The 15th Annual Waste Testing &
Quality Assurance Symposium
^^^B ^^^^v • B^LH ^^ ^^^^^
WIDA
+ £^f\
July 18-22,1999
Crystal Gateway Marriott • Arlington, VA
-------
WTQA'99
PROGRAM COMMITTEE
Symposium Co-chairs:
David Friedman
Gail Hansen
Larry Keith
Advisory Board:
Anthony Pagliaro
Jerry Parr
Organics:
Joan Fisk
Barry Lesnik
Inorganics:
Ollie Fordham
Quality Assurance:
Duane Geuder
Charles Sellers
General:
Kim Kirkland
U.S. EPA, Office of Research & Development
U.S. EPA, Office of Solid Waste
Waste Policy Institute
ACIL
Catalyst Information Resources, LLC
U.S. EPA, Office of Emergency and Remedial Response
U.S. EPA, Office of Solid Waste
U.S. EPA, Office of Solid Waste
U.S. EPA, Office of Emergency and Remedial Response
U.S. EPA, Office of Solid Waste
U.S. EPA, Office of Solid Waste
WPI
Waste Policy Institute
A Virginia Tech Affiliated Corporation
m.
(9
Virginia
Tech
-------
Proceedings
The Fifteenth Annual
Waste Testing & Quality Assurance
Symposium
(WTQA '99)
July 18-22, 1999
Crystal Gateway Marriott
Arlington, VA
-------
HIGHLIGHTS
15th Annual Waste Testing & Quality Assurance Symposium
(WTQA '99)
Preparing for Change Under PBMS
WTQA '99 is sponsored by the Waste Policy Institute under a cooperative agreement with the U.S. Environmental
Protection Agency. Co-Sponsors include ACIL, ACS Division of Environmental Chemistry, Virginia Tech Chemistry
Department, and EnviroAnalysis.
Highlights
The theme of WTQA '99 is "Preparing for Change Under PBMS." Performance Based Measurement System
(PBMS) is a new approach to compliance monitoring that has been initiated by EPA. PBMS will allow facilities to
use any scientifically valid technologies or methods to demonstrate compliance, rather than require EPA-approved
methods. PBMS' aims include reducing costs incurred by regulated entities that must demonstrate regulatory
compliance, and helping laboratories to improve their productivity. Although PBMS will add flexibility to the data
gathering process, it also mandates that all methods yield accurate compliance determinations. Some of those in
the environmental community who will be affected by PBMS include permit writers, state and federal enforcement
officials, and regulated entities. WTQA '99 will provide the latest information on the implementation of PBMS.
PBMS Status and Issues Session
The latest information on PBMS implementation from the RCRA, CERCLA, CWA, SDWA and CAA programs will
be presented. In addition reports from NELAC, ELAB, ACS, and the interagency Methods and Data Comparability
Board will be provided. This session will bring together the latest updates on who is doing what with PBMS and
when it will happen.
Environmental Business in the PBMS Paradigm Session
Features sub-sessions on Contracting (including frameworks for contracting under PBMS, changes in Superfund
contracting, and a model agreement) and Laboratory Management Issues (including managing labs under PBMS,
laboratory performance expectations, consensus standards roles, and more).
PBMS Implementation Session
Features sub-sessions on Ensuring Scientific and Legal Defensibility (including an overview of EPA's approach,
the Comparable Fuels Rule as a model, and perspectives on quality assurance, private laboratories, legal and
enforcement issues) and Field and Laboratory Implementation Issues (including how to develop DQOs,
developing project-specific MQOs, moving MQOs into commercial laboratories, and balancing error sources for
project planning).
Laboratory Auditing and Accreditation under PBMS
Highlights new roles of auditing and accreditation, laboratory compliance programs, changes in state auditor's
roles, documentation requirements, community issues, and expectations of government laboratories.
iii
-------
IV
-------
CONTENTS
QUALITY ASSURANCE
Paper Page
Number Number
1 Historical Perspective of Performance-Based Measurement Systems (PBMS) at Rocky 3
Mountain Arsenal (RMA). M. Wolf
2 Maintaining Control at a Rapid Response Field Analytical Support Project - A Case Study of 8
Performance-Based Measurement Systems. E.N. Amick
3 Suggestions for Reduction of Analytical Costs by Elimination of Unnecessary Quality Control 11
(QA) Samples. D.M. Chatham
4 Assessment of the Performance of Fourier Transform Infrared Spectroscopy for the 15
Determination of Volatile Organic Compounds in Waste Drum Headspace. W.F. Bauer, C.A.
Crowder, R.E. Evans, T. Dunder
5 Comparison of Laboratory Duplicate, Matrix Spike, and Field Duplicate Results for Mercury in 21
a Large Multi-State Pipeline Investigation. J.G. Head, M.P. Cohen, R.J. Vitale
6 Current Activities in Environmental Standard Reference Materials for Trace Organic 26
Contaminants. S.A. Wise, B.A. Benner Jr., M. Lopez de Alda, R.M. Parris, D.L. Poster, L.C.
Sander, M.M. Schantz
7 Reactive Sulfide Analysis: A Case Study in Auditing Waste Characterization Methodologies. 32
L.J. Dupes, R.J. Vitale
8 Lessons Learned from Performance Evaluation Studies. R.L. Forman, R.J. Vitale 38
9 A Contaminated Marine Sediment Standard Reference Material: SRM 1944, New York/New 47
Jersey Waterway Sediment. M.M. Schantz, E.S. Beaty, D.A. Becker, R. Demiralp, R.
Greenberg, M. Lopez de Alda, K.E. Murphy, R.M. Parris, D.L. Poster, L.C. Sander, S.A. Wise,
R. Turle, C. Chiu
10 An Application of USEPA's Data Quality Objective Process. K.A. Storne 52
11 New Sampling Device Provides Laboraory Verification - Part 1 - Preliminary Data Provides 59
Some Interesting Possibilities. T. Wayne
INORGANIC ANALYSIS
Paper Page
Number Number
12 Recent Developments in the Determination of Trace Level Perchlorate by Ion 63
Chromatography. P.E. Jackson, D.T. Tsui, H. Okamoto, F. Calovini
13 A Generic Leaching Procedure to Predict Environmental Impact of Reactive Materials Such 66
as Coal Combustion By-Products. D.J. Hassett
14 Effect of Zero Valent Iron on the Extraction of Lead, Zinc and Copper in the TCLP. D.S. 72
Kendall
15 New Developments of Method 7473. T.M. Serapiglia, H.M. Boylan, H.M. Kingston 75
16 Speciation of Mercury in Soil. S.J. Nagourney, B. Buckley, E. Fisher 77
17 Method Development for Speciation Analysis of Mercury and Tin Compounds in Standard 80
Reference Materials Using GC-AED and GC-MS. S. Tutschku, M.M. Schantz, S.A. Wise
-------
Paper
Number
18 A Universal ICP-OES Method for Environmental Analyses. Z.A. Grosser, L. Davidowski, J.
Latino, D. Sears
19 New Technologies for Metals Digestions for Environmental Samples. L. Orr
20 Magnesium Chloride in the Cyanide Distillation. R-K. Smith, J. Neuhaus
21 Decreasing Hydraulic Conductivity Behavior and Regulatory Compliance of Alternative
Hydraulic Barriers: An Exercise in Patience. J.D. Quiroz, T.F. Zimmie
22 PBMS: How Will Implementation Change the Analysis of Environmental Sample by ICP-MS?
R.E. Wolf
23 Determination of Mercury in the Range of 1-100 ng/L Using CV-AAS. M. Leyrer, G.
Schlemmer, Z. Grosser
24 Application of In-Situ Gamma Spectrometry in the Remediation of Radioactively
Contaminated Soil. C. Sutton, J.D. Yesso, R.J. Danahy, T. Cox
25 Effect of Environmental Variables Upon In-Situ Gamma Spectrometry Data. C. Sutton
26 Using Acid Mine Drainage to Detoxify Hexavalent Chromium Leachate Feasibility for Coal
Generated Electric Power. H.M. Kingston, D. Huo, R. Cain
Page
Number
81
86
87
91
91
99
101
107
113
ENVIRONMENTAL BUSINESS IN THE PBMS PARADIGM
Paper
Number
27 The Shell for Analytical Chemistry Requirements for USAGE Projects. C. Groenjes
Page
Number
117
ORGANIC ANALYSIS
Paper
Number
28
29
30
31
32
33
34
35
Questionable Practices in the Organic Laboratory: Part II. J.F. Solsky
Comparison of Five Soil Extraction Techniques for Pesticide and Semivolatile Analysis. R.
McMillin, D. Spencer, D. Gregg, L. Wool
Freezer Storage of Soil Samples Containing Volatile Organic Compounds. A.D. Hewitt
Performance of the Disposable EnCore® Sampler for Storing Soil for Volatile Organic
Analysis. S.S. Sorini, J.F. Schabron
An Easy, Cost-Effective Solution for Sampling Volatile Organic Compounds in Soils. M.J.
Ricker
Recovery of VOCs from Soils With and Without Methanol Preservation. J.H. Phillips, A.D.
Hewitt, J.P Glaser
PAH Separation and Detection by GC/FID: Bringing Method 8100 Into the 90's. D.R. Gere,
A.D. Broske, L. Green, G. Reed
Extraction of Diesel Range Organics (DRO) and Waste Oil Organics (WOO) from Soils and
Sediments; Expanding Method 3545A (Pressurized Fluid Extraction). B.E. Richter
Page
Number
121
125
125
129
134
163
170
170
VI
-------
Paper Pa9e
Number Number
36 The Analysis of Carbamates Using LC/MS. J. Krol, E. Block, M. Young, M. Benvenuti, J. 171
Romano
37 Novel Biosensors for Characterizing Environmental Endocrine Chemicals. O.A. Sadik, S. 176
Benda, M. Masila, F. Van, J. Krautova
38 Theory of Operation and Applications of the Pulsed Flame Photometric Detector (PFPD) for 177
Gas Chromatography. N.A. Kirshen
39 Simultaneous Measurement of Volatile and Semivolatile Compounds: Introducing Methods 182
3511 and 3570. D. Mauro, S. Emsbo-Mattingly
40 A Comparison of Static Headspace and Solid-Phase Microextraction for the Determination of 182
Volatile Organics in Water. N.A. Kirshen, Z. Penton
41 Evaluation of a Vacuum Distiller for Performing Method 8261 Analysis. M. Hiatt 187
42 Method 8261: Using Surrogates to Measure Matrix Effects and Correct Analytical Results. M. 188
Hiatt
43 Application of a Dioxin/Furan Immunaoassay Kit to Field Samples. R.O. Harrison, R.E. 189
Carlson
44 Volatile and Extractable Petroleum Hydrocarbons: A Round Robin Illustrates Essential 189
PBMS Standards. S. Emsbo-Mattingly, J. Fitzgerald
45 Fast and Effiecient Volatiles Analysis by Purge and Trap GC/MS. C.E. Boswell 190
46 A New Approach for Highly Complex Organic Analyses Using Simultaneous Selected Ion 194
and Full Ion Scanning. E.A. LeMoine, A. Patkin
47 Does Chemical lonization Have a Future in the Environmental Laboratory? E.A. LeMoine, A. 199
Patkin, H. Hoberecht
48 The Use of Sulfuric Acid Cleanup Techniques to Minimize Matrix Interferences for the 205
Analysis of Toxaphene in Soils and Sediments. F.J. Carlin Jr., R.J. Vitale
49 The Analysis of Army Chemical Agents: GB, VX, Mustard, and Lewisite in Soil at Rocky 211
Mountain Arsenal. D. Parks
50 Comparison of Sampling Protocols for the Zero Headspace Extractor (ZHE) for TCLP and 216
SPLP. D. Turriff, N. Melberg, C. Reitmeyer, B. Podhola
51 Field Application of a Portable Gas Chromatograph for Groundwater Headspace Sampling. 216
P.J. Ebersold
52 On-Site Determination of Volatile Organic Halides (VOH) in Water by UV-lnduced 221
Colorimetry. D. Chen, T.A. Jackson, D. Shattuck, J. McLean, M. Hines
LAB AUDITING AND ACCREDITATION
Paper Page
Number Number
54 The Role of a Compliance Program and Data Quality Review Procedure Under PBMS. A. 231
Rosecrance
Author Index 237
Notes 239
vii
-------
viii
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
QUALITY
ASSURANCE
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
HISTORICAL PERSPECTIVE OF PERFORMANCE-BASED MEASUREMENT SYSTEMS (PBMS)
AT ROCKY MOUNTAIN ARSENAL (RMA)
Mary K. Wolf
Chemist
Lockheed Martin Systems Support & Training Services, Rocky Mountain Arsenal Bldg. 130,
72nd and Quebec, Commerce City, CO 80022
ABSTRACT
Compliance monitoring under a performance-based measurement systems (PBMS) is an on-going process at the
Rocky Mountain Arsenal (RMA) in Commerce City, CO. RMA is a Superfund site where disposal of industrial and
military chemical wastes in unlined basins over a period of approximately 10 years during and following World War
II resulted in widespread contamination of soil and both surface and ground waters. The United States Army,
along with Shell Oil Company and the U.S. Fish and Wildlife Service, are in the process of remediating RMA. The
remediation effort involves the analysis of various matrices for a wide variety of analytes, some of which are
unique to RMA, and standard analytical methodologies are either not available or are not adequate to fulfill regula-
tory requirements in certain instances. Hence the requirement to develop methods which are specific to the RMA
and are performance-based.
In response to these site specific requirements and utilizing the Army Environmental Agency Guidelines, RMA
developed the RMA Chemical Quality Assurance Plan (CQAP), which addresses all activities from planning to
data verification related to the remediation of RMA. Compliance with the CQAP ensures that data produced are
legally defensible, cost effective, and scientifically sound. A strict proficiency demonstration process for methods is
prescribed by the CQAP to validate both standard and new or unproven methods.
Recently the Environmental Laboratory Advisory Board (ELAB) defined five critical elements for PBMS implemen-
tation. As recommended by ELAB, the data produced by laboratories should be legally defensible, cost effective,
scientifically sound, demonstrate good performance criteria, and achieve regulatory compliance monitoring
requirements. Historically, analogous criteria have been applied to the analytical work performed by laboratories
supporting the RMA remediation effort. ELAB has also recommended essential elements for PBMS implementa-
tion. This presentation discusses the analytical program at RMA, under the Comprehensive Analytical Laboratory
Services (CALS) contract (CALS contractor URS Greiner Woodward Clyde), in the context of these elements.
Utilizing the performance criteria, regulatory development, and analytical methods specific to RMA, the remedia-
tion of RMA has progressed at an accelerated rate.
INTRODUCTION
RMA was established in 1942 during World War II. It is located ten miles northeast of downtown Denver and
occupies 27 square miles. The U.S. Army manufactured military chemical weapons at the Arsenal until the 1960's.
Also, during that time and through the early 1980's chemical weapons were destroyed. Following World War II, in
an effort to increase economic growth in the area, offset costs, and maintain the facilities for national security,
private industry leased the facilities at RMA. One of the manufacturers operating under the lease program was
Julius Hyman and Company which produced pesticides. In 1952, Shell Chemical Company acquired Julius Hyman
and Company and continued to produce pesticides until 1982. Most of RMA was placed on the National Priorities
List (NPL) in 1987. As such, RMA is subject to compliance with CERCLA (Comprehensive Environmental
Response, Compensation, and Liability Act, also know as Superfund).
The Remediation Venture Office (RVO), formed in October 1996 to expedite the implementation of the remedia-
tion, is an innovative triparty arrangement consisting of personnel from the Army, Shell Chemical, and the Fish and
Wildlife Service. Members of the RVO work together to coordinate and provide oversight of the remediation
management based on best value concepts, including but not limited to, quality assurance (QA), health and safety,
regulatory compliance, fiscal monitoring, and community involvement. Today there is no manufacturing, weapons
production, or storage at RMA. As a Superfund site, RMA's only mission is environmental cleanup.
DISCUSSION
The production of the military chemical weapons, pesticides, insecticides, and herbicides generated many waste
streams. These wastes were disposed of using widely accepted practices of the time. Efforts to contain liquid
wastes began soon after the discovery that contaminated groundwater was causing damage to crops north of
RMA in the mid1950s. The Army and Shell Chemical began a systematic investigation into the contamination
problems at RMA. Beginning in 1974, Interim Response Actions (IRA) were implemented to protect offsite human
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
health and the environment from RMA pollution.
The United States Environmental Protection Agency (USEPA) has defined PBMS as "a set of processes wherein
the data quality needs, mandates or limitations of a program or project are specified, and serve as criteria Tor
selecting appropriate methods to meet those needs in a cost-effect manner."1 The unusual matrices and analytes
routinely found at RMA pose unique problems for the regulators and the analytical laboratories. Analyses for
analytes in matrices for which no standard method exists have been required. This has necessitated the modifica-
tion of existing approved standard methods, thus the formation of a PBMS at RMA.
The RMA Chemical Quality Assurance Plan (CQAP) was developed from the Army Environmental Agency Guide-
lines and provides the written guidance for operating the RMA QA program. The purpose of the RMA OUAh- is to
provide for consistent generation of analytical data, establish standard practices which permit mterlaboratory
comparisons of data, establish procedures for demonstrating that analytical systems are in control, and ensure
that the data produced by the laboratories is not only of highest quality, but scientifically and legally defensible.
The remediation efforts at RMA pose unique problems for project site evaluations as they are being defined. The
unusual matrices, along with analytes of raw chemicals, by-products, and break-down products cause unique
problems for project site specifications. The regulatory agencies, along with the RVO, meet and determine the
goals for the remediation effort, the critical health care risks, the analytes of interest, and the reporting limits for
those analytes.
The data quality objectives (DQO) are written detailing a clear objective of the project site evaluation, defining the
most appropriate type of data to collect and the most appropriate conditions from which to collect data, and speci-
fying acceptable levels of decision errors that will be used to establish the quantity and quality of data needed to
support the decision.
The DQOs may include analytes or matrices that may or may not have specific methods available to produce the
required analytical results. The laboratories, after reviewing the project site specified requirements select the
appropriate method, or, if needed, modify an existing method to analyze the samples. The specifications within the
CQAP allow the laboratories the flexibility to use their expertise to modify existing methods to achieve regulatory
compliance.
Laboratory standard operating procedures (SOP) provide specific instructions for the performance-based method
analysis. The SOPs include a summary of the performance-based method with information about the matrix,
analytes, and a short description of the procedure. The application of the method is stated along with tested
concentration range, instrument response, detection/reporting limits, interferences, and analysis rate. Other
aspects that are covered in SOPs are safety considerations, apparatus and reagents. Detailed and specific proce-
dures are stated for the preparation of standards including initial and daily calibration standards, instrument mass
tuning criteria and performance, and the analysis of calibration data. Acceptance criteria for all standards along
with corrective actions if criteria are not met are specified. A description of sample collection and storage condi-
tions is given. Also stated is a detailed procedure of the analytical process, including acceptance criteria for
sample analysis, calculations, and the preparation and analysis of quality control samples. The function of quality
control charts and acceptance criteria for controlling the method is outlined. Finally, references are given on which
the performance-based methods are based. The performance-based method SOPs prescribe strict quality control
(QC) and analytical requirements, ensuring that data generated are legally defensible, scientifically sound, and
meets good performance criteria based on historical laboratory performance.
A capacity/capabilities visit (CCV) is performed by the CALS contractor to determine whether the laboratory will be
able to support RVO with the analyses that are needed. Personnel from URS Greiner Woodward Clyde will visit
the laboratories to inspect the laboratory statement of qualifications, training files, facilities and equipment, data
management systems, analytical capabilities, and SOPs. The program and contract requirements of RMA are
discussed in detail during the visit. These requirements include performance audits, performance-based method
proficiency demonstration for analytical methods, participation in the Analytical Laboratory Performance Evalua-
tion System (ALPES), development of performance-based method SOPs, laboratory QA plan, laboratory QC plan
a data management plan, quality systems audits, and attendance at QA meetings as required'
Contract laboratories before performing analytical work, in support of the CALS contract, must demonstrate their
competence in meeting RVO specific QA/QC requirements through a performance-based method proficiency
demonstration. The purpose of performance-based method proficiency demonstration is to- a) establish the lowest
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
concentration at which a result may be reliably reported, b) define the working range of the analytical process, and
c) provide initial performance-based quality control acceptance criteria which will be used to control the analytical
process during sample analysis. Performance-based method proficiency demonstration provides evidence that a
laboratory is able to meet RVO DQOs.
CALS provides to the contract laboratories a reference method (if available), target analyte lists, and the target
reporting limits (TRLs). The TRLs are the reporting limits needed by RVO to support remediation goals. The TRL
information is used by the laboratories during the performance-based method proficiency demonstration to deter-
mine the dynamic concentration range of the method. The performance-based method proficiency demonstration
consists of three parts:
• Instrument calibration
• Preparation and analysis of proficiency samples
• Calculation of method reporting limits (MRL)
INSTRUMENT CALIBRATION
The performance-based method requires an initial calibration, prior to the analysis of samples. The initial calibra-
tion sequence includes, at a minimum, five calibration standards and a zero standard. The standards will bracket
the working range of the measurement system. The acceptability of the initial calibration will be reviewed using
appropriate QC criteria. Upon completion of an acceptable initial calibration, the laboratory proceeds with the
analysis of the proficiency samples.
PREPARATION AND ANALYSIS OF PROFICIENCY SAMPLES
RMA standard matrix, which includes RMA standard soil, standard water (ASTM Type II water, plus 100 milligrams
per liter of sulfate and chloride), or other matrices specific to RMA, must be used during the performance-based
method proficiency demonstration. Spiking solutions are prepared that are independent of the calibration stock
solutions. A minimum of five concentrations of the target analytes is prepared in the RMA standard matrix plus a
preparation blank sample. The concentrations of the target analytes are evenly distributed throughout the dynamic
concentration range. Two sets of performance-based method proficiency samples are prepared and analyzed
according to the specified performance-based method SOP. The proficiency samples are prepared and analyzed
on two separate days to introduce day-to-day laboratory variability.
CALCULATION OF MRLS
After the analysis of the performance-based method proficiency samples, the results of the analysis is evaluated
for the determination of the MRLs. The found concentrations of the target analytes for each spiking concentration,
including the blank sample, is entered into the MRL computer program. The MRL is extracted using confidence
bands as described by Habaux and Vos using 2-tail 90% confidence bands. The software program: a) plots the
found versus target concentration data, b) determines the confidence band about the resultant linear regression
curve, and c) calculates the MRL.
MRLs are the lowest reportable target analyte concentration in a sample using a specific analytical method.
Reporting MRL concentrations as performance-based method target concentrations considers both the measure-
ment precision and the method accuracy. Analyte concentrations in field samples are corrected for method recov-
ery efficiencies determined during the performance-based method proficiency demonstration.
Upon completion of the performance-based method proficiency demonstration, the laboratories will submit the
data to RVO for review. Method proficiency data includes the calibration data, sample preparation, sample analy-
sis, MRL calculations, and certificates of analysis for all reference materials assuring the purity and identification of
all analytes.
Upon method approval, the RVO will provide the laboratory with the a unique method number to be used when
reporting data. The pre-award performance evaluation (PE) sample is shipped to the laboratory ensuring method
performance criteria are met. The laboratory will analyze the pre-award PE sample and submit the data to RVO as
a RVO-required data package. The data package is reviewed and comments submitted to the laboratory. Correc-
tive action, if necessary, is implemented by the laboratory before the laboratory is awarded a contract by the CALS
contractor to perform work for RMA. If necessary, a second pre-award PE sample may be submitted to the
laboratory to demonstrate that the corrective actions have been implemented. If the laboratory fails two pre-award
PE samples a contract will not be awarded.
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
While performing analysis of samples, the laboratories analyze QC checks. These include at;a
blanks laboratory control samples (LCSs), matrix spikes, surrogates, and duplicates (when
results obtained from the QC samples must be evaluated against acceptance criteria per the
performance-based method SOP and historical laboratory QC performance. The results of the QC checks are
included in the electronic data file which is sent to PMRMA with the results of the field sample analysis.
A requirement of each laboratory is to control chart the LCS to demonstrate that the laboratory's process for
sample preparation and analysis is in control. The LCS matrix should be comparable to the sample matrix. RVO
identifies specific controlling analytes (RMA target analytes) contained in the LCS solution that are control charted
for each method. The recoveries of the analytes should be in a state of statistical control. The control charts are
used to monitor the variation of the analytical method and provide a mechanism for the laboratories to detect
out-of-control situations and to improve the analytical method. When an out-of-control situation is observed the
laboratories must investigate the method, determine a cause, and implement corrective action.
The laboratories generating data for RVO prepare data packages that are stand-alone compilations of all data
related to the analysis of a single analytical lot. An analytical lot is defined as the number of samples, including
QC, that can be processed through the rate limiting step of an analytical method. The data packages contain all
information necessary to verify the reported results and to completely document the quality control procedures
utilized during the analysis. Any deviations from the performance-based method SOP must be clearly noted in the
data package. This ensures that the data generated are accurate, defensible, and meets the project site-specific
DQOs.
Information contained in the data packages includes:
• reported sample results and associated MRLs;
• reported QC sample results;
• case narrative that explains deviations during the preparation and analysis of the samples, corrective
actions, manual integrations, and other observations identified and noted during the preparation or analy-
sis of the samples;
• standards preparation, including certificates of analyses of the standards;
• sample preparation and extraction;
• initial and continuing calibration information;
• copies of the chain-of -custodies; and
• quantitation reports and chromatograms of the calibration and sample analysis.
As part of the CALS contract, laboratories submit monthly quality assurance status reports (QASR). These reports
include: QA/QC changes, method changes, personnel changes, facility changes, data quality indicators (including
accuracy, precision, and completeness), revisions of MRLs, and non-conformance occurrences. Each of these
areas discuss, acceptance criteria, out-of-control situations, or modifications performed that relate to RMA
samples. The QASRs are reviewed by the CALS contractor. During the review the CALS contractor, determines if
any out-of-control situations have occurred and if the laboratories have addressed the situations. What caused the
situation and the types of corrective actions taken by the laboratories should be noted in the QASRs by the labora-
tories. The CALS contractor may request additional information concerning the laboratories' corrective actions to
more fully understand and evaluate the situation. If the severity of the situation is warranted. The CALS contractor
may conduct an unannounced audit or may issue a stop work order until all out-of-control issues have been
adequately addressed. This is an on-going performance-based assessment of the laboratories method profi-
ciency, accuracy, and data deliverables.
Audits are an essential part of the PBMS at RMA. The two types of audits performed by the CALS contractor are
quality systems audits and performance audits. Quality systems audits are audits of the operational functions of
the contract laboratories including the QA program. The performance audits monitor the laboratories' ability to
produce accurate analytical measurements of the specific RMA analytes through analysis of PE samples.
Quality systems audits provide RVO a performance-based assessment of the contact laboratories. Quality system
audits are either external or internal (self-assessment). The external quality systems audits, through on-site visits,
verify that the laboratories are complying with the CQAP's QA/QC requirements and determine if the QA/QC
procedures were implemented effectively and suitably to achieve technically sound and defensible analytical data
During the life of the laboratories' contract with the CALS contractor, a minimum of one quality systems audit is
conducted every six months. Additional quality system audits may be performed if there are QAyQC concerns
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
large sample volumes, PE sample results, changes in laboratory management and/or QA program, and/or results
of previous quality systems audits.
During the quality systems audit the CALS audit team will review QA plans and performance-based method SOPs
and verify that previous audit findings have been implemented. Specific data packages, both routine and PE
samples, will be inspected to verify reported results and verify conformance with QA and program requirements.
Interviews will be conducted, if necessary, to clarify concerns, substantiate auditor concerns, or verify the imple-
mentation of corrective actions. A walk-through of the laboratory is performed to evaluate the various areas of the
laboratory. This may include sample receipt, organic preparation and analysis, inorganic preparation and analysis,
data management and review, quality assurance, and training.
A quality systems audit report is prepared by the CALS audit team and submitted to the laboratory detailing the
findings and observations of the audit. The laboratory addresses the findings and observations presented in the
audit report and submits the response to the CALS contractor. RVO reviews the response and determines
whether the laboratory has addressed and implemented corrective actions appropriately.
Internal quality systems audits will be performed by the contract laboratories annually, at a minimum. These audits
are conducted by the QA department in order to assess the PBMS used by the laboratories. Any deficiencies
observed during the internal audits are documented and corrective actions implemented. Documentation of the
quality systems internal audits is retained by the laboratories and is reviewed by the CALS audit team during the
external quality systems audit. The corrective actions of the internal quality systems audit must have been satis-
factorily implemented or the associated deficiencies will become findings during the external quality systems audit.
ALPES, as an independent QA oversight, administers the performance audits for RVO performance-based
methods. These performance audits are conducted semiannually or whenever problems occur. The PE samples
may be prepared for special projects or in batches for distribution to multiple laboratories. The batch is submitted
in the form of samples ready for analysis. Double-blind PE samples may also be submitted to the laboratories to
further monitor the PBMS of the laboratories.
RMA analytes of interest are added to the required matrix to achieve the desired concentration. Matrices used are
RMA standard water, soil, quartz filters or passivated summa canisters, or other special matrices such as
concrete, waste material, or biota. The laboratories are notified of single-blind PE sample shipment and the
expected arrival date. Double-blind PE samples are included in the shipments of field samples. The PE samples
are analyzed in accordance with the laboratories' approved performance-based method SOP A RVO stand-alone
data package is generated and submitted for review. The data are reviewed for accuracy and completeness.
Contract laboratories may participate in performance audits conducted and evaluated by outside organizations
such as the National Institute of Occupational Safety and Health (NIOSH) Proficiency Testing Program or by state
certifications. The laboratories submit to the CALS contractor copies of the PE sample results, the acceptance
criteria, and any corrective action taken to address deficiencies. RVO may, after reviewing the corrective actions,
perform a quality systems audit.
SUMMARY
Compliance monitoring under the PBMS is an integral part of the analytical program at RMA. The PBMS's flexibil-
ity supports RVO's analytical program with methodologies that are scientifically sound, legally defensible, demon-
strate good performance criteria and meet regulatory compliance monitoring requirements.
The remediation effort will transform the former military chemical weapons and pesticides manufacturing facility
into one of the largest urban wildlife refuges in the country.
FOOTNOTES
1. Federal Register, Vol. 62, No. 193, October 6, 1997, Page 52098
ACKNOWLEDGMENTS
The Comprehensive Analytical Laboratory Services support team at the Rocky Mountain Arsenal comprised of the
RVO Support Team, URS Greiner Woodward Clyde, Lockheed Martin, Oak Ridge National Laboratory, and
Roybal Corporation. Northglenn and Westminister, CO Public Works and Utilities Departments.
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
1 Comprehensive Analytical Laboratory Services, Laboratory Quality Control Plan, April 1999.
2 Comprehensive Analytical Laboratory Services, Quality Assurance Management Plan, April 1999.
3^ Environmental Analytical Laboratory, Procurement of an Off-Post Analytical Laboratory in Support of CALS,
SOP 107, Rocky Mountain Arsenal, June 1997.
4 Federal Register, Vol. 62, No. 193, Octobers, 1997, Page 52098
5. Hubaux, A. and Vos, G., "Decision and Detection Limits for Linear Calibration Curves,' Analytical Chemistry,
6 Oak Ridge National Laboratory Analytical Support Group, Office of the Program Manager for Rocky Mountain
Arsenal Analytical Laboratory Performance Evaluation System, SOP ALPES-99-LSD-002, February 1999.
7 Program Manager Rocky Mountain Arsenal, Chemical Quality Assurance Plan, April 1996.
8. RVO Analytical Systems Manager, Standard Operating Procedure for Performing ASM Laboratory Audits,
SOP 2010, Rocky Mountain Arsenal, Draft 1999.
MAINTAINING CONTROL AT A RAPID RESPONSE FIELD ANALYTICAL SUPPORT PROJECT -
A CASE STUDY OF PERFORMANCE-BASED MEASUREMENT SYSTEMS
Edwin Neal Amick
Senior Scientist
Lockheed Martin Services Group, 7411 Beach Drive East, Port Orchard, WA
ABSTRACT
The Lockheed Martin Field Analytical Support Project (FASP) Team routinely uses performance-based analytical
methods to provide rapid results at environmental field sites using mobile laboratories. This work is performed
under the Environmental Services Assistance Team (ESAT) contract to the U.S. Environmental Protection Agency
(EPA) Region 10. This paper describes a case study for the development, validation, and application of a
performance-based analytical field method. An EPA Region 10 removal action project required quick turnaround
data to determine the extent of contamination and confirm removal action. Drinking-water wells in an agricultural
area were contaminated with high concentrations of the herbicide dinoseb. The source of the dinoseb was
adjacent to an agricultural irrigation canal, prompting quick action to avoid additional groundwater contamination.
A performance based analytical procedure for the herbicide dinoseb in soil was developed using gas chromatogra-
phy with electron capture detection. Available EPA methods for dinoseb did not meet the data quality objectives for
the project, or were not practical for use in a mobile laboratory. The primary objective was to provide reliable
analytical data for two action levels of dinoseb in soil (1.6 ug/Kg and 80 ug/Kg). The FASP team developed a
procedure two weeks prior to field deployment. The method used a small quantity of extraction solvent with direct
injection of the extract into a gas chromatograph with an electron capture detector. The method was validated
prior to field deployment, and quality assurance protocols were developed to assure project data quality objectives
were met. Field chemists analyzed a total of 820 soil samples at the field site. Quality control included analyzing
extraction blanks, extraction spikes, and matrix spikes. In addition, investigators shipped 10% of the samples to a
fixed laboratory for comparison analysis. The results of the quality control show the field method produced reliable
data. Overall, performance-based analytical methods for field screening allow for quicker, more cost effective site
investigations and remedial actions. This paper provides guidelines for establishing quality control procedures to
assure generation of data within project data quality objectives.
INTRODUCTION
On October 19, 1998, the FASP Team was tasked by the EPA to determine the feasibility of providing on-site
support for the analysis of the herbicide dinoseb (2-sec-butyl-4,6-dinitrophenol) in soil. The project was to support
an emergency removal project starting on November 2, 1998 in central Washington. The Region 10 FASP Team's
approach to field analysis is to use laboratory-grade instruments in a well-equipped mobile laboratory to provide
data of documented quality. If possible, the FASP Team follows standard operating procedures (SOPs) developed
tor compounds most likely to be encountered in Region 10. These SOPs contain generalized quality assurance
procedures, which may be modified depending upon site-specific data quality requirements. However, no field
bUP existed for dinoseb. A performance-based method for dinoseb was quickly developed with site-specific
quality control procedures incorporated into an SOP before field deployment. This paper describes the analytical
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
method developed, the method validation procedures, quality control procedures, problems encountered, and
corrective actions employed during the project.
METHOD DEVELOPMENT
The first step in preparing the analytical protocol was to obtain the data quality objectives for the removal project.
The objectives were well defined. Two action levels were defined for dinoseb. All soil having dinoseb concentra-
tions above 1.6 mg/Kg had to be removed. To track potential "hot spots," the investigators wanted detection limits
at least ten times lower than the action level, or at 0.16 mg/Kg. The other action level was at 80 mg/Kg. Soil above
this limit had to be segregated from lower level contaminated soil for efficient remediation. A quick analytical data
turnaround was an important objective, with an analytical capacity of at least 30 samples per day. Although false
positive results were not desired, a higher priority was assuring a minimum of false negative results.
Present EPA methods did not meet the data quality objectives of the project, or were not practical for field labora-
tory use. Dinoseb is listed as an analyte for EPA methods 8041 and 8151. Method 8041 is a gas chromatographic
method using a flame ionization detector for soil extracts. However, flame ionization would not meet the detection
levels required for the project. Method 8151 uses a derivatization procedure followed by electron capture gas
chromatography. Although this procedure produces low detection limits, the derivatization method is not practical
for field laboratory use. In addition, fixed-laboratory extraction procedures were not practical for a mobile labora-
tory. A method was developed to determine dinoseb by analyzing soil extracts without derivatization using electron
capture gas chromatography. The extraction procedure selected was the same as used previously for FASP soil
extractions using methyl-tert-butyl ether as an extraction solvent. The extraction method had been validated for
pentachlorophenol but not for dinoseb, so further validation studies were required for this project.
The gas chromatograph was a Hewlett Packard Model 5890 Series II equipped with an electron capture detector.
The column was a 30-meter J&W™ DB5-MS with a 0.53 mm ID and a 1.5 micron film. Helium was used for the
carrier gas at a constant flow of 7.0 milliliters per minute. The initial oven temperature was 100°C for three
minutes, ramped at 12°C per minute to 300°C, then held for 5.0 minutes. The injector temperature was 200°C and
two microliter injections were made in the splitless mode. The detector temperature was 340°C with 5%
methane/argon used as the make-up gas. Soil was extracted in disposable glass culture tubes with PTFE-lined
screw caps. Five grams of soil were extracted twice with five milliliters of methyl tert-butyl ether. Before adding the
solvent, the soil was spiked with a surrogate compound, and acidified with phosphoric acid. The extraction was
performed using a multi-tube vortexer followed by centrifuging to separate the phases.
METHOD VALIDATION
The FASP team follows a set of guidelines for method validation before field deployment, although the specific
validation steps depend upon project goals and historical method performance. As a minimum, method validation
includes verifying instrument response and linearity over the concentration range of interest for the target
compounds. In addition, method extraction blanks must show no interfering compounds and spiked matrix
extracts must demonstrate good recovery of target compounds.
Since the method developed for dinoseb was a new procedure, additional quality assurance validation was
performed. The precision of the method near the detection limit was found by analyzing a series of spiked soils.
These results were compared with a precision study using the EPA method 8151 laboratory procedure. Although a
thorough check for possible interfering compounds could not be done because of time limitations, several chlorin-
ated pesticides and herbicides were found not to interfere. The definitive validation procedure was analyzing
samples collected from the site using the developed method. These samples were also analyzed by a commercial
laboratory. The results of the split samples had to agree before the mobile laboratory deployment.
FIELD QUALITY CONTROL
Quality control protocols used in the field laboratory are generally the same as those used in fixed laboratories,
with the difference that criteria used in the field are not as strict as those used in fixed laboratories. This greater
flexibility allows sample analyses to continue thus avoiding project slowdowns without adversely affecting data
quality objectives.
The initial instrument calibration was performed with a minimum of five calibration levels. The calibration was
successful if concentrations were found within + 25% of the expected value and with a regression correlation
coefficient (r2) greater than 0.995. A calibration verification standard was analyzed once every 12 hours at a level
near the lower action value for dinoseb. The acceptance criterion was the result being within + 35% of expected
value, or a new calibration curve was prepared. Retention time windows for dinoseb standards and spikes were
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
established to be within ± 0.4% of the initial calibration retention time. Matrix spike and matrix spi^ke duplicates
were analyzed once per 20 samples. The acceptance criterion was recovery between 50% and 150/0 and the
Sale percent difference (RPD) less than or equal to 30%. Surrogate spike control limits were set at a recovery
between 50% and 150%. Reagent extraction blanks and spikes were analyzed once per day.
In addition, a minimum of 10% of the samples analyzed on site were shipped to a commercial laboratory for confir-
mation analysis by EPA Method 8151. Duplicate field samples were submitted blindly to the field laboratory.
Quality assurance included a peer review of all documentation, chromatograms, and results before reporting to the
site investigator.
RESULTS AND DISCUSSION
Method validation experiments before field deployment displayed to site investigators that the performance-based
method would meet the data quality objectives of the project. A method detection limit and precision study was
performed by spiking a soil sample with dinoseb at a level near the desired lower detection limit. A series of seven
replicate soil samples spiked at 0.2 mg/Kg was analyzed. The results showed a mean concentration of 0.24
mg/Kg with a standard deviation of 0.018 mg/Kg, or a percent relative standard deviation of 7.3%. The minimum
detection limit was set at 0.10 mg/Kg, which was below the objective of 0.16 mg/Kg. The field method was
compared with EPA method 8151 by extracting a series of seven soil samples spiked at 0.2 mg/Kg, then methy-
lating the extract as specified in the EPA method. Although the methylated dinoseb has a greater response with
sharper peaks, the precision was poorer using Method 8151 (a resulting mean concentration of 0.195 mg/Kg with
a standard deviation of 0.97 mg/Kg). The primary method validation was the comparison of 40 samples from the
site analyzed by the field method compared to split samples sent to a commercial laboratory for quick-turnaround
analysis by Method 8151. Of the 40 samples, 33 samples showed non-detects for dinoseb in each methods. The
seven samples with dinoseb showed good comparison, with an average percent difference of nine percent
between the two methods.
The only problem encountered during the method validation was poor recovery of the surrogate compound. The
surrogate selected for the method was 2,4,6-tribromophenol. This surrogate had successfully been used in previ-
ous projects as a surrogate for pentachlorophenol analyses. However, for the analysis of the 40 soil samples from
the site, nearly all recoveries were less than 50%. It was felt that the problem was due to a matrix effect specific to
the surrogate but not the dinoseb. The dinoseb recoveries from spiked site samples were good, and the confirma-
tion results for the dinoseb agreed with dinoseb results from the field method. Another compound (2,4,5,6-
tetrachloro-m-xylene) was selected as the surrogate, with this surrogate showing good recoveries from the on-site
samples.
After validation results verified the performance-based method would meet the project objectives, the mobile
laboratory was immediately moved to the site location, where 820 samples were analyzed over a six-week period.
Nearly all of the quality control samples were within the acceptance criteria established for the project; however,
one problem developed during the project. Four days into sample analyses, two batches of samples had many
surrogate recoveries below the target value of 50%, 24 of the 62 samples. The problem was believed to be due to
a high concentration of carbonates in the soil matrix. The extraction procedure called for adding acid to the soil
prior to extraction, to assure the dinoseb analyte (a weak acid) was extracted with the organic solvent. Upon
adding acid to some of the soil samples, a vigorous reaction was observed, producing excessive foaming suggest-
ing a high concentration of lime in the matrix. The first step of corrective action was to discuss the problem with
the project managers. The decision was made to continue with the sampling and analyses, with low surrogate
results flagged to assist with selection of samples for confirmation analyses. The other corrective action was to
modify the method, using a weaker buffering solution. The modified method was tested by analyzing six samples
containing high carbonate levels in duplicate, one set analyzed with the original buffer and the second set analyzed
with the weaker buffer. All six samples showed good surrogate recoveries using the modified procedure. The new
buffer solution was incorporated into the procedure, with acceptable surrogate recoveries found afterwards.
CONCLUSION
Performance-based analytical methods are practical and useful means to analyze environmental samples in the
field. Analytical methods can be optimized for the analytes of interest with quality control protocols specific for
project data quality objectives. For this project, quality control included continuing calibration levels and matrix
spme levels near the removal action defined for the cleanup, providing on-site investigators estimates of analytical
precision and accuracy. Verification of performance-based method results with laboratory-based methods is an
important component of quality assurance. Typically, 10 percent of the field samples are shipped to an off-site
10
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
laboratory for analysis. Confirmation results not only verily the results of the field project, but can be used to direct
method improvement.
SUGGESTIONS FOR REDUCTION OF ANALYTICAL COSTS BY ELIMINATION OF
UNNECESSARY QUALITY CONTROL (QC) SAMPLES
Douglas M. Chatham
5413 Forest Ridge Dr., Loganville, GA 30052-3437
Environmental Specialist and QC Chemist
J.M. Waller Associates, Inc. under contract to the U.S. Army Reserve Command,
1401 Deshler Street, SW, Fort McPherson, GA 30330-2000
Web page: http://chatham.home.mindspring.com; e-mail: chatham@mindspring.com
ABSTRACT
A rationale is presented for collecting fewer QC samples for hazardous-waste projects and to encourage project
managers and quality assurance project officers to question the need for every QC sample or activity. Many field
QC samples can be eliminated from hazardous waste site investigations, resulting in significant analytical cost
savings, without any effect on the quality of the overall investigation. The categories of QC samples or analyses
which could be reduced include second column confirmations, field blanks, matrix spike and matrix spike
duplicates, and duplicate samples. Additional cost reductions could be realized through careful selection of
analytical methods and the use of on-site methods, where feasible.
INTRODUCTION
Many field QC samples can be eliminated from hazardous waste site investigations with no effect on the quality of
the overall investigation. The QA/QC requirements for environmental investigations were derived under CERCLA
and RCRA with the purpose of generating legally defensible results. "The EPA Contract Laboratory Program
(CLP) is intended to provide analytical services for Superfund waste site samples. As discussed in the User's
Guide to the Contract Laboratory Program (EPA 1988), the program was developed to fill the need for legally
defensible results supported by a high level of quality assurance (i.e., data of known quality) and documentation."1
All analyses performed for CERCLA (Superfund) investigations were initially required to be conducted at DQO
Level IV (CLP). The initial (discovery) stage of a site investigation should be conducted at Level III or IV. Once the
origin and responsibilities are established for a site, the purpose of QA/QC should be adjusted to new DQOs.
Determining the extent of contamination, conducting RI/FS, and monitoring remediations may be successfully
accomplished with field screening methods, on-site Level II analyses, and fixed laboratory Level II, with some
samples (generally 10%) confirmed at Level III or Level IV.
Significant analytical cost reductions could be realized by eliminating unnecessary second column confirmations,
field blanks, matrix spike and matrix spike duplicates, and duplicate samples. Second column confirmations, field
blanks, matrix spikes/spike duplicates, and field duplicates can, in most cases, be reduced or eliminated. This
should result in a reduction in the number of QC samples, a better understanding of the effect of QC on the data,
and reduced costs in time and money for the work.
QUALITY CONTROL SAMPLES AND PROCEDURES
SECOND-COLUMN CONFIRMATIONS
Second column confirmations apply to organic analyses using GC methods, such as SW846 methods SW8010,
SW8020, SW8021, SW8080, SW8081 and SW8280. A second column confirmation often is billed by the
laboratory as a separate sample analysis. Method 8000A of SW846 states in Paragraph 7.6.9.1 "Tentative
identification of an analyte occurs when a peak from a sample extract falls within the daily retention time window.
Normally, confirmation is required: on a second GC column, by GC/MS if concentration permits, or by other
recognized confirmation techniques. Confirmation may not be necessary if the composition of the sample matrix is
well established by prior analyses."2 Methods SW8010B, SW8011, SW8015A, SW8020B, SW8021A, and
SW8030A, include the statement "if analytical interferences are suspected, or for the purpose of confirmation,
analysis using the second GC column is recommended."
11
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Many projects have specified that all positive results for GC methods will be confirmed by second column
confirmation only because the SW846 method provides for it. Many more projects have suffered from inflated
analytical costs because second column confirmations were not discussed in the work plan or the QAPP and the
laboratory performed these analyses because they were called for by the SW846 method. The large number of
confirmations resulting from this protocol is excessive and often results in an unnecessary inflation of the
analytical cost If good historical data exists, the only analytes requiring confirmation are compounds not
previously detected and confirmed. For example, if benzene was detected and confirmed by method SW8240 (a
GC/MS method) or by method SW8020 with second column confirmation during Superfund investigations, a
positive result for benzene in the RI/FS investigation does not need to be confirmed. Positive results less than
Quantitation Limits, MCLs, ARARs, or cleanup levels should not be confirmed. Sampling efforts involving
numerous samples at each site, e.g. grid sampling, should include only enough confirmations to confirm the
identity of each analyte found at the site.
BLANKS
The field blanks collected at a site could include trip blanks, ambient blanks, bottle blanks, source water blanks,
and equipment rinseate blanks. The reason for analyzing different types of blanks is to be able to trace the origin
of contamination in order to take corrective action. This requires that the results be available as field work is being
conducted. Generally, blank results are not available before the sample results are reported, which can be many
weeks after the field effort is completed. A multiplicity of blanks may be justified, but the project manager should
develop good reasons for them. Long-term programs involving numerous separate projects could benefit from
different types of blanks, since corrective action can be taken between projects. If on-site analytical equipment is
available, analysis of blanks on-site would allow corrective action to be taken rapidly and these are generally much
less expensive than fixed-base laboratory analyses. On-site analysis of blanks must be conducted with methods
which are analyte-specific, have quantitation limits lower than the action levels, and documented calibrations and
detection limits. Many of the blanks submitted to laboratories for analysis are probably not necessary.
In many cases, two or more blanks could be combined; e.g., an equipment rinseate blank taken to the sampling
site serves as an ambient blank and a bottle blank, and if this blank is shipped in a cooler with VOA analyses, it
also serves as a trip blank. Another approach might be to collect a full set of field blanks and analyze only the
most comprehensive (the equipment rinsate). As stated by Dr. Keith3, "Sample analysis is often expensive.
Sometimes it is prudent to collect a full suite of blanks but only analyze the field blanks. If the field blanks indicate
no problems, the other blanks may be discarded or stored as necessary. If a problem is discovered, the individual
blanks can be analyzed to determine its source. Resampling will still likely be necessary."
Data validation guidelines state that if a compound is found in any blank, positive sample results greater than the
quantitation limit and less than five times the blank concentration are qualified as not detected (U or ND) at a
quantitation limit (QL) equal to the sample result. If this adjusted QL is above the action level, it cannot be used to
demonstrate a concentration below the action level. There is no difference between a positive sample result
greater than an action level and a blank qualified result with a quantitation limit greater than the action level when
the purpose is to demonstrate a concentration below the action level. Thus, if the purpose of sampling is to
demonstrate that ARARs, MCLs, or cleanup levels have been met, or for monitoring remediation efforts, there
may be no reason to take any field blanks. Since the resulting corrective action (i.e., resampling) based on a
sample result above the action level is the same with or without blanks, the blanks are not necessary.
MATRIX SPIKE/MATRIX SPIKE DUPLICATES
It has been estimated that up to 90 percent of all environmental measurement variability can be attributed to the
sampling process.6 The matrix spiking protocol assumes that one sample out of a batch of twenty is adequate to
assess the effect of the matrix on accuracy and precision. Much of the variability of the sampling process is due to
the variability of environmental media and the contaminants within that media; likewise, the matrix effect is as
variable as each medium and its contaminants. To be effective in defining method accuracy and precision matrix
spiking would have to be done for all samples.
Since data validation based on MS/MSD results is applied only to the sample spiked, the QA/QC value of MS/MSD
samp es is much lower than the value of surrogate recoveries and of laboratory control sample/laboratory control
sample duplicate results (LCS/LCSD). Surrogates are added to every sample analyzed for organics and are the
best measure of accuracy and matrix effects for an individual sample. LCS/LCSD results for each batch and the
re/. r^n°0ntro1 ChartS ** {^e best measure of laboratory accuracy and precision for organic analyses The
LUb/LUbU program is also the best measure of accuracy and precision for metals analyses Laboratories do not
charge for surrogates or LCS samples. The digestion procedure for metals virtually destroys the matrix so that the
12
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
only interferences normally encountered in ICP and atomic absorption methods are from high concentrations of
other metals. Elimination of MS/MSD samples could reduce analytical costs by 10%. For a project with analytical
costs of $50,000, this represents a savings of $5,000.
FIELD DUPLICATES
The two types of Field Duplicates are split samples and co-located samples. A split sample is a sample which has
been thoroughly blended and split between two containers. Often, the split samples are sent to different
laboratories. Split samples are intended to measure the precision of the whole sampling and analysis procedure.
Most often, if they contain anything to measure, split samples are a measure of how thoroughly the sample was
blended before being split. There is no way to determine an effect on the rest of the samples at the site.
Co-located samples are samples taken in the same location but not blended. The intent of co-located samples is
to measure sampling precision or the variability of the matrix.
"When designing experiments or procedures, it is important to keep in mind that the overall objective is
accuracy. It naturally follows that those in charge of a project should ask whether additional measurements
really contribute to the accuracy of a method, or simply to its precision.
In today's business world cost is very important, and each extra measurement adds to the cost of a project.
We all know that precision is important, but we need to take a closer look at the costs and benefits to the
customer when expenses are increased for the sake of improving precision without necessarily increasing
accuracy."7
Often, the stated purpose of field duplicates is to measure the precision of the complete process from sampling
through analysis. This is nice-sounding phraseology in a work plan, but what can you do with the results? Due to
the potentially large variability inherent in the media being sampled particularly for soils and sediments, one
sample location out of twenty probably will not represent the sampling or matrix variability. The result is that these
measurements are often reported as measures of "precision", but they have no effect on the flagging or the use of
the data. As stated above, the source of the greatest variation in environmental analytical results is the variability
of the media. Comparable results (<40% RPD) are seldom achieved from co-located duplicate soil samples, even
with the best efforts of the best sampling technicians available. A statistical evaluation of all sample results at a
site should be used to measure the precision and representativeness of the sampling program. These statistical
measures may provide confidence intervals for establishing extent of contamination in a medium.
SUMMARY
Since the purpose of this paper is to encourage the use of performance-based criteria to the selection of QC
samples, the recommended guidelines listed in this section should not be used as a prescriptive set of guidelines.
Any and all QC which contributes to the quality of the data or are required for other reasons should be included
regardless of arguments presented in this paper. For each QC sample or analysis proposed, Project Managers
(PMs) and Quality Assurance Project Officers (QAPOs) should ask what that determination contributes to the
quality of the data and whether it helps meet the project DQOs. If a QC sample contributes nothing toward the
DQOs, an argument should be made against incurring the cost for that sample.
RECOMMENDATIONS
The following are recommended guidelines and uses for QA/QC samples:
Second-Column Confirmations
1. If historical data exist, the laboratory should be directed to conduct second-column confirmations only for
compounds not previously detected. When second-column confirmations are deemed necessary, the
laboratory should confer with the PM or the QAPO.
2. Positive results less than Quantitation Limits, MCLs, ARARs, or cleanup levels should not be confirmed.
3 . Sampling efforts involving numerous samples at each site, e.g. grid sampling, should have a limited
number of confirmations.
Blanks
1. For sampling efforts undertaken to demonstrate that ARARs, MCLs, or cleanup levels have been met,
eliminate all field blanks.
13
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
2. For projects which require blanks, use the following criteria for determining the frequency and type of
blanks to take:
1.Ambient blanks - Collect only in the event that the field team observes nearby activities that could
contaminate VOC samples.
2.Equipment blanks - Collect rinseates on bailers used to collect groundwater samples. Collect equipment
rinseates for each decontamination event. Do not collect rinseate blanks for soil or sediment samples.
3. Combine blanks (Equipment Rinseate, Ambient, and Trip Blanks) wherever possible. When equipment
rinseate or ambient blanks are taken, eliminate trip blanks and ship all sample VOCs in the same cooler
as the blank.
4. If sampling of multiple types of blanks cannot be avoided, analyze only the equipment rinseate. If a
problem is found, then analyze the remainder of the blanks.
5. If corrective actions are possible, submit source blanks as needed to implement those corrective actions.
During long-term programs, submit source water blanks from water purification systems either to a fixed
base laboratory or to an on-site chemist to maintain quality control of that system.
Matrix Spike/Matrix Spike Duplicates
1. Use surrogate recoveries to measure matrix effects for organic analyses.
2. Use Laboratory Control Spikes/Duplicates (LCS/LCSD) rather than MS/MSDs for determining precision
and accuracy.
3. Use control charts for warning and control limits on precision and accuracy.
4. Avoid MS/MSD for metal analyses; metal analyses do not generally require a measure of matrix effects
since the digestion and analytical methods destroy the matrix.
Field Duplicates
1. Collect and analyze field duplicates for Level IV (CLP) projects only. Eliminate or greatly reduce the
requirements for field duplicates for Levels I, II, and III projects, unless it is necessary to establish
statistical measures of uncertainty in the definition of extent of contamination.
LIST OF REFERENCES
1. EPA, Risk Assessment Guidance for Superfund, Volume I, Human Health Evaluation Manual (PartA), p. 5-5,
EPA/540/1-89/002, December 1989.
2. EPA, Test Methods for Evaluating Solid Waste, Physical/Chemical Methods, SW846, 3rd Edition, Final Update
1, November 1990.
3. Keith, Lawrence H., Ph.D., Environmental Sampling and Analysis, A Practical Guide, Lewis Publishers Inc ,
1992.
4. Taylor, John K., Ph.D., Quality Assurance of Chemical Measurements, Lewis Publishers, Inc., 1990.
5. USEPA Contract Laboratory Program, National Functional Guidelines for Organic Data Review, June, 1991.
6. Homsher, M.T.; Haeberer, Fred; Marsden, Paul J.; Mitchum, R.K.; Neptune, Dean; and' Warren, John,
Performance Based Criteria, A Panel Discussion, Environmental Lab, October/November 1991.
7. Phifer, Lyle H., Accuracy Versus Precision, Environmental Testing & Analysis, March/April 1995.
8. EPA, Data Quality Objectives Process for Superfund, Interim Final Guidance p 42 EPA540-R-93-071
September 1993.
14
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
ASSESSMENT OF THE PERFORMANCE OF FOURIER TRANSFORM INFRARED SPECTROSCOPY FOR
THE DETERMINATION OF VOLATILE ORGANIC COMPOUNDS IN WASTE DRUM HEADSPACE
William F. Bauer1. Catherine A. Crowder1, Robert E Evans1 and Thomas Dunder2
11daho National Engineering and Environmental Laboratory, 2525 Fremont, Idaho Falls, ID 83405-2208
ph: 208-526-1180, wlb@inel.gov
2Entropy, Inc. P.O. Box 90067, Raleigh, NC 27675-0067
ph: 800-486-3550.
ABSTRACT
Since the 1970's, the Department of Energy (DOE) has retrievably stored transuranic (TRU) radioactive wastes in
drums. Most of these drums are destined for final disposition in the Waste Isolation Pilot Plant (WIPP) in New
Mexico. Prior to transportation to and acceptance into the WIPP, each drum must meet a set of criteria, one of
which is to demonstrate that a set of volatile organic compounds (VOCs) do not exceed a specified concentration
in the headspace of the waste drum. Because of the large number of drums that must be sampled and the high
cost in time and money associated with sampling each drum and sending the sample to a laboratory for analysis,
a Fourier transform infrared spectroscopic (FTIRS) method was developed to provide near-real-time analysis of
waste drums as they are being processed through various facilities. Specifically, the method quantitatively deter-
mines 29 target VOCs, methane, and 12 other interfering inorganic and organic compounds that have been found
real TRU waste drum headspace. These 42 analytes are quantitatively determined from each sample spectrum
using the method of partial least squares (PLS). A single calibration for each analyte of interest was performed
using a set of 190 spectra and these calibrations transferred to each of the deployed instruments. The implemen-
tations of the FTIRS method have been in accordance with the WIPP Quality Assurance Program Plan (QAPP),
including participation in the Performance Demonstration Program (POP).
Overall, the method has been demonstrated to perform adequately via the POP, control samples, method
performance samples and direct comparison to gas chromatography-mass spectrometry (GC/MS) analyses of
duplicate samples. Precision and accuracy are within the respective ±25% and ±30% precision and accuracy
requirements of the QAPP. The long term precision is ±5%. The FTIRS method consistently achieves an overall
score of 90±5% and a score of 97±3% for a set of 8 critical target compounds in the PDP. Both of these scores
are sufficient to pass the PDP. The major errors encountered are primarily associated with analytes at low concen-
trations in the presence of other analytes or interferences at high concentrations.
INTRODUCTION
For nearly 30 years, the DOE has been retrievably storing TRU wastes in metal drums at its various facilities.
These drums are destined for final disposal at the WIPP in New Mexico. Before these drums may be sent to the
WIPP it must be determined that they comply with certain transportation requirements and the WIPP Waste
Acceptance Criteria (WAC). The headspace of each drum must be sampled and analyzed for a number of volatile
organic compounds (VOCs) and hydrogen to determine if the drum meets the WAC or these compounds.
Normally, this involves manual sampling the drum headspace with a SUMMA canister, transport of the canister to
a laboratory and subsequent analysis by gas chromatography/mass spectrometry. Due to the large number of
drums that must be sampled, a more cost effective and timely sampling and analysis alternative was needed that
could be used alone or readily interfaced with existing field equipment for drum venting and headspace sampling.
To meet the need for a rapid and cost effective method for at-line waste drum headspace analysis of VOCs, a
method employing Fourier transform infrared spectroscopy (FTIRS) was developed at the Idaho National
Engineering Laboratory (INEEL)1 Since that time, an FTIRS based VOC analysis system has been incorporated
into the Drum Vent Facility (DVF) at INEEL's Radioactive Waste Management Complex (RWMC) and a similar
mobile gas analysis system (MGAS) was fabricated which has been deployed to several field locations.
Fourier transform infrared spectroscopy was selected as a reasonable alternative to at-line GC/MS instrumenta-
tion for several reasons. Very rugged FTIRS instrumentation is commercially available and has been used for
several on-line applications2'3 Analysis times with FTIRS are generally in the range of seconds to minutes while
GC/MS analysis times are typically 10's of minutes. Most VOCs and many other compounds can be analyzed by
FTIRS. Furthermore, like GC/MS, the FTIR spectrum of a sample contains a historical record of the composition
of the sample. As methods improve, or as new analytes or interferences are identified, the FTIR spectra of all the
previous samples can be reanalyzed. The very nature of absorption spectroscopy makes this possible. It also
makes the instrument calibrations/standardizations universal within certain limits, i.e. once a calibration/ standardi-
zation is established it is possible to transfer that calibration to another instrument4
15
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
wall-mounted FTIRS system at the RWMC was supplied by Bomem, IncjQuebec Quebec
Canada) according to the specification supplied by the INEEL. This is a "turn-key FTIR system that is located in
the DVF in a potential contamination area near the containment silo where waste drums are vented^ To minimize
the possibility of damage to the instrument and the possibility of radioactive contamination of the hardware, the
instrument is housed in a NEMA 12 enclosure (48 in. tall x 36 in. wide x 16 in. deep). In the NEMA 12 enclosure, a
Bomem MB 100 series FTIRS is vertically mounted with a small manifold to control the gas handling operations
necessary for the FTIRS analysis. The FTIRS is equipped with a specially designed top plate with an Axiom
Analytical 1 meter pathlength LFT series gas cell and an EG&G Judson microcooler motor cooled MCT detector.
The gas cell has an internal volume of only 50 ml and is essentially a gold plated light pipe with zinc selenide
windows. This cell and the sample handling manifold within the NEMA 12 enclosure are heated to 110°C to
minimize carryover problems and evacuation times between samples. The optical bench is purged with hydrocar-
bon and CO2 free dry air that is vented into the NEMA 12 enclosure to help maintain a slight positive pressure
within the enclosure. Transducers are mounted in the cell to record the temperature and pressure of each sample.
Because of the heat load supplied by the instrumentation, heated gas cell and manifold components, the NEMA
12 enclosure is cooled to -28 C with a closed cycle air conditioner mounted directly to the cabinet. Operation of
the gas handling equipment and the FTIR are controlled via RS422 from a personal computer located in the DVF
control room over 60 feet away. Vacuum is supplied by a direct connection to the DVF facility vacuum manifold.
A complete description of the computerized DVF operations is beyond scope of this paper, however a brief
description of how the FTIRS interacts with this system follows. When the computer controlling the DVF opera-
tions is ready for an FTIRS analysis, it sends an analog trigger signal to the FTIRS computer. The FTIRS
computer then initiates a sequence to evacuate the internal manifold and gas cell up to an external valve on the
main sampling manifold controlled by the DVF computer. When the cell and internal manifold are evacuated, the
FTIRS computer signals the DVF computer to open the valve and allow the sample to flow into the FTIRS cell.
When a stable pressure is reached in the gas cell, the FTIRS sends another analog signal telling the DVF
computer that it has the sample and then begins to collect the first spectrum of 48 coadded scans. This spectrum
is then evaluated using individual PLS methods for each of the analytes. The spectral residuals are used as an
indicator of a potential problem. If the spectral residual is above a preset value for any of the analyte methods,
then the cell pressure is reduced and a second, "diluted" sample spectrum is acquired and analyzed for the
analytes that triggered the dilution. A report is generated which is stored on the hard disk of the FTIRS computer,
sent to the printer and sent to the DVF computer hard disk via a local area network connection. Total analysis time
is 4 to 6 minutes, depending upon the need for dilution.
The MGAS was supplied by Applied Automation, Inc. (Bartlesville, OK) according the specifications supplied by
the INEEL. This system contains a Bomem MB 100 series FTIR with an identical top plate as in the wall mounted
system described above, a VG Gaslab 300 quadrapole residual gas analyzer (RGA) for hydrogen analysis, a gas
handling manifold, and a vacuum pump. These components are housed in a specially designed stainless steel
cart. One compartment on this cart contains the controlling computers and associated electronics. Another
contains the vertically mounted FTIRS. A third contains the vacuum pumps, the RGA, and a cooling fan pulling air
through a HEPA filter and then dispersing it to the various cart compartments. A fourth compartment is an oven
containing the gas-handling manifold. A compartment with a lid that opens up contains the keyboards and
computer screens for the user interface. The oven and transfer lines within the cart are also maintained at 100°C.
Sample collection and analysis proceeds similar to that described for the wall mounted system.
Quantitative analysis methods. The wall-mounted FTIRS at the RWMC and the FTIRS in the MGAS use the same
quantitative analysis methods. These methods are based upon the partial least squares (PLS) algorithm5 and were
generated using Galactic Industries PLSplus add-on package to Grams/386 or Grams/32. Each analyte has its
own PLS method so that the spectral region can be optimized for that analyte. Table 1 lists the target analytes and
the PLS method parameters for each analyte. The calibration set consists of over 190 spectra collected on differ-
ent instruments with different detectors and cells. The first group consisted of pure component spectra of the 30
target analytes, carbon dioxide, ethane and propane collected on a MB 100 FTIRS by Bomem using a 20 cm cell
and a DTGS detector. This set of spectra was later expanded with spectra to represent interfering compounds
actually found in waste drum headspace by FTIRS analysis. These spectra of hydrocarbons >C6, trimethylamine,
nitrous oxide, ammonia, high concentrations of carbon dioxide, carbon monoxide, and methane were collected
using a 20 cm cell and a DTGS detector on the wall-mounted FTIRS at the RWMC. Others of mixed standards
ethanol, isopropanol, perfluorotributylamine, and 1,4-dioxane were collected using a 1 meter cell with an MCT
detector with either the FTIRS at the RWMC or on the MGAS. Artificial spectra containing offsets and sloping lines
16
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 1. Target analytes, PLS methods and QAOs7 for waste drum headspace analysis by FTIRS. Precision and
accuracy QAOs for all analytes are ±25% and ±30%, respectively.
QAOs
Analyte Spectral Region(s)
Acetone
Benzene
Bromoform
1-Butanol
MEK
Carbon Tetrachloride
Chlorobenzene
Chloroform
Cyclohexane
11-Dichloroethane
12-Dichloroethane
11-Dichloroethene
c-12-Dichlorothene
Ethylbenzene
Ethyl Ether
Methanol
Methylene Chloride
Methyl isobutyl ketone
1 122-Tetrachloroethane
Tetrachloroethene
Toluene
111-Trichloroethane
Trichloroethene
Freon 113
1 24-Trimethylbenzene
1 35-Trimethylbenzene
o-Xylene
m-Xylene
p-Xylene
Methane
1262-1160
712-670
1170-1120
1156-900
1240-1130
1820-1670
820-775
1105-988
806-731
1255-1181
2987-2825
1106-950
717-695
750-722
1260-1200
912-825
1169-1049
882-827
830-680
3140-3020
1225-1020
1100-935
784-744
1296-1237
1407-1330
1815-1703
845-742
940-870
775-689
1129-1003
1134-1040
968-920
864-830
1240-996
834-774
860-81 1
773-700
810-730
840-710
3026-3000
1310-1291
Points
212
88
105
266
115
156
95
337
157
155
337
325
47
59
125
181
249
115
156
125
214
343
84
124
81
117
214
147
179
132
196
101
74
507
126
103
152
167
270
55
41
Spectra PLS Factors SECV
191
187
191
191
191
191
191
191
185
191
192
191
191
183
191
191
191
191
191
191
191
191
191
191
191
191
191
191
191
181
19
15
9
26
24
12
21
34
29
21
28
25
14
23
25
24
25
12
21
14
31
17
17
31
13
14
23
21
32
13
3.6
0.6
1.9
7.4
2.0
1.6
2.6
0.6
0.4
4.5
3.5
3.4
0.9
7.7
1.2
1.5
1.9
3.4
2.8
0.7
1.8
2.0
1.7
1.2
3.2
0.9
2.3
2.5
2.8
5.9
R2
0.989
0.999
0.991
0.919
0.994
0.996
0.983
0.999
1.000
0.965
0.994
0.986
0.998
0.868
0.997
0.998
0.995
0.983
0.980
0.999
0.996
0.999
0.998
0.998
0.976
0.998
0.987
0.984
0.981
1.000
MDLa
50
5
5
50
50
5
5
5
5
5
5
5
5
10
5
50
5
50
5
5
5
5
5
5
5
5
5
5
5
500
PRQLb
100
10
10
100
100
10
10
10
10
10
10
10
10
20
10
100
10
100
10
10
10
10
10
10
10
10
10
10
10
1000
3MDL=Method detection limit
bPRQL=Program required quantitation limit
17
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
were also added to the set for simple background factor definition. The frequency regions for each analyte were
selected after evaluating the correlation spectra for that component calculated by a development aid in the
PLSolus package and the actual spectrum of the analyte. An optimum number of factors for each analyte method
were selected from the evaluation of the predicted residual error sum of squares (PRESS) values determined
using the cross-validation procedure in the PLSplus package.
RESULTS AND DISCUSSION
To date the FTIRS method has been applied to the analysis of VOCs in the headspace of over 600 actual waste
drums Prior to performing these analyses, the methodology was demonstrated to meet the WIPP quality assur-
ance objectives (QAOs) outlined in the Quality Assurance Program Plan (QAPP)6'7. The QAOs essentially consist
of ±25% precision, ±30% accuracy, 90% completeness, and the MDLs and PRQLs listed in Table 1. Table 2 lists
the quality control samples used to demonstrate that the QAOs are being met. An on-line batch is defined as a
12-hour period whereas an analytical batch would represent a set of 20 samples.
QC Sample
Method performance samples
Laboratory or on-line duplicates
Laboratory or on-line blanks
Laboratory or on-line control
samples
GC/MS comparison sample
Blind audit samples
Minimum Frequency
Seven initially and four semiannually
One per analytical or on-line batch
One per analytical or on-line batch prior
sample analysis
One per analytical or on-line batch prior to
sample analysis
One per analytical or on-line batch
Controlled by POP Plan
Acceptance Criteria
Meets Table 1 QAOs
RPD = 25%
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 3. Results from the
RWMC FTIRS.
initial analysis of replicate analyses of the method performance samples with the
Concentration (ppmv)
Analyte
Acetone
Benzene
Bromoform
1-Butanol
Methyl Ethyl Ketone
Carbon Tetrachloride
Chlorobenzene
Chloroform
Cyclohexane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
c-1 ,2-Dichloroethene
Ethyl benzene
Ethyl Ether
Methanol
Dichloromethane
Methyl Isobutyl Ketone
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethene
Toluene
1,1,1 -Trichloroethane
Trichloroethene
Freon-1 1 3
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
o-Xylene
m-Xylene
p-Xylene
Methane
Cylinder
AAL13812
ALM46028
ALM46028
ALM47039
AAL13812
ALM49467
ALM50465
AAL13812
ALM50465
AAL13812
ALM49467
ALM50465
ALM46028
ALM46028
ALM49467
AAL7127
ALM50465
ALM49467
ALM49467
ALM50465
AAL13812
AAL7127
ALM46028
ALM50465
ALM50465
AAL13812
ALM49467
AAL13812
ALM50465
ALM50465
True
100.2
10.1
9.9
101.0
100.2
9.9
9.9
9.9
9.7
10.1
9.9
9.9
9.9
19.9
10.2
100.0
9.9
99.5
9.9
9.8
10.1
10.1
10.0
10.2
9.9
10.0
9.9
10.1
9.8
997.0
Mean
98.6
11.3
9.1
89.5
97.3
9.8
9.5
10.0
9.7
8.1
11.7
9.2
10.0
18.0
10.5
98.5
9.3
88.1
12.3
9.8
9.7
9.7
10.1
10.2
9.8
8.9
9.9
9.7
11.5
996.2
ppmv
SD %RSD % Error
0.9
0.1
0.5
3.7
1.2
0.0
0.6
0.1
0.0
0.6
0.6
0.2
0.2
0.5
0.1
3.2
0.1
1.2
0.7
0.2
0.3
0.4
0.2
0.1
0.6
0.7
0.3
0.5
0.7
1.1
1.0
0.6
5.2
4.1
1.2
0.5
6.1
1.3
0.4
7.4
5.0
2.2
2.0
3.0
1.0
3.2
1.5
1.3
5.4
2.3
3.1
4.1
2.2
0.7
5.6
8.1
3.4
5.5
6.3
0.1
-1.5
11.5
-8.5
-11.4
-2.8
-0.6
-4.3
1.2
0.3
-19.1
18.1
-6.6
1.0
-9.7
3.1
-1.5
-6.3
-11.4
24.3
-0.2
-3.0
-4.0
1.0
-0.4
-1.0
-10.8
-0.2
-3.9
16.9
-0.1
DL
2.8
0.2
1.4
11.1
3.6
0.1
1.7
0.4
0.1
1.8
1.8
0.6
0.6
1.6
0.3
9.5
0.4
3.5
2.0
0.7
0.9
1.2
0.7
0.2
1.7
2.2
1.0
1.6
2.2
3.3
remaining VOCs, and methane are scored independently. Both FTIRS instruments have passed multiple POP
cycles with an average CTC score of 97±3% and an overall score of 90±5%. The POP samples are among the
most difficult samples analyzed by the FTIRS method to date. These samples can contain several 10's of
thousands of ppmv of carbon dioxide and may have several thousand ppmv of methane in the same sample along
with 5 to 10 VOC analytes. Not only is having more than 5 analytes unusual, but carbon dioxide and methane at
these high concentrations, and in the same sample is somewhat unusual. An analysis of the data from 231
unvented drums9 indicate that the higher levels of methane and of carbon dioxide usually found in the POP
samples represent only about 10±5% of the drums and there was no correlation between the them. Of the drums
analyzed so far with the FTIRS method, the frequency at which carbon dioxide at these levels occurs appears to
be about right but methane has only rarely been noted, and then only at low concentrations. Carbon dioxide and
methane at these high concentrations may be significant interferences for many analytes at low concentrations in
the same sample In particular, very high concentrations of carbon dioxide can significantly affect the results for
the tarqet aromatic VOCs The result is either a false positive or negative with an elevated detection limit caused
by the high spectral residual forced dilution. A more complete calibration set which includes many of the nonline-
arities which result in these errors may help to minimize this problem. Even with this issue, the performance of the
19
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
FTIRS method in the POP is quite good.
Table 4. Summary of 61 measurements of the same on-line control sample by the MGAS FTIRS in May, October
Concentration (ppmv)
Analyte
Acetone
Carbon Tetrachloride
Chloroform
1,1-Dichloroethane
Dichloromethane
Toluene
1,1,1-Trichloroethane
Trichloroethene
Freon-113
Methane
True
101.0
100.0
100.0
99.2
98.9
100.0
199.0
99.2
100.0
1000.0
Mean
100.8
101.8
100.3
96.7
96.9
101.6
200.0
100.4
97.5
1024.0
SD
5.6
4.7
4.7
4.2
4.3
5.1
9.7
5.0
4.2
46.0
%RSD
5.6
4.6
4.7
4.3
4.4
5.0
4.9
5.0
4.3
4.5
% Error
-0.2
1.8
0.3
-2.5
-2.0
1.6
0.5
1.2
-2.5
2.4
An additional requirement of the method is that there is some method by which unexpected compounds can be
identified. One of the additional advantages of this method is that as new compounds are encountered, they
generally show up as interferences causing an increase in the spectral residual. Evaluation of the spectra will
reveal the infrared spectrum of the unknown compound that can be identified by comparison to a spectral library
or by manual interpretation. In this way, 1,4,-dioxane, ethanol, isopropanol, trimethylamine, ammonia, nitrous
oxide, and perfluorotributylamine have been identified. More recently acetylene, ethylene, acetaldehyde and
carbonyl sulfide have been identified. Once appropriate spectra of the new compound are added to the calibration
set and new calibration methods are generated, they are used to reanalyze the spectra from all previous suspect
samples.
SUMMARY
The determination of VOCs and methane in TRU waste drum headspace with FTIRS has been demonstrated to
be generally a fast and reliable method. Accuracy and precision are well within the QAOs defined in the WIPP
QAPP. The ability of the method to meet the PRQL and MDL for each analyte has also been demonstrated. The
major problems encountered with the method occur when very high concentrations of a compound or interference
are present in the sample. This forces the need for a dilution and subsequently raises the detection limits for the
sample and causes inaccuracies due to high spectral residuals. Unknown compounds can be identified when
encountered, the calibrations adjusted accordingly and all previous samples can be reanalyzed.
ACKNOWLEDGMENTS
Work supported by the U.S. Department of Energy, Assistant Secretary for Environmental Management, under
DOE Idaho Operations Office Contract DE-AC07-94ID13223.
REFERENCES
1. Bauer, W.F.; Connolly, M.J.; Rilling, A.; Gravel, D.; Perry, S., Evaluation of Fourier transform infrared spectros-
copy for the determination of volatile organic compounds in transuranic waste drum headspace, Idaho
National Engineering Laboratory, INEL-95/-332, September 7, 1995.
2. Cronin, J.T. Spectroscopy 1992, 7, 33-39.
3. Russworm, G.M.; Kagann, R.H.; Simpson, O.A.; McClenny, W.A.; Hergot, W.F Journal of the Air and Waste
Management Association 1991, 41, 1062-1066.
4. Wang, Y ; Lysaght, M.J.; Kowalski, B.R.K. Analytical Chemistry 1992, 64, 562-564.
5. Haaland, D.M.; Thomas, E.V.T. Analytical Chemistry 1988, 60, 1193-1201.
6. USDOE, Transuranic Waste Characterization Quality Assurance Program Plan USDOE-Carlsbad Area
Office, CAO-94-1010, April 30, 1995.
7. USDOE, Transuranic Waste Characterization Quality Assurance Program Plan, US DOE Carlsbad Area
Office, Interim Guidance, February, 1996.
8. USDOE, Performance Demonstration Program Plan for the Analysis of Simulated Headspace Gases for the
TRU Waste Characterization Program, USDOE-Carlsbad Area Office, CAO-95-1077, June, 1995.
20
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
9. Clements, T.L.J.; Kudera, D.E., TRU Waste Sampling Program: Volume 1-Waste Characterization, Idaho
National Engineering Laboratory, EGG-WM-6503, September, 1985.
COMPARISON OF LABORATORY DUPLICATE, MATRIX SPIKE, AND FIELD DUPLICATE RESULTS FOR
MERCURY IN A LARGE MULTI-STATE PIPELINE INVESTIGATION
Joseph G. Head
Quality Assurance Chemist
Martin P Cohen
Senior Programmer
RockJ. Vitale, CPC
Technical Director of Chemistry/Principal
Environmental Standards, Inc., 1140 Valley Forge Road, P.O. Box 810, Valley Forge, PA 19482-0810
ABSTRACT
The collection, preparation, and analysis of laboratory duplicate, matrix spike, and field duplicate samples have
historically been considered important quality control measures to be used by the laboratory in the performance of
analyses for various environmental investigations. For an on-going pipeline investigation, involving the collection of
thousands of samples across six states, a significant number of laboratory duplicate, matrix spike, and field dupli-
cate samples have been collected by several sampling consultants for the characterization of mercury. The analy-
ses of field duplicates as well as the preparation and analysis of laboratory duplicates and matrix spikes for the
subject investigation have been performed by several commercial environmental laboratories. A formal description
of the comparative study of the statistical trends observed among the results of the field-prepared duplicate
samples, the laboratory-prepared duplicate samples, and the laboratory-prepared matrix spike samples for
mercury will be presented.
INTRODUCTION
As data quality indicators, laboratory duplicate, matrix spike, and field duplicate samples provide information to
data users relative to analytical precision, field sample collection precision, and perhaps, to a lesser extent,
sample representativeness. These quality control measures, when used in conjunction, can provide valuable infor-
mation specifically pertaining to how the analytical method may or may not be working for sample analysis. The
careful collection, preparation, and analysis of meaningful laboratory duplicate, matrix spike, and field duplicate
samples have historically been a challenge for environmental investigators.
Typically, investigators routinely collect a sufficient sample mass/volume, thoroughly homogenize the sample in
the field, and place the sample in laboratory-supplied bottleware for shipment to the laboratory for analysis. In the
case of single-blind field duplicates, a larger aliquot of sample form one location is homogenized; the aliquot is
subsequently split between two separate sets of fictitiously labeled bottles for shipment to the laboratory for analy-
sis. The preparation of laboratory duplicate and matrix spike samples may involve homogenizing the received
sample (to varying degrees) and subsequently splitting the sample into separate aliquots for the preparation and
analysis of background, laboratory duplicate, and matrix spike samples.
The sample data utilized for this study was collected as part of on-going natural gas pipeline investigations.
Approximately 1,100 solid and 30 aqueous sample collection events occurred from August 1996 to January 1999
as part of this pipeline investigation. These sample collection events were performed in six states (Ohio, West
Virginia, Virginia, Maryland, Delaware, and Pennsylvania). The solid samples were mostly shallow borings or
surface samples. The aqueous samples were mostly monitoring well samples.
PROCEDURE
A quality control sample utilized by almost all analytical methods to evaluate the accuracy of the analytical proce-
dure is a matrix spike sample. A matrix spike sample is an aliquot of a matrix (e.g., water or solid) fortified (spiked)
with known quantities of specific analytes and subjected to the entire analytical procedure. The percent recovery
for the matrix spike is calculated using the equation:
21
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
%Recovery = (Matrix Spike Sample Concentration - Sample Concentration)/(Spike Added Concentration) x 100
The recovery of the analyte provides the user with information about the effectiveness of the analytical procedure
in terms of the accuracy in determining the qualitative presence and quantitative concentration of the analyte in the
sample matrix being analyzed. The "true" recovery of the analyte can be inherently impacted by the effectiveness
(precision) of the sample collection, preparation, and analytical procedures, as demonstrated by the results of
laboratory and field duplicate analyses described below. The project-specified matrix spike recovery criteria for
mercury are 75-125% in solid and aqueous matrices.
During the evaluation of the matrix spike analyses, the concentration of the background sample in relation to the
concentration of the spike added was used to determine the usability of the matrix spike recoveries. The national
US EPA data evaluation guidelines indicate that a recovery from a matrix spike analysis is not considered
meaningful if four-times the concentration of spike added is less than the observed concentration in the
background sample. For this study, only matrix spike results where the concentration of the sample was less than
or equal to four-times the concentration of spike added were evaluated.
A quality control sample utilized to assess the precision of the method is the laboratory duplicate sample. A labora-
tory duplicate is a second aliquot of a sample that is treated in the same manner as the original sample in order to
determine the precision of the method (not withstanding any confounding effects of sample homogeneity). The
duplicate precision is expressed as the relative percent difference (RPD) and is calculated by the following
equation:
%RPD = (Sample Concentration - Laboratory Duplicate Concentration)/[(Sample Concentration + Laboratory
Duplicate Concentration)/2] x 100
The concentration of an analyte in the laboratory duplicate sample is compared to the concentration of that analyte
in the original sample to assess the precision between the results. The precision between the laboratory duplicate
results provides the data user with information regarding the homogeneity of the sample matrix and effectiveness
of the sample preparation and analysis procedures.
During the evaluation of the laboratory duplicate analyses, the concentrations of the background and laboratory
duplicate results were used to determine the precision criterion utilized. If both results were greater than or equal
to five-times the sample-specific project required detection limit (PRDL), the project laboratory duplicate precision
criterion for mercury in a solid matrix was the RPD between the results must be less than or equal to 35%. If at
least one of the results was less than five-times the sample-specific PRDL, the project laboratory duplicate preci-
sion criterion for mercury in a solid matrix was the difference between the results must be less than or equal to
twice the sample-specific PRDL. If both results were greater than or equal to five-times the sample-specific PRDL,
the project laboratory duplicate precision criterion for mercury in an aqueous matrix was the RPD between the
results must be less than or equal to 20%. If at least one of the results was less than five-times the sample-
specific PRDL, the project laboratory duplicate precision criterion for mercury in an aqueous matrix was the differ-
ence between the results must be less than or equal to the sample-specific PRDL.
A quality control sample that can be utilized to measure the precision of the field sampling and analytical method is
the field duplicate sample. A field duplicate sample is a sample that is thoroughly homogenized in the field, split
between two sets of bottleware, and submitted to the laboratory as two discrete samples. The duplicate precision
is expressed as the RPD and is calculated by the following equation:
%RPD = (Sample Concentration - Field Duplicate Concentration)/[(Sample Concentration + Field Duplicate
Concentration)/2] x 100
The concentration of an analyte in the field duplicate sample is compared to the concentration of that analyte in
the original sample to assess the precision between the results. The precision between the field duplicate results
provides the data user with information regarding the homogeneity of the sample matrix and effectiveness of the
sample collection and analysis procedures.
During the evaluation of the field duplicate analyses, the concentrations of the background and field duplicate
results were used to determine the precision criterion utilized. If both results were greater than or equal to five-
me"SPeCC
RPn h > PTCt fie'd dUpNcate Predsion criterion for mercurv in a solid matrix was the
RPD between the results must be less than or equal to 35%. If at least one of the results was less than five-times
22
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
the sample-specific PRDL, the project field duplicate precision criterion for mercury in a solid matrix was the differ-
ence between the results must be less than or equal to twice the sample-specific PRDL. If both results were
greater than or equal to five-times the sample-specific PRDL, the project field duplicate precision criterion for
mercury in an aqueous matrix was the RPD between the results must be less than or equal to 20%. If at least one
of the results was less than five-times the sample-specific PRDL, the project field duplicate precision criterion for
mercury in an aqueous matrix was the difference between the results must be less than or equal to the sample-
specific PRDL.
RESULTS
All of the matrix spike analyses performed on project samples were initially evaluated. Tables 1 and 2 summarize
the matrix spike recovery data.
Table 1. Solid Matrix Spike Results
Analyte
Mercury
Low Recovery (<75%)
427 (36%)
Acceptable Recovery
505 (43%)
High Recovery (>125%)
254(21%)
Table 2. Aqueous Matrix Spike Results
Analyte
Mercury
Low Recovery (<75%)
0 (0%)
Acceptable Recovery
26 (96%)
High Recovery (>125%)
1 (4%)
The data for the aqueous matrix spike analyses demonstrate that the sample collection, preparation, and analyti-
cal procedures utilized were acceptable for the vast majority of the samples collected for mercury. A small portion
of the aqueous matrix spike analyses displayed unacceptable recoveries indicating possible problems with the
matrix, homogeneity, collection, preparation, and analysis. The matrix spike recoveries are graphically presented
in Figure 1.
All of the laboratory duplicate analyses performed on project samples were initially evaluated. Tables 3 and 4
summarize the laboratory duplicate precision data.
Table 3. Solid Laboratory Duplicate Results
Analyte
Mercury
Acceptable Precision
T022"T86%"J
Unacceptable Precision
164 (14%7
Table 4. Aqueous Laboratory Duplicate Results
Analyte
Mercury
Acceptable Precision
27(100%)
Unacceptable Precision
0 (0%)
The data for the solid and aqueous laboratory duplicate analyses demonstrate that the sample collection, prepara-
tion, and analytical procedures utilized were acceptable for the vast majority of the samples collected for mercury.
A small portion of the solid laboratory duplicate analyses displayed unacceptable recoveries indicating possible
problems with sample homogeneity, collection, preparation, and analysis. The laboratory duplicate results are
graphically presented in Figure 2.
All of the field duplicate analyses performed on project samples were initially evaluated. Tables 5 and 6 summa-
rize the field duplicate precision data.
Table 5. Solid Field Duplicate_Results_
Analyte
Acceptable Precision
Mercury j §57(76%)
Unacceptable Precision
278 (24%)
Table 6. Aqueous Field Duplicate Results
Analyte
Mercury
Acceptable Precision
27 (96%)
Unacceptable Precision
1 (4%)
23
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
The data for the solid and aqueous field duplicate analyses demonstrates that the sample collection and analytical
procedures utilized were acceptable for the majority of the samples collected for mercury. A portion of the solid
field duplicate analyses displayed unacceptable recoveries indicating possible problems with sample homogeneity,
collection, and analysis. The field duplicate results are graphically presented in Figure 3.
In order to determine the most likely cause or causes for matrix spike recovery failures, laboratory duplicate
failures, and field duplicate precision failures, the correlation between the matrix spike recoveries, the laboratory
duplicate precision, and the field duplicate precision was evaluated. The correlation of the matrix spike, laboratory
duplicate, and field duplicate samples required that each of these quality control samples was present in the
sample collection event or sample delivery group (SDG). The project sample collection scheme was not devel-
oped to collect all of the quality control samples with each SDG. Therefore, a limited number of SDGs contained
all of these quality control samples. Tables 7 and 8 summarize the correlation of the quality control data.
Table 7. Solid Results Comparison
Mercury
MS Low
MS Acceptable
MS High
LD In/FD In
220(19.2%)
376 (32.9%)
152(13.3%)
LD In/FD Out
96 (8.4%)
77 (6.7%)
59 (5.2%)
LD Out/FD In
63 (5.5%)
25 (2.2%)
22(1.9%)
LD Out/FD Out
34 (3.0%)
4 (0.3%)
16(1.4%)
T?b'?_§- Aguejoys Results Comparison
Mercury
MS Low
MS Acceptable
MS High
LD In/FD In
0 (0.0%)
25 (96.2%)
1 (3.8%)
LD In/FD Out
0 (0.0%)
0 (0.0%)
0 (0.0%)
LD Out/FD In
0 (0.0%)
0 (0.0%)
0 (0.0%)
LD Out/FD Out
0 (0.0%)
0 (0.0%)
0 (0.0%)
The correlation of the solid quality control results has been graphically presented in Figure 4. The correlation of the
aqueous quality control results is graphically presented in Figure 5.
DISCUSSION
The individual quality control samples provide pieces of information about sample matrix, homogeneity, collection,
preparation, and analysis. In order to garner the most information about the project samples, the data user must
collectively utilize the information generated by the analyses of the quality control samples. The correlation of the
quality control samples placed each SDG into one of 12 categories. The data user can infer certain information
from each of the 12 categories.
An SDG is placed in Category A if the matrix spike (MS) recovery, laboratory duplicate (LD) precision, and field
duplicate (FD) precision are acceptable (In). In this case, all procedures are acceptable.
An SDG is placed in Category B if the MS recovery is low (Low) and the LD and FD precision are acceptable (In).
In this case, it may be inferred the sample matrix is binding the analyte (inhibiting the digestion procedure from
liberating the analyte for analysis), the sample matrix is inhibiting the instrumental sensitivity or instrument
response, or an error during the matrix spike sample preparation may have occurred.
An SDG is placed in Category C if the MS recovery is high (High) and the LD and FD precision are acceptable (In).
In this case, it may be inferred the sample matrix is potentially positively influencing the instrument sensitivity or an
error during the matrix spike sample preparation may have occurred.
An SDG is placed in Category D if the MS recovery and LD precision are acceptable (In) and the FD precision is
unacceptable Out). In this case, it may be inferred the sample collection procedure did not adequately homoge-
zatfo!! t'^h emP™b miSSi°n '° "* lab°ratOry °r the Sample matrix does not allow for adequate homogeni-
An SDG is placed in Category E if the MS recovery is low (Low), the LD precision is acceptable (In) and the FD
precision ,s unacceptable (Out). In this case, it may be inferred the low matrix spike recovery may be attributed I to
sample homogeneity problems. The inferences previously made for Categories B and DmaTa^o apply
24
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
An SDG is placed in Category F if the MS recovery is high (High), the LD precision is acceptable (In), and the FD
precision is unacceptable (Out). As is the case with Category E, it may be inferred the low matrix spike recovery
may be attributed to sample homogeneity problems. The inferences previously made for Categories C and D may
also apply.
An SDG is placed in Category G if the MS recovery is acceptable (In), the LD precision is unacceptable (Out), and
the FD precision is acceptable (In). In this case, it may be inferred the sample preparation procedure did not
adequately homogenize the sample prior to sample analysis or the sample matrix does not allow for adequate
homogenization of the sample matrix.
An SDG is placed in Category H if the MS recovery is low (Low), the LD precision is unacceptable (Out), and the
FD precision is acceptable (In). As is the case with Categories D and E, it may be inferred the low matrix spike
recovery may be attributed to sample homogeneity problems. The inferences previously made for Categories B
and G may also apply.
An SDG is placed in Category I if the MS recovery is high (High), the LD precision is unacceptable (Out), and the
FD precision is acceptable (In). As is the case with Categories D, E, and H, it may be inferred the low matrix spike
recovery may be attributed to sample homogeneity problems. The inferences previously made for Categories C
and G may also apply.
An SDG is placed in Category J if the MS recovery is acceptable (In), the LD precision is unacceptable (Out), and
the FD precision is unacceptable (Out). In this case, it may be inferred the sample collection and sample prepara-
tion procedures do not adequately homogenize the sample prior to analysis or the sample matrix does not allow
for adequate homogenization of the sample matrix.
An SDG is placed in Category K if the MS recovery is low (Low), the LD precision is unacceptable (Out), and the
FD precision is unacceptable (Out). In this case, it may be inferred the low matrix spike recovery may be attributed
to sample homogeneity problems, the sample collection and sample preparation procedures do not adequately
homogenize the sample prior to analysis, or the sample matrix does not allow for adequate homogenization of the
sample matrix.
An SDG is placed in Category L if the MS recovery is high (High), the LD precision is unacceptable (Out), and the
FD precision is unacceptable (Out). In this case, it may be inferred the high matrix spike recovery may be attrib-
uted to sample homogeneity problems, the sample collection and sample preparation procedures do not
adequately homogenize the sample prior to analysis, or the sample matrix does not allow for adequate homogeni-
zation of the sample matrix.
SUMMARY
The individual quality control analysis provides an indication of the sample collection, preparation, and analysis
procedures as well as the sample matrix. When more than one type of quality control sample is utilized, the data
user gains a better insight into the performance of the procedures used for collection and analysis of the sample
matrix. As demonstrated by the data presented, the vast majority of the aqueous data collected and a portion of
the solid data collected indicate that the procedures utilized were acceptable for the sample matrices collected. In
almost one-half of the solid sample SDGs, matrix effects interfered with analysis. As a result, sample matrices
were not adequately analyzed or the sample collection and analysis procedures were not adequately performed.
Also, in several solid sample SDGs, the sample collection and analysis procedures were not adequately
performed or sample matrices were not adequately analyzed due to sample homogeneity complications.
• %Low]
• %ln
D%High
solid
Figure 1. Matrix Spike Recoveries
aqueous
Matrix
25
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
150 T— -•:
•Acceptable
I Unacceptable
Figure 2. Laboratory Duplicate Results
solid
aqueous
Matrix
Figure 3. Field Duplicate Results
34%
13%
19%
Catrgory A
Category B
D Category C
D Category D
Category E
Category F
Category G
Q Category H
Category I
Category J
D Category K
Category L
Figure 5. Comparison - Mercury in Aqueous Samples
solid
aqueous
Matrix
Figure 4. Comparison - Mercury in Solid Samples
'•CatrgoryA
• Category B
D Category C
D Category D
Category E
Category F
Category G
Q Category H
Category I
Category J
D Category K
Category L
96%
CURRENT ACTIVITIES IN ENVIRONMENTAL STANDARD REFERENCE MATERIALS
FOR TRACE ORGANIC CONTAMINANTS
S_A Wise. B.A. Benner, Jr., M. Lopez de Alda, R.M. Parris, D.L. Poster, L.C. Sander, and M.M. Schantz
Analytical Chemistry Division, National Institute of Standards and Technology (NIST), Gaithersburg, MD 20899
ABSTRACT
In the past five years the National Institute of Standards and Technology (NIST) has issued several new
environmental matrix Standard Reference Materials (SRMs) with certified concentrations of polycyclic aromatic
hydrocarbons, polychlorinated biphenyls, and chlorinated pesticides. These materials include air and diesel
26
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
participate matter, sediments, mussel tissues, and cod liver oil. The certified values for these materials are
presented.
INTRODUCTION
Since 1980 the National Institute of Standards and Technology (NIST) has issued a number of Standard
Reference Materials (SRMs) for use in the determination of organic contaminants in environmental samples1
These SRMs include simple calibration solutions that contain a number of analytes and are useful for calibrating
the measurement system and natural matrix materials that are useful for validating the complete analytical
procedure and providing quality control of routine analyses. The recent natural matrix SRMs are described briefly
below with a summary of the certified values for several of these materials.
RESULTS AND DISCUSSION
The typical mode used for certification of natural matrix SRMs for organic contaminants has been the analysis of
the material using two or more "chemically independent" analytical techniques. The results of these multiple
technique analyses, if in agreement, are used to determine the "certified" concentrations for the measured
analytes. When results are obtained from only one analytical technique, the concentrations are typically reported
as reference values (previously denoted as noncertified values). A summary of the natural environmental matrix
SRMs for organic contaminants that have been issued by NIST during the past five years is provided in Table 1.
The recent natural matrix SRM activities have focused primarily on: (1) updating the certified and reference values
on existing materials (ie., recertification), (2) replacing materials that are no longer available (i.e., renewals), and
(3) producing new matrix materials. Tables 2-5 summarize the certified values for PAHs and PCB congeners in
several of these SRMs. The concentrations listed in Tables are the certified concentrations or reference
concentrations (denoted in parentheses) as determined by statistically combining the results from the different
analytical methods. For each SRM the method used for combining the data and the definition of the associated
uncertainties are given in detail in the Certificate of Analysis.
Several of the SRMs issued during the past 20 years have been reanalyzed (i.e., a new certification of the same
material) to provide certified and reference values for additional analytes. Environmental matrix SRMs need to be
updated or recertified as analytical measurement capabilities improve and/or as the need for more analytes
increases. The SRMs recently recertified are listed in Table 6 and include SRM 1588a, SRM 1939a, SRM 1649a,
and SRM 1650a. The number of certified and reference values for PAHs, PCBs, and pesticides determined in the
original certification are compared to those in the recertification. An excellent example of the need to update and
recertify an existing SRM is SRM 1649, Urban Dust/Organics. SRM 1649, the first particle-based natural matrix
material developed by NIST for organic contaminants, was issued in 1992 with certified concentration values for
only five PAHs and reference concentrations for nine additional PAHs. Since 1982 NIST has developed and
implemented improved analytical methods for the measurement and certification of a significantly greater number
of PAHs, as well as PCBs and pesticides, in environmental matrix SRMs. The recertified air particulate material
was reissued recently as SRM 1649a, Urban Dust, and the updated certificate lists certified values for 22 PAHs,
35 PCB congeners, and 8 chlorinated pesticides, as well as reference values for 22 PAHs, 1 chlorinated pesticide,
17 congeners of 2,3,7,8-polychlorinated dibenzo-p-dioxins and dibenzofurans, 32 inorganic constituents,
mutagenic activity, particle-size characteristics, total organic carbon, total extractable material, and carbon
composition. The certified concentrations for selected PAHs and PCB congeners in SRM 1649a are shown in
Tables 2 and 4, respectively.
Two renewal materials, SRM 1941 a (Organics in Marine Sediment) and SRM 1974a (Organics in Mussel Tissue),
were issued in 1994 and 1995 after the first issue of these materials was depleted after five years. As a
complement to the frozen mussel tissue (SRM 1974a), three freeze-dried mussel tissue materials are available:
SRM 2974, which is a freeze-dried version of the same mussel tissue homogenate used for SRM 1974a; RM
8045, which has similar concentrations of contaminants as SRM 2974; and SRM 2977, which has contaminant
concentrations 2-5 times lower than SRM 2974 (see Tables 3 and 5). A new marine sediment, SRM 1944 (NY/NJ
Waterway Sediment) was recently completed with concentrations of PAHs, PCB congeners, and chlorinated
pesticides that are approximately 10 times higher than in SRM 1941a (see Tables 2 and 4). SRM 1944 will also be
the first NIST SRM with values assigned for selected dibenzo-p-dioxin and dibenzofuran congeners. Two new
diesel particulate-related SRMs are available with certified values for PAHs, i.e., SRM 2975 Diesel Particulate
Matter (Industrial Forklift) and SRM 1975 Diesel Particulate Extract, which is a dichloromethane extract of the
diesel particulate material used in SRM 2975. The certification of a fish tissue material, SRM 1946 (Lake Superior
Fish Tissue) is in progress and will be issued as a frozen tissue homogenate (similar to SRM 1974a and 1945)
with certified values for PCBs, pesticides, and methylmercury.
27
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
REFERENCES
1. S.A. Wise., and M.M. Schantz, in R. Clement and M. Sin (editors), Reference Materials for Environmental
Analysis: Making and Using Them, Lewis Publishers, Boca Raton, FL, 1997, 143-186.
