&EPA
         United States
         Environmental Protection
         Agency
             Office of Research and
             Development
             Washington, DC 20460
EPA/600/R-01/089
September 2001
Innovative Technology
Verification Report
         Field Measurement
         Technologies for Total
         Petroleum Hydrocarbons in Soil

         Horiba Instruments Incorporated
         OCMA-350 Oil Content Analyzer

-------
                                   EPA/600/R-01/089
                                   September 2001
  Innovative Technology
     Verification  Report
 Horiba Instruments Incorporated
OCMA-350 Oil Content Analyzer
                Prepared by

               Tetra Tech EM Inc.
          200 East Randolph Drive, Suite 4700
              Chicago, Illinois 60601

             Contract No. 68-C5-0037
               Dr. Stephen Billets
          Characterization and Monitoring Branch
           Environmental Sciences Division
           Las Vegas, Nevada 89193-3478
          National Exposure Research Laboratory
          Office of Research and Development
          U.S. Environmental Protection Agency

-------
                                       Notice
This document was prepared for the U.S. Environmental Protection Agency (EPA) Superfund
Innovative Technology Evaluation Program under Contract No. 68-C5-0037. The document has
been subjected to the EPA's peer and administrative reviews and has been approved for publication.
Mention of corporation names, trade names, or commercial products does not constitute endorsement
or recommendation for use.

-------
                      UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                                      Office of Research and Development
                                           Washington, DC 20460
  V-     cT"
              ENVIRONMENTAL TECHNOLOGY VERIFICATION PROGRAM
                                VERIFICATION STATEMENT
 TECHNOLOGY TYPE:    FIELD MEASUREMENT DEVICE

 APPLICATION:           MEASUREMENT OF TOTAL PETROLEUM HYDROCARBONS

 TECHNOLOGY NAME:   OCMA-350 OIL CONTENT ANALYZER

 COMPANY:               HORIBA INSTRUMENTS INCORPORATED
 ADDRESS:                17671 ARMSTRONG AVENUE
                            IRVINE, CA 92614

 WEB SITE:                http://www.horiba.com

 TELEPHONE:             (800) 4HORIBA
VERIFICATION PROGRAM DESCRIPTION

The U.S. Environmental Protection Agency (EPA) created the Superfund Innovative Technology Evaluation (SITE) and
Environmental  Technology  Verification (ETV) Programs to facilitate deployment of innovative technologies through
performance verification and information dissemination.  The goal of these programs is to further environmental protection
by substantially accelerating the acceptance and use of improved and cost-effective technologies. These programs assist and
inform those involved in design, distribution, permitting, and purchase of environmental technologies. This document
summarizes results  of a demonstration of the OCMA-350 Oil Content Analyzer (OCMA-350) developed by Horiba
Instruments Incorporated (Horiba).

PROGRAM OPERATION

Under the SITE and ETV Programs, with the full participation of the technology developers, the EPA evaluates and
documents the performance of innovative technologies by developing demonstration plans, conducting field tests, collecting
and analyzing demonstration data, and preparing reports. The technologies are evaluated under rigorous quality assurance
(QA) protocols to produce well-documented data of known quality. The EPA National Exposure Research Laboratory, which
demonstrates field sampling, monitoring, and measurement technologies, selected Tetra  Tech EM Inc. as the verification
organization to assist in field testing seven field measurement devices for total petroleum hydrocarbons (TPH) in soil. This
demonstration was funded by the SITE Program.

DEMONSTRATION DESCRIPTION

In June 2000, the EPA conducted a field demonstration of the OCMA-3 50 and six other field measurement devices for TPH
in soil. This verification statement focuses on the OCMA-350; a similar statement has been prepared for each of the other
six devices. The performance and cost of the OCMA-350 were compared to those of an off-site laboratory reference method,
"Test Methods for Evaluating Solid Waste" (SW-846) Method 8015B (modified). To verify a wide range of performance
attributes, the demonstration had both primary and secondary objectives. The primary objectives included (1) determining
the method detection limit, (2) evaluating the accuracy and precision  of TPH measurement, (3) evaluating the effect of
interferents, and (4) evaluating the effect of moisture content on TPH measurement for each device. Additional primary
objectives were to measure sample throughput and estimate TPH measurement costs.   Secondary objectives included
(1) documenting the skills and training required to properly operate the device, (2) documenting the portability of the device,
(3) evaluating the device's durability, and (4) documenting the availability of the device and associated spare parts.

The OCMA-3 50 was demonstrated by using it to analyze 74 soil environmental samples, 89 soil performance evaluation (PE)
samples, and 36 liquid PE samples. In addition to these 199 samples, 9 extract duplicates prepared using the environmental
samples were analyzed.  The environmental samples were collected  in five areas contaminated with gasoline, diesel,
lubricating oil, or other petroleum products, and the PE samples were obtained from a commercial provider.
                           The accompanying notice is an integral part of this verification statement.                  September 2001
                                                   iii

-------
Collectively, the environmental and PE samples provided the different matrix types and the different levels and types of
petroleum hydrocarbon contamination needed to perform a comprehensive evaluation  of the OCMA-350. A complete
description of the demonstration and a summary of its results are available in the "Innovative Technology Verification Report:
Field Measurement Devices for Total Petroleum Hydrocarbons in Soil—Horiba Instruments Incorporated OCMA-350 Oil
Content Analyzer" (EPA/600/R-01/089).

TECHNOLOGY DESCRIPTION

The OCMA-350  developed by Horiba is based on  infrared analysis.  The OCMA-350 includes a single-beam, fixed-
wavelength, nondispersive infrared filter-based spectrophotometer. Infrared radiation from a tungsten lamp is transmitted
through a cylindrical, quartz cuvette containing a sample extract. The radiation that has passed through the extract enters
a detector containing a filter that isolates analytical wavelengths in the 3,400- to 3,500-nanometer range.

During the demonstration, Horiba dried soil by adding anhydrous sodium sulfate. Extraction of petroleum hydrocarbons in
a given soil sample was typically performed by adding 20 milliliters of Horiba's proprietary S-316 extraction solvent to
5 grams of the sample.  The mixture was agitated using an ultrasonic mixer. The sample extract was decanted into a beaker
through a filter-lined funnel, and then the filtrate was poured into a quartz cuvette.  The cuvette was placed in the
spectrophotometer, and the TPH concentration in milligrams per kilogram was read on the digital display.  Periodically,
Horiba recycled the extraction solvent using its Model SR-300 solvent reclaimer.

VERIFICATION OF PERFORMANCE

To ensure data usability, data quality indicators for accuracy, precision, representativeness, completeness, and comparability
were assessed for the reference method based on project-specific QA objectives.  Although the reference method results
generally exhibited a negative bias, based on the results for the data quality indicators, the reference method results were
considered to be of adequate quality.  The  bias was considered to be  significant primarily  for low- and medium-
concentration-range soil samples containing diesel, which made up only 13 percent of the total number of samples analyzed
during the demonstration. The reference method recoveries observed during the demonstration were typical of the recoveries
obtained by most organic analytical methods for environmental samples. In general, the user should exercise caution when
evaluating the accuracy of a field measurement device by comparing it to reference methods because the reference methods
themselves may have limitations. Key demonstration findings are summarized below for the primary objectives.

Method Detection Limit: Based on the TPH results for seven low-concentration-range diesel soil PE samples, the method
detection limits were determined to be 15.2 and 4.79 milligrams per kilogram for the OCMA-350 and reference method,
respectively.

Accuracy and Precision: Seventy-eight of 107 OCMA-350 results (73 percent) used to draw conclusions regarding whether
the TPH concentration in a given sampling area or sample type exceeded a specified action level agreed with those of the
reference method; 6 OCMA-350 conclusions were false positives, and 23 were false negatives.

Of 102 OCMA-350 results used to assess measurement bias, 36 were within 30 percent, 16 were within 30 to 50 percent, and
50 were not within 50 percent of the reference method results; 64 OCMA-350 results were biased low, and 38 were biased
high.

For soil environmental samples, the OCMA-350 results were statistically (1) the same as the reference method results for four
of the five sampling areas and (2) different from the reference method results for one sampling area. For soil PE samples,
the OCMA-350 results were statistically (1) the same as the reference method results for blank samples and (2) different from
the reference method results for medium- and high-concentration-range weathered gasoline samples and low-, medium-, and
high-concentration-range diesel samples. For liquid PE samples, the OCMA-350 results were statistically (1) the same as
the reference method results for weathered gasoline samples and (2) different from the reference method results for diesel
samples.

The OCMA-350 results correlated highly with the reference method results for three of the five sampling areas (the square
of the  correlation coefficient [R2] values were greater than or equal to 0.90, and F-test probability values were less than
5 percent).  The OCMA-350 results correlated moderately with the reference method results for diesel soil PE samples (the
R2 value was 0.68, and the F-test probability value was less than 5 percent). The OCMA-350 results correlated weakly with
the reference method results for two sampling areas and for weathered gasoline soil PE samples (R2 values ranged from 0.19
to 0.53, and F-test probability values ranged from 22.58 to 62.68 percent).

Comparison of the OCMA-3 50 and reference method median relative standard deviations (RSD) showed that the OCMA-3 50
and the reference  method exhibited similar overall precision. Specifically, the median RSD ranges were 1.5 to 20 percent
and 5.5 to 18 percent for the OCMA-350 and reference method, respectively. The analytical precision for the OCMA-350
(a median relative percent difference of 1) was better than that for the reference method (a median relative percent difference
of 4).
                             The accompanying notice is an integral part of this verification statement.                   September 2001
                                                       iv

-------
Effect of Interferents: The OCMA-350 showed no response to neat tetrachloroethene (PCE); neat 1,2,4-trichlorobenzene;
or soil spiked with humic acid. The device's mean responses for neat methyl-tert-butyl ether (MTBE),  Stoddard solvent,
and turpentine were 72.5,86, and 85 percent, respectively. The reference method showed varying mean responses for MTBE
(39 percent); PCE (17.5 percent); Stoddard solvent (85 percent); turpentine (52 percent); 1,2,4-trichlorobenzene (50 percent);
and humic acid (0 percent).  For the demonstration, MTBE and Stoddard solvent were included in the definition of TPH.

Effect of Moisture Content: The OCMA-350 showed a three-fold increase in TPH results when the soil moisture content
was increased from 9 to 16 percent for weathered gasoline soil  PE samples. However, the device showed a three-fold
decrease in TPH results when the soil moisture content was increased from less than 1 percent to 9 percent for diesel soil PE
samples. The reference method TPH results were unaffected when the soil moisture content was increased.

Measurement Time: From the time of sample receipt, Horiba required 46 hours, 15 minutes, to prepare a draft data package
containing TPH results for 199 samples and 9 extract duplicates compared to 30 days for the reference method, which was
used to analyze 4 additional extract duplicates.

Measurement Costs: The TPH measurement cost for  199 samples and  9 extract duplicates was estimated to be $15,750,
including the capital equipment purchase cost of $6,500, for the OCMA-350 compared to $42,050 for the reference method.

Key demonstration findings are summarized below for the secondary objectives.

Skill and Training Requirements: The OCMA-350 can be operated by one person with basic wet chemistry skills.  The
sample analysis procedure for the device can be learned in the field with a few practice attempts.  During the demonstration,
some of the items in the OCMA-350, including the narrow-mouth vials used for weighing and extracting soil samples and
the gravity filtration setup, made the TPH measurement procedure less simple and more time-consuming.

Portability: The OCMA-350 can be easily moved between sampling areas in the field, if necessary.  It can be operated using
a 110-volt alternating current power source or a direct current power source such as a 12-volt power outlet in an automobile.

Durability and Availability of the Device: During a 1 -year warranty period, Horiba will repair any malfunctioning items of
the device at no cost. During the demonstration, none of the device's reusable items malfunctioned or was damaged. Horiba
does not supply some items necessary for TPH measurement using the device, including a test tube rack, a funnel rack and
stand, a digital balance, and miscellaneous glassware and laboratory supplies; the availability of replacement or spare parts
not supplied by Horiba depends on their manufacturer or distributor.

In summary, during the demonstration, the OCMA-350 exhibited the following desirable characteristics of a field TPH
measurement device: (1) good precision, (2) sensitivity to interferents that are petroleum hydrocarbons, and (3) high sample
throughput. In addition, the OCMA-350 exhibited moderate measurement costs. In general, however, the OCMA-350 TPH
results for the PE samples did not compare well with the reference method results.  The device results were significantly
impacted by soil moisture content and by turpentine, an interferent that  is not a petroleum hydrocarbon. Collectively, the
demonstration findings indicated that the OCMA-350 may be considered for TPH screening purposes; however, the user
should exercise caution when considering the device for a field TPH measurement application requiring definitive results.

Original
signed by

Gary  J. Foley, Ph.D.
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
 appropriate quality assurance procedures. The EPA makes no expressed or implied warranties as to the performance of the technology
 and does not certify that a technology will always operate as verified. The end user is solely responsible for complying with any and
 all applicable federal, state, and local requirements.
                             The accompanying notice is an integral part of this verification statement.                   September 2001

                                                       V

-------
                                      Foreword
The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the
nation's natural resources. Under the 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, the EPA's Office of
Research and  Development provides  data and scientific  support that can be  used to  solve
environmental problems, build the scientific knowledge base needed to manage ecological resources
wisely, understand how pollutants affect public health, and prevent or reduce environmental risks.

The National Exposure Research Laboratory (NERL) is the agency's center for investigation of
technical and management approaches for identifying and quantifying risks to human health and the
environment. Goals of the laboratory's research program are to (1) develop and evaluate methods
and technologies for characterizing and monitoring air, soil, and water; (2) support regulatory and
policy decisions; and (3) provide the scientific support needed to ensure effective implementation
of environmental regulations and strategies.

The EPA's Superfund Innovative Technology Evaluation (SITE) Program evaluates technologies
designed for characterization and remediation of contaminated Superfund and Resource Conservation
and Recovery Act sites. The SITE Program was created to provide reliable cost and performance
data in order to speed acceptance and use of innovative remediation, characterization, and monitoring
technologies by the regulatory and user community.

Effective  measurement and  monitoring  technologies  are  needed  to  assess the  degree of
contamination at a site, provide data that can be used  to determine the risk to public health or the
environment, supply the  necessary cost and performance  data to select the most appropriate
technology, and monitor the success or failure of a remediation process. One component of the EPA
SITE  Program, the Monitoring and Measurement Technology (MMT) Program, demonstrates and
evaluates innovative technologies to meet these needs.

Candidate technologies can originate within the federal government or the private sector. Through
the SITE Program, developers are given the opportunity to conduct a rigorous demonstration of their
technologies under actual field conditions.  By completing the demonstration and distributing the
results, the agency establishes a baseline for acceptance and use of these technologies.  The MMT
Program is administered by the Environmental Sciences Division of NERL in Las Vegas, Nevada.
                                            Gary J. Foley, Ph.D.
                                            Director
                                            National Exposure Research Laboratory
                                            Office of Research and Development
                                           VI

-------
                                      Abstract
The OCMA-350 Oil Content Analyzer (OCMA-3 50) developed by Horiba Instruments Incorporated
(Horiba), was demonstrated under the U.S. Environmental Protection Agency Superfund Innovative
Technology Evaluation Program in June 2000 at the Navy Base Ventura County site in Port
Hueneme, California. The purpose of the demonstration was to collect reliable performance and cost
data for the OCMA-350 and six other field measurement devices for total petroleum hydrocarbons
(TPH) in soil.  In addition to  assessing ease  of device operation, the key objectives of the
demonstration included determining the (1) method detection limit, (2) accuracy and precision, (3)
effects of interferents and soil moisture content on TPH measurement, (4) sample throughput, and
(5)  TPH measurement costs for each device.  The demonstration involved  analysis  of both
performance evaluation samples and environmental samples collected in  five areas contaminated
with gasoline, diesel, lubricating oil, or other petroleum products. The performance and cost results
for a given  field measurement device were compared to those for an  off-site laboratory reference
method, "Test Methods for Evaluating Solid Waste" (SW-846) Method 8015B (modified). During
the demonstration, Horiba required 46 hours, 15 minutes, for TPH measurement of 199 samples and
9 extract duplicates. The TPH measurement costs for these samples were  estimated to be $15,750
for the OCMA-350 compared to $42,050 for the reference method.  The  method detection limits
were determined to be 15.2 and 4.79  milligrams per kilogram  for the OCMA-350 and reference
method, respectively.  During the demonstration, the OCMA-350 exhibited good precision and
sensitivity to interferents that are petroleum hydrocarbons (methyl-tert-butyl ether and Stoddard
solvent). However, the OCMA-350 TPH results (1) did not compare well with the reference method
results for the performance evaluation samples and (2) were significantly impacted by soil moisture
content and by turpentine, an interferent that is not a petroleum hydrocarbon. In addition,  some of
the items in the OCMA-350 made the TPH measurement procedure less simple and more time-
consuming  during the demonstration.  Collectively, the demonstration findings indicated that the
OCMA-350 may be considered  for TPH screening purposes; however, the user should exercise
caution when considering the device for a field TPH measurement application requiring definitive
results.
                                          vn

-------
                                      Contents


Chapter                                                                          Page

Notice	ii

Verification Statement	iii

Foreword 	  vi

Abstract 	vii

Figures	  xi

Tables	xii

Abbreviations, Acronyms, and Symbols  	  xiv

Acknowledgments  	  xvi

1       Introduction	1
        1.1     Description of SITE Program  	1
        1.2     Scope of Demonstration	4
        1.3     Components and Definition of TPH  	4
               1.3.1   Composition of Petroleum and Its Products	4
                       .3.1.1  Gasoline	6
                       .3.1.2  Naphthas  	6
                       .3.1.3  Kerosene  	6
                       .3.1.4  Jet Fuels	6
                       .3.1.5  Fuel Oils	7
                       .3.1.6  Diesel	7
                       .3.1.7  Lubricating Oils	7
               1.3.2   Measurement of TPH 	7
                       .3.2.1  Historical Perspective  	7
                       .3.2.2  Current Options for TPH Measurement in Soil	8
                       .3.2.3  Definition of TPH  	9

2       Description of Infrared Analysis and the OCMA-350  	11
        2.1     Description of Infrared Analysis 	11
        2.2     Description of OCMA-350  	13
               2.2.1   Device Description	13
               2.2.2   Operating Procedure	15
                                          Vlll

-------
                              Contents (Continued)



Chapter                                                                         Page

       2.3    Developer Contact Information	15

3      Demonstration Site Descriptions 	16
       3.1    Navy Base Ventura County Site  	17
              3.1.1   Fuel Farm Area	17
              3.1.2   Naval Exchange Service Station Area 	18
              3.1.3   Phytoremediation Area	18
       3.2    Kelly Air Force Base Site 	19
       3.3    Petroleum Company Site	19

4      Demonstration Approach  	21
       4.1    Demonstration Objectives 	21
       4.2    Demonstration Design	21
              4.2.1   Approach for Addressing Primary Objectives  	22
              4.2.2   Approach for Addressing Secondary Objectives  	26
       4.3    Sample Preparation and Management	30
              4.3.1   Sample Preparation	30
              4.3.2   Sample Management	32

5      Confirmatory Process  	36
       5.1    Reference Method Selection  	36
       5.2    Reference Laboratory Selection	38
       5.3    Summary of Reference Method	38

6      Assessment of Reference Method Data Quality	47
       6.1    Quality Control Check Results  	47
              6.1.1   GRO Analysis	47
              6.1.2   EDRO Analysis	50
       6.2    Selected Performance Evaluation Sample Results	56
       6.3    Data Quality	59

7      Performance of the OCMA-350 	60
       7.1    Primary Objectives	60
              7.1.1   Primary Objective PI: Method Detection Limit	62
              7.1.2   Primary Objective P2: Accuracy and Precision  	62
                     7.1.2.1  Accuracy 	63
                     7.1.2.2  Precision	72
              7.1.3   Primary Objective P3: Effect of Interferents  	77
                     7.1.3.1  Interferent Sample Results	77
                     7.1.3.2  Effects of Interferents on TPH Results for Soil Samples	77
              7.1.4   Primary Objective P4: Effect of Soil Moisture Content	85
              7.1.5   Primary Objective P5: Time Required for TPH Measurement	87
       7.2    Secondary Objectives	89
              7.2.1   Skill and Training Requirements for Proper Device Operation 	89
              7.2.2   Health and Safety Concerns Associated with Device Operation 	90

-------
                              Contents (Continued)
Chapter
                                                                  Page
              7.2.3   Portability of the Device	90
              7.2.4   Durability of the Device 	90
              7.2.5   Availability of the Device and Spare Parts	91

       Economic Analysis	92
       8.1     Issues and Assumptions	92
                 . 1   Capital Equipment Cost 	92
                 .2   Cost of Supplies  	92
                 .3   Support Equipment Cost	93
                 .4   Labor Cost	93
                 .5   Investigation-Derived Waste Disposal Cost	93
                 .6   Costs Not Included	93
       8.2     OCMA-350 Costs 	94
              8.2.1   Capital Equipment Cost 	94
              8.2.2   Cost of Supplies  	94
              8.2.3   Support Equipment Cost	94
              8.2.4   Labor Cost	94
              8.2.5   Investigation-Derived Waste Disposal Cost	96
              8.2.6   Summary of OCMA-350 Costs  	96
       8.3     Reference Method Costs  	96
       8.4     Comparison of Economic Analysis Results	96

9      Summary of Demonstration Results	98

10     References	103
Appendix
Supplemental Information Provided by the Developer	105

-------
                                       Figures
                                                                                  Page
       Distribution of various petroleum hydrocarbon types throughout boiling point
       range of crude oil  	5
5-1.    Reference method selection process	37
7-1.    Summary of statistical analysis of TPH results	61
7-2.    Measurement bias for environmental samples  	65
7-3.    Measurement bias for soil performance evaluation samples	66
7-4.    Linear regression plots for environmental samples	71
7-5.    Linear regression plots for soil performance evaluation samples	72
                                           XI

-------
                                      Tables
Table                                                                           Page
1-1.    Summary of Calibration Information for Infrared Analytical Method	8
1-2.    Current Technologies for TPH Measurement	9
2-1.    Infrared Absorbance Values of CH, CH2, and CH3 Groups  	12
2-2.    OCMA-350 Components and Supplies	14
3-1.    Summary of Site Characteristics	17
4-1.    Action Levels Used to Evaluate Analytical Accuracy 	23
4-2.    Demonstration Approach 	27
4-3.    Environmental Samples 	31
4-4.    Performance Evaluation Samples	33
4-5.    Sample Container, Preservation, and Holding Time Requirements 	35
5-1.    Laboratory Sample Preparation and Analytical Methods	38
5-2.    Summary of Project-Specific Procedures for GRO Analysis	40
5-3.    Summary of Project-Specific Procedures for EDRO Analysis	44
6-1.    Summary of Quality Control Check Results for GRO Analysis	51
6-2.    Summary of Quality Control Check Results for EDRO Analysis  	55
6-3.    Comparison of Soil and Liquid Performance Evaluation Sample Results 	57
6-4.    Comparison of Environmental Resource Associates Historical Results to
       Reference Method Results  	58
7-1.    TPH Results for Low-Concentration-Range Diesel Soil Performance Evaluation
       Samples	62
7-2.    Action Level Conclusions	64
7-3.    Statistical Comparison of OCMA-350 and Reference Method TPH Results for
       Environmental Samples 	68
7-4.     Statistical Comparison of OCMA-350 and Reference Method TPH Results for
       Performance Evaluation Samples	70
7-5.    Summary of Linear Regression Analysis Results	73
7-6.    Summary of OCMA-350 and Reference Method Precision for Field Triplicates
       of Environmental Samples  	74
                                         xn

-------
                              Tables (Continued)
Table                                                                          Page
7-7.    Summary of OCMA-350 and Reference Method Precision for Extract Duplicates .... 75
7-8.    Comparison of OCMA-350 and Reference Method Precision for Replicate
       Performance Evaluation Samples	76
7-9.    Comparison of OCMA-350 and Reference Method Results for Interferent
       Samples	78
7-10.   Comparison of OCMA-350 and Reference Method Results for Soil Performance
       Evaluation Samples Containing Interferents	80
7-11.   Comparison of Results for Soil Performance Evaluation Samples at Different
       Moisture Levels  	86
7-12.   Time Required to Complete TPH Measurement Activities Using the OCMA-350 .... 88
8-1.    OCMA-350 Cost Summary  	95
8-2.    Reference Method Cost Summary 	97
9-1.    Summary of OCMA-350 Results for the Primary Objectives 	99
9-2.    Summary of OCMA-350 Results for the Secondary Objectives  	102
                                        Xlll

-------
                   Abbreviations, Acronyms, and Symbols
±
(im
AC
AEHS
AFB
API
ASTM
bgs
BTEX
BVC
CCV
CFC
CFR
cm
DC
DER
DRO
EDRO
EPA
EPH
ERA
FFA
FID
GC
GRO
Horiba
HPLC
ICV
IDW
ITVR
kg
L
LCS
LCSD
MCAWW
MDL
Greater than
Less than or equal to
Plus or minus
Microgram
Microliter
Micrometer
Alternating current
Association for Environmental Health and Sciences
Air Force Base
American Petroleum Institute
American Society for Testing and Materials
Below ground surface
Benzene, toluene, ethylbenzene, and xylene
Base Ventura County
Continuing calibration verification
Chlorofluorocarbon
Code of Federal Regulations
Centimeter
Direct current
Data evaluation report
Diesel range organics
Extended diesel range organics
U.S. Environmental Protection Agency
Extractable petroleum hydrocarbon
Environmental Resource Associates
Fuel Farm Area
Flame ionization detector
Gas chromatograph
Gasoline range organics
Horiba Instruments Incorporated
High-performance liquid chromatography
Initial calibration verification
Investigation-derived waste
Innovative technology verification report
Kilogram
Liter
Laboratory control sample
Laboratory control sample duplicate
"Methods for Chemical Analysis of Water and Wastes'
Method detection limit
                                        xiv

-------
           Abbreviations, Acronyms, and Symbols (Continued)
Means
mg
min
mL
mm
MMT
MS
MSD
MTBE
n-Cx
NDIR
NERL
NEX
ng
nm
OCMA-350
ORD
ORO
OSWER
PC
PCB
PCE
PE
PHC
PPE
PRA
PRO
QA
QC
R2
RPD
RSD
SFT
SITE
STL Tampa East
SW-846
TPH
UST
VGA
VPH
R.S. Means Company
Milligram
Minute
Milliliter
Millimeter
Monitoring and Measurement Technology
Matrix spike
Matrix spike duplicate
Methyl-tert-butyl ether
Alkane with "x" carbon atoms
Nondispersive infrared
National Exposure Research Laboratory
Naval Exchange
Nanogram
Nanometer
OCMA-350 Oil Content Analyzer
Office of Research and Development
Oil range organics
Office of Solid Waste and Emergency Response
Petroleum company
Polychlorinated biphenyl
Tetrachloroethene
Performance evaluation
Petroleum hydrocarbon
Personal protective equipment
Phytoremediation Area
Petroleum range organics
Quality assurance
Quality control
Square of the correlation coefficient
Relative percent difference
Relative standard deviation
Slop Fill Tank
Superfund Innovative Technology Evaluation
Severn Trent Laboratories in Tampa, Florida
"Test Methods for Evaluating Solid Waste"
Total petroleum hydrocarbons
Underground storage tank
Volatile organic analysis
Volatile petroleum hydrocarbon
                                        xv

-------
                                Acknowledgments
This report was prepared for the U.S. Environmental Protection Agency (EPA) Superfund Innovative
Technology Evaluation Program under the direction and coordination of Dr. Stephen Billets of the
EPA National Exposure Research Laboratory (NERL)—Environmental Sciences Division in Las
Vegas, Nevada.  The EPA NERL thanks Mr. Ernest Lory of Navy Base Ventura County, Ms. Amy
Whitley of Kelly Air Force Base, and Mr. Jay Simonds of Handex of Indiana for their support in
conducting field activities for the project. Mr. Eric Koglin of the EPA NERL served as the technical
reviewer of this  report.  Mr.  Roger Claff of the American Petroleum Institute, Mr. Dominick
De Angelis of ExxonMobil Corporation, Dr. Ileana Rhodes of Equilon Enterprises, and Dr. Al
Verstuyft of Chevron Research and Technology Company served as the peer reviewers of this report.

This report was prepared for the EPA by Dr. Kirankumar Topudurti and Ms. Suzette Tay of Tetra
Tech EM Inc. Special acknowledgment is given to Mr. Jerry Parr of Catalyst Information Resources,
L.L.C., for serving as the technical consultant for the project.  Additional acknowledgment and
thanks are given to Ms.  Jeanne Kowalski, Mr. Jon Mann, Mr. Stanley Labunski, and Mr. Joe
Abboreno of Tetra Tech EM Inc. for their assistance during the preparation of this report.
                                         xvi

-------
                                              Chapter 1
                                             Introduction
The U.S. Environmental Protection Agency (EPA) Office
of Research and Development (ORD) National Exposure
Research Laboratory (NERL) conducted a demonstration
of seven innovative field measurement devices for total
petroleum hydrocarbons (TPH) in soil. The demonstration
was conducted as part of the EPA Superfund Innovative
Technology  Evaluation  (SITE)  Monitoring  and
Measurement Technology (MMT) Program using TPH-
contaminated soil from five areas located in three regions
of the United States. The demonstration was conducted at
Port Hueneme, California,  during the week of June  12,
2000.  The purpose of the  demonstration was to obtain
reliable performance and cost data on field measurement
devices in order to provide (1) potential users with a better
understanding of the devices'  performance and operating
costs  under well-defined field conditions and (2)  the
developers with documented re suits that will assist them in
promoting acceptance and use of their devices. The TPH
results obtained using the seven field measurement devices
were  compared  to the  TPH results  obtained  from a
reference laboratory chosen for the demonstration, which
used a reference method modified for the demonstration.

This innovative  technology verification  report (ITVR)
presents demonstration performance results and associated
costs   for  the  OCMA-350  Oil  Content  Analyzer
(OCMA-350)  developed  by   Horiba  Instruments
Incorporated (Horiba). Specifically, this report describes
the SITE Program, the scope of the demonstration, and the
components  and definition of TPH  (Chapter 1);  the
innovative field measurement device and the technology
upon   which  it  is  based  (Chapter  2);  the  three
demonstration sites  (Chapter  3); the  demonstration
approach (Chapter 4);  the selection of the reference
method and laboratory  (Chapter  5); the  assessment of
reference  method  data  quality  (Chapter  6);  the
performance of the field measurement device (Chapter 7);
the economic analysis for the field measurement device
and reference  method  (Chapter 8); the demonstration
results in summary form (Chapter 9); and the references
used to prepare the ITVR (Chapter 10). Supplemental
information  provided by  Horiba  is  presented  in  the
appendix.

1.1    Description of SITE Program

Performance verification of innovative environmental
technologies is an integral part of the regulatory and
research mission of the  EPA.  The SITE Program was
established by  the  EPA  Office of  Solid  Waste and
Emergency Response (OSWER) and  ORD under  the
Superfund Amendments and Reauthorization Act of 1986.
The overall  goal  of the SITE Program is  to conduct
performance verification studies and to promote  the
acceptance of innovative technologies that may be used to
achieve long-term protection of human health and  the
environment.   The program  is designed to meet three
primary objectives: (1) identify and remove obstacles to
the development  and  commercial use of innovative
technologies,   (2)  demonstrate  promising  innovative
technologies and gather reliable performance and cost
information to support site characterization and cleanup
activities,  and  (3) develop procedures and policies that
encourage the use of innovative technologies at Superfund
sites as well as at other waste sites or commercial facilities.

The  intent  of a  SITE  demonstration is to  obtain
representative, high-quality performance and cost data on
one or more innovative technologies so that potential users
can assess the  suitability of a given technology for a
specific application.  The SITE Program includes  the
following elements:

•   MMT Program—Evaluates innovative technologies
    that sample, detect, monitor, or measure hazardous

-------
    and toxic substances. These technologies are expected
    to  provide better,  faster,  or more cost-effective
    methods for producing real-time data during site
    characterization and remediation studies  than do
    conventional technologies.

    Remediation   Technology  Program—Conducts
    demonstrations of innovative treatment technologies to
    provide reliable performance,  cost, and applicability
    data for site cleanups.

•   Technology  Transfer  Program—Provides  and
    disseminates technical  information  in the form of
    updates,  brochures,  and  other publications  that
    promote  the  SITE  Program   and  participating
    technologies. The  Technology Transfer Program also
    offers technical assistance, training, and workshops to
    support the technologies.  A  significant  number of
    these activities are performed by EPA's Technology
    Innovation Office.

The TPH field measurement device demonstration was
conducted as part of the MMT Program,  which provides
developers  of  innovative  hazardous  waste  sampling,
detection, monitoring,  and measurement  devices with an
opportunity  to  demonstrate the  performance of their
devices under actual field conditions. These devices may
be used to sample, detect, monitor, or measure hazardous
and toxic substances in water, soil gas, soil, and sediment.
The technologies include chemical sensors for in situ (in
place) measurements, soil and sediment samplers, soil gas
samplers, groundwater samplers, field-portable analytical
equipment, and other systems that  support field sampling
or data acquisition and analysis.

The MMT Program promotes acceptance of technologies
that can be used to (1) accurately assess the degree of
contamination at a site, (2) provide data to evaluate
potential effects on human health and the environment, (3)
apply data to  assist in selecting  the  most appropriate
cleanup action, and (4) monitor the effectiveness  of a
remediation process. The program places a high priority
on  innovative  technologies  that provide more  cost-
effective, faster, and safer methods forproducing real-time
or near-real-time data than  do conventional, laboratory-
based technologies. These innovative technologies are
demonstrated under field conditions, and the results are
compiled, evaluated, published, and  disseminated by the
ORD. The primary objectives of the MMT Program are as
follows:
•   Test and verify the performance of innovative field
    sampling and analytical technologies that enhance
    sampling,  monitoring,  and  site  characterization
    capabilities

    Identify  performance   attributes   of   innovative
    technologies to address field sampling,  monitoring,
    and characterization problems in a more cost-effective
    and efficient manner

•   Prepare protocols, guidelines,  methods, and  other
    technical publications that enhance acceptance of these
    technologies for routine use

The MMT Program is administered by the Environmental
Sciences Division of the NERL in Las Vegas, Nevada.
The NERL is the EPA center for investigation of technical
and  management  approaches  for  identifying  and
quantifying risks to human health and the environment.
The NERL mission components  include (1) developing
and evaluating methods and technologies for sampling,
monitoring, and characterizing  water,  air,  soil,  and
sediment; (2) supporting regulatory and policy decisions;
and (3) providing the technical support needed to ensure
effective implementation of environmental regulations
and  strategies.   By demonstrating  innovative  field
measurement devices for TPH in soil, the MMT Program
is supporting the development and evaluation of methods
and  technologies   for  field  measurement  of  TPH
concentrations in a variety of soil types.   Information
regarding the selection of field measurement devices for
TPH is  available in American Petroleum Institute (API)
publications (API 1996, 1998).

The MMT Program's  technology verification process is
designed to conduct demonstrations that will generate
high-quality data so that  potential  users have  reliable
information regarding device performance and cost. Four
steps are inherent in the process: (1) needs identification
and technology selection, (2) demonstration planning and
implementation, (3) report preparation, and (4) information
distribution.

The first step of the  verification process  begins with
identifying technology needs of the EPA and the regulated
community.   The EPA   regional  offices,  the  U.S.
Department of Energy, the U.S. Department of Defense,
industry, and state environmental  regulatory  agencies are
asked  to   identify technology  needs  for  sampling,

-------
monitoring, and measurement of environmental media.
Once a need is identified, a search is conducted to identify
suitable technologies  that will address the need.  The
technology search and identification process consists of
examining  industry  and  trade publications, attending
related  conferences, exploring leads  from  technology
developers  and industry experts, and reviewing responses
to Commerce Business Daily announcements. Selection
of technologies for field testing includes  evaluation of
the candidate technologies based on several  criteria.  A
suitable technology for field testing

•   Is designed for use in the field

•   Is   applicable to  a  variety  of  environmentally
    contaminated sites

    Has potential for  solving problems that  current
    methods cannot satisfactorily address

•   Has estimated costs that  are  lower  than  those  of
    conventional methods

    Is likely to achieve better results than current methods
    in areas such as data quality and turnaround time

•   Uses techniques that are easier or safer than current
    methods

    Is commercially available

Once  candidate   technologies  are  identified,  their
developers   are  asked  to  participate  in a  developer
conference.  This conference gives the  developers  an
opportunity to describe their technologies' performance
and to learn about the MMT Program.

The second step of the verification process is to plan and
implement a demonstration that will generate high-quality
data to  assist potential users in selecting a technology.
Demonstration   planning   activities   include   a
predemonstration sampling and analysis investigation that
assesses existing conditions at the proposed demonstration
site or  sites.  The objectives of the  predemonstration
investigation are to (1) confirm available information on
applicable   physical,  chemical,   and   biological
characteristics of contaminated media at the sites to justify
selection of site areas for the demonstration;  (2) provide
the technology developers with an opportunity to evaluate
the areas, analyze representative  samples, and identify
logistical requirements; (3) assess  the overall logistical
requirements for  conducting  the demonstration;  and
(4) provide the reference laboratory with an opportunity to
identify any matrix-specific analytical problems associated
with the contaminated media and to propose appropriate
solutions.  Information generated through the predemon-
stration investigation  is  used  to develop  the  final
demonstration  design  and  sampling  and  analysis
procedures.

Demonstration planning activities also include preparing
a detailed demonstration plan that describes the procedures
to  be used to verify the performance and cost of each
innovative  technology.     The   demonstration  plan
incorporates  information  generated  during   the
predemonstration investigation as well  as input  from
technology developers, demonstration site representatives,
and technical peer reviewers. The demonstration plan also
incorporates  the  quality assurance  (QA)  and  quality
control (QC) elements needed to produce data of sufficient
quality to document the  performance and cost of each
technology.

During the demonstration, each innovative technology is
evaluated   independently  and,   when  possible   and
appropriate, is compared to a reference technology. The
performance and cost of one innovative technology are not
compared to those of another technology evaluated in the
demonstration.  Rather, demonstration data are used to
evaluate the  individual performance, cost, advantages,
limitations, and field applicability of each technology.

As part of the third step  of the verification process, the
EPA publishes a verification  statement  and  a detailed
evaluation of each technology in an ITVR. To ensure its
quality, the ITVR is published only after comments from
the technology developer and external peer reviewers are
satisfactorily addressed.  In addition, all demonstration
data used to evaluate each innovative technology are
summarized  in  a data evaluation report  (DER) that
constitutes a complete record of the demonstration. The
DER is not  published as  an EPA  document,  but  an
unpublished copy may be obtained from the EPA project
manager.

The fourth step of the verification process is to distribute
information regarding demonstration results. To  benefit
technology developers and potential technology users, the
EPA  distributes  demonstration  bulletins and   ITVRs
through direct  mailings,  at conferences,  and  on the
Internet.   The ITVRs and additional  information on the

-------
SITE Program are available on the EPA ORD web site
(http://www.epa.gov/ORD/SITE).

1.2    Scope of Demonstration

The purpose of the demonstration was to evaluate field
measurement devices for TPH in soil in order to provide
(1) potential users with a better  understanding of the
devices' performance and costs under well-defined field
conditions and (2) the developers with documented results
that will assist them in promoting acceptance and use of
their devices.

Chapter 2 of this ITVR describes both the technology upon
which the OCMA-350 is based and the field measurement
device itself.    Because TPH is a  "method-defined
parameter," the performance results for the  device are
compared to the  results  obtained using  an  off-site
laboratory measurement method—that is, a reference
method. Details on the selection of the reference method
and laboratory are provided in Chapter 5.

The  demonstration had both  primary and  secondary
objectives.   Primary objectives  were critical  to  the
technology verification and required the use of quantitative
results  to  draw  conclusions regarding  each  field
measurement device's performance as well as to estimate
the cost of operating the device.  Secondary objectives
pertained to information that  was useful but did not
necessarily require the use of quantitative results to draw
conclusions regarding the performance of each  device.
Both the primary and secondary objectives are discussed
in Chapter 4.

To meet the demonstration objectives,  samples were
collected from five individual areas at three sites. The first
site is referred to as the Navy Base Ventura County (BVC)
site; is located in PortHueneme, California; and contained
three  sampling areas.  The Navy BVC site lies in EPA
Region 9.  The second site is referred to as the Kelly Air
Force Base (AFB) site; is located in San Antonio, Texas;
and contained one sampling area. The Kelly AFB site lies
in EPA Region 6.  The third  site is  referred to as the
petroleum company (PC) site, is located in north-central
Indiana, and contained one sampling area. The PC site lies
in EPA Region 5.

In preparation for the demonstration, a predemonstration
sampling and analysis investigation was completed at the
three  sites in  January  2000.   The  purpose of  this
investigation was to assess whether the sites and sampling
areas  were appropriate  for evaluating  the  seven field
measurement  devices   based  on   the  demonstration
objectives.  Demonstration field activities were conducted
between June 5 and 18, 2000.  The procedures used to
verify the performance and costs of the field measurement
devices are documented in a demonstration plan completed
in June 2000 (EPA 2000). The plan also incorporates the
QA/QC elements that were needed to generate data of
sufficient quality to document field measurement device
and reference laboratory performance and costs. The plan
is  available  through  the   EPA  ORD   web  site
(http://www.epa.gov/ORD/SITE) or from the EPA project
manager.

1.3    Components and Definition of TPH

To  understand the term  "TPH,"   it is  necessary to
understand the composition of petroleum and its products.
This  section  briefly  describes the  composition  of
petroleum  and  its products  and defines  TPH from  a
measurement  standpoint.    The   organic  compounds
containing only hydrogen and carbon that are present in
petroleum and its derivatives are collectively referred to as
petroleum hydrocarbons  (PHC). Therefore, in this ITVR,
the term "PHC"  is used to identify sample constituents,
and the term "TPH" is used to identify analyses performed
and  the   associated   results  (for  example,   TPH
concentrations).

1.3.1   Composition of Petroleum and Its Products

Petroleum is essentially a mixture of gaseous, liquid, and
solid  hydrocarbons that occur  in  sedimentary  rock
deposits.  On the molecular level, petroleum is a complex
mixture of  hydrocarbons; organic compounds  of sulfur,
nitrogen, and oxygen; and compounds containing metallic
constituents, particularly vanadium, nickel, iron,  and
copper. Based on the limited data available, the elemental
composition of petroleum appears to vary over a relatively
narrow range: 83 to 87 percent carbon, 10 to 14 percent
hydrogen,  0.05 to 6  percent sulfur, 0.1  to 2 percent
nitrogen, and 0.05 to 1.5 percent oxygen.  Metals are
present in petroleum at concentrations of up to 0.1 percent
(Speight 1991).

Petroleum  in the crude state  (crude oil) is a mineral
resource, but  when  refined  it provides liquid  fuels,
solvents, lubricants, and many other marketable products.
The  hydrocarbon  components of crude  oil include

-------
paraffinic, naphthenic, and aromatic groups.  Paraffins
(alkanes)  are saturated,  aliphatic  hydrocarbons  with
straight or branched chains but without any ring structure.
Naphthenes  are  saturated,  aliphatic  hydrocarbons
containing one or more rings, each of which may have one
or more paraffinic side chains (alicyclic hydrocarbons).
Aromatic  hydrocarbons contain one or more aromatic
nuclei, such as benzene, naphthalene, and phenanthrene
ring systems, that may  be  linked with  (substituted)
naphthenic rings or paraffinic side chains. In crude oil, the
relationship  among  the  three primary   groups  of
hydrocarbon  components is a result of hydrogen gain or
loss between any two  groups.    Another  class  of
compounds that  is present in petroleum products such as
automobile gasoline but rarely in crude oil is known as
olefins.   Olefins  (alkenes)  are unsaturated, aliphatic
hydrocarbons.

