& EPA
United States
Environmental Protection
Agency
National Exposure
Research Laboratory
Research Triangle Park, NC 27711
EPA/600-R-98/161
September 1998
Research and Development
TECHNICAL ASSISTANCE
DOCUMENT FOR SAMPLING AND
ANALYSIS OF OZONE PRECURSORS
-------
Technical Assistance Document for
Sampling and Analysis of Ozone
Precursors
September 30, 1998
U.S. Environmental Protection Agency
National Exposure Research Laboratory
Human Exposure and Atmospheric Sciences Division
Research Triangle Park, North Carolina, NC 27711
-------
DISCLAIMER
This document has been reviewed in accordance with U.S. Environmental Protection
Agency policy and approved for publication. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
11
-------
TABLE OF CONTENTS
Page
Section 1 Introduction 1
1.1 Purpose 2
1.2 Organization 3
1.3 Summary of the Monitoring Regulations 4
1.4 References 8
Section 2 Methodology for Measuring Volatile Organic Compound Ozone Precursors in
Ambient Air 1
2.1 Network Monitoring Requirements 2
2.2 Target Volatile Organic Compound Ozone Precursors 4
2.2.1 Total Nonmethane Organic Compound (TNMOC) and PAMS
Hydrocarbons (PAMHC) 6
2.3 Chromatography Discussion and Issues 9
2.3.1 Gas Chromatography with Flame lonization Detection 9
2.3.2 Identification and Quantification Issues 12
2.3.3 Sample Moisture Issues 14
2.3.4 Calibration Standards 17
2.3.4.1 Primary Calibration Standard 17
2.3.4.2 Retention Time Calibration Standard 18
2.3.4.3 Calibration Standard Preparation 19
2.3.4.3.1 Procedure for Humidification 20
2.3.4.3.2 Calibration Standard Dilution Procedure 30
2.3.5 Column Configurations 36
2.3.6 Column Selection 37
2.3.7 Pre-measurement Chromatographic System Verification 46
2.3.7.1 Retention Times and Relative Retention Times 46
2.3.7.2 Internal Standards 49
2.3.7.3 Identification of Co-Eluting Compounds and
Matrix Effects 50
2.3.7.4 Detection Limits 51
2.4 Automated Method for Collecting and Analyzing Volatile Organic Compound
Ozone Precursor Samples 51
2.4.1 Sample Collection 52
2.4.1.1 Sample Probe and Manifold 53
2.4.1.2 Sample Introduction 55
2.4.1.3 Sample Conditioning 58
2.4.1.4 Sample Concentration 58
in
-------
TABLE OF CONTENTS (Continued)
Page
2.4.2 Sample Analysis 59
2.4.2.1 Sample Focusing or Cryofocusing 59
2.4.2.2 Gas Chromatography 60
2.4.2.3 Analytical System Calibration 61
2.4.3 System Operation 62
2.4.3.1 Initial System Set-up 63
2.4.3.2 Sampling Parameters 64
2.4.3.3 Field Operation 64
2.4.4 System Specifications 65
2.5 Manual Method for Collecting and Analyzing Volatile Organic Compounds . . 66
2.5.1 Sample Collection 67
2.5.1.1 Multiple-event Sample Collection Equipment 68
2.5.1.2 Multiple-event Sample Collection Procedure 71
2.5.1.3 Multiple-event System Specifications 75
2.5.1.4 Single-event Collection Equipment 77
2.5.1.5 Single-event Sample Collection Procedure 80
2.5.1.6 Single-event System Specifications 82
2.5.1.7 Canister Sampling System Certification 83
2.5.1.7.1 Certification Equipment 87
2.5.1.7.2 Certification Procedure 88
2.5.2 Canister Cleaning 90
2.5.2.1 Canister Cleaning Equipment 91
2.5.2.2 Canister Cleaning Procedure 94
2.5.2.3 Canister Blanking Procedure 96
2.5.2.4 Final Canister Evacuation Procedure 97
2.5.3 Canister Sampling Issues 98
2.5.3.1 Precautions in the Use of Canisters 99
2.5.3.2 Contamination 99
2.5.3.3 Sample Stability 101
2.5.3.4 Positive Pressure Samples 103
2.5.3.5 Diluted Samples 104
2.5.3.6 Canister Leakage 104
2.5.4 Sample Analysis 106
2.5.4.1 Sample Introduction 106
2.5.4.2 Sample Conditioning 106
2.5.4.3 Sample Concentration 108
2.5.4.4 Sample Focusing or Cryofocusing 109
2.5.4.5 Gas Chromatography 109
2.5.4.6 Analytical System Calibration 110
IV
-------
TABLE OF CONTENTS (Continued)
Page
2.5.5 System Operation Ill
2.5.5.1 Initial System Set-up Ill
2.5.5.2 Sampling Parameters 112
2.5.5.3 Operation 112
2.6 Data Processing Capabilities of Automated VOC Systems and Submittal of
VOC Data to the AIRS AQS Data Base 113
2.6.1 Data Processing Capabilities of Automated VOC Measurement
Systems 114
2.6.1.1 Data Acquisition and Processing 116
2.6.2 AIRS AQS Data Submittal 120
2.6.2.1 Initial AIRS AQS Setup 123
2.6.2.2 Site and Monitor File Updates 125
2.6.2.3 Raw Data Transactions 131
2.6.2.4 Submitting Data 149
2.7 Validating Data from Automated VOC Systems 151
2.7.1 Data Validation Approach 153
2.7.1.1 Routine Procedures 154
2.7.1.2 Tests for Internal Consistency 161
2.7.1.3 Historical Data Comparisons 169
2.7.1.4 Parallel Consistency Checks to Identify Systematic Bias . 173
2.7.2 Treatment of Outliers 173
2.8 Quality Control and Quality Assurance for VOC Measurements 174
2.8.1 Data Quality Objectives 175
2.8.2 Quality Control 183
2.8.2.1 Sample Collection 183
2.8.2.1.1 System Certification 185
2.8.2.1.2 Calibration of Manual Sampling System
Components 185
2.8.2.1.3 Collection of Field Duplicate Samples 187
2.8.2.1.4 Preventive Maintenance 187
2.8.2.2 Sample Handling and Custody 188
2.8.2.3 Sample Analysis 190
2.8.2.4 Data Documentation and Archives 194
2.8.3 Quality Assurance 196
2.8.3.1 Development of Standard Operating Procedures 197
2.8.3.2 QA Program Guidance 199
2.8.3.2.1 Audit Types 199
2.9 References 202
-------
TABLE OF CONTENTS (Continued)
Page
Section 3 Determination of Total Nonmethane Organic Compounds Using
Method TO-12 1
3.1 References 2
Section 4 Methodology for Measuring Oxides of Nitrogen and Total Reactive Oxides of
Nitrogen in Ambient Air 1
4.1 Oxides of Nitrogen 1
4.1.1 Measurement Principle 2
4.1.2 Method and Equipment Description 4
4.2 Total Reactive Oxides of Nitrogen 6
4.3 Measurement of Total Reactive Oxides of Nitrogen in the Atmosphere
(Gas Phase Chemiluminescence)-Draft Instrumental Method 7
4.3.1 Applicability 8
4.3.2 Principle of Measurement 8
4.3.3 Measurement Apparatus 9
4.3.3.1 Configuration 9
4.3.3.2 Reconfiguration 12
4.3.3.2.1 Shelter 12
4.3.3.2.2 Plumbing 13
4.3.3.2.3 Electronics 14
4.3.4 Calibration 15
4.3.4.1 Apparatus 16
4.3.4.1.1 Air Flow Controllers 16
4.3.4.1.2 NO Flow Controller 16
4.3.4.1.3 Air Flowmeters 16
4.3 A.I A NOFlowmeter 18
4.3.4.1.5 Pressure Regulator for Standard NO Cylinder . 18
4.3.4.1.6 Ozone Generator 18
4.3.4.1.7 Valve 18
4.3.4.1.8 Reaction Chamber 18
4.3.4.1.9 Mixing Chamber 19
4.3.4.1.10 Output Manifold 19
4.3.4.1.11 Valve 19
4.3.4.2 Reagents 19
4.3.4.2.1 NO Concentration Standard 19
4.3.4.2.2 Zero Air 20
4.3.4.3 Dynamic Parameter Specification 20
43 A A Calibration Procedure 23
VI
-------
TABLE OF CONTENTS (Continued)
Page
4.3.4.5 Determination of Converter Efficiency 29
4.3.4.6 Frequency of Calibration 29
4.3.4.7 Analyzer Challenge 30
4.4 Nitric Acid Measurement 30
4.5 References 30
Section 5 Methodology for Determining Carbonyl Compounds in Ambient Air 1
5.1 Ozone Scrubbers 4
5.1.1 Denuder Ozone Scrubber 5
5.1.1.1 Denuder Ozone Scrubber Equipment 6
5.1.1.2 Denuder Ozone Scrubber Operational Procedure 8
5.1.2 Cartridge Ozone Scrubber 8
5.1.2.1 Cartridge Ozone Scrubber Equipment 9
5.1.2.2 Cartridge Ozone Scrubber Operation Procedure 9
5.2 Multiple-event Sample Collection Systems 11
5.2.1 Multiple-event Collection System Equipment 11
5.2.2 Multiple-event Sampling Procedures 14
5.2.3 Sample Probe and Manifold 16
5.2.4 Multiple-event System Specifications 21
5.3 Process Blanks 25
5.3.1 Blank Criteria 25
5.3.2 Frequency of Collection 27
5.4 Breakthrough Analysis 28
5.5 Collection of Collocated Samples 28
5.6 Quality Assurance and Quality Control 28
5.7 General Cartridge Handling Guidelines 29
5.8 References 31
Section 6 Guidance for PAMS Meteorological Monitoring 1
6.1 Background 1
6.2 PAMS Site Types 3
6.3 Application of PAMS Meteorological Data 4
6.4 Surface Meteorological Monitoring 5
6.4.1 Siting and Exposure 5
6.4.2 Specifications 6
6.4.3 Wind Speed and Wind Direction 6
6.4.4 Temperature 8
6.4.5 Atmospheric Humidity 9
vn
-------
TABLE OF CONTENTS (Continued)
Page
6.4.6 Solar Radiation 10
6.4.7 Ultraviolet Radiation 11
6.4.8 Barometric Pressure 12
6.4.9 Precipitation 13
6.5 Upper-Air Meteorological Monitoring 13
6.5.1 Siting and Exposure 16
6.5.2 Aircraft 16
6.5.3 Tall Towers 17
6.5.4 Balloon Systems 17
6.5.5 Ground-Based Remote Sensors 19
6.5.6 Estimation of Mixing Height 22
6.6 References 23
APPENDICES
A Compendium Method TO-15. Determination of Volatile Organic Compounds
(VOCs) in Air Collected-Prepared Canisters and Analyzed by Gas
Chromatography/Mass Spectrometry (GC/MS).
B Humidity
C Method TO-12. Method for the Determination of Non-Methane Organic
Compounds (NMOC) in Ambient Air Using Cryogenic Preconcentration and
Direct Flame lonization Detection (PDFID)
D Compendium Method TO-11 A. Determination of Formaldehyde in Ambient Air
Using Adsorbent Cartridge Followed by High Performance Liquid
Chromatography (HPLC) [Active Sampling Methodology]
Vlll
-------
LIST OF FIGURES
Page
1-1 Isolated Area Network Design 7
2-1 Water Content of Air at 75% and 100% Relative Humidity Over a Range of
Temperatures 24
2-2 Configuration of Materials to Perform Direct Injections of Water into the Canister
Before Filling with Dry Calibration Gas 27
2-3 Configuration of Materials to Perform Injection of Water Through a Heated Tee While
Filling with Dry Calibration Gas 28
2-4 Calibration Standard Dilution System 31
2-5 Example Chromatogram for the PAMS Target Compounds from the PLOT Analytical
Column 39
2-6 Example Chromatogram for the PAMS Target Compounds from the BP1 Analytical
Column 40
2-7 Representative Ambient Air Sample Analyzed on a PLOT Column 41
2-8 Representative Ambient Air Sample (Same as Figure 2-6) Analyzed on a
BP-1 Column 42
2-9 Vertical Configuration 56
2-10 Horizontal Configuration 57
2-11 A Typical Multiple-Event Sample Collection System 69
2-12 A Typical Single-Event Sample Collection System 78
2-13 Dedicated Manifold for Zero Gas Certification 85
2-14 Dedicated Manifold for Challenge Gas Certification 86
2-15 Schematic of a Canister Cleanup System 92
2-16 Report Generating Process 115
IX
-------
LIST OF FIGURES
Page
2-17 Column Placement for a Type Fl Transaction 128
2-18 Column Placement for a Type A6 Transaction 132
2-19 Column Placement for an Hourly Data Transaction 140
2-20 Column Placement for a Daily Data Transaction 146
2-21 Flow of Data Validation Activities 155
2-22 Appearance of Calibration Data at East Hartford, CT, in June 1995. Example scatter
plot showing calibration data of about 30 ppbC. Data are level 0, preliminary
data, CT DEP 160
2-23 Example PE Turbochrom® Summary Report 163
2-24 Time Series Plot 166
2-25 Time Series Plot of Several Species Groups at Stafford, CT, in 1994. Example of
misidentification of a paraffin for an unidentified peak 167
2-26 Time Series Example of System Contamination 168
2-27 Example of Peak Misidentification Using a Scatter Plot. Typically, data points would
be present in the region of the plot between the two extreme edges 170
2-28 Example of a Typical "Fingerprint" Observed at a PAMS Site 171
2-29 Example of Calibration Gas "Fingerprint" Observed in Data Submitted to AIRS .... 172
4-1 Flow Schematic of a Typical NO-NO2 Instrument 5
4-2 Flow Schematic of the Reconfigured NOy Instrument for PAMS Application 11
4-3 Flow Schematic of a Typical Gas Phase Titration Calibration System 17
5-1 Cross-Sectional View of the Denuder O3 Scrubber 7
5-2 Cross-Sectional View of the Cartridge O3 Scrubber 10
-------
LIST OF FIGURES
Page
5-3 Schematic of a Typical Multiple-Event Carbonyl Cartridge Sampling System 12
5-4 Vertical Configuration 19
5-5 Horizontal Configuration 20
5-6 Isolated Area Network Design 23
XI
-------
LIST OF TABLES
Page
1-1 PAMS Minimum Monitoring Network Requirements 5
2-1 Target Volatile Organic Compounds 5
2-2 Target VOC Compound Classification 7
2-3 Vapor Pressure of Water Below 100°C, mm HG 22
2-4 Peak Identifications, Ambient Air Sample 43
2-5 AIRS Transaction Types 122
2-6 Current AIRS AQS Regional Coordinators 124
2-7 AIRS Sampling Frequency Codes 127
2-8 Configuration Comments for Type A6 or A7 Transactions 130
2-9 Target Volatile Organic Compounds 133
2-10 Carbonyl Target List 134
2-11 AIRS Method Codes 135
2-12 AIRS Interval Codes 137
2-13 AIRS Unit Codes 138
2-14 Hourly Sample Valid Start Hour Based on the Interval 138
2-15 Null Values 139
2-16 QC Objectives for VOC Sample Collection 184
2-17 VOC QC Procedures 191
2-18 Format for Standard Operating Procedures 198
5-1 PAMS Minimum Monitoring Network Requirements 2
xii
-------
LIST OF TABLES
Page
5-2 Example Schedule for the Collection of Blanks 27
5-3 Quality Assurance and Quality Control Criteria 30
6-1 Overview of PAMS Meteorological Monitoring Requirements 2
6-2 Application of the PAMS Meteorological Data 4
6-3 System Specifications for Surface Meteorological Measurement 6
6-4 Principles of Humidity Measurement 10
6-5 Classification of Pyranometers 12
6-6 Capabilities and Limitations of Meteorological Measurement Systems for Vertical
Profiling of the Lower Atmosphere 15
6-7 Manufacturers' Specifications for Sensors Used in Rawinsondes 17
6-8 Functional Precision of Rawinsonde Measurements 18
Xlll
-------
Section: 1
Revision: 1
Date: 09/30/98
Page: 1 of 8
Section 1.0
Introduction
Section 182 (c)(l) of the 1990 Clean Air Act Amendments (CAAA) required the
Administrator to promulgate rules for enhanced monitoring to obtain more comprehensive and
representative data on ozone air pollution. The Environmental Protection Agency (EPA) has
revised the ambient air quality surveillance regulations in Title 40 Part 58 of the Code of Federal
Regulations (40 CFR Part 58)1 to include provisions for enhanced monitoring of ozone (O3),
oxides of nitrogen (NOX), volatile organic compounds (VOCs), selected carbonyl compounds,
and monitoring of meteorological parameters. The revisions require States to establish
Photochemical Assessment Monitoring Stations (PAMS) as part of their existing State
Implementation Plan (SIP) monitoring networks in ozone nonattainment areas classified as
serious, severe, or extreme.
The principal reasons for requiring the collection of additional ambient air pollutant and
meteorological data are the lack of successful attainment of the National Ambient Air Quality
Standard (NAAQS) for O3, and the need to obtain a more comprehensive air quality data base for
O3 and its precursors. Analysis of the data will help the EPA understand the underlying causes of
ozone pollution, devise effective controls, and measure improvement. Data acquired from
enhanced ambient air monitoring networks will have a variety of uses, which may include:
• Developing, evaluating, and refining new O3 control strategies;
• Determining NAAQS attainment or non-attainment for O3;
• Tracking VOCs and NOX emissions inventory reductions;
• Providing photochemical prediction model input;
• Evaluating photochemical prediction model performance;
-------
Section: 1
Revision: 1
Date: 09/30/98
Page: 2 of 8
• Analyzing ambient air quality trends; and
• Characterizing population exposure to VOCs and O3.
1.1 Purpose
The Technical Assistance Document (TAD) for Sampling and Analysis of Ozone
Precursors was initially published in October 1991. The document was intended to provide
guidance to those responsible for implementing PAMS. Since the initial publication, there has
been a draft revision in October 1994 to Sections 2.0, 4.0, and 5.0 and a revision to Section 6.0 in
June 1995, all of which were included in Appendix N of the PAMS Implementation Manual.
Since these revisions, there have been significant advances in the methodology used to measure
the components and parameters of interest at PAMS. These advances have necessitated this
revision of the TAD.
The purpose of this document is to provide guidance in support to the enhanced ozone
monitoring revisions in 40 CFR Part 58. The document provides technical information and
guidance to Regional, State, and local Environmental Protection Agencies responsible for
measuring O3 precursor compounds in ambient air. Sampling and analytical methodology for
speciated VOCs, total nonmethane organic compounds (NMOC) and selected carbonyl
compounds (i.e., formaldehyde, acetaldehyde, and acetone) are specifically addressed. The
document also addresses methodology for measuring NOX, as required by PAMS, and discusses
issues associated with the collection of total reactive oxides of nitrogen (NOy) and
meteorological measurements.
The technical guidance provided for measuring O3 precursors is based on emerging and
developing technology. Guidance for automated applications, in particular, is based on
experience obtained from the application of this technology during the beginning years of PAMS
implementation. Because these methods are based on emerging technology and reflect state-of-
-------
Section: 1
Revision: 1
Date: 09/30/98
Page: 3 of 8
the-art, they will be subject to continuing evaluation and improvements or clarifications in the
future.
Users should consider this guidance a basic reference to assist in developing and
implementing their PAMS monitoring program. The technical assistance document is prepared
in a document control format to accommodate revisions that are anticipated as the emerging
technologies develop.
1.2 Organization
The guidance provided in Section 2 of this document addresses the measurement of
volatile organic O3 precursors and includes method descriptions for manual and automated
sample collection and analysis. Detailed discussions are presented on selected topics such as
which volatile organic O3 precursors to measure, critical chromatography issues, moisture
control, data validation, Quality Control and Quality Assurance, AIRS data entry, and how
canister sampling should be approached.
Section 3 discusses the measurement of total NMOC using Method TO-12 from the
Compendium of Methods for Sampling and Analysis of Toxic Organic Compounds in Ambient
Air.2 Measurement of total NMOC by Method TO-12 has limited application to the
implementation of the 40 CFR Part 58 requirements, but Method TO-12 is included because of
its applications to canister cleanliness verification, application to alternative monitoring
strategies, and use in O3 prediction models. Alternative monitoring strategies involve the use of
canister and/or automated Method TO-12 complemented with an extensive canister sampling and
manual VOCs speciation analysis program.
Section 4 addresses the measurement of NOX and issues associated with NOy. Section 5
addresses the measurement of selected carbonyl compounds using Compendium Method
TO-11A from the Compendium of Methods for Sampling and Analysis of Toxic Organic
Compounds in Ambient Air and includes new information regarding the methodology and issues
-------
Section: 1
Revision: 1
Date: 09/30/98
Page: 4 of 8
associated with the measurement of carbonyl compounds. Section 6 provides guidance for
PAMS meteorological monitoring, which is essential to the PAMS program. Note that all
sections of this Technical Assistance Document are intended to be independent. Figures, tables,
and text are therefore repeated as necessary.
1.3 Summary of the Monitoring Regulations
The 1990 CAAA required EPA to promulgate regulations to enhance existing ambient air
monitoring networks. Existing SIP stations are identified as State and Local Agency Monitoring
Stations (SLAMS) and National Air Monitoring Stations (NAMS). The enhanced O3 monitoring
stations are a subset to SLAMS and identified as Photochemical Assessment Monitoring Stations
(PAMS).
The monitoring revisions by EPA required changes to 15 separate Sections, Subparts, or
Appendices of 40 CFR Part 58, and varied in complexity and impact on State and local agencies.
The areas of the revised 40 CFR Part 58 regulations most relevant to enhanced ambient air
monitoring are operating schedules, PAMS methodology, and quality assurance. Section 58.13
of 40 CFR Part 58 contains the operating schedule for SLAMS, NAMS, and PAMS. This
section requires sampling for VOCs and carbonyl compounds according to the monitoring period
and minimum monitoring network requirements specified in Sections 4.3 and 4.4 of Appendix D
of the revised regulations.1
Unlike the SLAMS and NAMS design criteria which are pollutant-specific, PAMS
design criteria are site specific. Design criteria for the PAMS network are based on selection of
an array of site locations relative to O3 precursor sources and predominant wind direction
associated with peak O3 events. Four PAMS site types are described in the regulations. The
number and type of monitoring sites and sampling requirements is dependent on the population
of the Metropolitan Statistical Area (MSA) or Consolidated Metropolitan Statistical Area
(CMSA). The specified minimum sampling requirements for VOCs and carbonyl compounds for
each site type are presented in Table 1-1. Monitoring for O3 and NOX (including NO and NO2)
-------
Section: 1
Revision: 1
Date: 09/30/98
Page: 5 of 8
Table 1-1 . PAMS Minimum Monitoring Network Requirements
Population
ofMSA/CMSA1
Less than 500,000
500,000 to 1,000,000
1,000,000 to
2,000,000
More than 2,000,000
Required
Site Type
(1)
(2)
(1)
(2)
(3)
(1)
(2)
(2)
(3)
(1)
(2)
(2)
(3)
(4)
Minimum VOCs
Sampling
Frequency2
AorC
AorC
AorC
B
AorC
AorC
B
B
AorC
AorC
B
B
AorC
AorC
Minimum Carbonyl
Compounds Sampling
Frequency2
_
DorF
-
E
-
-
E
E
-
-
E
E
-
-
'Whichever area is larger.
frequency requirements are as follows:
A = Eight 3-hour samples every third day and one additional 24-hour sample every sixth day during
the monitoring period.
B = Eight 3-hour samples every day during the monitoring period and one additional 24-hour sample
every sixth day year-round.
C = Eight 3-hour samples on the 5 peak O3 days plus each previous day, eight 3-hour samples every
sixth day and one additional 24-hour sample every sixth day during the monitoring period.
D = Eight 3-hour samples every third day during the monitoring period.
E = Eight 3-hour samples on the 5 peak O3 days plus each previous day and eight 3-hour samples
every sixth day during the monitoring period.
F = Eight 3-hour samples on the 5 peak O3 days plus each previous day, eight 3-hour samples every
sixth day and one additional 24-hour sample every sixth day during the monitoring period.
-------
Section: 1
Revision: 1
Date: 09/30/98
Page: 6 of 8
requires continuous measurements. The sampling schedule applicable to a specific area is
dependent on population and PAMS site types. Specific monitoring objectives are associated
with each sampling location. An example of an isolated area network design shown in
Figure-1-1 identifies the location of the four PAMS site types referred to in Table 1-1.
The EPA has also prepared a guidance document on enhanced O3 monitoring network
design and siting criteria3 which provides assistance regarding the number of PAMS required,
station location, and probe siting criteria. The PAMS site types are described below.
Type (1) PAMS characterize upwind background and transported O3 and precursor
concentrations entering the MSA or CMS A and are used to identify those areas subjected to
overwhelming transport. Type (2) PAMS monitor the magnitude and type of precursor emissions
in the area where maximum O3 precursor emissions are expected and are also suited for
monitoring urban air toxic pollutants. Type (3) PAMS characterize O3 precursor concentrations
occurring downwind from the area of maximum emissions. Type (4) PAMS characterize
extreme downwind transported O3 and its precursor concentrations exiting the area and identify
those areas which are potential contributors.
Appendix A of 40 CFR Part 58 references the Quality Assurance Handbook for Air
Pollution Measurement Systems, Volume IV (Revised March, 1995) - Ambient Air Specific
Methods4 for general quality assurance recommendations for PAMS. Quality assurance
procedures for VOC, NOX, O3, and carbonyl and meteorological measurements must be
consistent with EPA guidance. This guidance, and other information from appropriate sources,
including Section 2.8 of this document, should be used by States in developing a Quality
Assurance program.
-------
Section: 1
Revision: 1
Date: 09/30/98
Page: 7 of 8
Isolated Area Network Design
(4)
(3)
Central business district
Urbanized fringe
U1
U3
Note:
U1 and U2 represent the first and second most
predominant high ozone day morning wind direction.
U3 represents the high ozone day afternoon wind direction.
o: s/g/m orr/3 7 9 7/p ams /i s ol ate d. p p t
(1), (2), (3), and (4) are different types of PAMS sites (See Table 1-1).
Figure 1-1. Isolated Area Network Design
-------
Section: 1
Revision: 1
Date: 09/30/98
Page: 8 of 8
Appendix C of 40 CFR Part 58 requires that methods used for O3 and NOX be reference
or equivalent methods. Because there are no reference or equivalent methods promulgated for
VOC and carbonyl measurements, Appendix C of the revisions refers agencies to this guideline
document for direction.
Appendix C of the revisions would also allow the use of approved alternative VOC
measurement methodology (including new or innovative technologies). This provision requires
States that pursue alternatives to the methodology described herein to provide details depicting
rationale and benefits of their alternative approach in their network description as required in
40 CFR Part 58, Section 58.40 - PAMS Network Establishment.
1.4 References
1. U.S. Environmental Protection Agency. Code of Federal Regulations. Title 40, Part 58.
Ambient Air Quality Surveillance, Final Rule Federal Register, Vol. 58, No. 28,
February 12, 1993.
2. Compendium of Methods for the Determination of Toxic Organic Compounds in
Ambient Air. Compendium Method TO-12. Method for the Determination of Non-
Methane Organic Compounds (NMOC) in Ambient Air Using Cryogenic
Preconcentration and Direct Flame lonization Detection (PDFID). EPA-600/4-89/017.
Research Triangle Park, NC: U.S. Environmental Protection Agency, 1988.
3. Photochemical Assessment Monitoring Stations Implementation Manual. EP A-454/B-
93-051. Research Triangle Park, NC: U.S. Environmental Protection Agency, 1994.
4. Quality Assurance Handbook for Air Pollution Measurement Systems, Volume IV -
Ambient Air Specific Methods. EPA 600/4-77-27a. U.S. Environmental Protection
Agency, 1977. Revised March, 1995.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 1 of 225
Section 2.0
Methodology for Measuring Volatile Organic Compound Ozone
Precursors in Ambient Air
In accordance with the provisions for the enhanced O3 ambient monitoring network
requirements specified in 40 CFR Part 58, Subpart E,l this section provides guidance and method
descriptions for measuring volatile organic compounds (VOCs) that are considered to contribute to the
formation of ozone in the right atmospheric conditions. Information and guidance are provided to assist
in the development, implementation, and use of these methods for designing a VOC measurement
program consistent with the requirements of the enhanced O3 monitoring rule. Areas addressed
include:
• A review of the network monitoring requirements;
• A list of target VOCs to be measured;
• Chromatography issues associated with peak identification and quantification;
• Automated and manual methodology for collecting and analyzing samples;
• The minimum requirements of a Quality Assurance (QA) and Quality Control (QC)
program;
• Guidance for validating data from automated GC systems; and
• Submitting data into the AIRS AQS data base.
Measuring VOCs is a complex process involving the application of gas chromatographic
techniques for qualitative and quantitative determination of individual hydrocarbon compounds and an
estimation of total non-methane organic compound (TNMOC) content in ambient air. Two methods
are presented for collecting and analyzing VOC samples: an automated method (Section 2.4) and a
manual method (Section 2.5). Ideally, agencies responsible for designing, implementing, and operating
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 2 of 225
their O3 monitoring networks will satisfy their monitoring requirements by using some combination of the
automated and manual gas chromatographic approaches. Even if agencies primarily choose the
automated methodology, manual sampling and analysis capability are needed to fulfill the 24-hour
sample requirement; verify the proper operation of the automated systems; characterize the quality of
the collected data; address the identification of unknown compounds; and enhance the
representativeness of the monitoring network.
Users are ultimately responsible for equipment selection, set-up, parameter optimization, and
preparation of Standard Operating Procedures (SOPs) for their specific network. Because of the
complexity of the measurement process and the numerous choices of instrumentation (e.g., sampling
equipment, gas chromatographs, data acquisition hardware and software, etc.), the method descriptions
presented are generic. Background information on the potential benefits and limitations of the methods
are also provided.
2.1 Network Monitoring Requirements
The minimum sampling frequency requirements for speciated VOC monitoring are prescribed
in 40 CFR Part 58, Subpart E, Appendix D - Network Design for State and Local Air Monitoring
Stations (SLAMS! National Air Monitoring Stations (NAMS). and Photochemical Assessment
Monitoring Stations (PAMSY Section 4.3 - Monitoring Period requires, at a minimum, that O3
precursor monitoring be conducted annually throughout the months of June, July, and August when
peak O3 values are expected. Section 4.4 - Minimum Monitoring Network Requirements specifies the
minimum required number and type of monitoring sites and sampling frequency requirements based on
the population of the affected MSA/CMSA or nonattainment area, whichever is larger. These
monitoring requirements are outlined in Table 1-1. The minimum speciated VOC sampling frequency
requirements are summarized by site type below:
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 3 of 225
Site Type 1 - Eight 3-hour samples every third day and one additional 24-hour sample
every sixth day during the monitoring period; or eight 3-hour samples on the five peak
O3 days plus each previous day and eight 3-hour samples and one 24-hour sample
every sixth day, during the monitoring period.
Site Type 2 - (population less than 500,000) - Same as Site Type 1.
Site Type 2 - (population greater than 500,000) - Eight 3-hour samples every day
during the monitoring period and one additional 24-hour sample every sixth day year
around.
Site Type 3 - (population greater than 500,000) - Same as Site Type 1.
Site Type 4 - (population more than 2,000,000) - Same as Site Type 1.
Either of the two VOC methods (automated or manual) described in this section is capable of
satisfying the sampling frequency and sample integration requirements. Samples collected for either
method should represent a time-integrated average for the required sampling period. It is important to
understand that the 3-hour sample integration period is a maximum requirement in the sense that
samples can be collected more frequently at shorter sampling intervals (i.e., three 1-hour periods) but
not less frequently for longer sampling intervals.
The manual methodology, where samples are collected in canisters, is primarily applicable to
the less frequent sampling required for site Types 1, 3, and 4 (i.e., eight 3-hour samples every third day
or during peak O3 events) and the 24-hour sample requirement. The automated method, which allows
for direct on-line sample collection, is primarily applicable to the more frequent sampling requirements
for Site Type 2 (eight 3-hour samples every day during the monitoring period). The automated method
provides a viable option for the continuous collection of hourly samples. Though not required,
continuous collection of hourly samples also offers a more definitive assessment of the temporal and
diurnal distribution of VOCs. Although it is possible to use the manual methodology for Site Type 2
sampling requirements, it is not practical due to the large number of SUMMA® canisters required.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 4 of 225
2.2 Target Volatile Organic Compound Ozone Precursors
For the purposes of this document, the term VOCs refers to gaseous aliphatic and aromatic
nonmethane organic compounds that have a vapor pressure greater than 0.14 mm Hg at 25 °C, and
generally have a carbon number in the range of C2 through C12. Many of these compounds play a
critical role in the photochemical formation of O3 in the atmosphere. Volatile organic compounds are
emitted from a variety of sources. In urban areas, the dominant source may be automobiles. Table 2-1
presents the target VOCs which could be measured and reported to satisfy the requirements of
40 CFR Part 58, Subpart E. Users should consider these target compounds in developing their
measurement systems and monitoring approach, and initially report and submit results for these
compounds into the Aerometric Information Retrieval System (AIRS) as described in Section 2.6.2 of
this document. The VOCs listed in Table 2-1 were selected primarily based on their abundance in
urban atmospheres and their potential role in the formation of O3. Polar compounds are not included
on the target list due to their surface adsorption characteristics and the difficulty in measuring these
compounds using the methodology designed for nonpolar hydrocarbons. The methodology described
in this document is designed to measure the more abundant non-polar hydrocarbons or VOCs.
The target list in Table 2-1 is not definitive or all-encompassing, but should be used as a
guideline for implementation that should evolve as the monitoring program matures. As experience is
gained in the collection of data regarding the abundance of specific VOCs at each site, target
compounds may be deleted from the list depending on the frequency of occurrence. If additional
compounds are identified and occur at high frequency, they should be added to the list of PAMS target
compounds.
The compounds listed in Table 2-1 are presented in the order of their expected
chromatographic elution from a J&W® DB™-1 non-polar dimethylsiloxane capillary analytical column.
The AIRS parameter code for each compound is also given in Table 2-1. Compounds
-------
Table 2-1. Target Volatile Organic Compounds
Section: 2
Revision: 1
Date: 09/30/98
Page: 5 of 225
AIRS
Parameter
Code
Target
Compound
Name
43203 Ethylene
43206 Acetylene
43202 Ethane
43205 Propylene
43204 Propane
43214 Isobutane
43280 1-Butene
43212 w-Butane
43216 /raws-2-Butene
43217 cH-2-Butene
43221 Isopentane
43224 1-Pentene
43220 w-Pentane
43243 Isoprene (2-methyl-1,3 -butadiene)
43226 /raws-2-Pentene
43227 cw-2-Pentene
43244 2,2-Dimethylbutane
43242 Cyclopentane
43284 2,3-Dimethylbutane
43285 2-Methylpentane
43230 3-Methylpentane
43245 1-Hexene*
43231 w-Hexane
43262 Methylcyclopentane
43247 2,4-Dimethylpentane
45201 Benzene
43248 Cyclohexane
43263 2-Methylhexane
43291 2,3-Dimethylpentane
AIRS
Parameter
Code
43249
43250
43232
43261
43252
45202
43960
43253
43233
45203
45109
45220
45204
43235
45210
45209
45212
45213
45207
45211
45208
43238
45225
45218
45219
43954
43141
43102
43000
Target
Compound
Name
3-Methylhexane
2,2,4-Trimethylpentane (isooctane)
w-Heptane
Methylcyclohexane
2,3,4-Trimethylpentane
Toluene
2-Methylheptane
3 -Methy Iheptane
w-Octane
Ethylbenzene
m/p-Xylene
Styrene
o-Xylene
w-Nonane
Isopropylbenzene (cumene)
w -Propy Ibenzene
w-Ethyltoluene (l-ethyl-3-methyIbenzene)
/>-Ethyltoluene (1 -ethyl-4-methyIbenzene)
1,3,5-Trimethylbenzene
o-Ethyltoluene (1 -ethy 1-2-methyIbenzene)
1,2,4-Trimethy Ibenzene
w-Decane
1,2,3-Trimethylbenzene
»2-Diethy Ibenzene
/>-Diethy Ibenzene
H-Undecane
w-Dodecane*
TNMOC**
PAMHC***
* These compounds have been added as calibration and retention time standards primarily for the purpose of retention
time verification. They can be quantitated at the discretion of the user.
** Total Nonmethane Organic Compounds
* * * PAMS Hydrocarbons
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 6 of 225
with lower boiling points typically elute first on this analytical column, followed by the heavier, higher
molecular weight components with higher boiling points. Concentrations of the target VOCs and
unknown compounds (unidentified peaks) are calculated in units of parts per billion Carbon (ppbC).
The concentration in ppbC for a compound can be divided by the number of carbon atoms for that
compound to estimate the concentration in parts per billion volume (ppbv). The target compound list in
Table 2-2 has also been separated and classified into categories based on structure. The categories
include paraffins (alkanes and cycloalkanes), olefins (alkenes and cycloalkenes), aromatics (arenes),
and alkynes. Because the compound proved to be unstable and decomposed in the calibration gas
cylinder, 2-methyl-l-pentene was replaced on the list of PAMS target volatile organic compounds by
1-hexene. w-Dodecane was added as a late-eluting retention time marker.
2.2.1 Total Nonmethane Organic Compound (TNMOC) and PAMS Hydrocarbons
(PAMHC)
The TNMOC measurement is the unspeciated total concentration of VOCs (typically C2
through C12) in ambient air. This measurement supplements the O3 precursor compound measurements
and is used for O3 models that do not require speciated hydrocarbon measurement input. This estimate
can be made using either the automated or manual techniques described in Sections 2.4 and 2.5,
respectively. An estimate of the TNMOC in ppbC is determined as the sum of all identified and
unidentified gas chromatographic peaks in the C2 through C12 range as eluted from the analytical column
and detected by the flame ionization detector (FID). The concentration in ppbC of TNMOC is
calculated by taking the total area count measured and applying the response factor for propane, the
primary calibration compound. The C2 through C12 retention time window should be established and
periodically verified by analyzing ethylene or acetylene (C2) and dodecane (C12). These compounds
may be incorporated in the retention time or calibration standard.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 7 of 225
Table 2-2. Target VOC Compound Classification
Alkyne
Acetylene
Aromatic
Styrene
m/p-Xylene
o-Xylene
Toluene
Ethylbenzene
w-Propylbenzene
1,2,4-Trimethylbenzene
1,3,5-Trimethylbenzene
1,2,3 -Trimethylbenzene
Benzene
Isopropylbenzene (cumene)
/w-Ethyltoluene (1 -ethyl-3-methylbenzene)
/>-Diethylbenzene
o-EthyItoluene (1 -ethy 1-2-methylbenzene)
/>-Ethyltoluene (1 -ethyl-4-methylbenzene)
/w-Diethylbenzene
Olefin
1-Hexene*
1-Butene
Isoprene (2-methyl-l,3-butadiene)
1-Pentene
trans-2-Butene
c/s-2-Butene
trans-2-Pentene
cis-2-Pentene
Propylene
Ethylene
Paraffin
Isopentane
3-Methylheptane
2-Methylheptane
7?-Octane
2,3,4-Trimethylpentane (isooctane)
Ethane
Propane
Isobutane
7?-Nonane
ft-Butane
2,2,4-Trimethylpentane
7?-Hexane
«-Pentane
3-Methylpentane
2-Methylpentane
Cyclopentane
2,3-Dimethylbutane
Methylcyclopentane
2,4-Dimethylpentane
2,2-Dimethylbutane
^-Heptane
3-Methylhexane
2,3 -Dimethylpentane
Cyclohexane
2-Methylhexane
Methylcyclohexane
«-Decane
«-Undecane
«-Dodecane*
*These compounds have been added as calibration and retention time standards primarily for the purpose
of retention time verification. They can be quantitated at the discretion of the user.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 8 of 225
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 9 of 225
Compendium Method TO-12, preconcentration direct flame ionization detector (PDFID)
techniques described in Section 3.0 of this document and in Appendix C, may also be used to
determine TNMOC. Method TO-12 measures carbon-containing compounds from the sample as
concentrated by cryogenic trapping and thermal desorption directly into a FID. The FID response is
typically calibrated using propane to give a per-carbon response in area counts per ppbC. Compounds
with a carbon number greater than C12 may be transferred and detected using the Method TO-12
technique. Because of inherent differences between the "summation of peaks" and PDFID
approaches, the two approaches do not provide equivalent TNMOC results and are not directly
comparable. Since the vapor pressure of carbon-containing compounds decreases with increasing
molecular weight, compounds with a carbon number above C12 are not expected to contribute
significantly (more than a few percent) to the TNMOC value.
A subgroup of TNMOC, PAMHC is the sum of peak areas for only the PAMS target
compounds. Both TNMOC and PAMHC are valuable data components and the ratio
PAMHC/TNMOC may indicate the conversion of ozone precursors to carbon-containing products
resulting from atmospheric chemistry.
The PAMHC parameter itself is of limited value because the PAMS target list may change by
geographic area. Also, PAMHC provides a broad measure of compounds that is often not
substantially different from TNMOC. PAMHC could be used by a state or agency measuring only
listed compounds, and then calculating the percent of unidentified compounds as:
r> +TT -A ^ A TNMOC - PAMHC * inn
Percent Unidentified * 100
TNMOC
Alternatively, PAMHC can be used to determine the percentage of the total made up by the listed
compounds.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 10 of 225
T> +TJ ^ A PAMHC * 1ft-
Percent Identified ^^=^^ * 100
TNMOC
This ratio for a given PAMS site usually stays within a range characteristic of the site, subject to
seasonal variation.
2.3 Chromatography Discussion and Issues
The following section discusses the basic operating principles of the gas chromatography with
flame ionization detection (GC/FID) methodology used to measure ambient VOCs either as an
independent analytical system or as part of an automated sampling/analytical system. Related
chromatography issues or concerns regarding peak identification and quantitation, sample moisture
removal, calibration, primary and retention time standard preparation and humidification, and analytical
column selection and configuration are also discussed.
2.3.1 Gas Chromatography with Flame Ionization Detection
Gas chromatography with flame ionization detection is the established analytical technique for
monitoring VOCs in ambient air. The sensitivity, stability, dynamic range, and versatility of GC/FID
systems make them extremely effective in measuring very low concentrations of VOCs. The gas
chromatograph may be an independent analytical system or a component of an automated
sampling/analytical system.
Typically, a sample taken from an urban environment contains more than 100 detectable
compounds that may reasonably be separated into quantifiable peaks. These compounds are generally
present at concentrations varying from less than 0.1 ppbC to greater than 500 ppbC with the typical
concentration ranging between 0.1 to 50 ppbC. Detection of typical urban concentration levels
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 11 of 225
generally requires that samples be passed through a preconcentration trap to concentrate the
compounds of interest and separate them from components of the sample that are not of interest
(i.e., air, methane, water vapor, and carbon dioxide).
The GC/FID systems required for VOC measurement consist of the following principal
components:
• Sample introduction;
• Sample conditioning for moisture removal (optional);
• Sample concentration;
• Sample focusing for optimal sample injection and improved chromatographic separation
(optional);
• Gas chromatography; and
• Flame ionization detection.
An air sample may be introduced to the measurement system directly from ambient air, an
integrated canister, or a calibration gas cylinder. The sample is optionally passed through a sample
conditioning system for moisture removal and then concentrated using an adsorbent or glass bead trap
that is cryogenically cooled using liquid nitrogen, liquid carbon dioxide, or thermoelectric closed-cycle
coolers. The concentrated sample is then thermally desorbed and introduced into the carrier gas prior
to being introduced to the analytical column. Sample refocusing is optional and may be performed
using a cryogenically or thermoelectrically cooled secondary trap. Sample refocusing may also occur at
the head of the cryogenically cooled analytical column. Sample focusing is used to concentrate the
desorbed sample into a narrow band for injection onto the capillary GC analytical column. The focused
sample is thermally desorbed rapidly and injected onto the analytical column of the gas chromatograph
as a "plug," which maximizes GC column resolution and results in improved C2 and C3
chromatographic separation and peak shape. Sample focusing is effective when low carrier gas flow
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 12 of 225
rates (1-2 mL/minute) are used. The analytical column separates the sample into individual components
based on the distribution equilibrium between the mobile (carrier gas) and stationary (liquid column
coating) phases. The separated components elute from the column and enter the FID, where a signal is
generated based on carbon response. The time of elution and detection (retention time) is the primary
basis for the identification of each compound. Retention time units are typically expressed in minutes
and are specific to the conditions of the GC system used. The identification of sample components is
determined by matching the known retention times of the components in a retention time standard with
those in the sample. It is desirable to confirm GC peak identification periodically using a mass
spectrometric detector, if available.
The FID is the most widely used, universal GC detector. As a general observation, the FID
provides good sensitivity and uniform response to w-alkanes based on the number of carbon atoms in
the compound. For unsaturated, cyclic, or aromatic hydrocarbons, the FID response is less
predictable. The FID is, therefore, well suited for ambient air analysis since a majority of VOCs in
ambient air are hydrocarbons. This uniformity of FID response to w-alkanes simplifies calibration in
that a single hydrocarbon compound (e.g., propane) can be used to calibrate the detector response for
all hydrocarbons.2'3 This FID response characteristic also provides for the unique capability of
estimating the concentrations of not only the target peaks (identified) but also the unidentified
components of the sample. Some automated GC systems require a two-component calibration mixture
(e.g., propane and benzene) due to the use of dual analytical columns. By summing all identified and
unidentified chromatographic peak areas, a useful estimate of the concentration of TNMOC is
provided. The FID also has a broad linear dynamic range of response, allowing for the analysis of
samples with concentrations ranging from picogram (using preconcentration) to microgram quantities of
hydrocarbons.
Modern GC technology, coupled with sophisticated data acquisition and processing software,
provides for reasonable estimates of both the identity and quantity of the target species to the extent that
the analytical column is capable of separating them and the system has been adequately characterized
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 13 of 225
and calibrated. The retention characteristics of the analytical column must be determined for each
target compound using pure components or mixtures of pure components diluted with a humidified inert
gas.
2.3.2 Identification and Quantification Issues
Although the peak identification and quantification expected with GC/FID systems is
acceptable for meeting the objectives of PAMS, the GC/FID technique has some inherent limitations.
Chromatographic systems using GC/FID rely primarily on the practical use of retention times to make
compound identifications for each chromatographic peak. Commercial GC/FID systems configured for
VOC analyses must be suitably designed to provide stability of system parameters to ensure consistent
retention times for confident peak identification.
Gas chromatographic peak misidentifications typically occur as a result of retention time
shifting and interferences due to co-eluting non-target compounds. Modern GC capillary columns are
generally capable of adequately separating the targeted compounds; however, co-elution of unidentified
species with the targeted species can and does occur. The identification and quantitative uncertainty
resulting from co-elution will depend on the type of unidentified compound and the abundance relative
to the affected target VOC. The target VOCs are exclusively hydrocarbons which are primarily
emitted into the atmosphere by mobile sources and generally dominate most urban samples.
Concentration estimates for substituted hydrocarbon species such as oxygenated or halogenated
hydrocarbons using FID are uncertain since these compounds do not respond to the FID solely on a
per carbon basis. Generally, the identification and quantification of a targeted compound will not be
significantly affected unless a substituted species, at a significant concentration, co-elutes with the target
compound.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 14 of 225
The potential for target compound identification errors can be reduced or eliminated by:
Ensuring that the measurement system is fully optimized and characterized as discussed
in Section 2.3.7, Pre-Measurement Chromatographic System Verification;
Designating chromatographic reference peaks and using relative retention times for
peak identification (Section 2.7.1, Data Validations);
Using dual-column configurations to provide improved resolution (Section 2.3.5,
Column Configuration);
Having an experienced chromatographer conduct visual inspection of the
chromatograms at some practical frequency to verify proper system operation;
Reviewing the chromatographic data using computer-based exploratory software
designed to improve and validate the GC data and determine outliers;
Periodically re-analyzing samples on a different well-characterized GC system to
identify co-eluting compounds; and
Periodically confirming peak identification using more definitive GC/MS techniques.
Quantitative errors can be reduced by careful attention to quality control (calibration details
and system blanks), frequent response checks using canister samples containing target compound
mixtures of known concentration, and periodic performance audits or proficiency studies using
independent reference materials. Analytical system blank analysis of humidified, ultra zero air is
performed to characterize the background concentration of VOCs present in the measurement system.
If unacceptable levels of background system contamination occur the data will be quantitatively
compromised. Sources of contamination can be related to the:
• Source of humidified, ultra zero air;
• Sample to trap transfer line;
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 15 of 225
• Carrier gas and filters; and
• Analytical columns.
The effort devoted to peak identification, confirmation, and quantification is important to the quality of
the collected data. Users must determine the appropriate level of effort to devote to this activity based
on their specific needs and capabilities.
2.3.3 Sample Moisture Issues
The accurate identification and quantitation of trace level VOCs in ambient air generally
require the use of sample concentration techniques for sample enrichment to enhance instrument
sensitivity. The effects of moisture must be considered in any measurement program where sample
concentration is required. Cryogenic concentration techniques are commonly used, especially for light
hydrocarbons. The vast difference in boiling points of the C2 and C12 hydrocarbons also may require
the use of sub-ambient chromatography to adequately separate the entire range of compounds. The
co-collection of moisture in the concentration trap and subsequent injection of water onto the analytical
column can cause a number of problems and adversely affect the overall quality of the data generated.
These problems include:
• Cryogenic trap freezing which results in reduced sample flow or trap blockage;
• Chromatographic column plugging due to ice formation and subsequent retention time
shifting, peak splitting, and poor peak shape and resolution which result in incorrect
peak identification and peak naming;
• Chromatographic column deterioration (especially with A12O3 columns);
• Baseline shifts due to elution of the water profile;
• FID flame extinction;
• Poor reproducibility and precision of the data generated;
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 16 of 225
• Competition for active sites and adverse effects on adsorbent concentration traps; and
• Suppression of the FID signal.
In addition, if "cold spots" exist in the sample concentration or transfer system, water can collect and
cause sample carryover or "ghost" peaks in subsequent sample analyses. This carryover may affect the
data by causing chromatographic interferences which affect the resolution, identification, and
quantitation of the components of interest.
Moisture removal from the sample stream prior to sample concentration minimizes these
problems and also allows larger sample volumes to be concentrated, thus providing greater detection
sensitivity. Moisture related problems can be alleviated by various water management methods that
include Nafion® driers (Perma-Pure® Inc., Toms River, NJ), selected condensation at reduced
temperatures, selective temperature desorption, non-cryogenic hydrophobic adsorbent sample
concentration traps, dry gas purging, and selective multibed sorbent trapping. However, some methods
used to remove moisture from the sample may result in the loss of polar VOCs which affects the
TNMOC measurement. This effect is variable, based on drier efficiency and compound selectivity. A
drier that minimizes both polar VOC loss and the potential for introducing contaminants into the system
should be considered.
Nafion® driers are commonly used for ambient air sample drying, and are discussed in
Compendium Method TO-14.4 The Nafion® membrane consists of a hygroscopic copolymer of
tetrafluoroethylene and a perfluorosulfonic acid that is coaxially mounted within a larger Teflon® or
stainless steel tube. The humid sample stream is passed through the membrane tube, allowing water to
pass through the walls by a process called "perevaporation" into a dry nitrogen (N2) or air purge stream
that is counter-currently flowing through the annular space between the membrane and the outer tube.
Variables that determine the drying efficiency include the surface area of the membrane used, sample
flow rate or sample residence time in the dryer, pressure or vacuum of the sample and purge flow rate,
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 17 of 225
temperature, and sample humidity. Depending upon the variables affecting drying efficiency, Nafion®
drier water removal efficiency ranges of 80-95% have been reported.5'6'7
Nafion® drying devices have shown demonstrated losses of certain polar VOCs (amines,
ketones, alcohols, and some ethers).8'9 This reduction in recovery of polar VOCs significantly affects
TNMOC measurements made using a Nafion® drier. Reduction in recovery of polar VOCs by drying
can reduce the TNMOC measurement by 20-30% in typical ambient air samples. Nafion® driers have
also caused rearrangement of several monoterpenes (a-pinene and a-pinene) but have no effect on the
recovery of isoprene.6 Hydrocarbons, chlorinated or fluorinated hydrocarbons, esters, aldehydes, and
some ethers are unaffected by the drier.8'5'10
Recent information11 discusses issues reported when using Nafion® driers shortly after heating
to regenerate the drier by removing residual water vapor and organic compounds, in order to improve
drier efficiency. Heating can significantly affect the sample integrity of the C4- C6 alkenes and cause
compound losses and rearrangement. The degree of loss and rearrangement is dependent on length of
time and the temperature used for drier regeneration, as well as sample humidity. Isoprene may be lost
without reappearance of an equivalent amount of carbon. In the case of C6 alkenes, new, unidentified
peaks may emerge in the retention time area of the original peaks. Heating had no effect on C2-C3
alkenes, C2-C10 alkanes, cycloalkenes, and aromatics. The effects of heating are reversed if the drier is
immediately purged with clean, dry nitrogen or air at a flow rate of 50 cc/minute for at least three hours.
Heating of Nafion® driers for regeneration should be avoided and is not recommended for PAMS. If
the drier shows a loss of efficiency as determined by recovery of the target compounds in the retention
time standard, the drier should be replaced. To improve efficiency and prevent memory effects, the
drier should be replaced at least seasonally or more frequently as needed. Information on the use of
Nafion® drying devices is presented in EPA Compendium Method TO-144 or EPA Compendium
Method TO-15, entitled Determination of Volatile Organic Compounds (VOCs) in Air Collected-
Prepared Canisters and Analyzed by Gas Chromatography/Mass Spectrometry (GC/MS) (see
Appendix A).
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 18 of 225
Sample drying using selective condensation at reduced temperatures is performed by
selectively condensing moisture at a reduced temperature from the sample stream during thermal
desorption from the sample concentration trap. This method of drying has been evaluated for
recoveries of compounds having a wide range of volatilities and was found to give good recovery,
reproducibility, and acceptable chromatography when operated at 15°C.12
Studies have been done incorporating the use of controlled vaporization of VOCs off glass
bead traps at ambient and reduced temperatures, non-cryogenic hydrophobic adsorbent sample
concentration traps, dry gas concentration trap purging at selected temperatures,13 and dual sorbent
trapping systems to selectively reduce sample moisture.14'15'16 Techniques for drying an ambient air
sample have been combined, including dry purging after collection on solid sorbent, loss of water by
breakthrough when collecting on solid sorbents, and sample splitting.17'18 These novel approaches to
mitigating the effects of moisture should be evaluated to determine any limitations or negative effects
prior to incorporating them into any VOC measurement system. EPA Method TO-15 (see Appendix
A) has recently been added to the EPA Compendium of Methods for the Determination of Toxic
Organics in Ambient Air and describes different techniques for drying ambient air samples. One of the
main goals of this method was to use drying methods that would not affect polar VOCs as drastically as
the Nafion® driers.
2.3.4 Calibration Standards
Calibrating a GC/FID system to measure VOCs requires two distinctly different types of
calibration mixtures: a primary standard to calibrate detector response for gas chromatographic peak
quantitation (primary calibration standard) and a qualitative mixture of known hydrocarbon compounds
to determine gas chromatographic peak retention times (retention time standard).
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 19 of 225
2.3.4.1 Primary Calibration Standard
The GC/FID response is calibrated in ppbC using a propane primary calibration standard
referenced to a National Institute of Standards and Technology (NIST) Standard. A propane and
benzene mixture is recommended for systems that utilize dual columns or column switching
configurations that use two FIDs. Standard Reference Materials (SRMs) from NIST and Certified
Reference Materials (CRM) from specialty gas suppliers are available for this purpose. NIST currently
has a fifteen component ambient non-methane organics in nitrogen SRM available (SRM 1800) for use
as a reference or primary calibration standard. SRM 1800 contains both propane and benzene. Less
expensive working standards needed for calibration verification over the range of expected
concentrations can be prepared by the user or purchased from a gas supplier, provided they are
periodically referenced to a primary SRM or CRM. The primary calibration standards must be
humidified to reflect the ambient air matrix being analyzed. A procedure for preparing humidified
standards is given in Section 2.3.4.3.1. A procedure for diluting standards is given in Section 2.3.4.3.2.
Based on the uniform carbon response of the FID to hydrocarbons, the response factor determined
from the propane or benzene primary calibration standard is used to convert area counts into
concentration units (ppbC) for every peak in the chromatogram.
It is also feasible to incorporate the primary calibration standard into the retention time
standard described below by confirming the concentration of propane and benzene in the retention time
mixture using a primary SRM or CRM.
2.3.4.2 Retention Time Calibration Standard
The retention time calibration standard is a multiple-component mixture containing all target
VOCs at varying concentration levels. The retention time calibration standard is a humidified working
standard used during the initial setup of the GC/FID system to optimize critical peak separation
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 20 of 225
parameters and determine individual retention times for each of the target compounds. The retention
time calibration standard is also used during the routine operation of the GC/FID system as a QC
standard for verifying these retention times.
The response of the GC/FID to selected hydrocarbons in this standard can be used to
monitor system performance and determine when system maintenance or recalibration of the FID using
the primary calibration standard is necessary. Proper operation of the FID according to the
manufacturer's specifications produces a linear response across the chromatographic range. The
concentration of each compound in the retention time standard need not be directly referenced to the
SRM or CRM (as is the case for the primary calibration standard); rather, the concentration of each
compound can be determined with reasonable accuracy using the FID propane or benzene carbon
response factor from the calibrated GC system. If the propane and benzene in the retention time
mixture are used for primary calibration, then both must be directly referenced to an SRM or CRM.
To reference a working standard to an SRM or CRM, the analytical system is calibrated with the SRM
or CRM, then the working standard is analyzed against the SRM or CRM calibration. If necessary, a
correction factor for the working standard is calculated.
A multiple-component high pressure mixture containing the target VOCs can be obtained
from a specialty gas supplier. Multiple-component mixtures can also be prepared by the user to
confirm the peak identifications using the retention time standard. The retention time standard must be
humidified for use as discussed in Section 2.3.4.3.
2.3.4.3 Calibration Standard Preparation
The primary propane and benzene calibration standards must be humidified to ensure integrity
and stability. Water vapor has been shown to improve the stability of low pressure VOC gas mixtures
in SUMMA® canisters.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 21 of 225
A stock multiple-component retention time calibration standard containing the compounds of
interest may be prepared at a concentration level approximately 100 times that of the anticipated
working standard concentration. The stock standard can be prepared by blending gravimetrically
weighed aliquots of neat liquids or by adding aliquots of gaseous standards with an inert diluent gas.
The aliquot of each compound should be introduced through a heated injector assembly into an
evacuated SUMMA® passivated stainless steel canister or other inert container. For the neat liquid
aliquots, the pre-injection and post-injection syringe weights are recorded, and the difference used to
determine the amount of liquid actually transferred to the canister. Following injection of all neat liquid
and gaseous components, the canister is pressurized to at least 2 atmospheres above ambient pressure
with clean, dry N2. Concentrations are calculated based on the amount of compounds and diluent
injected and the final canister pressure, using ideal gas law relationships.
The stock retention time calibration standard is used to prepare humidified retention time
working standards at the ppbC level. It is not necessary to determine exact component concentrations
in the multi-component mixture because the working retention time standard should not be used to
determine compound specific response factors. However, the approximate concentration of the stock
standard must be known in order to prepare the working retention time standards. Preparation of the
working standards is accomplished by syringe injection of a gaseous aliquot of the stock standard into a
SUMMA® passivated stainless steel canister or other inert canister, and subsequently humidifying for
use.
2.3.4.3.1 Procedure for Humidification
The relative humidity of the air in a canister is an important issue with respect to the storage
stability and recovery of VOCs. A study using SUMMA® passivated canisters under varying
pressures, relative humidities (RHs) and different VOC residence times has shown that humidification of
canisters improves the recovery of higher molecular weight, less volatile components.19 The study
showed that RH levels above 18% were required for improved compound recovery. Another study
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 22 of 225
using SUMMA® passivated canisters showed that a relative humidity of at least 15% was necessary to
ensure complete recovery of 41 chlorinated, brominated and aromatic compounds at concentrations of
2 to 4 ppbv.20 There is some evidence that canisters lined with fused silica (SilcoCan™, Restek, Inc.
Bellefonte, PA) do not have a minimum requirement for humidity, as do the SUMMA® polished
canisters.21
The relative humidity of air taken from a humidified canister can vary over a significant
range. For example, as shown in Figures B-l and B-2 of Appendix B (also see Reference 22),22 if
18L of air at 75% RH is sampled, the air subsequently released from the canister will vary from 33%
RH at 30 psig (first sample taken) to 100% RH at 0 psig. This knowledge is important since the
retention times of individual gas chromatographic peaks and the response factors of some types of gas
chromatographic detectors change appreciably with sample humidity. A second concern is the loss of
water-soluble VOCs either to condensed water or to water consolidated in drops on the canister
interior surface. For the example given above, after the canister is filled with 18L of ambient air at 75%
RH, 55% of the water in the fully pressurized canister (30 psig) will be condensed on the canister
interior surface and 45% will be in the gas phase. As sample is removed from the canister the water
adsorbed will be replenished by evaporation of the condensed water. The ratio of H2O in the gas
phase (maintained at the equilibrium vapor pressure by evaporation from the wall) to the amount of air
in the canister will increase and the RH will increase. If the RH of the ambient sample is high enough
(>70% RH) then there will still be condensed water inside the canister even when the canister pressure
is reduced to atmospheric pressure. The reader can gain a better appreciation for the variation in RH
of gas released from a canister by assuming various RH values for ambient air and using Figures B-l
and B-2 in Appendix B.
In general, the amount of water in a given volume of air at a specified RH is calculated by
using the ideal gas law and a table of water vapor pressures (Table 2-3 .)23 The ideal gas law applied
to calculating the amount of water required to humidify 6 liters of air to 100% RH at 21°C and one
atmosphere (zero psig) of pressure is:
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 23 of 225
PV nRT
(2-1)
Where:
PV
RT
n = moles of H2O
V = canister volume, 6 L
P = vapor pressure of H2O, atm
T = temperature in K, 21 ° C + 273 = 294K
R = ideal gas constant, 0.08205 L-atm/K mole
Converting the vapor pressure of H2O in mm at 21 °C to atm:
18.65mm
760mm/atm
0.02454 atm
-------
Section:
Revision:
Date:
Page:
2
1
09/30/98
24 of225
Table 2-3. Vapor Pressure of Water at Various Temperatures, mm Hg
Temp
°C
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
0.0
9.209
9.844
10.518
11.231
11.987
12.788
13.634
14.530
15.477
16.477
17.535
18.650
19.827
21.068
22.377
23.756
25.209
26.739
28.349
30.043
31.824
33.695
35.663
37.729
39.898
42.175
44.563
47.067
49.692
52.442
0.2
9.33
9.976
10.658
11.379
12.144
12.953
13.809
14.715
15.673
16.685
17.753
18.880
20.070
21.234
22.648
24.039
25.509
27.055
28.680
30.392
32.191
34.082
36.068
38.155
40.344
42.644
45.054
47.582
50.231
53.009
0.4
9.458
10.109
10.799
11.528
12.302
13.121
13.987
14.903
15.871
16.894
17.974
19.113
20.316
21.583
22.922
24.326
25.812
27.374
29.015
30.745
32.561
34.471
36.477
38.584
40.796
43.117
45.549
48.102
50.774
53.580
0.6
9.585
10.244
10.941
11.680
12.462
13.290
14.166
15.092
16.071
17.105
18.197
19.349
20.565
21.845
23.198
24.617
26.117
27.696
29.354
31.102
32.934
34.864
36.891
39.018
41.251
43.595
46.050
48.627
51.323
54.156
0.8
9.714
10.380
11.085
11.833
12.624
13.461
14.347
15.284
16.272
17.319
18.422
19.587
20.815
22.110
23.476
24.912
26.426
28.021
29.697
31.461
33.312
35.261
37.308
39.457
41.710
44.078
46.556
49.157
51.879
54.737
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 25 of 225
Substituting values in the above equation:
(0.02454 atm)(6L)
(0.08205 L atm/K mole)(294K)
n 0.00610 moles of H,O required for 100%RH in the canister.
To calculate the moles of H2O required for a given relative humidity multiply the value of n by the RH
expressed as a fraction (e.g., 20% = 0.2) and convert to the number of mg of H2O. Express the
number of mg as an equal number of |iL since 1.0 mg of water occupies 1.0 jiL. Thus:
0.00610moles x 0.2 x 18=12= x 1000==! x LOML 22.0 |oL
mole gm 1.0 mg
The number of jiL to be added for other sample volumes scales linearly with volume, e.g., for a sample
volume of 18L, multiply the number of |iL to be added to 6L by the ratio 18/6 = 3. Hence, the number
of jiL to be added to a canister in order to simulate the sampling of 18 L of sample air at 21°C and
20% RH is 66 jiL. Figure 2-1 can be used to approximate the amount of water in 1 L of air at
temperatures from -30°C to 40°C at 75% and 100% RH. The values in the figure also scale linearly
with sample volume.
Based on the studies of SUMMA®-passivated canisters, low pressure (30 psig) calibration
standards prepared in canisters ideally should have at least a certain minimum amount of water vapor
(>20% relative humidity) to ensure sample integrity but not enough water to cause condensation of
water vapor in the canister (< 33% relative humidity). Using Equation 2-1, the amount of liquid water
that must be added to a 6L canister (pressurized to 18 L with dry air) to achieve these conditions at
21°C (70°F) is between 66 and 110 jiL. This range will of course
-------
60
50
40
30
1 20
100%R.H.
75% R.H.
-30 -20
-10
0
10
20
30
40
Temperature, °C
Tl
Figure 2-1. Water Content of Air at 75% and 100% Relative Humidity Over a Range of Temperatures
Amount of water is expressed as mg/liter. The density of water under standard conditions is 1 g/mL. Thus 1 mg of water occupies a lo 0
volume of 1 uL. For a dry six-liter canister sample at 25°C, approximately (23 x 6) uL of water would be need to be added to o o
achieve 100% relative humidity. Pressure is measured at the exit of the canister.
oo
8 R
to
Figure adapted from Tipler, A. "Water Management in Capillary Gas Chromatographic Air Monitoring Systems." In Proceedings of
the 1994 U.S. EPA/A&WMA International Symposium: Measurement of Toxic and Related Air Pollutants, Research Triangle Park, NC,
1994.
VO
oo
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 27 of 225
vary slightly with the sample temperature (and atmospheric pressure) and should be recalculated for
specific conditions. If excess water is added to the canister, water will condense inside the canister.
However, the presence of condensed water is not observed to have any effect on the recovery of the
non-polar PAMS target compounds and an excess of water vapor has often been used in practice
when only non-polar compounds are of interest.
A detailed procedure for humidifying non-polar canister calibration standards prepared from
dry stock high-pressure cylinder gases is given below. Two simple methods can be used to humidify
calibration gas:
• Direct injection of water into the canister before filling with dry calibration gas; and
• Injection of water into the canister through a stainless steel union tee, then filling with
dry calibration gas.
Both procedures incorporate active temperature controlled heating of the gas transfer line to 90° C.
Heating ensures that the higher molecular weight compounds are transferred quantitatively and not
adsorbed onto the stainless steel tubing during gas transfer. Heat also keeps the water injected through
the stainless steel tee from condensing on the surfaces.
The following materials are needed:
Two-stage, non-corrosive, ultra high purity regulator - the regulator must have stainless
steel diaphragms and inert seats and seals to prevent air diffusion and adsorption of low
concentration trace level gases.
stainless steel tubing and union tee - chromatographic grade stainless steel, fused
silica-lined stainless steel, or nickel are all recommended tubing material choices. A stainless
steel union tee should be used.
High purity water- HPLC or spectrophotometric-grade high purity water.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 28 of 225
Cord heater -110 VAC rated for metal contact, with a 300-watt heat capacity minimum.
Temperature controller- active temperature controller operating on a thermocouple
feedback loop.
Figure 2-2 shows the configuration of the materials for direct injection and Figure 2-3 shows
the configuration of the materials for union tee injection.
Direct Injection of Water into the Canister—To humidify calibration standards by
direct injection, follow these steps:
1) Insert an inert 10-mm septum into the Winch nut on top of the canister valve and hand
tighten to seat the septum.
2) Fill a glass syringe with the desired volume of high purity water for the canister size
used. Open the canister valve slightly while quickly injecting the water, allowing the
vacuum to draw the water into the canister.
3) Close the valve and remove the cap. The canister is now ready to be filled with dry
calibration gas. Any remaining water droplets in the canister valve will be carried into
the canister by the flow of dry calibration gas.
4) Install the correct CGA type high purity regulator onto the calibration gas cylinder.
Install a Winch stainless steel male connector to connect the female NPT thread on the
regulator to the Winch stainless steel tubing. Install a length of Winch stainless steel
tubing to connect the canister to the connector fitting on the regulator.
5) Leak check the entire system by capping the Winch tube outlet and pressurizing the
system to the desired final canister pressure. Close the pressure regulator and monitor
pressure changes. If the pressure drops, check all fitting connections.
6) Loosely attach the canister so that a complete seal is not achieved. With the valve
closed, purge the entire system before use by opening and closing the pressure
regulator at least three times, allowing the excess gas to escape past the incomplete
seal. As an option, a stainless steel toggle shutoff valve may be installed between the
canister and gas transfer tube to vent the purge gas.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 29 of 225
7) Wrap the transfer tubing with a cord heater and plug it into the active temperature
controller. Activate the temperature controller and the system to equilibrate (for 5 to
10 minutes) at a setting of about 90°C.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 30 of 225
Steps 4 through 7
Two stage
high purity
regulator
1/4" stainless steel tubing
1/4" nut
Canister valve
High pressure
cylinder containing
dry calibration gas
Steps 1 through 3
Syringe filled with
high purity water
1/4" nut
10 mm septum
Canister
Figure 2-2. Configuration of Materials to Perform Direct Injections of Water into
the Canister Before Filling with Dry Calibration Gas
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 31 of 225
Heated zone
Two stage
high purity
regulator
\/ /'/ / / / /
Heater
power
(out)
High pressure
cylinder containing
dry calibration gas
Syringe filled with
high purity water
Thermocouple
Selectable
temperature
controller
' Tee with 10mm septum
in 1/4" nut on sideport
Canister valve
Canister
Figure 2-3. Configuration of Materials to Perform Injection of Water Through a
Heated Tee While Filling with Dry Calibration Gas
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 32 of 225
8) Tighten the canister nut and set the delivery pressure gauge on the regulator to the
desired final canister pressure. Fill the canister and allow it to sit overnight for static
equilibration.
Union Tee Injection of Water into the Canister—To humidify calibration standards
in canisters using a stainless steel union tee, follow these steps:
1) Install the correct CGA type high purity regulator onto the calibration gas cylinder.
Install a Winch stainless steel male connector to connect the female NPT thread on the
regulator to the Winch stainless steel tubing.
2) Install a length of Winch stainless steel tubing to connect the canister to the connector
fitting on the regulator. Install a Winch stainless steel union tee at the end of the tubing
and place an inert 10-mm septum in the Winch nut on the side of the tee. Hand tighten
the nut to seat the septum.
3) Leak check the entire system by capping the Winch union tee outlet and pressurizing
the system to the desired final canister pressure. Close the pressure regulator and
monitor pressure changes. If the pressure drops, check all fitting connections.
4) Loosely attach the canister so that a complete seal is not achieved. With the valve
closed, purge the entire system before use three times by opening and closing the
pressure regulator at least three times, allowing the excess gas to escape past the
incomplete seal. As an option, a stainless steel toggle shutoff valve may be installed
between the canister and gas transfer tube to vent the purge gas.
5) Tighten the canister nut and wrap the transfer line and union tee with the cord heater
and plug it into the active temperature controller. Activate the temperature controller
and the system to equilibrate (for 5 to 10 minutes) at a setting of about 90° C.
6) Fill a glass syringe with the desired volume of high purity water for the canister size
used. Open the canister valve slightly, insert the syringe into the septum and quickly
inject the water.
7) Set the delivery pressure gauge on the regulator to the desired final delivery pressure
and open the canister valve completely to allow the gas to fill the canister to the desired
pressure.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 33 of 225
8) When the final pressure is achieved, close the toggle valve, turn off the temperature
controller, and close the regulator. Allow the canister valve to cool before closing.
The valve may become hot due to thermal conductivity. Note: Do not close the
canister valve when it is hot because the Viton® ring will be distorted and the
valve damaged.
9) Allow the canister to sit overnight for static equilibration.
10) Calculate the volume of water to be added using equation 2-1.
2.3.4.3.2 Calibration Standard Dilution Procedure
In order to prepare multiple concentration levels from the primary calibration standard for
system calibration, the calibration gas may be diluted according to the basic procedure. This dilution
procedure involves volumetric dilution based on pressure and is provided here as a simplified, proven
means of accurately preparing diluted calibration standards. Calibration gases may also be diluted by
dynamic flow dilution, or by using commercially available dilution systems.
The primary calibration standard is initially humidified as described in Section 2.3.4.3.1. The
standard is diluted with ultra high purity nitrogen. Stainless steel fittings and chromatographic grade
stainless steel tubing are used for all connecting lines and fittings. The primary calibration gas is
transferred into a canister for dilution. The calibration gas must be humidified as described in Section
2.3.4.3.1. The initial pressure of the canister is measured. The canister is then diluted to the desired
pressure and the final canister pressure is measured. Equilibration and static mixing are allowed to take
place for at least 18 hours prior to analysis. The calculated dilution factor is used to determine the final
concentration value for the calibration standard.
Dilution equipment is commercially available; a dilution apparatus can also be assembled in
the laboratory. The dilution apparatus shown in Figure 2-4 requires the materials described below for
assembly.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 34 of 225
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 35 of 225
Chassis
Pressure
regulator
N2
High
pressure
cylinder
\\\\\\\\\\\\
Canister
belbws\;
valve
1/4" Stainless steel tubing
fsfe Rcw
Source central
shutoff valve
valve
Temperature controlled
M8' stainless steel tubing
\ \ \
Temperature controller power out
Selectable
Temperature
controller
\\\\\\\\
Temperature
controller
power switch
Figure 2-4. Calibration Standard Dilution System
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 36 of 225
Ultra high purity grade nitrogen or air - 99.9999% purity or equivalent, hydrocarbon
free.
Hydrocarbon trap - available from chromatographic supply vendors to remove trace
impurities from high pressure cylinder gases.
'Xt-inch and C -inch stainless steel tubing and union tees - chromatographic grade
stainless steel, fused silica-lined stainless steel, or nickel are all recommended tubing material
choices.
Cord heater -110 VAC rated for metal contact, with a 300-watt heat capacity minimum.
Temperature controller- active temperature controller operating on a thermocouple
feedback loop.
N-, source shutoff valve - a stainless steel bellows assembly designed valve. When the
shutoff valve is closed, the dilution gas cylinder and regulator are isolated during purge
evacuation of the system. When the shutoff valve is opened, the valve is used to apply
dilution gas to the system for controlled introduction.
Flow control valve - a stainless steel micro-metering needle designed valve, used to
introduce dilution gas into the system at a controlled flow rate.
Bellows vacuum isolation valve - a stainless steel bellows assembly designed valve. When
closed, the bellows vacuum valve isolates the high vacuum pump from the system. When
opened, the valve is used to apply vacuum, from the high vacuum pump, to the system.
High precision absolute pressure gauge - a compound pressure gauge used to measure
the pressure in the system and the calibration canister in both positive and negative pressure
modes. The pressure gauge must be able to measure pressure from 40 psig to 5 mm Hg
absolute.
High vacuum pump - An oil-less diaphragm pump used to apply vacuum to the system.
The pump must be able to create vacuum to 5 mm Hg absolute.
Once a dilution system is available, the basic steps for standard dilution are described below:
1) Turn the temperature controller on and allow the system to equilibrate at 100° C.
Open the main dilution gas cylinder valve. Set the delivery pressure using the second
gauge on the pressure regulator to approximately 10 pounds per square inch gauge
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 37 of 225
(psig) pressure over the desired final canister pressure using the pressure control knob
on the regulator.
2) Zero the absolute pressure gauge by adjusting to zero.
3) Connect the calibration gas canister to be diluted to the dilution apparatus.
4) Turn on the high vacuum pump and the active temperature controller.
5) With the canister bellows valve and the N2 source shut-off valve closed, open the
bellows vacuum isolation valve. Allow vacuum throughout the system to stabilize at the
lowest vacuum achievable by the pump to purge all residual gas from the system.
6) Once stabilized, close the bellows vacuum isolation valve. Open the canister bellows
valve and allow the pressure in the system to equilibrate to the initial canister pressure.
7) Measure the initial pressure of the canister from the absolute pressure gauge. Record
the initial canister pressure.
8) Close the flow control valve and open the N2 source shut-off valve. Slowly open the
flow control valve while monitoring the absolute pressure gauge. The slower the
canister is filled, the easier it is to meet the final target pressure.
9) Continue to fill the canister until the final set point is achieved. Allow the absolute
pressure gauge needle to equilibrate before reading the final pressure of the canister.
10) Once the canister has filled to the desired pressure, close the flow control valve, the N2
source shut-off valve, and lastly the canister bellows valve. Turn off the vacuum pump
and the active temperature controller.
11) Disconnect the canister, close the main valve on the dilution gas cylinder.
12) The canister should sit for at least 18 hours before analysis or further dilutions are
performed to allow for static mixing and equilibration.
Calculations—The following calculations are used to determine the target final pressure
(Equation 2-2) and dilution factor (Equation 2-3). The calculations do not account for barometric
pressure and temperature changes, which are expected to be negligible.
-------
+ 14.696)
fa
Section: 2
Revision: 1
Date: 09/30/98
Page: 38 of 225
(2-2)
where:
Pi
cf
14.696
Example:
Final Diluted Absolute Pressure, psia
Initial Concentration, ppbC
Initial Gauge Pressure, psig
Final or Target Diluted Concentration, ppbC
Atmospheric Pressure, psi
To dilute a 30 ppbC calibration standard in a canister with an original pressure of 5 psig
(19.696 psia) to a final concentration of 15 ppbC, what is the target diluted pressure?
Pfa
30ppbC(5psig + 14.696 psi)
15 ppbC
39.39 psia
To convert psia to psig (the measured value), subtract 14.696:
39.39 psia - 14.696 = 24.70 psig
DF =JL
Pfa
(2-3)
where:
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 39 of 225
Pi = Initial Gauge Pressure, psig
Pia = Initial Absolute Pressure, psia = P; + 14.696
Pf = Final Gauge Pressure, psig
Pfi = Pf+ 14.696
DF = Dilution Factor
Pfi = Final Absolute Pressure, psia
Example:
Continuing with the example above, what is the dilution factor for the 30 ppbC standard
which was diluted to a final pressure of 24.70 psig? What is the final concentration of the standard?
Pia P| + 14.696
5 psig + 14.696
19.696 psia
Pfa Pf + 14.696
24.7 psig + 14.696
39.40 psia
DF —12.
Pfa
19.696
39.40
0.499
Standard Concentration C; * DF
0.499 * 30ppbC
14.97 ppbC
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 40 of 225
where:
C; = Initial concentration, ppbC
2.3.5 Column Configurations
The chromatographic column configurations generally used for VOC monitoring programs
incorporate single-column, single-detector, or dual-column, dual-detector applications. The simplest
analytical column configuration involves the use of a single column with a single FID. However, this
configuration imposes limitations on the overall separation of the selected target VOCs. Analyzing the
full range of C2 through C12 target hydrocarbons using a single analytical column may result in less than
optimal separation for either the light or heavy hydrocarbons, depending on the analytical column
chosen. For example, to improve resolution of the C2 through C4 hydrocarbons, a thick liquid-phase
fused silica or Porous Layer Open Tubular (PLOT) column at sub-ambient column oven temperatures
may be desirable. However, PLOT columns generally result in less than optimal resolution of the C5
through C12 hydrocarbons. Likewise, PLOT columns increase retention times of the C10 through C12
hydrocarbons and require longer sample analysis time. If the heavier hydrocarbons are not eluted from
the thick phase or PLOT columns, the TNMOC measurement may be affected, and carryover and
ghost peaks may result.
In order to improve the separation characteristics for the light hydrocarbons (C2 through C4)
as well as the heavier hydrocarbons (C5 through C12), a dual-column, dual-detector configuration
should be considered. In this case, two columns can be judiciously selected to provide optimal
separation of both light and heavy hydrocarbons without sub-ambient column oven temperatures.
Because both columns are generally contained in one gas chromatographic oven for automated
applications, columns must be selected that will provide the desired separation with a single GC oven
temperature program. Dual column systems may be configured with the analytical columns in parallel,
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 41 of 225
operating either concurrently or sequentially. Pre-column and post-column switching valves and the
Deans®24 switch have been used to accommodate these dual-column configurations.
2.3.6 Column Selection
Column selection for analysis of the target VOCs is dictated by the target compound
resolution requirements and other practical and cost considerations, such as the need to minimize
cryogen consumption and total sample analysis time. Selecting columns that will provide the desired
separation of the C2 through C4 hydrocarbons without cooling the column oven to sub-ambient
temperature decreases cryogen consumption significantly.
Several columns suitable for either single- or dual-column applications are discussed below.
The columns described have been used in either a single- or dual-column configuration in conjunction
with a single- or dual-FID for separation of the C2 through C12 hydrocarbons. The column conditions
described are recommendations provided from laboratory applications or conditions determined by the
manufacturer to provide adequate separation of the VOCs of interest. However, these conditions must
be evaluated and optimized to verify acceptable peak resolution prior to use.
The C4 through C12 hydrocarbons may be resolved using a 0.32 or 0.22 millimeter (mm)
inside diameter (ID.), 50 meter (m) long SGE, Incorporated BP1 fused silica column with a
1-micrometer dimethyl polysiloxane coating. This column generally does not provide adequate
separation of the C2 and C3 hydrocarbons even at sub-ambient column oven temperatures. However,
the column can provide adequate separation of the C2 and C3 hydrocarbons if the coating is 3 jam thick
and Electronic Pressure Control is used along with sub-ambient column oven temperatures. Under
these conditions, a single column can be used for all of the target hydrocarbons. Other compatible
columns include the J&W DB™-1, Hewlett-Packard HP-1, Chrompack CP-SIL 5 CB, Restek RTx-
1, and the Supelco SPB-1. The DB™-1 column has been historically and extensively used in ambient
air applications. The SGE BP1 column can be used in conjunction with a 0.32 mm ID., 50 m, Porous
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 42 of 225
Layer Open Tubular (PLOT) fused silica analytical column with a 5-micrometer Hewlett-Packard
A12O3/KC1 or Al2O3/Na2SO4 coating. The Al2O3/Na2SO4 column is slightly more polar than the
A12O3/KC1 and provides optimal resolution of the C4 hydrocarbons. The PLOT column provides
acceptable light hydrocarbon separation under the same column oven temperature program conditions
used for the DB™-1 column but does not provide complete separation and elution of C9 through C12
hydrocarbons. Other compatible columns include the J&W GS-Alumina™ A12O3/KC1 and
Al2O3/Na2SO4. However, alumina PLOT columns from different manufacturers may not be directly
interchangeable and may require some method modification due to the variation in column selectivity.
Because the alumina layer is active, PLOT A12O3 analytical columns are very susceptible to
polar compounds such as water, which causes column deactivation and shifting of peak retention times.
Moisture and other polar compounds must be removed from the sample stream using a membrane drier
or other drying device. If manual sample analysis using a single PLOT A12O3 column is performed,
sequential analyses or the use of separate GC systems may be considered to optimize and obtain
complete C2 through C12 separation and elution. Figures 2-5 and 2-6 are example chromatograms of
retention time calibration standards containing the PAMS target compounds as eluted from the
0.32 mm ID., 50 m, 5 micrometer, Al2O3/Na2SO4 PLOT and 0.22 mm ID., 50 m, 1 micrometer,
SGE, Incorporated BP1 columns. Since these columns have been successfully used by others, users
should give primary consideration to these column types during their column selection process. Figure
2-7 shows a representative ambient air sample analyzed on a PLOT column; Figure 2-8 shows the
same sample analyzed on a BP1 column. Peaks are numbered on the chromatograms, identified peaks
are listed in Table 2-4.
Stationary phase selectivity is neither completely understood nor easily explained. Using a
simplification, selectivity can be considered the ability of the stationary phase to differentiate between
two compounds by virtue of a difference in their chemical and/or physical properties. Stationary phase
and solute factors such as polarizability, solubility, magnitude of dipoles and hydrogen bonding influence
selectivity. In many cases, more than one factor will be
-------
6
1
12
20
10
8'
13
7
Time (minutes)
22
8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46
Peak#
1
Column:
Carrier:
Initial Temp:
Rate 1 :
Temp 2:
Rate 2:
Final Temp:
HPAI2O3/Na2SO4
50m x 0.32mm x 5|xm
Helium, ~2.5ml/min
45°C, 15 minutes
5°C/min
170°C
15°C/min
200° C, 6 minutes
2
4
5
7
8
9
10
11
AIRS Code
43202
43203
43204
43205
43214
43212
43206
43216
43280
43217
43242
Compound Name
Ethane
Ethylene
Propane
Propylene
Isobutane
n-Butane
Acetylene
trans-2-Butene
1-Butene
cis-2-Butene
Cyclopentane
Peak#
12
13
14
15
16
17
18
19
20
21
22
AIRS Code
43221
43220
43226
43224
43227
43244
43284
43263
43230
43243
43246
Compound Name
Isopentane
n-Pentane
trans-2-Pentene
1-Pentene
cis-2-Pentene
2, 2-Dimethylbutane
2, 3-Dimethylbutane
2-Methylpentane
3-Methylpentane
Isoprene
2-Methyl-1-Pentene
Tf
CJQ
B
to
o
o
vo
o
VO
oo
o:s/g/morr/37974)ams/compounl.ppt
Figure 2-5. Example Chromatogram for the PAMS Target Compounds from the PLOT Analytical Column
-------
8
14
23 26
32
3!
, 2
14 16
Column:
Carrier:
Initial Temp:
Rate 1 :
Temp 2:
Rate 2:
Final Temp:
SGE, Inc. BPI,
50m x 0.22mm x Vm
Helium, ~2.5ml/min
45°C, 15 minutes
5°C/min.
170°C
15°C/min.
200°C, 6 minutes
o: s/g/morr/37974)ams/compoun2.ppt
6 •
6
12
i ',
> ' * 10.,., 13 1*
' Q 11 f-
jy] ° '
f [ ,f I
• b ' i
I 1-
''f: , i
-; 1' ' ! ! 'f •
, " '. : ' > ' ',
< ; H , « i
'i
1 '* ' ' '' 8 "
• ''\ .' ' ' ' i
- I-, .rrn-TT-i.oiii , ip'~r' r
£-1
. i ' i ' < " : , -I. ,
jL_..v ., ^.J1'-' .'—:• Jl1' '
Time (minutes)
•"IT. | i- ,1rn ".inr-,-Tn •"-.,«, ,, 1|ir.p_rrTT,,F
20 22 24 26 28 30 32 34 36 38 40 42 44
Peak # AIRS Code Compound Name peak # AIRS Code
1 43231
2 N/A
3 43262
4 43247
5 45201
1p-m 6 43248
1 in
lln 7 43263
8 43291
9 43249
10 43250
1 1 43232
12 43261
13 43252
14 45202
15 43960
16 43253
17 43233
Hexane 18 45203
Unknown 19 45109
Methylcyclopentane 20 45220
2, 4-Dimethylpentane 21 45204
Benzene 22 43235
Cyclohexane 23 45210
2-Methylhexane 24 45209
2, 3-Dimethylpentane 25 45212
3-Methylhexane 26 45213
2, 2, 4-Trimethylpentane 27 45207
n-Heptane 28 4521 1
Methylcyclohexane 29 45208
2, 3, 4-Trimethylpentane 30 43238
Toluene 31 45225
2-Methylheptane 32 45218
3-Methylheptane 33 45219
n-Octane 34 43954
46
Compound Name
Ethyl benzene
m/p-Xylene
Styrene
O-Xylene
n-Nonane
Isopropyl benzene
n-Propylbenzene
m-Ethyltoluene
p-Ethyltoluene
1, 3, 5-Trimethylbenzene
o-Ethyltoluene
1, 2, 4-Thmethylbenzene
n - Decane
1, 2, 3-Trimethylbenzene
m-Diethylbenzene
p-Diethylbenzene
n-Undecane
l§
to
O
o
VO
oo
Figure 2-6. Example Chromatogram for the PAMS Target Compounds from the BP1 Analytical Column
-------
I
O3 Oi
I I
12
;3
l\ ,
1 ^ "~*t 1"'*"^"^
o:
0-
1 I
I
IT- CD or-- cNPO'S-^J -r- f*} Op otq cq-tfcn <*2 c\pc>
r* IN ^ ^ COTO &O IT- -TNT ^r 1*7 ^CX& *>S2:2 £~? !*M..
fsj (TN fN*TM UM^SI OIPl -rn c*"> 1XJ CO €"rSv> C^tfifiJ. ^ "q"Tj"
13
7
6
14
j 20
ft 11
19 21 22
; A^ '
»*u
1
i-
10 12
1 . ' 1 ' >-^
•|— -vv^"l"T"T^~f"1'l"^" *"""T"''l" i !
X I
r
; 15 16 17 1
I
^ dO O CM32 N ^*- Q- hi i £M r^jrt ?A «
i2z< H?O
-------
^| T~
p^ ^
'13
-I
10
12'
18
IT
.3
16
14
15 /'
^rJA
v—I'h
20
1?
\
11
25
22 23
• . i
2g
111 32 33
S [/ '
III
ID OrJ
O "*C0
SD iii ffiy
•I
nE to 2
ucj r~J e*j
C^P- ™ rt
CN ^
IJ) CD
CM i-
PM ~: •'•' >4 it 2?; if.
]-:.- -."«• \
TT1 i I | i i I IJTj !
Tf
S
to
vo
oo
Figure 2-8. Representative Ambient Air Sample (same as Figure 2-6) Analyzed on a BP-1 Column
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 47 of 225
Table 2-4. Peak Identifications, Ambient Air Sample
PLOT Column
Peak Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Peak Identification
ethane
unidentified
ethene
propane
propylene
isobutane
n -butane
acetylene
trans -2-butQne
1-butene
unidentified
c/s-2-butene
cyclopentane
isopentane
trans -2-pentQnQ
1-pentene
cw-2-pentene
2,2-dimethylbutane
2,3 -dimethylbutane
2-methylpentane
3-methylpentane
isoprene
BP1 Column
Peak Number
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
Peak Identification
hexane
methylcyclopentane
2,4-dimethylpentane
benzene
cyclohexane
2-methylhexane
2,3 -dimethy Ipentane
3-methylhexane
2,2,4-trimethylpentane
n -heptane
methylcyclohexane
2,3,4-trimethylpentane
toluene
2-methylheptane
3-methylheptane
n -octane
ethylbenzene
m/p-xylene
styrene
o-xylene
7?-nonane
isopropylbenzene
w-propylbenzene
/w-ethyltoluene
/?-ethyltoluene
1,3,5-trimethylbenzene
o-ethyltoluene
1 ,2,4-trimethylbenzene
7?-decane
1 ,2,3-trimethylbenzene
/w-diethylbenzene
/>-diethylbenzene
«-undecane
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 48 of 225
significant, so there will be multiple selectivity influences. Unfortunately, information about most
compound characteristics, such as the strength of hydrogen bonding or dipoles, is not readily available
or easy to determine. This lack of physical data makes it difficult to accurately predict and explain the
separation obtained for a particular column and set of compounds. Some generalizations, however, can
be made. The DB1 nonpolar phase (dimethyl polysiloxane) is the most nonpolar siloxane stationary
phase available. In most cases, compounds will elute from this column primarily in order of increasing
boiling point. However, both vapor pressure and solubility in the stationary phase influence the exact
elution order.
PLOT chromatography is accomplished through the gas/solid adsorption interactions between
the solutes and the solid adsorbent coated on the column tubing wall. The aluminum oxide (A12O3)
surface is deactivated using KC1 or Na2SO4. Stationary phase polarity is based on the relative
retention of saturated and unsaturated hydrocarbons. The more polar column will result in unsaturated
compounds being more retained relative to the saturated hydrocarbons. The Na2SO4 deactivation of
the A12O3 results in a slightly more polar column than the KC1 deactivation.
There are some alternative columns that can be used to separate C2 through C12
hydrocarbons for both single- or dual-column approaches. The column selection process should be
based on the capability of the column to separate the VOCs listed in Table 2-1 in conjunction with
desired overall sample analysis time and cryogen use. The manufacturer-recommended conditions and
carrier gas flow rates should be evaluated and optimized to verify acceptable peak resolution prior to
use. When choosing alternate columns, the user should consult directly with the analytical column
manufacturer for advice regarding column characteristics, optimum gas chromatographic oven
temperature programs, carrier gas flow rates, and other operational considerations.
The following columns are alternatives for single-column, C2 through C4 hydrocarbon
separation and may require sub-ambient temperature conditions to achieve adequate separation:
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 49 of 225
1. J&W DB™-1 with a 5-micron dimethyl siloxane phase thickness, an internal diameter
of 0.32 mm, and a length of 60 m. The recommended oven temperature program is -
60 ° C for 2 minutes then to 180 ° C at 8 ° C per minute. The final oven temperature is
maintained for 13 minutes for a total analytical run time of 45 minutes.
2. J&W GS-Q® fused silica PLOT capillary column with an internal diameter of 0.53 mm
and a length of 30 m. The recommended oven temperature program is 40 °C for
4 minutes to 200°C at 10°C per minute. The final oven temperature is maintained for
5 minutes for a total analytical run time of 25 minutes. The GS-Q® column is not
affected by water.
The following columns are alternatives for single-column, C5 through C12 hydrocarbon
separation and may require sub-ambient oven temperature conditions to achieve adequate separation:
1. Restek® RTx-502.2 capillary fused silica column with a 3-micron phase thickness, an
internal diameter of 0.53 mm, and a length of 105 m. The recommended GC oven
temperature program is 35°C for 10 minutes to 200°C at 4°C per minute. The final
oven temperature is maintained for 7 minutes, which results in a total analytical run time
of 58 minutes. This column is capable of separating the C4 through C12 hydrocarbons
without the need for sub-ambient column oven temperatures.
2. J&W DB™-624 capillary fused silica column with a 3-micron stationary phase
thickness, an internal diameter of 0.53 mm, and a length of 75 m. The recommended
oven temperature program is 35 °C for 8 minutes to 200°C at 10°C per minute. The
final temperature of 200° C is maintained for 3 minutes, which results in a total analytical
run time of 27.5 minutes.
3. Restek® RTX-1 capillary fused silica column with a 3-micron dimethylsiloxane phase
thickness and an internal diameter of 0.32 mm and a length of 60 m. The
recommended oven temperature program is -25 °C for 4 minutes then to 175° C at
4 ° C per minute, then to 220° C at 22 ° C per minute. The final oven temperature is
maintained for 5 minutes for a total run time of about 60 minutes. The Electronic
Pressure Control program is 18.3 psi for 5 minutes then to 37.5 psi at 0.35 psi/minute.
Total program time is about 61 minutes.
