&EPA
Part 75 CEMS Field Audit Manual
Clean Air Markets Division
U.S. EPA
Washington, DC 20014
July 16, 2003
CLEANAIR
MARKET PROGRAMS
-------
-------
Acknowledgments
This Part 75 CEMS Field Audit Manual is the result of past U.S. EPA documents, as
well as significant support and reviews to provide updated information. The Clean Air Markets
Division (CAMD) of the U.S. EPA had prepared an "Acid Rain Program CEMS Field Audit
Manual" to assist with auditing continuous emission monitoring systems (CEMS) installed
under 40 CFR Part 75. In addition, Joseph Winkler, U.S. EPA Region VI, with contractor
assistance from Gerhard Gschwandtner, Comprehensive Monitoring Services, Inc., had
prepared a separate "Acid Rain Program: Continuous Emission Monitoring Systems Reference
Manual." Dr. James Jahnke, of Source Technology Associates, prepared EPA's 1994 document
entitled "An Operator's Guide to Eliminating Bias in CEM Systems" (EPA 430-R-94-016).
Additional references used in developing this manual are listed at the end of Section 2. The
Clean Air Markets Division acknowledges the authors of these existing documents for
providing a valuable foundation for preparing this document.
The CAMD team leader for this manual was Matthew Boze. Other CAMD staff
provided valuable support and comments, including Louis Nichols and others in the Emission
Monitoring Branch. Contract support for this document was provided by Perrin Quarles
Associates, Inc. under Purchase Order No. 2W-1382-NTSA. James Jahnke, Ph.D., Source
Technology Associates, and James Peeler, Emission Monitoring, Inc., provided a technical
review of a draft of this manual. Other reviews were provided by Joseph Winkler, EPA Region
VI, and Errin Pichard, Florida Department of Environmental Protection.
-------
Disclaimer
Any mention of trade names or commercial products in this document is not intended to
constitute endorsement or recommendation for use.
-------
July 16, 2003
Part 75 CEMS Field Audit Manual
Table of Contents
Page
Section 1: Introduction 1
1.1 Background 1
1.1.1 Importance of Monitoring for Emission Trading Programs 1
1.1.2 Structure of Part 75 Monitoring Provisions 3
1.2 What does the manual cover? 6
1.3 Part 75 Audit Program Overview 8
1.3.1 Part 75 Electronic Audit Program 8
1.3.2 Audit Targeting 8
1.3.3 Field Audits 8
1.4 Role of the Inspector 9
1.4.1 "Hands Off' Approach 10
1.4.2 Inspection Safety 11
1.4.3 Recommended Training Courses 11
1.5 Key Part 75 Materials with Internet Links 12
Section 2: Part 75 CEMS Overview 15
2.1 Introduction 15
2.2 Sampling Location 17
2.2.1 Gas Measurement Location 18
2.2.2 Flow Measurement Location 18
2.2.3 Sampling in Stratified and Swirling Flow Conditions 18
2.3 Extractive Gas CEMS 19
2.3.1 Source Level or Direct Extractive Systems 19
2.3.1.1 Sample Probe 20
2.3.1.2 Sample Transport and Conditioning Systems 22
2.3.2 Dilution Extractive Systems 23
2.3.2.1 Dilution Probe 24
2.3.2.2 Out-of-Stack Dilution System 25
2.3.2.3 Sample Transport and Dilution Air Cleaning Systems 26
2.3.3 Calibration Gas System 27
2.3.4 Analyzers 27
2.3.4.1 Absorption Spectroscopy 28
2.3.4.2 Luminescence Spectroscopy 29
2.3.4.3 Electro-Analytical Methods 30
2.3.4.4 Paramagnetic Techniques 30
2.4 In-Situ Gas Monitors 31
2.4.1 Path In-Situ CEMS 32
2.4.2 Point In-Situ CEMS 32
2.4.3 In-Situ Gas Analyzers 32
2.4.4 System Calibration 33
Part 75 Field Audit Manual - My 16, 2003 Page i
-------
Table of Contents (cont.)
Page
2.5 Flow CEMS 33
2.5.1 Sampling Location 34
2.5.2 Differential Pressure Flow Monitors 34
2.5.3 Thermal Mass Flow Monitors 35
2.5.4 Ultrasonic Flow Monitor 37
2.6 Data Acquisition and Handling System (DAHS) 38
2.6.1 CEMS Computer Systems 38
2.6.2 Emissions Data Processing 39
2.6.3 QA Test Data Processing 39
2.6.4 Part 75 Reporting 39
2.7 References 40
Section 3: Audit Preparation 41
3.1 Using Part 75 Electronic Data to Conduct Pre-Audit Reviews 41
3.1.1 Quarterly Electronic Data Reports 41
3.1.2 Quarterly Feedback Reports 44
3.1.3 Using MDC to Prepare for an Audit 44
3.1.3.1 MDC Overview 44
3.1.3.2 Getting Started with MDC 45
3.1.3.3 Review and Print Electronic Portion of Monitoring Plans ... 46
3.1.3.4 QA Tests, Exemptions, and Extensions 47
3.1.3.5 Recertification Events and Monitoring System Downtime
Reports 50
3.1.3.6 Using MDC Hourly to Check Emissions Data
and Calculations 50
3.2 Hardcopy File Review 53
3.2.1 Correspondence, Petitions, and Previous Audit/Inspection Reports ... 53
3.2.2 Linearity Test and RATA Reports 54
3.2.3 Permits 54
3.3 Scheduling and Coordinating the Audit 55
3.4 Materials to Bring 55
Section 4: On-Site CEMS Inspection 57
4.1 Pre-Audit Interview 58
4.2 Calibration Error Test 59
4.3 Probe/Sensors, Sample Lines, and Sample Conditioning Systems 62
4.3.1 Probe 62
4.3.2 Sample Lines 63
4.3.3 Dilution Air and Gas Sample Conditioning Systems 64
4.3.3.1 Dilution Extractive Systems 64
4.3.3.2 Source Level Extractive Systems 66
4.4 Gas Analyzers 67
68
4.5 Calibration Gases 68
4.6 Flow Monitors 71
Page ii Part 75 Field Audit Manual - My 16, 2003
-------
Table of Contents (cont.)
Page
4.7 DAHS 73
4.7.1 DAHS Certification and Verification Tests 74
4.7.2 Changes in Correction Factors 74
4.7.3 Manually Entered Data 74
4.8 Maintenance Log and Daily Checklists Review 76
4.9 QA/QC Plan Review 78
Section 5: CEMS Performance Test Observation 81
5.1 Introduction 81
5.2 Linearity Test 82
5.3 Relative Accuracy Test Audit (RATA) 83
Section 6: On-Site Inspection of Appendix D and Appendix E
Monitoring Systems 87
6.1 QA/QC Plan Review 87
6.2 DAHS and Supporting Records 88
6.3 Appendix D Fuel Flow Monitors 89
6.4 Appendix D Fuel Flow Monitor Quality Assurance 90
6.4.1 QA Testing 90
6.4.2 Maintenance and Inspection Records 90
Section 7: Reporting Audit Results 91
7.1 Exit Interview 91
7.2 Audit Report 91
7.3 Follow-up Activities 92
Section 8: Conducting Level 3 Audits 93
8.1 Overview 93
8.2 Tri-Blend or Single Blend Gases 93
8.3 "Hands Off Policy 95
8.4 Test Plan/Procedures 95
8.5 Single Gas Challenge 95
8.6 Linearity Test 96
8.7 Calibration Gases 96
8.8 Training 97
Part 75 Field Audit Manual - My 16, 2003 Page iii
-------
[Thispage intentionally left blank,]
Page iv Part 75 Field Audit Manual - My 16, 2003
-------
List of Tables
Page
Table 1-1: Part 75 Pollutants/Parameters and CEMS Components 4
Table 1-2: Part 75 Non-CEMS Methodologies 5
Table 1-3: Levels of Field Audits 9
Table 1-4: Available EPA Training Courses 12
Table 2-1: Common Extractive Gas CEM Analytical Methods 28
Table 2-2: In-Situ Gas Analyzer Methods 33
Table 3-1: Electronic Monitoring Plan Information 46
Table 3-2: Summary of MDC QA Test Checks 49
Table 3-3: Summary of MDC Hourly Checks 53
Table 4-1: Pre-Audit Interview Items 59
Table 4-2: Part 75 Calibration Error Test Data Validation Requirements 61
Table 4-3: Probe/Sensor Check Summary 63
Table 4-4: Summary of Dilution Air System Checks 65
Table 4-5: Summary of Source Level Extractive System Checks 66
Table 4-6: Summary of Analyzer Checks 68
Table 4-7: Part 75 Calibration Gases (Appendix A, ง 5.1) 70
Table 4-8: Summary of Calibration Gas Checks 71
Table 4-9: Summary of Flow Monitor Checks 73
Table 4-10: Summary of DAHS Checks 76
Table 4-11: Checks for QA/QC Plan Elements 79
Table 5-1: Linearity Test Specifications 83
Table 6-1: Appendix D - QA/QC Plan Review 87
Table 6-2: Appendix E - QA/QC Plan Review 88
Table 6-3: Summary of DAHS and Supporting Records Checks 89
Table 8-1: Effects of Gas Blends on Dilution System Measurements 94
Table 8-2: Elements of a Standard Operating Procedure for Performance Testing 95
Table 8-3: Available EPA Training Courses 97
Part 75 Field Audit Manual - My 16, 2003
Page v
-------
List of Illustrations
Pas
Illustration 1-1: Overview of Continuous Emission Monitoring in a Part 75 Trading
Program 3
Illustration 2-1: Basic CEMS Types (Jahnke and Peeler, 1997) 16
Illustration 2-2: Example of Continuous Emission Monitoring Systems
at a Part 75 Unit 17
Illustration 2-3: Typical Source-Level Extractive CEMS (Gschwandtner
and Winkler, 2001) 21
Illustration 2-4: Dilution Extractive CEMS (Gschwandtner and Winkler, 2001) 23
Illustration 2-5: In-Stack Dilution Probe (adapted from Jahnke, 2000) 24
Illustration 2-6: One Type of Out-of-Stack Dilution System (Gschwandtner
and Winkler, 2001) 26
Illustration 2-7: In-Situ Gas CEMS (Jahnke, 1992) 32
Illustration 2-8: Example of Multiple Probe Locations (Jahnke, 1994) 35
Illustration 2-9: Thermal Mass Flow Monitor Probe (adapted from Jahnke, 1992) 36
Illustration 2-10: Ultrasonic Flow Monitor 37
Illustration 3-1: Example EDR Data Format for Record Type 320 42
Illustration 3-2: Example Summary of Quarterly Report Content For Two Acid Rain
CEMS Units Emitting Through Common Stack 43
Illustration 3-3: MDC Screen Showing Multiple Linearity Tests in One Quarter 48
Illustration 3-4: Example MDC Hourly Graph of SO2 Concentration 52
Illustration 5-1: Asymptotic Calibration Check Response (Jahnke, 1994) 82
Page vi
Part 75 Field Audit Manual - My 16, 2003
-------
Section 1: Introduction
After reading this Introduction, the inspector (auditor) should understand the
organization of the manual and the topics it covers, the role of the field audit in the
Part 75 compliance program, the key components of the field audit, and where to
obtain the latest information on the regulation and on manual updates.
1.1 Background
1.1.1 Importance of Monitoring for Emission Trading Programs
The U.S. Environmental Protection Agency (EPA) has established monitoring
requirements at 40 CFR Part 75 as part of its efforts to develop cap and trade emission
reduction programs. A cap and trade program is an innovative, market-based approach to
reducing emissions. The "cap" sets a ceiling on emissions that is below an applicable baseline
level. Sources in the program receive emission "allowances," with each allowance authorizing a
source to emit one ton of the pollutant being controlled. Limiting the number of available
allowances ensures the cap's integrity. At the end of each year, every source must have enough
allowances to cover its emissions for that year. Unused allowances may be sold, traded, or
saved (banked) for future use. While this approach allows sources flexibility in deciding how
they achieve compliance, the cap ensures that the affected sources reduce emissions collectively
to the desired reduction goal.
The cornerstone for ensuring that sources achieve the required emission reductions is a
strong monitoring program Accurate monitoring of all emissions and timely reporting ensure
that a ton from one source is equal to a ton from any other source and that the integrity of the
cap is maintained. Under Part 75, participating sources must fully account for each ton of
emissions according to stringent, uniform protocols. The resulting compliance information is
unprecedented in its accuracy and comprehensiveness. All data are publicly available on the
Internet, providing complete transparency.
To date, the Part 75 monitoring requirements are used for two separate programs.
Under the Acid Rain Program, sources have had to meet Part 75 and emission reduction
requirements since 1995. EPA has had the lead in ensuring compliance with the Acid Rain
Program, although EPA has teamed with State and local agencies on various aspects of
implementing the Part 75 monitoring provisions.
In May 2002, State agencies began to take the lead role in implementing and ensuring
compliance with Part 75 for purposes of a separate nitrogen oxides (NOX) trading program that
many eastern States have adopted in response to EPA's 1998 NOX SIP Call. EPA believes that
a strong audit program is an essential component of an effective Part 75 compliance oversight
program. Given the increased role of State and local agencies in Part 75 implementation, EPA
has prepared this manual to assist agencies in implementing Part 75 and to ensure the ongoing
integrity of the new NOX trading program.
The manual begins with the premise that each link in the chain of the Part 75 program is
important in ensuring that the data ultimately used to measure emissions and account for the
Part 75 Field Audit Manual - My 16, 2003 Page 1
-------
Introduction Section 1
use of allowances in a trading program remain accurate. Illustration 1-1 depicts the major links
in the data quality chain for a continuous emissions monitoring system (CEMS). The process
starts with ensuring that the gas standards used to calibrate and test the monitoring equipment
are accurate. EPA adopted the Traceability Protocol for Assay and Certification of Gaseous
Calibration Standards for this purpose. The source must conduct the necessary quality
assurance tests following all appropriate procedures and report the results of those tests
accurately. These quality assurance activities are conducted initially for certification and then
on an ongoing basis to maintain a measure of the system's ability to accurately determine
emissions. Once the data measurements are quality-assured, the next step is to ensure that the
measured data obtained from the CEM analyzer are accurately recorded by the data acquisition
and handling system (DAHS) and appropriately reported in the quarterly electronic data reports
(EDR). The EDRs are submitted quarterly to EPA so that it can review and account for the
emissions data in the cap and trade program. EPA provides the necessary data management
systems to track emissions and allowance transfers.
The integrity of the overall trading program can break down anywhere along this chain
of activities, therefore EPA relies on a combination of electronic and field auditing to verify
overall data integrity. The field audit procedures in this manual are critical for examining these
links to verify proper performance of the monitoring systems and identify problems which may
lead to inaccurate emissions accounting.
Page 2 Part 75 Field Audit Manual - My 16, 2003
-------
Section 1
Introduction
Illustration 1-1:
Overview of Continuous Emission Monitoring in a Part 75 Trading Program
Electronic Data Reports
Emissions Tracking
System
Allowance Tracking
Systems
1.1.2 Structure of Part 75 Monitoring Provisions
Continuous emissions monitoring systems (CEMS) are the primary monitoring method
under Part 75. The Part 75 rule includes requirements for installing, certifying, operating, and
maintaining CEMS for SO2, NOX, CO2, O2, opacity, and volumetric flow. Appendices A and B
of Part 75 provide the technical specifications for the installation and performance of CEMS,
including certification and quality assurance test procedures. The rule also includes approved
non-CEMS options for certain gas and oil fired units, and procedures to account for missing
data.
Recordkeeping and reporting provisions require Acid Rain Program and NOX trading
program affected units to submit Part 75 hourly emission data and related quality assurance
data through electronic report formats to EPA's Emission Tracking System (ETS) which is
operated by the Clean Air Markets Division (CAMD). The ETS data in turn are used to
Part 75 Field Audit Manual - My 16, 2003
Page 3
-------
Introduction
Section 1
maintain the emission allowance accounts in the Allowance Tracking System and the NOX
Allowance Tracking System.
TIP!
Check www.epa.gov/airmarkets
for further regulatory information
The Part 75 requirements are
outlined below to introduce you to the
rule and some of the terminology used in
the manual. You can obtain copies of
Part 75 and determine whether EPA has
published further revisions to Part 75,
issued new monitoring guidance, or
revised the information in this manual by checking CAMD's website. Section 1.5 of the manual
provides a list of important regulations and policy guidance documents, with links to specific
pages on CAMD's website you may find helpful.
Monitoring Methods
The monitoring requirements for each type of unit subject to Part 75 are in Subpart B of
the rule. CEMS are required except for some gas and oil fired units. Table 1-1 summarizes
CEMS components that are required by pollutant, while Table 1-2 summarizes the non-CEMS
options.
Table 1-1:
Part 75 Pollutants/Parameters and CEMS Components
Pollutant/Parameter
S02 (Ib/hr)
NOX (Ib/mmBtu)1
NOX (Ib/hr)2
Opacity (%)3
CO2 (Ib/hr)4
Heat Input (mmBtu/hr)
Required CEMS Components
SO2
NOX
Flow
Opacity
Diluent
Gas (O2
or CO2)
Data
Acquisition
and Handling
System
(DAHS)
'Heat input in mmBtu/hr is also required.
2For units subject to NOX SIP Call trading program. Can monitor with or without diluent
monitor.
'Required only for coal and residual oil units.
"Alternative mass balance method may by used to monitor CO2.
Page 4
Part 75 Field Audit Manual - My 16, 2003
-------
Section 1
Introduction
Table 1-2:
Part 75 Non-CEMS Methodologies
Pollutant/
Parameter
S02 (Ib/hr)
SO2 (Ib/hr)
NOX (Ib/mmBtu),
NOX (Ib/hr)
SO2, CO2, NOX
(Ib/hr for all, and
Ib/mmBtu for NOX)
Heat Input
(mmBtu/hr)
Unit Type
natural gas
gas or oil
gas or oil
peaking units
gas or oil
gas or oil
Monitoring Methodology
Default SO2 emission rate combined with measured fuel
flow. (Part 75, Appendix D)
Fuel sampling and analysis combined with measured
fuel flow. (Part 75, Appendix D)
Estimate NOX rate by using site-specific emission
correlations with measured fuel flow if measuring Ib/hr.
(Part 75, Appendix E)
Conservative default values for units with low mass
emissions. (ง 75.19)
Measured fuel flow and GCV. (Part 75, Appendix D)
Monitoring Certification Requirements
The implementing agency must certify an allowable monitoring method before it can be
used for Part 75 monitoring. The source must perform certification tests and submit the results
to EPA and the appropriate State agency. Part 75 performance certification testing is outlined
in ง 75.20 and Appendix A, ง 6. Certification tests for a CEMS may include:
7-day calibration error test for each monitor
Linearity check for each pollutant concentration monitor
Relative Accuracy Test Audit (RATA) for each monitoring system
Bias test for each monitoring system
Cycle time test for each pollutant concentration monitor
Daily interference test for flow monitors
DAHS testing
There are also certification requirements for non-CEMS methods. These include
accuracy tests for fuel flow monitors (ง 75.20 and Appendix D, ง 2.1.5), stack tests to develop
NOX correlations for gas or oil peaking units (Appendix E, ง 2.1), or unit-specific default values
for low mass emissions units (งง 75.19-75.20).
Recertification may be required if the facility replaces, modifies, or changes a certified
CEMS in a way that may significantly affect the ability of the system to accurately measure
monitored parameters.
Quality Assurance/Quality Control Procedures
The source is required to develop and implement a written quality assurance/quality
control (QA/QC) plan for each monitoring system (ง 75.21). The QA/QC plan must include
procedures for system operation, as well as procedures for conducting quality assurance tests
(QA tests), preventive maintenance, and recordkeeping. Appendices A and B to Part 75
Part 75 Field Audit Manual - My 16, 2003
Page 5
-------
Introduction Section 1
describe the technical procedures for how and when to conduct periodic QA tests, which
include:
Daily calibration error tests: Challenge a gas CEMS at a zero and high level with
calibration gas.
Daily interference tests for flow monitors: Follow procedure to detect plugging or
other problems that could interfere with a flow monitor.
Quarterly linearity tests: Challenge a gas CEMS at 3 levels with calibration gases.
Quarterly flow-to-load evaluations: Compare flow monitor values to values from an
initial flow-to-load correlation as a means to check flow monitor data quality over time.
Semi-annual or annual RAT As: Compare monitored values to values measured by an
approved EPA reference method. Also, use RATA results to detect and, if necessary,
adjust for low bias.
Recordkeeping and Reporting
Part 75 includes requirements for notifications, recordkeeping, and reporting in
Subparts F and G. As noted earlier, most of the reporting to EPA is done electronically every
quarter in a standard electronic format, and much of the recording will be done automatically
using the DAHS. Some important records and reporting that you will want to review include:
Monitoring plan: Submitted electronically, although some information is submitted only
in hardcopy. Contains information describing the unit, CEMS, other monitoring
methodologies, and specific calculation procedures.
Hourly parameters, including emissions, flow, heat input, monitor availability, and other
information.
Periodic QA test results.
Recertification tests.
Other records that are required to be kept on-site such as:
Annual span/range evaluation.
SO2 scrubber parameters to verify proper control operation during a missing data
period.
Missing Data
Part 75 requires sources to account for emissions during periods when there are no valid
data (missing data periods) due to the monitor not operating or operating out of control. The
missing data methodologies are necessary so that a source accounts for emissions during each
hour of operation. The missing data algorithms become increasingly conservative as monitor
downtime increases so that sources have an incentive to maintain high data availability.
1.2 What does the manual cover?
This manual details recommended procedures for conducting a field audit of a Part 75
monitoring system. Included are: tools you can use to prepare for an audit; techniques you can
use to conduct the on-site inspection and review records; proper methods for observing
performance tests; and guidelines for preparing a final report. EPA has designed the audit
procedures in this document so that personnel with varying levels of experience can use them
Page 6 Part 75 Field Audit Manual - My 16, 2003
-------
Section 1 Introduction
While the manual is written primarily for State and local agency inspectors, industry personnel
may find some of the material useful for their internal data quality management activities.
The manual covers gas (SO2, NOX, and diluent) and flow monitoring systems it does
not cover opacity monitor audits. Although Part 75 requires opacity monitors for coal-fired
units subject to the Acid Rain Program, opacity data and quality assurance tests are not
reported to CAMD in quarterly emission data reports. Moreover, the source can comply with
Part 75 by satisfying performance specifications in Part 60 that are generally applicable to
opacity monitors and can follow a State's recording and reporting requirements. Thus, there
are no special Part 75 audit techniques for these systems.
The manual is organized into eight major sections, with one appendix:
Section 1 introduces cap and trade programs, Part 75, the role of field audits and the
inspector, CAMD's audit targeting role, the importance of inspector training, and a list
of key Part 75 materials with Internet links.
Section 2 provides a short introduction to the various types of CEMS and the major
components of a CEMS, including basic installation and operating principles.
Section 3 describes preparing for an audit prior to the plant visit, with emphasis on
using CAMD's Monitoring Data Checking (MDC) software to review the electronic
data.
Section 4 covers the on-site CEMS inspection, including what to look for and questions
to ask during a walk through of CEMS components, as well as how to review the
QA/QC plan and other in-plant records.
Section 5 describes how to observe CEMS performance tests (linearity and relative
accuracy test audits).
Section 6 outlines specific on-site review procedures for Appendix D and E monitoring
systems and records.
Section 7 guides you in conducting the exit interview and preparing a written audit
report.
Section 8 discusses issues that should be considered by a State or local agency in
developing a performance testing program, with an emphasis on single gas challenges
and linearity tests.
Appendix A to the Manual provides sample checklists for the field audit, RAT A, and
linearity observations. The checklists are based on the discussions and techniques in
Sections 3 through 6.
Part 75 Field Audit Manual - My 16, 2003 Page 7
-------
Introduction Section 1
1.3 Part 75 Audit Program Overview
The Part 75 audit program consists of both electronic audits and field audits. CAMD
uses automated tools such as the Monitoring Data Checking (MDC) system to conduct
automated checks of data submitted under Part 75 for potential problems. Also, it uses its data
systems and its ability to check data through automated information systems to target units for
follow-up data audits. On-site field audits performed to ensure that monitoring systems are
installed and operated properly are also essential in the Part 75 audit program.
1.3.1 Part 75 Electronic Audit Program
CAMD performs routine electronic audits on each quarterly report submittal using the
ETS and MDC software. EPA may also perform targeted electronic checks to find other
specific data reporting problems. The electronic audits identify errors in the quarterly electronic
data report, the monitoring plan, and the QA tests. An automated ETS feedback report that
focuses on the reported emissions data is sent to the source instantly upon electronic submittal
by ETS. EPA then uses MDC to analyze the monitoring plan and QA data, and sends an MDC
feedback report at the end of the quarterly submission period. The reports categorize errors as
critical and non-critical for critical errors, the source must correct and resubmit the quarterly
report.