The distribution of paraffins, naphthenes, and aromatic
hydrocarbons depends on the source of crude oil.  For
example, Pennsylvania crude oil contains high levels of
paraffins (about 50 percent), whereas Borneo crude oil
contains  less than  1  percent paraffins.  As shown in
Figure 1 -1, the proportion of straight or branched paraffins
                                             decreases with increasing molecular weight or boiling
                                             point fraction for a given crude oil; however, this is not
                                             true for naphthenes  or aromatic hydrocarbons.   The
                                             proportion  of  monocyclonaphthenes  decreases  with
                                             increasing molecular weight  or boiling point fraction,
                                             whereas the opposite is true for polycyclonaphthenes (for
                                             example, tetralin and decalin) and polynuclear aromatic
                                             hydrocarbons; the proportion of mononuclear aromatic
                                             hydrocarbons appears to be independent  of molecular
                                             weight or boiling point fraction.

                                             Various petroleum products consisting of carbon and
                                             hydrogen are formed when crude  oil is  subjected to
                                             distillation and other processes in a refinery.  Processing of
                                             crude oil results in petroleum products with trace quantities
                                             of metals and organic compounds that contain  nitrogen,
                                             sulfur,  and  oxygen.  These  products include  liquefied
                                             petroleum gas, gasoline, naphthas, kerosene, fuel  oils,
                                             lubricating  oils,  coke,  waxes,  and  asphalt.   Of these
                                             products,  gasoline, naphthas,  kerosene, fuel oils, and
                                             lubricating  oils  are  liquids  and may be  present  at
                                             petroleum-contaminated sites.  Except  for gasoline and
                                             some naphthas, these products are  made  primarily by
                                             collecting particular boiling point fractions of crude oil
Lighter oils
                                                                           Heavier oils and residues
   100
                               Increasing nitrogen, oxygen, sulfur, and metal content
                                                                  Polynuclear aromatic hydrocarbons
                          Mononuclear aromatic hydrocarbons
            Monocyclonaphthenes
                                                                            Polycyclonaphthenes
              Straight and branched paraffins
                          100
                                 200                300
                                       Boiling point, °C
400
500
Source: Speight 1991

Figure 1-1. Distribution of various petroleum hydrocarbon types throughout boiling point range of crude oil.

-------
from a distillation column. Because this classification of
petroleum products is based on boiling point and not on
chemical composition, the composition of these products,
including the ratio of aliphatic to aromatic hydrocarbons,
varies depending on the source of crude oil.  In addition,
specific information (such as boiling points and carbon
ranges) for different petroleum products, varies slightly
depending on the source of the information. Commonly
encountered forms and blends of petroleum products are
briefly described below.  The descriptions are primarily
based on information in books written by Speight (1991)
and Gary and Handwerk (1993). Additional information
is provided by Dryoff (1993).
1.3.1.1
Gasoline
Gasoline  is a  major  exception  to  the boiling  point
classification described  above  because  "straight-run
gasoline"  (gasoline directly recovered from a distillation
column) is only a small fraction of the blended gasoline
that  is commercially available as  fuel.   Commercially
available gasolines are complex mixtures of hydrocarbons
that boil below  180 °C or at most 225  °C and that contain
hydrocarbons with 4 to 12 carbon atoms per molecule. Of
the commercially available gasolines, aviation gasoline has
a narrower boiling range (38 to 170 °C) than automobile
gasoline (-1 to 200 °C).  In addition, aviation gasoline may
contain high levels of paraffins  (50  to 60 percent),
moderate  levels of naphthenes (20 to 30 percent), a low
level of aromatic hydrocarbons (10 percent),  and no
olefins, whereas automobile gasoline may contain up to
30  percent olefins  and  up to  40 percent aromatic
hydrocarbons.

Gasoline composition can vary widely depending on the
source of crude oil.  In addition, gasoline composition
varies from region to region because of consumer needs for
gasoline  with a high octane rating  to  prevent engine
"knocking."  Moreover, EPA regulations regarding the
vapor pressure of gasoline, the chemicals used to produce
a high octane  rating,  and cleaner-burning  fuels have
affected gasoline composition. For example, when use of
tetraethyl  lead  to produce gasoline with a  high octane
rating was banned by the EPA, oxygenated fuels came into
existence.  Production of these fuels included addition of
methyl-tert-butyl  ether (MTBE),  ethanol,  and  other
oxygenates.  Use  of oxygenated fuels  also  results  in
reduction  of air pollutant emissions (for example, carbon
monoxide and nitrogen oxides).
1.3.1.2     Naphthas

"Naphtha" is a generic term applied to petroleum solvents.
Under  standardized  distillation  conditions,  at least
10 percent of naphthas should distill below 175 °C, and at
least 95 percent of naphthas should distill below 240 °C.
Naphthas can be both aliphatic and aromatic and contain
hydrocarbons with 6 to 14 carbon atoms per molecule.
Depending on the intended use of a naphtha, it may be free
of aromatic hydrocarbons (to make it odor-free) and sulfur
(to make it less toxic and less corrosive). Many forms of
naphthas are commercially available, including Varnish
Makers' and Painters' naphthas (Types I and II), mineral
spirits (Types I through  IV), and aromatic  naphthas
(Types I and II).  Stoddard solvent is an example of an
aliphatic naphtha.
                                            1.3.1.3
           Kerosene
                                            Kerosene is a straight-run petroleum fraction that has a
                                            boiling point range of 205 to 260 °C.  Kerosene typically
                                            contains hydrocarbons with 12 or more carbon atoms per
                                            molecule.  Because of its use as an indoor fuel, kerosene
                                            must be free of aromatic and unsaturated hydrocarbons as
                                            well as sulfur compounds.
                                            1.3.1.4
           Jet Fuels
                                            Jet fuels, which are also known as aircraft turbine fuels,
                                            are  manufactured  by blending gasoline, naphtha, and
                                            kerosene in varying proportions. Therefore, jet fuels may
                                            contain a carbon  range  that  covers  gasoline through
                                            kerosene.   Jet fuels are used in both  military and
                                            commercial  aircraft.  Some examples of jet fuels include
                                            Type A, Type A-l, Type B, JP-4, JP-5,  and JP-8.  The
                                            aromatic hydrocarbon content of these fuels ranges from
                                            20 to 25 percent.   The military jet fuel JP-4 has  a wide
                                            boiling point range (65 to 290 °C), whereas commercial jet
                                            fuels,  including  JP-5 and Types A and  A-l, have  a
                                            narrower boiling point range (175 to 290 °C) because of
                                            safety considerations. Increasing concerns over combat
                                            hazards associated with JP-4 jet fuel led to development of
                                            JP-8 jet fuel, which has a flash point of 38 °C and a
                                            boiling  point range  of  165  to 275 °C.  JP-8 jet fuel
                                            contains hydrocarbons with  9 to  15 carbon atoms per
                                            molecule. Type B jet fuel has a boiling point range of 55
                                            to 230  °C and a  carbon  range of 5  to 13 atoms per
                                            molecule.   A  new  specification  is currently being
                                            developed by the American  Society for Testing and
                                            Materials (ASTM) for Type B jet fuel.

-------
1.3.1.5
Fuel Oils
Fuel oils  are divided into two classes: distillates and
residuals.  No. 1 and 2 fuel oils are distillates and include
kerosene, diesel, and home heating oil. No. 4,5, and 6 fuel
oils are residuals or black oils, and they all contain crude
distillation tower bottoms (tar) to which cutter stocks
(semirefined or refined distillates) have been added. No. 4
fuel oil contains the most cutter stock, and No. 6 fuel oil
contains the least.

Commonly available fuel oils include No. 1, 2, 4, 5, and 6.
The boiling points, viscosities, and densities of these fuel
oils increase with  increasing number designation.  The
boiling point ranges for No. 1,2, and 4 fuel oils are about
180 to 320, 175 to 340, and 150 to 480 °C, respectively.
No. 1 and 2 fuel oils contain  hydrocarbons with 10  to
22 carbon atoms per molecule; the carbon range for No. 4
fuel oil is  22 to 40  atoms per molecule. No.  5 and 6 fuel
oils have a boiling point range of 150 to 540 °C but differ
in the amounts of residue they contain: No. 5  fuel oil
contains a small amount of residue, whereas No. 6 fuel oil
contains a large amount. No.  5 and 6 fuel  oils contain
hydrocarbons with 28 to 90 carbon atoms per molecule.
Fuel oils  typically contain about  60 percent aliphatic
hydrocarbons and 40 percent aromatic hydrocarbons.
1.3.1.6
Diesel
Diesel is primarily used to operate motor vehicle and
railroad diesel engines. Automobile diesel is available in
two grades: No.  1 and 2. No.  1 diesel, which is sold in
regions with cold climates, has a boiling point range of 180
to 320 °C and a cetane number above 50. The  cetane
number is similar to the octane number of gasoline; a
higher number corresponds to less knocking. No. 2 diesel
is very similar to No. 2 fuel oil.  No. 2 diesel has a boiling
point range  of 175 to 340  °C and a minimum  cetane
number of 52. No. 1 diesel is used in high-speed engines
such as truck and bus engines, whereas No. 2 diesel is used
in other diesel engines. Railroad diesel is similar to No. 2
diesel but has a higher boiling  point (up to 370 °C) and
lower cetane number (40 to 45). The ratio of aliphatic to
aromatic hydrocarbons in diesel is about 5. The carbon
range for hydrocarbons present  in diesel is 10 to 28 atoms
per molecule.

1.3.1.7     Lubricating Oils

Lubricating oils can be distinguished from other crude oil
fractions by their high boiling points (greater than 400 °C)
and viscosities.   Materials  suitable  for  production  of
lubricating oils are composed principally of hydrocarbons
containing 25 to 35 or even 40 carbon atoms per molecule,
whereas residual stocks may contain hydrocarbons with 50
to 60 or more (up to 80 or so) carbon atoms per molecule.
Because it is difficult to isolate hydrocarbons from the
lubricant  fraction of petroleum,  aliphatic  to  aromatic
hydrocarbon ratios are not well documented for lubricating
oils. However, these ratios are expected to be comparable
to those of the source crude oil.

1.3.2  Measurement of TPH

As  described in Section  1.3.1, the  composition  of
petroleum and its products is complex and variable, which
complicates TPH measurement. The measurement of TPH
in soil is further complicated by weathering effects. When
a petroleum  product  is released to soil, the product's
composition  immediately  begins to  change.    The
components with lower boiling points are volatilized, the
more water-soluble components migrate to groundwater,
and biodegradation can affect many other components.
Within a short period, the contamination remaining in soil
may have  only some  characteristics in common with the
parent product.

This section  provides a historical perspective on TPH
measurement,  reviews  current  options   for  TPH
measurement in soil, and discusses the definition of TPH
that was used for the demonstration.

1.3.2.1     Historical Perspective

Most  environmental  measurements  are  focused  on
identifying and quantifying a particular trace element (such
as  lead)  or organic  compound  (such  as  benzene).
However,  for some  "method-defined" parameters,  the
particular  substance being measured may  yield different
results depending on the measurement  method  used.
Examples  of such parameters include oil and grease and
surfactants. Perhaps the most problematic of the method-
defined parameters is TPH. TPH arose as a parameter for
wastewater analyses in the  1960s because of petroleum
industry concerns that the  original "oil and grease"
analytical  method, which is gravimetric in nature, might
inaccurately characterize petroleum industry  wastewaters
that contained  naturally occurring  vegetable  oils and
greases  along with  PHCs.   These naturally occurring
materials are typically long-chain fatty acids (for example,
oleic acid, the major component of olive oil).

-------
Originally, TPH was defined as any material extracted with
a particular solvent that is not adsorbed by the silica gel
used to remove fatty acids and that is not lost when the
solvent is evaporated.  Although this definition  covers
most of the components of petroleum products, it includes
many  other  organic compounds  as  well,  including
chlorinated  solvents, pesticides, and  other synthetic
organic  chemicals.    Furthermore,   because  of  the
evaporation step in the gravimetric analytical method, the
definition excludes  most   of  the  petroleum-derived
compounds in gasoline that  are volatile in nature.  For
these reasons, an infrared analytical method was developed
to measure TPH.  In this method, a calibration standard
consisting of three components is analyzed at a wavelength
of 3.41  micrometers ((im),  which corresponds to an
aliphatic  CH2 hydrocarbon stretch. As shown in Table 1 -1,
the calibration standard is designed to  mimic a petroleum
product having a relative distribution  of  aliphatic and
aromatic  compounds as well as a certain percentage of
aliphatic  CH2 hydrocarbons.  The infrared  analytical
method indicates that any compound that is extracted by
the solvent, is not adsorbed by silica gel, and contains  a
CH2 bond is a PHC.  Both the gravimetric and infrared
analytical methods  include an optional,  silica  gel
fractionation step to remove  polar, biogenic compounds
such as fatty acids, but this cleanup  step can also remove
some  petroleum  degradation products that are polar in
nature.

In the 1980s, because  of the  change in focus from
wastewater analyses to characterization of hazardous waste
sites that  contained contaminated soil, many parties began
to adapt  the existing wastewater analytical methods for
application to soil.  Unfortunately, the term "TPH" was in
common  use, as many states had adopted this term (and the
wastewater analytical methods) for cleanup activities at
underground storage tank (UST) sites. Despite efforts by
the API and others to establish new analyte names (for
example, gasoline range organics  [GRO] and diesel range
organics  [DRO]), "TPH" is  still present in many state
regulations as a somewhat ill-defined term, and most state
programs still have cleanup criteria for TPH.

1.3.2.2     Current Options for TPH Measurement
           in Soil

Three widely used technologies measure some form of
TPH in soil to some degree. These technologies were used
as starting points in deciding how to define TPH for the
demonstration.  The three  technologies and the analytes
measured are summarized in Table 1-2.

Of the three technologies, gravimetry and infrared are
discussed in Section 1.3.2.1. The third technology, the gas
chromatograph/flame ionization detector (GC/FID), came
into use because of the documented shortcomings of the
other two technologies. The GC/FID had long been used
in the petroleum refining industry as a product QC tool to
determine the boiling point distribution of pure petroleum
products. In the 1980s, environmental laboratories began
to apply this technology along with sample preparation
methods developed for soil samples to measure PHCs at
environmental levels (Zilis,  McDevitt, and Parr  1988).
GC/FID methods measure all organic compounds that are
extracted by the solvent and that can be chromatographed.
However, because of method limitations, the very volatile
portion  of  gasoline compounds containing four or five
carbon atoms per molecule is not addressed by GC/FID
methods; therefore,  100  percent recovery  cannot be
achieved for  pure gasoline.   This  omission is not
considered  significant  because these low-boiling-point
aliphatic compounds (1) are not expected to be present in
environmental samples (because  of volatilization) and
(2) pose less environmental risk  than  the  aromatic
hydrocarbons in gasoline.

The primary limitation of GC/FID methods relates to the
extraction solvent used. The  solvent should not interfere
with the analysis, but to achieve environmental levels of
detection (in the  low milligram per kilogram [mg/kg]
Table 1-1. Summary of Calibration Information for Infrared Analytical Method
Standard
Constituent
Hexadecane
Isooctane
Chlorobenzene
Constituent Type
Straight-chain aliphatic
Branched-chain aliphatic
Aromatic
Portion of Constituent
in Standard
(percent by volume)
37.5
37.5
25
Number of Carbon Atoms
Aliphatic
CH3
2
5
0
CH2
14
1
0
CH
0
1
0
Aromatic
CH
0
0
5
Average
Portion of Aliphatic CH2 in
Standard Constituent
(percent by weight)
91
14
0
35

-------
Table 1-2. Current Technologies for TPH Measurement
Technology
Gravimetry
Infrared
Gas chromatograph/flame ionization detector
What Is Measured
All analytes removed from the sample by the
extraction solvent that are not volatilized
All analytes removed from the sample by the
extraction solvent that contain an aliphatic CH2
stretch
All analytes removed from the sample by the
extraction solvent that can be chromatographed
and that respond to the detector
What Is Not Measured
Volatiles; very polar organics
Benzene, naphthalene, and other aromatic
hydrocarbons with no aliphatic group attached;
very polar organics
Very polar organics; compounds with high
molecular weights or high boiling points
range) for soil, some concentration of the extract is needed
because the sensitivity of the FID is in the nanogram (ng)
range.   This  limitation  has resulted  in  three  basic
approaches for  GC/FID analyses for GRO, DRO, and
PHCs.

For GRO analysis, a GC/FID method was developed as
part of research  sponsored by API and was the subject of
an interlaboratory validation study (API 1994); the method
was first published in  1990.  In this method, GRO is
defined as the sum of the organic compounds in the boiling
point  range  of  60 to  170 °C, and the  method uses a
synthetic calibration standard as both a window-defining
mix and a quantitation standard.  The GRO method was
specifically incorporated  into EPA "Test Methods for
Evaluating Solid Waste" (SW-846) Method 8015B in 1996
(EPA 1996).  The GRO method uses the purge-and-trap
technique for sample preparation, effectively limiting the
TPH components to the volatile compounds only.

For DRO analysis, a GC/FID method was developed under
the sponsorship of API as a companion to the GRO method
and was interlaboratory-validated in 1994.  In the DRO
method,  DRO  is  defined as the  sum  of the  organic
compounds in the boiling point range of 170 to 430 °C. As
in the GRO  method,  a synthetic calibration standard is
used  for quantitation.   The DRO method was also
incorporated into SW-846 Method 8015B in 1996. The
technology used  in  the  DRO  method can measure
hydrocarbons with boiling points up to 540 °C. However,
the hydrocarbons with boiling points in the range of 430 to
540   °C  are   specifically  excluded  from  SW-846
Method 8015B  so  as not  to include the higher-boiling-
point petroleum products. The DRO method uses a solvent
extraction and concentration step, effectively limiting the
method to nonvolatile hydrocarbons.

For PHC analysis,  a GC/FID method was developed by
Shell  Oil Company (now Equilon Enterprises).   This
method was interlaboratory-validated along with the GRO
and DRO methods in an API study in  1994.  The PHC
method originally defined PHC  as  the sum of  the
compounds in the boiling point range of about 70 to
400  °C,  but  it now defines PHC as  the  sum of  the
compounds in the boiling point range of 70 to 490  °C.
The method provides options for instrument calibration,
including use of synthetic standards, but it recommends
use of products similar to the contaminants present at the
site  of concern.    The  PHC  method has  not  been
specifically incorporated  into  SW-846; however,  the
method has been used as the basis for the TPH methods in
several states, including Massachusetts,  Washington, and
Texas. The PHC method uses solvent microextraction and
thus has a higher detection limit than the GRO and DRO
methods. The PHC method also begins peak integration
after elution of the solvent peak for n-pentane.  Thus, this
method  probably  cannot  measure   some  volatile
compounds (for example,  2-methyl pentane and MTBE)
that are measured using the GRO method.
1.3.2.3
Definition of TPH
It is not possible to establish a definition of TPH that
would  include crude oil  and its refined products and
exclude other organic  compounds.  Ideally, the TPH
definition selected for the demonstration would have

•   Included compounds that are PHCs, such as paraffins,
    naphthenes, and aromatic hydrocarbons

    Included, to the  extent possible, the  major liquid
    petroleum products  (gasoline, naphthas, kerosene, jet
    fuels, fuel oils, diesel, and lubricating oils)

    Had little inherent bias based on the composition of an
    individual manufacturer's product

-------
•   Had  little  inherent  bias  based  on  the  relative
    concentrations of aliphatic and aromatic hydrocarbons
    present

•   Included much of the volatile portion of gasoline,
    including all weathered gasoline

•   Included MTBE

    Excluded crude oil residuals  beyond the extended
    diesel range organic (EDRO) range

•   Excluded  nonpetroleum organic  compounds  (for
    example,   chlorinated  solvents,  pesticides,
    poly chlorinated  biphenyls  [PCB],  and  naturally
    occurring oils and greases)

    Allowed TPH measurement using a widely accepted
    method

•   Reflected  accepted TPH measurement practice in
    many states

Several states, including Massachusetts, Alaska, Louisiana,
and North Carolina, have implemented or are planning to
implement a TPH contamination cleanup approach based
on the aliphatic and aromatic hydrocarbon fractions of
TPH.  The action levels  for the  aromatic hydrocarbon
fraction are more stringent than those for the aliphatic
hydrocarbon fraction.  The approach used in the above-
mentioned   states  involves  performing  a   sample
fractionation procedure and two analyses to determine the
aliphatic  and aromatic hydrocarbon concentrations  in a
sample. However, in most applications of this approach,
only a few samples are subjected to the dual aliphatic and
aromatic  hydrocarbon  analyses  because of the costs
associated with performing sample fractionation and two
analyses.

For the demonstration, TPH was not defined based on the
aliphatic and aromatic hydrocarbon fractions because

•   Such a definition is used in only a few states.

•   Variations  exist among the sample fractionation and
    analysis procedures used in different states.
•   The   repeatability  and   versatility  of   sample
    fractionation and  analysis  procedures are not well
    documented.

•   In some states, TPH-based action levels are still used.

•   The associated analytical costs are high.

As stated in Section 1.3.2.2, analytical methods currently
available for measurement  of TPH  each exclude some
portion  of TPH and are unable to measure TPH alone
while excluding all other organic compounds, thus making
TPH a method-defined parameter.  After consideration of
all the information  presented above,  the GRO and DRO
analytical methods were selected for TPH measurement for
the  demonstration.  However, because of the general
interest  in higher-boiling-point petroleum products, the
integration range of the DRO method was extended to
include  compounds with boiling points up to  540 °C.
Thus, for the demonstration, the TPH concentration was
the sum of all organic compounds that have boiling points
between 60 and 540 °C and that can be chromatographed,
or the sum of the results obtained using the GRO and DRO
methods.   This approach accounts  for most gasoline,
including  MTBE,   and virtually  all  other  petroleum
products and excludes a portion (25 to 50 percent) of the
heavy lubricating oils.  Thus, TPH measurement for the
demonstration included PHCs as well as  some organic
compounds that are not PHCs.   More specifically, TPH
measurement  did  not exclude nonpetroleum  organic
compounds such as  chlorinated solvents, other synthetic
organic  chemicals  such as pesticides and  PCBs,  and
naturally  occurring oils and  greases.   A  silica gel
fractionation   step   used to  remove  polar, biogenic
compounds such as fatty acids in some GC/FID methods
was not included in the sample preparation step because,
according to  the State of California, this step can also
remove some petroleum degradation products that are also
polar in nature  (California Environmental  Protection
Agency 1999). The step-by-step approach used to select
the  reference  method for  the  demonstration  and the
project-specific procedures implemented for soil sample
preparation and analysis using the reference method are
detailed in Chapter 5.
                                                    10

-------
                                              Chapter 2
                     Description of Infrared Analysis and the OCMA-350
Measurement of TPH in soil by field measurement devices
generally involves extraction of PHCs from soil using an
appropriate solvent followed by measurement of the TPH
concentration in the extract using an optical method. An
extraction solvent is selected that will not interfere with the
optical measurement of TPH in the extract.  Some field
measurement devices use light in the visible wavelength
range, and others use light outside the visible wavelength
range (for example, infrared radiation).

The  optical measurements made by field measurement
devices  may   involve   absorbance,  reflectance,  or
fluorescence.  In general, the optical measurement for a
soil extract is compared to a calibration curve in order to
determine the TPH concentration. Calibration curves may
be developed by (1) using a series of calibration standards
selected based on the type of PHCs being measured at a
site  or (2) establishing  a correlation between off-site
laboratory  measurements and field measurements for
selected, site-specific soil samples.

Field measurement devices  may  be  categorized  as
quantitative, semiquantitative,  and qualitative.   These
categories are explained below.

•   A quantitative measurement device measures TPH
    concentrations ranging from its reporting limit through
    its linear range. The measurement result is reported as
    a single, numerical  value that has  an  established
    precision and accuracy.

    A semiquantitative  measurement device measures
    TPH concentrations  above its  reporting limit. The
    measurement result may be reported as a concentration
    range with lower and upper limits.

    A  qualitative  measurement  device indicates the
    presence or  absence of PHCs above or below a
    specified value (for example, the reporting limit or an
    action level).

The OCMA-350 is a field measurement device capable of
providing quantitative TPH measurement results.  Optical
measurements made using the OCMA-350 are based on
infrared  analysis, which is  described  in Section  2.1.
Section  2.2  describes  the  OCMA-350   itself,  and
Section 2.3 provides Horiba contact information.   The
technology and device descriptions presented below are
not intended to provide complete operating procedures for
measuring  TPH   concentrations  in  soil  using  the
OCMA-350. Detailed operating procedures forthe device,
including soil extraction, TPH measurement, and TPH
concentration calculation procedures, are available from
Horiba. Supplemental information provided by Horiba is
presented in the appendix.

2.1    Description of Infrared Analysis

This section describes the technology, infrared analysis,
upon which the  OCMA-350 is based. This technology is
suitable for measuring aromatic and aliphatic hydrocarbons
independent of their carbon  range.  TPH measurement
using infrared analysis involves extraction of PHCs in soil
using an organic solvent. Light in the infrared  range is
used to  irradiate  the  extract and  measure  its TPH
concentration.

Infrared  spectrophotometry generally covers the 1- to
15-(im wavelength region.  All organic molecules  and
some inorganic ions absorb light energy in  this region.
Absorbance  of infrared energy  results in  changes in
vibrational motion in  a  molecule.    The  molecular
vibrations observed may  be divided into  two  types,
stretching and  bending.  Stretching  is the rhythmic
movement of the  atoms  back and forth along the bond
axis, whereas bending involves a change in bond angle
                                                    11

-------
between two atoms that are bonded to a third atom or the
movement of a group of atoms with respect to the rest of
the molecule.   Based  on the  unique absorbance  and
vibrational motion associated with each compound, the
type of compound in a sample can be identified using
infrared analysis (Fritz and Schenk 1987).

The energy associated with  shorter  or  near-infrared
wavelengths (less than 4 (im) used for TPH measurement
causes mostly  stretching vibrations of bonds between
hydrogen and heavier atoms.  TPH measurement using
infrared  analysis  involves   absorbance  measurement
because the carbon-hydrogen bonds in the hydrocarbons
absorb  infrared  light.    During infrared  analysis,
absorbances associated  with  CH,  CH2,   and  CH3
configurations are measured at a wavelength close to 3,400
nanometers (nm).   Specifically,  infrared devices  that
operate in the 3,380-to 3,500-nm wavelength range should
be able to measure CH (3,380 nm), CH2 (3,420 nm), and
CH3 (3,500 nm) configurations (Simard and others 1951).

Table 2-1 shows the absorbance values of a series of waste
oils, refinery products,  and pure compounds in samples
collected from a wide variety of locations at a petroleum
refinery. The absorbance variations shown in Table 2-1
are not great enough to seriously affect sample analytical
results when TPH concentrations below 10 milligrams per
liter (mg/L) are being measured unless high concentrations
of the lower boiling point  aromatic  hydrocarbons are
present.  Based on this information,  the  most  serious
shortcoming of infrared analysis is likely its inability to
detect the lower homologs of the  aromatic hydrocarbon
series. Because the aromatic  C-H absorption is relatively
weak,  only aromatic  hydrocarbons  with  appreciable
paraffin  or  cycloparaffin  side chains  exhibit  normal
behavior during TPH measurement (Simard and others
1951).

During infrared analysis, a sample extract is placed in a
glass  or quartz cuvette that is then  inserted into the
spectrophotometer and covered with an  opaque  light
shield.  A beam of infrared light is then passed through
the sample  extract.    Infrared sources  are generally
continuum sources, which emit radiation at intensities that
vary smoothly over ranges of wavelengths.  A typical
example  of  an   infrared  source  for  near-infrared
instruments used for TPH measurement is a tungsten lamp
(Fritz and Schenk  1987; Ewing 1969).
Table 2-1. Infrared Absorbance Values of CH, CH2, and CH3 Groups

                          Absorbance of 16-Milligram per Liter
Sample                      Sample in Carbon Tetrachloride
Waste oils
  From kerosene treating area              0.424
  From gasoline treating area              0.384
  From pipe stills and gasoline              0.374
  desulfurization area
  From solvent extraction area              0.415
  From thermal cracking area              0.318
  From catalytic cracking area              0.355
  From wax refining area                 0.399
  From crude oil handling area              0.409
Refinery products
  Light lubricating oil                     0.427
  Heavy lubricating oil                    0.491
  Light Refugio crude oil                  0.385
  Furnace oil from catalytic                0.339
  cracking
  No. 6 fuel oil                          0.287
  White oil (medicinal)                    0.417
  Kerosene                           0.440
  Gasoline base stock                    0.387
Pure compounds
  Cetane                             0.615
  Isooctane                           0.404
  n-Heptane                           0.613
  p-Di-tert-butyl cyclohexane               0.396
  Ethyl cyclohexane                      0.543
  Cyclohexane                         0.703
  1 -Tetradecene                       0.427
  Decene                             0.414
  Diisobutylene                         0.208
  High molecular weight alkyl              0.155
  benzene
  Cumene                            0.109
  Mixed xylenes                        0.070
  Benzene                            0.000


Source: Simard and others 1951
Some of the infrared radiation emitted by the source is
absorbed by compounds in the sample extract, and the rest
of the radiation passes through the extract.  Absorbance,
which is defined as the logarithm of the ratio of the radiant
power of the light source to that of the light that passes
through the sample extract, is measured by a photoelectric
detector in the spectrophotometer (Fritz and Schenk 1987).
Absorbance can be calculated using Equation 2-1.
                                                      12

-------
where
             A = log (I0/l)


       Absorbance

       Intensity of light source
(2-1)    2.2    Description of OCMA-350

        The OCMA-350 was developed by Horiba and has been
        commercially available since 1995. This section describes
        the device and summarizes its operating procedure.
        I   =  Intensity of light that passes through the
              sample extract

Therefore, the intensity of the light that passes through the
sample   extract  is  inversely  proportional  to  the
concentration of target compounds in the extract,  or the
intensity of the light absorbed by the extract  is directly
proportional to the concentration of target compounds in
the extract.  The absorbance of a  sample extract thus
measured is directly proportional to the concentration of
PHCs  present   in  the  extract  in accordance   with
Beer-Lambert's law, which allows Equation  2-1  to  be
expressed as shown in Equation 2-2.
                      A =
                                        (2-2)
where
    A
    b
    c
=  Absorbance
=  Molar absorptivity (centimeter [cm] per mole
   per L)

=  Light path length (cm)

=  Concentration of  absorbing  species  (mole
   per L)
Thus, according to Beer-Lambert's law, the absorbance of
hydrocarbons is directly proportional to the concentrations
of the absorbing hydrocarbons and the path length of the
infrared radiation that is  not absorbed by the  sample
extract and passes through the extract.  In Equation 2-2,
the molar absorptivity is a proportionality constant, which
is a characteristic of the  absorbing hydrocarbons and
changes as the wavelength  of the light irradiating the
sample extract changes. Therefore, Beer-Lambert's law
applies only to monochromatic light (light energy of one
wavelength).

After the absorbance of the sample extract is measured, the
TPH concentration in the  extract  can be determined by
comparing the absorbance reading to a calibration curve of
absorbance  values  and  corresponding  hydrocarbon
concentrations for a  series of known standards selected
based on the type  of PHCs being measured at a site.
        2.2.1  Device Description

        Infrared analysis using the OCMA-350 involves use of a
        single-beam,  fixed-wavelength, nondispersive infrared
        (NDIR) spectrophotometer to determine the concentration
        of PHCs in a  liquid sample extract.  NDIR spectro-
        photometers offer several advantages over conventional,
        scanning infrared spectrophotometers that disperse infrared
        light using a diffraction  grating component.   Whereas
        scanning infrared spectrophotometers take about 1 to 3
        minutes to scan a sample and have moderate sensitivity
        and stability, NDIR spectrophotometers  can  achieve a
        stable  reading  in  about  5 seconds and have greater
        sensitivity and stability (Fritz and Schenk  1987).

        In the OCMA-350, infrared radiation from atungsten lamp
        is  transmitted  through   a cylindrical, quartz  cuvette
        containing a sample extract. The radiation that has passed
        through the extract enters a detector containing a filter that
        isolates analytical wavelengths in the 3,400- to 3,500-nm
        range.  According to Horiba, the OCMA-350 can measure
        both aromatic and aliphatic hydrocarbons independent of
        their carbon range.  The  device's response to aliphatic
        hydrocarbons is greater  than its response to  aromatic
        hydrocarbons.

        According to Horiba, several solvents can be used to
        complete extraction of  soil  samples for OCMA-350
        analysis, including Horiba's proprietary S-316 extraction
        solvent,  Freon  113,  tetrachloroethene,  and  carbon
        tetrachloride.  For the demonstration, Horiba used its
        proprietary  S-316  extraction  solvent  to extract soil
        samples.  Horiba recommends use of the S-316 extraction
        solvent because it has a higher boiling point (134 °C) and
        lower  freezing  point (-143 °C) than other  extraction
        solvents.  In addition, the S-316 extraction  solvent is
        nonflammable,   nontoxic,  and  relatively  nonvolatile
        because of its low vapor pressure.

        The  OCMA-350  can   be used   to  measure  PHC
        concentrations in soil.   Because hydrocarbons absorb
        infrared radiation, the OCMA-350  responds to their
        concentrations in sample  extracts.  Standards of known
        concentrations can be used to develop a calibration curve
                                                     13

-------
of absorbance versus TPH concentration, that can in turn
be used to  calculate the TPH concentrations in actual
samples.   According  to Horiba, the  OCMA-350 can
achieve a method detection limit (MDL) of 1 mg/kg for
TPH and is linear up to 1,000 mg/kg in soil.  Horiba also
claims that the device can achieve repeatability of plus or
minus (±) 2 mg/kg from 0 to 99.9 mg/kg; ±4 mg/kg from
100  to  200  mg/kg;   and   ±10 mg/kg from  201  to
1,000 mg/kg.  No information is currently available from
Horiba regarding the  accuracy  of the  device.    An
evaluation of the MDL, accuracy, and precision achieved
by the OCMA-350 during the demonstration is presented
in Chapter 7.

According to  Horiba, the OCMA-350  has  an operating
temperature range of 0 to 40  °C and humidity range of 0 to
90 percent.  Horiba  does not specify a particular storage
temperature for the device.

Table 2-2 lists components of the OCMA-350 and supplies
that may or may not be required for measuring TPH in soil
using the device. The supplies are categorized according
to whether or not they are available from Horiba.

Supplies associated with TPH measurement using  the
OCMA-350 may also be categorized according to whether
they are expendable or reusable.  Expendable supplies
include 10 milliliters (mL) of B-heavy oil that comes with
the OCMA-350;  S-316  extraction solvent;  anhydrous
sodium sulfate;  40-mL, volatile organic analysis (VOA)
vials; 11-cm-diameter, No. 40 Whatman filter papers; and
10-mL,  disposable, glass  pipettes.   Horiba generally
recommends use of B-heavy oil as the default calibration
standard when the hydrocarbon content of soil is unknown.
According to  Horiba, B-heavy oil may be replaced with
diesel   as   the   calibration  standard   based  on  the
contamination at a site.

Reusable supplies include the infrared spectrophotometer
and all its accessories except B-heavy oil, Model VC-50
ultrasonic  mixer,  Model  SR-300 solvent  reclaimer,
Centronics printer cable, printer, balance, stainless-steel
spatulas, filter forceps, glass  funnels, glass beakers, pipette
bulb,    10-microliter  (|iL)  microsyringe,  bottle-top
dispenser, test tube rack, and funnel rack and stand.  Of
these  reusable supplies, items required  for TPH analysis
using the OCMA-350 include the balance, stainless-steel
spatulas, filter forceps, glass  funnels, glass beakers, pipette
bulb,  test tube rack, and funnel  rack and stand; Horiba
considers the other items to be optional.   During  the
demonstration, Horiba used the Model  GE-50 ultrasonic
Table 2-2. OCMA-350 Components and Supplies

Spectrophotometer and accessories
    Infrared spectrophotometer
    Proprietary, 10-mm, quartz cuvette with Teflon® cap
    25-uL microsyringe
    10-mL syringe
    10 ml of B-heavy oil
    Three 40-mL, volatile organic analysis vials
    Three 11-centimeter-diameter, No. 40 Whatman filter papers
    Two 10-mL, disposable, glass pipettes
    Instruction manual and simplified operating instruction sheet
Supplies available from Horiba
    S-316 extraction solvent (1.5- or 7-kg bottle)
    Model VC-50 ultrasonic mixer3
    Model SR-300 solvent reclaimer
    Centronics printer cable
Supplies not available from Horiba
    Anhydrous sodium sulfate, 10-60 mesh
    Printer
    Balance
    Stainless-steel spatulas
    Filter forceps
    Glass funnels
    Glass beakers
    Pipette bulb
    10-uL microsyringe
    Bottle-top dispenser
    Test tube rack
    Funnel rack and stand
Notes:

uL =  Microliter
kg =  Kilogram
mL  =  Milliliter
mm  =  Millimeter
    During the demonstration, Horiba used the Model GE-50 ultrasonic
    mixer, which is no longer available from the developer. According
    to Horiba,  the Model  GE-50 and VC-50 ultrasonic mixers are
    functionally the same.
mixer, which is no longer available from the developer.
According to Horiba, the Model VC-50 ultrasonic mixer
that is currently available  is functionally  the  same  as
Model GE-50.

The OCMA-350  infrared spectrophotometer  and  its
accessories come in an optional, hard-plastic carrying case
and weigh 36 pounds. The carrying case is 22 inches long,
16  inches wide, and 15 inches high.  The device can  be
operated using a 110- or 220-volt alternating current (AC)
power source.   It may also be operated using a direct
current (DC) power source such as a 12-volt power outlet
in an automobile  along  with a DC/AC  power  inverter.
During  the  demonstration,  Horiba   operated  the
OCMA-350 using AC power from the demonstration field
trailer.

The spectrophotometer presents results in units selected by
the user during calibration,  such  as mg/kg in soil  or
                                                       14

-------
absorbance values.  During the demonstration, Horiba
programmed the spectrophotometer to present results as
mg/kg in soil.  The spectrophotometer is also equipped
with a parallel printer port and an  RS232-C data port to
allow transfer of data to a computer or other data logger.
The spectrophotometer keeps a record of the time and date
of each measurement and a record  of each data set that it
generates. Spectrophotometer printouts contain the time
and date of each measurement along with the concentration
or absorbance value, which facilitates recordkeeping.

According to Horiba, about 20 samples can be analyzed by
one field technician using the OCMA-350 over an 8-hour
period.  The  device is relatively easy to use.  Normal
training for using the device is limited to  reading the
instruction manual and simplified operating instruction
sheet.  Horiba offers a  1-day  training course covering
operation, maintenance, and service of the OCMA-350, but
according to the developer, purchasers of the device rarely
request this training.   Horiba also provides technical
support over the telephone during regular business hours
at no additional cost.

According  to Horiba, the  OCMA-350 is  innovative
because it can reuse proprietary S-316 extraction solvent
that has  been recycled  by the Model SR-300  solvent
reclaimer.   According  to  Horiba,  the  reuse of this
extraction solvent  can reduce solvent costs  by  up to
90 percent.

2.2.2  Operating Procedure

The  OCMA-350 is calibrated using  known standards.
Horiba generally recommends use  of B-heavy oil, which
is provided with the OCMA-350, as the default calibration
standard when the hydrocarbon content of soil is unknown.
Based on the results of the predemonstration investigation,
Horiba used  diesel as the calibration standard  for the
demonstration. Horiba conducted a three -point calibration
using S-316 extraction solvent as a blank and 100- and
200-mg/L diesel standards at the beginning and end of
each day and after every batch of samples was analyzed.
Concrete sand was also used as a sample blank for a
calibration check once a day.

During the demonstration, Horiba dried soil by adding
anhydrous sodium sulfate.   Extraction of a given  soil
sample was typically performed by adding 20 mL of S-316
extraction solvent to 5.0 grams of the sample. The mixture
was agitated using the Model GE-50 ultrasonic mixer.  The
sample extract was decanted into a beaker through a filter-
lined funnel, and then the filtrate was poured into a quartz
cuvette. The cuvette was placed in the spectrophotometer
for TPH measurement. Horiba pressed the "Run" button
and read the concentration on the digital display.

2.3    Developer Contact Information

Additional  information about the OCMA-350  can be
obtained from the following source:

    Horiba Instruments Incorporated
    Mr. Jim Vance
    17671 Armstrong Avenue
    Irvine, CA 92614
    Telephone: (800) 4HORIBA, extension 170
    Fax: (949) 250-0924
    E-mail: jim.vance@horiba.com
    Internet: www.horiba.com
                                                    15

-------
                                               Chapter 3
                                 Demonstration Site Descriptions
This  chapter  describes  the  three  sites  selected  for
conducting the demonstration. The first site is referred to
as the Navy BVC site;  it is located in Port Hueneme,
California, and contains three sampling areas. The second
site is referred to as the Kelly AFB site; it is located in San
Antonio, Texas, and contains one sampling area. The third
site is referred to as the PC site; it  is located in north-
central Indiana and contains one sampling area.  After
review of the information available on these and other
candidate  sites, the Navy BVC, Kelly AFB, and PC sites
were selected based on the following criteria:

    Site Diversity—Collectively, the three sites contained
    sampling areas with the different soil types and  the
    different levels and  types  of  PHC contamination
    needed to  evaluate the seven  field measurement
    devices selected for the demonstration.

•   Access and Cooperation—The site representatives
    were interested in supporting the demonstration by
    providing site access for collection of soil samples
    required for the demonstration.  In addition, the field
    measurement devices were to be demonstrated at the
    Navy BVC site using soil samples from all three sites,
    and the Navy BVC site representatives were willing to
    provide the site support facilities required  for  the
    demonstration and to support a visitors' day during the
    demonstration.    As  a testing location  for   the
    Department  of Defense  National  Environmental
    Technology Test Site program, the Navy BVC site is
    used to  demonstrate technologies and systems  for
    characterizing  or  remediating  soil,  sediment,  and
    groundwater contaminated with fuel hydrocarbons or
    waste  oil.

To ensure that the sampling areas were selected based on
current  site   characteristics,  a   predemonstration
investigation was conducted.  During this investigation,
samples were collected from the five candidate areas and
were  analyzed for GRO  and EDRO using  SW-846
Method 8015B (modified)  by the reference laboratory,
Severn Trent Laboratories in Tampa, Florida (STL Tampa
East). The site descriptions in Sections 3.1 through 3.3 are
based  on   data   collected  during   predemonstration
investigation sampling activities, data collected during
demonstration  sampling  activities,   and information
provided  by   the  site   representatives.     Physical
characterization of samples was performed in the field by
a geologist during both predemonstration investigation and
demonstration activities.

Some of the predemonstration investigation samples were
also analyzed by the OCMA-350 developer, Horiba, at its
facility. Horiba used reference laboratory and OCMA-350
results to  gain a preliminary  understanding  of the
demonstration  samples   and  to  prepare   for  the
demonstration.

