United States
Environmental Protection
Agency
Office of Research and
Development
Washington, D.C. 20460
EPA/600/R-98/143
November 1998
Environmental Technology
Verification Report
Photoacoustic Infrared Monitor
Innova AirTech Instruments
Type 1312 Multi-gas Monitor
-------
-------
EPA/600/R-98/143
November 1998
Environmental Technology
Verification Report
Photoacoustic Infrared Monitor
Innova AirTech Instruments
Type 1312 Multi-gas Monitor
by
Wayne Einfeld
Sandia National Laboratories
Albuquerque, New Mexico 87185-0755
IAG DW89936700-01-0
Project Officers
Stephen Billets
Eric Koglin
National Exposure Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Las Vegas, Nevada 89193-3478
-------
Notice
The U.S. Environmental Protection Agency (EPA), through its Office of Research and Development (ORD),
funded and managed, under Interagency Agreement No. DW89936700-01-0 with the U.S. Department of
Energy's Sandia National Laboratory, the verification effort described in this document. This report has received
both technical peer and administrative policy reviews and has been approved for publication as an EPA
document. Mention of corporate names, trade names, or commercial products does not constitute endorsement or
recommendation for use.
11
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, D.C. 20460
ENVIRONMENTAL TECHNOLOGY VERIFICATION PROGRAM
VERIFICATION STATEMENT
TECHNOLOGY TYPE:
APPLICATION:
TECHNOLOGY NAME:
COMPANY
ADDRESS:
PHONE:
PHOTOACOUSTIC INFRARED MONITOR
MEASUREMENT OF CHLORINATED VOLATILE ORGANIC
COMPOUNDS IN WATER
Type 1312 Multi-gas Monitor
Innova AirTech Instruments
Energivej 30
2750 Ballerup, Denmark
(714) 974-5560
PROGRAM DESCRIPTION
The U.S. Environmental Protection Agency (EPA) created the Environmental Technology Verification Program
(ETV) to facilitate the deployment of innovative environmental technologies through verification of performance
and dissemination of information. The goal of the ETV Program is to further environmental protection by
substantially accelerating the acceptance and use of improved and cost-effective technologies. The ETV Program
is intended to assist and inform those involved in the design, distribution, permitting, and purchase of
environmental technologies.
Under this program, in partnership with recognized testing organizations, and with the full participation of the
technology developer, the EPA evaluates the performance of innovative technologies by developing
demonstration plans, conducting field tests, collecting and analyzing the demonstration results, and preparing
reports. The testing is conducted in accordance with rigorous quality assurance protocols to ensure that data of
known and adequate quality are generated and that the results are defensible. The EPA's National Exposure
Research Laboratory, in cooperation with Sandia National Laboratories, the testing organization, evaluated field-
portable systems for monitoring chlorinated volatile organic compounds (VOCs) in water. This verification
statement provides a summary of the demonstration and results for the Innova AirTech Instruments Type 1312
Multi-gas Monitor.
DEMONSTRATION DESCRIPTION
The field demonstration of the Type 1312 photoacoustic infrared monitor was held in September 1997. The
demonstration was designed to assess the instrument's ability to detect and measure chlorinated volatile organic
compounds in groundwater at two contaminated sites: the Department of Energy's Savannah River Site, near
Aiken, South Carolina, and the McClellan Air Force Base, near Sacramento, California. Groundwater samples
from each site were supplemented with performance evaluation (PE) samples of known composition. Both
sample types were used to assess instrument accuracy, precision, sample throughput, and comparability to
EPA-VS-SCM-28
The accompanying notice is an integral part of this verification statement
iii
November 1998
-------
reference laboratory results. The primary target compounds at the Savannah River Site were trichloroethene and
tetrachloroethene. At the McClellan Air Force Base, the target compounds were trichloroethene,
tetrachloroethene, 1,2-dichloroethane, 1,1,2-trichloroethane, 1,2-dichloropropane, and rrara-l,2-dichloropropene.
These sites were chosen because they contain varied concentrations of chlorinated VOCs and exhibit different
climatic and geological conditions. The conditions at these sites are typical, but not inclusive, of those conditions
under which this technology would be expected to operate. A complete description of the demonstration,
including a data summary and discussion of results, may be found in the report entitled Environmental
Technology Verification Report, Photoacoustic Spectrophotometer, Innova AirTech Instruments, Type 1312
Multi-gas Monitor. (EPA/600/R-98/143).
TECHNOLOGY DESCRIPTION
The Type 1312 utilizes photoacoustic spectroscopy for the detection of chlorinated VOCs in the headspace of a
water sample. The vapors from the equilibrium headspace of a stirred water sample are circulated through the
instrument's measurement cell. When a gas in the cell is irradiated with electomagnetic energy at frequencies
that correspond to resonant vibration frequencies of VOC compounds in the gas, a portion of the incident energy
is absorbed, causing some of the molecules of the gas to be excited to a higher vibrational energy state. These
molecules subsequently relax back to the lower-energy, vibrational state through a combination of radiative and
kinetic processes. The kinetic energy decay process results in increased heat energy of the gas molecules and a
corresponding temperature and pressure increase in the gas. The incident infrared source is modulated and the
resulting pressure is also modulated. The varying pressure in the cell produces an acoustic wave that is detected
with a high-sensitivity microphone. Compound specificity is achieved by using bandpass filters tuned to the
energy absorption bands of target compounds, and quantification is done by measuring the intensity of the
resulting acoustic signal.
The Type 1312 is a commercially available measurement system that provides groundwater analysis capabilities
in a field-portable package. The instrument weighs 30 pounds with accessories and is encapsulated in a weather-
resistant case. Required accessories include a motorized stir plate, a 2-L flask, and assorted connecting tubing.
The system can be easily transported and operated in the rear compartment of a minivan. Instrument detection
limits for TCE and PCE in water are in the vicinity of 5 u.g/L. Sample composition must be known since the
measurement technique is susceptible to interference from unknown VOCs in the sample. Sample processing and
analysis can be accomplished by a technician; however, method development and periodic instrument calibration
require a higher level of operator experience and training. About 1 day of training is recommended for a
technician to be able to perform routine sample processing. At the time of the demonstration, the baseline cost of
the Type 1312 was $28,000. Maintenance costs are less than $100 per year. And with the exception of a
disposable inlet air filter, the instrument uses no consumable items, such as carrier gases or calibration standards.
VERIFICATION OF PERFORMANCE
The following performance characteristics of the Type 1312 were observed:
Sample Throughput: Throughput was approximately one to two water samples per hour.
Completeness: The Type 1312 reported results for all but one of the 141 PE and groundwater samples provided
for analysis at the two demonstration sites. One PE sample was dropped by the Innova team during preparation
and handling.
Analytical Versatility: The Type 1312 was calibrated for and reported results for TCE and PCE. The Type
1312 reported results for 29 of 31 detects of TCE and PCE in groundwater samples from both sites that were
reported by the reference laboratory. The instrument also reported results for carbon tetrachloride, chloroform,
and cw-l,2-dichloroethene under appropriate circumstances. The instrument can report results for up to five
compounds from a single analysis; however, sample composition must be known to account for possible spectral
interferences from all sample components.
EPA-VS-SCM-28
The accompanying notice is an integral part of this verification statement
iv
November 1998
-------
Precision: Instrument precision was determined by analysis of sets of four replicate samples from a variety of FE
mixtures containing known concentrations of TCE and PCE. The range of relative standard deviations (RSDs)
for TCE was 4 to 22%, and 5 to 46% for PCE. The distribution of RSD values for combined TCE and PCE
measurements from both sites had a median value of 15% and a 95th percentile value of 34%. By comparison, the
compiled RSDs for TCE and PCE from the reference laboratory had a median value of 9% and a 95th percentile
value of 18%.
Accuracy: Instrument accuracy was evaluated by comparing Type 1312 results with the known concentrations of
TCE and PCE in PE mixtures. The range of absolute percent differences (APD) for TCE was 4 to 48%, and 2 to
48% for PCE. The distribution of APD values for combined TCE and PCE measurements at both sites had a
median value of 29% and a 95th percentile value of 47%. By comparison, the compiled APDs for TCE and PCE
from the reference laboratory had a median value of 10% and>a 95th percentile value of 25%.
Comparability: A comparison of Type 1312 and reference laboratory data was based upon 33 groundwater
samples analyzed at each site. The correlation coefficient (r) for TCE and PCE detected by both the Type 1312
and the reference laboratory below the 300 p.g/L concentration level was 0.984 at Savannah River and 0.892 at
McClellan. The number of data pairs above the 300 |ig/L concentration level was insufficient for a meaningful
correlation analysis. The observed correlation coefficients reveal a linear relationship between the Type 1312 and
laboratory data at both sites. The median absolute percent difference between mutually detected TCE and PCE by
the Type 1312 and the reference laboratory was 29% with a 95th percentile value in excess of 2000%.
Deployment: The system was ready to analyze samples within 30 minutes of arrival at the site. At both sites, the
instrument was transported in a rental vehicle and was powered by line or generator ac power. During this
demonstration, the system was set up and operated on a table. It can also be set up and operated in the rear
luggage compartment of a minivan or station wagon.
The results of the demonstration show that the Innova AirTech Instruments Type 1312 Multi-gas Monitor can
provide useful, cost-effective data for routine groundwater monitoring when the composition of the samples is
known. Since the composition of the sample must be known to avoid spectral interference, the instrument is not
well suited for site characterization applications where the VOC content of the samples is unknown. In the
selection of a technology for deployment at a site, the user must determine what is appropriate through
consideration of instrument performance and the project's data quality objectives.
GaryJ. FoleyfPh. D.
Director
National Exposure Research Laboratory
Office of Research and Development
Samuel G. Varnado
Director
Energy and Critical Infrastructure Center
Sandia National Laboratories
NOTICE: EPA verifications are based on an evaluation of technology performance under specific, predetermined
criteria and the appropriate quality assurance procedures. EPA makes no expressed or implied warranties as to the
performance of the technology and does not certify that a technology will always, under circumstances other than those
tested, operate at the levels verified. The end user is solely responsible for complying with any and all applicable
federal, state and local requirements.
EPA-VS-SCM-28
The accompanying notice is an integral part of this verification statement
V
November 1998
-------
Foreword
The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the nation's natural
resources. The National Exposure Research Laboratory (NERL) is the EPA center for the investigation of
technical and management approaches for identifying and quantifying risks to human health and the
environment. The NERL research goals are to (1) develop and evaluate technologies for the characterization and
monitoring of air, soil, and water; (2) support regulatory and policy decisions; and (3) provide the science
support needed to ensure effective implementation of environmental regulations and strategies.
The EPA created the Environmental Technology Verification (ETV) Program to facilitate the deployment of
innovative technologies through verification of performance and dissemination of information. The goal of the
ETV Program is to further environmental protection by substantially accelerating the acceptance and use of
improved and cost-effective technologies. It is intended to assist and inform those involved in the design,
distribution, permitting, and purchase of environmental technologies.
Candidate technologies for this program originate from the private sector and must be market ready. Through
the ETV Program, developers are given the opportunity to conduct rigorous demonstrations of their technologies
under realistic field conditions. By completing the evaluation and distributing the results, the EPA establishes a
baseline for acceptance and use of these technologies.
GaryJ. Foley,Ph. D.
Director
National Exposure Research Laboratory
Office of Research and Development
VI
-------
Executive Summary
The U.S. Environmental Protection Agency, through the Environmental Technology Verification Program, is
working to accelerate the acceptance and use of innovative technologies that improve the way the United States
manages its environmental problems. As part of this program, the Consortium for Site Characterization Technol-
ogy was established as a pilot program to test and verify field monitoring and site characterization technologies;
The Consortium is a partnership involving the U.S. Environmental Protection Agency, the Department of Defense,
and the Department of Energy. In 1997 the Consortium conducted a demonstration of five systems designed for the
analysis of chlorinated volatile organic compounds in groundwater. The developers participating in this demon-
stration were Electronic Sensor Technology, Perkin-Elmer Photovac, and Sentex Systems, Inc. (portable gas chro-
matographs); Inficon, Inc. (portable gas chromatograph/mass spectrometer, GC/MS); and Innova AirTech Instru-
ments (photoacoustic infrared analyzer). This report documents demonstration activities, presents demonstration
data, and verifies the performance of the Innova Type 1312 Multi-gas Monitor. Reports documenting the perform-
ance of the other technologies have been published separately.
The demonstration was conducted at two geologically and climatologically different sites: the U.S. Department
of Energy's Savannah River Site, near Aiken, South Carolina, and McClellan Air Force Base, near Sacramento,
California. Both sites have groundwater resources that are significantly contaminated with a variety of chlorinated
volatile organic compounds. The demonstrations designed to evaluate the capabilities of each field-portable system
were conducted in September 1997 and were coordinated by Sandia National Laboratories.
The demonstration provided adequate analytical and operational data with which to evaluate the performance
of the 1312 Multi-gas Monitor (hereafter, the Type 1312 or 1312). Instrument precision and accuracy were
determined from analysis of replicate samples from 13 standard mixtures containing trichloroethene (TCE) and
tetrachloroethene, also known as perchloroethene (PCE). The relative standard deviations obtained from an
analysis of replicate samples from each of the 13 two-component mixtures were used as measures of precision.
The distribution of relative standard deviations for combined results from TCE and PCE had a median value of
15% and a 95th percentile value of 34%. Accuracy was expressed as the absolute percent difference between the
1312 measured value and the true value of the component in the standard mixtures. The distribution of absolute
percent difference values for TCE and PCE in all mixtures had a median value of 29% and a 95th percentile value
of 47%. A comparison of Type 1312 and reference laboratory TCE and PCE results from 33 groundwater samples
at each site resulted in a median absolute percent difference of 29% with a 95th percentile value in excess of
2000%. The Type 1312 reported results for 29 of 31 detects of either TCE or PCE reported by the reference
laboratory in all groundwater samples from both sites. A correlation analysis of 1312 and laboratory TCE and PCE
groundwater results revealed a relatively high degree of linear correlation for Savannah River data (r >0.98) and
McClellan data (r > 0.89).
The Type 1312 costs between $28,000 and $35,000, depending on options, and can be operated by a field
technician with minimal training. Method development and analysis of multicomponent samples may require a
higher level of operator training and knowledge in use of the instrument. The throughput rate was in the range of
one to two samples per hour. Analytical results are available at the conclusion of a sample run. The results of the
demonstration show that the 1312 can provide useful, cost-effective data for routine groundwater monitoring in
circumstances where the sample composition is known. As with any technology selection, the user must
determine what is appropriate for the application by taking into account instrument performance and the project's
data quality objectives.
Vll
-------
Vlll
-------
Contents
Notice jj
Verification Statement ijj
Foreword vj
Executive Summary vii
F'gures • • xiii
Tables xiv
Acronyms and Abbreviations xv
Acknowledgments h xvjj
Chapter 1 Introduction j
Site Characterization Technology Challenge 1
Technology Verification Process 2
Identification of Needs and Selection of Technology 2
Planning and Implementation of Demonstration 2
Preparation of Report 3
Distribution of Information 3
The Wellhead VOC Monitoring Demonstration 3
Chapter 2 Technology Description 5
Technology Overview 5
Principle of Operation.... 5
History of the Technology g
Advantages 7
Limitations g
Improvements g
Applications g
Performance Characteristics g
Method Detection Limits and Practical Quantitation Limit 8
IX
-------
Accuracy
o
Precision
Instrument Working Range 9
Comparison with Reference Laboratory Analyses 9
Data Completeness ™
Specificity 10
Other Field Performance Characteristics 10
Instrument Setup and Disassembly Time 10
Instrument Calibration Frequency During Field Use 1°
Ancillary Equipment and Field Maintenance Requirements 10
Sample Throughput Rate 10
Operator Training Requirements and Ease of Operation 10
Chapter 3 Demonstration Design and Description 11
Introduction
Overview of Demonstration Design H
Quantitative Factors H
Qualitative Factors 12
Site Selection and Description I3
Savannah River Site 13
McClellan Air Force Base 15
Sample Set Descriptions 17
PE Samples and Preparation Methods 20
Groundwater Samples and Collection Methods 22
Sample Handling and Distribution 23
Field Demonstration Schedule and Operations 23
Site Operations and Environmental Conditions 23
Field Audits 24
Data Collection and Analysis 25
Demonstration Plan Deviations 25
-------
Chapter 4 Laboratory Data Results and Evaluation 27
Introduction 27
Reference Laboratory ... ..>...... 27
Laboratory Selection Criteria 27
Summary of Analytical Work by DataChem Laboratories 28
Summary of Method 8260A 28
Method 8260A Quality Control Requirements 28
Summary of Laboratory QC Performance 28
Target Compound List and Method Detection Limits.... 29
Sample Holding Conditions and Times 29
System Calibration 29
Daily Instrument Performance Checks 31
Batch-Specific Instrument QC Checks 31
Sample-Specific QC Checks.. 31
Summary of Analytical and QC Deviations 33
Other Data Quality Indicators 33
PE Sample Precision ....34
PE Sample Accuracy 34
Groundwater Sample Precision 39
Summary of Reference Laboratory Data Quality 40
Chapters Demonstration Results 41
Type 1312 Calibrated and Reported Compounds 41
Preanalysis Sample Information 41
Sample Completion 41
Blank Sample Results 42
Performance at Instrument Detection Limit 42
PE Sample Precision 42
PE Sample Accuracy 44
Comparison with Laboratory Results 46
XI
-------
Sample Throughput 49
Performance Summary 49
Chapter 6 Field Observations and Cost Summary 51
Introduction 51
Method Summary 51
Equipment 51
Sample Preparation and Handling 51
Consumables 52
Historical Use 52
Equipment Cost 52
Operators and Training 53
Data Processing and Output 53
Compounds Detected 53
Initial and Daily Calibration 53
QC Procedures and Corrective Actions 54
Sample Throughput 54
Problems Observed During Audit 54
Data Availability and Changes 54
Applications Assessment 54
Chapter 7 Technology Update 55
Chapter 8 Previous Deployments 56
References 57
xn
-------
Figures
2-1.
