&EPA
•United States
Environmental Protection
Agency
Office of Research
and Development
Washington DC 20460
EPA/600/R-97/020
February 1997
The Rapid Optical Screening
Tool (ROST™) Laser-Induced
Fluorescence (LIF) System
for Screening of Petroleum
Hydrocarbons in Subsurface
Soils
359ASB96
-------
Internet Address (URL) • http://www.epa.gov
Recycled/Recyclable • Printed with Vegetable Oil Based Inks on Recycled Paper (20% Postconsumer)
-------
The Rapid Optical Screening Tool (ROST™)
Laser-Induced Fluorescence (LIF) System for
Screening of Petroleum Hydrocarbons in
Subsurface Soils
Innovative Technology
Verification Report
by
Grace Bujewski
Brian Rutherford
Sandia National Laboratories
National Exposure Research Laboratory
Characterization Research Division
U.S. Environmental Protection Agency
Las Vegas, Nevada 89193
-------
Notice
The U.S. Environmental Protection Agency (EPA), through its Office of Research and Development
(ORD), partially funded and managed the extramural research described here. It has been peer reviewed
by the Agency and approved as an EPA publication. Mention of trade names or commercial products
does not constitute endorsement or recommendation by EPA for use.
11
-------
\
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
Washington, D.C. 20460
ENVIRONMENTAL TECHNOLOGY
VERIFICATION PROGRAM
ENVIRONMENTAL TECHNOLOGY VERIFICATION PROGRAM
VERIFICATION STATEMENT
TECHNOLOGY TYPE: CONE PENETROMETER-DEPLOYED SENSOR
APPLICATION: IN-SITU DETECTION OF PETROLEUM HYDROCARBONS
TECHNOLOGY NAME: RAPID OPTICAL SCREENING TOOL (ROST)
COMPANY: FUGRO GEOSCIENCES, INC.
ADDRESS: 6105 ROOKIN, HOUSTON, TEXAS 77074
PHONE:
713/778-5580
The U. S. Environmental Protection Agency (EPA) has created a program to facilitate the deployment of
innovative technologies through performance verification and information dissemination. The goal of the
Environmental Technology Verification (ETV) Program is to further environmental protection by
substantially accelerating the acceptance and use of improved and more cost effective technologies. The
ETV Program is intended to assist and inform those involved in the design, distribution, permitting, and
purchase of environmental technologies. This document summarizes the results of a demonstration of a
cone penetrometer-deployed Rapid Optical Screening Tool (ROST™) marketed by Fugro Geosciences,
Inc.
PROGRAM OPERATION
The EPA, in partnership -with recognized testing organizations, objectively and systematically evaluates
the performance of innovative technologies. Together, with the full participation of the technology
developer, they develop plans, conduct tests, collect and analyze data, and report findings. The
evaluations are conducted according to a rigorous demonstration plan and established protocols. EPA's
National Exposure Research Laboratory which conducts demonstrations of field characterization and
monitoring technologies, selected the U. S. Department of Energy's Sandia National Laboratories as a
testing organization.
DEMONSTRATION DESCRIPTION
In May and October, 1995, two cone penetrometer-deployed sensor systems were demonstrated to
evaluate how well they could measure subsurface petroleum hydrocarbon contamination. The
performance of each system was evaluated by comparing field analysis results to those obtained using
conventional sampling and analytical methods. These methods included using a hollow stem auger in
conjunction with a split spoon sampler and subsequent analysis of the collected sample by a reference
laboratory using EPA Method 418.1 for total recoverable petroleum hydrocarbons (TRPH) and California
Department of Health Services Method 8015-Modified for total petroleum hydrocarbons.
EPA-VS-SCM-01
The accompanying notice is an integral part of this verification statement
February
-------
The primary objectives of the demonstration were to (1) verify technology performance, (2) determine
how well the developer's field instrument performs hi comparison to conventional laboratory methods, (3)
determine the logistical and economic resources needed to operate the instrument, and (4) produce a
verified data set for use in considering the technology for future use in hazardous waste investigations.
Field demonstrations were conducted at two geologically and climatologically different sites: (1) the
Hydrocarbon National Test Site located at Naval Construction Battalion Center (NCBC) Port Hueneme,
California, and (2) the Steam Plant Tank Farm at Sandia National Laboratories (SNL), Albuquerque, New
Mexico. The conditions at each of these sites represent what are considered typical under which the
technology would be expected to operate, but are not considered all inclusive. Details of the
demonstration, including a data summary and a discussion of results may be found in the report entitled
"The Rapid Optical Screening Tool (ROST™) Laser Induced Fluorescence System for Screening
Petroleum Hydrocarbons hi Subsurface Soil." The EPA document number for this report is EPA/600/R-
97/020.
TECHNOLOGY DESCRIPTION
The ROST™ sensor evolved from the tunable laser instrumentation originally developed at North Dakota
State University (NDSU) with U.S. Air Force research support. The technology had been commercialized
and marketed by a consortium of government and industry led by Loral Corporation and Dakota
Technologies. ROST™ was acquired by Fugro Geosciences, Inc., in May 1996 and is now offered as an
integrated service with their cone penetrometer (CPT) systems worldwide.
The sensor uses a wavelength tunable ultraviolet laser source coupled with an optical detector to measure
fluorescence via optical fibers, a technique known as laser-induced fluorescence spectroscopy (LIF). The
measurement is made through a sapphire window on a probe that is pushed into the ground with a truck-
mounted cone penetrometer. The optical fibers are integrated with the geotechnical probe and umbilical of
the CPT system.
During the period of this demonstration, the ROST™ technology was available for use within the 48
contiguous United States for a cost of approximately $5,300 per day or site-specific footage rates, which
ncludes a CPT rig provided by a commercial vendor. Crew per diem and mobilization costs are additional
and site specific. The ROST™ subassembly can be integrated with any commercially available industry-
standard CPT rig. Typical crew members include a ROST™ system operator, CPT operator, and an
assistant. Under normal conditions, an average of 300 feet of pushes can be completed in a day. This
translates to a total cost of under $20 per foot. As of January 1997, Fugro has reduced the cost for the
ROST™ and a CPT to approximately $4250 per day.
VERIFICATION OF PERFORMANCE
fhe findings of the demonstration are as follows:
The ROST™ system was easily integrated with a conventional cone penetrometer truck. Full
integration was accomplished in less than two hours.
EPA-VS-SCM-01
The accompanying notice is an integral part of this verification statement
iv
February 1997
-------
Data was collected every 0.2 ft. or less if the cone slowed or stopped. Push rate is dependent on the
CPT. Standard data collection rate is one sample per 1.2 seconds.
At Port Hueneme, the correlation with conventional TPH analysis was 89.2% with 5.4% false
negatives. At the Sandia tank farm, the TPH correlation was 93.4% with 3.3% false negatives.
Real time data acquisition was achieved at both sites.
The results of this study satisfied the requirements of the demonstration plan. The ROST™ system
successfully located the perimeter of the plume and showed acceptable correlation to conventional
methods. The false negative rate combined from both demonstrations was less than 5% and was within the
performance specifications of the instrument. Any disagreements with the laboratory results were
primarily confined to regions where contaminant concentration levels were close to the detection
threshold. A portion of these discrepancies could be the result of variability in laboratory results where
random errors are estimated to be in the range of 10 to 15 percent.
The ROST™ system is an emerging technology worthy of consideration for site investigations where
aromatic hydrocarbons (e.g., petroleum, oils, lubricants, and coal tars) are suspected. The technology
offers a number of advantages over conventional drilling and sampling technologies for the purpose of
screening a site to determine the nature and extent of contamination. The information obtained from this
technology could provide a complete picture of the contamination and it can be used to predict optimal
sampling locations. As with any technology, there are some limitations which a prospective user should
be aware when designing an environmental investigation. Stratigraphy and unidentifiable fluorescent
interferences are issues that could prevent the sole use of the ROST™ LIF system. The technology has
been used to identify lighter fuels but this capability was not evaluated in these demonstrations. Because
the technology "does not provide species-specific quantitation, it should be used in conjunction with
conventional sampling and analysis if risk assessment or cleanup criteria must be met. As a screening
technology to identify the nature and extent of aromatic hydrocarbon contamination, this technology has
many advantages over conventional techniques.
Gary J.
I Director
National Exposure Research Laboratory
Office of Research and Development
NOTICE- EPA verifications are based on an evaluation of technology performance under specific, predetermined criteria and the
appropriate quality assurance procedures. EPA makes no expressed or implied warranties as to the performance of the technology and
does not certify that a technology will always, under circumstances other than those tested, operate at the levels verified. The end user
is solely responsible for complying with any and all applicable Federal, State and Local requirements.
EPA-VS-SCM-01
The accompanying notice is an integral part of this verification statement
February 1997
-------
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 Office of Research and
Development (ORD) provides data and science support that can be used to solve environmental problems
and to build the scientific knowledge base needed to manage'our ecological resources wisely, to
understand how pollutants affect our health, and to prevent or reduce environmental risks.
The National Exposure Research Laboratory (NERL), ORD, is the Agency's center for the investigation
of technical and management approaches for identifying and quantifying risks to human heath and the
environment. One focus of the Laboratory's research program is to develop and evaluate technologies
for the characterization and monitoring of air, soil, water and subsurface resources. This in turn, will
provide the scientific information needed by EPA to support regulatory and policy decisions; and to
provide the science support needed to ensure effective implementation of environmental regulations and
strategies.
Effective measurement and monitoring technologies are needed to assess the degree of contamination at a
site, to provide data which may be used to determine the risk to public health or the environment, to
supply the necessary cost and performance data to select the most appropriate technology, and to monitor
the success or failure of a remediation process.
Candidate technologies can originate from within the federal government or from the private sector.
Through this program, developers are given the opportunity to conduct a rigorous demonstration of their
technology's performance under realistic field conditions. By completing the evaluation and distributing
the results, the Agency establishes a baseline for acceptance and use of these technologies. The
Characterization and Monitoring portion of this program is administered by NERL's Characterization
Research Division in Las Vegas, Nevada.
Gary J. Foley, Ph.D.
Director
National Exposure Research Laboratory
VI
-------
Table of Contents
Section
Notice • ii
Verification Statement iii
Foreword vi
List of Figures • x
List of Tables *» x
List of Abbreviations and Acronyms xi
Acknowledgment xiii
EXECUTIVE SUMMARY 1
Technology Description) 1
Demonstration Objectives and Approach 2
Demonstration Results and Performance Evaluation 3
Cost Evaluation 3
INTRODUCTION 4
The Site Characterization Technology Challenge 4
Technology Demonstration Process 4
Technology Selection 4
Technology Demonstration 5
Technology Performance Assessment, Evaluation, and Verification 5
Information Distribution 6
.The CPT-LIF Sensor Demonstrations 6
ROST™ LBF TECHNOLOGY DESCRIPTION r 7
ROST™LIF Sensor 7
Laser Source 8
System Components 8
Dynamic Range 10
Sensitivity, Noise, and Background 10
Calculated Fluorescence and Concentration Thresholds 11
Mobilization and Installation of ROST™ for CPTLIF work 11
Deployment Costs 11
Technology Applications 12
Advantages of the Technology 12
Detectors and Data Acquisition Modes 13
Limits of the Technology 13
Response to Different Petroleum Hydrocarbons 13
Matrix Effects 13
Spectral Interferences 14
Truck-Mounted Cone Penetrometer Access Limits 14
Cone Penetrometer Advancement Limits 15
REFERENCE LABORATORY RESULTS AND EVALUATION 16
Selection of Reference Laboratory and Methods 16
vu
-------
Assessment of Laboratory Data Quality 17
Audits 17
Sample Holding Times 17
Sample Preparation 17
Sample Analysis 18
Detection Limits 18
Quality Control Procedures 18
Accuracy, Precision, and Completeness 18
Use of Qualified Data for Statistical Analysis 20
DEMONSTRATION DESIGN AND DESCRIPTION 21
Evaluation of ROST™LIF Sensor Performance 21
Description of Demonstration Sites 21
Port Hueneme Site Description 21
Demonstration Sampling Operations, Port Hueneme 24
Port Hueneme Sampling Locations 26
SNL Tank Farm Site Description 27
SNL Tank Farm Predemonstration Sampling 29
Demonstration Sampling Operations, SNL Tank Farm 30
Calibration Procedures, Quality Control Checks, and Corrective Action 31
ROST™ LIF Initial Calibration Procedures ."."".".31
ROST™ LIF Continuing Calibration Procedures 32
Method Blanks 32
Spike Samples 32
Instrument Check Standards 32
Performance Evaluation Materials 32
Duplicate Samples 32
Equipment Rinsate Samples 32
Data Reporting, Reduction, and Verification Steps 33
Data Reporting 33
Data Reduction and Verification Steps for the ROST™ LIF Data 33
Changes to the Demonstration Plan 34
TECHNOLOGY RESULTS AND EVALUATION 36
Developer Claims Presented 36
Technology Data Quality Assessment 36
Precision 37
Completeness 37
Accuracy 37
Port Hueneme Site Data Presentation and Results 39
Port Hueneme Detection Limit 39
Downhole Results for Port Hueneme 39
Port Hueneme Subsurface Contaminant Mapping 40
SNL Tank Farm Site Data Presentation and Results 43
SNL Tank Farm Detection Limit 43
Downhole Results for SNL Tank Farm .""45
SNL Tank Farm Subsurface Contaminant Mapping 46
Cost Evaluation .........41
Overall Performance Evaluation 48
vni
-------
APPLICATIONS ASSESSMENT 49
Advantages of the Technology 49
Real-Time Analysis 49
Continuous LIF Data Output 49
Continuous Lithological Logging 49
Cost Advantages 50
Enhanced Operator Safety 50
Performance Advantages 50
Limitations of the Technology 51
Applicability 51
Quantisation and Speciation 51
Push Limitations of CRT 52
Interferences ; 52
Conclusions 52
DEVELOPER FORUM , 53
High Mobility and Productivity Rate of Fugro CPT/ROST™ 53
Contaminant Applicability and Product Identification Capability 53
ROST™'s Delineation Capabilities 53
CPT/ROST™ Data Presentation 54
ROST™ Upgrades 54
Fugro's Existing and Emerging Technologies 54
PREVIOUS COMMERCIAL PROJECTS 55
REFERENCES 57
APPENDICES
Appendix A: Reference Laboratory Data A-l
Appendix B: ROST™ LIF Field Data Logs B-l
Appendix C: ROST™ LIF Draft EMMC Method C-l
IX
-------
List of Figures
3-1 ROST™ system components
5-1 Site vicinity map, Port Hueneme ,
5-2 Demonstration site and sampling locations, Port Hueneme
5-3 Site vicinity map, SNL Tank Farm
5-4 Demonstration site and sampling locations, SNL Tank Farm
6-1 Downhole results for Port Hueneme
6-2 Adjusted downhole results for Port Hueneme
6-3 Cross-sectional map of transect near Tank 5114 at Port Hueneme,
6-4 Downhole results for SNL Tank Farm
6-5 Subsurface contaminant map for SNL Tank Farm
....9
..22
..25
..28
,.29
,.41
.42
.43
.45
.46
List of Tables
Table
3-1 Total petroleum hydrocarbon analysis methods
4-1 Quality control results for TPH
4-2 Quality control results for TRPH
5-1 Port Hueneme boring and push summary table
5-2 SNL Tank Farm boring and push summary table
6-1 Summary of comparison of Port Hueneme LIF data with laboratory data.
6-2 Summary of comparison of SNL LIF data with laboratory data
6-3 Cost comparison
6-4 ROST™ LIF claims evaluation
7-1 Performance statistics
9-1 Summary of selected CPT/ROST™ commercial projects
..12
..19
,.19
,.26
.31
.39
.44
.47
.48
.51
.55
-------
List of Abbreviations and Acronyms
AEC
ASTM
ATI
bbl
bgs
CalEPA-DTSC
CAS
cm
CPT
CSC
CSCT
DFM
DHS
DoD
DOE
DOT
DQO
EMMC
EPA
ETI
ETV
ft
FVD
GC/FID
HNTS
HSA
Hz
row
IR
IRP
ITVR
LIF
LOD
m
mg/kg
mg/L
m/min
Army Envkonmental Center
American Society for Testing and Materials
Analytical Technologies, Inc.
Barrel (Equivalent to 42 U.S. Gallons)
Below Ground Surface
State of California Envkonmental Protection Agency, Department of
Toxic Substances Control
Chemical Abstracts Service
Centimeter
Cone Penetrometer Testing
Computer Sciences Corporation
Consortium for Site Characterization Technology
Diesel Fuel Marine
Department of Health Services (State of California)
Department of Defense
Department of Energy
Department of Transportation
Data Quality Objective
Envkonmental Monitoring Management Council
U.S. Environmental Protection Agency
Environmental Technology Initiative
Envkonmental Technology Verification
Feet
Fluorescence Versus Depth
Gas Chromatograph/Flame lonization Detector
Hydrocarbon National Test Site
Hollow Stem Auger
Hertz
Investigation Derived Waste
Infrared
Installation Restoration Program
Innovative Technology Verification Report
Laser-Induced Fluorescence
Limit of Detection
Meter
Micrometer
Milligrams per Kilogram
Milligrams per Liter
Meters per Minute
XI
-------
List of Abbreviations and Acronyms (Continued)
ml
mL
mm
MS
MSD
ms
msl
NCBC
NCCOSCRDT&E
NERL-CRD
nm
ns
PAH
PDA
PE
PPE
ppm
PRC
QA
QAPP
R
ROST
RPD
SCAPS
SNL
SOP
SPT
TER
TPH
TRPH
TSF
U.S.
uses
UV
WES
WTM
Millijoules
Milliliter
Millimeter
Matrix Spike
Matrix Spike Duplicate
Millisecond
Mean Sea Level
Naval Construction Battalion Center
Naval Command, Control, and Ocean Surveillance Center Research,
Development, Test, and Evaluation (Division)
National Exposure Research Laboratory-Characterization Research
Division
Nanometer
Nanosecond
Polycyclic Aromatic Hydrocarbons
Photodiode Array
Performance Evaluation
Personal Protective Equipment
Parts per Million
PRC Environmental Management, Inc.
Quality Assurance
Quality Assurance Project Plan
Quality Control
Correlation Coefficient
Remedial Investigation/Feasibility Studies
Rapid Optical Screening Tool™
Relative Percent Difference
Site Characterization and Analysis Penetrometer System
Sandia National Laboratories (Department of Energy)
Standard Operating Procedure
Standard Penetrometer Testing
Technology Evaluation Report
Total Petroleum Hydrocarbons
Total Recoverable Petroleum Hydrocarbons
Tons per Square Foot
United States
Unified Soil Classification System
Ultraviolet
Waterways Experimental Station (Army Corps of Engineers)
Wavelength-Time Matrix
XII
-------
Acknowledgment
We wish to acknowledge the support of all those who helped plan and conduct the demonstrations,
interpret data, and prepare this report. In particular, for demonstration site access and relevant
background information, Stephan McCarel and Ernest Lory (Naval Facilities Engineering Service
Center), and David Miller (SNL); for implementation of the demonstration plan and data evaluation,
Grace Bujewski, Brian Rutherford, Robert Knowlton, Robert Helgesen, and Peter Stang (PRC); for
logistical and health and safety support, Michael Skelly (Roy F. Weston, Inc.) and Michael Strosinsky
(SNL); for editorial and publication support, Merlyn Liberty (Tech Reps, Inc.); for peer and technical
reviews, Jerry Peace (SNL); Bruce LaBelle (State of California Environmental Protection Agency,
Department of Toxic Substances Control); and for U.S. EPA project management, Stephen Billets and
Bob Lien (U.S. EPA). In addition, we gratefully acknowledge the participation of the ROST™
technology developers, Greg Gillispie and Randy St. Germain, Dakota Technologies, Inc. (701-237-
4908); Mr. Andy Taer, Fugro Geosciences (713) 778-5580; and Mr. George Robitaille, Tri-Services
SCAPS Program of the U.S. Army Environmental Center (410-612-6865).
Xlll
-------
-------
Section 1
Executive Summary
The Consortium for Site Characterization Technology (CSCT) has established a formal program
to accelerate acceptance and application of innovative monitoring and site characterization
technologies that improve the way the nation manages its environmental problems. The CSCT is
a partnership program involving the U.S. Environmental Protection Agency (EPA), the
Department of Defense (DoD), and the Department of Energy (DOE). Its mission is to support
the demonstration and verify the performance of new and emerging technologies.
In 1995 the CSCT conducted a demonstration of two in situ laser-induced fluorescence-based
technologies using the Site Characterization and Analysis Penetrometer System (SCAPS) cone
penetrometer testing (CPT) platform. The two technologies were the Rapid Optical Screening
Tool™ (ROST™) developed by Loral Corporation and Dakota Technologies, Inc., and the
SCAPS LIF, developed through a collaborative effort of the Army, Navy, and Air Force under
the Tri-Services SCAPS program and by the Naval Command, Control, and Ocean Surveillance
Center, Research, Development Test, and Evaluation (NCCOSC RDT&E) Division. 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 distinguish contaminated and
noncontaminated areas. Results for the SCAPS LIF and CPT performance evaluation are
presented in a separate report.
The purpose of this Innovative Technology Verification Report (ITVR) is to document the
demonstration activities and present and evaluate the demonstration data hi order to verify the
performance of the ROST™ LIF sensing technology relative to developer claims as presented in
the approved demonstration plan.
Technology Description
The ROST™ LIF sensor provides real-time field screening of the physical characteristics of soil
and chemical characteristics of aromatic petroleum hydrocarbon contamination at hazardous
waste sites. The current configuration is designed to quickly and cost-effectively distinguish
aromatic hydrocarbon-contaminated areas from uncontaminated areas. The ROST™ system
mounted on a standard cone penetrometer truck is also capable of acquiring geologic information
and has the added benefit of reduced generation of investigation-derived waste. This capability
allows further investigation and remediation decisions to be made more efficiently and reduces
the number of samples that must be submitted to laboratories for analysis.
The ROST™ sensor evolved from the tunable laser instrumentation originally developed at North
Dakota State University (NDSU) with U.S. Air Force research support. The NDSU tunable laser
system was first deployed for LIF-CPT hi 1992 in a demonstration project at Tinker AFB. The
technology developers from NDSU then formed a small business, Dakota Technologies, Inc.
(DTI) and participated in additional demonstrations of LIF-CPT projects. The technology has
been commercialized and marketed by a consortium of government and industry led by Loral
Corporation. ROST™ was acquired by Fugro Geosciences, Inc., in May 1996 and is now offered
as an integrated service with their CPT systems worldwide. DTI provides ROST™ technical
support to Fugro.
-------
The ROST™ uses a wavelength tunable ultraviolet laser source coupled with an optical detector
to measure fluorescence via optical fibers. The measurement is made through a sapphire window
on a probe that is pushed into the ground with a truck-mounted CPT. The optical fibers are
integrated with the geotechnical probe and umbilical of a standard truck-mounted CPT system.
