vvEPA
United States
Environmental Protection
Agency
Technical Assistance Document (TAD) for
Precursor Gas Measurements in the NCore
Multi-pollutant Monitoring Network
Version 4
-------
EPA-454/R-05-003
September 2005
Technical Assistance Document (TAD)
For Precursor Gas Measurements in the NCore Multi-Pollutant
Monitoring Network
VERSION 4
Prepared by:
BATTELLE
505 King Avenue
Columbus, Ohio 43201-2693
Prepared for:
Vickie Presnell, Project Officer
Nealson Watkins, Work Assignment Manager
Emissions, Monitoring, and Analysis Division
Contract No. 68-D-02-061
Work Assignment 3-02
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Emissions, Monitoring, and Analysis Division
Research Triangle Park, North Carolina 27711
-------
DISCLAIMER
This document is a work prepared for the United States Government by
Battelle. In no event shall either the United States Government or Battelle have
any responsibility or liability for any consequences of any use, misuse, inability to
use, or reliance upon the information contained herein, nor does either warrant or
otherwise represent in any way the accuracy, adequacy, efficacy, or applicability
of the contents hereof.
in
-------
ACKNOWLEDGMENTS
Significant contributions to the TAD content were made by Kevin Cavender, Dennis
Mikel, Michael Papp, Joann Rice, Solomon Ricks, Nealson Watkins, and Lewis Weinstock of
the U.S. EPA, Office of Air Quality Planning and Standards, and Anna Kelly, of the Hamilton
County Department of Environmental Services (Ohio).
The U.S. Environmental Protection Agency wishes to acknowledge the assistance and
input provided by the following advisors in the preparation of this guidance document: George
Allen, of the Northeast States of Coordinated Air Use Management (NESCAUM); James
Schwab, of the State University of New York at Albany; William McClenny of the U.S. EPA
Office of Research and Development; and Eric Edgerton of Atmospheric Research and Analysis,
Inc.
IV
-------
TABLE OF CONTENTS
ACKNOWLEDGMENTS iv
ACRONYMS ix
1.0 INTRODUCTION 1
1.1 Development of the NCore Network 2
1.2 The Need for High Sensitivity Precursor Gas DatainNCore 3
1.3 Precursor Gas Monitoring in the NCore Network 5
1.4 Data Quality Objectives 6
1.5 Format and Purpose 7
1.6 References 8
2.0 HIGH SENSITIVITY CARBON MONOXIDE MEASUREMENTS 1
2.1 Introduction 1
2.1.1 Properties of Carbon Monoxide 1
2.1.2 Sources and Sinks of Carbon Monoxide 1
2.1.3 Historical Overview of CO Measurement Method 2
2.2 Summary of Method 3
2.2.1 CO Measurement by NDIR Spectrophotometry 3
2.2.2 Gas Filter Correlation (GFC) 5
2.3 Recommendations for NCore 6
2.3.1 Recommended Method Performance Criteria 7
2.3.2 Recommended Features for High Sensitivity CO Measurements 13
2.3.3 Commercial High Sensitivity GFC CO Monitors 14
2.3.4 Sampling Requirements 19
2.4 Potential Problems and Solutions 21
2.4.1 Interferences and Sources of Bias 21
2.4.2 Detector Stability 23
2.5 Supporting Equipment 23
2.5.1 Data Acquisition Device 23
2.5.2 Calibration Equipment 23
2.6 Reagents and Standards 25
2.6.1 Calibration Standards 25
2.6.2 Zero Air 26
2.7 Quality Control 26
2.7.1 Site Visit Checks and Remote Diagnostic Checks 26
2.7.2 Multipoint Calibrations 26
2.7.3 Level 1 Zero/Span Checks 27
2.7.4 Precision Checks 29
2.8 Preventive Maintenance and Troubleshooting 29
2.8.1 Preventive Maintenance 30
2.8.2 Troubleshooting 31
2.9 References 32
-------
3.0 HIGH SENSITIVITY SULFUR DIOXIDE MEASUREMENTS 1
3.1 Introduction 1
3.1.1 Properties of SO2 1
3.1.2 Historical Overview of Measurement Method 2
3.2 Summary of SO2 Measurement by UV Fluorescence 2
3.3 Recommendations for NCore 4
3.3.1 Recommended Method Performance Criteria 4
3.3.2 Recommended Features for High Sensitivity SO2 Measurements 10
3.3.3 Commercial High Sensitivity UV Fluorescence SO2 Monitors 12
3.3.4 Sampling Requirements 16
3.4 Potential Problems and Solutions 18
3.4.1 Sources of Positive Interference or Bias 19
3.4.2 Sources of Negative Interference or Bias 20
3.5 Supporting Equipment 20
3.5.1 Data Acquisition Device 21
3.5.2 Calibration Equipment 21
3.6 Reagents and Standards 23
3.6.1 Calibration Standards 23
3.6.2 Zero Air 24
3.7 Quality Control 24
3.7.1 Site Visit Checklists and Remote Diagnostic Checks 24
3.7.2 Multipoint Calibrations 24
3.7.3 Level 1 Zero/Span Checks 25
3.7.4 Precision Checks 27
3.8 Preventive Maintenance and Troubleshooting 27
3.8.1 Preventive Maintenance 28
3.8.2 Troubleshooting 29
3.9 References 30
4.0 HIGH SENSITIVITY TOTAL REACTIVE NITROGEN OXIDES MEASUREMENTS 1
4.1 Introduction 1
4.1.1 Properties of NOy 1
4.1.2 Sources of NOy 2
4.1.3 Historical Overview of NOy Measurement Method 3
4.2 Summary of NOy Measurement by Chemiluminescence Method 4
4.3 Recommendations for NCore 8
4.3.1 Recommended Method Performance Criteria 8
4.3.2 Recommended Features for High Sensitivity Ambient NOy Measurements 14
4.3.3 Commercial ChemiluminescentNOy Monitors 16
4.3.4 Sampling Requirements 22
4.4 Potential Problems and Solutions 24
4.4.1 Interferences 25
4.4.2 Converter Efficiency 26
4.5 Equipment and Supplies 28
4.5.1 Data Acquisition Device 28
4.5.2 Calibration Equipment 29
VI
-------
4.6 Reagents and Standards 30
4.6.1 Calibration Standards 30
4.6.2 Zero Air 31
4.7 Quality Control 31
4.7.1 Site Visit Checklists and Remote Diagnostic Checks 31
4.7.2 Multipoint Calibrations 31
4.7.3 Level 1 Zero/Span Checks 33
4.7.4 Precision Checks 35
4.8 Preventive Maintenance and Troubleshooting 35
4.8.1 Preventive Maintenance 35
4.8.2 Troubleshooting 36
4.9 References 38
5.0 QUALITY ASSURANCE/QUALITY CONTROL (QA/QC) PROCEDURES 1
5.1 Introduction 1
5.2 QA/QC Management 1
5.3 Network Calibration and Instrument Check Procedures 2
5.3.1 Multipoint Calibrations 2
5.3.2 Level 1 Zero/Span Checks 5
5.3.3 Precision Checks 7
5.4 Independent Audits and Assessments 7
5.4.1 Proficiency Test Samples 8
5.4.2 Technical Systems Audit 9
5.4.3 Audits of Data Quality 10
5.5 References 10
6.0 DATA ACQUISITION AND MANAGEMENT 1
6.1 Introduction 1
6.2 Data Acquisition and Analysis 1
6.2.1 Example Data Logger: ESC 8832 Data System Controller 2
6.2.2 Example Environmental Data System: ENVIDAS System 4
6.2.3 Summary Data Acquisition Process 6
6.3 Data Acquisition System Quality Assurance 7
6.3.1 Personnel 7
6.3.2 Security 8
6.3.3 Data Entry and Formatting 8
6.3.4 Data Review 9
6.3.5 Calibrations and Audits 10
6.4 Data and Records Management 11
6.4.1 Calibration Data 11
6.4.2 Electronic Data Files 12
6.4.3 Hard Copies 12
6.5 References 13
Vll
-------
Appendix A Sample Manifold Design for Precursor Gas Monitoring A-1
Appendix B Standard Operating Procedures for Selected Trace Level Precursor Gas
Monitoring and Calibration Equipment B-l
LIST OF TABLES
2-1 Ecotech Model EC9830T High Sensitivity CO Analyzer Specifications 16
2-2 Thermo Electron Model 48C TLE CO Analyzer Specifications 17
2-3 Teledyne API Model 300E High Sensitivity CO Analyzer Specifications 18
2-4 Example of a Preventive Maintenance Schedule for High Sensitivity CO Analyzers ..30
2-5 Instrument Troubleshooting for High Sensitivity CO Analyzers 31
3-1 Ecotech EC9850T SO2 Analyzer Specifications 13
3-2 Thermo Electron Model 43C TLE SO2 Analyzer Specifications 15
3-3 Teledyne API Model 100AS SO2 Analyzer Specifications 16
3-4 Example of a Preventive Maintenance Schedule for High Sensitivity SO2 Analyzers . 27
3-5 Instrument Troubleshooting for Precursor SO2 Analyzers 29
4-1 Thermo Electron Model 42C-YNOy Analyzer Specifications 17
4-2 API Model 200AU/501 NOy Analyzer Specifications 18
4-3 Ecotech EC9841-NOy Analyzer Specifications 20
4-4 ECO PHYSICS CLD 88 p and CLD 780 TR Analyzer Specifications 22
4-5 Example of a Preventive Maintenance Schedule for NOy Monitoring 35
4-6 Instrument Troubleshooting for High Sensitivity NOy Analyzers 36
5-1 Concentration Levels for Biweekly Precision Checks 6
5-2 Concentration Ranges for PT Samples 7
6-1 Example Internal Diagnostic Parameters of High Sensitivity Precursor Gas
Analyzers Accessible to a Digital Data Acquisition System 6
LIST OF FIGURES
2-1 General Schematic of a Typical GFC CO Analyzer 5
2-2 Ecotech EC9830T High Sensitivity CO Analyzer 15
2-3 Thermo Electron Model 48C TLE CO Analyzer 17
2-4 Teledyne API Model 3OOE CO Analyzer 18
3-1 Schematic Illustration of the Optical Chamber of a Precursor SO2 Analyzer 3
3-2 Ecotech EC9850T SO2 Analyzer 13
3-3 Thermo Electron Model 43C TLE SO2 Analyzer 14
3-4 Teledyne API Model 100AS SO2 Analyzer 16
4-1 General Schematic of a Typical Chemiluminescence NOy Instrument 7
4-2 Thermo Electron Model 42C-Y NOy Analyzer 17
4-3 API Model 200AU/501 NOy Analyzer 18
4-4 Ecotech EC9841-NOy Analyzer 20
4-5 ECO PHYSICS Model CLD 88 p and CON 765 NOy Converter 22
6-1 ESC 8832 Data System Controller 3
6-2 Flow of Data from Precursor Gas Analyzers to Final Reporting 7
Vlll
-------
ACRONYMS
CFR
CO
EPA
FPD
FRM
GFC
HCN
IR
LDL
MDL
Mo
N2
NAAMS
NAAQS
NAMS
NCore
NDIR
NH4+
NIST
NO
NO2
NOX
NOy
NPN
Os
Pd
Pt
Code of Federal Regulations
carbon monoxide
U.S. Environmental Protection Agency
flame photometric detection
Federal Reference Method
gas filter correlation
hydrogen cyanide
infrared
lower detectable limit
method detection limit
molybdenum
nitrogen
National Ambient Air Monitoring Strategy
National Ambient Air Quality Standards
National Air Monitoring Station
national core monitoring network
non-dispersive infrared
ammonia
ammonium
National Institute of Standards and Technology
nitrogen oxide
nitrogen dioxide
nitrogen oxides
reactive nitrogen oxides
n-propyl nitrate
ozone
palladium
platinum
IX
-------
PM particulate matter
PMT photomultiplier tube
ppb part per billion
ppm part per million
ppt part per trillion
QA quality assurance
QC quality control
SLAMS State and Local Air Monitoring Station
SC>2 sulfur dioxide
SOP standard operating procedure
TAD technical assistance document
VOC volatile organic compound
UV ultraviolet
-------
(This page has been intentionally left blank.)
XI
-------
Precursor Gas TAD
Section 1 Introduction
Version 4
September 2005
Page Iof8
TECHNICAL ASSISTANCE DOCUMENT
FOR PRECURSOR GAS MEASUREMENTS IN THE NCORE MULTI-POLLUTANT
MONITORING NETWORK
1.0 INTRODUCTION
The U.S. EPA is currently implementing the National Ambient Air Monitoring Strategy
(NAAMS).[1] The NAAMS goals include improvement of the scientific and technical
competency of the nation's air monitoring networks and increased value in protecting public
health and the environment. Monitoring of ambient air pollution is a critical part of the nation's
air program infrastructure. Monitoring data are used to characterize air quality and associated
health and ecosystem impacts, develop emission strategies to reduce impacts, and account for
progress over time. Substantial improvements in ambient air quality have been observed over
the last two decades, despite increases in the U.S. population, vehicle usage, and industrial
productivity. Ambient concentrations of several of the criteria air pollutants [specifically, lead
(Pb), carbon monoxide (CO), sulfur dioxide (802), and nitrogen dioxide (NC^)] are now well
below the applicable National Ambient Air Quality Standards (NAAQS).
While the obvious problems of widespread elevated concentrations have been largely
solved for some criteria pollutants, problems related to particulate matter (PM), ozone (O3), and
toxic air pollutants remain. It is now clear that even very low air pollution levels can be
associated with adverse environmental and human health effects. As a result, new approaches in
air monitoring are needed to measure these low levels and to incorporate these measurements
with other data into comprehensive assessments of human and environmental health.
One of the major areas of investment in the NAAMS is the use of highly sensitive
commercial air pollutant monitors for the characterization of the precursor gases CO, SO2, and
total reactive oxides of nitrogen (NOy) in a new national core monitoring network (NCore). The
high sensitivity CO and SO2 analyzers are fundamentally the same as those designated as Federal
Reference and Equivalent methods (http://www.epa.gov/ttn/amtic/criteria.html), but with
-------
Precursor Gas TAD
Section 1 Introduction
Version 4
September 2005
Page 2 of 8
modifications to improve sensitivity and accuracy or reduce interferences. The use of such
precursor gas analyzers in the NCore network will still allow determination of compliance with
the NAAQS, but will provide measurements at much lower detection limits than are achievable
by current monitors. This capability for accurate measurements at low concentrations will
support long-term epidemiological studies, reduce uncertainties in data for modeling of air
pollution episodes, and support source apportionment and observational analyses.
The implementation of high sensitivity monitoring for CO, SC>2, and NOy in the NCore
network will require installation of new analyzers at selected sites, and implementation of new
monitoring, calibration, and data acquisition procedures. The purpose of this Technical
Assistance Document (TAD) is to provide state, local, and tribal (S/L/T) agencies with guidance
on the equipment, procedures, data acquisition, and quality assurance/quality control (QA/QC)
efforts needed to properly implement high sensitivity precursor gas monitoring.
1.1 Development of the NCore Network
NCore is both a repackaging and an enhancement of existing networks. The emphasis on
the term "Core" reflects a multi-faceted, multi-pollutant national network that can be
complemented by more specific efforts, such as intensive field campaigns to understand
atmospheric processes, or personal and indoor measurements to assess human exposure and
health effects. The NCore network will replace the current National Air Monitoring Station
(NAMS) and State and Local Air Monitoring Station (SLAMS) programs, and leverages all of
the major existing networks to produce an integrated multi-pollutant approach to air monitoring.
Emphasis is placed on a backbone of multi-pollutant sites, continuous monitoring methods, and
measurement of important pollutants other than the criteria pollutants (e.g., ammonia and NOy).
When complete, NCore will meet a number of important data needs: improved flow and timely
reporting of data to the public, including supporting air quality forecasting and information
systems such as AIRNow; continued determination of NAAQS compliance; improved
development of emissions control strategies; enhanced accountability for the effectiveness of
-------
Precursor Gas TAD
Section 1 Introduction
Version 4
September 2005
Page 3 of 8
emission control programs; and more complete information for scientific, public health, and
ecosystem assessments. Structurally, NCore will establish three levels of monitoring sites:
• Level 1- a small number of research-oriented sites accommodating the greatest
diversity of instrumentation with specific targeted objectives, reasonably analogous to
the current PM Supersite program;
• Level 2 - the backbone network of approximately 75 long-term, nationwide
multi-pollutant sites, encompassing both urban (about 55 sites) and rural (about
20 sites) locations;
• Level 3 - sites focused primarily on specific pollutants of greatest concern (PM and
Os), with as few as one measured parameter. It is estimated that over 1,000 Level 3
sites will be part of NCore.
While each of these three levels has specific objectives, there will likely be a continuum of site
capabilities. Level 2 sites, for example, may meet a minimum level of multi-pollutant
measurements, or may be augmented as necessary with other measurements so that the most
heavily equipped Level 2 sites approach Level 1 in scope. Similarly, Level 3 sites may be single
pollutant sites, but as necessary, may be augmented by other monitors to approach Level 2 site
capabilities. These variations will be dictated by the needs of the particular area or agency
responsible for air monitoring programs. The Level 2 sites are the primary platform for new
implementation of high sensitivity precursor gas monitors.
1.2 The Need for High Sensitivity Precursor Gas Data in NCore
The precursor gases CO, 862, and NOy play important roles in the formation of
atmospheric ozone, air toxics, and PM, on both local and regional scales. This interconnection
among distinct air quality issues requires an integrated multiple pollutant air quality monitoring
and management approach. For example, multi-pollutant monitoring data can allow health
-------
Precursor Gas TAD
Section 1 Introduction
Version 4
September 2005
Page 4 of 8
studies to separate confounding effects, particularly in the face of varying ambient
concentrations and PM composition. Air quality models and source attribution methods can
benefit because multi-pollutant data allow more robust diagnostic evaluations (i.e., ensuring the
model produces correct results for the correct reasons). Many emission sources release multiple
pollutants, and consequently source apportionment models may yield more conclusive results
with the use of multi- pollutant monitoring data. The NCore Level 2 sites are focused on
providing multi-pollutant monitoring data to address these issues.
Equally important is that monitoring in NCore accurately quantify the low precursor gas
concentrations that often typify conditions across the U.S. National health assessments and air
quality model evaluations require data that are representative of broad urban (e.g., 5 to 40 km)
and regional/rural (> 50 km) spatial scales, and long-term epidemiological studies must represent
a variety of airshed characteristics across different population regimes. These requirements drive
the mixed urban/rural placement of the NCore Level 2 sites. The NCore sites thus should be
perceived as developing a representative report card on air quality across the nation, capable of
delineating differences among geographic and climatological regions. While relatively high
precursor gas concentrations may once have characterized all urban areas, emission reductions
have changed that situation. For example, even in Atlanta, GA, median precursor gas
concentrations are only a few times the detection limits of the conventional monitors currently in
use.[2] Characterization of rural/regional environments is also important to understanding
background conditions, transport corridors, regional-urban dynamics, and influences of global
transport, as air quality modeling domains continue to expand. Localized source-oriented
dispersion modeling evolved throughout the 1970's and 80's into broader urban scale modeling
(e.g., EKMA and Urban airshed modeling for ozone), then into Regional approaches in the
1980's and 1990's (e.g., the Regional Oxidant (ROM) and Acid Deposition (RADM) Models),
and currently into national scale approaches (e.g., the Community Multiscale Air Quality
(CMAQ) models). This movement toward broader spatial scale modeling coincides with
increased recognition of the importance of the regional/rural transport environment on urban
conditions. As peak urban air pollution levels decline, rural and regional levels also decline. For
example, in rural Centreville, AL, median precursor gas concentrations are at or below the
-------
Precursor Gas TAD
Section 1 Introduction
Version 4
September 2005
Page 5 of 8
detection limits of the conventional monitors currently in use.[2] Measurements of low precursor
gas concentrations thus are needed throughout the NCore network so that models can incorporate
these low concentrations accurately.
The overarching objective of the high sensitivity precursor gas monitoring in NCore is to
determine concentrations in well-mixed representative rural and urban atmospheres. NCore
provides an opportunity to address new directions in monitoring and to begin to fill measurement
and technological gaps that have accumulated in the existing networks. The NAAMS recognizes
that there are both nationally and locally oriented objectives in monitoring that require different
design approaches. The multipollutant high sensitivity monitoring approach in NCore addresses
national level needs and makes the most of available resources.
1.3 Precursor Gas Monitoring in the NCore Network
The use of high sensitivity precursor gas monitors for the characterization of CO, 862,
and NOy at NCore Level 2 monitoring stations is one of the major areas of investment for the
NAAMS. In most cases, the adoption of improved precursor gas monitoring methods and
associated calibration procedures will be necessary given the low levels of these pollutants at
many of the likely NCore Level 2 sites. The use of high sensitivity monitors will also be
valuable at Level 2 sites in urban areas, since at many such locations the concentrations of these
gases are not consistently at elevated levels. In addition, as emissions reductions are realized and
concentrations shift downward, high sensitivity monitors in urban areas will support the
detection of trends. The applicability of high sensitivity monitors to urban Level 2 sites will be
addressed on a site-by-site basis.
Precursor gas monitoring of CO, SO2, and NOy at NCore Level 2 sites will be
accomplished by use of commercially available, continuous high sensitivity monitors. This
document provides the technical guidance needed for implementation of these monitors in the
NCore network. The following is a brief summary of the measurement principles and limitations
of these monitors:
-------
Precursor Gas TAD
Section 1 Introduction
Version 4
September 2005
Page 6 of 8
• CO: Commercially available, non-dispersive infrared (NDIR) monitors will be used
that include modifications to enhance performance and sensitivity. The principal
constraints on lower detection limits of these devices are water vapor interference and
background drift. These limitations are reduced by drying the sample air and by
automated determination of the monitor's baseline reading (i.e., auto-zeroing) using
an appropriate CO oxidation system.
• SO2: Commercially available, ultraviolet (UV) fluorescence monitors will be used
that include modifications to enhance performance and sensitivity. A more intense
UV light source and improved optical filtering to minimize interference from nitric
oxide (NO), are modifications typically made to increase sensitivity of these
monitors. As with CO measurements, precursor SO2 measurements also may be
affected by water vapor and background drift.
• NOy: Commercially available monitors based on the chemiluminescent reaction of
NO with Os will be used. Since the chemiluminescence method detects only NO,
other trace nitrogen species including NO2 must first be converted to NO in order to
be measured using this method. Consequently, appropriate means of sampling and
converting the chemical species that constitute NOy are critical to accurate
measurements. Typical improvements made to these monitors for high sensitivity
measurements include increased sample flow rate, placement of the converter at the
sample inlet, improved cooling of the detector, and reduction of interferences through
a prereactor for baseline determination.
1.4 Data Quality Objectives
Data Quality Objectives (DQOs) are qualitative and quantitative statements that clarify
the monitoring objectives, define the appropriate type of data, and specify the tolerable levels of
measurement errors for the monitoring program. By applying the DQO process to the
-------
Precursor Gas TAD
Section 1 Introduction
Version 4
September 2005
Page 7 of 8
development of a quality system for NCore precursor gas monitoring, EPA guards against
committing resources to data collection efforts that might not support the decisions for which the
data are collected. The DQO process is iterative and allows for the incorporation of new
information and modifying outputs from previous steps as inputs for a subsequent step.
The final outcome of the DQO process is a design for collecting data (e.g., the type of
measurements or number of samples to collect, and when, where, and how to collect samples),
together with limits on the probabilities of making decision errors. During calendar year 2005 a
Workgroup made up of personnel representing decision makers, technical experts, quality
assurance manager, and statisticians will work their way through the seven step DQO process to
address requirements for precursor gas monitoring in NCore. The results of that process will be
reflected in additional guidance provided for the NCore network.
1.5 Format and Purpose
Section 2 of this TAD provides detailed procedures for the use and calibration of high
sensitivity CO monitoring equipment at Level 2 NCore sites. Sections 3 and 4 provide the same
information for high sensitivity SO2 and NOy measurements, respectively. Section 5 of this TAD
describes the QA/QC procedures needed to support precursor gas monitoring at the NCore
Level 2 sites, and Section 6 provides guidance about the acquisition and management of data
from the precursor gas monitors. Each major section concludes with a list of the references cited
in that section. Appendix A of this TAD is a description prepared by EPA of sampling manifold
designs for precursor gas monitoring.
Although this TAD is not intended to be a Standard Operating Procedure (SOP), several
topics addressed here are discussed in the 2001 EPA guidance on preparing Standard Operating
Procedures (SOPs).[3] S/L/T agencies will find this document useful in preparing SOPs for the
specific precursor gas analyzers employed at their sites. Furthermore, SOP's prepared by EPA
for selected precursor gas monitoring and calibration equipment are included as Appendix B.
Electronic versions of these SOP's are also located at
http ://www. epa. gov/ttn/amtic/precursop. html.
-------
Precursor Gas TAD
Section 1 Introduction
Version 4
September 2005
Page 8 of 8
This TAD was written to help field operators, data managers, and data users understand
why (not just how) key procedures are performed, what issues exist with the analyzers that they
should be aware of, and how to address these issues. Special attention is paid to interferences,
equipment selection, and calibration procedures. Users of this TAD should also consult Part 1,
General Principles, of EPA's Quality Assurance Handbook, Volume II,[4] which contains
detailed information pertinent to all measurement methods.
1.6 References
1. "National Ambient Air Monitoring Strategy," U.S. Environmental Protection Agency,
Office of Air Quality Planning and Standards, Research Triangle Park, NC 27711,
April 2004, Final Draft.
2. Data from sites in the Southeastern Aerosol Research and Characterization (SEARCH)
Study, summarized in personal communication by Eric Edgerton, Atmospheric Research
and Analysis, Inc., June 10, 2005. SEARCH data are available at
http://www.atmospheric-research.com/studies/SEARCH/index.html.
3. "Guidance for Preparing Standard Operating Procedures (SOPs)," EPA QA/G-6,
EPA/240/B-01/004, U.S. Environmental Protection Agency, Office of Environmental
Information, Washington, DC, 20460, March 2001.
4. "Quality Assurance Handbook for Air Pollution Measurements," Volume II, EPA-454/R-
98-004, U. S. Environmental Protection Agency, Research Triangle Park, NC 27711,
August 1998.
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page Iof33
2.0 HIGH SENSITIVITY CARBON MONOXIDE MEASUREMENTS
2.1 Introduction
The NAAMS[1] builds upon research from the late 1990s, which indicated that carbon
monoxide (CO) could be used as a background tracer for industrial processes and as a world-
wide tracer for long-range pollution transport.[2"4] The Strategy also calls for monitoring to
ascertain whether CO can be used as an indicator for volatile organic compounds (VOCs).
In response to the need for ambient CO monitoring, researchers and academia have
worked with commercial instrument manufacturers to create high sensitivity CO instruments,
with manufacturer-stated detection limits as low as 0.04 parts per million (ppm) (40 parts per
billion (ppb)) in air.
2.1.1 Properties of Carbon Monoxide
CO is a colorless, odorless, tasteless, and highly poisonous gas. It affects the oxygen
carrying capacity of the blood by diffusing through the alveolar walls of the lungs and competing
with oxygen for the four iron sites in the hemoglobin molecule. Since the affinity of the iron site
for CO is approximately 240 times greater than for oxygen[5] even low levels of CO can cause a
number of symptoms including headache, mental dullness, dizziness, weakness, nausea,
vomiting, and loss of muscular control. In extreme cases, collapse, unconsciousness, and death
can occur. CO is only slightly soluble in water (2.3 ml/100 ml H2O at 20 °C and 760 mm Hg)[5]
and consequently CO is not readily deposited or washed out of the atmosphere.
2.1.2 Sources and Sinks of Carbon Monoxide
CO has both natural and anthropogenic sources. The two primary CO production
mechanisms are the oxidation of hydrocarbons in the atmosphere and the combustion of
carbonaceous fuels. Global background concentrations of CO typically fall between 50 and
120 ppb and fluctuate seasonally, as well as geographically. In general, higher concentrations
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 2 of 3 3
occur during winter and are found in the northern hemisphere, due to the preponderance of
anthropogenic sources in that part of the world. [5] CO is also produced naturally from the
photochemical oxidation of methane (CFLt) and other hydrocarbons in the atmosphere.
In urban environments, human activity results in CO levels that can greatly exceed
background levels. Along with carbon dioxide (CO2), CO is a primary gaseous product of the
combustion of carbonaceous fuels and is emitted from both mobile and stationary sources. In
most urban areas, automobiles are a substantial source of CO. Consequently, CO levels tend to
build up during morning and evening commute times, and maximum mixing ratios exceeding 50
ppm have been observed in congested intersections. Typically, however, one-hour average
concentrations greater than 10 ppm are rare.
CO is relatively stable in the atmosphere and has an average global atmospheric lifetime
of between 1 and 4 months. [5] The primary mechanism for removal of CO from the atmosphere
is the oxidative reaction with hydroxyl radical to form CO2:
CO + OH'
The hydroxyl radical coincidently is part of a chain reaction that also produces CO from
hydrocarbons in the atmosphere.
2.1.3 Historical Overview of CO Measurement Method
The standard reference method for the determination of ambient CO is non-dispersive
infrared spectrophotometry (NDIR). The NDIR CO measurement principle is the absorption of
infrared (IR) radiation, with a wavelength of 4.7 micrometers (|J,m), by CO. The first
instrumental method introduced, the Luft-type instrument, was granted Federal Reference
Method (FRM) designation in 1976. In 1981, instrument manufacturers developed several
modifications of the NDIR FRM and submitted those modified instruments for FRM
designation. The gas filter correlation (GFC) method became the most popular modification in
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 3 of 3 3
the monitoring community because of its improved linearity and detection capabilities. EPA
sponsored the development of the GFC technique at Ford AeroNutronics in Newport Beach, CA.
Instrument manufacturers used the original analyzer developed in that work as a guide for
designing commercial versions. Today, FRM-designated instruments are available from several
manufacturers using both Luft-type and GFC methods, however in ambient applications the GFC
method is almost exclusively used.
The lower detectable limit (LDL) for an ambient monitor is defined as that minimum
concentration level that produces a signal of twice the baseline noise level (Code of Federal
Regulations, Volume 40, Part 53.23c, or, in the shortened format used hereafter, 40 CFR 53.23c).
Modifications in commercial GFC CO instruments have lowered estimated LDL values to 0.04
ppm (40 ppb). This chapter focuses on the GFC method and on those modifications to the GFC
method that allow for high sensitivity CO measurements.
2.2 Summary of Method
2.2.1 CO Measurement by NDIR Spectrophotometry
GFC analyzers operate on the principle that the CO molecule has a sufficiently
characteristic IR absorption spectrum that absorption can be used as a measure of CO
concentration in the presence of other gases. CO absorbs IR radiation maximally at a
wavelength of 4.7 |j,m, which is in a spectral region where few other atmospherically significant
species absorb to interfere with the accurate quantification of CO. The few potential
interferences are discussed in Section 2.5.1.
Since NDIR is a spectrophotometric method, the concentration of CO can be determined
based upon the Beer-Lambert Law. The Beer-Lambert law relates the concentration of an
absorbing species to the degree of light attenuation according to the equation shown below:
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 4 of 3 3
///0=e(
where:
/ = light intensity after absorption by absorbing species
I0 = light intensity before absorption by the absorbing species
a = absorption coefficient for absorbing species
x = path length between light source and detector
C = concentration of the absorbing species.
By measuring the degree of light attenuation through a sample cell of known length in both the
presence and absence of CO, the concentration can be accurately determined if the absorption
coefficient of CO is known.
In Luft-type instruments, IR radiation is passed alternately between a reference cell,
containing a non-absorbing gas, and the sample cell. Prior to the introduction of sample gas
containing CO to the sample cell, the intensity of the light passing through the sample cell is
adjusted to match the intensity passing through the reference cell. As sample gas containing CO
is introduced to the sample cell, an imbalance in the light transmitted through the two cells
develops. This imbalance results in a detectable signal that is related to the CO concentration in
the sample cell. However, although CO has a strong characteristic absorption at 4.7 |j,m, other
gaseous atmospheric species can contribute to the light attenuation by the sample and, thus,
interfere with the accurate quantification of CO. Consequently, gas filter correlation (GFC)
techniques were developed to address this problem.
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 5 of 3 3
2.2.2 Gas Filter Correlation (GFC)
A schematic of a GFC CO monitor is shown in Figure 2-1. In GFC monitors, there is
only one sample cell, which acts both as the sample and reference cell. In this technique, a broad
band of IR radiation is emitted from an IR source and enters the sample cell. Mirrors are used to
reflect the light across the length of the cell multiple times which increases the effective
pathlength and sensitivity of the monitor. Depending on instrument design, the light passes
IK Source
Sample Cell
Sample
Gas Filter
Wheel
Permeation
Dryer
n.
Lter
n
J
f '
( 1 N
Span Gas
Bandwidth
Filter
/
Display
Detector
Pimp
Analog RS-232
Output Digital
Output
Figure 2-1. General schematic of a typical GFC CO analyzer.
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 6 of 3 3
through a bandpass filter either before or after the sample cell. This bandpass filter blocks most
wavelengths from passing and allows only a narrow bandwidth of radiation centered on 4.7 |j,m
to reach the detector. In GFC monitors, a mechanical chopper assembly with cells containing
either pure nitrogen (Ni) or pure CO is positioned between the IR source and the sample cell, and
acts as a gas filter. As the gas filter assembly rotates, the IR energy alternately passes through
the cells containing the CO or N2 gases. When the light passes through a cell containing CO, all
the light at 4.7 |j,m is absorbed by the CO in the gas filter cell. This technique effectively "scrubs
out" any light at wavelengths where CO absorbs prior to the sample cell, but allows other
wavelengths of light to pass through the sample cell and reach the detector. As the chopper
wheel spins, the gas filter cell containing N2 in the chopper assembly then crosses the IR energy
beam. Since N2 is transparent to IR radiation, this gas filter cell allows all IR light to pass
through unattenuated by the N2. A neutral attenuator is used in the gas filter to reduce the overall
light intensity exiting the N2 cell to match the intensity exiting the CO cell.
In the absence of CO in the sample cell, no change in light intensity is detected as the gas
filter rotates. However, when CO is introduced to the sample cell, some of the light that passes
through the N2 cell of the wheel is absorbed by CO in the sample cell, resulting in a difference in
light intensity at the detector that is proportional to the concentration of CO in the sample cell.
As the gas filter rotates, the light intensity is modulated and creates a signal from the detector
that is electronically demodulated using phase-sensitive amplifiers and subsequently processed to
generate a CO concentration reading.
2.3 Recommendations for NCore
Since the high sensitivity analyzers deployed at NCore sites are intended to monitor low
ambient CO concentrations, it is important that they meet a variety of performance criteria as
described below. Many of these performance criteria are more stringent than those for routine
CO analyzers; consequently, there are a number of recommended features that the precursor CO
analyzers should have in order to achieve the performance criteria. This section describes the
recommended performance criteria and the analyzer features that are recommended in order to
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 7 of 3 3
achieve the performance criteria, and provides examples of commercial high sensitivity CO
analyzers that are available for deployment at the NCore sites. Additionally, this section
discusses some important sampling requirements that should be considered during the
installation of the analyzers.
2.3.1 Recommended Method Performance Criteria
The U.S. EPA has recently assessed the measurement quality objectives needed for high
sensitivity precursor gas monitoring in NCore, relative to the long-established statistics stated in
40 CFR 58.[6] In particular, EPA recommends that measurement quality objectives for bias and
precision be based on upper confidence limits at the monitoring site level, to provide a higher
probability of reaching appropriate conclusions (e.g., in comparisons to NAAQS). The intent of
this recommendation is to move S/L/T agencies to a performance-based quality system i.e.,
allowing organizations that show tight control of precision and bias to reduce the frequency of
certain QC checks, and to focus their quality system efforts where most needed.
The U.S. EPA recommends that the high sensitivity CO analyzers that are deployed at
NCore sites meet the following method performance criteria.
2.3.1.1 Precision
Precision is defined as the measure of agreement among individual measurements of the
same property taken under the same conditions. Precision is assessed from checks that are
performed at least once every two weeks (see Sections 2.7.4 and 5.3.3). Calculations to assess
precision are given below and should be used to assess precision on a quarterly basis. It is
recommended that high sensitivity CO analyzers have a 95 percent probability limit for precision
of ±15 percent or less.
Calculation of precision starts with the comparison of the known challenge concentration
used in the precision checks to the corresponding measured concentrations reported by the
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 8 of 3 3
analyzer. The resulting percent differences are then used as described below. For each single
point precision check, calculate the percent difference, dt, as follows:
Equation 1
reported - challenge
4 = . „ —-100
challenge
where reported'is the concentration indicated by the high sensitivity CO analyzer and challenge
is the concentration of the standard used in the precision check. The precision estimator is then
calculated as the coefficient of variation (CV) upper bound, using Equation 2 as follows:
Equation 2
where n is the number of data points (i.e., precision check comparisons), the dt values are the
resulting percent differences, and X o.i,n-i is the 10th percentile of a chi-squared distribution with
n-1 degrees of freedom.
2.3.1.2 Bias
Bias is defined as a systematic or persistent distortion of a measurement process that
causes errors in one direction. Bias is assessed from the degree of agreement between a
measured value and the true, expected, or accepted value. Analyzer bias is calculated using
comparisons of known challenge concentrations to the corresponding measured concentrations
reported by the analyzer. The challenge comparisons used to assess bias should be the same as
those used to assess precision (see Section 2.3.1.1 above). The bias estimator is an upper bound
on the mean absolute value of the percent differences as described in Equation 3 as follows:
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 9 of 3 3
Equation 3
\bias\=AB+t0.95,n-l —j=
v»
where n is the number of challenge comparisons being aggregated; to.95,n-i is the 95th quantile of
a t-distribution with n-1 degrees of freedom; the quantity AB is the mean of the absolute values
of the individual dt 's and is calculated using Equation 4 as follows:
Equation 4
and the quantity AS is the standard deviation of the absolute value of the df's and is calculated
using Equation 5 as follows:
Equation 5
AS=
Since the bias statistic as calculated in Equation 3 uses absolute values, it does not have a
direction or sign (negative or positive) associated with it. The sign of the calculated bias is to be
determined by rank ordering the percent differences of the QC check samples from a given
analyzer for a particular assessment interval. Calculate the 25th and 75th percentiles of the
percent differences for each analyzer. The absolute bias upper bound should be flagged as
positive if both the 25th and 75th percentiles are positive, and as negative if both these percentiles
are negative. The absolute bias upper bound would not be flagged if the 25th and 75th percentiles
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 10 of 33
are of different signs. It is recommended that high sensitivity CO analyzers have an upper bound
for the average bias of ± 15 percent or less.
2.3.1.3 Representativeness
Representativeness refers to whether the data collected accurately reflect the conditions
being measured. It is the data quality indicator most difficult to quantify. Unless the samples are
truly representative, the other indicators are meaningless. Representativeness for monitoring of
low ambient levels of CO in NCore is different than for routine monitoring, since the objectives
of the monitoring are much different. Representativeness can only be assured in terms of the
appropriate selection of the sampling site, proper implementation of ambient air sampling, and
reasonable coverage of the sampling schedule (i.e., 24 hours per day, 7 days per week, ideally).
2.3.1.4 Completeness
Completeness is defined as the amount of data collected relative to the total expected
amount. Ideally, 100 percent of the expected amount of data would always be collected; in
practice, completeness will be less for many reasons, ranging from calibration time and site
relocation to power outages and equipment failure. For monitoring of ambient CO
concentrations in NCore, EPA requires a minimum data completeness of 75 percent. In practice
typical completeness values can often approach 90 to 95 percent.
2.3.1.5 Comparability
Comparability is defined as the process of collecting data under conditions that are
consistent with those used for other data sets of the same pollutant. The goal is to ensure that
instruments purchased and operated by different states and local agencies produce comparable
data. To promote comparability, this TAD describes the recommended characteristics of high
sensitivity CO analyzers and the procedures for their installation and use. For example, all
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 11 of 33
monitoring agencies should purchase instruments that have the additional features described in
Section 2.3.2, and should adhere to the sampling requirements described in Section 2.3.3.
2.3.1.6 Method Detection Limit
The method detection limit (MDL) refers to the lowest concentration of a substance that
can be reliably determined by a given procedure. The MDL is typically not provided by the
vendor. Based on the objectives of the Precursor Gas Program, it is expected that most sites will
be measuring pollutant concentrations at lower ranges than the typical SLAMS/NAMS network.
Therefore, the ability to quantify concentrations at these lower levels will be very important.
The use of a vendors advertised LDL is sufficient to make intelligent purchasing decisions.
Vendors quantify LDLs under ideal conditions and therefore one might consider this value as the
best possible detection that can be achieved. As these monitors are deployed into monitoring
networks, where both environmental conditions, equipment (calibration, dilution devices,
sampling lines, gaseous standards) and operator activities can vary, it is important to estimate
what pollutant concentrations can truly be detected, above background noise (the potential
conditions mentioned above). The site specific MDL establishes an estimate based on the
routine operation (and conditions) of that instrument in the network and provides a more
meaningful evaluation of data as it is aggregated across the precursor gas network. By
establishing site specific MDLs, values less than the MDL can be flagged which would allow
data users a more informed decision on the use of that data.
The MDL should be established on-site by supplying the analyzer at least seven times
with a test atmosphere containing CO at a concentration that is approximately one to five times
greater than the estimated MDL, and recording the response. To perform the MDL test, run zero
air through the analyzer and establish an acceptable zero; dilute pollutant gas to the targeted
concentration (one to five times the estimated MDL) and collect 20 to 25 one minute
observations. Repeat this seven times over the course of 5 to 14 days. Average the concentration
from the 20-25 readings; calculate the standard deviation (S) of the average readings and
compute the MDL. The MDL is then calculated as the standard deviation of the response values
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 12 of 33
times the Student's t-value for the number of test measurements (40 CFR Part 136, Appendix B).
The MDL for high sensitivity CO analyzers should be established prior to putting the analyzers
into service, and should be 0.080 ppm (80 ppb) or lower over an averaging time of no more than
5 minutes.
2.3.1.7 Lower Detectable Limit
The LDL is the minimum pollutant concentration that produces a signal of twice the
noise level. To estimate the LDL, zero air is sampled and the noise level of the CO readings is
determined according to 40 CFR 53.23(b). The vendor-specified LDL for the most sensitive
range of high sensitivity CO analyzers should be 0.040 ppm (40 ppb) or lower, over an averaging
time of no more than 5 minutes.
2.3.1.8 Linear Range
The linear range of each high sensitivity CO analyzer should extend from approximately
0.040 ppm to at least 5 ppm. Users should determine if their range should exceed 5 ppm and
adjust accordingly. A range of 5 ppm may not be sufficient in all areas and situations. Note that
some high sensitivity CO analyzers can operate simultaneously on a number of ranges, with each
range recorded on a separate data logger channel with its own calibration curve. Although
requiring slightly more effort to calibrate and maintain, recording of multiple ranges would allow
capture of a wide range of CO concentrations.
2.3.1.9 Zero/Span Drift
Zero drift is defined as the change in response to zero pollutant concentration, over
12- and 24-hour periods of continuous unadjusted operation. Span drift is defined as the percent
change in response to an upscale pollutant concentration over a 24-hour period of continuous
unadjusted operation. Zero and span drift specifications should be obtained from the vendor
prior to putting a high sensitivity CO analyzer into service. Such CO analyzers should have 12-
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 13 of 33
and 24-hour zero drift less than 100 ppb, and should have a span drift of less than ±1 percent of
the full scale measurement range of the analyzer per 24 hours. Zero tests should be performed
with the internal zero engaged. It is suggested that the zero trap of the analyzer be initially and
periodically (annually) evaluated for efficiency or if the operator suspects a problem with the
zero trap. A suggested means of confirming the functionality of the zero trap is to sample
calibration air spiked with 1000 to 2000 ppb CO during the zero cycle, and review results for the
automatic zeroing periods. This approach tests the key components of the zeroing/drying system
and should meet the vendor-specified zero drift criterion.
2.3.2 Recommended Features for High Sensitivity CO Measurements
Continuous high sensitivity CO analyzers are commercially available from a number of
vendors. The design of these analyzers is similar among vendors with some slight variations. A
diagram of a typical GFC CO instrument is described in Section 2.2, and examples of specific
instruments are provided in Section 2.3.3. In general, each of the analyzers contains the
following systems:
• Pneumatic System: This portion of the analyzer consists of a sample probe, sample
inlet line, particulate filter, dryer, catalytic converter, flow meter, and pump, all used
to condition the ambient sample air and bring it to the analyzer.
• Analytical System: This portion of the analyzer consists of the IR source, the gas
correlation filter, motor, optical multipass cell, detector, and bandpass filter. Being a
mechanical device, the motor can and will wear out. The gas correlation filter can be
subject to leakage and the IR source will eventually burn out. Extra IR sources
should be stocked as replacement parts and the gas correlation filters should be
replaced as necessary.
• Electronic Hardware: This portion of the analyzer consists of the electronic
components that control the analyzer and process the signals. This part of the
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 14 of 33
analyzer generally requires little or no maintenance. However, if the instrument is
operated near the manufacturer's recommended upper temperature limit, individual
integrated chips can fail and cause problems with data storage or retrieval.
In addition to these general systems, the high sensitivity versions of the commercial GFC
CO instruments typically have four distinct features that allow them to measure CO at ppb
levels:
• The sample stream is dried using a permeation tube or Nafion® Dryer prior to
introduction to the sample cell;
• The analyzer baseline is determined and corrected automatically and frequently by
introducing into the sample cell CO-free air, that is generated using a heated, on-
board, converter that is based on palladium (Pd), platinum (Pt), or other composition;
• The temperature of the optical bench is tightly controlled (i.e., within ± 1 °C) to
maintain detector stability;
• The instrument uses an ultra-sensitive detector, in order to detect very small changes
in light intensity.
It is recommended that the high sensitivity CO analyzers deployed in NCore employ these
features. Examples of commercial analyzers with these features are presented below.
2.3.3 Commercial High Sensitivity GFC CO Monitors
Several vendors of commercial GFC CO analyzers supply instruments for both ambient
and high sensitivity monitoring. Only high sensitivity GFC CO analyzers with nominal LDLs of
40 ppb or below are discussed in detail in this document. Three such analyzers [from Ecotech,
Thermo Electron Corporation, and Teledyne Advanced Pollution Instrumentation (API)] are
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 15 of 33
described here as examples of available high sensitivity CO monitors (no endorsement should be
inferred). The descriptions provided for these instruments and the performance claimed for them
are based solely on information provided by the respective vendors, and confirmation by EPA
should not be inferred.
2.3.3.1 Ecotech Model EC9830T
The Model EC9830T (Figure 2-2) is the high-sensitivity version of the Model EC9830,
which has a U.S. EPA Reference Method designation of RFCA-0992-088.[7]
Figure 2-2. Ecotech EC9830T high sensitivity CO analyzer (courtesy of Ecotech).
This analyzer has a vendor-specified LDL of 20 ppb, which is achieved in part by the
implementation of the features described in Section 2.3.2 and through the use of a cell with a 6 m
path length. This LDL is achieved by use of a Kalman digital filter to provide a compromise
between response time and noise reduction. The 95% response time of the EC9830T is 300
seconds with the Kalman filter. The recommended operating temperatures for the
Model EC9830T are 20 °C to 30 °C, but it may be operated between 15 °C and 35 °C. Since
temperature stability of the analyzer is crucial to maintaining its high sensitivity, an automatic
background correction will be initiated if the internal temperature of the instrument changes by
more than 4 °C. The automatic zeroing feature allows the analyzer to periodically check and
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 16 of 33
correct for background light intensity, and significantly reduces zero drift. The Model EC9830T
incorporates a second order polynomial correction curve that is custom calibrated for each
analyzer in the range of 0 to 3,000 ppb (3 ppm), with linearity within 5 percent. Outputs are
provided in both analog and digital formats. Table 2-1 shows the specifications of the Model
EC9820T.[7]
Table 2-1. Ecotech Model EC9830T high sensitivity CO analyzer specifications
Parameter
Precision
Bias
Method Detection Limit
Lower Detectable Limit
Linearity
Zero Drift
Span Drift
EC9830T Specifications
± 5 % of reading for range 0-1 ppm
± 1% of reading for range 1-20 ppm
Not Available
Not Available
0.020 ppm (20 ppb) with Kalman 300 sec filter active
± 5% 0-1 ppm scale
± 1% of full scale from 1-20 ppm
Temperature dependence, 0.1% per degree C change.
24 hours; less than 0.020 ppm (20 ppb)
30 days; less than 0.020 ppm (20 ppb)
Temperature dependence, 0.05 % per degree C change.
24 hours less than 0.5% of reading
30 days less than 1 .0% of reading
2.3.3.2 Thermo Electron Corporation Model 48C-TLE
The Thermo Electron Corporation Model 48C-TLE Enhanced Trace Level CO
analyzer'8"101 (Figure 2-3) is an improved version of the standard Model 48C Ambient CO
analyzer (U.S. EPA Designation Method RFCA-0981-054). In addition to the features described
in Section 2.3.2, the primary modifications to the Model 48C-TLE analyzer that improve its
sensitivity over the Model 48C include the use of higher reflectance gold-coated mirrors,
incorporation of a baseline auto-zeroing function, and the implementation of ± 1 ° C control of
optical bench temperature. The recommended operating temperature for the instrument ranges
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 17 of 33
from 20 °C to 30 °C, but the Model 48C-TLE CO analyzer can be operated over the range of
5 °C to 45°C. The Model 48C-TLE CO analyzer has an LDL of 0.02 ppm (20 ppb) with a 30
second averaging time. The analyzer has ten operating ranges from 0 to 1 ppm through 0 to
1,000 ppm, including a 0 to 5 ppm range. Data can be provided in analog or digital formats.
Table 2-2 shows the specifications of the Model 48C-TLE CO.
Figure 2-3. Thermo Electron Model 48C-TLE CO analyzer (courtesy of
Thermo Electron).
Table 2-2. Thermo Electron Model 48C-TLE CO analyzer specifications
Performance Parameters
Precision
Bias
Method Detection Limit
Lower Detectable Limit
Linearity
Zero Drift
Span Drift
Model 48C-TLE CO Analyzer Specifications
±2% of reading or 0.02 ppm (20 ppb) (whichever is larger)
Not Available
Not Available
0.04 ppm (40 ppb) ; 60 sec averaging time
± 1% full-scale
< 0.100 ppm (100 ppb) (24 hour)
± 1% full-scale (24 hour)
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 18 of 33
2.3.3.3 Teledyne Advanced Pollution Instrumentation (API) Model 300E
The Teledyne/API Model 300E high sensitivity CO analyzer'11'121 (Figure 2-4) has a
U.S. EPA Reference Method designation of RFC A-1093-093. The Model 300E is a
high-sensitivity version of the Model 300, which has the same Reference Method designation,
and incorporates the recommended features described in Section 2.3.2 to achieve enhanced
sensitivity. The Model 300E CO instrument has selectable measurement ranges that can be set
anywhere from 0 to 1 ppm up to 0 to 1,000 ppm. The operational temperature range for the
instrument is between 5 °C and 40 °C. Table 2-3 shows the specifications of the Model 300E.
Figure 2-4. Teledyne API Model 300E CO analyzer (courtesy of Teledyne API).
Table 2-3. Teledyne API Model 300E high sensitivity CO analyzer specifications
Performance Parameters
Precision
Bias
Method Detection Limit
Lower Detectable Limit
Linearity
Zero Drift
Span Drift
Model 300E CO Analyzer Specifications
0.5% of reading
Not Available
Not Available
0.04 ppm (40 ppb); 30 second averaging time
1% full-scale
<0.1 ppm (100 ppb) per 24 hours; 0.2 ppm (200
ppb) per 7 days
<0.5% reading per 24 hours, 1% reading per
7 days
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 19 of 33
2.3.4 Sampling Requirements
Proper siting of the sampling equipment and sampling probes is necessary to ensure that
the precursor gas analyzers are obtaining representative samples of the ambient air. Likewise,
proper environmental control and proper sampling is critical to ensuring that the analyzers are
operating correctly and that the CO measurements are comparable to CO measurements recorded
at other sites.
2.3.4.1 Analyzer Siting
Analyzer siting should follow the criteria in 40 CFR 58, Appendix E. The installation of
the CO analyzer should allow for the sample manifold inlet to be located between 3 and
15 meters above ground level, with at least one meter of vertical and horizontal separation from
supporting structures. The probe should be positioned with at least 270 degrees of unrestricted
airflow including the predominant wind direction. The probe should be separated from the drip
line of nearby trees or structures by at least 20 meters, and should be positioned at least twice as
far from nearby obstacles as the height of the obstacles.
2.3.4.2 Instrument Shelter
To help ensure proper performance, the precursor gas analyzers and supporting
equipment should be installed and operated in a temperature controlled environment. An
insulated instrument shelter should be used to protect the analyzers from precipitation and
adverse weather conditions, maintain operating temperature within the analyzers' temperature
range requirements, and provide security and electrical power. The environmental control of the
shelter should be sufficient to minimize fluctuations in shelter temperature. The recommended
shelter temperature range is 20 °C to 30 °C, and daily changes in temperature should not exceed
5 °C over a 24-hour period. Condensation of moisture must be prevented, and it may be
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 20 of 3 3
necessary to impose seasonal temperature ranges to assure remaining above the ambient
dewpoint.
2.3.4.3 Sample Inlet or Manifold
Sample air for a high sensitivity CO analyzer should be drawn directly to the analyzer
through FEP (Fluoroethylpropylene) or PTFE (Polytetrafluoroethylene) tubing (e.g., 1A inch
outer diameter (OD)), equipped with a 1-micron Teflon® filter at the inlet to remove particles
from the sample air. The filter should be changed weekly, or more often in excessively dirty
conditions. In this configuration the CO analyzer does not share its inlet system with any other
analyzer.