A combination of these light and heavy hydrocarbon separation columns may be used to
accommodate dual-column approaches.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 50 of 225
2.3.7 Pre-measurement Chromatographic System Verification
Prior to making speciated VOC measurements using an automated GC system, the level of
system operation must be thoroughly documented. Information collected during this process is
important in characterizing the system operation and establishing a baseline for performance. The
information from the pre-measurement system verification is used to determine system specific target
analyte retention times, relative retention times, identification of co-eluting compounds and matrix
effects, internal standard retention times, interferences, and detection limits.
2.3.7.1 Retention Times and Relative Retention Times
The rigorous sampling frequency requirements and large data sets associated with PAMS
require the use of an automated GC system with FID, and presume the commercial availability of such
systems. These systems must rely on the practical use of retention times and relative retention times for
qualitative peak identification. Commercial GC/FID systems are designed to provide stable system
parameters that ensure adequate peak identification based on the use of retention times.
Retention time is the time at which the component elutes from the analytical column and
reaches the detection device. The retention of a compound will be determined by its distribution
equilibrium between the stationary and mobile phases, i.e., the distribution ratio. Retention time units
are typically expressed in minutes and this time is specific to the conditions of the GC system used.
When dealing with complex target analyte lists, as in the case of PAMS measurements,
preparing multiple retention time standards that contain 10-15 target analytes that are of known
retention order and well separated by retention time will simplify peak identification and retention time
assignment. These standards must be analyzed to determine specific retention times for the target
compounds and resolve chromatographic issues relative to the instrument conditions, analytical
column(s), and chromatographic conditions used. Retention time is widely applied in chromatography
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 51 of 225
and based on the information gathered from standards. When the retention times for a GC system are
verified, it is important for the system to be operated for a period to allow equilibration and retention
time stabilization to occur. Several standards should be analyzed over a period of days to assess
retention time variability and system stability. The retention time variability is used to establish retention
time windows for each component. It is very important that standards be prepared in humidified air, at
a relative humidity similar to the samples being analyzed.
The identification of sample components is determined by matching the retention times of the
components in the standard with those in the sample. This procedure provides the chromatographer
with a certain degree of confidence that the correct peak has been accurately identified. Peak
identification by retention time is adequate for the PAMS network requirements. A compound's
retention time is characteristic, though not unique. It is, therefore, possible for other compounds to
have the same retention time. The presence of co-eluting compounds or missed peak identifications
cannot be completely excluded. Periodic confirmation of peak identification and quantification using
more definitive techniques, such as GC/MS, is encouraged.
Retention times are typically stable and reproducible, but they are subject to system
variability. To account for any retention time variations, relative retention time (RRT) can be used to
aid in assigning peak identifications. Many commercial GC systems incorporate the use of relative
retention times for peak identification in their data acquisition and processing software. On most
commercial GC systems, the use of RRT for peak identification is easy to implement. An adjusted or
relative retention time can be determined by using both reference or internal standard peaks. Reference
peaks are those components of the sample that are typically present in the sample matrix (reference
peaks of opportunity). Internal standard peaks are components subsequently added to the sample that
are uncommon to the sample matrix. The relative retention time of a target compound (a), as compared
with a reference compound (b), may be calculated as:
-------
RRT -
RTb
Section: 2
Revision: 1
Date: 09/30/98
Page: 52 of 225
RT
(2-4)
where:
RTa = retention time of the target compound
RTb = retention time of the reference peak
The relative retention time of a compound determined in this manner will vary with temperature and the
analytical column stationary phase, but should otherwise be independent of other analytical conditions.
The relative retention time method of peak identification works well when the target compound elutes
relatively close to the reference peak used and retention time shifting is linear. The use of reference
peaks in several retention time windows is only recommended to compensate for retention time shifting
that is not linear. The use of too many reference peaks may actually compromise the ability of the data
system to adequately identify the target peaks consistently.
A retention time reference peak should be chosen that;
• Is always or typically present in the sample matrix;
• Is in the same general retention time area or carbon number range of the
chromatogram;
• Shows chromatographic behavior similar to target components (sharp peak shape); and
• Is well separated from other components in the sample matrix.
Suggested retention time reference peaks include propane, toluene, benzene, and butane, or other
compounds appropriate to the individual PAMS site.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 53 of 225
2.3.7.2 Internal Standards
When GC analysis is performed on a continuous basis at an often unattended or remote site,
fluctuations in ambient temperature and other factors can cause variations in instrument performance
and chromatographic retention times. Changes in ambient conditions can cause small changes or
variations in carrier gas flow rate, column temperature, detector response, sample injection volumes,
and sample moisture content. Use of internal standards can help to minimize the influence of GC
system variability. Internal standards are often also used as reference peaks for determining relative
retention times.
The internal standard should be added to the cryofocusing or adsorbent sample collection
system, concurrent with sample collection, to minimize the effects of the sample matrix. The chief
difficulty in using internal standards for VOC analysis lies in finding an internal standard that does not
interfere with the sample constituents. Characteristics that must be considered when choosing a suitable
internal standard include:
• Components that are uncommon in ambient air;
• Ease and reproducibility in handling and introducing into the GC system;
• Similar in chemical and physical properties to those compounds being analyzed;
• Moderate volatility and low vapor pressure comparable to the expected retention times
and concentrations of the sample hydrocarbons;
• Does not interfere with the measurement method;
• Complete resolution from all other components present in the sample;
• Stable under the conditions and method used; and
• Does not react with components of the measurement system.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 54 of 225
Perfluorotoluene (PFT) is a compound that meets these characteristics and has been used as an internal
standard for air monitoring programs.
Separation of the internal standard compound from other compounds normally found in the
sample must be accomplished using the measurement system and methods implemented by the user to
accomplish sample analyses. Typical ambient air samples are very complex and contain numerous
components. Verification of the internal standard performance and retention time characteristics using
the GC system chosen must be determined using actual ambient air samples. A suitable internal
standard can be analyzed concurrently with the sample to adjust for variations in retention time and
detector response.
2.3.7.3 Identification of Co-Eluting Compounds and Matrix Effects
Another important part of pre-measurement chromatographic system verification is the
determination and effect of possible co-eluting compounds and other sample matrix effects on the ability
to find reference peaks, make peak identification, and ultimately to quantitate target analytes.
Blank samples that contain humidified zero air should be analyzed to establish the GC system
background and determine the level of contamination or artifacts. Blank or zero air samples should not
contain the target VOCs at a concentration greater than the detection limit. Any significant levels of
contamination or artifacts that interfere with the retention times of target analytes must be addressed or
documented prior to sample analysis. Information from the analysis of standards containing target
analytes can then be used to determine where co-eluting compounds may occur. Co-elution issues can
be resolved by optimizing the chromatographic conditions of the system, such as carrier gas linear
velocity and column oven temperature.
Further information regarding co-eluting compounds in samples not identified by zero air
analyses may be obtained using GC/MS. When used under similar conditions (column type,
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 55 of 225
temperature program, etc.), the GC/MS provides valuable information to aid the user in confirming
peak identification and determining the presence of co-eluting compounds and other unknowns. When
GC/MS is used for confirmation, it is important to ensure that the system sensitivity or detection limits
are equivalent to the O3 precursor GC/FID system being used.
2.3.7.4 Detection Limits
The development of methods to measure trace levels of organic compounds in ambient air
and the need for the ability to measure extremely low concentration levels for risk assessment purposes
requires that the analytical system detection limits for the target compounds be established for the
analytical system used. The analytical detection limit must meet the measurement quality objectives
given in Section 2.8. The detection limit is one of the most important performance characteristics of an
analytical system. The GC system detection limit should not be determined until a complete, specific,
and well defined analytical method has been developed. All sample processing steps used in the
analytical method must be included in the experimental determination of the detection limit. Refer to
Section 2.8 for guidance on the approach to establishing VOC detection limits for PAMS. If the
analytical method detection limit does not meet the quality objectives, the sensitivity of the GC system
and methodology used may not be adequate and should be re-evaluated and improved prior to use for
O3 precursor monitoring programs.
2.4 Automated Method for Collecting and Analyzing Volatile Organic
Compound Ozone Precursor Samples
The minimum monitoring network requirements for enhanced O3 monitoring are described in
Section 4.4 of 40 CFR Part 58, Subpart E, Appendix D, and are also discussed in Section 2.1 of this
document. The rigorous sampling frequency requirements of enhanced O3 monitoring (e.g., eight 3-
hour samples every day during the monitoring period) makes automated GC methodology a viable,
cost-effective approach for obtaining VOC measurements at all sites within a network. An automated
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 56 of 225
GC system offers an additional advantage in its inherent capability to provide short-term (e.g., 1-hour)
measurements on a continuous basis for long time intervals.
The following description of automated methodology is based on currently available
commercial automated GC systems and is described in general terms. The intent is to provide guidance
on the configuration and operation of automated GC systems, not to serve as a Standard Operating
Procedure (SOP). Alternative approaches using custom fabricated automated systems are acceptable.
This guidance should be used to define equipment specifications and prepare system specific SOPs
consistent with the 40 CFR Part 58 enhanced O3 monitoring requirements. The users must recognize
that they are responsible for optimization and characterization of the critical parameters for their specific
GC system (consistent with the manufacturers' instructions, if applicable).
The GC system must be capable of automated sample collection, analysis, and data
acquisition on site and must be housed in a temperature-controlled shelter. The primary components of
an automated GC are a sample introduction system, sample conditioning system (for moisture removal),
sample concentration system (for sample enrichment), cryofocusing trap (as an option for improving
peak shape and resolution), gas chromatograph with FID(s), and a data acquisition and processing
system. Commercially available systems incorporate many variations of the primary components of an
automated GC system.
The purpose of Section 2.4 is to describe the sample collection, sample analysis, system
operation, system calibration, and system specifications for an automated GC system. The sample
probe and manifold, sample introduction, sample conditioning, and sample concentration systems are
discussed in Sample Collection, Section 2.4.1. Sample cryofocusing, gas chromatography, and data
acquisition and processing are discussed in Sample Analysis, Section 2.4.2. This guidance should be
used to define automated GC specifications for procurement and to develop and implement a network
monitoring program consistent with the 40 CFR Part 58 enhanced O3 monitoring requirements.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 57 of 225
2.4.1 Sample Collection
Samples collected for automated analysis should represent a time-integrated average for the
required sampling period. In the case where an integrating canister is used to collect the sample, the
canister should be filled at a constant flow rate over the full integration period minus the time required to
transfer a sample to the primary trap and purge and evacuate the canister. In the case where the
sample is collected directly onto the primary concentration trap, the sample should be collected at a
constant flow rate for the full integration period minus the time required to desorb the sample onto a
secondary trap or onto the analytical column and perform system operations to accommodate the next
sample collection. The minimal sample integration time required to constitute a 1-hour sample is 40
minutes. Additional provisions must be made to meet the 24-hour sample requirement. A manual
approach to 24-hour sample collection and analysis is discussed in Section 2.5.
The O3 precursor compounds are collected from a sample manifold with a probe and
introduced into the automated GC system. Water may be removed from the sample stream as
discussed in Section 2.3.3 and then the VOCs concentrated onto a primary sample collection trap.
The concentrated sample is thermally desorbed onto a secondary cryofocusing trap (optional) or onto
the head of the cooled GC column to focus the desorbed sample into a small volume or "plug." The
sample volume is then desorbed for analysis by the GC/FID system.
2.4.1.1 Sample Probe and Manifold
A sample probe and manifold assembly should be used to provide a representative air sample
for collection and subsequent analysis. Sample probe and manifold assemblies are commercially
available or can be custom fabricated. Examples of typical sample probe and manifold assemblies are
presented below. If automated calibration techniques that periodically flood the manifold with
calibration standards are to be applied for the criteria pollutants, a separate manifold would be required
to support the VOC and carbonyl components of the PAMS program.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 58 of 225
The sample probe is constructed of glass that is approximately 1 inch in outside diameter
(O.D.). The inlet of the sample probe is configured with an inverted funnel, approximately 4 inches
O.D. The sample manifold is constructed of glass, approximately 1 and /^ inches O.D. The manifold
has ports used for sample distribution. The number of ports located on the manifold must be equal to
or greater than the total number of monitoring systems to which sample will be delivered. To reduce
the potential for bias, the port nearest to the inlet of the manifold should be reserved for VOC sampling.
Teflon® bushings are used to connect sample lines to the manifold. Because the manifold and
ports are constructed of glass, care must be taken to not place excessive stress on the assembly to
avoid breakage. For VOC sampling, the sample lines should be constructed of 1/8 inch O.D. stainless
steel tubing. The 1/8 inch tubing is flexible and will accommodate the flow rates typically associated
with VOC sample collection. The sample lines should be kept as short as possible to reduce sample
transfer time.
A blower and bleed adapter are located at the exit end of the sample manifold. The blower is
used to pull sample air through the probe and manifold and the bleed adapter is used to control the rate
at which the sample air is pulled through the manifold. An excess of sample air is pulled through the
sample probe and manifold to prevent back diffusion of room air into the manifold and to ensure that
the sample air is representative of outside ambient air. Sample air flow through the sample probe and
manifold should be at least two times greater than the total air flow being removed for collection and
analysis by all systems on the manifold.
The vertical placement of the sample probe and inlet funnel should be at a height of 3 to 15
meters above ground level. Because the O3 monitoring requirements involve multiple-pollutant
measurements, this range serves as a practical compromise for probe position. In addition, the probe
inlet should be positioned more than 1 meter, both vertically and horizontally, away from the housing
structure. The probe inlet should be positioned away from nearby obstructions such as a forest canopy
or building. The vertical distance between the probe inlet and any obstacle should be a least two times
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 59 of 225
the height difference between the obstacle and the probe inlet. Unrestricted air flow across the probe
inlet should occur within an arc of at least 270 degrees. The predominant and second most
predominant wind direction must be included in this arc. If the probe inlet is positioned on the side of a
building, a 180 degree clearance is required. More specific details of probe positioning are presented
in the "PAMS Implementation Manual."25 The glass probe should be reinforced or supported along the
straight vertical axis of the assembly. Typically this support is provided by routing the probe shaft
through a rigid section of metal or plastic tubing that is secured to the housing structure.
The manifold can be positioned in either a horizontal or vertical configuration. Figure 2-9
presents the manifold assembly in the vertical configuration. Figure 2-10 presents the manifold
assembly in the horizontal configuration. If the horizontal configuration is used, the sample ports must
point upward so that material that may be present in the manifold will not be transferred into the sample
lines.
With continuous use the sample probe and manifold can accumulate deposits of particulate
material and other potential contaminants. The sample probe and manifold should be cleaned to remove
these materials. The recommended frequency for cleaning is quarterly. To clean the assembly,
disconnect the sample lines and blower from the manifold. The sample lines and blower are not
cleaned. For safety, electric power to the blower should be terminated until the cleaning process is
completed. Disassemble the individual components by disconnecting the probe, manifold, collection
bottle, and coupling devices from each other. The individual components should then be cleaned using
heated high purity distilled water and a long handled bottle brush. The components should then be
rinsed with the distilled water and allowed to dry completely before reassembling. If required, mild
glass cleaner or detergent can be used to clean particularly dirty components. However, care should
be taken to select cleaners and detergents that are advertised to have low organic compound content
and the number of rinses performed should be increased to ensure that all associated residues are
removed.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 60 of 225
2.4.1.2 Sample Introduction
The air sample can be introduced to the automated GC system directly from the air sample
manifold using a mass flow controller or other flow control device at a constant flow rate over the
prescribed sample integration time. As an alternative, the air sample may be collected into an
integrating canister at a constant flow rate over the prescribed sample integration time,
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 61 of 225
Sampling cane
Cane
support
Shelter roof
Shelter
wall
First port dedicated
to VOC sampling
Sample manifold
(with sufficient number of
ports to individually support
all monitoring conducted)
Bleed adapter
(flow control)
Blower
and mount
Collection
bottle
Exhaust
tube
Shelter
wall
To
atmosphere
Figure 2-9. Vertical Configuration
-------
Sampling cane
Cane
support
Shelter roof
Shelter
wall
First port dedicated
to VOC sampling
Bleed adapter
(flow control)
JMULJLJL
Sample manifold
(with suflident number of
ports to individually support
all monitoring conducted)
Blower
and mount
Exhaust
tube
Collection
bottle
\
Shelter
wall
To
atmosphere
oo
s §
ON O ^ to
to \o
Figure 2-10. Horizontal Configuration
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 63 of 225
and then supplied to the sample concentration trap at the end of the integrating period. For purposes of
calibration and proficiency studies, and to meet the 24-hour sampling requirements, samples may also
be introduced directly from pressurized SUMMA® canisters.
2.4.1.3 Sample Conditioning
Moisture is removed from the sample stream for automated GC analysis to prevent or reduce
the detrimental effects of moisture on the primary concentration trap, analytical column(s), and
detector(s) as described in Section 2.3.3. Moisture removal also allows for analysis of larger sample
volumes, which provides lower detection limits, and is crucial to the measurement of very low
concentration VOCs.
Some commercially available automated GC systems incorporate the use of Nafion®
membrane sample drying devices. New developments in moisture removal include controlled
temperature vaporization, selective temperature condensation, hydrophobic concentration traps, and
micro-scale purge-and-trap. The loss of polar VOCs may result from moisture removal using some of
these techniques and this loss of polar VOCs may significantly affect the TNMOC measurement. The
user must characterize the effects of their particular sample conditioning method on the TNMOC
measurement and target VOCs of interest.
2.4.1.4 Sample Concentration
Ambient air samples are primarily concentrated using multi-bed sorbent or
cryogenically-cooled deactivated glass bead traps. Sampling time and flow rate are typically used to
determine the total volume concentrated onto the primary trap. Multi-bed sorbent traps (Carbotrap®
and Carbosieve®) or cryogenically cooled glass bead traps are required to efficiently collect the
complete range (C2 through C12) of VOCs for O3 precursor monitoring.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 64 of 225
Samples are collected onto sorbent traps at ambient temperature or the traps are cooled
using liquid cryogen (H2, CO2, AT) or Peltier® electronic cooling devices to improve collection
efficiency. Ideally, sorbent traps selectively adsorb only the trace VOCs and do not interact with the
atmospheric constituents (i.e., CO2) or introduce any contaminants into the system. Sorbent traps may
also be designed to eliminate water vapor by using hydrophobic sorbent materials.
Sample concentration using glass bead traps requires a trapping temperature of-185°C.
Trapping at a temperature above -185°C will result in the loss of early-eluting C2 compounds such as
acetylene, ethane, and ethylene. Trapping at a temperature below -185°C can result in the collection
of methane and oxygen, and can have an adverse effect on chromatography. These traps are typically
cooled using liquid cryogen (N2 or Ar). This cooling process is commonly known as cryogenic
concentration or cryotrapping, and is the oldest and best known of the techniques for collecting C2
through C12 VOCs. The glass beads provide surface area for collection of the VOCs at the cryogenic
trapping temperature.
2.4.2 Sample Analysis
Following sample collection and concentration, the sample is thermally desorbed directly onto
the analytical column(s). The analytical column may be cryogenically cooled to aid in focusing the
desorbed sample into a narrow band prior to chromatographic separation. The analytical column
chromatographically separates the sample into components for subsequent detection by the FID. The
signal from the FID is then acquired and processed using a PC-based data acquisition and processing
system.
2.4.2.1 Sample Focusing or Cryofocusing
The sample cryofocusing step is optional and may not be employed in all commercially
available automated GC systems. The secondary cryofocusing trap is used to focus the desorbed
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 65 of 225
sample from the concentration trap into a "plug" for injection onto the analytical column. Cryofocusing
improves the peak separation and in particular the resolution of C2 and C3 hydrocarbons. This
technique is especially helpful when the sample is desorbed from the concentration trap at low flow
rates.
Cryofocusing traps incorporate the use of fused silica tubing that is cooled using liquid
cryogen. The fused silica tubing is wide-bore (0.32 mm ID.) or megabore (0.53-mm ID.) deactivated
fused silica tubing that is cooled to approximately -185°C. Cryofocusing traps may be packed to
increase the surface area and improve the focusing of the sample band.
2.4.2.2 Gas Chromatography
The gas chromatograph contains the analytical column(s) of choice for PAMS VOC analysis.
Refer to Sections 2.3.5 and 2.3.6 for guidance on column configuration and selection. Commercially
available GC systems are typically configured with the appropriate analytical column(s) to separate the
VOCs of interest. However, the user must determine if the system meets the enhanced O3 monitoring
requirements and specifications (Section 2.4.4) prior to procurement. The user must also characterize
the performance of the system operation prior to use. Commercial GC systems may incorporate the
use of single or dual-column configurations (in series or parallel) that may require sub-ambient oven
temperature programs. It is important to note that systems that eliminate the need for sub-ambient
column oven temperatures reduce the overall cryogen consumption of the system. New developments
in carrier gas electronic pressure programming and control have greatly improved peak resolution and
retention time stability for some automated GC systems.
Automated GC systems employ the use of a PC-based data acquisition and processing
system for peak integration and quantitation. Data acquisition and processing systems are comprised of
hardware and software that perform data acquisition, peak detection and integration, peak identification
by retention time, post-run calculations and quantitation, calibration, peak reintegration, user program
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 66 of 225
interfacing, and hard copy output. Data are automatically stored on magnetic media (e.g., hard disk or
floppy diskette).
The GC data acquisition and processing software is developed and supplied by the GC
manufacturer and should contain the necessary algorithms to acquire, integrate, and identify the
chromatographic peaks by retention time. The system should be capable of producing an electronic
and hard copy report file that contains the information needed to identify the sample and a listing of all
peaks detected in the chromatogram. This listing should contain the peak name if it is a target
compound. All detected peaks (both target and unidentified) should be reported with a concentration,
in ppbC, and a retention time. The listing should also contain the TNMOC estimate calculated by
summing the concentrations of all peaks (both target and unidentified) detected in the chromatogram.
See Section 2.6.1 for a more detailed discussion on data processing capabilities of automated GC
systems.
2.4.2.3 Analytical System Calibration
The detector response of the analytical system should be calibrated with multiple level
propane primary standards over the expected sample concentration range. Benzene is suggested as a
second primary standard to calibrate dual-column systems. These dual-column systems employ a
Deans®24 switch or other column switching techniques. Benzene may also be used to quantitate the
target compounds when using a single-column approach. The primary calibration standard is used to
generate a response factor per carbon atom for determining the concentration of each target VOC, as
well as the TNMOC. It is impractical and unnecessary to determine compound specific response
factors for each of the target VOCs presented in Table 2-1 because the carbon response of the FID to
these compounds is approximately linear.
For a known, fixed sample volume, concentration is proportional to the area under the
chromatographic peak. The area is converted to ppbC using the following equation:
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 67 of 225
CA RF(AC) (2-5)
where:
CA = Concentration (ppbC)
RF = Response Factor, ppbC/area count
AC = Area Counts
The response factor (RF) is an experimentally determined calibration constant (ppbC/area count), and
is used for all compound concentration determinations. The response factor is determined by the
analysis of the primary standard using the following equation:
3(CB)
RF =J= (2-6)
MAC
where:
Carbon Atoms in Propane (6 when benzene is used as a second calibration
standard)
CB = Concentration of the NIST Propane Standard (ppbv)
MAC = Mean Area Count, determined from the analyses of multiple levels or multiple
injections of the primary standard
The retention time of target compounds is determined by analyzing the retention time
calibration standard as described in Section 2.3.4.2. This standard is analyzed in triplicate, at a
minimum, to establish the correct retention times and retention time windows for the peaks of interest.
The primary standard (Section 2.3.4.1) is used to perform a calibration check of the
analytical system in order to determine system variability and overall performance. The calibration and
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 68 of 225
retention time checks may be performed concurrently using the retention time calibration standard. The
compound concentrations and retention times should compare within the limits of the data quality
objectives established for the monitoring program. If they do not, the analytical system should be
recalibrated.
2.4.3 System Operation
This section provides guidance and general operating considerations for initial system set-up,
optimization of sampling parameters, and field operation for automated GC systems.
2.4.3.1 Initial System Set-up
During the initial set-up of the automated system several parameters must be evaluated to
optimize the operating conditions. Critical parameters include, but are not limited to, the sample
collection flow rate and sample integration time, sample concentration and desorption conditions, oven
temperature program parameters, detector calibration, and the peak detection and integration methods
used by the data acquisition and processing system. These parameters are optimized by varying the
operating conditions to achieve the best resolution and detection of the target VOCs using primary
calibration and retention time calibration standards.
Prior to making VOC measurements using an automated GC system, the baseline
performance of the system must be thoroughly documented. The information from the system baseline
characterization is used to determine system specific target compound retention times, relative retention
times, identification of co-eluting compounds and matrix effects, internal standard retention times,
interferences, and detection limits. Subsequent calibrations and retention time QC checks should be
verified against the system baseline to identify trends or excursions from acceptable performance. See
Section 2.3.7 for a discussion of pre-measurement system characterization.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 69 of 225
Users should anticipate a minimum of six months for initial setup, configuration, familiarization,
and development of SOPs prior to the field implementation of an automated GC system. The system
should initially be set up by the manufacturer and demonstrate adequate system stability and
performance. Under terms of agreement for purchase, the manufacturer
should be required to provide a detailed instruction manual for system operation and to meet the
specifications as defined by the user. For a set of primary system specification guidelines see Section
2.4.4.
2.4.3.2 Sampling Parameters
Determination of optimum sampling parameters is dependent on field conditions
(i.e., expected compound concentration ranges, humidity, temperature, etc.), desired sensitivity,
cryogen consumption, and sample trapping efficiency. During the setup period, these sampling
parameters should be evaluated to determine the optimum conditions for each. Primary sampling
parameters are the sample collection frequency (1 sample each hour) and the minimum sample
collection or integration time (40 minutes).
For hourly sampling, the minimum sample collection or integration time is 40 minutes. A
sample collection volume of 200 to 600 mL is recommended. The sample volume used requires a
trade-off between the required detection limit and potential moisture interference problems. Longer
sample integration times may be implemented by using an intermediate sample collection or integration
device. This device usually consists of a sample integration vessel configured to provide integrated
collection of one sample while the previously collected sample is being analyzed. Advantages to using
an intermediate sample integration device include longer integration times and reduced cryogen use
during the concentration step of sample analysis.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 70 of 225
2.4.3.3 Field Operation
The automated GC system should be installed in a temperature-controlled shelter at the field
location. Detailed SOPs for field operation of the automated GC system must be developed. The
SOPs should be based on information obtained during the set-up and familiarization period and the
requirements of the monitoring program. Refer to QA/QC Section 2.8.3.1 for a more detailed
discussion of SOP development. The system should be maintained by a qualified operator who should
perform the routine operational and quality control functions as specified in the SOPs. Critical
operational checks should be performed as frequently as practical. Operational parameters should be
adjusted, if necessary, so that the data quality objectives are met. It is recommended that all
adjustments to the operational parameters be documented in a laboratory notebook. Primary
calibration and retention time checks should be performed routinely according to the minimum QC
requirements given in Section 2.8. Retention time calibration checks are performed to provide retention
time reference information for validating compound identifications. The retention time calibration
standard can also be used to track the FID response to determine when recalibration is necessary.
2.4.4 System Specifications
A set of primary specifications is provided below to conduct the evaluation for procurement
of an automated GC/FID system. It is imperative that the enhanced O3 monitoring network
requirements for this type of system be compared against vendor offers to ensure that appropriate
systems are procured. Primary system specifications are presented below. Additional system
specifications may be added at the discretion of the user.
The automated GC/FID system must be able to meet the sampling frequency
requirements as prescribed in 40 CFR Part 58, Subpart E, Appendix D, and the
sample integration requirements as discussed in Section 2.1 of this document (minimum
sample integration time of 40 minutes to comprise a 1-hour sample). The manual
methodology described in Section 2.5 is required for collection of the one 24-hour
sample every sixth day.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 71 of 225
Cryogen consumption is a primary consideration for system procurement. Transport
and delivery of liquid cryogen to the site may be impractical. The amount of liquid
cryogen consumed by the system will determine the frequency of site visits and impact
the cost for site operation and the level of data capture. Systems that utilize electronic
cooling devices should be strongly considered if all other user specified requirements
are satisfied.
To avoid cross contamination, the system must demonstrate system background levels
that are below the 0.2 ppbC estimated detection level for each VOC target species and
3 ppbC for TNMOC.
The system must demonstrate the ability to separate the target VOCs of interest (C2
through C12) and provide an adequate estimate of the TNMOC value. Refer to Section
2.2.1 for a discussion of TNMOC.
To ensure adequate peak identification and quantitation by retention time, the system
must incorporate operating parameters that provide stable retention times. Observed
retention time drift must be less than 0.1 minutes.
The sample conditioning device, used to remove moisture from the sample stream and
reduce the effects of moisture on the system, must minimize both polar VOC losses and
the potential for introducing contaminants into the analytical system.
The minimum level of quantitation (LOQ) must be 3.0 ± 0.2 ppbC for propane and
correspond to an FID signal that is 3 to 5 times the baseline noise.
The system should incorporate microprocessor control and battery backup capability to
ensure that all programmed control activities for sample collection and analysis will be
retained should the system power be interrupted. The system should automatically
resume all operations once power is restored to the system to improve the level of data
capture. Although not a requirement, the capability to log and report system
interruptions (date, time, and type of failure) is advantageous.
The system operation should be flexible enough to allow sample collection and analysis
parameters to be easily modified to meet changes in network monitoring frequency and
sample integration times as required.
Expedient and responsive vendor support is a key consideration. The user should
specify that the vendor maintain an adequate supply of replacement parts and a staff of
qualified service technicians to ensure that the minimum number of sampling events are
missed should a system failure occur. The user should specify that the vendor
guarantee that parts and components be delivered to the site within 2 working days
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 72 of 225
from the placement of the order. The user should also specify that the response to
automated GC system service calls be received within 24 hours of placement and the
system be placed in acceptable working order within 7 days of the service request.
The vendor must provide an in-depth, detailed manual covering all aspects of the
automated GC/FID system (i.e., operation, maintenance, etc.), initial system setup, user
training, and demonstrate adequate system performance.
2.5 Manual Method for Collecting and Analyzing Volatile Organic Compounds
The manual methodology for obtaining volatile organic compound (VOC) measurements
involves collecting time-integrated, whole air canister samples for subsequent analysis at a central
laboratory. Under the minimum network monitoring requirements in 40 CFR Part 58, Subpart E,
States must obtain 3-hour and 24-hour integrated measurements of VOCs at specified sample
collection frequencies based on individual PAMS site type requirements. The sample collection
frequencies range from one 24-hour sample every sixth day to eight 3-hour samples every day.
Specific sample collection frequencies are discussed in Section 2.1. A discussion of sample collection
methodology is provided in EPA Compendium Method TO-15 (Appendix A).
Application of the manual methodology to the enhanced O3 monitoring regulations requires
the collection and analysis of a large number of canister samples. The magnitude and success of the
manual monitoring program depends on the quantity of canisters available, the capabilities of the sample
collection system used, the analytical capacity of the central laboratory, and the availability and skill of
staff to address the needs of the specific program design. An integrated, well planned sample collection
and analysis program is necessary to address the numerous aspects of a canister-based monitoring
operation, which include canister cleaning and transport, sample collection procedures and frequency,
analysis procedures, and data acquisition and reporting. These details must address the needs of the
specific program. Users of manual methodology are responsible for the selection, set-up, and
optimization of their specific system(s), and for the preparation of SOPs that delineate the details of all
operations.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 73 of 225
The intent of this section is to provide general guidance on manual methodology. The
following sections generally describe multiple-event and single-event canister sampling equipment,
procedures, and operation. Recommended system specifications applicable to the procurement of
canister sampling systems are also presented.
2.5.1 Sample Collection
This section describes the configuration and use of SUMMA® passivated canisters and
associated multiple- and single-event sample collection systems. These systems provide samples for
subsequent analysis at a central laboratory using a GC/FID analytical system with computerized data
reduction and reporting capabilities.
Canister sample collection systems should be capable of unattended operation in order to
allow collection of samples in accordance with the network monitoring requirements presented in Table
1-1 (see Section 1). Procedures for collecting canister samples are described in Section 2.5.1.1 and
Section 2.5.1.4. Precautions pertaining to the use of canisters, canister cleaning procedures, and
sample collection system certification procedures are discussed in detail in Section 2.5.3.
Collecting time-integrated whole ambient air samples for subsequent analysis of target VOCs
is a widely accepted practice. Samples collected should represent a time-integrated average for the
required sampling period (i.e., collected at a constant flow rate over the full collection period). Sample
collection systems currently in use incorporate diverse operating approaches. The primary difference
among the various approaches is the technique and associated hardware used to perform time-
integration of canister sample collections and multiple- or single-event sample collection capabilities.
Time-integration techniques generally involve the use of electronic and/or mechanical devices. Canister
sampling systems are available commercially or can be custom built by the user for a specific
application.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 74 of 225
Multiple-event sample collection systems are needed to meet the 3-hour, around-the-clock
collection frequency. Back-to-back collection of the individual 3-hour samples may not be practical
using single-event systems due to the required attendance of an operator to change the sample canisters
between events.
2.5.1.1 Multiple-event Sample Collection Equipment
A typical multiple-event sample collection system configuration is presented in Figure 2-11.
The multiple-event canister sample collection system is comprised of the following primary components:
-------
I
Sample
Inlet
Section: 2
Revision: 1
Date: 09/30/98
Page: 75 of 225
Manifold
By-pass Pump
Microprocessor
Sample Canisters
Figure 2-11. A Typical Multiple-event Sample Collection System
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 76 of 225
Inlet probe and manifold assembly - Constructed of glass (see Figures 2-9 and 2-10) or
stainless steel. Used as a conduit to transport sample air from the atmosphere at the required
sampling height and distribute it for collection.
By-pass pump - A single- or double-headed diaphragm pump, or a caged rotary blower.
Used to continuously draw sample air through the inlet probe and manifold assembly at a rate
in excess of the sampling system total uptake. All excess sample air is exhausted back to the
atmosphere.
Sample pump - A stainless steel bellows pump, capable of 2 atmospheres above ambient
output pressure. Used to extract sample air from the manifold assembly and deliver it to the
sample canister during collection.
Sample inlet line - Chromatographic-grade stainless steel tubing. Used to connect the
sampler to the manifold assembly.
Sample canisters - SUMMA® passivated stainless steel sample vessels of desired internal
volume with a bellows valve attached at the inlet of each unit. Used to contain the collected
sample air for transportation and analysis.
Electronic pressure sensor - A pressure measurement device capable of measuring
vacuum (0-30 in Hg) and pressure (0-30 pounds per square-inch gauge). Used to measure
initial and final sample canister pressures.
Adjustable orifice and mass flow meter assembly or electronic mass flow controller -
An indicating flow control device(s). Used to maintain a constant flow-rate (± 10%) over a
specific sampling period under conditions of changing temperature (20-40 °C) and humidity
(0-100% relative).
Particulate filter - Two micron sintered stainless steel in-line filter. Used to remove
particulate material larger than 2 microns from the sample air being collected.
Microprocessor - An event control and data acquisition device. Used to allow unattended
operation (i.e., activation and deactivation of each sampling event) of the sampling system and
to record sampling event specific process data (i.e., start and end times, elapsed times, initial
and final sample pressures, etc.).
Solenoid valves or a multi-port rotary valve - Eight electric-pulse-operated or low
temperature coil, stainless steel body solenoid valves with Viton® plunger seat and o-rings or
one multi-port stainless steel body rotary valve with Viton® o-rings. Used to provide access
to or isolation of the sample canister(s).
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 77 of 225
Stainless steel tubing and fittings - Isolation and interconnection hardware. Used to
complete system interconnections. All tubing in contact with the sample prior to analysis
should be chromatographic grade stainless steel and all fittings should be 316 grade stainless
steel.
2.5.1.2 Multiple-event Sample Collection Procedure
Samples are collected in individual canisters using a single pump and one or more flow
control devices. A stainless steel metal bellows style pump draws in ambient air from the sampling
probe and manifold assembly at a constant flow rate to fill and pressurize each sample canister during
each specific sampling event.
A flow control device(s) is used to maintain a constant sample flow rate into each canister
over each specific sampling period. The flow rate used is a function of the final desired sample
pressure, the internal volume of the canister used, and the specified sampling period and assumes that
the canisters start at a pressure of 5 mm Mercury (Hg) absolute. The flow rate is calculated as follows:
(2-7)
Where:
F = flow rate (mL/min)
P = final canister pressure, atmospheres absolute
V = volume of the canister (mL) at one atmosphere pressure
T = sample period (hours)
60 = minutes in an hour
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 78 of 225
For example, if 6-L canisters are to be filled to 1.5 atmospheres absolute pressure each over individual
3-hour integration period (i.e., collection episode), the flow rate specific to each period is calculated as
follows:
1.5 atm x 6000mL/atm cn T / •
50mL/min
3 hr x 60 min/hr
During operation, the microprocessor control device is programmed to activate and
deactivate the components of the sample collection system, consistent with the beginning and end of
each individual sample collection period.
The use of individual electric-pulse-operated or low temperature coil solenoid valves avoids
any substantial temperature rise that would occur with conventional coil solenoid valves. The
temperature rise associated with conventional coil solenoid valves could cause outgassing of organic
compounds from valve components into the samples.
Electric-pulse-operated solenoid valves require only a brief electrical pulse to open or close
at specified start and stop times. The valve, therefore, experiences no temperature increase. The
pulses may be obtained either from the microprocessor directly, if it can be programmed for short (5 to
60 seconds) actuation pulses, or by incorporating an attached electric pulse circuit.
Low temperature coil solenoid valves incorporate a low current draw circuit design to lift the
plunger and valve seat assembly (i.e., open the valve) when the coil is energized. A spring returns the
plunger and valve seat assembly (i.e., closes the valve) automatically when the coil is not energized.
Because no electric-pulse circuit is required, the low temperature coil solenoid valves are much simpler
in design and considered more reliable.
Canister sampling systems can collect sample from a shared sample probe and manifold
assembly as described in Section 2.4.1.1 or from a dedicated stainless steel sample probe, manifold
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 79 of 225
assembly, and by-pass pump. If a dedicated probe, manifold assembly, and by-pass pump are used,
provisions should be made (i.e., using a separate timer device, etc.) to start the by-pass pump several
hours prior to the first sampling event of a multiple-event collection period to flush and condition the
sample collection system components. The connecting lines between the sample inlet and each canister
should be kept as short as possible to minimize internal surface area and system residence time.
The flow rate into each canister should remain relatively constant over the entire collection
period of each sampling event. If an adjustable orifice(s) is used as the flow control device(s), a drop in
the flow rate will occur near the end of the each sample collection period as pressure in the canister
increases. Typically this condition occurs when canister pressure exceeds one-half atmosphere above
ambient pressure. Consequently, care must be taken to select a sample flow rate that will yield a final
pressure that will not significantly exceed 22-24 psia (i.e., -8-10 psig) at the end of the sample
collection interval.
Prior to any field use, each sample collection system should be certified as nonbiasing,
meaning that the sample collection system does not add to or subtract from the concentrations of the
samples collected using it. (Refer to Section 2.5.1.7 for details pertaining to canister sampling system
certification). The canisters should also be determined to be clean before each use. (Refer to Section
2.5.2 for details pertaining to canister cleaning.) Each adjustable orifice and mass flow meter assembly,
or mass flow controller, used as a flow control device should be calibrated against a primary flow
measurement standard (i.e., a bubble flow meter, etc.). The calibration qualifies the relationship
between indicated flow versus measured flow. Multiple calibration points of comparison, spanning the
entire range of the flow control device in increments of 10% of the device range, should be used. For
example, if a mass flow meter having a 0-100 mL measurement range is being calibrated, comparisons
would be made at 0, 10, 20, 30, 40, 50, 60, 70, 80, 90 and 100 mL. Calibration curves that
mathematically define the relationship between the flow rates indicated by the flow control device and
corresponding actual flow rates measured by primary flow measurement standard are generated from
these comparisons. The calibration curves are used to set actual desired flow rates, based on the flow
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 80 of 225
rates indicated by the flow control devices. Pressure sensors should be calibrated against a primary
pressure measurement standard (i.e., manometer or absolute pressure gauge), and should also be
calibrated using the same process described above. Calibration of the flow control and pressure
measurement devices should be performed prior to any field deployment. A calibration check should
then be conducted periodically according to a program specific QA/QC schedule as developed by the
user. The calibration check should consist of performing a single point comparison at a representative
setting (e.g., a flow rate typically used for sample collection). The recommended frequency for
performing calibration checks is biannually (two calibration checks per year).
The following procedure provides generic steps for operating a typical multiple-event
sampling system:
1. Set the sample collection system to the desired sample collection flow rate(s)
(i.e., referencing the corresponding calibration curve(s) and considering the canister
volume and the desired final canister pressure).
2. Program the microprocessor event control system to begin and stop sampling consistent
with user specific collection frequency requirements.
3. Attach all sample canisters to the sample collection system.
4. Open all the canister bellows valves.
5. After sampling, record the initial and final sample pressures in each canister
(i.e., referencing the corresponding calibration curve(s)), and the start and end time of
each collection event onto the sampling field data sheet. The microprocessor event
control and data acquisition system should automatically store these data for each
collection event. Final sample pressure should be close to the desired calculated final
pressure.
6. Close all of the canister bellows valves.
7. Attach an identification tag to each canister documenting the canister serial number,
sample event number, sample type, location, and collection date.
8. Disconnect and remove each canister from the sample collection system.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 81 of 225
An in-depth SOP, specific to the exact sample collection system utilized, must be developed.
2.5.1.3 Multiple-event System Specifications
Multiple-event sample collection systems will be required if canisters are used to meet the
network monitoring requirements for VOC measurements in a practical, non-labor intensive manner.
A set of primary specifications is provided below to direct the evaluation and procurement of
multiple-event sample collection systems. It is imperative that the site requirements for this type of
system be compared against vendor offerings to ensure that appropriate systems are procured.
Primary collection system specifications are presented below. However, additional collection system
specifications may be added at the discretion of the user.
An in-depth, detailed manual describing all aspects of the sample collection system
(i.e., operation, maintenance, etc.) must be provided by the vendor.
The overall size of the sample collection system, including canisters, should be kept as
compact as possible. The sample collection systems are usually installed into existing
sampling site shelters where many other parameters (i.e., criteria pollutant
concentrations, meteorological conditions, etc.) are also being measured. Each of the
other parameters requires separate instrumentation and consequently the shelters can
become very crowded.
The sample collection system should meet all applicable electrical and safety codes,
operate on standard 110 Vac power, and incorporate a main power fuse or circuit
breaker. Specific potential electrical hazard and/or other safety considerations should
be detailed in a supplied users manual.
The overall configuration, and components comprising that configuration, should allow
for simple operation, maintenance, and service of the sample collection system.
Materials used in the construction of components of the sample collection system
should exhibit nonbiasing characteristics, i.e., the materials should neither contribute to
nor take away from the measured organic content. The components themselves should
generally conform to the descriptions presented in Section 2.5.1.1. All surfaces that
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 82 of 225
come in direct contact with sampled air should be constructed of glass, stainless steel,
or Viton®. The use of Teflon® or other plastics or polymers should be avoided because
the absorption/ desorption characteristics of these materials increase the potential for
sample bias.
The sample collection system should be able to be certified as nonbiasing. The user
should seek assurances and/or evidence that the sample collection system
design/configuration being considered will be, or has been certified according to the
recommended procedures presented in Section 2.5.1.7.
Ideally, the sample collection system should be able to accommodate the sample
collection event frequency presented in Section 2.1 and simultaneously allow a
duplicate 3-hour sample collection as recommended for QC purposes. The sampling
system should have the capability to collect the following during any given 24-hour
period:
Eight 3-hour time-integrated canister samples;
One 3-hour time-integrated duplicate canister sample, collected concurrently
with one of the eight 3-hour canister samples; and
One 24-hour time-integrated canister sample, collected concurrently with the
eight 3-hour samples, but not concurrently with the duplicate 3-hour canister
sample.
It is imperative that the sample system have the collection capabilities detailed above,
including the 24-hour sample collection. If not, a second sample collection system will
be required to address the 24-hour sample collection and, consequently, more overall
labor and space would be needed to fully address the network monitoring
requirements.
The ability of the sample collection system to perform sample collections as presented
above would allow the operator to visit the site only twice during the 24-hour period
being characterized; once to install cleaned evacuated canisters prior to sampling and
once to remove canisters containing the collected samples.
The sample collection system must be able to perform time-integration of the canister
sample collections. The sample collection system should allow for variable collection
flow rates so that canisters of different internal volume may be used (refer to Section
2.5.1.2 for specifics on the relationship between canister volumes, collection duration,
and collection flow rate).
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 83 of 225
The sample collection system should incorporate a microprocessor event control and
data acquisition device. At a minimum this microprocessor should be programmable to
control the start and stop times of every collection event within a 24-hour sampling
period. The microprocessor should also be able to simultaneously collect and store all
the sample collection process data pertaining to each sampling event as follows:
Start and stop times for each sample collection; and
Beginning and ending sample canister pressures for each sample collection.
The microprocessor should incorporate a battery backup system to address power
failure situations. Battery backup is necessary to ensure fewer invalidated samples and
a higher collection completion rate. The battery backup system would ensure that all
programmed control activities and collection process data are retained for a
predetermined interval in the event that standard power to the system is interrupted.
Retaining the programmed control activities would allow sample collection to resume
automatically at the next programmed event time when standard power is restored to
the sample collection system. Retaining the collection process data for samples
collected prior to the termination of standard power allows these samples to be
qualified as valid or invalid, based on sampling start and stop times and initial and final
pressures. Although not absolutely necessary, adding a miniature printer would allow
for the generation of a report listing all sample collection process data.
Expedient and responsive vendor support should be a mandatory requirement and
primary consideration when procuring a multiple-event canister sample collection
system. The user should specify that the vendor maintain an adequate supply of
replacement parts and qualified service technicians to ensure that the absolute minimal
number of sampling events is missed should a sample collection system failure occur.
The user should specify that the vendor guarantee that parts/ components be delivered
to the sampling site within two working days of order placement, and that a sample
collection system delivered to the vendor for repair or for other problems be serviced
and returned to the user within seven working days.
2.5.1.4 Single-event Sample Collection Equipment
A typical single-event sample collection system configuration is presented in Figure 2-12.
The single-event sample collection system consists of the following primary components:
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 84 of 225
Inlet probe and manifold assembly - Constructed of glass (see Figure 2-4) or stainless
steel. Used as a conduit to transport sample air from the atmosphere at the required sampling
height and distribute it for collection.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 85 of 225
Manifold
By-pass Pump
1
Sample
Inlet
To Atmosphere
\f Particulate
-£- Filter
Pressure
Measurement
Microprocessor
(Optional)
Adjustable Orifice
and Rotameter
or Mass Flow Meter
or
Electronic Mass
Flow Controller
Mechanical
Pressure Gauge
Electronic
Pressure Gauge
(Optional)
Figure 2-12. A Typical Single-event Sample Collection System
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 86 of 225
By-pass pump - A single- or double-headed diaphragm pump, or a caged rotary blower.
Used to continuously draw sample air through the inlet probe and manifold assembly at a rate
in excess of the sampling system total uptake. All excess sample air is exhausted back to the
atmosphere.
Sample pump - A stainless steel bellows pump, capable of 2 atmospheres above ambient
output pressure. Used to extract sample air from the manifold assembly and deliver it to the
sample canister during collection.
Sample inlet line - Chromatographic-grade stainless steel tubing. Used to connect the
sampler to the manifold assembly.
Sample canisters - SUMMA® passivated stainless steel sample vessels of desired internal
volume with a bellows valve attached at the inlet of each unit. Used to contain the collected
sample air for transportation and analysis.
Stainless steel vacuum/pressure gauge or electronic pressure sensor (optional) - A
pressure measurement device capable of measuring vacuum (0-30 in Hg) and pressure
(0-30 pounds per square-inch gauge). Used to measure initial and final sample canister
pressures.
Adjustable orifice and rotameter. or mass flow meter assembly, or electronic mass
flow controller - An indicating flow control device (or devices). Used to maintain a constant
flow rate (± 10%) over a specific sampling period under conditions of changing temperature
(20-40 °C) and humidity (0-100% relative).
Particulate filter - Two-micron sintered stainless steel in-line filter. Used to remove
particulate material larger than 2 microns from the sample air being collected.
Electronic timer or microprocessor (optional) - An event control device. Used to allow
unattended operation (activation and deactivation) of the collection system.
Solenoid valve - An electric-pulse-operated or low temperature coil, stainless steel body,
solenoid valve, with Viton® plunger seat and o-ring. Used to provide access to or isolation of
the sample canister(s).
Elapsed time indicator - A time measurement device used to measure the duration of the
sampling episode.
Stainless steel tubing and fittings - Isolation and interconnection hardware. Used to
complete system interconnections. All tubing in contact with the sample prior to analysis
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 87 of 225
should be chromatographic grade stainless steel and all fittings should be 316 grade stainless
steel.
2.5.1.5 Single-event Sample Collection Procedure
The sample is collected in a canister using a pump and flow control device. A stainless steel
metal bellows style pump draws in ambient air from the sampling probe and manifold assembly at a
constant flow rate to fill and pressurize the sample canister.
A flow control device is used to maintain a constant sample flow rate into the canister over a
specific sampling period. The flow rate used is a function of the final desired sample pressure, the
internal volume of the canister used, and the specified sampling period. A starting pressure of 5 mm
mercury (Hg) absolute for the canisters is assumed. The flow rate is calculated using the formula
presented in Section 2.5.1.2.
During operation, the timer is programmed to activate and deactivate the sample collection
system at specified times, consistent with the beginning and end of a sample collection period.
Single-event sample collection systems can collect sample from a shared sample probe and
manifold assembly as described in Section 2.4.1.1 or from a dedicated stainless steel sample probe,
manifold assembly, and by-pass pump. If a dedicated probe, manifold assembly, and by-pass pump
are used, a second electronic timer should be incorporated to start the by-pass pump several hours
prior to the sampling period to flush and condition the components. The connecting lines between the
sample inlet line and the canister should be as short as possible to minimize internal surface area and
system residence time.
The flow rate into the canister should remain constant over the entire sampling period. If an
adjustable orifice is used as the flow control device, a drop in the flow rate will occur near the end of
the sample collection period because the orifice size is no longer critical as pressure in the canister
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 88 of 225
increases. Typically this condition occurs when canister pressure exceeds one-half atmosphere above
ambient pressure. Consequently, care must be used to select a sample flow rate that will yield final
pressure that will not significantly exceed 22-24 psig (i.e., -8-10 psig) at the end of the sample
collection interval.
Prior to field use, each sample collection system should be certified as nonbiasing. (Refer to
Section 2.5.1.7 for details pertaining to canister sample collection system certification.) The canisters
should also be demonstrated to be clean before each use. (Refer to Section 2.5.2 for details pertaining
to canister cleaning.)
The following generic steps are provided for the operation of a typical single-event sample
collection system:
1. Activate the sample collection system and verify the correct sample flow rate using a
calibrated mass flow meter or rotameter. The flow can be measured directly at the inlet
of the system. The calibrated mass flow meter or rotameter is attached to the sample
inlet line, before the particulate filter. The sample collection system is activated and the
indicated flow rate is compared to the desired collection flow rate. The values should
agree within ± 10%. If a mass flow controller is being used as the sample collection
system flow control device, allow the system to equilibrate for two minutes. After the
two-minute equilibration, the desired sample flow rate is attained by adjusting the
system mass flow controller until the calibrated mass flow meter or rotameter indicates
the correct flow rate. If the sample collection system uses a adjustable orifice assembly
as the flow control device, adjust the orifice size until the correct flow rate is achieved.
If the sampling system uses a fixed orifice as the flow control device, ensure that the
orifice is the correct size to provide the desired flow rate.
2. Deactivate the sample collection system and reset the elapsed time indicator to show no
elapsed time.
3. Disconnect the calibrated mass flow meter or rotameter and attach a cleaned evacuated
canister to the sample collection system.
4. Open the canister bellows valve.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 89 of 225
5. Record the initial vacuum in the canister, as indicated by the sample collection system
vacuum gauge, on the canister sampling field data sheet.
6. Record the time of day and elapsed time indicator reading on the canister sampling field
data sheet.
7. Set the electronic timer to start and stop sampling at the appropriate times.
8. After sample collection, record the final sample pressure on the sampling field data
sheet. Final sample pressure should be close to the desired calculated final pressure.
Time of day and elapsed time indicator readings should also be recorded.
9. Close the canister bellows valve. Disconnect and remove the canister from the sample
collection system.
10. Attach an identification tag to the canister documenting the canister serial number,
sample number, sample type, location, and collection date.
An in-depth SOP, specific to the exact sample collection system utilized, must be developed.
2.5.1.6 Single-event System Specifications
Although different in complexity and general application, many of the collection system
specifications that apply to multiple-event collection systems also apply to single-event collection
systems. To ensure that a single-event collection system meets the user's program needs, the following
system specifications should be presented to, and addressed by, the candidate vendor(s) prior to
procurement.
An in-depth, detailed manual covering all aspects of the sample collection system
(i.e., operation, maintenance, etc.) must be provided by the vendor.
Although fewer canisters are utilized at a time, as compared to the multiple-event
systems, the overall size of the sampling system should still be kept as compact as
possible.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 90 of 225
The sampling system should meet all applicable electrical and safety codes, operate on
standard 110 Vac power, and incorporate a main power fuse or circuit breaker.
Specific potential electrical hazards and/or other safety considerations should be
detailed in a supplied user's manual.
The overall configuration, and components comprising that configuration, should allow
for simple operation, maintenance, and service of the sample collection system.
Materials used in the construction of components of the sample collection system
should exhibit nonbiasing characteristics. The components themselves should generally
conform to the descriptions presented in Section 2.5.1.4. All surfaces that will come in
direct contact with sampled air should be constructed of glass, stainless steel, or
Viton®. The use of Teflon® or other plastics or polymers should be avoided because
the absorption/desorption characteristics of these materials increase the potential for
sample bias.
The sample collection system should be able to be certified as nonbiasing. The user
should seek assurances and/or evidence that the sample collection system
design/configuration being considered can or has been certified according to the
recommended procedures presented in Section 4.0 of Compendium Method TO-12
(Appendix C).
The sample collection system should be able to perform time-integration of the canister
sample collections, and allow for variable collection flow rates so that different volume
canisters may be used.
Expedient and responsive vendor support should be a mandatory requirement and
primary consideration when procuring a single-event canister sample collection system.
The user should specify that the vendor maintain an adequate supply of replacement
parts and qualified service technicians to ensure that the absolute minimal number of
sampling events is missed should a sample collection system failure occur. The user
should specify that the vendor guarantee that parts/components be delivered to the
sampling site within two working days of order placement. The user should also specify
that a sample collection system delivered to the vendor for repair or for other problems
be serviced and returned to the user within seven working days.
An alternative procedure involves passive sampling into a SUMMA®-polished canister, with
subsequent pressurization of the canister to 30 psig at the central laboratory. This procedure reduces
the relative humidity in the canister but also dilutes the sample.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 91 of 225
2.5.1.7 Canister Sampling System Certification
Canister sampling systems should exhibit nonbiasing characteristics before being used to
collect samples. These sampling systems should be subjected to laboratory certification to quantify any
additive or subtractive biases that may be attributed directly to the sampling system. The following
procedure is recommended for certifying canister sampling systems. Alternative approaches are
acceptable provided they are properly described and documented.
A challenge sample, consisting of a blend of organic compounds at a known concentration in
clean humidified zero air, is collected through the sampling system. A reference sample is concurrently
collected using a dedicated mass flow controller that has been characterized prior to each use. The
samples are then analyzed using a GC system that is equivalent to or better than the GC system that will
be used to analyze field volatile organic O3 precursor samples. The percent recoveries for target
challenge compounds are calculated, based on the concentrations determined for the reference sample.
Recoveries of each of the challenge compounds should be in the range of 80-120% of the
concentrations determined for the reference sample. A system-specific overall recovery should also be
calculated. The overall recovery is the average of the individual compound recoveries. Each sampling
system should have an overall recovery of 85-115%. The challenge sample percent recoveries are
used to gauge potential additive and/or subtractive bias characteristics for each specific sampling
system.10
In addition to characterizing the sampling system with a blend of VOCs, the system should
also be characterized using humidified zero air. A humidified zero air blank sample is collected through
the sampling system to further gauge the potential for additive bias. The blank samples can be analyzed
for specific target analytes, TNMOC, or both, depending on individual program requirements. Two
criteria apply to the blank portion of the certification process: a determined concentration criterion of
0.2 ppbC or less for any individual target compound is required if speciation analysis of the blank
sample is performed, and a TNMOC concentration criterion of 10 ppbC or less is also required.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 92 of 225
Sampling is accomplished using dedicated manifolds for both the zero and challenge phases of
the certification procedure (Figures 2-13 and 2-14). Zero air supplied to the zero manifold should be
hydrocarbon-free and humidified to approximately 70% relative humidity.
The zero air should be supplied from a canister cleaning system similar to the one described in
Section 2.5.2.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 93 of 225
Cleaned humidified
zero dilution gas (in)
Sample
canisters [ 1
(1-4)
Canister sampling
systems (1 -4)
To atmosphere
Exit
rotameter
Zero manifold
(not temperature controlled)
Canister sampling
systems (5-8)
Sample
canisters
(5-8)
Figure 2-13. Dedicated Manifold for Zero Gas Certification
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 94 of 225
Certified challenge
stock gas (in)
Cleaned humidified
zero dilution gas (in)
Mixing chamber
Controlled temperature challenge manifold
Power out
Thermocouple
Programmable
temperature
controller
^sphere
•
Exit
rotameter
Reference sample collection
(characterized mass flow controller)
Reference
sample
canister
Sample
canisters
(1-5)
Canister sampling
systems (5-8)
Figure 2-14. Dedicated Manifold for Challenge Gas Certification
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 95 of 225
2.5.1.7.1 Certification Equipment
The equipment required to perform canister sampling system certification is described below.
The equipment listed is consistent with the systems presented in Figures 2-13 and 2-14.
Mass flow controllers - Mass flow controllers located at the inlets to the manifolds. Mass
flow controllers are used to regulate the certification pollutant, diluent, and zero air flow rates.
Also, a dedicated, characterized mass flow controller is used to collect reference samples.
Mixing chamber - A mixing chamber located between the outlets of the mass flow
controllers and the inlet of the challenge manifold. The mixing chamber is a stainless steel
vessel with opposed inlet and outlet ports that cause the blend of challenge gases and the
diluent gas to swirl and mix prior to entering the challenge manifold. The mixing chamber is
used to ensure that a homogeneous blend of challenge gas is delivered to the challenge
manifold. The zero manifold does not require a mixing chamber.
Challenge gas manifold - A challenge gas manifold constructed of 1/8-inch O.D.
chromatographic grade stainless steel tubing and 1/8-inch tee fittings. The challenge manifold
is used to distribute challenge gas to the individual sampling systems being certified. The
number of sample ports provided on the challenge gas manifold is determined by the number
of sampling systems to be certified simultaneously.
Zero air manifold - A zero air manifold constructed of 1/4-inch O.D. chromatographic
grade stainless steel tubing and 1/4-inch fittings. The zero manifold is used to distribute zero
air to the individual sampling systems being certified. The number of sample ports provided
on the zero air manifold is determined by the number of sampling systems to be certified
simultaneously.
Exit rotameter - An exit rotameter located at the outlet of both the challenge gas and zero
air manifolds. The exit rotameter is used to visually indicate that an excess of challenge gas or
zero air is present in the respective manifolds during certification sample collection.
Cord heater - A cord heater rated at 80 watts spiraled around the outside of the challenge
manifold. The cord heater is used to heat the challenge manifold to 80°C. Heating the
challenge manifold helps to reduce the potential for loss of challenge gas compounds to the
walls of the challenge manifold. The zero manifold is not heated.
Temperature controller- A temperature controller used in conjunction with the cord heater
to actively regulate the challenge manifold temperature at 80°C.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 96 of 225
2.5.1.7.2 Certification Procedure
The procedure to perform canister sampling system certification is presented below.
1. Perform a positive pressure leak check of all sampling system fittings. Attach source of
pressurized air to the inlet of the system. Coat the fittings with indicating bubble solution
to locate leaks. Repair any leaks found. Perform a negative pressure leak check.
Attach an evacuated canister to the exit of the sampling system. Open the canister
bellows valve and record the initial vacuum, indicated by the sample pressure gauge.
Close the canister bellows valve and view the sample pressure gauge and determine
whether vacuum is maintained. The system is leak free if the vacuum is maintained. If
vacuum is not maintained, the system is not leak free. Repair leaks and retest the
system.
2. Connect the sampling systems and the reference sample flow controller to the zero
manifold and purge them with humidified zero air for 48 hours. The purge air should
simultaneously be routed to the challenge manifold to clean and prepare it for challenge
sample collection. Terminate the humidified zero air flow at the end of the 48 hour
period.
3. Purge the sampling systems, reference system, and manifold with dry zero air for 1 hour
to removed accumulated moisture. During the dry purge, determine the certification
flow requirements using the following equation:
Qt (5sxN1) + (QRxN2)xF1 (2-8)
where:
Qt = Total required flow rate (mL/min)
Qs = Individual sampling system collection flow rate (mL/min)
N! = Number of sampling systems
QR = Reference system collection flow rate (mL/min)
N2 = Number of reference systems
Fj = Excess flow factor = 2.0
-------
4.
Section: 2
Revision: 1
Date: 09/30/98
Page: 97 of 225
Determine the pollutant and diluent flows required to generate the desired concentration
of challenge gas using the following equations:
Steo 1
(2-9)
where:
F2
Dilution factor (for use in next equation)
Desired challenge gas concentration (ppbv)
Concentration of the stock cylinder (ppbv)
Steo 2
Ster
Q F2 x QT
(2-10)
where:
QP
QT
Pollutant flow rate (mL/min)
Total required flow rate
QD QT - QP
(2-11)
where:
QD
Diluent flow rate (mL/min)
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 98 of 225
5. Generate and deliver the challenge gas to the challenge manifold and sampling systems.
Condition the challenge manifold with the challenge gas for 10 minutes, with the
sampling systems off. Condition the challenge manifold an additional 10 minutes with
the sampling systems on, and in the bypass mode. Connect a clean evacuated canister
to each sampling system.
6. Collect the challenge and reference samples. Conduct challenge sample collection
according to the normal specified operation of the sampling system.
7. Connect the sampling systems to the zero manifold and purge with zero air, humidified
to 100% relative humidity, for 48 hours. Dry the manifold and samplers with dry zero
air for 1 hour. Adjust the zero air stream to 70% relative humidity. Condition the zero
manifold for 10 minutes with the sampling systems off. Condition the zero manifold an
additional 10 minutes with the sampling systems on, and in the bypass mode. Connect
a clean evacuated canister to each sampling system.
8. Collect the humidified zero air blank samples. Conduct the blank sample collections
using the same sampling system operating procedures used during the challenge sample
collection.
9. Analyze the zero and challenge samples and calculate the percent recoveries.
The sampling system must be challenged with a known concentration of selected analytes
prior to deployment. Additional challenges at the middle and end of the sampling season are
recommended. Operator/analyst judgment is critical: a challenge should be performed whenever the
operation of the sampling system is questioned for any reason.
2.5.2 Canister Cleaning
The canister cleaning procedure and equipment described in this section are recommended
when obtaining integrated whole ambient air samples for subsequent analysis of VOCs. The cleaning
procedure involves purging the canisters with cleaned humidified air and then subjecting them to high
vacuum.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 99 of 225
The purpose of canister cleaning is to ensure that the canister interior surfaces are free of
contaminants and that the canister meets a predetermined cleanliness criterion (i.e., < 10 ppbC
NMOC). This level of cleanliness minimizes the potential for carryover of organic pollutants from one
sample to the next, and helps ensure that the samples collected are representative.
2.5.2.1 Canister Cleaning Equipment
The equipment required to clean canisters includes a source of clean, humidified air to
pressurize the canisters to a pressure of 20 psig, and a vacuum system for evacuating the canisters to
5 mm Hg absolute pressure. Air from a standard oil-less air compressor will contain pollutants from the
ambient air. In addition, various VOCs will be found in the compressed air because of the lubricants
used in the air compressor. Hydrocarbon-free air may be purchased in cylinders and humidified before
being used in the cleaning process. However, this approach may be cost-prohibitive. Figure 2-
15 presents the schematic of a canister cleanup system that is suitable for cleaning up to 16 canisters
concurrently. This, and any alternative system, must include a vacuum pump capable of evacuating the
canisters to an absolute pressure of 5 mm Hg. The equipment is designed so that one manifold of eight
canisters is undergoing the pressurization portion of the cleaning cycle while the other manifold of eight
canisters is undergoing the vacuum portion of the cleaning cycle.
The following equipment is incorporated in a canister cleaning system.
Air compressor - A shop or laboratory oil-less air compressor used to provide the air
supply for the canister cleanup apparatus.
Coalescing filter - A coalescing filter designed to remove condensed moisture or
hydrocarbon contaminants present in the air supplied from the air compressor.
Permeation driers - Permeation driers used to dry the air prior to introduction into the
catalytic oxidizers. Two permeation driers are installed in parallel.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 100 of 225
Moisture indicators - Visual moisture indicators installed in the transfer lines between the
permeation driers and the catalytic oxidizers to monitor the performance of the permeation
drier.
Catalytic oxidizers - Catalytic oxidizers installed in the clean-air system to oxidize any
hydrocarbon contaminants that may be present in the air supplied by the air compressor. For
best results and most efficient operation of the catalytic oxidizers, manufacturer's
specifications should be strictly followed.
-------
Filter Assemblies
Air Compressor
Air Flow
Rota meters
Air
Humidifier
Air Bypass
To Certification System
Routing
Valve
Absolute
Pressure
Gauge
o
Vacuum Source
Selector Valve
A* B
I
Catalytic
Oxidizers
Vacuum
Cryotrap
LJ
Turbomolecular
Pump
Roughing
pump
A* B
8-Port
Manifold
DDDDDDDD DDDDDDDD
A. Manifold Air Pressure Valve
B. Manifold Vacuum Valve
C. Manifold Pressure Release Valve
D. Manifold Port for Connecting Canisters to be Cleaned
o:/s/g/morr/3797/pams/hobson/graph5.ppt
Tf
oo
^ O
o vo
1—1 uS
8, ^
Figure 2-15. Schematic of a Canister Cleanup System
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 102 of 225
Filter assemblies - A 5-micron sintered stainless steel filter installed in the filter housing
assembly downstream of each catalytic oxidizer to trap any paniculate material that may be
present in the air stream leaving the catalyst bed of the oxidizer.
Air cryotrap and purge valves - The air cryotrap allows the cleaned air supply lines to be
subjected to cryogenic temperatures to condense (1) water formed during the oxidation of
hydrocarbons, (2) any remaining unoxidized hydrocarbons, and (3) other condensables. Air
cryotrap purge valves are used to purge these condensed components from the air cryotrap,
as described in the operating procedure described below.
Pressure regulators - A high purity dual stage pressure regulator installed in each branch of
the air supply line so that the maximum pressure attained during the cleanup procedure is
controlled at 20 psig.
Flow controllers - The flow control devices shown in the canister cleanup schematic (Figure
2-10) are metering valves. The flow rates are set not to exceed the maximum recommended
flow rate through the catalytic oxidizers.
Air flow rotameters - Rotameters installed in the air supply lines to allow monitoring of the
flow rates through the catalytic oxidizers.
Air humidifier - The air humidifier shown in Figure 2-15 is a SUMMA®-passivated,
double-valve stainless steel canister with an inlet dip tube that projects to the bottom of the
sphere. FffLC-grade water is placed in the canister prior to use. Two rotameters are
connected to control air flow so that about 80% of the flow rate can be directed to the
humidifier (to bubble through the water to become saturated), while the other 20% bypasses
the humidifier. This procedure allows the humidification apparatus to supply cleaned, dried
air that has been humidified to a relative humidity of-80%.
Manifold air pressure valves - Manifold air pressure valves used to isolate the air supply
system from the manifold, or to make the pressurized air available to the manifold.
Eight-port manifolds - Eight-port manifolds designed to allow up to eight canisters at a time
to be connected. Fewer canisters may be connected to the manifold if the vacant ports are
sealed off with a plug fitting.
Roughing pump - The roughing pump shown in Figure 2-15 is a high-capacity diaphragm
vacuum pump used to remove the moist cleaning air from the canisters while evacuating the
canisters to about 100 mm Hg absolute. The high moisture content of the cleaning air
contained in the canisters will not impede the function of this diaphragm style pump, but will
impede the performance of the high-vacuum pump.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 103 of 225
High-vacuum pump - A high-vacuum pump capable of reducing the pressure in the canisters
to 5 mm Hg absolute. High moisture content will impede the performance of the high-vacuum
pump.
Vacuum cryotrap - A U-shaped trap located in the vacuum manifold that is sized to fit inside
a Dewar flask filled with cryogen. The purpose of this trap is to condense water vapor from
the air that is pulled from the canisters during the vacuum cycle and prevent back-diffusion of
organic vapors from the high-vacuum pump into the canisters during the vacuum cycle of the
cleaning procedure.
Vacuum source selector valve - The vacuum source selector valve is a multiposition valve
used to route either the roughing pump or the high vacuum pump to the eight-port manifold
assemblies or isolate both pumps from the manifold assemblies.
Compound absolute pressure gauge - An absolute pressure gauge used to measure the
pressure attained in the canisters during the vacuum and pressurization cycles of the cleaning
procedure. The absolute pressure gauge must be able to measure absolute pressures from 40
psig down to 0.5 mm Hg absolute.
Air bypass valve - The air bypass valve is used to allow for a 1.0 L/min flow of air to be
maintained through the catalytic oxidizers when the cleaning system is not in use. This flow
prevents the oxidizers from overheating when the clean up system is not in use.
Manifold valves - The manifold vacuum valve and the manifold pressure valve are used to
apply vacuum or pressure to the canisters, as required during the cleaning procedure.
Manifold ports - The manifold ports permit connection of the canisters to the manifold.
Fittings that mate directly with the canister valve fittings are used. These connections will not
leak during the pressurization portion or the vacuum portion of the cleaning procedure.
2.5.2.2 Canister Cleaning Procedure
The cleanup system is prepared for use by checking the position of all the valves. All valves
should be closed initially, with the exception of the air bypass valve. Fill both the air source and
vacuum pump vacuum flasks with cryogen and actuate the high-vacuum pump. Ensure that these
vacuum flasks remain filled with cryogen throughout all cleanup activities. The inlet bellows valve on the
humidifier is opened and the valve on the wet air rotameter is also opened. Close the valve on the dry
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 104 of 225
air (bypass) rotameter to allow the air to become humidified. Allow the system to stabilize for 10
minutes. After preparing the cleanup system, canister cleaning is performed using the following
procedure.
1. Connect the canisters to be cleaned to the cleaning manifolds. Record the canister
numbers and pre-cleanup concentrations, if available, as determined by the last
analysis, in the appropriate cleanup and canister history log book. Record data
pertinent to the vacuum and pressure cleanup cycles as they are completed.
2. Remove collected moisture from the air cryotraps by opening and immediately closing
the air cryotrap purge valves. Removal of the collected moisture should be performed
at the beginning of each pressure cycle, so that the cryotraps do not plug with ice.
3. Release pressure from the canisters by opening all the canister bellows valves and then
opening the manifold pressure release valve. When venting is complete, leave the
canister bellows valves open and close the manifold pressure release valve.
4. Begin the first vacuum cycle by actuating the roughing pump, placing the vacuum source
selector valve in the roughing pump position, and opening the manifold vacuum valve.
5. Evacuate the canisters to approximately 100 mm Hg, as indicated by the absolute
pressure gauge.
6. Position the vacuum source selector valve in the high-vacuum pump position.
7. Evacuate the canisters to 5 mm Hg absolute pressure (or less) and maintain the vacuum
for 30 minutes.
8. Close the manifold vacuum valve after the 30-minute high-vacuum period has been
completed.
9. Begin the first pressure cycle by purging the air cryotraps (refer to Step 2) and then
closing the air bypass valve. Open the manifold air pressure valve. Using the air flow
control valves, adjust the air flow rate to the manufacturer's recommended optimum
flow rate for the oxidizers, as indicated by the air rotameters.
10. Check the pressure regulators to verify that they are set to deliver a final pressure of
20 psig. Fill the canisters to 20 psig. As the final pressure is attained, the flow rates
indicated on the air rotameters will drop to zero, regardless of the setting on the flow
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 105 of 225
controllers because the pressure in the canisters and the pressure at the exit of the
regulators reach equilibrium.
11. Close the manifold air pressure valve when filling is complete. Open the air bypass
valve and adjust the air flowmeters to 1.0 L/min.
12. Release the pressure from the canisters after they have been under a 20 psig pressure
for 30 minutes by opening the manifold pressure release valve.
13. Repeat steps 4, 5, 6, 7, and 8 for Vacuum Cycle 2.
14. Repeat steps 9, 10, 11, and 12 for Pressure Cycle 2.
15. Repeat steps 4, 5, 6, 7, and 8 for Vacuum Cycle 3.
16. Repeat steps 9, 10, and 11 for Pressure Cycle 3.
17. Close all of the bellows valves on the canisters.
2.5.2.3 Canister Blanking Procedure
Prior to initial use, the cleanliness of all canisters should be assessed. After the initial blanking
of 100% of the canisters, the blanking frequency can be reduced. One canister on a cleaning bank of
eight canisters is considered representative and should be blanked. The selection of the canister to be
blanked (from the bank of eight canisters) is determined by selecting the canister with the highest pre-
cleanup TNMOC concentration on the manifold. This canister is selected because the potential for
compound carryover is most likely to be the largest of any of the canisters on the manifold. The blank
sample is analyzed using the PDFID technique as described in Method TO-12 (Appendix C). If this
measurement meets the predetermined cleanliness criterion (i.e., < 10 ppbC), then the other canisters on
the manifold are considered clean. Blanking is a part of the overall canister cleanup procedure, and is
described below.
1. Select the canister to be blanked by referencing the cleanup history logbook to
determine the canister with the highest pre-cleanup TNMOC concentration.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 106 of 225
2. Verify that all the canister bellows valves are closed. Disconnect the canister selected
to be blanked.
3. Using the PDFID technique, if the canister analysis meets the predetermined
concentration criterion (i.e., < 10 ppbC), then the blanked canister and all the other
canisters on the bank of eight canisters are considered clean.
4. If the canister does not meet the cleanliness criterion (i.e., < 10 ppbC), it is reconnected
to the manifold. The entire bank of canisters is given another vacuum and pressure
cycle. After the additional cycle, the same canister is blanked again.
5. After the canister is blanked and has met the concentration acceptance criterion, it is
reconnected to the manifold.
2.5.2.4 Final Canister Evacuation Procedure
After cleaning and blanking, the canisters are ready for final evacuation in preparation for
sample collection. The procedure for final evacuation is described below.
1. Release the pressure from the canisters by opening the manifold pressure release valve
and opening all of the canister bellows valves. When venting is complete, close the
manifold pressure release valve.
2. Begin final evacuation of the canisters by actuating the roughing pump, placing the
vacuum source selector valve in the roughing pump position and opening the manifold
vacuum valve.
3. Evacuate the canisters to approximately 100 mm Hg, as indicated by the absolute
pressure gauge.
4. Activate the turbomolecular vacuum pump, checking to be sure there is liquid cryogen
in the vacuum cryotrap.
5. Switch the vacuum source selector valve to the high-vacuum pump position. Allow the
canisters to evacuate to 5 mm Hg, as indicated by the absolute pressure gauge.
6. Close the canister bellows valves on all of the canisters on the manifold. Close the
manifold vacuum valve.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 107 of 225
7. Disconnect the canisters from the manifold and remove any old identification tags.
Store the cleaned canisters in the designated storage area.
2.5.3 Canister Sampling Issues
The use of canister sampling for collecting and consequently determining concentrations of
VOCs in ambient air is an integral part of the sampling strategy and recommended monitoring
requirements specified in the proposed revisions to 40 CFR Part 58. The technology utilizes stainless
steel canisters with interior surfaces conditioned to minimize surface reactivity. Conditioning allows
stable storage for many of the compounds of interest. Currently, there are two processes used to
condition canister interior surfaces. They are the SUMMA® process and the Silcosteel® process. The
SUMMA® process is a proprietary electroplating treatment that passivates the internal steel surface of
the canister. The Silcosteel® process treats the internal surface by coating it with a thin layer of fused
silica. SUMMA® canisters have been used extensively for the collection of VOC samples since 1983,
and their use is well characterized. Silcosteel® canisters have been used since 1996 and although their
use is not as well characterized as the use of SUMMA® canisters, early evaluation suggests the
Silcosteel® canisters are suitable for use in ambient air sampling. Conditioned stainless steel canisters in
a variety of volumetric sizes are commercially available from several manufacturers.
An important advantage of the canister based methodology is that the collected whole air
sample can be divided into portions for replicate analyses (permitting convenient assessment of
analytical precision) and reanalyses using different analytical systems for specific peak identification and
confirmation. General canister sampling procedures are described in Section 2.5.1.
The presence of high levels of particulate matter can also pose problems in sampling VOCs.
If a filter is used to collect parti culate matter, ozone can interact with the parti culate material trapped on
the filter, resulting in the generation of artifacts. Also, use of a filter such as a 2 ji Teflon® filter on the
inlet to the manifold or on the inlet to the monitoring system means that the filter must be changed
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 108 of 225
frequently (i.e., daily). However, if the monitoring station is near a source of particulate matter (such as
industrial emissions), it may be necessary to use a filter and accept the necessity of more frequent visits
to the monitoring site to change the filter.
2.5.3.1 Precautions in the Use of Canisters
The canister sampling technique is not without potential problems. Primary problem areas
associated with canister sampling include contamination and sample stability. If not controlled, these
problems can significantly reduce the quality and usefulness of the data obtained using the canister
sampling technique. The general discussion and guidance presented below are intended to provide
users with information that should minimize these problems.
2.5.3.2 Contamination
Contamination may cause additional compounds to appear in the sample or increase the
concentrations of compounds present in the ambient air. Contamination may also cause loss of
sampled compounds or may introduce compounds that interfere with gas chromatographic sample
analysis. Contamination can originate from the sample canisters, canister cleanup systems, components
in the sampling systems or analytical system, and improper canister storage practices. These problems
become more significant as analytical sensitivities (detection limits) are lowered.
To minimize collection system contamination, canisters should be purchased from a reputable
supplier who uses high-quality manufacturing and final cleaning procedures. Purchase requirements
should specify contamination-free valves and criteria for maximum residual concentrations of target
compounds. New canisters should be inspected carefully for proper welding and fittings and should
always be blank checked (filled with humidified zero air and analyzed) before use to check for
contamination. Canisters with excessive contamination should be returned to the supplier or cleaned
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 109 of 225
repeatedly until acceptable. Some contaminated canisters may appear uncontaminated immediately
after cleaning but will outgas contaminants upon storage for several weeks. All canisters in routine use
should be blank checked frequently, and particularly after extended periods of storage, to ensure that
significant contamination does not appear.
Canisters used for ambient or low-level measurements should be segregated from those used
for higher-level concentrations or for higher-molecular-weight compounds. Higher-molecular-weight
compounds are more difficult to remove from the internal canister surface. Although the application of
heat during the cleaning process can facilitate the removal of the higher-molecular-weight compounds,
other potential problems may result. For example, if heat is repeatedly applied, it is difficult to maintain
valve integrity. Also, the effects of heat on the SUMMA® treated surfaces have not been clearly
established.
Canister cleanup systems should be constructed of clean, high-quality stainless steel
components, contain suitable cryogenic traps, and be operated systematically and meticulously to avoid
system contamination from vacuum pump oil, poor quality zero air, water used in humidification
systems, room air, or other sources.
Sampling and analytical systems should be constructed of clean, high quality components,
with particular attention paid to pumps, valves, flow controllers, or components having any non-metallic
surface. Before installation and at periodic intervals, samplers should be carefully tested for
contamination or compound loss by analyzing collected samples of zero air and known concentrations
of target compounds. This procedure is termed "certification" and allows the potential contamination
characteristics of each specific sampling system to be assessed. A canister sampling system
certification protocol, as presented in Section 2.5.1.7, should be implemented to ensure that the status
of each sampler is known prior to use.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 110 of 225
Equipment found to be contaminated should be tested further to attempt to identify the source
of the contamination. Contaminated components should be replaced or cleaned, and the system
recertified. Minor contamination can often be reduced by purging the system extensively with
humidified zero air.
The entire measurement system (sampling and analytical) should be checked regularly for
additive and subtractive biases to ensure that measurements obtained are representative. Such
checking involves extensive and continual testing of the analytical systems, sampling systems, sample
canisters, and canister clean up systems. Program checks also involve using humidified zero air and
standards of known concentration, to perform canister blanks and system audits. Collection of samples
from collocated systems and other quality assurance techniques should also be performed.
2.5.3.3 Sample Stability
Sample stability refers to the representativeness of the ambient air sample contained in a
canister after sample collection and storage. For the sample to be stable, the compound matrix and
concentrations of the sample must not change significantly with time. Some of the ways that the
concentration of target compounds in an ambient air sample may change after sampling are:
• Adsorption or desorption on the interior surfaces of the canister or on particulate matter
in the sample from the ambient air;
• Chemical reaction;
• Dilution of the sample with another gas after sampling; and
• Stratification of the sample in the canister.
A number of studies26'27'28'29 have shown that a wide range of VOCs are stable in canisters
for at least 30 days. Most of the reported studies were performed in SUMMA®-treated stainless steel
canisters at pressures above atmospheric pressure. SUMMA® passivation of the interior surfaces of
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 111 of 225
the canisters is designed to passivate the surfaces to minimize catalytic activity on the surface and to
reduce the number and activity of adsorptive sites on the canister's interior walls.
While many compounds have been shown to be stable in canisters, it is not known how these
results extend to the variety of conditions that may be encountered during the use of canisters for
PAMS. These conditions include variable quality of the canisters and their passivation process,
variable moisture content or humidity in the sample air, previous history of use or residual contamination
of the canister, sample pressure in the canister above or below atmospheric pressure, storage
temperatures, and canister age.
Current information indicates that hydrocarbon VOCs with vapor pressures above 0.5 mm
Hg at 25 °C store well in canisters. Substituted hydrocarbons, particularly the halogenated
hydrocarbons with similar vapor pressure properties, also store well in canisters. Laboratory tests
indicate that many oxygenated hydrocarbons such as aldehydes, ketones, and alcohols have limited
storage stability in canisters. Recent studies in aluminum canisters have indicated that the presence of
water in the sample has a great effect on the stability of polar organic compounds.30 Results from
another study confirm that the presence of water is a key factor in ensuring the stability of polar organic
compounds in stainless steel canisters.31 Target analytes for which there is little stability information or
for which storage stability characterization is questionable should be specifically tested for storage
stability in the canisters. These tests should be performed under typical conditions of use.
The potential for physical adsorption as a mechanism for loss of VOCs from the vapor phase
in canisters was evaluated using the Dubinin-Radushkevich isotherm, which provides a specific
relationship between the tendency for adsorption and compound or sample specific properties such as
polarizability, vapor concentration, temperature, and equilibrium vapor pressure.32 A computer-based
model has been developed for predicting adsorption behavior and vapor phase losses for
multicomponent systems. Solely on the basis of physicochemical properties of the compounds (i.e.,
independent of considerations of the properties of the surface), the model predicts displacement of the
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 112 of 225
more volatile VOCs from a canister surface by water vapor at relative humidities in the range of 1 to
20%. This prediction is generally consistent with experimental observations, but in most cases,
experimental conditions (i.e., canister surface properties) are not sufficiently well characterized to permit
detailed quantitative comparison with the model. The model has contributed to the following guidelines
for canister sampling of VOCs:
Relative humidity and temperature should be measured during the sampling process so
that water vapor can be added to the canister prior to analysis if the relative humidity is
low.
Sample pressure in the canister should be as high as possible without causing
precipitation of liquid water within the canister.
When considering the applicability of canister sampling to new compounds, the first
parameters to be evaluated should be chemical reactivity and vapor pressure of the
compound.
Since all species present in the canister participate in the competitive adsorption
process, consideration of the quality of data obtained from multiple canisters at the
same site should include at least semi-quantitative specification (such as total FID
response) of non-target species present in the samples.
2.5.3.4 Positive Pressure Samples
Samples obtained so that the final sample pressure is above atmospheric pressure (typically 5
to 20 psig) are considered positive pressure samples. Positive pressure samples are the least likely to
be affected by the attainment of adsorption equilibrium in the canister after sampling. The only
precaution recommended in this regard is that after sampling, no sample be withdrawn until the sample
has been in the canister for at least 24 hours to allow the adsorption equilibria to stabilize.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 113 of 225
2.5.3.5 Diluted Samples
Samples may be diluted by adding pressurized, clean air, N2, or other gaseous diluent
(Section 2.3.4.3.2). It is recommended that at least 24 hours elapse between dilution of a sample and
removal of an aliquot for analysis.
2.5.3.6 Canister Leakage
There are three potential sources of canister leakage. These sources are:
• Faulty canister welds;
• Leakage at the connection of the valve to the canister; and
• Leakage through the valve.
A faulty weld is a manufacturing defect. Faulty welds are fairly rare and can be detected by conducting
leakage acceptance tests. Canisters may also sustain physical damage. Damaged canisters should be
repaired and retested for leaks.
Leaks at the connection of the valve to the canister are the most troublesome type of leak.
Welding the valve to the canister virtually eliminates such leaks but makes subsequent valve
replacement impractical and expensive. Usually, the valve is connected to the canister using a standard
tubing compression fitting. Properly installed, these fittings are very reliable. However, these fittings
can loosen when an operator improperly opens and closes the valve. If the valve rotates with respect
to the canister during opening and closing, small leaks in this fitting can occur. Overtightening the fitting
in an attempt to prevent such movement exacerbates the problem, as does any other physical strain on
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 114 of 225
the connection. Short of welding the valve to the canister, vulnerability to leakage in this connection can
be greatly reduced by:
• Using an oversize fitting (e.g., 5/16-inch or 3/8-inch rather than 1/4-inch);
• Equipping the canister with a valve guard to protect the valve from physical strain; and
• Mechanically clamping or fastening the valve to the canister or valve guard to prevent
rotation during opening or closing.
These measures are offered by some canister manufacturers and should be specified. Even with these
precautions, periodic retesting of canisters is necessary to ensure that no significant leaks in the valve
connection develop with extended use.
Leaks through the valve can occur if the valve seat has become damaged through wear or
overtightening. The practice of installing a cap on the valve connection when the canister is not
connected to a sampling system effectively minimizes sample or vacuum loss during periods of storage.
A canister may quickly be tested for obvious leaks by pressurizing it with zero air and
submerging it in clean water to look for bubbles. To check for microleaks, the canister should be
evacuated and its pressure observed for several days with a sensitive absolute pressure gauge
connected. This test is performed with the canister valve open. To check the valve for leakage through
the bellows, evacuate the canister, check the absolute pressure, close the valve, disconnect the pressure
gauge, and do not cap the valve inlet fitting. Several days later, reconnect the pressure gauge and
check the pressure. The canister pressure should not increase more than a few mm Hg during that
period.
Canisters with excessive leaks must be repaired and repassivated or replaced, but those with
relatively minor microleaks can be used for many applications if precautions are taken. Canisters
determined to have microleaks can be prepared for use just prior to sample collection and analyzed
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 115 of 225
promptly after sample collection. Reduction of the pre- and post-sampling time reduces the potential
for bias. Between evacuation and analysis, the canister connection should be tightly capped, and the
canister should be stored in a well ventilated, non-contaminated area. Avoid storing the canisters in
automobiles and laboratories where organic materials are used. Storage or shipping containers should
be of an all-metal construction, well ventilated, and should avoid the use of foam padding and other
organic materials.
2.5.4 Sample Analysis
The methodology in Section 2.5.1 describes the collection of whole air samples into
SUMMA® canisters. The canisters are then sent to a central laboratory location for sample analysis.
When incorporating manual methods, the user must develop procedures to perform sample analysis.
Those using manual methods for sample collection may also incorporate the use of automated GC/FID
systems as described in Section 2.4 for sample analysis. If a manual sample analysis system is used, it
should incorporate the same basic components as discussed in the automated method which include a
sample introduction system, sample conditioning system (for moisture removal), sample concentration
system for sample enrichment, an optional cryo-focusing trap, a gas chromatograph which houses the
appropriate analytical column(s) and FID(s), and a data acquisition and processing system. Capillary
GC/FID is the recommended analytical system for PAMS but individual sites may use GC/MS or a gas
chromatograph with multiple detectors (i.e., both FID and MS).
2.5.4.1 Sample Introduction
The air sample is introduced to the primary sample concentration trap of the GC system
directly from a SUMMA® canister using a mass flow controller or other flow control device at a
constant flow rate. Samples may also be introduced for purposes of calibration and proficiency studies
directly from pressurized canisters.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 116 of 225
2.5.4.2 Sample Conditioning
Sample moisture must be removed from the sample stream to prevent or reduce the effects of
moisture on the primary concentration trap, analytical column(s), and detector(s) as described in
Section 2.3.4. Moisture removal allows for analysis of larger sample volumes, which provides lower
detection limits, and is crucial to the measurement of very low concentration VOCs.
Moisture may be removed from the sample stream using a Nafion® membrane sample
conditioning device. Some commercially available concentration systems incorporate the use of
Nafion® sample drying devices. New developments in this area using controlled temperature
desorption, selective temperature condensation, hydrophobic concentration traps, etc., are currently
being studied (see Section 2.3.4). The loss of polar VOCs may result from sample conditioning and
significantly affect the TNMOC measurement. The user must characterize the effects of the sample
conditioning method chosen on the TNMOC measurement and the target VOCs of interest.
Chromatographic retention times are changed by the presence of water vapor. Also, any MS analysis
is subject to response variability because of filament changes caused by the presence of water vapor.
In order to characterize effects, primary calibration and/or retention time standards can be analyzed
with and without the conditioning device. If a commercially available analytical system with sample
conditioning is used, vendors should provide information regarding the effects of their conditioning or
drying device.
Preconcentration techniques in use for sampling of atmospheric VOCs are susceptible to
reactions occurring during sampling involving the organic compounds of interest and other reactive
constituents of the air. The most widely recognized and significant interference results from reactions
with ozone. Chemical reactions in or on the concentration media may alter the quantities of VOCs and
may also contribute to the formation of artifacts which may be interpreted as atmospheric constituents.
If a solid sorbent is used for preconcentration, reactive constituents of the atmosphere may also react
with the solid sorbent bed to produce artifacts. In particular, unsaturated hydrocarbons such as
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 117 of 225
isoprene and monoterpenes can be depleted during cryogenic concentration in the presence of ozone.33
In sampling trace organic gases onto solid sorbents such as Tenax® TA, unsaturated compounds such
as styrene, cyclohexene, and monoterpenes were found to react with ozone during ambient sampling
with loss of anaiyte34'35'36'37'38'39'40'41 and formation of oxidized products.41'42 Artifacts such as
benzaldehyde, phenol, and acetophenone have been found to result from direct reaction of ozone with
solid sorbents, especially Tenax® XA.34'42'43'44'45'46'47 Reactions with oxidants can be reduced or
eliminated by selectively removing the oxidant from the sample flow prior to concentration of the
organic trace gases of interest. Although numerous options are available for either chemically or
physically scrubbing ozone,48 none of these options for ozone scrubbing are sufficiently or thoroughly
characterized enough for VOC sampling applications to ensure that the concentrations of the trace
organic compounds are not affected by use of the ozone scrubber. There is, therefore, no generally
accepted ozone scrubber for VOC sampling presently in use. Ozone scrubbing options for carbonyl
sampling are discussed in Section 5 of this document.
2.5.4.3 Sample Concentration
Ambient air samples are primarily concentrated using multi-bed sorbent or deactivated glass
bead traps. Sampling time and flow rate are used to determine the total volume concentrated onto the
primary trap. Multi-bed sorbent traps (Carbotrap® and Carbosieve®) are required to efficiently collect
the complete range (C2 through C12) of VOCs for O3 precursor monitoring.13
Samples are collected onto sorbent traps at sub-ambient or ambient temperatures to improve
collection efficiency. Ideally, sorbent traps selectively adsorb only the trace VOCs and do not interact
with the atmospheric constituents or introduce any contaminants into the system. Sorbent traps may
also be designed to eliminate water vapor by using hydrophobic sorbent materials.
Sample concentration using glass bead traps, which require concentrating temperatures
around -185 ° C, are cooled using liquid N2, CO2, or Ar. This process is commonly known as
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 118 of 225
cryogenic concentration or cryotrapping, and is the oldest and best known of the techniques for
collecting C2 through C12 VOCs. The glass beads provide surface area for collection of the VOCs at
the cryogenic trapping temperature. The most prevalent system being used in the PAMS uses Peltier
cooling of the concentrator to eliminate the use of liquid cryogens and make unattended automated
monitoring feasible.49
2.5.4.4 Sample Focusing or Cryofocusing
Following sample collection and concentration, the sample is thermally desorbed either
directly onto the analytical column(s) or onto a secondary cryofocusing trap. For concentrated samples
that are desorbed directly onto the analytical column(s) to separate the VOCs, the analytical column is
cryogenically cooled to aid in re-focusing the desorbed sample into a narrower band prior to
chromatographic separation.
Concentrated samples are optionally desorbed onto a cryofocusing trap to focus the
desorbed sample into a narrow "plug" for subsequent thermal desorption and injection onto the
analytical column. The sample cryofocusing step is optional. Cryofocusing improves the peak
separation and particularly the resolution of C2 and C3 hydrocarbons, and is especially helpful when the
sample is desorbed from the concentration trap at low flow rates. Cryofocusing traps incorporate the
use of fused silica tubing that is cooled using liquid cryogen. The fused silica tubing is wide-bore (0.32-
mm ID.) or megabore (0.53-mm ID.) deactivated fused silica tubing that is cooled to approximately -
185 °C. These traps may be packed with glass beads to increase the surface area and improve the
focusing of the sample band. Closed cycle coolers are available and have been used in field tests.17'18
2.5.4.5 Gas Chromatography
The gas chromatograph contains the analytical column(s) of choice for PAMS VOC analysis.
Refer to Sections 2.3.5 and 2.3.6 for guidance on column configuration and selection. The user must
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 119 of 225
configure the GC system to meet the enhanced O3 monitoring requirements and specifications, and must
also characterize the system operation prior to use. Gas chromatographic systems may incorporate the
use of single or dual-column configurations (in series or parallel) and may require sub-ambient oven
temperature programs. It is important to note that implementing systems that eliminate the need for
sub-ambient column oven temperatures will reduce the overall cryogen consumption of the system.
Gas chromatographic systems employ the use of a PC-based data acquisition and processing
system for peak integration and quantitation. Data acquisition and processing systems are comprised of
hardware and software that perform data acquisition, peak detection and integration, peak identification
by retention time, post-run calculations and quantitation, calibration, peak reintegration, user program
interfacing, and hard copy output. Data are automatically stored on magnetic media (e.g., hard disk or
floppy diskette).