CPT and standard penetrometer testing (SPT) have been widely used in the geotechnical industry
for determining soil strength and soil type from measurements of tip resistance and sleeve
friction on an instrumented probe.
The ROST™ LIF method provides data on the in situ distribution of petroleum hydrocarbons
from the fluorescence response induced hi the polycyclic aromatic hydrocarbon (PAH)
compounds that are components of the petroleum hydrocarbon. The methods detect PAHs in the
bulk soil matrix throughout the vadose, capillary fringe, and saturated zones. The methods
provide a screening of the relative petroleum hydrocarbon concentrations present. However, for
the purposes of this demonstration, only the detect/nondetect capability of ROST™ was
evaluated.
Demonstration Objectives and Approach
The primary objectives of the field demonstrations were to evaluate the ROST™ LIF technology
in the following areas: (1) its performance compared to conventional sampling and analytical
methods; (2) the logistical resources necessary to operate the technology; (3) the quality of the
LIF data; (4) the applications of the technology as determined by its performance in the CSCT
demonstrations; and (5) its performance relative to developer claims. Performance of the
ROST™ LIF sensor was evaluated to determine the agreement between LIF "detect/nondetect"
data and laboratory analyses for both total recoverable petroleum hydrocarbons by EPA Method
418.1 and total petroleum hydrocarbons by California Department of Health Services Method
8015-modified. A secondary objective for this demonstration was to evaluate the LIF technology
for cost, range of usefulness, and ease of operation.
In the approved demonstration plan, the developers presented several performance claims against
which they were evaluated. These claims included the ability to collect measurements up to 150
feet below the surface when the sensor is used with an industry-standard 20-ton CPT rig; the
ability to integrate the sensor subassembly with the rig in the field within a few hours, a standard
data collection rate of one sample every 1.2 seconds, providing a spatial resolution of less than
0.2 feet for a standard push rate of 1 meter per minute; the ability of the system to acquire
multidimensional data representations such as wavelength time matrices (WTMs) to identify fuel
or waste type (e.g., creosote); and the ability of the crew to utilize WTM information to eliminate
false positives from nonhydrocarbon fluorophores.
The demonstration was designed to evaluate the ROST™ technology as a field screening method
by comparing the LIF data to data produced by conventional sampling and analytical methods.
For both demonstrations, conventional sampling and analysis consisted of boring with a hollow
stem auger, collecting split spoon samples as closely as possible to the push cavity, and analyzing
discrete samples at an off-site commercial laboratory for petroleum hydrocarbons by EPA
Method 418.1 and California Department of Health Services Method 8015-modified. The
demonstrations were conducted at two geologically and climatologically different sites: (1) the
Hydrocarbon National Test Site located at Naval Construction Batallion Center (NCBC), Port
Hueneme, California, in May 1995, and (2) the Steam Plant Tank Farm at Sandia National
Laboratories (SNL), Albuquerque, New Mexico, in November 1995.
-------
Demonstration Results and Performance Evaluation
The ROST™ LIF technology was evaluated against the developer claims as presented in the
demonstration plan for the Port Hueneme demonstration and subsequent addendum for the SNL
arid site demonstration All developers' claims were met. The ROST™ technology was
integrated with the SCAPS cone penetrometer platform and was operated in both static and
dynamic modes. The data collection rate was measured to be 1 meter per minute with a vertical
spatial resolution of less than 2 cm. In static mode, the system acquired multidimensional data
representations to identify fuel or waste type. Wavelength-time matrices were used to eliminate
false positives from nonhydrocarbon fluorophores, specifically carbonates at the SNL. The
ROST™ system was demonstrated to be an effective, rapid in situ field screening method for
characterizing the subsurface distribution of diesel no. 2 and diesel fuel marine to depths of 55
feet in a variety of soil textures in unsaturated and saturated zones.
Cosf Evaluation
The ROST™ technology is available for use within the 48 contiguous United States for a cost of
approximately $5,300 per day or site-specific footage rates, which includes a CPT rig provided
by a commercial vendor such as Fugro Geosciences. Crew per diem and mobilization costs are
additional and site specific. The ROST™ subassembly can be integrated with any commercially
available industry-standard CPT rig. Typical crew members include a ROST™ system operator
(at a minimum), CPT operator, and assistant. Under normal conditions, an average of 300 feet of
pushes can be completed in a day. This translates to a cost of under $20 per foot. This compares
to conventional drilling costs, which range between $15 to $20 per foot for drilling and installa-
tion of monitoring wells and between $50 and $100 per foot for drilling and sampling for site
characterization. In addition, laboratory analysis costs, which range from $90 to $150 per sample
for TPH and TRPH, respectively, must be considered.
The main savings attributable to the ROST™ LIF system is that it can substantially reduce the
duration of the field investigation, quantity of costly sample collection and analyses, and
ultimately the number of soil borings and monitoring wells drilled at a site. In a general site
characterization effort, it can provide more data in less time and less expensively than
conventional drilling and sampling. Investigation-derived wastes are minimal, and worker
exposure to contaminants is reduced when using in situ technologies rather than conventional
drilling and sampling methods.
-------
Section 2
Introduction
The Site Characterization Technology Challenge
Rapid, reliable and cost effective field screening technologies are needed to assist in the complex
task of characterizing and monitoring of hazardous and chemical waste sites. However, some
environmental regulators and remediation site managers may be reluctant to use new site
characterization technologies that have not been validated in an EPA-sanctioned testing program,
since data from them may not be admissible in potential legal proceedings associated with a site
or its cleanup. Until characterization technology claims can be verified through an unbiased
evaluation, the user community will remain skeptical of innovative technologies, despite their
promise of better, less expensive and faster environmental analyses.
The Consortium for Site Characterization Technology was established as a component of the
Environmental Technology Innovation, Commercialization and Enhancement Program as
outlined in 1993 in President Clinton's Environmental Technology Initiative to specifically
address these concerns. The CSCT is a partnership between the EPA, the Department of Energy,
and the Department of Defense. As a partnership, the CSCT offers valuable expertise to support
the demonstration of new and emerging technologies. Through its organizational structure, it
provides a formal mechanism for independent third-party assessment, evaluation, and verification
of emerging site characterization technologies.
The mission of the CSCT is to identify, demonstrate, assess, and disseminate information about
innovative and alternative environmental monitoring, measurement, and characterization
technologies to developers, remediation site managers, and regulators. The Consortium is
intended to be a principal source of information and support with respect to the availability,
maturity, and performance of innovative environmental monitoring, measurement, and
characterization technologies.
Technology Demonstration Process
The CSCT provides technology developers a clearly defined performance assessment, evaluation
and verification pathway. The pathway is outlined in the following four components:
• technology selection;
• technology demonstration;
• technology performance assessment, evaluation, and verification, and
• information distribution.
These are discussed in more detail in the following paragraphs.
Technology Selection
The first step in the overall demonstration process is one of technology selection. The selection
process comprises two components. Beyond the initial identification of potential technologies, a
critical aspect of technology selection is an assessment of its field deployment readiness. Only
pre-production and production instrumentation with a history of successful laboratory or field
-------
operation are accepted into the program. Early, unproven prototype instrumentation systems
requiring extensive testing and modification prior to field deployment are not acceptable
demonstration candidates. The candidate technology must meet minimum technology maturity
criteria in order to participate in a demonstration. The degree of technology maturity may be
described by one to three levels:
Level 1
Technology has been demonstrated in a laboratory environment and ready for
initial field trials.
Level 2
Technology has been demonstrated hi a laboratory environment and in field trials.
Level 3
Technology has been demonstrated extensively both in the laboratory and in field trials
and is commercially available.
A second aspect of the technology selection process involves a determination of technology/field
requirements match. Because of limited resources, the Consortium must determine a
technology's suitability for demonstration in light of the current needs of the environmental
characterization and monitoring community. A technology may be given priority for demon-
stration and evaluation based on its environmental and fiscal impact and the likelihood that its
demonstration will fill information gaps which currently impede cost effective and efficient
environmental problem solving. The CSCT conducts surveys of EPA, DOE, DoD, state, local,
tribal and industry agencies to assist in determining the degree of match between the candidate
technology and the needs of the environmental restoration community.
Technology Demonstration
A technology demonstration plan is developed by the technology verification entity, according to
document preparation guidance provided by the Consortium for Site Characterization
Technology. The demonstration plan includes a technology description, the experimental design,
sampling and analysis plan, methods for evaluating the technology, a quality assurance project
plan, and a health and safety plan. After approval by the EPA and technology developers, the
demonstration plan is implemented at an appropriate field location. The CSCT provides
technical support to the technology developer during demonstration plan preparation and
execution and also audits the demonstration and data collection processes.
Technology Performance Assessment, Evaluation, and Verification
In this important component of the demonstration process, an objective comparison of demon-
stration technology data is carried out against a reference data set generated using conventional
analysis methodologies. The principal product of this phase of the project is the ITVR, prepared
by an independent third party. The report documents the demonstration technology data along
with an assessment of the technology's performance in light of the reference data. The degree of
data analysis hi the technology report is determined by the level of maturity of the technology
under evaluation, with the more mature technologies receiving more thorough analysis. The
CSCT provides Level 1 technologies with a fielding opportunity in which the system can be
tested. Evaluation of the system performance and comparison of field data with reference
laboratory data are the developer's responsibility. In the case of Level 2 technologies, the
performance evaluation is performed by the CSCT. The most extensive evaluation is done for
-------
the Level 3 technologies since these are considered market-ready. As part of the demonstration
objectives, the CSCT evaluates the developer claims regarding the capabilities of the Level 3
technology and prepares a technology evaluation report containing an assessment of the
technology's performance.
Information Distribution
Evaluation reports for Level 2 technologies are distributed to the technology developers, CSCT
partners, and the general public. In addition, for Level 3 technologies performance verification
statements are distributed to the developers for subsequent use in seeking additional develop-
mental funding or marketing.
Technology reports for Level 1 technologies are distributed as EPA project reports. There is no
technology evaluation contained hi these documents. Results are compiled and reference data is
provided so that the developer and reader can formulate an opinion regarding technology
performance.
The CPT-LIF Sensor Demonstrations
The developer of the ROST™ LIF technology was Loral Corporation and Dakota Technologies,
Inc. (DTI). PRC Environmental Management, Inc. (PRC), a contractor to the NCCOSC RDT&E
Division (the developer of the SCAPS LIF technology), prepared the demonstration plan for both
developers and conducted the predemonstration and demonstration field efforts, coordinated the
analyses of the soil samples, and provided the raw data to Sandia National Laboratories-New
Mexico (SNL), a DOE-owned laboratory operated by Lockheed Martin Corporation. SNL, as the
EPA's verification entity, reviewed and approved the demonstration plan and amendments and
reduced and analyzed the data generated during the two field demonstrations.
The ROST™ LIF is a CSCT Level 3 technology. For these demonstrations, the CSCT and the
developers selected the demonstration sites, participated in the demonstration planning process,
and jointly and separately evaluated the data generated during both demonstrations.
This report describes how the demonstration participants collected and analyzed samples,
provides the results of the demonstration, and describes how the performance of the ROST™ LIF
technology was verified. Section 5 discusses the experimental design for the demonstration.
Section 4 presents the reference laboratory results and evaluation. Section 3 describes the
ROST™ LIF technology. Section 6 presents the ROST™ LIF demonstration results and
evaluation. Section 7 is an assessment of recommended applications of the technology. Section
8 is a forum wherein the developer has the opportunity to discuss the technology results and
comment on the evaluation and future technology developments. Section 9 is a table
summarizing selected ROST™ commercial projects. In addition, there are appendices containing
the reference laboratory data, ROST™ LIF data, and proposed ROST™ LIF method.
-------
Section 3
ROST™ LIF Technology Description
The description of the ROST™ LIF technology and verification of its performance has been
divided into two sections, Section 3 and Section 6. Because this is an innovative technology,
evaluating its performance and comparing it to conventional laboratory methods with well-
established procedures is not as simple as the evaluation of the laboratory methods as presented
in the preceding section. Section 3 describes the ROST™ LIF sensor technology and includes
background information and a description of the equipment. General operating procedures,
training and maintenance requirements, and some preliminary information regarding the costs
associated with the technologies are also discussed. Much of this information was provided by
the technology developier and presented in the demonstration plan (Loral, 1995). Any claims
made in this section may or may not have been verified during this demonstration. Specifically,
the subsections regarding technology applications and limitations and advantages of the
technology were provided by the developer and may not have been verified. The verification of
technology performance at the two demonstration sites and evaluation of developer claims for
this program are presented in detail in Section 6.
ROST™LIF Sensor
Petroleum-based fuels, such as gasoline, diesel, and kerosene, and other hydrocarbons, such as
coal tar and creosote, contain compounds that fluoresce when excited by ultraviolet light. A soil
sample contaminated with petroleum substances will exhibit fluorescence intensity that is
proportional to the contaminant concentration. The concentration of the hydrocarbon fraction in
an unknown sample can be determined by comparing its fluorescence intensity to that of
calibration standards.
ROST™ detects the presence and quantitates the amount of aromatic petroleum hydrocarbons by
the laser-induced fluorescence in the sample. The Rapid Optical Screening Tool is a tunable dye
laser-induced fluorescence system designed as a field screening tool for detecting petroleum
hydrocarbons in the subsurface. The ROST™ LIF system uses a pulsed laser coupled with an
optical detector to make fluorescence measurements via optical fibers. The measurement is
made through a sapphire window on a probe that is pushed into the ground with a truck-mounted
cone penetrometer.
The ROST™ approach permits temporary or permanent installation of the LIF equipment on a
CPT truck or other direct push vehicle, although a dedicated ROST™ unit could be permanently
installed in a CPT. The CPT LIF system uses a steel probe containing the LIF sapphire optical
window as well as the cone and sleeve strain gauges. The excitation and emission optical fibers
are isolated from the soil system by a 6.35-mm diameter sapphire window located 60 cm from
the probe tip and mounted flush with the outside of the probe. The ROST™ LIF system uses
600-fJ.m diameter fibers that are up to 100 m in length.
-------
Laser Source
The ROST™ LIF primary laser uses a neodymium-doped yttrium aluminum garnet pump
(Nd:YAG) laser. It produces 532-nm light at 50 Hertz (Hz) with a pulse energy of 50 ml. The
light from the primary laser pumps a rhodamine 6G dye laser whose output is then frequency-
doubled to produce ultraviolet (UV) light. The laser system used in the ROST™ is capable of
generating wavelengths of light ranging from about 280 nm to about 300 nm, depending on the
dye being used. The wavelength of light produced by the ROST™ LIF laser is tunable within this
range. The ROST™ laser system is coupled to a silica clad silica ultraviolet/visible light
transmitting optical fiber. This fiber and the collection fiber are integrated with the geotechnical
probe and umbilical of a standard truck-mounted CPT system.
System Components
The Rapid Optical Screening Tool consists of the spectrometer rack and the control rack (Figure-
3-1). The spectrometer rack holds all the spectroscopic instrumentation, including the Nd:YAG
pump laser, tunable dye laser, emission monochromator, photomultiplier tube, and associated
power supplies and motion controllers. The control rack contains the control computer and a
digital oscilloscope signal processor. In operation the racks can be positioned independently and
separated from each other by up to 25 feet. The racks themselves are standard industrial models
with a 20-inch by 25-inch footprint and stand 25 inches high. The Nd:YAG pump laser and dye
laser are arranged on an optical breadboard affixed to the top of the spectrometer rack. When the
opaque plastic dye laser cover is in place, the total height of the spectrometer rack is 34 inches.
The computer monitor can be conveniently placed on top of either the control rack or the dye
laser cover.
Spectrometer Rack Components
The spectrometer rack holds modules for generation of pulsed ultraviolet light and detection of
the return fluorescence signal. The fiberoptic cables leading to and from the cone penetrometer
probe are interfaced at the back of the spectrometer rack through ST connectors. The generation
of exitation light in the ROST is based on a two-stage dye laser pumped by the 532 nm harmonic
of a compact pulsed Nd:YAG laser. The Nd:YAG laser head, Rhodamine 6G dye laser, and all
related optics are arranged on a 19-in. by 23-in. aluminum breadboard, which is affixed to the top
of the spectrometer rack. Light in the 280-300 nm wavelength range is generated via frequency
doubling of the dye laser output. A 266-nm exictation wavelength capability is employed for
direct detection of benzene, toluene, ethylbenzene, and xylenes (BTEX) and other single-ring
aromatic hydrocarbons. In this case, the 532-nm Nd:YAG beam is diverted around the dye laser
to the frequency doubling crystal for fourth harmonic generation (266 nm) To change between
the tunable and 266-nm configurations requires insertion (or removal) of two mirrors on
kinematic mounts.
All other mechanical operations are controlled through software. During normal operation, the
only time the cover need be removed is to change between the 266-nm and 280-3 00-nm configu-
rations. The frequency doubling crystal has been incorporated into a housing whose temperature
is held at 40° C for isolation from any temperature drift in the truck. Pyroelectric power meters
are built in for monitoring the 532-nm output of the Nd:YAG laser and the ultraviolet light
emerging from the doubling crystal. If the ultraviolet output relative to the 532-nm pump input
falls below specifications, an automated routine is initiated by the operator to re-optimize the
frequency doubling crystal position.
-------
(p)
Figure 3-1. ROST™ system components.
A. Control Rack
B. Spectrometer rack
C. Nd:YAG power supply
D. Optical breadboard
E. Frequency doubling crystal
F. Nd:YAG laser head
G. Dye laser
H. Monochromator
I. Stepper motor controller/driver
J. Photomultiplier tube and housing
K. High voltage power supply
L. Stepper motor reset button
M. Water reservoir
N. Dye reservoir
O. Monitor
P. Keyboard
Q. Utility Drawer
R. Computer
S. Oscilloscope
T. Temperature Controller
The ROST™ fluorescence detector consists of a monochromator and photomultiplier tube
(PMT). The monochromator selects a narrow wavelength interval of the pulsed, polychromatic
fluorescence light that is returned to it from the cone via the collection fiber. The normal
wavelength range of the monochromator setting is 300 to 500 nm. The fluorescence light pulse
lasts for tens to hundreds of nanoseconds. The photons emerging from the monochromator to a
pulsed electrical signal hi the PMT. The photoelectron stream is amplified during passage down
the photomultiplier dynode chain to the anode. The time profile of the electron current that is
collected at the dynode is slightly distorted in time by the time response characteristics of the
PMT.
Control Rack Components
The principal components of the control rack are the control computer, the digital oscilloscope
signal processor, and the slide-out computer keyboard. Signals passing from the computer to the
spectrometer rack are used to set the monochromator slit width, the wavelength passed by the
monochromator, the wavelength of the dye laser, and the position of the frequency doubling
crystal for the chosen dye wavelength. Information passing from the spectrometer rack to the
control computer includes the signal from the PMT, diagnostic information from the Nd:YAG
laser, and the outputs from the power meters.
-------
The pulsed electrical signal from the PMT is fed to a digital storage oscilloscope, which
digitizes, averages, and displays the fluorescence intensity versus time waveform. The user may
select the number of waveforms to be averaged in the digital storage oscilloscope. After com-
pletion of the specified number of acquisitions, the waveform is downloaded to the computer for
permanent storage and post-processing of the data. The digital storage oscilloscope and
computer communicate via a GPD3 bus.
Dynamic Range
The linear dynamic range of the ROST™ LIF detector depends on the specific hydrocarbon
analyte as well as the particular matrix. Generally, for in situ measurements, it has been found
that the linear portion of the response curves extends well beyond three orders of magnitude.
Nonlinearity tends to occur at concentrations greater than 10,000 mg/kg. In sandy soils, the
nonlinearity occurs at lower concentrations than in clay rich soils, possibly due to self absorption
or saturation. The linear dynamic range of the LIF sensor also depends on operator-controlled
instrumental parameters. The linear dynamic range may be extended to higher concentrations by
adjusting the slit width of the detector, but this results in decreased sensitivity at lower
concentrations.
Sensitivity, Noise, and Background
Three quantities are needed to determine the fluorescence LOD and concentration LOD limit:
noise, background, and sensitivity. Sensitivity is determined using the calibration samples
prepared, in most cases, immediately prior to the site visit using soil from the site and standard
analytical techniques. The noise is computed after the pushes have been performed and is
generally computed on a push by push basis.
The fluorescence intensity for each calibration sample is measured in triplicate each day prior to
the start of operations. The three measurements are averaged to provide a single measured
intensity for each concentration. The fluorescence data are regressed using the known
concentration values to establish a slope and intercept. The intercept is an estimate of the
intensity of the unspiked calibration standard (0 mg/kg). The slope is an estimate of the
"sensitivity" of the fluorescence measurement to changes in hydrocarbon contamination.:
intercept: b = intensity measured on 0 mg/kg calibration sample
slope :
=
where the sums are taken over the range of calibration samples. For these calibration soils, x is
given by the concentration of the target fuel, while y is the measured fluorescence intensity
adjusted to be a percentage of the M-l standard.
Following each push, a histogram is provided for the LIF responses showing a percentage of the
M-l standard. A subjective decision is made based on the belief that background counts (again
expressed as a percentage of M-l standard) should be somewhat normally distributed. This
decision results in an estimate of background noise and an estimate of the mean background
fluorescence level, expressed as a percentage of M-l standard and the background noise, the
10
-------
standard deviation of the fitted normal distribution. For pushes in uncontaminated areas, the
noise is directly reflected in the width of the histogram. It is possible to get a histogram that is
bi-modal (or multi-modal), complicating the noise evaluation. A single mode will not be
observed if different levels of background fluorescence are present, e.g., from two different types
of minerals, or if samples reflect information from uncontaminated regions and regions where
hydrocarbons are present.
Calculated Fluorescence and Concentration Thresholds
The ROST™ LIF fluorescence threshold can be qualitatively interpreted as the minimum signal
amplitude that is reliably associated with petroleum contamination. The fluorescence threshold
is affected by any fluorescence background arising from the fiber, window, or soil matrix. The
basis of the ROST™ fluorescence threshold determination is that the background signals should
be normally distributed. The center of the normal (Gaussian) distribution gives the background
value, and the standard deviation can be used to establish confidence intervals. For this
demonstration a 99% confidence interval was used, such that:
fluorescence threshold = mean background + 2.58 x standard deviation of the
background.
The concentration threshold is determined directly from the fluorescence threshold using the
estimated sensitivity provided by the calibration results. These results are based on the equation:
concentration threshold = 2.58 x standard deviation of the background/sensitivity
For the Port Hueneme demonstration, the ROST™ data was integrated over a 6-inch interval.
For the SNL demonstration, the data was averaged over a 3-inch interval. Any signal exceeding
the fluorescence threshold was considered a "detect."