Some existing air monitoring stations may provide air to multiple analyzers through a
common manifold. In such a case, it is recommended that manifolds for high sensitivity CO
measurements be made of glass. Since neither zero air nor sample air is totally particulate-free,
over time sample manifolds will collect particulate matter on the internal walls. A glass
manifold is transparent, and can be inspected easily and cleaned readily by rinsing with distilled
water and air drying. However, caution must be used with glass manifolds because of their
fragility. The CO analyzer should be located as close to the inlet manifold as possible, to
minimize the length of sampling lines, and sampling lines should be of constructed of FEP or
PTFE . The sample manifold must be of sufficient diameter that outside air is drawn into the
manifold at as close to atmospheric pressure as possible. The manifold must also allow for
excess gas flow to be exhausted without over-pressurization (e.g., during delivery of gases from
high-pressure cylinders.) If the pressure in the manifold differs from atmospheric pressure, the
CO readings obtained will not be representative. A manifold of 1 inch inner diameter should be
sufficient to avoid pressure differences. A detailed description of sample manifold designs is
presented in Appendix A.
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 21 of 33
2.4 Potential Problems and Solutions
This section describes several of the potential problems associated with high sensitivity
CO measurements, and discusses the practical solutions to these problems, some of which the
vendors have already implemented in their analyzers. In addition to these potential problems,
other problems may arise in the routine operation of high sensitivity CO analyzers. Operators
are encouraged to report any observations or anecdotal data that might add to the understanding
or awareness of interferences or other anomalies in CO measurements.
2.4.1 Interferences and Sources of Bias
Preventing interferences or biases is crucial to the accurate measurement of low ambient
levels of CO. The following sections describe several potential positive and negative sources of
interference or bias. Section 2.4.1.1 describes the most common positive interferences (water
vapor and CO2), and Section 2.4.1.2 describes the most common negative interferences and
sources of bias (incomplete removal of CO during instrumental auto-zeroing and loss of CO by
reaction or adsorption in dirty inlet lines or filters). In each section recommended procedures to
minimize these interferences or sources of bias are described.
2.4.1.1 Positive Interferences
GFC CO analyzers determine CO concentration by measuring the amount of light that is
absorbed at a select wavelength (4.7 |j,m) as it passes through a sample cell containing CO. Any
other gas in the air sample that also absorbs at those wavelengths could present an interference
that results in an inaccurate determination of CO concentration. Removal of potential
interferences must be done selectively such that these interferences are completely removed
without affecting the CO concentration. For CO measurements at low ambient levels, this is
particularly critical in order to achieve the desired sensitivity. Of particular concern are water
vapor and CO2, which are both generally present in the atmosphere at concentrations greatly
exceeding those of CO. Studies have shown conclusively that water vapor interferes with the
ability of NDIR analyzers to accurately quantify CO. Water absorbs very strongly in the 3.1, 5.0
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 22 of 3 3
to 5.5, and 7.1 to 10.0 |j,m regions of the IR spectrum. Since water vapor absorbs light in regions
in close proximity to the CO absorptions, removal of water vapor from the sample air is
necessary to avoid positive interferences in the determination of CO concentration. To achieve
this goal, high sensitivity CO analyzers are equipped with a permeation tube or Nafion™ drier
that selectively removes water vapor from the sample gas without removing CO.
CO2 absorbs in the IR spectrum at 2.7, 5.2, and 8.0 to 12.0 |j,m. These regions are also
very close to the regions of CO absorption. Since atmospheric carbon dioxide is typically much
higher in concentration than CO and has properties that are similar to CO, it is impractical to
selectively remove CO2 from the sample air without removing a fraction of the CO. Therefore, it
is important that the bandpass filter used to limit the interrogating radiation is sufficiently
selective to restrict the wavelengths to a small region centered on the CO absorption band of
4.7 |j,m. Manufacturers of high sensitivity CO analyzers select the bandpass filter to effectively
remove CO2 interference. An added benefit of such filters is that they also limit interference
from water vapor.
2.4.1.2 Negative Interferences and Biases
High sensitivity CO analyzers are equipped with a solenoid switching system to draw
sample air into a heated internal scrubber that converts all CO to CO2. The analyzer then
measures the light absorption of this CO-free air and uses that light intensity to establish the zero
reading. However, any CO that is not converted to CO2 would remain in the sample gas and
decrease the light intensity (i.e., absorb the light) used to establish the zero reading, resulting in
an artificially high zero reading and a negative bias when measuring the CO in ambient air. To
avoid this situation, it is important that the heated scrubber be maintained at the manufacturer's
recommended temperature. Scrubber efficiency must be checked periodically, e.g., every 30
days. A convenient means to check CO scrubber efficiency is to sample ambient air, then zero
air, then a CO calibration mixture, all with the internal heated CO scrubber engaged. Zero air
and sample air readings should be within ±0.010 ppm (10 ppb), and scrubber efficiency should
be >99%.
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 23 of 3 3
2.4.2 Detector Stability
The temperature of the detector in a high sensitivity CO analyzer must remain stable in
order to allow for ppb sensitivity. Commercial high sensitivity CO analyzers provide a display
of the detector temperature. This temperature should be checked periodically for compliance
with the vendor's required temperature setting. Bench temperature should be checked both with
and without the zero scrubber engaged, to ensure that scrubber effluent does not cause heating of
the optical bench.
2.5 Supporting Equipment
In addition to the precursor gas analyzers, several pieces of supporting equipment should
be maintained at each NCore site. At a minimum, this equipment includes a data acquisition
device and calibration equipment.
2.5.1 Data Acquisition Device
Many types of equipment can be used to record the concentration measurements obtained
from the analyzer. Recommended options for data acquisition are described in Section 6 of this
TAD.
2.5.2 Calibration Equipment
The following equipment is recommended for calibration of a high sensitivity CO
analyzer.
2.5.2.1 Calibration Standard and Standard Delivery System
The calibration standards used for the calibration of high sensitivity CO analyzers should
be generated by dilution of a commercially-prepared and certified compressed gas CO standard
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 24 of 3 3
using a mass flow controlled (MFC) calibration unit. The method of generating calibration
standards by dilution involves the accurate measurement of the analyte gas flow rate and the
diluent gas flow rate prior to blending these gas streams. The calibration unit includes mass flow
controllers that are based upon small thermistors that are sensitive to heat loss. A potential
voltage is applied to the thermistor and, as the gas flow increases across the thermistor, the
resistance of the thermistor changes proportionally with the flow rate. This change in resistance
can be measured very accurately by electronic circuitry and a feedback loop within the MFC
circuitry monitors the gas flow and controls the flow rate to maintain the desired rate. Using two
channels in parallel, the MFC calibrator unit controls the analyte gas flow rate and the diluent
gas flow rate such that upon mixing these gases generates a working standard with the desired
concentration. Typical flow ranges of the MFC units are up to 10 L/min for the diluent gas flow
and up to 100 cm3/min for the analyte gas flow. These systems allow for accurate dilution of CO
standard gases from high concentration (usually 200 to 250 ppm) to low ambient working
standard concentrations (e.g., from 0.04 to 0.8 ppm). When the analyte concentration in the
commercially-prepared standard cylinder is certified by reference to NIST standards, and the
MFCs are calibrated to NIST-traceable standards, the resulting working standard concentration is
considered to be NIST-traceable.
It is highly important when purchasing a MFC calibrator that it meet the 40 CFR 50
requirements of ±2 percent flow accuracy, and that the calibrations of both MFC channels be
checked periodically using a NIST traceable flow standard. Routine MFC checks must be
standard procedure. To accomplish such check, a NIST traceable flow standard must be on hand
as part of every calibration system.
2.5.2.2 Zero Air Source/Generator
Zero air is required for the calibration of high sensitivity CO analyzers. This air must
contain no detectable CO (i.e., CO content must be less than the LDL of the CO analyzer) and
must be free of paniculate matter. Suitable zero air may be supplied from compressed gas
cylinders of purified air, with additional external CO scrubbers (e.g. hopcalite or carulite) to
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 25 of 3 3
remove residual CO in the commercial product. However, it is likely too expensive and
impractical to maintain a sufficient supply of zero air cylinders to operate a high sensitivity CO
analyzer continuously. As an alternative, many commercially available zero-air generation
systems can greatly reduce CO levels in air. However, depending on the required zero air flow
rates, it may be difficult to reduce CO levels to 0.040 ppm or less, unless a Pd or Pt scrubber is
used. A recommended approach to test zero air quality is to compare the readings of the high
sensitivity CO analyzer in zero air in sample mode vs. the analyzer output in the "auto zero"
mode. This comparison should be done at least quarterly and can only be done with those
analyzers that provide a digital recording of the output in the "auto zero" mode.
2.6 Reagents and Standards
Routine operation of precursor CO analyzers requires the use of calibration standards and
zero air to conduct periodic calibrations and instrument checks. This section describes the
requirements for these gases.
2.6.1 Calibration Standards
The primary CO standards used must be certified, commercially-prepared compressed gas
standards with a certified accuracy of no worse than ±2 percent. Standards in the concentration
range of 200 to 250 ppm are suitable choices for dilution to prepare low concentration calibration
mixtures. The commercially-prepared CO standard may contain only CO in an inert gas
(e.g., N2), or may be a mixed component standard that also contains known concentrations of
other precursor gases (e.g.,SO2, NO). (Note that mixtures containing both SO2 and NO may not
be suitable for SO2 calibration, depending on the NO rejection ratio of the SO2 analyzer - see
Section 3 SO2 Measurements.)
Every gas standard used in precursor gas monitoring must be accompanied by a certificate
of calibration from the vendor stating the concentration of the standard, the uncertainty of that
certification, and the expiration date of the certification. Standards traceable to NIST are
preferred. Certification documents for all standards must be retained in a common location and
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 26 of 3 3
reviewed periodically so that standards for which the vendor's certification has expired may be
removed from service and replaced.
2.6.2 Zero Air
Zero air used as dilution gas for calibration purposes should have a CO concentration
below the LDL of the high sensitivity CO monitor. Commercial cylinder gas grades such as
Ultra Zero and CEM grade may be suitable as a starting point, provided additional cleanup is
employed as noted in Section 2.5.2.2. Commercial zero air further scrubbed of CO may be used
to crosscheck the purity of air provided by a commercial continuous air purification system.
2.7 Quality Control
A thorough quality control program is critical to the collection of high sensitivity CO
monitoring data, and must be implemented at each NCore site. Components of such a program
are described below.
2.7.1 Site Visit Checks and Remote Diagnostic Checks
To determine whether the CO analyzer is working properly, field operators should
conduct routine checks of instrument diagnostics and performance every time they visit the
monitoring station. Each agency needs to develop diagnostic or maintenance checklists or
electronic spreadsheets to document that all required checks have been made. Such lists and
sheets should be useful both for collecting diagnostic information and for assessing the quality of
the monitoring data. To the extent possible, diagnostic checks can be done remotely, provided
the data acquisition system allows remote access to instrument diagnostic information (see
Section 6).
2.7.2 Multipoint Calibrations
A multipoint calibration includes a minimum of four points (three spaced over the
expected range and a zero point), generated by the calibration system. Although more points
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 27 of 3 3
may be preferable, current high sensitivity CO analyzers typically provide linear response over
the entire operating range; therefore, four points should be sufficient. Multipoint calibrations
must be done prior to the high sensitivity CO analyzer being put into service and should be
repeated at least quarterly thereafter. An analyzer should be calibrated (or recalibrated) if any of
the following conditions occur:
• Upon initial installation;
• The Level 1 span check or precision check difference exceeds 15 percent;
• After repairs or service is conducted that may affect the calibration;
• Following physical relocation or an interruption in operation of more than a few days;
• Upon any indication that the analyzer has malfunctioned or a there has been a change
in calibration; or
• The measured concentration values during challenges with performance test samples
(Section 5.4.1) differ from the certified standard values by ±15 percent.
The analyzers should be calibrated in-situ without disturbing the normal sampling inlet system to
the degree possible.
2.7.3 Level 1 Zero/Span Checks
Level 1 zero and span calibrations are simplified, two-point calibrations used when
adjustments may be made to the analyzer. When no adjustments are made to the analyzer, the
Level 1 calibration may also be called a zero/span "check" and must not be confused with a level
2 zero and span check. Level 1 zero and span checks should be conducted nightly. They are used
to assess if the analyzers are operating properly and to assess if any drift in instrument response
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 28 of 3 3
has occurred. The level 1 check should not exceed ±15 percent. Zero drift is internally adjusted
by the analyzer. The zero check is used to verify that the internal zero is working properly.
They are conducted by challenging the analyzer with zero air and a test atmosphere containing
CO at a concentration of between 70 percent and 90 percent of the full measurement range in
which the analyzer is operating. The challenge gas should be sampled through as much of the
sampling inlet system as practical to mimic the actual sampling of ambient air. The results of the
Level 1 zero/span check should be plotted on control charts to graphically illustrate the trends in
the response of the analyzer to the challenge gases. If the measured concentrations fall outside
of the control limits, the accuracy of the MFC calibration system should be checked with a
NIST-traceable flow standard. If the MFC flow accuracy is confirmed, the data recorded since
the last successful Level 1 check should be flagged and the analyzer should be recalibrated using
the multipoint calibration procedures described in Section 2.7.2.
State-of-the-art calibration equipment now exists that is fully automated. These "new
generation" calibration units are fully integrated with computers, mass flow calibrators, and the
associated hardware and software where they can create test atmospheres manually or
automatically. For the precursor gas program, it is recommended that the NCore sites have fully
automated calibration capability. Below are a number of reasons why this is advantageous:
• By performing the calibrations or checks automatically, agencies no longer spend the
manpower needed to perform them.
• Automated calibrations or checks can be triggered internally or by a DAS. Since newer
DASs allow remote access, this allows a remote user to challenge the analyzers without
actually being present.
• High sensitivity precursor gas analyzers are expected to have more zero and span drift than
less sensitive analyzers; therefore, it is important that a zero and Level I check be performed
daily.
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 29 of 3 3
• New generation DASs can record calibration and check data and allow remote users to track
daily Level I check and zero drift. This is important for data validation, verification and
troubleshooting.
2.7.4 Precision Checks
At least once every two weeks a precision check should be conducted by challenging the
CO analyzer with a known low CO concentration to assess the performance of the analyzer. The
precision checks should be conducted by challenging the precursor CO analyzer with a standard
gas of known concentration between 0.25 and 0.50 ppm (250 and 500 ppb). After completion of
the precision check, the operator should calculate the percent difference between the measured
value and the known standard value. Precision should be calculated quarterly, using the
calculated percent differences from the precision checks, according to the equations provided in
Section 2.3.1.1. For acceptable precision to be maintained it is recommended that the calibration
system's gas flows be verified frequently against a NIST flow standard, and adjusted if necessary
before making any adjustments to the analyzer.
2.8 Preventive Maintenance and Troubleshooting
Long-term operation of continuous high sensitivity precursor gas analyzers requires a
preventive maintenance program to avoid instrument down-time and data loss. Despite active
preventive maintenance, occasional problems may arise with the high sensitivity CO analyzers.
This section briefly describes several key items that might be included in the preventive
maintenance program established for high sensitivity CO analyzers deployed atNCore sites, as
well as some of the troubleshooting activities that may be useful in resolving unexpected
problems with these analyzers. This discussion is not meant to be exhaustive or comprehensive
in detail. More thorough discussions can be found in the analyzer operation manuals, and should
be included in SOPs developed for these analyzers. Example SOP's prepared by EPA are
included as Appendix B of this TAD.
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 30 of 33
2.8.1 Preventive Maintenance
Routine preventive maintenance procedures should be in place to prevent down-time and
data loss. Management and field operators should jointly develop their preventive maintenance
program. A program designed by persons unfamiliar with analyzer operations may include
unnecessary items or omit mandatory ones. Several factors linked to shelter and sampling
manifold design can contribute to data loss. CO values can be low if the sample probe, manifold,
and lines are dirty, cracked, or leaky. The sample probe and manifold should be cleaned at least
every six months. FEP and PTFE sampling lines should be replaced every two years. Teflon®
filters used in the sampling train to remove fine particles should be replaced at least once per
month, but may need to be replaced as often as every week, depending on the condition of the
filter and the particulate loading around the monitoring site.
Table 2-4 illustrates items that monitoring agencies should include in their preventive
maintenance program for precursor CO monitoring.
In addition to a schedule, the preventive maintenance plan should also include more
detailed task descriptions, such as illustrated below:
• Because the analyzer pneumatic system requires so much preventive maintenance, the
tubing, solenoids, and pump should be inspected regularly. Cracked tubing or loose
fittings can cause the instrument to analyze room air rather than ambient air and lead
to invalid data. A faulty pump can also cause problems with pneumatic systems.
When oscillations in the flow rate force the operator to adjust the flow continually,
the pump is failing and should be either repaired or replaced.
• Check the instrument for vibration. When pumps get old, they sometimes will vibrate
more than is normal. If this occurs, it can cause cracks if the tubing is touching
another surface.
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 31 of 33
Consult the analyzer operations manual for complete details on operation and
maintenance.
Table 2-4. Example of a preventive maintenance schedule for high sensitivity CO
analyzers.
Item
Replace particle filter
Clean fan/fan filter
Inspect internal, external tubing; replace if necessary
Rebuild or replace pump
Replace IR source
Clean optic bench
Replace wheel motor
Replace gases in correlation wheel
Schedule
Weekly
Semi-annually
Annually
Every two years, or as needed
As needed based on
manufacturer's diagnostics
As needed
As needed
As needed
2.8.2 Troubleshooting
High sensitivity CO analyzers are subject to many factors that can cause inaccurate
measurements or down-time. Table 2-5 summarizes common problems seen with high
sensitivity CO analyzers, their possible causes, and possible solutions. More specific
information can be found in the manufacturer's operations manual.
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 32 of 33
Table 2-5. Instrument troubleshooting for high sensitivity CO analyzers.
Problem
Noisy output
High positive zero drift
No Response to Span Gas
Differential Signal at Zero
Zero output at ambient levels
No flow through analyzer
Reference signal at zero
Possible Cause
Defective DC power supply
Dirty optics
Defective bandpass filter
IR source is defective
IR power supply defective
IR source is defective
IR power supply is defective
CO leak from correlation wheel
Pump failure
IR source failure
IR power supply defective
Pump failure
N2 leak from correlation wheel
Possible Solution
Replace power supply
Clean optics bench
Replace filter
Replace IR source
Replace IR power supply
Replace IR source
Replace IR power supply
Replace wheel
Check pump
Replace IR source
Replace power supply
Replace/ rebuild pump head
Replace wheel
When troubleshooting, an operator must constantly be aware of environmental factors that may
affect the instruments. Environmental factors can also cause sporadic problems that can be
difficult to diagnose. Examples of factors that may affect the performance of the high sensitivity
CO analyzers are:
• Variable shelter temperature (fluctuations greater than several degrees);
• Excessive vibration from other equipment;
• Voltage instability; fluctuations in the 110 VAC line voltage;
Air conditioning system blowing on the instrument;
• Frequent opening of the door of the shelter.
2.9 References
"National Ambient Air Monitoring Strategy," U.S. Environmental Protection Agency,
Office of Air Quality Planning and Standards, Research Triangle Park, NC 27711,
April 2004, Final Draft.
-------
Precursor Gas TAD
Section 2 CO Measurements
Version 4
September 2005
Page 33 of 33
2. Parrish, D.D., Holloway, J.S., Fehsenfeld, F.C, "Routine Continuous Measurement of
Carbon Monoxide with Parts per Billion Precision," Environmental Science and
Technology, 28:9, 1615-1618, 1994.
3. Lobert, J.M., Harris, J.M., "Precursors and Air Mass Origin at Kaashidhoo, Indian
Ocean" Journal of Geophysical Research, Volume 107(D19), 8013, doi:
10.1029/2001JD000731, 2002.
4. Cogan, M., Lobert. J.M., "Instrumentation for Trace-Level Measurements of Carbon
Monoxide in Pristine Environments," Instrument Society of America, Proceedings of the
43rd Annual ISA Analysis Division Symposium, Volume 31, 1998.
5. "Air Quality Criteria for Carbon Monoxide," EPA/600/8-90/045F, U.S. Environmental
Protection Agency, Office of Research and Development, Washington, DC, 20460,
December 1991.
6. Rhodes, R.C., Guidelines on the Meaning and Use of Precision and Accuracy Data
Required by 40CFR Part 58, Appendices A and B, EPA60014-83-023, U.S.
Environmental Protection Agency, Research Triangle Park, NC, June 1983.
7. Ecotech Pty Ltd. (2004). "EC9830T Trace Carbon Monoxide Analyzer Product
Specification Sheet." Accessed September, 2004. Available at
http://www.ecotech.com.au/brochures_new/EC9830T.pdf.
8. Thermo Electron Corporation (2004a). "Instruction Manual, Model 48C Precursor Gas
Filter Correlation CO Analyzer." Accessed September, 2004. Available at
https://www.thermo.com/eThermo/CMA/PDFs/Various/156File 17817.pdf.
9. Thermo Electron Corporation (2004b). "Model 48C Gas Filter Correlation CO Analyzer
Data Sheet." Accessed September, 2004. Available at http://www.thermo.com/eThermo/
CM A/PDF s/Product/productPDF 12767.pdf
10. Thermo Electron Corporation (2004c). Personal Communication with Michael
Nemergut.
11. Teledyne Advanced Pollution Instrumentation (2004a). "Model 300E Gas Filter
Correlation CO Analyzer Data Sheet." Accessed September, 2004. Available at
http ://www.teledyne-api .com/products/model_3 OOe. asp.
12. Teledyne Advanced Pollution Instrumentation (2004b). "Instruction Manual, Model
300E Gas Filter Correlation CO Analyzer." Accessed September, 2004. Available at
http://www.teledyne-api.com/manuals/04288A5.M300E.Manual.pdf.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page Iof30
3.0 HIGH SENSITIVITY SULFUR DIOXIDE MEASUREMENTS
3.1 Introduction
Sulfur dioxide (862) has been identified as a key precursor of fine particulate matter
(PM2.5) and, thus, plays an important role in PM-related health effects. A variety of sulfur
compounds are released to the atmosphere from both natural and anthropogenic sources.
Ninety-five percent of sulfur emitted to the air from anthropogenic sources is released as sulfur
dioxide, mainly from stationary sources that burn fossil fuels (e.g., coal and oil) that contain
sulfur. The major human sources of SC>2 are coal-burning electrical utilities, refineries, and ore
smelters. The amount of SC>2 released depends on the sulfur content of the fuel; high-sulfur coal
may contain as much as six percent sulfur by weight.
3.1.1 Properties of SOg
SC>2 is a colorless gas that can be detected by taste and odor in ambient air at
concentrations as low as 0.3 ppm. Above 3 ppm, 862 has a pungent, irritating odor similar to a
struck match. In addition to having a bad odor, high concentrations of sulfur dioxide can affect
breathing, cause respiratory illnesses, and aggravate existing respiratory and cardiovascular
diseases.
SC>2 is found at appreciable levels (i.e., low ppb) in the lower atmosphere (troposphere),
where it is oxidized to sulfuric acid (H2SO4), by both gas-phase photochemical reactions and
aqueous-phase reactions with hydrogen peroxide and other oxidants in cloud water. The
gas-phase and aqueous-phase oxidation pathways are roughly equally important in converting
SO2 to H2SO4. Sulfuric acid is non-volatile and condenses into particulate form, typically with
partial to complete neutralization by atmospheric ammonia (NHa). Sulfates account for a
substantial fraction of suspended particulate matter in urban air. They can be transported long
distances and return to the earth as a major component of acid rain.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 2 of 31
3.1.2 Historical Overview of Measurement Method
The UV fluorescence monitoring method for atmospheric 862 was developed to improve
upon the flame photometric detection (FPD) method for 862, which in turn had displaced the
pararosaniline wet chemical method for 862 measurement. The pararosaniline method is still the
U.S. EPA's Reference Method for atmospheric SO2, but is rarely used because of its complexity
and slow response, even in its automated forms. Both the UV fluorescence and FPD methods are
designated as Equivalent Methods by EPA, but UV fluorescence has largely supplanted the FPD
approach because of the UV method's inherent linearity, sensitivity, and the absence of
consumables, such as the hydrogen gas needed for the FPD method. Numerous vendors supply
UV fluorescence 862 analyzers, including a few who supply high sensitivity analyzers
(examples are provided below).
3.2 Summary of SO2 Measurement by UV Fluorescence
The current method for the measurement of 862 is based on the principle that 862
molecules absorb ultraviolet (UV) light at one wavelength and emit UV light at a different
wavelength. This process is known as fluorescence, and involves the excitation of the SO2
molecule to a higher energy electronic state by light absorption. Once excited, the molecule
decays non-radiatively to a lower energy electronic state from which it then decays to the
original, or ground, electronic state by emitting a photon of light at a longer wavelength
(i.e., lower energy) than the original excitation light. The process can be summarized in the
following equations:
[Eq-3-1]
S07^S07+hv7 [Eq. 3-2]
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 3 of 31
where SO 2* represents the excited state of SC>2, hv} and hv2 represent the energy of the excitation
and fluorescence photons, respectively, and hv2 < hv}. The intensity of the emitted light is
proportional to the number of 862 molecules in the sample gas.
Figure 3-1 shows a schematic illustration of the optical chamber of a commercial SO2
analyzer. Light from a high intensity UV lamp passes through a bandwidth filter, allowing only
photons with wavelengths around the SC>2 absorption peak (near 214 nm) to enter the optical
chamber. The light passing through the source bandwidth filter is collimated using a UV lens
and passes through the optical chamber, where it is detected on the opposite side of the chamber
by the reference detector. A photomultiplier tube (PMT) is offset from and placed perpendicular
to the light path to detect the 862 fluorescence. Since the 862 fluorescence (330 nm) is at a
wavelength that is different from the excitation wavelength, an optical bandwidth filter is placed
in front of the PMT to filter out any stray light from the UV lamp. A lens is located between the
filter and the PMT to focus the fluorescence onto the active area of the detector and optimize the
fluorescence signal.
Window,1 Seal
Figure 3-1. Schematic illustration of the optical chamber of a precursor SO2 analyzer.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 4 of 31
3.3 Recommendations for NCore
Since the high sensitivity 862 analyzers deployed at NCore sites are intended to monitor
low ambient 862 concentrations, it is important that they meet a variety of performance criteria.
Many of these performance criteria are more stringent than those for routine SO2 analyzers;
consequently, a number of features are required in the high sensitivity SC>2 analyzers in order to
achieve the performance criteria. This section describes the recommended performance criteria
and the analyzer features that are recommended in order to achieve the performance criteria, and
provides examples of commercial high sensitivity SC>2 analyzers that are available for
deployment at the NCore sites. Additionally, this section discusses some important sampling
requirements that should be considered during the installation of the analyzers.
3.3.1 Recommended Method Performance Criteria
The U.S. EPA has recently assessed the measurement quality objectives needed for high
sensitivity precursor gas monitoring in NCore, relative to the long-established statistics stated in
40 CFR 58.[1] In particular, EPA recommends that measurement quality objectives for bias and
precision be based on upper confidence limits at the monitoring site level, to provide a higher
probability of reaching appropriate conclusions (e.g., in comparisons to NAAQS). The intent of
this recommendation is to move S/L/T agencies to a performance-based quality system i.e.,
allowing organizations that show tight control of precision and bias to reduce the frequency of
certain QC checks, and to focus their quality system efforts where most needed.
The U.S. EPA recommends that the high sensitivity SO2 analyzers that are deployed at
NCore sites meet the following method performance criteria.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 5 of 31
3.3.1.1 Precision
Precision is defined as the measure of agreement among individual measurements of the
same property taken under the same conditions. Precision is assessed from checks that are
performed at least once every two weeks (see Sections 3.7.4 and 5.3.3). Calculations to assess
precision are given below and should be used to assess precision on a quarterly basis. It is
recommended that high sensitivity 862 analyzers have a 95 percent probability limit for
precision of ±15 percent or less.
Calculation of precision starts with the comparison of the known challenge concentration
used in the precision checks to the corresponding measured concentrations reported by the
analyzer. The resulting percent differences are then used as described below. For each single
point precision check, calculate the percent difference, df, as follows:
Equation 1
reported - challenge
d,= —-100
challenge
where reported is the concentration indicated by the high sensitivity SC>2 analyzer and challenge
is the concentration of the standard used in the precision check. The precision estimator is then
calculated as the coefficient of variation (CV) upper bound, using Equation 2 as follows:
Equation 2
where n is the number of data points (i.e., precision check comparisons), the dt values are the
resulting percent differences, and X o.i,n-i is the 10th percentile of a chi-squared distribution with
n-1 degrees of freedom.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 6 of 31
3.3.1.2 Bias
Bias is defined as a systematic or persistent distortion of a measurement process that
causes errors in one direction. Bias is assessed from the degree of agreement between a
measured value and the true, expected, or accepted value. Analyzer bias is calculated using
comparisons of known challenge concentrations to the corresponding measured concentrations
reported by the analyzer. The challenge comparisons used to assess bias should be the same as
those used to assess precision (see Section 3.3.1.1 above). The bias estimator is an upper bound
on the mean absolute value of the percent differences as described in Equation 3 as follows:
Equation 3
AS
•-=
v«
where n is the number of challenge comparisons being aggregated; t0.9s,n-i is the 95th quantile of
a t-distribution with n-1 degrees of freedom; the quantity AB is the mean of the absolute values
of the individual dt 's and is calculated using Equation 4 as follows:
Equation 4
and the quantity AS is the standard deviation of the absolute value of the dt's and is calculated
using Equation 5 as follows:
Equation 5
AS=
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 7 of 31
Since the bias statistic as calculated in Equation 3 uses absolute values, it does not have a
direction or sign (negative or positive) associated with it. The sign of the calculated bias is to be
determined by rank ordering the percent differences of the QC check samples from a given
analyzer for a particular assessment interval. Calculate the 25th and 75th percentiles of the
percent differences for each analyzer. The absolute bias upper bound should be flagged as
positive if both the 25th and 75th percentiles are positive, and as negative if both these percentiles
are negative. The absolute bias upper bound would not be flagged if the 25th and 75th percentiles
are of different signs. It is recommended that high sensitivity SC>2 analyzers have an upper
bound for the average bias of ± 15 percent or less.
3.3.1.3 Representativeness
Representativeness refers to whether the data collected accurately reflect the conditions
being measured. It is the data quality indicator most difficult to quantify. Unless the samples are
truly representative, the other indicators are meaningless. Representativeness for monitoring of
low ambient levels of SC>2 in NCore is different than for routine monitoring, since the objectives
of the monitoring are much different. Representativeness can only be assured in terms of the
appropriate selection of the sampling site, proper implementation of ambient air sampling, and
reasonable coverage of the sampling schedule (i.e., 24 hours per day, 7 days per week, ideally).
3.3.1.4 Completeness
Completeness is defined as the amount of data collected relative to the total expected
amount. Ideally, 100 percent of the expected amount of data would always be collected; in
practice, completeness will be less for many reasons, ranging from calibration time and site
relocation to power outages and equipment failure. For monitoring of ambient 862
concentrations in NCore, EPA requires a minimum data completeness of 75 percent. In practice
typical completeness values can often approach 90 to 95 percent.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 8 of 31
3.3.1.5 Comparability
Comparability is defined as the process of collecting data under conditions that are
consistent with those used for other data sets of the same pollutant. The goal is to ensure that
instruments purchased and operated by different states and local agencies produce comparable
data. To promote comparability, this TAD describes the recommended characteristics of high
sensitivity 862 analyzers and the procedures for their installation and use. For example, all
monitoring agencies should purchase instruments that have the additional features described in
Section 3.3.2, and should adhere to the sampling requirements described in Section 3.3.3.
3.3.1.6 Method Detection Limit
The MDL refers to the lowest concentration of a substance that can be reliably
determined by a given procedure. The MDL is typically not provided by the vendor. Based on
the objectives of the Precursor Gas Program, it is expected that most sites will be measuring
pollutant concentrations at lower ranges than the typical SLAMS/NAMS network. Therefore,
the ability to quantify concentrations at these lower levels will be very important. The use of a
vendors advertised LDL is sufficient to make intelligent purchasing decisions. Vendors quantify
LDLs under ideal conditions and therefore one might consider this value as the best possible
detection that can be achieved. As these monitors are deployed into monitoring networks, where
both environmental conditions, equipment (calibration, dilution devices, sampling lines, gaseous
standards) and operator activities can vary, it is important to estimate what pollutant
concentrations can truly be detected, above background noise (the potential conditions
mentioned above). The site specific MDL establishes an estimate based on the routine operation
(and conditions) of that instrument in the network and provides a more meaningful evaluation of
data as it is aggregated across the precursor gas network. By establishing site specific MDLs,
values less than the MDL can be flagged which would allow data users a more informed decision
on the use of that data.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 9 of 31
The MDL should be established on-site by supplying the analyzer at least seven times
with a test atmosphere containing 862 at a concentration that is approximately one to five times
greater than the estimated MDL, and recording the response. To perform the MDL test, run zero
air through the analyzer and establish an acceptable zero; dilute pollutant gas to the targeted
concentration (one to five times the estimated MDL) and collect 20 to 25 one minute
observations. Repeat this seven times over the course of 5 to 14 days. Average the concentration
from the 20-25 readings; calculate the standard deviation (S) of the average readings and
compute the MDL. The MDL is then calculated as the standard deviation of the response values
times the Student's t-value for the number of test measurements (40 CFR Part 136, Appendix B).
The MDL for high sensitivity SC>2 analyzers should be 0.30 ppb or lower over an averaging time
of no more than 5 minutes.
3.3.1.7 Lower Detectable Limit
The LDL is the minimum pollutant concentration that produces a signal of twice the
noise level. To estimate the LDL, zero air is sampled and the noise level of the 862 readings is
determined according to 40 CFR 53.23(b). The vendor-specified LDL for the most sensitive
range of high sensitivity SO2 analyzers should be 0.20 ppb or lower, over an averaging time of
no more than 5 minutes.
3.3.1.8 Linear Range
The linear range of each high sensitivity SC>2 analyzer should extend from approximately
0.20 ppb to at least 100 ppb. Users should determine if their range should exceed 100 ppb and
adjust accordingly. A range of 100 ppb may not be sufficient in all areas and situations. Note
that some high sensitivity 862 analyzers can operate simultaneously on a number of ranges, with
each range recorded on a separate data logger channel with its own calibration curve. Although
requiring slightly more effort to calibrate and maintain, recording of multiple ranges would allow
capture of a wide range of SC>2 concentrations.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 10 of 31
3.3.1.9 Zero/Span Drift
Zero drift is defined as the change in response to zero pollutant concentration, over
12- and 24-hour periods of continuous unadjusted operation. Span drift is defined as the percent
change in response to an upscale pollutant concentration over a 24-hour period of continuous
unadjusted operation. Zero and span drift specifications should be obtained from the vendor
prior to putting a high sensitivity 862 analyzer into service. Such 862 analyzers should have 12-
and 24-hour zero drift less than 0.20 ppb, and should have a span drift of less than ±1 percent of
the full scale measurement range of the analyzer per 24 hours.
3.3.1.10 NO Rejection Ratio
The NO rejection ratio refers to the effectiveness with which fluorescent emission from
nitric oxide (NO) is blocked in a UV fluorescence SO2 analyzer. This interfering emission can
be greatly reduced by optical filtering of the light reaching the PMT. For high sensitivity SO2
monitoring in NCore, it is recommended that the NO rejection ratio of the SO2 analyzer be at
least 100 to 1, i.e., 100 ppb of NO must produce a response equivalent to that from no more than
1 ppb of SO2.
3.3.2 Recommended Features for High Sensitivity SC>2 Measurements
Continuous UV fluorescence SO2 analyzers are commercially available from a number of
vendors. The design of these analyzers is similar among vendors with some slight variations. A
diagram of the typical UV fluorescence instrument is shown above in Section 3.2, and examples
of specific instruments are provided in Section 3.3.3. In general, each of the analyzers contains
the following systems:
• Pneumatic System: This portion of the analyzer consists of sample probe,
sample inlet line, particulate filter, hydrocarbon scrubber/kicker, dryer (if
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 11 of 31
needed), sample cell, flow meter, and pump, all used to bring ambient air from
the inlet to the defector.
• Analytical System: This portion of the analyzer consists of the UV source
with the associated source filters, lenses, and optics, as well as the light
baffles, the detector (photomultiplier tube), and bandpass filters.
• Electronic Hardware: This portion of the analyzer consists of the electronic
components that control the analyzer and process the signals. This part of the
analyzer generally requires little or no maintenance. If the instrument is
operated outside the manufacturer's recommended temperature range,
however, individual integrated chips can fail and cause problems with
operation, data storage, or retrieval.
In addition to these general systems, the high sensitivity versions of the commercial
UV fluorescence 862 analyzers typically have the following features that allow them to measure
SO2 at sub-ppb levels:
• A high intensity pulsed UV light source that provides a greater degree of
sensitivity;
• Multiple reflective optical filters that allow only light at the wavelength
causing excitation of the SC>2 molecules to enter the optical chamber, while
excluding all light at wavelengths that may cause interference; and
• Optical filtering to maximize the rejection of fluorescence from NO
molecules.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 12 of 31
It is recommended that the high sensitivity SC>2 analyzers deployed in NCore employ these
features. Examples of commercial analyzers with these features are presented below.
3.3.3 Commercial High Sensitivity UV Fluorescence SCb Monitors
High sensitivity 862 analyzers are commercially available from a number of vendors.
Examples of such analyzers discussed in this section are closely similar to instrumentation
designated as Federal Equivalent Methods; however, modifications to improve measurement
sensitivity have been made. The following three instruments are shown as examples of
commercially available high sensitivity SC>2 analyzers (no endorsement should be inferred). The
descriptions provided for these instruments and the performance claimed for them are based
solely on information provided by the respective vendors, and confirmation by EPA should not
be inferred.
3.3.3.1 Ecotech Model EC9850T
The Ecotech Model EC9850T SC>2 analyzer^ shown in Figure 3-2 is a UV fluorescence
SC>2 analyzer specifically designed to measure background concentrations of SC>2 with a lower
detection limit of 200 ppt (0.2 ppb). The EC9850T is a high sensitivity version of the Ecotech
ML9850/EC9850 that has a U.S. EPA Equivalent Method Designation of EQSA-0193-092. The
higher sensitivity is achieved in part by using a specially selected high output UV lamp operating
at 214 nm, a Kalman digital filter that continuously provides a compromise between response
time and noise reduction, and a special high performance 862 scrubber impregnated with
Na2CC>3 solution. Scrubbed sample air is used for zeroing of the analyzer's response. The
EC9850T SO2 analyzer uses a 360 nm filter to reduce interference from NO and potential
interference from other species. The analyzer has multiple range settings up to 0 to 200 ppb full
scale, and has a recommended operating temperature range of 20 °C to 30 °C, but may be
operated between 15 °C and 35 °C. The EC9850T SC>2 analyzer has both analog and digital
outputs. Table 3-1 shows the specifications of the Model EC9850T.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 13 of 31
Figure 3-2. Ecotech EC9850T SO2 analyzer (courtesy of Ecotech).
Table 3-1. Ecotech EC9850T SO2 analyzer specifications.
Performance
Parameters
Precision
Bias
Method Detection Limit
Lower Detectable Limit
Linearity
Zero Drift
Span Drift
NO Rejection Ratio
EC9850T SO2 Analyzer Specifications
± 2% of reading
Not available
Not available
200 ppt (0.2 ppb) (with Kalman or 300 sec filter active)
Not available
Temperature dependence, 0.1% per degree C changes.
24 hours; less than 200 ppt (0.2 ppb)
Temperature dependence, 0.05 % per degree C changes.
24 hours less than 0.5% of reading
30 days less than 1 .0% of reading
Not available3
a: Vendor states rejection ratio is high, but difficult to quantify; dependent on water vapor content of air.
3.3.3.2 Thermo Electron Corporation Model 43C-TLE
The Model 43C-TLE enhanced high sensitivity pulsed fluorescence SC>2 analyzer ' is
shown in Figure 3-3. The Model 43C-TLE SC>2 analyzer is a successor to the Model 43C
analyzer, which has a U.S. EPA Equivalent Method Designation of EQSA-0486-060. As
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 14 of 31
recommended in Section 3.3.2, the Model 43C-TLE achieves higher sensitivity than the
Model 43C through the use of additional reflective bandpass filters to further reduce stray light,
and by operation of the UV flash lamp at a higher voltage, which increases the total output light
intensity. Additionally, the Model 43C-TLE also includes a longer optical bench that increases
the effective sample path length and improves sensitivity. Furthermore, optimization of the
optical properties of the detector filter allows for greater rejection of NO interference. This
feature allows an instrument originally designed for rural background monitoring applications to
also be used in urban monitoring programs such as NCore Level 2.
The Model 43C-TLE 862 analyzer has multiple range settings up to 1 ppm full scale and
may be safely operated in the temperature range of 0 °C to 45 °C. The Model 43C-TLE analyzer
displays the 862 concentration on a screen on the front panel of the analyzer, and has both
analog and digital outputs. An optional capability of the 43C-TLE is for generation of internal
zero and span mixtures using an SC>2 permeation tube, although this means of calibration is not
recommended for use in NCore. Table 3-2 shows the specifications of the Model 43C-TLE.
Figure 3-3. Thermo Electron Model 43C-TLE SO2 analyzer (courtesy Thermo
Electron).
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 15 of 31
Table 3-2. Thermo Electron Model 43C-TLE SO2 analyzer specifications.
Performance
Parameters
Precision
Bias
Method Detection Limit
Lower Detectable Limit
Linearity
Zero Drift
Span Drift
NO Rejection Ratio
Model 43C-TLE SO2 Analyzer Specifications
1% of reading or 0.2 ppb
Not available
Not available
0.2 ppb (10 sec avg. time)
0.10 ppb (60 sec avg. time)
0.05 ppb (300 sec avg. time)
± 1% of full scale
less than 0.2 ppb per day
± 1 % per week
<2.5 ppb equivalent with 500
ppb NO interference
3.3.3.3 Teledyne Advanced Pollution Instrumentation (API) Model 100AS
The Teledyne API Model lOOAS SC>2 analyzer ' ' shown in Figure 3-4 is based on the
Model 100A fluorescence SO2 Analyzer. Both the Model 100A and the Model 100AS analyzers
are designated as U.S. EPA Equivalent Method Number EQSA-0495-100. The sensitivity and
stability of the Model 100 AS analyzer is achieved in part through the use of an optical shutter to
compensate for PMT drift and a reference detector to correct for changes in UV lamp intensity.
A hydrocarbon "kicker" and advanced optical design combine to minimize inaccuracies due to
interferents, including NO. The Model 100AS has sensitivity ranges of 0 to 10 ppb full scale up
to 0 to 1 ppm full scale and can measure SC>2 with a lower detectable limit of 0.10 ppb. The
analyzer may be safely operated in the temperature range of 5 °C to 40 °C. The Model 100AS
has both analog and digital outputs. An optional capability of the Model 100AS is for generation
of internal zero and span mixtures, using an 862 permeation tube. Table 3-3 shows the
specifications of the Model 100AS.
Note: The Model 100AS is not API's "true high sensitivity " analyzer, and the Model lOOEUhigh
sensitivity analyzer is expected to be commercially available this year.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 16 of 31
Figure 3-4. Teledyne API Model 100AS SO2 analyzer (courtesy Teledyne API).
Table 3-3. Teledyne API Model 100AS SO2 analyzer specifications.
Performance
Parameters
Precision
Bias
Method Detection Limit
Lower Detectable Limit
Linearity
Zero Drift
Span Drift
NO Rejection Ratio
Teledyne API Model 100AS SO2 Analyzer
Specifications
0.5% of reading
Not available
Not available
100 ppt (0.1 ppb) (4 minute averaging time)
1% full scale
< 0.2 ppb/24 hours
< 0.5% reading/24 hours
100:1 standard; 250:1 with optional filter
3.3.4 Sampling Requirements
Proper siting of the sampling equipment and sampling probes is necessary to ensure that
the gas analyzers are obtaining representative samples of the ambient air. Likewise, proper
environmental control of the analyzer and proper sampling are critical to ensuring that the
analyzers are operating correctly and that the SO2 measurements are comparable to SO2
measurements recorded at other sites.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 17 of 31
3.3.4.1 Analyzer Siting
Analyzer siting should follow the criteria in 40 CFR58, Appendix E. The installation of
the precursor SC>2 analyzer should allow for the sample manifold inlet to be located between 3
and 15 meters above ground level, with at least one meter of vertical and horizontal separation
from supporting structures. The probe should be positioned with at least 270 degrees of
unrestricted airflow including the predominant wind direction. The probe should be separated
from the drip line of nearby trees or structures by at least 20 meters, and should be positioned at
least twice as far from of nearby obstacles as the height of the obstacles.
3.3.4.2 Instrument Shelter
To help ensure proper performance, the precursor analyzers and supporting equipment
should be installed and operated in a temperature controlled environment. An insulated
instrument shelter should be used to protect the analyzer from precipitation and adverse weather
conditions, maintain operating temperature within the analyzer's temperature range
requirements, and provide security and electrical power. The environmental control of the
shelter should be sufficient to minimize fluctuations in shelter temperature. The recommended
shelter temperature range is 20°C to 30°C, and daily changes in temperature should not exceed
5°C over a 24-hour period. Condensation of moisture must be prevented, and it may be
necessary to impose seasonal temperature ranges to assure remaining above the ambient
dewpoint.
3.3.4.3 Sample Inlet or Manifold
Sample air for a high sensitivity SCh analyzer should be drawn directly to the analyzer
through FEP or PTFE tubing (e.g., !/4 inch outer diameter (OD)), equipped with a 1-micron
Teflon® filter at the inlet to remove particles from the sample air. The filter should be changed
weekly, or more often in excessively dirty conditions. In this configuration the SC>2 analyzer
does not share its inlet system with any other analyzer.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 18 of 31
Some existing air monitoring stations may provide air to multiple analyzers through a
common manifold. In such a case it is recommended that manifolds for precursor 862
measurements be made of glass. Since neither zero air nor sample air is totally particulate-free,
over time sample manifolds will collect particulate matter on the internal walls. A transparent
glass manifold can be inspected easily and cleaned readily by rinsing with distilled water and air
drying. However, caution must be used with glass manifolds because of their fragility.
The SC>2 analyzer should be located as close to the inlet manifold as possible, to minimize
the length of sampling lines, and sampling lines should be of constructed of FEP or PTFE. The
sample manifold must be of sufficient diameter that outside air is drawn into the manifold at as
close to atmospheric pressure as possible. The manifold must also allow for excess gas to be
exhausted in the event of over pressurization (i.e., during delivery of gases from high-pressure
cylinders.) If the pressure in the manifold differs from atmospheric pressure, the SO2 readings
obtained will not be representative. A manifold of 1 inch inner diameter should be sufficient to
avoid pressure differences. A detailed description of sample manifold designs is presented in
Appendix A.
3.4 Potential Problems and Solutions
This section describes several of the potential problems associated with precursor SC>2
measurements, and discusses the practical solutions to these problems, many of which the SC>2
analyzer vendors have already implemented in their analyzers.
The following sections describe several potential positive and negative sources of
interference or bias. Section 3.4.1 describes the most common positive interferences (volatile
aromatic and poly-nuclear aromatic hydrocarbons, and NO), and the most common source of
positive bias (stray light). Section 3.4.2 describes the most common sources of negative bias
(collisional quenching of SC>2 and loss of SC>2 in sampling lines). In each section recommended
procedures to minimize these interferences or sources of bias are described.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 19 of 31
In addition to these potential problems, other problems may arise in the routine operation
of precursor 862 analyzers. Operators are encouraged to report any observations or anecdotal
data that might add to the understanding or awareness of interferences or other anomalies in 862
measurements with UV fluorescence analyzers in general.
3.4.1 Sources of Positive Interference or Bias
Positive interference in precursor SC>2 monitoring can result from other gases in the
sample that happen to fluoresce at the same wavelength as SC>2. Perhaps the most prevalent
sources of this type of interference are volatile aromatic (e.g., xylenes) and poly-nuclear aromatic
(PNA) (e.g., naphthalene) hydrocarbons. Such compounds absorb UV photons and fluoresce in
the region of the 862 fluorescence. Consequently, any such aromatic hydrocarbons that are in
the optical chamber can act as a positive interference. To remove this source of interference, the
high sensitivity SC>2 analyzers have hydrocarbon scrubbers or "kickers" to remove these
compounds from the sample stream before the sample air enters the optical chamber. Another
potential source of positive interference is nitric oxide (NO). NO fluoresces in a spectral region
that is close to the SO2 fluorescence. However, in high sensitivity SO2 analyzers, the bandpass
filter in front of the PMT is designed to prevent NO fluorescence from reaching the PMT and
being detected. Care must be exercised when using multicomponent calibration gases containing
both NO and SO2 that the NO rejection ratio of the SO2 analyzer is sufficient to prevent NO
interference.
The most common source of positive bias (as opposed to positive spectral interference) in
high sensitivity SO2 monitoring is stray light reaching the optical chamber. Since SO2 can be
excited by a broad range of UV wavelengths, any stray light with an appropriate wavelength that
enters the optical chamber can excite SO2 in the sample and increase the fluorescence signal.
Furthermore, stray light at the wavelength of the SO2 fluorescence that enters the optical
chamber may impinge on the PMT and increase the fluorescence signal. The analyzer
manufacturers incorporate several design features to minimize the stray light that enters the
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 20 of 31
chamber. These features include the use of light filters, dark surfaces, and opaque tubing to
prevent light from entering the chamber.
3.4.2 Sources of Negative Interference or Bias
Non-radiative deactivation (quenching) of excited 862 molecules can occur from
collisions with common molecules in air, including nitrogen, oxygen, and water. During
collisional quenching, the excited SC>2 molecule transfers energy kinetically allowing the SC>2
molecule to return to the original lower energy state without emitting a photon. Collisional
quenching results in a decrease in the SC>2 fluorescence and results in the underestimation of SC>2
concentration in the air sample. The concentrations of nitrogen and oxygen are constant in the
ambient air, so quenching from those species at a surface site is also constant, but the water
vapor content of air can vary. Despite this variability, in routine ambient monitoring the effect of
water vapor on SC>2 fluorescence measurements is negligible. Only if high or highly variable
water vapor concentrations were a concern (as in source sampling), should it be necessary to dry
the sample air using optional equipment available from the analyzer vendors. Condensation of
water vapor in sampling lines must be avoided, as it can absorb SC>2 from the sample air. The
simplest approach to avoid condensation is to heat sampling lines to a temperature above the
expected dew point, and within a few degrees of the controlled optical bench temperature. An
alternative approach would be to maintain all sampling lines at reduced pressure by locating the
analyzer's critical orifice at the sample inlet point.
3.5 Supporting Equipment
In addition to the precursor gas analyzers, several pieces of supporting equipment should
be maintained at each NCore site. At a minimum, this equipment includes a data acquisition
device and calibration equipment.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 21 of 31
3.5.1 Data Acquisition Device
Many types of equipment can be used to record the concentration measurements obtained
from the analyzer. Recommended options for data acquisition are described in Section 6 of this
TAD.
3.5.2 Calibration Equipment
The equipment required for calibration of a high sensitivity SC>2 analyzer include a MFC
calibrator unit, and a source of zero air. The following equipment is recommended for
calibration of a high sensitivity 862 analyzer.
3.5.2.1 Calibration Standard and Standard Delivery System
The calibration standards used for the calibration of high sensitivity 862 analyzers should
be generated by dilution of a commercially-prepared and certified compressed gas 862 standard
using a MFC calibration unit. That commercially-prepared standard may contain only 862 in an
inert gas (e.g., N2), or may be a mixed component standard that also contains known
concentrations of other precursor ambient gases (e.g., CO, NO). However, note the caution
stated in Section 3.4.1 regarding potential NO interference in mixed standards containing SO2
and NO.
It is important when purchasing a MFC calibrator that it meets the 40 CFR 50 requirements
of ±2 percent accuracy, and that the flow rates of both MFC channels are calibrated using a NIST
traceable flow standard.
When the analyte concentration in the commercially-prepared standard cylinder is certified
by reference to NIST standards, and the MFCs are calibrated to NIST-traceable standards, the
resulting working gas concentration is considered to be NIST-traceable.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 22 of 31
3.5.2.2 Zero Air Source/Generator
Zero air is required for the calibration of precursor SC>2 instruments. This air must
contain no detectable SC>2 (i.e., SC>2 content must be less than the LDL of the SC>2 analyzer) and
be free of particulate matter. Suitable zero air may be supplied from compressed gas cylinders of
purified air. However, it may be expensive to maintain a sufficient supply of zero air cylinders
to operate a precursor 862 analyzer continuously. As an alternative, many commercially
available zero air generation systems can supply suitably SCh-free air.
To ensure that the zero air used is free from contaminants, the SO2 analyzer should be
independently supplied with zero air from different sources. If the analyzer responds differently
to the different sources, generally the source with the lowest response is the highest quality
source. Confirmation of zero air quality can be achieved using various additional scrubbing
traps. For example, ambient air can be scrubbed of SC>2 using 24 x 7 purged activated carbon.
The carbon type used for scrubbing is important; Barnebey & Sutcliffe Corp. (formerly
Barnebey-Cheney) type GI (www.bscarbon.com, Columbus, Ohio) has been shown to work well.
As an alternative to using an activated carbon scrubber, a sodium carbonate coated denuder, such
as the Sunset Laboratory Model #DN-315 stainless steel concentric denuder, can be used.
Alternatively, a cartridge of soda lime attached to the outlet of the zero air system will last for
extended periods (potentially over one year) and maintain SC>2 at less than 0.05 ppb.
Note: For zero-air sources based on removing SO 2 by means of soda lime,
charcoal, or a denuder as described above, the inlet air must be outside
ambient air rather than instrument shelter air. Contaminant levels inside the
shelter may greatly exceed those in outside air. Also to the extent possible, the
components of the zero-air system should be free of materials that might outgas
hydrocarbons.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 23 of 31
3.6 Reagents and Standards
Routine operation of high sensitivity SO2 analyzers requires the use of calibration
standards and zero air to conduct periodic calibrations and instrument checks. This section
describes the requirements for these gases.