3-1.
3-2.
3-3.
3-4.
4-1.
4-2.
4-3.
4-4.
4-5.
4-6.
4-7.
4-8.
4-9.
4-10.
5-1.
5-2.
5-3.
5-4.
5-5.
5-6.
6-1.
A diagram of the Innova Type 1312 Multi-gas Monitor configured for headspace monitoring 6
The general location of the Savannah River Site in the southeast United States 13
A map of the A/M area at the Savannah River Site showing the subsurface TCE plume 14
A map of Sacramento and vicinity showing the location of McClellan Air Force Base 16
Subsurface TCE plumes at McClellan Air Force Base in the shallowest (A) aquifer layer 18
Laboratory control standard recovery values for SRS analyses 32
Laboratory control standard recovery values for MAFB analyses 32
Laboratory precision on SRS PE samples containing mix 1 35
Laboratory precision on SRS PE samples containing mix 2 35
Laboratory precision on MAFB PE samples containing mix 2 36
Laboratory precision on MAFB PE samples containing mix 3 36
Laboratory mean recoveries for SRS PE samples containing mix 1 37
Laboratory mean recoveries for SRS PE samples containing mix 2 37
Laboratory mean recoveries for MAFB PE samples containing mix 2 38
Laboratory mean recoveries for MAFB PE samples containing mix 3 38
Type 1312 PE sample precision at the SRS 43
Type 1312 PE sample precision at MAFB 43
Type 1312 PE sample recovery at the SRS 45
Type 1312 PE sample recovery at MAFB ; 45
Type 1312 groundwater results at the SRS relative to laboratory results 48
Type 1312 groundwater results at MAFB relative to laboratory results 48
The Innova Type 1312 Multi-gas Monitor. 52
xm
-------
Tables
2-1. MDL and PQL in Air and Water Samples for the Type 1302 Multi-gas Monitor 9
2-2. Working Range of the Innova Type 1312 Photoacoustic Monitor in Water.... 9
3-1. Quarterly Monitoring Results for SRS Wells Sampled in the Demonstration 15
3-2. Groundwater Contaminants at MAFB 19
3-3. Quarterly Monitoring Results for MAFB Wells Sampled in the Demonstration 19
3-4. Composition of PE Source Materials 21
3-5. PE Sample Composition and Count for SRS Demonstration 21
3-6. PE Sample Composition and Count for MAFB Demonstration 22
3-7. Weather Summary for the SRS and MAFB During Demonstration Periods.... 24
3-8. Innova PE Sample Composition and Count for SRS Demonstration 26
3-9. Innova PE Sample Composition and Count for MAFB Demonstration 26
4-1. Method 8260A Quality Control Summary 29
4-2. Reference Laboratory Method Detection Limits for Target Compounds ..30
4-3. Summary of Reference Laboratory Quality Control and Analytical Deviations 33
4-4. Sources of Uncertainty in PE Sample Preparation 34
4-5. Summary of SRS Groundwater Analysis Precision 39
4-6. Summary of MAFB Groundwater Analysis Precision 39
5-1. Type 1312 Calibrated and Reported Compounds 41
5-2. False Positive Rates from Blank Sample Analysis 42
5-3. False Negative Rates from Very Low-Level PE Sample Analysis 42
5-4. Precision for TCE and PCE at Both Site's 42
5-5. Summary of PE Sample Precision and Percent Difference Statistics for the SRS and MAFB 44
5-6. Target Compound Recovery at Both Sites 44
5-7. Type 1312 and Reference Laboratory Results for SRS Groundwater Samples 46
5-8. Type 1312 and Reference Laboratory Results for MAFB Groundwater Samples 47
5-9. Type 1312 Absolute Percent Difference Summary for Pooled Groundwater Results 49
5-10. Correlation Coefficients for Reference Laboratory and Type 1312 Groundwater Analyses 49
5-11. Type 1312 Performance Summary 50
6-1. Type 1312 Cost Summary 53
xiv
-------
Acronyms and Abbreviations
ac
APD
BNZN
°C
ccc
CCL4
CLFRM
cnT1
cm3
dc
11DCA
12DCA
DCE
11DCE
C12DCE
t!2DCE
DCL
DOE
EPA
ETV
OF
FTIR
eV
GC
GW
GC/MS
L
m
mg
mg/L
mL
mm
MAFB
MCL
MDL
MS
NERL
alternating current
absolute percent difference
benzene
degrees centigrade
calibration check compounds
carbon tetrachloride
chloroform
one per centimeter (wave number)
cubic centimeters
direct current
1,1 -dichloroethane
1,2-dichloroethane
dichloroethene
1,1-dichloroethene
cis-1,2-dichloroethene
trans-1,2-dichloroethene
DataChem Laboratories
Department of Energy
Environmental Protection Agency
Environmental Technology Verification Program
degrees Farenheit
Fourier-transform infrared
electron-volt
gas chromatograph
groundwater
gas chromatograph/mass spectrometer
liter
meter
milligram
milligrams per liter
milliliter
millimeter
McClellan Air Force Base
maximum concentration level
method detection limit
mass spectroscopy
National Exposure Research Laboratory
xv
-------
NR
PC
PCE
PE
ppb
ppbv
ppm
ppmv
PQL
PVC
QA
QC
r
RPD
RSD
SPCC
SRS
TCA
111TCA
TCE
V
V-A
Vac
Vdc
VOA
VOC
not reported
personal computer
tetrachloroethene(perchloroethene)
performance evaluation
parts per billion
parts per billion volume
parts per million
parts per million volume
practical quantitation limit
poly (vinyl chloride)
quality assurance
quality control
correlation coefficient
relative percent difference
relative standard deviation
system performance check compounds
Savannah River Site
trichloroethane
1,1,1 -trichloroethane
trichloroethene
volts
volts-ampere
volts alternating current
volts direct current
volatile organics analysis
volatile organic compound
microgram
micrograms per liter
microliter
xvi
-------
Acknowledgments
The author wishes to acknowledge the support of all those who helped to plan and conduct the demonstrations
analyze the data, and prepare this report. In particular, the technical expertise of Gary Brown, Robert Helgesen
Michael Hightower and Dr. Brian Rutherford of Sandia National Laboratories in the planning and conduct of the
study are recognized. The assistance of Dr. Timothy Jarosch and Joseph Rossabi of Westinghouse Savannah
River Co. in planning the demonstration and field activities at both Savannah River and McClellan is also
recognized The willingness of Phillip Mook and Timothy Chapman of the Environmental Directorate at
McClellan Air Force Base to host the McClellan phase of the study is also greatly appreciated. The availability
of funding from the Department of Defense's Strategic Environmental Research and Development Program
helped to make the McClellan phase of the study possible. The guidance and contributions of project technical
leaders Dr Stephen Billets and Eric Koglin of the EPA National Exposure Research Laboratory, Environmental
Sciences Division in Las Vegas, Nevada, during all phases of the project are also recognized.
The participation of personnel from Innova AirTech Instruments in this technology demonstration is also
acknowledged. Michael Vecht served as the primary point of contact at Innova. IngeLise Olsen and Charlotte
Kaarsberg operated the instruments during the demonstrations.
For more information on the wellhead monitoring demonstration, contact:
Stephen Billets, Project Technical Leader, Environmental Protection Agency
National Exposure Research Laboratory, Environmental Sciences Division
P.O. Box 93478, Las Vegas, Nevada, 89193-3478
(702) 798-2232
For more information on the Innova Type 1312 Multi-gas Monitor, contact:
Michael Vecht, Director of Marketing, Innova AirTech Instruments
Energivej 30, 2750 Ballerup
Denmark
International: 45 4420 01 00
Hal Peper, California Analytical Instruments Inc.
1238 West Grove Avenue, Orange, CA 92665
(714)974-5560
xvn
-------
XV111
-------
Chapter 1
Introduction
Site Characterization Technology Challenge
The U.S. Environmental Protection Agency (EPA) created the Environmental Technology Verification (ETV)
Program to facilitate the deployment of innovative environmental technologies through verification of
performance and dissemination of information. The goal of the ETV Program is to further environmental
protection by substantially accelerating the acceptance and use of improved and cost-effective technologies. It is
intended to assist and inform those involved in the design, distribution, permitting, purchase, and use of
environmental technologies. The ETV Program capitalizes on and applies the lessons that were learned in the
implementation of the Superfund Innovative Technology Evaluation Program to twelve pilot programs: Drinking
Water Systems, Pollution Prevention for Waste Treatment, Pollution Prevention for Innovative Coatings and
Coatings Equipment, Indoor Air Products, Advanced Monitoring Systems, EvTEC (an independent, private-
sector approach), Wet Weather Flows Technologies, Pollution Prevention for Metal Finishing, Source Water
.Protection Technologies, Site Characterization and Monitoring Technology, Climate Change Technologies and
Air Pollution Control. '
For each pilot, the EPA utilizes the expertise of partner "verification organizations" to design efficient
procedures for performance tests of the technologies. The EPA selects its partners from both public and private
sectors, including federal laboratories, states, and private sector entities. Verification organizations oversee and
report activities based on testing and quality assurance protocols developed with input from all major stakeholder
and customer groups associated with the technology area. The U.S. Department of Energy's (DOE's) Sandia
National Laboratories in Albuquerque, New Mexico, served as the verification organization for the
demonstration described in this report.
The performance verification reported here is based on data collected during a demonstration of technologies for
the characterization and monitoring of chlorinated volatile organic compounds (VOCs) in groundwater. Rapid,
reliable, and cost-effective field screening and analysis technologies are needed to assist in the complex task of
characterizing and monitoring hazardous and chemical waste sites. Environmental regulators and site managers
are often reluctant to use new technologies that have not been validated in an objective EPA-sanctioned testing
program or other similar process. Until the field performance of a technology can be verified through objective
evaluations, users will remain skeptical of innovative technologies, despite the promise of better, less expensive,
and faster environmental analyses. This demonstration was administered by the Site Characterization and
Monitoring Technology Pilot Program, which is also known as the Consortium for Site Characterization
Technology. The mission of the Consortium is to identify, demonstrate, and verify the performance of innovative
site characterization and monitoring technologies. The Consortium also disseminates information about
technology performance to developers, environmental remediation site managers, consulting engineers, and
regulators.
-------
Technology Verification Process
The technology verification process consists of the four key steps shown here and discussed in more detail in the
following paragraphs:
1. identification of needs and selection of technology;
2. planning and implementation of demonstration;
3. preparation of report; and
4. distribution of information.
Identification of Needs and Selection of Technology
The first aspect of the verification process is to determine the technology needs of the EPA and the regulated
community. The EPA, the U.S. Department of Energy, the U.S. Department of Defense, industry, and state
agencies are asked to identify technology needs for site characterization and monitoring. Once a need is
recognized, a search is conducted to identify suitable technologies that will address this need. This search and
identification process consists of reviewing responses to Commerce Business Daily announcements, searching
industry and trade publications, attending related conferences, and following up on suggestions from technology
developers and experts in the field. Candidate characterization and monitoring technologies are evaluated against
the following criteria:
• may be used in the field or in a mobile laboratory;
• has a regulatory application;
• is applicable to a variety of environmentally affected sites;
• has a high potential for resolving problems for which current methods are unsatisfactory;
• has costs that are competitive with current methods;
• has performance as good or better than current methods in areas such as data quality, sample preparation,
and/or analytical turnaround time;
• uses techniques that are easier and safer than current methods; and
• is a commercially available, field-ready technology.
Planning and Implementation of Demonstration
After a technology has been selected, the EPA, the verification organization, and the developer(s) agree on a
strategy for conducting the demonstration and evaluating the technology. A conceptual plan for designing a
demonstration for a site characterization technology has been published by the Site Characterization and
Monitoring Technology Pilot Program (EPA, 1996a). During the planning process, the following steps are
carried out:
• identification of at least two demonstration sites that will provide the appropriate physical or chemical
attributes in the desired environmental media;
• identification and definition of the roles of demonstration participants, observers, and reviewers;
• determination of logistical and support requirements (for example, field equipment, power and water
sources, mobile laboratory, communications network);
• arranging for field sampling and reference analytical laboratory support; and
-------
• preparation and implementation of a demonstration plan that addresses the experimental design, sampling
design, quality assurance and quality control (QA/QC), health and safety considerations, scheduling of field
and laboratory operations, data analysis procedures, and reporting requirements.
Preparation of Report
Each of the innovative technologies is evaluated independently and, when possible, against a reference
technology. The technologies are operated in the field by the developers in the presence of independent
observers who are provided by the EPA or the verification organization. Demonstration data are used to evaluate
the capabilities, limitations, and field applications of each technology. Following the demonstration, all raw and
reduced data used to evaluate each technology are compiled in a technology evaluation report, which is a record
of the demonstration. A data summary and detailed evaluation of each technology are published in an
environmental technology verification report. The report includes a verification statement, which is a concise
summary of the instrument's performance during the demonstration.
Distribution of Information
The goal of the information distribution strategy is to ensure that environmental technology verification reports
and accompanying verification statements are readily available to interested parties through traditional data
distribution pathways, such as printed documents. Related documents and updates are also available on the
World Wide Web through the ETV Web site (http://www.epa.gov/etv) and through a Web site supported by the
EPA Office of Solid Waste and Emergency Response Technology Innovation Office (http://clu-in.com).
Additional information at the ETV Web site includes a summary of the demonstration plan, test protocols (where
applicable), demonstration schedule and participants, and in some cases a brief narrative and pictorial summary
of the demonstrations.
The Wellhead VOC Monitoring Demonstration
In August 1996, the selection of a technology for monitoring chlorinated VOCs in water was initiated by
publication in the Commerce Business Daily of a solicitation and notice of intent to conduct such a technology
demonstration. Potential participants were also solicited through manufacturer and technical literature
references. The original demonstration scope was limited to market-ready in situ technologies; however, only a
limited response was obtained, so the demonstration scope was expanded to include technologies that could be
used to measure groundwater (GW) at or near the wellhead. The final selection of technologies was based on the
readiness of the technologies for field demonstration and their applicability to the measurement of chlorinated
VOCs in groundwater at environmentally affected sites.
For this demonstration, five instrument systems were selected. Three of them were field-portable gas
chromatographs with various detection systems: one with a surface acoustic wave detector from Electronic
Sensor Technology, one with dual electron capture and photoionization detectors from Perkin-Elmer Photovac,
and one with an argon ion/electron capture detector from Sentex Systems. The fourth instrument was a field-
portable gas chromatograph/mass spectrometer (GC/MS) from Inficon, and the fifth was a photoacoustic infrared
spectrometer from Innova AirTech Instruments. This report documents demonstration activities, presents
demonstration data, and verifies the performance of the photoacoustic infrared spectrometer from Innova
AirTech Instruments. Reports documenting the performance of the other four technologies have been published
separately.
The demonstration was conducted in September 1997 at the DOE Savannah River Site (SRS) near Aiken,
Georgia, and at McClellan Air Force Base (MAFB), near Sacramento, California. Both sites have subsurface
-------
plumes of chlorinated VOCs and extensive networks of groundwater monitoring wells. The demonstrations were
coordinated by Sandia National Laboratories with the assistance of personnel from the Savannah River Site.
The primary objective of this demonstration was to evaluate and verify the performance of field-portable
characterization and monitoring technologies for analysis of chlorinated VOCs in groundwater. Specific
demonstration objectives were to:
• verify instrument performance characteristics that can be directly quantified (such factors include response
to blank samples, measurement accuracy and precision, sample throughput, and data completeness);
• verify instrument characteristics and performance in various qualitative categories such as ease of operation,
required logistical support, operator training requirements, transportability, versatility, and other related
characteristics; and
• compare instrument performance with results from standard laboratory analytical techniques currently used
to analyze groundwater for chlorinated VOCs.
The goal of this and other ETV demonstrations is to verify the performance of each instrument as a separate
entity. Technologies are not compared with each other in this program. The demonstration results are
summarized for each technology independent of other participating technologies. In this demonstration, the
capabilities of the five instruments varied and in many cases were not directly comparable. Some of the
instruments are best suited for routine monitoring where compounds of concern are known and there is a
maximum contaminant concentration requirement for routine monitoring to determine regulatory compliance.
Other instruments are best suited for characterization or field-screening activities where groundwater samples of
unknown composition can be analyzed in the field to develop an improved understanding of the type of
contamination at a particular site. This field demonstration was designed so that both monitoring and
characterization technologies could be verified.
-------
Chapter 2
Technology Description
This chapter was provided by the developer and was edited for format and relevance. The data presented
include performance claims that may not have been verified as part of the demonstration. Chapters 5 and 6
report instrument features and performance observed in this demonstration. Publication of this material does
not represent EPA approval or endorsement.