Mobilization and Installation of ROST™ for CPT LIF work
ROST™ is transported to job sites, installed in the CPT truck for the duration of the job, and then
demobilized for transport to its next site. Several different transportation modes have already
been tested and proved satisfactory. For transportation by commercial entities (motor freight or
airlines), the spectrometer rack and control rack are placed in wooden shipping containers.
During the shipment, the racks rest in the boxes on the shock mountings that are affixed to the
floor of the CPT truck.
Deployment Costs
The daily rate for ROST™ is approximately $5300 per day, which includes the daily rate for the
CPT rig. Footage rates may be proposed on a job-specific basis. Per diem costs vary with each
deployment. Electronic data files are available to the client for $12 per push and color integrated
CPT/ROST logs are available for $12.50 per push. Mobilization fees are quoted on a job
specific basis. Additional crew members are available on an hourly basis.
11
-------
Technology Applications
Fugro Geosciences' ROST™ LIF system was developed in response to the need for real-time in
situ measurements of subsurface contamination at hazardous waste sites. The ROST™ LIF
system performs rapid field screening to determine either the presence or absence or relative
concentration of petroleum hydrocarbon contaminants within the subsurface of the site. The site
can be further characterized with limited numbers of carefully placed borings or wells. In
addition, remediation efforts can be directed on an expedited basis as a result of the immediate
availability of the LIF and soil matrix data.
Advantages of the Technology
The LIF sensing technology is an in situ field screening technique for characterizing the
subsurface distribution of petroleum hydrocarbon contamination before installing groundwater
monitoring wells or soil borings. The method is not intended to be a complete replacement for
traditional soil borings and monitoring wells, but is a means of more accurately placing a reduced
number of borings and monitoring wells in order to achieve site characterization.
The LIF technology using a CPT platform provides real-time field screening of the physical
characteristics of soil and chemical characteristics of petroleum hydrocarbons at hazardous waste
sites. The current configuration is designed to quickly and cost-effectively distinguish petroleum
hydrocarbon-contaminated areas from uncontaminated areas. This capability allows further
investigation and remediation decisions to be made more efficiently and reduces the number of
samples that must be submitted to laboratories for costly analysis. By achieving site characteri-
zation while expending a minimum amount of resources, remaining resources can be directed at
studying the actual risks posed by the hazardous waste site and for remediation if warranted.
Table 3-1 compares the important attributes of the ROST™ technology with those of traditional
laboratory methods. The major advantage of ROST™ is that it provides real-time data in the
field without the need for sample manipulation and the accompanying risk of sample degrada-
tion. ROST™ also provides a qualitative fingerprinting capability in a fraction of the time
required by gas chromatography. Under normal conditions, an average of 300 feet of pushes can
be reasonably advanced hi one day.
Table 3-1. Total petroleum hydrocarbon analysis methods.
Criterion
Basis of method
Applicability
Possibte interferences
Sample preparation
Analysis time
Fingerprint capability
Real-time///? silu
DHS Method 8015
Gas chromatography
Volatile organic compounds
Any volatile compound
Extract, filter
5-30 minutes
Yes
No
EPA Method 41 8.1
IR absorbance in C-H
Compounds with C-H bonds
Any species with C-H bonds
Extract, filter
Seconds
No
No
ROST™
Fluorescence
Aromatic hydrocarbons (single,
double and multi-rino)
Fluorescent minerals
None
Seconds
Yes
Yes
12
-------
Detectors and Data Acquisition Modes
ROST™ uses the digital oscilloscope to capture time-domain information about the pulsed
fluorescence signal resulting from the pulsed laser excitation. The areas of the time-integrated
waveforms are proportional to the total photon flux passed to the detector. The time dependence
of the fluorescence contains significant additional information, particularly about oxygen
quenching. Oxygen fluorescence quenching leads to a decrease in the emitted intensity at all
wavelengths. The fluorescence response as a function of fuel concentration (sensitivity) is
affected by variable oxygen levels in the soil matrix. Variability in the oxygen levels
encountered during a push can cause small changes to the FVD profile. Measurements
performed with continuous excitation sources, or with pulsed sources but not time resolved
detection, are unavoidably affected by the phenomenon. ROST™ operators and data interpreters
can make valid assessments of the extent of contamination even when oxygen content varies.
Limits of the Technology
This section discusses the limits of the ROST™ LIF as it is currently understood.
Response to Different Petroleum Hydrocarbons
The relative response of the ROST™ LIF sensor depends on the specific analyte being measured.
The instrument's sensitivity to different hydrocarbon compounds such as gasoline, diesel fueland
jet fuel are comparable. The sensitivity is not as great for coal tar and creosote although they are
readily detectable. These variations in sensitivity are primarily a reflection of the variations in
the PAH distribution. Other contributing factors such as optical density, self absorption, and
quenching are less important. The total observable fluorescence produced by any given
petroleum hydrocarbon sample depends on the mole fraction of fluorescing PAHs along with the
relative quantum efficiency of each of the fluorescing species. The fluorescence properties of a
hydrocarbon mixture in soil may also change after long-term exposure to and interaction with the
environment. A contaminant that has been in the ground for any period of tune will undergo
changes in chemical composition due to weathering, biodegradation, and volatilization.
The ROST™ LIF system often uses 290 nm as the excitation wavelength. This wavelength is
short enough to excite the fluorescence of all aromatic hydrocarbons with at least two conjugated
aromatic rings. Aliphatic species and single-ring aromatics do not contribute to the ROST™ LIF
signal from 290 nm. The ROST™ can also be configured for 266 nm excitation. The attenua-
tion of light passing through the optical fibers, however, is greater at 266 nm than at 290 nm or
longer wavelengths, so the possible length of the push may be restricted.
Matrix Effects
The in situ fluorescence response of the LIF sensor to hydrocarbon compounds is sensitive to
variations in the soil matrix. Matrix properties that affect LIF sensitivity include soil grain size,
mineralogy, moisture content, and surface area. Each of these factors influences the relative
amount of analyte that is adsorbed on or absorbed into the soil. Only the relative fraction of
analyte that is optically accessible at the window of the probe can contribute to the fluorescence
signal. Of the four influencing factors mentioned above, the dominant variable appears to be soil
surface area. LIF sensitivity to petroleum hydrocarbons on soil has been shown to be inversely
13
-------
proportional to the available surface area of the soil substrate. Sandy soils tend to have a much
lower total available surface area than clay soils. Hydrocarbon compounds in sandy soils
generally yield a correspondingly higher fluorescence response than they do in clay rich soils. In
one study, soil samples were prepared as a series of sand/clay (illite) mixtures with progressively
increasing clay content. The relative LIF response to DFM in each soil is essentially identical
once the response curves were normalized to the available surface area of each of the soils. The
moisture content of the soil matrix is another influencing factor. The LIF sensitivity to
petroleum hydrocarbons generally increases with greater soil moisture content, although in some
natural soils the effect appears to be small. LIF response curves representing the results of
fluorescence measurements on a soil with varying water content have also been generated. These
results suggest that the response is fairly insensitive to changes in moisture content. In another
study it was demonstrated that increasing the amount of water in a soil tends to narrow the
sensitivity difference between sandy and clay soils. It is thought that water physically displaces
the hydrocarbons from within the pore spaces of the matrix, effectively reducing the surface area
available to contaminants. The effects of soil grain size have also been examined in laboratory
studies. LIF sensitivity generally increases with increased grain size. The measured fluorescence
was shown to be substantially greater in the coarser mesh sizes.
Spectral Interferences
The ROST™ LIF sensor is sensitive to any material that fluoresces when excited by ultraviolet
wavelengths. Although intended to specifically target petroleum hydrocarbons, the excitation
energy produced by the LIF system's laser may cause other naturally occurring substances to
fluoresce as well. At some investigation sites, it is possible that LIF sensors could respond to
fluorescence originating from nonhydrocarbon sources. Many common fluorescent minerals can
produce a measurable LIF signal. Other non-hydrocarbon fluorescent material introduced
through human activity may be found in the subsurface environment. Deicing agents, antifreeze
additives, and many detergent products are all known to fluoresce very strongly. The potential
presence of fluorescence emission from nontarget (non-hydrocarbon) analytes within the soil
matrix must be considered when assessing LIF field screening data. In some instances, the
inability to discriminate between hydrocarbon fluorescence and nonhydrocarbon fluorescence
can lead to false positives for the presence of hydrocarbons. Nonhydrocarbon fluorescence can
mask the presence of hydrocarbon fluorescence, leading to reduced sensitivity or erroneous
estimation of the relative amount of hydrocarbon present. In the worst case, spectral interference
can lead to a false positive or false negative report of findings. However, because the LIF sensor
collects full spectral information, it is almost always possible to discriminate between
hydrocarbon and nonhydrocarbon fluorescence by analyzing the spectral features associated with
the data.
Truck-Mounted Cone Penetrometer Access Limits
The CPT support platform used to deploy the ROST™ LIF is typically a 20-ton all-wheel drive
diesel powered truck. The dimensions of the truck require a minimum access width of 10 feet
and a height clearance of 15 feet. Some sites, or certain areas of sites, might not be accessible to
a vehicle of this size. The access limits for a typical CPT truck are similar to those for
conventional drill rigs and heavy excavation equipment.
14
-------
Cone Penetrometer Advancement Limits
The CPT sensors and sampling tools may be difficult to advance in subsurface lithologies
containing cemented sands and clays, buried debris, gravel units, cobbles, boulders, and shallow
bedrock. As with all intrusive site characterization methods, it is extremely important that all
underground utilities and structures be located using reliable geophysical equipment operated by
trained professionals before undertaking activities at a site. Local utility companies should be
contacted for the appropriate information and approval.
15
-------
Section 4
Reference Laboratory Results and Evaluation
The purpose of this section is to address issues related to the reference laboratory used for these
demonstrations. Section 4 is divided into four subsections. The first subsection provides details
concerning the selection of ATI as the reference laboratory and the reference methods performed
on the soil samples at ATI for the purpose of comparison with results from the LIF technology.
The second subsection provides an assessment of data quality for the laboratory and gives a
description of the quality control procedures for TRPH (total recoverable petroleum hydro-
carbons by IR spectrophotometry) by EPA Method 418.1 and California DHS Method 8015-
modified for TPH (total petroleum hydrocarbons by GC-FID). These methods will be referred to
as TRPH and TPH throughout the remainder of this report. In the third subsection, the methods
used to estimate accuracy, precision, and completeness are discussed and results provided. The
final subsection provides a summary of the laboratory data quality evaluation and a brief
discussion of how the laboratory results will be used for comparison with the results of the LIF
technology.
Selection of Reference Laboratory and Methods
To assess the performance of the LIF technology as a field screening tool for petroleum
hydrocarbons in the subsurface, the data generated using the LIF technology was compared to
data obtained using conventional sample collection and analytical methods. The analytical
laboratory selected to provide reference analytical services, ATI, is certified in the state of
California. The laboratory is located in San Diego, California.
ATI was selected because of its experience with QA procedures, analytical result reporting
requirements, data quality parameters, and previous involvement with the Navy SCAPS program.
ATI is not affiliated with the U.S. Navy, Loral Corporation, DTI, or any of the demonstration
team members. ATI provided copies of the analytical results directly to SNL in order to maintain
independence of the data. Copies of all QA and analytical procedures were provided to SNL for
review prior to the demonstration and were included hi the approved demonstration plan.
After discussion between representatives of State of California EPA, SNL, and the U.S. EPA,
EPA Method 418.1 for TRPH and California DHS Method 8015-Modified for TPH were
selected as the reference methods for the LIF technologies. The TRPH and TPH methods were
chosen because of their widespread and generally accepted use in delineating the extent of
petroleum hydrocarbon contamination. The TRPH and TPH methods are currently used as
indicators of petroleum contamination hi leaking underground and aboveground fuel tank
investigations; as such they are the most comparable analytical methods corresponding to the
objective of demonstrating rapid field screening using LIF.
EPA Method 418.1 for total recoverable petroleum hydrocarbons (TRPH) is used for the
' measurement of Freon-113-extractable petroleum hydrocarbons from surface and saline waters,
soil, and industrial and domestic wastes. The sample is acidified to a low pH (<2) and serially
extracted with Freon-113 in a separatory funnel. Interferences from polar animal oils and greases
16
-------
are removed with silica gel adsorbent. Infrared analysis of the extract is performed, and its
absorption is directly compared to that measured on a standard mixture of hydrocarbons.
California Department of Health Services (DHS) Method 8015-modified for total petroleum
hydrocarbons (TPH) is based on EPA SW-846 Method 8015 for determination of ketones,
modified for determination of petroleum hydrocarbons in soil (EPA, 1995). It is used for the
determination of gasoline and diesel in contaminated groundwater, sludges, and soil. After
solvent extraction, a sample is injected into a gas chromatograph where compounds are
separated. Compounds in the GC effluent are identified and quantified using a flame ionization
detector. The chromatogram produced by this analysis covers the carbon range from C7 to C36
and can help to identify the product type using the n-alkane pattern distribution, pristane: phytane
ratios, and the width of the unresolved complex mixture.
Assessment of Laboratory Data Quality
Audits
As part of the cooperative agreement between the U.S. EPA and the State of California EPA
Department of Toxic Substances Control, a representative of the California EPA audited the ATI
laboratory in April 1995 and provided audit results to SNL. The audit found no irregularities and
verified the procedures used to homogenize and analyze the discrete soil samples. SNL reviewed
the ATI Quality Assurance Manual and all related procedures prior to the demonstrations (ATI,
1995).
Sample Holding Times
The holding time specification for EPA Method 418.1 is 28 days from the sampling date. The
holding time specification for California DHS Method 8015-modified is extraction within 14
days of sampling date. The required holding times per ATI SOP 105 from the date of sample
receipt to the date of extraction and analysis were met for the samples from both sites. However,
for the SNL samples, two samples (SNLDB11-5 and SNLDB11-10) were misplaced prior to
homogenization and were left unrefrigerated in a sealed container for five days before being
located. They were homogenized, extracted and analyzed per both methods within 14 days of the
sampling date (CEDVflC, 1996). The results are shown in Table A-2. These samples had large
concentrations (> 10,000 mg/kg) of hydrocarbons that exceeded the LIF detection limit. For this
verification study, the total concentration of the petroleum hydrocarbons in the sample was
unimportant for the comparison; the fact that bom samples showed contamination well above the
LIF detection limit (qualifying the samples as "detect") was important for the purpose of
comparison to the LIF method. For this reason, they were not excluded from the data set.
Sample Preparation
All soils were homogenized for five minutes prior to extraction and analysis per ATI SOP 421.
Preparation of soils for TPH analysis was performed per ATI SOP 400 by diluting in methylene
chloride. Preparation of soils for TRPH analysis was performed by extraction with Freon-113
for 45 minutes prior to analysis per ATI SOP 803.
17
-------
Sample Analysis
TRPH was determined by EPA Method 418.1 by calculating the linear regression of absorbance
versus concentration. The concentration thus derived tells only the concentration of oils in the
Freon-113 extract. This was then related back to the original sample. TPH was quantified by
DHS Method 8015-modified by sample peak area using the mean response factor of the curve.
The concentration was calculated using the response factor and the mean calibration factor
obtained from prepared diesel fuel standards and adjusting for volume and dilution factors. FID
was used for compound detection.
Detection Limits
The ATI method detection limit for TRPH is 1.0 mg/kg for soil. The method detection limit for
TPH is 5.0 mg/kg for soil.
Quality Control Procedures
For TPH, quality control procedures included preparation of a calibration curve for instrument
calibration using NIST-traceable standards. A reagent blank is extracted each time a batch of no
more than 20 samples is extracted. An additional reagent blank is extracted for each batch of 20
samples in any given day. A blank spike is extracted with each batch of no more than 20
samples. Surrogates are run with each soil sample and quality control sample. Matrix spikes and
matrix, spike duplicates are also prepared and associated to no more than 20 samples of a similar
matrix to check for precision and accuracy. Spiking is done directly into the sample prior to
extraction. Spiking levels for fuel hydrocarbons are 100 mg/kg for soils.
For TRPH, a reagent blank, blank spike, matrix spike, and matrix spike duplicate were analyzed
for each batch of 10 samples. Spiking level for petroleum hydrocarbons is 130 mg/kg for soils.
A laboratory control sample was analyzed to verify the working curve, and a midrange check
standard was run every tenth scan. The working calibration curve was prepared once per day.
Calibration standards were run at least every 10 samples to verify the calibration curve. In
addition, a laboratory control sample (a midrange reference standard) was run at least once
during each instrument run to verify the calibration curves. ATI did not report the actual results
to the developers or SNL, but did report that all calibration and control standards were within
acceptance limits.
Accuracy, Precision, and Completeness
This section discusses the accuracy, precision, and completeness of the reference method data.
Tables 4-1 and 4-2 display the results of the quality control samples used to estimate accuracy
and precision of the methods. The data from the reference laboratory were internally reviewed by
ATI QC personnel before the data were delivered to SNL and NCCOSC RDT&E Division. SNL
reviewed the raw data and quality control sample results and verified all calculations.
Accuracy
Accuracy and matrix bias of the reference methods were assessed using laboratory spiked
samples and, in the case of DHS Method 8015-modified, surrogate additions. Results of past PE
audits of ATI were also reviewed to verify laboratory performance for accuracy and precision.
18
-------
Table 4-1. Quality control results for TPH1.
QC Sample
Matrix Spike % Recovery
MS Duplicate % Recovery
MS Duplicate RPD
Surrogate Spikes
Blank Spike
Reagent Blanks
ATI Acceptance Limits
63-119% Recovery
63-1 19% Recovery
18%
69-132% Recovery
61-125% Recovery
<5.0 mg/kg
Port Hueneme Demo
Average Result
88 (range 80-100)
86 (range 77-1 00)
3 (range 0-8)
104 (range 97-126)
96 (range 90-100)
all < 5.0 mg/kg
SNL Demo Average
Result
lOO(onesampJe)
110 (one sample)
4 (range 0-10)
110 (range 100-126)
108 (range 100-110)
all < 5.0 mg/kg
(Total Petroleum Hydrocarbons by GC/FID, California DHS Method 8015-modified).
Table 4-2. Quality control results for TRPH .
QC Parameter
Matrix Spike % Recovery
MS Duplicates RPD
Blank Spike
Reagent Blank
ATI Acceptance Limits
74-126% Recovery
20%
88-1 18% Recovery
< 1.0 mg/kg
Port Hueneme Demo
Average Result
104 (range 79-118)
3 (range 0-20)
102 (range 90-1 18)
all < 1.0 mg/kg
SNL Demo Average
Result
104 (range 98-1 06)
4 (range 0-1 3)
104 (rang 100-1 10)
all < 1.0 mg/kg
(Petroleum Hydrocarbons by 1R Spectrophotometry, EPA Method 418.1).
To estimate accuracy, the percent recovery is calculated using the following equation:
.. „ Spiked sample result - Unspiked sample result
% Recovery = — x 100%
Spike concentration
Diesel fuel standard was the spiking compound for the TPH method, and the surrogate is bis-2-
ethylhexylphthalate. Surrogate recoveries were all well within laboratory acceptance limits (69-
132% recovery). Blanks were prepared using sterilized silica sand as the "soil." The spiking
compound for TRPH was a prepared mixture of fuel hydrocarbons containing hexadecane,
isooctane, and benzene. Blanks for both methods were prepared using sterilized silica sand as
the "soil."
The percent recoveries for the laboratory measurements of matrix spikes, blank spikes, and
duplicate spikes for both methods are presented in Tables 4-1 and 4-2.
Cal EPA-DTSC also obtained splits of samples to independently verify ATI's results at the State
of California Hazardous Materials Laboratory. There was excellent agreement between both
laboratories for TPH and TRPH.
Precision
Precision of the reference method results can be estimated using the field duplicates by
comparing the relative percent differences (RPD) for sample results and their respective field
duplicates, or results of a laboratory spiked sample prepared and analyzed in duplicate, using the
following equation:
I Sample result — Duplicate result I
RPD = - : ! x 100%
Average result
19
-------
Field duplicate samples were analyzed by both reference methods. After the soil samples were
homogenized, nine of the samples from the Port Hueneme site and one of the samples
(SNLDB11-40) from the SNL site were analyzed hi duplicate (see Table A-l). This subset was
selected randomly by the SNL verification entity in the field during the Port Hueneme demon-
stration, based on a visual assessment of the contamination of the sample; only the samples
containing visually detectable hydrocarbon contamination were analyzed in duplicate. The
sample for the SNL demonstration was selected after the demonstration based on inspection of
the LIF results. The mean precision estimate (RPD) for the 10 total field duplicates was 10.7%
for TPH and 16.4% for TRPH. Overall, these data show good agreement between the samples
and their respective field duplicates, indicating a high degree of precision by the reference
laboratory.
The precision for the laboratory duplicates (Table 4-1,4-2) was estimated by comparing the
results of 14 pairs of matrix spike/matrix spike duplicates for TPH and 23 pairs of matrix
spike/matrix spike duplicates for TRPH. Overall, those data shows good agreement between the
laboratory matrix spikes and their duplicates for both methods.
Completeness
Percent completeness is defined as follows for all measurements:
%C = 100% x -
where
number of sample measurements judged to be valid
total number of discrete sample measurements
Results were obtained for all of the soil samples. A total of 130 analytical soil sample results
plus nine field duplicate results using both TPH and TRPH methods were available from Port
Hueneme. A total of 92 soil sample results for both TPH and TRPH plus one field duplicate
sample result were available from the SNL Tank Farm demonstration data set. As mentioned
earlier, two samples from SNL that were left unrefrigerated for 5 days at the laboratory were
included in the data set because their suitability for comparison to the LIF measurements did not
appear to be compromised. Based on these results, the completeness of the data set was 100
percent
Use of Qualified Data for Statistical Analysis
As noted above, 100 percent of the reference laboratory results from Port Hueneme and SNL
samples were reported and validated. The data review indicated that all data were acceptable for
meeting the demonstration objectives. The results of these analyses are presented in tabular form
in Appendix A, Tables A-l and A-2, and graphically in Section 6.
20
-------
Section 5
Demonstration Design and Description
Evaluation ofROST™LIF Sensor Performance
The performance of the ROST™ LIF sensor was evaluated to determine the percentage
agreement between LIF "detect/nondetect" data and both TPH and TRPH results. Conventional
sampling and analysis consisted of boring adjacent to the push holes with a hollow stem auger,
collecting split spoon samples as close as possible to the push cavity, and analyzing the discrete
samples at the reference laboratory. The data from the laboratory analysis of soil samples which
showed TRPH or TPH contamination above the LIF detection limit were considered to show a
"detect." Similarly, if in situ LIF readings registered above the LIF site detection limit, they
would also indicate a "detect." The number of matches (detect/detect plus nondetect/nondetect)
were tallied and reported as percentage agreement. The misses were indicated as LIF "false
positives" or "false negatives." Because of natural interferences and fluorescent subsurface
minerals, a greater number of false positives than false negatives was expected during the
operation of the LIF technologies. Because the false positive data could be investigated with
additional data analysis, the goal was to keep the number of false negatives to no more than 5
percent.