Table 3-1 summarizes key site characteristics, including
the contamination type, sampling depth intervals,  TPH
concentration ranges, and soil type in each sampling area.
The TPH concentration ranges and soil types presented in
Table  3-1  and throughout this report are based on
reference  laboratory  TPH  results  for demonstration
samples and soil characterization completed  during the
demonstration, respectively. TPH concentration range and
soil type information obtained during the demonstration
was generally consistent with  the information  obtained
during the predemonstration investigation except for the
B-38  Area at Kelly AFB.   Additional information on
differences between demonstration and predemonstration
investigation  activities  and   results  is  presented in
Section 3.2.
                                                    16

-------
Table 3-1. Summary of Site Characteristics
Site
Navy Base
Ventura
County
Kelly Air
Force
Base
Petroleum
company
Sampling Area
Fuel Farm Area
Naval Exchange
Service Station
Area
Phytoremediation
Area
B-38 Area
Slop Fill Tank
Area
Contamination Type3
EDRO (weathered diesel with
carbon range from n-C10
through n-C40)
GRO and EDRO (fairly
weathered gasoline with carbon
range from n-C6 through n-C14)
EDRO (heavy lubricating oil
with carbon range from n-C14
through n-C40+)
GRO and EDRO (fresh
gasoline and diesel or
weathered gasoline and trace
amounts of lubricating oil with
carbon range from n-C6 through
n-C40)
GRO and EDRO (combination
of slightly weathered gasoline,
kerosene, JP-5, and diesel with
carbon range from n-C5 through
n-C32)
Approximate
Sampling Depth
Interval
(foot bgs)
Upper layerb
Lower layerb
7 to 8
8 to 9
9 to 10
10 to 11
1.5 to 2.5
23 to 25
25 to 27
2 to 4
4 to 6
6 to 8
8 to 10
TPH Concentration
Range (mg/kg)
44.1 to 93.7
8,090(015,000
28.1 to 280
144 to 2,570
61 7 to 3,030
9.56 to 293
1,130 to 2, 140
43.8 to 193
41 .5 to 69.4
6.16to3,300
37.1 to 3,960
43.9 to 1,210
52.4 to 554
Type of Soil
Medium-grained sand
Medium-grained sand
Silty sand
Sandy clay or silty sand and gravel
in upper depth interval and clayey
sand and gravel in deeper depth
interval
Silty clay with traces of sand and
gravel in deeper depth intervals
Notes:

bgs   = Below ground surface
mg/kg = Milligram per kilogram

a    The beginning or end point of the carbon range identified as "n-Cx" represents an alkane marker consisting of "x" carbon atoms on a gas
    chromatogram.
b    Because of soil conditions encountered in the Fuel Farm Area, the sampling depth intervals could not be accurately determined. Sample collection
    was initiated approximately 10 feet bgs, and attempts were made to collect 4-foot-long soil cores. This approach resulted in varying degrees of
    core tube penetration up to 17 feet bgs. At each location in the area, the sample cores were divided into two samples based on visual observations.
    The upper layer of the soil core, which consisted of yellowish-brown, medium-grained sand, made up one sample, and the lower layer of the soil
    core, which consisted of grayish-black, medium-grained sand and smelled of hydrocarbons, made up the second sample.
3.1     Navy Base Ventura County Site

The Navy BVC site in Port Hueneme, California, covers
about 1,600 acres along the south California coast.  Three
areas at the Navy BVC site were selected as sampling areas
for the demonstration: (1) the Fuel Farm Area (FFA),
(2) the Naval Exchange (NEX) Service Station Area, and
(3) the Phytoremediation Area (PRA).  These  areas are
briefly described below.

3.1.1   Fuel Farm Area

The FFA is a tank farm in the  southwest corner  of the
Navy BVC site.  The area contains five  tanks  and was
constructed to refuel ships and to supply heating fuel for
the Navy BVC site.  Tank No. 5114 along the south edge
of the FFA was used to store marine diesel.  After Tank
No. 5114  was deactivated in 1991, corroded pipelines
leading into and out of the tank leaked and contaminated
the surrounding soil with diesel.

The horizontal  area of contamination in the FFA  was
estimated  to be  about 20  feet wide  and  90 feet long.
Demonstration  samples  were collected within  several
inches  of the  three   predemonstration  investigation
sampling  locations  in  the FFA  using  a Geoprobe®.
Samples were collected at the three locations from east to
west and about 5 feet apart. During the demonstration, soil
in the area was found to generally consist of medium-
grained sand, and the soil cores contained two  distinct
                                                       17

-------
layers.  The upper layer consisted of yellowish-brown,
medium-grained sand with no hydrocarbon odor and TPH
concentrations ranging from 44.1 to 93.7 mg/kg; the upper
layer's  TPH   concentration  range   during   the
predemonstration investigation was 38 to 470 mg/kg. The
lower layer consisted of grayish-black, medium-grained
sand  with  a  strong  hydrocarbon  odor  and  TPH
concentrations ranging from 8,090 to 15,000 mg/kg; the
lower  layer's  TPH concentration  range  during  the
predemonstration   investigation   was   7,700   to
11,000 mg/kg.

Gas   chromatograms  from  the  predemonstration
investigation and the demonstration showed that FFA soil
samples contained (1) weathered diesel, (2) hydrocarbons
in the  n-C10 through  n-C28 carbon  range with  the
hydrocarbon   hump  maximizing  at  n-C17,   and
(3) hydrocarbons in the n-C12 through n-C40 carbon range
with the hydrocarbon hump maximizing at n-C20.

3.1.2  Naval Exchange Service Station Area

The NEX Service Station Area lies in the northeast portion
of the Navy BVC site.  About 11,000 gallons of regular
and unleaded gasoline was released from UST lines in this
area between September 1984 and March 1985. Although
the primary soil contaminant in this area is  gasoline,
EDRO is also of concern because (1) another spill north of
the area may have resulted in a  commingled plume of
gasoline  and diesel  and   (2) a  significant  portion  of
weathered gasoline is associated with EDRO.

The horizontal area of contamination in the NEX Service
Station Area was estimated to be about 450 feet wide and
750 feet long.  During the demonstration, samples were
collected  at the three  predemonstration investigation
sampling locations in the NEX Service Station Area from
south to north and about 60 feet apart using a Geoprobe®.
Soil  in the area was  found to generally  consist  of
(1)  brownish-black,  medium-grained  sand  in  the
uppermost depth interval and (2) grayish-black, medium-
grained sand in the three deeper depth intervals. Traces of
coarse sand were also present in the deepest depth interval.
Soil  samples collected  from the  area  had  a strong
hydrocarbon odor.   The  water table in the area was
encountered at about 9 feet below ground surface  (bgs).
During the demonstration, TPH  concentrations ranged
from 28.1 to  280 mg/kg  in the 7- to 8-foot bgs  depth
interval; 144 to  3,030 mg/kg in the 8- to 9- and 9- to 10-
foot bgs depth intervals; and 9.56 to 293 mg/kg in the 10-
to 11-foot bgs depth interval. During the predemonstration
investigation, the TPH concentrations in the (1) top two
depth intervals (7 to 8 and 8 to 9 feet bgs) ranged from 25
to 65  mg/kg and (2) bottom depth interval (10 to 11 feet
bgs) ranged from 24 to 300 mg/kg.

Gas  chromatograms   from  the   predemonstration
investigation and the demonstration showed that NEX
Service Station Area soil samples contained (1) fairly
weathered gasoline with a high aromatic  hydrocarbon
content and  (2) hydrocarbons in the n-C6 through n-C14
carbon range. Benzene, toluene, ethylbenzene, and xylene
(BTEX)   analytical   results  for   predemonstration
investigation samples from the 9-  to 10-foot bgs depth
interval at the middle  sampling  location revealed  a
concentration of 347 mg/kg; BTEX  made up  39 percent of
the total GRO and 27 percent of the TPH at this location.
During the predemonstration investigation, BTEX analyses
were  conducted at the  request of a few developers to
estimate the aromatic hydrocarbon content  of the GRO;
such  analyses  were not conducted for  demonstration
samples.

3.1.3  Phytoremediation Area

The PRA lies  north of the FFA and west  of the NEX
Service Station Area at the Navy  BVC site.  The PRA
consists of soil from a fuel tank removal project conducted
at the Naval Weapons Station in Seal Beach, California.
The  area  is contained  within concrete railings and is
60 feet wide, 100 feet long, and about 3 feet  deep.  It
consists of 12 cells of equal size (20 by 25 feet) that have
three different types of cover: (1) unvegetated cover, (2) a
grass  and legume mix, and (3) a native grass mix. There
are four replicate cells of each cover type.

In the PRA, demonstration samples were  collected from
the 1.5- to 2.5-foot bgs depth interval within several inches
of  the  six  predemonstration  investigation  sampling
locations  using a  split-core  sampler.    During  the
demonstration,  soil at four adjacent sampling locations
was found to generally consist of dark yellowish-brown,
silty sand with some clay and no hydrocarbon odor. Soil
at the two remaining adjacent sampling locations primarily
consisted of dark yellowish-brown, clayey sand with no
hydrocarbon odor, indicating the  absence of volatile
PHCs.  The TPH concentrations  in the  demonstration
samples ranged from  1,130 to 2,140 mg/kg; the TPH
concentrations   in  the  predemonstration  investigation
samples ranged from 1,500 to 2,700 mg/kg.
                                                   18

-------
Gas  chromatograms   from   the   predemonstration
investigation and the demonstration showed that PRA soil
samples  contained  (1)   heavy  lubricating  oil  and
(2) hydrocarbons in the n-C14 through n-C40+ carbon range
with the hydrocarbon hump maximizing at n-C32.

3.2    Kelly Air Force Base Site

The Kelly AFB site covers approximately 4,660 acres and
is  about 7 miles from the center of San Antonio, Texas.
One area at Kelly AFB, the B-38 Area, was selected as a
sampling area for the demonstration. The B-38 Area lies
along the east boundary of Kelly AFB and is part of an
active  UST farm  that serves  the government vehicle
refueling  station at the  base.   In December  1992,
subsurface soil contamination resulting from leaking diesel
and gasoline USTs and associated piping was discovered
in this area during UST removal and upgrading activities.

The B-38 Area was estimated to be about 150 square feet
in size. Based on discussions  with  site representatives,
predemonstration investigation  samples were collected in
the 13- to 17- and 29- to 30-foot bgs depth intervals at four
locations in the area using a Geoprobe®.  Based on
historical information, the water table in the area fluctuates
between  16   and  24  feet  bgs.    During  the
predemonstration investigation, soil in the area was found
to generally consist of (1) clayey silt in the upper depth
interval above the water table with a TPH concentration of
9 mg/kg and (2) sandy clay with significant gravel in the
deeper depth interval below the water table with TPH
concentrations ranging  from  9  to  18 mg/kg.    Gas
chromatograms from the predemonstration investigation
showed that B-38 Area soil samples contained (1) heavy
lubricating oil and (2) hydrocarbons in the n-C24 through
n-C30 carbon range.

Based  on the  low  TPH concentrations  and the type of
contamination detected  during the predemonstration
investigation  as  well   as   discussions  with  site
representatives who indicated that most of the contam-
ination in the B-38 Area can be  found at or near the water
table, demonstration samples were collected near the water
table.  During the demonstration, the water table  was
24 feet bgs.  Therefore, the demonstration samples were
collected in the 23- to 25- and 25- to 27-foot bgs depth
intervals  at  three locations in  the B-38  Area using a
Geoprobe®.  Air Force activities in  the area during the
demonstration  prevented the sampling  team   from
accessing the  fourth  location   sampled  during  the
predemonstration investigation.

During the demonstration, soil in the area was found to
generally consist of (1) sandy clay or silty sand and gravel
in the upper depth interval  with  a TPH concentration
between  43.8 and  193  mg/kg and (2) clayey sand and
gravel   in  the  deeper   depth   interval  with  TPH
concentrations between 41.5 and 69.4 mg/kg. Soil samples
collected in  the area had  little or no hydrocarbon odor.
Gas chromatograms from the demonstration showed that
B-38 Area soil samples contained either (1) fresh gasoline,
diesel, and hydrocarbons in the n-C6 through n-C25 carbon
range with the hydrocarbon hump maximizing at n-C17;
(2) weathered gasoline with trace amounts of lubricating
oil and hydrocarbons in the  n-C6 through n-C30 carbon
range  with   a   hydrocarbon hump representing  the
lubricating oil between n-C20 and n-C30; or (3) weathered
gasoline   with   trace  amounts   of  lubricating   oil
and hydrocarbons in the n-C6 through n-C40 carbon range
with a hydrocarbon hump representing the lubricating oil
maximizing  at n-C31.

3.3    Petroleum Company Site

One area at the PC site in north-central Indiana, the Slop
Fill Tank (SFT) Area, was selected as a sampling  area for
the demonstration.  The SFT Area lies in the west-central
portion of the PC site and is part  of an active fuel tank
farm.  Although the primary soil contaminant in this area
is gasoline, EDRO is also of concern because of a heating
oil release that occurred north of the area.

The SFT Area was estimated to be 20 feet long and 20 feet
wide.  In this area, demonstration samples  were collected
from 2 to 10 feet  bgs  at 2-foot depth  intervals within
several inches of the five  predemonstration investigation
sampling  locations using a  Geoprobe®.   Four  of the
sampling locations  were spaced 15 feet apart to form the
corners of a square, and the fifth sampling location was at
the center of the square.  During the demonstration, soil in
the area was found  to generally  consist of  brown to
brownish-gray, silty clay with traces of sand and gravel in
the deeper depth intervals.  Demonstration soil samples
collected in  the area had  little or  no hydrocarbon odor.
During the demonstration, soil in  the three upper depth
intervals had TPH concentrations ranging from  6.16 to
3,960 mg/kg, and soil in  the deepest depth interval had
TPH concentrations  ranging from 52.4 to 554  mg/kg.
During the predemonstration investigation,  soils in the
                                                    19

-------
three upper depth intervals and the deepest depth interval
had TPH concentrations ranging from 27 to 1,300 mg/kg
and from 49 to 260 mg/kg, respectively.

Gas   chromatograms  from   the   predemonstration
investigation and the demonstration showed that SFT Area
soil samples contained (1) slightly weathered gasoline,
kerosene, JP-5, and diesel  and (2) hydrocarbons in the
n-C5 through n-C20 carbon range. There was also evidence
of  an  unidentified   petroleum  product   containing
hydrocarbons in the n-C24  through n-C32 carbon range.
BTEX  analytical   results  for  predemonstration
investigation  samples  from the  deepest depth interval
revealed concentrations of 26, 197, and 67 mg/kg at the
northwest,  center,  and southwest sampling  locations,
respectively.  At the northwest location, BTEX made up
13 percent of the total GRO and 5 percent of the TPH. At
the center location, BTEX made up 16 percent of the total
GRO and 7 percent of the TPH. At the southwest location,
BTEX made up  23  percent of the total  GRO  and
18 percent of  the  TPH.   BTEX  analyses  were  not
conducted for demonstration samples.
                                                   20

-------
                                               Chapter 4
                                     Demonstration Approach
This chapter presents the objectives (Section 4.1), design
(Section 4.2), and sample preparation and management
procedures (Section 4.3) for the demonstration.

4.1    Demonstration Objectives

The primary goal of the SITE MMT Program is to develop
reliable performance  and cost data on innovative, field-
ready technologies. A SITE demonstration must provide
detailed and reliable  performance and cost data so that
potential technology users have adequate information to
make  sound  judgments  regarding  an   innovative
technology' s applicability to a specific site and to compare
the technology to conventional technologies.

The  demonstration  had  both  primary  and  secondary
objectives.  Primary objectives were  critical  to  the
technology evaluation and required the use of quantitative
results to  draw  conclusions regarding  a  technology's
performance.    Secondary  objectives   pertained  to
information that was useful but did not necessarily require
the use of quantitative  results  to  draw conclusions
regarding a technology's performance.

The  primary objectives  for the  demonstration of the
individual field measurement devices were as follows:

PI. Determine the MDL

P2. Evaluate  the  accuracy and  precision  of TPH
    measurement for a  variety  of contaminated soil
    samples

P3. Evaluate   the  effect  of  interferents  on  TPH
    measurement

P4. Evaluate the  effect of soil moisture content on TPH
    measurement
P5.  Measure the time required for TPH measurement

P6.  Estimate costs associated with TPH measurement

The secondary objectives for the demonstration of the
individual field measurement devices were as follows:

S1.  Document the skills and training required to properly
    operate the device

S2.  Document health and safety concerns associated with
    operating the device

S3.  Document the portability of the device

S4.  Evaluate  the durability of the device based on  its
    materials of construction and engineering design

S5.  Documentthe availability of the device and associated
    spare parts

The objectives forthe demonstration were developed based
on input from MMT Program stakeholders, general user
expectations of field measurement devices, characteristics
of the demonstration areas, the time available to complete
the   demonstration,  and  device  capabilities  that  the
developers intended to highlight.

4.2     Demonstration Design

A predemonstration sampling and analysis investigation
was conducted to assess existing conditions and confirm
available  information  on  physical  and   chemical
characteristics of soil in each demonstration area. Based
on information from the predemonstration investigation as
well as  available historical data, a demonstration design
was developed to address the demonstration objectives.
Input regarding the demonstration design was obtained
                                                    21

-------
from  the  developers  and   demonstration  site
representatives. The demonstration design is summarized
below.

The demonstration involved analysis of soil environmental
samples, soil performance evaluation (PE) samples, and
liquid PE samples.   The environmental samples were
collected from three  contaminated  sites, and the  PE
samples  were  obtained from a  commercial  provider,
Environmental Resource Associates  (ERA)  in Arvada,
Colorado. Collectively, the environmental and PE samples
provided the different matrix types and the different levels
and types of PHC contamination needed to perform  a
comprehensive demonstration.

The  environmental  samples  were  soil core samples
collected from the demonstration areas at the Navy BVC,
Kelly AFB, and PC sites described in Chapter 3. The soil
core samples collected at the Kelly AFB and PC sites were
shipped to the Navy  BVC site 5 days prior to the start of
the field analysis activities.   Each  soil core sample
collected from a specific depth interval  at a particular
sampling location in a given area was homogenized and
placed in individual sample containers. Soil samples were
then provided to the developers and reference laboratory.
In addition, the PE samples were obtained from ERA for
distribution to  the developers and reference laboratory.
Field analysis of all  environmental and PE samples was
conducted near the PRA at the Navy BVC site.

The  field measurement devices  were evaluated based
primarily on how they compared with the  reference
method  selected for the demonstration. PE samples were
used to  verify that reference  method performance was
acceptable. However, for the comparison with the device
results,  the reference method  results were not adjusted
based on the recoveries observed during analysis of the PE
samples.

The  sample collection and homogenization procedures
may have resulted in GRO losses of up to one order of
magnitude in environmental samples.  Despite any such
losses, the homogenized samples were expected to contain
sufficient levels of GRO to allow demonstration obj ectives
to be achieved.   Moreover,  the  environmental sample
collection and  homogenization procedures implemented
during the demonstration ensured that the developers and
reference laboratory received the same sample material for
analysis, which was  required   to  allow meaningful
comparisons of field measurement device and reference
method results.

To facilitate effective use of available information on both
the   environmental  and   PE  samples   during   the
demonstration, the developers and  reference laboratory
were  informed  of  (1) whether each sample was an
environmental or PE sample, (2) the area where each
environmental  sample  was  collected,  and  (3)  the
contamination type and  concentration  range of each
sample. This information was included in  each sample
identification number.  Each  sample  was identified as
having a low (less  than  100  mg/kg), medium (100 to
1,000 mg/kg), or high (greater than 1,000 mg/kg) TPH
concentration range. The concentration ranges were based
primarily on predemonstration investigation results or the
amount of weathered gasoline  or diesel added during PE
sample preparation. The concentration ranges were meant
to be used only as a guide by the developers and reference
laboratory.  The gasoline used for PE sample preparation
was 5 0 percent weathered; the weathering was achieved by
bubbling nitrogen gas into a known volume of gasoline
until the volume was reduced by 50 percent.  Some PE
samples also contained interferents specifically added to
evaluate the effect of interferents on TPH measurement.
The type of contamination and  expected TPH concen-
tration ranges were identified;  however,  the specific
compounds used as interferents were not identified.  All
PE samples were prepared in triplicate as separate, blind
samples.

During the  demonstration, a Horiba field technician
operated the OCMA-350, and EPA representatives made
observations to  evaluate the device.  All the developers
were  given  the  opportunity to  choose  not to analyze
samples collected in a particular area or a particular class
of samples, depending on the  intended uses of their
devices.  Horiba chose to analyze all the demonstration
samples.

Details of the approach used to address the  primary  and
secondary objectives for the demonstration are presented
in Sections 4.2.1 and 4.2.2, respectively.

4.2.1  Approach for A ddressing Primary
       Objectives

This section presents the approach used to address each
primary objective.
                                                   22

-------
Primary Objective PI: Method Detection Limit

To determine the MDL for each field measurement device,
low-concentration-range  soil  PE  samples  containing
weathered gasoline or diesel were to be analyzed.  The
low-range PE samples were prepared using Freon  113,
which facilitated preparation of homogenous samples.  The
target concentrations of the PE samples were set to meet
the  following criteria: (1) at  the minimum  acceptable
recoveries set by ERA, the samples contained measurable
TPH concentrations, and (2) when feasible, the sample
TPH  concentrations  were  generally  between  1  and
10 times the MDLs claimed by the developers and the
reference laboratory, as recommended by 40  Code of
Federal  Regulations (CFR)  Part 136,   Appendix B,
Revision 1.1.1.   Horiba  and  the reference  laboratory
analyzed seven weathered gasoline and seven diesel PE
samples to statistically determine the MDLs for GRO and
EDRO soil samples. However, during the preparation of
low-range weathered gasoline  PE samples,  significant
volatilization of PHCs occurred because of the matrix used
for  preparing these samples.  Because of the problems
associated  with  preparation  of low-range  weathered
gasoline PE samples, the results for these samples could
not be used to determine the MDLs.

Primary Objective P2: Accuracy and Precision

To  estimate the  accuracy and precision  of each field
measurement device, both environmental and PE samples
were analyzed.  The evaluation of analytical accuracy was
based on the assumption that a field measurement device
may  be  used to  (1) determine  whether the  TPH
concentration in a given area exceeds an action level or
(2) perform a preliminary characterization of soil  in a
given area. To evaluate whether the TPH concentration in
a soil sample exceeded an action level, the developers and
reference laboratory were asked to determine whether TPH
concentrations in a given area or PE sample type exceeded
the action levels listed in Table 4-1.  The action levels
chosen  for environmental samples were based on the
predemonstration investigation analytical results and state
action levels.  The action levels chosen for the PE samples
were based in part on the  ERA acceptance limits for PE
samples; therefore, each PE sample was expected to have
at least the TPH concentration  indicated in Table  4-1.
However, because  of the problems  associated  with
preparation of the  low-concentration-range  weathered
gasoline PE samples, the results  for these samples could
not be used to address primary objective P2.

In addition, neat (liquid)  samples of weathered gasoline
and diesel were analyzed by the developers and reference
laboratory to  evaluate accuracy and precision. Because
extraction of the neat samples  was  not necessary, the
results for these samples provided accuracy and precision
information strictly associated with the analyses and were
not affected by extraction procedures.
Table 4-1. Action Levels Used to Evaluate Analytical Accuracy
Site
Navy Base Ventura County
Kelly Air Force Base
Petroleum company
Area
Fuel Farm Area
NEX Service Station Area
Phytoremediation Area
B-38 Area
Slop Fill Tank Area
Performance evaluation samples (GRO analysis)
Performance evaluation samples (EDRO analysis)
Typical TPH Concentration Range3
Low and high
Low to high
High
Low
Medium
Medium
High
Low
Medium
High
Action Level (mg/kg)
100
50
1,500
100
500
200
2,000
15
200
2,000
Notes:

mg/kg = Milligram per kilogram

a    The typical TPH concentration ranges shown cover all the depth intervals in each area. Table 4-2 shows the depth intervals that were sampled
    in each area and the typical TPH concentration range for each depth interval. The action level for each area was used as the basis for evaluating
    sample analytical results regardless of the typical TPH concentration ranges for the various depth intervals.
                                                     23

-------
Sample  TPH   results  obtained  using  each   field
measurement device  and the  reference method  were
compared to the action levels presented in Table 4-1 in
order to determine whether sample TPH concentrations
were above the action levels.  The results obtained using
the device and reference  method were compared to
determine how many times the device's results agreed with
those of the reference method for a particular area or
sample type. In addition, the ratio of the TPH results of a
given device to the TPH results of the reference method
was calculated.  The ratio was used to develop a frequency
distribution in order to determine how many of the device
and reference  method results  were within 30  percent,
within 50 percent, and outside the 50 percent window.

To complete a preliminary characterization of soil in a
given area using a field measurement device, the user may
have to demonstrate to a regulatory agency that (1) no
statistically significant difference exists between the results
of the laboratory method selected for the project (the
reference method) and the device results, indicating that
the device may be used as a substitute for the laboratory
method, or (2) a consistent correlation exists between the
device and laboratory method results, indicating that the
device results  can  be  adjusted  using  the  established
correlation.

To evaluate whether a statistically significant difference
existed between a given field measurement device and the
reference method results, a two-tailed, paired Student's li-
test was performed.  To  determine whether a consistent
correlation existed between the TPH results of a given field
measurement device and the reference method, a linear
regression was performed to estimate the square of the
correlation coefficient (R2), the slope, and the intercept of
each regression equation.  Separate regression equations
were developed for each demonstration area and for the PE
samples that did not contain interferents. The reliability of
the regression equations was tested using the F-test; the
regression equation probability derived from the F-test was
used to evaluate whether the correlation between the TPH
results of the device and the reference method occurred
merely by chance.

To evaluate analytical precision,  one set of blind field
triplicate environmental samples was collected from each
depth interval at one location in each demonstration area
except the B-38 Area, where  site conditions  allowed
collection of triplicates in the top depth interval  only.
Blind triplicate low-, medium-, and high-concentration-
range PE samples were also used to evaluate analytical
precision because TPH concentrations in environmental
samples collected during the demonstration sometimes
differed from the analytical results for predemonstration
investigation samples.  The low- and medium-range PE
samples were prepared using Freon 113 as a carrier, which
facilitated preparation of homogenous samples.

Additional information regarding analytical precision was
collected  by  having  the  developers  and  reference
laboratory analyze extract duplicates. Extract duplicates
were  prepared by extracting  a soil sample once  and
collecting two aliquots of the extract. For environmental
samples,  one sample  from  each depth  interval was
designated  as  an  extract  duplicate.    Each  sample
designated as an extract duplicate was collected from a
location  where  field triplicates  were  collected.   To
evaluate  a given field measurement device's  ability to
precisely measure TPH, the relative standard  deviation
(RSD) of the device and reference method TPH results for
triplicate samples was calculated. In addition, to evaluate
the analytical precision of the device and  reference
method,  the relative  percent  difference  (RPD) was
calculated using the TPH results for extract duplicates.

Primary Objective P3: Effect of Interferents

To evaluate the effect  of interferents on each  field
measurement device's ability to accurately measure TPH,
high-concentration-range soil PE samples containing
weathered gasoline or diesel with or without an interferent
were analyzed.  As explained in Chapter 1, the definition
of TPH is quite variable.  For the purposes of addressing
primary objective P3, the term "interferent" is used in a
broad sense and is applied to both PHC and  non-PHC
compounds.   The six different interferents  evaluated
during the demonstration were MTBE; tetrachloroethene
(PCE); Stoddard solvent; turpentine (an alpha and beta
pinene mixture); 1,2,4-trichlorobenzene; and humic acid.
The boiling points and vapor pressures of (1) MTBE and
PCE are similar to those of GRO; (2) Stoddard solvent and
turpentine are similar to those of GRO and EDRO; and
(3) 1,2,4-trichlorobenzene and humic acid are similar to
those of EDRO.  The solubility, availability, and cost of
the interferents were also considered during interferent
selection.  Specific reasons for the selection of the six
interferents are presented below.

•   MTBE  is an oxygenated  gasoline  additive  that is
    detected  in  the   GRO   analysis  during   TPH
    measurement using a GC.
                                                     24

-------
•   PCE is not a petroleum product but is detected in the
    GRO analysis during TPH measurement using a GC.
    PCE may also be viewed  as a typical halogenated
    solvent that may be present in some environmental
    samples.

•   Stoddard solvent is  an aliphatic naphtha compound
    with a carbon range of n-C8 through n-C14 and is partly
    detected in both the GRO and EDRO analyses during
    TPH measurement using a GC.

    Turpentine is not a petroleum product but has a carbon
    range  of n-C9 through n-C15 and is partly detected in
    both the GRO and EDRO  analyses  during TPH
    measurement using  a GC.  Turpentine may also be
    viewed as a substance that behaves similarly  to a
    typical naturally occurring  oil or grease during TPH
    measurement using a GC.

•   The  compound  1,2,4-trichlorobenzene  is  not  a
    petroleum product  but is detected in  the EDRO
    analysis.  This compound may also be viewed  as a
    typical halogenated  semivolatile organic compound
    that behaves similarly to a chlorinated pesticide or
    PCB during TPH measurement using a GC.

•   Humic  acid  is  a  hydrocarbon  mixture  that  is
    representative of naturally occurring organic carbon in
    soil and  was suspected to be detected during EDRO
    analysis.

Based  on the  principles of operation  of the field
measurement devices, several  of the interferents were
suspected to be detected by the  devices.

The PE samples containing MTBE and PCE were not
prepared  with diesel and the PE samples  containing
1,2,4-trichlorobenzene and humic acid were not prepared
with weathered gasoline because these interferents were
not expected to impact the analyses and because practical
difficulties such as solubility constraints were associated
with preparation of such samples.

Appropriate  control  samples were also prepared  and
analyzed to address primary objective P3. These samples
included processed garden soil, processed garden soil and
weathered gasoline, processed garden soil and diesel, and
processed garden soil and humic acid samples.  Because of
solubility constraints, control samples containing MTBE;
PCE;   Stoddard  solvent;    turpentine;   or   1,2,4-
trichlorobenzene could not be prepared.  Instead,  neat
(liquid) samples of these interferents were prepared and
used as quasi-control samples to evaluate the effect of
each interferent on the field measurement device and
reference method results.  Each PE sample was prepared
in triplicate and submitted to the developers and reference
laboratory as blind triplicate samples.

To evaluate the effects of interferents on a given field
measurement device's ability to accurately measure TPH
under  primary  objective P3, the means and  standard
deviations of the TPH results for triplicate PE samples
were calculated. The mean for each group of samples was
qualitatively evaluated to determine whether the  data
showed any trend—that is, whether an increase in the
interferent  concentration resulted  in  an increase or
decrease in the measured TPH concentration. A one-way
analysis of variance was performed to determine whether
the group means were the same or different.

Primary Objective P4: Effect of Soil Moisture Content

To evaluate the effect of soil moisture  content, high-
concentration-range soil PE samples containing weathered
gasoline or diesel were analyzed.  PE samples containing
weathered gasoline were prepared at two moisture levels:
9 percent moisture  and 16 percent moisture.  PE samples
containing  diesel  were also prepared at two moisture
levels:  negligible  moisture  (less than 1 percent)  and
9 percent moisture. All the moisture levels were selected
based  on  the  constraints  associated  with  sample
preparation. For example, 9 percent moisture represents
the minimum moisture level  for containerizing samples in
EnCores and 16 percent moisture represents the saturation
level of the soil used to prepare PE  samples.  Diesel
samples with  negligible moisture could be prepared
because they did not require EnCores for containerization;
based on vapor pressure data for diesel and weathered
gasoline, 4-ounce jars were  considered to  be appropriate
for containerizing diesel samples but not for containerizing
weathered  gasoline  samples.   Each  PE sample  was
prepared in triplicate.

To measure the effect of soil moisture content on a given
field measurement  device's ability to accurately measure
TPH under primary objective P4, the means and standard
deviations of the TPH results for triplicate PE samples
containing weathered gasoline and diesel at two moisture
levels  were  calculated.    A two-tailed,  two-sample
Student's t-test was performed to determine whether the
device  and reference method results were impacted by
moisture—that  is,  to determine whether  an increase in
                                                    25

-------
moisture resulted in an increase or decrease in the TPH
concentrations measured.

Primary Objective P5: Time Required for TPH
Measurement

The sample throughput (the number of TPH measurements
per  unit  of  time)  was  determined  for  each  field
measurement device by measuring the time  required for
each activity associated with TPH measurement, including
device setup, sample extraction, sample analysis, and data
package preparation.  The EPA provided each developer
with  investigative samples  stored  in  coolers.    The
developer unpacked the coolers and checked the chain-of-
custody forms to verify that it had received the correct
samples. Time measurement began when the developer
began to set up its device.   The total time required to
complete  analysis of all  investigative  samples  was
recorded.  Analysis was considered to be complete and
time measurement stopped when the developer provided
the EPA with a summary table of results, a run log, and
any supplementary information that the developer chose.
The summary table listed all samples analyzed and their
respective TPH concentrations.

For the reference  laboratory, the total  analytical time
began to be measured when the laboratory received all the
investigative samples, and time measurement continued
until the EPA representatives received a complete data
package from the laboratory.

Primary Objective P6: Costs Associated with TPH
Measurement

To estimate  the costs associated with TPH measurement
for each field measurement device, the following five cost
categories were  identified: capital equipment, supplies,
support equipment, labor, and investigation-derived waste
(IDW) disposal. Chapter 8  of this ITVR discusses the
costs  estimated for the OCMA-350 based on these  cost
categories.

Table 4-2 summarizes the demonstration approach used to
address the primary objectives and includes demonstration
area characteristics, approximate sampling depth intervals,
and  the  rationale  for the analyses performed  by the
reference laboratory.
4.2.2  Approach for Addressing Secondary
       Objectives

Secondary  objectives were  addressed based  on  field
observations made during the demonstration. Specifically,
EPA representatives observed TPH measurement activities
and documented them in a field logbook. Each developer
was given the opportunity to review the field logbook at
the end of each day of the demonstration.  The approach
used to address each secondary objective for each field
measurement device is discussed below.

•  The skills and training required  for proper device
   operation (secondary objective SI) were evaluated by
   observing and noting the skills required to operate the
   device  and prepare the  data  package during the
   demonstration  and by discussing necessary  user
   training with developer personnel.

   Health  and safety concerns associated with device
   operation (secondary objective S2) were evaluated by
   observing  and noting possible health and safety
   concerns during the demonstration, such as the types
   of hazardous  substances handled  by  developer
   personnel during analysis, the number of times that
   hazardous  substances  were  transferred from  one
   container to another during the analytical procedure,
   and direct exposure  of developer  personnel to
   hazardous substances.

•  The portability of the device (secondary objective S3)
   was evaluated by observing and noting the weight and
   size of  the device and additional equipment required
   for TPH measurement as well as how easily the device
   was set up for use during the demonstration.

•  The durability of the device (secondary objective S4)
   was evaluated by noting the materials of construction
   of the device and additional equipment required for
   TPH measurement. In addition, EPA representatives
   noted likely device failures or repairs that may be
   necessary  during  extended  use of the  device.
   Downtime required to make device repairs during the
   demonstration was also noted.

   The availability of the device and associated spare
   parts (secondary  objective  S5) was evaluated  by
   discussing the availability of replacement devices with
   developer personnel and determining whether spare
   parts were available in retail stores or only from the
                                                    26

-------
u

o
Q.
Q.
o

Rationale for Analyses
by Reference Laboratory


_ c
X 0
CL = ^

_ •£ D)
o oj ra
i^1 O
K 0




,
c
O





o



E
&
 C







-o
1
T3
, c
'ro o
LL t







T3
1
T3
-B
68





^
D1
x


E
o
^
—
CO

and EDRO because samples
ined PHCs in both gasoline and
OS
0 8

CO
, ^~
"o .E
Q >
ro 2
CO D)






(5
m*
—
(5
1

LL
LL






o ^
00 O) T- o
o o o •"
•" •" •" O
1^ 00 O) T-


8 §
X > = ra
LLJ 0) JS 2
Z CO CO <



LO
csi
O
LO


ce
CL


> 0
ra >
Z CD




LO
CN
O



oo ro
CO (U
CO <


~OJ LL
^ <


o*
C
D)
O
^
cJ
c
E
o
d)
D)
C
2
c
o
J2
ro
o
±±

aj
CL

2
D)
T3
c
ra "ro
E £
(0 ^j
w g
^ "*"'
ro g-
O T3




E
, Q-
:^ (U
co -a








CN

LO
CN












O
f CD 00 T-
£ £ £ £
CN f CD 00


t s
CO <


0
CL

Rationale for Analyses
by Reference Laboratory


f-
X 0
CL =
1— ro -=
ra | o>

K 0




-B
68





o








T3
-B
68


m

E
^ ^
0) D)








^
±±
CO


CN
CL
"o?
Q.
E
CO
LIJ
n
1
c

-------
O
O
u
re

2
Q.
Q.
c
o
^
re

to
c
o

i)
Q
_

.Q
re

Rationale for Analyses
by Reference Laboratory
_ c
X 0
CL =
1- 2 "a.
ro c c

g-c -
li" O
•~ o




 (!)
•5 83
.0, 2
°l










X
^r
ra
E oS
ls|
<" c\T
£ ^S
ro LLJ S
-i O -
<, CL SjL











"3 "
^ S>
D) =t
I" "
° S 8
^8^
LU -Q ^.
i 	 ro QJ
^ T3 c
^ T3 ^
~o -2 oj
C CO Q.
ro >-
||r
S^> °.
F '""
Sof
-a o a)
QJ - t
jjj«0
t5 LLJ '*-
^ O in
^ CL ^










Q)

(D
b
!_
^^
D)
1)
E
o
in
(D
co"
~~^
"c
(D
"o ^>
° =S
"O ^
ro E
T3 0
•o in
£ 00
w co"
' ¥
ro ^§
"S c
S &
b 3.












"3

^>
8
CM
oo"

^^ ^-~.
'C D)

-> ^)
8o
ro CD
o ^<
W (D
ro oj
"ai 2-
«) 3
0) *-
b o


"3

E

it
^§
8) 
O)
T3
'o
ro
o
E
X











c
O m"
processed garden soil during the
predemonstration investigation was
considered to be insignificant evaluati
of the effect of humic acid interference
which occurs in the diesel range.

















O)
E
8
in
oT
T3
'o
ro
o
E
X











CO
CL
"oT
Q.
E
81
LLJ
* —
1
C
(D
•g
ro
D)
T3
83
$
O
CL





















































































                                                                    28

-------
•D
0)
O
O
u
re

2
Q.
Q.
C
o
^
re

to
C
o


0)
Q
_

.Q
re

Rationale for Analyses
by Reference Laboratory
_ C
X 0
CL =
1 — 2 **to
"CD 1 g5
,§0 "
"~ O







CD
CL
1—
C

.^
to
C
E
"c
o
O






1
to
"o
2
CD
6
'5
CO



g ID
=5 «
CD_ 2
O ^
<











X
-»-•
2
CD
CL
CD
CO
'o
GRO and EDRO because weathered
gasoline contains significant amounts
PHCs in both gasoline and diesel
ranges

Only EDRO because diesel does not
contain PHCs in gasoline range
o
Only GRO because MTBE and PCE d
not interfere with EDRO analysis





.C
X















CD
C
8
CD
O)
T3
CD
£
to
-32
§



















CD
in
CD
b



















LU
m

2


CD
_Q
CD
g
Z CD"




















LU

CL

GRO and EDRO because Stoddard
solvent contains PHCs in both gasolin
and diesel ranges





O)
X
















"c
CD
"o
CO
•g
CD
T3
T3

CO

GRO and EDRO because turpentine
interferes with both analyses
.c
Only EDRO because 1,2,4-
trichlorobenzene does not interfere wil
GRO analysis


CD
_Q
o
— Q.
0 Q.
Z CD


















CD
C
"c
8.
^
i-













CD
CD
N
C
CD
O
o

O
4
CM






0)
CD
O
"o.
CL
CD
"o
z
CD
^
^3
o
CL 9.



LU
CL

g-
CD
CD
C

 0)
CO CD
CO >


"D i—
oj oj
(0 Q"
Q.O)
2 -D
O C
co ra
^) ^1
c" ^

Is.
"CD ^~
CO C
CD (D
b £






T3
1
±±
CO




CL
"a?
Q.
E
ro
(/)
LJJ
•— '
1
C
CD
•g
CD
D)
T3
CO
O
g
CL

















































CO
<0
"o
valuation
CD
CD
O
C
CD
E
1
CD
CL
II

LU
CL







val Exchange
CD
II

X
LU
z











CD
E
LL.
"CD
II


LL.
LL.









CD
CO
m
O
LL.
II
m

-ocarbon
on Area
-Q -*--
-C ^
ll

ll.
"CD .c
CL CL
II II
0 <
x ce
CL CL






m ^
troleum comp;
Irachloroether
CD CD
CL 1-
ii M

LU
o o
CL CL






E CD
2£
D) CD
0 _

%.*?
roi-
ii ii

D)LU
"3)1-









3 °
co O
T3 CD
= Below gro
= Base Ven
CO °
E'm


C
CD
1=
Ll_
CL
O
CO
II
l_
CO














































c/i
CD
.^
CD_
O
£•
CD
T3
8
QJ
CO
G)
£
T3
CD
CD
CL
C
bjectives P5 a
o
£•
CD

CL
in
to
^
-Q
CD
-»-•
T3
1
CD
CD


o
s demonstratil
D)
C
3
T3
T3
CD
•5
T3
C
8
(/)
0)
(/)
"ro
c
ro
GJ
o
nsof all sam|
Field observatiol

ro
rations of less than
^
CD
O
C
o
o
X
CL
1-
0)
CD
O
C
CO
CD
D)
C
2
.C
gi
T3
C
CD
E
^
CD
E
3
CD
O
'o.
CD
K

d
—
2
to
C
o
E
CD
"O
CD
£
CO


CO
laboratory re;
CD
o
C
CD


C
g
Q
"o
O
CD
CL
E
CO
CD
^
1
CD
T3
"c?"
"S
8
CD
"o
C
2
o
o
CD
£
.2;
C
f~
"Q.
CD
T3
D)

"o.
E

CD



demonstration
CD
D)
C
T3
LL.
1 1
CD
£
^
T3
CD
D)"a 
-------
    developer. In addition, the availability of spare parts
    required during the demonstration was noted.

Field observations of the  analyses of all the samples
described in Table 4-2 were used to address the secondary
objectives for the demonstration.

4.3    Sample Preparation and Management

This section presents sample preparation and management
procedures used during the demonstration. Specifically,
this  section  describes how  samples  were  collected,
containerized,  labeled, stored,  and shipped  during  the
demonstration.   Additional  details  about  the sample
preparation and management procedures are presented in
the demonstration plan (EPA 2000).

4.3.1  Sample Preparation

The sample preparation procedures for both environmental
and PE samples are described below.

Environmental Samples

For  the  demonstration,  environmental  samples  were
collected  in  the   areas  that  were   used  for   the
predemonstration investigation: (1) the FFA, NEX Service
Station Area, and PRA at the Navy BVC site; (2) the B-38
Area at the Kelly AFB site; and (3) the SFT Area at the PC
site.  Samples were collected in all areas except the PRA
using a Geoprobe®; in the PRA, samples  were collected
using a Split Core Sampler.

