United States Environmental Monitoring Systems
Environmental Protection Laboratory
Agency Research Triangle Park NC 27711
Research and Developmen EPA/600/4-77/027b June 1987
f/EPA Quality Assurance
Handbook for
Air Pollution
Measurement
Systems:
Volume III. Stationary
Sources Specific
Methods
Sections 3.0.4, 3.0.7
3.0.9, 3.0.10,
3.14, and 3.15
•Note: Sections 3.0.4, 3.0.7, 3.0.9, and 3.0.10 are technical guidelines
to support the June 4, 1987, EPA Promulgated Final Rules for
Quality Assurance Requirements for Gaseous CEMS used for
Compliance Determination (Appendix F to 40 CFR 60).
-------
June 1987
Volume III
Table of Contents
Section Pages Date
Purpose and Overview of the Quality 3 1 -04-85
Assurance Handbook
3.0 General Aspects of Quality Assurance for
Stationary Source Emissions
Testing Programs
3.0.1 Planning the Test Program 11 5-01-79
3.0.2 General Factors Involved in Stationary 2 5-01-79
Source Testing
3.0.3 Chain-of-Custody Procedure for Source 7 5-01-79
Sampling
3.0.4 Procedure for NBS-Traceable Certification 14 6-09-87
of Compressed Gas Working Standards
Used for Calibration and Audit of
Continuous Source Emission Monitoring
(Revised Traceability Protocol No. 1
3.0.5 Specific Procedures to Assess 56 9-23-85
Accuracy of Reference Methods Used for
SPNSS
3.0.6 Specific Procedures to Assess 14 9-23-85
Accuracy of Reference Methods Used for
NESHAP
X3.0.7 Calculation and Interpretation of 14 11-05-85
Accuracy for Continuous Emission
Monitoring Systems (CEMS)
3.0.8 Audit Materials Available from 7 11-04-85
U.S.E.P.A.
^3.0.9 Continuous Emission Monitoring 47 6-01-86
Systems (CEMS) Good Operating
Practices
3.0.10 Guideline for Developing Quality 11 11-26-85
Control Procedures for Gaseous
Continuous Emission Monitoring
Systems
3.1 Method 2—Determination of Stack Gas
Velocity and Volumetric Flow Rate
3. .1 Procurement of Apparatus and Supplies
3. .2 Calibration of Apparatus
3. .3 Presampling Operations
3. .4 On-Site Measurements
3. .5 Postsampling Operations
3. .6 Calculations
3. .7 Maintenance
3. .8 Auditing Procedure
3.1.9 Recommended Standards for Establishing
Traceability
3.1.10 Reference Method
3.1.11 References
3.1.12 Data Forms
15
21
7
12
3
4
1
5
1
11
2 1
8 1
-15-80
-15-80
-15-80
-15-80
-15-80
-15-80
1-15-80
1-15-80
1-15-80
-15-80
-15-80
-15-80
3.2 Method 3—Determination of Carbon
Dioxide, Oxygen Excess Air, and Dry
Molecular Weight
-------
June 1987
Section
Table of Contents (continued)
Pages
Date
15
4
6
12
2
3
1
5
1
3
1
6
-15-80
-15-80
-15-80
-15-80
-15-80
-15-80
-15-80
-15-80
-15-80
-15-80
-15-80
-15-80
3.2.1 Procurement of Apparatus and Supplies
3.2.2 Calibration of Apparatus
3.2.3 Presampling Operations
3.2.4 On-Site Measurements
3.2.5 Postsampling Operations
3.2.6 Calculations
3.2.7 Maintenance
3.2.8 Auditing Procedure
3.2.9 Recommended Standards for
Establishing Traceability
3.2.10 Reference Method
3.2.11 References
3.2.12 Data Forms
3.3 Method 4—Determination of Moisture
in Stack Gases
3.3.1 Procurement of Apparatus and Supplies
3.3.2 Calibration of Apparatus
3.3.3 Presampling Operations
3.3.4 On-Site Measurements