Other ROST™ attributes evaluated included the ability to collect measurements up to 150 feet
below the surface when the sensor is used with an industry-standard 20-ton CPT rig; the ability
to integrate the sensor subassembly with the rig in the field within a few hours, a standard data
collection rate of one sample every 1.2 seconds, providing a spatial resolution of less than 0.2
feet for a standard push rate of 1 meter per minute; the ability of the system to acquire multi-
dimensional data representations such as wavelength time matrices (WTMs) to identify fuel type;
and the ability of the crew to utilize WTM information to eliminate false positives from
nonhydrocarbon fluorophores. These attributes were evaluated by observing them hi the field
during the demonstration.
Performance audits were conducted in the field to verify that the ROST™ LIF system was
operated according to the procedures outlined in the demonstration plan.
Description of Demonstration Sites
Field demonstrations were conducted at two sites: (1) the Hydrocarbon National Test Site
located at Naval Construction Battalion Center (NCBC) Port Hueneme, California, in May 1995,
and (2) the Steam Plant Tank Farm at Sandia National Laboratories (SNL), Albuquerque, New
Mexico, in November 1995.
Port Hueneme Site Description
The NCBC Port Hueneme site encompasses approximately 4,000 acres on the Pacific coast in
Ventura County, California. NCBC Port Hueneme is approximately 60 miles northwest of Los
Angeles and is located immediately to the west and northwest of the City of Port Hueneme
(Figure 5-1). NCBC Port Hueneme is an active Navy facility where remedial investigation/
feasibility studies (RI/FS) under the Navy's Installation Restoration Program (IRP) are currently
21
-------
Channel Islands Boulevard
Naval Facilities Engineering
Service Center (NFESC)
I • T I itlWiffi ffl '' »' i l *n^ » " i'MM'i ''"in
rdl 'liiiHii i V '.!'. f v:.;j.^:i!"''',' TITS S5 •!"" T^ • "•'' '.Ta^I
N
1000
2000 feet
J
Figure 5-1. Site vicinity map, Port Hueneme; NCBC Port Hueneme area is
delineated by the dashed perimeter.
22
-------
in progress. The demonstration area is located at Site 22, the aboveground fuel farm. Site 22 is
located in the southwestern portion of NCBC Port Hueneme, approximately 1,000 feet west of
Hueneme Harbor and approximately 2,000 feet north of the Pacific Ocean.
Port Hueneme Site History
Site 22 includes five decommissioned aboveground fuel storage tanks numbered 5021, 5022,
5025, 5113, and 5114. The tanks are surrounded by a series of asphalt-paved earthen berms that
restrict surface runoff and which were designed to contain the contents of each tank in the event
of failure. Based on investigative findings during remedial investigation/feasibility study
activities, it appears that all five tanks or their associated piping leaked.
Based on the contaminant type and distribution in the vicinity of Tank 5114, this area was
selected for the demonstration. Tank 5114, a 10,500-barrel capacity tank, was constructed in
1969 and used to store diesel fuel marine (DFM).
Port Hueneme Site Geology and Hydrogeology
The uppermost 1 to 2 feet of soil at Site 22 typically consist of orange-brown silty sand or silt.
Below this interval is a layer consisting of predominantly medium-grained sand, tan in color,
with some coarse and fine-grained sand. This sand layer is approximately 18 feet thick. Site 22
has been built up several feet higher than the surrounding region; the elevation of the ground
inside the berms averages about 17 feet above mean sea level (msl). A dark gray silt layer is
present below the sand layer corresponding approximately to 18.5 feet below ground surface
(bgs). Recent measurements of groundwater elevations in monitoring wells at Site 22 indicate a
groundwater flow direction to the south-southeast. Depth to groundwater is 11 to 13.5 feet bgs.
Port Hueneme Site Contaminants and Distribution
The soils and groundwater in the area around Tank No. 5114 have been contaminated by
petroleum hydrocarbons. The full extent of the contamination has not been assessed; however,
previous site investigations have indicated TPH levels exceeding 70,000 mg/kg to a depth of 20
feet bgs. Predemonstration sampling as part of this effort indicated TPH contamination at 24,000
mg/kg at a depth of 16 feet bgs. Laboratory analysis confirmed that DFM is present in the soil.
Contaminants appear to have migrated vertically and reached their greatest concentration near
the water table.
Predemonstration Sampling and Analysis
A predemonstration sampling and analysis event was performed in accordance with the
demonstration plan to evaluate the demonstration site and the standard analytical methods for
verifying the LIF technologies. The developers conducted predemonstration sampling between
April 4 and 12,1995. Representatives of SNL, U.S. EPA, and State of California EPA
Department of Toxic Substances Control (Cal EPA-DTSC) were present during the predemon-
stration event. During the sampling activities, a number of individual ROST™ pushes were
advanced at the site. Following select pushes, a borehole was advanced adjacent to the
penetrometer hole using a hollow stem auger with split spoon sampler, and discrete soil samples
were collected. The soil samples were shipped to Analytical Technologies, Inc., (ATI) for
23
-------
confirmatory analyses. Representatives of Cal EPA-DTSC collected duplicates for analysis at
the State of California Hazardous Materials Laboratory for verification of contaminants.
In addition to the soil samples submitted to ATI and the State of California Hazardous Materials
Laboratory for chemical analysis, one to two soil samples per boring were submitted to the
Law/Crandall geotechnical laboratory hi San Diego, California. These samples were subjected
to mechanical soil analysis for grain size estimation using ASTM Method 422 and for moisture
and density analysis using ASTM Method 2937.
The results of the:predemonstration sampling and analysis were used by the developers to assess
matrix effects or interferences, revise operating procedures where necessary, and finalize their
performance claims. The developers and representatives of Cal EPA-DTSC, SNL, and U.S. EPA
determined that the site and the contaminant type and distribution were acceptable for the
purposes of this demonstration.
Demonstration Sampling Operations, Port Hueneme
The objective of the sampling design at Port Hueneme was to collect in situ LIF and con-
ventional laboratory analytical data concurrently to demonstrate the ROST™ LIF sensor's
capability to delineate the boundary (field screening) of a petroleum hydrocarbon plume. To
accomplish this, a series of eight iterative pushes and comparison borings were advanced
between Tank 5114 and the expected plume boundary. After each push, a boring was drilled
adjacent to the push hole and sampled. The push and boring locations are depicted in Figure 5-2.
According to the demonstration plan, the SCAPS CPT platform alternatively pushed the SCAPS
LIF probe and ROST™ LIF probe, producing a pair of pushes located approximately 8 inches
apart, prior to the advancement of the comparison boring between the two push holes.
The SCAPS CPT platform was used to push the ROST™ LIF probe and acquire fluorescence
data to a total depth of 16 to 20 feet bgs. Following the pair of pushes, the rig was moved
completely away from the location and a hollow stem auger (HSA) drill rig was positioned with
its stem center approximately 4 inches from the push hole. A hole was drilled using an 8-inch
diameter hollow stem auger such that the internal diameter of the auger was parallel to, and
approximately 2 inches offset from, the LIF probe cavity. Operating within this drilling
geometry, the advancing auger flights destroyed the LIF probe's push hole while allowing for the
collection of split spoon soil samples within approximately 3 inches (horizontally) of the push
cavity. Soil samples were collected with a split spoon sampler lined with 6-inch long, 2.5-inch
diameter stainless steel tubes. The sampler was driven in advance of the lead auger using a 140-
pound slide hammer falling over a 30-inch distance, in accordance with the ASTM 1586
Standard Penetration Test.
Soil samples were collected from every 1 to 1.5 feet of boring starting at a depth of approxi-
mately 2 feet below ground surface. The sampler was overdrilled approximately 6 inches prior
to retrieval to reduce the amount of slough soils typically in the bottom of the borehole. Only
tubes containing sample soils that appeared relatively undisturbed were used.
24
-------
Berm
Berm
B21
B23 B25
• •
B28
•B24
B22
x/v
Berm
S22-W3
B09 SCAPS/Boring Location
I S22-W3 Existing Monitoring Well
N
? T r j
Figure 5-2. Demonstration site and sampling locations, Port Hueneme. B21-B28 are
the locations of the HSA borings associated with the CPT pushes. For example, B21 is the boring associated with
PHDR21, the initial ROST push.
The depth from which samples were collected was measured by lowering a weighted tape before
and after sample retrieval. This permitted identification of the depth from which the samples
were collected in the vadose zone to within approximately 3 inches. In the water saturated zone,
however, sloughing and hydraulic soil movement (flowing or heaving sand conditions) were
encountered which resulted in much greater uncertainty in identifying sample depth.
After each split spoon sampler was retrieved and the individual soil sample collection tubes were
visually inspected, each soil sample was handled as follows:
• The soil sample tube was sealed with Teflon swatches and plastic end caps. The tube was
labeled with the sample identification information.
• The end caps of the sealed, labeled soil sample tube were duct-taped in place. The samples
were placed into an insulated cooler with ice, recorded onto the chain-of-custody form, and
held for shipment to ATI for analysis. The PRC sample custodian and SNL verification
entity verified the accuracy and completeness of the soil sample chain-of-custody forms and
placed a custody seal on the cooler. Original field sheets and chain-of-custody forms
accompanied all samples shipped to the reference laboratory.
• In addition to those soil samples submitted to ATI for chemical analysis, one to two soil
samples per boring were submitted to Law/Crandall's geotechnical laboratory in San Diego,
California. These samples were subjected to mechanical soil analysis to determine grain size
distribution using ASTM Method 422 and for moisture and density analysis using ASTM
25
-------
Method 2937. Those samples determined by grain size analysis to contain a substantial
portion (>25 percent) of fine-grained material (defined as that passing through a #200 sieve)
were subjected to hydrometer testing by ASTM Method 422. Although not part of the
verification process, Law/Crandall, Inc., performed the geotechnical laboratory analyses on
selected soil samples to confirm the visual logging of the borings in the field.
Rinsate samples of the split spoon sampler were collected to check for cross-contamination
after decontamination of the sampler. The rinsate samples were submitted to ATI for
analysis.
Port Hueneme Sampling Locations
The sampling locations were in a line running west to east located south of Tank 5114 (Figure
5-2). The first ROST™ push was located in what was estimated to be an area within the plume
and identified as PHDR21, at 6 feet east of the 0-foot location (Table 5-1). The first boring was
advanced and sampled immediately after the probe was retrieved and the CPT rig was moved
away. A second push, designated as PHDR22, was then advanced in an area estimated to be
outside of the plume boundary. The second boring was advanced and sampled immediately after
the probe was retrieved and the CPT rig was moved away. The strategy was to advance the first
two pushes in locations that would bound the edge of the plume and then locate subsequent
pushes, PHDR23-PHDR28, in an effort to close in on the horizontal extent of the plume. The
distance between each successive push decreased until the edge of the subsurface hydrocarbon
plume had been defined within 9 feet, for a total of 8 borings. The number of sampling locations
was based on past use of the CPT and LIF technology to define hydrocarbon plume boundaries at
other sites and on demonstration budget constraints.
Table 5-1. Port Hueneme boring and push summary table.
Push or Boring
Identification
PHDR21
PHDB21
PHDR22
PHDB22
PHDR23
PHDB23
PHDR24
PHDB24
PHDR25
PHDB25
PHDR26
PHDB26
PHDR27A
PHDB27
PHDR28
PHDB28
Date
5-17-95
5-17-95
5-18-95
5-18-95
5-19-95
5-19-95
5-22-95
5-22-95
Comments
R21 located 6 feet 8 inches east of zero point; B21 located 4 inches west of
R21. Total of 15 samples collected; max depth 19 feet.
R22 located 200 feet east of zero point; B22 located 4 inches east of R22.
Total of 1 7 samples collected; max depth 1 9.5 feet.
R23 located 53 feet 8 inches east of zero point; B23 located 4 inches west of
R23. Total of 16 samples collected; max depth 19 feet.
R24 located 162 feet east of zero point; B24 located 4 inches east of R24.
Total of 21 samples collected; max depth 19.5 feet.
R25 located 81 feet 8 inches east of zero point; B25 located 4 inches west of
R25. Total of 16 samples collected; max depth 20 feet.
R26 located 142 feet east of zero point; B26 located 4 inches west of R26.
Total of 17 samples collected; max depth 20 feet.
R27A advanced through 6 feet pilot hole of approximately 6 feet bgs. R27A
located 156 feet 4 inches east of zero point. B27 located 4 inches east of
R27A. Note that push R27 was refused in upper 5 feet due to gravel and
cobble. Total of 19 samples collected; max depth 19.5 feet.
R28 advanced through 6-ft pilot hole 148 feet east of zero point. B28 located 4
inches east of R28. Total of 17 samples collected; max depth 18.5 feet.
Note: PHDR represents the ROST™ push at the Port Hueneme site. PHDB represents the hollow stem auger boring.
26
-------
Each boring using the hollow stem auger and split spoon sampler was identified with a unique
number assigned in the field. For example, PHDB21 identified the boring (B) that was
collocated with the initial ROST™ (PHDR21) push. Individual samples collected from each
boring were sequentially numbered as they were logged; for example, PHDB21-5 identified the
fifth soil sample collected from boring B21. Each sample was submitted for analysis
accompanied by the chain-of-custody documentation.
Note that PHDR27A represents the second ROST™ push attempted at the location indicated in
Figure 5-2. The first push was refused due to an impenetrable gravel/cobble layer within 6 feet
of the surface. PHDP27A was offset 8 inches from PHDR27 and was advanced without
difficulty. It was attempted only after a pilot hole was advanced using an uninstrumented
(dummy) probe. After extraction of the dummy probe, the ROST LEF probe was advanced
though the pilot hole and LIF measurements wefe collected throughout the push. Because
PHDR28 was located within 9 feet of PHDR27A and it was assumed that the subsurface
gravel/cobble layer would be encountered, a dummy probe was also used to prepush this
location.
SNL Tank Farm Site Description
The location for the second LIF demonstration was an active fuel tank farm for the Steam Plant
at Sandia National Laboratories, Albuquerque, New Mexico (Figure 5-3). This site was selected
because it represented a different climate, geology, and contaminant distribution than the Port
Hueneme demonstration site. The Tank Farm site is an SNL Environmental Restoration Site that
is currently being characterized and will begin a remediation feasibility investigation beginning
in 1998. It is located in the southwest portion of Technical Area I on the northeast corner of the
intersection of Hardin and Wyoming Boulevards (Figure 5-4). The 3-acre site is L-shaped and
contains five tanks. The area west and north of Tank 5 was the area for this demonstration.
Site History, SNL Tank Farm
The Steam Plant Tank Farm was constructed in the 1940s. All tanks contained #2 diesel fuel to
be used as a backup supply system for the Steam Plant when the primary fuel supply (natural
gas) was unavailable. The backup supply system has never been used and the fuel currently in
the tanks is the original product delivered. One documented release of fuel occurred in June
1991, when the main valve of Tank 5 was left open and more than 5,000 gal of fuel was
discharged into a holding tank at the Steam Plant (approximately one-half mile north of the tank
farm). During transfer operations from the holding tank to another storage tank south of Hardin
Boulevard, a leaking pipe was discovered. The pipe was then cut and capped, and the impacted
soils in the area were scheduled for excavation. A few weeks later during excavation operations,
it became evident that the fuel release was much greater than previously thought. Although the
full horizontal and vertical extent of the plume was not determined, the 50 foot by 35 foot by 15
foot deep excavation pit was backfilled with the original fuel-contaminated soil. Recent site
investigations using a Geoprobe® identified petroleum contamination down to at least 30 feet
bgs in the area of the excavation.
SNL Tank Farm Site Geology/Hydrogeology
SNL is located near the east-central ridge of the Albuquerque Basin. The basin is a rifted graben
within the Rio Grande Rift that is bounded on the east and west by north-south trending faults.
SNL lies on a partially dissected bajada formed by coalescing alluvial complexes. The deposits
27
-------
1-25
1-40
North
Albuquerque
Gibson Blvd
1
BD
Sandia National Laboratories
Hardin Blvd
Kirtland Air Force Base
Figure 5-3. Site vicinity map, SNL Tank Farm
on the surface are composed of alluvial fan deposits shed from the eastern uplifts that interfinger
with valley alluvium and consist of clayey to silty sands, with lesser amounts of silt, clay, and
sand. Surficial deposits are underlain by a thick sequence (greater than 5,000 feet) of basin-fill
deposits of interbedded gravels, sands, silts, and clays. Depth to groundwater is approximately
500 feet, with the potential for perched water at shallower depths. During the exploratory and
informal predemonstration investigations, the SCAPS CPT consistently met with refusal at a
depth of 52-57 feet, due to a consolidated gravel/caliche layer at this depth.
SNL Tank Farm Site Contaminants and Distribution
The SNL Geoprobe® investigations and the preliminary investigations using the SCAPS LIF
sensor indicated diesel contamination greater than 1000 mg/kg in the vadose zone down to 56
feet. The area that was excavated down to approximately 15 feet and subsequently backfilled
with the contaminated soil contains a somewhat homogenized mixture of diesel contaminated
soil and uncontaminated soil. A high concentration of subsuface fluorescing minerals, most
likely calcium carbonate, was identified prior to the demonstration. Calcium carbonate is
present to some degree throughout the vadose zone in this area; it is more concentrated near the
surface.
28
-------
I
eo
f
Tank 1
Tank 2
Tank 3
o
o
o
SNLDP12/B12
:: Tank 4
O-
\SNLDP10/B10
DP11/B11
APPROXIMATE
LOCATION OF
LEAKING FUEL
TRANSFER LINE
O
TankB
Hardin Boulevard
100
50
SCALE 1" = 1O01
Figure 5-4. Demonstration site and sampling locations, SNL Tank Farm
SNL Tank Farm Predemonstration Sampling
A formal predemonstration event was not conducted at the SNL Tank Farm site. The site was
evaluated for its suitability as a demonstration site during a site exploratory tour by the NCCOSC
RDT&E Division in August 1995. Two other arid locations were evaluated at this time and
determined to be unsuitable for this demonstration. Immediately prior to the field demonstration
in November, NCCOSC RDT&E Division personnel performed informal sampling to determine
sampling locations for the demonstration. Earthen berms had been removed to allow access to
contaminated areas. Following select pushes, stab samples (discrete soil samples collected using
the cone penetrometer soil sampling apparatus) were collected and shipped to ATI for overnight
confirmatory TPH and TRPH analysis. Laboratory analysis of the stab samples indicated TRPH
of 3380 mg/kg and TPH of 3300 mg/kg (as diesel) at a depth of 25 feet. Carbonate was
observed in all the discrete soil samples in varying concentrations by the professional geologist
29
-------
and confirmed by applying hydrochloric acid, causing release of carbon dioxide, to a few
representative samples.
Demonstration Sampling Operations, SNL Tank Farm
The sampling operations at the SNL Tank Farm were similar to the operations at Port Hueneme,
with the following changes.
Because tiie horizontal extent of the plume at Port Hueneme Site 22 had been delineated to
within 9 feet with 8 pushes during the field demonstration, this capability of the ROST™
technology was not the primary focus of the second demonstration. For the SNL Tank Farm
demonstration, the developers and representatives of SNL and U.S. EPA determined that it
would be preferable to collect more samples from areas expected to be contaminated to compare
the LIF technology with the results from the reference laboratory analysis of discrete soil
samples. The addendum to the demonstration plan reflected this change to the sampling strategy.
For the demonstration, three pairs of CPT pushes were advanced, followed by three overborings.
Based on the results of the informal predemonstration, the first pair of CPT pushes and boring
were located in an area that had contamination throughout the push, the second pair of pushes
and boring were advanced in an area that had contamination from approximately the 40 to 50 feet
depth, and the third pair of pushes and boring were advanced in an area expected to be
uncontaminated.
During the Port Hueneme demonstration, samples were collected throughout the contaminated
and uncontaminated areas at intervals of every 1 to 1.5 feet. The experimental design called for
several pushes to be located in clean areas in order to delineate the horizontal extent of the
plume. This resulted in a large quantity of clean samples in the data set (114 nondetects of 130
total samples as determined by the reference laboratory). For the demonstration at the SNL Tank
Farm, the experimental design was modified to focus discrete sampling in the impacted areas and
limit the number of samples in areas expected to be unimpacted. This conserved resources and
allowed for more comparisons of hydrocarbon-impacted samples (68 detects of 92 total samples
as determined by the reference laboratory).
During drilling operations, discrete soil samples for reference laboratory analysis were collected
using a California modified split-spoon sampler lined with 2.5-in diameter by 3-in stainless steel
tubes. The smaller size of the sample tube was selected to allow for a greater number of discrete
samples to be collected during a single 24-inch sample drive and would also permit finer scale
resolution of the comparison of the LIF response to the reference laboratory analytical results. In
addition, fewer samples were collected in the unimpacted boring. A total of 92 soil samples were
collected during this demonstration, compared to 130 for the Port Hueneme demonstration.
AH demonstration samples were collected and documented as previously described. Each
ROST™ push was identified with a unique number assigned in the field (Table 5-2). For
example, the tenth ROST™ push was identified as SNLDR10 (SNL Demonstration, ROST™
push 10). Each boring was uniquely identified, such as SNLB10 for the boring (B) that was
collocated with the SNLDR10 push. Individual samples collected from each boring were
sequentially numbered as they were logged; for example, SNLDB10-5 identified the fifth soil
sample collected from the tenth boring.
30
-------
Table 5-2. SNL Tank Farm boring and push summary table.
Push or Boring
Identification
SNLDR10
SNLDB10
SNLDR11
SNLDB11
SNLDR12
SNLDB12
Date
11-6-95
11-7-95
11-8-95
Comments
R10 located 2 feet 8 inches north and east of fuel transfer line. B10 located 4
inches offset from R10. Total of 53 samples collected; max depth 56.25 feet.
R1 1 located 9 feet 8 inches south and west of fuel transfer line. B1 1 located 4
inches offset from R1 1 . Total of 28 samples collected; max depth 55.25 feet.
R12 located 50 feet north and east of R10. B12 located 4 inches offset from
R12. Total of 20 samples collected; max depth 49.5 feet.
Note: SNLDR represents the ROST™ push at the SNL Tank Farm Demonstration. SNLDB represents the hollow stem
auger boring.
Calibration Procedures, Quality Control Checks, and Corrective Action
Calibration procedures, method-specific QC requirements, and corrective action associated with
nonconformance QC for the LIF technology are described in the following paragraphs.