1.3.2 Audit Targeting
In addition to performing electronic audits, EPA periodically compiles a recommended
field audit target list based on a review of all of the quarterly electronic data reports. This
national list attempts to identify trends based on a large population of units that may not be
identifiable from a smaller population at the State level alone. The target list is intended to help
States allocate their auditing resources on those units that are most likely to have data problems
based on the findings of EPA's electronic auditing efforts. States may use these
recommendations to focus their audit efforts, but may also choose other units for field audits
through State targeting approaches or at random.
1.3.3 Field Audits
EPA relies on State and local agencies to conduct field audits of monitoring systems to
assess the systems performance and a source's compliance with monitoring requirements. The
audits also encourage good monitoring practices by raising plant awareness of monitoring
requirements. The field audit consists of a thorough evaluation of the CEMS via pre-audit
record review, on-site inspection of equipment and system peripherals, record reviews, test
observations, and interviews with the appropriate plant personnel.
EPA has defined three levels (see Table 1-3) to describe field audit activities and
procedures and the objective of the audit.
Page 8 Part 75 Field Audit Manual - My 16, 2003
-------
Section 1
Introduction
Table 1-3:
Levels of Field Audits
Audit
Level
Level 1
Level 2
Level 3
Records
Review
On-site
Inspection
ofCEMS
Daily
Calibration
Observation
Linearity or
RATA
Observation
Performance
Test Audit
The Level 1 audit may be used to verily Part 75 recordkeeping requirements, emissions
data and monitoring plan information, and is recommended as a follow-up to a previous audit.
This audit consists of an on-site inspection, records review and daily calibration observation. A
Level 2 audit expands the audit to include a performance test observation. The test observation
is a critical element to ensure that CEMS are properly operating and performance test protocols
are being conducted in accordance with the required standards. For this reason, EPA
encourages agencies to perform Level 2 audits, which are the focus of this manual.
A Level 3 audit involves agency personnel conducting a performance test instead of
merely observing the test. Conducting a performance test such as a linearity test or relative
accuracy test audit provides an independent assessment of the monitoring systems. Because of
the equipment and expertise involved, EPA does not emphasize that State or local agencies
perform Level 3 audits. However, some agencies strongly support Level 3 audit programs, and
Section 8 of this document provides guidance on various Part 75-related issues for those
agencies that do conduct performance tests as part of their inspection program.
1.4 Role of the Inspector
Your primary task as an inspector conducting a field audit is to document whether the
monitoring at a facility is in compliance with the Part 75 requirements. To carry out this task,
you will need to understand the Part 75 rule and have a general understanding of CEMS
components and their function. You will also need to ask questions, carefully record your
observations and compile the information necessary to determine compliance.
Your role is not to provide technical advice or consulting on the operation of the
monitoring equipment. The source is responsible for operating the monitoring equipment, and
correcting any monitoring problems. At the same time, however, the field audit is an
opportunity to provide information to the source on Part 75 requirements. You might, for
example, clarify regulatory requirements, and you should share with the source your
observation of monitoring practices that may create regulatory problems.
Part 75 Field Audit Manual - My 16, 2003
Page 9
-------
Introduction
Section 1
Importance of Missing Data
Under Part 75
Because Part 75 monitoring is used to
account for total mass emissions,
when the monitor or monitoring
system fails to record a valid hour of
data, Part 75 uses a conservative
approach to substitute for missing
data. Audit findings may invalidate
data and require use of substitute data,
so the findings could have a significant
financial impact, independent of any
non-compliance issues.
If your findings indicate that data
from a monitor may be invalid, which would
require the source to use substitute data, it is
important to inform the source of the
problem during the field audit. Extensive
missing data generally will penalize a source
in allowance accounting and result in a
significant monetary penalty for the source.
For the same reason, it is important to notify
CAMD quickly of the potential for
invalidating data, so that the issue is resolved
prior to the end-of-year compliance process.
EPA's primary concern is to collect accurate
CEM data to ensure the integrity and fairness
of the trading program. EPA has no interest
in prolonging the length of time that a source
is considered out-of-control. Thus, the goal
of the audit should be to promptly identify
what needs to be corrected so that the source is once again reporting quality-assured, verified
emissions data. This issue is discussed further in Section 7.2.
1.4.1 "Hands Off Approach
EPA's policy is to use a "hands off approach when conducting the audit so that you do
not have any physical contact with the CEMS hardware. This approach avoids creating a
situation in which monitoring equipment may be damaged or the inspector's actions may be
questioned should the monitoring system fail to operate well during the audit. You should ask
authorized plant personnel to perform any actions with the CEMS equipment (for example,
initiating a daily calibration check or displaying analyzer range). Remember, it is more
important for you as the inspector to observe how the CEMS operator performs his/her duties,
as this will indicate whether the plant personnel are able to follow appropriate requirements and
procedures, and will help to identify any problems that occur. Have the operator explain what
he/she is doing and show you where the procedure is documented.
Page 10
Part 75 Field Audit Manual - My 16, 2003
-------
Section 1 Introduction
1.4.2 Inspection Safety
Any type of air pollution source inspection has potential health and safety problems, and
inspection safety is a serious concern. Appropriate safety training is imperative for all
inspectors, and each plant may have specialized training and/or safety equipment policies.
Before going on site for an audit, you must ensure that you have all necessary personal safety
equipment. Also, make sure to contact the site and ask for details on plant safety requirements.
Once on site, use the safety equipment properly, adhere to your agency's safety requirements,
and follow plant safety requirements as well. Some of the hazards you may encounter in
performing CEMS audits include:
Accessing CEMS equipment or platforms and working at elevations with fall hazards
Electrical shock when inspecting heated lines, pumps, or internal areas of CEMS
cabinets and enclosures
Hazards associated with use or transport of compressed gas cylinders
Hazards associated with poisonous calibration gases (NO)
Exposure to effluent gases
Entry of confined spaces
Typical hazards associated with working in an industrial environment (moving
equipment, vehicles, and machinery, trip and fall hazards, etc.)
1.4.3 Recommended Training
Air PoDution Training Institute
Courses Information on EPA courses and course
schedules are available at EPA's AIR
The following table lists EPA
Pollution Training Institute website:
http://www.epa.gov/air/oaqps/eog/apti.html
classes that you may find helpful in
developing a knowledge base for
performing Part 75 CEMS field audits at
stationary sources. State agencies,
regional organizations, or university professional development programs may provide similar
courses in your area.
Part 75 Field Audit Manual - My 16, 2003 Page 11
-------
Introduction
Section 1
Table 1-4:
Available EPA Training Courses
EPA Course
Number
APTI 445
APTI 446
APTI 450
APTI 474
SI 476B
T008-02
T468-02
Course Title
Baseline Inspection Techniques
Inspection Safety Procedures
Source Sampling for Pollutants
Continuous Emission Monitoring
Continuous Emission Monitoring Systems Operation
Monitors
and Maintenance of Gas
Safety and the Agency Inspector
Stack Testing/Stack Test Observation for Traditional
Pollutants
and Hazardous Air
1.5 Key Part 75 Materials with Internet Links
The following is a list of key Part 75 reference materials with internet links to the
webpage where the document can be found either on the EPA website or the Government
Printing Office website. To avoid a dead link, the links in most cases are not to the document
itself, but to the web page where a link to the document may be found. You will need to survey
the page to find the direct link.
40 CFR Part 75 - On the CAMD
website you will find an unofficial
consolidated version of the Part 72 and
Part 75 rules that contains the current
text of Part 75 (and งง 72.1 - 72.3, the
Acid Rain Program rule general
provisions: purpose, definitions,
measurements, abbreviations, and
acronyms) as amended by recent
revisions. You may find this unofficial version a helpful tool because it has an easy-to-
use format.
http://www.epa.gov/airmarkets/monitoring/consolidated/index.html
Recent Part 75 Revisions
June 12, 2002 (67 FR 40394),
and
August 16, 2002 (67 FR 53503)
While all reasonable steps have been taken to make this unofficial version accurate, the
Code of Federal Regulations (CFR) and Federal Register (FR) take precedence if there
are any discrepancies. Official versions of FR revisions are available on the EPA
Page 12
Part 75 Field Audit Manual - My 16, 2003
-------
Section 1 Introduction
website, http://www.epa.gov/fedrgstr/, and the official CFR is available at the
Government Printing Office website.
http://www.access.gpo.gov/nara/cfr/cfrhtml 00/Title 40/40cfr75 OO.html
Electronic Data Report Version 2.2 Reporting Instructions - The instruction manual
describes each data field element for the information that is recorded and reported to
EPA for Part 75, and provides the field auditor with a helpful summary of Part 75
requirements. The EDR v2.2 Instructions support the June 12, 2002 revised Part 75
rule.
http://www.epa.gov/airmarkets/monitoring/
Parts 75 and 76 Policy Manual - This manual contains a series of questions and
answers that can be used on a nationwide basis to ensure that the Part 75 provisions are
applied consistently for all sources affected by the rule. It is intended to be a living
document. EPA will issue new questions and answers as they arise and will revise
previously issued questions and answers as necessary to provide clarification. EPA
intends to release a revised verison of the manual in 2003.
http://www. epa. go v/airmarkets/monitoring/polman/index.html
Recertification and Diagnostic Testing Policy - Recertification is required whenever a
replacement, modification, or change in the certified continuous emissions monitoring
system or continuous opacity monitoring system occurs that may significantly affect the
system's ability to accurately measure or record the pollutant or parameter of interest.
EPA is preparing a document to clarify what types of changes to a monitoring system
may be considered significant. EPA expects that the document will describe various
events as either recertification events or diagnostic testing events, and describe the type
of certification or diagnostic testing that needs to be performed. You should check the
www.epa.gov/airmarkets website for release of this policy.
Monitoring Data Checking (MDC) Software - The MDC software, discussed in
Section 3 of this manual, allows regulated industry and State agencies to enter, analyze,
print, and export electronic monitoring plan, certification, and quality assurance data,
and to evaluate hourly emissions data required by Part 75. The software also allows
regulated sources to submit electronically monitoring plan and certification data to EPA
via ftp. The software provides a standard Windows-based, mouse-driven, point and
click user interface, and can be installed under Windows 95 (or higher), Windows NT,
or Windows 2000. The software and installation instructions can be downloaded from
the CAMD website.
http://www.epa.gov/airmarkets/monitoring/mdc/
40 CFR Part 60, Appendix A Reference Methods - The RATA reference methods
(and related information) are available from EPA's Emission Measurement Center
website. The website versions of the reference methods are the official CFR version.
http://www. epa. go v/ttn/emc/promgate.html
Part 75 Field Audit Manual - My 16, 2003 Page 13
-------
Introduction Section 1
An Operator's Guide to Eliminating Bias in CEM Systems - This EPA publication
is designed for CEMS operators as a tool for diagnosing and correcting the causes of
measurement bias in CEMS. It is also a useful CEMS reference guide for the field
auditor.
http://www.epa.gov/airmarkets/monitoring/
Observer's Checklists for Test Methods 2F, 2G, and 2H - These are detailed
observer checklists that can be used when observing a flow RATA using one of these
alternative flow reference methods.
http://www.epa.gov/airmarkets/monitoring/
Page 14 Part 75 Field Audit Manual - My 16, 2003
-------
Section 2: Part 75 CEMS Overview
This section provides a brief introduction to the various types of components of
continuous emission monitoring systems that facilities have installed to meet Part 75
requirements.
2.1 Introduction
The monitoring requirements of Part 75 are performance-based requirements that
generally do not require that a source use a particular type of CEMS. There are several types
of CEMS available. The differences in how these systems are designed and operate (in terms of
sample acquisition, sample handling, sample analysis, and data recording) can be important in
understanding what to look for in a field audit and how to interpret audit results.
This section provides only an overview of the major concepts related to the types of
CEMS and their principles of operation. For further detail on these complex systems, see
Section 2.7, which provides a list of in-depth references. EPA recommends that inspectors who
will conduct Part 75 CEMS audits should attend EPA's Air Pollution Training Institute course
on CEMS (APTI Course 474). The summary information in this section draws heavily from the
manual for that course (Jahnke, 1992), as well as from an EPA reference manual specifically
tailored to Part 75 CEMS (Gschwandtner and Winkler, 2001).
All CEMS perform the following basic functions:
Locate or extract a representative sample;
Analyze the sample; and
Compile and record the results.
CEMS are divided into two types based on the first of these basic functions. An
extractive system removes and transports the sample from the stack to the analyzer, often
conditioning the sample prior to the analyzer. An in-situ system analyzes the sample directly in
the stack. Illustration 2-1 identifies these two main system types. There are several variations
on these types, which Sections 2.3 through 2.5 review.
Part 75 Field Audit Manual - My 16, 2003 Page 15
-------
Part 75 CEMS Overview
Section 2
Illustration 2-1:
Basic CEMS Types (Jahnke and Peeler, 1997)
EXTRACTIVE
(see Section
2.3)
IN-SITU
(see Sections
2.4 and 2.5)
Single Pass
Double Pass
Illustration 2-2 shows a set of typical Part 75 CEMS at a Part 75 unit. The flow and
opacity CEMS are examples of in-situ systems. The opacity monitor measurement is taken
over a path, across the stack. Most continuous opacity monitors are double pass (light is
transmitted across the stack and back to the detector) to meet EPA quality assurance
requirements. The ultrasonic flow monitor in this example also provide an integrated
measurement along a path across the stack. Flow and opacity CEMS are always in-situ
systems. The gas CEMS (SO2, NOX, CO2) in Illustration 2-2 are dilution extractive systems. In
this illustration, the gas is extracted at a single point (the sampling probe) and diluted with clean
dry air. The diluted sample is transported through the sampling line and analyzed on a wet
basis.
Page 16
Part 75 Field Audit Manual - My 16, 2003
-------
Section 2
Part 75 CEMS Overview
Illustration 2-2:
Example of Continuous Emission Monitoring Systems at a Part 75 Unit
Sampling
Probe cz
I Opacity
Monitor
Flow
Monitor
Dilution Air
System
Stack
Calibration Gases
Analyzers
DAHS
In-situ gas CEMS (not shown in Illustration 2-2) are also used by Part 75 sources,
sometimes in conjunction with extractive gas CEMS. The use of in-situ gas CEMS for Part 75
compliance is far less common than the use of extractive systems in 2002 only about three
percent of the gas monitors used to meet Part 75 requirements were in-situ monitors. In-situ
gas CEMS can measure at a point (or short path) like an extractive gas CEMS or along a path
across the stack similar to an opacity monitor.
The following sections begin with a discussion of Part 75 requirements for the sample
measurement location, and then briefly describe the three basic types of systems under Part 75
(gas extractive CEMS, gas in-situ CEMS, and Flow CEMS). The final CEMS component
the data acquisition and handling system (DAHS) used for electronic data and reporting is a
critical element for Part 75 compliance and is discussed separately in Section 2.6. Finally,
Section 2.7 provides references that you can use to gain an in-depth understanding of how these
systems operate and what their limitations are.
2.2 Sampling Location
Whether the system is extractive or in-situ, the flue gas must be monitored at a location
where the pollutant concentration and emission rate measurements are directly representative of
the total emissions from the affected unit. Flowing gases are generally well mixed, but in some
cases gas stratification can be a concern for the gas measurement location. Stack flow, on the
other hand, is always stratified to some degree (lower velocities along the stack walls).
Cyclonic or swirling flow (flow that is not parallel to the stack center line) also will have a
negative impact on flow monitors and manual reference test methods. Thus, the selection of
sampling points and paths is an important concern for flow monitors. To obtain a
Part 75 Field Audit Manual - My 16, 2003
Page 17
-------
Part 75 CEMS Overview Section 2
representative measurement, and avoid problems due to stratification and cyclonic flow, Part 75
provides specific requirements for the CEMS sampling location in Appendix A, ง 1.
2.2.1 Gas Measurement Location
Part 75 requires that the representative sampling location be chosen based on the
procedures in 40 CFR Part 60, Appendix B, Performance Specification 2, which suggests a
measurement location: (1) at least two equivalent diameters downstream from the nearest
control device, the point of pollutant generation, or at another point at which a change in
pollutant concentration or rate may occur, and (2) at least a half equivalent diameter upstream
from the effluent exhaust or control device. Other Part 75 location requirements from
Appendix A are summarized below:
Locate the measurement location so that the gas CEMS (pollutant and diluent monitor)
passes the certification RAT A. (Note - while not required specifically, the diluent O2 or
CO2 monitor should sample at the same point as the pollutant monitor.)
Point Monitors - Locate the measurement point (1) within the centroidal area of the
stack or duct cross section, or (2) no less than 1.0 meter from the stack or duct wall.
Path Monitors - Locate the measurement path (1) totally within the inner area bounded
by a line 1.0 meter from the stack or duct wall, or (2) such that at least 70.0 percent of
the path is within the inner 50.0 percent of the stack or duct cross-sectional area, or (3)
such that the path is centrally located within any part of the centroidal area.
2.2.2 Flow Measurement Location
Part 75 establishes the following basic location criteria for flow monitors:
The location satisfies the minimum siting criteria of Part 60, Appendix A, Method 1
(i.e., the location is greater than or equal to eight stack or duct diameters downstream
and two diameters upstream from a flow disturbance, or, if necessary, two stack or duct
diameters downstream and one-half stack or duct diameter upstream from a flow
disturbance); or
The results of a flow profile study, if performed, are acceptable (i.e., there are no
cyclonic (or swirling) or unacceptable stratified flow conditions). Part 75 recommends
that if a flow profile study indicates unacceptable results, the facility should relocate the
monitor or add straightening vanes or other source modifications to correct the flow
patterns.
Regardless of whether these criteria are met, a flow monitor can be installed in any location if
the flow CEMS can meet the Part 75 performance specification requirements.
2.2.3 Sampling in Stratified and Swirling Flow Conditions
Stack flow is seldom ideal, and some degree of stratification and swirling flow will be
present at the monitoring location. Approaches to dealing with stratification, swirling or
cyclonic flow, and changing flow profiles due to load changes are discussed below.
Page 18 Part 75 Field Audit Manual - My 16, 2003
-------
Section 2 Part 75 CEMS Overview
Stratified Flow - Flow monitoring systems may locate a single point or path
representative of the reference method determined stack flow if the stratification is fairly
constant over varying loads. If stratification varies with load, an array of sampling
points can be placed across the stack to obtain a flow average instead of one single
sample point. For a path monitoring system that is already averaging over a line across
the stack, the source can select a path that is not as sensitive to the variation or can add
a monitor to provide multiple paths on the cross section.
Correction Factors - Part 75 allows the source to calibrate the flow monitor to the
reference method flow (pre-RATA). Sources commonly use this approach to enable a
flow CEMS to pass the multi-load flow RATA at a particular measurement location. A
flow RATA typically is performed at three loads to account for different flow profiles at
changing loads. The options described above for stratified flow can include application
of a correction factor for stratification based on the reference method RATA values. If
the source conducts a test using new Methods 2F, 2G, or 2H (developed to account for
non-parallel flow conditions and wall effects on flow measurement), calibrating to the
reference method also can account for effects due to swirling. Method 2 using the s-
type pitot tube will be subject to bias if swirling or stratification due to wall effects or
other factors are present. Calibration of flow monitors relative to Method 2 under such
conditions will not account for those effects.
2.3 Extractive Gas CEMS
There are two types of extractive gas CEMS:
The "source level" or "direct extractive" system extracts gas from the stack, and
conveys the sample to one or more analyzers. These extractive systems will include
filters to remove particulates and may include conditioning to remove moisture, acids,
condensible particulates, and interfering compounds. In the case of a hot wet system,
the sample lines and analytic components of the systems are heated to prevent
condensation of the stack moisture. Heated lines are also required for dry systems up to
the point where conditioning occurs.
A dilution extractive system filters the stack sample and dilutes the stack sample with
clean dry air. Dilution occurs either inside the sample probe in the stack or outside of
the stack, usually at the stack port. Dilution systems sample the gas at flow rates much
smaller than those used in source level systems. Using dry air to dilute the flue gas at
ratios of 50:1 to 300:1, the dew point of the diluted sample is reduced to levels where
condensation will not occur in the monitoring system. As a result, moisture removal
systems and heated sample lines may not be incorporated into the system. A dilution air
cleaning system, however, is required to clean the dilution air and remove CO2 water,
and any traces of the gases that are to be monitored.
2.3.1 Source Level or Direct Extractive Systems
A diagram of a source level extractive system is shown in Illustration 2-3. The
illustration shows both wet (heated sampling line by-passing the conditioning system) and dry
systems.
Part 75 Field Audit Manual - My 16, 2003 Page 19
-------
Part 75 CEMS Overview Section 2
2.3.1.1 Sample Probe
Probes for source level extractive systems are constructed with stainless steel or
Hastelloyฎ tubes. A coarse filter is commonly attached at the end of the probe to filter out
particulate matter before it can enter into the tube. Some designs place a deflector plate or
cylindrical sheath around the filter to provide protection from plugging. A coarse filter can also
be in-line in a housing outside of the stack prior to the sample line. Sometimes a combination
of filters, a coarse filter at the probe opening, and an in-line fine filter outside of the stack, are
used to ensure the removal of particulate matter.
Blowback or back purging is frequently used to keep the coarse filter from plugging.
This involves blowing pressurized air or steam back through the filter in an opposite direction
to the normal stack flow. The blowback occurs at regular intervals (typically from 15 minutes
to 8 hours) and typically lasts for 5 to 10 seconds.
Page 20 Part 75 Field Audit Manual - My 16, 2003
-------
Section 2
Part 75 CEMS Overview
Illustration 2-3:
Typical Source Level Extractive CEMS
(Gschwandtner and Winkler, 2001)
Coarse
Filter
Zero NO SO2 COj'Ol
(Diluent)
Heated
Pump
OD&
Unheated
Condensers Pump
Color Key
Red - Wet Flue Gas
Blue - Dry Flue Gas
Unheated
Sample
Line
Moisture
Sensor
1
Heated
Sample Line
Vent
1
Display/
Records
DAHS
Part 75 Field Audit Manual - My 16, 2003
Page 21
-------
Part 75 CEMS Overview Section 2
2.3.1.2 Sample Transport and Conditioning Systems
Most source level extractive systems used in Part 75 applications are dry systems which
remove moisture prior to the sample pump and analyzer. In a dry system the sample line from
the probe to the moisture removal system is heated to prevent water condensation. If the
facility uses a wet system that does not remove moisture prior to the analyzer, the entire length
of the sample line, sample pump, and analyzer must be heated. The sample line is usually
wrapped in a tube bundle or umbilical which includes the sample lines, blowback lines,
calibration gas lines, heating elements, and electric lines.
Dry source level CEMS used in conjunction with a flow CEMS for Part 75 SO2, NOX
and CO2 mass measurements must also determine the moisture content of the stack gas.
Illustration 2-3 shows a heated "wet" sample line connected to a wet O2 analyzer. This wet
system is used in conjunction with the dry system's dry O2 analyzer to determine moisture.
Another alternative is to use an in-situ "wet" O2 analyzer with the dry extractive O2 analyzer. A
less common approach is to use an H2O analyzer to measure the wet sample.
Moisture Removal Systems
There are two common types of moisture removal systems: condensers and permeation
dryers. Condensers cool the gas below the dew point (using refrigerated coils or cooled jet
stream condensers), and then remove the condensed liquid water from the gas stream. Water
removal is performed automatically to prevent filling the condensate trap and flooding the
sampling line. Absorption of SO2 and NO2 in the condensate is a concern, so systems are
designed to minimize contact time between the dried gas and liquid.
Permeation dryers are constructed using Nafionฎ, a material that selectively allows the
mass transfer of water vapor from the sample gas through the tube membrane to dry purge gas
flowing in an outer tube in the opposite direction. The gas entering the permeation dryer must
be heated above the dew point temperature. Permeation dryers avoid the problems of
condensate absorption of pollutants and do not have condensate traps. However, the dryers
can be subject to plugging problems from condensing liquids, particulates, or precipitates.
Pumps
Diaphragm pumps and ejector pumps are the most common pumps used in extractive
systems. Both operate without pump lubricating oils, which can cause sample contamination.
Both diaphragm pumps and ejector pumps can be heated and used in a hot wet system, or used
ahead of a conditioning system.
Fine filter
Many analyzers require removal of particulate larger than 0.5 |im, so systems will
usually have an additional fine filter near the analyzer inlet. There are two types: (1) a surface
filter, usually paper or other porous material which builds up a filter cake, and (2) a depth filter,
which consists of packed fibers of quartz wool or other material.
Page 22 Part 75 Field Audit Manual - My 16, 2003
-------
Section 2
Part 75 CEMS Overview
2.3.2 Dilution Extractive Systems
Most coal-fired units subject to Part 75 use
dilution extractive systems. As noted earlier dilution
ratios range from 50:1 to 300:1. A diagram of a dilution
extractive CEMS is provided below showing an in-stack
dilution probe, unheated sample lines and pumps, air
cleaning subsystem, calibration gases, analyzers, and
DAHS.
Dilution Ratio = Ql + Q2
Q2
where:
Q! = dilution air flow rate (L/min)
Q2 = sample flow rate (L/min)
Illustration 2-4:
Dilution Extractive CEMS
(Gschwandtner and Winkler, 2001)
Air Cleaning Subsystem
Air Intake
Vacuum
Probe
1 EE
"VYJ
F
*
1
t
1
;L
vacuum
Gauge
ซ kJTd
u1 piq
3
* -
;r ซ |
Charcoal MM
Filter MM
Ktซ_fXซ_r"~|_r i
P*Q PN 1 M
I II |l 1 1 1, | CO/COi I.
