluation, data produced by the
SCAPS LIF sensor was averaged over a 12-inch push
interval corresponding to intervals sampled for reference
method analysis. This data was used to determine a
mean fluorescence over that interval. This data was
compared to corresponding mean reference method con-
centrations for any given interval. To create these mean
reference method concentrations, PRC collected and
homogenized five replicate samples from the 12-inch
depth intervals identified as reference method sampling
intervals which were chosen based on the SCAPS and
ROST™ data. Each replicate sample was collected from
a randomly assigned section at each sample node.
The data developed by the SCAPS LIF sensor was
compared to reference method data for the following
compounds or classes of compounds: TPH, total BTEX,
VPH, total PAH, and individual compounds (BTEX,
naphthalene, 1-methylnaphthalene, 2-methylnaphthalene,
acenaphthene, fluoranthene, pyrene, benzo-a-pyrene,
and anthracene). These comparisons were described in
the August 1994 demonstration plan.
Method precision also was examined during the
demonstration. The SCAPS LIF sensor was required to
produce 10 replicate readings or measurements at given
depths without moving the sensor between readings.
From these 10 measurements at each discrete depth,
precision control limits were established.
For the quantitative evaluation, a total of 103 soil
samples were collected and analyzed by the reference
methods. The distribution of these samples was as
follows: 8 replicate sampling intervals producing
38 samples at the Atlantic site, 7 replicate sampling
intervals producing 35 samples at the York site, and
30 samples from 6 replicate sampling intervals at the
Fort Riley site. Only sample data reported as positive
values were used hi the evaluation. Sample data
reported as "not detected" was not used.
Deviations from the Approved Demonstra-
tion Plan
The primary deviation from the approved
demonstration plan dealt with the statistical analysis for
the quantitative evaluation.
Since the SCAPS sensors did not produce data
directly representing the concentration of contaminants,
or data in the same units as the reference method
analysis, the Wilcoxon Rank Sum Test could not be
used, .and the comparison of the technology's data to
reference method 99 percent confidence intervals was
not made. In addition, the effect of soil moisture was
not examined due to the fact that the bulk of the
contaminated zones at each site were at or near
saturation. Finally, the approved demonstration plan
identified a hydraulic probe sampler as the reference
method for collecting the soil samples used hi the
quantitative evaluations. However, due to sample matrix
affects, the hydraulic probe samples could not meet the
soil sampling objectives regarding sample volume. The
inability of this method to produce full sample recovery
was caused by the saturated fine sands encountered at
many of the target sampling depths. To allow for
adequate sample volume, PRC changed the reference
method for this soil sampling to hollow stem augering
and split spoon sampling.
Site Descriptions
The demonstration took place at three sites within
EPA Region 7. The three sites are the
(1) Atlantic-Poplar Street Former Manufactured Gas
Plant (FMGP) site (Atlantic site), (2) York FMGP site
(York site), and (3) the Fort Riley Building 1245 site
(Fort Riley site). Brief summaries for each site are
given below. Complete details are located in the August
1994 final demonstration plan.
The Atlantic site is located hi Atlantic, Iowa. The
site is surrounded by gas stations, grain elevators, a seed
supply company, and a railroad right-of-way. All
structures associated with the FMGP have been
demolished. A gas station now operates on the location
of the FMGP. The Atlantic Coal Gas Company
operated the FMGP from 1905 to 1925. During that
time, an unknown quantity of coal tar was disposed of on
site. In addition to the coal tar waste, more recent
releases of petroleum from two nearby gas stations also
have occurred. An investigation conducted at the site
from 1990 to 1992 identified the following primary
contaminants: BTEX and PAHs. The local ground-
water contains free petroleum product and pure coal tar.
The York site is located in York, Nebraska. The
site encompasses nearly a half acre hi an industrial
section, of the city. The site is surrounded by a former
railroad right-of-way, a concrete company, a seed
company, and a farm^supply store. The site is nearly
level, and several buildings occupied by the FMGP are
still present. The York Gas and Electric Company
operated the FMGP from 1899 to 1930. Coal tar waste
-------
was disposed of at the site. Current information on the
site suggests that coal tar waste and its constituents
should be the only waste encountered.
The Fort Riley site is located at Building 1245 on
the east side of the Camp Funston area at Fort Riley,
Kansas. Between 1942 and 1990, five 12,000-gallon
steel underground storage tanks were located at this site.
The tanks were used to store leaded and unleaded
gasoline, diesel fuel, and military operations gasoline.
Soil at the site is contaminated with gasoline and diesel
fuel believed to be the result of past petroleum fuel
releases from the underground storage tanks.
-------
Section 3
Reference Method Results
All soil samples collected during this demonstration
were submitted to PACE, Inc. (PACE), for chemical
and geotechnical analysis. The PACE laboratory in
Lenexa, Kansas, performed the 418.1 and Methods
8020 and OA-1 analyses, while the PACE laboratory hi
St. Paul, Minnesota, performed the Method 8310 an-
alysis. PACE subsequently subcontracted the
geotechnical analyses to Environmental Technical
Services (ETS), Petaluma, California. The chemical
data supplied by the reference laboratory, the
geotechnical data supplied by the geotechnical
laboratory, and the data produced by the on-site
professional geologist is discussed in this section.
Reference Laboratory Procedures
Samples collected during this demonstration were
homogenized and split for the following analyses:
• TPH by EPA Method 418.1 (EPA 1986)
• PAH by EPA SW-846 Method 8310 (EPA
1986)
• BTEX by EPA SW-846 Method 8020 (EPA
1986)
• Total VPH as gasoline by University of Iowa
Hygienic Laboratory Method OA-1 (University
Hygienic Laboratory 1991)
• Soil texture and TOC by the 90-3 Walk-
ley-Black Method (Page 1982)
The results of these analyses are summarized hi
Appendix A. The results are reported as wet weight
values as required hi the approved demonstration plan
(PRC 1994). The data is grouped by analytical method,
site, and whether the data is intended for qualitative or
quantitative data evaluation.
The data from the reference" laboratory was
internally reviewed by PACE personnel before the data
was delivered to PRC. PRC personnel conducted a data
review on the results provided by PACE following EPA
guidelines (EPA 1991). PRC reviewed the raw data and
checked the calculated sample values.
The following subsections discuss specific
procedures used to identify and quantitate TPHs, VPHs,
PAHs, BTEX, and TOC. Most of these procedures
involved requirements that were mandatory to guarantee
the quality of the data generated.
Sample Holding Times
The required holding tunes from the date of sample
receipt for each analytical method used to analyze the
soil samples were as follows: University of Iowa
Hygienics Laboratory Method OA-1 (Method OA-1),
14 days for extraction and analysis; EPA SW-846
Method 8020 (BTEX), 14 days for extraction and
analysis; EPA Method 418.1 (TPH), 14 days for
extraction and 40 days for analysis; EPA SW-846
Method 8310 (PAH), 14 days for extraction and 40 days
for analysis; and 90-3 Walkley-Black Method (TOC),
28 days for extraction and analysis.
All holding tunes for the samples were met during
this demonstration.
Sample Preparation
Preparation of soils for TPH analysis was performed
following EPA Method 418.1. This method uses a
Soxhlet extraction as stated in SW-846 Method
9071. The soil sample extracts were analyzed for TPH
using SW-846 Method 418.1.
-------
Extracts for VPH analysis were prepared following
Method OA-1. The BTEX sample preparation require-
ments were carried out as specified in that method.
The preparation of soil samples for TOC analysis
were carried out as specified in the 90-3 Walkley-Black
Method.
Sonication extraction, SW-846 Method 3550, was
used for the preparation of soil samples for SW-846
Method 8310 analysis. The preparation of samples for
PAH analysis by SW-846 Method 8310 were carried out
according to the method requirements.
Initial and Continuing Calibrations
Initial calibrations (ICAL) were performed before
sample analysis began. ICALs for SW-846 Methods
8020, 8310, and 418.1 consisted of the analysis of five
concentrations of standards. Method OA-1 required the
analysis of three concentrations of standards for the
ICAL. Linearity for these ICALs was evaluated by
calculating the percent relative standard deviation
(%RSD) of the calibration factors. The %RSD QC limit
for SW-846 Methods 8020 and 8310 and Method
OA-1 was 20 percent. The calibration factors were
calculated by dividing the response (measured as the area
under the peak or peak height) by the amount of
compound injected on the gas chromatograph (GC)
column. The 90-3 Walkley-Black Method for TOC
required a daily calibration to a reference sulfate
solution. This ICAL was performed in duplicate. All
initial calibrations met the respective method
requirements.
Continuing calibrations (CCAL) were performed on
a daily basis to check the response of the detector by
analyzing a mid-concentration standard and comparing
the calibration factor to that of the mean calibration
factor from the ICAL.
Calibration factors were monitored hi accordance
with the SW-846 and OA-1 Methods. No CCAL was
performed for the 90-3 Walkley-Black Method. Six
CCALs exceeded the 15 percent difference (%D) criteria
for various BTEX compounds. This resulted hi sample
results being qualified as estimated (J) and usable for
limited purposes. Various PAH compounds in six
SW-846 Method 8310 CCALs exceeded 15 %D for one
of the two detectors. Sample results for the compounds
exceeding 15 %D were qualified as estimated (J) and
usable for limited purposes. SW-846 Method 8310 uses
two detectors, an ultraviolet detector and a fluorescence
detector. Since one detector's CCAL response was
within QC guidelines, this data was considered useable.
Retention tunes of the single analytes were
monitored through the amount of retention time shift
from the CCAL standard as compared to the ICAL
standard. The retention tune windows for SW-846
Method 8310 were set by taking three tunes the standard
deviation of the retention times that were calculated from
the ICAL and CCAL standards. The retention tune
windows for SW-846 Method 8020 were set by PACE at
plus or minus 0.07 minutes for benzene, ethylbenzene,
m-xylene, and plus or minus 0.10 minutes for toluene.
No CCAL retention times for the individual PAH
analytes were outside the retention tune windows.
CCAL retention times for the individual BTEX analytes
were observed outside the retention time windows as set
by the ICAL. No samples were qualified based on this
QC criteria because the retention time shifts were
adjusted appropriately by PACE for sample identification
and quantitation.
Following the ICAL, a method blank was analyzed
to verify that the instrument met the method
requirements. Following this, sample analysis may
continue for 24 hours. As stated hi SW-846 Method
8000, a CCAL must be analyzed and the calibration
factor verified on each working day. Sample analysis
may continue as long as CCAL standards meet the
method requirements.
Sample Analysis
Specific PAH and BTEX compounds were identified
in a sample by matching retention times of peaks found
after analyzing the sample with those compounds found
hi PAH and BTEX standards. VPH was identified hi a
sample by matching peak patterns found after analyzing
the sample with those compounds found hi VPH
standards. Peak patterns may not always match exactly
because of the way the VPHs were manufactured or
because of the effects of weathering. When peak
patterns do not match, the analyst must decide the
validity of the identification of VPHs. For this reason,
peak pattern identification is highly dependent on the
experience and interpretation of the analyst.
Quantitation of PAHs, BTEX compounds, TPHs,
and VPHs was performed by measuring the response of
the peaks hi the sample to those same peaks identified hi
the ICAL standard. The reported results of this
calculation were based on wet weights (except for
PAHs). PAH data was reported on a dry-weight basis.
PRC converted this data to wet-weight based results.
Quantitation of TOC was performed by measuring the
volume of K2Cr2O7 titrated and calculating the
milliequivalents of K2Cr2O7 titrated. This value was
then multiplied by conversion factors and subsequently
10
-------
divided by the grams of sample. TOC results were
reported on a wet-weight basis.
Sample extracts can frequently exceed the
calibration range determined during the ICAL. When
this occurred, the extracts were diluted to obtain peaks
that fall within the linear range of the detector. For
BTEX compounds and VPHs, this linear range was
defined as the highest standard concentration response of
the analytes of interest analyzed during the ICAL. The
linear range for TPHs was defined as an absorbance
maximum of 0.8. For PAHs, as defined in SW-846
Method 8310, the linear range was from 8 tunes the
method detection limit (MDL) to 800 tunes the MDL
with the following exception: benzo(ghi)perylene
recovery at 80 tunes and 800 tunes the MDL are low.
Once a sample was diluted to within the linear range, it
was analyzed again. Dilutions were performed when
appropriate on the samples for this demonstration.
Detection Limits
The PACE reporting limit (PRL) for PAHs was
calculated by multiplying the calibration correction
factor based on dry weight, tunes the MDL for each
specific PAH. PRLs for BTEX compounds were
determined by the lowest concentration standard of the
ICAL. The BTEX ICAL concentration range was from
10 micrograms per liter (^g/L) to 100 jj,g/L. The PRL
for benzene, toluene, and ethyl benzene was 50 micro-
grams per kilogram C/g/kg) and 100 Mg/kg for total
xylene. The three levels of standard concentrations for
the VPH ICAL ranged from 2 milligrams per milliliter
(mg/mL) to 8 mg/mL. The PRL for VPH was 5 mil-
ligrams per kilograms (mg/kg). For TPH, the
calibration range was calculated by calibrating the
infrared detector using a series of working standards. A
plot was then prepared of absorbance versus milligram
petroleum hydrocarbons per 100 milliliter (mL) solution.
The PRL for TPH was 10 mg/kg. The MDL for TOC
analysis was 10 mg/kg wet weight.
Quality Control Procedures
A number of QC measures were used by PACE as
required by SW-846 Methods 8310 and 8020, EPA
Method 418.1, Method OA-1, and the 90-3 Walk-
ley-Black Method. These QC measures included the
analyses of method blanks, instrument blanks, laboratory
control samples (LCS), matrix spike (MS) and matrix
spike duplicates (MSD), and the use of sample surrogate
recoveries.
All method and instrument blanks met the
appropriate QC criteria, except for two method blanks
analyzed by Method 418.1. TPH was reported as
slightly above the PRL of 10 mg/kg in two method
blanks. Due to the low values reported in the two
method blanks, the sample results were not qualified.
Surrogate standards were added to all samples,
method blanks, MSs, and LCSs for the SW-846 Methods
8310 and 8020, and Method OA-1. All surrogate
recoveries for SW-846 Method 8020 were within the QC
acceptance criteria of 42 to 140 percent for soil. Seven
samples were qualified as estimated (J) and usable for
limited purposes based on surrogate recoveries for
Method OA-1. The QC acceptance criteria for surrogate
recovery for Method OA-1 was 67 to 127 percent.
Thirty soil samples for SW-846 Method 8310 analysis
were qualified as estimated (J) and usable for limited
purposes based on surrogate recoveries observed outside
the QC limits of 58 to 140 percent. Two surrogates
were used for Method 8310. Samples were qualified
only when both surrogates were outside the QC limits
and no dilution analysis was performed. Numerous soil
samples required dilution for the Method 8310 analysis
because of petroleum interference. Dilution of these
samples resulted in corresponding reductions in
surrogate concentrations. When this occurred, the
resultant concentration of surrogate was below its MDL.
In cases where dilution resulted in failure to detect the
surrogate, no coding of the data was implemented.
MS samples are samples to which a known amount
of the target analytes are added. There were 10 MSs
performed during the analysis by Method 418.1. Eight
of the MS samples were affected by high concentrations
of target analytes in the spiked samples. No samples
were qualified. Eleven MSs were performed during the
analysis by Method 8310. All but three MSs and MSDs
were outside the QC limits for percent recovery and
relative percent difference (RPD). These QC
exceedences were due to petroleum matrix interference.
The data associated with the QC samples was not
qualified because EPA guidelines state that samples
cannot be qualified based on MS and MSD results alone
(1991). There were seven MS and MSD samples
analyzed by Method 8020 and five by Method
OA-1. The MSs and MSDs analyzed by Method
8020 did not meet QC acceptance criteria for percent
recoveries or RPDs. No samples were qualified based
on these MS and MSD results due to the reasons stated
above. All MS and MSDs analyzed by Method
OA-1. and 90-3 Walkley-Black Method met all QC
acceptance criteria and were considered acceptable.
All LCSs met QC acceptance criteria and were
considered acceptable for all soil samples analyzed by
SW-846 Method 8310, Method OA-1, 90-3 Walk-
ley-Black Method, and Method 418.1. One soil LCS
11
-------
analyzed by Method 8020 was outside the QC control
limits. The soil LCS percent recovery for toluene was
below the QC limit. Twenty soil samples were qualified
as estimated (J) and usable for limited purposes.
Also, three equipment rinsate blanks and one trip
blank were analyzed to assess the efficiency of field
decontamination and shipping methods, respectively.
There was no contamination found above PRLs in any of
these type of blanks indicating decontamination pro-
cedures were adequate.
Confirmation of Analytical Results
Confirmation of positive results was not required by
any of the analytical methods performed except SW-846
Method 8310. Confirmation of positive PAH results by
Method 8310 was performed by the use of two types of
detectors. Both an ultraviolet detector and a
fluorescence detector were used in the analysis of PAHs.
The only requirement for using either detector for
quantisation was that they meet the QC criteria for
linearity (ICAL) and %D (CCAL). If either detector
failed either of these criteria, it could not be used for
quantitation, but it could be used for confirmation of
positive results.
Data Reporting
The results reported and qualified by the reference
method contained two types of qualifier codes. Some
data was coded with a "J," which is defined by PACE as
detected but below the PRL; therefore, the result is an
estimated concentration. The second code, "MI," was
defined as matrix interference. Generally, the effect of
a matrix interference is to reduce or enhance sample
extraction efficiency.
Quality Assessment of
Reference Laboratory Data
This section discusses the accuracy, precision, and
completeness of the reference method data.
Accuracy
Accuracy of the reference method was
independently assessed through the use of performance
evaluation (PE) samples purchased from Environmental
Resource Associates (ERA) containing a known quantity
of TPH. In addition, LCSs and past PE audits of the
reference laboratory were used to verify analytical
accuracy. Based on a review of this data, the accuracy
of the reference method was considered acceptable.
Precision
Precision for the reference method results was
determined by evaluating field duplicates, laboratory
duplicate, and MS and MSD sample results. Precision
was evaluated by determining the RPDs for sample
results and their respective duplicate sample results.