ROST™ LIF Initial Calibration Procedures
The time window (typically 250 ns wide) on the digital oscilloscope is adjusted to compensate
for the light transmit time through the optical fiber. For a 50-meter long push, the fluorescence
signal is received at the detector about 500 ns after the laser has actually fired. Once set, the
time delay needs to be adjusted only if the length of the fiber in the probe umbilical is changed.
The position of the ROST™ time window can be determined automatically with routines built
into the scope's software. The procedure was carried out prior to the demonstration.
A wavelength calibration for the emission monochromator was performed at the start of the
demonstration and thereafter during troubleshooting procedures. The 532 nm Nd:YAG second
harmonic light was used as a primary reference to verify the wavelength accuracy of the
monochromator. A small amount of 532 nm light was directed into the monochromator at a
narrow slitwidth, and the wavelength was scanned to verify that the signal maximizes at 532±0.2
nm. The monochromator can then be used as a secondary reference to calibrate the dye laser
wavelength.
The concentration calibration was performed using a set of calibration standards (DFM-spiked
site-specific soil samples) prepared by the serial addition method. The calibration standards
. were run in triplicate at the beginning of each day and again when equipment was changed.
These samples were sequentially presented to the sapphire window for measurement. After
measurement, the average and standard deviation was computed for each sample. If the standard
deviation exceeded 20 percent for replicate analyses of any single sample, that sample was rerun.
If deviation remained excessive, the system check standard was measured. If the check standard
was out of compliance, system checkout and debugging was required. A calibration curve was
generated by regressing fluorescence peak intensity expressed as a percentage of a reference
solution versus the concentration of fuel product added to the calibration soil sample.
ROST™ LIF Continuing Calibration Procedures
A fluorescence reference measurement was performed before and after each push for normali-
zation purposes and to check system performance. The reference material, referred to as M-l
31
-------
reference solution, is a selected mixture of hydrocarbons in solution. The M-l reference solution
is contained in a standard fluorescence cuvette that can be reproducibly strapped onto the cone
outside the sapphire window. If the reference intensity at the conclusion of the push differed by
more than 25 percent of the value immediately before the push, system troubleshooting
procedures were initiated.
Method Blanks
A clean sand blank was measured pre- and post-push as part of the standard data collection
procedure. If the clean sand blank LIF measurement varied beyond 50 percent of its pre-push
calibration value, troubleshooting procedures were initiated.
Spike Samples
Spiked samples were not used for monitoring the performance of the HOST™ LIF system. In situ
measurement precludes the presentation of spiked samples to the LIF measurement system.
Instrument Check Standards
A system check using the M-l reference solution was performed before and after ROST™ LIF
data collection. Both wavelength and intensity of the standard were monitored. If the
wavelength differed by greater than 5 nm from the known value, a wavelength calibration was
performed. If the intensity changed by more than 20 percent, system troubleshooting was
required.
Performance Evaluation Materials
Performance evaluation (PE) samples were not used for this demonstration. Because the LIF
technologies are in situ measurement techniques, PE samples cannot be inserted into these
dynamic measurement processes.
Duplicate Samples
Due to the nature of the in situ measurement, duplicate samples cannot be measured by LIF. In
an homogeneous environment, nearby pushes are a near duplicate measurement. Soil hetero-
geneity and variation in contaminant distribution can be significant over short distances both
horizontally and vertically. Therefore, the quality assurance plan included no in situ duplicate
measurements.
Equipment Rinsate Samples
To assess whether cross contamination was being introduced during equipment decontamination,
an equipment rinsate sample was collected daily. The source of the water for the equipment
rinsate sample was the deionized water used for the final rinse step of the equipment decon-
tamination process. Deionized water was poured over the sampler and into vials equipped with
Teflon seals in a manner so that headspace was minimized. The equipment rinsate samples were
sealed, labeled, and placed into an insulated cooler, logged on the chain-of-custody form, and
32
-------
submitted to ATI for analysis of TRPH and TPH using the reference analysis methods previously
described.
Data Reporting, Reduction, and Verification Steps
To maintain good data quality, specific procedures were followed by the developer and the SNL
verification entity during data reduction, validation, and reporting. These procedures are detailed
below.
Data Reporting
The following data were reported to SNL:
1. Data logs from all pushes, including ROST™ fluorescence as a percentage of M-l
fluorescence with respect to depth. Also provided were wavelength time matrices for select
positions along each push.
2. System check and calibration sample concentrations; tabulated raw system check and
calibration sample fluorescence data; average system check intensity and system check ratio
for each push; background, noise, and sensitivity calculated from calibration data.
3. Borehole logs indicating soil sample collection information, including sample numbers,
depth of samples, location of water table, and other relevant information concerning the
collection of the soil samples, and chain-of-custody documentation associated with soil
samples.
4. Laboratory results for TPH and TRPH measurements of soil samples, including the reference
method analytical results and quality control data.
Data Reduction and Verification Steps for,the ROST™ LIF Data
The LIF system records the fluorescence as a percentage of the M-l standard as a function of
depth as the probe is pushed into the ground. This raw data is calibrated using the system check
standard measured before and after each push, and the series of calibration samples measured on
a daily basis during the site operations. The raw data and daily calibration procedures were used
to make decisions in the field. Following the conclusion of site operations, the raw fluorescence
measurements were adjusted by a normalization factor, and a site-wide regression slope was
computed to the detection limits. This procedure is detailed below.
1. Each day, calibration curves were established using the DFM-spiked samples prepared prior
to the demonstration. The resulting sensitivity (the slope of the line estimated using all
calibration data) was used to determine the limit of detection (LOD) in mg/kg.
2. Each fluorescence versus depth (FVD) log was analyzed to determine if depth data from the
depth encoder were correct.
3. Each FVD was analyzed to determine the background signal for each push. Background
signals are assumed to generate a bell-shaped curve at the low end of the histogram centered
around the mean of the background generated signal. The calculated mean of the bell shaped
curve is then used to represent the background for that push. The background is subtracted
from each percent fluorescence measurement to produce a background-corrected data set.
The standard deviation is used as an estimate of noise. The LOD is calculated as 2.58
standard deviations (the 99th percentile) added to the background.
33
-------
4, To compare the in situ data with the soil sample analysis results, the percent fluorescence
measurements taken at depths from which the soil samples were gathered were tabulated.
Because the spacing between LEF data points is approximately 2.4 inches, the fluorescence
data from all points corresponding to the 6-inch interval of soil from Port Hueneme were
averaged to produce a single fluorescence intensity for a given sample. For the SNL demo,
the sample interval was 3 inches. Therefore, the percent fluorescence data corresponding to
the 3-inch sample interval were averaged to produce a single percent fluorescence measure-
ment to compare with the analytical results.
5. Fluorescence data were reduced to a detect or nondetect reading using the limit of detection
(LOD) determined in Step 3 above. The average percent fluorescence reading corresponding
to each soil sample was compared to the fluorescence threshold. Those exceeding the LOD
were recorded as detects; those falling below the LOD were recorded as nondetects.
6. Results from the reference laboratory were also reduced to a detect or nondetect reading.
The laboratory analytical result (TPH and TRPH) for each soil sample was compared to the
LOD in mg/kg. Those exceeding the LOD were recorded as detects; those falling below the
LOD were recorded as nondetects. This LOD in units of mg/kg was computed using the
fluorescence LOD (less background) divided by the sensitivity as described in Step 1 above.
7. Field notes and photographs were reviewed to verify that procedures outlined in the
demonstration plan were followed.
8. On-site system audits for field operations and procedural quality assurance audits were
conducted by SNL in the field while the demonstration was being conducted. Audit results
are reported in Section 6. Specifically, the ROST™ system and operators were audited for
compliance with the draft ROST™ method provided in Appendix C.
Changes to the Demonstration Plan
During the Port Hueneme demonstration, it was agreed that the developers would use the daily
calibration results from Port Hueneme and reduce the data after the demonstration according to
their standard procedure (as described above), which was slightly different than described in the
demonstration plan, to arrive at a site-specific detection limit for Port Hueneme. They used the
above-described method for data reduction for the SNL data set.
Information from single point tests (SPTs) on homogenized soil samples following reference
laboratory analysis were allowed for both developers for both demonstrations. SPTs are
measurements taken by placing a homogenized portion of a discrete sample (after laboratory
analysis was complete) on the LIF probe and recording the fluorescence intensity. This intensity
can be compared to the reported laboratory result for the original sample and to the in situ
fluorescence intensity to determine if the sample analyzed by the laboratory was collected at a
different depth than the depth of the in situ sensor measurement. The SPTs for the Port
Hueneme demonstration were performed by NCCOSC RDT&E Division. Results from the SPTs
were used to adjust sample depths for discrete samples; adjustments affected the results from
both technologies similarly.
For both demonstrations, calibration standards were prepared using site-specific soil. The
standards were measured daily at the start of operations. During the SNL Tank Farm
demonstration, it was determined that the soil collected at the surface for preparation of the
standards was not representative of the nonimpacted soil at the site. The soil down to a depth of
10-15 feet had been excavated near the leaking fuel transfer line in order to repair the line, and
then had been returned without remediation. In addition, the soil near the surface had a large
concentration of calcium carbonate, which fluoresces quite strongly under UV light. Because the
34
-------
calibration standards prepared prior to the demonstration showed a strong fluorescence signal, it
was agreed by all parties that a revised set of calibration standards would be prepared using soil
more representative of the subsurface environment. This soil was collected at a depth of 36 feet
bgs using the split spoon sampler during advancement of boring SNLDB12, the nonimpacted
location. The developer reported that the background signal produced using the newly prepared
calibration standards did not appreciably affect their data set, and elected not to use the new
calibration standards.
35
-------
Section 6
Technology Results and Evaluation
The purpose of this section is to present and evaluate the ROST™ LIF results from the two
demonstrations performed as part of this program. First, the developer claims are presented.
Second, the accuracy, precision, and completeness of the ROST™ LIF data set are provided.
Third, the ROST™ LIF results are compared to the laboratory results, and the performance of the
technology is evaluated against the developer claims. Finally, a summary of the performance
evaluation is given at the end of this section.
The in situ LIF results from both demonstrations are presented in Appendix B. The raw LIF data
have been analyzed by SNL and presented in this section in a variety of formats to compare them
with the reference laboratory results and to determine if the developer claims were met. The
graphical depictions of the ROST™ LIF data were developed from the original data set.
Developer Claims Presented
As stated in Section 5, the purpose of the demonstration was to generate appropriate field data to
verify the performance of the technology as a field screening tool for identifying petroleum
hydrocarbons in the subsurface. To accomplish this, two different sites were selected for
demonstration locations: a shallow, coastal site and a deep, arid site. The LIF data were
evaluated to determine the technology's performance relative to developer claims made in the
demonstration plan. The LIF sensor data were compared to the data from laboratory soil
analyses and the ROST™ CPT platform was compared to conventional sampling methods.
Specific claims for the ROST™ LIF sensor presented in the demonstration plan were:
1. The ROST™ system can be integrated with cone penetrometer trucks from all major
manufacturers. Field integration is routinely accomplished in a few hours.
2. Standard data collection rate is one sample every 1.2 second, providing a vertical spatial
resolution of 0.2 feet for a standard push rate of 1 meter per min.
3. The system can acquire multidimensional data representations, such as WTMs, to identify
fuel type or to eliminate false positives from non-hydrocarbon fluorophores.
4. Target of 80 percent agreement with conventional laboratory analysis of samples and 5
percent false negatives.
These claims were evaluated individually and collectively throughout the demonstration and in
post-demonstration data analysis. Results are summarized at the end of Section 6.
Technology Data Quality Assessment
Data generated by the ROST LIF technology were compared to the data generated from analysis
of soil samples using the two analytical methods. The quality of the reference laboratory data
has been previously discussed, and all laboratory data were determined to be acceptable for
comparison to the LIF technology data. The following LIF data quality indicators were closely
examined to determine if the technology data were of sufficient quality to be compared to the
36
-------
reference laboratory data. The indicators evaluated for the ROST™ LDF technology were
precision and completeness. The accuracy of the data was assessed upon comparison to the
laboratory results.
Precision
Precision refers to the reproducibility of measurements of the same characteristic, usually under
a given set of conditions. Unfortunately, the conditions can vary hi environmental data to an
extent that leaves the term ambiguous. Differences from site to site, sample to sample within a
site, and differences in results from repeated measurements from a single sample provide
examples. Because the ROST™ LIF sensor's primary utility is for in situ sensing as the probe is
pushed into the ground, it was not possible to obtain precision data for the sensor under
conditions that exactly duplicated the manner in which in situ measurements are made in the
subsurface.
During the Port Hueneme demonstration, an estimate of the instrumental precision was obtained
by placing a standard cuvette containing M-l reference standard in front of the sapphire window
and measuring the sample 20 times (50 laser shots for each analysis). This is the same as the
system check procedure used before and after each push. Because the system check standard is a
liquid, it was considered to be homogenous. This procedure provided an estimate of the precision
of the instrument. The standard deviation of the 20 measurements was 2.2% of the mean count.
Completeness
Completeness refers to the amount of data collected from a measurement process compared to
the amount that could be obtained under ideal conditions. For this demonstration, completeness
refers to the proportion of valid, acceptable data generated using each method. It was anticipated
that less than 100 percent completeness of both the LIF data and discrete sample analysis results
would occur. For LIF data collection, a push that was refused due to contact with cobbles or
other obstructions was disqualified. A substitute push was advanced in these cases, within 8
inches horizontally of the disqualified push. This occurred on ROST™ push 27 at Port Hueneme.
At this site, the refusals occurred near the surface, so the subsequent push (PHDR27A) allowed
for LIF data to be collected near the same location. As long as the substitute push was located
within 8 inches, the disqualified push was not counted against the completeness goal. Therefore,
the completeness was 100 percent for Port Hueneme. At SNL, preliminary pushes had indicated
an impenetrable gravel/caliche layer at approximately 50-58 feet bgs. While this was able to be
penetrated by the HSA rig, the cone penetrometer was not advanced past this depth. The pushes
were considered to be complete at the point of refusal. Therefore, the LIF data set was
considered 100 percent complete for the SNL site.
Accuracy
Accuracy refers to the degree of agreement of a measurement to the true value. For an in situ
field screening measurement technique such as LIF, determining the accuracy of the technique
presents a particular challenge. This is because it is not a simple matter to confidently assign a
"true" value to a subsurface contaminant distribution. When compared to conventional
laboratory-based measurements, the accuracy of the method is a function of both the sampling
errors and errors associated with the measurement method.
37
-------
Because there is no independent measure of the subsurface value of contaminant concentration,
the accuracy of the in situ measurement was assessed by comparing it to results from
conventional laboratory measurements. The percent agreement between TRPH or TPH and
fluorescence data and percent false negatives was calculated using the equations that follow.
Where:
x..=
% Agreement = —— -+• 100%
XT
Number of samples where fluorescence is less than the detection threshold and
the corresponding lab result is also less than the corresponding detection limit;
Number of samples where fluorescence is greater than the detection threshold
and the corresponding lab result is also greater than the corresponding detection
limit; and
Total number of samples collected for comparison.
% False Negatives =
100
XT
Where:
XH- — Number of samples where fluorescence is less than the detection threshold and
the corresponding lab result is greater than the corresponding detection limit.
The average of the ROST™ LIF measurements corresponding to a 6-inch interval (Port
Hueneme) or a 3-in interval (SNL Tank Farm) were compared to TRPH and TPH results for a
discrete sample collected at the same depth. The results of the comparison are shown in Table
6-1. The laboratory result for TPH and TRPH from each homogenized soil sample was com-
pared to the corresponding limit of detection established by the developer in the field. If the
laboratory result was above the LOD and the average LIF data from the push at the correspond-
ing depth exceeded the LIF fluorescence LOD, the result was a "detect/detect." If the average
LIF data were below the threshold and the corresponding analytical data were above the
corresponding LOD, the result was a "false negative." If the average LIF data were above the
LOD and the laboratory results were below the corresponding concentration LOD, the result was
a "false positive." If the average LIF data and laboratory results were below the fluorescence
LOD and corresponding concentration LOD, the result was "nondetect/nondetect" agreement.
This process was performed on each sample. The results provided the determination if the
developer's claims of 1) 80 percent agreement and 2) less than 5 percent false negatives were
achieved.
For the Port Hueneme demonstration, the accuracy achieved by the LIF technology was 87.7
percent agreement of LIF data with TRPH data, with 10 percent false negatives and 2.3 percent
false positives. Compared to TPH results, the technology achieved 89.2 percent agreement with
TPH results, with 5.4 percent false negatives and 5.4 percent false positives. For the SNL Tank
Farm demonstration, the accuracy achieved by the technology was 93.4 percent agreement with
either TRPH or TPH data, with 3.3 percent false negatives and 3.3 percent false positives when
compared to either TRPH or TPH.
38
-------
Port Hueneme Site Data Presentation and Results
The data presented in this section are used to assess of the ability of the ROST™ LIF to provide
field screening and mapping of subsurface contaminants in a shallow, coastal site with
contamination in the vadose zone, capillary fringe, and saturated zone. The percentage
agreement with the laboratory results of soil samples from the Port Hueneme demonstration site
is reported in this section.
Port Hueneme Detection Limit
As described in Section 5, the LOD was determined on a push-by-push basis in the field during
the demonstration, and a composite site sensitivity was calculated for determination of agreement
with the laboratory results. For the Port Hueneme site, the ROST™ site LOD was 5 mg/kg.
Because the soil samples were 6 inches long, the fluorescence for the 6-inch interval associated
with each sample was averaged and compared to the LOD. The reference method data were
considered to show a detect when the value exceeded the Port Hueneme LIF site detection limit
of 5 mg/kg. When the average in situ fluorescence result exceeded the fluorescence LOD, this
was designated a "detect." The actual results for the Port Hueneme demonstration are presented
in Table 6-1. The results indicate that the LIF data correlate better with the TPH results, which
may be due to the humic interferences common to TPRH analysis. The instances where matches
or misses occurred are listed in Appendix A, Table A-l.
Table 6-1. Summary of comparison of Port Hueneme LIF data with laboratory
data.
Category
LIF/Lab
Nondetect/Nondetect Match
Detect/Detect Match
Nondetect/Detect Miss ("FN")
Detect/Nondetect Miss ("FP")
Total Samples
Percent ND/ND of Total
Percent DID of Total
Percent Matches of Total
Percent ND/D Misses ("FN") of
Total
Percent D/ND Misses ("FP") of
Total
Compared to TRPH result
91
23
13
3
130
70%
17.7%
87.7%
10%
2.3%
Compared to TPH result
97
19
7
7
130
74.6%
14.6%
89.2%
5.4%
5.4%
Downhole Results for Port Hueneme
The LIF results obtained during five contaminated pushes at Port Hueneme have been plotted in
Figure 6-1. These five plots indicate the pushes and associated borings along the transect near
Tank 5114. The corresponding soil sample collection locations and results are also indicated.
The square symbols indicate the locations and results of the single point tests. As discussed in
Section 5, during the predemonstration event there was a depth discrepancy observed with the
hollow stem auger and split spoon sampling operation, believed to be due to sloughing of sands
in the saturated zone. This was also observed during the demonstration. The reference
laboratory provided splits of the homogenized samples from the demonstration to NCCOSC
39
-------
KDT&E Division to perform single point tests (SPTs) at their facility after the demonstration.
NCCOSC RDT&E Division personnel placed portions of the homogenates on the LIF probe
window, and the fluorescent intensity was measured. SPT results were compared to the in situ
measurements obtained during the demonstration.
On review of the SPT measurements and in situ results for both technologies demonstrated, SNL
determined that on two holes, a slight offset was apparent that affected the results of the
laboratory measurements when compared to the data from both technologies. Field notes were
reviewed to determine where sloughing of soils was most prominent. SNL determined that for
holes 23 and 28, depth adjustments of 6 niches and 4 inches, respectively, for the laboratory
samples collected in the saturated zone was appropriate. This adjustment supported (i.e.,
improved) the percentage agreement results from both LIF technologies. All downhole results,
including the adjusted data for holes 23 and 28, are presented in Figure 6-2.
Port Hueneme Subsurface Contaminant Mapping
The test area at Port Hueneme and the transect along which LIF pushes and hollow stem auger
borings were advanced to collect data for the performance evaluation are illustrated in Figure
6-3. Each symbol along a hole indicates a location where a soil sample was collected. The TPH
result of each data point was compared to the LIF measurement at the corresponding depth
interval. The area of the plume was estimated based on the laboratory measurements. This
figure illustrates several points:
• the contaminant plume was narrow, and the false positives and false negatives, in general,
were located at the plume boundaries,
• the LIF field screening technology was able to determine the horizontal extent of the plume
within 9 feet, based on the results of 8 pushes, and
• soil samples were collected at 1- to 1.5-foot intervals and often missed the boundaries of the
plume.
40
-------
HOLE21
£15000
*10000
f 5000
!> 0
748% :
10 15
Dq)th(ft)
20
HOLE23
HODB26
15,4000--
J3.2000
10
Depth (ft)
HOLE25
10
Depth (ft)
HOIB28
94%
E
0 1
20
Figure 6-1. Downhole results for Port Hueneme. Results from the five drilling locations where
single-point tests were evaluated are illustrated above. The laboratory measurements are indicated by the circles, the
single point test measurements are indicated with the square symbols, and the LIF results are indicated by the
continuous solid line. The horizontal axis is indexed by both concentration in mg/kg as measured by the average of the
analytical methods and in % fluorescence measured by the ROST™ LIF technology. Note: It is inappropriate to
compare the relative magnitude of the laboratory concentration to the LIF peaks as the LIF results are not linear at
higher concentrations.
41
-------
HOLE21
30KO
MOOD
Kcm
-HX>—CO—-OO--H>
<~oo-e^-
15 30
HOLE 25
•».
HOLE 22
-oo—ea—oo oo | oo oo op
•*•
aooo J.
feml
imol
1 20004-
HOLE 26
HOLE 23
HOLE 27
C"l
•»A
HOLE 24
no—"" | **** "•*
HOLE 28
2BQOT
. 2000 J
h 1B004
» 1000 j
- 500 4
Figure 6-2. Adjusted downhole results for Port Hueneme. This figure provides a summary of
all downhote results after adjustment for depth measurement inaccuracies due to sloughing that appears to have
affected the measurements in the saturated zone at boring locations 23 and 28. Note that the vertical axes for the
notes beyond the plume boundary, holes 22,24, and 27, have a smaller scale than holes 21,23,25,26, and 28,
where contamination was detected.
42
-------
LIF,T-ab Result
o Nondetect/Nondetect
• Detect/Detect
-A. Detect/Nondetect
^ Nondetect/Detect
Figure 6-3. Cross-sectional map of transect near Tank 5114 at Port Hueneme.
contour view of the results shown in Table 6-1, comparing the TPH results to those of the ROST™.