^~p~| CO 2 |~| Camtfta p
1 Back Purge Desiccant
Valve Dryer
* ^L, n 5
Purge^
Valve
i ;
i \
t \
< |l
3 1 1
Rotameter
Exhaust
Vent
fi
Prefilter Air Pump or
Plant Air
Filter
Supply
Manifold
Color Key
Blue- Stack Gas Sample
Green - Diluted Sample
C02
Calibration Gas Subsystem
Analyzer Rack
Exhaust
Manifold
Display/
Records
DAHS
Part 75 Field Audit Manual - My 16, 2003
Page 23
-------
Part 75 CEMS Overview
Section 2
The dilution of the sample can occur in the stack using a dilution probe, or outside the
stack using an out-of-stack dilution system. Both approaches use the same operating
principles. A critical orifice controls the sample gas flow rate, which is drawn through the
orifice by creating a vacuum at the outlet of the orifice with an ejector pump. As long as a
sufficient vacuum is present, the sample gas flow rate through the critical orifice is independent
of the downstream pressure. The ejector pump, also called an aspirator or an eductor pump, is
operated by the compressed, dry, clean dilution air. The dilution air flow through the venturi
nozzle of the pump (flow rates of 1 to 10 L/min) creates the vacuum pressure at the orifice
outlet. This vacuum pulls the gas sample through the orifice at rates of 50 to 500 mL/min
(limited by the orifice design), and into the ejector pump where it mixes with the dilution air.
2.3.2.1 Dilution Probe
The in-stack dilution probe combines a sonic orifice (a glass tube drawn to a point) with
an ejector pump inside the stack probe. (See Illustration 2-5.) The probe opening inside the
stack is screened and uses a quartz wool filter to prevent particulate matter from entering the
orifice. Plugging has been a problem in applications with wet saturated conditions (after a wet
scrubber) due to condensation that causes wetting of the filter and plugging of the orifice. In
some cases, heated dilution probes have been successfully used where entrained water droplets
are present.
Illustration 2-5:
In-Stack Dilution Probe (adapted from Jahnke, 2000)
Ejector Pump
Draws Flow Through
Sonic Orifice
Sample Gas Lin
Sonic Orifice
Controls Sample
Flow Rate
Diluted Sample Out
Dilution Air In
Calibration Gas Line
Screen Catches
Large Dust
Particles
Mounting Flange
Thermocouple
Calibration Gas Line
Quartz Wool
-Filters Small
Dust Particles
Page 24
Part 75 Field Audit Manual - My 16, 2003
-------
Section 2 Part 75 CEMS Overview
Gas Density Affects
The sonic flow of stack gas through the orifice is affected by the stack gas density and
viscosity, which in turn are dependent on molecular weight, stack pressure, and temperature. A
change in any of these factors will affect the sonic flow and dilution ratio. The primary means
for addressing theses issues include:
Molecular Weight Effect - Gas density changes that result from changing molecular
weight are mainly due to changes in-stack moisture or CO2 concentrations. These
parameters do not vary much in base load units, but both can be monitored and often
are as part of the Part 75 CEMS. With these measurements, empirical corrections can
be made to the dilution ratio.
The molecular weight effect is also a concern in choosing calibration gases. Gases used
in the initial and subsequent QA tests should have a consistent molecular weight,
otherwise a bias can be introduced due to the molecular weight differences.
Stack Pressure - Stack pressure (absolute stack pressure which includes ambient
pressure and stack static pressure) can vary due to changing load or ambient conditions.
Stack pressure can be monitored separately, with the DAHS applying pressure related
correction algorithms to the CEMS data.
Temperature - Temperature also can vary with load and can be monitored separately
with the DAHS correcting the data, as described above. Temperature correction factors
have been more difficult to develop, however, and have not worked well in situations
with temperature changes greater than 50ฐF. (Note - Calibration checks performed
when the source is not operating may not provide valid results due to the temperature
effect). In response, some vendors heat the dilution probe the heated probe is the
same as shown in Illustration 2-5, except electric heating coils are placed around the
probe and controlled to maintain a constant temperature. Another approach for
stabilizing the temperature is to use an out-of-stack dilution system, described in the
next section.
2.3.2.2 Out-of-Stack Dilution System
The out-of-stack dilution system dilutes the sample outside of the stack where it is
easier to maintain a constant temperature. As noted earlier, dilution in out-of-stack dilution
systems is performed in the same manner as with in-stack dilution probes, this time with a
critical orifice and ejector pump. Illustration 2-6 diagrams one manufacturer's out-of-stack
dilution system. The probe in this type of system is a simple tube similar to that used in a
source level system. Note that these systems are also affected by the changes in gas density
described above, and the use of stack moisture or CO2 corrections may be necessary.
Part 75 Field Audit Manual - My 16, 2003 Page 25
-------
Part 75 CEMS Overview
Section 2
Illustration 2-6:
One Type of Out-of-Stack Dilution System
(Gschwandtner and Winkler, 2001)
Educ-tor
Exhaust
Port
*.
Orifite
Protection
Filter
Calibration
and Blowback
Port
Eduttor Jet
o
Dilution
Eductor
Air
Met
^a Flow
' Control
f.i.i.i.i.i.ii-i-i-i-i-i-i-i-i-M.a
KKHHHHHHHHHHHKKKKHHH
V
Sampling
Probe
7=
\
Mounting.
Flanje
He at e d. Aluminum
Filter Chamber
Partitulate
Filter
2.3.2.3 Sample Transport and Dilution Air Cleaning Systems
Sample Lines
The sample line for a dilution system, as in the source extractive system, is often
wrapped in a tube bundle or umbilical which includes the sample lines, blowback lines,
calibration gas lines, heating elements, and electric lines. The sampling line often does not need
to be heated after the dilution air has been added, but heated "freeze protect" lines are used in
regions of the country where sub-zero temperatures may occur, or if the dilution ratio is low.
Dilution Air Cleaning System
The dilution air cleaning subsystem delivers dry, clean, pollutant-free air to the dilution
probe. It consists of a series of particulate and charcoal filters, dryers, and scrubbers, which
reduce CO2, NOX, SO2, moisture, organic compounds, and other compounds in ambient air to
sub-ppm levels.
The dilution air is compressed air provided by the plant's utilities, commonly referred to
as "plant air," or from a compressor dedicated to the CEMS. In either case the compressed air
enters an air cleaning subsystem where it is further cleaned and regulated. If the filters and
Page 26
Part 75 Field Audit Manual - My 16, 2003
-------
Section 2 Part 75 CEMS Overview
scrubbers are changed regularly, and there are no leaks in the subsystem, the dilution air will be
dry, clean, and free of contaminants, including CO2.
The flow of pressurized dilution air through the ejector pump moves the sample to the
analyzers, so a separate pump is not required. The dilution air pressure should be held
relatively constant because changes in the pressure will affect the dilution ratio. Some systems
include mass flow controllers to maintain the dilution flow rate at a constant level.
2.3.3 Calibration Gas System
Part 75 quality assurance requirements include daily calibration error tests and linearity
tests (usually quarterly), which challenge the extractive gas CEMS with calibration gases of
known concentrations. The calibration gases used in the tests include a zero level gas, as well
as low, mid, and high concentration levels based on the span of the monitor. The calibration
gas system consists of calibration gases, gas regulators, valves, and line filters. The gases must
meet the criteria specified in Part 75, Appendix A, ง 5.1.
Calibration gases for the daily calibration error test and linearity tests are injected as
close as possible to the probe (Part 75, App. A, ง 2.2.1). The calibration gas system must
include controls to ensure that calibration gases are injected at the same flow rate, pressure, and
temperature as the sample gas under normal system operation.
There are two common injection locations for source level extractive systems: (1) the
calibration gas is injected into the in-stack probe external filter housing and is drawn into the
sampling system, or (2) the calibration gas may be injected into an internal filter housing
between the probe and sample line at the stack flange. In dilution extractive systems, the
calibration gas must be injected into the dilution probe housing, with the calibration gas drawn
through the sonic orifice. In an out-of-stack dilution system, the calibration is injected prior to
the inlet of the critical orifice.
Calibration gases are also sometimes injected at the analyzers when performing certain
diagnostic tests or calibration adjustments.
2.3.4 Analyzers
Gas analysis methods for extractive systems can be divided into four major categories.
Those categories, with common Part 75 applications, are shown in Table 2-1 and are briefly
described below. More detailed discussion of analyzer operating principles are available in the
references listed in Section 2.7. The CEMS designer will choose the analytical method based
on the overall system design (e.g., dilution extractive versus source-level extractive, wet versus
dry systems). In source level extractive systems, the gas analyzers measure at stack
concentrations in the ppm range, while in the dilution extractive system gas analyzers read in
the ppm or ppb range similar to those of ambient monitors.
Part 75 Field Audit Manual - My 16, 2003 Page 27
-------
Part 75 CEMS Overview
Section 2
Table 2-1:
Common Extractive Gas CEM Analytical Methods
Techniques
Gas Measured
Absorption Spectroscopic Methods
Non-Dispersive Infrared (NDIR)
Gas Filter Cell Correlation (GFCIR)
Differential Absorption (UV) or (IR)
S02, NO,
S02, NO,
S02, NO,
CO2, H2O
CO2, H2O
C02
Luminescence Spectroscopic Methods
Chemiluminescence
Fluorescence
NO, NOX
S02
Electro- Analytical Methods
Electrocatalysis
02
Paramagnetic Techniques
Magnetodynamic
Thermomagnetic
02
02
2.3.4.1 Absorption Spectroscopy
Absorption Spectroscopic methods measure the amount of light absorbed by a pollutant
molecule. The analyzer has four main components: (1) radiation source to produce the light in
the desired range of the spectrum, (2) spectral limiters which further reduce the band width of
the light to specific wave lengths, (3) detectors which measure the light energy, (4) optical
components which direct and focus the light, and (5) components to correct for interfering
gases and drift (e.g., a reference gas cell).
Non-Dispersive Infrared (NDIR)
Non-Dispersive Infrared (NDIR) monitors are commonly used to measure CO2 in
dilution extractive systems. The analyzer measures the degree of absorption of infrared light by
molecules in the sample gas compared to a reference cell containing gas that does not absorb
infrared light in the wavelengths used by the instrument. The ratio of the detector signals from
the two cells is used to determine the light transmittance which is related to the CO2
concentration using calibration curves developed with known gas quantities. The monitors are
called non-dispersive because filters are used to narrow (not disperse) the infrared wavelength
to a small range centered on the absorption peak of the molecule of interest.
Gas Filter Cell Correlation (GFCIR)
Gas Filter Cell Infrared analyzers use a variation of the NDIR technique by using a
reference cell that contains 100 percent concentration of the pollutant measured instead of 0
percent. The reference cell will remove most of the light at the infrared wavelength absorbed
Page 28
Part 75 Field Audit Manual - My 16, 2003
-------
Section 2 Part 75 CEMS Overview
by the compound of interest. This method is more commonly used in in-situ applications, but is
used in extractive systems.
Differential Absorption
Differential Absorption analyzers perform measurements at two different light
frequencies. One frequency is absorbed by the target molecule, while the other reference
frequency is not. The ratio of the absorption at the two wavelengths is correlated to the target
gas concentration. Again, calibration curves are created using known gas concentrations. Part
75 sources use UV non-dispersive photometers for SO2 and NOX measurements. These types
of instruments can be used in wet extractive systems, as water vapor does not absorb light very
well in the UV region. There are also differential absorption analyzers that use light in the
infrared region, particularly for CO2 in wet source-level extractive systems.
A single differential absorption analyzer may also measure a number of component
gases by using multiple wavelengths of light.
2.3.4.2 Luminescence Spectroscopy
Luminescence spectroscopic methods measure the amount of light emitted by an excited
molecule. Analyzers using these methods are very specific for a given molecule and are more
sensitive than absorption spectroscopy or electrochemical methods. As in the absorption
spectroscopic methods, all of the instruments use calibration curves developed from known
target gas compositions to relate the measured light energy to gas concentration.
Chemiluminescence
Chemiluminescence monitors are commonly used for NOX in dilution extractive systems.
Chemiluminescence is the emission of light produced as a result of a chemical reaction, and a
Chemiluminescence NO - NOX monitor measures the amount of light generated by the reaction
of NO with O3. This monitor uses an ozone generator to produce O3 and a catalytic converter
to reduce NO2 in the sample gas to NO before reacting with O3. The monitor can measure both
NO or NOX by sequencing the NO-O3 reactions. First, the sample gas can bypass the converter
and go directly to the reaction cell, measuring the NO. Then, after this reaction, the gas goes to
the converter where the NO2 is reduced to NO and sent back to the reaction chamber to
measure NOX (NO and NO2). NO2 can be determined by subtracting the NO measured by the
first measurement from the total NOX (NO and NO2) measured in the second step.
Fluorescence
Fluorescence analyzers are used to measure SO2 in both dilution and source-level
extractive systems. Fluorescence occurs when a molecule absorbs light at one wavelength; as a
result of the absorbed energy, the molecule emits light at a different wavelength. The analyzer
uses light (either from a continuous or pulsed infrared light source) to irradiate the gas sample.
The light radiated back from the sample is measured by the sensor, after filtering to select a
narrow bandwidth of the fluorescent radiation.
Interference from quenching is a concern for both Chemiluminescence and fluorescence
analyzers. Quenching occurs when the excited molecules collide with other molecules, losing
energy as a result of the collision. This changes the energy state from the level caused by the
Part 75 Field Audit Manual - My 16, 2003 Page 29
-------
Part 75 CEMS Overview Section 2
analyzer chemical reaction or irradiation. For example, in a fluorescence analyzer the excited
SO2 molecule in the gas sample might collide with another molecule, changing its energy state
from what it would have been due to the analyzer irradiation. This can be a problem if the stack
gas composition changes, as different molecules have different quenching affects. It is also a
problem if the calibration gas background concentrations change (single blend - multiblend
calibration gases).
Quenching effects can be limited by using dilution extractive systems, which will result
in a constant background composition (the dilution air). Fluorescence analyzers can also use
ultraviolet light at lower wavelengths to shorten the fluorescence time to reduce quenching
probabilities. Chemiluminescence systems can increase the O3 flow into the reaction chamber to
provide a more constant background concentration.
2.3.4.3 Electro-Analytical Methods
The zirconium oxide (ZrO2) analyzer, an electrocatalytic analyzer, is the most common
O2 analyzer used by Part 75 sources. The analyzer can measure O2 on both a dry and wet basis,
and it is used with source level extractive systems and as an in-situ monitor.
This analyzer uses a heated ceramic material (ZrO2) with a thin platinum catalytic
coating as a solid electrolyte which allows the transfer of oxygen from the reference side of the
cell (maintained at 21 percent O2) to the sample side (continual flow of stack gas with lower O2
concentrations, e.g., 3-6 percent). The sample O2 concentration can be determined by
measuring the electromotive force of the O2 transfer, combined with a stable cell temperature
and reference cell partial O2 pressure.
The ZrO2 electolyte is heated to 850 ฐC. At that temperature O"2 ions catalyzed by the
platinum can move through the material. Combustibles materials in the stack gas sample (CO,
hydrocarbons), can burn at the operating temperatures of the analyzer consuming sample gas
O2. The combustible concentrations, however, are in much lower concentrations (ppm) than
the O2, and have a negligible impact on O2 measured on a percentage basis.
2.3.4.4 Paramagnetic Techniques
Paramagnetic techniques are also used by Part 75 sources to measure O2 Analyzers
using these techniques are only used in conjunction with source level extractive systems, and
water and particulate matter must be removed prior to the monitor.
Molecules that are attracted by a magnetic field are described as paramagnetic, while
those repelled are called diamagnetic. Most materials are diamagnetic, but O2 is paramagnetic
and strongly attracted to magnetic fields compared to most other gases (though NO and NO2
are also paramagnetic and may cause interference if present at high concentrations).
Magneto dynamic
A magnetodynamic analyzer makes use of the effect that O2 has on modifying a
magnet's magnetic field. In a "dumbbell type" of magnetodynamic analyzer, a torsion balance
dumbbell with diamagnetic glass spheres is suspended in a nonuniform magnetic field. The
dumbbell spheres are pushed away from the strongest part of the magnetic field. Oxygen alters
the field causing a change in the dumbbell position. A light source, mirror on the dumbbell, and
Page 30 Part 75 Field Audit Manual - My 16, 2003
-------
Section 2 Part 75 CEMS Overview
detector measure the dumbbell position. Current through a wire encircling the dumbbell creates
an electromagnetic counter-torque which restores the mirror to the position when O2 is not
present. The amount of current required to restore the dumbbell position is related to the
amount of O2 present.
Thermomagnetic
Thermomagnetic analyzers are often called "magnetic wind" analyzers, and are based on
the decrease in the paramagnetic attraction of O2 with increased temperature. The O2 in the
sample gas is drawn into a tube with a heated coil filament and magnetic field at one end. The
O2 enters the tube attracted by the magnetic field. As the molecules are heated the
paramagnetic attraction is decreased, and the heated molecules are pushed out by cooler
molecules with stronger paramagnetic attraction. The O2 flow through the tube creates the so
called "wind," and cools the heating coil reducing its resistance. The change in resistance is
measured and related to O2 concentration. The monitor can be affected by changes in the gas
composition which affect thermal conductivity and the filament temperature. Combustible
materials can also react on the heated filament changing the resistance.
2.4 In-Situ Gas Monitors
In-situ gas monitors are far less common at Part 75 sources. In-situ monitors were
initially designed for high concentration combustion gas applications, not for the lower
pollutant gas concentrations following pollution control devices. Some in-situ analyzers also
had difficulty meeting EPA certification and quality assurance requirements. However, in-situ
monitors do have some advantages over extractive systems. The monitoring system measures
concentrations at stack conditions and eliminates the need for the sample transport and
conditioning systems required by extractive CEMS. Newer designs offer a wider range of
analyzer options, and virtually all point and some path systems can now be calibrated with
calibration gases as required by Part 75.
In-situ monitors are classified as either path or point monitoring systems. (See
Illustration 2-6.) The in-situ point CEMS measures gas concentrations at a single point in the
stack, much like the single probe in a gas extractive system. The term "point" is used when the
sampling is over a short path, but still much less than the stack cross-section. The in-situ path
CEMS measures gas concentrations over an optical path equivalent to the internal stack
diameter by transmitting a light through the flue gas (single pass) and sometimes back (double
pass).
Part 75 Field Audit Manual - My 16, 2003 Page 31
-------
Part 75 CEMS Overview
Section 2
Illustration 2-7:
In-Situ Gas CEMS (Jahnke, 1992)
Transmitter
Receiver
Blower
Path (Single Pass)
Transmitter
Path (Double Pass)
Reflector
Transmitter
"\
Point
2.4.1 Path In-Situ CEMS
Path in-situ CEMS use spectroscopic analytical methods to measure pollutant
concentrations in the flue gas. The systems have the same principle components as an
absorption spectroscopic analyzer described in Section 2.3.4.1: (1) radiation source to produce
the light in the desired range of the spectrum, (2) spectral limiters which further reduce the band
width of the light to specific wave lengths, (3) detectors which measure the light energy, and
(4) optical components which direct and focus the light. In addition, blowers are required to
keep the optics clean of stack particulate.
In a single pass system, a light transmitter and detector are located on opposite ends of
the light path, and the light makes one "pass" along the measurement path. A double pass
system has the transmitter and light source on one end of the sample path and a retroreflector at
the opposite end to reflect the light back to the detector. The light makes two "passes" along
the measurement path.
2.4.2 Point In-Situ CEMS
The in-situ point CEMS typically consists of a measurement probe which contains a
cavity in which the sample gas can be measured either by a sensor or by light absorption. The
probe opening is protected by some sort of particulate filter (ceramic, sintered stainless steel, or
Hastelloyฎ filter). The sample concentrations within the cavity adjust to changing effluent
concentration via diffusion through the filter.
2.4.3 In-Situ Gas Analyzers
In-situ monitors use spectroscopic and electro-analytical techniques similar to extractive
systems described earlier in Section 2.3.4. In extractive system spectroscopic analyzers, the
light interacts with the sample within the analyzer instrument. For in-situ spectroscopic
analyzers, the light interacts with the sample in the sample probe (point in-situ) or across the
Page 32
Part 75 Field Audit Manual - My 16, 2003
-------
Section 2 Part 75 CEMS Overview
stack diameter (path). Analytical methods used by in-situ gas monitors at Part 75 sources are
shown below in Table 2-2.
Table 2-2:
In-Situ Gas Analyzer Methods
Techniques
Gas Measured
Absorption Spectroscopic Methods
Differential Absorption (UV)
Gas Filter Cell Correlation (GFCIR)
Second Derivative Spectres copy (UV)
SO2, NO
CO2,
SO2, NO
Electro-Analytical Methods
Electrocatalysis
02
2.4.4 System Calibration
Daily calibration error tests and linearity tests can be performed on point in-situ gas
CEMS in a manner similar to extractive gas CEMS, by flooding the probe with the calibration
gases. QA tests for path systems are more difficult. Path systems can use a flow through
calibration cell that is placed in the measurement path for linearity and calibration error tests.
The tests must use EPA Protocol gases, and the calibration cell must be located so as to
challenge the entire measurement system.
The system consists of a calibration cell, which is a flow-through gas cell for the zero
and other calibration gases, and a zero mirror to reflect the light back to the detector without
traveling through the stack. The calibration cell should be at the same temperature and pressure
as stack conditions. If the flow-through cell has a length shorter than that of the sample path,
calibration gas at high concentrations (percent levels) maybe necessary. For CO2, the only way
to perform the daily calibration and linearity tests is to have a flow-through cell with the same
path length as the sample path.
A single path system can not use a zero mirror. One approach for single pass systems is
to use a zero pipe combined with the flow through calibration cell. The zero pipe provides an
optical path not affected by the stack gas.
2.5 Flow CEMS
There are three types of flow CEMS in use today at Part 75 sources:
Differential pressure flow monitors,
Thermal mass flow monitors, and
Ultrasonic flow monitors
Part 75 Field Audit Manual - My 16, 2003 Page 33
-------
Part 75 CEMS Overview
Section 2
All of the flow monitors are in-situ monitors, determining flow on a wet basis based on
dynamic measurements of parameters that can be related to the stack velocity at a point or
along a path within the stack.
2.5.1 Sampling Location
As noted earlier in Section 2.2, the
measurement location must provide a
representative volumetric flow over all
operating conditions. Stratified flow
profiles, cyclonic flow, and flow profiles
that change with load all can impact the
choice of measurement location. Multiple
measurement points or paths and the use of
correction factors to calibrate the flow
CEMS to the reference method (pre-
RATA) may be required to meet the
location requirements in Appendix A, ง 1.2.
2.5.2 Differential Pressure
Flow Monitors
Factors Affecting Flow Data Accuracy
Representative sampling points
Sensor/monitor accuracy and stability
(performance specifications in
Appendix A, ง 3.3)
Accuracy of secondary parameter
values (gas temperature, measurement
path length)
Accurate duct dimensions (area
calculation)
Proper calibration
Differential pressure monitors sense the difference between the impact and wake
pressures at tube openings (pitot tubes or multi-point tubes) in the gas flow, and a differential
pressure transducer converts the pressure signals into electric current. The differential pressure
is the difference between the impact and wake pressures. The differential pressure is combined
with stack gas temperature, stack pressure, and molecular weight to determine velocity using
the pitot equation below. Volumetric flow is calculated by multiplying velocity by the stack or
duct cross sectional area.
v
s (abs)
P M
s s
where, vs = stack gas velocity
Kp = dimensionless constant
Cp = pitot tube coefficient
A/? = velocity pressure
Ts(abs)= absolute temperature of the stack gas
Ps = absolute pressure of the stack gas
Ms = molecular weight of the stack gas
Usually the molecular weight of the flue gas and stack pressure are assumed to be
constants and are not measured. Temperature measurements are made either with a
thermocouple or resistence temperature device (RTD). The temperature probe should also be
placed in a location representative of the stack flow profile.
Page 34
Part 75 Field Audit Manual - My 16, 2003
-------
Section 2
Part 75 CEMS Overview
The flow CEMS differential pressure probe may measure at one point in the stack, or at
multiple points using a multi-point averaging probe or multiple pitot tube assembly. (See
Illustration 2-7.)
Differential pressure monitors are sensitive to non-parallel or swirling flow. If the flow
is at an angle to the stack center line there will be a bias in the velocity measurement. The bias
is usually positive.
Illustration 2-8:
Example of Multiple Probe Locations (Jahnke, 1994)
Multi-Tube
Pitot Assembly
Exhaust
Stack
Averaging
Differential
Pressure Sensors
Quality Assurance Issues
Probe and line plugging is minimized by intermittently back purging the line and probe
openings with clean air. Back purging is required as a daily interference check (Part 75,
Appendix A, ง 2.2.2.2) and prevents extreme plugging in most situations. Daily calibration
error tests are performed after the probe and test the pressure transducer and system
electronics. The pressure side of the probe or transducer is pressurized to apre-set level to
check the span, and then both inputs are equalized to test the zero level. Because probes are
subject to corrosion and abrasion, they should be visually inspected periodically in addition to
these tests. Part 75 requires a quarterly leak check of all sample lines for a differential pressure
flow monitor.