Technology Overview
The Innova Type 1312 Multi-gas Monitor (hereafter referred to as the Type 1312 or the 1312) is a field-portable,
photoacoustic spectrophotometer designed for monitoring volatile organic compounds in the vapor phase. A
headspace sampling accessory extends the measurement capabilities of the instrument to water matrices as well.
The instrument's measurement technique utilizes a photoacoustic effect, which is based on the conversion of
light energy into sound energy by a gas, liquid, or soil. This effect was discovered and investigated by Alexander
Graham Bell in the late 1800s, but was little more than a curiosity until the 1970s, when interest in the technique
was renewed following the development of lasers and very sensitive detection techniques. Since that time,
instruments employing photoacoustic principles have been used to monitor a wide variety of chemicals in stack
and vent emissions, ambient air, and in the troposphere.
Principle of Operation
When a gas is irradiated with electromagnetic energy at a frequency that corresponds to a resonant vibration
frequency of the gas, some of the energy will be absorbed by the gas. The absorption causes some of the
molecules of the gas to be excited to a higher vibrational energy state. These molecules subsequently relax back
to the initial vibrational state through a combination of radiative and nonradiative processes. For vibrational
excitation, the primary relaxation process is a nonradiative, vibrational-to-translational energy transfer. An
increase in the translational energy of the gas molecules corresponds to a temperature and pressure increase in the
gas. The irradiating source is modulated, and the temperature and pressure response of the gas is also modulated.
The modulated pressure will result in an acoustic wave, which can be detected with a sound measuring device,
such as a microphone. The amplitude of the acoustic wave depends upon such factors as the geometry of the gas
cell, the incident light intensity, the absorbing gas concentration, the absorption coefficient, and the background
gas. For a nonresonant spherical gas cell under steady-state conditions, the amplitude of the acoustic wave can
be determined from the following equation:
P = K[(cp/Cv)-l]l0c(l/f)
In the above, P is the sound pressure—the measured parameter, 70 is the incident light intensity, c is the absorbing
gas concentration,/is the modulation frequency, Cp and Cv are heat capacities, and K is a cell- and gas-dependent
constant.
-------
The monitor, shown in Figure 2-1, is known as the Innova (formerly Briiel & Kjaer) Type 1312 Multi-gas
Monitor. The instrument uses a heated nichrome wire as its infrared radiation source. The light from the source
is focused by an ellipsoidal mirror, modulated with a mechanical chopper, and passed through an optical filter
before entering the photoacoustic gas cell. The acoustic signal is detected with a pair of condenser microphones.
The electrical signals from the microphones are amplified by preamplifiers mounted directly on the back side of
the microphones and added in a summation amplifier before being sent to an analog-to-digital converter for
further processing. The digitized signal is then converted to a concentration reading using a calibration factor
that is stored in the instrument.
The 1312 measures 6.9 x 15.6 x 11.8 inches and weighs 20 pounds. Its power consumption is 100 V-A and
power can be supplied by either a 110-V ac or a 12-V dc source. The cost of the system depends on the intended
use and expected sample matrix, but will fall in the range of $28,000 to $35,000.
The 1312 is a newer version of the Type 1302 that was previously tested at the Savannah River Site for the
measurement of chlorinated VOC compounds in air. The Type 1312 has a signal-to-noise ratio that is improved
over the 1302, resulting in lower detection limits. The 1312 can also work with lower sample volumes, making it
more suitable for headspace monitoring applications. During this demonstration, the instrument was connected
to a stirred recirculation flask, as shown in Figure 2-1, for improved response time.
To Inlet
Figure 2-1. A diagram of the Innova Type 1312 Multi-gas Monitor
configured for headspace monitoring.
History of the Technology
Both optical filter and Fourier-transform infrared (FTIR) photoacoustic instruments have been used and
evaluated for monitoring chlorinated volatile organic compounds in the air at soil remediation sites and from gas
wells and boreholes at contaminated sites. They have also been used to analyze chlorinated VOCs purged from
soil and water samples.
-------
One application of chlorinated VOC monitoring with the 1312 is the analysis of perchloroethene
(tetrachloroethene) (PCE) and trichloroethene (TCE) mixtures. The ability of the instrument to monitor the
compounds simultaneously has been evaluated in the laboratory and in the field at the Savannah River Site in
South Carolina. TCE and PCE are monitored using the 861 cm'1 and 900 cnf' spectral regions, respectively.
The optical bandpass of the filters used for these regions is approximately 60 cm'1. Since some interfering
absorbance from each compound is encountered in these spectral regions, a matrix method is used to determine
each concentration. In this method, the matrix equation KC = S is solved for C, where S is a column vector
containing the measured signals for each optical region, K is a matrix containing the response factors for each gas
in each of the different spectral regions, and C is a column vector containing the concentrations of each chemical.
Laboratory work has shown a linear photoacoustic sensor response for both PCE and TCE up to approximately
700 parts per million volume (ppmv) with a detection limit of approximately 0.07 ppmv for each. Above 700
ppmv, the response becomes nonlinear. Measurements on mixtures of TCE and PCE indicate that accurate
results can be obtained using the above matrix method.
The 1312 has previously been evaluated for field use in several different test scenarios at the Savannah River
Site. In one scenario, the instrument was used to monitor the concentration of PCE and TCE in the gas from a
horizontal extraction well. Measurement results were compared with the results from a gas chromatograph (HP
Model 5890). In general, the two instruments agreed to within 20% for both gases. The average relative percent
difference (RPD)1 for the 1312 compared with the GC was 10.0% (at concentrations less than 250 ppmv) and
11.4% (at concentrations greater than 250 ppmv) for PCE, with accuracy values of 9.8% and 7,4%, respectively,
for TCE. The precision (or stability) of the instrument was also evaluated over a 30-day interval by periodically
measuring 100 ppmv standards. The relative standard deviation (RSD)2 for five measurements over this time
period was 0.9% for PCE and 1.2% for TCE.
The instrument has also been used to monitor gas from vadose-zone wells, from off-gas treatment systems, and to
determine depth-discrete, soil-vapor concentrations of TCE and PCE with a cone penetrometer. A Fourier-
transform infrared photoacoustic instrument from Innova has also proven successful for laboratory-based purge-
and-trap analysis of PCE, TCE, carbon tetrachloride, and chloroform in water at Ames Laboratory, Iowa State
University. These investigations revealed detection limits for these compounds in the very low parts per billion
(ppb) range. Photoacoustic spectroscopy monitors from Innova have also been successfully used in indoor air
quality measurements (several units have been sold to the U.S. EPA), industrial hygiene applications,
fermentation emissions monitoring, and many more applications.
Advantages
Some of the advantages and characteristics of infrared photoacoustic spectroscopy as it pertains to monitoring
trace gases are as follows:
• High sensitivities can be obtained. Instruments using conventional infrared light sources, such as heated
nichrome wires, have demonstrated detection limits in the low parts per billion volume to parts per million
volume range for single gases.
The relative percent difference between two samples is the absolute value of their difference divided by their mean and
multiplied by 100.
The relative standard deviation is the sample standard deviation divided by the mean value and multiplied by 100.
-------
• Photoacoustic instruments are very stable, primarily as a result of the stability of the microphones.
Microphones are some of the most stable transducers known, with output drifts of <10% over hundreds of
years.
• A dynamic range of up to six orders of magnitude relative to the detection limit for a particular gas can be
achieved. Thus, very low as well as high contaminant concentrations can be measured with a single
instrument.
• Cell volume is very small (3 cm3), thus reducing the amount of sample and calibration gas needed. The
small cell volume also results in a compact instrument.
• Simple instrument and optical setups can be used, hi particular, multipass gas cells are not needed, thus
eliminating the problems of maintaining optical alignment through this type of cell.
• Photoacoustic systems measure the absorbance directly instead of indirectly as in transmission-type
instruments. Consequently, a good baseline stability results.
• No consumables are needed, keeping operational costs low.
• The instrument will function in temperatures ranging from 5 to 40 °C.
Limitations
The main limitation of the technology is the fact that most organic gases absorb energy over a wide range of the
infrared spectrum, making the measurement susceptible to interferences. The 1312 has a unique cross-
compensation algorithm that allows it to compensate for known interferences. Potential interferences include all
compounds that are active in the midinfrared region. In a sample matrix with unknown interferences,
measurement results could be erroneous. The technology cannot wholly replace laboratory testing, but it can
significantly reduce the amount of testing needed.
Improvements
Innova is investigating the use of various semipermeable silicone tubing configurations that will potentially
permit measurements to be made directly in the liquid sample, thus eliminating the need to do headspace
monitoring.
Applications
Photoacoustic spectroscopy technologies have been successfully used to measure various organic compounds in
air. By bringing the analyte of interest into the gas phase, the technology can also be used to monitor various
organic compounds in water, soil, and sludge.
As a result of high sensitivity combined with an extremely small sample cell, this technology is the only infrared-
based technology that can achieve the low detection limits needed for water and soil analysis applications. Since
the technology is extremely easy to use, even unskilled operators can achieve good results. The instrument is
best suited for routine monitoring applications where the composition of the contaminants in the water is known.
Performance Characteristics
Method Detection Limits and Practical Quantitation Limit
With this technology, almost any volatile chlorinated species that absorbs in the infrared spectrum can be
measured with a headspace sampling technique. In general, chlorinated species have strong infrared absorption
-------
and thus have low detection limits. In this demonstration, measurements were made for TCE, PCE, and
chloroform. Detection limits for chlorinated species in air are typically in the range of 50 to 100 ppbv. Detection
limits for the same compounds in water range from 1 to 10 ppb. Detection and quantitation limits for PCE and
chloroform have been determined for the Type 1302 monitor (a predecessor to the 1312) and are given in
Table 2-1. With the newer 1312 monitor, expectations are that the method detection limits (MDLs) and practical
quantitation limits (PQLs) will improve by a factor of three without affecting the upper limit of the working
range. The practical quantitation limits (defined as 3 x MDL) are also given in Table 2-1.
Table 2-1. MDL and PQL in Air and Water Samples for the Type 1302 Multi-
gas Monitor
Analyte
Tetrachloroethene
Chloroform
MDL Air (ppb)
70
70
MDL Water (ppb)
2
3
PQL Water (ppb)
7
10
Accuracy
Measurements have not been made on a certified standard with the headspace measurement technique. However,
as mentioned earlier, the results of the Type 1302 air sample measurements were within 20% of the gas
chromatography results.
Precision
Instrument precision for the Type 1312 is better than 20% RSD for concentrations ranging from the PQL to 0.5
ppm and better than 10% RSD over the upper end of the instrument's working range.
Instrument Working Range
For air samples, the working range for chloroform and tetrachloroethene, with a one-point calibration, is from the
PQL (0.2 ppmv) to 700 ppmv.
For water samples, a test of a 1-L volume of water in a 2.5-L flask gave the following results: 1 ppm chloroform
in water correlated to 35 ppm in the headspace, and 1 ppm tetrachloroethene in water correlated to 50 ppm in the
headspace. These results yielded a working range for chloroform and PCE as shown in Table 2-2, using a one-
point calibration of the instrument. With a two-point calibration, the range was extended as shown in Table 2-2.
Table 2-2. Working Range of the Innova Type 1312 Photoacoustic Monitor in Water
Analyte
Tetrachloroethene
Chloroform
Single-Point Calibration
7 ppb-15 ppm
10ppb-20ppm
Two-Point Calibration
7 ppb-150ppm
10ppb-200ppm
Comparison with Reference Laboratory Analyses
At the time of the demonstration, the developer had not performed this comparison.
-------
Data Completeness
Analysis and valid results will be reported for 95% or more of the samples presented for analysis during the
demonstration, provided that no unknown interferences are present.
Specificity
The Type 1312 has a built-in cross-compensation feature. The instrument can compensate for up to four known
interfering species and water vapor. If the instrument is not calibrated to perform compensation for an interfering
substance, the signals are added and the measurement result is dependent upon the degree of energy absorption of
the interfering substance at the wavelength range of the optical filter in use. Consider the following example:
• 10 ppm of chloroform and 100 ppm of xylene are present in the headspace.
• Without cross-compensation, 100 ppm of xylene are equivalent to 3.5 ppm of chloroform.
• With cross-compensation, the interference is reduced to a positive interference of 0.07 ppm.
Other Field Performance Characteristics
Instrument Setup and Disassembly Time
The setup time for the system is less than 15 minutes. The system can be air shipped in a container or,
alternatively, can be transported as a carry-on item.
Instrument Calibration Frequency During Field Use
The recommended instrument calibration interval is 3 months.
Ancillary Equipment and Field Maintenance Requirements
The system requires 110-V ac but can also be operated on 12-V dc through an external dc-to-ac inverter. A fine-
particle air inlet filter needs replacement once a month. No other consumables are required for routine field use
or maintenance.
Sample Throughput Rate
The expected throughput rate is three samples per hour.
Operator Training Requirements and Ease of Operation
Less than 1 hour of training is required to become proficient in operating the instrument.
10
-------
Chapter 3
Demonstration Design and Description
Introduction
This chapter summarizes the demonstration objectives and describes related field activities. The material is
condensed from the Demonstration Plan for Wellhead Monitoring Technology Demonstration (Sandia, 1997),
which was reviewed and approved by all participants prior to the field demonstration.
Overview of Demonstration Design
The primary objective was to test and verify the performance of field-portable characterization and monitoring
technologies for the analysis of chlorinated VOCs in groundwater. Specific demonstration objectives are listed
below:
• verify instrument performance characteristics that can be directly quantified; such factors include response
to blank samples, measurement accuracy and precision, data completeness, sample throughput, etc.;
• verify instrument characteristics and performance in various qualitative categories such as ease of operation,
required logistical support, operator training requirements, transportability, versatility, and other
considerations; and
• compare instrument results with data from standard laboratory analytical methods currently used to analyze
groundwater for chlorinated VOCs.
The experimental design included a consideration of both quantitative and qualitative performance factors for
each participating technology.
Quantitative Factors •
The primary quantitative performance factors that were verified included such instrument parameters as precision
and accuracy, blank sample response, instrument performance at sample concentrations near its' limit of detection,
sample throughput, and comparability with reference methods. An overview of the procedures used to determine'
quantitative evaluation factors is given below.
Precision
Measurement uncertainty was assessed over the instrument's working range by the use of blind replicate samples
from a number of performance evaluation (PE) mixtures. Eight PE mixtures containing chlorinated VOCs at
concentrations ranging from 50 jag/L to over 1000 ng/L were prepared and distributed at each site. The mixtures
were prepared from certified standard mixes with accompanying documentation giving mixture content and
purity. The relative standard deviation was computed for each compound contained in each set of replicate PE
samples and was used as a measure of instrument precision.
11
-------
Accuracy
Instrument accuracy was also evaluated by using results from the PE samples. A mean recovery was computed
for each reported compound in each PE mixture. The average instrument result for each compound, based on
four blind replicate sample analyses, was compared against the known concentration in the PE mixture and
reported as the average percent recovery and the absolute percent difference.
Blank Sample Response
At least two blank groundwater samples were analyzed with each instrument system per demonstration day.
These were distributed as blind samples in the daily set of samples provided to each instrument operator. The
results from these samples were used to assess the degree to which instrument contamination and sample-to-
sample carryover resulted in a false positive.
Low-Level Sample Response
The scope of this demonstration did not include an exhaustive determination of instrument detection limits.
However, 10 replicate spiked samples at concentrations near typical regulatory action limits were provided for
analysis at each site to validate the instrument performance at these low concentration levels. The results from
these analyses were compiled as detects and nondetects and were used to calculate the percentage of correct
determinations and false negatives.
Sample Throughput
Sample throughput takes into account all aspects of sample processing, including sample preparation, instrument
calibration, sample analysis, and data reduction. The multiday demonstration design permitted the determination
of sample throughput rates over an extended period. Thus the throughput rates are representative of those likely
to be observed in routine field use of the instrument.
Laboratory-Field Comparability
The degree to which the field measurements agree with reference laboratory measurements is a useful parameter
in instrument evaluation. In this demonstration, comparisons were made on groundwater samples by computing
the absolute percent difference between laboratory and field technology results for all groundwater contaminants
detected. Linear regression of the two data sets was also carried out to determine the strength of the linear
correlation between the two data sets.
Qualitative Factors
Key qualitative instrument performance factors observed during the demonstration were instrument portability,
logistical support requirements, operator training requirements, and ease of operation. Logistical requirements
include the technology's power requirements, setup time, routine maintenance, and the need for other equipment
or supplies, such as a computers, reagent solutions, or gas mixtures. Qualitative factors were assessed during the
demonstration by review of vendor information and on-site audits. Vendors provided information concerning
these factors during preparation of the demonstration plan. Vendor claims regarding these specifications and
requirements are included in Chapter 2. During the field demonstration phase, auditors from the verification
organization observed instrument operation and documented the degree of compliance with the instrument
specifications and methodology. Audit results are included in Chapter 6.
12
-------
Site Selection and Description
Two sites—the DOE Savannah River Site near Aiken, South Carolina, and McClellan Air Force Base near
Sacramento, California—were chosen for this demonstration. This section provides a brief history of each site, a
discussion of important geological features, and an outline of the nature and extent of contamination at each site.