3.3.5 Postsampling Operations
3.3.6 Calculations
3.3.7 Maintenance
3.3.8 Auditing Procedure
3.3.9 Recommended Standards for
Establishing Traceability
3.3.10 Reference Method
3.3.11 References
3.3.12 Data Forms
3.4 Method 5—Determination of Particulate
Emissions from Stationary Sources
3.4.1 Procurement of Apparatus and Supplies
3.4.2 Calibration of Apparatus
3.4.3 Presampling Operations
3.4.4 On-Site Measurements
3.4.5 Postsampling Operations
3.4.6 Calculations
3.4.7 Maintenance
3.4.8 Auditing Procedure
3.4.9 Recommended Standards for
Establishing Traceability
3.4.10 Reference Method
3.4.11 References
3.4.12 Data Forms
3.5 Method 6—Determination of Sulfur
Dioxide Emissions from Stationary Sources
3.5.1 Procurement of Apparatus and Supplies 6 5-01-79
3.5.2 Calibration of Apparatus 6 5-01-79
3.5.3 Presampling Operations 3 5-01-79
3.5.4 On-Site Measurements 7 5-01-79
3.5.5 Postsampling Operations 7 5-01-79
3.5.6 Calculations 2 5-01-79
3.5.7 Maintenance 1 5-01-79
3.5.8 Auditing Procedure 6 - 5-01-79
9
19
7
10
4
8
3
4
1
5
1
14
-15-80
-15-80
-15-80
-15-80
-15-80
-15-80
-15-80
-15-80
-15-80
-15-80
-15-80
-15-80
15
22
20
19
15
10
3
7
1
6
2
21
1-15-80
1-15-80
1-15-80
1-15-80
1-15-80
1-15-80
1-15-80
1-15-80
1-15-80
1-15-80
1-15-80
1-15-80
-------
June 1987
Table of Contents (continued)
Section Pages Date
3.5.9 Recommended Standards for 1 5-01-79
Establishing Traceability
3.5.10 Reference Method 4 5-01-79
3.5.11 References 1 5-01-79
3.5.12 Data Forms 13 5-01-79
3.6 Method 7—-Determination of Nitrogen
Oxide Emissions from Stationary Sources
3.6.1 Procurement of Apparatus and Supplies 5 5-01-79
3.6.2 Calibration of Apparatus 5 5-01-79
3.6.3 Presampling Operations 5 5-01-79
3.6.4 On-Site Measurements 8 5-01-79
3.6.5 Postsampling Operations 5 5-01-79
3.6.6 Calculations 4 5-01-79
3.6.7 Maintenance 1 5-01-79
3.6.8 Auditing Procedure 6 9-23-85
3.6.9 Recommended Standards for 1 5-01-79
Establishing Traceability
3.6.10 Reference Method 5 5-01-79
3.6.11 References 1 5-01-79
3.6.12 Data Forms 13 5-01-79
3.7 Method 8—Determination of Sulfuric Mist
and Sulfur Dioxide Emissions from
Stationary Sources
3.7.1 Procurement of Apparatus and Supplies 7 5-01-79
3.7.2 Calibration of Apparatus 10 5-01-79
3.7.3 Presampling Operations 4 5-01-79
3.7.4 On-Site Measurements 10 5-01-79
3.7.5 Postsampling Operations 9 5-01-79
3.7.6 Calculations 6 5-01-79
3.7.7 Maintenance 2 5-01-79
3.7.8 Auditing Procedure 3 5-01-79
3.7.9 Recommended Standards for 1 5-01-79
Establishing Traceability
3.7.10 Reference Method 5 5-01-79
3.7.11 References 1 5-01-79
3.7.12 Data Forms 17 5-01-79
3.8 Method 10—Determination of Carbon
Monoxide Emissions from Stationary
Sources
3.8.1 Procurement of Apparatus and Supplies 13 1-04-82
3.8.2 Calibration of Apparatus 18 1-04-82
3.8.3 Presampling Operations 6 1-04-82
3.8.4 On-Site Measurements 12 1-04-82
3.8.5 Postsampling Operations 5 1-04-82
3.8.6 Calculations 3 1-04-82
3.8.7 Maintenance 2 1-04-82
3.8.8 Auditing Procedure 7 1-04-82
3.8.9 Recommended Standards for 7 1-04-82
Establishing Traceability
3.8.10 Reference Method 3 1-04-82