The GC acquisition and processing software is typically developed and supplied by the GC
manufacturer and should contain the necessary algorithms to acquire, integrate, and identify the
chromatographic peaks by retention time. The system should be capable of producing an electronic
and hard copy report file that contains the information needed to identify the sample, and a listing of all
peaks detected in the chromatogram. This listing should contain the peak name if the peak is a target
compound. All detected target and non-target peaks should be reported with the associated
concentration in ppbC, and a retention time. The listing should also contain the TNMOC and PAMHC
estimates as calculated by summing the concentrations of peaks as described in Section 2.2. See
Section 2.6.1 for a more detailed discussion on data processing capabilities of automated GC systems.
2.5.4.6 Analytical System Calibration
The detector response of the analytical system should initially be calibrated with multiple level
propane standards over the expected sample concentration range. Benzene is suggested as a second
primary standard to address the needs of dual-column systems that employ the use of column switching
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 120 of 225
techniques. The primary calibration standard is used to generate a per Carbon response factor for
determining the concentration of each target VOC, as well as the TNMOC. It is unnecessary to
determine compound specific response factors for each of the target VOCs presented in Table 2-1,
because the Carbon response of the FID to these compounds is approximately linear. It is appropriate
to measure each compound concentration in terms of ppbC using the relative response factor
determined from the primary standard. Refer to Section 2.4.2.3 for a more detailed discussion of
analytical system calibration.
2.5.5 System Operation
This section provides guidance and general operating considerations for initial system set-up,
optimization of sampling parameters, and field operation for the GC system.
2.5.5.1 Initial System Set-up
During the initial set-up of the system, several parameters must be evaluated to optimize the
system operating conditions. Critical parameters include, but are not limited to, the sample collection
flow rate and sample integration time, sample concentration and desorption conditions, oven
temperature program parameters, detector calibration, and the peak detection and integration methods
used by the data acquisition and processing system. These parameters are optimized by varying the
operating conditions to achieve the best resolution and detection of the target VOCs using primary
calibration and retention time calibration standards.
Prior to making VOC measurements, the baseline performance of the system must be
thoroughly documented. The information from the system baseline characterization is used to determine
system specific target compound retention times, relative retention times, identification of co-eluting
compounds and matrix effects, internal standard retention times and interferences, and detection limits.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 121 of 225
Baseline characterization is a very important step in determining the overall quality and validity of the
measurement data. See Section 2.3.7 for a discussion of pre-measurement system characterization.
The system should initially be set up and tested for conformance with the manufacturer's
operational specifications. Under terms of agreement for purchase, the manufacturer should be
required to provide a detailed instruction manual for system operation and be required to provide initial
system setup, user training, and demonstration of adequate system performance. See System
Specifications for automated GC systems in Section 2.4.5 as guidance for procurement.
2.5.5.2 Sampling Parameters
Determination of optimum sampling conditions is dependent on field conditions (i.e., expected
compound concentration ranges, humidity, temperature, etc.), desired sensitivity (detection limit),
cryogen consumption, and sample trapping efficiency. During the setup period, these sampling
parameters should be evaluated to determine the optimum conditions. The primary sampling parameters
are the sample collection frequency and the sample collection or integration time.
2.5.5.3 Operation
The system should be operated in accordance with an SOP that is prepared by the user
based on the information obtained during the setup and familiarization period. The system should be
maintained by a qualified operator who should perform the routine operational and critical quality
control functions as specified in the SOPs. Operational parameters should be adjusted, if necessary, so
that the data quality objectives are met. Primary calibration checks should be performed at least once
each week or at an interval consistent with the individual site standard operating procedure. Retention
time calibration checks should be performed at least once per week or at an interval consistent with the
individual site standard operating procedure to provide retention time reference information for
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 122 of 225
validating compound identifications. The retention time calibration standard can also be used to track
the FID response and compound recovery to determine when recalibration is necessary.
Detailed SOPs for operation of the system must be developed. The SOP should be based
on information obtained during the set-up and familiarization period and the requirements of the
monitoring program. Refer to QA/QC Section 2.8.3.1 for a more detailed discussion of SOP
development.
2.6 Data Processing Capabilities of Automated VOC Systems and Submittal of
VOC Data to the AIRS AQS Data Base
As prescribed in Section 58.45 of 40 CFR Part 58, PAMS Data Submittal, the data from
VOC measurement systems is required to be submitted to EPA's Aerometric Information Retrieval
System (AIRS) Air Quality Subsystem (AQS) within 6 months following the end of each quarterly
reporting period. The agencies may ultimately use the data to make attainment/ nonattainment
decisions, track VOC emission inventory reductions, provide input to photochemical models, prepare
air quality trends, and characterize population exposure to O3. The data that are ultimately submitted to
AIRS must be uniform across all networks and consistent with enhanced O3 monitoring requirements.
The overall approach to generating data from automated GC systems involves the same basic
components (A/D converter, personal computer, acquisition and processing software, and data storage
module). These basic components comprise the automated GC data acquisition and processing
system. The VOC concentration data are ultimately generated in an electronic file and hard copy
format specific to the data acquisition and processing system used. The information generated must
finally be put in the required format for submittal to the AIRS data base.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 123 of 225
This section discusses how data acquisition and processing occurs, the data acquisition and
processing capabilities of automated VOC systems and guidance on formatting and submitting the data
to the AIRS data base.
2.6.1 Data Processing Capabilities of Automated VOC Measurement Systems
Automated VOC systems employ a variety of hardware and software strategies to collect
and process chromatographic data. Despite these differences, their overall approach to processing
data involves the same basic system components and procedures. This section provides background
and important details on automated VOC data processing.
The basic components of a typical data acquisition and processing system are an analog
signal generated by the chromatographic detector, an Analog/Digital (A/D) converter, a personal
computer for acquiring/processing data, and a data storage module, where data are stored on magnetic
media. After a sample is injected, the sample components are chromatographically separated and sent
to the detector. The detector electronics respond to the amount of the analyte passing through the
detector and generate an electronic signal which will be either analog or digital. Digital signals are sent
directly to the next step in processing; analog signals must first be digitally converted before further
processing can occur. Digitized signals are "filtered" by specified method acquisition parameters and
the resulting data stored to a file in the data acquisition system.
After the analysis is complete, the post-run method parameters further refine the raw data file
by applying integration algorithms and parameters, compound identification requirements, calibration
response factors and data reporting format to generate a processed data file. Processed data files
generally contain retention time data, peak height, peak area, and concentration data: all the elements
needed to compile a data report in a specified format. The entire process from signal generation is
repeated for each sample during automated operation. This report generating process is depicted in the
flowchart in Figure 2-16 and discussed in greater detail in Section 2.6.1.1.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 124 of 225
Processed
File Generated
Detect peaks
Establish baseline
Separate peaks
Integrate peaks
f RT
Identify < RRT
I RT window
Calculate concentration
Generate report
Chromatogram
To Database
Application
Electronic Transfer Format
o/s/g/m/3797pams/jr-fig2.ppt
Figure 2-16. Report Generating Process
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 125 of 225
2.6.1.1 Data Acquisition and Processing
This section presents a more detailed explanation of the steps involved in data acquisition and
processing.
Signal Output from Detector—An electronic signal, usually an analog voltage signal, is
produced by the detector during a chromatographic analysis and is measured by the detector
electrometer. The amplitude of the signal is related to the concentration of the sample. For the FID
response to hydrocarbon compounds, the relationship is a direct proportion. Depending on the type of
detector, the signal from the electrometer will be amplified before it is processed further. The
continuously varying analog signal must be converted to discrete bits for a computer data system to
process the data. These discrete bits are considered a "digital" signal. Digital signals can be
transmitted directly from the output terminal to the next step in the process.
A/D Conversion—Computer systems handling data for chromatographic systems require
the data to be in digitized form prior to storage. Analog-to-Digital converters, also known as A/D
converters, take the continuously variable analog signal and "cut" it into discrete slices to make a digital
signal for computer-based storage. These slices represent the detector response during a specific slice
in time. The sum of these signals is proportional to the sample component concentration. These digital
slices are the fundamental data stored for subsequent data processing.
Signal Filtering - Method Acquisition Parameters—To acquire data, the analytical
system must be given a data acquisition method. This method tells the instrument how to collect the
data and which data are important. The method controls the instrument by starting and stopping data
collection and by controlling how frequently the data processor takes data slices, which is called the
sampling rate. The peak threshold for data collection, or the minimum signal amplitude above
background, is also specified. The signals collected using this set of conditions are referred to as raw
data. As raw data are acquired, the data are kept in the CPU memory until acquisition is complete.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 126 of 225
Raw Data Storage—After sample collection is complete, the raw data are written to an
electronic storage module (e.g., computer disk) as a discrete data file. The raw data reside in memory
(CPU) until the PC controlling data acquisition is available to receive the data. Some systems have
capacity within memory to temporarily hold multiple raw data files and then to download the files to the
PC storage module at specified intervals. These systems must send data to the PC before memory is
exceeded or data may be lost. Once the data files are permanently stored to disk they can be backed
up and archived in long term storage.
Post-run Data Interpretation/Processed File Generation—Analytical methods contain
post-run instructions for interpreting raw data files. Accurate measurement of ozone precursors
depends heavily on accurate and consistent chromatographic data. Analytical methods should be
tailored to enhance the accuracy of output data. The term "peaks" will be used to describe the area
underneath the signal curve. In this section, the basic concepts and issues that the data system must
manage are presented. This information should be useful to operators setting initial conditions for post-
run data processing for their particular equipment. Information presented here should also help direct
corrective action when problems are identified with processed chromatographic data. Post-run
routines accomplish the following tasks:
• Peak detection;
• Baseline determination;
• Peak separation and integration;
• Timed events application - controlling peak detection and integration parameters at
preset times;
• Peak identification;
• Concentration calculation;
• Report production; and
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 127 of 225
Chromatogram generation.
Peaks are detected by applying criteria for discriminating changes in signal associated with
compound detection from the background signal. Certain parameters, such as minimum peak width
and signal sampling rate, shape the chromatogram during acquisition by affecting the amount of detail.
A higher sampling rate and smaller peak width will enhance the fine detail of the chromatogram. In
contrast, detail will be lost with smaller sampling rates and greater peak widths. Since detection
algorithms commonly monitor the rate of change of the signal - looking at slope changes and the rate of
slope change - to determine the start, apex and end of a peak, the number of points defining a peak is
very important for accurate detection. Some techniques, such as peak bunching, can be used to
"smooth" peaks by averaging adjacent signal slices and, therefore, screening unwanted noise.
Excessive bunching will also cause loss of chromatographic detail, biasing quantitation high for some
compounds and completely eliminating others.
To integrate detected peaks, the data system determines how closely-eluting peaks are to be
separated and draws a baseline accordingly. Well-separated peaks require simple integration treatment
- either drawing a baseline from the start of one peak to the start of the next or separating peaks by
adding drop lines to the baseline. For overlapping peaks, most systems have criteria based on peak
proximity and size comparisons that allow reasonable quantitation. "Timed events" can be used to
tailor integration schemes to a particular chromatogram or set of chromatographic conditions by
changing integration parameters to be modified according to a user-defined timetable. A timed events
table queues changes in integration and detection factors to enhance data accuracy. These
modifications can help compensate for changes in peak characteristics, such as peak width, peak
shape, and incomplete separation, resulting in more accurate and consistent data treatment.
Peak identification is based on retention time (RT) windows or retention time relative to a
reference peak or internal standard.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 128 of 225
Concentration calculations for most automated VOC systems use external standard-based
quantitation. A calibration curve is generated for the primary calibration compounds and each response
factor from the curve used for quantitation of the C2 through C12 VOCs. For unidentified peaks, the
primary response factor for propane is used to estimate concentration in ppbC.
Report Production/Chromatogram Generation—Many data systems allow great
flexibility with reporting format, allowing the user to custom-build hardcopy reports. The hardcopy can
be generated directly by the data system software or through a third-party data package. The
information compiled in the report includes, but is not limited to, retention time, peak area, type of
integration, and limits of concentration. The user can automate post-run data handling to save the
processed data file in a specified electronic format to facilitate electronic data file transfer to a
spreadsheet or data base.
Reprocessing Capabilities—Chromatographic data systems generally allow the user to
change and re-apply data interpretation parameters to adapt them to a given set of chromatographic
conditions. Frequently, changes in peak width, bunching factors, integration approach, and
identification parameters are needed to compensate for changing sample conditions, flow rates, and
column degradation (over time), and to generate more consistent accurate data. The data can also be
reprocessed to apply different calibration response factors as needed.
Data Format Options—Data can be presented in two forms - hardcopy report or
electronic data file. While hardcopy reports facilitate review, electronic data files allow the user to
manipulate results to a desired format and to transfer information quickly. The electronic data formats
available as part of the data system package vary from vendor to vendor. Most systems are capable of
providing ASCII output files; others can produce comma-delimited files or AIA format (Analytical
Instrument Association's Chromatography Data standard format), as either a routine analysis file or a
result from a post-run program. This file then needs to be reformatted for submittal to AIRS as
discussed in Section 2.6.2.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 129 of 225
Telemetry—Telemetry is defined as remote interaction with the PAMS analytical system,
including communication, operation, and data transfer. Telemetric capabilities may be a part of the
analytical system as purchased, or these capabilities may be acquired by purchase of a commercial
software package and any required associated hardware. At the very least, telemetry can be used to
communicate directly with the analytical system to assess or modify the operation, as well as to transmit
data files to a centrally-located host computer for further processing. Other potential telemetric
capabilities may include remote operation of devices at the PAMS site, such as a standards injection
system, or to perform a system malfunction auto-call should the system recognize that it has a problem.
2.6.2 AIRS AQS Data Submittal
This section provides guidance for the submittal of O3 precursor monitoring data into the Air
Quality Data Storage Subsystem of AIRS, which is a computer-based system for handling the storage
and retrieval of information pertaining to airborne pollutants. AIRS AQS is administered by the U.S.
Environmental Protection Agency, Office of Air Quality Planning and Standards, Information Transfer
and Program Integration Division (ITPID), Information Management Group. The AIRS Air Quality
Subsystem (AQS) contains data submitted by state and local agencies, and EPA reporting
organizations. These data correspond to the data previously handled by the SAROAD system. AIRS
AQS includes descriptions of air monitoring sites and monitoring equipment, measured concentrations
of air pollutants and related parameters, and calculated summary and statistical information. AIRS
AQS is currently being re-engineered, but no information regarding operation of a re-engineered AIRS
AQS system is presently available.
All VOC and carbonyl data collected for the enhanced O3 monitoring network must be
submitted to AIRS AQS in the form of transactions. A transaction provides the monitoring information
for a particular parameter (compound) for a given day. Because the VOC target list includes over 50
compounds, over 50 line transactions are necessary to input the results from one VOC sample.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 130 of 225
Because the carbonyl target list includes three compounds, three line transactions are necessary to input
the results from one carbonyl sample. Data can be input on-line one transaction at a time or in groups.
Transactions are used to provide data and control information for updating the AIRS data
base. Table 2-5 shows how the various transaction types are related.
Although transactions can be input into AIRS on-line, batch processing is recommended
considering the volume of data generated to meet the minimum network monitoring requirements. Prior
to submittal to AIRS, all data must be formatted to conform to the 80-character per line format as
required by AIRS and saved as an ASCII file. A group of transactions (e.g., data from one week) can
be saved as one file and loaded together into AIRS.
Within a transaction are fields corresponding to specific information. A field or "slot" is a
column or set of columns in the 80-character input line where a specific piece of data is placed. Four
general types of values are used to code the transactions: codes, dates, numeric data, and
alphanumeric data. Codes must be entered exactly as they are stored in the Geo-Common Subsystem
of AIRS. For example, a county code is three digits and all three digits of the code must be entered,
including any leading zeros. Dates are entered in the YYMMDD format, where YY is the last two
digits of the year, MM is the month number, and DD is the day of the month. Each part is two digits
and any leading zeros must be included. Numeric values must be entered right-justified and do not
have to include leading zeros. Alphanumeric values should be entered left-justified in the field.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 131 of 225
Table 2-5. AIRS Transaction Types
Transaction Type
Site(Al, A2, A3, A4, A5,
A6, A7)
Monitor (F1,F2,F3,F4,F5)
SLAMS (M, N, P, R, S, T, U,
V)
Related Transaction Records Have The Same
Site-ID (state, county, site codes)
Monitor-ID (site-ID, parameter, and POC codes)
Monitor-ID, year, transaction type (any of codes M through V)
Raw Data:
Hourly (1)
Daily (2)
Composite (3)
Monitor-ID, year, month, transaction type 1
AND type 4 transactions with the same monitor-ID
Monitor-ID, year, month, transaction type 2
AND type 4 transactions with the same monitor-ID
Monitor-ID, year, period, transaction type 3
Special:
Minimum Detectable Value
(Z)
Null Value (4)
Precision/ Accuracy (8, 9)
Monitor-ID, transaction type Z
Monitor-ID, transaction type 4
Monitor-ID, year, month, transaction type
Each field occupies a specific location in the 80-character per line format. For example, the
two digit code that identifies one of the 50 states, U.S. territories, or Washington, D.C., is called the
State Code and is placed in columns 2-3 of the 80-character per line format. There are also
international data, i.e., the code for Mexico is 80. The structure and placement of the fields in each
transaction is specific and must be carefully followed.
All data collected must be entered into AIRS. Quantifiable data which are below the
detection limits should be entered as the quantified value. The raw data values will be retained within
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 132 of 225
AIRS, but when any summary statistics are generated the system will automatically replace values
below the detection limits with a value that is one-half the detection limit prior to performing
calculations. Raw data listings will maintain the actual values. If the compound response is below the
detection limit and the data are not quantifiable, a sample value of zero is entered into AIRS. This value
will indicate to AIRS that the compound of interest was analyzed for but not present at a quantifiable
level. If a sample is missed or invalidated for any reason, appropriate null values or validity flags are
used.
Although sampling is conducted in accordance with 40 CFR Part 58 on local time, data are
reported to the AIRS in local standard time.
The following sections illustrate the specific formats required for the various types of
transactions.
2.6.2.1 Initial AIRS AQS Setup
A valid Time Sharing Option National Computing Center (TSO NCC) UserlD and
Password must be established prior to the submittal of any monitoring data into AIRS. UserlDs can be
obtained by contacting the appropriate AIRS AQS Regional Coordinator. A list of the current
coordinators is provided in Table 2-6.
The Air Quality (AQS) data storage processes operate on the input data transactions while
they reside in a "screening file." If data are resident on a PC, the file(s) must first be transferred to a
TSO dataset and then to an AIRS screening file. The screening file acts as a submittal buffer that
allows entry and review of data before inclusion into AIRS. Each data storage user group of AIRS has
a screening file for that organization's transactions. The records of an input transaction file are first
loaded into a screening file where they remain while the authorized users examine and modify them
through the AQS data storage processes. The transactions in the screening file are edited and
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 133 of 225
corrected by the user and examined by the AIRS Data Base Administrator prior to inclusion in AIRS.
Screening files should exist for state agencies. When the UserlD is established, a request should be
made to the AIRS AQS Regional Coordinator for access to the appropriate screening file.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 134 of 225
Table 2-6. Current AIRS AQS Regional Coordinators
Region I
Ms. Wendy McDougall
U.S. EPA, Region I
60 Westview Street
Lexington, MA 4323
(617) 860-4323
Region VI
Ms. Ruth Tatom
U.S. EPA, Region VI
1445 Ross Avenue
Dallas, TX 75202-2733
(214) 655-8355
Region II
Mr. Ed Finfer
U.S. EPA, Region II
290 Broadway - 20th Floor
New York, NY 10007-1866
(212) 637-4244
Region VII
Ms. Joyce Sousley
U.S. EPA, Region VII
25 Funston Road
Kansas City, KS 66115
(913)551-5050
Region III
Ms. Catherine Brown
U.S. EPA, Region III
841 Chestnut Street
Philadelphia, PA 19107
(215) 597-6149
Region VIII
Mr. Dale Wells
U.S. EPA, Region VIII (8ART-TO)
999 18th Street, Suite 500
Denver, CO 80202-2405
(303) 293-0967
Region IV
Region IX
Mr. Darren Palmer
U.S. EPA, Region IV
345 Courtland St., NE
Atlanta, GA 30365
(404) 347-3555 x4184
Mr. Jim Forrest
U.S. EPA, Region IX
75 Hawthorne Street
San Francisco, CA 94105
(415) 744-1291
Region V
Region X
Mr. William Damico
U.S. EPA, Region V (SQ-14J)
77 W. Jackson Boulevard
Chicago, IL 60604-3590
(312)353-8207
Mr. Bill Puckett
U.S. EPA, Region X (ES-095)
1200 6th Street
Seattle, WA 98101
(206) 553-1702
Minimum detectable limit values in AIRS may be specified for each of the PAMS target
compounds by the reporting agency. If no specific value is enumerated by the reporting agency, AIRS
will utilize a default value defined in the Geo-Common File for the parameter, generally 0.1 ppbC.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 135 of 225
2.6.2.2 Site and Monitor File Updates
Before any VOC or carbonyl data are submitted to AIRS AQS, transactions must be
submitted to update the monitor information in AIRS. These transactions, called Type F transactions,
provide information concerning the compounds of interest, the organizations involved, and the relative
monitor dates. Column 1 is the transaction type code. For the Type F transactions, the entry in
Column 1 is "F."
The monitor ID is placed in columns 2-16 and is comprised of the State, County, and Site
Code (columns 2-10) where the sample was collected, the parameter (columns 11-15), and Parameter
of Occurrence (POC) code (column 16). The POC code is used to distinguish between different
monitors at the same site that are measuring the same parameter. The first monitor at the site is
generally identified with POC = 1, with additional monitors at the same site measuring the same
compound identified with POC = 2, etc. If a new instrument were installed to replace the original
instrument used as the first monitor, that would be the same monitor and it would have POC = 1, even
if the sampling method or interval were changed.
The monitor type is indicated in column 17. For PAMS Meteorological and VOC sites, the
monitor type is "3," indicating other. Once the monitor information has been submitted to the AIRS
AQS, the Regional Coordinator should be contacted. The Regional Coordinator will then request that
the network be approved as a PAMS monitor.
The date when the monitor type became effective is recorded in columns 18-23. The date is
recorded in the YYMMDD format. The year, month, and day are each two digits and any leading
zeros must be included.
The next 9 columns indicate the organizations associated with analysis and collection of the
sample and reporting the subsequent data. Columns 24-26 are reserved for the code indicating the
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 136 of 225
analysis laboratory. Columns 27-29 indicate the collection laboratory. The reporting organization code
is stored in columns 30-32. These codes can be found in the AIRS Geo-Common Subsystem and
must be entered exactly as they appear in that subsystem.
The date when the reporting organization became effective is recorded in columns 33-38.
The date is recorded in the YYMMDD format. The year, month, and day are each two digits and any
leading zeros must be included.
The date when the sampling program began is recorded in columns 39-44. The date is
recorded in the YYMMDD format. The year, month, and day are each two digits and any leading
zeros must be included.
The 4-digit code indicating the urban area represented by the site is placed in columns 72-
75. The urban area codes can be found in the AIRS Geo-Common Subsystem and must be entered
exactly as in the subsystem.
The PAMS required sampling frequency code is placed in column 76. Table 2-7 presents
possible codes for the different sampling frequency requirements.
The transaction ID is placed in column 79. For a Type Fl transaction, the ID is "1."
The action code placed in column 80 indicates the type of data base processing to be
performed by this transaction. In order to insert new values into the database, "F is placed in column
80. Should modification to existing monitor information be necessary (e.g., a change in the reporting
organization, etc.), an "M" is placed in column 80. The deletion of existing monitor information is
indicated with a "D" in column 80.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 137 of 225
Table 2-7. AIRS Sampling Frequency Codes
Code
A
B
C
D
E
G
H
I
J
K
L
M
N
O
P
Sampling Frequency
Daily
Daily
Daily
Daily
Daily
Every
Every
Every
Every
Every
Every
Every
Every
Every
Every
24
83
1-Hr
Samples - PAMS
-Hr Samples
1 3-Hr Sample -
124-Hr
Sample
4 6-Hr Samples
3rd
3rd
3rd
3rd
3rd
3rd
6th
6th
6th
6th
Day
Day
Day
Day
Day
Day
Day
Day
Day
Day
-PAMS
PAMS
-PAMS
-PAMS
24 1-Hr Samples
83-Hr
1 3-Hr
Samples -
-PAMS
PAMS
Sample - PAMS
1 24-Hr Sample -
46-Hr
43-Hr
Samples -
Samples -
24 1-Hr Samples
83-Hr
13-Hr
Samples -
PAMS
PAMS
PAMS
-PAMS
PAMS
Sample - PAMS
1 24-Hr Sample -
PAMS
One Type Fl transaction is necessary for each monitor. Unique Type Fl transactions need
to be submitted only once prior to the submittal of monitoring data. Should any of the information
change, the Type Fl transactions must be modified reflecting the changes. If Type Fl transactions
have already been submitted and no modifications are necessary, it is not necessary to submit Type Fl
transactions again before the raw data are submitted.
Figure 2-17 shows the column placement of each of the fields for a Type Fl transaction.
The "F" in column 1 (transaction type) and the "1" in column 79 (transaction ID) denote the Type Fl
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 13 8 of 225
transaction. Below the figure is an example Type Fl transaction. For this example, site 25-009-2006
(columns 2-10) began monitoring for isoprene (indicated by 43243 in
-------
Monitor ID
F2500920064324313900601001001001940601940601
1120P II
o:/s/g/morr/3797/pams/hobson/graph2.ppt
Figure 2-17. Column Placement for a Type F1 Transaction
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 140 of 225
columns 11-15) on June 1, 1994 (indicated by 940601 in columns 18-23). The state agency (001)
analyzed and collected the samples, and reported the results. This site was in the Boston urban area
(as indicated by 1120 in columns 72-75). The required sampling frequency was one 24-hour sample
every sixth day (indicated by P in column 76). Because isoprene previously had not been monitored at
this site, this transaction was inserted into AIRS (indicated by "F in column 80). Had it been necessary
to modify information, a "M" would have been placed in column 80. It is important to note the column
placement of the codes in the transaction file. For example, because it is not required to input a
dominant source, column 64 is left blank. The second transaction indicates that the same site began
monitoring for formaldehyde (parameter 43502) on June 1, 1994, as well. Many of the necessary
codes are provided in subsequent sections of this document. These and other codes can also be found
in the Geo-Common Subsystem of AIRS.
In addition to Type Fl transactions, comments need to be added to the site information to
identify the configuration of the sampling/analysis equipment used. Comments can be added to the site
cards by submitting Type A6 (or A7) cards. Type A6 transactions add a comment to the Line 1
comment line, whereas the Type A7 transactions add a comment to the Line 2 comment line. If a
comment does not appear on Line 1, submit a Type A6 transaction. If a comment currently appears on
Line 1 and no comment appears on Line 2, submit a Type A7 transaction. If comments appear on both
lines, submit either transaction where the comment includes the original comment plus the comment
concerning the configuration of the sampling/ analysis equipment. If additional comment space is
needed, the Line 2 comment may be used as a continuation of Line 1. A list of possible configuration
comments appears in Table 2-8. Because of the limited space within the comment lines, it is important
to make the comments concise yet as informative as possible. Choose the combinations of codes that
best reflect the configuration of the sampling/analysis equipment.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 141 of 225
Table 2-8. Configuration Comments for Type A6 or A7 Transactions
Monitoring
voc
Carbonyl
Equipment
GC
(Choose one)
Dryer
(Choose one)
Concentrator
(Choose one)
Columns
Ozone
Scrubber
(Choose one)
Code
Single FID
Dual FID
W/Dryer
W/O Dryer
Sorbent
Cryo
BP1
GS-Q
RTx
Al203/Na2SO
Denuder
Cartridge
Description
GC configured with one FID
GC configured with two FIDs
Dryer is in line
No dryer is used
Sorbent sample concentrator
Cryogenic sample concentrator
SGE, Inc. BP1 (0.22 mm diameter, 50 m
length)
J&W® GS-Q (0.53 mm diameter, 30 m length)
Restek® RTx-502.2 (0.53 mm diameter, 25 m
length)
Hewlett Packard® Al2O3/Na2SO4 (0.32 mm
diameter, 50 m length)
KI Denuder
Sep-Pak® KI Cartridge
Column 1 is the transaction type and for Type A6 (or A7) the transaction type is "A."
Columns 2-10 are the State, County, and Site Code where the sample is collected. The
comment will be added to the Site File for this site.
Columns 11-78 are reserved for the site comment. If a Type A6 transaction is inserted, the
comment will appear on Line 1. A Type A7 transaction will place a comment on Line 2.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 142 of 225
Column 79 is the transaction ID. Type A6 transactions require a "6" in column 79, whereas
Type A7 transactions require a "7" in column 80. If a Type A6 transaction is submitted, a comment is
placed on Line 1. A comment is placed on Line 2 if a Type A7 transaction is submitted.
The action code placed in column 80 indicates the type of data base processing to be
performed by this transaction. In order to insert a new comment into the site files, an "F is placed in
column 80. Should modifications to existing values in the site files be necessary, a "M" is placed in
column 80 to indicate a modify transaction. For a modify transaction, any alteration to the current site
comments will completely replace that comment. Therefore, any change will require that the entire field
be re-entered. Deletions are not valid for Type A6 (or A7) transactions. To delete existing site
comments, fill columns 11-78 with asterisks on a modify transaction. Be sure to fill all columns with
asterisks or only asterisks will be stored in the Site File as the Site Comments.
Figure 2-18 illustrates the placement of information for a Type A6 transaction. Below the
figure is an example Type A6 transaction. The Type A6 is indicated with an "A" in column 1
(transaction type) and a "6" in column 79 (transaction ID). For this example, site 25-009-2006
(Columns 2-10) has the comment PAMS MONIT-CRYO-DUAL FID BPl/RTx W/Dryer -
DENUDER (Columns 11-78) inserted (column 80) as the Line 1 comment. Only one Type A6 (or
A7) transaction should be created for each site.
Type Fl and A6 (or A7) files should be saved as ASCII files (multiple lines can be saved as
one file). The files should then be transferred to the EPA mainframe for a subsequent loading into the
screening file. Once Type Fl and A6 (or A7) transactions have been loaded, they must be edited
(three edit levels). Any necessary corrections must be made and an update requested. This process is
described in Section 2.6.2.4.
2.6.2.3 Raw Data Transactions
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 143 of 225
Following the update of AIRS AQS with the Type Fl and A6 (or A7) transactions, AIRS
AQS will accept raw data transactions. As with the Type Fl transactions, codes have been assigned
to the various fields required for the raw data transactions. Tables 2-9 and 2-10 present the AIRS
codes assigned to the target VOC and carbonyl O3 precursor compounds, respectively. Table 2-11
presents the various method codes associated with PAMS sampling and analysis.
-------
c
.0
ro
^s
'E
o
O
c
o
't5
v
W
1 2
A
3
4
5
6
7
8
9
10
Site Comment
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
15
36
17
38
39
40
41
42
13
14
15
46
47
18
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
'7
8
79
80
A250092006PAMS MONIT-CRYO-DUAL FID DB-l/RTx W/DRYER -DENUDER
61
Tf O
CJQ S
S
&> ^
^O
o:/s/g/morr/3797/pams/hobson/graph1. ppt
Figure 2-18. Column Placement for a Type A6 Transaction
to
oo
-------
Table 2-9. Target Volatile Organic Compounds
Section: 2
Revision: 1
Date: 09/30/98
Page: 145 of 225
AIRS
Parameter
Code
Target
Compound
Name
43203 Ethylene
43206 Acetylene
43202 Ethane
43205 Propylene
43204 Propane
43214 Isobutane
43280 1-Butene
43212 w-Butane
43216 /raws-2-Butene
43217 cH-2-Butene
43221 Isopentane
43224 1-Pentene
43220 w-Pentane
43243 Isoprene (2-methyl-1,3 -butadiene)
43226 /raws-2-Pentene
43227 cw-2-Pentene
43244 2,2-Dimethylbutane
43242 Cyclopentane
43284 2,3-Dimethylbutane
43285 2-Methylpentane
43230 3-Methylpentane
43245 1-Hexene*
43231 w-Hexane
43262 Methylcyclopentane
43247 2,4-Dimethylpentane
45201 Benzene
43248 Cyclohexane
43263 2-Methylhexane
43291 2,3-Dimethylpentane
AIRS
Parameter
Code
43249
43250
43232
43261
43252
45202
43960
43253
43233
45203
45109
45220
45204
43235
45210
45209
45212
45213
45207
45211
45208
43238
45225
45218
45219
43954
43141
43102
43000
Target
Compound
Name
3-Methylhexane
2,2,4-Trimethylpentane (isooctane)
w-Heptane
Methylcyclohexane
2,3,4-Trimethylpentane
Toluene
2-Methylheptane
3 -Methy Iheptane
w-Octane
Ethylbenzene
m/p-Xylene
Styrene
o-Xylene
w-Nonane
Isopropylbenzene (cumene)
w -Propy Ibenzene
w-Ethyltoluene (l-ethyl-3-methyIbenzene)
/>-Ethyltoluene (1 -ethyl-4-methyIbenzene)
1,3,5-Trimethylbenzene
o-Ethyltoluene (1 -ethy 1-2-methyIbenzene)
1,2,4-Trimethy Ibenzene
w-Decane
1,2,3-Trimethylbenzene
»2-Diethy Ibenzene
/>-Diethy Ibenzene
H-Undecane
w-Dodecane*
TNMOC**
PAMHC***
* These compounds have been added as calibration and retention time standards primarily for the purpose of retention
time verification. They can be quantitated at the discretion of the user.
** Total Nonmethane Organic Compounds
* * * PAMS Hydrocarbons
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 146 of 225
Table 2-10. CarbonyI Target List
Compound
Formaldehyde
Acetaldehyde
Acetone
2, 5 -Dimethylbenzal dehy de
Acrolein
Benzaldehyde
Butyr/Isobutyraldehyde
Crotonaldehyde
Hexanaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
Methyl ethyl ketone
AIRS Parameter Code
43502
43503
43551
45503
43505
45501
43329
45316
43517
43513
43504
45504
43518
43552
Reporting Requirement for
PAMS
Required
Required
Required
Optional
Optional
Optional
Optional
Optional
Optional
Optional
Optional
Optional
Optional
Optional
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 147 of 225
Table 2-11. AIRS Method Codes
Category
VOC - Automated
VOC - Manual
Carbonyl
Analysis Method
Chrompack International Auto-TCT
Concentrator with Chrompack CP 9000 GC
Entech Laboratory Automation Model 2000
Concentrator with Hewlett-Packard 5890 Series
HOC
Nutech Model 3550A Concentrator with
Hewlett-Packard 5890 Series II GC
Perkin-Elmer Corporation Model ATD-400
with Perkin-Elmer Model 8700 GC
Varian Chromatography Systems Ozone
Precursor System with Varian Model 3600-CX
GC
Chrompack International GC
Hewlett Packard Company GC
Perkin-Elmer Corporation GC
Varian Chromatography Systems GC
Cartridges coated with DNPH on silica analysis
by HPLC with KI O3 scrubber
AIRS Method Code
124
122
123
128
129
125
126
127
130
202
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 148 of 225
Sample frequency codes and sample interval codes are listed in Table 2-7 and Table 2-12,
respectively. Units codes are listed in Table 2-13. Hourly data start hours are listed in Table 2-14.
Null values are listed in Table 2-15. Other codes which may be necessary can be found in the
Geo-Common Subsystem of AIRS.
There are three types of raw data transactions: hourly, daily, and composite. The hourly
raw data transaction is used when the sample is collected for any period of time less than 24 hours.
Hourly data are indicated with a "1" in the transaction type field (column 1). The daily raw data
transaction is used when the sample is collected for at least 24 hours. A transaction type code of "2" in
column 1 indicates daily data. If several samples obtained at different times are combined and analyzed
as one, the composite raw data transaction is used.
Raw data transactions are generally inserted into AIRS, but modifications to or deletions of
previously inserted data can also be performed. An insertion is indicated with an "F in the action field
(column 80), a "M" indicates a modification, and a "D" indicates a deletion.
Hourly Data—Hourly data are those which are collected for a period of time less than 24
hours. Because several hourly samples can be collected within a given day, AIRS allows for several
results for one parameter (compound) to be input with one transaction. Figure 2-19 illustrates the
format for hourly data.
Composite data are indicated with a "3" in the transaction type field and will not be used for PAMS.
The transaction type, "1," indicates hourly data and is placed in column 1. All the sample
values entered on a particular type 1 transaction line apply to the same day.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 149 of 225
Table 2-12. AIRS Interval Codes
Code
A
B
C
D
G
M
Q
X
Y
Z
1
2
3
4
5
6
7
8
9
Interval
1 Week
3 Hours
Composite Data
Yearly
Annual Geometric Mean
Annual Arithmetic Mean
Quarterly Arithmetic Mean
24-Hr Block Average
3-Hr Block Average
8-Hr Run Average
IHour
2 Hour
4 Hour
6 Hour
8 Hour
12 Hour
24 Hours
1 Month
3 Months
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 150 of 225
Table 2-13. AIRS Unit Codes
Code
078
008
Unit
ppbC
ppb
Category
VOC Compounds
Carbonyl Compounds
Table 2-14. Hourly Sample Valid Start Hour Based on the Interval
Sampling Interval
1 hour
2 hours
3 hours
4 hours
6 hours
8 hours
12 hours
Interval Code
1
2
B
3
4
5
6
Start Hours
00, 08, 16
00,01, 16, 17
00, 01, 02
00, 01, 02, 03
00, 01, 02, 03, 04, 05
00, 01, 02, 03, 04, 05, 06, 07
00, 01, 02, 03, 04, 05, 06, 07, 08, 09, 10, 11
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 151 of 225
Table 2-15. Null Values
Value
Reason
9967
9968
9969
9970
9971
9972
9973
9974
9975
9976
9977
9978
9979
9980
9981
9982
9983
9984
9985
9986
9987
9988
9989
9990
9991
9992
9993
9994
9995
9996
9997
9998
Sample Pressure Out of Limits
Technician Unavailable
Construction/Repairs in Area
Shelter Storm Damage
Shelter Temperature Outside Limits
Scheduled but NOT Collected
Sample Time Out of Limits
Sample Flow Rate Out of Limits
Insufficient Data (Can't Calculate)
Filter Damage
Filter Leak
Voided by Operator
Miscellaneous Void
Machine Malfunction
Bad Weather
Vandalism
Collection Error
Lab Error
Poor Quality Assurance Results
Calibration
Monitoring Waived
Power Failure (POWR)
Wildlife Damage
Precision Check (PREC)
QC Control Points (Zero/Span)
QC Audit (Audit)
Maintenance/Routine Repairs
Unable to Reach Site
Multipoint Calibration
Auto Calibration
Building/Site Repair
Precision/Zero/Span
-------
Monitor ID
8.
(D / _>•* p ^^ nj "^ (- T n3 — J12 _Q)
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 153 of 225
The monitor ID is placed in columns 2-16 and is comprised of the State, County, and Site
Code (columns 2-10) where the sample was collected, the parameter (columns 11-15), and Parameter
of Occurrence (POC) code (column 16). The POC code is used to distinguish between different
monitors at the same site that measure the same parameter. The first monitor at the site is generally
identified with POC = 1, with additional monitors at the same site measuring the same compound
identified with POC = 2, etc. If a new instrument were installed to replace the original instrument used
as the first monitor, that would be the same monitor and it would have POC = 1, even if the sampling
method or interval were changed.
Column 17 is reserved for the sampling interval code. Table 2-12 gives several interval
codes which may be used. It should be noted that the interval codes must represent a non-composite
sampling interval of less than 24 hours. Because there can be as many as eight sample values per
transaction, the temporal scope of a given type 1 transaction varies considerably depending on the
sampling interval. With some intervals, a single transaction holds all the sample values for a day, while
multiple transactions are needed with other intervals. For example, only four slots of one transaction
are used to report the four 6-hour observations for a day, while all eight slots of three transactions are
needed to report the eight 1-hour observations for a day.
The units code is placed in columns 18-20 and indicates the dimensional system in which the
parameter measurement is expressed. Table 2-13 presents several units codes.
Columns 21-23 indicate the method used to sample and analyze the parameter. A method
code is valid if it exists in combination with parameter, interval, and units in the AIRS Geo-Common
File. Several method codes are summarized in Table 2-11. Choose the method code that describes
the analytical system used. Special notes describing the analytical system should be made using Type
A6 (or A7) transactions.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 154 of 225
The date of the sample is placed in columns 24-29 in the YYMMDD format. The year,
month, and day are each two digits and any leading zeros must be included.
The start hour code is placed in columns 30-31. AIRS data base time is standard time.
This code indicates the beginning hour of the sampling period for the first sample value "slot" on the
transaction, given in standard time at the location of the monitoring site. The first hour of a day, 00,
begins at midnight, and the last hour, 23, begins at 11:00 p.m. The hour of a sample value is
determined by its position on the transaction (which "slot" it occupies). The first slot is for the hour
specified in the start hour field (columns 30-31) of the transaction. The second slot is for the start hour
plus the sampling interval, the third slot is for the start hour plus twice the sampling interval, and so on.
In general;
Hour of a slot (start hour) + [(slot number 1) * (hours in sampling interval)]
Table 2-14 presents valid start hours for hourly data based on the interval.
Column 32 is the decimal point (DP) indicator. Because sample values are entered as
integers, this code is used to indicate the number of digits to the right of the decimal point in the sample
value fields. The following equation can be used to determine the correct DP for the data:
Sample Concentration 10DP Sample Value entered into AIRS as an Integer
The DP applies to all sample value fields on the transaction. If the sample values require different DPs,
use multiple transactions to code the compatible DP and sample values.
Columns 33-72 are reserved for eight sample values. These fields contain the actual
numeric values of the hourly data. There are eight "slots" for eight observations for the sample date.
These slots are in columns 33-36, 38-41, 43-46, 48-51, 53-56, 58-61, 63-66, and 68-71. The slot
into which an observation is placed depends on the hour associated with the observation, the sampling
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 155 of 225
interval (in column 17), and start hour (in columns 30-31). For the day's observation, the following
relationship holds:
SLOT# 1 + (Hour Start Hour)
NH
where:
Hour is the hour associated with the observation;
Start Hour is the start hour of the observations for that day (columns 30-31); and
NH is the number of hours between observations.
When inserting a transaction, there must not be a value in the data base for the date and hour
corresponding to the slot used. In the sample value "slot" enter a true observation of a parameter
value. Leave blank any occurrence of the sample value for an hour when you do not wish to report a
true observation. Numeric integer values (no decimal points) that represent true observations must be
entered right justified in the fields. Leading zeros are not required, but recommended. The sample
value and the decimal point indicator together indicate the compound concentration reported in the units
indicated by the unit code.
Following each sample value field is a field for the validity flag. The validity flag is used to
indicate the reason for an abnormal observation. The presence of a validity flag with an observation
indicates that the observation is due to an "exceptional event" or its abnormal value has been checked
and found to be valid. The use of any validity flag (except "V") must be approved by state or EPA
regional personnel. If an observation results from an exceptional event (e.g., volcano, forest fire,
disaster clean-up, etc.), place a valid validity flag in column 37, 42, 47, 52, 57, 62, 67, and/or 72. A
valid validity flag is one that exists in the AIRS Geo-Common File for this parameter. If a validity flag is
not necessary, the validity flag field and/or fields should be left blank.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 156 of 225
The action code placed in column 80 indicates the type of data base processing to be
performed by this transaction. In order to insert new values into the raw data files, an "F is placed in
column 80. Should modifications to existing values in the raw data files be necessary, a "M" is placed
in column 80. The deletion of existing values from the raw data files is indicated with a "D" in column
80.
A protocol has been established to explain "missing data" should there be an instance where
data have not been collected. If there is a instance where data should have been collected, but were
not, the DP indicator (column 32) should be left blank. This procedure indicates to AIRS that the value
entered in the sample value field is a code explaining why the data are missing. A list of approved null
value codes appears in Table 2-15. Always use the most appropriate null value code. When a reason
for a null value exists that is not listed, use the miscellaneous void code: Number 9979.
An example of an hourly transaction appears in Figure 2-16. In the first transaction line, site
25-009-2006 is submitting data for benzene (parameter 45201) for the primary sample (POC =1).
The samples were collected for three hours (interval = B). The results are presented in ppbC (units =
078) for a Varian GC (method = 129). The sample was collected July 7, 1994 (columns 24-29).
Because the start hour is 00 (columns 30-31), the three hour samples were collected beginning at
midnight, 3:00 a.m., 6:00 a.m., 9:00 a.m., noon, 3:00 p.m., 6:00 p.m., and 9:00 p.m. local standard
time. The results are presented so that a decimal point appears two places to the left of the indicated
value (column 32). The sample collected at midnight (columns 33-36) has a value of 3.98 ppbC
whereas the sample collected at 3:00 p.m. (columns 58-61) has a value of 23.45 ppbC. This
transaction summarizes the data collected for one compound (benzene) for one day. Similar
transactions would be necessary for each of the over 50 compounds on the VOC target list for this
date.
The second line is the transaction for formaldehyde. Site 25-009-2006 is submitting data for
formaldehyde (parameter 43502) for the primary sample (POC=1). The samples were collected for
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 157 of 225
three hours (interval = B). The results are presented in ppbv (units = 008) for cartridges coated with
DNPH on silica using a potassium iodide ozone scrubber (method = 202). The sample was collected
July 7, 1994 (columns 24-29). Because the start hour is 00 (columns 30-31), the three hour samples
were collected beginning at midnight, 3:00 a.m., 6:00 a.m., 9:00 a.m., noon, 3:00 p.m., 6:00 p.m., and
9:00 p.m. local standard time. The results are presented so that a decimal point appears two places to
the left of the indicated value. The sample collected at 3:00 a.m. (columns 38-41) has a value of 3.33
ppbv whereas the sample collected at 9:00 p.m. (columns 68-71) has a value of 4.89 ppbv. This
transaction summarizes the formaldehyde data collected for one day. Similar transactions would be
necessary for each of the three compounds on the carbonyl target list for this date.
Daily Data—Daily data are those which are collected for a period of at least 24 hours or
more. These data are called transaction Type 2 and are indicated with a "2" in column 1 of the
transaction. The column placement for daily data is presented in Figure 2-20.
The monitor ID is placed in columns 2-16 and is comprised of the State, County, and Site
Code (columns 2-10) where the sample was collected, the parameter (columns 11-15), and Parameter
of Occurrence (POC) code (column 16). The POC code is used to distinguish between different
monitors at the same site that are measuring the same parameter. The first monitor at the site is
generally identified with POC = 1, with additional monitors at the same site measuring the same
compound identified with POC = 2. If a new instrument were installed to replace the original
instrument used as the first monitor, that would be the same monitor and it would have POC = 1, even
if the sampling method or interval were changed.
Column 17 is reserved for the sampling interval code. Table 2-12 gives several interval
codes which may be used. It should be noted that the interval codes must represent a non-composite
sampling interval of at least 24 hours.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 158 of 225
The units code is placed in columns 18-20 and indicates the dimensional system in which the
parameter measurement is expressed. Table 2-13 presents several units codes.
Columns 21-23 indicate the method used to sample and analyze the parameter. A method
code is valid if it exists in combination with parameter, interval, and units in the AIRS Geo-Common
File. Several method codes are summarized in Table 2-11. Choose the method code that describes
the analytical system used. Special notes about the analytical system should be made using Type A6
(or A7) transactions.
-------
Monitor ID
8.
>s
0) O
cr'w
%
'E
3
0
O
4
5
6
JD
W
7
8
9
10
Paramete
/
11
12
13
14
15
O
O
Q_
16
5
I
^
17
!
(/)
'If
! z>
18
19
20
Method
21
22
23
ro
X
24
25
O
26
27
>%
28
29
O
X
30
31
0)
E
%
32
Decimal I
33
Q.
E
(0
w VF
34
15
36
17
38
39
40
41
42
13
14
45
46
47
18
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
'7
'8
79
C
O
1
80
2250092006452011707812994070700P21864
2250092006435021700810294070700P2 280
I
I
Tf
CJQ
S R
^ o
(J\ \O
^ w
8, ^
VO
o:/s/g/morr/3797/pams/hobson/graph4.ppt
Figure 2-20. Column Placement for a Daily Data Transaction
to
oo
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 160 of 225
The date of the sample is placed in columns 24-29 in the YYMMDD format. The year,
month, and day are each two digits and any leading zeros must be included.
The start hour code is placed in columns 30-31. This entry indicates the beginning hour of
the sampling period, given in standard time at the location of the monitoring site. The first hour of a day,
00, begins at midnight, and the last hour, 23, begins at 11:00 p.m.
Sampling intervals of 24 hours or more should be reported on the day in which the majority
of the sampling occurred. Therefore, a 24-hour sample taken from midnight (00:00) to midnight
(24:00) local daylight time on June 5, would be reported with a start time of (00:00) on June 5 rather
than 23:00 on June 4.
The sample frequency code is placed in column 32 and indicates how much time elapses
between observations. This code is expressed in terms compatible with the associated Interval Code.
Table 2-7 presents sample frequency codes.
Column 33 is the decimal point (DP) indicator. Because sample values are entered as
integers, this code is used to indicate the number of digits to the right of the decimal point in the sample
value fields. The following equation can be used to determine the correct DP for the data:
Sample Concentration 10DP Sample Value entered into AIRS as an Integer
The sample value is placed in columns 34-37. The integer value represents the true
observation and is entered right justified. Leading zeros are optional.
The validity flag is placed in column 38 and is used to give the reason for an abnormal
observation. The presence of a validity flag with an observation indicates that the observation is due to
an "exceptional event" or its abnormal value was checked and found to be valid. The use of any
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 161 of 225
validity flag (except "V") must be approved by state or EPA regional personnel. If a validity flag is not
necessary, the validity flag field should be left blank.
The action code placed in column 80 indicates the type of data base processing to be
performed by this transaction. In order to insert new values into the raw data files, an "F is placed in
column 80. Should modifications to existing values in the raw data files be needed, an "M" is placed in
column 80. The deletion of existing values from the raw data files is indicated with a "D" in column 80.
A protocol has been established to explain "missing data" should there be an instance where
data have not been collected. If there is a instance where data should have been collected, but were
not, the DP indicator (column 32) should be left blank. This blank column indicates to AIRS that the
value entered in the sample value field is a code explaining why the data are missing. A list of approved
null value codes appears in Table 2-14.
An example of a daily transaction appears in Figure 2-20. In the first transaction line, site
25-009-2006 is submitting data for benzene (parameter 45201) for the primary sample (POC =1).
The sample was collected for 24 hours (Interval = 7). The results are presented in ppbC (units = 078)
for a Varian GC (method = 129). The sample was collected July 7, 1994 (columns 24-29). The
sample collection began at midnight (Start Hour = 00). The sample was collected on the PAMS
schedule of one 24-hour sample every 6 days (Sample Frequency = P). The results are presented such
that a decimal point appears two places to the left of the indicated value (column 33). The sample
value is 18.64 ppbC (columns 34-37). This transaction summarizes the data collected for one
compound (benzene) for one day. Similar transactions would be necessary for each of the over 50
compounds on the VOC list for this date.
The second line is the example transaction for formaldehyde. In this transaction, site 25-
009-2006 is submitting data for formaldehyde (parameter 43502) for the primary sample (POC = 1).
The sample was collected for 24 hours (interval = 7). The results are presented in ppbv (units = 008)
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 162 of 225
for cartridges coated with DNPH on silica (method = 102). The sample was collected July 7, 1994
(columns 24-29). The sample collection began at midnight local standard time (Start Hour = 00). The
sample was collected on the schedule of one 24-hour sample every 6 days (Sample Frequency = P).
The results are presented so that a decimal point appears two places to the left of the indicated value
(column 33). The sample value is 2.80 ppbv (columns 34-37). This transaction summarizes the data
collected for one compound (formaldehyde) for one day. Similar transactions would be necessary for
the three carbonyl compounds for this date.
2.6.2.4 Submitting Data
Once the raw data have been formatted in accordance with the required fixed block 80-
character per line formats, the process of submitting data to AIRS can begin. For batch processing,
groups of transactions should be saved as ASCII files which are then loaded onto the EPA mainframe.
However, the transactions must still be placed in the screening file.
Once the raw data ASCII files have been loaded onto the EPA mainframe, work within the
AIRS system can begin. After accessing AIRS, entering AQS, and selecting the screening file, a four
step process begins to submit the data to AIRS AQS. The process begins by choosing Submit from
the AQS Menu. The submenu should then include, among others, choices for LOAD, EDIT,
CORRECT, and SUBMIT.
The LOAD Process-Placing the Data in the Screening File—LOAD is a batch
process which is initiated from the Air Quality "Submit" menu. The transaction records are read from
the TSO dataset specified by the user and written into the user's screening file. A printed report is
generated that summarizes the processing performed in terms of the number of transaction records by
type or category. Suppression of the printed report can be achieved by selecting "N" for Print.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 163 of 225
The EDIT Process-Validating the Data—EDIT is a batch process which is initiated
from the Air Quality "Submit" menu. Transaction records in the user's screening file are examined and
edit/validation checks are applied. The purpose is to ensure that only "valid" data get into the data
base. A printed report is generated that lists each record that failed the edit checks and the reason(s)
why it failed. Suppression of the printed report can be achieved by selecting "N" for Print.
There are three (3) levels of edit/validation checks. Level 1 determines whether the
individual fields of a record have appropriate values (e.g., is the city code valid?). Level 2 checks the
relationship among the fields or records. Duplicate transaction records would result in a Level 2 error.
Level 3 statistically tests for anomalies in the data and checks for relationships among the transaction
records and the AIRS Data Base. All transaction data must pass this 3-level editing system prior to
admittance into the AIRS Data Base.
The CORRECT Process-Correcting the Data—Air Quality transaction records in the
user's screening file may be modified in order to correct edit/validation errors identified by the EDIT
process or to correct some other problem known to the user. CORRECT does not generate a printed
report.
The CORRECT process allows one to "browse" in the screening file to view its contents,
and to modify, delete, or add transaction records. A particular record or group of records may be
selected for display on the user's terminal screen. As each record is displayed, the user may leave the
data fields intact, alter the contents of the data fields, or delete the record. The user may also set (or
clear) an indicator that causes the record to be excluded temporarily from EDIT and UPDATE
processing, while retaining it in the file. The same kinds of operations may also be applied to groups of
records using "global" commands that specify the identities (keys) of the records to be affected and the
actions to be applied. New transaction records may also be inserted into the screening file. Level-1
edit checks are applied to inserted and modified records and any errors are indicated on the user's
screen. Edit checks are not applied to each record modified by a global command, but to the
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 164 of 225
command itself. Any records that have passed Level-2 or Level-3 edit checks and are interdependent
with a record affected by CORRECT processing will be reset to Level-1.
The NOTIFY Process-Requesting to Add Data to AIRS AQS—Once all records
have passed all three edit checks, the AIRS Data Base Administrator should be notified of the
impending AIRS AQS data by selecting NOTIFY from the submenu. NOTIFYing the AIRS Data
Base Administrator locks the screening file and prevents further use of it until the AIRS Data Base
Administrator releases it. The AIRS Data Base Administrator and the appropriate NAMS/PAMS
Coordinators review the data. The data are then added to the AIRS data base.
Should there be any suspect data, the AIRS Data Base Administrator may elect not to add
the data to AIRS, but instead leave the data in the user's screening file and request the user to double-
check the transaction. Upon inspection, should there be an error in the transaction, the user should
CORRECT the error, EDIT check/validate the data, and NOTIFY the AIRS Data Base Administrator
of the impending AIRS AQS data. If the data are valid, the user may want to indicate that the data
have been validated by placing a validity flag of "V" to indicate that the data are indeed valid. The
transaction should then proceed through the EDIT checks/validations and after successful completion of
the EDIT checks, the user should NOTIFY the AIRS Data Base Administrator of the data.
2.7 Validating Data from Automated VOC Systems
Although manual sampling methods may be used, automated GC techniques are currently
the most practical and cost effective way to comply with the rigorous speciated VOC sampling
frequency required for PAMS. Measuring VOCs in the atmosphere on a daily and hourly basis using
these systems produces extremely large and complex data sets. Managing, processing, and validating
the data requires technical expertise and an intensive effort to obtain reliable and consistent data for the
timely input into the AIRS AQS data base. The AIRS data base is used as the national repository for
PAMS data and can be used to assist State and local agencies in determining if the program objectives
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 165 of 225
and Data Quality Objectives (DQOs) described in the PAMS Implementation Manual25 are met. Data
submitted to AIRS by all agencies must be consistent with the PAMS monitoring DQOs and of
adequate quality to meet Clean Air Act Title I objectives.
PAMS data use is discussed elsewhere in this document, but it is pertinent to the data
validation process to reiterate the discussion. The data will allow the state and local agencies to
develop, evaluate, and refine new O3 control strategies; determine NAAQS attainment or non-
attainment for O3; track VOCs and NOX emissions inventory reductions; provide photochemical
prediction model input; evaluate photochemical prediction model performance; analyze ambient air
quality trends; and characterize population exposure to VOCs and O3. Data from VOC measurement
systems must be submitted to AIRS within six months following the end of each quarterly reporting
period.
Validating the measurement data can be as complex as the gas chromatographic techniques
and methods used to collect them. The evaluation of the quality and reliability of the data from GC
analyses is often referred to as the data validation process. The EPA defines data validation as a
systematic process consisting of data editing, screening, checking, auditing, verification, certification,
and review, for comparing a body of data to an established set of criteria to provide assurance that the
data are adequate for their intended use.50 Data validation is an element of quality assurance and
includes evaluation of the data quality and reliability and assurance that the data are consistent with
program data quality objectives.
Data validation is the final and most critical part of the process used to generate PAMS
data. The data must be validated and reviewed to ensure the overall quality of the measurement prior
to inclusion in the AIRS AQS data base. Data validation is used in conjunction with program
objectives, DQOs, and QA/QC to remove inconsistencies in the data set and improve data quality.
The key aspects of a complete data validation process are the pre-measurement chromatographic
system verification, development of QC procedures and measurement quality objectives; and validation
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 166 of 225
of the data prior to AIRS AQS data base entry. Although the data validation process is embodied in
the last of these aspects, all are of critical importance to the overall process. Prior to the development
of systematic data validation procedures, pre-measurement system validation and QA/QC
measurements for establishing the DQOs must be completed. QA/QC procedures pertain to the
techniques applied prior to obtaining data, such as calibration checks, system blanks and external
audits. Guidance for pre-measurement validation and QA/QC is given in Sections 2.3.7 and 2.8. This
process is necessary to set the stage for ensuring the overall quality of the monitoring network data.
2.7.1 Data Validation Approach
This guidance provides procedures for implementing a systematic data validation process for
those responsible for validating PAMS VOC data. Data validation in the context of this guidance refers
to the procedures performed after the data are collected. Pre-measurement chromatographic system
verification, QA/QC procedures and measurement quality objectives are established prior to collection
of the data to minimize the amount of faulty data generated.
Data validation is performed as a last step before the data are submitted to AIRS and is a
screening process to prevent additional unacceptable or questionable data from being submitted to
AIRS. Timely data validation is required to more easily resolve data issues and unusual events and take
the necessary corrective actions to minimize the generation of additional faulty data.
This data validation approach has been selected with respect to the PAMS DQOs and data
use; volume and type of data generated; anticipated computational and graphical capabilities of the state
and local agencies; and nature of the expected errors. Four categories have previously been
identified51 for validating air monitoring data:
Routine checks made during the initial processing and generation of the data, including
proper data result file identification, review of unusual events, review of
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 167 of 225
chromatography result reports, performance checks of the data processing system, and
deterministic relationships.
• Tests for internal consistency to identify values in the data set which appear atypical
when compared to values of the whole data set.
• Comparing the current data set with historical data to verify consistency over time.
• Tests for parallel consistency with data sets from the same population (region, period of
time, air mass, etc.) to identify systematic bias.
The data validation guidance presented here for VOC encompasses mainly routine checks,
tests for internal consistency, and historical data comparisons. Additional checks for parallel
consistency, which incorporate statistical evaluations, may be considered for data validation and are
only briefly discussed in this document. A flow chart of data validation activities is given in Figure 2-21.
2.7.1.1 Routine Procedures
Routine validation checks during the processing of the data include verification of proper
data file identification information, chromatography result report file review, identification of unusual
events, deterministic relationship checks, and performance checks of the automated data processing
systems. These routine checks should be done frequently (i.e., daily or weekly) to parcel the data into
manageable segments. The checks must be timely in order to address issues before the amount of
faulty or unusable data generated becomes too large, thereby significantly affecting the data
completeness.
Computerized data acquisition systems are used to collect the chromatographic information
from the analog output of the GC systems as discussed in Section 2.6. The resulting chromatographic
data files must be clearly and correctly identified, including the correct acquisition time and date,
sampling location, sample name or type, processing and calibration methods, and file naming
conventions. Examples of errors that may occur include: incorrect sampling locations, especially if
methods are copied from one site location to another; incorrect date and time stamp due to daylight
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 168 of 225
savings time change; or a sample name that indicates a normal sample when in fact a calibration or
blank check sample has been analyzed. Since these errors are mostly due to human error, an individual
other than the person originally generating the data should review the information.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 169 of 225
Review
Instrument QC
Information
Data Collection
Review Result File
Information
Review Sample
Chromatogram
Take Necessary
Corrective Action
Review and Update
Site Log Book
Deterministic
Relationships
Data Processing
Performance Review
Update
Acquisition
Method
Routine
Checks
Figure 2-21. Flow of Data Validation Activities
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 170 of 225
Control
Charts
Summary Report
Review
Time Series, Scatter
Plot, Finger Print,
Review for Outlier
Gross Limit Checks
Data Average, Mean,
and Range Checks
Sign, Wilcoxon
Signed-Rank,
Wilcoxon Rank Sum,
and Intersite
Correlations Tests
Treatment of Outlier
Internal
Consistency
Checks
Historical
Consistency
Checks
Parallel
Consistency
Checks
Figure 2-21. Continued
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 171 of 225
Due to the large volume of data generated from PAMS monitoring, close scrutiny of all
chromatograms and result reports is not practical. For example, a system making continuous, hourly
measurements of 56 target compounds can produce over 1,300 measurements per day. However, all
chromatograms should go through a cursory review by the station operator to determine if the quality of
the chromatography (i.e., appearance of the chromatogram, the peak shape, peak resolution, peak
integration, retention times, and baseline) is acceptable. This level of review can be done quickly by an
experienced chromatographer and will also determine if the chromatographic system is performing
properly. Comparison of the chromatogram to reference or historical information, such as calibration
and typical sample chromatograms, can simplify this process. Chromatograms are also reviewed to
determine if there are any gross errors present and whether chromatographic abnormalities, such as
electronic spikes, contamination, or levels of target analytes above the electrometer or calibration
ranges, exist. If the chromatography is acceptable, the chromatogram can then be further processed by
the data reduction and/or peak identification software as chosen by the user.
The cursory review of chromatograms may include the following determinations:
The signal from the FID or baseline is normal and the signal output is positive (on-
scale);
Chromatographic peaks are present, integrated correctly, and the peak-shape is sharp;
The peak resolution or separation is acceptable based on historical instrument
performance;
All components have been eluted from the analytical column as indicated by a flat or
normal baseline at the end of a run; and
No chromatographic abnormalities exist, such as large contamination or non-target co-
eluting compounds, and electronic spikes.
The result reports must also go through a more in depth review, in conjunction with the
chromatogram, to ensure that the key reference or internal standard compounds are identified correctly
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 172 of 225
and the resulting target peak retention times have not shifted. The results reports generated by the
automated GC system are subsequently reviewed and compared to calibration or standard analyses
information from the pre-measurement system verification discussed in Section 2.3.7. This review is
used to ensure that correct peak assignments or identifications are made and that the resulting
concentrations are correct. The information is also reviewed to determine whether the information used
to make the peak identifications (i.e., retention times, relative retention times, retention indices, etc.)
requires updating. The need for updating peak identification information in the acquisition method is
indicated by the frequency of missed or inaccurate peak identifications automatically made by the GC
system. Typically, a minimum of 10% of the data are processed through this level of review. The
minimum percentage of the chromatograms generated daily are selected by the station operator for
review. Chromatograms generated at the beginning and the end of an analysis day to bracket sample
analyses are selected. It can then be presumed that the data generated between bracketed result files
and chromatograms are acceptable and also accurate.
Routine review of the analytical instrument calibration is imperative for the ultimate quality and
usefulness of the data generated. As a QC function, the station operator must review this calibration
information on a regular basis to ensure that quality objectives are met and the system is operating
properly during the measurement process. If the GC system calibration has gone awry and QC sample
results do not agree with the specified quality objectives, the data should be appropriately qualified to
indicate any impact on the analytical results. If "real-time" (same day) calibration review is
implemented, quality issues can be identified and corrective actions, such as repeat analyses and
instrument maintenance, can be performed prior to continuing data collection. If post analysis
calibration information review is performed, the previously generated sample data should be flagged to
clearly qualify any uncertainty in the results.
An on-site logbook should be maintained to record information relative to instrument repair
and maintenance (replacing the cold trap) and unusual events (mowing the lawn, power outages, or
repairing the sampling shelter roof) that could explain variations or excursions in the data. This
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 173 of 225
information could be critical in the explanation of missing data files, high sample concentrations, or used
to reject outliers.
Deterministic relationships where two or more related parameters can be routinely checked
ensure that measured values of an individual parameter do not exceed that of an aggregate parameter.
The aggregate parameter must include the individual parameter. Also, relationships or ratios between
two source-related compounds can be compared in order to identify errors. For example, if the total
unidentified VOCs measured are greater than an established empirical percentage of the total NMOC
measurement at that site, a large number of target peaks were misidentified. A total unidentified VOC
value that comprises more than 30% of the total NMOC may indicate that retention times have shifted
and target peak identifications are incorrect. The concentration of benzene can be compared to
acetylene or toluene (all present in automobile exhaust) by developing scatter plots or calculating the
ratios and comparing this information to empirical relationships determined at a specific site. The
example in Figure 2-2252 shows a scatter plot of benzene and toluene that incorrectly includes
calibration information in the data set, as demonstrated by the data points around the concentration of
30 ppbC at the top part of the graph. Most of the ambient air data points in this scatter plot fall below
8 ppbC for benzene. Again, these relationships are site specific and empirical relationships must be
established.
Performance checks of the automated data processing system and supplemental procedures
developed to handle the data, including telemetry, should be implemented. All agencies should develop
data flow diagrams to indicate the steps taken to process the data following collection, including data
formatting, transmission, and processing or formatting for AIRS. The data are reviewed for
inconsistencies, missing data files, or nonsensical information. Any computerized programs put in place
to manipulate data should be validated at a minimum using methods for checking errors such as
developing a standard set of test output parameters, processing the test data set, and comparing the
results to the reference. Information regarding complex procedures required for testing and validating
computer systems can also be obtained from NIST.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 174 of 225
ill l
22.E
Ihl
O.C
0,0 12,5 250 :?,; K.Qp*bC
Figure 2-22. Appearance of Calibration Data at East Hartford, CT, in June 1995.
Example scatter plot showing calibration data of about 30 ppbC. Data are level 0,
preliminary data, CT DEP.
Main, H.H., P.T. Roberts, and M.E. Korc. Analysis of PAMS and NARSTO-Northeast Data -
Supporting Evaluation and Design of Ozone Control Strategies: A Workshop. Report and
presentation prepared for U.S. Environmental Protection Agency, Research Triangle Park, NC.
Presented at U.S. Environmental Protection Agency, Research Triangle Park, NC.
STI-94551/94581-1551-WD1, EPA WA-5-95 and WA-8-95, December 1995.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 175 of 225
Any software or calculation spreadsheet programs (Excel® or QuattroPro®) implemented by
the user must also be verified for accuracy. Errors in spreadsheets can occur during spreadsheet or
program development in the continued course of the program. Implementing a verification check of the
calculations in these spreadsheets and programs can help to avoid errors in the data generated.
Verification checks are accomplished by using the spreadsheet or program to calculate results for a
previously known data set. A more complete test of the user developed program would include testing
at the minimum and maximum expected input values. If the expected result is not obtained the program
may be in error and the appropriate corrective action should be taken.
2.7.1.2 Tests for Internal Consistency
Data review for internal consistency identifies values in the data set which appear atypical
when compared to values of the whole data set. Tests for internal consistency include the identification
of outliers and extreme differences in adjacent values that require further investigation. A number of
statistical tests can be used for internal consistency checks and determining outliers. Graphical and
visual presentation of the data, such as review of summary report file information, scatter plots, time-
series, or fingerprints can also be used for consistency checks.
Manual data review for internal consistency is impractical for the volume of continuous gas
chromatographic data generated by PAMS. The user should consider implementing a commercially
available peak processing or identification software package as an alternative to manual review.
Software packages, such as MetaChrom™ (Meta Four Software, Inc.) and VOCDat (developed by
Sonoma Technology, Inc., under sponsorship by EPRI53) have been designed to review large and
complex data sets for consistency.
Summary report file information, scatter plots, time-series, fingerprints, and control charts can
be generated using the Perkin-Elmer Turbochrom® Summary Report option, spreadsheet software
(Excel®, QuattroPro®, or Lotus®), or the VOCDat software developed for NARSTO-NE by Sonoma
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 176 of 225
Technology, Inc. VOCDat provides a useful graphical interface for generating time-series plots,
fingerprint plots, and control charts for VOC data generated by Perkin-Elmer Turbochrom® data
systems.
Qualitative comparison review for internal consistency may by done by generating a data
base or summary report file or "flat-file" of the concentration and peak identification results generated
over time. A typical summary file can contain a large number of measurements for the target PAMS
compounds: a system making continuous, hourly measurements of 56 target compounds can produce
over 1,300 measurements per day. The summary files can be reduced into manageable segments for
visual review. This information can assist the user in identifying system problems, target compound
misidentifications, system contamination, outliers, or missed information. Qualitative comparison review
of the final concentration results and peak identifications is valuable in checking for outliers or
inconsistencies in peak identification, retention times, calculations, and results. The information can also
be globally reviewed for clear changes in trends.
For example, the Peak Summary Option of the Perkin-Elmer Turbochrom® software can
easily be used to summarize the concentration and retention time results as shown in Figure 2-23. This
file can be generated weekly and the information reviewed to identify abrupt changes in concentration
or retention time, and the presence of ubiquitous compounds measured at the site. A criterion for
retention time shifts may also be established based on the percent relative standard deviation (%RSD)
information obtained for the site to pinpoint misidentification of target peaks.
When data are reviewed for internal consistency, dealing with values that fall below the
established detection limits for the system is at the discretion of the reviewer. For example,
concentration values below the detection limit may not be subject to data validation due to the inherent
uncertainty of these data. An action limit of 3 times the detection limit or a concentration value (i.e., 5
ppbC) may be used.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 177 of 225
I I I I I I I I
o o '—i^ors1/"!'—i ^r c^oooot""ooooc^c^
I I I I I I I I
*
*
*
*
*
*
*
*
H
S
o.
UJ
P5
>H
<
^?
Q
H
^
*
*
*
*
*
*
*
. :
o !
a !
0 |
° i
w
1
!
ti i
pi i
I
a 1
o !
1
w
W ca i
i-J u !
£^ i
S !
W
1
1
!' ' ' !' ' '' ' ' ' ' '' ' '' ' ' ' ' !
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 178 of 225
I I I I I I I I I I I I I I I
w
ffi
i i i i i i i i i i i i i
W
'Z
Pi
PH
, n
•Q Q> < < 1' < < 1' < 1' < < 1' < 1' < < 1' £-J
D ^^^^^^^^^^^^^^^^^"~
=
=
*c
=
O . m
r O • • • . : : : : : : ^ i/-i
hH r\rO(N(NOOOOOO(N(N'—' <—!'—' <—'^-X,
s °
o
^^ ri
?V P-l rrt
g w
s
* PH (^ <^ i^ <^ i^ i^ i^ i^ <^ <^ <^ <^ <^ <^ <^ <^ CZU
w
^
t~~
VO
H
W
T3
^JOOOOOOONOOONOOOOOOOOONONOOOOI>0000 0)
^(^(^(^rororororornrnrnrnrnrnrnrocD «
pHo6o6o6o6o6o6o6o6o6o6ooo6o6o6o6o6o «
"•5
C
<« o
^
H CO
Cl
M
UOOOOOOOOOOOOOOOOI)C« Q)
^;y 1_
pLH^/^l/^l/^l/^l/^l/^l/^l/^l/^l/^l/^l/^^/^l/^l/^^^ w
"Si "Si "Si "Si "Hi "Si "Si "Si "Si "Si "Si "Si "Si "Si 3 ° o)
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 179 of 225
The Turbochrom® Peak Summary Report may also be generated as an ASCII file for import
into Excel® or other spreadsheet software. The ASCII summary data base file can be loaded or
imported directly into spreadsheet software for further data processing or manipulation. Once the data
base file is loaded into the spreadsheet software, graphical representations of the continuous
measurements can easily be generated. Generation of diurnal graphs of hourly site measurement
information can be very useful in clearly identifying trends and determining if the potential for outliers
exists in the data base.
The VOCDat software53 can generate graphical presentations of the data in the form of time
series plots, scatter plots, and fingerprint plots for internal consistency determinations. Time series plots
as shown in Figure 2-2453 can be used to inspect each target compound, groups of target compounds,
and total NMOC. This information allows the identification of outliers, increased single-hour
concentrations, possible missed peak identifications, and extended periods of unusually high or low
concentrations. Experienced PAMS personnel frequently look for unusual "jumps" in the time series
plot between successive hourly data or departures from expected diurnal or seasonal patterns. Figure
2-2453 shows two items that warrant further investigation: first, the unusually high concentration of 62
ppbC for propane compared to the other data points for the period which are all below 20 ppbC; and
second, missing data for two periods (8/24 and 8/28). Review of the site log would be important in this
case to determine if downtime occurred or instrument maintenance was performed.
The time series plot in Figure 2-2554 shows an example of misidentification of a paraffin for an
unidentified VOC, indicated on the plot by the abrupt decrease in paraffin concentration with
concurrent increase in the unidentified VOC concentration. The time series plot in Figure 2-2654 shows
system contamination of the selected components that decreased in concentration, or were cleared out
of the system over a period of five days.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 180 of 225
&£fc HI Sapt, Whita WP ,-H|xJ
ll ....I .».1I...:-,...U ;...; I :..;:: ;..; :,.
*Sj TirwIESTj
46.5
31.0
15.5
8
[ I i . | . 1 ! f 1 r , [ i i r | r > ! | ^ ! . 1 r
-
-
I i i
• ' : . : /• '5 . J ':
t -• t|!-.-;/: '-.--•-•-- ---";' .--- *-'^",' , I''-' ;'-': ' .. '--""" -.'^'i.' *; '
acely
s B propa
nbula
• ispna
24 8/25 8/2i 6/2? 8/28 8/29 8/30
3vOCi!tl-|T<*eS»«Sri.,
Figure 2-24. Time Series Plot
Main, H.H., P.T. Roberts, J.D. Prouty, and M.E. Korc. Software for Display, Quality Control, and Analysis of Coni
VOC Data. Report prepared for Electric Power Research Institute, Palo Alto, CA by Sonoma Technology, Inc., •
Rosa, CA. STI-996142-1594, EPRI Research Project No. WO9108-01, June 1996.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 181 of 225
to TlKiST
Figure 2-25. Time Series Plot of Several Species Groups at Stafford, CT, in 1994. Example ol
misidentification of a paraffin for an unidentified peak.
Main, H.H., P.T. Roberts, and L.R. Chinkin. PAMS Data Analysis Workshop: Illustrating the Use of PAMS Data ti
Ozone Control Programs. Report and presentation prepared for U.S. Environmental Protection Agency, Resea
Triangle Park, NC. Presented at California Air Resources Board and EPA Region IX, Sacramento, CA. STI-9971i
WD7. May 1997.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 182 of 225
VOCDat - mini) Series Gmpfc iHJ
51 file &» graph BWow tjslp
Tima(EST)
6/7
Gary. IN
to!u-13-9ppbC
Figure 2-26. Time Series Example of System Contamination
Main, H.H., Roberts, P.T., and Chinkin, L.R. PAMS Data Analysis Workshop: Illustrating the Use of PAMS Data 1
Ozone Control Programs. Report and presentation prepared for U.S. Environmental Protection Agency, Resea
Triangle Park, NC. Presented at California Air Resources Board and EPA Region IX, Sacramento, CA. STI-9971i
WD7. May 1997.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 183 of 225
A scatter plot (shown in Figure 2-2754) can be used to compare pairs of target compounds or
target groups to identify outliers and excursions in the data such as the improper inclusion of calibration
data in the data set (discussed previously). The figure shows a comparison of 2-methylheptane to
toluene. The relationships between two species show a "cone" of data rather than the two distinct lines
observed in the figure. In this example, the toluene peak was misidentified as 2-methylheptane over a
period of several days.
A fingerprint plot shown in Figure 2-2852 allows further inspection of samples previously
flagged for more detailed review. The fingerprint plot shows the compound concentration for each
compound for a single hour. The fingerprints can quickly be scanned, hour-by-hour to allow the
observation of diurnal changes and inspection of hours surrounding suspect data to identify additional
effects. The time series plot in Figure 2-2952 clearly illustrates calibration data mistakenly included in
the data set. In this case, the calibration gas contained about 25 ppbC of the PAMS target
compounds.
Whatever methodology is used, the qualitative comparison review of summarized
measurement data can be a useful tool in verifying the internal consistency and overall accuracy of the
data generated from automated GC systems.
2.7.1.3 Historical Data Comparisons
Testing or comparing data for historical consistency uses many of the graphical techniques
discussed in Section 2.7.1.2 to compare the data set with previous data compiled from the monitoring
location. Pattern and successive value tests, parameter relationship tests, and control charts may also
be used.51 Gross limit checks are used to detect data that are unlikely and considered impossible.
Upper and lower limits are established based on historical data from the site. These limits are specific
to the monitoring site and should consider parameters and instrument characteristics. Values
representing pollutant behavior outside the specified limits are flagged for further investigation. Limits
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 184 of 225
on the individual concentration, difference in adjacent concentration or retention time values, difference
or percent difference between both adjacent
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 185 of 225
Example of Species Misidentification
35 -r
1 H—
10 15 20
2-Methylheptane (ppbC)
25
30
Figure 2-27. Example of Misidentification Using a Scatter Plot. Typically, data
points would be present in the region of the plot between the two extreme edges.
Main, H.H., Roberts, P.T., and Chinkin, L.R. PAMS Data Analysis Workshop: Illustrating the Use of
PAMS Data to Support Ozone Control Programs. Report and presentation prepared for U.S.
Environmental Protection Agency, Research Triangle Park, NC. Presented at California Air
Resources Board and EPA Region IX, Sacramento, CA. STI-997100-1719-WD7. May 1997.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 186 of 225
30
25
20
Example Typical Fingerprint
O
c
o
15
o>
o
c
o
o
10
Figure 2-28. Example of a Typical "Fingerprint" Observed at a PAMS Site
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 187 of 225
Example Calibration Gas
o
£1
Q.
a.
o
o
O
mTTTTTTTrrrm
20
10
Figure 2-29. Example of Calibration Gas "Fingerprint" Observed in Data
Submitted to AIRS
Main, H.H., P.T. Roberts, and M.E. Korc. Analysis of PAMS and NARSTO-Northeast Data -
Supporting Evaluation and Design of Ozone Control Strategies: A Workshop. Report and
presentation prepared for U.S. Environmental Protection Agency, Research Triangle Park, NC.
Presented at U.S. Environmental Protection Agency, Research Triangle Park, NC.
STI-94551/94581-1551-WD1, EPA WA-5-95 and WA-8-95, December 1995.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 188 of 225
values, and the average of four consecutive values are comparisons that may be considered.
Supplemental statistical tests can be used to identify data sets which have mean or range values that are
inconsistent with previous data sets.
2.7.1.4 Parallel Consistency Checks to Identify Systematic Bias
Tests to check for consistency with parallel data sets from the same population (region,
period of time, air mass, etc.) are used to identify systematic bias. Systematic bias is determined by
checking for the difference in average value or overall distribution values. The sign test, Wilcoxon
signed-rank test, Wilcoxon rank sum test, and intersite correlation test are recommended for testing
two parallel data sets.49 The first three tests are nonparametric and consequently can be used for
nonnormal data sets which frequently occur in air quality data.
2.7.2 Treatment of Outliers
Outliers may result from errors in the processing of a data set, instrument problems, and
calibration errors. Once an outlier has been identified using any of the approaches identified above,
treatment of the outlier must be decided. The outlier should not be arbitrarily dropped from the data
set. Outliers that are found to be errors should be corrected, if possible. If the correct value cannot be
obtained, the value may be appropriately flagged and excluded from the data set. Alternatively, if the
suspect data are retained in the data set the necessary qualifying information in the form of a "flag" must
be included with the value. There should be an explanation that warrants the exclusion or replacement
of data along with documentation reflecting the action taken. If no explanation is available, the outlier
should not be excluded.
The data point may be real. Data should only be excluded by the reporting agency when the values are
verified as not representative of ambient data, such as calibration runs, instrument malfunction,
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 189 of 225
contamination, etc. Verification should include site or operator log book, laboratory notes, etc. Later
data analysts may then choose to use the outlier or not, depending upon their analyses.
2.8 Quality Control and Quality Assurance for VOC Measurements
Data submitted to AIRS by all agencies will ultimately be used by the agencies to develop,
evaluate, and refine new O3 control strategies; determine NAAQS attainment or non-attainment for O3;
track VOCs and NOX emissions inventory reductions; provide photochemical prediction model input;
evaluate photochemical prediction model performance; analyze ambient air quality trends; and
characterize population exposure to VOCs and O3. Data from VOC measurement systems must be
submitted to the Aerometric Information Retrieval System (AIRS) within six months following the end
of each quarterly reporting period, and the data must be consistent with enhanced O3 monitoring
regulations and of sufficient quality to meet Clean Air Act Title I objectives.
The quality of the data submitted to the AIRS data base must be consistent across all
agencies. Because a significant investment of time and assets is expended to generate measurement
data, a quality control/quality assurance (QC/QA) program should be developed to ensure that the data
collection is consistent and that data quality objectives (DQOs) for the measurement program are met.
The quality program for VOC measurements, similar to programs for other air monitoring efforts,
incorporates quality control and quality assurance. These two systems work together to achieve the
goal of continuing quality in measurement efforts.
Quality assurance and quality control have been defined and interpreted in different ways.
Some sources differentiate between the two terms by stating that quality control is "the operational
techniques and the activities which sustain a quality of product or service (in this case, good quality
data) that meets the needs; also the use of such techniques and activities," whereas quality assurance is
"all those planned or systematic actions necessary to provide adequate confidence that a product or
service will satisfy given needs."55
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 190 of 225
Quality control may also be considered "internal quality control;" mainly, routine checks
included in normal internal procedures; for example, periodic calibrations and system blanks, duplicate
checks, and split samples. Quality assurance may also be viewed as "external quality control," those
activities that are performed on a more occasional basis, usually by a person outside of the normal
routine operations; for example, on-site system surveys, independent performance audits,
interlaboratory comparisons, and periodic evaluation of internal quality control data.56
EPA's Quality Assurance Handbook56 uses the term quality assurance collectively to include
both quality assurance and quality control. A description of the elements necessary for a PAMS QA
program are found in 40 CFR Part 58, Appendix A "Quality Assurance Requirements for State and
Local Air Monitoring Stations (SLAMS)." Existing SIP stations are SLAMS; National Air Monitoring
Stations (NAMS) and PAMS are considered a subset of SLAMS. Appendix A specifies the minimum
QA requirements applicable to SLAMS air monitoring data submitted to EPA. States are encouraged
to develop and maintain QA programs more extensive than the required minimum. The references
found at the end of this document are also found in the 40 CFR Part 58, Appendix A for reference
purposes in preparing quality assurance program documents.
This section is divided into three parts: DQOs, QC, and QA. Section 2.8.1 discusses the
development of data quality objectives. Section 2.8.2 deals with the systematic activities of a QC
program and Section 2.8.3 presents those assessment activities that are part of a QA system. The
QC/QA procedures discussed in the following sections are the minimum and must be tailored and
expanded for each PAMS monitoring network as appropriate. A clearly written QC/QA plan and
associated SOPs must be developed for each network.
2.8.1 Data Quality Objectives
Data quality objectives are defined in the PAMS Implementation Manual25 as "statements
that relate the quality of environmental measurements to the level of uncertainty that decision-makers
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 191 of 225
are willing to accept for results derived from the data." The development of DQOs starts with the
monitoring program objectives and goals. In order to develop DQOs for each program objective, it is
first necessary to narrow each program objective to one or more specific monitoring or data objectives.
Specific and often different program objectives are associated with each specific PAMS
network. The overall network should supply information sufficient to develop, evaluate, and refine new
O3 control strategies; determine NAAQS attainment or non-attainment for O3; track VOCs and NOX
emissions inventory reductions; provide photochemical prediction model input; evaluate photochemical
prediction model performance; analyze ambient air quality trends; and characterize population exposure
to VOCs and O3. The program objectives are discussed in detail in the PAMS Implementation
Manual25 and are classified into six general categories given below:
Category 1 Responsible and cost-effective control strategies;
Category 2 Photochemical modeling support;
Category 3 Reconciliation of emissions inventories;
Category 4 Ozone and precursor trends;
Category 5 Attainment and non-attainment decisions; and
Category 6 Population exposure analyses.
The DQOs must quantify the measurement variability in order for the risk in decision-making
to be adequately assessed. This quantification of variability can only occur if there exists a base level of
experience in use of the technologies and /or methods outlined for the program. In the case of PAMS,
until its implementation in 1993 there had never been a program of this scope, with these goals and
program objectives. Data compiled from the 1990 Atlanta Study57 and other studies of this type
provided the information for initial development of the PAMS DQOs.
It is important to note that all possible uses of the PAMS data are not known; therefore,
every practical attempt should be made to improve data quality beyond that necessary to satisfy the
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 192 of 225
DQOs specified for PAMS. The DQOs are considered preliminary and are expected to be revised as
improvements are made to the monitoring and statistical methods; as changes and/or additions are
made in the program objectives or in the data use; and/or as results of the monitoring indicate a need.
The PAMS DQOs are discussed in detail in the PAMS Implementation Manual25 and highlighted
below in relationship to each monitoring program objective category :
DQO 1.1 The data for any given pollutant measured at a PAMS site must be able to show
the presence of a diurnal pattern, if a pattern exists, with an 80% confidence level.
DQO 1.2 The data for any given pollutant measured at a PAMS site must be able to show
a change in the diurnal pattern, if a change exists, with an 80% confidence level.
DQO 2.1 The speciated VOC, ozone, NOX and meteorological data must satisfy the
regulations, including monitor siting, operation, and data quality criteria.
DQO 3.1 The monitoring data for total VOC concentrations collected at a Type 2 site must
be able to demonstrate a 3% annual trend (upward or downward) over a 5-year
monitoring period, if a trend exists, with an 80% confidence level.
DQO 3.2 The speciated VOC monitoring data collected at a Type 2 site, when composited
into categories, must be able to demonstrate a 20% change (upward or
downward) in the seasonal average between two consecutive years, if it exists,
with an 80% confidence level.
DQO 3.3 The speciated VOC data collected at a Type 2 site must be able to detect ratios
between key species that are indicative of specific sources. If the ratio between
the emissions of two species is N: 1, the compounds must be measured so that
this ratio can be estimated to within ±50% with an 80% confidence level.
DQO 4.1 The composite monitoring data for a given MSA, CMS A for ozone, NOX, and
speciated VOC must be able to demonstrate a yearly downward trend with an
80% confidence level until an area achieves attainment.
DQO 5.1 The ozone (and NO2 where appropriate) monitoring data must satisfy the criteria
specified in the NAMS and SLAMS monitoring regulations, including monitor
siting, operation, and data quality criteria.
DQO 6.1 The speciated VOC monitoring data must be able to provide annual average
concentration data at Type 2 sites within ±50%, with a confidence level of 80%.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 193 of 225
DQO 7.1 Meteorological data should be of sufficiently high quality that the relationship
between ozone and wind speed, direction, and solar radiation can be determined.
This level of quality can be shown if a functional relationship of ozone to
meteorology explains a definite percentage of the ozone variation. Wind direction
measurements must be of sufficient quality to develop a wind field for use in
trajectory analysis.
The set of DQOs presented above assumes that the physical location of the site remains the
same for the entire five-year period under consideration and that the area external to the site does not
change in such a way that the appropriateness as a Type 2 site is impacted. Should either of these
conditions occur, the DQOs must be modified to reflect the change in site conditions.
To meet the overall PAMS monitoring program DQOs above, specific quality objectives for
VOC measurements are expressed in terms of sensitivity (detection limit), accuracy, precision, and
completeness. The measurement quality objectives should not require more than the sampling and
analytical procedures can provide, nor should they be based solely on the performance capability of the
measurement system and methodology. The procedures used to assess the DQOs should be specified
so that QC and QA procedures can be assembled to determine DQO attainment. If DQOs are not
routinely met, a rapid resolution of the problem is required. Possible resolutions include redefinition of
the DQO, repair/replacement of instrumentation, and/or modification of the methodology and
procedures.
Detection limits reflect the smallest measured concentration of a compound that can be
measured with a known degree of certainty and should be based on empirical data from the sampling
and analytical system used for VOC measurements. The limit of detection is defined in the Federal
Register58 as the minimum concentration of a substance that can be measured and reported with 99%
confidence that the analyte concentration is greater than zero and is determined from analysis of a
sample in a given matrix containing the analyte. Definitions for the detection limit should be related to
the standard deviation of the measured values at or near zero concentration of the analyte.59 Detection
limit information should not be used to screen air measurement results. All results, even at
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 194 of 225
concentrations below the estimated detection limit, should be reported. More information is gained
when a result is reported even if the data are somewhat imprecise.
Method detection limits (MDLs) for PAMS are estimated using 40 CFR, Part 136,
Appendix B.60 Replicate analyses are performed at least seven times at very low ppbC concentrations
(within a factor of five times the estimated method detection limit). The standard deviation of the
concentration for the replicate samples is related to the Student's t-value at the 99% confidence level
with a standard deviation estimate having n-1 degrees of freedom. The detection limit is calculated by
multiplying the standard deviation by the appropriate Student's t-value. (Appropriate values for six,
seven, and eight replicates are 3.365, 3.143, and 2.998, respectively.) Specific MDLs are developed
for each measurement method at the start of a season by performing a detection limit study. The
estimated detection limit for specific target VOCs is 3 ppbC or better.
Accuracy is the relative closeness of a measurement to a known reference value. The
assessment of accuracy includes both accuracy and precision and is usually expressed as bias or
percent bias:
0/ TV Measured Value - True Value , „„
% Bias ^^^^^^=^^^^^^= x 100 (2-12)
True Value
Bias is thus a signed value: i.e., the bias may be positive or negative.
Accuracy can also be expressed as percent difference, relative percent difference, percent
recovery, and percentage deviation. Estimates of bias require the analysis of a reference material or
standard of known or established concentration. Reference standards are submitted in canisters as
"blind" audits by an internal or external laboratory. A "blind" audit usually refers to a sample that has
been submitted to the audited laboratory and identified as an audit sample of unknown concentration.
The reference value for an accuracy standard should be a certified reference material (CRM) or
traceable to a standard reference material such as a NIST Standard Reference Material (SRM).
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 195 of 225
Accuracy should be determined for as many target VOCs as practical based on individual testing needs
consistent with the standard operating procedure at the individual PAMS site. Federal regulations
require State and Local air monitoring agencies to perform annual accuracy checks. In the absence of
specified objectives, the absolute accuracy should be within ±25% of the reference value.
Precision refers to the agreement among a group of experimental measurements made under
identical conditions. The most commonly used estimate of precision for environmental measurements is
standard deviation, or the square root of the variance. The sample standard deviation (typically used
for less than 30 observations) is calculated as follows:
. (Y-Y)2
n- 1
where:
Y = means of all observations
Y; = ith value of observations
degrees of freedom
To make a comparison of two values (i.e., duplicates or replicates), Relative Percent
Difference (PRO) is a more meaningful statistic than RSD, since the number of values is only two.
V2- V,
RPD
x 100
Where:
V,
= second determination
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 196 of 225
= first determination
Other measures of precision include relative standard deviation, coefficient of variation, percent
difference, range, and relative range. Precision is an inverse relationship, i.e., the smaller the measure of
precision, the better the agreement among measurements. For manual canister sampling approaches,
two different evaluations of precision can be made: the repeated analyses of duplicate canister samples,
which provide an assessment of the total method precision including elements of imprecision in both the
sampling and analytical procedures; and/or replicate analysis of a single canister sample, which provides
only analytical precision. Use of these two precision estimates is valuable in determining the source of
imprecision in the measurement effort.
Since automated GC systems employ "real-time" sample collection and measurement
techniques, estimates of precision require repeat measurements of single or collocated SUMMA®
canister samples that have been collected using an external sampling device. A replicate sample is a
sample that has been divided into two or more portions, at some step in the measurement process.
Collocated samples are individual samples collected so they are equally representative of the variable(s)
of interest at a given point in space and time. Information gathered from collocated sample results
allows for estimation of sampling precision. Repeated analyses of collocated, or duplicate, samples
permits estimation of the sampling and analytical precision. Precision estimates should therefore
represent the variability of the entire measurement system. Collocated samples are recommended,
when possible, for assessing sampling and analytical precision.
Another measure of precision involves intra-laboratory and inter-laboratory precision.
Collocated samples that are processed and analyzed by the same organization provide intra-laboratory
precision information for the entire measurement process. Collocated samples that are processed and
analyzed by different organizations provide inter-laboratory precision information for the measurement
process. A sample exchange program that involves both inter-laboratory or intra-laboratory precision
gives important information concerning inconsistencies that may exist. Interpretation of these data must
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 197 of 225
be based on clear understanding and knowledge of how the data were obtained. Differences in the
methodologies (i.e., detection limits, analytical column, calibration procedures, etc.) used to analyze the
exchange sample must be clarified in order to interpret and resolve any inconsistencies in the results.
Precision for inter- and intra-laboratory exchange samples is calculated in the same manner as precision
for replicate analyses. In the absence of specified DQOs, objectives for precision should be
determined from the QA program, pre-measurement system verification, and historical information for
the target compounds of interest. In the absence of specified objectives, values for precision are
considered acceptable if they fall within ±25% RPD. This 25% target applies to both automated
methodology and manual sampling (duplicate prepared canisters or replicate analysis of a single
canister).
When evaluating the precision of VOC measurements, states or agencies must consider each
individual target compound because precision will be compound-dependent with an influence of
physical and chemical properties (such as vapor pressure and reactivity). In reviewing species data
pairs (primary and duplicate samples), the number of non-detects in both samples will probably be
significant. In these "non-detect pairs," the RPD will not be useful. Instead these pairs can be said to
have a qualitative precision. Data pairs where the compound is detected in both samples can be
evaluated for relative percent difference, with a goal of ±25%. In the evaluation of the data, it should
be noted that there will be a large range of concentrations and that compounds with an average
concentration near the method detection limit will probably exceed the goal of 25%.
Specific requirements for precision and accuracy of automated and manual methods are
contained in Sections 3.1 through 3.4 of Appendix A, 40 CFR 58. The calculations used for precision
and accuracy data quality assessments are given in Section 5 of 40 CFR 58, Appendix A.