Table 1. Recent NIST Natural Matrix SRMs for the Determination of Organic Contaminants in the Environmental
Samples
SRM
No.
Title
Date Certified Constituents
Issued
Reference (Noncertified)
Constituents
1588a Organics in Cod Liver Oil
1649a Urban Dust
1650a Diesel Particulate Matter
1939a PCBs in River Sediment
1941a Organics in Marine
Sediment
1998a
1998b
1999s
1998d
1994
1944 NY/NJ Waterway Sediment 1999
1945 Organics in Whale Blubber 1994
1974a Organics in Mussel Tissue 1999
1975 Diesel Particulate Extract 1999
2974 Organics in Freeze-Dried 1997
Mussel Tissue
2975 Diesel Particulate Matter 1999
(Industrial Forklisft)
2977 Mussel Tissue (Organic 1999
Contaminants and Trace
Elements)
8045 Mussel Tissue (Raritan Bay, 1999
NJ)
PCBs (24); Pesticides (14)
PAHs (22), PCBs (35);
Pesticides (8)
PAHs(19);Nitro-PAHs(1)
PCBs (20); Pesticides (3)
PAHs (23) PCBs (21);
Pesticides (6)
PAHs (24); PCBs (35);
Pesticides (4);
Trace Elements (9)
PCBs (27); Pesticides (15)
PAHs (15); PCBs (20);
Pesticides (7); Methyl-Hg
PAHs (8)
PAHs (14); PCBs (20)
PAHs (11)
PAHs (14); PCBs (25);
Trace Elements (8)
PAHs (17); PCBs (24);
Pesticides (12)
PCBs (34); Pesticides (3);
PCDDs/PCDFs (7)
PAHs (22); Pesticide (1);
PCDDs/PCDFs (17)
PAHs (25); Nitro-PAHs (3)
PCBs (4)
PAHs (14); PCBs (7);
Pesticides (4);
Trace Elements (27)
PAHs (32);
PCDDs/PCDFs (17);
Pesticides (7);
Trace Elements (20)
PCBs (2); Pesticides (2)
PAHs (18); PCBs (4);
Pesticides (4);
Trace Elements (32)
PAHs (-28); Nitro-PAHs (15)
PAHs (17); PCBs (4);
Pesticides (4);
Trace Elements (32)
PAHs (-25)
PAHs (16);
Trace Elements (8)
PAHs (10)
aOriginally issued in 1989; same material recertified in 1998.
"Originally issued in 1982; same material recertified in 1998.
"Originally issued in 1985; same material recertified in 1999.
"Originally issued in 1990; same material recertified in 1998.
28
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 2. Certified and Reference Concentrations of Selected PAHs in Sediment, Air Particulate, and Diesel
Particulate SRMsa
Naphthalene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo[c]phenathrene
Benz[a]nthracene
Chrysene
Triphenylene
Benzo[6]fluoranthene
Benzo[/]fluoranthene
Benzo[/r]fluoranthene
Benzo[a]fluoranthene
Benzo[e]pyrene
Benzo[a]pyrene
Perylene
Benzo[g/7/]perylene
Indeno [ 1,2,3-cd]pyrene
Dibenz[a,y]anthracene
Dibenz[a,c] anthracene
Dibenz[a,/7] anthracene
Pentaphene
Benzo[6]chrysene
Picene
aAII concentrations are certified
SRM 1941 a
(Mg/kg)
1010 ±140
489 ± 23
184 ± 14
981 ± 78
81 1 ± 24
(80 ± 39)
427 ± 25
380 ± 24
197 ±11
740 + 110
341 ± 22
361 ± 18
118 + 11
553 + 59
628 ± 52
452 ± 58
525 + 67
501 ± 72
74. 3 ±6.8
43.1 ±3.7
73.9 ±9.7
42 ±12
99 ±20
80.0 ± 9.0
values except those
Table 3. Certified and Reference Concentrations of
Naphthalene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benz[a]nthracene
Chrysene
Triphenylene
Benzo[£>]fluoranthene
Benzo[/]f I uora nthene
Benzo|7c]fluoranthene
SRM 1974a
(Mg/kg wet)
2.68 ± 0.50
2.53 + 0.28
0.69 ± 0.20
18.6 ± 1.0
17.26 + 0.74
3.71 + 0.54
5.04 ± 0.26
5.77 ± 0.67
5.28 ± 0.42
(2.33 ±0.20)
2.30 ±0.10
SRM 1944
(mg/kg)
1.65 ±0.31
5.27 ± 0.22
1.77 ±0.33
8.92 ± 0.32
9.70 + 0.42
0.76 + 0.10
4.72 ±0.11
4.86 + 0.10
1.04 + 0.27
3.87 ± 0.42
2.09 ± 0.44
2.30 + 0.20
0.78 ±0.12
3.28 + 0.11
4.30 ±0.13
1.1 7 ±0.24
2.84 ±0.10
2.78 ±0.10
0.500 + 0.044
0.335 ±0.013
0.424 ± 0.069
0.288 ± 0.026
0.63 ±0.10
0.518 ±0.093
in parentheses, which
SRM 1649a
(mg/kg)
4.14 + 0.37
0.432 ± 0.082
6.45 ±0.18
5.29 ±0.25
0.46 ± 0.03
2.21 ± 0.073
3.049 + 0.060
1.357 ±0.054
6.45 + 0.64
(1.5 + 0.4)
1.913 + 0.031
0.409 ± 0.035
3.09 + 0.19
2.509 ± 0.087
0.646 ± 0.075
4.01 +0.91
3.18 + 0.72
0.310 ±0.034
0.200 ± 0.025
0.288 ± 0.023
0.151 ±0.035
0.315 + 0.013
0.426 ± 0.022
are reference values.
SRM 1650a
(Mg/kg)
68.4 ± 8.5
(1.50 ±0.63)
49.9 ±2.7
47.5 ±2.7
2.75 + 0.64
6.33 + 0.77
14.4 ±0.8
11.4 + 1.6
8.81 + 0.60
3.52 + 0.40
2.64 + 0.31
7.44 + 0.53
1.33 + 0.35
(0.16 + 0.04)
6.50 + 0.94
5.62 + 0.53
0.52 + 0.10
0.500 + 0.063
0.890 ±0.21
(0.24 ±0.11)
0.316 + 0.038
0.620 + 0.081
PAHs in Mussel Tissue SRMsa
SRM 2974
(Mg/kg)
(9.63 ±0.61)
22.2 ±2.5
6.1 ±1.7
163.7 ±10.3
151.6 ±8.0
32.5 ± 4.8
44.2 ±2.7
50.7 ±6.1
46.4 ± 4.0
(20.5 ± 1.8)
20.2 ± 1.0
RM 8045
(Mg/kg)
31. 4 ±6.0
73.7 ±7.0
166 ±12
256 ± 21
25.3 ±2.3
59 ±10
63.1 ± 8.8
58 ±15
23.4+1.5
24.1 ±3.4
SRM 2977
(ug/kg)
18.8 ±4.8
35.1 ±3.8
8.1 ±4.2
38.7 ± 1.0
78.9 + 3.5
20.3 ± 0.8
(49 ± 2)
(38 + 1)
11.0 + 0.3
(4.6 ±0.2)
(4 + 1)
29
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 3. (Continued)
Benzo[e]pyrene
Benzo[a]pyrene
Perylene
Benzo[g/i/]perylene
Indeno [ 1,2,3-cc/]pyrene
Anthanthrene
Dibenz[a,/]anthracene
Dibenz[a,c + a,h] anthracene
Dibenz[a,/i] anthracene
Benzo[b]chrysene
Picene
aAII concentrations are certified values except those in parentheses, which are reference values.
Table 4. Certified and Reference Concentrations for Selected PCBs in Sediment, Air Particulate, and Cod Liver Oil
SRMsa
SRM 1974a
(ug/kg wet)
9.56 ± 0.21
1 .780 ± 0.073
0.874 ± 0.030
2.50 ± 0.25
1.62 ±0.32
(0.131 ±0.036)
(0.142 ±0.010)
(0.342 ± 0.022)
(0.182 + 0.016)
SRM 2974
(ug/kg)
84.0 ± 3.2
15.63 ±0.80
7.68 ± 0.35
22.0 ± 2.3
14.2 + 2.8
1.15 + 0.31
(1.247 ±0.084)
(3.00 + 0.22)
(1.60 + 0.16)
RM 8045
(Mg/kg)
89.3 ±6.3
(6.7 ± 2.6)
4.08 ±0.32
19.7 ±4.4
12.2 ±2.9
3.51 M 0.49
2.05 + 0.37
4.50 + 0.45
SRM 2977
(ug/kg)
13.1 ±1.1
8.35 + 0.72
3.50 ± 0.76
9.53 ± 0.43
4.84 + 0.81
2.0 + 0.2
1.41 ±0.19
1.07 + 0.15
2.29 + 0.27
PCB28
PCB31
PCB44
PCB49
PCB52
PCB66
PCB95
PCB87
PCB99
PCB 101/90
PCB 105
PCB 110
PCB 118
PCB 128
PCB 38/1 63/1 64
PCB 149
PCB 151
PCB 153
PCB 156
PCB 170/190
PCB 180
PCB 183
PCB 187/159/182
PCB 194
PCB 206
aAII concentrations are
SRM 1941 a
(ug/kg)
(9.8 ± 3.7)
(6.2 + 2.4)
4.80 ± 0.62
9.5 ±2.1
6.89 + 0.56
6.8 ± 1.4
7.5 + 1.1
6.70 ± 0.37
4. 17 ±0.51
11.0±1.6
3.65 + 0.27
9.47 ± 0.85
10.0 ±1.1
1.87 + 0.32
13.38 ±0.97
9.2 ±1.1
(2.62 ± 0.22)
17.6 + 1.9
0.93 ±0.14
3.00 ± 0.46
5.83 ±0.58
(1.63 ±0.15)
(7.0 ± 2.6)
1.78 ±0.23
3.67 ± 0.87
SRM 1944
(ug/kg)
80.8 ± 2.7
78.7+ 1.6
60.2 ± 2.0
53.0 ± 1.7
79.4 + 2.0
71. 9 ±4.3
65.0 + 8.9
29.9 ±4.3
37.5 ± 2.4
73.4 + 2.5
24.5 ± 1.1
63.5 ±4.7
58.0 ±4.3
8.47 ± 0.28
62.1 ± 3.0
49.7 + 1.2
16.93 ±0.36
74.0 ± 2.9
6.52 ± 0.66
22.6 ±1.4
44.3 ±1.2
12.19 ±0.57
25.1 ± 1.0
11. 2 ±1.4
9.21 ±0.51
SRM 1939a
(ug/kg)
(2461 + 78)
(6440 + 490)
1131 ±74
3740 + 280
4320 ±130
840+ 130
(1210 + 420)
380 ± 96
201 ± 28
1068 ±70
423 + 88
91.2 + 8.48
258.1 ±6.9
427 + 47
192.1 ±2.6
297 ± 19
37.0 + 6.6
107 ±17
140.3 ±6.1
47.3 ±2.3
156.4 + 2.6
35.5 + 4.1
29.7 ± 5.6
certified values except those in parentheses, which are
SRM 1649a
(Mg/kg)
18.5+ 1.2
17.3+ 1.4
15.4+ 1.6
12.2+ 1.5
24.65 ± 0.97
65 + 0.12
51.6 + 4.2
10.65 + 0.62
9.58 ± 0.69
52.9+ 1.0
8.63 ± 0.80
26.6 ± 1.6
25.7+ 1.5
6.35 ± 0.69
69.7 ± 7.5
75.7 ±1.3
34. 3 ±3.9
82.5 + 8.0
16.25 ±0.77
30.8 ± 2.2
78.7 ±8.2
20.34 + 0.95
40.1 ±2.5
28.9 + 3.6
20.6 + 4.6
reference values.
SRM 1588a
(ug/kg)
28.32 + 0.55)
8.33 ± 0.28
35.1 + 1.4
29.90 ± 0.84
83.3 + 2.3
54.7 ± 1.5
63.5+ 1.1
56.3 ± 1.1
126.5 ±4.3
60.2 + 2.3
76.0 ± 2.0
176.3 + 3.8
47.0 + 2.4
263.5 ±9.1
105. 7 ±3.6
54.8 + 2.1
273.8 ± 7.7
27.3+1.8
46.5 ±1.1
105.0 ±5.2
31.21 +0.62
35.23 ± 0.83
15.37 ±0.61
30
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 5. Certified and Reference Concentrations for Selected PCBs in Mussel Tissue and Whale Blubber SRMs3
SRM 1974a SRM 2974 RM 8045 SRM 2977 SRM 1945
(ug/kg wet) (ug/kg) (ug/kg) (ug/kg) (ug/kg)
PCB28
PCB31
PCB44
PCB49
PCB52
PCB66
PCB95
PCB87
PCB99
PCS 101/90
PCB 105
PCB110
PCB 118
PCB 128
PCB 138/1 63/1 64
PCB 149
PCB 151
PCB 153
PCB 156
PCB 170/1 90
PCB 180
PCB 183
PCB 187/159/182
PCB 194
PCB 195
aAII concentrations are
(9.0 ± 1.7)
(8.6 ± 2.4)
8.28 ± 0.84
10.12 + 0.59
13.1 ±1.3
11. 54 ±0.50
9.5 ±1.9
(6.1 ±1.6)
8.08 ± 0.46
14.6 ±1.1
(79 ±15)
(76 ±21)
72.7 ± 7.7
88.8 ± 5.7
115±12
101. 4 ±5.4
83 ±17
(54 ± 14)
70.9 ± 4.56
128 ±10.1
6.04 ± 0.39 53.0 ± 3.8.39
14.5 ±1.0
14.90 ± 0.40
2.50 ± 0.39
15.2±1.1
9.98 ± 0.27
2.91 ± 0.40
16.54 + 0.86
0.85 ±0.11
0.63 + 0.12
1.95 + 0.43
1.82 ±0.27
3.87 + 0.27
certified values except
127.3 + 9.4
130.8 + 5.3
22.0 ± 3.5
134 + 10
87.6 ± 3.5
25.6 ±3.6
145.2 ±8.8
74 + 1.011
5.5 + 1.4
17.1 ±3.83
16.0 + 2.47
34.0 ± 2.57
7.91 ± 0.90
21 .4 ±0.20
11. 8 ±0.64
16.9 ±0.9
17.7 ±2.8
18.4 ±1.5
20.8 ±2.1
10.2 ±0.3
18.85 ±0.44
35.9 ± 1.6
10.9 ±0.5
35.3 ± 0.5
35.1 ±1.0
5.24 + 0.17
35.7+ 1.5
34.7 ± 0.4
10.9 ±0.3
56.9 + 3.5
1.97 + 0.11
2.37 ± 0.56
7.81 ±0.63
5.25 ±0.14
16.9± 1.3
those in parentheses, which are
5.37 ± 0.44
3.92 ± 0.24
3.25 ± 0.63
8.37 ± 0.54
3.65 ± 0.32
5.39 ±0.59
2.15 + 0.08
1.59 + 0.20
11.2+ 1.2
3.76 ± 0.49
4.03 ± 0.20
10.51 ±0.81
2.49 + 0.28
16.6 + 1.6
9.23 + 0.12
14.1 ±1.0
0.96 ± 0.08
2.95 ± 0.23
6.79 ±0.67
1.33 + 0.10
4.76 + 0.38
28.9 ±3.6
9.63 ± 0.37
reference values.
(14.1 ±1.4)
(3.12 ±0.69)
12.2 ±1.4
20.8 + 2.8
43.6 ± 2.5
23.6+ 1.6
33.8 + 1.7
16.7 ±1.4
45.4 ± 5.44
65.2 ±5.6
30.1 ±2.3
23.3 ± 4.0
74.6 ±5.1
23.7 ±1.7
131. 5 ±7.4
106.6 ±8.4
28.7 ± 5.2
213±19
10.3+ 1.1
40.6 + 2.6
106.7 + 5.3
36.6 + 4.1
105.1 ±9.1
39.6 + 2.5
17.7 + 4.3
Table 6. Recertifications and Renewals of Previous Environmental Matrix SRMs
Original Certification
PAHsa
PCBs
Pesticides3
PAHsa
New Certification
PCBs
Pesticides3
0
5(9)
5(6)
-5
5
0
0
3(12)
10
0
0
0
0
22 (22)
19(25)
0
25
35
0
20(4)
14
8(1)
0
3
Recertifications
SRM1588a(1989-1998)b
SRM 1649a( 1982-1998)
SRM 1650a (1985-1999)
SRM 1939a( 1990-1998)
Renewals
SRM1941a(1989-1994)c 11(24) 0(15) 0(7) 23(14) 21(7) 6(4)
SRM 1974a (1990-1995) 9(19) 0(13) 0(12) 15(18) 20(4) 7(4)
3The first number indicates the number of certified constituents; the number in parenthesis indicates the number of
noncertified or reference values.
"The first date indicate the year of the original certification and the second date is the year of the reissue of the
material after recertification.
The first date indicate the year of the original certification and the second date is the year of the issue of the
renewal material.
31
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
REACTIVE SULFIDE ANALYSIS:
A CASE STUDY IN AUDITING WASTE CHARACTERIZATION METHODOLOGIES
Lester J. Dupes. CPC
Senior Quality Assurance Chemist II
RockJ. Vitale, CPC
Technical Director of Chemistry
Environmental Standards, Inc., 1140 Valley Forge Road, Valley Forge, PA 19482
ABSTRACT
The preparation and analysis of waste samples for reactive sulfides is defined in Chapter 7 of SW-846 as a
"method-defined parameter where the analytical result is wholly dependent on the process used to make the
measurement. ... Therefore, when the measurement of such method-defined parameters is required by regulation,
those methods are not subject to the flexibility afforded in other SW-846 methods." Changes made to the analyti-
cal preparation or analysis method may result in improper waste characterization and disposal.
The study involved the evaluation of eight commercial laboratories providing waste stream characterization
support for several industrial clientele. These laboratories received extensive full-day audits which included an
evaluation of both method compliance and the actual step-by-step analyst techniques used for the sample prepa-
ration and analysis of reactive sulfides. Audits of the reactive sulfide methods revealed both significant method
deviations and analyst error. Upon completion of the audits, a single-blind performance evaluation (PE) study was
conducted to determine the accuracy of the laboratory-reported results when compared to known values.
The PE study was conducted for reactive sulfide using both an aqueous phase and a solid sulfide salt as PE
samples. The PE study included a review of the reactive sulfide tests performed by SW-846 Chapter 7 preparation
and analysis by Method 9034. The results of this study exhibited a wide range of reactive sulfide concentrations,
some of which would have represented incorrect waste characterization if the PE samples were actual waste
stream samples.
This paper will focus on a discussion of the laboratory audit results, method compliance issues, analyst technique,
and a review of the PE sample results from these case studies.
INTRODUCTION
In an article in Chemical and Engineering News (July 20, 1998), it was announced that the US EPA is considering
omitting reactivity from the regulatory requirements associated with waste characterization. Significant historical
problems with the analytical methodology was cited as a reason for consideration of this action. The reactivity tests
are currently performed by preparation of waste samples using the reaction procedure detailed in SW-846 Chapter
7. Upon completion of the reaction step, reactive sulfides are analyzed using SW-846 Method 9034 and reactive
cyanides are analyzed using SW-846 Method 9012. Analytical results obtained for reactive sulfides are compared
against the interim guideline of 500 mg/kg for total releasable sulfide. Concentrations of reactive sulfides in a
waste sample that are greater than the interim guideline are considered hazardous, and disposal costs are signifi-
cantly greater than those for wastes with concentrations less than the guidelines which are classified as
non-hazardous.
A study was conducted by the authors that involved the evaluation of eight commercial laboratories providing
waste stream characterization support for several industrial clientele. The first case study involved the remediation
of a sludge basin at an industrial facility. Results of waste samples collected and submitted to several laboratories
resulted in a significant disparity among the reported results. This study involved detailed on-site audits of three
laboratories. The auditors witnessed the preparation and analysis of thoroughly homogenized split sludge samples
at each of the three laboratories and noted significant differences in the method performed and varying degrees of
compliance with the analytical methods. The second case study involved on-site audits of five commercial labora-
tories which are evaluated on a yearly basis as part of a corporate environmental laboratory program for waste
characterization. In addition, these laboratories participated in a single-blind performance evaluation (PE) study
which was performed to determine the accuracy of the laboratory-reported results when compared to known
values.
PRELIMINARY CASE STUDY PREPARATION
Prior to conducting these case studies, detailed step-by-step auditing checklists were created based on a thorough
review of the SW-846 Chapter 7 reaction method and SW-846 Method 9034. Since reactivity is defined in SW-846
32
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Chapter 7 as a "a method-defined parameter," variances and interpretation afforded other SW-846 methodologies
are not permitted. Therefore, these procedures require absolute compliance by the analytical laboratories and the
auditors remained stringent to this requirement during the auditing process.
CASE STUDY NUMBER ONE: SLUDGE BASIN
On a recent project involving the remediation of a sludge basin at a large industrial facility, the authors were
requested to identify the reasons for the significant disparity observed among split sample reactive sulfide results
from several laboratories. For split sludge samples submitted to one laboratory, reactive sulfide results were
consistently greater than 500 mg/kg (up to 2000 mg/kg). For split sludge samples submitted to another laboratory,
reactive sulfide results were consistently less than 30 mg/kg. For the same sludge samples submitted to a third
laboratory, reactive sulfide results were 300-600 mg/kg.
Because of the significant cost ramification of the hazardous classification due to reactive sulfides, it was impor-
tant to identify the reasons for the discrepancies. Through detailed on-site audits of all three laboratories and
witnessing the analysis of thoroughly homogenized split sludge samples at each of the three laboratories, a
number of very interesting observations resulted. The reaction set-up for all three laboratories varied significantly
for the supposedly "method-defined" parameter. The second laboratory, while performing the analysis adequately,
was observed to have a low-bias for sulfide due to poor technique. One laboratory had not obtained a positive
result for reactive sulfide from 1995 until the day of the on-site audit and witnessing of split sample analysis. On
the day of the on-site audit, after implementing changes suggested by the auditor, this laboratory obtained results
of 700-800 mg/kg; these results were comparable to the results obtained by the first laboratory on the same
samples. Once these technique problems were resolved, the third laboratory's reactive sulfide results were
comparable with the results of the other two laboratories.
The following method non-compliance issues were noted during the audits of these facilities.
Laboratory #1
Deviation from SW-846 Chapter 7.3.4
• The laboratory does not use a rotometer to monitor and control 60 mL/min of nitrogen, as stipulated in
Chapter 7 of SW-846.
Laboratory #2
Deviations from SW-846 Chapter 7.3.4
• The laboratory utilizes 50 mL of 2.5N NaOH scrubber solution. SW-846 Chapter 7 stipulates 50 ml of
0.25N NaOH scrubber solution.
• The laboratory does not use a rotometer to monitor and control 60 mL/min of nitrogen, as stipulated in
Chapter 7 of SW-846.
Deviations from SW-846 Method 9034
• For the "Standard Iodine Solution," the laboratory's SOP stipulates dissolving 20 to 25 g of Kl and 3.2 g of
iodine to 1 liter of reagent water. SW-846 Method 9034 stipulates the addition of 10 mL of 6N hydrochloric
acid (HCI) to this reagent. The laboratory did not add HCI to the iodine solution.
• The laboratory utilized the prepared iodine solution in a reagent blank to perform the iodine standardiza-
tion; however, SW-846 Method 9034 (Section 5.6) stipulates a very specific reagent to be prepared for the
iodine solution standardization .
• The laboratory acidified the 100 mL of scrubber solution with 6N HCI and then poured the acidified scrub-
ber on top of the iodine solution. SW-846 Method 9034 stipulates "[pipetting] the gas scrubber
solution...keeping the end of the pipette below the surface of the iodine solution."
Laboratory #3
Deviations from SW-846 Chapter 7.3.4
• The laboratory utilizes 300 mL of 0.25N NaOH scrubber solution. SW-846 Chapter 7 stipulates 50 mL of
0.25N NaOH scrubber solution.
• The laboratory did not use a rotometer to monitor and control 60 mL/min of nitrogen, as stipulated in
Chapter 7 of SW-846.
. The laboratory utilized 500 mL of 0.01N H2SO4 for sulfide reaction. SW-846 Chapter 7 stipulates adding
"enough sulfuric acid to fill the (500 mL boiling) flask half full" (250 mL).
• The laboratory utilized SW-846 Method 9030 as the sulfide determinative step. SW-846, Chapter 7
33
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
stipulates the use of SW-846 Method 9034 exclusively.
Deviations from SW-846 Method 9034
• The analyst stated that the hydrogen sulfide standard solution is typically prepared every two months and
stored in opaque brown plastic bottles at 4°C. SW-846, Method 9034, Section 5.7 states, "These
standards are unstable and should be prepared daily." The analyst also indicated that he does not verify
the hydrogen sulfide standard solution by direct titration techniques, and the true value is assumed to be
the theoretical value. Similarly, the analyst indicated that he does not empirically determine the normality
of the iodine solution; this determination is stipulated as a requirement in both SW-846 and the laboratory
SOP
• For the "Standard Iodine Solution," the laboratory's SOP stipulates dissolving 20 to 25 g of Kl and 3.2 g of
iodine to 1 liter of deionized water. SW-846 Method 9034 stipulates the addition of 10 mL of 6N HCI to this
reagent. The laboratory does not add HCI to the standard iodine solution.
• The laboratory SOP does not include the preparation of the sodium sulfide nonanhydrate stock solution
stipulated in Section 5.7 of SW-846 Method 9034.
• For titration, the laboratory utilized 200 mL of scrubber solution, 20 mL of iodine solution, and 8 mL of 6 N
HCI (total of approximately 228 mL). SW-846 Method 9034 requires that after combining the aforemen-
tioned scrubber solutions and reagents the laboratories should "add enough reagent water to bring the
volume to 100 mL."
• The laboratory poured the scrubber on top of the combined iodine/6N HCI solution. SW-846 stipulates
"[pipetting] the gas scrubber solution...keeping the end of the pipette below the surface of the iodine
solution."
ANALYST TECHNIQUE REVIEW
As previously discussed, method deviations may cause significant variances in obtaining results that are compara-
ble among laboratories and quantitatively accurate based on the prescribed analytical methods. However, the
actual laboratory techniques and procedures used by the analyst (which may not be method-defined) are as
important as method compliance in obtaining quantitative results. Improper techniques can be the cause of signifi-
cant problems when evaluating the comparability and usability of reactive sulfide data. Since these audits were
performed when samples were actually being prepared and analyzed, the authors were able to evaluate each
analyst's techniques in a step-by-step fashion.
The following are notable techniques observed during the audits of these facilities.
• Examination of one of the reaction glassware set-ups revealed a broken acid drop funnel connection
which was Teflon®-taped at the glassware break.
• The analyst proceeded to add 10 mL of reagent water for the method blank and 5 mL of the sulfide spike
solution (for the blank spike) to the respective open boiling flasks. At this point, the sodium hydroxide
scrubber solution and 0.01N sulfuric acid had not been measured and poured into the appropriate
vessels, nor had any glassware connections been made to minimize the time that samples would be
exposed to the ambient air.
• The sample was mostly free liquids and the analyst used a stainless steel spatula to administer approxi-
mately 0.5 g at a time into the boiling flask which was positioned on its side. After achieving a weight of
13.3 g, it was apparent that a great deal of sample was adhering to the side walls of the boiling flask. This
is problematic since the portion of sample coating the boiling flask walls does not directly react with the
0.01N sulfuric acid.
• It was noted that the stirring bars were already rigorously stirring the method blank, the blank spike, and
sample, without any glassware connections having been made. As evident by the smell of sulfide, labora-
tory personnel acknowledged that the stirring bars should not be rotating until the system is closed and the
sulfuric acid is dropped (also as specified in Section 7.5 of Chapter 7).
• The nitrogen gas flow was manually adjusted through the fritted glass connections immersed in the 50-mL
scrubber solutions. (Another of the laboratories maintained on approximate rate of 90-0.25 inch N2 gas
bubble per minute.) Once the flow was adjusted, a vacuum pump was placed on the exit air connection.
The observed pressure resulted in some of the glass connections separating to release pressure build-up
When the glassware joint connections appeared secure, the sulfuric acid was slowly dropped into the
boiling flasks. Once the drop funnels were empty, final adjustments were made to the stirring bar and
nitrogen flow, again resulting in pressure build-up and some of the glass connections separating to
release pressure build-up. The separation resulted in loss of reaction gas for some of the reactions In
addition, one of the gas losses was observed to be from a cracked piece of glassware that clearly had
34
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
observable gas bubbles being released from the crack. The analyst taped the cracked glassware with
Teflon® tape to minimize the continued loss of reaction gas. Finally, it was observed that both samples
coated the side-walls of the boiling flask.
• In the process of coordinating the shut-off of the nitrogen gas and the vacuum applied to the exit valve,
the scrubber solution for one sample backed up into the reaction boiling flask.
• The samples were removed and diluted to volume with reagent water in volumetric flasks. The sealed
flasks were inverted repeatedly to mix the solutions but in the process the scrubber solutions were purged
with the oxygen present in the neck of the flasks. Each of the 500-mL aerated scrubber solutions were
then poured into labeled disposable wide-mouth, 120-mL sludge cups, each with 1/4 1/2 inch of
headspace, capped, and stored at 4°C until titration. The auditors noted that the excessive aeration of
these final solutions caused a loss of sulfides, as evidence by the strong sulfide smell.
Titration of the Scrubber Solutions
• For the direct titration of the sulfide standard solution to empirically determine the "true" value of the
prepared sulfide stock solution (without being processed through the reaction procedure), the analyst
removed 5 mL of the stock standard and diluted this aliquot to 500 ml with reagent water. The analyst
utilized a squirt bottle for this dilution and squirted the reagent with a fast stream directly down the center
of the neck of the volumetric flask and directly into the sulfide standard/reagent water mixture. The fast
stream of reagent water into the solution resulted in the generation of large purging bubbles that formed at
the bottom of the open flask. The Laboratory QA Director also observed this procedure (as well as the
evident smell of sulfides) in the ambient air.
• The analyst performing the titration opened one of the sludge cups containing the reagent blank and
measured 100 ml of the solution using a glass graduated cylinder. The 100-mL of solution was then
poured into a second fresh 120-mL wide-mouth sludge cup and acidified with 6N HCI (the sample is not
pipetted under the surface of the iodine as required in Method 9034, Section 7.3.3).
• For all of the titrations, the rotation of the stirring bar applied by the analyst with the acidified, diluted scrub-
ber solutions imparted a notably significant aerating vortex. For some of the sample titrations, the distinct
smell of sulfide was noticeable prior to the actual titration of the acidified, diluted, vortexing scrubber
solutions. Loss of sulfide gas prior to titration was clearly evident.
• Once each sample to be titrated was placed on the stirring plate and the stirring bar was rigorously
rotated, the analyst added the starch indicator solution. This is not compliant with Method 9034, Section
7.3.6 which requires the titration of the scrubber solutions "...until the amber color fades to yellow. Add
enough starch indicator for the solution to turn dark blue and titrate until the blue disappears."
• Upon addition of the scrubber solution into the initial 3 mL of iodine, it was observed that the iodine was
totally consumed (the scrubber solutions went completely clear). For these samples, an additional 3 mL of
iodine was added directly to the diluted acidified scrubber solutions. The additional iodine aliquots
imparted a medium yellow color to the samples; however, the solutions were still not amber and additional
iodine should have been added. (Method 9034, Section 7.3.3 requires iodine to be added until the amber
color remains.)
ANALYTICAL RESULTS COMPARISON
The analysis for two site samples for reactive sulfides was performed by Laboratories #1 and #2 within a holding
time of 7 days from sample collection. A quantitative comparison of the reported results is as follows:
Field Sample Designation Laboratory #1 Results Laboratory #2 Results
Sample A 968 mg/kg 449 mg/kg
Sample B 857 mg/kg 415 mg/kg
Regarding Laboratory #2's analysis of samples A and B, the results of the blank spike (13.9%) and the matrix
spike (optimally, 17.6%) are clearly indicative of a reactive sulfide loss. The observed glassware reaction set-up,
the observed unrefined laboratory sample handling techniques, and the numerous unnecessary scrubber solution
agitations and transfers clearly demonstrated many opportunities for the loss of reactive sulfide during both the
preparatory and determinative procedures applied by Laboratory #2
A second group of samples were analyzed by Laboratories #1 and #3. This sample analyses included the analysis
of both raw sample provided to Laboratories #1 and #3 and analysis of several scrubber solutions previously
prepared by Laboratory #1. This approach was conceived to determine at which point (during the reaction or titra-
tion step) reactive sulfide was being lost.
35
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Field Sample Laboratory #1 Laboratory #3 Laboratory #3
Designation Results of Raw Samples Results of Raw Samples Results of Scrubber Solutions
Sample C 1070mg/kg 565 mg/kg 780 mg/kg
Sample C 860 mg/kg N/A 660 mg/kg
(Duplicate)
Regarding Laboratory #3's analysis of sample C, the results of the blank spike (66%) are clearly indicative of a
reactive sulfide loss. The observations and discoveries made during the audit provide compelling evidence by
which conclusions may be drawn. On the day of the audit of Laboratory #3 when the analyst prepared every
reagent fresh (with the authors correcting the incorrect preparation of certain reagents) and performed the entire
method step-by-step with the auditors present, the laboratory obtained its first reactive sulfide results >500 mg/kg
since 1995 which were comparable to laboratory #1's results. These reactive sulfide results were obtained on
investigative samples characterized by other laboratories to have been in excess of 500 mg/kg of reactive sulfide.
CASE STUDY NUMBER TWO: CORPORATE LABORATORY PROGRAM
The second case study involved the review of five commercial laboratories which are evaluated on a yearly basis
as part of a Corporate Environmental Laboratory program. These laboratories also underwent detailed on-site
audits. In addition, these laboratories participated in a single-blind performance evaluation (PE) study which was
performed to determine the accuracy of the laboratory-reported results when compared to known values. Since
these audits were not specific to the analysis of reactive sulfide as detailed in Case Study #1, analyst technique
could not be evaluated on a step-by-step basis. However, method deviations were present at each laboratory
audited. A summary of these finding is presented below.
SUMMARY OF AUDIT FINDINGS
• The volumes and concentrations of sodium hydroxide scrubber solution and sulfuric acid were not method
compliant.
• The flow rate of nitrogen was not monitored to 60 mL/min using a rotometer, and the size of the bubbles
varied between laboratories.
• The standardization of the potassium iodide solution and preparation of reagents used in the determina-
tive step were not documented for each titration batch.
• Method detection limit studies were not previously performed.
• A method blank, MS/MSD, and an LCS were not performed with each batch of 20 samples.
• Hydrogen sulfide standards must be prepared (and documented) daily. Several laboratories were prepar-
ing the standard annually and preserving the solutions with zinc acetate.
• Several of the laboratories were not using the correct determinative step for analysis of reactive sulfides.
Reactivity must be performed by SW-846 Method 9034 for sulfides.
• For titration of the reactive sulfides, the laboratory preserved the scrubber solution with zinc acetate. This
preservation step is not a requirement of Method 9034.
• The analyst indicated that the scrubber solution is added to the flask for titration and the standardized
iodine solution is pipetted.on top of the solution. This procedure is not method-compliant.
• The analyst indicated that the starch solution is added prior to beginning the titration. This procedure is not
method-compliant.
• Several of the analysts appeared to be improperly trained or to lack the knowledge and analytical
techniques to adequately perform the analytical method.
As an additional step, the authors conducted a single-blind performance evaluation (PE) study to monitor labora-
tory performance. Single-blind PE samples were simultaneously submitted to the five participating laboratories
being evaluated under this program. The PE samples were analyzed by each laboratory for reactive sulfide by
SW-846 Chapter 7 preparation followed by analysis by Method 9034.
A single-blind PE study sample is defined as a "test" sample in which the laboratory is aware that the sample
submitted is a PE sample but does not know the PE sample's true concentrations or results. A single-blind sample
permits the data user to better understand a laboratory's accuracy and precision capabilities and to draw conclu-
sions about the accuracy and precision of actual waste sample results.
The actual pure compounds used for the PE samples were chosen and prepared by Environmental Resource
Associates (ERA) of Arvada, Colorado. Two PE samples were chosen for the analysis of reactive sulfide One of
the reactive sulfide samples was provided as an aqueous solution and the second reactive sulfide PE sample was
36
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
provided as a gravimetrically determined sulfide salt contained in gelatin capsules. The capsule was to be placed
in the reaction vessel and would dissolve in the weak acid solution used as a reagent in the SW-846 Chapter 7
reaction/preparation method. The expected concentrations were calculated by ERA based on the assumed
10-gram sample weight used for solid sample preparation and analysis.
A review of reactive sulfide results for all laboratories, except for Laboratory B, exhibited low to very low recover-
ies. It was noted in the laboratory audit reports for all participating laboratories that significant issues relating to
method compliance were found by the authors during the on-site audits. Two laboratories are currently performing
method certification studies; the PE samples arrived during this method certification period. The authors directed
the laboratories to analyze the PE samples using the methods being certified with the understanding that the
methods may not be fully implemented by the laboratory. Additional correspondence received from one laboratory
indicated that the current reaction vessels being used are not of sufficient quality to maintain a leak tight seal
under the pressure requirements of the methods; the lack of a tight seal results in the loss of reactive sulfide.
It should be noted that the recoveries of reactive sulfide reported by Laboratory C are extremely low and may
represent a significant method or technique error by the analyst. The recoveries reported by Laboratory B appear
to be bias very high and it is recommended that the laboratory review the calculations to determine if a reporting
error occurred.
Compound/ True
Analyte Value
Reactive Sulfide 794
#1 (mg/l)
Reactive Sulfide #2
(mg/kg)
Laboratory A 1,010
Laboratory B 984
Laboratory C 976
Laboratory D 1,050
Laboratory E 1,010
Laboratory A
Reported Recovery
Value
265 33.38%
265 26.24%
Laboratory B
Reported Recovery
Value
1,250 157.43%
1,870 190.04%
Laboratory C
Reported Recovery
Value
ND o%
81.5
Laboratory D
Reported Recovery
Value
452 56.93%
520 49.52%
Laboratory E
Reported Recovery
Value
520 65.49%
690 68.32%
CONCLUSIONS
Based on the information obtained in these two case studies, it is evident that reactive sulfide analysis is being
performed by some commercial laboratories in a manner that is not compliant as mandated in a "method-defined
parameter." Laboratories that maintain strict adherence to the method, utilize correct techniques, and provide
adequate training can obtain acceptable results for reactive sulfide. Based on the collective studies, it was
observed that the analysis for reactive sulfides performed by one of the eight laboratories was method-compliant
and that this laboratory had demonstrated the capability of producing high quality reactive sulfide data.
REFERENCES
United States Environmental Protection Agency. "Chapter Seven: Characteristics Introduction and Regulatory
Definitions." Test Methods for Evaluating Solid Waste, Physical/Chemical Methods, SW-846, Third Edition,
Update III, Office of Solid Waste, December 1996.
United States Environmental Protection Agency. "Method 9034: Titrimetric Procedure for Acid-Soluble and Acid
Insoluble Sulfides." Test Methods for Evaluating Solid Waste, Physical/Chemical Methods, SW-846, Third
Edition, Update III, Office of Solid.
37
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
LESSONS LEARNED FROM PERFORMANCE EVALUATION STUDIES
Ruth L. Forman
Quality Assurance Specialist/Principal
RockJ.Vitale, CPC
Technical Director or Chemistry/Principal
Environmental Standards, Inc., 1140 Valley Forge Road, Valley Forge, PA 19482
ABSTRACT
Performance Evaluation (PE) samples are routinely utilized by both the regulatory and regulated communities to
demonstrate a laboratory's proficiency in performing a given analytical method. PE samples are submitted to
laboratories for a wide variety of regulatory programs and are typically prepared in deionized water, clean soil, or
other prepared media. The laboratory's reported results are compared to the known identities and concentrations
of target analytes in the PE samples. The evaluation of the laboratory's performance is typically based upon the
percentage of analytes the laboratory successfully recovered within a defined range of acceptance limits.
However, typically executed PE studies do not provide an indication of the laboratory's ability to successfully
identify and quantitate target analytes in a complex matrix or test other non-analytical aspects of the laboratory's
operation.
This presentation will focus on the authors' experience in conducting PE studies for a multi-state pipeline project
and will present the findings relative to these studies. Information gleaned from the PE studies relative to the
evaluation of the laboratory's performance will be discussed. Furthermore, observations regarding the laboratories'
performance in analyzing multi-phasic samples will be presented.
INTRODUCTION
Performance Evaluation (PE) samples are test samples that are prepared by spiking known concentrations of
select analytes into a well-characterized matrix. Typically, PE samples are made in a single matrix such as an
aqueous, solid, or an oil matrix. PE samples can be distributed as single-blind or as double-blind samples. For
single-blind PE samples, the laboratory is informed that they will be receiving a test sample. In the case of double-
blind PE samples, the test samples are given fictitious sample identifications and are submitted concurrently with
other project samples to the laboratory. That is, for double-blind PEs samples, the laboratory does not know that
the fictitiously labeled PE sample is a test sample. Typically, PE samples are utilized to determine a laboratory's
accuracy as it relates to the execution of a particular analytical methodology.
The authors have participated in the maintenance of a number of corporate laboratory programs in the capacity of
performing quality assurance/quality control (QA/QC) oversight for these programs. In these roles, the authors
have had experience in procuring, distributing, and evaluating the results from PE studies. However, this paper will
focus on the lessons learned from one particular project.
As the QA/QC oversight contractor on a 19,000 mile pipeline that stretches across nine of the United States,
quarterly PE samples have been submitted to the seven project laboratories for approximately three years. At the
onset of the project, a laboratory specification manual was prepared that identified prescribed SW-846 preparative
and analytical methods for the program execution. Where method ambiguities existed, program-specific method
requirements were established. In addition, the laboratory specification manual listed the target analytes, associ-
ated reporting limits, QC requirements (including frequency, QC limits, acceptance criteria and corrective action),
and data deliverable specifications (electronic and hard copy). By establishing a corporate laboratory specification
manual that all seven project laboratories were required to follow, data inconsistencies were minimized and data
comparability was enhanced.
Typically, PE samples are utilized to demonstrate method proficiency based upon the accuracy of the laboratory-
reported results compared to the known certified values. However, more information can be gleaned from a PE
study than a laboratory's demonstration of method proficiency, particularly in the case when a laboratory specifica-
tion manual is utilized for a laboratory program and when full data package deliverables are requested to substan-
tiate the reported analytical results. Information relative to the evaluation of the laboratory's technical and
administrative services, sample login and receipt, data package preparation, method compliance, and quality
assurance can also be evaluated.1 In addition, in this particular project, the authors were able to utilize the onqoinq
PE studies to identify laboratory specific trends, program specific trends, and to determine overall precision
amongst the project laboratories.2 These trends have been utilized to provide feedback to the project laboratories
to enhance their overall performance.
38
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Careful consideration was given to the preparation of the PE samples for the subject project. The intent was to test
the project laboratories' ability to analyze samples that were similar in matrix and composition to the project
samples for the analyses of interest. As such, the PE samples were custom-prepared by a reputable PE vendor
for the analytes of interest (volatiles, polyaromatic hydrocarbons [PAHs], polychlorinated biphenyls [PCBs], and
metals [including mercury]). The PE samples were soil samples that were carefully manufactured by mixing clay
and sand in proper proportion and sieve size such that the real world matrix would be stable, homogeneous, and
suitable for application of the spiked analytes. The PE samples were also moistened with deionized water to make
a multi-phasic test sample (viz., moist soil). The analytes were spiked into the PE samples at a concentration
roughly three to five times the reporting limits. The project reporting limits were based upon state cleanup action
levels.
Since the PE samples were custom-made for the subject project, verification of the manufacturing process was
important. Prior to distribution, the PE vendor verified (at their own production facility) that the recoveries of the
spiked analytes in the PE samples were acceptable for distribution to the project laboratories. In addition to the
distribution of the custom-made pre-moistened soil PE samples to the project laboratories, the PE samples were
submitted to three referee laboratories, with one of these referee laboratories receiving the PE samples in tripli-
cate. Use of the referee laboratories allowed for additional independent verification of the manufacturing process.
It should be noted that the project laboratory specification manual was distributed to the referee laboratories to
prescriptively follow for the analysis of the PE samples.
All PE samples were carefully shipped to the project laboratories and referee laboratories simultaneously. The PE
samples were shipped via overnight courier in an iced cooler, under Chain-of-Custody. For single-blind PE sample
rounds, the bottleware for the PE samples was provided by the PE provider. For double-blind PE sample rounds,
the bottleware for the PE samples originated from the project laboratory via a request from the project sampling
teams.
PE sample results are typically evaluated by comparing the laboratory-reported result to the certified true value
and determining the accuracy of the reported analytical results as a percentage relative to the true or certified
value. For the subject project, the PE sample results were evaluated in this manner and in two other ways. The
first way was to compare the laboratory-reported result to the mean result of the referee laboratories and deter-
mine a percentage. The second way was to compare the laboratory-reported result to the historical average result
and determine a percentage. The historical average result was based upon the large database of results obtained
from the PE supplier for the analyte of interest from previous PE samples that they prepared and distributed in a
similar manner.
The limits utilized for evaluating the PE samples were comparable to matrix spike limits typically observed for the
analytical methods. That is, for the volatile organic analysis, the recovery limits of 70-130% were utilized. For the
PAH analysis, recovery acceptance limits of 30-130% were utilized. For the PCB fraction, recovery acceptance
limits of 60-130% were utilized. For the metals fraction, recovery acceptance limits of 75-125% were utilized.
Finally, for the mercury fraction, recovery acceptance limits of 80-120% were utilized.
RESULTS
During the last two quarters of 1998, two single-blind PE sample studies were conducted for the subject project.
The results for the two studies are tabulated as follows. The first table in each of the two PE studies is a compari-
son of the laboratory-reported results against the certified true value. The second set of tables in each of the two
PE studies is a comparison of the laboratory-reported results against the mean referee-reported results. The third
set of tables in each of the two PE studies is a comparison of the laboratory-reported results against the historical
average (as previously discussed).
39
-------
ROUND 1. Summary of Laboratory Results and Recoveries
Compound/Analyte
(reporting units)4
benzene (pg/kg)
1,1,1-trichloroethane
(pg/kg)
1,1,2,2-tetrachloro-
ethane (pg/kg)
carbon tetrachloride
(pg/kg)
chlorobenzene (|jg/kg)
ethylbenzene (pg/kg)
4-methyl-2-pentanone
(Mg/kg)
tetrachloroethene
(Mg/kg)
trichloroethene (Mg/kg)
total xylenes (pg/kg)
Aroclor 1 248 (Mg/kg)
anthracene (pg/kg)
chrysene (Mg/kg)
benzo(k)fluoranthene
(M9/kg)
benzo(a)pyrene (pg/kg
indeno(1 ,2,3-cd)pyrene
(Mg/kg)
fluoranthene (pg/kg)
naphthalene (pg/kg)
antimony (mg/kg)
arsenic (mg/kg)
barium (mg/kg)
beryllium (mg/kg)
cadmium (mg/kg)
chromium (mg/kg)
lead (mg/kg)
mercury (mg/kg)
nickel (mg/kg)
silver (mg/kg)
True
Value
50
75
97
79
63
25
247
44
35
198
342
3,510
2,040
2,610
4,370
2,880
3,720
4,600
135
14.5
496
6.06
21.3
234
543
36.5
292
413
Lab A
Reported
Value
43
62
40
53
58
24
170
45
38
190
55
71
490
590
380
380
780
1,300
32.0
9.7
280
4.1
12
150
340
28.0
150
140
Recovery
86.00%
82.67%
41.24%
67.09%
92.06%
96.00%
68.83%
102.27%
108.57%
95.96%
16.08%
24.02%
22.61%
13.19%
20.97%
28.26%
23.70%
66.90%
56.45%
67.66%
56.34%
64.10%
62.62%
76.71%
51.37%
33.90%
LabB
Reported
Value
51
75
92
71
77
23
230
48
57
260
310
880
1,300
1,700
1,200
1,100
2,000
2,400
46.9
14.7
451
5.8
20.5
225
505
21.6
269
114
Recovery
102.00%
100.00%
94.85%
89.87%
122.22%
92.00%
93.12%
109.09%
162.86%
131.31%
90.64%
25.07%
63.73%
65.13%
27.46%
38.19%
53.76%
52.17%
34.74%
101.38%
90.93%
95.71%
96.24%
96.15%
93.00%
59.18%
92.12%
27.60%
LabC
Reported
Value
21
29
53
30
27
10
190
18
21
85
240
820
560
660
750
690
900
1,500
57.5
10.9
359
4.7
14.3
182
372
20.0
186
275
Recovery
42.00%
38.67%
54.64%
37.97%
42.86%
40.00%
76.92%
40.91%
60.00%
42.93%
70.18%
23.36%
27.45%
25.29%
17.16%
23.96%
24.19%
32.61%
42.59%
75.17%
72.38%
77.56%
67.14%
77.78%
68.51%
54.79%
63.70%
66.59%
LabD
Reported
Value
26
33
76
32
43
16
110
18
20
130
200
82
170
160
180
180
370
980
66.1
14.5
406
5.9
18.6
229
501
34.3
241
344
Recovery
52.00%
44.00%
78.35%
40.51%
68.25%
64.00%
44.53%
40.91%
57.14%
65.66%
58.48%
21.30%
48.96%
100.00%
81.85%
97.36%
87.32%
97.86%
92.27%
93.97%
82.53%
83.29%
LabE
Reported
Value
63
79
3
63
72
31
230
51
69
230
270
500
840
1,400
1,400
710
1,100
1,800
31.3
14.6
348
5.5
10.4
204
442
30.7
186
275
Recovery
126.00%
105.33%
79.75%
114.29%
124.00%
93.12%
115.91%
197.14%
116.16%
78.95%
14.25%
41.18%
53.64%
32.04%
24.65%
29.57%
39.13%
23.19%
100.69%
70.16%
90.76%
48.83%
87.18%
81.40%
84.11%
63.70%
66.59%
LabF
Reported
Value
37
ND
110
53
82
30
290
49
31
267
329
470
1,300
1,800
1,000
1,800
2,400
1,500
68.4
16.4
476
6.3
21.0
255
555
34.5
273
428
Recovery
74.00%
113.40%
67.09%
130.16%
120.00%
117.41%
111.36%
88.57%
134.85%
96.20%
13.39%
63.73%
68.97%
22.88%
62.50%
64.52%
32.61%
50.67%
113.10%
95.97%
103.96%
98.59%
108.97%
102.21%
94.52%
93.49%
103.63%
LabG
Reported
Value
50
27
45
33
59
24
82
48
33
210
264
1,300
1,500
2,100
2,000
2,000
2,600
2,700
47.0
14.4
401
5.6
17.5
220
472
30.4
239
233
Recovery
100.00%
36.00%
46.39%
41 .77%
93.65%
96.00%
33.20%
109.09%
94.29%
106.06%
77.19%
37.04%
73.53%
80.46%
45.77%
69.44%
69.89%
58.70%
34.81%
99.31%
80.85%
92.41%
82.16%
94.02%
86.92%
83.29%
81 .85%
56.42%
-------
ROUND 1 (Cont.) Summary of Laboratory Results and Recoveries
Compound/Analyte
(reporting units)4
benzene (pg/kg)
1,1,1-trichloroethane
(ug/kg)
1,1,2,2-tetrachloro-
ethane (ug/kg)
carbon tetrachloride
(ug/kg)
chlorobenzene (pg/kg)
ethylbenzene (M9/kg)
4-methyl-2-pentanone
(M9/kg)
tetrachloroethene
(Mg/kg)
trichloroethene (pg/kg)
total xylenes (pg/kg)
Aroclor1248(Mg/kg)
anthracene (pg/kg)
chrysene (pg/kg)
benzo(k)fluoranthene
(Mg/kg)
benzo(a)pyrene
(M9/kg)
indeno(1,2,3-
cd)pyrene (pg/kg)
fluoranthene (pg/kg)
naphthalene (Mg/kg)
antimony (mg/kg)
arsenic (mg/kg)
barium (mg/kg)
beryllium (mg/kg)
cadmium (mg/kg)
chromium (mg/kg)
lead (mg/kg)
mercury (mg/kg)
nickel (mg/kg)
silver (mg/kg)
Referee
Mean
28
39
51
39
40
16
90
24
21
138
242
820
1,447
1,737
1,755
1,575
2,250
2,325
47
12.1
402
5.40
15.2
194
476
34.5
208
190
Lab A
Reported
Value
43
62
40
53
58
24
170
45
38
190
55
71
490
590
380
380
780
1,300
32.0
9.7
280
4.1
12
150
340
28.0
150
140
Recovery
153.57%
158.97%
77.97%
136.95%
143.92%
150.00%
188.89%
189.87%
183.57%
137.98%
22.73%
33.86%
33.97%
21 .65%
24.13%
34.67%
55.91%
68.09%
80.17%
69.65%
75.93%
78.95%
77.32%
71.37%
81.16%
72.18%
73.53%
LabB
Reported
Value
51
75
92
71
77
23
230
48
57
260
310
880
1,300
1,700
1,200
1,100
2,000
2,400
46.9
14.7
451
5.8
20.5
225
505
21.6
269
114
Recovery
182.14%
192.31%
179.34%
183.46%
191.07%
143.75%
255.56%
202.53%
275.36%
188.82%
128.10%
107.32%
89.84%
97.87%
68.38%
69.84%
88.89%
103.23%
99.79%
121.49%
112.19%
107.41%
134.87%
115.98%
106.00%
62.61%
129.45%
59.87%
LabC
Reported
Value
21
29
53
30
27
10
190
18
21
85
240
820
560
660
750
690
900
1,500
57.5
10.9
359
4.7
14.3
182
372
20.0
186
275
Recovery
75.00%
74.36%
103.31%
77.52%
67.00%
62.50%
211.11%
75.95%
101.45%
61 .73%
99.17%
100.00%
38.70%
38.00%
42.74%
43.81%
40.00%
64.52%
122.34%
90.08%
89.30%
87.04%
94.08%
93.81%
78.09%
57.97%
89.51%
144.43%
LabD
Reported
Value
26
33
76
32
43
16
110
18
20
130
200
82
170
160
180
180
370
980
66.1
14.5
406
5.9
18.6
229
501
34.3
241
344
Recovery
92.86%
84.62%
148.15%
82.69%
106.70%
100.00%
122.22%
75.95%
96.62%
94.41%
82.64%
10.00%
11.75%
10.26%
1 1 .43%
16.44%
42.15%
140.64%
119.83%
101.00%
109.26%
122.37%
118.04%
105.16%
99.42%
115.98%
180.67%
LabE
Reported
Value
63
79
3
63
72
31
230
51
69
230
270
500
840
1,400
1,400
710
1,100
1,800
31.3
14.6
348
5.5
10.4
204
442
30.7
186
275
Recovery
225.00%
202.56%
162.79%
178.66%
193.75%
255.56%
215.19%
333.33%
167.03%
111.57%
60.98%
58.05%
80.60%
79.77%
45.08%
48.89%
77.42%
66.60%
120.66%
86.57%
101.85%
68.42%
105.15%
92.78%
88.99%
89.51%
144.43%
LabF
Reported
Value
37
ND
110
53
82
30
290
49
31
267
329
470
1,300
1,800
1,000
1,800
2,400
1,500
68.4
16.4
476
6.3
21.0
255
555
34.5
273
428
Recovery
132.14%
214.42%
136.95%
203.47%
187.50%
322.22%
206.75%
149.76%
193.90%
135.95%
57.32%
89.84%
103.63%
56.98%
114.29%
106.67%
64.52%
145.53%
135.54%
118.41%
116.67%
138.16%
131.44%
116.50%
100.00%
131.38%
224.79%
LabG
Reported
Value
50
27
45
33
59
24
82
48
33
210
264
1,300
1,500
2,100
2,000
2,000
2,600
2,700
47.0
14.4
401
5.6
17.5
220
472
30.4
239
233
Recover
y
178.57%
69.23%
87.72%
85.27%
146.40%
150.00%
91.11%
202.53%
159.42%
152.51%
109.09%
158.54%
103.66%
120.90%
113.96%
126.98%
115.56%
116.13%
100.00%
119.01%
99.75%
103.70%
115.13%
113.40%
99.08%
88.12%
115.01%
122.37%
NOTE:
ND - Not Detected.
-------
ROUND 1 (Cont.) Summary of Laboratory Results and Recoveries
Compound/Analyte
(reporting units)4
benzene (M9/kg)
1,1,1-trichloroethane
(Mg/kg)
1,1,2,2-tetrachloro-
ethane (Mg/kg)
carbon tetrachloride
(M9/kg)
chlorobenzene (Mg/kg
ethylbenzene (M9/kg)
4-methyl-2-pentanone
(Mg/kg)
tetrachloroethene
(Mg/kg)
trichloroethene
(Mg/kg)
total xylenes (Mg/kg)
Aroclor 1248 (jjg/kg)
anthracene (Mg/kg)
chrysene (Mg/kg)
benzo(k)fluoranthene
(M9/kg)
benzo(a)pyrene
(Mg/kg)
indeno(1,2,3-
cd)pyrene (M9/kg)
fluoranthene (pg/kg)
naphthalene (M9/kg)
antimony (mg/kg)
arsenic (mg/kg)
barium (mg/kg)
beryllium (mg/kg)
cadmium (mg/kg)
chromium (mg/kg)
lead (mg/kg)
mercury (mg/kg)
nickel (mg/kg)
silver (mg/kg)
Historica
Average
51
76
100
80
65
26
264
45
34
207
287
1,820
1,510
1,920
2,590
2,330
2,610
2,460
42
12.0
451
5.01
18.1
220
479
28.0
249
340
Lab A
Reported
Value
43
62
40
53
58
24
170
45
38
190
55
71
490
590
380
380
780
1,300
32.0
9.7
280
4.1
12
150
340
28.0
150
140
Recovery
84.81%
82.01%
40.20%
66.08%
89.37%
92.66%
64.39%
99.12%
1 1 1 .44%
91 .79%
19.16%
32.45%
30.73%
14.67%
16.31%
29.89%
52.85%
76.19%
80.83%
62.08%
81.84%
66.30%
68.18%
70.98%
100.00%
60.24%
41.18%
LabB
Reported
Value
51
75
92
71
77
23
230
48
57
260
310
880
1,300
1,700
1,200
1,100
2,000
2,400
46.9
14.7
451
5.8
20.5
225
505
21.6
269
114
Recovery
100.59%
99.21%
92.46%
88.53%
118.64%
88.80%
87.12%
105.73%
167.16%
125.60%
108.01%
48.35%
86.09%
88.54%
46.33%
47.21%
76.63%
97.56%
111.67%
122.50%
100.00%
115.77%
113.26%
102.27%
105.43%
77.14%
108.03%
33.53%
LabC
Reported
Value
21
29
53
30
27
10
190
18
21
85
240
820
560
660
750
690
900
1,500
57.5
10.9
359
4.7
14.3
182
372
20.0
186
275
Recovery
41.42%
38.36%
53.27%
37.41%
41.60%
38.61%
71.97%
39.65%
61.58%
41.06%
83.62%
45.05%
37.09%
34.38%
28.96%
29.61%
34.48%
60.98%
136.90%
90.83%
79.60%
93.81%
79.01%
82.73%
77.66%
71.43%
74.70%
80.88%
LabD
Reported
Value
26
33
76
32
43
16
110
18
20
130
200
82
170
160
180
180
370
980
66.1
14.5
406
5.9
18.6
229
501
34.3
241
344
Recovery
51.28%
43.65%
76.38%
39.90%
66.26%
61.78%
41 .67%
39.65%
58.65%
62.80%
69.69%
1 1 .26%
14.18%
39.84%
157.38%
120.83%
90.02%
117.76%
102.76%
104.09%
104.59%
122.50%
96.79%
101.18%
LabE
Reported
Value
63
79
3
63
72
31
230
51
69
230
270
500
840
1,400
1,400
710
1,100
1,800
31.3
14.6
348
5.5
10.4
204
442
30.7
186
275
Recovery
124.26%
104.50%
78.55%
110.94%
119.69%
87.12%
112.33%
202.35%
111.11%
94.08%
27.47%
55.63%
72.92%
54.05%
30.47%
42.15%
73.17%
74.52%
121.67%
77.16%
109.78%
57.46%
92.73%
92.28%
109.64%
74.70%
80.88%
LabF
Reported
Value
37
ND
110
53
82
30
290
49
31
267
329
470
1,300
1,800
1,000
1,800
2,400
1,500
68.4
16.4
476
6.3
21.0
255
555
34.5
273
428
Recovery
72.98%
110.55%
66.08%
126.35%
115.83%
109.85%
107.93%
90.91%
128.99%
114.63%
25.82%
86.09%
93.75%
38.61%
77.25%
91 .95%
60.98%
162.86%
136.67%
105.54%
125.75%
116.02%
115.91%
115.87%
123.21%
109.64%
125.88%
LabG
Reported
Value
50
27
45
33
59
24
82
48
33
210
264
1,300
1,500
2,100
2,000
2,000
2,600
2,700
47.0
14.4
401
5.6
17.5
220
472
30.4
239
233
Recovery
98.62%
35.71%
45.23%
41.15%
90.91%
92.66%
31.06%
105.73%
96.77%
101.45%
91.99%
71.43%
99.34%
109.38%
77.22%
85.84%
99.62%
109.76%
111.90%
120.00%
88.91%
1 1 1 .78%
96.69%
100.00%
98.54%
108.57%
95.98%
68.53%
NOTE:
ND - Not Detected.
-------
ROUND 2. Summary of Laboratory Results and Recoveries
Compound/Analyte True
(reporting units)' Value
benzene (pg/kg) 1 49
chlorobenzene (ug/kg) 44
1 ,2-dichloroethane 78
(Mg/kg)
ethylbenzene (|jg/kg) 1 08
tetrachloroethene 61
(M9/kg)
toluene (pg/kg) 89
1 , 1 ,2-trichloroethane 53
(pg/kg)
trichloroethene (ug/kg) 1 29
total xylenes (ug/kg) 328
Aroclor 1254 (pg/kg) 181
benzo(b)fluroanthene 4,240
(pg/kg)
benzo(k)fluoranthene 1,910
(Mg/kg)
benzo(a)pyrene 4,170
(pg/kg)
chrysene (pg/kg) 2,420
fluorene (pg/kg) 3,290
naphthalene (pg/kg) 3,780
phenanthrene (pg/kg) 1,570
pyrene (pg/kg) 4,940
antimony (mg/kg) 65
arsenic (mg/kg) 23.2
barium (mg/kg) 385
beryllium (mg/kg) 18.20
cadmium (mg/kg) 19.4
chromium (mg/kg) 191
lead (mg/kg) 511
mercury (mg/kg) 25.0
nickel (mg/kg) 222
silver (mg/kg) 391
Lab A
Reported Recovery
Value
100 67.11%
38 85.78%
68 87.63%
87 80.56%
42 68.63%
69 77.44%
52 98.30%
88 68.22%
260 79.27%
150 82.87%
3600 84.91%
1600 83.77%
2900 69.54%
2300 95.04%
2800 85.11%
2,400 63.49%
1400 89.17%
3900 78.95%
0.0
21 90.52%
330 85.71%
15 82.42%
16 82.47%
160 83.77%
410 80.23%
17.0 68.00%
170 76.58%
130 33.25% _,
LabB
Reported Recovery
Value
84 56.38%
31 69.98%
51 65.72%
67 62.04%
33 53.92%
58 65.10%
41 77.50%
69 53.49%
200 60.98%
130 71.82%
2700 63.68%
1,200 62.83%
2,200 52.76%
1 ,700 70.25%
2,300 69.91%
2,100 55.56%
1,100 70.06%
2,700 54.66%
58.3 90.11%
23.8 102.59%
351 91.17%
17.3 95.05%
19.4 100.00%
190 99.48%
494 96.67%
15.7 62.80%
206 92.79%
164 41.94%
LabC
Reported Recovery
Value
46 30.87%
24 54.18%
34 43.81%
44 40.74%
16 26.14%
37 41.53%
32 60.49%
33 25.58%
140 42.68%
160 88.40%
3500 82.55%
1400 73.30%
2300 55.16%
1700 70.25%
3000 91.19%
2,100 55.56%
1400 89.17%
4100 83.00%
12.2 18.86%
21.7 93.53%
331 85.97%
16.6 91.21%
16.7 86.08%
169 88.48%
426 83.37%
24.7 98.80%
186 83.78%
326 83.38%
LabD
Reported Recovery
Value
110 73.83%
36 81.26%
68 87.63%
86 79.63%
30 49.02%
72 80.81%
51 96.41%
83 64.34%
240 73.17%
110 60.77%
670 15.80%
350 18.32%
470 1 1 .27%
490 20.25%
740 22.49%
1000 26.46%
380 24.20%
1400 28.34%
19.4 29.98%
21.1 90.95%
346 89.87%
16.8 92.31%
17.6 90.72%
178 93.19%
448 87.67%
21.8 87.20%
198 89.19%
339 86.70%
LabE
Reported Recovery
Value
120 80.54%
37 83.52%
74 95.36%
85 78.70%
37 60.46%
72 80.81%
46 86.96%
90 69.77%
250 76.22%
130 71.82%
2900 68.40%
1,500 78.53%
2,500 59.95%
1700 70.25%
2,800 85.11%
2,100 55.56%
1,400 89.17%
3,600 72.87%
24 37.09%
20.8 89.66%
255 66.23%
11.9 65.38%
5.8 29.90%
190 99.48%
391 76.52%
18.4 73.60%
66.8 30.09%
208 53.20%
LabF
Reported Recovery
Value
76 51.01%
28 63.21%
51 65.72%
68 62.96%
21 34.31%
57 63.97%
38 71.83%
61 47.29%
210 64.02%
140 77.35%
5,000 117.92%
2,100 109.95%
3,800 91.13%
2,700 1 1 1 .57%
3,300 100.30%
2,300 60.85%
1,600 101.91%
5,500 111.34%
24.2 37.40%
21.8 93.97%
413 107.27%
19.0 104.40%
23.3 120.10%
181 94.76%
519 101.57%
27.5 110.00%
272 122.52%
333 85.17%
LabG
Reported Recovery
Value
88 59.06%
32 72.23%
54 69.59%
70 64.81%
33 53.92%
57 63.97%
38 71.83%
73 56.59%
210 64.02%
170 93.92%
3100 73.11%
1,400 73.30%
2,500 59.95%
1 ,800 74.38%
2,600 79.03%
2,300 60.85%
1,200 76.43%
3,500 70.85%
27.7 42.81%
24.2 104.31%
336 87.27%
18.1 99.45%
18.4 94.85%
197 103.14%
480 93.93%
25.9 103.60%
219 98.65%
361 92.33%
-------
ROUND 2 (Cont.) Summary of Laboratory Results and Recoveries
Compound/Analyte Histories
(reporting units)4 Average
benzene (pg/kg) 152
chlorobenzene (pg/kg) 45
1 ,2-dichloroethane 80
(pg/kg)
ethylbenzene (pg/kg) 112
tetrachloroethene 63
(pg/kg)
toluene (pg/kg) 90
1 , 1 ,2-trichloroethane 55
(pg/kg)
trichloroethene (pg/kg) 127
total xylenes (pg/kg) 342
Aroclor 1254(|jg/kg) 150
benzo(b)fluroanthene 2,780
(pg/kg)
benzo(k)fluoranthene 1 ,470
(pg/kg)
benzo(a)pyrene 2,470
(pg/kg)
chrysene (pg/kg) 1,840
fluorene (pg/kg) 2,240
naphthalene (pg/kg) 2,150
phenanthrene (pg/kg) 1,140
pyrene (pg/kg) 3,550
antimony (mg/kg) 21
arsenic (mg/kg) 19.3
barium (mg/kg) 350
beryllium (mg/kg) 15.00
cadmium (mg/kg) 16.5
chromium (mg/kg) 180
lead (mg/kg) 451
mercury (mg/kg) 19.2
nickel (mg/kg) 190
silver (mg/kg) 322
Lab A
Reported Recovery
Value
100 65.79%
38 83.70%
68 85.53%
87 77.68%
42 66.88%
69 76.41%
52 94.89%
88 69.29%
260 76.02%
150 100.00%
3600 129.50%
1600 108.84%
2900 117.41%
2300 125.00%
2800 125.00%
2,400 111.63%
1400 122.81%
3900 109.86%
0.0
21 108.81%
330 94.29%
15 100.00%
16 96.97%
160 88.89%
410 90.91%
17.0 88.54%
170 89.47%
130 40.37%
LabB
Reported Recovery
Value
84 55.26%
31 68.28%
51 64.15%
67 59.82%
33 52.55%
58 64.23%
41 74.82%
69 54.33%
200 58.48%
130 86.67%
2700 97.12%
1,200 81.63%
2,200 89.07%
1 ,700 92.39%
2,300 102.68%
2,100 97.67%
1,100 96.49%
2,700 76.06%
58.3 276.30%
23.8 123.32%
351 100.29%
17.3 115.33%
19.4 117.58%
190 105.56%
494 109.53%
15.7 81.77%
206 108.42%
164 50.93%
LabC
Reported Recovery
Value
46 30.26%
24 52.86%
34 42.77%
44 39.29%
16 25.48%
37 40.97%
32 58.39%
33 25.98%
140 40.94%
160 106.67%
3500 125.90%
1400 95.24%
2300 93.12%
1700 92.39%
3000 133.93%
2,100 97.67%
1400 122.81%
4100 115.49%
12.2 57.82%
21.7 112.44%
331 94.57%
16.6 110.67%
16.7 101.21%
169 93.89%
426 94.46%
24.7 128.65%
186 97.89%
326 101.24%
LabD
Reported Recovery
Value
110 72.37%
36 79.30%
68 85.53%
86 76.79%
30 47.77%
72 79.73%
51 93.07%
83 65.35%
240 70.18%
110 73.33%
670 24.10%
350 23.81%
470 19.03%
490 26.63%
740 33.04%
1000 46.51%
380 33.33%
1400 39.44%
19.4 91.94%
21.1 109.33%
346 98.86%
16.8 112.00%
17.6 106.67%
178 98.89%
448 99.33%
21.8 113.54%
198 104.21%
339 105.28%
LabE
Reported Recovery
Value
120 78.95%
37 81 .50%
74 93.08%
85 75.89%
37 58.92%
72 79.73%
46 83.94%
90 70.87%
250 73.10%
130 86.67%
2900 104.32%
1,500 102.04%
2,500 101.21%
1700 92.39%
2,800 125.00%
2,100 97.67%
1,400 122.81%
3,600 101.41%
24 113.74%
20.8 107.77%
255 72.86%
11.9 79.33%
5.8 35.15%
190 105.56%
391 86.70%
18.4 95.83%
66.8 35.16%
208 64.60%
LabF
Reported Recovery
Value
76 50.00%
28 61 .67%
51 64.15%
68 60.71%
21 33.44%
57 63.12%
38 69.34%
61 48.03%
210 61.40%
140 93.33%
5,000 179.86%
2,100 142.86%
3,800 153.85%
2,700 146.74%
3,300 147.32%
2,300 106.98%
1,600 140.35%
5,500 154.93%
24.2 114.69%
21.8 112.95%
413 118.00%
19.0 126.67%
23.3 141.21%
181 100.56%
519 115.08%
27.5 143.23%
272 143.16%
333 103.42%
LabG
Reported Recovery
Value
88 57.89%
32 70.48%
54 67.92%
70 62.50%
33 52.55%
57 63.12%
38 69.34%
73 57.48%
210 61.40%
170 113.33%
3100 111.51%
1,400 95.24%
2,500 101.21%
1,800 97.83%
2,600 116.07%
2,300 106.98%
1,200 105.26%
3,500 98.59%
27.7 131.28%
24.2 125.39%
336 96.00%
18.1 120.67%
18.4 111.52%
197 109.44%
480 106.43%
25.9 134.90%
219 115.26%
361 112.11%
-------
ROUND 2 (Cont.) Summary of Laboratory Results and Recoveries
Compound/Analyie Referee
(reporting units)4 Mean
benzene (|jg/kg) 1 1 8
chlorobenzene (pg/kg) 44
1,2-dichloroethane 80
(pg/kg)
ethylbenzene (pg/kg) 97
tetrachloroethene 45
(Mg/kg)
toluene (pg/kg) 83
1,1,2-trichloroethane 54
(pg/kg)
trichloroethene (pg/kg) 92
total xylenes (pg/kg) 295
Aroclor 1254 (pg/kg) 183
benzo(b)fluroanthene 2,880
(pg/kg)
benzo(k)fluoranthene 1,200
(M9/kg)
benzo(a)pyrene 2,220
(pg/kg)
chrysene (pg/kg) 1,780
fluorene (pg/kg) 2,700
naphthalene (pg/kg) 2,400
phenanthrene (pg/kg) 1,240
pyrene (pg/kg) 3,760
antimony (mg/kg) 25
arsenic (mg/kg) 21 .8
barium (mg/kg) 344
beryllium (mg/kg) 15.80
cadmium (mg/kg) 17.4
chromium (mg/kg) 170
lead (mg/kg) 424
mercury (mg/kg) 23.8
nickel (mg/kg) 1 93
silver (mg/kg) 290
Lab A
Reported Recovery
Value
100 84.75%
38 85.78%
68 85.53%
87 89.69%
42 93.75%
69 82.83%
52 96.65%
88 95.65%
260 88.14%
150 81.97%
3600 125.00%
1600 133.33%
2900 130.63%
2300 129.21%
2800 103.70%
2,400 100.00%
1400 112.90%
3900 103.72%
0.0
21 96.33%
330 95.93%
15 94.94%
16 91.95%
160 94.12%
410 96.70%
17.0 71.43%
170 88.08%
130 44.83%
LabB
Reported Recovery
Value
84 71.19%
31 69.98%
51 64.15%
67 69.07%
33 73.66%
58 69.63%
41 76.21%
69 75.00%
200 67.80%
130 71.04%
2700 93.75%
1,200 100.00%
2,200 99.10%
1,700 95.51%
2,300 85.19%
2,100 87.50%
1,100 88.71%
2,700 71.81%
58.3 236.99%
23.8 109.17%
351 102.03%
17.3 109.49%
19.4 111.49%
190 111.76%
494 116.51%
15.7 65.97%
206 106.74%
164 56.55%
LabC
Reported Recovery
Value
46 38.98%
24 54.18%
34 42.77%
44 45.36%
16 35.71%
37 44.42%
32 59.48%
33 35.87%
140 47.46%
160 87.43%
3500 121.53%
1400 116.67%
2300 103.60%
1700 95.51%
3000 111.11%
2,100 87.50%
1400 112.90%
4100 109.04%
12.2 49.59%
21.7 99.54%
331 96.22%
16.6 105.06%
16.7 95.98%
169 99.41%
426 100.47%
24.7 103.78%
186 96.37%
326 112.41%
LabD
Reported Recovery
Value
110 93.22%
36 81.26%
68 85.53%
86 88.66%
30 66.96%
72 86.43%
51 94.80%
83 90.22%
240 81.36%
110 60.11%
670 23.26%
350 29.17%
470 21.17%
490 27.53%
740 27.41%
1000 41.67%
380 30.65%
1400 37.23%
19.4 78.86%
21.1 96.79%
346 100.58%
16.8 106.33%
17.6 101.15%
178 104.71%
448 105.66%
21.8 91.60%
198 102.59%
339 116.90%
LabE
Reported Recovery
Value
120 101.69%
37 83.52%
74 93.08%
85 87.63%
37 82.59%
72 86.43%
46 85.50%
90 97.83%
250 84.75%
130 71.04%
2900 100.69%
1,500 125.00%
2,500 112.61%
1700 95.51%
2,800 103.70%
2,100 87.50%
1,400 112.90%
3,600 95.74%
24 97.56%
20.8 95.41%
255 74.13%
11.9 75.32%
5.8 33.33%
190 111.76%
391 92.22%
18.4 77.31%
66.8 34.61%
208 71.72%
LabF
Reported Recovery
Value
76 64.41%
28 63.21%
51 64.15%
68 70.10%
21 46.88%
57 68.43%
38 70.63%
61 66.30%
210 71.19%
140 76.50%
5,000 173.61%
2,100 175.00%
3,800 171.17%
2,700 151.69%
3,300 122.22%
2,300 95.83%
1,600 129.03%
5,500 146.28%
24.2 98.37%
21.8 100.00%
413 120.06%
19.0 120.25%
23.3 133.91%
181 106.47%
519 122.41%
27.5 115.55%
272 140.93%
333 114.83%
LabG
Reported Recovery
Value
88 74.58%
32 72.23%
54 67.92%
70 72.16%
33 73.66%
57 68.43%
38 70.63%
73 79.35%
210 71.19%
170 92.90%
3100 107.64%
1,400 116.67%
2,500 112.61%
1,800 101.12%
2,600 96.30%
2,300 95.83%
1 ,200 96.77%
3,500 93.09%
27.7 112.60%
24.2 111.01%
336 97.67%
18.1 114.56%
18.4 105.75%
197 115.88%
480 113.21%
25.9 108.82%
219 113.47%
361 124.48%
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
In the first PE study, some general observations can be made. Lower recoveries were observed for the volatile
fraction for laboratory C, D, and G. In addition, the project laboratories had higher recoveries in the volatiles
fraction than the referee laboratories. Furthermore, the recoveries of the PAH fraction were somewhat on the low
side across all laboratories, with laboratory D exhibiting very low PAH recoveries (<10%). In addition, laboratory A
exhibited a low recovery for the PCB fraction. Finally, the recoveries for antimony were observed to be low for both
the project laboratories and the referee laboratories.
At the conclusion of the first PE study, sanitized versions of the PE study results were provided to the project
laboratories as a mechanism for feedback. The project laboratories were requested to investigate the origin of the
problem areas that were identified in the study and take appropriate corrective action to identify and correct the
problem. In the second PE study, some general observations can be made. Lower recoveries were observed for
the volatile fraction for laboratory C, G, B, and F. The recoveries of the PAH fraction were greatly improved from
the previous round, with exception of laboratory D. Laboratory D still exhibited very low recoveries for the PAH
fraction. In addition, the recoveries of the PCB fraction were within acceptance limits (as previously defined) for all
project laboratories. Furthermore, the recoveries for antimony were observed to be low for both the project labora-
tories and the referee laboratories.
DISCUSSION
There are several reasons why a PE result could be outside the defined acceptance limits. First, there could be a
laboratory performance issue. This is usually observed when one laboratory performs very differently than all of
the other project laboratories for a given fraction or analyte. This was observed to be the case for laboratory D for
the PAH fraction. Second, there could be a method limitation for analyzing a sample for a given analyte. This was
observed to be the case for antimony where both the project laboratories and the referee laboratories exhibited
low recoveries for both rounds of PE samples. Furthermore, there could be a PE vendor preparation issue. This is
usually observed when all laboratories exhibit recoveries that are not within the defined acceptance range for most
of the analytes in a fraction. This was not observed to be the case for any of the PE samples issued to the project
laboratories.
LESSONS LEARNED/CONCLUSIONS
There were some inherent issues regarding the addition of water to the soil PE samples. Typically, dried, pulver-
ized sands are utilized for the solid matrix PE samples. Laboratories that perform well on the analysis of dried solid
PE samples do not necessarily perform well on the analysis of the multi-phasic PE samples (the latter being
samples that typically are submitted for analysis during environmental investigations), particularly for the PAH
fraction. Laboratories that utilized a single-solvent extraction for the preparation of the pre-moistened PAH soil PE
samples exhibited low recoveries. Laboratories that utilized a 1:1 mixture of methylene chloride/acetone for the
preparation of the pre-moistened PAH soil PE samples exhibited recoveries within the project-defined acceptance
limits. Another lesson learned is that often times laboratories pay close attention to the instrumentation and data
review but may not carefully evaluate the chemistry inherently embedded in the prescribed method. As such,
laboratory results can exceed the defined acceptance limits. For instance, between the first PE round and the
second PE round, laboratory A identified that their volumetric glassware had not been calibrated in the manufac-
turer's recommended frequency and the tolerance of the glassware they were using for the preparation of the PCB
fraction was outside the tolerance specifications.
Another observation is that the percent moisture that is added to each fraction of the PE sample should be as
consistent as possible. Often the laboratory will analyze an aliquot of sample from one of the designated analytical
fractions for percent moisture and cross-apply that one determination to all fractions from dry-weight calculations.
If the percent moisture is different for each fraction in the PE sample, the laboratory may unknowingly cross-apply
an incorrect percent moisture to the other fractions. Similarly, it was observed that the addition of water to the
volatile fraction of the soil PE sample resulted in a matrix that was not homogeneous. To overcome this problem
that was identified through the course of the PE studies, it was determined that a special coring tool was required
to properly subsample the volatile fraction of the PE sample to obtain acceptable (as previously defined) PE
results.
REFERENCES
1. Dupes, L.J., and G.R. Rose. 1995. "Conducting a Performance Evaluation Study It's Not Just Analytical
Results." Eleventh Annual Waste Testing & Quality Assurance Symposium, Washington, DC.
2. Vitale, R.J. and R.L. Forman. 1998. "A Comparison Of Single-Blind Versus Double-Blind Performance Evalua-
tion Sample Results From Commercial Environmental Laboratories." The 75th Annual American Institute of
Chemists National Meeting. Philadelphia, PA.
46
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
A CONTAMINATED MARINE SEDIMENT STANDARD REFERENCE MATERIAL:
SRM 1944, NEW YORK/NEW JERSEY WATERWAY SEDIMENT
M.M. Schantz, E.S. Beaty, D.A. Becker, R. Demiralp, R. Greenberg, M. Lopez de Alda,
K.E. Murphy, R.M. Parris, D.L. Poster, L.C. Sander and S.A. Wise
Analytical Chemistry Division, National Institute of Standards and Technology (NIST), Gaithersburg, MD 20899
R. Turle and C. Chiu
Environment Canada Environmental Technology Centre, Analysis and Air Quality Division, Ottawa, ON Canada
ABSTRACT
A Standard Reference Material (SRM) of contaminated marine sediment, SRM 1944 New York/New Jersey
Waterway Sediment, was recently issued by the National Institute of Standards and Technology (NIST) with
certified and reference values for over 100 organic and inorganic trace level constituents, along with total organic
carbon, total extractable material, and particle-size characteristics. The sediment material, which was collected
from multiple sites within the New York/New Jersey coastal waterways, has levels of polycyclic aromatic
hydrocarbons (PAHs), polychlorinated biphenyl (PCB) congeners, and chlorinated pesticides that are a factor of 5
to 10 higher than the previously issued SRM 1941a, Organics in Marine Sediment. SRM 1944 is the first NIST
natural matrix SRM with values assigned for selected dibenzo-p-dioxin and dibenzofuran congeners.
INTRODUCTION
For nearly two decades the National Institute of Standards and Technology (NIST) has been involved in the
development of Standard Reference Materials (SRMs) for the determination of organic contaminants such as
polycychc aromatic hydrocarbons (PAHs), polychlorinated biphenyls (PCBs), and chlorinated pesticides in natural
matrix environmental samples such as fossil fuels, air and diesel particulate material, coal tar, sediment, and
mussel tissue1"4. Recently NIST issued a contaminated marine sediment, SRM 1944 New York/New Jersey
Waterway Sediment, to meet the needs of laboratories involved in the testing of dredging materials from
waterways and harbors of contaminants to determine the appropriate disposal methods.
ANALYTICAL APPROACH
The general approach for the value assignment of the PAHs, PCB congeners, and chlorinated pesticides in
natural matrix SRMs has been described previously1'3 Briefly, the analytical approach for SRM 1944 consisted of
combining results from analyses using several combinations of different extraction techniques (Soxhlet and
pressurized fluid extraction) and extraction solvents, cleanup/isolation procedures (solid phase extraction and
normal-phase liquid chromatography), and chromatographic separation and detection techniques. For PAH
measurements reversed-phase liquid chromatography with fluorescence detection and gas chromatography/mass
spectrometry (GC/MS) using three different stationary phase columns were used. For the determination of PCBs
and pesticides, GC with electron capture detection and GC/MS on two different stationary phase columns were
used. In addition, for the PCB congeners and chlorinated pesticides, results from 19 laboratories that participated
in the 1995 NIST Intercomparison Exercise Program for Organic Contanrinants in the Marine Environment were
used as part of the value assignment4 The value assignment for the concentrations of selected trace elements
was accomplished by combining results from NIST [isotope dilution inductively coupled plasma mass spectrometry
(ID-ICPMS) or instrumental neutron activation analysis (INAA)], the National Research Council of Canada (NRCC)
(ID-ICPMS, graphite furnace atomic absorption spectrometry, and inductively coupled plasma atomic emission
spectrometry), the International Atomic Energy Agency (IAEA) (INAA), and seven selected laboratories using
several different analytical techniques that participated in an interlaboratory comparison exercise coordinated by
the NRCC5. Analytical measurements for the polychlorinated dibenzo-p-dioxins and dibenzofurans were the results
of an interlaboratory comparison study among 14 laboratories coordinated by NIST and Environment Canada,
Environmental Technology Centre, Analysis and Air Quality Division.
RESULTS AND DISCUSSION
The typical mode used at NIST for value assignment of natural matrix SRMs for organic contaminants has been
the analysis of the material using two or more "chemically independent" analytical techniques. The results of these
multiple technique analyses, if in agreement, are used to determine the "certified" concentrations for the measured
analytes. The requirement for using two or more analytical techniques is based on the assumption that the
agreement of the results from the independent methods minimizes the possibility of biases within the analytical
methods. When results are obtained from only one analytical technique (or multiple techniques that are not
sufficiently independent), the concentrations are typically reported as reference values and are considered as a
best estimate of the true value. The uncertainties associated with the reference values may reflect only
measurement precision, may not include all sources of uncertainty, or may reflect a lack of sufficient agreement
47
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
among multiple methods.
For value assignment for the PAHs, PCBs, and pesticides, the results from the various techniques (up to seven
sets of data for each group of compounds) were combined to provide the certified values listed in Tables 1-3. For
the 24 PAHs with certified values, the uncertainties range from 2% to 27% with 14 less than 10%. For the PCBs
and pesticides, the uncertainties for the certified values range from 2% to 14% with the majority in the 4% to 8%
range. Reference values are included in Table 3 for additional chlorinated pesticides. Reference values were also
determined for 32 additional PAHs including a number of methyl- and dimethyl-substituted PAHs (values not
shown). For the inorganic constituents in Table 4, the certified values were determined by combining results from
one NIST method with results from several outside laboratories. The reference values in Table 4 were determined
from results using only results from outside laboratories. Reference values for the concentrations of the seventeen
2,3,7,8-substituted polychlorinated dibenzo-p-dioxin and dibenzofuran congeners and the total tetra- through
hepta- substituted polychlorinated dibenzo-p-dioxins and dibenzofurans were assigned by combining results from
the analysis of SRM 1944 by 14 laboratories that participated in an interlaboratory comparison study (see Table 5).
These reference values represent the first natural matrix NIST SRM with values assigned for natural levels of
polychlorinated dibenzo-p-dioxin and dibenzofuran congeners. Reference values are also provided for total
organic carbon (4.4% ± 0.3%), extractable mass (1.15 % ± 0.04%), and particle size characteristics. With over 100
certified and reference values, SRM 1944 is one of the most characterized natural matrix SRMs available.
ACKNOWLEDGMENTS
We thank L. Rosman and P. Higgins, New York District of the U.S. Army Corp of Engineers (ACENYD) and the
crew of the Gelberman from the ACE Caven Point facility in Caven Point, NJ for their assistance in the
coordination and collection of this sediment material. We acknowledge M. Cronise and C. Fales (NIST) for their
assistance in the collection and homogenization of the sediment. Analytical measurements for selected trace
elements were provided by the International Atomic Energy Agency (Seibersdorf, Austria) by M. Makarewicz and
R. Zeisler. Trace element results were also used from seven laboratories that participated in an intercomparison
exercise coordinated by S. Willie of the Institute for National Measurement Standards, National Research Council
of Canada. We acknowledge M. Levenson and M. Vangel (NIST) for statistical evaluation of the results.
REFERENCES
1 . Wise, S.A., Schantz, M.M., Benner, B.A. Jr., Hays, M.J., and Schiller, S.B. Anal. Chem. 67, 1995, 1171-1178.
2. Schantz, M.M., Demiralp, R., Greenberg, R.R., Hays, M.J., Parris, R.M., Porter, B.J., Poster, D.L., Sander,
L.C., Schiller, S.B., Sharpless, K.S., and Wise, S.A. Fresenius J. Anal. Chem. 358, 1997, 431-440.
3. Schantz, M.M., Koster, B.J., Oakley, L.M., Schiller, S.B., and Wise, S.A. Anal. Chem. 67, 1995, 901-910.
4. Parris, R.M., Schantz, M.M., and Wise, S.A. NIST/NOAA NS&T/EPA EMAP Intercomparison Exercise
Program for Organic Contaminants in the Marine Environment: Description and Results of 1995 Organic
Intercomparison Exercises, NOAA Technical Memorandum NOS ORCA 104, Silver Spring, MD, Nov. 1996.
5. Willie, S. and Berman, S. NOAA National Status and Trends Program Tenth Round Intercomparison Exercise
Resultsfor Trace Metats in Marine Sediments and Biological Tissues, NOAA Technical Memorandum NOS
ORCA 106, Silver Spring, MD, Nov. 1996.
6. Paule, R.C. and Mandel, J. J. Research NBS, 87, 1982, 377-385.
7. Guide to the Expression of Uncertainty in Measurement, ISBN 92-67-10188-9, 1st Ed. ISO, Switzerland, 1993.
8. International Toxicity Equivalency Factor (I-TEF) Method of Risk Assessment for Complex Mixtures of Dioxins
and Related Compounds, North Atlantic Treaty Organization Committee on Challenges in the Modern Society,
Report No. 176, North Atlantic Treaty Organization (NATO), Brussels, Belgium (1988).
Table 1. Certified Concentrations for Selected PAHs in SRM 1944
mg/kg (dry-mass basis)3
Naphthalene 1.65 ±0.31
Phenanthrene 5.27 ± 0.22
Anthracene 1.77 ±0.33
Fluoranthene 8.92 ± 0.32
Pyrene 9.70 ± 0.42
Benzo[c]phenathrene 0.76 ±0.10
Benz[a]anthracene 4.72 ± 0.11
Chrysene 4.86 ±0.10
48
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 1. Certified Concentrations for Selected PAHs in SRM 1944(continued)
mg/kg (dry-mass basis)3
Triphenylene 1.04 ± 0.27
Benzo[£>]fluoranthene 3.87 ± 0.42
Benzo[/]fluoranthene 2.09 + 0.44
Benzo[/c]fluoranthene 2.30 ± 0.20
Benzo[a]fluoranthene 0.78 ±0.12
Benzo[e]pyrene 3.28 ± 0.11
Benzo[a]pyrene 4.30 ±0.13
Perylene 1.17 ±0.24
Benzo[g/?/]perylene 2.84 + 0.10
lndeno[1,2,3-cc/Jpyrene 2.78 ± 0.10
Dibenzfa, /janthracene 0.500 ± 0.044
Dibenz[a, c]anthracene 0.335 ± 0.013
Dibenz[a, ftjanthracene 0.424 ± 0.069
Pentaphene 0.289 ± 0.026
Benzo[6]chrysene 0.63 ±0.10
Picene 0.518 ±0.093
The results are expressed as the certified value ± the expanded uncertainty. Each certified value is a mean of the
means from two or more analytical methods, weighted as described in Paule and Mandel6. Each uncertainty,
computed according to the CIPM approach as described in the ISO Guide7, is an expanded uncertainty at the
95% level of confidence, which includes random sources of uncertainty within each analytical method as well as
uncertainty due to the drying study. The expanded uncertainty defines a range of values within which the true
value is believed to lie, at a level of confidence of approximately 95%.
Table 2. Certified Concentrations for Selected PCB Congeners in SRM 1944
ug/kg (dry-mass basis)3
PCB 8 (2,4'-Dichlorobiphenyl) 22.3 + 2.3
PCB 18 (2,2',5-Trichlorobiphenyl) 51.0 ±2.6
PCB 28 (2,4,4'-Trichlorobiphenyl) 80.8 ± 2.7
PCB 31 (2,4',5-Trichlorobiphenyl) 78.7 ±1.6
PCB 44 (2,2'3,5'-Tetrachlorobiphenyl) 60.2 ± 2.0
PCB 49 (2,2'4,5'-Tetrachlorobiphenyl) 53.0 ±1.7
PCB 52 (2,2',5,5'-Tetrachlorobiphenyl) 79.4 ± 2.0
PCB 66 (2,3',4,4'-Tetrachlorobiphenyl) 71.9 ±4.3
PCB 95 (2,2',3,5',6-Pentachlorobiphenyl) 65.0 ±8.9
PCB 87 (2,2',3,4,5'-Pentachlorobiphenyl) 29.9 + 4.3
PCB 99 (2,2',4,4',5-Pentachlorobiphenyl) 37.5 ± 2.4
PCB 101 (2,2',4,5,5'-Pentachlorobiphenyl) 73.4 + 2.5
90
PCB 105 (2,3,3',4,4'-Pentachlorobiphenyl) 24.5 ±1.1
PCB 110 (2,3,3',4',6-Pentachlorobiphenyl) 63.5 ±4.7
PCB 118 (2,3',4,4',5-Pentachlorobiphenyl) 58.0 ± 4.3
PCB 128 (2,2',3,3',4,4'-Hexachlorobiphenyl) 8.47 + 0.28
PCB 138 (2,2',3,4,4',5'-Hexachlorobiphenyl) 62.1 ±3.0
163 (2,3,3',4'5,6-Hexachlorobiphenyl)
164 (2,3,3',4',5',6-Hexachlorobiphenyl)
49
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 2. Certified Concentrations for Selected PCB Congeners in SRM 1944 (continued)
ug/kg (dry-mass basis)3
PCB 149 (2,2',3,4',5',6-Hexachlorobiphenyl) 49.7+1.2
PCB 151 (2,2',3,5,5',6-HexacWorobiphenyl) 16.93 ±0.36
PCB 153 (2,2',4,4',5,5'-Hexachlorobiphenyl) 74.0 + 2.9
PCB 156 (2,3,3',4,4',5-Hexachlorobiphenyl) 6.52 ± 0.66
PCB 170 (2,2',3,3',4,4',5-Heptachlorobiphenyl) 22.6 ±1.4
190
PCB 180 (2,2',3,4,41,5,5'-Heptachlorobiphenyl) 44.3 + 1.2
PCB 183 (2,2',3,4,4',5',6-Heptachlorobiphenyl) 12.19 + 0.57
PCB 187 (2,2',3,4',5,5',6-Heptachlorobiphenyl) 25.1 ± 1
159 (2,3,3',4,5,5'-Hexachlorobiphenyl)
182 (2,2',3',4,4',5,6'-Heptachlorobiphenyl)
PCB 194 (2,2',3,31,4,41,5,5'-Octachlorobiphenyl) 11.2 + 1.4
PCB 195 (2,2',3,3',4,4',5,6-Octachlorbiphenyl) 3.75 ± 0.39
PCB 206 (2,2',3,3',4,41,5>51,6-Nonachlorobiphenyl) 9.21 ± 0.51
PCB 209 Decachlorobiphenyl 6.81 ± 0.33
aSee uncertainty statement for Table 1.
Table 3. Certified and Reference Concentrations for Selected Chlorinated Pesticides in
SRM 1944
Certified Values3 M9/kg (dry-mass basis)
Hexachlorobenzene 6.03 + 0.35
cis-Chlordane (a-Chlordane) 16.51 ±0.83
trans-Nonachlor 8.20 + 0.51
4,4'-DDT 119+111
Reference Values"
a-HCH 2.0 ± 0.3
frans-Chlordane (y-Chlordane) 8 + 2
c/s-Nonachlor 3.7 + 0.7
2,4'-DDE 19 + 3
2,4'-DDD 38 + 8
4,4'-DDE 86+12
4,4'-DDD 108+16
aThe results are expressed as the certified value ± the expanded uncertainty. Each certified value is a mean of the
means from two or more analytical methods, weighted as described in Paule and Mandel6 Each uncertainty,
computed according to the CIPM approach as described in the ISO Guide7, is an expanded uncertainty at the
95% level of confidence, which includes random sources of uncertainty within each analytical method as well as
uncertainty due to the drying study. The expanded uncertainty defines a range of values within which the true
value is believed to lie, at a level of confidence of approximately 95%.
"The reference value for each analyte is the equally-weighted mean of the means from two or more analytical
methods or the mean from one analytical technique. The uncertainty in the reference value defines a range of
values that is intended to function as an interval that contains the true value at a level of confidence of 95% This
uncertainty includes sources of uncertainty within each analytical method, among methods, and from thedrvina
study. ' y y
50
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 4. Certified and Reference Concentrations for Selected Inorganic Constituents in SRM 1944
Certified Values (mg/kg, unless noted)3
Aluminum 5.33 ± 0.49(%) Lead 330 ±48
Arsenic 18.9 ±2.8 Manganese 505 ± 25
Cadmium 8.8 ±1.4 Nickel 76.1 ±5.6
Chromium 266 ± 24 Silver 6.4 +1.7
Iron 3.53±0.16(%) Zinc 656 ± 75
Reference Values (mg/kg, unless noted)3
Antimony 4.6 ±0.9 Rubidium13 75 + 2
Beryllium 1.6 ±0.3 Scandium" 10.2 ±0.2
Bromine 86 ±10 Selenium 1.4 ±0.2
Calcium" 1.0 ± 0.1 (%) Silicon 31 ± 3 (%)
Cesium" 3.0 ± 0.3 Sodium" 1.9 ± 0.1 (%)
Chlorine" 1.4±0.2(%) Thallium 0.59 ±0.1
Cobalt 14 ±2 Tin 42 ± 6
Copper 380 ± 40 Titanium" 4300 ± 300
Mercury 3.4 + 0.5 Vanadium" 100 ±9
Potassium" 1.6±0.2(%)
aThe results are expressed as the certified (or reference) value ± the expanded uncertainty. The certified (or
reference) value is based on the mean of available results from: (1) the mean of NIST INAA or ID-ICPMS
analyses, (2) the mean of two methods performed at NRC, and (3) the mean of results from seven selected
laboratories participating in the NRC intercomparison exercise, and (4) the mean results from INAA analyses at
IAEA. The expanded uncertainty in the certified value is equal to U = kuc where uc is the combined standard
uncertainty and k is the coverage factor, both calculated according to the ISO Guide24. The value of uc is
intended to represent at the level of one standard deviation the combined effect of all the uncertainties in the
certified value. Here uc accounts for both possible method biases, within-method variation, and material
inhomogeneity. The coverage factor, k, is the Student's t-value for a 95 % confidence interval with the
corresponding degrees of freedom. Because of the material inhomogeneity, the variability among the
measurements of multiple samples can be expected to be greater than that due to measurement variability
alone.
"This reference value is based only on NIST INAA measurements.
Table 5. Reference Concentrations for Selected Dibenzo-p-dioxin and Dibenzoluran Congeners in SRM 1944
ug/kg (dry-mass basis)3
2,3,7,8-Tetrachlorodibenzo-p-dioxin 0.133 + 0.009
1,2,3,7,8-Pentachlorodibenzo-p-dioxin 0.019 + 0.002
1,2,3,4,7,8-Hexachlorodibenzo-p-dioxin 0.026 ± 0.003
1,2,3,6,7,8-Hexachlorodibenzo-p-dioxin 0.056 ± 0.006
1,2,3,7,8,9-Hexachlorodibenzo-p-dioxin 0.053 ± 0.007
1,2,3,4,6,7,8-Heptachlorodibenzo-p-dioxin 0.80 + 0.07
Octachlorodibenzo-p-dioxin 5.8 + 0.7
2,3,7,8-Tetrachlorodibenzofuran 0.039 + 0.015
1,2,3,7,8-Pentachlorodibenzofuran 0.045 + 0.007
2,3,4,7,8-Pentachlorodibenzofuran 0.045 ± 0.004
1,2,3,4,7,8-Hexachlorodibenzofuran 0.22 + 0.03
1,2,3,6,7,8-Hexachlorodibenzofuran 0.09 +0.01
2,3,4,6,7,8-Hexachlorodibenzofuran 0.054 + 0.006
51
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 5. Reference Concentrations for Selected Dibenzo-p-dioxin and Dibenzoluran Congeners in SRM 1944
(continued)
|jg/kg (dry-mass basis)3
1,2,3,7,8,9-Hexachlorodibenzofuran 0.019 ± 0.018
1,2,3,4,6,7,8-Heptachlorodibenzofuran 1.0 ± 0.1
1,2,3,4,7,8,9-Heptachlorodibenzofuran 0.040 ± 0.006
Octachlorodibenzofuran 1.0 ± 0.1
Total Toxic Equivalents (TEQ)b 0.25 ± 0.01
Total Tetrachlorodibenzo-p-dioxins 0.25 ± 0.05
Total Pentachlorodibenzo-p-dioxins 0.19 ± 0.06
Total Hexachlorodibenzo-p-dioxins 0.63 ± 0.09
Total Heptachlorodibenzo-p-dioxins 1 -8 ± 0.2
Total Tetrachlorodibenzofurans 0.7 ± 0.2
Total Pentachlorodibenzofurans 0.74 ± 0.07
Total Hexachlorodibenzofurans 1.0 ± 0.1
Total Heptachlorodibenzofurans 1.5 ± 0.1
Total Dibenzo-p-dioxinsc 8.7 ± 0.9
Total Dibenzofurans0 5.0 ± 0.5
aEach reference value is the mean of the results from up to 14 laboratories participating in an interlaboratory
exercise. The expanded uncertainty in the reference value is equal to U = kuc where uc is the combined standard
uncertainty calculated according to the ISO Guide8 and k is the coverage factor. The value of uc is intended to
represent at the level of one standard deviation the combined effect of all the uncertainties in the reference
value. Here uc is the uncertainty in the mean arising from the variation among the laboratory results. The degrees
of freedom is equal to the number of available results minus one. The coverage factor, k, is the value from a
student's t-distribution for a 95 % confidence interval.