The sites chosen met the following selection criteria:
• presence of chlorinated VOCs in groundwater;
• multiple wells at the site with a variety of contaminants and depths;
• documented well-sampling history with characterization and monitoring data;
• convenient access; and
• support facilities and services at the site.
Savannah River Site
The Savannah River Site is operated under contract by the Westinghouse Savannah River Company. The
complex covers 310 square miles in western South Carolina, adjacent to the Savannah River, as shown in
Figure 3-1. The SRS was constructed during the early 1950s to produce the basic materials used in the
fabrication of nuclear weapons, primarily tritium and plutonium-239. Production of weapons material at the SRS
also produced unusable byproducts such as intensely radioactive waste. In addition to these high-level wastes,
other wastes at the site include low-level solid and liquid radioactive wastes, transuranic waste, hazardous
chemical waste, and mixed waste.
Figure 3-1. The general location of the Savannah River Site in
the southeast United States.
Geological Characteristics
The SRS is located on the upper Atlantic Coastal Plain. The site is underlain by a thick wedge (approximately
1000 feet) of unconsolidated Tertiary and Cretaceous'sediments that overlie Precambrian and Paleozoic
metamorphic rocks and consolidated Triassic sediments (siltstone and sandstone). The younger sedimentary
section consists predominantly of sand and sandy clay. The depth to the water table from the surface ranges from
50 to 170 feet for the wells used in this demonstration.
13
-------
Groundwater and Monitoring Wells
The wells selected for sampling in this demonstration were in the A/M area, located in the northwest section of
the site. This area encompasses an abandoned process transfer line that, beginning in 1958, carried wastewater
for 27 years from M-area processing facilities to a settling basin. Site characterization data indicate that several
leaks occurred in the transfer line, which is buried about 20 feet below the surface, producing localized
contamination. Past industrial operations resulted in the release of chlorinated solvents, primarily trichloroethene
(TCE), tetrachloroethene (PCE), and 1,1,1-trichloroethane, to the subsurface.
The A/M area monitoring-well network, shown in Figure 3-2, consists of approximately 400 wells. The dark
squares in the figure indicate soil borings and the light squares indicate monitoring wells. The largest group of
wells, comprising approximately 70% of the total, are associated with the plume originating from the process
transfer lines and the settling basin. The majority of these wells are constructed of 4-inch poly(vinyl chloride)
(PVC) casing with wire-wrapped screens varying in length from 5 to 30 feet. The wells are screened either in the
water-table aquifer (M-area aquifer, well depths ranging from 30 to 170 feet), the underlying tertiary aquifer
(Lost Lake aquifer, well depths ranging from 170 feet to 205 feet), or a narrow permeable zone within the
confining unit above the cretaceous aquifer (Crouch Branch Middle Sand, well depths ranging from 215 to 260
feet). The wells are all completed with approximately 2.5 feet of standpipe above ground and a protective
housing. Most wells are equipped with a dedicated single-speed centrifugal pump (1/2 hp Grundfos Model
10S05-9) that can be operated with a control box and generator. Wellhead pump connections also contain a flow
meter and totalizer for monitoring pumped volumes.
All the wells are measured quarterly for water levels. On a semiannual basis, all point-of-compliance wells (41),
plume definition wells (236), and background wells (6) are sampled to assess compliance with groundwater
protection standards. Other water quality parameters such as conductivity, turbidity, temperature, and pH are
Light Gray " High TCE Concentrations
DufcGray - LowwTCEConctirtnrtlon*
Each Grid Sown - 1000 Ft«t
ThslOwelb used In th» demonstration were located In the pluma shown.
The demonstration setup area was located v«y near the center of the flgunj.
Figure 3-2. A map of the A/M area at the Savannah
River Site showing the subsurface TCE plume.
14
-------
also measured. As a part of the monitoring program, VOCs are measured using EPA Method 8260A at an off-
site contract laboratory. The most recent (winter of 1996) quarterly water analysis results for the 10 wells used in
this demonstration are shown in Table 3-1. Well cluster numbers shown in the table include a letter designation
(A through D) that indicates the relative screening depth and aquifer zone. The A wells are the deepest of a
cluster, while the D wells mark the shallowest.
Table 3-1. Quarterly Monitoring Results for SRS Wells Sampled in the Demonstration
Sample Description
Very low 1
Very low 2
Low1
Low 2
Midi
Mid 2
Very high 1
Very high 2
Very high 1
Very high 2
Well Number
MSB 33B
MSB 33C
MSB 18B
MSB 37B
MSB4D
MSB 64C
MSB 4B
MSB 70C
MSB 14A
MSB8C
Compound
Trichloroethene
Tetrachloroethene
Trichloroethene
Tetrachloroethene
Trichloroethene
Tetrachloroethene
1,1-Dichloroethene
Trichloroethene
Tetrachloroethene
Carbon tetrachloride
Trichloroethene
Tetrachloroethene
Trichloroethene
Tetrachloroethene ,
1,1-Dichloroethene
Trichloroethene
Tetrachloroethene
Trichloroethene
Tetrachloroethene
1,1-Dichloroethane
1 , 1 , 1 -Trichloroethane
Trichloroethene
Tetrachloroethene
Trichloroethene
Tetrachloroethene
Qtrly. Results3 (uq/U
10
5
5
12
12
12
3
28
2
2
219
; 178
51
337
13
830
43
1290
413
61
17
3240
2440
3620
2890
McClellan Air Force Base
McClellan Air Force Base is located 7 miles northeast of downtown Sacramento, California, as shown in
Figure 3-3. The installation consists of about 3000 acres bounded by the city of Sacramento on the west and
southwest, the city of Antelope on the north, the unincorporated areas of Rio Linda on the northwest, and North
Highlands on the east.
McClellan has been an active industrial facility since its dedication in 1936, when it was called the Sacramento
Air Depot. Operations have changed from maintenance of bombers during Wprld War II and the Korean War, to
maintenance of jet aircraft in the 1960s, and now include the maintenance and repair of communications
equipment and electronics. McClellan currently operates as an installation of the Air Force Materiel Command
and employs approximately 13,400 military and civilian personnel.
15
-------
N
Figure 3-3. A map of Sacramento and vicinity showing the
location of McClellan Air Force Base.
Currently, most of the industrial facilities are located in the southeastern portion of the base. The southwestern
portion has both industrial and storage areas. In the far western part are vernal pools and wetland areas.
Between these wetlands and the engine test cells along the taxiways is an open area that was used for disposal
pits.
McClellan Air Force Base is listed on the EPA Superfund National Priorities List of hazardous waste sites. The
most important environmental problem at MAFB is groundwater contamination caused by the disposal of
hazardous wastes, such as solvents and oils, into unlined pits. Approximately 990 acres beneath McClellan are
contaminated with volatile organic compounds. Remediation activities at MAFB include an extensive
groundwater pump-and-treat network, as well as soil-vapor extraction systems.
McClellan has been designated a Chlorinated Hydrocarbons Remedial Demonstration Site as part of the National
Environmental Technology Test Sites program. The Strategic Environmental Research and Development
Program is the parent organization that provides support staff for the environmental technologies undergoing
development and testing at MAFB.
16
-------
Geological Characteristics
Surface features at MAFB include open grassland, creeks and drainages, and vernal pools, as well as industrial,
residential, and runway areas. The land surface is a relatively flat plain that slopes gently to the west. Surface
elevations range from about 75 feet above mean sea level on the eastern side of the base to about 50 feet above
mean sea level on the western side.
Surface soils at MAFB are variable, but are generally sediments that have formed from stream erosion of granite
rocks in the Sierra Nevada. Soil in the vadose zone — the unsaturated region between the surface and the
groundwater table — is composed of interbedded layers of sands, silts, and clays. The vadose zone ranges from 90
to 105 feet. Clays and hardpan layers in this zone slow,'but do not halt, infiltration of liquids into the underlying
aquifer.
The groundwater beneath MAFB behaves as one hydrogeologic unit. This single aquifer has been divided into
five groundwater monitoring zones, designated A, B, C, D, and E, from shallowest to deepest.
Groundwater and Monitoring Wells
An estimated 14 billion gallons of contaminated water underlie MAFB. Trichloroethene is the most frequently
detected contaminant in the subsurface groundwater. Over 90% of the contaminant mass is located in the A
zone, the shallowest portion of the aquifer. An estimated surface area of approximately 664 acres is underlain by
a plume in the A zone that exceeds the 5-ug/L maximum contaminant level for TCE, as shown in Figure 3-4.
Groundwater contaminants consistently detected above federal maximum concentration limits (MCLs) are shown
in Table 3-2.
Other detected compounds that are either below regulatory levels or are not currently regulated are also shown in
the table.
Monitoring wells at McClellan range from 2 to 8 inches in diameter. Well casings are Schedule 5 stainless steel
(304) and the well screen is Johnson stainless steel (304) with a 0.01- or 0.02-inch screen slot size. The screen is
surrounded by either 16 x 40 or 8 x 20 mesh gravel pack to a level about 3 feet above the screen. An
approximately 3-foot sand bridge and 3-foot bentonite seal are placed above the gravel pack. A concrete sanitary
seal containing about 3% bentonite powder is used to seal the well casing between the bentonite seal and the
ground surface.
For this demonstration, monitoring wells that penetrate both A and B aquifer zones in operational units A and B
were selected for sample collection. Quarterly monitoring data exist for 354 wells at the A and B zone aquifer
levels in these operational units. Monitoring results for TCE were used to select ten wells. Groundwater TCE
concentrations in the selected wells ranged from very low (-10 ng/L) to very high (>5000 fAg/L) levels.
Wells that had multiple contaminants or nonchlorinated contaminants were given selection preference over those
with only a few chlorinated hydrocarbons. The most recent (winter of 1996) monitoring results for the wells
chosen for this demonstration are shown in Table 3-3.
Sample Set Descriptions
The experimental design of the demonstration specified the preparation and collection of an approximately equal
number of PE samples and groundwater samples for distribution to the participants and reference laboratory.
Descriptions of the PE and groundwater samples and their preparation are given below.
17
-------
N
0 1000
I 1 1
Scale in Feet
Figure 3-4. Subsurface TCE plumes at McClellan Air Force Base in the
shallowest (A) aquifer layer. The circular lines enclose plume concentrations in
excess of 5 pg/L TCE. OU refers to operational units. Monitoring wells used in
the demonstration were primarily in OUs A and B. The demonstration setup
area was very near OU D (upper left in the figure).
18
-------
Table 3-2. Groundwater Contaminants at MAFB
Detected above MCLa
Benzene
Carbon tetrachloride
Chloroform
1 ,2-Dichlorobenzene
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
1,2-Dichloroethene (cis and trans)
Tetrachloroethene
1,1,1-Trichloroethane
Trichloroethene
Vinyl chloride
r)Afor»tarf hals\ur lUHfM
Bromodichloromethane
Trichlorofluoromethane
Detected - Not Requlated
Acetone
2-Butanone
1,1-Dichloroethane
4-Methyl-2-pentanone
Toluene
Table 3-3. Quarterly Monitoring Results for MAFB Wells Sampled in the Demonstration
sample Description
Very low 1
Very low 2
Low1
Low 2
Midi
Mid 2
Highl
High 2
Well Number
EW-86
MW-349
MW-331
MW-352
EW-87
MW-341
MW-209
MW-330
Compound
Trichloroethene
1,1 -Dichloroethene
Trichloroethene
Tetrachloroethene
Chloroform
Acetone
1,1-Dichloroethane
Carbon tetrachloride
Chloroform
Trichloroethene
c/s-1 ,2-Dichloroethene
1,1-Dichloroethane
Tetrachloroethene
Freonl 1
1,1,1 -Trichloroethane
1,1 -Dichloroethene
Trichloroethene
c/s-1 ,2-Dichloroethene
Trichloroethene
c/s-1 ,2-Dichloroethene
Chloroform
Trichloroethene
c/s-1 ,2-Dichloroethene
trans-1 ,2-Dichloroethene
Chloroform
Trichloroethene
c/s-1 ,2-Dichloroethene
frans-1 ,2-Dichloroethene
Qtrly. Results3 (ug/L)
8
13
9
5
8
9
16
5
7
19
41
6
5
115
17
334
220
5
350
18
53
586
80
13
44
437
64
9
19
-------
Table 3-3. Quarterly Monitoring Results for MAFB Wells Sampled in the Demonstration
(Continued)
Very high 1
Very high 2
Well Number
MW-334
MW-369
1,1-Dichloroethene
Benzene
Carbon tetrachloride
Chloroform
Dichloromethane
Trichloroethene
c/s-1 ,2-Dichloroethene
Xylene
1 ,2-Dichloroethane
Carbon tetrachloride
Chloroform
Tetrachloroethene
Trichloroethene
c/s-1 ,2-Dichloroethene
Qtrly. Results3 (ng/L)
1000
705
728
654
139
20,500
328
59
13
91
84
6
10,200
246
Winter 1996.
PE Samples and Preparation Methods .
Note- The description of PE samples in the following paragraphs pertains to the other four field technologies
that participated in the demonstration. The narrative is included here since the laboratory data validation was
based on analysis results from the PE sample set. Since the Innova 1312 is not capable of analyzing complex
mixtures, a separate set ofPE samples was prepared and distributed for analysis. A description of these Innova
PE samples is given in the "Deviations " section at the end of this chapter.
Three different commercially available (Supelco, Bellefonte, Pennsylvania) standard solutions of chlorinated
VOCs in methanol were used to prepare the PE mixtures. The standard solutions were supplied with quality
control documentation giving the purity and weight of the compounds in the mixture. The contents of the three
mixtures, termed mix 1, mix 2, and mix 3, are given in Table 3-4. VOC concentration levels in these standard
solutions were either 200 ^g/L or 2000 ng/L. The PE mixtures were prepared by dilution of these standard
solutions.
The number of replicate samples and the compound concentrations from each of the nine PE mixtures prepared at
each site are given in Table 3-5 for the SRS and Table 3-6 for MAFB. Ten replicates of the mixture with the
lowest concentration level were prepared so technology performance statistics near typical regulatory action
levels could be determined. Four replicates were prepared for each technology and the reference laboratory from
the other eight PE mixtures. The highest-level PE mixture, denoted "spike/low" in the tables, consisted of high-
level (>1000 ng/L) concentrations of TCE and PCE (and other compounds at MAFB as noted in the table) in the
presence of a low-level (50 or 100 ugfl,) PE mixture background. Eight blank samples were also provided to
each technology at each site. The blank samples were prepared from the same batch of deiomzed, carbon-filtered
water used to prepare the PE mixtures.
Performance evaluation mixtures were prepared in either 8-L or 10-L glass carboys equipped with bottom
spigots Stock PE solutions were dispensed with microsyringes into a known volume of deiomzed, carbon-
fntered water in the carboy. The mixture was gently stirred for 5 minutes with a Teflon-coated stir bar prior to
20
-------
Table 3-4. Composition of PE Source Materials
1,1-Dichloroethane
Dichloromethane
1,1-Dichloroethene
Chloroform
PE Mix 1 - Purgeable A
SupelcoCat. No. 4-8059
Lot LA68271
Trichloroflubromethane
Carbon tetrachloride
Trichloroethene
1,2-Dichloropropane
1,1,2-Trichloroethane
Tetrachloroethene
Dibromochloromethane
Chlorobenzene
1,2-Dichlorobenzene
2-Chloroethyl vinyl ether
PE Mix 2 - VOC 3
Supelco Cat. No. 4-8779
Lot LA64701
1,1-Dichloropropene
1,2-Dichloroethane
Trichloroethene
1,2-Dichloroproparie
1,1,2-Trichloroethane
1,3-Dichloropropane
1,2-Dibromoethane
.1,1,1,2-Tetrachloroethane
1,1,2,2-Tetrachloroethane"
1,2,3-Trichloropropane
1,2-Dibromo-3-chloropropane
c/s-1,3-Dichloropropene~~
frans-1,3-Dichloropropene
Hexachlorobutadiene
PE Mix 3 - Purgeable B
Supelco Cat No. 4-8058
Lot LA 63978
1,2-Dichloroethane
1,1,2,2-Tetrachloroethane"
c/s-1,3-Dichloropropene
trans-1,3-Dichloropropene
trans-1,2-Dichloroethene
1,1,1-Trichloroethane
Benzene
Bromodichloromethane
Toluene
Ethyl benzene
Bromoform
Table 3-5. PE Sample Composition and Count for SRS Demonstration
Sample Concentration Level
Very low level
Low level
Spike / low
Total number of samples
! Mixture - Mixture Concentration8
VOC Mix 1 -10ng/L
VOC Mix 1 - 50 ng/L
1.02 mg/L TCE spike + 50 ng/L mix 1
1.28 mg/L TCE and 1.23 mg/L PCE
spike +100 ng/L mix 2
' TCE = trichloroethene; PCE = tetrachloroethene.