SNL Tank Farm Site Data Presentation and Results
As described in the addendum to the demonstration plan, the purpose of the SNL Tank Farm
demonstration was to demonstrate the capabilities of the LIF technology at an arid site with a
deeper hydrocarbon plume in the vadose zone. Again, the percentage agreement of the LIF
technology data set with the laboratory analytical results of soil samples from the SNL Tank farm
site provides the basis for evaluation.
SNL Tank Farm Detection Limit
For the SNL Tank Farm site, a detection limit was determined on a push-by-push basis in the
field during the demonstration. For the SNL Tank Farm site, the LODs for the three pushes were
102.4 mg/kg (PHDR10), 77.8 mg/kg (PHDR11), and 41.0 mg/kg (PHDR12). The TRPH and
TPH measurements for each push were considered to show a detect when their values exceeded
these limits.
During the demonstration, it was realized that the site-specific background soil to be used for
preparation of calibration soils had been collected from the area that had been previously
excavated. This soil had a high concentration of fluorescent minerals and a high background
43
-------
reading. A second set of calibration soils was collected at 36 feet bgs from boring 12 (the
uncontaminated push/boring). This second set showed a more typical fluorescent response. It
was agreed by SNL and the developers that this soil could be used to prepare a second set of
calibration standards after the demonstration. However, the ROST™ developers elected not to
use this second set of calibration standards as their performance did not improve when using the
new calibration standards.
The results of the comparison are presented in Table 6-2. This table summarizes how well
ROST™ LIF qualitative results (detect or nondetect) matched those of the laboratory methods.
The table indicates better results than those of the Port Hueneme demonstration, in terms of
match and miss percentages. This is most likely due to the higher detection limits for this site,
which reflect the higher background fluorescence at SNL. The discrepancies are in regions that
are impacted at levels close to the ROST™ LIF detection limit and in the areas where high
carbonate fluorescence was observed. In the case of boring/push 10, the regions are separated
from the plume because of the excavation, which redistributed hydrocarbon contamination near
the surface.
Table 6-2. Summary of comparison of SNL LIF data with laboratory data.
Category
LIF/Lab
Nondetect/Nondetect Match
Detect/Detect Match
Nondetect/Detect Miss ("FN")
Detect/Nondetect Miss ("FP")
Total Samples
Percent ND/ND of Total
Percent D/D of Total
Percent Matches of Total
Percent ND/D Misses ("FN") of Total
Percent D/ND Misses ("FP") of Total
Compared to TRPH result
22
64
3
3
92
23.9%
69.5%
93.4%
3.3%
3.3%
Compared to TPH result
22
64
3
3
92
23.9%
69.5%
93.4%
3.3%
3.3%
At the Sandia Tank Farm demonstration, naturally occurring fluorescent minerals in the soil (i.e.,
carbonates) caused some false positive results for this test. Although carbonates occur naturally
throughout the vadose zone in desert environments and were observed in soil samples at all
depths, they were especially concentrated within 14 feet of the ground surface.
As a standard practice, in order to separate regions of mineral fluoresence from those of
hydrocarbon contamination, the ROST™ LIF operators evaluate the wavelength-time matrices
(WTMs) collected at several locations during the push. In addition, the on-site geologist
examines discrete samples collected from several additional locations. The discrete samples may
be collected with the CPT stab sampler or, in the case of this demonstration, with the hollow
stem auger and split spoon sampler. The carbonates can be distinguished from nonfluorescent
soil by examining the soil sample (carbonate-based minerals appear as white crystalline material)
and confirmed by pouring hydrochloric acid on the soil and observing release of carbon dioxide.
For this demonstration, the ROST™ LIF operators, relying on the evaluation of WTMs and
fluorescence lifetimes, determined that two locations in SNLDR12 had a high fluorescent
response from a nonhydrocarbon source. In order to evaluate their procedures, the verification
entity evaluated the WTMs independently and reviewed field notes to see if there was any
44
-------
difference in results. After independent analysis, the verification entity agreed with the Loral
results. This improved the matching percentage to 96.5% and reduced the false positives to 1.1%
when compared to either TPH and TRPH. WTMs and FVDs for all pushes are presented in
Appendix B.
Downhole Results for SNL Tank Farm
Figure 6-4 shows the downhole fluorescence measurements for pushes 10, 11, and 12. Again,
the developers determined that the area near the surface in SNLDR12 exhibited nonhydrocarbon
fluorescence, and the on-site geologist confirmed this by examining the soil samples collected
with the HSA after the push was completed.
1 ^ HOLE 10
g ^40000 T ,
0 ™\ ^^faj&fyt
0 10 20 30 40
Depth (ft)
1
250% S
r*V> 1 «
50 60
I-SS HOLE 11
§ B
g ^40000 -j-
U 20000 i jn 0 0°(p od
_. _, Cr o -*. /u
0 10 20 30 40
Depth (ft)
§
250 g
50 60
|
.1 ^1000 -L JV
1 S o J^rio-Ho—o
d o 10
HOLE 12
20 30 40
Depth (ft)
)
(O
In
luorescence
•~Vr- (a,
50 60
Figure 6-4. Downhole results for SNL Tank Farm.
45
-------
SNL Tank Farm Subsurface Contaminant Mapping
The test area at the SNL Tank Farm and the three collocated ROST™ LIF pushes and hollow
stem auger borings are shown in Figure 6-5. Each symbol along a hole indicates a location
where a soil sample was collected. The result of each sample was compared to the LIF
measurement at the corresponding depth interval. The horizontal boundary of the plume cannot
be estimated from the information obtained from the three pushes and borings; however, the
areas of strong carbonate fluorescence and hydrocarbon contamination are evident based on the
LIF and laboratory results. This figure shows several points:
• the contaminant plume was thick and migrated downward rather than laterally;
• the false negatives were confined to areas where the strong carbonate fluorescence signal
masked the hydrocarbon fluorescence signal; and
• the point of refusal for the CPT pushes was the gravel/caliche layer at 50-57 feet bgs.
Push/Boring Location
-10
-20
-30
-40
-SO
•60
#10
#12
55.25',
- Hydrocarbon
Impacted
Area
Area of Excavation
and Backfill
(Homogenized
Impacted Soils)
- Carbonate
Rich
Region
LIF/Lab Result
— Hydrocarbon
Impacted
Area
O Nondetect/Nondetect
zi> Nondetect/Dotect
-A. Detect/Nondetect
• Detect/Detect
Plan View
49.5V
Figure 6-5. Subsurface contaminant map for SNL Tank Farm. Cross-section view of the
results given in Table 6-2, comparing the ATI TPH results to those of the ROST™ LIF.
46
-------
Cost Evaluation
Table 6-3 provides a comparison of deployment costs for 1) the ROST™ LIF system, 2)
conventional drilling and sampling with a hollow stem auger drilling rig outfitted with a split
spoon sampler, and 3) a Geoprobe® for a typical POL investigation.
Table 6-3. Cost comparison.
Scenario: Define the lateral and horizontal extent of free-phase volatile organic petroleum hydrocarbons
and residual hydrocarbons. Depth to groundwater is 30 feet bgs. Soils will be continuously sampled from
the surface to 35 feet bgs. The soil samples will be logged for soil classification and screened for
petroleum hydrocarbon contamination in the field. The two soil samples from each boring that exhibit the
greatest response to the field screening will be submitted for laboratory analysis.
Hollow Stem Auger
Assumptions:
10 borings to 35 feet bgs
Production rate with continuous sampling, logging, and grouting is 70 feet per day
Drilling
Consultant Geologist
Organic Vapor Meter
Truck Rental
Disposal of Cuttings
Analytical Testing
Total
5days@$1500/day
50 hrs @ $65/hr
5 days @ $75/day
5 days @ $50/day
8 drums @ $100/drum
20 samples @ $60/sample
$7500
3250
375
250
800
1200
$13,375
Geoprobe®
Assumptions:
10 borings to 35 feet bgs
Production rate with continuous sampling, logging, and grouting is 100 feet per day
Geoprobing
Consultant Geologist
Organic Vapor Meter
Truck Rental
Analytical Testing
Total
3.5 days @$1200/day
35 hrs @ $65/hr
4 days @ $75/day
4 days @ $50/day
20 samples @ $60/sample
$4200
2275
300
200
1200
$8,175
ROST™/CPT
Assumptions:
10 pushes to 35 feet bgs
Production rate is ten locations per day (10 hr day)
Includes basic data report
ROST™/CPT
Per Diem/3 crew members
Per Diem/Consultant Geologist
Sampling (CPT)
Analytical Testing (confirmatory
samples from impacted zone)
Total
1 day @ $5300/day
1 day @ $225/day
15 hrs @ $65/hr
0.5 day @ $2500/day
5 samples @ $60/sample
$5300
225
975
1250
300
$8,050
47
-------
Overall Performance Evaluation
In summary, the results of the demonstrations satisfy the requirements set forth in the demon-
stration plan and addendum for the ROST™ LIF system. The system located the plume
accurately with higher matching percentage than the developer claimed. Matching percentages
for the Port Hueneme demonstration were nearly 90 percent when compared to TPH and TRPH.
Matching percentages for the SNL demonstration were 93.4% when compared to either labora-
tory method. The false negative rate for the Port Hueneme demonstration was 10 percent when
compared to TRPH and 5.4 percent when compared to TPH. The false negative rate for the SNL
demonstration was 3.3 percent when compared to either TRPH or TPH. Disagreements with the
laboratory results were primarily confined to regions where contaminant concentration levels
were close to the detection threshold. At Port Hueneme, an unusually low ROST™ detection
threshold of 5 mg/kg may have contributed to the large percentage of false negatives. A portion
of the ROST™ false negatives could be the result of variability in laboratory results where
random errors are estimated to be in the range of 10 to 15 percent in general and are quite
possibly higher near the TPH and TRPH detection limits.
As stated earlier, the performance of the ROST™ LIF was evaluated against the developer claims
made in the demonstration plan. Evaluation of the developer claims for the LIF system is
presented in Table 6-4.
Table 6-4. ROST™ LIF claims evaluation.
Claim
Result
Evaluation
The ROST™ system can be
integrated with cone penetrometer
trucks from all major manufacturers.
Field integration is routinely
accomplished in a few hours.
ROST™ system was integrated easily with
the SCAPS CRT truck and has been
successfully deployed on other CRT
trucks. Field integration was accomplished
in 2 hrs at Port Hueneme.
Met
Standard data collection rate is one
sample every 1.2 sec, providing a
vertical spatial resolution of 0.2 ft for
a standard push rate of 1 m/min.
Data was collected every 0.2 ft or less if
cone slowed or stopped. Push rate is
dependent upon CPT.
Met
System can acquire
multidimensional data
representations, such as WTMs, to
1) identify fuel type or 2) eliminate
false positives from nonhydrocarbon
fluorophores.
1) Only one class of hydrocarbon was
available at each site; therefore, the first
claim was not evaluated.
2) WTMs and fluorescence lifetimes were
used to distinguish hydrocarbon and
nonhydrocarbon fluorophores.
1) not evaluated
2) met
Target of 80 percent agreement
with conventional laboratory
analysis of samples and 5 percent
false negatives.
Result for Port Hueneme:
87.7% correlation with TRPH and 10%
false negatives; 89.2% correlation with
TPH and 5.4% false negatives.
Result for SNL site:
93.4% correlation with either TRPH or
TPH; 3.3% false negatives.
Both claims met when
TPH data is used for
comparison.
Both claims met
48
-------
Section 7
Applications Assessment
The ROST™ LIF technology is emerging as a supplement to and possible replacement for
conventional drilling and sampling methods. As demonstrated, the ROST™ LIF technology has
advantages and limitations. These advantages and limitations are described in the following
sections.
Advantages of the Technology
Real-Time Analysis
Through the use of a cone penetrometer system, the ROST™ LIF provides real-time analysis of
site conditions. This approach is faster than any competitive technology, and therefore quite
useful for real-time decision making in the field. This is especially important in guiding soil
sampling activities. For conventional field characterization, soil samples are collected using a
standard drill rig and sent to a commercial laboratory for analysis. It can take weeks, and
sometimes months, to get results. When the results are reviewed, a return trip to the field for
further drilling and sampling may be indicated. Real-time sampling and data analysis often
eliminates the expense and time delays of laboratory analysis and return trips to the field.
Continuous LIF Data Output
The ROST™ LIF has an advantage over conventional drilling and sampling methods in its ability
to provide nearly continuous spatial data. It is common practice in environmental investigations
to select a sampling interval (e.g., every 5 feet) to collect samples for analysis at a commercial
laboratory. Characterization of the contaminant zone may be severely impaired when the data
density is sparse as it commonly is with conventional drilling and sampling approaches due to
budget constraints. Areas of contamination may go wholly unnoticed in extreme cases. ROST™
allows a continuous record of possible contaminant locations and a more complete delineation of
the area of contamination. In addition, some drilling and sampling operations can be hindered by
an inability to produce core samples, due to flowing sands or limited cohesiveness of the soils to
be sampled, whereas an in situ method such as ROST™ could potentially retrieve readings from
these horizons.
Continuous Lithological Logging
The cone penetrometer affords continuous logging of the subsurface lithology with on-board
geotechnical sensors used in conjunction with the LIF sensor. This allows a user to target
stratigraphy of interest, which may influence contaminant flow and transport or have potential
interfering influences on the LIF readings. A conventional drilling and sampling program would
require continuous core collection and a dedicated geologist to get the same level of detail. The
geologist may be able to define finer scale attributes of the media, but only through a much more
labor-intensive effort. Compared to the conventional approach of sampling at regular intervals
(e.g., every 5 feet), the CPT offers much greater resolution. Although the CPT was not the focus
49
-------
of this evaluation of the ROST™ technology, the features of the cone penetrometer enhance the
usefulness of any sensor used with the CPT.
Cost Advantages
The ROST™ provides nearly continuous data at a fraction of the cost of discrete sampling and
analysis of the same area. The cost effectiveness of the Department of Energy SCATS (without
LIF) compared to conventional drilling and sampling techniques has been evaluated indepen-
dently by Booth et al. (LANL, 1991). They concluded that the SCAPS technology has a 30 to 50
percent cost savings for various scenarios analyzed.
Enhanced Operator Safety
The ROST™ LIF used in conjunction with a standard CPT rig is safer than a conventional
drilling and sampling program. There is little chance of contacting contaminated soils, because
soil samples are only occasionally brought to the surface and the sensor is driven into the
subsurface to take measurements. CPT and ROST™ operators are located in the CPT truck, and
generally are not in contact with the subsurface soil. Grouting of the push hole can be done to
minimize any potential cross-contamination of geologic units in the subsurface. With drilling and
sampling methods, the soil cuttings are brought to the surface and potentially come in contact
with workers and also must be disposed of as investigation-derived waste. The samples are
handled by multiple individuals for packaging and transport, and for subsequent laboratory
analysis, again providing an opportunity for exposure. Decontamination of the sampling and
drilling equipment is most often done manually by drilling personnel rather than automated. The
ROST™ LIF and CPT offer a clear advantage over conventional drilling and sampling in the area
of health and safety of the crew.
Performance Advantages
The ROST™ LIF technology works well in both the unsaturated and saturated zone. This may be
important at sites with a relatively shallow water table or perched zone to delineate the continuity
of the contamination across the interface.
The developer's performance claims were generally met in these demonstrations. Table 7-1
summarizes the performance statistics for the technology relative to the ability of the LIF to
locate the presence of hydrocarbons. The developer claimed an overall detect/nondetect success
rate of 80 percent. The technology exceeded this claim in all instances. The developer also
claimed a false negative rate of no more than 5 percent. They met this claim with one exception,
when the data were compared to TRPH data for the Port Hueneme demonstration resulting in 10
percent false negatives. A probable cause is the differences in analytical methods and matrix
interferences.
50
-------
Table 7-1. Performance statistics.
Demonstration
Site
Port Hueneme
SNL
Percent Agreement
Claim (>80%)
87.7% (TRPH)
89.2% (TPH)
93.4% (TRPH and
TPH)
Percent False
Positive (implicit
claim <20%)
2.3% (TRPH)
5.4% (TPH)
3.3% (TRPH and
TPH)
Percent False
Negative (claim <5%)
10% (TRPH)
5.4% (TPH)
3.3% (TRPH and
TPH)
The ROST™ LIF system should meet the expectations of regulators or site owners interested in
compliance with EPA sampling guidance (U.S. EPA, 1989b). In designing sampling strategies,
the EPA has acknowledged the concepts of uncertainty and potential errors in analysis. They
have incorporated these expectations in their guidance on allowable false positive and negative
rates when comparing confirmatory sampling data to screening data. The EPA guidance on
statistical sampling typically accepts a 5 to 10 percent false negative rate, which is within the
range of the ROST™ LIF based on the results of these demonstrations. In addition, they allow a
higher percentage of false positives, typically up to 20 percent. The ROST™ LIF system appears
to be capable of meeting EPA's guidance of performance criteria for comparison of laboratory
versus screening data.
Limitations of the Technology
Applicability
The ROST™ LIF system is applicable only to fuels and wastes containing nonchlorinated multi-
ring aromatic hydrocarbon molecules. The detection capabilities for ROST™ include, but are
not limited to, jet fuel, gasoline, diesel, lubricating oils, coal tar, and creosote. Other common
compounds such as chlorinated hydrocarbons would require separate sensors.
ROST™ has been used to detect two-ring aromatic compounds (naphthalenes) on commercial
projects involving jet fuel. In addition, ROST™ can readily detect mixtures of fuels and other
materials; however, the technology may not distinguish them in the presence of the other. These
capabilities were not evaluated as part of the CSCT demonstrations.
Quantitation and Speciation
The ROST™ LIF does not allow for the direct quantitation of specific constituents in the
petroleum contaminant. The regulatory requirements for determining cleanup requirements for
RCRA or CERCLA sites are established on the basis of individual constituent concentrations
(e.g., naphthalene concentrations) through comparisons with background, or established through
the use of risk assessment techniques.
ROST™ has been calibrated to TPH in soil, which is appropriate for underground storage tank
investigations. For RCRA or CERCLA investigations, it is best used as screening measure to
pinpoint optimal locations for conventional sampling and analysis. The RCRA and CERCLA
requirements are based on constituent-specific concentration thresholds and not aggregate
measures of a total class of products such as TPH. TPH is affected by many interferants and is
not readily correlated to individual constituents. For leaking underground fuel tank applications,
51
-------
the guidance often refers to an action level of 100 mg/kg TPH for delineation of areas of
potential concern. The ROST™ detection limits are site dependent and may exceed the 100
mg/kg action level at a given site, as shown in the SNL Tank Farm demonstration.
Push Limitations of CPT
A cone penetrometer system is limited in its ability to hydraulically push through certain
stratigraphies (e.g., boulders, cobbles, caliche). The maximum depth is governed by site-specific
stratigraphy and the method is limited to sites where the cone penetrometer can be pushed to the
depth of concern through primarily unconsolidated sedimentary deposits or formations. This can
limit the applicability of the ROST™ LIF deployment to sites which have less severe geor
technical characteristics. It should also be noted that the sensor location for the LEF is some
distance above the cone tip (i.e., 36.5-60 cm, depending on the probe used), and when refusal
occurs due to a stratigraphy change, the sensor does not actually get to that depth horizon. This
can be problematic if the stratigraphic layer is also an impedance to flow and transport of the
contaminants, thereby offering an opportunity for the contaminant to become concentrated at the
interface boundary. In this case, the LIF sensor would not be able to address the issue unless the
constituent concentrations were elevated 60 cm above the interface or refusal depth.
Interferences
The LIF system is subject to interferences which can make data reduction complicated, and limit
the real-time nature of data analysis and decision making. Moisture in the soil, oxygen, and
fluorescing compounds or minerals (e.g., carbonates) are examples of naturally occurring
constituents which affect the LIF readings and influence performance statistics. In many cases,
if site-specific interferences are identified prior to or during the field investigation, the WTMs
can provide information to distinguish fluorescent artifacts from actual hydrocarbon
contaminants in the subsurface.
Conclusions
The ROST™ LIF system is an emerging technology worthy of pursuit in site investigations where
polycyclic aromatic hydrocarbons (e.g., petroleum, oils, and lubricants, coal tars, and creosote)
are suspected. The technology offers a number of advantages over conventional drilling and
sampling technologies for the purpose of screening a site for the nature and extent of con-
tamination. It does not entirely take the place of a conventional sampling program, but adds
significant benefits in terms of cost savings and more thorough characterization. This infor-
mation, when used properly, could provide a more complete picture of the contamination and can
be used to predict optimal sampling locations. As noted above, there are some limitations of
which a prospective user should be aware when designing an environmental investigation.
Stratigraphy and unidentifiable fluorescent interferences are issues that could prevent the sole
use of the ROST™ LIF system. The technology has been used to identify lighter fuels but this
capability was not evaluated in these demonstrations. Because the technology does not provide
species-specific quantitation, it should be used in conjunction with conventional sampling and
analysis if risk assessment or cleanup criteria must be met. As a'screening technology to identify
the nature and extent of polycyclic aromatic hydrocarbon contamination,.this technology has
many advantages over conventional techniques.
52
-------
Section 8
Developer Forum
The following information was provided by Fugro Geosciences.
Fugro Geosciences acquired the technology from Loral (now Lockheed Martin) in May 1996.
Since ROST™'s introduction in 1994, Fugro has worked closely with Loral, providing CPT
services on the majority of Loral's ROST™ projects. Fugro now provides ROST™ worldwide
directly to consultants and site owners as an integrated service with our extensive direct push
capabilities.
Overall, Fugro Geosciences is pleased with the design and conclusions of the EPA CSCT
evaluation of ROST™. However, some significant features of ROST™ were not fully evaluated
by CSCT, due to ROST™,s deployment from the Navy's SCAPS CPT truck, the presence of only
a single contaminant in test site soils, and the detect/nondetect evaluation criteria. Specifically,
the features not evaluated are the high mobility and productivity rate of Fugro's CPT/ROST™,
ROST™'s contaminant applicability and product identification capability, and ROST™,s
delineation capabilities. Each of these important features are detailed in the following sections.
High Mobility and Productivity Rate of Fugro CPT/ROST™
Deployment of ROST™ from Fugro Geosciences' truck or all-terrain vehicle-mounted CPT rigs
would have allowed demonstration of our high site mobility and productivity rate. Fugro's
production rate on ROST™/CPT projects typically exceeds 300 linear feet per day for pushes
averaging 30 feet or greater in depth. Typically, 10 to 12 ROST™/CPT pushes per day can be
completed for projects involving shallower push depths.