3.8.11 References 2 1-04-82
3.8.12 Data Forms 11 1-04-82
-------
June 1987
Section
Table of Contents (continued)
Pages
Date
3.9 Method 13B—Determination of Total
Fluoride Emissions from Stationary
Sources (Specific-Ion Electrode Method)
3.9.1 Procurement of Apparatus and Supplies 20 1-04-82
3.9.2 Calibration of Apparatus 25 1-04-82
3.9.3 Presampling Operations 6 1-04-82
3.9.4 On-Site Measurements 21 1-04-82
3.9.5 Postsampling Operations 19 1-04-82
3.9.6 Calculations 7 1-04-82
3.9.7 Maintenance 3 1-04-82
3.9.8 Auditing Procedure 8 1-04-82
3.9.9 Recommended Standards for 1 1-04-82
Establishing Traceability
3.9.10 Reference Method 2 1-04-82
3.9.11 References 1 1-04-82
3.9.12 Data Forms 22 1-04-82
3.10 Method 13A—Determination of Total
Fluoride Emissions from Stationary
Sources (SPADNS Zirconium Lake
Method)
3.10.1 Procurement .of Apparatus and Supplies
3.10.2 Calibration of Apparatus
3.10.3 Presampling Operations
3.10.4 On-Site Measurements
3.10.5 Postsampling Operations
3.10.6 Calculations
3.10.7 Maintenance
3.10.8 Auditing Procedure
3.10.9 Recommended Standards for
Establishing Traceability
3.10.10 Reference Method
3.10.11 References
3.10.12 Data Forms
13
5
3
3
18
7
2
1
1
5
1
6
1-04-82
1-04-82
1-04-82
1 -04-82
1-04-82
1-04-82
1-04-82
1-04-82
1-04-82
1-04-82
1-04-82
1-04-82
3.11 Method 17—Determination of Paniculate
Emissions from Stationary Sources
(In-Stack Filtration Method)
3.11.1 Procurement of Apparatus and Supplies
3.11.2 Calibration of Apparatus
3.11.3 Presampling Operations
3.11.4 On-Site Measurements
3.11.5 Postsampling Operations
3.11.6 Calculations
3.11.7 Maintenance
3.11.8 Auditing Procedure
3.11.9 Recommended Standards for
Establishing Traceability
3.11.10 Reference Method
3.11.11 References
3.11.12 Data Forms
3.12 Method 9—Visible Determination of
the Opacity Emissions from
Stationary Sources
9
2
3
6
1
1
2
2
1
11
1
1
-04-82
-04-82
-04-82
-04-82
-04-82
-04-82
-04-82
-04-82
-04-82
-04-82
-04-82
-04-82
-------
June 1987
Table of Contents (continued)
Section Pages Date
3.12.1 Certification and Training of Observers 5 4-20-83
3.12.2 Procurement of Apparatus and Supplies 2 4-20-83
3.12.3 Preobservation Operations 2 4-20-83
3.12.4 On-Site Field Observations 18 4-20-83
3.12.5 Postobservation Operations 2 4-20-83
3.12.6 Calculations 7 4-20-83
3.12.7 Auditing Procedure 2 4-20-83
3.12.8 Reference Method 5 4-20-83
3.12.9 References and Bibliography 1 4-20-83
3.12.10 Data Forms 9 4-20-83
3.13 Methods 6A and 6B—Determinations
of Sulfur Dioxide, Moisture, and Carbon
Dioxide Emissions from Fossil Fuel
Combustion Sources
3.13.1 Procurement of Apparatus and Supplies 18 7-01-86
3.13.2 Calibration of Apparatus 14 7-01-86
3.13.3 Presampling Operations 6 7-01-86
3.13.4 On-Site Measurements 25 7-01-86
3.13.5 Postsampling Operations 15 7-01-86
3.13.6 Calculations 9 7-01-86
3.13.7 Maintenance 3 7-01-86
3.13.8 Auditing Procedure 11 7-01-86
3.13.9 Recommended Standards for 1 7-01-86
Establishing Traceability
3.13.10 Reference Method 5 7-01-86
3.13.11 References 2 7-01-86
3.13.12 Data Forms 18 7-01-86
3.14 Method 7A—Determination of Nitrogen