The   liners  containing  environmental   samples  were
transported to the sample management trailer at the Navy
BVC site, where the liners were cut open  longitudinally.
A geologist then profiled the samples  based on soil
characteristics to determine where the soil  cores had to be
sectioned.  The soil characterization performed for each
demonstration area is summarized in Chapter 3.

Each core sample section was  then transferred to a
stainless-steel bowl. The presence of any unrepresentative
material such as sticks, roots, and stones  was  noted in a
field logbook, and such material was removed to the extent
possible using gloved hands.  Any lump of clay in the
sample that was greater than about 1/8  inch in diameter
was   crushed   between  gloved   fingers  before
homogenization.  Each soil sample was homogenized by
stirring it  for at  least 2  minutes  using a stainless-steel
spoon  or gloved hands until the  sample  was visibly
homogeneous.    During  or  immediately  following
homogenization, any free  water was poured from the
stainless-steel  bowl containing  the soil  sample into  a
container designated for IDW.  During the demonstration,
the field sampling team used only nitrile gloves to avoid
the possibility of phthalate contamination from handling
samples  with plastic gloves.  Such contamination had
occurred during the predemonstration investigation.

After sample homogenization, the samples were placed in
(1) EnCores of approximately 5-gram capacity for GRO
analysis; (2) 4-ounce, glass jars provided by the reference
laboratory for EDRO and percent moisture analyses; and
(3) EnCores of approximately 25-gram capacity for TPH
analysis.  Using a quartering technique, each sample
container was filled by alternately spooning soil from one
quadrant of the mixing bowl and then from the opposite
quadrant until the container was full. The 4-ounce, glass
jars were filled after all the EnCores for a given sample
had been filled. After a sample container was filled, it was
immediately  closed  to  minimize  volatilization  of
contaminants.  To minimize the time required for sample
homogenization and filling of sample containers, these
activities  were  simultaneously  conducted  by  four
personnel.

Because of the large number of containers  being filled,
some time elapsed between the filling of the first EnCore
and the  filling of the last.  An attempt was made to
eliminate any bias by alternating between filling EnCores
for the developers and filling EnCores for the reference
laboratory.  Table  4-3  summarizes the  demonstration
sampling depth intervals, numbers of environmental and
QA/QC samples collected, and numbers of environmental
sample analyses associated with the demonstration of the
OCMA-350.

Performance Evaluation Samples

All PE samples for the  demonstration were prepared by
ERA and shipped to the sample management trailer at the
Navy BVC site. PE samples consisted of both soil samples
and liquid samples. ERA prepared soil PE samples using
two soil matrixes: Ottawa sand and processed garden soil
(silty sand).

To prepare the soil PE samples, ERA spiked the required
volume of soil based on the number of PE samples and the
quantity of soil per PE sample requested.  ERA then
                                                   30

-------
Table 4-3. Environmental Samples
Site
Navy
BVC
Kelly
AFB
PC
Area
FFA
NEX
Service
Station
Area
PRA
B-38
Area
SFT
Area
Depth
Interval
(foot bgs)
Upper layer
Lower layer
7 to 8
8 to 9
9 to 10
1 0 to 1 1
1.5 to 2.5
23 to 25
25 to 27
2 to 4
4 to 6
6 to 8
8 to 10
Number of
Sampling
Locations
3
3
3
3
3
3
6 (4 vegetated
and
2 unvegetated)
3
3
5
5
5
5
Total
Total Number of Samples,
Including Field Triplicates,
to Horiba and Reference
Laboratory3
5
5
5
5
5
5
8
5
3
7
7
7
7
74
Number of
MS/MSDb
Pairs
1
1
1
1
1
1
1
1
1
1
1
1
1
13
Number of
Extract
Duplicates0
1
1
1
1
1
1
1
1
1
1
1
1
1
13
Number of
TPH Analyses
by Horiba
6
5
6
5
5
6
9
6
4
8
8
8
7
83
Number of Analyses
by Reference
Laboratory11
GRO
0
0
8
8
8
8
0
8
6
10
10
10
10
86
EDRO
8
8
8
8
8
8
11
8
6
10
10
10
10
113
Notes:

AFB   =  Air Force Base
bgs   =  Below ground surface
BVC   =  Base Ventura County
FFA     =  Fuel Farm Area
MS/MSD =  Matrix spike and matrix spike duplicate
NEX     =  Naval Exchange
PC   =  Petroleum company
PRA  =  Phytoremediation Area
SFT  =  Slop Fill Tank
    Field triplicates were collected at a frequency of one per depth interval in each sampling area except the B-38 Area. Because of conditions in the
    B-38 Area, triplicates were collected in the top depth interval only. Three separate, blind samples were prepared for each field triplicate.
    MS/MSD samples were collected at a frequency of one per depth interval in each sampling area for analysis by the reference laboratory. MS/MSD
    samples were not analyzed by Horiba.
    Because of site conditions, Horiba did not analyze extract duplicates for the lower layer in the FFA, the 8- to 9- and the 9- to 10-foot bgs depth
    intervals in the NEX Service Station Area, and the 8-to 10-foot bgs depth interval in the SFT Area. Therefore, Horiba analyzed only nine extract
    duplicates.
    All environmental samples were also analyzed for moisture content by the reference laboratory.
homogenized the soil by manually mixing it.  ERA used
weathered gasoline or diesel as the spiking material, and
spiking was done at three levels to depict the three TPH
concentration  ranges:   low,  medium,  and  high.   A
low-range sample was  spiked  to correspond to a TPH
concentration of less than 100 mg/kg; a medium-range
sample was spiked to correspond to a TPH concentration
range of 100 to 1,000 mg/kg; and a high-range sample was
spiked to correspond to a TPH concentration of more than
1,000 mg/kg.  To spike each low- and medium-range soil
sample, ERA used Freon 113 as a "carrier" to distribute the
contaminant evenly  throughout the  sample.   Soil  PE
samples were  spiked with interferents at two  different
levels  ranging  from 50  to  500  percent of the TPH
                            concentration expected to be present. Whenever possible,
                            the interferents were added at levels that best represented
                            real-world  conditions.    ERA  analyzed  the  samples
                            containing weathered gasoline before shipping them to the
                            Navy  BVC  site.   The analytical  results  were used to
                            confirm sample concentrations.

                            Liquid PE samples  consisted  of neat materials.  Each
                            liquid PE sample  consisted  of approximately  2  mL of
                            liquid in a  flame-sealed, glass  ampule.   During  the
                            demonstration, the developers and reference laboratory
                            were given a table informing them of the amount of liquid
                            sample to be used for analysis.
                                                        31

-------
ERA grouped like PE samples together in a resealable bag
and placed all the PE samples in a cooler containing ice for
overnight shipment to the Navy BVC site. When the PE
samples arrived at the site, the samples were labeled with
the appropriate sample identification numbers and placed
in appropriate coolers for transfer to the developers on site
or for shipment to the reference laboratory as summarized
in Section 4.3.2. Table 4-4  summarizes the contaminant
types  and concentration ranges as well as the numbers of
PE samples and analyses associated with the demonstration
of the OCMA-350.
4.3.2
Sample Management
Following sample containerization, each environmental
sample was assigned a unique sample designation defining
the sampling  area, expected  type  of  contamination,
expected concentration range, sampling location, sample
number,  and QC identification,  as appropriate.  Each
sample container was labeled  with the  unique  sample
designation,  date, time, preservative, initials of personnel
who had filled the container, and analysis to be performed.
Each  PE  sample was also assigned a  unique  sample
designation that identified it as a PE sample.  Each PE
sample   designation  also  identified  the   expected
contaminant type and range, whether the sample was soil
or liquid, and the sample number.

Sample custody began when samples were placed in iced
coolers in the possession of the designated field sample
custodian. Demonstration samples were divided into two
groups to allow adequate  time for the developers and
reference laboratory to extract and analyze samples within
the method-specified holding times presented in Table 4-5.
The  two  groups  of samples  for reference laboratory
analysis were placed in coolers containing ice and chain-
of-custody forms and were shipped by overnight courier to
the reference laboratory on the first and third days of the
demonstration.  The two groups of samples for developer
analysis were placed in coolers containing ice and chain-
of-custody forms and  were   hand-delivered  to  the
developers at the Navy BVC site on the same days that the
reference laboratory received its two groups of samples.
During the demonstration,  each developer was provided
with a tent to provide shelter from direct sunlight during
analysis of demonstration samples. In addition,  at the end
of each day, the developer placed any samples or sample
extracts in its custody in coolers, and the coolers  were
stored in a refrigerated truck.
                                                    32

-------
Table 4-4. Performance Evaluation Samples
Sample Type
Typical TPH
Concentration
Range3
Total Number of
Samples to
Horiba and
Reference
Laboratory
Number of
MS/MSDb
Pairs
Number of
Analyses by
Horiba
Number of
Analyses by Reference
Laboratory0
GRO
EDRO
Soil Samples (Ottawa Sand)
Weathered gasoline
Diesel
Low
7
7
0
0
7
7
7
0
7
7
Soil Samples (Processed Garden Soil)
Weathered gasoline
Diesel
Blank soil (control sample)
MTBE (1 ,100 mg/kg) and weathered gasoline
MTBE (1 ,700 mg/kg) and weathered gasoline
PCE (2,810 mg/kg) and weathered gasoline
PCE (13,100 mg/kg) and weathered gasoline
Stoddard solvent (2,900 mg/kg) and
weathered gasoline
Stoddard solvent (15,400 mg/kg) and
weathered gasoline
Turpentine (2,730 mg/kg) and weathered
gasoline
Turpentine (12,900 mg/kg) and weathered
gasoline
Stoddard solvent (3,650 mg/kg) and diesel
Stoddard solvent (18,200 mg/kg) and diesel
Turpentine (3,850 mg/kg) and diesel
Turpentine (19,600 mg/kg) and diesel
1 ,2,4-Trichlorobenzene (3,350 mg/kg) and
diesel
1 ,2,4-Trichlorobenzene (16,600 mg/kg) and
diesel
Humic acid (3,940 mg/kg) and diesel
Humic acid (1 9,500 mg/kg) and diesel
Humic acid (3,940 mg/kg)
Humic acid (1 9,500 mg/kg)
Weathered gasoline at 16 percent moisture
Diesel at negligible moisture (less than
1 percent)
Medium
High
Medium
High
Trace
High
Trace
High
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
0
1
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
5
0
0
5
3
3
3
3
3
3
3
3
3
3
3
3
0
0
0
0
0
0
5
0
3
5
3
5
5
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
5
5
Liquid Samples (Neat Material)
Weathered gasoline
Diesel
MTBE
PCE
High
Not applicable
3
3
6
6
1
0
0
0
3
3
6
6
5
0
6
6
5
3
0
0
                                                          33

-------
Table 4-4. Performance Evaluation Samples (Continued)
Sample Type
Typical TPH
Concentration
Range3
Total Number of
Samples to
Horiba and
Reference
Laboratory
Number of
MS/MSDb
Pairs
Number of
Analyses by
Horiba
Number of
Analyses by Reference
Laboratory0
GRO
EDRO
Liquid Samples (Neat Material) (Continued)
Stoddard solvent
Turpentine
1 ,2,4-Trichlorobenzene
High
Not applicable
Total
6
6
6
125
0
0
0
6
6
6
6
125
6
6
0
90
6
6
6
125
Notes:

mg/kg    =  Milligram per kilogram
MS/MSD =  Matrix spike and matrix spike duplicate
MTBE  =
PCE   =
Methyl-tert-butyl ether
Tetrachloroethene
a   The typical TPH concentration range was based on reference laboratory results for the demonstration. The typical low, medium, and high ranges
    indicate TPH concentrations of less than 100 mg/kg; 100 to 1,000 mg/kg; and greater than 1,000  mg/kg, respectively.  The typical TPH
    concentration range for the liquid sample concentrations was based on the definition of TPH used for the demonstration and knowledge of the
    sample (neat material).

b   MS/MSD samples were analyzed only by the reference laboratory.

0   All soil performance evaluation samples were also analyzed for moisture content by the reference laboratory.
                                                               34

-------
Table 4-5. Sample Container, Preservation, and Holding Time Requirements
Parameter3
GRO
EDRO
Percent moisture
TPH
GRO and EDRO
Medium
Soil
Soil
Soil
Soil
Liquid
Container
Two 5-gram EnCores
Two 4-ounce, glass jars with Teflon™-lined lids
Two 4-ounce, glass jars with Teflon ™-lined lids
One 25-gram EnCore
One 2-milliliter ampule for each analysis
Preservation
4±2°C
4±2°C
4±2°C
4±2°C
Not applicable
Holding Time
Extraction
2b
14b
Not applicable
(days)
Analysis
14
40
7
Performed on site0
See note
d
Notes:

±  = Plus or minus

a    The reference laboratory measured percent moisture using part of the soil sample from the container designated for EDRO analysis.

b    The extraction holding time started on the day that samples were shipped.

0    If GRO analysis of a sample was to be completed by the reference laboratory, the developers completed on-site extraction of the corresponding
     sample within 2 days. Otherwise,  all on-site extractions and analyses were completed within 7 days.

d    The reference laboratory cracked open each ampule and immediately added the specified aliquot of the sample to methanol for GRO analysis and
     to methylene chloride for EDRO analysis. This procedure was performed in such a way that the final volumes of the extracts for GRO and EDRO
     analyses were 5.0 millilitersand I.Omilliliter, respectively. Once the extracts were prepared, the GRO and EDRO analyses were performed within
     14 and 40 days, respectively.
                                                               35

-------
                                              Chapter 5
                                       Confirmatory Process
The performance results for each field measurement device
were  compared to those  for  an off-site laboratory
measurement method—that is, a reference method. This
chapter describes  the rationale for the selection of the
reference method  (Section 5.1) and reference laboratory
(Section  5.2) and summarizes project-specific sample
preparation and analysis procedures associated with the
reference method (Section 5.3).

5.1    Reference Method Selection

During the demonstration, environmental and PE samples
were analyzed for TPH by the reference laboratory using
SW-846  Method  8015B (modified).    This  section
describes the  analytical  methods  considered  for  the
demonstration and provides  a rationale for the reference
method selected.

The reference method used was  selected  based  on  the
following criteria:

•   It is not a field screening method.

•   It is widely used and accepted.

•   It measures  light (gasoline) to heavy (lubricating  oil)
    fuel types.

    It can provide separate  measurements of GRO  and
    EDRO fractions of TPH.

•   It meets project-specific reporting limit requirements.

The analytical methods considered for the demonstration
and the reference  method selected based on the above-
listed criteria are illustrated  in  a flow  diagram  in
Figure 5-1.  The reference method selection process is
discussed below.

Analytical methods considered for the demonstration were
identified based on a review of SW-846, "Methods for
Chemical Analysis of Water and Wastes" (MCAWW),
ASTM, API, and state-specific methods.  The analytical
methods considered collectively represent six different
measurement technologies.  Of the methods  reviewed,
those identified as field screening methods, such as SW-
846   Method  4030,  were  eliminated  from  further
consideration in the reference method selection process.

A literature review was conducted to determine whether
the remaining methods are widely used and accepted in the
United States (Association for Environmental Health and
Sciences [AEHS] 1999). As a result of this review, state-
specific methods such as the Massachusetts Extractable
Petroleum Hydrocarbon (EPH) and Volatile Petroleum
Hydrocarbon (VPH) Methods (Massachusetts Department
of Environmental Protection 2000), the Florida Petroleum
Range Organic (PRO)  Method (Florida Department of
Environmental Protection 1996), and Texas Method 1005
(Texas Natural Resource Conservation Commission 2000)
were  eliminated  from  the  selection  process.   Also
eliminated were  the  gravimetric and infrared methods
except for MCAWW Method 418.1 (EPA 1983). The use
and acceptability of MCAWW Method 418.1 will likely
decline because the extraction solvent used in this method
is Freon  113,  a chlorofluorocarbon (CFC), and  use  of
CFCs will eventually be phased out under the  Montreal
Protocol.  However, because several states still accept the
use of MCAWW Method 418.1 for measuring  TPH, the
method was  retained for  further  consideration  in  the
selection process (AEHS 1999).
                                                   36

-------
                                                           2  oi
                                                                        •E
                                                                         
-------
Of the remaining methods, MCAWW Method 418.1, the
API PHC Method, and  SW-846 Method 8015B can  all
measure  light (gasoline) to heavy (lubricating oil) fuel
types.  However, GRO  and EDRO fractions cannot  be
measured separately using MCAWW Method 418.1. As
a result,  this method was eliminated from the selection
process.

Both the API PHC Method and SW-846 Method 8015B
can be used to separately measure the GRO and DRO
fractions of TPH. These methods can also be modified to
extend the  DRO range to EDRO  by using a calibration
standard that includes  even-numbered  alkanes in the
EDRO range.

Based on a review of state-specific action levels for TPH,
a TPH  reporting limit  of  10  mg/kg was used  for the
demonstration.  Because the TPH reporting limit for the
API PHC  Method (50  to  100 mg/kg)  is greater than
10 mg/kg, this method was eliminated from the selection
process (API 1994).  SW-846 Method 8015B (modified)
met the reporting limit requirements for the demonstration.
For  GRO,  SW-846  Method  8015B (modified) has a
reporting limit of 5 mg/kg, and for EDRO, this method has
a  reporting limit of 10 mg/kg.   Therefore, SW-846
Method  8015B (modified)  satisfied  all  the  criteria
established for selecting the reference  method.  As  an
added benefit,  because this  is a GC  method,  it also
provides   a   fingerprint  (chromatogram)  of  TPH
components.
                                        5.2     Reference Laboratory Selection

                                        This section provides the rationale for the selection of the
                                        reference laboratory. STL Tampa East was selected as the
                                        reference laboratory because it (1) has been performing
                                        TPH  analyses  for  many years,  (2) has  passed many
                                        external audits by successfully implementing a variety of
                                        TPH  analytical methods,  and (3) agreed  to  implement
                                        project-specific analytical requirements. In January 2000,
                                        a project-specific audit of the  laboratory was conducted
                                        and  determined  that  STL Tampa  East  satisfactorily
                                        implemented   the   reference  method  during  the
                                        predemonstration investigation. In addition, STL Tampa
                                        East successfully analyzed double-blind PE samples and
                                        blind field triplicates for GRO and EDRO  during the
                                        predemonstration investigation. Furthermore, in 1998 STL
                                        Tampa  East was one of four recipients and in 1999 was
                                        one of  six recipients  of the Seal of Excellence Award
                                        issued  by  the  American  Council  of  Independent
                                        Laboratories.  In each instance, this award  was issued
                                        based on the results of PE sample analyses and client
                                        satisfaction surveys. Thus, the selection of the reference
                                        laboratory was based  primarily on performance and not
                                        cost.

                                        5.3     Summary of Reference Method

                                        The laboratory sample preparation and analytical methods
                                        used for the demonstration are summarized in Table 5-1.
                                        The  SW-846 methods listed in Table 5-1  for GRO and
Table 5-1.  Laboratory Sample Preparation and Analytical Methods
Parameter
                          Method Reference (Step)
                                                                             Method Title
GRO
EDRO
Based on SW-846 Method 5035 (extraction)

Based on SW-846 Method 5030B (purge-and-trap)
Based on SW-846 Method 8015B (analysis)

Based on SW-846 Method 3540C (extraction)
Based on SW-846 Method 8015B (analysis)
Percent moisture   Based on MCAWW Method 160.3a
Closed-System Purge-and-Trap and Extraction for Volatile Organics
in Soil and Waste Samples
Purge-and-Trap for Aqueous Samples
Nonhalogenated Volatile Organics by Gas Chromatography

Soxhlet Extraction
Nonhalogenated Volatile Organics by Gas Chromatography

Residue, Total (Gravimetric, Dried at 103-105 °C)
Notes:

MCAWW = "Methods for Chemical Analysis of Water and Wastes"
SW-846  = "Test Methods for Evaluating Solid Waste"

a   MCAWW Method 160.3 was modified to include calculation and reporting of percent moisture in soil samples.
                                                     38

-------
EDRO analyses were tailored to meet the definition of
TPH for the  project (see Chapter  1).  Project-specific
procedures for soil sample preparation and analysis for
GRO and EDRO are summarized in Tables 5-2 and 5-3,
respectively.  Project-specific procedures were applied
(1) if amethodused offered choices (forexample, SW-846
Method 5035  for GRO extraction states that samples may
be collected with or without use of a preservative solution),
(2) if a method used did not provide specific details (for
example, SW-846 Method 5035 for GRO extraction does
not specify how  unrepresentative  material  should  be
handled  during  sample   preparation),  or  (3)  if  a
modification to  a  method  used was required in order to
meet demonstration objectives  (for example, SW-846
Method 8015B for EDRO analysis states that quantitation
is performed by summing the areas of all chromatographic
peaks eluting between the end of the  1,2,4-trimethyl
benzene or n-C10 peak, whichever occurs later, and the
n-octacosane  peak; however, an additional quantitation
was performed to sum the areas of all chromatographic
peaks eluting from the  end of the n-octacosane peak
through  the  tetracontane  peak   in  order  to  meet
demonstration objectives).

Before analyzing a liquid PE sample, STL Tampa East
added an aliquot of the liquid PE sample to the extraction
solvent used for soil samples. A specified aliquot of the
liquid PE sample  was  diluted in  methanol for GRO
analysis and in methylene chloride for EDRO analysis
such that the  final volume of the solution for GRO  and
EDRO analyses was 5.0 and 1.0 mL, respectively.  The
solution was then analyzed for GRO and EDRO using the
same procedures as are used for soil sample extracts.
                                                   39

-------
Table 5-2.  Summary of Project-Specific Procedures for GRO Analysis
SW-846 Method Reference (Step)
Project-Specific Procedures
5035 (Extraction)
Low-level (0.5 to 200 micrograms per kilogram) or high-level (greater
than 200 micrograms per kilogram) samples may be prepared.
Samples may be collected with or without use of a preservative
solution.
A variety of sample containers, including EnCores, may be used when
high-level samples are collected without use of a preservative.
Samples collected in EnCores should be transferred to vials containing
the extraction solvent as soon as possible or analyzed within 48 hours.
For samples not preserved in the field, a solubility test should be
performed using methanol, polyethylene glycol, and hexadecane to
determine an appropriate extraction solvent.
Removal of unrepresentative material from the sample is not discussed.
Procedures for adding surrogates to the sample are inconsistently
presented. Section 2.2.1 indicates that surrogates should be added to
an aliquot of the extract solution. Section 7.3.3 indicates that soil
should be added to a vial containing both the extraction solvent
(methanol) and surrogate spiking solution.
Nine ml of methanol should be added to a 5-gram (wet weight) soil
sample.
When practical, the sample should be dispersed to allow contact with
the methanol by shaking or using other mechanical means for 2 min
without opening the sample container. When shaking is not practical,
the sample should be dispersed with a narrow, metal spatula, and the
sample container should be immediately resealed.
Because the project-specific reporting limit for GRO was 5 milligrams
per kilogram, all samples analyzed for GRO were prepared using
procedures for high-level samples.
Samples were collected without use of a preservative.
Samples were containerized in EnCores.
Samples were weighed and extracted within 2 calendar days of their
shipment. The holding time for analysis was 14 days after extraction. A
full set of quality control samples (method blanks, MS/MSDs, and
LCS/LCSDs) was prepared within this time.
Because the reference laboratory obtained acceptable results for
performance evaluation samples extracted with methanol during the
predemonstration investigation, samples were extracted with methanol.
During sample homogenization, field sampling technicians attempted to
remove unrepresentative material such as sticks, roots, and stones if
present in the sample; the reference laboratory did not remove any
remaining unrepresentative material.
The soil sample was ejected into a volatile organic analysis vial, an
appropriate amount of surrogate solution was added to the sample, and
then methanol was quickly added.
Five ml of methanol was added to the entire soil sample contained in a
5-gram EnCore.
The sample was dispersed using a stainless-steel spatula to allow
contact with the methanol. The volatile organic analysis vial was then
capped and shaken vigorously until the soil was dispersed in methanol,
and the soil was allowed to settle.
5030B (Purge-and-Trap)
Screening of samples before the purge-and-trap procedure is
recommended using one of the two following techniques:
Use of an automated headspace sampler (see SW-846 Method 5021)
connected to a GC equipped with a photoionization detector in series
with an electrolytic conductivity detector
Extraction of the samples with hexadecane (see SW-846 Method 3820)
and analysis of the extracts using a GC equipped with a flame
ionization detector or electron capture detector
SW-846 Method 5030B indicates that contamination by carryover can
occur whenever high-level and low-level samples are analyzed in
sequence. Where practical, analysis of samples with unusually high
concentrations of analytes should be followed by an analysis of organic-
free reagent water to check for cross-contamination. Because the trap
and other parts of the system are subject to contamination, frequent
bake-out and purging of the entire system may be required.
Samples were screened with an automated headspace sampler (see
SW-846 Method 5021) connected to a GC equipped with a flame
ionization detector.
According to the reference laboratory, a sample extract concentration
equivalent to 10,000 ng on-column is the minimum concentration of
GRO that could result in carryover. Therefore, if a sample extract had a
concentration that exceeded the minimum concentration for carryover,
the next sample in the sequence was evaluated as follows: (1) if the
sample was clean (had no chromatographic peaks), no carryover had
occurred; (2) if the sample had detectable analyte concentrations
(chromatographic peaks), it was reanalyzed under conditions in which
carryover did not occur.
                                                           40

-------
Table 5-2.  Summary of Project-Specific Procedures for GRO Analysis (Continued)
SW-846 Method Reference (Step)
Project-Specific Procedures
5030B (Purge-and-Trap) (Continued)
The sample purge device used must demonstrate adequate
performance.
Purge-and-trap conditions for high-level samples are not clearly
specified. According to SW-846, manufacturer recommendations for
the purge-and-trap devices should be considered when the method is
implemented. The following general purge-and-trap conditions are
recommended for samples that are water-miscible (methanol extract):
Purge gas: nitrogen or helium
Purge gas flow rate: 20 mL/min
Purge time: 15 ± 0.1 min
Purge temperature: 85 ± 2 °C
Desorb time: 1 .5 min
Desorb temperature: 180 °C
Backflush inert gas flow rate: 20 to 60 mL/min
Bake time: not specified
Bake temperature: not specified
Multiport valve and transfer line temperatures: not specified
A Tekmar 201 6 autosampler and a Tekmar LSC 2000 concentrator
were used. Based on quality control sample results, the reference
laboratory had demonstrated adequate performance using these
devices.
The purge-and-trap conditions that were used are listed below. These
conditions were based on manufacturer recommendations for the purge
device specified above and the VOCARB 3000 trap.
Purge gas: helium
Purge gas flow rate: 35 mL/min
Purge time: 8 min with 2-min dry purge
Purge temperature: ambient temperature
Desorb time: 1 min
Desorb temperature: 250 °C
Backflush inert gas flow rate: 35 mL/min
Bake time: 7 min
Bake temperature: 270 °C
Multiport valve and transfer line temperatures: 1 1 5 and 1 20 °C
8015B (Analysis)
GC Conditions
The following GC conditions are recommended:
Column: 30-meter x 0.53-millimeter-inside diameter, fused-silica
capillary column chemically bonded with 5 percent methyl
silicone, 1 .5-micrometer field thickness
Carrier gas: helium
Carrier gas flow rate: 5 to 7 mL/min
Makeup gas: helium
Makeup gas flow rate: 30 mL/min
Injector temperature: 200 °C
Detector temperature: 340 °C
Temperature program:
Initial temperature: 45 °C
Hold time: 1 min
Program rate: 45 to 100 °C at 5 °C/min
Program rate: 1 00 to 275 °C at 8 °C/min
Hold time: 5 min
Overall time: 38.9 min
The HP 5890 Series II was used as the GC. The following GC
conditions were used based on manufacturer recommendations:
Column: 30-meter x 0.53-millimeter-inside diameter, fused-silica
capillary column chemically bonded with 5 percent methyl
silicone, 1 .5-micrometer field thickness
Carrier gas: helium
Carrier gas flow rate: 1 5 mL/min
Makeup gas: helium
Makeup gas flow rate: 15 mL/min
Injector temperature: 200 °C
Detector temperature: 200 °C
Temperature program:
Initial temperature: 25 °C
Hold time: 3 min
Program rate: 25 to 120 °C at 25 °C/min
Hold time: 4 min
Program rate: 1 20 to 245 °C at 25 °C/min
Hold time: 5 min
Overall time: 20.4 min
Calibration
The chromatographic system may be calibrated using either internal or
external standards.
Calibration should be performed using samples of the specific fuel type
contaminating the site. When such samples are not available, recently
purchased, commercially available fuel should be used.
The chromatographic system was calibrated using external standards
with a concentration range equivalent to 1 00 to 1 0,000 ng on-column.
The reference laboratory acceptance criterion for initial calibration was a
relative standard deviation less than or equal to 20 percent of the
average response factor or a correlation coefficient for the least-
squares linear regression greater than or equal to 0.990.
Calibration was performed using a commercially available,
10-component GRO standard that contained 35 percent aliphatic
hydrocarbons and 65 percent aromatic hydrocarbons.
                                                           41

-------
Table 5-2.  Summary of Project-Specific Procedures for GRO Analysis (Continued)
SW-846 Method Reference (Step)
Project-Specific Procedures
8015B (Analysis) (Continued)
Calibration (Continued)
Initial calibration verification is not required.
CCV should be performed at the beginning of every 1 2-hour work shift
and at the end of an analytical sequence. CCV throughout the 12-hour
shift is also recommended; however, the frequency is not specified.
CCV should be performed using a fuel standard.
According to SW-846 Method 8000, CCV should be performed at the
same concentration as the midpoint concentration of the initial
calibration curve; however, the concentration of each calibration point is
not specified.
A method sensitivity check is not required.
Initial calibration verification was performed using a second-source
standard that contained a 10-component GRO standard made up of
35 percent aliphatic hydrocarbons and 65 percent aromatic
hydrocarbons at a concentration equivalent to 2,000 ng on-column. The
reference laboratory acceptance criterion for initial calibration
verification was an instrument response within 25 percent of the
response obtained during initial calibration.
CCV was performed at the beginning of each analytical batch, after
every tenth analysis, and at the end of the analytical batch. The
reference laboratory acceptance criteria for CCV were instrument
responses within 25 percent (for the closing CCV) and 15 percent (for
all other CCVs) of the response obtained during initial calibration.
CCV was performed using a commercially available, 10-component
GRO standard that contained 35 percent aliphatic hydrocarbons and
65 percent aromatic hydrocarbons.
CCV was performed at a concentration equivalent to 2,000 ng
on-column.
A method sensitivity check was performed daily using a calibration
standard with a concentration equivalent to 100 ng on-column. The
reference laboratory acceptance criterion for the method sensitivity
check was detection of the standard.
Retention Time Windows
The retention time range (window) should be established using
2-methylpentane and 1 ,2,4-trimethylbenzene during initial calibration.
Three measurements should be made over a 72-hour period; the results
should be used to determine the average retention time. As a minimum
requirement, the retention time should be verified using a midlevel
calibration standard at the beginning of each 12-hour shift. Additional
analysis of the standard throughout the 1 2-hour shift is strongly
recommended.
The retention time range was established using the opening CCV
specific to each analytical batch. The first eluter, 2-methylpentane, and
the last eluter, 1 ,2,4-trimethylbenzene, of the GRO standard were used
to establish each day's retention time range.
Quantitation
Quantitation is performed by summing the areas of all chromatographic
peaks eluting within the retention time range established using
2-methylpentane and 1,2,4-trimethylbenzene. Subtraction of the
baseline rise for the method blank resulting from column bleed is
generally not required.
Quantitation was performed by summing the areas of all
chromatographic peaks from 2-methylpentane through
1 ,2,4-trimethylbenzene. This range includes n-C10. Baseline rise
subtraction was not performed.
Quality Control
Spiking compounds for MS/MSDs and LCSs are not specified.
According to SW-846 Method 8000, spiking levels for MS/MSDs are
determined differently for compliance and noncompliance monitoring
applications. For noncompliance applications, the laboratory may spike
the sample (1) at the same concentration as the reference sample
(LCS), (2) at 20 times the estimated quantitation limit for the matrix of
interest, or (3) at a concentration near the middle of the calibration
range.
The spiking compound mixture for MS/MSDs and LCSs was the 1 0-
component GRO calibration standard.
MS/MSD spiking levels were targeted to be between 50 and
150 percent of the unspiked sample concentration. The reference
laboratory used historical information to adjust spike amounts or to
adjust sample amounts to a preset spike amount. The spiked samples
and unspiked samples were prepared such that the sample mass and
extract volume used for analysis were the same.
                                                           42

-------
Table 5-2. Summary of Project-Specific Procedures for GRO Analysis (Continued)
SW-846 Method Reference (Step)
Project-Specific Procedures
8015B (Analysis) (Continued)
Quality Control (Continued)
According to SW-846 Method 8000, in-house laboratory acceptance
criteria for MS/MSDs and LCSs should be established. As a general
rule, the recoveries of most compounds spiked into a sample should fall
within the range of 70 to 130 percent, and this range should be used as
a guide in evaluating in-house performance.
The LCS should consist of an aliquot of a clean (control) matrix that is
similar to the sample matrix.
No LCSD is required.
The surrogate compound and spiking concentration are not specified.
According to SW-846 Method 8000, in-house laboratory acceptance
criteria for surrogate recoveries should be established.
The method blank matrix is not specified.
The extract duplicate is not specified.
The reference laboratory acceptance criteria for MS/MSDs and LCSs
were a relative percent difference less than or equal to 25 with 33 to
115 percent recovery. The acceptance criteria were based on
laboratory historical information. These acceptance criteria are similar
to those of the methods cited in Figure 5-1 .
The LCS/LCSD matrix was Ottawa sand.
The spiking compound mixture for LCSDs was the 1 0-component GRO
calibration standard.
The surrogate compound was 4-bromofluorobenzene. The reference
laboratory acceptance criterion for surrogates was 39 to 1 63 percent
recovery.
The method blank matrix was Ottawa sand. The reference laboratory
acceptance criterion for the method blank was less than or equal to the
project-specific reporting limit.
The extract duplicate was analyzed. The reference laboratory
acceptance criterion for the extract duplicate was a relative percent
difference less than or equal to 25.
Notes:

±      =  Plus or minus
CCV   =  Continuing calibration verification
GC    =  Gas chromatograph
LCS   =  Laboratory control sample
LCSD  =  Laboratory control sample duplicate
min      = Minute
mL      = Milliliter
MS      = Matrix spike
MSD     = Matrix spike duplicate
ng       = Nanogram
SW-846  = "Test Methods for Evaluating Solid Waste"
                                                              43

-------
Table 5-3. Summary of Project-Specific Procedures for EDRO Analysis
SW-846 Method Reference (Step)
Project-Specific Procedures
3540C (Extraction)
Any free water present in the sample should be decanted and
discarded. The sample should then be thoroughly mixed, and any
unrepresentative material such as sticks, roots, and stones should be
discarded.
Ten grams of soil sample should be blended with 1 0 grams of
anhydrous sodium sulfate.
Extraction should be performed using 300 ml of extraction solvent.
Acetone and hexane (1:1 volume per volume) or methylene chloride
and acetone (1:1 volume per volume) may be used as the extraction
solvent.
Note: Methylene chloride and acetone are not constant-boiling
solvents and thus are not suitable for the method. Methylene
chloride was used as an extraction solvent for method
validation.
The micro Snyder column technique or nitrogen blowdown technique
may be used to adjust (concentrate) the soil extract to the required final
volume.
Procedures for addressing contamination carryover are not specified.
During sample homogenization, field sampling technicians attempted to
remove unrepresentative material such as sticks, roots, and stones. In
addition, the field sampling technicians decanted any free water present
in the sample. The reference laboratory did not decant water or remove
any unrepresentative material from the sample. The reference
laboratory mixed the sample with a stainless-steel tongue depressor.
Thirty grams of sample was blended with at least 30 grams of
anhydrous sodium sulfate. For medium- and high-level samples, 6 and
2 grams of soil were used for extraction, respectively, and proportionate
amounts of anhydrous sodium sulfate were added. The amount of
anhydrous sodium sulfate used was not measured gravimetrically but
was sufficient to ensure that free moisture was effectively removed from
the sample.
Extraction was performed using 200 ml of extraction solvent.
Methylene chloride was used as the extraction solvent.
Kuderna Danish and nitrogen evaporation were used as the
concentration techniques.
According to the reference laboratory, a sample extract concentration of
1 00,000 micrograms per ml is the minimum concentration of EDRO
that could result in carryover. Therefore, if a sample extract had a
concentration that exceeded the minimum concentration for carryover,
the next sample in the sequence was evaluated as follows: (1) if the
sample was clean (had no chromatographic peaks), no carryover
occurred; (2) if the sample had detectable analyte concentrations
(chromatographic peaks), it was reanalyzed under conditions in which
carryover did not occur.
8015B (Analysis)
GC Conditions
The following GC conditions are recommended:
Column: 30-meter x 0.53-millimeter-inside diameter, fused-silica
capillary column chemically bonded with 5 percent methyl
silicone, 1 .5-micrometer field thickness
Carrier gas: helium
Carrier gas flow rate: 5 to 7 mL/min
Makeup gas: helium
Makeup gas flow rate: 30 mL/min
Injector temperature: 200 °C
Detector temperature: 340 °C
Temperature program:
Initial temperature: 45 °C
Hold time: 3 min
Program rate: 45 to 275 °C at 1 2 °C/min
Hold time: 12 min
Overall time: 34.2 min
An HP 6890 GC was used with the following conditions:
Column: 30-meter x 0.53-millimeter-inside diameter, fused-silica
capillary column chemically bonded with 5 percent methyl
silicone, 1 .5-micrometer field thickness
Carrier gas: hydrogen
Carrier gas flow rate: 1 .9 mL/min
Makeup gas: hydrogen
Makeup gas flow rate: 23 mL/min
Injector temperature: 250 °C
Detector temperature: 345 °C
Temperature program:
Initial temperature: 40 °C
Hold time: 2 min
Program rate: 40 to 345 °C at 30 °C/min
Hold time: 5 min
Overall time: 17.2 min
                                                          44

-------
Table 5-3. Summary of Project-Specific Procedures for EDRO Analysis (Continued)
SW-846 Method Reference (Step)
Project-Specific Procedures
8015B (Analysis) (Continued)
Calibration
The chromatographic system may be calibrated using either internal or
external standards.
Calibration should be performed using samples of the specific fuel type
contaminating the site. When such samples are not available, recently
purchased, commercially available fuel should be used.
ICV is not required.
CCV should be performed at the beginning of every 1 2-hour work shift
and at the end of an analytical sequence. CCV throughout the 12-hour
shift is also recommended; however, the frequency is not specified.
CCV should be performed using a fuel standard.
According to SW-846 Method 8000, CCV should be performed at the
same concentration as the midpoint concentration of the initial
calibration curve; however, the concentration of each calibration point is
not specified.
A method sensitivity check is not required.
The chromatographic system was calibrated using external standards
with a concentration range equivalent to 75 to 7,500 ng on-column. The
reference laboratory acceptance criterion for initial calibration was a
relative standard deviation less than or equal to 20 percent of the
average response factor or a correlation coefficient for the least-
squares linear regression greater than or equal to 0.990.
Calibration was performed using a commercially available standard that
contained even-numbered alkanes from C10 through C40.
ICV was performed using a second-source standard that contained
even-numbered alkanes from C10 through C40 at a concentration
equivalent to 3,750 ng on-column. The reference laboratory
acceptance criterion for ICV was an instrument response within
25 percent of the response obtained during initial calibration.
CCV was performed at the beginning of each analytical batch, after
every tenth analysis, and at the end of the analytical batch. The
reference laboratory acceptance criteria for CCV were instrument
responses within 25 percent (for the closing CCV) and 15 percent (for
all other CCVs) of the response obtained during initial calibration.
CCV was performed using a standard that contained only even-
numbered alkanes from C10 through C40
CCV was performed at a concentration equivalent to 3,750 ng
on-column.
A method sensitivity check was performed daily using a calibration
standard with a concentration equivalent to 75 ng on-column. The
reference laboratory acceptance criterion for the method sensitivity
check was detection of the standard.
Retention Time Windows
The retention time range (window) should be established using
C10 and C28 alkanes during initial calibration. Three measurements
should be made over a 72-hour period; the results should be used to
determine the average retention time. As a minimum requirement, the
retention time should be verified using a midlevel calibration standard at
the beginning of each 1 2-hour shift. Additional analysis of the standard
throughout the 1 2-hour shift is strongly recommended.
Two retention time ranges were established using the opening CCV for
each analytical batch. The first range, which was labeled diesel range
organics, was marked by the end of the 1 ,2,4-trimethylbenzene or n-C10
peak, whichever occurred later, through the n-octacosane peak. The
second range, which was labeled oil range organics, was marked by the
end of the n-octacosane peak through the tetracontane peak.
Quantitation
Quantitation is performed by summing the areas of all chromatographic
peaks eluting between n-C10 and n-octacosane.
Quantitation was performed by summing the areas of all
chromatographic peaks from the end of the 1 ,2,4-trimethylbenzene or
n-C10 peak, whichever occurred later, through the n-octacosane peak.
A separate quantitation was also performed to sum the areas of all
chromatographic peaks from the end of the n-octacosane peak through
the tetracontane peak. Separate average response factors for the
carbon ranges were used for quantitation. The quantitation results were
then summed to determine the total EDRO concentration.
All calibrations, ICVs, CCVs, and associated batch quality control
measures were controlled for the entire EDRO range using a single
quantitation performed over the entire EDRO range.
                                                           45

-------
Table 5-3. Summary of Project-Specific Procedures for EDRO Analysis (Continued)
SW-846 Method Reference (Step)
Project-Specific Procedures
8015B (Analysis) (Continued)
Quantitation (Continued)
Subtraction of the baseline rise for the method blank resulting from
column bleed is appropriate.
Because phthalate esters contaminate many types of products
commonly found in the laboratory, consistent quality control should be
practiced.
The reference laboratory identified occurrences of baseline rise in the
data package. The baseline rise was evaluated during data validation
and subtracted when appropriate based on analyst discretion.
Phthalate peaks were not noted during analysis.
Quality Control
Spiking compounds for MS/MSDs and LCSs are not specified.
According to SW-846 Method 8000, spiking levels for MS/MSDs are
determined differently for compliance and noncompliance monitoring
applications. For noncompliance applications, the laboratory may spike
the sample (1) at the same concentration as the reference sample
(LCS), (2) at 20 times the estimated quantitation limit for the matrix of
interest, or (3) at a concentration near the middle of the calibration
range.
According to SW-846 Method 8000, in-house laboratory acceptance
criteria for MS/MSDs and LCSs should be established. As a general
rule, the recoveries of most compounds spiked into a sample should fall
within the range of 70 to 130 percent, and this range should be used as
a guide in evaluating in-house performance.
The LCS should consist of an aliquot of a clean (control) matrix that is
similar to the sample matrix.
No LCSD is required.
The surrogate compound and spiking concentration are not specified.
According to SW-846 Method 8000, in-house laboratory acceptance
criteria for surrogate recoveries should be established.
The method blank matrix is not specified.
The extract duplicate is not specified.
The spiking compound for MS/MSDs and LCSs was an EDRO standard
that contained even-numbered alkanes from C10 through C40.
MS/MSD spiking levels were targeted to be between 50 and
150 percent of the unspiked sample concentration. The reference
laboratory used historical information to adjust spike amounts or to
adjust sample amounts to a preset spike amount. The spiked samples
and unspiked samples were prepared such that the sample mass and
extract volume used for analysis were the same.
The reference laboratory acceptance criteria for MS/MSDs and LCSs
were a relative percent difference less than or equal to 45 with 46 to
1 24 percent recovery. The acceptance criteria were based on
laboratory historical information. These acceptance criteria are similar
to those of the methods cited in Figure 5-1 .
The LCS/LCSD matrix was Ottawa sand.
The spiking compound for LCSDs was the EDRO standard that
contained even-numbered alkanes from C10 through C40.
The surrogate compound was o-terphenyl. The reference laboratory
acceptance criterion for surrogates was 45 to 1 43 percent recovery.
The method blank matrix was Ottawa sand. The reference laboratory
acceptance criterion for the method blank was less than or equal to the
project-specific reporting limit.
The extract duplicate was analyzed. The reference laboratory
acceptance criterion for the extract duplicate was a relative percent
difference less than or equal to 45.
Notes:

CCV   =
GC
ICV
LCS   =
LCSD  =
min    =
Continuing calibration verification           mL
Gas chromatograph                       MS
Initial calibration verification                MSD
Laboratory control sample                  n-Cx
Laboratory control sample duplicate         ng
Minute                                  SW-846
Milliliter
Matrix spike
Matrix spike duplicate
Alkane with "x" carbon atoms
Nanogram
"Test Methods for Evaluating Solid Waste"
                                                               46

-------
                                              Chapter 6
                        Assessment of Reference Method Data Quality
This chapter assesses reference method data quality based
on QC check results and PE sample results. A summary of
reference method data quality is included at the end of this
chapter.