"TEQ is the sum of the products of each of the 2,3,7,8-substituted congeners multiplied by their individual toxic
equivalency factors recommended by the North Atlantic Treaty Organization9
Total of tetra- through octachlorinated congeners.
AN APPLICATION OF USEPA'S DATA QUALITY OBJECTIVE PROCESS
Karen A. Storne
O'Brien & Gere Engineers, Inc., 5000 Brittonfield Parkway, Syracuse, NY 13221
ABSTRACT
The United States Environmental Protection Agency (USEPA) states that all collected data have error, no one can
afford absolute certainty about the data, and uninformed decisions associated with data collection tend to be
conservative and expensive.1 The USEPA proposed that, before an environmental data collection project begins,
criteria should be established for decision making that is defendable. To accomplish this, the USEPA developed
the data quality objective, or DQO, process. This is a systematic planning tool used to establish criteria for data
quality, to define tolerable error rates and to develop a data collection design. Gathering the information for the
DQO process is time-consuming and may negatively impact the project budget and schedule. Therefore, a
computerized worksheet that summarizes the DQO steps was developed and distributed for review by a team of
consultant specialists.
Based on comments received from the consultant specialists, the limitations of the DQO process, from the
consultant's aspect, were outlined. This paper presents a streamlined approach to the DQO process, involving use
of a computerized worksheet to aid a project team through the DQO process. Comments pertaining to the
worksheet and the DQO process, which were solicited from the consultant specialists, are described, including the
52
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
limitations outlined by the consultant specialists.
INTRODUCTION
The Quality Assurance Management Staff (QAMS) of the USEPA developed the DQO process to improve effec-
tiveness, efficiency, and defensibility of decisions related to environmental data collection, while minimizing expen-
ditures by eliminating unnecessary duplication or overly precise data.2 The DQO process is presented in the
USEPA's Guidance for the Data Quality Objectives Process, EPA QA/G-4, EPA/600/R-96/055, September 1994.
The DQO process results in qualitative and quantitative statements that are developed through a multi-step
process that includes the following:
• Step 1. State the problem to be resolved. Identify the team members, the general problem, the project
budget, the time for the study, and the social/political issues that may impact the project.
• Step 2. Identify the decision to be made. Identify the main issue to be resolved, the alternative actions
that would result from each resolution, and the specific decision statement that must be resolved to
address the project problem.
• Step 3. Identify the inputs to the decision. Identify the variables to be measured and the basis for the
action level.
• Step 4. Define the boundaries of the study. Define the geographical area, the media of concern, the
homogeneous strata, the time frame, the start and ending time periods, the scale of the decision, and the
practical constraints for the project.
• Step 5. Develop a decision rule. The decision rule involves the population parameter of interest, the scale
of the decision making, the action level, and the alternative action. Develop the test of the hypothesis and
decision error.
• Step 6. Specify the tolerable limits on decision errors. Determine the consequences of each decision
error, the quantitation limits of the error, the range of the parameter of interest, the grey region, and the
acceptable probability of committing decision errors, or how much error is acceptable before the data
becomes unusable.
• Step 7. Optimize the design for obtaining the data. Choose a sampling design that meets the DQO
requirements and the budget.
The statements from the DQO process are summarized and presented in the Project Management Section A5 of
the Quality Assurance Project Plan (QAPP).
CONSULTANT SPECIALISTS FEEDBACK
Since each step of the DQO process is critical in choosing a sampling design, electronic worksheets that prompt
team members for responses were developed, in order to efficiently and cost effectively gather the information for
each step from busy and remotely located consultant specialists. Examples of appropriate responses to each
request were included in the worksheets. The electronic worksheets were distributed to a team of consultant
specialists in the environmental consulting firm. The consultant specialists consisted of project managers, risk
assessors, quality control officers, project officers, hydrogeologists, field samplers, and data validators.
Comments, which were based on.practical experience in the environmental field, were obtained from each of the
consultant specialists.
In general, the initial response from the team to the DQO process worksheets was positive. The team indicated
that the process of gathering project information together in a form that can be shared with the project team early
on is very critical, and is not always done properly or completely. This worksheet could be used to effectively
accomplish this task. Several comments received from the team requested clarification of some of the steps or
that additional information be requested in the steps. Based on these responses, the worksheets were modified.
However, the team had significant concerns with the DQO process, as formulated, since this process anticipates
having an idealized situation for an environmental project. The process appears to be relatively in-flexible with
respect to application of the process to real-life situations involving consent order schedules, information gaps,
large number of target constituents, and tight budgets. The team indicated that picture-perfect projects that neatly
fit the requirements of this process rarely occur.
The following sections present some of the comments received from the various consultant specialists.
Comment: The consultant specialists were unsure about how the worksheets would be used if little background
data, such as the target compounds to be measured, is known.
53
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Resolution: The worksheets may not be applicable to projects where information is not known. A statement to this
effect was added to the introduction.
Comment: Decision errors are typically not evaluated. Rather, if sample data is questionable, the data is
validated, and samples are recollected and reanalyzed. In addition, sufficient number of samples is collected to
support the project decision.
Resolution: The goal of the USEPA is to minimize costs related to data collection by decreasing unnecessary
duplication samples, and overly precise data. Utilizing the DQO process may help to decrease the number of
samples collected, thereby decreasing costs. A statement to this effect was added to the introduction.
Comment: The consultant specialists commented that following the worksheets alone to develop a sampling
design for a project could potentially leave out important issues. The worksheets attempt to put the real world in
an organized box. This structured approach typically does not work in environmental projects.
Resolution: Environmental professionals, who can use a broad breath of knowledge, experience, and complex
data to solve DQO problems, are needed to evaluate the information in the worksheets. The worksheets are to
only be used as a guide to gather the information and the DEFT software is only used to evaluate the feasibility of
the chosen sampling design. The professionals must choose the sampling design that meets the DQO needs
based on the information gathered. A statement to this effect was added to the introduction.
Comment: Some of the consultant specialists may not be able to provide information requested by the DQO
Steps.
Resolution: The worksheets would be distributed to a core group of consultant specialists, consisting of the
project manager, the risk assessor(s), and the Quality Assurance Officer, for completion. After the worksheets are
completed, the remaining consultant specialists would receive a copy for information purposes. Asterisks indicated
the consultant specialists identified as responsible for completing the worksheet.
Comment: Step 1 should include the regulatory agencies and the client name.
Resolution: These requests were added to Step 1.
Comment: Less time should be spent on the alternative actions requested in Step 2B since these actions are
often not relevant until a basic understanding of the site has been developed.
Resolution: The assumption is that the background information is available to the project team. A statement to
this effect was added to the introduction.
Comment: There may be several variables identified in Step 3.
Resolution: The worksheets are intended to be used for only one constituent. Separate worksheets must be
used for each constituent for a project. A statement to this effect was added to the introduction.
Comment: The action levels may not be defined until the risk assessment has been performed.
Resolution: It is assumed that the action levels are fixed such as regulatory thresholds and standards. A state-
ment to this effect was added to Step 3.
Comment: The information requested in Step 3B, the basis for the action levels require prior agreement between
the consultant and the client before the action levels can be presented in a QAPP
Resolution: The action levels that will be used to evaluate the sample data are critical to a project, and should
always be included in the QAPP. If the action levels are not established, the methods that can provide method
detection limits that are appropriate for the action levels may not be chosen. In addition, the data user may
compare the sample results to incorrect action levels, resulting in incorrect decisions being made and the need for
resampling. The importance of the action level was noted in Step 3B.
Comment: The information requested in Step 4 fails to consider the complexities of sampling soil, groundwater,
sediment, and surface water, potential sources, and how contaminants reside in subsurface soil. Without consid-
ering these observations, the quality of the investigation may be low.
Resolution: The DQO software makes assumptions that there are no temporal issues associated with the project
and that the sample locations can be randomized. The DQO process and software is to be used only to evaluate
the feasibility of a sample design. Statements to this effect were added to Step 7.
Comment: The significance of the y-axis in the decision performance goal diagram and how the limits of tolerable
54
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
probability are established are unclear.
Resolution: The y-axis represents the probability that a decision error will be made; deciding that the parameter of
interest is on one side of the action level when the true value is on the other side of the action level. The grey area
is where the consequences of a decision error are minimal. Below the action level, a decision error will result in
unneeded actions and increased costs. Above the action level, error will result in human health and environmental
hazard issues. The probability of decision error is set above and below the grey area to indicate the tolerable error
limits. The limits of tolerable probability are established by the project team. Clarification of these issues was
presented in Steps 6A and 6B.
Comment: The worksheets are not clear with respect to how the DQO outputs are incorporated into the sampling
design.
Resolution: Step 7 was expanded to demonstrate how the DQO outputs were utilized to choose the sampling
design.
Comment: The worksheets don't explain how information from the DQO process is added to the QAPP.
Resolution: The DQO process results in qualitative and quantitative statements summarizing the project objec-
tive, which are added to Section A5, Problem Definition and Background, of the QAPP. An example of the infor-
mation added to the project QAPP was added to Section 7. and a statement to this effect was added to the
introduction.
Based on the previously discussed comments, the worksheets were edited. The final version of the worksheets,
with edits in bold print, is presented at the end of the paper. An example of the Decision Performance Goal
Diagram from the DEFT program is presented at the end of the paper.
CONCLUSION
The DQO process, as presented in USEPA Guidance for the Data Quality Objectives Process, EPA QA/G-4, is a
good planning tool for environmental projects. Electronic worksheets that summarize the various inputs required
for the DQO steps help to decrease the time required from each team member for the information gathering
process.
After review by the consultant specialist team, it was determined that there are limitations associated with the
DQO process. All consultant specialists agreed that the process of gathering and clarifying important project infor-
mation, including the action level, that is requested by the DQO process, and having this summarized for consult-
ant specialists before the project begins, is advantageous. However, the remaining steps of the DQO process may
not be applicable to all projects. In some cases, historical background is not available, and there is an abundance
of target analytes. The application of the remaining steps of the DQO process under these circumstances would
lead to increased time and budget demands which would not be beneficial to the overall project.
The team also concluded that only a core team of consultant specialists would be responsible for filling out the
worksheets. Also, the DQO process makes assumptions, including that there are no temporal issues associated
with the project and that the sample locations can be randomized. The team also noted that the DQO process is
only to be used as a guide to determine the sampling design. Environmental professionals, who can use a broad
breath of knowledge and experience, are needed to evaluate the information in the worksheets. The DEFT
software is only used to evaluate the feasibility of the chosen sampling design. The professionals must choose the
sampling design that meets the DQO needs based on the information gathered. These limitations must be consid-
ered when implementing the DQO process.
Data Quality Objective Worksheet 1999
This worksheet is a project-planning tool, based on the Data Quality Objective (DQO) process, presented in
USEPA QA/G-4. This process is used to establish criteria for data quality and sampling designs for each
constituent at the site, so that decisions made are reasonable, defendable, and represent a logical approach to
solving the project problem, while minimizing unnecessary duplication or overly precise data. This worksheet
should be used to organize project information and is intended to be used in projects for which the basic site
problem is known and background information is available to the project team. The DQO process results
in qualitative and quantitative statements that are presented in the project QAPP. This worksheet, along
with experience, should be used by professionals to establish the data quality and sampling design.
The steps of the DQO process are presented as well as examples of appropriate responses to each request.
55
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Please fill out appropriate steps and return to K. Storne within 5 working days.
Example provided: Investigation of possible soil contamination with trichloroethene (TCE). Early sampling activi-
ties indicate that there is a low concentration area (0-50 ppm) and a high concentration area (0-80 ppm); TCE is
not detected off-site; Future land use is residential; Total budget is $100,000; Remediation must take place within
one year.
Step 1. State the problem to be resolved:
A. Who are the team members? (* indicates member responsible for completing worksheet)
Project Manager* Risk Assessor*
Quality Control Officer* Data Validator_
Data User Laboratory Project Manager_
Field sampler Client
Regulatory agencies
B. What is the general problem? (Contamination of TCE in soil. Affects human health and the environment. Low activity area is
0-50 ppm and high activity area is 0-80 ppm)
C. What project budget is available? ($100,000)
D. What time is available? (One year for remediation)
E. What social/political issues have an impact? (Future land use is residential.)
Step 2. Identify the decision to be made:
A. What is the main issue to be resolved? (Does the TCE contamination pose unacceptable danger to human health or the environ-
ment?)
B. Specify alternative actions that would result from each resolution. (ActionA - Remediate soil; Action B - Do not remediate
soil)
C. Combine main issue and the alternative actions into a specific decision statement that must be resolved to
address the problem: (Determine whether or not TCE contamination in soil poses a danger that requires remediation)
Step 3. Identify inputs for the decision:
A. What are the variables/characteristics to be measured? (TCE)
B. What is the basis for the action level (regulatory threshold or standard), that must be established and
included in the QAPP before sample collection? (Risk assessor/toxicologist set site-specific exposure assessment at 50
ppm)
Step 4. Define the boundaries of the investigation:
A. What are the spacial boundaries?
1. What is the geographical area? (property boundary; none detected off site)
2. What is the media of concern? (TCE in surface soil to depth of 15cm)
3. What are the homogeneous Strata? (Area of high concentration to 80 ppm, area of low concentration to 50 ppm)
B. What are the temporal boundaries?
1. What is the time frame? (Results represent future conditions at sites)
2. When will the investigation start and end? (Starts in 1 month and ends in 1 year)
C. What is the scale of decision to be made? (For each residential lot-sized acre.)
D. What are the practical constraints on data collection? (Existing structures exist)
Step 5A. Develop a decision rule or if/then statement that includes:
1. The population parameter of interest (do not consider sample depth) (average mean)
2. The scale of the decision making (resident lot size)
3. The action level (50 ppm)
4. The alternative action (remediate / do not remediate) (If the true mean TCE concentration in the residential lot is greater than
50 ppm, the soil is remediated. If not, the soil will be left in place)
Step 5B. Develop a test of hypothesis and decision error:
1. If the assumption is that the site is clean:
Null Hypothesis: Site is clean; true mean level <50 ppm
Alternative Hypothesis: Site is not dean; true mean level >50 ppm
• False positive (F+) Type 1 Error: Decide that the site is not clean when it is which results in action when none
was required, which is an overreaction to a situation, wasted resources, unnecessary expenditure and cleanup
56
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
• False negative (F-), Type II Error: Decide the site is clean when it is not which results in no action when some
was required, which is a missed opportunity for correction, allows a hazard to public health or environment.
Step 6A. Specify limits on decision error; how much error is acceptable:
1. Determine consequences of each decision error; how sensitive is each decision?
(health/ecological/political/social/resource risk)
2. Set quantitation limits of false positive/negative error (o-20ppm, 20-35ppm, 35-50ppm, so-eoppm, eo-iooppm, ioo-200ppm,
200-250ppm)
3. Determine range of parameter of interest; should fall within range of possible concentration (0-250 ppm)
4. Specify grey region (see table - *), where consequence of decision/error are minor; grey area is bounded by:
A. the action level (50 ppm)
B. The value where the consequences of making decision begins to be significant (60 ppm)
Step 6B. Develop the "what/if" table:
1. Specify limits on probability of committing decision errors. (For 0.3 tolerable probability, at 30% of the time a
wrong decision will be tolerated); (SOppm- 30%, 35ppm - 20%, 20ppm -10%, eoppm - 30%, iboppm - 20%, 200ppm -10%)
What/If Table I
Reported TCE
Concentration
>50ppm
>50ppm
>50ppm
<50ppm
<50ppm
<50ppm
<50ppm
Decision
Made
Cleanup
Cleanup
Cleanup
No action
No action
No action
No action
True
Concentration#
0-20ppm
20-35ppm
35-50ppm
50-60ppm
60-100ppm
100-200ppm
200-250ppm
Error
Type
F(+)
F(+)
F(+)
F(-)
F(-)
F(-)
F(-)
Aversion
Severe (Cost high)
Moderate
Minor
Minor
Moderate
Severe
Very Severe (Risk to human
health and environment)
**Tolerable
Probability
10%
20%
30%
*Grey Region
30%
20%
10%
Note: Null hypothesis - the site is clean
This completes the question section of the worksheet. The information gathered and professional experience is
then used to generate the sampling design. This process is described below.
Step 7. Based on the DQO outputs and historical information develop a sampling design. The sampling design
must be cost-effective and balance sample size with method performance and decision error tolerance. For
example, it may be more cost effective to use less expensive and less precise methods in cases of high variability
in samples exist, so that a large number of samples can be taken and so that the sample design error can be
controlled. If less variability in samples exists, more expensive and precise methods can be used to collection
fewer samples to control the measurement error.
The USEPA DEFT software is used only as a guide to develop the sampling design alternatives. DEFT does not
account for the difference between media, spacial or temporal boundaries. Inputs for the DEFT program include:
A. Parameter of interest; assumption is that the population mean is used (mean)
B. Limits on probability of committing decision errors (SOppm- 30%, 35ppm - 20%, 20ppm - 10%, 60ppm - 30%,
100ppm - 20%, 200ppm - 10%)
C. Action level (SOppm)
D. Possible range of parameter (250ppm)
E. Unit cost of sample collection and analysis per sample ($30, $220)
F Location and width of grey region; range of possible parameter values where consequences of F(-) error are
minor; bounded by action level and parameter value where consequences of F(-) begin to become significant
(50-60ppm)
G. Estimated standard deviation (default is used; max concentration - minimum concentration /6 )
H. Null hypothesis; which error is F(+) and which is F(-) (Site is clean)
57
-------
WTQA '99 - 15th Annual Waste Testing & Qualify Assurance Symposium
Three basic sampling designs available in the DEFT program include:
A. Simple random; Many samples are taken and total costs are high. Every possible point at the site has equal
chance of being sampled. Simple random is used when variability is small and field and analytical costs are low
to detect peak concentrations.
B. Composite; Multiple samples are collected and combined; subsamples are collected for additional analysis.
Composite is used when the average concentration and sampling of a large number of sample sites at a
reduced cost is desired.
C. Stratified random; The site is divided into two or more subsets. Each subset is sampled separately with one of
the designs previously described. Stratified random is used to improve the precision of the design.
The previously listed inputs, the initial sample design, and the sample size are entered into the DEFT program,
and a performance goal diagram is drawn. Altering the inputs, the design, or the sampling size may change the
decision performance goal. The performance of the design is evaluated by the performance curve, which is based
on the graph of the power function, and which is overlaid onto the performance goal diagram. The design that
produces a very steep performance curve is preferred over one that is flatter. The power function is the probability
that the null hypothesis is rejected when the null hypothesis is false. Ideally, the power function would be zero if the
null hypothesis were true and one if the null hypothesis were false. Due to imperfect data, it is not possible to
achieve the ideal power function. However, the power function will yield values that are small when the null
hypothesis is true and large when the null hypothesis is false.
If the design fails to meet the DQOs, increase the budget, increase the width of the grey area, or increase the
tolerable decision error rates.
The statements resulting from the DQO process are presented in the Project Management, Section A5, of the
QAPP. (A simple random sample design should be used to compare concentrations of samples collected for TCE
analysis from the site to the action level of 50 ppm. 20 samples shall be collected from each sample location. Each
sample location will be generated randomly.)
The Decision Performance Goal Diagram for the example provided in the DQO worksheets
0.
00 <-
50.00 10000 15000 20000 25000
Concentration
DECISION PERFORMANCE GOAL DIAGRAM
Simple Random Sampling
Action Leuel - 50.00
Cost - $5000.00
Sample Size - 20
concentration prob. type
20.00 0.1000 F(+)
35.00 0.2000 F(*)
50.00 0.3000 F(+)
60.00 0.2S27 F(-)
100.00 0.2000 F(-)
200.00 0.1000 F(-)
Press any key to return to the Design/DQO Summary Screen.
58
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Footnotes
1. USEPA 1997. Introduction to Data Quality Objectives, Quality Assurance Division, Washington D.C., page 4.
2. USEPA 1994. Guidance for the Data Quality Objectives Process, EPA QA/G-4, Washington D.C., page 1.
References
USEPA. 1994a. Data Quality Objectives Decision Error Feasibility Trials (DQO/DEFT), Version 4.0, EPA QA/G-
4D, Washington, D.C.
USEPA. 1994b. Guidance for the Data Quality Objectives Process, EPA QA/G-4, EPA/600/R-96/055, Washington,
D.C.
USEPA. 1997a. Introduction to the Data Quality Objectives, Quality Assurance Division, Washington, D.C.
USEPA. 1997b. Overview of EPA Quality System Requirements for Environmental Programs, Quality Assurance
Division, Washington, D.C.
NEW SAMPLING DEVICE PROVIDES LABORATORY VERIFICATION - PART 1
PRELIMINARY DATA PROVIDES SOME INTERESTING POSSIBILITIES
Thomas Wayne:. Kabis
VP Research & Development
SIBAK Industries Limited, Inc., Solana Beach, California State
ABSTRACT
Laboratory accuracy and efficacy with regards to groundwater testing has long been a concern for environmental
investigators. Bench standards used in laboratory certification offer some degree of reliability, however, as many
investigators have learned, laboratory equipment is prone to radical and sudden departures from expected results.
These departures often go undetected by the laboratory chemist operating the analysis equipment until the report
is issued and questioned by the investigator. Usually, by the time a question has been raised, the samples are
out-of-date for viability and the question of validity falls to the QA/QC sheets and the calibration log for the instru-
ment. Since recalibration, again under ideal conditions and with bench standards, almost always indicates a
proper functioning of the instrument, erroneous data are accepted as valid and improper and false assumptions
are made based on such data. A new groundwater-sampling instrument has been invented which may, however,
provide immediate warning of aberrant analysis results. Preliminary data indicate a remarkable 98.75% reproduci-
bility of sampling results is possible using the KABIS Sampler.
59
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
60
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
INORGANIC
ANALYSIS
61
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
62
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
RECENT DEVELOPMENTS IN THE DETERMINATION OF TRACE LEVEL
PERCHLORATE BY ION CHROMATOGRAPHY
Peter E. Jackson
Dionex Corporation, 1228 Titan Way, Sunnyvale, CA 94088-3606
(408)481-4262
David T. Tsui
USAF, AFRL/HEST - BLDG 79, 2856 G. Street, Wright-Patterson AFB, OH 45433
Howard Okamoto
CDHS, 2151 Berkeley Way, Room 119, Berkeley, CA 94704
Frank Calovini
SAIC, 7615 State Road, Parma, OH 44134
ABSTRACT
Ammonium perchlorate, a key ingredient in solid rocket propellants, has recently been found in ground and
surface waters in a number of states in the U.S. Perchlorate poses a health risk and preliminary data from the U.S.
EPA reports that exposure to less than 4 18 ug/L provides adequate human health protection. Ion chroma-
tographic (1C) methods, based on either lonPac AS5 or AS11 columns, a large loop injection, and suppressed
conductivity detection have been developed for the determination of low ug/L levels of perchlorate in drinking and
ground waters. These methods provide similar freedom from common anion interferences, are linear in the range
of 2 -100 ug/L, and quantitative recoveries are obtained for low ug/L levels of perchlorate in spiked drinking and
ground water samples. MDLs obtained using 1C permit quantification of perchlorate below the levels which ensure
adequate health protection.
INTRODUCTION
Ammonium perchlorate, a key ingredient in solid rocket propellants, has recently been found in drinking water
wells in regions of the U.S. where aerospace material, munitions and fireworks were developed, tested, or
manufactured. To date, perchlorate has been found in ground and surface waters in California, Nevada, Utah,
Texas, New York, Maryland, Arkansas and West Virgina, although the total extent of the contamination problem is
not known1 The presence of perchlorate in drinking water poses a considerable health risk, even at trace levels2.
While perchlorate is listed on the EPA Contaminant Candidate List as a research priority, it is not currently
regulated under the Federal Safe Drinking Water Act. The California Department of Health Services (CDHS)
developed an ion chromatographic (1C) method for the analysis of trace perchlorate in 1997 to support the CDHS
action level of 18 ug/L in drinking water3. The CDHS method uses a large loop injection with an lonPac ASS
column and a hydroxide eluent containing p-cyanophenol. Detection is by suppressed conductivity using a chemi-
cally regenerated AMMS suppressor.
An updated 1C method employing an lonPac AS11 column, hydroxide eluent, and suppressed conductivity detec-
tion with a self regenerating ASRS suppressor, was developed in 19984. Draft Update IVB Method 9058, titled
"Determination of Perchlorate by Ion Chromatography" includes conditions for using either the lonPac AS5 or
AS11 columns5
This paper will report on a recent developments for the determination of trace level perchlorate using ion chroma-
tography. The performance of the ASS and AS11 methods will be discussed and their application to the analysis of
perchlorate in a variety of environmental samples, including drinking water, ground and surface waters, soils and
contaminated wastes will be demonstrated. The application an EG40 automated eluent generator with a new
polarizable anion analysis column, the lonPac AS16, for the determination of perchlorate in high ionic strength
samples will be also presented.
EXPERIMENTAL
Instrumentation
Either 4500 or DX-500 ion chromatographs (Dionex Corporation, Sunnyvale, CA) were used for this work. Separa-
tions were carried using Dionex lonPac® ASS, AS11 and AS16 (250 x 4.0 mm) analytical columns and lonPac
AG5, AG11 and AG16 (50 x 4.0 mm) guard columns. Anions were detected by suppressed conductivity detection
using either an Anion Micro-Membrane Suppressor, AMMS® with a regenerant of 35 mN sulfuric acid at 10
mL/minute; or an Anion Self-Regenerating Suppressor, ASRS®-ULTRA operated at 300 mA in the external water
mode.
63
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Reagents
All water used was deionized water, Type I reagent grade, 18 MQ-cm resistance or better. Sodium hydroxide, 50%
w/w aqueous solution was obtained from Fisher Scientific (Pittsburgh, PA). Sodium perchlorate, 99% ACS reagent
grade was obtained from Aldrich (Milwaukee, Wl), as was 95% p-cyanophenol. ACS reagent grade chemicals
were used for the preparation of the standards for the interference and recovery studies, with the exception of
humic acid and selenate standards, which were prepared from technical grade reagents.
RESULTS AND DISCUSSION
In order to quantify perchlorate at low ug/L levels, it is essential to optimize chromatographic conditions in terms of
retention time, peak shape and baseline noise. The perchlorate ion is a "polarizable" anion, consequently it should
be chromatographed on a hydrophilic anion exchanger to minimize peak tailing. In addition, perchlorate is highly
retained on anion exchange resins and requires a strong mobile phase to elute it within a reasonable timeframe,
which is desirable for lower detection limits. Initial investigations on an lonPac ASS column showed that an eluent
of 120 mM hydroxide containing an organic modifier, such as p-cyanophenol, was required to elute perchlorate
from the AS5 column. The effect of p-cyanophenol over the range of 0 - 3 mM on perchlorate retention was inves-
tigated, with an eluent of 120 mM NaOH containing 2.0 mM p-cyanophenol providing optimal peak shape and a
retention time for perchlorate of approximately 7 minutes3 The perchlorate anion is well resolved from common
inorganic anions, which essentially elute at the void volume under these conditions.
A large loop injection (740 uL) is required for this application in order to achieve sub-ppb detection limits for
perchlorate. The method detection limit (MDL) using the lonPac ASS column was determined by spiking perchlo-
rate at concentrations of 1.0, 2.5, and 4.0 ug/L into reagent water, as shown below in Table I.
TABLE I. Method Detection Limit in Reagent Water.
Perchlorate
Spike Cone. (pg/L)
1.0
2.5
4.0
No of Spiked
Replicates
14
16
16
Mean
Recovery (ug/L)
0.87
2.3
3.9
Standard Deviation
(ug/L)
0.11
0.12
0.11
Pooled MDL (df= 43)
MRL (5 x MDL)
Calculated MDL
(M9/L)
0.6
0.8
0.7
0.7 ug/L
4 ug/L
A linearity study was performed to ensure accurate quantification of perchlorate in the low ug/L range. A correla-
tion coefficient of 0.9998 was obtained for a plot of peak area versus concentration in the 2-100 ug/L range,
demonstrating that calibration is linear at the levels required for the quantification of perchlorate in drinking and
ground waters.
In addition to the ASS column, it has also been shown that perchlorate can be successfully chromatographed on
an lonPac AS11 column4. The major advantage of using this more hydrophilic column is that p-cyanophenol is not
required in the eluent in order to achieve good peak shape for perchlorate. This enables the use of electrolytic self
regererating suppressors (e.g. ASRS-ULTRA), which add considerable convenience to the operation of the ion
chromatograph, as SRS devices are not recommended for use with eluents containing electroactive modifiers,
such as p-cyanophenol.
The lonPac AS11 column with an eluent of 100 mM sodium hydroxide permits the elution of perchlorate in less
than 10 minutes. Figure 1 shows a typical chromatogram of a 20 ug/L perchlorate standard obtained using the
AS11 column.
The method detection limit was determined for the AS11 column using seven replicates of 2.5 mg/L perchlorate
spiked into reagent water according to the procedure outlined in U.S. EPA Method 300.06. The single operator
MDL was was calculated to be 0.3 pg/L using the conditions shown in Figure 1.
64
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Both the lonPac ASS and AS11 columns were tested for interferences by injecting low ug/L levels of perchlorate in
the presence of 100 |jg/L solutions of 22 common anions. Of the anions investigated, only cyanide, iodide, and
thiocyanate display any significant retention on either
0.6
US
0.0
2.0 4.0 6.0 8.0 10.0 12.0
Minutes
column using the elution conditions described above.
Perchlorate is resolved by at least 2 minutes from the
nearest eluting anion, thiocyanate, which would not be
typically found at high levels in drinking or ground waters.
Figure 1. Perchlorate standard at 20 ug/L Conditions:
guard column, Dionex lonPac AG11; analytical column,
Dionex lonPac AS11; eluent, 100 mM sodium hydroxide;
flow-rate, 1.0 mL/min; detection, suppressed conductivity;
injection volume, 1000 uL; peak 1 - perchlorate.
The effect of mg/L levels of common anions on perchlorate recovery was investigated by injecting solutions of low
ug/L levels of perchlorate in the presence of 50, 200, 600 and 1000 mg/L chloride, carbonate and sulfate, respec-
tively. Quantitative recoveries were obtained for perchlorate in all cases, demonstrating that mg/L levels of
common anions have no significant effect on the recovery of low ug/L levels of perchlorate.
Essentially, the ASS and AS11 columns give similar method performance, in terms of linearity, MDLs, freedom
from interferences, and spiked recoveries, as was demonstrated in the recent IPSC collaborative study7. The IPSC
study, which involved 19 laboratories, was organized to quantitatively evaluate the performance of ion chroma-
tographic methods for the measurement of perchlorate in drinking and ground water. The study samples consisted
of well water at three total dissolved solids levels of 72, 144, and 288 mg/L, which were spiked with perchlorate at
concentrations of 6, 18 ppb and 36 ug/L. Both the ASS and AS11 columns were found satisfactory for perchlorate
analysis in typical ground and surface water samples.
Tables II and III show examples of single operator accuracy and precision obtained using the ASS column for
perchlorate standard solutions and matrix spikes into ground water.
TABLE II. Single Operator Accuracy and Precision for Perchlorate Standard Solutions.
Sample Type
IPC Standard
QCS
LFB
Sample
Matrix
RW
RW
RW
Known Cone.
(M9/L)
5.0
100
4.0
100
4.0
Number
of Replicates
48
47
16
4
22
Mean Recovery
(ug/L)
4.9
100
4.0
100
3.9
(%)
98
100
100
100
98
SD
(ug/L)
0.35
4.2
0.31
2.8
0.33
RSD(%)
7.1
4.2
7.8
2.8
8.5
RW = reagent water
TABLE III. Single Operator Accuracy and Precision for Perchlorate Matrix Spikes.
Sample Type
Matrix Spike/Matrix
Spike Duplicate
Sample
Matrix
GW
Spike Cone.
(ug/L)
4.0
Number of
Spiked Pairs
20
Duplicate Spike
Mean Recovery
(M9/L)
3.8
(%)
95
SD of Mean
RPD (%)
0.02
Mean
RPD (%)
2.1
GW = ground water
65
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Dionex has recently developed a new column for the analysis of polarizable anions, such as perchlorate. The
lonPac AS16 column is more hydrophilic and has a significantly higher ion exchange capacity than either the AS5
or AS11 columns. This column will allow the injection of higher ionic strength samples and is also compatible with
the EG40 automated KOH eluent generator. Current work on perchlorate analysis by 1C involves extending the
range of applications to more complex samples, such as wastewaters containing solvents, and to high ionic
strength samples (> 2000 uS/cm) using either direct injection or with appropriate sample pretreatment.
CONCLUSION
The use of ion chromatography with the lonPac ASS or AS11 columns, large loop injection and suppressed
conductivity detection provides a simple, interference free method for the determination of perchlorate at low ug/L
levels in drinking and ground waters. The method is linear over the range of 2 -100 ug/L and quantitative recover-
ies were obtained for perchlorate in spiked drinking and ground water samples. The MDLs permit quantification of
perchlorate below the levels which ensure adequate health protection (4 -18 ug/L), as recommended by the U.S.
EPA. The new AS16 column provides similar performance to the ASS and AS11 columns for drinking water
samples, although its higher capacity makes it more suitable for the analysis of trace perchlorate in high ionic
strength matrices.
REFERENCES
1. California Department of Health Services (1998), Perchlorate in California Drinking Water, October Update.
2. Jarabek, A. (1999), Toxicology and Health Effects, paper presented at the Perchlorate Conference,
Ontario, CA.
3. California Department of Health Services (1997), Determination of Perchlorate by Ion Chromatography.
4. Wirt, K., Laikhtman, M., Rohrer, J. and Jackson, P.E. (1998), Am. Env. Lab., 10(4), 1.
5. Proposed Method 9058 in U.S. EPA "SW-846 Test Methods for Evaluating Solid Waste Physical/Chemical
Methods", Draft Update IVB, U.S. EPA, Washington, DC.
6. U.S. EPA. Method 300.0 (1993), The Determination of Inorganic Anions in Water by Ion Chromatography.
7. Chaudhuri, S., Okamoto, H., Pia, S. and Tsui, D. (1999), Inter-Agency Perchlorate Steering Committee
Analytical Subcommittee Report.
A GENERIC LEACHING PROCEDURE TO PREDICT ENVIRONMENTAL IMPACT OF
REACTIVE MATERIALS SUCH AS COAL COMBUSTION BY-PRODUCTS
David J. Hassett
Senior Research Advisor
Energy & Environmental Research Center, University of North Dakota,
15 North 23rd Street, Grand Forks, ND 58203
ABSTRACT
Leaching characterization of many materials can give misleading data if materials are reactive or undergo
chemical, physical, or mineralogical transformations upon contact with water. Additionally, if a reactive material
can undergo mineralogical transformations that take up to 30 days or more, short-term leaching tests with
equilibration times of 18 hours may not provide data relevant to what is likely to occur under field conditions. Many
coal combustion by-products (CCBs) fall into this category, and for this reason, leaching must take into account
site-specific conditions and must be based on a thorough and fundamental understanding of the nature of the
materials being characterized. This paper describes a generic leaching test developed for the characterization of
CCBs that is site-specific and also takes into account the reactivity and unique characteristics of many CCBs. The
test is also appropriate for numerous other waste materials likely to be disposed in monofills or under conditions
other than codisposal in sanitary landfills. The test incorporates a long-term leaching component. The original
leaching test, the synthetic groundwater leaching procedure (SGLP), was developed in 1982 at the University of
North Dakota. Since that time, the test has been used in numerous research projects and environmental
assessments. In many cases, the test provides significantly different data from that generated through the use of
the toxicity characteristic leach procedure (TCLP). In addition to leaching tests of real process and CCB streams
pilot plant studies have been used in conjunction with this test to predict potential for environmental impact from
materials to be generated in units either in the planning or permitting stages. Studies of leached materials usina
66
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
x-ray diffraction have provided insights into mechanisms responsible for unexpected leaching behavior. In the
case of some alkaline CCBs, initial leachate solution concentration increases of select trace elements have been
followed by dramatic concentration decreases of more than 1 or 2 orders of magnitude. Elements for which this
type of behavior have been noted are oxyanionic species of environmentally sensitive elements such as boron,
chromium, selenium, and vanadium. The phenomenon is referred to as anomalous leaching. Supporting data as
well as examples of mineralogical characterizations of leached residues to support hypotheses and conclusions
are provided.
INTRODUCTION
The determination of potential for environmental impact of disposed materials is a topic important enough to
warrant the application of only the best in characterization methods. Decisions based on waste characterization
prior to disposal are by necessity long-term decisions and must be based on data generated using scientific
protocols that are the most predictive of what can be expected to occur under actual disposal conditions. The
nature of laboratory leaching imposes certain limitations on the information that it is possible to generate. Current
protocols allow the measurement of total mass of analyte in any given waste form (bulk chemical analysis), and
leaching techniques allow the determination of mass of easily leached analyte as well as an assessment of
leachate concentration trends with respect to time. Normal leaching trends would predict the concentration of
many trace analytes to increase rapidly during the first few hours of leaching, with a subsequent gradual increase
to an equilibrium concentration. As this discussion will demonstrate, concentration in certain reactive solids may
actually decrease during the course of long-term leaching with equilibration times of up to 60, or even 90, days.
The reactive solid wastes under consideration here are coal combustion solid residues, primarily fly ash. Currently
over 40 million tons of fly ash are disposed yearly in the United States1, making the topic of ash disposal an
important and timely one. Ash disposal is a focus of U.S. Environmental Protection Agency (EPA) efforts to
determine the Resource Conservation and Recovery Act (RCRA) status of wastes from the combustion of fossil
fuels, including ash from combustion fuels (coal and other fuels), oil ash, and small-volume wastes2, as well as
being a hot topic for several state and environmental groups.
The accepted practice of using the toxicity characteristic leaching procedure (TCLP) for determination of
hazardousness of disposed wastes employs a technique that has been shown to be effective in determining
leaching trends from materials disposed of in sanitary landfills. The procedure utilizes acidic leaching conditions
with acetic acid and simulates what occurs in a landfill where decomposition of cellulosic materials and other
garbage produces acidic conditions, with acetic acid as a major component.
The TCLP is sometimes used to predict the nature of leachate generation in disposed coal combustion
by-products (CCBs). Although this test is adequate for its intended purpose, the simulation of leaching under
codisposal conditions in a sanitary landfill, its use for evaluation of CCBs disposed in monofills is clearly
inappropriate. Many CCBs, especially those from low-rank coals, are highly alkaline, and even those that are not
alkaline are unlikely to contact acetic acid with monofill disposal. Under most monofill disposal scenarios, it is the
ash itself that will likely determine the major composition of leaching solution regardless of the initial chemistry of
the water contacting the material. This is due to the relatively high mobility of calcium, sodium, and sulfate in many
ash types, which is enhanced in a monofill because of the tendency for there to be a condition where a relatively
large mass of ash would be mixed with a relatively small volume of pore water, either through infiltration of
rainwater prior to capping the facility or through infiltration of groundwater if the monofill is situated below the water
table. In the event that the monofill is wetted, the normally low permeability of liners or geologic materials in
well-situated monofills would produce conditions where the contact time of ash and water would be relatively long
prior to the generation of substantial volumes of leachate outside of the facility. Thus it is envisioned that most
scenarios of leachate generation in ash monofills would entail a long contact/equilibration period where ash and
water would have an opportunity to react together, should reactions be possible. In the case of CCBs from low-
rank coal or CCBs from advanced coal combustion techniques, it is likely that secondary hydrated phases will form
upon contact of the ash with water.
EXPERIMENTAL
Recognizing the inadequacy of TCLP to simulate leaching in ash monofills, researchers began an effort to fill this
gap and provide a protocol for estimating trace element mobility in this unique but important situation. A long-term
leaching project begun in 1987 provided data that indicated that not only was there a problem from the standpoint
of leaching solution chemistry, but a problem in the short-term nature of currently available protocols such as the
American Society for Testing and Materials (ASTM) water leach procedure was also uncovered3 It was observed
that certain important leachate components from ash-water leaching systems increased in concentration initially,
as would be expected, then decreased suddenly. At the same time, x-ray diffraction characterization of the waste
67
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Ca6Si2(SO4)2(CO3)2(C>H)i2'24l-l2O
Ca6Mn2(SO4)2(CO3)2(OH)12-24H2O
materials indicated the formation of a mineral called ettringite. It was also observed that the formation of this
mineral paralleled the decrease in concentration of several key trace components of the ash- water systems.
These elements were arsenic, boron, and selenium, and it was later determined that these and several other
important elements can substitute into the ettringite structure and thus be incorporated into a highly insoluble
phase formed from the hydration of ash. This was true in many low-rank CCBs and in all CCBs from advanced
coal combustion processes such as fluidized-bed combustion (FBC), duct injection acid gas control processes,
and lime injection multiburner (LIMB) processes. The following is a summary of the research that led to the
development of a synthetic groundwater leaching procedure (SGLP) and long-term leaching (LTL) procedure.
An investigation of the leachability of trace elements from CCBs was conducted at the University of North Dakota
Energy & Environmental Research Center (EERC). The primary objectives of the investigation were to develop
protocols for the estimation of trace element mobility and to begin to understand the processes in ash hydration
responsible for the anomalous leaching behavior observed in numerous cases. For purposes of discussion, it is
assumed here that "normal" leaching behavior is a rapid rise in concentration followed by a more gradual rise
leading to a stable but increased concentration of analyte. "Anomalous" leaching is defined as the situation where
analyte concentration initially rises, as would be expected, but then decreases with time, often to extremely low
levels. In most cases where anomalous leaching behavior was observed, ettringite was detected or conditions for
ettringite formation were met, although in some cases ettringite was not detected by x-ray diffraction.
Ettringite is a mineral with the nominal composition Ca6Al2(SO4)3(OH)i2'26H2O. Ettringite is also the family name
for a series of related compounds, as is the case for many mineral families. Included in this family are the following
minerals4:
Ettringite Ca6Al2(SO4)3(OH)12-26H2O
Charlesite Ca6(Si,AI)2{SO4)2[B(OH)4](OH,O)i2-26H2O
Sturmanite
Thaumasite
Jouravskite
Bentorite
Ettringite has fairly unique characteristic structural features. The structure comprises calcium aluminate columns
{Ca6AI2(OH)i224H2O}6*, with the channels between these columns containing the other components, which include
an oxyanion such as sulfate with hydroxide and water {(SO4)2^ (OH)cM (H2O)0-6}6". The structure of ettringite is
shown in Figure 15'7.
Although ettringite was reported in the scientific literature in the early 1930s, it has only been recently recognized
just how great a potential impact this mineral could have on human activities. Ettringite is relatively easy to
synthesize in the laboratory. All that is necessary is an aqueous solution containing calcium, aluminum, and sulfate
at a pH between 1 1 .5 and 12.5. If the proper concentrations of components are provided along with high alkalinity,
ettringite forms readily. These conditions are often
met when low-rank CCBs contact water. The ash in
most cases has all of the potential ingredients for
ettringite formation, and it has been found that many
low-rank CCBs do form ettringite as a primary
hydration product. The availability of alkaline con-
stituents to provide the required high pH conditions
are often the limiting factor with CCBs. Extensive
research into ettringite formation has been carried out
at the EERC in conjunction with North Dakota State
University. In this study, numerous substituted et-
tringites were synthesized in the laboratory. The sub-
stituents were elements that tend to exist as
oxyanions in aqueous solution, and they enter into the
ettringite structure by substituting for sulfate. Ettrin-
gites substituted with arsenic, boron, chromium,
molybdenum, vanadium, and selenium have been
Figure 1. Ettringite Structure
EtRC DH!631SCOR
68
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
prepared in the laboratory8. Thus ettringite formation has the potential to influence the solution concentrations of
these and probably numerous other elements, including aluminum, calcium, and sulfate, major constituents of the
ettringite structure. It is also important to note that the rate of formation of ettringite in CCBs is dependent on the
availability of the key ingredients in the structure. Since many of these are leached from the ash from various
crystalline and amorphous phases, the formation of ettringite can take from hours to months, depending on the
characteristics of the individual ash. Each ash, due to the variability of the phases making up these materials, is
unique in this manner.
A short-term leaching test of 18-hour duration such as the TCLP or ASTM water shake test may be concluded
before ettringite has even begun to form in some CCBs. This could resultin highly misleading information
regarding leachability of several very important and potentially problematic trace elements.
The following protocols were developed to fill the gap in leaching procedures for the characterization of materials
for which the TCLP was clearly inappropriate and for materials for which short-term leaching was not predictive of
field phenomena. The protocol for the TCLP is given for reference.
TOXICITY CHARACTERISTIC LEACHING PROCEDURE
The TCLP9 is the EPA regulatory leaching procedure. The TCLP has also been adopted by many state regulatory
agencies to provide leaching information on solid wastes (not hazardous) which are not federally regulated. This
test uses end-over-end agitation and a 20- to-1 liquid-to-solid ratio with an 18-hour equilibration time. Two leaching
solutions are specified for use with this test. Leaching Solution No. 1 is an acetate buffer prepared with 5.7 mL of
glacial acetic acid per liter of distilled deionized water which is adjusted to pH 4.93 with 1 N sodium hydroxide
solution. Leaching Solution No. 2 is an acetic acid solution prepared by diluting 5.7 mL of glacial acetic acid to one
liter with distilled deionized water. This solution will have a pH of 2.88. The TCLP specifies a test to determine the
alkalinity of the waste to be leached which, in turn, determines which leaching solution should be used. More
alkaline materials utilize Solution No. 2, while less alkaline materials are leached with Solution No. 1.
SYNTHETIC GROUNDWATER LEACHING PROCEDURE
The SGLP10 was developed as a generic leaching test to be applied to materials to simulate actual field leaching
conditions. Since the TCLP was designed to simulate leaching in a sanitary landfill under codisposal conditions, it
is not appropriate to evaluate leaching of CCBs in typical disposal or utilization scenarios. To provide more
appropriate and predictive information for CCBs and other unique materials, a leaching test was developed using
the same basic protocol as the TCLP, but allowing for the appropriate leaching solution chemistry. Test conditions
are end-over-end agitation, a 20-to-1 liquid-to-solid ratio, and an 18-hour equilibration time. The leachate often
used is distilled deionized water.
For certain predictive applications, this may not be totally appropriate, since mercury, for example, would likely be
influenced by the presence of chloride, leading to the formation of an extremely stable mercury chloride complex.
Local, site-specific factors, such as the presence of significant halide concentrations or other geochemical factors
likely to influence trace element mobility, would have to be considered in any real disposal setting. Additionally,
because of the extremely alkaline nature of most low-rank coal combustion ash and their high acid neutralization
capacity beyond the simple high pH, acidity from the impact of varying acid precipitation concentrations is
generally not considered to be an important factor, although, as with every imaginable factor, it would, no doubt,
influence results to some small degree. The purpose of this test is to provide data that are not influenced by the
presence of acetate ion or the initial acid impact when sample and leaching solution are mixed. The composition
of leaching solution is site-specific. In the original test applied to disposal in sites in central and western North
Dakota, a composition of leaching solution designed to simulate sodium sulfate bicarbonate-buffered water was
used. The solution used for leaching was prepared by dissolving 0.50 grams of sodium sulfate and 1.00 gram of
sodium bicarbonate in 1 liter of distilled deionized water.
The analysis of this synthetic groundwater leaching solution is as follows:
Na 436 mg/L
SO4 338 mg/L
HCO3 726 mg/L
pH 8.3-8.7
This composition is typical of groundwater in central and western North Dakota where the water is slightly alkaline
as a result of bicarbonate buffering, and primary mineralization is from sodium sulfate.
69
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
In another research project designed to simulate leaching of coal conversion solid residues in Indiana11, a solution
was prepared with the following nominal composition:
Na
Mg
Ca
CO3
SO4
PH
120 mg/L
310 mg/L
500 mg/L
300 mg/L
2810 mg/L
7
Although the alkalinity is expressed as carbonate, at this pH it would be present primarily as bicarbonate.
In practice, with many CCBs it is not necessary to add calcium, since the solubility of calcium in the leaching
system is determined by the ash contributions from leaching solution, which are negligible compared to the mass
of calcium available from the ash at a 20-to-1 liquid-to-solid ratio.
LONG-TERM LEACHING
A LTL procedure, also using distilled deionized water or a synthetic groundwater, can be used to identify effects
associated with any mineralogical changes that may occur in the waste forms upon long-term contact with water.
It was found in a previous research project3 that on long-term contact with water, certain coal conversion solid
waste materials form secondary hydrated phases with mineralogical and chemical compositions different from any
of the material in the original ash. It was also demonstrated that the formation of these hydrated phases was often
accompanied by dramatic decreases in solution concentrations of oxyanionic species such as borate, chromate,
selenate, and vanadate. The decrease in concentration of these elements would not be predicted from the results
of short-term leaching tests.
In the context of the SGLP, the LTL procedure is simply a continuation of the SGLP. In practice, several SGLP
leaching containers are prepared and rotated. One is sampled at 18 hours, thus fulfilling the SGLP requirement;
another is sampled at 30 days; and a final container is sampled at 60 days. In practice, additional containers can
be started and continued for 90 days, 120 days, or any time duration that is desired. The containers are placed on
the rotator in stages so that all of the
containers are equilibrated at the same
time. Thus one container is started; 30
days later, a second container is started;
and 18 hours before the test is to end, a
final container is started. This simplifies
the analytical process and results in
considerable time savings.
25.00
13.50
c
g
o>
o
<§
0.50
0.25 -
o.oo
Zn Zr
ISGLP EZaTCLPNo. 1 EUD TCLP No. 2
I Max.
RESULTS AND DISCUSSION
The results of short-term and long-term
leaching of several CCBs are shown in
Figures 2 and 3.
Figure 2. Comparative Leaching of
Bituminous Fly Ash
Results for these two ash types are included to illustrate several important leaching trends. Figure 2 illustrates two
of them. First, it can be seen in the results for vanadium that the acidic solutions used in the TCLP produced
solutions with concentrations of vanadium lower than those of the alkaline SGLP This is not an isolated case and
indicates that acid is not a worst-case scenario; rather, it is the phase location of trace elements that determine
acid or base solubility of the analytes of interest. Second, it can be seen that although analyte was mobilized for
most of the trace elements measured at above detection limits, concentrations were still well below the maximum
calculated concentrations indicated by the designation "Max." in the figure. Maximum calculated concentrations in
these figures are theoretical maximum concentrations, assuming total dissolution of the ash at a 20'1
70
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
liquid-to-solid ratio. Total dissolution of most ash samples is highly unlikely, even within geological time intervals.
Figure 3 shows long-term leaching trends of a lignite fly ash. This highly alkaline ash sample with measured pH
values of over 11.5 is what would be considered an ettringite-forming ash. This is suggested by the trends of
decreasing concentration for boron and selenium. These are two of the trace elements that exist as oxyanions in
aqueous solution that are known to
substitute into the ettringite structure.
Although vanadium, chromium, and to
a lesser extent, arsenic, are also
known to be removed by ettringite, it
appears that this was not significant in
this example. It is not known, how-
ever, what the concentrations of these ==-
three elements would have been in E
the absence of ettringite formation, so
this is merely an assumption.
20.00
16.18
12.36
8.54
Leaching of
18 hours
30 days
60 days
Calculated Maximum
RCRA Limit
0.30
0.15
0.00
Figure 3. Long-Term
Lignite Fly Ash
As B Cr Se V pH
SUMMARY
One of the more important conclusions that can be drawn from leaching to predict environmental impact is that
there are currently no laboratory leaching tests available that will reliably and consistently predict the concentration
of analytes in field leachates at coal ash monofill sites. This does not reduce the value of laboratory leaching;
rather, it should influence the way in which laboratory leaching is interpreted and perhaps used in future studies.
Laboratory leaching is a means of generating input data for models to predict field leachate concentrations.
Laboratory leaching will provide information regarding the mass of easily mobilized analyte as well as leachate
concentration trends. Concentrations of analytes in field leachate at the source could be calculated using
information on water infiltration and the permeability of the disposed material. Leachate concentrations at the
source of generation are of limited value, considering the effects of sediment attenuation, dispersion, diffusion,
and dilution as leachates travel through the subsurface environment. Leaching information combined with batch
sediment attenuation experiments to determine numerical values for chemical and physical attenuation, along with
factors for other phenomena leading to decreases in solution concentration of analytes, suggest a means to
predict field leaching without the complication of thermodynamic models that often omit important information such
as secondary hydrated phase formation and sorption of iron and aluminum oxide- hydroxides, as is the case with
coal ash leaching models. Considering the complications involved with predicting sorption and chemical
incorporation of analytes in these important concentration-reducing mechanisms, thermodynamic models may not
always be the best approach for predicting field effects. This is not to say that thermodynamic models can not be
developed for reliable prediction.
The implication for leaching characterization is straightforward. In ash characterization where secondary hydrated
phase formation reactions and release of iron and aluminum that can form highly sorptive materials may ultimately
control the concentrations of numerous environmentally important trace elements, short-term leaching is clearly
insufficient for predicting leachate concentrations and, thus, trace element mobility. Because of the importance of
decisions made on the leachability of potentially problematic trace elements such as those with the potential to
substitute into the ettringite phase, decisions made on the basis of short-term leaching are likely flawed.
Considering the importance of the potential impact of these trace elements on the environment, the overestimation
or underestimation of their mobility could be an invitation to disaster — either environmental, in the case of
overestimation where a problem is missed, or financial, where a nonexistent problem is projected, leading to costly
and unnecessary efforts to protect the environment from a nonexistent problem.
REFERENCES
1 . 1997 Coal Combustion Product (CCP) Production and Use; American Coal Ash Association, 2760 Eisenhower
Avenue, Suite 304, Alexandria, VA 22314-4569.
2. Waste from the Combustion of Fossil Fuels - Volume 1, Executive Summary; U.S. EPA Report to Congress
EPA 530-5-99-010; March 1999.
3. Stevenson, R.J.; Hassett, D.J.; McCarthy, G.J.; Manz, O.E. Solid Waste Codisposal Screening Study; Topical
report to the Gas Research Institute and Radian Corporation, North Dakota Mining and Mineral Resources
71
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Research Institute: Grand Forks, ND, 1986.
4. Dunn, P.J.; Peacor, D.R; Leavens, P.B.; Baum, J.L. Am. Mineralog. 1983 68 1033-1037.
5. Moore, A.E.; Taylor, H.F.W. Nature 1968, 218, 1085.
6. Moore, A.E.; Taylor, H.F.W. Acta Cryst. 1970, 626, 383-393.
7. Moore, A.E.; Taylor, H.F.W. Mm. Mag. 1973, 39, 377-389.
8. Hassett, D.J.; Thompson, J.S.; McCarthy G.J. Ettringite Formation as a Fixation Technology for Immobilizing
Trace Elements; Internal Report 5089-253-1829 for the Gas Research Institute; EERC publication, June 1992.
9. U.S. Environmental Protection Agency. Federal Register 1986, 51 (9), 1750-1758.
10. Hassett, D.J. A Generic Test of Leachability: The Synthetic Groundwater Leaching Procedure. In
Proceedings of the Waste Management for the Energy Industries Conference, University of North Dakota,
April 29-May1, 1987.
11. Hassett, D.J. Evaluation of Leaching Potential of Solid Coal Combustion Wastes; final report to the Indiana
Coal Council, Dec. 4, 1991; 40 p.
EFFECT OF ZERO VALENT IRON ON EXTRACTION OF LEAD, ZINC AND COPPER IN THE TCLP
Douglas S. Kendall
National Enforcement Investigations Center, Office of Enforcement and Compliance Assurance (OCEFT),
US EPA, Building 53, PO Box 25227, Denver Federal Center, Denver, CO 80225
Abstract
The presence of metallic iron in a TCLP extraction can dramatically change the concentration of lead and other
metals in the extract. Wastes, which exhibit the characteristic of toxicity due to the presence of extractable lead,
can pass the test if iron is added to the waste before TCLP testing. The reason for this is the reduction by iron of
lead (II) ions. Using results from TCLP tests of waste casting sand from a brass foundry and related experiments,
this paper will discuss this redox reaction. The possibility of hydroxide precipitation, and adsorption by hydrous
ferric oxide, will also be addressed. pH is important to each of these three possibilities. Lead, copper, and zinc
behave differently with respect to oxidation/reduction, adsorption, and hydroxide precipitation, and their measure-
ment allows deductions as to which mechanism is operating. Iron treatment does not result in long term stabiliza-
tion of a waste placed in the ground, and this will be illustrated by results from actual landfill samples. Wastes
which are treated to pass the TCLP test, but are not permanently stabilized, are an area of concern.
Introduction
The TCLP test is used to determine if wastes exhibit the characteristic of toxicity. If an extract of the waste
contains a regulated element or compound in a concentration greater than the limits in the regulation, then the
waste exhibits the characteristic of toxicity. As the principal test for toxicity, it is obvious that many waste streams
have been subjected to the TCLP test, and that the results of this testing have a large economic impact. Since the
test can be so important, it is not surprising that many waste treatments, and additions to industrial processes,
have been designed to affect the outcome of the TCLP test.
As a technical support center for EPA enforcement, the laboratory of the National Enforcement Investigations
Center (NEIC) has examined a number of additions to wastes designed to "beat" or pass the TCLP test. Not all
these additions can be considered treatments in the sense of conferring long term stability. This account will
describe one such treatment, the addition of iron to brass foundry waste.
Brass Casting Using Sand Molds
Brass can be formed into many useful items by casting the molten metal into sand molds. Numerous foundries
make a wide variety of metal parts this way. Information and data from one foundry is presented in this report, but
the results are applicable to many foundries using the same process and generating the same wastes. This infor-
mation was generated in support of an enforcement investigation, but only the technical aspects will be discussed.
The brass foundry under study prepared brass valves for use in applications such as drinking water systems The
brass used at the foundry had a composition as follows:
72
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
copper 80% to 88%
zinc 5% to 9%
tin 3% to 6%
lead 1.5% to 7.0%
The source of the lead and some of the rest of the brass was recycled automobile radiators. The lead is a desir-
able component since it aids in machinability of the brass. It is surprising that lead is still allowed in the brass used
for drinking water valves. Lead is not allowed in the solder used for drinking water systems. It is possible to make
brass valves without lead, but the foundry indicated that it would increase their costs.
Waste Sand
The molding sand used in brass casting is used many times, but eventually it must be replaced. Thus waste sand
is continually being generated. Since the sand has been in contact with molten brass, and exposed to metallic
vapors, it becomes contaminated with the components of the brass. In particular it contains significant amounts of
the principal components as follows:
lead 1500 to 6000 mg/kg
copper 1%to3%
zinc 1%to6%
TCLP extractions performed in our laboratory showed that the extract typically contained about 50 mg/L lead.
Copper and zinc were even more concentrated in the extract.
Iron Addition
As many foundries have discovered, the addition of iron metal, of zero valent iron, can profoundly affect the
outcome of a TCLP extraction involving lead1. The recipe used in this particular case was 10% iron by weight. The
iron was added to the waste sand in the form of filings or shavings procured as waste from a machining operation.
The TCLP extracts of the waste after the iron addition contain less than the regulatory limit of 5 mg/L of lead.
Copper was similarly diminished in the extract. Zinc was not significantly diminished by the presence of iron.
The explanation for the observed effects of iron addition are straightforward. A partial listing of the electromotive
series follows:
Na
Al
Zn
Fe
Cd
Pb
Cu
The higher an element is on the list, the easier it is to oxidize, and the harder to reduce. The nearer an element is
to the bottom of the list, the harder it is to oxidize, and the easier to reduce. The electromotive series is a way of
summarizing electrode potentials, and predicting oxidation reduction reactions. From the list, or from electrode
potentials, it is clear that metallic iron will reduce lead (II) or copper (II) ions to the zero valent states, which are
essentially insoluble. Iron will not reduce zinc (II) ions. The iron itself will be oxidized. As long as metallic iron is
present, either in the TCLP bottle or in a landfill, any lead ions appearing in solution will be reduced by the iron.
The concentration of lead ions will not reach the regulatory limit of 5 mg/L.
Why is iron treatment not a long term solution? Simply because iron metal cannot be expected to remain in a
landfill without oxidizing. The time required to completely oxidize all the iron present will depend upon the climate.
Oxidation will be faster in a wetter and warmer location than in a drier and colder one.
This theoretical prediction that oxidation will occur, and that lead will eventually become extractable, is confirmed
by actual samples and testing. Waste sand from the foundry, which is being described in this report, was placed in
a municipal landfill. It was placed in separate cells, so that it was not mixed with municipal waste. A drill rig was
used to collect samples of the waste sand after it had been in the ground for several years. The cores were
divided into sections, and TCLP tests and total elemental measurements were made. The TCLP results for one
core follow, listed in order from top to bottom.
73
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Location
Top of Core
Bottom of Core
Lead in TCLP extract (mg/L)
5.9
0.22
0.75
2.4
5.7
65.
Three of the samples exceeded the limit for extractable lead. All of them would have passed the TCLP test easily
when they were first placed in the landfill - total iron measurements showed that they had been treated with iron. It
is thus clear that, as expected, the iron treatment was by no means a permanent means of stabilizing lead. Also,
iron treatment has little effect on zinc, whose leachability should be of concern.
Adsorption by Iron Oxides
It is well known and well studied that iron oxides and hydroxides can adsorb lead, copper, and zinc ions under
certain conditions2. If ferric ions are in solution, and the pH is raised, then hydrous ferric oxide will precipitate. This
is an amorphous phase which incorporates considerable amounts of water. As it ages, hydrous ferric oxide (HFO)
converts to crystalline iron oxides, but not in the time period of a TCLP test. If lead, zinc, or copper ions are
present, they can compete with hydrogen ions for sites on the surface of the HFO. Thus as pH is increased the
fraction of the metal ions adsorbed increases. When the fraction of a particular ion which is adsorbed is plotted as
a function of pH, there is a sharp transition, a pH edge, from no adsorption to complete adsorption. Studies cited in
the reference show that HFO adsorbs lead more strongly than zinc, and zinc more strongly than copper. In other
words, lead is adsorbed at a lower pH than zinc or copper. Tests in our laboratory confirmed this under TCLP
conditions. A distinction can thus be made as to which mechanism is operating. A redox reaction reduces copper
to lower levels than lead, while the opposite is true for adsorption on HFO.
Will HFO be formed from iron filings during a TCLP test? This is unlikely, for the following reasons. Iron metal
added to a waste in a TCLP test will surely oxidize. However, in the absence of oxygen, ferrous iron is the most
stable state of iron. And in a well sealed TCLP bottle, all the oxygen will soon be used up by the iron oxidation.
Dissolved iron will be in the ferrous state. There will be no ferric iron to form HFO. This has been observed in
experiments at NEIC. TCLP extracts to which iron has been added typically show several hundred ppm of
dissolved iron. This must have been in the form of iron (II) during the extraction, since at the observed pH's ferric
iron would have been much less soluble. As the extracts are filtered, exposing them to air, visual observation
shows that the ferrous ions are rapidly oxidized, and HFO forms.
Perhaps the more important question is what happens to the iron treated waste as it sits in a landfill. If the water
which percolates through the waste is oxygenated, the iron filings could very well form HFO, and the HFO could
adsorb lead, zinc, and copper ions. There are several problems with this scenario of permanent treatment. If the
pH gets much below five the ions will desorb. If the local environment becomes anaerobic, the iron will be reduced
to the ferrous state, and the dissolved iron and other ions will be carried away by the groundwater flow. There may
be a place for iron oxide adsorption, but certainly not as a permanent treatment of lead-containing wastes.
Hydroxide Precipitation
The hydroxides of many metals, including copper, lead, and zinc, will precipitate if the pH is raised higher than
about 7 3 Hydroxide precipitation can lower the concentration of lead to below the TCLP regulatory limit. The
exact pH at which hydroxides will precipitate depends on the metal and other factors. Complexing agents can
raise the pH at which hydroxides will precipitate. The acetate buffer present in the TCLP test does tend to solubi-
lize metals at higher pHs than otherwise would be possible.
Hydroxide precipitation does not explain the observed effects of iron treatment of lead-contaminated wastes
While the presence and oxidation of iron metal does raise the pH of TCLP extraction fluid number one above 5 it
does not raise it nearly high enough to precipitate lead, copper or zinc as the hydroxides, especially with acetate
74
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
present. Of course, with TCLP extraction fluid number two the pH would be even lower than with fluid one, and
hydroxide precipitation would be even less of a factor.
Other Treatments
Other treatments used on wastes have as their main effects an influence on the outcome of the TCLP test. Lime
treatment is a common example. If enough lime is added to the waste, then any lead, zinc, copper, and a number
of other elements will precipitate as hydroxides. Whether this lime treatment will also serve as a permanent treat-
ment for the landfilled waste is far from certain. The length of time which the waste remains stabilized surely
depends on many factors, such as the amount and acidity of water which percolates through the waste. A large
burden is placed on the TCLP test, requiring its use as the principal determinant of the suitability of waste treat-
ment. Whether materials containing well over 1000 mg/kg of lead should be considered nontoxic just because
they pass the TCLP test is an open question. Perhaps it is time to consider additional tests to determine toxicity.
An obvious and simple method is to consider total amounts. In addition to leachable amounts determined in the
TCLP test, total concentrations in a waste should be used to determine the characteristic of toxicity.
Summary
The reason iron addition to lead-contaminated waste reduces the lead level in the TCLP extract to below the
regulatory limit (copper is also diminished) has been shown to be an oxidation reduction reaction. Two other
mechanisms which can reduce solution levels of lead, copper, and zinc in some situations are adsorption by
hydrous ferric oxide (HFO) and hydroxide precipitation. The relative concentrations of lead, copper and zinc, as
well as the pH, can be used to distinguish between these mechanisms.
The TCLP test certainly has a role in characterizing hazardous wastes. The question is whether it should be the
sole test. Experience has shown that the TCLP test by itself is not sufficient to establish long term stability of
treated hazardous waste placed in a landfill. Additional tests, and perhaps additional regulations, may be neces-
sary to compel the regulated community to concentrate on permanent, long-term treatment of hazardous wastes,
rather than focusing on passing the TCLP test.
Acknowledgments
Helpful discussions and cooperation with John Drexler of the University of Colorado in Boulder and with Joe Lowry
of NEIC are gratefully acknowledged. Laboratory assistance from Robin Ingamells is very much appreciated.
References
1. T.R. Ostrom and P.K. Trojan, "Lead Compound Formation in Alloy CDA 84400 and the Effect of Iron on
Leachable Lead.MFS Transactions, 94, pp. 725-734, 1986.
2. D.A. Dzombak and P.M. Morel, Surface Complexation Modeling, Hydrous Ferric Oxide, John Wiley, 1990.
3. J.F. Pankow, Aquatic Chemistry Concepts, Lewis Publishers, 1991, pp.219-242.
NEW DEVELOPMENTS OF METHOD 7473 FOR MERCURY ANALYSIS
Theresa M. Serapiglia
Graduate Assistant
Helen M. Boylan
Graduate Assistant
H.M. 'Skip' Kingston
Professor
Bayer School of Natural and Environmental Sciences, Department of Chemistry and Biochemistry,
Duquesne University, 600 Forbes Avenue, 308 Mellon Hall, Pittsburgh, PA 15282
ABSTRACT
EPA draft Method 7473 is a new technique based on traditional methodologies operating on the basis of thermal
decomposition, amalgamation, and atomic absorption spectrometry. With Method 7473, sample preparation and
analysis are essentially integrated into a single analytical instrument, allowing for direct analysis of both solid and
liquid samples. Direct analysis gives Method 7473 the capability to be applied in either laboratory or field settings.
75
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
This method has been previously validated for use with traditional environmental matrices in both the lab and
field1. There is a need, however, to extend this method to include the analysis of unique sample types and mercury
species. This paper will discuss the determination of total mercury in coal and other fossil fuels, fish tissue, and
additional significant matrices. Extraction techniques to be coupled with Method 7473 for the analysis of
operationally-defined mercury species will also be detailed.
INTRODUCTION
As a RCRA-regulated element, mercury is routinely analyzed in soil, water, and other environmental matrices. The
most common method for mercury analysis used today is the cold vapor technique, which originated in the late
1960s2 The cold vapor technique is problematic, as it is a very time-consuming and labor-intensive process and
also requires a significant amount of reagents for sample preparation. Method 7473 is able to minimize these
problems associated with the traditional technique and is very adept at routine environmental mercury monitoring.
The next phase of method development is expansion of Method 7473 to include other matrices and mercury
speciation.
Regulations are typically a main driving force behind advancements in method development. Recent EPA initia-
tives have sparked an interest in analyzing coal and its by-products for mercury content. Information Collection
Request no. 1858.01 requires coal-powered electric utilities to report the mercury content in coal, fly ash, and
stack gases on a monthly basis. Some states are proposing mercury emission regulations in other fossil fuels,
such as oil and gasoline. A simple and rapid method for mercury analysis in a broad range of fuel sources is there-
fore desired.
Since the Minamata Bay tragedy of the 1950s where hundreds of people were poisoned as a result of consump-
tion of mercury-contaminated seafood, mercury has been regulated in food products. An action level of 1 ppm
mercury in fish has been set by the FDA. With the ability to perform 'dockside' analysis, Method 7473 has tremen-
dous potential in the fishery industry.
Because the toxicity and mobility of an element is dependent on its chemical form, the trend in environmental
monitoring is shifting from total to species-based measurements. The EPA has recognized the need for a reliable
measurement technique for mercury speciation. Use of selective extraction for separation of operationally-defined
mercury species has been reported3 Coupling such an extraction procedure to analysis by Method 7473 will allow
for rapid (and potentially on-site) characterziation of mercury speciation.
SUMMARY
The analysis of coal and other combustible materials containing a high organic content by Method 7473 is not as
straightforward as that of standard environmental samples. An exothermic oxidation occurs during decomposition
due to the presence of oxygen as a carrier gas. Modifications to the analytical parameters have been evaluated,
including the use of nitrogen and air as a carrier gas. These less oxidizing gases eliminated the pyrotechnics, but
also changed the chemistry of both the catalyst and the amalgamation processes of the instrument. A discussion
of the chemical parameters that control these processes will be presented to evaluate the use of alternative carrier
gases. The optimization of Method 7473 for the analysis of such difficult matrices will be discussed. Data collected
on a variety of coal samples, fish tissue, and other matrices will be presented.
Selective extraction coupled to Method 7473 will be evaluated for the characterization of mercury species. The
species will be operationally-defined based on their respective solubilities. Examples of operational definitions are
provided in Table 1.
Table 1. Operationally-Defined Mercury Species
Operational Definition Individual Species Example Relative Toxicity Relative Mobility
Soluble in Organics Methylmercury High ~Low
Water-Soluble Mercuric nitrate Low High
Acid-Soluble Mercuric sulfide Low Low
While selective extraction will not provide results based on individual species, it can provide information on groups
of species, allowing for a relatively quick risk assesment based on mercury speciation. Refinement of the extrac-
tion procedure for the matrices and species of interest will be discussed.
76
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
REFERENCES
1. Boylan, H.M., H.M. Kingston, and R.C. Richter. Direct Mercury Analysis: Field and Laboratory Applications, in
Waste Testing and Quality Assurance Symposium. 1998. Arlington, VA.
2. Hatch, W.R. and W.L. Ott, Determination of Sub-Microgram Quantities of Mercury by Atomic Absorption
Spectrophotometry. Anal. Chem., 1968. 40(14): p. 2085-2087.
3. Miller, E.L., Draft Method: Determination of Organic, Inorganic, and Total Mercury in Soils. SW-846, 1993.
SPECIATION OF MERCURY IN SOIL
Stuart J. Nagourney
New Jersey Department of Environmental Protection, Division of Science, Research and Technology
Brian Buckley and Eric Fisher
Rutgers University, Environmental Occupational Health Sciences Institute
Introduction
The species of a metal contaminant will often determine its fate and transport. Species can mean the oxidation
state, crystal structure, or mineral form of the metal. In the environment, organic mercury species can both
bio-accumulate and bio-magnify through the ecosystem. Methylmercury is the most prevalent form and can be
found at concentrations more than 106 times as concentrated in a fish as it is in the water in which it lives. Mercury
species associated with natural organic agents such as humic acids seem to be difficult to extract even under
vigorous conditions.
The contaminant species can also play a role in its fate and transport within the organism. Chromium in the dichro-
mate species can be taken up by the body at a much faster rate than the Cr3+ form which is thought to be a micro-
nutrient for humans. The hexavalent form is a known carcinogen. The procedure for determining the concentration
and nature of mercury species in soils is to expose the soil to a series of solvents, with the intent of having each
solvent selectively remove a different mercury form. The literature reports more than 10 different combinations of
extracting media, each applied to a different soil type. Results are "operationally defined"; e.g., the part of the
extraction procedure intended to remove organic mercury is not stated to remove all of the organic species; rather,
the amount removed is defined as the organic fraction. Several studies report the re-distribution of mercury
species during extraction, further complicating data interpretation. A reliable, predictable method for identifying the
in situ amounts of various forms of mercury in soil is highly desirable from both a scientific and a regulatory
perspective. The objectives of this study are to:
• compare existing methods for the sequential extraction of mercury in soil using various aqueous and
mixed liquid phases to speciate mercury in contaminated soil, and
• develop a method for the sequential extraction of mercury in contaminated soil with defined values for the
precision and accuracy for the elemental, inorganic and mercury species.
An evaluation of various solvent mixtures and chromatographic elution schemes was conducted to explore the
capability of extracting various mercury species. Ion chromatography (1C) was employed to separate the various
mercury species. Analysis of the eluent was performed by direct coupling to an Inductively Coupled Plasma Mass
Spectrometer (ICPMS ), which allowed the concurrent measurement of various mercury forms. The procedure can
correct for extraction and pre-concentration efficiencies as low as 30% using stable enriched isotopes, and adding
two isotopes simultaneously can be used to monitor specie interconversion.
Experimental
The chromatographic system consisted an ion chromatograph fitted with a 400 ul sample loop. An 0.45 urn inline
filter is placed between the injection loop and the column to prevent column clogging. The mobile phase consists
of a methanol/HCI (1M) mixture in a ratio of 55:45 (v/v). A flow rate of 1 ml/min was used for all determinations.
After the analyses were completed the column was stored in a 10 % solution of the mobile phase diluted with 18
Mohm water.
The output of the 1C was connected directly to the Meinhard nebulizer. The detection unit was an ICPMS. Argon
77
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
was used as the plasma support gas. The ICPMS sensitivity was optimized with a solution of 10 ppb Hg(NO3)2 with
the mobile phase used as the solvent. The ICP was optimal operated at 1.3 kW forward power and the coolant,
auxiliary and nebulizer argon flow rates were 14, 0.8 and 0.8 l/min, respectively. The data were collected unless
otherwise indicated by single ion monitoring of mass 202 (the most abundant isotope for mercury) and the signals
were calculated using a peak integration time of 21 sec.
Methanol, hydrochloric acid and nitric acid were of analytical grade and were used without further purification.
Water was of HPLC grade and was delivered by a Millipore water purification system. Standard solutions of
mercury (100 ppm Hg(NO3)2 in 5 % hydrochloric acid) and methylmercury (100 ppm in 55 % methanol and 2 %
hydrochloric acid) were used. The analytical solutions with different concentrations were made from these stock
solutions by diluting with the mobile phase. Isotopically-labeled mercury standards were prepared from enriched
mercury oxide. The 200Hg2+ standard solution was made by reacting the mercury oxide powder with concentrated
HCI and diluted with 5% HCI. Methylmercury was formed from 202Hg using a methylcobalamine reaction.
The microwave extractions of soil samples were carried out using a microwave digestion system. 500 mg of a soil
sample, known to ± 0.01 g, was weighed in a polypropylene volumetric flask and subsequently 5 ml of methanol
and 100 ul of 6 M hydrochloric acid were added. A microwave digestion program consisting of 30 W for 20 min
was applied. After cooling to room temperature (25 °C), the sample was diluted to 10 ml using 18-M water. This
solution was then injected through a 0.45 urn syringe filter into the IC-ICPMS analytical system. Recoveries were
determined by adding known amounts of the mercury species to the soil.
Results and Discussion
Methanol fractions between 50 and 70% and HCI concentrations ranging from 0.4 to 1.2 M were used to minimize
analysis time while achieving complete separation. The acid concentration effected the Hg2+ to a much greater
extent than the methylmercury. Going from 0.4 M to 1.2 M HCI eluent decreased the Hg2+ ion retention time 4
times and increased the signal 5 times, whereas the retention time and the sensitivity of the organomercury signal
were only slightly changed. This was expected, because the inorganic mercury ion is doubly charged. All further
studies used a 1 M HCI eluent concentration. While the inorganic ion retention time remained unchanged, the
methylmercury retention time decreased with increasing methanol fractions. Increasing of the organic content also
resulted in a decrease in analyte sensitivity. Nebulization of volatile solvents such as methanol extinguishes the
plasma or causes it to be unstable. A mobile phase composition of 1 M HCI and 55% methanol enabled the
species to be fully resolved with complete separation in less than 7 minutes.
The linear range and the limits of detection of the ICPMS response for inorganic and methylmercury were deter-
mined by using a mobile phase composition of 1 M HCI and 55% methanol and a integration time of 21 sec. The
calibration graph is linear from the detection limit to the low parts per million level by using single ion monitoring of
mass 202. The linear range could be extended further by switching to a less abundant isotope of mercury, but
wasn't done because of contamination concerns. The limit of detection, defined as three times the standard devia-
tion of 6 repeated scans of 500 ppt of inorganic and 500 ppt of methylmercury was 15 and 50 ppt, respectively.
The determination of inorganic and methylmercury in soil samples was performed using the techniques described
above. Mercury-free soil was spiked with 1 ppm of each isotopically labeled mercury species (200Hg2+ and Me202
Hg+) and left untouched for at least 3 days. The influence of various parameters such as amount and composition
of the extraction solvent as well as the power and time program of the extraction procedure was investigated to
obtain the best recovery rates of the mercury species. Preliminary studies found that using either hydrochloric or
nitric acid for the methanolic extraction eluent, the same results were obtained. For further experiments, a
methanol/HCI mixture was used. With increasing methanolic content the recovery for the organic mercury is
increasing constantly whereas the rates for the inorganic species decreased similarly. Using 2 ml of HCI as the
extraction solvent, the calculated recoveries for inorganic mercury were 200%, whereas the methylmercury could-
n't be recovered at all. The reason for this result is that all the methylmercury got converted into the inorganic form
as the separate monitoring of mass 200 and 202 showed. Experiments performed using 100% of methanol for the
extraction recovered 75% of the organic species and 25% of the inorganic mercury. The recovery rate for the
organic species increase from 60 to 85% with increasing amounts of methanol up to 5 ml; the recovery of the
inorganic mercury increases even more, from 10 to 50% before it leveling off. Decreased recoveries for the
organic species using more than 5 ml of methanol is due the saturation of the methylmercury in the solvent. For
subsequent experiments, 5 ml of methanol was used as extraction solvent. The chromatogram is shown in Figure
In order to improve the efficiency of the method for the inorganic species, different amounts of 6 M HCI were
78
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
added to 5 ml of methanol, the soils extracted and the recovery rates calculated. Going from up to 100 ul HCI the
recovery rates for the inorganic and organic mercury increase from 50% to 107% and from 80% to 93%, respec-
tively. Adding more acid (1 ml of 6 M HCI) converts more organic mercury to the inorganic form, leading to
elevated recoveries for inorganic mercury (140 %) and decreased methylmercury recoveries (75%). Using a
mixture of 5 ml of methanol and 100 pi of 6 M HCI as extraction solvent, only 5% of methylmercury was converted
to inorganic mercury, and recovery rates were found for the both species of 93 and 107 %, respectively. The
power of the microwave was varied between 10 and 80 W and each level was run for periods of time between 5
and 30 minutes. The obtained recovery rates under the different microwave conditions are summarized in Figure
2. The best results were achieved by running the microwave for 20 min. at 30 W power. Under these conditions
the inorganic and the organic mercury could be recovered with 98 and 99% efficiency and only 1% of the organic
mercury was converted to the inorganic form. It was found that the conversion process increases drastically by
running the microwave with more than 40% power and especially longer than 20 min.
The linear range and the limits of detection for the determination of the two mercury species are determined by
spiking soil samples with different concentrations (between 50 ppb and 50 ppm) of the mercury species and
extract the samples under the conditions described in the experimental section. For each concentration, at least 3
soil samples were spiked and their extracts were injected into the analytical system 3 times. Under these condi-
tions the inorganic and organic mercury could be recovered with 97% and 96% efficiency respectively over a
concentration range of 50 ppb to 50 ppm. Using higher volumes of the extraction solvent could extend the linear
range. The deviation of the retention time was less than 5%. Calibration graphs based on peak areas were linear
(correlation coefficients better than 0.999) for each compound in the range tested. The detection limit, defined as
three times the standard deviation of 9 repeated scans (3 injections per soil extract) of 50 ppb each of inorganic
and organic mercury was 3 and 15 ppb, respectively.
The method described above was applied to three different soils. The accuracy of the method was tested by
analysis of three different certified soil samples (obtained from the National Institute of Standards and Technology,
NIST). All three soils contained certified concentrations of inorganic mercury in a range between 1.4 and 32 ppm
(there is no certified soil that contains methylmercury). The soils were extracted and the mercury species deter-
mined using the optimized microwave technique in combination with the IC-ICPMS method. The found concentra-
tions (n=6) for inorganic mercury compared very well with the certified values, as shown in Figure 3.
Conclusion
A rapid and efficient procedure is described for the
quantification of inorganic and methylmercury in soil
samples by employing microwave extraction and
subsequent IC-ICPMS detection. Under the optimal
microwave conditions the inorganic and organic
mercury could be recovered with 97% and 96%
efficiency over a concentration range of 50 ppb to 50
ppm. The limits of detection for the inorganic and
organic species using a methanol/HCI (1M) eluent
(55:45, v/v) were found to be 3 and 10 ppb, respec-
tively. The method was successfully employed for the
determination of inorganic mercury in 3 certified soil
samples.
Figure 1.
79
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Figure 2.
Figure 3. Comparison of the found concentrations of inorganic mercury in 3 different soil samples with the certi-
fied NIST values. (For chromatogaphic, ICPMS and extraction conditions see Experimental)
NIST#
2710
2704
2709
Soil type
Montana Soil High Traces
Buffalo River Sediment
San Joaquin Soil Baseline
CERTIFIED
Hg2+ cone, (ppm)
32.6 ppm
1 .47 ppm
1 .40 ppm
FOUND
Hg2+ cone, (ppm)
32.1 ±0.5
1.32 ±0.1
1.30 + 0.1
METHOD DEVELOPMENT FOR SPECIATION ANALYSIS OF MERCURY AND TIN COMPOUNDS IN
STANDARD REFERENCE MATERIALS USING GC-AED AND GC-MS
Silke Tutschku. Michele M. Schantz and Stephen A. Wise
Analytical Chemistry Division, NIST, Gaithersburg, MD 20899
Email: silke.tutschku@nist.gov
Speciation analysis of mercury and tin organic compounds has been a topic of concern among analytical chemists
for several years. Tin compounds, used in anti-fouling coatings and as stabilizing agents in polymers, show a very
high toxicity and are subject to restriction on their use in a number of countries. One of the concerns with respect
to mercury pollution is the investigation of the pathways for its conversion in the environment.
To understand the pathways of these elements in the environment and to avoid the health hazards associated with
them, it is necessary to develop methods for the determination of these compounds at very low concentrations in
different matrices. The hyphenation of high resolution separations, available with modern chromatographic
techniques, coupled with the high sensitivity and selectivity of atomic spectroscopic detection, provides a powerful
tool for speciation analysis. For the monitoring and investigating of those compounds in a wide field of samples
routinely, it is necessary to provide Standard Reference Materials (SRMs).
Analytical methods for the determination of methylmercury, mercury(ll) and buty-tin-compounds in different SRMs
like mussels, sediments, fish and blood samples have been developed. For the separation and detection of the
analytes, gas chromatography (GC) with atomic emission detection (AED) and GC with mas spectrometric
detection (MSD) were used. After optimization of the instrumental parameters, determination of mercury and tin
compounds in the low mg/L-level is possible.
80
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
However, the biggest problem for the analysis of environmental samples is sample preparation. In most cases,
sample preparation is time consuming and extraction recoveries are low. In biological and sediment samples, the
analytes are strongly bound to the matrix, and they have to be released prior to their determination. At the same
time, losses or changes in the composition can occur. Different methods of leaching were used and derivatization
of the species of interest with Sodium tetraethylborate and Sodium tetraphenylborate were investigated. Also
conventional liquid-liquid extraction, Solid-Phase-Micro-Extraction (SPME) - a new sample extraction and
enrichment technique was used and optimized. Analytical variables of the extraction such as fiber coatings,
sorption time and desorption time have been investigated. The methods developed provide rapid and sensitive
determination of mercury and tin organic compounds in sediments and marine organisms.
A UNIVERSAL ICP-OES METHOD FOR ENVIRONMENTAL ANALYSES
Zoe A. Grosser. Lee Davidowski, John Latino, and Douglas Sears
The Perkin-Elmer Corporation, 50 Danbury Road MS-219, Wilton, CT 06897
email: grosseza@perkin-elmer.com
Abstract
Environmental analyses are performed on a variety of matrices such as drinking water, wastewater, and solid and
hazardous waste materials. The metals of interest can vary and the concentrations can range from trace levels to
higher. The methods developed by the various environmental programs have differed in the quality control
required and only slightly in the analyte list. With the move toward method streamlining and performance-based
measurements, it now is possible to consider a universal method that will include a superset of the analytes most
often determined.
The concentration range requirements will vary by element and is very different among matrices. For the method
to be truly universal it must cover the full concentration range requirements covered by several methods or by
different sets of conditions used today. The dual-view capability of an ICP-OES spectrometer can be used to
extend the dynamic range for elements expected to exceed the range offered by analyses using either a radial or
axial view exclusively.
This paper will explore the utility of a universal method for ICP-OES environmental analysis. The analyte list will be
developed and compared with environmental requirements in other countries and US. The linear dynamic range
will be evaluated for a dual-view spectrometer and the precision, time for analysis, and interference correction will
be demonstrated.
Reference materials and real samples will be used to test the capabilities of the developed method. Low level
concentrations in drinking water and wastewater will challenge the method for detection capabilities. Soils and
digested waste samples will challenge the spectral overlap correction abilities.
Once the method is fully developed and characterized, the parameters necessary for speeding up the analysis will
be evaluated. The sample introduction system, washing parameters, and autosampler set up will be optimized and
the general procedures described.
Introduction
The trend in recent years has been for laboratories to push the limits of efficiency and productivity. With the move
towards a performance-based measurement system (PBMS), laboratories will have additional opportunities to
optimize methodology for analytical performance and range of application. This work describes work performed to
develop a universal inductively coupled plasma optical emission (ICP-OES) method for the analysis of a wide
range of environmental matrices.
ICP-OES has been used for environmental measurements for many years, and the applicability expanded with the
use of accessories and new instrumental capabilities. The speed and flexibility of systems have increased, while at
the same time the systems have become more widely available. Methods such as US EPA 200.7, USEPA 6010,
and EN ISO 11885 have been developed by different programs to take advantage of ICP technology. The
methods are similar, but differ in quality control requirements and the European method differs in the list of
81
-------
analytes. Table 1 compares the methods in general terms.
Ideally, in the environmental laboratory, one ICP-OES method would handle all elements and most of the matrices
encountered. The quality control requirements would be uniform and the reporting requirements would be the
same for all samples. The method developed for the instrument should be as fast as possible for two replicate
measurements while meeting the data quality objectives for precision and detection limits. The method should
provide accurate results and the QC should be built-in. Data transfer to third party software or LIMS should be
easily accomplished.
The goals for this study were to develop an ICP-OES method that incorporates the USEPA elements (European
elements will be added later) that covers the full concentration range expected. The performance of the method
was demonstrated on a variety of matrices containing a range of concentrations.
Table 1. ICP-OES Method Comparison
Elements
QC
Concentration Range
US EPA 200.7
32
Initial and Continuing
Water, typically low
US EPA 601 OB
Same 31, no Ce
Initial and Continuing
Solid and Hazardous
Waste, low and high
concentrations
EN ISO 11 885
Same 29, no Tl, Ce, Hg;
Bi, W, Zr, S included
Varies by state in Germany
Water and sludge, used for
other matrices, low and high
concentrations
Experimental
The Perkin-Elmer Optima 3300™ DV ICP-OES equipped with a low flow GemConea nebulizer and cyclonic spray
chamber was used for all determinations. The Optima 3000 DV ICP-OES is a simultaneous ICP with an echelle
polychromator and Segmented-Array Charge-Coupled Detector (SCO). Simultaneous measurement of the
background and analyte emission allows for accurate correction of transient background fluctuations. The instru-
ment can collect data from either the radial viewing configuration or axial configuration or a combination of the two
during a single analysis, this study was performed using the full dual view capability.
The instrumental conditions used for all determinations are shown in Table 2. The nebulizer flow was optimized for
the best detection limits. Calibration standards were prepared from PE Pure multielement and single element
standards. Table 3 lists the wavelengths chosen for the method, background correction points used, and the
standard concentration used for calibration. The background correction points are typical and provide a starting
point for method development. It is likely that individual instruments will require minor adjustment to the points to
optimize the correction. In several cases, two wavelengths are included for evaluation. In method development this
is a common procedure, when the evaluation is complete the more appropriate wavelength or view can be chosen
and the extra wavelength eliminated. Alternatively, since the analysis time is not increased by the addition of
wavelengths, a second wavelength may be retained for confirmation, if desired.
Table 2. Instrumental Conditions
Parameter
RF Power
Nebulizer Flow
Auxiliary Flow
Plasma Flow
Sample Pump Flow
Plasma Viewing
Processing Mode
Auto Integration
Read delay
Rinse
Replicates
Background correction
Optima 3300 DV
1450 watts
0.55 L/min
0.5 L/min
15.0 L/min
1.8 mL/min
Axial
Area
5 sec min -20 sec max
45 sec
10 sec
2
Manual selection of one or two points
82
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Yttrium was used as the internal standard and added on-line to the blanks, standards, and samples. The rinse
solution contained 2% HNO3+ 0.1% Triton-X 100™.
EPA method 200.7, revised in 1994, and EPA method 601 OB, revised in 1996, were used for guidance in develop-
ing the method.1'2
Table 3. Elements, Wavelengths, Correction Intervals, and Calibrations Standards
Element and Wavelength
(nm)
Ag 338.289
Al 308.21 5R
As 188.979
B 182.527
Ba 233.527R
Be 313.107
Ca 227.546
Ca315.887R
Cd 226.502
Ce 41 3.765
Co 228.61 6
Cr 205.560R
Cr 267.71 6
Cu 324.754
Fe 238.204R
Fe 273.955R
Hg 194.168
K 766.490R
Li 670.784R
Mg 279.079R
Mn 257.610
Mo 202.030
Na 589.592R
Na 330.237
Ni 23 1.604
P 178.221
Pb 220.353
Sb 206.833
Se 196.026
Si 251. 61 1R
Sn 189.933
Sr460.733R
Ti 334.441
Tl 190.800
V 292.402
Zn 206.200
Lower Correction Interval
(nm)
-0.052
-0.024
-0.026
-0.055
-0.048
-0.042
-0.054
0.024
-0.019
-0.038
-0.022
-0.031
-0.016
-0.136
-0.124
-0.031
0.036
-0.019
-0.074
-0.030
-0.021
-0.021
-0.020
-0.012
-0.015
-0.023
0.017
0.046
-0.027
Upper Correction Interval
(nm)
0.031
0.045
0.011
0.020
0.034
0.045
0.038
0.019
0.031
0.030
0.025
0.033
0.012
0.032
0.074
0.021
0.013
0.023
0.044
0.041
0.027
0.027
Standard Concentration
(mg/L)
5.0
250
1.0
5.0
5.0
5.0
250
250
5.0
5.0
5.0
5.0
5.0
5.0
1.0, 100
100
0.5
1.0, 100
5.0
250
5.0
5.0
100
100
5.0
5.0
1.0
5.0
1.0
5.0
5.0
5.0
5.0
1.0
5.0
5.0
Samples consisted of soil and sediment digests and wastewater reference materials from High-Purity Standards,
Inc. (Charleston, SC). NIST Drinking Water reference material 1643D, Trace Elements in Water was used for
83
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
validation at low levels.
Results and Discussion
The choice of axial or radial viewing was chosen based on the range of concentrations expected and the detection
limit needed to meet the quality objectives. The detection limits and linear range are shown in Table 4 for a variety
of elements.
Table 4. Detection Limits and Linear Ranges Dual View Method
Element
Fe
Al
Ca
Mg
Na
Pb
Tl
Se
As
IDL(mg/L)
2.2
36
6.4
14
19
1.6
4.2
5.0
6.3
Linear Range (mg/L)
1000
1000
1000
1000
1000
100
100
100
100
Observation
Radial
Radial
Radial
Radial
Radial
Axial
Axial
Axial
Axial
Freedom from interferences is also a consideration and the easily-ionizable element effect (EIE) was considered
for Na and K. Easily ionizable elements such as the alkalis can vary in signal intensity depending upon the concen-
tration of other easily ionizable elements present in the sample. This can cause inaccurate measurements of Na
and K in samples that contain varying amounts of these elements or are calibrated with single element standards.
This type of interference is enhanced in axial-viewing and can be resolved in several ways. An ionization buffer
can be used to minimize the differences between samples (matrix matching). The element chosen for the buffer
must not be required as an analyte element, ruling out the most commonly used element, Li. In addition, high
concentrations of one alkali often contain trace contamination of other alkalis, which may cause unacceptable
inaccuracies in analytical measurements at low concentrations. Another solution is to use an element as an inter-
nal standard that shows a similar effect. Radial viewing of the plasma does not show the same effect and can be
used as an alternative for these elements. Rubidium was evaluated as an internal standard element for K and
compared with yttrium as an internal standard, matrix matching and radial viewing. Table 5 summarizes the results
and shows that, although Rb is a better internal standard than Y for K, if radial viewing is an option it will require
less method development to implement.
Table 5. EIE Compensation Recoveries and (standard deviation)
Sample
2 K, 250 Na
2 K, 250 Na,
250 Ca
River Sediment B
(Certified 200 K)
Axial View
No IS,
No matrix matching
4.46 (0.035)
4.72 (0.042)
230(1.0)
Matrix matching
(250 Na added)
1.98(0.031)
2.08 (0.006)
212(1.6)
RbIS
2.29(0.011)
2.33(0.003)
165(0.1)
Radial View
YIS
4.87
(0.009)
5.23
(0.081)
266(0.1)
No IS,
No matrix matching
2.08(0.012)
2.12(0.026)
184(0)
Once the method was developed and characterized reference materials were used for validation. Drinking water
and wastewater reference materials were used to test the method at low concentrations. Table 6 shows the
results, including the certified values and recoveries of the certified values. Recovery values of 80-120% of the
certified value are generally acceptable and the values are well within this range.
84
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 6. Drinking Water and Wastewater Results (standard deviation)
Element
As 188.979
Be 313.107
Cd 226.502
K 766.490R
Mn 257.610
Pb 220.353
Sb 206.833
Sr 460.733R
1643D
Certified
(mg/L)
0.05602
(0.00073)
0.01253
(0.00028)
0.00647
(0.00037)
2.356
(0.035)
0.03766
(0.00083)
0.01815
(0.00064)
0.0541
(0.0011)
0.2948
(0.0034)
1643D
Measured (mg/L)
0.0581 (0.0001)
0.0122(0.00006)
0.0060 (0.00004)
2.21 (0.033)
0.038 (0.0003)
0.0184(0.0007)
0.0521 (0.002)
0.288(0.001)
1643D
Recovery
104
97.7
92.4
93.8
101
102
96.3
97.7
HPSWW
Certified (mg/L)
0.15(0.00075)
0.15(0.00075)
0.15(0.00075)
-
0.5 (0.0025)
0.5 (0.0025)
0.15(0.00075)
-
HPSWW
Measured
(mg/L)
0.150(0.001)
0.147(0.0001)
0.147 (0.0002)
-
0.497 (0.0005)
0.504 (0.003)
0.152(0.003)
-
HPSWW
%
Recovery
99.9
97.7
97.8
-
99.4
101
101
-
High level and mixed concentrations were tested with soil and sediment digests. The results are shown in Tables
7, 8, and 9.
Table 7. Soil Results (standard deviation)
Element
Al 308.21 5R
Cu 324.754
Fe 238.204R
K 766.490R
P 178.221
Pb 220.353
V 292.402
HPS Soil A Certified
(mg/L)
500 (2.5)
0.3 (0.002)
200(1)
200(1)
10(0.05)
0.4 (0.002)
0.1 (0.0005)
HPS Soil A Measured
(mg/L)
492(11)
0.318(0.0005)
195(4)
196(4)
11.2(0.09)
0.371 (0.003)
0.10(0.0002)
HPS Soil A
% Recovery
98.5
106
97.5
97.8
112
92.8
100
Table 8. Sediment Results (standard deviation)
Element
AI308.215R
Cd 226.502
Fe 238.204R
K 766.490R
P 178.221
Pb 220.353
V 292.402
HPS Sediment B Certified
(mg/L)
600 (3)
0.03 (0.0002)
400 (2)
200(1)
10(0.05)
2(0.01)
1 (0.005)
HPS Sediment B Measured
(mg/L)
605 (8)
0.025 (0.0002)
396 (6)
203 (3)
11.1 (0.03)
1 .94 (0.0002)
1 .00 (0.0003)
HPS Sediment B
% Recovery
101
83.3
99
102
111
97.0
101
85
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 9. Estuarian Sediment Results (standard deviation)
Element
Al 308.21 5R
As 188.979
Fe 238.204R
K 766.490R
P 178.221
Pb 220.353
V 292.402
HPS Est. Sed. Certified
(mg/L)
700 (3.5)
0.1 (0.0005)
350 (2)
150(0.75)
5 (0.03)
0.3(0.0015)
1 (0.005)
HPS Est. Sed. Measured
(mg/L)
711 (0.77)
0.093 (0.0026)
355(0.31)
160(0.08)
5.81 (0.03)
0.276(0.001)
1 .02 (0.0009)
HPS Est. Sed.
% Recovery
102
93.0
101
106
116
92.0
102
The results show that the method is operating properly for a variety of matrices and concentrations. The standard
deviation for two replicates is excellent. A more thorough study of interferences is required to ensure that
adequate compensation is built into the method. For the low-level samples, no interferences were observed. For
the higher concentration samples, interferences were observed, but could be compensated with algorithms such
as interfering element corrections (IEC) or multicomponent spectral fitting (MSF).
The method was evaluated for productivity. The rinse between each sample was maintained, but shortened from
the usual 45-60 seconds to 10 seconds. This provides a wash of the probe, but more rinsing of the tubing is
accomplished with the rinse-in of the next sample. The fast pumping speed option was not used for the washing
since the 1.6-1.8 mL/min provides an adequate wash in a reasonable time. The exact time for wash-in was evalu-
ated with a study of different read delays and monitoring of the precision to see when stability was achieved. A
rinse station was added to the system to allow longer unattended runs. The internal standard was added on-line
with a mixing block, reducing the need to pipet the solution into individual samples.
The overall measurement was documented based on an average run of samples with varying elemental composi-
tions. The average time required for each sample, including wash, rinse-in, and two replicates was 3 minutes and
23 seconds.
Conclusions
In this work we have explored the possibility of a universal ICP-OES method fort he measurement of 32 elements
in a variety of matrices at low and high concentration levels. Preliminary assessment of the method indicates that
this is a viable approach, incorporating both views of the plasma for the optimal detection limit and linear range
combinations. The time of analysis for 32 elements including trace and part-per-million concentrations was less
than 3.5 minutes and provided excellent precision. Further work will include the evaluation of additional matrices.