2.5.3 Thermal Mass Flow Monitors
Thermal mass flow monitors are based on the principle of thermal heat transfer due to
convection. Gas flowing across a heated surface tends to cool the surface. The molecules of
the flowing gas interact with the heated boundary layer surrounding the surface and remove
heat. The greater the flow rate, the more heat is removed. There are two types of sensors used
by thermal flow monitors: the constant power anemometer and the constant temperature
anemometer.
Both types of anemometers use two RTDs, one heated and the other unheated. The
RTDs are usually protected by a stainless steel tube and mounted together in the stack. The
constant power anemometer measures the temperature difference between the RTDs at a
constant current. The temperature difference responds proportionally to changes in velocity.
The temperature difference will be high at low velocities (less cooling of the heated RTD) and
Part 75 Field Audit Manual - My 16, 2003
Page 35
-------
Part 75 CEMS Overview
Section 2
low at high velocities (more cooling of the heated RTD). The constant temperature
anemometer maintains a constant temperature difference between the RTDs by varying the
current. The change in current is proportional to the change in gas velocity. The higher the
velocity, the higher the current required to maintain the heated RTD temperature; conversely, at
lower velocities a lower current is required.
The sensors can be placed at a single point if the flow profile or in a multiple point array
depending on flow conditions.
Illustration 2-9:
Thermal Mass Flow Monitor Probe (adapted from Jahnke, 1992)
pV -
(Flow)
Th
Heat Removed
Ta
(Th = Temperature of Heated RTD, Ta = Temperature of Unheated RTD)
The output from a thermal flow monitor (an empirical function dependent on the
measured temperature difference, gas composition, and power to the anemometers) is
proportional to mass flow. Information on the flue gas density is required to convert mass flow
to volumetric flow. The gas density is dependent on fuel gas composition, so changes in
moisture or CO2 can affect the system's response.
f(heatloss)*f(Tv-TJ = pVsAs
where, vs = stack gas velocity
f(Tv - TJ = heat loss function
Tv = temperature of velocity sensor
Ts = stack temperature
p = gas density
As = stack cross sectional area
Thermal flow monitors can not be used in locations where there are entrained water
droplets. The water droplets evaporate on the sensor causing a dramatic temperature loss due
to the heat of evaporation, which would be interpreted as caused by the flowing gas. Corrosion
and particulate build up on the sensors can be a problem.
Page 36
Part 75 Field Audit Manual - My 16, 2003
-------
Section 2
Part 75 CEMS Overview
Daily Interference and Calibration Error Tests
The daily interference test check requires a means to ensure on a daily basis that the
probe is sufficiently clean to prevent interference (Part 75, Appendix A, ง 2.2.2.2). The sensors
are usually cleaned by self heating at temperatures of 700ฐF or higher to burn off any adhering
particles or chemical compounds. Daily calibration error tests check that the sensors are
operational and that the system electronics are functioning correctly.
2.5.4 Ultrasonic Flow Monitor
Ultrasonic flow monitors measure the time required for an ultrasonic pulse to travel
across a stack or duct at an angle to the stack flow. The monitors are also called transit time
monitors. The monitor consists of downstream and upstream transducers located opposite of
each other on the stack wall which send and receive ultrasonic pulses. The difference in the
time that it takes for the pulse to go in the different directions between the two transducers (the
time it takes to go downstream is shorter than the time to go upstream) and the distance
traveled are used to calculate velocity. The equation for an ideal flow profile is shown below.
L
Va =
2 - cos0
[1/tl- l/t'2]
where, va = average velocity of the flue gas;
L = distance between the transducers;
0 = angle of inclination;
tj = traverse time in upstream direction;
t2 = traverse time in downstream direction
The most common angle of inclination 0, or path angle, between the transducers at Part
75 sources is 45ฐ. Other angles are possible, and various designs and approaches are used to
optimize signal recognition for better transmission, reception, reliability, and accuracy.
Illustration 2-10:
Ultrasonic Flow Monitor
Part 75 Field Audit Manual - My 16, 2003
Page 37
-------
Part 75 CEMS Overview Section 2
The velocity of the sound pulse through the stack gas is affected by the stack gas
density, so changes in the stack temperature due to load changes, and/or temperature variations
across the sonic path, will affect the flow measurement. In some cases correction factors can
be developed using the pre-RATA testing to account for these changes.
Non-parallel flow will also affect the ultrasonic monitor measurements. If the angle or
pitch of the non-parallel flow is in the same direction from the stack center line as the
measurement path inclination, the flow measurement will have a positive bias. The flow
measurement will have a negative bias if the flow pitch angle and inclination angle are in
opposite directions. Two ultrasonic monitors measuring along different path cross sections (an
x-pattern) can be used in situations with a stratified flow profile that varies with load. One path
is usually sufficient if the stratification is stable.
Daily Interference and Calibration Error Tests
Part 75 requires a means to ensure on at least a daily basis that the probe remains
sufficiently clean to prevent interference (Part 75, Appendix A, ง 2.2.2.2). Blowers are
provided to keep the sensors clean much like on an opacity monitor. The blower air is heated
to the temperature of the flue gas to prevent condensation, which may damage the transducer.
Daily calibrations of this type of flow CEMS are indirect. The electronic calibration process
verifies that the upstream and downstream sensors are working in both the transmitting and
receiving modes. The procedure also verifies that the signal to noise ratio is acceptable.
2.6 Data Acquisition and Handling System (DAHS)
The term "data acquisition and handling system" (DAHS) refers to the CEMS hardware
and software components that take the output from the analyzers, combine it with other
information, and compute hourly emissions. The DAHS acquires and stores the necessary data
(Part 75 requires that the DAHS automatically record all emissions data and the daily
calibration error checks - Appendix A, ง 4). It also computes the emissions and quality
assurance test results in terms of the regulatory requirements, displays the data and produces
the quarterly reports required by Part 75. The DAHS software for Part 75 is highly specialized
to match the electronic data reporting (EDR) format.
In addition to a DAHS, a CEMS requires software/hardware components for system
control functions. These functions include automatic calibration, probe blowback, analyzer
sequencing for time-shared analyzers, error detection, diagnostic routines, and similar tasks.
2.6.1 CEMS Computer Systems
In general there are two approaches for a CEMS computer system. In one approach, a
single computer handles both the CEMS control functions and the DAHS functions. In the
other, there is a separate control system that handles the CEMS functions and may manipulate
data as an interface between the analyzers and a separate DAHS computer. Because of the
complexity of the control and DAHS functions in Part 75 CEMS, it is more typical to have
separate systems for control and data acquisition and handling.
System control hardware options include computers, programmable logic controllers
(PLCs), data loggers, and embedded microprocessors. Besides providing CEMS control
Page 38 Part 75 Field Audit Manual - My 16, 2003
-------
Section 2 Part 75 CEMS Overview
functions described above, controllers also can provide some analyzer data processing prior to
the DAHS. These may include converting analog data to digital, and performing some of the
automatic correction calculations and emission calculations.
The DAHS computer is most often a stand alone personal computer, although some
plants are using the plant's distributive control system (DCS).
2.6.2 Emissions Data Processing
The analyzers may send digital signals directly to the DAHS, or send analog signals that
must be converted to a digital value and scaled. The real time, second-to-second data is first
averaged (1-15 minutes) by the analyzer, control interface, or DAHS. Corrections may be
applied to the data (e.g., pressure/temperature compensation, molecular weight, flow and
moisture monitoring polynomials, sonic velocity correction factors, NOX quenching correction
factors, and dilution ratio settings), and the data then converted to the proper units for the Part
75 formula calculations.
The data must be averaged again to one-hour averages, as required by Part 75. A valid
measurement must be recorded in each quadrant of the hour, except when performing required
Part 75 QA testing (e.g., a daily calibration or quarterly linearity test). For hours in which Part
75 QA testing is performed, only two quadrants of the hour need to be captured
(ง 75.10(d)(l)). The hourly data are then used to calculate the Part 75 emissions, parameters,
and rates based on formulas in Part 75, Appendices D - G, with the CEMS-based formulas in
Appendix F. A table of the required formulas is provided in the EDR v2.2 Reporting
Instructions (August 2002).
2.6.3 QA Test Data Processing
As noted previously, Part 75 requires automatic data capture and calculation of daily
calibration error test results. Other QA test data and results may be entered manually.
Automatic and manual data entry for a Part 75 DAHS are discussed further in Section 4.6 of
this manual, and in Section II.C. 3 of the EDR v2.2 Reporting Instructions.
Some CEMS use the daily calibration error test results to automatically correct the
analyzer data. These mathematical adjustments are similar to an internal bias adjustment factor
in that the DAHS evaluates the response in comparison to the reference value and assigns from
that time forward an adjustment factor which is used to adjust the data for reporting.
2.6.4 Part 75 Reporting
The DAHS generates the quarterly electronic data report (EDR) submitted by the
source to EPA's Clean Air Markets Division. The EDR is in ASCII text format with each line
representing a separate record, which includes plant and unit information, monitoring plan
information, hourly monitoring data, and QA test results. The reports are sent directly to EPA
by electronic data transfer. There is more discussion of the EDR in Section 3.1.1 of this
manual.
Part 75 Field Audit Manual - My 16, 2003 Page 39
-------
Part 75 CEMS Overview Section 2
2.7 References
Gschwandtner, G. and Winkler, J. (2001). U.S. EPA, Acid Rain Program: Continuous
Emissions Monitoring Systems Reference Manual. U. S. Environmental Protection Agency,
Acid Rain Division.
Jahnke, J. A.. (1992). APTI Course 474, Continuous Emissions Monitoring Systems, Student
Manual (Revised). U.S. Environmental Protection Agency, Office of Air Quality and Planning
Standards.
Jahnke, J. A. (1994). An Operator's Guide to Eliminating Bias in CEMSystems. U.S.
Environmental Protection Agency, Office of Air and Radiation. EPA 430-R-94-016.
Jahnke, J. A., and Peeler, J. W. (1997). Handbook: Continuous Emission Monitoring Systems
for Non-Criteria Pollutants. U.S. Environmental Protection Agency, Office of Research and
Development. EPA/625/R-97/001.
Jahnke, J. A. (2000). Continuous Emission Monitoring, Second Edition. John Wiley & Sons,
Inc.
Shigehara, R T., and Peeler, J. W. (1998). Evaluation of Daily Calibration Techniques of
Continuous Flow Monitoring Systems. U. S. Environmental Protection Agency, Acid Rain
Division.
Page 40 Part 75 Field Audit Manual - My 16, 2003
-------
Section 3: Audit Preparation
Section 3 covers the data that are available to help you prepare for the field audit and
explains how to use that data. Preparing for the audit can increase your efficiency by
allowing you to target issues for auditing while at the plant. You also gain credibility
with plant personnel if you come prepared. This section emphasizes the use of the
MDC software to review quarterly electronic data.
3.1 Using Part 75 Electronic Data to Conduct Pre-Audit Reviews
You will have available two types of source file information in preparing for a Part 75
field audit. Hardcopy file information may include correspondence, petition responses, portions
of the monitoring plan (such as system diagrams), reports of previous inspections or audits,
performance test reports, and permits. This type of information may be similar to background
file information available for other types of air compliance inspections and audits.
The second type of information is the quarterly electronic report required under Part 75.
This wealth of electronic data generally is not available under other programs, and provides a
valuable resource that you can use prior to the plant visit to conduct monitoring checks and to
identify potential source monitoring problems. The electronic information includes data on the
unit's monitoring plan, certification and recertification events, QA tests, and emissions and
operating data for each hour of the quarter.
This section first introduces you to the quarterly Electronic Data Report (EDR). The
section then describes how to review reports that EPA sends to affected sources to provide
feedback on each EDR submitted. Finally, the section focuses on how to use EPA's Monitoring
Data Checking (MDC) software to review EDRs.
The format and volume of the electronic data can be daunting at first, especially for
inspectors who have responsibilities for multiple air programs. The MDC software will be a
vital asset to help you access and analyze the EDR, and you should ensure that you take
advantage of this tool as you prepare for an audit.
3.1.1 Quarterly Electronic Data Reports
A facility usually submits each Part 75 quarterly EDR on a unit basis, so one facility may
submit multiple EDRs each quarter. In other cases, one report may be submitted for all units
that exhaust to a common monitored stack or other complex stack configuration. The EDR is
in ASCII text format, with each line representing a separate record referred to as a "record
type" or "RT." An example of the layout of information in a report line or record type is
provided in Illustration 3-1. In this example, a line of hourly NOX emission rate data (RT 320)
is presented. If you want to explore the structure and content of the EDR in more detail, you
should review the most up-to-date version of the EDR Reporting Instructions and Formats
see Section 1.5 for guidance on how to access the EDR guidance documents.
Part 75 Field Audit Manual - My 16, 2003 Page 41
-------
Audit Preparation
Section 3
Illustration 3-1:
Example EDR Data Format for Record Type 320
COLUMN POSITION
14 10 13 19 21 26
36 42 48 50 53
320**1 A1099050100 98.5 1800 0,331 0.33107H0101
Type
Code
Unit Stack ID
1
Hour N
Percent
Date Availability
F -factor
Monitoring
System ID
T^ 7
Ox Emission Load
Rate Range
for Hour _.
Fon
mil a
|n
Adjusted NOx
Emission Rate
for Hour
Method of
Determination
Code
To show you the various data elements you will see in an EDR, Illustration 3-2, below,
provides an example summary of the content of an EDR by record type. The example is for
two Acid Rain CEMS units monitored at a common stack. A number of these record types,
like RT 320 in Illustration 3-1, above, pertain to emissions or operating data that are reported
for each hour of operation, so a quarterly EDR will have many lines of data. A quarterly EDR
for a unit with a single stack may have over 20,000 lines. An EDR in text file format therefore
may look overwhelming, but the MDC software tool allows you to streamline your review and
analysis of the data.
Page 42
Part 75 Field Audit Manual - My 16, 2003
-------
Section 3
Audit Preparation
Illustration 3-2:
Example Summary Of Quarterly Report Content For Two Acid Rain
CEMS Units Emitting Through Common Stack
FACILITY INFORMATION
Type 100 Record
Type 102 Record
(Facility and report data)
(Facility information)
COMMON STACK FOR UNITS 1 & 2
Type 200 Records
Type 201 Records
Type 202 Records
Type 210 Records
Type 220 Records
Type 230 Records
Type 231 Records
Type 300 Records
Type 301 Record
Type 310 Records
Type 320 Records
Type 330 Records
Type 503 Records
Type 510 Records
Type 520 Records
Type 530 Records
Type 535 Records
Type 536 Record
Type 556 Records
Type 601 Records
Type 602 Records
Type 603 Records
Type 605 Records
Type 606 Records
Type 610 Records
Type 611 Records
Type 623 Records
Type 699 Record
(SO2 concentration data: by date and hour)
(NOX concentration data: by date and hour)
(CO2 concentration data: by date and hour)
(Diluent data: by date and hour)
(Volumetric flow data: by date and hour)
(Daily calibration test data: by date and hour)
(Flow interference data: by date and hour)
(Stack operating parameters: by date and hour)
(Quarterly and cumulative emission data)
(SO2 mass emissions data: by date and hour)
(NOX emission rate data: by date and hour)
(CO2 mass emissions data: by date and hour)
(For Unit 1 (Common stack definition table))
(For Unit 2 (Common stack definition table))
(Monitoring systems/analytical components table)
(Formula table)
(Span table)
(Stack operating load data)
(Range of Operation, normal load, and load usage)
(Monitoring system recertification events)
(Quarterly linearity test data)
(Quarterly linearity check results)
(Flow quarterly leak check results)
(Reference data for flow-to-load ratio or GHR evaluation)
(Quarterly flow-to-load or GHR check)
(RATA and bias test data)
(RATA and bias test results)
(On-line/Off-line calibration demonstration)
(QA test extension claim based on grace period)
UNIT 1 (MONITORED AT COMMON STACK)
Type 300 Records
Type 301 Record
Type 504 Record
Type 505 Record
Type 585 Records
Type 586 Record
Type 587Record
(Unit operating parameters: by date and hour)
(Quarterly and cumulative emission data)
(Unit information)
(Unit/program information)
(Monitoring methodology information)
(Control equipment information)
(Unitfuel type)
UNIT 2 (MONITORED AT COMMON STACK)
Type 300 Records
Type 301 Record
Type 504 Record
Type 505 Record
Type 585 Records
Type 586 Record
Type 587Record
CERTIFICATIONS
Type 900 Records
Type 901 Records
Type 999 Record
(Unit operating parameters: by date and hour)
(Quarterly and cumulative emission data)
(Unit definition table)
(Unit/program information)
(Monitoring methodology information)
(Control equipment information)
(Unitfuel type)
(Certification electronic signature)
(Certification statement)
(Contactperson information)
Part 75 Field Audit Manual - My 16, 2003
Page 43
-------
Audit Preparation
Section 3
3.1.2 Quarterly Feedback Reports
CAMD conducts two separate automated reviews of each quarterly EDR submission.
The first is generated by the Emission Tracking System (ETS) as "instant feedback" when the
source submits the file to the EPA mainframe. This instant ETS feedback provides the EPA
"Status Code," which indicates whether the file was accepted and whether any errors were
identified. Certain errors are considered critical and necessitate a resubmission of a corrected
file. Other errors are considered informational and should be corrected by the source in future
quarters but need not necessarily be addressed immediately. Any errors that are identified are
listed in the ETS feedback report with the relevant error code, description, number of hours and
identification of the first hour in which the error occurred. ETS checks include recalculations of
hourly and cumulative emission values, as well as range checks for acceptable values and codes
in various fields.
As part of preparing for a field audit, you can download the ETS feedback report from
the EPA mainframe (if you have access rights) or you can request a copy of the feedback from
your CAMD contact. If errors were identified on the latest ETS feedback, you should review
this report with the source during the on-site inspection and verify that the source has corrected
those errors.
The second automated review is generated by the Monitoring Data Checking (MDC)
software and is sent by email to a facility's Designated Representative shortly after the quarterly
EDR submission period. The MDC feedback consists of monitoring plan and QA test data
evaluations. As with the ETS errors, certain MDC errors are considered critical and must be
corrected by a resubmission of the quarterly file. Note that since MDC is available to sources
to download from the CAMD website, they can run these same evaluations on their quarterly
files prior to submission. The MDC software is discussed in more detail in the following
sections, with guidance on how you can use the software to prepare for the plant visit.
3.1.3 Using MDC to Prepare
for an Audit
3.1.3.1 MDC Overview
EPA developed the MDC
software to allow affected sources,
State agencies, and EPA staff to
enter, analyze, print and export
electronic monitoring plan,
certification and quality assurance
data, and to evaluate hourly
emissions data for Part 75
monitoring. The software also
allows industry users to submit
monitoring plan and certification
data to EPA through standard
electronic data transfer protocols.
Key Checks: Using MDC to Prepare
for the Field Audit
Print a copy of the monitoring plan report
to take with you to the plant.
Evaluate QA Test results (e.g., linearity,
RAT A, flow-to-load), and check for
duplicate tests. Identify if the source has
claimed an exemption or grace period.
Print and review recertification reports.
Also print missing data reports if available.
Check hourly emissions, calculations, and
missing data periods using the MDC Hourly
feature.
Page 44
Part 75 Field Audit Manual - My 16, 2003
-------
Section 3 Audit Preparation
To prepare for an audit, you can use MDC to perform the following tasks in reviewing
the quarterly EDR data:
View monitoring plan, QA test, extension/exemption, and compliance certification
records directly on screen.
Evaluate monitoring plans, quality assurance tests, and extension/exemption records.
Print out reports of monitoring plans, QA test data, or evaluations.
Analyze hourly data with the MDC Hourly data module.
Analyze, view and chart hourly data records for SO2, CO2, O2, NOX, and heat input.
Also, calculate and display summary data for daily, monthly, and quarterly time periods.
Check hourly emissions calculations. (MDC relies on the calculation procedures
identified in a unit's monitoring plan together with RATA test data for these
calculations.) You can also calculate cumulative emissions from the hourly data to
compare to the quarterly and annual values that the source reports in the EDR (see RTs
301 and 307).
Check to determine the quality assurance status (with respect to RAT As and linearity
tests) of measurements reported for specific hours and monitors. (MDC uses the
ongoing reported quality assurance test data and extension/exemption records for this
analysis.)
Archive hourly data calculations and QA status for quick retrieval.
Use a utility in MDC to modify parameter tolerances for comparisons of reported versus
calculated hourly emissions.
The following subsections address these uses in greater detail.
3.1.3.2 Getting Started with MDC
Obtaining MDC Software
EPA's MDC Software and supporting
information can be downloaded at:
www. epa. gov/airmarkets/monitoring/mdc.
The most current version of MDC
along with instructions and training
materials are available on CAMD's website.
Once you obtain MDC and install it on your
computer the next step is to import recent
EDRs for the units to be audited. You will
need first to obtain the quarterly EDRs,
which can be downloaded from CAMD's
website or received on a CD from CAMD.
The EDR files should be saved in the data folder of the MDC directory. You may need
as many as four quarters of the most recent EDRs for the units, and may need to go back
additional quarters depending on when the most recent RATA was performed. QA tests are
included in the EDR for the quarter during which the tests were performed.
Part 75 Field Audit Manual - My 16, 2003 Page 45
-------
Audit Preparation
Section 3
3.1.3.3 Review and Print Electronic Portion of Monitoring Plans
The electronic portion of a monitoring plan provides important background information.
The plan identifies the affected units (e.g., type, rated capacity, fuels combusted), control
equipment, what types of monitoring systems and components are in use, monitor span values,
DAHS formulas, missing data procedures, and other information. (See Table 3-1.)
The plan should be printed out using the MDC reports function and brought to the
plant. The plan printout provides a simple template and check sheet for the plant visit. For
example, when verifying monitor serial numbers to ensure that equipment has been properly
certified you may simply check off on the printed monitoring plan each monitor system
component that matches, and include this as a check sheet in the audit report.
Table 3-1:
Electronic Monitoring Plan Information
Electronic Monitoring Plan Records
Unit Operation Information (RT 504)
Monitoring Systems/Analytical (RT 510)
Components
Emission Formulas (RT 520)
Span Values (RT 530)
Load Range (RT 535)
Range of Operation (RT 536)
Monitoring Methodologies (RT 585)
Control Information (RT 586)
Fuel Type (RT 587)
Fuel Flow Meter Data (Appendix D)
(RT 540)
Fuel Usage Qualification (Appendix E)
(RT 507)
NOX Correlation Segments (Appendix E)
(RT 560)
Description
Unit ID, Boiler Type, Maximum Heat Input, Areas at
Flow Monitor and Stack Exit
System Parameter and ID; Component ID, Type,
Sample Method, Manufacturer, Model, Serial
Number
Parameter, Formula Code, Formula
Parameter, Scale, MPC/MEC/MPF, Max. NOX Rate,
Span Value, Full-Scale Range, Units of Measure
Maximum Hourly Load
Upper and Lower Bounds of Operating Range, Most
Frequently Used Loads, Normal Load
Parameter, Methodology, Fuel Type, Missing Data
Approach
Parameter, Type of Control, Control Dates
Primary/Secondary Fuels
Parameter, Fuel Type, Maximum Fuel Flow, Initial
Accuracy Test Method
Capacity or Gas Usage, Qualification Type, Method
Test Date, Test Number, Operating Level, Segment,
Heat Input, NOX Rate, Fuel Type
Page 46
Part 75 Field Audit Manual - My 16, 2003
-------
Section 3
Audit Preparation
3.1.3.4 QA Tests, Exemptions, and Extensions
Part 75 QA Tests
In the context of Part 75 and this
manual, QA tests refer to quality
assurance tests required by the rule.
The primary tests for electronic data
evaluation using MDC are:
Linearity Tests
Relative Accuracy Test Audits
(RATAs)
Flow-to-Load Tests
The MDC program evaluates QA test data
and provides detailed test reports. You can view
a list of the QA tests (linearity, flow-to-load, and
RATAs) as well as any test grace periods and
exemptions --in the EDRs that you have
imported into MDC by selecting Certification and
QA Tests or Test Extensions and Exemptions
from the Edit or Report pull down menus.
QA Tests
MDC will evaluate QA test data, and
provide brief error messages if errors are found.
For example, low, medium, and high gas levels
are compared to the instrument span to determine
if the linearity was performed at the appropriate levels. The program also will re-calculate
linearity and RATA results including RATA bias adjustment factors (BAF). These are the same
evaluations that are performed by CAMD at the end of each quarter, described above in Section
3.1.2.
If you have not received a copy of CAMD's evaluation, you should run QA test
evaluations of all the QA tests performed in the quarters you have downloaded. Make a note of
errors; also, you can print out specific pages of the MDC evaluation report for follow up with
the plant during the audit visit.