3.6.1 Calibration Standards
The primary SO2 standards used must be certified, commercially-prepared compressed
gas standards, with a certified accuracy of no worse than ±2 percent. SO2 gas standards of 10 to
20 ppm are conveniently diluted with a MFC calibrator down to working concentrations of 20
ppb or less. The commercially-prepared standard may contain only SO2 in an inert gas (e.g., N2),
or may be a mixed component standard that also contains known concentrations of other
precursor ambient gases (e.g., CO, NO). The potential for NO interference must be kept in mind
if a standard containing both SO2 and NO is used for SO2 calibration, as noted in Sections 3.4.1
and 3.5.2.1. It is critical when placing an SO2 gas standard into service that the cylinder
regulator be fully purged to avoid the effect of trace moisture on the delivered SO2 concentration.
Evacuating the regulator by means of a vacuum line attached to the regulator outlet, before
purging for a few minutes with the cylinder gas, is an effective procedure to dry and condition
the regulator.
Every gas standard used in precursor gas monitoring must be accompanied by a
certificate of calibration from the vendor stating the concentration of the standard, the
uncertainty of that certification, and the expiration date of the certification. Standards traceable
to NIST are preferred. Certification documents for all standards must be retained in a common
location and reviewed periodically so that standards for which the vendor's certification has
expired may be removed from service and replaced.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 24 of 31
3.6.2 Zero Air
Zero air used as dilution gas for calibration purposes should have an 862 concentration
below the LDL of the 862 analyzer. Multiple zero air sources should be checked, and soda lime,
charcoal, or sodium carbonate scrubbers may be necessary to achieve adequate zero air quality,
as noted in Section 3.5.2.2. A canister of soda lime followed by a particle filter on the outlet of
the zero air source will remove SC>2 for extended periods. Breakthrough can be tested by
temporarily adding a carbonate denuder and observing zero gas readings.
3.7 Quality Control
3.7.1 Site Visit Checklists and Remote Diagnostic Checks
To determine whether the SC>2 analyzer is working properly, field operators should
conduct routine checks of instrument diagnostics and performance every time they visit the
monitoring station. Each agency needs to develop diagnostic or maintenance checklists or
electronic spreadsheets to document that all required checks have been made. Such lists and
sheets should be useful both for collecting diagnostic information and for assessing the quality of
the monitoring data. To the extent possible, diagnostic checks can be done remotely, provided
the data acquisition system allows remote access to instrument diagnostic information (see
Section 6).
3.7.2 Multipoint Calibrations
A multipoint calibration includes a minimum of four points (three spaced over the
expected range and a zero point), generated by the calibration system. Although more points
may be preferable, current high sensitivity 862 analyzers provide inherently linear response over
their entire operating range; therefore, four points should be sufficient. Multipoint calibrations
must be done prior to the precursor 862 analyzer being put into service and at least every
six months thereafter. An analyzer should be calibrated (or recalibrated) if any of the following
conditions occur:
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 25 of 31
• Upon initial installation;
• The Level 1 span check or precision check difference exceeds 15 percent;
• After repairs or service is conducted that may affect the calibration;
• Following physical relocation or an interruption in operation of more than a few days;
• Upon any indication that the analyzer has malfunctioned or a there has been a change
in calibration; or
• The measured concentration values during challenges with performance test samples
(Section 5.4.1) differ from the certified standard values by ±15 percent.
The analyzers should be calibrated in-situ without disturbing the normal sampling inlet system to
the degree possible.
3.7.3 Level 1 Zero/Span Checks
Level 1 zero and span calibrations are simplified, two-point calibrations used when
adjustments may be made to the analyzer. When no adjustments are made to the analyzer, the
Level 1 calibration may also be called a zero/span "check" and must not be confused with a level
2 zero and span check. Level 1 zero and span checks should be conducted nightly if the
calibration system and 862 analyzers used can be programmed to automatically perform these.
They are used to assess if the analyzers are operating properly and to assess if any drift in
instrument response has occurred. They are conducted by challenging the analyzer with zero air
and a test atmosphere containing SC>2 at a concentration of between 70 percent and 90 percent of
the full measurement range in which the analyzer is operating. The challenge gas should be
sampled through as much of the sampling inlet system as practical to mimic the actual sampling
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 26 of 31
of ambient air. The results of the Level 1 zero/span check should be plotted on control charts to
graphically illustrate the trends in the response of the analyzer to the challenge gases. The span
check should not exceed ±15 percent and the zero drift should not exceed ±0.5 percent of full
scale. If the measured concentrations fall outside of the control limits, the accuracy of the MFC
calibration system should be checked with a NIST-traceable flow standard. If the MFC flow
accuracy is confirmed, the data recorded since the last successful Level 1 check should be
flagged and the analyzer should be recalibrated using the multipoint calibration procedures
described in Section 3.7.2.
State-of-the-art calibration equipment now exists that is fully automated. These "new
generation" calibration units are fully integrated with computers, mass flow calibrators, and the
associated hardware and software where they can create test atmospheres manually or
automatically. For the precursor gas program, it is recommended that the NCore sites have fully
automated calibration capability. Below are a number of reasons why this is advantageous:
• By performing the calibrations or checks automatically, agencies no longer spend the
manpower needed to perform them.
• Automated calibrations or checks can be triggered internally or by a DAS. Since newer
DASs allow remote access, this allows a remote user to challenge the analyzers without
actually being present.
• High sensitivity precursor gas analyzers are expected to have more zero and span drift than
less sensitive analyzers; therefore, it is important that a zero and Level I check be performed
daily.
• New generation DASs can record calibration data and allow remote users to track daily Level
I check and zero drift. This is important for data validation, verification and troubleshooting.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 27 of 31
3.7.4 Precision Checks
At least once every two weeks a precision check should be conducted by challenging the
SC>2 analyzer with a known low 862 concentration to assess the performance of the analyzer.
The precision checks should be conducted by challenging the 862 analyzer with a standard gas
of known concentration between 10 and 50 ppb. After completion of the precision check, the
operator should calculate the percent difference between the measured value and the standard
value. Precision should be calculated quarterly, using the calculated percent differences from the
precision checks, according to the equations provided in section 3.3.1.1. For acceptable
precision to be maintained, it is recommended that the calibration system's gas flows be verified
frequently against a NIST flow standard, and adjusted if necessary before making any
adjustments to the analyzer.
3.8 Preventive Maintenance and Troubleshooting
Long-term operation of continuous high sensitivity precursor gas analyzers requires a
preventive maintenance program to avoid instrument down-time and data loss. Despite active
preventive maintenance, occasional problems may arise with the precursor SO2 analyzers. This
section briefly describes several key items that might be included in the preventive maintenance
program established for precursor SC>2 analyzers deployed atNCore sites, as well as some of the
troubleshooting activities that may be useful in resolving unexpected problems with these
analyzers. This discussion is not meant to be exhaustive or comprehensive in detail. More
thorough discussions can be found in the analyzer operation manuals, and should be included in
SOPs developed for these analyzers. Example SOP's prepared by EPA are included as
Appendix B of this TAD.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 28 of 31
3.8.1 Preventive Maintenance
Routine preventive maintenance procedures should be in place to prevent down-time and
data loss. Management and field operators should jointly develop their preventive maintenance
program. A program designed by persons unfamiliar with analyzer operations may include
unnecessary items or omit mandatory ones. Several factors linked to shelter and sampling
manifold design can contribute to data loss. SC>2 values can be low if the sample probe,
manifold, and lines are dirty, cracked, or leaky. The sample probe and manifold should be
cleaned at least every six months. Sampling lines should be replaced every two years. Teflon®
filters used in the sampling train to remove fine particles may need to be replaced as often as
every week, depending on the condition of the filter and the particulate loading around the
monitoring site.
Table 3-4 illustrates items that monitoring agencies should include in their preventive
maintenance program for high sensitivity SC>2 monitoring.
Table 3-4. Example of a preventive maintenance schedule for high sensitivity SO2
analyzers.
Item
Replace particle filter
Replace internal span permeation tube (if applicable)
Perform pneumatic system leak check
Inspect internal, external tubing; replace if necessary
Rebuild or replace pump
Replace UV lamp
Clean optic bench
Replace PMT
Schedule
Weekly
Annually
At least quarterly
At least quarterly
Annually
As needed
As needed
As needed
In addition to a schedule, the preventive maintenance plan should also include more
detailed task descriptions, such as illustrated below:
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 29 of 31
• Because the analyzer pneumatic system requires so much preventive maintenance, the
tubing, solenoids, and pump should be inspected regularly. Cracked tubing or loose
fittings can cause the instrument to analyze room air rather than ambient air and lead
to invalid data. A faulty pump can also cause problems with pneumatic systems.
When oscillations in the flow rate force the operator to adjust the flow continually,
the pump is failing and should be either repaired or replaced. The pump should be
rebuilt or replaced when it is unable to maintain a vacuum of at least 25 inches of Hg.
• Check the instrument for vibration. When pumps get old, they sometimes will vibrate
more than is normal. If this occurs, it can cause cracks if the tubing is touching
another surface.
• Consult the analyzer operations manual for complete details on operation and
maintenance.
3.8.2 Troubleshooting
High sensitivity 862 analyzers are subject to many factors that can cause inaccurate
measurements or down-time. Table 3-5 summarizes common problems seen with high
sensitivity 862 analyzers, their possible causes, and possible solutions. More specific
information can be found in the manufacturer's operations manual.
When troubleshooting, an operator must constantly be aware of environmental factors
that may affect the instruments. Environmental factors can also cause sporadic problems that
can be difficult to diagnose. Examples of factors that may affect the performance of the high
sensitivity 862 analyzers are:
• Variable shelter temperature (fluctuations greater than several degrees)
• Excessive vibration from other equipment
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 30 of 31
Voltage instability; fluctuations in the 110 VAC line voltage
Air conditioning system blowing on the instrument
Frequent opening of the door of the shelter.
Table 3-5. Instrument troubleshooting for precursor SO2 analyzers.
Problem
Noisy output
High positive zero drift
No response to span gas
Zero output at ambient levels
No flow through analyzer
Possible Cause
Defective DC power supply
Dirty optics
PMT failure
Defective bandpass filter
PMT failure
UV source is defective
UV power supply defective
PMT failure
Pump failure
UV lamp failure
UV power supply defective
PMT failure
Pump failure
Possible Solution
Replace power supply
Clean optics bench
Replace PMT
Replace filter
Replace PMT
Replace UV lamp
Replace UV power supply
Replace PMT
Check pump
Replace UV lamp
Replace power supply
Replace PMT
Replace/ rebuild pump head
3.9 References
1. Rhodes, R.C., Guidelines on the Meaning and Use of Precision and Accuracy Data
Required by 40CFR Part 58, Appendices A and B, EPA60014-83-023, U.S.
Environmental Protection Agency, Research Triangle Park, NC, June 1983.
2. Ecotech Pty Ltd. (2004). "EC9850T Trace SO2 Analyzer Product Specification Sheet."
Accessed September, 2004. Available at http://www.ecotech.com.au/brochures_new/
EC9850T.pdf.
3. Thermo Electron Corporation (2004a). "Instruction Manual, Model 43C Precursor SC>2
Analyzer." Accessed September, 2004. Available at https://www.thermo.com/
eThermo/CMA/PDFs/Various/File 20810.pdf.
-------
Precursor Gas TAD
Section 3 SO2 Measurements
Version 4
September 2005
Page 31 of 31
4. Teledyne Advanced Pollution Instrumentation (2004a). "Model 100AS Ultra Sensitivity
SO2 Analyzer Data Sheet." Accessed September, 2004. Available at
http://www.teledyne-api.com/products/Model.100AS.pdf.
5. Teledyne Advanced Pollution Instrumentation (2004b). "Instruction Manual,
Model 100EUV Fluorescence SO2 Analyzer." Accessed September, 2004.
Available at http://www.teledyne-api.com/manuals/04515B2.M100E.Manual.pdf.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page Iof39
4.0 HIGH SENSITIVITY TOTAL REACTIVE NITROGEN OXIDES MEASUREMENTS
4.1 Introduction
Measurement of ambient nitrogen oxides differs from measurement of CO or SC>2 in that
the target air pollutant is not a single chemical but a group of chemicals of differing properties,
and is not a criteria air pollutant. Nitrogen oxides released from emission sources are primarily
nitric oxide (NO) with lesser amounts of nitrogen dioxide (NO2), which collectively are termed
NOX (i.e., NOX = NO + NO2). These primary emitted species are converted by atmospheric
processes to numerous other inorganic and organic nitrogen oxides, which collectively are called
NOZ, and the total of all reactive gaseous nitrogen species present in ambient air is called NOy
(i.e.,NOy = NOx + NOz).
Precursor gas monitoring in the NCore network builds upon capabilities of EPA's
Photochemical Assessment Measurement Stations (PAMS) network and Southern Oxidants
Study to measure ozone precursors, including total reactive oxides of nitrogen (NOy). Measuring
NOy is a valuable adjunct to NO and NOX monitoring because the individual species comprising
NOZ include numerous organic and inorganic nitrogen oxide compounds, that are difficult to
measure individually, but collectively contribute to a more complete and conservative measure
of nitrogen oxides. Determining NOy concentrations is useful in establishing nitrogen oxide
emission patterns and temporal trends, and in assessing the photochemical age and reactivity of
air masses.(e'g''1-4) NOy measurements are a critical tool in accounting for progress in large-scale
nitrogen emission reduction programs, providing input for a variety of source apportionment and
observation based models, and assisting in the evaluation of air quality models. [e'g''l'5'6 ]
4.1.1 Properties of NOY
NOy includes all of the nitrogen oxide compounds that react or are formed in the lower
atmosphere and that contribute to the photochemical formation of Os and the transport and
ultimate fate of nitrogen oxides. [1"4] NOy compounds include NOX (NO + NO2) and NOZ, which
include nitrogen acids [nitric acid (HNOs) and nitrous acid (HONO)], organic nitrates
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 2 of 40
[e.g., peroxyl acetyl nitrate (PAN), methyl peroxyl acetyl nitrate (MPAN), and peroxyl propionyl
nitrate (PPN)], other organic nitrogen oxides, particulate nitrates, nitrate radical (NOs), nitrogen
trioxide (TSPzOs), nitrogen pentoxide (^Os) and halogen-nitrogen species (e.g., C1ONO2 and
BrONO2). In typical urban environments, the principal NOy compounds include NO, NC>2,
HNO3, and PAN, and in some cases particulate nitrate.
In terms of precursor monitoring in NCore, a key factor is that the numerous species
making up the total NOy differ widely in their physical properties and chemical reactivity. For
example, some species, such as NO2 and HONO, are readily photolyzed, whereas others, such as
PAN, decompose rapidly at moderate temperatures. NO and NO2 are chemically reactive but
have relatively low solubility in water, whereas the key product species HNOs is highly soluble
and relatively unreactive. Consequently, physical removal of HNOs from the atmosphere is a
key removal process for NOy. Organic nitrogen oxides can vary widely in volatility and
stability, and HNOs is known to be highly "sticky"; that is, adsorptive on surfaces. In addition,
particulate ammonium nitrate (NFLjNOs) is volatile under certain ambient conditions, and can
decompose to release HNOs and ammonia (NHa) into the gas phase. These factors make
accurate sampling and measurement of atmospheric NOy much more challenging than
determination of CO or SO2. A discussion of sampling and measurement issues that must be
addressed in order to make more useful measurements of NOy is provided in Section 4.3.2.
4.1.2 Sources of
Nitrogen oxides are emitted to the atmosphere principally as NO and NO2, by both
natural and man-made sources. Important natural sources include lightning and natural fires.
The major man-made emissions result from transportation and combustion of fossil fuels for
energy production. Once released into the atmosphere, NO and NO2 are oxidized by
photochemical processes to a wide variety of products. Oxidation of NO to NO2 can occur by
reaction with atmospheric oxygen (only at high NO concentrations that may exist very near the
emissions source), or by reaction with atmospheric ozone (Os) and free radical species. When
NOX is mixed with hydrocarbon air pollutants and exposed to sunlight, a complex set of reactions
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 3 of 40
occurs that constitutes the phenomenon known as photochemical smog. This photochemical
process involves free radicals generated by photolysis and maintained through chain reactions,
and results in the production of large amounts of ozone. Depending on the nature of co-
pollutants, this process can also produce fine paniculate matter containing nitrate, sulfate, and
organic material, and the more complex nitrogen oxide species that make up NOZ (and, in turn,
NOy). The extent of conversion of NOX species to NOy species is a measure of the
"photochemical age" of an air mass; i.e., a measure of the time of transport and the reactivity of
the mix of pollutants in that air mass.(2"4)
4.1.3 Historical Overview of NO^Measurement Method
Ambient NOy must be measured in a practical, standardized manner, as it is not possible
to measure individually all the compounds that comprise NOy. Instruments used to measure NOy
must be sensitive enough to measure the low concentrations typically encountered in rural
locations as well as the higher concentrations encountered in urban smog. The standard
reference method for the determination of NO and NOX at ambient levels is chemiluminescence
(40 CFR Part 53), with several manufacturers offering EPA-approved instruments.
Instrumentation designated as Reference or Equivalent methods for measuring ambient
concentrations of NO2 is listed in 40 CFR Part 53.[7] Instruments designated as Reference
methods for NO2 are also approved for measuring NO. It must be noted that the designated
instruments may not truly measure NOX (i.e., NO plus NO2) in urban areas where photochemical
processes have occurred, butNOx plus some poorly defined fraction of NOZ.
For NOy, a standard reference method has not yet been designated; however, EPA has
suggested a modification of the NOX chemiluminescence monitoring approach that uses a heated
converter to reduce all reactive nitrogen species to NO, followed by detection of that NO by its
chemiluminescence reaction with an excess of Os. The original ambient NO is measured by
bypassing the converter. This procedure is similar to the current methodology used to monitor
NOX except that, in the NOy methodology, the converter has been moved to the sample inlet to
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 4 of 40
avoid line losses of adsorbent NOy species like HNOs, and additional calibration procedures
recommended for adequate measurement of the various NOZ species.
4.2 Summary of NOy Measurement by Chemiluminescence Method
NOy concentrations are determined by photometrically measuring the light intensity at
wavelengths greater than 600 nm from the chemiluminescent reaction of NO with O3.(8) This
principle is identical to that on which the measurement of NO and NOX is based, which is
designated by EPA as the Reference method for determining NO2 in ambient air.[9]
The chemiluminescence approach is based on the gas-phase reaction of NO and Os,
which produces a characteristic near-infrared luminescence (broad-band radiation from 500
to 3,000 nm, with a maximum intensity at approximately 1,100 nm) with an intensity that is
proportional to the concentration of NO. Specifically,
NO + 03 -> N02 + 02 (Eq 4_la)
or -> NO2 * + O2 (Eq. 4-lb)
N02 * -^ N02 (Eq 4_2a)
or -> N02 + hv (Eq.4-2b)
where:
NO2* = an electronically excited NO2 molecule
hv = E, the emitted photon energy (where h is Planck's constant, and v is the
frequency of the emitted photon)
M = inert molecules, predominantly N2 and Q^ in air.
The reaction of NO with Os produces predominantly ground state NO2 molecules, as in
reaction (4-la), but in a small fraction of the reactions, the NO2 produced is in an excited state,
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 5 of 40
as in reaction (4-lb). The resulting excited NC>2 molecules lose their excess energy either
through collisions with other inert molecules (M), as in reaction (4-2a), or by releasing a photon
of light, as in reaction (4-2b). Monitoring instruments based on chemiluminescence are designed
to enhance the emission and detection of the light produced in reaction (4-2b). For example, to
minimize the unproductive pathway in reaction (4-2a), the concentration of other molecules (M)
is kept low by carrying out the NO/Os reaction at low pressure.
To determine the concentration of NO by chemiluminescence, the sample gas flow is
mixed with O?, in a reaction chamber causing reactions (4-1) and (4-2) to occur. The
chemiluminescence that results from the reaction is monitored by an optically filtered
high-sensitivity photomultiplier, that responds to NC>2 chemiluminescence emission at
wavelengths longer than 600 nm. The electronic signal produced in the photomultiplier is
proportional to the NO concentration in the sample air.
To measure NOy, sample air is passed through a chemical reductant (molybdenum)
converter placed at the extreme sample air inlet point, and the nitrogen oxide compounds present
are reduced to NO.[10] The NO resulting from the reduction of these nitrogen oxide compounds,
plus any native NO, is reacted with Os, and the resulting chemiluminescent light is measured as
an indication of 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
nitrogen oxide compounds to NO occurs. The NO (i.e., native NO only) is reacted with Os, and
the resulting chemiluminescent light intensity is proportional to the NO concentration.
The primary differences between this method, as implemented for NOy monitoring and as
implemented for conventional NOX monitoring, are in the location of the molybdenum (Mo)
converter and in the calibration procedures required. The converter location at the extreme inlet
of the sampling system is designed to convert all NOy species to NO immediately upon entry of
sample air into the sampling system. This approach minimizes loss of NOy constituents such as
in sampling, and help to assure complete capture of the total NOy. Calibration procedures
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 6 of 40
for NOy monitoring go beyond those used for NOX monitoring, in that more stringent tests of
converter efficiency are required (see Section 4.4.2 for a discussion of converter efficiency).
Figure 4-1 shows a schematic illustration of a typical NOy instrument. Sample air is
drawn either from the ambient air or from calibration sources (i.e., zero/span gas), using a
three-way solenoid valve (not shown). At the sample inlet, the sample flow is either directed
through a heated molybdenum converter to reduce the reactive oxides of nitrogen to NO, or
directed around the converter to allow detection of only NO. The sample air flow then passes
through a filter to remove particulate matter and then through a flow control capillary to another
three-way valve. This three-way valve directs the sample flow either directly to the reaction
chamber (RX) where it is mixed with Os and the resultant chemiluminescence is measured, or
the sample is directed to a prereactor vessel where it is mixed with Os before passage into the
reaction chamber. The use of the prereactor allows the NO/O3 chemiluminescence to occur out
of view of the PMT, providing for an accurate measurement of background chemiluminescence
resulting from reactions other than the NO/Os reaction (e.g., reactions of hydrocarbons and Os).
The PMT is housed in a thermoelectric (TE) cooler to minimize thermal noise.
As shown in Figure 4-1, separate sample transfer lines downstream of the sample inlet
point are used for the NOy and NO measurement channels, and a third transfer line is used to
deliver calibration and converter efficiency assessment standards from the gas phase titration
(GPT) calibration system to the sample inlet. Because of the remote location of the converter
relative to the analyzer itself, these transfer lines may be of considerable length (i.e., up to 20 m).
The length of the sample transfer line presents no problem in the NOy measurement mode of the
NOy instrument, since all NOy species are converted to NO in the heated converter, and since
that same converter destroys any ozone present in the sample air. However, in the instrument's
NO mode, the ambient air drawn down the sample transfer line contains both ambient NO and
ambient Os. These two species can react [by reactions 4-la and b] to decrease the NO reaching
the chemiluminescence detector, resulting in an under-estimation of the ambient NO level. This
-------
Air
Dryer
Dry Air
Ozone
Generator
Capillary
Exhaust
Bypass
Pump
^ —
H
Filter
(NO Only) |-|
TT
(NO,) n
(Ca
Molybdenum
Converter
1 * 1
NO C Capillary
M NC C
ibration) U Filter
UGas Phase
Titralion
Calibration
System
Sample
In
Weatherproof
Enclosure
(Positioned al
Probe Inlet Height)
Prereactor
Chamber
Pressure
Gauge
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 7 of 40
Exhaust
—*>
Pump
Figure 4-1. General schematic of a typical chemiluminescence NOy instrument.
effect can be substantial: assuming 100 ppbv of 63, loss of 10 percent of the NO can occur
within a few seconds. One way to counteract this effect, is for sample air to be drawn rapidly
through the sample transfer line. For example, at a sample flow rate of 6 L/min, a sampling line
with an inner diameter of 4 mm and a length of 15 m (50 feet) would result in a residence time of
less than 2 seconds. Rapid transport of the sample can best be accomplished using an auxiliary
sampling pump (not shown in Figure 4-1) to draw sample down the transfer line to a "T" fitting
at the back panel of the NOy analyzer. The sample flow to the chemiluminescence detector is
then drawn from that "T" by the analyzer's internal sample pump. If implemented, it is
recommended that this approach be implemented on both the NO and NOy sample lines, to
achieve consistent residence times in the two lines. However, implementing rapid sample
transport through the Mo converter in this way may reduce converter efficiency and/or lifetime.
Consequently, a preferable approach may be to reduce sample transport time in both the NO and
NOy flow paths by reducing the gas pressure.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 8 of 40
4.3 Recommendations for NCore
Since the high sensitivity analyzers deployed at NCore sites are intended to monitor low
ambient NOy concentrations, it is important that they meet a variety of performance criteria as
described below. Many of these performance criteria for high sensitivity NOy analyzers are
more stringent than those for routine NOX analyzers; consequently, there are a number of
recommended features that the NOy analyzers should have in order to achieve the performance
criteria. This section describes the recommended performance criteria and the analyzer features
that are recommended in order to achieve the performance criteria, and provides examples of
commercial high sensitivity NOy analyzers that are available for deployment at the NCore sites.
Additionally, this section discusses some important sampling requirements that should be
considered during the installation of the analyzers.
4.3.1 Recommended Method Performance Criteria
The U.S. EPA has recently assessed the measurement quality objectives needed for high
sensitivity precursor gas monitoring in NCore, relative to the long-established statistics stated in
40 CFR 58.[11] In particular, EPA recommends that measurement quality objectives for bias and
precision be based on upper confidence limits at the monitoring site level, to provide a higher
probability of reaching appropriate conclusions (e.g., in comparisons to NAAQS). The intent of
this recommendation is to move S/L/T agencies to a performance-based quality system i.e.,
allowing organizations that show tight control of precision and bias to reduce the frequency of
certain QC checks, and to focus their quality system efforts where most needed.
The U.S. EPA recommends that the high sensitivity NOy analyzers that are deployed at
NCore sites meet the following method performance criteria. It is to be expected that these
criteria may be more difficult to meet for NOy than for NO.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 9 of 40
4.3.1.1 Precision
Precision is defined as the measure of agreement among individual measurements of the
same property taken under the same conditions. Precision is assessed from checks that are
performed at least once every two weeks (see Sections 4.7.4 and 5.3.3). Calculations to assess
precision are given below and should be used to assess precision on a quarterly basis. It is
recommended that high sensitivity NOy analyzers have a 95 percent probability limit for
precision of ±15 percent or less.
Calculation of precision starts with the comparison of the known challenge concentration
used in the precision checks to the corresponding measured concentrations reported by the
analyzer. The resulting percent differences are then used as described below. For each single
point precision check, calculate the percent difference, df, as follows:
Equation 1
reported - challenge
d,= —-100
challenge
where reported is the concentration indicated by the high sensitivity NOy analyzer and challenge
is the concentration of the standard used in the precision check. The precision estimator is then
calculated as the coefficient of variation (CV) upper bound, using Equation 2 as follows:
Equation 2
where n is the number of data points (i.e., precision check comparisons), the dt values are the
resulting percent differences, and X o.i,n-i is the 10th percentile of a chi-squared distribution with
n-1 degrees of freedom.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 10 of 40
4.3.1.2 Bias
Bias is defined as a systematic or persistent distortion of a measurement process that
causes errors in one direction. Bias is assessed from the degree of agreement between a
measured value and the true, expected, or accepted value. Analyzer bias is calculated using
comparisons of known challenge concentrations to the corresponding measured concentrations
reported by the analyzer. The challenge comparisons used to assess bias should be the same as
those used to assess precision (see Section 4.3.1.1 above). The bias estimator is an upper bound
on the mean absolute value of the percent differences as described in Equation 3 as follows:
Equation 3
AS
\b'as\=AB+t0.95,n-l —j=
v»
where n is the number of challenge comparisons being aggregated; to.95,n-i is the 95th quantile of
a t-distribution with n-1 degrees of freedom; the quantity AB is the mean of the absolute values
of the individual dt 's and is calculated using Equation 4 as follows:
Equation 4
and the quantity AS is the standard deviation of the absolute value of the dt's and is calculated
using Equation 5 as follows:
Equation 5
I II I
1-IKHIKI
^ = 1
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 11 of 40
Since the bias statistic as calculated in Equation 3 uses absolute values, it does not have a
direction or sign (negative or positive) associated with it. The sign of the calculated bias is to be
determined by rank ordering the percent differences of the QC check samples from a given
analyzer for a particular assessment interval. Calculate the 25th and 75th percentiles of the
percent differences for each analyzer. The absolute bias upper bound should be flagged as
positive if both the 25th and 75th percentiles are positive, and as negative if both these percentiles
are negative. The absolute bias upper bound would not be flagged if the 25th and 75th percentiles
are of different signs. It is recommended that high sensitivity NOy analyzers have an upper
bound for the average bias of ± 15 percent or less.
4.3.1.3 Representativeness
Representativeness refers to whether the data collected accurately reflect the conditions
being measured. It is the data quality indicator most difficult to quantify. Unless the samples are
truly representative, the other indicators are meaningless. Representativeness for monitoring of
low ambient levels of NOy in NCore is different than for routine monitoring, since the objectives
of the monitoring are much different. Representativeness can only be assured in terms of the
appropriate selection of the sampling site, proper implementation of ambient air sampling, and
reasonable coverage of the sampling schedule (i.e., continuous).
4.3.1.4 Completeness
Completeness is defined as the amount of data collected relative to the total expected
amount. Ideally, 100 percent of the expected amount of data would always be collected; in
practice, completeness will be less for many reasons, ranging from calibration time and site
relocation to power outages and equipment failure. For monitoring of ambient NOy
concentrations in NCore, EPA requires a minimum data completeness of 75 percent. In practice
typical completeness values can often approach 90 to 95 percent.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 12 of 40
4.3.1.5 Comparability
Comparability is defined as the process of collecting data under conditions that are
consistent with those used for other data sets of the same pollutant. The goal is to ensure that
instruments purchased and operated by different states and local agencies produce comparable
data. To promote comparability, this TAD describes the recommended characteristics of high
sensitivity NOy analyzers and the procedures for their installation and use. For example, all
monitoring agencies should purchase instruments that minimally have the features described in
Section 4.3.2, and should adhere to the sampling requirements described in Section 4.3.3.
4.3.1.6 Method Detection Limit
The MDL refers to the lowest concentration of a substance that can be reliably
determined by a given procedure. The MDL is typically not provided by the vendor. Based on
the objectives of the Precursor Gas Program, it is expected that most sites will be measuring
pollutant concentrations at lower ranges than the typical SLAMS/NAMS network. Therefore,
the ability to quantify concentrations at these lower levels will be very important. The use of a
vendors advertised LDL is sufficient to make intelligent purchasing decisions. Vendors quantify
LDLs under ideal conditions and therefore one might consider this value as the best possible
detection that can be achieved. As these monitors are deployed into monitoring networks, where
both environmental conditions, equipment (calibration, dilution devices, sampling lines, gaseous
standards) and operator activities can vary, it is important to estimate what pollutant
concentrations can truly be detected, above background noise (the potential conditions
mentioned above). The site specific MDL establishes an estimate based on the routine operation
(and conditions) of that instrument in the network and provides a more meaningful evaluation of
data as it is aggregated across the precursor gas network. By establishing site specific MDLs,
values less than the MDL can be flagged which would allow data users a more informed decision
on the use of that data.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 13 of 40
To perform the MDL test, run zero air through the analyzer and establish an acceptable
zero; dilute pollutant gas to the targeted concentration (one to five times the estimated MDL) and
collect 20 to 25 one minute observations. Repeat this seven times over the course of 5 to 14
days. Average the concentration from the 20-25 readings; calculate the standard deviation (S) of
the average readings and compute the MDL. The MDL is then calculated as the standard
deviation of the response values times the Student's t-value for the number of test measurements
(40 CFR Part 136, Appendix B). The MDL for high sensitivity NOy analyzers should be
established prior to putting the analyzers into service, and should be 0.20 ppb or lower over an
averaging time of no more than 5 minutes.
4.3.1.7 Lower Detectable Limit
The LDL is the minimum pollutant concentration that produces a signal of twice the
noise level. To estimate the LDL, zero air is sampled and the noise level of the readings is
determined according to 40 CFR 53.23(b). The vendor-specified LDL for the most sensitive
range of high sensitivity NOy analyzers should be 0.10 ppb or lower, over an averaging time of
no more than 5 minutes.
4.3.1.8 Linear Range
The linear range of each high sensitivity NOy analyzer should extend from approximately
0.10 ppb to at least 200 ppb. Users should determine if their range should exceed 200 ppb and
adjust accordingly. A range of 200 ppb may not be sufficient in all areas and situations. Note
that some high sensitivity NOy analyzers can operate simultaneously on a number of ranges, with
each range recorded on a separate data logger channel with its own calibration curve. Although
requiring slightly more effort to calibrate and maintain, recording of multiple ranges would allow
capture of a wide range of NOy concentrations.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 14 of 40
4.3.1.9 Zero/Span Drift
Zero drift is defined as the change in response to zero pollutant concentration, over
12- and 24-hour periods of continuous unadjusted operation. Span drift is defined as the percent
change in response to an upscale pollutant concentration over a 24-hour period of continuous
unadjusted operation. Zero and span drift should be obtained from the vendor prior to putting a
high sensitivity NOy analyzer into service. Such NOy analyzers should have 12- and 24-hour
zero drift less than 0.10 ppb, and should have a span drift of less than ±1 percent of the full scale
measurement range of the analyzer per 24 hours. Zero tests should be performed with the
internal zero prereactor engaged.
4.3.2 Recommended Features for High Sensitivity Ambient NO¥ Measurements
Continuous chemiluminescence NOy analyzers are commercially available from a
number of vendors. The design of these analyzers is similar among vendors with some slight
variations. A diagram of the typical chemiluminescence analyzer is described in Section 4.2, and
examples of specific instruments are provided below in Section 4.3.3. In general, each of the
analyzers contains the following systems:
• Pneumatic System: This portion of the analyzer consists of a sample inlet
incorporating a heated converter, sample inlet line, paniculate filter, gas phase
titration calibration unit, ozone generator, prereactor, flow meter, and pump, all
used to bring ambient air samples to the analyzer inlet.
• Analytical System: This portion of the analyzer consists of the reaction chamber,
photomultiplier, and bandpass filters.
• Electronic Hardware: This portion of the analyzer consists of the electronic
components that control the analyzer and process the signals. This part of the
analyzer generally requires little or no maintenance. If the instrument is operated
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 15 of 40
outside the manufacturer's recommended temperature range, however, individual
integrated chips can fail and cause problems with operation, data storage, or
retrieval.
In operation of these systems, the following recommendations should be followed with
precursor NOy analyzers to allow them to measure NOy at levels well below 1 ppb.
• Locate the sample inlet at 10 meters to avoid the physical removal of HNO3. The
inlet should face the prevailing wind direction, be as short as possible, and be
constructed of PFA Teflon®. Half of a Teflon® filter holder with the filter
support used as a "bug screen" should have a negligible effect on NOy
measurements, and provides a practical solution to ward off larger insects.
• Locate the site in an area that is not obstructed by nearby trees and obstacles (see
Section 4.3.4.1)
• Ensure that the sample residence time in the NO sample transfer line is less than 2
seconds to address the Os/NO reaction and subsequent loss of NO in the line, and
protect the sample transfer lines from light through the use of opaque conduit
normally provided by the vendor.
• A heated molybdenum converter rather than a heated gold/reactant converter is
recommended, since the latter requires a supply of either a toxic reductant gas (CO)
or a flammable reductant gas (H2), and provides no clear advantage in determining
total NOy in urban and suburban air.[10]
• The temperature of the molybdenum converter should be maintained at 350 °C.
Higher temperatures than recommended may result in converting significant
amounts of non-NOy species such as ammonia, organic amines, or particulate
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 16 of 40
ammonium. If a manufacturer recommends a converter temperature above 350 °C,
he should show evidence that such non-NOy species are not converted. It is equally
important that the converter not be operated below 350 °C to ensure optimal
conversion of NOy species.
• Automatic over-range capabilities are used to track the rapid changes that may
occur in ambient NOy levels. High sensitivity analyzers often have an analog
output range limited to 200 ppb full scale; digital ranges of up to 400 ppb may be
needed to track peak concentrations in urban areas.
It is recommended that the NOy analyzers deployed in NCore include these additional siting and
operational features in order to ensure useful measurements.
4.3.3 Commercial Chemiluminescent NO¥ Monitors
Vendors of commercial NOy analyzers typically supply instruments for both routine
ambient and high sensitivity monitoring. In this case, only analyzers with an LDL of 0.1 ppb or
less for NO are considered to have high sensitivity. (Note that the LDL for NO may be lower
than that for NOy.) Analyzers from Thermo Electron Corporation, Teledyne Advanced Pollution
Instrumentation (API), Ecotech, and ECO PHYSICS are described here as examples (no
endorsement should be inferred), though not all have both the requisite detection limit and NOy
measurement capabilities. A summary of each monitor follows. The descriptions provided for
these instruments, and the performance claimed for them, are based solely on information
provided by the respective vendors, and confirmation by EPA should not be inferred.
4.3.3.1 Thermo Electron Corporation Model 42C-Y
The Model 42C-Y NOy analyzed12'13] shown in Figure 4-2 is based on the Model 42C
NO-NO2-NOX analyzer, and employs an external molybdenum converter that is typically
operated at 325 °C. The Model 42C-Y NOy analyzer has a lower detection limit of 0.05 ppb, and
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 17 of 40
has adjustable range settings from 0 to 5 ppb up to 0 to 200 ppb. The recommended operating
temperature range is between 15 °C and 35 °C. The Model 42C-Y NOy analyzer uses increased
pumping speed to minimize residence time and enhance chemiluminescent detection, and has
increased ozone production capabilities.
Figure 4-2. Thermo Electron Model 42C-Y NOy analyzer (courtesy Thermo Electron).
Table 4-1 shows the vendor specifications of the Model 42C-YNOy analyzer.
Table 4-1. Thermo Electron Model 42C-Y NOy Analyzer Specifications.
Performance
Parameters
Precision
Bias
Method Detection Limit
Lower Detectable Limit
Linearity
Zero Drift
Span Drift
Model 42C-Y NOy
Analyzer Specifications
Not available
Not available
Not available
50 ppt (0.05 ppb) (120 second average time)
± 1% of full scale
Negligible (24-hour)
± 1 % of full scale (24-hour)
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 18 of 40
4.3.3.2 API Model 200AU with Model 501Y Converter
The API Model 200AU high sensitivity analyzer^14 16] shown in Figure 4-3 is based on the
Model 200A NO/NO2/NOX analyzer, and uses the optional Model 501Y molybdenum NOy
converter that places the converter at the sample inlet point.
Figure 4-3. API Model 200AU/501Y NOy analyzer (courtesy Teledyne API).
The Model 200AU has operating ranges from 0 to 5 ppb up to 0 to 2,000 ppb (2 ppm),
with a lower detectable limit of 0.05 ppb, and may be safely operated in the temperature range of
5 °C to 35 °C. Table 4-2 shows the vendor specifications for the Model 200AU/501 NOy
Analyzer.
Table 4-2. API Model 200AU/501 NOy Analyzer Specifications.
Performance Parameters
Precision
Bias
Method Detection Limit
Lower Detectable Limit
Linearity
Zero Drift
Span Drift
API Model 200AU/501 NOy Analyzer Specifications
0.5% of reading
Not available
Not available
50 ppt (0.05 ppb); 120 second averaging time
1% full scale
<0.1 ppb/24 hours, <0.2 ppb/7 days
<0.5% full scale/7 days or 50 ppt/7 days, whichever is
greater
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 19 of 40
4.3.3.3 Ecotech Model EC9841
High sensitivity NOy capabilities are only partly established in the Ecotech line of
analyzers. Although an Ecotech high sensitivity nitrogen oxides analyzer exists, it is not yet
available in an NOy configuration.
The high sensitivity Ecotech analyzer is the Model EC9841T Trace Nitrogen Oxides
Analyzer, which has a LDL of 0.05 ppb, and a maximum full scale reading of 200 ppb.
However, at this time the Model EC9841T is not available in a configuration consistent with the
NOy requirements illustrated in Figure 4-1. On the other hand, Ecotech's Model EC9841-NOy
does incorporate the needed NOy configuration, but has an LDL of 0.5 ppb. On the assumption
that Ecotech may later adapt the high sensitivity EC9841T analyzer for NOy detection, the
capabilities of both the EC9841-T and the EC9841-NOy are summarized below.
The EC9841-NOy analyzer^17'18] shown in Figure 4-4 has a lower detection limit of
500 ppt (0.5 ppb). It uses a programmable temperature-controlled molybdenum (Mo) converter
that is typically operated at 350 °C and is mounted directly onto the system's sample manifold
(external to the analyzer). The sample manifold and inlet lines are heated to ensure there is no
condensation as the sample gas is transported from the converter to the analyzer. The EC9841-
NOy analyzer also uses a high-vacuum pump to minimize residence time, and to decrease
reaction cell pressure and thus increase the sensitivity of the chemiluminescent detection.
Interferences are minimized in the EC9841-NOy through the use in the converter of Mo chips
that have a greater gas contact area than solid Mo granules to provide greater converter
efficiency with low flow restriction at atmospheric pressure. Under typical operating conditions,
the converter has less than a 5 percent conversion efficiency for ammonia, cyanides, and amines,
thereby limiting interference from these non-NOy species. The analyzer has automatically
selected range settings up to 20 ppm full scale, and has a recommended operating temperature
range of 20 °C to 30 °C, but may be operated between 15 °C and 35 °C. The Model EC9841T
differs from the EC9841-NOy primarily in its higher sensitivity and lower zero and span drift.
Table 4-3 shows the vendor specifications of both analyzers.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 20 of 40
Figure 4-4. Ecotech EC9841 -NOy analyzer (courtesy of Ecotech).
Table 4-3. Ecotech EC9841 -NOy Analyzer Specifications.
Performance
Parameters
Precision
Bias
Method Detection
Limit
Lower Detectable
Limit
Linearity
Zero Drift
Span Drift
EC9841-NOy
± 2% of reading
Not available
Not available
500 ppt (0.5 ppb) (with Kalman or
300 sec filter active)
Not available
Temperature dependence, 0.1% per
degree C changes.
24 hours; less than 1 ppb
Temperature dependence, 0.1 % per
degree C changes.
24 hours less than 1 .0% of reading
30 days less than 1 .0% of reading
EC9841T
±2% of reading
Not available
Not available
50 ppt (0.05 ppb) (with Kalman or
300 sec filter active)
Not available
Temperature dependence, 0.1% per
degree C changes.
24 hours; less than 50 ppt (0.05 ppb)
Temperature dependence, 0.05 % per
degree C changes.
24 hours less than 0.5% of reading
30 days less than 1 .0% of reading
4.3.3.4 ECO PHYSICS Models CLD 88 p or CLD 780 TR with CON 765 NOy Converter
ECO PHYSICS produces two different chemiluminescent analyzers with the requisite
high sensitivity for trace level measurements.[19] The ECO PHYSICS Model CLD 88 p analyzer
has a lower detection limit of 50 ppt (0.05 ppb), and the CLD 780 TR is a research grade NO
instrument capable of a lower detection limit of 10 ppt (0.01 ppb) with a 60 second integration
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 21 of 40
time. The CLD 88 p has four selectable operating ranges from 5 to 5,000 ppb (5 ppm) and may
be safely operated in the temperature range of 5 °C to 40 °C. The CLD 780 TR has five
selectable operating ranges from 5 to 500 ppb, and may be safely operated in the temperature
range of 5 °C to 50 °C. However, at this time these analyzers are not available in a configuration
consistent with the NOy requirements illustrated in Figure 4-1. On the assumption that the
manufacturer may adapt one or both analyzers to the required configuration, the capabilities of
both the CLD 88 p and the CLD 780 TR are summarized below.
The limitation of the ECO PHYSICS analyzers lies in the ECO PHYSICS CON 765 NOy
converter, which is designed to be used with either of the CLD 88 p or CLD 780 TR NO
analyzers.[19] The CON 765 reduces NOy species to NO by chemical reaction with CO on a gold
surface maintained at 315 °C. This converter has a reported NOy conversion efficiency that
exceeds 95%, and due to the nature of the reduction process the presence of water vapor in the
sample air is said to keep interference from NH? and HCN to a negligible level. Despite these
attractive features, the CON 765 is not the type of NOy converter recommended for use in
precursor gas monitoring in NCore. The need for a constant supply of highly toxic high purity
CO is a serious disadvantage of this converter, especially since the gold converter offers
essentially the same NOy conversion efficiency as the molybdenum converter recommended for
use in NCore.
Figure 4-5 shows the Model CLD 88 p with the CON 765 converter. Table 4-4 shows the
information available on the specifications of both the CLD 88 p and CLD 780 TRNO
analyzers.
-------
:J i
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 22 of 40
Figure 4-5. ECO PHYSICS Model CLD 88 p and CON 765 NOy Converter
Table 4-4. ECO PHYSICS CLD 88 p and CLD 780 TR Analyzer Specifications.
Performance
Parameters
Precision
Bias
Method Detection Limit
Lower Detectable Limit
Linearity
Zero Drift
Span Drift
CLD 88 p
Not available
Not available
Not available
50 ppt (0.05 ppb)
1% of full scale
0.05% of full scale (24 hours)
1% of full scale (24 hours)
CLD 780 TR
Not available
Not available
Not available
10 ppt (0.01 ppb) w. 60 sec response
1% of full scale
None (i.e., zero)
Not available
4.3.4 Sampling Requirements
Proper siting of the sampling equipment and sampling probes is necessary to ensure that
the gas analyzers are obtaining representative samples of the ambient air. Likewise, proper
environmental control of the analyzer and proper sampling are critical to ensuring that the
analyzers are operating correctly and that the NOy measurements are comparable to NOy
measurements recorded at other sites.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 23 of 40
4.3.4.1 A nalyzer Siting
Analyzer siting should follow the criteria in 40 CFR 58, Apendix E. The installation of
the precursor NOy analyzer should allow for the sample probe inlet to be approximately 10
meters above ground level with the inlet facing the prevailing wind direction. The probe should
be positioned with at least 270 degrees of unrestricted airflow including the predominant wind
direction. The probe should be separated from the drip line of nearby trees or structures by at
least 20 meters, and should be positioned at least twice as far from of nearby obstacles as the
height of the obstacles.
4.3.4.2 Instrument Shelter
To help ensure proper performance, the precursor analyzers and supporting equipment
should be installed and operated in a temperature controlled environment. An insulated
instrument shelter should be used to protect the analyzer from precipitation and adverse weather
conditions, maintain operating temperature within the analyzer's temperature range
requirements, and provide security and electrical power. The environmental control of the
shelter should be sufficient to minimize fluctuations in shelter temperature. The recommended
shelter temperature range is 20 °C to 30 °C, and daily changes in temperature should not exceed
5 °C over a 24-hour period. Condensation of moisture must be avoided and it may be necessary
to impose seasonal temperature ranges to assure remaining above the ambient dewpoint.
4.3.4.3 Sampling Issues
Studies of NOy sampling inlet issues have focused primarily on airborne NOy
measurements, where it is not feasible to locate the converter directly at the sample inlet
point. [20"22] The purpose of these studies was to identify the material that causes the least
adsorptive loss of NOy components during sampling. Nitric acid, as both a key component of
NOy and a strongly adsorbed species, has generally been the target compound in these studies.
Adsorption of NO and NO2 is of much less concern. Numerous tubing materials, including TFE,
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 24 of 40
PFA, and FEP Teflon®, have been investigated for use in sampling inlets.[20] In testing these
materials for HNOs adsorption, less than 5 percent of the HNOs was lost with Teflon® tubing,
while greater than 70 percent was lost with tubing made of stainless steel, glass, fused silica,
aluminum, nylon, silica-steel, and silane-coated steel. HNOs transmission through aluminum,
steel, and nylon tubes did not increase in over 1 hour of HNO3 exposure. HNO3 loss on
aluminum and steel tubes heated to 50 °C was irreversible. However, HNOs adsorption on glass
decreased over time, so that over a period of several hours of continuous HNOs exposure, glass
will be passivated to HNOs adsorption. Furthermore, heated glass tubing passivates faster than
room temperature tubing, and larger diameter glass tubing takes longer than smaller diameter
glass tubing to passivate with HNOs. PFA Teflon® causes the least adsorption of HNOs, and so
is recommended for sampling inlets.
Although PFA Teflon® is far superior to other materials in minimizing HNOs adsorption,
it has the disadvantage that any previously adsorbed HNOs can be released back into the air
stream by changes in temperature and/or relative humidity.[20] Thus, the best approach to NOy
monitoring is to expose the incoming sample air to as little surface area as possible upstream of
the heated converter. Therefore, the best approach is to minimize the length of PFA tubing at the
inlet.
The inlet system must also be configured to allow calibration through the same inlet
plumbing used in monitoring. As shown in Figure 4-1, this is easily accomplished by means of a
PFA cross fitting on the inlet of the converter, with one arm of the cross connected to the GPT
calibration system.
4.4 Potential Problems and Solutions
In addition to the potential problems with sampling described above, there are other
potential problems with the high sensitivity measurement of NOy in ambient air.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 25 of 40
4.4.1 Interferences
Interferences in NOy measurements are of two types. One potential interference is the
presence of nitrogen-containing species in ambient air that are not components of NOy, but that
can potentially be converted to NO by the heated converters used to achieve NOy measurement.
The primary examples of such an interferent are ammonia (NH3), and particulate ammonium
(NH4+), but other amines and even cyanide compounds (e.g., hydrogen cyanide, HCN) could be
present. This type of interference is addressed in the discussion of converter efficiency in
Section 4.4.2.
The other type of potential interferent consists of non-NOy species that can react with Os
to produce chemiluminescence in the relevant wavelength region, thereby artificially increasing
the apparent signal from NO in the sample air. The most important such interferents in ambient
air are unsaturated hydrocarbons (e.g., ethylene, propylene, and naturally emitted species such as
terpenes). Interference from such compounds in ambient NOy monitoring is minimized by the
use of a prereactor vessel in the NOy monitor (see Figure 4-1). The prereactor is a part of the
normal flow path of ozone to the reaction chamber in the monitor. When the sample air flow is
diverted into the prereactor, the NO/Os reaction occurs rapidly and the resulting NO2
chemiluminescent emission occurs entirely within the prereactor, where it cannot be detected by
the photomultiplier. However, the Os reactions with unsaturated hydrocarbons occur more
slowly, so light emission from these reactions is not completed within the prereactor volume. As
the sample/Os mixture flows from the prereactor into the reaction chamber, the photomultiplier
detects the background chemiluminescence from the hydrocarbon interferents, without emission
from the NO/Os reaction. Commercial high sensitivity NOy analyzers typically determine their
background readings automatically using this prereactor mode and, thus, this type of interference
is automatically accounted for by the analyzer software through subtraction of the background
readings.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 26 of 40
4.4.2 Converter Efficiency
4.4.2.1 Overview
The heated molybdenum converters used in commercial high sensitivity NOy analyzers
have undergone extensive testing and intercomparison in both laboratory and field studies to
confirm the wide variety of species that can be converted to NO and measured as part of the NOy
total.[eg'10'23] These studies indicate that the molybdenum converters can provide accurate
measurements of NOy. The goal with such converters is to achieve 100 percent conversion
efficiency of NOy species to NO, while approaching zero percent conversion of other non-NOy
nitrogen-containing species. Note that, as used in commercial high sensitivity NOy monitors, the
molybdenum converters are designed to convert particulate nitrate compounds, as well as the
numerous gaseous components of NOy, to NO for detect!on.[10]
As noted in Section 4.4.1, non-NOy species such as ammonia, particulate ammonium, or
hydrogen cyanide can also be oxidized to NO, although this conversion can be minimized (to a
few percent conversion or less) by maintaining the converter temperature at 350 °C. At sub-ppb
NOy concentrations, interference from such compounds can be substantial, and even at higher
levels the potential for interferences must be kept in mind. Sampling near a large source of
ammonia, for example, could produce erroneously high NOy readings, even though the
conversion efficiency for ammonia is much less than that for NO2 or the various NOZ
compounds. In general, the efficiency of a converter system at sub-ppb levels may vary
depending on the mix of NOy species present, the age and condition of the converter, the
converter temperature, ambient humidity, or ozone levels and cannot be entirely predicted even
from the behavior of a similar system. For these reasons, converter efficiency must be evaluated.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 27 of 40
4.4.2.2 Challenge Species for Converter Efficiency Checks
Studies of converter efficiencies have established that among NOy species, NO2 is
relatively easily reduced to NO. As a result calibration with NO and NO2, as described in
Section 4.7.2, is a necessary but not entirely sufficient approach to characterizing an NOy
monitoring system. A more stringent approach is to also calibrate with an NOy species that is
both more difficult to convert to NO and relatively easy to prepare in known concentrations. The
most common choice for such an additional compound to determine NOy converter efficiency is
n-propyl nitrate (NPN). This organic nitrate is used in the form of compressed gas standards that
are readily diluted to near-ambient NOy levels. Diluted NPN mixtures (Scott-Marrin, Inc.,
Riverside, CA; www.scottmarrin.com) are supplied to the monitor through the calibration line to
the monitor's inlet (Figure 4-1), and provide a more challenging test of the conversion efficiency
than testing with NO2 alone.
Conversion efficiency testing with NPN is in addition to, not in place of, routine
calibration with NO and NO2. Changes in pollutant levels and meteorological conditions over
time can significantly alter the instrument's conversion efficiency. Thus, NOy monitoring
requires routine NPN converter efficiency checks and consistent procedures to maintain or repair
the converter when its efficiency falls below acceptable levels. A single-point conversion
efficiency check with NPN is recommended every month in continuous NOy monitoring. An
NPN conversion efficiency of 95 percent or greater is considered acceptable for NOy monitoring,
converters falling below 95 percent efficiency should be replaced. Note that a new converter
should be allowed to "burn in" over one to three days of use before performance of an NPN test.