The MS and MSD RPD results for the PAH
compounds averaged 25 percent for all of the 11 MS and
MSD sample pairs. However, there was one MS and
MSD sample pair that had a RPD of 99.9 for
1-methyl-naphthalene. If this point was removed, the
overall average would decrease to 20 percent. The
average RPD for the seven BTEX MS and MSD sample
pairs was less than 25 percent. Only four BTEX RPDs
were outside advisory QC guidelines defined by the
PACE'S laboratory control charts. All five VPH MS
and MSD sample RPDs met advisory QC guidelines set
by the reference laboratory's control charts. The
10 TPH MS and MSD sample pairs were considered
acceptable.
Laboratory duplicate samples are two separate
analyses performed on the sample. During the analysis
of demonstration samples, 10 TPH laboratory duplicate
samples were prepared and analyzed. All TPH
laboratory duplicate RPD result values were less than
25 percent. This was considered to be acceptable.
Completeness
Results were obtained for all of the soil samples.
PACE J-coded values that were detected below the PRL,
but above the MDL. As discussed above, samples were
J-coded based on one or more of the advisory QC
guidelines not being met (i.e., surrogate and spike
recoveries)., Also, some samples were J-coded based on
BTEX CCALs not meeting QC guidelines. PRC did not
consider this serious enough to preclude the use of this
data because the %Ds for the CCALS did not exceed the
QC guidelines by more than 10 percent of the acceptable
range. The analytes with %Ds outside the QC guidelines
were not detected in most of the samples associated with
the CCALs. The J-coded data is valid and usable for
statistical analysis because the QC guidelines were based
on advisory control limits set by either the Method or by
PACE and this data set should be considered
representative of data produced by conventional
technologies. For this reason, the actual completeness
of data used was 100 percent.
Use of Qualified Data for Statistical Analysis
As noted above, 100 percent of the reference
laboratory results were reported and validated by
12
-------
approved QC procedures. The data review indicated
that J-coded data was acceptable for meeting the
demonstration objective of providing baseline data to
compare against die demonstrated technologies.
None of the QA/QC problems were considered
serious enough to preclude the use of J-coded data for
this demonstration. The surrogate and spike recovery
control limits were for advisory purposes only, and no
corrective action was required for the surrogate
recoveries that were outside of this range. RPD results
for MSs and MSDs that did not meet advisory QC
control limits were common when the matrix contained
a high concentration of petroleum. Again, these were
advisory limits and no corrective action was required.
These same general results would be seen by any
laboratory using the reference analytical methods on
such highly contaminated samples.
Also, rejection of a large percentage of data would
increase the apparent variation between the reference
laboratory data and the data from the technology. This
apparent variation would probably be of a similar
magnitude to that introduced by using the data. For
these reasons, the J-coded data was used.
Chemical Cross Sections
Chemical cross sections were created from the
reference analytical data produced for the qualitative
data evaluation (see Appendix A). These samples were
collected by a professional geologist on site during the
logging of boreholes. The cross sections were hand
contoured, and the contour intervals were selected to
best represent the range of contamination detected.
These cross sections were intended to represent a
conventional approach to the delineation of subsurface
contamination. The cross sections are presented on
Figures 3-1 to 3-6. A written interpretation of these
cross sections is presented below.
Atlantic Site
The five sampling nodes formed a northwest to
southeast trending transect across the site (Figure 3-1).
Node 1 on the far northwest edge of the cross section
represented an area that was not impacted by the
contamination from the Atlantic site. Just southeast of
this location at Node 2, two distinct layers of
contamination were identified. The upper zone extended
from approximately 1 foot to 5 feet below ground
surface (bgs). This zone was characterized by TPH
contamination ranging from 100 to 10,000 ppm. The
lower zone of contamination extended from
approximately 22 feet to 28 feet bgs. The TPH
concentrations in flu's zone ranged from 100 to greater
than 10,000 ppm. These two zones expanded and
blended together as Node 3 was approached. Around
Nodes 3 and 4 the thickness of the TPH plume remained
fairly constant, extending from approximately 1 foot to
31 feet bgs. The central portion of this zone exhibited
TPH contamination greater than 1,000 ppm. The
remainder of this zone exhibited TPH contamination in
the range of 100 to 1,000 ppm. As the far southeastern
edge of the transect was approached at Node 5, the
highest concentrations in the center of the plume pinched
out, leaving a contamination zone that extended from just
below the ground surface to approximately 27 feet bgs.
This zone exhibited contamination in the range of 100 to
1,000 ppm.
The total PAH cross section along this same transect
exhibited a slightly different distribution (Figure 3-2).
As with the TPH cross section, the total PAH cross
section began at Node 1 in an area exhibiting no signs of
contamination. At Node 2, again two zones of
contamination were detected. The upper zone extended
from the ground surface to approximately 7 feet bgs.
This 2:one deepened toward the east. The concentrations
of total PAHs in this zone ranged from 10 to greater
than 100 ppm. The lower zone extended from
approximately 14 to 30 feet bgs. The concentrations of
total PAHs in this zone ranged from 10 to greater than
100 ppm. Concentrations greater than 100 ppm were not
exhibited at this depth in the nodes occurring further
east. The distribution of the 10 to 100 ppm dipped
below the ground surface at progressive depths farther
east of Node 2. At Node 5, this upper zone began at
approximately 7 feet bgs. This zone also reached its
maximum depth around Nodes 3 and 4, approximately
30 feet bgs. Around Nodes 3 and 4 were two lenses of
total PAH contamination in excess of 100 ppm. The
largest of these zones appeared to be thickest around
Node 3, extending from approximately 7 to 16 feet bgs.
This zone thinned to the east and pinched out between
Nodes; 4 and 5. A smaller lens of greater than 100 ppm
total PAH contamination was exhibited at Node 4. This
zone extended between 7 to 9 feet bgs. This zone was
not detected in Nodes 3 or 5.
York Site
The five sampling nodes formed a north to south
trending transect. The TPH and total PAH distributions
appeared to be similar, except at Node 5, at the York
site (Figures 3-3 and 3-4). At Node 5, the TPH
contamination was more extensive, extending from 1 to
25 feet bgs. At this same location, the PAH
contamination extended from 13 to 21 feet bgs.
All of the nodes for this transect occurred in areas
that were impacted by the contamination associated with
13
-------
FIGURE 3-1. TPH REFERENCE METHOD CHEMICAL CROSS SECTION—ATLANTIC SITE
SOUTHEAST
NODES
0-
-4-
-3-
-8-
-8-
-9-
-10-
-11-
-14-
-13-
-18-
-17-
-18-
-19-
-20-
-21-
-22-
-23-
-I?:
-28-
-29-
-30-
-31-
-32-
-33-
-34-
-35-
-38 —
-37-
NODE+
NOOE3
NOOE2
DISTANCE (FEET)
NORTHWEST
NOOE1
::§
-5
-6
7
a
-»
-10
11
— 12
—13
-11
-IS
-10
17
-10
-IB
-22
23
--27
--21
--29
— 30
— 31
--32
--33
--34
--35
--35
--37
LEGEND
75
1< 100 PPM
• 100 - 1,000 PPM
125 1SO
1.000 - 10.000 PPM
> 10,000 PPM
175
225
ND - . NOT DETECTED
PPM - PARTS PER MILLION
27S
« - QUANTITATIVE
REFERENCE DATA
NOTE: QUANTITATIVE REFERENCE DATA USED BECAUSE OF POOR SAMPLE RECOVERY
FIGURE 3-2. PAH REFERENCE METHOD CHEMICAL CROSS SECTION—ATLANTIC SITE
SOUTHEAST
NODES
0-
-4
-5
-8
-10
-II
-12
-13
-14
-13
-18
-17
-18
-19
-20
-21
IE
-30-
-31-
-34-
-37-
NODE4
NOOE3
NODE2
NORTHWEST
NODE1
-0
—2
::?
—8
•-•
::j?
-14
—18
-17
-W
DISTANCE (FEET)
LEGEND
a< 10 PPM
l-_-_-J 10-100 PPM
H> 100 PPM
230
T - < 1 PPM
PPU - PARTS PER MILLION . - QUANITATIVE REFERENCE DATA
NOTE: QUANITATIVE REFERENCE DATA USED BECAUSE OF POOR SAMPLE RECOVERY
:»
—29
—30
—31
—32
—33
—34
—33
14
-------
FIGURE 3-3. TPH REFERENCE METHOD CHEMICAL CROSS SECTION—YORK SITE
NORTH
0-
-1-
-2-
-3-
-4-
-5-
-8-
-7-
-8-
-9-
-10-
-11-
-12-
-13-
-14-
-15-
-18-
-17-
-18-
-19-
-20-
-21-
-22-
-23-
-24-
-25-
-26-
NODE1
NODE2
NOOE3
NODE4
NODES
377
DISTANCE (FEET)
80
SOUTH
0
1
2
3
4
5
8
7
8
-»
10
11
12
13
14
15
18
17
18
It
20
21
22
23
24
25
26
120
LEGEND
< 10 PPM
> 10.OOO PPM
WpgiO - 100 PPM
NO - NOT DETECTED
r-_-_-|ioo - i.ooo PPM F
r%-— ~i t
PPM - PARTS PER MILLION
11.000 - 10,000 PPM
FIGURE 3-4. PAH REFERENCE METHOD CHEMICAL CROSS SECTION—YORK SITE
NORTH
o-
-1-
-2-
-3-
-4-
-5-
-8-
-7-
-8-
-e-
-10-
-11-
-12-
-13-
-14-
-15-
-18-
-17-
-18-
-18-
-20-
-21-
-22-
-23-
-24-
-25-
-28-
NODE1
NODE2
NOOE3
NODE4
NODES
DISTANCE (FEET)
SOUTH
-o
—1
—3
—4
--5
—8
—7
—8
—9
—10
—11
—12
—13
—14
—15
—18
—17
—18
—19
—20
—21
—22
—23
—24
—25
—28
80
70
8O
90
100
110
120
LEGEND
• < 10 PPM
;
•10-100 PPM
1OO - I.OOO PPM
> 1.000 PPM
T - < 1 PPM PPM - PARTS PER MILLION
NO - NOT DETECTED
15
-------
FIGURE 3-5. TPH REFERENCE METHOD CHEMICAL CROSS SECTION—FORT RILEY SITE
SOUTH
o-
-1-
-2-
-3-
-4-
-5-
-6-
-7-
-8-
-9-
-10-
-11-
-12-
-13-
-14-
-15-
-18-
-17-
-18-
-19-
-20-
-21-
-22-
-23-
-24-
-25-
-28-
-27-
-28-
-29-
-30-
-31-
NODE2
NODES
NODE3
NODE4
NORTH
-o
—i
—2
—3
—4
--5
—6
—7
—8
--9
—10
--11
—12
—13
—14
--15
—18
—17
—18
—19
—20
--21
—22
—23
—24
—25
—28
—27
—28
—29
—30
—31
DISTANCE (FEET)
ISO 180
LEGEND
< 10 PPM
> 10.000 PPM
gg$10 - 100 PPM
NO - NOT DETECTED
q 1.000 - 10.000 PPM
PPM - PARTS PER MILLION
FIGURE 3-6. PAH REFERENCE METHOD CHEMICAL CROSS SECTION—FORT RILEY SITE
SOUTH
0-
-1-
-2-
-3-
-4-
-3-
-«-
-7-
-8-
-9-
-10-
-11-
-12-
-13-
-14-
-15-
-18-
-17-
-18-
-19-
-20-
-21-
-22-
-23-
-24-
-25-
-28-
-27-
-28-
-29-
-30-
-31-
NODE2
NODES
NOOE3
NODE4
NORTH
•0
-1
-2
-3
-4
-5
•-8
-7
-I
-9
-10
-11
-12
-13
-14
-15
-18
—17
—18
•-19
—20
—21
—22
-23
-24
—25
—28
—27
—28
—29
—30
—31
' DISTANCE (FEET)
110
120
130
140
150
180
•170
180
LEGEND
2»P4<:10PPM r-_-_-lio - 100 PPM m±tt|ioo - i.ooo PPM
ND - NOT DETECTED PPM - PARTS PER MILLION
T - < 1 PPM
16
-------
this site. The contamination at this site appeared to
occur in a single band extending from approximately
10 to 22 feet bgs for total PAH and 2 to 25 feet bgs for
TPH contamination. This band of contamination thinned
from south to north across the transect. At the north end
of the transect, the TPH contamination thinned to a zone
extending from 12 to 19 feet bgs. Concentrations of
TPH in this zone ranged from 10 to 10,000 ppm, and the
concentrations of total PAH contamination range from
10 to 1,000 ppm. TPH contamination exhibited its
maximum concentrations in a lens around Nodes 3 and
4. This lens extended from approximately 12 to 16 feet
bgs, and exhibited TPH concentrations greater than
1,000 ppm. This lens tended to thin and deepen from
south to north. Node 4 exhibited the greatest TPH and
total PAH contamination. Two narrow lenses of total
PAH concentrations greater than 1,000 ppm existed at
approximately 14 to 16 feet bgs and 18 to 20 feet bgs.
A narrow lens of TPH contamination greater than
10,000 ppm was detected at approximately 18 to 19 feet
bgs at Node 4.
Fort Riley Site
The five sampling nodes formed a south to north
trending transect. The TPH and total PAH distributions
appeared to be similar at the Fort Riley site (Figures
3-5 and 3-6). Node 4, situated at the far north end of
the transect, was not affected by contamination. All of
the remaining nodes for this transect occurred in areas
that were impacted by the contamination associated with
this site. The contamination at this site appeared to
occur in a single zone extending from approximately
1 to 25 feet bgs for total PAH and 0 to 30 feet bgs for
TPH contamination. This zone of contamination
exhibited relatively constant thickness across Nodes
2, 3, and 5. Concentrations of TPH in this zone ranged
from 100 to greater than 10,000 ppm, and ithe
concentrations of total PAH contamination ranged from
10 to 300 ppm. Total PAH contamination exhibited its
maximum concentrations hi a lens around Nodes
2, 3, and 5. This lens extended from approximately 5 to
8 feet bgs at Node 3, becoming thicker and deeper at
Node 2 where it extended from 10 to 20 feet bgs. The
TPH concentrations exhibited two lenses of
concentration at greater than 10,000 ppm. These lenses
did not appear to be extensive and their occurrence was
limited to the areas around single nodes. Node 5 in the
center of the transect exhibited one of these lenses of
highest TPH contamination extending from 10 to 13 feet
bgs. Node 2 has the other such lens which extended
from 17 to 19 feet bgs.
Quality Assessment of
Geotechnical Laboratory Data
This section discusses the data quality of the
geotechnical laboratory results, the data quality of the
borehole logging conducted by the on-site professional
geologist, and the soil sampling depth control.
Geotechnical Laboratory
Soil samples submitted for textural determination
were analyzed by ASTM Method D-422 (1990). ETS,
Petaluma, California, conducted these analyses. ASTM
Method D-422 does not define specific QAXQC criteria,
however, it specifies the use of certified sieves, and
calibrated thermometers and hydrometers. ETS
followed the approved method and complied with all the
equipment certification and calibration requirements of
the method. Based on this, the geotechnical data was
detennined acceptable.
Borehole Logging
The data quality of the on-site professional
geologist's borehole logs was checked through on-site
audits by a soil scientist, and by comparison of the
geologist's descriptions for intervals corresponding to
samples analyzed by ASTM Method D-422. This
comparison is discussed later hi this report.
Sampling Depth Control
At each site, random checks of the reference
sampling intervals were made. These checks consisted
of stopping drilling operations just before inserting the
split spoon sampler into the hollow stem auger to collect
samples. At this tune, a weighted tape measure was
used to measure the top of the sampling interval. The
measurement was checked against the intended sampling
depth. If the difference between the intended and actual
sampling depth had varied by more than 1 inch, the
borehole would have been redrilled. Depth checks were
made at a minimum of once per sampling day. None of
these depth checks resulted hi data exhibiting a greater
than 1 inch difference between intended and actual
sampling depth. Based on this, the reported sample
intervals were considered accurate.
Stratigraphic Cross Sections
Stratigraphic cross sections were based on the data
produced by a professional geologist during the logging
of boreholes during the demonstration. The cross
17
-------
FIGURE 3-7. REFERENCE METHOD STRATIGRAPHIC CROSS SECTION—ATLANTIC SITE
SOUTHEAST
NODES
0-
-I-
-4-
-S-
-0-
-7-
-n-
-9-
-10-
-II-
-14-
-13-
-IS-
-17-
-1B-
-19-
-20-
-21-
-21-
-23-
-24 —
-25-
-28-
-27-
-28-
-29-
-30-
-31-
NODE4
NODE3
NODE2
NORTHWEST
NODE1
-1
-2
-3
-4
-6
-8
-7
-8
-9
.-10
-11
•-12
•-13
•-14
-15
-IB
-17
-18
•-19
-20
-21
-22
-23
—24
—25
—28
-27
;LOAM)
;LOAM)
(SAND)
(CLAY
.LOAM;
DISTANCE (FEET)
--30
—31
—32
—33
—34
—35
--38
—37
125
150
LEGEND
; SILTY CLAY (CH>
SILTY CLAY (CL)
SILT
SILTY SAND
250 275
WELL GRADED SAND
POORLY GRADED SAND
( ) - LABORATORY CLASSIFICATION (USDA)
sections were intended to represent a conventional
approach to the delineation of subsurface stratigraphy.
The cross sections are presented on Figures 3-7 to
3-9. A verbal interpretation of these cross sections is
presented below by site. QA/QC consisted of the
collection of samples for textural analysis by a
geotechnical laboratory (see Appendix A). These
samples are discussed at the end of each site-specific
discussion.
Atlantic Site
The Atlantic site is located on the flood plain of the
Nishnabotna River, which is located about 0.7 mile west
of the site. The flood plain is nearly level. The surface
soil at the site is a silty clay. These soils have most
likely formed from alluvium. Stratigraphic cross
sections produced during the demonstration from soil
borings indicated that the subsurface soil at the site
consisted of silts and clays and silty clay interfingered
with each other to a depth of approximately 21 feet bgs
on the northwest end (Node 1) and to 28 feet bgs on the
southeast (Node 5). See Figure 3-7 for a graphical
representation of the cross section. A layer of sand was
present from 18 to 36 feet bgs at Node 2. This zone
remained relatively uniform from Node 2 to Node 5.