Contaminant Applicability and Product Identification Capability
ROST™'s application to a wide range of petroleum contaminants and the technology's product
differentiation capability make it a powerful site characterization tool. However, these
capabilities were not demonstrated, since diesel fuel was the only contaminant present at both
evaluation sites. ROST™ has been used successfully on commercial projects to delineate and
differentiate materials including jet fuel/kerosene, gasoline, diesel fuel, lubricating oil, crude oil,
bunker oil, coal tar, and creosote. The ability to differentiate between these materials in real-
time using ROST™'s WTM function allows multiple sources to be recognized and delineated.
rTM
ROST'""s Delineation Capabilities
The demonstration only evaluated the detect/nondetect agreement between ROST™ and the
reference method. However, ROST™ provides significantly more value than simply a
detect/nondetect field screening tool. Since fluorescence intensity is generally proportional to in
53
-------
situ concentration, ROST™ can effectively delineate not only the presence, but the relative
concentration of contaminants. Our commercial clients typically use this proportional feature of
ROST™ data to pinpoint the zones of highest contaminant concentration and screen the variation
in concentration as they map the three-dimensional extent across a site.
CPT/ROST™ Data Presentation
CPT/ROST™ data are typically presented in a basic data report containing integrated logs
illustrating fluorescence intensity versus depth, stratigraphy, and contaminant WTM fingerprints.
As an option, the data may also be delivered to clients on a floppy disk in spreadsheet format.
This method of data delivery provides significant benefit to consultants and site owners planning
to input the data into three-dimensional graphic or modeling programs. Important zones can be
readily selected for interpretation and graphic presentation with minimal effort.
POST™ Upgrades
Dakota Technologies, Inc. (DTI), co-developers of ROST™, provide research and development
and technical support to Fugro. DTI has developed a ROST™ upgrade that will allow
simultaneous monitoring of fluorescence versus depth at four separate wavelengths during a
push. The systems will be upgraded to the multi-wavelength function in the near future. This
feature will allow detection of a wider range of contaminants simultaneously and will provide
continuous product differentiation without the need to pause and collect WTMs. Fugro and DTI
will continue to evaluate and upgrade the ROST™ system to make it as robust as possible.
Fugro's Existing and Emerging Technologies
Specialized CPT sensors and sampling tools developed by Fugro for site characterization
include:
• Standard Cone Penetrometer - identifies stratigraphy
• Piezocone - identifies stratigraphy and measures saturated pore pressure. Allows
identification of water table and estimation of hydraulic conductivity and refined
interpretation of fine-grained soils
Conductivity Cone - identifies stratigraphy and soil/groundwater conductivity
Supercone - combined standard, piezo, and conductivity cone
Natural Gamma Probe
Seismic Cone
Depth Discrete Groundwater and Soil Samplers
CPT Installed piezometers from 1/2-inch to 2-inch diameter
Fugro is currently an active participant in the development of the next generation of laser-
induced fluorescence in situ technology under the Advanced Applied Technology Demonstration
Facility sponsored by the U.S. Department of Defense in partnership with Tufts University and
Rice University. Fugro is also pursuing development of new sensors including probes for in situ
metals and chlorinated hydrocarbon screening. Following development, we anticipate
participating in evaluation of each of these tools under EPA's CSCT verification program.
54
-------
Section 9
Previous Commercial Projects
The following information was provided by Fugro. The investigations included industrial plants, oil
production facilities, refineries, railyards, and military bases in both the United States and Europe.
Further information on these deployments may be obtained from Fugro Geosciences.
Table 9-1. Summary of Selected CPT/ROST™ Commercial Projects.
Facility Type
Refinery Landfarm
Industrial Plant
Industrial Plant
Oil Production Field
Oil Production Field
Natural Gas
Production Plant
Refinery
Petrochemical Plant
Manufactured Gas
Plant
Degasification Plant
Refinery Stormwater
Impoundment
Air Force Base
Refinery
Industrial Plant
Industrial Plant
Paint Manufacturing
Plant
Wood Preserving
Plant
Retail Service Station
Wood Preserving
Plant
Railroad Yard
Site Location
Texas City, TX
Everett, MA
Tennessee
Guadalupe, CA
Lost Hills, CA
Refugio, TX
Beaumont, TX
Seadrift, TX
England, Wales, and
Scotland
Paris, France
Beaumont, TX
San Bernadino, CA
Westville, NJ
Vernon, CA
Indianapolis, IN
Anaheim, CA
Green Spring, WV
Valencia, CA
Visalia, CA
California
Contaminant of
Concern
Petroleum
Hydrocarbons
Naphthalene
Petroleum
Hydrocarbons
Kerosene, Diesel,
Crude Oil
Compressor
Lubricants
Natural Gas
Condensate
Polyaromatic
Hydrocarbons
Petroleum
Hydrocarbons
Coal Tar
Petroleum
Hydrocarbons
Petroleum
Hydrocarbons
Gasoline, Diesel, Jet
Fuel
Petroleum
Hydrocarbons
Petroleum
Hydrocarbons
Diesel, Fuel Oil,
Lubricants, Naphtha,
Gasoline, Kerosene
Petroleum
Hydrocarbons
Creosote
Gasoline
Creosote
Diesel, Bunker Oil
CPT/ROST
Soundings
Completed
23
72
29
319
10
17
21
19
54
13
56
105
30
41
47
11
40
6
30
41
Total Linear
Footage of
Testing
485
640
1100
7,458
430
625
845
549
623
115
635
1,610
1,075
2,101
1,372
624
653
188
3433
858
55
-------
Facility Type
Wood Preserving
Plant
Railroad Yard
Department of Energy
Refinery
Refinery
Retail Service Station
Oil Production Field
Naval Station
Oil Production Field
Refinery
Pipeline
Railroad Yard
Oil Production Field
Refinery
Refinery
Wood Preserving
Plant
Refinery
Refinery
Army Base
Refinery
Refinery
Refinery Stormwater
Impoundment
Refinery
Air Force Base
Site Location
Seattle, WA
Arizona
Aiken, SC
Germany
Carson, CA
Escondido, CA
Casmalia, CA
China Lake, CA
Los Angeles, CA
Ponca City, OK
Albert Lea, MN
Los Angeles, CA
Guadalupe, CA
Houston, TX
Wilmington, CA
Houston, TX
Cincinnati, OH
Carson, CA
Rock Island, IL
Wales, UK
Shreveport, LA
Lockport, IL
Toledo, OH
Edwards AFB, CA
Contaminant of
Concern
Creosote
Kerosene, Bunker
Oil
Diesel Fuel
Petroleum
Hydrocarbons
Cat Cracker Feed,
Petroleum
Hydrocarbons
Gasoline
Kerosene, Diesel,
Crude Oil
Multiple Petroleum
Products
Crude Oil
Gasoline
Jet Fuel
Lubricating Oil,
Diesel
Kerosene, Diesel,
Crude Oil
Benzene
Petroleum
Hydrocarbons
Creosote
Gasoline
Petroleum
Hydrocarbons
Diesel
Petroleum
Hydrocarbons
Petroleum
Hydrocarbons
Petroleum
Hydrocarbons
Petroleum
Hydrocarbons
Gasoline, Diesel, Jet
Fuel
CPT/ROST
Soundings
Completed
18
10
23
151
8
15
68
33
256
31
28
37
42
17
7
46
18
15
37
41
28
27
66
31
Total Linear
Footage of
Testing
2,082
679
970
4,218
430
128
2,268
1,553
6,031
1,401
458
1,238
2,056
410
482
2,188
548
1,127
882
129
753
293
1,344
1,400
56
-------
Section 10
References
Analytical Technologies, Inc (ATI). 1995. ATI-SD SOP 605. Hydrocarbon Characterization/Fuel
Fingerprint Analysis by GC-FID Standard Operating Procedure. Revision 6.0. Revised March
1995.
ATI. 1994. ATI Quality Assurance Manual. Revision 5.0. Revised November 1994.
ATI. 1993a. ATI-SD SOP 421. Sub-Sampling and Compositing Soil Samples for VOA and Non-VOA
Analyses. Standard Operating Procedure. Revision 3. Revised June 1993.
ATI. 1993b. ATI-SD SOP 803. Determination of Total Recoverable Petroleum Hydrocarbons by
Infrared Spectrophotometry, Standard Operating Procedure. Revision 1. Revised September 1993.
CEEvflC. 1996. Letter to Stang, PRC Environmental Management, Inc., re: two samples lost at ATI-
SD. From Leslie Getman, CEIMIC (formerly ATI-SD). February 28, 1996.
Loral Corporation. 1995. ROST™ Data Report, Real-Time Continuous Measurement of Subsurface
Contamination with the Rapid Optical Screening Tool (ROST™). Loral Corporation. September
28, 1995.
Los Alamos National Laboratory (LANL), 1991. Schroeder, J.D., Booth, S.R., and Trocki, L.K., "Cost
Effectiveness of the Site Characterization and Analysis Penetrometer System." Los Alamos
National Laboratory Report LA-UR-91-4016, submitted to Department of Energy. December 1991.
PRC Environmental Management, Inc (PRC). 1994. "Draft Site Inspection Report for Installation
Restoration Program Site 22 - Aboveground Fuel Farm," NCBC Port Hueneme. July 1, 1994.
PRC. 1995a. Laser Induced Fluorometry/Cone Penetrometer Technology Demonstration Plan at the
Hydrocarbon National Test Site, Naval Construction Battalion Center Port Hueneme, California.
Prepared for U.S. Environmental Protection Agency, Office of Research and Development,
Environmental Systems Monitoring Laboratory-Las Vegas, Consortium for Site Characterization
Technology. May 1995.
PRC. 1995b. LIF/CPT Technology Predemonstration and Demonstration Summary Report, NCBC Port
Hueneme Fuel Farm-County of Ventura Boring Permit E-575, Contract N66001-94-D0141, Delivery
Order 0003. July 10, 1995.
PRC. 1995c. Addendum to the Laser Induced Fluorometry/Cone Penetrometer Technology
Demonstration Plan at the Arid Demonstration Site, Sandia National Laboratories Steam Plant Tank
Farm, Albuquerque, New Mexico. Prepared for U.S. Environmental Protection Agency, Office of
Research and Development, National Exposure Research Laboratory, Characterization Research
Division-Las Vegas, Consortium for Site Characterization Technology. October 1995.
PRC. 1995d. LIF/CPT Technology Predemonstration and Demonstration Summary Report, Arid
Demonstration Site, Sandia National Laboratories Above Ground Fuel Farm, Contract N66001-94-
D0141, Delivery Order 0006. December 20,1995.
57
-------
U.S. Environmental Protection Agency (EPA). 1995. Rapid Optical Screening Tool (ROST™),
Innovative Technology Evaluation Report. Superfund Innovative Technology Evaluation Program,
Office of Research and Development, Washington, DC. EPA/540/R095/519.
U.S. Environmental Protection Agency (EPA). 1995. Test Methods for Evaluating Solid Waste,
Physical/Chemical Methods (SW-846). Revision 3. January 1995.
U.S. Environmental Protection Agency (EPA). 1994. "Guidance Manual for the Preparation of Site
Characterization Technology Demonstration Plans - Protocol I." Version 3.0. December 25,1994.
U.S. Environmental Protection Agency (EPA). 1989a. "Preparing Perfect Project Plans." Risk Reduction
Engineering Laboratory, Office of Research and Development. Cincinnati, Ohio. EPA/600/9-
89/087. October 1989.
U.S. Environmental Protection Agency (EPA). 1989b. "Methods for the Attainment of Cleanup
Standards, Volume 1: Soils and Solid Media." Office of Policy, Planning, and Evaluation, EPA
220/02-89-042. October 1989.
58
-------
Appendix A
Reference Laboratory Data
A-l
-------
Table A-1
Reference Laboratory Results of Soil Samples
NCBC Port Hueneme
Sample
Number
PHDB21-1
PHDB21-2
PHDB21-3
PHDB21-4
PHDB21-5
PHDB21-6
PHDB21-7
PHDB21-8
PHDB21-9
PHDB21-10
PHDB21-11
PHDB21-12
PHDB21-13
PHDB21-14
PHDB21-15
PHDB22-1
PHDB22-2
PHDB22-3
PHDB224
PHDB22-5
PHDB22-6
PHDB22-7
PHDB22-8
PHDB22-9
PHDB22-10
PHDB22-11
PHDB22-12
PHDB22-13
PHDB22-14
PHDB22-15
PHDB22-16
PHDB22-17
PHDB23-1
PHDB23-2
PHDB23-3
PHDB234
PHDB23-5
PHDB23-6
PHDB23-7
PHDB23-8
PHDB23-9
PHDB23-10
PHDB23-11
PHDB23-12
PHDB23-13
Depth
2.5-3.0'
3.0-3.5'
4.5-5.0'
5.0-5.5'
6.5-7.0'
7.0-7.5'
8.5-9.0'
9.0-9.5'
10.5-11.0'
11.0-11.5'
12.5-13.0'
13.0-13.5'
15.5-16.0'
16.0-16.5'
18.5-19'
2.5-3.0'
3.0-3.5'
4.5-5.0'
5.0-5.5'
7.0-7.5'
7.5-8.0'
8.5-9.0'
9.0-9.5
10.5-11.0'
11.0-11.5'
12.5-13.0'
13.0-13.5'
14.5-15.0'
15.0-15.5'
16.5-17.0'
17.0-17.5'
19.0-19.5'
2.5-3.0'
3.0-3.5*
4.5-5.0'
6.5-7.0'
7.0-7.5'
8.5-9.0'
9.0-9.5'
10.5-11.0'
11.0-11.5'
12.5-13.0'
13.0-13.5'
14.0-14.5'
14.5-15.0'
Date
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/17/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
TRPH Concentration
mg/kg
<1
4
<1
<1
4
<1
<1
2
2
2
21 900 (Dup 22500)
18500 (Dup 17400)
28
18
11
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
4
<1
<1
<1
<1
<1
6
<1
<1
14
<1
<1
<1
<1
53
<1
16200 (Dup 18300)
24200 (Dup 26500)
6460 (Dup 61 60)
22
TPH
mg/kg
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
18000 (Dup 18000)
15000 (Dup 4000)
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
9.6
<5.0
16000 (Dup 16000)
19000 (Dup 23000)
7000 (Dup 5800)
29
LIF/TRPH
Result
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
D/ND
DID
DID
DID
DID
ND/D
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/D
ND/ND
ND/ND
ND/D
ND/ND
ND/ND
ND/ND
ND/ND
ND/D
ND/ND
DID"
DID*
DID*
DID*
LIF/TPH
Result
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
D/ND
DID
DID
D/ND
D/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND .
ND/ND
ND/D
ND/ND
DID*
DID*
DID*
DID*
A-2
-------
Table A-1 (continued)
Reference Laboratory Results of Soil Samples
NCBC Port Hueneme
Sample
Number
PHDB23-14
PHDB23-15
PHDB23-16
PHDB24-1
PHDB24-2
PHDB244
PHDB24-5
PHDB24-7
PHDB24-8
PHDB24-9
PHDB24-10
PHDB24-12
PHDB24-13
PHDB24-14
PHDB24-15
PHDB24-17
PHDB24-18
PHDB24-19
PHDB24-20
PHDB24-21
PHDB25-1
PHDB25-2
PHDB25-3
PHDB25-4
PHDB25-5
PHDB25-6
PHDB25-7
PHDB25-8
PHDB25-9
PHDB25-10
PHDB25-11
PHDB25-12
PHDB25-13
PHDB25-14
PHDB25-15
PHDB25-16
PHDB26-1
PHDB26-2
PHDB26-3
PHDB26-4
PHDB26-5
PHDB26-6
PHDB26-7
PHDB26-8
PHDB26-9
PHDB26-10
Depth
17.0-17.5'
17.5-18.0'
18.5-19.0'
2.5-3.0'
3.0-3.5'
4.5-5.0'
5.0-5.5'
6.5-7.01
7.0-7.5'
8.5-9.0'
9.0-9.5'
1 0.5-1 1.01
11.0-11.5'
12.5-13.0'
13.0-13.5'
14.5-15.0'
15.0-15.5'
16.0-16.5'
16.5-17.0'
19.0-19.5'
3.0-3.5'
3.5-4.01
4.5-5.0'
5.0-5.5'
6.5-7.01
7.0-7.51
8.5-9.01
9.0-9.5'
11.0-11.5'
12.5-13.0'
13.0-13.5'
14.5-15.0'
15.0-15.5'
17.0-17.5'
17.5-18.0'
19.5-20.0'
2.5-3.0'
4.5-5.0'
5.0-5.5'
6.5-7.0'
7.0-7.5'
8.5-9.0'
9.0-9.5'
1 0.5-1 1.01
11.0-11.5'
12.5-13.0'
Date
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/18/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
TRPH Concentration
mg/kg
224
2
5
81
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
17
<1
<1
11
<1
<1
<1
<1
<1
<1
<1
<1
<1
25
748
5620
9340 (Dup 13600)
172(Dup264)
28
1
9
31
<1
<1
<1
<1
<1
<1
<1
<1
<1
TPH
mg/kg
89
<5.0
<5.0
77
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
51
1100
6400
16000 (Dup 15000)
150 (Dup 190)
16
11
<5.0
11
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
LIF/TRPH
Result
ND/D
ND/ND
ND/ND
ND/D
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/D
ND/ND
D/ND
DID
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/D
DID
DID
DID
DID
DID
ND/ND
MD/D
ND/D
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
LIF/TPH
Result
ND/D
ND/ND
ND/ND
ND/D
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
D/ND
D/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/D
DID
DID
DID
DID
DID
ND/D
ND/ND
ND/D
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
A-3
-------
Table A-1 (continued)
Reference Laboratory Results of Soil Samples
NCBC Port Hueneme
Sample Number
PHDB26-11
PHDB26-12
PHDB26-13
PHDB26-14
PHDB26-15
PHDB26-16
PHDB26-17
PHDB27-1
PHDB27-2
PHDB27-3
PHDB27-4
PHDB27-5
PHDB27-6
PHDB27-7
PHDB27-8
PHDB27-10
PHDB27-11
PHDB27-12
PHDB27-13
PHDB27-15
PHDB27-16
PHDB27-17
PHDB27-19
PHDB28-1
PHDB28-2
PHDB28-3
PHDB28-4
PHDB28-5
PHDB28-6
PHDB28-7
PHDB28-8
PHDB28-9
PHDB28-11
PHDB28-12
PHDB28-13
PHDB28-14
PHDB28-15
PHDB28-16
PHDB28-17
Depth
13.0-13.5'
15.0-15.5'
15.5-16.0'
17.0-17.5'
17.5-18.0'
19.0-19.5'
19.5-20.0'
2.5-3.0'
3.0-3.5'
4.5-5.0'
5.0-5.5'
6.5-7.0'
7.0-7.5'
8.5-9.0'
9.0-9.5'
10.5-11.0'
11.0-11.5'
12.5-13.0'
13.0-13.5'
14.5-15.0'
15.0-15.5'
16.5-17.0'
19.0-19.5'
2.5-3.0'
4.5-5.0'
5.0-5.5'
6.5-7.0'
7.0-7.5'
8.5-9.0'
9.0-9.5'
10.5-1 1.01
11.0-11.5'
12.5-13.0'
13.0-13.5'
14.5-15.0'
15.0-15.5'
15.5-16.0'
17.5-18.0'
18.0-18.5'
Date
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/19/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
5/22/95
TRPH Concentration
mg/kg
36
8600
3540
229
145
<1
<1
2
2
<1
<1
4
9
<1
1
<1
<1
<1
<1
<1
3
1
2
34
2
2
3
4
3
3
5
<1
2
<1
1100{Dup800)
1800 (Dup 2100)
100
13
9
TPH Concentration
mg/kg
41
7900
2800
250
170
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
9.8
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
780 (Dup 920)
2900 (Dup 3400)
250
<5.0
<5.0
LIF/TRPH
Result
ND/D
DID
DID
DID
DID
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/D
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
D/ND
ND/ND
DID
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
D/D*
D/D*
D/D*
D/D
ND/D
LIF/TPH
Result
ND/D
DID
DID
DID
DID
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
D/ND
ND/ND
D/D
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
D/D*
D/D*
D/D*
D/ND
ND/ND
Notes:
1. TRPH indicates total recoverable petroleum hydrocarbons, analyzed by EPA method 418.1. 2. TPH indicates total petroleum
hydrocarbons, analyzed by the Calif. DHS method 8015-modified. 3. mg/kg = milligrams per kilogram. 4. Dup indicates duplicate
analysis performed by separate analysis of split sample following homogenization. 5. Accuracy in depth is estimated to be within 3
Inches in the vadose zone, and 6 inches in the saturated zone. 6. * indicates samples for which single point test measurement
results were used to determine depth discrepancy between discrete soil samples and in-situ measurements. Depth of discrete
samples was adjusted 4 inches to correlate with in-situ LIF measurements.