In addition to the test evaluations, you will want to view the list of QA tests performed
by the source to see if a particular QA test (RATA or linearity) has been repeated in a quarter.
This would indicate a failed or aborted test that might be due to CEMS problems. You should
investigate these failed/aborted tests during the plant visit to find out what adjustments (if any)
were made to the system and why they were necessary. To perform this check, view the list of
QA tests for a source by using the 'Edit' pull down menu in MDC. An example is shown in
Illustration 3-3, where the SO2 linearity test has been repeated multiple times for a unit during
the second quarter of 2000.
Part 75 Field Audit Manual - My 16, 2003
Page 47
-------
Audit Preparation
Section 3
Illustration 3-3:
MDC Screen Showing Multiple Linearity Tests in One Quarter
m '
Fie Edit
" ' ' ,'',->'
Reports LME Ollies MDC Hourly Window Help
ll -"'.' .1, k> '"f -" (' . ,,"',,,,V! ': I -' .' ."ป-'.-* 1, '"
Ma 1' " ' - Delete [' '. , Edit f , ' ">',-. . Filter j "t ; , Clear ft
Unit/Stack ID (x)
2
"2
2
"2
"2
"2
>2
2
"2
~2
2
2
2
SysID
203
203
203
203
203
203
203
203
203
203
203
203
203
203
203
203
203
203
203
203
203
203
203
203
203
203
203
Param
302
S02
S02
S02
S02
S02
S02
S02
S02
S02
302
302
302
302
302
302
302
302
302
S02
S02
S02
S02
S02
302
302
302
Comp ID
014
014
014
014
014
014
014
014
014
014
014
014
014
014
014
014
014
014
014
014
014
014
014
014
014
014
014
Type
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
S02A
Test Type ()
Linearity (RT 601/602)
Linearity (RT 601/602)
Linearity (RT 601/602)
Linearity (RT 601/602)
Linearity (RT 601/602)
Linearity (RT 601,602)
Linearity (RT 601,602)
Linearity (RT 601,602)
Linearity (RT 601,602)
Linearity (RT 601/602)
Linearity (RT 601/602)
Linearity (RT 601/602)
Linearity (RT 601/602)
Li ea ty(RT 601/602)
Li ea ty (RT 601/602)
Li ea ty (RT 601/602)
Li ea ty (RT 601/602)
Li ea ty (RT 601/602)
Li ea ty (RT 601/602)
Li ea ty(RT 601/602)
Li ea ty(RT 601/602)
Li ea ty(RT 601/602)
Lines ty(RT 601/602)
Linea ty (RT 601/602)
Linea ty (RT 601/602)
Linea ty (RT 601/602)
Linea ty (RT 601,602)
Reason
D
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
D
0.
a
Q
Q
Q
Q
Test Date ( )
02/24/2000
04/27/2000
05/23/2000
05/26/2000
05/27/2000
06/05/2000
06/27,2000
06/27,2000
06.97,2000
0697/2000
06/27/2000
06/27/2000
06/28/2000
06,98/2000
06,98/2000
06,98/2000
06,98/2000
06,98/2000
06,98/2000
06/25/2000
04,97/2000
05,93/2000
05,96/2000
05,97,9000
06,97,9000
06,97,9000
06,98,9000
Test*
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
IB
20
21
22
23
24
25
26
27
Group ID
'' JSJJ
' 'J,- '
V//A t f close |tl
'"~*
Detail Records Deleted^
.i.
~
r
Right dick
i*stปt
o select and de-seiect records
g] Monitoring Data Checking .
B 51 44 am
.*) 8 51 AM
You should also print out copies of a number of RAT A and linearity test reports to
bring with you to the plant so that you may compare the results reported in the quarterly EDR
to on-site test report records. You should perform this check (see Section 4.6) to ensure that
the source properly transferred Q A test data not automatically recorded by a CEMS to the
CEMS data acquisition and handling system (DAHS).
Extensions and Exemptions
The Part 75 rule allows for the extension of QA test deadlines, and, in some cases,
exemptions from QA test requirements based on specific unit circumstances. Test extensions
are available for RAT As and quarterly Appendix D fuel flowmeter accuracy tests. There is also
a general "grace period" extension for all quarterly QA tests and RAT As, which allows time in
which to complete the testing for a short period after the end-of-quarter deadline.
Exemptions are available from the multi-load flow RATA requirement (flow RATA may
be allowed at a single operating level), SO2 RATA requirement (for units burning a low sulfur
fuel), and the requirement to perform a single load RATA at normal load. In addition, quarterly
QA test exemptions are available for linearity tests, leak checks, and flow-to-load ratio tests in
any quarter which does not qualify as a QA operating quarter (a QA operating quarter has at
least 168 hours of stack or unit operation). Quarterly QA test exemptions are also available for
linearity tests of one range of a dual range monitor (range not used in the quarter), linearity
tests of SO2 or NOX monitors with spans of 30 ppm or less, and flow-to-load ratio tests for
Page 48
Part 75 Field Audit Manual - My 16, 2003
-------
Section 3
Audit Preparation
complex stack configurations that have been approved by petition. The EDR v2.2 Instructions
(see sections that discuss RT 695 - RT 699) provide useful background descriptions of the
available extensions and exemptions and qualification requirements. A link to the EDR
Instructions is provided in Section 1.5 of this manual.
You can use MDC to determine
whether the source has claimed any
extensions or exemptions. Select Test
Extensions' and 'Exemptions' from MDC's
Report pull down menu. Then, run an
evaluation report for all of the extensions
and exemptions to determine if there are any
errors that require follow-up during the field
audit.
Extension and Exemption Record Types
RT 695: Single-Load Flow Rata Claim
RT 696: Fuel Flowmeter Accuracy Test
Extension
RT 697: RAT A Extension or
Exemption
RT 698: Quarterly QA Test Exemption
A common test exemption is the use of a 1-load level RATA for flow RAT As.
Generally, a 2-load test is required for QA testing. At least once every 5 years, a 3-load test is
required for quality assurance purposes. In addition, a 3-load test is required for certification
and recertification.
There are also a few situations in which a 1-load test is allowed without a special
exemption: (1) peaking units and bypass stacks automatically qualify; (2) a source that
conducts flow RATAs on a semiannual basis can alternate between a 2-load and 1 -load test; and
(3) cement kilns and other non-load based units can conduct tests at a single level if
representative of normal operations. For other sources to qualify to use a 1-load level flow
RATA, the source must submit the results of the single load analysis required by Appendix B,
ง 2.3.1.3(c) in RT 695. Check to make sure that flow RATAs were performed at multiple loads
(or that a unit otherwise qualified for a single load test) if the MDC exemption report indicates
that the source did not report RT 695.
Table 3-2:
Summary of MDC QA Test Checks
QA Test Checks
QA test evaluations
Repeated QA tests
Print RATA and linearity reports for
comparison with on-site hardcopy
data
Extensions and grace periods
Description
MDC will evaluate QA test data, recalculate results, and
provide error messages if errors are found.
Look for Failed/Repeated QA Tests (multiple tests in a
quarter).
Bring the reports to the plant to make sure the electronic data
match hardcopy data. For linearity tests, make sure the
reference cylinder gas concentrations match those on site.
Identify if the source has claimed an extension or grace
period.
(cont.)
Part 75 Field Audit Manual - My 16, 2003
Page 49
-------
Audit Preparation
Section 3
Table 3-2:
Summary of MDC QA Test Checks (cont.)
QA Test Checks
Description
Single load flow RATAs (RT 695)
Check for the exemption analysis, single load flow RATA
claim record (RT 695), which will show the results of the
single load analysis required by App. B, 2.3.1.3(c). Make
sure flow RATAs were performed at multiple load levels for
units (other than peaking units) without the single load
exemption request. Note: Non-peaking units performing
RATAs semiannually can alternate between single-load and
multiple-load test without requesting an exemption or
reporting RT 695. Certain non-load based units (such as
cement kilns) also may be exempt from multiple-level tests.
3.1.3.5 Recertification Events and Monitoring System Downtime Reports
Sources must report certification, recertification, and certain maintenance events in the
quarterly EDR (RT 556) in the report for the quarter in which the event occurred. The MDC
report function (currently under the Reports drop down menu) can provide a printout report of
recertification and maintenance events. This report identifies certification, recertification, and
maintenance events and what QA tests were required. If any of these records have been
reported, you should print a copy of this report and bring it with you on the plant visit to
compare against the plant records. It is important that all monitoring components are properly
certified. If there have been changes to equipment, proper diagnostic or recertification testing
should have occurred as required by ง 75.20(b). Note that MDC does not verify that all
required certification, recertification or diagnostic testing is included in the report. EPA
therefore relies on the inspector to make this verification as part of the audit.
A Monitoring System Downtime or Missing Parameter Report (RT 550) may also be
available, though unlike the Recertification Report (RT 556), this record is optional. This
useful report shows the start and end time of all missing data periods by monitoring system,
reason for missing data, and the corrective action taken. You can print out a copy of this
report, and use the report to: (1) target monitoring system problems for further investigation
with plant personnel during the plant visit, and (2) identify missing data periods that can be used
in reviewing plant maintenance plan records or control device parameter records at the plant.
Alternatively, if this report has not been submitted, the MDC Hourly function can be used to
identify missing data periods, as discussed in the following section.
3.1.3.6 Using MDC Hourly to Check Emissions Data and Calculations
The MDC Hourly module analyzes, calculates, views and graphs the hourly emissions
data records for SO2, CO2, O2, NOX flow and heat input. It also calculates and displays
summary data for daily, monthly and quarterly time periods. The program can analyze up to 4
quarters of calendar year data. MDC Hourly was designed for use with EDR v2.1. Use of
EDR v2.1 began in the second quarter of 2000.
Page 50
Part 75 Field Audit Manual - My 16, 2003
-------
Section 3 Audit Preparation
Getting Started with MDC Hourly
In order to check the hourly calculations with MDC Hourly you will need to make sure
that you have imported quarterly EDR files that include the quarter with data from the most
recent pollutant and flow RATAs so that you can take full advantage of the MDC data checks.
If you do not import all of the historical monitoring plan and quality assurance data into MDC
prior to running MDC Hourly, MDC Hourly may mistake hourly data as being out of control
(OOC) and will not provide any further analysis on those hours.
Hourly Calculations and Error Messages
View the Detailed Emissions Data tables for the different CEMS. The program
recalculates emissions for each hour using the monitoring plan formula and provides an error
message in the far right column if there is a discrepancy. There is no need to perform the
emission calculations yourself. One issue is the bias adjustment factor (BAF). If a source uses
the wrong BAF, they could underreport emissions and undermine the integrity of the trading
program. This recalculation feature in MDC can assist you in spotting this problem and
understand the potential magnitude of the impact on emission allowances.
Error messages will also identify OOC periods, which indicate that a QA test (RATA or
Linearity) was expired, failed, or not performed. MDC Hourly does not currently determine the
control status of the unit with respect to daily calibrations. Be careful of the 'out-of-control'
error messages you may generate when you run MDC Hourly. As noted in the Getting Started
discussion above, you can receive these error messages if you do not import the data into MDC
for the quarter in which the RATA was performed prior to running the analysis on the hourly
data.
Missing Data Periods
Use the filter function to identify missing data periods. This is especially important if the
source did not submit the optional RT 550 record described above in Section 3.1.3.5. The
reasons for the missing data are not provided, but this information may be used to identify
systems with high amounts of missing data to target during the plant visit, and also to select
missing data periods for an in-plant records review.
Appendix D Units
For oil- or gas-fired units that use Appendix D fuel monitoring to determine heat input
and SO2 emissions, view and record the fuel gross caloric value (GCV) and sulfur content
across quarters so that you can check these values against on-site fuel sampling and analysis
results.
Graphing
MDC Hourly's graphing function allows you to graph the hourly emission data as well
as flow and heat input. Graphing emissions data can be an invaluable tool for preparing for a
field audit. Graphing allows the inspector to verify that the reported emissions look reasonable
for the type of unit being evaluated. Abnormal emission trends can often be quickly identified
and should be investigated as part of the field audit to identify if the abnormality is due to a
problem with the monitoring system or to a change in the way the unit is operated. An abrupt
Part 75 Field Audit Manual - My 16, 2003 Page 51
-------
Audit Preparation
Section 3
or unusual change could indicate a modification to unit operation or the CEMS that may require
recertification or diagnostic testing. Consistent data with very little change (flat line) may
indicate that the CEMS is not operating properly, or that the unit is using a missing data
routine. Either situation should be investigated with the source during the on-site visit.
Illustration 3-4:
Example MDC Hourly Graph of SO2 Concentrations
Rpt. Adj. SO2 (ppm)
ORIS Code : Slack/Unit ID :
04/12/01 -04/21/01
Apr 17 01 - ll:oo
19.80 (ppm}
Stall! jgJMonitoring Data Checking...] ฉ MDC Hourly Checking Soil.
The graphing function also can be used to compare the hourly emissions or parameter
(flow, heat input) against the monitor span values listed in the monitoring plan. You should
note if the emissions or parameter exceed the relevant maximum potential concentration
(MFC), maximum expected concentration (MEC), or maximum potential stack gas flow rate
(MPF). The emissions should typically be between 20 and 80 percent of the monitor range
identified in the monitoring plan. (See Appendix A, ง 2.1.) If a majority of the
emission/parameter readings are outside 20 to 80 percent of the monitoring range, you should
make sure that the source has performed the MFC, MEC, span, and range checks that are
required at least annually by Appendix A, ง 2.1.
Page 52
Part 75 Field Audit Manual - My 16, 2003
-------
Section 3
Audit Preparation
Table 3-3:
Summary of MDC Hourly Checks
MDC Hourly Checks
Check hourly calculations and error
messages
Check BAFs used in hourly
calculations
Check missing data periods
For oil or gas units using Appendix
D, note fuel GCV and sulfur content
Check spans
View graphs of data trends
Description
Check recalculated hourly emissions for error messages,
which are provided for each hour in the far right column.
Error messages also identify out of control periods, which
indicate that a QA test was failed or was not performed.
Hourly error messages may need to be ignored if the most
recent RATA for a parameter was not imported.
MDC Hourly will use the BAF in the most recent RATA to
recalculate hourly emissions.
Use the filter function to identify missing data periods.
Identify systems with high amounts of missing data, and to
select missing data periods for in-plant records review.
View and record these values across quarters to check against
on-site fuel sampling and analysis results.
Use the graph function to compare the hourly emissions or
parameter (flow, heat input) against the monitor span values
in the monitor plan report. Follow-up with the source if
emissions/parameters are higher than the span. The source is
required to perform a span/range check at least annually
(App.Aง2.1).
Use the graph function to plot emissions and flow. Note any
abrupt changes in the data or consistent data with no change.
3.2 Hardcopy File Review
In addition to the electronic review of monitoring plan and quarterly report data, you
should also review the source's file for written correspondence and information pertaining to
Part 75 monitoring. These documents would include any petitions, previous audit/inspection
reports, linearity and RATA reports, and the source's acid rain permit or operating permit.
3.2.1 Correspondence, Petitions, and Previous Audit/Inspection Reports
Correspondence and any petitions provide background on any recent issues or unusual
monitoring situations that may come up during the plant visit. Previous audit or inspection
reports serve the same purpose. Also, if a prior report identified problems that required follow-
up, you should confirm (either through your pre-audit review or on-site audit) that the source
took appropriate corrective action.
Part 75 Field Audit Manual - My 16, 2003
Page 53
-------
Audit Preparation
Section 3
3.2.2 Linearity Test and RATA Reports
Some States require a hardcopy submittal of linearity test and RATA reports. If this is
the case in your State, and you have access to the reports in your office, you should compare
the hardcopy results to those reported electronically and accessed with the MDC program as
described in Sections 3.1.3 and 4.5, so that you will not have to conduct the comparison while
at the plant. You can also review the RATA test report as described in Table 3-4. The
hardcopy RATA test report review should be focused on reference method documentation that
can not be checked electronically (e.g., reference method analyzer bias/drift checks, cylinder gas
certifications, reference method equipment calibrations, traverse points, etc.).
Table 3-4:
RATA Report Review
RATA Hardcopy Review
Explanation
For a NOX RATA using the
instrumental method 7E, is the NOX
converter efficiency documented
through a performance test?
The NOX converter converts NO2 to NO. NO is
measured by the analyzer. If theNO2 concentration is
greater than 5 ppm, a NOX converter efficiency test
(RM 20) is required.
If an instrumental RM was used, were
cylinder gas certificates included in
the test report?
The calibration gas certificates should be provided in
the test report and show:
Certified gases (EPA Protocol, NIST, etc., see
Section 4.5).
Concentrations which match those used in the
bias/drift check calculations.
Expiration date after the RATA.
For instrumental methods (6C, 7E,
3A), were the appropriate bias and
drift corrections made for the RM
data?
Run results are corrected for bias based on the average
of the before and after bias checks.
For flow reference methods, is the
calibration date of the pitot tube
within 6 months of the RATA?
The most recent calibration should have been
performed within 6 months of the test.
Are CEMS and RM data on same
moisture basis?
Make sure the comparison between the CEMS and RM
are on the same basis.
3.2.3 Permits
You should make sure that the source has an Acid Rain permit or NOX budget permit, if
required. Check any relevant source specific provisions related to Part 75 monitoring in the
permit, and bring a copy of those provisions with you on the audit. The permit conditions for
Part 75 monitoring, however, may only list a general requirement to follow Part 75, and not be
that helpful.
Page 54
Part 75 Field Audit Manual - My 16, 2003
-------
Section 3 Audit Preparation
3.3 Scheduling and Coordinating the Audit
Level 2 audits require coordination with the source to obtain their schedule for
performing RATA and linearity performance tests. The source is required to provide written
notice of the date of periodic RAT As no later than 21 days prior to the first scheduled day of
testing (ง 75.61(a)(5)). Written notification may be provided by mail or by facsimile, and may
be provided by electronic mail if the EPA Regional Office or State agency determines that
electronic mail is acceptable (ง75.61(a)(5)(i)). However, the date can be changed and often is
due to operating constraints at the plant or with the CEMS. Part 75 allows performing a test on
a different date as long as notice of the new date is provided as soon as practicable, but no later
than twenty-four (24) hours in advance of the new test date.
Notice is not required prior to the quarterly linearity tests. So, you will need to
coordinate with the facility staff to schedule your audit to coincide with a linearity test date, if
desired.
Because the audit policy is "hands-off," you should review all of the inspection
procedures and requirements at the time of scheduling to allow the facility the opportunity to
gather the necessary information and arrange for the appropriate personnel to be available.
(The plant environmental contact may also not be authorized to access the CEMS.) Prior
knowledge of all procedures and the audit objectives will help to avoid problems and will allow
the audit to proceed as planned and in a timely manner. You should also check plant-specific
safety requirements with the source, and what safety equipment you will need.
Level 1 unannounced audits of a Part 75 monitoring system are often performed as part
of the overall unannounced air compliance inspection of the facility. If you conduct a Level 1
audit as part of an unannounced inspection, you should also go over the inspection procedures
and requirements as outlined in the previous paragraph when you arrive at the facility.
Unannounced audits may not be as productive since the sources often rely on outside
contractors or personnel from other sites to perform certain tasks. It is therefore more
productive to coordinate with the source before the audit to make sure that the proper
personnel are present to answer any questions that might arise during the audit.
3.4 Materials to Bring
The inspector should bring personal
as provided by agency policy. Useful
materials and data sheets to bring on the field regulatory text for Part 60 and Part 75, and the
audit include:
Note!
safety equipment (e.g., hard hat, safety
glasses, safety shoes, hearing protection, etc.) Sectlon 1 '5 of thls ^^ contams mfฐrmation to
help you find related documents, including
Parts 75 and 76 Policy Manual.
MDC printout of the monitoring plan
MDC printout of linearity and RATA reports
MDC printout of recertification and missing data reports
Appropriate Part 60 reference method requirements if observing a RATA
Part 75 Field Audit Manual - My 16, 2003 Page 55
-------
Audit Preparation Section 3
Checklists and data entry forms (see Appendix A of this manual)
Part 75 Rule (see easy-to-use version maintained on CAMD's website)
Parts 75 and 76 Policy Manual (bring and leave in car for reference as needed)
Copy of Acid Rain permit or operating permit monitoring provisions
Page 56 Part 75 Field Audit Manual - My 16, 2003
-------
Section 4: On-Site CEMS Inspection
Outline of Section 4
4.1 Pre-Audit Interview
4.2 Calibration Error Test
4.3 Probe/Sensors, Sample Lines, and
Sample Conditioning Systems
4.4 Gas Analyzers
4.5 Calibration Gases
4.6 Flow Monitors
4.7 DAHS
4.8 Maintenance Log and Daily Checklists
Review
4.9 QA/QC Plan Review
This section describes the
primary on-site activities that will apply
to any field audit for a Part 75 CEMS.
The field audit will consist of an initial
interview (Section 4.1), the walk through
and inspection of the various
components of CEMS (Sections 4.2 -
4.7), and a review of QA and
maintenance records (Sections 4.8 and
4.9). If you are conducting a Level 2
audit, you will also observe performance
tests (Section 5). For a Level 3 audit,
you will conduct certain performance
tests (Section 6). In each case, you will
complete your audit with an exit
interview and final report (Section 7).
You should not consider the organization of this section as suggesting a specific order
to the audit. For example, this section discusses conducting a walk-through visual observation
of all equipment first, followed by a review of the QA/QC plan and recordkeeping. In many
audits, you may go back and forth between visual observations and records review.
In budgeting your time you should spend more time on items that can not be checked or
verified off-site from EDR submittals or other file information. In addition, focus on any
potential problems flagged during the pre-audit review. For example, focus on CEMS and time
periods with the most missing data, multiple failed QA tests, unusual data trends, or mistakes in
emissions calculations. If you analyzed information using MDC or MDC Hourly, bring a
specific list of the questions and issues based on these pre-audit activities.
In conjunction with your own visual observations and records review, interviews with
plant staff provide important information on how the monitoring systems work and are
operated, as well as on compliance with Part 75 requirements. You should ask the staff how
they operate the equipment and perform QA activities, and how the QA/QC plan is used in
relation to the different CEMS and quality assurance components. A key auditing technique is
to establish a dialogue with personnel at the source and to let them answer your questions by
going through their procedures and showing you where they document the procedures.
You will want to confirm that actual equipment settings, monitor operations, and
quality assurance activities match the QA/QC plan. The QA/QC plan is an important resource
and a Part 75 requirement. The QA/QC plan is the document that provides detailed procedures
for how the CEMS is to be operated. If a source is following a complete QA/QC plan that
covers all of the Part 75 quality assurance requirements and the operating parameters for the
monitoring systems, then the data can be considered valid at all times that the system is
operating within those parameters since the system is being operated as it was during the
RATA. If the QA/QC plan is not followed or is incomplete, then it is questionable whether the
emissions data are of the same quality as was demonstrated during the RATA. You may also
Part 75 Field Audit Manual - My 16, 2003
Page 57
-------
On-Site CEMS Inspection Section 4
find that the procedures currently in use are appropriate, but the QA/QC plan has not been
updated. Many of the recommended checks in this manual refer back to the QA/QC plan.
One critical step is to verify that the CEMS components match those in the monitoring
plan. If changes have been made, you need to determine if the proper recertification testing or
diagnostic testing has been performed. Part 75 quality assurance begins with the initial
certification of the CEMS, and any changes since that certification need to be evaluated for their
effect on CEMS data.
Identifying Monitoring Hardware Recertification Issues
Recertification is required for a replacement, modification, or change that may significantly
affect the ability of a CEMS to accurately measure monitored parameters:
Complete System Replacement
Analyzer Replacement
Change in Orientation or Location of Sampling Probe or Site
Change in Flow Monitor K factor or Polynomial
This is a key issue for audits. See Section 1.5 of this manual for a link to CAMD's
guidance on recertification under Part 75.
4.1 Pre-Audit Interview
You should conduct a pre-audit meeting when you arrive at the plant. The meeting
should include the plant contact, the plant CEMS technician assigned to assist with the audit,
and may include plant management personnel. At this time, it is important for you to make sure
the plant personnel understand the general scope of the audit, and to agree upon a tentative
audit schedule so that necessary personnel will be available when needed. Table 4-1 identifies
several items you should discuss in the interview.
Page 58 Part 75 Field Audit Manual - My 16, 2003
-------
Section 4
On-Site CEMS Inspection
Table 4-1:
Pre-Audit Interview Items
Topic
Purpose
Audit purpose and agenda
Inform facility of the purpose and scope of the
audit.
Streamline subsequent activities.
Allow for meetings with necessary plant personnel.
Inspection and audit techniques to be used
including hands-off policy and the need for
a CEMS technician
Streamline subsequent activities.
Identify plant contacts for different parts of the
audit.
Specific areas to be observed
Identify any constraints on visiting locations in the
plant.
Safety requirements
Identify necessary safety equipment and plant safety
issues.
Records to be reviewed and copying needs
Streamline subsequent activities.
Provide source an opportunity to collect information
during the other audit activities.