Interference algorithms will be more completely evaluated and compared.
References
1. Method 200.7, Methods for the Determination of Metals in Environmental Samples, Supplement 1,
EPA/600/R-94/111 (1994).
2. Method 601 OB, Methods for Chemical and Physical Testing (SW-846), Revision 2, December 1996, US EPA.
NEW TECHNOLOGIES FOR METALS DIGESTIONS FOR ENVIRONMENTAL SAMPLES
L Orr
Technical Representative
Environmental Express
For the last 40 years hot plate acid digestion methodology has provided an adequate digestion of samples for low
level metals analysis. As detection limits of modern elemental analysis instrumentation have decreased however
sample contamination resulting from hot plate digestion has become a serious issue. Glass beakers used for
digestion, naturally carry many of the elements that are currently analyzed, and digesting on hotplates made of
86
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
metal parts such as Aluminum, Iron, Chromium, Lead, Copper, etc only add to the contamination problem. Hot
spots on the hotplates create problems including uneven evaporation, incomplete digestion, accidental boiling and
sample dryness. Microwave digestion addresses some of these contamination issues; however, microwave is only
EPA approved for up to 13 elements, which have restricted limits and still requires multiple transfer steps. With the
development of the Environmental Express HotBlock digestion system, clean disposable cups are used for
digestion, volume addition, filtration and sample storage without transferring the sample. Construction components
of the HotBlock are non-corrosive, and the system delivers uniform temperature heating. Experiments show that
this technology has solved most of the negative issues involving hotplate digestion while still allowing for the
digestion of all elements at all levels. Experiments include digestion recoveries, heating uniformity, contamination
and cross contamination, followed independent user findings of this technology.
MAGNESIUM CHLORIDE IN THE CYANIDE DISTILLATION
Dr. Roy-Keith Smith
Analytical Methods Manager
Jon Neuhaus
QA Specialist
Analytical Services, Inc., 110 Technology Parkway, Norcross, GA 30092
r-ksmith@asi-lab.com
Introduction
In the EPA wastewater and solid waste methods (EPA Method 335.2 and 901 OB) and Standard Methods (SM
4500-CN- C) for distillation of total cyanide, the addition of magnesium chloride solution to the distillation pot is
mandated. In attempting to chemically describe the actual function of the magnesium chloride, one is normally at a
loss to present a mechanism where it helps the isolation procedure. This paper presents the history of the magne-
sium chloride requirement and then evaluates the requirement based on the results of laboratory experiments
designed to delineate its utility in current laboratory practice.
There are a number of chemical additives that are used in the distillation pot to reduce interferences in the cyanide
isolation process, including ethylene diamine and sulfamic acid. Ethylene diamine is used to eliminate interfer-
ences from aldehydes by preventing cyanohydrin formation. Sulfamic acid is present to give an alternate substrate
for nitrites to chew upon1 Magnesium chloride is specifically required in the table of approved methods in 40 CFR
136.3 as a distillation additive2 A literature search has revealed that the specification for use of magnesium
chloride is due to an evolutionary process over several editions of Standard Methods, that culminated in the
Fifteenth Edition with the present requirement. The appearance of this reagent dates back to a paper by Serfass3
in the 1950's where it was reported that addition of magnesium chloride and mercuric chloride moderated the
evolution of hydrogen cyanide from the acidified solution allowing for better recovery.
The moderating effect was ascribed to a slower release of the cyanide through formation and then degradation of
tetracyanomercurate. Mercury forms a more stable cyanide complex than most metals, except iron and cobalt. If
hydrogen cyanide is released by a metal, the cyanide is grabbed by the mercury. The magnesium chloride was
added as a convenient source of chloride which effectively competes with cyanide for complexation to the
mercury, resulting in a slow movement of hydrogen cyanide to the absorbing flask over the course of the one hour
distillation. The combination of the two reagents was published in the 12th Edition of Standard Methods (1965).
In a paper in 1968", it was demonstrated that use of cuprous chloride, Cu2CI2, with sulfuric acid (the Williams distil-
lation procedure), was as effective as the mercury chloride/magnesium chloride mixture. The 13th Edition (1971)
of Standard Methods repeated the use of the HgCI2-MgCI2 reagent. The 14th Edition of Standard Methods (1975)
described use of cuprous chloride/magnesium chloride as a suitable catalyst. There seems to be no experimental
justification for this combination of reagents. Rather it appears as though that the replacement of mercury by
copper was a response to a growing awareness (or panic) of the hazards of mercury, and the revealing of the
horrors of the Minamata Bay massive mercury poisoning incident. The 15th Edition of Standard Methods (1980)
dropped the cuprous chloride and dictated use of only magnesium chloride catalyst5. Editions of Standard Methods
since the 15th Edition have repeated the same information about magnesium chloride.
87
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
The EPA wastewater monitoring methods were developed for the 1983 publication of Methods for the Chemical
Analysis of Water and Wastes (MCAWW), largely from the methods presented in Standard Methods The use of
magnesium chloride was not questioned, simply repeated. The EPA has for the most part simply copied the
cyanide distillation from the Fifteenth Edition and MCAWW in other methods such as 901 OB for SW-846. Of note
is that methods for cyanide isolation developed independently of the Standard Methods Committee lack magne-
sium chloride as a reagent. In the automated continuous flow distillation-colorimeteric procedure (EPA Method
335.3), a mixture of hypophosphorus and phosphoric acids are used for the acidification and no magnesium
chloride is required.
Materials and Methods
The cyanide distillations were performed in a 10-place midi-distillation unit (Kimble-Kontes, Vineland, NJj The
sulfate salts of silver, copper, nickel, manganese (II), cadmium, and mercury and the nitrates of cobalt, gold,
silver, and mercury were purchased as ACS Reagent grade solids from Fisher Scientific, Pittsburgh PA, Aldnch
Chemical Company, Milwaukee, Wl, or J.T. Baker. Mixed metals standards was purchased from High-Purity
Standards, Charleston, SC, as solutions of the nitrate salts. Magnesium chloride and sodium chloride were Fisher
Scientific ACS Reagent Grade. Reagents were prepared with doubly-deionized water in Class A volumetric glass-
ware.
The test metal solution and the cyanide spike were mixed in the reaction tube. Sulfamic acid was added to each
tube. Sulfuric acid was added through the vent tube, followed by magnesium chloride or sodium chloride solution.
The vacuum pump was turned on to generate an air flow through the system and mix the sample, then heating
was begun. Samples were distilled with a reflux rate of at least 60 drops/min for a period of one hour. Evolved
hydrogen cyanide was swept into a sodium hydroxide trap. The contents of the trap were assayed using the
pyridine-babituric acid colorimetric reaction (EPA Method 9034, EPA Method 335.2, Standard Methods 18th
Edition 4500-CN' C). Calibration checks were run daily to verify the calibration curve. Blanks and laboratory control
samples were run daily to assess laboratory contamination and method performance.
Results and Discussion
The results are presented in the Table and list the concentration of the cyanide spike, the metal ion, the presence
or absence of magnesium chloride and the percent recovery of the cyanide spike. The data are assessed as
relative percent recovery6 and presented as a bar graph in the Figure. Bars above the centerline indicate
decreased recovery on the addition of magnesium chloride. Bars below the centerline indicate increased recovery
of cyanide on the addition of magnesium chloride.
Although the Figure might suggest that there is a small but persistent positive effect due to the magnesium
chloride addition, the normal lab performance on cyanide distillation of duplicates exhibits 0-13 RPD. Any bars
falling between plus or minus 13 of the centerline should be attributed to normal variation, and thus are not
significant.
The first observations to make is that there are very few of the tested metals that exhibit any effect from magne-
sium chloride addition to the distillation. Cobalt gives poor recovery, 25 and 27%, regardless of any addition of
magnesium chloride. Other metals give good recoveries.
The second observation is that any silver present in the sample is going to adversely react with magnesium
chloride and cyanide recovery is significantly reduced. This was seen with both the sulfate and nitrate counterions.
The effect also persists if chloride is added to the sample as sodium chloride. The conclusion is that it is the
chloride ion that is the cause of the effect. Silver is well known to complex quite tightly with chloride, the [Ag(CI):]'
complex being important for solubilization of silver in acid digestions for total metals analysis7 Possibly what is
being seen is a tightly bound mixed complex of chloride and cyanide associated with the silver.
A third observation is that addition of magnesium chloride to samples containing mercury gives marginally
improved recoveries when sulfate is the counterion. Based on the paper by Serfass, one would expect dramati-
cally increased recoveries, but this is not seen. This same observation was duplicated when the mercury, as the
sulfate, concentration was increased five-fold. However significantly decreased recoveries are observed if nitrate
is present with the mercury. Addition of sodium chloride, instead of magnesium chloride, in the presence of
mercury nitrate, produces no effect, comparable to the results obtained from simple acidification and distillation
These observations are not easy to rationalize.
On the other hand, if gold and or palladium are present in the sample, the addition of magnesium chloride is
beneficial As is the case with silver, the effect is due to the presence of the added chloride with addition of
88
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
sodium chloride almost doubling the recovery of cyanide from gold solutions. The increased efficacy of sodium
chloride over magnesium chloride may be attributable to the amount of free ionic chloride in the solution. Magne-
sium chloride exhibits considerable covalent character even in solution, while sodium chloride is completely
ionized.
Table. Results of cyanide experiments
Cyanide spike mg/L Metal mg/L MgCI2 %R
1.0 - N 94
1.0 - Y 96
0.21 AgSO42.00 N 84,72
0.21 AgSO42.00 Y 61,50
0.42 AgNO32.00 N 73,81
0.42 AgNO32.00 Y 55,48
0.42 AgNO32.00 NaCI 51
0.21 Cu2.00 N 82
0.21 Cu2.00 Y 95
0.21 NI2.00 N 88
0.21 Ni2.00 Y 92
0.21 Mn2.00 N 89
0.21 Mn2.00 Y 90
2.00 Ferrocyanide N 90
2.00 Ferrocyanide Y 95
2.00 Ferricyanide N 88
2.00 Ferricyanide Y 90
0.42 Cd 2.00 N 68, 74
0.42 Cd2.00 Y 71,75,74
0.21 HgSO42.00 N 71
0.21 HgSO42.00 Y 81
0.21 HgS0410 N 83
0.21 HgSO410 Y 94
0.42 Hg(NO3)22.00 N 76
0.42 Hg(NO3)22.00 Y 50
0.42 Hg(NO3)22.00 NaCI 71
0.21 Pd2.00 N 46
0.21 Pd2.00 Y 55,58
0.42 Metals 1 N 81
0.42 Metals 1 Y 83
0.42 Metals 3 N 88
0.42 Metals 3 Y 92
0.42 Metals 4 N 83
0.42 Metals 4 Y 92
0.21 Co 2.00 N 25
0.21 Co 2.00 Y 27
0.42 Au 2.00 N 42, 38, 31
0.42 Au2.00 Y 61,59
0.42 Au2.00 NaCI 8,176
Metals 1 = Al (4 ppm), Sb (1 ppm), As (4 ppm), Ba (4 ppm), Be (0.5 ppm), Cd (0.4 ppm), Cr (0.8 ppm), Co (1
ppm), Cu (0.5 ppm), Fe (2 ppm), Pb (2 ppm), Mn (1 ppm), Ni (1 ppm), Se (4 ppm), Tl (4 ppm), V (1 ppm), Zn (1
pm), Y (8 ppm)
Metals 3 = Sn (2 ppm), Ti (0.8 ppm), Mo (0.4 ppm), Si (2 ppm)
Metals 4 = Sr (0.2 ppm), Ca (4 ppm), Mg (4 ppm), Li (0.04 ppm), K (4 ppm), Na (4 ppm), B 0.4 ppm), P (4 ppm)
89
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Conclusion
Considering the universe of samples received by a commercial laboratory for cyanide analysis, an analyst is
probably going to encounter silver in samples much more frequently that gold or palladium. Addition of chloride ion
is shown to be detrimental to cyanide recovery when silver is present in the sample. For samples with gold or
palladium, chloride addition improves recovery. Analysts should add chloride to samples containing either of these
metals.
For most analyses, however, the recommendation is to eliminate the addition of any chloride, magnesium or
sodium, to samples for distillation of cyanide.
60
Figure. Bar chart of RPD of cyanide distillation recoveries from experiments without and with added magnesium
chloride.
References
1. Smith, R.-K., 1999. Lectures on Wastewater Analysis and Interpretation, Genium Publishing, Schenectady,
NY 1-800-243-6486.
2. Guidelines establishing test procedures for the analysis of pollutants, 40 CFR 136, 1 July 1998, USEPA.
3. Serfass, E.J. et al, 1952. Analytical Method for the determination of cyanides in plating wastes and in effluents
from treatment processes. Plating 39:267.
4. Elly, C.T., 1968. Recovery of cyanides by modified Serfass distillation. WPCF Journal Vol 40 No 5 pg
848-856.
5. In the balloting of the 15th Edition, there was a least one negative vote that strenuously objected to the used of
magnesium chloride. The ballot was ignored.
6. Smith, R.-K., 1999. Handbook of Environmental Analysis, Fourth Edition, Genium Publishing Schenectady
NY 1-800-243-6486.
7. Kimbrough, D.E., and J. Wakakuwa, 1992. "A study of the linear dynamic ranges of several acid digestion
procedures", Environ. Sci. Technol. 26(1):173-178; Cohen, R.J., A.J. Meyer, E. O'Bryan, J. Kunze, and S.
Kunze, 1996. "An improved digestion method for silver analysis in solid samples", Am. Environ Lab 6/96'
28-29; EPA Methods 3010 and 3050, SW-846.
90
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
DECREASING HYDRAULIC CONDUCTIVITY BEHAVIOR AND REGULATORY COMPLIANCE OF
ALTERNATIVE HYDRAULIC BARRIERS: AN EXERCISE IN PATIENCE
Juan D. Quiroz
Graduate Research Assistant
Civil Engineering Department (JEC 4049), Rensselaer Polytechnic Institute, 110 8h St, Troy, NY 12180-3590
Ph: (518) 276-8143; Fax: (518) 276-4833; email: quiroj@tpi.edu
Thomas F. Zimmie
Professor
Civil Engineering Department (JEC 4049), Rensselaer Polytechnic Institute, 110 8th St., Troy, NY 12180-3590
Ph: (518) 276-6939; Fax: (518) 276-4833; email: zimmit@rpi.edu
Paper mill sludge used as landfill cover barrier material offers a viable alternative to compacted clay. Paper mill
sludge is the residual material from the paper making process and is characterized by high water contents, high
organic contents, high compressibilities and low shear strengths. Geotechnical research on this material has led to
the successful construction of a number of paper mill sludge landfill covers in the Northeastern United States. A
major challenge when dealing with paper sludge landfill capping projects is the process of educating regulatory
officials about paper sludge properties and behavior. The purpose of this paper is to present typical geotechnical
properties and discuss the hydraulic conductivity behavior of paper sludge landfill covers. Special emphasis will be
placed on how the hydraulic conductivity of paper sludge decreases with time after placement. The typical
hydraulic conductivity requirement of 1 x 10'7 cm/s is often accomplished by paper sludge hydraulic barriers,
however, there are instances where the hydraulic conductivity is slightly above the maximum. The amount of
barrier layer settlement typical of paper sludge landfill covers can range from 20% to 35% as compared to the 2%
to 3% for compacted clays. During this period of consolidation, large reductions in void ratio occur which affect
density, water content, shear strength and hydraulic conductivity. Hydraulic conductivity will be expected to
decrease during consolidation. In general, the hydraulic conductivity of a paper sludge landfill cover can decrease
about one order of magnitude over a period of one year. This is a reasonable time period when compared to the
design life of the landfill. There has been a considerable amount of data accumulated over the years to show
general trends of decreasing hydraulic conductivity to values lower than 1 x 10-7cm/s. Field data from three paper
sludge landfills in New York and Massachusetts will be presented. For instance, hydraulic conductivity tests from
the Corinth (NY) Landfill decreased from 1 x 10~7 cm/s to 2 x 10"8 cm/s, a decrease of about one order of
magnitude, during the post construction period. Other paper sludge landfill covers show similar trends. Due to the
changing properties of the paper sludge barrier layer, a long-term monitoring plan is essential to completely
evaluate the performance of paper sludge landfill cover. Also, the measurement of several geotechnical properties
(such as organic content, specific gravity, density, hydraulic conductivity and shear strength) often not required for
compacted clay C vers will be discussed.
Therefore, the hydraulic conductivity behavior of paper mill sludge landfill covers shows improvement over time.
Compliance of hydraulic conductivity can be achieved in about one year after placement for initially marginal
values of hydraulic conductivity. The paper sludge will behave accordingly, however Will regulatory agencies
permit the use of paper mill sludge and exercise patience allowing the paper sludge barrier layer to consolidate
and decrease its hydraulic conductivity?
PBMS: HOW WILL IMPLEMENTATION CHANGE THE ANALYSIS OF
ENVIRONMENTAL SAMPLES BY ICP-MS?
Ruth E. Wolf
Senior Scientist
The Perkin Elmer Corporation
50 Danbury Road, MS 219, Wilton, CT 06897
email: wolfre(g)perkin-elmer.com
Abstract
The results from the analysis of a variety of environmental samples by ICP-MS utilizing traditional methodologies,
such as US EPA Methods 6020 and 6020A and under Performance Based Measurement System (PBMS)
91
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
principles will be discussed. Data will be presented illustrating the performance of ICP-MS for elements recently
added to Method 6020A. In addition, the requirements regarding Interference Check Standards A and AB in
Methods 6020 and 6020A will be discussed in relation to their relevance on data quality. Data showing the
performance of Internal Standards will also be discussed in terms of what kind of drift is realistic for performing
cost effective analyses, while maintaining acceptable data quality. Recommendations will be given for laboratories
developing PBMS based methods for ICP-MS as well as for individuals auditing these laboratories on key issues
affecting data quality in ICP-MS.
Introduction
Currently, laboratories are required to use strict prescriptive methodologies for the analysis of environmental
samples. Although the methods contained in SW-846 are intended to be guidance methods and attempts have
been made by the EPA to clarify this intent, many regulators see these methods not as a guide, but as a rule.
Many samples being analyzed by environmental laboratories do not fit into the traditional water, soil, waste catego-
ries and the original SW-846 method may not give good results for a particular sample type without some modifi-
cations. Depending on the regulator, agency, or client the laboratory is reporting the data to, the laboratory's ability
to make needed method modifications may be restricted or even forbidden. The movement of the EPA to a
performance based measurement system will allow laboratories to use methodologies that will give reliable,
accurate, and meaningful results for specific sample matrices, not just follow the rules.
Review of Limitations of Method 6020
Approved Elements. The version of Method 6020 that was originally promulgated in SW-846 Update II (January
1994) only contained a partial list of the elements normally analyzed under most regulatory programs. Noticeably
missing from the method were the following elements: Se, Mo, V, Na, Ca, Mg, K, and Fe. The non-inclusion of
these elements in Method 6020 limited the usefulness of ICP-MS in certain environmental applications. Laborato-
ries wishing to use ICP-MS for these additional elements were either hesitant to do so, because the elements in
question were not originally included in the method or regulators and/or state agencies would not allow modifica-
tion of the method to include them, even with submission of relevant performance data. Application data published
(see Table 1) by Perkin Elmer includes method performance data for these and several other additional elements.
The data included in Table 1 show the detection limits and linear ranges attainable on modern ICP-MS instru-
ments, such as the ELAN 6000/6100. New advances in detector technology and operation, including the use of the
dual-range discreet dynode detectors, allow much higher concentrations, as much as 200 ppm Na, to be deter-
mined by ICP-MS. A recent revision of the method, Method 6020A, to be published in Update IVA, finally includes
the alkali metals, Se, and Hg as analytes for which the EPA has demonstrated the acceptability of Method 6020.
The inclusion of these elements in the updated Method 6020A will extend the applicability of the method until
PBMS is fully approved and implemented.
Interference Check Standards. Although the number of interferences in ICP-MS is limited, there are some well
known interferences that if not properly corrected for can lead to significant errors in the resulting data. The
composition of the Interference Check Standards A and AB in both Methods 6020 and Method 6020A are
designed to test the more common interferences encountered in environmental samples. These interferences
include the argon-chloride and argon-carbide interferences that interfere with the determination of As, Se, and Cr,
respectively. It should be recognized that the interference check standards required by Methods 6020 and 6020A
have the same intent although the exact concentration of the elements in the matrix may vary slightly between the
two different versions of the method as is illustrated by Table 2. For best results, the concentration of the matrix
elements should be indicative of those in the types of samples analyzed by the laboratory. Both the analyst and the
auditor should recognize that some of the interferences may be concentration dependent. This is particularly true
of molecular interferences. It is more important that the limitations of the interference corrections be tested and
documented by the laboratory than assuming that if the exact concentrations listed in the reference method are
used, no interference problems will exist. The laboratory should determine to what concentration a correction is
valid and establish a policy for samples exceeding that limit. Under the current two EPA methods, there are no
requirements set for the pass/fail conditions of these standards. The only requirement is that the solutions are run
at the beginning and end or every 12 hours, whichever is more frequent. Most laboratories have tried to follow the
QC Limits for the ICSA and ICSAB solutions from the ICP-OES Method 6010 for Method 6020. However the
detection capabilities of ICP-MS are so low that it is very difficult, if not impossible, to find a source for ICSA where
the measured concentrations of the analytes are below the MDLs of the analytes determined by Method 6020.
92
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 1. ELAN 6000/6100 IDLs, MDLs, and Linear Ranges for Method 6020
Analyte
Be
Al
Cr
Mn
Co
Ni
Cu
Zn
As
Ag
Cd
Sb
Ba
Tl
Pb
Hg*
Se*
Mo
Th
U
V*
Na*
Ca*
Mg*
K*
Fe*
Mass
9
27
52
55
59
60
65
66
75
107
114
123
135
205
208
201
82
98
232
238
51
23
44
24
39
54
IDL(|ig/L)
0.02
0.04
0.1
0.004
0.002
0.007
0.005
0.015
0.06
0.03
0.002
0.003
0.01
0.0003
0.001
0.020
0.09
0.004
0.002
0.0009
0.03
0.6
15
0.02
9
4
MDL (MQ/L)
0.02
0.12
0.5
0.009
0.002
0.04
0.02
0.03
0.2
0.03
0.004
0.03
0.01
0.0003
0.009
0.020
0.09
0.004
0.002
0.0009
1.2
4
20
0.04
9
4
Linear Range (mg/L)
10
10
10
10
10
10
10
5
5
5
5
10
10
10
10
20**
5
5
10
10
10
100
200**
100
100
200**
* included
** highest
in Method 6020A •
level standard ran
- SW-846 Update IVA.
for linearity test.
Table 2. Composition of Method 6020 ICSA and ICSAB solutions.
Analytes
Al Mg, P, K, S
Ca
FeNa
C (carbon)
Cl (chloride)
MoTi
As Cd, Zn
CrCo, Cu, Mn, Ni,
Ag
Hg
Se
V
Method 6020
Concentration in
ICSA
100 mg/L
100 mg/L
100 mg/L
200 mg/L
1000 mg/L
2 mg/L
Omg/L
Omg/L
Omg/L
Omg/L
Omg/L
Omg/L
Method 6020
Concentration in
ICSAB
100 mg/L
100 mg/L
100 mg/L
200 mg/L
1000 mg/L
2 mg/L
0.020 mg/L
0.020 mg/L
0.020 mg/L
0 mg/L
0 mg/L
0 mg/L
Method 6020A
Concentration in
ICSA
100 mg/L
300 mg/L
250 mg/L
200 mg/L
2000 mg/L
2 mg/L
Omg/L
Omg/L
Omg/L
Omg/L
Omg/L
Omg/L
Method 6020A
Concentration in
ICSAB
100 mg/L
300 mg/L
250 mg/L
200 mg/L
20000 mg/L
2 mg/L
0.1 00 mg/L
0.200 mg/L
0.050 mg/L
0.020 mg/L
0.1 00 mg/L
0.200 mg/L
93
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Analysts either following Methods 6020 or 6020A or developing his/her own methods under the PBMS system
should consider the use of such a solution to test the interference equations used in the method. It should also be
realized by both analysts and auditors that the prescribed ICSA and ICSAB solutions listed in Methods 6020 and
6020A do not test for a common interference found in some environmental samples from wastewater treatment
procedures and brackish or saline waters or sediments - bromide. Bromide is commonly used for disinfecting
wastewaters and drinking waters. In addition, brackish and ground waters can have significant concentrations of
bromide present. Bromine has naturally occurring isotopes at mass 79 and mass 81. The presence of part per
million levels of bromide in a sample can form the molecular species 1H-81Br* in the plasma and interfere with the
determination of Se at mass 8. Figure 1 shows a scan of 10 ppb Se superimposed over that of 10 ppm Bromide
(from an ion chromatography standard). The signal at mass 82 in the bromide solution is equivalent to 27 ppb Se
and can lead to an elevated result for selenium if this interference is not recognized and corrected for. The analyst
must also be cognizant of the fact that the correction that can be done for the formation of H-Br+ in the plasma is
not a dynamic correction and the concentration may only be valid over a limited concentration range. A dynamic
correction is one where the actual interfering molecular species can be measured at a different mass (e.g. 1H-79Br*
at mass 80) and the intensity due to the interfering species subtracted from that of the analyte at the desired
mass. In the case of bromide interference, the amount of 1H-79Br formed at mass 80 cannot be distinguished
from the high background at mass 80 due to the 40Ar-40Ar dimer. The only way to correct for this interference is to
measure the amount of formation of 1H-81Br at mass 82 using a clean (selenium free) bromide standard and
perform a correction very similar to an interelement correction in ICP-OES.
50000
40000
30000
20000
10000
IDppmBi-
IDppbSe
79 80 81
Vo."
82
83
Figure 1. Interference of 10 ppm Br- on Se.
Quality Control Limits on Internal Standards. The use of internal standards in ICP-MS is a well documented and
generally accepted practice used to compensate for signal drift caused by the gradual build-up of material on the
interface cones. Both methods 6020 and 6020A require that the internal standard intensities in all samples and
quality control standards be monitored throughout the course of the run and suitable actions carried out if either of
the established control limits are exceeded. Method 6020 requires that the intensity of the internal standards in the
subsequent continuing calibration check standards and blanks not vary more than ± 20% from the intensities origi-
nally monitored in the calibration blank, while the intensities in the actual samples are allowed to vary between
30-120%. It is common during the course of the analysis of real samples for the interface cones to become slightly
clogged while performing analyses over several hours. The degree to which this happens is entirely dependent on
the amount of dissolved material present in the samples. For digested soil samples, for example, it is not uncom-
mon to observe drift between 10-40% over the course of several hours due to deposition of calcium, aluminum,
and silicon oxides on the interface cones. To limit this deposition and the drift of the internal standards, samples
are routinely diluted to reduce the amount of dissolved solids to less than 0.1 - 0.2% TDS. However, this amount
of dilution may lead to unacceptably high detection limits for some determinations. The data shown in Table 3
shows that even with the internal standard recoveries less than the Method 6020 limit of 80%, acceptable results
were obtained for a 1 ppb instrument check standard. This data indicates that the internal standards are correctly
functioning and compensating for the signal drift.
94
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 3. Accuracy of low level calibration check standard.
Analyte
Be
Cr
Ni
Cu
As
Se
Mo
Ag
Cd
Sb
Ba
Hg
Tl
Pb
True Concentration
(re/L)
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
Measured Concentration
(|ig/L)
1.01
0.995
1.039
1.065
0.995
0.958
0.982
0.983
1.024
1.035
0.991
0.981
1.062
1.058
Internal Standard Recovery
%
84
71
71
71
67
67
71
71
71
71
71
69
69
69
Fortunately, in Method 6020A, this dual level requirement on the Internal Standard responses has been removed.
The new requirement allows the internal standard response in the samples to drop to 30% of that in the original
sample at which time, if similar recoveries are found in a calibration blank, corrective action must take place. Low
internal standard recoveries at this point in a standard with no matrix (a blank) indicates the interface cones are
becoming clogged. Either cleaning the cones or re-calibration is indicated. If, however, the calibration blank inter-
nal standards are not suppressed, the poor recoveries in the sample matrix indicates that the matrix is causing
some interference and the sample should be diluted and re-analyzed. The new single limit in Method 6020A gives
the analyst some flexibility in deciding what should be done and at what level. It should be stressed, however, that
caution should be used when reporting concentration values obtained from readings where the internal standard
response is very low (e.g. < 30-40% ), as significant error could occur. The laboratory should decide upon the
most prudent policy for the particular types of samples being analyzed and the data quality needed.
Sample Results for Method 6020
The data presented in Table 4 demonstrates the results obtained for NIST SRM 2711- Montana Soil using
Method 6020, including the additional analytes for which method performance data was generated. Perkin-Elmer
obtained SRM 2711 and processed it using U.S. EPA Method 3050 using the hydrochloric acid finish and analyzed
this digested sample in order to assess performance of Method 6020 using the ELAN 6000/6100 ICP-MS.
Because of the relatively high levels of many of the constituents in the SRM 2711 digestate and the relatively high
acid content (15% total with 5% HCI), the digestate was diluted tenfold before analysis by Method 6020. The
average values obtained by seventeen laboratories participating in a NIST round-robin study and the reported
ranges are also given in Table 4. The SRM 2711 digestate was analyzed in duplicate and the Relative Percent
Difference (RPD) between the duplicate measurements is well within the Method 6020 requirement of 20% RPD.
The largest RPD observed was 5.3%. As Table 4 shows many of the elements analyzed using Method 6020 on
the ELAN 6000/6100 are very close to the average values obtained for this SRM in the NIST study. Furthermore,
all values (except for sodium) are also within the reported range from the NIST study. The sodium level obtained
on the ELAN 6000/6100 is slightly higher than the high end of the NIST range; however, this difference is small
and is probably due to contamination considering the ubiquitous nature of sodium. An analytical spike of 100 ppb
in the diluted digestate was also analyzed and the spike recoveries calculated. The spike recoveries for all
elements, except lead, were between 96-110% recovery. Lead was recovered at 132%; however, the spike value
of the lead in solution was ten times less than the actual level of lead present in the digestate. As a result, accept-
able spike recovery of between 75-125% was not expected.
95
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 4. Resi lltS for MIST SRM ?71 1 - Moderately Contaminated Montana Soil
Analyte
Be
Al
V
Cr
Mn
Co
Ni
Cu
Zn
As
Se
Mo
Ag
Cd
Sb
Ba
Tl
Pb
Na
Mg
K
Ca
Fe
Measured
Cone
mg/kg
1.1
20066.5
48.2
23.7
493.0
8.1
17,1
104.1
315.8
94.0
2.2
1.2
4.3
40.0
3.9
192.2
1.8
1087.3
320.3
7726.8
5064.4
20742.0
21662.2
RPD
3.59
4.38
4.79
1.35
4.25
2.03
0.11
4.92
3.94
2.91
6.96
2.09
0.94
3.20
5.25
4.88
5.07
4.58
2.57
3.69
3.59
1.14
3.44
MIST Leach
Value
mg/kg
18000.0
42.0
20.0
490.0
8.2
16.0
100.0
310.0
90.0
NR
<2
4.0
40.0
<10
200.0
1100.0
260.0
8100.0
3800.0
21000.0
22000.0
Rai
low
12000.0
34.0
15.0
400.0
7.0
14.0
91.0
290.0
88.0
2.5
32.0
170.0
930.0
200.0
7200.0
2600.0
20000.0
17000.0
ige
high
23000.0
50.0
25.0
620.0
12.0
20.0
110.0
340.0
110.0
5.5
46.0
260.0
1500.0
290.0
8900.0
5300.0
25000.0
26000.0
Spike Amount
(ppb)
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
Spike
Recovery
(%)
104
—
98
97
110
98
96
99
111
103
109
105
102
102
98
98
105
132
—
—
—
—
—
Analysis of Samples by ICP-MS Using PBMS Principles - A Case Study
The following case study is presented to illustrate how sample analyses may be carried out by ICP-MS using
PBMS principles. The client is a manufacturer of calcium supplements, antacid tablets, and vitamin supplements.
Under California Proposition 65 (a.k.a. the Safe Drinking Water & Toxic Enforcement Act of 1985), lead is one of
the elements identified by the State of California as both a cancer causing agent and reproductive toxin1. Under
California Proposition 65 requirements, manufacturers of supplements and the raw materials used in their
manufacture are now required to test these materials for lead content. A "no significant risk level" or NSRL estab-
lished by the California Office of Environmental Health Hazard Assessment for lead exposure has been estab-
lished at 0.5 ug /day2. Since the actual dose may vary due to intake rate, the level of lead present in a material is
generally reported in units of mg lead per gram material (ug/g)- The client in question has several raw materials
that need to be routinely tested for lead content and would also like to obtain concentrations of 11 other elements
of interest in the raw materials.
Since the NSRL level established for lead is given as a total exposure of 0.5 ug/day, it is necessary to determine
what detection levels would be suitable to meet monitoring for this requirement. For example, the US RDA
(Recommended Daily Allowance) for calcium in the adult diet is 1000 mg or 1g. If the entire RDA were to be
obtained from a single calcium-containing supplement, the lead concentration in that supplement must be less
than 0.5ug/g of supplement material. In order to state a material has a Pb concentration less than 0.5ug/g, the
detection limit of Pb by the selected analytical technique must be significantly below 0.5ug/g to ensure reliable and
accurate results. The client has requested that the method used have a practical quantitation limit (PQL) for Pb of
0.05ug/g or lower. The client has defined the PQL as the level equal to ten times the standard deviation of the
blank. If these samples were to be run according to the strict Quality Control requirements in Method 6020 the
samples would need to be diluted after preparation by 50-fold in order to keep the internal standard responses
96
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
within the limits. This 50-fold dilution would lead to a PQL of 0.055 ug/g, which is above the stated requirements of
the client. Under the PBMS scheme, the method and quality control requirements would be developed to meet the
stated data quality objectives of the client.
Analytical Objectives: To determine Pb in the samples at a PQL of 0.05ug/g or better. The results must be
accurate to within 5% and have a minimum occurrence of false positives. The results are required to be reported
under California's Safe Drinking Water & Toxic Enforcement Act of 1985. It is also desirable to determine the
concentrations of As, Sb, Cd, Cr, Cu, Hg, Se, Tl, Sn, and Zn in all sample matrices. If possible, the client would
like all elements to be determined in a single analytical run.
Sample matrices and preparation: The clients sample include antacid tablets, calcium carbonate, tri calcium
phosphate, and magnesium oxide. NIST SRM 1400 - Bone Ash will be used to validate the method as the matrix
is similar to the samples. Since the samples were relatively simple chemical compounds, a rigorous digestion
method was not necessary. The samples were simply dissolved using nitric acid in the following manner: A 0.5 g
portion of sample was accurately weighed into precleaned 50 ml polypropylene autosampler tube. Approximately
20 ml of de-ionized water was added to each tube wash down the sides of the tube and form a slurry. Five mL of
concentrated Ultrex grade Nitric acid was added. The tubes were capped and shaken gently to mix. After dissolu-
tion, the samples were diluted to a final volume of 50mL using the graduated markings on the tubes.
Instrumental Method. The ICP-MS was set-up according to the manufacturers daily performance procedures. Due
to the low level concentrations expected for the elements of interest, the instrument was calibrated for Pb, As, Sb,
Cd, Cr, Cu, Se, Tl, Sn, and Zn at 0.1, 1.0, and 10.0 ppb. Mercury was calibrated at 0.2, 1.0, and 2.0 ppb. The
isotopes used for the determination of the elements of interest were selected based on the analysts knowledge of
the sample matrix and the possible interferences that could occur.
Determination of Practical Quantitation Limits (PQLs). In order to determine the practical quantitation limit, the
standard deviation of 7 readings from the continuing check blank that was run every 10 samples throughout the
course of the run was multiplied by 10 to determine the practical quantitation limit. The PQL in the solid was then
determined by converting back to the units of ug/g using the sample preparation weight (0.5g), sample preparation
volume (50mL), and dilution factor (10). The results are given below:
Element
Pb
As
Sb
Cd
Cr
Cu
Se
Tl
Sn
Zn
Hg
Standard Deviation
of Blank
0.0011
0.0119
0.0222
0.0007
0.0777
0.0082
0.0313
0.0011
0.0600
0.0102
0.0283
PQL = 10*STDDEV
(M9/L)
0.011
0.119
0.222
0.007
0.777
0.082
0.313
0.011
0.600
0.102
0.283
PQL in Solid
(ug/g)
0.011
0.119
0.222
0.007
0.777
0.082
0.313
0.011
0.600
0.102
0.283
Method Validation. NIST SRM 1400 - Bone Ash was selected as a reference material which could be used to
validate this method because the calcium and phosphate matrix in this SRM is similar to the calcium matrices
submitted for analysis by the client. Of particular importance is the accuracy of the Pb determination in SRM 1400,
as the client wants an accurate determination for Pb. The other elements determined in SRM 1400 will be
compared to both certified and reference values where applicable to evaluate the accuracy of the method for the
other analytes. Matrix spikes will also be used to evaluate the effect of interferences and matrix effects on the
results. The element of priority, Pb, is shown to be accurately determined (within 3%) as compared to the certified
NIST value for this element using the simple dissolution and analytical method described above. The results
obtained for arsenic, cadmium copper, selenium, and zinc also agree with the values reported by NIST to within
97
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
25% in the worst case. The pre-digestion spikes and post digest spikes also show recoveries within 10% of the
spike value in all cases where a spike was performed, indicating no severe matrix effects are present.
Results for NIST SRM 1400 - Bone Ash
Analyte
Pb
As
Sb
Cd
Cr
Cu
Se
Tl
Sn
Zn
Hg
Measured
Concentration
(ug/g)
8.89
0.50
0.65
0.03
1.48
2.3
<0.3
<0.01
<0.6
181
<0.3
NIST Value
* Certified Value
(M9/9)
9.07 (*)
0.4
0.03 (*)
2
0.08
181 (*)
1 ppb Post Digest
Spike Recovery
(%)
109
105
94
100
107
93
96
107
100
10 ug/g Pre-Digestion
Spike Recovery
(%)
106
103
99
100
107
97
94
107
Sample Results. The samples are then analyzed according to the developed method. Matrix spikes are performed
in order to assess data quality. The results for the calcium phosphate matrix are given below:
Calcium Phosphate
Analyte
Pb
As
Sb
Cd
Cr
Cu
Se
Tl
Sn
Zn
Hg
Measured
Concentration (ug/g)
0.097
2.84
0.4
0.134
33.8
0.725
0.5
0.01
<0.6
16.83
<0.3
1 ppb Post Digest
Spike Recovery (%)
107
112
100
108
123
95
109
110
99
102
10 ug/g Pre-Digestion Spike
Recovery (%)
106
107
105
106
96
93
114
110
Summary
The differences between US EPA Method 6020 and 6020A have been discussed. Data illustrating why changes
were made in Method 6020A to make it more flexible were presented. The limitations of the Interference Check
Standards as presented in Methods 6020 and 6020A were described and recommendations made regarding the
limitations and possible modification of the content of these solutions. The example of the bromide interference on
selenium was presented as a situation where the interference check solution in Methods 6020 and 6020A are not
adequate. The use of internal standards in ICP-MS was discussed in regards to the accuracy of the analysis when
the internal standard recoveries are low. Data for a certified reference material was presented showing the appli-
cability ICP-MS for the determination of many elements, including those that are traditionally run by ICP-OES.
Finally, an example of method development was briefly discussed under PBMS principles and data presented that
validated the method developed to satisfy the clients stated data quality needs.
References
1. List of Chemicals Known to the State to Cause Cancer or Reproductive Toxicity, California Code of Pecula-
tions, Title 22, Section 12000, August 26, 1997
98
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
2. "Method of Detection Argument in the Crystal Glassware Case", Prop 65 News, July 1994, Vol. 8, No. 7.
DETERMINATION OF MERCURY IN THE RANGE OF 1 - 100 ng/L USING CV-AAS
Manfred Leyrer, Gerhard Schlemmer and Zoe Grosser
The Perkin-Elmer Corporation, 50 Danbury Road MS-219, Wilton, CT 06897
email: grosseza@perkin-elmer.com
Abstract
Mercury continues to be an environmentally relevant element. It must be determined in the low nanogram per liter
range to measure background levels in ambient waters or groundwaters. The deposition of mercury is a global
issue and trace contamination must be distinguished from background levels. Bioaccumulation in the food chain
from even trace contamination can provide significant health hazards. The determination of extremely low mercury
levels (below a few ppt) can be achieved by collecting the analyte on an adsorption agent. Usually gold/platinium
or iridium is used to trap the mercury. The detection limits are mainly restricted by the level of the blank rather than
by the photometric noise of the instrument at these low levels. Extreme care has to be taken not to contaminate
the samples during sample handling, stabilization and measurement.
The automated measurements shown in this paper where performed under standard laboratory conditions. The
detection limits can be further improved if the samples and standards are handled using more rigorous clean
sampling and handling techniques. This paper shows how mercury in water can be analyzed in a range between 1
and 100 ng/L. An automated cold vapor technique atomic absorption technique and amalgamation on a
gold/platinium gauze have been used to obtain these data. Ambient water and soil samples have been measured
using this technique.
Introduction
Mercury pollution has decreased in the United States as sources of mercury have been controlled. Mercury is a
global pollutant and can be spread through the air to even the most remote areas. This confounds the
determination of the source of pollution and can bias the evaluation of local control effects. The measurement of
mercury continues to be of interest; however, at lower levels. As the interest in speciated forms of mercury and the
analysis of potential endocrine disrupting effects increases, measurement at lower concentrations will continue to
grow in importance. Table 1 shows the current regulatory levels for mercury in a variety of matrices, in the U.S.
and in Europe. The values are all listed in parts-per-trillion, unless otherwise noted, to allow for easy comparison.
The solution concentrations from solid values were obtained by assuming a typical digestion using 1 gram of solid
material and dilution to 100mL of final solution.
Ambient water is the single category currently requiring measurements at ultratrace levels.
Table 2 summarizes the methods for determination of mercury with AA and cold vapor generation. Flow injection
can be used to automatically prepare small samples or can be used in the continuous flow mode for larger
samples. Preconcentration of the vapor can be accomplished with amalgamation techniques or by collection in
graphite tube. Detection can be performed with an atomic absorption spectrometer or a dedicated system.
The detection limits achieved with most of the listed techniques is more than sufficient to give a confident result at
the decision-making concentration listed in Table 1. For ambient water concentrations, preconcentration using
amalgamation or collection on a graphite tube coupled with a sensitive detector are necessary to achieve the
desired results.
The scope of this work is to explore the factors involved in implementing the determination of mercury using flow
injection-continuous flow with amalgamation and a dedicated mercury detection system.
99
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 1. Summary of Mercury Regulatory Levels
Medium
Drinking Water
Wastewater (Chlor-Alkali-
Mercury Cells) (new)
Universal Treatment Stds
TCLP Extracts
Soils
Sludges
Ambient water
US Maximum Contaminant Level (ng/L)
2000
1 10,000 max for one day
48,000 avg. over 30 days
150,000 (wastewater) or
25,000-200,000 (nonwastewater)
200,000
1-21 mg/kg cleanup goal
(10,000-210,000 ng/L in solution, based
on 1g sample)
12 (freshwater cont. criteria)
1 .8 (Quality Guidance for the Great
Lakes)
EU Regulatory Limit (ng/L)
1000
50,000 ng/L before it is mixed with
other wastewater
0.5- 10 mg/kg
( 1 mg/kg for Agricultural soil )
(5,000 -10,000 ng/L in solution,
based on 1g sample)
16-25 mg/kg
(160,000-250,000 ng/L in solution,
based on a 1g sample)
Natural waters such as Lake
Constance, Germany carry around
0.8 ng/L Hg
Table 2. Methods for the Determination of Mercury with CVAAS
Technique
Flow Injection
Flow Injection-Continuous
Flow Injection
Flow Injection-Continuous
Flow Injection
Preconcentration
None
Au/Pt Gauze
None
Au/Pt Gauze
Graphite Tube
Detector
AAS
AAS
FIMS
FIMS
AAS
Detection Limit (ng/L)
100
10
4
0.5
0.5
Experimental
All work was performed using the Perkin-Elmer FIMS™ 400, with an automated amalgamation accessory. Figure
1 shows a schematic of the system.
When the amalgamation accessory is
used, the Fl valve provides a repro-
ducible and defined preconcentration
time, preventing sample to sample
carry-over in the continuous flow mode.
Ultrapure chemicals were used to mini-
mize contamination. Sample prepara-
tion and analysis was done in a clean
hood. The concentrations used are
documented in the Perkin-Elmer imple-
mentation of EPA method 245.1, ap-
proved through the alternate testing
procedure.1 The conditions for bromate
digestion and cleaning of reagents were
taken from draft EPA method 1631.2
AAS, OES, MS
specific detector
Figure 1.
System
Schematic of Flow Injection
quartz cell
long path glass cell
Au/ Pt gauze
graphite tube
plasma
volume selection
control of liquids
control of gases
Results and Discussion
Analysis at ultratrace levels requires careful sample collection and handling. The evaluation of the blank values
from different conditions using amalgamation are summarized in Table 3. A 60-second amalgamation using 10 mL
of sample is used in each case.
100
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 3. Blank Values
SnCI2, not purified
SnCI2, Purified 1 hour with Argon
ASTM Type 1 water with 0.5% HNO3
ASTM Type 1 water with 0.5% HNO3 and KMnO4
Peak Height
0.0075
0.0004
0.0005
0.0008
Concentration (ng/L)
15.6
0.8
1.0
1.6
%RSD
3.4
9.8
3.7
8.2
IMO -
11
Amalgamation typically improves
detection limits by a factor of ten.
Bromate reagent acts more quickly
than KMnO4 and may be cleaner.
This also can contribute to lower
detection limits. The time of amalga-
mation can be varied and increased
times yield increased preconcentra-
tion and lower detection limits. Figure
2 demonstrates peaks obtained from
standards preconcentrated for 180
seconds, using 30 ml of solution.
Figure 2. Peak profiles of mercury in
standards and samples
Conclusions
Most current regulatory levels are met satisfactorily with existing methodology. Bromate digestion can increase
laboratory productivity and provide less contamination for ultratrace samples. The time-savings aspect may be
useful for analyses at all concentration levels and should be further evaluated for incorporation into existing
methods. Amalgamation can improve detection limits to measure mercury at ambient water levels. As the move
towards a performance-based measurement system continues, the ability to match the available tools more
closely to the problem to be solved will be achieved. Techniques for ultratrace analysis require extra care at every
step of the collection, sample handling and analysis processes. Although an automated system, such as flow
injection, can help tremendously in isolating the sample from sources of contamination, additional skill will be
required compared to analyses at higher concentrations.
References
1. S. Mclnotosh and B. Welz, The Application of Flow Injection Analysis to Automating Cold Vapor Mercury
Analyses, ENVA-100, The Perkin-Elmer Corporation, 761 Main Ave, MS-10, Norwalk CT 06859.
2. Method 1631: Mercury in Water by Oxidation, Purge and Trap, and Cold Vapor Atomic Fluorescence
Spectrometry, EPA821-R-95-027, April 1995.
APPLICATION OF IN-SITU GAMMA SPECTROMETRY IN THE REMEDIATION OF
RADIOACTIVELY CONTAMINATED SOIL
Chris Sutton
Senior Technical Expert
John D. Yesso
Senior Technical Expert
Raymond J. Danahy
Environmental Scientist
Thomas Cox
Environmental Scientist
Soil and Water Division, Fluor Daniel Fernald, P.O. Box 538704, Cincinnati, OH 45253-8704
chris_sutton@fernald.gov
101
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
ABSTRACT
The Fernald Environmental Management Project (FEMP) is a U.S. Department of Energy site that is undergoing
total remediation and closure. Most of the remediation effort entails massive excavation of soil for disposal, both
offsite and onsite, at an engineered disposal facility. In-situ gamma spectrometry is routinely used to support soil
excavation operations to accurately and quickly identify soil areas as being above or below regulatory remediation
criteria.
Two different in-situ gamma spectrometry systems are used. The first is a sodium iodide (Nal) detector mounted
either on a tractor or a jogging stroller, depending on the terrain to be measured. The Nal system allows the
collection of a gamma energy spectrum which can be analyzed to identify and quantify radioactive isotopes which
are present within the detector's viewing area. Each energy spectrum is tagged by location coordinates provided
by an on-board global positioning system (GPS) to precisely locate elevated contamination areas. The second is a
tripod-mounted, high purity germanium detector (HPGe) gamma spectrometry system that is functionally similar to
the Nal system. The principal advantage of the HPGe is its superior resolution, which allows much more accurate
identification and quantification of radionuclide contaminants in soils.
In order to effectively utilize the data quality objective process with these systems, three quality assurance (QA)
elements had to be performed. First, method validation studies demonstrated comparability with conventional
radiochemistry methods and established performance-based acceptance criteria for key quality control parameters
at various data quality levels. The method validation studies for the HPGe system stressed accuracy and compa-
rability, while method validation studies for the Nal systems stressed quantifying measurement uncertainty and
detection limits. Second, a "User's Manual" was developed that specifies measurement approaches, provides data
interpretation guidelines, and discusses operational and environmental factors that could adversely affect in-situ
gamma spectrometry measurements. This manual is primarily designed for environmental scientists responsible
for remediating soils rather than for analytical chemists who perform the measurements. Third, an in-situ gamma
spectrometry QA program was implemented to address programmatic QA elements, to ensure legal defensibility
of the data, and to specify quality control (QC) criteria, their frequency of measurement, their acceptance limits
and whether or not they are to be control charted.
INTRODUCTION
The FEMP is a U.S. Department of Energy site that is undergoing total remediation and closure. Most of the
remediation effort entails massive excavation of soil for disposal, both offsite and onsite at an engineered disposal
facility. In-situ gamma spectrometry is routinely used in support of soil excavation operations to accurately and
quickly identify soil areas as being above or below regulatory remediation criteria. Two different in-situ gamma
spectrometry systems are used. The first is a sodium iodide (Nal) detector system, while the second is a high-
purity germanium (HPGe) detector system. The former system is mounted on either a tractor (RTRAK) or a
jogging stroller (RSS), depending on the terrain, while the latter system is tripod-mounted.
Both RSS and RTRAK have a measurement system consisting of a 4x4x16 inch Nal detector and associated
electronics to provide high-speed pulse height analysis. This system allows the collection of a gamma ray energy
spectrum, which can be analyzed to identify and quantify radioactive isotopes that may be present within the
detector's viewing area. The RTRAK and RSS are each equipped with a GPS operated in a real-time differential
mode to provide location coordinates. Each energy spectrum is tagged with the location coordinates provided by
the GPS. All energy and location data are stored on magnetic media by an on-board computer system. This infor-
mation is used to accurately locate and subsequently map radiological data within the measurement area.
On the RTRAK, the detector is positioned on the tractor horizontal to the ground and perpendicular to the direction
of travel at a height of approximately 31 cm above the ground. The detector on the RSS is mounted horizontal to
the ground and parallel to the direction of travel at a height of approximately 31 cm. The normal operation of the
RTRAK and RSS consists of moving the systems over the measurement area at a predetermined speed. Spectra
are continuously collected at regular intervals, typically a few seconds. The viewing area size is a function of the
tractor speed, the acquisition time, and the detector's geometrical configuration. For example, for the 4x4x16 inch
detector at the 31 cm height, the viewing area is 8.8. m2 for a single measurement when the system is moving at
one mile per hour, with a 4-second data acquisition time (typical operating parameters).
The HPGe detectors are mounted on tripods at heights ranging from 15 cm to 1.0 m above the ground surface.
The detectors are connected to 8192 channel multi-channel analyzers which allow the collection of a high resolu-
tion gamma ray spectrum. The superior resolution of HPGe detectors relative to Nal detectors allow it to
accurately quantify a wide variety of isotopes with minimal interferences. Data acquisition times typically are 15
102
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
minutes. The HPGe field of view ranges from over 100 m2 at a 1.0 m detector height to 3.1 m2 at a 15 cm detector
height.
METHOD VALIDATION STUDIES
The method validation study for HPGe entails determining the similarity between data generated by HPGe
measurements and data generated by laboratory analysis of physical samples. It also delineates acceptance crite-
ria for key QC elements and data quality elements. Three radiological contaminants of concern were measured by
HPGe and laboratory methods: total uranium, thorium-232 and radium-226. Method validation studies for Nal
systems stressed quantifying measurement uncertainty and detection limits. Such assessments were performed
as a function of vehicle speed and data acquisition time in order to determine preferred operating parameters.
HPGe Comparability Studies
One part of the method validation study for HPGe entailed assessing the comparability between HPGe measure-
ments and laboratory data. To accomplish this, a series of physical samples were collected from different areas of
widely varying concentrations of contaminants. In each area, samples were collected in a "bullseye" pattern to
mimic the averaging done by the field HPGe detector. That is, the area from which physical samples were taken
can be envisioned as a circle, with the HPGe detector located above the center. The HPGe detector records
gamma ray photons from every point within the circle; however, it records more gamma rays from soil closer to the
detector than from soil further from the detector.
For comparison with HPGe measurements, a weighted average (weighted based upon gamma photon fluence
contributions) of all laboratory data for a given area was calculated. Figures 1 and 2 show plots of HPGe measure-
ments vs weighted average laboratory data for total uranium and thorium-232. High correlation coefficients (R2
value), line slopes near one, and line intercepts close to 0.0 demonstrate comparability of data. The width of the
error bars for laboratory data in Figures 1 and 2 primarily reflect the degree of heterogeneity among samples in a
given area rather than laboratory precision.
Nal Method Validation
A major portion of the method validation studies for Nal systems addressed the total system measurement uncer-
tainty for moving systems. Data were acquired experimentally via repeated measurement profiles, which involved
moving the RTRAK or RSS back and forth along a given track for 20 iterations. Each track was divided into
segments and the mean and standard deviation of the measurements in each segment was determined. Table 1
shows the results of the precision studies for one area with the RTRAK moving at a speed for 0.5 mph, with a
2-second data acquisition time. Such precision studies were carried out in different areas, using a combination of
different speeds and data acquisition times in each area. The results of these studies demonstrated that:
1. The uranium-238 measurements display low degrees of precision. This limits the usability of the data for
low-concentration measurements. The low degree of precision (high uncertainty) occurs because of the low
photon yield at the energy of interest, the high spectrum background, and interferences from thorium-232 and
radium-226 daughter gamma rays.
2. The thorium-232 measurements display the highest degree of precision of the three radionuclides of interest.
The high degree of precision (small uncertainty) occurs because of a relatively high photon yield at the energy
of interest, the low spectrum background, and because of only limited interference from a low intensity radium-
226 peak.
3. The radium-226 measurements display a degree of precision similar to that of uranium or between that of the
other two radionuclides of interest. This is in part because both the photon yield and the detection efficiency at
the energy of interest fall between those of the thorium and uranium.
Knowledge of the overall precision from studies such as the one outlined above was a key factor in ascertaining a
priori minimum detectable concentrations, determining error rates, and setting trigger levels.
USER'S MANUAL
Early in the remediation process at the FEMP, it became clear that a critical need existed to bridge the gap
between primarily analytical information contained in method validation studies and programmatic remediation
design documents. The User's Manual bridges that gap by providing user guidelines, data interpretation guide-
lines, and measurement strategies and approaches; by discussing operational and technical factors that could
adversely affect data; and by delineating strengths and limitations of in-situ gamma spectrometry. While the
103
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
document is beneficial to anyone involved with any aspect of in-situ gamma spectrometry, it is primarily aimed
toward FEMP project personnel who:
• plan soil remediation projects;
• collect in-situ gamma spectrometry data for soil remediation projects;
• interpret in-situ gamma spectrometry data for soil remediation projects;
• integrate in-situ gamma spectrometry data with other data sets or into engineering designs; and
• make decisions based upon in-situ gamma spectrometry data.
The User's Manual has four sections: 1) Investigation Approaches; 2) Measurement Approaches; 3) Data Interpre-
tation Guidelines; and 4) Technical Issues. Section 1 deals with broader-scale issues such as how in-situ gamma
spectrometry is used in pre-design investigations and in soil excavation operations. Section 2 deals with smaller-
scale issues such as how in-situ gamma spectrometry is used to detect, confirm, and identify hot spots. Section 3
addresses such issues as climatic/weather effects upon in-situ gamma measurements, topographic effects, total
activity data interpretation, and mapping conventions. Section 4 addresses technical issues such as data review
checklists, minimum detectable concentrations, positioning and surveying, and the effects of radon-222 on
radium-226 measurements.
QUALITY CONTROL/QUALITY ASSURANCE
All in-situ gamma spectrometry operations, whether method validation studies or field measurements in support of
remediation operations, are governed by a comprehensive QA/QC program. The QA program contains all of the
same quality elements as a traditional environmental laboratory QA program. It has ten criteria: 1) QA program; 2)
personnel training/qualification; 3) quality improvement; 4) documents and records; 5) work processes; 6) method
design, 7) procurement/control of materials and services; 8) facilities and equipment/calibration and maintenance;
9) management assessment; and 10) external assessments and audits.
Of particular interest is the QC program, which is centered around performance-based measurements. In this
regard, acceptance criteria of key quality control elements are specified, while the mechanism of how such
measurements are obtained are not specified in either the QA plan or QC plans. Table 2 contains such criteria for
two data quality levels called Analytical Support Levels (ASLs) at the FEMP. ASL B corresponds generally to the
US EPA "screening data" category, while ASL D corresponds to the US EPA's "definitive data" category.
Information from the method validation studies, the User's Manual, and the QA/QC plans are incorporated into
Project Specific Plans (PSPs) and project Data Quality Objectives (DQOs) to support specific remediation activi-
ties. In-situ gamma spectrometry data are validated to ensure that they satisfy the requirements and needs speci-
fied by the PSPs and DQOs.
Fable 1. RTRAK precision si
Segment
1
2
3
4
5
6
7
ROAD
8
9
10
11
Averages
Minimum
Maximum
No
Measurements
129
217
206
205
216
225
200
120
231
232
240
193
udies at 0.5 MPH with a 2.0 second data acquisition time
Uranium-238 (pCi/g)
Mean
12.4
14.1
15.6
15.2
16.8
14.5
16.5
12.2
17.0
18.0
17.2
15.2
15.7
12.4
18.0
Std Dev
9.3
9.1
9.0
8.3
8.7
9.4
9.6
7.3
9.2
9.3
9.8
8.6
9.1
8.3
9.8
%Std
Dev
75
65
58
55
52
65
58
60
54
51
57
56
59
51
75
Thorlum-232 (pCi/g)
Mean
0.75
0.77
0.75
0.80
0.73
0.76
0.78
0.48
0.75
0.75
0.73
0.75
0.76
0.48
0.80
Std Dev
0.26
0.32
0.27
0.31
0.29
0.29
0.31
0.29
0.34
0.32
0.31
0.28
0.30
0.26
0.34
%Std
Dev
35
42
36
39
40
38
40
60
45
43
42
37
40
35
60
Radium-226 (pCi/g)
Mean
0.72
0.79
0.82
0.76
0.82
0.76
0.80
0.59
0.82
0.87
0.77
0.76
0.79
0.72
0.87
Std Dev
0.50
0.51
0.47
0.53
0.54
0.52
0.54
0.45
0.59
0.51
0.48
0.50
0.52
0.45
0.59
%Std
Dev
70
64
57
70
66
68
68
76
72
58
63
65
66
57
76
104
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
SUMMARY
Routine utilization of in-situ gamma spectrometry in remediation at Fernald rests upon three programmatic
elements. Method validation studies carried out to delineate key measurement quality control elements such as
comparability, representativeness, accuracy, uncertainty, and detection limits; a User's Manual which specifies to
environmental engineers and scientists how in-situ gamma spectrometry should be used in remediation opera-
tions; and a comprehensive QA program to ensure that in-situ gamma spectrometry data are of sufficient quality
for their intended usage and are legally defensible.
Table 2. Tabulation of quality control criteria and requirements
RTRAK and RSS Nal Detector QC Criteria and Requirements
QC Element
Energy
Calibration
Detector Counting
Efficiency Check
Nuclide
TI-208
Pb-212
TI-208
Gamma
Energy
2614.5 keV
238.6 keV
2614.5 keV
QC Criteria
Channel 447±2
Channel 40±2
Predetermined check
source value (decay
corrected) T ± 3 sigma
Frequency
Days used, prior to
and following use
Days used, prior to
and following use
Control
Chart
No
Yes
HPGe Detector QC Criteria and Requirements
QC Element j Nuclide
Energy
Calibration
Detector
Resolution
Detector Counting
Efficiency Check
Am-241
Cs-137
Co-60
Co-60
Co-60
Gamma 1 QC Criteria
Energy |
59.5 keV
661.6keV
1 332.5 keV
1332.5
1332.5
Channel 158±1
Channel 1763±2
Channel 3553±3
Measured mean value
T± 3 sigma
pre-determined check
source value (decay
corrected) T ± 3 sigma
Frequency
Days used, prior to
and following use
Days used, prior to
and following use
Days used, prior to
and following use
Control
Chart
No
Yes
Yes
HPGe Field Measurements QC Criteria and Requirements
QC Element
Field
Measurement
Interference
Field Control
Station
Field Control
Station
Minimum
Detectable
Concentration
Gamma Energy
Nuclide or Basis
1460.8 keV
Total U
Th-232
Ra-226
K-40
Temperature
Humidity
Soil Moisture
Free Release
Levels for
Nuclides
of Concern
QC Acceptance Frequency
Criteria
keV = 1460.8
FWHM < 3.0 keV
or
Channel = 3895.0
FWHM < 8 Channels
ASL-D
measured value ±3 sigma
measured value ±3 sigma
measured value±3 sigma
measured value±3 sigma
No Criteria
For ASL-D
95% UCL1
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
HPGe Field Measurements QC Criteria and Requirements (continued}
QC Element
Measurement
Accuracy
Measurement
Bias
Precision of
Duplicates
Detector
Counting
Efficiency
Determination
Gamma Energy
Nuclide or Basis
Compared to
weighted average
of physical samples
Compared to
weighted average
of physical samples
At least one per every
20 HPGe
measurements.
Determination of
conversion (efficiency)
factors.
QC Acceptance
Criteria
ASL-D - weighted average
of physical sample ±20%
ASL-B - weighted average
of physical sample ±35%
Bias acceptable unless it
produces errors resulting
in accuracy being
exceeded
measured value >(5 x
MDC) then RPD <±20%
measured value <(5 x
MDC) then measurement
difference < ±MDC
initial conversion factor
±10% for each gamma
energy2
Frequency
Annually
Annually
At least one per
every
20 HPGe
measurements.
Annually
Control
Chart
No
No
No
No
Note 1. Upper confidence level (UCL) for MDC.
Note 2. Nuclide and Gamma energies measured:
Cs-137 32.2 Eu-152 39.5 Am-241 59.5
Eu-152 121.8 Eu-152 244.7 Eu-152 344.3
Eu-152 411.1 Eu-152 444.0 Cs-137 661.6
Eu-152 778.9 Eu-152 964.0 Co-60 1173.7
Co-60 1332.5 Eu-152 1408.0
1200-1
1000
800
E
o.
o
0.
X
400
200
= 110719X+0.6484
2 = 0.9929
z:
s^
80
40
20
iZ
20 40 60 80
200
400 600 800 1000
Laboratory Total Uranium Data (ppm)
1200
1400
Figure 1. Correlation Between HPGe and Laboratory Data for Total Uranium at a 31 cm Detector Height
106
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
3.0 4.0 5.0 6.0
Laboratory Thorium-232 Data (pCl/g)
7.0
8.0
9.0
Figure 2. Correlation Between HPGe and Laboratory Data for Th-232 at a 31 cm Detector Height
EFFECT OF ENVIRONMENTAL VARIABLES UPON IN-SITU GAMMA SPECTROMETRY DATA
Chris Sutton
Senior Technical Expert
Soil and Water Division, Fluor Daniel Fernald, Mail Stop 35, P.O. Box 538704, Cincinnati, OH 45253-8704
chris_sutton@fernald.gov
ABSTRACT
The Fernald Environmental Management Project (FEMP) is a U.S. Department of Energy site that is undergoing
total remediation and closure. Fernald is a former uranium refinery which produced high quality uranium metal.
Soil in the Fernald site is pervasively contaminated with uranium and secondarily with thorium and radium
isotopes. In-situ gamma spectrometry is routinely utilized in soil excavation operations at Fernald to provide high
quality and timely analytical data on radionuclide contaminants in soil.
To understand the effect of environmental conditions upon in-situ gamma spectrometry measurements, twice daily
measurements were made, weather permitting, with a tripod-mounted high purity germanium detector (HPGe) at a
single field location (field quality control station) at the Fernald Environmental Management Project. Such
measurements are the field analogue of a laboratory control standard. The basic concept is that measurement
variations over an extended period of time at a single location can be related to environmental parameters.
Trends, peaks, and troughs in data might be correlative to both long-term and short-term environmental condi-
tions. In this paper environmental variables/conditions refer to weather related phenomena such as soil moisture,
rainfall, atmospheric humidity, and atmospheric temperature.
Based upon data collected over a year, the effect of soil moisture, humidity, temperature, various weather condi-
tions such as fog, time of day, and season upon HPGe measurements can be delineated. This has resulted in a
set of operating guidelines for field personnel and data interpretation guidelines for environmental scientists using
HPGe data. Further, the data set allows the long-term measurement uncertainty (precision) for each individual
107
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
analyte to be ascertained. For example, the mean of 250 total uranium measurements (dry weight basis) taken
throughout the year is 93.4 ppm with a standard deviation of 5.6 ppm. The standard deviation is 6.0% of the mean.
Based upon such means and standard deviations for each analyte of interest, control charts have been estab-
lished in which the warning and control limits are derived from the standard deviations.
Of particular interest is the behavior of radium-226. Because the HPGe actually measures gamma photons
emitted by radon-222 daughters to calculate radium-226, weather conditions leading to the buildup and dissipation
of radon-222 (a gas) in surface soils greatly affect the concentration of radium-226 determined from HPGe
measurements. Typically, morning radium-226 concentrations as determined from HPGe measurements average
over 25% higher than afternoon concentrations with a high degree of variability associated with that average.
INTRODUCTION
The Fernald Environmental Management Project (FEMP) is a U.S. Department of Energy site that is undergoing
total remediation and closure. Fernald is a former uranium refinery which produced high quality uranium metal.
Soil in the Fernald site is pervasively contaminated with uranium and secondarily with thorium and radium
isotopes. In-situ gamma spectrometry is routinely utilized in soil excavation operations at Fernald to provide high
quality and timely analytical data on radionuclide contaminants in soil.
To understand the effect of environmental conditions upon in-situ gamma spectrometry measurements, twice daily
measurements were made, weather permitting, with a tripod-mounted high purity germanium detector (HPGe) at a
single field location (field quality control station, or FCS).
To delineate the effect of weather and climatic conditions upon HPGe measurements, the field analogue of a
laboratory control standard was adopted. The basic concept is that measurements over an extended period of
time at a single field location can be related to weather and climatic variables. Trends, peaks, and valleys in data
may be related to both long term and short term weather and climatic conditions. In this report, such conditions
refer to weather related phenomena such as soil moisture, rainfall, atmospheric temperature, and humidity. FCS
measurements thus offer the possibility of normalizing all in-situ gamma spectrometry measurements to a
standard set of conditions, thereby enabling in-situ gamma spectrometry project personnel to tell when HPGe
measurements are "in control."
This paper presents results of twelve months (April 8, 1997 through March 31, 1998) of morning and afternoon
HPGe measurements at a FCS. A field location with a total uranium content of approximately 90 to 100 ppm (dry
weight basis) was chosen as the FCS. This location was selected over other possible locations because of the
closeness of its total uranium concentration to the FEMP final remediation level (FRL) of 82 ppm for total uranium.
Measurements were performed at a 1.0 meter detector height using a 15-minute data acquisition time. Data were
collected for total uranium, thorium-232, radium-226, and potassium-40. In this paper, only total uranium and
radium-226 data are discussed for the sake of brevity.
EFFECT OF SOIL MOISTURE ON HPGe MEASUREMENTS
When total uranium is plotted as a function of soil moisture on a wet weight basis, there is a distinct trend of
decreasing concentration with increasing soil moisture. This is not surprising as water acts as a diluent. However,
when wet weight concentrations are converted to dry weight concentrations (Figure 1), there is still a slight trend of
decreasing dry weight concentrations with increasing soil moisture content. Although the dry weight concentration
dependency upon soil moisture is evidenced by a very low correlation coefficient (shown as an R2 value) of 0.22 in
Figure 1, the upper and lower 95% confidence limits for the slope do not bound zero. Hence, the slope of the line
in Figure 1 is significantly different than zero. The slight trend of increasing dry weight concentration with decreas-
ing soil moisture content may reflect the fact that a soil moisture depth gradient usually exists. In drying periods,
the surface soil is usually drier than soil a few inches deeper. After periods of rain, surface soil is usually wetter
than soil a few inches deeper. Because a soil moisture measurement represents an average, the surface soil is
usually a little drier or wetter than the average. Since a majority of the gamma photons are emitted from surface
soils, it is not surprising that concentrations derived from abundances of these photons still show a residual
dependency upon moisture even following correction from wet weight to dry weight.
EFFECT OF ATMOSPHERIC TEMPERATURE ON HPGe MEASUREMENTS
Figure 2 is a plot of total uranium concentration as a function of temperature. A regression line indicates a slight
trend of increasing measured HPGe concentrations with increasing temperature. Although the trend in Figure 2 is
slight, it is real; the slope of the line of dry weight concentrations vs. temperature is significantly different than
zero.
108
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
The origin of the trend (albeit slight) of increasing measured concentration with increasing temperature is not
clear. Discussions with gamma spectroscopists suggest that it is not instrumental in origin. Speculation is that the
trend results from soil moisture gradients. At higher temperatures, more of a gradient between surface soils (drier)
and soils at depth (wetter) may exist. At lower temperatures, less of a gradient may exist. Because most of the
gamma photons are emitted from surface soils, they reflect radionuclide concentrations less diluted with water
than in bulk soils. Hence, higher apparent concentrations are measured at higher temperatures.
To summarize, an average higher temperature will result in higher HPGe measurements. However, the effect is
small, and the variation in measured concentrations due to other factors greatly exceeds any temperature effect
on measured HPGe concentrations. Thus, for all practical purposes, temperature can be ignored as having a
significant effect upon HPGe data.
EFFECT OF HUMIDITY ON HPGe MEASUREMENTS
Regression lines fitted to plots of concentration as a function of humidity for total uranium, thorium-232,
potassium-40, and radium-226 have slopes very near zero and extremely low correlation coefficients (expressed
as R2 values). Further, the slopes of concentration vs humidity are generally not significantly different than zero.
These facts demonstrate that humidity has little effect upon HPGe measurements.
CONTROL CHART FOR TOTAL URANIUM
Parameters other than temperature, humidity and soil moisture could also possibly affect HPGe measurements.
However, rather than collect a voluminous amount of data for multiple parameters, the use of control charts is
employed instead to evaluate the cumulative effect of environmental and weather conditions upon HPGe
measurements. Initial "means" control charts were constructed using typical conventions (warning limits are ±2
standard deviations from the mean; control limits are ±3 standard deviations from the mean). All of the data
collected between April 8, 1997 and March 31, 1998 were utilized in calculating standard deviations in order that
the standard deviations represent data collected over a wide range of environmental, climatic, and weather condi-
tions. Table 1 shows values of means, standard deviations, standard deviations as percentages of means,
warning limits, and control limits on both a wet weight and dry weight basis.
One significant aspect of the data in Table 1 is that the standard deviation as a percent of the mean for the two
radionuclide averages approximately 6% on a dry weight basis. The standard deviations shown in Table 1 are
interpreted to represent the long-term total system uncertainty, and this longterm total system uncertainty is very
good, typically less than 10%.
An example control chart displaying data resulting from all of the HPGe measurements performed between April 8,
1997 and March 31, 1998 is presented in Figure 3 for total uranium on a dry weight basis. The trends of increasing
total uranium concentrations in June and in July, and in August and September, for example, represent the
periods of drier soil. Figure 3 also clearly shows that total uranium for the winter months of November, December,
January and February is lower than for the summer months. This results from soil moistures being consistently
higher for the winter months than for the summer months.
Note that the x axis of Figures 3, 4, and 5 is entitled "Data Index." A given indice is merely an abbreviation of the
date and time the measurements was taken. For example, an indice of 41 signifies April 1. Indices of 513a and
513p indicate that the measurements were made on May 13 in the morning and in the afternoon. Lowercase "a"
and "p" in Figures 3, 4, and 5 indicate a.m. and p.m., respectively.
CONTROL CHARTS FOR RADIUM-226
Whereas data points for total uranium, thorium-232, and potassium-40 are predominately within warning and
control limits, the situation for radium-226 appears quite different. As shown in Figure 4, numerous radium-226
measurements fall outside warning and control limits.
Table 2 compares the mean and standard deviation of radium-226 measurements taken in the morning and after-
noon. Clearly, the means and standard deviations of morning measurements are substantially greater than means
and standard deviations of afternoon measurements. More specifically, morning means are 25% higher than after-
noon means, and morning standard deviations are approximately three times greater than afternoon standard
deviations. "F" tests indicate that morning standard deviations are statistically significantly different than afternoon
standard deviations at the 95% confidence level, while "t" tests indicate that differences between morning and
afternoon means are statistically significant at the 95% confidence limits. Examination of an expanded control
chart (Figure 5) demonstrates very well that for radium-226 measurements taken on the same day, very often the
109
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
morning measurements are higher than the afternoon measurements. Because radium-226 is determined from
gamma rays emitted by radon-222 daughters, the differences between morning and afternoon measurements are
related to radon buildup and its subsequent dissipation from soils. Typically, at the FEMP weather conditions in the
morning are favorable for "bad radon days." That is, morning weather conditions are not favorable for the dissipa-
tion and dispersion of radon accumulations from very near to the surface of soils to the atmosphere. Conversely,
by late morning or early afternoon weather conditions are such that near surface radon has dissipated and
dispersed. Usually, mornings with fog also had high measured concentrations of radium-226; thus, one indicator
as to whether HPGe measurements for radium-226 should be carried out is the presence of fog.
The effect of environmental influences on measurements for radium-226 is an important issue and has major
practical ramifications. Morning measurements for radium-226 can be anomalously high due to radon accumula-
tions near the ground surface, and afternoon radium226 measurements generally have a much lower degree of
variation among them than morning measurements (Table 2). These observations were important considerations
in developing a methodology to compensate for radon disequilibrium.
SUMMARY
1. Soil moisture has a significant effect upon the magnitude of HPGe measurements when concentrations of
radionuclides are calculated on a wet weight basis. Soil moisture has a minor effect upon HPGe measure-
ments when concentrations are calculated on a dry weight basis. This effect is likely related to gradients of
moisture from the soil surface to depth (10 inches).
2. Temperature has a minor effect upon HPGe measurements over the range of 14° F to 93° F. This effect
may be related to gradients of moisture from the surface of soils to soils at depth (10 inches).
3. Humidity has little observable effect upon HPGe measurements.
4. Weather conditions have significant effects upon HPGe measurements to determine radium-226 concen-
trations. Because HPGe actually measures gamma photons emitted by radon-222 daughters to calculate
radium-226, weather conditions leading to the buildup and dissipation of radon in surface soils greatly
affect the concentration of radium-226 calculated from HPGe measurements.
5. Typically, morning radium-226 concentrations are higher than afternoon radium-226 concentrations as
calculated from HPGe measurements. From April 8, 1997 through March 31, 1998, morning radium-226
concentrations averaged over 25% higher than afternoon concentrations with a high degree of variability
associated with that average.
6. Control charts were established for total uranium based upon the standard deviation of all measurements
made at the FCS from April 8, 1997 to March 31, 1998. Excellent long-term precision was observed for
this analyte as the standard deviation of the measurement population averaged 6% of the population
mean.
Table 1. Statistical calculations for control charts
Parameter
N=
Mean=
Std Dev.=
Std Dev. as % of Mean
UCL*=
UWL*=
LCL*=
LWL*=
Total Uranium
(ppm)
Wetwt.
250
74.4
7.70
10.4
97.4
89.8
51.2
58.9
Dry wt.
250
93.4
5.56
5.96
110.0
104.5
76.7
82.2
Thorium-232
(pCi/g)
Wetwt.
250
0.91
0.09
10.3
1.19
1.09
0.63
0.72
Dry wt.
250
1.14
0.07
5.83
1.34
1.28
0.94
1.01
UCL = upper control limit
UWL = upper warning limit
LCL = lower control limit
LWL = lower warning limit
110
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 2. Means and standard deviations of morning and afternoon Radium-226 concentrations
Time
Morning
Afternoon
Mean, Wet Wt.
(pCi/9)
1.04
0.84
Std Dev., Wet Wt.
(pCi/g)
0.28
0.08
Mean, Dry Wt.
(pCi/g)
1.30
1.05
Std Dev., Dry
Wt. (pCi/g)
0.31
0.10
110.0
105.0
65.0
60 D
Figure 1. Total
Uranium (Dry
Wt.) as a
Function of Soil
Moisture
Content
10.0 120 140 16.0 180 200 220
Soil Moisture (%)
240 260 280 300
Figure 2. Total
Uranium (Dry
Wt.) as a
Function of
Atmospheric
Temperature
1100
100D
900
80.0
Si
§
70.0
BOD
500
40.0
30 0
10
20
30
I*2-0.1881
40
50 60
Temperature (F)
70
80
90
100
111
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
120.0
1TOO
100.0'
E
a SOD-
e
80.0 •
70.0-
60.0-
UCL-110JD
g:..f
'...J */•*.
r
LCL= 76 J
50.0
Date Index
Figure 3. Control Chart for Total Uranium (Dry Wt. Basis)
2.50
2.30 -
2.10 -
190 -
» 1 70 -
8
I? 1.50 |
£ 1.30"
1 10 •
090 -
070 -
0.50
UCL=Q.77
5 S R
(0 CM ^T Q- CO Q. 0-
-—
tn UD
oo CD a>
^- m •«- ?-
Date Index
Figure 4. Control Chart for Radium-226 (Dry Wt. Basis)
112
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
2.50
2.50
-•2.00
- 1:50
0.00
••i.oo
• • 0.50
0.00
I
Figure 5. Expanded Control Chart for Radium-226 (Dry Wt. Basis)
USING ACID MINE DRAINAGE TO DETOXIFY HEXAVALENT CHROMIUM LEACHATE
FEASIBILITY FOR COAL GENERATED ELECTRIC POWER
H. M. 'Skip' Kingston
Professor
Dengwei Huo
Graduate Assistant
Bayer School of Natural and Environmental Science, Center for Microwave and Analytical Chemistry,
Duquesne University, 600 Forbes Avenue, 308 Mellon Hall, Pittsburgh, PA 15282
Randy Cain
Environmental Coordinator
Allegheny Power, 800 Cabin Hill Drive, Greensburg, PA 15601
A direct link between the production of Cr (VI) in coal fired electric power generation waste has been established.
This is one of the first studies to link the production of Cr (VI) in the process of coal fired electrical power plants to
the combustion conditions found in the facility. The study also evaluated aged and buried waste for stability and
leaching of Cr (VI). Raw material and cored material were evaluated to assess contribution and species contents
including stability. Run-off from many different sources was evaluated for stability and transportation of the
chromium species.
Field and bench scale tests, were completed that demonstrated the effectiveness of acid mine drainage (AMD) in
remediation. It was demonstrated that this waste stream can be used effectively to reduce the hexavalent
chromium in leachate from a coal combustion fly ash landfill. Speciated isotope dilution mass spectrometry
(SIDMS), was used to fully characterize the chromium species in many materials and leachates and to profile the
chemical interconversions of the chromium species when the leachate and AMD were combined.
Comparison of this remediation scenario against conventional methods like direct chemical treatment or passive
113
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
wetland treatment proved to be economically and environmentally favorable. The study focuses not only on direct
evaluation of the problem but includes the economic and scientific evaluation of the methods of measurement and
remediation.
This study may be a useful demonstration of the use of one waste stream to detoxify another that is economically
and scientifically feasible. It may be a viable solution for as many as half of the coal fired electrical generation
stations each of which have some form of this problem that needs to be addressed in the future.
114
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
ENVIRONMENTAL
BUSINESS IN
THE PBMS PARADIGM
115
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
116
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
THE SHELL FOR ANALYTICAL CHEMISTRY REQUIREMENTS FOR USAGE PROJECTS
Cheryl Groenjes
Chemist
U.S. Army Corps of Engineers, CENWO-HX-C, 12565 West Center Road, Omaha, NE 68144
The purpose of the 'Shell' is to establish the basic approach to be used when performance-based methods,
especially the SW-846 methods, are employed by the U.S. Army Corps of Engineers (USAGE) for the analytical
testing of environmental samples. These methods are flexible and can be readily adapted to individual project-
specific requirements. The chemistry data generated for USAGE projects must be produced by a process or
system of known quality to withstand scientific and legal challenge relative to the use for which the data are
obtained. The 'Shell' outlines such a process. Additionally, the 'Shell' applies the concepts to specific SW-846
methods for relatively critical data uses.
Project-specific data quality objectives (DQOs) must be established for both the field and laboratory operations.
Any differences between project DQOs and lab operational criteria must be reconciled before project execution.
For each project, data quality must be demonstrated for the analytes of concern at the levels of concern. However,
in order to promote flexibility while maintaining some degree of consistency, when no project specific DQOs exist,
the 'Shell' is used to establish the project analytical requirements.
Laboratories are required to maintain written, approved SOPs for all methods and operations. The demonstration
of method proficiency begins with establishing the basic sensitivity of the method by determining the method
detection limit (MDL). The relationship between the MDL, the method quantitation limit, the initial multi-point
calibration curve, and the laboratory's method reporting limit is established. A laboratory cannot claim to reliably
quantitate values below the low standard or above the high standard. A given method is suitable when the labora-
tory's low standard is below the site action level for each analyte of concern.
The 'Shell' describes the requirements for instrument calibration, calibration verification with standards from an
independent source, and continuing calibration procedures, while maintaining a level of flexibility, which may be
exercised, based on analyst judgement. Each preparation batch is to contain a method blank and a laboratory
control sample containing all of the project-specific analytes of concern spiked at the levels of concern to monitor
laboratory performance. Each preparation batch would typically contain additional QC samples to monitor the
effect of the matrix on the method. Corrective actions are carefully detailed and involve interaction with project
managers to avoid the generation of a significant amount of flagged or unusable data.
The intent of the 'Shell' is to ensure the generation of chemistry data of known quality. Laboratories employ
chemists and others who are experts in the interpretation of analytical data. Much is to be gained by enhancing the
interaction between the laboratory and project personnel. The 'Shell' encourages this.
117
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
118
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
ORGANIC
ANALYSIS
119
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
120
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
QUESTIONABLE PRACTICES IN THE ORGANIC LABORATORY: PART II
Joseph F. Solsky
Chemist
US Army Corps of Engineers (CENWO-HX-C), 12565 W Center Road, Omaha, NE 68144
ABSTRACT
During recent environmental laboratory audits conducted by the USAGE, certain 'questionable practices' have
been observed, especially in the organic analysis areas.
Most people have a relatively good idea of what constitutes a fraudulent activity today. The concepts of
'dry-labing,' 'peak shaving,' 'peak enhancing,' or 'time-traveling' are well understood. These practices clearly
involve the deliberate manipulation and/or alteration of data, often to achieve or meet method QC criteria. Unfortu-
nately, these practices are still being observed today. In addition, there are a new group of 'questionable practices'
now being observed that often involve the selective exclusion of data to achieve or meet method QC criteria.
Examples of some of these practices include the following: (1) Dropping points during initial calibration to meet
method criteria. (2) Reporting very tight QC performance ranges when actual lab control charts show a signifi-
cantly wider range. (3) Dropping points to achieve a lower Method Detection Limit (MDL). (4) Performing tunes by
picking the scan or series of scans that will meet the desired criteria after the original tune had failed. (5) Perform-
ing initial calibration curves but never verifying that the peaks used for the calibration actually represented the
target analyte.
These practices are often described as 'the common approach used by everyone,' yet when described to people
within EPA (e.g., the MICE Hotline), the clear response is that these approaches were never intended within the
context of SW-846, although not explicitly addressed nor prohibited.
INTRODUCTION
The US Army Corps of Engineers (USAGE) currently executes remedial and compliance activities under several
environmental regulatory programs. The analytical testing of various environmental samples is often a significant
part of these activities. The data must be produced by a process or system of known quality to withstand scientific
and legal challenge relative to the use for which the data are obtained. To give the USAGE programs the greatest
flexibility in the execution of its projects, the SW-846 methods, as published by EPA, are generally the methods
employed for the analytical testing of environmental samples. These methods are comprehensive and flexible and
can be readily adapted to individual project-specific requirements. As stated in the Final Rule that incorporated the
Third Edition of SW-846 (and its updates) into the RCRA regulations, this appendix is required to be used for
certain activities in the RCRA program. In other situations, this EPA publication functions as a guidance document
setting forth acceptable, although not required, methods to be implemented by the user, as appropriate, in satisfy-
ing RCRA-related sampling and analysis requirements.
During recent laboratory audits conducted by the USAGE, certain 'questionable practices' have been observed,
especially in the organic analysis areas. Prior to project execution, the USAGE may conduct a review of the
laboratory that was proposed for use on that specific project. This review typically consists of three phases: (1)
documentation review; (2) analysis of Performance Evaluation (PE) samples; and (3) on-site laboratory audit.
Additional follow-up audits can also be conducted. These 'questionable practices' have been noted during all
phases of these laboratory reviews.
The concepts of 'dry labbing', 'peak shaving', 'peak enhancing', or 'time traveling' are well understood. These
practices clearly involve the deliberate direct manipulation and/or alteration of data, often to achieve or meet
method QC criteria. Laboratory professionals clearly recognize these practices as inappropriate since no profes-
sional reason exits to employ them other than to meet specific contractual requirements and avoid potential penal-
ties. There is no technical basis that can justify the use of these practices. The impact on data usability must be
determined on a project by project basis. Unfortunately, these practices are still being observed today. When fraud
is detected in conjunction with USAGE projects, the Corps is attempting to separate any criminal/civil charges from
the actual impact of the fraud on data usability (e.g. to separate legal from technical issues).
As the nation moves away from the use of strict method protocols to a more performance based approach, the
laboratories will have more discretion as to how methods are actually implemented. This will allow the laboratory
community to take faster advantage of new technologies to cut costs and improve data quality. This move will
121
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
place pressure on the laboratory community to employ knowledgeable experts to properly implement these newer
technologies in a scientifically justifiable manner and to provide the enhanced documentation that will be needed.
Current market over capacity has caused bidding wars and corner cutting. This move will place pressure on the
regulator community to properly define what a performance based measurement system is and how its quality
should be defined. This move will place pressure on the buyer of analytical services to better define the Data
Quality Objectives (DQOs) such that the appropriate data can be obtained for any given project at a fair and
appropriate cost. At the present time, issues exist in all these areas that can and are compromising data quality.
During this transition, USACE is observing a new group of 'questionable practices'. Many of these practices
involve the selective exclusion of data to achieve or meet current method QC criteria rather than the direct
manipulation of any single data point.
QUESTIONABLE PRACTICES
The first example of such 'questionable practices' involves laboratory documentation, including Quality Control
Plans and Standard Operating Procedures (SOPs), that do not accurately reflect what the laboratory actually does.
Many of these plans contain statements that are misleading, in error, or simply incomplete. These laboratory
documents are often directly incorporated into project specific Quality Assurance Project Plans (QAPPs) or Work
Plans. Often, these laboratory documents are not carefully read or reviewed before incorporation. They should be.
Do misleading, erroneous, or incomplete statements justify these practices? Probably not.
The second example of such 'questionable practices' involves establishing initial calibration curves. Laboratories
have been observed running six or more standards for methods that state 'a minimum of five points should be
used to establish the initial calibration curve'. Points are then discarded, while maintaining a minimum of five
calibration points, throughout the curve until the appropriate QC criteria can be met. No technical justification
existed for the deletion of these points other than to meet the method QC criteria. This practice is often justified by
using the rationalization that a 'better curve' is generated. Another reason heard is that 'everyone is doing it1.
Points can only be rejected for inclusion in the curve if a known error was made or if a statistical evaluation
indicates that the point can be discarded. When multiple target analytes are included in each calibration standard,
it may become necessary to discard selected upper or lower points for individual target analytes. Points can be
discarded at the upper end of the curve if the linear range of the detector has been exceeded. For these cases,
the laboratory must dilute samples that exceed the highest point of the calibration curve. Points can be discarded
at the lower end of the curve if the detector is not producing a response. For these cases, the laboratory quantita-
tion limit must be adjusted accordingly. Under no other circumstances can points be discarded. If QC criteria
cannot be met, the instrument system may be unstable or the calibration solutions may be incorrectly prepared.
The 'best curve' is obtained when all valid points are included in the initial calibration curve.
The third example of such 'questionable practices' involves the verification of initial calibration curves through the
use of continuing calibration verification (CCV) solutions. Laboratories have been observed averaging the % differ-
ence or % drift across all target analytes even when several of the target analytes exceed the criteria by a signifi-
cant amount such that the average still meets the criteria as stated in the method. For example, when method
8021 is used, it is often difficult for laboratories to meet the CCV criteria for many of the gaseous target analytes.
Method 8000B states the following: '..., if the average of the responses for all analytes is within 15%, then the
calibration has been verified'. This language was chosen to make it easier for laboratories to implement this
method when certain problem analytes, i.e. the gases in method 8021, marginally exceed the stated method crite-
ria. It was never intended to allow the inclusion of obviously 'bad' data to make it 'acceptable'. Method 8000B goes
on to say: '..., and the data user must be provided with the calibration verification data or a list of those analytes
that exceeded the 15% limit'. If the QC criteria cannot be met, the instrument system may be unstable or the
calibration verification solution may be incorrectly prepared.
The fourth example of such 'questionable practices' involves the reporting of acceptance ranges for laboratory
control samples (to include surrogates). Laboratories have been observed reporting a very tight range for these
QC samples on laboratory report sheets, indicating that they have good method control. However, an examination
of actual control charts maintained by the laboratory shows a significantly wider range, if control charts are even
available. This practice is often justified by using the rationalization that 'but the LCS was within the QC range,
therefore, it must be okay'. Method 8000B stresses the importance of control charts to track laboratory perform-
ance. The ranges generated should then be compared to method established criteria. If a 'match' is not obtained,
then the laboratory should consider modifying their method to improve its performance. Simply reporting data
under this circumstance since the LCS 'met the method criteria' is unacceptable since it misleads the user of the
data and misrepresents the laboratory's reported data quality. Control chart ranges must also be reasonable. The
122
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
issue of control charts as related to what analytes need to be charted (all target analytes or just a subset), in what
QC samples (LCSs, MSs, LCSs and MSs combined, etc.), at what spiking levels (action levels or mid-level), and
appropriate recovery ranges (what would be considered a reasonable range for a given method) needs further
clarification in the SW-846 methods. Many laboratories do not understand the significance of these charts and
how to properly implement and use them.
The fifth example of such 'questionable practices' involves the reporting of wide matrix spike (MS) recovery
ranges. This item is related to the fourth example as given above. Laboratories have been observed reporting
very wide ranges for these QC samples on laboratory report sheets. However, an example of the actual ranges as
derived in the laboratory shows a significantly narrower range. This practice is often justified by using the rationali-
zation that 'by widening the ranges, less of our data is rejected'. No method is immune to all possible interferences
and not all interferences can be predicted. Therefore, it is important to monitor for these effects. The purpose of
the matrix spike (MS) is to see if a possible matrix effect is impacting the data quality. When the MS QC range is
exceeded, clients would normally be contacted to see if data flagging is appropriate, sample(s) should be rerun,
the method should be modified (i.e., add a clean-up step) to better deal with the interference, or a different method
chosen that is not affected by the interference. Data users should not penalize a laboratory, or its data, due to the
presence of reported potential matrix interferences. At the same time, laboratories should not flag all poor recover-
ies as possible matrix effects, especially in blanks, LCSs, etc. Good judgment should be used by all parties
involved.
The sixth example of such 'questionable practices' involves the determination of the method detection limit (MDL).
Laboratories have been observed running eight or more standards and then discarding points to achieve a lower
MDL. No technical justification existed for the deletion of these points other than to achieve a lower MDL. This
practice is often justified by using the rationalization that a 'better (lower) MDL' is generated. Points can only be
rejected if a known error was made or if a statistical evaluation indicates that the point can be discarded. Under no
other circumstances can points be discarded. The MDL study must be performed at the appropriate level with a
reasonable recovery of the target analyte(s) noted. The 'best MDL' is obtained when all valid points are included. It
appears that the industry is placing too much emphasis on this concept. Laboratories are being driven to report
lower and lower levels of contaminants. Perhaps the industry would be best served by using the performance-
based concept to demonstrate what a given method run by a given lab could actually 'see' (the concept of the MDL
check sample). The issue of the 'not detected' target analyte has caused great confusion ('detection limit1 versus
'quantitation limit' versus 'reporting limit').
The seventh example of such 'questionable practices' involves tuning a GC/MS detector. Laboratories have been
observed performing tunes in an inconsistent manner, such as picking a single scan or a series of scans that meet
the desired criteria. Single scans have been observed being used at various locations across the peak, including
single points being used on the peak tail. The use of an average of two or more scans have been observed over
various parts of the peak (front, tail, over apex), to even include more background scans than peak scans in the
average. These various schemes would be used when the recommended approach (average of three scans over
the peak apex minus a background scan) would fail the desired criteria. No technical justification existed for using
these various approaches other than to meet the method QC criteria. This practice is often justified by using the
rationalization that 'as long as a scan(s) can be found that passes, the instrument is in tune'. Different tune
parameters may be needed to optimize instruments from a given manufacturer. However, a consistent approach
must be used to evaluate whether the instrument is 'in tune'. A laboratory cannot simply pick and choose whatever
scan(s) happens to meet criteria on any given day.
The eighth example of such 'questionable practices' involves the misidentification of GC/MS peaks during initial
calibrations and during continuing calibration verifications. Laboratories have been observed performing these
calibrations but never verifying the identity of the peaks observed. These systems can make errors in the identifi-
cation of target analytes especially when more than one peak is present in the retention time window. As a conse-
quence, laboratories can generate calibration curves for the wrong target analyte. This has been observed for
certain Appendix IX compounds and for certain poor performing target analytes in methods 8260 and 8270. Instru-
ment raw data must be reviewed by the analyst to ensure that all peaks have been correctly identified, all peaks
are clearly visible and all peak shapes are appropriate for the target analyte being measured.
The ninth example of such 'questionable practices' involves performing continuing calibration verifications where
the majority of the target analytes have missed their assigned retention time windows. Laboratories are performing
unnecessary manual integrations to 'find' peaks that have missed these windows. This is very dangerous since
peaks can be easily missed during the analysis of samples resulting in the reporting of false negative data. The
123
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
SW-846 methods directly address retention time window criteria for the internal standards for the GC/MS methods
but do not address any requirements for these windows for the target analytes. When such windows are missed,
this should be a clear signal to the analyst that the system is out of control and corrective action is required. Such
corrective action should include a system inspection along with repeating the initial calibration or updating the
retention time windows for the target analytes.
Should the above 'questionable practices' be considered as examples of fraudulent activities? Some of the labora-
tories have described these practices as 'the common approach used by everyone', yet when described to people
within EPA (e.g., the MICE Hotline), the clear response is that these approaches were never intended. Potential
solutions might include the following: 1) More prescriptive methods (probably not) or more clearly written
guidance? 2) Training for lab staff on GLP to include statistics? 3) Rethinking the way laboratory services are
contracted for? 4) Collecting additional data from the laboratory for more detailed data validations? 5) The use a
standardized data reporting format and better data validation software? 6) Etc.
SUMMARY
The problems/issues noted above are very serious and directly impact on the usability of the data generated.
Oftentimes the impact is equivalent to the impacts observed during past demonstrated cases of fraudulent data
manipulation. How did we get to this point? There probably is no one single cause. One certain contributing factor
is the price paid for these services. It is not uncommon to encounter projects that were bid low simply 'to get one's
foot in the Federal door' Simply put, the price paid for these services was not sufficient to cover the costs of
producing the product. The fault lies both with the laboratory community for bidding in this manner and the govern-
ment for accepting bids based on low price only without considering the quality factor ('best value' procurement
strategy). However, this is a free market economy. The age old adages let the buyer beware' or 'you get what you
pay for' certainly apply here.
Another factor is the level of expertise that now exists at the laboratory level. Some laboratories have let go their
most experienced staff since they could no longer 'afford' them. Many people feel that the computer attached to
the instrument in use will give them the correct answer without additional thought. If anything, more expertise is
needed to evaluate the larger magnitude of data moving through the laboratory and the complexity of today's
instrumentation. Laboratories should not be treated as black boxes. It is not uncommon for this author to visit a
given laboratory and find that laboratory staff know very little about the fundamental chemistry of the method (or
the software) in use. One common phrase heard often is 'but the method doesn't specify the approach to use'.
This raises the question as to whether or not very prescriptive methods should be written. Yet each of the SW-846
methods typically state the following: This method is restricted to use by, or under the supervision of, analysts
experienced in the use of gas chromatography/mass spectrometers, and skilled in the interpretation of mass
spectra and their use as a qualitative tool'. Additional training of laboratory staff should be emphasized. Peer
review of raw data should also be emphasized. Audit trails should be 'turned on' when available and reviewed on a
regular basis.
More review of raw data would be encouraged. Most of the data generated within a laboratory is generated in an
electronic format. Yet much of the data is still manually managed and reviewed. A greater emphasis should be
placed on receiving data electronically and for the electronic screening/review of this data. To assist this process,
standardized electronic data reporting formats should be used. Standard file formats have been developed for
several of the instrumental methods that can transfer data electronically in a standard file format between an
instrument, or its data station, and a laboratory LIMS system. However, this standard is not often used. No
standard file format has been developed for the transfer of information from a laboratory, or its LIMS system, to
the data user. The use of a common data dictionary along with a common file structure, such as that proposed by
the Department of Energy Environmental Management Electronic Data Deliverable Master Specification (DEEMS)
would be encouraged.
Certainly, other contributing factors are involved. The 'CLP' prescriptive mentality is still with us. Data validation is
still often performed using a modified version of the National Functional Guidelines. This is not appropriate for the
SW-846 methods and further emphasizes the prescriptive, rather than performance based, approach. The move
to a performance based approach for the analysis of environmental samples is a welcome one. This move is also
being greeted with uneasiness. The approach will place additional burdens on the laboratory community, regulator
community, and the buyer of analytical services. Good communication will be very important to ensure that the
needs of everyone involved have been met. The writers of the methods must work together with the users of the
methods to minimize misunderstandings. It would be recommended that EPA revise Chapter One of SW-846 to
more clearly describe this approach. The 'questionable practices' as described above are serious issues that must
124
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
be resolved. Their timely resolution will give data users the confidence they need to make appropriate project
decisions while at the same time using our tax dollars wisely.
COMPARISON OF FIVE SOIL EXTRACTION TECHNIQUES FOR PESTICIDE AND SEMIVOLATILE ANALYSIS
Rick McMillin. David Spencer, Diane Gregg and Lisa Wool
The objective of this study is to compare some relatively new environmental soil extraction techniques to each
other and to the older techniques. This study will be looking at precision and recovery data for each technique
using a certified spiked soil sample. This is a continuation of previous work which brought up some questions that
are expected to be answered with this new study and a modified experimental design. A sixth procedure will also
be included (which is really a modification of a method rather than a completely new method) which will be the
abbreviated microwave modification. This simple modification eliminates or reduces the concentration step and
has great potential for lab use by significantly reducing extraction time and solvent consumption (pollution
prevention).
Certified spiked soils will be extracted in replicate by the various techniques. The replicate extractions will be split
over several days with each technique being performed the same day. The single extract from a 10 gram sample
will be split between semivolatile and pesticide analysis, effectively resulting in a 5 gram extraction for each.
Extraction techniques will include microwave, pressurized fluid extraction (using the ASE™), automated soxhlet
(using the Soxtherm™), standard soxhlet, and sonication. Also included will be the abbreviated microwave
modification. Extraction and analysis will be by standard EPA methodology. Precision and recovery data will be
presented in addition to time comparisons of the different techniques.