Completeness is the percentage of valid data reported compared to the total number of
attempted field samples, minus blanks, standards, and scheduled audit analyses. Completeness is
determined after the data validation process is complete, and after precision and accuracy are
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 198 of 225
determined and evaluated against the quality objectives. The objective for completeness is 85% and is
determined on an annual basis.
2.8.2 Quality Control
General QC guidance can be found in the EPA QA Handbook.56 Quality control for
measurement programs covers topics from preventive maintenance to corrective actions. Four areas of
particular importance to VOC measurements described in this section are sample collection, sample
handling and custody, sample analysis, and data documentation and archiving.
2.8.2.1 Sample Collection
Quality Control for sample collection should address: certification of the sample collection
system, calibration of the system components, field acquisition of duplicate samples, and preventive
maintenance efforts. A table of QC objectives for sample collection is given in Table 2-16. Technical
information pertaining to manual multiple-event and single-event VOC sample collection systems is
presented in Section 2.5. Similar information pertaining to automated GC systems is presented in
Section 2.4.
-------
Table 2-16. QC Objectives for VOC Sample Collection
Assessment
Sampling System
Carry-over
Sampling System
Background or
Contamination
Accuracy of
Collection Period
Sampling Integration
Period
Sampling System
Pressure/V acuum
Measuring Device
Accuracy
Duplicate Sample
Correction Precision
QC Procedure
Challenge with target
compounds
Humid zero air blank
Elapsed time meter
or timing device
check
On/off timer check
Flow control check
Pressure/vacuum
gauge or electronic
sensor check
Comparison of
duplicate canister
sample results
Frequency
Annual
Annual
6 Months
Quarterly
Weekly
Annual
10% of field
samples
Acceptance Criteria
80-120% recovery for target
compounds, overall
compound recovery of 85-
115%
2 ppbC or the MDL,
whichever is less for target
species or < 10 ppbC
TNMOC
Gain or loss in time
< 2 minutes per 24-hour
period
Measured transfer standard
flow within 10% of indicated
flow
< 10% difference between
field and lab measured
canister pressure
Agreement within ±25%
RPD.
Corrective Action
1) Additional system purge
with humid zero air
2) Repeat challenge
1) Additional system purge
with humid zero air
2) Repeat zero air
collection
Adjust or replace the timing
device
Adjust or replace timer
Adjust or replace flow control
device
1) Adjust for differences in
pressure/vacuum
measurement technology
2) Repeat check
1) Perform sampling
system PM
2) Repeat duplicate sample
collection
3) Check analytical system
precision
4) Check canisters for
leaks.
OQ 5-
8 R
^ o
VO VO
o o
H-h ^
ho ^>
00
^ to
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 200 of 225
2.8.2.1.1 System Certification
Canister sampling systems should exhibit non-biasing characteristics before being used to
collect samples. These sampling systems should be subjected to laboratory certification to quantify any
additive or subtractive biases that may be attributed directly to the sampling system. The procedure is
described in Section 2.5.1.7, Canister Sampling System Certification. Collection system certification
should be conducted prior to and after each PAMS season or at the start and end of each calendar
year. The percent recoveries for target challenge compounds are calculated, based on the determined
reference sample concentrations. Recoveries of each of the challenge compounds should be in the
range of 80-120% of the concentrations determined for the reference sample. A system-specific
overall recovery should also be calculated. The overall recovery is the average of the individual
compound recoveries. Each sampling system should have an overall recovery of 85-115%.
In addition to characterizing the sampling system with a blend of VOCs, the system should
also be characterized using humidified zero air. A humidified zero air blank sample is collected through
the sampling system to further gauge the potential for additive bias. The blank samples can be analyzed
for specific target analytes, total NMOC, or both, depending on individual program requirements. Two
criteria apply to the blank portion of the certification process: a determined concentration criterion of
3 ppbC or less for any individual target compound is required if speciation analysis of the blank sample
is performed, and a total NMOC concentration criterion of 10 ppbC or less is also required.
2.8.2.1.2 Calibration of Manual Sampling System Components
A QC check for the timing device in most manual sample collection systems is described in
the Quality Assurance Handbook.56 Every six months the elapsed-time meter should be checked
against a timepiece of known accuracy. If the meter shows any signs of temperature-dependence, it
should be checked on site during each season of the year. A gain or loss greater than 2 min/24-hour
period warrants adjustment or replacement of the indicator.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 201 of 225
The on/off timer should be calibrated and adjusted quarterly by using a calibrated
elapsed-time meter as the reference. Specific procedures for QC of the elapsed-time meter and on-off
timer calibration checks are provided in Section 2.2.2 of the Quality Assurance Handbook.56
The accurate and consistent control of flow into the sample canister is required to ensure that
the sample collected is time-integrated. To check the sample collection flow rate, a flow meter
(e.g., calibrated transfer standard) is installed at the inlet to the manual collection system and the system
activated. The flow rate measured by the transfer standard should agree within ±10% of the flow rate
indicated by the flow control device housed in the sample collection system. This check should be
performed annually or as needed based on performance. The ability of the collection system to
consistently perform an accurate time-integration can be assessed indirectly by making a control chart
of the sample pressure collected over each sampling event. If the collection flow rate and sample
collection period are not varied, the total pressure of samples collected should not differ by more than
±10%, assuming that the canisters all start with the same initial vacuum.
The vacuum/pressure measuring devices (e.g., gauges or electronic sensors) used in the
manual sample collection systems must be checked for accuracy. To perform this check, a series of
comparisons between a primary pressure measuring standard and the sample collection system device
should be conducted prior to installation and then annually. The comparisons should cover the range of
the device operation at increments representing 10% of the total range. Ensuring the accuracy of the
pressure measuring device also allows the sample canister to be assessed for leaks. The final sample
pressure in a canister as measured in the field should be within 10% of that measured when the canister
is received for analysis at the laboratory. Note that differences in pressure may occur due to
temperature and barometric pressure changes, and differences in the accuracy of the pressure
measurement technology used.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 202 of 225
2.8.2.1.3 Collection of Field Duplicate Samples
Field duplicate samples should be collected at a scheduled frequency of at least 10% (or at
the frequency specified by the site standard operating procedure) and are used to estimate the precision
of the manual sample collection method. Most commercially available manual sample collection
systems allow for collection of duplicate samples. Care should be taken to ensure that the duplicate
samples represent the same parcel of air over the same sample integration period.
Automated sample collection and analysis configurations have the same general QC
requirements for sample collection as manual systems. All of the QC checks for automated systems
should be done at the field station. Because true duplicates cannot be collected and analyzed with an
automated GC system, precision is estimated using repeated analyses of the target VOCs at ambient
concentrations. A humidified canister standard or sample is typically used as the source instead of the
sample manifold. This check is actually a measure of analytical precision (replicate analysis of a single
canister rather than duplicate analysis of multiple canisters). Replicate results should agree within 25%
RPD.
2.8.2.1.4 Preventive Maintenance
Preventive maintenance is an important part of the overall QC program for both manual and
automated sample collection systems. Maintenance items are generally specified by each sample
collection system manufacturer in the operating manual. These items may include any moving parts such
as valves or pumps. Most manual sample collection systems have an in-line paniculate filter which
needs to be replaced on a regular basis. The location and physical conditions of the sample collection
system may dictate other maintenance activities that are necessary to reduce the effects of heat, dust,
corrosion or other concerns.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 203 of 225
Any maintenance activity that involves the disassembly of hardware and replacement of parts
should be viewed as a potential change to the performance of the system. Replacement of major
sample collection system components (e.g., a flow control device) may warrant recertification of the
sample collection system. Duplicate analysis of multiple component calibration standard samples can
be used to assess whether changing a major component has affected the performance of the collection
system. If the duplicate analysis results compare within the quality objectives for the program, the
sample collection system does not require recertification. If duplicates do not meet the quality
objectives, then the sample collection system should undergo full challenge and blank recertification.
Repeated analyses of a multiple component calibration standard for the automated GC should also be
conducted and reviewed to check for shifts in retention time or changes in response factors that may be
caused by a maintenance activity.
Quality Control activities should be thoroughly documented in a log book dedicated to the
monitoring site. In addition to the technical details of the site maintenance activity, the time, date,
sample collection system or instrument ID, and monitoring site ID should be recorded.
2.8.2.2 Sample Handling and Custody
The QC procedures for canister preparation are vital to manual sample collection and
calibration standard preparation because all of these activities rely on leak-free uncontaminated
canisters. All canisters prepared for field use should initially be checked for leaks by pressurizing the
canister with zero air to a known measured pressure. Gross leaks are located by coating critical
canister surfaces considered to have leak potential (e.g.,valves and fittings) with a leak detection
solution. More subtle leaks are indicated by pressure changes in the canister over time. All canisters
should be cleaned and checked for contamination and certified clean to 10 ppbC TNMOC. Canister
cleaning procedures are described Section 2.5.2.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 204 of 225
Sample documentation includes chain-of-custody for canister samples and proper sample
identification and labeling. A chain-of-custody protocol should be developed so that at any point
between the canister's initial cleaning and its disposition after analysis the sample custodian can identify
and track the status of the canister. A unique identification is required for each canister at each point in
the sampling event. The canister should have a permanently assigned serial number and, after the
sample is collected, a unique sample identification number. A tag or label should be attached to the
canister so that it can easily be identified. The chain-of-custody form should always accompany the
canister and provide the means for each person responsible for custody to relinquish the canister to the
next person handling that canister. For example, the laboratory technician who provides the cleaned
canister to the field technician should initiate the chain-of-custody form and sign the canister over to the
field technician. After the sample is taken, the field technician returns the canister to the laboratory, with
the appropriate custody forms indicating the shipper and destination. The final completed form, upon
receipt at the laboratory, is signed by the sample custodian. This record allows the history of each
sample to be reconstructed if a problem arises with the analytical results.
A communication protocol should be established between the field sampling personnel and
the analytical laboratory personnel to ensure that sample canisters arrive at the monitoring locations
ahead of the scheduled sampling date. Sufficient numbers of canisters should be available to collect all
required samples, including any blank or duplicate samples that may be scheduled. The communication
protocol should include how to return the sample canister to the laboratory after collection.
For automated GCs, sample documentation can be accomplished using instrument specific
data collection software. Each chromatogram should have a header that uniquely identifies the sample
(e.g., filename and sample ID) as well as notation of the analysis conditions and column(s) used. Good
maintenance records are very important for automated GCs due to the large volume of data produced.
An injection or sample collection logbook should be maintained to provide a history for each analysis
so that any questions about results can be resolved.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 205 of 225
A standardized approach should be followed for identifying samples, blanks, calibration runs,
audits, and other analyses. All samples collected and analyzed with an automated GC should have a
unique file name designated to identify the site, instrument used to collect the sample, and the sampling
date and time. A data system should automatically append a characters) to the end of the electronic
storage file which corresponds to the order in which the sample was analyzed. The user is typically
limited to eight characters.
Electronic copies of all original and reprocessed files should be maintained to provide a
record of any and all changes made to the data. The original raw data files should never be changed
and should contain, if possible, a file name extension (such as .RAW) to differentiate them from other
files. The .RAW files should be archived with any processed data files. Data processing can be
checked at a later date.
2.8.2.3 Sample Analysis
Several steps are taken to ensure that the analytical system is in control, and these steps apply
to GC operations whether the sample is collected using an automated or a manual method. A summary
of these quality objectives is shown in Table 2-17. These objectives are the minimum QC procedures
pertaining to VOC analyses. States are strongly encouraged to develop more detailed, site-specific
SOPs.
During the initial analytical system set-up a multiple point calibration check using propane
and/or benzene is performed, and retention time windows are determined using the retention time
standard for each target compound. Calibration is the single most important operation in the
measurement process. Calibration is the process of establishing the relationship between the output of
a measurement process and a known input. For routine operation, the retention time calibration check
samples are analyzed to demonstrate that the retention times for each target VOC are within the
established window, to monitor the detector response drift, and verify the target compound recoveries.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 206 of 225
Results outside of the expected retention time window or detector response range may indicate the
need for maintenance, adjustment or recalibration of the analytical system before further sample
analysis. A humidified zero air blank should be analyzed after the highest level calibration standard to
assess analytical system carry-over of any target analytes.
-------
Table 2-17. Continued
Table 2-17. VOC QC Procedures
Assessment
System
Background and
Carry-over
Calibration
Quantitative and
Qualitative
Performance
Qualitative
Performance
Detection Limit
QC Procedure
System Blank
Analysis, Humidified
Zero Air
Multiple Point
Calibration (3 points
minimum).
Propane/benzene
bracketing the
expected sample
concentration
Retention
Time/Calibration
check using
mid-point of
calibration curve
Canister cleaning
certification
40CFR136PartB
Frequency
Weekly, following
retention time/calibration
check and after multiple-
point calibration curve
Prior to analysis at start
of season and when
system maintenance is
performed
Weekly
All canisters prior to use
Prior to analysis at start
of season
Acceptance
Criteria
20 ppbC total, both
analytical columns,
or lOppbC per
column
Correlation
Coefficient > 0.995
RF within 10% RPD
of calibration curve
average RF
RT within ±0.1
minutes of target
% recovery for
targets 80-120%
< 10 ppbC total
2 ppbC or better,
specific target peaks
selected
Corrective Action
1) Repeat analysis
2) Check system for leaks
3) Clean system with wet
air
4) Condition sample trap
1) Repeat individual
sample analysis
2) Repeat linearity check
3) Prepare new
calibration standards
and repeat
1) Repeat check
2) Repeat calibration
curve
Reclean canister and
reanalyze
N/A
s
to o
O VO
o o
*+> ;>
^ to
-------
Table 2-17. Continued
Assessment
Precision
Accuracy
QC Procedure
Replicate sample
analysis, manual or
automated
Performance
evaluation or NPAP
sample analysis
Frequency
10% of samples
Prior to start of season,
and monthly during
monitoring season
Acceptance
Criteria
Within ± 25% RPD
for target
concentrations > 5
times the MDL
20% absolute bias
Corrective Action
Repeat sample analysis
Repeat sample analysis
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 209 of 225
After the system is optimized as described in Section 2.4.3.1, analytical system blanks are
analyzed to verify that the analytical system contributes no more than the maximum specified
background level of a total 20 ppbC (10 ppbC per analytical column) to the analytical results. System
blanks are typically run with each analysis batch (or at a specific time of day for continuous GCs but not
the same time every day to avoid loss of data for the same time period for the entire season), and
usually in a predetermined order, such as the first or last analyses in a batch, after a calibration check
standard, etc. The hour chosen for system blank analysis on automated systems should be varied so
the data from the same hour are not consistently lost. Each blank must be humidified to ensure that the
analytical system responds similarly to ambient samples.
A calibration control standard or calibration curve should be analyzed to assess qualitative
and quantitative performance of the sampling system. Sample analyses should not be attempted until
the results demonstrate performance within the program's acceptance criteria. This type of control
sample should also be analyzed with each batch (or at various times of day for automated GC) and in a
predetermined order, usually before the analytical system blank. Again, the analysis hour should be
varied for automated GC systems. A calibration check sample is analyzed at least weekly to verify
calibration control.
An additional QC check of retention time windows for all target analytes should be
performed on both manual and automated GC systems using a retention time standard. This standard
should be used at program start up and weekly or as needed thereafter to provide a comprehensive
check of retention time windows and peak identifications for all target VOCs. The retention times
should not vary by more than 0.1 minutes. The results from repetitive analysis of this standard can
indicate trends in quantitative performance of the GC. If independent verification of the vendor
supplied cylinder can be performed, then the cylinder can also be used as the quantitative QC check
discussed in the preceding paragraph.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 21 Oof 225
The QC check standard may also be used to measure percent recovery for the target
compounds. Compound recovery can be used to verify sample trapping and transfer efficiency. At a
minimum, one target compound representing each carbon group should be checked for recovery
between 80 and 120%.
Replicate analyses are valuable in providing a means to estimate analytical precision. From
these data, control charts may be generated and simple statistical analyses performed on the analytical
results for each compound to estimate precision. The calculations are typically made in units of relative
percent difference or relative standard deviation (percent coefficient of variation). Variability in the
measurement data will be higher at low concentrations, which is also the range in which most ambient
VOC data are measured. For these reasons, precision measurements are made at ambient
concentrations, even though variability (imprecision) increases as the detection limit is approached.
Manual sampling systems provide sufficient sample in each canister to perform replicate analyses.
Repeated analysis should be performed on 10% of the samples to determine if the analytical precision is
within ±25% RPD for compounds greater than 5 times the MDL.
To determine precision for automated GC systems, analysis of a humidified gas standard or
typical field sample is required. The sample is composed of target compounds at ambient
concentrations in humid air. A control check sample from a gas cylinder or a diluted aliquot in a
SUMMA® canister can be used for this purpose.
2.8.2.4 Data Documentation and Archives
Data documentation protocols are vital for reporting results in a consistent manner and for
providing an audit trail. Identification of samples must be consistent throughout the documentation
process. Logbooks should be maintained for each system in the laboratory, including logs for
calibration standard, instrument run logs, and maintenance logs. Good laboratory practices should be
followed for logbook entries, including the use of indelible ink, corrections made only by single line
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 211 of 225
strikeout (initialed and dated), and identification of authorship for each entry. These logbooks should
be periodically peer reviewed by one person and evidence of the review should be written in the
logbooks.
Electronic logbooks may be used if certain precautions are taken. A prime concern for QC
of data documentation files is the ability to recover data in the event of equipment failure. Logbooks
may be kept in electronic format if desired, but a protocol specific to the use and backup of electronic
log entries must be established. All software and data files used in recording and archiving raw data
should be backed up so that an equipment hardware failure or tampering does not destroy the logbook
information. Once entered, electronic logbook files should have file attributes set to "Read Only."
Bound hard copy indices of each electronic notebook should also be maintained. Index entries should
include file name, date of entry, author's name and subject.
Logbooks can be kept in electronic form for calibrations, computer access (recording logins),
computer boot logs, etc. Among the advantages of using electronic logs is the capability for remote
access to the information contained in the logbook. An automated GC system can be designed to allow
personnel remote access to all monitoring data, including all supporting documentation.
As part of a QC program, the data should periodically be reviewed to look for anomalies in
the data. Examples of items that should be examined in the periodic QC or validation of data include
the following:
• Missing data;
• Consistency of identification of target compound between standards and samples;
• Concentrations of compounds within expected ranges and variances according to
weather conditions, traffic patterns, etc.;
• Identification of target compounds in the retention time check standard;
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 212 of 225
• Response factors from the primary calibration check standard within acceptable limits;
and
• Compounds detected in the blank within acceptable levels.
A complete discussion on data validation is found in Section 2.7.
Data validation cannot practically be performed on 100% of the compound identifications
and concentrations. The analyst should be responsible for cursory review of all chromatograms,
primarily for anomalies. The analyst should also review a small percentage of the files in depth to
confirm peak identifications and look for changes in methods that could affect automated data
processing. The key element to this activity is consistency to ensure that the same peak is identified the
same way to the extent allowable by the automated software in a batch reprocess mode. Exploratory
data processing software packages like MetaChrom® and VOCDat® may be used to validate the data
in more depth.
The processed results from raw data should be stored in a different directory on the
computer system. Reprocessed files should be compressed into storage files and given a unique file
name. The compressed file name may be used to designate that the automated software files have been
reprocessed and the level of reprocessing.
2.8.3 Quality Assurance
The procedures and activities associated with QC need periodic independent review and
assessment to ensure that the data meet the quality expectations of the program. This independent
review or auditing function should be done by an independent source with no daily involvement in the
operation of the sampling or analysis systems. In addition, independent review should be done on
DQOs and SOPs to ensure that the quality goals of the agency are reflected by the QC program. After
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 213 of 225
SOPs and the QC program are put into place, the QA function provides review of their adequacy and
implementation in meeting the DQOs of the monitoring program.
2.8.3.1 Development of Standard Operating Procedures
An SOP is written so that procedures may be performed consistently by everyone involved in
the monitoring program. The decision of whether an SOP is needed for a particular procedure is made
by answering two questions:
• Does the procedure significantly affect data quality?
• Is the procedure repetitive or routine?
If the answer to both questions is yes, an SOP is needed.
Few routine laboratory or field projects can be described completely in just one SOP;
several may be needed. In general, an SOP for each of several smaller segments should be written
instead of one large SOP for an entire operation.
At a minimum, written SOPs to support the VOC monitoring effort should be prepared for
the following activities:
• Sample collection;
• Sample analysis;
• Sample canister handling;
• SUMMA® canister preparation and blanking;
• Data handling;
• Sample identification and labeling requirements; and
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 214 of 225
Automated or manual data reporting.
Table 2-18 shows the suggested format for an SOP, including examples of items that should
be included in each section. The examples shown are only a few of the many that may be
-------
Section:
Revision:
Date:
Page:
2
1
09/30/98
215 of 225
Table 2-18. Format for Standard Operating Procedures
A. Technical Sections
Section
1 . Scope and Application
2. Summary of Method
3. Definitions
4. Interferences
5. Personnel Requirements
6. Facilities Requirements
7. Safety Precautions
8. Apparatus
9. Reagents/Materials
10. Samples/Sampling Procedures
1 1 . Calibration/Standardization
12. Analysis Procedures
13. Calculations
14. Data Reporting
15. Corrective Action
16. Method Precision and Accuracy
B. Quality
Section
1. QC Checks
2. QC/QA Controls
Typical Examples
Overview outlining purpose, range, sensitivity, acceptance
cntena
Overview describing sampling criteria and analytical methods,
method and instrumentation detection limits, reasons for
deviations from Federal Register methods
All acronyms, abbreviations, specialized terms
Sources of contamination
Educational level and training of intended SOP users, number
of operators required
Mobile analytical laboratory, air conditioning, type of
electricity
Special handling procedures; i.e., handling compressed gases,
hazard warnings, placed immediately before relevant part of
text
Larger items such as a meteorological tower, audit device, gas
chromatograph
All chemicals used, including distilled or deionized water;
grades of reagents and materials
Sample preparation, collection, storage, transport, and data
sheets
Preparation of standards and calibration curves, frequency,
and schedule
Standard and custom-tailored methods for all target analytes
Data reduction, validation, and statistical treatment, including
confidence levels and outliers
Selection criteria, format, equations, units
Criteria for initiation; individuals responsible
Tabular or narrative summary of DQOs
Control/Quality Assurance Sections
Typical Examples
Precision, accuracy, reproducibility, blanks, replicates, criteria,
and frequency summarized in tables
Audits, notebook checks, control charts and graphs; actions
to be taken when QC data approaches or exceeds QC limits
C. Reference Section
Standard reference methods, reports, SOPs, journal articles; avoid citing unpublished documents
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 216 of 225
covered. The SOP must be comprehensive and cover every step of the procedure in detail. The
author of a SOP tends to be extremely familiar with the material and may omit "obvious" information.
The written SOP should be reviewed extensively and executed by persons not intimately familiar with
the procedure to evaluate the completeness of the instructions presented in the SOP. Once the DQOs
have been established and SOPs are in place, QA measures can be planned and implemented to ensure
that DQOs are met. In the remainder of this section, program audits are described that guide QA for
sample collection, analysis, and reporting. Additional guidance for development of a VOC monitoring
QA program are discussed in the EPA QA Handbook,56 and this information should be used in
conjunction with the guidance presented here.
2.8.3.2 QA Program Guidance
The QA program plan and activities should be developed to catalog and assess the data
quality requirements of the program. The QA plan ensures that procedures are being implemented
correctly and that all of the QC and SOP requirements are being followed to collect, analyze, and
report samples that meet agency DQOs. The QA plan also provides guidance for regular assessment
of the quality program for the monitoring network. An integral part of the QA program is a regular
series of audits. The QA plan should provide guidance for audits, performed by an independent
person or group to formally evaluate if systems are in place and being followed in order to meet the
DQOs. The QA program should be a cooperative effort between independent auditors and program
auditees so that quality issues can be identified and, if necessary, appropriate corrective actions
implemented.
2.8.3.2.1 Audit Types
Three audit types are described in the following sections: technical systems audits,
performance audits, and data quality audits. These audits are used to determine whether data quality
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 217 of 225
objectives are being met. An audit may also uncover additional findings useful in improving the
monitoring program.
As defined in the QA Handbook (Volume n)61 the Technical Systems Audit is a
systematic, qualitative on-site review of the adequacy of the facilities, equipment, training, record
keeping, validation, reporting, sampling, analysis, and QC/QA procedures. This type of audit should
be used throughout the VOC monitoring program to ensure that the QA/QC program elements are in
place and being followed. Systems audits are normally done immediately before, or shortly after,
measurement systems are operational and should be repeated on a regularly scheduled basis, at least
annually.
Technical systems audits should cover areas specific to VOC monitoring networks, including
sampling system certification, QC check procedures, and record keeping. Sampling system
certification critical to the success of a VOC monitoring program is often overlooked. Sampling system
QC checks should be reviewed to determine if proper blank and challenge gas tests have been
performed on the sampling systems before they are used to collect samples. Checks for leaks in
sampling systems and canisters, vacuum/pressure gauge calibration, and flow meter testing should be
reviewed to confirm that the QC procedures were implemented and systems are ready for use.
Vacuum and pressure measurement checks should be audited with a MST-traceable high resolution
gauge. Changes in canister pressure can then be corrected for barometric pressure differences and final
corrected canister pressures can be reviewed to identify leaking canisters.
The technical systems audit investigates the specific elements of the measurement system to
ensure that established procedures are followed. With an emphasis on improving the program, the
technical systems audit evaluates the documentation and paper trail of a particular site to show that the
VOC monitoring is conducted according to the QA/QC program elements. An audit of this type is
conducted by an auditor familiar with VOC monitoring and the required types of record keeping used
in a VOC program. The auditor uses QA/QC guidance documents such as this document as well as
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 218 of 225
the site-specific SOPs to evaluate the site's measurement system. A specific checklist of required
elements is prepared and used as a guide for the auditor's review. Significant deficiencies are noted
and discussed at the time of the audit and any additional deficiencies are noted and resolved by the
generation of a list of corrective actions to be taken by the site. Additional suggestions for improvement
are made by the audit team and delivered as part of their systems audit report.
A Performance Evaluation Audit is a quantitative evaluation of the measurement system
and includes all associated data collection and analysis procedures. A Performance Evaluation Audit
involves the analysis of a reference material of known value. For both manual and automated VOC
sampling systems, the audit sample is usually a canister or compressed gas cylinder containing
humidified air with VOC target analytes at known concentrations. Performance evaluation audits
should be conducted on a regular basis specific to the program, at least once per season.
The EPA initiated a national performance audit program (NPAP) which applies to the VOC,
carbonyl, O3, and NO2 measurement systems used at PAMS monitoring stations. This program
mirrors the present EPA national performance audit program for SLAMS criteria pollutants. Agencies
collecting data in the PAMS monitoring network are required to participate in the national audit
program. The audit program uses performance evaluation samples which were field tested by the EPA
in 1993-1994. The EPA has established limits for acceptable performance on each type of
Performance Evaluation sample based on a statistical analysis of the results obtained from pilot tests.
The VOC and Performance Evaluation samples were pilot tested through a series of
proficiency tests. The VOC samples consist of 1.5 L compressed gas cylinders containing 10 to 50 of
the target VOC compounds. Participants in the audit program receive the Performance Evaluation
samples, determine the concentrations and report the results to the EPA. The EPA compares the
reported results to the expected results and issues a report to all participants.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 219 of 225
A Data Quality Audit exhaustively evaluates the information generated from data collection
through reporting. Procedures that are evaluated include raw data recording and transfer, calculations
and equations, documentation of data handling, reporting and completeness, comparability, and
discussion of QC indicators such as precision, accuracy, and representativeness. A Data Quality Audit
for PAMS should include an audit of the data entered into the AIRS. Data Quality Audits should be
conducted routinely on various components of the data generation system, and a comprehensive Data
Quality Audit should be combined with Technical Systems Audits once the program is generating data
on a regular basis.
The QA/QC Procedures outlined in the above sections serve as a guide to the process of
ensuring that the VOC measurements are conducted properly. Collection of data and the subsequent
submission of the data to the AIRS data base must be consistent across all agencies. The defined
sections of DQOs: Sample collection, certification, calibration, duplicates, maintenance, sample
handling and custody, and sample analysis are a necessary framework and starting point for site-
specific SOPs. Coupled with the documentation of the data, the technical and system audit program
provides a regular evaluation of the system, data collection, and analysis process. The QA/QC
program in this document establishes the program elements which are necessary for the successful
operation of a VOC measurements analytical program.
2.9 References
1. U.S. Environmental Protection Agency. Code of Federal Regulations. Title 40, Part 58.
Ambient Air Quality Surveillance, Final Rule Federal Register, Vol. 58, No. 28, February 12,
1993.
2. David, DJ. Gas Chromatographic Detectors. John Wiley and Sons, New York, NY,
1974.
3. Dietz, W.A. Response Factors for Gas Chromatographic Analyses. J. Gas Chromatogr.,
1967.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 220 of 225
4. Compendium of Methods for the Determination of Toxic Organic Compounds in Ambient Air.
Compendium Method TO-14. The Determination of Volatile Organic Compounds (VOCs)
in Ambient Air Using SUMMA ® Passivated Canister Sampling and Gas
Chromatographic Analysis. EPA-600/4-89/017, Research Triangle Park, NC: U.S.
Environmental Protection Agency, 1988.
5. Foulger, B. E. and P. G. Simmonds. Drier for Field Use in Determination of Trace
Atmospheric Gases. Anal. Chem., 51: 1089-1090(1979).
6. Burns, W. F., D. T. Tingey, R. C. Evans, and E. H. Bates. Problems with a Nafion®
Membrane Dryer for Drying Chromatographic Samples. J. Chromatogr. 269: 1-9(1983).
7. Pleil , J. D., K. D. Oliver, and W. A. McClenny. Enhanced Performance of Nafion® Dryers in
Removing Water from Air Samples Prior to Gas Chromatographic Analysis. J. Air Pollut.
Control Assoc. 37: 244-248 (1987).
8. Baker, B. B. Measuring Trace Impurities in Air by Infrared Spectroscopy at 20 Meters Path
and 10 Atmospheres Pressure. Amer. Ind. Hyg. Assoc. J. 35: 735-747 (1974).
9. Campbell, N. T., G. A. Bere, T. J. Blasko, and R. H. Groth. Effect of Water and Carbon
Dioxide in Chemiluminescent Measurement of Oxides of Nitrogen. J. Air Pollut. Control
Assoc. 32: 533-535 (1982).
10. Rasmussen, R. A. and A. K. Khalil. Atmospheric Methane (CFy: Trends and Seasonal
Cycles. J. Geophys. Res. 86: 9826-9832 (1981).
11. Gong, Qing, and K. L. Demerjian. Hydrocarbon Losses on a Regenerated Nafion®
Drier. J. Air Waste Manage. Assoc. 45: 490-493 (1995).
12. Ogle, L.D., D. A. Brymer, C. J. Jones, and R. L. Carlson. "Moisture Management Techniques
Applicable to Whole Air Samples Analyzed by Method TO-14, U; GC/MS Considerations."
In Proceedings of the 1993 U.S. EPA/A&WMA International Symposium: Measurement
of Toxic and Related Air Pollutants, Research Triangle Park, NC, 1993.
13. McClenny, W.A., K.D. Oliver, and E.H. Daughtry, Jr. Analysis of VOCs in Ambient Air
Using Multisorbent Packings for VOC Accumulation and Sample Drying. J. Air Waste
Manage. Assoc. 45: 792-800 (1995).
14. Cardin, D. B. and J. T. Deschenes. "A Cryogenless AutoGC System for Enhanced Ozone
Monitoring Using a Simplified Single Detector Approach." In Proceedings of the 1993 U.S.
EPA/A&WMA International Symposium: Measurement of Toxic and Related Air
Pollutants, Research Triangle Park, NC, 1993.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 221 of 225
15. Soroka, J. M., R. Isaacs, G. Ball, R. Singhvi, and T. Pritchett. "The Effect of Water on
Recoveries in Sorbent Tube and SUMMA® Canister Analysis." In Proceedings of the 1992
U.S. EPA/A&WMA International Symposium: Measurement of Toxic and Related Air
Pollutants, Research Triangle Park, NC, 1992.
16. Levaggi, D. A., W. Oyung, and R. V. Zerrudo. "Noncryogenic Concentration of Ambient
Hydrocarbons for Subsequent Nonmethane and Volatile Organic Compound Analysis." In
Proceedings of the 1992 U.S. EPA/A&WMA International Symposium: Measurement of
Toxic and Related Air Pollutants, Research Triangle Park, NC, 1992.
17. Oliver, K.D., J.R. Adams, E.H. Daughtry, Jr., W.A. McClenny, MJ. Yoong, M.A. Pardee,
E.B. Almasi, and N. A. Kirshen. Technique for Monitoring Toxic VOCs in Air: Sorbent
Preconcentration, Closed Cycle Cooler Cryofocusing, and GC/MS Analysis. Environ. Sci.
Technol. 30: 1939-1945 (1996).
18. Oliver, K.D., J.R. Adams, E.H. Daughtry, Jr., W.A. McClenny, M.J. Yoong, and M.A.
Pardee. Techniques for Monitoring Ozone Precursor Hydrocarbons in Air at Photochemical
Assessment Monitoring Stations: Sorbent Preconcentration, Closed-cycle Cooler
Cryofocusing, and GC-FID Analysis. Atmos. Environ. 30, 15: 2751-2757 (1995).
19. Miguel, M.G "A Comparison Study to Determine the Effects of Pressure, Relative Humidity,
and Canister Residence Time on NMHC Recovery Rates from Stainless Steel Canisters." In
Proceedings of the 1995 U.S. EPA/A&WMA International Symposium: Measurement of
Toxic and Related Air Pollutants, Research Triangle Park, NC, 1995.
20. Holdren, M.W., and D.L. Smith. Development of Procedures for Performance Evaluation
of Ambient Air Samplers for Volatile Organic Compounds. Final Report, EPA Contract
68-02-4127. Research Triangle Park, NC: U.S. Environmental Protection Agency, 1987.
21. Ochiai, N., S. Daishima, M. Morikawa, and C. Watanabe. "Stablility of Polar Volatile Organic
Compounds in Stainless Steel Canisters with Multi-Layer Pretreatment." Poster presented at
the 7997 U.S. EPA/A&WMA International Symposium: Measurement of Toxic and
Related Air Pollutants, Research Triangle Park, NC, 1997.
22. McClenny, W.A., S.M. Schmidt, and K.G. Kronmiller. "The Variation of the Relative
Humidity of Air Released from Canisters after Ambient Sampling." In Proceedings of the
1997 U.S. EPA/A&WMA International Symposium: Measurement of Toxic and Related
Air Pollutants, Research Triangle Park, NC, 1997.
23. Handbook of Chemistry and Physics. CRC Press: Boca Raton, FL (1974).
24. Deans, D. A New Technique for Heart Cutting in Gas Chromatography. Chromatographia,
1: 18-22(1968).
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 222 of 225
25. Photochemical Assessment Monitoring Stations Implementation Manual. EPA-454/B-
93-051. Research Triangle Park, NC: U.S. Environmental Protection Agency, 1994.
26. Oliver, K.D., J.D. Pleil, and W.A. McClenny. Sample Integrity of Trace Level Volatile
Organic Compounds in Ambient Air Stored in SUMMA®-Polished Canisters. Atmos.
Environ. 20: 1403-1411 (1986).
27. Holdren, M., D. Smith, and W. McClenny. Storage Stability of Volatile Organic
Compounds in SUMMA®-PolishedStainless Steel Canisters. Final Report, EPA Contract
No. 68-02-4127. Research Triangle Park, NC: U.S. Environmental Protection Agency,
1986.
28. McAllister R, D-P. Dayton, J. Rice, P. O'Hara, D. Wagoner, and R. Jongleux. Stability
Study - Final Report. Report to Scientific Instrument Specialists, Moscow, ID, 1988.
29. Kelly, T.J. and M.W. Holdren. Applicability of Canisters for Sample Storage in the
Determination of Hazardous Air Pollutants. Atmos. Environ. 29: 2595-2608 (1986).
30. Gohlson, A.R., R.K.M. Jayanty, and J.F. Storm. Evaluation of Aluminum Canisters for the
Collection and Storage of Air Toxics. Anal. Chem. 62: 1899-1902(1990).
31. Pate, Bruce, R.K.M. Jayanty, and M.R. Peterson. Temporal Stability of Polar Organic
Compounds in Stainless Steel Canisters. J. Air Waste Manage. Assoc. 42 (4): 460-462
(1992).
32. Coutant, R.W. and W.A. McClenny. "Competitive Adsorption Effects and the Stability of
VOC and PVOC in Sampling Canisters." In Proceedings of the 1991 U.S. EPA/A&WMA
International Symposium: Measurement of Toxic and Related Air Pollutants, Research
Triangle Park, NC, 1991.
33. Goldan, P.O., W.C. Kuster, F.C. Fehsenfeld, and S.A. Montzka. Hydrocarbon
Measurements in the Southeastern United States: The Rural Oxidants in the Southern
Environment (ROSE) Program 1990. J. Geophys. Res. 100:25, 945-25, 963 (1995).
34. Roberts, J.M., F.C. Fehsenfeld, D.L. Albritton, and R.E. Sievers. "Sampling and Analysis of
Monoterpene Hydrocarbons in the Atmosphere with Tenax Gas Chromatographic Porous
Polymer." In Identification and Analysis of Organic Pollutants in Air, Butterworth, Boston,
1984,371-387.
35. Pellizzari, E.D., and KJ. Krost. Chemical Transformations During Ambient Air Sampling for
Organic Vapors. Anal. Chem., 56: 1813-1819(1984).
36. Pellizzari, E., B. Demian, and K. Krost. Sampling of Organic Compounds in the Presence of
Reactive Inorganic Gases with Tenax GC. Anal. Chem., 56: 793-798(1984).
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 223 of 225
37. Jiittner, F. A Cryotrap Technique for the Quantitation of Monoterpenes in Humid and
Ozone-Rich Forest Air. J. Chromatogr. 442: 157-163 (1988).
38. Bufler, U. and K. Wegmann. Diurnal Variation of Monoterpene Concentrations in Open-Top
Chambers and in the Welzheim Forest Air, F.RG. Atmos. Environ. 25A: 251-256 (1991).
39. Janson, R., and J. Kristensson. Sampling and Analysis of Atmospheric Monoterpenes.
Report CM-79, Department of Meteorology, Stockholm University, International
Meteorological Institute of Stockholm. Distribution by Library, Department of Meteorology,
Stockholm University, Arrhenium Laboratory, S-10691, Stockholm, Sweden.
40. Hoffmann, T. Adsorptive Preconcentration Technique Including Oxidant Scavenging for the
Measurement of Reactive Natural Hydrocarbons in Ambient Air. Fesenius Journal of
Analytical Chemistry, 351: 41-47 (1995).
41. Calogirou, A., B.R. Larson, C. Brussol, M. Duane, and D. Kotzias. Decomposition of
Terpenes by Ozone During Sampling on Tenax. Anal. Chem., 68: 1499-1506 (1996).
42. Helmig, D., and J. Arey. Organic Chemicals in the Air at Whitaker's Forest/Sierra Nevada
Mountains, California. Sci. Total Environ. 112:233-250(1992).
43. Mattsson, M. and G. Petersson. Trace Analysis of Hydrocarbons in Air Using Standard Gas
Chromatographic and Personal Sampling Equipment. Int. J. Environ. Anal. Chem. 11: 211-
219(1982).
44. Ciccioli, P., E. Brancaleoni, M. Possanzini, A. Brachetti, and C. DiPalo. Sampling,
Identification and Quantitative Determination of Biogenic and Anthropogenic Hydrocarbons in
Forestal Areas. In Physico-Chemical Behaviour of Atmospheric Pollutants, Proceedings
of 3rd European Symposium, Varese, Italy. 10-12 April 1984. Reidel, Dordrecht, pp 62-73
(1984).
45. Walling, J.F., I.E. Bumgarner, DJ. Driscoll, C.M. Morris, AE. Riley, and L.H Wright.
Apparent Reaction Products Desorbed from Tenax Used to Sample Ambient Air. Atmos.
Environ. 20: 51-57(1986).
46. Cao, X.-L., and C.N. Hewitt. Study of the Degradation by Ozone of Adsorbents and of
Hydrocarbons Adsorbed During the Passive Sampling of Air. Environ. Sci. Technol. 28:
757-762 (1994).
47. Bowyer, J.R. and J.D. Pleil. Supercritical Fluid Extraction as a Means of Cleaning and
Desorbing Common Air Sampling Sorbents. Chemosphere 31: 2905-2918 (1995).
48. Helmig, D. Ozone Removal Techniques in the Sampling of Atmospheric Volatile Organic
Trace Gases. Atmos. Environ. 31: 3635-3651 (1997).
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 224 of 225
49. McClenny, W.A. "Future Monitoring Techniques for VOCs." In Chemistry and Analysis of
Volatile Organic Compounds in the Environment. H.J. Th. Bloemen and J. Burn, Editors,
Blackie Academic and Professional, London, UK 1993, 237-267.
5 0. Interim Guidelines and Specifications for Preparing Quality Assurance Project Plans.
QAMS-005/80. Washington, D.C.: U.S. Environmental Protection Agency, 1980.
51. Validation of Air Monitoring Data. EPA-600/4-80-030. U.S. Environmental Protection
Agency, Research Triangle Park, NC, 1996.
52. Main, H.H., P.T. Roberts, and M.E. Korc. "Analysis of PAMS and NARSTO-Northeast
Data - Supporting Evaluation and Design of Ozone Control Strategies: A Workshop." Paper
presented at U.S. Environmental Protection Agency, Research Triangle Park, NC, December
1995.
53. Main, H.H., P.T. Roberts, J.D. Prouty, and M.E. Korc. Software for Display, Quality
Control, and Analysis of Continuous VOC Data. Report prepared for Electric Power
Research Institute, Palo Alto, CA, EPRI Research Project No. WO9108-01, June 1996.
54. Main, H.H., P.T. Roberts, and L.R Chinkin. "PAMS Data Analysis Workshop: Illustrating the
Use of PAMS Data to Support Ozone Control Programs." Paper presented at California Air
Resources Board and EPA Region IX, Sacramento, CA, May 1997.
55. Juran, J. M. Quality Control Handbook. Third Edition, McGraw-Hill, New York, 1974.
Section 2.
5 6. Quality Assurance Handbook for Air Pollution Measurement Systems Volume I -A Field
Guide to Environmental Quality Assurance. EPA/600/R-94/038a, U.S. Environmental
Protection Agency, 1994.
57. Holdren, M.W. and D. L. Smith. "Performance of Automated Gas Chromatographs Used in
the 1990 Atlanta Ozone Study." In Proceedings of the 1991 U.S. EPA/A&WMA
International Symposium: Measurement of Toxic and Related Air Pollutants, Research
Triangle Park, NC, 1991.
58. U.S. Environmental Protection Agency. Federal Register, Vol. 49, No. 209, October 26,
1984.
59. Parsons, M. L. The Definition of Detection Limits. J. Chem. Ed. 46: 290-292 (1969).
60. U.S. Environmental Protection Agency. Code of Federal Regulations. Title 40, Chapter 1,
Part 136, Appendix B. Office of the Federal Register, July 1, 1987.
-------
Section: 2
Revision: 1
Date: 09/30/98
Page: 225 of 225
61. Quality Assurance Handbook for Air Pollution Measurement Systems. Volume II -
Ambient Air Specific Methods (Interim Edition). EPA/600/R-94/038a. U.S. Environmental
Protection Agency, 1994.
-------
Section: 3.0
Revision: 1
Date: 09/30/98
Page: 1 of 2
Section 3.0
Determination Of Total Nonmethane Organic
Compounds Using Method TO-12
Qualitative and quantitative determinations of individual VOCs and measurement of total
NMOC using the GC based methodology described in Section 2.0 requires instrumentation that
is expensive, complex, and difficult to operate and maintain. Method TO-121 provides a similar
measurement of total NMOC, but does not provide information on the individual VOCs
comprising the total. Method TO-12 is part of the "Compendium of Methods for the
Determination of Toxic Organic Compounds in Ambient Air," and is presented in Appendix C.
Method TO-12 involves a simple preconcentration procedure with subsequent direct flame
ionization detection and provides accurate and sensitive measurements of total NMOC
concentrations. The instrumentation for this method can be configured for either automated
in situ measurements or for analyzing integrated samples collected in canisters.
Although Method TO-12 is not directly applicable to PAMS, the method is included here
because:
Method TO-12 is a viable, practical, and effective method of post clean-up
determinations of canister cleanliness;
Method TO-12 can be used for ambient total NMOC measurements as input into
O3 predictive models that do not require speciated VOC information; and
Used in combination with the manual (canister) methodology described in
Section 2.5 or in an automated form, Method TO-12 can be applied (i.e., with the
approval of the EPA Administrator) as a viable alternative monitoring approach to
the automated methodology described in Section 2.4.
Total NMOC data, resulting from measurements made using Method TO-12, should be
entered into the AIRS data base. These NMOC data are entered under Parameter Code 43102.
The abbreviation for parameter code 43102 is "TNMOC." The Method Code is 012. Method
-------
Section: 3.0
Revision: 1
Date: 09/30/98
Page: 2 of 2
Code 012 describes the sum of data gathered by preconcentrated direct flame ionization detection
PDFID). PDFID is the TO-12 EPA-approved method which includes not only the sum of C2-C12
data, but also any compounds larger than C12 that are detected.
3.1 References
Compendium of Methods for the Determination of Toxic Organic Compounds in
Ambient Air. Compendium Method TO-12. Method for the Determination of
Non-Methane Organic Compounds (NMOC) in Ambient Air Using Cryogenic
Preconcentration and Direct Flame Ionization Detection (PDFID). EPA-600/4-89/017.
Research Triangle Park, NC: U.S. Environmental Protection Agency. 1988.
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 1 of 32
Section 4.0
Methodology for Measuring Oxides of Nitrogen
And Total Reactive Oxides of Nitrogen in Ambient Air
Measurement of ambient concentrations of nitric oxide (NO) and nitrogen dioxide
(NO2) is a requirement of the 40 CFR Part 58, Subpart E,1 enhanced O3 network monitoring
program. The NO and NO2 measurements are used to better characterize the nature and extent of
the O3 problem, track oxides of nitrogen emission inventory reductions, assess air quality trends,
and make attainment/nonattainment decisions. Information on measuring NO and NO2,
including method and equipment descriptions, is presented in Section 4.1.
Although not specifically required under 40 CFR Part 58, Subpart E, measurement of
total reactive oxides of nitrogen (NOy) is strongly encouraged by the EPA. Measurements of
NOy constitute a valuable adjunct to current NO and NO2 monitoring because the individual
species comprising NOy include not only NO and NO2 but also other organic nitroxyl compounds
that have recently been shown to play a significant role in the photochemical O3 formation
process. Information on measuring NOy, including measurement principle and procedures and
equipment descriptions, is presented in Section 4.2.
4.1 Oxides of Nitrogen
Oxides of nitrogen, defined here as the sum of the concentrations of NO and NO2 at the
same point in time, are principal precursors to the formation of O3. The Urban Airshed Model
(UAM), another type of mathematical O3 prediction model, requires NO and NO2, total NMOC,
and speciated VOC concentrations as inputs.
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 2 of 32
4.1.1 Measurement Principle
NO and NO2 are typically measured using a chemiluminescence instrument. The
principle of operation of this instrumentation is based on the gas-phase reaction of NO and O3.
This reaction produces a characteristic near infrared luminescence with an intensity that is
linearly proportional to the concentration of NO present. Specifically,
(NO + 03) - (N02 + 02 + hv) (4-1)
where:
hv = emitted photon energy (function of Planck's constant and the
frequency of radiation)
Prior to measurement, nitrogen dioxide is converted into NO using a molybdenum (Mo) reducing
surface heated to 325 °C. Specifically,
(NO2 + Mo) - (NO + MoO3) (4_2)
where:
Mo = molybdenum chemical reductant
The reaction results in electronically excited NO2 molecules which revert to their ground state,
resulting in an emission of light or chemiluminescence.
To determine the concentration of NO, the sample gas is blended with O3 in a reaction
chamber causing the reaction to occur. The chemiluminescence that results from the reaction is
monitored by an optically filtered high-sensitivity photomultiplier. The optical filter and
photomultiplier respond to light in a narrow-wavelength band unique to the NO and O3 reaction.
The electronic signal produced in the photomultiplier is proportional to the NO concentration.
To determine the concentration of NOX (i.e., NO + NO2), the sample gas is routed
through an NO2-to-NO chemical reductant converter and the NO2 converted to NO+O2. The
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 3 of 32
NO+O2 is blended with O3. The chemiluminescent response is proportional to the concentration
of NOX entering the converter. The NO2 determinations are not the result of direct measurement.
The concentration of NO2 is calculated as the difference between a measured NOX value and a
measured NO value representing the same point in time. There are basically two types of
NO2-to-NO converters; the chemical reductant converter and the photolytic converter. The
chemical reductant converter uses a reducing agent such as molybdenum or gold/carbon
monoxide to convert NO2 to NO. The photolytic converter uses high energy light to perform the
conversion.
Instruments using chemical reductant conversion permit accurate measurement of NO,
NO2, and NOX as long as there are no other nitroxyl compounds present in the sampled
atmosphere. Peroxyacetyl nitrate (PAN) and nitric acid (HNO3) are primary interferents to the
accurate measurement of NO2 when a chemical reductant converter is used. Chemical reductant
converters may partially or completely convert PAN, HNO3, and/or other nitroxyl compounds to
NO. The conversion of compounds other than NO2 causes artificially high values for NO2. The
least biased measurements are typically early morning measurements in urban areas where NO
and/or NO2 concentrations are high and concentrations of PAN, HNO3, and/or other nitroxyl
compounds are low. The potential for biasing the NO2 measurement due to the chemical
reductant conversion of PAN, HNO3, and/or other nitroxyl compounds is greatest in urban areas
during late afternoon because the ambient concentration of NO is negligible (i.e., at or below the
detection limit of most conventional instruments) and PAN and HNO3 and/or nitroxyl
compounds other than NO and NO2 comprise a significant percentage of the total airborne
nitroxyl compounds. Therefore, the conversion of the interfering compounds has a greater
biasing effect yielding values that are more closely related to NOy (i.e., organic nitroxyl
compounds including NO and NO2) than to NOX. Although the potential for measurement bias
exists, measurements of NO and NO2 are required under 40 CFR Part 58, Subpart E using EPA
approved instrumentation.
The photolytic converter uses high energy light (i.e., wavelengths from 300 to
430 nanometers) to specifically convert NO2 to NO+O2. The NO+O2 are then mixed with O3 and
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 4 of 32
measured by a chemiluminescence detector. Although the photolytic converter specifically
converts NO2 to NO, it is not 100% efficient and must be calibrated regularly to determine and
quantify conversion efficiency. Because the photolytic converter converts only NO2 to NO, bias
caused by the partial or complete conversion of other nitroxyl compounds is not experienced.
The photolytic converter is not as well developed as the chemical reductant converter and is
therefore more expensive to purchase, operate, and maintain.
4.1.2 Method and Equipment Description
Instrumentation approved or designated as reference or equivalent methods for
measuring ambient concentrations of NO2 are listed in 40 CFR Part 53.2 Subject to any
limitations specified in the applicable designation, each instrument is acceptable for use in
enhanced O3 monitoring networks, unless the applicable designation is subsequently canceled.
Instruments designated as reference methods for NO2 are also approved for measuring NO. The
detailed procedure for measuring ambient concentrations of NO2 using approved instrumentation
is contained in 40 CFR Part 50, Subpart C, Appendix F.3
Figure 4-1 presents the flow schematic of a typical NO-NO2 instrument (i.e., the
Thermo Environmental Instruments, Inc., Model 42, Designated Reference Method Number
RFNA-1289-074). Sample enters the instrument through a flow control capillary. A solenoid
valve routes the sample either through the converter (i.e., NO2 measurement mode) or around the
converter (i.e., NO measurement mode). When the sample flow is routed through the converter,
the chemiluminescence measured in the reaction chamber represents the concentration of NO +
NO2 and any other oxides of nitrogen compounds that are converted. When the sample flow is
routed around the converter, only NO is measured. The concentration of NO2 is calculated
automatically by a microprocessor in the instrument as the difference between a measured NOX
value and a measured NO value for the sample point in time.
-------
Air Dryer
Ozone
Generator
Ozone
Cleanser Capillary
Sample1
In
Exhaust
Chamber
Pressure
Gauge
^rj Q ^d oo
£3 rvj n> CD
OQ S- < O
ft ft H-. ^-.
' ' ' ' 2. O
3 ' '
L/i O ^^ -^.
2> §
oj o
oo
o:/s/g/morris/3797/pams/fig4-1 .ppt
Figure 4-1. Flow Schematic of a Typical NO-NO2 Instrument
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 6 of 32
4.2 Total Reactive Oxides of Nitrogen
The nitroxyl compounds in ambient air included in the group of specific compounds
referred to as NOy have not been specifically defined. This group contains all of the nitroxyl
compounds that react in the troposphere to any significant extent and, therefore, contribute to the
photochemical formation of O3.
Identified NOy constituents include:
NO;
NO2;
• nitrogen trioxide, N2O3;
• nitrogen pentoxide, N2O5;
• nitrous acid, HNO2;
HNO3;
• peroxynitrate;
PAN;
• other organic nitrates; and
• other aerosol nitrates.
In typical urban environments the principal NOy compounds are NO, NO2, PAN, and HNO3.
Measurements of NOy are a valuable metric serving multiple purposes. Speciated measurements
of NOy compounds provide valuable information relevant to understanding photochemical cycles
and evaluating the behavior of chemical mechanisms applied in O3 prediction models. Because
NOy is a conservative determination of all nitrogen emissions releases, excluding losses due to
deposited nitrogen, NOy should be an excellent indicator of NO and NO2 emissions trends.
However, speciated NOy measurements (i.e., analyzing separately for each NOy compound) on a
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 7 of 32
routine basis are presently impractical because they require that the user know the identity of all
the compounds to be measured and that appropriate individual methods can be applied.
One of the more important uses of NOy data is predicting tropospheric O3 and assessing
the importance of NO and NO2 and VOC levels to O3 production and control. Observational
based models (OEMs) assess the age of air masses to evaluate control strategies. Generally, air
masses that contain predominantly "fresh" NO emissions are more likely to be
hydrocarbon-deficient and require VOCs to produce O3. Air masses that contain nitroxyl
compounds are aged and are NOX deficient, requiring NOX for maximum O3 production. In either
case, it is critical to know NO, NO2 and NOy levels to decide the best regional control strategy for
03.
Since the measurement of individual reactive nitroxyl compounds is technically
difficult, time consuming, and expensive, it is currently impractical to require routine monitors of
these species at PAMS stations. However, a practical instrument based total NOy measurement
procedure has been developed. This total NOy measurement principle and procedure is presented
in Section 4.3. Measuring total NOy is not required by 40 CFR, Part 58, Subpart E.1 However,
for the reasons discussed above, it is strongly encouraged by the EPA as part of enhanced O3
monitoring programs.
4.3 Measurement of Total Reactive Oxides of Nitrogen in the Atmosphere
(Gas Phase Chemiluminescence) - Measurement principle and
Procedures
The measurement principle and procedures addresses the need to measure total NOy in
a practical, standardized manner. The total NOy measurement principle and procedures is based
on reconfiguration and operation of a commercially available NOy instrument (i.e., the Thermo
Environmental Instruments, Inc., Model 42s)—the only NOy instrument commercially available
at the time of preparation of this guidance. Although the total NOy measurement principle and
procedures are based on this specific instrument, the approach is sufficiently generic to be used
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 8 of 32
with other similar equipment. Any instruments used to measure NOy must have the sensitivity to
measure the low concentrations typically encountered during late afternoon periods and also be
able to measure the high concentrations encountered during early morning periods. Automatic
range change capabilities will probably be required to accommodate the wide range of NOy
concentrations experienced in a typical urban atmosphere. The total NOy measurement principle
and procedures are structured according to Federal Register format and includes the same topics
and level of detail as the EPA NO2 instrumental method. The EPA recognizes that NOy
measurement is an emerging technology. As advances in the technology are made, the NOy
instrumental method will be updated.
Applicability of the measurement of total NOy is presented in Section 4.3.1 and the
operational principle of measurement is presented in Section 4.3.2. Information on modification
and final configuration of the NOy instrument is presented in Section 4.3.3. Calibration
procedures, including generation of calibration curves and determination of converter efficiency,
are found in Section 4.3.4.
4.3.1 Applicability
The total NOy measurement principle and procedures are applicable to the
measurement of ambient concentrations of NO and NOy as part of PAMS networks. The data
obtained using these procedures are intended for use in mathematical models that predict
tropospheric O3 trends.
4.3.2 Principle of Measurement
Ambient NO and NOy concentrations are determined by photometrically measuring the
light intensity at wavelengths greater than 600 nanometers from the chemiluminescent reaction of
NO with O3.4'5'6 This principle is identical to that used for the measurement of NO and NO2.
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 9 of 32
To measure NOy, sample air is passed through a probe-mounted chemical reductant
converter and the nitroxyl compounds present are reduced to NO.7'8'9'10'11 NO, which commonly
exists in ambient air, passes through the converter unchanged. The NO resulting from the
reduction of these nitroxyl compounds, plus any native NO, is reacted with O3 and the resulting
chemiluminescent light is measured as the total NOy concentration.
To measure NO separately and specifically, sample air is by-passed around the
chemical reductant converter so that no reduction of the other nitroxyl compounds10 to NO
occurs. The NO (i.e., native NO only) is reacted with O3 and the resulting chemiluminescent
light is measured as the NO concentration.
It should be noted that field and laboratory studies have been conducted to compare
various NOy measurement techniques.11'12 Application of the total NOy measurement principle
and procedures using a heated chemical reductant surface is supported by these studies.
However, qualifiers do apply.
4.3.3 Measurement Apparatus
The NOy instrument presented in these procedures is based on the design of an
approved NO2 instrument. Details on the differences between the configuration of the NOy
instrument and the EPA-approved NO2 instrument are presented below.
4.3.3.1 Configuration
The configuration of the NOy instrument is very similar to that of the EPA-approved
NO2 instrument. The primary differences between the configurations are the locations of the
converter, particulate filter(s), flow control capillary, and the 3-way solenoid valve used for mode
control. Differences between the NOy instrument configuration and the EPA-approved NO2
instrument configuration can be better understood by comparing the NO instrument flow
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 10 of 32
schematic presented in Figure 4-2 to the flow schematic of the approved NO2 instrument
presented in Figure 4-1.
The NOy instrument uses a heated molybdenum chemical reductant converter that also
serves as the sample probe inlet to the instrument. The converter is positioned at a program
specific height above ground level within the range of 3 to 15 meters as required for enhanced O3
monitoring programs. Because the converter serves as the sample probe inlet, conversion of
PAN, HNO3, and other nitroxyl compounds to NO is maximized because the surface area, and
consequently surface adsorption of these compounds that typically occurs prior to reaching the
converter, is minimized. It is critical to the performance and longevity of the reductant material
that it not be overheated. The converter operating temperature is actively controlled at 325 °C
(±20°C). Separate sample transfer lines are used for the NOy and NO channels, and another
separate transfer line is used to deliver calibration and converter efficiency assessment standards
to the sample inlet.
The EPA-approved NO2 instrument incorporates a particulate filter that is located prior
to the flow control capillary and the chemical reductant converter. In the NOy instrument
configuration, the particulate filter is located after the chemical reductant converter in the NOy
sample transfer line to further minimize surface area and consequently surface adsorption of
nitroxyl compounds in the sample air. Because the NO channel uses a separate sample transfer
line, a second separate particulate filter is installed.
The EPA-approved NO2 instrument uses a 3-way solenoid valve (i.e., mode control
valve), located prior to the chemical reductant converter, to route sample gas through or around
the chemical reductant converter depending on whether the instrument is in the NO or NO2
mode. In the NOy instrument configuration, the mode control valve performs the same function
but is located after the chemical reductant converter to reduce surface area and surface effects.
-------
Air
Dryer
Dry Air
Ozone
Generator
Capillary
(NO Only)
(NOY)
(Calibration)
f
0-
Weatherproof
Enclosure
(Positioned at
Probe Inlet Height)
Filter
Filter
Prereactor
Canillarv _
— 1
_l\
Chamber
Pressure
Gauge
Pump
Tf O 73
£ ft 3
ft> ft i-l •
' '
oo
a
C±.
O
o:/s/g/morris/3797/pams/fig4-2.ppt
Figure 4-2. Flow Schematic of the Reconfigured NO Instrument for PAMS Application
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 12 of 32
The EPA approved NO2 instrument incorporates a glass capillary to control the sample
air flow rate through the instrument. The capillary is located prior to the mode control valve and
chemical reductant converter. In the NOy instrument configuration, the flow control capillary is
located after the chemical reductant converter and mode control valve. Positioning the capillary
after the mode control valve allows the flow rate through the separate NO and NOy sample
transfer lines to be controlled equally using only one flow control device.
4.3.3.2 Reconfiguration
The reconfiguration of the commercially available NOX instrument into the NOy
instrument applicable to PAMS requires modification of elements of the plumbing (i.e., the flow
pattern) and electronics (i.e., the converter temperature control system) of the instrument.
Modifications to the sampling station must also be made to accommodate the PAMS NOy
instrument configuration. General information on these modifications is presented below.
Specific details on modifications must be obtained from individual manufacturers.
Note: Refer as needed to Figure 4-2 and the instrument manual supplied by the
manufacturer while performing the modifications. All modifications should be made in
strict accordance with applicable ordinances (e.g., OSHA Regulations, electrical codes,
zoning requirements, etc.).
4.3.3.2.1 Shelter
Attach an open-bottomed weatherproof enclosure (i.e., unpainted aluminum or stainless
steel), sized (i.e., approximately 12 in. high, 6 in. wide, and 7 in. deep) to accommodate the
converter assembly to a 4 ft. long sampling tower mast. Attach the mast with enclosure to the
sampling tower at the sample probe inlet height specified in the monitoring program
requirements. Attach one end of a length of 1-1/2 in. O.D. flexible conduit to the weatherproof
enclosure. Route the flexible conduit along the mast, down the tower, and into the shelter,
securing as required to provide adequate support.
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 13 of 32
4.3.3.2.2 Plumbing
The following modifications are performed outside the instrument chassis:
(a) Remove the converter assembly from the chassis of the instrument and mount it in
the weatherproof enclosure so that the inlet of the converter faces down toward the open bottom
of the enclosure. Connect a 3/8 in. "Cross" fitting to the inlet of the converter using ancillary
fittings as required to insure proper mating.
(b) Extend a length of uniquely identified 1/4 in. O.D. thick wall Teflon® tubing
through the flexible conduit from the shelter into the weatherproof enclosure. Connect this
uniquely identified Teflon® tube to the outlet of the converter to serve as the NOy sample transfer
line.
(c) Extend a length of uniquely identified 1/4 in. O.D. thick wall Teflon® tubing
through the flexible conduit from the shelter into the weatherproof enclosure. Connect this
uniquely identified Teflon® tube to one of the side ports of the 3/8 in. "Cross" fitting using
ancillary fittings as required to insure proper mating to serve as the NO sample transfer line.
(d) Extend a length of uniquely identified 3/8 in. O.D. thick wall Teflon® tubing
through the flexible conduit from the shelter into the weatherproof enclosure. Connect this
uniquely identified Teflon® tube to the remaining side port of the "Cross" fitting to serve as the
calibration and converter efficiency assessment standards transfer line. Connect the end of the
Teflon® tube located in the shelter to the outlet of the calibration system (refer to
Section 4.3.4.1).
Note: The bottom port of the "Cross" fitting is left open. This open port serves
as the sample air inlet during monitoring and as the excess calibration standard gas vent
during calibration.
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 14 of 32
The following modifications are performed inside of the instrument chassis:
(a) Remove and replumb the flow rate control capillary so that it is connected to the
common port (C) of the mode control solenoid valve.
(b) Remove and replumb the particulate filter assembly so that the outlet of the
particulate filter assembly is connected to the normally closed port (NC) of the mode control
solenoid valve. Attach the appropriate Teflon® tube leading from the sampling probe to the inlet
of the parti culate filter assembly.
(c) Plumb a second paniculate filter assembly so that the outlet of the paniculate
filter assembly is connected to the normally open port (NO) of the mode control solenoid valve.
Attach the appropriate Teflon® tube leading from the sampling probe to the inlet of the
particulate filter assembly.
4.3.3.2.3 Electronics
The following modifications are performed outside of the instrument chassis:
(a) Extend a length of thermocouple wire (i.e., type K - Chromel/Alumel, 16 gauge
solid core, twisted pair with stainless steel overbraiding) through the flexible conduit from the
shelter into the weatherproof enclosure. Connect the thermocouple wire to the thermocouple
located in the converter. Ensure that the chromel wires are mated and that the alumel wires are
mated.
(b) Extend a length of power wire (i.e., two conductors with ground, 14 gauge solid
core copper, rated for outdoor use) through the flexible conduit from the shelter into the
weatherproof enclosure. Connect the two conductors of the power wire to the heater located in
the converter. Connect the ground wire directly to the weatherproof enclosure using a screw and
nut.
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 15 of 32
The following modifications are performed inside of the instrument chassis:
(a) Connect the thermocouple wire (i.e., end located in the shelter) to the converter
temperature control board. Ensure that the chromel and alumel wires are attached to the correct
terminals (refer to the manufacturer's instrument manual).
(b) Connect the power wire (i.e., end located in the shelter) to the converter
temperature control board. Ensure that the two conductors of the power wire are attached to the
correct terminals (refer to the manufacturer's instrument manual). Connect the ground wire
directly to the chassis ground of the instrument.
4.3.4 Calibration
Calibration of NOy instruments is accomplished by gas phase titration (GPT) of an NO
standard with O3. Major equipment/components required are a stable O3 generator, a
chemiluminescence NOy instrument with strip chart recorder(s), and an NO concentration
standard. The principle of this calibration technique is based upon the rapid gas phase reaction
between NO and O3 to produce stoichiometric quantities of NO2 in accordance with the
following equation:13
(NO + 03) - (N02 + 02) (4-3)
When the NO concentration is known, the concentration of NO2 can be determined. Ozone is
added to excess NO in a dynamic calibration system, and the NO channel of the
chemiluminescence NOy instrument is used as an indicator of changes in NO concentration.
Upon the addition of O3, the decrease in NO concentration observed on the calibrated NO
channel is equivalent to the concentration of NO2 produced. The amount of NO2 generated may
be varied by adding variable amounts of O3 from a stable uncalibrated O3 generator.14
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 16 of 32
Note: Because the principle, apparatus, and procedures used to calibrate the NOy
instrument are consistent with calibration of NOX and NOy instruments, it should be
possible to calibrate both types of instruments simultaneously. However, the large quantity
of calibration gas required may render simultaneous calibration of NOX and NOy
instruments impractical.
4.3.4.1 Apparatus
Figure 4-3 presents a schematic of a typical GPT calibration system and shows the
suggested configuration of the components listed below. All connections between components
in the calibration system downstream from the O3 generator should be constructed of glass,
Teflon®, or other nonreactive material.
4.3.4.1.1 Air Flow Controllers
Devices capable of maintaining constant air flows within ±2% of the required flow rate.
4.3.4.1.2 NO Flow Controller
A device capable of maintaining constant NO flows within ±2% of the required flow
rate. Component parts in contact with the NO should be made of a nonreactive material.
4.3.4.1.3 Air Flowmeters
Calibrated flowmeters capable of measuring and monitoring air flow rates with an
accuracy of ±2% of the measured flow rate.
-------
Flow
Controller
Flowmeter
Fo
03
Generator
(7
A v*_
1 '
NO
Standard
,* .
Flow
Controller
Flowmeter
Valv
/
\
NO
Vent to Atmosphere
To
,. , . .,.,_, Instrument
To Instrument (NO) /^Q *
Sample
Inlet
From
Calibration
System
Output Manifold
>
4
Mixing
Chamber
Reaction
Chamber
Vent to Atmosphere
tii iL5 I
Extra Outlets Capped
When Not in Use *
To Inlet of NOX
Instrument
Being Calibrated
£ ft B
ft> ft i-2- C±.
' ' ' ' %. O
o ri
o:/s/g/morris/3797/pams/fig4-3.ppt
FD = flow of diluent gas
F0 = flow of ozonated gas
FMn = flow of NO
NO
FT = total flow
Figure 4-3. Flow Schematic of a Typical Gas Phase Titration Calibration System
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 18 of 32
4.3.4.1.4 NO Flowmeter
A calibrated flowmeter capable of measuring and monitoring NO flow rates with an
accuracy of ±2% of the measured flow rate. Rotameters have been reported to operate unreliably
when measuring low NO flows and are not recommended.
4.3.4.1.5 Pressure Regulator For Standard NO Cylinder
The regulator must have a stainless steel diaphragm and internal parts and a suitable
delivery pressure.
4.3.4.1.6 Ozone Generator
The generator must be capable of generating sufficient and stable levels of O3 for
reaction with NO to generate NO2 concentrations in the range required. Ozone generators of the
electric discharge type may produce NO and NO2 and are not recommended. Photolytic O3
generators are recommended.
4.3.4.1.7 Valve
A valve used to divert the NO flow when zero air is required at the manifold. The
valve should be constructed of glass, Teflon®, or other nonreactive material.
4.3.4.1.8 Reaction Chamber
A chamber, constructed of glass, Teflon®, or other nonreactive material, for the
quantitative reaction of O3 with excess NO. The chamber should be of sufficient volume such
that the residence time meets the requirements specified in Section 4.3.4.3. For practical reasons,
residence time should be less than 2 minutes.
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 19 of 32
4.3.4.1.9 Mixing Chamber
A chamber constructed of glass, Teflon®, or other nonreactive material and designed to
provide thorough mixing of reaction products and diluent air. The residence time is not critical
when the dynamic parameter specification given in Section 4.3.4.3 is met.
4.3.4.1.10 Output Manifold
The output manifold should be constructed of glass, Teflon®, or other non-reactive
material and should be of sufficient diameter to ensure an insignificant pressure drop at the
analyzer connection. The system must have a vent designed to ensure atmospheric pressure at
the manifold and to prevent ambient air from entering the manifold.
4.3.4.1.11 Valve
A valve used to isolate the calibration line (i.e., Teflon® line used to deliver standard
gases to the NOy instrument sample inlet during calibration) when the instrument is monitoring.
The valve must be sized to ensure atmospheric pressure at the manifold and should be
constructed of glass, Teflon®, or other nonreactive material.
4.3.4.2 Reagents
4.3.4.2.1 NO Concentration Standard
Gas cylinder standard containing 50 to 100 ppm NO in N2 with less than 1 ppm NO2.
This standard must be traceable to a National Institute of Standards and Testing (NIST) NO in N2
Standard Reference Material (SRM 1683 or SRM 1684), an NBS NO2 Standard Reference
Material (SRM 1629), or an NBS/ EPA-approved commercially available Certified Reference
Material (CRM).15 A recommended protocol16 for certifying NO gas cylinders against either a
NO SRM or CRM is given in Section 2.0.7 of the reference. Procedures17 for certifying a NO gas
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 20 of 32
cylinder against a NIST NO2 SRM and for determining the amount of NO2 impurity in a NO
cylinder are presented in the reference.
4.3.4.2.2 Zero Air
Zero air is free of contaminants that are detectable on the NO/NOy analyzer or that react
with either NO, O3, or NO2 in the gas phase titration. A procedure for generating zero air is
given in the reference.14
4.3.4.3 Dynamic Parameter Specification
1. The O3 generator air flow rate (F0) and NO flow rate (FNO) (see Figure 4-3) must
be adjusted such that the following relationship holds:
PR = [NO]RC x tR = 2.75 ppm-minutes (4_4)
[NO]RC = [NO]
NO
F + F
ro rNO,
(4-5)
VRC
tR = < 2 minutes (4.5)
TT + TT ^ '
ro rNO
where:
PR = dynamic parameter specification, determined empirically, to ensure
complete reaction of the available O3, ppm-minute
[NO]RC = NO concentration in the reaction chamber, ppm
tR = residence time of the reactant gases in the reaction chamber, minute
[NO]STD = concentration of the undiluted NO standard, ppm
FNO = NO flow rate, scm3/min
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 21 of 32
generator air flow rate, scm /min
VRC = volume of the reaction chamber, scm3
2. The flow conditions to be used in the GPT system are determined by the
following procedure:
(a) Determine the total flow required at the output manifold (FT=analyzer
demand plus 50% excess). If a conventional NOX instrument is being
calibrated simultaneously with the NOy instrument, FT must reflect the
demand of both.
(b) Establish [NO]OUT as the highest NO concentration (ppm) which will be
required at the output manifold. [NO]OUT should be approximately
equivalent to 90% of the upper range limit (URL) of the NO2 concentration
range to be covered.
(c) Determine FNO as follows:
[NO]OUT x Ft
N0 = ~
(d) Select a convenient or available reaction chamber volume. Initially, a trial
VRC may be selected to be in the range of approximately 200 to 500 scm3.
(e) Compute F0 as follows:
(4-8)
\
[NO]STD x FNO x VRC
2.75
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 22 of 32
(f) Compute tR as follows:
O NO
Verify that tR < 2 minutes. If not, select a reaction chamber with a smaller VRC.
(g) Compute the diluent air flow rate as follows:
FD = FT - Fo - FNO (4-10)
where:
FNO = NO flow rate, scm3/min
[NOJoux = the highest NO concentration (ppm) that will be required at the output
manifold
FT = analyzer demand plus 50% excess
[NO]STD = concentration of the undiluted NO standard, ppm
FO = Os generator air flow rate, scm3/min
VRC = volume of the reaction chamber, scm3
tR = residence time of the reactant gases in the reaction chamber, minute
FD = diluent air flow rate, scm3/min
(h) If F0 turns out to be impractical for the desired system, select a reaction
chamber having a different VRC and recompute F0 and FD.
Note: A dynamic parameter lower than 2.75 ppm-minutes may be used if it
can be determined empirically that quantitative reaction of O3 with NO occurs. A
procedure for making this determination as well as a more detailed discussion of
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 23 of 32
the above requirements and other related considerations is included in the
reference.14
4.3.4.4 Calibration Procedure
The calibration procedure is presented below.
1. Assemble a dynamic calibration system configured like the one shown in
Figure 4-3.
2. Ensure that all flowmeters are calibrated under the conditions of use against a
primary standard (i.e., bubble flow meter or wet-test meter). All volumetric flow
rates should be corrected to 25 °C and 760 mm Hg.
3. Precautions must be taken to remove O2 and other contaminants from the NO
pressure regulator and delivery system prior to the start of calibration to avoid any
conversion of the standard NO to NO2. Failure to do so can cause significant
errors in calibration. This problem may be minimized by (1) carefully evacuating
the regulator after the regulator has been connected to the cylinder and before
opening the cylinder valve; (2) thoroughly flushing the regulator and delivery
system with NO after opening the cylinder valve; and (3) not removing the
regulator from the cylinder between calibrations. If the regulator is removed,
steps 1 and 2 should be repeated.
4. Select the operating range of the NO/NOy instrument to be calibrated. In order to
obtain maximum precision and accuracy, the NO and NOy channels of the
instrument should be set to the same range.
5. Connect the recorder output cable(s) of the NO/NOy instrument to the input
terminals of the strip chart recorder(s). All adjustments to the instrument should
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 24 of 32
be performed based on the appropriate strip chart readings. References to
instrument responses in the procedures given below refer to recorder
responses.
6. Determine the GPT flow conditions required to meet the dynamic parameter
specification as indicated in Section 4.3.4.3.
7. Adjust the diluent air and O3 generator air flows to obtain the flows determined in
Section 4.3.4.3, step 2. The total air flow must exceed the total demand of the
instrument(s) connected to the output manifold to ensure that no back diffusion of
ambient air occurs. Allow the instrument to sample zero air until stable NO and
NOy responses are obtained. After the responses have stabilized, adjust the
instrument zero control(s). Record the stable zero air responses as ZNO and
ZNOy.
Note: Some instruments may have separate zero controls for NO and
NOy, while still others may have only one zero control common to both
channels. Offsetting the instrument zero adjustments to +5% of scale is
recommended to facilitate observing negative zero drift.
8. Prepare the NO and NOy calibration curves as follows:
(a) Adjustment of NO span control. Adjust the NO flow from the standard NO
cylinder to generate a NO concentration of approximately 80% of the URL
of the NO range. The exact NO concentration is calculated from:
FXTn x [NOLTn
[NO]OUT = N° I JSID (4-11)
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 25 of 32
where:
= the highest NO concentration (ppm) that will be required at the output
manifold
FNO = NO flow rate, scm3/min
[NO]STD = concentration of the undiluted NO standard, ppm
FO = Os generator air flow rate, scm3/min
Fn = diluent air flow rate, scm3/min
D
Sample this NO concentration until the NO and NOy responses have stabilized. Adjust
the NO span control to obtain a recorder response (percent scale) as indicated below:
Recorder Response = — x 100 + ZNO (4-12)
URL
where:
[NO]OUT = diluted NO concentration at the output manifold, ppm
URL = nominal upper range limit of the NO channel, ppm
ZNO = stable zero air response for NO, percent scale
Note: Some instruments may have separate span controls for NO and
NOy, while still others may have only one span control common to both
channels. When only one span control is available, the span adjustment is
made on the NO channel of the instrument.
If substantial adjustment of the NO span control is required, it may be necessary
to recheck the zero and span adjustments by repeating Section 4.3.4.4, step 7 and
step 8(a). Record the NO concentration and the instrument's NO response.
-------
where:
Section: 4
Revision: 1
Date: 09/30/98
Page: 26 of 32
(b) Adjustment of NOy span control. When adjusting the instrument's NOy span
control, the presence of any NO2 impurity in the standard NO cylinder must
be taken into account. The exact NOy concentration is calculated from:
[NOJ
([NO]
STD
[NOJTMP)
yJOUT
-NO
F + F
rO rD
(4-13)
[NOy]
yJOUT
[NO]STO
[NQ2]iMp
= diluted NOy concentration at the output manifold, ppm
= NO flow rate, scm3/min
= concentration of the undiluted NO standard, ppm
= concentration of NO2 impurity in the standard NO cylinder, ppm
= O3 generator air flow rate, scm3/min
= diluent air flow rate, scm3/min
Adjust the NOy span control to obtain a recorder response (percent scale) as indicated below:
where:
URL
ZNO,
(4-14)
= diluted NOy concentration at the output manifold, ppm
URL = nominal upper range limit of the NO channel, ppm
ZNOy = stable zero air response for NOy, percent scale
Note: If the instrument has only one span control, the span adjustment
is made on the NO channel and no further adjustment is made for NOy. If
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 27 of 32
substantial adjustment of the NOy span control is required, it may be
necessary to recheck the zero and span adjustments by repeating
Section 4.3.4.4, step 7 and step 8(a). Record the NOy concentration and the
instrument's NOy response.
(c) Generate several additional concentrations (at least five evenly spaced points
across the remaining scale are suggested to verify linearity) by decreasing
FNO or increasing FD. For each concentration generated, calculate the exact
NO and NOy concentrations using Equations 4-11 and 4-13, respectively.
Record the instrument's NO and NOy responses for each concentration. Plot
the instrument responses versus the respective calculated NO and NOy
concentrations and plot the NO and NOy calibration curves and calculate the
linear regression. For subsequent calibrations where linearity can be
assumed, these curves may be checked with a two-point calibration
consisting of a zero air point and NO and NOy concentrations of
approximately 80% of the URL.
9. Preparation of NO2 Converter Efficiency Assessment Standards.
(a) Assuming the NOy zero has been properly adjusted while sampling zero air
as described in Section 4.3.4.4, step 7, adjust F0 and FD as determined in
Section 4.3.4.3, step 2. Adjust FNO to generate a NO concentration near 90%
of the URL of the NO range. Sample this NO concentration until the NO
and NOy responses have stabilized. Using the NO calibration curve obtained
in Section 4.3.4.4, step 8, measure and record the NO concentration as
[NO]orig. Using the NOy calibration curve obtained in Section 4.3.4.4, step 8,
measure and record the NOy concentration as [NOy]orig.
(b) Adjust the O3 generator to generate sufficient O3 to produce a decrease in the
NO concentration equivalent to approximately 80% of the URL of the NO2
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 28 of 32
range. The decrease must not exceed 90% of the NO concentration
determined in Section 4.3.4.4, step 9(a). After the instrument responses
have stabilized, record the resultant NO and NOy concentrations as [NO],,
and [N0y]rem.
(c) Calculate the resulting NO2 concentration from:
[N02]OUT = ([N0]ong - [N0]rem)
FNO ([N02]IMP)
F + F + F
rNO rO rD
(4-15)
where:
[NO2]OUT = diluted NO2 concentration at the output manifold, ppm
[NO]orig = original NO concentration, prior to addition of O3, ppm
[NO]rem = NO concentration remaining after addition of O3, ppm
FNO = NO flow rate, scm3/min
mip = concentration of NO2 impurity in the standard NO cylinder, ppm
F0 = O3 generator air flow rate, scm3/min
FD = diluent air flow rate, scm3/min
(d) Maintaining the same FNO, F0, and FD as in Section 4.3.4.4, step 9(a), adjust
the ozone generator to obtain several other concentrations of NO2 over the
NOy range (at least five evenly spaced points across the remaining scale are
suggested). Calculate each NO2 concentration using Equation 4-15 and
record the corresponding instrument NO responses.
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 29 of 32
4.3.4.5 Determination of Converter Efficiency
For each NO2 concentration generated (see Section 4.3.4.4, step 9), calculate the
concentration of NO2 converted from:
[N02]CONV = [N02]OUT - ([N0y]ong- [NOy]reJ (4-16)
where:
[NO2]CONv = concentrati°n of NO2 converted, ppm
[NO2]0uT = Diluted NO2 concentration at the output manifold, ppm
[NOy]orig = original NOy concentration prior to addition of O3, ppm
[NOy]rem = NOX concentration remaining after addition of O3, ppm
Plot [NO2]CONV (y-axis) versus [NO2]OUT (x-axis) and plot the converter efficiency curve
and calculate the linear regression. The slope of the curve times 100 is the average converter
efficiency (EC). The EC must be greater than 96%; if it is less than 96%, replace or service the
converter.
4.3.4.6 Frequency of Calibration
The frequency of calibration, as well as the number of points necessary to establish the
calibration curve and the frequency of other performance checks, will vary from one instrument
to another. The user's quality control program should provide guidelines for initial establishment
of these variables and for subsequent verification and alteration as operational experience is
accumulated. Manufacturers of instruments should include in their instruction/operation manuals
information and guidance as to these variables and on other matters of operation, calibration, and
quality control.
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 30 of 32
4.3.4.7 Analyzer Challenge
The instrument must be challenged annually, at the start of the sampling season.
Additional challenges, at the beginning, middle, and end of the sampling season, are suggested.
The optimum challenge would consist of a mixture of representative NOy compounds, but there
is presently no comprehensive commercial mixture covering a broad range of the compounds of
potential interest available to be used as a challenge gas. To assess general instrument
performance, the instrument should be challenged with NO, since this is the gas used for
calibration. Additionally, the system should be assessed for its ability to measure other NOy
constituents by challenging the instrument with other candidate analytes such as 50 ppm w-propyl
nitrite or nitric acid.
4.4 Nitric Acid Measurement
The NOy measurement instrument described above can be obtained from the vendor in
a modified configuration that allows for the separate measurement of total NOy and nitric acid.
This configuration is called the Model 42CW. The Model 42CW cycles between two separate
converters. One of the converters has a nylon filter located at the inlet. Ambient nitric acid is
absorbed by the nylon filter. The nitric acid concentration is determined as the difference
between unfiltered total NOy value and the value determined for the filtered channel.
4.5 References
1. U.S. Environmental Protection Agency. Code of Federal Regulations. Title 40, Part 58,
Subpart E. Enhanced O3 Monitoring Regulations. Office of the Federal Register,
February 12, 1993.
2. U.S. Environmental Protection Agency. Code of Federal Regulations. Title 40, Part 53,
Office of the Federal Register, July 1, 1987.
3. U.S. Environmental Protection Agency. Code of Federal Regulations. Title 40, Part 50,
Subpart C, Appendix F. Measurement Principle and Calibration Procedure for the
Measurement of NO2 in the Atmosphere (Gas Phase Chemiluminescence). Office of the
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 31 of 32
Federal Register, December 1, 1976.
4. Fontijn, A., A. J. Sabadell, and R. J. Ronco. Homogeneous Chemiluminescent
Measurement of Nitric Oxide with Ozone. Anal. Chem. 42: 575 (1970).
5. Stedman, D.H., E.E. Daby, F. Stuhl, and H. Niki. Analysis of Ozone and Nitric Oxide by
a Chemiluminiscent Method in Laboratory and Atmospheric Studies of Photochemical
Smog. J. Air Poll. Control Assoc. 22: 260 (1972).
6. Martin, B.E., J.A. Hodgeson, and R.K. Stevens. "Detection of Nitric Oxide
Chemiluminescence at Atmospheric Pressure." Paper presented at 164th National ACS
Meeting., New York City, August 1972.
7. Hodgeson, J.A., K.A. Rehme, B.E. Martin, and R.K. Stevens. "Measurements for
Atmospheric Oxides of Nitrogen and Ammonia by Chemiluminescence." Paper
presented at 1972 APCA Meeting, Miami, FL, June 1972.
8. Stevens, R.K. and J.A. Hodgeson. Applications of Chemiluminescence Reactions to the
Measurement of Air Pollutants. Anal. Chem. 45: 443A(1973).
9. Breitenbach, L.P and M. Shelef. Development of a Method for the Analysis of NO2 and
NH3 by NO-Measuring Instruments. J. Air Poll. Control Assoc. 23: 128 (1973).
10. Winer, A.M., J.W. Peters, J.P. Smith, and J.N. Pitts, Jr. Response of Commercial
Chemiluminescent NO-NO2 Analyzers to Other Nitrogen-Containing Compounds.
Environ. Sci. Technol. 8: 1118 (1974).
11. Williams, E.J., K. Baumann, J.M. Roberts, S.B. Bertman, R.B. Norton, F.C. Fehsenfeld,
S.R. Springston, L.J. Nunnermaker, L. Newman, K. Olszyna, J. Meagher, B. Hartsell,
E. Edgerton, J. Pearson, and M.O. Rodgers. Intercomparison of NOy Measurement
Techniques. Submitted for Publication, J. Geophys. Research, Special Issue, Southern
Oxidant Study (1997).
12. Dahv, A.V., B.C. Daube, J.D. Burley, and S.C. Wofsy. Laboratory Investigation of the
Catalytic Reduction Technique for Measurement of Atmospheric NOy. J. Geophys.
Research, 102: 10759-10776 (1997).
13. Rehme, K.A., B.E. Martin, and J.A. Hodgeson. Tentative Method for the Calibration of
Nitric Oxide, Nitrogen Dioxide, and Ozone Analyzers by Gas Phase Titration.
EPA-R2-73-246, Research Triangle Park, NC: U.S. Environmental Protection Agency,
1974.
14. Hodgeson, J.A., R.K. Stevens, and B.E. Martin. A Stable Ozone Source Applicable as a
Secondary Standard for Calibration of Atmospheric Monitors. ISA Transactions 11: 161
(1972).
-------
Section: 4
Revision: 1
Date: 09/30/98
Page: 32 of 32
15. A Procedure for Establishing Traceability of Gas Mixtures to Certain National Bureau of
Standards Standard Reference Materials. EPA-600/7-81-010. Joint publication by NBS
and EPA. 1981.
16. Quality Assurance Handbook for Air Pollution Measurement Systems Volume II -
Ambient Air Specific Methods (Interim Edition). EPA-600/R-94/03 8a. U. S.
Environmental Protection Agency. 1994.
17. Ellis, E.G. Technical Assistance Document for the Chemiluminescence Measurement of
Nitrogen Dioxide. EPA-E600/4-75-003 Research Triangle Park, NC: U.S. Environmental
Protection Agency. 1992.
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 1 of 31
Section 5.0
Methodology for Determining Carbonyl Compounds
in Ambient Air
Determination of ambient concentrations of carbonyl compounds is a requirement of
40 CFR Part 58,l Subpart E, enhanced O3 network monitoring programs. Carbonyl compounds
have been shown to contribute to the formation of photochemical O3. Formaldehyde,
acetaldehyde, and acetone are specifically required target compounds for PAMS; however, other
carbonyl compounds may be added to the target list consistent with individual program
objectives. The methodology used to accomplish carbonyl compounds monitoring is
Compendium Method TO-11 A.2 Method TO-11 A, presented in Appendix D, provides sensitive
and accurate measurements of carbonyl compounds and involves sample collection and analysis
procedures. In this method, a cartridge(s) containing a solid sorbent is used to capture the target
compounds. Information on solid sorbents used is presented in Section 4.4 of Method TO-11 A.
Ozone has been identified as an interferent in the measurement of carbonyl compounds when
using Method TO-11A. To eliminate this interference, removal or scrubbing of O3 from the
sample air stream is mandatory. Section 5.1 presents information on O3 scrubbers. Sample
analysis is accomplished using high performance liquid chromatography (HPLC) with
ultraviolet/visible detection.
Under 40 CFR Part 58,1 Subpart E, States are required to obtain 3-hour and 24-hour
integrated measurements of carbonyl compounds at specified collection frequencies based on
individual enhanced O3 monitoring site type requirements. The sample collection frequencies
range from one 24-hour sample every sixth day to eight 3-hour samples every day. Specific
sample collection frequencies and minimum network monitoring requirements for carbonyl
compounds are presented in Table 5-1. (Note: This section is intended to be independent of other
sections. Figures, tables, and text from other sections are repeated as required.) The sample
collection frequencies necessitate the use of an automated multiple-event sample collection
approach. Section 5.2 presents information on multiple-event sample collection
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 2 of 31
Table 5-1 . PAMS Minimum Monitoring Network Requirements
Population
ofMSA/CMSA1
Less than 500,000
500,000 to 1,000,000
1,000,000 to
2,000,000
More than 2,000,000
Required
Site Type
(1)
(2)
(1)
(2)
(3)
(1)
(2)
(2)
(3)
(1)
(2)
(2)
(3)
(4)
Minimum VOCs
Sampling
Frequency2
AorC
AorC
AorC
B
AorC
AorC
B
B
AorC
AorC
B
B
AorC
AorC
Minimum Carbonyl
Compounds Sampling
Frequency2
_
DorF
E
-
E
E
-
E
E
-
-
'Whichever area is larger.
frequency Requirements Are As Follows:
A = Eight 3-hour samples every third day and one additional 24-hour sample every sixth day during
the monitoring period.
B = Eight 3-hour samples every day during the monitoring period and one additional 24-hour sample
every sixth day year-round.
C = Eight 3-hour samples on the 5 peak O3 days plus each previous day, eight 3-hour samples every
sixth day and one additional 24-hour sample every sixth day during the monitoring period.
D = Eight 3-hour samples every third day during the monitoring period.
E = Eight 3-hour samples on the 5 peak O3 days plus each previous day and eight 3-hour samples
every sixth day during the monitoring period.
F = Eight 3-hour samples on the 5 peak O3 days plus each previous day, eight 3-hour samples every
sixthe day and one additional 24-hour sample every sixth day during the monitoring period.
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 3 of 31
Table 5-1 . PAMS Minimum Monitoring Network Requirements (Continued)
The minimum sampling frequency requirements for speciated VOC monitoring are prescribed in 40 CFR
Part 58, Subpart E, Appendix D - Network Design for State and Local Air Monitoring Stations (SLAMSX
National Air Monitoring Stations (NAMS). and Photochemical Assessment Monitoring Stations (PAMSV
Section 4.3 - Monitoring Period requires, at a minimum, that O3 precursor monitoring be conducted
annually throughout the months of June, July, and August when peak O3 values are expected. Section 4.4 •
Minimum Monitoring Network Requirements specifies the minimum required number and type of
monitoring sites and sampling frequency requirements based on the population of the affected
MSA/CMSA or nonattainment area, whichever is larger. The minimum speciated VOC sampling
frequency requirements are summarized by site type below:
• Site Type 1 - Eight 3-hour samples every third day and one additional 24-hour sample every sixth day
during the monitoring period; or eight 3-hour samples on the 5 peak O3 days plus each previous day
and eight 3-hour samples and one 24-hour sample every sixth day, during the monitoring period.
• Site Type 2 - (population less than 500,000) - Same as Site Type 1.
• Site Type 2 - (population greater than 500,000) - Eight 3-hour samples every day during the
monitoring period and one additional 24-hour sample every sixth day year around.
• Site Type 3 - (population greater than 500,000) - Same as Site Type 1.
• Site Type 4 - (population more than 2,000,000) - Same as Site Type 1.
Samples collected should represent a time-integrated average for the required sampling period. It is
important to understand that the 3-hour sample integration period is a maximum requirement in the sense
that samples can be collected more frequently at shorter sampling intervals (i.e., three 1-hour periods) but
not less frequently for longer sampling intervals.
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 4 of 31
systems, including a generic equipment description and operating procedure and recommended
specifications applicable to evaluation and procurement.
5.1 Ozone Scrubbers
The EPA has determined through laboratory tests that O3 present in ambient air
interferes with the measurement of carbonyl compounds when using Method TO-11 A. Ozone
can interfere with carbonyl analyses in three ways:
The ozone reacts with the 2,4-dinitrophenylhydrazine (DNPH) on the cartridge,
making the DNPH unavailable for derivatizing carbonyl compounds;
The ozone also degrades the carbonyl derivatives formed on the cartridge during
sampling; and
If the analytical separation is insufficient, the DNPH degradation products can
coelute with target carbonyl derivatives.
The extent of interference depends on the temporal variations of both the ozone and the carbonyl
compounds and the duration of sampling. Carbonyl compound losses have been estimated to be
as great as 48% on days when the ambient O3 concentration reaches 120 ppbv. Eliminating this
measurement interference problem by removing or scrubbing O3 from the sample air stream prior
to collection of the carbonyl compounds is a mandatory facet of carbonyl compounds sample
collection for enhanced O3 monitoring programs. Two types of O3 scrubbers, the Denuder O3
scrubber and the Cartridge O3 scrubber, have been developed. Both the Denuder and Cartridge
O3 scrubbers use potassium iodide (KI) as the scrubbing agent. Scrubbing is based on the
reaction of O3 with KI, specifically:
O3 + 21 + H2O - I2 + O2 + 2OH (5-1)
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 5 of 31
where:
O3 = ozone (ambient)
H2O = water (ambient)
the iodide ion from potassium iodide forming molecular iodine (I2),
oxygen (O2), and the hydroxide ion (OH~)
Both O3 scrubber designs effectively remove O3 at sample collection flow rates up to 1 L/minute
and have sufficient scrubbing capacity to meet the needs of carbonyl compounds measurement
for enhanced O3 monitoring programs.
This section presents details of the two types of O3 scrubber equipment and
recommended procedures for their use.
5. 1. 1 Denuder Ozone Scrubber
The Denuder O3 Scrubber is a copper tube coated internally with a saturated solution
of KI. The tube is coiled and housed in a temperature controlled chamber that is heated to, and
maintained at, 66° C during sample collection. Heating prevents condensation from occurring in
the tube during sampling. The scrubber is connected to the inlet of the sample collection system.