Also, the NPN standard may not be certified to better than ±5%, so it is recommended to track
conversion over time and use 95% of the original efficiency as the performance cutoff.
The most rigorous way to perform NO or NO2 calibrations and NPN conversion
efficiency checks would be by standard addition (i.e., by adding known concentrations of these
species to the ambient air sampled by the monitor). This approach would provide the most
accurate and realistic calibration because it includes the effects of water vapor and other
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 28 of 40
constituents in the sample air. However, this procedure may not fit well into the routine data
acquisition used at many monitoring sites. Often, data acquisition is keyed to recording one-hour
average data. Over that period, the ambient air background NO and NOy levels are not likely to
remain constant enough for an accurate standard addition calibration or efficiency check. As a
result, it is recommended to perform calibrations and conversion efficiency checks by diluting
appropriate standards with high purity air using a MFC dilution system. Personnel at NOy
monitoring sites are advised to consider the calibration approaches that fit best within their
programs.
Although HNOs is a key component of NOZ and in turn of NOy, and is known to be
especially difficult to sample, it is not advisable to attempt routine calibration checks with HNOs
in the field. The complexities of maintaining an HNOs source and delivering accurate HNOs
levels to the sample inlet outweigh the potential benefits. The best way to assure adequate
sampling of HNOs and other NOy species is to use a properly configured NOy monitor, as
described in Section 4.3.2 and 4.3.4. However, an annual or more frequent challenge of the
monitor with multiple compounds may be a valuable test of instrument performance. If
performed, such a challenge should involve several different tests, i.e., calibrations with NO and
NO2, a converter efficiency check with NPN, and perhaps a test of the conversion efficiency for
NH3 (the most likely gas-phase interferent) using a certified permeation source of NH3. An NPN
conversion efficiency of at least 95 percent, and a simultaneous ammonia conversion efficiency
of at most 5 percent, should be the target performance criteria for such a challenge.
4.5 Equipment and Supplies
4.5.1 Data Acquisition Device
Many types of equipment can be used to record the concentration measurements obtained
from the analyzer. Recommended options for data acquisition are described in Section 6 of this
TAD.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 29 of 40
4.5.2 Calibration Equipment
The equipment required for calibration of a precursor NOy analyzer include a MFC
calibrator unit as described in Section 2.5.2, with gas phase titration capability, and a source of
zero air. The following equipment is recommended for calibration of a high sensitivity NOy
analyzer.
4.5.2.1 Calibration Standard and Standard Delivery System
The calibration standards used for the calibration of precursor NOy analyzers should be
generated by dilution of a commercially-prepared and certified compressed gas NO standard
using a MFC calibration unit. It is important when purchasing a MFC calibrator that it meet the
40 CFR 50 requirements of ±2 percent accuracy, and that the flow rates of both MFC channels
be calibrated using a NIST-traceable flow standard.
When the analyte concentration in the commercially-prepared standard cylinder is vendor-
certified by reference to NIST standards, and the MFCs are calibrated to NIST-traceable
standards, the resulting working gas concentration is considered to be NIST-traceable.
4.5.2.2 Zero Air Source/Generator
Zero air is required for the calibration of high sensitivity NOy instruments. This air must
contain no detectable NOy species (i.e., NOy content must be less than the LDL of the analyzer)
and must be free of paniculate matter. Suitable zero air may be supplied from compressed gas
cylinders, with additional scrubbing by passage through a soda lime trap, sodium carbonate trap,
or carbonate coated denuder. However, it is likely too expensive and impractical to maintain a
sufficient supply of zero air cylinders to operate a high sensitivity NOy analyzer continuously.
As an alternative, many commercially-available zero-air generation systems can supply suitably
NOy-free air, provided additional external scrubbing is provided as noted above.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 30 of 40
4.6 Reagents and Standards
Routine operation of precursor NOy analyzers requires the use of calibration standards
and zero air to conduct periodic calibrations and instrument checks. This section describes the
requirements for these gases.
4.6.1 Calibration Standards
The primary NO standards must be certified commercially-prepared compressed gas
standards of NO in N2, with a certified accuracy of no worse than ±2 percent. NO gas standards
of 5 to 20 ppm (with less than 1 ppm NO2) are conveniently diluted with a MFC calibrator down
to working concentrations of 10 ppb or less. The commercially-prepared standard may contain
only NO, or may be a mixed component standard that also contains known concentrations of
other non-reactive precursor gases (e.g., CO, 802). This standard must be traceable to a NIST
NO in N2 Standard Reference Material (SRM 1683 or SRM 1684), NIST NO2 Standard
Reference Material (SRM 1629), or a NIST/EPA-approved commercially available Certified
Reference Material (CRM). Section 2.0.7 of EPA's Quality Assurance Handbook[24] gives a
recommended protocol for certifying NO gas cylinders against either a NO SRM or CRM.
Procedures for certifying a NO gas cylinder against a NIST NO2 SRM and for determining the
amount of NO2 impurity in a NO cylinder are presented by Ellis.[25] Commercial gas standards
for NO2 and NPN should be obtained with a certified accuracy no worse than ±2 percent, and ±5
percent, respectively.
Every gas standard used in precursor gas monitoring must be accompanied by a
certificate of calibration from the vendor stating the type of traceability, concentration of the
standard, the uncertainty of that certification, and the expiration date of the certification.
Standards traceable to NIST are preferred. Certification documents for all standards must be
retained in a common location and reviewed periodically so that standards for which the
vendor's certification has expired may be removed from service and replaced.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 31 of 40
4.6.2 Zero Air
Zero air used as dilution gas for calibration purposes should have an NOy concentration
below the LDL of the NOy monitor. Commercial grades of zero air may be suitable as a starting
point, provided additional cleanup is employed as noted in Section 4.5.2.2. Commercial zero air
further scrubbed of NOy may be used to crosscheck the purity of air provided by a commercial
continuous air purification system, or a rapid check of the purity of a zero air source can be made
by intercomparison of zero air readings when sampled directly vs. through the prereactor mode
of the NOy analyzer.
4.7 Quality Control
4.7.1 Site Visit Checklists and Remote Diagnostic Checks
To determine whether the high sensitivity NOy analyzer is working properly, field
operators conduct many routine checks of instrument diagnostics and performance every time
they visit the monitoring station. Each agency needs to develop maintenance checklists or
electronic spreadsheets to document that all required checks have been made. The lists and
sheets should be useful both for collecting data and for assessing the quality of that data.
Management must review them regularly and change them if necessary. To the extent possible,
diagnostic checks can be done remotely, provided the data acquisition system allows remote
access to instrument diagnostic information (see Section 6).
4.7.2 Multipoint Calibrations
Calibration procedures for high sensitivity NOy analyzers are more complicated than for
other high sensitivity precursor gas analyzers (i.e., for CO and SO2), in that they include
calibration with NO and NO2,[9] as well as checks of the converter efficiency for NOZ species and
potential interferents. A basic requirement is for a multipoint NO calibration that includes a
minimum of four points (three spaced over the expected range and a zero point), generated by
the calibration system. Although more points may be preferable, current high sensitivity
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 32 of 40
analyzers typically provide linear response over their entire operating range; therefore, four
points should be sufficient. Multipoint calibrations must be done prior to the high sensitivity
NOy analyzer being put into service and at least quarterly thereafter. An analyzer should be
calibrated (or recalibrated) if any of the following conditions occur:
• Upon initial installation;
• The Level 1 span check or precision check difference exceeds 15 percent;
• After repairs or service is conducted that may affect the calibration;
• Following physical relocation or an interruption in operation of more than a few days;
• Upon any indication that the analyzer has malfunctioned or a there has been a change
in calibration; or
• The measured concentration values during challenges with performance test samples
(Section 5.4.1) differ from the certified standard values by ±15 percent.
The analyzers should be calibrated in-situ without disturbing the normal sampling inlet system to
the degree possible.
A second requirement is for multipoint calibration with NO2, as a check of the conversion
efficiency of the molybdenum converter for NO2. This calibration is conducted by gas phase
titration of NO with Os. MFC calibration systems in common use at ambient monitoring sites
have GPT capability. The multipoint NO2 calibration should be done at approximately the same
three concentration levels as the NO calibration noted above. The major equipment/components
required for the GPT NO2 calibration are: a stable Os generator, a data acquisition and display
device, and the NO concentration standard used for the multipoint NO calibration. The principle
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 3 3 of 40
of this calibration technique is the rapid gas phase reaction of NO with Os to produce equal
stoichiometric quantities of NC>2 in accordance with the following equation:
This is the same overall reaction detailed in reactions (4-la and b) and (4-2a and b) above. For
calibration purposes, ozone is added to a stable and excess concentration of NO in a dynamic
calibration system, and the NO reading of the chemiluminescence NOy instrument is used as an
indicator of changes in NO concentration. The NO standard is diluted sufficiently to produce an
upscale NO reading on the measurement range of interest, and upon addition of O3 the decrease
in NO reading observed is equivalent to the concentration of NO2 produced. The amount of NO2
generated may be varied by adding variable amounts of Os from a stable Os generator, which is a
component of the GPT system of the calibrator. Comparison of the NO and NOy responses of
the analyzer then allows determination of the ratio of NO2 response to NO2 generated, which
indicates the converter efficiency for NO2. Maintenance or replacement of the converter should
be undertaken whenever the NO2 conversion efficiency falls below 96 percent.
4.7.3 Level 1 Zero/Span Checks
Level 1 zero and span calibrations are simplified, two-point calibrations used when
adjustments may be made to the analyzer. When no adjustments are made to the analyzer, the
Level 1 calibration may also be called a zero/span "check" and must not be confused with a level
2 zero and span check. Level 1 zero and span checks should be conducted nightly if the
calibration system and NOy analyzers used can be programmed to automatically perform these.
They are used to assess if the analyzers are operating properly and to assess if any drift in
instrument response has occurred. The level 1 span check should not exceed ±15 percent. They
are conducted by challenging the analyzer with zero air and a test atmosphere containing NOy at
a concentration of between 70 percent and 90 percent of the full measurement range in which the
analyzer is operating. The challenge gas should be sampled through as much of the sampling
inlet system as practical to mimic the actual sampling of ambient air. The results of the Level 1
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 34 of 40
zero/span check should be plotted on control charts to graphically illustrate the trends in the
response of the analyzer to the challenge gases. If the measured concentrations fall outside of
the control limits, the accuracy of the MFC calibration system should be checked with a NIST-
traceable flow standard. If the MFC flow accuracy is confirmed, the data recorded since the last
successful Level 1 check should be flagged and the analyzer should be recalibrated using the
multipoint calibration procedures described in Section 4.7.2.
State-of-the-art calibration equipment now exists that is fully automated. These "new
generation" calibration units are fully integrated with computers, mass flow calibrators, and the
associated hardware and software where they can create test atmospheres manually or
automatically. For the precursor gas program, it is recommended that the NCore sites have fully
automated calibration capability. Below are a number of reasons why this is advantageous:
• By performing the calibrations or checks automatically, agencies no longer spend the
manpower needed to perform them.
• Automated calibrations or checks can be triggered internally or by a DAS. Since newer
DASs allow remote access, this allows a remote user to challenge the analyzers without
actually being present.
• High sensitivity precursor gas analyzers are expected to have more zero and span drift than
less sensitive analyzers; therefore, it is important that a zero and Level I check be performed
daily.
• New generation DASs can record calibration or check data and allow remote users to track
daily Level I check and zero drift. This is important for data validation, verification and
troubleshooting.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 3 5 of 40
4.7.4 Precision Checks
At least once every two weeks a precision check should be conducted by challenging the
NOy analyzer with a known low NO concentration to assess the performance of the analyzer.
The precision checks should be conducted by challenging the precursor NOy analyzer with a
calibration mixture of a known NO concentration near 20 ppb. After completion of the precision
check, the operator should calculate the percent difference between the measured value and the
standard value. Precision should be calculated quarterly, using the calculated percent differences
from the precision checks, according to the equations provided in Section 4.3.1.1. For
acceptable precision to be maintained, it is recommended that the calibration system's gas flows
be verified frequently against a NIST flow standard, and adjusted if necessary before making any
adjustments to the analyzer.
4.8 Preventive Maintenance and Troubleshooting
Long-term operation of continuous gas analyzers requires a preventive maintenance
program to avoid instrument down-time and data loss. Despite active preventive maintenance,
occasional problems may arise with the high sensitivity NOy analyzers. This section briefly
describes several key items that might be included in the preventive maintenance program
established for high sensitivity NOy analyzers deployed atNCore sites, as well as some of the
troubleshooting activities that may be useful in resolving unexpected problems with these
analyzers. This discussion is not meant to be exhaustive or comprehensive in detail. More
thorough discussions should be included in the analyzer operation manuals and SOPs developed
for these analyzers. Example SOP's prepared by EPA are included as Appendix B of this TAD.
4.8.1 Preventive Maintenance
Routine preventive maintenance procedures should be in place to prevent down-time and
data loss. Management and field operators should jointly develop their preventive maintenance
program. A program designed by persons unfamiliar with analyzer operations may include
unnecessary items or omit mandatory ones.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 36 of 40
NOy values can be erroneous if the sample inlet and lines become dirty, cracked, or leaky.
PFA lines should be inspected at least quarterly and replaced as needed, but at least every two
years. Teflon® filters used in the sampling train to remove fine particles may need to be
replaced as often as every week, depending on the condition of the filter and the particulate
loading around the monitoring site. The NOy inlet should be inspected every time the NOy filter
is changed.
Table 4-5 illustrates items that monitoring agencies should include in their preventive
maintenance program for high sensitivity NOy monitoring.
Table 4-5. Example of a preventive maintenance schedule for NOy monitoring.
Item
Maintain air dryer
Replace particle filter
Perform pneumatic system leak check
Inspect internal, external tubing; replace if necessary
Clean optical bench
Replace PMT
Monitor NO2 conversion efficiency
Monitor NPN conversion efficiency
Schedule
As needed
Weekly
At least quarterly
Inlet, weekly; other,
quarterly
As needed
As needed
At least every 6 months
At least every month
4.8.2 Troubleshooting
Table 4-6 summarizes common problems seen with precursor NOy analyzers, possible
causes, and possible solutions. More specific information can be found in the manufacturer's
operations manuals. When troubleshooting, an operator must constantly be aware of
environmental factors that may affect the instruments. Environmental factors can also cause
sporadic problems that can be difficult to diagnose. Examples of factors that may affect the
performance of the precursor NOy analyzers are:
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 37 of 40
• Variable shelter temperature (fluctuations greater than several degrees)
• Excessive vibration from other equipment
• Voltage instability; fluctuations in the 110 VAC line voltage
• Air conditioning system blowing on the instrument
• Frequent opening of the door of the shelter.
• Leaks
Table 4-6. Instrument troubleshooting for high sensitivity NOy analyzers
Problem
Noisy output
High positive zero drift
High Prereactorzero reading
No response to span gas
Low or declining response to
span gas
Zero output at ambient levels
Low NO2 or NPN efficiency
No flow through analyzer
Possible Cause
Defective DC power supply
Dirty reaction cell
PMT failure
Defective bandpass filter
PMT failure
Moisture in PMT housing
PMT failure
Voltage failure
No O3 supply
O3 source failing
Dirty reaction cell window
Pump failure
PMT failure
Aging or dirty converter
Pump failure
Possible Solution
Replace power supply
Clean cell
Replace PMT
Replace filter
Replace PMT
Allow PMT housing to warm up;
purge with dry gas, reassemble
Replace PMT
Replace high voltage source
Clean or replace O3 generator
Clean or replace O3 generator
Clean window
Check pump
Replace PMT
Replace converter
Replace/ rebuild pump head
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 3 8 of 40
4.9 References
1. U.S. Environmental Protection Agency. Code of Federal Regulations. Title 40, Part 58,
Subpart E. "Enhanced Os Monitoring Regulations." Office of the Federal Register,
February, 1993.
2. Parrish, D.D., et al., "The Total Reactive Oxidized Nitrogen Levels and the Partitioning
Between the Individual Species at Six Rural Sites in Eastern North America", J.
Geophys. Research, 98; 2927-2939. 1993.
3. Trainer, M., et al., "Correlation of Ozone with NOy in Photochemically Aged Air", J.
Geophys. Research, 98; 2917-2925. 1993.
4. Kleinman, L.I., et al., "Photochemical Age Determinations in the Phoenix Metropolitan
Area", J. Geophys. Research, 108(D3); 4096. 2003.
5. Arnold, J.R., R.L. Dennis, and G.S. Tonnesen, "Diagnostic Evaluation of Numerical Air
Quality Models with Specialized Ambient Observations: Testing the Community
Multiscale Air Quality Modeling System (CMAQ) at Selected SOS 95 Ground Sites",
Aimos. Environ., 37; 1185-1198. 2003.
6. Li, Q., D.L. Jacob, J.W. Munger, R. M. Yantosca, and D.D. Parrish, "Export of NOy
from the North American Boundary Layer: Reconciling Aircraft Observations and Global
Model Budgets",/. Geophys. Research, 109(D2); D02313. 2004.
7. U.S. Environmental Protection Agency. Code of Federal Regulations. Title 40, Part 53,
Office of the Federal Register, July 1, 1987.
8. Fontijn, A., A.J. Sabadell, and R.J. Ronco, "Homogeneous Chemiluminescent
Measurement of Nitric Oxide with Ozone", Anal. Chem., 42; 575-579. 1970.
9. 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
Federal Register, December 1, 1976.
10. Williams, E. J., et al., "Intercomparison of Ground-Based NOy Measurement
Techniques." J. Geophys. Research, 103:22; 261-280. 1998.
11. Rhodes, R.C., Guidelines on the Meaning and Use of Precision and Accuracy Data
Required by 40CFR Part 58, Appendices A and B, EPA60014-83-023, U.S.
Environmental Protection Agency, Research Triangle Park, NC, June 1983.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 3 9 of 40
12. Thermo Electron Corporation (2004). "Product Specifications, Model 42C-Y NOy
Analyzer." Accessed December, 2004. Available at http://www.thermo.com/eThermo/
CM A/PDF s/Product/productPDF 24 5 3 3. pdf.
13. Personal Communication with Thermo representative, Michael Nemergut.
November 2004.
14. Teledyne Advanced Pollution Instrumentation (2004). "M200AU NOy Converter
Option." Manual Addendum.
15 Teledyne Advanced Pollution Instrumentation (1999). "Model 200AU Nitrogen Oxides
Analyzer (1999)." Instruction Manual.
16. Teledyne Advanced Pollution Instrumentation. "Model 200AU Ultra Sensitivity
NO/NO2/NOx Analyzer Data Sheet." Accessed November 2004. Available at
http://www.teledyne-api.com/products/model_200au.asp.
17. Ecotech Pty Ltd. (2004). "EC9841-NOy Analyzer Product Specification Sheet."
Accessed December, 2004. Available at http://www.ecotech.com.au/brochures new/
EC9841A-NOv.pdf.
18. Personal Communication with Ecotech representative, Andy Tolley. September 2004.
19. Information on ECO PHYSICS CLD 88 p and CLD 780 TR NO analyzers, and on CON
765 NOy converter, provided by ECO PHYSICS AG, Duernten, Switzerland and Ann
Arbor, Michigan, at www.ecophysics.com and www.ecophysics-us.com.
20. Neuman, J.A., L.G. Huey, T.B. Ryerson, and D.W. Fahey. "Study of Inlet Materials for
Sampling Atmospheric Nitric Acid." Environ. Sci. TechnoL, 33:7; 1,133-1,136. 1999.
21. Ryerson, T.B., L.G. Huey, K. Knapp, J.A. Neuman, D.D. Parrish, D.T. Sueper, and
F.C. Fehsenfeld. "Design and Initial Characterization of an Inlet for Gas-Phase NOy
Measurements from Aircraft." J. Geophys. Research, 104; 5,483-5,492. 1999.
22. Kondo, Y., S. Kawakami, M. Koike, D.W. Fahey, H. Nakajima, Y. Zhao, N. Toriyama,
M. Kanada, G.W. Sachse, and G.L. Gregory. "Performance of an Aircraft Instrument for
the Measurement of NOy." J. Geophys. Research, 102; 28,663-28,671. 1997.
23. 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; 10,759-10,776. 1997.
-------
Precursor Gas TAD
Section 4 NOy Measurements
Version 4
September 2005
Page 40 of 40
24. 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.
25. 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.
-------
Precursor Gas TAD
Section 5 QA/QC
Version 4
September 2005
Page 1 of 10
5.0 QUALITY ASSURANCE/QUALITY CONTROL (QA/QC) PROCEDURES
5.1 Introduction
One of the primary objectives of the NAAMS is to accurately measure CO, SO2, and NOy
at low concentrations.[1] Therefore, a rigorous quality assurance (QA) program, with appropriate
quality control (QC) activities, must be maintained to ensure that the precursor monitoring data
collected in the NCore network are of suitable quality to meet the objectives of the program.
Quality assurance is defined by EPA as "an integrated system of management activities
involving planning, implementation, assessment reporting and quality improvement to ensure
that a process, item or service is of the type and quality needed and the client expects."[2] QA is
an overall process, described in a management plan, to guarantee the integrity of the data, and
QA activities ensure that the process is appropriately defined and implemented. Quality control
(QC) is defined as "the overall system of technical activities that measures the attributes and
performance of a process, item or service against defined standards to verify that they meet the
stated requirements established by the customer."[2] QC activities are a series of analytical
measurements taken throughout the air monitoring process that are used to assess the quality of
the monitoring data. This section of the NCore TAD describes the QA/QC activities that should
be implemented for precursor gas monitoring in NCore.
5.2 QA/QC Management
The overall QA/QC program should be defined in a Quality Assurance Project Plan
(QAPP). Each State or Local monitoring agency should develop a QAPP in accordance with the
guidance document prepared by EPA.[3] The primary purpose of the QAPP is to "provide an
overview of the project, describe the need for the measurements, and define QA/QC activities to
be applied to the project, all within a single document. The QAPP should be detailed enough to
provide a clear description of every aspect of the project and include information for every
member of the project staff, including samplers, lab staff, and data reviewers."[4]
-------
Precursor Gas TAD
Section 5 QA/QC
Version 4
September 2005
Page 2 of 10
5.3 Network Calibration and Instrument Check Procedures
The following sections describe the instrument calibration and routine instrument check
procedures for precursor analyzers that should be implemented within NCore. In general, these
procedures may be followed for all the precursor pollutant monitors (i.e., CO, 862, NOy) unless
noted.
5.3.1 Multipoint Calibrations
A multipoint calibration includes a minimum of four points (three spaced over the
expected range and a zero point), generated by the calibration system. Although more points
may be preferable, current high sensitivity analyzers typically provide linear response over the
entire operating range; therefore, four points should be sufficient. Multipoint calibrations must
be done prior to the high sensitivity analyzer being put into service and should be repeated at
least quarterly thereafter. An analyzer should be calibrated (or recalibrated) if any of the
following conditions occur:
• Upon initial installation;
• The Level 1 span check or precision check difference exceeds 15 percent;
• After repairs or service is conducted that may affect the calibration;
• Following physical relocation or an interruption in operation of more than a few days;
• Upon any indication that the analyzer has malfunctioned or a there has been a change
in calibration; or
• The measured concentration values during challenges with performance test samples
(Section 5.4.1) differ from the certified standard values by ±15 percent.
-------
Precursor Gas TAD
Section 5 QA/QC
Version 4
September 2005
Page 3 of 10
The analyzers should be calibrated in-situ without disturbing the normal sampling inlet system to
the degree possible.
5.3.1.1 Precautions
A primary precaution is to check the calibration of the mass flow controllers in the
calibration system before calibration of the precursor gas analyzers, rather than relying solely on
the vendor-supplied calibration. This check should be conducted by comparison of readings over
the full range of each flow controller, using an independent and NIST-traceable flow
measurement device, such as an electronically timed bubble or piston-type flow meter.
Commonly used MFC calibrators allow the input of regression data from such comparisons, so
that corrections to the flow readings are made automatically by the calibration system.
Once the calibration system has been inspected, tested, and NIST traceability established, the
precursor analyzers can be calibrated in a field situation. The following additional precautions
should be observed.
• FEP or PTFE is the recommended material for all components and lines throughout
the calibration system, and all tubing and connection from the gas standard cylinders.
It is also preferred for any surfaces contacting the gas flows in solenoid valves in the
calibration system.
• When connecting cylinders to a MFC, make sure that the cylinder regulator is purged.
Fill and vent the regulator at least three successive times before connecting the
regulator to the delivery tubing. If a cylinder is changed, regulator purging is needed,
and the resulting calibration should be monitored very closely. If the response of the
analyzer is lower than before the cylinder change, then it can most likely be traced to
improper regulator purging. Purging and drying of a regulator can be aided by the
application of vacuum, as described in Section 3.6.1.
-------
Precursor Gas TAD
Section 5 QA/QC
Version 4
September 2005
Page 4 of 10
• It is recommended that stainless steel regulators with internal diaphragms that are
coated with Teflon® or other inert material be used with all calibration and audit
cylinders.
5.3.1.2 Calibration Procedures
Analyzer-specific SOPs should be developed based on the manufacturer's recommended
calibration procedures. Example SOP's prepared by EPA are included in Appendix B of this
TAD. However, the following steps outline the multipoint calibration procedure for the
precursor gas analyzers. If appropriate, a NIST-traceable multicomponent gas mixture (i.e., CO,
SO2, NO in N2) may be used to calibrate multiple precursor gas analyzers.
1. Allow both the calibration system and the analyzer to warm up properly. Consult the
manufacturer's instruction manual for specific details.
2. Record the normal QC check information. Especially note the zero and span
dial/readout values before starting the calibration.
3. Start introduction of zero air to the sampling manifold or the NOy inlet, and flag this
event in the data. With the zero air supplied at a constant flow rate, allow the
analyzer readings to stabilize on zero air. When stability is satisfactory, record the
response of the analyzer.
4. Set the MFC calibration unit to allow diluted calibration gas to flood the manifold or
NOy inlet at a concentration that is 80 to 90 percent of full scale of the analyzer
measurement range.
5. Allow the analyzer readings to stabilize at the working concentration. Record the
response from the analyzer.
-------
Precursor Gas TAD
Section 5 QA/QC
Version 4
September 2005
Page 5 of 10
6. Lower the concentration of the diluted calibration standard to 40 to 50 percent of the
analyzer's full scale. Repeat the previous step.
7. Lower the concentration of the diluted calibration standard to 15 to 20 percent of the
analyzer's full scale. Repeat Step 5.
8. Once three concentration levels are tested and all values recorded, allow the analyzer
to sample zero air again (Step 3), and again record the reading. Once the calibration
is completed, return the instrument to its normal ambient operational status.
The responses of the precursor analyzer should be analyzed by linear regression to assess
the results of the calibration. Acceptance criteria for the linear regressions are left to the
discretion of the monitoring agency, but the following are suggested: slope, 1 ± 0.10; intercept,
zero ± 1 x analyzer LDL or ±1% of the tested range (whichever is greater); and correlation
coefficient (r), > 0.995, where the ± values represent 95% confidence intervals. Regardless of
what criteria are selected, the instrument still must also pass audit tests, which require an
absolute difference between the analyzer reading and the standard gas concentration of no more
than 15 percent.
5.3.2 Level 1 Zero/Span Checks
Level 1 zero and span calibrations are simplified, two-point calibrations used when
adjustments may be made to the analyzer. When no adjustments are made to the analyzer, the
Level 1 calibration may also be called a zero/span "check" and must not be confused with a level
2 zero and span check. Level 1 zero and span checks should be conducted nightly. They are used
to assess if the analyzers are operating properly and to assess if any drift in instrument response
has occurred. The level 1 check should not exceed ±15 percent. If zero drift is internally
adjusted by the analyzer, the zero check is used to verify that the internal zero is working
properly. Zero checks are conducted by challenging the analyzer with zero air and a test
-------
Precursor Gas TAD
Section 5 QA/QC
Version 4
September 2005
Page 6 of 10
atmosphere containing concentrations between 70 percent and 90 percent of the full
measurement range in which the analyzer is operating. The challenge gas should be sampled
through as much of the sampling inlet system as practical to mimic the actual sampling of
ambient air. The results of the Level 1 zero/span check should be plotted on control charts to
graphically illustrate the trends in the response of the analyzer to the challenge gases. If the
measured concentrations fall outside of the control limits, the accuracy of the MFC calibration
system should be checked with a NIST-traceable flow standard. If the MFC flow accuracy is
confirmed, the data recorded since the last successful Level 1 check should be flagged and the
analyzer should be recalibrated using the multipoint calibration procedures described in Section
5.3.1.
State-of-the-art calibration equipment now exists that is fully automated. These "new
generation" calibration units are fully integrated with computers, mass flow calibrators, and the
associated hardware and software where they can create test atmospheres manually or
automatically. For the precursor gas program, it is recommended that the NCore sites have fully
automated calibration capability. Below are a number of reasons why this is advantageous:
• By performing the calibrations or checks automatically, agencies no longer spend the
manpower needed to perform them.
• Automated calibrations or checks can be triggered internally or by a DAS. Since newer
DASs allow remote access, this allows a remote user to challenge the analyzers without
actually being present.
• High sensitivity precursor gas analyzers are expected to have more zero and span drift than
less sensitive analyzers; therefore, it is important that a zero and Level I check be performed
daily.
-------
Precursor Gas TAD
Section 5 QA/QC
Version 4
September 2005
Page 7 of 10
• New generation DASs can record calibration data and allow remote users to track daily Level
I check and zero drift. This is important for data validation, verification and troubleshooting.
5.3.3 Precision Checks
At least once every two weeks a precision check should be conducted by challenging
each precursor analyzer with a known concentration of a standard gas mixture to assess the
ability of the analyzers to measure a gas under reproducible conditions. The precision checks
should be conducted by challenging the precursor analyzer with a standard gas of known
concentration in the ranges shown in Table 5-1. The gas must be supplied through all filters,
scrubbers, and other conditioners and should be supplied through as much of the sample inlet
system as possible. After completion of the precision check, the actual concentration of the
working standard and the measured concentration indicated by the analyzer should be reported
along with the percent difference between these values. Precision should be calculated at the end
of each calendar quarter as described in Section 2.3.1.1, 3.3.1.1, or 4.3.1.1.
Table 5-1. Concentration levels for biweekly precision checks.
Target
CO
S02
NOy
Concentration Range
250 to 500 ppb
10 to 50 ppb
Approx. 20 ppb
5.4 Independent Audits and Assessments
The effectiveness of the QA/QC procedures in place at each site should be evaluated
routinely through a series of independent audits and assessments. These assessments should
include:
• Proficiency test samples;
• Technical systems audits; and
-------
Precursor Gas TAD
Section 5 QA/QC
Version 4
September 2005
Page 8 of 10
• Audits of data quality.
5.4.1 Proficiency Test Samples
Proficiency test (PT) samples are used to ensure the performance of the precursor gas
analyzers used to collect the monitoring data. PT samples are challenges of the gas analyzers
with standards of known concentration that are independent of those standards used to calibrate
the analyzers. Generally this challenge is conducted as a blind audit, such that the site operator
is not aware of the gas standard concentrations delivered to the analyzers. Clearly, the
appropriate concentration values to be used for PT samples will be different for the different
precursor gases (CO, 862, and NOy). In addition, for any one of these gases, the appropriate PT
concentrations may vary with the analyzer operating range, which is selected based on the
characteristics of the monitoring site. Consequently, the recommended concentration ranges for
PT samples are given in Table 5-2 relative to the full scale range of the analyzer, rather than in
concentration units. At least one PT sample of known concentration is to be delivered to the
analyzer from each of the applicable ranges shown in Table 5-2. The indicated ranges are
consistent with the requirements of 40 CFR 58 Appendix A, Section 3.2.1.
Table 5-2. Concentration ranges for PT samples.
Audit Point
1
2
3
4
Percent of Full Scale Range3
3 to 8
15 to 20
35 to 45
80 to 90
a: Applies to operating range of CO, SO2, or NOy analyzer.
PT sample challenges should be conducted at least annually on each analyzer, and can be
conducted (a) by a person outside of the agency, or an independent QA group within the agency,
or (b) by having an independent audit device, such as used in the National Performance Audit
Program (NPAP), sent to the monitoring station. In the former case, an independent audit system
-------
Precursor Gas TAD
Section 5 QA/QC
Version 4
September 2005
Page 9 of 10
or standard is brought to the monitoring station and used to produce working standards of the
target gases that are supplied to the analyzers by the auditor. In the latter case, the audit device
provided to the monitoring agency produces working standards of the calibration gases that are
supplied to the analyzers. The operators and auditor do not know the concentrations of the
standards that are produced by the audit equipment. Responses of the precursor analyzers are
recorded and provided to the agency that supplied the audit device. That agency compares the
responses of the analyzers to the calculated concentrations from the audit device and provides an
audit report to the monitoring agency. In both cases, the PT sample audit should be conducted
by supplying the analyzer with the PT sample gas in its normal sampling mode such that the
audit gas passes through all sample inlet components used during normal ambient sampling.
Both the actual concentration of the PT sample gases and the concentration measured by
the analyzer being audited should be reported, along with the percent differences between these
concentrations for each audit point. The calculated percent differences are used to confirm the
analyzer precision and bias estimates obtained from routine checks.
The PT audit should also include an independent check of the gas flow controllers in the
calibration system, using a NIST-traceable flow standard.
5.4.2 Technical Systems Audit
A technical systems audit (TSA) is an on-site review and inspection of the operation of
an air monitoring station to assess its compliance with established QA/QC procedures and any
applicable regulations. TSAs assess whether all procedures for the monitoring program are
being followed and documented. A TSA should be conducted immediately before or shortly
after the start of monitoring and should be repeated at least every three years. TSAs should be
performed by a regional auditor who is knowledgeable of the monitoring program but
independent of routine operations.
-------
Precursor Gas TAD
Section 5 QA/QC
Version 4
September 2005
Page 10 of 10
5.4.3 Audits of Data Quality
An audit of data quality (ADQ) process is an examination of data after they have been
collected and verified. ADQs are conducted to determine how well the measurement system
performed with respect to goals specified in the Quality Assurance Project Plan (QAPP) and
whether the date were accumulated, transferred, calculated, summarized, and reported correctly.
The ADQ documents and evaluates the methods by which decisions were made during the data
treatment process.
5.5 References
1. "National Ambient Air Monitoring Strategy," U.S. Environmental Protection Agency,
Office of Air Quality Planning and Standards, Research Triangle Park, NC 27711,
April 2004, Final Draft.
2. EPA Order 5360.1 A2, "Policy and Program Requirements for the Mandatory Quality
Assurance Program," U.S. Environmental Protection Agency, Washington, D.C.,
May 2000.
3. "EPA Requirements for Quality Assurance Project Plans," EPA/QA R-5,
U.S. Environmental Protection Agency, Office of Environmental Information,
Washington, D.C., 20460, March 2001.
4. "Quality Assurance Handbook for Air Pollution Measurement Systems, Volume II:
Part 1," EPA-454/R-98-004, U. S. Environmental Protection Agency, Office of Air
Quality Planning and Standards, Research Triangle Park, NC, 27711, August 1998.
5. "Volume 1: A Field Guide to Environmental Quality Assurance," EPA/600/R-94/038a,
U.S. Environmental Protection Agency, Office of Research and Development,
Washington, D.C. 20460, April 1994.
-------
Precursor Gas TAD
Section 6 Data Acquisition
Version 4
September 2005
Page 1 of 13
6.0 DATA ACQUISITION AND MANAGEMENT
6.1 Introduction
The ambient pollutant data generated by the high sensitivity precursor gas analyzers in
the NCore network must be captured, organized, and verified in order to be useful. The process
of capturing the precursor gas data is known as data acquisition, whereas the organization of the
data is known as data management. Within both of these areas, quality assurance efforts and
data reviews must be carried out to verify the quality of the ambient data. This chapter of the
NCore TAD provides guidance in these areas, including identification of advanced equipment
and procedures that are recommended for implementation in NCore. The recommended
procedures rely on digital communication by the data acquisition system to collect a wider
variety of information from the precursor gas analyzers, to control instrument calibrations, and to
allow for more routine, automated, and thorough data quality efforts.
6.2 Data Acquisition and Analysis
Many S/L/T agencies currently perform precursor gas data acquisition by recording an
analog output from each precursor gas analyzer, using an electronic data logger. The analog
readings are converted and stored in digital memory in the data logger, for subsequent automatic
retrieval by a remote data management system. This approach can reliably capture the
monitoring data, but does not allow complete control of monitoring operations, and the recorded
analog signals are subject to noise that limits the detection of low concentrations. Furthermore,
with this data acquisition approach, the data review process is typically labor-intensive and not
highly automated. For these reasons, EPA discourages this approach, and instead strongly
recommends adoption of digital data acquisition methods. In that regard, the common analog
data acquisition approach often does not fully utilize the capabilities of the electronic data
logger. For example, a data logger used by many agencies for analog data acquisition is the ESC
Model 8816 Data Logger/Controller (Environmental Systems Corporation,
Knoxville, Tennessee). This and similar data loggers have the capability to acquire data in
-------
Precursor Gas TAD
Section 6 Data Acquisition
Version 4
September 2005
Page 2 of 13
digital form, and to control some aspects of calibrations and analyzer operation, but these
capabilities are not exercised in typical analog data acquisition approaches.
The recommended data acquisition approach for precursor gas monitoring in NCore is a
system that records analyzer readings and diagnostic information in digital form, and allows
direct digital communication between the data system and the precursor analyzers and calibration
systems, at one or more monitoring sites. Digital data acquisition reduces noise in the recording
of precursor gas monitoring data, thereby improving sensitivity, and also allows recording and
control of the instrument settings, internal diagnostics, and programmed activities of monitoring
and calibration equipment. Such data acquisition systems (DAS) also typically provide
automated data quality assessment as part of the data acquisition process.
It may be cost-effective for S/L/T agencies to adopt digital data acquisition and
calibration control simply by more fully exploiting the capabilities of their existing electronic
data loggers, such as the ESC 8816 noted above. For example, many high sensitivity gas
analyzers are capable of being calibrated under remote control. The opportunity to reduce travel
and personnel costs through automated calibrations is a strong motivator for S/L/T agencies to
make greater use of the capabilities of their existing data acquisition systems. Alternatively,
Section 6.2.1 presents one example of a new commercial data logger that is capable of
performing the recommended digital data acquisition. Also, Section 6.2.2 presents an example
of a highly sophisticated environmental data system capable of performing such digital data
acquisition and equipment control at multiple monitoring sites. Section 6.2.3 summarizes the
process by which precursor gas data is transferred from the high sensitivity analyzers to ultimate
use in public data systems such as EPA's Air Quality System (AQS).[1]
6.2.1 Example Data Logger: ESC 8832 Data System Controller
The ESC Model 8832 Data System Controller^2'3] (Environmental Systems Corporation,
Knoxville, Tennessee), shown in Figure 6-1, is an example of a digital data logger suitable for
-------
Precursor Gas TAD
Section 6 Data Acquisition
Version 4
September 2005
Page 3 of 13
Figure 6-1. ESC 8832 data system controller.
use in NCore precursor gas monitoring. The ESC 8832 is the most recent version in a line of
ESC digital data loggers that have been used in ambient monitoring, such as the ESC 8816,
which has been widely used by S/L/T agencies for recording analog monitoring data. Relative to
the Model 8816, the Model 8832 uses a 32-bit rather than 16-bit processor, provides more
memory capacity and somewhat faster data transfer, and allows comma-delimited parsing of
serial data. The ESC 8832 uses a 32-bit 50 MHz processor, and is designed for acquisition or
calculation of up to 99 data channels, with the capability of updating each channel once per
second. "Calculated" data channels include vector wind speed and direction and sigma theta. In
the standard configuration, the ESC 8832 can store over 100,000 data records and, with an
optional memory expansion card, can store over 300,000 data records. The ESC 8832
incorporates high-speed serial ports for data polling, and for downloads of configuration and
software upgrades without changing the data logger's EPROM. The ESC 8832 can record
internal diagnostic parameters of precursor gas analyzers, and program the performance of
zeroing and calibration procedures. Digital output relays can control valve switching or other
devices, with up to 32 events programmable. Up to 64 programmable alarms can be used to flag
recorded data values, and up to 32 calibrations can be applied to recorded data. The ESC 8832
can record analog inputs with 14-bit resolution over voltage ranges from ±100 mV to ±10 V full
-------
Precursor Gas TAD
Section 6 Data Acquisition
Version 4
September 2005
Page 4 of 13
scale. An Ethernet port and VGA analyzer output port are also available. The ESC 8832 is
capable of network communication with analyzers that have network ports installed (e.g.,
Thermo Environmental and API analyzers can provide this feature). Once configured as a
network client, the data logger can gather data from the instrument as if via a serial port.
Obtaining the maximum benefit from using a digital data logger, as exemplified by the
ESC 8832, requires using suitable software to manage the collected data. This requirement
especially applies when operating multiple ambient monitoring sites. To continue this example,
the E-DAS/ATX Ambient software developed by ESC is capable of polling one or more data
loggers at widely dispersed sites to process, edit, archive, and report the ambient data. The
Windows-based software can be used to perform data retrieval with error checking, automatic
time synchronization for multiple sites, graphical display of data, data storage and archiving, data
editing, and reporting of data in specified formats such as for EPA's AQS.[1]
6.2.2 Example Environmental Data System: ENVIDAS System
An example of the type of data system recommended for precursor gas monitoring at
NCore sites is the ENVIDAS environmental data acquisition system[4] (Envitech Ltd.,
Ramat Gan, Israel). The ENVIDAS system uses non-proprietary desktop or industrial personal
computer (PC) hardware to run flexible, Windows-based data acquisition software. A wide
variety of data acquisition cards may be installed for use with the ENVIDAS system. Typically,
the ENVIDAS system can accept up to 64 inputs, either analog or digital, with analog ranges
from ±1.25 V to ±10 V, and supports up to eight RS232C serial communication channels. Data
sampling rates (1, 2, 5, or 10 seconds) and averaging periods (e.g., 1, 5, 6, 15, or 60 minutes) are
user-selectable. Data storage includes calculated parameters such as vector wind speed and
direction and sigma theta. The multi-tasking software assures that data are not lost during data
polling or user interaction with the software. Data can be averaged, archived, displayed, and
reported in various formats including wind roses, pollution roses, histograms, and time series
plots.
-------
Precursor Gas TAD
Section 6 Data Acquisition
Version 4
September 2005
Page 5 of 13
The advantage of systems such as ENVIDAS, relative to data logging systems, is that
ENVIDAS is designed to communicate with and control a variety of precursor analyzers and
dilution calibrators at multiple monitoring sites. Thus, ENVIDAS facilitates remote control of
analyzer and calibrator operations at a monitoring site, as well as data acquisition. This
communication is achieved by means of RS232C protocols specific to each manufacturer's
analyzer or calibrator. For example, ENVIDAS supports RS232C protocols for analyzers made
by Thermo Environmental, API, Environnement S.A., and Monitor Labs, and for calibrators
made by Thermo Environmental, API, Horiba, Sabio, and Environics. Digital communication
and data acquisition means that the ENVIDAS system can access the internal software of an
analyzer or calibrator, and control the device's internal settings and parameters. Control of such
a device can be accomplished as fully through the ENVIDAS system as by an operator through
the front panel push buttons of the analyzer or calibrator itself. Basic analyzer settings, such as
the time and date, measurement range, units of concentration, time constant or averaging time,
and zero and calibration factors, can be controlled by the ENVIDAS system.
As in the data loggers exemplified in Section 6.2.1, a key feature for precursor
monitoring is recording of internal diagnostic information from the analyzers, to document the
status of the analyzer or diagnose data quality problems. For example, Table 6-1 lists several
internal status parameters for a typical CO, 862, and NOy analyzer that could be recorded and
controlled through the ENVIDAS (or a digital data logging) system.
EPA recommends that a high-performance data acquisition system such as ENVIDAS, or
a system such as the ESC 8832 data logger with digital data transfer and suitable software, be
used for precursor gas monitoring at NCore sites. Implementation of such a system offers a
monitoring agency the most up-to-date capabilities in instrument control, data acquisition, and
data analysis, especially when multiple monitoring sites are in operation.
-------
Table 6-1.
Precursor Gas TAD
Section 6 Data Acquisition
Version 4
September 2005
Page 6 of 13
Example Internal Diagnostic Parameters of High Sensitivity Precursor Gas
Analyzers Accessible to a Digital Data Acquisition System.
CO Analyzer
Optical chamber temperature
Optical chamber pressure
Temperature correction
Pressure correction
Sample flow rate
Power supply voltage
Sample/reference intensity ratio
Auto gain control frequency
Chopper motor speed
Alarms and max/min settings
SO2 Analyzer
Optical chamber temperature
Optical chamber pressure
Temperature correction
Pressure correction
Sample flow rate
Power supply voltage
Flash lamp voltage
PMT high voltage
Alarms and max/min settings
NOy Analyzer
Reaction chamber temperature
Reaction chamber pressure
P/T correction
Converter temperature
PMT cooler temperature
Sample flow rate
Power supply voltage
Alarms and max/min settings
Prereactor zero reading
Internal temperature
6.2.3 Summary Data Acquisition Process
Figure 6-2 illustrates the recommended digital data acquisition approach, in the form of a
schematic of the data flow from NCore precursor gas monitors through a local digital data
acquisition system, to final reporting of the data in various public databases. This schematic
shows several of the key capabilities of the recommended approach. A basic capability is the
acquisition of digital data from multiple analyzers and other devices, thereby reducing noise and
minimizing the effort needed in data processing. Another capability is two-way communication,
so that the data acquisition system can interrogate and/or control the local analyzers, calibration
systems, and even sample inlet systems, as well as receive data from the analyzers. Data transfer
to a central location is also illustrated, with several possible means of that transfer shown. S/L/T
agencies are urged to take advantage of the state of the art in this part of the process, as even
sophisticated technologies such as satellite data communication are now well established,
commercially available, and inexpensive to implement for monitoring operations. Finally, it is
important that data are reported in formats of immediate use in public data bases such as AQS,
and air quality index sites such as the multi-agency AIRNow site.[5] An advantage of software
such as the ESC E-DAX/ATS or ENVIDAS systems introduced above is their ability to facilitate
the assembly, formatting, and reporting of monitoring data to these data databases.
-------
Precursor Gas TAD
Section 6 Data Acquisition
Version 4
September 2005
Page 7 of 13
Manifold/external valves
Station Desktop
System
Calibrator
Zero Air Supply
Data pushed or pulled
from multiple stations
Figure 6-2. Flow of data from precursor gas analyzers to final reporting.
6.3 Data Acquisition System Quality Assurance
The use of a data acquisition system such as those described above is strongly
encouraged for the NCore precursor gas monitoring community. However, several practices
should be followed to ensure the quality of the data that are collected.
6.3.1 Personnel
Each organization conducting precursor gas monitoring should identify one or more
persons within the organization to implement and oversee the data acquisition and management
system. These personnel must have adequate education, training, and experience to perform
-------
Precursor Gas TAD
Section 6 Data Acquisition
Version 4
September 2005
Page 8 of 13
these functions, and must recognize the performance of these functions as a key (if not primary)
component of their responsibilities. These personnel contribute to quality assurance of the data
acquisition system through reviews and audits of the DAS and the collected data.
6.3.2 Security
The data acquisition system must be safeguarded against accidental or deliberate actions
that could result in the following potential problems:
Modification, destruction, or unwanted disclosure of data. The integrity of the data
should be maintained by (e.g.) implementing password protection and user authorization for
setup of the DAS, control of monitoring and calibration equipment, data acquisition and editing,
and data reporting.
Unavailability of the DAS or collected data. Protection against data loss may be avoided
by redundant data storage, and assurance of DAS operation may require hardware maintenance
or upgrades, surge protection, or backup systems.
6.3.3 Data Entry and Formatting
Electronic data acquisition systems such as those described above can record, average,
and compile the monitoring data in a variety of reporting formats. The personnel responsible for
the DAS should assure that the reported data are in the formats required for reporting to
databases such as AQS. Information on AQS requirements is available in the AQS Users
Guide[1] (Volume II, Air Quality Data Coding, and Volume III, Air Quality Data Storage).
Precursor gas monitoring data from NCore sites are to be reported to AQS as hourly
average values. However, it is suggested that S/L/T agencies also consider recording and
archiving data with shorter time resolution, e.g., as five minute averages. Such data can easily be
used to compute averages over longer time periods, and are valuable for diverse data analyses
beyond the purposes of AQS. For example, short time period data can be used to assess the
-------
Precursor Gas TAD
Section 6 Data Acquisition
Version 4
September 2005
Page 9 of 13
variability and uncertainty in hourly or longer time period data, and to evaluate temporal trends,
source impacts, and special research topics. The availability of high time resolution data will be
valuable to the data user community, and is likely to foster analyses of air quality that could not
be attempted with hourly or longer data periods.
6.3.4 Data Review
The review of collected data is the most important means to assure data quality in
ambient monitoring. The review process has multiple stages, beginning with observations in the
field, continuing through the analysis of electronic data, and ending with the reporting of final
data. Data review should be the subject of an SOP that defines the criteria an agency will apply
in processing and reporting the monitoring data.
Data review in the field should involve the observations and records of site operators on
topics such as the operational status of precursor analyzers, the need for maintenance or repair,
the occurrence of unexpected or unexplained readings, the existence of difficult or unusual
meteorological conditions, and the observation of ambient data outside the normal range for the
site. At a minimum, such observations must be recorded in a site notebook or other document.
Preferably, such observations should be recorded by electronic word processor, and ideally such
records are associated with the ambient data through the data acquisition system. Data review in
the field is the first step in flagging suspect data for subsequent review.
Data review is a key component of the data analysis process. Electronic data acquisition
systems allow automatic flagging of data based on the status (i.e., alarms, internal diagnostics,
calibration results) of the precursor analyzer, or based on other criteria such as expected data
ranges. However, review of the data by experienced personnel is still necessary. This review
should be carried out promptly after data collection and should take into account any field
observations such as those noted above. The aim of this review is to identify and remove suspect
data, and to identify and retain valid data based on the variety of information recorded. Software
associated with an electronic data acquisition system can be used to automatically compare
-------
Precursor Gas TAD
Section 6 Data Acquisition
Version 4
September 2005
Page 10 of 13
various types of data to flag or confirm the validity of the ambient measurements. This
capability should be exploited in data review for NCore.
The final step of data review is conducted in appropriately averaging and formatting data
to be reported. The usefulness of the AQS database, or other such publicly accessible data
repositories, is dependent on the consistency and accuracy of the processed data submitted to it.
Careful review of the data should take place to assure submission of complete and correctly
formatted data sets.
6.3.5 Calibrations and Audits
Quality assurance of the DAS is based on the system being operated within some range
of performance; i.e., that the data collected on the DAS and reported to the central database is the
same as that generated by the monitoring equipment. Among the practices used to document
DAS performance are routine calibration checks of the data acquisition system itself, data trail
audits, and performance audits.
6.3.5.1 DAS Calibration
In the case where analog signals from monitoring equipment are recorded by the DAS,
the calibration of a DAS is similar to the approach used for calibration of a strip chart recorder.
To calibrate the DAS, known voltages are supplied to each of the input channels and the
corresponding measured response of the DAS is recorded. Specific calibration procedures in the
DAS owner's manual should be followed when performing such DAS calibrations. The DAS
should be calibrated at least once per year.
In addition, precursor gas analyzers typically have an option to set output voltages to full
scale, or to ramp the analog output voltages supplied by the analyzer over the full output range.
Such a function can be used to check the analog recording process from the analyzer through the
DAS.
-------
Precursor Gas TAD
Section 6 Data Acquisition
Version 4
September 2005
Page 11 of 13
6.3.5.2 Data Trail A udit
A data trail audit consists of following one or more data values from the analyzer through
collection by the DAS, and through data processing to reporting to the central data repository.
This audit should be conducted by those personnel assigned to manage the data acquisition
hardware and software, and should be conducted at least annually. The procedure to be followed
is that one or more data points or data averages reported from the analyzer (e.g., hourly values)
should be collected by the DAS and checked on the DAS storage medium, and, in the final
format, reported to the data repository. The same values must be traceable through all steps of
the data acquisition and reporting process.
6.4 Data and Records Management
All raw data, averaged or processed data, operator documentation, correspondence on
data recording or processing issues, and other records should be retained for an appropriate
period of time after data have been collected, processed, and reported in the required format.
The following sections summarize specific data management recommendations for key types of
records.
6.4.1 Calibration Data
Calibration data should be recorded in the same manner as all other precursor gas
analyzer data, but appropriately flagged to distinguish it from routine ambient data. Any hard
copy records associated with calibration data must be linked or cross-referenced to the electronic
data, and the personnel who prepared the hard copy data must be fully identified. Calibration
data should be sequestered from ambient data in the data review process, and the appropriate
calibration results accurately applied to the relevant time periods of ambient monitoring.
Digital data acquisition systems such as those exemplified in Sections 6.2.1 and 6.2.2
offer a great advantage over analog systems in the tracking of calibration data, because of their
-------
Precursor Gas TAD
Section 6 Data Acquisition
Version 4
September 2005
Page 12 of 13
ability to control and record the internal readings of gas analyzers and calibration systems. That
is, a digital data acquisition system not only can record the analyzer's output readings, but can
schedule and direct the performance of analyzer calibrations, and record calibrator settings and
status. Thus flagging of calibration data to distinguish them from ambient monitoring data is
conducted automatically during data acquisition, with no additional effort or post-analysis.
These capabilities greatly reduce the time and effort needed to organize and quantify calibration
results.
6.4.2 Electronic Data Files
Electronic data files from different stages of the data collection and reporting process
should be retained for an appropriate period of time. The files to be retained should at first
include raw data files, intermediate files (e.g., edited to remove calibration or suspect data), and
the final data files submitted to the data repository. Any supporting information (including hard
copy records) should also be retained. Subsequent to final reporting of the data, a judgment may
be made to reduce the extent of the files retained. However, any data files that directly support
decisions made in the date review and editing process should be retained. In storing the
electronic data files, use should be made of file compression methods that reduce the size of files
for storage.