FREEZER STORAGE OF SOIL SAMPLES CONTAINING VOLATILE ORGANIC COMPOUNDS
Alan D. Hewitt
Research Physical Scientist
U.S. Army Cold Regions Research and Engineering Laboratory, 72 Lyme Road, Hanover, NH 03755-1290
Telephone: 603-646-4388; E-mail: ahewitt@crrel.usace.army.mit
ABSTRACT
This study evaluates freezer storage (-12±3°C) as a sample preservation method for volatile organic compounds
(VOCs) in soil. Five different soil matrices, spiked with several aromatic and/or chlorinated hydrocarbons at less
than 0.2 mg/kg, frequently showed no significant change in concentration after being frozen and stored for up to
12 days. Furthermore, with the exception of garden soil, 88% or more of the analyte concentrations were retained
after a two-day transportation period when held at 4±2°C.
INTRODUCTION
The options that are currently recommended by the U.S. Environmental Protection Agency (EPA) (Method 5035)
and the American Society for Testing and Materials (ASTM) (D4547-98) for soil sample collection and
preparation/preservation are (1) the immediate infield transfer of a sample into a weighed volatile organic analysis
(VOA) vial that either contains VOC-free water acidified to a pH of 2 so that a vapor partitioning (purge-and-trap or
headspace) analysis can be performed without re-opening or that contains methanol for analyte extraction in
preparation for analysis, or (2) the collection and up to two-day storage of intact samples in an airtight container
before initiating one of the aforementioned sample preparation/preservation procedures1'2 In both cases samples
should be held at 4±2°C while being transported or stored.
At the time these recommendations were made, two chemical preservation procedures, methanol immersion and
acidification to a pH of 2 or less with sodium bisulfate, received the most attention. It was recommended that
125
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
methanol preservation be used only when samples were anticipated to contain concentrations of VOCs in excess
of 0.2 mg/kg, and acidification be used when the concentrations were expected to be less than this value. Once
preserved, the pre-analysis holding period could be extended up to 14 days after sample collection. Other means
of biological preservation, such as lowering the storage temperature to below 0°C, although briefly mentioned, did
not receive as much attention as these chemical preservation procedures, because of insufficient validation.
The first sampling option described has the field personnel initiate sample preparation/preservation during the
collection activity, and may require that they handle solutions and weigh the sample collection vessels3. The
second option, which is the focus of this paper, allows for the transportation and storage of samples, so that
preparation/preservation can be performed in a laboratory setting. This study evaluates the extended storage of
discrete (5-g) samples at -12±3°C (commercial freezer) as a means of preserving samples prior to either methanol
extraction or analysis by vapor partitioning (i.e., purge-and-trap or headspace), subsequent to a 48-hour
transportation period during which samples were held at 4±2°C.
EXPERIMENTAL METHODS
After obtaining 5.0±0.1 g of the soil with a modified syringe, a pilot hole was made with a needle into the middle of
the sample plug. Then a 10-uL glass syringe was used to transfer into this cavity a small aliquot of aqueous
solution containing approximately 50 mg/L of some or all of the following analytes: trans-1,2-dichloroethene
(TDCE), cis-1,2-dichloroethene (CDCE), trichloroethene (TCE), tetrachloroethene (PCE), benzene (Ben), toluene
(Tol), ethylbenzene (E-Ben), p-xylene (p-Xyl), and o-xylene (o-Xyl). The resulting sample concentration was less
than 0.2 mg/kg for each analyte. Immediately after spiking, the syringe barrel was wiped clean, inserted into the
mouth of a 40-mL VOA vial, the sample extruded, and then the vial was capped. For these experiments a variety
of soil types, replicate samples, storage periods, and conditions were used (see Table 1).
After all the samples had been prepared, 5.00 mL of methanol was introduced to the first, iniddle, and last
replicate samples to estimate the day-zero (DO), extracted within an hour of spiking) concentrations. The methanol
was added by piercing each septum with a 23-gauge Luer Lok needle attached to a 5.00-mL glass syringe with a
Luer connector. For the remaining samples, methanol was introduced to triplicate samples after various periods of
refrigeration, or refrigeration and freezer storage (Table 1).
ANALYSIS
All of the samples were analyzed by equilibrium headspace analysis of a 0.500-mL aliquot of the methanol
supemate after a 24-hour extraction period (Method 5021). The preparation of working standards, the
instrumentation, and instrumental setting have been reported elsewhere. Samples prepared by methanol
extraction were corrected for the increase in extraction solution volume, caused by the soil moisture (10 to 20% by
dry wt.). The results of these experiments were evaluated using a one-way analysis of variance (ANOVA) and
least significance difference tests (Fishers Protected LSD), at the 95% confidence level (Table 1).
RESULTS
The relative standard deviation for the concentrations established for the sample triplicates was typically less than
5%. Table 1 shows the percent recovery for each analyte after the various holding times and conditions, relative to
the DO (initial concentration or control) values. For all five experiments there was often no significant difference in
the established analyte concentrations between the D2 and D14 results. Therefore, independent of soil type, once
placed in the freezer, losses were abated even though storage was extended for at least an additional 11 days.
With the exception of the fifth experiment, there also was little or no loss of VOCs relative to DO for the samples
held under these two sets of conditions (temperature and storage period). However, there were large losses of
both Ben and Tol after 48 hours at 4±2°C for the garden loam (fifth experiment). The additional refrigerated
storage periods (D3 and D5) used in the last two experiments show that there was a slow, continuous decrease of
Ben for the CRREL silt/clay, and a fairly rapid loss of all the aromatic compounds for the garden loam, when
stored at 4±2°C.
DISCUSSION
The experiments presented here are part of an ongoing evaluation of various transportation and storage protocols
so that VOC samples can be prepared/preserved within a laboratory setting. Previously, it was shown that when
samples were stored and transported in either a VOA vial or the En Core Sampler (En Novative Technologies,
Inc., Green Bay, Wisconsin), recoveries were often more than 80% after a 48-hour transportation period when
held at 4±2°C. Moreover, often no further significant losses occurred after samples held in these two vessels were
preserved by placing in a freezer (-12±3°C) for up to an additional 12 days4. In addition, it was observed that
losses due to biological degradation have greater impact (i.e., larger reduction in percentage of the initial
126
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
concentration) at lower VOC concentrations4. This earlier effort, however, assessed only one type of soil (CRREL
silt/clay). Here, the CRREL soil and four other soil types were assessed using a VOA vial as a storage vessel
(Table 1). Consistent with the earlier experiments, losses of VOCs were abated when samples were frozen, and
with the exception of the garden soil, recoveries relative to DO were more than 80% after a 48-hour transportation
period when held at 4±2°C. The garden soil, however, showed a 50% or greater reduction in Ben and Tol after 48
hours when refrigerated, independent of being initially frozen for 24 hours.
Clearly, samples held in an airtight vessel and stored in a freezer have been effectively preserved in these studies.
This method of preservation offers several advantages over the recommended infield chemical preservation
option, e.g., no prior knowledge of the VOC concentrations is necessary, few Department of Transportation (DOT)
regulatory requirements must be met, and field personnel don't have to handle chemicals or weigh samples.
Moreover, preservation by acidification cannot be used indiscriminately; that is, this technique cannot be used with
carbonaceous soils or when styrene is a VOC of interest5. An additional concern is that by lowering the pH of
some matrices, the formation of acetone, a regulated compound itself, has been observed4.
Based on the findings for the garden soil, it would not be advisable to hold samples for more than 48 hours at
4±2°C when they are taken from sites where nonhalogenated VOCS are the principal analytes of concern, and
where the data quality objectives (DQOs) call for the determination of low analyte concentrations. Indeed, future
studies may show that this period should be shortened for samples taken to fulfill these objectives. In particular, it
would not be advisable to hold samples for any period of time without chemical or physical (freezing) preservation
if taken from a location receiving a biological amendment. Although the experimental evidence was not shown
here, for sites where halogenated compounds are the analytes of interest and no biological amendment is being
applied, samples are not as susceptible to biodegradation, and therefore would be less likely to be compromised if
held for more than two days at 4±2°C prior to being chemically or physically preserved4'6. Additional information on
how to incorporate freezing as a method of sample preservation into a site sampling plan can be found
elsewhere4.
SUMMARY
Within the last couple of years new guidance has come from the U.S. EPA and ASTM regarding how soil samples
acquired for VOC characterization should be collected and handled in preparation for instrumental analysis. To
assist with the implementation of this new guidance, two very different protocols have been developed. In one
case, all of the steps leading up to those associated with the analysis process are performed in the field. The
other, more traditional approach, has all of the steps associated with sample preparation and analysis occur in a
laboratory. The focus of this paper was to continue the process of evaluating secure transporting and storing of
samples so that the laboratory protocol could be used. This study showed that, with the exception of a garden soil,
the storage of samples in an airtight vessel was found to be consistent with the intent of the new guidance, and in
general 80% or greater of the analyte concentrations were retained over a two-day storage period at4±2°C. The
large losses seen over this period for low concentrations of BTEX compounds in the garden soil, however, indicate
that temperate storage conditions are inappropriate for samples removed from sites receiving biological treatment
or that have biological characteristics similar to a garden soil. Independent of soil type, samples transferred to a
freezer (-12±3°C) often showed no significant change in VOC concentrations over an additional 11 or 12 days of
storage. For several reasons, this method of sample preservation appears to be better suited for VOCs in soil
matrices than acidification. For instance, acidification is incompatible with carbonates, causes the decomposition
of styrene and perhaps other target analytes, and has the potential to cause the formation of acetone. These
findings and observations support the effort to include storage at -12±3°C as a method of sample preservation in
future revisions of these guidance documents.
ACKNOWLEDGMENTS
Funding for this work was provided by the U.S. Army Environmental Center, Martin H. Stutz, Project Monitor. The
author thanks Dr. Thomas F. Jenkins and Marianne E. Walsh for critical review of the text.
This publication reflects the view of the author and does not suggest or reflect policy, practices, programs, or
doctrine of the U.S. Army or of the Government of the United States.
REFERENCES
1. U.S. Environmental Protection Agency (1986) Test methods for evaluating solid waste, Vol. 1B, SW-846.
2. American Society for Testing and Materials (in press) Standard guide for sampling waste and soils for volatile
organic compounds, ASTM D4547-98.
3. Hewitt, A.D., T.F Jenkins, and C.L. Grant (1995) Collection, handling, and storage: Keys to improved data
127
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
quality for volatile organic compounds in soil. American Environmental Laboratory, Vol. 7: 25-28.
4. Hewitt A.D. (in press) Storage and preservation of soil samples for volatile organic compound analysis. U.S.
Army Cold Regions Research and Engineering Laboratory, Hanover, New Hampshire, Special Report.
5. Hewitt, A.D. (1995) Enhanced preservation of volatile organic compounds in soil with sodium bisulfate. U.S.
Army Cold Regions Research and Engineering Laboratory, Hanover, New Hampshire, Special Report 95-26.
Table 1. Storage times, conditions, and % recoveries of analyte concentrations for triplicate samples relative to
the DO (initial) values.
Storage Percent recovery
1st
2nd
3rd
4th
5th
(Days)
2
14
2
14
2
14
2
3
5
14
2
3
5
14
Soil Type
Fort
Edwards
clay
Fort
Edwards
clay
WES
silt/sand
WES
silt/sand
Wisconsin
sand
Wisconsin
sand
CRREL
silt/clay
CRREL
silt/clay
CRREL
silt/clay
CRREL
silt/clay
Garden
loam
Garden
loam
Garden
loam
Garden
loam
Storage
conditions
4°C
4°C,
2D**/-12°C,
12D
4°C
4°C,
2D/-12°C,
12D
4°C
4°C,
2D/-12°C,
12D
4°C
4°C
4°C
-12°C,
1D/4°C,
2D/-12°C,
11D
4°C
4°C
4°C
-12°C,
1D/4°C,
TDCE CPE Ben
c_
97af 96a,b 100a
82b 90b 95a
92a 99a 96a
92a 100a 99a
97a 98a 97a
93a 99a 98a
91b
85c
78d
92b
42b
27d
11
37c
ICE. Tal
103a 101a
94b 94b
93a 94a
98a 99a
97a,b 101a
94b 96b
98a
91b
91b
102a
50b
35c
16d
47b
PCE E-Ben
98a 100a
89b 95a
94a 92b
98a 101a
98a 98a,b
86b 92b
101a
97b
97b
100a9
8a
80b
72c
40d
81b
p-Xyl o-Xyl
99a 102a
93b 98a
97a 95a
101 a 98a
97a 98a,b
88b 95b
98a
92c
95b
84b
78b
54c
84b
2D/-12°C,
11D
*Experiment
fValues with common letter are not significantly different at the 95% confidence interval (ANOVA and Fishers
protected LSD) for each experiment. The letter "a" was assigned to the DO values.
** Days
128
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
PERFORMANCE OF THE DISPOSABLE EN CORE® SAMPLER
FOR STORING SOIL FOR VOLATILE ORGANIC ANALYSIS
Susan S. Sorini
Principal Scientist
John F. Schabron
Manager
Chemical Monitoring Division, Western Research Institute, 365 N 9th Street, Laramie, WY 82072
ABSTRACT
Soil sampling and storage practices for volatile organic analysis must minimize loss of volatile organic compounds
(VOCs) from samples. The En Core® sampler is designed to collect and store soil samples for volatile organic
analysis in a manner that minimizes loss of contaminants due to volatilization and/or biodegradation. The sampler
consists of a coring body/storage chamber, O-ring sealed plunger, and O-ring sealed cap, all of which are
constructed of an inert composite polymer making the device chemically compatible with soil matrices and
contaminants. The devices are designed to collect and hold a soil sample of either 5-grams or 25-grams during
shipment to the laboratory for analysis. After the sample is collected in the En Core sampler, the coring body is
sealed with the slide-on, locking cap and immediately becomes a sample storage chamber.
The En Core sampler has undergone extensive testing during development to determine design specifications,
and after development to evaluate performance under various storage conditions. This paper discusses (1) testing
that was performed as part of the developmental work to select an O-ring design to minimize VOC loss from the
device; (2) testing that was conducted to generate performance data on the device for inclusion in a new American
Society for Testing and Materials (ASTM) standard practice for using the En Core sampler; and (3) testing to
evaluate performance of the device to store soils containing low level VOC concentrations. Results show that the
En Core sampler performs well for storing VOC-contaminated soil.
INTRODUCTION
A major concern in sampling soil for volatile organic analysis is maintaining sample integrity during collection and
shipment of soil samples to the laboratory for analysis. Laboratory data can greatly underestimate volatile organic
compound (VOC) concentrations in soil if attention is not paid to sampling and handling techniques. The dispos-
able En Core® device is a soil sampling/storage tool that is designed to collect and store a soil sample for volatile
organic analysis in a manner that minimizes loss of contaminants due to volatilization and/or biodegradation. The
En Core sampler has three components: (1) the coring body/storage chamber, which is volumetrically designed to
collect and store a soil sample of either approximately 5 grams or 25 grams; (2) an O-ring sealed plunger for
nondisruptive extrusion of the sample into an appropriate container for analysis or preservation; and (3) a slide-on
cap having an O-ring seal and locking arm mechanism. After the sample is collected in the En Core sampler, the
coring body is sealed with the slide-on, locking cap and immediately becomes a sample storage chamber. The
seals of the device are provided by three Tefon™-coated Viton™ O-rings. There is an American Society for Testing
and Materials (ASTM) practice, Standard Practice for Using the Disposable En Core Sampler for Sampling and
Storing Soil for Volatile Organic Analysis, that gives guidance on using the device and includes an appendix
showing data on the performance of the sampler to store VOC-spiked soils1.
The En Core sampler has undergone extensive testing during development to determine design specifications,
and after development to evaluate performance of the device under various storage conditions. This paper
discusses (1) testing that was performed as part of the developmental work to select an O-ring design to minimize
VOC loss from the device; (2) testing that was conducted to generate performance data on the device for inclusion
in the ASTM practice for using the sampler; and (3) testing to evaluate performance of the device to store soils
containing low level VOC concentrations.
DISCUSSION OF THE WORK
Performance of Teflon-Coated Viton O-Rings in the En Core Device to Minimize TCE Loss
Testing was performed by Alan Hewitt at the U.S. Army Cold Regions Research and Engineering Laboratory
(CRREL) in Hanover, NH to evaluate the performance of various O-ring designs based on their ability to minimize
loss of trichloroethylene (TCE) from soil samples stored in the 5-gram En Core device2. In this study, silty, clay-
type soil samples were collected from a site contaminated with TCE. Five soil samples were collected and immedi-
ately transferred to methanol for extraction and analysis. These were the time-zero samples. Samples from the
same site were also collected using En Core samplers having various O-ring designs for 48-hour and seven-day
129
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
storage at 4 ±2°C prior to analysis. For each O-ring design and storage condition, five samples were collected and
stored. After storage for the specified length of time, the samples were extruded into methanol and analyzed by
automated headspace analysis using an auto sampler coupled to a gas chromatograph with sequential photoioni-
zation flame ionization detectors.
To evaluate O-ring performance, the mean concentrations of TCE in the samples stored for 48 hours and seven
days at 4 ±2°C were compared to the mean concentrations of TCE in the time-zero samples. This was done by
performing a statistical calculation to determine if there is a difference between the mean values at a 95% confi-
dence level. In this evaluation, the actual difference between the time-zero mean concentration of TCE and the
mean concentration of TCE in the stored samples (experimental difference) is calculated and compared to a
computed difference3. If the experimental difference between the mean values is less than the computed threshold
difference, it can be concluded that the mean concentration of TCE in the stored samples is statistically the same
as the mean concentration of TCE in the time-zero samples, and as a result, the performance of the En Core
device is acceptable.
The concentrations of TCE in the time-zero samples and samples stored in the En Core devices having Teflon-
coated Viton O-rings, which is the O-ring design selected for the En Core device, are shown in Table 1. The
reason the TCE concentrations vary between storage times is because the samples were collected at different
depths. The experimental difference and computed difference for the mean values for 48-hour and seven-day
storage are also shown in Table 1. For both storage conditions, the mean concentration of TCE in the stored
samples is not statistically different from the mean concentration of TCE in the time-zero samples. This shows
excellent performance of the Teflon-coated Viton O-rings to minimize loss of TCE from soil collected and stored in
the En Core device. Before the O-ring design was finalized, additional testing involving a full list of VOCs was
performed to ensure that Teflon-coated Viton O-rings give acceptable performance for use in the En Core device.
Table 1. Evaluation of the Performance of Teflon-Coated Viton O-Rings in the En Core Device to Minimize TCE
Loss
Teflon-Coated Viton O-Rings Used in En Core Device for 48-Hour Storage at 4 ±2°C
TCE Concentrations in TCE Concentrations in
Time-Zero Samples, mg/Kg En Core Samples. mg/Kg
295 295
222 222
280 280
266 266
254 Experimental Ti - T2 = 2 254.
~x~i : 263 Computed TI-1^2 = 4 ~x~2 : 24
sa: 28 20<41 s: 38
Mean values are not statistically different
Teflon-Coated Viton O-Ring Used in En Core Device for 7-Day Storage at 4 ±2°C
TCE Concentrations in TCE Concentrations in
Time-Zero Samples. mg/Kg En Core Samples. mg/Kg
400 314
352 352
434 313
364 359
351 Experimental Ti - ~x~2 = 4 340
~x~t : 38 Computed T-i - ~*2 = 5 "x~2 : 33
s: 36 44<52 s: 28
Mean values are not statistically different
3 s = standard deviation
130
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Testing to Evaluate the Performance of the En Core Sampler for the ASTM Practice
A study was conducted to evaluate the performance of the 5-gram and 25-gram En Core samplers to store three
different soil types spiked with an aqueous solution containing nine VOCs. The soils used in the study are repre-
sentative of different environments and contained native microbial populations. They were (1) a river bank soil
having 49% sand, 26% silt, 24% clay, 5.3% organic material, -14% moisture, and a dehydrogenase (microbial)
activity of 22 mg total product formed (TPF)/g/24 hours; (2) a mountain soil having 75% sand, 13% silt, 12% clay,
4.3% organic material, -12% moisture, and a dehydrogenase activity of 11 mg TPF/g/24 hours; and (3) a prairie
soil having 67% sand, 17% silt, 16% clay, 1.5% organic material, -8% moisture, and a dehydrogenase activity of
17 mg TPF/g/24 hours. The VOCs used in the study were cis-dichloroethylene (CDCE), benzene, TCE, toluene,
perchloroethylene (PCE), ethylbenzene, m/p-xylene, o-xylene, and methylethylketone (MEK). These compounds
were selected as the analytes of interest because they are representative of halogenated, aromatic, and polar
VOCs typically found in contaminated soils.
In the study, soil samples were collected in the En Core samplers from a large container of loose soil and then
spiked with an aqueous solution containing the compounds listed above to give an approximate concentration of
2.5 ug/g of each analyte of interest in the samples. This analyte concentration in the soil was selected to limit the
influence of the analytical method on the data. After all samples were spiked and capped, five random samples for
each soil type were extruded from each size of En Core sampler into methanol for analysis to give time-zero
concentrations of the analytes of interest. Five each of the remaining samples were stored under the following
storage conditions: on ice at 4 ±2°C for 48 hours; 4 ±2°C for four days (on ice for 48 hours then refrigerated for 48
hours); on ice at 4 ±2°C for 48 hours followed by storage for 5 days in a freezer at -12 ±2°C; and in a freezer at -12
±2°C for seven days. After the samples were held for the appropriate times, they were extruded into methanol for
extraction and analysis. The methanol extracts of the samples were analyzed using EPA Method 8021B"
To evaluate the data, the mean values of the analytes of interest in the stored samples were compared to their
mean values in the time-zero samples by calculating average percent recovery. The average percent recoveries of
the VOCs of interest from samples of the river bank, mountain, and prairie soils stored in the 5-gram and 25-gram
En Core samplers for 48 hours at 4 ±2°C are shown in Table 2. These values range from 69 to 102%. The overall
average percent recoveries for the analytes of interest from the samples of the three soils stored at 4 ±2°C for 48
hours range from 83 to 98%. The overall average percent recoveries of the nine VOCs of interest from samples of
the three soil types stored under the other storage conditions used in the study are shown in Table 3. These
values range from 90 to 98% for the river bank soil, 77 to 91% for the mountain soil, and 55 to 79% for the prairie
soil. The data generated in this study are included in the appendix of the ASTM practice.
Table 2. Average Percent Recoveries of VOCs from Soil Samples Stored in En Core Samplers for 48 Mrs. at 4
±2°C
VOCs
CDCE
Benzene
TCE
Toluene
PCE
Ethylbenzene
m\p-Xylene
o-Xylene
MEK
River Bank Soil
91a(15)b/91c(1)b
93 (3) / 90 (4)
97 (1)/ 92 (3)
99(1)794(3)
100(1)796(3)
101 (3)798(3)
102(2)798(1)
99(1)798(3)
100(0)796(1)
Mountain Soil
87a(10)b/87c(8)b
86(11)790(11)
91 (8) 7 94 (6)
90 (5) 7 93 (6)
96 (4) 7 98 (6)
92 (7) 7 96 (4)
92 (2) 7 93 (4)
97 (2) 7 96 (4)
83 (0) 7 86 (2)
Prairie Soil
82a(9)b/76c (17)b
75(13)769(25)
79(10)772(22)
82(8)779(14)
91 (5)786(14)
91 (3) 7 89 (6)
90(1)788(6)
92 (2) 7 94 (4)
92 (0) 7 97 (3)
Overall Average % 98 (4)" 7 95 (3)d 90 (5)d 7 93 (4)d 86 (8)d/83 (12)d
Recovery
a Average percent recovery for the 5-gram sampler
b Value in parentheses is the percent relative standard deviation of the concentration values in the stored samples.
The percent relative standard deviation of the concentration values in the time-zero samples ranged from 0-11%.
c Average percent recovery for the 25-gram sampler
" Value in parentheses is the percent relative standard deviation of percent recovery values use to calculate overall
average percent recovery.
131
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 3. Overall Average Percent Recoveries of the Nine VOCs of Interest from Soil Samples Stored in En Core
Samplers .
Storage Condition River Bank Soil Mountain Soil Prairie Soil
4 ±2"C for 4 Days 98a (2)b/97c (3)b 83a (8)b/88c(6)b 71a(25)b/71c (21 )b
4±2°Cfor48Hrs. 94 (5) / 90 (7) 83 (10) / 77 (18) 59 (33) / 55 (43)
then-12±2°Cfor
5 Days
-12 ±2°C for 5 Days 98 (4) / 97 (6) 91 (11)/86(13) 76(29)779(21)
a Overall average percent recovery for the 5-gram sampler
b Value in parentheses is the percent relative standard deviation of percent recovery values used to calculate
overall average percent recovery.
c Overall average percent recovery for the 25-gram sampler
Based on the results of this testing, the following conclusions can be made.
• For storage at 4 ±2°C on ice for 48 hours, the En Core sampler performs well for storing the VOC-spiked
river bank, mountain, and prairie soils. Overall average percent recovery values for the analytes of interest
range from 83 to 98% (Table 2).
• The En Core sampler performs well for storing the VOC-spiked river bank and mountain soils under all of
the storage conditions used in the study. Overall average percent recovery values range from 77 to 98%
for the analytes of interest (Table 3).
• Slightly higher percent recovery values for the VOCs of interest from the river bank soil as compared to
those for the mountain soil (Tables 2 and 3) are most likely due to the difference in the composition of the
soils. The higher percent sand and lower percent clay composition of the mountain soil is a less favorable
soil matrix for holding VOCs.
• Percent recovery values for the spiked prairie soil stored in the En Core samplers are generally lower than
those for the river bank and mountain soils. It appears that in some cases, loose particles of the drier
prairie soil may have scattered when the En Core samplers were capped causing the seals of the device
to be compromised. This may be a result of the experimental design of the testing, in which the soil was
loose and had no structure when it was collected in the device.
Testing to Evaluate Performance of the En Core Device to Store Soils Containing Low Level VOC Concentrations
Two soils were used in this study, which was performed by En Chem, Inc. One soil was predominantly sand (83%
sand, 17% silt and clay, and 7% moisture), and the other was a biologically active loamy garden soil (63% sand,
24% silt, 13% clay, and 12% moisture). Five-gram En Core samplers were filled with soil and frozen. After freez-
ing, the samplers were removed from the freezer and allowed to thaw just enough to allow the spiking solution to
be injected. The spiking sblution was prepared by saturating deionized water with gasoline and then injecting a
VOC stock standard into the gasoline-saturated water. Five replicates of time-zero samples (spiked and immedi-
ately extruded into methanol) and five replicates of spiked samples for storage at 4 ±2°C for two and seven days
were prepared for each soil type. Low level samples are defined by EPA Method 50355 as having concentrations
between 1-200 ug/Kg. |n this study, all compounds in the soil samples were within this range, except methylene
chloride (-210 ug/Kg), benzene (-300 ug/Kg), and toluene (-360 ug/Kg). Benzene and toluene concentrations
were high because of their presence in the gasoline-saturated water that was used in the spiking solution.
Average percent recovery values for the VOCs of interest in this study for two-day and seven-day storage are
listed in Table 4. For the sandy soil at two-day storage, recovery values range from 82% for vinyl chloride to 94%
for bromoform. At seven-day storage, the values range from 66% for vinyl chloride and bromoform to 155% for
naphthalene. The lower percent recovery of vinyl chloride is expected because of its low boiling point. The high
recovery for naphthalene at seven-day storage is most likely due to analytical factors.
As shown in Table 4, recovery values for the garden soil at two-day storage, range from 76% for benzene to 101%
for chloroform. At seven-day storage, the values range from 46% for vinyl chloride to 116% for methylene chloride.
Lower percent recovery values for the aromatic compounds, such as benzene and toluene, at seven-day storage
are most likely due to biodegradation as chlorinated compounds with similar volatility do not show this loss at
seven days. In the near future, data showing the performance of the device to store soil containing low levels of
VOCs will be included in the ASTM practice.
132
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 4. Average Percent Recovery Values from Soils Spiked with Low Level VOC Concentrations and Stored in
En Core Samplers for 2 Days and 7 Days at 4 ±2°C
Sandy Soil
Garden Soil
Compound
Average %
Recovery
for 2-Day Storage
Average %
Recovery
for 7-Day Storage
Average %
Recovery
for 2-Day Storage
Average %
Recovery
for 7-Day Storage
Vinyl Chloride
Methylene Chloride
Methyl tert-butyl Ether
Chloroform
Carbon Tetrachloride
Bromodichloromethane
Benzene
1 , 1 ,2-Trichloroethane
Ethyl Dibromide
Toluene
Ethyl benzene
Styrene
Bromoform
m/p-Xylene
o-Xylene
1 ,3,5-Trimethylbenzene
1 ,2,4-Trimethylbenzene
Napthalene
Overall Average % Recovery
82 (7)a
89(3)
92(5)
91(3)
90(5)
88(4)
87(3)
90(8)
90(5)
90(2)
87(4)
83(4)
94(8)
90(2)
88(1)
90(5)
89(3)
86(8)
89 (3)b
66 (4)a
137(6)
91(3)
76(7)
76(9)
76(13)
73(2)
93(11)
97(9)
74(5)
71(4)
67(6)
66 (20)
77(1)
76(6)
102(5)
112(3)
155(10)
88 (28)b
82 (9)a
92(12)
98(5)
101 (8)
96(9)
98(3)
76(6)
95(3)
92(6)
86(12)
95(5)
86(7)
83(21)
96(6)
91(4)
98(4)
99(5)
96(10)
92 (7)b
46(13)a
116(9)
90(10)
97(14)
111(11)
106(13)
59(16)
107(11)
80 (14)
65(17)
93 (14)
65(13)
91(6)
89(12)
86(11)
106(11)
102(9)
108(8)
90 (22)b
aValue in parentheses is the percent relative standard deviation of the concentration values in the stored samples.
The percent relative standard deviation of the concentration values in the time-zero samples ranged from 3-15%.
"Value in parentheses is the percent relative standard deviation of the percent recovery values used to calculate
overall average percent recovery.
SUMMARY OF RESULTS
Results of the testing described in this paper can be summarized as follows.
• The Teflon-coated Viton O-rings show excellent performance in minimizing loss of TCE from soil collected
and stored in the En Core sampler at 4 ±2°C for 48 hours and seven days.
• For storage at 4 ±2°C on ice for 48 hours, the En Core sampler performs well for storing VOC-spiked river
bank, mountain, and prairie soils. Overall average percent recovery values for typical VOC contaminants
in soil range from 83 to 98%.
• The En Core sampler performs well for storing the VOC-spiked river bank and mountain soils under
storage conditions of 4 ±2°C for four days; 4 ±2°C for 48 hours followed by storage for 5 days at -12 ±2°C;
and -12 ±2°C for seven days. Overall percent recovery values range from 77 to 98% for typical VOC
contaminants in soil.
• Results show that when drier, loose soils are to be stored in the En Core samplers, care should be taken
to prevent scattering of particles during capping, which may compromise the seal of the device.
• The En Core sampler performs well for storing sandy and biologically active loamy garden soils containing
low levels of 18 VOCs. Overall average percent recovery values for two-day and seven-day storage at 4
±2°C range from 88 to 92%.
ACKNOWLEDGMENTS
Funding for this work was provided by the U.S. Department of Energy, Federal Energy Technology Center, under
cooperative agreement DE-FC26-98FT40323 and by En Chem, Inc., Green Bay, Wisconsin. The authors also
133
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
wish to acknowledge Alan D. Hewitt for the O-ring evaluation work and En Chem, Inc. for providing low level
performance data.
DISCLAIMER
This paper was prepared as an account of work sponsored by an agency of the United States Government.
Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty,
express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of
any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately
owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark,
manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring
by the United States Government or any agency thereof. The views and opinions of authors expressed herein do
not necessarily state or reflect those of the United States Government or any agency thereof.
REFERENCES
1. American Society for Testing and Materials, 1999, Standard Practice for Using the Disposable En Core
Sampler for Sampling and Storing Soil for Volatile Organic Analysis. In press.
2. Sorini, S.S., and J.F. Schabron, 1998, Validation of a New Soil VOC Sampler: Soil Sampling and Storage
Practices for Volatile Organic Analysis. Laramie, WY, WRI Report WRI-98-R007
3. Skoog, D.A., and D.M. West, 1976, Fundamentals of Analytical Chemistry, Holt, Rinehart and Winston: New
York, NY, pp. 67-69.
4. U.S. EPA, 1996, Method 8021B: Aromatic and Halogenated Volatiles by Gas Chromatography Using
Photoionization and/or Electrolytic Conductivity Detectors. Test Methods for Evaluating Solid Waste:
Physical/Chemical Methods (SW-846), Vol. 1B, Final Update III.
5. U.S. EPA, 1996, Method 5035: Closed-System Purge-and-Trap and Extraction for Volatile Organics in Soil
and Waste Samples. Test Methods for Evaluating Solid Waste: Physical/Chemical Methods (SW-846), Vol.
1B, Proposed Update III.
AN EASY, COST-EFFECTIVE SOLUTION FOR SAMPLING VOLATILE ORGANIC COMPOUNDS IN SOILS
Michael J. Ricker
U.S. Analytical Laboratory, 1090 Kennedy Avenue, Kimberly, Wl 54136, (920) 735-8295
Mike.Ricker@USOIL.com
ABSTRACT
This study evaluates a standard 40-ml vial as a suitable storage container for volatiles when samples are stored at
4±2°C. This method is consistent with SW-846 method 5035, which calls for a "Hermetically-sealed sample vial,
the seal of which is never broken from the time of sampling to the time of analysis". Method 5035 further
discusses field preservatives as a means of retarding biological degradation. This study also demonstrates that
the addition of chemical preservatives such as methanol or acid can be delayed for up to 48 hours without signifi-
cant losses due to biological degradation. Samples are taken utilizing a coring tool to rapidly delivering an approxi-
mate 5-gram sample to a 40-ml vial pretared with stirring bar. Sample preservation and weight determinations are
performed at the laboratory.
Introduction
Soil sampling options for volatiles following EPA's Method 5035 or ASTM's D4547-98 can be both time-consuming
and expensive. At the time of these recommendations, sample takers typically haul field balances and chemical
preservatives to the field or use expensive coring/storing devices which typically add $25-$30 to the cost of each
sample. New study data demonstrates that an empty 40-ml vial can be used as a sample storage container for at
least 10 days without any significant loss in VOC recoveries due to volatilization. Recoveries of 85 % or more were
retained for 55 of 63 analytes tested after a ten-day period when samples were stored at 4°C without preservative.
A 10-ul methanol standard was used to spike 13 grams of soil with the 63 volatile analytes. Subsequent studies by
Alan Hewitt - U.S. Army Cold Regions Research and Engineering Laboratory and U.S. Analytical suggest that the
10 ul of methanol were enough to sterilize the vial which minimized losses due to biological degradation.
134
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
The effects of biological degradation have been studied on several matrixes by both U.S. Analytical - presented
here and Alan Hewitt - U.S. Army Cold Regions Research and Engineering Laboratory2'3 Results have shown that
a 48-hour transportation period where samples were stored at 4±2°C without preservative can be incorporated
with at least 73% volatile retention for most soil matrixes.
The focus of this paper is to present data which supports the use of an empty 40-ml VOC vial as a suitable
transport/storage container for volatiles and to suggest a sampling protocol, which is less expensive, and more
user friendly than the most widely practiced options.
Experimental
Two separate studies were performed to determine the appropriate holding time for soils collected in empty VOC
vials. The first study made use of a methanol spiking solution, which is believed to have sterilized the soil, and
thus, prevented biological degradation. The significance of this data is that it can be used to show the integrity of a
40-ml vial as a storage container and displays the efficiency of recapturing and extracting volatiles from both dry
Ottawa sand and a complex matrix such as garden soil by adding methanol through the vial septum. The second
study made use of an aqueous spiking solution of low-level BTEX compounds so as not to sterilize the soil.
Results from study two could then be used to determine the level of biological degradation over time in order to
determine how long samples can be held without preservation.
For study one, a list of 63 volatiles was chosen that would demonstrate broad range applicability for this proce-
dure. In addition, gasoline was chosen to provide information on a common contamination mixture and to validate
the procedure for Wisconsin's GRO method. VOC's were spiked at a low (154 ug/kg) and a high (769 ug/kg) level.
The low-level spike would provide information on possible matrix interference and the high level would more
accurately determine procedural integrity. The gasoline was spiked at 153 mg/kg. The procedure was tested on
both dry Ottawa laboratory sand and biologically active garden soil. The six methanol addition time intervals were
0, 24, 48, 72, 168, and 240 hours. Five replicates were analyzed at each time interval to provide the necessary
statistical data. The steps of the procedure are outlined below:
1. Weigh 13 grams of soil matrix into 92- 40ml VOC vials, 30 for VOC low-level spike, 30 for VOC high-level
spike, and 30 for gasoline spike, and 2 for blanks.
2. Add 13 mis of methanol to the blanks and cap. Store with all other samples until ready for analysis.
3. Spike 1 replicate for the VOC low-level 0-hour time interval and immediately cap the vial. Add 13 mis of
methanol through the septum using a 25-ml Gaslight™ syringe equipped with a 22-gauge needle. Holding
the plunger down, pull the syringe out, shake the vial for 15 seconds, and vent. Record the time.
4. Repeat step 3 for the remaining 4 VOC low-level 0 hour replicates.
5. Spike the rest of the VOC low-level time intervals, record the time and store at 4±2°C.
6. Repeat steps 3-5 for both the high level and the gasoline middle level spikes.
7. After 24 hours add methanol through the septum to each of the five 24 hour replicates for all three spiking
schemes. Repeat this for all other time intervals.
8. Sonicate extract and analyze all VOCs using SW846 method 8260. Sonicate extract and analyze all
gasoline spikes using the WDNR GRO September 1995 method.
Samples were spiked as follows:
VOC low-level (154-ug/kg) spike 10 ul of a 200-mg/l solution below the soil surface.
VOC High-level (769 ug/kg) spike 50 ul of a 200-mg/l solution below the soil surface.
Gasoline (154 mg/kg) spike 20 ul of a 100,000 mg/l solution below the soil surface.
For Study two, the list of analytes were the BTEX compounds, benzene, toluene, ethylbenzene, o-xylene,
m-xylene, and p-xylene. This list of analytes was chosen since the aromatic compounds have been shown to be
more susceptible to biological degradation than other VOCs2 This study was designed to compare biological
degradation in three different soil matrixes. Dry Ottawa sand was used as a baseline to test the experimental
procedure. Biologically active garden soil and soil taken from a UST site mildly contaminated with fuel oil were the
test soils. After adding 5 g. of soil to an empty 40-ml VOC vial, 30 ul of an approximately 80-mg/l aqueous solution
of each analyte were added to the soil subsurface. The resulting sample concentration was less than 400 ug/kg for
each analyte. Three different soil types were spiked in duplicate and held from 0-5 days without preservative.
Samples were preserved with methanol through the septum using the same procedure as study 1.
Instrumentation
Volatiles. SW846-Method 8260
Hewlett Packard 6890 GC, Hewlett Packard 5973 MSD, Tekmar 3000 Sample Concentrator, Dynatech PTA
135
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
SOWS Autosampler.
WDNR GRQ September 1995
Varian 3400 GC, Tracer PID/FID Detectors, Tekmar LSC 2000 Sample Concentrator, Archon 5100
Autosampler.
Results
Study one results are compiled in Tables 1-25. Study two results are compiled in Table 26.
Discussion
Study one clearly demonstrates that in the absence of biological degradation a Teflon®-lined Silicone Septa 0.125"
40ml VOC vial is an excellent storage container for volatiles. For the Ottawa sand matrix, 58 of 63 analytes recov-
ered at 85% or better at the low-level spike after a 10-day holding period and 59 of 63 analytes recovered at 85%
or better at the high-level spike after a 10-day holding period. For the biologically active garden soil, (which proba-
bly was partially sterilized by methanol from the spiking solution) 42 of 63 recovered at 85% or better at the
low-level spike after a 10-day holding period and 55 of 63 analytes recovered at 85% or better at the high-level
spike after a 10-day holding period. Gasoline was recovered at 94% or higher in both Ottawa sand and biologically
active garden soil after a 10 day holding period.
High-level biologically active soil recoveries were significantly higher than low-level biologically active soil recover-
ies. The difference in these recoveries can be attributed to the degree of biological sterilization, which occurred in
the study. The amount of methanol used to spike the low-level concentrations was 10 pi, while 50 pi was used for
the high-level. Also, biological degradation rates are probably concentration dependent with low concentrations
having a shorter half-life than high.
Study two results (Table 28) indicate that soil samples can be held for 48 hours at 4±2°C without a biological
preservative. 73% or more of the analyte concentrations were retained after a two-day storage period without
preservation. These findings corroborate other recent findings by Alan Hewitt of the U.S. Army Cold Regions
Research and Engineering Laboratory. Recoveries of actual subsurface samples may be underestimated, since
the soils used in this study were allowed to come in contact with oxygen for several weeks before being used.
Studies using freshly exhumed soils should be conducted to determine if there is a necessary acclimation period
for the soil bacteria.
Summary with suggested sampling protocol
Over the last few years there has been a lot of discussion regarding the implementation of SW846, Method 5035
for volatiles. Some laboratories and government agencies have been reluctant to implement since; there is both
increased costs and new complexities in this procedure Vs the predecessor 5030. Laboratories must invest in new
autosamplers which can cost between $20,000-$30,000 each. There are new problems in the field as well, sample
takers either use a coring device to collect samples and than either transfer the sample to a 40 ml vial which
contains a chemical preservative or they transport the sample in this coring device and than it is up to the labora-
tory to transfer the sample from the coring device to a 40 ml vial. The first option requires hazardous chemicals to
be taken to the field, which is questionable under DOT regulations. The second option of using a coring/transport
device typically adds $25-$30 to the cost of each sample. Using an empty VOC vial and an inexpensive coring tool
addresses both DOT and cost concerns.
EPA method 5035 was developed to improve volatile data by reducing analyte losses caused by both volatilization
and biological degradation. Study one has shown that those analyte losses due to volatilization can virtually be
eliminated by using an empty VOC vial as a storage container. Study two indicates that losses due to biological
degradation are less significant if samples are preserved within 48 hours. This means preservation can be delayed
until back at the laboratory in an environment more conducive to precise measurement. Using the empty VOC vial
procedure simplifies a rather complex method and could help facilitate the implementation of method 5035.
If the empty VOC vial option is used, here is U.S. Analytical^ recommended sampling protocol. For each sampling
point use three 40-ml vials, pre-tare two with stirring bars and one without. Collect 5 g. samples using the
Easydraw™ syringe and Powerstop™ handle or equivalent and immediately transfer to the VOC vial. Cap the vials
taking care not to get dirt on the threads. Place on ice and ship to the laboratory within 48 hours. Based on studies
by Alan Hewitt, it is recommended that the laboratory immediately freeze the two samples with stirring bars and
add 5 mis of methanol through the septum of the third. The laboratory has the choice of running either the metha-
nol preserved sample first for medium to high-level contamination or risk gross instrument contamination and run
a low-level sample. In either case the methanol-preserved sample is necessary for high level contamination since
there is no way to dilute vials collected for low level analysis.
136
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
References:
1. U.S. Environmental Protection Agency (1986) Test Methods for Evaluating Solid Waste, Vol. 1B. SW-846.
2. Hewitt, A.D. (8/97) Biodegradation of Volatile Organic Compounds in Soil Samples. American Environmental
Laboratory, Vol. 9, Number 7 Pages 1, 5-7.
3. Hewitt, A.D. Freezer Storage of Soil Samples Containing Volatile Organic Compounds. 15th Annual Waste
Testing and Quality Assurance Symposium, July 18-22, 1999.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
VOC STUDY RESULTS
Matrix - Clean Laboratory Sand
VOC METHOD 8260 STANDARD 153.8 UG/KG / 3.08 UG/L
Rep, frl Rep. #2 Rep. #3 Rep. #4 Rep. #5
Table 1
ANALVTE
1,1,2-Trichtoroethane
1 ,2,4-Trimethylbenzene
1 ,3.5-TrirrBthylbenzene
Benzene
Bromodichloromethane
Bromoform
Carton Tetrachloride
Chloroform
cis - 1,3 - Dichloropropene
EDB (1,2-Dibromoethane)
Bhylbenzene
Methylene chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1.3 - Dichloropropene
Vinyl Chloride
1,1 - Dichloropropene
1,1-Dichloroethane
1,1-Dichloroethene
1,1,1-Trichloroethane
1,1,1,2 - Tetrachloroethane
1 , 1 ,2,2-Tetrachloroethane
1 ,2-DibrorrD-3-Chloropropane
1 ,2-Dichtorobenzene
1,2-Dichloroethane
1 ,2-Dichloropropane
1,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dtehlorobenzene
1 ,3-Dichloropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Allyl Chloride
Brorrobenzene
Bromochlororrethane
Chlorobenzene
Chloroethane
Chloromethane
cis-1,2-Dichloroethene
Dibrorrochlororrethane
Dibromorrethane
Dichlorofluoromethane
Di-bopropyl ether
Bhyl Bher
Hexachlorobutadiene
bopropylbenzene
n-Butylbenzene
n-Ropylbenzene
p-bopropyltoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1,2-Dichloroethene
Trichloroethene
Trichlorofluoromethane
Trichlorotrifluoroethane
0 Hour
Absol
146
156
162
156
161
149
186
181
164
148
158
157
166
320
131
159
149
158
168
135
187
165
178
200
152
141
138
155
178
141
127
137
137
149
149
142
154
192
162
156
158
169
154
190
132
162
146
16B
214
150
172
14/
167
138
157
156
152
166
184
177
149
19/
154
Rflc. %
95
101
105
101
105
97
121
117
107
96
102
102
108
104
85
103
97
103
109
88
122
107
116
130
99
92
90
100
115
92
83
89
89
97
97
92
100
125
105
101
103
110
100
124
86
105
94
109
139
98
112
96
109
yo
102
101
99
108
120
115
y/
128
100
0 Hour
Absol
141
153
150
154
160
151
183
168
149
157
154
162
167
307
127
154
149
149
168
138
176
166
169
191
149
150
138
145
182
134
147
140
135
149
148
142
148
184
153
166
14B
190
149
179
119
151
14/
1b/
205
148
1/8
161
157
134
153
147
154
159
1(55
157
144
209
164
Rec %
92
99
98
100
104
98
119
109
97
102
100
105
109
100
83
100
97
97
109
90
114
108
110
124
97
98
89
94
118
87
96
91
88
97
96
92
96
120
99
108
96
124
97
116
77
98
95
102
133
96
116
104
102
Ut
99
95
100
103
107
102
93
136
107
0 Hour
Absol
151
145
144
147
150
129
176
167
149
140
151
144
153
301
116
145
144
150
158
124
154
151
152
178
143
133
129
146
172
132
142
114
116
145
148
139
138
183
1b1
1bb
142
162
W
180
115
14/
148
16U
223
143
167
VU
148
136
147
146
143
151
1?6
155
144
185
142
Rec. %
98
94
93
95
98
84
114
108
97
91
98
93
99
98
75
94
94
97
103
81
100
98
99
116
93
86
84
95
112
86
92
74
75
94
96
90
90
119
98
101
92
105
95
117
75
95
96
104
145
93
109
110
96
88
96
95
93
98
114
100
94
120
92
0 Hour
Absol
144
145
148
148
151
143
173
171
155
145
150
154
154
299
126
150
146
149
158
128
167
157
158
183
156
147
159
150
171
133
141
126
128
142
146
145
146
176
1bb
152
142
1bb
1b2
176
115
160
139
161
199
142
161
1b8
153
133
147
145
143
148
M^
143
151
189
161
Roc. %
94
94
96
96
98
93
112
111
101
94
98
100
100
97
82
98
95
97
103
83
109
102
103
119
101
95
103
98
111
86
92
82
83
92
95
94
95
114
100
99
92
101
99
114
74
104
90
105
129
92
105
102
99
86
96
94
93
96
111
93
98
123
105
0 Hour
Absol
141
143
150
149
148
136
172
170
138
143
147
164
153
294
116
157
146
153
157
141
153
149
161
185
150
131
116
137
174
128
141
123
108
144
145
137
143
188
147
155
142
1b8
146
189
119
160
148
154
207
141
173
140
152
118
142
142
137
145
160
152
157
18U
142
R8C. %
91
93
98
97
96
88
112
111
90
93
95
107
99
95
75
102
95
99
102
92
99
97
104
120
98
85
75
89
113
83
92
80
70
93
94
89
93
122
95
101
92
103
9b
123
77
104
96
100
135
92
112
91
99
77
92
92
89
94
104
99
102
117
92
Absolute
AVG.
144
164
168
151
154
142
178
171
151
146
152
156
159
304
123
153
147
152
162
133
167
157
163
187
150
140
136
147
175
133
140
128
125
146
147
141
146
185
153
157
146
167
150
183
120
156
145
160
210
145
170
155
155
132
149
147
146
154
171
156
149
192
153
Average
% REC.
93.8
106.6
109.2
97.9
99.9
92.1
115.5
111.2
98.2
95.2
98.6
101.4
103.1
98.8
80.0
99.3
95.4
98.6
105.2
86.6
108.8
102.3
106.2
121.7
97.5
91.2
88.3
95.3
113.8
86.7
908
83.1
81.0
94.7
95.4
91.4
94.7
120.0
99.5
102.0
95.1
108.5
97.2
118.7
77.8
101.2
94.2
104.0
1 36.3
94.1
110.7
100.7
101.0
856
96.9
95.5
94.8
99.9
111.2
101.7
96.8
124.6
99.2
%
RSD.
2.8
3.4
4.0
2.5
4.0
6.5
3.4
3.2
6.3
4.6
2.8
5.1
4.5
3.3
5.7
3.7
1.5
2.6
3.5
5.3
8.7
49
62
4.5
3.2
6.0
116
4.4
2.6
3.7
54
8.2
10 1
2.1
1.0
2.3
4.0
3.2
3.6
3.4
4.9
8.4
2.3
3.6
5.8
4.4
2.6
3.3
4.4
2.7
3.8
7.6
4.7
6.0
3.8
3.7
4.8
5.6
5.4
7.9
3.7
5.8
6.8
* Denotes analyte required to be added to this study by the WDNR and w hich mist pass the imposed criteria to gain
acceptance for this method in Wisconsin.
137
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS
Matrix - Clean Laboratory Sand
VOC METHOD 8260 STANDARD 153.8 UG/KG / 3.08 UG/L
Table 2
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
ANALYTE
1 ,1 ,2-Trichloroethane
1 ,2,4-trimethylbenzene
1 ,3,5-Trimethylbenzene
Benzene
BromodichloromBthane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1 ,3 - Dichloropropene
SB (1,2-Dibrorroethane)
Hhylbenzene
Methylene chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1,3 - Dichloropropene
Vinyl Chloride
1,1 - Dichloropropene
1,1-Dichloroethane
1,1-Dichloroethene
1,1,1-Trichloroethane
1,1,1,2 - Tetrachloroethane
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dbromo-3-Chloropropane
1 ,2-Dichlorobenzene
1,2-Dichloroethane
1 ,2-Dichloropropane
1,2,3- Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dichlorobenzene
1 ,3-Dichloropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Qchloropropane
4-Chlorotoluene
Allyl Chloride
Brorrobenzene
Brorrochloromethane
Chlorobenzene
Chloroethane
Chlororrethane
cis-1,2-Dichloroethene
Dbromochlororrethane
Dibromomethane
Dichlorofluororrethane
Di-teopropyl ether
Shyl Bher
Hexachlorobutadiene
Is opropy Ibenzene
n-Butylbenzene
n-FVopylbenzene
p- Is opropy Itoluene
sec-Butylbenzene
tert-Buty Ibenzene
Tetrachloroethene
trans-1 ,2-Qchloroethene
Trichloroethene
Trichlorofluoromethane
Trichlorotrifluoroethane
24 Hour
Absol.
143
143
147
14B
157
143
158
173
144
142
149
154
158
299
109
145
146
149
157
124
153
154
144
170
155
120
129
144
167
127
139
120
121
138
141
142
142
156
146
145
146
159
149
182
109
144
147
155
201
139
156
142
148
132
139
139
13b
149
157
136
149
182
143
Rec %
93
93
96
96
102
93
102
112
93
92
97
100
102
97
71
94
95
97
102
80
99
100
93
111
100
78
84
94
108
82
90
78
78
90
92
92
92
101
95
94
95
103
97
118
71
93
96
100
131
90
101
92
96
86
90
90
88
97
102
88
97
118
93
24 Hour
Absol
147
142
14/
148
142
131
169
164
141
1bb
1bO
1b2
163
293
111
1b/
144
146
148
115
146
150
159
18b
148
130
127
144
1/2
124
129
117
116
136
147
130
143
169
141
1b/
141
165
147
178
114
159
14b
142
210
141
163
150
142
122
137
13b
136
1bO
1b2
142
145
19b
1b9
RdC %
96
92
96
9b
92
85
110
10/
92
100
98
99
106
95
72
102
93
95
96
75
95
98
103
120
96
85
83
93
112
81
84
76
75
88
96
84
93
110
91
102
91
107
95
116
74
103
94
92
137
91
106
98
92
79
89
87
88
97
99
92
94
126
103
24 Hour
Absol
144
148
144
13/
143
144
1/9
180
148
142
1b1
153
157
29b
115
150
145
144
146
123
152
150
149
183
146
132
136
141
169
135
142
127
125
140
145
133
141
162
148
145
142
166
147
177
106
151
146
137
213
137
160
150
149
130
142
144
143
149
1bb
142
150
181
147
Rec %
93
96
93
89
93
93
116
117
96
92
98
99
102
96
75
97
94
94
95
80
99
98
97
119
95
86
88
92
110
88
92
82
81
91
94
86
92
105
96
94
92
108
95
115
69
98
95
89
138
89
104
97
97
85
92
93
93
97
101
92
97
118
96
24 Hour
Absol
148
146
146
139
149
127
1/9
163
149
149
146
157
154
281
112
143
143
151
137
127
151
154
152
177
137
124
121
141
178
125
142
115
124
139
140
132
135
151
144
144
135
163
145
170
116
158
144
151
197
139
165
139
144
123
139
137
140
147
151
134
149
191
151
Rec %
96
95
95
90
97
82
116
106
97
97
9b
102
100
91
72
93
93
98
89
83
98
100
99
115
89
80
78
91
116
81
92
74
81
90
91
86
88
98
94
93
88
106
94
111
75
102
93
98
128
90
107
90
94
80
90
89
91
95
98
87
97
124
98
24 Hour
Absol
137
142
143
142
141
134
1/0
161
146
133
145
1b/
159
289
113
148
144
143
153
123
153
145
161
1/4
153
129
131
143
168
123
141
123
118
152
144
145
141
154
143
148
141
159
138
174
105
147
144
137
208
142
163
150
144
123
139
13/
132
143
149
132
143
182
148
Rec %
89
92
93
92
91
8/
110
104
95
86
94
102
103
94
73
96
93
93
99
80
99
94
104
113
99
84
85
93
109
80
92
80
76
99
93
94
91
100
93
96
91
103
89
113
68
96
93
89
135
92
106
97
94
80
90
89
86
93
97
86
93
118
96
Absolute
AVG.
144
144
145
143
146
135
171
168
145
144
148
155
158
291
112
148
144
146
148
122
151
151
153
178
148
127
129
142
171
127
139
120
121
141
143
136
140
158
144
148
141
162
145
176
110
152
145
144
206
139
161
146
145
126
139
138
137
147
153
137
147
186
150
AVG. % REC
of 0 Hour
99.5
97.1
96.5
94.7
95.0
95.6
96.0
98.2
96.2
98.2
97.6
99.1
99.6
95.8
90.7
97.2
98.0
96.5
91.5
91.7
90.1
95.7
93.5
94.9
98.5
90.4
94.7
97.1
97.5
95.0
99.3
93.9
96.7
96.7
97.5
96.7
96.3
85.5
94.2
94.1
96.2
97.2
96.9
964
91.5
97.3
99.9
90.2
98.1
96.3
94.7
94.2
93.5
95.5
93.2
93.9
93.9
95.8
89.2
87.4
98.7
97.1
98.0
%
RSD.
3.0
1.9
1.3
3.6
4.5
5.5
5.2
4.9
2.1
5.8
1.8
1.6
2.0
2.4
2.0
3.7
0.8
2.3
5.0
3.6
1.9
2.4
4.6
3.5
4.7
3.8
4.3
1.1
2.6
3.8
40
4.0
3.4
4.5
2.1
5.0
2.2
4.5
1.9
3.7
2.8
2.1
2.9
2.5
4.4
4.4
1.0
5.6
3.3
1.2
2.1
3.7
2.1
3.9
1.2
2.5
3.2
1.8
2.2
3.4
2.1
3.3
4.0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
acceptance for this method in Wisconsin.
138
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS Table 3
Matrix - Clean Laboratory Sand
VOC METHOD 8260 STANDARD 153.8 UG/KG / 3.08 UG/L
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
1 *
2 *
3 '
4 *
5 *
6 *
7 *
8 *
9 *
10 *
11 *
12 *
13 *
14 *
15 *
16 *
17 *
18 *
19 *
20 *
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
ANALVTE
1 , 1 ,2-Trichloroethane
1 ,2,4-Trirrethylbenzene
1 ,3,5-Trirrethylbenzene
Benzene
Brorrodtehloromathane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1 ,3 - Dichloropropene
EDB (1,2-Dibrorroethane)
Bhylbenzene
Methylene chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1,3 - Dichloropropene
Vinyl Chloride
1,1 - Dichloropropene
1,1-Dichtoroethane
1,1-Dichloroethene
1 ,1 ,1-Trichloroethane
1,1,1,2 - Tetrachloroethane
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dibrom>3-Chloropropane
1 ,2-Dichlorobenzene
1,2-Dtehtoroethane
1 ,2-Dichloropropane
1,2,3 - Tricnloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dichlorobenzene
1 ,3-Dichtoropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Allyl Chloride
Bromobenzene
Bromochloromethane
Chlorobenzene
Chloroethane
Chloromethane
cis-1 ,2-Dtehloroethene
Dibromochloromethane
Dibromorrethane
Dichlorofluoromethane
Di-bopropyl ether
Bhyl Bher
Hexachlorobutadiene
bopropylbenzene
n-Butylbenzene
n-Rx>pylbenzene
p-bopropyttoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1,2-Dichloroethene
Trichloroethene
Trichlorofluoromethane
Trichlorotrifluoroethane
48 Hour
Absol.
129
139
136
132
151
129
160
164
135
158
139
144
144
278
106
134
138
132
128
105
133
148
138
172
139
127
111
133
161
112
124
122
117
134
129
128
133
140
135
131
137
147
136
166
98
142
137
136
190
126
150
143
138
117
137
131
129
136
138
119
142
178
120
Rec.%
84
90
88
86
98
84
104
106
87
103
90
93
94
90
69
87
90
86
83
68
86
96
89
112
90
83
72
86
104
73
81
79
76
87
84
83
86
91
88
85
89
95
88
108
64
92
89
88
123
82
97
93
89
76
89
85
84
88
89
77
92
115
78
48 Hour
Absol.
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
omtte
omtte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
omtte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
Rec.%
i
i
J
6
1
1
d
J
i
J
J
J
J
i
i
J
i
i
J
i
d
1
d
J
J
d
j
d
d
d
d
d
J
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
48 Hour
Absol.
136
140
144
143
143
133
166
158
141
137
148
156
154
290
113
138
147
143
141
119
150
149
159
180
151
130
113
139
171
130
145
123
120
140
143
133
137
146
141
143
136
156
142
179
109
144
145
14/
212
138
151
126
146
128
139
141
138
148
159
137
146
175
126
Roc. %
88
91
93
93
93
86
108
103
91
89
96
101
100
94
73
90
95
93
91
77
98
97
103
117
98
85
73
90
111
85
94
80
78
91
93
86
89
95
92
93
88
101
92
116
71
93
94
96
138
89
98
82
95
83
90
92
90
96
103
89
95
114
82
48 Hour
Absol
174
143
140
147
136
138
174
169
138
137
150
151
155
272
109
143
141
143
149
131
163
151
157
187
156
118
151
137
166
120
138
109
121
137
135
131
132
143
137
141
138
166
143
201
114
1b3
138
147
218
137
165
124
140
121
134
136
135
13/
156
13b
165
212
VI)
Rec.%
113
93
91
96
88
89
113
110
89
89
97
98
101
88
71
93
92
93
97
85
106
98
102
121
101
76
98
89
108
78
90
71
78
89
87
85
86
93
89
91
89
108
93
131
74
99
89
96
142
89
107
80
91
79
87
88
88
89
101
88
107
138
111
48 Hour
Absol.
144
144
140
143
154
154
179
177
147
147
146
168
162
292
119
151
144
150
139
125
165
169
156
195
144
138
130
147
168
126
133
134
127
142
146
135
138
147
146
139
142
180
147
173
113
166
147
149
211
147
172
152
143
127
139
140
134
139
166
158
162
Ml
150
Rec.%
93
93
91
93
100
loo
116
115
95
95
95
109
105
95
77
98
94
97
90
81
107
110
101
126
94
89
84
95
109
82
86
87
82
92
95
88
90
96
95
90
92
117
96
112
73
108
96
97
137
9b
112
99
93
82
90
91
87
90
108
102
105
115
97
Absolute
AVG.
146
141
140
141
146
138
169
167
140
144
145
155
154
283
111
141
142
142
139
120
153
154
152
183
147
128
126
139
166
122
135
122
121
138
138
131
135
144
140
138
138
162
142
179
108
151
142
145
207
137
159
136
142
123
137
137
134
140
154
137
153
185
141
AVG. % REC
of 0 Hour
100.8
95.3
92.8
93.7
94.7
97.5
95.3
97.5
92.5
98.6
95.8
99.1
97.0
930
90.5
92.5
97.0
93.4
85.8
90.1
91.2
97.9
93.2
97.8
98.1
91.3
92.7
94.6
95.0
91.5
96.6
95.4
97.0
94.8
941
93.4
92.5
77.8
91.2
88.1
94.4
97.1
94.8
98.2
90.5
96.9
97.7
90.4
98.9
94.4
93.6
87.9
91.2
93.6
91.8
93.2
91.9
91 0
90.2
87.7
103.0
96.6
92.5
RSD.
13.5
1.7
2.2
4.5
5.6
8.1
4.9
4.9
3.7
7.1
34
6.7
4.8
3.4
5.0
5.0
2.6
5.3
6.4
93
9.8
6.5
6.5
5.3
5.1
6.5
14.7
4.1
2.6
6.4
6.4
8.4
3.4
2.3
5.6
2.4
2.3
2.4
3.3
3.9
2.0
8.7
3.2
8.6
6.8
7.3
3.5
40
6.0
6.1
6.7
10.1
2.6
4.1
1.7
3.4
2.8
3.8
7.7
11.5
7.4
9.5
16.4
* Denotes analyte required to be added to this study by the WDNR and w hich rrust pass the imposed criteria to gain
acceptance for this method in Wisconsin.
139
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS Table 4
Matrix - Clean Laboratory Sand
VOC METHOD 8260 STANDARD 153.8 UG/KG / 3.08 UG/L
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
1 •
2 *
3 *
4 "
5 '
6 *
7 *
8 *
9 *
10 *
11 *
12 *
13 *
14 *
15 *
16 *
17 *
18 *
19 *
20 *
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
ANALYTE
1,1,2-Trichloroethane
1 ,2,4-Trimethy Ibenzene
1,3,5-Trirrethy Ibenzene
Benzene
Bromodichloromethane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1,3 - Dichloropropene
H» (1,2-Dibromoethane)
Bhy Ibenzene
Methylene chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1 ,3 - Dichloropropene
Vinyl Chloride
1,1 -Dichloropropene
1,1-CSchioroethane
1,1-Dichloroethene
1,1,1-Trichloroethane
1,1,1,2- Tetrachloroethane
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dbrom>3-Chloropropane
1 ,2-Dichlorobenzene
1,2-Dichloroethane
1 ,2-Dichloropropane
1,2,3- Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dichlorobenzene
1 ,3-Dichloropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Ally I Chloride
Bromobenzene
Bromochloromethane
Chlorobenzene
Chloroethane
Chloromethane
cis-1 ,2-Dichloroetnene
Dibromochloromethane
Dibromo methane
Dichlorofluoromethane
Di-teopropyl ether
Bhyl Bher
Hexachlorobutadiene
teopropylbenzene
n-Buty Ibenzene
n-Ropy Ibenzene
p-lsopropyltoluene
sec-Butylbenzene
tert-Buty Ibenzene
Tetrachloroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trichlorofluoromethane
Trichlorotrifluoroethane
72 Hour
Absol.
1bU
135
143
139
143
130
170
160
140
138
146
136
159
292
104
142
144
147
139
120
144
148
155
181
147
121
106
140
164
124
135
118
102
138
135
132
133
161
137
138
142
163
146
181
109
148
140
134
192
140
140
136
143
115
136
130
132
142
155
141
147
173
142
Rec.%
98
88
93
90
93
85
110
104
91
90
95
88
103
95
67
92
93
95
90
78
94
96
101
118
95
79
69
91
10/
81
88
77
66
90
8/
86
86
105
89
89
92
106
95
118
71
96
91
87
125
91
91
88
93
75
U8
84
86
92
101
91
95
112
92
72 Hour
Absol.
150
142
145
153
149
126
173
176
142
145
145
145
163
289
124
139
144
145
151
128
151
154
156
188
151
141
122
142
175
143
145
126
126
139
146
132
137
147
141
145
143
163
143
189
117
144
153
149
220
143
171
137
141
118
136
134
133
143
158
134
152
174
143
Roc. %
98
92
94
99
9/
82
112
114
92
94
94
94
106
94
81
90
93
94
98
83
98
100
101
122
98
91
79
92
114
93
94
82
82
90
95
86
89
95
92
94
93
106
93
123
K
94
99
97
143
93
111
89
92
77
88
8/
86
93
102
8/
99
113
93
72 Hour
Absol.
136
136
142
144
140
137
165
157
130
137
141
146
147
283
100
141
139
147
140
12b
164
147
160
176
151
113
116
136
1/3
120
129
123
111
134
127
128
131
145
138
134
135
165
141
185
113
151
133
142
212
133
165
135
145
120
133
134
131
142
14b
139
1b/
19/
154
Roc %
88
88
92
93
91
89
107
102
84
89
91
95
95
92
65
91
90
96
91
81
107
95
104
114
98
73
75
88
112
78
84
80
72
87
83
83
85
94
89
87
88
107
91
120
73
98
86
92
138
86
107
88
94
78
86
87
8b
92
94
90
102
128
100
72 Hour
Absol.
144
141
136
1bO
145
130
181
175
140
138
142
144
165
287
115
147
142
146
145
131
150
161
145
177
146
127
143
138
1/0
126
142
132
126
138
145
132
136
140
14b
144
137
164
147
1/6
118
153
152
158
223
140
161
135
142
122
139
140
138
146
146
149
15/
19/
145
Rec %
94
91
88
97
94
85
117
113
91
90
92
93
107
93
74
95
92
95
94
85
97
104
94
115
95
82
93
89
111
82
92
86
82
90
94
86
88
91
94
94
89
106
95
114
76
99
99
102
145
91
104
88
92
79
90
91
89
95
95
97
102
128
94
72 Hour
Absol.
149
150
145
152
151
136
169
169
13/
153
154
148
162
313
115
149
147
158
150
110
1b8
164
160
177
154
122
111
140
182
139
132
132
121
143
148
143
144
146
14«
151
138
168
147
170
116
161
158
1/6
221
143
170
132
147
126
143
149
140
151
153
151
155
188
150
Rac %
97
98
94
99
98
88
110
110
89
99
100
96
105
102
74
97
96
102
97
72
102
106
104
115
100
79
72
91
118
90
86
86
79
93
96
93
93
95
96
98
90
109
95
110
75
104
102
114
143
93
110
86
96
82
93
9/
91
98
99
98
100
122
98
Absolute
AVG.
146
141
142
147
146
132
171
167
138
142
145
144
159
293
111
143
143
148
145
123
153
154
155
180
150
124
120
139
173
130
136
126
117
138
140
133
136
148
142
142
139
164
145
180
114
151
147
152
213
140
161
135
144
120
137
137
135
144
151
143
153
186
147
AVG. % REC
of 0 Hour
101.0
95.1
94.4
97.8
94.7
92.9
96.4
97.7
91.1
97.1
95.7
92.2
100.3
96.3
90.4
93.8
97.3
97.8
89.5
92.0
91.5
98.2
95.0
96.0
99.7
88.7
88.0
94.9
98.7
97.7
97.7
98.6
94.0
94.9
95.4
94.7
93.3
79.8
92.4
90.6
95.1
98.5
96.7
98.6
95.5
97.0
101.4
94.7
101.8
96.5
945
87.1
92.4
91.2
92.1
93.5
92.4
93.9
88.5
91.1
103.0
96.9
96.1
%
RSD.
4.1
4.2
2.6
4.1
3.0
3.3
3.4
5.2
3.5
4.8
3.5
3.2
4.6
4.0
8.6
3.0
2.1
3.5
3.8
6.7
5.0
4.9
4.0
2.8
2.2
8.3
12.0
1.7
3.8
7.6
4.9
4.7
8.9
2.2
6.4
4.1
3.6
5.4
3.4
4.7
2.4
1.3
1.8
4.2
3.2
4.1
6.9
10.6
6.0
2.9
7.8
1.5
1.6
3.3
2.7
5.5
2.9
2.7
3.7
5.0
2.9
6.3
3.3
* Denotes analyte required to be added to this study by the WDNR and w hich must pass the imposed criteria to gain
acceptance for this method in Wisconsin.
140
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS Table 5
Matrix - Clean Laboratory Sand
VOC METHOD 8260 STANDARD 153.8 UG/KG / 3.08 UG/L
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
1 *
2 *
3 *
4 *
5 '
6 *
7 *
8 *
9 *
10 *
11 *
12 '
13 *
14 *
15 '
16 *
17 *
18 *
19 '
20 *
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
ANALYTE
1,1,2-Trichloroethane
1 ,2,4-Trimethylbenzene
1 ,3.5-Trimethylbenzene
Benzene
Bromodichloromethane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1,3 - Dichloropropene
BDB (1,2-Dibrorroethane)
Bhylbenzene
Methylene chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1,3 - Dichloropropene
Vinyl Chloride
1,1 - Dichloropropene
1,1-Dichloroethane
1,1-Dichloroethene
1,1,1-Trichloroethane
1,1,1,2- Tetrachloroethane
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dibrorro-3-Chloropropane
1 ,2-Dichlorobenzene
1,2-D'chloroethane
1 ,2-Dichloropropane
1,2,3 - Trichloropropane
1,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dichlorobenzene
1 ,3-Dichloropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Allyl Chloride
Bromobenzene
BromDchloromethane
Chlorobenzene
Chloroethane
Chloromethane
cis-1 ,2-Dichloroetnene
Dibromochlororrethane
D'bromomethane
Dichlorofluoromethane
Di-teopropyl ether
Bhyl Bher
Hexachlorobutadiene
bopropylbenzene
n-Butylbenzene
n-R-opylbenzene
p-teopropyltoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trichlorofluororrethane
Trichlorotrifluoroethane
7 Day
Absol
141
144
144
145
134
128
138
148
138
134
135
147
149
282
127
149
137
138
140
124
148
142
130
152
143
132
124
139
143
132
154
132
121
135
150
128
137
135
139
154
132
141
142
123
117
150
138
147
148
145
149
140
146
134
140
139
142
146
141
131
158
118
124
Rec. %
91
93
94
94
87
83
90
96
89
87
88
96
97
92
82
97
89
89
91
80
96
92
84
99
93
86
80
90
93
86
100
86
79
87
98
83
89
87
90
100
86
92
92
80
76
97
90
96
96
94
97
91
95
87
91
90
92
95
91
85
103
77
B1
7 Day
Absol
146
146
150
143
139
126
145
148
143
141
145
150
152
274
127
143
135
145
144
128
147
154
142
165
146
121
174
143
144
147
138
136
132
134
141
128
143
139
138
158
141
158
138
141
124
148
136
156
156
148
15U
15/
152
13/
142
145
148
149
148
141
161
13/
130
Rac %
9!)
95
98
93
90
82
94
96
93
92
94
97
99
89
82
93
88
94
93
83
95
100
92
107
95
78
113
93
94
96
90
88
86
tif
92
83
93
90
90
103
91
102
90
92
80
96
88
101
101
96
97
102
99
89
92
94
96
97
96
92
104
89
8b
7 Day
Absol.
152
154
151
153
155
136
158
160
163
164
1b3
149
1b/
296
121
148
142
156
157
126
158
162
154
161
162
137
13/
1bO
151
155
149
137
135
144
152
137
149
141
152
159
151
162
150
144
128
157
151
160
168
151
160
152
162
142
150
14/
152
155
161
148
1/b
154
148
Rec. %
99
100
98
99
101
88
103
104
106
100
99
9/
102
96
rt»
96
92
101
102
82
102
105
100
105
10b
89
89
9/
98
101
97
89
88
93
99
89
9/
91
99
103
98
105
98
94
83
102
98
104
109
98
104
99
105
92
9/
96
99
100
105
96
113
100
96
7 Day
Absol
133
138
136
133
122
121
129
138
135
136
130
131
135
260
123
130
132
132
133
104
133
135
126
148
139
119
129
136
131
120
143
130
124
12^
128
128
130
120
134
146
135
128
130
131
110
129
131
137
150
130
133
152
146
129
13b
130
139
142
135
116
141
127
128
Rec %
86
89
88
86
79
78
84
90
88
88
85
85
88
84
80
85
86
86
86
68
86
88
82
96
90
77
84
88
86
fU
93
Hb
80
83
83
83
84
/8
8/
95
tif
83
84
85
71
84
85
89
97
85
86
99
95
84
8/
85
90
92
88
75
91
82
83
7 Day
Absol
132
132
134
127
134
126
140
143
144
140
133
134
134
254
111
134
127
142
137
108
127
137
125
148
135
112
119
129
133
128
127
125
117
127
13/
120
133
119
139
136
133
129
129
128
97
13b
127
140
153
130
129
126
138
123
131
132
135
13/
140
121
155
138
12b
Rec. %
86
86
87
82
87
82
91
93
93
91
86
87
87
83
72
87
82
92
89
70
82
89
81
96
88
73
77
84
86
83
83
81
76
82
89
/8
86
77
90
88
86
84
84
83
63
8/
83
91
99
85
84
82
89
80
85
86
88
89
91
78
100
89
82
Absolute
AVG.
141
142
143
140
137
127
142
147
144
141
139
142
145
273
121
140
134
142
142
118
142
146
135
155
145
124
136
139
140
136
142
132
126
133
141
128
138
130
140
150
138
143
138
133
115
143
136
148
95
155
141
144
145
149
133
139
139
143
146
145
131
158
135
AVG. % REC
of 0 Hour
97.4
96.2
94.8
93.0
88.8
89.7
79.8
86.0
95.5
96.2
91 7
91.1
91.5
89.8
98.6
91.9
91.4
93.8
87.6
88.5
85.0
92.8
82.7
62.6
96.5
88.4
100.3
95.1
80.1
102.2
101.7
103.1
100.7
91.3
96.3
91.0
94.8
70.6
91.6
96.0
945
860
92.0
72.9
96.0
92.1
94.1
L 923
63.6
73.8
97.2
34.4
93.7
95.8
100.9
93.5
94.3
98.0
94.7
34.7
33.8
105.8
70.2
%
RSD.
6.0
5.7
5.6
7.3
8.9
4.3
7.6
5.4
7.5
5.7
6.8
62
7.1
6.1
5.4
5.9
4.2
6.2
6.4
9.4
8.7
7.9
9.3
5.0
7.0
8.1
16.1
55
5.9
10.7
7.3
3.6
6.1
52
6.9
4.6
57
80
4.9
6.3
58
10.9
6.4
6.8
10.5
8.0
6.6
6.6
8.4
5.2
7.1
9.0
8.7
6.1
5.7
5.1
5.5
4.7
4.7
7.0
10.4
7.7
10.0
* Denotes analyte required to be added to this study by the WDNR and w hich mist pass the irrposed criteria to gain
acceptance for this method in Wisconsin.
141
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS
Matrix - Clean Laboratory Sand
VOC METHOD 8260 STANDARD 153.8 UG/KG / 3.08 UG/L
Table 6
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
ANALVTE
1,1,2-Trichloroethane
1 ,2,4-TrimBthylbenzene
1,3,5-Trimethylbenzene
Benzene
Bromodichlororrethane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1,3 - Dichloropropene
B» (1,2-Dbromoethane)
Bhylbenzene
Methylene chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1,3 - Dchloropropene
Vinyl Chloride
1,1 - Dchloropropene
1,1-Dichloroethane
1,1-Dchloroethene
1,1,1 -Trichloroethane
1,1,1,2 - Tetrachloroethane
1 ,1 ,2,2-Tetrachloroethane
1 ,2-Dibrorro-3-Chloropropane
1 ,2-Dichlorobenzene
1,2-D'chloroethane
1 ,2-Dichloropropane
1,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dichlorobenzene
1 ,3-Dichloropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Allyl Chloride
Bromobenzene
Brorrochloromethane
Chlorobenzene
Chloroethane
Chloromethane
cis-1,2-Dichloroethene
Dibrorrochlororrethane
Dibrorromethane
Dchlorofluoromethane
Di-teopropyl ether
Bhyl Bher
Hexachlorobutadiene
teopropylbenzene
n-Butylbenzene
n-Propylbenzene
p-feopropyltoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trichlorofluoromethane
Trichlorotrifluoroethane
10 Day
Absol
135
140
139
145
140
130
160
154
138
145
143
151
159
281
121
144
137
148
139
122
154
150
159
165
139
120
148
137
150
141
147
127
112
132
147
126
138
137
136
149
141
162
138
151
127
149
139
139
168
149
165
140
150
130
141
137
145
152
144
136
161
146
141
Rec %
87
91
90
94
91
85
104
100
89
94
93
98
103
91
/9
94
89
96
90
79
100
98
103
107
90
/8
96
89
98
92
96
82
72
86
96
82
90
89
88
97
91
105
90
98
83
97
90
90
109
97
107
91
97
84
91
89
94
99
94
88
105
9b
91
10 Day
Absol.
143
136
135
152
139
128
146
154
141
141
138
149
151
269
115
140
136
147
137
134
163
153
141
166
151
111
108
135
158
140
148
128
117
133
149
127
137
138
138
159
13/
158
142
168
121
151
136
151
172
150
160
127
148
121
138
138
140
14/
144
132
172
157
138
Rec %
93
88
88
99
90
83
94
100
92
91
90
9/
98
87
/4
91
88
95
89
87
99
99
92
108
98
12
70
88
102
91
96
83
76
86
97
82
89
90
89
103
89
102
92
109
78
98
88
98
112
98
104
83
96
79
90
90
91
96
94
86
112
102
90
10 Day
Absol
135
132
130
135
125
130
134
143
130
128
130
133
140
254
121
128
127
137
131
111
138
139
138
148
123
115
124
130
145
123
139
130
112
133
131
127
127
126
130
132
126
134
124
143
108
127
123
138
155
135
142
156
141
120
134
136
138
145
133
124
146
144
119
Rec. %
88
8b
84
87
81
85
8/
93
84
83
85
86
91
83
/9
83
82
89
85
72
89
90
90
96
80
/£>
81
85
94
80
90
85
72
86
85
83
82
82
84
86
82
87
80
93
70
82
80
90
100
88
92
101
92
78
8/
88
89
94
86
80
95
94
77
10 Day
Absol.
148
132
134
147
134
125
15/
153
13b
148
139
144
158
262
119
138
138
151
144
125
158
150
148
169
143
105
153
138
153
144
142
133
114
129
142
123
136
138
131
153
127
161
134
161
118
146
139
152
164
145
162
137
140
123
136
131
139
141
142
146
169
152
151
Rec. %
96
86
87
95
87
81
102
99
88
96
90
93
102
85
77
90
89
98
94
81
103
98
96
110
93
68
99
89
99
93
92
86
74
84
92
80
88
90
85
99
83
104
87
105
76
95
90
99
106
94
105
89
91
80
88
85
90
92
92
95
110
99
98
10 Day
Absol
150
141
141
15b
139
130
160
156
143
143
142
150
158
280
122
140
138
148
131
124
163
159
143
170
141
117
152
138
158
136
14/
140
126
134
145
127
139
129
138
161
141
155
142
159
130
156
143
152
186
149
156
147
148
129
141
141
143
151
145
140
1/5
149
142
Rec %
98
92
91
101
90
84
104
101
93
93
92
98
102
91
79
91
89
96
85
81
106
103
93
111
91
76
99
90
102
08
95
91
82
87
94
83
90
84
90
105
91
100
92
103
84
101
93
99
121
97
101
95
96
84
91
91
93
98
94
91
114
9/
92
Absolute
AVG.
142
136
136
147
135
129
151
152
137
141
138
145
153
269
119
138
135
146
136
123
153
150
146
164
139
113
137
136
153
137
145
131
116
132
142
126
135
133
134
151
134
154
136
156
121
146
136
146
169
146
157
141
145
125
138
136
141
147
142
136
164
149
138
AVG. % REC
of 0 Hour
98.3
91.9
90.1
97.4
87.8
90.7
85.1
88.7
90.9
96.1
91.2
93.0
96.4
88.5
97.0
90.4
91.8
96.1
84.2
92.5
91.3
95.4
89.2
87.3
92.7
!0.9
100.7
92.5
87.1
102.5
103.5
102.7
93.0
90.6
97.0
89.5
, 92.8
72.1
87.7
96.0
91.7
92.0
90.9
85.5
100.7
93.4
93.7
91.4
80.4
100.6
92.1
91.1
93.5
94.6
92.4
92.8
96.5
95.6
82.8
86.6
110.4
77.9
90.4
%
RSD.
5.0
3.2
3.2
5.4
4.6
1.7
7.6
3.3
3.9
5.6
3.7
5.2
5.1
4.3
2.5
4.3
3.5
3.8
4.3
6.7
6.2
4.7
5.7
5.4
7.2
5.1
14.8
2.4
3.6
5.9
2.7
4.0
5.0
1.4
5.0
1.3
3.7
4.4
3.0
7.7
5.5
7.5
5.7
6.3
7.0
7.7
5.5
4.8
6.8
4.2
5.8
7.6
3.0
3.6
2.1
2.7
2.1
2.9
3.5
6.3
7.2
3.4
8.5
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
* Denotes analyte required to be added to this study by the WDNR and w hich mist pass the in-posed criteria to gain
acceptance for this method in Wisconsin.
142
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS Table 7
Matrix - Clean Laboratory Sand
VOC METHOD 8260 STANDARD 769.2 UG/KG /15.38 UG/L
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
ANALYTE
1,1,2-Trichloroethane
1 ,2,4-TrimBthylbenzene
1 ,3,5-Trimethylbenzene
Benzene
Brorrodichloromethane
Brorroform
Carbon Tetrachloride
Chloroform
cis - 1,3 - Dichloropropene
H3B (1,2-Dibromoethane)
Bhylbenzene
Methylene chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1,3 - Dichloropropene
Vinyl Chloride
1,1 - Dichloropropene
1,1-Dichloroethane
1,1-Dichloroethene
1,1,1 -Trichloroethane
1,1,1,2- Tetrachloroethane
1,1,2,2-Tetrachloroethane
1 ,2-Dibromo-3-Chloropropane
1 ,2-Dichlorobenzene
1 ,2-Dichloroethane
1 ,2-Dichloropropane
1,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dichlorobenzene
1 ,3-Dichtoropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chtorotoluene
Allyl Chloride
Bromobenzene
Bromochloromethane
Chlorobenzene
Chloroethane
Chloromethane
cis- 1 ,2-Dichloroetnene
Dibromochlororrethane
Dibromorrethane
Dichlorofluoromethane
Di-bopropyl ether
Bhyl Bher
Hexachlorobutadiene
(sopropylbenzene
n-Butylbenzene
n-R"opylbenzene
p-bopropy toluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-Dchloroethene
Trichloroethene
Trichlorofluoromethane
Trichlorotrifluoroethane
0 Hour
Absol
772
734
742
762
819
674
876
844
792
818
781
788
817
1503
689
758
764
768
802
681
822
829
768
874
805
542
788
702
846
735
738
735
696
703
766
696
703
686
745
649
697
883
/61
791
631
789
821
853
841
/bti
823
724
770
714
734
750
740
767
814
777
967
773
664
Rec %
100
95
96
99
106
88
114
110
103
106
102
102
106
98
90
99
99
100
104
89
107
108
100
114
105
70
102
91
110
96
96
96
90
91
100
90
91
89
97
84
91
115
99
103
82
103
107
111
109
99
107
94
100
93
95
97
96
100
106
101
126
100
86
0 Hour
Absol
774
771
767
772
810
692
840
853
805
821
774
780
796
1524
689
780
776
791
814
673
821
813
784
886
828
642
823
725
842
760
769
773
736
715
786
710
722
672
755
652
732
893
/6b
768
628
818
8/2
841
845
76 1
844
738
784
732
755
762
761
795
827
803
893
723
649
Rec %
101
100
100
100
105
90
109
111
105
107
101
101
103
99
90
101
101
103
106
87
107
106
102
115
108
83
107
94
109
99
100
100
96
93
102
92
94
87
98
85
95
116
99
100
82
106
113
109
110
100
110
96
102
95
98
99
99
103
107
104
116
94
84
0 Hour
Absol.
812
767
766
776
842
727
867
863
826
826
798
807
826
1568
718
790
773
802
818
678
832
830
784
895
842
576
811
728
870
752
737
745
724
722
775
702
721
686
/M
641
738
901
790
779
641
818
890
849
861
770
8b4
766
787
740
751
772
774
790
82 /
843
9b3
/92
/23
Rec %
106
100
100
101
109
95
113
112
107
107
104
105
107
102
93
103
100
104
106
88
108
108
102
116
109
75
105
95
113
98
96
97
94
94
101
91
94
89
98
83
96
117
103
101
83
106
116
110
112
100
111
100
102
96
98
100
101
103
108
110
124
103
94
0 Hour
Absol
745
720
723
/29
/9b
675
829
812
769
776
734
752
770
1465
659
737
732
731
774
664
779
790
779
843
786
541
782
688
796
714
694
693
672
670
740
659
679
618
720
615
676
838
727
778
610
772
813
82 /
83 /
/33
8Ub
706
742
678
709
720
720
740
/V4
758
914
/60
68b
Rec %
97
94
94
95
103
88
108
105
100
101
95
98
100
95
86
96
95
95
101
86
101
103
101
110
102
70
102
89
103
93
90
90
87
87
96
86
88
80
94
80
88
109
95
101
79
100
106
107
109
95
105
92
96
88
92
94
94
96
101
99
119
99
89
0 Hour
Absol.
777
758
769
766
827
683
873
864
781
824
790
787
810
1555
709
777
779
783
811
681
823
814
810
885
842
578
810
709
858
746
757
748
719
722
783
696
724
659
749
651
744
906
771
765
629
817
856
836
8b2
768
(•68
728
788
712
/bb
757
761
784
81/
823
948
IH
682
Rec %
101
98
100
100
107
89
113
112
102
107
103
102
105
101
92
101
101
102
105
88
107
106
105
115
109
75
105
92
111
97
98
97
93
94
102
90
94
86
97
85
97
118
100
99
82
106
111
109
111
100
100
95
102
93
98
98
99
102
106
107
123
101
89
Absolute
AVG.
776
750
753
761
818
690
857
847
795
813
775
782
803
1523
693
768
765
775
804
675
815
815
785
876
820
576
803
710
842
741
739
739
709
706
770
692
710
664
744
641
717
884
763
776
628
803
850
841
847
759
818
732
774
715
740
752
751
775
812
801
935
764
680
Average
%RBC.
100.8
97.5
97.9
98.9
106.4
89.7
111.4
110.1
103.3
105.7
100.8
1017
1044
990
900
99.9
99.4
100.7
1045
87.8
1060
1059
102.0
113.9
106.6
748
104.4
92.3
109.5
96.4
960
96.0
92.2
91.8
100 1
900
922
863
968
834
932
114.9
99.1
1009
81 C
1043
1105
1093
110 1
98.6
1064
952
100 C
930
963
977
97.6
100.8
105.5
104 1
121.5
993
884
RSD.
3.1
2.9
2.7
2.5
2.2
3.2
2.5
2.5
2.8
2.6
3.2
2.5
2.7
2.7
3.3
2.7
2.5
3.6
2.2
1.1
2.6
2.0
2.0
2.3
30
7.1
2.1
2.3
3.3
2.4
3.9
4.0
3.6
3.1
2.4
2.9
2.7
4.3
1.9
2.4
4.1
3.1
3.0
1.3
1.8
2.6
3.9
1.2
1.1
20
42
3.0
2.5
3.4
2.7
2.6
2.8
2.9
2.7
4.3
3.3
3.4
4.1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
63
* Denotes analyte required to be added to this study by the WDNR and w hich must pass the imposed criteria to gain
acceptance for this method in Wisconsin.
143
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS
Matrix - Clean Laboratory Sand
VOC METHOD 8260 STANDARD 769.2 UG/KG /15.38 UG/L
Table 8
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
ANALYTE
1 , 1 ,2-Tnchloroethane
1 ,2,4-Trimethylbenzene
1,3,5-Trimethylbenzene
Benzene
Bromodichloromethane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1 ,3 - Dichloropropene
BDB (1,2-Dibromoethane)
Bhylbenzene
Methylene chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1,3 - Dichloropropene
Vinyl Chloride
1,1 - Dichloropropene
1,1-Dichloroethane
1,1-Dichloroethene
1,1,1-Trichloroethane
1,1,1,2 - Tetrachloroethane
1 , 1 ,2,2-Tetrachloroethane
1,2-Dibrorro-3-Chloropropane
1 ,2-Dichlorobenzene
1,2-Dichloroethane
1 ,2-Qchloropropane
1,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dichlorobenzene
1 ,3-Dichloropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Allyl Chloride
Bromobenzene
Brorrochloromethane
Chlorobenzene
Chloroethane
Chlororrethane
cis-1,2-Dichloroethene •
Dibrorrochlororrethane
Dibrorromethane
Dichlorofluoromethane
Di-bopropyl ether
Bhyl Bher
Hexachlorobutadiene
teopropylbenzene
n-Butylbenzene
n-FYopylbenzene
p-teopropyltoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trichlorofluororrethane
Trichlorotrifluoroethane
24 Hour
Absol.
745
731
708
715
797
649
8/8
815
799
756
744
714
762
1492
643
744
700
729
818
617
750
770
746
8/6
/81
633
739
692
804
701
665
709
697
681
713
66b
683
805
715
602
68 /
/84
742
723
531
762
841
824
854
717
781
747
738
710
697
/32
727
747
77b
/42
806
82b
644
Rec %
97
95
92
93
104
84
114
106
104
98
97
93
99
97
84
97
91
95
106
80
97
100
97
114
101
82
96
90
105
91
86
92
91
89
93
86
89
105
93
78
89
102
96
94
69
99
109
107
111
93
102
97
96
92
91
9b
94
97
101
96
105
107
84
24 Hour
Absol.
741
715
719
712
793
703
888
824
/b8
749
/40
/b9
/79
1481
686
735
741
/30
794
629
779
783
741
869
782
685
843
69 /
810
694
/03
/29
704
691
747
6//
683
795
716
624
694
829
726
710
534
770
842
781
842
722
798
807
740
709
/08
743
746
749
771
/34
801
836
666
Rec %
96
93
93
92
103
91
115
107
98
97
96
99
101
96
89
96
96
95
103
82
101
102
96
113
102
89
110
91
105
90
91
95
92
90
97
88
89
103
93
81
90
108
94
92
69
100
109
101
109
94
104
105
96
92
92
97
97
97
100
9b
104
109
87
24 Hour
Absol.
728
720
721
602
754
660
69b
716
801
732
/09
564
624
1432
678
724
711
681
795
363
757
614
508
723
781
622
721
702
728
636
676
740
717
691
705
683
680
626
710
588
681
803
710
467
313
644
837
807
819
620
773
758
724
700
711
740
723
740
708
b6/
742
49b
680
Roc %
95
94
94
78
98
86
90
93
104
95
92
73
81
93
88
94
92
89
103
47
98
80
66
94
101
81
94
91
95
83
88
96
93
90
92
89
88
81
92
76
88
104
92
61
41
84
109
105
106
81
100
98
94
91
92
96
94
96
92
74
96
64
88
24 Hour
Absol
746
707
/29
704
817
682
859
819
789
761
707
711
759
1408
614
693
725
738
802
620
765
747
723
845
807
628
786
676
802
696
703
697
679
677
719
659
682
744
703
609
671
831
721
760
534
745
835
826
857
699
825
738
728
679
701
721
710
734
797
708
847
/9b
666
Rec %
9/
92
95
92
106
89
112
106
103
99
92
92
99
92
80
90
94
96
104
81
99
97
94
110
105
82
102
88
104
90
91
91
88
88
93
86
89
97
91
79
87
108
94
99
69
97
109
107
111
91
107
96
95
88
91
94
92
95
104
92
110
103
87
24 Hour
Absol
743
733
739
722
803
668
843
841
/4b
765
/44
/49
/BO
1460
682
739
711
/3b
781
613
737
780
749
8/b
757
641
774
687
830
697
686
740
733
683
/29
679
690
740
722
b84
701
777
735
716
552
770
852
766
800
714
736
718
752
715
717
/44
731
752
756
740
816
785
61 /
Roc. %
743
733
739
722
803
668
843
841
745
765
744
749
780
95
682
739
711
735
781
613
737
780
749
875
757
641
774
687
830
697
686
740
733
683
729
679
690
740
722
584
701
777
735
716
552
770
852
766
800
714
736
718
752
715
717
744
731
752
756
740
816
/8b
61 /
Absolute
AVG.
740.4
721.1
722.9
690.7
792.4
672.3
832.5
802.9
778.1
752.3
728.6
699.2
740.7
1454.2
660.4
726.8
717.3
722.5
797.7
568.3
757.2
738.5
693.1
837.3
781.2
641.5
772.4
690.4
794.6
684.6
686.3
722.9
705.6
684.5
722.4
672.4
683.5
741.8
713
601.1
686.6
804.6
726.6
674.9
492.6
738.1
841.2
800.6
834
694.2
782.5
753.4
736
702.3
706.5
735.9
727
744.3
761
698
802.2
747
654.3
AVG % REC
of 0 Hour
95.4
96.2
96.0
90.8
96.8
97.4
97.2
94.8
97.9
92.6
94.0
89.4
92.2
95.5
95.4
94.6
93.8
93.3
99.3
J4.2
92.9
90.6
88.3
95.5
953
111.5
96.2
97.2
94.3
92.4
92.9
97.9
99.5
96.9
93.8
971
96.3
111.7
95.8
93.7
95.7
91.0
95.3
87.0
78.5
92.0
99.0
95.2
98.5
91.5
95.6
102.9
95.1
98.2
95.4
97.9
96.8
96.0
938
87.2
85.8
97.7
96.2
%
RSD.
1.0
1.5
1.6
7.2
3.0
3.1
9.5
6.2
3.3
1.7
2.6
11.3
8.9
2.4
4.7
2.8
2.2
3.3
1.7
20.2
2.1
9.7
15.0
7.8
2.3
3.9
6.1
1.4
4.9
4.0
2.4
2.7
2.9
0.9
2.2
1.5
0.6
9.6
1.0
2.7
1.7
3.1
1.7
17.5
20.5
7.3
0.8
3.3
2.9
6.1
4.2
4.4
1.5
2.0
1.2
1.3
1.8
1.0
4.4
10.7
4.8
9.1
3.7
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
* Denotes analyte required to be added to this study by the WDNF and w hich mjst pass the imposed criteria to gain
acceptance for this method in Wisconsin.
144
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS Table 9
Matrix - Clean Laboratory Sand
VOC METHOD 8260 STANDARD 769.2 UG/KG /15.38 UG/L
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
ANALYTE
1,1,2-Trichloroethane
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
Benzene
Bromodichloromethane
Brorroform
Carbon Tetrachloride
Chloroform
cis - 1 ,3 - Dichloropropene
E3B (1.2-Dibromoethane)
Bhylbenzene
Methylene chloride
MTBE
m&p-Xylene
naphthalene
o-Xylene
Styrene
Toluene
trans - 1,3 - Dichloropropene
Vinyl Chloride
,1 - Dichloropropene
,1-Dichloroethane
,1-Dichloroethene
,1,1 -Trie hloroethane
,1,1,2 - Tetrachloroethane
,1,2,2-Tetrachloroethane
1 ,2-Dibrom>3-Chloropropane
1,2-0ichloroben2ene
1,2-Dichloroethane
1 ,2-Dichloropropane
1,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Qchtorobenzene
1 ,3-Dichloropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Allyl Chloride
Brorrobenzene
BrorrDchloromethane
Chlorobenzene
Chloroethane
Chtoromethane
cis-1 ,2-Dichloroethene
Dibrorrochloromethane
Dibrorrorrethane
Dichtorofluoromethane
Di-bopropyl ether
Bhyl Bher
Hexachlorobutadiene
bopropylbenzene
n-Butylbenzene
n-R-opylbenzene
p-feopropyltoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trichlorofluoromethane
Trichlorotrifluoroethane
48
Absol.
730
711
727
691
821
649
825
812
799
762
726
720
749
1395
638
704
700
731
818
588
750
758
720
829
781
676
696
670
807
714
665
719
707
668
741
654
659
800
699
602
673
784
712
743
529
734
838
824
854
712
781
727
731
695
686
713
710
714
776
721
784
752
644
Hour
Roc. %
95
92
95
90
107
84
107
106
104
99
94
94
97
91
83
91
91
95
106
76
97
98
94
108
101
88
90
87
105
93
86
93
92
87
96
85
86
104
91
78
87
102
92
97
69
95
109
107
111
92
102
95
95
90
89
93
92
93
101
94
102
98
84
48
Absol
744
726
730
708
784
703
849
824
758
788
742
753
776
1466
684
744
741
729
794
582
779
768
702
842
782
743
754
668
799
675
703
731
692
667
745
663
676
793
715
624
680
829
730
738
543
768
820
781
842
701
798
729
738
688
699
727
716
737
798
733
714
715
666
Hour
R»c %
97
94
95
92
102
91
110
107
98
102
96
98
101
95
89
97
96
95
103
76
101
100
91
109
102
97
98
87
104
88
91
95
90
87
97
86
88
103
93
81
88
108
95
96
71
100
107
101
109
91
104
95
96
89
91
94
93
96
104
9b
93
93
87
48
Absol
751
723
715
701
835
660
821
802
801
760
718
710
725
1419
644
718
711
731
795
571
757
746
712
844
781
690
702
683
785
702
676
718
678
681
729
648
688
780
702
588
685
803
712
736
521
739
840
807
819
697
773
729
734
692
708
730
707
729
795
715
747
745
680
Hour
Rec %
98
751
93
91
109
86
107
104
104
99
93
92
94
92
84
93
92
95
103
74
98
97
93
110
101
90
91
89
102
91
88
93
88
89
95
84
89
101
91
76
89
104
93
96
68
96
109
105
106
91
100
95
95
90
92
95
92
95
103
93
97
97
88
48
Absol.
757
727
729
703
813
682
885
806
789
766
748
737
781
1452
672
737
725
745
802
620
765
764
732
865
807
762
771
668
822
702
703
725
695
676
743
660
682
787
713
609
697
831
731
749
561
765
868
826
857
704
825
693
747
692
705
725
724
736
794
710
731
780
666
Hour
Rec %
98
94
95
91
106
89
115
105
103
100
97
9b
102
94
87
96
94
97
104
81
99
99
95
112
105
99
100
87
107
91
91
94
90
88
97
86
89
102
93
79
91
108
95
97
73
99
113
107
111
92
107
90
97
90
92
94
94
96
103
92
95
101
87
48
Absol
712
690
708
673
754
668
803
801
745
730
709
691
750
1430
650
706
711
699
781
545
737
738
658
826
757
680
737
648
775
663
686
709
686
659
696
642
652
731
694
584
664
777
706
684
506
722
790
766
800
685
736
719
709
678
676
695
694
692
727
no
723
692
617
Hour
Rac. %
93
90
92
87
98
87
104
104
97
95
92
90
97
93
85
92
92
91
102
71
96
96
85
107
98
88
96
84
101
86
89
92
89
86
90
83
85
95
90
76
86
101
92
89
66
94
103
100
104
89
96
93
92
88
88
90
90
90
94
92
94
90
80
Absolute
AVG.
739
715
722
695
801
672
836
809
778
761
728
722
756
1432
657
722
717
727
798
581
757
754
704
841
781
710
732
667
797
691
686
720
691
670
731
653
671
778
704
601
680
805
718
730
532
745
831
801
834
700
783
719
732
689
694
718
710
721
778
718
739
737
654
AVG. % REC
of 0 Hour
95.2
95.4
95.8
91.4
97.9
97.4
97.6
95.5
97.9
93.7
93.9
92.3
94.1
94.1
94.9
93.9
93.8
93.8
99.3
86.0
92.9
926
898
960
953
123.3
91.2
94.0
94.7
93.2
92.9
97.5
97.5
94.9
94.9
94.3
94.6
117.1
94.6
93.7
94.8
91.0
94.2
94.1
84.7
92.9
978
952
98.5
92.2
95.6
98.2
94.5
96.3
93.8
95.5
94.5
93.1
95.8
89.6
79.1
96.4
96.2
/>
RSD.