Sample air is extracted from a sample probe and distribution manifold (see Section 5.2.3) and
pulled through the scrubber by an oilless vacuum pump. Ozone in the sample air is converted
(i.e., scrubbed) by the chemical reaction previously described in Section 5.1.
The Denuder O3 Scrubber is reusable. The copper tube should be recoated with a
saturated solution of KI after each six months of use. The Denuder O3 Scrubber prepared as
described in TO-1 1 A has been found to effectively remove ozone from the air stream for up to
100,000 ppb-hours. Thus, the scrubber will last for six months of 24-hour sampling on every
sixth day when sampling air with an average ozone concentration of 120 ppbv. To recoat the
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 6 of 31
denuder, fill the copper tube with a saturated solution of KI in water. Allow the solution to
remain in contact with the tube for a few minutes. Then, drain the tube. Dry the tube by blowing
a stream of clean air or nitrogen through the tube for about one hour.
An alternative to using a KI coated copper tube is to use a modified Dasibi ozone
scrubber device. Replace the manganese dioxide coated screens with 15 KI coated copper or
stainless steel screens assembled in a cartridge holder. Wash the screens in pure water in a sonic
bath. Dry the screens. Then, coat the screens by dipping them into a saturated KI solution in
water. Air dry the KI coated screens. This procedure deposits about 4 mmoles or about 700 mg
of KI over a sandwich of 15 two-inch diameter screens. Assemble the coated screens in the
Dasibi encasement with a fiberglass filter at each end. Close and seal the encasement including
the O-rings with the screws. Based on this removal capacity, this scrubber will last
approximately 300 days when sampling air with an average ozone concentration of 120 ppbv at a
rate of 1 L/min.
5.1.1.1 Denuder Ozone Scrubber Equipment
Figure 5-1 presents a cross-sectional view of the Denuder O3 Scrubber. The scrubber is
comprised of the following components:
Copper tubing - A 3 foot length of 1/4-inch O.D. copper tubing, coiled into a spiral
approximately 2 inches in diameter. Used as the body of the O3 scrubber.
Potassium iodide - The inside surface of the copper coil is coated with a saturated
solution of ACS Reagent Grade KI. Used to provide the O3 scrubbing mechanism.
Cord heater - A 2 foot long cord heater, rated at approximately 80 watts, wrapped
around the outside of the copper coil. Used to provide heat to prevent condensation of
water or organic compounds from occurring within the coil.
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 7 of 31
Sectional View
3' Potassium Iodide
Coated 1/4" O.D.
Copper Tubing
Aldehyde Probe
1/4" Brass
Buildhead
Union
110V Heater Plug
2'Glas-Col® I
Heater
Cord
Reducer 1/4"
Brass Buildhead
~7/8"
~ 1 1/4"
4"
Side View (cutaway)
Front View
o:/s/g/mom>/37«7/pams/hobson/Mib1 .ppt
Figure 5-1. Cross-Sectional View of the Denuder O3 Scrubber
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 8 of 31
Thermocouple - A Chromel-Alumel (Type K) thermocouple located between the
surface of the copper coil and the cord heater. Used to provide accurate temperature
measurement for temperature control.
Temperature controller - A Type K active temperature controller. Used to maintain
the O3 scrubber at 66°C as referenced by the Type K thermocouple.
Fittings - Bulkhead unions attached to the entrance and exit of the copper coil. Used to
allow connection to other components of the sampling system.
Chassis box - Conveniently sized aluminum enclosure. Used to contain the fittings,
coated copper tube, heater, and thermocouple.
5.1.1.2 Denuder Ozone Scrubber Operational Procedure
Recommended procedural steps for operation of the Denuder O3 Scrubber are as
follows:
(1) Connect the inlet of the Denuder O3 scrubber to the sample probe and
distribution manifold (see Figure 5-1).
(2) Connect the outlet of the Denuder O3 scrubber to the sample collection system
inlet.
(3) Set the temperature controller to maintain the scrubber at 66°C.
(4) Conduct sampling in accordance with the recommended procedures for operating
multiple-event sample collection systems as described in Section 5.2.2 and/or
Method TO-11A sampling procedures as described in Section 5.11 (see
Appendix D).
5.1.2 Cartridge Ozone Scrubber
The Cartridge O3 Scrubber is a standard Sep-Pak® Plus cartridge (i.e., identical in size
and shape to the precoated DNPH Silica Sep-Pak® cartridge) filled with approximately 1 gram of
ACS Reagent Grade KI. The scrubber is positioned at the inlet of the sample collection system.
Sample air is extracted from the sample probe and distribution manifold (see Figure 5-1) and
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 9 of 31
pulled through the O3 scrubber by an oilless vacuum pump. Ozone in the sample air is converted
(i.e., scrubbed) by the chemical reaction previously described in Section 5.1.
The Cartridge O3 Scrubber is commercially available (i.e., Waters Corporation) and is
disposable. The theoretical removal capacity of the scrubber, based on 100% consumption of KI,
is 200 mg of O3. Based on experience in the field, the cartridge O3 scrubber should be replaced
every three weeks.
5.1.2.1 Cartridge Ozone Scrubber Equipment
Figure 5-2 presents a cross-sectional view of the Cartridge O3 Scrubber. The scrubber
is comprised of the following components:
Cartridge housing - A two-part plastic vessel with an O.D. of approximately 1A inches
and an overall length of approximately 1-5/8 inches. One of the parts has a female Luer
style connector that serves as the scrubber inlet. The other part has a male Luer style
connector that serves as the scrubber outlet. Used to contain the scrubber media.
Potassium iodide - The scrubber medium is granular ACS Reagent Grade KI. Used to
provide the ozone scrubbing mechanism.
Inlet and outlet filters - Polyethylene fritted filters located inside the cartridge housing
at the inlet and outlet ends. Used to retain the scrubber media inside the cartridge
housing during sampling.
Compression ring - An aluminum ring sized to fit around the outside of the two
cartridge housing parts and seal them through compression. Used to provide a secure
leak-free seal between the two cartridge housing parts.
5.1.2.2 Cartridge Ozone Scrubber Operational Procedure
Recommended procedural steps for operation of the Cartridge O3 Scrubber are as
follows:
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 10 of 31
Connector (Luer Style)
Filter (Polyethylene)
Scrubber Media (Potassium Iodide)
Filter (Polyethylene)
Compression Ring (Aluminum)
Cartridge Housing
Connector (Luer Style)
o:/s/g/morris/3797/pams/hobson/hob2.ppt
Figure 5-2. Cross-Section View of the Cartridge O3 Scrubber
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 11 of 31
1. Connect the inlet of the Cartridge O3 scrubber to the sample probe and
distribution manifold (see Section 2.4.1.1).
2. Connect the outlet of the Cartridge O3 scrubber to the sample collection system
inlet.
3. Ensure that a leak-free connection is obtained.
4. Conduct sampling in accordance with the recommended procedures for operating
multiple-event sample collection systems as described in Section 5.2.2 and/or
Method TO-11A sampling procedures as described in Section 5.10 of Method
TO-11A (See Appendix D). Note: Heating of the cartridge ozone scrubbers
to 35 °C may be advisable under certain circumstances to prevent
condensation of water.
5.2 Multiple-event Sample Collection Systems
The use of solid sorbent cartridge sample collection systems to satisfy the sample
collection frequencies specified in Table 5-1 necessitates the use of multiple-event sample
collection systems. Multiple-event collection systems should be capable of unattended operation
in order to allow for multiple sample collection in a practical, non-labor intensive manner.
Multiple-event sampling systems are manufactured commercially or can be custom manufactured
by the user for a specific application. Several multiple-event sampling systems are commercially
available.
The following sections generally describe multiple-event sampling equipment,
procedures, and specifications. Also, recommended system specifications applicable to the
evaluation and procurement of multiple-event sampling systems are presented.
5.2.1 Multiple-event Collection System Equipment
A typical multiple-event sampling system configuration is presented in Figure 5-3. The
multiple-event cartridge sampling system is comprised of the following primary components:
-------
Manifold
By-Pass Pump
3-hr Duplicate
or
24-hr Sample
Mass Flow
Controller
j
To
Atmosphere
Tl
CKJ
oo
ft>
ss. o
O 3
Oilless Vacuum Pump
o:/s/g/morris/3797/pams/hobson/hob3.ppt
^ O
to vo
O
^o
oo
Figure 5-3. Schematic of a Typical Multiple-Event Carbonyl Cartridge Sampling System
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 13 of 31
Inlet probe and manifold assembly - Constructed of glass (see Figure 5-1) or stainless
steel. Used as a conduit to extract sample air from the atmosphere at the required
sampling height and distribute it for collection.
By-pass pump - A single- or double-headed diaphragm pump, or a caged rotary
blower. Used to continuously draw sample air through the inlet probe and manifold
assembly at a rate in excess of the sampling system total uptake. All excess sample air
is exhausted back to the atmosphere.
Sample pump - An oilless vacuum pump, capable of achieving an inlet pressure of
-25 inches Hg continually. Used to extract sample air from the manifold assembly and
pull it through the sample cartridges during collection.
Sample inlet line - Chromatographic-grade stainless steel tubing. Used to connect the
sampler to the manifold assembly. This line should be kept as short as possible.
Ozone scrubber - A Denuder or Cartridge type of O3 scrubber. Used to remove
ambient O3 from the sample air stream prior to exposure to the sample cartridge.
Sample cartridges - A plastic housing containing silica gel or CIS solid sorbent (see
Section 4.4 of Method TO-11A in Appendix D) coated with DNPH. Used to contain
the collected sample for transportation and analysis.
Adjustable orifice and mass flow meter assembly, or electronic mass flow
controller - An indicating flow control device(s). Used to maintain a constant flow rate
(± 10%) over a specific sampling period under conditions of changing temperature
(20-40°C) and humidity (0-100% relative).
Microprocessor - An event control and data acquisition device. Used to allow
unattended operation (i.e., activation and deactivation of each sampling event) of the
collection system, and to record sampling event specific process data (i.e., start and end
times, elapsed times, collection flow rates, etc.).
Check valves, solenoid valves, or a multi-port rotary valve - Eight stainless steel
check valves, eight solenoid valves with electric-pulse-operated or low temperature
coils, stainless steel bodies, and Viton® plunger seats and o-rings, or 1 multi-port
stainless steel body rotary valve with Viton® o-rings. Used to provide access to or
isolation of the inlet side of the sample cartridges.
Solenoid valves or a multi-port rotary valve - Eight solenoid valves with
electric-pulse-operated or low temperature coils, stainless steel bodies, and Viton®
plunger seat and o-rings, or 1 multi-port stainless steel body rotary valve with Viton®
o-rings. Used to provide access to or isolation of the outlet side of the sample
cartridges.
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 14 of 31
Tubing and fittings (Stainless steel or Teflon®) - Hardware for isolation and
interconnection of components. Used to complete system interconnections. All
stainless steel tubing in contact with the sample prior to analysis should be
chromatographic grade stainless steel and all fittings should be 316 grade stainless steel.
Note that if the manifold is heated, stainless steel tubing should be used because of the
potential of off-gassing of the tubing.
Note: Elapsed-time indicators installed in-line with sample pumps can provide
backup documentation that all samples ran for 180 minutes and can indicate that
a malfunction occurred with the programmable timers or that power was
interrupted.
5.2.2 Multiple-event Sampling Procedures
Samples are collected on individual solid sorbent sample cartridges using a single pump
and one or more flow control devices. An oil-less vacuum pump draws ambient air from the
sampling probe and manifold assembly through the sample cartridge at a constant flow rate
during each specific sampling event.
A flow control device(s) is used to maintain a constant sample flow rate through each
sample cartridge over each specific sampling period. The flow rate used is a function of the
desired total volume of ambient air sampled and the specified sampling period. The flow rate is
calculated as follows:
VxlOOO
F = - <5-2)
where:
F = flow rate (milliliters/minute)
V = desired total volume of ambient air sampled (liters)
1000 = milliliters in a liter
T = sample period (hours)
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 15 of 31
60 = minutes in an hour
For example, if the desired total volume of ambient air to be sampled is 168 L over each
individual 3-hour cartridge collection episode, the flow rate specific to each cartridge collection
episode is calculated as follows:
„ 168x1000 -.. ...... . . .
F = = 933 millihters/mmute TS-31
3x60 pjj
During operation, the microprocessor control device is programmed to activate and
deactivate the components of the sample collection system, consistent with the beginning and
end of each individual sample collection period.
Cartridge sampling systems can collect sample from a shared sample probe and
manifold assembly as described in Section 5.2.3 or from a dedicated stainless steel sample probe,
manifold assembly, and by-pass pump. If a dedicated probe, manifold assembly, and by-pass
pump are used, a separate timer device should be incorporated to start the by-pass pump several
hours prior to the first sampling event of a multiple-event collection period to flush and condition
the probe and manifold assembly components. The connecting lines between the manifold
assembly and the sampling system should be kept as short as possible to minimize the system
residence time.
The flow rate through each sample cartridge should remain relatively constant over the
entire collection period of each sampling event. Each adjustable orifice and mass flow meter
assembly, or mass flow controller, used as a flow control device should be calibrated against a
primary flow measurement standard (i.e., a bubble flow meter, etc.). Calibrations should include
multiple points of comparison (i.e., indicated flow versus measured flow), across the entire range
of the flow control device at increments reflecting 10% of the range. Calibration curves are
generated from these comparisons and are used to set actual desired flow rates based on the flow
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 16 of 31
rates indicated by the flow control devices. Calibration of the flow control devices should be
repeated periodically according to program specific QA/QC schedules as developed by the user.
Generic steps for operating a typical multiple-event sample collection system are as
follows:
1. Set the sampling system to the desired sample collection flow rate(s)
(i.e., referencing the corresponding calibration curve(s) and considering the
desired total volume of ambient air to be sampled and the sampling period for
each sampling event).
2. Program the microprocessor event control system to start and stop sample
collection consistent with program specific collection frequency requirements.
3. Attach all sample cartridges to the sampling system.
4. Record the start and end time of each collection event and the corresponding
flow rate onto the sampling field data sheet and calculate an average flow rate.
The microprocessor event control and data acquisition system should
automatically store these data for each collection event. The final total volume
of ambient air sampled should be close to the desired total volume.
5. Remove each sample cartridge (i.e., one at a time), cap both ends, and attach an
identifier to each (i.e., again, one at a time to avoid mislabeling). Sample event
number, sample type, location, collection date, should be recorded on the field
data sheet.
6. Place cartridges in tightly enclosed transport containers and transport the samples
and corresponding information to the central laboratory for preparation and
analysis.
5.2.3 Sample Probe and Manifold
A sample probe and manifold assembly should be used to provide a representative air
sample for collection and subsequent analysis. Sample probe and manifold assemblies are
commercially available or can be custom fabricated.
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 17 of 31
The sample probe is constructed of glass that is approximately 1 inch in outside
diameter (O.D.). The inlet of the sample probe is configured with an inverted funnel,
approximately 4 inches O.D. The sample manifold is constructed of glass, approximately 1 and
!/2 inches O.D. The manifold has ports used for sample distribution. The number of ports located
on the manifold must be equal to or greater than the total number of monitoring systems that
sample will be delivered to. The port nearest to the inlet of the manifold should be reserved for
VOC sampling; the second port or any other port may be used for carbonyl sampling.
Teflon® bushings are used to connect sample lines to the manifold. Because the
manifold and ports are constructed of glass, care must be taken to not place excessive stress on
the assembly to avoid breakage. For VOC sampling, the sample lines should be constructed of
1/8 inch O.D. stainless steel tubing. The 1/8 inch tubing is flexible and will accommodate the
flow rates typically associated with VOC sample collection. The sample lines should be kept as
short as possible to reduce sample transfer time. For carbonyl sampling, the sample lines should
be constructed of 1/4 inch O.D. stainless steel tubing; the scrubber and the carbonyl sample
cartridge holder assembly should be positioned as close to the manifold as possible.
A blower and bleed adapter are located at the exit end of the sample manifold. The
blower is used to pull sample air through the probe and manifold and the bleed adapter is used to
control the rate at which the sample air is pulled through the manifold. An excess of sample air
is pulled through the sample probe and manifold to prevent back diffusion of room air into the
manifold and to ensure that the sample air is representative of outside ambient air. Sample air
flow through the sample probe and manifold should be at least two times greater than the total air
flow being removed for collection and analysis by all systems on the manifold.
The vertical placement of the sample probe and inlet funnel should be at a height of 3 to
15 meters above ground level. Because the O3 monitoring requirements involve multiple-
pollutant measurements, this range serves as a practical compromise for probe position. In
addition, the probe inlet should be positioned more than 1 meter, both vertically and horizontally,
away from the housing structure. The probe inlet should be positioned away from nearby
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 18 of 31
obstructions such as a forest canopy or building. The vertical distance between the probe inlet
and any obstacle should be a least two times the height difference between the obstacle and the
probe inlet. Unrestricted air flow across the probe inlet should occur within an arc of at least
270 degrees. The predominant and second most predominant wind direction must be included in
this arc. If the probe inlet is positioned on the side of a building, a 180 degree clearance is
required. More specific details of probe positioning are presented in the "Enhanced Ozone
Monitoring Network Design and Siting Criteria Guideline Document."3 The glass probe should
be reinforced or supported along the straight vertical axis of the assembly. Typically this support
is provided by routing the probe shaft through a rigid section of metal or plastic tubing that is
secured to the housing structure.
The manifold can be positioned in either a horizontal or vertical configuration.
Figure 5-4 presents the manifold assembly in the vertical configuration. Figure 5-5 presents the
manifold assembly in the horizontal configuration. If the horizontal configuration is used, the
sample ports must point upward so that material that may be present in the manifold will not be
transferred into the sample lines.
With continuous use the sample probe and manifold can accumulate deposits of
particulate material and other potential contaminants. The sample probe and manifold should be
cleaned to remove these materials. The recommended frequency for cleaning is quarterly. To
clean the assembly, disconnect the sample lines and blower from the manifold. The sample lines
and blower are not cleaned. For safely, electric power to the blower should be terminated until
the cleaning process is completed. Disassemble the individual components by disconnecting the
probe, manifold, collection bottle, and coupling devices from each other. The individual
components should then be cleaned using heated high purity distilled water and a long handled
bottle brush. The components should then be rinsed with the distilled water and allowed to dry
completely before reassembling. If required, mild glass cleaner or detergent can be used to clean
particularly dirty components. However, care should be taken to select cleaners and detergents
that are advertised to have low organic compound content and the number of rinses performed
should be increased to ensure that all associated residues are removed.
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 19 of 31
Sampling cane
Cane
support
Shelter roof
Shelter
wall
First port dedicated to VOC sampling
Second port dedicated to Carbonyl sampling
Sample manifold
(with sufficient number of
ports to individually support
all monitoring conducted)
Bleed adapter
(flow control)
Blower
and mount
Collection
bottle
Shelter
wall
To
atmosphere
Exhaust
tube
o/l/B/motr/3797/paml/^lll.ppt
Figure 5-4. Vertical Configuration
-------
Sampling cane
Sample
Cane
support
Shelter roof
Shelter
wall
First port dedicated to VOC sampling
Second port dedicated to Carbonyl sampling
Bleed adapter
(flow control)
Blower
and mount
Sample manifold
(with sufficient number of
ports to individually support
all monitoring conducted)
Exhaust
tube
Collection
bottle
Shelter
wall
atmosphere
Tf
pa oo
<. |
2. o
O 3
o/s/g/morr/3797/pams/shelt2.ppt
Figure 5-5. Horizontal Configuration
to o
o vo
O oo
*~^5 CZ>
^ ^
^ oo
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 21 of 31
5.2.4 Multiple-event System Specifications
The use of sample cartridges to practically address the sampling frequency and schedule
for carbonyl compounds specified in Table 5-1 requires the use of multiple-event cartridge
sampling systems. The use of a single-event system to collect eight back-to-back 3-hour
cartridge samples would require that an operator be physically present on site to manually
complete the activities associated with the start and stop of each sampling event.
To ensure that a multiple-event sample collection system will meet the user's program
needs, system specifications and other pertinent general considerations should be presented to,
and addressed by, the candidate vendor(s) prior to procurement. Primary system specifications
are presented below. However, additional system specifications and considerations may be
added at the discretion of the user.
An in-depth, detailed manual covering all aspects of the sample collection system
(i.e., operation, maintenance, etc.) must be provided by the vendor.
The overall size of the sampling system should be kept as compact as possible.
The sampling systems are usually installed into existing sampling site shelters
where many other parameters (i.e., criteria pollutants concentrations,
meteorological conditions, etc.) are also measured. Each of the other parameters
requires separate instrumentation and consequently the shelters can become very
crowded.
The sample collection system should meet all applicable electrical and safety
codes, operate on standard 110 Vac power, and incorporate a main power fuse or
circuit breaker. Specific potential electrical hazards and/or other safety
considerations should be detailed in a supplied user's manual.
The overall configuration, and components comprising that configuration, should
allow for simple operation, maintenance, and service of the sample collection
system. Materials used in the construction of components of the sample
collection system should exhibit nonbiasing characteristics. The components
themselves should generally conform to the descriptions presented in
Section 5.2.1. All surfaces that come in direct contact with sampled air should
be constructed of glass, stainless steel, or Viton®.
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 22 of 31
To avoid cross-contamination, the sample collection system must have
provisions to isolate the inlet and outlet of each sample cartridge when that
cartridge is not collecting sample.
The sampling system must incorporate or provide for removal of O3, consistent
with the O3 scrubber designs detailed in Section 5.1.
Ideally, the sampling system should be able to accommodate the most intensive
sample collection event frequency presented in Figure 5-6 on an automated
unattended basis, and simultaneously accommodate a duplicate collection for one
of the 3-hour sampling events as recommended for quality control purposes.
These requirements mean that the sampling system should have the capability to
collect the following during any given 24-hour period:
Eight 3-hour cartridge samples;
One 3-hour duplicate cartridge sample, collected concurrently with one of
the eight 3-hour cartridge samples; and
One 24-hour cartridge sample, collected concurrently with the eight 3-hour
cartridge samples, but not concurrent with a duplicate 3-hour cartridge
sample.
It is imperative that the sample collection system have the collection capabilities
detailed above. If not, a second sampling system would be required to address
the 24-hour sample collection, and consequently more overall labor and space
would be needed to fully address the network monitoring requirements.
The ability of the sampling system to perform sample collections as presented
above would require the operator to visit the site only twice during the 24-hour
period being characterized; once to install sample cartridges prior to sampling
and once to remove sample cartridges containing the collected samples. Each
24-hour period is scheduled to begin at 12:00 A.M. (i.e., midnight) and end at
11:59 P.M. of the day being characterized. The sampling system must be able to
automatically address these periods (i.e., must be able to start and stop at the
specified times without requiring an operator to go to the site and manually
actuate the system).
The sampling system should incorporate a microprocessor event control and data
acquisition device. At a minimum this microprocessor should be able to be
programmed to control the start and stop times of every collection event within a
given 24-hour sampling duration. The microprocessor should also be able to
simultaneously collect and store all the sample collection process data pertaining
to each sampling event as follows:
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 23 of 31
Isolated Area Network Design
(4)
(3)
Central business district
Urbanized fringe
U1
U3
Note:
U1 and U2 represent the first and second most
predominant high ozone day morning wind direction.
U3 represents the high ozone day afternoon wind direction.
o: s/g/m orr/3 7 9 7/p ams /i s ol ate d. p p t
(1), (2), (3), and (4) are different types of PAMS sites (See Table 5-1).
Figure 5-6. Isolated Area Network Design
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 24 of 31
Start and stop times for each sample collection; and
Beginning and ending collection flow rates for each cartridge collection.
The microprocessor should incorporate a battery backup system to address power
failure situations. Incorporation of a battery backup system should result in
fewer invalidated sample collections and a higher sample collection completion
rate. The battery backup system would ensure that all programmed control
activities and collection process data would be retained for a predetermined
interval should standard power to the system be interrupted. Retaining the
programmed control activities would allow sampling to resume automatically at
the next programmed event time when standard power is once again established
to the sampling system. Retaining the collection process data obtained for
samples collected prior to the termination of standard power would allow these
samples to be qualified as valid or invalid based on sampling start and stop times
and initial and flow rates. Although not absolutely necessary, the incorporation
of a miniature printer that would allow for a report style listing of all sample
collection process data would be advantageous.
• Expedient and responsive vendor support should be a mandatory requirement and
primary consideration when procuring a multiple-event cartridge sampling
system. The user should specify that the vendor will maintain an adequate
supply of replacement parts and a staff of qualified service technicians to ensure
that the absolute minimum number of sample collection events are missed should
a sample collection system failure occur. The user should specify that the vendor
guarantee that parts/components be delivered to the sampling site within two
working days of order placement. The user should also specify that a sample
collection system delivered to the vendor for repair be serviced and returned to
the user within seven working days.
The manufacturer of a carbonyl sampling device and experience at some of the PAMS
sites indicate that the carbonyl sampler should not be located inside a shelter but outside to
alleviate the possibility of off-gassing from the shelter interfering with the samples. The
sampling methodology itself does not specify the location of the sampler. The wisdom of
locating the carbonyl sampler outside rather than inside a shelter would depend upon the
composition of the shelter and the security of the sampler if it is located in an outside area.
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 25 of 31
5.3 Process Blanks
To ensure data quality and obtain quantitative carbonyl compound concentrations, the
collection of blanks is necessary. For the purposes of PAMS, there are three types of blanks used
to ensure data quality: certification blanks, field blanks, and trip blanks. The guidance given here
should be considered a minimum and users are encouraged to build upon this guidance as
necessary.
Certification blanks consist of a minimum of three laboratory blank cartridges
that are eluted with acetonitrile and analyzed to verify the acceptability of a
specific cartridge lot from a commercial vendor. Certification blanks are
analyzed for each specific lot used for sampling. The mean mass plus 3 standard
deviations (x + 3s) for the group of three laboratory blanks is used to assess
acceptability.
Field blanks are blank cartridges which are sent to the field, connected to the
sampling system and treated identically to the samples except that no air is drawn
through the cartridge. Field blanks are used to assess the background carbonyl
levels for cartridges used during the ambient sample collection process.
Trip blanks are blank cartridges of the same lot that are sent to the field, stored,
and returned to the laboratory with the sample cartridges. Trip blanks are
optional and may be used to resolve contamination problems determined from
the field blanks. Trip blanks can be used to determine whether the contamination
occurred during the sampling process or during the shipping and storage process.
5.3.1 Blank Criteria
The acceptance criteria for blanks are discussed below. The criteria for certification are
considered conservative; most certification blank results will be well below these criteria. If the
mean mass plus 3 standard deviations (x ± 3s) for the group of three laboratory blanks meets the
criteria, then no further certification or laboratory blanks are required for a particular lot. If large
differences are observed for the 3 laboratory blank samples, additional laboratory blanks should
be analyzed to obtain values for the mean and standard deviation. For the certification blanks to
be acceptable, the following criteria should be met:
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 26 of 31
• Formaldehyde: <0.15 g/cartridge*
• Acetaldehyde: <0.10 g/cartridge
• Acetone: <0.30 g/cartridge
• Other aldehydes or ketones, concentration (per individual component):
<0.10 jig/cartridge.
* The equivalent formaldehyde concentration in ppbv as taken from Table 3 in EPA Compendium Method TO-11A
(see Appendix D) is 0.679 ppbv for a 180 L sample volume.
Using good techniques and collection systems (not mixing lots or vendors), field blanks
should consistently be at levels that are less than 2 times the average measured laboratory blank
value for a specific lot. The laboratory blank is a cartridge blank used for lot certification that
has never been shipped to the field. If field blanks do not meet these criteria, corrective action is
required. Sites that are unable to achieve these levels for field blanks must determine the source
of contamination. An assessment of the air in the sampling shelter may also provide useful
information in the determination of sources for field blank and sample contamination.
As a minimum, a sampling system blank sample should be collected at least on an
annual basis before initiation of sampling. Collection of a pre- and post-sampling blank is
strongly recommended to aid in the qualification of data. If the sampler is subjected to only a
single blank audit, a failure to meet QA/QC limits will leave open the question of whether the
previous year's data should be flagged or not. It is possible for a sampler to become
contaminated (or appear to become contaminated) during the down season, in which case there
would be no reason to invalidate the data from the previous year. Pre- and post-season audits
remove the ambiguity. Collect a sampler blank using carbonyl-free air when possible. Generate
carbonyl free air by purging air through acidic DNPH solution in a bubbling device or DNPH-
coated cartridge. Alternatively, measure the carbonyl content of the air using a DNPH-coated
cartridge and subtract the carbonyl content in the air from that in the sampler blank. Before
collecting the sampler blank, flush the system using the same procedures as used for collecting a
sample.
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 27 of 31
5.3.2 Frequency of Collection
At least one field blank, or the square root of the field sample size, whichever is larger,
should be collected and analyzed with each sample lot collected at the site. The square root of the
sample size is used to result in more field blanks for a smaller sample size and fewer field blanks
for a larger sample size. For example, if 100 field samples will be taken at the site, then 10 field
blank samples (the square root of 100) are collected and analyzed. If multiple lots are used,
ensure that each lot has the necessary number of associated field blanks. Certification blanks are
not included in the number of field blanks. Certification blanks are analyzed in addition to field
blanks to verify acceptability of a specific cartridge lot from the vendor. At a minimum, three
laboratory blanks from each lot are used for certification. Table 5-2 gives an example collection
schedule for a field samples from a single lot.
Table 5-2. Example Schedule for the Collection of Blanks
Field Sample Size
50
100
200
Lab Blanks for Lot
Certification
3
3
3
Field Blanks (square root of
the sample size)
7
10
16
Since field blank samples may not be collected on every sampling day, the issue of
maintaining consistency in the overall data treatment using blank subtraction is a challenge. For
PAMS blank subtraction must be performed using the average field blank mass obtained for each
field sample lot. Using the information in Table 5-2 as an example, for a sample size of 100, the
average mass for the 10 field blank samples is subtracted from each of the 100 samples. Again,
it is important that cartridge lot be tracked and the appropriate number of field blanks be
collected and subtracted from the samples for each lot used.
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 28 of 31
5.4 Breakthrough Analysis
Method TO-11A requires the use of a back-up cartridge during the first sampling event.
If less than 10% of the analyte is collected on the back-up cartridge, then back-up cartridges are
only required for 10% of the field samples. If more than 10% of the analyte is collected on the
back-up cartridge, then use back-up cartridges for all sampling events. Breakthrough is more
likely to occur when sampling at high flow rates, when sampling very dry or very humid air,
when sampling air containing high levels of oxides, and when sampling air containing high levels
of carbonyl compounds. Perform breakthrough analyses on the 24-hour sample or on the
duplicate 3-hour sample. Be careful in determining the flow rate because two cartridges installed
in series create a higher pressure drop, decreasing the sampling rate. If breakthrough occurs,
minimize the breakthrough by replacing the ozone scrubber more frequently, sampling at a lower
flow rate, using larger capacity cartridges, or heating the cartridges slightly to prevent moisture
condensation when sampling very humid air.
5.5 Collection of Collocated Samples
A collocated sample is collected from one manifold by two independent sampling
devices in the same sampling period. Collect collocated samples as indicated in Table 5-2.
Analyze the collocated samples in replicate. The replicate analyses should agree to within ±10%
and the means of the replicate analyses for the collocated samples should agree to within ±20%.
If the collocated samples do not agree to within ±20% and the replicate analyses are within
±10%, check the samples to ensure that they are truly collocated and check the sample flow rates
to ensure that the sampler is working correctly. Also verify that the sampler is not leaking by
performing a leak check as described in Section 10.2 of TO-11A (see Appendix D).
5.6 Quality Assurance and Quality Control
General quality assurance and quality control requirements are provided in Section 13.6
of Method TO-11A (see Appendix D). Each laboratory should develop SOPs for the sampling
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 29 of 31
and analysis of carbonyls and should develop criteria for sampling and analysis that are specific
to the laboratory. Table 5-3 provides the quality assurance and quality control procedures
consistent with Method TO-11 A.
5.7 General Cartridge Handling Guidelines
Unintentional exposure of the DNPH cartridges and eluted samples to aldehyde and
ketone sources can result in contamination of the samples, creating a positive bias in the
collected data. Various aldehydes and ketones are ubiquitous in the environment. For example,
biological processes can produce formaldehyde, acetone, and acetaldehyde on peoples' skin and
in peoples' breath. Wear polyethylene gloves at all times when handling the DNPH cartridges
during sampling collection and analysis. In addition, laboratory air often holds high
concentrations of acetone (and sometimes formaldehyde). Measure background levels of
carbonyls in the laboratory air using a DNPH cartridge and sample pump. If high background
levels are present, handle the cartridges in a nitrogen-purged glove box or under a purge of
carbonyl free air. Labeling inks, adhesives, and packing containers are all additional sources of
contamination. Avoid packing cartridges in old newspapers, writing directly on the cartridges
with ink, or placing adhesive labels directly on the cartridges. Additionally, DNPH is light
sensitive. Always protect the cartridges from direct sunlight.
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 30 of 31
Table 5-3. Quality Assurance and Quality Control Criteria
Parameter
Flow calibration
Mass flow meter
calibration factor
Leak check
Sampler blank
Collocated samples
Back-up cartridges
Trip blanks
Field blanks
Spiked cartridges
Multi-point
calibration
Continuing
calibration standard
Method detection
limits
Replicate injections
Performance
evaluation sample
Frequency
Each sampling event,
pre- and post-checks
Every quarter
Each sampling event,
pre- and post-checks
Pre- and post-seasons
10% of field samples
10% of field samples
10% of field samples
10% of field samples
10% of field samples
Every 6 months
Every analytical run
Annually or after
each instrument
change
10% of samples
Before and after
samples
Limits
±10%
1.0±0.1
No air flow
>MDL
±20%
10% of total on back-up
cartridge
<0.15 |ig
formal dehy de/cartri dge
<0.15 |ig
formal dehy de/cartri dge
80 to 120% recovery
0.999
±10%
<0.1 ppbvfor 180L
sample volume
±10%
±15%
Corrective Action
Mark sample as
suspect
Repair mass flow
meter
Check for leaks
Clean sampler,
qualify data if
required
Mark sample as
suspect
Use back-up
cartridges for all
samples
Blank correct data
Blank correct data
Flag data
Recalibrate
Recalibrate
Modify instrument
as needed
Reanalyze samples
Reanalyze samples
-------
Section: 5
Revision: 1
Date: 09/30/98
Page: 31 of 31
5.8 References
1. U.S. Environmental Protection Agency. Code of Federal Regulations. Title 40, Part 58.
Ambient Air Quality Surveillance, Final Rule Federal Register, Vol. 58, No. 28,
February 12, 1993.
2. Compendium of Methods for the Determination of Toxic Organic Compounds in
Ambient Air. Compendium Method TO-11 A. Determination of Formaldehyde in
Ambient Air Using Adsorbent Cartridge Followed by High Performance Liquid
Chromatography (HPLC). EPA-625/R-96/010b. Cincinnati, OH: U.S. Environmental
Protection Agency, 1997.
3. Grassick, D. and R. Jongleux. Enhanced Ozone Monitoring Network—Design and Siting
Criteria Guideline Document. Contract No. 68-DO-0125. Research Triangle Park, NC:
U.S. Environmental Protection Agency, 1991.
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 1 of 24
Section 6.0
Guidance for PAMS Meteorological Monitoring
6.1 Background
Title 40 Part 58 of the Code of Federal Regulations1 required the States to establish a
network of Photochemical Assessment Monitoring Stations (PAMS) in ozone nonattainment
areas which were classified as serious, severe, or extreme. The regulation states that each
PAMS program must include provisions for enhanced monitoring of ozone, the precursors to
ozone, and both surface and upper-air meteorological conditions. Although the PAMS rule
establishes a requirement for meteorological monitoring, it does not provide specifics; e.g., a list
of the meteorological variables to be monitored. Discussions to develop such a list took place in
the Spring of 1994. Recommendations based on these discussions were issued in April 19942
and incorporated in an early draft of this document. The list of variables has since been revised
to reflect input from the review of this and subsequent drafts. Currently, the list of
meteorological variables includes: wind direction, wind speed, temperature, humidity,
atmospheric pressure, precipitation, solar radiation, UV radiation, and mixing height. Table 6-1
provides an overview of the requirements for monitoring these variables.
The remainder of this section is organized as follows: Section 6.2 describes the PAMS
site types as established by the PAMS rule. Section 6.3 provides material on the application of
PAMS meteorological data. Sections 6.4 and 6.5 provide details related to measurement of
surface and upper-air meteorological variables, respectively. Section 6.6 provides a list of
references.
Users are referred to the "Quality Assurance Handbook for Air Pollution Measurement
Systems, Volume IV: Meteorological Measurements"3 for recommended procedures for quality
assurance and audit activities. The procedures provided in "On-Site Meteorological Program
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 2 of 24
Table 6-1. Overview of PAMS Meteorological Monitoring Requirements
QUESTION
Where to monitor?
How many sites?
When to monitor?
How long?
What variables?
What interval?
What levels?
ANSWER
All serious, severe, and extreme ozone nonattainment areas
2 to 5 surface sites per network plus one upper-air site
Routine continuous monitoring during the PAMS monitoring
season (3 months per year minimum)
Until the area is redesignated as attainment for ozone
Wind Directiona
Wind Speed3
Air Temperature"
Humidity a
Solar Radiation13
Ultraviolet Radiation13
Barometric Pressure13
Precipitation13'0
Surface measurements should be continuous and should be
reported hourly.
Upper-air measurements (profiles of wind and temperature)
be made at least 4 times per day.
should
Surface measurements should be made at 2 meters (temperature
and humidity) or 10 meters (wind direction and wind speed).
Other surface measurements are nominally made at about 2
meters.
a A required measurement for all PAMS sites.
b A required measurement for at least one site per PAMS area.
0 Precipitation data from other sources (National Weather Service or others) are acceptable on a case-by-case basis.
Guidance for Regulatory Modeling Applications"4 should be followed for processing of
meteorological measurements.
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 3 of 24
6.2 PAMS Site Types
The PAMS requirements were designed to provide as much information as practicable on
the roles of ozone precursors, pollutant transport, and local meteorology in the photochemical
process. Specific provisions of the Rule require the establishment and operation of four
monitoring sites in each PAMS area; each of the sites has specific data quality objectives
(DQOs). The four sites are as follows:
Site #1 - Upwind and background characterization site These sites are
established to characterize upwind background and transported ozone and
its precursor concentrations entering the area and will identify those areas
which are subjected to overwhelming incoming transport of ozone. The
#1 Sites are located in the predominant morning wind direction from the
local area of maximum precursor emissions and at a distance sufficient to
obtain urban scale measurements. Typically, these sites will be located
near the upwind edge of the photochemical grid model domain.
Site #2 - Maximum ozone precursor emissions impact site. These sites are
established to monitor the magnitude and type of precursor emissions in
the area where maximum precursor emissions representative of the
Metropolitan Statistical Area (MSA)/Consolidated Metropolitan Statistical
Area (CMSA) are expected to impact and are suited for the monitoring of
urban air toxic pollutants. The #2 Sites are located immediately
downwind (using the same morning wind direction as for locating Site #1)
of the area of maximum precursor emissions and are typically placed near
the downwind boundary of the central business district (CBD) or primary
area of precursor emissions mix to obtain neighborhood scale
measurements. Additionally, a second #2 Site may be required
depending on the size of the area, and it should be placed in the second-
most predominant morning wind direction.
Site #3 - Maximum ozone concentration site. These sites are intended to monitor
maximum ozone concentrations occurring downwind from the area of
maximum precursor emissions. Locations for #3 Sites should be chosen
so that urban scale measurements are obtained. Typically, these sites are
located 10 to 30 miles from the fringe of the urban area.
Site #4 - Extreme downwind monitoring site. These sites are established to
characterize the extreme downwind transported ozone and its precursor
concentrations exiting the area and will identify those areas which are
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 4 of 24
potentially contributing to overwhelming ozone transport into other areas.
The #4 Sites are located in the predominant afternoon downwind direction
from the local area of maximum precursor emissions at a distance
sufficient to obtain urban scale measurements. Typically, these sites will
be located near the downwind edge of the photochemical grid model
domain.
6.3 Applications of PAMS Meteorological Data
Meteorology is a critical element in the formation, transport, and ultimate disposition of
both ozone and its precursors. Consequently, meteorological data are essential to the
development and evaluation of ozone control strategies. Other types of evaluations which
depend on meteorological data include photochemical modeling, receptor modeling, emissions
tracking, and trends analysis. Other application areas associated with the PAMS meteorological
measurements are indicated in Table 6-2. Examples of applications can be found in the Draft
Guidelines for Quality Assurance and Management of PAMS Upper-Air Meteorological Data.5
Table 6-2. Application of the PAMS Meteorological Data
Variable
Wind Direction"
Wind Speed3
Air Temperature3
Humidity3
Pressure13
Precipitation13
Solar Radiation13
UV Radiation13
Photochemical
Modeling
/
/
/
/
/
/
/
Diagnostic
Analysis
/
/
/
/
/
/
/
/
Receptor
Modeling
/
/
/
a To be measured at multiple sites in each PAMS area.
b To be measured at one representative site in each PAMS area.
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 5 of 24
6.4 Surface Meteorological Monitoring
A minimum level of surface meteorological monitoring is required for each PAMS site
regardless of the site type (see Section 6.2). The minimum level includes measurements of wind
direction, wind speed, ambient temperature, and humidity (e.g., dew point or relative humidity).
In addition, measurements of solar radiation, ultraviolet radiation, barometric pressure, and
precipitation are required for at least one site in each PAMS network.
6.4.1 Siting and Exposure
The selection of an appropriate site for the surface meteorological measurements depends
on the intended use of the data; i.e., the Data Quality Objectives (DQOs). Ideally, for general
application, the site should be located in a level open area away from the influence of
obstructions such as buildings or trees. The area surrounding the site should have uniform
surface characteristics.6'7'8 Although it may be desirable to collocate the surface meteorological
measurements with the ambient air quality measurements, collocation of the two functions may
not be possible at all PAMS sites without violating one or more of the above criteria. Siting and
exposure requirements specific to each of the PAMS surface meteorological variables are
discussed in subsequent sections.
Surface meteorological measurements in urban areas, where compliance with siting and
exposure criteria may be precluded by the close proximity of buildings and other structures,
present special difficulties. In such cases, the individual involved in the site selection needs to
assess the likelihood that the data which may be collected at a given location will conform to the
DQOs. In all cases, the specific site characteristics should be well documented. This
documentation of the specific site characteristics is especially important in areas where surface
characteristics and/or terrain are not uniform and whenever standard exposure and siting criteria
cannot be met.
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 6 of 24
6.4.2 Specifications
System specifications for the surface measurements are given in Table 6-3. The
recommended sampling interval of the meteorological sensors by the data acquisition system is
10 seconds. Data for all variables should be processed to obtain one hour averages. The data
acquisition system clock should have an accuracy of ±1 minute per week.
Table 6-3. System Specifications for Surface Meteorological Measurements3
Variable
Wind Speed
Wind Direction
Air Temperature
Dew Point
Relative Humidity
Solar Radiation
UV Radiation
Barometric Pressure
Precipitation
Range
0.5 to 50 m/s
0 to 360 deg.
-20 to 40 °C
-30 to +30 °C
0 to 100 %RH
0 to 1200 W m'2
0 to 12 W m'2
SOOtollOOhPa
0 to 30 mm/hr
Accuracy
±0.2 m/s + 5%
±5 deg.
±0.5 °C
±1.5°C
±3%RH
±5%
±5%
±3hPa
±10%
Resolution
0.1 m/s
Ideg.
0.1 °C
0.1°C
0.5 %RH
10 W m'2
0.01 Wm'2
0.5 hPa
0.25 mm
Time/Distance
Constants
5 m (63% response)
5 m (50% recovery)
60 s (63% response)
30 minutes
60 s (63% response)
60 s (99% response)
60 s (99% response)
60 s (63% response)
60 s (63% response)
aQuality assurance guidance for auditing these values is provided in Quality Assurance Handbook for Air Pollution
Measurement Systems. Volume IV - Meteorological Measurements. EPA/600/R-94/038d. U.S. Environmental
Protection Agency, 1995.
6.4.3 Wind Speed and Wind Direction
Wind speed and direction are essential to the evaluation of transport and dispersion
processes of all atmospheric pollutants. Wind speed is typically measured with a cup or propeller
anemometer; wind direction is typically measured with a wind vane.
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 7 of 24
The standard height for surface layer wind measurements is 10 m above ground level.3'7'9
The location of the site for the wind measurements should ensure that the horizontal distance to
obstructions (e.g., buildings, trees, etc.) is at least ten times the height of the obstruction.9 In
urban areas, where the "ten times" criterion may not be met, one should provide a protocol for
invalidating the measurements for the problem directions. Evans and Lee10 provide a discussion
of the validity of 10-meter wind data in an urban setting where the average obstruction height is
of the same order as the wind measurement height.
An open lattice tower is the recommended structure for monitoring of meteorological
variables at the 10-meter level. In the case of wind measurements, certain precautions are
necessary to ensure that the measurements are not significantly affected by turbulence in the
immediate wake of the meteorological tower. To avoid such tower effects, the wind sensor
should be mounted on a mast a distance at least one tower width above the top of the tower, or if
the tower is higher than 10 m, on a boom projecting horizontally from the tower. In the latter
case, the boom should extend a distance at least twice the diameter/diagonal of the tower from
the nearest point on the tower. The boom should project into the direction which provides the
least distortion for the most important wind direction (i.e., into the prevailing wind).
There are several types of open lattice towers: Fixed, tilt-over, and telescopic. A fixed
tower is usually assembled as a one-piece structure from several smaller sections. This type of
tower must be sturdy enough so that it can be climbed safely to install and service the
instruments. Tilt-over towers are also one-piece structures, but are hinged at ground level. This
type of tower has the advantage of allowing the instruments to be serviced at the ground.
Telescopic 10m towers are usually composed of three sections, each approximately 4 m in
length. The top section is the smallest in diameter and fits inside the middle section which, in
turn, fits inside the base section. The tower can be extended to a height of 10 m by use of a hand
crank located at the lowest section. The top of the tower can be lowered to a height of about 4 m
providing easy access to the wind sensors. Telescopic and tilt-over towers are not generally
recommended for heights above 10m. Regardless of which type of tower is used, the structure
should be sufficiently rigid and properly guyed to ensure that the instruments maintain a fixed
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 8 of 24
orientation at all times. Instrumentation for monitoring wind speed and direction should never be
mounted on or near solid structures such as buildings, stacks, water storage tanks, cooling
towers, etc., because all such structures create significant distortions in the flow field.
A sensor with a high accuracy at low wind speeds and a low starting threshold is
recommended for PAMS applications. Light weight materials (e.g., molded plastic or
polystyrene foam) should be employed for cups and propeller blades to achieve a starting
threshold (lowest speed at which a rotating anemometer starts and continues to turn and produce
a measurable signal when mounted in its normal position) of < 0.5 m s"1. Wind vanes or tail fins
should also be constructed from light weight materials. The starting threshold (lowest speed at
which a vane will turn to within 5° of the true wind direction from an initial displacement of 10°)
should be < 0.5 m s"1. Overshoot must be < 25% and the damping ratio should lie between 0.4
and 0.7. The above information is summarized in Table 6-3.
6.4.4 Temperature
Temperature affects photochemical reaction rates and consequently, is an essential
variable for PAMS applications. Sensors used for monitoring ambient temperature include: wire
bobbins, thermocouples, and thermistors. Platinum resistance temperature detectors (RTD) are
among the more popular sensors used in ambient monitoring; these sensors provide accurate
measurements and maintain a stable calibration over a wide temperature range.
The standard height for surface layer ambient temperature measurements is 2 meters
above ground level.7 If a tower is used, the temperature sensor should be mounted on a boom
which extends at least one tower width/diameter from the tower. The measurement should be
made over a uniform plot of open, level ground at least 9 m in diameter. The surface should be
covered with non-irrigated or un-watered short grass or, in areas which lack a vegetation cover,
natural earth. Concrete, asphalt, and oil-soaked surfaces and other similar surfaces should be
avoided to the extent possible. The sensor should be at least 30m from any paved area. Other
areas to avoid include large industrial heat sources, rooftops, steep slopes, hollows, high
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 9 of 24
vegetation, swamps, snow drifts, standing water, and air exhausts. The distance to obstructions
for accurate temperature measurements should be at least four times the obstruction height. In
urban areas, one should be especially conscious of and avoid extraneous energy sources
(e.g., tunnels and subway entrances, rooftops, etc.).
Temperature measurements should be accurate to ±0.5°C over a range of-20 to +40°C
with a resolution of 0.1°C. The time constant (63.2%) should be < 60 seconds. Solar heating is
usually the greatest source of error and, consequently, adequate shielding is needed to provide a
representative ambient air temperature measurement. Ideally, the radiation shield should block
the sensor from view of the sun, sky, ground, and surrounding objects. The shield should reflect
all incident radiation and not reradiate any of that energy towards the sensor. The best type of
shield is one which provides forced aspiration at a rate of at least 3 m s"1 over a radiation range of
-100 to +1100 W m"2. Errors in temperature should not exceed ±0.25°C when a sensor is placed
inside a forced aspiration radiation shield. The sensor must also be protected from precipitation
and condensation, otherwise evaporative effects and other forms of radiational heating or cooling
will lead to a depressed temperature measurement (i.e., wet bulb temperature). Temperatures may
also be reported to AIRS AQS in °F, but metric is the preferred and recommended system of
units for meteorological measurements.
6.4.5 Atmospheric Humidity
Measurements of atmospheric humidity are essential to understanding chemical reactions
involving ozone precursors and water vapor. Measures of atmospheric humidity include vapor
pressure, dew point temperature, specific humidity, absolute humidity, and relative humidity.
There are several ways to measure the water vapor content of the atmosphere. The classical
measurement methods can be classified in terms of six scientific principles; examples are
provided in Table 6-4.
The standard height for humidity measurements is 2 m above ground level. The humidity
sensor should be installed using the same siting criteria as used for temperature. If possible, the
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 10 of 24
Table 6-4. Principles of Humidity Measurement3
Principle
Reduction of temperature by evaporation
Dimensional changes due to absorption of
moisture, based on hygroscopic properties of
materials
Chemical or electrical changes due to absorption
or adsorption
Formation of dew or frost by artificial cooling
Diffusion of moisture through porous membranes
Adsorption spectra of water vapor
Instrument/Method
psychrometer
hygrometers with sensors of hair, wood,
natural and synthetic fibers
electric hygrometers such as the Dunmore
Cell; lithium, carbon, and aluminum
oxide stirps; capacitance film
cooled mirror surfaces
diffusion hygrometers
infrared and UV absorption; Lyman-alpha
radiation hygrometers
aMiddleton, W.E.K. and A.F. Spillhaus, Meteorological Instruments, University of Toronto Press (1953).
humidity sensor should be housed in the same aspirated radiation shield as the temperature
sensor. The humidity sensor should be protected from contaminants such as salt, hydrocarbons,
and other particulates. The best protection is the use of a porous membrane filter which allows
the passage of ambient air and water vapor while keeping out particulate matter.
6.4.6 Solar Radiation
Solar radiation refers to the electromagnetic energy in the solar spectrum (0.10 to 4.0 jim
wavelength). The latter is commonly classified as ultraviolet (0.10 to 0.40 jim), visible light
(0.40 to 0.73 jim), and near-infrared (0.73 to 4.0 jim) radiation. About 97% of the solar radiation
reaching the earth's outer atmosphere lies between 0.29 and 3.0 jim.7 A portion of this energy
penetrates through the atmosphere and is either absorbed or reflected at the earth's surface. The
rest of the solar radiation is scattered and/or absorbed in the atmosphere before reaching the
surface. Solar radiation measurements are used in heat flux calculations, for estimating
atmospheric stability, and in modeling photochemical reactions.
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 11 of 24
Energy fluxes in the spectrum of solar radiation are measured using a pyranometer.
These instruments are configured to measure what is referred to as global solar radiation;
i.e., direct plus diffuse (scattered) solar radiation. The sensing element of the typical
pyranometer is protected by a clear glass dome to prevent entry of energy (wavelengths) outside
the solar spectrum (i.e., long-wave radiation). The glass domes used on typical pyranometers are
transparent to wavelengths in the range of 0.28 to 2.8 |im.
Solar radiation measurements should be taken in a location with an unrestricted view of
the sky in all directions. In general, locations should be avoided where there are obstructions that
could cast a shadow or reflect light on the sensor; light colored walls or artificial sources of
radiation should also be avoided. The horizon as viewed from the pyranometer should not
exceed 5 degrees. Sensor height is not critical for pyranometers; consequently, tall platforms or
rooftops are typical locations. Regardless of where the pyranometer is sited, it is important to
ensure that the level of instrument is maintained,11 and that the glass dome is cleaned as
necessary. To facilitate leveling, the pyranometers should be equipped with an attached circular
spirit level.
Manufacturer's specifications should match the requirements of the World Meteorological
Organization7 for either a secondary standard or first class pyranometer (see Table 6-5) especially
if the measurements are to be used for estimating heat flux. Photovoltaic pyranometers (which
usually fall under second class pyranometers) may be used for PAMS applications on a case-by-
case basis. The cost of photovoltaic sensors is significantly less than that of thermocouple-type
sensors; however, their spectral response is limited to the visible spectrum.
6.4.7 Ultraviolet Radiation
Ultraviolet (UV) radiation may be divided into three sub-ranges: UV-A (0.315
to 0.400 |im), UV-B (0.280 - 0.315 jam), and UV-C (0.100 - 0.280 |im). Due to absorption by
stratospheric ozone, the UV radiation reaching the surface of the earth consists primarily of
wavelengths longer than 0.28 |im (UV-A and UV-B ranges). The most important
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 12 of 24
Table 6-5. Classification7 of Pyranometers3
Characteristic
Resolution
Stability
Cosine Response
Azimuth Response
Temperature Response
Nonlinearity
Spectral Sensitivity
Response Time (99%)
Units
Wm-2
%FS year1
%
%
%
%FS
%
seconds
Secondary
Standard
±1
±1
<±3
<±3
±1
±0.5
±2
<25
First
Class
±5
±2
<±7
<±5
±2
±2
±5
<60
Second
Class
±10
±5
<±15
<±10
±5
±5
±10
<240
aQuality Assurance guidance for auditing these parameters is provided in Quality Assurance Handbook for
Air Pollution Measurement Systems. Volume IV - Meteorological Measurements. EPA/600/R-94/03 8d.
U.S. Environmental Protection Agency, 1995.
photochemically active chemical species at these wavelengths are ozone, nitrogen dioxide, and
formaldehyde. All three of these chemical species are important in the formation of ozone.
Pyranometers with a spectral response covering both the UV-A and UV-B (0.280 to 0.400 m)
ranges are recommended for PAMS applications. The same siting criteria used for "all wave"
global solar radiation measurements apply.
6.4.8 Barometric Pressure
Barometric pressure (station pressure) is used in the calculation of fundamental
thermodynamic quantities (e.g., air density). The type of sensor used to measure pressure is
called a pressure transducer. There are numerous commercially available pressure transducers
which meet the specifications in Table 6-3. Ideally, the pressure sensor should be located in a
ventilated shelter about 2 m above ground level. The height of the station above mean sea level
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 13 of 24
and the height of the pressure sensor above ground level should be documented. If needed, the
pressure can then be adjusted to a standard height.
If the pressure sensor is placed indoors, accommodations should be made to vent the
pressure port to the outside environment. One end of a tube should be attached to the sensor's
pressure port and the other end vented to the outside of the trailer or shelter so that pressurization
due to the air conditioning or heating system is avoided. The wind can often cause dynamical
changes of pressure in a room where a sensor is placed. These fluctuations may be on the order
of 2 to 3 hPa when strong or gusty winds prevail.
6.4.9 Precipitation
Precipitation should be measured with a recording rain gauge such as a tipping bucket or
weighing bucket. The rain gauge should be located on level ground in an open area.
Obstructions should not be closer than two to four times their height from the instrument. The
area around the rain gauge should be covered with natural vegetation. The mouth of the rain
gauge should be level and should be as low as possible while still precluding in-splashing from
the ground (30 cm above ground level is the recommended minimum height). A wind
shield/wind screen (such as an Alter-type wind shield,3 consisting of a ring with approximately
32 free-swinging separate metal leaves) should be employed to minimize the effects of high wind
speeds.
6.5 Upper-Air Meteorological Monitoring
The design of the upper-air monitoring program will depend upon region specific factors
such that the optimal design for a given PAMS region is expected to be some combination of
remote sensing and conventional atmospheric soundings - in special cases, the upper-air
monitoring plan may be augmented with data from aircraft and/or tall towers. Data from existing
sources, e.g., the National Weather Service (NWS) upper-air network, should be considered and
integrated with the PAMS monitoring plan.
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 14 of 24
Remote sensing systems (e.g., doppler SODAR) provide continuous measurements of
wind speed and wind direction as a function of height. These data are needed to provide wind
data with the necessary temporal and vertical resolution to evaluate changes in transport flow
fields coincident with the evolution of the convective boundary layer. Such evaluations will aid
in the diagnosis of conditions associated with extreme ozone concentrations. Remote sensing
platforms for use in obtaining these data are discussed in Section 6.3.5.
Conventional atmospheric soundings obtained using rawinsondes or their equivalent are
needed to provide atmospheric profiles with the necessary vertical resolution for estimating the
mixing height (see Section 6.3.6) and for use in initializing the photochemical grid models used
for evaluating ozone control strategies. Such soundings should extend to the top of the CBL or
1000 meters, whichever is greater, and should include measurements of wind speed, wind
direction, temperature, and humidity. Four soundings per day are needed to adequately
characterize the development of the atmospheric boundary layer. These soundings should be
acquired just prior to sunrise when the atmospheric boundary layer is usually the most stable; in
mid-morning when the growth of the boundary layer is most rapid; during mid-afternoon when
surface temperatures are maximum; and in late-afternoon when the boundary layer depth is
largest. Soundings obtained from a NWS upper-air station may be used to fulfill part of this
requirement depending on the time of the sounding and the location of the NWS site.
Rawinsondes for use in obtaining these data are discussed in Section 3.4.
The information presented in Sections 6.3.2 (Aircraft), 6.3.3 (Tall Towers), 6.3.4
(Balloon Systems), and 6.3.5 (Ground-Based Remote Sensors) provides background for use in
designing an upper air monitoring plan for PAMS. The capabilities of the various platforms for
upper-air meteorological monitoring (towers, balloon systems, and remote sensors) are compared
in Table 6-6.
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 15 of 24
Table 6-6. Capabilities and Limitations of Meteorological Measurement Systems
for Vertical Profiling of the Lower Atmosphere
Typical Maximum Height/Range (meters agl)a
Variable
Wind Speed
Wind Direction
Wind Sigmas0
Relative Humidity
Temperature
Measurement System
Tower
100b
100b
100b
100b
100b
SODAR
600
600
600
d
d
Mini-
SODAR
300
300
300
d
d
RADAR
2-3 km
2-3 km
2-3 km
d
d
RADAR
with
RASS
2-3 km
2-3 km
2-3 km
d
1.2km
Radio-
sonde
>10km
>10km
d
>10km
>10km
Tether-
sonde
1000
1000
d
1000
1000
Typical Minimum Height (meters agl)a
Variable
Wind Speed
Wind Direction
Wind Sigmas0
Relative Humidity
Temperature
Measurement System
Tower
10
10
10
2
2
SODAR
50
50
50
d
d
Mini-
SODAR
10
10
10
d
d
RADAR
100
100
100
d
d
RADAR
with
RASS
100
100
100
d
100
Radio-
sonde
10
10
d
10
10
Tether-
sonde
10
10
d
10
10
Typical Resolution (meters)
Variable
Wind Speed
Wind Direction
Wind Sigmas0
Relative Humidity
Temperature
Measurement System
Tower
2-10
2-10
2-10
2-10
2-10
SODAR
25
25
25
d
d
Mini-
SODAR
10
10
10
d
d
RADAR
60-100
60-100
60-100
d
d
RADAR
with
RASS
60-100
60-100
60-100
d
60-100
Radio-
sonde
5-10
5-10
d
5-10
5-10
Tether-
sonde
10
10
d
10
10
a Meters above ground level
b Typically meteorological towers do not exceed 100 m. However, radio and TV towers may exceed 600 m.
0 The standard deviation of horizontal and vertical wind components.
d No capability for this variable
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 16 of 24
6.5.1 Siting and Exposure
The upper-air measurements are intended for more macro-scale application3 than are the
surface meteorological measurements. Consequently, the location of the upper-air site need not
be associated with any particular PAMS surface site. Factors that should be considered in
selecting a site for the upper-air monitoring include whether the upper-air measurements for the
proposed location are likely to provide the necessary data to characterize the meteorological
conditions associated with high ozone concentrations, and the extent to which data for the
proposed location may augment an existing upper-air network. Near lake shores and in coastal
areas, where land/sea/lake breeze circulations may play a significant role in ozone formation and
transport, additional upper-air monitoring sites may be needed; this consideration would also
apply to areas located in complex terrain. All of the above are necessary components of the
DQOs for an upper-air monitoring plan.
6.5.2 Aircraft
Aircraft (both airplanes and helicopters) are a prime example of a mobile observation
station. They are capable of traversing large horizontal and vertical distances in a relatively short
period of time. An aircraft platform equipped with meteorological instrumentation can provide
detailed atmospheric observations over large areas. Traditionally, aircraft are used for episodic
field studies which often require extensive data sets for model evaluation. Lenschow12 provides
an excellent overview of aircraft measurements in boundary layer applications. While an aircraft
can provide detailed atmospheric observations over large areas, the total sampling time per flight
(typically 6 to 8 hours) is relatively short because of fuel considerations. Aircraft may also be
subject to Federal Aviation Administration (FAA) restrictions on flight paths over urban areas.
In addition, the operating cost for this type of platform is extremely expensive.
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 17 of 24
6.5.3 Tall Towers
In some instances it may be possible to use existing towers which may be located in
PAMS areas to acquire vertical profiles of atmospheric boundary layer data. Radio and
television transmission towers, which may be as tall as 600 m, can be equipped with in situ
meteorological sensors at many levels. An advantage to using a tower is the ability to run an
unattended data acquisition system. Also, data can normally be collected under all weather
conditions. However, the main disadvantage of using a tower is the inability to determine the
mixed layer height during most of the day. When moderate to strong convective conditions exist,
the mixed layer height easily exceeds that of the tallest towers. Another disadvantage is the
potentially high cost of maintenance, especially during instances when the instrumentation needs
to be accessed for adjustments or repairs.
6.5.4 Balloon Systems
Balloon based systems include rawinsonde (sometimes called radiosonde) and
tethersonde systems. The rawinsonde consists of a helium filled balloon, an instrument package,
a radio transmitter, and a tracking device. The instrument package includes sensors for
measuring atmospheric temperature, relative humidity, and barometric pressure. Data from
ground-based radar, which is used to track the balloon, are processed to determine wind speed
and direction. The height of the instrument package is determined by the ascent rate of the
balloon. Typical specifications for the sensors used in rawinsondes are given in Table 6-7.
Table 6-7. Manufacturers' Specifications for Sensors Used in Rawinsondes
Sensor
Pressure
Temperature
Relative Humidity
Range
1080 to 3 mb
-90 to +60 °C
5 to 100%RH
Accuracy
±0.5 mb
±0.2 °C
Resolution
O.lmb
0.1 °C
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 18 of 24
Unlike surface measurements, there is no equivalent to system accuracy for upper-air
meteorological measurements from rawinsondes. Consequently, to assess the quality of
rawinsonde measurements, the NWS uses a special statistical parameter called the "functional
precision," defined as the root-mean-square (rms) difference between measurements made by
identical instruments at as nearly as possible the same time and same point in the atmosphere.13
The functional precision of NWS radiosonde measurements is given in Table 6-8.
Table 6-8. Functional Precision of Rawinsonde Measurements
13
Variable
Wind Speed3
Wind Direction"
Temperature13
Dew Point Depression13
Height13
Functional Precision
±3.1 m/s
±18deg[ < 3.1 m/s]
±14deg[5.1 m/s]
± 9 deg [10.3 m/s]
±6deg[15.4m/s]
± 5 deg [20.6 m/s]
±0.6°C
±3.3°C
±24 m
a at the same height
b at the same pressure
A tethersonde system is comprised of a tethered balloon with one or more instrument
packages attached to the tether. The instrument package includes a radio transmitter and sensors
to measure atmospheric temperature, relative humidity, barometric pressure, wind speed, and
wind direction. Data are telemetered to the ground by radio or by conductors incorporated within
the tethering cable. Tethersondes are capable of providing data up to about 1000 m in good
conditions. Use of a tethersonde is limited by wind speed; they can only be used reliably in light
to moderate wind conditions (5 m/s at the surface to 15 m/s aloft). Tethered balloons are also
considered a hazard to aviation and thus are subject to FAA regulations. A permit is required to
operate such a system.
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 19 of 24
6.5.5 Ground-Based Remote Sensors
Ground-based remote sensors have become effective tools for acquiring upper-air
information and have played an increasingly important role in atmospheric boundary layer
studies. There are two basic types of remote sensing systems used to acquire three-component
wind velocity profiles: Radar (radio detection and ranging) and sodar (sound detection and
ranging). Radars (also called wind profilers) transmit an electromagnetic signal (-915 MHZ)
into the atmosphere in a predetermined beam width which is controlled by the configuration of
the transmitting antenna. Sodars (also called acoustic sounders) transmit an acoustic signal
(~ 2 to 5 KHz) into the atmosphere in a predetermined beam width which is also controlled by
the transmitting antenna. The radar has a range of approximately 150 to 3000 m with a
resolution of 60 to 100 m. The sodar has a range of about 50 to 1500 m with a resolution of
about 25 to 50 m.
Both systems transmit their respective signals in pulses. Each pulse is both reflected and
absorbed by the atmosphere as it propagates upwards. The vertical range of each pulse is
determined by how high it can go before the signal becomes so weak that the energy reflected
back to the antenna can no longer be detected. As long as the reflected pulses can be discerned
from background noise, meaningful wind velocities can be obtained by comparing the doppler
shift of the return signal to that of the output signal. A positive or negative doppler shift
indicates whether the radial wind velocity is moving towards or away from the transmitting
antenna. The attenuation of a transmitted pulse is a function of signal type, signal power, signal
frequency, and atmospheric conditions. Radar signal reflection depends primarily on the
presence of an index of refraction gradient in the atmosphere which varies with temperature and
humidity. Sodar signal reflection depends primarily on the presence of small scale atmospheric
turbulence. The reflected signals received by either a radar or sodar are processed in a system
computer by signal conditioning algorithms.
In order to obtain a profile of the three-component wind velocity (U, V, W), one vertical
beam and two tilted beams are needed. The two tilted beams are usually between 15° and 30°
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 20 of 24
from the vertical. These two beams are also at right angles to each other in azimuth. For
example, one tilted beam may be oriented towards the north while the second tilted beam points
east. Each antenna transmits a pulse and then listens for the reflected signal in succession. After
all three antennas perform this function, enough information is available to convert the radial
velocities into horizontal and vertical wind velocities by using simple trigonometric
relationships.
There are two types of antenna configurations for radars and sodars: Monostatic and
phased array. Monostatic systems consist of three individual transmit/receive antennas. Phased
array consist of a single antenna array which can electronically steer the beam in the required
directions. Vertical panels (also known as clutter fences) are usually placed around the antennas.
This placement effectively acts to block out any stray side-lobe echoes from contaminating the
return signal of a radar. For sodars, these panels cut down on the side-lobe noise which may be a
nuisance to nearby residents and also prevents any background noise which may contaminate the
return signal.
A RASS (radio acoustic sounding system) utilizes a combination of electromagnetic and
acoustic pulses to derive a virtual air temperature profile. A RASS usually consists of several
acoustic antennas placed around a radar system. The antennas transmit a sweep of acoustic
frequencies vertically into the atmosphere. As the sound pulses rise, the speed of the acoustic
wave varies according to the virtual air temperature. Concurrently, a radar beam is emitted
vertically into the atmosphere. The radar beam will most strongly reflect off the sound wave
fronts created by the acoustic pulses. The virtual air temperature is computed from the speed of
sound which is measured by the reflected radar energy. The typical range of a RASS is
approximately 150 to 1500 m with a resolution of 60 to 100 m.
Unlike in situ sensors which measure by direct contact, remote sensors do not disturb the
atmosphere. Another fundamental difference is that remote sensors measure a volume of air
rather than a fixed point in space. The thickness of the volume is a function of the pulse length
and frequency used. The width of the volume is a function of beam spread and altitude. Siting
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 21 of 24
of these profilers is sometimes a difficult task. Artificial and natural objects located near the
sensors can potentially interfere with the transmission and return signals, thereby contaminating
the wind velocity data.
Since sodars utilize sound transmission and reception to determine the overlying wind
field, a clear return signal with a sharply defined atmospheric peak frequency is required. Thus,
consideration of background noise may put limitations on where a sodar can be located. External
noise sources can be classified as active or passive, and as broad-band (random frequency) or
narrow-band (fixed frequency). General background noise is considered active and is broad-
band. If loud enough, it can cause the sodar software to reject data because it can not find a peak
or because the signal-to-noise ratio is too low. The net effect is to lower the effective sampling
rate due to the loss of many transmission pulses. A qualitative survey should be conducted to
identify any potential noise sources. A quantitative noise survey may be necessary to determine
if noise levels are within the instrument's minimum requirements.
Examples of active, broad-band noise sources include highways, industrial facilities,
power plants, and heavy machinery. Some of these noise sources have a pronounced diurnal,
weekly, or even seasonal pattern. A noise survey should at least cover diurnal and weekly
patterns. Examination of land-use patterns and other sources of information may be necessary to
determine if any seasonal activities may present problems.
Examples of active, fixed-frequency noise sources include rotating fans, a back-up beeper
on a piece of heavy equipment, birds, and insects. If these noise sources have a frequency
component in the sodar operating range, they may be misinterpreted as good data by the sodar.
Some of these sources can be identified during the site selection process. One approach to
reducing the problem of fixed frequency noise sources is to use a coded pulse, i.e., the transmit
pulse has more than one peak frequency. A return pulse would not be identified as data unless
peak frequencies were found in the return signal the same distance apart as the transmit
frequencies.
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 22 of 24
Passive noise sources are objects either on or above the ground (e.g., tall towers, power
transmission lines, buildings, trees) that can reflect a transmitted pulse back to the sodar antenna.
While most of the acoustic energy is focused in a narrow beam, side-lobes do exist and are a
particular concern when antenna enclosures have degraded substantially. Side-lobes reflecting
off stationary objects and returning at the same frequency as the transmit pulse may be
interpreted by the sodar as a valid atmospheric return with a speed of zero. It is not possible to
predict precisely which objects may be a problem. Anything in the same general direction in
which the antenna is pointing and higher than 5 to 10m may be a potential reflector. It is
therefore important to construct an "obstacle vista diagram" prior to sodar installation that
identifies the direction and height of potential reflectors in relation to the sodar. This diagram
can be used after some data have been collected to assess whether or not reflections are of
concern at some sodar height ranges. Note that reflections from an object at distance Xfrom an
antenna will show up at height Xcos(a), where a is the tilt angle of the antenna from the vertical.
The radar, sodar, and RASS antennas should be aligned and tilted carefully as small
errors in orientation or tilt angle can produce unwanted biases in the data. True North should
also be established for antenna alignment. Installation of the antennas should not be permanent
since problems are very likely to arise in siting the profilers in relation to the tower and other
objects that may be in the area. One final consideration is the effect of the instrument on its
surroundings. The sound pulse from a sodar and RASS is quite audible and could become a
nuisance to residents who might happen to live near the installation site. This audible pulse
should be a consideration in the siting process because of the potential irritation to nearby
residents.
6.5.6 Estimation of Mixing Height
In addition to the directly measured meteorological variables, estimates are also required
of the depth of the mixed layer (i.e., mixing height). The mixing height is a derived variable
indicating the depth through which vertical mixing of pollutants occurs. Reliable estimates of
the mixing height are essential to dispersion modeling in support of PAMS.
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 23 of 24
The EPA recommended method for estimating mixing height requires measurements of
the vertical temperature profile.14'15 In this method, the afternoon mixing height is calculated as
the height above the ground of the intersection of the dry adiabatic extension of the maximum
surface temperature with the 12 z morning temperature profile. This concept of a mixing layer in
which the lapse rate is roughly dry adiabatic is well founded on general theoretical principles and
on operational use in regulatory dispersion modeling over the last two decades. Comparisons of
mixing height estimates based on the Holzworth method with several other techniques indicate
that all methods perform similarly in estimating the maximum afternoon mixing depth.16'17 The
Holzworth method is normally preferred because of its simplicity.
6.6 References
1. U.S. Environmental Protection Agency. Code of Federal Regulations. Title 40, Part 58,
Ambient Air Quality Surveillance, Final Rule Federal Register, Vol. 58, No. 28,
February 12, 1993.
2. Recommendations for Meteorological Monitoring Requirements in Support ofPAMS,
Internal Memorandum, Desmond Bailey, Emissions Monitoring and Analysis Division,
Research Triangle Park, NC: U.S. Environmental Protection Agency, April 1994.
3. Quality Assurance Handbook for Air Pollution Measurement Systems. Volume IV -
Meteorological Measurement, EPA-600/R-94/038d. U. S. Environmental Protection
Agency, 1994.
4. On-site Meteorological Program Guidance for Regulatory Modeling Applications, EP A-
450/4-87-013. Research Triangle Park, NC: U. S. Environmental Protection Agency,
1987b.
5. Draft Guidelines for Quality Assurance and Management ofPAMS Upper-Air
Meteorological Data. Sonoma Technology, Inc., under EPA Contract No. 68-D3-0020.
December, 1995.
6. On-Site Meteorological Instrumentation Requirements to Characterize Diffusion from
Point Sources, EPA-600/9-81-020. Research Triangle Park, NC: U.S. Environmental
Protection Agency, 1981.
7. Guide to Meteorological Instruments and Methods of Observation (Fifth Edition), WMO
No. 8 Geneva, Switzerland: World Meteorological Organization, 1983.
-------
Section: 6
Revision: 1
Date: 09/30/98
Page: 24 of 24
8. Instructor's Handbook on Meteorological Instrumentation, NC AR/TN-23 7+IA. Boulder,
Colorado: National Center for Atmospheric Research, 1985.
9. Standard Practice for Characterizing Surface Wind Using a Wind Vane and Rotating
Cup Anemometer, ASTM Designation: D 5741-96, Annual Book of ASTM Standards,
Section 11, 1996.
10. Evans, R.A., and B.E. Lee. The Problems of Anemometer Exposure in Urban Areas - A
Wind-Tunnel Study. Meteorological Magazine, 110, 188-199(1981).
11. Katsaros, K. B., and DeVault, J. E. On Irradiance Measurement Errors at Sea Due to Tilt
of Pyranometers. J. Atmos. Oceanic Technol. 3, 740-745 (1986).
12. Lenschow, D. H. Aircraft Measurements in the Boundary Layer. Probing the
Atmospheric Boundary Layer, American Meteorological Society, Boston, pp. 39-55
(1986).
13. Hoehne, W. E. Precision of National Weather Service Upper Air Measurements, NOAA
Technical Memorandum NWS T&ED-16, U.S. Department of Commerce, Sterling, VA.
(1980).
14. Holzworth, G. C. Estimates of Mean Maximum Mixing Depths in the Contiguous United
States. Monthly Weather Review, 92, 235-242 (1964).
15. Holzworth, G. C. Mixing Heights, Wind Speeds, and Potential for Urban Air Pollution
Throughout the Contiguous United States, Publication No. AP-101. Research Triangle
Park, NC: U. S. Environmental Protection Agency, 1972.
16. Hanna, S. R. The Thickness of the Planetary Boundary Layer. Atmos. Environ. 3, 519-
536(1969).
17. Irwin, J. S., and J. O. Paumier. Characterizing the Dispersive State of Convective
Boundary Layers for Applied Dispersion Modeling. Boundary-Layer Meteorology, 53,
267-296 (1990).
-------
APPENDIX A
Method TO-15
Determination of Volatile Organic Compounds (VOCs) in Air Collected-Prepared
Canisters and Analyzed by Gas Chromatography/Mass Spectrometry (GC/MS)
-------
EPA/625/R-96/010b
Compendium of Methods
for the Determination of
Toxic Organic Compounds
in Ambient Air
Second Edition
Compendium Method TO-15
Determination Of Volatile Organic
Compounds (VOCs) In Air Collected In
Specially-Prepared Canisters And
Analyzed By Gas Chromatography/
Mass Spectrometry (GC/MS)
Center for Environmental Research Information
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
January 1997
-------
Method TO-15
Acknowledgements
This Method was prepared for publication in the Compendium of Methods for the Determination of Toxic
Organic Compounds in Ambient Air, Second Edition (EPA/625/R-96/010b), which was prepared under
Contract No. 68-C3-0315, WA No. 3-10, by Midwest Research Institute (MRI), as a subcontractor to
Eastern Research Group, Inc. (ERG), and under the sponsorship of the U.S. Environmental Protection
Agency (EPA). Justice A. Manning and John O. Burckle, Center for Environmental Research
Information (CERI), and Frank F. McElroy, National Exposure Research Laboratory (NERL), both in the
EPA Office of Research and Development, were the project officers responsible for overseeing the
preparation of this method. Additional support was provided by other members of the Compendia
Workgroup, which include:
John O. Burckle, EPA, ORD, Cincinnati, OH
James L. Cheney, Corps of Engineers, Omaha, NB
• Michael Davis, U.S. EPA, Region 7, KC, KS
Joseph B. Elkins Jr., U.S. EPA, OAQPS, RTP, NC
Robert G. Lewis, U.S. EPA, NERL, RTP, NC
Justice A. Manning, U.S. EPA, ORD, Cincinnati, OH
• William A. McClenny, U.S. EPA, NERL, RTP, NC
Frank F. McElroy, U.S. EPA, NERL, RTP, NC
Heidi Schultz, ERG, Lexington, MA
• William T. "Jerry" Winberry, Jr., MRI, Gary, NC
This Method is the result of the efforts of many individuals. Gratitude goes to each person involved in
the preparation and review of this methodology.
Author(s)
• William A. McClenny, U.S. EPA, NERL, RTP, NC
Michael W. Holdren, Battelle, Columbus, OH
Peer Reviewers
Karen Oliver, ManTech, RTP, NC
• Jim Cheney, Corps of Engineers, Omaha, NB
Elizabeth Almasi, Varian Chromatography Systems, Walnut Creek, CA
Norm Kirshen, Varian Chromatography Systems, Walnut Creek, CA
Richard Jesser, Graseby, Smyrna, GA
Bill Taylor, Graseby, Smyrna, GA
Finally, recognition is given to Frances Beyer, Lynn Kaufman, Debbie Bond, Cathy Whitaker, and Kathy
Johnson of Midwest Research Institute's Administrative Services staff whose dedication and persistence
during the development of this manuscript has enabled it's production.
DISCLAIMER
This Compendium has been subjected to the Agency's peer and administrative review, and it has been
approved for publication as an EPA document. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
-------
METHOD TO-15
Determination of Volatile Organic Compounds (VOCs) In Air Collected In
Specially-Prepared Canisters And Analyzed By Gas Chromatography/
Mass Spectrometry (GC/MS)
TABLE OF CONTENTS
Page
1. Scope 15-1
2. Summary of Method 15-2
3. Significance 15-3
4. Applicable Documents 15-4
4.1 ASTM Standards 15-4
4.2 EPA Documents 15-4
5. Definitions 15-5
6. Interferences and Contamination 15-6
7. Apparatus and Reagents 15-6
7.1 Sampling Apparatus 15-7
7.2 Analytical Apparatus 15-8
7.3 Calibration System and Manifold Apparatus 15-10
7.4 Reagents 15-10
8. Collection of Samples in Canisters 15-11
8.1 Introduction 15-11
8.2 Sampling System Description 15-11
8.3 Sampling Procedure 15-13
8.4 Cleaning and Certification Program 15-14
9. GC/MS Analysis of Volatiles from Canisters 15-17
9.1 Introduction 15-17
9.2 Preparation of Standards 15-18
10. GC/MS Operating Conditions 15-21
10.1 Preconcentrator 15-21
10.2 GC/MS System 15-22
10.3 Analytical Sequence 15-23
10.4 Instrument Performance Check 15-23
10.5 Initial Calibration 15-24
10.6 Daily Calibration 15-27
10.7 Blank Analyses 15-28
10.8 Sample Analysis 15-29
-------
TABLE OF CONTENTS (continued)
Page
11. Requirements for Demonstrating Method Acceptability for VOC Analysis from
Canisters 15-31
11.1 Introduction 15-31
11.2 Method Detection Limit 15-32
11.3 Replicate Precision 15-32
11.4 Audit Accuracy 15-32
12. References 15-33
-------
METHOD TO-15
Determination of Volatile Organic Compounds (VOCs) In Air Collected In
Specially-Prepared Canisters And Analyzed By Gas Chromatography/
Mass Spectrometry (GC/MS)
1. Scope
1.1 This method documents sampling and analytical procedures for the measurement of subsets of the 97
volatile organic compounds (VOCs) that are included in the 189 hazardous air pollutants (HAPs) listed in
Title III of the Clean Air Act Amendments of 1990. VOCs are defined here as organic compounds having
a vapor pressure greater than 10"1 Torrat25°Cand760mmHg. Table 1 is the list of the target VOCs along
with their CAS number, boiling point, vapor pressure and an indication of their membership in both the list
of VOCs covered by Compendium Method TO-14A (1) and the list of VOCs in EPA's Contract Laboratory
Program (CLP) document entitled: Statement-of-Work (SOW) for the Analysis of Air Toxics from Superfund
Sites (2).
Many of these compounds have been tested for stability in concentration when stored in specially-prepared
canisters (see Section 8) under conditions typical of those encountered in routine ambient air analysis. The
stability of these compounds under all possible conditions is not known. However, a model to predict
compound losses due to physical adsorption of VOCs on canister walls and to dissolution of VOCs in water
condensed in the canisters has been developed (3). Losses due to physical adsorption require only the
establishment of equilibrium between the condensed and gas phases and are generally considered short term
losses, (i.e., losses occurring over minutes to hours). Losses due to chemical reactions of the VOCs with
cocollected ozone or other gas phase species also account for some short term losses. Chemical reactions
between VOCs and substances inside the canister are generally assumed to cause the gradual decrease of
concentration over time (i.e., long term losses over days to weeks). Loss mechanisms such as aqueous
hydrolysis and biological degradation (4) also exist. No models are currently known to be available to
estimate and characterize all these potential losses, although a number of experimental observations are
referenced in Section 8. Some of the VOCs listed in Title III have short atmospheric lifetimes and may not
be present except near sources.
1.2 This method applies to ambient concentrations of VOCs above 0.5 ppbv and typically requires VOC
enrichment by concentrating up to one liter of a sample volume. The VOC concentration range for ambient
air in many cases includes the concentration at which continuous exposure over a lifetime is estimated to
constitute a 10"6 or higher lifetime risk of developing cancer in humans. Under circumstances in which many
hazardous VOCs are present at 10"6 risk concentrations, the total risk may be significantly greater.
1.3 This method applies under most conditions encountered in sampling of ambient air into canisters.
However, the composition of a gas mixture in a canister, under unique or unusual conditions, will change
so that the sample is known not to be a true representation of the ambient air from which it was taken. For
example, low humidity conditions in the sample may lead to losses of certain VOCs on the canister walls,
losses that would not happen if the humidity were higher. If the canister is pressurized, then condensation
of water from high humidity samples may cause fractional losses of water-soluble compounds. Since the
canister surface area is limited, all gases are in competition for the available active sites. Hence an absolute
storage stability cannot be assigned to a specific gas. Fortunately, under conditions of normal usage for
sampling ambient air, most VOCs can be recovered from canisters near their original concentrations after
storage times of up to thirty days (see Section 8).
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-1
-------
Method TO-15 VOCs
1.4 Use of the Compendium Method TO-15 for many of the VOCs listed in Table 1 is likely to present two
difficulties: (1) what calibration standard to use for establishing a basis for testing and quantitation, and (2)
how to obtain an audit standard. In certain cases a chemical similarity exists between a thoroughly tested
compound and others on the Title III list. In this case, what works for one is likely to work for the other in
terms of making standards. However, this is not always the case and some compound standards will be
troublesome. The reader is referred to the Section 9.2 on standards for guidance. Calibration of compounds
such as formaldehyde, diazomethane, and many of the others represents a challenge.
1.5 Compendium Method TO-15 should be considered for use when a subset of the 97 Title III VOCs
constitute the target list. Typical situations involve ambient air testing associated with the permitting
procedures for emission sources. In this case sampling and analysis of VOCs is performed to determine the
impact of dispersing source emissions in the surrounding areas. Other important applications are prevalence
and trend monitoring for hazardous VOCs in urban areas and risk assessments downwind of industrialized
or source-impacted areas.
1.6 Solid adsorbents can be used in lieu of canisters for sampling of VOCs, provided the solid adsorbent
packings, usually multisorbent packings in metal or glass tubes, can meet the performance criteria specified
in Compendium Method TO-17 which specifically addresses the use of multisorbent packings. The two
sample collection techniques are different but become the same upon movement of the sample from the
collection medium (canister or multisorbent tubes) onto the sample concentrator. Sample collection directly
from the atmosphere by automated gas chromatographs can be used in lieu of collection in canisters or on
solid adsorbents.
2. Summary of Method
2.1 The atmosphere is sampled by introduction of air into a specially-prepared stainless steel canister. Both
subatmospheric pressure and pressurized sampling modes use an initially evacuated canister. A pump
ventilated sampling line is used during sample collection with most commercially available samplers.
Pressurized sampling requires an additional pump to provide positive pressure to the sample canister. A
sample of air is drawn through a sampling train comprised of components that regulate the rate and duration
of sampling into the pre-evacuated and passivated canister.
2.2 After the air sample is collected, the canister valve is closed, an identification tag is attached to the
canister, and the canister is transported to the laboratory for analysis.
2.3 Upon receipt at the laboratory, the canister tag data is recorded and the canister is stored until analysis.
Storage times of up to thirty days have been demonstrated for many of the VOCs (5).
2.4 To analyze the sample, a known volume of sample is directed from the canister through a solid
multisorbent concentrator. A portion of the water vapor in the sample breaks through the concentrator during
sampling, to a degree depending on the multisorbent composition, duration of sampling, and other factors.
Water content of the sample can be further reduced by dry purging the concentrator with helium while
retaining target compounds. After the concentration and drying steps are completed, the VOCs are thermally
desorbed, entrained in a carrier gas stream, and then focused in a small volume by trapping on a reduced
Page 15-2 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs Method TO-15
temperature trap or small volume multisorbent trap. The sample is then released by thermal desorption and
carried onto a gas chromatographic column for separation.
As a simple alternative to the multisorbent/dry purge water management technique, the amount of water
vapor in the sample can be reduced below any threshold for affecting the proper operation of the analytical
system by reducing the sample size. For example, a small sample can be concentrated on a cold trap and
released directly to the gas chromatographic column. The reduction in sample volume may require an
enhancement of detector sensitivity.
Other water management approaches are also acceptable as long as their use does not compromise the
attainment of the performance criteria listed in Section 11. A listing of some commercial water management
systems is provided in Appendix A. One of the alternative ways to dry the sample is to separate VOCs from
condensate on a low temperature trap by heating and purging the trap.
2.5 The analytical strategy for Compendium Method TO-15 involves using a high resolution gas
chromatograph (GC) coupled to a mass spectrometer. If the mass spectrometer is a linear quadrupole system,
it is operated either by continuously scanning a wide range of mass to charge ratios (SCAN mode) or by
monitoring select ion monitoring mode (SIM) of compounds on the target list. If the mass spectrometer is
based on a standard ion trap design, only a scanning mode is used (note however, that the Selected Ion
Storage (SIS) mode for the ion trap has features of the SIM mode). Mass spectra for individual peaks in the
total ion chromatogram are examined with respect to the fragmentation pattern of ions corresponding to
various VOCs including the intensity of primary and secondary ions. The fragmentation pattern is compared
with stored spectra taken under similar conditions, in order to identify the compound. For any given
compound, the intensity of the primary fragment is compared with the system response to the primary
fragment for known amounts of the compound. This establishes the compound concentration that exists in
the sample.
Mass spectrometry is considered a more definitive identification technique than single specific detectors such
as flame ionization detector (FID), electron capture detector (BCD), photoionization detector (PID), or a
multidetector arrangement of these (see discussion in Compendium Method TO-14A). The use of both gas
chromatographic retention time and the generally unique mass fragmentation patterns reduce the chances for
misidentification. If the technique is supported by a comprehensive mass spectral database and a
knowledgeable operator, then the correct identification and quantification of VOCs is further enhanced.
3. Significance
3.1 Compendium Method TO-15 is significant in that it extends the Compendium Method TO-14A
description for using canister-based sampling and gas chromatographic analysis in the following ways:
• Compendium Method TO-15 incorporates a multisorbent/dry purge technique or equivalent (see
Appendix A) for water management thereby addressing a more extensive set of compounds (the VOCs
mentioned in Title III of the CAAA of 1990) than addressed by Compendium Method TO-14A.
Compendium Method TO-14A approach to water management alters the structure or reduces the
sample stream concentration of some VOCs, especially water-soluble VOCs.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-3
-------
Method TO-15 VOCs
• Compendium Method TO-15 uses the GC/MS technique as the only means to identify and quantitate
target compounds. The GC/MS approach provides a more scientifically-defensible detection scheme
which is generally more desirable than the use of single or even multiple specific detectors.
• In addition, Compendium Method TO-15 establishes method performance criteria for acceptance of
data, allowing the use of alternate but equivalent sampling and analytical equipment. There are several
new and viable commercial approaches for water management as noted in Appendix A of this method
on which to base a VOC monitoring technique as well as other approaches to sampling (i.e., autoGCs
and solid adsorbents) that are often used. This method lists performance criteria that these alternatives
must meet to be acceptable alternatives for monitoring ambient VOCs.
• Finally, Compendium Method TO-15 includes enhanced provisions for inherent quality control. The
method uses internal analytical standards and frequent verification of analytical system performance
to assure control of the analytical system. This more formal and better documented approach to quality
control guarantees a higher percentage of good data.
3.2 With these features, Compendium Method TO-15 is a more general yet better defined method for VOCs
than Compendium Method TO-14A. As such, the method can be applied with a higher confidence to reduce
the uncertainty in risk assessments in environments where the hazardous volatile gases listed in the Title III
of the Clean Air Act Amendments of 1990 are being monitored. An emphasis on risk assessments for human
health and effects on the ecology is a current goal for the U.S. EPA.
4. Applicable Documents
4.1 ASTM Standards
• Method D1356 Definitions of Terms Relating to Atmospheric Sampling and Analysis.
• Method E260 Recommended Practice for General Gas Chromatography Procedures.
• Method E355 Practice for Gas Chromatography Terms and Relationships.
• Method D5466 Standard Test Method of Determination of Volatile Organic Compounds in
Atmospheres (Canister Sampling Methodology).
4.2 EPA Documents
• Quality Assurance Handbook for Air Pollution Measurement Systems, Volume II, U. S. Environmental
Protection Agency, EPA-600/R-94-038b, May 1994.
• Technical Assistance Document for Sampling and Analysis of Toxic Organic Compounds in Ambient
Air, U. S. Environmental Protection Agency, EPA-600/4-83-027, June 1983.
• Compendium of Methods for the Determination of Toxic Organic Compounds in Ambient Air: Method
TO-14, SecondSupplement, U. S. Environmental Protection Agency, EPA-600/4-89-018, March 1989.
• Statement-of-Work (SOW) for the Analysis of Air Toxics from Superfund Sites, U. S. Environmental
Protection Agency, Office of Solid Waste, Washington, B.C., Draft Report, June 1990.
• Clean Air Act Amendments of 1990, U. S. Congress, Washington, D.C., November 1990.
Page 15-4 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs Method TO-15
5. Definitions
[Note: Definitions used in this document and any user-prepared standard operating procedures (SOPs)
should be consistent with ASTMMethods D1356, E260, andE355. Aside from the definitions given below,
all pertinent abbreviations and symbols are defined within this document at point of use.}
5.1 Gauge Pressure—pressure measured with reference to the surrounding atmospheric pressure, usually
expressed in units of kPa or psi. Zero gauge pressure is equal to atmospheric (barometric) pressure.
5.2 Absolute Pressure—pressure measured with reference to absolute zero pressure, usually expressed in
units of kPa, or psi.
5.3 Cryogen—a refrigerant used to obtain sub-ambient temperatures in the VOC concentrator and/or on
front of the analytical column. Typical cryogens are liquid nitrogen (bp -195.8°C), liquid argon (bp -
185.7°C), and liquid CO2 (bp -79.5°C ).
5.4 Dynamic Calibration—calibration of an analytical system using calibration gas standard concentrations
in a form identical or very similar to the samples to be analyzed and by introducing such standards into the
inlet of the sampling or analytical system from a manifold through which the gas standards are flowing.
5.5 Dynamic Dilution—means of preparing calibration mixtures in which standard gas(es) from pressurized
cylinders are continuously blended with humidified zero air in a manifold so that a flowing stream of
calibration mixture is available at the inlet of the analytical system.
5.6 MS-SCAN—mass spectrometric mode of operation in which the gas chromatograph (GC) is coupled
to a mass spectrometer (MS) programmed to SCAN all ions repeatedly over a specified mass range.
5.7 MS-SIM—mass spectrometric mode of operation in which the GC is coupled to a MS that is
programmed to scan a selected number of ions repeatedly [i.e., selected ion monitoring (SIM) mode].
5.8 Qualitative Accuracy—the degree of measurement accuracy required to correctly identify compounds
with an analytical system.
5.9 Quantitative Accuracy—the degree of measurement accuracy required to correctly measure the
concentration of an identified compound with an analytical system with known uncertainty.
5.10 Replicate Precision—precision determined from two canisters filled from the same air mass over the
same time period and determined as the absolute difference between the analyses of canisters divided by their
sum and expressed as a percentage (see Section 11 for performance criteria for replicate precision).
5.11 Duplicate Precision—precision determined from the analysis of two samples taken from the same
canister. The duplicate precision is determined as the absolute difference between the canister analyses
divided by their sum and expressed as a percentage (see Section 11 for performance criteria for duplicate
precision).
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-5
-------
Method TO-15 VOCs
5.12 Audit Accuracy—the difference between the analysis of a sample provided in an audit canister and the
nominal value as determined by the audit authority, divided by the audit value and expressed as a percentage
(see Section 11 for performance criteria for audit accuracy).
6. Interferences and Contamination
6.1 Very volatile compounds, such as chloromethane and vinyl chloride can display peak broadening and
co-elution with other species if the compounds are not delivered to the GC column in a small volume of
carrier gas. Refocusing of the sample after collection on the primary trap, either on a separate focusing trap
or at the head of the gas chromatographic column, mitigates this problem.
6.2 Interferences in canister samples may result from improper use or from contamination of: (1) the
canisters due to poor manufacturing practices, (2) the canister cleaning apparatus, and (3) the sampling or
analytical system. Attention to the following details will help to minimize the possibility of contamination
of canisters.