Oxide Emissions from Stationary
Sources (Grab Sampling—Ion
Chromatographic Method)
3.14.1 Procurement of Apparatus and Supplies 10 7-01-86
3.14.2 Calibration of Apparatus 14 7-01-86
3.14.3 Presampling Operations 6 7-01-86
3.14.4 On-Site Measurements 7 7-01-86
3.14.5 Postsampling Operations 11 7-01-86
3.14.6 Calculations 6 7-01-86
3.14.7 Maintenance 2 7-01-86
3.14.8 Auditing Procedure 6 7-01-86
3.14.9 Recommended Standards for 1 7-01-86
Establishing Traceability
3.14.10 Reference Method 3 7-01-86
3.14.11 References 2 7-01-86
3.14.12 Data Forms 12 7-01-86
3.1 5 Method 7D—Determination of Nitrogen
Oxide Emissions from Stationary
Sources (Alkaline-Permanganate—Ion
Chromatographic Method)
3.15.1 Procurement of Apparatus and Supplies 18 7-01-86
3.15.2 Calibration of Apparatus 20 7-01-86
3.15.3 Presampling Operations 6 7-01-86
3.15.4 On-Site Measurements 10 7-01-86
3.15.5 Postsampling Operations 13 7-01-86
-------
June 1987
Table of Contents (continued)
Section Pages Date
3.15.6 Calculations 5 7-01-86
3.15.7 Maintenance 3 7-01-86
3.15.8 Auditing Procedure 6 7-01-86
3.15.9 Recommended Standards for 1 7-01-86
Establishing Traceability
3.15.10 Reference Method 9 7-01-86
3.15.11 References 2 7-01-86
3.15.12 Data Forms 11 7-01-86
-------
Section 3.0.4
Rev. 6/9/87
Page 1
3.0.4. PROCEDURE FOR NBS-TRACEABLE CERTIFICATION OF COMPRESSED
GAS WORKING STANDARDS USED FOR CALIBRATION AND
AUDIT OF CONTINUOUS SOURCE EMISSION MONITORS
(Revised TraceabiHty Protocol No. 1)
CONTENTS
Subsection Title
3.0.4.0 General Information
3.0.4.1 Procedure Gl; Assay and Certification of
a Compressed Gas Standard Without Dilution
3.0.4.2 References
4.0 GENERAL INFORMATION
4.0.1 Purpose and Scope of the Procedure
Section 3.0.4 describes a procedure for assaying the concentration of gaseous
pollutant concentration standards and certifying that the assay concentrations are
traceable to an authoritative reference concentration standard. This procedure is
recommended for certifying the local working concentration standards required by the
pollutant monitoring regulations of 40 CFR Part 601'2 for the calibration and audit
of continuous source emission monitors. The procedure covers certification of com-
pressed gas (cylinder) standards for CO, C02, NO, N02, and S02 (Procedure Gl).
4.0.2 Reference Standards
Part 60 of the monitoring regulations1'2 require that working standards used
for calibration and audit of continuous source emission monitors be traceable to
either a National Bureau of Standards (NBS) gaseous Standard Reference Material
(SRM) or a NBS/EPA-approved Certified Reference Material (CRM)3. Accordingly, the
reference standard used for assaying and certifying a working standard for these
purposes must be an SRM, a CRM, or a suitable intermediate standard (see the next
paragraph). SRM cylinder gas standards available from NBS are listed in Table 7.2
at the end of subsection 4.0. A current list of CRM cylinder gases and CRM vendors
is available from the Quality Assurance Division (MD-77), Environmental Monitoring
Systems Laboratory, U. S. EPA, Research Triangle Park, NC 27711.