Seven soil samples were collected to verify the
geologist's borehole logging at the Atlantic site. The
geologist's classification of soils matched the
geotechnical laboratory's classifications six out of seven
tunes (Table 3-1). This one point of disagreement was
sample DR16 from the 2- to 3-foot interval at Node 1.
In this sample, the geologist identified silt as the
predominant size fraction of the sample, while the
geotechnical laboratory identified clay as the
predominant size fraction. This is a common point of
variance between field soil classification and laboratory
classification. These common differences are magnified
in grossly contaminated soils, and when the geologist is
forced to wear plastic gloves during classification. This
difference was noted, and geologist's Stratigraphic
borehole logs meet the demonstration DQOs for
screening level data.
York Site
The York site is located on the flood plain of Beaver
Creek, which is located 0.1 mile southwest of the site.
The site is situated on a nearly flat lying terrace above
the river. The surface soils is a silt loam. These soils
most likely formed hi alluvium on stream terraces. A
Stratigraphic cross section based on soil borings during
the demonstration was prepared (Figure 3-8). The top
18
-------
FIGURE 3-8. REFERENCE METHOD STRATIGRAPHIC CROSS SECTION—YORK SITE
NORTH
0
-1
-2
-3
-4
-5
-6
-7
-8
-9
-10
-11
-12
-13
-14
-15
-15
-17
-18
-19
-20
-21
-22-
-23-
-24-
-25-
-26-
NODE1
NODE2
NODE3
NODE4
NODES
DISTANCE (FEET)
LEGEND
50 80
CLAYEY SILT ft SILT (ML)
WELL GRADED SAND
SILTY CLAY (CL)
POORLY GRADED SAND
I
SOUTH
-0
—1
—2
--3
—4
--5
--8
—7
—8
—9
—10
—11
—12
—13
—14
—15
—18
—17
—18
—19
—20
--21
--22
—23
--24
—25
--26
120
SILT
SILTY CLAY (CH)
( ) - LABORATORY CLASSIFICATION (USOA)
FIGURE 3-9. REFERENCE METHOD STRATIGRAPHIC CROSS SECTION—FORT RILEY SITE
SOUTH
0-
-1-
-2-
-3-
-4-
-5-
-6-
-7-
-8-
-9-
-10-
-11-
-12-
-13-
-14-
-15-
-16-
-17-
-18-
-19-
-20-
-21-
-22-
-23-
-24-
-25-
-28-
-27-
-28-
-29-
-3O-
-31-
NODE1
NODE2
NODES
NODE3
NODE4
(LOAM)
(SANDY
LOAM)
NORTH
-I
-2
-3
-4
•-5
•-6
-7
-8
•-»
-10
-11
-12
-13
-14
-15
-16
—17
—IB
—19
—20
—21
—22
-23
-24
-25
—26
-27
-28
-29
—30
DISTANCE (FEET)
100 110 120 130 140 150 160 17O 180 190- 200 210 220
LEGEND
SILTY CLAY (CL)
SILTY CLAY (CH)
( ) - LABORATORY CLASSIFICATION (USDA)
. WELL GRADED SAND
POORLY GRADED SAND
19
-------
TABLE 3-1. COMPARISON OF GEOLOGIST DATA AND GEOTECHNICAL LABORATORY DATA-
ALL SITES
Site
Geologist Classification
Geotechnical Laboratory Classification
Match
Atlantic Silty Clay (ML)
Clayey Silt (ML)
Silt (ML)
Well Graded Sand (SW)
Clay (CL)
Silty Clay (CL)
Silty Clay (CL)
York Clayey Silt (ML)
Silty Clay (CL)
Well Graded Sand (SW)
Poorly Graded Sand (SP)
Clayey Silt (ML)
Sand (SW)
Fort Riley Poorly Graded Sand (SP)
Fill
Poorly Graded Sand (SP)
Clayey Silt (ML)
Poorly Graded Sand (SP)
Poorly Graded Sand (SP)
Poorly Graded Sand (SP)
Well Graded Sand (SW)
Sandy Lean Clay (CL)
Clay or Silt (CL or ML)
Silt or Clay (ML or CL)
Well or Poorly Graded Sand (SW or SP)
Sandy Lean Clay or Sandy Silt (CL or ML)
Sand Lean Clay or Sand Silt (CL or ML)
Silt or Clay (ML or CL)
Silt or Clay (ML or CL)
Silt or Clay (ML or CL)
Silty to Clayey Sand (SM or SC)
Poorly Graded Sand with Silt or Clay (SW-SC or
Silt or Lean Clay (CL or ML)
Silty or Clayey Sand (SM or SC)
Silty or Clayey Sand (SM or SC)
Silty or Clayey Sand (SM or SC)
Silty or Clayey Sand (SM or SC)
Silty or Clayey Sand with Gravel (SC or SM)
Silty or Clayey Sand (SM or SC)
Silty or Clayey Sand (SM or SC)
Silty or Clayey Sand (SM or SC)
Silty or clayey sand (SM or SM)
Noa
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Noa
Yes
Yes
Noa
No
Yes
Noa
No
Noa
Noa
Noa
Noa
Notes:
These failures to match were due to the geologist underestimating the percentage of fines in the sample.
Unified Soil Classification System two-letter code.
1 to 2 feet of the cross section was fill. From 2 to 14
feet bgs, the cross section consisted of clayey silt with
some lenses of silty clay and silt. At approximately 14
feet bgs, there were thick lenses of silt, sandy silt, and
sand. These lenses were approximately 7 feet thick and
were interfingered with each other. At approximately 21
feet bgs, the material became primarily well graded sand
to the bottom of the section at 25 feet bgs.
Six soil samples were collected to verify the
geologist's borehole logging at the York site. The
geologist's classification of soils matched the
geotechnical laboratory's classifications four out of six
times (Table 3-1). The two points of disagreement were
samples DR27 (Node 1, 15 to 15.5 feet bgs) and DR 29
(Node 3, 12 to 13 feet bgs). In both cases, the geologist
underestimated the percentage of silt and clay size
20
-------
particles hi the samples. This is a common point of
variance between field soil classification and laboratory
classification. These differences are magnified in
grossly contaminated soils, and when the geologist is
forced to wear plastic gloves during classification
activities. The variances described above are not
uncommon hi environmental studies, and thus, the
geologist's stratigraphic borehole logs, while exhibiting
some disagreement with the laboratory data, are
considered to meet the demonstration DQOs for
screening level data.
Fort Riley Site
The Fort Riley site is located on the flood plain of
the Kansas River, which is located 0.1 mile southeast of
the site. The site is situated on a nearly flat lying terrace
above the river. The surface soil is a silt loam. This
soil is most likely formed from deep alluvium. A
stratigraphic cross section based on soil borings
conducted during the demonstration is presented on
Figure 3-9. This cross section showed typical deposition
hi an alluvial setting with interfingered beds of clay, silt,
silty clay, clayey silt and sand. In the center of the cross
section, the top 8 feet was fill. The northern and
southern edges of the cross section were silt or silty clay
at the surface. In the northern half of the cross section,
poorly graded sand was present from 5 to 17 feet bgs.
In the southern half of the cross section from 8 to
approximately 18 feet bgs, the cross section consisted of
interfingered lenses of clay, silty sand, and sand. Below
20 feet bgs, the cross section became primarily sand
with silt and clay lenses of various thickness intermixed
to the terminal depth of the cross section.
Eight soil samples were collected to verify the
geologist's borehole logging at the Fort Riley site. The
geologist's classification of soils matched the
geotechnical laboratory's classifications one out of eight
times (Table 3-1). Both classifications did correctly
identify the dominant particle size fraction. In all of the
cases of disagreement, the geologist underestimated the
percentage of silt and clay size particles. Small shifts hi
the estimation of these particles can alter the descriptive
modifier used hi classification. The variances described
above affect the accuracy of the reference stratigraphic
cross sections as far as the secondary classification
modifiers are concerned. The baseline classification as
to the dominant particle size is accurate. This data met
the demonstration's DQOs, however, decisions based
solely on differences in classification modifiers should be
qualified as semiqualitative.
21
-------
Section 4
Site Characterization and Analysis Penetrometer System
This section describes the SCAPS sensors that were
evaluated during this demonstration. The description
provided is based on information provided by the
developer, on information PRC obtained from reports
and journal articles written about the technology, and on
observations made during the demonstration. The
description includes background information on the
technology and its components, general operating
procedures, training and maintenance requirements, and
the cost of the technology discussed.
Background Information
The SCAPS LIF sensor was developed by the Army
(U.S. Army Corps of Engineers, Waterways Experiment
Station [WES] and the Army Environmental Center
[AEC]), Navy (Naval Command, Control and Ocean
Surveillance Center), and the Air Force (Armstrong
Laboratory). This system uses laser light to cause
compounds in soils to fluoresce and measures the
resulting fluorescence. Currently, this technology is
most commonly used to detect PAH compounds
associated with petroleum fuels. The U.S. Army holds
a patent for this combination of a sapphire window and
cone penetrometry. The LIF sensor was modified from
a design developed by the Navy for use in detecting
petroleum, oils, and lubricants in seawater.
The SCAPS CP sensor is a standard sensor
commercially available.
Components
This section describes the components of the SCAPS
LIF and CP system, which consists of a cone
penetrometer truck, modified CP, sampling tools, a
nitrogen (N^ laser, and a fluorescence detection system.
Cone Penetrometer Sensor
A complete CP system consists of a truck, hydraulic
rams and associated controllers, push rods, samplers,
and the CP sensor itself. The weight of the truck
provides a static reaction force, typically 20 tons, against
which the hydraulic system works to advance
1-meter-long segments of 3.57-centimeter-diameter
threaded push rod into the ground. The CP, which is
mounted on the end of the series of push rods, contains
sensors that continuously log tip stress and sleeve
friction. The data from these sensors is used to map
subsurface stratigraphy. Conductivity or pore pressure
sensors can be driven into the ground simultaneously
with the tip resistance and sleeve friction sensors. The
conductivity and pore pressure sensors are used to
further define subsurface stratigraphy.
Soil, groundwater, and soil gas sampling tools can
also be used with the CP system. These capabilities are
discussed in greater detail hi the general ITER.
Generally, sampling tools and sensors cannot be used
concurrently.
In favorable stratigraphies, push depths of 50 meters
or greater have been achieved. The CP can be pushed
through asphalt, but concrete must be cored prior to
advancing the CP. Advancing sensors and sampling
tools with the cone penetrometer truck may be difficult
in the following subsurface environments:
Gravel units
Cemented sands and clays
Buried debris
Boulders
Bedrock
The cone penetrometer truck used with the SCAPS
sensors is fitted with a steam cleaner to decontaminate
the push rods as they are withdrawn from the ground.
The decontamination water is contained in the
decontamination apparatus and it can be directly
discharged into a storage container. In addition, the
combination CP and LIF sensors used in the SCAPS is
modified to provide automatic grouting of die CP hole
during the retraction of the push rods. The decontami-
22
-------
nation water, pressure sprayer, and grouting pump are
mounted in a trailer that can be towed behind the cone
penetrometer truck.
The SCAPS system is mounted on a specially
engineered 20-ton truck designed with protected work
spaces which provide additional health and safety
protection to SCAPS workers at hazardous waste sites.
LIF Sensor
The SCAPS LIF sensor's main components are the
N2 laser, fiber optic cable, and the fluorescence
detection system, and the computer system. The N2
laser creates laser light of a known wavelength. The
laser light passes along a fiber optic cable and into the
soil through a sapphire window, 2 millimeter (mm) thick
and 6.35 mm in diameter, mounted 65 centimeters (cm)
above the terminal end of the CP probe hi which it is
mounted. Induced fluorescence from the soil is returned
to the fluorescence detector along a second fiber optic
cable. The fiber optic cables are all silica fiber optic
cables, 365 micrometers 0/m) hi diameter. A
photodiode array (PDA) and optical multichannel
analyzer (OMA) is used as the fluorescence detector,
and the data is processed by a computer system. The
return fluorescence data and soil stratigraphy data (from
the CP) are collected and interpreted by the same
computer system. A diagram of the SCAPS sensor
configuration is shown on Figure 4-1.
To operate the SCAPS sensors, the cone
penetrometer truck must be positioned over a designated
penetration point. At this tune, the LIF sensor's
response is checked using a standard rhodamine solution
held against the sapphire window. This procedure is
carried out before and after each push. The CP and LIF
sensor are then advanced into the soil at a rate of
2 centimeters per second (cm/s) or approximately 4 feet
per minute.
The LIF sensor is operated with a N2 laser that
provides excitation pulses at a rate of 10 pulses per
second (Hz). The PDA accumulates the fluorescence
emission response over 10 laser shots, and then an
emission spectrum of the soil fluorescence is retrieved
from the PDA by the OMA and computer system.
Therefore, at the data acquisition rate of 10 Hz and a
penetration rate of 2 cm/s, the spectral resolution of the
LIF detection system under these operating conditions is
2 cm. The fluorescence intensity at peak emission
wavelength for each stored spectrum is displayed hi real
tune on a panel plot, which also includes the soil
classification data from the CP sensor (Figure 4-2).
This sensor is described hi detail in the general ITER.
LIF Sensor Components
The mam SCAPS LIF sensor components are:
« N2 laser
« Fiber optic cable (365 pm diameter) and
modified CP fitted with a sapphire window
• Fluorescence detection system
« Computer system
Each SCAPS LIF sensor component is discussed hi
more detail below.
N2 Laser
Laser radiation excitation is produced by a pulsed
nitrogen laser made by Photon Technology, Inc. (PTI).
The laser produces light at a wavelength of 337 nano-
meters (nm) with an intensity of approximately 1 mega-
joule (Mj). The emitted laser radiation is focused
through a lens and directed into the excitation fiber.
Fiber Optic Cable
Each laser pulse is focused through a lens and
directed into an Ensign-Bickford hard coat, all-silica
optical fiber with a core diameter of 365 jum. The
core/cladding diameter is approximately 400 /um. The
optical fiber along with a return fiber (same
specifications), instrumentation cables, and a grout line
are all protected by a neoprene shrink tubing jacket
forming the sensor umbilical, which is passed through
the center of each push rod. The transmit fiber is
terminated at a 2-mm-thick, 6.3-mm-diameter sapphire
window, which is coated with an anti-reflective material
to reduce 337 nm light backscatter into the return fiber.
This sapphire window is removable to facilitate periodic
replacement as necessary. The sapphire window passes
the laser light onto the soil surface adjacent to the
window. The fluorescence signature of the soil is
returned by another optical fiber with the same
specifications of the transmit fiber. The return fiber
passes the returned light into the monochromator
(EG&G Princeton Applied Research Company [PARC],
Modd 1229 Monochromator).
Fluorescence Detection System
When return fluorescence travels up through the
return fiber, it first enters the monochromator. The
monochromator contains mirrors and a grating so that
23
-------
r
FIGURE 4-1. TRI-SERVICES SCAPS
Space Far Extra
Proboa
FIGURE 4-2. SCAPS PANEL PLOT - NODE 4 ATLANTIC SITE
Cone Resistance
qc (tons/ft')
1 100
1000
SlMv* Friction
f, (tons/ft1)
0 Z 4 6 B 10
CPT based SOIL
CLASSIFICATION
Ss.S 2
*> 4J » 4J a a o
e •-• c C c
O 300 1000 1300 MOO
«r«it*nc* Inttntuv Hcvtltngttl
=•«"» - SMB1» •«« at Piak
-------
the returned light is diffracted into its component
wavelengths. The light then exits the monochromator
and enters an EG&G PARC Model 1421B-1024-G
intensified silicon PDA detector, which is attached
directly to the monochromator. The detector is capable
of being gated and provides a blue to mid-spectrum
response using a 1024 element array. The intensity of
the returned light causes the internal diodes to produce
an electrical signal directly proportional to the intensity
of the incident light for each of the 1024 elements of the
PDA. Each element corresponds to a particular
wavelength. The PDA detector is controlled by a
EG&G PARC Model 1460 OMA. The OMA receives
the data and displays the spectral signature of the
returned signal. The OMA can be used as a stand-alone
processor using its display and touch screen technology
to control the detector and communicate with external
devices. However, the system also can be controlled by
an external computer via a GPID interface.
Computer System
The computer system is comprised of two Hewlett
Packard 486DX33 Vectra computers. One computer is
used as the data acquisition computer and the second
computer is used for post-acquisition processing. The
data acquisition computer is used to communicate and
transfer data from the OMA and record measurements of
soil stratigraphy. Both the soil classification and LIF
sensor response are displayed in real time during the
advancement of the CP. Once the push is completed,
the data is transferred (through a local area network) to
the post-processing computer where the data is
manipulated and plotted. It should be noted that
although normal sensor data consists of the fluorescence
intensity response at peak emission wavelength, SCAPS
LIF sensor is configured to collect and store the entire
fluorescence spectrum from approximately 300 to
800 nm.
General Operating Procedures
Four people are needed to operate the SCAPS as
currently deployed. The crew chief, the LIF sensor
operator, the hydraulic ram operator, and the rod
handler. The crew chief is an experienced engineer that
plans and manages the total deployment of the SCAPS.
This involves predeployment and post-deployment
logistics, push rod decontamination, and grouting. The
actual collection of data on site is handled by the three
other crew members. The hydraulic ram operator
operates the hydraulics of the cone penetrometer truck
and monitors CP depth and soil stratigraphy data. The
rod handler screws the push rods into place as the CP is
advanced. The LIF sensor operator monitors both the
LIF sensor response and the soil classification data as the
push is executed. The LIF sensor operator also handles
the post-acquisition data processing between penetration
events, and produces the final chemical and physical
characterization reports.
Cost
The SCAPS LIF and CP sensors are commercially
available. However, there are a number of SCAPS units
currently available to various Federal agencies under
cooperative work agreements with the U.S. Army Corps
of Engineers. WES has produced five SCAPS units:
one for research, three for the U.S. Army Corps of
Engineers, and one for the Department of Energy. The
Navy has produced two units for deployment and one for
research use. Similar LIF and CP technology is
available from either Hogentogler or Applied Research
Associates, Inc., both of which have non-exclusive
licenses from WES to use LIF technology with cone
penetrometry.