6.4.3 Hard Copies
As is evident from the discussion above, hard copy records are discouraged for precursor
gas monitoring in NCore, in favor of electronic data acquisition systems. However, some hard
copy records may be unavoidable; e.g., a site operator's personal laboratory notebook. There are
two key requirements for hard copy records. First, these records must be carefully reviewed by
the staff who recorded them and the personnel who oversee the electronic data acquisition to
assure that relevant information from the hard copy records is recognized and used in the data
analysis process. Second, in storing hard copy records, care must be taken to properly and
completely identify the records, and link them with the corresponding electronic data records.
-------
Precursor Gas TAD
Section 6 Data Acquisition
Version 4
September 2005
Page 13 of 13
6.5 References
1. "AQS User Guide: Air Quality System", U.S. Environmental Protection Agency, Office of
Air Quality Planning and Standards, Research Triangle Park, NC 27711, November 15,
2004; available at http://www.epa.gov/ttnairs 1 /airsaqs/manuals/AOSUserGuide.pdf.
2. Information on ESC Model 8832 data system controller, Environmental Systems
Corporation, Knoxville, Tennessee, available at http://www.escdas.com/cem/cem 8832.html.
3. Information on ESC EDAS data software, Environmental Systems Corporation, Knoxville,
Tennessee, available at http://www.escdas.com/ambient/atx.html.
4. Information on ENVIDAS data acquisition system, Envitech Ltd., Givataim, Israel, available
at http: //www. envitech. co. il/Envi dasF W. aspx.
5. AIRNow Web site, developed by the U.S. Environmental Protection Agency,
State/Local/Tribal Agencies, National Park Service, and National Oceanic and Atmospheric
Administration, with international and media partners, (http://www.epa.gov/airnow)
-------
APPENDIX A
SAMPLE MANIFOLD DESIGN FOR PRECURSOR
GAS MONITORING
-------
Sample Manifold Design for Precursor Gas Monitoring
Many important variables affect sampling manifold design for ambient precursor gas monitoring:
residence time of sample gases, materials of construction, diameter, length, flow rate, and
pressure drop. Considerations for these parameters are discussed below.
Residence Time Determination: The residence time of air pollutants within the sampling system
(defined as extending from the entrance of the sample inlet above the instrument shelter to the
bulkhead of the precursor gas analyzer) is critical. Residence time is defined as the amount of
time that it takes for a sample of air to travel through the sampling system. This issue is
discussed in detail for NOy monitoring in Section 4.2, and recommendations in Section 4 for the
arrangement of the molybdenum converter and inlet system should be followed. However,
residence time is also an issue for other precursor gases, and should be considered in designing
sample manifolds for those species. For example, Code of Federal Regulations (CFR), Title 40
Part 58, Appendix E.9 states, "Ozone in the presence of NO will show significant losses even in
the most inert probe material when the residence time exceeds 20 seconds. Other studies indicate
that 10-second or less residence time is easily achievable."1 Although 20-second residence time
is the maximum allowed as specified in 40 CFR 58, Appendix E, it is recommended that the
residence time within the sampling system be less than 10 seconds. If the volume of the
sampling system does not allow this to occur, then a blower motor or other device (such as a
vacuum pump) can be used to increase flow rate and decrease the residence time. The residence
time for a sample manifold system is determined in the following way. First the total volume of
the cane (inlet), manifold, and sample lines must be determined using the following equation:
Total Volume = Cv + Mv + Lv Equation 1
Where:
Cv = Volume of the sample cane or inlet and extensions
Mv = Volume of the sample manifold and moisture trap
Lv = Volume of the instrument lines from the manifold to the instrument bulkhead
A-l
-------
The volume of each component of the sampling system must be measured individually. To
measure the volume of the components (assuming they are cylindrical in shape), use the
following equation:
V = 7i * (d/2)2 * L Equation 2
Where:
V = volume of the component, cm3
7i = 3.14
L = Length of the component, cm
d = inside diameter of the component, cm
Once the total volume is determined, divide the total volume by the total sample flow rate of all
instruments to calculate the residence time in the inlet. If the residence time is greater than 20
seconds, attach a blower or vacuum pump to increase the flow rate and decrease the residence
time.
Laminar Flow Manifolds: In the past, vertical laminar flow manifolds were a popular design.
By the proper selection of a large diameter vertical inlet probe and by maintaining a laminar flow
throughout, it was assumed that the sample air would not react with the walls of the probe.
Numerous materials such as glass, plastic, galvanized steel, and stainless steel were used for
constructing the probe. Removable sample lines constructed of FEP or PTFE were placed to
protrude into the manifold to provide each instrument with sample air. A laminar flow manifold
could have a flow rate as high as 150 L/min, in order to minimize any losses, and large diameter
tubing was used to minimize pressure drops. However, vertical laminar flow manifolds have
many disadvantages which are listed below:
• Since the flow rates are so high, it is difficult to supply enough audit gas to provide an
adequate independent assessment for the entire sampling system;
• Long laminar flow manifolds may be difficult to clean due to size and length;
• Temperature differentials may exist that could change the characteristics of the gases, e.g., if
a laminar manifold's inlet is on top of a building, the temperature at the bottom of the
building may be much lower, thereby dropping the dew point and condensing water.
A-2
-------
For these technical reasons, EPA strongly discourages the use of laminar flow manifolds in the
national air monitoring network. It is recommended that agencies that utilize laminar manifolds
migrate to conventional manifold designs that are described below.
Sampling Lines as Inlet and Manifold: Often air monitoring agencies will place individual
sample lines outside of their shelter for each instrument. If the sample lines are manufactured
out of Polytetrafluoroethylene (PTFE) or Fluoroethylpropylene (FEP) Teflon®, this is
acceptable to the EPA. The advantages to using single sample lines are: no breakage and ease
of external auditing. However, please note the following caveats:
1. PTFE and FEP lines can deteriorate when exposed to atmospheric conditions, particularly
ultraviolet radiation from the sun. Therefore, it is recommended that sample lines be
inspected and replaced regularly.
2. Small insects and particles can accumulate inside of the tubing. It has been reported that
small spiders build their webs inside of tubing. This can cause blockage and affect the
response of the instruments. In addition, particles can collect inside the tubing, especially at
the entrance, thus affecting precursor gas concentrations. Check the sampling lines and
replace or clean the tubing on a regular basis.
3. Since there is no central manifold, these configurations sometimes have a "three-way" tee,
i.e., one flow path for supplying calibration mixtures and the other for the sampling of
ambient air. If the three-way tee is not placed near the outermost limit of the sample inlet
tubing, then the entire sampling system is not challenged by the provision of calibration gas.
It is strongly recommended that at least on a periodic basis calibration gas be supplied so
that it floods the entire sample line. This is best done by placing the three-way tee just
below the sample inlet, so that calibration gas supplied there is drawn through the entire
sampling line.
4. The calibration gas must be delivered to the analyzers at near ambient pressure. Some
instruments are very sensitive to pressure changes. If the calibration gas flow is excessive,
the analyzer may sample the gas under pressure. If a pressure effect on calibration gas
response is suspected, it is recommended that the gas be introduced at more than one place
in the sampling line (by placement of the tee, as described in item #3 above). If the response
A-3
-------
to the calibration gas is the same regardless of delivery point, then there is likely no pressure
effect.
Conventional Manifold Design - A number of "conventional" manifold systems exist today.
However, one manifold feature must be consistent: the probe and manifold must be constructed
of borosilicate glass or Teflon® (PFA or PTFE). These are the only materials proven to be inert
to gases. EPA will accept manifolds or inlets that are made from other materials, such as steel or
aluminum, that are lined or coated with borosilicate glass or the Teflon® materials named above.
However, all of the linings, joints and connectors that could possibly come into contact with the
sample gases must be of glass or Teflon®. It is recommended that probes and manifolds be
constructed in modular sections to enable frequent cleaning. It has been demonstrated that there
are no significant losses of reactive gas concentrations in conventional 13 mm inside diameter
(ID) sampling lines of glass or Teflon® if the sample residence time is 10 seconds or less. This is
true even in sample lines up to 38 m in length. However, when the sample residence time
exceeds 20 seconds, loss is detectable, and at 60 seconds the loss can be nearly complete.
Therefore, EPA requires that residence times must be 20 seconds or less (except for NOy).
Please note that for particulate matter (PM) monitoring instruments, such as nephelometers,
Tapered Element Oscillating Microbalance (TEOM) instruments, or Beta Gauges, the ambient
precursor gas manifold is not recommended. Particle monitoring instruments should have
separate intake probes that are as short and as straight as possible to avoid particulate losses due
to impaction on the walls of the probe.
T-Type Design: The most popular gas sampling system in use today consists of a vertical
"candy cane" protruding through the roof of the shelter with a horizontal sampling manifold
connected by a tee fitting to the vertical section (Figure 1). This type of manifold is
commercially available. At the bottom of the tee is a bottle for collecting particles and moisture
that cannot make the bend; this is known as the "drop out" or moisture trap bottle. Please note
that a small blower at the exhaust end of the system (optional) is used to provide flow through
the sampling system. There are several issues that must be mitigated with this design:
• The probe and manifold may have a volume such that the total draw of the precursor gas
analyzers cannot keep the residence time less than 20 seconds (except NOy), thereby
requiring a blower motor. However, a blower motor may prevent calibration and audit
A-4
-------
gases from being supplied in sufficient quantity, because of the high flow rate in the
manifold. To remedy this, the blower motor must be turned off for calibration.
However, this may affect the response of the instruments since they are usually operated
with the blower on.
Horizontal manifolds have been known to collect water, especially in humid climates.
Standing water in the manifold can be pulled into the instrument lines. Since most
monitoring shelters are maintained at 20-30 °C, condensation can occur when warm
humid outside air enters the manifold and is cooled. Station operators must be aware of
this issue and mitigate this situation if it occurs. Tilting the horizontal manifold slightly
and possibly heating the manifold have been used to mitigate the condensation problem.
Water traps should be emptied whenever there is standing water.
Screw Type Opening
Blower Motor
Sample Cane
roof line
Teflon Connectors -
Bushing
n n
Modular Manifold
Moisture Trap
adaptor
Figure 1. Conventional T-Type Glass Manifold System
California Air Resources Board "Octopus" Style: Another type of manifold that is being
widely used is known as the California Air Resources Board (CARB) style or "Octopus"
manifold, illustrated in Figure 2. This manifold has a reduced profile, i.e., there is less volume in
the cane and manifold; therefore, there is less need for a blower motor. If the combined flow
rates of the gas analyzers are high enough, then an additional blower is not required.
A-5
-------
roof line
n
Sample Cane
8-port "Octopus"
Manifold
Teflon Connectors -
Bushing
Screw Type Opening
Moisture Trap
Figure 2. CARB or "Octopus" Style Manifold
Placement of Tubing on the Manifold: If the manifold employed at the station has multiple
ports (as in Figure 2) then the position of the instrument lines relative to the calibration input line
can be crucial. If a CARB "Octopus" or similar manifold is used, it is suggested that sample
connections for analyzers requiring lower flows be placed towards the bottom of the manifold.
A-6
-------
Also, the general rule of thumb states that the calibration gas delivery line (if used) should be in
a location so that the calibration gas flows past the analyzer inlet points before the gas is
evacuated out of the manifold. Figure 3 illustrates two potential locations for introduction of the
calibration gas. One is located at the ports on the "Octopus" manifold, and the other is upstream
near the air inlet point, using an audit or probe inlet stub. This stub is a tee fitting placed so that
"Through-the-Probe" audit line or sampling system tests and calibrations can be conducted.
roof line
Instrument
inlet lines
Audit and probe
inlet stub
Sample Cane
Calibration
outlet line
Instrument
inlet lines
Figure 3. Placement of Lines on the Manifold
A-7
-------
Measurements and Features
1. Knurled Connector
2. O-ring
3. Threaded opening
4. Top extension - 56 mm
5. Overall Length - 304 mm
6. Outside diameter- 24 mm
7. Top and bottom shoulder - 50 mm
8. Length of inlet tube - 30 mm
9. Distancebetween inlet tubes -16 mm
10. Length of internal tube -145 mm
11. Width of inlet tube OD - 6 mm
12. Distance from inner tube to wall - 18mm
13. Inside width of outer tube 60 mm
14. Down tube length 76 mm
15. Width Down tube OD - 24 mm
16 Overall Width- 124 mm
Figure 4. Specifications for an 'Octopus" Style Manifold
Figure 4 illustrates the specifications of an Octopus style manifold. Please note that EPA-
OAQPS has used this style of manifold in its precursor gas analyzer testing program. This type
of manifold is commercially available.
A-8
-------
Vertical Manifold Design: Figure 5 shows a schematic of the vertical manifold design.
Commercially available vertical manifolds have been on the market for some time. The issues
with this type of manifold are the same with other conventional manifolds, i.e., when sample air
moves from a warm humid atmosphere into an air-conditioned shelter, condensation of moisture
can occur on the walls of the manifold. Commercially available vertical manifolds have the
option for heated insulation to mitigate this problem. Whether the manifold tubing is made of
glass or Teflon®, the heated insulation prevents viewing of the tubing, so the interior must be
inspected often. The same issues apply to this manifold style as with horizontal or "Octopus"
style manifolds: additional blower motors should not be used if the residence time is less than 20
seconds, and the calibration gas inlet should be placed upstream so that the calibration gas flows
past the analyzer inlets before it exits the manifold.
Glass Manifold
roof line
Insulation
Heater Power Cord
=D-
Support Pipe
Sample Ports
Exhaust Hose
"T" Connector
Manifold Support
>>
*fl»
Blower Motor
-, /
Floor
Figure 5. Example of Vertical Design Manifold
A-9
-------
Manifold/Instrument Line Interface: A sampling system is an integral part of a monitoring
station, however, it is only one part of the whole monitoring process. With the continuing
integration of advanced electronics into monitoring stations, manifold design must be taken into
consideration. Data Acquisition Systems (DASs) are able not only to collect serial and analog
data from the analyzers, but also to control Mass Flow Calibration (MFC) equipment and solid
state solenoid switches, communicate via modem or Ethernet, and monitor conditions such as
shelter temperature and manifold pressure. As described in Chapter 6, commercially available
DASs may implement these features in an electronic data logger, or via software installed on a
personal computer. Utilization of these features allows the DAS and support equipment to
perform automated calibrations (Autocals). In addition to performing these tasks, the DAS can
flag data during calibration periods and allow the data to be stored in separate files that can be
reviewed remotely.
Figure 6 shows a schematic of the integrated monitoring system at EPA's Burden Creek NCore
monitoring station. Note that a series of solenoid switches are positioned between the ambient
air inlet manifold and an additional "calibration" manifold. This configuration allows the DAS
to control the route from which the analyzers draw their sample. At the beginning of an Autocal,
the DAS signals the MFC unit to come out of standby mode and start producing zero or
calibration gas. Once the MFC has stabilized, the DAS switches the analyzers' inlet flow (via
solenoids) from the ambient air manifold to the calibration manifold. The calibration gas is
routed to the instruments, and the DAS monitors and averages the response, flagging the data
appropriately as calibration data. When the Autocal has terminated, the DAS switches the
analyzers' inlet flow from the calibration manifold back to the ambient manifold, and the data
system resets the data flag to the normal ambient mode.
The integration of DAS, solenoid switches, and MFC into an automated configuration can bring
an additional level of complexity to the monitoring station. Operators must be aware that this
additional complexity can create situations where leaks can occur. For instance, if a solenoid
switch fails to open, then the inlet flow of an analyzer may not be switched back to the ambient
manifold, but instead will be sampling interior room air. When the calibrations occur, the
instrument will span correctly, but will not return to ambient air sampling. In this case, the data
collected must be invalidated. These problems are usually not discovered until there is an
A-10
-------
Sampling Cane
Burden's Creek Sampling Station - OAQPS/MQAG
Manifold Fan ( F j
v-
Charcoal
Scrubber
Notes:
S - Teflon
P-Pump
F- Manifold Fan/Blower
V - Vent
Q] - Participate Filti
Sam pie tubing lengths < 3-ft
INC
J-. Calibration Manifold
T
V
Solenoid
Blower
Control Outputs
Cal Sys
fl
DAS Data |_| 1
, \~~\\ Analog Inp
Jwc
ts-
TNC
ru
^\ c
3 r-
L
TEC042CYTL
NOx
ntrol Out- To Solenoids
Environics
9100 Cal Sys
Zero Air
Source
TECO 43CTL
SO2
TECO 48CTL
CO
Other Monitor
(O3 etc)
Figure 6. Example of a Manifold/Instrument Interface
external "Through-the Probe" audit, but by then extensive data could be lost. It is recommended
that the operator remove the calibration line from the calibration manifold on a routine basis and
challenge the sampling system from the inlet probe. This test will discover any leak or switching
problems within the entire sampling system.
Figure 7 shows a close up of an ambient/calibration manifold, illustrating the calibration
manifold - ambient manifold interface. This is the same interface used at EPA's Burden's Creek
monitoring station. The interface consists of three distinct portions: the ambient manifold, the
solenoid switching system and the calibration manifold. In this instance, the ambient manifold is
a T-type design that is being utilized with a blower fan at the terminal. Teflon® tubing connects
the manifold to the solenoid switching system. Two-way solenoids have two configurations.
Either the solenoid is in its passive state, at which time the ports that are connected are the
normally open (NO) and the common (COM). In the other state, when it is energized, the ports
that are connected are the normally closed (NC) and the COM ports. Depending on whether the
A-ll
-------
solenoid is 'active' or not, the solenoid routes the air from the calibration or ambient manifold to
the instrument inlets. There are two configurations that can be instituted with this system.
1. Ambient Mode: In this mode the solenoids are in "passive" state. The flow of air (under
vacuum) is routed from the NO port through the solenoid to the COM port.
2. Calibration Mode: In this mode, the solenoids are in the "active" state. An external
switching device, usually the DAS, must supply direct current to the solenoid. This
causes the solenoid to be energized so that the NO port is shut and the NC port is now
connected to the COM port. As in all cases, the COM port is always selected. The
switching of the solenoid is done in conjunction with the MFC unit becoming active;
generally, the MFC is controlled by the DAS. When the calibration sequences have
finished, the DAS stops the direct current from being sent to the solenoid and switches
automatically back to the NO to COM (inactive) port configuration. This allows the air
to flow through the NO to COM port and the instrument is now back on ambient mode.
Calibration
Gas from the Mass
Flow Calibrator
Air Flow to the Analyzers
Figure 7. Ambient - Calibration Manifold Interface
A-12
-------
Reference
1. Code of Federal Regulations, Title 40, Part 58, Appendix E.9
A-13
-------
APPENDIX B
STANDARD OPERATING PROCEDURES FOR
SELECTED TRACE LEVEL PRECURSOR GAS
MONITORING AND CALIBRATION EQUIPMENT
SOP's also available at: http://www.epa.gov/ttn/amtic/precursop.html
-------
STANDARD OPERATING PROCEDURES
THERMO ELECTRON CORPORATION MODEL 48C-TLE
TRACE LEVEL CO INSTRUMENT
Version 2
-------
Section 1.1 Acknowledgments
This Standard Operating Procedure (SOP) for Carbon Monoxide - Trace Level is the product of EPA's
Office of Air Quality, Planning and Standards. The following individuals are acknowledged for their
contributions.
Principal Author
Dennis K. Mikel, OAQPS-EMAD, Research Triangle Park, NC 27711
Reviewers
Joann Rice, Trace Gas Team Lead, OAQPS-EMAD, Research Triangle Park, NC 27711
Anna Kelly, OAQPS-EMAD, Research Triangle Park, NC 27711
Keith Kronmiller, Mantech, Inc. Research Triangle Park, NC 27711
Comments and questions can be directed to:
Joann Rice
EPA-OAQPS
Emission, Analysis and Monitoring Division
Mail Drop D243-02
Research Triangle Park, NC 27711
Email: ricejoMin@epa.gov
Phone: (919)-541-3372
-------
Section 2.0 Table of Contents
Section
1.0 Title Page
1.1 Acknowledgements
2.0 Table of Contents
2. 1 List of Tables and Figures
3.0 Procedures
3.1 Scope and Applicability
3.2 Summary of Method
3.3 Definitions
3.4 Health and Safety Warnings
3.5 Cautions
3.6 Interferences
3.7 Personal Qualifications
3.8 Equipment and Supplies
3.9 Procedure
3.9.1 Sample Collection
3.9.2 Sample Handling and Preservation
3.9.3 Instrument Operation, Start up and Maintenance
3.9.3.1 Startup
3.9.3.2 Operation and Range Setting
3.9.9.3 Diagnostic Checks/Manual Checks
3.9.3.4 Preventive Maintenance
3.9.3.5 Instrument Trouble shooting
3.9.4 Calibration and Standardization
3.9.4.1 Adj ustment to Zero Air
3.9.4.2 Adjustment to Calibration Gas
3.10 Data Analysis and Calculations
4.0 Quality Control and Quality Assurance
4.1 Precision
4.2 Bias
4.3 Representativeness
4.4 Completeness
4.5 Comparability
4.6 Method Detection Limit
5.0 References
Page
1
2
3
4
5
5
5
7
7
7
8
8
8
10
10
10
10
10
11
11
13
14
14
14
14
15
15
15
15
16
16
16
16
17
-------
List of Tables
Table Name and Number
3-1 Definition of Key Terms
3-2 Diagnostic Checks
3-3 Preventive Maintenance Schedule for the TECO 48C-TLE
4-1 Measurement Quality Assurance Objectives
4-2 Operating Parameters for the TECO 48C-TLE Trace Gas Instrument
Page
7
13
13
16
17
List of Figures
Figure Name and Number
3-1 Schema of the TECO 48C-TLE
3-2 Menu Tree Schema
A-l Monthly Quality Control and Maintenance Record
Page
9
12
18
-------
TECO 48C-TLE CO SOP
Version No. 2
2/3/2005
Page 5 of 18
STANDARD OPERATING PROCEDURES
THERMO 48C-TLE TRACE-LEVEL CO INSTRUMENT
3.0 PROCEDURES
3.1 Scope and Applicability
Carbon Monoxide (CO), a colorless, odorless, tasteless, highly poisonous gas has detrimental effects on
human health. CO originates from the partial oxidation of hydrocarbon fuels, coal and coke1. CO affects
the oxygen carrying capacity of the blood. CO can diffuse through the alveolar walls of the lungs and
compete with oxygen for one of the four iron sites in the hemoglobin molecule. The affinity of the iron
site for CO is approximately 210 times greater than oxygen2. Low levels of CO can cause a number of
symptoms including headache, mental dullness, dizziness, weakness, nausea, vomiting and loss of
muscular control. In extreme cases, collapse, unconsciousness and death can occur.
The Thermo Electron Corporation (TECO) model 48C-TLE is a state of the science instrument for the determination
of trace levels of CO by Non-Dispersive Infrared Spectrophotometry (NDIR) using Gas Filter Correlation (GFC).
This SOP will detail the operation, preventive maintenance, cautions and health warnings.
The Detection Limit (DL) for non-trace levels of CO is 1.0 parts per million (ppm) (Code of Federal
Regulations, Volume 40, Part 53.23c, or, in the shortened format used hereafter, 40 CFR 53.23c)3.
However, the 48C-TLE has a DL to 20 parts per billion (ppb), which is accomplished by modifications to
the Federal Reference Method (FRM) instruments. This document will discuss the Trace Level (TL)
operating procedures in detail.
3.2 Summary of Method
The analytical principle is based on absorption of IR light by the CO molecule. NDIR-GFC analyzers
operate on the principle that CO has a sufficiently characteristic IR absorption spectrum such that the
absorption of IR by the CO molecule can be used as a measure of CO concentration in the presence of
other gases. CO absorbs IR maximally at 2.3 and 4.6 um. Since NDIR is a spectrophotometric method, it
is based upon the Beer-Lambert Law. The degree of reduction depends on the length of the sample cell,
the absorption coefficient, and CO concentration introduced into the sample cell, as expressed by the
Beer-Lambert law shown below:
T = I/Io = e('axc) (equation 1)
where:
T = Transmittance of light through the gas to the detector
I = light intensity after absorption by CO
lo = light intensity at zero CO concentration
a = specific CO molar absorption coefficient
x = path length, and
C = CO concentration
-------
TECO 48C-TLE CO SOP
Version No. 2
2/3/2005
Page 6 of 18
For Gas Filter Correlation, there is only one sample cell. This cell acts as the sample and reference cell.
The broad band of IR radiation is emitted from an IR source. The IR light passes through a very narrow
band pass filter which screens out most wavelengths and allows only the light that CO absorbs to enter
the sample cell. The GFC analyzer has a chopper wheel with two pure gases: Nitrogen and CO. As the
chopper wheel rotates and allows the IR energy to enter "CO side" of the wheel, all IR energy that could
be absorbed by CO in the sample stream is absorbed by the CO in the wheel. This technique effectively
"scrubs out" any light that could possibly be attenuated. The single detector records the light level (lo).
As the wheel spins, the "N2 side" of the wheel reaches the IR energy beam. This side of the wheel allows
all IR light to pass through the wheel and be absorbed by any CO that might be in the sample gas. This
light level is CO sensitive (I). The detector records the attenuation of this light, compares the two light
levels (I/Io) and sends a signal to the electrometer board that calculates the concentration. The voltage is
related to the CO concentration according to the Beer-Lambert law in equation 1 shown above. Thus,
TECO 48C-TLE can be measured continuously. The 48C-TLE version has four distinct features that
allow it to measure CO at ppb levels:
• Required sample stream dried using permeation dryer;
• Analyzer baseline determined and corrected using heated Carolite catalytic converter;
• Frequent auto-zero, at a minimum once per hour, through the catyalytic converter;
• The instrument has an ultra-sensitive or "hot" detector.
The 48C-TLE instrument operates in the following fashion:
1. In sample mode, ambient air is allowed to enter through the rear bulkhead sample port. Solenoid #1
is in its Normally Open (NO) mode. The ambient air flows through the solenoid to the permeation
dryer, which removes the moisture and water from the air stream.
2. The sample stream then passes through a sample filter, which removes particles that can build up on
the mirror and sample chamber and attenuate the IR beam.
3. The sample then enters the sample cell. A major difference between a non-TL and TL instrument is
the detector. The TL instrument has a detector that is more sensitive to the light emitted and
absorbed in the sample cell. This detector must be more sensitive because the amount of attenuation
by the CO gas in the sample stream is much lower. Therefore, the detector must be sensitive at lower
ambient levels. Temperature of the sample cell is also critical. The sample cell and detector must be
maintained at a constant temperature in order for the detector to keep a stable background.
Fluctuations of more than 1° Centigrade can cause the baseline to drift, giving false readings at low
levels.
4. The detector sends the signal to the demodulator which interprets the signal. The demodulator sends
a digital value to the Central Processor Unit (CPU).
5. At the end of the hour, the CPU sends a voltage signal to the Solenoid #1 and switches it to the
"Normally Closed" (NO position. This allows room air to be drawn into the instrument and to pass
through the catalytic converter. The catalytic converter uses a Carolite bed heated to 50° Centrigrade
to convert all CO to Carbon Dioxide (2CO + O?_—> 2CO?). This effectively "scrubs" all CO from
the sample stream. The CO "free" air flows through the sample cell and the CPU interprets the
signal from the demodulator as the "background" or "baseline" value. The baseline is then adjusted
at that time. The baseline adjustment takes 10 minutes.
6. The CPU then switches Solenoid #1 to its NO position and ambient air then drawn into the analyzer.
-------
TECO 48C-TLE CO SOP
Version No. 2
2/3/2005
Page 7 of 18
3.3 Definitions
Here are some key terms for this method.
Table 3-1 Definitions of Key Terms
Term
DAS
Interferences
Definition
Data acquisition system. Used for automatic collection and
recording of CO concentrations.
Physical or chemical entities that cause CO measurements to be
higher (positive) or lower (negative) than they would be without the
entity. (See Section 3.6).
3.4 Health and Safety Warnings
To prevent personal injury, please heed these warnings concerning the 48C-TLE.
1. CO is a poisonous gas. Vent any CO or calibration span gas to the atmosphere rather than into the
shelter or other sampling area. If this is impossible, limit exposure to CO by getting fresh air every
5 to 10 minutes. If the operator experiences light headedness, headache or dizziness, leave the area
immediately.
2. The IR source is a filament resistor that has an electrical current running through it. The IR source
can become very hot. When troubleshooting, allow the instrument to cool off especially if you
suspect the IR source as the cause of trouble.
3. Always use a third ground wire on all instruments.
4. Always unplug the analyzer when servicing or replacing parts.
5. If it is mandatory to work inside an analyzer while it is in operation, use extreme caution to avoid
contact with high voltages. The analyzer has a 110 volt Volts Alternating Current (VAC) power
supply. Refer to the manufacturer's instruction manual and know the precise locations of the VAC
components before working on the instrument.
6. Avoid electrical contact with jewelry. Remove rings, watches, bracelets, and necklaces to prevent
electrical bums.
3.5 Cautions
To prevent damage to the 48C-TLE. all cautions should immediately precede the applicable step in this
SOP. The following precautions should be taken:
1. Normally, if Teflon™ filters are used in the sample train, cleaning the optical bench will not be
required. However, in the event that the bench is cleaned, be careful to avoid damaging the interior of
the sample chamber. In addition, some GFC instruments have a series of mirrors that deflect the light
in order to increase the path length. The mirrors are aligned at the factory. If the mirrors become
misaligned, the IR light beam will not be directed to the detector. Use extreme caution when cleaning
or servicing the sample chamber(s). In addition the mirrors are very fragile Avoid dropping the
instrument. This may damage, misalign or crack the mirrors and cause expensive repairs.
2. Keep the interior of the analyzer clean.
3. Inspect the system regularly for structural integrity.
-------
TECO 48C-TLE CO SOP
Version No. 2
2/3/2005
Page 8 of 18
4. To prevent major problems with leaks, make sure that all sampling lines are reconnected after
required checks and before leaving the site.
5. Inspect tubing for cracks and leaks. The permeation dryer may rest upon parts that vibrate, such as the
air pump. Check the areas of the permeation dryer where they come into contact with other parts.
6. It is recommended that the analyzer be leak checked after replacement of any pneumatic parts.
7. If cylinders are used in tandem with Mass Flow Control (MFC) calibrators, use and transport of
cylinders are a major concern. Gas cylinders can sometimes contain pressures as high as 2000
pounds per square inch. Handling of cylinders must be done in a safe manner. If a cylinder is
accidentally dropped and valve breaks off, the cylinder can become explosive or a projectile.
8. Transportation of cylinders is regulated by the Department of Transportation (DOT). It is strongly
recommended that all agencies contact the DOT or Highway Patrol to learn the most recent
regulations concerning transport of cylinders.
9. CO is a highly poisonous gas. Long term exposure can cause problems with motor coordination and
mental acuity. It is strongly recommended that all agencies have Material Safety Data Sheets
(MSDS) at all locations where CO cylinders are stored or used. MSDS can be obtained from the
DOT or from your vendor.
10. It is possible (and practical) to blend other compounds with CO. If this is the case, it is recommended
that MSDS for all compounds be made available to all staff that use and handle the cylinders or
permeation tubes.
11. Shipping of cylinders is governed by the DOT. Contact the DOT or your local courier about the
proper procedures and materials needed to ship high-pressure cylinders.
3.6 Interferences
Water Vapor: Studies have shown conclusively that NDIR analyzers have interference from water
vapor. Water absorbs very strongly across several bands of IR spectra. Water vapor interference occurs
because water vapor absorption of light in the region of 3.1, 5.0 -5.5 and 7.1 -10.0 um in the IR region.
Since water vapor absorbs light in this region, this has a quenching effect on the reaction of CO. The
TECO 48C-TLE is equipped with a permeation drier, which effectively scrubs all water and water vapor.
No maintenance is required on the dryer.
Carbon Dioxide: CO2 absorbs in the IR spectrum at 2.7, 5.2, and 8.0 to 12.0 um. This is very close to
the regions that CO absorbs within as well. However, since atmospheric carbon dioxide is much higher in
concentration than CO, this UV spectral range must be avoided. To prevent light in this spectral region,
the TECO 48C-TLE analyzer has a band pass filter that blocks these wavelengths.
3.7 Personal Qualifications
The person(s) chosen to operate the TECO 48C-TLE should have a minimum of qualifications. The
understanding of basic chemistry and electronics are a must. The understanding of digital circuitry is
helpful, but not required. Also, courses in data processing and validation are also welcome.
3.8 Equipment and Supplies
Monitoring Apparatus: The design of the 48C-TLE is identical to the 48C, with several major
variations. A diagram of the TECO 48C-TLE instrument is described in Figure 3-1. The three main
components are:
• Pneumatic System: Consists of sample inlet line, particulate filter, filter holder, permeation dryer,
reaction chamber, flowmeter, and pump, all used to bring ambient air samples to the analyzer inlet.
-------
TECO 48C-TLE CO SOP
Version No. 2
2/3/2005
Page 9 of 18
• Analytical System: This portion of the instrument consists of the IR source, the correlation wheel,
motor, mirrors, detector and band pass filter.
• Electronic Hardware: The part of the analyzer that generally requires little or no maintenance. The
brain of the 48C series is the CPU. It monitors and regulates motor speed, temperatures, flows and
pressure. It also monitors and stores diagnostic information. If the 48C-TLE is operated above the
manufacturer's recommended temperature limit, however, individual integrated chips can fail and
cause problems with data storage or retrieval.
Other apparatus and equipment includes the following.
Instrument Shelter: A shelter is required to protect the analyzer from precipitation and adverse weather
conditions, maintain operating temperature within the analyzer's temperature range requirements, and
provide security and electrical power. The recommended shelter temperature range is 20-30°C.
Spare Parts and Incidental Supplies: See the TECO 48C-TLE operating manual, Section 5-lfor
specific maintenance and replacement requirements.
Calibration System: A system that creates concentrations of CO of known quality is necessary for
establishing traceability. This is described in detail in the "Environics 9100 SOP." Please reference this
document.
DAS: A data acquisition system is necessary for storage of ambient and ancillary data collected by the
48C-TLE.
IR Source
Sairple
Span Gas
Sample Cell -(Jgggggg£> - Wheel
Display analog HS-232
Output Digital
Output
Figure 3-1 Schema of the TECO 48C-TLE
-------
TECO 48C-TLE CO SOP
Version No. 2
2/3/2005
Page 10 of 18
Wiring, Tubing and Fittings: Teflon™ and borosilicate glass are inert materials that should be used
exclusively throughout the intake system. It is recommended that Polytetrafluoroethylene (PTFE) or
Fluoroethylpropylene (FEP) Teflon™ tubing be used. PTFE and FEP are the best choice for the
connection between the intake manifold and the 48C-TLE bulkhead fitting. Examine and discard if
particulate matter collects in the inlet tubing. All fittings and ferrules should be made of Teflon™ or
stainless steel. Connection wiring to the DAS should shielded two strand wire or RS-232 cables for
digital connections.
Reagents and Standard: The TECO 48C-TLE does not require any reagents since the instrument uses
photometry to analyze for CO. All standards for the CO method can be obtained in compressed cylinders
and must be NIST traceable. Please see the "Calibration of Trace Gas Analyzers" SOP.
3.9 Procedure
3.9.1 Sample Collection: Sampling for Trace Level CO is performed by drawing ambient air through a
sample manifold directly into the analyzer continuously via a vacuum pump. All inlet materials must be
constructed from Teflon™ or borosilicate glass as detailed in 40 CFR 58. The siting criteria for CO
Trace Level instruments in detailed in 40 CFR 58, appendix A4.
3.9.2 Sample Handling and Preservation: CO samples receive no special preparation prior to analysis.
Therefore this SOP does not need a section on Sample Handling and Preservation.
3.9.3 Instrument Operation, Startup and Maintenance
This section discusses startup, operation and maintenance of the 48C-TLE. The TECO 48C-TLE series
instrument has a digital front panel screen with selection switches below. This allows the user to check
functions, switch operating parameters, adjust zero and span and read warnings messages. It is
extremely important that the users familiarize themselves with the menus available. Inadvertently
changing parameters within the analyzer can damage the instrument and possibly invalidate data
as well. Please reference the TECO 48C-TLE owner's manual and read it carefully before
adjusting any parameters that are set by the factory.
3.9.3.1 Start up
The following text is taken from the TECO 48C-TL manual. It is the identical for the 48C-TLE model.
The Model 48C Trace Level is shipped complete in one container. If, upon receipt of the analyzer, there is
obvious damage to the shipping container, notify the carrier immediately and hold for inspection. The
carrier, and not Thermo Environmental Instruments Inc., is responsible for any damage incurred during
shipment. Follow the procedure below to unpack and inspect the instrument.
1. Remove the instrument from the shipping container and set on a table or bench that allows easy
access to both the front and rear of the instrument.
2. Remove the instrument cover to expose the internal components and remove any packing material.
3. Check for possible damage during shipment and check that all connectors and printed circuit boards
are firmly attached.
4. Re-install the instrument cover.
5. Connect the sample line to the sample bulkhead on the rear panel. Ensure that the sample line is not
contaminated by dirty, wet or incompatible materials. All tubing should be constructed of Teflon™ or
borosilicate glass with an OD of %" and a minimum ID of 1/8".
-------
TECO 48C-TLE CO SOP
Version No. 2
2/3/2005
Page 11 of 18
6. The length of the tubing should be less than 10 feet.
7. All gas must be delivered to the instrument at atmospheric pressure. It may be necessary to employ an
atmospheric bypass plumbing arrangement or attach the instrument inlet line to a manifold that is
vented to the atmosphere.
8. Connect the exhaust bulkhead to a suitable vent. The exhaust line should be %" OD (outside
diameter) with a minimum ID of 1/8" OD. The length of the exhaust line should be less than 10 feet.
Verify that there is no restriction in this line.
9. Connect a suitable recording device to the rear panel terminals. The EPA recommends, but does not
require, recording of the data digitally. The TECO 48C-TLE has this option. Please refer to the "Data
Acquisition and Management" SOP. If the DAS system that you have does not have the RS-232
capabilities, then proceed to the next section, Diagnostic Checks/Manual Checks. If you have
connected the 48C-TLE to a computer or DAS, review the Diagnostic Check from your computer
screen. TECO offers TECO communication software, a computer program that allows the operator to
log the diagnostic data that is collected by the 48C-TLE CPU. Several DAS manufacturers offer this
type of software as well.
10. Plug the instrument into an outlet of the appropriate voltage and frequency. The Model 48C Trace
Level is supplied with a three-wire grounding cord. Under no circumstances should this grounding
system be defeated.
11. Turn the power on.
3.9.3.2 Operation and Range Setting
1. The exhaust fan will start and the display will come on. The Central Processing Unit (CPU) will boot
the system and load the firmware.
2. The display has a 4 line by 20 character alphanumeric display that shows the sample concentration,
instrument parameters, instrument controls and help/warning messages. The menus for access (as
described in sections 3.9.3.4 and 3.9.3.5 of this SOP) are performed using the display and the 6 push-
buttons just below the display.
3. Once the instrument loads the firmware, the display will display "CO PPM XX.XX, below this
value is the time and "REMOTE." This is the "RUN" menu and should always be left in the "RUN"
menu when it is collecting ambient data.
4. To access the Main Menu, press the "Menu." This will put you into the Main Menu.
5. From the Main Menu, Use the | and J, pushbuttons move the cursor down so that the arrow is next to
"Range" selection. Press "Enter."
6. The Range menu contains the gas units, CO ranges, and the custom ranges. In the upper right corner
of the display, the words "SINGLE, DUAL, or AUTO" is displayed to indicate the active mode. The
"Range" menu in the dual and auto-range modes appear the same except for the word DUAL or
AUTO, displayed in the upper right corner. For more information about the SINGLE, DUAL, or
AUTO-range modes, see page 3-6 of the owner's manual. The default is set to single range.
7. To set the upper scale range, use the t and J, pushbuttons to move the cursor down so that the arrow is
next to "CO Range" selection. Press "Enter."
8. Use the "f^-^^-" buttons on the front panel to enter in the desired concentration, this should be
"5." Press "Enter."
9. Press "Menu" and then press "Run." The full scale range has now been set to 5.00 ppm. The default
for the units is "ppm." It is recommended that this not be changed.
10. In single range mode, there is one range, one averaging time, and one span coefficient. The two CO
analog outputs are arranged on the rear panel terminal strip as shown in Figure 3-3 of the owner's
manual. To use the single range mode, set option switches 4 and 5 off. For more information about
-------
TECO 48C-TLE CO SOP
Version No. 2
2/3/2005
Page 12 of 18
setting the internal option switches, see .Internal Option Switches, on page 3-7.
3.9.3.3 Diagnostic Checks/Manual Checks
To determine whether the 48C-TLE is working properly, the field operators should perform the
Diagnostic Checks every time they visit the monitoring station. It is good practice for the operator to
check these Diagnostic Checks either by the computer or manually. Figure 3-2 of this SOP has the menu
"tree" that was taken from the TECO 48 C-TL manual. By pressing the "menu" button and following the
tree, an operator can easily get to the location needed.
Power-Up 5cr?cn
Salf-Taat Scraan
Screen
Main Menu
Tims
Fact DM
Santa Mmdu 'On1
Calibration
Instrument
Qin train
Alarm
Cat unit*
CO range
CUttom imcri
Avg,
Hand.
-------
TECO 48C-TLE CO SOP
Version No. 2
2/3/2005
Page 13 of 18
4. Once the operator has reviewed all of the diagnostic information, return to the Main Menu by
pressing "Menu" followed by "Run."
Table 3-2 Diagnostic Checks
Check
Program Number
Voltages
Temperatures
Pressure
Flow
S/R Ratio
AGC Intensity
Motor Speed
Explanation
The current software version used by the instrument
There are four voltages that should be recorded.
There are two temperatures that should be recorded: internal and chamber.
Reference Page 3-47 for details.
This is the pressure inside of the optics bench.
This is the actual flow rate through the optics bench. Reference page 3-53 for
specifications.
Sample/Reference Ratio. This is the ratio of the intensities of the light source
through the sample side and the reference side of the correlation wheel. Please
See page 3-38 for the optimal values.
Automatic Gain Control. The AGC displays the intensity (in Hertz) of the
reference channel. The AGC circuit optimizes the noise and resolution level of
the analyzer. Please see page 3-39 for details.
This displays, in percentage, the status of the chopper motor. This value should
be 100%.
Once the Diagnostic checks have been established and recorded for the 48C-TLE, it is time to calibrate
the instrument. Please refer to section 3.9.4 of this SOP.
3.9.3.4 Preventive Maintenance
Preventive maintenance should prevent down-time and data loss. Table 3.3 lists the preventive
maintenance items that are should be performed. Please see section 5-1 of the owner's manual. Section
5-1 also has a list of the spare parts that the operator should keep in stock.
Table 3-3 Preventive Maintenance Schedule the TECO 48C-TLE
Item
Replace particle filter
Diagnostics Checks
Perform Level I calibration
Replace IR source
Leak Check and Pump Check Out
Inspect Pneumatic Lines
Clean inside of Chassis
Schedule
Weekly
Weekly
Daily
As needed
Annually
Semi-annually
As needed
-------
TECO 48C-TLE CO SOP
Version No. 2
2/3/2005
Page 14 of 18
Rebuild or replace pump
Clean optic bench
Replace wheel motor
Replace gases in correlation wheel
As needed
As needed
As needed
As needed
3.9.3.5 Instrument Troubleshooting
The TECO 48C-TLE manual has an excellent troubleshooting guide in Section 6-1 of the manual. For
details on using the Test Functions for predicting failures, please reference this section.
3.9.4 Calibration and Standardization
The calibration of the TECO 48C-TLE is performed by comparing the digital or analog output of the
instrument against standardized gases of known quality. Generation of these gases is detailed in the
"Calibration of Trace Gas Analyzers" SOP. The recommended ranges for the calibration are detailed in
Table 4-2. This section will detail how to adjust the 48C-TLE to the standardized gases. Once the
calibration has been performed, compare the response of your DAS to the calculated "source" value. If
this is outside of +/-10%, then adjust the instrument response as detailed in the next sections of this SOP.
3.9.4.1 Adjustment to Zero Air
In order to adjust the output of the 48C-TLE to zero air, perform the following:
1. Allow the instrument to sample zero air from a manifold that is at near atmospheric pressure for a
minimum of 15 minutes.
2. On the front panel press the "Menu" button. This will bring up the main menu. Using the ^
arrow until the cursor is on the "Calibration" selection. Press "Enter."
3. This next screen is the "Calibration" screen. In this screen press the f^ buttons until you align
the cursor at the "Calibrate Zero" selection. Press "Enter."
4. The next screen will show a "CO PPM X.X above the words "SET TO ZERO?" If the
analyzer has stabilized to zero air, press "Enter." Then Press "Menu" and then press "Run." This
will adjust the baseline to the zero air. If you decide to adjust the higher range response,
continue onto Section 3.9.4.2.
3.9.4.2 Adjustment to Calibration Gas
In order to adjust the output of the 48C-TLE to NIST traceable calibration gas, perform the following:
1. Switch the calibration unit to generate a known concentration of CO. Allow the instrument to sample
calibration gas from a manifold that is at near atmospheric pressure for a minimum of 15 minutes.
2. On the front panel press the "Menu" button. This will bring up the main menu. Using the ^ arrow
until the cursor is on the "Calibration" selection. Press "Enter."
3. This next screen is the "Calibration" screen. In this screen press the f^ buttons until you align the
cursor at the "Calibrate CO" selection. Press "Enter."
4. The first line of the display shows the current CO concentration reading. The second line shows the
instrument range and the third line states, "SET TO XX.XX." The next line shows "
INC/DEC." The cursor should be under one of the digits in the third line. Use the "
buttons on the front panel to enter in the desired concentration, that is being generated by the
calibration system. Press "Enter."
-------
TECO 48C-TLE CO SOP
Version No. 2
2/3/2005
Page 15 of 18
5. Then Press "Menu" and then press "Run." This will adjust the response of the instrument to the
calibration gas concentration. The instrument is now calibrated.
3.10 Data Analysis and Calculations
Data analysis for this analyzer is detailed in "Data Acquisition and Management" SOP. For the TECO
48C-TLE, there is one design detail of which the operator must be aware; the auto-zero function. As
detailed in Section 3.1 of this SOP, the TECO 48C-TLE has an auto-zero sequence that occurs at the end
of the hour (default). During this period, the 48C-TLE analog output will be at or close to zero, since the
detector is sampling air with the CO "scrubbed out." While this occurs the display will illustrate "ZERO."
If the operator records the data during this sequence via the analog output, then the operator must be
aware and flag this data in the DAS. The digital output via the RS-232 is flagged; therefore, no other
flagging is required. The auto-zero function can be modified from once per hour to any increment up to
once per day. It is recommended that the factory default not be changed from once per hour at this time.
4.0 QUALITY CONTROL AND QUALITY ASSURANCE
The following section has brief definitions of the QA/QC indicators. Table 4-1 has the Measurement
Quality Objectives (MQOs) of the TECO 48C-TLE. Please note that this section details primarily with
the QA indicators. Quality Control for continuous electronic instruments, such as the TECO 48C-TLE
consists of performing the diagnostic checks, maintenance and calibrations. These procedures are
detailed in sections 3.9.3 and 3.9.4: Instrument Operation, Startup and Maintenance and Calibration and
Standardization. Appendix A has an example of a Quality Control and Maintenance Record developed by
the EPA for this instrument. In addition, please review Table 4-2, which has the recommended
operation parameters for the TECO 48C-TLE. The operation parameters include recommended operating
full scale range, calibration ranges and recommended cylinder concentrations.
4.1 Precision
Precision is defined as the measure of agreement among individual measurements of the same property
taken under the same conditions. For CO, this refers to testing the CO analyzer in the field at
concentrations between 0.250 and 0.500 ppm. The test must be performed, at a minimum, once every two
weeks. Calculations for Precision can be found in Reference 4.
4.2 Bias
Bias is defined as the degree of agreement between a measured value and the true, expected, or accepted
value. Quantitative comparisons are made between the measured value and the true, standard value
during audits. Generally, three upscale points and a zero point are compared. Two audit types commonly
used for CO, direct comparison and blind, are discussed below. The SOP should discuss plans for each
type of audit.
• Direct Comparison Audits: An independent audit system is brought to the monitoring location and
produces gas concentrations that are assayed by the monitoring station's CO analyzer. In most cases,
a person outside of the agency or part of an independent QA group within the agency performs the
audit. The responses of the on-site analyzer are then compared against the calculated concentration
from the independent audit system and a linear regression is generated
• Blind Audits: In blind audits (also called performance evaluation audits) State or Local agency staff
are sent an audit device, such as the National Performance Evaluation Program (NPEP). The agency
-------
TECO 48C-TLE CO SOP
Version No. 2
2/3/2005
Page 16 of 18
staff does not know the CO concentrations produced by the audit equipment. Responses of the on-
site analyzer are then compared against those of the generator and a linear regression is calculated.
4.3 Representativeness
Representativeness refers to whether the data collected accurately reflect the conditions being measured.
It is the data quality indicator most difficult to quantify. Unless the samples are truly representative, the
other indicators are meaningless. Since the NCORE Level I and II siting criteria are urban and regional,
the CO Trace Level criteria are the same. Please reference the National Monitoring Strategy5 for a
discussion of NCORE Level II CO monitoring scale.
4.4 Completeness
Completeness is defined as the amount of data collected compared to a pre-specified target amount. For
CO, EPA requires a minimum completeness of 75% (40 CFR 50, App.H.3). Typical completeness with
the TECO 48C-TLE values can approach 90-93%.
4.5 Comparability
Comparability is defined as the process of collecting data under conditions that are consistent with those
used for other data sets of the same pollutant. The TECO 48C-TLE meets the MQOs for a Trace Level
CO instrument. Please see Table 4-1.
4.6 Method Detection Limit
The method detection limit (MDL) or detectability refers to the lowest concentration of a substance that
can be determined by a given procedure. The TECO 48C-TLE must be able to detect a minimum value of
0.040 ppm of CO.
Table 4-1 Measurement Quality Assurance Objectives
Requirement
Bias
Precision
Completeness
Representativeness
Comparability
Method Detection
Limit
Frequency
NCORE,
once per
year
1 every 2
weeks
Quarterly,
Annually
N/A
N/A
NA
Acceptance Criteria
To be Determined from Data
Quality Objectives
Concentration: 0.250 -0.500 ppm,
Coefficient of Variance: To be
determined
NCORE, 75%
Neighborhood, Urban or Regional
Scale
Must be a Trace Level instrument.
See Sections 3.1 and 3.2 of this
document.
0.040 ppm
Reference
40 CFR Pt. 5 8
40 CFR Pt. 58
Appendix A
National
Monitoring
Strategy.
40 CFR 58
National
Monitoring
Strategy.
National
Monitoring
Strategy
Information or Action
Use of NIST generated gas
concentrations with Mass
Flow Calibration unit that is
NIST traceable
To be determined
If under 75%, institute
Quality Control Measures
N/A
N/A
Testing is performed at the
factory.
-------
TECO 48C-TLE CO SOP
Version No. 2
2/3/2005
Page 17 of 18
Table 4-2 Operating Parameters for the TECO 48C-TLE Trace Gas Instrument
Item
Full Scale Range
Units
Compressed Gas
Cylinder
Calibration Ranges
a. zero
b. Level I Span
c. Mid Point Span
d. Precision Level
Range
0 to 5. 000 ppm
Part per million (ppm)
200 - 250 ppm
0- 0.010 ppm
4.000 -5. 000 ppm
2. 000 -2. 500 ppm
0.250 -0.500 ppm
Comments
Suggested Range. Reduce to 1.000 ppm if rural site
Recommended
NIST Traceable Protocol #1 cylinder with CO
concentration between 200 - 250 ppm.
There are a number of commercially available vendors.
NIST Traceable Protocol # 1 cylinder with CO
concentration between 200 - 250 ppm. Recommended
gas flow range 75 - 90 cc/min. Zero air flow 4.80 -
5.00 liters/mm.
NIST Traceable Protocol #1 cylinder with CO
concentration between 200 - 250 ppm. Recommended
gas flow range 75 - 90 cc/min. Zero air flow 8.00
10.00 liters/min.
NIST Traceable Protocol # 1 cylinder with CO
concentration between 200 - 250 ppm. Recommended
gas flow range 20 - 35 cc/min. Zero air flow 18.00 -
20.00 liters/min.
5.0 REFERENCES
1. Merck Index, twelfth edition 1996, page 296
2. Seinfeld,, John H., Atmospheric Chemistry and Physics of Air Pollution, 1986, page 54
3. Code of Federal Regulations, Title 40, Part 53.23c
4. Code of Federal Regulation, Title 40, Part 58, Appendix A
5. The National Air Monitoring Strategy, Final Draft, 4/29/04,
http: //www .epa. gov/ttn/amtic/monstratdoc .html
-------
Appendix A
Environmental Protection Agency
Monthly Quality Control and Maintenance Records
TECO 48C-TLE CO Analyzer
TECO 48C-TLE CO SOP
Version No. 2
2/3/2005
Page 18 of 18
Site Name/Location
Technician
Month/Year
Serial Number
_Range
Parameter
Program Number
Bias Voltage
+5 volt supply
+15 volt supply
-15 volt supply
Battery
Internal Temp
Chamber Temp
Pressure
Flow
S/R Ratio
AGC Intensity
Motor Speed
Test Analog
Outputs*
Option
Switches**
Date
Date
Date
Acceptance Criteria
48 TR007 00
Communications
48LTR007 00
-105-115V
NA
NA
NA
NA
8.0-47deg. C
48.0- 52.0 deg. C
250 -1000 mm Hg
0.35- 2.5 LPM
1.14-1.18
200,000 - 300,000 Hz
100%
See note below
See note below
* When the operator needs to set the analog output against the DAS, this function should be
utilized. Please refer to page 3-41 to 3-44 of the owner's manual to initiate this feature.