2.4
2.2
1.3
2.0
4.0
3.1
3.8
1.1
3.3
2.7
2.2
3.3
3.0
2.0
3.0
2.5
2.2
2.3
1.7
4.7
2.1
1.7
4.0
1.8
2.3
5.6
4.5
1.8
2.3
3.1
2.4
1.1
1.6
1.3
2.8
1.3
2.3
3.5
1.3
2.7
1.8
3.1
1.6
3.6
3.9
2.7
3.5
3.3
2.9
1.4
4.2
2.1
1.9
0.9
1.9
2.0
1.6
2.6
3.8
1 4
3.7
4.6
3.7
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
* Denotes analyte required to be added to this study by the WDNR and w hich rrust pass the Imposed criteria to gain
acceptance for this method in Wisconsin.
145
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS
Matrix - Clean Laboratory Sand
VOC METHOD 8260 STANDARD 769.2 UG/KG /15.38 UG/L
Table 10
1 *
2 *
3 *
4 *
5 *
6 *
7 *
8 *
9 *
10 *
11 *
12 *
13 *
14 *
15 '
16 *
17 *
18 *
19 *
20 *
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
ANALYTE
1,1,2-Trfchloroethane
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
Benzene
Bromodichlororrethane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1 ,3 - Dichloropropene
BDB (1,2-Dibromoethane)
Bhylbenzene
Methylene chloride
MTBE
m&p-Xylene
Naphthalene
c-Xylene
Styrene
Toluene
trans - 1 ,3 - Dichloropropene
Vinyl Chloride
,1 - Dichloropropene
,1-Dichloroethane
,1-Dichloroethene
,1,1-Trichloroethane
,1,1,2 - Tetrachloroethane
, 1 ,2,2-Tetrachloroethane
,2-Dibromo-3-Chloropropane
,2-Dichlorobenzene
,2-Dichloroethane
,2-Dichloropropane
,2,3 - Trichloropropane
,2,3-Trichlorobenzene
,2,4-Trichlorobenzene
1 ,3-Dichlorobenzene
1 ,3-Dichloropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Allyl Chloride
Bromobenzene
Bromochloromethane
Chlorobenzene
Chloroethane
Chloromethane
cis-1 ,2-Dichloroethene
Dibrorrochloromethane
Pbrorrorrethane
Dichlorofluoromethane
Di-bopropyl ether
Bhyl Bher
Hexachlorobutadiene
teopropylbenzene
n-Butylbenzene
n-R-opylbenzene
p-tsopropyttoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trichlorofluororrethane
Trichlorotriftuoroethane
72 Hour
Absol
781
728
744
717
849
684
882
838
823
/75
741
/23
782
1457
6b3
748
711
746
820
611
806
773
//8
886
798
696
72b
682
811
737
675
/12
702
682
754
665
689
769
718
605
695
836
740
767
523
767
866
832
871
726
793
751
749
/12
720
/43
734
764
808
732
808
866
'fbf
Rec %
101
95
9/
93
110
89
11b
109
10/
101
96
94
102
95
8b
9/
92
97
1U/
79
105
100
101
115
104
90
94
89
105
96
88
93
91
89
98
86
9U
100
93
79
90
109
96
100
68
100
113
108
113
94
103
98
9/
92
94
9/
95
99
105
9b
105
111
98
72 Hour
Absol
/13
710
718
695
758
669
836
801
743
742
719
696
762
1422
666
711
711
710
751
598
773
743
726
840
756
740
771
677
793
654
710
710
674
672
715
661
673
734
707
604
700
795
699
715
536
743
814
764
836
685
771
714
732
685
696
738
722
743
747
689
686
794
711
Rec.%
93
92
93
90
98
87
109
104
97
96
93
90
99
92
87
92
92
92
98
78
100
97
94
109
98
96
100
88
103
85
92
92
88
87
93
86
87
95
92
78
91
103
91
93
70
97
106
99
109
89
100
93
9b
89
90
96
94
97
97
90
89
1U3
92
72 Hour
Absol
730
711
713
711
833
668
844
843
788
760
735
721
754
1456
618
722
728
744
799
612
772
772
747
879
813
667
748
687
815
700
660
682
685
675
747
668
667
734
717
605
687
831
721
747
512
762
842
825
864
716
787
71b
742
684
706
72b
703
728
794
736
801
809
718
Rec %
98
751
93
92
108
87
110
110
102
99
95
94
98
95
80
94
95
97
104
80
100
100
97
114
106
87
97
89
106
91
86
89
89
88
97
87
87
95
93
79
89
108
94
97
67
99
109
107
112
93
102
93
96
89
92
94
91
95
103
96
104
105
93
72 Hour
Absol
747
709
71b
683
8U3
677
833
803
7/0
"774
722
709
759
1417
6bb
711
700
718
779
560
740
749
696
840
774
666
743
665
811
690
693
703
699
663
717
656
674
698
719
588
717
818
712
745
529
754
832
782
837
697
784
706
737
690
703
720
712
722
781
753
788
770
bb3
Rec. %
97
92
93
89
104
88
108
104
100
101
94
92
99
92
85
92
91
93
101
73
96
97
90
109
101
87
97
86
105
90
90
91
91
86
93
85
88
91
93
76
93
106
93
97
69
98
108
102
109
91
102
92
96
90
91
94
93
94
102
98
102
100
86
72 Hour
Absol.
759
722
723
712
816
719
889
844
785
764
756
725
793
1b09
703
764
7b9
752
792
606
759
786
706
892
798
702
7/0
697
825
688
720
746
/12
686
753
679
677
739
/16
634
683
855
744
771
524
/84
861
840
871
/27
809
736
/26
698
/OO
/46
/27
736
81 /
/41
782
788
704
Rec %
99
94
94
92
106
93
116
110
102
99
98
94
103
98
91
99
99
98
103
79
99
102
92
116
104
91
100
91
107
89
94
97
92
89
98
88
88
96
93
82
89
111
97
100
68
102
112
109
113
94
105
96
94
91
91
97
95
96
106
96
102
102
91
Absolute
AVG.
746
716
722
703
812
683
857
826
782
763
734
714
770
1452
659
731
722
734
788
597
770
765
731
867
788
694
751
681
811
693
691
711
694
675
737
666
676
734
715
607
696
827
723
749
525
762
843
808
856
710
789
724
737
693
705
734
720
738
789
730
773
803
710
AVG % REC
of 0 Hour
96.1
95.5
95.9
92.5
99.2
99.0
100.0
97.5
98.4
93.8
94.7
91.3
95.8
95.4
95.1
95.1
94.4
94.7
98.1
884
94.5
93.8
93.1
99.0
96.0
120.5
93.6
95.9
96.3
93.6
93.6
96.2
97.9
95.6
95.8
96.1
95.3
110.6
96.1
94.6
97.1
93.5
94.8
96.5
83.6
949
99.1
96.2
101.1
93.6
96.3
98.9
95.2
97.0
95.2
97.6
95.8
95.3
97.2
91.2
82.7
105.1
104.4
%
RSD.
3.5
1.2
1.8
2.0
4.3
3.0
3.1
2.6
3.7
1.8
2.0
1.7
2.2
2.5
4.6
3.3
3.2
2.5
3.3
3.6
3.1
2.4
4.5
2.9
2.9
4.4
2.6
1.8
1.4
4.3
3.6
3.2
2.1
1.3
2.6
13
1 2
3.5
0.7
2.8
1.9
2.7
2.6
3.0
1.6
2.0
2.5
4.2
2.1
2.6
1.8
2.5
1.2
1.7
1.3
1.5
1.7
2.2
3.4
3.3
64
41
47
* Denotes analyte required to be added to this study by the WDMR and w hich mist pass the irrposed criteria to gain
acceptance for this method in Wisconsin
146
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS Table 11
Matrix - Clean Laboratory Sand
VOC METHOD 8260 STANDARD 769.2 UG/KG / 15.38 UG/L
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
1 *
2 '
3 '
4 *
5 *
6 '
7 '
8 *
9 '
10 *
11 '
12 *
13 '
14 '
15 '
16 '
17 *
18 *
19 '
20 '
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
ANALYTE
1,1,2-Trichloroethane
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
Benzene
Brorrodichlororrethane
Brorroform
Carbon Tetrachloride
Chloroform
cis - 1,3 - Dichloropropene
EDB ( 1 ,2-Dibromoethane)
Bhylbenzene
Methylene chloride
Mlbt
mSVp-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1,3 - Dchloropropene
Vinyl Chloride
1,1 - Dichloropropene
1.1-Dichtoroethane
1.1-Dichloroethene
1 , 1 , 1-Trichloroethane
1,1,1,2- Tetrachloroethane
1 , 1 ,2.2-Tetrachloroethane
1 ,2-Dibromo-3-Chloropropane
1 ,2-Dichlorobenzene
1,2-Dichloroethane
1 ,2-Dichloropropane
1,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-D'chlorobenzene
1 ,3-Dichloropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chtorotoluene
Ally I Chloride
Bromobenzene
Brorroc hloromethane
Chlorobenzene
Chloroethane
Chlororrethane
cis-1 ,2-Dichloroethene
Dibromochloromethane
Dibromomethane
Dichlorofluoromethane
Di-bopropyl ether
Bhyl Bher
Hexachlorobutadiene
teopropylbenzene
n-Butylbenzene
n-Ropylbenzene
p-bopropy toluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1,2-Dchloroethene
Trichloroethene
Trichlorofluororrethane
Trichlorotrifluoroethane
7 Day
Absol.
650
604
611
586
711
582
707
714
705
653
605
614
643
1192
555
600
617
614
679
481
625
647
593
703
675
544
638
577
689
597
604
624
622
584
5/4
631
563
571
576
598
497
702
606
654
414
633
734
706
773
602
685
642
614
579
585
608
606
613
642
598
693
705
546
Roc %
85
78
79
76
92
76
92
93
92
85
79
80
84
77
72
78
80
80
88
63
81
84
77
91
88
71
83
75
90
78
78
81
81
76
75
82
73
74
75
78
65
91
79
85
54
82
95
92
100
78
89
83
80
75
76
79
79
80
83
78
90
92
71
7 Day
Absol
729
724
731
698
776
684
863
805
759
751
735
713
769
1426
708
713
720
707
757
585
764
747
741
881
798
667
816
679
802
662
728
733
700
686
684
721
665
686
694
/18
606
818
729
760
526
733
823
(78
820
696
792
730
740
/02
711
741
727
752
744
696
762
782
709
Roc %
95
94
95
91
101
89
112
105
99
98
96
93
100
93
92
93
94
92
98
76
99
97
96
115
104
87
106
88
104
86
95
95
91
89
89
94
86
89
90
93
79
106
95
99
68
95
107
101
107
90
103
95
96
91
92
96
94
98
97
90
99
102
92
7 Day
Absol.
702
661
661
663
790
659
834
776
743
728
677
690
731
1315
599
662
668
686
748
573
723
725
718
821
746
569
671
623
768
674
623
658
652
635
629
673
603
624
639
658
571
769
669
729
496
740
/89
/66
826
669
757
688
671
627
643
665
651
66 /
745
682
784
805
698
Rttc. %
98
751
86
86
103
86
108
101
97
95
88
90
95
85
78
86
87
89
97
74
94
94
93
107
97
74
87
81
100
88
81
85
85
83
82
87
78
81
83
85
74
100
87
95
64
96
103
100
107
87
98
89
87
82
84
86
85
87
97
89
102
105
91
7 Day
Absol
699
695
690
585
744
650
695
708
681
715
678
551
630
1328
625
667
667
660
726
379
628
597
528
698
762
640
774
657
714
607
690
693
661
668
643
668
640
658
542
69 1
442
694
676
513
327
621
ftiti
715
650
590
608
703
715
659
673
698
699
712
693
535
712
581
523
Rec %
91
90
90
76
97
84
90
92
88
93
88
72
82
86
81
87
87
86
94
49
82
78
69
91
99
83
101
85
93
79
90
90
86
87
84
87
83
86
70
91
57
90
88
67
43
81
102
93
85
77
79
91
93
86
87
91
91
92
90
69
93
75
68
7 Day
Absol
739
712
713
706
780
667
836
810
754
754
721
720
758
1426
675
725
712
718
741
583
750
747
718
863
777
597
719
668
809
662
687
735
693
677
669
692
6t>5
661
649
695
590
841
714
758
522
721
808
803
824
702
812
760
727
708
691
718
711
729
752
679
814
763
680
Rsc %
96
92
93
92
101
87
109
105
98
98
94
94
99
93
88
94
92
93
96
76
98
97
93
112
101
78
93
87
105
86
89
95
90
88
87
90
86
86
84
90
77
109
93
99
68
94
105
104
107
91
105
99
95
92
90
93
92
95
98
88
106
99
88
Absolute
AVG.
704
679
681
647
760
648
787
762
728
720
683
657
706
1337
632
673
677
677
730
520
698
692
659
793
751
603
724
641
756
640
666
688
665
650
639
677
627
640
620
673
541
765
679
683
457
689
788
753
779
652
731
705
693
655
660
686
679
694
715
638
753
727
631
AVG. % REC
of 0 Hour
90.7
90.5
90.4
85.1
92.9
93.9
91.8
90.0
91.6
88.6
88.1
J4.0
87.9
87.8
91.3
87.6
88.5
87.4
90.9
77.0
85.6
85.0
J4.0
90.5
91.6
104.8
90.1
902
89.8
864
90.2
93.2
93.8
92.0
J3.0
97.8
88.4
964
833
1049
75.4
86.5
890
88.0
728
85.9
92.7
89.6
91.9
85.9
89.3
96.2
89.6
91 6
892
91.2
90.3
896
88.1
79.6
50.5
95.1
92.8
RSD.
4.9
7.1
69
9.1
4.3
6.0
10.1
6.4
4.7
5.7
7.4
11.1
9.3
7.2
96
7.4
6.1
6.1
4.2
17.3
9.6
9.8
14.2
11.0
6.3
8.3
10.1
6.4
7.0
5.6
7.7
7.0
4.8
6.4
6.7
4.9
7.0
6.9
9.7
7.1
12.8
8.7
7.0
15.3
18.7
8.4
4.3
5.5
9.7
8.1
11.5
6.3
7.4
8.2
7.4
7.6
7.3
8.0
6.6
10.9
6.6
12.4
14.2
* Denotes analyte required to be added to this study by the WDNR and w hich rrust pass the imposed criteria to gain
acceptance for this method in Wisconsin.
147
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS
Matrix - Clean Laboratory Sand
VOC METHOD 8260 STANDARD 769.2 UG/KG /15.38 UG/L
Table 12
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
ANALYTE
1,1,2-Trichtoroethane
1,2,4-Trimethylbenzene
1,3,5-Trimethylbenzene
Benzene
BromodichloromBthane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1,3 - Dichloropropene
H3B (1,2-Dibrorrt>ethane)
Bhylbenzene
Methylene chloride
MTBE
rr&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1,3 - Dichloropropene
Vinyl Chloride
1,1 - Dichloropropene
1,1-Dichloroethane
1,1-Dichloroethene
1,1,1 -Trichloroethane
1,1,1,2 - Tetrachloroethane
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dibromo-3-Chloropropane
1 ,2-Dichlorobenzene
1,2-Dichloroethane
1 ,2-Dichloropropane
1,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Qchlorobenzene
1 ,3-Dichloropropane
1 ,4-Dchlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Allyl Chloride
Brorrobenzene
Bromochlororrethane
Chlorobenzene
Chloroethane
Chloromethane
cis-1 ,2-Dchloroethene
Dibromochloromethane
Dbromomethane
Dichlorofluoromethane
Di-bopropylether
Bhyl Bher
Hexachlorobutadiene
bopropylbenzene
n-Butylbenzene
n-Propylbenzene
p-lsopropytoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trichlorofluoromethane
Trichlorotrifluoroethane
10 Day
Absol
685
686
697
691
764
681
847
812
763
720
717
712
759
1415
564
717
712
698
785
609
759
743
761
906
785
662
693
652
816
641
621
655
639
648
691
629
648
877
685
599
663
807
701
842
564
741
827
799
992
678
818
702
697
673
671
701
691
703
767
710
692
919
669
R«c. %
89
89
91
90
99
89
110
105
99
94
93
92
99
92
73
93
92
91
102
79
99
97
99
118
102
86
90
85
106
83
81
85
83
84
90
82
84
114
89
/a
86
105
91
109
73
96
108
104
129
88
106
91
91
8/
87
91
90
91
100
92
90
119
8/
10 Day
Absol
m
697
712
704
/9/
700
901
849
772
752
735
741
814
1434
659
725
726
703
811
610
792
783
739
919
774
745
774
671
860
682
668
726
677
658
700
660
647
983
698
668
661
854
711
818
550
743
835
799
951
715
840
720
700
673
678
712
689
699
/86
/21
723
824
651
Rec %
93
91
92
92
104
91
117
110
100
98
96
96
106
93
86
94
94
91
105
79
103
102
96
119
101
9/
101
87
112
89
8/
94
88
86
91
86
B4
128
91
8/
86
111
92
106
72
97
109
104
124
93
109
94
91
B/
88
93
90
91
102
94
94
W
85
10 Day
Ateol
682
658
6/9
677
767
655
893
795
725
681
696
667
722
1371
606
692
690
691
748
595
768
723
727
885
765
647
650
631
765
627
606
664
674
618
669
623
647
917
666
616
649
7b7
680
771
499
715
766
730
901
663
776
763
693
663
663
682
6/2
694
/32
6/3
681
934
744
Rec %
98
751
88
88
100
85
116
103
94
89
90
8/
94
89
79
90
90
90
97
77
100
94
94
115
99
84
84
82
99
81
79
86
88
80
87
81
84
119
87
80
84
98
88
100
65
93
100
95
117
86
101
99
90
86
86
89
87
90
95
87
88
121
97
10 Day
Absol
662
626
648
634
720
579
822
746
688
670
6b3
659
682
1294
583
659
649
651
701
560
699
699
686
806
699
657
668
607
759
594
604
633
605
603
618
b8b
601
853
629
575
609
/39
644
743
486
666
752
725
852
632
730
684
645
626
627
658
63 /
658
691
629
623
834
6/4
Rec. %
86
81
84
82
94
/b
107
97
89
87
85
86
89
84
76
86
84
85
91
73
91
91
89
105
91
Bb
87
79
99
77
79
82
79
78
80
76
78
111
82
75
79
96
84
97
63
87
98
94
111
82
9b
89
84
81
81
85
83
86
90
82
81
108
88
10 Day
Absol
6/2
6b3
662
6/3
746
628
883
/98
719
692
691
684
/b1
1346
603
700
674
686
/33
612
768
/33
/b2
86 /
745
6bb
724
64b
795
609
640
6/9
651
638
649
621
631
888
666
635
633
799
690
810
528
710
774
730
925
665
795
731
681
64/
64/
673
66b
6/3
/26
6/8
683
870
717
Rec %
87
85
86
8/
97
82
115
104
93
90
90
89
98
87
78
91
88
89
95
80
100
95
98
113
97
85
94
84
103
79
83
88
85
83
84
81
82
115
87
82
82
104
90
105
69
92
101
95
120
86
103
95
88
84
84
87
86
87
94
8«
89
113
93
Absolute
AVG.
683
664
679
676
759
648
869
800
733
703
698
692
746
1372
603
698
690
686
755
597
757
736
733
876
754
673
701
641
799
630
628
671
649
633
665
623
635
904
669
618
643
791
685
797
525
715
791
756
924
670
792
720
683
656
657
685
671
685
740
682
680
876
691
AVG % REC
of 0 Hour
88.0
88.6
90.2
888
927
94.0
101.4
944
92.3
86.5
90.1
88.5
92.8
90.1
87 1
909
90.2
88.5
940
88.4
928
903
934
100.0
91,9
1169
87.4
90.3
94.8
85.0
850
90.9
91 5
896
86.4
90.0
89,4
360
898
964
807
89.5
89.8
102.7
83.6
891
930
899
109.1
88.4
967
98.3
88.2
91 7
88.7
91 1
893
88 4
91.2
851
Til
1146
101.5
%
RSD.
2.8
4.2
3.8
3.9
3.7
7.3
3.9
4.6
4.7
4.7
4.4
4.9
6.5
4.1
5.9
3.7
4.4
3.0
5.7
3.7
4.6
4.2
4.0
5.1
4.5
6.0
7.0
3.8
5.2
5.4
4.3
5.2
4.5
3.5
5.0
4.3
3.2
5.5
3.9
5.7
3.5
5.7
3.8
4.9
6.3
4.4
4.8
5.1
5.7
4.5
5.3
4.2
3.3
3.1
3.1
3.2
3.3
2.8
5.0
5.3
5.3
5.6
5.5
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
* Denotes analyte required to be added to this study by the WDNR and w hich must pass the imposed criteria to gain
acceptance for this method in Wisconsin.
148
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Rep. #1
Rep. #2
Rep. #3
Rep. #4
Rep. #5
Average
% Recovery of 0 Hour.
% RSD
Table 13
GASOLINE STANDARD (153.8) MG/KG or 3077 UG/L
Matrix - Clean Laboratory Sand
Unspiked
Absolute
0.21
0.4
NR
NR
NR
%Rec.
NA
MA
NA
NA
NA
NA
NA
NA
0 Hour
Absolute
95.9
94.8
97.4
98.2
95.6
96.3
%Roc.
62.3
61.6
63.3
63.8
62.2
62.6
NA
1.4
24 Hour
Absolute
98.2
100
100
102
102
100
%Rec.
63.8
65.1
65.3
66.2
66.1
65.3
104
1.5
48 Hour
Absolute
99.4
101
101
103
103
101
%Rec.
64.6
65.7
65.8
66.7
66.9
65.9
105
1.4
72 Hour
Absolute
97.4
98.9
102
101
99.6
99.6
%Rec.
63.3
64.3
66.2
65.4
64.7
64.8
104
1.7
7 Day
Absolute
97.3
97.7
96.8
101
98.4
98.2
% Rec
63.3
63.5
62.9
65.7
64.0
63.9
102
1.7
10 Day
Absolute
95.6
97.8
91.6
93.3
95.8
94.8
%Rec
62.1
63.6
59.5
60.6
62.3
61.6
98
2.6
x./users/mike/agasvial
NR = Not Run
NA = Not Applicable
GASOLINE STANDARD (153.8) MG/KG or 3077 UG/L
Matrix - Biologically Active Garden Soil
Rep.#1
Rep. #2
Rep. #3
Rep. #4
Rep. #5
Average
% Recovery of 0 Hour.
% RSD
Unspiked
Absolute
0.2
NR
NR
NR
NR
%Rec *
NA
NA
NA
NA
NA
NA
NA
NA
0 Hour
Absolute
87.0
85.7
84.2
87.8
86.2
86.2
% Rec *
70.8
69.7
68.5
71.5
70.2
70.1
NA
1.6
24 Hour
Absolute
84.6
82.2
82.9
85.2
83.9
83.8
% Rec *
68.9
66.9
67.5
69.3
68.3
68.2
97.3
1.5
48 Hour
Absolute
85.8
61.5
87.1
86.6
82.1
80.6
%Rec '
69.8
50.1
70.9
70.4
66.8
65.6
93.6
13.4
72 Hour
Absolute
83.2
87.5
83.8
87.4
85.2
85.4
% Rec. '
67.7
71.2
68.2
71.2
69.3
69.5
99.2
2.3
7 Day
Absolute
84.4
85.8
84.5
84.9
85.9
85.1
% Rec *
68.7
69.9
68.8
69.1
69.9
69.3
988
0.9
10 Day
Absolute
80.3
80.3
81.3
82.8
81.5
81.3
% Rec •
65.4
65.4
66.1
67.4
66.3
66.1
94.3
1.3
* % Recoveries were corrected for moisture using a
dry weight value of 79.9 %. ( See Table 14 ).
X/users/Mike/bngasvia
NR = Not Run
NA = Not Applicable
149
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS
Matrix - Biologically Active Garden Soil
VOC METHOD 8260 STANDARD 153.8 UG/KG / 3.08 UG/L
Table 14
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
ANALVTE
1,1,2-Trichloroethane
1 ,2,4-Trimethylbenzene
1 ,3,5-Trirrethylbenzene
Benzene
Bromodichloromethane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1 ,3 - Dichloropropene
SB (1,2-Dibromoethane)
Hhylbenzene
Methylene Chloride
MTBE
rr&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1 ,3 - Dichloropropene
Vinyl Chloride
1,1 - Dichloropropene
1,1-Dichloroethane
1,1-Dichloroethene
1 , 1 , 1-Trichloroethane
1,1,1,2 - Tetrachloroethane
1 , 1 ,2,2-Tetrachloroethane
1 ,2-DibromD-3-Chloropropane
1 ,2-Dichlorobenzene
1,2-Dichloroethane
1 ,2-Dichloropropane
1,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dichlorobenzene
1 ,3-Dichloropropane
1 ,4-Dchlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Allyl Chloride
Bromobenzene
Bronnochlororrethane
Chlorobenzene
Chloroethane
Chloromethane
cis-1 ,2-Dichloroethene
Qbromochloromethane
Dibromo methane
Dichlorofluoromethane
Di-isopropyl Bher
Bhy I Hher
Hexachlorobutadiene
bopropylbenzene
n-Butylbenzene
n-R-opylbenzene
p-teopropyltoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trichlorofluoromethane
Trichlorotrifluoroethane
0 Hour
Absol
133
130
134
136
123
115
116
133
127
131
134
130
136
268
109
134
125
132
127
166
126
131
119
127
121
131
143
133
135
132
138
119
112
127
134
128
131
118
133
128
129
132
127
182
16/
134
114
131
199
135
140
138
135
131
132
130
132
133
128
12b
143
1/0
1/5
Rec %
86
84
87
88
80
75
75
86
83
85
87
85
88
87
71
87
81
86
83
108
82
85
77
82
78
85
93
86
87
86
90
77
73
83
87
83
85
76
86
83
84
86
83
118
108
87
74
85
129
87
91
89
88
85
86
85
86
86
83
81
93
111
113
0 Hour
Absol.
126
119
123
129
112
102
122
122
115
121
123
123
128
249
100
123
115
123
117
106
123
125
130
124
115
118
115
121
123
122
124
110
106
122
122
122
121
109
125
123
119
117
123
124
93
128
103
119
103
128
126
122
125
123
12b
119
123
123
124
125
135
114
10b
Rec %
82
77
80
84
73
66
/9
79
75
fti
80
80
83
81
65
80
75
80
76
69
80
81
84
81
75
77
74
78
80
79
81
72
69
79
79
79
78
71
81
80
77
76
80
80
60
83
67
77
67
83
82
79
81
80
81
77
80
80
81
81
87
74
68
0 Hour
Absol.
144
136
138
140
130
111
126
13/
126
132
139
143
146
2/5
111
13/
128
145
132
106
138
145
128
136
128
121
132
135
139
139
143
125
119
135
138
131
133
120
135
134
131
130
133
122
103
140
113
13/
11/
144
143
140
138
140
138
139
138
137
135
13/
153
10/
9/
Rec. %
94
8/
90
91
84
72
82
89
82
86
90
93
95
89
12
89
83
94
86
69
90
94
83
88
83
79
86
88
90
90
93
81
77
87
89
85
86
78
87
87
85
84
86
79
67
91
/3
89
/6
93
93
91
90
91
90
90
90
89
87
89
99
/O
63
0 Hour
Absol
138
134
138
138
129
108
120
139
124
136
137
138
145
278
112
136
131
136
127
103
133
136
133
132
127
129
125
134
135
140
133
123
121
137
138
131
13/
119
137
137
133
137
134
106
101
142
116
138
106
143
143
130
136
133
135
134
135
13/
138
134
144
10/
98
Rec. %
90
87
89
90
84
70
78
90
81
88
89
90
94
90
73
88
85
88
83
67
86
88
86
86
83
84
81
87
88
91
86
80
79
89
89
85
89
77
89
89
86
89
87
69
66
92
75
89
69
93
93
85
88
86
88
87
88
89
90
87
94
69
63
0 Hour
Absol
130
138
138
139
125
109
128
138
125
132
134
141
133
2/5
103
133
126
136
130
105
138
139
132
136
129
114
122
134
135
132
134
115
119
133
134
131
136
118
135
134
132
136
137
121
101
138
117
133
104
141
139
132
142
132
139
135
137
139
132
142
149
112
100
Rec %
86
89
89
90
81
71
83
89
81
86
87
91
86
89
67
86
82
88
84
68
89
90
86
88
84
74
79
87
87
86
87
75
77
86
87
85
88
77
88
87
86
88
89
78
66
89
76
86
67
91
90
86
92
86
90
88
89
90
86
92
9/
73
65
Absolute
AVG.
134
131
134
136
124
109
122
133
123
130
133
135
137
269
107
132
125
134
126
117
132
.135
128
131
124
122
127
131
133
133
134
118
115
131
133
128
131
117
133
131
128
130
131
131
113
136
112
131
125
138
138
132
135
132
134
131
133
134
131
132
145
122
115
Average
% REC. **
109.1
106.5
109.0
110.8
1005
88.6
99.5
108.6
100.3
105.9
108.3
109.6
111.8
109.4
86.9
107.8
101.5
109.3
102.8
95.1
107.0
110.0
104.2
106.5
100.7
99.6
103.4
1068
108.2
108.0
109.2
96.2
93.8
106.2
108.2
1043
106.9
94.8
108.2
106.7
1045
106.0
106.2
106.4
91 8
110.7
91.5
106.9
102.0
112.1
112.2
107.6
110.0
107.0
108.7
106.8
108.1
1087
106.9
1076
117.7
99.2
93.3
%
RSD.
5.2
5.5
4.7
3.1
5.9
4.4
4.1
5.2
3.8
4.4
4.7
6.2
5.5
4.4
4.9
41
4.7
5.9
4.6
23.3
5.1
5.5
4.4
4.1
4.8
5.7
8.4
4.6
4.6
5.4
51
5.0
5.5
4.8
4.8
3.1
4.9
37
3.5
4.3
4.3
6.2
4.3
226
268
42
49
56
32.9
49
50
53
4.6
4.5
4.2
57
4.7
4.7
4.2
5.7
47
222
29.3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
* Denotes analyte required to be added to this study by the WDNR and w hich must pass the imposed criteria to gain
acceptance for this method in Wisconsin.
** Recoveries have been corrected for % moisture using a dry w eight value of 79.9%. x te,rs;
150
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS Table 15
Matrix - Biologically Active Garden Soil
VOC METHOD 8260 STANDARD 153.8 UG/KG / 3.08 UG/L
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
ANALYTE
1 ,1 ,2-Trichloroethane
1 ,2,4-Trimethylbenzene
1 ,3,5-TrinBthylbenzene
Benzene
Bromodichloromethane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1 ,3 - Dichloropropene
H3B (1.2-Dibromoethane)
Bhylbenzene
Methylene Chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1,3 - Dichloropropene
Vinyl Chloride
1,1 - Dichloropropene
1,1-Dichloroethane
1,1-Dichloroethene
1,1,1 -Trichloroethane
1 , 1 , 1 ,2 - Tetrachloroethane
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dibromo-3-Chloropropane
1 ,2-Dichlorobenzene
1,2-Dichloroethane
1 ,2-Dichloropropane
1 ,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dichlorobenzene
1 ,3-Dchloropropane
1,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Ally I Chloride
Bromobenzene
Brorrochloromethane
Chtorobenzene
Chloroethane
Chloromethane
cis-1 ,2-Dchloroeihene
Dibromochloromethane
Dibromomethane
Dichlorofluoromethane
Di-isopropyl Bher
Bhyl Bher
Hexachlorobutadiene
teopropylbenzene
n-Butylbenzene
n-Fropylbenzene
p-bopropyltoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trichlorofluoromethane
Trichlorotrifluoroethane
24 Hour
Absol
135
127
132
124
123
102
116
128
45
118
122
124
140
254
110
134
100
122
81
89
133
131
111
127
121
134
121
129
133
133
139
122
116
128
133
124
126
106
125
109
127
127
127
113
77
131
106
138
99
138
134
133
132
111
122
113
127
128
130
120
140
92
95
Rue %
88
83
86
81
80
66
75
83
29
77
79
81
91
82
71
87
65
79
52
58
86
85
72
83
79
87
79
84
86
86
90
79
75
83
86
81
82
69
81
71
82
82
83
73
50
85
69
89
64
90
87
86
86
72
79
73
82
83
84
78
91
60
61
24 Hour
Absol
135
131
133
128
124
103
119
139
49
119
128
129
137
255
105
132
101
122
83
97
130
135
117
126
121
116
121
131
130
132
133
122
112
129
134
125
129
108
128
105
131
126
130
104
82
131
110
133
106
136
134
136
131
11b
124
114
128
133
129
122
144
96
82
Rec %
87
85
86
83
81
67
77
90
32
77
83
84
89
83
68
86
65
79
54
63
84
88
76
82
78
75
79
85
84
86
86
79
73
84
87
81
84
70
83
68
8b
82
84
68
53
85
71
86
69
88
87
88
85
75
81
74
83
86
84
79
93
62
53
24 Hour
Absol.
128
125
126
122
114
99
120
133
43
114
120
125
131
241
109
125
95
118
76
101
123
127
117
127
119
115
110
123
123
129
130
127
113
122
127
121
124
102
121
99
121
122
125
119
83
125
103
137
104
132
129
120
126
109
117
111
122
127
124
123
141
106
101
Rec %
83
81
82
79
74
64
78
86
28
74
78
81
85
78
71
81
61
77
49
66
80
82
76
83
77
74
72
80
80
84
84
83
73
79
82
78
81
66
79
64
79
79
81
77
54
81
67
89
67
86
84
78
82
71
76
72
79
82
81
80
92
69
66
24 Hour
Absol
128
125
129
124
117
97
113
131
41
111
118
121
134
243
107
122
96
114
77
86
114
130
109
124
118
109
117
12/
128
129
127
121
112
126
12/
121
124
101
120
92
123
124
123
87
80
121
103
134
96
131
123
122
125
1U8
117
108
121
125
118
118
138
96
85
Rec %
83
81
84
80
76
63
73
85
27
72
76
78
87
79
70
79
62
74
50
56
74
84
71
80
76
71
76
82
83
84
83
79
73
82
82
79
80
65
78
60
80
80
80
57
52
79
67
87
62
85
80
79
81
70
76
70
78
81
76
76
90
62
55
24 Hour
Absol
130
125
132
126
126
98
115
135
38
120
123
124
135
248
107
128
10U
120
79
91
129
135
118
125
124
112
123
129
129
134
134
120
111
126
129
121
128
102
124
100
125
127
128
111
80
127
106
133
108
132
133
121
129
110
121
111
126
130
124
124
147
103
91
Rec %
84
81
86
82
82
64
75
88
24
78
80
81
88
81
69
83
65
78
51
59
84
87
76
81
81
73
80
84
84
87
87
78
72
82
84
79
83
66
81
65
81
83
83
72
52
82
69
86
70
86
86
79
84
71
79
72
82
84
80
81
96
67
59
Absolute
AVG.
131
126
130
125
121
100
116
133
43
116
122
124
135
248
107
128
98
119
79
93
126
131
114
126
120
117
118
128
128
131
132
122
113
126
130
122
126
104
123
101
125
125
126
107
80
127
105
135
102
134
130
126
128
110
120
111
125
128
125
121
142
98
91
AVG. % REC
of 0 Hour
97.7
96.6
97.2
91.5
97.6
91.4
95.2
99.8
348
894
916
923
984
923
1004
96.7
787
88.7
624
792
95.4
97.1
89.1
96.1
973
95.6
93.1
97.2
965
98.8
98.5
103.6
977
96.6
97.6
954
95.9
88.9
928
77.0
975
95.9
968
81 6
708
93.1
93.7
102.7
81 7
97.1
94.6
955
950
84.0
89.7
850
93.8
96.1
949
91.7
98 1
30.7
791
%
RSD.
2.7
2.0
22
1.7
4.2
2.4
2.3
3.1
9.7
3.2
3.1
2.3
2.7
2.6
1.8
37
27
30
3.5
6.7
6.1
2.7
37
1.2
2.0
8.4
44
25
2.8
1.7
3.3
22
1.6
2.0
2.8
1.7
1 7
3.1
2.4
64
31
1 7
2.0
11 4
29
32
28
1 6
49
2.4
3.6
59
2.3
25
2.5
2.1
2.6
24
38
2.2
2.5
56
8.3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
* Denotes analyte required to be added to this study by the WDNR and w hich must pass the imposed criteria to gain
acceptance for this method in Wisconsin.
151
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS
Matrix - Biologically Active Garden Soil
VOC METHOD 8260 STANDARD 153.8 UG/KG / 3.08 UG/L
Table 16
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
ANALYTE
1,1,2-Trichloroethane
1 ,2,4-Trirrethylbenzene
1 ,3,5-Trirrethylbenzene
Benzene
BromDdichloromethane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1,3 - Dichloropropene
BDB (1,2-Dbromoethane)
Hhylbenzene
Methylene Chloride
MTBE
rr&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1 ,3 - Dichloropropene
Vinyl Chloride
1,1 - Dichloropropene
1,1-Dichloroethane
1,1-Dichloroethene
1,1,1 -Trichloroethane
1,1,1,2 - Tetrachloroethane
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dibromo-3-Chloropropane
1 ,2-Dichlorobenzene
1,2-Dichloroethane
1 ,2-Dichloropropane
1,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dichlorobenzene
1 ,3-Dichloropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Allyl Chloride
Brorrobenzene
Brorrochloromethane
Chlorobenzene
Chloroethane
Chlororrethane
cis-1,2-Dichloroethene
Dbrorrochloromethane
Dibromorrethane
Dichlorofluoromethane
D-isopropyl Hher
Bhyl Hher
Hexachlorobutadiene
bopropylbenzene
n-Burylbenzene
n-FVopylbenzene
p-lsopropy toluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trie hlorof luoromethane
Trichlorotrifluoroethane
48 Hour
Absol.
127
130
134
122
120
92
117
136
21
114
117
127
139
244
114
128
87
118
62
89
114
127
112
130
119
120
131
129
128
130
132
129
120
130
131
128
127
96
121
99
123
132
125
108
68
128
109
145
105
132
134
137
128
105
116
111
126
129
123
124
139
100
86
Roc %
82
84
87
79
/8
60
76
88
13
74
76
83
90
79
74
83
57
76
40
58
74
82
73
84
77
78
85
84
83
84
86
84
ftt
84
85
83
82
62
(•9
64
80
86
81
/O
44
83
n
94
68
86
87
89
83
b8
/b
t2
82
84
80
81
90
6b
bb
48 Hour
Absol
129
119
129
120
118
97
114
128
33
109
118
123
129
242
104
123
85
118
65
93
114
129
113
124
122
123
123
125
127
128
130
123
105
123
128
121
126
91
119
97
119
121
123
109
71
123
103
138
109
130
124
122
124
101
112
103
119
124
122
123
129
104
92
Rec %
84
77
84
/8
76
63
/4
83
21
71
76
80
84
79
67
80
bb
77
42
60
M
84
73
80
79
80
80
81
82
83
84
80
68
8U
83
IV,
82
59
77
63
77
78
80
71
46
80
67
90
71
85
80
/y
80
6b
73
67
77
80
79
BU
84
67
60
48 Hour
Absol.
133
124
125
123
119
91
119
132
30
116
120
126
133
248
100
127
87
119
62
97
123
133
118
131
119
109
122
127
132
128
133
117
108
124
130
125
121
93
121
104
125
126
126
118
76
127
104
145
102
135
131
126
126
102
113
103
123
128
127
128
146
109
99
Rec. %
86
81
81
80
77
b9
77
86
19
/b
78
82
86
81
6b
83
56
77
40
63
80
86
77
85
77
71
79
83
86
83
86
76
70
81
84
81
79
60
79
68
81
82
82
76
49
83
6/
94
66
88
8b
82
82
66
73
67
80
83
83
83
95
71
64
48 Hour
Absol
136
126
128
122
120
97
118
134
24
119
119
129
137
245
101
131
85
119
65
85
113
128
107
130
122
122
118
129
128
128
134
117
106
124
137
124
125
90
123
94
129
130
128
119
67
126
105
139
107
132
129
123
126
100
113
102
122
127
127
119
134
98
94
Rec %
88
82
83
79
78
63
76
87
16
77
11
84
89
79
66
8b
bb
77
42
55
73
83
69
84
79
7y
76
84
83
83
at
76
69
81
89
80
81
59
80
61
84
85
83
77
43
82
68
90
69
86
84
80
82
6b
73
66
79
82
82
n
87
64
61
48 Hour
Absol
129
123
128
122
120
y3
115
129
32
116
116
131
134
243
100
129
87
119
66
87
107
124
105
128
125
108
108
125
127
129
127
118
111
126
128
125
125
89
123
99
127
126
126
111
74
128
105
136
105
130
129
119
12b
101
114
102
120
126
128
122
139
93
86
Rec. %
84
80
83
79
78
60
/b
84
20
75
7b
8b
87
79
6b
84
b/
77
43
b6
70
80
68
83
81
70
70
81
83
84
82
77
72
82
83
81
81
58
80
64
83
82
82
72
48
83
68
88
68
8b
84
77
81
6b
74
66
78
82
83
/9
90
60
56
Absolute
AVG.
131
124
129
122
119
94
116
132
28
115
118
127
134
244
104
127
86
118
64
90
114
128
111
128
121
116
120
127
128
128
131
121
110
125
131
124
124
92
121
98
124
127
126
113
71
126
105
141
105
132
129
125
125
101
114
104
122
126
125
123
137
101
91
AVG % REC
of 0 Hour
97.3
95.0
96.1
89.5
96.5
86.0
95.2
98.7
22.4
88.1
88.6
94.4
97.5
90.8
97.0
96.2
59.0
88.2
50.5
77.0
867
94.6
865
97.9
97.9
94.9
94.4
96.7
96.4
96.8
97.6
101.9
95.1
96.0
98.3
96.8
94.7
78.8
91.1
75.0
96.9
97.3
96.2
86.3
32.7
92.8
931
106.9
54.0
95.6
93.7
94.7
92.8
77.1
85.0
79.3
91.7
94.6
95.2
93.1
94.8
32.5
79.6
%
RSD.
2.7
3.1
2.6
0.9
0.9
2.8
1.7
2.4
18.7
3.1
1.4
2.3
2.9
1.0
5.5
2.3
1.2
0.6
3.1
5.6
5.0
2.6
4.7
2.2
2.2
6.3
7.0
1.7
1.7
0.6
2.3
4.2
5.5
2.1
2.7
2.0
1.7
3.0
1.4
3.9
3.0
35
1.4
4.5
5.4
1.8
2.2
29
2.5
16
2.9
5.6
1.2
1.8
1.3
3.8
23
1.6
2.2
2.8
4.6
5.8
6.0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
* Denotes analyte required to be added to this study by the WDNR and w hich mist pass the irrposed criteria to gain
acceptance for this method in Wisconsin
152
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS Table 17
Matrix - Biologically Active Garden Soil
VOC METHOD 8260 STANDARD 153.8 UG/KG / 3.08 UG/L
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
1 *
2 *
3 *
4 *
5 *
6 *
7 '
8 *
9 '
10 *
11 *
12 *
13 '
14 *
15 *
16 *
17 *
18 '
19 *
20 '
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
ANALYTE
1,1,2-Trichloroethane
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
Benzene
Brorrodichloromethane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1,3 - Dichloropropene
HDB (1,2-Dibromoethane)
Bhylbenzene
Methylene Chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1,3 - Dichloropropene
Vinyl Chloride
1,1 - Dichloropropene
1,1-Dichloroethane
1,1-Dichloroethene
1 , 1 , 1-Trichloroethane
1,1,1.2 - Tetrachloroethane
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dibromo-3-Chloropropane
1 ,2-Dichlorobenzene
1,2-Dichloroethane
1 2-Pchloropropane
1,2,3 - Trichloropropane
1,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dichlorobenzene
1 ,3-Dichloropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Allyl Chloride
Bromobenzene
Bromochlororrethane
Chlorobenzene
Chloroethane
Chloromethane
cis-1 ,2-Dichloroethene
Dibromochlororrethane
Dibromorrethane
Dichlorofluoromethane
Di-isopropyl Bher
Shyl Bher
Hexachlorobutadiene
bopropylbenzene
n-Butylbenzene
n-FVopylbenzene
p-teopropyKoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trichlorofluororrethane
Trichlorotrifluoroethane
72 Hour
Absol
133
128
138
128
119
91
123
137
11
10/
117
134
141
244
102
127
80
118
56
89
114
132
121
136
126
139
119
132
131
131
139
120
111
131
131
125
134
13/
123
87
129
127
12^
121
62
127
106
143
105
136
138
132
131
yy
114
107
128
134
130
123
136
113
108
Roc %
86
83
89
83
77
59
80
89
7.2
70
76
87
91
79
66
82
52
76
36
58
74
86
78
88
82
90
77
86
85
85
90
78
72
85
85
81
87
89
80
57
84
83
82
79
40
83
69
93
68
88
89
86
85
64
74
69
83
87
84
80
88
73
70
72 Hour
Absol
130
128
130
128
124
88
122
140
19
116
119
123
137
242
101
130
78
122
61
99
132
138
123
136
124
135
112
130
132
136
136
122
106
132
131
128
129
137
123
92
132
132
131
127
67
125
108
141
122
140
130
136
130
99
115
100
124
131
133
126
134
117
105
Rec %
84
83
84
83
80
57
79
91
12
75
77
80
89
79
65
85
50
79
39
64
86
90
80
88
81
87
73
85
86
88
88
79
69
86
85
83
84
89
80
59
86
86
85
82
43
81
70
91
79
91
85
88
84
64
75
65
80
8b
86
82
tif
76
68
72 Hour
Absol
123
126
134
124
121
89
12b
135
18
108
120
130
131
242
y4
128
/2
120
55
91
115
133
120
134
123
133
111
131
129
134
130
115
104
128
130
124
123
135
123
y/
131
133
127
115
61
128
103
141
114
134
129
127
129
y/
120
100
126
132
132
123
126
114
102
Ree %
80
82
87
80
79
58
81
8/
12
70
78
84
85
79
61
83
46
78
36
59
75
86
78
87
80
86
72
85
84
87
85
75
67
83
84
80
80
87
80
63
85
86
83
75
40
83
67
92
74
87
84
82
84
63
78
65
82
86
86
80
82
74
66
72 Hour
Absol.
136
125
135
126
127
93
125
140
19
113
122
133
135
249
104
133
77
121
62
98
121
139
121
139
123
13/
123
131
13b
13/
132
120
110
128
131
126
134
137
126
97
134
127
131
121
67
132
108
141
112
139
134
130
132
101
118
102
126
133
129
134
137
119
110
Rec %
88
81
87
82
83
60
81
91
12
73
79
86
88
81
67
86
50
79
40
64
79
90
78
90
80
89
80
85
87
89
86
78
72
83
85
82
87
89
82
63
87
83
85
79
43
86
70
92
73
90
87
84
86
65
76
66
82
86
84
87
89
77
71
72 Hour
Absol
137
12b
134
126
127
90
126
140
15
10b
121
134
133
248
y/
131
76
12b
56
91
120
130
122
140
124
127
120
132
129
132
132
118
109
128
134
126
133
133
126
87
128
126
132
128
64
129
109
142
111
134
134
134
132
97
116
100
125
133
139
129
138
121
103
Rec %
89
81
87
82
82
58
82
91
9.4
68
78
87
86
81
63
8b
49
81
36
59
78
85
79
91
80
83
78
86
84
86
86
76
71
83
87
82
86
86
82
56
83
82
86
83
42
84
71
92
72
87
87
87
86
63
75
65
81
86
90
84
90
78
67
Absolute
AVG.
132
126
134
126
123
90
124
138
16
110
120
131
135
245
99
130
76
121
58
94
120
134
121
137
124
134
117
131
131
134
133
119
108
129
131
126
131
136
124
92
130
129
129
122
64
128
106
141
113
136
133
131
130
98
116
102
125
132
132
127
134
117
105
AVG. % REC
of 0 Hour
98.1
96.3
100.0
92.6
99.8
32.6
101.5
103.5
13.1
34.2
89.9
96.9
98.2
91.0
92.9
97.8
31.2
89.9
$5.8
80.0
91.6
99.2
94.5
104.7
100.1
109.4
91.8
99.8
98.4
100.7
99.4
100.3
936
988
98.8
98.0
99.4
116.3
93.2
39.9
101 5
98 9
99.1
935
56.6
94.0
94.7
107.5
89.8
99.0
96.4
994
965
747
87 1
77.4
944
99 1
100.8
960
92.7
956
91 8
RSD.
4.5
1.2
2.1
1 4
29
2.1
1 5
1.7
20.5
40
1.6
3.5
2.8
1.4
4.0
1.8
3.9
2.1
54
4.9
5.9
29
1.1
1.7
1.0
3.3
4.5
0.6
1.9
1.9
2.6
2.1
2.8
1.6
1.3
1.2
3.7
1.5
12
53
1.8
2.4
19
41
4.1
2.0
2.1
0.5
5.3
20
26
2.7
0.9
1 6
1.9
29
1.1
0.8
3.1
37
3.5
2.8
32
* Denotes analyte required to be added to this study by the WDNR and w hich rmst pass the Imposed criteria to gain
acceptance for this method in Wisconsin.
153
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS
Matrix - Biologically Active Garden Soil
VOC METHOD 8260 STANDARD 153.8 UG/KG / 3.08 UG/L
Table 18
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
ANALYTE
1 , 1 ,2-Trichloroethane
1 ,2,4-Trimethylbenzene
1 3,5-Trimethylbenzene
3enzene
Brorrodichloromethane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1.3 - Dichloropropene
H3B(1,2-Dibrorroethane)
Bhylbenzene
Methylene Chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1,3 - Dichloropropene
Vinyl Chloride
1,1 - Dchloropropene
1,1-Qchloroethane
1,1-Dchloroethene
1, 1,1 -Trie hloroethane
1,1,1,2 - Tetrachloroethane
1 , 1 ,2,2-Tetrachloroethane
,2-Dibromo-3-Chloropropane
,2-Dichlorobenzene
,2-Dichloroethane
,2-Dichloropropane
,2,3 - Trichloropropane
,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dichlorobenzene
1 ,3-Dichloropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Allyl Chloride
Bromobenzene
BromDchloromethane
Chlorobenzene
Chloroethane
Chloromethane
cis-1 ,2-Dichloroethene
Dibromochloromethane
DibrorrDmethane
Dichlorofluoromethane
Q-isopropyl Ether
Hhy Bher
Hexachlorobutadiene
bopropylbenzene
n-Butylbenzene
n-FVopy (benzene
p-bopropyltoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trichlorofluoromethane
Trichlorotrifluoroethane
7 Day
Absol
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
omtte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritte
orritta
orritte
orritte
orritta
orritte
orritte
orritte
orritte
orritte
oniltt
orritte
orritte
orritte
Rec %
d
d
d
d
d
d
d
d
d
d
d
d
d
d
1
J
1
i
1
J
j
d
j
d
J
J
d
d
d
d
J
J
i
J
d
J
J
d
J
d
J
d
J
d
d
d
d
j
J
J
J
d
J
j
J
J
j
i
j
j
J
i
i
7 Day
Absol
136
119
129
124
119
77
121
141
12
105
111
12/
133
232
98
127
51
116
44
91
127
135
114
133
124
136
115
127
127
131
134
115
107
127
130
125
126
127
117
85
125
127
127
120
44
129
102
154
120
132
125
131
125
79
105
86
117
133
130
122
129
108
100
Rec %
88
77
84
80
ft
50
78
91
/.b
68
/2
83
86
fi>
63
82
33
75
28
59
82
88
74
86
81
88
/b
83
82
85
87
75
fO
83
85
81
82
83
76
bb
81
82
83
78
28
84
66
100
/8
86
81
8b
81
b1
68
56
fti
86
85
/9
84
to
65
7 Day
Absol
130
116
129
122
122
72
124
137
4
93
109
121
131
229
100
123
41
114
27
91
134
130
117
133
125
138
121
130
134
131
135
117
104
128
131
123
126
125
117
68
125
129
126
112
40
124
101
1bb
121
132
130
120
12b
77
101
78
11b
126
134
127
133
123
109
Rec %
84
75
84
79
/9
46
80
89
2.3
60
/1
/8
85
f4
65
80
26
74
17
59
8/
85
/6
86
81
89
t6
84
87
85
87
/6
68
83
85
80
82
81
fti
44
81
84
82
73
26
81
65
101
/9
86
84
78
81
50
66
50
74
82
tif
82
86
80
71
7 Day
Afasol
133
115
132
121
126
70
127
140
2
89
110
130
130
229
91
123
40
120
25
94
130
139
122
138
123
13/
121
127
130
131
137
112
102
131
129
124
129
122
118
64
127
131
127
130
41
130
104
16b
124
134
131
12/
126
/b
101
/6
115
130
128
127
124
117
104
Rec. %
86
75
86
79
82
46
82
91
1.3
58
72
84
85
74
59
80
26
fU
16
61
85
90
79
89
80
89
78
82
85
85
89
73
66
85
84
80
84
79
77
42
83
85
83
84
27
84
68
10/
80
87
85
82
82
48
65
49
75
84
83
82
81
76
68
7 Day
Absol
124
112
126
119
119
75
120
131
b
94
109
122
12/
224
86
121
40
114
30
83
131
124
112
128
121
134
108
124
131
128
130
109
99
128
126
120
127
120
117
/5
123
128
124
111
38
125
101
148
111
130
124
112
122
74
99
/6
111
123
126
117
131
114
101
Rec %
81
73
82
77
77
49
78
85
2.9
61
71
/9
83
73
56
79
26
74
19
54
85
80
73
83
79
87
70
80
85
83
84
71
64
83
82
78
83
78
76
48
80
83
80
72
24
81
66
96
72
85
81
72
/9
48
64
49
72
80
82
76
85
74
65
Absolute
AVG.
130
115
129
121
121
73
123
137
5
95
110
125
130
229
93
123
37
116
31
90
130
132
116
133
123
136
116
127
130
130
134
113
103
128
129
123
127
124
117
73
125
129
126
118
41
127
102
155
119
132
127
122
124
76
101
79
114
128
129
123
129
115
103
AVG % REC
Of 0 Hour
97.2
88.0
96.1
891
98.1
57.3
100.3
102.7
4.4
73.0
32.4
92.5
94.7
85.0
87.4
931
28.1
86.3
245
76.6
99.0
97.6
906
101.4
99.6
111 0
91.3
96.5
978
98.0
99.5
95.7
89.3
98.4
96.9
95.5
96.5
106.0
87.9
"556
97.2
98.6
963
90.3
35.9
933
90.5
118.2
94.7
95.8
92.4
92.4
92.0
57.7
75.7
59.9
86.0
95.5
985
929
89.4
94.4
90.0
%
RSD.
3.8
2.3
2.1
1.7
2.8
4.1
2.5
3.4
78.3
7.0
1.0
3.5
1.8
1.5
6.8
1.9
24.2
2.5
27.7
5.5
2.2
5.1
3.6
2.9
1.4
1.1
5.1
1.9
22
1.3
2.2
2.9
3.5
1.4
1.6
1.8
1.3
2.5
06
12.6
1.3
1.5
1.3
7.4
6.1
2.2
1.5
4.5
4.8
1.3
2.7
6.8
1.4
3.0
25
6.4
2.4
3.6
2.7
3.7
3.0
56
4.2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
* Denotes analyte required to be added to this study by the WDNR and w hich mist pass the inposed criteria to gain
acceptance for this method in Wisconsin.
154
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS Table 19
Matrix - Biologically Active Garden Soil
VOC METHOD 8260 STANDARD 153.8 UG/KG / 3.08 UG/L
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
1 •
2 *
3 *
4 *
5 *
6 '
7 '
8 *
9 '
10 *
11 *
12 *
13 *
14 '
15 *
16 *
17 *
18 *
19 *
20 '
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
ANALYTE
1 , 1 ,2-Trichloroethane
1,2,4-Trimethylbenzene
i, 3, 5-Trirrethy (benzene
Benzene
Bromodichloromethane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1,3- Dichloropropene
BDB (1,2-Dibrorroethane)
Bhylbenzene
Methylene Chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1 .3 - Dichloropropene
Vinyl Chloride
1.1 - Dichloropropene
1 . 1 -Die hloroethane
1,1-Dichloroethene
1,1,1 -Trichloroethane
1 , 1 , 1 ,2 - Tetrac hloroethane
1 .1 ,2,2-Tetrachloroethane
1 ,2-Dibrorro-3-Chloropropane
1 ,2-Dichlorobenzene
1,2-Dichloroethane
1 ,2-D'chloropropane
1,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichtorobenzene
1 ,3-Dicnlorobenzene
1 ,3-Dichtaropropane
1 ,4-Dchlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Ally I Chloride
Bromobenzene
Brorrochloromethane
Cnlorobenzene
Chloroethane
Chlororrethane
cis-1 ,2-Dichloroethene
Dibromochloromethane
Dibromorrethane
Dichlorofluororrethane
Di-isopropyl Bher
Bhyl Bher
Hexachlorobutadiene
bopropylbenzene
n-Butylbenzene
n-R-opylbenzene
p-teopropy toluene
sec-Butylbenzene
tert-Butylbenzene
Tetrac hloroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trichlorofluoromethane
Trichlorotrifluoroethane
10 Day
Absol
129
110
123
123
119
69
123
136
4
Bb
10/
141
133
218
93
124
29
112
24
88
108
131
114
139
122
129
113
128
129
139
135
114
104
127
126
122
121
121
113
bb
12b
135
126
123
27
126
100
163
122
132
123
122
121
70
93
66
110
124
127
126
136
119
104
Rec %
84
71
80
80
77
45
80
88
26
55
69
92
86
71
60
80
19
73
15
57
70
85
74
90
79
84
73
83
84
90
87
74
68
82
82
79
79
79
73
36
81
88
82
80
18
B2
65
106
79
86
80
79
fti
46
60
43
71
81
82
82
88
77
68
10 Day
Absol
129
110
125
122
121
70
121
136
2
83
104
129
134
215
87
118
29
113
18
86
107
132
116
134
126
129
109
125
128
130
132
113
100
127
122
122
124
119
109
63
122
127
124
118
29
12B
10b
1b9
121
132
121
129
119
67
94
67
110
124
127
132
135
115
101
Rec %
84
71
81
79
78
46
79
88
1
54
67
84
87
140
56
77
19
73
12
56
69
86
75
87
82
84
71
81
83
84
86
73
65
82
79
79
81
77
71
41
79
83
80
76
19
83
68
103
78
86
79
84
77
44
61
43
71
80
83
86
61
75
66
10 Day
Abaol.
125
103
119
114
113
71
111
133
3
82
101
114
129
214
90
113
25
108
21
64
102
118
92
122
123
132
115
125
129
125
135
112
95
120
129
115
122
100
112
62
128
127
120
101
20
117
97
158
94
127
119
126
119
6b
92
bb
108
120
13B
119
123
91
M
Rue %
81
67
77
74
73
46
72
86
2
53
65
74
84
139
58
73
16
70
13
42
66
77
59
79
80
86
74
81
84
81
88
72
61
78
84
75
79
65
73
40
83
83
78
65
13
76
63
102
61
82
77
82
77
42
59
42
70
78
89
77
80
59
48
10 Day
Absol
131
111
125
117
115
69
112
130
2
/9
103
120
132
222
99
125
26
108
19
63
102
120
94
121
122
133
108
129
130
125
136
117
109
124
126
123
129
100
116
50
128
123
12/
93
21
119
98
157
102
128
118
125
123
69
yb
71
112
12b
130
112
132
98
84
Rec %
85
72
81
76
75
45
72
84
1.3
51
67
78
86
144
64
81
17
70
12
41
• 66
78
61
79
79
86
70
84
84
81
88
76
71
81
82
80
84
65
75
33
83
80
83
60
14
77
64
102
66
83
77
81
80
45
62
46
72
81
85
73
86
63
54
10 Day
Absol
131
110
128
122
119
66
117
138
2
81
107
126
136
217
91
125
27
111
20
77
106
129
100
135
123
134
112
129
130
131
134
114
99
127
127
124
127
104
115
61
126
134
126
108
24
123
96
166
108
130
124
123
124
68
9b
63
110
126
123
113
130
103
83
Rec %
85
71
83
79
77
43
76
89
1
52
b9
82
88
141
59
81
18
72
13
50
69
84
65
88
80
6f
72
84
84
8b
8/
74
64
83
83
81
82
67
75
40
82
87
82
/O
15
80
62
108
70
85
81
80
80
44
61
41
71
82
80
n
84
at
54
Absolute
AVG.
129
109
124
119
117
69
117
134
2
82
104
126
132
217
92
121
27
110
20
75
105
126
103
130
123
131
111
127
129
130
134
114
101
125
126
121
124
109
113
58
126
129
124
108
24
122
99
160
109
130
121
125
121
68
94
66
110
124
129
120
131
105
89
AVG % REC
Of 0 Hour
96.0
S2.9
92.4
87.5
94.8
33.2
95.4
100.6
1.9
52.7
781
93.5
96.3
i07
85.9
91.2
21 7
32.0
158
54.4
795
93.1
iO.4
99.3
99.4
107.1
87.2
96.7
97.0
97.7
99.9
96.2
87.8
95.5
94.7
94.4
947
93 1
850
44.5
97.9
992
95.2
52.9
21.3
90.0
88.1
I220
870
94.0
87.7
94.3
89.3
51.5
700
50.4
J2.4
92.5
98.1
90.9
90.4
86.1
77.5
RSD
2.0
29
2.6
3.3
2.8
2.8
4.7
2.3
45.2
3.0
2.5
8.1
2.1
1.4
5.1
4.4
7.5
2.1
109
15.7
26
5.2
10.7
62
1 4
1.9
2.6
1 6
05
4.4
1.0
1 8
55
2.5
2.0
2.9
2.6
9.8
2.5
9 1
20
40
2.2
11.3
154
38
3.5
2.4
10.9
1.9
2.0
2.2
1 9
2.8
1.6
4.3
1.3
1.7
4.2
7 1
4.0
11 2
147
* Denotes analyte required to be added to this study by the WDNR and w hich rrust pass the imposed criteria to gain
acceptance for this method in Wisconsin.
155
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS
Matrix - Biologically Active Garden Soil
VOC METHOD 8260 STANDARD 769.5 UG/KG / 15.38 UG/L
Table 20
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
ANALYTE
1,1,2-Trichloroethane
1 ,2,4-Trirrethylbenzene
1 ,3,5-Trimethylbenzene
Benzene
Bromodichloromethane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1,3 - Dichloropropene
BDB (1,2-Dibrorroethane)
Bhylbenzene
Methylene Chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1,3 - Dichloropropene
Vinyl Chloride
1,1 - Dichloropropene
1,1-Dichloroethane
1,1-Dichloroethene
1,1,1 -Trichloroethane
1,1,1,2-Tetrachloroethane
1 ,1 ,2,2-Tetrachloroethane
1 ,2-Dibrorro-3-Chloropropane
1 ,2-Dichlorobenzene
1,2-Dichloroethane
1 ,2-Dichloropropane
1,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dichlorobenzene
1,3-Dichloropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Allyl Chloride
Brorrobenzene
Brorrochloromethane
Chlorobenzene
Chloroethane
Chloromethane
cis-1,2-Dichloroethene
Dibromochloromethane
Dibromorrethane
Dichlorofluororrethane
Di-isopropyl Bher
Bhyl Bher
Hexachlorobutadiene
teopropylbenzene
n-Butylbenzene
n-Propylbenzene
p-bopropyltoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trichlorofluoromsthane
Trichlorotrifluoroethane
0 Hour
Absol
645
672
664
655
655
580
678
699
624
646
661
650
660
1318
557
653
609
663
648
600
668
680
689
700
658
517
612
617
670
675
597
534
639
606
656
593
610
583
614
579
610
673
639
701
543
669
617
676
655
665
705
583
649
590
650
671
677
673
648
665
752
668
575
Roc %
84
87
86
85
85
/5
88
91
81
84
86
84
86
171
72
85
/y
86
84
78
ur
88
90
91
86
67
79
80
87
88
78
69
83
79
85
77
79
76
80
75
79
87
83
91
71
87
80
88
85
86
92
76
84
77
84
87
88
87
84
86
98
Ut
75
0 Hour
Absol
678
676
671
673
667
625
680
705
637
665
658
654
696
1316
616
657
625
661
655
597
670
696
676
701
659
585
691
623
682
671
641
565
663
609
664
597
610
566
617
591
597
683
643
710
553
690
626
692
662
6/9
721
589
650
612
6b3
698
682
679
643
674
743
ti2/
521
Roc %
88
88
87
8/
87
81
88
92
83
86
86
85
90
171
80
85
81
86
85
78
87
90
88
91
86
76
90
81
89
87
83
73
86
79
86
78
79
74
80
77
78
89
83
92
72
90
81
90
85
88
94
76
84
79
85
91
89
88
84
88
96
81
68
0 Hour
Absol
669
697
680
671
672
622
684
700
639
672
671
658
684
1344
602
666
633
676
661
585
670
687
687
703
660
519
649
629
679
681
609
572
675
616
666
601
612
563
632
571
603
669
6b4
709
649
703
640
701
636
681
707
594
64/
631
662
720
701
693
652
686
79/
64b
529
Rec %
87
91
88
87
87
81
89
91
83
87
87
85
89
175
78
87
82
88
86
76
87
89
89
91
86
67
84
82
88
88
79
74
88
80
86
78
79
73
82
74
78
87
85
92
71
91
83
91
83
88
92
77
84
82
86
94
91
90
8b
89
104
84
69
0 Hour
Absol
623
640
624
618
614
bb6
637
655
591
631
617
616
632
1222
564
610
585
618
606
555
620
638
642
652
630
503
655
594
631
619
598
526
599
578
624
564
578
509
586
532
576
620
607
641
516
642
580
643
618
630
666
565
612
570
621
653
638
635
600
631
711
609
512
Rec %
81
83
81
80
80
72
83
85
77
82
80
80
82
159
73
79
/6
80
79
72
81
83
83
85
82
65
85
77
82
80
78
68
78
75
81
73
75
66
76
69
75
81
79
83
67
83
75
84
80
82
87
73
79
74
81
85
83
82
78
82
92
79
67
0 Hour
Absol
674
713
695
682
690
645
710
717
647
684
686
678
696
1370
626
677
643
683
672
614
698
712
692
721
672
552
704
640
696
684
627
584
716
620
683
610
622
558
637
593
614
683
660
732
565
706
649
704
671
691
723
619
665
65 /
683
742
720
709
66 /
715
774
630
523
Roc %
88
93
90
89
90
84
92
93
84
89
89
88
90,
178
81
88
83
89
87
80
91
92
90
94
87
72
91
83
90
89
81
76
93
81
89
79
81
73
83
77
80
89
86
95
73
92
84
91
87
90
94
80
86
85
89
96
94
92
87
93
101
82
68
Absolute
AVG.
658
679
667
660
659
605
678
695
627
659
658
651
673
1314
593
652
619
660
648
590
665
682
677
695
656
535
662
620
671
666
614
556
658
606
658
593
606
556
617
573
600
665
640
698
545
682
622
683
646
669
704
590
644
612
654
697
683
678
642
674
755
636
532
Averaqe
7o REC %*
1070
110 5
108.4
1073
1072
985
110.2
1130
102.0
1073
107 1
1059
109.5
106.8
964
1061
100.7
1073
105.4
96.0
1062
11 ! 0
110 1
113 1
106.6
S7.0
107.7
1009
109.2
1083
99.9
901
107.1
985
107 1
964
936
90.4
100.3
93.2
97.6
1082
104.1
1135
88.6
110.9
101.2
111.1
1051
1088
114.5
959
104.8
99.5
1063
1133
111 1
1102
104.4
1096
1228
1034
86. 5
%
RSD.
3.5
4.1
4.0
3.9
4.3
6.0
3.9
3.4
3.5
3.2
3.9
3.4
41
4.2
5.2
3.9
3.6
39
39
3.8
4.2
4.0
3.0
37
2.4
62
5.5
28
3.7
4.0
3.1
4.5
6.6
2.7
33
2.9
2.7
5.0
3.2
43
25
3.9
3.2
4.9
3.3
39
4.3
3.6
3.1
35
32
33
30
5.6
3.4
5.2
4.5
4.1
3.9
4.6
4.3
3.5
4.6
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
* Denotes analyte required to be added to this study by the WDNR and w hich mist pass the imposed criteria to gain
acceptance for this method in Wisconsin.
156
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS Table 21
Matrix - Biologically Active Garden Soil
VOC METHOD 8260 STANDARD 769.5 UG/KG /15.38 UG/L
ANALYTE
1 , 1 ,2-Trichloroethane
1,2,4-TrimBthylbenzene
1 ,3,5-Trimethylbenzene
Benzene
Bromodichlororrethane
Bromoform
Carbon Tetrachtoride
Chloroform
cis - 1,3 - Dichloropropene
HDB (1,2-Dibromoethane)
Bhylbenzene
Methylene chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
toluene
trans - 1 ,3 - Dchloropropene
Vinyl Chloride
1,1 - Dchloropropene
1, -Dchloroethane
1, -Dchloroethene
1, ,1-Trichloroethane
1, ,1,2-Tetrachtoroethane
1 , ,2,2-Tetrachloroethane
1 ,2-Dibromo-3-Chloropropane
1,2-Dchlorobenzene
1,2-Dchloroethane
1 ,2-Dchloropropane
1,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dchlorobenzene
1 ,3-Dchloropropane
1 ,4-Dchlorobenzene
2-Chtorotoluene
2,2-Dchloropropane
4-Chlorotoluene
Allyl Chloride
Bromobenzene
Bromochloromethane
Chlorobenzene
Chloroethane
Chlororrethane
cis-1 ,2-Dchloroethene
Dbromoc hlororrethane
Dbromomethane
Dchlorofluoromethane
D-bopropyl ether
Bhyl Bher
Hexachlorobutadiene
bopropylbenzene
n-Butylbenzene
n-R-opylbenzene
p-bopropyltoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-Dchloroethene
Trichloroethene
Trichlorofluoromethane
Trichlorotrifluoroethane
Rep.
24
Absol
658
662
655
637
640
554
630
689
458
623
628
640
674
1282
565
641
537
634
549
539
632
684
628
66/
637
538
613
619
657
667
613
551
645
600
641
594
595
564
605
517
592
641
630
648
494
663
588
6/2
594
675
686
580
624
5/0
623
662
658
655
620
643
737
567
476
#1
Hour
Rec. %
i_ 85
86
85
83
83
72
82
90
60
81
82
83
88
83
73
83
70
82
71
70
82
89
82
8/
83
70
80
80
85
87
80
72
84
78
83
77
77
/3
/9
6/
77
83
82
84
64
86
76
8/
77
88
89
75
81
74
81
86
86
85
81
84
96
74
62
Rep.
24
Absol
676
679
677
652
660
587
678
697
492
643
646
642
6/2
1304
6UO
652
555
643
5/0
563
637
683
6/1
704
649
537
636
627
663
679
618
563
658
613
659
598
615
571
615
526
596
666
643
654
489
680
612
669
620
683
694
612
650
603
640
685
693
6/8
643
643
746
651
582
#2
Hour
Rec %
88
88
88
85
86
76
88
91
64
83
84
83
8/
85
/8
85
72
84
74
/3
83
89
87
91
84
70
83
81
86
88
80
73
85
80
86
78
80
74
80
68
77
86
83
85
63
88
79
8/
81
89
90
79
84
78
83
89
90
88
83
84
97
85
76
Rep.
24
Absol.
681
6/1
660
642
659
613
643
/OO
495
665
646
640
692
1298
606
660
557
637
578
553
638
677
644
677
640
579
670
630
675
677
624
570
672
60 /
6/4
602
602
54/
611
526
601
663
634
650
488
680
608
713
616
680
688
606
634
585
636
677
674
6/4
627
64/
/45
582
507
#3
Hour
Rec %
88
8/
86
83
86
80
84
91
64
86
84
83
90
84
79
86
72
83
75
72
83
88
84
88
83
75
8/
82
88
88
81
74
87
79
88
78
78
71
79
68
/8
86
82
84
63
88
79
93
80
88
89
/9
82
76
83
88
88
88
81
84
9/
76
66
Rep.
24
Absol.
665
704
685
653
669
594
665
705
466
649
647
632
671
1308
623
654
56 /
652
564
564
654
681
656
687
666
582
606
638
659
665
585
601
722
616
667
608
611
548
624
529
609
663
640
653
498
686
613
681
625
687
688
633
644
637
654
721
/16
693
64/
654
698
637
529
#4
Hour
Rec.%
86
91
89
85
87
77
86
92
60
84
84
82
87
85
81
85
74
85
73
73
85
88
85
89
86
76
79
83
86
86
76
78
94
80
87
79
79
71
81
69
79
86
83
85
65
89
80
88
81
89
89
82
84
83
85
94
93
90
84
85
91
83
69
Rep.
24
Absol
668
661
655
638
653
596
626
686
4/8
655
633
62/
690
1270
625
635
557
630
564
538
626
675
629
673
649
609
705
629
658
661
637
560
680
611
673
599
593
522
603
510
602
652
626
628
489
643
599
690
606
674
699
624
624
579
622
670
65 /
660
610
633
698
562
493
#5
Hour
Rec •/,
87
86
85
83
85
77
81
89
62
85
82
81
90
82
81
83
72
82
73
/O
81
88
82
87
84
79
92
82
85
86
83
73
88
79
87
78
77
68
78
66
78
85
81
82
63
84
78
90
79
88
91
81
81
75
81
87
85
66
79
82
91
/3
64
Absolute
AVG.
669
675
666
644
656
589
648
695
478
647
640
636
680
1292
604
648
554
639
565
551
637
680
645
681
648
569
646
628
662
669
615
569
675
609
663
600
603
550
611
522
600
657
634
646
491
670
604
685
612
680
691
611
635
595
635
683
679
672
629
644
725
600
517
AVG % REC
of 0 Hour
101.8
99.4
99.9
97.7
99.5
97.3
95.6
100.0
76.1
98.1
97.1
97.7
100.9
98.4
101.9
99.3
89.6
96.8
87.1
L 93.4
95.8
996
95.3
98.0
98.8
106.3
976
101.3
98.6
100.5
100.2
102.3
102.6
100.6
100.7
101.2
99.5
99.1
99.1
91 0
100.0
98.7
99.1
92.6
90.1
98.3
97.0
100.3
94.7
101.6
98.1
103.6
98.6
97.2
97.1
98.0
99.4
99.1
98.0
95.5
96.0
94.3
97.2
RSD.
1.4
2.6
2.1
1.2
1.6
3.7
3.5
1.1
3.4
2.5
1.4
1.0
1.5
1.2
4.0
1.6
2.0
1.4
1 9
2.3
1.6
0.6
2.8
2.2
1.7
5.5
64
1.1
1.1
1.2
3.1
3.3
4.3
1.0
2.1
0.8
1.6
3.4
1 4
1.5
1.1
1 6
1.1
1.7
0.9
2.6
1.7
2.6
2.0
0.8
0.8
3.3
1.8
4.5
2.1
3.4
3.7
2.3
2.5
1.2
3.4
6.9
8.0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
* Denotes analyte required to be added to this study by the WDNR and w hich trust pass the imposed criteria to gain
acceptance for this method in Wisconsin.
157
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS
Matrix - Biologically Active Garden Soil
VOC METHOD 8260 STANDARD 769.5 UG/KG /15.38 UG/L
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
Table 22
1 *
2 *
3 *
4 *
5 '
6 *
7 *
8 *
9 *
10 '
11 *
12 *
13 *
14 *
15 *
16 *
17 *
18 *
19 '
20 *
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
ANALYTE
1 , 1 ,2-Trichloroethane
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
Benzene
Bromodichloromethane
Brorroform
Carbon Tetrachloride
Chloroform
cis - 1,3 - Dichloropropene
B3B (1,2-Dibromoethane)
Bhylbenzene
Methylene chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1,3 - Dichloropropene
Vinyl Chloride
1,1 - Dichloropropene
1,1-Dichloroethane
1,1-Dichloroethene
1,1,1 -Trichloroethane
1,1,1,2-Tetrachloroethane
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dibromo-3-Chloropropane
1 ,2-Dichlorobenzene
1,2-Dichloroethane
1 ,2-Dichloropropane
1,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dichlorobenzene
1 ,3-Dichloropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dchloropropane
4-Chlorotoluene
Allyl Chloride
Brorro benzene
Brorrochloromethane
Chlorobenzene
Chloroethane
Chlororrethane
cis-1,2-Dichloroethene
Dbrorrochloromethane
Dibrorromethane
Dichlorofluoromethane
Di-lsopropyl ether
Bhyl Bher
Hexachlorobutadiene
bopropylbenzene
n-Butylbenzene
ri-R-opylbenzene
p-bopropyltoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-D'chloroethene
Trichloroethene
Trichlorofluororrethane
Trichlorotrifluoroethane
48 Hour
Absol
650
688
664
637
646
587
663
6/4
413
636
634
628
667
1280
620
643
516
636
528
556
639
656
642
669
656
591
639
634
649
659
61 /
595
703
609
653
596
602
523
607
504
594
64/
631
624
487
655
598
6/3
600
665
6/5
63 /
632
604
634
699
691
6/1
629
B50
/08
627
545
Rec %
84
89
86
83
84
76
86
88
54
83
82
82
87
166
81
84
67
83
69
72
83
85
83
87
85
77
83
82
84
86
80
77
91
79
85
77
78
68
79
65
77
84
82
81
63
85
/8
87
78
86
88
83
82
78
82
91
90
87
82
84
92
81
71
48 Hour
Absol
651
677
666
635
655
568
644
691
435
632
623
627
676
1244
599
640
495
623
535
529
628
661
632
672
651
640
636
626
652
654
593
585
681
600
642
580
605
502
604
504
592
648
612
650
481
661
604
671
619
673
690
622
633
583
621
683
682
671
615
61 /
65B
58/
505
Rec %
85
88
86
83
85
74
84
90
57
82
81
81
88
162
78
83
64
81
69
69
82
86
82
87
85
83
83
81
85
85
77
76
88
78
83
75
79
65
78
65
77
84
79
84
62
86
78
87
80
8/
90
81
82
76
81
89
89
8/
80
80
85
/6
66
48 Hour
Absol
678
690
686
664
674
609
664
712
442
662
660
651
705
1323
638
664
535
659
546
555
666
693
644
699
6/5
639
675
653
679
686
628
617
704
62 /
683
615
621
528
628
525
612
665
656
655
493
686
628
719
637
698
/03
626
654
614
643
703
700
693
640
654
706
594
528
Rec %
88
90
89
86
88
79
86
93
57
86
86
85
92
172
83
86
70
86
71
72
87
90
84
91
88
83
88
85
88
89
82
80
91
81
89
80
81
69
82
68
80
86
85
85
64
89
82
93
83
91
91
81
85
80
83
91
91
90
83
85
92
77
69
48 Hour
Absol
641
678
667
640
647
569
654
686
402
621
626
624
660
1265
610
644
515
630
522
561
640
669
651
681
644
553
632
621
649
666
603
584
692
613
646
600
597
501
604
510
591
645
628
652
492
671
600
669
6l3
665
675
608
628
598
632
68 /
685
671
624
655
/10
59 /
503
Rec. %
83
88
87
83
84
74
85
89
52
81
81
81
86
164
/9
84
67
82
68
73
83
87
85
88
84
72
82
81
84
86
78
76
90
80
84
78
78
65
78
66
77
84
82
85
64
87
78
87
80
86
88
/9
82
78
82
89
89
87
81
85
92
78
65
48 Hour
Absol
661
667
661
645
646
604
654
679
413
649
634
624
690
1285
626
652
520
634
539
549
642
669
644
684
651
594
677
631
665
667
630
575
685
607
666
592
605
494
613
503
603
637
633
645
482
665
606
683
609
671
682
62 /
627
572
616
674
663
661
637
646
724
607
553
Rec %
86
87
86
84
84
78
85
88
54
84
82
81
90
167
81
85
68
82
70
71
83
87
84
89
85
77
88
82
86
87
82
75
89
79
87
77
79
64
80
65
78
83
82
84
63
86
79
89
79
87
89
81
81
74
80
88
86
86
83
84
94
79
72
Absolute
AVG.
656
680
669
644
653
587
656
688
421
640
635
631
679
1279
618
648
516
636
534
550
643
669
642
681
655
603
652
633
659
666
614
591
693
611
658
596
606
510
611
509
598
648
632
645
487
667
607
683
615
674
685
624
635
594
629
689
684
673
629
644
701
602
527
AVG. % REC.
of 0 Hour
99.7
100.1
100.3
97.6
99.1
97.0
96.7
99.0
37.1
97.0
96.5
96.9
100.9
97.4
1043
99.4
J3.4
96.4
324
93.2
96.7
98.1
94.9
98.0
99.9
112.7
98.5
102.0
98.1
100.0
100.0
106.3
105.2
100.9
99.9
100.6
100.0
91.7
99.1
88.8
99.7
97.4
98.7
924
89.3
97.9
97.6
1000
95.2
1008
97.2
105.7
98.5
97.1
96.2
98.9
100.1
993
98.0
95.6
92.8
94.7
99.0
RSD.
2.2
1.4
1.5
1.8
1.9
3.2
1.3
2.1
4.0
2.5
23
1 8
2.6
2.3
2.4
1.5
2.8
2.1
1.7
2.2
2.2
2.1
1.1
1.7
1.8
6.1
3.4
2.0
2.0
1.8
2.6
2.7
1.5
1.6
2.5
2.2
1.5
3.0
1.6
1.8
1.5
1.6
2.5
1.9
1.1
1.8
2.0
31
2.2
2.0
1.7
1.7
1.7
2.8
1.7
1.7
2.0
1.7
1.6
2.5
3.7
2.6
43
* Denotes analyte required to be added to this study by the WDNR and w hich rrust pass the irrposed criteria to gain
acceptance for this method in Wisconsin.
158
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS » Table 23
Matrix - Biologically Active Garden Soil
VOC METHOD 8260 STANDARD 769.5 UG/KG /15.38 UG/L
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
1 •
2 *
3 '
4 *
5 *
6 '
7 '
8 *
9 *
10 *
11 *
12 *
13 *
14 *
15 *
16 *
17 *
18 *
19 *
20 *
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
ANALYTE
1 , 1 ,2-Trichloroethane
1 ,2,4-TrimBthylbenzene
1 ,3,5-Trirrethylbenzene
Benzene
Bromodichloromethane
Brorroform
Carbon Tetrachloride
Chloroform
cis - 1,3 - Dichloropropene
HX (1,2-Dibromoethane)
Bhylbenzene
Methytene chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1 ,3 - Dichloropropene
Vinyl Chloride
1,1 - Dichloropropene
1,1-Dichloroethane
1,1-Dichloroethene
1 ,1 ,1-Trichloroethane
1,1,1.2 - Tetrachloroethane
1,1,2,2-Tetrachloroethane
1 ,2-Dibrom>3-Chloropropane
1 ,2-Dichlorobenzene
1,2-Dichloroethane
1 ,2-Dicnloropropane
1 ,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-D'chlorobenzene
1 ,3-Q'chtoropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Ally! Chloride
Brorrobenzene
Bromochlororrethane
Chlorobenzene
Chloroethane
Chloromethane
cis-1 ,2-Dichloroethene
Dibrorrochloromethane
Dibrorromethane
Dichlorofluororrethane
Di-bopropyi ether
Bhyl Bher
Hexachlorobutadiene
Isopropylbenzene
n-Butylbenzene
n-Ropylbenzene
p-teopropy toluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachtoroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trichlorofluoromethane
Trichlorotrifluoroethane
72 Hour
Absol
651
660
652
622
645
563
628
679
389
626
619
596
661
1224
574
633
477
608
505
492
602
652
582
655
639
568
635
620
641
651
609
554
651
593
635
577
605
460
594
472
584
620
620
605
447
632
590
649
587
655
6/3
588
621
544
605
651
Bb/
6b2
599
593
695
512
425
Rtc %
85
86
85
81
84
73
82
88
51
81
80
77
86
80
75
82
62
79
66
64
78
85
76
85
83
74
82
81
83
85
79
72
85
77
82
75
79
60
77
61
76
81
81
79
58
82
77
84
76
85
87
76
81
71
79
85
85
85
78
77
90
67
bb
72 Hour
Abm
659
666
667
629
659
572
666
664
388
631
627
626
666
1270
575
651
492
628
493
538
633
666
633
686
665
606
663
632
649
667
617
541
666
610
645
602
616
490
615
480
605
652
626
650
456
6/0
620
680
626
6/4
681
b90
641
561
629
669
6/6
668
635
639
690
601
b43
Rac %
86
87
87
82
86
74
87
89
50
82
81
81
87
83
75
85
64
82
64
70
82
87
82
89
86
79
86
82
84
87
80
70
86
79
84
78
80
64
80
62
79
85
81
84
59
87
81
88
81
88
88
77
83
73
82
8/
88
87
82
83
90
78
71
72 Hour
Absol
694
702
700
680
692
589
679
720
444
665
672
654
684
1335
579
686
515
669
564
566
679
701
666
710
692
671
630
653
678
705
607
576
709
637
682
621
630
736
641
548
621
675
659
6bb
bU/
/OB
638
/08
646
/03
/Ob
61b
666
602
670
716
713
/(JO
66 /
66 /
690
B04
529
Roc %
90
91
91
88
90
76
88
94
58
86
87
85
89
87
75
89
67
87
73
73
88
91
87
92
90
87
82
85
88
92
79
75
92
83
89
81
82
96
83
71
81
88
86
85
66
92
83
92
84
91
92
8u
8/
78
87
93
93
91
a/
87
90
78
69
72 Hour
Absol.
658
677
672
637
651
583
651
690
407
623
631
624
668
1266
599
643
476
630
521
541
627
667
650
676
659
658
640
629
651
680
617
575
681
614
656
608
611
694
625
509
611
656
b2b
658
474
654
595
bb9
613
670
677
59b
641
57 f
632
b/b
683
6^0
631
643
646
BUB
547
Rec. %
86
88
87
83
85
76
85
90
53
81
82
81
87
82
78
84
62
82
66
TO
81
87
84
88
86
86
83
82
85
88
80
75
88
80
85
79
79
90
81
66
79
85
81
85
62
85
77
87
60
87
88
77
83
75
B2
88
89
68
82
83
84
79
71
72 Hour
Absol.
672
677
676
656
664
580
678
697
427
643
644
634
669
1292
578
650
495
647
536
558
663
678
669
717
661
663
623
635
671
659
625
566
668
619
647
605
628
702
626
534
61b
6/0
637
657
480
682
616
696
629
680
684
623
656
577
640
680
685
t>82
651
645
6b9
658
568
KK %
87
88
88
85
86
75
88
91
55
83
84
82
87
84
75
84
64
84
70
73
86
88
87
93
86
86
81
83
87
86
81
73
87
80
84
79
82
91
81
69
80
87
83
85
62
89
80
90
82
88
89
81
85
75
83
88
89
89
85
84
86
86
74
Absolute
AVG.
667
676
673
645
662
577
660
694
411
637
639
627
669
1277
581
652
491
636
524
539
641
673
640
689
663
633
638
634
658
672
615
562
675
614
653
602
618
616
620
508
607
654
633
645
473
669
612
680
620
676
684
602
645
572
635
678
683
676
636
637
676
597
522
AVG '•, StC
of 0 Hour
101.4
99.6
101.0
97.7
100.4
95.4
97.4
99.9
35.5
96.6
97.0
96.3
99.4
97.2
98.0
100.0
79.3
964
i08
91 3
963
986
94.5
99.1
101.1
118.3
964
I022
98.0
1009
100 1
101.1
102.5
101.5
992
101 6
101.9
1109
100.5
887
101.2
983
989
924
867
98 1
98.3
99.6
95.9
101.1
971
1021
1001
935
97.2
974
99.9
998
99.2
946
895
338
982
RSD.
2.5
2.4
2.6
3.6
2.7
1.7
3.2
2.3
5.9
2.7
3.3
3.3
1.3
3.2
1.8
3.0
3.2
3.6
5.3
5.3
4.7
2.7
5.5
3.7
2.9
7.1
2.4
1.9
2.4
3.2
1.2
2.6
3.2
2.6
2.8
2.7
1.7
21.2
2.8
6.5
2.3
3.3
2.4
3.5
4.9
4.2
3.2
3.4
3.5
2.6
1.9
2.6
2.6
3.8
3.7
3.5
2.9
2.6
4.0
4.2
3.3
8.8
10.7
* Denotes analyte required to be added to this study by the WDIvF and w hlch rrust pass the imposed criteria to gain
acceptance for this method in Wisconsin.
159
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS
Matrix - Biologically Active Garden Soil
VOC METHOD 8260 STANDARD 769.5 UG/KG /15.38 UG/L
Table 24
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
ANALYTE
1 , 1 ,2-Trichloroethane
1 ,2,4-Trimethylbenzene
1 ,3,5^Trimethylbenzene
Benzene
Bromodichloromethane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1 ,3 - Dichloropropene
KB (1,2-Dibromoethane)
Bhylbenzene
Methylene chloride
MTBE
m&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1,3- Dichloropropene
Vinyl Chloride
1,1 - Dichloropropene
1,1-Dichloroethane
1,1-Qchloroethene
1 , 1 , 1 -Trie hloroethane
1,1,1,2 - Tetrachloroethane
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dibrorro-3-Chloropropane
1 ,2-Dichlorobenzene
1,2-Dic hloroethane
1,2-Dichloropropane
1,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Qchlorobenzene
1,3-Dichloropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Ally I Chloride
Brorrobenzene
Brorrochloromethane
Chlorobenzene
Chloroethane
Chloromethane
cis-1,2-Dichloroethene
Dibrorrochlororrethane
Dibromomethane
Dichlorofluororrethane
Di-teopropyl ether
Bhyl Bher
Hexachlorobutadiene
teopropylbenzene
n-Butylbenzene
n-R"opylbenzene
p-bopropyltoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trichlorofluororrethane
Trichlorotrifluoroethane
7 Day
Absol
675
664
673
653
666
581
658
697
2«b
635
631
634
697
1280
607
6b8
384
641
431
508
630
680
629
686
660
708
649
640
682
694
645
569
679
629
667
609
615
678
618
451
616
657
641
624
428
677
601
696
608
687
696
596
643
b28
618
649
666
663
642
62b
652
563
492
Roc. %
88
86
87
85
86
75
86
91
37
82
82
82
91
83
79
85
50
83
56
66
82
88
82
89
86
92
84
83
89
90
84
74
88
82
87
79
80
88
80
59
80
85
83
81
56
88
78
90
79
89
90
77
84
69
80
84
86
86
83
81
85
73
64
7 Day
Absol .
662
669
666
638
645
557
677
689
288
622
621
632
661
1266
585
648
400
629
445
542
634
676
662
690
652
66H
643
632
6bb
672
605
562
672
602
653
589
607
666
600
469
594
638
634
6b/
459
662
607
692
615
669
677
606
639
548
613
668
680
6/3
633
636
667
62b
b83
Rec %
86
87
86
83
84
72
88
90
3/
81
81
82
86
82
/6
84
52
82
58
/O
82
88
86
90
85
87
84
82
85
87
79
73
87
78
85
76
79
87
78
61
77
83
82
8b
60
86
79
90
80
87
88
79
83
71
80
86
88
87
82
83
87
81
76
7 Day
Absol
685
684
684
667
680
603
671
715
263
638
636
641
693
1291
b98
660
400
651
434
bb1
644
699
663
698
675
694
675
645
676
693
653
574
686
624
674
619
62b
6b9
622
471
620
668
6bb
6b2
483
696
618
710
631
69b
698
603
653
544
632
666
685
683
649
663
6bb
b99
b41
Rec. %
89
89
89
87
88
78
87
93
34
83
83
83
90
84
/8
86
52
85
56
72
84
91
86
91
88
90
88
84
88
90
85
75
89
81
88
80
81
86
81
61
81
87
85
8b
63
90
80
92
82
90
91
78
85
71
82
86
89
89
84
86
86
78
70
7 Day
Absol
667
665
665
553
635
574
517
609
26/
613
604
529
641
1252
591
647
389
593
423
271
501
542
414
533
652
680
635
631
620
631
602
564
686
611
650
602
613
487
604
340
604
599
62 /
391
246
581
b96
674
416
631
638
589
625
b30
608
649
662
666
602
48 /
b83
34/
322
Rec. %
87
86
86
72
83
/b
67
79
35
80
78
69
83
81
77
84
51
77
55
35
65
70
54
69
85
88
83
82
81
82
78
73
89
79
84
78
80
63
78
44
78
78
81
51
32
75
77
88
54
82
83
76
81
69
79
84
86
87
78
63
76
45
42
7 Day
Absol.
671
677
677
652
650
575
674
688
308
619
628
615
676
1272
611
649
394
638
438
545
632
685
656
686
662
697
6bb
642
653
681
632
574
680
621
656
607
618
638
625
463
611
658
639
645
465
673
612
697
620
6/3
684
613
64b
544
62b
669
685
676
643
620
641
621
b69
Rec. %
87
88
88
85
84
fb
88
89
40
80
82
80
88
83
79
84
51
83
57
71
82
89
85
89
86
91
85
83
85
88
82
75
88
81
85
79
80
83
81
60
79
85
83
84
60
87
79
91
81
87
89
80
84
71
81
87
89
88
83
81
83
81
74
Absolute
AVG.
672
672
673
633
655
578
639
680
282
625
624
610
674
1272
598
652
393
630
434
483
608
656
605
659
660
689
651
638
657
674
532
568
680
617
660
605
615
625
614
439
609
644
639
594
416
658
607
694
578
671
678
601
641
539
619
658
675
672
634
606
642
551
501
AVG. % REC
of 0 Hour
102.2
98.9
100.9
95.9
99.4
95.5
94.3
97.8
M.9
94.8
94.7
93.7
100.0
96.8
100.9
100.0
53.5
95.5
56.9
51 .9
91.4
96.2
89.3
94.7
100.7
128.9
98.4
102.8
97.9
101.2
59.1
102.2
103.4
101.9
100.2
1020
101.5
112.6
99.5
76.6
101.5
96.7
99.8
85.0
76.3
96.4
97.5
101.6
89.5
100.3
96.3
101.9
99.5
880
947
945
98.8
992
987
89.9
85.0
86.7
943
/O
RSD.
1.3
1.3
1.2
7.3
2.7
2.9
10.8
6.0
6.4
1.7
2.0
7.6
3.4
1.1
1.8
09
1.7
3.6
1 8
24.8
9.9
9.8
17.8
10.7
1.4
2.2
2.3
1.0
37
3.8
4.6
1.0
0.9
1.8
1 6
1.8
1.1
12.6
1.8
12.8
1.7
4.3
1 6
19.2
23.4
6.8
1.4
1.8
157
3.7
3.6
1.6
16
1.6
1 6
1 4
16
1 2
29
11.3
5.4
212
21.2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
* Denotes analyte required to be added to this study by the WDNR and w hich rrust pass the irrposed criteria to gain
acceptance for this method in Wisconsin
160
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
VOC STUDY RESULTS Table 25
Matrix - Biologically Active Garden Soil
VOC METHOD 8260 STANDARD 769.5 UG/KG / 15.38 UG/L
Rep. #1 Rep. #2 Rep. #3 Rep. #4 Rep. #5
1 •
2 '
3 '
4 *
5 '
6 *
7 *
8 *
9 *
10 *
11 *
12 *
13 '
14 *
15 *
16 *
17 *
18 '
19 '
20 '
21
22
23
24
25
26
2t
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
61
52
53
b4
bb
56
57
58
59
60
61
62
63
ANALYTE
1,1,2-Trichloroethane
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
Benzene
Bromodichloromethane
Bromoform
Carbon Tetrachloride
Chloroform
cis - 1,3 - Dichloropropene
BDB (1,2-DibromDethane)
Bhylbenzene
Methylene chloride
MTBE
rr&p-Xylene
Naphthalene
o-Xylene
Styrene
Toluene
trans - 1,3 - Dichloropropene
Vinyl Chloride
1,1 - Dichloropropene
,1-Dichloroethane
,1-Dichloroethene
,1,1-Trichloroethane
,1,1,2 - Tetrachloroethane
, 1 ,2,2-Tetrachloroethane
1 ,2-Dibromo-3-Chloropropane
1,2-Dichlorobenzene
1,2-Dichloroethane
1 ,2-Dichloropropane
1,2,3 - Trichloropropane
1 ,2,3-Trichlorobenzene
1 ,2,4-Trichlorobenzene
1 ,3-Dichlorobenzene
1 ,3-Dichloropropane
1 ,4-Dichlorobenzene
2-Chlorotoluene
2,2-Dichloropropane
4-Chlorotoluene
Allyl Chloride
Bromobenzene
Bromochloromethane
Chlorobenzene
Chloroethane
Chloromethane
cis-1 ,2-Dichloroethene
Dibromochloromethane
Dibromomethane
Dichlorofluoromethane
Di-bopropyl ether
Bhyl Bher
Hexachlorobutadiene
teopropylbenzene
n-Butylbenzene
n-FYopylbenzene
p-bopropyltoluene
sec-Butylbenzene
tert-Butylbenzene
Tetrachloroethene
trans-1 ,2-Dichloroethene
Trichloroethene
Trichlorofluoromsthane
Trichlorotrifluoroethane
10
Absol
680
677
667
643
660
573
655
688
268
616
621
628
674
1268
586
646
343
633
398
515
631
672
629
678
652
671
638
645
659
686
633
560
685
623
655
602
608
615
609
434
610
655
632
634
436
668
604
688
604
679
676
588
644
b40
617
651
674
670
634
634
647
597
537
Day
Rec y.
88
88
87
83
86
74
85
89
35
80
81
82
88
82
76
84
45
82
52
6/
82
87
82
88
85
87
83
84
86
89
82
73
89
81
85
78
79
80
79
56
79
85
82
82
57
87
78
89
78
88
88
76
84
70
80
85
88
87
82
82
84
78
70
10
Absol
651
6bb
665
641
656
529
681
687
227
598
610
631
643
1246
551
633
324
616
365
554
645
673
658
691
651
651
606
624
645
656
591
543
645
608
638
598
616
607
605
433
610
646
625
658
468
660
595
684
611
667
655
577
642
522
613
646
666
667
632
627
640
636
569
Day
Rac. %
85
85
86
83
85
69
88
89
29
78
79
82
84
81
72
82
42
80
47
72
84
87
86
90
85
85
79
81
84
85
77
71
84
79
83
78
80
79
79
56
79
84
81
86
61
86
77
89
79
87
85
75
83
68
80
84
87
87
82
81
83
83
74
10
Absol.
670
667
674
650
642
575
676
697
233
607
620
650
678
1258
582
646
328
631
382
543
621
688
651
701
654
697
681
639
661
661
622
553
664
623
663
606
624
602
604
438
614
659
635
638
446
671
612
688
608
681
681
600
647
520
615
632
669
682
626
h525
639
604
548
Day
Rec %
87
87
88
84
83
75
88
91
30
79
81
84
88
82
76
84
43
82
50
71
81
89
85
91
85
91
88
83
86
86
81
72
86
81
86
79
81
78
78
57
80
86
83
83
58
87
79
89
79
88
88
78
84
68
80
82
87
89
81
81
83
78
71
10
Absol
693
701
690
670
687
589
669
712
226
643
635
662
696
1308
612
678
350
656
391
509
629
696
630
703
678
691
673
661
685
706
657
583
696
641
688
633
638
601
635
434
627
674
650
621
455
695
630
739
609
/(If
713
620
652
552
630
663
690
/02
644
642
668
5/2
514
Day
Rec %
90
91
90
87
89
77
87
93
29
84
83
86
90
85
79
88
45
85
51
66
82
90
82
91
88
90
87
86
89
92
85
76
90
83
89
82
83
78
83
56
81
88
84
81
59
90
82
96
79
92
93
81
85
72
82
86
90
91
84
83
87
74
tif
10
Absol
666
659
667
653
660
579
673
693
233
608
628
642
668
1265
569
645
340
643
380
531
630
686
641
690
662
663
635
637
656
680
612
534
635
619
662
607
613
584
610
433
616
645
641
632
444
680
616
698
613
678
683
591
640
513
613
638
661
673
647
632
664
596
538
Day
Rec %
86
86
87
85
86
75
87
90
30
79
82
83
87
82
74
84
44
84
49
69
82
89
83
90
86
86
82
83
85
88
79
69
82
80
86
79
80
76
79
56
80
84
83
82
58
88
80
91
80
88
89
77
83
67
80
83
86
87
84
82
86
77
70
Absolute
AVG.