10
4
42
d spensmg samples from the bottom of the carboy. A twofold excess volume of PE mixture was prepared in
s±n.l°HenSUre " TP 6 VOlUmu Wdl ln 6XCeSS °f the required Volume' The mixture ™ ^ stirred during
sample dispensmg to minimize headspace losses in the lower half of the carboy. Headspace losses that did occur
^^^1^
Samples were dispensed into bottles specified by participants (40 mL, 250 mL, and 1 L) with zero
were
21
-------
Table 3-6. PE Sample Composition and Count for MAFB Demonstration
Sample Concentration Level
Very low level
Low level
Mid level
High level
Spike / low
Total number of samples
PE Mixture - Mixture Concentration3
VOC Mix 3- 10ua/L
VOC Mix 3 - 50 ua/L
VOCMJX2- 100ua/L
VOC Mix 3 - 200 ug/L
vnn Mix 2 - 300 ua/L
VOC Mix 1 - 600 ug/L
VOCMix2- 800 ug/L
1.22 mg/LTCE, 1.00 mg/L PCE, 0.50 mg/L 11DCA,
and 0.50 mg/L BNZN spike + 100 ng/L mix 3
1.04 mg/L 11DCA, 0.86 mg/L BNZN, 0.57 mg/L
TCE, and 0.51 mg/L PCE spike + 50 ug/L mix 2
No. of Replicates
10
4
4
4
4
4
4
4
4
42
* TCE - trichtoroethene; PCE = tetrachloroethene; 11DCA
Reference laboratory samples were preserved by acidification as specified in Method 8260A. Following
preparation, all samples were kept under refrigeration until they were distributed to participants. All PE mixtures
were prepared and dispensed on the weekend before the demonstration week.
Groundwater Samples and Collection Methods
A total of 33 groundwater samples were provided to each participant and reference laboratory at each
demonstration site. These samples were collected from 10 wells selected to cover TCE concentrations ranging
from 10 ug/L to >1000 ug/L. The presence of other groundwater contaminants was also considered in well
selection, as noted previously. Samples from each well were prepared in either triplicate or quadruplicate to
allow statistical evaluation of instrument precision and accuracy relative to the reference laboratory results.
Groundwater at both sites was sampled by the same contract personnel who conduct sampling for quarterly well
monitoring. Site-specific standard operational procedures, published in the demonstration plan, were followed at
both sites. The sampling procedure is briefly summarized in the next paragraph.
The wells were purged with three well volumes using a submersible pump. During the purge, pH, temperature,
and conductivity were monitored. Following well purge, pump flow was reduced and the purge line was used to
fill a 10-L glass carboy. This initial carboy volume of groundwater was discarded. The carboy was tilled to
between 9 and 10 L a second time at a fill rate of 2 to 3 L/minute with the water stream directed down the side of
the carboy for minimal agitation. The filled carboy was gently mixed with a Teflon stir bar for 5 minutes. Zero-
headspace samples were immediately dispensed from the carboy while it was at the wellhead in the same manner
as PE samples. Either three or four replicate samples were prepared for each technology and the reference
laboratory. Following dispensing, the sample bottles were placed in a cooler and held under refrigeration until
they were distributed to the participants. Groundwater sampling was completed during the first 2 days of each
demonstration. Lists of the sampled wells and quarterly monitoring results are given in Tables 3-1 and 3-3 tor
the SRS and MAFB, respectively.
22
-------
Sample Handling and Distribution
The distribution and status of all samples were tracked with chain-of-custody forms. Samples were dispensed to
participants in small coolers containing a supply of blue ice. Normally, two sets of either 10 or 11 samples were
distributed to participants each day during the 4 days of the demonstration, for a total of 83 samples, including
blanks, at each site.
Some of the participants required information concerning the content of the samples prior to carrying out an
analysis. This information was noted on the chain-of-custody form for each PE and groundwater sample, and
was made available to the participants. Recorded information included:
• number of contaminants in the sample; ; .
• list of contaminants in the sample;
• boiling point range of sample constituents; and
• approximate concentration range of contaminants in sample (low, mid, high). .
The type of information provided during this demonstration would be required by the technology as a part of its
normal operational procedure and did not compromise the results of the test. The information provided to each
of the participants is documented in Chapter 5.
Field Demonstration Schedule and Operations
The following schedule was followed at both sites. The field team arrived on the Thursday prior to the
demonstration week. Performance evaluation samples were prepared on Friday, Saturday, and Sunday.
Technology participants arrived at the site on Monday morning and immediately began instrument setup. The
first set of PE samples was normally distributed to all participants by midday Monday. The groundwater
sampling crew, consisting of at least two on-site contractors and at least one ETV field-team member, carried out
sampling of the 10 wells on Monday and Tuesday. The first groundwater samples were distributed on
Wednesday. Thursday was reserved as a visitor day during which local and regional regulatory personnel and
other potential instrument users were invited to hear presentations about instrument capabilities as well as to
view the instruments in operation. Sample analysis was also performed on Thursday. On Friday, the final day of
the demonstration, participants finished sample analysis, packed up, and departed by midafternoon.
Site Operations and Environmental Conditions
Instruments were deployed in parking lots or open fields adjacent to the well networks sampled during each
demonstration. All participants came to the site self-equipped with power and shelter. Some came with field-
portable generators and staged under tent canopies; others operated their instruments inside vehicles and used dc-
to-ac power inverters connected to the vehicle's battery. Tables were provided for those participants who
required a work space. Each team provided its own instrument operators. Specifics regarding instrument setup
and the qualifications, training, and experience of the instrument operators are given in Chapter 6.
The SRS demonstration took place on September 8 through 12, 1997, and the MAFB demonstration on
September 22 through 26, 1997! The verification organization team staged its operations out of a tent at the SRS
and out of a mobile laboratory at MAFB. The PE mixtures at the SRS were prepared at a nearby SRS laboratory
facility and in the mobile laboratory at MAFB. Refrigerators at on-site facilities of the groundwater sampling
contractors were used to store the samples at both sites prior to their distribution.
23
-------
Environmental conditions at both sites are summarized in Table 3-7. Conditions at SRS were generally hot and
humid. Sporadic rain showers were encountered on one of the test days, but did not impede demonstration
activities. Conditions at MAFB were initially hot and progressed to unseasonably hot. Moderately high winds
were also encountered during the last 2 days at MAFB.
Table 3-7. Weather Summary for SRS and MAFB During Demonstration Periods
Site/Parameters
SRS
Temperature range (°C)
Relative humidity range (%)
Mon
20-34
25-68
Tue
Wed
21-33
28-67
21-28
51-71
MAFB
Temperature range (°C)
Relative humidity range (%)
Wind speed range (knots)
17-33
17-72
0-7
18-36
25-47
3-6
18-37
15-59
1-6
Thu
Fri
18-30
40 - 70
19-33
26-70
24 - 35
17-67
4-13
24-35
31-83
2-11
Note: Ranges are given for the 7a.m. to 7 p.m. time interval.
Field Audits ^ .
Field auditors were used to observe and record specific features of technology operations. The demonstration
goal was to have at least two auditors observe each technology over the course of the two field demonstrations.
Audit results are documented in Chapter 6. The following checklist was used by the audit team as a guideline for
gathering information during the audit:
• description of equipment used;
• logistical considerations, including size and weight, shipping and power requirements, other required
accessories;
• historical uses and applications of the technology;
• estimated cost of the equipment and its field operation;
• number of operators required;
• required operator qualifications;
• description of data produced;
• compounds that the equipment can detect;
• approximate detection limits for each compound, if available;
• initial calibration criteria;
• calibration check criteria;
• corrective actions for unacceptable calibrations;
• specific QC procedures followed;
• QC samples used;
• corrective action for QC samples;
• sample throughput rate;
• time requirements for data analysis and interpretation;
• data output format and description;
24
-------
• specific problems or breakdowns occurring during the demonstration;
• possible sample matrix interference; and
• other auditor comments and observations.
Data Collection and Analysis
The analytical results were collected in hardcopy format at the end of each day. These results were used to
document sample completion and throughput. The participants also provided a compilation of their results on
computer disks at the conclusion of each demonstration week. No feedback on analytical results or performance
was given to the participants during the course of either demonstration week. Following the SRS demonstration,
and only after all results were submitted, was qualitative verbal feedback given to each participant concerning
their accuracy and precision on SRS PE sample results. This was reasonable since a well-defined monitoring
plan would use preliminary samples to determine control limits and to make system modifications or refinements
prior to advancing to the next phase of sampling and analysis. Three weeks following the MAFB demonstration,
copies of all submitted data were entered into spreadsheets by the verification organization and transmitted to
participants for final review. This gave each participant the opportunity to detect and change calculation or
transcription errors. If other more substantive changes were proposed, they were submitted to the verification
organization, along with documentation outlining the rationale for the change. Following this final data review
opportunity, no other data changes were permitted. The extent and nature of any changes are discussed in
Chapter 6. -
Demonstration Plan Deviations
The following deviations from the written demonstration plan were recorded during the field demonstration. The
impact of each deviation on the overall verification effort, if any, is also included.
• Five blank samples were submitted to the reference laboratory from the SRS demonstration instead of the
8 samples specified in the demonstration plan. The impact on the verification effort was minimal since a
total of 13 blanks (8% of the total field sample count) were analyzed by the reference laboratory. •
• During groundwater sampling of SRS well MSB 14A, two 250-mL sample bottles were not filled.
Omission of this sample resulted in a double replicate sample set instead of a triple replicate for Electronic
Sensor Technology and Sentex. The impact on the study was insignificant since this omission accounted for
only 1 sample out of a total groundwater sample count of 33.
• The demonstration plan specified that only two VOC mixtures would be used at each demonstration site. In
fact, three mixtures were used at the MAFB demonstration (Table 3-6) to add complexity to the sampling.
This change caused some minor confusion with one of the developers, who was not expecting this particular
set of compounds at MAFB. The most significant impact of this change was a loss of time for the affected
developer as a result of extended data review of the unanticipated mixture. The misunderstanding was
verbally clarified and no further problems were encountered. The results from the high-level VOC mix 1
were not used in the statistical analyses.
• A different set of PE samples was prepared for the Innova 1312. The data processing algorithm used in the
Innova instrument could not accommodate the complex mixtures contained in the PE mixtures distributed to
the other participants. The Innova instrument was limited to the analysis of 5 compounds while the PE
mixtures contained in excess of 10 chlorinated VOCs. To accommodate these special needs, 2-component
PE mixtures were prepared from stock TCE and PCE methanol solutions at the SRS demonstration. The
composition of these mixtures is given in Table 3-8 for SRS and Table 3-9 for MAFB.
25
-------
Table 3-8. Innova PE Sample Composition and Count for SRS Demonstration
Sample Cone. Range
Very low
Low
Mid
High
Very high
Mix SVL1
Mix SL1
MixSL2
Mix SM1
Mix SM2
Mix SH1
Mix SH2
Mix SVH
PE Mixture and Composition (wg/L)
10 (trichloroethene and tetrachloroethene)
50 (trichloroethene and tetrachloroethene)
100 (trichloroethene and tetrachloroethene)
onn (trichloroethene and tetrachloroethene)
256 trichloroethene
245 tetrachloroethene
400 (trichloroethene and tetrachloroethene)
693 trichloroethene
713 tetrachloroethene
1278 trichloroethene
1223 tetrachloroethene
Replicates
6
3
3
3
4
3
4
4
Table 3-9. Innova PE Sample Composition and Count for MAFB Demonstration
Very low
Low
Mid
High
Very high
PE Mixture and Composition (ug/L)
Mix MVL1
Mix ML1
MixML2
Mix MM1
Mix MM2
MixMHI
Mix MVH1
9 trichloroethene
1 1 tetrachloroethene
42 trichloroethene
47 tetrachloroethene
80 trichloroethene
89 tetrachloroethene
169 trichloroethene
188 tetrachloroethene
31 9 trichloroethene
355 tetrachloroethene
758 trichloroethene
845 tetrachloroethene
401 1,1-dichloroethene
392 benzene
1434 trichloroethene
1598 tetrachloroethene
761 1,1-dichloroethene
741 benzene
Replicates
5
4
4
4
4
4
4
Eight blank water samples were also included in the sample set at each site and were submitted blind. A
total of 38 and 37 PE and blank samples were prepared and submitted to the Innova team at the SRS and
MAFB sites, respectively.
The groundwater samples submitted to the Innova team for analysis were the same as those submitted to the
other demonstration participants.
26
-------
Chapter 4
Laboratory Data Results and Evaluation
Introduction
A reference laboratory was used to verify PE sample concentrations and to generate analytical results for all
groundwater samples using EPA Method 8260A. This chapter includes a brief description of the reference
laboratory and its data quality control program; the methodology and accompanying quality control procedures
employed during sample analysis; and laboratory results and associated measures of data quality for both
demonstration sites.
Reference Laboratory
DataChem Laboratories (DCL) in Salt Lake City, Utah, was chosen as the reference laboratory for both phases of
this demonstration. This is a full-service analytical laboratory with locations in Salt Lake City and Cincinnati
Ohio. It provides analytical services in support of environmental, radiological, mixed-waste, and industrial
hygiene programs. DataChem's qualifications include U.S. EPA Contract Laboratory Program participation in
both inorganic and organic analysis and American Industrial Hygiene Association accreditation, as well as U S
Army Environmental Center and U.S. Army Corps of Engineers (Missouri River Division) certification State-
specific certifications for environmental analytical services include Utah, California, Washington, New Jersey
New York, Florida, and others. . y>
Laboratory Selection Criteria
Selection criteria for the reference laboratory included the following: relevant laboratory analytical experience
adequacy of QC documentation, turnaround time for results, preselection audit results, and cost Early
discussions with DCL revealed that the laboratory conducts a high number of water analyses using Method
8260A. Prior to laboratory selection, a copy of the DataChem Quality Assurance Program Plan (DataChem
1997) was carefully reviewed. This document outlines the overall quality assurance program for the laboratory
and provides specific quality control measures for all the standard analytical methods used by the laboratory
Laboratory analysis and reporting time for sample analysis was 21 days, with a per-sample cost of $95.
In June 1997, Sandia sent several PE water samples to DCL for evaluation. Laboratory performance on these
samples was reviewed during an audit in June 1997. The laboratory detected all compounds contained in the PE
mixtures. Reported concentration levels for all compounds in the mixtures were within acceptable error margins
The audit also indicated that the laboratory conducted its operations in accordance with its QA plan The results
of this preliminary investigation justified the selection of DCL as the reference laboratory and provided ample
evidence of the laboratory's ability to correctly use Method 8260A for the analysis of demonstration samples
27
-------
Summary of Analytical Work by DataChem Laboratories
In addition to the preselection audit samples noted above, DCL also analyzed predemonstration groundwater
samples collected at SRS in August 1997. During the demonstration phase, DCL was sent split samples of all PE
and groundwater samples given to the demonstration participants from both the Savannsth River and McClellan
sites, A total of 90 and 91 samples from the SRS and MAFB demonstrations, respectively, were received and
analyzed by the laboratory. Over the course of 1 month, demonstration samples were run in 9 batches of
approximately 20 samples per batch. The results were provided in both hardcopy and electronic format. The
hard copy included all paperwork associated with the analysis, including the mass spectral information for each
compound detected and complete quality control documentation. The electronic copy was provided in
spreadsheet format and included only the computed result for each target compound in each sample.
Preselection evaluation of DCL established their competence in the use of Method 8260A. In light of these
findings and in an effort to expedite laboratory analysis of demonstration samples, an estimate of the
concentration levels of target compounds in both PE and groundwater samples was provided to the laboratory
with each batch of samples. With a knowledge of the approximate concentration range of the target compounds,
the analyst was able to dilute the sample appropriately, thereby eliminating the need to do multiple dilutions in
order to obtain a suitable result within the calibrated range of the instrument.
Summary of Method 8260A
Method 8260A, which is included in the EPA SW-846 compendium of methods, is used to measure volatile
organic compounds in a variety of solid waste matrices, including groundwater (EPA, 1996b). The method can
be used to quantify most volatile organic compounds with boiling points below 200 °C that are either insoluble or
only slightly soluble in water. The method employs a chromatography/mass spectrometric procedure with purge-
and-trap sample introduction. An inert gas is bubbled through a vessel containing the water sample. The volatile
organic compounds partition into the gas phase and are carried to a sorbent trap, where they are adsorbed.
Following the purge cycle, the sorbent trap is heated and the volatile compounds are swept into the GC column,
where they are separated according to their boiling points. The gas chromatograph is interfaced directly to a
mass spectrometer that bombards the compounds with electrons as they sequentially exit the GC column. The
resulting fragments, which possess charge and mass characteristics that are unique for each compound, are
detected by the spectrometer's mass detector. The signal from the mass detector is used to build a compound
mass spectrum that is used to identify the compound. The detector signal intensities for selected ions unique to
each target compound are used to quantify the amount of the compound in the sample.
Method 8260A Quality Control Requirements
Method 8260A specifies a number of quality control activities to be carried out in conjunction with routine
sample analysis. These activities are incorporated into DCL QA documentation and are summarized in Table 4-1
(DataChem, 1997). Corrective actions are specified in the event of failure to meet QC criteria; however, for the
sake of brevity they are not given in the table. In most cases the first corrective action is a calculation check.
Other corrective actions include system recalibration, sample rerun, batch rerun, or flag data.
Summary of Laboratory QC Performance
The following sections summarize the QC activities and results that accompanied the analysis of each sample
batch.
28
-------
Table 4-1. Method 8260A Quality Control Summary
Activity
Spectrometer tune check
System performance
check
System calibration check
Lab method blank
Field blank
Laboratory control
standard
Matrix spike
Matrix spike duplicate
Surrogate standards
Internal standards
Frequency
Bromofluorobenzene
standard every 12 hours
SPCCa sample every 12
hours
CCCB sample every 12
hours
One or more per batch
(approx. 20 samples)
One or more per batch
One or more per batch
One or more per batch
One or more per batch
Included in every sample
Included in every sample
Data Acceptance Criteria
Relative abundance; range of characteristic mass
fragments meets specifications.