WES has produced an operations manual for the
SCAPS and has limited training for U.S. Army Corps of
Engineers SCAPS operators.
Currently, WES estimates the daily rate for use of
the SCAPS LIF and CP sensors would be $3,500. This
cost represents operating costs. The cost does not
include normal resources associated with commercial
application, such as marketing, research, and profit.
Mobilization and operator per diem costs are included in
the daily rate. Based on the daily use charge of the LIF
and CP sensors, a total cost of approximately $20,000
was realized for the three site characterization activities.
This cost includes the initial mobilization and the
subsequent inter-site mobilization required for two days
of travel. The data was generally generated in two days
at each site and a total of two days of travel between all
3 sites was used. For comparison, the predemonstration
activities used conventional field screening and produced
similar data at the three sites; however, it required more
personnel and on-site analytical capabilities. The
approximate three site characterization cost was
$43,000. This effort resulted hi fewer data points,
relative to the continuous data output of the SCAPS
sensors. In addition, the predemonstration activity only
produced one borehole log at each site. Another cost
comparison can be made relative to the costs accrued
producing the reference cross sections for this
demonstration. The reference cross sections cost
approximately $55,000, including approximately
$30,000 for drilling services, approximately $8,000 for
an on-site geologist and a sample, approximately
$12.,000 for off-site analytical services, and
approximately $5,000 for handling and disposal of
investigation derived waste.
25
-------
Observations
Observations recorded during the demonstration of
the SCAPS LIF and CP sensors are briefly summarized
below.
The SCAPS conducted a total of 30 grid pushes and
6 non-grid pushes during the demonstrations at the three
demonstration sites. The following discussion reflects
observations made by Dr. Harry Ellis of PRC during the
demonstration of the SCAPS. Dr. Ellis did not operate
the SCAPS equipment and required no training prior to
the demonstration. Because WES was responsible for
operations, Dr. Ellis was an observer only.
The crew operating the SCAPS unit was a
developmental group, rather than a general operational
crew. The only difference between these crews involved
the number of personnel. The developmental group
crew included a dedicated person to post-process the
data.
In some cases, the size of the SCAPS truck made
on-site maneuvering in confined spaces difficult.
Once at a demonstration site, it took 3 to 4 hours to
convert the SCAPS unit from a road travel mode to
operating mode. This included unpacking the
computers, LIF sensor, and other sensitive components,
connecting and testing these components, connecting the
trailer to the truck, and so on. Moving about the site
from one push location to another required no
adjustments except lifting up the access ladder. The
decontamination and grouting trailer can be moved
separately from the truck hi close quarters. If this is
done, the connecting lines will usually have to be
disconnected and reconnected which takes a few
minutes. Demobilization in preparation for road
movement to a new site takes about 2 hours.
Based on the progress of work during the
demonstration, it is possible to estimate the speed of
operations. The average push was 35 feet below grade.
These estimates are as follows:
• About 1 hour per day for watering, fueling, and
minor maintenance
• Approximately 1 to 1.5 hours per 35 foot push.
This includes all operations from placement on
location to placement on the location for the
next push. This includes truck movement
between points, rod advancement and
withdrawal, grouting, and decontamination.
If the deepest pushes of the demonstration are
considered, 75 feet below grade, it would add 10 to
15 minutes per push and require more frequent filling of
the clean water storage tank. This suggests that
additional depths can be achieved with minimal impact
on throughput. The automatic decontamination and hole
grouting while the push rod is being withdrawn are
advantages of this technology and greatly increase
sample throughput.
Rock, debris, and similar items downhole may stop
advancement of the push rods and sensors. It is
necessary to have a good idea of the actual subsoil
conditions before estimating production rates for the
SCAPS at a given site. The terrain can limit the use of
the SCAPS. It needs about 20 feet of overhead
clearance. Side slopes and rough terrain can limit its
use. The leveling jacks can compensate within limits.
The N2 laser used by the SCAPS LIF sensor
consumes nitrogen gas. Currently, the nitrogen gas
cylinders are mounted on the decontamination and
grouting trailer. Whenever the clean water tanks on the
trailer need refilling, operation of the LIF sensor is
stopped because the nitrogen source for the N2 laser is
attached to the decontamination and grouting trailer.
This potential downtime could be decreased by mounting
the nitrogen cylinders on the cone penetrometer truck
itself rather than the trailer. However, this would create
an ergonomic problem of lifting these heavy items into
position. If this were done, the trailer could be taken for
water refill during a calibration and push, and part of the
1 hour per day for replenishment would be eliminated.
Normal wear and tear does slow down operations.
Two such items were noted during this demonstration.
The fiber optic cable hi use during initial pushes was
nearing the end of its useful life (about 150 pushes is
estimated). This made it difficult to achieve optimal
operating conditions during the daily ICALs. Also, the
jaws which grip the probe rod were worn and caused
intermittent problems with probe withdrawal due to
internal slippage.
Of the 36 pushes done during this demonstration,
three resulted in catastrophic unit failure. This equates
to an 8 percent catastrophic failure rate. Catastrophic
failure either resulted in the physical loss of the CP and
LIF sensor, LIF sensor down-tune in excess of 8 hours,
or the disabling of one of the SCAPS components.
Specifically:
• While at Grid 5 at the York site, the grout
pump seized due to concrete clotting in the
26
-------
helical pump. This pump was original
equipment (about 5 years old). Hand grouting
was used until the pump could be temporarily
repaired. This caused a work delay of
approximately 4 hours and an added 1 to
1.5 hours to the completion tune for each
subsequent push due to the tune associated with
hand grouting.
• During the first nongrid push at the York site,
the probe stopped producing a response. The
probe was pulled, and neither a post calibration
or a flashlight shining directly into the sapphire
window would elicit any response from the
fluorescence detector. The OMA appeared to
be functioning normally, so it was concluded
that the fiber optic cable had broken.
Therefore, the crew rigged a new probe and
umbilical and resumed the push. The old
umbilical was returned to WES for repair. The
repair was estimated to cost $2,000 and was
expected to require a day's labor from two
instrumentation technicians. This cable break
resulted hi a work delay of approximately
2 hours. This delay was minimized by the fact
that the SCAPS is deployed with a second
sensor, and umbilical cord which is already
threaded through a second set of push rods.
• During the last nongrid push at the Fort Riley
site, the sensor array was lost downhole due to
push rod breakage during retrieval. The broken
end of the push rods that were retrieved
exhibited a fracture along the male threads.
The primary maintenance practice with the SCAPS
LIF and CP sensors is to inspect and repair as necessary.
It will be useful to accumulate the experience necessary
to predict the useful life of various SCAPS components
and set up a more detailed schedule for overhaul or
replacement of components.
Data Presentation
To qualitatively assess the abilities of the SCAPS
CP sensor in identifying the subsurface textural
properties of a site, it was required to collect soil texture
data during its advancement at each of the five sample
nodes at each site. The nodes were arranged hi a
transect line across a known area of subsurface soil
contamination identified during predemonstration
sampling and previous investigations conducted at each
site. Sampling at a node was continuous from the
surface of the soil a depth of 50 feet.
The soil texture data generated by the technology
was used to produce soil texture cross sections along
each transect line. A comparison of its data to that of
the reference methods is discussed in Section 5. The
following sections present chemical and physical data for
the SCAPS site.
Chemical Data
Two types of SCAPS data are presented hi this
section. The data used in the qualitative evaluation is
presented and discussed as cross sections. The SCAPS
data used in the quantitative evaluation are listed hi
Tables 4-1, 4-2, and 4-3. This data is discussed hi
Section 5.
The SCAPS LIF logs are used to describe the
relative distribution of subsurface contaminants and
produce contaminant cross sections. Although the
SCAPS produced its own cross sections, PRC
transferred the data and plotted it on a scale that matched
the ones used for the reference method. The
transformed SCAPS chemical cross sections are
presented on Figures 4-3, 4-4, and 4-5. An example of
the standard SCAPS LIF graphic is shown on Figure
4-6.
The LIF sensor data was reported as intensity at the
peak wave length. Theoretically, changes hi intensity
relative to background can be used to assess relative
changes hi the concentration of subsurface fluorescing
materials. In theory, as the LIF sensor intensity
increases, the concentration of contaminants
(fluorescing) also may increase. One objective of this
demonstration was to evaluate this relationship. The
following data presentation was produced by the SCAPS
operator and represents a typical narrative data
evaluation provided by SCAPS. These narratives can
often be generated hi the field within 24 hours of data
acquisition.
Atlantic Site
The standard operating procedures for the SCAPS
LIF sensor data interpretation include review of panel
plots. These plots include soil stratigraphy, fluorescence
peak intensity, and wavelength at peak intensity. An
example panel plot is seen on Figure 4-2. Based on the
fluorescence response at various depths, the fluorescence
emission spectra for a particular depth was inspected to
determine if different contaminants were present.
At the Atlantic site, the LIF sensor response
indicated that sampling Node 1 was at background
fluorescence levels. Sampling Node 2 showed the
presence of a fluorescing contaminant at 20.5 to 24 feet
27
-------
TABLE 4-1. QUANTITATIVE EVALUATION DATA FOR THE ATLANTIC SITE
Node
2
2
3
4
4
4
5
5
Depth (feet)
21-22
24-25
16-17
6.5 - 7.5
10-11
27.5 - 28.5
16-17
23.5 - 24.5
Number of
Readings
5
6
5
5
5
5
5
6
Maximum
Fluorescence
Readinq
9,078.00
10.33
1,638.0
425.2
1,195.0
13,623.0
21,225.0
36.25
TABLE 4-2. QUANTITATIVE EVALUATION DATA
Node
2
2
3
4
4
4
5
Depth (feet)
15-16
13.5 - 14.5
17-18
17-18
14-15
18-19
1.5-2.5
Number of
Readinqs
5
5
5
5
5
6
6
Maximum
Fluorescence
Readinq
331.9
1,154.0
923.2
1,527.0
1,526.0
992.8
313.9
TABLE 4-3. QUANTITATIVE EVALUATION DATA
Node
1
1
2
2
5
5
Depth (feet)
2-3
13-14
6-7
17-18
10.5-11.5
16-17
Number of
Readings
5
6
5
6
5
6
Maximum
Fluorescence
Readinq
1,979.0
453.7
1,787.0
12,561.0
4,289.0
3,684.0
Minimum
Fluorescence
Reading
4,058.0
1.50
682.71
35.33
543.7
1,127.0
13,557.0
25.20
FOR THE YORK
Minimum
Fluorescence
Reading
118.1
493.9
33.83
518.1
219.9
166.4
275.5
FOR THE FORT
Minimum
Fluorescence
Readinq
306.1
265.7
54.82
1,245.0
1,731.0
1,620.0
Mean
5,809.0
5.37
1,323.0
213.4
837.1
6,310.0
19,574.0
29.88
SITE
Mean
221.3
723.8
268.5
908.4
775.7
412.5
297.1
RILEY SITE
Mean
1,143.0
366.6
1,136.0
4,853.0
2,923.0
3,036.0
Standard
Deviation
2,007.0
3.38
391.5
154.2
326.6
5,765.0
3,420.0
3.73
Standard
Deviation
93.15
263.6
370.1
388.9
642.6
311.1
12.91
Standard
Deviation
761.6
76.21
648.7
4,147.0
990.0
1,402.0
28
-------
FIGURE 4-3. SCAPS CHEMICAL CROSS SECTION—ATLANTIC SITE
SOUTHEAST
NODES
0-
-1-
-2-
-3-
-4-
-5-
-6-
-7-
-8-
-9-
-10-
-11-
-12-
-13-
-14-
-15-
-16-
-17-
-18-
-18-
-20-
-23-
-24-
-25-
-28-
-27-
-28-
-29-
-30-
-31-
-32-
-33-
-34-
-35-
-35-
-37-
NODE4
NODE3
NODE2
NORTHWEST
NODE1
-0
—1
—2
—4
--5
—«
—7
--»
--10
—11
::!§
—14
--15
—16
—17
—18
—19
—20
--21
—22
--23
—24
—25
--28
"27
—28
—2«
—30
--31
—32
—33
—34
--35
—38
—37
DISTANCE (FEET)
150
250 275 300
LEGEND
30 - 100 (COUNTS)
.-_-_-] 100 - 1.OOO
] 1.000 - 10.000
FIGURE 4-4. SCAPS CHEMICAL CROSS SECTION—YORK SITE
NORTH
0-
-1-
-2-
-3-
-4-
-5-
-8-
-7-
-8-
-9-
-10-
-11-
-12-
-13-
-14-
-15-
-18-
-17-
-18-
-19-
-20-
-21-
-22-
-23-
-Z4-
-25-
-28-
NODE1
NODE2
NODE3
NODE4
NODES
SOUTH
-0.
—1
—2
—3
—4
—5
--6
—7
—10
—11
—12
—13
--14
—15
—IS
—17
—IB
—19
—20
—21
—22
—23
—24
—25
—28
DISTANCE (FEET)
60
LEGEND
10 - 100 (COUNTS)
- 1,000
31.000 - 10,000
29
-------
FIGURE 4-5. SCAPS CHEMICAL CROSS SECTION—FORT RILEY SITE
SOUTH
0-
-1-
-2-
-3-
-4-
-5-
-8-
-7-
-8-
-9-
-10-
-11-
-12-
-13-
-14-
-15-
-18-
-17-
-18-
-19-
-20-
-21-
-22-
-23-
-24-
-25-
-28-
-27-
-28-
-29-
-30-
-31-
NODE1
NODE2
NODES
NODE3
NODE4
NORTH
-7
•-8
-9
•-10
-11
-12
-13
-14
-15
-18
•-17
-18
-19
-20
-21
—22
•-23
—24
-25
—28
•-27
-28
-29
-30
-31
DISTANCE (FEET)
100 — 110
120
130
ISO
180
170
180
190
200
210
220
LEGEND
- 100 (COUNTS)
- i.ooo
11.000 - 10.00O
FIGURE 4-6. SCAPS STRATIGRAPHIC CROSS SECTION—ATLANTIC SITE
SOUTHEAST
NODES
0-
-5-
-7-
-8-
-9-
-10-
-11-
-12-
-13-
-14-
-15-
-18-
-17-
-23-
-%'-
-a-
IE
-34-
-35-
NODE4
NOOE3
NODE2
NORTHWEST
NODE1
-0
—1
— 2
— 3
— 4
— 5
—8
~7
— 8
—9
—10
— 11
17
IB
DISTANCE (FEET)
-20
21
22
23
28
27
-28
29
30
31
32
33
34
35
38
37
25 50 75 100 125 150 175 200
LEGEND »CAY F^jsAND Mis™ MIX M\™
250
300
30
-------
bgs that had a fluorescence emission peak at
approximately 480 nm. The same pattern was observed
at both pushes in this sampling node. Sampling Node
3 indicated fluorescence at 10 feet which continued until
approximately 25.5 feet. Inspection of the emission
spectra for this contaminant zone indicated contaminants
with different spectral features from those observed at
Node 2 and, therefore, the possible presence of two
different products. The fluorescence response for Node
4 indicated significant contamination areas at 8 to
30 feet. The emission spectra changed with depth. Near
the top of the zone, the spectra indicated peak
fluorescence at approximately 425 nm. The spectra at
28.7 feet indicated a possible mixture. The spectra at
30 feet had an emission maximum at 480 nm. This was
similar to the spectra from Node 2 at 22 feet. The
significant fluorescence response for Node 5 began at
approximately 8 feet and continued to 22 feet. The
emission spectra indicated a product that was
significantly different from the contaminant at the other
nodes. This product had an emission maximum at
400 nm.
The above descriptions are based on the real-tune
outputs from the LIF sensor. Thus, the operator can
identify different wastes as pushes are made. It is
important to note that this data is only used to identify
differences in wastes and not identify specific wastes or
classes of contaminants. Figure 4-2 shows the panel plot
for Node 4 at the Atlantic site. Shifts in "Wavelength at
Peak (nm)" seen in the far right of the panel plot clearly
shows the emission wavelength shifts discussed above,
and used to identify changes hi waste type
characteristics. Figure 4-9 shows three individual plots
of fluorescence intensity versus wavelength for select
depths from the Node 4 push. These plots can be
produced after a push and represent waveforms at
distinct depths during a push. This data is used to
confirm the conclusions regarding waste type differences
based on the panel plots.
York Site
The initial review of the panel plots during and
immediately after pushes indicated the potential presence
of three distinct contaminant types based on the different
wavelengths observed for the peak fluorescence response
at various depths. The spacial distribution of these
different contaminants on site was relatively consistent
across the five sampling nodes. However, the intensity
of the fluorescence response for each contaminant did
vary significantly between a number of the sampling
nodes. The contaminant near the ground surface was
consistently found to have an emission spectrum that
peaked at approximately 400 nm. The contaminants
detected at greater depths always yielded emission
spectra with peak wavelength that increased with depth.
Review of individual spectra at various depths indicated
a contaminant with an emission spectrum peaking at a
lower wavelength (450 nm) overlaying a contaminant
with an emission spectrum peaking at a longer
wavelength (480 to 500 nm). These observations will be
discussed hi detail below for the individual pushes.
The two Node 1 pushes yielded panel plots that were
similar. The stratigraphy was similar, as was the
fluorescence response. This sampling node yielded low
fluorescence response (300 to 500 counts) at 14 to
18 feet. The emission spectra of this fluorescence
response indicated a contaminant with peak fluorescence
at approximately 490 to 500 nm.
The two Node 2 pushes yielded panel plots that were
similar except for two observations. First, the second
push 'indicated the presence of a contaminant near the
surface (4 to 7 feet) that had an emission maximum at
410 nm that was not observed in the first push. This
was the first observation of this fluorescence response on
this site. The second difference between these two panel
plots was that the fluorescence response observed at
greater depths in the second push was resolved into two
bands (12 to 16 feet and 17 to 20 feet), while no similar
spacial resolution was observed for the first push.
However, the spectra for the two contaminant regions hi
the second push were very similar to those observed
from the top to the bottom of the band observed between
11 to 18 feet for the first push. The fluorescence
response of the upper zone hi this band indicated a
contaminant with a spectrum peaking between 410 and
430 nm. The middle of the band had a spectrum that
peaked at 450 nm, and the bottom of the band had a
spectrum that peaked at 480 nm. It should be noted that
the change hi the emission spectra could be inferred
during the push by observing the change hi the
wavelength at peak fluorescence panel that was
generated in real tune during a push event. As discussed
earlier, the pattern observed at this sampling node was
generally repeated at the other sampling nodes.