672
671
672
651
661
569
671
695
237
614
623
642
672
1269
580
649
337
636
383
530
631
683
642
692
659
674
646
641
661
678
532
554
665
623
661
609
620
602
612
434
615
656
637
636
450
675
611
699
609
682
681
532
645
529
617
646
672
679
636
632
651
601
541
AVG %R£C
of 0 Hour
102.1
98.9
100.9
98.7
1002
94.0
990
100 1
378
93.2
94.6
98.7
997
966
978
995
54.4
96.3
59.1
898
949
1000
948
996
1005
1260
977
103 3
98.5
101 8
6S.1
99.7
101.0
102.8
1004
102.7
1023
1083
993
758
1025
983
99.4
91.2
32.5
989
98.2
1023
942
1020
967
69.1
100 1
86.5
945
927
98.3
100.1
99.1
93 8
863
O'lO
101 7
%
RSD.
2.3
2.7
1.5
1.8
2.4
4.1
1.5
1.5
7.3
2.8
1.5
2.2
2.9
1.8
3.9
2.6
3.2
2.3
33
35
4.6
1.5
2.0
1.4
1.7
2.8
47
2.1
2.2
3.0
46
34
3.9
1.9
2.7
23
1.9
1.9
2.1
0.5
1.1
1.8
1.5
2.1
27
20
22
3.3
0.5
2.2
30
3.1
0.8
30
1.2
1.9
1.6
2.1
1.3
1.1
2.1
3.9
3.7
* Denotes analyte required to be added to this study by the WDNR and w hich rrust pass the imposed criteria to gain
acceptance for this rrethod in Wisconsin.
161
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 26
Biological Degradation in Soils Using an Aqueous Spiking Solution
Ottawa Sand
Benzene
Toluene
Ethyl benzene
M-P Xylene
O- Xylene
Benzene
Toluene
Ethyl benzene
M-P Xylene
O- Xylene
spike
level ug/kg
372
329
197
384
219
% rec.
1 Day
99.1
98.2
98.6
97.1
97.1
% rec.
2 Dav
114
112
112
107
104
Ust Contaminated
spike
level ug/kg
354
329
201
404
210
% rec.
1 Day
66.4
73.8
81.3
80.6
78.5
% rec.
2 Dav
108.6
99.8
99.1
89.4
94.8
% rec.
3 Dav
74.6
76.3
83.7
81.3
78.3
Soil
% rec.
3 Dav
53.8
49.9
83.6
55.3
104.1
Biologically Active Garden
spike
level ug/kg
292
269
167
319
188
% rec.
1 Day
91.4
95.4
122.8
132.7
130.9
% rec.
2 Dav
75.8
73.1
104.4
106.2
119.5
% rec.
3 Dav
75.3
77.5
97.6
100.5
100.5
% rec.
4 Dav
70.8
68.5
71.0
66.4
68.6
% rec.
4 Dav
45.2
34.9
64.7
34.4
86.9
Soil
% rec.
4 Dav
54.2
56.7
85.6
88.7
92.3
% rec.
7 Dav
72.4
71.2
78.4
70.8
75.1
% rec.
7 Dav
13.7
10.0
42.0
15.7
84.0
% rec.
7 Dav
33.2
35.4
82.0
92.9
106.2
162
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
RECOVERY OF VOCS FROM SOILS WITH AND WITHOUT METHANOL PRESERVATION
John H. Phillips
Environmental Quality Office, Ford Motor Company, One Parklane Blvd., PTE 1400, Dearborn, Ml 48126
Alan D. Hewitt
U.S. Army Cold Regions Research and Engineering Laboratory, 72 Lyme Rd., Hanover, NH 03755
Jeffery P Glaser
TriMatrix Laboratories, Inc., 5560 Corporate Exchange Ct., Grand Rapids, Ml 49512
ABSTRACT
Four sterilized reference soils were fortified with nine Volatile Organic Compounds (VOCs), then aged from three
to nineteen months in sealed glass ampoules. Prior to analysis ampoules were placed in sealed vials with water or
methanol and then shattered, to allow the solvent to contact the soil. Samples were then analyzed in triplicate by
dynamic headspace equilibrium (purge and trap) or static headspace equilibrium. Non preserved soil recoveries
are compared to soil preserved with methanol for up to 32 days. Recovery of VOCs from soil were found to be
dependent upon both the physiochemical properties of the VOC and the soil. Fortification technique and time had
a significant impact on recovery of VOCs from soil, indicating that the mechanism and duration of soil contamina-
tion is a key factor in VOC recovery. Increasing contact time of the soil with methanol resulted in a corresponding
increase in VOC recovery. Sonication and temperature effects were minor in comparison to solvent contact time.
We strongly recommend that if the high level (methanol preservation) option of Method 5035 is selected the
methanol/soil contact time be held constant and all sample vials be prepared in the same manor. The fugacity
model generally over predicted the concentration of VOCs in the soil phase when evaluating soil/vapor equilibrium.
The fugacity model generally over predicted the concentration VOCs in the aqueous phase when evaluating
soil/water equilibrium. There were exceptions to both of these trends depending upon soil type and analyte. The
fugacity model commonly employed for risk assessment is a poor predictor of VOC phase distributions for any soil
type other than sand.
INTRODUCTION
Methanol preservation of soil samples has gained considerable interest with the promulgation of EPA SW-846
method 5035.1 Methanol is a preservative which both eliminates biodegradation and substantially reduces the
volatilization of VOCs during sample transport, handling and storage.2 Many states are in the process of reevaluat-
ing their policy on VOC sample handling, and methanol field preservation is one of the techniques under consid-
eration. Numerous researchers have shown higher recoveries of VOCs, when soil samples are either preserved or
extracted with methanol.3'4 The scientific literature has also sighted that slow sorption/desorption mechanisms play
a significant role in the aging of soils, which impacts the recoverability of VOCs.5'6
By eliminating both biodegradation and volatilization of VOCs from the equation, one can evaluate the impact of
methanol preservation on the recovery of VOCs from soil. The mechanisms of VOC partitioning can then be
accessed with fugacity models which are commonly used in risk assessment. Analytical techniques can also be
compared on an unbiased basis. A critical question to be answered is whether or not methanol preservation/
extraction will lead to an overestimation of groundwater concentrations when employing standard risk assessment
techniques.
EXPERIMENTAL
Four soils, Ottawa Sand, Yokene Clay, Ft. Edwards Clay and CRREL Silt/Sand were dried, passed through a 30
mesh sieve then sent for physical characterization (Table 1). Each soil was then submitted to the Phoenix
Memorial Laboratory at the University of Michigan for sterilization by gamma irradiated in the Ford nuclear reactor.
It should be noted that PCE was a background contaminate in the CRREL soil.
Two sets of quadruplicate ampoules were prepared of each soil type, for the aqueous fortification study. One gram
dry weight was transferred to one milliliter ampoules for direct vapor partitioning analysis and three grams were
transferred to two milliliter ampoules for methanol preservation. The ampoules were then spiked with trans-1,2-
dichloroethylene (TDCE), cis-1,2-dichloroethylene (CDCE), trichloroethylene (TCE), tetrachloroethylene (PCE),
benzene, toluene, ethylbenzene and p-xylene, and treated with distilled VOC free water to create samples of
identical analyte concentrations and moisture contents. The target concentration for each analyte was 0.2 mg/Kg.
The ampoules were then flame sealed in a laminar flow hood. All ampoules were subjected to three freeze/thaw
cycles over a 98 day period prior to analysis.
Twenty-two vials containing two grams of CRREL silt/sand were allowed to vapor fortify at 20°C in a 5.6 L
163
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
desiccator for 593 days. The fortification solution contained 170-320 ug each; TDCE, CDCE, TCE, PCE, benzene,
toluene, ethylbenzene, o-xylene and p-xylene, in one milliliter of tetraglyme plus 40 mg of methanol. The expected
analyte concentration from the vapor fortification process was between two and five mg/Kg. As with the aqueous
fortified samples the ampoules were then flame sealed after the treatment period.
Samples were analyzed at the U.S. Army Cold Regions Research and Engineering Laboratory (CRREL) by
headspace gas chromatography with photo ionization detection (HS/GC/PID) and at TriMatrix Laboratory in Grand
Rapids, Michigan by purge and trap gas chromatography with mass spectrometry (PT/GC/MS). Samples were
analyzed in triplicate by both direct vapor equilibrium and methanol extraction techniques. The one milliliter
ampoule was placed inside the headspace or purge and trap vessel along with the appropriate volumes of water
and internal standards. The vessels were then shaken by hand until the ampoules fragmented completely. In the
case of methanol extraction the three milliliter ampoule was placed in a vessel containing three ml of VOC free
methanol. The vessel was then hand shaken until the vial fragmented, exposing the soil to the methanol. Approxi-
mately 100-200 pL of methanol was then removed from the soil and injected in a prepared vapor equilibrium vial
for analysis by either the headspace or purge and trap technique.
Samples were analyzed on day zero by both soil/aqueous vapor equilibrium techniques and by methanol extrac-
tion. A series of predefined methanol contact times ranging from 30 minutes to 30 days were established at the
onset of the study. All vials for methanol extraction/preservation were prepared on day zero, but methanol was not
withdrawn for analysis until their scheduled analysis time. Several of the vials were sonicated for 30 minutes at
40°C prior to analysis at pre-selected time periods.
MODELING
An inter-phase partitioning model was evaluated against experimental results to determine how accurately
standard fugacity models predict phase partitioning. Inter-phase fugacity models are the basis for fate, transport
and risk assessment predictions and form the foundation from which regulatory decisions are made. Modeling
was completed by Tim Mayotte with Colder Associates in Lansing, Michigan.7
RESULTS
Results from the aqueous fortification study and vapor fortification study are summarized in tables two and three
respectively. Results from the PT/GC/MS and HS/GC/PID vapor partitioning techniques correlated very closely.
Values derived from the two techniques were nearly always within the standard deviation of the analysis technique.
The average aqueous, low level purge and trap results were slightly greater than those for headspace, indicating
that dynamic vapor partitioning may be causing a slight shift in the soil/water equilibrium.
DICUSSION
Fugacity modeling was used to predict the phase distribution of VOCs in the sealed aqueous spiked ampoules.
Results indicated that for Ottawa sand 50% to 80% of the VOCs would volatilize to the headspace of the vial, 15%
to 30% of the VOCs would partition into the aqueous phase and 1% to 20% would remain on the soil. Compounds
with higher vapor pressures partitioned more readily into the vapor phase. This model simulation shows how easily
VOCs can be lost to the atmosphere if great care in not taken during sample collection and handling. For all other
soil types the fugacity model predicted from 40% to 98% of the VOC would be associated with the soil phase. This
shows the significant impact of organic matter (%carbon) on the fugacity models. Compounds with greater
octanol/water partition coefficients (K0w) such as PCE and benzene, toluene, ethylbenzene and xylene (BTEX) are
predicted to partition greater than 80% into the soil phase. The predominate factors influencing fugacity models
commonly used for transport, fate and risk assessment are the VOC's vapor pressure, the VOC's K0w and the
percent carbon of the soil. Surface adsorption or other physio-chemical mechanisms which are know to occur on
soil particle surfaces and in interstitial spaces have no influence on fugacity models commonly in use.
Aqueous fortification study results showed 100% recovery of VOCs from Ottawa sand. Recoveries decreased to
the range of 15% to 50% for the CRREL silt/sand. The Ft. Edwards and Yokene clays showed VOC recoveries
falling somewhere between those observed in the Ottawa sand and CRREL soil. Fugacity models predicted
Ottawa sand recoveries within experimental error. Fugacity predictions generally followed the same compound to
compound recovery trend for all other soil types, however model predictions generally varied from 10% to 100% of
the experimental values. Prediction accuracy was dependent upon soil type and compound. The purgable fraction
concentration for TDCE and CDCE was over predicted by 8 fold and 4 fold respectively for the CRREL soil.
A previous 14 day aqueous fortification of BTEX compounds on CRREL silt/sand yielded recoveries of 50% to
60% for all compounds. The current 98 day aqueous fortification study showed recoveries decreasing with
164
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
compound hydrophobicity ranging from 50% for benzene down to only 30% for xylene. This indicates that
VOC/soil contact time, often referred to as "aging", retards the movement of VOCs from the soil phase to the
aqueous and vapor phases.
A preliminary 7 day vapor fortification study with benzene, toluene, TCE and TDCE on CRREL silt/sand showed an
uptake of only about 10 ug of VOC per gram of soil. The 593 day vapor fortification study showed soil uptakes
increasing with VOC hydrophobicity and ranging from 120 ug/g for TDCE to 300 ug/g for toluene. This data once
again indicates that VOC/soil contact time or "aging" plays an important role in VOC uptake and release for certain
soil types, due to the slow sorption/desorption mechanism.
Samples which were preserved with methanol and then analyzed within 1-2 hours showed recoveries equivalent to
direct aqueous vapor equilibrium for aqueous fortified samples aging 98 days. Near 100% recovery was observed
for all VOCs in Ottawa sand with recoveries gradually decreasing by soil type (Ottawa sand > Ft. Edwards Clay >
Yokene clay > CRREL silt/sand). The CRREL silt/sand showed VOC recoveries ranging from 25% to 50%.
Methanol contact time ranging from 30 minutes to 30 days was evaluated with vapor fortified CRREL soil after
aging 593 days. All VOC recoveries steadily increased with methanol contact time, in some cases more than an
order of magnitude. The rate of increase decreased with compound hydrophobicity. These results are in agree-
ment with previous research on less volatile organic compounds. The data demonstrates that the transport of
VOCs from certain soil matrices can be extremely slow and is very much dependent upon contamination age.
Sonication is sometimes recommended in conjunction with method 5035 to reduce the variability in results due to
methanol contact time. Sonication imparts both mechanical and thermal energy into the sample which should have
a significant impact on the thermodynamics of the system. We performed a series of parallel studies where both
sonicated (30 minutes at 40°C) and static samples were extracted and analyzed at the same methanol contact
times. Contact times ranged from 3.5 hours to 101 days. Some improvement was noted in extraction efficiency at
the 3.5 hour equilibration period, however any added benefits of sonication were lost after only 24 hours of equili-
bration. No statistical difference could be detected between samples that were and were not sonicated when
methanol contact times were 24 hours or greater.
Standard inter-phase model techniques were used to predict partitioning of VOCs from CRREL silt/sand into the
aqueous and gas phases. Experimental aqueous vapor equilibrium results for the 593 day vapor fortified samples
were compared to fugacity calculations. An initial soil VOC mass as determined by the 30 day methanol contact
time was used, since the true soil concentration is not known. We recognize that the 30 day methanol contact time
result will be conservative, therefore this is the minimum mass of VOC present in the soil. We do not believe that
the total mass of VOCs on the soil is dramatically greater than the 30 day result, as the rate of VOC increase had
decreased substantially by 30 days. The fugacity model was found to dramatically over predict the mass of VOCs
expected to partition into the aqueous and gas phases (Figure 1). The over prediction ranged from 100% to more
than 23 fold depending upon the compound. The over prediction is conservative since more VOC may actually be
present in the soil.
Since standard fugacity calculations failed significantly using a three compartment model we took a step back and
used a two compartment inter-phase model to predict the vapor fortification process. The model predicted fairly
closely the gas/soil partitioning of benzene and TCE, but under estimated the movement of TDCE and CDCE from
the gas phase to the soil (Figure 2). Toluene, ethylbenzene, p-xylene and PCE movement from the gas phase to
the soil phase was either consistently overestimated or we only achieved 50% recovery of VOCs from the CRREL
silt/sand even after a 30 day methanol contact period.
CONCLUSIONS
We highly recommend that a fixed methanol soil contact time be established for method 5035 and that the metha-
nol be decanted from the soil after the established contact period. If this is not done test results will simply not be
comparable! We also question the use of methanol preservation/extraction for applications other than identifying
contamination locations. The common practice of assuming infinite depletion for contamination sources in risk
assessment and setting cleanup limits is even less appropriate when using methanol extraction data. Methanol
extraction data should never be used for transport, fate or risk assessment, since in the real world we are dealing
with groundwater and not "ground-methane!" If equilibrium can not be reached after 30 days of contact with an
organic solvent, how long will it take to reach equilibrium between groundwater and soil? The extremely slow
release of organic constituents form aged soil may allow time for biota to acclimate and degrade many organic
contaminates. The could this be why natural attenuation is showing so much promise.
165
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
We have demonstrated that standard fugacity based inter-phase partitioning models do a good job of predicting
the movement of VOCs between air, water and soil, when the soil is sand and the contamination is recent. Unfor-
tunately for the majority of sites this is not the case. The CRREL silt/sand was found to have greater retention of
VOCs than Yokene clay, which had three times more organic mater, surface area and cation exchange capacity.
This shows that these properties are not the only factors effecting gas/soil and water/soil partitioning. Yet, existing
fugacity based inter-phase partitioning models place nearly 100% of the soil sorption capacity on the percent
organic matter it contains. The CRREL silt/sand retained VOCs to a much greater extent than predicted by the
fugacity model, which indicates that other significant factors must be included in our models to get accurate
estimates of VOC fate and transport in the environment. Additional research is needed in this area. Silicate, alumi-
num and clay mineral surface adsorption along with the nature of organic mater present in the soil must be taken
into consideration at a minimum. How can we make risk assessment and regulatory policy decisions related to
VOC contamination and exposure when our basic assumptions are flawed?
ACKNOWLEDGEMENTS
The authors would like to give a special thanks to Dr. Ed Baum and Dr. Richard Rediske in the Department of
Chemistry and Water Resources Institute at Grand Valley State University, Allendale, Michigan.
REFERENCES
1. Test Methods for Evaluating Solid Waste; U.S. Environmental Protection Agency, National Technical Informa-
tion Service: Washington, DC, 1996
2. Hewitt, A.D. "Evaluation of Methanol and NaHSO4 for Preservation of Volatile Organic Compounds in Soil
Samples", American Environmental Laboratory, 8: pp. 16-18, 1995
3. Minnich, M.M.; Schumacher, B.A. "Comparison of soil Gas, Heated Headspace, and Methanol Extraction
Techniques for Soil Volatile Organic Compound Quantification", 12th Annual Waste Testing & Quality Assur-
ance Symposium, July, 1996
4. Minnich, M.M., "Behavior and Determination of Volatile Organic Compounds in Soil A Literature Review",
EPA600/R-93/140, 1993
5. Pignatello, J.; Xing, B. "Mechanisms of Slow Sorption of Organic Chemicals to Natural Particles", Environmen-
tal Science & Technology, Vol. 30 No. 1, 1996
6. Ball, W.P. et al. "Hot Methanol Extraction for the Analysis of Volatile Organic Chemicals in Subsurface Core
Samples from Dover Air Force Base, Delaware", Water Monitoring & Remediation, Vol. 17, No. 1, 1997
7 Mayotte, T.J. "Observations from Theoretical Interphase Mass Partitioning Modeling and Comparisons with
Analytical Data", Colder Associates Report 993-8491, 26 March, 1999.
Table 1. Soil Physical/Chemical Properties
pH CEC BET C OM Ca Mg
Units s.u. melOOg m2/g % % ppm ppm
Ottawa Sand 7.7 0.1 0.09 <0.1 <0.1 150 7
Ft Edwards Clay 8.3 13 43.7 0.8 1.4 4350 797
Yokene Clay 7.2 35 14.9 2.3 3.9 2705 841
CRREL Silt/Sand 7.5 10 5.2 0.7 1.2 1000 20
166
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Methanol D Fugacity Model • Aqueous
Figure 1
7000 T
6000 --
5000 --
o> 4000 -
= 3000
Vapor Fortification Model
E3Fugacity Projection
iMethanol Extraction
Figure 2
167
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 2. Three Month Aqueous Fortification of soils (ng/g)
Soil Type
cis-1,2-DCE
CRREL
Yokene
Ft. Edw.
Ottawa
trans-1,2-DCE
CRREL
Yokene
Ft. Edw.
Ottawa
Benzene
CRREL
Yokene
Ft. Edw.
Ottawa
TCE
CRREL
Yokene
Ft. Edw.
Ottawa
Toluene
CRREL
Yokene
Ft. Edw.
Ottawa
PCE
CRREL
Yokene
Ft. Edw.
Ottawa
Ethylbenzene
CRREL
Yokene
Ft. Edw.
Ottawa
m+p-Xylene
CIRREL
Yokene
Ft. Edw.
Ottawa
P&T Low
28.5
102.5
139.5
163.0
P&T Low
57.3
144.3
191.8
214.3
P&T Low
108.0
145.0
188.3
209.0
P&T Low
520.3
153.5
209.5
224.3
P&T Low
84.5
147.3
204.3
218.7
P&T Low
104.8
114.0
119.6
173.7
P&T Low
78.0
110.3
160.3
200.7
P&T Low
57.5
101.5
154.8
191.3
DayO
P&T High
18.8
104.5
113.0
163.8
P&T High
21.8
135.0
141.0
201.5
P&T High
92.5
133.5
145.5
202.5
P&T High
436.3
139.8
153.0
216.5
P&T High
90.5
162.3
179.0
215.5
P&T High
97.8
139.3
116.0
170.3
P&T High
90.0
127.5
123.0
176.0
P&T High
74.5
131.8
129.5
185.8
Day 0
HSLow
53.2
116.0
137.0
179.0
HSLow
24.3
87.8
113.0
145.0
HSLow
92.0
122.0
153.0
174.0
HSLow
379.0
140.0
179.0
219.0
HSLow
67.8
128.0
175.0
202.0
HSLow
102.0
109.0
154.0
184.0
HSLow
63.8
92.3
148.0
184.0
HSLow
44.2
89.0
155.0
188.0
DayO
HS High
71.0
132.0
176.0
185.0
HS High
71.1
109.0
165.0
162.0
HS High
96.4
130.0
174.0
175.0
HS High
460.0
161.0
215.0
224.0
HS High
85.5
144.0
189.0
196.0
HS High
124.0
131.0
185.0
172.0
HS High
84.1
118.0
145.0
159.0
HS High
67.6
113.0
158.0
160.0
Day 2
HS Sonc
114.0
137.0
176.0
174.0
HS Sonc
118.0
115.0
171.0
157.0
HS Sonc
120.0
134.0
175.0
171.0
HS Sonc
735.0
166.0
216.0
213.0
HS Sonc
118.0
145.0
186.0
191.0
HS Sonc
139.0
133.0
188.0
173.0
HS Sonc
106.0
117.0
141.0
157.0
HS Sonc
98.5
118.0
151.0
158.0
Day 1
HS Sonc
130.0
137.0
176.0
181.0
HS Sonc
131.0
157.0
HS Sonc
127.0
172.0
HS Sonc
897.0
213.0
HS Sonc
127.0
189.0
HS Sonc
143.0
167.0
HS Sonc
112.0
156.0
HS Sonc
110.0
159.0
Day 2
HS Sonc
159.0
181.0
HS Sonc
136.0
159.0
HS Sonc
148.0
176.0
HS Sonc
1350.0
213.0
HS Sonc
153.0
193.0
HS Sonc
162.0
172.0
HS Sonc
124.0
162.0
HS Sonc
122.0
162.0
168
-------
01
<0
539 DAY VAPOR FORTIFIED CRREL SOIL STUDY
S/V1MS Revision - Corrected lor Stamiirt DHtarance* i
48
PT
HS 30 min
TWmln
HS4l»ur
L
1
Pf4rtour
HSIdty
TIMy
HS1SOOC
____.
FTltont
HS4d*y
»T4dty
HSTdoy
PTTday
48 16 day
PT14(tay
48 23 div
PTJIday
48S2d«y
PTJ2d«y
r
TDCE
123
6.13
B
141
44.30
A
285
38.68
•
290
2838
a
864
130.50
3
779T
6239
6
1MOI
23459
C
2402
206.37
Ct.9
1940
21548
C
2450~"
Z73J>1
466
38 18
C
447
10.79
C
763
43.10
d,ejf
820
20.79
d,a,f,g
776
31.34
iStf
813
26.39
d,a,f,fl
|27|
2274
d.e.l,g
875
16.6S
g
849
26.10
is
871
24.66
la
1003
0.00
n
813
39.17
f.fl
1363
64.29
i
971
104.43
h
1517
65.57
i 1906
9623
i k
TCE
197
B.39
a i
221
37.16
a
336
2702
a
254
32.53
«
647
62.27
C.
4aT
54.84
b
1322
83.27
d,e
1329
12878
d,«,T
1368
60.00
W
1244
4464
d,«
1541
61.96
iia.
1456
4678
«,f
1671
98.49
A....
1617
81.01
(
2021
102.14
n
1501
88.37
I
3037
156.31
n
2198
168.82
h
3314
112.40
4278
156,98
k
(ugfto)
to)
301
30.37
a,t>
399
58.00
e,b,c
440
35.66
a.b.c
465
36.47
D,C
646
70.50
d
632
80.34
d
1006
47.26
ej
1235
106-51
I.B.I
1028
66.06
e,l
1137
60.67
o.f.i)
1169
66.06
is
1264
1716
f.a.^
1213
62.45
<,g,n
1336
26.31
g.hj
1410
52.92
hl
1383
270.08
h,|
1959
90-74
t
1781
77,69
k
2106
40.41
k
2807
163,78
Value* wMi common Mnr art not donrMemtly dlffmnt «t ttw S5% confidence Moral (ANOVA Bid Itehere proMcMd LSO)
PCE
738
5.00
*,b
954
121.66
a,b
1213
104.08
c
1152
132.56
»,c
1661
100.66
d.e
1492
180.30
a
2601
140.12
fliL
2526
116-94
fl.h
2548
96.39
z£
2390
60,47
g
2611
60,33
" '
Ijj^h
TOSv
8640
g
2649
51.32
4h
2373
116.50
g
2656
6926
1962
10443
1
3546
126.03
k
1830
255.59
*,l
3824
34.64
k
4789
493.28
E-tai
433
1.15
a,b
563
63,85
a.b.c
676
36.13
t£
730
97.89
C
1006
83.05
d
962
147.73
d
1436
40.00
e
"T762
211-09
1
1457
49.33
B
1606
63.11
e
1544
76.08
e
1S63
57.01
e
1573
41.63
a
1567
24.03
•
1746
04-29
f
!" 1432
33.95
8
2176
98.49
9
1544
137.89
«
2266
116.47
4
3072
435.92
g
D-Xyi
286
2.31
>
i 370
i 60.62
'""
a
539
46.50
b
572
109.89
6
822
93.74
c
""""~83S
269.37
C
1250
26.17
d
1495
63.65
e,f
1269
70 £4
d
"""1456
29.67
e.f
1 1364
I 36.06
IdA. .
T 1496
83.66
e,f
1383
45.83
a.e.t
1456
29.67
»,f
... 1666
fBioe
Si.
"Issa"
39.32
d,e
2096
M-^
B_.
1516
97.09
0,1
2120
121.24
2927
458.00
g
.„,__ — ..,
9-Xvt
672
5.00
a,b
910
101.80
fl.b
1134
55.66
"-"TiTg-
171.64
B,C
1446
105.83
d
1460
293.79
a
; 1867
65.68
Jf
.. -4 2*«
277.16
g
1680
76.38
a
2142
102.21
f.h
2108
112.69
LL
2166
60.36
f,h
2158
S5-6B
i.h
2162
12.34
f,n
£346
76.3£
i*-h
1974
62.63
e.f.h
271!
160.73
y.
2183
130.09
fl.h
273C
117,90
L
3609
407.78
j
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
PAH SEPARATION AND DETECTION BY GC/FID. BRINGING METHOD 8100 INTO THE 90'S.
Dennis R. Gere and Alan D. Broske
Hewlett-Packard, 2850 Centerville Rd., Wilmington, DE 19808
Linda Green
ChromaSkills, Inc., PO Box 80558, Baton Rouge, LA 70898-0558
(504) 928-3087; fax (504) 928-0278
Email: Leslie@chroma-skills.com
Gail Reed
Virginia Tech University, Dept. of Chemistry Hahn Hall, Blacksburg, VA 24061
(540) 231-8253; (540) 231-3255
Email: greed@vt.edu
Method 8100 Revision 0 (the original version) was promulgated in September of 1986. This method was based
upon a packed column GC method that did not have the capability of adequately resolving 4 critical pairs of PAHs.
We will describe part of a two-year study, which has resulted in a revised capillary GC method. We will describe
the multilab validation, which was carried out on an optimized PAH, GC capillary column method. The result was
12 methods which provided adequate resolution of PAHs regulated by the US EPA, including the four pairs of
compounds not resolved in the original Method 8100. Further, 4 of the methods will yield results adequate for the
EPA method 8270 (semi-volatiles) where a GC/MS is used. Using arbitrarily defined as the best set of conditions
(best method) in one of the 12 methods, we placed 10 columns in three different laboratories and carried out a
round robin study. The results of this round robin will be presented.
Column#
1-gr3
2-gr4
3-gr1
4-gr2
5-lg2
6-lq1
7-dg1
8-dg2
9-dg3
10-lg3
Mean
Standard Deviation
% RSD
Retention time
Peakl
4.273
5.148
4.311
4.342
4.306
4.273
4.417
4.340
4.379
4.311
4.410
0.342
7.7
Retention Time
Peak 16
42.143
45.138
42.143
42.343
42.143
42.143
42.663
42.229
42.430
42.143
42.625
1.487
3.5
Resolution
Factor 11 -12
1.58
1.65
1.59
1.62
1.58
1.58
1.60
1.56
1.61
1.61
1.60
0.033
2.0
Resolution
Factor 14-1 5
3.47
3.29
3.50
3.54
3.54
3.53
3.46
3.41
3.36
3.48
3.46
0.090
2.6
The table shows only a brief amount of the data acquired in the different labs. We will describe the features and
benefits of this revision of Method 8100 in detail, including some comparisons between the HPLC method and the
GC (GC/MS) methods with detection limits and other data.
EXTRACTION OF DIESEL RANGE ORGANICS (DRO) AND WASTE OIL ORGANICS (WOO) FROM SOILS
AND SEDIMENTS: EXPANDING METHOD 3545A (PRESSURIZED FLUID EXTRACTION)
Bruce E. Richter
Dionex, SLCTC, 1515 W. 2200 S., Suite A, Salt Lake City, UT 84119
801-972-9292 phone
bruce_richter@worldnet.att.net
Method 3545A specifies the use of Pressurized Fluid Extraction (PFE) for the extraction of organic compounds
from soils and other solid wastes. This technique uses conventional liquid solvents at elevated pressures and
170
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
temperatures to obtain rapid and complete extractions. PFE has been compared to Soxhlet and ultrasonic
extraction for the extraction of compounds covered by RCRA, and in all cases, PFE gives equivalent or superior
results. Currently, Method 3545A covers the following compounds: bases/neutrals and acids (BNAs), organo-
chlorine pesticides (OCPs), polychlorinated biphenyls (PCBs), organophosphorus pesticides (OPPs), chlorinated
phenoxy herbicides, polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) from
soils, clays, sediments, sludges, and solid wastes. However, validation data for the extraction of hydrocarbons
have not been submitted for inclusion in Method 3545A. The purpose of this presentation is to discuss the method
development results and validation data for the extension of Method 3545A to diesel range organics (DRO) and
waste oil organics (WOO).
Prior to this work, data had been reported1 showing that two or more complete extractions with PFE were needed
to get quantitative recovery of TPH from wet clay samples when using IR-transparent solvents such as
perchloroethylene (PERC). For the validation data set, we did not want to use two or more extractions, and we
wanted to use a GC method for the determinative step. This necessitated the use of other solvents and operating
conditions. As part of the method development, six different solvents were investigated: hexane, heptane,
methylene chloride, and 1:1 mixtures of each of these solvents with acetone. Temperatures ranging from 100 °C
to 200 °C were investigated. Wet as well as dry samples were investigated. The conditions that were identified
from the method development phase are methylene chloride/acetone (1:1), 175 °C with 15 mL total solvent and 15
minutes total time for 10-g samples. The recovery from certified soils using these conditions averaged 116% with
3.6% RSD (relative standard deviation).
After the method development phase, we conducted a validation phase. This consisted of two parts. First, we
determined the bias and precision of the method using three different matrices (clay, loam, and sand) and at two
different concentrations (5 and 2000 mg/kg). These samples were spiked with both #2 diesel and 30w motor oil.
GC was the determinative step in all cases. Sample extracts were treated with standard clean up procedures
using silica gel and Na2SO4 and concentrated to 1 mL. Second, portions of real-world samples were extracted
using PFE, automated Soxhlet, and ultrasonic extraction. In all cases, the bias and precision using PFE were
comparable or superior to the results obtained using the other techniques. The complete data set will be
discussed in this presentation.
These data have been submitted for consideration to be included in Update IVb of the SW-846 Methods Manual
so that Method 3545A can be extended to include DRO and WOO.
Reference
1. M.L. Bruce, Proceedings of the Eleventh Annual Waste Testing and Quality Assurance Symposium, American
Chemical Society, 1995, 114-120.
THE ANALYSIS OF CARBAMATES USING LC/MS
Jim Krol
Sr. Applications Chemist
Eric Block
Sr. Mass Spec Applications Chemist
Michael Young
Sr. Applications Chemist
Mark Benvenuti
Applications Chemist
Joe Romano
Environmental Market Manager
Waters Corporation, 34 Maple St, Milford, MA 01757
Office 508/482-2131, Email Jim_Krol@Waters.com.
ABSTRACT
The analysis of carbamates has received renewed interest recently in light of their implication as potential
endocrine disrupters, and their use as common pesticides for food products. Before their use, carbamates must
be manufactured from various raw materials that are themselves potential endocrine disrupters, and the
171
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
manufacturing waste must be characterized prior to disposal. All total, carbamates and related products are enter-
ing into eco-system with potential adverse effects.
The EPA Office of Solid Waste has recently published a final rule covering the analysis of 40 carbamate waste
constituents.1 To monitor all these currently requires 6 different analytical methods from GC to LC. Several of the
listed carbamate methods utilize mass spectrometry detection.
This presentation will discuss the analysis of several commonly used carbamates using HPLC-Positive
Electrospray Mass Spectrometry, a preliminary new method similar to Method 8321A using thermospray mass
spec detection. Linearity will be demonstrated from 5 to 1000 ppb which covers the general calibration range, with
an LOD of 1 to 2 ppb and with %RSD of response of less than 10%. Supporting the utility of the LC/MS method is
the analysis of drinking water, including spiked carbamate recovery, and a vegetable matrix.
This work will shed insight into how new mass spectrometry technology can be applied to enhance the monitoring
of environmental carbamate pollutants as well as other organics.
INTRODUCTION
Carbamates are commercially available pesticides derived from carbamic acid. Highly effective and having a
broad spectrum of activity, carbamates are used worldwide to protect crops and other vegetation from the ravages
of insect pests.
Carbamates, their intermediaries, their degradation products, and their metabolites are of great concern to
members of the regulatory and scientific communities as more and more drinking water sources are testing
positive for the presence of carbamates. They find their way into the aquifers and surface water through agricul-
ture runoff after being directly applied to food crops such as grains, fruits, and vegetables. If food crops are
harvested too soon after application, residues and their byproducts may remain on the produce. Additionally
buyers of grain, fruits, and vegetables are becoming increasingly vigilant for pesticide residues due to their toxic
nature.
In an effort to protect drinking water resources, the US Environmental Protection Agency and other international
governing bodies now regulate pesticide use and require routine monitoring of drinking and raw source water. This
effort has been extended to solid waste products such as soil and hazardous waste disposal, all of which could
potentially comtaminate the drinking water supply.
EXPERIMENTAL
In this study, various instrumental conditions were examined and optimized for the analysis of a 10 carbamate
component standard mixture without the use of pre or post column derivatization.
System: Waters Alliance® LC/MS with MassLynx™ system control & data processing
Mass Spec: Waters ZMD Detector (4000 amu mass range)
MS Interface: Positive Electrospray (ESI+)
Column: Waters Symmetry® Ci8, 1 mm x 150 mm
Temperature: 35° C
Mobile Phase: Linear Gradient from 10%-80% MeOH in 10 mM NH4OAc
Flow Rate: 75 uL/min
Injection Vol: 10 uL/min
Analysis Time: 18 minutes
For comparison purposes, a similar carbamate standard mixture and samples were analyzed using post column
derivatization, the accepted method of analysis.
System: Waters Alliance® System for Carbamate Analysis and
Millennium32 system control & data processing
Column: Waters Carbamate Column, 3.9 mm x 150 mm
Temperature: 30° C
Mobile Phase: Multistep Gradient using MeOH / AcCN / Water
Flow Rate: 1.5 mL/min
Injection Vol: 400 uL/min
Post Column: Dual post column reaction with NaOH and OPA @ 0.5 ml_/min
172
-------
Detection:
Analysis Time:
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Fluorscence, 339 nm excitation, 445 nm emission
30 minutes
The 10 carbamates plus beta Naphthol working standards were prepared from an AccuStandard (New Haven,
Conn.), M531 Carbamate Mixture, 0.1 mg/mL concentration in AcCN. Dilutions were made with 1 mM HCI (pH 3)
for the post column method, and 100% MeOH for the mass spec method.
DISCUSSION
Figure 1 shows the total ion chromatogram from the full scan analysis of 20 ng (10 uL of 2 ug/L) of each of the 10
component carbamate mixture. Another carbamate standard at 10 ug on column was analyzed using the post
column fluorescence method is shown in Figure 2. Note that using the MS separation conditions, the first 4 carba-
mates coelute, but are fully resolved with the post column method. For conventional carbamate identification and
quantitation, carbamate resolution is critical. However, for mass spectrometry, resolution is not critical.
By extracting from the full scan, the [M+H]+ or [M + NH4]+ ions specific to each coeluting compound, individual
chromatograms can be resolved by the mass spectrometer, as shown in the insert of figure 1. This allows for other
unknown carbamates or organics in a complex matrix, such as wastewater or solid waste, to be selectively
detected, identified, and quantitated without the need for chromatographic resolution using complex gradient
methods.
Figures 3 and 4 show the full scan mass spectra of aldicarb sulfoxide and aldicarb sulfone. Note that although
similar in structure and in retention characteristics, they give unique mass spectra. The mass spectrometer was
optimized using the cone voltage programmability capability of the Waters ZMD for response of the carbamate
[M+H]* ion, except for oxamyl and aldicarb where the carbamate [M + NH4]+ ion was used. This feature allows the
operator to change ionization cone voltage to maximize the response, and to switch between positive and negative
electrospray within the same run, although negative electrospray was not used for this study.
The carbamate mixture was re-analyzed and acquired in the SIR (single ion recording) mode, where the MS
detector was set to detect only a single [M+H]* ion value. Each chromatogram only shows the single, individual
carbamate in the mixture, and demonstrates the selectivity of mass spec detection. Concurrently, acquisition in the
SIR mode also enhances sensitivity. This is a primary benefit of mass spec detection, which is shown in figure 5.
A series of 6 carbamate working calibration standards between 5 and 1000 ng/mL (ppb), representing between 50
and 10,000 pg on column, were analyzed in triplicate, and calibration curves generated using SIR response and a
1/x weighting. The 1/x weighting was used to minimize the statistical effect of the higher concentrations on the
linear regression. Figure 6 shows the calibration curve for Carbofuran, a carbamate that coelutes with propoxur.
Again, this demonstrates that resolution is not as important with MS detection as it is with conventional detection.
The coefficient of determination for the weighted regressions is given in Table 1.
Table 1. Coefficient of Determination (r2) Linearity
Carbamate
Aldicarb Sulfoxide
Aldicarb Sulfoxide
Oxamyl
Methomyl
3-OH Carbofuran
Coeff. Of Determination
0.9969
0.9982
0.9990
0.9959
0.9970
Carbamate
Aldicarb
Propoxur
Carbofuran
Carbaryl
Methiocarb
Coeff. Of Determination
0.9963
0.9963
0.9981
0.9994
0.9995
The lowest carbamate calibration standard, 5 ng/mL (ppb) or 50 pg on column, was analyzed 5 times to calculate
the limit of detection, defined as 3 times the standard deviation, the limit of quantitation, defined as 10 times the
standard deviation, and the precision, defined as %RSD = (mean)(100)/std dev). This data is tabulated in Table 2.
In real samples such as drinking water and vegetables, carbamates are typically present at concentrations near
the limit of detection described above. Solid waste and aqueous samples are typically extracted using a methylene
chloride liquid-liquid partitioning, and the methyene chloride is taken to dryness and resolubilized with methanol.2
An alternative sample prep enrichment, eliminating the use of methylene chloride, was employed for a carbamate
recovery from a typical drinking water sample.
173
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 2. Mass Spec Carbamate Sensitivity and Precision
Carbamate
Aldicarb Sulfoxide
Aldicarb Sulfone
Oxamyl
Methomyl
3-OH Carbofuran
Aldicarb
Propoxur
Carbofuran
Carbaryl
Methiocarb
Limit of Detection
ng/mL (ppb)
0.8
1.8
0.7
1.6
0.4
0.2
0.7
0.9
0.3
0.4
Limit of Quantitation
ng/mL (ppb)
2.6
6.1
2.2
5.5
1.4
0.5
5.5
3.0
1.1
1.4
Response at 50 pg
%RSD
3.5
9.6
3.2
10.0
2.1
0.7
11.3
7.5
1.8
2.0
Milford, Mass drinking water was spiked with 500 ng/L (500 ppt) of each carbamate. Two hundred and fifty (250)
mLs was passed through a Waters Oasis® HLB Cartridge to retain the carbamates. The carbamates were eluted
from the cartridge using 6 ml of 10% MeOH / MTBE. This solution was taken to dryness and resuspended in 1 mL
of acetonitrile, which was used for MS analysis. This represents a 250-fold enrichment. The same sample was
analyzed using the conventional post column fluorescence method. The recovery data and method comparison
data are given in Table 3.
Table 3. Comparison of MS and Post Column (PCFD) Method Carbamate Recovery
Carbamate
Aldicarb Sulfoxide
Aldicarb Sulfone
Oxamyl
Methomyl
3-OH Carbofuran
Aldicarb
Propoxur
Carbofuran
Carbaryl
Methiocarb
Mass Spec
% Recovery
74.8
88.7
83.2
92.3
101
79.4
103
95.6
97.7
81.2
Mass Spec
%RSD
i 19
16
18
8.0
8.6
9.3
13
7.5
14
14
PCFD
% Recovery
54.7
98.7
90.8
99.9
98.7
90.7
97.5
97.2
89.6
91.6
PCFD
%RSD
0.5
4.0
7.0
6.4
2.3
9.3
5.6
4.7
2.2
2.2
The mass spec carbamate recoveries observed in this experiment are consistent with the recovery data using
method 8231A published in SAIC Carbamate Method Evaluation Report.3 Although the mass spec recoveries are
lower than those determined by the conventional post column florescence method, this data indicates that both the
mass spec method and the conventional post column fluorescence method yield acceptable results.
A solid vegetable sample, bell pepper, was prepared by an outside source using the State of Florida
Modified-CDFA Multiresidue Method effective as of July 1998.4 Fifty grams of vegetable is homogenized and
extracted with acetonitrile. This extracted is passed over an aminopropyl solid phase extraction cartridge, the efflu-
ent taken to dryness and resolubilized in 1 mL of methanol.
The same extracted bell pepper sample was analyzed using both the LC/MS method and the PCFD method for
comparison purposes. The PCFD method was performed several days after the LC/MS analysis. The results are
summarized in Table 4, and the SIR chromatograms shown in figure 7 through 9.
These analyses indicated that the solid phase extraction procedure and both methods yield acceptable results.
174
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 4. Analysis of Bell Pepper, data expressed as ppb
Methomyl
PCFD
46.0
411.5
32.5
ND
ND
LC/MS
46.5
342.5
40.5
ND
ND
Oxamyl
PCFD
ND
ND
ND
32.4
54.1
LC/MS
ND
ND
ND
48.0
76.0
Carbaryl
PCFD
ND
ND
ND
14.2
136.7
LC/MS
ND
ND
ND
13.5
154.5
Propoxur
PCFD
293.8
319.7
298.6
280.6
290.5
LC/MS
276.5
323.5
241.0
263.5
341.5
CONCLUSION
These data indicate that the use of positive electrospray mass spec detection is a viable technique for the analysis
of carbamates in drinking water and on vegetables. The detection limit, precision, and %recovery show that this
electrospray method gives equivalent results to the thermospray method described in 8321 A. However, for best
sensitivity, the conventional post column fluorescence method, providing detection below 0.5 ug/l (ppb), is the
technique of choice.
REFERENCES
1. Emergency Revision of the Land Disposal Restrictions (LDR) Treatment Standards for Listed Hazardous
Wastes from Carbamate Production. (63FR4740), Sept 4, 1998.
2. Method 8318, N-MethylCarbamates by High Performance Liquid Chromatography (HPLC). Rev 0, Sept 1994.
3. Carbamate Method Evaluation Report, prepared by SAIC, Dunn Loring VA, 22027, Aug 20, 1998, prepared for
B. Lesnik, EPA Office of Solid Waste.
4. Chemical Residue Laboratory, Florida Dept of Agriculture & Cons Services, Florida Modified-CDFA Multiresi-
due Method. Rev 4, SOP# CR Method 42, July 1998.
Figure 1: Carbamates, Mass Spec
Total Ion Chromatogram
Figure 2: Carbamates, Post Column Fluorescence
Ctiromatogram
Figure 3: Background-subtracted full-scan ESI* Spectrum
of Mdlcerb Sulfonide (Peak 1 in Figure 1)
2O7|M
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Figure 5: Consult* SIR Chromrtoy»m» or 10 CvfewMta
SensKMly: M pg On-Cohimn
iml MIBtiirt »^t
™1 i
"i]
"-".
aj» u» TH» injn 1200 nj» iftjo m>
Figura 0: Carbofuran callbratfon cum; 5-1000 ngftnL
Figure 7: Carbamate Analysis of Bel P»pper
Oxamyl
1
Figure 8: Carfaamtto Analyoto of Ball Peppar
ktottioniyl
Figure 9: Carbamate Analysts of Bel Pepper
Carbaryl
KwatemM A
1»n«»C»«a
a*.
NOVEL BIOSENSORS FOR CHARACTERIZING ENVIRONMENTAL ENDOCRINE CHEMICALS
Omowunmi A. Sadik*. Sharin Benda, Miriam Masila, Fei Van, Jenny Krautova
Department of Chemistry, State University of New York-Binghamton, FOB 6016, Binghamton, NY 13902
Corresponding Author: Tel- (607) 777-4132, e-mail: osadik@binghamton.edu
Accumulating evidence strongly indicates that certain pesticides, environmental pollutants, industrial chemicals
and naturally occurring phytoestrogens can dramatically alter normal physiological functioning of the endocrine
system. Obtaining information on which chemicals in the environment should be labeled as EDCs is critical,
requiring a significant amount of time and efforts. This poses a great challenge to current methodologies such as
GC/MS due to their cost and long turn-around time.
Several factors have been identified for making the monitoring and surveillance studies of EDCs difficult. First, the
176
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
cost and time involved in screening (or testing) a wide variety of synthetic EDCs and their metabolites. Secondly,
EDCs generally exhibit an unusual dose-response behavior unlike most toxic substances. The dose-response
curves for most toxic substances usually increase with increasing levels of the chemical compound and eventually
levels off. Whereas, the response curves for environmental estrogens exhibit an inverted U-shape and the
greatest response is produced at extremely low doses. Thirdly, there are concerns that the effects of different
EDCs are additive or even synergistic and hence regulations concerning individual compounds may not be
adequate. Consequently, scientists and regulators are looking for very simple, fast and least costly screening
methods that are capable of identifying and classifying all EDC as well as potentially contaminated sites within the
environment. Once screening methods are developed, further laboratory re-analysis, confirmation and long-term
studies would be necessary to identify endocrine disrupting characteristics among the population.
We have utilized the concept of renewable inummosensing and electrochemical microassembly technologies to
analyze a range of suspected EDCs, including polychlorinated biphenyls (PCBs), chlorinated phenols, atrazines,
and heavy metals. Experiments performed with PCB antibodies resulted in a detection limit of 0.1 ng/ml for
selected Aroclors, for a total analysis time of about 20 minutes. This paper discusses the potential of affinity
biosensors and immunocytochemistry for characterizing potential EDCs.
THE THEORY OF OPERATION AND APPLICATIONS OF THE PULSED FLAME
PHOTOMETRIC DETECTOR (PFPD) FOR GAS CHROMATOGRAPHY
Norman A. Kirshen
Varian, Inc., 2700 Mitchell Drive, Walnut Creek, CA 94598
Introduction
The Pulsed Flame Photometric Detector (PFPD) was developed in the early 1990's by Dr. Aviv Amirav.1'3 Unlike
the traditional flame photometric detector which has a continuous flame, the PFPD is based on a pulsed flame for
the generation of flame chemiluminescence. The detector operates with a fuel rich mixture of hydrogen and air.
This mixture is ignited and then propagates into a combustion chamber three to four times per second where the
flame front extinguishes. Carbon light emissions and the emissions from the hydrogen/oxygen combustion flame
are complete in two to three milliseconds, after which a number of heteroatomic species give delayed emissions
which can last from four to 20 milliseconds. These delayed emissions are filtered with a wide band pass filter,
detected by an appropriate photomultiplier tube, and electronically gated to eliminate background carbon emission.
Twenty-eight elements can be detected with the PFPD, thirteen of which give delayed emissions, and therefore
infinite selectivity. These latter elements include environmentally and industrially important S, P, As, Sn, and N.
Applications of the PFPD hi the Sulfur mode for the analysis of sulfur compounds in petrochemical products as
well as in beverages are shown. Several petrochemical applications of interest are follows: 1) thiophene in
benzene, 2) sulfur gases in natural gas, and 3) COS in propylene. The Phosphorus mode of operation is very
sensitive and is applicable to the detection of organophosphorus pesticides. The use of the PFPD as an elemental
specific detector used in concert with a mass spectral detector is shown to be very helpful in providing additional
information to more easily identify target pesticides.
High speed data acquisition firmware and software enables one to easily set up the PFPD and to review the
pulsed emission data emanating from each chromatogram. This allows the qualitative confirmation of target
compounds. Dual channel data processing also provides the ability to qualitatively analyze two elemental modes
simultaneously.
Experimental
In a conventional flame photometric detector (FPD), a sample containing heteroatoms of interest is burned in a
hydrogen-rich flame to produce molecular products that emit light (i.e., chemiluminescent chemical reactions). The
emitted light is isolated from background emissions by narrow bandpass wavelength-selective filters and is
detected by a photomultiplier and then amplified. The detectivity of the FPD is limited by light emissions of the
continuous flame combustion products including CH*, C2*, and OH* Narrow bandpass filters limit the fraction of
the element-specific light which reaches the PMT and are not completely effective in eliminating flame background
and hydrocarbon interferences. The solution to this problem, conceived by Professor Amirav of Tel Aviv University
was to set the fuel gas (H2) flow into the FPD so low that a continuous flame could not be sustained. But by
177
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
inserting a constant ignition source into the gas flow, the fuel gas would ignite, propagate back through a quartz
combustor tube to a constriction in the flow path, extinguish, then refill the detector, ignite and repeat the cycle.
The result was a pulsed flame photometric detector (PFPD) shown in figure 1.
Vent
Ignitor
Ignitor
Chamber
Combustor
Chamber
Needle Valve
The background emissions from the hydrogen-rich airhydrogen flame
(approximately 10 mL/min H2 and 40mL/min Air) is a broad band
chemiluminescence. The combustion of hydrocarbons is highly
exothermic, rapid and irreversible, producing a light emission by the
hydrocarbon products equal to the time for the flame to propagate
through the combustor or 2 to 3 milliseconds. Many of the chemilumi-
nescent reactions of other elements such as S (S2*), P (HPO*), N
(HNO*) etc., are less energetic and more reversible, and proceed
after the temperature behind the propagating flame has dropped.
These heteroatom emissions are therefore delayed from the
background emissions. By using the leading edge of the flame
background emission to trigger a gated amplifier with an adjustable
delay time, heteroatomic emissions can be amplified to the virtual
exclusion of the hydrocarbon background emission. The selective
amplification of the element-specific emissions is the basis of the
PFPD's unique sensitivity and selectivity (see figure 2).
Figure 1. Schematic Cross Section of the PFPD
The PFPD pulses approximately 3 times per second so
that in a period of about 330 milliseconds the detector fills
with the mixture of fuel gases and column effluent. When
the flame propagates through this mixture, all the light
emission from a given flux of some element, sulfur, for
example, is concentrated into a period of only 20 millisec-
onds following each flame pulse. This light intensity is
approximately 16 times brighter than the steady state
emission from a conventional FPD where the emission
would be spread over a period of 330 milliseconds. This
effect plus the fact that the gated amplifier is only active
during a 20 millisecond period for sulfur combines to
greatly improve the signal to noise ratio in the PFPD.
Figure 2. Flame Background and Sulfur Emission Time
Profiles
OH*, CH
6 8 10 12
TIME (Milliseconds)
14 16 18
PFPD EMISSION
EARLY-OH*. CH
Of equal importance is the ability to resolve the emissions of the heteroa-
toms from the flame background. The delayed sulfur or phosphorus
emissions are integrated after the flame background has dropped to a
negligible level. This delay permits the use of much wider bandpass optical
filters that no longer must filter the background but can be selected to
target the wavelength range of the desired element specific emissions.
The result is lower overall noise levels and therefore greater detectivity.
PFPD Specifications
The PFPD detects 28 elements:
S, P, N, As, Sn, Se, Mn, B, Br, Ga, Ge, Pb, Si, Te, V Al Bi Cr Cu,
Eu, Fe, Ni, Rh, Ru, W, In, Sb
Figure 3. Hydrocarbon and Sulfur Emission Profiles as a function of
Wavelength. Filter used for S is the BG-12.
600 500 400 300
WAVELENGTH (nm)
200
178
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Thirteen of these elements have delayed emissions from the background Carbon emission and therefore exhibit
infinite selectivity:
PFPD Element Selective Detection s, P, N, AS, se, sn,
Ge, Ga, Sb, Te, Br,
Cu, In
Sensitivity (MDL) (pgX/sec)
Detectivities and selec-
tivities are shown in
Figure 4.
Control of the PFPD
parameters is either
from the GC or worksta-
tion (Figure 5).
Cm
Tb
Bk
Dy
Cf
Ho
Es
Er
Fm
Tm
Md
Yb
No
Lu
Lr
Figure 4. The PFPD
periodic chart: Detectivi-
ties and Selectivities of
Elements
ftort
&
1
• '2'
3
4
__5_
Adlut
PN
Betectej Middle Detecta
MK.CMNiot.2it)* | PFPD _
TenvewlwelGJ J300 •
•'.. f&Kie j Hangs
Initial 10 i
1
tRWw
jtamuWpliei Voltage (V> ^01
Sate Dd^i (msec): 6
Gate Width (msecl 20
Tfissw level (mV): 20
Rear Detector i
»J Detector Overe f*1 On
•M Electronics-. P On
Square Root Mode: (* On
1 Autozeio —
KJyes
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
1
2
3
1. Hjdrogw Sttfde
2. Carbonyl yfide
3. Methyl Mercaptan
llbthsMdttw
4
*
lH 20 3D 40 50 60 70 90 S>0 100 110
Figure 6. 1 ppm ea. Suflur Gases in Natural Gas
Figure 7. COS in Propylene
-40 ppb
COS
~6ppb
H,S
Propylene
elutes
Flame
propagation
disrupted
Unidentified
mercaptans and sulfides
I I I r I
Column: 30m x 0.53 mm GSQ Carrier: H2 @ 28 cm/sec.
200 uL injected via SPI
J
C
Figure 9.
Phosphorus
Pesticides in a
Lettuce Extract
Matrix at 10 ppb
1. Dimethyl Sulfide
2. Diethyl Sulfide
3. ?
4. Dipropyl Sulfide
5. ?
6. ?
7. 3-methylthio-1-propanol
Figure 8. Sulfur Compounds in Beer
by SPME
180
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
The PFPD has good arsenic selectivity and detectivity that allows it to be used for the monitoring of catalyst
poisoning gases such as arsine. The PFPD
can also simultaneously detect AsH3 and
PH3 as shown in Figure 10 where propylene
is the major component.
Phosphinc Det. Limit = 2.0 PPB
I
Arsine Del. Limit = 8.4 PPB
0.96 ppm
Analytical Software
Analytical software available for the PFPD
permits one to view the emissions of the
PFPD on a scope like window (Figure 11).
This allows for quick set up and optimization
of the detector flows. It also allows the user
to view the emission profile of the
background and eluting peak for qualitative
information. Finally, the emission data from
the complete chromatogram can be saved
as a data file and viewed (Figure 12). The
resulting data may also be manipulated to
provide dual elemental chromatograms.
Figure 10. Simultaneous Determination of
PH3 and AsH3 in Propylene
Figure 11.
Conclusions
The PFPD is a highly sensitive and selective flame photometric detector capable of detecting 28 elements, 13 of
these with infinite selectivity. Analytical software is capable of providing elemental qualitative information from the
pulse emission data for one or two channels. The detector has many applications is the petrochemical, industrial,
environmental and food industries.
References
1. Amirav, "Pulsed Flame Detector Method and Apparatus". USA, Patent No. 5153673, Israel Patent No. 95617,
European patents approved, Japan patent pending.
2. Atar, S. Cheskis and A. Amirav, "Pulsed Flame - A Novel Concept for Molecular Detection", Anal.Chem., 63,
2061-2064(1991).
3. Cheskis, E. Atar and A. Amirav, "Pulsed Flame Photometer - A Novel Gas Chromatography Detector", Anal.
Chem., 65, 539-555(1993).
181
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Figure 12. PFPD Data
File Viewer Showing
Phosphorus and Sulfur
Chromatograms
H 1M 125 1» 17S
52S »« !K «0 4M «SO <» 5SO 5M 6SO 57S B»
SIMULTANEOUS MEASUREMENT VOLATILE AND SEMIVOLATILE COMPOUNDS:
INTRODUCING METHODS 3611 AND 3670
David Mauro
META Environmental, Inc., 49 Clarendon Street, Watertown, MA 02172
Tel: 617-923-4662; Fax: 617-923-4610
Stephen Emsbo-Mattingly
META Environmental, Inc., 49 Clarendon Street, Watertown, MA 02172
Tel: 617-923-4662; Fax: 617-923-4610
The simultaneous measurement of volatile and semivolatile compounds will be embraced by the USEPA with the
release of SW-846 later this year. Methods 3511 and 3570 provide protocols for the extraction of waters and soils,
respectively, developed by the Electric Power Research Institute (EPRI) and used for a decade at former
manufactured gas plant (MGP) sites around the country. With few exceptions, these methods simultaneously
extract volatiles (8260), semivolatiles (8270), polycyclic aromatic hydrocarbons (8100), pesticides (8081),
polychlorinated biphenyls (8082), chlorophenols, total petroleum hydrocarbons (8015B), volatile petroleum
hydrocarbons (MADEP), extractable petroleum hydrocarbons (MADEP), and any other organic compound
extractable by dichloromethane. These microextraction methods are fast, flexible, field compatible, inexpensive,
sensitive, and suitable for GC/FID, GC/PID, GC/ECD, and GC/MS. These methods can be substituted for many
sites requiring Method 5035 for methanol preserved volatiles in soil. The method description, validation data, and
application history will be presented.
A COMPARISON OF STATIC HEADSPACE AND SOLID-PHASE MICROEXTRACTION
FOR THE DETERMINATION OF VOLATILE ORGANICS IN WATER
Norman A. Kirshen and Zelda Penton
Varian, Inc., 2700 Mitchell Drive, Walnut Creek, CA 94598
182
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Introduction
EPA method 5021 is a static headspace (SHS) method for volatiles in soils. This method is part of the SW-846
manual and may be used with either GC or GC/MS. It covers a wide range of compounds over a concentration
range of 20 to 200 parts per billion (PPB). The method lists 58 compounds from dichlorodifluromethane to
naphthalene. Solid-Phase Microextraction (SPME) is a relatively simple technique where an adsorbent coated
fused silica fiber may be exposed to the headspace above an aqueous sample for a fixed period of time. The fiber
is then injected into and desorbed in the injection port of a GC or GC/MS. The objective of this initial study was to
compare various quantitative parameters of these two techniques to determine the possible applicability of SPME
for the determination of VOC's in soils. Soils were not used in these initial studies, only the matrix modifiers.
The behavior of organic standards extracted from a matrix modifier are compared for a set of model compounds
using the headspace technique and three different SPME fibers. This study was performed with a CTC
Combi PAL AutoSampler which is capable of headspace and SPME analysis. Headspace samples of 200 ml
were extracted and injected after shaking a sample for 30 minutes at 60°C. SPME fibers were exposed to the
sample headspace for 30 minutes with shaking at 40°C and then desorbed for 5 minutes. A Saturn ion-trap
GC/MS was used for chromatographic analysis.
The following studies and results are shown: 1) a comparison of the absolute response to gaseous VOCs with
SPME and SHS showing improved response with SPME, 2) a comparison of mid to late eluting VOCS showing
variable response comparisons between SHS and the three SPME fibers, 3) linearity studies indicating that the
SPME fiber can become overloaded if too many compounds are adsorbed simultaneously, 4) good area count
precision exhibited for selected compounds for all SPME and SHS, and 5) calculations of minimum detectable
quantities using SPME, SHS, and Purge and Trap proving that SPME has comparable detection limits for many
analytes and may be an acceptable alternative for VOC's in soils or waste water.
Experimental
CTC Combi PAL AutoSampler equipped with 10 ml vial tray:
Headspace injection: Heated headspace gas tight syringe (1.0 mL)
• 5 mL sample equilibrated In 10 mL vial with shaking for 30 minutes at 60°C
• Injection volume: 200 mL
SPME injection: SPME fiber holder and fiber
• Headspace: 5 mL in 10 mL vial
• Adsorption: 30 minutes with shaking at 40°C, desorption 5 minutes
• Fibers: PDMS (100 mM); Carboxen™ 1006-PDMS; Carboxen™ 1006-DVB-PDMS
Varian Saturn 2000 GC/MS:
• Column: 60M x 0.32mm DB-624
• Injection: Splitless at 230°C
• Ion-trap: 150°C
In the EPA method 5021 static headspace method, a soil sample is added to a 22 mL vial containing 10 milliliters
of matrix modifier. After this vial Is sealed, internal and surrogate standards are injected. In the present study
10 mL vials were used with 5 mL of modifier.
Standards were added to the vial via syringe injec-
tions through the septum cap prior to analysis. Vial
were then loaded into the Combi PAL 10 mL vial
tray. The autosampling procedure was then started.
During the sampling process each sample is trans-
ferred to the incubator for heating and mixing (by
shaking) for 30 minutes at 60°C (see figure 1 and
2). While the sample continues to be heated and
mixed, a gas tight syringe withdraws 200 mL which
is then injected into the GC/MS injector.
Figure 1. Conventional Static Headspace Sampling
After equilibrium
Original sample
183
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Figure 2.
Combi PAL in
SPME Mode
Moving syringe holder
Incubator for heating and mixing
Sample tray
Solid phase microextraction is an equilibrium technique where analytes are not completely extracted from the
headspace matrix. The recovery is dependent on the partitioning or equilibrium of analytes among the three
phases present in the vial: the aqueous sample, the headspace, and the fiber coating on the fused silica needle.
Adsorbent coatings can be films (30-100 uM) of polymer, copolymer, carbonaceous adsorbent, or a combination
of these. The coated fused silica fiber is attached to a metal rod and the entire assembly comprises the SPME
syringe assembly. The fiber is within a protective sheath in the standby mode. The sheath is pushed through the
vial septum by the autosampler and lowered into the headspace. The fiber is then inserted into the headspace and
adsorption is commenced. Following this step, the fiber is pulled back into the sheath, withdrawn from the
autosampler vial and injected into the GC injector (see figures 3 and 4). In the present studies fibers were held in
the 5 ml headspace for 30 minutes at 40°C.
absorb
desorb
Figure 3. SPME Headspace Sampling
Figure 4. SPME Adsorb and Desorb Steps
Results and Discussion
Following the procedures outlined above, 200 ug/L solutions of three gaseous VOCs were prepared by spiking 5
mL of martrix modifier in each of several 10 mL septum capped vials. These analytes were then analyzed by both
headspace and SPME. Three different fibers were used in the SPME experiments. The results are shown in
Figure 5. The Carboxen SPME fiber gave the best response to the three VOC's followed by the headspace
technique. The PDMS fiber is ineffective with the low boiling VOC's while the three-phase fiber Carboxen-DVB-
PDMS, falls in the middle.
184
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
70000
60000
40000
30000
DSHS
• Carfaoxen
• PDMS
• Tri-phase
Chlorome thane
Chloroethene
Figure 5. Absolute
Response to Gaseous
VOC's with Headspace
and SPME
This same type of study was performed with higher boiling VOC's, the results shown in Figure 6. The Carboxen
fiber is effective in adsorbing the higher boiling VOC's. The three-phase fiber also adsorbs the higher boiling
analytes efficiently.
Figure 6. The
absolute
Response to
Later Eluting
VOC's with
Headspace
and SPME
3500000
3000000
2500000
2000000
1500000
1000000
500000
DSHS
• Carboxen
• PDMS
• Tri-phase
1,2,3-TrichloroPropane
1,2-dibromo-3-chloropropane 1,3-butadiene,1,1,2,3,4,4 hexachlorine
Early SPME linearity studies with 60 target compounds of EPA method 5021 indicated that the fiber was being
185
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
saturated. When the number of compounds screened was reduced to 20, this observation was confirmed (see
figures 7 and 8).
2500000
2000000
1500000
1000000
SOOOOO
Carboxen fiber
60 compounds
0 20 40 60 80 100 120 140 160 180
Jnearity studies were conducted
From 3.3 to 176 ppb with two test
mixtures, one containing 60 com-
pounds and the other containing
20 compounds. With the 60 com-
pound standard, linearity flattens
aut as the concentrations in-
crease. With the 20 component
-mixture, linearity is quite good for
Doth compounds over the tested
•ange. These results are consis-
ent with the supposition that the
ibers will saturate if too many
components occur at these levels
n the headspace.
Figure 7. SPME Calibration
Curves for
frichlorofluoromethane
Statistical data was collected for a
representative list of VOCs from
the EPA method 5021 target com-
pound list. The results for SPME
(Carboxen) and static headspace
are shown in Table 1.
Purge and trap studies for the
same compound list were also
performed on a separate system.
The minimum detectable quanti-
ties of these compounds were
determined using a S/N equal to
five as an estimation. These re-
sults are shown in Table 2.
Figure 8. SPME Calibration
Curves for 1,1,2-Trichloroethane
7000000
6000000
3000000
Carboxen fiber
20 compounds
60 compounds
20
60
80 100 120 140 160 180
Table 1. Precision for Selected VOC's at 33 ppb, n=4
Compound %RSD SPME %RSD Headspace
Vinyl Chloride
CCI3F
TCE
1,1,2-TCA
1,2,3-TCP
4.3
3.4
1.5
2.0
2.2
5.4
8.0
2.4
2.4
6.2
Minimum detectable quantities obtained for SPME are comparable or lower than those obtained by static
headspace.
186
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 2. Minimum Detectable Quantities of Selected VOC's, S/N=5, Average of n=4
Compound SPME-Car (ppb) Headspace (ppb) P and T (ppb)
Vinyl Chloride
CCI3F
TCE
1,1,1-TCA
1,2,3-TCP
0.2
0.1
0.01
0.1
0.03
4.5
0.8
0.5
2.8
3.3
0.06
0.02
0.02
0.2
0.09
Conclusions
The results of these preliminary studies comparing SPME to static headspace indicate that SPME could be a
reasonable alternative for determining volatiles in water or soils. SPME could also make an excellent screening
tool prior to Purge and Trap analyses.
Linearity experiments gave good results although it is possible to saturate the SPME fiber if too many compounds
are being monitored simultaneously.
Precision was good and the detection limits of SPME are lower than headspace.
While these studies were carried out with a matrix modifier, the results would be expected to be similar to those
obtained with soil samples.
EVALUATION OF A VACUUM DISTILLER FOR PERFORMING METHOD 8261 ANALYSES
Michael Hiatt
National Exposure Research Laboratory, US Environmental Protection Agency,
944 East Harmon Ave., Las Vegas, NV 89119
702-798-2381
hiatt.mike@epamail.epa.QOv
The Environmental Protection Agency's Office of Research and Development has developed a vacuum distillation
method to determine volatile organic compounds in difficult matrices. The developed method is intended for use
by both the Superfund and RCRA Programs and incorporates a novel approach to establish data quality. The
resultant method (SW-846 Method 8261, Update IVB) uses surrogate compounds to measure matrix effects and
to compensate for their biases.
This poster presents the results of an evaluation of the first commercial version of the ORD vacuum distillation
apparatus, the VD1000 (produced by Cincinnati Analytical Instruments, Cincinnati OH under a license agreement
with the Environmental Protection Agency). The vacuum distiller combined with an HP 5972 GC/MS is tested for
compliance with calibration criteria identified within Method 8261. In addition, method detection limits for the
vacuum distiller for water, soil, and fish tissue are presented. The potential for contamination of samples by a high-
concentration sample (by each Method 8261 analyte) is also presented as the percent of a high concentration
standard detected in a subsequent blank.
The review of Method 8261 analytical data is simplified by graphical presentation of method performance and the
impact that a sample matrix imparts on analyte recoveries. The use of surrogates to monitor matrix effects on a
given sample provides the means to determine analyte recovery as a function of critical chemical properties and to
present the functions graphically. Additional check surrogates are monitored to quantify the accuracy of the recov-
ery functions. Examples of data and graphic presentations for review are presented.
Sample throughput is evaluated on the basis of average Superfund sample sets. The comparison of analytical
costs associated with Method 8261 and current Superfund requirements are presented. Estimates of how the use
of Method 8261 could impact Superfund is presented.
187
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Notice
The U. S. Environmental Protection Agency (EPA) through its Office of Research and Development (ORD),
funded this research and approved this abstract as a basis for a poster presentation. The actual presentation has
not been peer reviewed by EPA. Mention of trade names or commercial products does not constitute endorsement
or recommendation by EPA for use.
METHOD 8261: USING SURROGATES TO MEASURE MATRIX EFFECTS
AND CORRECT ANALYTICAL RESULTS
Michael Hiatt
National Exposure Research Laboratory, US Environmental Protection Agency,
944 East Harmon Ave., Las Vegas, NV 89119
702-798-2381
hiatt.mike(5)epamail. epa.gov
The Environmental Protection Agency's Office of Research and Development has developed a vacuum distillation
method to determine volatile organic compounds in difficult matrices. With the understanding that such a method
would be intended for use by both the Superfund and RCRA Programs with a need to establish data quality, a
novel approach to optimize QA requirements is incorporated. The resultant method (SW-846 Method 8261,
Update IVB) uses surrogate compounds representing the range of chemical properties of the method's analytes in
order to measure matrix effects and to compensate for their biases. Method 8261 eliminates the need for matrix
spike/matrix spike duplicates as well as calibration of instrumentation by matrix type. This poster presents the
theory behind the surrogate corrections incorporated within the method.
There are primarily three main chemical properties of volatile organic compounds that define their behavior and
recovery during vacuum distillation. These properties are the compounds' vapor pressure (measured as boiling
point, BP), partition coefficient between air and water (Kaw), and partition coefficient between an organic phase and
air (Kao). By adding surrogate compounds to measure recoveries as a function of these properties, the impact of
any matrix (e.g., biota) on recovery of analytes is predicted. The measurement of matrix effects by sample elimi-
nates the need for matrix spike/matrix spike duplicates as well as the need to calibrate instrumentation by matrix
(i.e., Method 5030 for water and Method 5035 for soil).
The impact of Method 8261 corrections allows for an expanded list of analytes that include the volatile organic
compounds (VOCs): polar compounds such as dioxane and pyridine, the nitrosamines and aniline, and
compounds that are considered semi-volatile such as naphthalene. With the streamlining of analytical require-
ments and expanded analyte list, the productivity of using Method 8261 is greatly superior to alternative methods.
The measurement of matrix effects by sample simplifies the review of Method 8261 analytical data. The relation-
ship of the chemical properties (BP, Kaw, and Kao) to recovery are displayed graphically. The mysterious and
dubious "matrix effects" disclaimer provided by other methods when an analysis does not perform as anticipated is
not a hindrance to Method 8261. Extreme matrix effects are accurately compensated. Additional use of "check"
surrogates allows the evaluation of matrix corrections effectiveness. Examples of data and graphic presentations
for review are presented.
Notice
The U. S. Environmental Protection Agency (EPA) through its Office of Research and Development (ORD),
funded this research and approved this abstract as a basis for a poster presentation. The actual presentation has
not been peer reviewed by EPA. Mention of trade names or commercial products does not constitute endorsement
or recommendation by EPA for use.
188
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
APPLICATION OF A DIOXIN/FURAN IMMUNOASSAY KIT TO FIELD SAMPLES
Robert O. Harrison
CAPE Technologies, L.L.C, 3 Adams St., South Portland ME 04106
Robert E. Carlson
ECOCHEM Research, Inc., 1107 Hazeltine Blvd., Chaska MN 55318
Several immunoassay screening methods have been approved under the 4000 series of Field Screening Methods
within SW-846 and are now widely used in the field. In less than a decade since their introduction, the commercial
immunoassay kits behind these methods have significantly changed the process of assessment and remediation
of hazardous waste sites. The major advantages include dramatically accelerated turnaround, decreased cost,
and improved statistical reliability of site assessments because of the greater number of samples analyzed.
For the first time, these advantages are now also available to dioxin analysts. A previously described enzyme
immunoassay (EIA) for polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans (PCDD/Fs) has been
applied to a variety of soil, sediment, and ash samples. Several simple sample preparation methods have been
developed specifically for use with this immunoassay. The resulting protocols can be used in a simple field
laboratory without extensive equipment. The methods are simple enough to be learned quickly without
face-to-face training.
Sample throughput for a single analyst can be 15 or more samples per 8 to 10 hour day, which includes 4 hours
for incubations during which no handling is needed. Use of an optional overnight incubation allows staggered
processing of multiple batches. This approach can boost productivity to 30 or more samples per 12 to 14 hours
over 2 days.
Results from several comparison studies will be described which demonstrate strong correlation between EIA
results and TEQ, ranging from low ppt to high ppb levels. Special attention will be given to the data package for
Method 4025 (Dioxin in Soil by Immunoassay). Issues relating to implementation of this technology in both fixed
and field lab situations will be discussed, including QA requirements and limitations of the method. Selected
customer experiences will be used to demonstrate related points, such as data interpretation and troubleshooting.
The most recent results from an ongoing program of method improvement will be described.
VOLATILE AND EXTRACTABLE PETROLEUM HYDROCARBONS:
A ROUND ROBIN ILLUSTRATES ESSENTIAL PBMS STANDARDS
Stephen Emsbo-Mattingly. M.S.
Chairman, Technical Committee Independent Testing Laboratory Association, and Laboratory Director
META Environmental, Inc., 49 Clarendon Street, Watertown, MA 02172
Tel: 617-923-4662; Fax: 617-923-4610
John Fitzgerald, P.E.
Deputy Regional Engineer
MADEP Bureau of Waste Site Cleanup, 205A Lowell Street, Wilmington, MA 01887
Tel: 978-661-7702; Fax: 978-661-7615
The Massachusetts Department of Environmental Protection (MADEP) formally released the Volatile Petroleum,
Hydrocarbon (VPH) and Extractable Petroleum Hydrocarbon (EPH) methods in January, 1998. These methods
offer lexicologically meaningful replacements for traditional measurements of Total Petroleum Hydrocarbons
(TPH) which employ infrared (IR) or gas chromatography (GC) techniques. The VPH and EPH methods
simultaneously measure two range aromatics, four range aliphatics, and twenty-three individual, aromatics
including BTEX, MTBE, and PAHs. These methods are also two of the first finalized and widely employed
"performance based" methods. Unlike traditional EPA protocols, analytical modifications are allowed, provided
specific performance criteria are satisfied. Before releasing the VPH and EPH protocols for regulatory monitoring,
the MADEP conducted a round robin in order to evaluate the effectiveness of unmodified and modified versions of
the method.
The round robin results demonstrated a promising performance among participating laboratories, as well as
189
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
several potential problems for interlaboratory comparability. The MADEP used the results of the round robin to
incorporate specific performance criteria for the final release of the methods. These included criteria for sample
preparation procedures, alternative capillary columns and detectors, surrogates, chromatographic integrations,
method proficiency standards, reporting limits, detection limits, report content, and self-certification statements.
These performance criteria guard against analytical short-cuts which compromise the public health orientation of
the VPH and EPH methods.
This presentation will describe the MADEP VPH and EPH methods, the round robin results, several common
method modifications, the importance of reference materials, and the performance criteria necessary for assuring
data accuracy and comparability.
FAST AND EFFICIENT VOLATILES ANALYSIS BY PURGE AND TRAP GC/MS
C. Eric Boswell
National Air and Radiation Environmental Laboratory, Mixed Waste Analytical Program,
540 South Morris Avenue, Montgomery, AL 36115
(334) 270-7071
Boswell.eric@epa.gov
ABSTRACT
Recent changes in environmental regulatory paradigms, such as EPA's performance-based measurement
systems (PBMS), are lowering method compliance barriers for laboratories working under the Resource Conser-
vation and Recovery Act (RCRA). One of the stated goals of PBMS is to educate the regulators and the regulated
community on the inherent and intended flexibility of SW-846 methods. Operating under EPA's PBMS guidelines,
laboratories could employ the flexibility of SW-846 methods to simplify and improve purge and trap GC/MS volatile
organic analyses (P/T GC/MS VOAs). Laboratories performing Method 8260B for P/T GC/MS VOAs have two
basic GC configuration options: wide bore columns connected to the mass spectrometer through a jet separator or
narrow bore columns directly interfaced to the mass spectrometer.
SW-846 methodology recognizes both approaches as valid. The narrow bore column/direct interface approach is
the better of the two techniques for most analyses when certain modifications are made. When newer purge and
trap concentrator designs are employed and when several Method 8260B instrument parameters are modified
dramatic performance benefits result. This "enhanced" narrow bore column/direct interface approach produces
results such as reduced susceptibility to column contamination by high level samples, improved chromatographic
behavior of early eluting and closely eluting compounds, analysis times under 20 minutes, and improved hardware
ruggedness. The outcome is better quality data, higher sample throughput, and fewer instrument mechanical
failures.
INTRODUCTION
Connecting the purge and trap concentrator to the GC inlet is one of the major challenges in P/T GC/MS VOAs.
The challenge stems from vastly different flow rate requirements of the purge and trap concentrator, the capillary
column, and the mass spectrometer. Method 8260B1 describes GC/MS systems equipped with either cryogenic
cooling devices attached to narrow bore (0.25 mm and 0.32 mm) capillary columns or wide bore (0.53 mm) capil-
lary columns connected to enrichment devices such as jet separators. Many laboratories choose wide bore capil-
lary columns with jet separators when running Method 8260B because they can easily accept the high flow rates
required to efficiently desorb the trap. The jet separator provides the necessary decrease in carrier gas flow rate
prior to entering the mass spectrometer. The wide bore column/jet separator approach has been the traditional
approach to P/T GC/MS VOAs for some time. The wide bore column/jet separator approach has a host of
problems. The problems include susceptibility to column contamination by high level samples, poor chroma-
tographic behavior of early eluting and closing eluting compounds, long analysis times (run times approaching 40
minutes), and frailty of the jet separator. Narrow bore capillaries, which potentially offer better chromatography,
have not been used as much for volatiles analysis primarily because they cannot easily handle the relatively high
flow rates coming from the purge and trap concentrator. Method 8260B suggests cryofocusing the analytes on a
capillary pre-column interface situated between the purge and trap concentrator and the GC capillary column. This
device condenses the desorbed sample components and focuses them into a narrow band that can be transferred
190
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
to the analytical capillary column. However, this is an additional capital expense and it adds to the total analysis
time. Newer purge and trap concentrator designs allow a much simpler interface. A conventional split/splitless
injector usually already installed on the GC/MS system can be plumbed in series with the purge and trap concen-
trator. The operating principle is quite simple: the excess flow coming from the purge and trap is vented at the
column inlet allowing a reduction in carrier gas flow rate to one more suitable for high resolution chromatography.
Feyerherm and Neal2'3 have described how this is done with a Hewlett Packard 5890 GC. Aside from this instru-
ment modification, the concentrator desorb time and the GC oven temperature program should be optimized to
improve the chromatographic behavior of method compounds and shorten analysis time. The concentrator desorb
time may be as short as 30 seconds depending on the trap material. Shortening the desorb time reduces the
amount of water transferred to the GC system and thus improves chromatography. The GC oven temperature
program for P/T GC/MS VOAs must accommodate compounds with a relatively wide boiling point range. The
initial oven temperature will determine how well-behaved the gases (Dichlorodifluoromethane, Chloromethane,
Vinyl Chloride, Bromomethane, and Chloroethane) are. Once the compounds are on the GC column, the higher
boilers are not difficult to resolve. Fast (50-60°C/min.) GC oven temperature ramps can be used to save time
without any loss in resolution. This paper describes a series of modifications to Method 8260B for P/T GC/MS
analysis of VOA samples. The method performance has been tested primarily with spiked water (Method 5030) in
a single laboratory.
EXPERIMENTAL
Instrumentation and materials
All work was performed with an Ol (College Station, TX) MPM-16 autosampler/4560 Purge and Trap concentrator.
An Ol tenax, silica gel, and charcoal trap (Ol trap #10) was used as the sorbent trap. To connect the purge and
trap, perform the following operations. Cut the total flow line to the split/splitless inlet about 3 4 cm from the
septum nut. Using a 1/16" stainless steel union, connect the supply end to the "CARRIER IN" fitting on the purge
and trap concentrator. Using another 1/16" stainless steel union, connect the transfer line from the purge and trap
concentrator to the split/splitless GC inlet. These connections allow the use of the GC total flow controller to
control the purge and trap desorb flow rate. All other connections are identical to other purge and trap installations.
Figure 1 contains a plumbing diagram of the purge and trap concentrator-GC inlet connections. A Hewlett Packard
(Palo Alto, CA) 5890 GC with EPC/Hewlett Packard 5971 MSD was employed as the GC/MS system. The analyti-
cal column used was a Restek (Bellefonte, PA) Rtx-5 (30m x 0.25mm x 1.0um) with no guard column. Analytical
standards were purchased from Ultra Scientific (N. Kingstown, Rl) and were prepared by dilution with purge and
trap grade methanol. All samples were 5 ml_ water samples prepared by spiking stock solutions into organic-free
reagent water.
Operating Conditions
The purge and trap conditions and the GC/MS conditions are listed in Tables I and II respectively. After an 11 min.
purge, the trap was heated to 180°C for 0.5 min. for sample desorption. Following the desorption step, the trap
was baked at 200°C for 7.00 min. to complete the autosampler cycle. The injector was operated in the split mode
with PURGE A (or B) ON all the time. A single taper 4 mm ID glass liner without glass wool was used in the GC
inlet. An injector temperature of 200°C produced the best overall results. Liquid nitrogen was used to cool the oven
to the initial temperature of 10°C. The GC temperature was ramped faster at the beginning and at the end of the
GC oven program where the compounds exhibit a wide range of boiling points. The total carrier gas flow was 20
mL/min. and the split ratio was set at 40:1. The column flow was set at 0.5 mL/min (26.2 cm/sec.). We used a
GC/MS interface temperature of 280°C.
RESULTS AND DISCUSSION
BFB may be directly injected to save time, but the injector should be operated in the splitless mode. BFB solutions
are typically made up in methanol. Due to the solvent effect in splitless injections, standards made up in methanol
do not give good peak shapes. Purging the BFB takes a little more time, but solves all of the above problems. We
used a typically short GC oven temperature ramp for the BFB run.
Figure 2 is a total ion chromatogram of a 200 ug/L VOA standard on a narrow bore capillary column/direct inter-
face GC/MS system. The chromatographic run time is 17 minutes with a total GC cycle time of 20 minutes. There
are no noticeable water effects in the chromatogram. Notice the gaussian peak shapes of the five gases (DCDFM,
Cloromethane, Vinyl Chloride, Bromomethane, and Chloroethane). The gases give an indication of the system's
overall chromatographic performance. These compounds are usually difficult to separate and typically produce
poor peak shapes on 0.53 mm column/jet separator systems. Ethyl Benzene and the m,p-Xylene pair which are
typically unresolved on a 0.53 mm column. Styrene and o-Xylene usually coelute on a 0.53 mm column. We
achieved baseline resolution on Ethyl Benzene and the m,p-Xylene pair and partial resolution on Styrene and
191
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
o-Xylene. Because of the large number of analytes, we do have several resolution challenges. Bromochlorometh-
ane and Chloroform coelute at 6.3 min. Bromoform elutes between Styrene and o-Xylene at 11.2 min. and is diffi-
cult to see on the total ion chromatogram. A similar close elution occurs with sec-Butylbenzene and 1,3-
Dichlorobenzene at 13.93 and 13.94 min. None of the coeluting targets share common ions so their ion chroma-
tograms are easily identified and quantified. For our system, a 0.5 min. desorb time dramatically reduced the
amount of desorbed water while giving good chromatographic responses. With a tenax, silica gel, and charcoal
trap, all compounds easily desorb at 180°C within 30 seconds with minimal carryover into subsequent blank water
QC samples. For a tenax, silica gel, and charcoal trap, the purge flow rate didn't seem to affect chromatographic
peak responses as much as other parameters. Purge and trap valve and transfer line temperatures around 100°C
gave better results than hotter temperatures in the 180-200°C range. There was no apparent condensation of the
higher boiling volatiles in the 100°C transfer line. The trap bake time was set so the purge and trap cycle time
corresponded to the 20 minute GC cycle time.
The narrow bore capillary column system was calibrated by running a five-point curve with standards at 10, 20, 50,
100, and 200 ug/L (50, 100, 250, 500, and 1000 ng of standard injected). Table III is a summary of mean relative
response factors (RRF), percent relative standard deviation (%RSD), method detection limits (MDLs), and
estimated quantitation limits (EQLs) for selected compounds. Three of the four ketones (acetone, 2-butanone, and
4-methyl-2-pentanone) exhibit typically low RRFs, but the overall purging efficiencies are comparable to other
methods. The linearity data of Table III suggest that a wider calibration range is possible for most of the VOA
targets. The MDL and EQL data exhibit exceptional sensitivity for 5 mL samples. These data reflect a very simple
and robust system that can generate accurate and reproducible results.
The one potential disadvantage to this approach is the requirement for sub-ambient GC oven cooling to reach the
initial temperature of 10°C. This could be overcome by choosing a different GC column or a thicker film.
CONCLUSIONS
Performing split injections with P/T GC/MS VOAs allows a narrow bore column to handle the relatively high flow
rates coming from the purge and trap concentrator. Narrow bore columns can be interfaced to purge and trap
concentrators via split/splitless injectors by performing a relatively simple hardware modification. Combining this
hardware modification with method optimizations in the concentrator desorb time and the GC oven temperature
program produces dramatic performance improvements. This easy alternative to the more traditional wide bore
column/jet separator approach to P/T GC/MS VOAs results in reduced susceptibility to column contamination by
high level samples, improved chromatographic behavior of early eluting and closely eluting compounds, analysis
times under 20 minutes, and improved hardware ruggedness.
Heated Transfer Line
Septum
Purge
Carrier Gas
Supply
P&T
Figure 1. Basic plumbing diagram for a back pressure regulated split/splitless injector with a P/T autosampler.
192
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Figure 2. Total ion chromatogram of a 200 ug/L VGA calibration standard using a narrow bore capillary/direct
interface system.
REFERENCES
1. USEPA SW-846 "Test Methods for Evaluating Solid Waste," Method 8260B Volatile Organic Compounds By
Gas Chromatography/Mass Spectrometry (GC/MS), December 1996.
2. Feyerherm, Fred; Capillary Direct VOA's, Hewlett Packard Company, Houston, TX, 1991.
3. Neal, Barney; EPA Volatiles Analysis Using Narrow Bore Capillary Columns, Hewlett Packard Company,
Analytical Education Center, Atlanta, GA, 1992.
TABLE I. Purge and Trap Conditions.
Trap Material
Sample Volume
Purge Flow Rate
Purge Temperature
Purge Time
Desorb Temperature
Desorb Time
Bake Temperature
Bake Time
Valve Temperature
Transfer Line Temperature
Tenax, Silica Gel, and Charcoal (Ol trap #10)
5mL
40 mL/min.
ambient
11 min.
180°C
0.5 min.
200 °C
7 min.
100°C
100°C
193
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
TABLE II. GC/MS Conditions
Injector Mode
GC Inlet Liner
GC Inlet Temperature
Total Flow
Septum Purge
Column
Column Linear Velocity
GC Oven Ramp
GC-MS Interface temperature
Scan Range
Solvent Delay
Split
Single taper,
4 mm ID, no glass wool
200 °C
20 mL/min.
3 mL/min.
Rtx-5, 30 m x 0.25 mm, 1 urn film
26.2 cm/sec.
Hold 2.0 min
10-90°C
90-140°C
140 - 240 °C
Hold 1.5 min
(0.5 mL/min.)
@10°C
@ 20 °C/min.
@ 6 °C/min.
@ 60 °C/min
@ 240 °C
280 °C
35-300 amu
2 min.
TABLE III. Summary of mean RRFs, %RSDs, MDLs, and EQLS for selected compounds.
COMPOUND
Chloromethane
Bromomethane
Acetone
2-Butanone
Chloroform
Benzene
Ethyl Benzene
1 ,3-Dichlorobenzene
Hexachlorobutadiene
1 ,2,3-Trichlorobenzene
MEAN RRF
1.09818
0.48876
0.07834
0.09202
0.95677
1.50253
1.71616
1 .46463
0.41990
0.83833
%RSD
1.336
1.467
2.717
4.150
2.073
3.086
1.807
2.014
4.389
4.532
MDL (ppb)
0.14
0.36
1.23
1.10
0.18
0.05
0.06
0.12
0.19
0.35
EQL (ppb)
0.45
1.20
4.11
3.66
0.60
0.18
0.21
0.40
0.64
1.17
A NEW APPROACH FOR HIGHLY COMPLEX ORGANIC ANALYSES USING
SIMULTANEOUS SELECTED ION AND FULL ION SCANNING
Elaine A. LeMoine and Adam Patkin
The Perkin Elmer Corporation, 50 Danbury Road, Wilton, CT 06897-0219
ABSTRACT
Regulated semivolatile organic compounds are present in our environment at widely diverse concentrations with
an equally disparate range of sensitivity requirements. Nowhere is this more evident than in hazardous waste and
site remediation samples. The number of target analytes in a Gas Chromatography/Mass Spectrometric (GC/MS)
analysis is one of the most extensive. The ability to identify and quantify all of these compounds with one analysis
pushes the method and technology to extremes. In most instances additional preparatory work and analyses are
required to address the range of concentrations and the number of analytes.
Full scan ion monitoring is by far the most prevalent mode of operation for these types of samples. It covers the
necessary mass range and provides classical spectra that can be library searched for positive identification.
Selected Ion Recording (SIR also called SIM) is a GC/MS mode of operation where only the ions of interest are
monitored, independent from surrounding interferences and coelutions, providing dramatic increases in sensitivity.
194
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
SIR has not been widely accepted in environmental testing labs. It requires prior knowledge of the sample matrix.
Only targeted masses are detected and all others go unreported.
The combination of simultaneous Selected Ion and Full Ion (SIFI) scanning is a new approach providing the
advantages of both techniques in a single chromatographic run. Using SIFI, samples comprised of analytes with
wide response variations can be identified and quantified in the same chromatographic run. Analytical problems
due to interferences can be greatly minimized. During a chromatographic analysis using the full scan mode,
analytes requiring very low level detection in a complex matrix can be quantified using the added sensitivity of
Selected Ion Recording (SIR) and still display library searchable spectra obtained from the full scan mode. Both
the full scan and SIR scan functions are combined into one analysis providing low level detection in complex matri-
ces while retaining all the functionality of full scan for the more responsive analytes.
This paper presents semivolatile organic data from analyses using a simultaneous Selected Ion and Full Ion (SIFI)
scanning method. It also discusses some of the many productivity gains possible.
INTRODUCTION
There are well over 100 compounds that can be detected using EPA Method 82701. Some compounds are
particularly sensitive to these analytical conditions and respond with a strong signal while others are more difficult
to detect, particularly at low levels. Calibration standards are generally at the same concentrations for the majority
of target analytes. This often requires a system optimized for detection of the less sensitive compounds, while
sacrificing signal at the higher concentrations for the more responsive ones. Consequently, compounds that don't
have rigorous sensitivity requirements, are quantitated at unnecessarily low levels forfeiting quantitation in the
more useful upper ranges.
By implementing selected ion recording only where the additional sensitivity is needed, the analytical range of the
remaining analytes is not affected. Additionally, the compounds monitored utilizing selected ion recording can use
the full ion scan if higher concentrations are encountered. This does not require an additional analysis, since both
scan modes are implemented at the same time.
EXPERIMENT AND RESULTS
Six standards containing a number of semivolatile organic compounds listed in EPA method 8270 were analyzed
using the GC/MS conditions listed in Tables 1 & 2. The selected compounds represent the various sensitivities
encountered when analyzing for semivolatiles using GC/MS. Standards representing final concentrations of 1, 2, 5,
10, 20, and 40 ppb were selected as reasonable levels reflecting a range of sufficient sensitivities for soils and
groundwater. Each contained internal standards at a constant concentration of 40 ppb.
Table 1. Chromatography Conditions.
Gas Chromatograph:
Column:
Oven Temperature Program:
Capillary Splitless Injection:
Programmable Pneumatic Control
(PPC):
Split Vent:
Injection Volume:
Perkin-Elmer AutoSystem XL
PE-5MS 30 m x 0.25 mm; 0.25 mm film
thickness
40 °C for 1 min;
45 °C/min. to 160 °C for 3 min;
6 °C/min to 320 °C for 2 min
250 °C
Helium @ 1.0 mL/min.
Splitless -1.00 to 1.00 min; Split 20 mL @ 1.00
min.
1.0 |JL
The specific analytes, Estimated Quantitation Levels (EQLs) as listed in EPA Method 8270, along with the
regulated drinking water Maximum Contaminant Levels (MCLs) and the Maximum Contaminant Level Goals
(MCLGs) are listed in Table 3. Method 8270 is obviously not intended to be sensitive enough for drinking water
determinations, but appropriate for more complex matrices. We will demonstrate how compounds with widely
varying sensitivity requirements can be combined into one analysis for project specific needs. This is in no way
intended to recommend mixing drinking water samples with more highly contaminated samples, but sensitivity
195
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
requirements can vary from compound to compound for any particular project, requiring greater method flexibility.
A method combining full ion and selected ion scanning provides this flexibility and is illustrated in Figure 1.
Function 1 contains the full scan parameters. After a 4-minute filament delay, scanning of m/z 45 to m/z 450
proceeds for the duration of the chromatographic run. Function 2 contains selected ion recording (SIR) scanning
parameters specific to 2-methyl-4,6-dinitrophenol. Mass 198 is monitored at this compound's expected retention
time. Function 3 contains similar SIR information for pentachlorophenol using m/z 266 and function 4 at m/z 149
for bis-(ethylhexyl)phthalate. All four functions are executed during each analysis.
Table 2. Mass spectrometer method.
Perkin-Elmer TurboMass Mass Spectrometer
Mass Scan Range:
Scan Time(sec):
Inter-scan Delay (sec):
Filament Delay:
Ion Source Temperature:
Transfer Line Temperature:
lonization Mode:
Full Scan
Monitoring
Function 1
45 -450 m/z
0.30
0.20
Selected Ion Monitoring
Function 2
198 m/z
0.02
0.02
4 min
175°C
275 °C
El
Function 3
266 m/z
0.02
0.02
Function 4
149 m/z
0.02
0.02
Table 3. Analytes, Method 8270 EQLs, drinking water MCLs and MCLGs.
Analytes
2-Methyl-4,6-dinitrophenol
Acenphthene-dlO(ISTD)
Aminobiphenyl
Anthracene
Benz(a)anthracene
Benzidine
bis-(ethylhexyl)phthalate
Bromophenyl phenyl ether
Butyl benzyl phthalate
Chrysene
Chrysene-d12(ISTD)
Dibutyl phthalate
Dichlorobenzidine
Dimethylaminoazobenzene
Fluoranthene
Hexachlorobenzene
Naphthalene-d8 (ISTD)
Pentachlorophenol
Perylene-d12(ISTD)
Phenanthrene
Phenanthrene-d10 (ISTD)
Pyrene
EPA Method 8270
EQLs
(Ground water ug/L)
50
20
10
10
10
10
10
10
10
20
10
10
10
50
10
10
Drinking Water
MCL
(M9/L)
6
1
1
Drinking Water
MCLG
(mg/L)
zero
zero
zero
196
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
ft'
Full Scan 4
2 SIR of 198.00, (Ek)
3 SIR of 266.00 , (EI+)
4 SIR of 149.00, (Ek
Figure 1. Simul-
taneous Selected
Ion and Full Ion
Scanning.
The percent Relative Standard Deviations (% RSDs) were calculated for every compound using all six calibration
standards. This was demonstrated for one of the compounds for which full ion scanning and selected ion scanning
was implemented, bis-(ethylhexyl) phthalate. Table 4 lists the initial calibration results for this compound. Relative
Response Factors (RRFs) and %RSDs are presented for both full ion and selected ion scanning. Each demon-
strates compliance with the 30% RSD maxima method criteria. However, the selected ion mode of operation can
be used to identify and quantify at much lower levels if and when necessary.
Table 4. Initial Calibration using full ion monitoring and selected ion recording.
Initial Calibration for bis-(ethylhexyl)phthalate
Standards
1 ppb
2ppb
5 ppb
10 ppb
20 ppb
40 ppb
Average RRF
% RSD
Full Ion Scanning (RRFs)
0.794
0.801
0.850
0.737
0.864
1.272
0.886
19.987
Selected Ion Scanning
(RRFs)
42.342
43.397
44.162
38.044
43.873
63.550
45.895
17.775
Figure 2 shows the total ion chromatogram (TIC), the extracted ion current (EIC), and the selected ion recording
(SIR) from one analysis. The signal produced in the SIR mode for bis-(ethylhexyl)phthalate at 1ppb is approxi-
mately four times that of the extracted ion in the full scan mode as measured by signal-to-noise.
197
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
0510_4
100n
S/N:PtP-200.99
bis-(ethylhexyl) phthalate
Selected km Recording
i I 11 i 11 | i i i i l
S/N:PtP-48.70
4: SIR of 1 Channel EI+
1,67e6
0510_4
100n
Extracted Ion Current
1: Scan EI+
3.13e4
42
Total Ion Current
23.00 23.50
24.00
24.50
25.00
25.50
I ' ' i ' I '
26.50 27.00
27.50
1: Scan EI+
TIC
2.02e5
28.00
Time
Figure 2. SIR and full scan signal-to-noise comparison.
The spectra obtained from these traces are displayed in Figure 3. The SIR mode provides increased sensitivity by
scanning for longer periods of time on a few specific masses, in this case m/z 149, while the full scan mode
produces a spectra which can be library searched for accurate identification.
OS10_4 210 (24 190) Cm [205:2)6)
1001
4: SIR of 1 Channel EH
05IO_42315<24.1B5)
1001
Selected km Moratomg
1:ScanEH-
2.4Se4
Ful Ion Mentoring
uo
,*«
IfiB — '9"
Z2B — 340 .28 — 262 — BIB — 320 _ 3*0
380 400
Figure 3. Spectra and signal intensity for bis-(ethylhexyl)phthalate using SIR and full scan mode.
198
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
05 ID.
100-
*•
a
UHethythtx
4
H
yt)pMlutete
4. SR ol 1 Chanrml Bf
11 149.00
183.8
U
SO 100 1S\ 2DO #0 300 3SO 400
2200 2260
2350 / MOO 2460 250D
Figure 4. Both SIR and EIC yield a sufficient number of scans across the peak.
There must be a sufficient number of scans across each peak to ensure accurate integration. Figure 4 displays
the SIR and the EIC for bis-(ethylhexyl)phthalate from one analysis. The scan number can be read across the
x-axis below each peak and clearly shows more than enough scans, assuring accurate integration.
SUMMARY
A GC/MS analysis can be optimized to take advantage of the highest concentrations necessary for all the analytes
using the full scan mode, while at the same time increasing the sensitivity only where it is needed using selected
ion recording. Compounds can be identified and quantified in either scan mode in the same analysis. This added
flexibility can expand the overall analytical range across a widely disparate group of compounds by selectively
choosing when and where increased sensitivity is needed.
REFERENCE
1. Test Methods for Evaluating Solid Waste, Physical/Chemical Methods, SW-846, 3rd edition; Final Update III;
Method 8082, rev 0, Dec 1996.
DOES CHEMICAL IONIZATION HAVE A FUTURE IN THE ENVIRONMENTAL LABORATORY?