The EPA regulations define a "traceable" standard as one which "...has been
compared and certified, either directly or via not more than one intermediate stan-
dard, to a primary standard such as a...NBS [gaseous] SRM or...CRM"4'5. Certifica-
tion of a working standard directly to an SRM or CRM primary standard is, of course,
preferred and recommended because of the lower error. However, an intermediate
reference standard is permitted, if necessary. In particular, a Gas Manufacturer's
Intermediate Standard (see subsection 4.0.2.1) that has been referenced directly to
an SRM or a CRM according to Procedure Gl 1s an acceptable Intermediate standard and
could be used as the reference standard on that basis. However, purchasers of com-
-------
Section 3.0.4
Rev. 6/9/87
Page 2
merdal gas standards referenced to an Intermediate standard such as a GMIS should
be aware that, according to the above definition, such a standard would have to be
used directly for calibration or audit. Since a second Intermediate standard Is not
permitted, such a standard could not be used as a reference standard to certify
other standards.
4.0.2.1 Gas Manufacturer's Intermediate Standard (GMIS). A GMIS 1s a compressed
(cylinder) gas standard that has been assayedwith direct reference to an SRM or
CRM and certified according to Procedure Gl, and also meets the following re-
quirements:
1. A candidate GMIS must be assayed a minimum of three (3) times, uniformly
spaced over a three (3) month period.
2. Each of the three (or more) assays must be within 1.0 percent of the mean
of the three (or more) assays.
3. The difference between the last assay and the first assay must not exceed
1.5 percent of the mean of the three (or more) assays.
4. The GMIS must be recertified every three months, and the reassay must be
within 1.5 percent of the previous certified assay. The recertified concentra-
tion of the GMIS 1s the mean of the previous certified concentration and the
reassay concentration.
4.0.2.2 Recert1f1cation of Reference Standards. Recert1f1cat1on requirements
for SRMs and CRMs are specifiedByNBSand NBS/EPA, respectively. See 4.0.2.1
for GMIS recertlflcation requirements.
4.0.3 Using the Procedure
The assay/certification procedure described here 1s carefully designed to mini-
mize both systematic and random errors 1n the assay process. Therefore, the proce-
dure should be carried out as closely as possible to the way 1t 1s described. Simi-
larly, the assay apparatus has been specifically designed to minimize errors and
should be configured as closely as possible to the design specified. Good labora-
tory practice should be observed 1n the selection of Inert materials (e.g. Teflon,
stainless steel, or glass, 1f possible) and clean, non-contaminating components for
use 1n portions of the apparatus 1n contact with the candidate or reference gas
concentrations.
4.0.4 Certification Documentation
Each assay/certification must be documented 1n a written certification report
signed by the analyst and containing at least the following Information:
1. Identification number (cylinder number).
2. Certified concentration of the standard, 1n ppm or mole percent.
3. Balance gas 1n the standard mixture.
-------
Section 3.0.4
Rev. 6/9/87
Page 3
4. Cylinder pressure at certification.
5. Date of the assay/certification.
6. Certification expiration date (see 4.0.6.3).
7. Identification of the reference standard used: SRM number, cylinder number,
and concentration for an SRM; cylinder number and concentration for a CRM or
GMIS.
8. Statement that the assay/certification was performed according to this Sec-
tion 3.0.4.
9. Identification of the laboratory where the standard was certified and the
analyst who performed the certification.
10. Identification of the gas analyzer used for the certification, Including the
make, model, serial number, the measurement principle, and the date of the last
multipoint calibration.
11. All analyzer readings used during the assay/certification and the calcula-
tions used to obtain the reported certified value.