The spacial distribution and spectral characteristics
of the fluorescence response observed hi the two pushes
at Node 3 were very similar to those observed hi Node
2. However, the intensity of the fluorescence response
hi the second push at Node 3 was lower than that
observed for the first push at Node 3. The wavelength
of peak fluorescence intensity for the contaminants
detected at 11 to 15 feet and 16 to 20 feet were very
similar to the spectra obtained hi the Node 2 pushes at
similar depths. The two different fluorescence spectral
responses were well resolved spatially hi both the Node
3 pushes.
31
-------
The two pushes at Node 4 again indicated a
fluorescence response near the surface (0 to 1.5 feet).
The spectrum of this contaminant was the same as that
observed near the surface hi the second push at Node
2 (maximum fluorescence at 410 nm). The pattern
observed for Node 2 and Node 3; increasing wavelength
of peak fluorescence response with increasing depth was
observed for the two pushes at Node 4. The intensity of
the fluorescence response at this sampling node was
similar to those observed at Nodes 2 and 3.
The fluorescence response observed for the Node
5 pushes indicated the low wavelength fluorescence
(410 nm) response near the surface and the longer
wavelength fluorescence (400 to 500 nm) response at
greater depths. The intensity of the fluorescence at this
node was significantly less than that observed at Nodes
2, 3, and 4.
The extra push at Node 6 (not part of the formal
demonstration) was very similar to the general pattern
observed on this site. The fluorescence response
indicated three different contaminants at different depths.
The low wavelength (410 nm) contaminant was observed
near the surface and the longer wavelength response was
observed at depth (450 nm and 480 to 500 nm). Again,
the differences in the spectra for the different
fluorescence responses observed at various depths can be
inferred from the wavelength at fluorescence peak panel
of the standard panel plot.
Fort Riley Site
The results obtained at the Fort Riley site indicated
a fairly homogeneous distribution of a single contaminant
(emission spectra with peak fluorescence at about
410 nm). Node 1 indicated low level fluorescence near
the surface and from approximately 10 to 20 feet. The
first push at sampling Node 1 indicated a higher level of
contaminant in a narrow band at about 21 feet. In the
area around sampling Nodes 2, 3, and 5, the panel plots
indicated low to high level contamination beginning at
approximately 7 to 10 feet and continuing to about 20 to
22 feet. Sampling Node 4 indicated essentially no
fluorescence contamination.
Textural Data
The SCAPS CP uses ASTM methods to generate the
subsurface textural data. The individual CP logs were
used to construct stratigraphic cross sections for each of
the demonstration sites. Although the SCAPS produced
its own cross sections, PRC transferred the data and
plotted it on a scale that matched the one used for the
reference method. The transformed SCAPS strati-
graphic cross sections are presented on Figures
4-6, 4-7, and 4-8. The SCAPS data package did not
include a narrative of the stratigraphic cross sections,
therefore, a PRC geologist provided the descriptions
presented below. The SCAPS CP software uses the
term "mixed" to modify the soil classification. When
the "mixed" modifier is used, the dominant particle size
is named and the term mixed is added to indicate a
significant percentage of different size particles are
present.
Atlantic Site
The SCAPS CP sensor identified an unclassified unit
hi the surface 1 foot across the cross section. Figure
4-2 is the SCAPS stratigraphic cross section for the site.
The SCAPS CP sensor identified a thin layer of mixed
silt from 1 to 2 feet bgs across the cross section. From
2 feet bgs to approximately 21 feet bgs the SCAPS
sensor identified primarily clay. The SCAPS sensor
identified a silt mix lens in the northern three nodes of
the cross section at 8 to 15 feet bgs. This lens thinned
from 7 feet thick at Node 3 to 2 feet thick at Node 1. In
the southern two nodes clay was present from 3 to
28 feet bgs. A 2-foot-thick silt mix layer was below the
clay in the southern two nodes. This is followed by sand
to the bottom of the section. In the northern three nodes
from 21 feet bgs to the bottom of the cross section, the
SCAPS sensor identified primarily sand with several thin
clay, silt mix, and sand mix lenses throughout. The
SCAPS sensor also identified a 2-foot-thick peat layer at
19.5 feet bgs hi Node 5.
York Site
The SCAPS CP sensor identified sand, sand mix,
and silty mix hi the upper 2 feet of the cross section.
From 2 feet to 17 feet bgs, the CP sensor logged thick
lenses of clays and silt mix at the York site. Figure
4-4 is a SCAPS stratigraphic cross section for the site.
From 17 to 25 feet bgs (the bottom of the section), the
SCAPS sensor logged many thin beds of silt, clay, sandy
silt, silt mix, and sand.
Fort Riley Site
The SCAPS CP sensor identified extensive layers of
clays, silts, sands, and mixtures throughout this cross
section. Figure 4-5 is a SCAPS sensor stratigraphic
cross section for the site. From the surface to a depth of
10 feet bgs, the SCAPS sensor identified silt mix and
clay hi the south three quarters of the cross section. In
Node 4, the SCAPS sensor identified sand and sand mix
32
-------
FIGURE 4-7. SCAPS STRATIGRAPHIC CROSS SECTION—YORK SITE
NORTH
0-
-1-
-2-
-3-
-4-
-5-
-6-
-7-
-8-
-9-
-10-
-11-
-12-
-13-
-14-
-15-
-16-
-17-
-18-
-19-
-20-
-21-
-22-
-23-
-24-
-25-
-28-
NODE1
NODE2
NODE3
NODE4
NODE5
DISTANCE (FEET)
80
SOUTH
-0
—1
—2
—3
—4
--5
—8
—7
--»
—9
"10
—11
—12
--13
—14
—15
—18
—17
—18
—19
—20
--21
—22
—23
—24
—25
—28
120
LEGEND
[,'•:.; .[SAND
CLAY TO SANDY SILT
1 SAND MIX TO SANDY SILT
FIGURE 4-8. SCAPS STRATIGRAPHIC CROSS SECTION—FORT RILEY SITE
SOUTH
0-
-1-
-2-
-3-
-4-
-5-
-8-
-7-
-8-
-B-
-10-
-11-
-12-
-13-
-14-
-19-
-16-
-17
-18-
-19-
-20-
-21-
-22-
-23-
-24-
-25 H
-28-
-27-
-28-
-29-
-30-
-31-
90
NODE1
NODE2
NODES
NODE3
NODE4
NORTH
-0
-1
-2
-3
-4
5
•-8
-7
-8
-9
-10
-11
-12
-13
-14
-15
-18
-17
-18
-19
-20
-21
-22
-23
-24
-25
—28
-27
-28
—29
—30
—31
DISTANCE (FEET)
100 110 120 130 140 ISO 160 170 180 190 200 210
220
LEGEND
JStLTY CLAY
•I SILT MIX
I SAND MX
l.V.-SAND
33
-------
FIGURE 4-9. FLUORESCENCE INTENSITY VS. WAVELENGTH—NODE 4 ATLANTIC SITE
01
4-1
C
-------
Section 5
Data Comparison
The data produced by SCAPS were evaluated using
the criteria described hi Section 2. The qualitative and
quantitative data evaluations are discussed separately..
The qualitative evaluation compares the chemical and
stratigraphic cross sections produced by SCAPS relative
to cross sections from the reference methods. The
quantitative evaluation statistically compares the SCAPS
data with analytical data produced by the reference
methods.
Qualitative Assessment
The qualitative assessment presents the evaluation of
both the stratigraphic and chemical mapping capabilities
of the SCAPS sensors relative to the reference methods.
In addition, the potential affects of TOC on the system's
measurements are examined. Both the reference and
technology cross sections were produced from collocated
sampling areas as discussed in Section 2. Since these
methods were sampling spatially different locations,
matrix heterogeneity will impact the comparisons of both
the physical and chemical cross sections. Based on a
review of the demonstration data, this impact appears to
have had a minimal impact on the qualitative data
evaluation.
Stratigraphic Cross Sections
The following sections present descriptions of the
similarities and differences observed between the
stratigraphic cross sections produced by the SCAPS CP
sensor and the reference methods. For this comparison,
PRC used the SCAPS cross sections shown in Section
4. These cross sections were produced directly from the
technology's raw data, however, they are scaled to
match the reference method cross sections shown in
Section 3. These comparisons are qualitative and, as
such, are subjective in nature. However, these
comparisons were made by a certified professional
geologist (American Institute of Professional Geologists)
with over 17 years of experience in this field.
Atlantic Site
The SCAPS sensor and the reference method's
stratigraphic cross sections exhibited good correlation.
The surface materials identified as silts and silty clays by
SCAPS were identified as fill and silty clay by the
reference methods. Fill is defined as a man-made
deposit of rock and/or soil. Sand was identified hi the
northern (Node 1) portion of the cross section by both
the SCAPS and the reference methods at approximately
21 feet bgs. A 7-foot-thick silt mix lens was identified
by both the SCAPS and the reference methods hi the
center of the cross section extending from 8 to 15 feet
bgs. The SCAPS and reference methods showed
relatively good correlation at Nodes 1 and 2, except that
the different strata were mapped at slightly shallower
depths by the SCAPS relative to the reference methods.
The SCAPS also identified a 2-foot-thick peat layer at
19.5 feet bgs in Node 5, while the field geologist saw no
evidence of peat hi the soil core.
One notable variation between the SCAPS sensor's
and reference methods cross sections was observed. The
reference method had trouble collecting samples for
logging purposes in the running sands that generally
occurred from 20 feet bgs to the termination of the
reference borehole. This lack of complete sample
recover/ is common for this method of borehole logging,
and caused the geologist to use circumstantial evidence
to fill in the resultant gaps in the borehole logs at depth.
The circumstantial evidence used was direct feedback
from the driller on changes in drilling characteristics,
cuttings, and interpolation based on what was recovered.
The SCAPS did not need to physically collect soil
samples to produce borehole logs, and thus, is not as
affected by running sands. This explains the greater
detail shown hi the SCAPS cross section below
approximately 20 feet bgs.
Seven samples were collected at the Atlantic site for
geotechnical analysis. The results of these analyses were
35
-------
compared to the corresponding SCAPS stratigraphic
data. Four out of the seven samples showed intermethod
agreement. The remaining samples were not matched
due to the SCAPS lack of reporting or detecting
increases in sand content. This resulted in the SCAPS
identifying intervals as clays when they were identified
by the reference method as sandy clays or silts. This
indicates that the SCAPS may not be sensitive to small
shifts in particle size distribution.
York Site
The SCAPS CP sensor and the reference method's
cross sections exhibited fairly good correlation. The
SCAPS identified sand, sand mix, and silty mix in the
top 2 feet of the cross section, while the reference
method identified the same interval as fill. The SCAPS
identified components of fill and, therefore, for this
zone, the SCAPS and the reference method are most
likely identifying the same material. From 2 to
17 feet bgs, the SCAPS identified thick lenses of
mixtures of clays and silt, while the reference method
identified thick layers of clayey silt with lenses of silt
and silty clay. From 17 to 25 feet bgs (the bottom of the
cross section), the SCAPS identified thin beds of silt,
clay, sandy silt, silt mix, and sand. The reference
method identified this interval as being composed of
primarily lenses of sand with sandy silt and silt. The
small lenses of silty clay were not identified in the
reference method's logs. The lack of correlation
relative to the thin sand, silt, and clay lenses may be
more representative of the reference method's inability
to resolve thin strata. The detail of the reference method
can be increased by spending more time examining
sample cores, however, time and cost factors often
prohibit fine detailed examination of sample cores. The
SCAPS produces the same level of detail whenever it is
used. Running sands were not a problem at this site.
Six samples were collected at the York site for
geotechnical analysis. The results of these analyses were
compared to the corresponding SCAPS stratigraphic
data. Four out of the six samples showed intermethod
agreement. The remaining two samples did not show
good agreement. This was due to the SCAPS apparent
inability to detect small increases of decreases in coarse
or fine particle sizes. This indicates that the SCAPS
sensor may not be sensitive to small shifts hi abundance
in secondary particles sizes.
Fort Riley Site
The SCAPS CP sensor and the reference method
cross sections are generally well correlated when the
cross sections are considered as a whole, however,
minor differences occurred when individual layers were
examined. The SCAPS identified many more variable
clays, silts, sands, and mixture layers than the reference
method. Nodes 1 and 4 were quite similar in both cross
sections with the exception mat the lower sand was
logged at different depths in each. (At Node 1, the
SCAPS located the beginning of the sand at 20.5 feet
bgs, while the reference method located its upper limit
as 19 feet bgs). In Node 4, the SCAPS logged sand
from 1 feet bgs to the terminal depth of the push, while
the reference method logged sand from 10 feet bgs to the
termination of the borehole. Below 19 feet bgs, across
the cross section, the SCAPS identified numerous thin
lenses of silt, silt mix, and sand, while the reference
method identified primarily sand. This may be due to
the occurrence of running sands below 19 feet bgs. This
is similar to the differences observed at the Atlantic site.
Eight samples were collected at the Fort Riley site
for geotechnical analysis. The results of these analyses
were compared to the corresponding SCAPS
stratigraphic data. Only two of the samples showed
intermethod matches. However, both methods identified
the dominant particle size as sand. This lack of
intermethod agreement was due to the SCAPS lack of
sensitivity to small changes hi particle size distributions
for minority constituents in a given strata. In all cases
of poor intermethod matching, the SCAPS identified the
sample as a sand when the reference method laboratory
identified the sample as a silty or clayey sand. This type
of disagreement was also seen between the geologist's
classifications and the reference method laboratory
classifications (see Section 3).
Summary
The SCAPS CP sensor and the reference method
produced similar geologic cross sections; however, the
SCAPS data showed more detailed spatial resolution. In
addition, limited QC checks of the SCAPS stratigraphic
data showed good correlation with the reference method.
The SCAPS was not as sensitive to small shifts hi
particle size distribution relative to the reference method.
The SCAPS provided a finer resolution of thin strata by
identifying more thin stratigraphic units than the
reference method. This difference was magnified when
running sands were encountered at the Atlantic and Fort
Riley sites. This may be due to the CP sensor's ability
to continuously acquire soil textural data during a push
and the common limitations of a geologist's logs where
strata are less than several niches thick. It is possible
that the SCAPS cross sections are more representative of
the actual site stratigraphy below 19 feet bgs at the Fort
Riley and Atlantic sites. An additional difficulty with
the reference method was its inability to retrieve samples
from running sands. This caused significant data gaps at
depth. The SCAPS does not require active soil sampling
36
-------
to log a hole, and therefore, it is not as affected by
running sands, and may be more representative of
subsurface stratigraphy than the reference method in
running sands.
Chemical Cross Sections
The following sections present descriptions of the
similarities and differences observed between the
chemical cross sections produced by the SCAPS LIF
sensor and the reference method. Unless otherwise
specified the comparisons are made in consideration of
both reference cross sections for TPH and total PAH.
PRC used the SCAPS LIF sensor's cross sections shown
in Section 4. These cross sections were produced
directly from the SCAPS raw data, however, they are
scaled to match the reference method cross sections
shown in Section 3. These comparisons are qualitative,
and as such are subjective hi nature. The effects of
heterogeneity may influence this data comparison,
however, the qualitative nature of this comparison should
greatly reduce the potential impact of heterogeneity in
contaminant distribution. These comparisons were made
by a soil scientist with over 9 years of experience hi site
characterization activities.
Atlantic Site
Both the SCAPS LIF sensor and the reference
method showed good correlation for background
characterization. This is exhibited by the data from both
SCAPS and the reference method showing Node 1 to be
outside the area of contamination. Both reference cross
sections detected the zone of contamination at Node
2, which extended from approximately 20 to 28 feet bgs
for TPH and from 16 to 31 feet bgs for total PAH. The
SCAPS identified this zone being from 2 to 9 feet
thinner than the reference method for TPH and total
PAH, respectively. The SCAPS identified the zone as
beginning almost 2 feet shallower and being 2 feet
thinner than the reference method, relative to the TPH
cross section. This difference is acceptable and can be
explained as an artifact of data interpolation, which was
used for the reference method to create the reference
method cross section. This is common when relatively
few samples are used to define zones of contamination.
The major difference between SCAPS and the reference
method in Node 2 dealt with the failure of the SCAPS to
detect the zone of elevated contamination 1.5 feet bgs
identified by the reference method. This difference may
have been due to spatial variability exhibited across the
node. The size of the shallow contaminated zone may be
an artifact of data interpolation. Overall, the zones of
elevated SCAPS LIF data corresponds well with general
zones of contamination shown in both reference method
cross sections. The shape of each cross section is
heavily influenced by the contour intervals used, and
therefore, it is not possible to say which reference cross
section shows the closest match to the SCAPS cross
section. The quantitative data evaluation will answer
this question. Interpolation can often lead to the over-
estimation of layer thicknesses.
Another way to examine the relationship between
the LIF sensor's data and the qualitative reference
method data is to superimpose the two data types on a
single plot of fluorescence intensity and reference
method concentration against depth. To make the plot
scales meaningful, the SCAPS LIF data and the
reference method data had to be normalized. The
reference method data was normalized to the highest
TPH and total PAH concentrations measured during the
qualitative sampling. The LIF data was normalized to
the average high reading measured over a qualitative
method reference sampling point at this site. This
normalization allows a general comparative evaluation of
the data.
Figures 5-1, 5-2, and 5-3 show the normalized data
plots. A review of this data shows that the qualitative
reference method data and the LIF sensor's data
generally agree in their identification of zones of high,
medium, and low contamination. The major exception
to this is exhibited in Node 2 (1 to 1.5 feet bgs). In this
zone, the reference method exhibited both TPH and total
PAH contamination hi the range of 50 percent of the
high reading for the site. This is opposite of the SCAPS
LIF data which exhibited contamination hi the range of
1 percent of the high LIF reading. This difference may
have been an artifact of the heterogeneity of the
contaminant distribution, the relative constituent
distribution of the waste, or it could reflect a false
negative reading. Minor differences were seen in the
relative readings produced by both data sets for the
zones of lowest contamination. In these cases the LIF
data was generally lower. This is probably an artifact of
the normalization of the LIF data. In these cases the LIF
sensor did detect increased fluorescence, however,
relative to the high, this fluorescence was generally less
than 1 percent.