6.2.1 Canisters should be manufactured using high quality welding and cleaning techniques, and new
canisters should be filled with humidified zero air and then analyzed, after "aging" for 24 hours, to determine
cleanliness. The cleaning apparatus, sampling system, and analytical system should be assembled of clean,
high quality components and each system should be shown to be free of contamination.
6.2.2 Canisters should be stored in a contaminant-free location and should be capped tightly during
shipment to prevent leakage and minimize any compromise of the sample.
6.2.3 Impurities in the calibration dilution gas (if applicable) and carrier gas, organic compounds
out-gassing from the system components ahead of the trap, and solvent vapors in the laboratory account for
the majority of contamination problems. The analytical system must be demonstrated to be free from
contamination under the conditions of the analysis by running humidified zero air blanks. The use of
non-chromatographic grade stainless steel tubing, non-PTFE thread sealants, or flow controllers with Buna-N
rubber components must be avoided.
6.2.4 Significant contamination of the analytical equipment can occur whenever samples containing high
VOC concentrations are analyzed. This in turn can result in carryover contamination in subsequent analyses.
Whenever a high concentration (>25 ppbv of a trace species) sample is encountered, it should be followed
by an analysis of humid zero air to check for carry-over contamination.
6.2.5 In cases when solid sorbents are used to concentrate the sample prior to analysis, the sorbents
should be tested to identify artifact formation (see Compendium Method TO-17 for more information on
artifacts).
7. Apparatus and Reagents
[Note: Compendium Method To-14A list more specific requirements for sampling and analysis apparatus
which may be of help in identifying options. The listings below are generic.]
Page 15-6 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs Method TO-15
7.1 Sampling Apparatus
[Note: Subatmospheric pressure and pressurized canister sampling systems are commercially available and
have been used as part of U.S. Environmental Protection Agency's Toxic Air Monitoring Stations (TAMS),
Urban Air Toxic Monitoring Program (UATMP), the non-methane organic compound (NMOC) sampling
and analysis program, and the Photochemical Assessment Monitoring Stations (PAMS).]
7.1.1 Subatmospheric Pressure (see Figure 1, without metal bellows type pump).
7.1.1.1 Sampling Inlet Line. Stainless steel tubing to connect the sampler to the sample inlet.
7.1.1.2 Sample Canister. Leak-free stainless steel pressure vessels of desired volume (e.g., 6 L), with
valve and specially prepared interior surfaces (see Appendix B for a listing of known manufacturers/resellers
of canisters).
7.1.1.3 Stainless Steel Vacuum/Pressure Gauges. Two types are required, one capable of measuring
vacuum (-100 to 0 kPa or 0 to - 30 in Hg) and pressure (0-206 kPa or 0-30 psig) in the sampling system and
a second type (for checking the vacuum of canisters during cleaning) capable of measuring at 0.05 mm Hg
(see Appendix B) within 20%. Gauges should be tested clean and leak tight.
7.1.1.4 Electronic Mass Flow Controller. Capable of maintaining a constant flow rate (± 10%) over
a sampling period of up to 24 hours and under conditions of changing temperature (20-40°C) and humidity.
7.1.1.5 Particulate Matter Filter. 2-^m sintered stainless steel in-line filter.
7.1.1.6 Electronic Timer. For unattended sample collection.
7.1.1.7 Solenoid Valve. Electrically-operated, bi-stable solenoid valve with Viton® seat and O-rings.
A Skinner Magnelatch valve is used for purposes of illustration in the text (see Figure 2).
7.1.1.8 Chromatographic Grade Stainless Steel Tubing and Fittings. For interconnections. All
such materials in contact with sample, analyte, and support gases prior to analysis should be chromatographic
grade stainless steel or equivalent.
7.1.1.9 Thermostatically Controlled Heater. To maintain above ambient temperature inside insulated
sampler enclosure.
7.1.1.10 Heater Thermostat. Automatically regulates heater temperature.
7.1.1.11 Fan. For cooling sampling system.
7.1.1.12 Fan Thermostat. Automatically regulates fan operation.
7.1.1.13 Maximum-Minimum Thermometer. Records highest and lowest temperatures during
sampling period.
7.1.1.14 Stainless Steel Shut-off Valve. Leak free, for vacuum/pressure gauge.
7.1.1.15 Auxiliary Vacuum Pump. Continuously draws air through the inlet manifold at 10 L/min.
or higher flow rate. Sample is extracted from the manifold at a lower rate, and excess air is exhausted.
[Note: The use of higher inlet flow rates dilutes any contamination present in the inlet and reduces the
possibility of sample contamination as a result of contact with active adsorption sites on inlet walls.]
7.1.1.16 Elapsed Time Meter. Measures duration of sampling.
7.1.1.17 Optional Fixed Orifice, Capillary, or Adjustable Micrometering Valve. May be used in
lieu of the electronic flow controller for grab samples or short duration time-integrated samples. Usually
appropriate only in situations where screening samples are taken to assess future sampling activity.
7.1.2 Pressurized (see Figure 1 with metal bellows type pump and Figure 3).
7.1.2.1 Sample Pump. Stainless steel, metal bellows type, capable of 2 atmospheres output pressure.
Pump must be free of leaks, clean, and uncontaminated by oil or organic compounds.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-7
-------
Method TO-15 VOCs
[Note: An alternative sampling system has been developed by Dr. R. Rasmussen, The Oregon Graduate
Institute of Science and Technology, 20000 N. W. Walker Rd., Beaverton, Oregon 97006, 503-690-1077, and
is illustrated in Figure 3. This flow system uses, in order, a pump, a mechanical flow regulator, and a
mechanical compensation flow restrictive device. In this configuration the pump is purged with a large
sample flow, thereby eliminating the need for an auxiliary vacuum pump to flush the sample inlet.]
7.1.2.2 Other Supporting Materials. All other components of the pressurized sampling system are
similar to components discussed in Sections 7.1.1.1 through 7.1.1.17.
7.2 Analytical Apparatus
7.2.1 Sampling/Concentrator System (many commercial alternatives are available).
7.2.1.1 Electronic Mass Flow Controllers. Used to maintain constant flow (for purge gas, carrier
gas and sample gas) and to provide an analog output to monitor flow anomalies.
7.2.1.2 Vacuum Pump. General purpose laboratory pump, capable of reducing the downstream
pressure of the flow controller to provide the pressure differential necessary to maintain controlled flow rates
of sample air.
7.2.1.3 Stainless Steel Tubing and Stainless Steel Fittings. Coated with fused silicate minimize
active adsorption sites.
7.2.1.4 Stainless Steel Cylinder Pressure Regulators. Standard, two-stage cylinder regulators with
pressure gauges.
7.2.1.5 Gas Purifiers. Used to remove organic impurities and moisture from gas streams.
7.2.1.6 Six-port Gas Chromatographic Valve. For routing sample and carrier gas flows.
7.2.1.7 Multisorbent Concentrator. Solid adsorbent packing with various retentive properties for
adsorbing trace gases are commercially available from several sources. The packing contains more than one
type of adsorbent packed in series.
7.2.1.7.1A pre-packed adsorbent trap (Supelco 2-0321) containing 200 mg Carbopack B (60/80 mesh)
and 50 mg Carbosieve S-III (60/80 mesh) has been found to retain VOCs and allow some water vapor to pass
through (6). The addition of a dry purging step allows for further water removal from the adsorbent trap.
The steps constituting the dry purge technique that are normally used with multisorbent traps are illustrated
in Figure 4. The optimum trapping and dry purging procedure for the Supelco trap consists of a sample
volume of 320 mL and a dry nitrogen purge of 1300 mL. Sample trapping and drying is carried out at 25 °C.
The trap is back-flushed with helium and heated to 220°C to transfer material onto the GC column. A trap
bake-out at 260°C for 5 minutes is conducted after each run.
7.2.1.7.2An example of the effectiveness of dry purging is shown in Figure 5. The multisorbent used
in this case is Tenax/Ambersorb 340/Charcoal (7). Approximately 20% of the initial water content in the
sample remains after sampling 500 mL of air. The detector response to water vapor (hydrogen atoms
detected by atomic emission detection) is plotted versus purge gas volume. Additional water reduction by
a factor of 8 is indicated at temperatures of 45 °C or higher. Still further water reduction is possible using
a two-stage concentration/dryer system.
7.2.1.8 Cryogenic Concentrator. Complete units are commercially available from several vendor
sources. The characteristics of the latest concentrators include a rapid, "ballistic" heating of the concentrator
to release any trapped VOCs into a small carrier gas volume. This facilitates the separation of compounds
on the gas chromatographic column.
7.2.2 Gas Chromatographic/Mass Spectrometric (GC/MS) System.
7.2.2.1 Gas Chromatograph. The gas chromatographic (GC) system must be capable of temperature
programming. The column oven can be cooled to subambient temperature (e .g., -5 0 ° C) at the start of the gas
Page 15-8 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs Method TO-15
chromatographic run to effect a resolution of the very volatile organic compounds. In other designs, the rate
of release of compounds from the focusing trap in a two stage system obviates the need for retrapping of
compounds on the column. The system must include or be interfaced to a concentrator and have all required
accessories including analytical columns and gases. All GC carrier gas lines must be constructed from
stainless steel or copper tubing. Non-polytetrafluoroethylene (PTFE) thread sealants or flow controllers with
Buna-N rubber components must not be used.
7.2.2.2 Chromatographic Columns. 100% methyl silicone or 5%phenyl, 95% methyl silicone fused
silica capillary columns of 0.25- to 0.53-mm I.D. of varying lengths are recommended for separation of many
of the possible subsets of target compounds involving nonpolar compounds. However, considering the
diversity of the target list, the choice is left to the operator subject to the performance standards given in
Section 11.
7.2.2.3 Mass Spectrometer. Either a linear quadrupole or ion trap mass spectrometer can be used as
long as it is capable of scanning from 35 to 300 amu every 1 second or less, utilizing 70 volts (nominal)
electron energy in the electron impact ionization mode, and producing a mass spectrum which meets all the
instrument performance acceptance criteria when 50 ng or less of p-bromofluorobenzene (BFB) is analyzed.
7.2.2.3.1Linear Quadrupole Technology. A simplified diagram of the heart of the quadrupole mass
spectrometer is shown in Figure 6. The quadrupole consists of a parallel set of four rod electrodes mounted
in a square configuration. The field within the analyzer is created by coupling opposite pairs of rods together
and applying radiofrequency (RF) and direct current (DC) potentials between the pairs of rods. Ions created
in the ion source from the reaction of column eluates with electrons from the electron source are moved
through the parallel array of rods under the influence of the generated field. Ions which are successfully
transmitted through the quadrupole are said to possess stable trajectories and are subsequently recorded with
the detection system. When the DC potential is zero, a wide band of m/z values is transmitted through the
quadrupole. This "RF only" mode is referred to as the "total-ion" mode. In this mode, the quadrupole acts
as a strong focusing lens analogous to a high pass filter. The amplitude of the RF determines the low mass
cutoff. A mass spectrum is generated by scanning the DC and RF voltages using a fixed DC/RF ratio and
a constant drive frequency or by scanning the frequency and holding the DC and RF constant. With the
quadrupole system only 0.1 to 0.2 percent of the ions formed in the ion source actually reach the detector.
7.2.2.3.2Ion Trap Technology. An ion-trap mass spectrometer consists of a chamber formed between
two metal surfaces in the shape of ahyperboloid of one sheet (ring electrode) and a hyperboloid of two sheets
(the two end-cap electrodes). Ions are created within the chamber by electron impact from an electron beam
admitted through a small aperture in one of the end caps. Radio frequency (RF) (and sometimes direct
current voltage offsets) are applied between the ring electrode and the two end-cap electrodes establishing
a quadrupole electric field. This field is uncoupled in three directions so that ion motion can be considered
independently in each direction; the force acting upon an ion increases with the displacement of the ion from
the center of the field but the direction of the force depends on the instantaneous voltage applied to the ring
electrode. A restoring force along one coordinate (such as the distance, r, from the ion-trap's axis of radial
symmetry) will exist concurrently with a repelling force along another coordinate (such as the distance, z,
along the ion traps axis), and if the field were static the ions would eventually strike an electrode. However,
in an RF field the force along each coordinate alternates direction so that a stable trajectory may be possible
in which the ions do not strike a surface. In practice, ions of appropriate mass-to-charge ratios may be
trapped within the device for periods of milliseconds to hours. A diagram of a typical ion trap is illustrated
in Figure 7. Analysis of stored ions is performed by increasing the RF voltage, which makes the ions
successively unstable. The effect of the RF voltage on the ring electrode is to "squeeze" the ions in the xy
plane so that they move along the z axis. Half the ions are lost to the top cap (held at ground potential); the
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-9
-------
Method TO-15 VOCs
remaining ions exit the lower end cap to be detected by the electron multiplier. As the energy applied to the
ring electrode is increased, the ions are collected in order of increasing mass to produce a conventional mass
spectrum. With the ion trap, approximately 50 percent of the generated ions are detected. As a result, a
significant increase in sensitivity can be achieved when compared to a full scan linear quadrupole system.
7.2.2.4 GC/MS Interface. Any gas chromatograph to mass spectrometer interface that gives
acceptable calibration points for each of the analytes of interest and can be used to achieve all acceptable
performance criteria may be used. Gas chromatograph to mass spectrometer interfaces constructed of
all-glass, glass-lined, or fused silica-lined materials are recommended. Glass and fused silica should be
deactivated.
7.2.2.5 Data System. The computer system that is interfaced to the mass spectrometer must allow the
continuous acquisition and storage, on machine readable media, of all mass spectra obtained throughout the
duration of the chromatographic program. The computer must have software that allows searching any
GC/MS data file for ions of a specified mass and plotting such ion abundances versus time or scan number.
This type of plot is defined as a Selected Ion Current Profile (SICP). Software must also be available that
allows integrating the abundance in any SICP between specified time or scan number limits. Also, software
must be available that allows for the comparison of sample spectra with reference library spectra. The
National Institute of Standards and Technology (NIST) or Wiley Libraries or equivalent are recommended
as reference libraries.
7.2.2.6 Off-line Data Storage Device. Device must be capable of rapid recording and retrieval of data
and must be suitable for long-term, off-line data storage.
7.3 Calibration System and Manifold Apparatus (see Figure 8)
7.3.1 Calibration Manifold. Stainless steel, glass, or high purity quartz manifold, (e.g.,1.25-cm I.D.
x 66-cm) with sampling ports and internal baffles for flow disturbance to ensure proper mixing. The
manifold should be heated to ~50°C.
7.3.2 Humidifier. 500-mL impinger flask containing HPLC grade deionized water.
7.3.3 Electronic Mass Flow Controllers. One 0 to 5 L/min unit and one or more 0 to 100 mL/min units
for air, depending on number of cylinders in use for calibration.
7.3.4 Teflon Filter(s). 47-mm Teflon® filter for particulate collection.
7.4 Reagents
7.4.1 Neat Materials or Manufacturer-Certified Solutions/Mixtures. Best source (see Section 9).
7.4.2 Helium and Air. Ultra-high purity grade in gas cylinders. He is used as carrier gas in the GC.
7.4.3 Liquid Nitrogen or Liquid Carbon Dioxide. Used to cool secondary trap.
7.4.4 Deionized Water. High performance liquid chromatography (HPLC) grade, ultra-high purity (for
humidifier).
8. Collection of Samples in Canisters
8.1 Introduction
8.1.1 Canister samplers, sampling procedures, and canister cleaning procedures have not changed very
much from the description given in the original Compendium Method TO-14. Much of the material in this
Page 15-10 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs Method TO-15
section is therefore simply a restatement of the material given in Compendium Method TO-14, repeated here
in order to have all the relevant information in one place.
8.1.2 Recent notable additions to the canister technology has been in the application of canister-based
systems for example, to microenvironmental monitoring (8), the capture of breath samples (9), and sector
sampling to identify emission sources of VOCs (10).
8.1.3 EPA has also sponsored the development of a mathematical model to predict the storage stability
of arbitrary mixtures of trace gases in humidified air (3), and the investigation of the SilcoSteel™ process
of coating the canister interior with a film of fused silicate reduce surface activity (11). A recent summary
of storage stability data for VOCs in canisters is given in the open literature (5).
8.2 Sampling System Description
8.2.1 Subatmospheric Pressure Sampling [see Figure 1 (without metal bellows type pump)].
8.2.1.1 In preparation for subatmospheric sample collection in a canister, the canister is evacuated to
0.05 mm Hg (see Appendix C for discussion of evacuation pressure). When the canister is opened to the
atmosphere containing the VOCs to be sampled, the differential pressure causes the sample to flow into the
canister. This technique may be used to collect grab samples (duration of 10 to 30 seconds) or time-
weighted-average (TWA) samples (duration of 1-24 hours) taken through a flow-restrictive inlet (e.g., mass
flow controller, critical orifice).
8.2.1.2 With a critical orifice flow restrictor, there will be a decrease in the flow rate as the pressure
approaches atmospheric. However, with a mass flow controller, the subatmospheric sampling system can
maintain a constant flow rate from full vacuum to within about 7 kPa (1.0 psi) or less below ambient
pressure.
8.2.2 Pressurized Sampling [see Figure 1 (with metal bellows type pump)].
8.2.2.1 Pressurized sampling is used when longer-term integrated samples or higher volume samples
are required. The sample is collected in a canister using a pump and flow control arrangement to achieve
atypical 101-202 kPa (15-30 psig) final canister pressure. For example, a 6-liter evacuated canister can be
filled at 10 mL/min for 24 hours to achieve a final pressure of 144 kPa (21 psig).
8.2.2.2 In pressurized canister sampling, a metal bellows type pump draws in air from the sampling
manifold to fill and pressurize the sample canister.
8.2.3 All Samplers.
8.2.3.1 A flow control device is chosen to maintain a constant flow into the canister over the desired
sample period. This flow rate is determined so the canister is filled (to about 88.1 kPa for subatmospheric
pressure sampling or to about one atmosphere above ambient pressure for pressurized sampling) over the
desired sample period. The flow rate can be calculated by:
F = PxV
T x 60
where:
F = flow rate, mL/min.
P = final canister pressure, atmospheres absolute. P is approximately equal to
kPa gauge
101.2
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-11
-------
Method TO-15 VOCs
V = volume of the canister, mL.
T = sample period, hours.
For example, if a 6-L canister is to be filled to 202 kPa (2 atmospheres) absolute pressure in 24 hours, the
flow rate can be calculated by:
„ 2 x 6000 _. T . .
F = = 8.3 mL/min
24 x 60
8.2.3.2 For automatic operation, the timer is designed to start and stop the pump at appropriate times
for the desired sample period. The timer must also control the solenoid valve, to open the valve when
starting the pump and to close the valve when stopping the pump.
8.2.3.3 The use of the Skinner Magnelatch valve (see Figure 2) avoids any substantial temperature rise
that would occur with a conventional, normally closed solenoid valve that would have to be energized during
the entire sample period. The temperature rise in the valve could cause outgassing of organic compounds
from the Viton® valve seat material. The Skinner Magnelatch valve requires only a brief electrical pulse
to open or close at the appropriate start and stop times and therefore experiences no temperature increase.
The pulses may be obtained either with an electronic timer that can be programmed for short (5 to 60
seconds) ON periods, or with a conventional mechanical timer and a special pulse circuit. A simple electrical
pulse circuit for operating the Skinner Magnelatch solenoid valve with a conventional mechanical timer is
illustrated in Figure 2(a). However, with this simple circuit, the valve may operate unreliably during brief
power interruptions or if the timer is manually switched on and off too fast. A belter circuit incorporating
a time-delay relay to provide more reliable valve operation is shown in Figure 2(b).
8.2.3.4 The connecting lines between the sample inlet and the canister should be as short as possible
to minimize their volume. The flow rate into the canister should remain relatively constant over the entire
sampling period.
8.2.3.5 As an option, a second electronic timer may be used to start the auxiliary pump several hours
prior to the sampling period to flush and condition the inlet line.
8.2.3.6 Prior to field use, each sampling system must pass a humid zero air certification (see
Section 8.4.3). All plumbing should be checked carefully for leaks. The canisters must also pass a humid
zero air certification before use (see Section 8.4.1).
8.3 Sampling Procedure
8.3.1 The sample canister should be cleaned and tested according to the procedure in Section 8.4.1.
8.3.2 A sample collection system is assembled as shown in Figures 1 and 3 and must be cleaned
according to the procedure outlined in Sections 8.4.2 and 8.4.4.
[Note: The sampling system should be contained in an appropriate enclosure.]
8.3.3 Prior to locating the sampling system, the user may want to perform "screening analyses" using
a portable GC system, as outlined in Appendix B of Compendium Method TO-14A, to determine potential
volatile organics present and potential "hot spots." The information gathered from the portable GC screening
analysis would be used in developing a monitoring protocol, which includes the sampling system location,
based upon the "screening analysis" results.
Page 15-12 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs Method TO-15
8.3.4 After "screening analysis," the sampling system is located. Temperatures of ambient air and
sampler box interior are recorded on the canister sampling field test data sheet (FTDS), as documented in
Figure 9.
[Note: The following discussion is related to Figure 1]
8.3.5 To verify correct sample flow, a "practice" (evacuated) canister is used in the sampling system.
[Note: For a subatmospheric sampler, a flow meter and practice canister are needed. For the pump-driven
system, the practice canister is not needed, as the flow can be measured at the outlet of the system.]
A certified mass flow meter is attached to the inlet line of the manifold, just in front of the filter. The
canister is opened. The sampler is turned on and the reading of the certified mass flow meter is compared
to the sampler mass flow controller. The values should agree within ±10%. If not, the sampler mass flow
meter needs to be recalibrated or there is a leak in the system. This should be investigated and corrected.
[Note: Mass flow meter readings may drift. Check the zero reading carefully and add or subtract the zero
reading when reading or adjusting the sampler flow rate to compensate for any zero drift.]
After 2 minutes, the desired canister flow rate is adjusted to the proper value (as indicated by the certified
mass flow meter) by the sampler flow control unit controller (e.g., 3.5 mL/min for 24 hr, 7.0 mL/min for 12
hr). Record final flow under "CANISTER FLOW RATE" on the FTDS.
8.3.6 The sampler is turned off and the elapsed time meter is reset to 000.0.
[Note: Whenever the sampler is turned off, wait at least 30 seconds to turn the sampler back on.]
8.3.7 The "practice" canister and certified mass flow meter are disconnected and a clean certified (see
Section 8.4.1) canister is attached to the system.
8.3.8 The canister valve and vacuum/pressure gauge valve are opened.
8.3.9 Pressure/vacuum in the canister is recorded on the canister FTDS (see Figure 9) as indicated by
the sampler vacuum/pressure gauge.
8.3.10 The vacuum/pressure gauge valve is closed and the maximum-minimum thermometer is reset to
current temperature. Time of day and elapsed time meter readings are recorded on the canister FTDS.
8.3.11 The electronic timer is set to start and stop the sampling period at the appropriate times. Sampling
starts and stops by the programmed electronic timer.
8.3.12 After the desired sampling period, the maximum, minimum, current interior temperature and
current ambient temperature are recorded on the FTDS. The current reading from the flow controller is
recorded.
8.3.13 At the end of the sampling period, the vacuum/pressure gauge valve on the sampler is briefly
opened and closed and the pressure/vacuum is recorded on the FTDS. Pressure should be close to desired
pressure.
[Note: For a subatmospheric sampling system, if the canister is at atmospheric pressure when the field final
pressure check is performed, the sampling period may be suspect. This information should be noted on the
sampling field data sheet.]
Time of day and elapsed time meter readings are also recorded.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-13
-------
Method TO-15 VOCs
8.3.14 The canister valve is closed. The sampling line is disconnected from the canister and the canister
is removed from the system. For a subatmospheric system, a certified mass flow meter is once again
connected to the inlet manifold in front of the in-line filter and a "practice" canister is attached to the
Magnelatch valve of the sampling system. The final flow rate is recorded on the canister FTDS (see
Figure 9).
[Note: For a pressurized system, the final flow may be measured directly.]
The sampler is turned off.
8.3.15 An identification tag is attached to the canister. Canister serial number, sample number, location,
and date, as a minimum, are recorded on the tag. The canister is routinely transported back to the analytical
laboratory with other canisters in a canister shipping case.
8.4 Cleaning and Certification Program
8.4.1 Canister Cleaning and Certification.
8.4.1.1 All canisters must be clean and free of any contaminants before sample collection.
8.4.1.2 All canisters are leak tested by pressurizing them to approximately 206 kPa (30 psig) with zero
air.
[Note: The canister cleaning system in Figure 10 can be used for this task.]
The initial pressure is measured, the canister valve is closed, and the final pressure is checked after 24 hours.
If acceptable, the pressure should not vary more than ± 13.8 kPa (± 2 psig) over the 24 hour period.
8.4.1.3 A canister cleaning system may be assembled as illustrated in Figure 10. Cryogen is added
to both the vacuum pump and zero air supply traps. The canister(s) are connected to the manifold. The vent
shut-off valve and the canister valve(s) are opened to release any remaining pressure in the canister(s). The
vacuum pump is started and the vent shut-off valve is then closed and the vacuum shut-off valve is opened.
The canister(s) are evacuated to <0.05 mm Hg (see Appendix B) for at least 1 hour.
[Note: On a daily basis or more often if necessary, the cryogenic traps should be purged with zero air to
remove any trapped water from previous canister cleaning cycles.]
Air released/evacuated from canisters should be diverted to a fume hood.
8.4.1.4 The vacuum and vacuum/pressure gauge shut-off valves are closed and the zero air shut-off
valve is opened to pressurize the canister(s) with humid zero air to approximately 206 kPa (30 psig). If a zero
gas generator system is used, the flow rate may need to be limited to maintain the zero air quality.
8.4.1.5 The zero air shut-off valve is closed and the canister(s) is allowed to vent down to atmospheric
pressure through the vent shut-off valve. The vent shut-off valve is closed. Repeat Sections 8.4.1.3 through
8.4.1.5 two additional times for a total of three (3) evacuation/pressurization cycles for each set of canisters.
8.4.1.6 At the end of the evacuation/pressurization cycle, the canister is pressurized to 206 kPa (30
psig) with humid zero air. The canister is then analyzed by a GC/MS analytical system. Any canister that
has not tested clean (compared to direct analysis of humidified zero air of less than 0.2 ppbv of targeted
VOCs) should not be used. As a "blank" check of the canister(s) and cleanup procedure, the final humid zero
air fill of 100% of the canisters is analyzed until the cleanup system and canisters are proven reliable (less
than 0.2 ppbv of any target VOCs). The check can then be reduced to a lower percentage of canisters.
Page 15-14 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs Method TO-15
8.4.1.7 The canister is reattached to the cleaning manifold and is then reevacuated to <0.05 mm Hg
(see Appendix B) and remains in this condition until used. The canister valve is closed. The canister is
removed from the cleaning system and the canister connection is capped with a stainless steel fitting. The
canister is now ready for collection of an air sample. An identification tag is attached to the inlet of each
canister for field notes and chain-of-custody purposes. An alternative to evacuating the canister at this point
is to store the canisters and reevacuate them just prior to the next use.
8.4.1.8 As an option to the humid zero air cleaning procedures, the canisters are heated in an
isothermal oven not to exceed 100 ° C during evacuation of the canister to ensure that higher molecular weight
compounds are not retained on the walls of the canister.
[Note: For sampling more complex VOC mixtures the canisters should be heated to higher temperatures
during the cleaning procedure although a special high temperature valve would be needed].
Once heated, the canisters are evacuated to <0.05 mm Hg (see Appendix B) and maintained there for 1 hour.
At the end of the heated/evacuated cycle, the canisters are pressurized with humid zero air and analyzed by
a GC/MS system after a minimum of 12 hrs of "aging." Any canister that has not tested clean (less than 0.2
ppbv each of targeted compounds) should not be used. Once tested clean, the canisters are reevacuated to
<0.05 mm Hg (see Appendix B) and remain in the evacuated state until used. As noted in Section 8.4.1.7,
reevacuation can occur just prior to the next use.
8.4.2 Cleaning Sampling System Components.
8.4.2.1 Sample components are disassembled and cleaned before the sampler is assembled.
Nonmetallic parts are rinsed with HPLC grade deionized water and dried in a vacuum oven at 50°C.
Typically, stainless steel parts and fittings are cleaned by placing them in a beaker of methanol in an
ultrasonic bath for 15 minutes. This procedure is repeated with hexane as the solvent.
8.4.2.2 The parts are then rinsed with HPLC grade deionized water and dried in a vacuum oven at
100°Cfor 12 to 24 hours.
8.4.2.3 Once the sampler is assembled, the entire system is purged with humid zero air for 24 hours.
8.4.3 Zero Air Certification.
[Note: In the following sections, "certification" is defined as evaluating the sampling system with humid zero
air and humid calibration gases that pass through all active components of the sampling system. The system
is "certified" if no significant additions or deletions (less than 0.2 ppbv each of target compounds) have
occurred when challenged with the test gas stream.]
8.4.3.1 The cleanliness of the sampling system is determined by testing the sampler with humid zero
air without an evacuated gas sampling canister, as follows.
8.4.3.2 The calibration system and manifold are assembled, as illustrated in Figure 8. The sampler
(without an evacuated gas canister) is connected to the manifold and the zero air cylinder is activated to
generate a humid gas stream (2 L/min) to the calibration manifold [see Figure 8(b)].
8.4.3.3 The humid zero gas stream passes through the calibration manifold, through the sampling
system (without an evacuated canister) to the water management system/VOC preconcentrator of an
analytical system.
[Note: The exit of the sampling system (without the canister) replaces the canister in Figure 77.]
After the sample volume (e.g., 500 mL) is preconcentrated on the trap, the trap is heated and the VOCs are
thermally desorbed and refocussed on a cold trap. This trap is heated and the VOCs are thermally desorbed
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-15
-------
Method TO-15 VOCs
onto the head of the capillary column. The VOCs are refocussed prior to gas chromatographic separation.
Then, the oven temperature (programmed) increases and the VOCs begin to elute and are detected by a
GC/MS (see Section 10) system. The analytical system should not detect greater than 0.2 ppbv of any
targeted VOCs in order for the sampling system to pass the humid zero air certification test. Chromatograms
(using an FID) of a certified sampler and contaminated sampler are illustrated in Figures 12(a) and 12(b),
respectively. If the sampler passes the humid zero air test, it is then tested with humid calibration gas
standards containing selected VOCs at concentration levels expected in field sampling (e.g., 0.5 to 2 ppbv)
as outlined in Section 8.4.4.
8.4.4 Sampler System Certification with Humid Calibration Gas Standards from a Dynamic
Calibration System
8.4.4.1 Assemble the dynamic calibration system and manifold as illustrated in Figure 8.
8.4.4.2 Verify that the calibration system is clean (less than 0.2 ppbv of any target compounds) by
sampling ahumidified gas stream, without gas calibration standards, with a previously certified clean canister
(see Section 8.1).
8.4.4.3 The assembled dynamic calibration system is certified clean if less than 0.2 ppbv of any
targeted compounds is found.
8.4.4.4 For generating the humidified calibration standards, the calibration gas cylinder(s) containing
nominal concentrations of 10 ppmv in nitrogen of selected VOCs is attached to the calibration system as
illustrated in Figure 8. The gas cylinders are opened and the gas mixtures are passed through 0 to 10 mL/min
certified mass flow controllers to generate ppb levels of calibration standards.
8.4.4.5 After the appropriate equilibrium period, attach the sampling system (containing a certified
evacuated canister) to the manifold, as illustrated in Figure 8(b).
8.4.4.6 Sample the dynamic calibration gas stream with the sampling system.
8.4.4.7 Concurrent with the sampling system operation, realtime monitoring of the calibration gas
stream is accomplished by the on-line GC/MS analytical system [Figure 8(a)] to provide reference
concentrations of generated VOCs.
8.4.4.8 At the end of the sampling period (normally the same time period used for experiments), the
sampling system canister is analyzed and compared to the reference GC/MS analytical system to determine
if the concentration of the targeted VOCs was increased or decreased by the sampling system.
8.4.4.9 A recovery of between 90% and 110% is expected for all targeted VOCs.
8.4.5 Sampler System Certification without Compressed Gas Cylinder Standards.
8.4.5.1 Not all the gases on the Title III list are available/compatible with compressed gas standards.
In these cases sampler certification must be approached by different means.
8.4.5.2 Definitive guidance is not currently available in these cases; however, Section 9.2 lists several
ways to generate gas standards. In general, Compendium Method TO-14A compounds (see Table 1) are
available commercially as compressed gas standards.
9. GC/MS Analysis of Volatiles from Canisters
9.1 Introduction
9.1.1 The analysis of canister samples is accomplished with a GC/MS system. Fused silica capillary
columns are used to achieve high temporal resolution of target compounds. Linear quadrupole or ion trap
mass spectrometers are employed for compound detection. The heart of the system is composed of the
sample inlet concentrating device that is needed to increase sample loading into a detectable range. Two
Page 15-16 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs Method TO-15
examples of concentrating systems are discussed. Other approaches are acceptable as long as they are
compatible with achieving the system performance criteria given in Section 11.
9.1.2 With the first technique, a whole air sample from the canister is passed through a multisorbent
packing (including single adsorbent packings) contained within a metal or glass tube maintained at or above
the surrounding air temperature. Depending on the water retention properties of the packing, some or most
of the water vapor passes completely through the trap during sampling. Additional drying of the sample is
accomplished after the sample concentration is completed by forward purging the trap with clean, dry helium
or another inert gas (air is not used). The sample is then thermally desorbed from the packing and
backflushed from the trap onto a gas chromatographic column. In some systems a "refocusing" trap is placed
between the primary trap and the gas chromatographic column. The specific system design downstream of
the primary trap depends on technical factors such as the rate of thermal desorption and sampled volume, but
the objective in most cases is to enhance chromatographic resolution of the individual sample components
before detection on a mass spectrometer.
9.1.3 Sample drying strategies depend on the target list of compounds. For some target compound lists,
the multisorbent packing of the concentrator can be selected from hydrophobic adsorbents which allow a high
percentage of water vapor in the sample to pass through the concentrator during sampling and without
significant loss of the target compounds. However, if very volatile organic compounds are on the target list,
the adsorbents required for their retention may also strongly retain water vapor and a more lengthy dry purge
is necessary prior to analysis.
9.1.4 With the second technique, a whole air sample is passed through a concentrator where the VOCs
are condensed on a reduced temperature surface (cold trap). Subsequently, the condensed gases are thermally
desorbed and backflushed from the trap with an inert gas onto a gas chromatographic column. This
concentration technique is similar to that discussed in Compendium Method TO-14, although a membrane
dryer is not used. The sample size is reduced in volume to limit the amount of water vapor that is also
collected (100 mL or less may be necessary). The attendant reduction in sensitivity is offset by enhancing
the sensitivity of detection, for example by using an ion trap detector.
9.2 Preparation of Standards
9.2.1 Introduction.
9.2.1.1 When available, standard mixtures of target gases in high pressure cylinders must be certified
traceable to a NIST Standard Reference Material (SRM) or to a NIST/EPA approved Certified Reference
Material (CRM). Manufacturer's certificates of analysis must be retained to track the expiration date.
9.2.1.2 The neat standards that are used for making trace gas standards must be of high purity;
generally a purity of 98 percent or better is commercially available.
9.2.1.3 Cylinder(s) containing approximately 10 ppmv of each of the target compounds are typically
used as primary stock standards. The components may be purchased in one cylinder or in separate cylinders
depending on compatibility of the compounds and the pressure of the mixture in the cylinder. Refer to
manufacturer's specifications for guidance on purchasing and mixing VOCs in gas cylinders.
9.2.2 Preparing Working Standards.
9.2.2.1 Instrument Performance Check Standard. Prepare a standard solution of BFB in humidi-
fied zero air at a concentration which will allow collection of 50 ng of BFB or less under the optimized con-
centration parameters.
9.2.2.2 Calibration Standards. Prepare five working calibration standards in humidified zero air at
a concentration which will allow collection at the 2,5, 10, 20, and 50 ppbv level for each component under
the optimized concentration parameters.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-17
-------
Method TO-15 VOCs
9.2.2.3 Internal Standard Spiking Mixture. Prepare an internal spiking mixture containing bromo-
chloromethane, chlorobenzene-d5, and 1,4-difluorobenzene at 10 ppmv each in humidified zero air to be
added to the sample or calibration standard. 500 ^L of this mixture spiked into 500 mL of sample will result
in a concentration of 10 ppbv. The internal standard is introduced into the trap during the collection time
for all calibration, blank, and sample analyses using the apparatus shown in Figure 13 or by equivalent
means. The volume of internal standard spiking mixture added for each analysis must be the same from run
to run.
9.2.3 Standard Preparation by Dynamic Dilution Technique.
9.2.3.1 Standards may be prepared by dynamic dilution of the gaseous contents of a cylinder(s)
containing the gas calibration stock standards with humidified zero air using mass flow controllers and a
calibration manifold. The working standard may be delivered from the manifold to a clean, evacuated
canister using a pump and mass flow controller.
9.2.3.2 Alternatively, the analytical system may be calibrated by sampling directly from the manifold
if the flow rates are optimized to provide the desired amount of calibration standards. However, the use of
the canister as a reservoir prior to introduction into the concentration system resembles the procedure
normally used to collect samples and is preferred. Flow rates of the dilution air and cylinder standards (all
expressed in the same units) are measured using a bubble meter or calibrated electronic flow measuring
device, and the concentrations of target compounds in the manifold are then calculated using the dilution
ratio and the original concentration of each compound.
A/r -r- , i ^ (Original Cone.) (Std. Gas Flowrate)
Manifold Cone. = -^—^ L^ L
(Air Flowrate) + (Std. Gas Flowrate)
9.2.3.3 Consider the example of 1 mL/min flow of 10 ppmv standard diluted with 1,000 mL/min of
humid air provides a nominal 10 ppbv mixture, as calculated below:
Manifold Cone = (10 PPm)0 mL/min)(1000 ppb/1 ppm) = 1Q
(1000 mL/min) + (1 mL/min)
9.2.4 Standard Preparation by Static Dilution Bottle Technique
[Note: Standards may be prepared in canisters by spiking the canister with a mixture of components
prepared in a static dilution bottle (12). This technique is used specifically for liquid standards.]
9.2.4.1 The volume of a clean 2-liter round-bottom flask, modified with a threaded glass neck to accept
a Mininert septum cap, is determined by weighing the amount of water required to completely fill up the
flask. Assuming a density for the water of 1 g/mL, the weight of the water in grams is taken as the volume
of the flask in milliliters.
9.2.4.2 The flask is flushed with helium by attaching a tubing into the glass neck to deliver the helium.
After a few minutes, the tubing is removed and the glass neck is immediately closed with a Mininert septum
cap.
9.2.4.3 The flask is placed in a 60° C oven and allowed to equilibrate at that temperature for about
15 minutes. Predetermined aliquots of liquid standards are injected into the flask making sure to keep the
flask temperature constant at 60°C.
9.2.4.4 The contents are allowed to equilibrate in the oven for at least 30 minutes. To avoid
condensation, syringes must be preheated in the oven at the same temperature prior to withdrawal of aliquots
to avoid condensation.
Page 15-18 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs Method TO-15
9.2.4.5 Sample aliquots may then be taken for introduction into the analytical system or for further
dilution. An aliquot or aliquots totaling greater than 1 percent of the flask volume should be avoided.
9.2.4.6 Standards prepared by this method are stable for one week. The septum must be replaced with
each freshly prepared standard.
9.2.4.7 The concentration of each component in the flask is calculated using the following equation:
n (
Concentration, mg/L =
where: Va = Volume of liquid neat standard injected into the flask, \)L.
d = Density of the liquid neat standard, mg/(iL.
Vf = Volume of the flask, L.
9.2.4.8 To obtain concentrations in ppbv, the equation given in Section 9.2.5.7 can be used.
[Note: In the preparation of standards by this technique, the analyst should make sure that the volume of
neat standard injected into the flask does not result in an overpressure due to the higher partial pressure
produced by the standard compared to the vapor pressure in the flask. Precautions should also be taken to
avoid a significant decrease in pressure inside the flask after withdrawal ofaliquot(s).]
9.2.5 Standard Preparation Procedure in High Pressure Cylinders
[Note: Standards may be prepared in high pressure cylinders (13). A modified summary of the procedure
is provided below.]
9.2.5.1 The standard compounds are obtained as gases or neat liquids (greater than 98 percent purity).
9.2.5.2 An aluminum cylinder is flushed with high-purity nitrogen gas and then evacuated to better
than 25 in. Hg.
9.2.5.3 Predetermined amounts of each neat standard compound are measured using a microliter or
gastight syringe and injected into the cylinder. The cylinder is equipped with a heated injection port and
nitrogen flow to facilitate sample transfer.
9.2.5.4 The cylinder is pressurized to 1000 psig with zero nitrogen.
[Note: User should read all SOPs associated with generating standards in high pressure cylinders. Follow
all safety requirements to minimize danger from high pressure cylinders.]
9.2.5.5 The contents of the cylinder are allowed to equilibrate (-24 hrs) prior to withdrawal of aliquots
into the GC system.
9.2.5.6 If the neat standard is a gas, the cylinder concentration is determined using the following
equation:
Concentration, ppbv = —° umestandard x 1Q9
Volumedllutlon gas
[Note: Both values must be expressed in the same units.]
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-19
-------
Method TO-15
VOCs
9.2.5.7 If the neat standard is a liquid, the gaseous concentration can be determined using the following
equations:
V=
and:
n _ (mL)(d)
MW
where: V = Volume of injected compound at EPA standard temperature (25 °C) and pressure
(760 mm Hg), L.
n = Moles.
R = Gas constant, 0.08206 L-atm/mole °K.
T = 273 °K (standard temperature).
P = 1 standard pressure, 760 mm Hg (1 atm).
mL= Volume of liquid injected, mL.
d = Density of the neat standard, gm/mL.
MW = Molecular weight of the neat standard expressed, gm/gm-mole.
The volume of injected compound is divided by the cylinder volume at STP and then multiplied by 109 to
obtain the component concentration in ppb units.
9.2.6 Standard Preparation by Water Methods.
[Note: Standards may be prepared by a water purge and trap method (14) and summarized as follows].
9.2.6.1 A previously cleaned and evacuated canister is pressurized to 760 mm Hg absolute (1 atm) with
zero grade air.
9.2.6.2 The air gauge is removed from the canister and the sparging vessel is connected to the canister
with the short length of 1/16 in. stainless steel tubing.
[Note: Extra effort should be made to minimize possible areas of dead volume to maximize transfer of
analytesfrom the water to the canister.]
9.2.6.3 A measured amount of the stock standard solution and the internal standard solution is spiked
into 5 mL of water.
9.2.6A This water is transferred into the sparge vessel and purged with nitrogen for 10 mins at
100 mL/min. The sparging vessel is maintained at 40°C.
9.2.6.5 At the end of 10 mins, the sparge vessel is removed and the air gauge is re-installed, to further
pressurize the canister with pure nitrogen to 1500 mm Hg absolute pressure (approximately 29 psia).
9.2.6.6 The canister is allowed to equilibrate overnight before use.
9.2.6.7 A schematic of this approach is shown in Figure 14.
9.2.7 Preparation of Standards by Permeation Tubes.
9.2.7.1 Permeation tubes can be used to provide standard concentration of a trace gas or gases. The
permeation of the gas can occur from inside a permeation tube containing the trace species of interest to an
Page 15-20
Compendium of Methods for Toxic Organic Air Pollutants
January 1997
-------
VOCs Method TO-15
air stream outside. Permeation can also occur from outside a permeable membrane tube to an air stream
passing through the tube (e.g., a tube of permeable material immersed in a liquid).
9.2.7.2 The permeation system is usually held at a constant temperature to generate a constant
concentration of trace gas. Commercial suppliers provide systems for generation and dilution of over
250 compounds. Some commercial suppliers of permeation tube equipment are listed in Appendix D.
9.2.8 Storage of Standards.
9.2.8.1 Working standards prepared in canisters may be stored for thirty days in an atmosphere free
of potential contaminants.
9.2.8.2 It is imperative that a storage logbook be kept to document storage time.
10. GC/MS Operating Conditions
10.1 Preconcentrator
The following are typical cryogenic and adsorbent preconcentrator analytical conditions which, however,
depend on the specific combination of solid sorbent and must be selected carefully by the operator. The
reader is referred to Tables 1 and 2 of Compendium Method TO-17 for guidance on selection of sorbents.
An example of a system using a solid adsorbent preconcentrator with a cryofocusing trap is discussed in the
literature (15). Oven temperature programming starts above ambient.
10.1.1 Sample Collection Conditions
Cryogenic Trap Adsorbent Trap
Set point -150°C Set point 27°C
Sample volume -uptolOOmL Sample volume -uptol,OOOmL
Carrier gas purge flow - none Carrier gas purge flow - selectable
[Note: The analyst should optimize the flow rate, duration of sampling, and absolute sample volume to be
used. Other preconcentration systems may be used provided performance standards (see Section 11) are
realized.]
10.1.2 Desorption Conditions
Cryogenic Trap Adsorbent Trap
Desorb Temperature 120°C Desorb Temperature Variable
Desorb Flow Rate ~ 3 mL/min He Desorb Flow Rate ~3 mL/min He
Desorb Time <60 sec Desorb Time <60 sec
The adsorbent trap conditions depend on the specific solid adsorbents chosen (see manufacturers'
specifications).
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-21
-------
Method TO-15 VOCs
10.1.3 Trap Reconditioning Conditions.
Cryogenic Trap Adsorbent Trap
Initial bakeout 120°C(24hrs) Initial bakeout
Variable (24 hrs)
After each run 120°C (5 min) After each run Variable (5 min)
10.2 GC/MS System
10.2.1 Optimize GC conditions for compound separation and sensitivity. Baseline separation of benzene
and carbon tetrachloride on a 100% methyl polysiloxane stationary phase is an indication of acceptable
chromatographic performance.
10.2.2 The following are the recommended gas chromatographic analytical conditions when using a 50-
meter by 0.3-mm I.D., 1 (jm film thickness fused silica column with refocusing on the column.
Item Condition
Carrier Gas: Helium
Flow Rate: Generally 1-3 mL/min as recommended by manufacturer
Temperature Program: Initial Temperature: -50°C
Initial Hold Time: 2 min
Ramp Rate: 8° C/min
Final Temperature: 200°C
Final Hold Time: Until all target compounds elute.
10.2.3 The following are the recommended mass spectrometer conditions:
Item Condition
Electron Energy: 70 Volts (nominal)
Mass Range: 35-300 amu [the choice of 35 amu excludes the detection of some target compounds
such as methanol and formaldehyde, and the quantitation of others such as ethylene
oxide, ethyl carbamate, etc. (see Table 2). Lowering the mass range and using
special programming features available on modern gas chromatographs will be
necessary in these cases, but are not considered here.
Scan Time: To give at least 10 scans per peak, not to exceed 1 second per scan].
A schematic for atypical GC/MS analytical system is illustrated in Figure 15.
10.3 Analytical Sequence
10.3.1 Introduction. The recommended GC/MS analytical sequence for samples during each 24-hour
time period is as follows:
• Perform instrument performance check using bromofluorobenzene (BFB).
• Initiate multi-point calibration or daily calibration checks.
Page 15-22 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs Method TO-15
• Perform a laboratory method blank.
• Complete this sequence for analysis of <20 field samples.
10.4 Instrument Performance Check
10.4.1 Summary. It is necessary to establish that a given GC/MS meets tuning and standard mass
spectral abundance criteria prior to initiating any data collection. The GC/MS system is set up according to
the manufacturer's specifications, and the mass calibration and resolution of the GC/MS system are then
verified by the analysis of the instrument performance check standard, bromofluorobenzene (BFB).
10.4.2 Frequency. Prior to the analyses of any samples, blanks, or calibration standards, the Laboratory
must establish that the GC/MS system meets the mass spectral ion abundance criteria for the instrument
performance check standard containing BFB. The instrument performance check solution must be analyzed
initially and once per 24-hour time period of operation.
The 24-hour time period for GC/MS instrument performance check and standards calibration (initial
calibration or daily calibration check criteria) begins at the injection of the BFB which the laboratory records
as documentation of a compliance tune.
10.4.3 Procedure. The analysis of the instrument performance check standard is performed by trapping
50 ng of BFB under the optimized preconcentration parameters. The BFB is introduced from a cylinder into
the GC/MS via a sample loop valve injection system similar to that shown in Figure 13.
The mass spectrum of BFB must be acquired in the following manner. Three scans (the peak apex scan and
the scans immediately preceding and following the apex) are acquired and averaged. Background subtraction
is conducted using a single scan prior to the elution of BFB.
10.4.4 Technical Acceptance Criteria. Prior to the analysis of any samples, blanks, or calibration
standards, the analyst must establish that the GC/MS system meets the mass spectral ion abundance criteria
for the instrument performance check standard as specified in Table 3.
10.4.5 Corrective Action. If the BFB acceptance criteria are not met, the MS must be retuned. It may
be necessary to clean the ion source, or quadrupoles, or take other necessary actions to achieve the
acceptance criteria.
10.4.6 Documentation. Results of the BFB tuning are to be recorded and maintained as part of the
instrumentation log.
10.5 Initial Calibration
10.5.1 Summary. Prior to the analysis of samples and blanks but after the instrument performance
check standard criteria have been met, each GC/MS system must be calibrated at five concentrations that
span the monitoring range of interest in an initial calibration sequence to determine instrument sensitivity
and the linearity of GC/MS response for the target compounds. For example, the range of interest may be
2 to 20 ppbv, in which case the five concentrations would be 1, 2, 5, 10 and 25 ppbv.
One of the calibration points from the initial calibration curve must be at the same concentration as the daily
calibration standard (e.g., 10 ppbv).
10.5.2 Frequency. Each GC/MS system must be recalibrated following corrective action (e.g., ion
source cleaning or repair, column replacement, etc.) which may change or affect the initial calibration criteria
or if the daily calibration acceptance criteria have not been met.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-23
-------
Method TO-15 VOCs
If time remains in the 24-hour time period after meeting the acceptance criteria for the initial calibration,
samples may be analyzed.
If time does not remain in the 24-hour period after meeting the acceptance criteria for the initial calibration,
a new analytical sequence shall commence with the analysis of the instrument performance check standard
followed by analysis of a daily calibration standard.
10.5.3 Procedure. Verify that the GC/MS system meets the instrument performance criteria in
Section 10.4.
The GC must be operated using temperature and flow rate parameters equivalent to those in Section 10.2.2.
Calibrate the preconcentration-GC/MS system by drawing the standard into the system. Use one of the
standards preparation techniques described under Section 9.2 or equivalent.
A minimum of five concentration levels are needed to determine the instrument sensitivity and linearity. One
of the calibration levels should be near the detection level for the compounds of interest. The calibration
range should be chosen so that linear results are obtained as defined in Sections 10.5.1 and 10.5.5.
Quantitation ions for the target compounds are shown in Table 2. The primary ion should be used unless
interferences are present, in which case a secondary ion is used.
10.5.4 Calculations.
[Note: In the following calculations, an internal standard approach is used to calculate response factors.
The area response used is that of the primary quantitation ion unless otherwise stated.]
10.5.4.1 Relative Response Factor (RRF). Calculate the relative response factors for each target
compound relative to the appropriate internal standard (i.e., standard with the nearest retention time) using
the following equation:
RRF = ^£*
where: RRF = Relative response factor.
Ax = Area of the primary ion for the compound to be measured, counts.
A1S = Area of the primary ion for the internal standard, counts.
C1S = Concentration of internal standard spiking mixture, ppbv.
Cx = Concentration of the compound in the calibration standard, ppbv.
[Note: The equation above is valid under the condition that the volume of internal standard spiking mixture
added in oilfield and QC analyses is the same from run to run, and that the volume of field and QC sample
introduced into the trap is the same for each analysis. Cis and Cx must be in the same units.]
10.5.4.2 Mean Relative Response Factor. Calculate the mean RRF for each compound by averaging
the values obtained at the five concentrations using the following equation:
RRF = -
Page 15-24 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs Method TO-15
where: RRF = Mean relative response factor.
X; = RRF of the compound at concentration i.
n = Number of concentration values, in this case 5.
10.5.4.3 Percent Relative Standard Deviation (%RSD). Using the RRFs from the initial calibration,
calculate the %RSD for all target compounds using the following equations:
-
%RSD = x 100
RRF
and
- RRF)Z
N - 1
where: SD^ = Standard deviation of initial response factors (per compound).
j = Relative response factor at a concentration level i.
RRF = Mean of initial relative response factors (per compound).
10.5.4.4 Relative Retention Times (RRT). Calculate the RRTs for each target compound over the
initial calibration range using the following equation:
RT
RRT =
where: RTC = Retention time of the target compound, seconds
RT1S = Retention time of the internal standard, seconds.
10.5.4.5 Mean of the Relative Retention Times (RRT). Calculate the mean of the relative retention
times (RRT) for each analyte target compound over the initial calibration range using the following
equation:
RRT =
where: RRT = Mean relative retention time for the target compound for each initial calibration
standard.
RRT = Relative retention time for the target compound at each calibration level.
10.5.4.6 Tabulate Primary Ion Area Response (Y) for Internal Standard. Tabulate the area
response (Y) of the primary ions (see Table 2) and the corresponding concentration for each compound and
internal standard.
10.5.4.7 Mean Area Response (Y) for Internal Standard. Calculate the mean area response (Y)
for each internal standard compound over the initial calibration range using the following equation:
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-25
-------
Method TO-15VOCs
where: Y= Mean area response.
Y = Area re sponse for the primary quantitation ion for the internal standard for each initial
calibration standard.
10.5.4.8 Mean Retention Times (RT). Calculate the mean of the retention times (RT) for each
internal standard over the initial calibration range using the following equation:
RT.
RT =
where: RT = Mean retention time, seconds
RT = Retention time for the internal standard for each initial calibration standard, seconds.
10.5.5 Technical Acceptance Criteria for the Initial Calibration.
10.5.5.1 The calculated %RSD for the RRF for each compound in the calibration table must be less
than 30% with at most two exceptions up to a limit of 40%.
10.5.5.2 The RRT for each target compound at each calibration level must be withiin 0.06 RRT units
of the mean RRT for the compound.
10.5.5.3 The area response Y of at each calibration level must be within 40% of the mean area
response Y over the initial calibration range for each internal standard.
10.5.5.4 The retention time shift for each of the internal standards at each calibration level must be
within 20 s of the mean retention time over the initial calibration range for each internal standard.
10.5.6 Corrective Action.
10.5.6.1 Criteria. If the initial calibration technical acceptance criteria are notmet, inspect the system
forproblems. Itmay be necessary to clean the ion source, change the column, ortake other corrective actions
to meet the initial calibration technical acceptance criteria.
10.5.6.2 Schedule. Initial calibration acceptance criteria must be met before any field samples,
performance evaluation (PE) samples, or blanks are analyzed.
10.6 Daily Calibration
10.6.1 Summary. Prior to the analysis of samples and blanks but after tuning criteria have been met,
the initial calibration of each GC/MS system must be routinely checked by analyzing a daily calibration
standard to ensure that the instrument continues to remain under control. The daily calibration standard,
which is the nominal 10 ppbv level calibration standard, should contain all the target compounds.
10.6.2 Frequency. A check of the calibration curve must be performed once every 24 hours on a
GC/MS system that has met the tuning criteria. The daily calibration sequence starts with the injection of
the BFB. If the BFB analysis meets the ion abundance criteria for BFB, then a daily calibration standard may
be analyzed.
10.6.3 Procedure. The mid-level calibration standard (10 ppbv) is analyzed in a GC/MS system that
has met the tuning and mass calibration criteria following the same procedure in Section 10.5.
10.6.4 Calculations. Perform the following calculations.
Page 15-26 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs Method TO-15
[Note: As indicated earlier, the area response of the primary quantitation ion is used unless otherwise
stated.]
10.6.4.1 Relative Response Factor (RRF). Calculate a relative response factor (RRF) for each target
compound using the equation in Section 10.5.4.1.
10.6.4.2 Percent Difference (%D). Calculate the percent difference in the RRF of the daily RRF
(24-hour) compared to the mean RRF in the most recent initial calibration. Calculate the %D for each target
compound using the following equation:
RRF - RRF
O/OD = £ L x 100
RRF
where: RRFC = RRF of the compound in the continuing calibration standard.
RRF = Mean RRF of the compound in the most recent initial calibration.
10.6.5 Technical Acceptance Criteria. The daily calibration standard must be analyzed at the
concentration level and frequency described in this Section 10.6 and on a GC/MS system meeting the BFB
instrument performance check criteria (see Section 10.4).
The %D for each target compound in a daily calibration sequence must be within ±30 percent in order to
proceed with the analysis of samples and blanks. A control chart showing %D values should be maintained.
10.6.6 Corrective Action. If the daily calibration technical acceptance criteria are not met, inspect the
system for problems. It may be necessary to clean the ion source, change the column, or take other corrective
actions to meet the daily calibration technical acceptance criteria.
Daily calibration acceptance criteria must be met before any field samples, performance evaluation (PE)
samples, or blanks are analyzed. If the % D criteria are not met, it will be necessary to rerun the daily
calibration sample.
10.7 Blank Analyses
10.7.1 Summary. To monitor for possible laboratory contamination, laboratory method blanks are
analyzed at least once in a 24-hour analytical sequence. All steps in the analytical procedure are performed
on the blank using all reagents, standards, equipment, apparatus, glassware, and solvents that would be used
for a sample analysis.
A laboratory method blank (LMB) is an unused, certified canister that has not left the laboratory. The blank
canister is pressurized with humidified, ultra-pure zero air and carried through the same analytical procedure
as a field sample. The injected aliquot of the blank must contain the same amount of internal standards that
are added to each sample.
10.7.2 Frequency. The laboratory method blank must be analyzed after the calibration standard(s) and
before any samples are analyzed.
Whenever a high concentration sample is encountered (i.e., outside the calibration range), a blank analysis
should be performed immediately after the sample is completed to check for carryover effects.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-27
-------
Method TO-15 VOCs
10.7.3 Procedure. Fill a cleaned and evacuated canister with humidified zero air (RH >20 percent, at
25°C). Pressurize the contents to 2 atm.
The blank sample should be analyzed using the same procedure outlined under Section 10.8.
10.7.4 Calculations. The blanks are analyzed similar to a field sample and the equations in
Section 10.5.4 apply.
10.7.5 Technical Acceptance Criteria. A blank canister should be analyzed daily.
The area response for each internal standard (IS) in the blank must be within ±40 percent of the mean area
response of the IS in the most recent valid calibration.
The retention time for each of the internal standards must be within ±0.33 minutes between the blank and
the most recent valid calibration.
The blank should not contain any target analyte at a concentration greater than its quantitation level (three
times the MDL as defined in Section 11.2) and should not contain additional compounds with elution
characteristics and mass spectral features that would interfere with identification and measurement of a
method analyte.
10.7.6 Corrective Action. If the blanks do not meet the technical acceptance criteria, the analyst should
consider the analytical system to be out of control. It is the responsibility of the analyst to ensure that
contaminants in solvents, reagents, glassware, and other sample storage and processing hardware that lead
to discrete artifacts and/or elevated baselines in gas chromatograms be eliminated. If contamination is a
problem, the source of the contamination must be investigated and appropriate corrective measures need to
be taken and documented before further sample analysis proceeds.
If an analyte in the blank is found to be out of control (i.e., contaminated) and the analyte is also found in
associated samples, those sample results should be "flagged" as possibly contaminated.
10.8 Sample Analysis
10.8.1 Summary. An aliquot of the air sample from a canister (e.g., 500 mL) is preconcentrated and
analyzed by GC/MS under conditions stated in Sections 10.1 and 10.2. If using the multisorbent/dry purge
approach, adjust the dry purge volume to reduce water effects in the analytical system to manageable levels.
[Note: The analyst should be aware that pressurized samples of high humidity samples will contain
condensed water. As a result, the humidity of the sample released from the canister during analysis will vary
In humidity, being lower at the higher canister pressures and Increasing In humidity as the canister pressures
decreases. Storage integrity of water soluble compounds may also be affected.]
10.8.2 Frequency. If time remains in the 24-hour period in which an initial calibration is performed,
samples may be analyzed without analysis of a daily calibration standard.
If time does not remain in the 24-hour period since the injection of the instrument performance check
standard in which an initial calibration is performed, both the instrument performance check standard and
the daily calibration standard should be analyzed before sample analysis may begin.
10.8.3 Procedure for Instrumental Analysis. Perform the following procedure for analysis.
10.8.3.1 All canister samples should be at temperature equilibrium with the laboratory.
Page 15-28 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs Method TO-15
10.8.3.2 Check and adjust the mass flow controllers to provide correct flow rates for the system.
10.8.3.3 Connect the sample canister to the inlet of the GC/MS analytical system, as shown in
Figure 15 [Figure 16 shows an alternate two stage concentrator using multisorbent traps followed by a trap
cooled by a closed cycle cooler (15)]. The desired sample flow is established through the six-port chromato-
graphic valve and the preconcentrator to the downstream flow controller. The absolute volume of sample
being pulled through the trap must be consistent from run to run.
10.8.3.4 Heat/cool the GC oven and cryogenic or adsorbent trap to their set points. Assuming a six-
port value is being used, as soon as the trap reaches its lower set point, the six-port chromatographic valve
is cycled to the trap position to begin sample collection. Utilize the sample collection time which has been
optimized by the analyst.
10.8.3.5 Use the arrangement shown in Figure 13, (i.e., a gastight syringe or some alternate method)
introduce an internal standard during the sample collection period. Add sufficient internal standard
equivalent to 10 ppbv in the sample. For example, a 0.5 mL volume of a mixture of internal standard
compounds, each at 10 ppmv concentration, added to a sample volume of 500 mL, will result in 10 ppbv of
each internal standard in the sample.
10.8.3.6 After the sample and internal standards are preconcentrated on the trap, the GC sampling
valve is cycled to the inject position and the trap is swept with helium and heated. Assuming a focusing trap
is being used, the trapped analytes are thermally desorbed onto a focusing trap and then onto the head of the
capillary column and are separated on the column using the GC oven temperature program. The canister
valve is closed and the canister is disconnected from the mass flow controller and capped. The trap is
maintained at elevated temperature until the beginning of the next analysis.
10.8.3.7 Upon sample injection onto the column, the GC/MS system is operated so that the MS scans
the atomic mass range from 35 to 300 amu. At least ten scans per eluting chromatographic peak should be
acquired. Scanning also allows identification of unknown compounds in the sample through searching of
library spectra.
10.8.3.8 Each analytical run must be checked for saturation. The level at which an individual
compound will saturate the detection system is a function of the overall system sensitivity and the mass
spectral characteristics of that compound.
10.8.3.9 Secondary ion quantitation is allowed only when there are sample matrix interferences with
the primary ion. If secondary ion quantitation is performed, document the reasons in the laboratory record
book.
10.8.4 Calculations. The equation below is used for calculating concentrations.
c - A
where: Cx = Compound concentration, ppbv.
Ax = Area of the characteristic ion for the compound to be measured, counts.
A1S = Area of the characteristic ion for the specific internal standard, counts.
C1S = Concentration of the internal standard spiking mixture, ppbv.
RRF = Relative response factor from the analysis of the continuing calibration standard
or the mid level standard of the initial calibration.
DF = Dilution factor calculated as described in section 2. If no dilution is performed,
DF= 1.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-29
-------
Method TO-15 VOCs
[Note: The equation above is valid under the condition that the volume (-500 juL) of internal standard
spiking mixture added in all field and QC analyses is the same from run to run, and that the volume
(-500 mL) of field and QC sample introduced into the trap is the same for each analysis.]
10.8.5 Technical Acceptance Criteria.
[Note: If the most recent valid calibration is an initial calibration, internal standard area responses and
RTs in the sample are evaluated against the corresponding internal standard area responses andRTs in the
mid level standard (10 ppbv) of the initial calibration.]
10.8.5.1 The field sample must be analyzed on a GC/MS system meeting the BFB tuning, initial
calibration, and continuing calibration technical acceptance criteria at the frequency described in
Sections 10.4, 10.5 and 10.6.
10.8.5.2 The field samples must be analyzed along with a laboratory method blank that met the blank
technical acceptance criteria.
10.8.5.3 All of the target analyte peaks should be within the initial calibration range.
10.8.5.4 The retention time for each internal standard must be within ±0.33 minutes of the retention
time of the internal standard in the most recent valid calibration.
10.8.6 Corrective Action. If the on-column concentration of any compound in any sample exceeds the
initial calibration range, an aliquot of the original sample must be diluted and reanalyzed. Guidance in
performing dilutions and exceptions to this requirement are given below.
• Use the results of the original analysis to determine the approximate dilution factor required to get the
largest analyte peak within the initial calibration range.
• The dilution factor chosen should keep the response of the largest analyte peak for a target compound
in the upper half of the initial calibration range of the instrument.
• Do not submit data for more than two analyses, i.e., the original sample and one dilution, or, if the
screening procedure was employed, the most concentrated dilution analyzed and one further dilution.
[Note: Analysis involving dilution should be reported with a dilution factor and nature of the dilution gas.]
10.8.6.1 Internal standard responses and retention times must be evaluated during or immediately after
data acquisition. If the retention time for any internal standard changes by more than 20 sec from the latest
daily (24-hour) calibration standard (or mean retention time over the initial calibration range), the GC/MS
system must be inspected for malfunctions, and corrections made as required.
10.8.6.2 If the area response for any internal standard changes by more than ±40 percent between the
sample and the most recent valid calibration, the GC/MS system must be inspected for malfunction and
corrections made as appropriate. When corrections are made, reanalysis of samples analyzed while the
system was malfunctioning is necessary.
10.8.6.3 If, after reanalysis, the area responses or the RTs for all internal standards are inside the
control limits, then the problem with the first analysis is considered to have been within the control of the
Laboratory. Therefore, submit only data from the analysis with SICPs within the limits. This is considered
the initial analysis and should be reported as such on all data deliverables.
11. Requirements for Demonstrating Method Acceptability for VOC Analysis from Canisters
Page 15-30 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs Method TO-15
11.1 Introduction
11.1.1 There are three performance criteria which must be met for a system to qualify under
Compendium Method TO-15. These criteria are: the method detection limit of < 0.5 ppbv, replicate precision
within 25 percent, and audit accuracy within 30 percent for concentrations normally expected in
contaminated ambient air (0.5 to 25 ppbv).
11.1.2 Either SIM or SCAN modes of operation can be used to achieve these criteria, and the choice of
mode will depend on the number of target compounds, the decision of whether or not to determine tentatively
identified compounds along with other VOCs on the target list, as well as on the analytical system
characteristics.
11.1.3 Specific criteria for each Title III compound on the target compound list must be met by the
analytical system. These criteria were established by examining summary data from EPA's Toxics Air
Monitoring System Network and the Urban Air Toxics Monitoring Program network. Details for the
determination of each of the criteria follow.
11.2 Method Detection Limit
11.2.1 The procedure chosen to define the method detection limit is that given in the Code of Federal
Regulations (40 CFR 136 Appendix B).
11.2.2 The method detection limit is defined for each system by making seven replicate measurements
of the compound of interest at a concentration near (within a factor of five) the expected detection limit,
computing the standard deviation for the seven replicate concentrations, and multiplying this value by 3.14
(i.e., the Student's t value for 99 percent confidence for seven values). Employing this approach, the
detection limits given in Table 4 were obtained for some of the VOCs of interest.
11.3 Replicate Precision
11.3.1 The measure of replicate precision used for this program is the absolute value of the difference
between replicate measurements of the sample divided by the average value and expressed as a percentage
as follows:
|x. - x,
percent difference = = x 100
where: Xj = First measurement value.
x2 = Second measurement value.
x = Average of the two values.
11.3.2 There are several factors which may affect the precision of the measurement. The nature of the
compound of interest itself such as molecular weight, water solubility, polarizability, etc., each have some
effect on the precision, for a given sampling and analytical system. For example, styrene, which is classified
as a polar VOC, generally shows slightly poorer precision than the bulk of nonpolar VOCs. A primary
influence on precision is the concentration level of the compound of interest in the sample, i.e., the precision
degrades as the concentration approaches the detection limit. A conservative measure was obtained from
replicate analysis of "real world" canister samples from the TAMS and UATMP networks. These data are
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-31
-------
Method TO-15 VOCs
summarized in Table 5 and suggest that a replicate precision value of 25 percent can be achieved for each
of the target compounds.
11.4 Audit Accuracy
11.4.1 A measure of analytical accuracy is the degree of agreement with audit standards. Audit accuracy
is defined as the difference between the nominal concentration of the audit compound and the measured
value divided by the audit value and expressed as a percentage, as illustrated in the following equation:
. ,.. , „, Spiked Value - Observed Value ,„„
Audit Accuracy, % = — x 100
Spiked Value
11.4.2 Audit accuracy results for TAMS and UATMP analyses are summarized in Table 6 and were used
to form the basis for a selection of 30 percent as the performance criterion for audit accuracy.
Page 15-32 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs Method TO-15
12. References
1. Compendium of Methods for the Determination of Toxic Organic Compounds in Ambient Air: Method
TO-14A, Second Edition, U. S. Environmental Protection Agency, Research Triangle Park, NC,
EPA 600/625/R-96/010b, January 1997.
2. Winberry, W. T., Jr., et al., Statement-of-Work (SOW) for the Analysis of Air Toxics From Superfund
Sites, U. S. Environmental Protection Agency, Office of Solid Waste, Contract Laboratory Program,
Washington, D.C., Draft Report, June 1990.
3. Coutant, R.W., Theoretical Evaluation of Stability of Volatile Organic Chemicals and Polar Volatile
Organic Chemicals in Canisters, U. S. Environmental Protection Agency, EPA Contract No. 68-DO-0007,
Work Assignment No. 45, Subtask 2, Battelle, Columbus, OH, June 1993.
4. Kelly, T.J., Mukund, R., Gordon, S.M., and Hays, M.J., Ambient Measurement Methods and Properties
of the 189 Title III Hazardous Air Pollutants, U. S. Environmental Protection Agency, EPA Contract No. 68-
DO-0007, Work Assignment 44, Battelle, Columbus, OH, March 1994.
5. Kelly T. J. and Holdren, M.W., "Applicability of Canisters for Sample Storage in the Determination of
Hazardous Air Pollutants," Atmos. Environ., Vol. 29, 2595-2608, May 1995.
6. Kelly, T.J., Callahan, P.J., Pleil, J.K., and Evans, G.E., "Method Development and Field Measurements
for Polar Volatile Organic Compounds in Ambient Air," Environ. Sci. Technol.,Vol.21, 1146-1153, 1993.
7. McClenny, W.A., Oliver, K.D. and Daughtrey, E.H.., Jr. "Dry Purging of Solid Adsorbent Traps to
Remove Water Vapor Before Thermal Desorption of Trace Organic Gases," J. Air and Waste Manag. Assoc.,
Vol. 45, 792-800, June 1995.
8. Whitaker, D.A., Fortmann, R.C. and Lindstrom, A.B. "Development and Testing of a Whole Air Sampler
for Measurement of Personal Exposures to Volatile Organic Compounds," Journal ofExposure Analysis and
Environmental Epidemiology, Vol. 5, No. 1, 89-100, January 1995.
9. Pleil, J.D. and Lindstrom, A.B., "Collection of a Single Alveolar Exhaled Breath for Volatile Organic
Compound Analysis," American Journal of Industrial Medicine,Vol. 28, 109-121, 1995.
10. Pleil, J.D. and McClenny, W.A., "Spatially Resolved Monitoring for Volatile Organic Compounds Using
Remote Sector Sampling," Atmos. Environ., Vol. 27A, No. 5, 739-747, August 1993.
11. Holdren, M.W., et al., Unpublished Final Report, EPA Contract 68-DO-0007, Battelle, Columbus, OH.
Available from J.D. Pleil, MD-44, U. S. Environmental Protection Agency, Research Triangle Park, NC,
27711,919-541-4680.
12. Morris, C.M., Burkley, R.E. and Bumgarner, J.E., "Preparation of Multicomponent Volatile Organic
Standards Using Dilution Bottles," Anal. Letts., Vol. 16 (A20), 1585-1593, 1983.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-33
-------
Method TO-15 VOCs
13. Pollack, A.J., Holdren, M.W., "Multi-Adsorbent Preconcentration and Gas Chromatographic Analysis
of Air Toxics With an Automated Collection/Analytical System," in the Proceedings of the 1990
EPA/A&WMA International Symposium of Measurement of Toxic and Related Air Pollutants, U. S.
Environmental Protection Agency, Research Triangle Park, NC, EPA/600/9-90-026, May 1990.
14. Stephenson, J.H.M., Allen, F., Slagle, T., "Analysis of Volatile Organics in Air via Water Methods" in
Proceedings of the 1990 EPA/A&WMA International Symposium on Measurement of Toxic and Related Air
Pollutants, U. S. Environmental Protection Agency, Research Triangle Park, NC, EPA 600/9-90-026,
May 1990.
15. Oliver, K. D., Adams, J. R, Davehtrey, E. H., Jr., McClenny, W. A., Young, M. J., and Parade, M. A.,
"Techniques for Monitoring Toxices VOCs in Air: Sorbent Preconcentration Closed-Cycle Cooler
Cryofocusing, and GC/MS Analysis," Environ. Sci. Technol.,Vol. 30, 1938-1945, 1996.
Page 15-34 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs
Method TO-15
APPENDIX A.
LISTING OF SOME COMMERCIAL WATER
MANAGEMENT SYSTEMS USED WITH AUTOGC SYSTEMS
Tekmar Dohrman Company
7143 East Kemper Road
Post Office Box 429576
Cincinnati, Ohio 45242-9576
(513)247-7000
(513) 247-7050 (Fax)
(800)543-4461
[Moisture control module]
Entech Laboratory Automation
950 Enchanted Way No. 101
Simi Valley, California 93065
(805) 527-5939
(805) 527-5687 (Fax)
[Microscale Purge and Trap]
Dynatherm Analytical Instruments
Post Office Box 159
Kelton, Pennsylvania 19346
(215) 869-8702
(215) 869-3885 (Fax)
[Thermal Desorption System]
XonTech Inc.
6862 Hayenhurst Avenue
VanNuys, CA 91406
(818)787-7380
(818) 787-4275 (Fax)
[Multi-adsorbent trap/dry purge]
Graseby
500 Technology Ct.
Smyrna, Georgia 30082
(770)319-9999
(770) 319-03 3 6 (Fax)
(800)241-6898
[Controlled Desorption Trap]
Varian Chromatography System
2700 Mitchell Drive
Walnut Creek, California 94898
(510)945-2196
(510) 945-23 3 5 (FAX)
[Variable Temperature Adsorption Trap]
January 1997
Compendium of Methods for Toxic Organic Air Pollutants
Page 15-35
-------
Method TO-15 VOCs
APPENDIX B.
COMMENT ON CANISTER CLEANING PROCEDURES
The canister cleaning procedures given in Section 8.4 require that canister pressure be reduced to <0.05mm
Hg before the cleaning process is complete. Depending on the vacuum system design (diameter of
connecting tubing, valve restrictions, etc.) and the placement of the vacuum gauge, the achievement of this
value may take several hours. In any case, the pressure gauge should be placed near the canisters to
determine pressure. The objective of requiring a low pressure evacuation during canister cleaning is to
reduce contaminants. If canisters can be routinely certified (<0.2 ppbv for target compounds) while using
a higher vacuum, then this criteria can be relaxed. However, the ultimate vacuum achieved during cleaning
should always be <0.2mm Hg.
Canister cleaning as described in Section 8.4 and illustrated in Figure 10 requires components with special
features. The vacuum gauge shown in Figure 10 must be capable of measuring O.OSmmHg with less than
a 20% error. The vacuum pump used for evacuating the canister must be noncontaminating while being
capable of achieving the 0.05 mm Hg vacuum as monitored near the canisters. Thermoelectric vacuum
gauges and turbomolecular drag pumps are typically being used for these two components.
An alternate to achieving the canister certification requirement of <0.2 ppbv for all target compounds is the
criteria used in Compendium Method TO-12 that the total carbon count be <10ppbC. This check is less
expensive and typically more exacting than the current certification requirement and can be used if proven
to be equivalent to the original requirement. This equivalency must be established by comparing the total
nonmethane organic carbon (TNMOC) expressed in ppbC to the requirement that individual target
compounds be <0.2 ppbv for a series of analytical runs.
Page 15-36 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
VOCs Method TO-15
APPENDIX C.
LISTING OF COMMERCIAL MANUFACTURERS AND RE-SUPPLIERS OF
SPECIALLY-PREPARED CANISTERS
BRC/Rasmussen
17010NW Skyline Blvd.
Portland, Oregon 97321
(503)621-1435
Meriter
1790 Potrero Drive
San Jose, CA 95124
(408) 265-6482
Restek Corporation
110 Benner Circle
Bellefonte, PA 16823-8812
(814)353-1300
(800)356-1688
Scientific Instrumentation Specialists
P.O. Box 8941
815 Courtney Street
Moscow, ID 83843
(208) 882-3860
Graseby
500 Technology Ct.
Smyrna, Georgia 30082
(404)319-9999
(800)241-6898
XonTech Inc.
6862 Hayenhurst Avenue
VanNuys, CA 91406
(818)787-7380
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-37
-------
Method TO-15 VOCs
APPENDIX D.
LISTING OF COMMERCIAL SUPPLIERS OF PERMEATION TUBES AND SYSTEMS
Kin-Tek
504 Laurel St.
Lamarque, Texas 77568
(409) 938-3627
(800) 326-3627
Vici Metronics, Inc.
2991 Corvin Drive
Santa Clara, CA 95 051
(408) 737-0550
Analytical Instrument Development, Inc.
Rt. 41 and Newark Rd.
Avondale, PA 19311
(215)268-3181
Ecology Board, Inc.
9257 Independence Ave.
Chatsworth, CA91311
(213)882-6795
Tracer, Inc.
6500 Tracer Land
Austin, TX
(512)926-2800
Metronics Associates, Inc.
3201 Porter Drive
Standford Industrial Park
Palo Alto, CA 94304
(415)493-5632
Page 15-38 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
c_
88
88
i
•5-
i?
88
era
(Jl
TABLE 1. VOLATILE ORGANIC COMPOUNDS ON
MEMBERSHIP IN COMPENDIUM METHOD TO
THE TITLE III CLEAN AIR AMENDMENT LIST--
14A LIST AND THE SOW-CLP LIST OF VOCs
Compound
Methyl chloride (chloromethane); CH3C1
Carbonyl sulfide; COS
Vinyl chloride (chloroethene); C2H3C1
Diazomethane; CH2N2
Formaldehyde; CH2O
1,3-Butadiene; C4H6
Methyl bromide (bro mo methane); CH3Br
Phosgene; CC12O
Vinyl bromide (bromoethene); C2H3Br
Ethylene oxide; C2H4O
Ethyl chloride (chloroethane); C2H5C1
Acetaldehyde (ethanal); C2H4O
Vinylidene chloride (1,1-dichloroethylene); C2H2C12
Propylene oxide; C3H6O
Methyl iodide (iodomethane); CH3I
Methylene chloride; CH2C12
Methyl isocyanate; C2H3NO
Allyl chloride (3-chloropropene); C3H5C1
Carbon disulfide; CS2
Methyl ter-butyl ether; C5H12O
Propionaldehyde; C2H5CHO
Ethylidene dichloride (1,1-dichloroethane); C2H4C12
Chloroprene (2-chloro- 1,3 -butadiene); C4H5C1
CAS No,
74-87-3
463-58-1
75-01-4
334-88-3
50-00-0
106-99-0
74-83-9
75-44-5
593-60-2
75-21-8
75-00-3
75-07-0
75-35-4
75-56-9
74-88-4
75-09-2
624-83-9
107-05-1
75-15-0
1634-04-4
123-38-6
75-34-3
126-99-8
BP fC)1
-23.7
-50.0
-14.0
-23.0
-19.5
-4.5
3.6
8.2
15.8
10.7
12.5
21.0
31.7
34.2
42.4
40.0
59.6
44.5
46.5
55.2
49.0
57.0
59.4
v.p.
(mmHg)1
3.8 xlO3
3.7 xlO3
3.2 xlO3
2.8 x 103
2.7 x 103
2.0 x 103
1.8 xlO3
1.2 xlO3
1.1 x 103
l.lxlO3
1.0 xlO3
952
500
445
400
349
348
340
260
249
235
230
226
MW1
50.5
60.1
62.5
42.1
30
54
94.9
99
107
44
64.5
44
97
58
141.9
84.9
57.1
76.5
76
86
58.1
99
88.5
TO-14A
X
X
X
X
X
X
X
X
CLP-SOW
X
X
X
X
X
X
X
X
O
o
o
o.
H
-------
TABLE 1. (continued)
4-
O
i
•5-
i?
e
Compound
Acrolein (2-propenal); C3H4O
1,2-Epoxy butane (1,2-butylene oxide); C4H8O
Chloroform; CHC13
Ethyleneimine (aziridine); C2H5N
U-Dimethylhydrazine; C2H8N2
Hexane; C6H14
1 ,2-Propyleneimine (2-methylaziridine); C3H7N
Acrylonitrile (2-propenenitrile); C3H3N
Methyl chloroform (1,1,1-trichloroethane); C2H3C13
Methanol; CH4O
Carbon tetrachloride; CC14
Vinyl acetate; C4H6O2
Methyl ethyl ketone (2-butanone); C4H8O
Benzene; C6H6
Acetonitrile (cyanomethane); C2H3N
Ethylene dichloride (1,2-dichloroethane); C2H4C12
Triethylamine; C6H15N
Methylhydrazine; CH6N2
Propylene dichloride (1,2-dichloropropane); C3H6C12
2,2,4-Trimethyl pentane C8H18
1,4-Dioxane (1,4-Diethylene oxide); C4H8O2
Bis(chloromethyl) ether; C2H4C12O
Ethyl acrylate; C5H8O2
Methyl methacrylate-C5H8O2
1,3-Dichloropropene; C3H4C12 (cis)
Toluene; C7H8
CAS No.
107-02-8
106-88-7
67-66-3
151-56-4
57-14-7
110-54-3
75-55-8
107-13-1
71-55-6
67-56-1
56-23-5
108-05-4
78-93-3
71-43-2
75-05-8
107-06-2
121-44-8
60-34-4
78-87-5
540-84-1
123-91-1
542-88-1
140-88-5
80-62-6
542-75-6
108-88-3
BP OC)1
52.5
63.0
61.2
56
63
69.0
66.0
77.3
74.1
65.0
76.7
72.2
79.6
80.1
82
83.5
89.5
87.8
97.0
99.2
101
104
100
101
112
111
(mm&g)1
220
163
160
160.0
157.0
120
112
100
100
92.0
90.0
83.0
77.5
76.0
74.0
61.5
54.0
49.6
42.0
40.6
37.0
30.0
29.3
28.0
27.8
22.0
MW1
56
72
119
43
60.0
86.2
57.1
53
133.4
32
153.8
86
72
78
41.0
99
101.2
46.1
113
114
88
115
100
100.1
111
92
TO-14A
X
X
X
X
X
X
X
X
X
X
CLP-SOW
X
X
X
X
X
X
X
X
X
X
X
X
X
O
O.
H
9
h^
(Jl
O
O
-------
c_
88
88
i
•5-
i?
88
era
(Jl
TABLE 1. (continued)
Compound
Tetrachloroethylene; C2C14
Epichlorohydrin (l-chloro-2,3-epoxy propane); C3H5C1O
Ethylene dibromide (1,2-dibromoethane); C2H4Br2
N-Nitroso-N-methylurea; C2H5N3O2
2-Nitropropane; C3H7NO2
Chlorobenzene; C6H5C1
Ethylbenzene; C8H10
Xylenes (isomer & mixtures); C8H10
Styrene; C8H8
p-Xylene; C8H10
m-Xylene; C8H10
Methyl isobutyl ketone (hexone); C6H12O
Bromoform (tribromomethane); CHBr3
1,1,2,2-Tetrachloroethane; C2H2C14
o-Xylene; C8H10
Dimethylcarbamyl chloride; C3H6C1NO
N-Nitrosodimethylamine; C2H6N2O
Beta-Propiolactone; C3H4O2
Cumene (isopropylbenzene); C9H12
Acrylic acid; C3H4O2
N,N-Dimethylformamide; C3H7NO
1,3-Propane sultone; C3H6O3S
Acetophenone; C8H8O
Dimethyl sulfate; C2H6O4S
CAS No.
127-18-4
106-89-8
106-93-4
684-93-5
79-46-9
108-90-7
100-41-4
1330-20-7
100-42-5
106-42-3
108-38-3
108-10-1
75-25-2
79-34-5
95-47-6
79-44-7
62-75-9
57-57-8
98-82-8
79-10-7
68-12-2
1120-71-4
98-86-2
77-78-1
BP OC)1
121
117
132
124
120
132
136
142
145
138
139
117
149
146
144
166
152
Decomposes at
162
153
141
153
180730mm
202
188
v.p.
(mmgHg)1
14.0
12.0
11.0
10.0
10.0
8.8
7.0
6.7
6.6
6.5
6.0
6.0
5.6
5.0
5.0
4.9
3.7
3.4
3.2
3.2
2.7
2.0
1.0
1.0
MW1
165.8
92.5
187.9
103
89
112.6
106
106.2
104
106.2
106.2
100.2
252.8
167.9
106.2
107.6
74
72
120
72
73
122.1
120
126.1
TO-14A
X
X
X
X
X
X
X
X
X
X
CLP-SOW
X
X
X
X
X
X
X
X
X
X
O
o
o
o.
H
-------
TABLE 1. (continued)
i
•5-
i?
c-
88
e
ss
Compound
Bis(2-Chloroethyl)ether;C4H8C12O
Chloroacetic acid; C2H3C1O2
Aniline (aminobenzene); C6H7N
1,4-Dichlorobenzene (p-); C6H4C12
Ethyl carbamate (urethane); C3H7NO2
Acrylamide; C3H5NO
N,N-Dimethylamlme; C8H1 IN
Hexachloroethane; C2C16
Hexachlorobutadiene; C4C16
Isophorone; C9H14O
N-Nitrosomorpholine; C4H8N2O2
Styrene oxide; C8H8O
Diethyl sulfate; C4H10O4S
Cresylic acid (cresol isomer mixture)
o-Cresol; C7H8O
Catechol (o-hydroxyphenol); C6H6O2
Phenol; C6H6O
1,2,4-Trichlorobenzene; C6H3C13
Nitrobenzene; C6H5NO2
CAS No.
111-44-4
79-11-8
62-53-3
106-46-7
51-79-6
79-06-1
121-69-7
67-72-1
87-68-3
78-59-1
59-89-2
96-09-3
64-67-5
1319-77-3
95-48-7
120-80-9
108-95-2
120-82-1
98-95-3
BP OC)1
178
189
184
173
183
125/25 mm
192
Sublimes at
186
215
215
225
194
208
202
191
240
182
213
211
(mmHg)1
0.71
0.69
0.67
0.60
0.54
0.53
0.50
0.40
0.40
0.38
0.32
0.30
0.29
0.26
0.24
0.22
0.20
0.18
0.15
MW1
143
94.5
93
147
89
71
121
236.7
260.8
138.2
116.1
120.2
154
108
108
110
94
181.5
123
TO-14A
X
X
X
CLP-SOW
X
X
X
O
O.
H
9
h^
(Jl
'Vapor pressure (v.p.), boiling point (BP) and molecular weight (MW) data from:
(a) D. L. Jones and J. bursey, "Simultaneous Control of PM-10 and Hazardous Air Pollutants II: Rationale for Selection of Hazardous Air Pollutants as
Potential Particulate Matter," Report EPA-452/R-93/013, U. S. Environmental Protection Agency, Research Triangle Park, NC. October 1992;
(b) R. C. Weber, P. A. Parker, and M. Bowser. Vapor Pressure Distribution of Selected Organic Chemicals, Report EPA-600/2-81-021, U. S.
Environmental Protection Agency, Cincinnati, OH, February 1981; and
(c) R. C. Weast, ed., "CRC Handbook of Chemistry and Physics," 59th edition, CRC Press, Boca Raton, 1979.
O
O
-------
VOCs
Method TO-15
TABLE 2. CHARACTERISTIC MASSES (M/Z) USED FOR QUANTIFYING
THE TITLE III CLEAN AIR ACT AMENDMENT COMPOUNDS
Compound
Methyl chloride (chloromethane); CH3C1
Carbonyl sulfide; COS
Vinyl chloride (chloroethene); C2H3C1
Diazomethane; CH2N2
Formaldehyde; CH20
l,3-Butadiene;C4H6
Methyl bromide (bromomethane); CH3Br
Phosgene; CC12O
Vinyl bromide (bromoe hene); C2H3Br
Ethylene oxide; C2H4O
Ethyl chloride (chloroethane); C2H5C1
Acetaldehyde (ethanal); C2H4O
Vinylidene chloride (1,1-dichloroethylene); C2H2C12
Propylene oxide; C3H6O
Methyl iodide (iodomethane); CH3I
Methylene chloride; CH2C12
Methyl isocyanate; C2H3NO
Allyl chloride (3-chloropropene); C3H5C1
Carbon disulfide; CS2
Methyl ter-butyl ether; C5H12O
Propionaldehyde; C2H5CHO
Ethylidene dichloride (1,1-dichloroethane); C2H4C12
Chloroprene (2-chloro-l,3-butadiene); C4H5C1
Chloromethyl methyl ether; C2H5C1O
Acrolein (2-propenal); C3H4O
1,2-Epoxy butane (1,2-butylene oxide); C4H8O
Chloroform; CHC13
Ethyleneimine (aziridine); C2H5N
1,1-Dimethylhydrazine; C2H8N2
Hexane; C6H14
1 ,2-Propyleneimine (2-methylazindine); C3H7N
Acrylonitrile (2-propenenitrile); C3H3N
Methyl chloroform (1,1,1 trichloroethane); C2H3C13
Methanol; CH4O
Carbon tetrachloride; CC14
Vinyl acetate; C4H6O2
Methyl ethyl ketone (2-butanone); C4H8O
Benzene; C6H6
Acetonitrile (cyanomethane); C2H3N
CAS No.
74-87-3
463-S8-1
7S-01-4
334-88-3
50-00-0
106-99-0
74-83-9
75-44-5
593-60-2
75-21-8
75-00-3
75-07-0
75-35-4
75-56-9
74-88-4
75-09-2
624-83-9
107-05-1
75-15-0
1634-04-4
123-38-6
75-34-3
126-99-8
107-30-2
107-02-8
106-88-7
67-66-3
151-56-4
57-14-7
110-54-3
75-55-8
107-13-1
71-55-6
67-56-1
56-23-5
108-05-4
78-93-3
71-43-2
75-05-8
Primary Ion
50
60
62
42
29
39
94
63
106
29
64
44
61
58
142
49
57
76
76
73
58
63
88
45
56
42
83
42
60
57
56
53
97
31
117
43
43
78
41
Secondary Ion
52
62
64
41
30
54
96
65
108
44
66
29,43
96
57
127
84,86
56
41,78
44,78
41,53
29,57
65,27
53,90
29,49
55
41,72
85,47
43
45,59
41,43
57,42
52
99,61
29
119
86
72
77,50
40
January 1997
Compendium of Methods for Toxic Organic Air Pollutants
Page 15-43
-------
Method TO-15
VOCs
TABLE 2. (continued)
Compound
Ethylene dichloride (1,2-dichloroethane); C2H4C12
Triethylamine; C6H15N
Methylhydrazine; CH6N2
Propylene dichloride (1,2-dichloropropane); C3H6C12
2,2,4-Trimethyl pentane; C8H18
1,4-Dioxane (1,4 Diethylene oxide); C4H8O2
Bis(chloromethyl) ether; C2H4C12O
Ethyl acrylate; C5H8O2
Methyl methacrylate; C5H8O2
1,3-Dichlompropene; C3H4C12 (cis)
Toluene; C7H8
Trichloethylene; C2HC13
1,1,2-Trichloroethane; C2H3C13
Tetrachloroethylene; C2C14
Epichlorohydrin (l-chloro-2,3-epoxy propane); C3H5C1O
Ethylene dibromide (1,2-dibromoethane); C2H4Br2
N-Nitrso-N-methylurea; C2H5N3O2
2-Nitropropane; C3H7NO2
Chlorobenzene; C6H5C1
Ethylbenzene; C8H10
Xylenes (isomer & mixtures); C8H10
Styrene; C8H8
p-Xylene; C8H10
m-Xylene; C8H10
Methyl isobutyl ketone (hexone); C6H12O
Bromoform (tribromomethane); CHBr3
1,1,2,2-Tetrachloroethane; C2mC14
o-Xylene; C8H10
Dimethylcarbamyl chloride; C3H6C1NO
N-Nitrosodimethylamine; C2H6N2O
Beta-Propiolactone; C3H4O2
Cumene (isopropylbenzene); C9H12
Acrylic acid; C3H4O2
N,N-Dimethylformamide; C3H7NO
1,3-Propane sultone; C3H6O3S
Acetophenone; C8H8O
Dimethyl sulfate; C2H6O4S
Benzyl chloride (a-chlorotoluene); C7H7C1
1 ,2-Dibromo-3-chloropropane; C3H5Br2Cl
Bis(2-Chloroethyl)ether; C4H8C12O
Chloroacetic acid; C2H3C1O2
CAS No.
107-06-2
121-44-8
60-34-4
78-87-5
540-84-1
123-91-1
542-88-1
140-88-5
80-62-6
542-75-6
108-88-3
79-01-6
79-00-5
127-18-4
106-89-8
106-93-4
684-93-5
79-46-9
108-90-7
100-41-4
1330-20-7
100-42-5
106-42-3
108-38-3
108-10-1
75-25-2
79-34-5
95-47-6
79-44-7
62-75-9
57-57-8
98-82-8
79-10-7
68-12-2
1120-71-4
98-86-2
77-78-1
100-44-7
96-12-8
111-44-4
79-11-8
Primary Ion
62
86
46
63
57
88
79
55
41
75
91
130
97
166
57
107
60
43
112
91
91
104
91
91
43
173
83
91
72
74
42
105
72
73
58
105
95
91
57
93
50
Secondary Ion
64,27
58,101
31,45
41,62
41,56
58
49,81
73
69, 100
39,77
92
132,95
83,61
164,131
49,62
109
44, 103
41
77,114
106
106
78, 103
106
106
58, 100
171,175
85
106
107
42
43
120
45,55
42,44
65, 122
77, 120
66,96
126
155,157
63,95
45,60
Page 15-44
Compendium of Methods for Toxic Organic Air Pollutants
January 1997
-------
VOCs
Method TO-15
TABLE 2. (continued)
Compound
Aniline (aminobenzene); C6H7N
1,4-Dichlorobenzene (p-); C6H4C12
Ethyl carbamate (urethane); C3H7NO2
Acrylamide; C3H5NO
N,N-Dimethylaniline; C8H1 IN
Hexachloroethane; C2C16
Hexachlorobutadiene; C4C16
Isophorone; C9H14O
N-Nitrosomorpholine; C4H8N2O2
Styrene oxide; C8H8O
Diethyl sulfate; C4H10O4S
Cresylic acid (cresol isomer mixture)
o-Cresol; C7H8O
Catechol (o-hydroxyphenol); C6H6O2
Phenol; C6H6O
1,2,4-Trichlorobenzene; C6H3C13
Nitrobenzene; C6H5NO2
CAS No.