Notify plant personnel of intent to take
photographs or videos to document
observed conditions
The inspector should ask for permission to take
photographs on the plant property and explain the
purpose for the photos. The source can make a
confidentiality claim, but emission-related
information is public information under the Clean
Air Act. You should follow your agency's policy on
the treatment of confidential business information
claims. This is not likely to be an issue if photos or
videos are limited to the CEMS components.
4.2 Calibration Error Test
Part 75 requires a daily calibration error test. This test provides a simple check of most
CEMS components, and can detect many, but not all, sources of potential error in the
monitoring system.
If possible you should observe routine daily calibration error tests as part of any CEMS
field audit, but often the tests are initiated automatically at a specific time of day. Following the
calibration sequence, the DAHS will typically generate a calibration summary report (presenting
the CEMS responses and calculated calibration error results) to record the data for each
operating day. Daily calibration check results are also reported electronically in the EDR.
If the daily checks have occurred before you arrived at the plant, review the results of
the tests for that day and request the plant contact or CEMS technician to initiate a routine daily
calibration test. Ask for an explanation of how daily calibrations are performed, how the results
are used, what responses will trigger adjustments, and how adjustments are made. The specific
procedures used to conduct this calibration routine should be described in the QA/QC plan
available on-site. Adjustments are not to be made during the calibration error test. When
Part 75 Field Audit Manual - My 16, 2003
Page 59
-------
On-Site CEMS Inspection Section 4
adjustments are made following a calibration test, another calibration error test is required after
all adjustments are completed to validate that the adjustment was appropriate.
The Daily Calibration Summary form included in Appendix A can be used to compile
the daily calibration check data while on-site. You should also request a printout from the
source's DAHS and compare these values to the values you recorded from the analyzer's
display. For each monitor:
Record the gas monitor responses from the zero and upscale calibration gas injections
(for dilution systems multiply the monitor data by the dilution ratio to obtain the actual
concentration).
For flow monitors record the results of the electronic tests for ultrasonic and thermal
flow monitors, or the pressure transducer checks for differential pressure monitors.
Note the start and stop times, and whether the response is stable when the system
records the calibration test response. Also check that the time taken for the test still
provides at least two valid CEMS data points per hour to meet the valid hour
requirements of Part 75.
Check that the calibration gas flow rates and pressure match sampling conditions.
Calculate the results from the monitor data and retrieve the daily calibration error (CE)
results from the daily report generated by the DAHS. Compare your results to those
generated by the DAHS. If there is a discrepancy ask about correction factors in the
DAHS, such as pressure correction factors for dilution probes or similar factors.
Compare the calibration error at the zero and high levels to the Part 75 data validation
requirements in Appendix B, ง 2.1.4(a) (see Table 4-2).
If the calibration error check is failed, the data from the monitor are invalid and the
monitor is out of control until it successfully passes a subsequent calibration error test. Missing
data routines must be used until the monitor is adjusted and successfully passes a subsequent
calibration error test.
Page 60 Part 75 Field Audit Manual - My 16, 2003
-------
Section 4
On-Site CEMS Inspection
Table 4-2:
Part 75 Calibration Error Test Data Validation Requirements
Monitored Parameter
SO2orNOx
CO2 or O2
H2O
Flow
Calibration Error Requirement (App. B, ง 2.1.4(a))
< 5.0% of the Span Value, or
< 5 ppm absolute value of the difference between the monitor response
and the reference value if the span value of the monitor is less than 50
ppm, or
< 10 ppm absolute value of the difference between the monitor
response and the reference value if the span value of the monitor is
greater than 50 ppm but less than 200 ppm.
< 1.0%CO2orO2
< 6.0% of the Span Value. Moisture monitor systems composed of wet
and dry O2 monitors must meet the O2 calibration error requirement of
< 1.0%.
< 6.0% of the span value, or
< 0.02 inches of water absolute value of the difference between the
monitor response and the reference value if the monitor is a differential
pressure type.
You should also compare the CE results you obtain to the results of the most recent
automatic calibration error test that occurred earlier in the day of your audit or on previous
days. Previous calibration error tests usually can be brought up on the DAHS visual display
screen. If there is a significant shift in the CE results from that previous test, ask the facility
contact if they can explain the shift. Were adjustments made to the monitoring system(s) that
could cause the shift? Does the shift indicate that adjustments are necessary?
Routine adjustments of the monitor are allowed after successful calibration error tests.
These adjustments are to be made to bring the monitor as close as practicable to the calibration
gas tag value or flow reference signal. If the monitor is physically adjusted by a person, a
follow-up calibration error test is required to verify that the adjustment was performed properly.
No follow-up calibration error testing is required when only a mathematical adjustment is made
automatically by the DAHS. These mathematical adjustments are similar to an internal bias
adjustment factor in that the DAHS evaluates the response in comparison to the reference value
and assigns from that time forward an adjustment factor which is used to adjust the data for
reporting. If this type of adjustment is used, the inspector should determine at what point the
source physically adjusts the calibration of the analyzer (i.e., what the range of acceptable
calibration error adjustment factors is). Typically, sources that use the auto adjustment in this
manner physically recalibrate when the mathematical adjustment exceeds 2.5 percent of the
span. The criteria that are used should be documented in the source's QA/QC plan.
Part 75 Field Audit Manual - My 16, 2003
Page 61
-------
On-Site CEMS Inspection Section 4
4.3 Probe/Sensors, Sample Lines, and Sample Conditioning Systems
Most of the suggested checks of the probe/sensors and the sample lines do not require a
visual check of the actual probe/sensors or lines. The probe/sensors and sample lines may not
be easily accessible and may be located in a confined space that your agency's (or the plant's)
safety policies prohibit you from entering. Access to the gas sampling probe and flow monitor
probe/sensors also may require climbing ladders or working at substantial heights.
4.3.1 Probe/Sensors
You should ensure that the probe/sensors are in the same location as when the unit was
certified/recertified (as detailed in the monitoring plan) changes in probe/sensors orientation
or location require recertification of a CEMS. A hardcopy submittal of the monitoring plan will
contain a diagram with the probe/sensors location. A copy should have been submitted to the
State agency. If it is not available in your files, it should be on file at the plant. In addition, ask
how frequently the probe/sensors are inspected, if there have been any problems (plugging for
example), and if the probe/sensors or a component have been recently changed. The change
might also require recertification or diagnostic testing.
Flow Monitors
In addition, for flow monitors, you could ask the source representative to perform a
daily interference check as required by App. B ง 2.1.2, or review the results of the most recent
interference tests. The daily interference check tests the flow monitor probe or sensors, sample
lines, and temperature transceiver for plugging or malfunctioning.
The interference test procedure should be identified in the QA/QC Plan. Go over the
interference test procedures with the source representative, and ask if there have been any
recent failures, and what corrective action was taken. Some interference checks for various
types of flow monitors are described below:
Differential Pressure - Regular back purge of probe and sample lines. Back purge flow
or measured flow is monitored and compared to baseline values to indicate if there is a
plugging problem. There may also be a moisture removal system, or a heated sample
line.
Ultrasonic - Purge air blowers may be installed to prevent particulate build up on the
transceivers, and purge air flow rate is monitored. Units without purge air blowers can
monitor signal strength to identify particulate build up.
Thermal - Temperature sensors may have an auto self clean feature which heat the
sensors to burn off particulate build-up and help avoid moisture condensation. Another
approach uses short high pressure blasts of air. Interference can be tested by checking
the stack temperature measured by both sensors, or by comparing the sensor
measurements against calibration data.
Page 62 Part 75 Field Audit Manual - My 16, 2003
-------
Section 4
On-Site CEMS Inspection
Table 4-3:
Probe/Sensor Check Summary
What to Check
Description
All Probe/Sensor Types
Is the probe in the same stack or duct
location as in the monitoring plan?
Has the duct or stack location been modified?
(Dimensions)
How often are the probes/sensors inspected,
have there been problems, and have the
probe/sensors or components been changed?
Visually compare probe location to location on the
diagram submitted with the monitoring plan. A
measurement of distances from disturbances or
stack/duct diameter are not necessary.
Ask the source representative, and note during a
check of CEMS maintenance logs. These duct/stack
changes could affect CEMS measurements, and may
require recertification.
Ask the source and note during a check of CEMS
QA/QC Plan and maintenance logs.
Flow Monitors
Ask the source to perform a daily
interference check, or review the results of
the most recent interference check.
Appendix B ง 2. 1 .2 requires a daily interference
check for flow monitors. Each flow monitor is to be
designed to provide a means for checking interference
from plugging of each sample line and sensing port,
and malfunction of each resistance temperature
detector (RTD), transceiver or equivalent.
4.3.2 Sample Lines
Extractive Gas CEMS Sample Lines
The daily calibration error test (see Section 4.2) provides the best check on whether gas
sample line leaks are diluting the sample gas measured by the analyzer. The daily calibration
error test should also detect most problems from condensation. However, for source level
extractive NOX systems that do not have an SO2 analyzer, if practical, you should also visually
check the umbilical line as it enters the CEM shelter. Look for loops or sags in the line, as well
as moisture droplets. Because calibration gases do not have significant amounts of NO2, the
daily calibration error test for a NOX monitor may not detect a negative bias from absorption of
NO2 by condensed water.
Differential Pressure Flow CEMS Sample Lines
Manual or automatic quarterly leak checks are required for differential pressure flow
monitors (Appendix B ง 2.2.2). The test criteria are not specified by Part 75. Most often the
test checks the line after the probe flange, by closing the line and holding a steady pressure for a
specified time. Check how and when the tests were performed. The test date and results
(pass/fail) are reported in the EDRs, so you can confirm the reported electronic results with on-
site data. If a recent test failed, find out what was done to correct the problem. A failed leak
Part 75 Field Audit Manual - My 16, 2003
Page 63
-------
On-Site CEMS Inspection Section 4
test results in out of control data until the leak is fixed and a subsequent leak check is passed
(Appendix B ง 2.2.3(g)).
4.3.3 Dilution Air and Gas Sample Conditioning Systems
The majority of Part 75 gas CEMS are extractive systems that transport the stack gas
from the stack to an analyzer. The checks described in the next two sections relate to the
systems that condition the stack gas for analysis.
4.3.3.1 Dilution Extractive Systems
Dilution systems are frequently used for Part 75 compliance. These systems dilute the
stack gas with clean dry air prior to the analyzer. The ratio of sample taken from the stack to
dilution air is the dilution ratio and is an important parameter of a dilution system. How that
ratio is maintained, or what mathematical compensations are made to account for changes in the
ratio, is important to cover during an audit.
Dilution Air Ratio
The following questions maybe useful for auditing dilution extractive CEMS systems:
What is the dilution ratio and how is it verified? This should be described in the QA/QC
plan.
Is the dilution probe ejector pump vacuum at or below the certification value? The
ejector vacuum affects sample flow if it decreases below the values which creates critical
flow through the orifice. How often does the source check the ejector vacuum, as well
as the dilution air and analyzer flow settings? Some systems automatically adjust the
dilution air pressure from a pressure transducer signal, so ask how this works. Again,
these activities should be described in the QA/QC plan.
Check if any corrections are applied to the dilution ratio for changes in pressure,
temperature, or molecular weight (See Section 2.3.3.1 on gas density effects). These
corrections are typically applied by the DAHS, and should be described in the QA/QC
plan. Also find out if there have been any changes in the factors since the last RATA
because a RATA is required following a change in these factors.
You should also ask the source if the dilution probe orifice has been changed. The
orifice controls the sample flow rate; if the orifice has changed it may have changed the
dilution ratio. You can also check this during the maintenance record reviews.
If the orifice has been changed, ask the source if there is a procedure for doing this in
the QA/QC plan. In addition, ask what prompts making a change. Replacing a dilution
probe orifice with one of the same size requires diagnostic testing, while replacement by
a different sized orifice requires recertification. Check to see that the appropriate testing
has been completed.
Page 64 Part 75 Field Audit Manual - My 16, 2003
-------
Section 4
On-Site CEMS Inspection
Dilution Air Cleaning System
The dilution air supplied to the dilution probe must be treated to remove contaminants
that could interfere with the analyzers. The dilution air cleaning system removes moisture, CO2,
particulate matter, hydrocarbons, and other contaminants that may be present in ambient air. A
good place to start is simply to ask the source representative how the system works. You
should also investigate maintenance practices. For the air cleaning system, ask the source how
often the filters are changed, and at what point drying agents are replaced. Again, this should
be verified against the Q A/QC plan documentation. You can also ask to see a copy of the filter
replacement documentation in either the maintenance log or a copy of the daily checklist. You
can check the inlet and outlet pressures of the CO2 filter and compare to the appropriate range
for the inlet and outlet pressure in the QA/QC plan. Find out from the source how often these
pressures are checked and how these checks are documented.
Table 4-4:
Summary of Dilution Air System Checks
What to Check
Description
Has the dilution probe orifice been
changed?
The orifice controls the sample flow rate, so a change in the
orifice will affect the dilution air ratio. Ask the source
CEMS operator if the orifice has been changed since the last
RAT A, and note maintenance records for any evidence of
changes (Section 4.7).
If the orifice has been changed ask if there is a procedure for
changing the orifice in the QA/QC Plan, what prompts a
change, and how is the dilution ratio verified.
If it has been changed, confirm that necessary
diagnostic/recertification testing was performed
Is the dilution probe ejector pump
vacuum at or below the certification
value?
Are the dilution ratio, and dilution
air and analyzer flow settings
properly set?
Ask the CEMS operator how often they check the ejector
vacuum, as well as the dilution air and analyzer flow
settings, and ask whether the values are recorded, and what
are the proper settings.
Also ask the operator to show that the pressure gauges or
rotameters are set in accordance with the QA/QC plan, and
ask how often the values are verified and documented.
Are correction factors applied to the
dilution ratio for changes in
pressure, temperature, or molecular
weight? Have these been changed
since the last RAT A?
Some systems apply correction factors to the dilution ratio
to account for changes in gas density. Changes to the
correction factors should be recorded in the maintenance
log, and the QA/QC plan should outline the procedures for
changing the correction factors. A RATA should be
performed following any change.
(cont.)
Part 75 Field Audit Manual - My 16, 2003
Page 65
-------
On-Site CEMS Inspection
Section 4
Table 4-4:
Summary of Dilution Air System Checks (cont.)
What to Check
Description
Check the inlet and outlet pressures
of CO, air cleaner filter.
Check CO2 air cleaner.
Ask the CEMS operator how often they check the CO2 air
cleaner filter inlet and outlet pressures, if the values are
recorded, and the proper setting.
Also ask how often the filters are changed, and at what point
drying agents are replaced. Again this should be verified
against the QA/QC plan documentation
4.3.3.2 Source Level Extractive Systems
Source level extractive systems transport the stack gas to the analyzer without dilution.
These systems also are described as direct extractive or non-dilution systems. These systems
may measure emissions on either a wet or dry basis. Wet systems need to maintain the
extracted sample at a temperature above the dew point of the sample to avoid condensation in
the umbilical lines. Therefore, each component of the sampling and analysis system in a wet
extractive system must be heated. Dry systems require some sort of moisture removal, either
through condensation or permeation driers, as well as heating of the components upstream of
the moisture removal system. The daily calibration error test may show problems related with
the conditioning system (except for NOX systems as noted in Section 4.3.2).
In addition to observing a daily calibration error test, ask the source representative to
describe what maintenance is performed on the conditioning system, and check that information
against the maintenance log and QA/QC plan. You can also check the chiller temperature, if
water is removed by condensation, and compare the temperature to the acceptable ranges in the
QA/QC plan. These system checks are summarized in Table 4-5.
Table 4-5:
Summary of Source Level Extractive System Checks
What to Check
Observe a daily calibration error test.
Check the umbilical lines entering the
CEM shelter for condensation.
For a dry system using chillers check the
chiller temperature.
Check general conditioning system
maintenance practices.
Description
Ask the source to perform a test. This can be initiated
by the source representative from the analyzer or
DAHS. (See Section 4.2)
Visual check as described earlier in Section 4.3.2. Are
there water droplets visible in the line?
Ask the CEMS operator what the proper temperature
range is, and how often it is verified. Compare against
the QA/QC plan.
Ask the source representative how the conditioning
system is maintained. Compare to the QA/QC plan,
and check the maintenance records.
Page 66
Part 75 Field Audit Manual - My 16, 2003
-------
Section 4
On-Site CEMS Inspection
4.4 Gas Analyzers
Analyzers for extractive systems often are located in a CEM shelter and generally will be
rack-mounted and connected to the conditioning system, usually with a sample manifold. In-
situ analyzers are located on the stack or duct or in the annulus between the stack and stack
liner.
In checking the analyzers, it is
important that you first verify that each
analyzer in use is the same as the one that
was previously certified or recertified.
Analyzers may have been changed without
the notification and recertification required
by Part 75 (Subpart C, ง 75.20(b)). As
noted in previous sections, it is important
that the source certifies the equipment
actually in use.
Reminder - MDC Monitoring
Plan Report
A printout of the MDC monitoring plan
report can be used as a convenient check off
sheet when verifying the serial numbers of
CEMS components.
Analyzer Displays/Alarms
Displays and alarms will vary by analyzer model. You should ask the plant contact what
displays and alarms are in the model that the source uses, and what they indicate. The displays
should show that the analyzer is operating properly. An alarm light may show a potential or
developing problem which should be addressed as described in the QA/QC plan. Ask the
source when an alarm last occurred and what caused the alarm. This information should be
recorded in the maintenance log.
Flow Rate
Check the sample flow rate to each analyzer.
rotameter. The proper flow rates should be
identified in the QA/QC plan.
Range Settings
Ask the source what the range
settings are for each analyzer and how these
settings are displayed. Compare the settings
to the monitoring plan (RT 530, Column
49). You should also check to see if the
source has conducted an annual span/range
check as required by Part 75, Appendix A,
ง2.1. The range may need to be changed if
that setting is not representative of the
concentrations being measured or if the
stack concentrations have changed (such as
a change in fhel supply or new control
equipment). The rule requires that the
evaluation of the MPC, MEC, span and
range for each gas monitor be conducted at
least once a year.
It may be displayed digitally or by a
Span versus Range Under Part 75
The span is the calculated, quality-
assured portion of a monitor's
measurement range.
The span is equal to 1.0 to 1.25 times the
Maximum Potential Concentration
(MPC) or Maximum Potential Stack Gas
Flow Rate (MPF).
The range is the actual setting of the
monitor.
The range is to be set so that the
majority of readings fall between 20% to
80% of the range selected.
Range is always > Span
Part 75 Field Audit Manual - My 16, 2003
Page 67
-------
On-Site CEMS Inspection
Section 4
Correction Factors
A correction factor may be applied to NOX analyzer data to account for NOX quenching
effects. Corrections may also be applied to moisture monitors. The correction is applied to the
analyzer data by the DAHS. You should ask if there have been any correction factor changes
since the last RATA. Any change should be recorded in the maintenance record, and the
procedures for making the change in the QA/QC plan. Verify that a RATA has been performed
following the change (see ง 75.20(b)).
Table 4-6:
Summary of Analyzer Checks
What to Check
Description
Have analyzers been changed without the
notification and recertification required by
Part 75? (Subpart C, ง 75.20(b))
Ask the source if there have been any changes since the
last audit or certification/recertification. Document the
serial numbers and compare to those in the monitoring
plan.
Status of the control panel lights,
indicators and alarms? Displays should
show that the analyzer is on and operating
properly.
The displays will vary by analyzer, so ask the source
what the displays are and what they mean. An alarm
light could indicate a potential problem that needs to be
addressed (check QA/QC plan).
Check range setting, and whether the
source has performed the annual
span/range check required by Appendix A,
ง2.1.
Ask the operator what the range settings are and how
they are displayed. Compare the range setting to the
value in the monitoring plan (RT 530), and the results
of the recent span/range check.
Check the sample flow rate.
Compare the sample flow rate if displayed by a
rotameter or a digital reading to the QA/QC plan.
Have there been any changes to correction
factors (NOX quenching and moisture
monitors) since the last RATA?
Changes to the correction factors should be recorded in
the maintenance log, and the QA/QC plan should
outline the procedures for changing the correction
factors. A RATA should be performed following any
change. QA testing may also be required (see
ง 75.20(b)).
4.5 Calibration Gases
Calibration gas cylinders are used for daily zero and span calibrations and/or linearity
checks. You should spot check a number of the calibration gases to verify that the gases in use
meet protocol gas requirements in Part 75, that the calibration gas concentrations meet Part 75
quality assurance test requirements, and that the values are entered correctly in the DAHS.
You also can visually check the delivery system.
Make a note of the cylinder gas numbers (an engraved ID number stamped on the
cylinder) and check the gas certificate for each ID number for the following information (the
certificate should be on the cylinder, on file electronically, or inhardcopy at the facility):
Page 68
Part 75 Field Audit Manual - My 16, 2003
-------
Section 4 On-Site CEMS Inspection
Meets Calibration Gas Certification Requirements
Expiration Date. Cylinder gases are not certified after the expiration date. The use of
an expired cylinder is not in compliance with ง 75.2l(c).
Check the type of gas certification. Part 75 requires the use of a calibration gas as
defined in ง 72.2; Table 4-7, below, describes the types of permissible Part 75
calibration gases.
Meets Gas Concentration Requirements
Check that the zero air calibration gas concentrations are certified by the supplier to
meet the concentration limits in ง 72.2. (See Table 4-7.) Zero air material is a
calibration gas that may be used to zero an SO2, NOX or CO2 analyzer. Zero air material
has an effective concentration of 0.0% for the component being zeroed (SO2, NOX and
Total Hydrocarbons < 0.1 ppm, CO < 1 ppm, or CO2 < 400 ppm), and is free of certain
other interfering gaseous species. A zero air cylinder containing a multi-component
mixture should be certified that it meets the concentrations above, and that other
components do not interfere with the CEMS reading. For more on zero air calibration
gas or zero air materials, see Questions 10.2 and 10.3 in the Parts 75 and 76 Policy
Manual.
Determine the cylinder gas concentration values and verify that the values are in the
correct range for the instrument span. You also should record the concentrations to
make sure the values are consistent with the values entered into the DAHS for daily
calibration error and linearity tests. The ranges for the low, medium and high linearity
points are:
-- Low level: 20% - 30% of span
-- Mid level: 50% - 60% of span
-- High level: 80% - 100% of span
Part 75 Field Audit Manual - My 16, 2003 Page 69
-------
On-Site CEMS Inspection
Section 4
Table 4-7:
Part 75 Calibration Gases (Appendix A, ง 5.1)
Calibration Gas Type
Acronym
Description
NIST - standard reference
material
SRM
Calibration gas obtained from the National Institute
of Standards and Technology (NIST).
NIST - standard reference
material-equivalent compressed
gas primary reference material
PRM
Gas mixtures listed in a declaration of equivalence
in accordance with section 2.1.2 of the "EPA
Traceability Protocol for Assay and Certification of
Gaseous Calibration Standards," September 1997,
EPA-600/R-97/121 (EPA Traceability Protocol).
NIST - traceable reference
material
NTRM
Calibration gas mkture tested by and certified by
NIST to have a certain specified concentration of
gases.
NIST/EPA-approved certified
reference materials
CRM
Calibration gas mkture that has been approved by
EPA and NIST as having specific known chemical
or physical property values, certified by a
technically valid procedure as evidenced by a
certificate or other documentation issued by a
certifying standard-setting body.
Gas manufacturer's intermediate
standard
GMIS
Compressed gas calibration standard that has been
assayed and certified by direct comparison to an
SRM, an SRM-equivalent PRM, a CRM, or a
NTRM, in accordance with section 2.1.2.1 of the
EPA Traceability Protocol.
EPA protocol gas
Vendor-certified to be within 2.0 percent of the
concentration specified on the cylinder label (tag
value), using the uncertainty calculation procedure
in section 2.1.8 of the EPA Traceability Protocol.
Zero air material
Calibration gas certified by gas vendor:
SO2, NOX and Total Hydrocarbons < 0.1 ppm,
CO < 1 ppm, or CO2 < 400 ppm. If a mixture, the
other components are certified not to interfere with
the CEM readings for the target compound.
Research gas mixture
ROM
Calibration gas mixture developed by agreement of
a requestor and NIST that NIST analyzes and
certifies as "NIST traceable."
In addition to the certificates, you should also check the calibration gas line pressure
gauges to determine the cylinder gas pressure. The cylinder should not be used if the cylinder
gas pressure is below 150 psi. (This requirement is in the EPA Traceability Protocol for Assay
and Certification of Gaseous Calibration Standards, ง 2.1.6.4.) You should check to see if the
facility checks cylinder pressure as part of its QA/QC plan. At what point do they replace a
calibration standard? Also check the outlet regulator pressure or calibration gas flow rate to
see if it matches the QA test procedures in the QA/QC plan.
Page 70
Part 75 Field Audit Manual - My 16, 2003
-------
Section 4
On-Site CEMS Inspection
Table 4-8:
Summary of Calibration Gas Checks
What to Check
Description
Check the certificates for the expiration
date of the cylinder gas.
Gas cylinders may have expired dates. Cylinder gases are
not certified after expiration date, and the use of an
expired cylinders is not in compliance with
ง 75.21(c).
Check the calibration gas type (certified,
EPA Protocol, or other).