To ensure that the reference method results were of known
and adequate  quality, EPA representatives performed a
predemonstration audit and an in-process audit of the
reference laboratory. The predemonstration audit findings
were used in developing the predemonstration design. The
in-process audit was performed when the laboratory had
analyzed a sufficient number of demonstration samples for
both GRO and EDRO and had prepared its first  data
package.   During  the  audit,  EPA  representatives
(1) verified that the laboratory had properly implemented
the EPA-approved demonstration plan and (2) performed
a critical review of the first data package.  All issues
identified during the  audit were fully addressed by the
laboratory  before  it  submitted  the subsequent  data
packages to the EPA. The laboratory also addressed issues
identified during  the EPA final review of  the  data
packages. Audit findings are summarized in the DER for
the demonstration.

6.1    Quality Control Check Results

This section summarizes QC check results for GRO and
EDRO analyses performed using the reference method.
The QC checks associated with soil sample analyses for
GRO  and EDRO included  method blanks,  surrogates,
matrix spikes and matrix spike duplicates (MS/MSD), and
laboratory control samples and laboratory control sample
duplicates (LCS/LCSD).  In addition, extract duplicates
were analyzed for soil environmental samples.  The QC
checks associated with liquid PE sample analysis for GRO
included method  blanks,  surrogates,  MS/MSDs,  and
LCS/LCSDs.   Because liquid PE sample analyses for
EDRO did not include  a preparation  step,  surrogates,
MS/MSDs, and LCS/LCSDs were not analyzed; however,
an instrument blank was analyzed as  a method blank
equivalent. The results for the QC checks were compared
to project-specific acceptance criteria. These criteria were
based on the  reference laboratory's historical QC limits
and its  experience in analyzing the predemonstration
investigation  samples using the reference method.  The
reference  laboratory's QC limits  were established as
described  in   SW-846  and were  within the  general
acceptance criteria recommended by SW-846 for organic
analytical methods.

Laboratory duplicates were also analyzed to evaluate the
precision associated with percent moisture analysis of soil
samples.  The acceptance criterion for the laboratory
duplicate results was an RPD less than or equal to 20. All
laboratory duplicate results met this criterion.  The results
for the laboratory duplicates are not separately discussed
in this  ITVR because  soil sample TPH results  were
compared  on a wet weight basis except for those used to
address primary object P4 (effect of soil moisture content).

6.1.1       GRO Analysis

This section summarizes the results for QC checks used by
the reference  laboratory during GRO analysis, including
method blanks, surrogates, MS/MSDs, extract duplicates,
and LCS/LCSDs. A summary  of the QC check results is
presented at the end of the section.

Method Blanks

Method blanks were analyzed  to verify that steps in the
analytical  procedure did not introduce contaminants that
affected analytical results.  Ottawa sand and deionized
water were used as method blanks for soil and liquid
                                                   47

-------
samples, respectively.  These blanks underwent all the
procedures required for sample preparation. The results
for all method blanks met the acceptance criterion of being
less than or equal to the required project-specific reporting
limit (5 mg/kg). Based on method blank results, the GRO
analysis results were considered to be valid.

Surrogates

Each soil investigative and QC sample for GRO analysis
was  spiked with  a surrogate, 4-bromofluorobenzene,
before extraction to determine whether significant matrix
effects existed within the sample and to estimate the
efficiency of analyte recovery during sample preparation
and analysis. A diluted, liquid PE sample was also spiked
with the surrogate during sample preparation. The initial
surrogate spiking  levels for soil and liquid PE samples
were 2 mg/kg and 40 (ig/L, respectively. The acceptance
criterion was 39 to 163 percent surrogate  recovery.  For
samples analyzed at a dilution factor greater than four, the
surrogate concentration was diluted to a level below the
reference laboratory's reporting limit for the reference
method; therefore, surrogate recoveries for these samples
were not used to assess impacts on data quality.

A total of 101 surrogate measurements were made during
analysis of environmental and associated QC samples.
Fifty-six of these samples were  analyzed at a dilution
factor less than or equal to four. The surrogate recoveries
for these 56 samples ranged from 43 to 345 percent with a
mean recovery of 150 percent and a median recovery  of
136 percent.  Because the mean  and median recoveries
were greater than 100 percent, an overall positive bias was
indicated.

The surrogate recoveries for 16 of the 56 samples did not
meet the acceptance criterion.  In each case, the surrogate
was recovered at a concentration above the upper limit  of
the acceptance criterion.   Examination of  the  gas
chromatograms for the 16 samples revealed that some
PHCs or naturally occurring interferents present in these
environmental samples coeluted  with  the surrogate,
resulting in higher surrogate recoveries. Such coelution is
typical for hydrocarbon-containing samples analyzed using
a GC/FID technique, which was the technique used in the
reference  method.  The  surrogate  recoveries for QC
samples such as method blanks and LCS/LCSDs met the
acceptance criterion, indicating that the laboratory sample
preparation  and  analysis  procedures were in control.
Because  the   coelution   was   observed  only  for
environmental  samples  and  because  the  surrogate
recoveries for QC samples met the acceptance criterion,
the  reference   laboratory  did   not   reanalyze  the
environmental samples with high surrogate recoveries.
Calculations performed to evaluate whether the coelution
resulted  in  underreporting  of  GRO  concentrations
indicated an insignificant impact of less  than 3 percent.
Based  on the surrogate results for environmental and
associated  QC  samples, the GRO analysis results for
environmental samples were considered to be valid.

A total of 42 surrogate measurements were made during
the analysis  of soil PE and associated QC samples.
Thirty-four of these samples were analyzed at a dilution
factor less than or equal to four. The surrogate recoveries
for these 34 samples ranged from 87 to 108 percent with a
mean recovery of 96 percent and a median recovery  of
95 percent.  The surrogate recoveries for all 34 samples
met the acceptance criterion.  Based on the  surrogate
results for soil PE and associated QC samples, the  GRO
analysis results for soil PE samples were considered to be
valid.

A total of 37 surrogate measurements were made during
the analysis of liquid PE and associated QC samples. Six
of these samples were analyzed at a dilution factor less
than or equal to four.  All six samples were QC samples
(method blanks  and  LCS/LCSDs).    The  surrogate
recoveries  for these  six samples ranged  from  81  to
84 percent, indicating a small negative bias. However, the
surrogate recoveries for all six samples met the acceptance
criterion. Based on the surrogate results for liquid PE and
associated QC samples, the GRO analysis results for liquid
PE samples were considered to be valid.

Matrix Spikes and Matrix Spike Duplicates

MS/MSD results were evaluated to determine the accuracy
and precision of the analytical results with respect to the
effects of the sample matrix. For GRO analysis, each soil
sample designated as an MS or MSD was  spiked with the
GRO calibration standard  at an initial spiking level  of
20 mg/kg.  MS/MSDs were also prepared for  liquid PE
samples. Each diluted, liquid PE sample designated as an
MS or MSD was spiked with the GRO calibration standard
at an initial spiking level of 40 (ig/L.  The acceptance
criteria for MS/MSDs were 33 to 115 percent recovery and
an RPD less than or equal to 25.  When the MS/MSD
percent recovery acceptance criterion was  not met, instead
of attributing the  failure  to  meet the  criterion to an
                                                    48

-------
inappropriate spiking level, the  reference laboratory
respiked the sample at a more appropriate and practical
spiking level. Information on the selection of the spiking
level and calculation of percent recoveries for MS/MSD
samples is provided below.

According  to  Provost  and Elder (1983), for percent
recovery data to be reliable, spiking  levels should be at
least five times the unspiked sample  concentration. For
the demonstration,  however, a  large number  of the
unspiked sample concentrations were expected to range
between 1,000 and  10,000 mg/kg, so use of such high
spiking levels was  not  practical.  Therefore, a target
spiking level of 50 to 150 percent of the unspiked  sample
concentration was used for the demonstration. Provost and
Elder (1983) also  present an alternate  approach for
calculating percent recoveries for MS/MSD samples (100
times the ratio of the measured concentration in a spiked
sample to the calculated concentration in the sample).
However, for the demonstration, percent recoveries were
calculated using the traditional approach  (100 times the
ratio of the amount recovered to the amount spiked)
primarily because the alternate approach is not commonly
used.

For environmental samples, a total of 10 MS/MSD pairs
were analyzed.  Four sample pairs collected in the NEX
Service Station Area were designated as MS/MSDs. The
sample matrix in this area primarily consisted of medium-
grained sand. The percent recoveries for all but one of the
MS/MSD  samples ranged from 67 to 115 with RPDs
ranging  from 2 to  14.  Only one MS sample  with a
162 percent recovery did not meet the percent recovery
acceptance  criterion;  however,  the  RPD  acceptance
criterion for the MS/MSD and the percent recovery and
RPD acceptance criteria for the LCS/LCSD associated
with the analytical batch for this sample were met. Based
on the MS/MSD results, the GRO analysis results for the
NEX Service Station Area samples were considered to be
valid.

Two sample pairs  collected in  the B-38 Area were
designated as MS/MSDs.  The sample matrix in this area
primarily  consisted of  sand and clay.   The  percent
recoveries for the MS/MSD samples ranged from 60 to 94
with RPDs of 1 and 13. Therefore, the percent recoveries
and RPDs for these samples met the  acceptance criteria.
Based on the MS/MSD results, the GRO analysis results
for the B-38 Area samples were considered to be valid.
Four  sample pairs  collected in the SFT Area were
designated as MS/MSDs.  The sample matrix in this area
primarily consisted of silty clay. The percent recoveries
for the MS/MSD samples ranged from 0 to 127 with RPDs
ranging from 4 to 21.  Of the  four sample pairs, two
sample pairs metthe percent recovery acceptance criterion,
one sample pair exhibited percent recoveries less than the
lower acceptance  limit, and one sample pair exhibited
percent recoveries greater than the upper acceptance limit.
For the two sample pairs that did not meet the percent
recovery acceptance  criterion, the  RPD acceptance
criterion for the MS/MSDs and the percent recovery and
RPD acceptance criteria for the LCS/LCSDs associated
with the analytical batches for these samples were met.
Because of the varied percent recoveries for the MS/MSD
sample pairs, it was not possible to conclude whether the
GRO analysis  results for the  SFT Area samples had a
negative  or positive bias.   Although one-half of the
MS/MSD  results  did not meet the percent  recovery
acceptance criterion, the out-of-control situations alone did
not constitute adequate grounds for rejection of any of the
GRO analysis results for the SFT Area samples. The out-
of-control  situations may have  been  associated  with
inadequate  spiking levels  (0.7 to 2.8 times the unspiked
sample  concentrations  compared  to  the  minimum
recommended value of 5 times the concentrations).

Three soil PE sample pairs were designated as MS/MSDs.
The  sample matrix for these samples consisted of silty
sand.  The  percent recoveries for these  samples ranged
from 88 to 103 with RPDs ranging from 4 to  6.  The
percent recoveries and RPDs for these samples met the
acceptance  criteria.  Based on the MS/MSD results, the
GRO analysis  results for the  soil  PE samples were
considered to be valid.

Two  liquid PE  sample pairs were  designated as
MS/MSDs.  The  percent recoveries for these samples
ranged from 77 to 87 with RPDs of 1 and 5. The percent
recoveries and RPDs for these samples metthe acceptance
criteria. Based on the MS/MSD results, the GRO analysis
results for the liquid PE samples were considered to be
valid.

Extract Duplicates

For GRO analysis, after soil sample extraction, extract
duplicates  were  analyzed  to evaluate the  precision
associated  with the  reference  laboratory's  analytical
procedure.  The reference laboratory sampled duplicate
                                                   49

-------
aliquots of the GRO extracts for analysis. The acceptance
criterion for extract duplicate precision was an RPD less
than or equal to 25. Two or more environmental samples
collected in each demonstration area whose samples were
analyzed for GRO (the NEX  Service  Station, B-38, and
SFT Areas) were designated as extract duplicates. A total
of  10  samples designated as extract duplicates  were
analyzed for GRO. The RPDs for these samples ranged
from 0.5 to 11. Therefore, the RPDs for all the extract
duplicates met the acceptance criterion.  Based on the
extract duplicate results, the  GRO analysis results were
considered to be valid.

Laboratory Control Samples and Laboratory
Control Sample Duplicates

For GRO analysis, LCS/LCSD results were evaluated to
determine  the  accuracy and precision associated with
control samples prepared by the reference laboratory.  To
generate a soil LCS or  LCSD, Ottawa sand was spiked
with the GRO calibration standard at a spiking level of
20 mg/kg.  To generate an LCS or LCSD for liquid PE
sample analysis, deionized water was spiked with the GRO
calibration standard at a spiking level of 40  (ig/L.  The
acceptance criteria for LCS/LCSDs were 33 to 115 percent
recovery  and an RPD  less than  or  equal to 25.   The
LCS/LCSD  acceptance criteria  were  based on the
reference laboratory's historical data.

Ten pairs of soil LCS/LCSD  samples were prepared and
analyzed. The percent recoveries for these samples ranged
from 87 to 110 with RPDs  ranging  from 2 to  14.  In
addition, two pairs of liquid LCS/LCSD samples were
prepared and analyzed.  The percent recoveries for these
samples ranged from 91 to 92 with RPDs equal to 0 and 1.
Therefore, the percent recoveries and RPDs for the soil and
liquid LCS/LCSD samples met the acceptance criteria,
indicating that the GRO analysis procedure was in control.
Based on the LCS/LCSD results, the GRO analysis results
were considered to be valid.

Summary of Quality Control Check Results

Table 6-1  summarizes the QC check results for  GRO
analysis.  Based on the QC check results, the conclusions
presented below were drawn  regarding the accuracy and
precision of GRO analysis results for the demonstration.
The project-specific percent recovery acceptance criteria
were met for most environmental samples and all  PE
samples.  As expected, the percent recovery ranges were
broader for the environmental samples than for the  PE
samples.  As indicated by the mean and median percent
recoveries, the QC check results generally indicated a
slight negative bias  (up  to  20 percent)  in  the  GRO
concentration measurements; the exceptions were  the
surrogate recoveries for environmental samples and  the
LCS/LCSD recoveries for soil PE samples.  The observed
bias  did  not  exceed the  generally  acceptable bias
(± 30 percent) stated in SW-846 for organic analyses and
is  typical for  most organic  analytical  methods  for
environmental  samples.  Because the percent recovery
ranges were sometimes above and sometimes below 100,
the observed bias did not appear to be systematic.

The project-specific RPD acceptance criterion was met for
all samples. As expected, the RPD range and the mean and
median RPDs  for  MS/MSDs associated with the soil
environmental  samples were greater than those for other
QC checks and matrixes listed in Table 6-1.  The low
RPDs observed indicated good precision in the  GRO
concentration  measurements  made  during  the
demonstration.

6.1.2     EDRO Analysis

This section summarizes the results for QC checks used by
the reference laboratory during EDRO analysis, including
method and  instrument blanks, surrogates, MS/MSDs,
extract duplicates, and LCS/LCSDs.  A  summary of the
QC check results is presented at the end of the  section.

Method and Instrument Blanks

Method and instrument blanks were analyzed to verify that
steps in  the  analytical  procedures  did not  introduce
contaminants that affected analytical results. Ottawa sand
was used as a method blank for soil samples. The method
blanks underwent all the procedures required for sample
preparation. For liquid PE samples, the extraction solvent
(methylene chloride)  was used as an instrument blank.
The results for all method and instrument blanks met the
acceptance criterion of being less than  or equal to  the
required  project-specific  reporting  limit  (10 mg/kg).
Based on the method and instrument blank results,  the
EDRO analysis results were considered to be valid.
                                                   50

-------
i

O
U)

0)
O
o
o
re
D
O

"S


re


E




I
-
ra ^ o O
(D <
^

'" D)
-Q ra
< ce

o c
c o
JS -^
Q. <0
8 o
£ S
d i ^ to g
z  .Hj .2 0
ra -1 ra
(D >
o
0) .c

•ago
2 §0
.C
^
o
 -o E
o c c ra
CO OJ (0 U)

2
CL
CM



S "a.
f!
Q
W
O
CO
O
























































^
ro
o
-Q.
3
T3
(D
Q.
E
S
o
•E
8
£•
° (D
g s
«f
-o V
C (D
ra ^
(L) Q.
-Q. m j-
E||
-^ (/) "ro "ro
— — P ^
O ~m
111" o
°^^l^
gf 'g.| o
llpf
_l _l 2 CL O

II II II II II
Q
.. §g
 =^ ^
-Q . O ^ LLJ O
ra
(5
±±
o
S
c
ro
•5.
ra
o
v=
'o
(D
CL
(/)
'o
,
(0
C
ra
£
o3
3
"w"
(D
Q.
E
S
T3
'^
g-

£
CM
T3
C
(3
(/)
(D
CL
E
(3
(/)

1
£
O
CO
^
c
ra
.a
T3
o

"5
E
CM
C
O
demonstrati
(D
£
D)
c
Q

                                                             51

-------
Surrogates

Each soil investigative and QC sample for EDRO analysis
was spiked with a surrogate, o-terphenyl, before extraction
to determine whether significant matrix effects existed
within the sample and to estimate the efficiency of analyte
recovery during sample preparation and analysis.  For a
30-gram sample, the spike concentration was 3.3 mg/kg.
For samples with higher EDRO concentrations, for which
smaller sample amounts were used during extraction, the
spiking  levels  were  proportionately  higher.   The
acceptance criterion  was 45  to  143 percent surrogate
recovery.  Liquid PE samples for EDRO analysis were not
spiked with a surrogate because the analysis  did not
include a sample preparation step.

A total of 185 surrogate measurements were made during
analysis of environmental and associated QC samples. Six
of these samples did not  meet the percent  recovery
acceptance criterion.   Four of the six samples  were
environmental  samples.  When the reference laboratory
reanalyzed the four samples, the surrogate recoveries for
the samples met the acceptance criterion; therefore, the
reference  laboratory reported  the EDRO concentrations
measured during  the  reanalyses.   The remaining two
samples for which the surrogate recoveries  did not meet
the acceptance criterion were LCS/LCSD samples; these
samples had low surrogate recoveries. According to the
reference laboratory, these low recoveries were due to the
extracts  going  dry  during the extract concentration
procedure. Because  two samples were  laboratory QC
samples, the reference laboratory reanalyzed them as well
as all  the other  samples  in the QC  lot; during the
reanalyses, all surrogate recoveries met the acceptance
criterion. The surrogate recoveries for all results reported
ranged from 45 to 143 percent with mean and median
recoveries of 77 percent, indicating  an overall negative
bias.  The surrogate  recoveries  for all reported sample
results met the  acceptance criterion.   Based on the
surrogate  results for  environmental  and  associated QC
samples, the EDRO analysis results were considered to be
valid.

A total of 190 surrogate measurements were made during
analysis of soil PE and associated QC samples.  Five of
these   samples  did  not  meet the  percent  recovery
acceptance criterion.   In  each case, the surrogate was
recovered at a concentration below the lower limit of the
acceptance criterion.  Three of the five samples were soil
PE samples,  and the remaining  two  samples  were
LCS/LCSDs.  The reference laboratory reanalyzed the
three soil PE samples and the LCS/LCSD pair as well as
all the other samples in the QC lot associated with the
LCS/LCSDs;  during  the  reanalyses,  all  surrogate
recoveries met the acceptance criterion.  The surrogate
recoveries for all  results reported ranged from  46 to
143  percent with  mean and  median  recoveries  of
76 percent,  indicating  an overall negative bias.  The
surrogate recoveries for all reported sample results met the
acceptance criterion.  Based on the surrogate results for
soil PE and  associated QC samples, the EDRO  analysis
results were  considered to be valid.

Matrix Spikes and Matrix Spike Duplicates

MS/MSD results were evaluated to determine the accuracy
and precision of the analytical results with respect to the
effects of the sample matrix. For EDRO analysis, each soil
sample designated as an MS or MSD was spiked  with the
EDRO calibration standard at an initial spiking  level of
50 mg/kg when a 30-gram sample  was  used during
extraction. The initial spiking levels were proportionately
higher when smaller  sample amounts were used during
extraction. The acceptance criteria for MS/MSDs were 46
to 124 percent recovery and an RPD less than or equal to
45.  When  the MS/MSD percent recovery acceptance
criterion was not met, instead of attributing the failure to
meet the  criterion to  an inappropriate spiking level, the
reference  laboratory  respiked the samples at  a  target
spiking level between 50 and 150 percent of the unspiked
sample concentration. Additional information on spiking
level selection for MS/MSDs is presented in Section 6.1.1.
No MS/MSDs were prepared for liquid PE samples for
EDRO analysis because the  analysis  did  not include  a
sample preparation step.

For environmental  samples, a total of 13 MS/MSD pairs
were analyzed.  Two sample pairs collected in the FFA
were designated as  MS/MSDs. The sample matrix in this
area primarily consisted of medium-grained sand.  The
percent recoveries for the MS/MSD samples ranged from
0 to 183 with RPDs of 0 and 19. One of the two sample
pairs  exhibited percent  recoveries less than the  lower
acceptance limit. In the second sample pair, one sample
exhibited  a  percent  recovery  less  than the  lower
acceptance limit, and one sample exhibited a percent
recovery greater than the upper acceptance limit. For both
sample pairs, the  RPD acceptance  criterion  for the
MS/MSDs and the percent recovery and RPD acceptance
criteria for the LCS/LCSDs associated with the analytical
                                                   52

-------
batches for these samples were met. Because of the varied
percent recoveries for the  MS/MSD sample pairs, it was
not possible to conclude whether the EDRO analysis
results for the FFA samples had a negative or positive
bias.  Although the MS/MSD results did not meet the
percent recovery acceptance criterion, the out-of-control
situations alone did not constitute adequate grounds for
rejection of any of the EDRO analysis results for the FFA
samples.  The out-of-control situations may have been
associated with inadequate spiking levels (0.1 to 0.5 times
the unspiked  sample  concentrations  compared  to  the
minimum   recommended  value  of  5  times   the
concentrations).

Four sample pairs collected in the NEX Service Station
Area were designated as MS/MSDs. The sample matrix in
this area primarily consisted of medium-grained sand. The
percent recoveries for the MS/MSD samples ranged from
81 to  109 with RPDs ranging from 4 to 20. The percent
recoveries and RPDs for these samples met the acceptance
criteria.  Based  on the  MS/MSD results, the EDRO
analysis results for the NEX Service Station Area samples
were considered to be valid.

One sample pair collected in the PRA was designated as an
MS/MSD.   The sample  matrix in this area primarily
consisted of silty sand. The percent  recoveries for the
MS/MSD samples were 20 and 80 with an RPD equal to
19. One sample exhibited  a percent recovery less than the
lower acceptance limit, whereas the percent recovery for
the other sample met the acceptance criterion. The RPD
acceptance criterion for the  MS/MSD and the percent
recovery and RPD acceptance criteria for the LCS/LCSD
associated with the analytical batch for this sample pair
were  met.   Although the  percent recoveries for  the
MS/MSD  sample pair may indicate a  negative bias,
because the  MS/MSD results for only one sample pair
were available, it was not possible to conclude that the
EDRO analysis results for the PRA samples had a negative
bias.   Although  one of the  percent recoveries for the
MS/MSD did not meet the acceptance criterion, the out-of-
control situation alone did not constitute adequate grounds
for rejection of any of the EDRO analysis results for the
PRA samples. The out-of-control situation may have been
associated with inadequate spiking levels (0.4 times the
unspiked sample concentration compared to the minimum
recommended value of 5 times the concentration).

Two  sample pairs  collected in the  B-38 Area were
designated as MS/MSDs.  The sample  matrix in this area
primarily  consisted of sand  and clay.   The  percent
recoveries for the MS/MSD samples ranged from 25 to 77
with  RPDs of 6 and 11.  Of the two sample pairs, one
sample pair met the percent recovery acceptance criterion,
and one sample pair exhibited percent recoveries less than
the lower acceptance limit.  For the sample pair that did
not meet the percent recovery  acceptance criterion,  the
RPD  acceptance criterion for the  MS/MSDs and  the
percent recovery and RPD acceptance criteria for  the
LCS/LCSDs associated with the analytical batch for the
sample pair were met. Although the percent recoveries for
one MS/MSD sample pair indicated  a  negative bias,
because the percent recoveries for the other sample pair
were acceptable, it was not possible to conclude that the
EDRO analysis results for the B-38 Area samples had a
negative bias.  Although one-half of the MS/MSD results
did not meet the percent recovery acceptance criterion, the
out-of-control situations alone did not constitute adequate
grounds for rejection of any of the EDRO analysis results
for the B-38 Area samples. The out-of-control situations
may have been associated with inadequate spiking levels
(1.4 times the unspiked sample concentrations compared
to the minimum  recommended value of 5  times  the
concentrations).

Four  sample  pairs  collected in  the  SFT  Area  were
designated as MS/MSDs.  The sample matrix in this area
primarily consisted of silty clay. The percent recoveries
for the MS/MSD samples ranged from 0 to 223 with RPDs
ranging from 8 to 50.  Of the four sample pairs, three
sample pairs had one sample each that exhibited a percent
recovery  less than the lower acceptance limit and one
sample pair had one sample that exhibited a percent
recovery greater than the upper acceptance limit. The RPD
acceptance criterion was met for  all but one of  the
MS/MSDs.  The percent recovery and RPD acceptance
criteria for the LCS/LCSDs associated with the analytical
batches for these samples were met. Because of the varied
percent recoveries for the  MS/MSD sample pairs, it was
not possible to conclude whether the EDRO analysis
results for the SFT Area samples had a negative or positive
bias.  Although one-half of the MS/MSD results did  not
meet the percent recovery acceptance criterion and one of
the four sample pairs did  not meet the RPD acceptance
criterion,  the  out-of-control situations  alone did  not
constitute  adequate  grounds for rejection of any of the
EDRO analysis results for the SFT Area samples. The out-
of-control situations may have  been associated with
inadequate spiking levels  (0.4 to 0.7 times the unspiked
                                                   53

-------
sample  concentrations  compared to  the  minimum
recommended value of 5 times the concentrations).

Five soil PE sample pairs were designated as MS/MSDs.
The sample matrix for these samples primarily consisted of
silty sand.   The percent recoveries for these samples
ranged from 0 to 146 with RPDs ranging from 3 to 17. Of
the five  sample pairs, three sample pairs met the percent
recovery acceptance criterion, one  sample pair exhibited
percent recoveries less than the lower acceptance limit, and
one sample pair exhibited percent recoveries greater than
the upper acceptance limit. For the two sample pairs that
did not meet the percent recovery acceptance criterion, the
RPD  acceptance criterion for the MS/MSDs and the
percent  recovery and  RPD acceptance  criteria for the
LCS/LCSDs associated with the analytical batches for
these samples were met.  Because of the varied percent
recoveries for the  MS/MSD sample  pairs, it was not
possible to conclude whether the EDRO analysis results
for the soil PE samples had a negative or  positive  bias.
Although the percent recoveries for two of the five sample
MS/MSD pairs did not meet the acceptance criterion, the
out-of-control situations alone did not constitute adequate
grounds for rejection of any of the EDRO analysis results
for the soil PE samples.

Extract Duplicates

For EDRO  analysis, after soil sample extraction, extract
duplicates  were  analyzed  to  evaluate the  precision
associated  with  the  reference  laboratory's  analytical
procedure.  The reference laboratory sampled duplicate
aliquots  of the  EDRO  extracts  for  analysis.   The
acceptance criterion for extract duplicate precision was an
RPD less than or equal to 45.  One or more environmental
samples  collected  in each  demonstration  area were
designated as extract duplicates. A total of 13 samples
designated as extract duplicates were analyzed for EDRO.
The RPDs for these samples ranged from 0 to 11 except for
one extract duplicate pair collected in the SFT Area that
had an RPD equal to 34.  The RPDs  for all the extract
duplicates met the acceptance criterion.  Based on the
extract duplicate results, all EDRO results were considered
to be valid.

Laboratory Control Samples and Laboratory
Control Sample Duplicates

For EDRO analysis, LCS/LCSD results were evaluated to
determine the accuracy and precision  associated  with
control samples prepared by the reference laboratory. To
generate a soil LCS or LCSD, Ottawa sand was spiked
with the EDRO calibration standard at a spiking level of
50 mg/kg. The acceptance criteria for LCS/LCSDs were
46 to 124 percent recovery and an RPD less than or equal
to 45. The LCS/LCSD acceptance criteria were based on
the reference laboratory's historical data. No LCS/LCSDs
were prepared for liquid PE samples for EDRO analysis
because the analysis did not include a sample preparation
step.

Twenty-two pairs of LCS/LCSD samples were prepared
and analyzed.  The percent recoveries for these samples
ranged from 47 to 88 with RPDs ranging from 0 to 29.
Therefore, the percent recoveries and RPDs for these
samples met the  acceptance criteria, indicating that the
EDRO analysis procedure was in control.  Based on the
LCS/LCSD  results, the EDRO  analysis  results were
considered to be valid.

Summary of Quality Control Check Results

Table 6-2 summarizes the QC check results for EDRO
analysis.  Based on the QC check results, the conclusions
presented below were drawn regarding the accuracy and
precision of EDRO analysis results for the demonstration.

The project-specific percent recovery acceptance criteria
were met for  all surrogates and LCS/LCSDs.  About
one-half  of the  MS/MSDs did not meet the percent
recovery acceptance criterion. As expected, the MS/MSD
percent recovery range was broader for environmental
samples than for PE samples.  The mean and median
percent recoveries for all the QC check samples indicated
a  negative  bias  (up  to  33  percent)  in the EDRO
concentration measurements. Although the observed bias
was  slightly greater than the  generally acceptable  bias
(±30 percent) stated in SW-846 for organic analyses, the
observed recoveries were not  atypical for most organic
analytical methods for environmental samples.  Because
the percent recovery ranges were sometimes above and
sometimes below 100, the observed bias did not appear to
be systematic.

The project-specific RPD acceptance criterion was met for
all samples except one environmental MS/MSD sample
pair.  As expected, the RPD  range and the mean and
median RPDs for MS/MSDs associated with the  soil
environmental samples were greater than those for other
QC checks and matrixes listed in Table 6-2.  The low
                                                   54

-------
i
Q
UJ

£
U)
0)
a:
o
c
o
o
re
D
O

"5

re

E
CM

(D


"o?
c

o E c | .1
° §£££
-7 -^ (D (D ™
ro ^ o O
0) <



§8,
o ro
< o:

 ^j 8 .h;
ra ^ o O
(D <
^

— GJ
™ D)
o ro
< o:

8
l-i
Q. (D
|5
£ jS
O £ "*"* "l~~l

o ^  Q
ro — ' ro
0) >
^ LU
o
0) _c

•ago
5|I
.
o
—

Q.
ID
0




CO
LJJ 0)
CL "a.

Q
w
w



 -o E
o c c ra
W o) ra «)
Q
W
O
W
O






















































£
(0
o
-Q.
3
T3

-------
RPDs observed indicated good precision in the EDRO
concentration   measurements   made  during   the
demonstration.

6.2    Selected Performance Evaluation Sample
       Results

Soil and liquid PE samples were analyzed during the
demonstration to  document the reference method's
performance  in analyzing  samples  prepared under
controlled conditions. The PE sample results coupled with
the QC check results were used to establish the reference
method's performance  in such a  way that  the  overall
assessment of  the reference  method  would  support
interpretation of the OCMA-350's performance, which is
discussed in Chapter 7. Soil PE samples were prepared by
adding weathered gasoline or diesel to  Ottawa sand or
processed garden soil.  For each sample, an amount of
weathered gasoline or diesel was added to  the  sample
matrix in order to prepare a PE sample with a low (less
than 100 mg/kg), medium (100 to 1,000 mg/kg), or high
(greater than 1,000 mg/kg) TPH concentration. Liquid PE
samples consisted of neat materials. Triplicate samples of
each type of PE sample were analyzed by the reference
laboratory  except  for the low-concentration-range PE
samples, for which seven replicate samples were analyzed.

As described in Section 4.2,  some  PE  samples  also
contained interferents.  Section 6.2 does not discuss the
reference method  results  for  PE  samples  containing
interferents  because the  results address  a  specific
demonstration objective.  To facilitate comparisons, the
reference   method  results  that   directly   address
demonstration objectives are discussed along with the
OCMA-350 results in Chapter 7.  Section 6.2  presents a
comparison of the reference method's mean TPH results
for selected  PE samples to  the  certified values  and
performance  acceptance  limits provided by ERA,  a
commercial PE sample provider that prepared the PE
samples for the demonstration. Although the reference
laboratory reported sample results for GRO  and EDRO
analyses separately, because ERA provided certified values
and performance acceptance limits, the reference method's
mean TPH results (GRO plus EDRO analysis results) were
used for comparison.

For soil samples  containing weathered  gasoline,  the
certified values used for comparison to the reference
method results were based  on mean TPH results for
triplicate samples analyzed  by ERA using a GC/FID
method. ERA extracted the PE samples on the day that PE
samples  were  shipped  to  the  Navy  BVC site  for
distribution to the reference  laboratory and developers.
The reference laboratory completed methanol extraction of
the demonstration samples within 2  days  of receiving
them. Between 5 and 7 days elapsed between the time that
ERA and the time that the reference laboratory completed
methanol extractions of the demonstration samples. The
difference in extraction times is not believed to have had
a significant effect on the reference method's TPH results
because the samples  for GRO analysis were containerized
in EPA-approved EnCores  and were stored at 4 ± 2 °C to
minimize volatilization. After methanol extraction of the
PE samples,  both  ERA and the reference  laboratory
analyzed the  sample  extracts within the  appropriate
holding times for the extracts.

For soil samples containing  diesel, the  certified values
were established by calculating the TPH concentrations
based  on the amounts  of diesel  spiked  into known
quantities of soil; these  samples were not analyzed by
ERA.  Similarly, the densities of the neat materials were
used as the certified  values  for the liquid PE samples.

The performance acceptance limits for soil PE  samples
were based on ERA's historical data on percent recoveries
and RSDs from  multiple laboratories  that had analyzed
similarly prepared ERA PE samples using a GC method.
The performance acceptance limits were determined at the
95 percent confidence level using Equation 6-1.

Performance Acceptance Limits = Certified Value x
(Average Percent Recovery + 2(Average RSD))     (6-1)

According to SW-846, the 95 percent confidence  limits
should be treated as warning limits, whereas the 99 percent
confidence limits should be treated as control limits. The
99 percent confidence limits are calculated by using three
times the average RSD in  Equation 6-1  instead of two
times the average RSD.

When  establishing the performance acceptance limits,
ERA did not account  for variables among the multiple
laboratories, such as different extraction and analytical
methods,  calibration  procedures,  and  chromatogram
integration ranges (beginning and end points).  For this
reason, the performance acceptance limits should be used
with caution.
                                                   56

-------
Performance acceptance limits for liquid PE samples were
not  available because  ERA did  not have  historical
information on percent recoveries and RSDs for the neat
materials used in the demonstration.

Table 6-3 presents the PE sample types, TPH concentration
ranges, performance acceptance limits, certified values,
reference method mean TPH concentrations, and ratios of
reference method mean TPH concentrations to certified
values.

In addition to the samples listed in Table 6-3, three blank
soil PE samples (processed garden soil) were analyzed to
determine whether the soil PE sample matrix contained a
significant TPH concentration.  Reference method GRO
results for all triplicate samples were below the reporting
limit of 0.54 mg/kg. Reference method EDRO  results
were calculated  by adding the results  for DRO and oil
range organics (ORO) analyses.  For one of the triplicate
samples, both the DRO  and ORO results were below the
reporting limits of 4.61 and 5.10 mg/kg, respectively. For
the remaining two triplicates, the DRO and ORO  results
were 1.5 times greater than the reporting limits. Based on
the  TPH   concentrations in the  medium-  and  high-
                              concentration-range soil PE samples listed in Table 6-3,
                              the contribution of the processed garden soil to the TPH
                              concentrations was insignificant and ranged between 0.5
                              and 5 percent.

                              The reference method's mean TPH results for the soil PE
                              samples listed in Table 6-3 were within the performance
                              acceptance limits except for the low-concentration-range
                              diesel samples. For the low-range diesel samples, (1) the
                              individual TPH concentrations for all seven replicates were
                              less than  the lower performance  acceptance  limit and
                              (2) the upper 95 percent confidence limit for TPH results
                              was also less than the lower performance acceptance limit.
                              However, the reference method mean and individual TPH
                              results for the low-range diesel samples were within the
                              99 percent  confidence interval of  10.8 to 54.6 mg/kg,
                              indicating that the reference method results met the control
                              limits but not the warning limits.  Collectively, these
                              observations  indicated   a  negative  bias  in  TPH
                              measurements for low-range diesel samples.

                              As noted above, Table 6-3 presents ratios of the reference
                              method mean TPH concentrations to the certified values
                              for PE samples.  The ratios  for  weathered gasoline-
Table 6-3. Comparison of Soil and Liquid Performance Evaluation Sample Results
Sample Type3
Soil Sample (Ottawa Sand)
   TPH
Concentration
   Range
  Performance
Acceptance Limits
    (mg/kg)
Certified Value
               Reference Method
                  Mean TPH
                 Concentration
Reference Method Mean
  TPH Concentration/
Certified Value (percent)
Diesel
                            Low
                                         18.1 to 47.4
                                                            37.3 mg/kg
                                                    14.7 mg/kg
Notes:

mg/kg = Milligram per kilogram
mg/L  = Milligram per liter

a    Soil samples were prepared at 9 percent moisture unless stated otherwise.
                                                                                                   39
Soil Samples (Processed Garden Soil)
Weathered gasoline

Weathered gasoline at
16 percent moisture
Diesel

Diesel at less than 1 percent
moisture
Medium
High
High
Medium
High
High
196 to 781
1,1 10 to 4,430
992 to 3,950
220 to 577
1 ,900 to 4,980
2,1 00 to 5,490
550 mg/kg
3,120 mg/kg
2,780 mg/kg
454 mg/kg
3,920 mg/kg
4,320 mg/kg
344 mg/kg
2,030 mg/kg
1 ,920 mg/kg
281 mg/kg
2,720 mg/kg
2,910 mg/kg
62
65
69
62
69
67
Liquid Samples
Weathered gasoline
Diesel
High
High
Not available
Not available
81 4,1 00 mg/L
851 ,900 mg/L
648,000 mg/L
1,090,000 mg/L
80
128
                                                     57

-------
containing soil samples ranged from 62 to 69 percent and
did not appear to depend on whether the samples were
medium-  or  high-range  samples.   The  ratio for neat,
weathered gasoline (liquid sample) was 80 percent, which
was 11 to 18  percentage points greater than the ratios for
the soil samples.  The difference  in the ratios may be
attributed to  (1) potential loss of volatiles  during soil
sample  transport and  storage and during soil  sample
handling when extractions were performed and (2) lower
analyte  recovery during soil  sample extraction. The less
than 100 percent ratios observed indicated a negative bias
in  TPH  measurement  for  soil  and  liquid  samples
containing weathered gasoline. The observed bias for the
liquid samples did not exceed the generally acceptable bias
(±30 percent) stated in SW-846 for most organic analyses.
However, the bias for soil samples exceeded the acceptable
bias by up to  8 percentage points.

The ratios for diesel-containing soil samples ranged from
39 to 69 percent and increased with increases in the TPH
concentration range.  The ratio for neat diesel (liquid
sample) was 128 percent, which was substantially greater
than the ratios for soil samples. Collectively, the negative
bias observed for soil  samples and the positive bias
observed  for liquid  samples indicated  a  low  analyte
recovery during soil sample extraction because the soil and
liquid samples were analyzed using the same calibration
procedures but only the soil samples required extraction
before analysis. The extraction procedure used during the
demonstration is an EPA-approved method that is widely
used by commercial laboratories in the United States.
Details  on the  extraction procedure are presented in
Table 5-3 of this ITVR.
The  positive bias  observed for liquid samples did not
exceed the generally acceptable bias stated in SW-846.
The negative bias observed for high-concentration-range
soil samples exceeded the acceptable bias by an average of
2 percentage points. However, the negative bias observed
for low- and  medium-range  samples exceeded the
acceptable  bias  by   31  and  8  percentage  points,
respectively, indicating a negative bias.

Because the reference method results exhibited a negative
bias for soil PE samples when compared to ERA-certified
values, ERA's historical data on percent recoveries and
RSDs  from  multiple   laboratories  were   examined.
Table 6-4 compares ERA's historical percent recoveries
and RSDs to the reference method percent recoveries and
RSDs obtained during the demonstration.  Table 6-4 shows
that ERA's historical recoveries also exhibited a negative
bias  for  all sample types except weathered gasoline in
water and that the reference method recoveries were less
than ERA's historical recoveries  for all sample types
except diesel in water.  The  ratios of reference method
mean recoveries to ERA historical mean recoveries  for
weathered gasoline-containing samples indicated that the
reference method TPH results were 26 percent less than
ERA's historical  recoveries.   The reference method
recoveries for diesel-containing (1) soil samples were
32 percent less than the ERA historical recoveries and
(2) water samples were 63 percent greater than the ERA
historical recoveries.   In all cases, the RSDs for the
reference method were significantly lower than ERA's
historical RSDs,  indicating that the  reference method
achieved  significantly greater  precision.  The  greater
precision observed for the reference method during the
Table 6-4. Comparison of Environmental Resource Associates Historical Results to Reference Method Results
ERA Historical Results
Sample Type
Weathered gasoline in soil
Diesel in soil
Weathered gasoline in water
Diesel in water
Mean Recovery
(percent)
88.7
87.7
109
78.5
Mean Relative
Standard Deviation
(percent)
26.5
19.6
22.0
22.8
Mean
Recovery3
(percent)
65
59
80
128
Reference Method Results
Reference Method Mean
Recovery/ERA Historical
Mean Recovery (percent)
74
68
73
163
Mean Relative
Standard Deviation3
(percent)
8
7
5
6
Notes:

ERA  = Environmental Resource Associates

3    The reference method mean recovery and mean relative standard deviation were based on recoveries and relative standard deviations observed
    for all concentration ranges for a given type of performance evaluation sample.
                                                     58

-------
demonstration may be associated with the fact that the
reference method was implemented by a single laboratory,
whereas ERA's historical RSDs were based on results
obtained from multiple laboratories that may have used
different analytical protocols.