* *The option switches are set at the factory. Please reference owner's manual, Page 3-62
"Service Mode Menu" on changing these options switches.
Date
Comments and Notes
Figure A-l TECO 48C-TLE Quality Control and Maintenance Record
-------
STANDARD OPERATING PROCEDURES
TELEDYNE - ADVANCED POLLUTION INSTRUMENTS
MODEL 300EU TRACE LEVEL
CARBON MONOXIDE INSTRUMENT
Version 2
-------
Section 1.1 Acknowledgments
This Standard Operating Procedure (SOP) for Carbon Monoxide - Trace Level is the product of EPA's
Office of Air Quality, Planning and Standards. The following individuals are acknowledged for their
contributions.
Principal Author
Dennis K. Mikel, OAQPS-EMAD, Research Triangle Park, NC 27711
Reviewers
Office of Air Quality, Planning and Standards
Joann Rice, Trace Gas Team Lead, OAQPS-EMAD, Research Triangle Park, NC 27711
Anna Kelly, OAQPS-EMAD, Research Triangle Park, NC 27711
Keith Kronmiller, Mantech, Inc. Research Triangle Park, NC 27711
Comments and questions can be directed to:
Joann Rice
EPA-OAQPS
Emission, Analysis and Monitoring Division
Mail Drop D243-02
Research Triangle Park, NC 27711
Email: rice.joann@epa.gov
Phone: (919)-541-3372
-------
Section 2.0 Table of Contents
Section
1.0 Title Page
1.2 Acknowledgements
2.0 Table of Contents
2. 1 List of Tables and Figures
3.0 Procedures
3.1 Scope and Applicability
3.2 Summary of Method
3.3 Definitions
3.4 Health and Safety Warnings
3.5 Cautions
3.6 Interferences
3.7 Personal Qualifications
3.8 Equipment and Supplies
3.9 Procedure
3.9.1 Sample Collection
3.9.2 Sample Handling and Preservation
3.9.3 Instrument Operation, Start up and Maintenance
3.9.3.1 Startup
3.9.3.2 Operation and Range Setting
3.9.9.3 Diagnostic Checks/Manual Checks
3.9.3.4 Preventive Maintenance
3.9.3.5 Instrument Trouble shooting
3.9.4 Calibration and Standardization
3.9.4.1 Adj ustment to Zero Air
3.9.4.2 Adjustment to Calibration Gas
3.10 Data Analysis and Calculations
4.0 Quality Control and Quality Assurance
4.1 Precision
4.2 Bias
4.3 Representativeness
4.4 Completeness
4.5 Comparability
4.6 Method Detection Limit
5.0 References
Page
1
2
3
4
5
5
5
7
7
7
8
8
8
10
10
10
10
10
10
11
13
13
13
13
14
14
14
14
15
15
15
15
15
17
-------
List of Tables
Table Name and Number
3-1 Definition of Key Terms
3-2 Diagnostic Checks
3-3 Preventive Maintenance Schedule for the TAPI 300EU
4-1 Measurement Quality Assurance Objectives
4-2 Operating Parameters for the TAPI 300EU Trace Gas Instrument
Page
7
12
13
16
17
List of Figures
Figure Name and Number
3-1 Schema of the Teledyne API 300EU
A-l Teledyne API Quality Control and Maintenance Record
Page
9
18
-------
Teledyne API CO SOP
Version No. 2
2/3/2005
Page 5 of 18
STANDARD OPERATING PROCEDURES
TELEDYNE - ADVANCED POLLUTION INSTRUMENTS
MODEL 300EU TRACE LEVEL CARBON MONOXIDE INSTRUMENT
3.0 PROCEDURES
3.1 Scope and Applicability
Carbon Monoxide (CO), a colorless, odorless, tasteless, highly poisonous gas has detrimental effects on
human health. Carbon Monoxide originates from the partial oxidation of hydrocarbon fuels, coal and
coke1. Carbon Monoxide affects the oxygen carrying capacity of the blood. Carbon Monoxide can
diffuse through the alveolar walls of the lungs and compete with oxygen for one of the four iron sites in
the hemoglobin molecule. The affinity of the iron site for CO is approximately 210 times greater than
oxygen2. Low levels of CO can cause a number of symptoms including headache, mental dullness,
dizziness, weakness, nausea, vomiting and loss of muscular control. In extreme cases, collapse,
unconsciousness and death can occur.
The Teledyne - Advanced Pollution Instruments (TAPI) model 300EU is a state of the science instrument
for the determination of trace levels of Carbon Monoxide by Non-Dispersive Infrared Spectrophotometry
(NDIR) using Gas Filter Correlation (GFC). This SOP will detail the operation, preventive maintenance,
cautions and health warnings.
The Detection Limit (DL) for non-trace levels of CO is 1.0 parts per million (ppm) (Code of Federal
Regulations, Volume 40, Part 53.23c, or, in the shortened format used hereafter, 40 CFR 53.23c)3.
However, the 300EU has a DL to 20 parts per billion (ppb), which is accomplished by modifications to
the Federal Reference Method (FRM) instruments. This document will discuss the Trace Level (TL)
operating procedures in detail.
3.2 Summary of Method
The analytical principle is based on absorption of IR light by the CO molecule. NDIR-GFC analyzers
operate on the principle that CO has a sufficiently characteristic IR absorption spectrum such that the
absorption of IR by the CO molecule can be used as a measure of CO concentration in the presence of
other gases. Carbon Monoxide absorbs IR maximally at 2.3 and 4.6 um. Since NDIR is a
spectrophotometric method, it is based upon the Beer-Lambert Law. The degree of reduction depends on
the length of the sample cell, the absorption coefficient, and CO concentration introduced into the sample
cell, as expressed by the Beer-Lambert law shown below:
T = I/Io = e('axc) (equation 1)
where:
T = Transmittance of light through the gas to the detector
I = light intensity after absorption by Carbon Monoxide
lo = light intensity at zero Carbon Monoxide concentration
a = specific Carbon Monoxide molar absorption coefficient
x = path length, and
C = Carbon Monoxide concentration
-------
Teledyne API CO SOP
Version No. 2
2/3/2005
Page 6 of 18
For Gas Filter Correlation, there is only one sample cell. This cell acts as the sample and reference cell.
The broad band of IR radiation is emitted from an IR source. The IR light passes through a very narrow
band pass filter which screens out most wavelengths and allows only the wavelength of light that CO
absorbs to enter the sample cell. The GFC analyzer has a chopper wheel with two pure gases: Nitrogen
and Carbon Monoxide. As the chopper wheel rotates and allows the IR energy to enter "CO side" of the
wheel, all IR energy that could be absorbed by CO in the sample stream is absorbed by the CO in the
wheel. This technique effectively "scrubs out" any light that could possibly be attenuated. The single
detector records the light level (lo). As the wheel spins, the "N2 side" of the wheel reaches the IR energy
beam. This side of the wheel allows all IR light to pass through the wheel and be absorbed by any CO
that might be in the sample gas. This light level is CO sensitive (I). The detector records the attenuation
of this light, compares the two light levels (I/Io) and sends a signal to the electrometer board that
calculates the concentration. The voltage is related to the CO concentration according to the Beer-
Lambert law in equation 1 shown above. Thus, TAPI 300EU can measure CO continuously. The 300EU
version has four distinct features that allow it to measure CO at ppb levels:
• The sample stream is dried using permeation or Nafion™ Dryer;
• The analyzer baseline is determined and corrected using heated palladium catalytic converter;
• The baseline is frequently auto-zeroed, at a minimum once per hour, through the palladium converter;
• The instrument has an ultra-sensitive or "hot" detector.
The 300 EU instrument operates in the following fashion:
1. In sample mode, ambient air is allowed to enter through the rear bulkhead sample port. Solenoid #1
is in its Normally Open (NO) mode. The ambient air flows through the solenoid to the permeation
dryer, which removes the moisture and water from the sample air stream.
2. The sample stream then passes through a sample filter, which removes particles that can build up on
the mirror and sample chamber and attenuate the IR beam.
3. The sample then enters the sample cell. A major difference between a non-TL and TL instrument is
the detector. The TL instrument has a detector that is more sensitive to the light emitted and
absorbed in the sample cell. This detector must be more sensitive because the amount of attenuation
by the CO gas in the sample stream is much lower. Therefore, the detector must be sensitive at lower
ambient levels. Temperature of the sample cell is also critical. The sample cell and detector must be
maintained at a constant temperature in order for the detector to keep a stable background.
Fluctuations of more than 1° Centigrade can cause the baseline to drift, giving false readings at low
levels.
4. The detector sends the signal to the demodulator which interprets the signal. The demodulator sends a
digital value to the Central Processor Unit (CPU).
5. At the end of the hour, the CPU sends a voltage signal to the Solenoid #1 and switches it to the
"Normally Closed" (NO position. This allows room air to be drawn into the instrument and to pass
through the catalytic converter. The catalytic converter uses a palladium bed heated to 50°
Centrigrade to convert all Carbon Monoxide to Carbon Dioxide (2CO + O? —> 2CO2). This
effectively "scrubs" all CO from the sample stream. The CO "free" air flows through the sample cell
and the CPU interprets the signal from the demodulator as the "background" or "baseline" value.
The baseline is then adjusted at that time. The baseline adjustment usually takes between 7-10
minutes.
6. The CPU then switches Solenoid #1 to its NO position and ambient air then drawn into the analyzer.
-------
Teledyne API CO SOP
Version No. 2
2/3/2005
Page 7 of 18
3.3 Definitions
Here are some key terms for this method.
Table 3-1, Definitions of Key Terms
Term Definition
DAS Data acquisition system. Used for automatic collection and
recording of CO concentrations.
Interferences Physical or chemical entities that cause CO measurements to be
higher (positive) or lower (negative) than they would be without the
entity. (See Section 3.6).
3.4 Health and Safety Warnings
To prevent personal injury, please heed these warnings concerning the 300EU.
1. Carbon Monoxide is a poisonous gas. Vent any CO or calibration span gas to the atmosphere
rather than into the shelter or other sampling area. If this is impossible, limit exposure to CO by
getting fresh air every 5 to 10 minutes. If the operator experiences light headedness, headache or
dizziness, leave the area immediately.
2. The IR source is a filament resistor that has an electrical current running through it. The IR source
can become very hot. When troubleshooting, allow the instrument to cool off especially if you
suspect the IR source as the cause of trouble.
3. Always use a third ground wire on all instruments.
4. Always unplug the analyzer when servicing or replacing parts.
5. If it is mandatory to work inside an analyzer while it is in operation, use extreme caution to avoid
contact with high voltages. The analyzer has a 110 volt Volts Alternating Current (VAC) power
supply. Refer to the manufacturer's instruction manual and know the precise locations of the VAC
components before working on the instrument.
6. Avoid electrical contact with jewelry. Remove rings, watches, bracelets, and necklaces to prevent
electrical bums.
3.5 Cautions
To prevent damage to the 300EU, all cautions should immediately precede the applicable step in
this SOP. The following precautions should be taken:
1. Normally, if Teflon™ filters are used in the sample train, cleaning the optical bench will not be
required. However, in the event that the bench is cleaned, be careful to avoid damaging the interior of
the sample chamber. In addition, some GFC instruments have a series of mirrors that deflect the light
in order to increase the path length. The mirrors are aligned at the factory. If the mirrors become
misaligned, the IR light beam will not be directed to the detector. Use extreme caution when cleaning
or servicing the sample chamber(s). In addition the mirrors are very fragile. Avoid dropping the
instrument. This may damage, misalign or crack the mirrors and cause expensive repairs.
2. Keep the interior of the analyzer clean.
3. Inspect the system regularly for structural integrity.
4. To prevent major problems with leaks, make sure that all sampling lines are reconnected after
required checks and before leaving the site.
-------
Teledyne API CO SOP
Version No. 2
2/3/2005
Page 8 of 18
5. Inspect tubing for cracks and leaks.
6. It is recommended that the analyzer be leak checked after replacement of any pneumatic parts.
7. If cylinders are used in tandem with Mass Flow Control (MFC) calibrators, use and transport is a
major concern. Gas cylinders can sometimes contain pressures as high as 2000 pounds per square
inch (psi). Handling of cylinders must be done in a safe manner. If a cylinder is accidentally dropped
and valve breaks off, the cylinder can become explosive or a projectile.
8. Transportation of cylinders is regulated by the Department of Transportation (DOT). It is strongly
recommended that all agencies contact the DOT or Highway Patrol to learn the most recent
regulations concerning transport of cylinders.
9. Carbon Monoxide is a highly poisonous gas. Long term exposure can cause problems with motor
coordination and mental acuity. It is strongly recommended that all agencies have Material Safety
Data Sheets (MSDS) at all locations where CO cylinders are stored or used. MSDS can be obtained
from the DOT or from your vendor.
10. It is possible (and practical) to blend other compounds with CO. If this is the case, it is recommended
that MSDS for all compounds be made available to all staff that use and handle the cylinders or
permeation tubes.
11. Shipping of cylinders is governed by the DOT. Contact the DOT or your local courier about the
proper procedures and materials needed to ship high-pressure cylinders.
3.6 Interferences
Water Vapor: Studies have shown conclusively that NDIR analyzers have interference from water
vapor. Water absorbs very strongly across several bands of IR spectra. Water vapor interference occurs
because water vapor absorption of light in the region of 3.1, 5.0 5.5 and 7.1 -10.0 um in the IR region.
Since water vapor absorbs light in this region, this has a quenching effect on the reaction of CO. The
TAPI 3000EU is equipped with a Nafion™ drier, which effectively scrubs all water and water vapor. No
maintenance is required on the dryer.
Carbon Dioxide: Carbon dioxide absorbs in the IR spectrum at 2.7, 5.2, and 8.0 to 12.0 um. This is very
close to the regions that CO absorbs within as well. However, since atmospheric CO2 is much higher in
concentration than CO, this UV spectral range must be avoided. To prevent light in this spectral region,
the TAPI 300EU analyzer has a band pass filter that blocks these wavelengths.
3.7 Personal Qualifications
The person(s) chosen to operate the TAPI 300EU should have a minimum of qualifications. The
understanding of basic chemistry and electronics are a must. The understanding of digital circuitry is
helpful, but not required. Also, courses in data processing and validation are also welcome.
3.8 Equipment and Supplies
Monitoring Apparatus: The design of the 300EU is identical to the 300E with several major variations.
A diagram of the TAPI 300EU instrument is described in Figure 3-1. The three main components are:
• Pneumatic System: Consists of sample inlet line, particulate filter, permeation dryer, reaction
chamber, flowmeter, and pump, all used to bring ambient air samples to the analyzer inlet.
• Analytical System: This portion of the instrument consists of the IR source, the correlation wheel,
motor, mirrors, detector and band pass filter.
-------
Teledyne API CO SOP
Version No. 2
2/3/2005
Page 9 of 18
• Electronic Hardware: The part of the analyzer that generally requires little or no maintenance. If the
300EU is operated above the manufacturer's recommended temperature limit, however, individual
integrated chips can fail and cause problems with data storage or retrieval.
Other apparatus and equipment includes the following.
Instrument Shelter: A shelter is required to protect the analyzer from precipitation and adverse weather
conditions, maintain operating temperature within the analyzer's temperature range requirements, and
provide security and electrical power. The recommended shelter temperature range is 20-30°C.
Spare Parts and Incidental Supplies: See the TAPI 300EU operating manual, Appendix B for specific
maintenance and replacement requirements.
Calibration System: A system that creates concentrations of CO of known quality is necessary for
establishing traceability. This is described in detail in the "Environics 9100 SOP." Please reference this
document.
DAS: A data acquisition system is necessary for storage of ambient and ancillary data collected by the
300EU.
IE Source
Sanple Cell
^ Filter
Wheel
Sanfile
(jgy^y
Permeation.
ryer
-
(
/
\
Span Gas
Display
analog ES-232
Output Digital
Output
Figure 3-1 Schema of the Teledyne API 300EU
Wiring, Tubing and Fittings: Teflon™ and borosilicate glass are inert materials that should be used
exclusively throughout the intake system. It is recommended that Polytetrafluoroethylene (PTFE) or
Fluoroethylpropylene (FEP) Teflon™ tubing be used. PTFE and FEP are the best choice for the
connection between the intake manifold and the 300EU bulkhead inlet. Examine and discard if
particulate matter collects in the tubing. All fittings and ferrules should be made of Teflon™ or stainless
-------
Teledyne API CO SOP
Version No. 2
2/3/2005
Page 10 of 18
steel. Connection wiring to the DAS should shielded two strand wire or RS-232 cables for digital
connections.
Reagents and Standard: The TAPI 300EU does not require any reagents since the instrument uses
photometry to analyze for CO. All standards for the CO method can be obtained in compressed cylinders
and must be NIST traceable. Please see the "Calibration of Trace Gas Analyzers" SOP.
3.9 Procedure
3.9.1 Sample Collection: Sampling for Trace Level CO is performed by continuously drawing ambient
air through a sample manifold directly into the analyzer via a vacuum pump. All inlet materials must be
constructed from Teflon™ or borosilicate glass as detailed in 40 CFR 58. The siting criteria for CO
Trace Level instruments in detailed in 40 CFR 58, Appendix A4.
3.9.2 Sample Handling and Preservation: Carbon Monoxide samples receive no special preparation
prior to analysis. Therefore this SOP does not have a section on Sample Handling and Preservation.
3.9.3 Instrument Operation, Startup and Maintenance
This section discusses startup, operation and maintenance of the 300EU. The TAPI 300EU series
instrument has a digital front panel screen with toggle switches below. This allows the user to check
functions, switch operating parameters, adjust zero and span and read warnings messages. It is
extremely important that the user familiarize themselves with the menus available. Inadvertently
changing parameters within the analyzer can damage the instrument and possibly invalidate data
as well. Please reference the TAPI 300EU owner's manual and read it carefully before adjusting
any parameters that are set by the factory.
3.9.3.1 Start up
Before the instrument is operated, inspect the instrument for any damage. If damage is observed to the
shipping box or the instrument, contact your shipping personnel.
Carefully remove the cover and check for internal damage. Please see Section 3.1 of the TAPI 300EU
manual.
Remove the 6 red shipping screws that hold down the internal bench and parts. See Section 3.1 of the
TAPI 300EU manual.
Once you have removed the shipping screws and performed your inspection, replace the cover.
Plug the instrument into a grounded power strip that has surge protection. It is also advisable to purchase
an Uninterrupted Power Supply (UPS). An UPS will protect the 300EU from power surges and keep
the unit operating until an operator can shut it down.
Check to see that the 300EU has enough clearance so that it gets proper ventilation. Check the TAPI
300EU manual Section 3.1.
Connect the output of the analog to a DAS via shielded two wire cable. Please see EPA SOP on "Data
Management" for details.
Connect the digital RS-232 port to an appropriate cable and connect it to the DAS. Please see EPA SOP
on "Data Management" for further details.
Connect the sample inlet port to the station intake manifold.
Press the power rocker switch to "ON."
-------
Teledyne API CO SOP
Version No. 2
2/3/2005
Page 11 of 18
3.9.3.2 Operation and Range Setting
1. The exhaust fan will start and the display will come on. The Central Processing Unit (CPU) will boot
the system and load the firmware. You will see in the upper right hand corner that the fault warning
light will be flashing red. This is letting you know that the analyzer has been off.
2. To clear the fault warning, press the "CLR" button below the display. This will clear the warning and
the sample green light will flash. If the red fault light continues to flash, and clearing it does not
change this condition, then reference section 6.2.1 of the TAPI 300EU manual for instruction.
3. Once the red fault light is cleared, the operator will see the main menu. At this time, the time of day
and date must be verified and reset if necessary.
4. From main menu press the toggle button under the "SETUP" label.
5. In this menu, you will see "8 1 8" on the bottom of the display. This is the default password. In
addition, you will see on the top of the display the words "ENTER SETUP PASS:818."
6. At this time, press the toggle under "ENTR."
7. This will bring up the "PRIMARY SETUP MENU" screen.
8. In this menu, press the toggle button under "CLK."
9. In the next menu, press the "TIME" toggle switch.
10. In the next menu, you will see the time above 3 or 4 toggle switches. Adjust these toggles so that the
time is correct. Press the "ENTR" toggle switch.
11. This returns you to the "TIME OF DAY CLOCK" menu. Press the toggle switch under "DATE."
12. This will put you in the date menu. You should see the day (digits), month and year (digits) above 5
toggle switches. Adjust these toggles until the correct date is obtained. Press the "ENTR" toggle
switch. Press the "EXIT" toggle switch twice. This returns you to the main menu.
13. The range should be illustrated in the top middle of the main menu. This value should be set to 5000
ppb. If it is not set to this range, then it must be reset.
14. To change the range of the instrument, press the toggle under "SETUP." If the password is correct,
then press "ENTR."
15. In this menu, you will see "8 1 8" on the bottom of the display. This is the default password. In
addition, you will see on the top of the display the words "ENTER SETUP PASS:818."
16. At this time, press the toggle under "ENTR."
17. This will bring up the "PRIMARY SETUP MENU" screen.
18. Press the toggle switch under the "RNGE."
19. This is the "RANGE CONTROL MENU." Press the toggle switch under the "UNIT."
20. This display will show the range options. Press the toggle under the "PPB." Press the toggle under
the "ENTR."
21. This will put you into the "RANGE CONTROL MENU." Press the toggle switch under the "SET."
22. This display will show the full scale range value. Press the toggles under the digits to adjust the
instrument to the full scale value desired. Press the toggle under the "ENTR."
23. Press the "EXIT" toggle switch twice.
24. The instrument is now set with the appropriate time, date and full scale range.
25. It is recommended that you allow the 300EU 24 hours before you attempt function checks or
calibration.
26. If the DAS system that you have does not have the RS-232 capabilities, then proceed to the next
section, Diagnostic Checks/Manual Checks. If you have connected the 300EU to a computer or DAS,
review the Diagnostic Check from your computer screen. TAPI offers API.COM, a computer
program that allows the operator to log the diagnostic data that is collected by the 300EU CPU.
Several DAS manufacturers offer this type of software as well.
-------
3.9.3.3 Diagnostic Checks/Manual Checks
Teledyne API CO SOP
Version No. 2
2/3/2005
Page 12 of 18
To determine whether the 300EU is working properly, the field operators should perform the Diagnostic
Checks every time they visit the monitoring station. It is good practice for the operator to check these
Diagnostic Checks either by the computer or manually. Below are instructions on how to perform this
manually. Please note that the TAPI 300EU has set upper and lower ranges for some of these Diagnostic
checks. Please reference the owner's manual for these ranges.
1. If you observe the display, it should show "Sample" in the left hand corner, "Range" in the middle of
the display and "CO= XX.XX." Below this line there should be one line that read "",
"CAL" and "SETUP."
2. There is a series of toggle switches/ buttons below the display. These correspond to the bottom row
of the display.
3. If you press the button below the left hand "," will allow you to access the function check tree
in the opposite direction.
4. Toggle through the function check tree. The following table illustrates the functions that should be
recorded. Please see section 6.2.2 of the TAPI 300EU manual for more details. A manual check list
on maintenance check sheet is attached in Appendix A of this SOP.
Table 3-2 Diagnostic Checks
Check
Range
Stabil
CO Meas
CORef
MR ratio
Azero ratio
Sample Pres
Sample FL
Sample Temp
Bench Temp
Wheel Temp
Box Temp
PHT Drive
Slope
Offset
Time
Explanation
The full scale range of the instrument
The standard deviation of CO concentrations for the last 25 readings
The demodulated peak of the IR detector output on the measure side of the wheel
The demodulated peak of the IR detector on the reference side of the wheel
The result of the CO meas/CO Ref
The result of the CO meas/CO Ref during the Azero cycle
The absolute pressure of the sample gas in the sample chamber
Mass flow rate of sample air
The temperature of the gas inside the sample chamber
Optical bench temperature
Filter wheel temperature
The temperature inside the instrument chassis.
The voltage needed to the thermoelectric coolers of the IR detector board
The sensitivity of the instrument as calculated during the last calibration.
The overall offset of the instrument calculated during the last calibration
Displays current time.
Once the Diagnostic checks have been established and recorded for the 300EU, it is time to calibrate the
instrument. Please refer to section 3.9.4 of this SOP.
3.9.3.4 Preventive Maintenance
Preventive maintenance should prevent down-time and data loss. Table 3.3 lists the preventive
maintenance items that are listed in the Model 300EU manual, section 9.1.
-------
Teledyne API CO SOP
Version No. 2
2/3/2005
Page 13 of 18
Table 3-3 Preventive Maintenance Schedule the TAPI 300EU
Item
Replace particle filter
Verify Test Function
Perform Level I calibration
Pump Diaphragm
Perform Leak Check
Inspect Pneumatic Lines
Clean inside of Chassis
Rebuild or replace pump
Replace IR source
Clean optic bench
Replace wheel motor
Replace gases in correlation wheel
Schedule
Weekly
Weekly
Daily
Bi-annually
Annually
Annually
As needed
As needed
As needed
As needed
As needed
As needed
3.9.3.5 Instrument Troubleshooting
The TAPI 300EU manual has an excellent troubleshooting guide in Section 9.2. Please reference page
143 of the manual for details on using the Test Functions for predicting failures.
3.9.4 Calibration and Standardization
The calibration of the TAPI 300EU is performed by comparing the output of the instrument against
standardized gases of known quality. Generation of these gases is detailed in the "Calibration of Trace
Gas Analyzers" SOP. This section will detail how to adjust the 300EU to the standardized gases. Once
the calibration has been performed, compare the response of your DAS to the calculated "source" value.
If this is outside of +/-10%, then adjust the instrument response as detailed in the next sections.
3.9.4.1 Adjustment to Zero Air
In order to adjust the output of the 300EU to zero air, perform the following:
1. Allow the instrument to sample zero air from a manifold that is at near atmospheric pressure for a
minimum of 15 minutes.
2. On the bottom of the front panel screen there is a toggle switch/ button that is beneath the "CAL"
label. Press this button.
3. This next screen is the "M-P CAL" screen. In this screen press the button below the "ZERO"
label.
4. The next screen will show an "ENTR, SPAN and CONC" above the toggle switches on the
bottom of the panel. Press the button below the "ENTR" label. This operation changes the
calculation equation and zeros the instrument.
5. Press the button below the "EXIT" label. This returns the operator to the main "SETUP" menu.
3.9.4.2 Adjustment to Calibration Gas
In order to adjust the output of the 300EU to NIST traceable calibration gas, perform the following:
-------
Teledyne API CO SOP
Version No. 2
2/3/2005
Page 14 of 18
1. Switch the calibration unit to generate a known concentration of CO. Allow the instrument to
sample calibration gas from a manifold that is at near atmospheric pressure for a minimum of
15 minutes.
2. On the bottom of the front panel screen of the main menu there is a button that is beneath the
"CAL" label. Press this button.
3. This next screen is the "M-P CAL" screen. The next screen will show an "ENTR, SPAN and
CONC" above the toggle switches on the bottom of the panel. Press the button below the
"CONC" label. On the bottom line, there will be digits below each toggle button. In order to
change the concentration, toggle each digit before and after the decimal place to get the
concentration that is being generated in the manifold by the calibrator. At this time, press the
"ENTR."
4. At this time, press the toggle below the "SPAN" switch.
5. This operation changes the calculation equation and adjusts the slope of the instrument.
6. Press the button below the "EXIT" label. This returns the operator to the main "SETUP"
menu.
3.10 Data Analysis and Calculations
Data analysis for this analyzer is detailed in the "Data Acquisition and Management" SOP. For the TAPI
300EU, there is one design detail of which the operator must be aware; the auto-zero (Azero) function.
As detailed in Section 3.1, the TAPI 300EU has an Azero sequence that occurs at the end of the hour.
During this period, the 300EU "freezes" the output to the CPU at the last value calculated by the CPU and
the display will illustrate "AZERO." During the auto-zero sequence, the display and analog output are
"frozen" on one value. If the operator records the data via the analog output, then the operator must be
aware of this sequence and flag this data in the DAS. The digital output via the RS-232 is flagged;
therefore, no other flagging is required. The Azero function can be modified from once per hour to any
increment up to once per day. It is recommended that the factory default not be changed from once per
hour at this time.
4.0 QUALITY CONTROL AND QUALITY ASSURANCE
The following section has brief definitions of the QA/QC indicators. Table 4-1 has the Measurement
Quality Objectives (MQOs) of the TAPI 300EU. Please note that this section deals primarily with the
data quality indicators. Quality Control for continuous electronic instruments, such as the TAPI 300EU
consists of performing the diagnostic checks, maintenance and calibrations. These procedures are
detailed in sections 3.9.3 and 3.9.4: Instrument Operation, Startup and Maintenance and Calibration and
Standardization. Appendix A has an example of a Quality Control and Maintenance Record developed by
the EPA for this instrument.
4.1 Precision
Precision is defined as the measure of agreement among individual measurements of the same property
taken under the same conditions. For CO, this refers to testing the CO analyzer in the field at
concentrations between 0.250 and 0.500 ppm (250 - 500 ppb). The test must be performed, at a
minimum, once every two weeks. Calculations for Precision can be found in Reference 4.
-------
Teledyne API CO SOP
Version No. 2
2/3/2005
Page 15 of 18
4.2 Bias
Bias is defined as the degree of agreement between a measured value and the true, expected, or accepted
value. Quantitative comparisons are made between the measured value and the true, standard value
during audits. Generally, three upscale points and a zero point are compared. Two audit types commonly
used for CO, direct comparison and blind, are discussed below. The SOP should discuss plans for each
type of audit.
• Direct Comparison Audits: An independent audit system is brought to the monitoring location and
produces gas concentrations that are assayed by the monitoring station's CO analyzer. In most cases,
a person outside of the agency or part of an independent QA group within the agency performs the
audit. The responses of the on-site analyzer are then compared against the calculated concentration
from the independent audit system and a linear regression is generated
• Blind Audits: In blind audits (also called performance evaluation audits), State or Local Agency
staff are sent an audit device, such as done in the National Performance Audit Program (NPAP). The
agency staff does not know the CO concentrations produced by the audit equipment. Responses of
the on-site analyzer are then compared against those of the audit device and a linear regression is
calculated.
4.3 Representativeness
Representativeness refers to whether the data collected accurately reflect the conditions being measured.
It is the data quality indicator most difficult to quantify. Unless the samples are truly representative, the
other indicators are meaningless. Since the NCORE Level I and II siting criteria are urban and regional,
the TL-CO criteria are the same. Please reference the National Monitoring Strategy5 for a discussion of
NCORE Level II CO monitoring scale.
4.4 Completeness
Completeness is defined as the amount of data collected compared to a pre-specified target amount. For
CO, EPA requires a minimum completeness of 75% (40 CFR 50, App.H.3). Typical completeness with
the TAPI 300EU values can approach 90-93%.
4.5 Comparability
Comparability is defined as the process of collecting data under conditions that are consistent with those
used for other data sets of the same pollutant. The TAPI 300EU meets the MQOs for a TL-CO
instrument. Please see Table 4-1.
4.6 Method Detection Limit
The method detection limit (MDL) or detectability refers to the lowest concentration of a substance that
can be determined by a given procedure. The TAPI 300EU must be able to detect a minimum value of
0.020 ppm of CO.
-------
Teledyne API CO SOP
Version No. 2
2/3/2005
Page 16 of 18
Table 4-1 Measurement Quality Assurance Objectives
Requirement
Bias
Precision
Completeness
Representativene
ss
Comparability
Method
Detection Limit
Frequency
NCORE,
once per
year
1 every 2
weeks
Quarterly,
Annually
N/A
N/A
NA
Acceptance Criteria
To be Determined from
Data Quality Objectives
Concentration: 0.250 -
0.500 ppm, Coefficient of
Variance: To be
determined
NCORE, 75%
Neighborhood, Urban or
Regional Scale
Must be a Trace Level
instrument. See Sections
3.1 and 3.2 of this
document.
0.020 ppm
Reference
40CFRR58
40CFRR58
Appendix A
National
Monitoring
Strategy.
40CFR58
National
Monitoring
Strategy.
National
Monitoring
Strategy
Information or Action
Use of NIST generated
gas concentrations with
Mass Flow Calibration
unit that is NIST
traceable
To be determined
If under 75%, institute
Quality Control
Measures
N/A
N/A
Testing is performed at
the factory.
-------
Teledyne API CO SOP
Version No. 2
2/3/2005
Page 17 of 18
Table 4-2 Operating Parameters for the TAPI 300EU Trace Gas Instrument
Item
Full Scale Range
Units
Compressed Gas
Cylinder
Calibration Ranges
a. zero
b. Level I Span
c. Mid Point Span
d. Precision Level
Range
0 to 5000 ppb
Part per billion (ppb)
200 - 250 ppm
0- 10 ppb
4000 - 5000 ppb
2000 - 2500 ppb
250 - 500 ppb
Comments
Suggested Range. Reduce to 1000 ppb if rural site
Recommended
NIST Traceable Protocol #1 cylinder with CO
concentration between 200 - 250 ppm.
There are a number of commercially available vendors.
NIST Traceable Protocol # 1 cylinder with CO
concentration between 200 - 250 ppm. Recommended
gas flow range 75 - 90 cc/min. Zero air flow 4.80 -
5.00 liters/mm.
NIST Traceable Protocol #1 cylinder with CO
concentration between 200 - 250 ppm. Recommended
gas flow range 75 - 90 cc/min. Zero air flow 8.00
10.00 liters/min.
NIST Traceable Protocol # 1 cylinder with CO
concentration between 200 - 250 ppm. Recommended
gas flow range 20 - 35 cc/min. Zero air flow 18.00 -
20.00 liters/min.
5.0 REFERENCES
1. Merck Index, twelfth edition 1996, page 296
2. Seinfeld,, John H., Atmospheric Chemistry and Physics of Air Pollution, 1986, page 54
3. Code of Federal Regulations, Title 40, Part 53.23c
4. Code of Federal Regulation, Title 40, Part 58, Appendix A
5. The National Air Monitoring Strategy, Final Draft, 4/29/04,
http: //www .epa. gov/ttn/amtic/monstratdoc .html
-------
Appendix A
Environmental Protection Agency
Monthly Quality Control and Maintenance Records
Teledyne API 300EU CO Analyzer
Teledyne API CO SOP
Version No. 2
2/3/2005
Page 18 of 18
Site Name/Location
Technician
Month/Year
Serial Number
_Range
Parameter
Offset
Slope
PHT Drive
Box Temp
Wheel Temp
Bench Temp
Sample Temp
Sample Flow
Sample Pressure
A-zero Ratio
MR Ratio
CO Reference
CO Measured
Stability
Range
Other Tests
Dark Current
Date
Date
Date
Acceptance
Criteria
Date
Comments and Notes
Figure A-l Teledyne API 300 EU Quality Control and Maintenance Record
-------
STANDARD OPERATING PROCEDURES
TELEDYNE - ADVANCED POLLUTION INSTRUMENTATION
MODEL 100AS TRACE LEVEL
SULFUR DIOXIDE INSTRUMENT
Version 2
-------
Section 1.1 Acknowledgments
This Standard Operating Procedure (SOP) for Sulfur Dioxide - Trace Level is the product of EPA's
Office of Air Quality, Planning and Standards. The following individuals are acknowledged for their
contributions.
Principal Author
Solomon Ricks, OAQPS-EMAD, Research Triangle Park, NC 27711
Reviewers
Office of Air Quality, Planning and Standards
Joann Rice, Trace Gas Team Lead, OAQPS-EMAD, Research Triangle Park, NC 27711
Anna Kelly, OAQPS-EMAD, Research Triangle Park, NC 27711
Keith Kronmiller, Mantech, Inc. Research Triangle Park, NC 27711
Comments and questions can be directed to:
Solomon Ricks
EPA-OAQPS
Emission, Analysis and Monitoring Division
Mail Drop D339-02
Research Triangle Park, NC 27711
Email: ricks.solomon(giepa.gov
Phone: (919)-541-5242
-------
Section 2.0 Table of Contents
Section
1.0 Title Page
1.3 Acknowledgements
2.0 Table of Contents
2. 1 List of Tables and Figures
3.0 Procedures
3.1 Scope and Applicability
3.2 Summary of Method
3.3 Definitions
3.4 Health and Safety Warnings
3.5 Cautions
3.6 Interferences
3.7 Personal Qualifications
3.8 Equipment and Supplies
3.9 Procedure
3.9.1 Sample Collection
3.9.2 Sample Handling and Preservation
3.9.3 Instrument Operation, Start up and Maintenance
3.9.3.1 Startup
3.9.3.2 Operation and Range Setting
3.9.9.3 Diagnostic Checks/Manual Checks
3.9.3.4 Preventive Maintenance
3.9.3.5 Instrument Trouble shooting
3.9.4 Calibration and Standardization
3.9.4.1 Adj ustment to Zero Air
3.9.4.2 Adjustment to Calibration Gas
3.10 Data Analysis and Calculations
4.0 Quality Control and Quality Assurance
4.1 Precision
4.2 Bias
4.3 Representativeness
4.4 Completeness
4.5 Comparability
4.6 Method Detection Limit
5.0 References
Page
1
2
3
4
5
5
5
6
6
7
7
8
8
9
9
9
9
10
10
11
12
13
13
13
13
14
14
14
14
15
15
15
15
16
-------
List of Tables
Table Name and Number
3-1 Definition of Key Terms
3-2 Diagnostic Checks
3-3 Preventive Maintenance Schedule for the T-API 100AS
4-1 Measurement Quality Assurance Objectives
Page
6
12
13
16
List of Figures
Figure Name and Number
3-1 T-API Model 100AS Flow Diagram
A-l T-API Model 100AS Quality Control and Maintenance Record
Page
9
17
-------
Teledyne-API SO2 SOP
Version No. 2
4/26/2005
Page 5 of 17
STANDARD OPERATING PROCEDURES
TELEDYNE - ADVANCED POLLUTION INSTRUMENTATION
MODEL 100AS TRACE LEVEL SULFUR DIOXIDE INSTRUMENT
3.0 PROCEDURES
3.1 Scope and Applicability
Sulfur Dioxide (SO2) is a colorless, nonflammable gas that has a strong suffocating odor. SO2 originates
from fuel containing sulfur (mainly coal and oil) burned at power plants and during metal smelting and
other industrial processes. High levels of SO2 can result in temporary breathing impairment for asthmatic
children and adults who are active outdoors. Long-term exposure to high levels of SO2, in the presence of
high levels of particulate matter, may aggravate existing cardiovascular disease and respiratory illness.
The Teledyne-Advanced Pollution Instrumentation (T-API) model 100AS combines proven detection technology for
the determination of trace levels of SO2. This SOP will detail the operation, preventive maintenance, cautions and
health warnings.
The Detection Limit (DL) for a non-trace level SO2 analyzer is 10 parts per billion (ppb) (Code of Federal
Regulations, Volume 40, Part 53.23c, or, in the shortened format used hereafter, 40 CFR 53.23c)3.
However, the T-API model 100AS has an estimated DL of 100 parts per trillion (ppt), which is
accomplished by an increased detector sensitivity, as well as increasing the length of the standard
instrument's optics bench. This document will discuss the Trace Level (TL) operating procedures in
detail.
3.2 Summary of Method
The Model 100AS Trace Level operating principle is based on measuring the emitted fluorescence of SO2
produced by the absorption of ultraviolet (UV) light. The UV lamp emits ultraviolet radiation which
passes through a 214 nm band pass filter, excites the SO2 molecules, producing fluorescence which is
measured by a photomultiplier tube (PMT) with a second UV band pass filter. SO2 absorbs in the 190 nm
- 230 nm region free of quenching by air and relatively free of other interferences. The equations
describing the above reactions are as follows:
SO2 + hvjIa -> SO2*
The excitation ultraviolet light at any point in the system is given by:
Ia=I0[l-exp(-ax(SO2))]
where:
I0 = UV light intensity,
a = the absorption coefficient of SO2
x = the path length
SO2 = concentration of SO2
-------
Teledyne-API SO2 SOP
Version No. 2
4/26/2005
Page 6 of 17
The excited SO2 decays back to the ground state emitting a characteristic fluorescence:
SO2* "*-> SO2 + hv2
When the SO2 concentration is relatively low, the path length of exciting light is short and the background
is air, the above expression reduces to:
F = k(SO2)
where:
F = amount of fluorescent light given off
k = rate at which SO2* decays into SO2
The 100AS instrument operates in the following fashion:
1. In sample mode, the sample is drawn into the analyzer through the SAMPLE bulkhead. The sample
flows through a hydrocarbon "kicker." which operates on a selective permeation principle, allowing
only hydrocarbon molecules to pass through the tube wall. The driving force for the hydrocarbon
removal is the differential partial pressure across the wall. This differential pressure is produced
within the instrument by passing the sample gas through a capillary tube to reduce its pressure and
feeding it into the shell side of the hydrocarbon kicker. The SO? molecules pass through the
hydrocarbon "kicker" unaffected.
2. The sample flows into the fluorescence chamber, where UV light is focused through a narrow 214 nm
band pass filter into the reaction chamber, exciting the SO? molecules; the molecules then give off
their characteristic decay radiation. A second filter allows only the decay radiation to fall on the
PMT. The PMT transfers the light energy into the electrical signal which is directly proportional to
the light energy in the sample stream being analyzed. The preamp board converts this signal into a
voltage which is further conditioned by the signal processing electronics.
3. The UV light source is measured by a UV detector. Software calculates the ratio of the PMT output
and the UV detector in order to compensate for variations in the UV light energy. Stray light is the
background light produced with zero ppb SOZ. Once this background light is subtracted, the CPU
will convert this electrical signal into the SO? concentration which is directly proportional to the
number of SO? molecules.
3.3 Definitions
Here are some key terms for this method.
Table 3-1 Definitions of Key Terms
Term Definition
DAS Data acquisition system. Used for automatic collection and
recording of Sulfur Dioxide concentrations.
Interferences Physical or chemical entities that cause Sulfur Dioxide
measurements to be higher (positive) or lower (negative) than they
would be without the entity. (See Section 3.6).
-------
Teledyne-API SO2 SOP
Version No. 2
4/26/2005
Page 7 of 17
3.4 Health and Safety Warnings
To prevent personal injury, please heed these warnings concerning the T-API 100AS.
1. Always use a third ground wire on all instruments.
2. Always unplug the analyzer when servicing or replacing parts.
3. If it is mandatory to work inside an analyzer while it is in operation, use extreme caution to avoid
contact with high voltages. The analyzer has a 110 volt Volts Alternating Current (VAC) power
supply. Refer to the manufacturer's instruction manual and know the precise locations of the VAC
components before working on the instrument.
4. Avoid electrical contact with jewelry. Remove rings, watches, bracelets, and necklaces to prevent
electrical bums.
3.5 Safety Precautions
To avoid damaging internal components of the T-API 100AS, the following precautions should be taken:
1. Wear an anti-static wrist strap that is properly connected to earth ground (note that when the analyzer
is unplugged, the chassis is not at earth ground);
2. If an anti-static wrist strap is not available, be sure to touch a grounded metal object before touching
any internal components;
3. Handle all printed circuit boards by the edge;
4. Carefully observe the instructions in each procedure specified in Section 8 of the manual;
5. Normally, if Teflon™ filters are used in the sample train, cleaning the optical bench will not be
required. However, in the event that the bench is cleaned, be careful to avoid damaging the interior of
the sample chamber. Use extreme caution when cleaning or servicing the sample chamber(s). In
addition the mirrors are very fragile; avoid dropping the instrument. This may damage, misalign or
crack the mirrors and cause expensive repairs;
6. Keep the interior of the analyzer clean;
7. Inspect the system regularly for structural integrity;
8. To prevent major problems with leaks, make sure that all sampling lines are reconnected after
required checks and before leaving the site;
9. Inspect tubing for cracks and leaks;
10. It is recommended that the analyzer be leak checked after replacement of any pneumatic parts;
11. If cylinders are used in tandem with Mass Flow Control (MFC) calibrators, use and transport is a
major concern. Gas cylinders can sometimes contain pressures as high as 2000 pounds per square
inch (psi). Handling of cylinders must be done in a safe manner. If a cylinder is accidentally dropped
and valve breaks off, the cylinder can become explosive or a projectile;
12. Transportation of cylinders is regulated by the Department of Transportation (DOT). It is strongly
recommended that all agencies contact the DOT or Highway Patrol to learn the most recent
regulations concerning transport of cylinders;
13. It is possible (and practical) to blend other compounds with SO2. If this is the case, it is
recommended that MSDS for all compounds be made available to all staff that use and handle the
cylinders or permeation tubes; and
14. Shipping of cylinders is governed by the DOT. Contact the DOT or your local courier about the
proper procedures and materials needed to ship high-pressure cylinders.
-------
Teledyne-API SO2 SOP
Version No. 2
4/26/2005
Page 8 of 17
3.6 Interferences
The most common source of interference is from other gases that fluoresce in a similar fashion to SO2
when exposed to UV light. The most significant of these is a class of hydrocarbons called polynuclear
aromatic hydrocarbons (PAH); of which naphthalene is a prominent example. Xylene is another
hydrocarbon that can cause interference. These hydrocarbons are removed via the hydrocarbon "kicker"
(see section 3.2 for explanation on the hydrocarbon "kicker").
Nitrogen oxide (NO) fluoresces in a spectral range close to SO2. Interference from NO is addressed by
the presence of the band pass filter, which allows only the wavelengths emitted by the excited SO
molecules to reach the PMT.
3.7 Personnel Qualifications
The person(s) chosen to operate the T-API 100AS should have a minimum of qualifications. The
understanding of basic chemistry and electronics are a must. The understanding of digital circuitry is
helpful, but not required. Also, courses in data processing and validation are also welcome.
3.8 Equipment and Supplies
Monitoring Apparatus: The T-API 100AS combines proven detection technology with advanced
diagnostics for greater flexibility and reliability. A diagram of the T-API 100AS instrument is described
in Figure 3-1. The three main components are:
• Pneumatic System: Consists of sample inlet line, particulate filter, reaction chamber, flowmeter, and
pump, all used to bring ambient air samples to the analyzer inlet.
• Analytical System: This portion of the instrument consists of the UV lamp, mirrors, photo-detector
and band pass filter.
• Electronic Hardware: The part of the analyzer that generally requires little or no maintenance. If the
T-API 100AS is operated above the manufacturer's recommended temperature limit, however,
individual integrated chips can fail and cause problems with data storage or retrieval.
Other apparatus and equipment includes the following.
Instrument Shelter: A shelter is required to protect the analyzer from precipitation and adverse weather
conditions, maintain operating temperature within the analyzer's temperature range requirements, and
provide security and electrical power. The recommended shelter temperature range is 20-30°C.
Spare Parts and Incidental Supplies: See the T-API 100AS manual, Section 10, for a list of
recommended spare parts.
Calibration System: A system that creates concentrations of SO2 of known quality is necessary for
establishing traceability. This is described in detail in the Environics Series 9100 Computerized Ambient
Monitoring Calibration System SOP. Please reference this document.
DAS: A data acquisition system is necessary for storage of ambient and ancillary data collected by the T-
API 100AS.
-------
Teledyne-API SO2 SOP
Version No. 2
4/26/2005
Page 9 of 17
E'ilir.uit
Sample Pum?
Figure 3-1 T-API Model 100AS Flow Diagram
Wiring, Tubing and Fittings: Teflon™, stainless steel, and borosilicate glass are inert materials that
should be used exclusively throughout the intake system. Stainless steel tubing should be avoided
because it is expensive, hard to clean, and can develop micro-cracks that are difficult to detect. Teflon™
tubing is the best choice for the connection between the intake manifold and the T-API 100AS inlet.
Examine and discard if particulate matter collects in the tubing. All fittings and ferrules should be made
of Teflon™ or stainless steel. Connection wiring to the DAS should shielded two strand wire or RS-232
cables for digital connections.
Reagents and Standard: The T-API 100AS does not require any reagents since the instrument uses
photometry to analyze for sulfur dioxide. All standards for the SO2 method can be obtained in
compressed cylinders and must be NIST traceable. Please see the Environics Series 9100 Computerized
Ambient Monitoring Calibration System SOP.
3.9 Procedure
3.9.1 Sample Collection: Sampling for trace level SO2 is performed by drawing ambient air through a
sample manifold directly into the analyzer continuously via a vacuum pump. All inlet materials must be
constructed from Teflon™, or borosilicate glass. The siting criteria for SO2 Trace Level instruments in
detailed in 40 CFR 58, appendix A4.
3.9.2 Sample Handling and Preservation: SO2 samples receive no special preparation prior to analysis.
Therefore this SOP does not need a section on Sample Handling and Preservation.
3.9.3 Instrument Operation , Startup and Maintenance
This section discusses startup, operation and maintenance of the T-API 100AS. The T-API 100AS has a
digital front panel screen with pushbuttons below. This allows the user to check functions, switch
operating parameters, and adjust zero and span. Recommend at start-up of instrument and after a warm-
up period for the instrument, run through the menu items and record the current settings. It is extremely
important that the user familiarize themselves with the menus available. Inadvertently changing
parameters within the analyzer can damage the instrument and possibly invalidate data as well.
-------
Teledyne-API SO2 SOP
Version No. 2
4/26/2005
Page 10 of 17
Please reference the T-API100AS owner's manual and read it carefully before adjusting any
parameters that are set by the factory.
3.9.3.1 Start up
Before the instrument is operated, inspect the instrument for any damage. If damage is observed to the
shipping box or the instrument, contact your shipping personnel.
Carefully remove the cover and check for internal damage. Please see Section 2 of the T-API 100AS
manual.
Remove the red shipping screws that hold down the internal bench and parts. See Section 2 of the T-API
100AS manual.
Once you have removed the shipping screws and performed your inspection, replace the cover.
Plug the instrument into a grounded power strip that has surge protection. It is also advisable to purchase
an Uninterrupted Power Supply (UPS). An UPS will protect the T-API 100AS from power surges and
keep the unit operating until an operator can shut it down.
Check to see that the 100AS has enough clearance so that it gets proper ventilation. Check the T-API
100AS manual, Section 2.
Connect the output of the analog to a DAS via shielded two wire cable. Please see EPA SOP on "Data
Management" for details.
Connect the digital RS-232 port to an appropriate cable and connect it to the DAS. Please see EPA SOP
on "Data Management" for further details.
Connect the sample inlet port to the station intake manifold.
Press the power rocker switch to "ON."
3.9.3.2 Operation and Range Setting
1. The exhaust fan will start and the display will come on. The Central Processing Unit (CPU) will boot
the system and load the firmware. You will see in the upper right hand corner that the fault warning
light will be flashing red. This is letting you know that the analyzer has been off.
2. To clear the fault warning, press the "CLR" button below the display. This will clear the warning and
the sample green light will flash.
3. Once the red fault light is cleared, the operator will see the main menu. At this time, the time of day
and date must be verified and reset if necessary.
4. From main menu press the toggle button under the "SETUP" label.
5. In this menu, you will see "8 1 8" on the bottom of the display. This is the default password. In
addition, you will see on the top of the display the words "ENTER SETUP PASS: 818."
6. At this time, press the toggle under "ENTR."
7. This will bring up the "PRIMARY SETUP MENU" screen.
8. In this menu, press the toggle button under "CLK."
9. In the next menu, press the "TIME" toggle switch.
10. In the next menu, you will see the time above 3 or 4 toggle switches. Adjust these toggles so that the
time is correct. Press the "ENTR" toggle switch.
11. This returns you to the "TIME OF DAY CLOCK" menu. Press the toggle switch under "DATE."
12. This will put you in the date menu. You should see the day (digits), month and year (digits) above 5
toggle switches. Adjust these toggles until the correct date is obtained. Press the "ENTR" toggle
switch. Press the "EXIT" toggle switch twice. This returns you to the main menu.
13. The range should be illustrated in the top middle of the main menu. This value should be set to 100
ppb. If it is not set to this range, then it must be reset.
14. To change the range of the instrument, press the toggle under "SETUP." If the password is correct,
then press "ENTR."
-------
Teledyne-API SO2 SOP
Version No. 2
4/26/2005
Page 11 of 17
15. In this menu, you will see "8 1 8" on the bottom of the display. This is the default password. In
addition, you will see on the top of the display the words "ENTER SETUP PASS: 818."
16. At this time, press the toggle under "ENTR."
17. This will bring up the "PRIMARY SETUP MENU" screen.
18. Press the toggle switch under the "RNGE."
19. This is the "RANGE CONTROL MENU." Press the toggle switch under the "UNIT."
20. This display will show the range options. Press the toggle under the "PPB." Press the toggle under
the "ENTR."
21. This will put you into the "RANGE CONTROL MENU." Press the toggle switch under the "SET."
22. This display will show the full scale range value. Press the toggles under the digits to adjust the
instrument to the full scale value desired. Press the toggle under the "ENTR."
23. Press the "EXIT" toggle switch twice.
24. The instrument is now set with the appropriate time, date and full scale range.
25. It is recommended that you allow the T-API 100AS 24 hours before you attempt function checks or
calibration.
If your DAS does not have RS-232 capabilities, then proceed to the next section, Diagnostic
Checks/Manual Checks. If you have connected the T-API 100AS to a computer or DAS, review the
Diagnostic Check from your computer screen. T-API offers API.COM, a computer program that allows
the operator to log the diagnostic data that is collected by the T-API 100AS CPU. Several DAS
manufacturers offer this type of software as well.
3.9.3.3 Diagnostic Checks/Manual Checks
To determine whether the T-API 100AS is working properly, the field operators should perform the
Diagnostic Checks every time they visit the monitoring station. It is good practice for the operator to
check these Diagnostic Checks either by the computer or manually. Below are instructions on how to
perform this manually. Please note that the T-API 100AS has set upper and lower ranges for some of
these Diagnostic checks. Please reference the owner's manual for these ranges.
1. If you observe the display, it should show "Sample" in the left hand corner, "Range" in the middle of
the display and "SO2= XX.X." Below this line there should be one line that read "",
"CAL" and "SETUP."