Gas certifications must meet the definitions in ง72.2.
Check the type of cylinder gas against the descriptions in
Table 4-7 and ง 72.2.
Check the zero air material
documentation to ensure that it is
properly certified.
Calibration gas used to zero a gas analyzer. See Table 4-
7 for the zero air material concentrations defined in
ง 72.2. (Also see Policy Manual questions 10.2 and
10.3.)
Check the concentration values for each
cylinder, and that the cylinder calibration
tags are withing the correct
concentration range for the span.
Linearity test point ranges are shown below:
Low level: 20% - 30% of span
Mid level: 50% - 60% of span
High level: 80%-100% of span
Record the concentration values to check
against the values recorded by the
DAHS for calibration error and linearity
tests.
This check is to ensure that the proper cylinder gas
concentration values have been entered into the DAHS for
calculation of daily calibration error tests.
Read the cylinder regulator pressure.
The cylinder should not be used if the cylinder gas
pressure is below 150 psi (EPA Traceability Protocol for
Assay and Certification of Gaseous Calibration Standards
ง2.1.6.4).
Check the regulator outlet pressure.
The pressure or calibration gas flow rate should be set as
specified in the QA test procedures in the QA/QC plan.
4.6 Flow Monitors
In addition to the daily calibration error tests (Section 4.2) and daily interference tests
(Section 4.3.1), there are a number of other flow monitor issues you can investigate.
Flow-to-Load
There is no equivalent to a linearity test for a flow CEMS. In response, EPA developed
a quarterly comparison of flow CEMS data to unit load data. Each quarter, the source must
conduct a comparison of the hourly flow and load data for any hour in which the unit load is
within ฑ 10 of the average load during the most recent RAT A. As part of your pre-audit
checks, you should note whether the source conducted this test as required and whether the test
was passed. If there are issues with the test conducted at the facility, go over those issues with
the source representative while on site.
Part 75 Field Audit Manual - My 16, 2003
Page 71
-------
On-Site CEMS Inspection Section 4
Temperature/Pressure Monitoring
Check to see if there are monitors for stack temperature and/or stack absolute pressure
associated with the flow monitor. Ask the source what type of QA/QC procedures are
followed for these monitors (what is checked and on what schedule), and if there have been any
problems with these components. You can compare to the QA/QC plan or maintenance
records.
K--factors
A k-factor refers to a correction factor or polynomial coefficient used by the DAHS to
correct the flow monitor measurement for flow variables not measured by the monitoring
method, as well as for measured changes in stack pressure and temperature. The factors are set
based on a "pre-RATA" test, in which the source correlates the monitor measurement to the
RATA flow reference method results.
Changing the k-factors or polynomial coefficients does not require recertification, but
does require that a three load RATA be performed (ง 75.20(b)) as a diagnostic test event.
You should ask when the last time the correction factors were changed and why, and check that
the RATA was performed. The changes should be recorded in the maintenance log, and a
procedure outlined in the QA/QC plan.
Displays/A larms
As in the case of gas analyzers, the displays and alarms will vary by manufacturer. You
should ask the plant contact what displays and alarms are in the model that the source uses, and
what they indicate. Information on alarms and corrective action should be recorded in the
maintenance log.
Span/Range
Ask the source what the range settings are for the flow monitor and how these settings
are displayed. Compare the settings to the monitoring plan (RT 530, Column 49). You should
also check to see if the source has conducted an annual span/range check as required by Part
75, Appendix A, ง 2.1. (See the span/range discussion in Section 4.4.)
Page 72 Part 75 Field Audit Manual - My 16, 2003
-------
Section 4
On-Site CEMS Inspection
Table 4-9:
Summary of Flow Monitor Checks
What to Check
Description
Observe and/or review daily
calibration error test and
interference test.
See Sections 4.2 and 4.3.1.
Review any issues observed in
reported results of flow-to-load
tests.
See Appendix B ง 2.2.5 for requirements of flow-to-load tests.
Make sure the test was performed, and that the results were
properly calculated. If the source is excluding hours from the
analysis as allowed for in Appendix B, check to see that the
exclusions meet Part 75 criteria.
Identify if temperature and stack
pressure are monitored and what
QA/QC procedures apply.
Temperature and pressure may be monitored in conjunction
with the flow monitor. Note the QA and preventive
maintenance procedures and schedule. Ask if there have been
any problems, and corrective actions. Check the QA/QC plan
and maintenance records.
Check if there have been any
changes to k-factors and
polynomial coefficients since the
last RATA.
Changes to the correction factors should be recorded in the
maintenance log, and the procedures for changing the correction
factor outlined in the QA/QC plan. A three load RATA is
required following any change to flow monitor k-factors
(ง 75.20(b)).
Status of the control panel lights,
indicators and alarms? Displays
should show that the analyzer is on
and operating properly.
The displays will vary by analyzer, so ask the source what the
displays are and what they meaa An alarm light could indicate
a potential or developing problem that needs to be addressed
(check the QA/QC plan).
Check range setting, and whether
the source has performed the
annual span/range check required
by Appendix A, ง 2.1.
Ask the operator what the range settings are and how they are
displayed. Compare the range setting to the value in the
monitoring plan (RT 530), and the results of the recent
span/range check.
4.7 DAHS
The data acquisition and handling system (DAHS) consists of all the hardware and
software used to comply with Part 75 electronic recordkeeping and reporting requirements. It
is a critical component of the monitoring system, as it converts the analyzer signal to reported
emissions data. The on-site audit of the DAHS focuses on data handling issues that can not be
checked electronically. These checks fall into three areas:
DAHS certification and verification of missing data routines and emission calculations
Changes in correction factors
Manually entered data
Part 75 Field Audit Manual - My 16, 2003
Page 73
-------
On-Site CEMS Inspection
Section 4
4.7.1 DAHS Certification and Verification Tests
As for other components of the plant's CEMS, make sure that the DAHS version in use
has been previously certified/recertified by checking against the monitoring plan report you
printed out from MDC.
Also you should ask to see a copy of the DAHS missing data and formula verification
test results. Part 75 requires a verification test for the missing data routines and emission
calculation formulas as part of the initial certification (ง 75.20(c)(9)), and whenever the DAHS
is replaced, upgraded to support a new EDR version, or when the missing data algorithm has
been changed (diagnostic testing under ง 75.63(a)(2)(iii)).
The source is not required to submit the results of the verification tests, and there are no
MDC or ETS electronic checks of the routines. The source is, however, required to keep these
test results on-site and available for inspection, so you should note that the tests have been
performed, and what the results were (including any vendor certification). The missing data
routines may be verified either by performing tests using a checking software, or by a
certification by the DAHS software developer that the software package meets all of the
missing data requirements of Part 75. Question 14.96 in the Parts 75 and 76 Policy Manual
provides more information on the verification testing requirements.
4.7.2 Changes in Correction Factors
Previous sections on dilution extractive systems, analyzers, and flow monitors
recommended that you investigate what types of correction factors, if any, are applied to raw
data (e.g., pressure/temperature compensation, molecular weight, flow and moisture monitoring
polynomials, sonic velocity correction factors, NOX quenching correction factors, and dilution
ratio settings). In addition you should ask how changes to these correction factors are entered
into the DAHS. As noted earlier, changes
to correction factors should be recorded in
the maintenance log, and the QA/QC plan
should outline the procedures for changing
the correction factors (Appendix B, ง
1.1.3). Diagnostic QA testing is required
for many of these changes (see ง 75.20(b)).
4.7.3 Manually Entered Data
Part 75 requires that the DAHS
automatically record all emissions data and
the daily calibration error checks (Appendix
A, ง 4). There are a few exceptions which
allow manual entry or editing of data, which
are shown in the text box. Check how
various types of data are entered manually
by asking the source to explain their
procedures, and by reviewing the QA/QC
plan and the hardcopy supporting
Manual Data Entry Allowed by Part 75
Negative (< 0) emission values
Erroneous emission values (if significant
must be approved by EPA)
SO2 concentration < 2.0 ppm
Reference method back-up data
RATA reference method data and RATA
results
Leak checks, 7-day calibration error tests,
and cycle time tests
Operating data (load and time)
Missing data periods
Add-on control equipment operation
during missing data periods
For more information on manual entry or
editing, see Section II.C.3 of the EDR v2.2
Reporting Instructions (August 2002).
Page 74
Part 75 Field Audit Manual - My 16, 2003
-------
Section 4 On-Site CEMS Inspection
information for the manually entered data. A number of areas you may emphasize are described
below.
QA Tests
Ask the source how the QA test results (daily calibration error test, linearity
test, and RAT As) are entered into the DAHS. If you have not compared a RAT A report from
the MDC program to plant hardcopy RATA results in your pre-audit review, ask for a hardcopy
of the report and check that the dates and time, relative accuracy, and bias adjustment factors
(BAF) match. Also if you identified problems with QA test calculations, go over how that data
are entered into the DAHS, as well as the applicable QA/QC plan procedures. For example, the
BAF may not have been applied correctly to the emission data. The BAF should be applied
starting with the hour following completion of the RATA.
Missing Data
Ask how the source records and reviews missing data. How do they check that the
substitute values appear correct (e.g., do the substituted values appear to be correct in view of
the percent monitor data availability (PMA) and the length of the missing data period; do the
substitute NOX and flow rate values change when the load range changes during a missing data
period; are maximum potential values substituted when the PMA drops below 80.0 percent)?
Compare the procedures described to those in the QA/QC plan. Pick out a recent missing data
period (one you may have identified in your pre-audit review), and spot check the electronic
data by comparing against supporting hardcopy documentation.
Optional Missing Data Routines - Units with Add-on Control Equipment
If the source is using an optional
missing data routine for units with add-on
control equipment, you also may need to ง 75'34' ฐPtl0na! MlS.Smง Dat*
, , , , ... Procedures for Units with Add-On
review control device parameter monitoring
records. There are four missing data options
for units with add-on controls, three of
which require parameter monitoring to
demonstrate the level of control achieved
during the missing data period (ง 75.34).
You will first need to check the
QA/QC plan. The facility must identify add-
on SO2 or NOX control equipment
parameters and acceptable ranges in the plan
if the source is using add-on control equipment missing data options (see Part 75, Appendix B,
ง 1.1.1). Ask the source for the parameter monitoring records for a number of missing data
periods. You may have identified specific missing data periods from your pre-audit preparation.
Compare the parameter data to the acceptable ranges in the QA/QC plan and identify any
periods when the parameter range is exceeded. Check how the missing data period was then
flagged in the DAHS (control operating properly or not operating properly).
Control Equipment
(1) Standard Missing Data Routines with
Parametric Supporting Data
(2) No Parameter Data
(3) Parametric Missing Data Substitution
Method
(4) Parameter Data Used to Support Use of
Maximum Controlled Emission Rate
Part 75 Field Audit Manual - My 16, 2003 Page 75
-------
On-Site CEMS Inspection
Section 4
Table 4-10:
Summary of DAHS Checks
What to Check
Description
Check that DAHS verification tests
have been performed for the missing
data routines and calculations.
Check the DAHS against the monitoring plan. Verification
tests for missing data routines and emissions calculations are
required for initial certification and if the DAHS is replaced
or there have been changes to the software. Verification test
results for the missing data routines and emissions
calculations are required to be kept on site by งง 75.20(c)(9)
and 75.63(a)(2)(iii). A vendor certification that the software
meets Part 75 requirements is sufficient for missing data
routines.
Ask the source what type of correction
factors are applied to raw data, and
how they are entered into the DAHS.
Also ask if any changes have been
made.
Changes to correction factors should be recorded in the
maintenance log, and the QA/QC plan should outline the
procedures for changing the correction factors (Appendix B,
ง 1.1.3). QA testing may also be required (see ง 75.20(b)).
Ask what data are entered manually,
and how. Spot check the electronic
data with the manual hardcopy
You want to make sure there is documentation supporting
the data that is added manually to the DAHS, and that
QA/QC plan procedures are followed. Procedures for
manually entering data should be documented in the QA/QC
plan.
How are data for the daily calibration,
linearity, and RATA tests recorded?
Review some recent records to verify
that hardcopy and electronic data
match.
Compare a RATA or linearity report from the MDC
program to the plant hardcopy. Check that the dates and
time, relative accuracy (or linearity error), and bias
adjustment factors (BAF), if applicable, match.
Review parameter monitoring records
for units using the optional missing
data procedures for add-on control
equipment.
The QA/QC plan will identify control equipment parameters
and acceptable ranges (Appendix B, ง 1.1.1 and
ง 75.58(b)). Compare the control equipment parameters to
the ranges for a missing data period.
4.8 Maintenance Log and Daily Checklists Review
The maintenance logs should detail any maintenance performed on the system and
should reference all preventive maintenance performed. Appendix B, Section 1.1.3 requires the
facility to keep the following maintenance records:
Date, time, and description of any testing, adjustment, repair, replacement, or preventive
maintenance action performed on any monitoring system;
Records of any corrective actions associated with a monitor's outage period;
Any adjustment that changes a system's ability to record and report emissions data must
be recorded (e.g., changing of flow monitor or moisture monitoring system polynomial
Page 76
Part 75 Field Audit Manual - My 16, 2003
-------
Section 4 On-Site CEMS Inspection
coefficients, K factors or mathematical algorithms, changing of temperature and
pressure coefficients and dilution ratio settings); and
A written explanation of the procedures used to make the adjustment(s).
During the inspection you should begin by asking the CEMS technicians to describe
what goes wrong with equipment, what breaks, and what maintenance is required. Ask how
they knew something was wrong, and what they did about it. Also ask what QA tests were
done to show that the problem was resolved. Make sure you understand what they are saying,
as their terminology may not always match the Part 75 rule terminology. For example, the plant
personnel may refer to a daily calibration error test as a "system cal," "overboard cal," or "span
check."
Then check the maintenance log book to see how the information is recorded. You
should point out any entries that you do not understand, and explain why you do not understand
the entry. Compare to the QA/QC plan to determine whether the source is implementing the
preventive maintenance procedures. You should also review the maintenance log to identify
recertification events and adjustments that have been made to the systems which could affect
the monitored data. The maintenance log may also provide another check on handling of
missing data periods, and it can verify that linearity tests and RATAs were performed on the
dates identified in the EDR.
Some checks may include:
Do the log entries sufficiently describe the action taken? Are the entries understandable?
Does the log show maintenance checks at the frequency identified in the QA/QC plan?
Are there recurring failures or malfunctions recorded in the log?
Are malfunctions resolved as specified in the QA/QC plan? Are calibration error tests or
other required QA tests performed before the CEMS is returned to service?
Do events in the maintenance log correspond to reported missing data periods in the
quarterly reports?
Are there repeated adjustments to the zero or span?
Have system parts or components been replaced? If so, has the proper recertification or
diagnostic testing been performed?
Are corrective actions recorded for malfunctions or as a result of daily calibration error
tests or other performance tests?
Are the name/initials of the person performing task or logging data provided? (While this
is not required, it is important for the log entries to be traceable to the person who makes
the entry so that further questions can be answered if needed.)
Part 75 Field Audit Manual - My 16, 2003 Page 77
-------
On-Site CEMS Inspection Section 4
4.9 QA/QC Plan Review
The QA/QC plan has been referenced throughout the previous sections. At a minimum,
a QA/QC plan should describe the detailed procedures and operations for: calibration error and
linearity tests; calibration and linearity adjustments; preventive maintenance and any adjustments
to the system; spare parts list; a troubleshooting matrix; and recordkeeping and reporting. For
units with add-on SO2 or NOX emission controls, the QA/QC plan may also contain (for missing
data purposes) add-on emission control parameters and the range for each parameter
representative of the normal operating conditions. Required elements in the QA/QC plan are
summarized in Table 4-11. You should note that Part 75 allows electronic storage of the
information in the QA/QC plan, provided that the information can be made available in
hardcopy upon request during an audit. The plan may also reference manufacturer's operating
manuals.
Reviewing QA/QC plan information is an important aspect of the on-site inspection
activities described throughout the previous sections. A recommended approach to inspecting
various aspects and components of a plant's CEMS QA program is to:
Ask the plant staff to explain how they operate the equipment and perform QA and
maintenance activities.
Observe and record your own observations of procedures, plant documentation, and
equipment settings.
Compare the QA/QC plan description of the equipment settings, monitoring operating
procedures, and QA test procedures to actual operations at the plant.
You should note instances where actual plant operations are different from the
operations described in the QA/QC plan, and ask for an explanation. Determine and discuss
whether the actual procedure or the QA plan procedures meet Part 75 requirements. In some
cases the actual operations may be appropriate, but the QA/QC plan may need to be updated.
Clearly, you should also note areas where the actual operations and QA/QC plan match, but
neither meet Part 75 requirements.
In addition to examining specific elements of the QA/QC plan where there are
discrepancies or potential problems you may also do a general review of the QA/QC plan to
check that the plan covers the Part 75 elements as shown in Table 4-11. Particular areas to
focus on are the QA test procedures, CEMS adjustment procedures, and preventive
maintenance procedures.
Page 78 Part 75 Field Audit Manual - My 16, 2003
-------
Section 4
On-Site CEMS Inspection
Table 4-11:
Checks for QA/QC Plan Elements
What to Check
Is there a written QA/QC plan (it may be stored electronically but
should be available), and when was it last updated?
Are calibration error test and linearity test procedures outlined in the
plan?
Are calibration and linearity test adjustment procedures outlined?
Are RATA test procedures provided?
Are emissions and QA test recordkeeping and reporting procedures,
including missing data procedures included?
If using add-on control equipment missing data options, are control
equipment parameters identified?
Are procedures for preventive maintenance and Part 75
recordkeeping/reporting identified?
Part 75 Requirement
Appendix B, ง 1
Appendix B, ง 1.2.1
Appendix B, ง 1.2.2
Appendix B, ง 1.2.3
Appendix B, ง 1.1.2
Appendix B, ง 1.1.1
Appendix B, งง 1.1.1 and 2
Part 75 Field Audit Manual - My 16, 2003
Page 79
-------
On-Site CEMS Inspection Section 4
[Thispage intentionally left blank.]
Page 80 Part 75 Field Audit Manual - My 16, 2003
-------
Section 5: CEMS Performance Test Observation
Section 5 covers how you should observe CEMS performance tests (linearity tests and
RAT As). You will generally observe a performance test in conjunction with the on-site
CEMS audit activities outlined in Section 4.
5.1 Introduction
In addition to the basic step of observing a daily calibration error test (see Section 4.2),
EPA encourages all States and Regional Offices to conduct level 2 audits that include a RATA
or linearity test observation. These performance tests are a critical element of the Part 75 Q A
program. The RATA, in particular, serves as the primary measurement of CEMS accuracy.
The presence of an agency observer can serve as an effective tool for ensuring that the source
carries out the testing properly and that the results of the tests provide a valid assessment of
data quality.
During the performance test observation you should document any deviations from the
test protocol. You should provide copies to the source to avoid misunderstandings about any
decisions you may make. The observer should notify the tester immediately if there is a
question as to whether or not any test procedures comply with Part 75 requirements, so that the
tester can take corrective action and, if necessary, restart the testing.
Once the performance test is started and until it is concluded, the only adjustments that
may be made to the monitor are routine calibration adjustments towards the calibration gas or
reference signal value that are the result of a regularly scheduled calibration error test. This is
more likely to occur during a RATA than during a linearity test.
The source can make non-routine calibration adjustments (done by physically adjusting
the instrument response using analyzer controls) before the linearity or RATA, and at other
times, provided the QA/QC plan includes specific criteria for the adjustment. A calibration
error test must be performed following the adjustment (Part 75, Appendix B, ง 2.1.3). You
should note if any adjustments were made prior to the performance test, and review plant
maintenance records for documentation of the adjustment (Part 75, Appendix B, ง 1.1.3).
Calibration error tests are often performed prior to a linearity test or RATA. As
discussed earlier with respect to monitor adjustments, you should ask the source if a calibration
error test was performed prior to the linearity or RATA to be observed, what the results were,
and if any adjustments were made. RATA and linearity tests should not be conducted on a
system that is out-of-control with respect to the calibration error requirements.
Appendix A contains observer checklists to assist you in observing the performance
tests. You should also be familiar with the Part 75 performance test requirements in Appendix
A งง 6.2 and 6.5, and the Part 60 reference methods used in the RATAs.
Part 75 Field Audit Manual - My 16, 2003 Page 81
-------
Performance Test Observation
Section 5
5.2 Linearity Test
A linearity test for a gas CEMS is conducted by challenging the CEMS with three audit
gases: one having a value of 20 to 30 percent a monitor's span, the second a value of 50 to 60
percent of span, and the third a value of 80 to 100 percent of span. Flow and moisture
monitors are not subject to the linearity test requirements, and SO2 and NOX monitors with
spans < 30 ppm also are exempt. The source must conduct a linearity test every quarter
(although exceptions for low operating hours and grace periods may apply). For a dual range
analyzer, the test is performed for each range used during the applicable quarter.
The CEM system is challenged three times with each audit gas. The same gas
concentration or level is not to be run twice in succession. The audit gas is injected at the gas
injection port required in Part 75, Appendix A, ง 2.2.1.
The response to the zero and upscale gases is not immediate, but approaches an
asymptotic value (See Illustration 5-1). It is important that sufficient time is allowed for the
concentration reading to stabilize for each injection of gas. Watch the analyzer's instantaneous
response to see if enough time is allowed to reach a stable fully equilibrated response. The time
should not exceed 15 minutes. Pay the most attention to SO2, since it is the most adsorptive
and absorptive (this will delay the equilibration time).
Illustration 5-1:
Asymptotic Calibration Check Response (Jahnke, 1994)
CEM Measurement
(ppm)
Asymptotic Value
Time
Also note the time required compared to the gas injections for the daily calibration error
tests. Both times should be about the same. If they are not ask the source why this is the case.
Record the analyzer's stabilized response and compare to the value recorded by the DAHS.
The values should be the same. The average CEMS response will be taken from the DAHS as
the official component for recording the CEMS data. These linearity test procedures and
requirements are specified in detail in Part 75, Appendix A, ง 6.2.
Page 82
Part 75 Field Audit Manual - My 16, 2003
-------
Section 5
Performance Test Observation
Time Shared Systems
Time shared systems use the same gas
analyzers for multiple emission units. For
time shared systems, the 15 minute
maximum cycle time requirement for the
analyzer gas concentration to stabilize
includes the cycle time of each probe
location served by the system.
Appendix A contains an observer
checklist. Inspect the cylinder gases as
outlined in Section 4, record the linearity
test results using the checklist, and calculate
the linearity result for each level. Compare
the results to the appropriate performance
specification listed in Table 5-1.
Table 5-1:
Linearity Test Specifications
Monitored
Parameter
Linearity Test Performance Specification
(Part 75, Appendix A, ง 3.2)
SCXorNC)
< 5.0% linearity error, or
< 5 ppm absolute value of the difference between the average of the
monitor response values and the average of the reference values
IR-AI
CO, or O2
< 5.0% linearity error for each calibration gas concentration, or
< 0.5% absolute value of the difference between the average of the
monitor response values and the average of the reference values
IR-A
5.3 Relative Accuracy Test Audit (RATA)
A Relative Accuracy Test Audit
(RATA) compares a unit's CEMS
measurements to that of reference method
stack tests. The reference method tests
yield results representative of the pollutant
concentration, emission rate, moisture,
temperature, and flue gas flow rate from the
unit. The RATA compares these results
directly to CEMS measurements. The
RATA test is the primary measurement of
CEMS accuracy for a Part 75 CEMS.
RATA Reference Materials
Reference Methods in Part 60, Appendix A,
are available at the Emission Measurement
Center website: www.epa. gov/ttn/emc.
Answers to common Part 75 RATA
questions are provided in the Parts 75 and
76 Policy Manual at
www.epa.gov/airmarkets.
Part 75 requires semi-annual or
annual RAT As, depending on the relative accuracy achieved in the preceding RATA. Most
units qualify for the annual RATA frequency (Appendix B, ง 2.3.1.2).
Gas RATAs performed at single load levels are often conducted simultaneously, and
may take about 7 hours. The flow RATAs may be conducted at one, two, or three loads. A
Part 75 Field Audit Manual - My 16, 2003
Page 83
-------
Performance Test Observation Section 5
multi-load RAT A may therefore be performed over 2 to 3 days, with each load of the flow
RATA taking about 3 to 5 hours. One load of the flow RATA testing often is conducted at the
same time as the gas RAT As. Part 75 requires that each RATA be completed within 168
consecutive unit or stack operating hours. For multi-load flow RAT As, up to 720 consecutive
unit or stack operating hours are allowed to complete the testing at all load levels (Part 75,
Appendix A, ง 6.5).