12. Chronological record of all certifications for the standard.
Certification concentrations should be reported to 3 significant digits. Certifica-
tion documentation should be maintained for at least 3 years.
4.0.5 Certification Label
A label or tag bearing the Information described 1n Items 1 through 9 of sub-
section 4.0.4 must be attached to each certified gas cylinder.
4.0.6 Assay/Certification of Compressed Gas (Cylinder) Standards
4.0.6.1 Aging of newly-prepared gas standards. Freshly prepared gas standard
concentrations and newly filled gascylindersmust be aged before being assayed
and certified. S02 concentrations contained 1n steel cylinders must be aged at
least 15 days; other standards must be aged at least 4 days.
4.0.6.2 Stability test for reactive gas standards. Reactive gas standards,
Including nitric oxide (NO),nitrogendioxide(N02)• sulfur dioxide ($03), and
carbon monoxide (CO), that have not been previously certified must be tested for
stability as follows: Reassay the concentration at least 7 days after the first
assay and compare the two assays. If the second assay differs from the first as-
say by 1.5% or less, the cylinder may be considered stable, and the mean of the
two assays should be reported as the certified concentration. Otherwise, age the
cylinder for a week or more and repeat the test, using the second and third as-
says as 1f they were the first and second assays. Cylinders that are not stable
may not be sold and/or used for calibration or audit purposes.
-------
Section 3.0.4"
Rev. 6/9/87
Page 4
4.0.6.3 Recert1f1cation of compressed gas standards. Compressed gas standards
must be recertified according to this Section 3.0.4 within the time limits speci-
fied in Table 7.13«6«7. The reassay concentration must be within 5% of the pre-
vious certified concentration. If not, the cylinder must be retested for stabil-
ity (subsection 4.0.6.2). The certified concentration of a recertified standard
should be reported as the mean of all assays, unless a clear trend or substantial
change suggests that previous assays are no longer valid.
Table 7.1 Recertification limits for compressed gas standards.
Maximum months until
recertification for
Pollutant
Carbon monoxide
Nitric oxide
Sulfur dioxide
Nitrogen dioxide
Carbon dioxide
Oxygen
Sulfur dioxide and
carbon dioxide
Propane
Others not specifically
Balance
gas
N2 or air
N2
N2
\\2 or air
N2 or air
N2
N2
N2 or air
listed
Concentration
range
I 5 ppm
I 10 ppm
I 10 ppm
£ 10 ppm
I 300 ppm
I 2 percent
* 200 ppm S02,
* 10 percent C02
£ 5 ppm
cylinder material :
Al or SS other
18
18
18
6
18
18
18
18
6
6
6
6
6
18
18
6
6
6
4.0.6.4 Minimum cylinder pressure. No compressed gas cylinder standard should
be used when its gas pressureis~ below 700 kPa (100 psi), as indicated by the
cylinder pressure gauge.
4.0.6.5 Assay/certification of multi-component compressed gas standards. Proce-
dure Gl may be used to assay and certify individual components of multi-component
gas standards, provided that none of the components other than the component
being assayed cause a detectable response on the analyzer.
-------
Section 3.0.4
Rev. 6/9/87
Page 5
4.0.7 Analyzer Calibration
4.0.7.1 Basic analyzer calibration requirements. The assay procedure described
in this Section 3.0.4 employs a direct ratio referencing technique that inherent-
ly corrects for minor analyzer calibration variations (drift) and DOES NOT depend
on the absolute accuracy of the analyzer calibration. What is required of the
analyzer is as follows: 1) it must have a linear response to the pollutant of
interest (see subsection 4.0.7.5), 2) it must have good resolution and low noise,
3) its response calibration must be reasonably stable during the assay/certifica-
tion process, and 4) all assay concentration measurements must fall within the
calibrated response range of the analyzer.