York Site
The SCAPS LIF sensor's cross section showed little
correlation to the reference method cross sections at
Node 5. Both the TPH and total PAH reference cross
sections exhibited zones of elevated contaminant
concentrations at Node 5. The zone of elevated total
PAH contamination extended from approximately 13 to
22 feet bgs, and the TPH contamination extended from
approximately 1 to 24 feet bgs. The LIF sensor only
37
-------
r
FIGURE 5-1. NORMALIZED LIF AND QUALITATIVE REFERENCE DATA—ATLANTIC SITE
Node -2 Node -3 Node -4 Node -5
Percentage of High Reading Percentage of High Reading Percentage of High Reading Percentage of High Reading
0 50 100150200250 0 100 200 300 400 0 100 200 300 400 100 300 500
n OR
1.96
3.92
5.78
7.74
9.61
—11.34
£.13.33
t 15.08
18.94
20.94
2469
26.69
28,58
i i i i
—•.
1 — « — _
*"
0.07
2.1
4.1
5.85
7.83
9.58
— 11.3
£13.31
£ 15.13
§"17 1"5
Q1'-10
18.92
20.94
2284
24.77
26.81
28.61
1 t 1
T
j"
L
e- ==
007
2.12
4.11
6.04
7.9
9.76
—11.52
£13.49
£15.35
1733
Qi9:o9
21.05
22.91
24.65
26.62
28.52
i i i
E=s:
fe=
£^
J
c_
£
—
/Total PAH oTPH
-LIP Data
U ZUU 4UU
0.06
1.9
3.9
5.65
7.65
— 11.25
£13.26
flB.12
*
•fefc^
^L
r
y
0.06
1.81
3.78
5.61
7.6
8.47
11.48
g13.47
|.16.48
* 17.6
18.28
21.17
23.19
2527
2724
29
1 1 1
~^— P
f
K
IS.
1
38
-------
FIGURE 5-3. NORMALIZED LIF AND QUALITATIVE REFERENCE DATA—FORT RILEY SITE
Node-1 Node -2 Node -3 Node -5
Percentage of High Reading Percentage of High Reading Percentage of High Reading Percentaqe of High Readinu
0 50 100 150 200 250 20 60 100 140
nnft
U.UD
1.83
3.83
5.86
7.82
9.58
11.43
§13.41
5^15.19
,§"17.18
19.03
21.04
22.9
24.59
26.58
28.46
i I i i
r2"
>
?•
*~
0 100 200 300 400 <
0.06
1.78
3.74
5.48
7.46
9.34
11.35
§13.13
t~~ 14.9
16.89
18.65
20.63
22.5
24.5
26.38
28.25
i i i
7
^
f~
'fr
^r-S"
p>
0.06
1.86
3.85
5.68
7.67
9.56
^11.41
gl3.41
§15.29
17.27
19.01
21
22.83
24.68
26.71
28.61
) 40 80 120
I
jl
= \
v~
?~
i
~ ">•
g=-
"
t-f
0 100 200 300 400
33.95
1.46
3.28
5.14
7.11
8.86
10.87
g.12.76
£ 14.7
g 16.58
18.31
20.31
22.17
24.18
26.07
27.91
nTotal PAH nTPH .LIF Data \
' '
i
^•'^
^__
^^ir
..
identified limited fluorescence from 1 to 3 feet bgs.
These differences may have been due to spatial
variability in contaminant distribution, however, the
vertical extent of this contamination probably indicates
more than isolated spots of contamination. Nodes
1, 2, 3, and 4 showed better correlation between the
reference method cross sections and the SCAPS cross
section. Overall, the SCAPS LIF cross section was
relatively similar to the two reference cross sections.
Since the contour intervals strongly influenced the shape
of the contours, it is not possible to identify which
reference cross section most closely match the SCAPS
LIF cross section. The quantitative data evaluation will
answer this question. The differences between the
reference method cross sections and the SCAPS cross
section could be the combination of an artifact of data
interpolation for the reference method cross sections,
and the finer definition provided by the SCAPS, which
produces continuous profiles with a 2 cm resolution.
The zone of low SCAPS readings shown around Node
3 may be a reflection of the spatial variability of the
contamination or the small elementary sample volume
used by the technology.
Figure 5-2 shows the normalized line graphs of the
five SCAPS LIF sensor pushes at the York site. The
qualitative reference data for TPH and total PAH is
superimposed on these line graphs, at the sample depths
they represent. The reference data has been normalized
to the high average LIF reading measured at the
qualitative reference method sampling depths. This
normalization makes the data comparable on a relative
scale.
A review of this data shows that generally the
relative magnitudes between the two types of data were
hi agreement. Zones of high reference readings
corresponded to zones of high LIF readings. This
relationship appears to hold for medium and low zones
of contiimination. At the low end of this comparison it
appears as if the LIF data is much less than the reference
data. This is an artifact of the normalization procedures.
In several cases, the LIF data produce relatively much
higher readings. This can be seen in Node 3 (17 to
18.5 feet bgs) and hi Node 4 (17 to 18 feet bgs). In
these cases, the LIF data was 100 to 50 percent of the
high reading, while the reference method data was at
approximately 1 percent of the high reading. These are
examples of false positive readings for the LIF data.
However, heterogeneity of contaminant distribution, or
the constituent composition, could have influenced this
data.
Fort Riley Site
The SCAPS LIF sensor's cross section showed little
correlation to the reference method's cross sections at
Node 4,, The reference method cross sections exhibited
39
-------
an isolated zone of elevated contaminant concentrations
at Node 4. This isolated detect may be an artifact of
limited reference sampling at this node. Examination of
the drilling logs for this node indicate that this was the
only depth interval at Node 4 to exhibit elevated (above
background) readings on the portable photoionization
detector (PID). The SCAPS detected three zones of
elevated fluorescence readings along Node 4. These
relatively small zones of fluorescence may be
representative of the spatial variability of contamination
at Node 4 and the small representative elementary
sampling volume for the technology. Aside from Node
4 on the northernmost end of the transect, the remaining
nodes produced cross section data that showed a
relatively good match between Jhe technology and the
reference methods. The greater definition of potential
contaminant lenses in the SCAPS cross sections is most
probably due to the 2 cm sampling resolution provided
by the technology. The need to interpolate data for the
reference method reduces the potential for identifying
distinct smaller lenses of contamination. Overall, the
SCAPS cross section exhibited a good match with the
reference method cross sections. Since the shape of the
cross sections is heavily influenced by the selected
contour intervals, it is not possible to identify which
reference cross section exhibited the closest match to the
SCAPS LIF cross sections. The quantitative data
evaluation will answer this question.
Figure 5-3 shows the normalized line graphs of the
five SCAPS LIF sensor pushes at the Fort Riley site.
The qualitative reference data for TPH and total PAH is
superimposed on these line graphs, at the sample depths
they represent. The reference data has been normalized
to the highest TPH and total PAH concentrations
detected. The SCAPS LIF data has been normalized to
the high average LIF reading measured at the qualitative
reference method sampling depths. This normalization
makes the data comparable on a relative scale. A review
of this data shows that for all pushes the general
contamination trends identified by the technology match
the trends detected by the qualitative reference data.
Similar zones of low, medium, and high contamination
were identified by the technology and the reference
method.
Summary
Generally, the SCAPS LIF sensor showed a good
relative correlation with the reference method's cross
sections. The closest match was exhibited when
technology's cross section was compared to the total
PAH reference method's cross sections. The TPH
reference method cross sections generally appeared to
show more resolution than either the technology's or
total PAH reference method cross sections. In addition,
the SCAPS data and qualitative reference method data
were well correlated hi their identification of zones of
low, medium, and high contamination.
The observed differences between the cross sections
for the SCAPS LIF sensor and reference method could
have been caused by several factors. The SCAPS
sampling volume covered a circle less than 0.5 cm hi
diameter, and approximately one micrometer thick
(approximately 0.2 cubic centimeters). This makes the
SCAPS hyper-sensitive to the natural spatial variability
of contaminant distribution. The reference method's use
subsample from a homogenized 12-inch sampling
interval, approximately 1,000 grams of soil. This base
sample volume is several thousand times larger than the
sample volume used by SCAPS. This larger sample
volume may average out the smaller heterogeneities
detected by the SCAPS sensor. Some of this relative
sample volume effect is canceled out by the fact that the
technology collects much more data. In the case of this
demonstration, the reference method used a total of
76 samples, compared to the over 1,300 sample points
the SCAPS produced.
Total Organic Carbon
PRC compared the SCAPS sensor's intensity
measurements for areas free from contamination to the
corresponding TOC concentrations. This evaluation
examined the potential for gross humics to affect LIF
sensor intensity measurements. SCAPS data from the
York, Atlantic, and Fort Riley sites were reviewed.
This evaluation focused on contamination-free zones to
eliminate the carbon from the site contaminants from
biasing the results. The samples collected for this
evaluation exhibited TOC concentrations ranging from
not detected to over 3,000 ppm. Based on the limited
data base (11 samples), there appears to be no affect of
TOC concentrations on LIF data at any of the three sites.
This is based on the fact that although the TOC
concentrations varied over three orders of magnitude,
the LIF intensity measurements remained relatively
constant. However, it is possible that TOC becomes a
potential interferant in the presence of organic solvents
or petroleum products. This interference may be created
by the contaminants' activation of fluorescent properties
hi the TOC, specifically humics. Isolation and
examination of the potential for this activation of
fluorescence in humics was beyond the scope of this
demonstration.
Quantitative Assessment
This section presents the comparative evaluation of
the SCAPS LIF sensor's data and the reference method's
analytical data, and an evaluation of the technology's
40
-------
precision and resolution. The precision and resolution
discussion will be presented after the regression analysis
discussion.
The reference method sampling and analysis
identified considerable heterogeneity in the distribution
of contaminants in the soil matrix. The experimental
design of this demonstration expected heterogeneity and
intended to define it through collocated replicate
sampling. This sampling did define the heterogeneity.
However, in many cases the heterogeneity was greater
than expected. In almost 50 percent of the 21 quanti-
tative sample intervals the heterogeneity produced ranges
between maximum and minimum concentrations in
excess of one order of magnitude. This heterogeneity
coupled with the developers inability to specifically
identify the compounds they are measuring, the lack of
a reference analytical method that monitors the exact
suite of the compounds measured by the technology, the
mixed distribution of constituents hi the contamination,
and the varied age of the contaminants cause uncertainty
to be introduced into the point by point comparison of
data in the quantitative evaluation. Therefore, any
conclusions stated in this section should be considered as
trend indicators and not definitive statements on
technology performance. However, the conclusions are
likely to be duplicated if similar field in situ verification
is attempted.
The quantitative assessment evaluated SCAPS LIF
sensor's data at distinct intervals relative to
corresponding data from the reference method. This
evaluation is intended to quantify relationships between
the technology's data and compound or class-specific
analytical data produced by the reference methods. The
target compounds for this evaluation were TPH, VPH,
BTEX, total BTEX, naphthalene, 1-methylnaphthalene,
2-methylnaphthalene, acenaphthene, fluoranthene,
phenanthrene, pyrene, benzo-a-pyrene, total PAH, and
total naphthalene. The TPH, VPH, total naphthalene,
total PAH, and total BTEX groupings were made in an
effort to more closely match the technology's data. The
developers felt that classes of compounds would show
the closest match to the technology's data.
This data evaluation involved regression analysis of
the SCAPS LIF data against the corresponding reference
method data. As defined in the approved demonstration
plan, a correlation coefficient (r2) of 0.80 or better
defines a useable predictive model.
The SCAPS LIF sensor made two collocated pushes
at each node. The first push was intended to produce the
primary data for both the qualitative and quantitative
evaluations. The second push was intended to examine
the teclinology's precision. The second push also
produced continuous LIF data to depth. The primary
data evaluation focused on the data from the first push at
each node, however, PRC did examine the possible
impact of averaging the two pushes for the regression
evaluation. This averaging had very limited impact on
the outcome of the regression analysis and will only be
discussed where its findings differ from the first push
data.
The data sets were initially examined as a whole and
then post-hoc techniques were used to eliminate data
outliers. The total data set for this evaluation consisted
of 21 sampling intervals, 8 at the Atlantic site, 7 at the
York site, and 6 at the Fort Riley site. Each one of
these intervals produced one data point for the regression
analysis,, however, each of these data points represented
the mean concentration from five collocated samples.
Therefore, this evaluation was based on the analytical
results of 105 individual samples and analyses. The data
presented is based on non-transformed data. PRC
mirrored the analyses discussed below with
log-transformed data, however, in no case did the
correlations improve. This suggests that the high and
low concentration points did not disproportionally bias
the regression.
PRC also examined the data hi its raw form, prior
to averaging the reference method data. This approach
did not improve the correlation of the data.
The initial regression analysis examined the data set
of mean concentrations as a whole. From this
evaluation, no i^s of greater than 0.20 were observed
(Table 5-1). An examination of the maximum and
minimum concentrations for each set of collocated
samples indicated that several locations at each site
exhibited considerable heterogeneity. This was expected
and is normal for environmental sampling. Using data
pouits from reference sampling depths that exhibited
wide ranges hi contamination introduced additional
uncertainty into the data evaluation. In these cases, it
was hard to define representative mean concentration.
Concentrations were highly location dependant. In an
effort to reduce the impact of this heterogeneity on the
data evaluation, all data pouits exhibiting a greater than
1 order of magnitude range between the maximum and
minimum were eliminated. Ranges, which are
nonparaimetric statistics, were selected for this post-hoc
data reduction since they are not dependent on data
distribution. In most cases, this resulted hi at least a
50 percent reduction hi useable data. For this reason,
the subsequent data analyses should be considered
indicators of trends hi correlation, and not well-defined
predictive models.
41
-------
TABLE 5-1. REGRESSION ANALYSIS RESULTS FOR SCAPS AND THE REFERENCE METHODS-
ALL SITES
Initial Regression
Compound
TPH
VPH
Benzene
Toluene
Ethylbenzene
Xylene
Naphthalene
1 -Methylnaphthalene
2-Methylnaphthalene
Alenaphthene
Fluoranthene
Phenanthrene
Pyrene
Benzo-a-Pyrene
Total Naphthalene
Total PAH
Total BTEX
Notes:
n
21
20
16
19
20
20
21
18
20
12
19
21
17
19
21
21
20
r2
0.01
0.02
0.02
0.03
0.00
0.01
0.03
0.07
0.01
0.00
0.12
0.02
0.01
0.03
0.01
0.01
0.02
n Number of sample points, each
r2 Coefficient of determination.
slope
0.12
0.01
0.20
0.85
0.19
0.67
-0.18
0.01
-0.00
-0.00
0.00
-0.00
-0.00
0.00
-0.01
-0.00
2.20
sample
y-intercept
(ppm)
3,340
274
4,943
13,023
11,391
30,510
46.4
47.0
34.2
45.2
2.19
18.2
8.03
1.25
202
117
56,539
Final Post-Hoc Data Reduction
n
7
9
8
10
10
11
9
9
9
4
12
10
7
8
8
9
10
r2
0.89
0.94
0.34
0.41
0.88
0.94
0.01
0.29
0.43
0.31
0.10
0.02
0.76
0.50
0.50
0.06
0.94
slope
2.2
0.16
1.9
9.6
7.0
17.4
0.00
0.02
0.01
0.09
-0.00
-0.00
0.01
0.00
0.07
0.02
38.7
x-intercept
y-intercept (fluorescence
(ppm) intensity)
-346
-53.8
3,028
4,376
607
-3,377
14.6
38.4
10.7
0.90
1.80
15.2
-0.65
0.69
50.0
101
-11,370
point is the mean concentration from five collocated
-157
336
-1,594
-518
-87
194
No Data
-1,920
-1,070
-10
No Data
No Data
0.02
No Data
-714
-5,050
294
samples.
ppm Parts per million.
After these data points were removed, the regression
analysis was run again. No significant changes hi the
regression parameters were observed (Table 5-1).
However, a post-hoc examination of the residuals
identified several outliers for each regression.
A final regression analysis was conducted on the
data sets after the outliers were removed. This
regression showed considerable improvements in the
data correlation (Table 5-1). TPH, VPH, ethylbenzene,
xylene, and total BTEX all exhibited i^s above the
0.80 criteria for acceptance. Pyrene had an r2 of
0.76, almost meeting the acceptance criteria for
correlation. The slope data cannot be used to assess data
quality since the LIF data was not in the same units as
the reference method data. However, the slope data can
indicate trends hi relative fluorescence. The slopes of
the ethyl benzene, xylene, and total BTEX regressions
were all much greater than 1.0. This indicates that
relatively large changes hi contaminant concentration are
required to cause changes in LIF data. Conversely, the
VPH regression had a slope much less than 1.0,
indicating that small changes in VPH can cause relatively
larger changes in LIF data. This can be translated into
a general conclusion regarding the LIF sensor's
sensitivity. Based on the slope data, the LIF sensor
appears to be most sensitive to the compounds measured
hi the VPH analysis relative to the TPH, ethylbenzene,
xylene, and total BTEX analyses.
42
-------
Although the r2 data at this point indicates that the
concentrations of the above compounds appear to be
correlated, the small size of the data set limits the
usefulness of any predictive models based on these
regression parameters. The regression parameters for
TPH and VPH could best be used to produce general
predictive models for concentration based on LIF data.
Due to the negative y-intercepts, these models could not
be applied to LIF data for intensities below 157 for TPH
and 336 for VPH. These intensities correspond to the x-
intercepts for the respective regression models when the
concentration of contaminants is 0 mg/kg.
The number of compounds that exhibited acceptable
correlations suggests that these relationships are real.
However, the lack of correlation observed for many of
the compounds may not be wholly attributable to
technology performance. Rather, poor correlations are
likely due to a combination of effects such as matrix
heterogeneity, the lack of a definitive match between
reference analytical methods and the suite of compounds
measured by the technology, the variable distribution of
contaminant constituents, and the variable ages of the
contaminants.