In summary, compared to ERA-certified values, the TPH
results for all PE sample types except neat diesel exhibited
a negative bias to a varying degree; the TPH results for
neat diesel  exhibited a positive bias of 28 percent.  For
weathered gasoline-containing soil samples, the bias was
relatively independent of the TPH concentration range and
exceeded the  generally acceptable bias stated in SW-846
by up to 8 percentage points. For neat gasoline samples,
the bias did not exceed the acceptable bias. For diesel-
containing soil samples, the bias increased with decreases
in the TPH concentration range, and the bias for low-,
medium-,  and  high-range  samples   exceeded  the
acceptable  bias  by  31,  8,  and  2  percentage  points,
respectively.   For neat  diesel  samples,  the  observed
positive bias did not exceed the  acceptable bias. The low
RSDs (5 to  8 percent)  associated  with  the  reference
method indicated good precision in analyzing both soil and
liquid samples. Collectively, these observations suggest
that caution should be exercised during comparisons of
OCMA-350 and reference method results for low- and
medium-range soil samples containing diesel.
6.3    Data Quality

Based on the reference method's performance in analyzing
the QC check  samples and selected PE  samples, the
reference  method results were considered  to  be of
adequate quality  for the  following  reasons: (1)  the
reference  method was implemented with  acceptable
accuracy (±30 percent) for all samples except low- and
medium-concentration-range  soil  samples  containing
diesel, which made up only 13 percent of the total number
of samples analyzed during the demonstration, and (2) the
reference method was implemented with good precision
for all samples (the overall RPD range was 0 to 17).  The
reference method results generally exhibited a negative
bias. However, the bias was considered to be significant
primarily  for  low-  and  medium-range  soil samples
containing diesel because the bias exceeded the generally
acceptable  bias of ±30 percent stated  in SW-846 by
31 percentage points for  low-range  and  8  percentage
points for medium-range samples.  The reference method
recoveries  observed  were  typical   of  the  recoveries
obtained by  most  organic analytical  methods  for
environmental samples.
                                                    59

-------
                                              Chapter 7
                                  Performance of the OCMA-350
To  verify a wide range of performance attributes, the
demonstration had both primary and secondary objectives.
Primary  objectives were  critical  to the  technology
evaluation  and were intended to produce quantitative
results regarding a technology' s performance. Secondary
objectives provided information that was useful but did not
necessarily  produce  quantitative results  regarding  a
technology's performance.  This chapter discusses the
performance  of the OCMA-350  based on the  primary
objectives  (excluding   costs  associated  with   TPH
measurement) and secondary objectives. Costs associated
with  TPH measurement  (primary  objective   P6)  are
presented in Chapter 8. The demonstration results for both
the  primary and secondary objectives are  summarized in
Chapter 9.

7.1    Primary Objectives

This section  discusses  the performance  results for the
OCMA-350 based on primary  objectives PI through P5,
which are listed below.

PI. Determine the MDL

P2. Evaluate  the  accuracy   and  precision  of  TPH
    measurement  for  a variety  of contaminated soil
    samples

P3. Evaluate   the  effect  of  interferents  on   TPH
    measurement

P4. Evaluate the effect  of soil moisture content on TPH
    measurement

P5. Measure the time required for TPH measurement

To  address primary objectives PI through P5,  samples
were  collected from five different  sampling areas.  In
addition, soil and liquid PE samples were prepared and
distributed to Horiba and the reference laboratory.  The
numbers and types of environmental samples collected in
each sampling area and the numbers and  types of PE
samples prepared are discussed in Chapter 4.  Primary
objectives PI through P4 were addressed using statistical
and  nonstatistical  approaches, as  appropriate.   The
statistical tests performed to address these objectives are
illustrated in the flow diagram in Figure 7-1.  Before a
parametric test was performed, the Wilk-Shapiro test was
used to determine  whether the OCMA-350 results and
reference method  results or,  when  appropriate,  their
differences  were normally distributed at a significance
level of 5 percent. If the results or their differences  were
not normally  distributed, the Wilk-Shapiro  test  was
performed on transformed results (for example, logarithm
and square  root transformations) to verify the normality
assumption. If the  normality assumption was not met, a
nonparametric test was  performed. Nonparametric  tests
are not as powerful as  parametric tests  because the
nonparametric tests do not account for the magnitude of
the difference between sample results.   Despite this
limitation, when the normality assumption was not met,
performing a nonparametric test was considered to  be a
better   alternative  than   performing   no  statistical
comparison.

For the OCMA-350, when the TPH concentration in a
given sample was reported as below the reporting limit,
one-half the reporting  limit  was  used as the  TPH
concentration for that sample, as is commonly done, so that
necessary   calculations   could  be  performed  without
rejecting the data.  The  same approach was used for the
reference method except that the appropriate reporting
limits were used in calculating the  TPH concentration
depending  on which TPH measurement  components
(GRO, DRO, and ORO) were reported at concentrations
below  the reporting limits.  Caution was  exercised to
ensure that these necessary data manipulations did not
alter the conclusions.
                                                   60

-------
(O
a>


I
Q_
       OM
       CL
     O O

     'in .92,

     '
ct £•
  ra
0)

15
0)

T3
.2
ro
o
ro
O
c
g
^ro


CL



to

S
in
$
-a
c
ra
comparison of means
(nonparametric) to
determine whether
increase in moisture
content resulted in
increase or decrease in
a>
~3
in
0)
X
CL
                                                                        w ,_ .2
                                                                        •c o =j
                                                            2  S.TS o
                                                                   HI
                                                         •i  0-1
                                                                2.|£i:
                                                                  ra & m °
                                                                E E g   ^
                                                                     =! £
                                                                     ra o
                                                                     0) 0)
                                                          ro  o-S'c o X «> U
                                                          3c.Erora.o2i;
                                                          '  m "^ o C I-T Q_ ZJ
                                                          qj .£

                                                           ro
                                                        "o ^ ro oj
                                                         c '(- Q- c
                                                    •u - 5 -112
                                                    a) o o *" o ro
          g 3
          T3 £

          O X
                                                     r

                                                    •§ >-
                                                    a)
                                                            ra CD
                                                       ro g_
^ o ra a)

I gg-i
                               T3 ^v

                                £ -2 2 T3
                               — "- — C X
                                      .^ oj ro c

                           i S«gE^ I
                           -£ -Y c p oj OJ w
                           -o *- 'c g o ^
                           OJ (/) t i_ c >
                           E =E JB jjj £ „
                               l^^s
                                flj _2 O P
                               CL W •"
                                     ra
                                     01
-a
o
E
T3

 Q_   R -Q -^
                                                                           in
                                                                           ±;
                                               t5 .2
                                               •° ts
                                               O (D
                                                11
                                                ro  E
~ o) >,

Sf t

lit

I g E „

11 s -
E oj 3 j?
"S "S c °
a; T3 g
E o ;s
iP -i-i Q)
t W ^

«?§
                                                                                                       U)
                                                                                                       ±i
                                                   61

-------
The  reference method GRO  results  were adjusted  for
solvent dilution associated with the soil sample moisture
content because the method required use of methanol, a
water-miscible solvent, for extraction of soil samples. In
addition, based  on discussions  with Horiba and  the
reference laboratory, all TPH results for the OCMA-350
and the reference method were rounded to three significant
figures.

7.1.1  Primary Objective PI: Method Detection
       Limit

To determine the MDLs forthe OCMA-350 and reference
method, both Horiba and the reference laboratory analyzed
seven low-concentration-range soil PE samples containing
weathered gasoline and seven low-concentration-range soil
PE samples containing diesel.  As discussed in Chapter 4,
problems  arose  during preparation  of the low-range
weathered gasoline samples; therefore, the results for the
soil PE samples containing weathered gasoline could  not
be used to determine MDLs.

Because the OCMA-350 and  reference  method results
were both normally distributed, the MDLs for the soil PE
samples  containing  diesel   were  calculated using
Equation  7-1  (40  CFR Part   136,  Appendix  B,
Revision 1.1.1). An MDL thus calculated is influenced by
TPH concentrations because the standard deviation will
likely decrease with a decrease in TPH concentrations.  As
a result, the MDL will be lower when low-concentration
samples are  used for MDL determination.  Despite this
limitation, Equation 7-1 is commonly used and provides a
reasonable estimate of the MDL.
Table 7-1.  TPH Results for Low-Concentration-Range Diesel Soil
         Performance Evaluation Samples
                MDL = (S)t
                           n-1, 1-«=0.99)
(7-1)
where
    S  =  Standard deviation of replicate TPH results

    Vi. 1-0=0.99) =  Student's t-Value appropriate for a
                  99  percent confidence level and a
                  standard deviation estimate with n-1
                  degrees of freedom (3.143 for n = 7
                  replicates)

Because GRO compounds were not expected to be present
in the soil  PE samples containing  diesel, the reference
laboratory  performed  only  EDRO  analysis of these
samples and reported the  sums of the DRO and ORO
concentrations as the TPH results.  The OCMA-350 and
reference method results for these samples are presented in
Table 7-1.
OCMA-350 Result (mg/kg)
                                   Reference Method Result (mg/kg)
28.8
21.0
30.6
24.2
33.4
23.2
32.2
16.4
16.4
13.2
16.0
14.2
14.1
12.8
MDL
         15.2
                                     4.79
Notes:

MDL   =  Method detection limit
mg/kg =  Milligram per kilogram
Based on the TPH results for the low-concentration-range
diesel soil PE samples, the MDLs were determined to be
15.2 and 4.79 mg/kg for the OCMA-350 and reference
method, respectively. Because the ORO concentrations in
all these samples were below the reference laboratory's
estimated reporting limit (5.1 mg/kg), the  MDL for the
reference  method was also calculated using only  DRO
results.  The MDL for the reference method based on the
DRO results was 4.79 mg/kg, which was the same as the
MDL forthe reference method based on the EDRO results,
indicating  that  the ORO  concentrations  below the
reporting limit did not impact the MDL for the reference
method.

The MDL of 15.2 mg/kg for the OCMA-350 was greater
than the MDL of 1 mg/kg claimed by Horiba. The MDL
of 4.79  mg/kg for the reference method compared well
with the  MDL  of 4.72  mg/kg  published in  SW-846
Method  8015C  for diesel samples  extracted using a
pressurized fluid extraction method  and  analyzed for
DRO.

7.1.2  Primary Objective P2: Accuracy and
       Precision

This section discusses the ability of the OCMA-350 to
accurately and precisely measure TPH concentrations in a
variety  of contaminated  soils.  The  OCMA-350  TPH
results were  compared  to  the reference method  TPH
results.    Accuracy and  precision  are   discussed in
Sections 7.1.2.1 and 7.1.2.2, respectively.
                                                    62

-------
7.1.2.1     Accuracy

The accuracy of OCMA-350 measurement of TPH was
assessed by determining

•  Whetherthe conclusion reached using the OCMA-350
   agreed with that reached using the reference method
   regarding whether the TPH concentration in a given
   sampling area or soil type exceeded a specified action
   level

   Whether the OCMA-350 results were biased high or
   low compared to the reference method results

•  Whether the OCMA-350 results were different from
   the reference method results at a statistical significance
   level of 5 percent when a pairwise  comparison was
   made

•  Whether a significant correlation existed between the
   OCMA-350 and reference method results

During examination of these four factors, the data quality
of the reference method and OCMA-350 TPH results was
considered.  For example, as discussed in Chapter 6, the
reference  method generally  exhibited  a  low  bias.
However,  the bias observed for all samples except low-
and medium-concentration-range diesel  soil samples did
not exceed the generally acceptable bias of ±30 percent
stated in SW-846 for organic analyses.  Therefore, caution
was exercised during comparison of the  OCMA-350 and
reference method results, particularly those for low- and
medium-range diesel  soil  samples.   Caution was also
exercised   during   interpretation  of  statistical  test
conclusions  drawn based on a small number of samples.
For example, only three samples were used  for each type
of PE sample except the low-range diesel  samples; the
small number of samples used increased the probability
that the results being compared would be found to  be
statistically the same.

As discussed in Chapter 2,  during the demonstration, the
OCMA-350 was calibrated using diesel. The following
sections discuss how the OCMA-350 results compared
with the reference method results by addressing each of the
four factors identified above.

Action Level Conclusions

Table 7-2 compares action level conclusions reached using
the  OCMA-350  and  reference  method results for
environmental and soil PE samples.  Section 4.2 of this
ITVR explains how the action levels were selected for the
demonstration.    Of the  environmental  samples,  the
percentage of samples for which the conclusions agreed
ranged from 71 to 100. Of the PE samples, the percentage
of samples for which the conclusions agreed ranged from
0 to 100.  Overall, the conclusions were the same for
73 percent of the samples.

The  least agreement  observed for the environmental
samples was  for those from the SFT Area (71 percent).
Regarding the PE samples, none of the OCMA-350 action
level  conclusions  agreed  with those  of the  reference
method for the medium-concentration-range weathered
gasoline and  diesel samples and the  high-concentration-
range diesel samples with 9 percent moisture content.  The
0 percent agreement observed for these samples cannot be
explained.

When the action level conclusions did not agree, the TPH
results were  further interpreted to  assess  whether the
OCMA-350 conclusion was conservative. The OCMA-
350 conclusion was considered to be conservative when
the OCMA-350 result was  above the action level and the
reference method result was below the action level. A
regulatory agency would likely favor a field measurement
device whose results are conservative; however, the party
responsible for a site cleanup might not favor a device that
is overly conservative because of the cost associated with
unnecessary  cleanup.   For  the  OCMA-350,  29  of
107  action level conclusions  did  not agree with the
reference method conclusions.  Of the 29 OCMA-350
conclusions,  11 of the 13 conclusions for environmental
samples and 12 of the 16 conclusions for PE samples were
not conservative.

Measurement Bias

To determine the  measurement bias,  the ratios of the
OCMA-350 TPH results to the reference method TPH
results were calculated.  The observed bias values were
grouped to identify the number of  OCMA-350 results
within the following  ranges  of the reference method
results: (1) greaterthan 0 to 30 percent, (2) greaterthan 30
to 50 percent, and (3) greater than 50 percent.

Figure 7-2 shows the distribution of measurement bias for
the environmental samples. Of the  five sampling areas,
good agreement between the OCMA-350 and reference
method results was observed for samples collected from
the FFAandNEX Service Station Area;  for these samples,
                                                    63

-------
Table 7-2. Action Level Conclusions
Sampling Area or Sample Type
Fuel Farm Area
Naval Exchange Service Station Area
Phytoremediation Area
B-38 Area
Slop Fill Tank Area
PE sample
PE sample
Soil PE sample
containing
weathered
gasoline in
Soil PE sample
containing diesel in
Blank soil
(9 percent moisture content)
Blank soil and humic acid
(9 percent moisture content)
Medium-concentration range
(9 percent moisture content)
High-concentration range
(9 percent moisture content)
High-concentration range
(16 percent moisture content)
Low-concentration range
(9 percent moisture content)
Medium-concentration range
(9 percent moisture content)
High-concentration range
(less than 1 percent moisture
content)
High-concentration range
(9 percent moisture content)
Action
Level
(mg/kg)
100
50
1,500
100
500
10
200
200
2,000
2,000
15
200
2,000
2,000
Total
Total Number
of Samples
Analyzed
10
20
8
8
28
3b
6
3
3
3
7
3
3
3
108
Percentage of Samples for
Which OCMA-350 and
Reference Method
Conclusions Agreed
100
95
75
75
71
100
100
0
33
67
43
0
100
0
73
When Conclusions Did Not Agree,
Were OCMA-350 Conclusions
Conservative or Not
Conservative?3

Not conservative
Not conservative (one of two
conclusions for each area)
Not conservative

Not conservative
Conservative
Not conservative

Not conservative

Notes:

mg/kg = Milligram per kilogram
PE    = Performance evaluation


    the action level. A conservative conclusion may also be viewed as a false positive.
b    Action level conclusions could be drawn for only two of three samples; the OCMA-350 TPH result for the remaining sample was reported as a "less
    than" value—specifically, less than 15.2 mg/kg, which was greater than the action level.
80 and 85 percent of the OCMA-350 results, respectively,
were within 50 percent of the reference method results.
Moderate agreement between the device and reference
method results was observed for samples collected from
the PRA and B-38 Area;  for these  samples,  62 and
50 percent of the OCMA-350 results, respectively, were
within 50 percent of the reference method results.  For
samples collected from the SFT Area, 28 percent of the
OCMA-350 results were within 50 percent of the reference
method results.

Most of the OCMA-350 results were biased low except
those for the PRA and FFA samples. Low and high biases
were split equally for the samples collected from the PRA.
The results for 70 percent of the samples from the FFA
were  biased  high, perhaps because  of the significant
negative bias associated with reference method results for
diesel soil PE samples (see Chapter 6).  The reason that
most of the OCMA-350 results for the other sampling
areas were biased low was unclear.

Figure 7-3 shows the distribution of measurement bias for
selected soil  PE samples.  Of the five sets of samples
containing  PHCs  and the  one set of blank  samples,
moderate  agreement  between  the  OCMA-350 and
reference method  results was observed  for the  blank
samples and  the  high-concentration-range weathered
gasoline and diesel samples; for these samples, 67,50, and
                                                      64

-------
                         Fuel Farm Area
                    Total number of samples: 10
              >0 to 30
                             >30 to 50
                           Bias, percent
                                                >50
B-38 Area
Total number of samples: 8
Q
< «,
2 ±: ^
P
*£,
.0 1- 1
3
Z












r
>0 to 30 >30 to 50 >50
Bias, percent
               Naval Exchange Service Station Area
                   Total number of samples: 20
              >0 to 30
                             >30 to 50
                           Bias, percent
                                                >50
Slop Fill Tank Area
Total number of samples: 28
T)
*? 15
i|
° 1 10
ii- E iu
II
El=.
3
0 -













^m
>0 to 30 >30 to 50 >50
Bias, percent
                    Phytoremediation Area
                   Total number of samples: 8
              >0 to 30
                             >30 to 50
                           Bias, percent
                                                >50
                                                                   Notes:

                                                                   > = Greater than

                                                                   I  I   OCMA-350  result biased  low  compared to reference
                                                                        method result
                                                                   •   OCMA-350 result biased high compared to reference
                                                                        method result
Figure 7-2. Measurement bias for environmental samples.
                                                              65

-------
                          Blank soil
                   Total number of samples: 3
              >0 to 30
                              >30 to 50
                            Bias, percent
                                                >50
  Diesel in low-concentration range
      Total number of samples: 7
>0 to 30
                >30 to 50
              Bias, percent
                                   >50
 O
                     Weathered gasoline in
                  medium-concentration range
                   Total number of samples: 3
              >0 to 30          >30 to 50           >50
                            Bias, percent
Diesel in medium-concentration range
Total number of samples: 3
2
3
«?
in
IF
0> 0-
£ i-
E
3
Z










































>0 to 30 >30 to 50 >50
Bias, percent
Weathered gasoline in
high-concentration range
Total number of samples: 6
§
i"
O 3
o 8
* =
!r, £ „
£ H 1
1
z
Q











>0 to 30 >30 to 50 >50
Bias, percent
                                                                               Diesel in high-concentration range
                                                                                   Total number of samples: 6
                                                                             >0 to 30
                                                                                             >30 to 50
                                                                                           Bias, percent
                                                                                                                >50
Notes:    > = Greater than; EH OCMA-350 result biased low compared to reference method result;
         compared to reference method result
               OCMA-350 result biased high
Figure 7-3.  Measurement bias for soil performance evaluation samples.
                                                            66

-------
50 percent of the OCMA-350 results, respectively, were
within 50 percent of the reference method results.  For
medium-range diesel samples, 33 percent of the OCMA-
350 results were within 50 percent of the reference method
results.  For the OCMA-350, only one of the seven low-
range diesel sample results was within 30 percent of the
reference  method result, and the  remaining six  sample
results exhibited a bias  of greater than 50 percent.  All
three  OCMA-350 results  for medium-range  weathered
gasoline   samples exhibited  a bias  of  greater than
50 percent.

The  OCMA-350 results were  biased high for  all PE
samples except the high-concentration-range  weathered
gasoline samples, the medium-concentration-range diesel
samples,  and  three   high-concentration-range   diesel
samples.  The high bias  observed for the OCMA-350
results for low-range diesel samples might be explained by
the significant negative bias associated with the reference
method re suits for low-range die sel sample s. However, the
high bias of greater than 50 percent observed  for the
remaining OCMA-350 results cannot be  explained based
solely on the negative bias associated with the reference
method (see Chapter 6).

Pairwise Comparison of TPH Results

To evaluate whether a statistically significant difference
existed between the OCMA-350 and  reference method
TPH  results,  a parametric test  (a  two-tailed,   paired
Student's  t-test)  or a nonparametric test (a Wilcoxon
signed rank test) was selected based on  the approach
presented in Figure  7-1.  Tables  7-3  and 7-4  present
statistical  comparisons of the OCMA-350  and reference
method results  for  environmental  and  PE samples,
respectively.  The tables present the OCMA-350 and
reference  method results for each sampling area or PE
sample  type, the statistical test performed and the
associated null hypothesis used to compare the  results,
whether the results were  statistically the same or different,
and the probability that the results were the same.

Table 7-3 shows that  the OCMA-350 and  reference
method results were statistically the same at a significance
level of 5 percent for all sampling areas except the SFT
Area. Based on a simple comparison of the results, this
conclusion appeared to be reasonable for  all sampling
areas.  For example, for the SFT Area,  the OCMA-350
results were biased low by up to one order of magnitude
for 26 of 28 samples, which justified the  statistical test
conclusion.
Table  7-4  shows  that the  OCMA-350  and  reference
method results were statistically the same at a significance
level of 5 percent for blank soil PE samples and neat
weathered gasoline PE samples.  The OCMA-350  and
reference method results for all other PE sample types
were statistically different. Based on a simple comparison
of the results, these conclusions appeared to be reasonable,
except for high-concentration-range weathered gasoline
(16 percent moisture content) and high-concentration-
range  diesel (less than  1  percent  moisture  content).
However, the  statistical conclusions  for high-range
weathered gasoline (16  percent moisture content)  and
diesel (less than 1 percent moisture content) are justified
considering that the OCMA-350 exhibited a consistent bias
and that the differences between the OCMA-350  and
reference method TPH results were considerably different
than zero.

Of the OCMA-350 PE sample results that were statistically
different  from  the  reference  method  results,  the
OCMA-350 results for (1) all soil samples containing low-
concentration-range diesel, and high-concentration-range
diesel (less than 1 percent moisture content) were biased
high by  up to  a  factor of two;  (2) all soil samples
containing medium- and high-concentration-range diesel
(9 percent moisture  content) and medium-  and high-
concentration-range weathered gasoline were biased low
by up to a factor of three; and (3) neat diesel samples were
biased low by up to 35 percent. In addition, the OCMA-
350 results  for the liquid PE  samples were biased  low
when compared to the sample densities. Specifically, the
OCMA-350 results were biased low by 32 percent for neat
weathered gasoline samples and by 13 percent for neat
diesel samples.

Correlation of TPH Results

To  determine whether a significant  correlation existed
between  the OCMA-350 and reference  method  TPH
results, linear regression analysis was performed. A strong
correlation between the OCMA-350 and reference method
results would indicate that the device results could be
adjusted using the established correlation and that field
decisions could be made using the  adjusted  results in
situations where the device results may not be the same as
off-site laboratory results. Figures 7-4 and 7-5 show the
linear  regression plots for environmental  and soil PE
samples, respectively. Table 7-5 presents the  regression
model, square  of the correlation coefficient  (R2),  and
probability that the slope of the regression line is equal to
                                                     67

-------
Table 7-3. Statistical Comparison of OCMA-350 and Reference Method TPH Results for Environmental Samples
Sampling Area
Fuel Farm Area
Naval Exchange
Service Station
Area
Phytoremediation
Area
TPH Result (mg/kg)
OCMA-350
87.6
14,600
69.1
17,200
94.4
15,300
1,570
8,990
44.4
14,400
32.6
241
111
360
314
1,440
1,530
Less than 15.2
380
1,540
1,100
Less than 15.2
266
997
1,200
Less than 15.2
43.2
1,840
2,390
Less than 15.2
2,050
3,070
2,840
786
798
1,590
771
1,480
Reference
Method
68.2
15,000
90.2
12,000
44.1
13,900
1,330
8,090
93.7
12,300
28.8
144
617
293
280
1,870
1,560
9.56
270
881
1,120
14.2
219
1,180
1,390
15.2
54.5
2,570
3,030
15.9
2,140
1,790
1,390
1,420
1,130
1,530
1,580
1,300
Statistical Analysis Summary
Statistical Test
and Null Hypothesis
Statistical Test
Wilcoxon signed rank test
(nonparametric)
Null Hypothesis
The median of the differences
between the paired observations
(OCMA-350 and reference method
results) is equal to zero.
Statistical Test
Two-tailed, paired Student's t-test
(parametric)
Null Hypothesis
The mean of the differences
between the paired observations
(OCMA-350 and reference method
results) is equal to zero.
Statistical Test
Two-tailed, paired Student's t-test
(parametric)
Null Hypothesis
The mean of the differences
between the paired observations
(OCMA-350 and reference method
results) is equal to zero.
Were OCMA-350 and Reference
Method Results Statistically the
Same or Different?
Same
Same
Same
Probability of Null
Hypothesis Being
True (percent)
10.54
72.75
65.14
                                                          68

-------
Table 7-3.  Statistical Comparison of OCMA-350 and Reference Method TPH Results for Environmental Samples (Continued)
Sampling Area
B-38 Area
Slop Fill Tank
Area
TPH Result (mg/kg)
OCMA-350
47.5
31.4
30.4
48.6
80.6
118
28.8
20.6
Less than 15.2
111
92.0
271
44.7
Less than 15.2
Less than 15.2
Less than 15.2
1,560
571
389
576
434
172
146
156
240
78.5
96.7
16.7
510
123
254
20.4
35.1
1,540
235
52.6
Reference
Method
79.0
41.5
61.4
67.3
193
69.4
43.8
51.6
105
269
397
339
6.16
37.1
43.9
52.4
3,300
1,270
588
554
834
501
280
185
1,090
544
503
146
938
517
369
253
151
3,960
1,210
121
Statistical Analysis Summary
Statistical Test
and Null Hypothesis
Statistical Test
Two-tailed, paired Student's t-test
(parametric)
Null Hypothesis
The mean of the differences
between the paired observations
(OCMA-350 and reference method
results) is equal to zero.
Statistical Test
Two-tailed, paired Student's t-test
(parametric)
Null Hypothesis
The mean of the differences
between the paired observations
(OCMA-350 and reference method
results) is equal to zero.
Were OCMA-350 and Reference
Method Results Statistically the
Same or Different?
Same
Different
Probability of Null
Hypothesis Being
True (percent)
14.99
0.00
Note:



mg/kg = Milligram per kilogram
                                                           69

-------
Table 7-4.  Statistical Comparison of OCMA-350 and Reference Method TPH Results for Performance Evaluation Samples
Sample Type
TPH Result
OCMA-350
Reference
Method
Statistical Analysis Summary
Statistical Test
and
Null Hypothesis
Were OCMA-350 and
Reference Method
Results Statistically the
Same or Different?
Probability of Null
Hypothesis Being
True (percent)
Soil Samples (Processed Garden Soil) (TPH Results in Milligram per Kilogram)
Blank (9 percent moisture content)
Weathered
gasoline
Diesel
Medium-concentration range
(9 percent moisture content)
High-concentration range
(9 percent moisture content)
High-concentration range
(16 percent moisture
content)
Low-concentration range
(9 percent moisture content)
Medium-concentration range
(9 percent moisture content)
High-concentration range
(9 percent moisture content)
High-concentration range
(less than 1 percent moisture
content)
Less than 15.2
20.9
18.7
107
108
104
657
595
638
1,460
1,710
1,720
28.8
21.0
30.6
24.2
33.4
23.2
32.2
124
143
144
992
995
958
3,540
3,570
3,460
5.12
13.1
13.5
350
346
336
1,880
2,020
2,180
1,740
1,980
2,050
16.4
16.4
13.2
16
14.2
14.1
12.8
276
273
295
2,480
2,890
2,800
2,700
2,950
3,070
Statistical Test
Two-tailed, paired
Student's t-test
(parametric)
Null Hypothesis
The mean of the
differences between the
paired observations
(OCMA-350 and
reference method
results) is equal to zero.
Same
Different
Different
Different
Different
Different
Different
Different
7.83
0.02
0.44
0.40
0.12
0.25
0.53
4.16
Liquid Samples (Neat Materials) (TPH Results in Milligram per Liter)
Weathered gasoline
Diesel
541 ,000
566,000
560,000
750,000
737,000
758,000
656,000
61 1 ,000
677,000
1 ,090,000
1 ,020,000
1,160,000
Statistical Test
Two-tailed, paired
Student's t-test
(parametric)
Null Hypothesis
The mean of the
differences between the
paired observations
(OCMA-350 and
reference method
results) is equal to zero.
Same
Different
5.99
1.00
                                                          70

-------
             Comparison of Fuel Farm Area results
                     5,000     10,000    15,000    20,000
                  Reference method TPH result (mg/kg)
                                                                o
     Comparison of B-38 Area results
        50      100      150      200      250
      Reference method TPH result (mg/kg)
           Comparison of Naval Exchange Service Station
                          Area results
       3,000-
                     1,000     2,000      3,000      4,000
                  Reference method TPH result (mg/kg)
Comparison of Slop Fill Tank Area results
      1,000    2,000    3,000   4,000    5,000
     Reference method TPH result (mg/kg)
           Comparison of Phytoremediation Area results
       3,500
                    500    1,000    1,500    2,000   2,500
                  Reference method TPH result (mg/kg)
                                                                         Notes:
                                                                         mg/kg = Milligram per kilogram
                                                                         R2    - Square of the correlation coefficient
Figure 7-4. Linear regression plots for environmental samples.
                                                            71

-------
     2,000
              Comparison of weathered gasoline
            performance evaluation sample results
                 500     1,000    1,500   2,000   2,500
               Reference method TPH result (mg/kg)
     4,000
     3,000
                  Comparison of diesel
           performance evaluation sample results
 M
     2,000
     1,000
               500   1,000   1,500  2,000   2,500  3,000
               Reference method TPH result (mg/kg)
        Comparison of weathered gasoline and diesel
          performance evaluation sample results
     4,000
                  1,000      2,000      3,000     4,000
               Reference method TPH result (mg/kg)
Notes:

mg/kg  =  Milligram per kilogram
R2    =  Square of the correlation coefficient

Figure 7-5. Linear regression plots for soil performance evaluation
         samples.
zero (F-test probability) for each sampling area and soil PE
sample type.

Table 7-5 shows that R2 values for  (1) environmental
samples except PRA and B-38 Area samples ranged from
0.90 to 0.96 and (2) soil PE samples ranged from 0.53 to
0.68.  The R2 values for the PRA and B-38 Area samples
were  0.19  and 0.23,  respectively.   The R2  values  for
separate  regression models for weathered  gasoline and
diesel soil PE samples were lower and higher, respectively,
than the  R2 value for a combined regression model for
these  PE samples.   For environmental samples,  the
probabilities  of the slopes of the regression lines being
equal to zero were 0.00 percent for the FFA, NEX Service
Station Area, and SFT Area samples; 27.92 percent for the
PRA  samples; and 22.58 percent for the B-38  Area
samples. These probabilities  indicated that there was (1)
a less than 5 percent probability for the FFA, NEX Service
Station Area, and SFT Area samples and (2) a greater than
5 percent probability for the PRA and B-38 Area samples
that the  OCMA-350  and  reference  method  results
correlated only  by chance.  For soil PE  samples,  the
probabilities  of the slopes of the regression lines being
equal to  zero were 0.01 and  0.00 percent for the diesel
samples and the combined weathered gasoline and diesel
samples, respectively, and  62.68 percent for the weathered
gasoline  samples. These probabilities indicated that there
was (1) a less than 5 percent probability for the diesel
samples and the combined weathered gasoline and diesel
samples and (2) agreaterthan 5 percent probability forthe
weathered  gasoline samples  that the OCMA-350 and
reference method results correlated only by chance.

Based on the R2 and probability values, the OCMA-350
and reference method results were considered to be (1)
highly correlated for FFA, NEX Service Station Area, and
SFT Area samples; (2)  moderately correlated for diesel
and combined weathered gasoline and  diesel soil  PE
samples; and (3) weakly  correlated for PRA and  B-38
Area samples and for weathered gasoline soil PE samples.
                                                        7.1.2.2
           Precision
Both  environmental and PE samples were analyzed to
evaluate the precision associated with TPH measurements
using the OCMA-350 and reference method. The results
of this evaluation are summarized below.
                                                     72

-------
Table 7-5. Summary of Linear Regression Analysis Results
                                      Regression Model
                                  (y = OCMA-350 TPH result,
        Square of Correlation
 Probability That Slope of
Regression Line Was Equal
Sampling Area or Sample Type
x = reference method TPH result)
Coefficient
to Zero (percent)
Environmental Samples
Fuel Farm Area
Naval Exchange Service Station Area
Phytoremediation Area
B-38 Area
Slop Fill Tank Area
y = 1.13x + 128.27
y = 0.77x + 123.21
y = 1 .28x - 286.80
y = 0.32x + 26.13
y = 0.41x + 2.22
0.96
0.92
0.19
0.23
0.90
0.00
0.00
27.92
22.58
0.00
Soil Performance Evaluation Samples
Weathered gasoline
Diesel
Weathered gasoline and diesel
y = 0.59x - 61 .65
y = 0.82x - 33.84
y = 0.77x - 93.46
0.53
0.68
0.64
62.68
0.01
0.00
Environmental Samples

Blind field triplicates were analyzed to evaluate the overall
precision of the sampling, extraction, and analysis steps
associated with TPH  measurement.  Each  set of field
triplicates was collected from a well-homogenized sample.
Also,  extract duplicates  were  analyzed to  evaluate
analytical precision only.  Each set of extract duplicates
was  collected by  extracting a given soil  sample and
collecting two aliquots  of the  extract.   Additional
information  on  field triplicate  and extract  duplicate
preparation is included in Chapter 4.

Tables 7-6 and 7-7 present the OCMA-350 and reference
method results for field triplicates and extract duplicates,
respectively. Precision was estimated using RSDs for field
triplicates and RPDs for extract duplicates.

Table 7-6 presents  the TPH results and RSDs for 12 sets
of field triplicates  analyzed using the OCMA-350 and
reference  method.    For the   OCMA-350, the  RSD
calculated for one  set of field  triplicates from the SFT
Area (field  triplicate set 12)  was not considered in
evaluating the device's precision because the TPH result
for one of the three  samples (156 mg/kg) was one order of
magnitude  higher  than  the results for  the other two
samples and  was  thus considered to be an analytical
outlier.  The  RSDs for the remaining 11 triplicate sets
ranged from 0 to 49 percent with a median of 20 percent.
For the reference method, the  RSDs ranged from 4 to
39 percent with a median of 18 percent.  Comparison of
the OCMA-350 and  reference method  median  RSDs
showed that the  OCMA-350 exhibited  less overall
precision than the reference method. The OCMA-350 and
reference method RSDs did not exhibit consistent trends
based on  soil type, PHC contamination type, or TPH
concentration.

Table 7-7 presents the TPH results and RPDs for 13 sets of
extract duplicates analyzed using the OCMA-350 and
reference method. For the OCMA-350, the RPDs ranged
from  0 to 3  with a median of  1.  The RPDs  for the
reference method ranged from 0 to 11 with a median of 4.
Comparison of the median RPDs for the OCMA-350 and
reference method indicated that the OCMA-350 achieved
a higher level of precision than the reference method. The
OCMA-350 and  reference method RPDs did not exhibit
consistent trends based on PHC  contamination type or
TPH concentration.  As expected, the median RPDs for
extract duplicates were less than the median RSDs for field
triplicates for both the OCMA-350 and reference method.
These  findings  indicated that greater precision  was
achieved  when  only the analysis  step   could  have
contributed to TPH measurement error than when all three
steps  (sampling, extraction, and  analysis)  could have
contributed to such error.

Performance Evaluation Samples

Table 7-8 presents the OCMA-350 and reference method
TPH results and RSDs for eight sets of replicates for soil
PE samples  and two sets  of  triplicates for liquid  PE
samples.

For the OCMA-350, the RSD calculated for the blank soil
samples was  not considered in evaluating the device's
precision because one of the  three blank  soil  sample
results was  below  the   MDL (15.2  mg/kg) and the
                                                    73

-------
Table 7-6. Summary of OCMA-350 and Reference Method Precision for Field Triplicates of Environmental Samples
Sampling Area
Fuel Farm Area
Naval Exchange Service
Station Area
Phytoremediation Area
B-38 Area
Slop Fill Tank Area
Field Triplicate
Set
1
2
3
4
5
6
7
8
9
10
11
12
OCMA-350
TPH Result
(milligram per kilogram)
87.6
69.1
94.4
1 4,600
17,200
15,300
314
380
267
1,440
1,540
997
1,530
1,100
1,200
Less than 15.2
Less than 15.2
Less than 15.2
2,050
3,070
2,840
47.5
30.4
48.6
434
240
510
172
78.5
123
146
96.7
254
156
16.7
20.4
Relative Standard
Deviation (percent)
16
9
18
22
18
0
20
24
35
38
49
123
Reference Method
TPH Result
(milligram per kilogram)
68.2
90.2
44.1
15,000
12,000
13,900
280
270
219
1,870
881
1,180
1,560
1,120
1,390
9.56
14.2
15.2
2,140
1,790
1,390
79
61.4
67.3
834
1,090
938
501
544
517
280
503
369
185
146
253
Relative Standard
Deviation (percent)
34
11
13
39
16
23
21
13
14
4
29
28
                                                          74

-------
Table 7-7. Summary of OCMA-350 and Reference Method Precision for Extract Duplicates
Sampling Area
Fuel Farm Area
Naval Exchange Service
Station Area
Phytoremediation Area
B-38 Area
Slop Fill Tank Area
Extract
Duplicate
Set
1
2
3
4
5
6
7
8
9
10
11
12
13
OCMA-350
TPH Result
(milligram per kilogram)
94.6
94.2
15,300
Not analyzed3
266
267
997
Not analyzed3
1,200
Not analyzed3
Less than 15.2
Less than 15.2
3,090
3,060
48.2
46.8
32.5
28.6
33.2
439
429
174
170
147
145
156
Not analyzed3
Relative Percent
Difference
0
Not calculated3
0
Not calculated3
Not calculated3
0
1
3
2 percent"
2
2
1
Not calculated3
Reference Method
TPH Result
(milligram per kilogram)
44.1
44.1
13,700
14,000
226
213
1,190
1,170
1,420
1,360
15.5
14.9
1,710
1,860
79.6
78.4
41.4
41.5

829
838
528
473
271
289
189
181
Relative Percent
Difference
0
2
6
2
4
4
8
2
0
1
11
6
4
Notes:

3     The extract was disposed of before an extract duplicate sample was analyzed; therefore, a relative percent difference could not be calculated.

b     Two extract duplicates of the sample extract were analyzed. Therefore, the relative standard deviation is reported instead of the relative percent
     difference.
                                                                 75

-------
Table 7-8. Comparison of OCMA-350 and Reference Method Precision for Replicate Performance Evaluation Samples
Sample Type
Replicate Set
OCMA-350
TPH Result
Relative Standard
Deviation (percent)
Reference Method
TPH Result
Relative Standard
Deviation (percent)
Soil Samples (Processed Garden Soil) (TPH Results in Milligram per Kilogram)
Blank (9 percent moisture content)
Weathered
gasoline
Diesel
Medium-range TPH
concentration
(9 percent moisture
content)
High-range TPH
concentration
(9 percent moisture
content)
High-range TPH
concentration
(16 percent moisture
content)
Low-range TPH
concentration
(9 percent moisture
content)
Medium-range TPH
concentration
(9 percent moisture
content)
High-range TPH
concentration
(9 percent moisture
content)
High-range TPH
concentration (less
than 1 percent
moisture content)
1
2
3
4
5
6
7
8
Less than 15.2
20.9
18.7
107
108
104
657
595
638
1,460
1,710
1,720
28.8
21
30.6
24.2
33.4
23.2
32.2
124
143
144
992
995
958
3,540
3,570
3,460
45
2
5
9
17
8
2
2
5.12
13.1
13.5
350
346
336
1,880
2,020
2,180
1,740
1,980
2,050
16.4
16.4
13.2
16.0
14.2
14.1
12.8
276
273
295
2,480
2,890
2,800
2,700
2,950
3,070
45
2
7
8
10
4
8
6
Liquid Samples (Neat Materials) (TPH Results in Milligram per Liter)
Weathered
gasoline

Diesel
9
10
541 ,000
566,000
560,000
750,000
737,000
758,000
2
1
656,000
611,000
677,000
1 ,090,000
1 ,020,000
1,160,000
5
6
                                                          76

-------
remaining two results were above the MDL. The RSDs for
the remaining seven  replicate  sets ranged  from 2 to
17 percent with a median of 5 percent.  The RSDs for the
two triplicate sets of liquid samples were 1 and 2 percent
with a median of 1.5 percent.

For the reference method, the RSD calculated for the blank
soil samples was not considered in evaluating the method's
precision because one of the three blank soil sample results
(5.12 mg/kg) was  estimated by adding one-half  the
reporting limits for the GRO, DRO, and ORO components
of TPH measurement.  The RSDs for the remaining seven
replicate sets ranged from 2 to 10 percent with a median of
7 percent. The RSDs for the two triplicate sets of liquid
samples were 5 and 6 percent with a median of 5.5 percent.
Comparison of the OCMA-350 and reference  method
median RSDs showed that the OCMA-350 exhibited better
precision than the reference method for both soil and liquid
PE samples. Finally, for the reference method, the median
RSD for the soil PE samples (7 percent) was less than that
forthe environmental samples (18 percent), indicating that
greater precision was achieved for the samples prepared
under  more  controlled conditions  (the  PE samples).
Similarly, forthe OCMA-350, the median RSD forthe soil
PE samples  (5  percent) was  less than that  for  the
environmental samples (20 percent).

7.1.3  Primary Objective P3: Effect of Interferents

The effect of interferents on TPH measurement using the
OCMA-350 and reference method was evaluated through
analysis of high-concentration-range soil PE samples that
contained weathered gasoline or diesel with or without an
interferent. The  six interferents used were MTBE; PCE;
Stoddard solvent; turpentine; 1,2,4-trichlorobenzene; and
humic acid.  In  addition, neat (liquid)  samples  of each
interferent except humic acid were used as quasi-control
samples to evaluate the effect of each interferent on the
TPH  results  obtained  using the  OCMA-350  and  the
reference  method.   Liquid interferent samples  were
submitted for analysis as blind triplicate samples. Horiba
and the reference laboratory were provided with flame-
sealed ampules of each interferent and were given specific
instructions to prepare dilutions of the liquid interferents
for analysis.   Two dilutions of each  interferent were
prepared;  therefore, there  were  six  OCMA-350  and
reference method TPH results for each interferent.  Blank
soil was mixed with humic  acid at two levels to prepare
quasi-control samples  for this  interferent.   Additional
details  regarding  the  interferents  are provided  in
Chapter 4.  The results for the quasi-control interferent
samples are discussed first below, followed by the effects
of the interferents on the TPH results for soil samples.