2. There is a series of toggle switches/ buttons below the display. These correspond to the bottom row
of the display.
3. If you press the button below the left hand "," will allow you to access the function check tree
in the opposite direction.
4. Toggle through the function check tree. The following table illustrates the functions that should be
recorded. Please see the T-API 100AS manual for more details. A maintenance checklist is attached
in Appendix A of this SOP.
-------
Table 3-2 Diagnostic Checks
Teledyne-API SO2 SOP
Version No. 2
4/26/2005
Page 12 of 17
Check
Range
Stability
Pressure
Sample flow
PMT
Norm PMT
UV lamp
Lamp ratio
Stray light
Dark PMT
Dark lamp
Slope
Offset
HVPS
DCPS
RCell temp
Box temp
PMT temp
Explanation
The full scale limit at which the reporting range of the instrument's analog outputs
are currently set.
The standard deviation of SO2 concentrations for the last 25 readings.
The current pressure of the sample gas as it enters the sample chamber.
The flow rate of the sample gas through the sample chamber.
The raw output voltage of the PMT.
The output voltage of the PMT after normalization.
The output voltage of the UV reference detector.
The current output of the UV reference detector divided by the reading stored in
the CPU's memory from the last time a UV lamp calibration was performed.
The offset due to stray light recorded by the CPU during the last zero-point
calibration.
The PMT output reading recorded the last time the UV source lamp shutter was
closed.
The UV reference detector output reading recorded the last time the UV source
lamp shutter was closed.
The sensitivity of the instrument as calculated during the last calibration.
The overall offset of the instrument as calculated during the last calibration.
The PMT high voltage power supply.
The composite of the +5 and + 15 VDC supplies.
The current temperature of the sample chamber.
The ambient temperature of the inside of the analyzer case.
The current temperature of the PMT.
Once the Diagnostic checks have been established and recorded for the T-API100AS, it is time to
calibrate the instrument. Please refer to section 3.9.4 of this SOP.
3.9.3.4 Preventive Maintenance
Preventive maintenance should prevent down-time and data loss. Table 3.3 lists the preventive
maintenance items that are listed in the T-API 100AS manual, Section 8.
Table 3-3 Preventive Maintenance Schedule for the 100AS
Item
Visual inspection and cleaning
Sample particulate filter inspection
TEST Functions
Perform Level I calibration
Pneumatic sub-system
PMT sensor hardware calibration
Sample chamber windows and filters
Leak check
Pump diaphragm
Factory calibration
Schedule
Bi-annually
Weekly
Weekly
Daily
Annually, or after repairs
On PMT or preamp changes
As necessary
Annually
Change annually
Annually, or after repairs
-------
Teledyne-API SO2 SOP
Version No. 2
4/26/2005
Page 13 of 17
3.9.3.5 Instrument Troubleshooting
The T-API 100AS manual has an excellent troubleshooting guide in Section 9.
3.9.4 Calibration and Standardization
The calibration of the T-API 100AS is performed by comparing the output of the instrument against
standardized gases of known quality. Generation of these gases is detailed in the Environics Series 9100
Computerized Ambient Monitoring Calibration System SOP. This section will detail how to adjust the
T-API 100AS to the standardized gases. Once the calibration has been performed, compare the response
of your DAS to the calculated "source" value. If this is outside of+/-10%, then adjust the instrument
response as detailed in the next sections.
3.9.4.1 Adjustment to Zero Air
Before running the zero air through the T-API 100AS, ensure that the zero air is free from contaminants.
One way to determine if the zero air is free from contaminants is through the use of multiple zero air
sources, and determining which source produces the lowest response.5
To adjust the output of the T-API 100AS to zero air, perform the following:
1. Allow the instrument to sample zero air from a manifold that is at or near atmospheric pressure
for a minimum of 15 minutes.
2. On the bottom of the front panel screen there is a toggle switch/ button that is beneath the "CAL"
label. Press this button.
3. This next screen is the "M-P CAL" screen. In this screen press the button below the "ZERO"
label.
4. The next screen will show an "ENTR, SPAN and CONC" above the toggle switches on the
bottom of the panel. Press the button below the "ENTR" label. This operation changes the
calculation equation and zeros the instrument.
Press the button below the "EXIT" label. This returns the operator to the main "SETUP" menu.
3.9.4.2 Adjustment to Calibration Gas
It is desirable, but not essential, to calibrate the T-API 100AS at low SO2 levels (typically, less than 20
ppb). However, fluorescence SO2 analyzers have been shown to be inherently linear over a wide dynamic
range. If low concentration SO2 calibration cylinders (less than 50 ppm) are used, there is the potential
for contamination by back diffusion from a poorly purged regulator. Contamination with even a small
amount of moisture from back diffusion can cause the SO2 concentration to become unstable; the lower
the cylinder concentration, the more susceptible it is to any contamination from "abuse" in the field5. The
best way to ensure low concentration cylinders are not contaminated by back diffusion is to make sure
whenever the cylinder valve is open, there is gas flow out of the cylinder.
To adjust the output of the T-API 100AS to NIST traceable calibration gas, perform the following:
1. Switch the calibration unit to generate a known concentration of SO2. Allow the instrument to
sample calibration gas from a manifold that is at near atmospheric pressure for a minimum of 15
minutes.
-------
Teledyne-API SO2 SOP
Version No. 2
4/26/2005
Page 14 of 17
2. On the bottom of the front panel screen of the main menu there is a button that is beneath the
"CAL" label. Press this button.
3. This next screen is the "M-P CAL" screen. The next screen will show an "ENTR, SPAN and
CONC" above the toggle switches on the bottom of the panel. Press the button below the
"CONC" label. On the bottom line, there will be digits below each toggle button. In order to
change the concentration, toggle each digit before and after the decimal place to get the
concentration that is being generated in the manifold by the calibrator. At this time, press the
"ENTR."
4. At this time, press the toggle below the "SPAN" switch.
5. This operation changes the calculation equation and adjusts the slope of the instrument.
Press the button below the "EXIT" label. This returns the operator to the main "SETUP" menu.
3.10 Data Analysis and Calculations
Data analysis for this analyzer is detailed in the "Data Acquisition and Management SOP."
4.0 QUALITY CONTROL AND QUALITY ASSURANCE
The following section has brief definitions of the QA/QC indicators. Table 4-1 has the Measurement
Quality Objectives (MQOs) of the T-API 100AS. Please note that this section details primarily with the
QA indicators. Quality Control for continuous electronic instruments, such as the T-API 100AS consists
of performing the diagnostic checks, maintenance and calibrations. These procedures are detailed in
sections 3.9.3 and 3.9.4: Instrument Operation, Startup and Maintenance and Calibration and
Standardization. Appendix A has an example of a Quality Control and Maintenance Record developed by
the EPA for this instrument.
4.1 Precision
Precision is defined as the measure of agreement among individual measurements of the same property
taken under the same conditions. For SO2, this refers to testing the SO2 analyzer in the field at
concentrations between 0.0003 and 0.005 ppm. The test must be performed, at a minimum, once every
two weeks. Calculations for Precision can be found in Reference item 3.
4.2 Bias
Bias is defined as the degree of agreement between a measured value and the true, expected, or accepted
value. Quantitative comparisons are made between the measured value and the true, standard value
during audits. Generally, three upscale points and a zero point are compared. Two audit types commonly
used for SO2, direct comparison and blind, are discussed below. The SOP should discuss plans for each
type of audit.
• Direct Comparison Audits: An independent audit system is brought to the monitoring location and
produces gas concentrations that are assayed by the monitoring station's SO2 analyzer. In most cases,
a person outside of the agency or part of an independent QA group within the agency performs the
audit. The responses of the on-site analyzer are then compared against the calculated concentration
from the independent audit system and a linear regression is generated
• Blind Audits: In blind audits (also called performance evaluation audits); agency staff are sent an
audit device, such as the National Performance Evaluation Program (NPEP). The agency staff does
-------
Teledyne-API SO2 SOP
Version No. 2
4/26/2005
Page 15 of 17
not know the SO2 concentrations produced by the audit equipment. Responses of the on-site analyzer
are then compared against those of the generator and a linear regression is calculated.
4.3 Representativeness
Representativeness refers to whether the data collected accurately reflect the conditions being measured.
It is the data quality indicator most difficult to quantify. Unless the samples are truly representative, the
other indicators are meaningless. Since the NCORE Level I and II siting criteria are urban and regional,
the trace level SO2 criteria are the same. Please reference the National Monitoring Strategy5 for a
discussion of NCORE Level II SO2 monitoring scale.
4.4 Completeness
Completeness is defined as the amount of data collected compared to a pre-specified target amount. For
SO2, EPA requires a minimum completeness of 75% (40 CFR 50, App.H.3). Typical completeness with
the T-API 100AS values can approach 90-93%.
4.5 Comparability
Comparability is defined as the process of collecting data under conditions that are consistent with those
used for other data sets of the same pollutant. The MQOs for a trace level SO2 instrument are still to be
determined. Please see Table 4-1.
4.6 Method Detection Limit
The method detection limit (MDL) or detectability refers to the lowest concentration of a substance that
can be determined by a given procedure. The T-API 100AS must be able to detect a minimum value of
300pptofSO2.
-------
Teledyne-API SO2 SOP
Version No. 2
4/26/2005
Page 16 of 17
Table 4-1 Measurement Quality Assurance Objectives
Requirement
Bias
Precision
Completeness
Representativen
ess
Comparability
Method
Detection Limit
Frequency
NCORE,
once per
year
1 every 2
weeks
Quarterly,
Annually
N/A
N/A
N/A
Acceptance Criteria
To be determined.
Concentration: 0.0003
-0.005 ppm,
Coefficient of
Variance: To be
determined.
NCORE, 75%
Neighborhood, Urban
or Regional Scale
Must be a Trace Level
instrument. See
Sections 3.1 and 3.2 of
this document.
300 ppt
Reference
40CFR
Pt.58
40CFR
Pt.58
Appendix
A
National
Monitoring
Strategy.
40 CFR 58
National
Monitoring
Strategy.
National
Monitoring
Strategy
Information or
Action
To be determined.
To be determined.
If under 75%,
institute Quality
Control Measures
N/A
N/A
Testing is
performed at the
factory.
5.0 REFERENCES
1. Code of Federal Regulations, Title 40, Part 53.23c
2. Code of Federal Regulation, Title 40, Part 58, Appendix A
3. The National Air Monitoring Strategy, Final Draft, 4/29/04,
http: //www .epa. gov/ttn/amtic/monstratdoc .html
4. Instruction Manual, T-API Model 100AS Trace Level SO2 Analyzer.
5. Trace SO2 Monitoring Guidance for the MANE-VU Regional Aerosol Intensive Network (RAIN)
program, Draft (dated March 7, 2005).
-------
Appendix A
Environmental Protection Agency
Monthly Quality Control and Maintenance Records
Teledyne - API 100AS Ultra-Sensitivity SO2 Analyzer
Teledyne-API SO2 SOP
Version No. 2
4/26/2005
Page 17 of 17
Site Name/Location
Technician
Month/Year
Serial Number
_Range
Date
Time
Range
Stability
Pressure
Sample flow
PMT
Norm PMT
UV lamp
Lamp ratio
Stray light
Dark PMT
Dark lamp
Slope
Offset
HVPS
DCPS
RCell temp
Box temp
PMT temp
Other Tests
ETest
OTest
Parameters
Acceptance
Criteria
Date
Comments and Notes
Figure A-l. T-API100AS Quality Control and Maintenance Record
-------
STANDARD OPERATING PROCEDURES
THERMO ELECTRON CORPORATION
MODEL 43C-TLE TRACE LEVEL
SULFUR DIOXIDE INSTRUMENT
Version 3
-------
Section 1.1 Acknowledgments
This Standard Operating Procedure (SOP) for Sulfur Dioxide - Trace Level is the product of EPA's
Office of Air Quality, Planning and Standards. The following individuals are acknowledged for their
contributions.
Principal Author
Solomon Ricks, OAQPS-EMAD, Research Triangle Park, NC 27711
Reviewers
Office of Air Quality, Planning and Standards
Joann Rice, Trace Gas Team Lead, OAQPS-EMAD, Research Triangle Park, NC 27711
Anna Kelly, OAQPS-EMAD, Research Triangle Park, NC 27711
Keith Kronmiller, Mantech, Inc. Research Triangle Park, NC 27711
Comments and questions can be directed to:
Solomon Ricks
EPA-OAQPS
Emission, Analysis and Monitoring Division
Mail Drop D339-02
Research Triangle Park, NC 27711
Email: ricks.solomon(giepa.gov
Phone: (919)-541-5242
-------
Section 2.0 Table of Contents
Section
1.0 Title Page
1.4 Acknowledgements
2.0 Table of Contents
2. 1 List of Tables and Figures
3.0 Procedures
3.1 Scope and Applicability
3.2 Summary of Method
3.3 Definitions
3.4 Health and Safety Warnings
3.5 Cautions
3.6 Interferences
3.7 Personal Qualifications
3.8 Equipment and Supplies
3.9 Procedure
3.9.1 Sample Collection
3.9.2 Sample Handling and Preservation
3.9.3 Instrument Operation, Start up and Maintenance
3.9.3.1 Startup
3.9.3.2 Operation and Range Setting
3.9.9.3 Diagnostic Checks/Manual Checks
3.9.3.4 Preventive Maintenance
3.9.3.5 Instrument Trouble shooting
3.9.4 Calibration and Standardization
3.9.4.1 Adj ustment to Zero Air
3.9.4.2 Adjustment to Calibration Gas
3.10 Data Analysis and Calculations
4.0 Quality Control and Quality Assurance
4.1 Precision
4.2 Bias
4.3 Representativeness
4.4 Completeness
4.5 Comparability
4.6 Method Detection Limit
5.0 References
Page
1
2
3
4
5
5
5
6
6
6
7
7
8
9
9
9
9
9
9
10
11
12
12
12
12
12
12
13
13
13
13
13
14
15
-------
List of Tables
Table Name and Number
3-1 Definition of Key Terms
3-2 Diagnostic Checks
3-3 Preventive Maintenance Schedule for the 43C-TLE
4-1 Measurement Quality Assurance Objectives
Page
6
11
11
15
List of Figures
Figure Name and Number
3-1 Model 43C-TLE Flow Diagram
A-l Model 43C-TLE Quality Control and Maintenance Record
Page
8
16
-------
STANDARD OPERATING PROCEDURES
THERMO ELECTRON CORPORATION
MODEL 43C-TLE TRACE LEVEL SULFUR DIOXIDE INSTRUMENT
3.0 PROCEDURES
3.1 Scope and Applicability
Sulfur Dioxide (SO2) is a colorless, nonflammable gas that has a strong suffocating odor. SO2 originates
from fuel containing sulfur (mainly coal and oil) burned at power plants and during metal smelting and
other industrial processes. High levels of SO2 can result in temporary breathing impairment for asthmatic
children and adults who are active outdoors. Long-term exposure to high levels of SO2, in the presence of
high levels of particulate matter, may aggravate existing cardiovascular disease and respiratory illness.
The Thermo Electron Corporation model 43C-TLE combines proven detection technology and advanced diagnostics
for the determination of trace levels of SO2. This SOP will detail the operation, preventive maintenance, cautions
and health warnings.
The Detection Limit (DL) for a non-trace level SO2 analyzer is 10 parts per billion (ppb) (Code of Federal
Regulations, Volume 40, Part 53.23c, or, in the shortened format used hereafter, 40 CFR 53.23c)3.
However, the 43C-TLE has an estimated DL of 100 parts per trillion (ppt), which is accomplished by an
increased detector sensitivity, as well as increasing the length of the standard instrument's optics bench.
This document will discuss the Trace Level (TL) operating procedures in detail.
3.2 Summary of Method
The Model 43C-TLE Trace Level operating principle is based on measuring the emitted fluorescence of
SO2 produced by the absorption of ultraviolet (UV) light. Pulsating UV light is focused through a narrow
band-pass filter mirror allowing only light wavelengths of 190 to 230 nm to pass into the fluorescent
chamber. SO2 absorbs light in this region without any quenching by air or most other molecules found in
polluted air. The SO2 molecules are excited by UV light and emit a characteristic decay radiation. A
second filter allows only this decay radiation to contact a photomultiplier tube (PMT). Electronic signal
processing transfers the light energy impinging on the PMT into a voltage which is directly analyzed.
Specifically,
SO2 + hvi -» SO2* -» SO2 + hv2
where:
hvi = incidence light,
hv2 = fluoresced light, and
SC>2* = SC>2 in its excited state
The 43C-TLE instrument operates in the following fashion:
-------
1. In sample mode, the sample is drawn into the analyzer through the SAMPLE bulkhead. The sample
flows through a hydrocarbon "kicker." which operates on a selective permeation principle, allowing
only hydrocarbon molecules to pass through the tube wall. The driving force for the hydrocarbon
removal is the differential partial pressure across the wall. This differential pressure is produced
within the instrument by passing the sample gas through a capillary tube to reduce its pressure and
feeding it into the shell side of the hydrocarbon kicker. The SO? molecules pass through the
hydrocarbon "kicker" unaffected.
2. The sample flows into the fluorescence chamber, where pulsating UV light excites the SO?
molecules. The condensing lens focuses the pulsating UV light into the mirror assembly. The mirror
assembly contains eight selective mirrors that reflect only the wavelengths which excite SO?
molecules.
3. As the excited SO? molecules decay to lower energy states they emit UV light that is proportional to
the SO? concentration. The band pass filter allows only the wavelengths emitted by the excited SO?
molecules to reach the PMT. The PMT detects the UV light emission from the decaying SO?
molecules. The photo detector, located at the back of the fluorescence chamber, continuously
monitors the pulsating UV light source and is connected to a circuit that compensates for fluctuations
in the UV light.
4. The sample then flows through a flow sensor, a capillary, and the shell side of the hydrocarbon
"kicker." The model 43C-TLE trace level outputs the SO? concentration to the front panel display
and the analog or digital outputs.
3.3 Definitions
Here are some key terms for this method.
Table 3-1 Definitions of Key Terms
Term Definition
DAS Data acquisition system. Used for automatic collection and
recording of Sulfur Dioxide concentrations.
Interferences Physical or chemical entities that cause Sulfur Dioxide
measurements to be higher (positive) or lower (negative) than they
would be without the entity. (See Section 3.6).
3.4 Health and Safety Warnings
To prevent personal injury, please heed these warnings concerning the 43C-TLE.
1. Always use a third ground wire on all instruments.
2. Always unplug the analyzer when servicing or replacing parts.
3. If it is mandatory to work inside an analyzer while it is in operation, use extreme caution to avoid
contact with high voltages. The analyzer has a 110 volt Volts Alternating Current (VAC) power
supply. Refer to the manufacturer's instruction manual and know the precise locations of the VAC
components before working on the instrument.
4. Avoid electrical contact with jewelry. Remove rings, watches, bracelets, and necklaces to prevent
electrical bums.
3.5 Safety Precautions
To avoid damaging internal components of the 43C-TLE. the following precautions should be taken:
-------
1. Wear an anti-static wrist strap that is properly connected to earth ground (note that when the analyzer
is unplugged, the chassis is not at earth ground);
2. If an anti-static wrist strap is not available, be sure to touch a grounded metal object before touching
any internal components;
3. Handle all printed circuit boards by the edge;
4. Carefully observe the instructions in each procedure specified in Chapter 7 of the manual;
5. Normally, if Teflon™ filters are used in the sample train, cleaning the optical bench will not be
required. However, in the event that the bench is cleaned, be careful to avoid damaging the interior of
the sample chamber. Use extreme caution when cleaning or servicing the sample chamber(s). In
addition the mirrors are very fragile; avoid dropping the instrument. This may damage, misalign or
crack the mirrors and cause expensive repairs;
6. Keep the interior of the analyzer clean;
7. Inspect the system regularly for structural integrity;
8. To prevent major problems with leaks, make sure that all sampling lines are reconnected after
required checks and before leaving the site;
9. Inspect tubing for cracks and leaks;
10. It is recommended that the analyzer be leak checked after replacement of any pneumatic parts;
11. If cylinders are used in tandem with Mass Flow Control (MFC) calibrators, use and transport is a
major concern. Gas cylinders can sometimes contain pressures as high as 2000 pounds per square
inch (psi). Handling of cylinders must be done in a safe manner. If a cylinder is accidentally dropped
and valve breaks off, the cylinder can become explosive or a projectile;
12. Transportation of cylinders is regulated by the Department of Transportation (DOT). It is strongly
recommended that all agencies contact the DOT or Highway Patrol to learn the most recent
regulations concerning transport of cylinders;
13. It is possible (and practical) to blend other compounds with SO2. If this is the case, it is
recommended that MSDS for all compounds be made available to all staff that use and handle the
cylinders or permeation tubes; and
14. Shipping of cylinders is governed by the DOT. Contact the DOT or your local courier about the
proper procedures and materials needed to ship high-pressure cylinders.
3.6 Interferences
The most common source of interference is from other gases that fluoresce in a similar fashion to SO2
when exposed to UV light. The most significant of these is a class of hydrocarbons called polynuclear
aromatic hydrocarbons (PAH); of which naphthalene is a prominent example. Xylene is another
hydrocarbon that can cause interference. These hydrocarbons are removed via the hydrocarbon "kicker"
(see section 3.2 for explanation on the hydrocarbon "kicker").
Nitrogen oxide (NO) fluoresces in a spectral range close to SO2. Interference from NO is addressed by
the presence of the band pass filter, which allows only the wavelengths emitted by the excited SO
molecules to reach the PMT.
-------
3.7 Personnel Qualifications
The person(s) chosen to operate the Thermo 43C-TLE should have a minimum of qualifications. The
understanding of basic chemistry and electronics are a must. The understanding of digital circuitry is
helpful, but not required. Also, courses in data processing and validation are also welcome.
3.8 Equipment and Supplies
Monitoring Apparatus: The design of the 43C-TLE combines proven detection technology with
advanced diagnostics for greater flexibility and reliability. A diagram of the 43C-TLE instrument is
described in Figure 3-1. The three main components are:
• Pneumatic System: Consists of sample inlet line, particulate filter, reaction chamber, flowmeter, and
pump, all used to bring ambient air samples to the analyzer inlet.
• Analytical System: This portion of the instrument consists of the UV lamp, mirrors, photo-detector
and band pass filter.
• Electronic Hardware: The part of the analyzer that generally requires little or no maintenance. If the
43C-TLE is operated above the manufacturer's recommended temperature limit, however, individual
integrated chips can fail and cause problems with data storage or retrieval.
Other apparatus and equipment includes the following.
Instrument Shelter: A shelter is required to protect the analyzer from precipitation and adverse weather
conditions, maintain operating temperature within the analyzer's temperature range requirements, and
provide security and electrical power. The recommended shelter temperature range is 20-30°C.
Spare Parts and Incidental Supplies: See the 43C-TLE manual, Chapter 5, for a list of recommended
spare parts.
Calibration System: A system that creates concentrations of SO2 of known quality is necessary for
establishing traceability. This is described in detail in the Environics Series 9100 Computerized Ambient
Monitoring Calibration System SOP. Please reference this document.
DAS: A data acquisition system is necessary for storage of ambient and ancillary data collected by the
43C-TLE.
-------
Figure 3-1 Model 43C-TLE Flow Diagram
Wiring, Tubing and Fittings: Teflon™, stainless steel, and borosilicate glass are inert materials that
should be used exclusively throughout the intake system. Stainless steel tubing should be avoided
because it is expensive, hard to clean, and can develop micro-cracks that are difficult to detect. Teflon™
tubing is the best choice for the connection between the intake manifold and the 43C-TLE inlet. Examine
and discard if particulate matter collects in the tubing. All fittings and ferrules should be made of
Teflon™ or stainless steel. Connection wiring to the DAS should shielded two strand wire or RS-232
cables for digital connections.
Reagents and Standard: The 43C-TLE does not require any reagents since the instrument uses
photometry to analyze for sulfur dioxide. All standards for the SO2 method can be obtained in
compressed cylinders and must be NIST traceable. Please see the Environics Series 9100 Computerized
Ambient Monitoring Calibration System SOP.
3.9 Procedure
3.9.1 Sample Collection: Sampling for trace level SO2 is performed by drawing ambient air through a
sample manifold directly into the analyzer continuously via a vacuum pump. All inlet materials must be
constructed from Teflon™, or borosilicate glass. The siting criteria for SO2 Trace Level instruments in
detailed in 40 CFR 58, appendix A4.
3.9.2 Sample Handling and Preservation: SO2 samples receive no special preparation prior to analysis.
Therefore this SOP does not need a section on Sample Handling and Preservation.
3.9.3 Instrument Operation , Startup and Maintenance
This section discusses startup, operation and maintenance of the 43C-TLE. The 43C-TLE has a digital
front panel screen with pushbuttons below. This allows the user to check functions, switch operating
parameters, and adjust zero and span. Recommend at start-up of instrument and after a warm-up period
for the instrument, run through the menu items and record the current settings. It is extremely important
that the user familiarize themselves with the menus available. Inadvertently changing parameters
within the analyzer can damage the instrument and possibly invalidate data as well. Please
reference the Thermo 43C-TLE owner's manual and read it carefully before adjusting any
parameters that are set by the factory.
-------
3.9.3.1 Start up
Before the instrument is operated, inspect the instrument for any damage. If damage is observed to the
shipping box or the instrument, contact your shipping personnel.
Carefully remove the cover and check for internal damage. Please see Chapter 2 of the 43C-TLE
manual.
Check that all connectors and printed circuit boards are firmly attached.
Once you have removed any packing material and performed your inspection, replace the cover.
Connect the sample line to the SAMPLE bulkhead on the rear panel of the instrument. See Chapter 2 of
the 43C-TLE manual.
Connect the EXHAUST bulkhead to a suitable vent.
Connect a suitable recording device to the rear panel terminals. Please see EPA SOP on "Data
Management" for details.
Plug the instrument into a grounded power strip that has surge protection. It is also advisable to purchase
an Uninterrupted Power Supply (UPS). An UPS will protect the 43C-TLE from power surges and
keep the unit operating until an operator can shut it down.
Press the power rocker switch to "ON."
3.9.3.2 Operation and Range Setting
1. The exhaust fan will start and the Power-Up and Self-Test screens will be displayed. These screens
are displayed each time the instrument is turned on, and will continue to be displayed till the
instrument has completed its warm up and self-checks. You should allow 30 minutes for the
instrument to stabilize.
2. After the warm-up period the Run screen, the normal operating screen, is displayed. This screen is
where the SO2 concentration is displayed.
3. From the Run screen, the Main Menu, which contains a list of submenus, can be displayed by
pressing the MENU pushbutton. If the instrument is in REMOTE mode, press ENTER and select
LOCAL mode in order to be able to change parameters.
4. Instrument parameters and features are divided into the submenus according to their function. Use
the UP/DOWN ARROW pushbuttons to move the cursor to each submenu. Note: When the Main
Menu is entered directly from the Run screen, the LEFT ARROW pushbutton may be used to jump
to the most recently displayed submenu screen.
5. To set the range for the instrument, press the DOWN ARROW pushbutton till the cursor is on
"Range." Press the ENTER pushbutton to display the Range Menu.
6. In the upper right corner of the display, the word SINGLE, DUAL, or AUTO is displayed to indicate
the active mode. For a detailed explanation about the SINGLE, DUAL, or AUTORANGE mode, see
Chapter 3 (page 3-7) of the manual. This SOP addresses setting the range for a single range.
7. Press the ENTER pushbutton for the Gas Units screen. Use the DOWN ARROW pushbutton to
select "PPB" and press ENTER. Press MENU to return to the Range Menu.
8. Use the DOWN ARROW pushbutton to display the Range screen and press ENTER.
9. Use the UP/DOWN ARROW pushbuttons to scroll through the preset ranges. Select "100.0" and
press ENTER. Press MENU to return to the Range Menu.
10. Press RUN to return to the Run screen.
11. To set the correct time and date on the instrument, press MENU to return to the Main Menu. Press
the DOWN ARROW pushbutton till the cursor is on Instrument Controls. Press ENTER to display
the Instrument Controls screen.
12. Use the UP/DOWN ARROW pushbuttons to scroll through the choices. Select "Time" and press
ENTER
13. Use the UP/DOWN ARROW pushbuttons to increase/decrease the hours and minutes; use the
LEFT/RIGHT ARROW pushbuttons to move the cursor left and right. Set the appropriate time and
press ENTER. Press MENU to return to the Instrument Controls screen.
-------
14. Select "Date" and press ENTER.
15. Use the UP/DOWN ARROW pushbuttons to increase/decrease the month, day, and year; use the
LEFT/RIGHT ARROW pushbuttons to move the cursor left and right. Set the appropriate date and
press ENTER. Press RUN to return to the Run screen.
16. The instrument is now set with the appropriate time, date, full scale range and units.
17. It is recommended that you allow the 43C-TLE 24 hours before you attempt function checks or
calibration.
18. If your DAS does not have RS-232 capabilities, then proceed to the next section, Diagnostic
Checks/Manual Checks. If you have connected the 43C-TLE to a computer or DAS, review the
Diagnostic Check from your computer screen.
3.9.3.3 Diagnostic Checks/Manual Checks
To determine whether the 43C-TLE is working properly, the field operators should perform the
Diagnostic Checks every time they visit the monitoring station. It is good practice for the operator to
check these Diagnostic Checks either by the computer or manually. Below are instructions on how to
perform this manually. Please note that the 43C-TLE has set upper and lower ranges for some of these
Diagnostic checks. Please reference the owner's manual for these ranges.
1. To display the Diagnostics menu, from the Run screen press the MENU pushbutton to display the
Main Menu. Use the UP/DOWN ARROW pushbuttons to move the cursor to "Diagnostics." Press
ENTER for the Diagnostics screen.
2. Use the UP/DOWN ARROW pushbuttons to toggle through the function check tree. The following
table illustrates the functions that should be recorded. Please see Chapter 3 (page 3-34) 43C-TLE
manual for more details. A maintenance check sheet is attached in Appendix A of this SOP.
3. On the Program Number screen, the version numbers of the program installed are displayed. Prior to
contacting the factory with any questions regarding the instrument, note the program numbers.
-------
Table 3-2 Diagnostic Checks
Check
Voltages
Temperatures
Pressure
Flow
Lamp intensity
Optical Span Test
Test Analog Outputs
Explanation
The current DC power supply and PMT power supply voltages
The current internal instrument and chamber temperatures
The current chamber pressure
The current sample flow rate
The current UV lamp intensity
A quick way of checking the optics and electronics for span drift
Enable analog outputs to be set to zero and full scale to adjust analog outputs to
agree with the front panel display
Once the Diagnostic checks have been established and recorded for the 43C-TLE, it is time to calibrate
the instrument. Please refer to section 3.9.4 of this SOP.
3.9.3.4 Preventive Maintenance
Preventive maintenance should prevent down-time and data loss. Table 3.3 lists the preventive
maintenance items that are listed in the model 43C-TLE manual, Chapter 5. The maintenance procedures
described in Chapter 5 of the manual should be performed every six months.
Table 3-3 Preventive Maintenance Schedule for the 43C-TLE
Item
Visual inspection and cleaning
Sample particulate filter inspection
Verify Test Function
Perform Level I calibration
Capillary inspection and replacement
Perform flow check
Fan filter inspection and cleaning
Lamp voltage check
Schedule
Bi-annually
Weekly
Weekly
Daily
Bi-annually
Bi-annually
Bi-annually
Bi-annually
3.9.3.5 Instrument Troubleshooting
The 43C-TLE manual has an excellent troubleshooting guide in Chapter 6.
3.9.4 Calibration and Standardization
The calibration of the 43C-TLE is performed by comparing the output of the instrument against
standardized gases of known quality. Generation of these gases is detailed in the Environics Series 9100
Computerized Ambient Monitoring Calibration System SOP. This section will detail how to adjust the
43C-TLE to the standardized gases. Once the calibration has been performed, compare the response of
your DAS to the calculated "source" value. If this is outside of+/-10%, then adjust the instrument
response as detailed in the next sections.
3.9.4.1 Adjustment to Zero Air
Before running the zero air through the 43C-TLE, ensure that the zero air is free from contaminants. One
way to determine if the zero air is free from contaminants is through the use of multiple zero air sources,
and determining which source produces the lowest response5. Appendix D of the 43C-TLE manual
provides suggestions for generating trace-quality zero air.
-------
In order to adjust the output of the 43C-TLE to zero air, perform the following:
1. Allow the instrument to sample zero air from a manifold that is at near atmospheric pressure for a
minimum of 15 minutes.
2. From the Main Menu select the Calibration menu; select the "Calibrate Zero" screen.
3. Press ENTER to set the SO2 reading to zero.
4. Press MENU to return to the Calibration menu.
5. Press RUN to return to the Run screen.
3.9.4.2 Adjustment to Calibration Gas
It is desirable, but not essential, to calibrate the 43C-TLE at low SO2 levels (typically, less than 20 ppb).
However, fluorescence SO2 analyzers have been shown to be inherently linear over a wide dynamic
range. If low concentration SO2 calibration cylinders (less than 50 ppm) are used, there is the potential
for contamination by back diffusion from a poorly purged regulator. Contamination with even a small
amount of moisture from back diffusion can cause the SO2 concentration to become unstable; the lower
the cylinder concentration, the more susceptible it is to any contamination from "abuse" in the field5. The
best way to ensure low concentration cylinders are not contaminated by back diffusion is to make sure
whenever the cylinder valve is open, there is gas flow out of the cylinder.
Appendix D of the 43C-TLE manual offers additional discussion on trace level calibration issues.
In order to adjust the output of the 43C-TLE to NIST traceable calibration gas, perform the following:
1. Switch the calibration unit to generate a known concentration of SO2. Allow the instrument to
sample calibration gas from a manifold that is at near atmospheric pressure for a minimum of 15
minutes.
2. From the Main Menu select the Calibration menu; select the "Calibrate SO2" screen.
3. On the bottom line, there will be individual digits with which the span value can be set. In order
to change the concentration, use the UP/DOWN ARROW pushbuttons to increase/decrease
each digit; use the LEFT/RIGHT ARROW pushbuttons to move the cursor left and right.
4. Press ENTER to calibrate the SO2 reading to the SO2 calibration gas.
5. This operation changes the calculation equation and adjusts the SO2 span coefficient of the
instrument.
6. Press MENU to return to the Calibration menu.
7. Press RUN to return to the Run screen.
3.10 Data Analysis and Calculations
Data analysis for this analyzer is detailed in the "Data Acquisition and Management SOP."
4.0 QUALITY CONTROL AND QUALITY ASSURANCE
The following section has brief definitions of the QA/QC indicators. Table 4-1 has the Measurement
Quality Objectives (MQOs) of the Thermo 43C-TLE. Please note that this section details primarily with
the QA indicators. Quality Control for continuous electronic instruments, such as the 43C-TLE consists
of performing the diagnostic checks, maintenance and calibrations. These procedures are detailed in
sections 3.9.3 and 3.9.4: Instrument Operation, Startup and Maintenance and Calibration and
Standardization. Appendix A has an example of a Quality Control and Maintenance Record developed by
the EPA for this instrument.
-------
4.1 Precision
Precision is defined as the measure of agreement among individual measurements of the same property
taken under the same conditions. For SO2, this refers to testing the SO2 analyzer in the field at
concentrations between 0.0003 and 0.005 ppm. The test must be performed, at a minimum, once every
two weeks. Calculations for Precision can be found in Reference item 3.
4.2 Bias
Bias is defined as the degree of agreement between a measured value and the true, expected, or accepted
value. Quantitative comparisons are made between the measured value and the true, standard value
during audits. Generally, three upscale points and a zero point are compared. Two audit types commonly
used for SO2, direct comparison and blind, are discussed below. The SOP should discuss plans for each
type of audit.
• Direct Comparison Audits: An independent audit system is brought to the monitoring location and
produces gas concentrations that are assayed by the monitoring station's SO2 analyzer. In most cases,
a person outside of the agency or part of an independent QA group within the agency performs the
audit. The responses of the on-site analyzer are then compared against the calculated concentration
from the independent audit system and a linear regression is generated
• Blind Audits: In blind audits (also called performance evaluation audits); agency staff are sent an
audit device, such as the National Performance Evaluation Program (NPEP). The agency staff does
not know the SO2 concentrations produced by the audit equipment. Responses of the on-site analyzer
are then compared against those of the generator and a linear regression is calculated.
4.3 Representativeness
Representativeness refers to whether the data collected accurately reflect the conditions being measured.
It is the data quality indicator most difficult to quantify. Unless the samples are truly representative, the
other indicators are meaningless. Since the NCORE Level I and II siting criteria are urban and regional,
the trace level SO2 criteria are the same. Please reference the National Monitoring Strategy5 for a
discussion of NCORE Level II SO2 monitoring scale.
4.4 Completeness
Completeness is defined as the amount of data collected compared to a pre-specified target amount. For
SO2, EPA requires a minimum completeness of 75% (40 CFR 50, App.H.3). Typical completeness with
the 43C-TLE values can approach 90-93%.
4.5 Comparability
Comparability is defined as the process of collecting data under conditions that are consistent with those
used for other data sets of the same pollutant. The 43C-TLE meets the MQOs for a trace level SO2
instrument. Please see Table 4-1.
4.6 Method Detection Limit
The method detection limit (MDL) or detectability refers to the lowest concentration of a substance that
can be determined by a given procedure. The 43C-TLE must be able to detect a minimum value of 300
ppt of SO2.
-------
Table 4-1 Measurement Quality Assurance Objectives
Requirement
Bias
Precision
Completeness
Representativen
ess
Comparability
Method
Detection Limit
Frequency
NCORE,
once per
year
1 every 2
weeks
Quarterly,
Annually
N/A
N/A
N/A
Acceptance Criteria
To be determined.
Concentration: 0.0003
-0.005 ppm,
Coefficient of
Variance: To be
determined.
NCORE, 75%
Neighborhood, Urban
or Regional Scale
Must be a Trace Level
instrument. See
Sections 3.1 and 3.2 of
this document.
300 ppt
Reference
40CFR
Pt.58
40CFR
Pt.58
Appendix
A
National
Monitoring
Strategy.
40 CFR 58
National
Monitoring
Strategy.
National
Monitoring
Strategy
Information or
Action
To be determined.
To be determined.
If under 75%,
institute Quality
Control Measures
N/A
N/A
Testing is
performed at the
factory.
5.0 REFERENCES
1. Code of Federal Regulations, Title 40, Part 53.23c
2. Code of Federal Regulation, Title 40, Part 58, Appendix A
3. The National Air Monitoring Strategy, Final Draft, 4/29/04,
http://www.epa.gov/ttn/amtic/monstratdoc.html
4. Instruction Manual, Model 43 C Trace Level SO2 Analyzer
5. Trace SO2 Monitoring Guidance for the MANE-VU Regional Aerosol Intensive Network (RAIN)
program, Draft (dated March 7, 2005).
-------
Appendix A
Environmental Protection Agency
Monthly Quality Control and Maintenance Records
Thermo 43C-TLE SO2 Analyzer
Site Name/Location
Technician
Month/Year
Serial Number
_Range
Parameter
DC Voltage
PMT Voltage
Internal Temp
Chamber Temp
Chamber
Pressure
Sample Flow
UV Lamp
Intensity
Other Tests
Optical Span
Test
Date
Date
Date
Acceptance
Criteria
Date
Comments and Notes
Figure A-l Thermo 43C-TLE Quality Control and Maintenance Record
-------
STANDARD OPERATING PROCEDURES
THERMO ENVIRONMENTAL INSTRUMENTS
42CY NOy TRACE LEVEL
REACTIVE NITROGEN COMPOUNDS INSTRUMENT
Version 3
-------
Section 1.1 Acknowledgments
This Standard Operating Procedure (SOP) for Trace Level Reactive Nitrogen Compounds is the product
of EPA's Office of Air Quality, Planning and Standards. The following individuals are acknowledged for
their contributions.
Principal Author
Kevin A. Cavender, OAQPS-EMAD, Research Triangle Park, NC 27711
Reviewers
Office of Air Quality, Planning and Standards
Joann Rice, Trace Gas Team Lead, OAQPS-EMAD, Research Triangle Park, NC 27711
Anna Kelly, OAQPS-EMAD, Research Triangle Park, NC 27711
Keith Kronmiller, Mantech, Inc. Research Triangle Park, NC 27711
Comments and questions can be directed to:
Kevin A. Cavender
EPA-OAQPS
Emission, Analysis and Monitoring Division
Mail Drop C339-02
Research Triangle Park, NC 27711
Email: cavender.kevin(giepa.gov
Phone: (919)-541-2364
-------
Section 2.0 Table of Contents
Section
1.0 Title Page
1.5 Acknowledgements
2.0 Table of Contents
2. 1 List of Tables and Figures
3.0 Procedures
3.1 Scope and Applicability
3.2 Summary of Method
3.3 Definitions
3.4 Health and Safety Warnings
3.5 Cautions
3.6 Interferences
3.7 Personal Qualifications
3.8 Equipment and Supplies
3.9 Procedure
3.9.1 Sample Collection
3.9.2 Sample Handling and Preservation
3.9.3 Instrument Operation, Start up and Maintenance
3.9.3.1 Startup
3.9.3.2 Operation and Range Setting
3.9.9.3 Diagnostic Checks/Manual Checks
3.9.3.4 Preventive Maintenance
3.9.3.5 Instrument Trouble shooting
3.9.4 Calibration and Standardization
3.9.4.1 Adj ustment to Zero Air
3.9.4.2 Adjustment to Calibration Gas
3.10 Data Analysis and Calculations
4.0 Quality Control and Quality Assurance
4.1 Precision
4.2 Bias
4.3 Representativeness
4.4 Completeness
4.5 Comparability
4.6 Method Detection Limit
5.0 References
Page
1
2
3
4
5
5
5
7
7
7
8
8
8
10
10
10
10
10
11
12
13
13
13
14
14
15
15
15
15
16
16
16
16
17
-------
List of Tables
Table Name and Number
3-1 Definition of Key Terms
3-2 Diagnostic Checks
3-3 Preventive Maintenance Schedule for the 42CY
4-1 Measurement Quality Assurance Objectives
Page
7
13
13
17
List of Figures
Figure Name and Number
3-1 Simplified Flow Diagram of 42CY Monitor
3-2 Schematic of plumbing and wiring for 42CY External Converter
Page
6
11
-------
Thermo 42CY-NOy SOP
Version No. 3
9/28/2005
Page 5 of 18
STANDARD OPERATING PROCEDURES
THERMO ENVIRONMENTAL INSTRUMENTS
42CY NOy TRACE LEVEL
REACTIVE NITROGEN COMPOUNDS INSTRUMENT
3.0 PROCEDURES
3.1 Scope and Applicability
Reactive nitrogen compounds (NOy) have been identified as precursors for both ozone and fine particulate
matter (PM2 5). 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.
NOy consists of all oxides of nitrogen in which the oxidation state of the N atom is +2 or greater, ie, the
sum of all reactive nitrogen oxides including NOX (NO + NO2) and other nitrogen oxides referred to as
NOZ. The major components of NOZ include nitrous acids [nitric acid (HNO3), and nitrous acid
(HONO)], organic nitrates [peroxyl acetyl nitrate (PAN), methyl peroxyl acetyl nitrate (MPAN), and
peroxyl propionyl nitrate, (PPN)], and particulate nitrates.
NO + NO2 + NO7 = NO,
y
The Thermo Environmental Instruments (TECO) model 42CY is an instrument for the determination of trace levels
of NOy by its chemiluminescent reaction with O3. This SOP will detail the operation, calibration, preventive
maintenance, cautions and health warnings.
3.2 Summary of Method
The analytical principle is based on the chemiluminescent reaction of NO with an excess of O3. This
reaction produces a characteristic near infrared luminescence with an intensity that is linearly proportional
to the concentration of NO present. Specifically,
NO + O3 -> NO2 + O2 + hv
where:
hv = emitted photon energy
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 measure NOy, sample air is passed through a probe-mounted chemical reduction converter and the
nitroxyl compounds present are reduced to NO. The sample is then blended with O3 and the
-------
Thermo 42CY-NOy SOP
Version No. 3
9/28/2005
Page 6 of 18
chemiluminescent response is proportional to the concentration of NOy entering the converter. The
chemical reduction converter uses heated molybdenum to convert non-NO NOy species to NO.
Specifically,
where:
NO2/z + Mo -» NO + MoO3
Mo = heated molybdenum reductant
The concentration of NO2 + NOZ is calculated as the difference between a measured NOy value and a
measured NO value representing approximately the same point in time. This procedure is similar to the
current methodology used to measure NOX, however, the converter temperature is higher in order to
enhance conversion of NOZ species. In addition, the converter has been moved to the sample inlet to
avoid line losses of "sticky" NOy species such as HNO3.
A diagram of the 42CY instrument is presented in Figure 3-1. For the 42CY, ambient sample is first
drawn through a short Teflon sample line and split into two parallel flow channels using a 1A inch PFT
Teflon tee. Channel 1 passes through a Teflon filter and then directly to the monitor. Channel 2 first
passes through a catalytic converter before going through a Teflon filter to the monitor. Flow from each
channel is alternately fed to the reaction chamber to detect the NO. The converter is operated outside of
the analyzer, close to the ambient sampling point. This allows for a short flow path upstream of the
converter and minimizes the loss of species such as HNO3. In addition to alternating flows from Channel
1 and Channel 2 to the reactor, the analyzer also alternates a flow of internal zero air, produced by pre-
reacting the sample flow with a high concentration of ozone before reaching the chemiluminescent
detector. The signal from this zero air stream is used to correct for analyzer drift, and allows the analyzer
to achieve very low detection limits (0.05 ppb) compared with standard NOX analyzers.
Prereactor
Exhaust
Sample
In
Weatherproof
Enclosure
(Positioned at
Probe Inlet Height)
Pump
Figure 3-1. Simplified Flow Diagram of 42CY NOy Monitor
-------
Thermo 42CY-NOy SOP
Version No. 3
9/28/2005
Page 7 of 18
3.3 Definitions
Here are some key terms for this method.
Table 3-1 Definitions of Key Terms
Term Definition
NOy Total reactive nitrogen including NO, NO2, HNO3, and other
reactive organic nitrogen compounds.
NOZ Reactive nitrogen compounds other than NO and NO2.
DAS Data acquisition system. Used for automatic collection and
recording of Carbon Monoxide concentrations.
Interferences Physical or chemical entities that cause NOy measurements to be
higher (positive) or lower (negative) than they would be without the
entity. (See Section 3.6).
3.4 Health and Safety Warnings
To prevent personal injury, please heed these warnings concerning the 42CY.
1. Nitrogen oxides are a poisonous gas. Vent any nitrogen oxide or calibration span gas to the
atmosphere rather than into the shelter or other sampling area. If this is impossible, limit exposure
to nitrogen oxide by getting fresh air every 5 to 10 minutes. If the operator experiences light
headedness, headache or dizziness, leave the area immediately.
2. Always use a third ground wire on all instruments.
3. Always unplug the analyzer when servicing or replacing parts.
4. If it is mandatory to work inside an analyzer while it is in operation, use extreme caution to avoid
contact with high voltages. The analyzer has a 110 volt Volts Alternating Current (VAC) power
supply. Refer to the manufacturer's instruction manual and know the precise locations of the VAC
components before working on the instrument.
5. Avoid electrical contact with jewelry. Remove rings, watches, bracelets, and necklaces to prevent
electrical bums.
3.5 Cautions
To prevent damage to the 42CY. all cautions should immediately precede the applicable step in this SOP. The following precautions
should be taken:
1. Keep the interior of the analyzer clean.
2. Inspect the system regularly for structural integrity.
3. To prevent major problems with leaks, make sure that all sampling lines are reconnected after
required checks and before leaving the site.
4. Inspect tubing for cracks and leaks.
5. It is recommended that the analyzer be leak checked after replacement of any pneumatic parts.
6. If cylinders are used in tandem with Mass Flow Control (MFC) calibrators, use and transport is a
major concern. Gas cylinders can sometimes contain pressures as high as 2000 pounds per square
inch (psi). Handling of cylinders must be done in a safe manner. If a cylinder is accidentally dropped
and valve breaks off, the cylinder can become explosive or a projectile.
-------
Thermo 42CY-NOy SOP
Version No. 3
9/28/2005
Page 8 of 18
7. Transportation of cylinders is regulated by the Department of Transportation (DOT). It is strongly
recommended that all agencies contact the DOT or Highway Patrol to learn the most recent
regulations concerning transport of cylinders.
8. Low levels of nitrogen oxides in the air can irritate your eyes, nose, throat, and lungs, possibly
causing you to cough and experience shortness of breath, tiredness, and nausea. Exposure to low
levels can also result in fluid build-up in the lungs 1 or 2 days after exposure. Breathing high levels of
nitrogen oxides can cause rapid burning, spasms, and swelling of tissues in the throat and upper
respiratory tract, reduced oxygenation of body tissues, a build-up of fluid in your lungs, and death.
9. It is possible (and practical) to blend other compounds with NO. If this is the case, it is recommended
that MSDS for all compounds be made available to all staff that use and handle the cylinders or
permeation tubes.
10. Shipping of cylinders is governed by the DOT. Contact the DOT or your local courier about the
proper procedures and materials needed to ship high-pressure cylinders.
3.6 Interferences
Ammonia: Depending on the converter temperature, the converter may convert a small amount of
ammonia (NH3) to NO, resulting in increased NO readings. However, under normal circumstances NH3
concentrations are low compared to NO and this positive interference is negligible. Nonetheless, care
should be taken when siting the monitor to be sure that it is not located near significant NH3 sources
which could cause elevated NH3 concentrations (e.g., concentrated animal feeding operations).
3.7 Personal Qualifications
The person(s) chosen to operate the 42CY should have a minimum of qualifications. The understanding
of basic chemistry and electronics are a must. The understanding of digital circuitry is helpful, but not
required. Also, courses in data processing and validation are also welcome.
3.8 Equipment and Supplies
Monitoring Apparatus: The design of the 42CY is similar to the 42C with several major variations. A
diagram of the 42CY instrument is presented in Figure 3-1. The four main components are:
• Remote Inlet and Converter: This component consists of a weather resistant enclosure that houses the
molybdenum converter and supports the sample inlet.
• Pneumatic System: Consists of sample inlet lines, sample bypass pump, particulate filters, reaction
chamber, flowmeters, and vacuum pump, all used to bring ambient air samples to the analyzer.
• Analytical System: This portion of the instrument consists of the ozone generator, pre-reaction
chamber, reaction chamber, and photomultiplier tube.
• Electronic Hardware: The part of the analyzer that generally requires little or no maintenance. If the
42CY is operated above the manufacturer's recommended temperature limit, however, individual
integrated chips can fail and cause problems with data storage or retrieval.
Other apparatus and equipment includes the following.
Instrument Shelter: A shelter is required to protect the analyzer from precipitation and adverse weather
conditions, maintain operating temperature within the analyzer's temperature range requirements, and
provide security and electrical power. The recommended shelter temperature range is 20-30°C.
-------
Thermo 42CY-NOy SOP
Version No. 3
9/28/2005
Page 9 of 18
Spare Parts and Incidental Supplies:. See the 42CY operating manual, Chapter 5, for specific
maintenance and replacement requirements.
Calibration System: A system that creates concentrations of nitrogen oxide of known quality is
necessary for establishing traceability. The calibration system must also include a high precision ozone
generator in order to generate concentrations of nitrogen dioxide by gas phase titration with nitrogen
oxide. This is described in detail in the "Environics 9100 SOP." Please reference this document.
DAS: A data acquisition system is necessary for storage of ambient and ancillary data collected by the
42CY. The 42CY requires a minimum of two analog outputs, one each for NO and NOy outputs. A third
output is also needed if the monitor is to be run in auto-ranging mode to capture the range information.
Note, a digital DAS is preferred because diagnostic information can also be collected which will greatly
help troubleshooting and validation of data.
Wiring, Tubing and Fittings: PFT Teflon™ should be used exclusively throughout the intake system.
Examine the tubing and discard if particulate matter is collects in the tubing. All fittings and ferrules
should be made of Teflon™ or stainless steel. Connection wiring to the DAS should be shielded two
strand wire or RS-232 cables for digital connections.
Reagents and Standard: The 42CY does not require any reagents since the instrument uses photometry
to analyze for NOy. All standards for the NOy method can be obtained in compressed cylinders and must
be NIST traceable. Please see the "Calibration of Trace Gas Analyzers" SOP.
3.9 Procedure
3.9.1 Sample Collection: Ambient air is drawn through a sample inlet located on the remote
inlet/converter. The sampling point should be located 3 to 5 meters above ground level, at least 1 meter
from all obstructions, and at least 10 meters from obstructions over a range of at least 180 degrees. These
requirements necessitate mounting the catalytic converter outside the shelter. The sample bypass lines
from the converter to the instrument should not exceed 10 meters in length.
3.9.2 Sample Handling and Preservation: NOy samples receive no special preparation prior to analysis.
Therefore this SOP does not need a section on Sample Handling and Preservation.
3.9.3 Instrument Operation, Startup and Maintenance
This section discusses startup, operation and maintenance of the 42C-NOY. The 42CY series instrument
has a digital front panel screen with control buttons below. This allows the user to check functions,
switch operating parameters, adjust zero and span and read warnings messages. It is extremely
important that the user familiarize themselves with the menus available. Inadvertently changing
parameters within the analyzer can damage the instrument and possibly invalidate data as well.
Please reference the 42CY owner's manual and read it carefully before adjusting any parameters
that are set by the factory.
3.9.3.1 Installation and Start up
1. Before the instrument is operated, inspect the instrument for any damage. If damage is observed
to the shipping box or the instrument, contact your shipping personnel.
2. Remove the instrument from its shipping container and set on a table or bench that allows easy
access to both the front and read or the instrument.