4.0.7.2 Analyzer multipoint calibration. The gas analyzer used for the assay/
certification must have had a multipoint calibration within 3 months of Its use
when used with this procedure. This calibration is not used to quantitatively
interpret analyzer readings during the assay/certification of the candidate gas
because a more accurate, direct ratio comparison of the candidate concentration
to the reference standard concentration is used. However, this multipoint cali-
bration is necessary to establish the calibrated range of the analyzer and its
response linearity.
The multipoint calibration should consist of analyzer responses to at least
5 concentrations, including zero, approximately evenly spaced over the concentra-
tion range. Analyzer response units may be volts, millivolts, percent of scale,
or other measurable analyzer response units. The upper range limit of the cali-
brated range is determined by the highest calibration point used. If the analyz-
er has a choice of concentration ranges, the optimum range for the procedure
should be selected and calibrated. Plot the calibration points and compute the
linear regression slope and intercept. See subsection 4.0.7.5 for linearity re-
quirements and the use of a mathematical transformation, 1f needed. The Inter-
cept should be less than 1 percent of the upper concentration range limit, and
the correlation coefficient (r) should be at least 0.999.
4.0.7.3 Zero and span check and adjustment. On each day that the analyzer will
be used for assay/certification, its response calibration must be checked with a
zero and at least one span concentration near the upper concentration range Urn-
it. If necessary, the zero and span controls of the analyzer should be adjusted
so that the analyzer's response (i.e. calibration slope) 1s within about ±5
percent of the response indicated by the most recent multipoint calibration. If
a zero or span adjustment is made, allow the analyzer to stabilize for at least
an hour or more before beginning the assay procedure, since some analyzers drift
for a period of time following zero or span adjustment. If the analyzer is not
in continuous operation, turn it on and allow 1t to stabilize for at least 12
hours before the zero and span check.
4.0.7.4 Pollutant standard for multipoint calibration and zero and span adjust-
ment. The pollutantstandardorstandardsused for multipoint calibration or
zero and span checks or adjustments must be obtained from a compressed gas stan-
dard certified traceable to an NBS SRM or a NBS/EPA CRM according to Procedure Gl
of this Section 3.0.4. This standard need not be the same as the reference stan-
dard used in the assay/certification. The zero gas must meet the requirements 1n
subsection 4.0.8.
-------
Section 3.0.'
Rev. 6/9/87
Page 6
4.0.7.5 Linearity of analyzer response. The direct ratio assay technique used
In Procedure 61 requires tnat the analyzer have a linear response to concentra-
tion. Linearity 1s determined by comparing the quantitative difference between a
smoothly-drawn calibration curve based on all calibration points and a straight
line drawn between zero and an upper reference point (see Figure 1). This dif-
ference 1s measured in concentration units, parallel to the concentration axis,
from a point on the calibration curve to the corresponding point for the same
response on the straight line.
For the general linearity requirement, the straight line is drawn between
zero and the highest calibration point (Figure la). Linearity is then acceptable
when no point on the smooth calibration curve deviates from the straight line by
more than 1.5 percent of the value of the highest calibration concentration. An
alternative linearity requirement is defined on the basis of the actual reference
and candidate concentrations to be used for the assay. In this case, the refer-
ence and candidate concentrations are plotted on the calibration curve, and the
straight line is drawn from zero to the reference concentration and extrapolated,
1f necessary, beyond the candidate concentration (Figure Ib). The deviation of
the smooth calibration curve from the straight line at the candidate concentra-
tion point then must not exceed 0.8 percent of the value of the reference concen-
tration. This latter specification may allow the use of an analyzer having
greater nonlinearlty when the reference and candidate concentrations are nearly
the same.
For analyzers having an inherently non-linear response, the response can
usually be linearized with a simple mathematical transformation of the response
values, such as R'= square root(R) or R'= log(R), where R' is the transformed
response value and R 1s the actual analyzer response value. Using the trans-
formed response values, the multipoint calibration should meet one of the above
linearity requirements as well as the requirements for intercept and correlation
coefficient given 1n subsection 4.0.7.2.