Elaine A. LeMolne, Adam Patkin and Herman Hoberecht
The Perkin Elmer Corporation, 50 Danbury Road, Wilton, CT 06897-0219
ABSTRACT
Electron Capture Negative Chemical lonlzatlon Mass Spectrometry (EC NCI MS) is a technique which, for the
most part, has been overlooked by the environmental testing community. It offers advantages in selectivity and
sensitivity for a variety of environmental applications.
Pesticide and PCB analyses are routinely performed using an electron capture detector (ECD) or electron ioniza-
tion mass spectrometer (El MS), each of which offers unique advantages and disadvantages. ECDs afford greater
selectivity for the identification of halogenated compounds along with increased sensitivity, although confirmatory
analyses are required. While El MS offers a higher confidence level for accurate compound identification, it is
199
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
usually at the expense of sensitivity. EC NCI MS combines many of the advantages from each of these
techniques. As with ECD applications, chemical ionization offers a high degree of selectivity for specific classes of
compounds while affording even greater sensitivity.
El MS can produce highly fragmented spectra with a concomitant reduction of base peak intensity. Alternatively,
EC NCI uses a moderator gas for low-energy electron capture as with an ECD. This leads to minimal fragmenta-
tion, maximizing signal intensity. Sensitivity gains can also be realized by scanning for longer periods of time on a
few specific masses, using Selected Ion Recording (SIR also called SIM).
NCI using Full Ion scanning provides enhanced sensitivity and characteristic spectral patterns yielding valuable
compound information. At the same time, Selected Ion scanning affords even further sensitivity gains. The acqui-
sition of data using simultaneous Selected Ion and Full Ion (SIFI) scanning in the El NCI mode combines the
advantages of both techniques for a dramatic increase in sensitivity. Figure 1 displays the chromatograms of an
El NCI simultaneous SIFI analysis.
PCB 1254
ncirepie Selected Ion Chromatogram a: SIR of 1 channel ci-
19.22
10CH
] m/z 394
19.1
,qo
ncirepiG Extracted Ion Chromatogram 1922 1:ScanS4
1001 I 19.93
%4 m/z 394 17K7 J I
1 Ji 1B.33)H
Qq ,....,,'... i -. .'^f,-'Vh y.^i JV-.-..,-.
ncirepie 1: Scan Cl-
1QO Total Ion Chromatogram 16.26 TIC
1 ' 16.71
16.17 H A •-..- |3.^^ 1Q no
15.07 M .All Al7,67 N ' 19'93
LHI.,..._. . -,
f r*rM^ WH* 17 'i^ i i') i i ri i | i i i i i i i i i i i i i i i i i i i ) i i i i | i Time
12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00
Figure 1. Aroclor 1254 acquired with NCI and simultaneous SIFI
The results of PCB analyses using NCI and simultaneous SIFI scanning will be shown. Sensitivity data is
presented as well as a discussion on the merits of the use of this technique in the high throughput environmental
laboratory.
INTRODUCTION
To date, most environmental labs have determined PCB concentrations by averaging the areas of a few selected
characteristic peaks from a total ion Chromatogram. It is generally recognized that the measurement of the PCB
individual biphenyl congeners provides greater overall quantitative accuracy. It is preferable for determining PCB
containing samples with unrecognizable chromatography patterns such as highly weathered samples and samples
containing multiple Arochlors. SW-846 states "The PCB congener approach affords greater quantitative accura-
cy...the congener method is of particular value in determining weathered Arochlors." Pesticide and PCB analyses
are routinely performed using an electron capture detector (ECD) or electron ionization mass spectrometer (El
MS), each of which offers unique advantages and disadvantages. ECDs afford greater selectivity for the identifica-
tion of halogenated compounds along with increased sensitivity, although confirmatory analyses are required.
While El MS offers a higher confidence level for accurate compound identification, it is usually at the expense of
sensitivity. Chemical Ionization Mass Spectrometry (Cl MS) combines many of the advantages from each of these
techniques. As with ECD applications, chemical ionization (Cl) offers a high degree of selectivity for individual
classes of compounds while affording even greater sensitivity. These features make chemical ionization an excel-
lent candidate for low level biphenyl identification and quantification.
EXPERIMENT and RESULTS
The list of biphenyl congeners is quite extensive. For the purposes of this paper we will focus on ten congeners
representing each successive level of halogenation. Standards containing the specific congeners listed in Table 1
were analyzed using electron ionization and negative chemical ionization (NCI).
200
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 1. Individual Biphenyls
• 2-Chlorobiphenyl (Cl)
• '3,3'-Dichlorobiphenyl (2 Cl)
• 2,4,5-Trichlorobiphenyl (3 Cl)
• '2,2',4,4'-Tetrachlorobiphenyl (4 Cl)
• '2,3',4,5',6-Pentachlorobiphenyl (5 Cl)
• '2,2',3,3',6,6'-Hexachlorobiphenyl (6 Cl)
• '2,2',3,4,5,5',6- Heptachlorobiphenyl (7 Cl)
• '2,2',3,3',4,4',5,5'-Octachlorobiphenyl (8 Cl)
• '2,2',3,3',4,4',5,5',6-Nonachlorobiphenyl (9 Cl)
• '2,2',3,3',4,4',5,5',6,6'-Decachlorobiphenyl (10 Cl)
Table 2 lists the chromatographic conditions used for both ionization modes. Table 3 lists the Perkin-Elmer Turbo-
Mass mass spectrometer parameters used for each ionization mode. Methane was selected as the reagent gas
for NCI. None is used for El. The electron energy was kept constant for both modes of operation. In NCI the injec-
tion of reagent gas increases the gas pressure in the ion source. The ion source temperature was kept constant
while the electron multiplier was run at a slightly higher setting in NCI mode. The selected masses using NCI
reflect the most abundant mass and are usually from the molecular ion cluster, while those for El were the most
abundant and do not necessarily reflect the molecular ion owing to greater fragmentation.
Table 2. Chromatography Conditions
GC:
Column:
Oven Temperature Program:
Capillary Splitless Injection:
Programmable Pneumatic Control (PPC):
Split Vent:
Injection Volume:
Perkin-Elmer AutoSystem XL
PE-5MS 30 m x 0.25 mm; 0.25 mm film thickness
55°C for 5 min.,
45°C/min. to 160°C;
6°C/min to 320°C
250°C
Helium @ 1.0 mL/min.
Splitless -1.00 to 1.00 min; Split 50 ml_ @ 1.00 min.
1.0 |jL
Table 3. El and NCI Parameters
TurboMass Parameters
Reagent Gas:
Electron Energy:
Pressure:
Ion Source Temperature:
Electron Multiplier:
Full Scan:
Analytes:
Selected Ion Scans:
Electron Ionization
None
70 eV
6.9 x 10-6Torr
150°C
524V
m/z 160tom/z600
10 individual biphenyls
m/z 188, 222, 256, 292, 326, 360,
394, 430, 464, & 498
Negative Chemical Ionization
Methane
70 eV
ISxIO^Torr
150°C
651 V
m/z 1 60 to m/z 600
10 individual biphenyls
m/z 187, 221, 255, 292, 326, 360,
394, 430, 464, & 498
A mixed concentration standard containing the previously named 10 biphenyls was analyzed using electron ioniza-
tion and negative chemical ionization. Figure 2 shows an overlay offset of both total ion current chromatograms.
201
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Specific analyte concentrations are annotated over the individual peaks. The El chromatogram shows an inverse
response to increasing levels of halogenation, with chlorobiphenyl the most responsive and decachlorobiphenyl
the least. The overall response is much less than that of the same compounds using NCI. In the NCI mode of
operation we see a general increase in response to increasing levels of halogenation with the exception of penta-
chlorobiphenyl and hexachlorobiphenyl.
PCB Mix (100,1C, SHE)
Electron
lonization
TIC
Negative
Chemical
lonization
TIC
Figure 2. TICs of a
Biphenyl Standard
Mix using El & NCI
Decachlorobiphenyl shows a dramatic increase in NCI response whereas chlorobiphenyl exhibits a reduced signal.
It is believed that the inverse response of penta- and hexachlorobiphenyl is related to the isomeric structure. The
2,4,6 conjugation of the chlorines in the pentachlorobiphenyl may enable the ring to stabilize the negative charge
better, reducing fragmentation and increasing signal. Figure 3 shows chlorobiphenyl and decachlorobiphenyl
comparing the integrated areas obtained using El and NCI. El shows greater sensitivity for the less halogenated
compound, although the NCI signal is quantifiable. NCI shows even greater sensitivity gains with increasing
halogenation while the decachlorobiphenyl peak using El shows a barely discernible signal.
The optimal technique will depend on the particular biphenyls targeted and the sensitivity levels required. NCI
offers greater sensitivity for a larger number of the biphenyls examined here than El offers. The ability to change
between modes of operation, i.e., NCI and El is crucial for optimum versatility and instrument utilization. The
combination of these two techniques can provide more information relative to sample content and realization of
dramatic increases in sensitivity. Ion sources can be easily interchanged without disrupting the chromatography
conditions or column retention times. The chromatogram shown previously in Figure 2 exemplifies this. The
TurboMass ion source can be changed in about 1 minute (exclusive of cool down and heat up times) allowing for
confirmation using El and the NIST library along with the sensitivity gains realized using NCI. The two modes of
operation are complimentary.
If greater sensitivity is needed for less halogenated biphenyls using NCI, selected ion monitoring is an attractive
alternative. The TurboMass mass spectrometer can perform Selected Ion Recording (SIR also called SIM)
scanning while also acquiring in the Full Scan mode. A mass spectrometer data acquisition method was created
combining both Full scan and SIR modes. The selected masses and retention time windows are listed in Figure 4.
Full ion scanning is performed throughout the entire chromatographic run, while the individual biphenyl signature
masses are scanned at the appropriate chromatographic elution times.
The SIR peak can then be used for quantification while the Full Scan provides spectra for accurate identification.
In this manner biphenyls with varying responses can be combined in one analysis. Quantification can be
performed using Full Scan or SIR on a compound-by-compound basis. Simultaneous Selected Ion & Full Ion
(SIFI) Scanning combines all the benefits from both modes of operation into one chromatographic run.
202
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
0222_15Sm(Mn, 3x1)
100-
Area
PCB Mix (20, 2, Ing)
Decachlorobiphenyl
NCI Mode
1:ScanCI-
498
1.15e6
0219_7Sm(Mn, 3x1)
100-
Area
19
Decachlorobiphenyl
El Mode
1: Scan EI+
498
398
0222JI5 Sm (Mn, 3x1)
100-
Area
1: Scan Cl-
187
9.32e4
Chbrobiphenyl
NCI Mode
0219_7Sm(Mn, 3x1)
100-
Area
101875
1:ScanEI+
188
1.46e6
Chtorobiphenyl
El Mode
6.00
8.00
10.00
12.00
14.00
16.00
18.00
20.00
22.00
24.00
26.00
28.00
30.00
•Time
Full Scan 160,8fe6K*.f
Figure 3. Monochlorinated and Decachlorinated biphenyls using El & NCI.
SUMMARY
Electron capture negative chemical
ionization provides a high level of
discrimination. It can selectively detect
halogenated compounds, such as
polychlorinated biphenyls. Selectivity
also helps to greatly minimize interfer-
ences. Generally, for the congeners
examined in this paper, NCI shows a
much higher response for the more
highly chlorinated compounds, al-
though this will depend on the degree
of isomerization. The dramatic in-
creases in sensitivity realized for the
more highly chlorinated biphenyls
using NCI ordinarily reflects the in-
creasing level of PCB chlorination.
Figure 4. Simultaneous Full Ion and
Selected Ion Scanning using Negative
Chemical Ionization
The less highly chlorinated biphenyls, and particularly tri- and tetra-chlorobiphenyl, are not as responsive to EC
NCI. El may provide a better solution for these congeners. An even more accurate compound profile can be
obtained by the efficient use of both forms of ionization. Fast and simple switching between El and Cl modes of
operation affords the most efficient technique for optimum analyte characterization. However, Cl is not without it's
drawbacks. More frequent source maintenance is required for Cl and incurs the added cost of reagent gas. Since
chemical ionization is new to the environmental laboratory, it will require additional method development and
increased operator skill level. With a sensitive detector and an efficient technique for alternating sources, NCI can
be a powerful tool for use in the environmental laboratory, providing the ability to use both electron and chemical
ionization for the determination and characterization of PCB contaminated material.
2 SIR of 2 masses (Cl-j
3 SIR o(255 00. (CI-)
4 SIR of 2 masses (CI-)
5 SIR of 3 masses (CI-)
6 SIR of 2 masses (CI-)
203
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
PCB Mix (100, 10, 5ng)
0217 34
1001 Decachlorott
%\ SIR
0217_34
0217 34
,„ 14.46
0217 34
100,
%] 12.25
o' /
0217 34
100 1147
0217_34
1°°] 17.93
*1« .15 14,48 I
As | \
6.00 8.00 1000 12.00 1400 16.00 1800
6: SIR of 2 Channels Cl-
22 03 498.00
phenyl 1-34e8
5 SIR of 3 Channels Cl-
21 01 430.00
| 1.3408
IT
4: SIR of 2 Channels Cl-
32600
1.34e8
3: SIR of 1 Channel Cl-
255.00
3.44 e6
2 SIR of 2 Channels Cl-
221.00
1.40e7
1 ScanCI-
21 59 TIC
U5.08e7
20 00 22 00 24 00 26.00 28 00 30 00
Figure 5. Simultaneous
Selected Ion and Full Ion
Scans for a biphenyl
standard mixture.
The selected ion currents for five of the biphenyls are show in Figure 5 along with the total ion chromatogram
acquired at the same time. Figure 6 emphasizes the sensitivity gains realized using SIFI and NCI. The less
responsive chlorobiphenyl at a concentration of 10 ng/uL shows over a factor of 40 increase in area using the
selected ion monitoring mode as compared to the full ion scannning mode. In one analysis, selected ion monitor-
ing can be implemented for chlorobiphenyl and other poorly responding analytes, while full ion monitoring is more
than sufficient for the more responsive biphenyls, such as Decachlorobiphenyl.
Figure 6.
Chlorinated biphenyl
peak areas using
simultaneous SIFI in
NCI mode.
0222.
100 -,
Am'
*-
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
THE USE OF SULFURIC ACID CLEANUP TECHNIQUES TO MINIMIZE MATRIX INTERFERENCES
FOR THE ANALYSIS FOR TOXAPHENE IN SOIL, SEDIMENT, AND GROUNDWATER
Francis J. Carlin Jr.
Analytical Science Division Hercules Incorporated, Research Center,
500 Hercules Road, Wilmington, DE 19808-1599
Rock J. Vitale, CPC
Environmental Standards, Inc., 1140 Valley Forge Road, Valley Forge, PA 19482
ABSTRACT
Toxaphene is a chlorinated camphene insecticide, containing more than 170 components, which yields a complex,
multi-component gas chromatogram. Accurate identification and measurement of toxaphene in environmental
samples depend on the removal of matrix interferences from the sample extracts before analyses by gas chroma-
tography using an electron capture detector (GC-ECD). The successful analysis of complex soil and sediment
samples, which contain natural and synthetic substances, by gas chromatographic techniques, has long been a
challenge for analytical chemists. The use of selective detectors and cleanup techniques has assisted in this
endeavor. With respect to the analyses for extractable chlorinated pesticides and polychlorinated biphenyls (PCB)
in environmental samples, the U.S. EPA has published a variety of cleanup techniques. The intent of these proce-
dures is to minimize co-extracted interfering substances, while maintaining the qualitative and quantitative integrity
of the target analyte(s) of interest. On two projects involving the collection of varying soil, sediment, and groundwa-
ter samples, multi-laboratory method validation studies were performed to assess the viability of using sulfuric acid
cleanup techniques for the analyses for toxaphene. Extraction techniques, analytical conditions, method validation
data, and on-going laboratory control recovery data indicate that sulfuric acid is an effective cleanup technique
which significantly reduces co-extracted matrix interferences and that sulfuric acid does not affect the qualitative
and quantitative integrity of toxaphene. Based on these studies, the authors conclude that SW-846 Method 3665A
should include toxaphene as a validated analyte.
INTRODUCTION
Toxaphene had been a widely used pesticide in the United States until most uses were banned in 19821 Over the
past several years, Hercules has collected environmental samples as part of monitoring and remediation activities
and has submitted those samples to contract laboratories for the determination of toxaphene. Because toxaphene
is a chlorinated camphene insecticide that contains more than 170 components, a multi-component gas chroma-
togram is obtained2 Therefore, in order to identify and measure toxaphene accurately, it is particularly important to
eliminate from the sample matrix interferences such as single-response organochlorine pesticides and other
electron capture-sensitive compounds3'4'5
With respect to the analyses for extractable chlorinated pesticides and PCBs in environmental samples, the U.S.
EPA has published6 a variety of cleanup techniques. The intent of these cleanup procedures is to minimize
co-extracted interfering substances, while maintaining the qualitative and quantitative integrity of the target
analyte(s) of interest. In particular, the U.S. EPA has published SW-846 Method 3665A, which is a sulfuric acid
cleanup procedure. This procedure is identified in the method as being specific to the analyses for PCBs, since
these compounds have been shown to be relatively unaffected by this cleanup technique. Conversely, as noted in
the method, this cleanup technique "cannot be used to cleanup extracts for other target analytes, as it will destroy
most organic chemicals including the pesticides Aldrin, Dieldrin, Endrin, Endosulfan (I and II), and Endosulfan
sulfate." As a result, sulfuric acid has been exclusively relegated for use as a cleanup technique for the analysis
for PCBs during many environmental investigations.
Two projects involved the collection of a significant number of soil, sediment, and groundwater samples collected
at sites in the Southeast and on the West Coast. Due to the nature of these sites, significant chromatographic
interferences were suspected in the form of co-extracted non-target analytes. Because of this concern, both of the
laboratories retained to perform the analyses were first required to perform independent formal method validation
studies, inclusive of method detection limit (MDL) studies and precision and accuracy studies, to assess the viabil-
ity of using sulfuric acid cleanup techniques for the analysis for toxaphene.
EXPERIMENTAL METHODS
Soil, sediment, and groundwater samples were extracted using SW-846 Methods 3550A and 351OC. The resulting
extracts were exchanged to hexane and concentrated to a final volume of 10 mL. The final hexane solutions were
cleaned up by shaking with concentrated sulfuric acid (SW-846 Method 3665A). The hexane layer was then
analyzed by GC-ECD, following SW-846 Method 8081A with additional, project-specific data quality objectives
205
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
(DQOs) and analytical requirements. All samples were injected into two different GC columns for quantitation and
confirmation: DB-1701 and DB-5, respectively.
RESULTS AND DISCUSSION
The QA/QC samples from the investigations at two different commercial laboratories working on samples from two
different sites were examined for evidence of changes in the GC profile of toxaphene. Comparisons of the
chromatograms of the toxaphene calibration standard; the soil, sediment, or groundwater samples; the matrix
spikes (MS) and matrix spike duplicates (MSD); and the laboratory control samples (LCS) were made to deter-
mine if changes in the GC patterns were generated as a result of the H2SO4 cleanup. A secondary comparison
was made to ascertain if the use of H2SO4 had any effect on the recoveries of the two surrogate compounds:
tetrachloro-meta-xylene (TCMX) and decachlorobiphenyl (DCB).
Comparisons of the chromatograms in Figure 1 and Figure 2 demonstrate that there were no discernable changes
in the toxaphene after exposure to the H2SO4 cleanup step. In Figure 1, the chromatograms of the soil samples
(980672-006) show no changes in the GC peak pattern for the MS and MSD samples. The few peaks that are
different from the toxaphene standard originate in the soil extract, and they are not removed by the H2S04 treat-
ment. They are not related to changes in the toxaphene during the acid cleanup step. In Figure 2, the chroma-
togram of the water sample (980674-001) after H2SO4 cleanup shows only traces of peaks. In the MS and MSD
samples, the GC patterns of the toxaphene after cleanup clearly demonstrate a peak-for-peak match for the
components in the toxaphene standard. The toxaphene standard is a calibration solution, which had no contact
with H2SO4.
The quantitative recoveries for toxaphene in the MS and MSD samples, the LCS's, and the surrogate compounds
TCMX and DCB from a number of representative samples collected from Site 1 are presented in Table 1. Excel-
lent recoveries were achieved for toxaphene in the MS and MSD samples and the LCS's, ranging from 88 % to
110%. Similarly, acceptable recoveries were observed for TCMX and DCB, which ranged for 73% to 123%. The
quantitative recoveries for toxaphene in the MS and MSD samples, the LCS's, and the surrogate compounds
TCMX and DCB from a number of representative samples collected from Site 2 are presented in Table 2. As was
the case for Site 1, excellent recoveries were also obtained for Site 2 for toxaphene in the MS and MSD samples
and the LCS's, ranging from 78% to 110%. Similarly, acceptable recoveries were observed for TCMX and DCB,
which ranged from 67% to 97%.
SUMMARY
The work presented within this study is a summary of research performed on a significant number of soil,
sediment, and groundwater samples collected from two different sites and analyzed at two independent commer-
cial laboratories. This research has demonstrated that sulfuric acid is an excellent cleanup option for the analysis
for toxaphene. There are no discernable changes in the toxaphene GC-ECD profile after sulfuric acid cleanup.
The recovery of toxaphene, TCMX, and DCB indicate that there is no destruction of those compounds as a result
of contact with concentrated sulfuric acid. The application of the research presented herein, complete with
thoughtful project planning, project-specific DQOs, and analytical requirement specifications, represents the
essence and appropriate application of the U.S. EPA performance-based measurement systems (PBMS). It is
recommended that, during the next update period, toxaphene be added to the list of analytes that have been
validated for use in SW-846 Method 3665.
REFERENCES
1. United States Public Health Service, Agency for Toxic Substances and Disease Registry, "Toxicological Profile
for Toxaphene (Update)," August, 1996.
2. M.A. Saleh, "Toxaphene Chemistry, Biochemistry, Toxicity, and Environmental Fate," Rev. Environ. Contam.
Toxicol. 118: 1-85(1991).
3. J.W. Gooch and F. Matsumura, "Evaluation of Toxic Components of Toxaphene in Lake Michigan Trout," J.
Agr. Food Chem. 33: 844-848 (1985).
4. F Matsumura, R.W. Howard, and J.O. Nelson, "Structure of the Toxic Fraction of Toxaphene " Chemosphere
5:271-276(1975).
5. J.O. Nelson and F Matsumura, "Separation and Comparative Toxicity of Toxaphene Components," J. Agr.
Food Chem. 23: 984-990 (1975).
6. United States Environmental Protection Agency, "Test Methods for Evaluation of Solid Waste, Volume 1B:
Laboratory Manual, Physical/Chemical Methods, (SW-846)," Method 8080A, Organochlorine Pesticides and
Polychlorinated Biphenyls by Gas Chromatography, Revision 1, November 1992.
206
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
600-
200-
Soll 980672-006
Sulfuric Acid Cleanup
I TCMX
DCB
400-
200-
0-
11I II I|11II|I I I I11 I II|IIII|I I I I|I I II|11 I I| II I I | I III | II 11 11 [ 1 I | I I I I | I I M | M M | I I I I | I I I I | I I I I | I I I I | I I I I | I I I 11 I I I I | I I I I | I I I I | 11 I I | I
7 8 9 10 11 12 13 14 15 16 17 18 19
980672-006 MS
Sulfuric Acid Cleanup
I TCMX
JLJ „
DCB
600—I
400-
200—
400-
300-
200—
100-
10 11 12 13 14 15 16
8 & 10 11 12 13 14 15 16 17 18 19
Figure 1. Comparison of Chromatograms of a Toxaphene Standard with Extracts of a Soil Sample and Matrix
Spike Samples
207
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
400-
300-
2GO-E
100—
Rinse Water 980674-001
Sullurio Acid Cleanup
TCMX
600—
400—
200-
600—
400—
200—
TCMX
400—
300—
100—
T ] i rnym 111111 [ i M 11 TTTT| 11 M | n n 111111-11
7 3 9 10 11
Toxaphene Standard
No Sullunc Acid Cleanup
TCMX
12 13 14 15 15
DCB
9 10 11 12 13 14 15 16 17 18 19
980674-001 MSD
Sulfurlc Acid Cleanup
DCB
8 9 10 11 12 13 14 15 16 17 18 19
DCB
DCB
I | I II I [III l|lll l| llll| Illl |llll|llll|ll II |l III |MM |llll[l III] III l| III! | Illljl I ll|lll I |l III | I IM| II ll| II |
7 8 9 10 11 12 13 14 15 16 17 10 19
Figure 2. Comparison of Chromatograms of a Toxaphene Standard with Extracts of a Water Sample and Matrix
Spike Samples
208
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 1. Summary of Recovery Data from Laboratory 1 and Site 1
SAMPLE NUMBER
9020127-BLK1
90201 27-BS1
90201 27-MS1
90201 27-MSD1
9020258-BLK1
9020258-BS1
9020258-MS1
9020258-MSD1
SAMPLE TYPE
Blank
LCS
MS
MSD
Blank
LCS
MS
MSD
MATRIX
Soil
Soil
Soil
Soil
Soil
ANALYTE
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
% RECOVERY
ND
77.8
72.7
90.9
88.9
83.8
88.0
90.7
80.8
91.0
92.8
83.8
ND
77.2
79.3
95.4
79.0
91.0
122
123
114
106
111
101
ND = Not Detected
209
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 2. Summary of Recovery Data from Laboratory 2 and Site 2
SAMPLE NUMBER
980675-005
980788-007
980674-001
SAMPLE TYPf/J
MS
MSD
LCS 1
LCS2
MS
MSD
LCS 17
LCS 18
MS
MSD
LCS 98
piUATRIX
Sediment
Sediment
Sediment
Sediment
Water
Water
ANALYTE
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
Toxaphene
TCMX
DCB
% RECOVERY
88
78
85
92
79
86
78
67
80
86
68
84
92
87
97
94
88
92
91
85
94
95
89
92
110
82
78 i
107
93
78
100
80
91
ND = Not Detected
210
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
THE ANALYSIS OF ARMY CHEMICAL AGENTS:
GB, VX, MUSTARD, AND LEWISITE IN SOIL AT ROCKY MOUNTAIN ARSENAL
Dwight Parks
Chemist
Lockheed Martin Systems Support and Training Services, Rocky Mountain Arsenal Building 130,
72nd and Quebec, Commerce City, CO 80022-1748
ABSTRACT
A critical need in the ongoing remediation of Rocky Mountain Arsenal (RMA) is the ability to quickly determine
whether various solid materials are free of military chemical agent contamination. The analysis must be performed
and results available within two hours of receipt of the sample at the onsite laboratory. The reported results must
exhibit a high degree of confidence and accuracy. Suitable methods for bis(2-Chloroethyl) sulfide, (Mustard,H),
Isopropylmethylphosphonofluridate, (Sarin.GB), O-ethylS(2-diisopropylaminoethyl)methylphosphonothiolate(VX),
and 2-chlorovinyl dichloroarsine (Lewisite.L) have been validated at the Environmental Analytical Laboratory (EAL)
at RMA under the Comprehensive Analytical Laboratory Services (CALS) contract (CALS contractor URS Greiner
Woodward Clyde). These techniques are used in support of the remediation at RMA.
GB, VX and H are first extracted from solid matrices in a fluid containing chloroform and 2-(diisopropyl amino)
ethanol. The extract is decanted, centrifuged, and analyzed on two separate gas chromatographs (GC) equipped
with dual flame photometric detectors (FPD). The first GC is configured in the phosphorous mode with a 525-nm
filter for the analysis of GB and VX. The second GC is configured in the sulfur mode with a 393-nm filter for the
analysis of Mustard (H). Simultaneous second column confirmation is used on both gas chromatographs to
provide confirmatory analysis.
Lewisite is first extracted in a fluid containing 0.01% 1,3-propanedithiol in hexane. The 1,3-propanedithiol derivat-
izes Lewisite and its breakdown products into a similar derivative (LD) which is chromatographically stable. The
derivatization of Lewisite and its breakdown products allow for a quick qualitative and quantitative analysis by a
gas chromatograph for the presence or former presence of Lewisite in solid matrices. The extract is analyzed on a
GC equipped with dual flame photometric detectors configured in the sulfur mode, dual columns and injectors for
simultaneous qualitative and quantitative analysis.
These methodologies have been subjected to a performance-based validation process which for GB, VX, H and L
resulted in the following Method Reporting Limits (MRL) and accuracies. Method validations for these four Army
chemical agents were performed using non-agent RMA standard soil:
• The MRL for Mustard was determined to be 0.250 ug/g, with a method accuracy of 113%.
• The MRL for GB was determined to be 0.320 ug/g, with a method accuracy of 85.1 %.
• The MRL for VX was determined to be 0.353 ug/g, with a method accuracy of 74.5%.
• The MRL for Lewisite was determined to be 0.275 ug/g, with a method accuracy of 80.3%.
The following poster material will explain in detail how these methods are performed. These methodologies have
been critical in furthering the remediation at RMA by providing a means to establish that various solid materials are
free of military chemical agent contamination.
INTRODUCTION
Chemical agent production at RMA occurred under a variety of different programs for numerous years. Mustard
was manufactured at the facility from December 1942 until May of 1943. Mustard was also found at RMA at
various other times for many different projects including the filling of munitions and demilitarization. The production
of Lewisite began in April of 1943 and ended in November of 1943. The production of nerve agent GB (Sarin),
occurred at the Arsenal from 1953-1958. Demilitarization occurred in 1972,1973, and 1976. Finally VX was stored
at RMA in what is now known as the Toxic Storage Yard. These activities lead to the need to develop analytical
techniques that quickly determine that various solid matrices are free of these four chemical agents. The analyses
must be performed and results available within two hours of receipt at the onsite laboratory. These analyses can
be performed within the necessary framework using two different methods on three different GCs.
THE ANALYSIS OF GB, VX, AND H IN SOIL
Reagents:
1. Chloroform, CHCI3, Residue grade (assay 99.95%) or better
211
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
2. 2-(diisopropylamino)ethanol. [(CH3)2CH]2-N-CH2-CH2-OH
3. Extraction Mixture, (Prepare by placing 500 ml of chloroform in a 1L volumetric flask. Add 25 ml of
2-(diisopropylamino)ethanol and 5.0 mL of distilled water. Shake until all components are dissolved. Dilute to
final volume with chloroform. The solution is stable indefinitely and is stored at room temperature in it's
volumetric flask.)
4. Hypochloric Acid (bleach)
5. Non-agent RMA standard soil
Analytes: Chemical Agent Standard Analytical Reference Materials (CASARM) are received as dilute solutions
sealed in 5-mL glass ampules under a blanket of dry inert gas from the Chemical Research and Development and
Engineering Center (CRDEC) in Aberdeen, Maryland.
1. Mustard: bis(2-Chloroethyl) sulfide, CI-CH2-CH2-S-CH2.CI (CAS No 505-60-2). A dilute CASARM stock solution
in hexane is obtained from CRDEC at approximately 2000ug/mL.
2. GB (Sarin): Isopropylmethylphosphonfluoridate, CH3P(O)(F)-OCH(CH3)2,(CAS No. 107-44-8). A dilute
CASARM stock solution in isopropyl alcohol is obtained from CRDEC at approximately 250 ug/mL.
3. VX: O-ethyl S-(2-diisopropylaminoethyl) methyphosophonothiolate (CAS No. 50782-69-9), CH3-P(O)(OC2H5)-
S-CH2-CH2-N[CH(CH3)2]2. A dilute CASARM stock solution in isopropyl alcohol is obtained from CRDEC at
approximately 100ug/mL
Sample Preparation: GB, VX, and H are extracted together in one vial.
1. Weigh 5.0 grams of soil into a 40 milliliter (ml) volatile (VGA) vial with a TeflonO lined screw cap.
2. Add 5.0 ml of 2-(diisopropyl amino)ethanol/chloroform extraction fluid.
3. Place the sample on a vortex mixer for 15 seconds
4. Allow it to stand for approximately one minute.
5. Transfer the liquid extract into autosampler vials. If necessary, to settle out particulate matter decant the
extract into a centrifuge tube and centrifuge at 1800 revolutions per minute (RPM) for 1 minute.
6. Analyze on both the sulfur and phosphorus mode GC's.
Instrumentation: Sulfur-mode GC (Mustard)
A Hewlett Packard 5890 GC equipped with dual FPDs and a 7673 autosampler. The FPDs are operated in the
sulfur mode and are each equipped with a 393-nm filter (purple). The primary column is a Restek RTx-5, 0.53 mm
I.D. 30 meters, 0.50 urn film thickness and the secondary column is a Restek RTx-200, 0.53 mm I.D. 30 meters, 1
urn film thickness. The primary and secondary columns are joined by a "Y" adaptor. The instrument parameters
are: initial oven temperature of 120°C for 2.0 minutes, then temperature ramped at a rate of 35°C/minute to a final
temperature of 300°C and held for 1.0 minute. The injectors and detectors are held constant at 250°C.
Instrumentation: Phosphorus-mode GC (GB and VX)
A Hewlett Packard 5890 GC equipped with dual FPD and a 7673 autosampler. The FPDs are operated in the
phosphorus mode and are each equipped with a 525-nm phosphorus filter(yellow). The primary column is a
Restek RTx-5, 0.53 mm I.D., 30 meters, 0.50 urn film thickness and the secondary column is a Restek RTx-200,
0.53mm I.D., 30 meters, 1 urn film thickness. The primary and secondary columns are joined by a "Y" adaptor.
The instrument parameters are: initial oven temperature of 50°C for 3.0 minutes, then temperature ramped at a
rate of 50°C/minute to a final temperature of 300°C and held for 3.50 minutes. The injectors and detectors are
held constant at 250°C.
THE ANALYSIS OF LEWISITE IN SOIL
Reagents:
1. Hexane, HPLC grade or equivalent
2. 0.01% 1,3-Propanedithiol extraction fluid (Using a gastight syringe, add 100uL of 1,3-Propanedithiol (1,3-PDT,
Aldrich, 99%, Cat. No. P5, 060-9) to a final volume of 1.0 L HPLC grade hexane. Invert to mix. This solution
has an expiration date of six months after date of preparation. It should be stored in a hood since 1,3-PDT is a
stench irritant.)
3. Hypochlorite Acid
4. Non-agent RMA standard soil
Analyte: The CASARM is received as a dilute solution sealed in 5-mL glass ampule under a blanket of dry inert
gas from CRDEC. A new stock is received annually.
1. Lewisite: 2-chlorovinyldichloroarsine, CI-CH=CH-AsCI2, (CAS No. 541-258-3). A dilute CASARM stock in
hexane is obtained from CRDEC at approximately 2000ug/mL.
212
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Sample Preparation: Sample preparation for the analysis of Lewisite is slightly more complex. Lewisite rapidly
hydrolyzes to form a variety of products and hydrochloric acid and does not readily lend itself for analysis byGC.
The analysis of Lewisite includes a step to derivatize the Lewisite in a stable product (LD) which can be easily
analyzed by GC.
1. Weigh 10.0 grams of soil into a 40 mL VOA vial.
2. Pipet 5.00 mL of the 0.01% 1,3-propanedithiol in hexane into the VOA vial.
3. Gently swirl to mix the soil into the solution.
4. Place the sample on a vortex mixer for 10 seconds to throughly mix the solvent and soil.
5. Place in an ultrasonic bath and sonicated for 6 minutes.
6. Let the sample sit undisturbed for 30 minutes. During this time the derivatization of Lewisite and Lewisite
Oxide into the stable derivative LD occurs. This reaction has been observed by various authors investigating
the derivative and derivative detection.1 The derivative, LD, is chromatographically stable, and enables the
analysis to proceed by normal chromatographic techniques.
7. Transfer the liquid extract into autosampler vials.
8. Analyze on the sulfur-mode GC.
Instrumentation: Sulfur Mode GC (Lewisite)
A Hewlett Packard 5890 GC equipped with dual FPDs and a 7673 autosampler. The FPDs are operated in the
sulfur mode with 393-nm filter(purple). The primary column is a Restek RTx-1701, 30 meters, 0.53 mm ID, and a
0.50um film thickness and the secondary column is a Restek RTx -5, 30 meters, O.SOum film thickness. The
instrument parameters are: initial oven temperature of 120°C for 3.0 minutes, then temperature ramped at
35°C/minute to a final temperature of 270°C and held for 2.71 minutes. The injectors and detectors are held
constant at 250°C.
GB/VX/H RESULTS
The FPDs display good selectivity and sensitivity to the analytes of interest in comparison to the wide range of
background contamination present at the RMA. (See chromatograms #1 and #2 pages 7 and 8). No problems
have been experienced with false positives due to non-agent contamination. During the performance-based
method proficiency process the following results were obtained. The method proficiency matrix was non-agent
RMA standard soil:
Tested Concentration Range: The tested concentration ranges for GB, VX, and H on both columns are 0.25 to 2.5
M9/9-
Sensitivity:
Column Analyte MRL Accuracy
Restek RTx-5 GB 0.130ug/g 0.957
VX 0.320ug/g 0.847
H 0.220ug/g 1.1
Restek RTx-200 GB 0.320ug/g 0.849
VX 0.350ug/g 0.745
H 0.200ug/g 0.920
Method Reporting Limits (MRL): The MRLs will be the worst case scenario obtained from the performance-based
method proficiency listed above in order to include the performance of each column. Mustard is assigned the MRL
of the lowest proficiency spike that was analyzed, since the calculated MRL for Mustard was less than the lowest
spike analyzed (see table below).
Analvte MRL Upper Limit Accuracy
GB 0.320ug/g 2.50ug/g 0.851
VX 0.353ug/g 2.50ug/g 0.745
H o.250ug/g 2.50ug/g 1.13
LEWISITE
The FPDs displayed good sensitivity and selectivity for Lewisite in comparison to the wide range of background
contamination and interferences (Chromatogram #3 page 8). No problems have been experienced with false
positives due to non-agent contamination. During the performance-based method proficiency process the
213
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
following results were obtained. The method proficiency matrix was non-agent RMA standard soil.
Tested Concentration Range: The tested concentration for Lewisite on both columns was 0.125 to 1.25 ug/g.
Sensitivity:
Column Analyte MRL Accuracy
RestekRTx-1701 LewisiteO. 121 ug/g 0.803
Restek RTx-5 LewisiteO.275ug/g 0.924
Method Reporting Limit (MRL): The MRL will be the worst case scenario obtained from the performance-based
method proficiency listed above in order to include the performance of each column.
Analyte MRL URL Accuracy
Lewisite 0.275ug/g 1.25ug/g 0.803
SUMMARY
These methods provide a quick qualitative and quantitative means to determine if various solid matrices are free
of these four chemical agents. Some of the matrices that have been tested using these procedures are soil,
concrete, chlorinated parafilm, fume hood filter remnants, and various other waste materials. In each case the
methodologies have been able to determine that these matrices are free of GB, VX, H and L. The use of the
FPDs with simultaneous confirmatory analysis has filled a critical need in providing high quality, defensible data
that has enabled to cleanup of RMA to move forward.
FOOTNOTES
1. Albro, Thomas., Field Determination of Lewisite and It's Breakdown Products by Flame Photometric and
Photoionization (PID) Detectors in Soil and Water Matrices: September 1994, Page 2
ACKNOWLEDGMENTS:
The Comprehensive Analytical Laboratory Services Team at the Rocky Mountain Arsenal comprised of the RVO
Support Team. URS Greiner Woodward Clyde, Lockheed Martin, Oak Ridge National Laboratory, and Roybal
Corporation.
REFERENCES
1. Albro, Thomas., Field Determination of Lewisite and It's Breakdown Product by Flame Photometric and
Photoionization (PID) Detectors in Soil and Water Matrices: September 1994.
2. Chemical Agent Standard Analytical Reference Material Quality Assurance Plan for Chemical Agent Air
Monitoring, Revision II, US Army CBDCOM Quality Directorate, August 1995.
3. Comprehensive Analytical Laboratory Services, Laboratory Quality Control Plan, April 1999.
4. Comprehensive Analytical Laboratory Services, Quality Assurance Management Plan, April 1999.
5. Environmental Analytical Laboratory, Rocky Mountain Arsenal, SOP 301, Analytical Method for the Determi-
nation of Mustard, Sarin and VX in Soil using 2-(Diisopropyl amino) ethanol Extraction Followed by Gas
Chromatography with Flame Photometric Detection Method Number LOOS, October 1998.
6. Environmental Analytical Laboratory, Rocky Mountain Arsenal, SOP 310, Lewisite in Soil by 1,3-Propanedithiol
Derivatization Certified Method LL13, December 1998.
7. Kuznear, Casimir and Trautmann William, L. History of Pollution Sources and Hazards at Rocky Mountain
Arsenal. September 1980.
8. Program Manager Rocky Mountain Arsenal, Chemical Quality Assurance Plan, April 1996.
9. Program Manager Rocky Mountain Arsenal, Chemical Hygiene Plan, Revision 4, October 1996
10. Toxic Chemical Agent Safety Standards, US Army, 31 March, 1997.
214
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
t>
0
-1^
0
6
K
b
ft-515
(RH-6)4924
iMB
wan
Chromatogram #2. Mustard @ 1.0 ug/9. sample
B05403 MS, Restek RTx-5 0.53mm ID, 0.50 pm
film thickness, 30.0 meter capillary column.
Chromatogram #1. GB and VX @ 1.0 ug/g,
sample B05403 MS, Restek RTx-5 0.53 mm ID,
0.50|jm film thickness, 30.0 meter capillary
column.
H
b
0
0
a—fl KlISWD (TO-5) J1K
0
n ' • •
^ B U ^ 01 0)
o o b o o o
1 H It 1} it Hi
ty y y ty If jp
;^-lA&
Chromatogram #3. Derivatized Lewisite @1.0
pg/g, sample B05403 MS, Restek RTx-1701
0.53mm ID, 0.50pm film thickness, 30.0 meter
capillary column.
215
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
COMPARISON OF SAMPLING PROTOCOLS FOR THE ZERO
HEADSPACE EXTRACTION (ZHE) FOR TCLP AND SPLP
David Turriff. Nils Melberg, Chris Reitmeyer and Brandon Podhola
En Chem Inc, 1241 Bellevue St., Green Bay, Wl 54304
Telephone: 920-469-2436
Email: turriff@enchemgb.enchem.com
With the introduction of SW846 Method 5035 for soil volatile organic analysis, the large method bias that existed
with SW846 Method 5030 was mostly eliminated. However, protocols for sampling soil and waste for TCLP and
SPLP still rely on the bulk sampling protocol. In regulatory programs that allow for either TCLP/SPLP or Method
5035 to assess contamination risk, there may be a very large discrepancy between the two options. Since soils
cannot be preserved for TCLP/SPLP, there are limited sampling options that will maintain soil integrity.
We compared two sampling protocols, the bulk sampling option (Option A) and the En Core™ sampler (Option B).
Samples were prepared by mixing a soil consisting of 5 Ibs sand, 2 Ibs farm topsoil and 18 Ib of garden soil with a
TCLP standard mixture. The soil was rotated in a mixing drum for 18 hours and then sampled for the TCLP
protocol. Two oz jars with new, high-performance sealing Teflon™ inserts obtained from QEC, Inc. were used for
Option A and the 25 gm En Core™ was used for Option B.
Spiked samples were analyzed at zero time to obtain the initial total concentration. Sets of samples prepared by
Options A or B were leached by TCLP protocol on Day 4, Day 7 and Day 14. Leachates were analyzed by SW846
Method 5030.
Table 1
Day
Day Two
Day Seven
1,1 Dichloroethene
Bulk
7.3
ND
En Core™
12.1
12.7
Benzene
Bulk
28.1
10.1
En Core™
52.4
50.6
Tetrachloroethene
Bulk
37.7
25.5
En Core™
51.4
56.8
Results of representative compounds are presented in Table 1. Data are the mean of five replicates and are
expressed as ug/l of leachate.
As expected, the more volatile compounds showed significantly more time dependent losses than the less volatile
compounds. Data will be presented on the completed time study to fourteen days of soil hold time and for the full
list of TCLP volatile compounds. In addition, two types of bottles with different Teflon™ seals will be compared.
FIELD APPLICATION OF A PORTABLE GAS CHROMATOGRAPH
FOR GROUNDWATER HEADSPACE SAMPLING
Peter J. Ebersold
Technical Specialist
Perkin-Elmer Photovac, 761 Main Avenue, M/S 219, Norwalk, CT 06859
Abstract
For many years portable instrumentation has been widely accepted as a screening tool for a variety of environ-
mental monitoring needs. Technology advances have made accurate identification and quantification of numerous
contaminants possible in the field. End users, however, remain skeptical of field data except as a screening tool.
As a result, most groundwater samples are still collected, preserved in the field and sent to a laboratory for analy-
sis using EPA Method 8260B, Volatile Organic Compounds by Gas Chromatoaraphv/ Mass Spectrometry. While
this is a widely accepted practice, problems associated with sample handling and storage can lead to erroneous
results. Advantages of field sampling and analysis are an immediate answer to questions about the presence and
concentration of VOCs, reduced sampling costs, and the ability to quickly respond to a spill with remediation
techniques appropriate to the area, concentration and content of a spill. This paper will demonstrate the
216
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
applicability of a field portable gas chromatograph for the characterization of volatile organic compounds (VOCs) in
groundwater headspace, thus building confidence in field quantitative data.
In 1997, Perkin-Elmer Photovac participated in EPA's Environmental Technology Verification Program for Field
Portable GCs. In the ETV program field analysis results obtained with the Voyager portable gas chromatograph
were compared to laboratory analysis generated using EPA Method 8260B. As a result of PE Photovac's participa-
tion at two ETV field sites and subsequent data review in the ETV Program, a field sample handling method was
developed to minimize errors, allowing very accurate readings down to ppb levels while maintaining high sample
throughput. A second comparative project was executed in early 1998 to improve accuracy and repeatability.
Excellent correlation between lab and field data was demonstrated in this study.
Introduction
It is a common practice to check for the presence of volatile organic compounds such as BTEX or TCE in ground-
water samples during site assessment or site remediation. The groundwater samples are collected and sent to a
lab for analysis often using Method 8260B, Volatile Organic Compounds by Gas Chromatography/Mass
Spectrometrv. As with any lab method the major disadvantages can be sample handling prior to analysis and
turnaround time. Samples often deteriorate during collection or transport reducing the accuracy of the lab data.
Site managers may be faced with critical project decisions while waiting for lab data which could result in project
delays or unnecessary costs associated with additional sampling. This paper will present a means to sample and
analyze groundwater samples quickly and accurately in the field. Field analysis often results in significant cost or
time savings.
Environmental Field Sampling
At most groundwater assessment or remediation sites that involve groundwater sampling, the samples are
collected in 40 mL VOA vials. The samples are preserved using an appropriate method to prevent sample deterio-
ration, transported as quickly as possible to a lab, stored at the lab, and analyzed using EPA Method 8260B,
Volatile Organic Compounds by Gas Chromatography/ Mass Spectrometrv. There are many potential problems
when collecting field samples of groundwater. The field technician can aerate the sample while filling the VOA vial
causing a loss or dilution of the analytes. Poor sampling technique can allow the analytes to escape into ambient
air. Headspace may remain in the vials causing dilution of the sample with trapped ambient air resulting in low
concentrations of analytes in the vial compared to the source groundwater. If samples are not properly preserved
with a compound such as hydrochloric acid then bacteria in the sample that were inactive due to the lack of
oxygen in groundwater can now "bioremediate" the sample in the vial causing a breakdown or alteration of the
sample compounds. During transport the vials can offgas volatile organic compounds if the vial is not properly
capped or the temperature during transport reaches high levels. Vials can be broken resulting in a total loss of that
sample. Numerous sampling factors can determine the accuracy of the lab results.
Traditional Field Sampling Methods
There are a wide variety of portable analysis kits and instrumentation that have been long accepted as screening
tools in the field. These include portable photoionization detectors and flame ionization detectors for non specific
analysis of total VOCs in the headspace over groundwater, immunoassay field kits, colorometric tests, and detec-
tor tubes. These screening tools often have several significant drawbacks. The PID or FID cannot speciate
compounds so the technician or engineer in the field will learn the total concentrations of VOCs in a sample but
not identify specific compounds or concentrations in the sample. Certain techniques that do speciate are often less
accurate or subject to interference from other compounds in the matrix. Training of the sampling technician as well
as ambient conditions during sampling can also affect accuracy.
Field Sampling Today
Despite these limitations field screening remains an important tool to delineate the extent of contamination at a
site. Valuable time and money can be saved using field screening. Over the past ten to twenty years several new
ways have evolved to accurately assess samples on site with a more rapid turnaround. At larger sites a mobile lab
can be used. A mobile lab consists of lab instrumentation installed in a large trailer or van. The groundwater is
placed into VOA vials, brought to the mobile lab and analyzed quickly. This allows the project manager to make
rapid decisions about the site remediation. There Is also a new generation of portable or transportable instruments
available in the market. Field portable gas chromatographs as well as transportable gas chromatograph/mass
spectrometers can be brought to a site to perform near real time analysis on a wide variety of samples. This type
of instrumentation provides an immediate answer to questions about which compound is present and at what
concentration. Advantages of portable field instrumentation are reduced sampling costs and the ability to respond
quickly to contamination. This quick response can provide substantial savings to a remediation project by reducing
217
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
idle time for field personnel and equipment. Immediate response to new contamination allows more rapid comple-
tion of projects at a lower cost.
Field Sampling Reality
Unfortunately, field data is often seen as suspect or unacceptable by regulators or clients. The quality control
methods for field data is perceived to be less rigorous than the quality control normally practiced by a certified lab.
Lab data is seen as "the gold standard". Yet by implementing many of the same quality control techniques used in
a certified lab it is possible to attain the same accuracy and data confidence using field instrumentation. As the
advantages of performing sample analysis in the field have become more widely recognized, regulators, environ-
mental consultants and PRPs are all becoming more interested in implementing new technology to remediate
sites more rapidly and at lower cost. The move to field sampling using portable instrumentation seems logical. Yet
this has not happened to date. The main reason is a lack of confidence in the data quality.
EPA ETV Program
The US EPA's Environmental Technology Verification (ETV) program was established to facilitate deployment of
innovative environmental technology, provide a verification of performance and disseminate this information to
potential users of environmental technology. ETV is not an approval program but a verification of vendor claims
about their technology. The ETV program was intended to give potential end users of field technology, such as
environmental consultants, a higher level of confidence in the accuracy of field technology.
Wellhead Demonstration Program
In September 1997 the ETV program sponsored a demonstration project to verify the performance of several
instruments designed to analyze volatile organic compounds in groundwater. Two sites were selected to provide
the groundwater samples, Savannah River Site (SRS) in Aiken, SC and McClellan Air Force Base in Sacramento,
CA. Certain compounds were targeted for detection and analysis at each site. At SRS those compounds were
trichloroethylene (TCE) and tetrachloroethylene (PCE). At MAFB the compounds were 1,2 DCA, 1,1,2, TCA, 1,2
Dichloropropane, trans-1,3 Dichloropropene. Other compounds were also expected to be present. The Perkin-
Elmer Photovac Voyager portable gas chromatograph was one of the portable instruments selected to participate
in the demonstration project at both sites. By inviting the Voyager to participate, both the ETV sponsors and Photo-
vac hoped to validate the field performance claims for the Voyager as well as elevate field instrumentation from a
screening technique to a quantitative technique that could be used in place of a fixed lab. This would allow the
Voyager to meet user needs for a more rapid accurate field method for VOC analysis in groundwater. Samples
possibly containing VOCs in groundwater were provided at each site by the ETV program and were analyzed on
the Voyager portable GC. Duplicate samples of the groundwater matrix obtained at each site were sent to a certi-
fied lab for analysis. This would allow direct comparison of data obtained using the Voyager with data obtained by
a lab using Method 8260B.
Technology Description
The Voyager portable gas chromatograph uses a three column design that allows the separation of compounds on
one of three built-in columns. The three built-in columns allow analysis of a wide range of compounds without
having to physically change columns. The Voyager is equipped with a dual detection system. Most volatile organic
compounds are detected using a photoionization detector equipped with a 10.6 eV lamp. Some chlorinated VOCs
such as carbon tetrachloride are detected on an electron capture detector (ECD) with a 15 millicurie Nickel-63
source. Samples are injected or pumped into a heated injection port and then introduced onto the isothermally
heated column set. The user selectable temperature range of the injection port and oven is 30 to 80 degrees
Centigrade. Ultra high purity nitrogen is used as the carrier gas. The Voyager is completely self contained and
weighs 15 pounds. Figure 1 shows the column type and configuration used in the Voyager.
Assay Development
An assay was developed for the Voyager that allowed for detection and accurate quantification of the compounds
expected at the SRS and MAFB sites. The assay included a single column analysis for optimum separation of up
to twenty four compounds. A three point calibration curve for each compound was incorporated into the method.
Original Sample Method
At the Savannah River Site, the following sampling method was used for performing the analysis on the Voyager
GC. A three point calibration curve was established for each compound. A calibration was performed daily before
the first analysis. The ETV personnel provided each groundwater sample to the Voyager field technicians at the
SRS site. The vial was placed in a water bath for 15 minutes at 30 degrees C to allow equilibration. After
218
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
equilibration the vial was uncapped and 20 mL of the sample was poured into another VOA vial. The second vial
with 20 mL of sample and 20 mL of headspace was shaken for two minutes and placed in the water bath at 30
degrees for five minutes. After five minutes the vial was removed from the water bath, a 500 uL gas tight syringe
was used to remove 500 uL of headspace in the vial and the sample was injected onto the Voyager columns.
Voyager results from each completed analysis were automatically stored in the Voyager's internal datalogger. The
contents of the datalogger were later downloaded to a PC for archiving the results and printing the chroma-
tograms. At McClellan Air Force Base, a slightly modified method was used. A three point calibration curve was
run for each compound. A calibration was performed daily before the first analysis. A 40 mL VOA was filled with a
groundwater sample and provided to the Voyager field technicians. The vial was placed in a water bath for 15
minutes at 30 degrees C to allow equilibration. In this modified method, after equilibration 20 mL of sample was
removed using a 20 mL glass syringe. The 20 mL withdrawn from the vial was discarded. The vial, which now
contained 20 mL of sample and 20 mL of headspace, was returned to the water bath at 30 degrees for five
minutes. After five minutes the vial was removed from the water bath, a 500 uL gas tight syringe was used to
remove 500 uL of headspace in the vial and the sample was injected onto the Voyager columns. After the Voyager
completed the analysis results were automatically stored in the Voyager's internal datalogger. The contents of the
datalogger were later downloaded to a PC in order to archive the results and print the chromatograms. The results
of the field analysis at both sites for selected compounds are shown in Figure 2. Overall mean percent recovery is
an average of how closely all reported Voyager concentrations for each sample matched the concentrations found
in the laboratory analysis as a percentage. One hundred percent would indicate perfect Voyager concentration
correlation with laboratory concentrations.
Carrier
Gas (N2)
Column A
'-uiurr.," Mate-Up
Carrier Gas
Column B
Pre-Columri
Column C
Precolumn:
Column A (HEAVY):
ColiimnB (MID RANGE):
Column C (LIGHT):
*ii x 0.53mm x 2.0 urn SPB-35
»n x 0.25rnrn BLANK Fused Silica
20rn x 0.32mm x 1.0 urn SupeJcowaxlO (PEG)
15m x 0.32mm x 12 urn Quadrex 007-1
Figure 1. Voyager
Configuration
Figure 2. Results of SRS/MAFB ETV Demonstration
Target Compound
Trichloroethylene
1,2Dichloroethane
1,2 Dichloropropane
1 ,2 Dichloropropane
1,1,2Trichloro-ethane
Tetrachloro-ethylene
t-1 ,2 Dichloropropene
Site
SRS
MAFB
SRS
MAFB
SRS
MAFB
SRS
MAFB
SRS
MAFB
SRS
MAFB
Overall Mean % Recovery
92-164
231-344
55-86
0-170
55-86
0-170
50-116
1-124
106-162
95-143
219
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
The results did not correlate as well as the Photovac had hoped. Possible reasons for the poor correlation include
loss of analytes due to sample handling technique used by the vendor field personnel. In particular, uncapping the
vial and pouring off part of the sample as was done at SRS most likely resulted in loss of analytes. The stability of
the water bath temperature was also called into question. During sampling the Voyager and water bath were inside
a minivan. Since ambient temperatures were high, the air conditioning was running in the minivan but the doors
were frequently opened and closed. Some of the compounds may have coeluted on the columns leading to false
high readings for some compounds and no detection indications for other compounds. Since the concentration of
compound varied widely in each sample matrix, there could have been errors induced by using only a three point
calibration curve.
Improved Sample Methodology
The outcome of the field sampling at SRS and MAFB prompted development of a new methodology to improve
the accuracy, repeatability, and detection of all compounds. A five point calibration curve was developed for each
compound at 0, 0.005, 0.05, 0.5 and 5 mg/L. More frequent calibrations were performed throughout the day. A
temperature block was substituted for the water bath. The temperature block maintained a more stable tempera-
ture during the VOA vial hold time. A reduced volume of headspace sample was injected into the Voyager to
reduce the possibility of coelutions. Modifying the Voyager assay reduced analysis time and increased sample
throughput.
Redesigned Method testing
This modified sample method was tested by mixing Supelco prepared reference standards of benzene, toluene,
ethylbenzene, m-xylene, trichloroethylene, tetrachlorethylene, bromodichloromethane, and dibromochloromethane
with organic free deionized water. A sample matrix of all eight compounds was prepared at concentrations of 7,
30, 700, 3000 ug/L. A spike matrix of 300 ug/L of benzene, toluene, ethylbenzene, m-xylene, bromodichlormeth-
ane, and dibromochloromethane and 5000 ug/L of trichloroethylene and tetrachlorethylene was also prepared.
Four samples of each concentration were analyzed on the Voyager using the modified sample method and assay.
Duplicate samples were sent to a reference lab for analysis using Method 8260B. The results of the redesigned
test method are shown in Figure 3.
Figure 3. Redesigned Method Results
Compounds
Benzene
Toluene
Ethylbenzene
m-Xylene
Trichloroethylene
Tetrachloro-ethylene
Dibromochloromethane
Bromodichloromethane
% Recovery
at 7 ug/L
82
139
11
11
89
136
121
93
% Recovery
at 30 ug/L
139
115
155
157
145
56
75
206
% Recovery
at 700 ug/L
112
106
98
74
102
115
101
97
% Recovery
at 3000 ug/L
94
92
93
93
89
98
102
102
% Recovery
at 300/5000
ug/L
66
85
60
52
66
95
79
73
Overall Mean
Percent
Recovery
99
107
83
77
98
100
96
114
Conclusions
The Voyager portable gas chromatograph can be used to analyze the headspace over groundwater for the
presence and quantification of volatile organic compounds. Participation in the EPA's ETV program allowed Photo-
vac to compare Voyager data with certified lab data for samples provided by the ETV managers at two field sites.
The initial correlation was not as close as Photovac had expected. A new assay was developed for the Voyager to
target the specific compounds and a new field sampling technique was implemented. Data obtained using the new
method showed improved recovery (Voyager versus lab) and the elimination of coelutions. The new Voyager
assay and sampling methodology increased sample throughput by reducing total analysis time required to run the
samples on the Voyager portable GC. This reduced analysis time provides quicker results to field personnel and
increases the number of samples per day that can be analyzed. Most Importantly, the Voyager now meets the
vendor performance goals established for the ETV program. Meeting the performance goals set for the Voyager
should lead to increased confidence in Voyager field data by regulators and end users. Ultimately, increased field
sampling can reduce the cost of environmental site assessment or remediation and shorten the time needed to
complete a project.
220
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
References
Voyager Portable Digital Gas Chromatograph User's Manual, Revision C, December 1997
Environmental Technology Verification Report, Field Portable Gas Chromatograph, Perkin-Elmer Photovac,
Voyager by Wayne Einfeld, Sandia National Laboratories, Albequerque, NM EPA/600/R-98/144 November
1998
Phoenix Lab Report, Phoenix Laboratories, Manchester, CT
ON-SITE DETERMINATION OF VOLATILE ORGANIC HALIDES
(VOH) IN WATER BY UV-INDUCED COLORIMETRY
Dong Chen. Ph. D.
Research Director
Thomas A. Jackson
Research Chemist
Envirol, Inc., 708 W 1800 North Suite 10 A, Logan, UT 84321
Phone: (435) 753-7946, Fax: (435) 787-2878
David Shattuck
Former President of Envirol, Inc.
145 Buckingham, Providence, UT 84332
Joan McLean
Associate Professor
Utah Water Research Laboratory, Utah State University, Logan, UT 84322
Mark Mines
Former Project Manager of Envirol, Inc.
MSE-HKM, Inc, Sheridan, WY 82801
Abstract
A novel UV-induced colorimetric field test kit, Quick Test® VOH Water Test Method, for the quantitation of volatile
organic halides in water has been developed by Envirol Inc. (Logan, UT). An average method detection limit (MDL)
of 4 g/L (ppb) was achieved for TCE, PCE, chloroform and carbon tetrachloride with a dynamic range up to 200
Og/L. With dilution, the dynamic range can be up to 200 mg/L (ppm). The accuracy and precision results for the
analysis of TCE in water were comparable to standard laboratory methods validated by SW-846. The independent
performances of Quick Test VOH Water Test Method in the field were compared with laboratory results. Statisti-
cal analysis by linear regression and non-parametric t-test (Wilcoxon test) confirms that the Quick Test VOH
Water Test Method meets U.S. EPA Superfund Innovation Technology Evaluation (SITE) Level 2 criteria for field
testing.
Introduction
Trichloroethylene (TCE) is widely used in industry as a degreasing solvent and perchloroethylene (PCE) is used
as a cleaning agent in dry-cleaning facilities. From 1987 to 1993, the TCE and PCE releases into water and land
were estimated to be more than 291,000 Ibs. for TCE and more than one million Ibs. for PCE. The U.S. EPA has
classified TCE and PCE as possible carcinogens and has set the MCL (Maximum Contaminant Level) at five
parts per billion (ppb) for TCE and PCE in drinking water.1'2
Current approaches for evaluating TCE, PCE and other volatile organic halides in water at field sites involve
obtaining and preserving field samples for transport to a laboratory where samples are stored until analysis by gas
chromatography at a cost of approximately $80-$200 per sample. Storage and time constraints for samples taken
in the field often limit the number of samples that can be processed and therefore limit the number of results that
can be obtained. The lag time between sample collection and quantification, and reporting of results can often be
from many days to several weeks. The traditional approach is limited with regard to 1) the number of samples that
can be analyzed due to cost and time, 2) the statistical validation due to the number of samples taken, 3)
decisions concerning site management (removal actions, treatment technologies) are delayed or postponed due to
the relatively long time required from sampling to analyzing results, and 4) evaluation of treatment effectiveness
cannot be determined until results are available.
221
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
A novel UV-induced colorimetric field Quick Test kit, Quick Test VOH Water Test Method, for the quantitation of
TCE, PCE and other volatile organic halides in water has been developed by Envirol, Inc. The process is based on
a photochemical-induced oxidation-reduction reaction between the organic halide and the chemical reagent. The
purpose of this study was to characterize the performance of the Quick Test VOH Water Test Method for the
analysis of VOH in water and to test the suitability of this new method for field investigation of VOH-contammated
sites. The characterization study includes quality-control parameters specified in Test Methods for Evaluating Solid
Waste (SW-846)3, in Lesnik and Marsden4 and our previous research paper,5 including detection limit, dynamic
range, accuracy and precision, interference analysis, and matrix specificity. The Quick Test VOH Water Test
Method was also performed under field conditions, and the results for TCE/PCE in water were compared with
approved U.S. EPA procedures for analysis of TCE/PCE by an independent, certified laboratory.
Experiment
Material and Method
All inorganic chemicals used in this study were reagent grade and the organic chemicals used were optical or
HPLC grade. Quick Test VOH Water Test Method contains all components for water extraction and solution
preparation analysis. The procedure started with a 290 ml water sample being extracted with 2.0 mL of octane
and 30 inches of Teflon® tape. The mixture is shaken manually for three minutes. After this shaking, the Teflon
tape is removed from the solution and placed into a 10-cc syringe. The syringe plunger is used to force the extrac-
tion solvent (octane) and the analyte from the Teflon tape to the extraction solvent vial. The clear extraction
solvent is transferred to a drying vial containing 50 mg sodium sulfate, eliminating residual water, and then trans-
ferred from the drying vial to the liquid/liquid transfer vial containing 1.0 mL acetonitrile. After one minute of
shaking, 0.60 mL of acetonitrile (bottom layer) is pipetted into a vial containing 0.4 mL of the reagent. The mixture
is placed directly into Envirometer™, a field instrument developed by Envirol, Inc.,5 for UV exposure and quantita-
tion. The kit also provides two sets of premeasured standards (5, 90, 190 3>g/L) of VOH (TCE, PCE, CCI4 or
CHCI3) for instrument calibration and two calibration verification samples (90 Og/L).
The spiked concentrations of VOH (TCE, PCE, carbon tetrachloride, chloroform) were verified by GC/ECD
(SHIMADZU, GC-17A) with purge and trap (Tekmer™ 3000) (Method 5030). For performance of the Envirol VOH
test kit the user needs the Envirometer and an adjustable mechanical pipetter capable of measuring 0.60 mL
solution with less than 1 percent absolute error (equivalent to Wheaton No. 851268).
Results and Discussion
Method Detection Limit
The method detection limit (MDL) for the Quick Test VOH Water Test Method was determined with the method
specified in SW-846.3 TCE, PCE, carbon tetrachloride and chloroform were each tested individually. Type II water
(organic free water) was spiked with each of the test chemicals individually at several levels to determine a
primary spiking concentration where the signal/noise ratio was in the range of 2.5-5.0. The primary spiking
concentration was then multiplied by a number from 3-5 to obtain the secondary spiking concentration. In this
study, the multiplier value chosen was four. Once the appropriate secondary spiking concentration was deter-
mined, Type II water was spiked at that concentration and then16 replicates were extracted and analyzed using
the Quick Test VOH Water Test Method. The mean and standard deviation of the TCE, PCE, carbon tetrachloride
and chloroform concentration for the16 samples was determined. The standard deviation was then multiplied by
the appropriate t-statistic to determine the method detection limit of each chemical. To determine the method
quantitation limit (MQL), the same data was used, but with a t-statistic for the 99.9 percent confidence level. The
results of this analysis are summarized in Table 1.
The MDL listed in Table 1 are appropriate for determination of TCE, PCE, carbon tetrachloride and chloroform in
water at regulatory levels above the MDL. The average MDL across all analytes is 4 d>g/L, which is below the MCL
of TCE and PCE (5 ppb) set by the U.S. EPA. The dynamic range of this method is 4 to 200 g/L(ppb). Figure 1
shows the standard curve for TCE in water. With dilution, the dynamic range may be extended up to 200 mg/L
(ppm). Three points (5, 90, and 190 <5g/L) all fall within the dynamic range of the method, and thus were chosen
as the standardization points for the Envirometer.
Specificity of Reaction
The photochemical reaction which is utilized for the detection of volatile organic halides is interspecific towards
various organo halides, thus the method displays varying sensitivity toward different test compounds. This neces-
sitates understanding of the dominant analyte of interest for proper screening and quantitation. Relative sensitivi-
ties of various volatile organic halides to TCE, PCE, carbon tetrachloride and chloroform are given in Table 2.
222
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Sensitivity is clearly related to the extent of halogenation, with carbon tetrachloride being the most sensitive
compound and vinyl chloride the least.
Table 1. Quick Test VOH Water Test Method Detection and Quantitation Limits.
Chemical
TCE*
PCE
Carbon
Tetrachloride
Chloroform
Primary
Spike
Level
Og/L
5
4
4
5
Multiplier
Value
4
4
4
4
Secondary
Spike Level
Og/L
20
16
16
20
Standard
deviation
for (8) 16
replicate
analyses
1.4
1.7
1.1
1.8
t-statistic
multiplier
for (8) 16
replicates
3.00
2.60
2.60
2.60
MDL
Og/L
4
4
3
5
t-statistic
multiplier
for (8) 16
replicates
5.405
4.073
4.073
4.073
MQL
g/L
8
7
4
7
*TCE spiked at two concentrations and each concentration was
analyzed 8 times.
The MDL was calculated for each of the spiked concentrations
and the two MDLs were averaged.
All other chemicals were spiked at one concentration and
analyzed 16 times.
G7
£
•o .0
I*
ra
U
300
200
100
50 100 150 200
Spiked [TCE], ppb
Table 2. Method Performance Data as Percent Relative Sensitivity to TCE, PCE, Carbon Tetrachloride and
Chloroform
Trichloroethene
Perchloroethene
Carbon tetrachloride
Chloroform
1,1-Dichloroethene
Vinyl chloride
trans-1 ,2-Dichloroethene
cis-1 ,2-Dichloroehtene
Dichloromethane
1,1,1-Trichloroethane
1 , 1 ,2-Trichloroethane
1,2-Dichloroethane
Bromoform
Bromodichloromethane
Chlorobibromomethane
Compared with
TCE
100
82
114
82
69
0.8
61
43
20
112
80
15
77
75
71
Compared with
PCE
122
100
139
100
84
1.0
74
52
24
137
98
18
94
91
87
Compared with carbon
tetrachloride
88
72
100
72
61
0.7
54
38
18
98
70
13
68
65
63
Compared with
chloroform
122
100
139
100
84
1.0
74
52
24
137
98
18
94
91
87
Method Accuracy (Bias)
Method accuracy was determined by evaluating the percent recovery of TCE spiked in Type II water. Data was
generated using the field instrument for the spike concentrations shown in Table 3.
Method accuracy, as recovery, for the Quick Test ranged from 91-110 percent. The reported method accuracy for
halogenated volatiles by Method 8021B (U.S. EPA, 1996) is 96 percent for TCE and 86-109 percent for other
volatile organic halides. The recovery data obtained with the Quick Test Method for TCE exceed that reported
using Method 8021 B. The reported method accuracy range using the Quick Test is within limits for other volatile
organic halides as reported in Method 8021 B.
223
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 3. Method Accuracy Determined by Spiking Type II Water with 20 and 150 g/L TCE, and
was analyzed using the Quick Test procedure. The results obtained from this study are shown in Table 4.
Table 4. Method Precision Determined by Spiking Type II Water with 20 and 150 Og/L TCE.
TCE
Fortification
Og/L
20
150
Number of
Samples
10
10
Mean TCE Concentration
Determined by Quick Test
g/L
22
137
Standard
Deviation (s)
1.6
8.8
Coefficient of
Variation
(100*s/mean)
7.3
6.4
Range of
Concentration
g/L
19-24
126-156
For Method 8021B, precision was reported as 3.5 of the average recovery for a single operator using GC/HECD
for TCE and 1.5 to 9.9 for other volatile organic halides (U.S. EPA, 1996). For the Quick Test the standard devia-
tion of recovery for TCE was 7.3 and 6.4 for the two concentration levels. It is concluded that the method precision
for Quick Test VOH Water Test Method is comparable to standard method precision and is acceptable.
Chemical Interferences
An analysis of chemical interferences was performed using 250 mL of Type II water to assess the degree to which
other related or pertinent compounds would affect the measured TCE concentration. Table 5 is a summary of this
analysis.
No significant interference was observed for the compounds tested. 2,2,2-trichloroethanol had an interference
effect at 2,000 Og/L (100-fold) but no significant interference at 200 Og/L (10-fold).
Table 5. Results of Interference Analysis for the Quick Test for TCE Interfering Substance. TCE Spiked in Type II
Water at a Concentration of 20 Og/L.
lnterference=s
Concentration Required for a Detectable Interference (Og/L)
Benzene
Methanol
Toluene
Oxalic Acid
Sodium Trichloroacetate
Sodium Dichloroacetate
2,2,2-trichloroethanol
>2,000
>2,000
>2,000
>2,000
>2,000
>2,000
>200
False Positive/False Negative Study
False positive analysis for the Quick Test was performed using Type II water as the clean test matrix. The
suggested concentration of TCE for the false positive test is one-half of the MDL or 2.2 <&g/L (Lesnik, 1992).
Twenty replicate samples of the fortified water were analyzed using the Quick Test. The results of the false
positive analysis indicate that this method meets the criteria of no more than 10 percent false positive. The false
negative analysis for the Quick Test was performed using Type II water spiked at two times the MDL or 8.8
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
method meets the criteria of zero false negatives.
Matrix Suitability
Matrix specific performance data was evaluated using the Quick Test VOH Water Test Method. Four water matri-
ces: Type II water, Type II water with 20 mg/L humic acid, Type II water with 1,000 mg/L suspended solids and
buffered (pH 9) Type II water. Prior to fortification, each matrix was analyzed using the Quick Test and with
GC/ECD. All matrices were uncontaminated with respect to volatile organic halides. Matrix specific performance
data was generated by spiking the water matrices at 20 Og/L and 150 <5g/L TCE. The four water matrices were
spiked at two concentrations and analyzed using Method 8021B for verification of the spiking concentration. The
results obtained with the Quick Test VOH Water Test Method and Method 8021B were comparable and both are
subject to variability. There is no evidence of matrix interferences with the sample types tested.
Correlations Study
An independent evaluation of the test kit was preformed by the University of Waterloo. Samples were prepared by
spiking Type II water with TCE only or TCE with PCE. The prepared samples were analyzed in duplicate using the
test kit and standard methods. Results are presented in Table 6. For comparison of results, the reported amount
of PCE quantified using the standard method was adjusted by its sensitivity relative to TCE (Table 2).
Table 6. Inter-method Comparison between the Standard Laboratory Method and the Quick Test VOH Water
Test Method
Sample ID
TCE-1
TCE-2
TCE-3
TCE-4
TCE-5
TCE/PCE-1
TCE/PCE-2
TCE/PCE-3
TCE/PCE-4
TCE/PCE-5
Test Kit
(*9/L)
Replicate 1
30
57
192
1410
O/L
35
98
610
NA
O/L
Replicate 2
31
62
65
1270
O/L
35
100
540
2400
O/L
Mean
31
60
129
1340
O/L
35
99
575
2400
O/L
Standard Method
(a>g/L)
Replicate 1
16
68
309
1541
2834
22
135
577
3283
6287
Replicate 2
12
65
319
1528
2867
26
133
602
3228
5245
Mean
14
66
314
1534
2850
24
134
590
3256
5766
Percent
Difference
121
9
59
13
46
26
2.5
26
The Quick Test VOH Water Test Method has been used to measure volatile organic halides at the field site
contaminated with TCE. Table 7 outlines comparison data between the Quick Test and Method 8260 for measur-
ing volatile organic halides in water samples. For this comparison, the concentrations determined using GC/MS
were adjusted by their relative sensitivities compared with TCE as determined for the Quick Test (Table 2).
Regression analyses were used to determine if there was a relationship between the Quick Test and the confirma-
tory laboratory procedure. Similar analyses have been used by the U.S. EPA Superfund Innovative Technology
Evaluation (SITE) program to evaluate intermethod comparisons.6 Three components of the regression were
evaluated, the y-intercept, the slope and the coefficient of determination, r2 To meet Level 3 accuracy require-
ments, the r2 value must be between 0.85 and 1.0 and the slope and y-intercept must be within the 90 percent
confidence interval of their ideal values of 1.0 and zero, respectively.6 To meet Level 2 accuracy the r2 values
must be between 1.0 and 0.75 when the slope and intercept do not meet their ideal values. A Level 2 accuracy
requirement indicates a consistent relationship between the test and the confirmatory method but the relationship
is not 1:1. Table 8 displays the results from the regression analysis by Wilcoxon test method using data from
Tables 6 and 7. The Quick Test VOH Water Test Method meets Level 2 criteria, that is, there is a relationship
between the methods but the relationship is not 1:1.
225
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Table 7. Total VOH Quick Test Kit Results Comparison with Laboratory Data
Sample
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Total Volatile Organic Halide
by the Quick Test
Og/L
890
270
<4
6.4
99
172
153
57
1670
3025
2900
300
119000
12.9
_1 . ; * — _
Total Volatile Organic Halide
by Standard Methods
#g/L
959
206
<10
<10
102
145
165
66
1703
4858
4040
339
98208
<10
Percent Difference
7.2
31
NC
NC
2.9
19
10
14
1.9
38
28
12
21
NC
Table 8. Statistical comparison of Quick Test and Standard Methods Results for Data in Tables 6 and 7
concentration range
*g/i
5 to 4000
n
18
r2
0.974
y intercept
91.7*
slope
0.678
* y-intercept was not statistically different from 0 at a 90 percent confidence limit
Meeting Level 2 accuracy requirement indicates that there is a consistent relationship within the samples tested,
however the relationship between methods cannot be assumed to be statistically equivalent. Regression analysis
can be performed on select samples from a site to determine the relationship between methods, and therefore
Quick Test VOH results can be corrected using the regression equation generated.
Conclusions
The performance characteristics of a new field test, based on a photo-induced oxidation-reduction reaction
producing coloration proportional to the concentration of VOH (volatile organic halides) present, have been evalu-
ated. The average method detection limit (MDL) for the Quick Test VOH Water Test Method for volatile organic
halides is 4 Og/L in water. The dynamic range is 4 to 200 Og/L (ppb), which is useful for sites where cleanup
levels are within the stated dynamic range. With dilution the dynamic range can be extended up to 200 mg/L
(ppm). The Quick Test VOH Water Test Method meets standard method=s criteria, set by the U.S. EPA for
accuracy and precision. The matrix suitability study results show that the Quick Test VOH Water Test Method is
not subject to matrix effects. Independent correlation studies between Quick Test results and those reported for
field water samples analyzed by standard methods confirm that the Quick Test VOH Water Test System meets
U.S. EPA Superfund Innovation Technology Evaluation (SITE) Level 2 criteria for field testing. The quality control
(QC) procedures prescribed for the Quick Test are adequate and flexible to accommodate the intended uses of
this method. The Quick Test procedure is simple, easy to use and is optimized for quantitation of volatile organic
halides.
Reference
1. National Primary Drinking Water Regulations, TRICHLOROETHYLENE, US EPA, Office of Ground Water and
Drinking Water, Revised December 23, 1998.
National Primary Drinking Water Regulations, TETRACHLOROETHYLENE, US EPA, Office of Ground Water
and Drinking Water, Revised December 23, 1998.
3. U.S. Environmental Protection Agency, Test Method for Evaluating Solid Waste, Physical/Chemical Method
SW-846, U.S. Government Printing Office, Washington, DC, 1992.
4. B. Lesnik and P. Marsden, A Demystifying Methods Development @ Environ. Lab. Mag. July 1995.
5. Dong Chen, David Shattuck, Mark Hines and Joan McLean, Field Analytical Chemistry and Technology
1998.
2.
226
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
6. U.S. Environmental Protection Agency, APCP Immunoassay Technologies, Innovative Evaluation Report,
Report No. EPA/540/R-95/514, 1995.
227
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
228
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
LAB AUDITING AND
ACCREDIDATION
229
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
230
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
THE ROLE OF A COMPLIANCE PROGRAM AND DATA QUALITY REVIEW PROCEDURE UNDER PBMS
Ann Rosecrance
Core Laboratories, 5295 Hollister Road, Houston, TX 77040
(713)329-7414
INTRODUCTION
The trend away from sole reliance on method specified quality control (QC) to a performance based measure-
ments system (PBMS) creates the need for a broader based oversight program to ensure that environmental
project and regulatory program requirements are met. A strict QC program based on method compliance will not
be sufficient to ensure compliance with PBMS guidelines. Further, strict QC programs have not always been effec-
tive in ensuring method and project compliance and in preventing ethics violations.
Under PBMS, a comprehensive compliance program is warranted to help ensure compliance of all activities and
ethical performance of work, regardless of the method or project requirements. New approaches to data review
are needed to ensure that performance standards can be met. This paper provides guidance on key elements that
should be included in an effective compliance program and presents a data quality review procedure to use for
determining if data of acceptable quality can be generated.
IMPLEMENTING A COMPLIANCE PROGRAM
Ethics Policy or Statement
A compliance program must have an ethics policy or statement. This policy or statement should define the
company or organization's position on ethics and state what is expected of its employees or members with regards
to ethical behavior.
For example, a company's ethics policy may be the following:
"All employees at all times shall conduct themselves and the business of the Company in an honest and
ethical manner. Compliance with this policy shall be strictly enforced."
The ethics policy should be documented and posted for all employees to view. Companies may wish to further
affirm and document employee commitment to compliance with the ethics policy through an Employee Ethics
Agreement that each employee must sign as a condition of their employment.
Compliance Program Management
The compliance program should be managed by a senior management employee with the authority, skills and
availability to perform such an assignment. The compliance program manager should report to upper manage-
ment on a regular basis on the status of ethics activities within the organization. Companies may also elect to form
an Ethics Committee with members from their upper management staff or Board of Directors that meets on a
regular basis to set ethics policy and discuss ethics related matters.
Ethics Procedures
Policies and procedures for ethical conduct and for reporting and investigating suspected ethics violations should
be developed and included in the company's policy and procedures manual. An ethics procedure should define
ethical conduct and what constitutes unethical behavior and how it is handled. Disciplinary action for ethics viola-
tions, up to and including termination, should be stated in the ethics procedure. Fair procedures for reporting and
investigating alleged unethical behavior should be included in an ethics reporting and investigation procedure.
These procedures as well as other company procedures should be accessible to all employees.
Zero Tolerance Policy
Companies should have a zero tolerance policy on unethical activities and non-compliance with required proce-
dures. Unethical behavior or fraud may be defined as intentional falsification of data or records, such as sampling
or sample handling records, laboratory worksheets or logbooks, instrument settings or data, sample results or
data, and laboratory analysis reports. Unacceptable behavior may be defined as deliberate lack of adherence to
company and method requirements, such as procedures for instrument calibration, quality control, standards and
reagents preparation, sample handling, and sample preparation and analysis.
Laboratories may wish to go one step further and issue a policy that defines specific unacceptable and fraudulent
activities. Since most laboratory procedures define what employees are required to do, this policy ensures that
231
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
employees are educated as to what they are not allowed to do. Such a policy may include the following unaccept-
able and fraudulent activities: 1) making up data (dry labbing) or other sampling and analysis information; 2)
misrepresentation of QC samples and spikes as being extracted or digested when in fact they were not extracted
or digested; 3) improper clock setting (time traveling) or improper date/time recording; 4) improper peak integra-
tion (peak shaving or enhancing); 5) improper GC/MS tuning; 6) improper calibration/QC analysis; 7) file substitu-
tion; 8) deletion of non-compliant data; 9) improper alteration of analytical conditions; 10) unwarranted
manipulation of computer software; and 11) lack of notification to management on identified sample or data errors.
Laboratories that are proactive in informing employees of what constitutes unacceptable and fraudulent behavior
have a better chance of preventing fraud than laboratories that do not.
Ethics Assistance and Reporting Mechanism
Companies should a have a single point of contact for assisting employees with questions on ethics related
matters and for reporting observations of suspected unethical behavior or business conduct. A Helpline or Hotline
is such a mechanism where phone calls, faxes or other correspondence on ethics concerns, questions or reports
of suspected unethical behavior can be directed and then addressed appropriately. The phone numbers and
addresses for the Helpline or Hotline should be documented and readily available to all employees. The Helpline
or Hotline can be manned by a senior management employee, such as the compliance program manager, or by
an outside service.
Compliance Plan
A compliance plan should include or refer to all of the procedures used by an organization for ensuring compliance
with company, client and government requirements. The compliance plan should include or refer to company
policies and procedures on business conduct, especially ethics. Also include or refer to technical and quality
assurance procedures used by the laboratory and required by client, method or regulatory agencies to ensure that
data are accurate and traceable. The compliance plan should further include or refer to environmental manage-
ment activities and procedures used for chemical and waste handling to comply with federal, state and local
regulations. A compliance plan may also include a quality management program such as ISO 9002.
Compliance Training
Compliance training should be provided to all employees and include, at a minimum, training on the ethics policy
and procedures. Ethics training should be documented on training forms and included in the employee training or
personnel files. Training on laboratory procedures should be ongoing and based on each individual and their work
assignments.
Compliance Audits
Adherence to the compliance plan and associated procedures/requirements should be checked on a regular basis
via on-site audits. The compliance officer, quality assurance staff or outside consultants may conduct compliance
audits. Any findings of non-compliance with company, client or government requirements should be documented
and provided to management. Prompt and effective corrective action should be taken on any findings and reported
back to the auditing body for review and approval.
DATA QUALITY REVIEW
Despite the number of laboratory audits that are conducted at environmental testing laboratories, many of these
audits do not address data quality and thus do not identify data quality problems. Traditional audits tend to focus
on laboratory procedures and QC criteria rather than data quality. Probably the most important area that affects
the usability of sample data is not receiving the critical attention it should have.
A data quality review should be performed to determine if data of acceptable quality can be and are being gener-
ated by a given laboratory. This review does not replace on-site assessments that evaluate method compliance or
tape audits that evaluate the accuracy of reported data. The following items should be included in a data quality
review of organic analysis data, whether for PBMS methods or traditional methods. Similar principles apply to
inorganic analysis data.
Initial Demonstration of Competency Data
An initial demonstration of competency (IDC) study (also referred to as initial demonstration of capability or profi-
ciency study) demonstrates the ability of each analyst and instrument to achieve acceptable accuracy and preci-
sion for each analyte in each test method performed. It should be performed prior to performing sample analyses
and whenever there is a new analyst or major change in the instrumentation. An IDC study involves the
232
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
preparation and analysis of a minimum of four spiked samples at concentrations of 20 ug/L for volatiles, 100 ug/L
for semivolatiles and 2-50 ug/L for pesticides and PCBs.
First determine if IDC studies have been performed for each analyst and instrument. If not performed, note which
studies are needed for immediate action. If performed, review the data from each study and determine if each
target analyte was included. For each analyte, evaluate the spike value, found values, average percent recovery
and standard deviation (SD). Compare the average percent recovery and SD for each analyte to the method or
project specified acceptance range or values. If the average percent recovery is within the acceptance range, then
acceptable accuracy can be achieved. If the SD is less than the maximum allowable value, then acceptable preci-
sion can be achieved. If either criteria were not met, then note the analytes that require immediate action (repeat
of study.)
Method Detection Limits
A method detection limit (MDL) determination or study establishes the lowest concentration that the laboratory can
measure an analyte with 99% confidence. Using the procedure in 40 CFR Part 136 Appendix B, a MDL study
involves the preparation and analysis of a minimum of seven spiked samples at a concentration 1-5 times the
estimated MDL. The MDL is calculated by multiplying the standard deviation obtained for the seven measure-
ments by 3.14.
First determine if MDL studies have been performed for each method and analyte. If not performed, note which
studies are needed for immediate action. If performed, evaluate each study to determine if each target analyte
was included. For each analyte, evaluate the spike value, found values, average percent recovery, standard devia-
tion (SD) and calculated MDL. Compare the calculated MDL and the spike value. If the calculated MDL is greater
than the spike concentration, then the study should be repeated at a higher spike concentration. If the spike
concentration is greater than 10 times the calculated MDL, then the study should be repeated at a lower spike
concentration.
Laboratory Reporting Limits
Laboratory reporting limits (RLs) are the minimum values used by the laboratory to report sample data. Laborato-
ries typically use quantitation limits or values that are generally 5 to 10 times the MDLs for their RLs. For samples
that are diluted, the RLs must be multiplied by the sample dilution factor. Target analytes found in samples at
concentrations greater than the RLs are reported as numerical values. Target analytes not detected above the
corresponding RLs are reported as "not detected" or at a qualified value greater than the MDL.
First obtain and review the laboratory's RLs for each method, matrix and analyte. Then evaluate the RLs in water
for each method and analyte to determine if the laboratory RLs are greater than the MDLs (data for other matrices
may also be reviewed.) If any RLs are less than the associated MDLs, then note which analytes require immediate
action (Note: an error here means that the laboratory may be reporting data lower than it can actually measure.) If
the RLs are greater than or equal to the associated MDLs, then it can be expected that the laboratory's reports will
provide values that can be detected or backed up by laboratory measurements. Alternately, if MDLs are not avail-
able for certain analytes, the lowest calibration standard may be evaluated and compared to the laboratory RLs. If
any RLs are less than the lowest concentration calibration standard, then note the analytes that require immediate
action. If the RLs are greater than or equal to the lowest concentration calibration standard, then it can be
expected that the laboratory's reports will provide values that can be detected by calibration standards.
Initial Calibration Data
Initial calibration is performed to establish the calibration curve and range for each analyte.
Analyte Presence and Standard Concentration. First review recent initial calibration data for each method and
analyte. Also review the source and concentration for each initial calibration standard. Determine if all target
analytes were included in the calibration standards. If not, note any missing analytes for immediate action. Next
determine if the concentration values used for each analyte in the calibration table or curve match the actual
concentrations provided with the calibration standards. If the concentrations do not match, then note any analytes
that require immediate action (Note: this error could result in incorrect concentrations in samples.) If the values do
match, then the calibration table or curve can be considered accurate with regards to assigned standard concen-
tration. Also evaluate if surrogates were analyzed at multiple concentrations. Previous EPA SW-846 methods
allowed single concentrations but recent updates to SW-846, i.e., Update III and Method 8000B, require multi-
point concentration for surrogates as well as target analytes. If surrogates were not analyzed at multiple concen-
trations, then note which analyses are affected for immediate action.
233
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Analyte Identification. Evaluate the data for the lowest concentration standard analysis to determine if the identifi-
cation data for each target analyte is representative of that analyte, such as GC/MS mass spectrum or characteris-
tic ions, GC/MS "Q" value, GC retention time, elution order, etc. If not, note which analytes are questionable and
require'immediate action (Note: this error could result in incorrect analyte identification in samples.) If all analytes
are included and the data are representative, then the laboratory should be able to correctly identify target analytes
in samples.
Analyte Response. Evaluate the analyte response in each calibration standard to determine if the responses are
acceptable and proportionate to concentration. For GC/MS analyses, determine if the relative response factors
(RRFs) for each analyte are above the minimum required value. For each target analyte, evaluate if the responses
increase with concentration (e.g., the area for benzene in a 100 ppb standard should have twice the area as a 50
ppb standard.) If RRFs are below the minimum value or if responses are not proportionate to concentration, then
note the analytes that require immediate action. If the analyte responses are acceptable, then it can be expected
that the laboratory can acceptably measure responses for target analytes in samples.
Calibration Accuracy. Evaluate the calibration table or curve to determine if all data were used and that no points
in the middle of the calibration table or curve were deleted to force the calibration to meet certain criteria. Also
evaluate if manual integrations appear to be acceptable. The only points (concentrations) that should be deleted
from the calibration are low or high points that are outside the calibration range or points with a known error. If any
analytes were deleted from the middle of the calibration or if manual integration appears to be improper, then note
the analytes that require immediate action.
Next evaluate the %RSD for average RFs or RRFs for each analyte in the initial calibration and determine the
method used for sample quantitation. If the %RSD value for each analyte is less than or equal to 15%, it is accept-
able by EPA SW-846 methods to use RRF or RF for quantitation. If the %RSD is greater than 15% for any analyte,
evaluate if a linear or higher order calibration curve was used for quantitation and if the minimum number of
standards (5 for 1st order, 6 for 2nd order and 7 for 3rd order) were included in the calibration. If not, note the
analytes that require immediate action. If the correct number of standards were analyzed and the appropriate
technique is used for quantitation, then the initial calibration can be considered acceptable for sample quantitation.
Analytical Conditions. Also evaluate the conditions used for initial calibration to determine if the same conditions
were used for sample analysis (such as purging temperature for volatiles). If not, note the analyses that are
affected for immediate action.
Calibration Verification
Calibration verification is performed at a regular frequency (every 12 hours for GC/MS analysis and at the begin-
ning, end, and 5 to 10% of the runs for GC analysis) to verify that the current instrument performance is still
acceptable in comparison to performance during the initial calibration.
First review recent calibration verification data for each analysis. Also review the source and concentration for the
calibration verification analysis. Determine if the concentration values used in the calibration verification matches
the actual concentration provided with the calibration standard. If the concentrations do not match, then note any
analytes that require immediate action. If the values do match, then the calibration can be considered accurate
with regards to assigned standard concentration. Next evaluate the data for the calibration verification standard
analysis to determine if all of the target analytes were included and detected in the standard. If any target analytes
were not included or not detected, then note the analytes that require immediate action.
Evaluate the % difference (%D) from the expected value or the % recovery compared to the known value for each
target analyte in the calibration verification. Determine if the %D or % recovery for each analyte was within the
method or project specified acceptance values, generally +/- 15 to 20%. If not, note the analytes that require
immediate action. (Note: Action may not be necessary if the analyte(s) in question was not detected in any associ-
ated samples and the standard indicates that the analyte could be detected if it was present in a sample.) If the
%D or % recovery for each analyte was within the allowable values, then the calibration verification can be consid-
ered valid with regard to the initial calibration.
Laboratory Control Sample
A laboratory control sample (LCS) is a purchased or prepared sample with a known concentration of target
analytes taken through the entire sample preparation and analysis procedure and used to measure recovery.
234
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
First evaluate the analytes that were included in the LCS and their concentration values. Determine if the method
or project required analytes were included in the LCS and if the concentration was at the required value(s). Review
the source data for the LCS to determine if the LCS was from a different source or lot than the calibration
standards and if the concentration values assigned by the laboratory match the values from the source. If any
analytes or concentrations are incorrect, note the analytes that require immediate action. For each spiked analyte,
evaluate the spike value, found values and percent recovery. Compare the percent recovery for each analyte to
the method or project specified acceptance values. If the percent recovery is within the acceptance range, then
acceptable accuracy can be achieved. If not, note the analytes that require immediate action.
Laboratory Blanks
Laboratory blanks are analyzed to measure any background contamination introduced by the laboratory during the
sample preparation or analysis procedures. Laboratory blanks include method blanks, reagent blanks, calibration
blanks and holding or storage blanks.
Review blank data to determine if any analytes are present and at what concentrations. If target analytes are
present in the blank, review associated sample data to determine if the background in the blank could have a
significant affect on the sample values. If there are no detects for the affected analyte(s) in the sample or if the
analyte concentration is the sample is high, then low level background contamination will not have a significant
affect. If there are low level concentrations in the sample slightly above or near the blank level, then the sample
may be affected. Also review surrogate data in the blank to establish a baseline level with which to compare the
sample data. If surrogate recovery is acceptable in the blank, then unacceptable recovery in samples is probably
due to the sample and not laboratory performance. Note any unacceptable recovery of surrogates in blanks for
immediate action.
Sample Data
Last but not least are the sample data. Review sample data for surrogate recovery, internal standard response (if
internal standards are used), and analyte identification and quantitation. Determine if surrogates and internal
standards (if applicable) were added to each sample and if the surrogate recovery and internal standard
responses were within method or project specifications. If not, determine if corrective action was taken or if
additional analyses were performed. If reanalysis data still are not acceptable, then note the impact (low or high
bias) on sample results. Evaluate reported analytes in samples to determine if identification characteristics and
criteria were satisfied, such as GC/MS mass spectrum, GC/MS "Q" value, GC retention time and elution order. If
not, the analyte identification and presence may be suspect and sample results should be handled accordingly
(i.e., reprocessed or rejected.) Next determine if concentrations for found analytes were calculated and reported
correctly. If not, the analyte concentration may be incorrect and sample results should be handled appropriately
(i.e., recalculated or rejected.) Also review matrix spike and duplicate data if available for the same sample to
determine if the results for found analytes correlate between each analysis. Determine if non-spiked analytes
found in the original sample are also found in the matrix spike and duplicate at similar concentrations. If not, there
may be a lack of precision or an error in one or more of the analyses; sample results should be handled appropri-
ately (i.e., qualified or rejected.) Also review all sample documentation to determine if complete and consistent. If
not, note what is needed for immediate action.
For any of the items that require action, consult with the laboratory manager for correction and resolution. Data of
acceptable quality can be achieved when all of the above criteria are satisfied.
CONCLUSION
With PBMS on the horizon, environmental professionals may wonder what will happen to control of laboratory data
quality if adherence to strict method requirements is no longer mandatory. Data quality has not been guaranteed
by the traditional focus on method QC limits, and in fact many unethical practices have occurred in environmental
laboratories in order to meet QC limits. Change is disconcerting but necessary for improvement. By implementing
an effective compliance program and by conducting data quality review with the guidance provided in this paper,
ethics awareness and environmental data quality can be improved.
235
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
236
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Author
Amick, E.N.
Beaty, E.S.
Bauer, W.F.
Becker, D.A.
Benda, S.
Benner, Jr., B.A.
Benvenuti, M.
Block, E.
Boswell, C.E.
Boylan, H.M.
Broske, A.D.
Buckley, B.
Cain, R.
Calovini, F.
Carlin, Jr., F.J.
Carlson, R.E.
Chatham, D.M.
Chen, D.
Chen, M.T.
Chiu, C.
Crowder, C.A.
Cox, T.
Danahy, R.J.
Davidowski, L.
Demiralp, R.
Dunder, T.
Dupes, L.J.
Ebersold, P.J.
Emsbo-Mattingly, S.
Emsbo-Mattingly, S.
Evans, R.E.
Fisher, E.
Fitzgerald, J.
Forman, R.L.
Gere, D.R.
Glaser, J.P
Green, L.
Greenberg, R.
Gregg, D.
Groenjes, C.
Grosser, Z.
Grosser, Z.A.
Harrison, R.O.
Hassett, D.J.
Head, J.G.
Hewitt, A.D.
AUTHOR INDEX
Paper Page
No. No.
2
9
4
9
8
47
15
47
37 176
6
26
36 171
36 171
45 190
15 75
34 170
16 77
26 113
12 63
48 205
43 189
3
11
52 221
5
9
4
21
47
15
24 101
24 101
18 81
9
4
7
51
47
15
32
216
39 182
44
4
16
44
8
34
33
34
9
29
27
23
18
43
13
5
30
189
15
77
189
38
170
163
170
47
125
117
99
81
189
66
21
125
Author
Hewitt, A.D.
Hiatt, M.
Hiatt, M.
Mines, M.
Hoberecht, H.
Huo, D.
Jackson, P.E.
Jackson, T.A. '
Kendall, D.S.
Kingston, H.M.
Kingston, H.M.
Kirshen, N.A.
Kirshen, N.A.
Krautova, J.
Krol, J.
Latino, J.
LeMoine, E.A.
LeMoine, E.A.
Leyrer, M.
Lopez de Alda, M.
Lopez de Alda, M.
Masila, M.
Mauro, D.
McLean, J.
McMillin, R.
Melberg, N.
Murphy, K.E.
Nagourney, SJ.
Neuhaus, J.
Okamoto, H.
Orr, L.
Parks, D.
Parris, R.M.
Parris, R.M.
Patkin, A.
Patkin, A.
Penton, Z.
Phillips, J.H.
Podhola, B.
Poster, D.L.
Poster, D.L.
Quiroz, J.D.
Reed, G.
Reitmeyer, C.
Paper
No.
33
41
42
52
47
26
12
52
14
15
26
38
40
37
36
18
46
47
23
6
9
37
39
52
29
50
9
16
20
12
19
49
6
9
46
47
40
33
50
6
9
Page
No.
163
187
188
221
199
113
63
221
72
75
113
177
182
176
171
81
194
199
99
26
47
176
182
221
125
216
47
77
87
63
86
211
26
47
194
199
182
163
216
26
47
21
34
50
91
170
216
237
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
Author
Richter, B.E.
Ricker, M.J.
Romano, J.
Rosecrance, A.
Sadik, O.A.
Sander, L.C.
Sander, L.C.
Schabron, J.F
Schantz, M.M.
Schantz, M.M.
Schantz, M.M.
Schlemmer, G.
Sears, D.
Serapiglia, T.M.
Shattuck, D.
Smith, R-K.
Solsky, J.F.
Sorini, S.S.
Spencer, D.
Storne, K.A.
Sutton, C.
Sutton, C.
Tsui, D.T.
Turle, R.
Turriff, D.
Tutschku, S.
Vitale, R.J.
Vitale, R.J.
Vitale, R.J.
Vitale, R.J.
Wayne, T.
Wise, S.A.
Wise, S.A.
Wise, S.A.
Wolf, M.
Wolf, R.E.
Wool, L.
Van, F.
Yesso, J.D.
Young, M.
Zimmie, T.F.
Paper
No.
35
32
36
54
37
6
9
31
6
9
17
23
18
15
52
20
28
31
29
10
24
25
12
9
50
17
5
7
8
48
11
6
9
17
1
22
29
37
24
36
21
Page
No.
170
134
171
231
176
26
47
129
26
47
80
99
81
75
221
87
121
129
125
52
101
107
63
47
216
80
21
32
38
205
59
26
47
80
3
91
125
176
101
171
91
238
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
NOTES
239
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
240
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
241
-------
WTQA '99 - 15th Annual Waste Testing & Quality Assurance Symposium
U.S-
242
------- |