-------
Thermo 42CY-NOy SOP
Version No. 3
9/28/2005
Page 10 of 18
3. Carefully remove the instrument cover and remove any packing material.
4. Check for possible internal damage.
5. Check that all connectors and printed circuit boards are firmly attached.
6. Once you have performed your inspection, replace the cover.
7. Remove the remote inlet enclosure and place on a bench that allows easy access.
8. Open the enclosure and check for internal damage.
9. Using % inch PFT Teflon™ tubing, plumb the inlet as shown in Figure 3-2.
10. Mount the enclosure at the desired sampling location being sure to meet the requirements detailed
in 3.9.1, above.
11. Determine the distance from the point where the sample tubing exits the converter enclosure to
the point where the tubing will enter the monitoring station, including ample length to account for
any curves or obstructions. Cut the two plastic conduits to this length.
12. Run the two sample lines, calibration line, and thermocouple wire through one conduit, and the
converter power cable through the second.
13. Connect the two sample lines, calibration line, thermocouple wire, and power cable inside the
converter enclosure as shown in Figure 3-2, and secure the conduits to the converter enclosure.
14. Run the tubing, thermocouple wire, and power cable in to the monitoring station.
15. Mount the 42CY in its rack being sure that the 42CY has enough clearance so that it gets proper
ventilation.
16. Connect the sample lines to the bypass pump, filters, and analyzer as shown in Figure 3-1.
17. Connect the Dryrite™ air dryer to the dry air bulkhead.
18. Connect the intake of the vacuum pump to the exhaust bulkhead.
19. Connect the charcoal container to the outlet of the vacuum pump.
20. Connect the converter power cable and thermocouple wire to the back of the 42CY.
21. Connect the power cable and plug the instrument, vacuum pump, and bypass pump into a
grounded power strip that has surge protection. It is also advisable to purchase an Uninterrupted
Power Supply (UPS). An UPS will protect the 42CY from power surges and keep the unit
operating until an operator can shut it down.
22. Connect the output of the analog to a DAS via shielded two wire cable. Please see EPA SOP on
"Data Management" for details.
23. Connect the digital RS-232 port to an appropriate cable and connect it to the DAS. Please see
EPA SOP on "Data Management" for further details.
24. Press the power rocker switch to "ON."
25. After an adequate warm-up period, run through the menu and record factory/start-up settings.
Compare to recommended limits located in the manual and listed on the daily worksheet.
26. Perform a leak check. Press the MENU pushbutton to display the Main Menu. Use the
UP/DOWN ARROW pushbuttons to move the cursor to "Diagnostics." Press ENTER for the
Diagnostics screen. Use the UP/DOWN ARROW pushbuttons to select Flow. The sample flow
rate should be approximately 11pm. Cap or otherwise plug the sample inlet on the remote
converter. The sample flow rate should drop to less than 0.1 1pm with in 15 seconds. Slowly
remove the cap. Do not leave the inlet capped for more than a minute as it can cause damage to
the monitor. If the sample flow does not drop to below 0.1 1pm, tighten all connections and
repeat leak test.
-------
Thermo 42CY-NOy SOP
Version No. 3
9/28/2005
Page 11 of 18
Figure 3-2. Schematic of plumbing and wiring for 42CY External Converter
3.9.3.2 Operation and Range Setting
1. The exhaust fan will start and the Power-Up and Self-Test screens will be displayed. These screens
are displayed each time the instrument is turned on, and will continue to be displayed till the
instrument has completed its warm up and self-checks. You should allow 30 minutes for the
instrument to stabilize.
2. After the warm-up period the Run screen, the normal operating screen, is displayed. This screen is
where the NO and NOy concentration is displayed. The display for the model that EPA is using
currently mis-reports the data. The data reported as NOX is actually NOy. The data reported as NO2 is
actually NO2 + NOZ. Future versions should correct this, and are expected to report NO, NOy and
"DIFF" which would represent NO2 + NOZ.
3. In the bottom right hand corner the word "LOCAL" or "REMOTE" will be displayed. The analyzer
must be in local mode to adjust the configuration using the keys on the front panel. Press the ENTER
button until the analyzer is in local mode.
4. From the Run screen, the Main Menu, which contains a list of submenus, can be displayed by
pressing the MENU pushbutton.
5. Instrument parameters and features are divided into the submenus according to their function. Use
the UP/DOWN ARROW pushbuttons to move the cursor to each submenu. Note: When the Main
Menu is entered directly from the Run screen, the LEFT ARROW pushbutton may be used to jump
to the most recently displayed submenu screen.
6. To set the range for the instrument, press the DOWN ARROW pushbutton till the cursor is on
"Range." Press the ENTER pushbutton to display the Range Menu.
7. In the upper right corner of the display, the word single, dual, or auto is displayed to indicate the
active mode. For a detailed explanation about the single, dual, or autorange mode, see Chapter 3
(page 3-7) of the manual. This SOP addresses setting the range for a single range.
-------
Thermo 42CY-NOy SOP
Version No. 3
9/28/2005
Page 12 of 18
8. Press the ENTER pushbutton for the Gas Units screen. Use the DOWN ARROW pushbutton to
select "PPB" and press ENTER. Press MENU to return to the Range Menu.
9. Use the DOWN ARROW pushbutton to display the Range screen and press ENTER.
10. Use the UP/DOWN ARROW pushbuttons to scroll through the preset ranges. Select "100.0" and
press ENTER. Press MENU to return to the Range Menu. Note, a higher range may be required in
areas with higher NOy concentrations.
11. Press RUN to return to the Run screen.
12. To set the averaging time, press the MENU button to return to the Main Menu. Press the DOWN
ARROW pushbutton till the cursor is on Averaging Time.
13. Press ENTER for the averaging time screen. Use the DOWN ARROW pushbutton to select the
desired averaging time and press ENTER. Press RUN to return to the Range Menu.
14. To set the correct time and date on the instrument, press MENU to return to the Main Menu. Press
the DOWN ARROW pushbutton till the cursor is on Instrument Controls. Press ENTER to display
the Instrument Controls screen.
15. Use the UP/DOWN ARROW pushbuttons to scroll through the choices. Select "Time" and press
ENTER
16. Use the UP/DOWN ARROW pushbuttons to increase/decrease the hours and minutes; use the
LEFT/RIGHT ARROW pushbuttons to move the cursor left and right. Set the appropriate time and
press ENTER. Press MENU to return to the Instrument Controls screen.
17. Select "Date" and press ENTER.
18. Use the UP/DOWN ARROW pushbuttons to increase/decrease the month, day, and year; use the
LEFT/RIGHT ARROW pushbuttons to move the cursor left and right. Set the appropriate date and
press ENTER. Press RUN to return to the Run screen.
19. The instrument is now set with the appropriate time, date and full scale range.
20. It is recommended that you allow the 42CY 24 hours before you attempt function checks or
calibration.
21. If the DAS system that you have does not have the RS-232 capabilities, then proceed to the next
section, Diagnostic Checks/Manual Checks. If you have connected the 42CY to a computer or DAS,
review the Diagnostic Check from your computer screen.
3.9.3.3 Diagnostic Checks/Manual Checks
To determine whether the 42CY is working properly, the field operators should perform the Diagnostic
Checks every time they visit the monitoring station. It is good practice for the operator to check these
Diagnostic Checks either by the computer or manually. Below are instructions on how to perform this
manually. Please note that the 42CY has set upper and lower ranges for some of these Diagnostic checks.
Please reference the owner's manual for these ranges.
1. To display the Diagnostics menu, from the Run screen press the MENU pushbutton to display the
Main Menu. Use the UP/DOWN ARROW pushbuttons to move the cursor to "Diagnostics." Press
ENTER for the Diagnostics screen.
2. Use the UP/DOWN ARROW pushbuttons to toggle through the function check tree. The following
table illustrates the functions that should be recorded. Please see Chapter 3 (page 3-36) 42CY manual
for more details. A manual checklist on maintenance is attached in Appendix A of this SOP.
3. On the Program Number screen, the version numbers of the program installed are displayed. Prior to
contacting the factory with any questions regarding the instrument, note the program numbers.
-------
Thermo 42CY-NOy SOP
Version No. 3
9/28/2005
Page 13 of 18
Table 3-2 Diagnostic Checks
Check
Voltages
Temperatures
Pressure
Flow
Test Analog Outputs
Explanation
The current DC power supply and PMT power supply voltages
The current internal instrument and chamber temperatures
The current chamber pressure
The current ozonator and sample flow rate
Enable analog outputs to be set to zero and full scale to adjust analog outputs to
agree with the front panel display
Reference the owners manual or the worksheet in Appendix A of this SOP for
acceptable limits. Once the Diagnostic checks have been established and recorded for
the 42CY, it is time to calibrate the instrument. Please refer to section 3.9.4 of this
SOP.
3.9.3.4 Preventive Maintenance
Preventive maintenance should prevent down-time and data loss. Table 3.3 lists the preventive
maintenance items that are listed in the 42CY manual, section 5.
Table 3-3 Preventive Maintenance Schedule the TECO 42CY
Item
Replace ozonator air feed drying column
Inspect and replace sample filters
Inspect and replace capillaries
Digital to analog converter test
Inspect and clean cooler fins
Inspect and clean fan filters
Schedule
As needed
Weekly
Quarterly
As needed
Semi-annually
Semi-annually
3.9.3.5 Instrument Troubleshooting
The 42CY manual has an excellent troubleshooting guide in Chapter 6. Please reference the manual for details on
troubleshooting the 42CY analyzer.
3.9.4 Calibration and Standardization
The calibration of the 42CY is performed by comparing the output of the instrument against standardized
gases of known quality. Generation of these gases is detailed in the "Calibration of Trace Gas Analyzers"
SOP. This section will detail how to adjust the 42CY to the standardized gases. Once the calibration has
been performed, compare the response of your DAS to the calculated "source" value. If this is outside of
+/-10%, then adjust the instrument response as detailed in the next sections.
3.9.4.1 Adjustment to Zero Air
In order to adjust the output of the 42CY to zero air, perform the following:
1. Connect the calibration line to a source of zero air. Supply a sufficient amount of zero air to
supply the analyzer and to ensure that a small amount of excess zero air exits from the sample
inlet.
-------
Thermo 42CY-NOy SOP
Version No. 3
9/28/2005
Page 14 of 18
2. Allow the analyzer to sample zero air for at least 15 minutes or until stable NO, NOy, and
NO2+NOZ readings are obtained.
3. From the Main Menu, choose Calibration. From the calibration menu, choose Calibrate Zero.
4. The Calibrate Zero screen displays the NO, NOy, and Prereactor readings.
5. Press ENTER to set the NO, NOy, and Prereactor readings to zero.
6. The message "Savings Parameters(s)" is briefly displayed to indicate that the background
readings have been set to zero.
7. Press MENU to return to the Calibration menu.
8. Press RUN to return to the Run screen.
3.9.4.2 Adjustment to Calibration Gas
In order to adjust the output of the 42CY to NIST traceable calibration gas, perform the following:
1. Switch the calibration unit to generate a known concentration of NO corresponding to
approximately 80% of full scale.
2. Supply a sufficient amount of calibration air to supply the analyzer and to ensure that a small
amount of excess calibration air exits from the sample inlet.
3. Allow the instrument to sample calibration gas for a minimum of 15 minutes, or until stable NO,
NOy, and NO2+NOZ readings are obtained.
4. From the Main Menu select the Calibration menu; select the "Calibrate NO" screen.
5. On the bottom line, there will be individual digits with which the span value can be set. In order
to change the span value, use the UP/DOWN ARROW pushbuttons to increase/decrease each
digit; use the LEFT/RIGHT ARROW pushbuttons to move the cursor left and right.
6. Change the span value to reflect the known concentration of NO in the calibration gas being
sampled.
7. Press ENTER to calibrate the NO reading to the NO calibration gas.
8. This operation changes the calculation equation and adjusts the NO span coefficient of the
instrument.
9. Press MENU to return to the Calibration menu.
10. Select "Calibrate NOy".
11. Change the span value to reflect the known concentration of NO plus any known NO2 impurity
in the calibration gas being sampled.
12. Press ENTER to calibrate the NOy reading to the NO calibration gas.
13. This operation changes the calculation equation and adjusts the NOy span coefficient of the
instrument.
14. Press MENU to return to the Calibration menu.
15. Adjust the calibrator to add a known concentration of ozone to the calibration gas corresponding
to approximately 60% of full scale. The ozone will react with the NO in the calibration gas to
form NO2.
16. Supply a sufficient amount of calibration air to supply the analyzer and to ensure that a small
amount of excess calibration air exits from the sample inlet.
17. Allow the instrument to sample calibration gas for a minimum of 15 minutes, or until stable NO,
NOy, and NO2+NOZ readings are obtained.
18. Select "Calibrate NO2" from the Calibration menu.
19. Change the span value to reflect the known concentration of ozone added to the calibration gas
plus any known NO2 impurity in the calibration gas being sampled.
20. Press ENTER to calibrate the NO2 reading to the NO calibration gas.
21. This operation changes the calculation equation and adjusts the NO2 span coefficient of the
instrument.
-------
Thermo 42CY-NOy SOP
Version No. 3
9/28/2005
Page 15 of 18
22. Press RUN to return to the Run screen.
3.10 Data Analysis and Calculations
Data analysis for this analyzer is detailed in "Data Acquisition and Management" SOP.
4.0 QUALITY CONTROL AND QUALITY ASSURANCE
The following section has brief definitions of the QA/QC indicators. Table 4-1 has the Measurement
Quality Objectives (MQOs) of the 42CY. Please note that this section details primarily with the data
quality indicators. Quality Control for continuous electronic instruments, such as the 42CY consists of
performing the diagnostic checks, maintenance and calibrations. These procedures are detailed in
sections 3.9.3 and 3.9.4: Instrument Operation, Startup and Maintenance and Calibration and
Standardization. Appendix A has an example of a Quality Control and Maintenance Record developed by
the EPA for this instrument.
4.1 Precision
Precision is defined as the measure of agreement among individual measurements of the same property
taken under the same conditions. For NOy, this refers to testing the NOy analyzer in the field at
concentrations between 0.2 and 100 ppb (note, higher test levels may be required in areas with higher
NOy concentrations). The test must be performed, at a minimum, once every two weeks. Calculations for
Precision can be found in Reference 4.
4.2 Bias
Bias is defined as the degree of agreement between a measured value and the true, expected, or accepted
value. Quantitative comparisons are made between the measured value and the true, standard value
during audits. Generally, three upscale points and a zero point are compared. Two audit types commonly
used for NOy, direct comparison and blind, are discussed below. The SOP should discuss plans for each
type of audit.
• Direct Comparison Audits: An independent audit system is brought to the monitoring location and
produces gas concentrations that are assayed by the monitoring station's NOy analyzer. In most cases,
a person outside of the agency or part of an independent QA group within the agency performs the
audit. The responses of the on-site analyzer are then compared against the calculated concentration
from the independent audit system and a linear regression is generated
• Blind Audits: In blind audits (also called performance evaluation audits); agency staff are sent an
audit device, such as the National Performance Evaluation Program (NPEP). The agency staff does
not know the NOy concentrations produced by the audit equipment. Responses of the on-site analyzer
are then compared against those of the generator and a linear regression is calculated.
4.3 Representativeness
Representativeness refers to whether the data collected accurately reflect the conditions being measured.
It is the data quality indicator most difficult to quantify. Unless the samples are truly representative, the
other indicators are meaningless. Since the NCORE Level I and II siting criteria are urban and regional,
the TL-NOy criteria are the same. Please reference the National Monitoring Strategy5 for a discussion of
NCORE Level II NOy monitoring scale.
-------
Thermo 42CY-NOy SOP
Version No. 3
9/28/2005
Page 16 of 18
4.4 Completeness
Completeness is defined as the amount of data collected compared to a pre-specified target amount. The
EPA does not have a minimum completeness requirement for NOy sampling. However, for NOX, EPA
requires a minimum completeness of 75% (40 CFR 50, App.H.3). Typical completeness with the 42CY
values can approach 90-93%.
4.5 Comparability
Comparability is defined as the process of collecting data under conditions that are consistent with those
used for other data sets of the same pollutant. The 42CY meets the MQOs for a Trace Level NOy
instrument. Please see Table 4-1.
4.6 Method Detection Limit
The method detection limit (MDL) or detectability refers to the lowest concentration of a substance that
can be determined by a given procedure. The 42CY must be able to detect a minimum value of 50 ppt of
NOy.
Table 4-1 Measurement Quality Assurance Objectives
Requirement
Bias
Precision
Completeness
Representativ
eness
Comparability
Method
Detection
Limit
Frequency
NCORE,
once per
year
1 every 2
weeks
Quarterly,
Annually
N/A
N/A
NA
Acceptance Criteria
Slope: 1.00+/-0.15
Intercept: +/- 3% of
full scale
Regression: <0.9950
Concentration: 0.2 -
600 ppb, Coefficient of
Variance less than
15%.
NCORE, 75%
Neighborhood, Urban
or Regional Scale
Must be a Trace Level
instrument. See
Sections 3.1 and 3.2 of
this document.
50 ppt
Reference
40 CFR
Pt.58
40 CFR
Pt.58
Appendix A
National
Monitoring
Strategy.
40 CFR 58
National
Monitoring
Strategy.
National
Monitoring
Strategy
Information or
Action
UseofNIST
generated gas
concentrations
with Mass Flow
Calibration unit
that is NIST
traceable
If CV is greater
than 15%, institute
Quality Control
measures
If under 75%,
institute Quality
Control Measures
N/A
N/A
Testing is
performed at the
factory.
-------
Thermo 42CY-NOy SOP
Version No. 3
9/28/2005
Page 17 of 18
5.0 REFERENCES
2. Merck Index, twelfth edition 1996, page 296
3. Seinfeld, John H., Atmospheric Chemistry and Physics of Air Pollution, 1986, page 54
4. Code of Federal Regulations, Title 40, Part 53.23c
4. Code of Federal Regulation, Title 40, Part 58, Appendix A
5. The National Air Monitoring Strategy, Final Draft, 4/29/04,
http: //www .epa. gov/ttn/amtic/monstratdoc .html
6. Model 42CY Instruction Manual, Thermo Environmental Instruments
-------
Appendix A
Environmental Protection Agency
Monthly Quality Control and Maintenance Records
Thermo 42C Trace Level NO/NOy Analyzer
Thermo 42CY-NOy SOP
Version No. 3
9/28/2005
Page 18 of 18
Site Name/Location:
Technician:
Date/Time:
Instrument:
Serial Number:
Range:
Date
Time
PMT voltage
+5 volt supply
+ 15 volt supply
Internal temp
Chamber temp
Cooler temp
Converter temp
Chamber pressure
Ozonator flow
Sample flow
Other Tests
Parameters Acceptance Range
5 +/- 0.5
15+/- 1.5
15-45 C
47-51 C
-20 - -1 C
300 - 350 C
200 - 400 mm Hg
0.05 -0.15 1pm
0.1 -2 1pm
-------
STANDARD OPERATING PROCEDURES
ENVIRONICS SERIES 9100
COMPUTERIZED AMBIENT MONITORING CALIBRATION
SYSTEM
Version 1
-------
Section 1.1 Acknowledgments
This Standard Operating Procedure (SOP) for the Environics Series 9100 Computerized Ambient
Monitoring Calibration System is the product of EPA's Office of Air Quality, Planning and
Standards. The following individuals are acknowledged for their contributions.
Principal Author
Anna Kelley, Hamilton County Department of Environmental Services, Cincinnati, OH 45219
On an Intergovernmental Personnel Act (IPA) assignment with OAQPS-EMAD, Research
Triangle Park, NC 27711
Reviewers
Office of Air Quality, Planning and Standards
Joann Rice, Precursor Gas Monitoring Team Lead, OAQPS-EMAD, Research Triangle Park, NC
27711
Michael Papp, QA Team Lead, OAQPS-EMAD, Research Triangle Park, NC 27711
Keith Kronmiller, Mantech, Inc. Research Triangle Park, NC 27711
Comments and questions can be directed to:
Joann Rice
EPA-OAQPS
Emissions, Monitoring, and Analysis Division
Mail Drop D243-02
Research Triangle Park, NC 27711
Email: rice.joann@epa.gov
(919)541-3372
-------
2.0 Table of Contents
Section Page
1.0 Title Page 1
1.1 Acknowledgements 2
2.0 Table of Contents 3
2.1 List of Figures 3
3.0 Scope and Procedures 4
3.1 Scope and Applicability 4
3.2 Summary of Method 4
3.2.1 Summary of Specific Procedures 4
3.2.2 Summary of Analyzer Operation 4
3.3 Definitions 5
3.4 Health and Safety Warnings 6
3.5 Cautions 6
3.6 Interferences 7
3.7 Personnel Qualifications 7
3.8 Equipment and Supplies 7
3.9 Procedures 9
3.9.1 Start-up 9
3.9.1.1 Unpack the Instrument 9
3.9.1.2 Warm-up 10
3.9.1.3 Leak Check 10
3.9.1.4 Flow Verification 11
3.9.2 Identifying Calibrator Ports to Standards 12
3.9.3 Calibrations - Multi-point 13
3.9.3.1 Determination of Concentrations for Multi-Point Calibration 13
3.9.3.2 Performing the Multi-Point Calibration 15
3.9.4 Verifications - Zero, Span and One-point Quality Checks 16
3.9.4.1 Set-up a Sequence 16
3.9.4.2 Set-up Timer Control 17
3.9.4.3 Run a Timer Control 18
3.9.4.4 Zero, Span and One-Point Quality Checks 18
3.9.4.5 Calibrator Automation 18
4.0 Quality Assurance and Quality Control 19
5.0 References 19
List of Figures
Figure 1.0 Flow Diagram: Environics 9100 5
Figure 2.0 Sampling System Set-up 7
Figure 3.0 Calibration Graph of Environics 9100 Mass Flow Controller 12
List of Tables
Table 1.0 Proposed Audit Level Concentrations 15
Table 2.0 MQO Table for Calibrator 19
-------
Environics 9100 SOP
Version No. 1
March 2, 2005
Page 4 of 19
3.0 Scope and Procedures
3.1 Scope and Applicability
This Standard Operating Procedure (SOP) is applicable for the Environics Series 9100
Computerized Ambient Monitoring Calibration System integrated with several ambient air
monitoring analyzers. The SOP outlines the steps to run calibrations and verifications either
automatically through programming sequences or through manual commands.
3.2 Summary of Method
3.2.1 Summary of Specific Procedures
This SOP will use the following methods to unpack, check and set-up, and execute a multi-point
calibration, and daily zero, span and one point QC checks.
3.9.1.1 Unpack the Instrument
3.9.1.2 Warm up the Environics
3.9.1.3 Leak check the Environics
3.9.1.4 Flow Verification
3.9.2 Assign the Calibrator Ports to the Gaseous Standards
3.9.3 Multi-point Calibrations
3.9.4 Verifications: Zero, Span and One Point Quality Checks
3.9.4.1 Setting up a Sequence
3.9.4.2 Setting up a Timer Control
3.9.4.3 Running the Timer Control
3.9.4.4 Zero, Span, and One-Point Quality Checks
3.9.4.5 Calibrator Automation
3.2.2 Summary of Analyzer Operation Refer to the Flow Diagram, Figure 1.
In the Environics 9100 Calibrator, Port 1 is connected to mass flow controller (MFC) 1.
Port 1 is used for Zero Air. MFC 2 can choose any of the remaining available ports.
When nitrogen dioxide (NO2) is required pollutant, ozone flows into the reaction
chamber along with gases from the other ports except 1. It is here that ozone reacts with
the gas. The resultant gas is then sent to the Mixing Chamber and is diluted with zero air
from Port 1 before exiting the calibrator and sent to the appropriate analyzer. If diluting
gases for CO and SO2, this process is by-passed.
-------
Environics 9100 SOP
Version No. 1
March 2, 2005
Page 5 of 19
Output
Port 2-5
Portl
Input pressure
20-30 psig
MF2
Critical
Orifice
Flow
Meter
Input pressure
30-35 psig
Mixing
Chamber
Backpressure
Regulator
Reaction
Chamber
9100 System Flow Diagram
3.3 Definitions
MFC
Seem
DAS
Span Check
One Point Quality Check
Figure 1.0 Flow Diagram: Environics 9100
Mass Flow Controller
Standard Cubic Centimeter per minute
Digital Acquisition System
a one point verification of the monitoring system challenging an
analyzer with a known concentration of gas and measuring the
analyzer response. Span checks are the highest concentration
test point approximately 80 to 90% of full scale range.
a one point verification of the analyzer made by challenging the
analyzer with a known concentration of gas and measuring the
analyzer response. This concentration ideally should be in the
range of daily observed acceptable quality. A response outside
the limits initiates a series of actions to determine data quality.
-------
Environics 9100 SOP
Version No. 1
March 2, 2005
Page 6 of 19
3.4 Health and Safety Warnings
1. Some gases can be explosive or otherwise reactive when blended. Users must check gas
compatibility before blending. Please consult a gas handbook, a specialty gas manufacturer or
other competent source for information about gas compatibility. Failure to observe these
precautions may result in damage to the instrument, serious injury or death.
2. Pressurized cylinders are extremely dangerous if improperly handled. Proper regulators, use
of safety caps and proper restraints are mandatory. Avoid cross contamination when attaching
regulators or making manifold connections. Always consult your gas supplier for proper
safety procedures. Failure to observe these precautions may result in serious injury or death.
3. Rules and regulations regarding the transportation of gas cylinders are governed by the
Department of Transportation. Each agency should familiarize themselves with these
regulations and follow them when involved in transporting gas cylinders. Failure to follow
these precautions may result in serious injury or death.
4. Secure Material Safety Data Sheets (MSDS) for all gases and keep in specially designated
binder in any easily accessible location.
5. Power to the unit should be disconnected before working on it to prevent injury or electrical
shock.
6. If it is necessary to work on the inside of the unit with power connected, extreme caution
should be taken. Failure to do so can result in injury or electrical shock.
7. Because the electronic circuit boards contain static sensitive components, always use static
discharge equipment when working on the unit and handling circuit boards.
8. A line should be run to the outside or a fume hood, whichever is more readily available, to
vent the outflow of gases from the entire system: Environics 9100, the Zero Air Generator,
and the analyzers. This should be done to prevent the possibility of a build-up of gases in a
closed room/monitoring station and the potential for gaseous poisoning to the unaware station
operator.
3.5 Cautions
1. The Environics 9100 operates at a line voltage between 100-120 VAC (optional 220-240
VAC). Operating at an incorrect line voltage will damage the instrument and void the
manufacturer's warranty.
2. Because the RS-232 ports and the parallel printer port are especially susceptible to damage if
the unit is powered ON while any computer or printer equipment is being connected or
disconnected, the unit must be powered OFF before connecting or disconnecting any cables,
wiring harnesses or other sources of potential electrical impulses.
-------
Environics 9100 SOP
Version No. 1
March 2, 2005
Page 7 of 19
3. If routine maintenance is performed, prior to putting the unit back into service check the
following items:
• Inspect the power cord and internal wiring
• Check all fittings for tightness and leak check if possible
• Clean circuit boards using a vacuum or air gun to remove dust
• Check all tubing for splits, kinks, or cuts
3.6 Interferences
The calibrator as well as the zero air supply and any gaseous cylinders should be stored and
operated in a temperature and humidity controlled shelter as extremes of either can effect the final
concentration. Condensation of humidity in the sampling and associated lines can also dilute the
concentration of gases delivered to the analyzers.
3.7 Personnel Qualifications
Persons setting up and operating this system should have an understanding of gaseous systems
and/or a good background in chemistry, physics, and the scientific process. For someone
unfamiliar with gaseous systems and/or setting up analyzers and related equipment at a
monitoring site, an instructional course would be beneficial. Other sources of instruction include
any US EPA guidance documents and training with other experienced personnel.
3.8 Equipment and Supplies
Sampling Manifold
Zero Air Generator
Environics 9100
Calibrator
Pollutant Analyzer
> S Fan
Cal Standard #1 Cal Standard #2
Figure 2.0: Sampling System Set-up
-------
Environics 9100 SOP
Version No. 1
March 2, 2005
Page 8 of 19
Environics Series 9100: used to blend gases to desired concentration to run check calibration of
ambient air analyzers through not only a calibration of each but also through the daily zero, span,
and one point quality checks.
Zero Air Source: For this study, an API Model 701 Zero Air Generator is used. Carbon monoxide
and hydrocarbon scrubbers are available as an option when purchasing. For precursor gas
applications, the CO scrubber is required. Please see the manual for a recommended maintenance
schedule on the unit. The purpose of the zero air source is to dilute the gases to achieve the
desired final concentration of gaseous pollutants in question. Zero Air should be filtered before
entering the calibrator to ensure it is free of any particulate contaminants. A cylinder of zero air
may also be used. However, the latter may not be cost effective or practical for monitoring
organizations given the frequency of use of the zero air supply. It is also imperative for the
monitoring organization to verify/recertify their zero air supply annually using an independent
zero air source. The independent source should be of a higher quality than what is routinely used.
Most high purity cylinder air sources are not of the quality needed as an independent source.
High quality cylinders can be used, but they require external scrubbers for the precursor gases of
interest.
It is not recommended that the API 701 be used to generate air at a flow rate greater than 10
LPM. The internal CO scrubber is not as efficient at the higher flow rates (e.g., 20 LPM);
therefore, caution should be taken when exceeding 10 LPM. Decreased scrubber efficiency
presents itself as an increased CO analyzer baseline. An external Hopcalite or Carulite scrubber
can be added to the API 701 to remove any CO not eliminated by the scrubber at higher flow
rates. CO scrubbers can fail. The manual recommends replacement on an annual schedule
depending on CO levels in the source air.
It is also recommended that ambient air, not air from the inside of the monitoring station, is used
as a source for the API zero air generator.
Reagents and Standards: Gas Cylinders are used as the source of the target pollutant
concentration. Cylinders used should be of the highest purity available, specifically for multi-
point calibrations and performance evaluations. All gases used should be either NIST traceable or
EPA Protocol Gases having acceptance criteria of 2%. Three different gas cylinders were used in
the initial testing project. The concentrations used to obtain the desired blend for the routine span
and precision checks follow:
Concentration of gases, cylinder one:
• NO 50 ppm nominal in N2 balance @ 2000 psig Tolerance: ±1% EPA
• CO 5000 ppm nominal in N2 balance @ 2000 psig Tolerance: ±1% EPA
• NOX 50 ppm nominal in N2 balance @ 2000 psig
• NO2 <0.2 ppm in N2 balance @ 2000 psig
• SO2 50 ppm nominal in N2 balance @ 2000 psig Tolerance: ±1% EPA
Concentration of gases used in cylinder two:
-------
Environics 9100 SOP
Version No. 1
March 2, 2005
Page 9 of 19
• CO 350 ppm nominal in N2 balance @ 2000 psig Tolerance ± 1% EPA
• NO 10 ppm nominal in N2 balance @ 2000 psig Tolerance ±1% EPA
• NO2 <0.05 ppm in N2 balance @ 2000 psig
• NOX 10 ppm nominal in N2 balance @ 2000 psig
• SO2 10 ppm nominal in N2 balance @ 2000 psig Tolerance ±1% EPA
Cylinder three (span and precision checks)
• CO 60 ppm nominal
Targeted concentrations of the span and precision checks:
Pollutant Range Span Precision Check
CO 5000 ppb 4000 ppb 250-500 ppb
SO2 100 ppb 80 ppb 20 ppb
NOy 100 ppb 80 ppb 20 ppb
Data Acquisition, Storage and Communication Package: A data acquisition package is
necessary to retrieve the data from a remote site and store it for further analysis. Two data
acquisition systems were used in this project: Environmental Data Acquisition System (EDAS)
for Windows both models 8816 and 8832, Environmental Services Corporation and Envidas, by
Envotech. For further information, please refer to the Data Management SOP.
3.9 Procedures
3.9.1 Start-up
3.9.1.1 Unpack the instrument
1. Remove the unit from the shipping container and inspect for damages.
2. Note any damage to the shipping case and report to freight carrier immediately.
3. Removing the screws on each side of the top removes the top cover.
4. Inspect the interior of the calibrator for any loose parts or visible damage.
5. Check for any loose circuit boards. If loose, press down to reseat them before connecting the
power.
6. Report any instrument damage to Environics or the local distributor.
7. If no damage is found, replace the cover and screws.
The above procedure should be done upon immediate receipt of the instrument or can be moved
to the location the calibrator will be sited and then inspected for damage. However, do not wait
too long to do the inspection to ensure validation of the warranty period. Please refer to inside the
front cover of the manual for specifics on the warranty period and specific coverage.
-------
Environics 9100 SOP
Version No. 1
March 2, 2005
Page 10 of 19
3.9.1.2 Warm-up
When bringing zero air supply and calibrator together for initial checks, suggested warm up time
is 24 hours. If moving to sampling location after initial checks, suggested warm up time would be
3-4 hours.
1. Set-up Calibrator in place and connect zero air supply to Port 1.
Note: Pressure of zero air supply should be between 30-35 PSI
2. Secure the appropriate electrical power supply to the Environics and switch to the "ON"
position
3. Allow calibrator and zero air supply to warm-up and stabilize before proceeding.
3.9.1.3 Leak Check Refer to Section 5.14 in the Environics Manual for further information on
the Leak Check
A leak check of the calibrator should be performed before beginning other work. A leak in the
calibrator will result an incorrect dilution of gases delivered to analyzers yielding incorrect
calibration and verifications of concentrations by the analyzer(s). It is strongly recommended to
perform the leak check even if the calibrator is brand new. This ensures integrity of the system
and prevents aggravation to the operator.
1. Set the pressure of the zero air supply to 10 PSI
2. At the READY screen, press the MORE key (F8) in the lower right hand corner of the
screen. This will take you to the second screen.
3. Once on the second screen, press the LEAK CHECK key (F7)
4. Adjust the PSI reading on the external pressure gage on the front of the zero air supply source
as necessary.
Note: The PSI should NOT be above 13. A PSI over 13 will activate the pressure safety.
5. Press the START key (F1) on the front panel.
6. Allow the Environics to run through the leak test cycle. The factory acceptance criterion for a
leak check is less than 3 seem.
Automatic sequence of events during the LEAK CHECK cycle:
1. Acquire atmospheric pressure (approximately 59 seconds)
2. Pressurize system - calibrator and zero air supply (approximately 89 seconds)
3. Shut down solenoids in calibrator - actual leak check of the calibrator and zero
air supply
Note: The test may be cancelled at any time by pressing any key.
7. Once the LEAK CHECK is complete, the Environics will display the results. If less than 3
seem, the results are acceptable.
8. Record results in appropriate lab notebook
9. Press the EXIT (F8) key located in the lower right hand corner of the screen.
10. Adjust the zero air supply to read 30 PSI on the external pressure gage
-------
Environics 9100 SOP
Version No. 1
March 2, 2005
Page 11 of 19
3.9.1.4 Flow Verification
Although the mass flow controllers have been factory set, it is strongly recommended to verify
the flows through the calibrator using aNIST traceable source. This ensures accuracy as the flow
measurement standard must be more accurate than the specified flow accuracy of the MFC. For
this study, a BIOS Dry-Cal device was used. Flow devices must be certified annually. Flow
checks should be verified at the start-up of the sampling system. Additional verifications should
be performed annually or when the gaseous pollutant flow is questioned.
To complete flow verification of Port 1, zero air, and Port 2, cylinder gases, a BIOS Dry-Cal
device was used. If another type of flow device is used, these procedures may differ. Consult the
manual that accompanies that device for specific instructions.
1. Attach the appropriate BIOS cell to the base of the dry-cal.
2. Attach tubing to the exhaust end of the dry-cal.
3. Attach the calibration output tubing to the dry-cal. Note: the tubing from the calibrator output
may not be the correct size to connect to the input of the dry-cal. Therefore, it may be
necessary to have a small adapter/connector between two different sizes of tubing.
4. Turn on dry-cal
5. From the READY screen on the front panel of the Environics calibrator, press the CONC
MODE (Fl) soft key
6. When the Concentration Mode Screen appears, move cursor using the arrow keys on the soft
key pad to Port 1.
7. Enter Port 1 and the cursor will move to the Target Gas area
8. Once in the Target Gas area, the concentration should be 0 for all gases. If not, change to 0 in
those pollutant concentration fields as needed using the numeric key pad and the arrow keys.
9. After work in Target Gas area is complete, cursor should be in Total Flow area. If not move
cursor to that area and enter total flow desired: 5.0 SLPM.
10. Move the cursor to another field using the arrow keys. This will update and confirm on the
screen the current requested information.
11. Press START (F1) on the soft key pad at the bottom of the screen. This initiates the calibrator
to open the appropriate solenoids and start the air flow through the calibrator.
12. Observing the display on the dry cal and the front of the calibrator, begin to record the actual
readings from the dry cal when the flows are constant on both. It is recommended to record
eight to ten readings.
13. After the readings are completed for a specific flow, select STOP (F ) on the bottom of the
front panel of the calibrator, repeat steps 9-12. Five or more Target Flows up to 20 LPM
should be used to create a line.
14. When completed with Port 1, EXIT (F8) to the READY screen and switch to the cell for
flows on the dry cal device. The flow of Port 2 will now be verified.
15. At the READY screen, select the FLOW MODE (F2) at the bottom of the screen.
16. Cursor should appear at Port 1. Move to Target Flow area and enter 0 SLPM.
17. Move cursor to Port 2 and enter desired SCCM: 90.
18. Move the cursor to another field using the arrow keys. This will update and confirm on the
screen the current requested information.
19. Turn on dry-cal
-------
Environics 9100 SOP
Version No. 1
March 2, 2005
Page 12 of 19
20. Select the START key (Fl) on the soft key pad at the bottom of the screen. This initiates the
calibrator to open the appropriate solenoids and start the air flow through the calibrator.
21. Observing the display on the dry cal and the front of the calibrator, begin to record the actual
readings from the dry cal when the flows are constant on both. It is recommended to record
eight to ten readings.
22. After the readings are completed for a specific flow, select EXIT (F8) on the bottom of the
front panel of the calibrator, repeat steps 17, 18, and 20. Five or more Target Flows down to
10 SCCM should be used to create a line.
23. When completed with Port 2, EXIT (F8) to the READY screen.
24. Turn off dry-cal, disconnect tubing, and replace all tubing to original set-up as necessary.
25. Average each of the flow points.
26. Perform a linear regression analysis and plot the actual vs. target concentration flows.
Environics M FC 1 ZA Calibration 12/16/04
LPM (less 20 Ipm setpoint)
Figure 3: Calibration Graph of Environics 9100 Mass Flow
Controller
3.9.2 IDENTIFYING THE CALIBRATOR PORTS TO THE GASEOUS STANDARDS
Refer to Section 5.8 in the Environics Series 9100 Manual
The purpose of this section is to assign gaseous cylinders to a specific port located on the rear of
the Environics 9100 Calibrator.
1 From the READY SCREEN, select MAINTAIN PORTS (F6)
2. Using the soft number key pad on the right front screen, choose the desired gas port to be
configured. Once a port number is selected, the cursor will appear in front of the word
BALANCE
3. Using the Right Arrow key, move to the Gas ID column. Lines 1 will be used to name the
BALANCE of the gas in the cylinder. Lines 2-6 will be used to name the individual gases in
that cylinder.
4. Enter the first gas to be configured on line 2
-------
Environics 9100 SOP
Version No. 1
March 2, 2005
Page 13 of 19
Note: If it is intended to use nitric oxide to perform a Gas Phase Titration, "NO" must be
used as the GAS ID. The system will compute the NO2 while running. System recognizes
NO only for this process.
5. Use the "Exit" key to return to the CGC area or use the "Up and/or Down" keys to move to a
different CGC field.
6. Enter the CGC for each GAS ID
Note: The cylinder concentration can either be entered as parts per million (ppm)
or percent (%).Percent of a concentration is more typically used for higher
blending higher concentration of gases.
Note: Concentrations of up to five (5) gases may be added for one multi blend gas
cylinder. Beneath the CGC, the first cell is labeled BALANCE. If using a multi blend gas
and contains nitric oxide (NO), this must be the first gas entered.
7. Repeat steps 2-4 for all gases to be entered for that cylinder
8. Press CYL ID (F6), to move to cylinder ID field and enter a name for that cylinder.
Suggestion: use the cylinder number listed on the cylinder tag
9. After completing set-up for a desired port, press EXIT (F8) to select a different port. Press
EXIT(F8) again to go to the Main Menu.
3.9.3 CALIBRATIONS - Multi-point
Running a multipoint calibration, a zero point and three separate concentrations are to be run to
determine linearity of the instrument. The proposed audit ranges concentrations found in 40 CFR
Part 58 Appendix A Section 3.2.2 combined with actual ambient concentrations observed would
serve as a guideline of concentrations to choose for a multi-point calibration. The recommended
sequence when performing the multi-point calibration is to perform a zero first, making any
necessary adjustments to the zero, then move to the highest concentration working backwards
ending with the low concentration and a final zero. Performing the points in this order more
completely coats the exposed surfaces allowing for a more stable concentration. An added benefit
to performing the concentration points in this sequence is the time savings element.
3.9.3.1 Determination of Concentrations for the Multi-point Calibration
Before performing the actual multi-point calibration, calculate the desired concentration of each
point based upon the gas cylinder concentrations and gas flow. A recommendation: calculate
these values before purchasing your pollutant gas standards to ensure having the desired
concentrations when ready to perform the multi-point calibration.
Based on the flow verification of the mass flow controllers in the calibrator performed earlier, use
the actual flow of the zero air and gases in the multi-blend cylinder to calculate what
concentrations are needed for each pollutant for the multi-point calibration. Refer to Example #1
as you follow the step by step process.
1. Determine the range of the each pollutant, example CO is 0-5000 ppb
-------
Environics 9100 SOP
Version No. 1
March 2, 2005
Page 14 of 19
2. Next, determine 10 and 90 percent of the range, 500 and 4500 ppb respectively. By solving
for the minimum and maximum concentrations, you will be able to determine what
concentration of pollutant gas cylinder(s) is needed.
Note: It will be necessary to determine your total air flow needed through your system.
This will be based on the set-up of the entire sampling system including manifold.
3. Using the following equation and referring to Example #1, Steps 2 & 3, determine what flow
is needed to generate these concentrations.
Concentration of pollutant in cylinder x pollutant flow = desired concentration
Total flow through system: pollutant flow + zero air flow
4. In this example, make the pollutant flow, the unknown value and solve the equation for it.
5. Once the 10 and 90 percent of range have been determined, solve for several more
concentration values between those two.
Example #1 - Calculations for actual one point quality checks
Actual flow of gases through the system:
Zero air: 6.843 = 6.999.6 ccm; gaseous concentrations: 100 ccm = 98.05 ccm, 16.73 ccm = 13.63
ccm
Step 1: Zero air is run through the system to flush out any sample air before beginning the
precision check
Step 2: Zero air is mixed with a gas containing
X concentration of CO at a flow of 100 ccm:
10.07 ppm of CO resulting concentration of CO: 139.1 ppb
[10.07 ppm x 98.057 (98.05 + 6999.6)] x 1000 = 139.1 ppb
X concentration of SO2 at a flow of 100 ccm:
13.13 ppm of SO2 resulting concentration of SO2: 181.4 ppb
[13.13 ppm x 98.057 (98.05 +6999.6)] x 1000 = 181.4 ppb - Over Range
X concentration of NO at a flow of 100 ccm:
10.26 ppm of NO resulting concentration of NO: 141.7 ppb
[10.26 x 98.057 (98.05 +6999.6)] x 1000 = 141.7 ppb - Over Range
Step 3: Zero air is mixed with a gas containing 13.13 ppm SO2, 10.26 ppm NO, and 10.07 ppm
CO resulting in 25.5 ppb of SO2, 19.9 ppb NO, and 19.6 ppb.
X concentration of CO at a flow of 16.73 ccm:
10.07 ppm of CO resulting concentration of CO: 19.6 ppb
[10.07 ppm x 13.63.7 (13.63 +6999.6)] x 1000= 19.6 ppb - Under Range
X concentration of SO2 at a flow of 16.73 ccm
-------
Environics 9100 SOP
Version No. 1
March 2, 2005
Page 15 of 19
13.13 ppm of SO2 resulting concentration of SO2: 25.5
[13.13 ppm x 13.637 (13.63)+ 6999.6)] x 1000 = 25.5 ppb
X concentration of NO at a flow of 16.73 ccm
10.26 ppm of NO resulting concentration of NO: 19.9 ppb
[10.26 ppm x 13.637 (13.63)+ 6999.6)] x 1000 = 19.9 ppb
Step 4: Zero air only is run through the system
Audit Level
1
2
3
4
5
03
0 02-0 05
0 . 06-0 . 10
0 . 11-0 . 20
0 . 21-0 . 30
0 . 31-0 . 90
Conce
S02,
0 0003-0 005
0 . 006-0 . 01
0 . 02-0 . 10
0 . 11-0 . 40
0 . 41-0 . 90
sntration Range,
NO2
0 0002-0 002
0 . 003-0 . 005
0 . 006-0 . 10
0 . 11-0 . 30
0 . 31-0 . 60
PPM
CO
0 08-0 10
0 . 50-1 . 00
1 . 50-4 . 00
5-15
20-50
Table 1.0 Proposed Audit Levels, 40 CFR, Part 58 Appendix A, Section 3.2.2
Example concentrations used for Multipoint Calibration
CO
421 ppb
855 ppb
1078 ppb
2660 ppb
2990 ppb
3534 ppb
4081 ppb
SO2
8.69 ppb
18.57 ppb
40.56 ppb
59.98 ppb
82.5 ppb
93.0 ppb
NO/NOy
12.54 ppb
28.6 ppb
50.04 ppb
71.3 ppb
92.28 ppb
3.9.3.2 Performing the Multi-point Calibration
1 From the READY screen, select Concentration Mode screen. This goes into the MANUAL
Mode of Calibrations. Calibrations will occur for each analyzer and each point at a time. The
cursor will appear at the first TARGET GAS concentration.
2. Enter the desired gas concentration in the TARGET GAS area by moving the cursor using
the arrows.
3. Using the arrow keys, move to the TOTAL FLOW area and enter the flow, 10 LPM
4. Using the arrow keys, cursor to the OZONE area and enter desired concentration if applicable
for the pollutant needed for NO2
5. Press START. This will initiate the blending of the gases to the analyzer.
6. Allow the analyzer to stabilize.
7. Once stable, wait ten minutes, record the reading for that concentration.
8. Continue to the next concentration level repeating steps 1 through 5.
-------
Environics 9100 SOP
Version No. 1
March 2, 2005
Page 16 of 19
9. The above steps can be automated by programming each step into a Sequence and assigning a
designated time to Run.
3.9.4 VERIFICATIONS - Zero, Span and One Point Quality Checks
These steps will assess and confirm the upper range of the multi-point calibration, the zero base
line and the ambient working range of the instrument. The one point quality checks will
determine both precision and bias of the analyzers.
This step may be either automated to occur as directed or manually. If performing a manual
operation, follow the same procedure in Section 3.9.3.2 Performing the Multi-Point Calibrations.
To operate in AUTOMATIC Mode, it is first necessary to program the Sequences and name the
Programs.
3.9.4.1 SET-UP A SEQUENCE Refer to Section 5.6.1 in the Environics Series 9100 Manual
In this step, the steps necessary to automate the zero and span check will be set up. Below
is an automated Sequence used in the project. Refer to Example # 1 for a detailed
explanation of each step.
Zero
Run Time Mode Port Air Flow (ccm) Gas (ccm) Total O3 Gas
ID Cone.
1 10 (min) Cone. 2 6.843 0.0 6.843 0.0 NO 0.0
2. 20 Cone. 2 6.843 100 6.9430.0 NO 0.142
3. 20 Cone. 2 6.843 16.73 6.85970.0 NO 0.025
4. 10 Cone. 2 6.843 0.0 6.843 0.0 NO 0.0
1. From the READY, select Program Mode (F4)
2. Once in Program Mode, select Sequence (F2)
3. At the blank screen, the cursor is in the RUNTIME column
4. Enter the desired run time for the first step Refer to Row 1-10 minutes
Note: If "0" is entered, the step will be skipped and not displayed.
Valid run times are 1 to 60 minutes.
Twenty lines (20) are available for programming one SEQUENCE
5. Move the cursor to the MODE column. Using the function keys, choose the appropriate
MODE
Note: For zero, span, and precision checks use Concentration (CONC) Mode
6. Move the cursor to the PORT column; enter the desired span gas port which has been
established in section 3.9.2 (In this case, it is not necessary to use Port 1. Using Port 2 and
telling it not to send any gas from port 2 serves the same purpose as using Port 1 only)
7. Move to the Total Flow column and enter the desired flow.
8. Move to the Gas ID column; when at this column, the GAS ID should default to those
provided in section 3.9.2
Note: Gas ID number refers to the order of gases entered in the
MAINTAIN PORTS screen
-------
Environics 9100 SOP
Version No. 1
March 2, 2005
Page 17 of 19
9. Move to CONC column; enter desired concentration for that gas. Enter concentration in ppm.
Note: Moving cursor from column to column initiates a check to
determine if the value entered is legal. If the value is not possible, an Error
message is displayed at the bottom of the screen.
Note: When using the CONC MODE, the calibrator will automatically
calculate the amount of gas needed to produce the required concentration. It is
considered Good Laboratory Practice to check to determine calculations are correct. This
can be a random check on one or more of the calculations.
10. Follow steps 3-9 for additional rows in the Sequence. Example 1 has four rows: zero, one
quality check - CO; one point quality check - NO and SO2, and a zero.
11. After entering all steps in a specific sequence, SAVE (F2) the Sequence in the Register. Use
the function keys to type the name of the Sequence. We'll call Example # 1 PC-1 and saved
as sequence line number 01
12. When the SAVE is complete, press EXIT (F8) to return to the Program Mode
13. Pressing EXIT again, returns to the READY screen.
3.9.4.2 SET-UP TIMER CONTROL (TIMER CTL) Section 5.6.2 in the Environics Series
9100 Manual
This is used to automate a number of sequences which allow for automating zero, span and one-
point quality checks.
1. If this is the first set-up and no previous Timer CTL are stored in the register, the cursor will
be in the Sequence line and it will be necessary to enter the desired sequence for the specific
day(s) and time (s) the Timer CTL
1. From the Main Menu, select Program Mode
2. From Program Mode, select TIMER CTL
3. Go to the day you want a specific Sequence to Run.
4. Select the Sequence, 01, to be run.
5. Enter the time to start the Sequence. Once the time is entered, the system will display the
duration of the Sequence and calculate the end time.
Note: Time is based on a 24 hour clock with 12:00 being noon and 24:00
being midnight. Entering 0:00 will cause the Sequence to be skipped and
will not be displayed; therefore must enter time as low as 00:01
Note: Three sequences per day may be entered for each day of the week.
6. Once the desired sequences are in the desired order, press the SAVE (F2) key.
7. Enter the desired register number. A name may also be entered.
8. Press EXIT (F8)
-------
Environics 9100 SOP
Version No. 1
March 2, 2005
Page 18 of 19
3.9.4.3 RUN TIMER CTL
1. If a program has been previously saved press the RECALL (F3) key to bring the program to
the screen
2. Press the START (Fl) key.
Note: The day and time of the next sequence will be displayed in the lower
right hand corner of the screen.
3.9.4.4 Zero, Span, and One-Point Quality Checks
1. At the Concentration Mode, Manual Operation screen, enter the TARGET GAS
concentration for the desired gas. The system will determine the actual flows for each port to
achieve the requested gas concentration(s). Remember: Port 1 is Zero Gas.
2. Once the TARGET GAS concentrations have been set for the desired gas, press the START
(Fl) key, to begin the operation.
Note: A similar process is to go to the FLOW MODE and enter the desired flow. The
Environics will calculate the actual concentration based on the flow.
If the desire is to automate the zero, span and one point quality checks, set-up a Sequence and
SAVE the sequence. Then to Run the Sequence, go to Timer Control and set the desired time to
run your Sequence. Below is an example of a Sequence set-up for Precision Checks used in the
testing program. Refer to previous sections 3.9.4.2 and 3.9.4.3. Note: Atri-blend standard was
used in the testing process. An example of the gases used and the concentrations are found on
pages 8 and 9.
3.9.4.5 Calibrator Automation
A Data Acquisition System can control the running of a programmed sequence. To use this
function, the Environics calibrator must have the optional status board. The INPUTS option
provides the ability to trigger various actions within the system. The OUTPUTS option will
signal external devices when certain operating conditions occur. In order for the data acquisition
system to run the calibrator, the INPUTS option is needed.
1. Using the key pad, press the STATUS SETUP - INPUT soft key at the Main Menu screen.
2. At the INPUT screen add the desired function(s), Sequence, Timer Control, etc to be run in
the Status Line. The desired function(s) can be chosen from the soft keys at the bottom of the
screen.
3. When the desired function(s) have been entered, select the EXIT soft key returning to the
Main Menu and the READY mode.
Note: the calibrator must be in the READY mode for the automation to occur
-------
Environics 9100 SOP
Version No. 1
March 2, 2005
Page 19 of 19
For further detailed information on communications between the data acquisition system, calibrator and
analyzers, refer to the Data Acquisition SOP.
4.0 Quality Control and Quality Assurance
After confirming the flows through the calibrator, a multi-point calibration for each analyzer should be
performed. A different certified gas cylinder should be used other than the cylinder used for span and
precision check gas.
Operating the MFC below 10% of its rated full-scale flow may result in flow inaccuracies.
The system may produce gas blends that fall outside the accuracy specifications of the system if the gas
pressures do not stay within the ranges specified for each gas port.
The operator of the system should not be the same person who performs the multi-point calibrations, but
MQO Table for Environics 9100 Calibrator
Criteria
Mass Flow Controllers
Zero Air Source
Acceptable Range
Minimum Frequency
Reference
in the case of the Trace Gas Analyzers, it could be.
5.0 References
1. 40 CFR Part 58, Appendix A
2. Environics Series 9100 Computerized Ambient Monitoring Calibration System, Operating Manual,
Revision 6, October 1999
-------
United States Office of Air Quality Planning and Standards Publication No. EPA-454/R-05-003
Environmental Protection Emissions, Monitoring and Analysis Division September 2005
Agency Research Triangle Park, NC
------- |