Compound relative response factors must exceed
required minimums.
Response factor of CCC varies by no more than ±25%
from initial calibration.
Internal standard retention time within 30 seconds of
last check.
Internal standard area response within -50 to 100% of
last check.
< 3x Detection limit.
< 3x Detection limit.
Compound recovery within established limits.0
Spike recovery within established limits.c
Relative percent difference of check compounds <50%.
Recovery within established limits.
Recovery within established limits.c
SPCC = system performance check compounds.
CCC = calibration check compounds.
Target Compound List and Method Detection Limits
The method detection limits and practical quantitation limits for the 34 target compounds used in this
demonstration are given in Table 4-2. The PQL marks the lower end of the calibrated working range of the
instrument and indicates the point at which detection and reported results carry a 99% certainty. Detects reported
between the MDL and PQL carry less certainty and are flagged accordingly in the tabulated results.
Sample Holding Conditions and Times
Method 8260A specifies a maximum 14-day holding time for refrigerated water samples. All samples prepared
m the field were kept under refrigeration before and during shipment to the laboratory. Upon receipt at the
laboratory, they were held under refrigeration until analysis. All samples were analyzed within the 14-day time
period following their preparation or collection.
System Calibration
Method 8260A stipulates that a five-point calibration be carried out using standard solutions for all target
compounds across the working range of the instrument. Each mix of compounds is run five times at each of the
five points in the instrument range. For an acceptable calibration, precision from these multiple analyses as
29
-------
Table 4-2. Reference Laboratory Method Detection Limits for Target Compounds
Target Compound
Trichlorofluoromethane
1,1-Dichloroethane
Methylene chloride
1,1-Dichloroethene
Chloroform
Carbon tetrachloride
1,1-Dichloropropene
1 ,2-Dichloroethane
Trichloroethene
1 ,2-Dichloropropane
1,1,2-Trichloroethane
Tetrachloroethene
1 ,3-Dichloropropane
Dibromochloromethane
1,2-Dibromoethane
Chlorobenzene
1,1,1 ,2-Tetrachloroethane
1 ,1 ,2,2-Tetrachloroethane
1 ,2,3-Trichloropropane
1 ,2-Dibromo-3-chloropropane
Hexachlorobutadiene
c/s-1 ,3-Dichloropropene
frans-1 ,3-Dichloropropene
1 ,2-Dichlorobenzene
frans-1 ,2-Dichloroethene
1 ,1 ,1-Trichloroethane
Benzene
Bromodichloromethane
Toluene
Ethyl benzene
Bromoform
c/s-1 ,2-Dichloroethene
orfftc-Xylene
Method Detection Limit
(ua/U
0.15
0.08
0.10
0.08
0.07
0.10
0.10
0.04
0.14
0.04
0.09
0.10
0.06
0.08
0.09
0.06
0.05
0.07
0.50
0.62
0.10
0.17
0.08
0.17
0.17
0.26
0.12
0.11
0.15
0.14
0.10
0.14
0.11
2.9
Practical Quantitation
Limit ^ti/L)
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
5
Notes- Detection limits are given for an undiluted 5-mL sample volume. Detection limits are determined annually
using the method outlined in 40 CFR Part 1 36 Appendix B (seven replicates of deiomzed water spiked at
1 ua/L concentration level). Dilutions of the original sample raise the MDL and PQL values accordingly.
Surrogate standards used in the analyses were 1 ,2-dichloroethane-d4, toluene-da, and 4-bromofluorobenzene.
Internal standards were fluorobenzene, chiorobenzene-ds, and 1,4-dichlorobenzene-d4.
30
-------
given by the relative standard deviation, must be 30% or less. A minimum instrument response factor1 is also
fSr J? ™y !- d f?ua desi§nated subset of Compounds termed system performance check compounds
(bPCC). The five-point calibration curve from the most recent instrument calibration met the specified precision
criteria. The system performance check compound response factors also met method criteria.
Daily Instrument Performance Checks
Daily mass spectrometer tune checks as well as other system performance and calibration checks noted in
Table 4-1 were carried out for each of the nine sample batches and met Method 8260A on quality control criteria.
Batch-Specific Instrument QC Checks
Method Blanks
All method blank analyses met established criteria (Table 4-1), with one exception. Hexachlorobutadiene one of
the demonstration target compounds, was detected in two of the method blanks at levels in excess of 3 times the
MDL. This compound was a component in one of the standard mixes used in preparing the PE samples because
reference laboratory data for this compound were not used in the study. Only one of the participating
technologies was calibrated to detect this particular compound. Occasional detection of this compound as a
minor instrument contaminant does not adversely affect the analytical results for other target compounds.
Laboratory Control Standard
At least one laboratory control standard was run with each of the nine batches of samples. Recovery values for
each component in the mixture are given in Figure 4-1 for SRS analyses and Figure 4-2 for MAFB analyses
Recovery values were all within the laboratory-specific control criteria. '
Matrix Spike and Matrix Spike Duplicate
The compounds in the matrix spike were the same as those in the laboratory control standard. Computed matrix
spike and matrix spike duplicate recoveries were all within the recovery ranges noted in Table 4-1. The relative
percent differences (RPDs)2 calculated for the matrix spike and matrix spike duplicate samples also met the
a?d0SaCnrifw f 122^ ^° ValU6S fr°m ^^ Splke analyS6S W6re leSS than 10% for the SRS
and less than 13% for MAFB samples.
Sample-Specific QC Checks
Internal Standard
All samples met internal standard acceptance criteria except one. All three internal standards in sample SP3 1
failed to meet area response criteria and results from that sample were not included in the reference data set.
1 ™e™f°™e^tm 1S th& rati° °f instrument resP°nse for a P«ticular target compound to the instrument response for an
: between two samples is the absolute value of their difference divided by their mean and
31
-------
DCL Laboratory Control Standard Recoveries
Savannah River Data Set
120
Batch 1
Batch 2 Batch 3
Analysis Batch No.
Batch 4
Batch 5
Figure 4-1. Laboratory control standard recovery values for SRS analyses.
DCL Laboratory Control Standard Recoveries
McClellan Data Set
120
80
Batch 1
Batch 2 Batch 3 Batch 4 Batch 5
Analysis Batch No.
Figure 4-2. Laboratory control standard recovery values for MAFB analyses.
32
-------
Surrogate Standard
With the following exceptions, surrogate standard recoveries met the criteria established by the laboratory as
noted in Table 4-1. Six samples (SP12, SP16, SP26, SP29, SP33, and SP65) failed surrogate recovery criteria for
l,2-dichloroethane-d4 and passed recovery criteria for 4-bromofluorobenzene and toluene-dg. The actions taken
are noted in Table 4-3.
Summary of Analytical andQC Deviations
A summary of QC deviations as well as other analytical errors or omissions is given in Table 4-3 The actions
taken with regard to the affected data and the reference data set are also tabulated, along with a brief rationale.
Table 4-3. Summary of Reference Laboratory Quality Control and Analytical Deviations
Deviation or QC Criteria Failure
Required dilution not made on two samples (SP20 and
SP21 ). Some compounds were present above
instrument linear range.
Three field blanks were not sent to DCL from SRS
demonstration.
Calculation error in original DCL report. Dilution factors
applied incorrectly in two samples (SP55 and SP57).
Sample SP31 failed internal standard recovery limits.
The following samples failed one or more surrogate
standard recovery limits: SP12, SP16, SP26 SP29
SP33, and SP65.
Hexachlorobutadiene detected as a contaminant in
selected blanks and samples.
Chloroethyl vinyl ether was not detected in PE samples
known to contain this compound.
Three sample results (MG20, MG51, and MG59) are
from a second withdrawal from the original zero-
headspace sample vial.
Action
Data Included: Data values for affected samples fall in
the range of the other three replicate samples.
No Action: Five field blanks and 10 method blanks
were run, yielding an adequate data set
Data Corrected and Included: The correct dilution
factors were applied following a teleconference with the
DCL analyst.
Data Not Included.
Data Not Included: SP12; results clearly fall outside of
the range of other three replicate samples.
Data Included: All others; nearly all target compounds
fall within the range of concentration reported for the
other three replicate samples.
No Action: This compound was not a target compound
for any of the technologies. Its presence as a low-level
contaminant does not affect the results of other target
compounds.
No Action: The GC/MS was not calibrated for this
compound. None of the technologies included this
compound in their target compound lists
Data Included: The original volume withdrawn from the
vial was 0.05 mL, resulting in an insignificant headspace
volume and no expected impact on the composition of
the second sample.
Other Data Quality Indicators
The demonstration design incorporated nine PE mixtures of various target compounds at each site that were
prepared in the field and submitted in quadruplicate to each technology as well as to the laboratory. Laboratory
accuracy and precision checks on these samples were assessed. Precision on replicate analysis of groundwater
samples was also evaluated. The results of these assessments are summarized in the following sections
33
-------
PE Sample Precision
The relative standard deviation from quadruplicate laboratory analyses of each PE 'mixture prepared in the field
was computed for each target compound in the mixture. As noted in Chapter 3, care was taken to ensure the
preparation and distribution of homogeneous samples from each PE mixture. The RSD values represent an
overall estimate of precision that takes into account field handling, shipping, storage, and analysis of samples.
The precision data are shown in Figures 4-3 and 4-4 for SRS and Figures 4-5 and 4-6 for MAFB. (See Tables
3-5 and 3-6 for the composition and concentration level of each PE mixture.) The compiled RSDs for all PE
sample results had a median value of 7% and a 95th percentile value of 25%. In selected instances, precision in
excess of Method 8260A specifications (^30% RSD) is observed for tetrachloroethene, trichloroethene, c/s-1,3-
dichloropropene, 1,2,3-trichloropropane, and 1,1,2,2-tetrachloroethane. Precision well in excess of method
specifications is observed for l,2-dibromo-3-chloropropane, fran.s-l,3-dichloropropene, and 1,1-dichloropropene.
The implications of these results with respect to evaluation of the technology performance are discussed, when
applicable, in Chapters 5 or 7.
PE Sample Accuracy
An error propagation analysis was carried out to estimate the degree of uncertainty in the stated "true"
concentration level of the PE samples prepared in the field. The sources of uncertainty and their magnitude
encountered during PE sample preparation are listed in Table 4-4. These errors are combined using the
methodology described by Bevington (1969) to arrive at a combined uncertainty in the PE sample value of ±5%.
Thus, for a 100-u.g/L PE mix, the true value is known with 99% certainty to be within the range of 95 to
105 u.g/L.
Table 4-4. Sources of Uncertainty in PE Sample Preparation
Type of Uncertainty
Weight of component in PE mix
ampule.
Volume of methanol solvent used
to dilute neat compounds.
Volume of PE solution (from
ampule) used in final PE solution.
Volume of water diluent in final
PE solution.
Magnitude
0.5 mg in 1200 mg
0.2 ml in 600 ml_
±5% of microsyringe volume;
e.g., 25 uL for a 500-uL syringe
5 ml in 10 L
Source of Estimate
Gravimetric balance uncertainty included
in PE mix certification documents
Published tolerances for volumetric flasks
(Fisher Catalog)
Published tolerances in certificates
shipped with microsyringes
Published tolerances for volumetric flasks
(Fisher Catalog)
The laboratory results for PE samples are compared with the "true" value of the mixture to provide an additional
measure of laboratory performance. A mean recovery3 was computed for each PE compound in each of the four
sample splits analyzed from each mixture. The SRS recovery values are shown in Figures 4-7 and 4-8, and
MAFB recoveries are shown in Figures 4-9 and 4-10. Acceptable mean percent recovery values, specified in
Method 8260A, fall within the range of 70 to 130% with exceptions for a few compounds that pose analytical
difficulties. With the following exceptions, all PE compounds at all concentration ranges met the Method 8260A
recovery criteria. The exceptions are 1,2,3-trichloropropane, 1,1-dichloropropene, l,2-dibromo-3-chloropropane,
3 Recovery is the ratio of the mean concentration level from analysis of the four sample splits to the reference or "true"
concentration levels of the target compounds in each PE mix.
34
-------
Target Compound
1,2-Dichlorobenzene
Chlorobenzene
Dibromochloromethane
Tetrachloroethene
1,1,2-Trichloroethane
1,2-Dichloropropane
Trichloroethene
Carbon Tetrachloride
Chloroform
1,1-Dichloroethene
Methylene Chloride
1,1-Dichloroethane
Trichlorofluoromethane
DataChem PE Sample Precision
Site: Savannah River Mix 1
10
20 30
Relative Standard Deviation, %
40
50
Figure 4-3. Laboratory precision on SRS PE samples containing mix 1.
Trichloroethene was spiked into the spike/low samples.
Target Compound
Tetrachloroethene
trans-1,3-Dichloroprppene
cis-1,3-DichIoropropene
1,2-Dlbromo-3-Chloropropane
1,2,3-Trichloropropane
1,1,2,2-Tetrachloroethane
1,1,1,2-Tetrachloroethane
1,2-Dibromoethane
1,3-Dlchloropropane
1,1,2-Triohloroethane
1,2-Dichloropropane
Trichloroethene
1,2-Dlchloroethane
1,1-DIchloropropene
DataChem PE Sample Precision
Site: Savannah River Mix 2
10
20 30
Relative Standard Deviation,'
50
Figure 4-4. Laboratory precision on SRS PE samples containing mix 2.
Tetrachloroethene was spiked into the mix 2 samples. Trichloroethene and
tetrachloroethene were spiked into the spike/low samples.
35
-------
Target Compound
Benzene
trana-1,3-Dichloropropene
ds-1,3-DlcNoropropene
1,2-Dfbrofno-3-Chtoropropane
15.3-Trichloropropan«
1,1,2,2-Telrachtoroetnana
1,1,1,2-Tetractiloroethane
1,2-Dibromoethane
1,3-Dichloropropane
Telrachloroelhane
1,1,2-Trichtoroethana
1,2-DicMoropropane
Trichloroethene
1,2-Dtehtoroethane
1,1-Dichloropropene
1.1-Dfchloroethane
DataChem PE Sample Precision
Site: McClellan Mix 2
Relative Standard Deviation, %
Figure 4-5. Laboratory precision on MAFB PE samples containing mix
2. Trichloroethene, tetrachloroethene, 1,1-dichloroethane, and bemzene
were spiked into the spike/low samples.
Target Compound
Bromoform
Ethylbenzeno
Toluene
Bromodtehtoromethane
Benzene
1,1,1-Trichtoroethane
lrans-1,2-Dfchtoroethene
trans-1,3-Dichloropropene
cls-1,3-Dichk5ropropene
1,1,2,2-Tetrachloroethane
Tetrachloroethene
Trichtoroethene
1,2-Dtehtoroe thane
1,1-Dichloroethane
DataChem PE Sample Precision
Site: McClellan Mix 3
20 30
Relative Standard Deviation, %
Figure 4-6. Laboratory precision on MAFB PE samples containing mix 3.
Trichloroethene, tetrachloroethene, 1,1-dichloroethane, and benzene were
spiked into the spike/low samples.
36
-------
Target Compound
1,2-Dichlorobenzene
Chlorobenzene
Dibromochloromethane
Tetrachloroethene
1,1,2-Trichloroethane
1,2-Dichloropropane
Trichloroethene
Carbon Tetrachloride
Chloroform
1,1-Dichloroethene
Methylene Chloride
1,1-Dichloroethane
Trichlorofluoromethane
DataChem PE Sample Recovery
Site: Savannah River Mix 1
70
80 90 100 110 120
Average Percent Recovery
130
140
150
Figure 4-7. Laboratory mean recoveries for SRS PE samples containing mix
1. Trichloroethane was spiked into the spike/low samples.
Target Compound
trans-1,3-Dichloropropene
cis-1,3-Dichloropropens
1,2-Dibromo-3-Chloropropane
1,2,3-Triohloroprapane
1,1,2,2-Tetrachloroethane
1,1,1,2-TetrachloroethanB
1,2-Dibromoethane
1,3-Dichloropropane
Tetrachloroethene
1,1,2-Triohloroethane
1,2-Dichloropropane
Trichloroethene
1,2-Dichloroethane
1,1 -Dichloropropene
DataChem PE Sample Recovery
Site: Savannah River Mix 2
ID Spike/Low
0High
E3Mid
Low
50 60 70 80 90 100 110 120
Average Percent Recovery
130
—I—
140
150
Figure 4-8. Laboratory mean recoveries for SRS PE samples containing mix
2. Trichloroethane and tetrachloroethene were spiked into the spike/low
samples.
37
-------
Target Compound
Benzene
trans-1,3-Dfchtoropropene
cis-1,3-Dichloropropeno
1£-Dibromo-3-Chloropropan9
1.2,3-Trichtofopropane
1.1,2,2-Tetrachloroelhane
1,1,1,2-Telrachtoroolhane
1,2-Dibromoethana
1,3-Dtchtoropropana
TetracNoroalhane
1.1,2-Trichtoroathan*
1,2-Dtehtoropropane
Trichtofoathena
1,2-Dfchloroalhane
1,1-DfcNoropropene
1,1-Dtchloroethane
DataChem PE Sample Recovery
Site: McClellan Mix 2
D3 Spike/Low
0High
I Mid
I Low
50
60 70 80 90 100 110 120
Average Percent Recovery
130
140
150
Figure 4-9. Laboratory mean recoveries for MAFB PE samples containing
mix 2. Trichloroethene, tetrachloroethene, 1,1-dichIoroethane, andl benzene
were spiked into the spike/low samples.