Flow RAT As are conducted at load levels in the range of operation that extends from
the "minimum safe, stable load" to the "maximum sustainable load." Three levels
apply in this range:
Low: First 30% of range Mid: > 30.0% and <60.0% High: > 60.0% of range
You should ensure that the source performs the RATA according to Part 75 and
reference method requirements. RATA issues of specific importance to Part 75 requirements
include:
Unit operating conditions. The RATA should be performed while the unit is burning a
normal fuel listed in the monitoring plan (App. A ง 6.5 (a)). Gas RAT As should be
performed at normal load. Check the load against the monitoring plan normal load. If
you have questions about the normal load identified in the plan, you may also check load
against the most recent load data submitted in the EDR to verify that the current load
designation is representative of the loads reported. Flow RAT As for peaking units or
bypass stacks can always be performed at one load. For all other situations, the RATA
for initial certification or recertification of a flow monitor must be a 3-load test. A 3-load
flow RATA also is required at least once every 5 years as part of the ongoing QA
requirements. For units that must conduct a flow RATA on an annual basis, the standard
QA flow RATA is a 2-load test. For units that must conduct a flow RATA on a semi-
annual basis, the source can alternate between a 1-load and a 2-load test. Finally, units
that operate at 1-load consistently (at least 85 percent of the time) can qualify for 1-load
testing instead of the 2-load test (see Part 75, Appendix B, ง 2.3.1.3(c)).
Check if the source conducted a daily calibration error test on the CEMS prior to the
testing or pre-RATA adjustments. While Part 75 allows a source to make pre-RATA
non-routine adjustments, adjustments may not be made between runs at a load level or
between load levels except for routine adjustments as a result of the calibration error test
(Part 75, Appendix B, ง 2.1.3).
Has the source performed and passed a stratification test? Stratification testing is
required for units wishing to use fewer traverse points under the alternatives allowed in
Part 75, Appendix A, ง 6.5.6(a) and (b). For details on stratification testing, see Part 75,
Appendix A, ง 6.5.6.1 - 6.5.6.3.
Rake probes should not be used as they do not distinctly capture the required traverse
points to ensure that a representative stack sample is obtained for analysis. See Policy
Manual Question 8.39.
Page 84 Part 75 Field Audit Manual - My 16, 2003
-------
Section 5 Performance Test Observation
Moisture measurements, used to correct dry basis measurements in determining
emission rate, may not be made using the wet bulb-dry bulb technique. Wet-bulb-dry
bulb measurements, however, may be used to determine molecular weight. See Policy
Manual Question 3.10.
EPA recently promulgated new flow reference methods (2F, 2G, and 2H). Part 75
sources had raised concerns that Method 2 measurements could be biased high in some
situations due to stratified or non-parallel flow. In response, Method 2F measures yaw
and pitch angle adjusted velocity, 2G adjusts for yaw angle, and 2H accounts for wall
effects. A detailed observer checklist is available for these methods at CAMD's website,
and you should refer to that observer checklist if you are observing a flow RATA using
one of these new methods. Also available for downloading from the website is software
(FLOW-CALC) that you can use to enter the reference method data and calculate
results. See http://www. epa. gov/airmarkets/monitoring/index.html
Check that the traverse point locations for the reference method tests meet the
requirements of Appendix A ง, 6.5.6, and the sampling location dimensions. Gas tests
should typically use at least 3 traverse points. Check Performance Specification (PS) 2 in
Part 60, Appendix B. Units with wet scrubbers may use a shorter measurement line than
required by PS 2 if minimal stratification is demonstrated, and moisture and gas systems
may use a single point if the stratification test is passed (Part 75, Appendix A, ง 6.5.6).
The minimum number of traverse points for a flow test is 12, unless also using Reference
Method 2H, which requires at least 16 points (Part 60, Appendix B). A source can use
more than the minimum number of sample points.
Check the reference method calibration gases used for instrumental test methods. The
calibration gas certificate should show: EPA Protocol gases or other certified gases;
concentrations that match those used in the bias/drift check calculations; and an
expiration date after the RATA. The regulator gauge should show a cylinder pressure
>150psi.
Verify the relative accuracy and reference method calculations yourself by doing the
calculations on site using the raw data. You should also obtain a copy or record the
results of the RATA tester's calculations. Make sure that the CEMS and reference
method data are for the same runs.
Observation forms for flow and gas RAT As are provided in Appendix A. The forms
have more detailed checks than provided here. In addition, as noted above, an observer
checklist for flow reference methods 2F, 2G, and 2H is available at CAMD's website.
Part 75 Field Audit Manual - My 16, 2003 Page 85
-------
Performance Test Observation Section 5
RATA Reference Methods (Part 75, Appendix A, ง 6.5.10)
The following methods from Appendix A to Part 60 or their approved
alternatives are the reference methods for performing relative accuracy test audits:
Method 1 or 1A for determining the appropriate test locations;
Method 2 or its allowable alternatives including in Appendix A to Part 60 (except
for Methods 2B and 2E) for stack gas velocity and volumetric flow rate;
Methods 3, 3A, or 3B for O2 or CO2;
Method 4 for moisture;
Methods 6, 6A, or 6C for SO2;
Methods 7, 7 A, 7C, 7D, or 7E for NOX, excluding the exception in section 5.1.2 of
Method 7E. When using Method 7E for measuring NOX concentration, total NOX,
both NO and NO2, must be measured. Notwithstanding these requirements, Method
20 may be used as the reference method for relative accuracy test audits of NOX
monitoring systems installed on combustion turbines.
Page 86 Part 75 Field Audit Manual - My 16, 2003
-------
Section 6: On-Site Inspection of Appendix D and Appendix E
Monitoring Systems
Section 6 focuses on the on-site review of fuel monitoring and operating records
that are required for units that use the excepted monitoring provisions available
to gas and oil fired units. The audit emphasis is on verifying recordkeeping.
Part 75, Appendix D provides an optional monitoring protocol that may be used by gas-
or oil-fired units instead of SO2 and flow CEMS. It includes procedures for measuring oil or
gaseous fuel flow using a fuel flowmeter and procedures for conducting sampling and analysis
to determine sulfur content, density, and gross calorific value (GCV) of fuel oil or gaseous
fuels.
Part 75, Appendix E is available to qualifying gas or oil fired peaking units as an
optional NOX emissions rate estimation procedure that maybe used instead of a NOX CEMS.
Baseline stack testing is performed at four operating levels to establish a NOX rate - heat input
curve with NOX rate the dependent variable.
6.1 QA/QC Plan Review
The QA/QC plan provides a template for performing a review of fuel monitoring and
sampling records, and is the place to start the field audit for the excepted monitoring provisions
of Part 75, Appendices D and E. As described in Section 4.9, you should determine whether
the QA/QC plan meets the requirements of Part 75, Appendix B, and whether the facility is
implementing the plan. Checks of the QA/QC plan and rule requirements for Part 75, Appendix
D are identified in Table 6-1, and the checks for Part 75, Appendix E are in Table 6-2.
Table 6-1:
Appendix D - QA/QC Plan Review
What to Check
Part 75 Requirements
Review the fuel sampling methods and analysis
procedures. Compare the sampling methods and
frequencies to the rule requirements outlined in
the sample Appendix D field audit sheets (see
Appendix A of this manual).
The QA/QC plan should include standard
sampling and analysis procedures used by the
source or its fuel supplier (Appendix B, ง 1.3.5).
Are the fuel flow meter test procedures and the
transducer or transmitter accuracy test
procedures outlined in the plan? Ask the source
how often (and how) the tests are performed, and
compare to the plan and Part 75 requirements.
Test procedures are required in the QA/QC plan
(Appendix B, ง 1.3.2).
(cont)
Part 75 Field Audit Manual - My 16, 2003
Page 87
-------
Appendix D and Appendix E Monitoring
Section 6
Table 6-1:
Appendix D - QA/QC Plan Review (cont.)
What to Check
Part 75 Requirements
Check the fuel flowmeter, transducer, or
transmitter calibration and maintenance records.
Records are required by the QA/QC plan
provisions in Appendix B, ง 1.3.3. Those
records are to include adjustments, maintenance,
or repairs performed on the fuel flowmeter
monitoring system as well as records of the data
and results for Appendix D fuel flowmeter
accuracy tests and transducer accuracy tests.
Table 6-2:
Appendix E - QA/QC Plan Review
What to Check
Check the recommended range of Q AQC
parameters, and hourly records of these
parameters. Make sure the parameters are
identified and recorded.
Request that the source identify any parameter
deviations, and ensure proper missing data
procedures are used for those hours.
Check written Appendix E NOX emission rate
testing procedures .
Part 75 Requirement (App. B ง 1.3.6)
The QA/QC plan must identify recommended
ranges of quality assurance- and quality control-
related operating parameters that are recorded
each hour. There are to be at least 4 parameters
for turbines or reciprocating engines, and oxygen
for boilers. (Appendix E, ง 2.3.2).
The source is required to redetermine the
Appendix E correlation if a single deviation
period exceeds 16 operating hours
This is another required component of the
QA/QC plan. The procedures should match the
test requirements in Appendix E, ง 2.1 .
6.2 DAHS and Supporting Records
Elements of the data acquisition and handling system (DAHS) for an Appendix D or
Appendix E monitoring systems should be checked in a manner similar to that described under
the CEMS on-site inspection in Section 4 of this manual.
The DAHS consists of all the hardware and software used to comply with all electronic
recordkeeping and reporting requirements. Make sure that the version used is previously
certified/recertified by checking against the monitoring plan, and ask to see and check the
DAHS verification test for the missing data routines. The source is not required to submit these
results, and there are no electronic checks of the routines. Verification test results for the
missing data routines are required by งง 75.20(c)(9) and 75.63(a)(2)(iii). The latest test should
have occurred no earlier than when the unit began using EDR v2.1 (or v2.2, when applicable).
Page 88
Part 75 Field Audit Manual - My 16, 2003
-------
Section 6
Appendix D and Appendix E Monitoring
Additional checks are necessary to ensure that data entered manually into the DAHS
match hardcopy supporting information. Fuel sampling data (sulfur content, density, GCV)
may be entered manually into the DAHS. Ask the source how these data are entered. You
should also compare the electronic fuel data in the DAHS to the on-site hardcopy fuel sampling
and analysis data to make sure they match. The source may pull the electronic data up for you
using the DAHS, or you may use data that you printed out from MDC Hourly.
You should also investigate how emissions data, missing data periods, and operating
parameters are recorded. If data are entered by hand, you should similarly spot check the
electronic data with the hardcopy. For Appendix E units, check to see that the results from the
plant's copy of the Appendix E NOX emission rate tests match the curve in the monitoring plan
and DAHS. This should include checks on the dates and times, NOX load, and fuel flow values.
You can use MDC's Appendix E test report function to review the electronic data. You should
also check the source's hardcopy capacity factor documentation against the capacity factor
submitted in the EDR (see Record Type 507).
Table 6-3:
Summary of DAHS and Supporting Records Checks
What to Check
Check that the current DAHS version is certified
and that a DAHS verification test was conducted
for the missing data routines.
How is fuel sampling data (sulfur content,
density, GCV) entered? Do the data match the
fuel analysis results?
How are emissions data, missing data periods,
and operating parameters recorded?
Does the hardcopy report of the Appendix E NOX
emission rate tests match the curve in the
monitoring plan and DAHS?
How to Check
Compare the DAHS to that identified in the
monitoring plan. Missing data routine
verification test results are required by
งง 75.20(c)(9) and75.63(a)(2)(iii). The latest
test should have occurred no earlier than when
the unit began using EDR v2. 1 (or v2.2, when
applicable).
Ask the source to pull up the fuel data in the
DAHS and compare to hardcopy fuel sampling
and analysis data.
If the data are entered by hand, spot check the
data with the hardcopy.
Compare the plant's copy of the test report to the
Appendix E test report from the MDC program.
Check that dates and times, NOX, load, and fuel
flow values match.
6.3 Appendix D Fuel Flow Monitors
If practical, visually check the fuel flow monitors to verify that the fuel flow monitors
match those in the monitoring plan. At a minimum, the monitoring plan should at least show
the flowmeter component type, manufacturer, model/version, and serial number. Some units
may also report each auxiliary component (pressure and temperature transducers and
transmitters) in the monitoring plan.
Part 75 Field Audit Manual - My 16, 2003
Page 89
-------
Appendix D and Appendix E Monitoring Section 6
6.4 Appendix D Fuel Flow Monitor Quality Assurance
Billing meters are exempt from QA accuracy testing and other quality assurance
requirements (Part 75, Appendix D, ง 2.1.4). You should first ask if the monitor is used for
commercial billing and has been designated as such. This should be identified in the monitoring
plan. If it is, no further checks are necessary.
6.4.1 QA Testing
After the initial certification tests, fuel flow monitors are to be tested once every four
fuel flowmeter QA operating quarters. Extensions up to 20 operating quarters are available
based on quarterly fuel flow-to-load test results. For these QA tests, make the following
checks:
Are quarterly fuel flow-to-load tests performed? This test is optional. Test results
are reported in the EDR -- you may have reviewed the data prior to the visit using
MDC. Verify that on-site test results match those in the electronic report.
Check fuel flow monitor accuracy test reports. Compare the hardcopy accuracy test
results against those reported electronically. MDC can provide a copy of the electronic
report. You can also verify the calculations. Ask the source for a copy of the report,
which should be in a format similar to Part 75, Appendix D, Table D-l or D-2.
6.4.2 Maintenance and Inspection Records
The maintenance logs should detail any maintenance performed on the system and
should reference all preventive maintenance performed. During the inspection, you should look
at the maintenance log book, which is required as part of the QA/QC plan. The maintenance
conducted should match the maintenance procedures in the plan. Also verify that there have
been no changes to monitoring equipment without appropriate recertification testing. See
Section 4.8 for more discussion of maintenance logs.
Page 90 Part 75 Field Audit Manual - My 16, 2003
-------
Section 7: Reporting Audit Results
Section 7 provides a brief overview of reporting results in the exit interview and
audit report. The interview should highlight issues suggesting data invalidity so
that the source can take steps to minimize allowance penalties from missing data.
CAMD should receive prompt notice of issues that may affect the validity of mass
emissions data.
7.1 Exit Interview
You should conduct an informal summary of the audit findings (exit briefing) before you
leave the facility. You should discuss the audit findings with reference to specific regulatory
requirements, or in terms of potential problems. Evidence should be available to support a
finding or observation, although you should raise any questions on a monitoring method or rule
requirement. If a source disputes a finding, you should give it an opportunity to provide
adequate alternate information. Further investigation after the meeting may be required to
achieve the resolution of questioned items. In such cases, you should table the discussion at the
briefing.
If you perform the audit periodically, you should address any progress or lack thereof
since the previous audit. This is readily done by reviewing the findings of the last audit and
noting how the issues have been resolved. If there has been no progress, the review should
emphasize the ongoing problems and stress the need for immediate resolution.
7.2 Audit Report
The audit report organizes and coordinates information gathered during the audit in a
usable manner -- it is the compilation of factual information and professional judgment resulting
from the audit. The report also serves to record the procedures used in gathering the data and
gives factual observations and evaluations from the audit. Information in the report must be
accurate, relevant, complete, objective, and clear. You should avoid discussions of general
topics, and should link all compliance issues directly to regulatory requirements. Include any
follow-up actions in the audit report.
You also should prepare a cover letter summarizing the audit results and follow-up
activities. Send the cover letter and audit report to the Designated Representative or the
source's contact person and a copy to the EPA Regional and/or State agencies. The Clean Air
Markets Division should receive a copy of the cover letter. If any findings are likely to affect
reported mass emissions used for allowance true-up activities, notify CAMD immediately. You
should complete the audit report within one month following the audit, while observations are
still fresh, and to provide a quick response to any problems. Notify the source if the report is
delayed.
Part 75 Field Audit Manual - My 16, 2003 Page 91
-------
Reporting Audit Results Section 7
7.3 Follow-up Activities
The audit team should keep documentation of any outstanding issues from the audit. A
follow-up review should be scheduled within a reasonable time after the audit.
Page 92 Part 75 Field Audit Manual - My 16, 2003
-------
Section 8: Conducting Level 3 Audits
Section 8 discusses issues that an agency should consider in developing a performance
testing program (a Level 3 audit capability). This section emphasizes single gas
challenges and linearity tests.
8.1 Overview
This section focuses on issues that an agency should consider if it chooses to develop a
Level 3 audit program that includes single gas challenges and linearity testing of gas CEMS.
The emphasis in this section is on the calibration gases and development of a testing program,
rather than on a step-by-step description of how to perform a test.
Agencies that currently do not perform Level 3 performance tests may have in-house
expertise and experience available in the agency's ambient air monitoring staff. Single gas
challenges and linearity testing of Part 75 gas CEMS require the same knowledge base as
calibration testing of ambient gas monitors. An agency can take advantage of this existing
resource in developing a Level 3 program for Part 75 audits.
8.2 Tri-Blend or Single Blend Gases
Part 75 affected sources commonly use multi-component calibration gases. Multi-
component gases provide cost savings over single component gases by reducing the number of
cylinder gases and the equipment necessary for calibration tests. These calibration gases also
allow for calibrating analyzers simultaneously. The tri-blend (or triple blend) contains SO2, NO,
and CO2. The SO2 and NO are at relatively low concentrations in the ppm range, while the CO2
concentration is in the percent range. The single component (or single blend) calibration gas
consists of the target compound SO2, NO, or CO2 blended with N2 or air. Compared to a
single blend, a tri-blend replaces a portion of the N2 or air with CO2 (up to 20 percent by
volume for high range testing).
Tri-blends with a CO2 concentration of 11 percent (representative of most stack
conditions) are recommended for agency auditors. CO2 concentrations representative of stack
gas conditions will minimize molecular weight or interference effects, described below, that can
occur if the CO2 concentration in the calibration gas is
different from the stack gas.
Molecular Weight Effects - Dilution Systems
CO2 weighs more than N2 or air, so a tri-blend
calibration gas containing CO2 will be heavier than a single
blend calibration gas in which the balance is either N2 or air.
This will impact a single gas challenge or linearity test of
dilution extractive systems if the inspector uses a single
blend to conduct a test on units that normally use tri-blends
for calibration and quality assurance. The sample flow in a
dilution extractive system is controlled by the critical orifice,
Molecular Weight of
Major Calibration Gas
Component Compounds
CO2 44
N2 28
Air 29
Part 75 Field Audit Manual - My 16, 2003 Page 93
-------
Single Gas Challenges and Linearity Tests
Section 8
which is dependent on molecular weight. Dilution extractive systems, therefore, will produce
different test results depending on whether a single blend or tri-blend gas is used. Your best
approach to dealing with this issue is to use a tri-blend with CO2 concentrations representative
of stack conditions. Table 8-1 summarizes the effects of calibration gas blend molecular
weights on dilution system emission measurements and QA testing measurements.
Table 8-1:
Effects of Gas Blends on Dilution System Measurements
(from Table 3-2 in Jahnke, 1994)
Activity Performed
Calibration Gas Blend
Used
Possible Resulting Measurement Biases
CEM system calibrated
with CO2 triple blend
Emissions
Measurements
Emission measurements bias minimized
(because CO2 present in both flue gas and
calibration gas).
CEM system calibrated
with single blend (e.g.,
SO2 in nitrogen)
Emission measurements are biased (because
CO2 is present in flue gas).
CEM system calibrated
with single blend
Calibration Error Test
and Linearity Check
Calibration error test conducted with CO2
triple blend will show a bias.
Linearity check conducted with CO2 triple
blends will show bias.
CEM system calibrated
with CO2 triple blend
Calibration error test conducted with single
blend will show a bias.
Linearity check conducted with single blends
will show bias.
CEM system calibrated
with single blend
RATA conducted with Reference Method 6C
calibrated with a CO2 triple blend will show
bias.
CEM system calibrated
with CO2 triple blend
RATA conducted with Reference Method 6C
calibrated with a single blend will show bias.
RATA
CEM system calibrated
with a single blend
RATA conducted with Reference Method 6C
calibrated with a single blend will minimize
bias.
CEM system calibrated
with CO2 triple blend
RATA conducted with Reference Method 6C
calibrated with a CO2 triple blend will
minimize bias.
Interference Effects
CO2 can interfere with chemiluminescent NOX monitors ("quenching") and cause a
negative error in the NOX emission measurement. If the auditor uses a NO single blend
calibration gas without CO2, the result reported by the DAHS will be higher than the calibration
gas concentration, as the CEMS results are corrected to account for the CO2 concentration in
Page 94
Part 75 Field Audit Manual - My 16, 2003
-------
Section 8
Single Gas Challenges and Linearity Tests
the stack gas. As in the case of molecular weight effects, using a tri-blend with CO2
concentrations representative of stack conditions will minimize this problem.
8.3 "Hands Off" Policy
EPA's "hands off approach, discussed in Section 1.4.1, also applies to conducting a
single gas challenge or linearity test. You do not need to physically handle a facility's CEMS
hardware. You should ask qualified plant personnel to perform any actions with the CEMS
equipment. This includes connecting cylinder gases to the plant calibration gas manifolds or
your gas delivery line to the CEMS calibration gas injection point, and obtaining results from
the DAHS.
8.4 Test Plan/Procedures
The agency should prepare standard operating procedures for performing the single gas
challenge or linearity test. A general outline of items to cover is provided in Table 8-2. A
sample procedure for linearity testing is provided in Appendix A of this manual.
Table 8-2:
Elements of a Standard Operating Procedure for Performance Testing
Procedure Element
Pre-Test Survey
Pre-Test Equipment Preparations
Pre-Test Meeting
Equipment Set-up
Test Procedure
Pack-up Procedure
Post-Test Meeting
Description/Purpose
Contact the source to schedule the test, verify CEMS
calibration information, and review test logistics.
Select and prepare test equipment.
Plant meeting the day of the test. Review of test
procedures and any special circumstances.
Outline of steps to set up the test equipment.
Step by step procedures for the test itself, including
calculations.
Description of equipment breakdown after the test is
completed.
Provide source with the test results.
8.5 Single Gas Challenge
The single gas challenge uses one protocol gas to challenge the CEMS at one point in
the measurement range. It has a logistical advantage over performing a linearity test. One tri-
blend gas cylinder can cover CO2, SO2, and NOX, and there is no need for a manifold or trailer.
The disadvantage is that only one point in the measurement range is tested. EPA recommends
that if a single gas challenge is performed, a mid level gas be selected (~ 40 - 60 percent of
span). Have the source perform a daily calibration error test so that the full scale can be
evaluated. This approach minimizes the resources required, but most closely approximates a
full linearity test.
Part 75 Field Audit Manual - My 16, 2003
Page 95
-------
Single Gas Challenges and Linearity Tests Section 8
When you perform the single gas challenge, connect your calibration gas cylinder to the
source's calibration gas manifold. Keep in mind the "hands off' policy. You should perform
three runs. For each run:
Inject gas, wait for the response to fully stabilize, and record value from the DAHS; and
Allow system to sample stack gas and equilibrate before the next run.
Based on this test, the percent error is calculated in the same manner as the linearity
error:
LE = -x 100
R
where:
LE = Percent linearity error
R = Reference value of calibration gas
A = Average of monitoring system responses
8.6 Linearity Test
Linearity test procedures and requirements are outlined in Part 75, Appendix A, ง 6.2.
There are two approaches your agency might take with regards to the logistics of on-site
linearity testing. The first approach is similar to the single gas challenge in that you bring your
own calibration gases to the site and have the source connect each of your three cylinder gases
(low, mid, and high) to the plant calibration gas manifold. The other approach is to provide
your own gas delivery manifold set-up with your cylinders in a truck or trailer and have the
source attach your gas delivery line to the CEMS injection point.
You will challenge the CEMS three times with each audit gas. The audit gas is injected
at the gas injection port required in Part 75, Appendix A, ง 2.2.1. The same gas concentration
or level is not used twice in succession. Sufficient time should be allowed for the concentration
reading to stabilize for each gas injection. The average CEMS response will be taken from the
DAHS as the official component for recording the CEMS data. You then calculate the percent
error using the same equation identified in Section 8.5, above.
8.7 Calibration Gases
As noted earlier, a tri-blend gas containing SO2, NOX, and 11 percent CO2 is
recommended for agency audits of SO2 and NOX monitors. The number of different calibration
gases that your agency will need is dependent on the spans of the CEMS installed on units in
your area. At least three audit gases will be needed to conduct the linearity test at a unit: one
having a value 20 to 30 percent of span, the second with a value of 50 to 60 percent of span and
the third with a value of 80 to 100 percent of span. Review the monitoring plans for the units in
your area to identify which gases you will need. Your agency should develop a standard
operating procedure for maintaining the calibration gas inventory.
Page 96 Part 75 Field Audit Manual - My 16, 2003
-------
Section 8
Single Gas Challenges and Linearity Tests
8.8 Training
Table 8-3 lists available EPA training courses pertaining to gas CEMS and analyzer
calibration. Two of the courses, APTI 435 and APTI 464, are geared to operating an ambient
monitoring system, but include instructions on how to calibrate gas analyzers that are directly
applicable to performing a single gas challenge or linearity test on a Part 75 gas CEMS.
Table 8-3:
Available EPA Training Courses
EPA Course
Number
APTI 435
APTI 464
APTI 450
APTI 474
SI 476B
Course Title
Atmospheric Sampling
Analytical Methods for Air Quality Standards
Source Sampling for Pollutants
Continuous Emission Monitoring
Continuous Emission Monitoring Systems Operation and Maintenance of Gas
Monitors
Part 75 Field Audit Manual - My 16, 2003
Page 97
------- |