62-53-3
106-46-7
51-79-6
79-06-1
121-69-7
67-72-1
87-68-3
78-59-1
59-89-2
96-09-3
64-67-5
1319-77-3
95-48-7
120-80-9
108-95-2
120-82-1
98-95-3
Primary Ion
93
146
31
44
120
201
225
82
56
91
45
108
110
94
180
77
Secondary Ion
66
148,111
44,62
55,71
77, 121
199,203
227, 223
138
86,116
120
59,139
107
64
66
182, 184
51,123
January 1997
Compendium of Methods for Toxic Organic Air Pollutants
Page 15-45
-------
Method TO-15
VOCs
TABLE 3. REQUIRED BFB KEY IONS AND
ION ABUNDANCE CRITERIA
Mass
50
75
95
96
173
174
175
176
177
Ion Abundance Criteria1
8.0 to 40.0 Percent of m/e 95
30.0 to 66.0 Percent of m/e 95
Base Peak, 100 Percent Relative Abundance
5.0 to 9.0 Percent of m/e 95 (See note)
Less than 2.0 Percent of m/e 174
50.0 to 120.0 Percent of m/e 95
4.0 to 9.0 Percent of m/e 174
93.0 to 101.0 Percent of m/e 174
5.0 to 9.0 Percent of m/e 176
:A11 ion abundances must be normalized to m/z 95, the
nominal base peak, even though the ion abundance of m/z
174 may be up to 120 percent that of m/z 95.
Page 15-46
Compendium of Methods for Toxic Organic Air Pollutants
January 1997
-------
VOCs
Method TO-15
TABLE 4. METHOD DETECTION LIMITS TMDLV
TO-14A List
Benzene
Benzyl Chloride
Carbon tetrachloride
Chlorobenzene
Chloroform
1 ,3-Dichlorobenzene
1 ,2-Dibromoethane
1 ,4-Dichlorobenzene
1 ,2-Dichlorobenzene
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis-1 ,2-Dichloroethene
Methylene chloride
1 ,2-Dichloropropane
cis-1 ,3-Dichloropropene
trans- 1 ,3-Dichloropropene
Ethylbenzene
Chloroethane
Trichlorofluoromethane
1 ,1 ,2-Trichloro- 1 ,2,2-trifluoroethane
1 ,2-Dichloro-l ,1 ,2,2-tetrafluoroethane
Dichlorodifluoromethane
Hexachlorobutadiene
Bromomethane
Chloromethane
Styrene
1 ,1 ,2,2-Tetrachloroethane
Tetrachloroethene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 ,1 ,2-Trichloroethane
Trichloroethene
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
Vinyl Chloride
m,p-Xylene
o-Xylene
Lab #1, SCAN
0.34
—
0.42
0.34
0.25
0.36
—
0.70
0.44
0.27
0.24
—
—
1.38
0.21
0.36
0.22
0.27
0.19
—
—
—
—
0.53
0.40
1.64
0.28
0.75
0.99
—
0.62
0.50
0.45
—
—
0.33
0.76
0.57
Lab #2, SIM
0.29
—
0.15
0.02
0.07
0.07
0.05
0.12
—
0.05
—
0.22
0.06
0.84
—
—
—
0.05
—
—
—
—
—
—
—
—
0.06
0.09
0.10
0.20
—
0.21
—
0.07
—
—
0.48
0.08
0.28
'Method Detection Limits (MDLs) are defined as the product of the standard
deviation of seven replicate analyses and the student's "t" test value for 99%
confidence. For Lab #2, the MDLs represent an average over four studies.
MDLs are for MS/SCAN for Lab #1 and for MS/SIM for Lab #2.
TABLE 5. SUMMARY OF EPA DATA ON REPLICATE PRECISION (RP)
January 1997
Compendium of Methods for Toxic Organic Air Pollutants
Page 15-47
-------
Method TO-15
VOCs
FROM EPA NETWORK OPERATIONS1
Monitoring Compound
Identification
Dichlorodifluoromethane
Methylene chloride
1 ,2-Dichloroethane
1,1,1 -Trichloroethane
Benzene
Trichloroethene
Toluene
Tetrachloroethene
Chlorobenzene
Ethylbenzene
m-Xylene
Styrene
o-Xylene
p-Xylene
1 ,3 -Dichlorobenzene
1 ,4-Dichlorobenzene
EPA's Urban Air Toxics Monitoring
Program (UATMP)
%RP
16.3
36.2
14.1
12.3
12.8
14.7
36.2
20.3
14.6
14.7
22.8
49.1
14.7
#
07
31
44
56
08
76
12
21
32
75
592
06
14
ppbv
4.3
1.6
1.0
1.6
1.3
3.1
0.8
0.9
0.7
4.0
1.1
0.6
6.5
EPA's Toxics Air Monitoring
Stations (TAMS)
%RP
13.9
19.4
—
10.6
4.4
—
3.4
—
—
5.4
5.3
8.7
6.0
~
#
47
47
—
47
47
—
47
—
—
47
47
47
47
-
ppbv
0.9
0.6
—
2.0
1.5
—
3.1
—
—
0.5
1.5
0.22
0.5
~
'Denotes the number of replicate or duplicate analysis used to generate the statistic. The replicate precision is
defined as the mean ratio of absolute difference to the average value.
2Styrene and o-xylene coelute from the GC column used in UATMP. For the TAMS entries, both values were
below detection limits for 18 of 47 replicates and were not included in the calculation.
TABLE 6. AUDIT ACCURACY (AA) VALUES1 FOR SELECTED
COMPENDIUM METHOD TO-14A COMPOUNDS
Selected Compounds From TO-14A List
Vinyl chloride
Bromomethane
Trichlorofluoromethane
Methylene chloride
Chloroform
1 ,2-Dichloroethane
1,1,1 -Trichloroethane
Benzene
Carbon tetrachloride
1 ,2-Dichloropropane
Trichloroethene
Toluene
Tetrachloroethene
Chlorobenzene
Ethylbenzene
o-Xylene
FY-88 TAMS AA(%), N=30
4.6
—
6.4
8.6
—
6.8
18.6
10.3
12.4
~
8.8
8.3
6.2
10.5
12.4
16.2
FY-88 UATMP AA(%), N=3
17.9
6.4
—
31.4
4.2
11.4
11.3
10.1
9.4
6.2
5.2
12.5
—
11.7
12.4
21.2
1 Audit accuracy is defined as the relative difference between the audit measurement result and its nominal value divided
by the nominal value. N denotes the number of audits averaged to obtain the audit accuracy value. Information is not
available for other TO-14A compounds because they were not present in the audit materials.
Page 15-48
Compendium of Methods for Toxic Organic Air Pollutants
January 1997
-------
VOCs
Method TO-15
To AC
Inlet
Metal Bellows
Type Pump
for Pressurized
Sampling
Vent
To AC
FIGURE 2. SAMPLER CONFIGURATION FOR SUBATMOSPHERIC
Figure 1. Sampler configuration for subatmospheric pressure or pressurized canister sampling.
January 1997
Compendium of Methods for Toxic Organic Air Pollutants
Page 15-49
-------
Method TO-15
VOCs
100K
££\
RED
115 V AC
40ptd, 450 V DC
R2 100K
IMGNELA1CH
SOLENOID
VAIVE
Copocilor Ci ond C2 - «0 uf. 430 VOC (Sprogue Mom TVA 1712 or eqwonenl)
RnWv Ri ond R2 - 0.3 Mil, 5X toMronc*
DMt Di ond 02 - 1000 PRV. 2.5 A (RCA. SK 3081 (X taimoHnl)
(o). Simple Circuit for Operating Magnelatch Volve
FIMER
WITCH
-0^
^
(PU
£QMEC
*«?•
>v
MPl
ttEUK
Recliffe
- 200 PRV
AC
BRIDGE
RECTFIER
AC
1.5 A (RCA <
D
/,
.
/T '
12.7K 2.7K /
RT" 1 R^ /^oc-x,
j. 1 Ci I 7
XlX 200 ul J 1/w
--r-200^ ]£__
20
40
>K 3109 or eqinoieni)
1
^
2
^
C2
I
I
ul
01
NO
RED
BLACK
1 : —
f WHITE
k
rot
N— PTJIAPJZC
MAGNELATCH
SOLENOID
VALVE
n
Copocilor Ci - 200 uT. 290 VOC (Sprogm Atom 1VA 1928 or tquMOMnl)
CapoeiUx C2 - 20 u<, 400 VOC Non-Potorind (Sprogut Atom TVAN 1652 or tawnoHnl)
RMy - 10JX» Ohm CO». « mo (AUF Potltr ond Brumlitld, KCP 9. or •QinOHnl)
RcsMtr Ri ond R} - O.S •Oil. SX loKrone*
(b). Improved Circuit Designed to Handle Power Interruptions
FIGURE 9. ELECTRICAL PULSE_ CIRCUITS FOR DRIVING _
Figure 2. Electrical pulse circuits for driving Skinner magnelatch solenoid valve with
mechanical timer.
Page 15-50
Compendium of Methods for Toxic Organic Air Pollutants
January 1997
-------
VOCs
Method TO-15
Inlet
I Auxilliory
Vacuum
Pump
To AC
FIGURE 3. ALTERNATIVE SAMPLER CONFIGURATION FOR
Figure 3. Alternative sampler configuration for pressurized canister sampling.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants
Page 15-51
-------
Method TO-15
VOCs
778
STAGE 1: SAMPLE TRANSFER TO THE PRECONCENTRATION TRAP
ADSORBENTTRAP
AT NEAR AMBIENT
TEMPERATURE
SAMPLER INLET
^
AIR SAMPLE IN
SAMPLE GAS
FLOW CARRIER
GAS IN
STAGE 2: DRY PURGING
DRY HELIUM
ADSORBENT TRAP
PURGE GAS
rr
AT NEAR AMBIENT
TEMPERATURE
PURGE GAS
PLUS WATER
CARRIER
GAS IN
STAGE 3: TRAP DESORPTION - ANALYTE TRANSFER TO GC COLUMN
CARRIER GAS IN'
ADSORBENTTRAP
(HOT)
Figure 4. Illustration of three stages of dry purging of adsorbent trap.
Page 15-52
Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
(O
o
X
z
D
O
O
OC
Ul
1
[2
oc
5
CD
O
OC
o
X
150 i
140J
130
120
110
I I I I I I I I I I
TEMPERATURE, °C -
1 B45
• 55
A 65
-
I
100ip -
90 [- -
80 U -
I •
f
60 h
50
40
30
20
10
- • -
_ —
A • =
A • . •
"^ ^^ A ^^ ' ^H
A 1 $ -
_ i II II II II II
0 100 200 300 400 500 600 700 800 900 1000 1100
PURGE VOLUME, ml
o
o
Figure 5. Residual water vapor on VOC concentrator vs. dry He purge volume.
n
rt-
o
O.
H
-------
Method TO-15
VOCs
GC
Column
Effluent
Ion
Source/
Filament
Figure 6. Simplified diagram of a quadrupole mass spectrometer.
GC
Column
Effluent
Filament
End Cap
Ring
Electrode
End Cap
Supplementary
rf Voltage
\~1 Electron Multiplier
Figure 7. Simplified diagram of an ion trap mass spectrometer.
Page 15-54
Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
o
o
(o) Reol Time
GC-FID-ECD-PID
or GC-MS
Calibration Cos
Cylinder
Moss Flow
Controller
(0-50 mL/min)
Internal
Baffles
Teflon
Filter
Zero Air
Cylinder
Moss Flow
Controller
(0-50 L/min)
I
V
Vacuum/Pressure
Gauge
Heated Calibration Manifold
Teflon
Filter
Pump
Shut Off
Valve
Flow
Control
Valve
(b) Evacuated or Pressurized
Canister Sampling System
500 ml
Round-Bottom
Flask
(c) Canister Transfer
Standard
Humidifier
Figure 8. Schematic diagram of calibration system and manifold for
(a) analytical system calibration, (b) testing canister sampling system and (c) preparing canister transfer standards.
n
rt-
o
O.
H
-------
Method TO-15
VOCs
B.
COMPENDIUM METHOD TO-15
CANISTER SAMPLING FIELD TEST DATA SHEET
A. GENERAL INFORMATION
SITE LOCATION:
SITE ADDRESS:
SAMPLING DATE:
SAMPLING INFORMATION
TEMPERATURE
SHIPPING DATE:
CANISTER SERIAL NO.:
SAMPLER ID:
OPERATOR:
CANISTER LEAK
CHECK DATE:
START
STOP
INTERIOR
AMBIENT
MAXIMUM
MINIMUM
PRESSURE
CANISTER PRESSURE
SAMPLING TIMES
FLOW RATES
START
STOP
LOCAL TIME
ELAPSED TIME METER
READING
MANIFOLD
FLOW RATE
CANISTER FLOW
RATE
FLOW
CONTROLLER
READOUT
SAMPLING SYSTEM CERTIFICATION DATE:
QUARTERLY RECERTIFICATION DATE:
C.
LABORATORY INFORMATION
DATA RECEIVED:
RECEIVED BY:
INITIAL PRESSURE: _
FINAL PRESSURE:
DILUTION FACTOR: _
ANALYSIS
GC-FID-ECD DATE: _
GC-MSD-SCANDATE:
GC-MSD-SIMDATE: _
RESULTS*:
GC-FID-ECD: _
GC-MSD-SCAN:
GC-MSD-SIM:
SIGNATURE/TITLE
Figure 9. Canister sampling field test data sheet (FTDS).
Page 15-56
Compendium of Methods for Toxic Organic Air Pollutants
January 1997
-------
VOCs
Method TO-15
Vent
Valve
3-Port
Cos
Valve
Pressure
Regulator
Exhaust
Exhaust
Vacuum Pump
Shut Off Valve
^3M*J
x*~*\ V v«n* Sh"t
r ^ A Off Valve
Zero \
Air
Supply
Pressure
Regulator
n
. Dewor
Flosk
Cryogenic
Trap Cooler
(Liquid Argon)
Trap
Cryogenic
Trap Cooler
(Liquid Argon)
Humidifier
Vacuum
Shut Off
Valve
Exhaust
Vacuum
Gauge
Vent
Shut Off
Valve
Vacuum
Gauge
Shut Off
Valve
Zero
Shut Off
Valve
Flow
Control
Valve
Manifold
Optional
Isothermal
Oven
Figure 10. Canister cleaning system.
January 1997
Compendium of Methods for Toxic Organic Air Pollutants
Page 15-57
-------
Method TO-15
VOCs
Pressure
Regulator
Vent
Carrier
Gas
Water Management
System and
Main Preconcentrator
Optional
Pressure
Gauge
Vent
(Excess)
-\
\oJ
(
6-Port
Chromotographic
Valve
Cryogenic
Trapping Unit
I"
I
I
_±_
Flame lonization j
Detector (FID) J
OV-1 Capillary Column
(0.32 mm X 50 m)
1 Low Dead-Volume
Tee (Optional)
I I Flow Restrictor
J ' (Optional)
Mass Spectrometer
in SCAN or SIM Mode
Figure
h 6-port
Page 15-58
Compendium of Methods for Toxic Organic Air Pollutants
January 1997
-------
VOCs
Method TO-15
TIME
(o). Certified Sompler
TIME
(b). Contominoted Sompler
Figure 12. Example of humid zero air test results for a clean sample canister
(a) and a contaminated sample canister (b).
January 1997
Compendium of Methods for Toxic Organic Air Pollutants
Page 15-59
-------
Method TO-15
VOCs
MASS
FLOW
CONT.
PRECONCENTRATOR
(-160° C)
SAMPLE
30 cc/min
VACUUM
PUMP
CRYOTRAP
ON/OFF
VALVES
0.25 cc
LOOP
INSULATED INTERNAL STANDARD
VALVE BOX (45 ± 2° C)
MASS
FLOW n.Bce/mln
INT. STD.
Figure 13. Diagram of design for internal standard addition.
Page 15-60
Compendium of Methods for Toxic Organic Air Pollutants
January 1997
-------
VOCs
Method TO-15
3-HAY VALVE
K
AIR GAUGE /TS
^v*"""
FLOW *
GLASS
SPARGING
VESSEL
FLOWMEfER
2-WAY VALVE
NITROGEN
Figure 14. Water method of standard preparation in canisters.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 15-61
-------
Method TO-15
VOCs
Humidifier
Exhaust
Calibration Zero Air
Gas Cylinder Cylinder
Calibration Manifold i
T 5= Thermocouple
F a Zero Dead Vol. Fit.
FC = Flow Controller
S = Solenoid Valve
Heated Enclosure
To
Auto. Temp.
Control
Figure 15. Diagram of the GC/MS analytical system.
Page 15-62
Compendium of Methods for Toxic Organic Air Pollutants
January 1997
-------
STATUS:
TRAP 1: Sampling
TRAP 2: Desorbing
CLOSED
A °pENtS]
MFC PUMP
MFCF
.i MFC [•
SAMPLE PUMP
SAMPLE INLET
^
CAL/INT STD
^.VENT
PURGE GAS
^*
CAL GAS
INTERNAL STD
g
A..
c
.
I
I
1
_l
-tj^b-
;v-ifl
>v iy
....tOC** f
;v-2l
hT-i i
t-**sl '
iV"3J*-
>V-4l
tXi i
"*&-
1^"^!
l_
1
,
^»v
...X TftS?
^ *
\ fl^»— ^
/**
51^
L_k=%
JGE GAS §AM_PL_E___PURGE_
VENT
SOLID SORBENT CONCENTRATOR
_..i
/
MFC-V
HELIUM
OPEN
OPEN
,••
x*
IlililililililJIil
TO GC/
• » I «M.« *«>•*•• *>»|to
DETECTOR
STIRLING CYCLE COOLER
-------
APPENDIX B
Humidity
-------
Appendix B
Humidity1
Dalton's Law of partial pressures and the hypothesis that water vapor equilibrium above a
canister surface is identical to that established above liquid water can be used to predict the
variation of the percent relative humidity (%RH) of air released from canisters used in ambient
air sampling, typically 6L canisters pressurized with 18L of air. During sampling, water vapor
partial pressure increases as air enters the canister. When (and if) the water vapor partial
pressure exceeds its saturation vapor pressure, the rate of water vapor condensation on the
canister walls equals its sampling rate into the canister. Under constant temperature conditions,
the %RH of air subsequently released from the canister can be calculated. This development
shows that the air released from the canister is initially less humid than the original sample, if the
ambient %RH >33% RH and increases as air is released from the canister according to the
relationship:
%RH = 100% — forV >V
V
where:
Vs = Residual air volume in canister;
Vr = The residual air volume at which water is predicted to be completely
removed (using the assumptions) from the canister wall.
For Vs < Vr, the %RH is constant and equal to its value at Vr. Vr is shown to depend on
the %RH of the ambient air sample. Experimental values are shown to agree reasonably well
'McClenny, W.A., S.M. Schmidt, and K.G. Kronmiller. "Variation of the Relative
Humidity of Air Released from Canisters After Ambient Sampling." In Proceedings of the
Measurement of Toxic and Related Air Pollutants International Symposium, Research Triangle
Park,NC, 1997.
B-l
-------
with predictions. However, experimental values were systematically lower than predicted
especially when ambient air with mid-range %RH was sampled. The difference appears to be
related to the mass of water vapor condensed on the sampling apparatus upstream of the canister.
Near Vr, some effects due to strongly adsorbed water vapor may also be present.
Note: The results of mathematical prediction of the %RH of air released from a
canister are summarized below and graphical information is provided to allow predictions
to be made for a variety of situations.
Three ranges of %RH of the ambient air that is initially sampled into the canister are
conveniently treated:
0%RH < original ambient %RH < 33.3%RH
No condensation occurs and (neglecting any adsorbed water vapor) the relative
humidity of air released from the canister should be constant at its original
ambient air value although in practice some water vapor will be adsorbed on the
canister wall.
33.3%RH < original ambient %RH < 70%RH
Locate the ambient sample %RH on the abscissa (the value of 60%RH was chosen
in Figure B-l as an example) and the point of intersection of this value with the
curve A.
Identify the value of canister volume on the ordinate scale corresponding to the
point of intersection (8L in the example). This value is the volume Vr at which all
condensed water vapor on the canister walls has been evaporated during the
process of releasing the sample air from the canister.
Use Vr to locate a point of intersection on curve B. This point Vr = Vs divides the
curve B into two sections, (1) Vs < Vr and (2) Vs > Vr.
Note: In general (for any ambient %RH), as the sample is released from the
canister (at any stage having a volume Vs remaining in the canister), the
curve B indicates the %RH (on the abscissa scale) of the released sample air
over the range 18L < Vs < 6L.
B-2
-------
'(0
a.
(0
(0
O
Figure B-1. A - Original %RH of the sample air vs the remaining sample
volume, Vr, in the canister when all condensed water is
evaporated; B - residual sample volume, Vs vs %RH of air
released from canister.
-------
For the example, for Vs < 8L, since all condensed water in the canister will have
been evaporated, the mixture of water vapor and air will be constant and the %RH
of the remaining sample air will be released at its value at Vs = 8L. In the example
this value if 74%RH.
For Vs > Vr, the abscissa value (obtained with curve B) corresponding to Vs
indicates the predicted %RH value of released air. In the example, this applies for
any remaining sample volume between 18L and 8L.
Figure B-2 shows the modification of Figure B-l (curve B) to predict the %RH of
released sample air for the example (when the %RH of the ambient air sample
was 60%). To do this locate the ordinate value corresponding to the volume Vs
remaining in the canister [between 18L (SOpsig) and 6L (ISpsig)] and read the
abscissa value of %RH for the point of intersection of Vs and the curve.
70%RH < original ambient %RH < 100%RH
Refer to Figure B-l and note from curve A that Vr < 6L which is equivalent to
stating that some condensed water is always on the canister wall for 18L < Vs
< 6L. The %RH of released air is determined by using curve B in Figure B-l to
find the value of %RH corresponding to any residual volume Vs over the entire
range of values of Vs > 6L.
Results of Experimental Determination of the %RH of Sample Air Released
from Pressurized Canisters—Figures B-3 and B-4 show the %RH of air released from
canisters initially pressurized to 18L when 34%, 61% and 90%RH (23 ± 3°C) ambient air
samples were made available to the sampling manifold. The predicted values are shown for
comparison. Some of the difference between experimental and predicted values appears to be
due to condensation of water vapor on the sampling apparatus. This condensation causes a
systematic displacement of the entire set of experimental points such as seen in Figure B-4.
The curves in Figure B-4 correspond to a 61% Relative Humidity value for ambient air
and involve both water vapor condensation followed by evaporation of available water from the
canister wall as the sample air is released. Two experimental approaches (Can 1 and Can 2) are
shown. The two differ as the sample value remaining approaches Vr. The Can 2 values show the
general features of the predicted characteristic. However, the value of Vr (constant % Relative
Humidity) occurs at a higher value (58% Relative Humidity) than predicted (76%), probably due
B-4
-------
18
30
td
-------
td
s Remaining i
> c
Sample Volumi
> N:
•
\
^ 4
\
!4% RH
V
Note:
Arrows i
of Ambk
*^
^
ndicate
jnt Sam
E:
P
^5<
%RH
pie
90C
/oRH
i
30
15
.5*
'55
3
(0
(0
p
CO
O
30 40 50 60 70 80 90
% Relative Humidity (From Can)
100
Figure B-3. Comparison of Predicted and Experimental %RH Values of Released Air Versus Volume of Sample
Remaining in Canister; 34% RH and 90% RH Ambient Air Sample
B-6
-------
td
5 Remaining (L
^ _
J 0
(ample Volume
•> N
°*\
o ^
a
i \^
O N.
o
Qm
m
m
X
o
Note:
Arrows indice
of Ambient S
O »
•\
O
(
3
W O
30 40 50 60 70
te % RH
ample
. Can 1
. Can 2
» Predicted
-
Vr
ou
35
A f- ~*
low
(/)
£
Q_
C
(0
O
n
80 90 100
% Relative Humidity (From Can)
Figure B-4. Comparison of Predicted and Experimental %RH Values of Released Air Versus Volume of Sample
Remaining in Canister; 61% RH Ambient Air Sample
-------
to the condensation of water vapor on the inlet lines during sampling. Can 1 shows a
monotonically increasing value of % Relative Humidity as residual canister volume decreases.
Recent experimental work indicates that different canisters exhibit different behaviors near Vr.
Additional experimental work is being carried out to investigate whether these differences may
be related to the condition of the interior surface of the canister.
Adjustments must be made to the predicted values when ambient conditions of
temperature change appreciably between sampling and release of air. Obviously, more or less
water is condensed in the canister when the ambient temperature at which the canister is held
becomes lower or higher than the temperature at which the sample was taken. Consideration
should also be given to the mass of water in the canister since the condensation of water in the
canister for the same %RH but different temperatures may lead to droplets with various surface
to volume ratios. Another factor that could make a difference in the response profile of %RH
versus canister pressure is the manner in which water is introduced into the canister. Water added
to synthetic samples for humidification by using a certain number of |iL probably has a different
surface distribution in the canister than humidified samples introduced directly from the ambient
air.
-------
APPENDIX C
Method TO-12
Method for the Determination of Non-Methane Organic Compounds (NMOC) in
Ambient Air Using Cryogenic Preconcentration and Direct Flame lonization
Detection (PDFID)
-------
METHOD TO-12
METHOD FOR THE DETERMINATION OF NON-METHANE ORGANIC COMPOUNDS (NMOC)
IN AMBIENT AIR USING CRYOGENIC PRECONCENTRATION AND DIRECT FLAME
IONIZATION DETECTION (PDFID)
1. Scope
1.1 In recent years, the relationship between ambient concentrations
of precursor organic compounds and subsequent downwind
concentrations of ozone has been described by a variety of
photochemical dispersion models. The most important application
of such models is to determine the degree of control of precursor
organic compounds that is necessary in an urban area to achieve
compliance with applicable ambient air quality standards for ozone
(1,2).
1.2 The more elaborate theoretical models generally require detailed
organic species data obtained by multicomponent gas chromatography
(3). The Empirical Kinetic Modeling Approach (EKMA), however,
requires only the total non-methane organic compound (NMOC)
concentration data; specifically, the average total NMOC
concentration from 6 a.m. to 9 a.m. daily at the sampling
location. The use of total NMOC concentration data in the EKMA
substantially reduces the cost and complexity of the sampling and
analysis system by not requiring qualitative and quantitative
species identification.
1.3 Method T01, "Method for The Determination of Volatile Organic
Compounds in Ambient Air Using Tenax® Adsorption and Gas
Chromatography/Mass Spectrometry (GC/MS)", employs collection of
certain volatile organic compounds on Tenax® GC with subsequent
analysis of thermal desorption/cryogenic preconcentration and
GC/MS identification. This method (T012) combines the same type
of cryogenic concentration techniques used in Method T01 for high
sensitivity with the simple flame ionization detector (FID) of the
GC for total NMOC measurements, without the GC columns and complex
procedures necessary for species separation.
1.4 In a flame ionization detector, the sample is injected into a
hydrogen-rich flame where the organic vapors burn producing
ionized molecular fragments. The resulting ion fragments are then
collected and detected. The FID is nearly a universal detector.
However, the detector response varies with the species of
[functional group in] the organic compound in an oxygen
atmosphere. Because this method employs a helium or argon carrier
gas, the detector response is nearly one for all compounds. Thus,
-------
T012-2
the historical short-coming of the FID involving varying detector
response to different organic functional groups is minimized.
1.5 The method can be used either for direct, in situ ambient
measurements or (more commonly) for analysis of integrated samples
collected in specially treated stainless steel canisters. EKMA
models generally require 3-hour integrated NMOC measurements over
the 6 a.m. to 9 a.m. period and are used by State or local
agencies to prepare State Implementation Plans (SIPs) for ozone
control to achieve compliance with the National Ambient Air
Quality Standards (NAAQS) for ozone. For direct, in situ ambient
measurements, the analyst must be present during the 6 a.m. to 9
a.m. period, and repeat measurements (approximately six per hour)
must be taken to obtain the 6 a.m. to 9 a.m. average NMOC
concentration. The use of sample canisters allows the collection
of integrated air samples over the 6 a.m. to 9 a.m. period by
unattended, automated samplers. This method has incorporated both
sampling approaches.
Applicable Documents
2.1 ASTM Standards
D1356 - Definition of Terms Related to Atmospheric Sampling
and Analysis
E260 - Recommended Practice for General Gas Chromatography
Procedures
E355 - Practice for Gas Chromatography Terms and
Relationships
Other Documents
U. S. Environmental Protection Agency Technical Assistance
Documents (4,5)
Laboratory and Ambient Air Studies (6-10)
Summary of Method
3.1 A whole air sample is either extracted directly from the ambient
air and analyzed on site by the GC system or collected into a
precleaned sample canister and analyzed off site.
The analysis requires drawing a fixed-volume portion of the sample
air at a low flow rate through a glass-bead filled trap that is
cooled to approximately -186°C with liquid argon. The cryogenic
trap simultaneously collects and concentrates the NMOC (either via
condensation or adsorption) while allowing the methane, nitrogen,
oxygen, etc. to pass through the trap without retention. The
system is dynamically calibrated so that the volume of sample
-------
T012-3
passing through the trap does not have to be quantitatively
measured, but must be precisely repeatable between the calibration
and the analytical phases.
3.3 After the fixed-volume air sample has been drawn through the trap,
a helium carrier gas flow is diverted to pass through the trap, in
the opposite direction to the sample flow, and into an FID. When
the residual air and methane have been flushed from the trap and
the FID baseline restablizes, the cryogen is removed and the
temperature of the trap is raised to approximately 90°C.
3.4 The organic compounds previously collected in the trap revola-
tilize due to the increase in temperature and are carried into the
FID, resulting in a response peak or peaks from the FID. The area
of the peak or peaks is integrated, and the integrated value is
translated to concentration units via a previously-obtained
calibration curve relating integrated peak areas with known
concentrations of propane.
3.5 By convention, concentrations of NMOC are reported in units of
parts per million carbon (ppmC), which, for a specified compound,
is the concentration of volume (ppmV) multiplied by the number of
carbon atoms in the compound.
3.6 The cryogenic trap simultaneously concentrates the NMOC while
separating and removing the methane from air samples. The
technique is thus direct reading for NMOC and, because of the
concentration step, is more sensitive than conventional continuous
NMOC analyzers.
Significance
4.1 Accurate measurements of ambient concentrations of NMOC are
important for the control of photochemical smog because these
organic compounds are primary precursors of atmospheric ozone and
other oxidants. Achieving and maintaining compliance with the
NAAQS for ozone thus depends largely on control of ambient levels
of NMOC.
4.2 The NMOC concentrations typically found at urban sites may range
up to 5-7 ppmC or higher. In order to determine transport of
precursors into an area, measurement of NMOC upwind of the area
may be necessary. Upwind NMOC concentrations are likely to be
less than a few tenths of 1 ppm.
4.3 Conventional methods that depend on gas chromatography and
qualitative and quantitative species evaluation are excessively
difficult and expensive to operate and maintain when speciated
measurements are not needed. The method described here involves a
-------
T012-4
simple, cryogenic preconcentration procedure with subsequent,
direct, flame ionization detection. The method is sensitive and
provides accurate measurements of ambient NMOC concentrations
where speciated data are not required as applicable to the EKMA.
Definitions
[Note: Definitions used in this document and in any user-prepared
Standard Operating Procedures (SOPs) should be consistent with
ASTM Methods D1356 and E355. All abbreviations and symbols are
defined within this document at point of use.]
5.1 Absolute pressure - Pressure measured with reference to absolute
zero pressure (as opposed to atmospheric pressure), usually
expressed as pounds-force per square inch absolute (psia).
5.2 Cryogen - A substance used to obtain very low trap temperatures in
the NMOC analysis system. Typical cryogens are liquid argon
(bp-185.7) and liquid oxygen (bp-183.0).
5.3 Dynamic calibration - Calibration of an analytical system with
pollutant concentrations that are generated in a dynamic, flowing
system, such as by quantitative, flow-rate dilution of a high
concentration gas standard with zero gas.
5.4 EKMA - Empirical Kinetics Modeling Approach; an empirical model
that attempts to relate morning ambient concentrations of non-
methane organic compounds (NMOC) and NOX with subsequent peak,
downwind ambient ozone concentrations; used by pollution control
agencies to estimate the degree of hydrocarbon emission reduction
needed to achieve compliance with national ambient air quality
standards for ozone.
5.5 Gauge pressure - Pressure measured with reference to atmospheric
pressure (as opposed to absolute pressure). Zero gauge pressure
(0 psig) is equal to atmospheric pressure, or 14.7 psia (101 kPa).
5.6 In situ - In place; In situ measurements are obtained by direct,
on-the-spot analysis, as opposed to subsequent, remote analysis of
a collected sample.
5.7 Integrated sample - A sample obtained uniformly over a specified
time period and representative of the average levels of pollutants
during the time period.
5.8 NMOC - Nonmethane organic compounds; total organic compounds as
measured by a flame ionization detector, excluding methane.
5.9 ppmC - Concentration unit of parts per million carbon; for a
specific compound, ppmC is equivalent to parts per million by
volume (ppmv) multiplied by the number of carbon atoms in the
compound.
-------
T012-5
5.10 Sampling - The process of withdrawing or isolating a repre-
sentative portion of an ambient atmosphere, with or without the
simultaneous isolation of selected components for subsequent
analysis.
Interferences
6.1 In field and laboratory evaluation, water was found to cause a
positive shift in the FID baseline. The effect of this shift is
minimized by carefully selecting the integration termination point
and adjusted baseline used for calculating the area of the NMOC
peak(s).
6.2 When using helium as a carrier gas, FID response is quite uniform
for most hydrocarbon compounds, but the response can vary
considerably for other types of organic compounds.
Apparatus
7.1 Direct Air Sampling (Figure 1)
7.1.1 Sample manifold or sample inlet line - to bring sample
air into the analytical system.
7.1.2 Vacuum pump or blower - to draw sample air through a
sample manifold or long inlet line to reduce inlet
residence time. Maximum residence time should be no
greater than 1 minute.
7.2 Remote Sample Collection in Pressurized Canisters (Figure 2)
7.2.1 Sample canister(s) - stainless steel, Summa®-polished
vessel(s) of 4-6 L capacity (Scientific Instru-
mentation Specialists, Inc., P.O. Box 8941, Moscow, ID
83843), used for automatic collection of 3-hour
integrated field air samples. Each canister should
have a unique identification number stamped on its
frame.
7.2.2 Sample pump - stainless steel, metal bellows type
(Model MB-151, Metal Bellows Corp., 1075 Providence
Highway, Sharon, MA 02067) capable of 2 atmospheres
minimum output pressure. Pump must be free of leaks,
clean, and uncontaminated by oil or organic compounds.
7.2.3 Pressure gauge - 0-30 psig (0-240 kPa).
7.2.4 Solenoid valve - special electrically-operated,
bistable solenoid valve (Skinner Magnelatch Valve, New
Britain, CT), to control sample flow to the canister
with negligible temperature rise (Figure 3). The use
of the Skinner Magnelatch valve avoids any substantial
temperature rise that would occur with a conventional,
-------
T012-6
normally closed solenoid valve, which would have to be
energized during the entire sample period. This
temperature rise in the valve could cause outgasing of
organics from the Viton valve seat material. The
Skinner Magnelatch valve requires only a brief
electrical pulse to open or close at the appropriate
start and stop times and therefore experiences no
temperature increase. The pulses may be obtained with
an electronic timer that can be programmed for short
(5 to 60 seconds) ON periods or with a conventional
mechanical timer and a special pulse circuit. Figure
3[a] illustrates a simple electrical pulse circuit for
operating the Skinner Magnelatch solenoid valve with a
conventional mechanical timer. However, with this
simple circuit, the valve may operate unpredictably
during brief power interruptions or if the time is
manually switched on and off too fast. A better
circuit incorporating a time-delay relay to provide
more reliable valve operation is shown in Figure 3[b].
7.2.5 Stainless steel orifice (or short capillary) - capable
of maintaining a substantially constant flow over the
sampling period (see Figure 4) .
7.2.6 Particulate matter filter - 2 micron stainless steel
sintered in-line type (see Figure 4).
7.2.7 Timer - used for unattended sample collection.
Capable of controlling pump(s) and solenoid valve.
7.3 Sample Canister Cleaning (Figure 5)
7.3.1 Vacuum pump - capable of evacuating sample canister(s)
to an absolute pressure of <5 mm Hg.
7.3.2 Manifold - stainless steel manifold with connections
for simultaneously cleaning several canisters.
7.3.3 Shut off valve(s) - seven required.
7.3.4 Vacuum gauge - capable of measuring vacuum in the
manifold to an absolute pressure of 5 mm Hg or less.
7.3.5 Cryogenic trap (2 required) - U-shaped open tubular
trap cooled with liquid nitrogen or argon used to
prevent contamination from back diffusion of oil from
vacuum pump, and to provide clean, zero air to sample
canister(s).
7.3.6 Pressure gauge - 0-50 psig (0-345 kPa), to monitor
zero air pressure.
-------
T012-7
7.3.7 Flow control valve - to regulate flow of zero air into
canister(s).
7.3.8 Humidifier - water bubbler or other system capable of
providing moisture to the zero air supply.
7.4 Analytical System (Figure 1)
7.4.1 FID detector system - including flow controls for the
FID fuel and air, temperature control for the FID, and
signal processing electronics. The FID burner air,
hydrogen, and helium carrier flow rates should be set
according to the manufacturer's instructions to obtain
an adequate FID response while maintaining as stable a
flame as possible throughout all phases of the
analytical cycle.
7.4.2 Chart recorder - compatible with the FID output
signal, to record FID response.
7.4.3 Electronic integrator - capable of integrating the
area of one or more FID response peaks and calculating
peak area corrected for baseline drift. If a separate
integrator and chart recorder are used, care must be
exercised to be sure that these components do not
interfere with each other electrically. Range
selector controls on both the integrator and the FID
analyzer may not provide accurate range ratios, so
individual calibration curves should be prepared for
each range to be used. The integrator should be
capable of marking the beginning and ending of peaks,
constructing the appropriate baseline between the
start and end of the integration period, and
calculating the peak area.
Note: The FID (7.4.1), chart recorder (7.4.2),
integrator (7.4.3), valve heater (7.4.5), and a trap
heating system are conveniently provided by a standard
laboratory chromatograph and associated integrator.
EPA has adapted two such systems for the PDFID method:
a Hewlett-Packard model 5880 (Hewlett-Packard Corp.,
Avondale, PA) and a Shimadzu model GC8APF (Shimadzu
Scientific Instruments Inc., Columbia, MD; see
Reference 5). Other similar systems may also be
applicable.
7.4.4 Trap - the trap should be carefully constructed from a
single piece of chromatographic-grade stainless steel
-------
T012-8
tubing (0.32 cm O.D, 0.21 cm I.D.) as shown in Figure
6. The central portion of the trap (7-10 cm) is
packed with 60/80 mesh glass beads, with small glass
wool (dimethyldichlorosilane-treated) plugs to retain
the beads. The trap must fit conveniently into the
Dewar flask (7.4.9), and the arms must be of an
appropriate length to allow the beaded portion of the
trap to be submerged below the level of liquid cryogen
in the Dewar. The trap should connect directly to the
six-port valve, if possible, to minimize line length
between the trap and the FID. The trap must be
mounted to allow the Dewar to be slipped conveniently
on and off the trap and also to facilitate heating of
the trap (see 7 . 4 . 13) .
7.4.5 Six-port chromatographic valve - Seiscor Model VIII
(Seismograph Service Corp., Tulsa, OK), Valco Model
9110 (Valco Instruments Co., Houston, TX), or
equivalent. The six-port valve and as much of the
interconnecting tubing as practical should be located
inside an oven or otherwise heated to 80 - 90°C to
minimize wall losses or adsorption/desorption in the
connecting tubing. All lines should be as short as
practical.
7.4.6 Multistage pressure regulators - standard two-stage,
stainless steel diaphram regulators with pressure
gauges, for helium, air, and hydrogen cylinders.
7.4.7 Pressure regulators - optional single stage, stainless
steel, with pressure gauge, if needed, to maintain
constant helium carrier and hydrogen flow rates.
7.4.8 Fine needle valve - to adjust sample flow rate through
trap.
7.4.9 Dewar flask - to hold liquid cryogen to cool the trap,
sized to contain submerged portion of trap.
7.4.10 Absolute pressure gauge - 0-450 mm Hg, (2 mm Hg [scale
divisions indicating units]), to monitor repeatable
volumes of sample air through cryogenic trap (Wallace
and Tiernan, Model 61C-ID-0410, 25 Main Street,
Belleville, NJ).
7.4.11 Vacuum reservoir - 1-2 L capacity, typically 1 L.
7.4.12 Gas purifiers - gas scrubbers containing Drierite® or
silica gel and 5A molecular sieve to remove moisture
-------
T012-9
and organic impurities in the helium, air, and
hydrogen gas flows (Alltech Associates, Deerfield,
IL). Note: Check purity of gas purifiers prior to use
by passing zero-air through the unit and analyzing
according to Section 11.4. Gas purifiers are clean
if produce [contain] less than 0.02 ppmC hydrocarbons.
7.4.13 Trap heating system - chromatographic oven, hot water,
or other means to heat the trap to 80° to 90°C. A
simple heating source for the trap is a beaker or
Dewar filled with water maintained at 80-90°C. More
repeatable types of heat sources are recommended,
including a temperature-programmed chromatograph oven,
electrical heating of the trap itself, or any type of
heater that brings the temperature of the trap up to
80-90°C in 1-2 minutes.
7.4.14 Toggle shut-off valves (2) - leak free, for vacuum
valve and sample valve.
7.4.15 Vacuum pump - general purpose laboratory pump capable
of evacuating the vacuum reservoir to an appropriate
vacuum that allows the desired sample volume to be
drawn through the trap.
7.4.16 Vent - to keep the trap at atmospheric pressure during
trapping when using pressurized canisters.
7.4.17 Rotameter - to verify vent flow.
7.4.18 Fine needle valve (optional) - to adjust flow rate of
sample from canister during analysis.
7.4.19 Chromatographic-grade stainless steel tubing (Alltech
Applied Science, 2051 Waukegan Road, Deerfield, IL,
60015, (312) 948-8600) and stainless steel plumbing
fittings - for interconnections. All such materials
in contact with the sample, analyte, or support gases
prior to analysis should be stainless steel or other
inert metal. Do not use plastic or Teflon® tubing or
fittings.
7.5 Commercially Available PDFID System (5)
7.5.1 A convenient and cost-effective modular PDFID system
suitable for use with a conventional laboratory
chromatograph is commercially available (NuTech
Corporation, Model 8548, 2806 Cheek Road, Durham, NC,
27704, (919) 682-0402).
-------
T012-10
7.5.2 This modular system contains almost all of the
apparatus items needed to convert the chromatograph
into a PDFID analytical system and has been designed
to be readily available and easy to assemble.
Reagents and Materials
Gas cylinders of helium and hydrogen - ultrahigh purity grade.
Combustion air - cylinder containing less than 0.02 ppm
hydrocarbons, or equivalent air source.
8.3 Propane calibration standard - cylinder containing 1-100 ppm (3-
300 ppmC) propane in air. The cylinder assay should be traceable
to a National Bureau of Standards (NBS) Standard Reference
Material (SRM) or to a NBS/EPA-approved Certified Reference
Material (CRM).
8.4 Zero air - cylinder containing less than 0.02 ppmC hydrocarbons.
Zero air may be obtained from a cylinder of zero-grade compressed
air scrubbed with Drierite® or silica gel and 5A molecular sieve
or activated charcoal, or by catalytic cleanup of ambient air.
All zero air should be passed through a liquid argon cold trap for
final cleanup, then passed through a hydrocarbon-free water
bubbler (or other device) for humidification.
8.5 Liquid cryogen - liquid argon (bp -185.7°C) or liquid oxygen, (bp
-183°C) may be used as the cryogen. Experiments have shown no
differences in trapping efficiency between liquid argon and liquid
oxygen. However, appropriate safety precautions must be taken if
liquid oxygen is used. Liquid nitrogen (bp -195°C) should not be
used because it causes condensation of oxygen and methane in the
trap.
9. Direct Sampling
9.1 For direct ambient air sampling, the cryogenic trapping system
draws the air sample directly from a pump-ventilated distribution
manifold or sample line (see Figure 1). The connecting line
should be of small diameter (1/8" O.D.) stainless steel tubing and
as short as possible to minimize its dead volume.
9.2 Multiple analyses over the sampling period must be made to
establish hourly or 3-hour NMOC concentration averages.
10. Sample Collection in Pressurized Canister(s)
For integrated pressurized canister sampling, ambient air is sampled by
a metal bellows pump through a critical orifice (to maintain constant
flow), and pressurized into a clean, evacuated, Summa®-polished sample
canister. The critical orifice size is chosen so that the canister is
pressurized to approximately one atmosphere above ambient pressure, at a
-------
T012-11
constant flow rate over the desired sample period. Two canisters are
connected in parallel for duplicate samples. The canister(s) are then
returned to the laboratory for analysis, using the PDFID analytical
system. Collection of ambient air samples in pressurized canisters
provides the following advantages:
! Convenient integration of ambient samples over a specific
time period
! Capability of remote sampling with subsequent central
laboratory analysis
! Ability to ship and store samples, if necessary
! Unattended sample collection
! Analysis of samples from multiple sites with one analytical
system
! Collection of replicate samples for assessment of
measurement precision
With canister sampling, however, great care must be exercised in
selecting, cleaning, and handling the sample canister (s) and sampling
apparatus to avoid losses or contamination of the samples.
10.1 Canister Cleanup and Preparation
10.1.1 All canisters must be clean and free of any
contaminants before sample collection.
10.1.2 Leak test all canisters by pressurizing them to
approximately 30 psig [200 kPa (gauge)] with zero air.
The use of the canister cleaning system (see Figure 5)
may be adequate for this task. Measure the final
pressure - close the canister valve, then check the
pressure after 24 hours. If leak tight, the pressure
should not vary more than +_ 2 psig over the 24-hour
period. Note leak check result on sampling data
sheet, Figure 7.
10.1.3 Assemble a canister cleaning system, as illustrated in
Figure 5. Add cryogen to both the vacuum pump and
zero air supply traps. Connect the canister(s) to the
manifold. Open the vent shut off valve and the
canister valve(s) to release any remaining pressure in
the canister. Now close the vent shut off valve and
open the vacuum shut off valve. Start the vacuum pump
and evacuate the canister(s) to < 5.0 mm Hg (for at
least one hour). [Note: On a daily basis or more
often if necessary, blow-out the cryogenic traps with
zero air to remove any trapped water from previous
canister cleaning cycles.]
10.1.4 Close the vacuum and vacuum gauge shut off valves and
open the zero air shut off valve to pressurize the
-------
T012-12
canister(s) with moist zero air to approximately 30
psig [200 kPa (gauge)]. If a zero gas generator
systems is used, the flow rate may need to be limited
to maintain the zero air quality.
10.1.5 Close the zero shut off valve and allow canister(s) to
vent down to atmospheric pressure through the vent
shut off valve. Close the vent shut off valve.
Repeat steps 10.1.3 through 10.1.5 two additional
times for a total of three (3) evacuation/
pressurization cycles for each set of canisters.
10.1.6 As a "blank" check of the canister(s) and cleanup
procedure, analyze the final zero-air fill of 100% of
the canisters until the cleanup system and canisters
are proven reliable. The check can then be reduced to
a lower percentage of canisters. Any canister that
does not test clean (compared to direct analysis of
humidified zero air of less than 0.02 ppmC) should not
be utilized.
10.1.7 The canister is then re-evacuated to < 5.0 mm Hg,
using the canister cleaning system, and remains in
this condition until use. Close the canister valve,
remove the canister from the canister cleaning system
and cap canister connection with a stainless steel
fitting. The canister is now ready for collection of
an air sample. Attach an identification tag to the
neck of each canister for field notes and chain-of-
custody purposes.
10.2 Collection of Integrated Whole-Air Samples
10.2.1 Assemble the sampling apparatus as shown in Figure 2.
The connecting lines between the sample pump and the
canister(s) should be as short as possible to minimize
their volume. A second canister is used when a
duplicate sample is desired for quality assurance (QA)
purposes (see Section 12.2.4) . The small auxiliary
vacuum pump purges the inlet manifold or lines with a
flow of several L/min to minimize the sample residence
time. The larger metal bellows pump takes a small
portion of this sample to fill and pressurize the
sample canister(s). Both pumps should be shock-
mounted to minimize vibration. Prior to field use,
each sampling system should be leak tested. The
-------
T012-13
outlet side of the metal bellows pump can be checked
for leaks by attaching the 0-30 psig pressure gauge to
the canister (s) inlet via connecting tubing and
pressurizing to 2 atmospheres or approximately 29.4
psig. If pump and connecting lines are leak free
pressure should remain at +2 psig for 15 minutes. To
check the inlet side, plug the sample inlet and insure
that there is no flow at the outlet of the pump.
10.2.2 Calculate the flow rate needed so that the canister(s)
are pressurized to approximately one atmosphere above
ambient pressure (2 atmospheres absolute pressure)
over the desired sample period, utilizing the
following equation:
F _
00(60)
where:
F = flow rate (cmVmin)
P = final canister pressure (atmospheres absolute)
= (Pg/Pa) + 1
V = volume of the canister (cm3)
N = number of canisters connected together for
simultaneous sample collection
T = sample period (hours)
Pg = gauge pressure in canister, psig (kPa)
Pa = standard atmospheric pressure, 14.7 psig
(101 kPa)
For example, if one 6-L canister is to be filled to 2
atmospheres absolute pressure (14.7 psig) in 3 hours, the
flow rate would be calculated as follows:
„ 2 x 6000 x 1 ,_ 3. .
F = = 67 cm /mm
3 x 60
10.2.3 Select a critical orifice or hypodermic needle
suitable to maintain a substantially constant flow at
the calculated flow rate into the canister(s) over the
desired sample period. A 30-gauge hypodermic needle,
-------
T012-14
2.5 cm long, provides a flow of approximately 65
cmVmin with the Metal Bellows Model MBV-151 pump (see
Figure 4). Such a needle will maintain approximately
constant flow up to a canister pressure of about 10
psig (71 kPa), after which the flow drops with
increasing pressure. At 14.7 psig (2 atmospheres
absolute pressure), the flow is about 10% below the
original flow.
10.2.4 Assemble the 2.0 micron stainless steel in-line
particulate filter and position it in front of the
critical orifice. A suggested filter-hypodermic
needle assembly can be fabricated as illustrated in
Figure 4.
10.2.5 Check the sampling system for contamination by filling
two evacuated, cleaned canister(s) (See Section 10.1)
with humidified zero air through the sampling system.
Analyze the canisters according to Section 11.4. The
sampling system is free of contamination if the
canisters contain less than 0.02 ppmC hydrocarbons,
similar to that of humidified zero air.
10.2.6 During the system contamination check procedure, check
the critical orifice flow rate on the sampling system
to insure that sample flow rate remains relatively
constant (±10%) up to about 2 atmospheres absolute
pressure (101 kpa). Note: A drop in the flow rate
may occur near the end of the sampling period as the
canister pressure approaches two atmospheres.
10.2.7 Reassemble the sampling system. If the inlet sample
line is longer than 3 meters, install an auxiliary
pump to ventilate the sample line, as illustrated in
Figure 2.
10.2.8 Verify that the timer, pump(s) and solenoid valve are
connected and operating properly.
10.2.9 Verify that the timer is correctly set for the desired
sample period, and that the solenoid valve is closed.
10.2.10 Connect a cleaned, evacuated canister(s) (Section
10.1) to the non-contaminated sampling system, by way
of the solenoid valve, for sample collection.
10.2.11 Make sure the solenoid valve is closed. Open the
canister valve(s). Temporarily connect a small
rotameter to the sample inlet to verify that there is
-------
T012-15
no flow. Note: Flow detection would indicate a
leaking (or open) solenoid valve. Remove the
rotameter after leak detection procedure.
10.2.12 Fill out the necessary information on the Field Data
Sheet (Figure 7).
10.2.13 Set the automatic timer to start and stop the pump or
pumps to open and close the solenoid valve at the
appropriate time for the intended sample period.
Sampling will begin at the pre-determined time.
10.2.14 After the sample period, close the canister valve(s)
and disconnect the canister(s) from the sampling
system. Connect a pressure gauge to the canister (s)
and briefly open and close the canister valve. Note
the canister pressure on the Field Data Sheet (see
Figure 7). The canister pressure should be
approximately 2 atmospheres absolute [1 atmosphere or
101 kPa (gauge)]. Note: If the canister pressure is
not approximately 2 atmospheres absolute (14.7 psig) ,
determine and correct the cause before next sample.
Re-cap canister valve.
10.2.15 Fill out the identification tag on the sample
canister(s) and complete the Field Data Sheet as
necessary. Note any activities or special conditions
in the area (rain, smoke, etc.) that may affect the
sample contents on the sampling data sheet.
10.2.16 Return the canister(s) to the analytical system for
analysis.
11. Sample Analysis
11.1 Analytical System Leak Check
11.1.1 Before sample analysis, the analytical system is
assembled (see Figure 1) and leak checked.
11.1.2 To leak check the analytical system, place the six-
port gas valve in the trapping position. Disconnect
and cap the absolute pressure gauge. Insert a
pressure gauge capable of recording up to 60 psig at
the vacuum valve outlet.
11.1.3 Attach a valve and a zero air supply to the sample
inlet port. Pressurize the system to about 50 psig
(350 kPa) and close the valve.
-------
T012-16
11.1.4 Wait approximately 3 hrs. and re-check pressure. If
the pressure did not vary more than +_ 2 psig, the
system is considered leak tight.
11.1.5 If the system is leak free, de-pressurize and
reconnect absolute pressure gauge.
11.1.6 The analytical system leak check procedure needs to be
performed during the system checkout, during a series
of analysis or if leaks are suspected. This should be
part of the user-prepared SOP manual (see Section
12.1).
11.2 Sample Volume Determination
11.2.1 The vacuum reservoir and absolute pressure gauge are
used to meter a precisely repeatable volume of sample
air through the cryogenically-cooled trap, as follows:
With the sample valve closed and the vacuum valve
open, the reservoir is first evacuated with the vacuum
pump to a predetermined pressure (e.g., 100 mm Hg).
Then the vacuum valve is closed and the sample valve
is opened to allow sample air to be drawn through the
cryogenic trap and into the evacuated reservoir until
a second predetermined reservoir pressure is reached
(e.g., 300 mm Hg). The (fixed) volume of air thus
sampled is determined by the pressure rise in the
vacuum reservoir (difference between the predetermined
pressures) as measured by the absolute pressure gauge
(see Section 12 .2.1) .
11.2.2 The sample volume can be calculated by:
(AP)(Vr)
"
where:
Vs = volume of air sampled (standard cm3)
• P = pressure difference measured by gauge (mm Hg)
Vr = volume of vacuum reservoir (cm3)
usually 1 L
Ps = standard pressure (760 mm Hg)
For example, with a vacuum reservoir of 1000 cm3 and a
pressure change of 200 mm Hg (100 to 300 mm Hg), the volume
-------
T012-17
sampled would be 263 cm3. [Note: Typical sample volume
using this procedure is between 200-300 cm3.]
11.2.3 The sample volume determination need only be performed
once during the system check-out and shall be part of
the user-prepared SOP Manual (see Section 12.1).
11.3 Analytical System Dynamic Calibration
11.3.1 Before sample analysis, a complete dynamic calibration
of the analytical system should be carried out at five
or more concentrations on each range to define the
calibration curve. This should be carried out
initially and periodically thereafter [may be done
only once during a series of analyses]. This should
be part of the user-prepared SOP Manual (See Section
12.1). The calibration should be verified with two or
three-point calibration checks (including zero) each
day the analytical system is used to analyze samples.
11.3.2 Concentration standards of propane are used to
calibrate the analytical system. Propane calibration
standards may be obtained directly from low
concentration cylinder standards or by dilution of
high concentration cylinder standards with zero air
(see Section 8.3). Dilution flow rates must be
measured accurately, and the combined gas stream must
be mixed thoroughly for successful calibration of the
analyzer. Calibration standards should be sampled
directly from a vented manifold or tee. Note:
Remember that a propane NMOC concentration in ppmC is
three times the volumetric concentration in ppm.
11.3.3 Select one or more combinations of the following
parameters to provide the desired range or ranges
(e.g., 0-1.0 ppmC or 0-5.0 ppmC): FID attenuator
setting, output voltage setting, integrator resolution
(if applicable), and sample volume. Each individual
range should be calibrated separately and should have
a separate calibration curve. Note: Modern GC
integrators may provide automatic ranging such that
several decades of concentration may be covered in a
single range. The user-prepared SOP manual should
address variations applicable to a specific system
design (see Section 12.1).
-------
T012-18
11.3.4 Analyze each calibration standard three times
according to the procedure in Section 11.4. Insure
that flow rates, pressure gauge start and stop
readings, initial cryogen liquid level in the Dewar,
timing, heating, integrator settings, and other
variables are the same as those that will be used
during analysis of ambient samples. Typical flow
rates for the gases are: hydrogen, 30 cmVminute;
helium carrier, 30 cmVminute; burner air,
400 cmVminute.
11.3.5 Average the three analyses for each concentration
standard and plot the calibration curve(s) as average
integrated peak area reading versus concentration in
ppmC. The relative standard deviation for the three
analyses should be less than 3% (except for zero
concentration). Linearity should be expected; points
that appear to deviate abnormally should be repeated.
Response has been shown to be linear over a wide range
(0-10,000 ppbC). If nonlinearity is observed, an
effort should be made to identify and correct the
problem. If the problem cannot be corrected,
additional points in the nonlinear region may be
needed to define the calibration curve adequately.
11.4 Analysis Procedure
11.4.1 Insure the analytical system has been assembled
properly, leaked checked, and properly calibrated
through a dynamic standard calibration. Light the FID
detector and allow to stabilize.
11.4.2 Check and adjust the helium carrier pressure to
provide the correct carrier flow rate for the system.
Helium is used to purge residual air and methane from
the trap at the end of the sampling phase and to carry
the re-volatilized NMOC from the trap into the FID. A
single-stage auxiliary regulator between the cylinder
and the analyzer may not be necessary, but is
recommended to regulate the helium pressure better
than the multistage cylinder regulator. When an
auxiliary regulator is used, the secondary stage of
the two-stage regulator must be set at a pressure
higher than the pressure setting of the single-stage
-------
T012-19
regulator. Also check the FID hydrogen and burner air
flow rates (see 11.3.4) .
11.4.3 Close the sample valve and open the vacuum valve to
evacuate the vacuum reservoir to a specific
predetermined valve (e.g., 100 mm Hg).
11.4.4 With the trap at room temperature, place the six-port
valve in the inject position.
11.4.5 Open the sample valve and adjust the sample flow rate
needle valve for an appropriate trap flow of 50-100
cm3/min. Note: The flow will be lower later, when
the trap is cold.
11.4.6 Check the sample canister pressure before attaching it
to the analytical system and record on Field Data
Sheet (see Figure 7). Connect the sample canister or
direct sample inlet to the six-port valve, as shown in
Figure 1. For a canister, either the canister valve
or an optional fine needle valve installed between the
canister and the vent is used to adjust the canister
flow rate to a value slightly higher than the trap
flow rate set by the sample flow rate needle valve.
The excess flow exhausts through the vent, which
assures that the sample air flowing through the trap
is at atmospheric pressure. The vent is connected to
a flow indicator such as a rotameter as an indication
of vent flow to assist in adjusting the flow control
valve. Open the canister valve and adjust the
canister valve or the sample flow needle valve to
obtain a moderate vent flow as indicated by the
rotameter. The sample flow rate will be lower (and
hence the vent flow rate will be higher) when the trap
is cold.
11.4.7 Close the sample valve and open the vacuum valve (if
not already open) to evacuate the vacuum reservoir.
With the six-port valve in the inject position and the
vacuum valve open, open the sample valve for 2-3
minutes [with both valves open, the pressure reading
won't change] to flush and condition the inlet lines.
11.4.8 Close the sample valve and evacuate the reservoir to
the predetermined sample starting pressure (typically
100 mm Hg) as indicated by the absolute pressure
gauge.
-------
T012-20
11.4.9 Switch the six-port valve to the sample position.
11.4.10 Submerge the trap in the cryogen. Allow a few minutes
for the trap to cool completely (indicated when the
cryogen stops boiling). Add cryogen to the initial
level used during system dynamic calibration. The
level of the cryogenic liquid should remain constant
with respect to the trap and should completely cover
the beaded portion of the trap.
11.4.11 Open the sample valve and observe the increasing
pressure on the pressure gauge. When it reaches the
specific predetermined pressure (typically 300 mm Hg)
representative of the desired sample volume (Section
11.2), close the sample valve.
11.4.12 Add a little cryogen or elevate the Dewar to raise the
liquid level to a point slightly higher (3-15 mm) than
the initial level at the beginning of the trapping.
Note: This insures that organics do not bleed from
the trap and are counted as part of the NMOC peak(s).
11.4.13 Switch the 6-port valve to the inject position,
keeping the cryogenic liquid on the trap until the
methane and upset peaks have diminished (10-20
seconds). Now close the canister valve to conserve
the remaining sample in the canister.
11.4.14 Start the integrator and remove the Dewar flask
containing the cryogenic liquid from the trap.
11.4.15 Close the GC oven door and allow the GC oven (or
alternate trap heating system) to heat the trap at a
predetermined rate (typically, 30°C/min) to 90°.
Heating the trap volatilizes the concentrated NMOC
such that the FID produces integrated peaks. A
uniform trap temperature rise rate (above 0°C) helps
to reduce variability and facilitates more accurate
correction for the moisture-shifted baseline. With a
chromatograph oven to heat the trap, the following
parameters have been found to be acceptable: initial
temperature, 30°C; initial time, 0.20 minutes
(following start of the integrator); heat rate,
30°/minute; final temperature, 90°C.
11.4.16 Use the same heating process and temperatures for both
calibration and sample analysis. Heating the trap too
quickly may cause an initial negative response that
-------
T012-21
could hamper accurate integration. Some initial
experimentation may be necessary to determine the
optimal heating procedure for each system. Once
established, the procedure should be consistent for
each analysis as outlined in the user-prepared SOP
Manual.
11.4.17 Continue the integration (generally, in the range of
1-2 minutes is adequate) only long enough to include
all of the organic compound peaks and to establish the
end point FID baseline, as illustrated in Figure 8.
The integrator should be capable of marking the
beginning and ending of peaks, constructing the
appropriate operational baseline between the start and
end of the integration period, and calculating the
resulting corrected peak area. This ability is
necessary because the moisture in the sample, which is
also concentrated in the trap, will cause a slight
positive baseline shift. This baseline shift starts
as the trap warms and continues until all of the
moisture is swept from the trap, at which time the
baseline returns to its normal level. The shift
always continues longer than the ambient organic
peak(s). The integrator should be programmed to
correct for this shifted baseline by ending the
integration at a point after the last NMOC peak and
prior to the return of the shifted baseline to normal
(see Figure 8) so that the calculated operational
baseline effectively compensates for the water-shifted
baseline. Electronic integrators either do this
automatically or they should be programmed to make
this correction. Alternatively, analyses of
humidified zero air prior to sample analyses should be
performed to determine the water envelope and the
proper blank value for correcting the ambient air
concentration measurements accordingly. Heating and
flushing of the trap should continue after the
integration period has ended to insure all water has
been removed to prevent buildup of water in the trap.
Therefore, be sure that the 6-port valve remains in
the inject position until all moisture has purged from
the trap (3 minutes or longer).
-------
T012-22
11.4.18 Use the dynamic calibration curve (see Section 11.3)
to convert the integrated peak area reading into
concentration units (ppmC). Note that the NMOC peak
shape may not be precisely reproducible due to
variations in heating the trap, but the total NMOC
peak area should be reproducible.
11.4.19 Analyze each canister sample at least twice and report
the average NMOC concentration. Problems during an
analysis occasionally will cause erratic or
inconsistent results. If the first two analyses do
not agree within +_ 5% relative standard deviation
(RSD), additional analyses should be made to identify
inaccurate measurements and produce a more accurate
average (see also Section 12.2) .
12. Performance Criteria and Quality Assurance
This section summarizes required quality assurance measures and provides
guidance concerning performance criteria that should be achieved within
each laboratory.
12.1 Standard Operating Procedures (SOPs)
12.1.1 Users should generate SOPs describing and documenting
the following activities in their laboratory: (1)
assembly, calibration, leak check, and operation of
the specific sampling system and equipment used; (2)
preparation, storage, shipment, and handling of
samples; (3) assembly, leak check, calibration, and
operation of the analytical system, addressing the
specific equipment used; (4) canister storage and
cleaning; and (5) all aspects of data recording and
processing, including lists of computer hardware and
software used.
12.1.2 SOPs should provide specific stepwise instructions and
should be readily available to, and understood by, the
laboratory personnel conducting the work.
12.2 Method Sensitivity, Accuracy, Precision and Linearity
12.2.1 The sensitivity and precision of the method is
proportional to the sample volume. However, ice
formation in the trap may reduce or stop the sample
flow during trapping if the sample volume exceeds 500
cm3. Sample volumes below about 100-150 cm3 may cause
increased measurement variability due to dead volume
in lines and valves. For most typical ambient NMOC
-------
T012-23
concentrations, sample volumes in the range of 200-400
cm3 appear to be appropriate. If a response peak
obtained with a 400 cm3 sample is off scale or exceeds
the calibration range, a second analysis can be
carried out with a smaller volume. The actual sample
volume used need not be accurately known if it is
precisely repeatable during both calibration and
analysis. Similarly, the actual volume of the vacuum
reservoir need not be accurately known. But the
reservoir volume should be matched to the pressure
range and resolution of the absolute pressure gauge so
that the measurement of the pressure change in the
reservoir, hence the sample volume, is repeatable
within 1%. A 1000 cm3 vacuum reservoir and a pressure
change of 200 mm Hg, measured with the specified
pressure gauge, have provided a sampling precision of
+_ 1.31 cm3. A smaller volume reservoir may be used
with a greater pressure change to accommodate absolute
pressure gauges with lower resolution, and vice versa.
Some FID detector systems associated with laboratory
chromatographs may have autoranging. Others may
provide attenuator control and internal full-scale
output voltage selectors. An appropriate combination
should be chosen so that an adequate output level for
accurate integration is obtained down to the detection
limit; however, the electrometer or integrator must
not be driven into saturation at the upper end of the
calibration. Saturation of the electrometer may be
indicated by flattening of the calibration curve at
high concentrations. Additional adjustments of range
and sensitivity can be provided by adjusting the
sample volume use, as discussed in Section 12.2.1.
12.2.3 System linearity has been documented (6) from 0 to
10,000 ppbC.
12.2.4 Some organic compounds contained in ambient air are
"sticky" and may require repeated analyses before they
fully appear in the FID output. Also, some adjustment
may have to be made in the integrator off time setting
to accommodate compounds that reach the FID late in
the analysis cycle. Similarly, "sticky" compounds
from ambient samples or from contaminated propane
-------
T012-24
standards may temporarily contaminate the analytical
system and can affect subsequent analyses. Such
temporary contamination can usually be removed by
repeated analyses of humidified zero air.
12.2.5 Simultaneous collection of duplicate samples decreases
the possibility of lost measurement data from samples
lost due to leakage or contamination in either of the
canisters. Two (or more) canisters can be filled
simultaneously by connecting them in parallel (see
Figure 2(a)) and selecting an appropriate flow rate to
accommodate the number of canisters (Section 10.2.2).
Duplicate (or replicate) samples also allow assessment
of measurement precision based on the differences
between duplicate samples (or the standard deviations
among replicate samples).
13. Method Modification
13.1 Sample Metering System
13.1.1 Although the vacuum reservoir and absolute pressure
gauge technique for metering the sample volume during
analysis is efficient and convenient, other techniques
should work also.
13.1.2 A constant sample flow could be established with a
vacuum pump and a critical orifice, with the six-port
valve being switched to the sample position for a
measured time period. A gas volume meter, such as a
wet test meter, could also be used to measure the
total volume of sample air drawn through the trap.
These alternative techniques should be tested and
evaluated as part of a user-prepared SOP manual.
13.2 FID Detector System
13.2.1 A variety of FID detector systems should be adaptable
to the method.
13.2.2 The specific flow rates and necessary modifications
for the helium carrier for any alternative FID
instrument should be evaluated prior to use as apart
of the user-prepared SOP manual.
13.3 Range
13.3.1 It may be possible to increase the sensitivity of the
method by increasing the sample volume. However,
limitations may arise such as plugging of the trap by
ice.
-------
T012-25
13.3.2 Any attempt to increase sensitivity should be
evaluated as part of the user-prepared SOP manual.
13.4 Sub-Atmospheric Pressure Canister Sampling
13.4.1 Collection and analysis of canister air samples at
sub-atmospheric pressure is also possible with minor
modifications to the sampling and analytical
procedures.
13.4.2 Method TO-14, "Integrated Canister Sampling for
Selective Organics: Pressurized and Sub-atmospheric
Collection Mechanism," addresses sub-atmospheric
pressure canister sampling. Additional information
can be found in the literature (11-17) .
-------
T012-26
1. Uses, Limitations, and Technical Basis of Procedures for Quantifying
Relationships Between Photochemical Oxidants and Precursors, EPA-450/2-
77-21a, U.S. Environmental Protection Agency, Research Triangle Park,
NC, November 1977.
2. Guidance for Collection of Ambient Non-Methane Organic Compound (NMOC)
Data for Use in 1982 Ozone SIP Development, EPA-450/4-80-011, U.S.
Environmental Protection Agency, Research Triangle Park, NC, June 1980.
3. H.B. Singh, Guidance for the Collection and Use of Ambient Hydrocarbons
Species Data in Development of Ozone Control Strategies, EPA-450/4-80-
008, U.S. Environmental Protection Agency, Research Triangle Park, NC,
April 1980.
4. R.M. Riggin, Technical Assistance Document for Sampling and Analysis of
Toxic Organic Compounds in Ambient Air, EPA-600/4-83-027, U.S.
Environmental Protection Agency, Research Triangle Park, NC, 1983.
5. M.J. Jackson, et al., Technical Assistance Document for Assembly and
Operation of the Suggested Preconcentration Direct Flame lonization
Detection (PDFID) Analytical System, publication scheduled for late
1987; currently available in draft form from the Quality Assurance
Division, MD-77, U.S. Environmental Protection Agency, Research Triangle
Park, NC 27711.
6. R.K.M. Jayanty, et al., Laboratory Evaluation of Non-Methane Organic
Carbon Determination in Ambient Air by Cryogenic Preconcentration and
Flame lonization Detection, EPA-600/54-82-019, U.S. Environmental
Protection Agency, Research Triangle Park, NC, July 1982.
7. R.D. Cox, et al., "Determination of Low Levels of Total Non-Methane
Hydrocarbon Content in Ambient Air," Environ. Sci. Technol., 16 (1):57.
1982.
8. F.F. McElroy, et al., A Cryogenic Preconcentration - Direct FID (PDFID)
Method for Measurement of NMOC in the Ambient Air, EPA-600/4-85-063,
U.S. Environmental Protection Agency, Research Triangle Park, NC, August
1985.
9. F.W. Sexton, et al., A Comparative Evaluation of Seven Automated Ambient
Non-Methane Organic Compound Analyzers, EPA-600/54-82-046, U.S.
Environmental Protection Agency, Research Triangle Park, NC, August
1982.
10. H.G. Richter, Analysis of Organic Compound Data Gathered During 1980 in
Northeast Corridor Cities, EPA-450/4-83-017, U.S. Environmental
Protection Agency, Research Triangle Park, NC, April 1983.
11. Cox, R.D. "Sample Collection and Analytical Techniques for Volatile
Organics in Air," presented at APCA Specialty Conference, Chicago, IL,
March 22-24, 1983.
12. Rasmussen, R.A. and Khalil, M.A.K. "Atmospheric Halocarbons:
Measurements and Analyses of Selected Trace Gases," Proc. NATO ASI on
Atmospheric Ozone, 1980, 209-231.
13. Oliver, K.D., Pleil, J.D. and McClenny, W.A. "Sample Integrity of Trace
Level Volatile Organic Compounds in Ambient Air Stored in "SUMMA8"
Polished Canisters," accepted for publication in Atmospheric Environment
as of January 1986. Draft available from W.A. McClenny, MD-44, EMSL,
EPA, Research Triangle Park, NC 27711.
-------
T012-27
14. McClenny, W.A. Pleil, J.D., Holdren, J.W. and Smith, R.N.; 1984.
"Automated Cryogenic Preconcentration and Gas Chromatographic
Determination of Volatile Organic Compounds," Anal. Chem. 56:2947.
15. Pleil, J.D. and Oliver, K.D., 1985, "Evaluation of Various
Configurations of Nafion Dryers: Water Removal from Air Samples Prior
to Gas Chromatographic Analysis." EPA Contract No. 68-02-4035.
16. Oliver, K.D.; Pleil, J.D. and McClenny, W.A.; 1986. "Sample Integrity
of Trace Level Volatile Organic Compounds in Ambient Air Stored in
"SUMMA®" Polished Canisters," Atmospheric Environ. 20:1403.
17. Oliver, K.D. and Pleil, J.D., 1985, "Automated Cryogenic Sampling and
Gas Chromatographic Analysis of Ambient Vapor-Phase Organic Compounds:
Procedures and Comparison Tests," EPA Contract No. 68-02-4035, Research
Triangle Park, NC, Northrop Services, Inc. - Environmental Sciences.
-------
T012-28
CWSIER
FIGURE 1. SCHEMATIC OF ANALYTICAL SYSTEM FOR
NMOC-TWO SAMPLING MODES
-------
T012-29
SAMPLE
IN
PRESSURE
GAUGE
CRITICAL
ORIFICE
AUXILIARY
VACUUM
PUMP
METAL
BELLOWS
PUMP
CANISTER(S)
FIGURE 2. SAMPLE SYSTEM FOR AUTOMATIC COLLECTION
OF 3-HOUR INTEGRATED AIR SAMPLES
-------
T012-30
TIMER
SWITCH
115 V AC
Rl
100K
RED
40| RZ 100K U1
4<>Mfd, 450 v DC
D2
WHITE
Cl OM Cl - « «1. 490 VDC (Sprogut AUxrPlW 1712 or MuMM)
OM R> - Uiou, Uuinm
and 02 - 1000 mv. 2A A (RCA. 9C JOB1 or oquhannl)
MAGNELATCH
SOLENOID
VALVE
FIGURE 3[o]. SIMPLE CIRCUIT FOR OPERATING MAGNELATCH VALVE
115 V AC
Bntot Ibclliir - 200 HW. 1.9 * OKA. SIC 3109 o> oquMiM)
DUO Di ona Di - 1000 PIW U A (MX » mi or oquMM)
Copador Cl - 200 »I. »0 VOC (Sprogu. AloirAlVA 19» or oquMnQ
Copoclor Cl - 20 f>. «0 VOC Hor,-Pl»i««l (Sproguo Atom* NAN 1992 or oquhoM)
RW- IIM100 onm crA U ™ {«"T naur and BnMm Kd> 5. or oquMM)
NON-POLARIZED
FIGURE 3[b]. IMPROVED CIRCUIT DESIGNED TO HANDLE POWER INTERRUPTIONS
FIGURE 3. ELECTRICAL PULSE CIRCUITS FOR DRIVING
SKINNER MAGNELATCH SOLENOID VALVE
WITH A MECHANICAL TIMER
-------
T012-31
•F' SERIES COMPACT. MLINE FILTER
W/2 ion SS SINTERED ELEUENT
FEMALE CONNECTOR. 0.25 in O.D. TUBE TO
0.25 in FEMALE NPT
HEX NIPPLE. 0.25 in MALE NPT BOTH ENDS
30 GAUGE x 1.0 in LONG HYPODERMIC
NEEDLE (ORIFICE)
FEMALE CONNECTOR. 0.25 in O.D. TUBE TO
0.25 in FEMALE NPT
9THERMOGREEN LBI 6 mm (0.25 in)
SEPTUM (LOW BLEED)
0.25 in PORT CONNECTOR W/TWO 0.25 in NUTS
FIGURE 4. FILTER AND HYPODERMIC NEEDLE
ASSEMBLY FOR SAMPLE INLET FLOW
CONTROL
-------
T012-32
ZERO AIR
SUPPLY
VENT SHUT OFF
./ VALVE
VACUUM SHUT OFF
VALVE
VENT
VENT SHUT OFF
VALVE
ZERO AIR
SUPPLY
3-PORT
GAS
VENT VALVE VALVE
VSfiuoM VACUUM PUMP
™i!!!!' SHUT OFF VALVE VENT VALVE
CRYOGENIC
f TRAP
ZERO SHUT OFF
VALVE
VACUUM
GAUGE
/W VH FLOW
£-.
r» X. /\ CONTROL
X L^
VACUUM GAUGE
SHUT OFF VALVE
hd
-i VALVE
H-l CANISTER VALVE
SAMPLE CANISTERS
FIGURE 5. CANISTER CLEANING SYSTEM
-------
T012-33
TUBE LENGTH: -30 cm
O.D.: 0.32 cm
I.D.: 0.21 cm
CRYOGENIC LIQUID LEVEL
60/80 MESH GLASS BEADS
GLASS WOOL
<«4 cm
(TO FIT DEWAR)
•vl3 cm
FIGURE 6. CRYOGENIC SAMPLE TRAP DIMENSIONS
-------
T012-34
PRESSURIZED CANISTER SAMPLING DATA SHEET
GENERAL INFORMATION:
PROJECT:.
SITE:
LOCATION:
MONITOR STATION NUMBER:
PUMP SERIAL NUMBER:
OPERATOR:
ORIFICE IDENTIFICATION:.
FLOW RATE:
CALIBRATED BY:.
LEAK CHECK
Pass
Fail
FIELD DATA:
Date
Canister
Serial
Number
Sample
Number
Sample Time
Start
Stop
Average Atmospheric Conditions
Temperature
Pressure
Relative Humidity
Canister pressure
Final, Laboratory
Comments
Date
Title Signature
FIGURE 7. EXAMPLE SAMPLING DATA SHEET
-------
T012-35
NMOC
PEAK
END
INTEGRATION
CONTINUED HEATING
OF TRAP
WATER-SHIFTED
BASELINE
OPERATIONAL BASELINE
CONSTRUCTED BY INTEGRATOR
TO DETERMINE CORRECTED AREA
NORMAL BASELINE
TIME (MINUTES)
FIGURE 8. CONSTRUCTION OF OPERATIONAL BASELINE
AND CORRESPONDING CORRECTION OF
PEAK AREA
-------
APPENDIX D
Compendium Method TO-11A
Determination of Formaldehyde in Ambient Air Using Adsorbent Cartridge
Followed by High Performance Liquid Chromatography (HPLC)
[Active Sampling Methodology]
-------
EPA/625/R-96/010b
Compendium of Methods
for the Determination of
Toxic Organic Compounds
in Ambient Air
Second Edition
Compendium Method TO-11A
Determination of Formaldehyde in Ambient Air
Using Adsorbent Cartridge Followed by High
Performance Liquid Chromatography (HPLC)
[Active Sampling Methodology]
Center for Environmental Research Information
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
January 1997
-------
Method TO-11A
Acknowledgements
This Method was prepared for publication in the Compendium of Methods for the Determination of Toxic
Organic Compounds in Ambient Air, Second Edition (EPA/625/R-96/010b), which was prepared under
Contract No. 68-C3-0315, WA No. 3-10, by Midwest Research Institute (MRI), as a subcontractor to Eastern
Research Group, Inc. (ERG), and under the sponsorship of the U.S. Environmental Protection Agency (EPA).
Justice A. Manning and John O. Burckle, Center for Environmental Research Information (CERI), and Frank
F. McElroy, National Exposure Research Laboratory (NERL), both in the EPA Office of Research and
Development, were the project officers responsible for overseeing the preparation of this method. Additional
support was provided by other members of the Compendia Workgroup, which include:
John O. Burckle, U.S. EPA, ORD, Cincinnati, OH
• James L. Cheney, Corps of Engineers, Omaha, NB
Michael Davis, U.S. EPA, Region 7, KC, KS
Joseph B. Elkins Jr., U.S. EPA, OAQPS, RTP, NC
Robert G. Lewis, U.S. EPA, NERL, RTP, NC
Justice A. Manning, U.S. EPA, ORD, Cincinnati, OH
• William A. McClenny, U.S. EPA, NERL, RTP, NC
Frank F. McElroy, U.S. EPA, NERL, RTP, NC
• Heidi Schultz, ERG, Lexington, MA
• William T. "Jerry" Winberry, Jr., MRI, Gary, NC
Method TO-11 was originally published in March of 1989 as one of a series of peer reviewed methods in the
second supplement to "Compendium of Methods for the Determination of Toxic Organic Compounds in
Ambient Air," EPA 600/4-89-018. In an effort to keep these methods consistent with current technology,
Method TO-11 has been revised and updated as Method TO-11A in this Compendium to incorporate new or
improved sampling and analytical technologies.
This Method is the result of the efforts of many individuals. Gratitude goes to each person involved in the
preparation and review of this methodology.
Author(s)
William T. "Jerry" Winberry, Jr., Midwest Research Institute, Gary, NC
Silvestre Tejada, U.S. EPA, NERL, RTP, NC
Bill Lonneman, U.S. EPA, NERL, RTP, NC
Ted Kleindienst, ManTech, RTP, NC
Peer Reviewers
Robert G. Lewis, U.S. EPA, NERL, RTP, NC
Sucha S. Parmar, Atmospheric Analysis and Consulting, Ventura, CA
Joette Steger, Eastern Research Group, Morrisville, NC
Finally, recognition is given to Frances Beyer, Lynn Kaufman, Debbie Bond, Cathy Whitaker, and Kathy
Johnson of Midwest Research Institute's Administrative Services staff whose dedication and persistence during
the development of this manuscript has enabled it's publication.
in
-------
DISCLAIMER
This Compendium has been subjected to the Agency's peer and administrative review, and it has been
approved for publication as an EPA document. Mention of trade names or commercial products does
not constitute endorsement or recommendation for use.
IV
-------
Method TO-11A
Determination of Formaldehyde in Ambient Air Using Adsorbent Cartridge Followed by High
Performance Liquid Chromatography (HPLC) [Active Sampling Methodology]
TABLE OF CONTENTS
Page
1. Scope 11A-1
2. Applicable Documents 11A-3
2.1 ASTM Standards 11A-3
2.2 Other Documents 11A-3
2.3 Other Documents 11A-3
3. Summary of Method 11A-3
4. Significance 11A-4
5. Definitions 11A-6
5.1 CIS 11A-6
5.2 HPLC 11A-6
5.3 Method Detection Limit (MDL) 11A-6
5.4 Photochemical Reaction 11A-6
5.5 Photochemical Smog 11A-6
5.6 ppbv 11A-6
5.7 ppmv 11A-6
5.8 Silica Gel 11A-7
5.9 Denuder 11A-7
5.10 Certification Blank 11A-7
5.11 Cartridge Blank 11A-7
5.12 Scrubber 11A-7
6. Extended Methodology and Common Interferences 11A-7
7. Apparatus 11A-8
7.1 Isocratic HPLC 11A-8
7.2 Cartridge sampler 11A-9
7.3 Sampling system 11A-10
7.4 Stopwatch 11A-11
7.5 Polypropylene shipping container
with polyethylene-air bubble padding 11A-11
7.6 Thermometer 11A-11
7.7 Barometer (optional) 11A-11
7.8 Volumetric flasks 11A-11
7.9 Pipets 11A-11
7.10 Erlenmeyer flask, 1 L 11A-11
7.11 Graduated cylinder, 1 L 11A-11
7.12 Syringe, 100-250 ^L 11A-11
7.13 Sample vials 11A-11
-------
TABLE OF CONTENTS (continued)
Page
7.14 Melting point apparatus 11A-11
7.15 Rotameters 11A-11
7.16 Calibrated syringes 11A-11
7.17 Soap bubble meter or wet test meter 11A-11
7.18 Mass flow meters and mass flow controllers 11A-11
7.19 Positive displacement 11A-12
7.20 Cartridge drying manifold 11A-12
7.21 Liquid syringes 11A-12
7.22 Syringe rack 11A-12
7.23 Luer® fittings/plugs 11A-12
7.24 Hot plates, beakers, flasks, measuring and
disposable pipets, volumetric flasks, etc 11A-12
7.25 Culture tubes (20 mm x 125 mm) with polypropylene screw caps 11A-12
7.26 Polyethylene gloves 11A-12
7.27 Dry test meter 11A-12
7.28 User-prepared copper tubing for ozone scrubber 11A-12
7.29 Cord heater and Variac 11A-12
7.30 Fittings 11A-12
8. Reagents and Materials 11A-12
8.1 2,4-Dinitrophenylhydrazine (DNPH) 11A-13
8.2 DNPH coated cartridges
8.3 High purity acetonitrile 11A-13
8.4 Deionized-distilled water 11A-13
8.5 Perchloric acid 11A-13
8.6 Ortho-phosphoric acid 11A-13
8.7 Formaldehyde 11A-13
8.8 Aldehydes and ketones, analytical grade, best source 11A-13
8.9 Carbonyl hydrazone
8.10 Ethanol ormethanol 11A-13
8.11 Nitrogen 11A-13
8.12 Charcoal 11A-13
8.13 Helium 11A-13
8.14 Potassium Iodide 11A-13
9. Preparation of Reagents and Cartridges 11A-13
9.1 Purity of the Acetonitrile 11A-13
9.2 Purification of 2,4-Dinitrophenylhydrazine (DNPH) 11A-14
9.3 Preparation of DNPH-Formaldehyde Derivative 11A-16
9.4 Preparation of DNPH-Formaldehyde Standards 11A-16
9.5 Preparation of DNPH-Coated Cartridges 11A-16
9.6 Equivalent Formaldehyde Cartridge Concentration 11A-19
10. Sampling Procedure 11A-19
11. Sample Analysis 11A-22
vi
-------
TABLE OF CONTENTS (continued)
Page
11.1 Sample Preparation 11A-22
11.2 Sample Extraction 11A-23
11.3 HPLC Analysis 11A-24
11.4 HPLC Calibration 11A-25
12. Calculations 11A-26
13. Performance Criteria and Quality Assurance 11A-28
13.1 Standard Operating Procedures (SOPs) 11A-28
13.2 HPLC System Performance 11A-29
13.3 Process Blanks 11A-29
13.4 Method Precision and Accuracy 11A-29
13.5 Method Detection Limits 11A-30
13.6 General QA/QC Requirements 11A-30
14. Detection of Other Aldehydes and Ketones 11A-31
14.1 Introduction 11A-31
14.2 Sampling Procedures 11A-32
14.3 HPLC Analysis 11A-32
15. Precision and Bias 11A-33
16. References 11A-33
vn
-------
METHOD TO-11A
Determination of Formaldehyde in Ambient Air Using Adsorbent
Cartridge Followed by High Performance Liquid
Chromatography (HPLC)
[Active Sampling Methodology]
1. Scope
1.1 This document describes a method for the determination of formaldehyde and other carbonyl compounds
(aldehydes and ketones) in ambient air utilizing a coated-solid adsorbent followed by high performance liquid
chromatographic detection. Formaldehyde has been found to be a major promoter in the formation of
photochemical ozone. In particular, short term exposure to formaldehyde and other specific aldehydes
(acetaldehyde, acrolein, crotonaldehyde) is known to cause irritation of the eyes, skin, and mucous membranes
of the upper respiratory tract.
1.2 Over the last several years, numerous methods have been developed for the sampling and analysis of
carbonyl compounds. Because of the role which formaldehyde plays in photochemistry, most of the more
recent methods were designed to quantitate formaldehyde specifically. Early methods centered around wet
chemical technology involving a bubbler or impinger containing a reactive reagent (1). In some cases the
reactive reagent produced a color in the presence of formaldehyde. Examples of the more commonly used
reagents were: 3-methyl-2-benzothiazolone hydrazone (MBTH), sodium sulfite, 4-hexylresorcinol, water,
sodium tetrachloromercurate, and chromatropic acid. These reagents demonstrated high collection efficiency
(>95%), provided fairly stable non-volatile products and minimized formation of undesirable by-products.
Indeed, as part of U. S. Environmental Protection Agency's (EPA's) effort to quantitate atmospheric
concentrations of formaldehyde, the National Air Sampling Network utilized the impinger technique for
several years containing chromatrophic acid specifically for formaldehyde. However, impinger sampling had
numerous weaknesses which eventually lead to its demise. They were:
• Labor intense.
• Used acidic/hazardous reagents.
• Lacked sensitivity.
• Prone to interferences.
• Poor reproducibility at ambient concentration levels.
As EPA's interest focused upon formaldehyde and it's sources, the development of passive personal
sampling devices (PSDs) developed (2). These devices were mainly used by industrial hygienists to assess
the efforts of respiratory exposure for formaldehyde on workers. However, because of the design and flow rate
limitation, they require long exposures (up to 7 days) to the atmosphere to meet traditional bubbler technique
sensitivities. Consequently, the passive PSD had limited application to ambient monitoring.
To address the need for a monitoring method to sample carbonyl compounds in the air at sensitivities
needed to reach health-base detection limits (10~6 risk level), a combination of wet chemistry and solid
adsorbent methodology was developed (3-6). Activating or wetting the surface of an adsorbent with a
chemical specific for reacting with carbonyl compounds allowed greater volumes of air to be sampled, thus
enabling better sensitivity in the methodology. Various chemicals and adsorbents combinations have been
utilized with various levels of success. The most commonly used technique is based on reacting airborne
carbonyls with 2,4-dinitrophenylhydrazine (2,4-DNPH) coated on an adsorbent cartridge followed by
separation and analysis of the hydrazone derivative by high performance liquid chromatography (HPLC) with
ultraviolet (UV) detection.
1.3 Historically, Compendium Method TO-5, "Method For the Determination of Aldehydes and Ketones in
Ambient Air Using High Performance Liquid Chromatography (HPLC)" was used to quantitate formaldehyde
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-1
-------
Method TO-11A Formaldehyde
in ambient air. This method involved drawing ambient air through a midget impinger sampling train
containing 10 mL of 2N HC1/0.05% 2,4-DNPH reagent. Formaldehyde (and other aldehydes and ketones)
readily formed a stable derivative with the DNPH reagent, and the DNPH derivative is analyzed for aldehydes
and ketones utilizing HPLC. Compendium Method TO-11 modifies the sampling procedures outlined in
Method TO-5 by introducing a coated adsorbent. Compendium Method TO-11 is based on the specific
reaction of organic carbonyl compounds (aldehydes and ketones) with DNPH-coated silica gel cartridges in
the presence of a strong acid, as a catalyst, to form a stable color hydrazone derivative according to the
following reaction:
N02
R1 | R1
N02 —1±± ». ^C = N-NH-
-------
Formaldehyde Method TO-11A
liters/minute, allowing compatibility with pumps used in personal sampling equipment. These pre-coated
commercial cartridges have generally lower and more consistent background (7) concentration of carbonyls
than cartridges prepared under normal chemical laboratory environment, as specified in the original
Compendium Method TO-11.
1.7 The commercially-prepared pre-coated cartridges are used as received and are discarded after use. The
collected and uncollected cartridges are stored in culture tubes with polypropylene caps and placed in cold
storage when not in use.
1.8 This method may involve hazardous materials, operations, and equipments. This method does not purport
to address all the safety problems associated with its use. It is the responsibility of whoever uses this method
to consult and establish appropriate safety and health practices and determine the applicability of regulatory
limitations prior to use.
2. Applicable Documents
2.1 ASTM Standards
• D1193 Specification for Reagent Water
• D1356 Terminology Relating to Atmospheric Sampling and Analysis
• D3195 Practice for Rotameter Calibration
• D3631 Method for Measuring Surface Atmospheric Pressure
• D5197 Determination of Formaldehyde and Other Carbonyl Compounds in Air (Active Sampler
Methodology)
• El 77 Practice for Use of the Terms Precision and Bias in ASTM Test Methods
• E682 Practice for Liquid Chromatography Terms and Relationships
2.2 Other Documents
• Technical Assistance Document for Sampling and Analysis Toxic Organic Compounds in Ambient Air,
U. S. Environmental Protection Agency, EPA-600/4-83-027, June 1983.
• Quality Assurance Handbook for Air Pollution Measurement Systems, U. S. Environmental Protection
Agency, EPA-600/R-94-D38b, May 1994.
• Compendium of Methods for the Determination of Toxic Organic Compounds in Ambient Air: Method
TO-ll, Second Supplement, U. S. Environmental Protection Agency, EPA-600/4-89-018, March 1989.
2.3 Other Documents
• Existing Procedures (8-10).
• Ambient Air Studies (11-15).
3. Summary of Method
3.1 A known volume of ambient air is drawn through a prepacked cartridge coated with acidified DNPH at
a sampling rate of 100-2000 mL/min for an appropriate period of time. Sampling rate and time are dependent
upon carbonyl concentration in the test atmosphere.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-3
-------
Method TO-11A Formaldehyde
3.2 After sampling, the sample cartridges and field blanks are individually capped and placed in shipping
tubes with polypropylene caps. Sample identifying tags and labels are then attached to the capped tubes. The
capped tubes are then placed in a polypropylene shipping container cooled to subambient temperature (~4°C),
and returned to the laboratory for analysis. Alternatively, the sample vials can be placed in a thermally-
insulated styrofoam box with appropriate padding for shipment to the laboratory. The cartridges may either
be placed in cold storage until analysis or immediately washed by gravity feed elution with 5 mL of acetonitrile
from a plastic syringe reservoir to a graduated test tube or a 5 mL volumetric flask.
3.3 The eluate is then diluted to a known volume and refrigerated until analysis.
3.4 For determining formaldehyde, the DNPH-formaldehyde derivative can be determined using isocratic
reverse phase FiPLC with an ultraviolet (UV) absorption detector operated at 360 nm. To determine
formaldehyde and 14 other carbonyls, the HPLC system is operated in the linear gradient program mode.
3.5 For quantitative evaluation of formaldehyde and other carbonyl compounds, a cartridge blank is likewise
desorbed and analyzed.
3.6 Formaldehyde and other carbonyl compounds in the sample are identified and quantified by comparison
of their retention times and peak heights or peak areas with those of standard solutions. Typically, Q-C7
carbonyl compounds, including benzaldehyde, are measured effectively to less than 0.5 ppbv.
4. Significance
4.1 Formaldehyde is a major compound in the formation of photochemical ozone (16). Short term exposure
to formaldehyde and other specific aldehydes (acetaldehyde, acrolein, crotonaldehyde) is known to cause
irritation of the eyes, skin, and mucous membranes of the upper respiratory tract (19). Animal studies indicate
that high concentrations can injure the lungs and other organs of the body (19). In polluted atmospheres,
formaldehyde may contribute to eye irritation and unpleasant odors that are common annoyances.