Similar conclusions are drawn if the data from the
two SCAPS LIF sensor pushes are used in the regression
analysis. The only exceptions are for ethylbenzene and
pyrene. In this data set, ethylbenzene no longer exhibits
an acceptable r2, but pyrene does (Table 5-2). The same
trends in slopes are observed for this data set, the LIF
sensor seems to be more sensitive to the VPH, and hi
this data set the PAH compound pyrene. The TPH and
VPH shows the most conducive data for creating a
predictive model, just as above.
The quantitative determination of a detection limit
for the SCAPS LIF sensor was not possible given the
data produced from this demonstration.
Qualitative observations regarding the detection
limits of this technology can be made with the data
produced from this demonstration. Measurable
fluorescence was reported for TPH concentrations as low
as 60 mg/kg and VPH concentrations as low as
19 mg/kg. At no point during the demonstration did the
SCAPS LIF sensor report no fluorescence above
background for soils exhibiting contamination detectable
by the reference methods. Another qualitative method
for assigning a detection threshold is to determine the x-
intercept for the TPH and VPH regression models
discussed above. The x-intercept for these models
represents the point at which TPH or VPH
concentrations are 0 mg/kg. For TPH the fluorescence
intensity at the x-intercept is 157 and for VPH it is
336. Cross checking these pseudo thresholds against the
information hi Table 5-3, for data remaining after the
initial removal of outliers based on heterogeneity, shows
that hi most cases SCAPS LIF readings in these
threshold ranges corresponded to the lowest contaminant
concentrations.
To examine the potential for site-induced effects on
the data evaluation, the data was divided by site and
regression analyses were run on the resultant three data
sets. This regression analysis began with data sets
whose gross outliers had been removed. These outliers
were defined as data points where the maximum and
minimum values varied by over one order of magnitude.
This site-specific regression showed that only
naphthalene and fluoranthene exhibited acceptable
correlations at the Atlantic site; no compounds showed
acceptable correlations (r2 greater than 0.80) at the York
site; and acceptable correlations for toluene, VPH, and
total BTEX were found at the Fort Riley site. The
number of samples resulting hi these acceptable
correlations ranged from 3 to 4, out of a maximum of
6 to 8. This small sample set greatly limits the use of
this data to form predictive models. However, these
regressions exhibited the same trends in their slopes, as
exhibited hi the data set as a whole. The slopes for the
VPH arid PAHs were all less than 1.0, and the BTEX
compounds produced regression equations with slopes
greater than 1.0.
Inherent instrument precision for the SCAPS LIF
sensor measurements at the York and Atlantic sites was
evaluated by calculating the percent RSD of each set of
10 replicate measurements, taken at single depths (Table
4-1). The SCAPS took no precision measurements at
the Fort Riley site. These precision measurements were
taken from intervals where peak wavelengths ranged
from 350 to 650 nm. The percent RSD was calculated
by dividing the standard deviation by the mean, then
multiplying the result by 100. The range of RSDs at the
Atlantic site were 1.1 to 4.1. The range of RSDs for the
York site were 1.0 to 1.6. Based on this data, the
standard deviations noted on Table 4-1 are most likely
due to heterogeneity hi contaminant distribution. The
maximum inherent instrument precision of the SCAPS
LIF sensor observed during this demonstration was
4.1 and 1.7 percent for the Atlantic and York sites.
With this high degree of inherent instrument precision,
it is possible to identify the cause of the wide range of
measurement standard deviations exhibited hi Tables
4-1, 4-2, and 4-3. All but 1 to 5 percent of this variance
can be attributed to matrix heterogeneity hi the vertical
direction. The small area and volume of the SCAPS LIF
measurements tend to accentuate matrix heterogeneity hi
the soil matrix.
43
-------
TABLE 5-2. REGRESSION ANALYSIS RESULTS FOR THE AVERAGE OF BOTH SCAPS PUSHES
AND THE REFERENCE METHODS—ALL SITES
Initial Regression
Compound
TPH
VPH
Benzene
Toluene
Ethylbenzene
Xylene
Naphthalene
1-Methylnaphthalene
2-Methylnaphthalene
Alenaphthene
Fluoranthene
Phenanthrene
Pyrene
Benzo-a-Pyrene
Total Naphthalene
Total PAH
Total BTEX
Notes:
n
21
20
16
19
20
20
21
18
20
12
19
21
17
19
21
21
20
r2
0.09
0.11
0.14
0.11
0.05
0.05
0.00
0.16
0.01
0.00
0.19
0.00
0.00
0.09
0.00
0.01
0.09
n Number of sample points, each
r2 Coefficient of determination.
ppm Parts per million.
slope
0.63
0.05
0.86
3.0
1.1
3.5
0.001
0.01
0.00
0.00
0.00
-0.00
0.00
0.00
0.00
0.01
9.1
y-intercept
(ppm)
2,485
205
3,619
9,251
9,725
25,403
37.6
41.6
30.0
44.0
1.76
17.1
7.27
1.05
183
103
44,820
Final Post-Hoc Data Reduction
n
7
10
8
10
10
11
11
11
9
4
12
9
7
7
10
10
10
r2
0.84
0.95
0.50
0.42
0.69
0.88
0.04
0.09
0.72
0.29
0.08
0.01
0.92
0.32
0.32
0.08
0.89
slope
2.1
0.22
2.6
9.7
10.7
17.0
0.00
0.01
0.02
0.12
-0.00
-0.00
0.02
0.00
0.05
0.02
37.7
y-intercept
(ppm)
375
-72.9
1,874
5,248
1,891
1,120
14.3
37.2
2.6
-15.8
1.76
16.1
-1.38
0.37
39.8
96.9
392
x-intercept
(fluorescence
intensity)
-179
331
-721
-541
-177
-66
No Data
-3,720
-130
132
No Data
No Data
69
-185
-796
-4,845
-10
sample point is the mean concentration from five collocated samples.
The wavelength resolution of the SCAPS LIF sensor
was also examined during this demonstration. During
the precision measurements at the Atlantic site, the
deviation reported peak wavelengths for the York site
ranged from 0.6 to 1.7 percent. Based on an exam-
ination of the spectral wave forms produced by SCAPS
during this demonstration, PRC determined that the
reported peak wavelength could vary by approximately
plus or minus 5 percent before significantly affecting the
reported intensity. The inherent instrument peak
wavelength resolution is less than 5 percent and, thus, it
should not affect instrument performance.
44
-------
TABLE 5-3
Site
Atlantic
York
Fort Riley
Notes:
a r>at=
. DATA
Node
2
2
3
4
4
4
5
5
1
2
2
3
4
4
5
1
1
2
2
5
5
a no into re
FOR MEAN SCAPS,
Depth
(feet)
21-22
24-25
16-17
6.5 - 7.5
10-11
27.5 - 28.5
16-17
23.5 - 24.5
15-16
13.5 - 14.5
17-18
17-18
14-15
18-19
1.5-2.5
2-3
13-14
6-7
17-18
10.5-11.5
16-17
imoininri aftar lha initial
TPH, AND VPH— ALL
SITES
SCAPS
Fluorescence
Intensity TPH
(mean) (mean mq/kq)
5,809
5.37
1,323
213.4
837.1
6,310
19,674
29.88
221.3
723.8
268.5
908.4
775.7
412.5
297.1
1,143
366.6
1,136
4,853
2,923
3,036
rammf^l f\f *"\ii+li»r^ I^Qi^Afl
11,090a'b
4,044
425
255
2,436a'b
1,094
201 a
239a
773
1,539
497
778a,b
2,281 a'b
1,878
60a'b
5,728
1,416
2,169
13,150a'b
22,480a
3,926a'b
^"if» rv^^vimi trvi *
VPH
(mean mq/kq)
1,402a
538
112a,b
43
1,320a
452
96a
77a,b
No Data
25a,b
20a'b
19a.b
64a,b
175
ND
48
184
42
790a,b
334a'b
442a,b
•i»l*"l rv-itriirvti irv% s*s%t>v
Total PAH
(mq/kq)
673
291
5.8
8.3
78
121
2.4
3.0
260
246
160
230
515
799
0.45
89
31
11
154
246
65
Data point used in the final regression analysis.
ND Not detected.
mg/kg Milligram per kilogram.
45
-------
Section 6
Applications Assessment
The SCAPS technology is designed to be operated
by trained technicians from the AEC, Army Corps of
Engineers, U.S. Navy, WES, and other licensees. The
SCAPS technology is available for use by private
citizens or corporations, although it is available to state
and federal agencies. Hogentogler and Applied
Research Associates, Inc., have nonexclusive licenses
from WES to use the LIF sensor with cone
penetrometry. A similar technology is operated by Loral
Corporation, under secondary license from Hogentogler.
The SCAPS technology's current usage has been focused
on contamination detection and delineation at military
installations. The target contaminants are primarily
PAHs, and most often this technology is applied at
petroleum fuel release sites. As demonstrated, this
technology can rapidly acquire and plot data defining
zones of general contamination if the contamination has
a fluorescent signature. This data can greatly facilitate
site characterization activities.
The qualitative assessment portion of this
demonstration showed that this technology is comparable
to reference methods in its ability to map subsurface
contaminant plumes at petroleum fuel and coal tar
contamination sites. This demonstration showed that
both the SCAPS LIF sensor and the reference methods
identified similar zones of subsurface petroleum and coal
tar contamination at each of the three sites. Many of the
differences between the SCAPS and the reference
methods can be explained by their respective methods of
data collection. The technology produces a continuous
profile, while the reference methods take a few selective
samples and target boundaries and zones of
contamination. In addition, the reference methods had
difficulty retrieving samples in running sands, adding
potential data gaps. The technology produced
continuous data without the need to physically retrieve
samples. The SCAPS technology can produce relatively
continuous data on petroleum or coal tar contaminant
distribution over a 35-foot depth in approximately 1 to
1.5 hours. The reference methods would be able to
collect samples over this interval, however, definitive
analytical services would require, at best, several days,
and the costs associated with analyzing continuous
samples collected every 2 niches would be prohibitive.
Even if the reference methods used on-site analysis and
produced only screening level data, it would take several
hours to provide data on the samples. Therefore, on-
time critical projects that can use screening level data, or
on projects where it is more critical to cover large areas
in greater detail, the SCAPS technology seems to have
distinct advantages. The cost of this technology is
comparable to conventional approaches, except that this
technology produces greater resolution for similar cost.
However, this resolution is at a lower data quality level
than the reference methods.
Another powerful aspect of this technology is that it
provides continuous descriptions of the subsurface soil
concurrently with the chemical data. This demonstration
found that the subsurface logging capabilities of the
SCAPS CP sensor was of comparable accuracy to the
reference methods, however, it appeared to exhibit
greater resolution. Site-specific calibration borings were
not used for this demonstration, and the technology still
produced acceptable accuracy for subsurface
stratigraphic logging.
The quantitative data assessment for this technology
indicated that the resultant LIF data may be correlated to
VPH, TPH, ethylbenzene, xylene, and total BTEX
concentrations. In addition, this data suggests that a
detection threshold for the SCAPS may be around
157 fluorescence units for TPH and 336 fluorescence
units for VPH. These values generally matched the
lowest TPH and VPH concentrations measured. The
lowest TPH and VPH concentrations measured by the
reference methods were 60 and 19 mg/kg, respectively.
Both of these samples exhibited fluorescence above
background. Due to the data set sizes, the predictive
models based on this data should only be used for the
most general estimates. The original reference data sets
were reduced by as much as 50 percent when data points
exhibiting excessive heterogeneity were eliminated. The
46
-------
qualitatively identify changes in waste characteristics and
possibly types. The regression analysis showed some
correlation between the technology's results and individual
compounds, however, this may have been an artifact of
then- relatively systematic distribution within a larger class
of compounds, TPH or VPH, most closely being
monitored. Based on the results of this demonstration, the
use of site-specific calibration samples for the application
of the SCAPS LIF sensor may increase its performance in
a qualitative node, however, it seems unlikely that they
would improve its quantitative performance due to matrix
and contaminant interferences. Site-specific calibration
samples were not used during this demonstration and the
technology still produced similar contaminant distributions
to the reference methods. Even with site-specific
calibration, in the configuration deployed hi this
demonstration, it is not likely that the technology can
produce definitive data, however, site-specific calibration
may allow an estimation of relative contaminant
concentrations. This would be true if the observed
correlations were real.
Based on this demonstration, this technology appears
to produce screening level data for both physical and
chemical characterization sensors. The failure to achieve
better quantitative correlations for the chemical data may
not be wholly attributable to the technology performance.
This may have been due to the relatively small reference
method data set size, the lack of a reference method that
measures the same suite of compounds as the SCAPS LIF
sensor monitors, the complex interactions between the
fluorescing compounds and the soil matrix which resulted
in the observed heterogeneity. The first two factors can
be addressed with changes hi experimental design and
innovations in analytical methods, however, the final
factor will require more research to isolate specific matrix
interactions, and the heterogeneity issue may not be
solvable given current technology.
If the SCAPS LIF sensor performance is to be
evaluated in the field, this demonstration has shown that
on a point-by-point quantitative basis, it is possible that
little to no correlation to reference data will be observed.
This is due to a combination of heterogeneity effects,
limitations hi conventional sampling and analysis, and the
complex interaction of waste aging and constituent
distribution of relative fluorescence. Therefore, based on
the results of this demonstration, field evaluations of this
technology should be restricted to qualitative evaluations
consisting of cross section comparisons and comparisons
of normalized LJF and to verify that LIF highs correspond
to higher levels of contamination. This latter comparison
will also be affected by effects listed above.
In the configuration used during this demonstration,
the SCAPS LIF and CP sensors provided screening level
chemical and stratigraphic data in real time, at a rate faster
than conventional approaches, and with apparently greater
resolutions. The LIF data was relatively correlated to the
reference chemical data hi that both data sets tended to
identify the same zones of high, medium, and low
contamination. The added benefit of sensors mat function
without physical sampling allows them to produce data hi
subsurface environments that prohibit conventional
sampling. An example of such an environment are the
running sands encountered at the Atlantic and Fort Riley
sites. The cost of this technology is comparable to
reference methods, hi fact, on a per-data point basis, this
technology is much less expensive than reference methods.
Although there are many advantages to this
technology, a potential user should be aware of several
disadvantages. This technology has a sampling volume
several thousand times smaller than conventional sampling
analysis. This makes the technology more sensitive to
matrix heterogeneity. Some of this sensitivity is reduced
(vertically) by the averaging of
10 data, points every 2 cm. This effect can also be
minimized by the sampling of more push locations to
reduce ithe sensitivity hi a horizontal orientation. At a
developear-claimed data collection rate up to 400 linear feet
per day (6,096 data points), additional pushes can be
conducted without greatly increasing project duration.
The LIF results can be influenced by the age and
constituent distribution of wastes. This coupled with
heterogeneity effects, and a lack of instrument calibration,
makes quantitation or field verification of LIF results
difficult. The use of the LIF and CP sensors is restricted
to the maximum push depth of the cone penetrometer
truck. This depth can be as much as 150 feet, or in the
case of this demonstration, 30 to 70 feet. These shallow
depths were realized when deeper strata exhibited
increased cone tip resistance and sleeve friction, and at
locations where strata at shallower depths would not
provide adequate lateral support for the push rod. These
condition greatly increase the chance for push rod
breakage and sensor loss.
This technology can currently provide rapid
assessment of the distribution of fluorescent material hi the
subsurface. When these materials are PAHs or petroleum
fuels, the technology can be used to map the general extent
of subsurface contamination. This data can be used to
guide critical conventional soil sampling, and the
placement of groundwater monitoring wells. All of this
data can be produced and interpreted hi the field. This
real-time sampling and analysis allows the use of
contingency-based sampling which assists hi character-
izing a site with a single mobilization. These aspects
coupled with its low volume waste production during
decontamination make this technology a powerful and
effective, site characterization tool.
47
-------
can be used to guide critical conventional soil sampling,
and the placement of groundwater monitoring wells. All
of this data can be produced and interpreted in the field.
This real-tune sampling and analysis allows the use of
contingency-based sampling which assists in character-
izing a site with a single mobilization. These aspects
coupled with its low volume waste production during
decontamination make this technology a powerful and
effective site characterization tool.
48
-------
Section 7
Developer Comments and Technology Status Update
The developer of SCAPS submitted both editorial
and technical comments on the draft ITER. Where
appropriate, the editorial comments were addressed.
The developer's technical comments are presented
verbatim below in italics. PRC's response to the
comments is presented below each developer comment
in plain type.
1. The graphical representations, produced by PRC, of
the physical and chemical cross sections may be
sufficient to represent "tradition data," but it is a
poor representation of what was produced by the
our system while it was in the field.
Panel plots from the SCAPS LIF and CP sensors
have been included in the ITER. The data from
both of these sensors is often plotted in color cross
sections to assist in the interpretation of the data.
Color plots for this demonstration were submitted
by the SCAPS operator. These plots generally
show greater resolution than the ones used in the
ITER. The developer's color plots are hi the TER;
they were not added to the ITER due to the
complexities and costs associated with reproducing
color graphics.
2. There is a general editorial comment concerning the
"negative" tone to the discussions. There are
numerous examples of paragraphs starting with a
negative sentence and then followed with several
positive comments. The report could just as easily
be written to highlight the positive aspects of the
technology.
The ITER was reviewed regarding its tone. Where
the tone disproportionately stressed either the
negative or positive, the text was altered to present
a more uniform presentation of the data.
3. Considering the lack of precision and accuracy in
the reference "quantitative" methods, it does not
seem appropriate to judge SCAPS correlation with
those methods. We have never claimed to be more
than a screening tool, and therefore should not be
judged by a tougher standard.
The ITER has been clarified. It now indicates that
the developer claimed the technology demonstrated
was designed to produce screening level data. In
addition, the inclusion of the quantitative evaluation
was; explained as an attempt to develop baseline data
on the current quantitative capabilities of the
technology.
The developer's comments regarding the precision
and accuracy of the reference methods is noted.
The; ITER has been modified to explain and consider
the impact of heterogeneity in the soil matrix, and
the problems observed with the reference methods,
primarily sample collection methods.