A-4
-------
Table A-2
Reference Laboratory Results of Soil Samples
SNL Tank Farm
Sample
Number
SNLDB10-1
SNLDB10-2
SNLDB10-3
SNLDB10-4
SNLDB10-5
SNLDB10-6
SNLDB10-7
SNLDB10-8
SNLDB10-9
SNLDB10-10
SNLDB10-11
SNLDB10-12
SNLDB10-13
SNLDB10-14
SNLDB10-15
SNLDB10-16
SNLDB10-17
SNLDB10-18
SNLDB10-19
SNLDB10-20
SNLDB10-21
SNLDB10-22
SNLDB10-23
SNLDB10-24
SNLDB10-25
SNLDB10-26
SNLDB10-27
SNLDB10-28
SNLDB10-29
SNLDB10-30
SNLDB10-31
SNLDB10-32
SNLDB10-33
SNLDB10-34
SNLDB10-35
SNLDB10-36
SNLDB10-37
Depth
2.75-3.0'
3.25-3.5'
4.75-5.01
5.25-5.5'
6.75-7.0'
7.25-7.5'
8.75-9.0'
9.25-9.5'
10.75-11.0'
11.25-11.5'
12.75-13.0'
13.25-13.5'
14.75-15.0'
15.25-15.5'
1 6.75-1 7.01
17.25-17.5'
18.75-19.0'
19.25-19.5'
20.75-21.0'
21.25-21.5'
22.75-23.0'
23.25-23.5'
24.75-25.01
25.25-25.5'
26.75-27.0'
28.75-29.0'
29.25-29.5'
30.75-31.0'
31.25-31.5'
32.75-33.0'
33.25-33.5'
34.75-35.0'
35.25-35.5'
36.75-37.0'
37.25-37.5'
38.75-39.0'
39.25-39.5'
Date
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
TRPH Concentration
mg/kg
60
25
89
42
T
162
17
11
27
22
206
1470
487C
7600
14300
850C
25600
25800
14700
5790
6530
8560
5100
5400
11200
20400
24900
7330
3520
1340
28400
18200
9620
26200
32200
21700
TPH Concentration
mg/kg
25
<5
99
&
7(
150
<5
1^
2i
27
270
1500
500C
660C
21000
13000
26000
28000
14000
6300
6900
9100
4200
4500
9800
20000
23000
6600
3100
1400
35000
24000
18000
10000
21000
28000
21000
LIF/TRPH
Result
ND/ND
ND/ND
ND/ND
ND/ND
D/ND
ND/D
ND/ND
ND/ND
MD/ND
ND/ND
DID
ND/D
D/D
D/D
D/D
DID
D/D
D/D
D/D
D/D
DID
DID
DID
)/D
D/D
)/D
)/D
)/D
)/D
)/D
D/D
)/D
)/D
D/D
D/D
)/D
D/D
LIF/TPH
Result
ND/ND
ND/ND
ND/ND
ND/ND
D/ND
ND/D
ND/ND
ND/ND
ND/ND
ND/ND
DID
ND/D
D/D
D/D
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
D/D
D/D
D/D
D/D
DID
DID
DID
A-5
-------
Table A-2 (continued)
Reference Laboratory Results of Soil Samples
SNL Tank Farm
Sample
Number
SNLDB10-38
SNLDB10-39
SNLDB10-40
SNLDB1041
SNLDB10-42
SNLDB1CM3
SNLDBKM4
SNLDB10-45
SNLDB1046
SNLDB1047
SNLDB1048
SNLDB1049
SNLDB10-50
SNLDB10-51
SNLDB11-1
SNLDB11-2
SNLDB11-3
SNLDB11-4
SNLDB11-5
SNLDB11-6
SNIDB11-7
SNLDB11-8
SNLDB11-9
SNLDB11-10
SNLDB11-11
SNLDB11-12
SNLDB11-13
SNLDB11-14
SNLDB11-15
SNLDB11-16
SNLDB11-17
SNLDB11-18
SNLDB11-19
SNLDB11-20
SNLDB11-21
SNLDB11-22
SNLDB11-23
SNLDB11-24
SNLDB11-25
SNLDB11-26
SNLDB11-27
SNLDB11-28
Depth
40.75-41.0'
41.2541.5'
42.7543.0'
43.2543.5'
44.7545.0'
45.2545.5'
46.7547.0'
47.2547.51
48.7549.0'
49.2549.5'
50.75-51.0
51.25-51.5'
52.75-53.0'
53.25-53.5'
6.0-6.25'
10.75-11.0'
11.25-11.5'
16.0-16.25'
20.75-21.0'
21.25-21.5'
25.75-26.0'
26.25-26.5'
30.75-31.0'
33.25-33.5'
35.75-36.0'
36.25-36.5'
40.7541.0'
41.2541.5'
42.7543.0'
43.2543.5'
44.7545.0'
45.2545.5'
46.7547.0'
47.2547.5'
48.2548.5'
48.7549.0'
49.2549.5'
50.75-51.0'
51.25-51.5'
52.75-53.0'
53.25-53.5'
55.0-55.25'
Date
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/6/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
11/7/95
33548
33548
11/7/95
11/7/95
11/7/95
TRPH Concentration
mg/kg
15800
8440
9,500 (Dup 9,1 60)
15000
7500
11000
13000
19000
26000
8200
13000
15000
17000
5500
9.7
9
<1
3470
13000
15200
12000
22300
18200
31000
19800
22200
26200
5160
20600
18300
7030
6240
11900
25400
17200
44600
7340
14700
23600
16100
13600
21400
TPH Concentration
mg/kg
14000
9700
12,000 (Dup 12,000)
18000
12000
9900
15000
23000
32000
14000
14000
27000
12000
8500
19
<5
<5
2700
11000
21000
10000
21000
17000
21000
19000
21000
24000
4200
22000
22000
14000
10000
13000
29000
29000
39000
8900
14000
25000
16000
13000
20000
LIF/TRPH
Result
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
ND/ND
ND/ND
ND/ND
ND/D
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
LIF/TPH
Result
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
ND/ND
ND/ND
ND/ND
ND/D
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
DID
A-6
-------
Table A-2 (continued)
Reference Laboratory Results of Soil Samples
SNL Tank Farm
Sample
Number
SNLDB12-1
SNLDB12-2
SNLDB12-3
SNLDB12-5
SNLDB12-7
SNLDB12-9
SNLDB12-11
SNLDB12-12
SNLDB12-13
SNLDB12-15
SNLDB12-17
SNLDB12-19
SNLDB12-20
Depth
2.75-3.0'
3.25-3.5'
6.0-6.25
11.25-11.5'
16.0-16.25'
21.0-21.25'
26.25-26.5'
26.75-27.0'
31.0-31.25'
36.5-36.75'
41.0-41.25'
43.0-43.25'
49.0-49.5'
Date
11/8/95
11/8/95
11/8/95
11/8/95
11/8/95
11/8/95
11/8/95
11/8/95
11/8/95
11/8/95
11/8/95
11/8/95
11/8/95
TRPH Concentration
mg/kg
3
2
<1
<1
<1
2
2
<1
<1
<1
<1
<1
<1
TPH Concentration
mg/kg
<5
<5
<5
<5
<5
<5
<5
<5
<5
<5
<5
<5
<5
LIF/TRPH
Result
D/ND
D/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
LIF/TPH
Result
D/ND**
D/ND**
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
ND/ND
2. TPH indicates total petroleum hydrocarbons, analyzed by the Calif. DHS Method 8015-modified.
3. mg/kg = milligrams per kilogram.
4. ** indicates where WTM review indicated a nonhydrocarbon fluorophore.
A-7
-------
-------
Appendix B
ROST™ LIF Field Data Logs
B-l
-------
PHDR21
DRAFT
0 50 100 150 200
Intensity (% of Reference)
WPHDR21C
120
100
80
60
40
20
300 340 380 420 460 500
Wavelength (nm)
WPHDR21A
120
100
'. 80
60
40
20
300 340 380 420 460 500
Wavelength (nm)
WPHDR21B
1ZU
80
i5?&-»-! ', ;
K5>^=* ! ! •
: : :
300 340 380 420 460
Wavelength (nm)
500
-------
PHDR22
-0.5 0 0.5
Intensity (% of Reference)
-------
PHDR23
20
DRAFT
0 50 100 150
Intensity (% of Reference)
WPHDR23A
120
100
? 80
| 60
t- 40
20
300 340 380 420 460 500
Wavelength (nm)
WPHDR23B
120
100
1 80
| 60
t- 40
20
300 340 380 420 460 SOO
Wavelength (nm)
-------
PHDR24
DRAFT
WPHDR24A
120
** flO
o en
E 60
1- 40
"0
;
—
-
300 340 380 420 460 SOO
Wavelength (nm)
-5 0 5 10 15 20
Intensity (% of Reference)
-------
PHDR25
DRAFT
1
Wphdr25A
120
100
JT 80
I 60
i= 40
20
300 340 380 420 460 500
Wavelength (nm)
Wphdr25B
120
100
80
60
40
20
300 340 380 420 460 500
Wavelength (nm)
Wphdr25C
300 340 380 420 460 500
Wavelength (nm)
0 50 100 150
Intensity (% of Reference)
-------
PHDR26
DRAFT
0 50 100 150
Intensity (% of Reference)
WPHDR26A
300 340 380 420 460 500
Wavelength (nm)
WPHDR26B
300 340 380 420 460 SOO
Wavelength (nm)
-------
PHDR27A
13
14
15
16
17
18
19
20
21
DRAFT
WHDR27AA
120
fe , —
i?F^
i
i
^"*^*—
-Ci
300 340 380 420 460 500
Wavelength (nm)
WHDR27AB
300 340 380 420 460 500
Wavelength (nm)
-505
Intensity (% of Reference)
-------
PHDR28
-------
o
o
•J
I
H
H
GO
§
Q
B
O
•n
I
o
•n
I
.is
8
>* -S
fc a
S
-S
B
o
>n o
cs en
Q W O, H BG <
s
cs
-------
O
o
•J
I
fc
CJ
§
88
Q
. £2
88
o o
88
Q a H~SM£
I
§8
II
-------
88
8
Q
-------
o
3
1
fc
U
H
O)
§
88
O
9
I
i — • • - : • .... ....
i ; . L .. ... . i... .1 . .
i ! '
_._!_.. ., ,. ' ..... .._ . . 1 . .j.
i • '. i
i • ••
i '• !
1 ..'...].. i i .... .L.,., .__
r 1 : ;
: I i \. i -
[ ; — 1 1 — r !
S3 j ' ! i : !
_/ ^..r-Jv-M-., \JJ--P ^4— .,,lll '•„,„* , , ,1- — •—*"-*—%. .u.L_,L
f 1 1
; i
! i-
i i
; ;
•
h. i.. 1
f I
V% ! '
* 1 i. .
r CQ i
. — -*/vJ~— .—. '
c<
.18
- -O
Q w eu f- EC
-------
-------
Appendix C
ROST™ LIF Draft EMMC Method
c-i
-------
DRAFT METHOD
FIELD SCREENING OF SUBSURFACE PETROLEUM HYDROCARBONS
WITH THE RAPID OPTICAL SCREENING TOOL (ROST™)
1.0 SCOPE AND APPLICATION
1.1 This field screening method provides rapid determination of the location and distribution of subsurface
petroleum hydrocarbon contamination. The method can be used to detect contaminants in the vadose, capillary
fringe, and saturated zones to depths greater than 50 meters below ground surface. The measurements are
performed in situ and physical sampling is not required. The list of the petroleum products for which this method
is appropriate includes, but is not limited to:
mineral oil
kerosene
fuel oil
gasoline
diesel fuel
lubricating oil
tar
asphaltum
hydraulic oil
jet fuel
aviation fuel
petroleum distillates
1.2 The method detection limit depends on several factors including soil matrix properties, excitation source
wavelength, length of fiber optic probe, optical collection efficiency, and petroleum product type. The detection
limits can be as low as a few parts-per-million (ppm) and the method can be applied up to contamination levels of
10 percent or greater.
1.3 The method yields qualitative (type) and quantitative (amount) information on subsurface petroleum, oil,
and lubricant (POL) contamination, making it appropriate for preliminary assessments of contaminant
distribution as in environmental field screening applications.
2.0 SUMMARY OF METHOD
2.1 This method provides an overview and guidelines for the use of Laser Induced Fluorescence/Cone
Penetrometer Testing (LIF/CPT) with the ROST system to obtain in situ measurements of hydrocarbon
contamination in soil. Procedures for calibration and data analysis are also provided.
2.2 The ROST instrument senses POLs via the fluorescence response to ultraviolet wavelength laser
excitation of their aromatic hydrocarbon constituents. The fluorescence measurements are carried out remotely
over fiber optic cables. Excitation light is delivered by an optical fiber to a sapphire window located in a sub-
assembly near the penetrometer tip. One or more collection fibers transmit the return fluorescence signal back to
the surface for analysis.
2.3 The ROST is deployed on a standard cone penetrometer truck, which provides a mobile platform for
moving from one push location to another. The ROST system has been integrated with cone penetrometer trucks
from all major manufacturers. Fluorescence measurements can be obtained at depths as great as 50 meters below
ground surface when the sensor is used in conjunction with a standard 20-ton penetrometer vehicle.
2.4 Geotechnical sensors are normally integrated with the LIF sensor probe to facilitate geotechnical and
statigraphic analyses of the soil matrix.
C-2
-------
3.0 Definitions
3.1 LIF: laser induced fluorescence.
3.2 Penetrometer: an instrument in the form of a conically-tipped cylindrical rod that is hydraulically
advanced into soil to acquire subsurface measurements of penetration resistance. Used for cone penetrometer
testing (CPT). Also called cone penetrometer, friction-cone penetrometer.
3.3 POL: petroleum, oil, lubricant. Used in reference to any petroleum product or derivative.
3.4 Push rods: cylindrical rods with threaded tips that are joined to advance the penetrometer probe into the
ground.
3.5 UV: ultraviolet
3.6 PMT: photomultipliertube
3.7 DSO: digital storfage oscilloscope
4.0 BNTTERFERENCES
4.1 In addition to the aromatic hydrocarbon constituents of the specifically targeted petroleum hydrocarbons,
other substances may fluoresce when excited by the laser light source and interfere with the POL determination.
Possible interfering species include fluorescent minerals, naturally occurring organic material, de-icing agents,
antifreeze additives, and detergent products.
4.2 The possibility of fluorescence emission from nontarget (non-POL) analytes, leading to false positive
assignment of POL contamination, must be considered. The fluorescence of the POL species of interest can be
distinguished from non-POL fluorescence on the basis of spectral and temporal (fluorescence decay) information
acquired at selected (or all) depths during the push. Past experience indicates that POL species have
characteristic fluorescence patterns (wavelength-time matrices) that allow them to be identified and distinguished
from potential interferents.
4.3 There are several background sources caused by the laser light separate from the petroleum or soil matrix
fluorescence. Their signal amplitudes occur on the same time scale as the petroleum fluorescence and can
therefore contribute to the total intensity. The possibilities include window fluorescence, cladding/buffer
fluorescence, Raman signals generated within fiber, stray light hi monochromator. These can be distinguished
from the true fluorescence signals by appropriate control experiments.
5.0
SAFETY
5.1 The ROST LIF sensor involves high-power pulsed laser beams that represent a potential eye hazard. Eye
protection precaution similar to those which apply to the use of pulsed lasers hi laboratory situations must be
observed.
5.2 Components of the ROST system are at sufficiently high voltage to present a shock hazard. However,
these components are not accessible during normal operation.
C-3
-------
6.0 EQUIPMENT AND SUPPLIES
6.1 The main ROST components are the Nd:YAG pump laser, tunable dye laser, fiber optic cable,
monochromator, photomultiplier tube (PMT), digital storage oscilloscope (DSO), control/analysis computer and
software. The system components in the current version of ROST fit into two half-height instrumentation racks,
each of which is approximately 26" high x 20" wide x 24" deep. The pulsed laser excitation source operates
either at a selected wavelength in the range 280-300 run or at 266 nm. The fluorescence emission which is
transmitted back to the surface is spectrally analyzed with a monochromator. The intensity of the emission signal
passed by the monochromator is quantitatively measured by a photomultiplier tube detector and digital
oscilloscope. ROST records voltage-time waveforms created by pulsed fluorescence light striking a
photomultipher tube detector. Typically, waveforms are averaged over 50 laser shots, which requires one second
at the standard 50 pulses/second repetition rate. Approximately 0.2 seconds is required to transfer the waveform
to the host computer in preparation for acquisition of the next waveform.
6.2 The industry standard CPT systems employ a hydraulic ram mounted to a truck chassis so that a series of
attached threaded rods can be advanced into the ground through an opening in the floor of the vehicle.
6.3 For ROST measurements, a sub-assembly is positioned between the standard penetrometer cone and the
first push rod. A sapphire window view port is mounted on the side of the sub-assembly. An optical module that
holds the ends of the optical fibers and other optics is inserted firmly within the sub-assembly. The electrical
cables for the geotechnical sensors pass around the optical module to the cone penetrometer. Field integration is
routinely accomplished in a few hours.
6.4 The ROST system can be integrated with and deployed on commercially available CPT vehicles. ROST
has a depth encoder system independent of the CPT depth measurement device so the fluorescence data can be
acquired independently of the CPT measurements. One person operates the LIF sensor, taking measurements of
the calibration and control standards, and monitoring the actual real-time push data.
7.0 REAGENTS AND STANDARDS
7.1 Reagent-grade chemicals shall be used in all tests. Unless otherwise indicated, it is intended that all
reagents shall conform to the specifications of the Committee on Analytical Reagents of the American Chemical
Society, where such specifications are available. Other grades may be used, provided it is first ascertained that
the reagent is of sufficient high purity to permit its use without lessening the accuracy of the determination.
7.2 A check standard is used to verify satisfactory instrument performance on a continuing basis. The check
standard should fluoresce in the same wavelength range as the target species. Other desirable attributes of the
check standard are that it possess a high quantum efficiency, be chemically stable, easily prepared, and exhibit
minimal photdegmdation. The appropriate concentration of the check standard will depend on system sensitivity.
7.3 A method blank may be prepared from a sample of clean dry soil. Fine to medium-grain sea sand is
appropriate.
8.0 PROCEDURE
8.1 Before the LIF/CPT system is deployed, the site is visited to determine location of obstructions that
would limit access by the CPT truck. These obstructions may include buildings, cement platforms, and fence
lines. The site is also surveyed for possible underground obstructions such as utilities, pipelines, and existing
storage tanks. At this time, information on possible contaminants and prior efforts at characterization or
remediation is also obtained. Soil samples can be collected for preparation of site-specific calibration standards.
C-4
-------
8.2 The CPT truck is positioned over the push location and then elevated and leveled on hydraulic jacks.
Following a short series of measurements to establish ROST quality control, the sensor is pushed into the ground
at a rate of 1 meter/minute. The push rods are 1 meter in length, and rods are added approximately once a minute
as the sensor is advanced. A 30-meter push will typically require about 40 minutes to reach full depth.
Approximately 300 linear feet of push data can be accomplished in a routine day's operation.
8.3 The ROST sensor measures fluorescence signal as a function of depth as the penetrometer is pushed into
the ground, thereby providing a fluorescence vs. depth (FVD) log. The ordinate of the FVD is presented relative
to the check reference intensity.
8.4 As the next push rod is being added, a wavelength-time matrix can be acquired for contaminant
identification purposes. Alternatively, the ROST operator can signal to the hydraulics operator to temporarily
interrupt the push for WTM measurement.
9.0 QUALITY CONTROL AND SYSTEM CHECKOUT
9.1 The fluorescence intensity value is typically reported relative to the fluorescence intensity of a reference
solution, which is measured just prior to the initiation of each push. The M-l reference solution, a selected fluid
hydrocarbon mixture, is contained in a standard 1-cm pathlength cuvette, which can be strapped onto the sapphire
window. The procedure provides an end-to-end system check and normalizes the data for any variation in the
power of the laser light used to excite the contaminant, length of cable carrying the excitation and emission light,
background noise, and other instrument settings such as monochromator slitwidth.
9.2 If the reference check intensity varies by more than 25 percent from the average of the previous values,
the probe window and sample cuvette should be cleaned and the measurement repeated. If compliance cannot be
achieved, the system operator should begin troubleshooting procedures as per the system's maintenance manual.
9.3 The time window (typically 250 ns wide) on the digital oscilloscope is adjusted to compensate for the
light transit tune through the optical fiber; for a 50 meter long fluorescence signal is received at the detector
about 500 ns after the laser has actually fired. Once set, the time delay need by adjusted only if the length of
fiber in the probe umbilical is changed. The position of the ROST time window can be determined automatically
with routines built into the scope's software.
9.4 A wavelength calibration for the emission monochromator is performed at the start of the job and
thereafter during troubleshooting procedures. The 532 nm Nd:YAG 2nd harmonic light is used as a primary
reference to verify the wavelength accuracy of the monochromator. A small amount of 532 nm light is directed
into the monochromator at a narrow slitwidth and the wavelength is scanned to verify that the signal maximizes
at 532 ± 0.2 nm. The monochromator can then be used as a secondary reference to calibrate the dye laser
wavelength.
10.0
CALIBRATION OF CONTAMINANT CONCENTRATION
10.1 At present, there is no standard procedure for calibrating the LIF sensor. Depending on data objectives,
fluorescence intensity alone may be reported as a relative indicator of POL presence. The reference fluorescence
intensity data format is well-suited for field screening applications, in which the goal is to delineate contaminant
plume boundaries and to define the relative distribution of contamination over the site. The fluorescence
intensity is proportional to POL concentration over a wide range of concentration. The reliability of LIF-CPT for
screening sites in this fashion, i.e., without any formal calibration procedure, has been demonstrated on many
occasions.
C-5
-------
10.2 When called for, a calibration curve can be generated to establish the LIF sensor response, dynamic
range, and limit of detection. Depending on the objectives of the investigation, the following options should be
considered:
10.2.1 A common POL contaminant (gasoline, diesel fuel, coal tar, etc.) and soil type (sand, silt, clay) that is
though most representative of the site conditions is designated. Then the relative fluorescence intensities can be
converted to concentration units with response tables determined in laboratory studies. These tables exist
currently only for common fuels on sand.
10.2.2 A contaminant from the site is spiked onto a specified reference soil type and analyzed by ROST™. A set
of standards is prepared by inoculating the soil samples with a series of increasing amounts of the target analyte.
The spiked samples are tumbled for 24-48 hours to ensure uniform distribution of the fuel.
10.2.3 The contaminant from the site is spiked onto clean soil samples from the site. The soil is gathered from
below the surface at a depth of 1-2 feet, to reduce hydrocarbon contamination from aerosols and other airborne
particulates. This option is the most specific of the synthetic calibration standard approaches, but still assumes
that the soil and product used in the calibration is representative of the site.
10.3 The calibration standards can be obtained directly from the ground by soil borings, which are submitted
for analysis for approved laboratory methods. The influence of confounding variables such as weathering, soil
moisture, soil matrix, and other changes, are eliminated. The disadvantage of the in situ calibration standard is
the difficulty in obtaining a sample for the conventional analysis from actually the same spot as surveyed by
ROST. There are two options: 1) use the ROST data as measured during the active pushes; 2) place the sample
material on the window and rerun.
11.0 SAMPLE COLLECTION
This is an in situ method. Spectroscopic measurements are obtained directly without physical sampling. Sample
collection is not a part of the normal method procedure. The vertical spatial resolution is less than 4 cm when the
penetrometer is driven at the standard 1 m/min push rate.
12.0
DATA ANALYSIS AND CALCULATIONS
12.1 The amplitude of the logs is the area under the voltage vs. time waveforms, which is proportional to the
total light received (within the wavelength interval set by the monochromator) per laser pulse. The standard
display format is to plot the area under the fluorescence intensity vs. time waveform as a function of depth. This
is referred to as a Fluorescence vs. Depth (FVD) log or plot.
12.2 The raw voltage-time waveforms (voltage proportional to fluorescence light intensity) are subjected to
various data processing and analysis procedures. The first type of manipulation is to remove any DC offset from
current leakage through the amplifiers of the scope input stages, from ambient light (not induced by the laser) that
reaches the detector, or dark current from the photomultiplier tube. The basis for removing the DC offset is that
true light-induced signals, cannot occur in advance of the laser pulse itself. The DC offset which is automatically
subtracted from the averaged waveforms before they are transferred to the system computer.
12.3 The light-induced background signals are eliminated by an analysis of the baseline. One expects the true
background (including noise components) to be normally distributed, i.e., to follow a Gaussian distribution. The
C-6
-------
center of the Gaussian falls at the true background amplitude and the width corresponds to the noise level
(uncertainty). We generate a histogram of the intensities measured during the course of a push.
13.0 METHOD PERFORMANCE
13.1 The detection limit, accuracy, and precision obtained through use of the method are dependent on the soil
matrix, target analyte, and choice of laser wavelength, as well as instrumental conditions such as fiber length and
monochromator slitwidth. They must be established on a case-by-case basis.
14.0 REFERENCES
References are to be provided by Loral.
C-7
-------
------- |