4.2 Over the last several years, carbonyl compounds including low molecular weight aldehydes and ketones
have received increased attention in the regulatory community. This is due in part to their effects on humans
and animals as primary irritation of the mucous membranes of the eyes, the upper respiratory tract, and the
skin. Animal studies indicate that high concentrations of carbonyl compounds, especially formaldehyde, can
injure the lungs, may contribute to eye irritation and effect other organs of the body. Aldehydes, either directly
of indirectly, may also cause injury to plants. Sources of carbonyl compounds into the atmosphere range from
natural occurrences to secondary formation through atmospheric photochemical reactions. Consequently,
carbonyl compounds are both primary (directly emitted) and secondary (formed in the atmosphere) air
pollutants (19).
4.2.1 Natural Occurrence. Natural sources of carbonyls do not appear to be important contributors to
air pollution. Acetaldehyde is found in apples and as a by-product of alcoholic fermentation process. Other
lower molecular weight aliphatic aldehydes are not found in significant quantities in natural products. Olefmic
and aromatic aldehydes are present in some of the essential oils in fruits and plants. These include citronella,
in rose oil; citral, in oil of lemongrass; benzaldehyde, in oil of bitter almonds; and cinnamaldehyde, in oil of
cinnamon.
4.2.2 Production Sources. Aldehydes are commercially manufactured by various processes, depending
on the particular aldehyde. In general, they are prepared via oxidation reactions of hydrocarbons,
hydroformulation of alkenes, dehydrogenation of alcohols, and addition reactions between aldehydes and other
Page 11A-4 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde Method TO-11A
compounds. Formaldehyde is manufactured from the oxidation of methanol as illustrated in the following
equation:
[cat.]
CH3OH - CH2O + H2
Formaldehyde and other aldehyde production in the United States has shown a substantial growth over the last
several years. This is due, in part, to their use in a wide variety of industries, such as the chemical, rubber,
tanning, paper, perfume, and food industries. The major use is as an intermediate in the synthesis of organic
compounds, including, alcohols, carboxylic acids, dyes, and medicinals.
4.2.3 Mobile Combustion Sources. A major source of carbonyl compounds in the atmosphere may be
attributed to motor vehicle emissions. In particular, formaldehyde is the major carbonyl in automobile exhaust,
accounting for 50-70 percent of the total carbonyl burden to the atmosphere (19). Furthermore, motor vehicles
emit reactive hydrocarbons that undergo photochemical oxidation to produce formaldehyde and other
carbonyls in the atmosphere.
4.3 Secondary Pollutant. As a secondary pollutant (formed in the atmosphere), carbonyls are formed by very
complex photo-oxidation mechanism involving volatile organic compounds (VOCs) with nitrogen oxide
(20,21). Both anthropogenic and biogenic (e.g., isoprene) hydrocarbons leads to in situ formation of carbonyls,
especially formaldehyde compounds. Aldehydes are both primary pollutants and secondary products of
atmospheric photochemistry.
The complete photo-oxidation mechanism is indeed complex and not well understood. However, a brief
discussion is warranted (22). When VOCs and oxides of nitrogen (NOX) are in the atmosphere and are
irradiated with sunlight, their equilibrium in the photostationary state is changed. The photostationary state
is defined by the equilibrium between nitrogen dioxide (NO2), nitrous oxide (NO) and ozone (Q ). This
equilibrium is theoretically maintained until VOCs are introduced. Various reactions occur to produce OH
radicals. The VOCs react with the OH radicals and produce RO2 radicals that oxidizes NO to NO2, destroying
the photostationary state. Carbonyls react with OH to produce RO2 radicals. Likewise carbonyls, particularly
formaldehyde in sunlight, are sources of the OH radicals.
The results of these processes lead to the following:
• Accumulation of ozone.
• Oxidation of hydrocarbons (HCs) to aldehydes and ketones which lead to the continued production of
HO2- and OH- radicals, the real driving force in photochemistry smog.
Consequently, the determination of formaldehyde and other carbonyl compounds in the atmosphere is of
interest because of their importance as precursors in the production of photochemical smog, as photochemical
reaction products and as major source of free radicals in the atmosphere.
4.4 Historically, DNPH impinger techniques have been widely used to determine atmospheric carbonyls.
However, due to the limitation of applying this technique to remote locations, the solid adsorbent methodology
has become a convenient alternative to impinger sampling. A number of solid adsorbents have been used over
the years to support the DNPH coating. They are: glass beads, glass fiber filters, silica gel, Chromosorb® P,
Florisil®, Carbopack® B, XAD-2, and CIS. Several of these adsorbents are available commercially as pre-
packed cartridges. The commercially available cartridges provide convenience of use, reproducibility and low
formaldehyde blanks. Two of the more widely used pre-packed adsorbents are silica gel and CIS.
4.4.1 Silica Gel. Silica gel is a regenerative adsorbent, consisting of amorphous silica (SiO2) with surface
OH groups, making it a polar material and enhancing surface absorption. DNPH-coated silica gel cartridges
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-5
-------
Method TO-11A Formaldehyde
have been used by numerous investigators since 1980 for sampling formaldehyde in ambient air. Tejada (3,4)
evaluated several adsorbents, including CIS, Florsil, silanized glass wool, and silica gel as possible supports
for the DNPH coating. Results indicated that silica gel provided the best support with minimum interferences.
The studies did document that olefmic aldehydes such as acrolein and crotonaldehyde degraded partially and
formed unknown species. For stable carbonyls such as formaldehyde, acetaldehyde, propionaldehyde,
benzaldehyde, and acetone, correlation with an DNPH-impinger technique was excellent. However, further
studies by Arnts and Tejada identified a severe loss of carbonyl-DNPH derivative due to the reaction of
atmospheric ozone on DNPH-coated silica gel cartridges while sampling ambient air. This bias was eliminated
when sampling continued with the application of an ozone scrubber system (KI denuder) preceding the
cartridge.
4.4.2 CIS Cartridge. CIS is an octadecylsilane bonded silica substrate which is non-polar, hydrophobic,
and relatively inert, whose surface has been passivated with non-polar paraffmic groups. Because of these
qualities, CIS has been used historically as an adsorbent trap for trace organics in environmental aqueous
samples through hydrophobic interactions. The adsorbed trace organic molecules are then eluted from the
adsorbent with various organic solvents. In early 1990, CIS was used in an ambient air study as the support
for DNPH. While CIS showed promising results (23), it's use today as the support for DNPH is limited.
4.5 Both adsorbents have historically performed adequately as the support for the DNPH coating. The
comparison between silica gel and CIS as the adsorbent for the DNPH is illustrated in Table 1. The user is
encouraged to review the weaknesses and strengths outlined in Table 1 for using silica gel or CIS as the
adsorbent for the DNPH coating.
5. Definitions
[Note: Definitions used in this document and any user-prepared Standard Operating Procedures (SOPs)
should be consistent with those used in ASTMD1356. All abbreviations and symbols are defined within this
document at the point of first use.]
5.1 CIS— CIS is an octadecylsilane bonded silica substrate, which is non-polar, hydrophobic, and relatively
inert.
5.2 HPLC— high performance liquid chromatography.
5.3 Method Detection Limit (MDL)— the minimum concentration of an analyte that can be reported with
95% confidence that the value is above zero, based on a standard deviation of at least seven repetitive
measurements of the analyte in the matrix of concern at a concentration near the low standard.
5.4 Photochemical Reaction— any chemical reaction that is initiated as a result of absorption of light.
5.5 Photochemical Smog— air pollution resulting from photochemical reactions.
5.6 ppbv— a unit of measure of the concentration of gases in air expressed as parts of the gas per billion (109)
parts of the air-gas mixture, normally both by volume.
5.7 ppmv— a unit of measure of the concentration of gases in air expressed as parts of the gas per million
(106) parts of the air-gas mixture, normally both by volume.
Page 11A-6 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde Method TO-11A
5.8 Silica Gel—silica gel is a regenerative adsorbent consisting of amorphous silica (SiO2) with OH surface
groups making it a polar material and enhancing surface reactions.
5.9 Denuder— A device designed to remove gases from an air sampling stream by the process of molecular
diffusion to a collecting surface.
5.10 Certification Blank— certification blank is defined as the mean value of the cartridge blank plus three
standard deviations. For Compendium Method TO-11A, the Certification Blank should be less than
0.15 (jg/cartridge for formaldehyde.
5.11 Cartridge Blank— cartridge blank is the measured value of the carbonyl compounds on an unsampled,
DNPH-coated cartridge. This is the value used in the calculations delineated in section 12.
5.12 Scrubber— to remove a specific gas from the air stream by passing through a pack bed.
6. Extended Methodology and Common Interferences
6.1 This procedure has been written specifically for the sampling and analysis of formaldehyde. Other
carbonyl compounds found in ambient air are also observed in the HPLC analysis. Resolution of these
compounds depend upon column and mobile phase conditions during FiPLC analysis. Organic compounds
that have the same retention time and significant absorbance at 360 nm as the DNPH derivative of
formaldehyde will interfere. Such interferences (24) can often be overcome by altering the separation
conditions (e.g., using alternative HPLC columns or mobile phase compositions). In addition, other aldehydes
and ketones can be detected with a modification of the basic procedure. In particular, chromatographic
conditions can be optimized to separate acetone and propionaldehyde and 12 other higher molecular weight
aldehydes and ketones (within an analysis time of about one hour), as identified below, by utilizing one or two
Zorbax ODS columns in series under a linear gradient program:
Formaldehyde Isovaleraldehyde Propionaldehyde p-Tolualdehyde
Acetaldehyde Valeraldehyde Crotonaldehyde Hexanaldehyde
o-Tolualdehyde Butyraldehyde 2,5-Dimethylbenzaldehyde Methyl ethyl ketone
Acetone m-Tolualdehyde Benzaldehyde
The linear gradient program varies the mobile phase composition periodically to achieve maximum resolution
of the C-3, C-4, and benzaldehyde region of the chromatogram.
6.2 Formaldehyde may be a contamination of the DNPH reagent. If user- prepared cartridges are employed,
the DNPH must be purified by multiple recrystallizations in UV grade carbonyl-free acetonitrile.
Recrystallization is accomplished at 40-60 °C by slow evaporation of the solvent to maximize crystal size. The
purified DNPH crystals are stored under UV grade carbonyl-free acetonitrile until use. Impurity levels of
carbonyl compounds in the DNPH are determined by HPLC prior to use and should be less than the
Certification Blank value of 0.15 ^g/cartridge.
6.3 The purity of acetonitrile is an important consideration in the determination of allowable formaldehyde
blank concentration in the reagent. Background concentrations of formaldehyde in acetonitrile will be
quantitatively converted to the hydrazone, adding a positive bias to the ambient air formaldehyde
concentration. Within the project quality control procedures, the formaldehyde in the acetonitrile reagent
should be checked on a regular basis (see Section 9.1).
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-7
-------
Method TO-11A Formaldehyde
6.4 Ozone at high concentrations has been shown to interfere negatively by reacting with both the DNPH and
its carbonyl derivatives (hydrazones) on the cartridge (25,26). The extent of interference depends on the
temporal variations of both the ozone and the carbonyl compounds and the duration of sampling. Significant
negative interference from ozone was observed even at concentrations of formaldehyde and ozone typical of
clean ambient air (i.e., 2 and 40 ppb, respectively).
6.5 Exposure of the DNPH-coated sampling cartridges to direct sunlight may produce artifacts and should be
avoided.
6.6 The presence of ozone in the sample stream is readily inferred from the appearance of new compounds
with retention times different from the other carbonyl hydrazone compounds.
6.7 The most direct solution to the ozone interference is to remove the ozone before the sample stream reaches
the coated cartridge. This process entails constructing an ozone denuder (9) or scrubber and placing it in front
of the cartridge. The denuder can be constructed of 1 m of 0.64-cm outside diameter (O.D.) by 0.46-cm inside
diameter (I.D.) copper tubing, that is filled with a saturated solution of KI, allowed to stand for a few minutes,
drained and dried with a stream of clean air or nitrogen for about 1 h. The capacity of the ozone denuder as
described is about 100,000 ppb-hour of ozone. Packed-bed granular potassium iodide (KI) scrubbers can also
be used in place of the denuder and are commercially available. Very little work has been done on long term
usage of a denuder or KI scrubber to remove ozone from the ambient air gas stream. The ozone removal
devices should be replaced periodically (e.g., monthly) in the sample train to maintain the integrity of the data
generated.
6.8 Test aldehydes or carbonyls (formaldehyde, acetaldehyde, acrolein, propionaldehyde, benzaldehyde, and
p-tolualdehyde) that were dynamically spiked into an ambient sample air stream passed through the KI denuder
with practically no losses (7). Similar tests were also performed for formaldehyde (26).
6.9 Ozone scrubbers (cartridge filled with granular KI) are also available from suppliers of pre-coated DNPH
cartridges. These scrubbers are optimized when the ambient air contains a minimum of 15% relative humidity.
7. Apparatus
7.1 Isocratic HPLC. System consisting of a mobile phase reservoir a high pressure pump; an injection valve
(automatic sampler with an optional 25-yL loop injector); a Zorbax ODS (DuPont Instruments, Wilmington,
DE) reverse phase (RP) column, or equivalent (25-cm x 4.6-mm ID); a variable wavelength UV detector
operating at 360 nm; and a data system, as illustrated in Figure 1.
[Note: Most commercial HPLC analytical systems will be adequate for this application.]
7.2 Cartridge sampler. Prepacked, pre-coated cartridge (see Figure 2), commercially available or coated in
situ with DNPH according to Section 9.
[Note: This method was developed using the Waters Sep-Pak cartridge, coated in situ with DNPH on silica
gel by the users, as delineated in the original Compendium Method TO-11 as a guideline. EPA has
experience in use of this cartridge during various field monitoring programs over the last several years. Other
manufacturer's cartridges should work as well. However, modifications to these procedures may be necessary
if another commercially available cartridge is selected.]
Page 11A-8 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde Method TO-11A
Major suppliers of pre-coated cartridges are:
• Supelco, Supelco Park, Bellefonte, PA 16823-0048, (800) 247-6628.
• SKC Inc., 334 Valley View Road, Eighty Four, PA 15330-9614, (800) 752-8472.
• Millipore/Waters Chromatography, P.O. Box 9162, Marlborough, MA 01752-9748,
(800) 252-4752.
• Atmospheric Analysis and Consulting (AAC) Inc., 4572 Telephone Rd., Suite 920, Ventura, CA 93003,
(805) 650-1642.
[Note: The SKC cartridge (see Figure 2) is an example of a dual bed tube. The glass cartridge contains a
front bed of 300 mg DNPH-coated silica gel with the back bed of 150 mg DNPH-coated silica gel. Airflow
through the tube should be from front to back bed, as indicated by the arrows enscribed on the cartridge. The
dual bed tube cartridge may be used in atmospheres containing carbonyl concentrations in excess of the
American Conference of Government Industrial Hygienists (ACGIH) 8-hour exposure limit, where
breakthrough of carbonyls on the adsorbent might occur. If used in routine ambient air monitoring
applications, the tube is recovered as one unit, as specified in Section 11.2.]
If commercially prepared DNPH-coated cartridges are purchased, ensure that a "Certification Blank for
Formaldehyde " is provided for the specific batch of which that cartridge is a member. For a commercial
cartridge to be acceptable, the following criteria must be met:
• Formaldehyde concentration: <0.15 (jg/cartridge.
If the enhanced carbonyl analysis is being performed, the following Certification Blank criteria must also be
met:
• Speciated carbonyl concentration:
- Acetaldehyde: <0.10 (jg/cartridge
- Acetone: <0.30 (jg/cartridge
- Other: <0.10 (jg/cartridge
Typical physical and chemical characteristics of commercial cartridge adsorbents are listed in Table 2 and
illustrated in Figure 2.
7.3 Sampling system, the DNPH-cartridge approach is capable of accurately and precisely sampling
100-2000 mL/min of ambient air. The monitoring of carbonyl compounds has recently been enhanced by the
promulgation of new ambient air quality surveillance regulations outlined in Title 40, Part 58. These
regulations require States to establish additional air monitoring stations as part of their existing State
Implementation Plan (SIP) monitoring network as part of EPA's Photochemical Assessment Monitoring
Stations (PAMS) to include provisions for enhanced (1) monitoring of ozone and oxides of nitrogen (NOX),
(2) monitoring of volatile organic compounds (VOCs), (3) monitoring of meteorological parameters, and (4)
monitoring selected carbonyl compounds (formaldehyde, acetone, and acetaldehyde). Specifically, monitoring
for carbonyl involves:
• 8, 3 h sequential samples starting at midnight.
• 1, 24 h time-integrated "reality check" sample.
Consequently, the sampler must be able to accommodate numerous regulatory and practical needs. Practical
needs would include:
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-9
-------
Method TO-11A Formaldehyde
• Ability to sequence two cartridges in series for breakthrough volume confirmation for a 24-hour
sampling event.
• Ability to collocate with any of the 8, 3 h samples.
Traditionally, three sampling approaches have been used to monitor carbonyl compounds in the ambient air.
They are:
• Manual single-port carbonyl sampler.
• Programmable single-port carbonyl sampler.
• Automated multi-port sampler.
Components of the single-port carbonyl sampler, for both manual and semi-automatic, are illustrated in
Figure 3. Components usually include a heated manifold/sample inlet, a denuder/cartridge assembly, a flow
meter, a vacuum gauge/pump, a timer and a power supply. In operation, ambient air is drawn through the
denuder/cartridge assembly with a vacuum pump at a fixed flow rate between 0.1 to 2 Lpm. The vacuum
gauge is used to measure the net vacuum in the system for all flow-rate corrections. Controlling the system
is usually a 7-day, 14-event timer to coordinate sampling events to allow a sample to be extracted continuously
or intermittently over a period of time. Finally, an elapsed-time counter is employed to measure the actual time
the sampling took place. This is particularly suitable for unattended sampling when power fails for short
periods.
The automated multi-port sampler is especially designed to collect numerous short-term (2 to 3 hours) sample
sequentially over a 24 hour, 7 day a week, nighttime and weekend monitoring period. This arrangement allows
for the sampling of short periods where the objectives of the project are to identify progress of atmospheric
reactions involving carbonyls. As illustrated in Figure 4, components of the fully automated multi-port
carbonyl sampler includes a heated inlet, ozone denuder (or scrubber) inlet manifold assembly, inlet check
valves, DNPH multi-port cartridge assembly, exhaust manifold, mass flow controller and sample pump. The
multi-port sampler automatically switches between sampling ports at preselected times, as programmed by the
user. Typically, a sequential air sampler contains a microprocessor timer/controller that provides precise
control over each sampling event. The microprocessor allows the user to program individual start date and
time, sample duration, and delays between samples. The timer also allows activation of the flow system prior
(approximately 10 min) to sequencing to allow purging of the sampler inlet with fresh sample. Finally, the
automated sequential sampler can be operated from an external signal, such as an ozone monitor, so that
sampling starts above certain preset ozone levels or via a modem. As a final option, various manufacturers
provide wind sensor instrumentation (wind speed and direction) which is connected to the automated
sequential sampler so that sampling begins when the wind is from a preset direction and speed.
Major suppliers of commercially available carbonyl samplers are:
• Supelco, Supelco Park, Bellefonte, PA 16823-0048, (800) 247-6628.
• SKC Inc., 334 Valley View Road, Eighty Four, PA 15330-9614, (800) 752-8472.
• Millipore/Waters Chromatography, P.O. Box 9162, Marlborough, MA 01752-9748, (800) 252-4752.
• XonTech, Inc. 6862 Hayvenhurst Avenue, Van Nuys, CA 91406, (818) 787-7380.
• ATEC Atmospheric Technology, P.O. Box 8062, Calabasas, CA 91372-8062, (310) 457-2671.
• Atmospheric Analysis and Consulting (AAC) Inc., 4572 Telephone Road, Suite 920, Ventura, CA
93003,(805) 650-1642.
• Scientific Instrumentation Specialists, P.O. Box 8941, Moscow, ID, (209) 882-3860.
7.4 Stopwatch.
Page 11A-10 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde Method TO-11A
7.5 Polypropylene shipping container (see Figure 5) with polyethylene-air bubble padding. To hold
sample cartridges.
7.6 Thermometer. To record ambient temperature.
7.7 Barometer (optional).
7.8 Volumetric flasks. Various sizes, 5-2000 mL.
7.9 Pipets. Various sizes, 1-50 mL.
7.10 Erlenmeyer flask, 1 L. For preparing HPLC mobile phase.
7.11 Graduated cylinder, 1 L. For preparing FiPLC mobile phase.
7.12 Syringe, 100-250 ,wL. For HPLC injection, with capacity at least four times the loop value.
7.13 Sample vials.
7.14 Melting point apparatus (optional).
7.15 Rotameters.
7.16 Calibrated syringes.
7.17 Soap bubble meter or wet test meter.
7.18 Mass flow meters and mass flow controllers. For metering/setting air flow rate through sample
cartridge of 100-2000 mL/min.
[Note: The mass flow controllers are necessary because cartridges may develop a high pressure drop and
at maximum flow rates, the cartridge behaves like a "critical orifice. "Recent studies have shown that critical
flow orifices may be used for 24-hour sampling periods at a maximum rate of 2 L/minfor atmospheres not
heavily loaded with particulates without any problems.]
7.19 Positive displacement. Repetitive dispensing pipets (Lab-Industries, or equivalent), 0-10 mL range.
7.20 Cartridge drying manifold. With multiple standard male Luer® connectors.
7.21 Liquid syringes. 10 mL (polypropylene syringes are adequate) for preparing DNPH-coated cartridges.
7.22 Syringe rack. Made of an aluminum plate (0.16 cm x 36 cm x 53 cm) with adjustable legs on four
corners. A matrix (5 cmx 9 cm) of circular holes of diameter slightly larger than the diameter of the 10-mL
syringes was symmetrically drilled from the center of the plate to enable batch processing of 45 cartridges for
cleaning, coating, and/or sample elution.
7.23 Luer® fittings/plugs. To connect cartridges to sampling system and to cap prepared cartridges.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-11
-------
Method TO-11A Formaldehyde
7.24 Hot plates, beakers, flasks, measuring and disposable pipets, volumetric flasks, etc. Used in the
purification of DNPH.
7.25 Culture tubes (20 mm x 125 mm) with polypropylene screw caps. Used to transport coated cartridges
for field applications (see Figure 5), Fisher Scientific, Pittsburgh, PA, or equivalent.
7.26 Polyethylene gloves. Used to handle cartridges, best source.
7.27 Dry test meter.
7.28 User-prepared copper tubing for ozone scrubber (see Figure 6a). A 36 inch length of %-inch O.D.
copper tubing is used as the body of the ozone scrubber. The tubing should be coiled into a spiral
approximately 2 inches in O.D. EPA has considerable field experience with the use of this denuder.
[Note: Ozone scrubbers (cartridge filled with granular KI) are also available from suppliers ofpre-coated
DNPH cartridges, as illustrated in Figure 6(b).]
7.29 Cord heater and Variac. A 24 inch long cord heater, rated at approximately 80 watts, wrapped around
the outside of the copper coil denuder, controlled by a Variac, to provide heat (~50°C) to prevent condensation
of water or organic compounds from occurring within the coil.
7.30 Fittings. Bulkhead unions are attached to the entrance and exit of the copper coil to allow attachment
to other components of the sampling system.
8. Reagents and Materials
[Note: Purity of Reagents—Reagent grade chemicals shall be used in all tests. Unless otherwise indicated,
it is intended that all reagents conform to the specifications of the Committee on Analytical Reagents of the
American Chemical Society where such specifications are available; Purity of Water—Unless otherwise
indicated, references to water shall be understood to mean reagent water as defined by Type II ofASTM
Specifications D 1193.]
8.1 2,4-Dinitrophenylhydrazine (DNPH). Aldrich Chemical or J.T. Baker, reagent grade or equivalent.
Recrystallize at least twice with UV grade acetonitrile before use.
8.2 DNPH coated cartridges. DNPH coated cartridge systems are available from several commercial
suppliers.
8.3 High purity acetonitrile. UV grade, Burdick and Jackson "distilled-in-glass," or equivalent. The
formaldehyde concentration in the acetonitrile should be < 1.5 ng/mL. It is imperative (mandatory) that the
user establish the purity of the acetonitrile before use (see Section 9.1).
8.4 Deionized-distilled water. Charcoal filtered.
8.5 Perchloric acid. Analytical grade, best source, 60%, specific gravity 1.51.
8.6 Ortho-phosphoric acid. Analytical grade, best source, 36.5-38%, specific gravity 1.19.
Page 11A-12 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde Method TO-11A
8.7 Formaldehyde. Analytical grade, best source, 37% solution (w/w).
8.8 Aldehydes and ketones, analytical grade, best source. Used for preparation of DNPH derivative
standards (optional).
8.9 Carbonyl hydrazones. Formaldehyde and other carbonyl hydrazones are available for use as standards
from commercial sources at various levels of purity.
8.10 Ethanol or methanol. Analytical grade, best source.
8.11 Nitrogen. High purity grade, best source.
8.12 Charcoal. Granular, best source.
8.13 Helium. High purity grade, best source.
8.14 Potassium Iodide. Analytical grade, best source. Used for coating inside of copper tubing of denuder
system to remove ozone interference.
9. Preparation of Reagents and Cartridges
9.1 Purity of the Acetonitrile
9.1.1 The purity of acetonitrile is an important consideration in the determination of the formaldehyde
blank concentration. Formaldehyde in the reagent will be quantitatively converted to the hydrazone and
measured as part of the blank. The contribution to the blank from the reagent is dependent on the
formaldehyde concentration in the reagent and the amount of the reagent used for extraction. Some examples
will illustrate these considerations.
Example A
• Silica gel DNPH cartridge has a blank level of 60 ng.
• Cartridge is eluted with 5-mL of acetonitrile reagent containing a formaldehyde of 3 ng/mL.
• Analyst measures a blank level of 75 ng of which 80% comes from the cartridge and 20% comes from
the reagent.
Example B
• Silica gel DNPH cartridge has a blank level of 30 ng.
• Cartridge is eluted with 5 mL of acetonitrile reagent containing a formaldehyde of 6 ng/mL.
• Analyst measures a blank level of 60 ng of which 50% comes from the cartridge and 50% comes from
the reagent.
9.1.2 As a quality control procedure, the formaldehyde in the acetonitrile reagent should be checked on
a regular basis. This can be done by mixing known proportions of the acetonitrile reagent and a DNPH
solution having a measured formaldehyde blank. (The extract from a blank cartridge can serve as the DNPH
solution.) After analyzing the resultant solution, a mass balance is performed on the observed formaldehyde
level and the contribution from the DNPH reagent as shown in the following example.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-13
-------
Method TO-11A Formaldehyde
• 1 mL of a DNPH solution containing 2.1 ng/mL of formaldehyde (as carbonyl) is mixed with 9 mL of
acetonitrile reagent containing as unknown formaldehyde blank. The analyst measures a resultant
solution concentration of 1.55 ng of formaldehyde. This data can be used to calculate the formaldehyde
in the reagent:
TT/^TT/-V / T (1.55ng/mL x 10ml-2.1ng/mL x ImL) , .„ , T
HCHO ng/mL = -^ ^ & L = l.49ng/mL
9mL
The formaldehyde contribution to the cartridge blank should be as low as possible but certainly less than 20%
of the total measured blank. Using a cartridge blank level of 30 ng/cartridge, the formaldehyde concentration
in the reagent would have to be less than 1.5 ng/mL (i.e., 50 nM) to give a blank level less than 20% of the
measured blank.
9.2 Purification of 2,4-Dinitrophenylhydrazine (DNPH)
[Note: This procedure should be performed under a properly ventilated hood, as inhalation of acetonitrile
can result in nose and throat irritation. Various health effects are resultant from the inhalation of acetonitrile.
At 500 ppm in air, brief inhalation has produced nose and throat irritation. At 160 ppm, inhalation for 4
hours has caused flushing of the face (2 hour delay after exposure) and bronchial tightness (5 hour delay).
Heavier exposures have produced systemic effects with symptoms ranging from headache, nausea, and
lassitude to vomiting, chest or abdominal pain, respiratory depression, extreme weakness, stupor, convulsions
and death (dependent upon concentration and time).]
[Note: Purified DNPH, suitable for preparing cartridges, can be purchased commercially.]
9.2.1 Prepare a supersaturated solution of DNPH by boiling excess DNPH in 200 mL of acetonitrile for
approximately one hour.
9.2.2 After one hour, remove and transfer the supernatant to a covered beaker on a hot plate and allow
gradual cooling to 40-60 °C.
9.2.3 Maintain the solution at this temperature (40-60 °C) until 95% of solvent has evaporated.
9.2.4 Decant solution to waste, and rinse crystals twice with three times their apparent volume of
acetonitrile.
9.2.5 Transfer crystals to another clean beaker, add 200 mL of acetonitrile, heat to boiling, and again let
crystals grow slowly at 40-60°C until 95% of the solvent has evaporated.
9.2.6 Repeat rinsing process as described in Section 9.2.4.
9.2.7 Take an aliquot of the second rinse, dilute 10 times with acetonitrile, acidify with 1 mL of 3.8 M
perchloric acid per 100 mL of DNPH solution, and analyze by HPLC.
[Note: An acid is necessary to catalyze the reaction of the carbonyls with DNPH. Most strong inorganic
acids such as hydrochloric, sulfiiric, phosphoric, or perchloric acids will do the job. Perchloric or phosphoric
acids are the preferred catalyst for using acetonitrile solution of DNPH as the absorbing solution. The DNPH
derivatives do not precipitate from solution as readily as when hydrochloric or phosphoric acids are used as
the catalyst. This is an ideal situation for an HPLC analytical finish as this minimizes sample handling. For
most ambient air sampling, precipitation is not a problem because the carbonyl concentration is generally
in the ppb range.]
Page 11A-14 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde Method TO-11A
9.2.8 An impurity level of <0.15 ^g/cartridge of formaldehyde in DNPH-coated cartridge is acceptable
(based on the Certification Blank section 5.10). An acceptable impurity level for an intended sampling
application may be defined as the mass of the analyte (e.g., DNPH-formaldehyde derivative) in a unit volume
of the reagent solution equivalent to less than one tenth (0.1) the mass of the corresponding analyte from a
volume of an air sample when the carbonyl (e.g., formaldehyde) is collected as DNPH derivative in an equal
unit volume of the reagent solution. An impurity level unacceptable for a typical 10 L sample volume may be
acceptable if sample volume is increased to 100 L. If the impurity level is not acceptable for intended
sampling application, repeat recrystallization.
9.2.9 If the impurity level is not satisfactory, pipet off the solution to waste, then add 25 mL of acetonitrile
to the purified crystals. Repeat rinsing with 20 mL portions of acetonitrile until a satisfactorily low impurity
level in the supernatant is confirmed by HPLC analysis.
9.2.10 If the impurity level is satisfactory, add another 25 mL of acetonitrile, stopper and shake the reagent
bottle, then set aside. The saturated solution above the purified crystals is the stock DNPH reagent.
9.2.11 Maintain only a minimum volume of saturated solution adequate for day to day operation. This will
minimize wastage of purified reagent should it ever become necessary to re-rinse the crystals to decrease the
level of impurity for applications requiring more stringent purity specifications.
9.2.12 Use clean pipets when removing saturated DNPH stock solution for any analytical applications.
Do not pour the stock solution from the reagent bottle.
9.3 Preparation of DNPH-Formaldehyde Derivative
[Note: Purified crystals or solutions ofDNPH-derivatives can be purchased commercially.]
9.3.1 To a portion of the recrystallized DNPH, add sufficient 2N HC1 to obtain an approximately saturated
solution. Add to this solution formaldehyde (other aldehydes or ketones may be used if their detection is
desirable), in molar excess of the DNPH. Allow it to dry in air.
9.3.2 Filter the colored precipitate, wash with 2N HC1 and water and let the precipitate air dry.
9.3.3 Check the purity of the DNPH-formaldehyde derivative by melting point determination or HPLC
analysis. The DNPH-formaldehyde derivative should melt at 167°C ± 1°C. If the impurity level is not
acceptable, recrystallize the derivative in ethanol. Repeat purity check and recrystallization as necessary until
acceptable level of purity (e.g., 99%) is achieved.
9.3.4 DNPH derivatives of formaldehyde and other carbonyls suitable for use as standards are
commercially available both in the form of pure crystals and as individual or mixed stock solutions in
acetonitrile.
9.4 Preparation of DNPH-Formaldehyde Standards
9.4.1 Prepare a standard stock solution of the DNPH-formaldehyde derivative by dissolving accurately
weighed amounts in acetonitrile.
9.4.2 Prepare a working calibration standard mix from serial dilution of the standard stock solution. The
concentration of the DNPH-formaldehyde compound in the standard mix solutions should be adjusted to
reflect relative distribution in a real sample.
[Note: Individual stock solutions of approximately 100 mg/L are prepared by dissolving 10 mg of the solid
derivative in 100 mL of acetonitrile. The individual solution is used to prepare calibration standards
containing the derivative of interest at concentrations of 0.5-20 i^g/mL, which spans the concentration of
interest for most ambient air work.]
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-15
-------
Method TO-11A Formaldehyde
9.4.3 Store all standard solutions in a refrigerator. They should be stable at least one month.
9.4.4 DNPH-formaldehyde standards can also be purchased from various commercial suppliers. If
purchased, ensure that a "Certification of Concentration" is provided.
9.5 Preparation of DNPH-Coated Cartridges
[Note: This procedure must be performed in an atmosphere with a very low aldehyde background. All
glassware and plastic ware must be scrupulously cleaned and rinsed with deionized water and carbonylfree
acetonitrile. Contact of reagents with laboratory air must be minimized. Polyethylene gloves must be worn
when handling the cartridges. If the user wishes to purchase commercially prepared DNPH-coated
cartridges, they are available from various vendors. If commercial prepared DNPH-coated cartridges are
purchased, ensure that a "Certification Blank for Formaldehyde " is provided for the specific batch of which
that cartridge is a member. For a commercial cartridge to be acceptable, the following criteria must be met:
• Formaldehyde concentration: <0.15 ^g/cartridge.
If the enhanced carbonyl analysis is being performed, the following Certification Blank criteria must also be
met:
• Speciated carbonyl concentration:
- Acetaldehyde: <0.10 jug/cartridge
- Acetone: <0.30 ^g/'cartridge
- Other: <0.10 pg/cartridge
One who is not experienced in the preparation of DNPH-coated cartridge is strongly advised to use certified
commercially available cartridges.]
9.5.1 DNPH Coating Solution
9.5.1.1 Pipet 30 mL of saturated DNPH stock solution to a 1000 mL volumetric flask, then add 500 mL
acetonitrile.
9.5.1.2 Acidify with 1.0 mL of ortho-phosphoric acid (H3PO4).
[Note: The atmosphere above the acidified solution should preferably be filtered through a DNPH-coated
cartridge to minimize contamination from laboratory air. Shake solution, then make up to volume with
acetonitrile. Stopper the flask, invert and shake several times until the solution is homogeneous. Transfer the
acidified solution to a reagent bottle with a 0-10 mL range positive displacement dispenser.]
9.5.1.3 Prime the dispenser and slowly dispense 10-20 mL to waste.
9.5.1.4 Dispense an aliquot solution to a sample vial, and check the impurity level of the acidified
solution by HPLC according to Section 9.2.
9.5.1.5 The impurity level should be less than the Certification Blank of <0.15 ,wg/cartridge for
formaldehyde, similar to that in the DNPH coating solution.
9.5.2 Coating of Cartridges
9.5.2.1 Open the pre-packed cartridge package, connect the short end to a 10-mL syringe, and place it
in a syringe rack (see Figure 7).
[Note: Prepare as many cartridges (-100) and syringes as possible.]
Page 11A-16 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde Method TO-11A
9.5.2.2 Using a positive displacement repetitive pipet, add 10 mL of acetonitrile to each of the syringes
(see Figure 7).
9.5.2.3 Let liquid drain to waste by gravity.
[Note: Remove any air bubbles that may be trapped between the syringe and the silica cartridge by displacing
them with the acetonitrile in the syringe.]
9.5.2.4 Set the repetitive dispenser containing the acidified DNPH coating solution to dispense 7 mL
into the cartridges.
9.5.2.5 Once the effluent flow at the outlet of the cartridge has stopped, dispense 7 mL of the DNPH
coating reagent into each of the syringes (see Figure 7).
9.5.2.6 Let the coating reagent drain by gravity through the cartridge until flow at the other end of the
cartridge stops.
9.5.2.7 Wipe the excess liquid at the outlet of each of the cartridges with clean tissue paper.
9.5.2.8 Assemble a drying manifold with a scrubber or "guard cartridge" connected to each of the ports
(see Figure 7). These "guard cartridges" are DNPH-coated and serve to remove any trace of formaldehyde in
the nitrogen gas supply.
9.5.2.9 Insert cartridge connectors (flared at both ends, 0.64 by 2.5-cm outside diameter TFE-
fluorocarbon FEP tubing with inside diameter slightly smaller than the outside diameter of the cartridge port)
onto the long end of the scrubber cartridges.
9.5.2.10 Remove the cartridges from the syringes and connect the short ends to the exit end of the
scrubber cartridge.
9.5.2.11 Pass nitrogen through each of the cartridges at about 300-400 mL/min for 5-10 minutes.
9.5.2.12 Within 10 minutes of the drying process, rinse the exterior surfaces and outlet ends of the
cartridges with acetonitrile using a Pasteur pipet.
9.5.2.13 Stop the flow of nitrogen after 15 minutes, wipe the cartridge exterior free of rinsed acetonitrile
and remove the dried cartridge.
9.5.2.14 Plug both ends of the coated cartridge with standard polypropylene Luer® male plugs, place
the plugged cartridge in a shipping tube with polypropylene screw caps.
9.5.2.15 Put a serial number and a lot number label on each of the individual shipping tubes.
9.5.2.16 Store shipping tubes containing the DNPH-coated cartridges in a refrigerator at 4°C until use.
[Note: Plugged cartridges may also be placed in screw-capped glass culture tubes and placed in a
refrigerator until use. Cartridges will maintain their integrity for up to 90 days stored in refrigerated, capped
shipping tubes.]
9.5.2.17 Take a minimum of 3 blank cartridges from the cartridge batch and analyze for formaldehyde,
as delineated in Section 11. The batch of user-prepared DNPH-coated cartridges is acceptable if the following
criteria are met:
• Formaldehyde Certification Blank: <0.15 (jg/cartridge.
If the enhanced carbonyl analysis is being performed, the following certification criteria must also be met:
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-17
-------
Method TO-11A Formaldehyde
• Speciated carbonyl concentration:
- Acetaldehyde: <0.10 ng/cartridge
- Acetone: <0.30 (jg/cartridge
- Other: <0.10 (jg/cartridge
9.5.2.18 If analysis meets the above criteria, provide documentation with all cartridges associated with
that batch involving "Certification Blank for Formaldehyde. " This certificate must be part of the project
records.
9.5.2.19 If the cartridge results are close to, but above the Certification Blank, run a few more blank
cartridges to check background level.
9.5.2.20 If analysis indicates failure of the cartridge, then all cartridges in that batch are unacceptable.
Prepare anew batch of cartridges according to Section 9.5 until certification is achieved.
9.5.2.21 Store all certified cartridges in a refrigerator at 4°C until use.
9.5.2.22 Before transport, remove the shipping container (or screw-capped glass culture tubes)
containing the adsorbent tubes from the refrigerator and place culture tubes in a friction-top metal can
containing 1-2 inches of charcoal for shipment to sampling location. Alternately, acidified DNPH-coated filters
can be used in place of charcoal filters to remove impurity carbonyl compounds in the air.
9.5.2.23 As an alternative to friction-top cans for transporting sample cartridges, the coated cartridges
could be shipped in their individual glass containers (see Figure 5a). A batch of coated cartridges may also
be packed in a polypropylene shipping container for shipment to the field (see Figure 5b). The container
should be padded with clean tissue paper or polyethylene-air bubble padding. Do not use polyurethane foam
or newspaper as padding material.
9.5.2.24 The cartridges should be immediately stored in a refrigerator or freezer (<4°C) upon arrival
in the field.
9.6 Equivalent Formaldehyde Cartridge Concentration
9.6.1 One can calculate the equivalent formaldehyde background concentration (ppbv) contributed from
a commercial or user-prepared DNPH-coated cartridge following exposure to formaldehyde-free air.
9.6.2 The equivalent formaldehyde background concentration includes the contribution of formaldehyde
from both the acetonitrile and the cartridge.
9.6.3 Knowing the equivalent background concentration, as determined by the user (see Section 9.5.2)
or supplied by the commercial supplier (see Note. Section 9.5), of formaldehyde in the cartridge (ng/cartridge),
the formaldehyde background concentration contributed by the DNPH-coated cartridge (thus the method
minimum detection limits) can be related to the total sample volume, as identified in Table 3.
9.6.4 For example, if the averaged background formaldehyde concentration supplied by the manufacturer
is 70 ng/cartridge, then that cartridge can add 0.95 ppbv of equivalent formaldehyde, to the final ambient air
concentration value, as delineated in Table 3 for a total air volume of 60 L.
9.6.5 The user should use DNPH-coated cartridges with the lowest background concentration to improve
accuracy and detection limits.
10. Sampling Procedure
10.1 The sampling system is assembled and should be similar to that shown in Figures 3 and 4.
[Note: Figures 3 and 4 illustrate different tube/pump configurations. The tester should ensure that the pump
is capable of constant flow rate throughout the sampling period.]
Page 11A-18 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde Method TO-11A
It is recommended that the sampling system employ a heated inlet (~50°C) coupled to an ozone denuder or
scrubber to minimize water and ozone interference associated with the DNPH-coated adsorbent tube.
Historically, the coated cartridges have been used as direct probes and traps for sampling ambient air when
the ambient temperature was above freezing.
[Note: As illustrated in Figure 8, the ozone denuder has been effective for up to 80 hours without
breakthrough at ozone levels of approximately 700 ppb. Other studies have evaluated both denuders and
scrubbers at ozone concentrations between 125 and 200ppbv and found they have effectively removed ozone
from the air stream for up to 100,000 ppb-hours; however, moisture was required (-10% RH) in the gas
stream (26). The user should evaluate the length of time of the application of the denuder or scrubber to his
field work. Caution should be utilized when using these devices for extensive periods of time at high humidity
(>65%). Regarding the 24 hour samples, special caution shoud be taken while sampling nighttime periods
when relative humidities approaching 100% are frequently encountered. It is recommended that routine
schedule of ozone removal device replacement should be implemented as part of the sampling program.]
[Note: For sampling ambient air below freezing, a short length (30-60 cm) of heated (50-60 °F) stainless steel
tubing must be added to condition the air sample prior to collection on the DNPH-coated cartridges.]
10.2 Before sample collection, the system must be checked for leaks. Plug the inlet of the system so no flow
is indicated at the output end of the pump. The mass flow meter should not indicate any air flow through the
sampling apparatus.
10.3 Air flow through the DNPH-adsorbent cartridge may change during sampling as airborne particles
deposit on the front of the cartridge. The flow change could be significant when sampling particulate-laden
atmospheres. Particle concentrations greater than 50 ug/m3 are likely to represent a problem. For unattended
or extended sampling periods, a mass flow controller is highly recommended to maintain constant flow. The
mass flow controller should be set at least 20% below the maximum air flow through the cartridge.
10.4 The entire assembly (including a "test" sampling cartridge) is installed and the flow rate checked at a
value near the desired sampling rate. In general, flow rates of 1,000-2,000 mL/min should be employed. The
total sample volume should be selected to ensure that the collected formaldehyde concentration exceeds the
background formaldehyde DNPH-cartridge concentration, as illustrated in Table 3. The total moles of
carbonyl in the volume of air sampled should not exceed that of the DNPH concentration (i.e., 2 mg cartridge).
In general, a safe estimate of the sample size should be 75% of the DNPH loading of the cartridge.
[Note: If the user suspects that there will be breakthrough of a DNPH-coated cartridge during the sampling
event, a backup cartridge should be used during the first sampling event. One would analyze the back-up
cartridge for formaldehyde. If the back-up cartridge concentration exceeds 10% of the formaldehyde
concentration on the front cartridge, then continue to use back-up cartridges in the monitoring program.
However, if formaldehyde is not detected above the average blank level in the back-up cartridge after the first
sampling event, then one can continue to use only one cartridge under normal representative conditions.]
[Note: The SKC tube is a dual bed configuration, allowing one to analyze the back bed (see Figure 2) for
quantifying breakthrough.]
Generally, calibration is accomplished using a soap bubble flow meter or calibrated wet test meter connected
to the flow exit, assuming the system is sealed.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-19
-------
Method TO-11A Formaldehyde
[Note: ASTM Method D 3686 describes an appropriate calibration scheme that does not require a sealed flow
system downstream of the pump.]
10.5 The operator must measure and record the sampling flow rate at the beginning and end of the sampling
period to determine sample volume. A dry gas meter may be included in the system to measure total sample
volume and to compare against the in-line mass flow controller. Some commerical systems use flow monitors
with data loggers to make these measurements.
10.6 Before sampling, flush the inlet (denuder/manifold, etc.) for approximately 15 min at the established flow
rate to condition the system. Remove the glass culture tube from the friction-top metal can or styrofoam box.
Let the cartridge warm to ambient temperature in the glass tube before connecting it to the sample train.
10.7 Using polyethylene gloves, remove the DNPH-coated cartridge from the shipping container and connect
it to the sampling system with a Luer® adapter fitting. Most commercially available cartridges are
bidirectional. However, review manufacturer suggestions for orientation of the cartridge to the inlet of the
; OO O
sampler.
[Note: If using the SKC dual bed tube, ensure the ambient air is pulled through the tube in the direction
enscribed on the tube by an arrow.]
Record the following parameters on Compendium Method TO-11A field test data sheet (FTDS), as illustrated
in Figure 9: date, sampling location, time, ambient temperature, barometric pressure (if available), relative
humidity (if available), dry gas meter reading (if appropriate), flow rate, rotameter setting, cartridge batch
number, and dry gas meter pump identification numbers.
10.8 The sampler is turned on and the flow is adjusted to the desired rate. A typical flow rate through one
cartridge is 1.0 L/min and 0.8 L/min for two tandem cartridges.
10.9 The sampler is operated for the desired period, with periodic recording of the variables listed in Figure 9.
10.10 If the ambient air temperature during sampling is below 15 °C, a heated inlet probe is recommended.
However, no pronounced effect of relative humidity (between 25% - 90%) has been observed for sampling
under various weather conditions—cold, wet, and dry winter months and hot and humid summer months.
However, a negative bias has been observed when the relative humidity is <25%. At high humidity, the
possibility of condensation must be guarded against, especially when sampling is an air conditioned trailer.
10.11 At the end of the sampling period, the parameters discussed in Section 10.7 are recorded and the sample
flow is stopped. If a dry gas meter is not used, the flow rate must be checked at the end of the sampling
interval. If the flow rates at the beginning and end of the sampling period differ by more than 10%, the sample
should be marked as suspect.
10.12 Immediately after sampling, remove the cartridge (using polyethylene gloves) from the sampling system,
cap with Luer® end plugs, and place it back in the original labeled glass shipping container or culture tube.
Cap, seal with TFE-fluorocarbon tape, and place it in appropriate padding. Refrigerate at 4°C until analysis.
Refrigeration period prior to analysis should not exceed 2 weeks. If a longer storage period is expected, the
cartridge should be extracted with 5 mL of acetonitrile (see Section 11.2.4 and 11.2.5) and the eluant placed
in a vial for long term storage.
Page 11A-20 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde _ Method TO-11A
[Note: If samples are to be shipped to a central laboratory for analysis, the duration of the non-refrigerated
period should be kept to a minimum, preferably less than two days.]
10.13 If a dry gas meter or equivalent total flow indicator is not used, the average sample flow rate must be
calculated according to the following equation:
Q! + Q2 + QN
N
where:
QA = average flow rate, L/min.
Qi, Q2 QN = flow rates determined at beginning, end, and intermediate points during sampling,
L/min.
N= number of points averaged.
10.14 The total flow rate is then calculated using the following equation:
Vm = (T2-T1)xQA
where:
Vm = total volume sampled at measured temperature and pressure, L.
T2 = stop time, minutes.
Tj = start time, minutes.
T2 - Tj = total sampling time, minutes.
QA = average flow rate, L/min.
10.15 The total volume (Vs) at EPA standard conditions, 25°C and 760 mm Hg, is calculated from the
following equation:
„ „ „ PA „ 298
X
m 760 273 + T
where:
A
Vs = total sample volume at 25 °C and 760 mm Hg pressure, L.
Vm = total sample volume at measured temperature and pressure, L.
P A = average ambient pressure, mm Hg.
TA = average ambient temperature, °C.
11. Sample Analysis
11.1 Sample Preparation
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-21
-------
Method TO-11A Formaldehyde
11.1.1 The samples (trip blank, field blank and field samples) are returned to the laboratory in a shipping
container and stored in a refrigerator at (<4°C) until analysis. Alternatively, the samples may also be stored
alone in their individual containers.
11.1.2 The time between sampling and extraction should not exceed 2 weeks. Since background levels
may change during storage, always compare field samples to those associates with field and trip to a blank
samples, stored under the same conditions.
11.2 Sample Extraction
[Note: Beware of unintentional exposure of samplers and eluted samples to aldehyde and ketone sources.
Laboratory air often holds high concentrations of acetone. Labeling inks, adhesives, and packaging
containers (including vials with plastic caps) are all possible sources on contamination.]
[Note: Contamination is most likely to occur during sample extraction. Before eluting derivatives, clean all
glassware by rinsing with acetonitrile, then heating in a 60° C vacuum oven for at least 30 minutes. Eluting
the samples in a nitrogen-purged glove bag further reduces the risk of contamination.
The acetonitrile used to elute the DNPH derivatives is a typical source of contamination. Formaldehyde-free
acetonitrile used to elute samples should be used only for this purpose, and stored in a carbonyl free
environment. A concentrations of 10 ^g/L of any aldehyde or ketone in the acetonitrile adds 0.05 ^g of that
carbonyl to sampler blank values if using 5 mL extraction volumes.]
11.2.1 Remove the sample cartridge from the labeled shipping tube or container. Connect the sample
cartridge to a clean syringe.(Some commerical cartridges do not require the addition of a syringe for elution.)
[Note: The liquid flow during desorption should be in the reverse direction of air flow during sample
collection.]
11.2.2 Place the sample cartridge syringe in the syringe rack (see Figure 7).
[Note: If the two beds in the SKC tube are being recovered separately for breakthrough studies, break the
tube and place the beds in separate vials. Proceed with recovery, as specified in Section 11.2.3 through
Section 11.2.6.]
11.2.3 Backflush the cartridge (gravity feed) by passing 5 mL of acetonitrile from the syringe through the
cartridge to a 5-mL volumetric flask. The backflush elution approach may add particulate particles also
collected on the cartridge to the acetonitrile solution which can cause sample valve failure and increase column
back pressure. To minimize this, frontflush the cartridge contents with the acetonitrile reagent rather than
blackflush. The use of 5mL of acetonitrile is sufficient for quantitative cartridge sample elution in either mode.
[Note: A dry cartridge has an acetonitrile holdup volume of about 0.3 mL. The eluantflow may stop before
the acetonitrile in the syringe is completely drained into the cartridge because of air trapped between the
cartridge filter and the syringe Luer® tip. If this happens, displace the trapped air with the acetonitrile in the
syringe using a long-tip disposable Pasteurpipet.]
11.2.4 Dilute to the 5-mL mark with acetonitrile. Label the flask with sample identification. Store in
refrigerated conditions until the sample is analyzed by HPLC. Pipet two aliquots into sample vials with TFE-
Page 11A-22 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde Method TO-11A
fluorocarbon-lined septa. Analyze the first aliquot for the derivative carbonyls by HPLC. Store the second
aliquot in the refrigerator until the results of the analysis of the first aliquot are complete and validated. The
second aliquot can be used for confirmatory analysis, if necessary.
11.2.5 Sample eluates are stable at 4°C for up to one month.
11.3 HPLC Analysis
11.3.1 The HPLC system is assembled and calibrated as described in Section 11.4. The operating
parameters are as follows when formaldehyde is the only carbonyl of interest:
Column: Zorbax ODS (4.6-mm ID x 25-cm), or equivalent.
Mobile Phase: 60% acetonitrile/40% water, isocratic.
Detector: ultraviolet, operating at 360 nm.
Flow Rate: l.OmL/min.
Retention Time: 7 minutes for formaldehyde with one Zorbax ODS column. Thirteen
minutes for formaldehyde with two Zorbax ODS columns.
Sample Injection Volume: 25 i\L.
Before each analysis, the detector baseline is checked to ensure stable conditions.
11.3.2 The HPLC mobile phase is prepared by mixing 600 mL of acetonitrile and 400 mL of water. This
mixture is filtered through a 0.22-/mi polyester membrane filter in an all-glass and Teflon® suction filtration
apparatus. The filtered mobile phase is degassed by purging with helium for 10-15 minutes (100 mL/min) or
by heating to 60°C for 5-10 minutes in an Erlenmeyer flask covered with a watch glass. A constant back
pressure restrictor (350 kPa) or short length (15-30 cm) of 0.25-mm (0.01 inch) ID Teflon® tubing should be
placed after the detector to eliminate further mobile phase outgassing.
11.3.3 The mobile phase is placed in the HPLC solvent reservoir and the pump is set at a flow rate of 1.0
mL/min and allowed to pump for 20-30 minutes before the first analysis. The detector is switched on at least
30 minutes before the first analysis, and the detector output is displayed on a strip chart recorder or similar
output device. The isocratic flow of 60% acetonitrile/40% water is adequate for the analysis of formaldehyde;
however, sufficient time between air sample analyses is required to assure that all other carbonyl compounds
are eluted from the HPLC column prior to the next sample. The gradient flow approach ,mentioned later (see
Section 14.3) is properly programmed to elute other carbonyl compounds.
11.3.4 A 100-(iL aliquot of the sample is drawn into a clean HPLC injection syringe. The sample injection
loop (25-(iL) is loaded and an injection is made. The data system, if available, is activated simultaneously with
the injection. If a strip chart recorder is used, mark the point of injection on the chart paper.
11.3.5 After approximately one minute, the injection valve is returned to the "load" position and the
syringe and valve are rinsed or flushed with acetonitrile/water mixture in preparation for the next sample
analysis.
[Note: The flush/rinse solvent should not pass through the sample loop during flushing.]
The loop is cleaned while the valve is in the "load" mode.
11.3.6 After elution of the DNPH-formaldehyde derivative (see Figure 10), data acquisition is terminated
and the component concentrations are calculated as described in Section 12.
11.3.7 After a stable baseline is achieved, the system can be used for further sample analyses as described
above. Be sure to examine the chromatogram closely to ensure that background DNPH-formaldehyde
derivative peaks are not on the solvent slope of the DNPH peak.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-23
-------
Method TO-11A Formaldehyde
[Note: After several cartridge analyses, background buildup on the column may be removed by flushing with
several column volumes of 100% acetonitrile.]
11.3.8 If the concentration of analyte exceeds the linear range of the instrument, the sample should be
diluted with mobile phase, or a smaller volume can be injected into the HPLC.
11.3.9 If the retention time is not duplicated (±10%), the acetonitrile/water ratio may be increased or
decreased to obtain the correct elution time. If the elution time is too long, increase the ratio; if it is too short,
decrease the ratio. If retention time is not reproducing, the problem may be associated with the HPLC flow
system. A control chart is recommended to evaluate retention time changes.
[Note: The chromatographic conditions described here have been optimized for the detection of formaldehyde.
Analysts are advised to experiment with their HPLC system to optimize chromatographic conditions for their
particular analytical needs. If a solvent change is necessary, always recalibrate before running samples.]
11.4 HPLC Calibration
11.4.1 Calibration standards can be prepared by the user in acetonitrile from the solid
DNPH-formaldehyde derivative or liquid standards can be purchased from various manufacturers. From the
solid compound, individual stock solutions of 100 ug/mL are prepared by dissolving 10 mg of solid derivative
in 100 mL of acetonitrile. Since the MW of HCHO-hydrazone is 210 g/mol, and the MW of HCHO is 30
g/mol, the stock solution concentration converts to 14.3 ug/mL as formaldehyde (30/210 x lOOmg/mL). The
solid compound is weighed using a 5-place analytical balance and liquid dilutions are made with volumetric
glassware. Stock solutions obtained from commercial suppliers generally range from 1 to 50 ug/mL as the
carbonyl compound. These stock solutions are typically provided in 1 mL ampules.
11.4.2 Using the stock solution, working calibration standards are produced. To generate the highest
concentration working standard, use a pipette to quantitatively transfer 1.00 ml of the stock solution to a 25
mL volumetric flask. For example, using a 14.3 ug/mL stock solution produces a working standard solution
of 570 ng/mL ( 14300 ng/mL x 1/25 ). The high concentration working standard diluted serially, using 1 to
5 mL pipettes and volumetric flasks, can produce working standards ranging between 28.5 and 570 ng/mL.
11.4.3 Each calibration standard (at least five levels) is analyzed three times and area response is tabulated
against mass concentration injected (see Figure 11). All calibration runs are performed as described for sample
analyses in Section 11.3. The results are used to prepare a calibration curve, as illustrated in Figure 12. The
slope of the calibration curve gives the response factor, RF. Linear response is indicated where a correlation
coefficient of at least 0.999 for a linear least-squares fit of the data (mass concentration versus area response)
is obtained. The intercept of the calibration curve should pass through the origin. If it does not, check your
reagents and standard solutions preparation procedure for possible contamination. If the calibration curve does
not pass through the origin, the equation for the calibration curve should include the intercept.
11.4.4 Once linear response has been documented, a concentration standard near the anticipated levels
of each carbonyl component, but at least 10 times the detection limit, should be chosen for daily calibration.
The day to day response for the various components should be within 10% of the calibration value. If greater
variability is observed, prepare a fresh calibration check standard. If the variability using a freshly prepared
calibration check standard is greater than 15% , a new calibration curve must be developed from fresh
standards. A plot of the daily values on a Quality Control Chart (day versus concentration) is helpful to check
for long term drift of the concentration value.
11.4.5 The response for each component in the daily calibration standard is used to calculate a response
factor according to the following equation shown for formaldehyde:
Page 11A-24 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde Method TO-11A
RF
JxrHCHO
where:
RFHCHO = response factor for formaldehyde given as area counts per ng/mL.
CHCHO = concentration of analyte in the calibration standard in units of ng/mL.
P = peak area counts for the formaldehyde standard.
P0 = calibration curve intercept; in most cases this is zero.
11.4.2 The RF for each carbonyl compound is determined in the same way as that given for formaldehyde.
The concentration of HCHO and other carbonyl compounds is determined with the calibration curves for each
component in the analyzed sample. Example calculation for HCHO is given in section 12.
12. Calculations
Determination of the carbonyl compound air concentration requires three steps: (1) determination of the
average blank and the standard deviation of the blank; (2) determination of the collected carbonyl compound
mass of the cartridge; (3) calculation of the carbonyl compound air concentration. The following discusion
provides these steps for formaldehyde.
12.1 Blank Determination
Since the blank level for any arbitary cartridge is unknown, an average value for the blank is used in the
calculation. As noted earlier, the average blank value is determined for each lot of cartridges. For a given lot
size, N, a minimum of /N cartridge blanks (rounded to the next whole number) should be analyzed; i.e., for
a lot size of 200, a minimum of V~200 or 14 cartridge blanks should be analyzed. A minimum of 3 of these
blanks are used for the Certification Blank, and the remaining 1 1 are used for field blanks. The mass of HCHO
on each cartridge is determined by multiplying the observed peak area for blank cartridge solution by the
acetonotrile extract volume ( typically 5 mL ) and dividing by the response factor as provided in the following
equation:
x * E
where:
MBL_HCHOi = the blank HCHO mass for cartridge , i.
= HCHO response factor calculated in Section 1 1.4.5.
Oi = area counts for HCHO in blank sample extract.
VE = extract volume in mL ( usually 5 mL).
Once all blank cartridges have been measured, the average blank value is determined by the following
equation:
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-25
-------
Method TO-11A Formaldehyde
MBL-HCHO = T7 X E M
BL-HCHO;
where:
MBL-HCHO = me average HCHO mass for all cartridges.
MBL.HCHOi = blank HCHO mass for cartridge, i.
N = the number of blank cartridges.
[Note: Measurement of cartridge blanks should be destributed over the period that this particular cartridge
lot is used for ambient air sampling. It is recommended that a trend plot of blank results be constructed to
evaluate background carbonyl results over the period of cartridge lot utilization in the sampling program. If
significant drifting is observed, blank average values should be segmented to be more representative of
carbonyl background.]
12.2 Carbonyl Analyte Mass
The calculation equation for the mass of the collected carbonyl compound mass for an individual
cartridge is the as that for the cartrigde blanks. The gross measured carbonyl mass is determined with an
equation analogous to that given in section 12.1. The equation for formaldehyde is given as:
M
_1_V_I_Q A
RF
JxrHCHO
where:
MSAl = gross HCHO mass for cartridge, i.
PSAl = HCHO peak area counts for cartridge, i.
RFHCHO = the response factor for HCHO.
VE = acetonitrile extract volume in mL (typically 5 mL).
The net HCHO mass for an individual cartridge is determined by substracting the average blank value from
the gross HCHO mass obtained for sample i, and is given as:
~ ^SA, ~ ^BL-HCHO
Page 11A-26 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde Method TO-11A
12.3 Carbonyl Compound Concentration
The sample air concentration for carbonyl compounds cannot be determined directly from the mass
measurement and requires conversion to units of volume. The conversion calculation for HCHO is determined
using the ideal gas law and is given by the following equation:
MHCHO, _ _ N 760
x (R x TAMB) x
"AMB
where:
VHCHOi = gas volume of HCHO on cartridge, i.
MHCHOI = mass of HCHO on cartridge, i.
MW = molecular weight of HCHO, 30.03 g/mole.
R = gas constant, 0.082 L-atm/mol-deg.
TPMB = ambient air temperature in degrees Kelvin, 273 + T (C°).
PAMB = ambient air pressure in torr.
For an ambient air temperature of 25°C and a pressure of 760 torr, the ideal law equation reduces to:
= 1.2276 x MHCHQ
In this equation, the HCHO mass in ng is converted to a volume in nL. The volume of air that was passed
through the cartridge was measured by either a mass flow controller or dry test meter calibrated at a known
temperature and pressure. To determine HCHO concentration in the units of ppbv, apply the following
equation:
V
CHCHoPPbv = -
VAIR
where:
H0i = volume of formaldehyde in nL
= volume of sample air through the cartridge
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-27
-------
Method TO-11A Formaldehyde
13. Performance Criterial and Quality Assurance
This section summarizes required quality assurance measures and provides guidance concerning performance
criteria that should be achieved within each laboratory.
13.1 Standard Operating Procedures (SOPs).
13.1.1 Users should generate SOPs describing the following activities in their laboratory: (1) assembly,
calibration, and operation of the sampling system, with make and model of equipment used; (2) preparation,
purification, storage, and handling of sampling reagent and samples; (3) assembly, calibration, and operation
of the HPLC system, with make and model of equipment used; and (4) all aspects of data recording and
processing including lists of computer hardware and software used.
13.1.2 SOPs should provide specific stepwise instructions and should be readily available to and
understood by the laboratory personnel conducting the work.
13.2 HPLC System Performance
13.2.1 The general appearance of the HPLC system should be similar to that illustrated in Figure 1.
13.2.2 HPLC system efficiency is calculated according to the following equation:
N = 5.54
where:
N = column efficiency, theoretical plates.
tr = retention time of analyte, seconds.
W1/2 = width of component peak at half height, seconds.
A column efficiency of >5,000 theoretical plates should be utilized.
13.2.3 Precision of response for replicate HPLC injections should be ±10% or less, day to day, for analyte
calibration standards at 150 ng/mL or greater levels (as the carbonyl compound). At 75 ng/mL levels and
below, precision of replicate analyses could vary up to 25%. Precision of retention times should be ±7% on
a given day.
13.3 Process Blanks
13.3.1 At least one field blank should be used for each day of field sampling, shipped and analyzed with
each group of samples. The number of samples within a group and/or time frame should be recorded so that
a specified minimum number of blanks is obtained for a given cartridge lot used for field samples. The field
blank is treated identically to the samples except that no air is drawn through the cartridge. The performance
criteria described in Section 9.2 should be met for field blanks. It is also desirable to analyze trip and
laboratory blank cartridges as well, to distinguish between possible field and lab contamination.
[Note: Remember to use the field blank value for each cartridge lot when calculating concentration. Do not
mix cartridge lots in the blank value determinations ]
Page 11A-28 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde Method TO-11A
13.4 Method Precision and Accuracy
13.4.1 At least 50% of the sampling events should include a collocated sample. A collocated sample is
defined as a second sampling port off the common sampling manifold. If more than five samples are collected
per sampling event, a collocated sample should be collected for each sampling event. Precision for the
collocated samples should be ±20% or better. EPA historical data has demonstrated effectiveness in reaching
±20%, as illustrated in Figure 13.
13.4.2 Precision for replicate HPLC injections should be ±10% or better, day to day, for calibration
standards.
13.4.3 Cartridges spiked with analytes of interest can be used in round-robin studies to intercompare
several laboratories performing carbonyl analyses. The spiked samples are prepared in the laboratory by
spiking a blank cartridge with a solution of derivatized carbonyls in acetonitrile. The laboratory preparing the
spike samples should analyze at a minimum 3 of the prepared spiked samples to evaluate the consistency of
prepared samples.
13.4.4 Before initial use of the method, each laboratory should generate triplicate spiked samples at a
minimum of three concentration levels, bracketing the range of interest for each compound. Triplicate
nonspiked samples must also be processed. Spike recoveries of >80 ±10% and blank levels should be achieved.
13.4.5 For ambient air sampling, an ozone denuder must be used as part of the sampling system. As
discussed in Section 6.4, ozone effects the ultimate method precision and accuracy by reacting with its
carbonyl derivative (hydrazones) on the cartridge. To illustrate this point, Figure 14 documents the
concentration of formaldehyde captured on collocated DNPH-cartridges, one with a denuder (see Figure 14a)
and the other without a denuder (see Figure 14b). The formaldehyde peak is considerably higher with use of
an ozone denuder.
13.5 Method Detection Limits
13.5.1 Determine method detection limits using the procedures in 40 CFR Part 136B. Prepare a low level
standard of the carbonyl derivatives at a concentration within two to five times the estimated method detection
limit. Inject the standard into the analytical system seven times.
13.5.2 Calculate the measured concentration using the calibration curve.
13.5.3 Determine the standard deviation for the seven analyses and use the standard deviation to calculate
the detection limit as described in 40 CFR Part 136B.
13.6 General QA/QC Requirements
13.6.1 General QA/QC requirements associated with the performance of Compendium Method TO-11A
include:
Sampling
• Each sampling event, flow calibration with bubble meter, both pre- and post-checks.
• Mass flow meter calibration factor determined every quarter.
• Each sampling event, leak check, both pre- and post-checks.
• 10 percent of field samples collocated to help calculate method precision and evaluate biases.
• 10 percent of field samples operated with back-up cartridge to evaluate analyte breakthrough.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-29
-------
Method TO-11A Formaldehyde
• Field and trip (optional) blank cartridges are included with each field sample collection program.
• Sample volumes calculated and reviewed project QA officer.
Reagents
• Coating solution prepared from concentrated stock solution immediately before each coating.
• Solution analyzed before each coating to determine acceptability (less than 0.10 (jg/cartridge for
each aldehyde), control chart of contaminant concentration maintained.
• Three blank cartridges per lot for immediate elution/analysis to determine Certification Blank for
the carbonyl compounds.
Analysis
• Multi point calibration curve performed each six months.
• Continuing calibration standard (mid-level) analyzer every analytical run to evaluate precision, peak
resolution and retention time drift.
• Method detection limits (MDLs) verified annually or after each instrument change.
• Replicate analysis of approximately 10 percent of sample eluents to evaluate precision.
• Samples quantitated against least squares calibration line.
• Performance evaluation (PE) sample acquired from independent sources analyzed prior to and after
field samples.
• Random collocated samples shipped to independent laboratory for analysis and compared to in-
house collocated sample.
• Testing of acetonitrile used for sample extraction for background carbonyl evaluation.
Data Acquisition
• Sample chromatograms and standards checked daily for peak shape and integration quality,
resolution of carbonyls, overall sensitivity and retention time drift.
• Separate tape backups made of raw data immediately after completion of each analysis.
• Peaks in each sample checked for correct ID and integration using system software before export
to ASCII file.
• Final results checked and edited by project QA officer before producing final report.
• Tape backups of final data files produced.
13.6.2 All results should be reviewed by the project QA officer, independent of the field and
laboratory operations, to evaluate the overall adherence to the methodology in meeting the program data quality
objectives (DQOs).
14. Detection of Other Aldehydes and Ketones
14.1 Introduction
14.1.1 The procedure outlined above has been written specifically for the sampling and analysis of
formaldehyde in ambient air using an adsorbent cartridge and HPLC. Ambient air contains other aldehydes
and ketones. Optimizing chromatographic conditions by using two Zorbax ODS columns in series and varying
the mobile phase composition through a gradient program will enable the analysis of other aldehydes and
ketones. Alternatively, other aldehydes and ketones may also be analyzed using a single C-18, reverse phase
column and a ternary gradient as described by Waters or Smith, et al. (J. Chromatography, 483, 1989, 431-
436). Thus, other aldehydes and ketones can be detected with a modification of the basic procedure.
Page 11A-30 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde Method TO-11A
14.1.2 In particular, chromatographic conditions can be optimized to separate acetaldehyde, acetone,
propionaldehyde, and some higher molecular weight carbonyls within an analysis time of about 1 h by utilizing
two Zorbax ODS columns in series, and a linear mobile phase program. Operating the HPLC in a gradient
mode with one Zorbax ODS column may also provide adequate resolution and separation. Carbonyl
compounds covered within the scope of this modification include:
Formaldehyde Crotonaldehyde
o-Tolualdehyde
Aceteldehyde Butyraldehyde
ra-Tolualdehyde
Acetone Benzaldehyde
/>-Tolualdehyde
Propionaldehdye Isovaleraldehyde
Hexanaldehyde
Valeraldehyde 2,5-Dimethylbenzaldehyde Methyl ethyl ketone
14.1.3 The linear gradient program varies the mobile phase composition periodically to achieve maximum
resolution of the C-3, C-4 and benzaldehyde region of the chromatogram. The following gradient program
was found to be adequate to achieve this goal: Upon sample injection, linear gradient from 65% acetonitrile
(ACN)/35% water to 55% ACN/45% water in 36 min; to 100% ACN in 20 min; 100% ACN for 5 min; reverse
linear gradient from 100% ACN to 60% ACN/40% water in 1 min; maintain at 60% ACN/40% water for 15
min.
14.2 Sampling Procedures
Same as Section 10.
14.3 HPLC Analysis
14.3.1 The FiPLC system is assembled and calibrated as described in Section 11. The operating parameters
are as follows:
Column: Zorbax ODS, two columns in series
Mobile Phase: Acetonitrile/water, linear gradient
Step 1. 60-75% acetonitrile/40-25% water in 30 minutes.
Step 2. 75-100% acetonitrile/25-0% water in 20 minutes.
Step 3. 100% acetonitrile for 5 minutes.
Step 4. 60% acetonitrile/40% water reverse gradient in 1 minute.
Step 5. 60% acetonitrile/40% water, isocratic, for 15 minutes.
Detector: Ultraviolet, operating at 360 nm
Flow Rate: l.OmL/min
Sample Injection Volume: 25 yL
14.3.2 The gradient program allows for optimization of chromatographic conditions to separate
acetaldehyde, acetone, propionaldehyde, and other higher molecular weight aldehydes and ketones in an
analysis time of about one hour.
14.3.3 The chromatographic conditions described here have been optimized for a gradient HPLC
system equipped with a UV detector (variable wavelength), an automatic sampler with a 25-^\L loop injector
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-31
-------
Method TO-11A Formaldehyde
and two DuPont Zorbax ODS columns (4.6 x 250-mm), a recorder, and an electronic integrator. Analysts are
advised to experiment with their HPLC systems to optimize chromatographic conditions for their particular
analytical needs. Highest chromatographic resolution and sensitivity are desirable but may not be achieved.
The separation of acetaldehyde, acetone, and propionaldehyde should be a minimum goal of the optimization.
14.3.4 The carbonyl compounds in the sample are identified and quantified by comparing their
retention times and area counts with those of standard DNPH derivatives. Formaldehyde, acetaldehyde,
acetone, propionaldehyde, crotonaldehyde, benzaldehyde, and o-, m-, p-tolualdehydes can be identified with
a high degree of confidence. The identification of butyraldehyde is less certain because it coelutes with
isobutyraldehyde and is only partially resolved from methyl ethyl ketone under the stated chromatographic
conditions. A typical chromatogram obtained with the gradient HPLC system for detection of other aldehydes
and ketones is illustrated in Figure 15.
14.3.5 The concentrations of individual carbonyl compounds are determined as outlined in Section 12.
14.3.6 Performance criteria and quality assurance activities should meet those requirements outlined
in Section 13.
15. Precision and Bias
15.1 This test method has been evaluated by round robin testing. It has also been used by two different
laboratories for analysis of over 1,500 measurements of formaldehyde and other aldehydes in ambient air for
EPA's Urban Air Toxics Program (UATP), conducted in 14 cities throughout the United States.
15.2 The precision of 45 replicate HPLC injections of a stock solution of formaldehyde-DNPH derivative
over a 2-month period has been shown to be 0.85% relative standard deviation (RSD).
15.3 Triplicate analyses of each of twelve identical samples of exposed DNPH cartridges provided
formaldehyde measurements that agreed within 10.9% RSD.
15.4 A total of 16 laboratories in the U.S., Canada, and Europe participated in a round robin test that
included 250 blank DNPH-cartridges, three sets of 30 cartridges spiked at three levels with DNPH derivatives,
and 13 sets of cartridges exposed to diluted automobile exhaust gas. All round robin samples were randomly
distributed to the participating laboratories. A summary of the round robin results is shown in Table 4.
15.5 The absolute percent differences between collocated duplicate sample sets from the 1988 UATP
program were 11.8% for formaldehyde («=405), 14.5% for acetaldehyde («=386), and 16.7% for acetone
(n=346).
15.6 Collocated duplicate samples collected in the 1989 UATP program and analyzed by a different
laboratory showed a mean RSD of 0.07, correlation coefficient of 0.98, and bias of-0.05 for formaldehyde.
Corresponding values for acetaldehyde were 0.12, 0.95 and -0.54, respectively. In the 1988 UATP program,
single laboratory analyses of spiked DNPH cartridges provided over the year showed an average bias of+6.2%
for formaldehyde («=14) and +13.8% for acetaldehyde («=13).
15.7 Single laboratory analyses of 30 spiked DNPH cartridges during the 1989 UATP program showed an
average bias of +1.0% (range -49 to +28%) for formaldehyde and 5.1% (range -38% to +39%) for
acetaldehyde.
16. References
Page 11A-32 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde Method TO-11A
1. Riggin, R. M., "Determination of Aldehydes and Ketones in Ambient Air: Method TO-5," in
Compendium of Methods for the Determination of Toxic Organic Compounds in Ambient Air, U. S.
Environmental Protection Agency, EPA-600/4-84-041, Research Triangle Park, NC, April 1984.
2. Winberry, W. T. Jr., et al., "Determination of Formaldehyde and Other Aldehydes in Indoor Air Using
Passive Sampling Device, Method IP-6C," in Compendium of Methods for the Determination of Air
Pollutants in Indoor Air, U. S. Environmental Protection Agency, EPA-600/4-90-010, May 1990.
3. Tejada, S.B., "Standard Operating Procedure For DNPH-coated Silica Cartridges For Sampling
Carbonyl Compounds In Air And Analysis by High-performance Liquid Chromatography,"
Unpublished, U. S. Environmental Protection Agency, Research Triangle Park, NC, March 1986.
4. Tejada, S.B., "Evaluation of Silica Gel Cartridges Coated in situ with Acidified
2,4-Dinitrophenylhydrazine for Sampling Aldehydes and Ketones in Air," Intern. J. Environ. Anal.
Chem., Vol. 26:167-185, 1986.
5. Winberry, W. T. Jr., et al., "Determination of Formaldehyde in Ambient Air Using Adsorbent
Cartridge Followed by HPLC: Method TO-11," in Compendium of Methods for the Determination
of Toxic Organic Compounds in Ambient Air, Second Supplement, U. S. Environmental Protection
Agency, EPA-600/4-89-018, Research Triangle Park, NC, March 1989.
6. Winberry, W. T. Jr., et al., "Determination of Formaldehyde and Other Aldehydes in Indoor Air Using
a Solid Adsorbent Cartridge: Method IP-6A," in Compendium of Methods for the Determination of
Air Pollutants in Indoor Air, U. S. Environmental Protection Agency, EPA-600/4-90-010, May 1990.
7. Nolan, L., et al., "Monitoring Carbonyls in Ambient Air Using the New Supelclean™ LPD (Low
Pressure Drop) DNPH Cartridge," in Proceedings of the 1995 EPA/AWMA International Symposium
on Measurement of Toxic and Related Air Pollutants, VTP-50, pp. 279, May 1995.
8. Quality Assurance Handbook for Air Pollution Measurement Systems, Volume II - Ambient Air
Specific Methods, EPA-600/R-94-038b, U. S. Environmental Protection Agency, Research Triangle
Park,NC, May 1994.
9. Technical Assistance Document for Sampling and Analysis ofO3 Precursors, U. S. Environmental
Protection Agency, EPA-600/8-9-215, Research Triangle Park, NC, October 1991.
10. Ahonen, I., Priha, E., and Aijala, M-L, "Specificity of Analytical Methods Used to Determine the
Concentration of Formaldehyde in Workroom Air," Chemosphere,Vo\. 13:521-525, 1984.
11. Levin, J. O., et al., "Determination of Sub-part-per-Million Levels of Formaldehyde in Air Using
Active or Passive Sampling on 2,4-Dinitrophenylhydrazine-Coated Glass Fiber Filters and
High-Performance Liquid Chromatography," Anal. Chem., Vol. 57:1032-1035, 1985.
12. Perez, J. M., Lipari, F., and Seizinger, D. E., "Cooperative Development of Analytical Methods for
Diesel Emissions and Particulates - Solvent Extractions, Aldehydes and Sulfate Methods", presented
at the Society of Automotive Engineers International Congress and Exposition, Detroit, MI,
February-March 1984.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-33
-------
Method TO-11A Formaldehyde
13. Kring, E. V., et al., "Sampling for Formaldehyde in Workplace and Ambient Air Environments-
Additional Laboratory Validation and Field Verification of a Passive Air Monitoring Device
Compared with Conventional Sampling Methods," J. Am. Ind. Hyg. Assoc., Vol. 45:318-324, 1984.
14. Sirju, A., and Shepson, P. B., "Laboratory and Field Evaluation of the DNPH-Cartridge Technique
for the Measurement of Atmospheric Carbonyl Compounds," Environ. Sci. Technol. Vol. 29:384-392,
1995.
15. Chasz, E. et al., "Philadelphia Air Management Lab, Summary of Procedures and Analytical Data for
Enhanced Ambient Monitoring of PAMS Carbonyls," in Proceedings of the 1995 EPA/AWMA
International Symposium on Measurement of Toxic and Related Air Pollutants, VIP-50, pp. 293, May
1995.
16. Grosjean, D., "Ambient Levels of Formaldehyde, Acetaldehyde, and Formic Acid in Southern
California: Results of a One-Year Base-Line Study," Environ. Sci. Technol., Vol. 25, 710-715, 1991.
17. Bufalini, J.J., and Brubaker, K.L., "The Photooxidation of Formaldehyde at Low Pressures." In:
Chemical Reaction in Urban Atmospheres, (C.S. Tuesday), American Elsevier Publishing Co., New
York, pp. 225-240. 1971.
18. Altshuller, A.P., and Cohen, I.R., "Photooxidation of Formaldehyde in the Pressence of Aliphatic
Aldehydes", Science, Vol. 7:1043-1049, 1963.
19. "Formaldehyde and Other Aldehydes," Committee on Aldehydes, Board of Toxicology and
Environmental Hazards, National Research Council, National Academy Press, Washington, DC, 1981.
20. Altshuller, A. P., "Production of Aldehydes as Primary Emissions and Secondary Atmospheric
Reactions of Alkenes and Alkanes During the Night and Early Morning Hours," Atmos. Environ., Vol.
27A:21-31, 1993.
21. Tanner, R L., et al., "Atmospheric Chemistry of Aldehydes; Enhanced PAN Formation From Ethanol
Fuel Vehicles," Environ. Sci. Technol., Vol. 22:1026-1034, 1988.
22. Ciccioli, P., and Cecinato, A., "Advanced methods for the Evaluation of Atmospheric Pollutants
Relevant to Photochemical Smog and Dry Acid Deposition: Chapter 11" in Gaseous Pollutants:
Characterization and Cycling, edited by Jerome O. Nriagu, ISBN 0-471-54898-7, John Wiley and
Sons, Inc., 1992.
23. Parmar, S. S., et al., "A Study of Ozone Interferences in Carbonyl Monitoring Using DNPH Coated
C18 and Silica Cartridges," in Proceedings of the 1995 EPA/AWMA International Symposium on
Measurement of Toxic and Related Air Pollutants, VIP-50, pp. 306, May 1995.
24. Parmar, S. S., et al., "Effect of Acidity on the Sampling and analysis of Carbonyls Using DNPH
Derivatization Method," in Proceedings of the 1996 EPA/AWMA International Symposium on
Measurement of Toxic and Related Air Pollutants, VIP-64, pp. 311, May 1996.
Page 11A-34 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde Method TO-11A
25. Arnts, R. R, and Tejada, S. B., "2,4-Dinitrophenylhydrazine-Coated Silica Gel Cartridge Method for
Determination of Formaldehyde in Air: Identification of an Ozone Interference," Environ. Sci.
Technol. Vol. 23:1428-1430, 1989.
26. Kleindienst, T. E., et al., "Measurement of Q-C4 Carbonyls on DNPH-Coated Silica Gel and Cg
Cartridges in the Presence of Ozone," in Proceedings of the 1995 EPA/AWMA International
Symposium on Measurement of Toxic and Related Air Pollutants, VIP-50, pp 29T, May 1995.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants Page 11A-35
-------
Method TO-11A
Formaldehyde
TABLE 1. COMPARISON OF DNPH COATED CARTRIDGES: SILICA GEL VS. CIS
Topic
Background
Breakthrough
Ozone interference
Extraneous chromato-
graphic peaks
Comparison
Silica gel < CIS
Silica gel < CIS
Silica gel CIS
Silica gel CIS
Discussion
Silica gel is purer, therefore less background contamination
from acetone and formaldehyde as compared to CIS.
CIS allows carbonyl compounds to breakthrough easier with
longer sampling periods, thus causing bias results. CIS has a
lower capacity for carbonyls in general. Loading of DNPH on
CIS plays an important role in breakthrough for carbonyls.
Ozone interference with silica gel is documented. Ozone
interference with CIS is not clear at this time. Therefore, must
use denuder with both systems.
Researchers have detected extraneous peaks in the
chromatography of both CIS and silica gel when ozone is
present.
TABLE 2. TYPICAL DNPH-CARTRIDGE SPECIFICATIONS
Category
Adsorbent
Particle size
DNPH loading1
Bed weight2
Capacity
Background
(per cartridge)
Pressure drop
Sampling temperature
Collection efficiency
Solvent hold-up volume
Tube dimensions
Typical Specifications
chromatographic grade silica or CIS coated with 2,4-
dinitrophenylhydrazine (DNPH)
150-1000 jim (60/100 mesh to 18/35 mesh)
0.3-0.9% (-1-3 mg/cartridge)
approx. 350 mg
approx. 75 jig formaldehyde, assuming a 50% consumption of DNPH
<0.15 fig formaldehyde
<0.10 jig acetaldehyde
<0.10 //g other carbonyls
<0.30 jig acetone
7 inches of water @ 0.5 L/min
15 inches of water @ 1.0 L/min
37 inches of water (fb, 2.0 L/min
10°CtolOO°C
>95% for formaldehyde for sampling rates up to 2.0 L/min
-1.0 mL
From -2 inches to -5 inches in length
-1 inch O.D. at widest point
'Loading is variable among commercial suppliers.
2The SKC tube is a dual bed cartridge with 300 mg of DNPH-coated silica gel in the front bed and 150 mg of
DNPH-coated silica gel in the back bed.
Page 11A-36
Compendium of Methods for Toxic Organic Air Pollutants
January 1997
-------
Formaldehyde
Method TO-11A
TABLE 3. EQUIVALENT FORMALDEHYDE CONCENTRATION (ppbv) RELATED TO
BACKGROUND FORMALDEHYDE CONCENTRATION (ng/cartridge)
Equivalent formaldehyde concentration
(ppbv)
Background formaldehyde
cartridge concentration,
ng/cartridge 70
100
150
Sample volume, L
60
0.950
1.358
2.037
120
0.475
0.679
1.018
180
0.317
0.453
0.679
1440
0.040
0.057
0.085
TABLE 4. ROUND ROBIN TEST RESULTS3
Sample Type
Blank cartridges:
ug aldehyde
(% RSD)
n
Spikedb cartridges:
% recovery (%
RSD)
low
medium
high
n
Exhaust samples:
jig aldehyde
% RSD
n
Formaldehyde
0.13
46
33
89.0 (6.02)
97.2 (3.56)
97.5 (2.15)
12
5.926
12.6
31
Acetaldehyde
0.18
70
33
92.6 (13.8)
97.8 (7.98)
102.2 (6.93)
13
7.990
16.54
32
Propionaldehyde
0.12
47
23
108.7 (32.6)
100.9 (13.2)
100.1 (6.77)
12
0.522
26.4
32
Benz aldehyde
0.06
44
8
114.7 (36.1)
123.5 (10.4)
120.0 (8.21)
14
0.288
19.4
17
tistics shown after removal of outliers.
aSixteen participating laboratories. Sta
bNormal spiking levels were approximately 0.5, 5 and 10 ug of aldehyde, designated as low, medium, and high in this table.
January 1997
Compendium of Methods for Toxic Organic Air Pollutants
Page 11A-37
-------
Method TO-11A
Formaldehyde
INJECTION VALVE
WATER
ACN
GUARD COLUMN
WASTE
- SAMPLE LOOP
AUTOSAMPLER
ANALYTICAL
COLUMN
DETECTOR
360 nm
RECORDER
INTEGRATOR
WASTE
Figure 1. Basic high-performance liquid chromatographic (HPLC) system used for carbonyl analysis.
Page 11A-38
Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde
Method TO-11A
SKC
Precision-
Sealed Tips
Glass
Tube
High-Purity
Glass Wool
DNPH-Coated Adsorbent
(Back Bed)
DNPH-Coated Adsorbent
(Front Bed)
Precision
Lockspring
NIOSH-Approved
Sealing Caps
SUPELCO Luer®
Fitting
AAC
DNPH-C-18 Cartridge
Luer®
Fitting
Luer®
Fitting
Luer®
Fitting
DNPH- Cartridge
DNPH - Coaled Adsorbent
WATERS
DNPH-Silica Gel Cartridge
Luer®
Fitting
DNPH-
Luer®
Fitting
Cartridge
Figure 2. Example of commercially available DNPH-cartridges.
January 1997
Compendium of Methods for Toxic Organic Air Pollutants
Page 11A-39
-------
Method TO-11A
Formaldehyde
Mass Flow
Controller
Sampling Pump
DNPH-Cpated
Sampling
Cartridge
Shut-off
Valve
Flushing Pump
Ozone
Denuder or
Scrubber
Needle Valve
a
c
05
05
-------
Formaldehyde
Method TO-11A
Sample Inlet
Heated Sample Inlet
Heated Zone
Ozone l_r
Denuder
Timer-controlled
Solenoid Valves
0.5-1.0LPM
A
DNPH Cartridge
1 LPM
Figure 4. Example of components of an automated multi-port sampler for
carbonyls monitoring using DNPH-coated cartridges.
January 1997 Compendium of Methods for Toxic Organic Air Pollutants
Page 11A-41
-------
Method TO-11A
Formaldehyde
Culture Tube
and Cap
Glass
Wool
DNPH
Adsorbent
Tube
(a) DNPH-cartridge in culture shipping tube
Packing
Foam
DNPH Polypropylene
Adsorbent Shipping Cap
Tubes
Polypropylene
Shipping
Container
(b) DNPH-cartridge in polypropylene shipping container
Figure 5. Example of commercially available shipping containers for DNPH cartridges.
Page 11A-42 Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde
Method TO-11A
3' Potassium
Iodide Coated
1/4" O.D. Copper
Tubing
Top View (cutaway)
(a) Cross-sectional view of EPA's ozone denuder assembly
Female Luer
Potassium Iodide __^_^
Cartridge Ozone \ | -4
Scrubber
Polyethylene Frit
Male Luer
(b) Commercially available packed granular potassium iodide (Kl) ozone scrubber
Figure 6. Example of (a) cross-sectional view of EPA's ozone denuder assembly, and
(b) commercially available packed granular potassium iodide (KI) ozone scrubber.
January 1997
Compendium of Methods for Toxic Organic Air Pollutants
Page 11A-43
-------
Method TO-11A
Formaldehyde
10 ml Glass
Syringe
Uncoated
Sample Cartridges
1
2
1
^
\
fs
i
1,
*//jh
w
>fc
y
Vff\
w
A K
n n
yy/////A
u u
,
y
fe?
M
y
-^
\
t
1
I
1
%
^ Test Tu
Rack
Waste
Beaker
V//////A
(a) Rack for Coating Cartridges
Syringe Fitting
Guard Cartridge
Sample Cartridge-
Waste Vial
'•^
J
1_
(b) Rack for Drying DNPH-Coated Cartridges
'N2Gas Stream
DNPH-Coated
Cartridges
Figure 7. Example of a typical syringe rack for coating (a) and drying (b) sample cartridges.
Page 11A-44
Compendium of Methods for Toxic Organic Air Pollutants
January 1997
-------
Formaldehyde
Method TO-11A
5
Q.
Q_
0
*j
2
§
c
0
O
(D
1
O
&\JV
800
700
600
500
400
300
200
100
0
•mn
r^nnf* <^t inniit f^nrl nf rl^niiHp*r
*
ozone at output end of denuder .
\ /
/
/
0
20 40 60 80
Sampling Time (hours)
100
Figure 8. Example of capacity of 3' x 0.25" O.D. x 4.6-mm ID. copper KI ozone
denuder at 2 L/min flow.
January 1997
Compendium of Methods for Toxic Organic Air Pollutants
Page 11A-45
-------
Method TO-11A
Formaldehyde
I. GENERAL INFORMATION
PROJECT:
SITE:
COMPENDIUM METHOD TO-11A
CARBONYL SAMPLING FIELD TEST DATA SHEET
(One Sample per Data Sheet)
DATES(S) SAMPLED: _
LOCATION:
INSTRUMENT MODEL NO.:
PUMP SERIAL NO.:
TIME PERIOD SAMPLED:
OPERATOR:
CALIBRATED BY:
OZONE DENUDER USE TIME (Hr):
HEATED INLET: YES NO
ADSORBENT CARTRIDGE INFORMATION:
Type:
Adsorbent:
Serial Number:
Sample Number:
II. SAMPLING DATA INFORMATION
Start Time:
Stop Time:
Time
Avg.
Dry Gas
Meter
Reading
Rotameter
Reading
Flow Rate,
*Q mL/min
Ambient
Temperature,
°C
Barometric
Pressure,
mm Hg
Relative
Humidity, %
Comments
* Flow rate from rotameter or soap bubble calibrator (specify which).
Total Volume Data (¥„,) (use data from dry gras meter, if available)
Vm = (Final - Initial) Dry Gas Meter Reading, or
or
1
1000 x (Sampling Time in Minutes)
III. COMMENTS
Figure 9. Example of Compendium Method TO-11A field test data sheet (FTDS).
Page 11A-46
Compendium of Methods for Toxic Organic Air Pollutants
January 1997
-------
Formaldehyde
Method TO-11A
OPERATING PARAMETERS
HPLC
Column: Zorbax ODS or C-18 RP
Mobile Phase: 60% Acetonitrile/40% Water
Detector: Ultraviolet, operating at 360 nm
Flow Rate: 1 mL/min
Retention Time: ~ 7 minutes for formaldehyde
Sample Injection Volume: 25 L
o
CD
"c"
CD
_C
"CD
c/>
CO
CD
.CD
_Q
£
V)
1
0
10
TIME, min
20
Figure 10. Example of chromatogram of DNPH-formaldehyde derivative.
January 1997
Compendium of Methods for Toxic Organic Air Pollutants
Page 11A-47
-------
Method TO-11A
Formaldehyde
OPERATING PARAMETERS HPLC
Column: Zorbax ODS or C-18 RP
Mobile Phase: 60% Acetontrile/40% Water
Detector: Ultraviolet, operating at 360 nm
Flow Rate: 1 mL/min
Retention Time: ~ 7 minutes for formaldehyde
Sample Injection Volume: 25 juL
Peak
a
b
c
d
e
Cone.
MS/mL
0.61
1.23
6.16
12.32
18.48
Area Counts
226541
452186
2257271
4711408
6053812
o
0
o
0
O
0
O
0
O
0
Time
Figure 11. Example of HPLC chromatogram of varying
concentration of DNPH-formaldehyde derivative.
Page 11A-48
Compendium of Methods for Toxic Organic Air Pollutants
January 1997
-------
Formaldehyde
Method TO-11A
O
o
o
§•
o
o
<
LU
o
o
8-
I
3
CORRELATION COEFFICIENT:
0.9999
OPERATING PARAMETERS
HPLC
Column: Zorbax ODS or C-18 RP
Mobile Phase: 60% Acetonitrile/40% Water
Detector: Ultraviolet, operating at 360 nm
Flow Rate: 1 mL/min
Retention Time: ~ 7 minutes for formaldehyde
Sample Injection Volume: 25 uL
i
6
12 15
!
18
DNPH - Formaldehyde Derivative (ug/mL)
Figure 12. Example of calibration curve for formaldehyde.
January 1997
Compendium of Methods for Toxic Organic Air Pollutants
Page 11A-49
-------
Method TO-11A
Formaldehyde
+20% precision
limits
Figure 13. Historical data associated with collocated samples for
formaldehyde (ppbv) in establishing 20% precision.
Page 11A-50
Compendium of Methods for Toxic Organic Air Pollutants January 1997
-------
Formaldehyde
Method TO-11A
0
x = Unknown
0 = DNPH
1 = Formaldehyde
I
03
JD
0
CO
_Q
-------
DNPH
PEAK IDENTIFICATION
>
Ul
!
I
I
1 2
14
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Compound
Formaldehyde
Acetaldehyde
Acrolein
Acetone
Propionaldehyde
Crotonaldehyde
Butyraldehyde
Benzaldehyde
Isovaleraldehyde
Valeraldehyde
o-Tolualdehyde
m-Tolualdehyde
p-Tolualdehyde
Hexaldehyde
2,4-Dimethylben-
zaldehyde
Concentration
ng/mL
1.140
1.000
1.000
1.000
1.000
1.000
0.905
1.000
0.450
0.485
0.515
0.505
0.510
1.000
0.510
o
o.
H
O
i
10
20
30
40
50
TIME, min
e
o
•«
Figure 15. Typical chromatogram of a linear gradient program for analyzing other aldehydes/ketones from a DNPH-coated cartridge.
O.
CD
------- |