7.1.3.1     Interferent Sample Results

Table 7-9 presents the OCMA-350 and reference method
TPH results, mean TPH results, and mean responses for
triplicate sets of liquid PE samples and soil PE samples
containing  humic  acid.    Each  mean   response  was
calculated by dividing the mean TPH result for a triplicate
set by the interferent concentration and  multiplying by
100. For liquid PE samples, the interferent concentration
was estimated using its density and purity.

The mean responses forthe OCMA-350 ranged from 0 to
94 percent for the  liquid interferent samples; the mean
response for humic acid was 0 percent. The TPH results
for a given triplicate set and between the triplicate sets
showed good agreement except for Stoddard solvent. The
mean responses for MTBE (72.5 percent) and Stoddard
solvent (86 percent) indicated that these compounds can be
measured as TPH using the OCMA-350.   The mean
response for  turpentine  (85  percent)  indicated  that
turpentine would result in  false positives during  TPH
measurement.  The mean response of 0 percent for PCE;
1,2,4-trichlorobenzene; and humic acid indicated that these
compounds would  not result in either false positives or
false negatives during TPH measurement.

The mean responses forthe reference method ranged from
17 to 92 percent for the liquid interferent samples; the
mean response for humic acid was  0 percent.  The TPH
results for a given triplicate set and between the triplicate
sets showed good agreement.  The mean responses for
MTBE (39 percent)  and  Stoddard  solvent (85 percent)
indicated that these compounds can be measured as TPH
using the reference method.  The mean responses for PCE
(17.5   percent);   turpentine   (52  percent);   and
1,2,4-trichlorobenzene (50 percent) indicated that these
interferents will likely result in false positives during TPH
measurement.  The mean response of 0 percent for humic
acid indicated that humic acid would not result in either
false positives or false negatives during TPH measurement.

7.1.3.2     Effects of Interferents on TPH Results for
           Soil Samples

The effects of interferents on TPH measurement for soil
samples containing weathered gasoline  or diesel  were
                                                    77

-------
Table 7-9. Comparison of OCMA-350 and Reference Method Results for Interferent Samples
Interferent and Concentration3
OCMA-350
TPH Result
Mean TPH
Result
Mean Response11
(percent)
Reference Method
TPH Result
Mean TPH
Result
Mean Response11
(percent)
Liquid Interferent Samples (TPH Results in Milligram per Liter)
Methyl-tert-butyl ether
(740,000 milligrams per liter)
Tetrachloroethene
(1 ,621 ,000 milligrams per liter)
Stoddard solvent
(771 ,500 milligrams per liter)
Turpentine
(845,600 milligrams per liter)
1 ,2,4-Trichlorobenzene
(1 ,439,000 milligrams per liter)
544,000
526,000
520,000
547,000
510,000
562,000
Less than 6,080
Less than 6,080
Less than 6,080
Less than 1 ,220
Less than 1 ,220
Less than 1 ,220
511,000
824,000
847,000
463,000
762,000
576,000
722,000
711,000
720,000
685,000
746,000
721,000
Less than 6,080
Less than 6,080
Less than 6,080
1,690
1,600
1,780
530,000
540,000
3,040
610
730,000
600,000
718,000
712,000
3,040
1,690
72
73
0
0
94
78
85
85
0
0
309,000
272,000
270,000
303,000
313,000
282,000
269,000
270,000
277,000
290,000
288,000
307,000
561 ,000
628,000
606,000
703,000
Not reported
713,000
504,000
459,000
442,000
523,000
353,000
349,000
711,000
620,000
732,000
754,000
756,000
752,000
284,000
299,000
272,000
295,000
598,000
708,000
468,000
408,000
688,000
754,000
38
40
17
18
78
92
55
48
48
52
Interferent Samples (Processed Garden Soil) (TPH Results in Milligram per Kilogram)
Humic acid at 3,940 milligrams per
kilogram
Humic acid at 1 9,500 milligrams
per kilogram
Less than 15.2
Less than 15.2
Less than 15.2
Less than 15.2
Less than 15.2
Less than 15.2
7.6
7.6
0
0
8.99
8.96
8.12
69.3
79.1
78.5
9.00
76.0
0
0
Notes:
a    A given liquid interferent concentration was estimated using its density and purity.
b    The mean response was calculated by dividing the mean TPH result for a triplicate set by the interferent concentration and multiplying by 100.
                                                                78

-------
examined through analysis  of PE  samples containing
(1)  weathered  gasoline  or  diesel  (control)  and
(2) weathered gasoline or diesel plus a given interferent at
two levels. Information on the selection of interferents is
provided in Chapter 4.

Triplicate sets of control samples and samples containing
interferents were prepared for analysis using the OCMA-
350 and reference method. A parametric or nonparametric
test was  selected for statistical evaluation of the results
using the approach presented in Figure  7-1.

TPH results for samples with  and without interferents,
statistical tests performed, and statistical test conclusions
for both  the  OCMA-350  and  reference  method  are
presented in  Table 7-10.  The null hypothesis for the
statistical tests was that mean TPH results for samples with
and without interferents were equal. The statistical results
for each interferent are discussed below.

Effect of Methyl-Tert-Butyl Ether

The effect of MTBE was evaluated  for soil PE samples
containing weathered gasoline.  Based on the liquid PE
sample  (neat material)  analytical results,  MTBE was
expected to bias both the  OCMA-350 and reference
method results high.

For the OCMA-350, at the interferent levels used, MTBE
was expected to bias the TPH results high by 127 percent
(low level) and 196 percent  (high level).  The expected
bias would be lower (37 and 58 percent, respectively) if
MTBE in soil samples was assumed to be  extracted as
efficiently  as  weathered  gasoline  in soil  samples.
Table 7-10 shows that MTBE biased the OCMA-3 5 0 TPH
results high at both low and high interferent levels, which
confirmed the conclusions drawn from the results of the
neat MTBE  analysis.    Although the statistical tests
confirmed the  observed bias, they indicated that  the
observed bias was  statistically the same at both low and
high MTBE levels.

For the reference method, at the interferent levels used,
MTBE was expected to  bias the TPH results high by
21 percent (low level) and 33 percent (high level).  The
expected  bias would be lower  (17  and  27 percent,
respectively) if MTBE in soil samples was assumed to be
extracted  as  efficiently  as  weathered gasoline in soil
samples. However, no effect on TPH measurement was
observed  for  soil  PE  samples  analyzed during  the
demonstration.  A significant amount of MTBE, a highly
volatile compound, may have been lost during PE sample
preparation, transport, storage, and handling, thus lowering
the MTBE concentrations to levels that would not have
increased the TPH results beyond the reference method's
precision (7 percent).

Effect of Tetrachloroethene

The effect of PCE was evaluated for soil  PE samples
containing weathered gasoline.  Based on the liquid PE
sample (neat  material)  analytical  results, PCE  was
expected to have no effect on the TPH results for the
OCMA-350; however, it was expected to bias the reference
method results high.

Forthe OCMA-350, PCE appeared to have biased the TPH
results high. However, the bias decreased with an increase
in the PCE level. Specifically, the bias was 144 percent at
the low level and 90 percent at the high level. For this
reason, no conclusion was drawn regarding the effect of
PCE on TPH measurement using the OCMA-350.

For the reference method, at the  interferent levels used,
PCE  was expected  to bias the TPH results high by
24 percent (low level) and 113 percent (high level).  The
expected  bias would  be lower  (20  and  92 percent,
respectively) if PCE in soil samples was assumed to be
extracted  as efficiently as weathered gasoline in  soil
samples. The statistical tests showed that the probability
of the three means being equal was less than 5 percent.
However, the tests also showed that at the high level, PCE
biased the TPH results high,  which appeared  to be
reasonable based on the  conclusions drawn from  the
analytical results for neat PCE. As to the reason for PCE
at the low level  having  no  effect on the TPH results,
volatilization during PE  sample  preparation, transport,
storage,  and  handling  may have  lowered  the PCE
concentrations to levels that would not have increased the
TPH  results beyond the reference method's precision
(7 percent).

Effect of Stoddard Solvent

The  effect  of  Stoddard solvent was  evaluated  for
weathered gasoline and diesel soil PE samples. Based on
the liquid PE  sample (neat material)  analytical results,
Stoddard  solvent  was  expected to bias  both   the
OCMA-350 and reference method results high.

For the  OCMA-350,  at the  interferent  levels used,
Stoddard  solvent was expected to bias the  TPH  results
                                                    79

-------
V)
•£
c
o
o
Q.

re
c
o
0)
u
0)
Q.
'o
CO
£
U)
U)
0)
•o
o
0)
u
•D
C
re
O
O
i^
o
c
o


a


o










T3
t5
2
CD
O
c
CD
'CD















o
in
O
0


























CD
'S-L- £ ~ -5 j/> E ^
^t^-2 51 w ^
•p; — (D ?» 0) ^ O
-S 1 1 ff "Srog.
CL $ ra — 'CD
CO
C .2 CO j/, £
CD "7? "o. C "3 C CD "r^
^xws§'St/5!t
**
CO
to
K
8
:«
"ro
CO

X ^

(/) "•-
g« !>
CD D- £,
** "3

CL CO "gj
Ho: £
D)
-^C^^CD^CfnO
-S S | 'c"? cS <" ill
£ co ra JD **

«
CD
1-
1
"ro
CO
X
n -i_i -^
h- "5 S
(/) ~~-
IKE
If
Q^ ~g.
x E
CL ^^
i 	
1






T3
C
CO
X
"S s^
^ c
CD 2
Q..2
E oi
co -^
w £








































U)
•£
£
•g
cu
'c


D
O
^

^
U)
cu
Q.
E

-
'o




















°
o
CN"
0
00
00
^








0
CO
CD

£
CD









(D
C

O
(/)
ro
O)
-D
OJ
OJ
"ro
"^
>







CD
_Q
8
Q.
Q.
CO
'o
Z










0 0
CN 00
O T-
CN" CN"




Not applicable







£ S
in CD


























CD
_Q
8
Q.
Q.
CO
'o
Z







CN
[ —
CN"
000
00 O) O
f 00 00
CN" CN" CN"




Not applicable



CN
00
O)

ass
O) O) O)
















1
CD
Q










































U)
*-
C
CU
rtl
•c
cu

£
•£

^
U)
cu
Q.
E

CO
'o



,_
CN
^




CD
E
CO
CO
o
CO
'co
~m
CO
i ^
CD .-
6 s

En
O)

0
o
O)
^
O)
CN
O

Mean without
interferent

ra".  (1
> C












T3
CO t/
^— V C
.0 c
"CD -^
Et
co c
co -^
SP




0
in

^




was different
t
c
n
variance





CN
00
00





iS

0










0











i

I !
'c't
>S|
r-= £
: <= J
D) ^
- CO 1




0

CN
CN




) 1
i t
! !
ili
:
J to
^ c
: ° t
) fc, i
i SH
m





0
CD
T-~





























!! '
s •
i°
?.L
fa'
u F
- tr
= O
3 0

°.
CO
CN"
0
in

CN




i 1
S 3~
= 5

difference)

Q

ID

g
in
T-~




"£
\f

~c
£
LLJ C

i 	 1
^ T—

















j"
[)
6
|
fl
TJ
i)





0 0
CN CD
co in
CN CN




/)
U
>
V
'o
comparison
means
(parametric)





° S
CO CD
T-~ T-"




)

i)













0
o
0


CO
r- •*->
lean witl
iterferen
s .E








g
CO
CN"
o

in
CN
CO

CN

Mean without
interferent
CO
5 >•
CO fO
I§

Q

m_

8
in
T-~



f 	 ^
D)

D)
E
o
LLJ J5

CL Cl













•E
CD
gig
^^
D) ra
'n 3











0
CD

CN




was same as
mean with

analysis of
variance





o
in
T-~

























CO
c »
CO C
E?l
ii]
fc 3 .E











0

^~
CN




00
ill
a -a ^
s^>S
c r f
t .!
(nonparame
and Kruskal
\A/^llif n-^irlAr





0
in
T-"

























n

I "c
> £ CD
^3 5
iE-2
= co .E








0
^-
"
Q

h--
"*




interferent at
low level was
! 'o
_
„ CO CO
2 Q. C
5 E n

0
o
CN

8
CN
T-"



"3



O
o
LLJ "~-
O <">
CL C-















w level
_52











0

in
"*




same as
mean with
^o
(nonparame





8
CO
T-~








































0

o
"*




interferent at
high level







S
O)



















                                                                                   80

-------

1
o
o.
U)
•£
c
o
o
Q.


re
C
o
i)
U
a)
Q.
'o
00
£
i/i
•D
O
II)
U
•D
C
re
o
in


S
O
O
i^
O
c
o



1


o





T3
.c
^
o

01
,2
(U
LL















0
U)
CO
<£
o
0






















u- -C £
0-,- £ ~ -5 j/> | ^
^"t 'w -^ S w c
!Q "'""' CD °5» a) ^ o
l^Jli}1
c J2 oo j» o ^
lj|j|f |1
in
0)
1-

S
to

to
w

X

KH"
£ 0> D>
1" rv P
0) "- -S-
_ = "3
x S =¥
P- S D>
*~ CC E
D)
-Q<-— CfnO
-S S | p"5 a <" ill
£ w ra .2 **


•ffi
01
1-
1
to
co
X
P: = "3
,- 00 ^
«K E
o) °- -S
!§
Q^ ~g.
x E
CL *-'







T3
C
X
to «
0) 2
"Q.,2
F ^

co £





































•D
D
C
O


0

"
0
in
CO
"*
o
o
o

All three
means (with




One-way
analysis o

0
O)

R
O)
T-








Stoddard
solvent
T3

> D)





and without
interferents)

T3
c -
ra to
^~* 0)
O c
i: o
0) £,
E ^,
ra oT
ra-^
•&H




0
to

"*




and without
interferents)
T3
c
ra
"o"
variance
(parametri




R
o
C\l





f 	 ^
D)

D)
O
8
:£•
•* i^ (
OJ ~ ^
S i>!t
— 00 7




o
r^
T-




















5 oj
i "o
) ^
5 (D
= c
3 O
5"

(18,200m

























0
00
C\l
00




interferent at
high level









R
O)
to
*•
















0
o
0
! =£• i
3 ° c i
(D "-'.C (D 0
lillsi
< E ra .E 3 'o










f~)
CM
*
000

^t~ Uj ^t~
^" CO ^"
o
o
o

£-3S~ i
30^ t
(D "-'.C (D 0
lilid
< E ra .== 3 'o
-Q
c -
ra to
14- ^ C
•^ ° S ° t
m oo •«
if Illi

0
OM
00

O O O
r^ in co
00 (D O)
T- T- T-





f 	 ^
D)

1?
ig
Kd;
T3
 D)





01a 	 ^aniiy
different from










0
o
O)
C\l
0
o
00
C\l




01a 	 vanuy
different from




ola 	 ^allL
difference

0
o

C\l
R
in
C\l
v




c
^
„, c
Turpentine
C\9 900 m











3
3
0
V
5













0
o
C\l
^




3
3
0
V
5


o
c
osuedmoo
aQiAAiiprl




R
C\l
OM
v




0

0
'

























0
o
to
2.








o
means
(parametri




R
CO
CO
v














                                                                                 81

-------

1
o
o.

U)
•s
c
o
o
Q.


re
C
o
i)
u
'o
CO
£
ui
•D

O
•D
C
re

o
in



S
O
O
i^
O

c
o



a



o





T3
.c
"CD
rence IV
£
a:












o
in
<£
o
0










U- -r: £
0-,- £ ~ -g jfl E ^
^"t 'w -^ S w c
!Q -»-• CO ^ ^- gj OJ
l^Jli}1
c J2 co .co o ex.
Ijjjjlf 11
CO
8
to
"lO
w
X
^1
J2E
x^J?
Pi 0, D>
<* .§
D)
="£: 'w ^ £ co E c
-Q<-— 0);>CfnO
-S S I f-? Is jfi S.
£ CO ffl a **

1— »- *-
«
0)
1—
8
"(0
w
X
H 3 ™

c 0) 0>
aj LJ- -S-
If
^ "ra
x E
CL *-'




T3
C
CO
X
to «
S jjj
CD S:
Q.,2
E ID
CO ~
w £































•D
D
C
O
I/I
•s
Interferi
1
I/I
Q.
E
re
CO
'o
CO


in
CD
CN
Mean without
interferent
CD
° u-
co o
= CO
^ 'CO
_L CO
Kruska
way an

8
in
0
CD
CO
in

o
o
o

si
0,1
0, CO
£ g
— (D
< b

1(
m" w
? g.
c ra
0 g

0
O)
CD
CN

0





"O)
it
Is
1— G-



OJ
Sm>raSmOJ
COC^OJC^CDOJC^ —
raS^sS^sESfe-g,
3 E .E ° E .E ° co E .E S
T3
C
to
"o~
•^
"Z
F
0) CO
si
si



0
CO
in





nd without
terferents)
to .£
T3
C
to
"o"
oj t5
= Is
to 2
b to
> S





0
r--.














o
CO C ^
> o E
* o) •" 2
^ .CO to CO CO



0
CD
in

0
o
O)
CO
000
o o o
O CO CO
in co co





^ O OJ
'c ^ £
to -K o
.Efig
fc&^^
3 CO T3 O


,_
o
0




















"to "c co "to
.- „ (DC „ „
ic^into c c_
'3£>£,o 53 ra «
3 S^1! fc era
= CO "5 D) t S 0)
SSK-s^ 8 E



0
in
CO








0
in
in
CO
0
CD
O
O)
^





0
CD
in
CD





0
O)
CD
CD





to ci co to
£"c7n!nto "c "c_
'>aj>i*i-aj aj-cajaj
c|-^;iEl|l|S
SoJ-c,coE^=a3-Do3>
OJ^ii D)to o^^-iii t- -S >
2 .£
W o
c c -Q"
l|8«-l «
>••= g'5 S c E
S c .2 .1 E" g |
— 'CO T3 Q- O E -S=





0









0
o
CO
CN

888





D)
gf
'c °

1— C-
= -C




3 t 3 .c




0
o
CO
O)




"richloro-
0
CN




"55
B ?
8






i (D O
^r N m
CN c °°-
T- J2 C2-













0
CO
T-

0
O)




rrichloro-
m '









0
CO




D)
1)
I O









0
CO





i (D 0
•* N CD
,— o i 	


















                                                                            82

-------

1
o
o.
U)
•s
c
o
o
Q.


re
c
o
U
c
re
Q.


'5




I

U)


"5
U)


£

•D
O
•D
C
re
o
in



S
O
O
i^
O
c
o



1



o





T3
.C
'S
2
 "c r n O
5<°w'5-2(1,a3

(/)
-1— '
 1 £ g ra
S — C ^ ^) (D
jj ~5\ S) QJ > E
ES E I ° co
1 I
? OJ
ra "5 E
Q. ro

= § s.™
z oj E ro
co ^ o E
8
CN
CN"
0
in

CN"
o
o
o
All three
means (with




S7.<»
? g.
s g
0 ra

0)
CN
in


CN

CD




"3
It
o o
!*
X d
^
0)
3
c
OJ ^3
OJ O
09.
0
00
o
CN"



and without
interferents)
T3
c
ra
'S
^ E
II







in













0
CD
00
CN"




m
-c -^ —
mean wit
interferer
high leve






0
CN"
0 0
CD CN
CD t
CN" CN"



were
significantly
different from
one another

_
w *o
— c
g^ ? 8
S ra £ .£    ^

if-3£
oj -9 oj
Q. -t; E
   Ed) O    QJ
   -i-i —    i-
^£ 2

ii?

M  M M

D)LU
=* m LU
ral- o
E 2 CL
                                                                                     CL
                                                                83

-------
high by 396 percent (low level) and 2,100 percent (high
level) for weathered gasoline soil PE samples and  by
320 percent (low level) and 1,600 percent (high level) for
diesel soil PE samples.  The expected bias would be lower
(116 and 618 percent, respectively, for weathered gasoline
soil PE samples and 92 and 458 percent, respectively, for
diesel soil PE samples) if Stoddard solvent in soil samples
was  assumed to be extracted as efficiently as weathered
gasoline and  diesel  in soil samples. The statistical tests
showed that the mean TPH results with  and without the
interferent were different for weathered gasoline soil PE
samples, which confirmed the conclusions drawn from the
analytical results for neat Stoddard solvent. However, for
diesel soil PE samples, (1) the mean TPH result without
the  interferent and the mean  TPH  result  with  the
interferent at the low level  were equal and (2) the mean
TPH results with the interferent at the low and high levels
were equal, indicating  that Stoddard solvent at the low
level did not affect the  TPH results for the diesel soil PE
samples but that Stoddard  solvent at the high level did
affect the TPH results.  The conclusion reached for the
interferent at the low level was unexpected and did not
seem reasonable based  on a simple comparison of means
that differed by a factor of three. The anomaly might have
been associated  with  the  nonparametric test used  to
evaluate the effect of Stoddard solvent on TPH results for
diesel soil PE samples, as nonparametric tests do not
account for the magnitude of the difference between TPH
results.

For the reference method, at the interferent levels used,
Stoddard solvent was expected to bias the TPH results high
by 121 percent (low level) and 645 percent (high level) for
weathered gasoline  soil PE samples and by  114 percent
(low level) and 569 percent (high level) for diesel soil PE
samples.  The expected bias would be lower (99 and
524 percent, respectively, for weathered gasoline soil PE
samples and 61 and 289 percent, respectively, for diesel
soil PE samples) if Stoddard solvent in soil samples was
assumed to be  extracted  as  efficiently as  weathered
gasoline and  diesel  in soil samples. The statistical tests
showed that the mean TPH results with  and without the
interferent were different for both weathered gasoline and
diesel soil PE samples, which confirmed the conclusions
drawn from  the  analytical results for  neat Stoddard
solvent.

Effect of Turpentine

The  effect of turpentine was evaluated for weathered
gasoline and diesel soil PE samples. Based on the  liquid
PE sample (neat material)  analytical results, turpentine
was expected to bias both the OCMA-350 and reference
method results high.

For  the  OCMA-350,  at the interferent levels  used,
turpentine was expected to  bias the TPH results high by
368 percent (low level) and 1,740 percent (high level) for
weathered gasoline soil PE samples and by 333 percent
(low level) and 1,700 percent (high level) for diesel soil PE
samples.   The expected bias would be lower (108  and
512 percent, respectively, for weathered gasoline soil PE
samples and 96 and 488 percent, respectively, for diesel
soil PE samples) if turpentine in soil samples was assumed
to be extracted as  efficiently as weathered gasoline  and
diesel in soil samples. The statistical tests showed that the
mean TPH results  with and without the interferent were
different for both weathered gasoline and  diesel soil PE
samples, which confirmed the conclusions drawn from the
analytical results for neat turpentine.

For the reference method, at the interferent levels used,
turpentine was expected to  bias the TPH results high by
69 percent (low level)  and 327 percent (high level) for
weathered gasoline soil PE samples and by 72 percent (low
level) and  371 percent (high  level) for  diesel soil PE
samples.   The  expected bias would be lower  (56  and
266 percent, respectively, for weathered gasoline soil PE
samples and 39 and 200 percent, respectively, for diesel
soil PE samples) if turpentine in soil samples was assumed
to be extracted as  efficiently as weathered gasoline  and
diesel in soil samples. The statistical tests showed that the
mean TPH results  with and without the interferent were
different for weathered gasoline soil PE samples, which
confirmed  the  conclusions  drawn from  the analytical
results for neat turpentine.  However, for  diesel soil PE
samples,  (1) the mean TPH result without the interferent
and the mean TPH result with the interferent at the  low
level were  equal and (2) the mean TPH results with the
interferent  at  the low and high  levels were equal,
indicating that turpentine at the low level did not affect the
TPH  results  for  the diesel soil  PE samples but  that
turpentine at the high level did affect the TPH results.  The
conclusion reached for the interferent at the low level  was
unexpected and did not seem reasonable based on a simple
comparison of means that differed by a factor of three.
The anomaly might have been associated with  the
nonparametric test used to evaluate the effect of turpentine
on  TPH  results  for  diesel  soil  PE  samples,   as
nonparametric tests do  not account for the magnitude of
the difference between TPH results.
                                                     84

-------
Effect of 1,2,4-Trichlorobenzene

The  effect of 1,2,4-trichlorobenzene was evaluated for
diesel soil PE samples.  Based on the liquid PE sample
(neat material) analytical results, 1,2,4-trichlorobenzene
was expected to have no effect on the TPH results for the
OCMA-350; however, itwas expected to bias the reference
method results high.

For the OCMA-350,  1,2,4-trichlorobenzene appears to
have biased the TPH results high; the bias was statistically
significant only at the high interferent level. The observed
bias contradicted the conclusions drawn from the results of
the neat 1,2,4-trichlorobenzene analysis, which cannot be
explained.

For the reference method, at the  interferent levels used,
1,2,4-trichlorobenzene  was expected  to  bias the TPH
results high by 62 percent (low level) and 305 percent
(high level).  The expected bias would be lower (33 and
164 percent, respectively) if 1,2,4-trichlorobenzene in soil
samples was assumed  to  be extracted as efficiently as
diesel in soil samples. The statistical tests showed that the
probability of three  means being  equal  was less than
5 percent. However, the tests also showed that when the
interferent was present at the high level, TPH results were
biased high.   The effect observed at the  high level
confirmed  the conclusions drawn from  the  analytical
results for neat 1,2,4-trichlorobenzene. The statistical tests
indicated that the mean TPH result with the interferent at
the low level was not different from the mean TPH result
without the interferent, indicating that the low level of
1,2,4-trichlorobenzene did not affect TPH measurement.
However, a simple comparison of the mean TPH results
revealed  that  the  low level  of 1,2,4-trichlorobenzene
increased the TPH result to nearly the result based on the
expected bias of 33 percent. Specifically, the mean TPH
result with the interferent at the low level was 3,510 mg/kg
rather than the expected  value of  3,620 mg/kg.  The
conclusions drawn from the statistical tests were justified
when the variabilities  associated with the mean TPH
results were taken into account.

Effect of Humic Acid

The effect of humic acid was evaluated for diesel soil PE
samples.   Based on the analytical results  for  soil  PE
samples  containing  humic acid, this interferent  was
expected to have no effect on the TPH  results for the
OCMA-350 and reference method.
For the OCMA-350, humic acid biased the TPH results
low; the bias was statistically significant at both the low
and high interferent levels. This observation appeared to
contradict the conclusions drawn from  the  analytical
results for soil PE samples containing humic acid (quasi-
control samples); however, the apparent contradiction was
attributable  to the  fact that the quasi-control  sample
analyses could predict only a positive bias (a negative bias
is equivalent to a negative concentration).

For the reference method, humic acid appeared to have
biased the TPH results low. However, the bias decreased
with an increase in the humic acid level. Specifically, the
negative bias  was  19  percent  at  the  low  level and
10 percent at the high level. For this reason, no conclusion
was  drawn regarding the effect  of humic acid on TPH
measurement using the reference  method.

7.1.4   Primary Objective P4: Effect of Soil
        Moisture Content

To measure the  effect of soil moisture content on  the
ability  of the  OCMA-350  and reference method  to
accurately measure TPH, high-concentration-range soil PE
samples containing  weathered gasoline or diesel at two
moisture levels were analyzed.   The OCMA-350 and
reference method results were converted from a wet weight
basis to a dry weight basis in order to evaluate the effect of
moisture content on the sample TPH results. The OCMA-
350  and reference method dry  weight TPH results for
weathered gasoline samples and the reference method dry
weight TPH results for diesel  samples were  normally
distributed; therefore, a two-tailed, two-sample Student's
t-test was performed to determine whether the device and
reference method results were impacted by soil moisture
content—that is, to determine whether an increase in soil
moisture content resulted in an increase or decrease in the
TPH concentrations measured. Because the OCMA-350
dry  weight  TPH results for diesel  samples  were not
normally distributed, aKruskal-Wallis one-way analysis of
variance (a nonparametric test) was performed. The null
hypothesis for the t-test and the analysis of variance was
that  the two means were equal or that  the  difference
between the means was equal to zero. Table 7-11 shows
the sample moisture levels, TPH results, mean TPH results
for sets of triplicate samples, whether the mean TPH
results at different moisture levels were the same, and the
probability of the null hypothesis being true.
                                                    85

-------
i
to
'o

re

V)
_0)

Q.


re
00


o


1


I
UJ
a.


'o
"5
U)
O
c
o



1


o












T3
O
r—

t5

§
OJ
OJ
o:

















s
CO
2
O
O























CO
"o '<" "
(D Q) *— v
= 'o H a>
ro >i D) ^~
o ^ c qj
O '(D o:
ct|-
I 111
t o3 w S
c ~ > 5

(D •" -I Q
2 "£ £
OJ ±± 3
1_ -5 4± Q)
^ w <" E
> <" ° ra
ce 2 w

x <°
Q- = ^""E"
1 — ^ ^ ro
(/) (O l—
S ce il'.i'
5 1.^
o ra oi
4_i pn Q. ' — *
"5 *- ^ E
(D ^) 2 I)
x |lfl
K Q ^
'o '<» "
0)0)^^
= 'o H -
in .c co
(D CO.™ O
o: -5; = g
x > ^,
D_
13
— (U
-^ 3
to to
^ 'o
§ E



















o
s
C\l


000
r^ o) ^~
O CO t
OM OM OM













0
in
O)




o o o
f CD in
r^ o o
T-~ rsT CN

oj
0
CL
to
to


'o
CO
'in
c
ra
ra"


-------
Table 7-11 shows that OCMA-350 results for weathered
gasoline and diesel soil samples at different moisture levels
were  statistically different at a significance  level of
5 percent; therefore, the OCMA-350 results were impacted
by soil moisture content. Based on a simple comparison of
the results, this  conclusion appeared to be reasonable.
Specifically, on average at the higher moisture level, the
mean TPH result (1) for weathered gasoline soil samples
was higher by a factor of three and (2) for diesel soil
samples was  lower by a  factor of three.  The  opposite
effect of  soil moisture  content on TPH results for
weathered gasoline and diesel soil samples could not be
explained.

Table 7-11 also shows that reference method results for
weathered gasoline soil samples and diesel soil samples at
different moisture levels were statistically the same at a
significance level of 5 percent; therefore, the reference
method results were not impacted by soil moisture content.
Based on  a  simple  comparison of  the results,  this
conclusion appeared to be reasonable.

7.1.5  Primary Objective P5: Time Required for
       TPH Measurement

During the demonstration, the time required for TPH
measurement activities,  including   OCMA-350  setup,
sample   extraction,   sample  analysis,  OCMA-350
disassembly, and data package preparation, was measured.
During the demonstration, one field technician performed
the TPH measurement activities using the OCMA-350.
Time  measurement  began  at  the   start  of  each
demonstration day when the technician began to set up the
OCMA-350 and ended when he disassembled the OCMA-
350. Time not measured included (1) the time spent by the
technician  verifying  that he  had  received  all  the
demonstration samples  indicated on  chain-of-custody
forms and (2) the times when he took breaks. In addition
to the total time required for TPH measurement, the time
required to extract and analyze the first and last analytical
batches of soil samples was measured.  The number and
type of samples in a batch were selected by Horiba.

The time required to complete TPH measurement activities
using the OCMA-350 is shown in Table 7-12.  The time
required for  each activity was rounded to the nearest
5 minutes.

Overall, Horiba required 46 hours, 15 minutes, for TPH
measurement of 74 soil environmental samples, 89 soil PE
samples, 36 liquid PE samples, and 9 extract duplicates.
Information  regarding  the  time  required for  each
measurement  activity  during   the   entire   6-day
demonstration and for extraction and analysis of the first
and last batches of soil samples is provided below.

The setup time for the OCMA-350 was measured at the
beginning of each day during  the  6-day demonstration
period.   Setup activities  included OCMA-350  setup;
analysis  of QC  check standards;  and  organization of
extraction,  dilution,  analysis,  and  decontamination
supplies.  OCMA-350 setup required a total of 1 hour,
45 minutes, during the 6-day demonstration, or an average
of 17.5 minutes per day. The setup times on the days when
the first and last soil sample batches were analyzed were
25  and  10 minutes, respectively, or an  average  of
17.5 minutes per day.  The total setup time for the entire
demonstration and the setup  time on the days when the
first and last soil sample batches were analyzed suggested
that the field technician gradually became more familiar
with the OCMA-350 setup procedure as the demonstration
progressed.

Extraction of all  163 soil samples required 25  hours,
45 minutes, resulting in an  average extraction time of
9 minutes per sample.  However, the field technician also
completed  other activities   associated  with   sample
extraction  during   the   demonstration,  such  as
decontaminating metal spatulas and glass beakers. Thus,
the average sample extraction time of 9 minutes included
the time required to complete these activities.

The time required for extraction  of the first  and last
batches of soil samples was  also recorded.  In general,
Horiba designated 24 samples for each analytical batch.
The number of samples was based on the capacity of the
test tube rack used during extraction. However, fewerthan
24 samples comprised an analytical batch in some cases;
for example, the last batch of soil  samples consisted of
only 17 samples.  The first batch of soil samples (which
consisted of 24 samples) required 2 hours, 50 minutes, for
extraction, or an average of 7 minutes per sample. The last
batch of soil samples (which consisted of 17 samples)
required  1 hour, 30 minutes, for extraction, or an average
of 5 minutes per sample.  The decrease in the average
extraction time between the first and last batches of soil
samples suggested that the field technician became more
familiar with the OCMA-350 extraction procedure as the
demonstration progressed. The difference between the
9-minute average extraction time for all the soil samples
and the  average  extraction time for the first  and last
batches of soil samples (6 minutes) may be attributable to
                                                    87

-------
Table 7-12. Time Required to Complete TPH Measurement Activities Using the OCMA-350
                                                                 Time Required3
Measurement Activity
First Soil Sample Batchb
Last Soil Sample Batchb
6-Day Demonstration Period
OCMA-350 setup0
Sample extraction
Sample analysis
OCMA-350 disassembly0
Data package preparation
25 minutes
2 hours, 50 minutes
2 hours
20 minutes
Not measured
10 minutes
1 hour, 30 minutes
1 hour, 10 minutes
10 minutes
Not measured
1 hour, 45 minutes
25 hours, 45 minutes
16 hours, 45 minutes
1 hour, 30 minutes
30 minutes
Total
                                      5 hours, 35 minutes
                                                               3 hours
                                                  46 hours, 15 minutes
Notes:
    The time required for each activity was rounded to the nearest 5 minutes.
    The first sample batch required 29 TPH analyses (24 sample extract analyses, 3 extract duplicate analyses, and 2 reanalyses). The last sample
    batch required 22 TPH analyses (17 sample extract analyses and 5 reanalyses).
    The OCMA-350 setup and disassembly times were measured at the beginning and end of each day, respectively. The times reported for the first
    soil sample batch were the times recorded on the first day and the times reported for the last soil sample batch were the times recorded on the
    fifth day of the 6-day demonstration period.
the additional extraction activities mentioned above that
were conducted by the field technician.

A total of 16 hours, 45 minutes, was required to obtain and
report 208 TPH results using the OCMA-350, resulting in
an average analysis time of 5  minutes per sample.  In
addition to performing the 208 analyses required to report
208 TPH results, the field technician performed extract
dilutions and reanalyses for the high-concentration-range
samples. The average analysis  time does not include the
time taken to perform these additional activities.

The time required to analyze the first and last batches of
soil samples  was also recorded. A total of 2 hours was
required to  analyze the first batch of samples, which
required 29 TPH analyses (24 sample extract analyses,
3 extract duplicate analyses, and 2 reanalyses); therefore,
an average of 4 minutes was required to complete one
analysis. A total of 1 hour, 10 minutes, was required to
analyze the last batch of soil  samples, which required
22  TPH  analyses  (17  sample   extract  analyses and
5 reanalyses); therefore, an  average of 3  minutes was
required to complete one analysis. The decrease in the
average analysis time between the first and last batches of
soil samples  suggested that the field technician became
more familiar with the OCMA-350 analysis procedure as
the demonstration progressed. The difference between the
5-minute  average sample  analysis time for the  entire
demonstration and the average sample analysis time for the
first and last batches (3.5 minutes) may be attributable to
the additional analysis  activities mentioned above that
were conducted by the field technician.
                    The disassembly time was measured at the end of each day
                    of the 6-day demonstration. Disassembly included packing
                    up the OCMA-350 and associated  supplies required for
                    TPH measurement. OCMA-350 disassembly required 10
                    to 20 minutes each day, totaling 1 hour, 30 minutes, for the
                    6-day demonstration. Thus, the average disassembly time
                    was 15 minutes per day.  The disassembly times on the
                    days  when the first and last soil sample  batches were
                    analyzed were 20  and 10 minutes, respectively,  or an
                    average of 15 minutes per day. The total disassembly time
                    for the entire  demonstration and the setup times on the
                    days  when the first and last soil sample  batches were
                    analyzed suggested that the  field  technician gradually
                        J       OO                              O       J
                    became more  familiar  with the OCMA-350 disassembly
                    procedure as the demonstration progressed.

                    Preparation of the OCMA-350  data package required
                    30 minutes in the field. Preparation of the data package
                    submitted  to the EPA at the end of the demonstration
                    involved   compiling  printouts  of TPH  results   and
                    photocopying Horiba's log sheets, which included sample
                    identification  numbers, the  mass of soil and volume of
                    solventused, and TPH results. Additional information was
                    provided in cases  when dilutions  had  been necessary.
                    Although  the  data package  preparation  time  was
                    30 minutes in the field, during the weeks  following the
                    demonstration, Horiba spent additional time revising the
                    data package to address EPA comments.  The revisions
                    primarily consisted of identifying in the data package TPH
                    results below the reporting  limit based on  the revised
                    MDL.  The amount of additional time that Horiba spent

-------
finalizing the data package could not be quantified and
was not included as part of the time for TPH measurement.

For the reference method, time measurement began when
the reference  laboratory received  all the  investigative
samples and continued until the EPA received the  first
draft data package from the laboratory. The reference
laboratory took 30 days to deliver the first draft  data
package  to the EPA.   Additional time taken by the
reference laboratory to address EPA comments on all the
draft laboratory data packages was not included as part of
the time required for TPH measurement.

7.2    Secondary Objectives

This section discusses the performance results for the
OCMA-350 in terms of the secondary objectives stated in
Section 4.1. The  secondary objectives were addressed
based on (1)  observations of the  performance of the
OCMA-350 during the demonstration and (2) information
provided by Horiba.

7.2.1  Skill and Training Requirements for
       Proper Device  Operation

Based on observations made during the demonstration, the
OCMA-350 is easy  to operate,  requiring  one  field
technician with basic wet chemistry  skills acquired on the
job or in a university.   Because the amount of sample
drying agent  (sodium  sulfate)  used and the  sample
extraction time may need to be adjusted based on soil type
and moisture content, basic knowledge of soil types is also
recommended  so that the  technician can differentiate
among sand, silt,  and clay soil types and modify the
sample drying and extraction steps  accordingly.  During
the demonstration, Horiba chose  to  conduct  sample
analyses using one technician.

According to  Horiba,  normal  training for  using the
OCMA-350 is limited to reading the instruction manual
and simplified operating instruction sheet that are included
with the device. The sample analysis procedure for the
OCMA-350 can be learned in the field with a few practice
attempts.  In  addition,  during  regular business hours,
Horiba provides technical support over the telephone at no
additional cost. Horiba also offers a 1-day training course
at its facility or at a user-specified site for $960; for this
cost,  Horiba  provides  one OCMA-350  for training
purposes and one instructor to train up to 10 people. If the
training is conducted at a user-specified site,  the  user
would  have to pay an additional cost to  cover the
instructor's travel and per diem for 1 day. Horiba does not
provide a training video for the OCMA-350.

With  the  OCMA-350,  minimal effort  is  required to
calculate a TPH concentration because the mass of soil and
volume of solvent used during sample extraction can be
programmed into the device, and the concentration can be
read from the digital display in units of mg/kg TPH on a
wet weight basis. For a diluted sample extract, the user
must multiply the digital display reading by an appropriate
dilution  factor.    During the  demonstration,  Horiba
connected the OCMA-350 to a printer that printed a hard
copy of the TPH concentrations.  After the demonstration,
Horiba made minimal revisions to the TPH results reported
in the field. Specifically, of the 208 TPH results reported
in the field at the end of the demonstration, fewer than
5 percent were corrected based on EPA review of the data
package.   The corrections primarily involved use of
inappropriate reporting limits.

During the demonstration, Horiba used supplies and took
steps  that  significantly  increased the time and effort
associated with sample preparation and analysis activities.
For example, to calculate TPH  concentrations, Horiba
used the built-in function of the OCMA-350 that specified
the  standard mass  of soil and volume of  solvent used
during sample extraction.  As a result, Horiba had to
measure  exactly 5.0 grams of  each  soil   sample and
carefully place the measured amount in a narrow-mouthed,
40-mL vial without losing any mass; this  step took a
significant  amount of  the  field technician's  time (an
average of 5 minutes per sample). Taking full advantage
of the device features that allow use of varying amounts of
soil and solvent during sample extraction as well as using
vials with wider mouths would have  made the extraction
process less time-consuming.  In addition, Horiba used a
gravity  filtration setup  that required significant  time,
space, and supplies-such as glass funnels, a runnel rack,
Whatman No. 40 filter paper, and glass beakers-and that
may have resulted in loss of some volatiles.   Use of
disposable filter syringes may have resulted in savings of
time, space, and supplies. Horiba also used its Model SR-
300 solvent reclaimer to  recycle spent solvent. Although
the solvent reclaimer was easy to set up and use, it took a
significant  amount  of time to  generate  solvent.   In
summary, some of the TPH measurement steps described
above reduced the quantity of waste generated during the
demonstration but made the TPH measurement procedures
less simple and more time-consuming.
                                                    89

-------
7.2.2  Health and Safety Concerns Associated
       with Device Operation

The  chemical hazards associated  with operating the
OCMA-350 are considered to be insignificant. The two
chemicals  required  for use  of the  OCMA-350 are
anhydrous  sodium  sulfate drying agent and Horiba's
proprietary S-316 extraction  solvent, both of which are
nontoxic.

During the demonstration, the Horiba  field  technician
operated the OCMA-350 in modified Level D personal
protective equipment (PPE) to prevent eye and skin contact
with chemicals.  The PPE included safety glasses, work
boots, work clothes with long sleeves and long pants, and
occasionally disposable gloves when soil samples  were
wet.  Sample analyses were performed outdoors in a well-
ventilated area; therefore, exposure to volatile chemicals
through inhalation was not a concern.

During the demonstration, Horiba used the Model GE-50
ultrasonic mixer, which is no longer available from the
developer. According to Horiba, the ultrasonic mixer that
is currently available (Model VC-50) is functionally the
same as  Model  GE-50.  A potential physical  hazard
observed during the demonstration involved  use of the
ultrasonic mixer, which emitted a very high-frequency
sound; Horiba's field technician wore  ear plugs while
using the mixer to prevent ear damage.