Target Compound
Bromoform
Ethylbenzene
Toluene
BromodlcNoromethane
Benzene
1,1,1-Trichloroethane
trans-1,2-Dfchloroethene
trans-1,3-D!chloropfopene
dv1,3-Dictiloropropene
1,1,2,2-Tatrachloroethane
TrtracWoroethene
Trichloroethene
1,2-DfchIoroethane
1,1-Dichloroethane
DataChem PE Sample Recovery
Site: McClellan Mix 3
80 90 100 110
Average Percent Recovery
Figure 4-10. Laboratory mean recoveries for MAFB PE samples containing
mix 3. Trichloroethene, tetrachloroethene, 1,1-dichloroethane, andl benzene
were spiked into the spike/low samples.
38
-------
t T !nZ6ne a*Se!ected c<™*ntration levels. The implications of these exceptions for the
technology evaluation are further discussed, if applicable, in Chapter 5. The compiled absolute percent
differences (APDs)4 for all PE sample results had a median value of 7% and a 95th percentile vaLo?25%.
Groundwater Sample Precision
Relative standard deviations are given in Table 4-5 for compound concentrations in excess of 1 ng/L in
fon± t T7f,rOI±6 SRS demonstration- Trichloroethene and tetrachloroethene were Ae only
contaminants detected in SRS groundwater samples. A similar compilation of RSD values from the MAFB
groundwater samples is included in Table 4-6. These values are based on analytical resute from ehheT^ee or
four replicate samples. With three exceptions, all tabulated values are less than 20%
Table 4-5. Summary of SRS Groundwater Analysis Precision
Very low 2
Low1
Low 2
Midi
Mid 2
Highl
High 2
Very high 1
Very high 2
Relative Standard Deviation (%)
TCE
10.6
34.4
5.4
7.1
9.4
7.3,
0.8
11.8
8.4
6.2
PCE
14.3
12.4
5.7
8.7
11.6
4.2
1.8
7.9
5.7
6.3
Table 4-6. Summary of MAFB Groundwater Analysis Precision
Sample
Description
Very low 1
Very low 2
Low1
Low 2
Midi
Mid 2
Highl
High 2
Very high 1
Very high 2
Notes: 11DCE = 1,1
11DCE
9.1
2.6
6.8
11.5
12.0
2.5
-dichloroethe
TCE
5.0
<0.1
3.7
5.2
10.5
3.6
2.4
5.3
5.4
8.0
ne; TCE = trie
CLFRM
1.3
2.0
4.9
20.9
5.3
5.2
6.4
tlloroethene: I
rveictuve O
CCL4
4.2
1.9
4.0
4.9
SLFRM = rhlf
[anaara u
PCE
5.7
22.3
13.9
irnfnrm- r*f"M j
eviation (-5
11DCA
<0.1
4.1
.9.4
*)
C12DCE
3.8
12.6
3.8
4.1
5.1
6.5
10.1
t12DCE
3.8
BNZN
4.9
, . - , = caron eracore- PCE = tetrachi
ecel,^^
4 The absolute percent difference is the absolute value of the percent difference between a measured value and a true value.
39
-------
Summary of Reference Laboratory Data Quality
With the exceptions noted below, a review of DCL analytical data showed that all Method 8260A QC criteria
were met. Internal standard recovery limits were not met for one sample. The results for this sample were
markedly different from the other three samples in the replicate set and the sample was omitted from the data set.
Six samples failed one or more surrogate standard recovery criteria. These sample results were compared with
replicate sample results. Five of the six samples were comparable and were included in the reference data set.
The data for the remaining sample were not comparable and were omitted from the reference data set. Other
quality control deviations, which are summarized in Table 4-3, did not significantly affect the quality of the
laboratory data.
A review of DCL precision and accuracy on field-prepared PE mixtures corroborates laboratory internal QC
results. A similar precision evaluation on groundwater samples from both sites further supports these
observations. Overall, the internal and external QC data reveal appropriate application and use of Method
8260A by DataChem Laboratories. The laboratory results for groundwater samples from both sites are
considered suitable for use as a reference data set.
40
-------
Chapter 5
Demonstration Results
Type 1312 Calibrated and Reported Compounds
The 1312 was calibrated primarily for TCE and PCE in this demonstration, and PE mixtures containing only
these two compounds were provided to the Innova team for analysis. In selected groundwater samples, other
compounds were reported when the appropriate bandpass filters and calibration data for the instrument were
available in the field. Table 5-1 lists the compounds for which the 1312 reported results in either PE or
groundwater samples.
Table 5-1. Type 1312 Calibrated and Reported Compounds
Trichloroethene
Compounds Reported at SRS
Tetrachloroethene
1,1,1-Trichloroethane
Trichloroethene
Compounds Reported at MAFB
Tetrachloroethene
Carbon tetrachloride
c/s-1,2-Dichloroethene
Chloroform
Preanalysis Sample Information
As noted in Chapter 2, successful use of the Type 1312 requires that the composition of the sample be known so
that spectral interferences can be accounted for. Consequently, both PE and groundwater samples were
accompanied by information on the chain-of-custody forms indicating the contaminants that were present in the
sample. For groundwater samples, the most recent quarterly monitoring results from each sampled well (given in
Chapter 3) were used for information on sample composition.
Sample Completion
All but one of the 141 PE and groundwater samples submitted for analysis to the Innova team were completed at
both demonstration sites. A PE sample from the very low category at the SRS was lost by the Innova team during
handling.
41
-------
Blank Sample Results
Eight blank samples were provided for analysis at each demonstration site. False positive detects were counted
only for compounds reported at concentration levels greater than 1 ng/L. A list of false positive detects is given
for both sites in Table 5-2.
Table 5-2. False Positive Rates from Blank Sample Analysis
SRS Blank Samples
Trichloroethene
Tetrachloroethene
False Positive
1 of 8 (13%)
Oof 8
MAFB Blank S.
Compound
Trichloroethene
Tetrachloroethene
imples
False Positives
1 of 8 (13%)
Oof 8
Performance at Instrument Detection Limit
Ten replicate samples of a PE mixture at a concentration level of 10 jxg/L were provided for analysis at each site.
Reported nondetects were compiled and are given as percent false negatives in Table 5-3. Vendor-provided
compound detection limits are also shown in the table for comparison.
Table 5-3. False Negative Rates from Very Low Level PE Sample Analysis
SRS PEMixl
10uq/L(TCEandPCE)
Trichloroethene (2)
Tetrachloroethene (2)
False Negative
Oof 6
3 of 6 (50%)
MAFB PEMixS
9 uq/L TCE and 11 wa/L PCE
Compound
Trichloroethene (2)
Tetrachloroethene (2)
False Negative
1 of 4 (25%)
1 of 4 (25%)
Note: Vendor-provided detection limits (in ng/L) are shown in parentheses after each compound
PE Sample Precision
Precision results from each of the replicate sample sets provided from seven PE mixtures at the SRS and six
mixtures at MAFB are shown in Figures 5-1 and 5-2, respectively. The figures show the relative standard
deviation for TCE and PCE at the concentration levels used in the study. (The composition and concentrations of
each of these mixtures are given in Table 3-8 for SRS and Table 3-9 for MAFB.) Note that precision and
accuracy were not determined for the "very low" concentration level. The data are also presented in tabular form
in Table 5-4.
Table 5-4. Precision for TCE and PCE at Both Sites
Compound
Trichloroethene
Tetrachloroethene
Site
SRS
MAFB
SRS
MAFB
Relative Standard Deviation (%)
Low 1
22
16
46
38
Low 2
18
17
21
22
Midi
13
20
11
13
Mid 2
7
14
8
22
Highl
19
4
8
16
High 2
9
10
Very Highl
22
8
7
5
Note: Blank cells Indicate that no data were reported,
42
-------
Compound
Innova Multi-Gas Monitor PE Sample Precision
Site: Savannah River
Tetrachloroethene
Trichloroethene
IHV High
H High 2
0High 1
0Mid2
lEMidl
BLow 2
Low1
10 20 30
Relative Standard Deviation, %
40
Figure 5-1. Type 1312 PE sample precision at the SRS.
Innova Multi-Gas Monitor PE Sample Precision
Compound Site: Savannah River
Tetrachloroethene
50
Trichloroethene
10
20 30
Relative Standard Deviation,'
40
50
Figure 5-2. Type 1312 PE sample precision at MAFB.
43
-------
Overall instrument precision is summarized in Table 5-5 for PE mixtures used at each site. For this summary,
RSD values from all PE sample analyses for all compounds at each site were pooled and the median and 95
percentile values of the distribution were computed.
Table 5-5. Summary of PE Sample Precision and Percent Difference Statistics for the
SRS and MAFB
Parameter
RSD, %
Absolute percent
difference
Percentile
50th
95m
Number in pool
50th
95th
Number in pool
SRS
12
30
14
36
48
14
MAFB
16
29
12
9
22
12
Combined Sites
15
34
26
29
47
26
PE Sample Accuracy
The accuracy of the Type 1312 in analyzing PE samples was determined by comparing the average value trom
each of the replicate sample sets with the known concentration of the PE mixture (Tables 3-8 and 3-9 for SRS
and MAFB, respectively). These comparisons are shown as percent recoveries1 in Figures 5-3 and 5-4 for the
SRS and MAFB, respectively. To assist in assessing the sign of the difference, the percent recovery data are
plotted as either a positive or negative difference from the 100% recovery line. The percent recovery values are
also expressed as absolute percent difference (APD) values2 and are shown in Table 5-6. (For example, a 90%
recovery is equivalent to a -10% difference; a 120% recovery is equivalent to a +20% difference.) Table 5-5
contains a summary of overall 1312 absolute percent differences relative to the true or reference value of the PE
mixtures, along with the precision summary. These summaries are from pooled TCE and PCE data from each
site. The median and 95th percentiles of the absolute values of these pooled values were computed and are
reported under the absolute percent difference category in Table 5-5.
Table 5-6. Target PE Compound Recovery at Both Sites
Target Compound
Trichloroethene
Tetrachloroethene
Site
SRS
MAFB
SRS
MAFB
Low1
39
20
42
24
Low 2
39
15
34
21
Midi
34
4
25
8
Mid 2
35
11
33
2
Highl
35
5
34
3
High 2
38
36
Very Highl
48
10
48
7
Note: Blank cells indicate that no data were reported.
1 Percent recovery is the Type 1312 value divided by the true value, multiplied by 100.
2 The absolute percent difference is the absolute value of the percent difference between a field and reference (in this case the
reference laboratory) measurement. As an example, the percent difference between a field measurement of 85 and a
laboratory measurement of 110 is -22.7% and the absolute percent difference is 22.7%.
44
-------
Compound
Innova - Mult-Gas Monitor PE Sample Recovery
Site: Savannah River
Tetrachloroethene
Trichloroethene
IDV High
S High 2
fflMidl
SLow 2
Low 1
20 40 60 80 100 120 140
Average Percent Recovery
160 180 200
Figure 5-3. Type 1312 PE sample recovery at the SRS.
Compound
Innova Mult-Gas Monitor PE Sample Recovery
Site: McClellan
Tetrachloroethene
OH V High 1
El High 1
0Mid2
El Midi
ID Low 2
El Low 1
Trichloroethene
20 40
Figure 5-4. Type 1312 PE sample recovery at MAFB.
60 80 100 120 140 160 180 200
Average Percent Recovery
45
-------
Comparison with Laboratory Results
At each demonstration site, a total of 33 samples collected from 10 wells were provided to the participants and to
the reference laboratory. Replicate sample sets were composed of either 3 or 4 samples from each well. Average
laboratory results from each replicate set were used as the reference values for comparison with technology
results. A side-by-side comparison of laboratory and Type 1312 results for all groundwater samples is given in
Table 5-7 for the SRS and Table 5-8 for MAFB. The RSD values and their statistical summaries are included in
the table. Well designation (very low, low, mid, high, and veiy high) is based on TCE concentration levels;
however, other compounds were present in the groundwater samples at concentration levels noted in the tables.
The average percent difference between average Type 1312 and laboratory results for TCE and PCE at SRS and
TCE only at MAFB is shown in Figures 5-5 and 5-6, respectively. Average laboratory results for groundwater
contaminants reported at levels less than 1 |Hg/L are not included in the comparison. The SRS groundwater
comparison in Figure 5-5 includes only TCE and PCE. Two well samples at the SRS were also contaminated
with 1,1-dichloroethene, chloroform, and carbon tetrachloride, as noted in Table 5-7. The groundwater samples
at MAFB were more complex, as indicated by the additional compounds shown in Table 5-8. As noted
previously, the Type 1312 was configured for analysis of only a few compounds, principally TCE and PCE, and
thus was unable to detect the other compounds in the samples.
Table 5-7. Type 1312 and Reference Laboratory Results for SRS Groundwater Samples
Sample
Description
Very low 1
Very tow 2
Low1
Low 2
Midi
Mid 2
Highl
High 2
Very high 1
Very high 2
Range
Well
Number
MSB 33B
MSB 33C
MSB 18B
MSB 37B
MSB4D
MSB64C
MSB4B
MSB70C
MSB14A
MSB8C
Compound
Trichloroethene
Tetrachloroethene
Trichloroethene
Trichloroethene
Tetrachloroethene
Trichloroethene
Tetrachloroethene
Chloroform
Carbon tetrachloride
Trichloroethene
Tetrachloroethene
Trichloroethene
Tetrachloroethene
1,1-Dichloroethene
Trichloroethene
Tetrachloroethene
Trichloroethene
Tetrachloroethene
1,1-Dichloroethene
Trichloroethene
Tetrachloroethene
Trichloroethene
Tetrachloroethene
Replicates
3
3
3
4
4
3
3
4
3
3
Lab.
Avg.
(MB/L)
9.0
3.5
2.4
11
27
27
22
1.3
1.0
150
87
35
240
12
747
33
1875
520
32
1367
800
4933
3668
Median
95" Percenfite
Lab.
RSD
(%)
11
14
34
5
6
7
9
0
15
9
12
7
4
8
1
2
12
8
8
8
6
6
6
0-34
8
15
Type 1312"
Avg.
(ug/L)
28
37
13
29
20
31
17
NR
NR
124
73
45
203
NR
453
34
1345
451
NR
961
816
3519
3835
Type 1312a
RSD
(%)
50
87
41
13
63
12
23
NR
NR
7
6
16
16
NR
53
18
13
21
NR
5
4
15
5
4-87
16
65
* NR = not reported.
46
-------
Table 5-8. Type 1312 and Reference Laboratory Results for MAFB Groundwater Samples
Sample
Description
Very low 1
Very low 2
Low 1
Low 2
Mid 1
Mid 2
High 1
High 2
Very high 1
Very high 2
Range
Median
well
Number
EW-86
MW-349
MW-331
MW-351
EW-87
MW-341
MW-209
MW-330
MW-334
MW-369
Replicates
3
3
4
3
4
3
3
4
3
3
Compound
Trichloroethene
1,1-Dichloroethene
Trichloroethene
Tetrachloroethene
Chloroform
1,1-Dichloroethene
Carbon tetrachloride
1,1-Dichloroethene
1,1-Dichloroethane
Carbon tetrachloride
Chloroform
Trichloroethene
Freon11
1,1-Dichloroethene
1,1-Dichloroethane
c/s-1 ,2-Dichloroethene
Carbon tetrachloride
Trichloroethene
1,1-Dichloroethene
1,1-Dichloroethane
c/s-1 ,2-Dichloroethene
1,1,1 -Trichloroethane
Trichioroethene
Tetrachloroethene
c/s-1 ,2-Dichloroethene
Chloroform
Trichloroethene
c/s-1 ,2-Dichloroethene
Chloroform
Trichloroethene
trans-1 ,2-Dichloroethene
c/s-1 ,2-Dichloroethene
Chloroform
1 ,2-Dibromochloropropane
Trichloroethene
1,1-Dichloroethene
c/s-1 ,2-dichloroethene
Chloroform
Benzene
Trichloroethene
Carbon tetrachloride
c/s-1 ,2-Dichloroethene
Chloroform
Carbon tetrachloride
frichloroethene
Lab.
Avg.
(U9/L)
4.6
7.7
13
2.0
9.0
3.8
137
2.5
15
7.5
4.8
16
20
1.5
5.1
1.5
1.4
22
180
3.0
3.3
6.8
114
1.2
15
3.5
280
38
6.9
238
7.7
66
42
6.1
380
690
237
397
283
10,667
350
207
63
51
6167
95m Percentile
Lab.
RSD
{%)
5
9
0
6
1
3
4
7
0
2
2
4
6
12
4
4
4
5
12
9
13
12
11
14
4
5
4
4
21
2
4
5
5
6
5
3
7
5
5
5
5
10
6
5
8
0-21
5
13
Type 1312a
Avg.