4. The site descriptions do not adequately address the
heterogeneous contaminant distributions that were
observed. This can be illustrated, by the variation
observed in some of the replicates of the reference
samples. This variance represents a horizontal
heterogeneity at these sights. In addition, the
vertical heterogeneity observed over the one foot
averaged area in the SCAPS data, indicates that a
nonhomogeneous distribution of the stratigraphy and
contamination exists.
The ITER has been rewritten to address the issue of
heterogeneity at all levels of data comparison.
5. Precision data indicated a high level of precision for
the SCAPS technology, while statements in the
report imply that fluorescence intensity variations
were due to the technology rather than the
heterogeneous distribution of the contaminant.
The ITER has been rewritten to consider the effects
of heterogeneity on all levels of data comparison.
49
-------
Section 8
References
American Society for Testing and Materials (ASTM). 1990. "Standard Test Method for Particle-Size Analysis of
Soils."
Environmental Protection Agency (EPA). 1986. "Laboratory Manual Physical/Chemical Methods." la. Test Methods
for Evaluating Solid Waste. Volume IB.
EPA. 1991. National Functional Guidelines for Organic Data Review. Contract Laboratory Program. June.
EPA. 1993. Data Quality Objectives Process for Superfund - Interim Final Guidance.
Page, A. L., ed. 1982. "Chemical and Microbiological Properties." In. Methods of Soil Analysis. 2nd Edition.
Number 9. Part 2.
PRC Environmental Management, Inc. 1994. "Final Demonstration Plan for the Evaluation of Cone Penetrometer-
and Geoprobe*-Mounted Sensors." August.
University Hygienic Laboratory. 1991. "Method OA-1 for Determination of Volatile Petroleum Hydrocarbons
(Gasoline)." University of Iowa. Iowa City, Iowa.
50
-------
APPENDIX A
Qualitative, Quantitative, Geotechnical, and TOC Data
A-1. Qualitative Reference Laboratory Data for TPH and PAH - Atlantic Site A-1
A-2. Qualitative Reference Laboratory Data for TPH and PAH - York Site A-2
A-3. Qualitative Reference Laboratory Data for TPH and PAH - Fort Riley Site A-3
A-4. Quantitative Reference Laboratory Data - Atlantic Site A-4
A-5. Quantitative Reference Laboratory Data - York Site A-5
A-6. Quantitative Reference Laboratory Data - Fort Riley Site A-6
A-7. Geotechnical and TOC Data - Atlantic Site A-7
A-8. Geotechnical and TOC Data - York Site .. > A-8
A-9. Geotechnical and TOC Data - Fort Riley Site A-8
51
-------
TABLE A-1. QUALITATIVE REFERENCE LABORATORY
DATA FOR TPH AND PAH—ATLANTIC SITE
Node
Number
2
2
2
2
2
3
3
3
3
4
4
4
4
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
Depth
(feet)
1-1.5
8-9
13-14
25-26
35-36
1-2
10-11
20.5-21.5
33.5 - 34.5
2-2.5
6-6.5
6.5-7
7-7.5
7.5-8
8-8.5
8.5-9
9-9.5
9-10
9.5-10
10-10.5
10.5-11
15-16
27-28
1-2
5-6
7-8
27-28
33.5 - 34.5
TPH
(ppm)
1,680
15
24.7
NS
19.3
55.4
1,130
222
54.2
149
330
614
1,650
4,170
541
73.7
1,680
897
2,880
2,960
3,820
1,170
118
399
ND
275
146
ND
PAH
(ppm)
88.19
3.59
0
0
.06
11.40
158.37
6.06
.99
13.39
.258
5.267
21.494
44.205
128.811
71.760
55.482
40.644
80.999
104.487
107.437
62.091
48.879
7.007
0.020
18.496
3.481
0.030
Notes:
ppm Part per million.
NS Not sampled.
ND Not detected.
52
-------
TABLE A-2. QUALITATIVE REFERENCE LABORATORY
DATA FOR TPH AND PAH—YORK SITE
Node
Number
1
1
1
1
2
2
2
2
3
3
3
3
3
3
3
3
4
4
4
4
4
4
5
5
5
5
Notes:
ppm
ND
Depth
(feet)
12-13
14-15
17-18
22 - 22.5
8.5-9
10.5-11
14-14.5
20-21
10-11
12-13
16-16.5
16.5-17
17-17.5
17.5-18
18-18.5
21.5-22.5
8-9
11-12
14-15
17-18
18-18.5
21.5-22
10-11
12-13
17-18
21-22
Part per million.
Not detected.
TPH
(ppm)
26.1
345
13.7
ND
ND
417
855
10.2
10
259
2,570
3,650
57.5
12.7
27.8
ND
115
174
8,150
137
13,100
74.2
23.7
66
377
ND
PAH
(ppm)
1.09
48.81
.88
.01
0
7.72
127.62
.060
0
37.62
134.67
313.97
1.90
0.23
0.20
0.01
0.66
182.09
1,412.16
10.33
1,130.18
14.35
0
9.31
128.09
0.165
53
-------
TABLE A-3. QUALITATIVE
DATA FOR TPH AND PAH-
REFERENCE LABORATORY
-FORT RILEY SITE
Node
Number
1
1
1
2
2
2
2
2
3
3
3
3
4
5
5
5
5
5
5
5
5
5
5
5
Notes:
ppm
ND
Depth
(feet)
1-1.5
18-19
28.5 - 30
5-6
6-7
15-16
23.5 - 25
28.5-30
1.5-2.5
5.5 - 6.5
15-16
23.5 - 24
15-16
2.5 - 3.5
5-6
10-10.5
10.5-11
11-11.5
11.5-12
12-12.5
12.5-13
13-13.5
24-25
29-30
Part per million.
Not detected.
TPH
(ppm)
47.1
482
ND
37.4
233
6,720
89.3
96.9
112
2,670
1,850
ND
37
ND
1,280
6,730
32,800
19,300
9,360
12,700
2,830
2,550
9.94
ND
PAH
(ppm)
0.667
12.964
0.036
0.108
0.142
137.885
1.707
2.713
1.358
118.471
18.730
0
0
0
1.655
143.622
338.566
344.984
206.888
190.693
87.639
64.711
0
0
54
-------
TABLE A-4. QUANTITATIVE REFERENCE LABORATORY DATA—ATLANTIC SITE
Chemical Minimum
Node 2 (21 to 22 feet)
Ethyl-
benzene 25,000.00
TPH 8,850.00
VPH 910.00
Total
PAH 70.42
Total
BTEX 15,400.00
Chemical Minimum
Node 3 (16 to 17 feet)
Ethyl-
benzene 3,600.00
TPH 104.00
VPH 88.00
Total
PAH 4.43
Total
BTEX 25,590.00
Chemical Minimum
Node 4 (10 to 11 feet)
Ethyl-
benzene 29,000.00
TPH 959.00 '
VPH 1,200.00
Total
PAH 12.26
Total
BTEX 218,700.00
Chemical Minimum
Node 5 (16 to 17 feet)
Ethyl-
benzene 390.00
TPH 87.80
VPH 36.00
Total
PAH 0.48
Total
BTEX 5,490.00
Maximum
42,000.00
15,400.00
2,000.00
918.32
293,000.00
Maximum
4,600.00
1,290.00
130.00
6.84
33,550.00
Maximum
35,000.00
3,780.00
1,400.00
148.10
307,000.00
Maximum
2,100.00
516.00
160.00
4.72
28,700.00
Standard
Mean Deviation
ISliilBfi 7,035.62
lilgOJJI 2,689.89
427.22
672.69 354.08
221,200.00 53,049.03
Standard
Mean Deviation
IllPJJQjpJI 469.04
425.25 577.00
21.35
HI 1-05
mafflg 3,728.07
Standard
Mean Deviation
S1}600.g§ 2,408.32
1,129.72
83.67
77.91 60.07
fHI^7,4,QTOIJ 34,028.49
Standard
Mean Deviation
811.80
177.36
59.36
1.79
11,794.98
Minimum Maximum
Node 2 (24 to 25 feet)
250.00 39,000.00
36.00 9,880.00
7.90 1,400.00
3.99 691.99
1,250.00 260,000.00
Minimum Maximum
Node 4 (6.5 to 7.5 feet)
250.00 3,600.00
20.70 412.00
10.00 110.00
3.91 13.73
320.00 37,300.00
Minimum Maximum
Node 4 (27.5 to 28.5 feet)
1,300.00 23,000.00
117.00 3,030.00
42.00 970.00
29.51 270.81
13,700.00 193,000.00
Minimum Maximum
Node 5 (23.5 to 24.5 feet)
940.00 1,700.00
48.20 893.00
52.00 160.00
1.96 5.11
14,330.00 22,200.00
Mean
14,690.00
4,004.40
537.58
290.96
93,950.00
Mean
1,410.00
254.93
43.33
HB
10,435.00
10,520.00
1,093.60
452.40
96,740.00
Mean
238.50
Standard
Deviation
16,663.45
4,592.73
579.81
324.69
109,210.58
Standard
1,897.71
166.39
57.74
4.63
17,931.46
Standard
Deviation
10,146.77
1,156.83
461.91
98.83
85,721.28
Standard
Deviation
301.30
366.76
46.38
1.26
3,340.21
Notes:
Values used in the final regression equations.
55
-------
TABLE A-5. QUANTITATIVE REFERENCE LABORATORY DATA—YORK SITE
Chemical
Minimum
Maximum
Mean
Standard
Deviation
Node 1(1 5 to 16 feet)
Ethyl-
benzene
TPH
VPH
Total
PAH
Total
BTEX
Chemical
200.00
53.00
a
64.65
580.00
Minimum
1,900.00
2,270.00
a
755.70
4,700.00
Maximum
smm
773.20
a
260.31
Mean
*
948.75
905.13
a
284.20
2,317.79
Standard
Deviation
Node 2 (17 to 18 feet)
Ethyl-
benzene
TPH
VPH
Total
PAH
Total
BTEX
Chemical
1,600.00
88.70
5.40
40.41
1,900.00
Minimum
7,200.00
1,380.00
33.00
252.47
14,100.00
Maximum
1SS1S2S
496.94
8,946.0Q
Mean
2,846.58
535.84
12.51
98.64
5,131.92
Standard
Deviation
Node 4 (14 to 15 feet)
Ethyl-
benzene
TPH
VPH
Total
PAH
Total
BTEX
Chemical
2,200.00
647.00
37.00
166.06
4,540.00
Minimum
19,000.00
6,450.00
97.00
1,048.82
36,300.00
Maximum
ilfiSSfl
£!l»§i
19,542.00
Mean
6,412.49
2,361.49
24.87
342.68
11,891.11
Standard
Deviation
Node 5 (1.5 to 2.5 feet)
Ethyl-
benzene
TPH
VPH
Total
PAH
Total
BTEX
Notes:
a
ND
ND
15.20
ND
0.01
ND
No data.
Not detected.
ND
138.00
ND
0.95
ND
ND
ND
0.45
ND
ND
67.83
ND
0.40
ND
Minimum
Maximum Mean
Standard
Deviation
Node 2 (13.5 to 14.5 feet)
1,800.00
156.00
14.00
159.85
1,800.00
Minimum
Node 3 (17 to 18 feet)
290.00
261.00
7.50
97.45
1,090.00
Minimum
Node 4 (18 to 19 feet)
92.00
18.20
6.00
1.40
13,000.00 IflgJlgfl
2,710.00 1,539.20
45.00
466.41 Msum
23,280.00 7,556.00
Maximum Mean
2,700.00 W§i$i§M
1 ,450.00 SSBUfl
30.00 Jg5J
359.67
4,900.00 msam
Maximum Mean
57,000.00 21,810.50
4,000.00 1,878.15
280.00: 175.33
2,332.21 798.91
274.00 128,800.00 42,528.80
4,714.69
1,035.82
11.90
129.85
8,853.57
Standard
Deviation
874.25
513.04
9.30
116.76
1,641.76
Standard
Deviation
27,363.23
2,155.00
148.01
1,116.53
59,970.23
Values used in the final regression equations.
56
-------
TABLE A-6. QUANTITATIVE REFERENCE LABORATORY DATA—FORT RILEY SITE
Chemical
Minimum
Maximum
Mean
Standard
Deviation
Node 1 (2 to 3 feet)
Ethyl-
benzene
TPH
VPH
Total
PAH
Total
BTEX
Chemical
79.00
27.50
6.00
0.97
339.00
Minimum
3,700.00
15,800.00
110.00
260.60
20,610.00
Maximum
1,075.80
5,727.98
47.50
89.15
6,412.40
Mean
1,502.51
6,698.52
44.90
105.57
8,295.04
Standard
Deviation
Node 2 (6 to 7 feet)
Ethyl-
benzene
TPH
VPH
Total
PAH
Total
BTEX
Chemical
107.50
48.60
9.00
0.05
89.00
Minimum
2,000.00
7,720.00
98.00
42.05
10,070.00
Maximum
762.50
2,169.32
41.83
10.98
2,956.63
Mean
1,072.32
3,186.75
48.87
18.15
4,756.48
Standard
Deviation
Node 5 (10.5 to 11. 5 feet)
Ethyl-
benzene
TPH
VPH
Total
PAH
Total
BTEX
1,400.00,
17,700.00
250.00
157.14
23,270.00
29,000.00
32,800.00
430.00
340.26
96,300.00
13,680.00
waflfwm
SfeSbEHS
24M8
11,238.86
5,934.81
80.81
67.41
27,580.73
Minimum
Maximum Mean
Standard
Deviation
Node 1(1 3 to 14 feet)
100.00
32.90
9.20
0.02
230.00
Minimum
20,000.00 8,595.00
3,110.00 1,416.00
320.00 183.55
96.62 31.44
70,000.00 26,497.60
Maximum Mean
10,009.87
1,607.95
152.78
44.47
35,204.28
Standard
Deviation
Node 2 (17 to 18 feet)
28,000.00
7,050.00
530.00
60.66
147,000.00
Minimum
60,000.00
16,900.00 SBIIiiPJl
1,200.00 SM
224.76
254.ooo.oo wmmm
Maximum Mean
13,464.77
4,182.11
259.71
72.14
47,704.30
Standard
Deviation
Node 5 (16 to 17 feet)
20,000.00
1,090.00
170.00
18.23
63,100.00
55,000.00 ESiHi
9,630.00 £9JffipJ
930.00
162.28 gg^l
219,700.00
13,612.49
3,470.96
289.34
58.45
62,975.05
Notes:
Values used in the final regression equations.
57
-------
TABLE A-7. GEOTECHNICAL AND TOC DATA—ATLANTIC SITE
Node/
Grid
1/F
1/F
1/F
1/F
1/F
4/C
4/C
Notes:
mg/kg
mm
Depth
(feet)
2-3
10-11
20.5-21
30.5 - 31
35 - 35.5
9-10
15-16
TOC
(mq/kq) %>2 mm
4.000 .03
ND 0
600 1
200 28.38
400 0
3,800 0
3,200 0
% Sand
(0.5-2 mm)
12.43
36
50.84
62.78
24.72
19.34
44.79
% Silt
(2-50 «m)
58.33
43.78
34.17
4.72
44.73
51.24
30.57
% Clay
(<2 urn)
29.21
20.22
13.99
4.12
30.55
29.42
24.64
USDA
Classification
Silty clay loam
Loam
Loam
Sand
Clay loam
Silty clay loam
Loam
uses
Classification
Sandy lean clay (CL)
Silt or clay (CL or ML)
Silt or clay (CL or ML)
Well to poorly graded
sand (SW or SP)
Sandy lean clay or
sandy silt (CL or ML)
Sandy lean silt or
sandy lean clay (CL or
ML)
Silt or clay (CL or ML
Milligram per kilogram.
Millimeter.
urn Micrometer.
USDA United States Department of Agriculture.
USCS Unified Soil Classification System, ( ) two-letter classification code.
ND Not detected.
58
-------
TABLE A-8. GEOTECHNICAL AND TOC DATA—YORK SITE
Node/
Grid
1/G
1/G
1/G
1/G
3/C
3/C
Depth
(feet)
5-6
7-8
15-15.5
18.5-19
12-13
16.5-17
TOC
(ma/ka)
ND
2,800
1,400
490
3,200
2,600
%>2 mm
0.00
0.05
0.23
30.54
0.00
6.69
% Sand
(0.5-2 mm)
13.66
26.08
60.24
46.38
8.90
52.48
% Silt
(2-50 Mm)
58.94
51.05
20.93
12.09
60.43
17.92
% Clay
(<2 itm)
27.40
22.82
18.60
10.99
30.67
22.91
USDA
Classification
Silly clay loam
Silt loam
Sandy loam
Sandy loam
Silty clay loam
Sandy clay loam
uses
Classification
Clay or silt with sand
(CL or ML)
Clay or silt with sand
(CL or ML)
Silty to Clayey sand
(SM or SC)
Poorly graded sand
with silt or clay
(SW-SC or SP-SC)
Silt or lean clay with
sand (CL or ML)
Clayey or silly sand
(SM or SC)
Notes:
mg/kg Milligram per kilogram.
mm Millimeter.
MID Micrometer.
USDA United States Department of Agriculture.
USCS Unified Soil Classification System, ( ) two-letter classification code.
ND Not detected.
TABLE A-9. GEOTECHNICAL AND TOC DATA—FORT RILEY SITE
Node/
Grid
4/H
4/H
4/H
4/H
2/E
3/G
Notes:
mg/kg
mm
Mm
USDA
USCS
Depth TOC
(feet) (ma/kg)
2 - 3 3,400
7.5 - 8.5 600
15-16 800
29 - 30 300
15-16 4,600
5.5 - 6.5 9,000
%>2 mm
0.00
.16
0.00
20.36
.11
.10
% Sand
(0.5-2 mm)
31.32
60.76
62.44
57.48
55.13
47.61
% Silt
(2-50 urn)
43.48
22.08
19.16
10.46
25.87
36.02
% Clay
25.20
17.00
18.40
11.70
18.89
16.27
USDA
Classification
Loam
Sandy loam
Sandy loam
Sandy loam
Sandy loam
Loam
Milligram per kilogram.
Millimeter.
Micrometer.
United States Department of Agriculture.
Unified Soil Classification System, ( ) two-letter classification code.
*U.S. GOVERNMENT PRINTING OFFICE: 1995-653-424
59
-------
------- |