In general, a user of the OCMA-350 should employ  good
laboratory practices during sample analysis to minimize
exposure to potential chemical and physical hazards.
Example  guidelines  for good  laboratory practices are
described   in  ASTM's  "Standard  Guide   for Good
Laboratory Practices in Laboratories Engaged in Sampling
and Analysis of Water" (ASTM 1998).

7.2.3  Portability of the Device

The  OCMA-350 infrared  spectrophotometer  and  its
accessories come in an optional, hard-plastic carrying case
and weigh 36 pounds. The carrying case is 22 inches  long,
16 inches wide, and 15 inches high. The OCMA-350 is
operated using a 110- or 220-volt AC power source or a
DC power  source such as a  12-volt power outlet in  an
automobile along with a DC/AC power inverter. Because
the OCMA-350 comes in an optional carrying case, and
because an AC power source is not required to operate the
device, it is easily transported between sampling areas in
the field.
Other supplies that Horiba used during the demonstration
come in several boxes but not in carrying cases that would
facilitate transport of the supplies between sampling areas.
As a result, significant effort would be required for the
user to move these supplies between sampling areas in the
field.  For example, the Model SR-300 solvent reclaimer
and 7-kg bottle of S-316 extraction solvent each come in
a  cardboard  box and weigh  15  and  18  pounds,
respectively.  Glassware and other laboratory supplies-
such as  40-mL vials, a test tube rack, runnels, a funnel
rack,  and beakers-also  come  in separate boxes.   In
addition, the Model GE-50  ultrasonic mixer (6  pounds)
and the  Epson LQ-500  dot-matrix printer  (10  pounds)
come in separate boxes and  require an AC power source
for operation; however, because a typical user does not
require these two items for TPH measurement using the
OCMA-350, they are considered to be optional.   In
summary,  the portability of the OCMA-350 and other
required supplies depends on the number of samples to be
analyzed, site characteristics (for example, the availability
of AC power), and the needs of the user (for example, a
printer is needed only if a printout of the TPH results is
required).

To operate the  OCMA-350, a sample preparation and
analysis area is required. The area must be large enough
to accommodate the infrared spectrophotometer and  its
accessories, the  Model  SR-300 solvent reclaimer, and
glassware  and other laboratory supplies. The size of the
area depends on the number of samples to be analyzed and
is thus project-specific. During the demonstration, Horiba
performed sample preparation and analysis under one 8- by
8-foot tent that housed two 8-foot-long, folding tables; two
folding  chairs;  one  20-gallon  laboratory  pack  for
flammable waste; and one  55-gallon drum for general
refuse. Horiba used the AC power source available at the
demonstration site to  operate the OCMA-350, ultrasonic
mixer, and printer.

7.2.4  Durability of the Device

Horiba  used   several  reusable   items   during  the
demonstration. Of the reusable items that Horiba provides,
the   infrared  spectrophotometer  and  Model   GE-50
ultrasonic mixer contain mechanical or electronic parts that
could potentially malfunction during sample analysis.
Based on  observations made during the demonstration,
these  reusable   items  are  durable;  neither of them
malfunctioned or  was  damaged.    These  items  are
manufactured by Horiba and are distributed by Horiba and
authorized distributors of Horiba products.  The  infrared
                                                   90

-------
spectrophotometer is housed in an optional, hard-plastic
carrying case to prevent damage during transport, and the
Model GE-50 ultrasonic  mixer comes in a padded,
cardboard box.

The Mettler-Toledo digital balance (Model BD202) and
the Epson LQ-500 dot-matrix printer are reusable supplies
that were used during the demonstration but that Horiba
does not provide.  These supplies also proved to be durable
and did not malfunction or become damaged.

Based on observations made during the demonstration, the
operation of the OCMA-350 was unaffected by the varying
temperature and humidity conditions encountered between
8:00 a.m. and 5:00 p.m. on any given day.  During the
daytime, the temperature ranged from about 17 to 24 °C,
and the relative humidity ranged from 53 to 88 percent.
During sample analysis, wind speeds up to 20 miles per
hour did not affect the operation of the reusable items.

7.2.5  A variability of the Device and Spare Parts

During the demonstration, none  of the reusable  items
provided  by   Horiba-specifically,   the   infrared
spectrophotometer, Model SR-300  solvent  reclaimer,
Model GE-50  ultrasonic  mixer, and  Centronics printer
cable-required repair  or  replacement.  Except for the
Centronics printer cable, replacement parts for these items
would not have been available in stores.  The Centronics
printer cable can be replaced with other brands of cable
that are compatible with the OCMA-350 and are available
in computer  hardware stores.  According  to  Horiba,
replacement parts for the other items can be obtained from
Horiba or an authorized distributor of Horiba products
within 2 days, depending on the shipping procedures and
site location. Except for an extra 3.15 -ampere fuse for the
infrared spectrophotometer, spare parts for reusable items
are  not  provided  with  the OCMA-350.   Horiba
recommends that the infrared spectrophotometer, Model
SR-300 solvent reclaimer, and Model GE-50 ultrasonic
mixer be returned to Horiba or an authorized distributor of
Horiba products for service if they malfunction. Because
Horiba provides a 1-year warranty for these  items, they
will be repaired at no additional cost during the warranty
period.   According to  Horiba,  repairs are generally
completed in 1 week  or less, depending on the repairs
needed.

All reusable items not provided by Horiba can be obtained
from a scientific equipment or computer hardware supplier,
as appropriate.

Of the expendable items used during the demonstration,
the S-316 extraction solvent can be obtained from Horiba
or  an  authorized  distributor of  Horiba products.
Expendable items not provided by Horiba can be obtained
from a scientific equipment supplier.
                                                   91

-------
                                              Chapter 8
                                         Economic Analysis
As discussed throughout this ITVR, the OCMA-350 was
demonstrated by using it to analyze  soil environmental
samples, soil PE samples, and liquid PE samples.  The
environmental  samples were  collected  from  three
contaminated sites, and the PE samples were obtained from
a  commercial  provider,   ERA.    Collectively,  the
environmental and PE samples provided the different
matrix types and the different levels and types of PHC
contamination needed to  perform  a  comprehensive
economic analysis for the OCMA-350.

During the demonstration, the OCMA-350 and the off-site
laboratory reference method were each used to perform
more  than 200  TPH analyses.   The purpose of  the
economic analysis was to estimate the total cost of TPH
measurement for the  OCMA-350 and then compare this
cost to that for the reference method. The cost per analysis
was not estimated for the OCMA-350 because the cost per
analysis would increase as the number of samples analyzed
decreased. This increase would be primarily the result of
the distribution of the initial capital equipment cost across
a smaller number of samples.  Thus,  this increase in the
cost per analysis cannot be fairly compared to the reference
laboratory's fixed cost per analysis.

This chapter provides information  on  the issues and
assumptions  involved  in  the  economic   analysis
(Section 8.1), discusses the costs associated with using the
OCMA-350  (Section 8.2), discusses the costs associated
with using the reference method (Section 8.3), and presents
a comparison of the economic  analysis results for  the
OCMA-350 and the reference method (Section 8.4).

8.1    Issues and Assumptions

Several factors affect TPH measurement costs. Wherever
possible in this chapter, these factors are identified in such
a way that decision-makers can independently complete a
project-specific economic analysis.   The following five
cost categories were included in the economic analysis for
the demonstration: capital  equipment, supplies, support
equipment, labor, and IDW disposal.   The issues and
assumptions associated with these categories and the costs
not included in the analysis are briefly discussed below.
Because the reference method costs were  based on a fixed
cost per analysis, the issues and assumptions discussed
below apply only to the OCMA-350 unless otherwise
stated.

8.1.1   Capital Equipment Cost

The capital equipment cost was the cost associated with the
purchase  of  the   OCMA-350   used  during   the
demonstration. This cost was obtained  from a standard
price list provided by Horiba. Because the device must be
purchased, no salvage value was included in the capital
equipment cost.

8.1.2   Cost of Supplies

The cost of supplies was estimated based on the supplies
required to analyze all demonstration samples using the
OCMA-350  that were  not  included  in the  capital
equipment cost category.  The supplies that Horiba used
during the demonstration fall into two general categories:
expendable  and  reusable.   Examples  of expendable
supplies include anhydrous sodium sulfate and solvent as
well as disposable gloves and pipettes; however, Horiba
promotes  solvent recovery and reuse by  providing an
optional accessory, the Model SR-300 solvent reclaimer.
Examples of reusable supplies include beakers and a
digital balance. During the demonstration, the types and
quantities of all supplies used by Horiba were noted each
day.

For supplies provided by Horiba during the demonstration,
Horiba's costs were used to estimate the cost of supplies.
The costs for supplies  not provided by  Horiba were
                                                   92

-------
estimated based on price quotes from independent sources.
Because a user cannot typically return unused supplies, no
salvage value for supplies that were not used during the
demonstration was included in the cost of supplies.

8.1.3  Support Equipment Cost

During the demonstration, the OCMA-350, the Model
GE-50 ultrasonic mixer, and the Epson LQ-500 dot-matrix
printer were operated using an AC power source.  The
costs associated with providing the power supply and the
electrical energy consumed were not included in the
economic analysis; the  demonstration site provided AC
power at no cost. Of the three items mentioned above, the
OCMA-350 can also be operated using a DC power source
such as an  automobile cigarette lighter.  A typical user
does not require the other two items for TPH measurement
using the OCMA-350, so these items are considered to be
optional.

Because of the large number of samples analyzed during
the demonstration, the EPA provided support equipment,
including  a  tent,   tables,  and chairs,  for the  field
technician's  comfort during  sample  extraction and
analysis.    For  the economic  analysis,  the  support
equipment  costs were estimated based on price quotes
from independent sources.

8.1.4  Labor Cost

The labor cost was estimated based on the time  required
for OCMA-350 setup, sample preparation, sample analysis,
and summary data package preparation. The data package
included, at a minimum, a result summary table, a run log,
and any supplementary information submitted by Horiba.
The measurement of the time required for  Horiba to
complete all analyses and submit the data package to the
EPA was  rounded  to the nearest half-hour.   For the
economic analysis, it was assumed that a field technician
who had worked for a fraction of a day would be paid for
an entire 8-hour  day. Based on this assumption, a daily
rate for a field technician was used in the analysis.

During the demonstration, EPA representatives evaluated
the skill level required for the field technician to complete
analyses and calculate TPH concentrations. Based on the
field observations,  a field technician with  basic wet
chemistry skills acquired on the job or in a university and
a few hours of device-specific training was considered to
be qualified to operate the OCMA-350. For the economic
analysis, an hourly  rate of $16.63 was used  for a field
technician (R.S. Means Company [Means] 2000), and a
multiplication factor of 2.5 was applied to labor costs in
order to account for overhead costs.  Based on this hourly
rate and multiplication factor, a daily rate of $332.60 was
used for the economic analysis.

8.1.5  Investigation-Derived Waste Disposal Cost

During the demonstration, Horiba was provided with two
20-gallon laboratory packs for collecting hazardous wastes
generated (one for flammable wastes and one for corrosive
wastes) and was charged for each laboratory pack used.
Unused samples and sample extracts, used EnCores, and
unused chemicals that could not be returned to Horiba or
an independent vendor were disposed of in a laboratory
pack. Items such as used PPE and disposable glassware
were disposed of with municipal garbage in accordance
with demonstration site waste disposal guidelines. Horiba
was  required to provide any containers necessary  to
containerize individual wastes prior to their placement in
a laboratory pack.  The cost for these containers was not
included in the IDW disposal cost estimate.

8.1.6   Costs Not Included

Items whose costs were not  included in the economic
analysis are identified below along with a rationale for the
exclusion of each.

Oversight of Sample Analysis Activities. A typical user
of the  OCMA-350 would not be  required  to  pay for
customer  oversight  of   sample  analysis.     EPA
representatives audited all activities associated with sample
analysis during the demonstration, but costs  for EPA
oversight were not included  in the economic analysis
because these  activities  were project-specific.  For the
same reason, costs for EPA oversight of the reference
laboratory were also not included in the analysis.

Travel  and Per Diem  for Field  Technicians.  Field
technicians may  be available  locally.   Because  the
availability of field technicians is primarily a function of
the location of the project site, travel and per diem costs
for field technicians were not included in the economic
analysis.

Sample Collection and Management. Costs for sample
collection  and management activities,  including sample
homogenization and labeling, were not included  in the
economic analysis because these activities were project-
                                                    93

-------
specific and  were  not  device-  or reference  method-
dependent.

Shipping.  Costs for shipping (1)  the OCMA-350 and
necessary supplies to the demonstration site and (2) sample
coolers to the reference laboratory were not included in the
economic analysis because such costs vary depending on
the shipping distance and the service used (for example, a
courier or overnight shipping versus economy shipping).

Items Costing Less Than $10.  The cost of inexpensive
items such as ice used for sample preservation in the field
was  not included in  the economic  analysis because the
estimated cost was less than $10.

8.2    OCMA-350 Costs

This section presents information on the individual costs of
capital equipment, supplies, support equipment, labor, and
IDW disposal for the OCMA-350 as well as a summary of
these costs.  Additionally, Table  8-1  summarizes the
OCMA-350 costs.

8.2.1  Capital Equipment Cost

The capital equipment cost was the cost associated with the
purchase of the OCMA-350.  Horiba does not rent the
OCMA-350.  As discussed in Chapter 2, the OCMA-350
consists  of the  infrared  spectrophotometer  and  its
accessories, including (1)  a proprietary, 10-mm,  quartz
cuvette with a Teflon® cap; (2) a 25-^L microsyringe; (3)
a 10-mL syringe; (4) 10 mL of B-heavy oil for calibration;
(5) three 40-mL, VOA vials; (6) three  11-cm-diameter,
No.  40 Whatman filter  papers; and  (7) two  10-mL,
disposable, glass pipettes. The OCMA-350 also contains
an  instruction  manual  and  a  simplified   operating
instruction sheet. The OCMA-350 can be purchased from
Horiba for $6,500.

8.2.2  Cost  of Supplies

The  supplies that Horiba used during the demonstration
fall into two general categories: expendable and reusable.
Table 8-1 lists all the expendable and reusable supplies that
Horiba used during the demonstration. Of the expendable
supplies, Horiba's proprietary  S-316 extraction solvent
was used for extracting PHCs from soil  and for cleaning
glassware, anhydrous sodium sulfate was used for drying
wet soil samples, and the VOA vials were used for soil
sample extraction; the use of the other expendable supplies
is self-explanatory.   Of  the  reusable  supplies,  the
Model GE-50 ultrasonic mixer was used to disperse soil
samples in the extraction solvent, and the Model SR-300
solvent reclaimer was used to recycle spent solvent; the use
of the other reusable supplies is self-explanatory.

Cost information  for  all the supplies used during the
demonstration is presented in Table 8-1.  The total cost of
the supplies used by Horiba during the demonstration was
$6,715 (the cost of each item was rounded to the nearest
$1).   Of these  supplies, the S-316 extraction solvent,
Model GE-50 ultrasonic mixer, Model SR-300 solvent
reclaimer, and Centronics printer cable are available from
Horiba. The other supplies have to be purchased from a
retail  vendor of laboratory  supplies.   The Centronics
printer cable can be replaced with other brands that are
compatible with the OCMA-350 and are available in
computer hardware stores.

During the demonstration, Horiba also used the following
supplies that cost less than $10 each: (1) a 280-count box
of Kim wipes® ($3.60) for wiping the quartz cuvette and (2)
a 50-pound bag of concrete sand ($1.89) used as blank
material for QC checks (only about 50 grams of sand was
used during the  demonstration).

Other supplies  that can be  purchased from Horiba if
additional quantities are needed include the proprietary,
10-mm, quartz cuvette ($535) with the Teflon® cap ($60);
the  25-(iL microsyringe ($210); 10-mL syringe  ($34);
10 mL of B-heavy oil for calibration ($11); a 3.15-ampere
fuse ($3); a power supply cable ($17); and the instruction
manual ($25). Horiba provides an electronic copy of the
simplified operating instruction sheet at no cost to the user.
During  the  demonstration,  Horiba did  not require
additional quantities of these supplies; therefore, the costs
of these items are not listed in Table 8-1.

8.2.3   Support Equipment Cost

Horiba was provided with  one  8- by  8-foot  tent  for
protection   from   inclement  weather  during  the
demonstration as well as two tables and two  chairs for use
during sample preparation and analysis activities.  The
purchase cost for the tent ($159) and the rental cost for two
tables and two chairs for 1 week ($39) totaled $198.

8.2.4  Labor Cost

One field technician was required for 6 days during the
demonstration to complete all sample analyses and prepare
the  summary data package.   Based on a  labor rate of
                                                    94

-------
Table 8-1. OCMA-350 Cost Summary
Item
Capital equipment
Purchase of infrared spectrophotometer and its accessories
Supplies
Expendable
S-316 extraction solvent (7-kilogram bottle)
Anhydrous sodium sulfate, 10-60 mesh (500-gram bottle)
40-mL, VOA vials (72 per box)
1 1 -cm-diameter, No. 40 Whatman filter papers (100 per box)
10-mL, disposable, glass pipettes (12 per case)
Disposable, latex gloves (100 per box)
Reusable
Model GE-50 ultrasonic mixer
Model SR-300 solvent reclaimer
Epson LQ-500 dot-matrix printer
Centronics printer cable
Mettler-Toledo digital balance (Model BD202)
Stainless-steel spatulas (three per pack)
Filter forceps
Glass funnel
50-mL, glass beakers (12 per pack)
Pipette bulb
10-microliter microsyringe
VWR Model 51 1 bottle-top dispenser
Test tube rack (24 holes)
Funnel rack and stand (holds four small funnels)
Support equipment
Tent
Tables and chairs (two each)
Labor
Field technician
Investigation-derived waste disposal
Total Cost"
Quantity

1 unit


1 unit
2 units
3 units
2 units
1 unit
1 unit

1 unit
1 unit
1 unit
1 unit
1 unit
1 unit
1 unit
4 units
1 unit
1 unit
1 unit
1 unit
1 unit
1 unit

1 unit
1 set for 1 week

6 person-days
1 20-gallon container

Unit Cost ($)

6,500


1,040
19.75
62.50
26
75.60
20.50

2,080
1,650
275
40
715
20.30
20
17.67
38.80
19.50
19.80
234.90
22.50
90

159
39

332.60
345

Itemized Cost3 ($)

6,500


1,040
40
188
52
76
21

2,080
1,650
275
40
715
20
20
71
39
20
20
235
23
90

159
39

1,996
345
15,750
Notes:

cm     =   Centimeter
ml     =   Milliliter
VOA   =   Volatile organic analysis

a    Itemized costs were rounded to the nearest $1.

b    The total dollar amount was rounded to the nearest $10.
                                                              95

-------
$332.60 per day, the total labor cost for the OCMA-350
was $1,996 (rounded to the nearest $1).

8.2.5  Investigation-Derived Waste Disposal Cost

Horiba used one  laboratory pack to collect  flammable
hazardous waste, including unused soil and liquid samples
that contained PHCs  and used EnCores and ampules,
generated during the demonstration.  Horiba collected all
the proprietary S-316 extraction solvent, including unused
solvent;  reclaimed  solvent;  and to-be-reclaimed, used
solvent,  in  different  bottles.   At the  end  of the
demonstration, Horiba took all the solvent to its laboratory
for reuse. As a result, the IDW disposal cost included only
the purchase cost of the laboratory pack ($38) and the cost
associated with hazardous waste disposal in a landfill
($307) (Means 2000). The total IDW disposal cost was
$345.

8.2.6  Summary of OCMA-350 Costs

The total cost for performing more than 200 TPH analyses
using the OCMA-350 and for preparing a summary data
package was $15,750 (rounded to the nearest $10). The
TPH analyses were  performed for 74 soil environmental
samples, 89 soil PE samples, and 36 liquid PE samples. In
addition to these 199 samples,  9 extract duplicates were
analyzed for specified soil environmental samples. When
Horiba  performed  multiple extractions,  dilutions,  or
reanalyses for a sample, these  were not included in the
number of samples analyzed. During the demonstration,
the  multiple  extractions,  dilutions, and  reanalyses
collectively required only about 6 percent more supplies
than would otherwise have been needed.  The total cost
included $6,500 for capital equipment; $6,715 for supplies;
$198  for support equipment; $1,996 for labor; and $345
for IDW disposal.  Of these five costs, the two largest were
the capital cost (41 percent of the total cost) and the cost of
supplies (43  percent of the total cost).

A significant portion of the cost of supplies was associated
with the Model GE-50 ultrasonic mixer (31 percent) and
Model SR-300 solvent reclaimer (25 percent). Although
these two items were useful during the demonstration, they
are considered to be optional,  and their purchase  is
justified  only when a large number of samples will be
analyzed. When a small number of samples will be
analyzed, the ultrasonic mixer can  be replaced with a
cheaper means of agitation to disperse samples in solvent.
Purchase of the solvent reclaimer should be  considered
only after evaluation of the disposal cost for spent solvent,
purchase  cost  for  virgin  solvent,   and  potential
environmental concerns associated with hazardous waste
disposal.

8.3     Reference Method Costs

This  section  presents the costs associated  with the
reference  method  used to analyze the demonstration
samples for TPH.  Depending on the nature of a given
sample, the reference laboratory analyzed the sample for
GRO,   EDRO,  or  both  and  calculated  the  TPH
concentration   by  adding   the   GRO  and  EDRO
concentrations, as appropriate.  The reference method costs
were  calculated using unit cost information  from the
reference  laboratory invoices.   To allow  an  accurate
comparison of the OCMA-350 and reference method costs,
the reference method costs were estimated for the same
number of samples as was analyzed  by Horiba.  For
example,  although  the  reference  laboratory  analyzed
MS/MSD samples for TPH and all soil samples for percent
moisture, the associated sample analytical costs were not
included in the reference method  costs because Horiba did
not analyze MS/MSD samples  for TPH or soil samples for
percent moisture during the demonstration.

Table 8-2  summarizes the reference method costs, which
totaled  $42,050.    This  cost covered  preparation of
demonstration samples and their analysis for  TPH.  In
addition, at no additional cost, the reference laboratory
provided (1) analytical  results  for internal QC  check
samples such as method blanks and LCS/LCSDs and (2) an
electronic data deliverable and two paper copies of full,
EPA Contract Laboratory Program-style data packages
within  30 calendar days of the  receipt  of the  last
demonstration sample by the reference laboratory.

8.4     Comparison of Economic Analysis Results

The total  costs for the OCMA-350 ($15,750) and the
reference method ($42,050) are  listed in Tables 8-1 and
8-2, respectively. The total TPH measurement cost for the
OCMA-350 was 62 percent less than that for the reference
method. Although the OCMA-350 analytical results did
not have the same level of detail (for example, carbon
ranges) as the reference method analytical results or
comparable QA/QC data, the OCMA-350 provided TPH
analytical  results on site at significant cost  savings. In
addition, use of the OCMA-350 in the field will likely
produce additional cost savings because the results will be
available  within  a few  hours  of sample collection;
therefore,  critical decisions  regarding  sampling  and
                                                    96

-------
Table 8-2. Reference Method Cost Summary
Item
Soil environmental samples
GRO
Extract duplicates
EDRO
Extract duplicates
Soil performance evaluation samples
GRO
EDRO
Liquid performance evaluation samples
GRO
EDRO
Total Cost"
Number of Samples Analyzed

56
7
74
9

55
89

27
24

Cost per Analysis ($)

111
55.50
142
71

111
142

111
106.50

Itemized Cost3 ($)

6,216
388
10,508
639

6,105
12,638

2,997
2,556
42,050
Notes:

a    Itemized costs were rounded to the nearest $1.
b    The total dollar amount was rounded to the nearest $10.
analysis  can be made in the field, resulting  in a more
complete data set.  However, these  savings  cannot be
accurately estimated and thus were not included in the
economic analysis.
                                                       97

-------
                                               Chapter 9
                               Summary of Demonstration Results
As discussed throughout this ITVR, the OCMA-350 was
demonstrated by using it to analyze 74 soil environmental
samples, 89 soil PE samples, and 36 liquid PE samples. In
addition  to  these  199  samples,  9 extract  duplicates
prepared using the environmental samples were analyzed.
The  environmental  samples were  collected  from five
individual areas at three contaminated sites, and the PE
samples were obtained from a commercial provider, ERA.
Collectively, the environmental and PE samples provided
the different matrix types and the different levels and types
of PHC contamination needed to perform a comprehensive
evaluation of the OCMA-350.

The OCMA-350 performance and cost data were compared
to those for an off-site laboratory reference  method,
SW-846 8015B (modified). As discussed in Chapter 6, the
reference method results were considered to be of adequate
quality forthe following reasons: (1) the reference method
was implemented with acceptable accuracy (±30 percent)
for all the samples except low- and medium-concentration-
range soil samples containing diesel, which made up only
13 percent of the total number of samples analyzed during
the  demonstration, and (2)  the reference method was
implemented with good precision  for all samples.  The
reference method  results generally exhibited  a negative
bias. However, the bias was considered to be  significant
primarily  for  low-  and medium-range  soil  samples
containing  diesel.    The reference  method  recoveries
observed during the  demonstration were typical of the
recoveries obtained by most organic analytical methods for
environmental  samples.
This chapter compares the performance and cost results for
the OCMA-350 with those for the reference method, as
appropriate.   The  performance and  cost  results  are
discussed  in  detail  in Chapters 7 and 8, respectively.
Tables 9-1 and 9-2 summarize the results forthe primary
and secondary objectives, respectively. As shown in these
tables, during the demonstration, the OCMA-350 exhibited
the following desirable  characteristics  of a  field TPH
measurement  device: (1)  good precision, (2) sensitivity to
interferents that are PHCs (MTBE and Stoddard solvent),
and  (3)  high  sample  throughput.   In  addition,  the
OCMA-350 exhibited moderate measurement costs.

In general, the OCMA-350 TPH results forthe PE samples
did not compare well with the reference method results. In
addition, turpentine  biased the OCMA-350 TPH results
high, whereas humic acid biased the results low.  These
findings indicated that the accuracy of TPH measurement
using the  device  will likely be impacted by naturally
occurring oil and grease and organic matter present in soil
samples. PCE and 1,2,4-trichlorobenzene did not impact
the OCMA-350 results in a consistent manner when the
effect of these interferents was tested using neat materials
and spiked soil samples.  Finally, the device results were
significantly  impacted  by  soil  moisture  content:  a
7 percentage  point  increase  in soil moisture content
resulted in (1) a three-fold increase in  TPH  results for
weathered gasoline soil PE samples and (2) a three-fold
decrease in TPH results for diesel  soil PE samples.
Collectively, the demonstration findings indicated that the
OCMA-350  may be considered for  TPH  screening
purposes; however, the user should exercise caution when
considering the device  for a field  TPH measurement
application requiring definitive results.
                                                   98

-------
 £
ll

1 8
o c
Of the 108 OCMA-350 results, 1 result was inc
78 (73 percent) agreed with those of the refere
positives, and 23 were false negatives.



CO
OJ "O
£ g
a ° ra
j£ «> £
c .c ^
Comparison of project-specific actic
conclusions of the OCMA-350 with
reference method for 74 soil enviror
soil PE samples
^
0 X
2 CL
=: 1—
o „_
O o
ra ^^
0) § S
£ 'in E
d) O d)
ro i— 3
3 Q- (/)
g -o g
LJJ ra E

C\l
CL
in


CD "5 in"
ni ^
in" ^ 
£ . £ . g_J
*— ^ LO ^ O "O
8__| s^ ^|
O m O m •— ™
CO (/) CO (/) -^ Q


^ al ^ al ^ co
36 of 102 OCMA-350 results (35 percent) were
OCMA-350 results were biased high, and 20 v\
16 of 102 OCMA-350 results (16 percent) were
OCMA-350 results were biased high, and 1 1 v\
50 of 102 OCMA-350 results (49 percent) were
17 OCMA-350 results were biased high, and 3
00
»- C\l
O -n
tn ra
o —
£ £
j- C
1 §
|I

Comparison of OCMA-350 TPH res
the reference method for 74 soil em
soil PE samples


















m ® T3 , 0) ° 0)
S. 3 o -c o) £ y>
d) " ^ E

I'D 13'SE 'S -
 £ '5 o-Q
**-(/) (/) CO
0) (/) — ^_^ ^ Q.
-^ — W "f~ C -!-•
^ -. OJ m ^ "(JO
° QJ ^ ^ S -^
s 5 fi 8. i^
£ <5 w m -^ c
o — ^ S £ ra
« s g £ JS S
^ £r .
0 21
si - o.
~u m E
c c ... — ro

Q) C £ (/)(/)
ra £ !S LJJ "dj
(u S CL tn
!~ o) ra — .0)
Correlation (as determined by lineal
analysis) between OCMA-350 and r
TPH results for (1 ) soil environment
collected from five areas and (2) so
including weathered gasoline and d















                                                                       99

-------
I
c
o
o.

U)

.1
CL

0)
U)
±i
o
in
O
O
re

E









ance Results
p
o
t
 %

 c P 1
LU ro E 3-

C\l
CL








C
BE: 39 percent
E: 17.5 percent
iddard solvent: 85 percent
•pentine: 52 percent
,4-Trichlorobenzene: 50 perce
mic acid: 0 percent
i- o .2 3 *l
i^.-c
= •"• £
0 T3 ^
C C 3
— ro -a
in • - 
2 to ^



T3 £
(D 'S
(/) .n
^ -»-•
ro ro
MTBE, a petroleum hydrocarbon, <
statistically significant interference
levels.
ro
_>.
c
o
(D
O
c
(D
1
GJ
"c

E caused statistically significal
high level.
o &
Q- £







PCE results were inconclusive.

T3 i-
QJ O
(/)
8 1j
c —
O £
O ^_i
>- O (/)

T3 0) E
iddard solvent, a petroleum hy
tistically significant interfereno
athered gasoline and diesel Sc
£ ro a)
W to 3


cf
O (U
.a o „, >-
-tl c a) o
ro Q) .^
^ m "o "OJ
2 «Si (/) >
Stoddard solvent, a petroleum hyd
caused statistically significant inter
(1) at both levels for weathered ga
samples and (2) only at the high le
diesel samples.
(D T3
o c

2 in
(U p
•^ c
ro (/)
^ W 0)
C QJ o
SC c
— c
c o ro
'c w w
n\ ro —
•pentine caused statistically si(
at both levels for weathered g
only at the high level for diese
^ *~* *~*
i- ^Cl



„
ro 2
£ 0)
•F -c
Turpentine caused statistically sigr
interference at both levels for weat
gasoline and diesel samples.

ro
o

'c
D1

8


,4-Trichlorobenzene caused si
irference only at the high level
c\i .2
^_- c


-QJ
= >
.8 ~
to D)
-*-1 ~<~
1 ,2,4-Trichlorobenzene caused sts
significant interference only at the







0)

mic acid results were inconclu:
^
X



•£
8
ti=
'c
Humic acid caused statistically sig
interference at both levels.
g
LU '"> £
CL ^ — 3
«*— d) Q. T3 -*-
sis i« 11
S 7K o m ~ -a ra
•^ 0) 3 LLJ C 0 b)~
c .2 r ^ g. M g
™ 2 !^ o ^  t!
>-^ J2 in — CD <=
ro ^ c /D T3 Q. o
> ro QJ c c P o

111 il s|
-= D) j- "do := J.
^ £ 0) (fl o •*
S S 3 tj
.9 ro ro •? C •? o 'o
O>ro
'o

8= "~
(D C „
QJ O C
£ £ p
.£ £ 2

3 ,2 in
"ro oi ra

LJJ I E

CO
CL




ro
g
ro
to
ro
m
I moisture content did not havi
nificant impact.
O D)
W 'in



=^
>, w
— OJ
O 0) m
Soil moisture content had a statist!
significant impact on both weather
gasoline sample and diesel sampli

Q
in "-

|=|-
?« Q Q. C
	 (O ro
flii
ro ^ ro  'o c
LLJ  0

^-
CL
                                                                 100

-------
I
c
o
o.

0)

.1
0)
±i
o
in
O
O
ra


E





















in
±±
in

CL

























T3
O
.C
^

d>
O
C
2

>
t3
(D
!Q
0
£•
ra
E

CL
T3
"ra <=
-B ra
o of
E ^
c "o.
2 E
•:r ro
> in
 —
m c CD
- (D CO „
<» -5 of J2
3 o) Q> ro
0 _,. Q..9
•=? E Q.
co 14— ro ^
•^ O 0) T3
O

O

"2

CL
2
CL
D)
O

•"
Q.
'(D
O
2
"Q- ra
F ra
c ^*
E "-
o ro
*= "ro
d) "O
E a±
^z; ro
15 -o
?l
(D
Q.
E
= °- ^
3 0) 3 -§,
in ± in =;
ro ^ ro o
a) cr a) c!

in
CL

























En
o
CM"

&
=5 °
11
"^^ OJ
O (/)
in ro

in £
tf> Q.
1 0
Q- O) ro
1- oo S
_ i_ - X
(/) O f QJ
.92 X— O)
"Q- "ro c ~o
o m £ t~
W Q. S ro_
•£ =o 5 <£
^ ^ S °"
^ Q E
•§"~ c w
O" "5 2 LU
^ (0 > D-
ro s_" oj "o
|ll|
*o "c h- °°
"S E ^ ^
o • c ^~
•ii- 3 OJ ^
o> S" | g $
O ^r 3 LLJ ro
'to °- ro °" =
1— in E o) T3


0)
to
£1
!- E
d) d)
E   =>
                                                         ^3 ^3    >
                                                    oj cr OJ  <1J
                                                    CL w ce ce
                                                    M  M  M  M


                                                         Q Q
                                                    LLJ CM  CL W
                                                    CL ce ce ce
ved waste

d>
c
o
ra
D1
ID
c

II
'in
1
o

CL
E
ro
.9>
1

M
1

">,
•c
•v
£*
2

II


d>
"S
o
Z
2
K

M
DJLLJ
ns were m;
o

ro
CL
E
8
8
1
ro
to
<
                                                                       101

-------
Table 9-2. Summary of OCMA-350 Results for the Secondary Objectives
Secondary Objective
S1 Skill and training
requirements for proper
device operation
S2 Health and safety concerns
associated with device
operation
S3 Portability of the device
S4 Durability of the device
S5 Availability of device and
spare parts
Performance Results
The device can be operated by one person with basic wet chemistry skills.
The device's instruction manual and simplified operating instruction sheet are considered to be adequate
training materials for proper device operation. The sample analysis procedure for the device can be learned
in the field with a few practice attempts.
The device's digital display provides a sample's TPH concentration in milligrams per kilogram TPH on a wet
weight basis; minimal effort is required to calculate a TPH concentration. At the end of the demonstration,
Horiba reported 208 TPH results. Of these, fewer than 5 percent required corrections, which primarily
involved use of inappropriate reporting limits.
Some of the device's items, including the narrow-mouth vials used for weighing and extracting soil samples
and the gravity filtration setup, made the TPH measurement procedure less simple and more time-
consuming.
No significant health and safety concerns were noted; when the device is used in a well-ventilated area,
basic eye and skin protection (safety glasses, disposable gloves, work boots, and work clothes with long
sleeves and long pants) should be adequate for safe device operation.
The device can be easily moved between sampling areas in the field, if necessary.
The device can be operated using a 1 1 0-volt alternating current power source or a direct current power
source such as a 1 2-volt power outlet in an automobile.
The primary component of the device, the infrared analyzer, is provided in an optional, hard-plastic carrying
case to prevent damage to the analyzer. During the demonstration, none of the device's reusable items
malfunctioned or was damaged. The moderate temperatures (1 7 to 24 °C) and high relative humidities (53
to 88 percent) encountered during the demonstration did not affect device operation.
During a 1 -year warranty period, Horiba will repair any malfunctioning items of the device at no cost.
Horiba does not supply some items necessary for TPH measurement using the device, including a test tube
rack, a funnel rack and stand, a digital balance, and miscellaneous glassware and laboratory supplies; the
availability of replacement or spare parts not supplied by Horiba depends on their manufacturer or distributor.
                                                         102

-------
                                             Chapter 10
                                             References
AEHS.  1999. "State Soil Standards Survey." Soil &
    Groundwater.  December 1999/January 2000.

API. 1994. "Intel-laboratory Study of Three Methods for
    Analyzing Petroleum   Hydrocarbons  in Soils."
    Publication Number 4599. March.

API. 1996. "Compilation of Field Analytical Methods for
    Assessing Petroleum Product Releases."  Publication
    Number 463 5. December.

API.  1998.  "Selecting Field  Analytical Methods:  A
    Decision-Tree Approach." Publication Number 4670.
    August.

ASTM.   1998.  "Standard  Guide  for Good Laboratory
    Practices in Laboratories Engaged in  Sampling and
    Analysis of Water." Designation: D 3856-95. Annual
    Book of ASTM Standards. Volume 11.01.

California Environmental  Protection Agency.  1999.
    Memorandum Regarding Guidance  for  Petroleum
    Hydrocarbon Analysis.  From Bart Simmons, Chief,
    Hazardous Materials  Laboratory.   To  Interested
    Parties. October 21.

Dryoff, George V. Editor. 1993.  "Manual of Significance
    of Tests for Petroleum Products."  ASTM Manual
    Series:  MNL 1.  6th Edition.

EPA.  1983. "Methods for Chemical Analysis of Water
    and Waste."  Revision.  Environmental Monitoring
    and  Support  Laboratory.  Cincinnati, Ohio.  EPA
    600-4-79-020. March.

EPA. 1996. "Test Methods for Evaluating Solid Waste."
    Volumes 1A through 1C. SW-846.  Third Edition.
    Update III. OSWER. Washington, DC. December.
EPA.  2000.  "Field Measurement Technologies for Total
    Petroleum  Hydrocarbons  in  Soil—Demonstration
    Plan." ORD. Washington, DC. EPA/600/R-01/060.
    June.

Ewing,  Galen  W.   1969.  Instrumental Methods  of
    Chemical Analysis.   Third Edition.   McGraw-Hill
    Book Company.  New York.

Florida Department of Environmental Protection.  1996.
    "FL-PRO Laboratory Memorandum."  Bureau  of
    Waste Cleanup.  Accessed on April  21.  On-Line
    Address: www.dep.state.fl.us/labs/docs/flpro.htm

Fritz,  James S.,  and  George H.  Schenk.    1987.
    Quantitative Analytical Chemistry. Allyn and Bacon,
    Inc.  Boston, Massachusetts. Fifth Edition.

Gary,  J.H.,  and G.E.  Handwerk.  1993.  Petroleum
    Refining: Technology and Economics. Marcel Dekker,
    Inc.  New York, New York.

Massachusetts Department of Environmental Protection.
    2000. "VPH/EPH Documents." Bureau of Waste Site
    Cleanup. Accessed  on April 13.  On-Line Address:
    www. state .ma.us/dep/bwsc/vp_eph .htm
Means.    2000.    Environmental  Remediation
   Data- Unit Price. Kingston, Massachusetts.
Cost
Provost,  Lloyd  P.,  and  Robert S.  Elder.    1983.
    "Interpretation of Percent Recovery Data." American
    Laboratory. December. Pages 57 through 63.

Simard, R.G., Ichiro Hasegawa, William Bandaruk, and C.
    E. Headington. 1951.  "Infrared Spectrophotometric
    Determination of Oil  and  Phenols  in Water."
    Analytical Chemistry. Volume 23, No. 10.  October.
    Pages 1384 to 1387.
                                                  103

-------
Speight, J.G.  1991. The Chemistry and Technology of   Zilis, Kimberly, Maureen McDevitt, and Jerry Parr.  1988.
   Petroleum.  Marcel Dekker, Inc. New York, New       "A Reliable  Technique for Measuring Petroleum
   York.                                                Hydrocarbons in the Environment." Paper Presented
                                                        at the Conference on Petroleum Hydrocarbons and
Texas Natural Resource Conservation Commission. 2000.       Organic Chemicals in Groundwater. National Water
   "Waste Updates."  Accessed on April 13.  On-Line       Well Association (Now Known as National Ground
   Address:   www.tnrcc.state.tx.us/permitting/       Water Association). Houston, Texas.
   wastenews .htm#additional
                                                  104

-------
                                               Appendix
                    Supplemental Information Provided by the Developer
This appendix contains supplemental information provided
by Horiba.  Specifically, updates and improvements to the
OCMA-350 resulting from the SITE demonstration are
summarized below.

Sampling Technique

During the SITE demonstration, the weight of each sample
was measured and adjusted to 5.0 ±0.04 grams to  take
advantage  of the  preset  value  in the  OCMA-350.
Significant time was used to adjust the sample weight to
exactly 5 grams.  It was noted that much time could be
saved by quickly weighing the sample to obtain a nominal
value of 5 grams (say ±0.5 gram) but measuring its weight
to the nearest 0.1 gram. It is a simple matter to reset the
sample weight in the microprocessor of the OCMA-350.

The balance used during the SITE demonstration measured
sample weight to within ±0.01 gram.  The precision of the
balance measurement and the balance' s sensitivity to gusty
winds encountered during the demonstration suggest that
the  OCMA-350  instruction  manual   is  correct in
recommending use  of a balance  with a resolution of
±0.1 gram.

During the  SITE demonstration, a given soil sample was
extracted using Horiba's  proprietary  S-316  extraction
solvent. The extract was then filtered through a Whatman
No. 40 filter by gravity. It has been noted that use  of a
syringe filter could reduce the filtering time by more than
50 percent.
Efficient Use of Resources

The Horiba Model SR-300 Solvent Reclaimer was used
during the SITE demonstration and reduced the cost of
solvent used by reclaiming spent solvent. However, during
the demonstration, only a small portion of the spent solvent
was regenerated because of time constraints.  Greater
efficiency and cost-effectiveness would be achieved by
collecting larger volumes of spent solvent and regenerating
larger batches of spent solvent using the Model SR-300
solvent reclaimer.

Consistency in Sample Preparation

The GE-50 ultrasonic disrupter was used to facilitate
extraction of PHCs in soil. The sonication time varied
from  30 seconds  to  1  minute.   After reviewing the
extraction process, Horiba believes that 1 minute should
have been a fixed value  for the  sonication time for all
samples to be consistent.

Safety Issue

While  using the GE-50  ultrasonic disrupter  during the
SITE   demonstration,  Horiba made  no provision for
donning ear protection.   Horiba now believes that ear
protection is necessary during use of the GE-50 ultrasonic
disrupter.
   This appendix was written solely by Horiba. The statements presented in this appendix represent the developer's point of view and summarize
   the claims made by the developer regarding the OCMA-350. Publication of this material does not represent the EPA's approval or
   endorsement of the statements made in this appendix; performance assessment and economic analysis results for the OCMA-350 are discussed
   in the body of this IT VR.
                                                   105

-------
Efficient Use of Time                                     350's printer port allowed Horiba to get an estimated value
                                                           for subsequent dilution purposes. However, it was learned
During the SITE demonstration, the sample concentration    that  the  time  used to get the printout for the off-scale
was initially measured by allowing the microprocessor of    reading would have been belter spent by simply estimating
the OCMA-350 to achieve a stable reading even if it was    a dilution factor and not waiting for the printout.
off the  scale.   The reasoning was that the  OCMA
   This appendix was written solely by Horiba. The statements presented in this appendix represent the developer's point of view and summarize
   the claims made by the developer regarding the OCMA-350. Publication of this material does not represent the EPA's approval or
   endorsement of the statements made in this appendix; performance assessment and economic analysis results for the OCMA-350 are discussed
   in the body of this IT VR.
                                                        106

-------