(ng/L)
13
NR
34
149
NR
NR
NR
NR
NR
9.8
57
40
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
146
17
NR
130
244
NR
64
262
NR
NR
NR
NR
NR
NR
NR
NR
NR
13,864
NR
NR
NR
NR
6443
L
Type 1312a
RSD
(%)
22
NR
27
32
NR
NR
NR
NR
NR
3
97
5
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
12
5
NR
23
8
NR
21
8
NR
NR
NR
NR
NR
NR
NR
NR
NR
4
NR
NR
NR
NR
11
3-97
12
55
^— ^— ••^•^
47
-------
Compound
Model 1312 GW Sample Difference
Site: Savannah River Ref: Laboratory
Tetrachloroethone
Trichloroethene
v High 2
HVHighl
D High 2
• Highl
DMid2
Midi
DLow2
Low1
DV Low 2
•V Low1
-200
-150
-100
-50 0 50
Average Percent Difference
100
150
Figure 5-5. Type 1312 groundwater results at the SRS relative to laboratory results.
Innova - Multi-Gas Monitor GW Sample Difference
Compound Site: McClellanAFB Ref: Laboratory
HVHIgh2
HV Highl
DHigh2
• Highl
DMid2
Midi
DLow2
Low1
nV Low2
•V Low1
i1 ., "IB —IH l|!'
-200
-150
-100
-50 0 50
Average Percent Difference
100
150
Figure 5-6. Type 1312 groundwater results at MAFB relative to laboratory results.
48
-------
The median and 95th percentiles of the distribution of absolute percent differences between 1312 and laboratory
results for all groundwater samples are given in Table 5-9.
Table 5-9. Type 1312 Absolute Percent Difference Summary for
Pooled Groundwater Results
Percentile
50th
95th
Number of samples in pool
SRS
26
493
19
MAFB
156
4922
14
29
2236
33
To assess the degree of linear correlation between some of the 1312 and laboratory groundwater data pairs shown
in Tables 5-7 and 5-8, correlation coefficients (r) were computed. The correlation analysis was carried out for all
TCE and PCE data pairs with laboratory values less than or equal to 300 ug/L. Only a few data pairs exist above
this concentration level and including them in these analyses would result in spuriously high r values (Havlicek
and Grain, 1988). The computed correlation coefficients are shown in Table 5-10.
Table 5-10. Correlation Coefficients for Reference Laboratory and
Type 1312 Groundwater Analyses
Data Set
SRS Laboratory (1 through 300 ug/L)
MAFB Laboratory (1 through 300 ug/L)
Correlation
Coefficient
0.984
0.892
Number of
Data Pairs
12
8
Sample Throughput
The throughput rate for the Type 1312 was in the range of one to two water samples per hour. Two identical
instruments were used at the MAFB site; however, these sample rates assume operation with one instrument.
Performance Summary
Table 5-11 contains a summary of instrument performance parameters and operational features of the 1312 that
were verified in this demonstration. For groundwater samples, the precision results for the reference laboratory
are given alongside the Type 1312 performance results to facilitate comparison of the two methodologies.
49
-------
Table 5-11. Type 1312 Performance Summary
Blank sample
Detection limit sample
Instrument
Feature/Parameter
:alse positives detected at low (<15%) rates for TCE.
False negatives reported at low rates (<25%) for TCE and higher (<50%) rates for PCE at
concentration levels of approximately 10 ng/L.
'E sample precision
•CE, RSD range: 4 to 22%
PCE, RSD range: 5 to 46%
•CE and PCE combined, Type 1312 median RSD: 15%; QS^percentile RSD: 34%
'CE and PCE combined, reference laboratory median RSD: £)%; 95th percentile RSD:
18%
PE sample accuracy
TCE, APD range: 4 to 48%
PCE, APD range: 2 to 48%
TCE and PCE combined, Type 1312 median APD: 29%; 95th percentile APD: 47%
TCE and PCE combined, reference laboratory median APD: 10%; 95th percentile APD:
25%
Type 1312 comparison
with laboratory results
or groundwater
samples
Analytical versatility
Sample throughput
Support requirements
Operator requirements
Total system weight,
including accessories
Portability
Total system cost
Shipping requirements
Performance Summary
Type 1312 median RSD: 15%
Type 1312 95th percentile RSD: 73%
Laboratory median RSD: 6%
Laboratory 95th pejrcentile RSD: 14%
Type 1312: laboratory median APD: 29%; 95th percentile APD: >2000%
Type 1312: laboratory correlation:
SRS (5300 ng/L) r= 0.984
MAFB (<300ng/L) r= 0.892
PE samples: calibrated for TCE and PCE; up to three components can be detected if
sample composition is known.
GW samples: calibrated for TCE and PCE; also reported results for carbon tetrachloride,
chloroform, and c/s-1,2-dichloroethene when possible. Reported results for 29 of 31
detects of TCE and PCE in all GW samples reported by the laboratory.
1-2 samples per hour
110-V ac power or 12-V dc-to-ac with inverter
Sample processing: minimally trained technician
Data processing and review: B.S. chemist or equivalent
30 pounds
Transportable - best suited for use in vehicle at the wellhead
$28,000 - $35,000 (depending upon options selected)
Air freight, luggage check, carry-on
50
-------
Chapter 6
Field Observations and Cost Summary
Introduction
The following subsections summarize the audit findings obtained while observing instrument operation at both
sites. The purpose of the audits was to observe the instrument in operation as well as to verify that the analytical
procedures used during the demonstration were consistent with the written procedures submitted to the
verification organization prior to the field demonstration. An instrument cost summary and an applications
assessment are also provided.
Method Summary
The 1312 employs a static (equilibrium) headspace method with no temperature control. It uses photoacoustic
infrared analysis of the headspace vapors at preselected wavelengths to measure selected chlorinated VOCs in
water. Compounds are quantified by using a classical least-squares matrix analysis at multiple wavelengths to
account for spectral overlaps when more than one chlorinated VOC is encountered in the sample. The method
requires that the composition of the sample be known so that spectral interferences from other chlorinated
compounds can be avoided by selecting the appropriate bandpass filter in the instrument.
Equipment
The 1312, shown in Figure 6-1, has dimensions of 7 inches x 16 inches x 11 inches and weighs 20 pounds
without accessories. A magnetic stir plate accessory has dimensions of 4 inches x 5 inches x 5 inches and
weighs 6 pounds. A glass, 2-L flask with a dual-inlet cap and Teflon tubing for connection to the instrument is
also employed. The system did not include a printer. It can be powered from line ac or from a 12-V dc auto
battery. The system was powered with 110 V ac from a portable generator at the SRS demonstration and by line
ac from the local power grid at McClellan. The equipment was transported as air freight to both demonstration
sites. The system could also be checked as baggage in a shipping case.
Sample Preparation and Handling
At the request of the Innova team, PE and groundwater samples were provided in 1-L amber bottles with zero
headspace. Sample preparation was begun by pouring the entire contents of the cold 1-L sample into the 2-L
flask. This flask, containing a magnetic stir bar, was immediately sealed with a cap equipped with two inlets and
microvalves in the closed position. The flask was placed on a motor-driven stir plate and was stirred at moderate
speed under ambient temperature conditions. After 30 minutes, Teflon tubing leads from the two inlets on the
cap were connected to the inlet and outlet ports on the rear of the 1312 instrument and the microvalves near the
cap were opened. An air pump inside the instrument circulated headspace vapors through the analysis cell for 20
to 30 seconds prior to a 11A minute measurement period. The analysis cycle consisted of ten consecutive
51
-------
Figure 6-1. The Innova Type 1312 Multi-gas Monitor.
measurements of the headspace vapor. An average value from the ten measurements was used to compute the
results for each sample. The entire measurement period, including equilibration time., lasted for about 45
minutes. The Innova team chose to use a 60-minute equilibration time at MAFB. Consequently, the total
analysis time at that site was 75 minutes. The analysis result is displayed in units of parts per million volume on
a display panel. This result is then converted to micrograms per liter (ppb) in water through the use of a
calibration curve. The instrument also measures the temperature of the gas in the cell and the water vapor
content in the sample. These data are used in the data processing algorithm to correcl: for water vapor spectral
interferences. Following each analysis, the gas analysis cell in the instrument was flushed with ambient air and
the headspace flask was rinsed with distilled water before refilling it with another sample.
Consumables
The only consumable used in operating the instrument is a fine-particle air inlet filter that needs replacement
once a month.
Historical Use
The 1312 has been used extensively for chlorinated VOC measurements in soil gas and air at various
environmental sites, including extensive testing and routine operation at SRS. This is the first demonstration of
the 1312 for analysis of volatile organics in water using this headspace technique.
Equipment Cost
The 1312 has a price range of $28,000 to $35,000, depending upon the number of instrument options selected.
Instrument costs are summarized in Table 6-1. For the purposes of comparison, reference laboratory costs were
$95 per sample in addition to overnight shipping costs of approximately $30 per batch of 12 samples. Sample
throughput for the 1312 was on the order of 1 to 2 samples per hour.
52
-------
Table 6-1. Type 1312 Cost Summary
Instrument/Accessory
Instrument (131 2)
Instrument accessories (flasks, stir motor, tubing)
Sample handling accessories
(syringes, vials, standards)
Maintenance costs
Cost
$28,000 to 35,000
$300 to $500
None required
$100/year
Operators and Training
One operator is required for system operation. For wellhead monitoring operations where operator and the
instrument would follow a well-sampling team, one operator is probably sufficient since only one or two samples
would be provided per hour. Sample preparation and instrument operation could be carried out by a field
technician with several hours of training.
Data Processing and Output
The instrument, as configured for the demonstration, reported analytical results in units of parts per million
volume for up to five calibrated compounds. Results were read from the instrument display and manually
recorded by the instrument operator. The results were converted to units of micrograms per liter in water by
reference to a series of calibration curves produced during instrument calibration at the manufacturer's facility in
Denmark before the demonstration. Thus, considerable operator involvement in the data analysis and reporting
process was required. The instrument is capable of communication with a laptop computer and in fact can be
operated from a computer using Innova software developed for the instrument. However, these features were not
employed at this demonstration.
Compounds Detected
The instrument was calibrated for the following five compounds prior to the demonstration: TCE, PCE,
chloroform, carbon tetrachloride, and cw-l,2-dichloroethene. As noted in Chapter 2, the method utilizes an
infrared spectrophotometric method with many potential spectral interferences in multicomponent mixtures. The
instrument is only capable of analyzing five or fewer compounds simultaneously and prior knowledge of the
composition of the sample is required to avoid spectral interferences and possible erroneous readings.
Initial and Daily Calibration
The instrument was calibrated at the factory before the demonstration by injecting known amounts of the target
compounds into 1 L of water and recording the resulting concentration in the headspace. The analyses were
repeated at three or four points over the desired calibration range. Linear regression was used to determine the
relationship between water concentration (u.g/L) and headspace vapor concentration (ppmv). Three sets of
calibration curves were prepared for each of the five target analytes over the entire working range of the
instrument in the following concentration categories: low, 0 to 500 u,g/L; medium, 500 to 1000 |iig/L; and high
1000u.g/Landup.
The instrument's calibration is reported to be highly stable over a period of a year or more. Consequently, daily
calibration checks are not specified in the field method.
53
-------
QCProcedures and Corrective Actions
No QC checks were included in the field method.
Sample Throughput
Maximum throughput is about one sample per 45 minutes. Typical throughput is between one and two samples
per hour.
Problems Observed During Audit . .
Data from PE samples of known composition reported initially at the SRS site showed a consistent negative bias.
Further investigation into this discrepancy revealed that data were being reported in units of microhters per liter
in water instead of micrograms per liter in water. Liquid density values for the target compounds were obtained
by the Innova team and the data were converted to the appropriate units. No temperature control of the sample
was done at either site. The temperature of the sample was inferred from a dewpoint measurement that was made
with each sample. At the SRS, the temperature of the sample was typically below thai: used during calibration of
the instrument. This temperature effect would translate to a negative bias in the reported results since
equilibrium headspace concentrations increase with temperature. Sample temperatures at MAFB were in closer
agreement with those used during calibration, resulting in less bias in the reported results.
Data Availability and Changes
Data from the 1312 were obtained at the end of each demonstration day m hard copy. Data were provided in
spreadsheet format at the conclusion of each demonstration week. Several typographical errors were corrected at
the final data review; however, with the exception of the density correction factor noted above, no substantive
data changes were made. The instrument operators also reported instances where specific interferents might
influence reported TCE and PCE results in groundwater samples. Such information would be useful in data
interpretation by the final user if previous information on groundwater composition is available.
Applications Assessment ...,.* 1^-1+1-
This demonstration was intended to provide an assessment of the instrument's suitability for analytical tasks in
site characterization and routine site monitoring. Site characterization refers to those instances where subsurface
contamination is suspected, but information on specific compounds and their concentration level is not available.
The instrument best suited for this application is one that can screen a wide array of compounds in a timely and
cost-effective manner. Analytical precision and accuracy requirements may be relaxed in these instances since a
general description of the site characteristics is adequate for remediation planning. At the other end of the
spectrum is a monitoring application where contaminant compounds and their subsurface concentrations are
known with some certainty. Periodic monitoring requirements imposed by local regulatory agencies may specify
that analyses be carried out for specific contaminant compounds known to be present in the water. Quarterly
well-monitoring programs fall into this category.
Based on its performance in this demonstration, the Innova Type 1312 Multi-gas Monitor is best suited for
monitoring applications where the contaminant content of the samples is known. For example, the instrument
could be used in a routine quarterly monitoring program to analyze TCE in groundwater samples. The instrument
is not well suited for characterization or screening applications where the contaminants at the site are not known
since interferences from unknown contaminants in the samples could cause considerable error in the reported
results.
54
-------
Chapter 7
Technology Update
Note: The following comments were submitted by the technology developer. They have been edited for format
consistency with the rest of the report. The technical content in the following comments has not been verified by
the verification organization.
During the testing of PE samples at the SRS, we encountered some low recoveries that we suspected were caused
by temperature. Our instrumentation allows us to read out the dewpoint, which was 58 to 70 °F at SRS and 75 to
80 °F during the calibration in the laboratory. For that reason we were more concerned about temperature at
MAFB and therefore allowed the samples to equilibrate for a longer time before analysis. At MAFB we obtained
dewpoints of 70 to 85 °F, which were closer to those observed during the calibration, and noted analysis
recoveries closer to 100%.
We are currently working on procedures that will be less dependent on sample temperature, including
thermostatting and temperature compensation algorithms. We are also working on a system for permanent on-
line monitoring of chlorinated organic solvents using the same technology. The first installations are already in
place in England.
_
55
-------
Chapter 8
Previous Deployments
The following documents report on previous use of the 1312 Multi-gas Monitor.
L. C. Waters, R. A. Jenkins, R. W. Counts, and L. Hernandez, "A Photoacoustic Infrared Method for the
Detection of Selected Chlorinated Volatile Organic Chemicals (VOC) in Water: Method OSO30," in DOE
Methods for Evaluating Environmental and Waste Management Samples, pp. OSO30-1 to OSO30-15, DOE/EM-
0089T; U.S. Department of Energy, Washington, DC, 1994.
J. E. Sollid, "Soil Pore-Gas Sampling by Photoacoustic Radiometry," paper presented at the Air.and Waste
Management Association/Society of Photo-optical Instrumentation Engineers (AWMA/SPIE) Conference on
Optical Sensing for Environmental and Process Monitoring, Air and Waste Management Association, Pittsburgh,
PA, 1994.
K. Gunn, Z. Quo, and B. A. Tichenor, "Tracer Gas Measurement of Indoor-Outdoor Air Exchange Rates," paper
presented at the EPA/Air and Waste Management Association Specialty Conference on Measurements of Toxic
and Related Air Pollutants, Air and Waste Management Association, Pittsburgh, PA, 1994.
W. Buttner, P. Wagner, A. Husain, S. Pfeifenrot, K. Dooly, and S. Barrie, "In-situ Sampling of Aqueous-Phase
Contamination of Chlorinated Solvents," paper presented at the Air and Waste Management Association
Specialty Conference on Field Analytical Methods for Hazardous Wastes and Toxic Chemicals, Air and Waste
Management Association, Pittsburgh, PA, 1997.
56
-------
References
Bevington, P. R., 1969, Data Reduction and Error Analysis for the Physical Sciences, pp. 52-60. McGraw-Hill,
New York.
DataChem, 1997, "DataChem Laboratories Environmental Chemistry/Radiochemistry Quality Assurance
Program Plan," 1997 Revision, DataChem Laboratories, Salt Lake City, UT.
EPA, 1996a, "A Guidance Manual for the Preparation of Site Characterization and Monitoring Technology
Demonstration Plans," Office of Research and Development, National Exposure Research Laboratory, Las
Vegas, NV. [Available at the ETV Web Site (www.epa.gov/etv) in pdf format.]
EPA, 1996b, "Test Methods for Evaluating Solid Waste: Physical/Chemical Methods; Third Edition; Final
Update III," Report No. EPA SW-846.3-3, Government Printing Office Order No. 955-001-00000-1, Office of
Solid Waste and Emergency Response, Washington, DC.
Havlicek, L. L., and R. D. Grain, 1988, Practical Statistics for the Physical Sciences, pp. 80-93. American
Chemical Society, Washington, DC.
Sandia, 1997, "Demonstration Plan for Wellhead Monitoring Technology Demonstration; Sandia National
Laboratories," Albuquerque, NM. [Available at the ETV Web Site (www.epa.gov/etv) in pdf format.]
57
-------
------- |