4.0.8 Zero Gas
Zero gas used for dilution of any candidate or reference standard should be
clean, dry, zero-grade air or nitrogen containing a concentration of the pollutant
of Interest equivalent to less than 0.5 percent of the analyzer's upper range limit
concentration. The zero gas also should contain no contaminant that causes a de-
tectable response on the analyzer or that suppresses or enhances the analyzer's
response to the pollutant. The oxygen content of zero air should be the same as
that of ambient air.
4.0.9 Accuracy Assessment of Commercially Available Standards
Periodically, the USEPA will assess the accuracy of commercially available
compressed gas standards that have been assayed and certified according to this
Section 3.0.4. Accuracy will be assessed by EPA audit analysis of representative
actual commercial standards obtained via an anonymous agent. The accuracy audit
results, Identifying the actual gas manufacturers or vendors, will be published as
public Information.
-------
Section 3.0.4
Rev. 6/9/87
Page 7
o
CL
_>
to
c
i
Concentration difference
between calibration curve
and straight line must
not exceed 1.5% of
Cmax at any point
between 0 and
— Crnax
Concentration
points
Smooth calibration
curve based on all
calibration
points
V
i
Straight line
between zero
and highest
calibration
point
Concentration
at highest
libration
f point (Cmax)
Concentration, ppm or percent
a) General linearity requirement
a
5
ce
Concentration difference
between calibration
curve and straight line
at candidate
concentration point \.
must not exceed 0.8% J^
of the reference
concentration
Candidate
concentration
Smooth calibration
curve based on all
calibration points
Calibration
points
Straight line
between zero
and reference
concentration
Reference
concentration
<
Concentration difference
between calibration
curve and straight line
at candidate concentration
point must not v.
exceed 0.8% of the >^
reference concentration
Smooth calibration
curve based on all
calibration points
Candidate
concentration
Calibration
points
Reference
concentration
Straight line between zero and
reference concentration, extrapolated
to beyond the candidate concentration
Concentration, ppm or percent
Concentration, ppm or percent
b) Alternative linearity requirement
Figure 1. Illustration of linearity requirements.
-------
Table 7.2. NBS SRM reference gases.
Section 3.0.4
Rev. 6/9/87
Page 8
SRM
number
2627
2628
2629
1683b
1684b
1685b
1686b
1687b
2630
2631
2653
2654
2655
2656
2612a
2613a
2614a
1677C
2635
1678c
1679c
2636
1680C
1681c
2637
2638
2639
2640
2641
2642
2657
2658
2659
NBS-SRM
Type
NO/N2
NO/N2
NO/N2
NO/N2
NO/N2
NO/N2
NO/N2
NO/N2
NO/N2
NO/N2
N02/Air
N02/Air
N02/Air
N02/A1r
C0/A1r
C0/A1r
CO/Air
CO/N2
CO/N2
CO/N2
CO/N2
CO/N2
CO/N2
CO/N2
CO/N2
CO/N2
CO/N2
CO/N2
CO/N2
CO/N2
02/N2
02/N2
02/N2
cylinders contain
Nominal
concentration
5
10
20
50
100
250
500
1000
1500
3000
250
500
1000
2500
10
20
45
10
25
50
100
250
500
1000
2500
5000
1
2
4
8
2
10
21
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
percent
percent
percent
percent
percent
percent
percent
SRM
number
1693
1694
1661a
1662a
1663a
1664a
1696
1670
1671
1672
2632
2633
2634
2619a
2620a
2621a
2622a
2623a
2624a
2625a
2626a
1674b
1675b
1665b
1666b
1667b
1668b
1669b
2643
2644
2645
2646
2647
2648
2649
2650
approximately 870 liters of
Type
S02/N2
S02/N2
S02/N2
S02/N2
S02/N2
S02/N2
S02/N2
C02/Air
C02/Air
C02/Ai r
C02/N2
C02/N2
C02/N2
C02/N2
C02/N2
C02/N2
C02/N2
C02/N2
C02/N2
C02/N2
C02/N2
C02/N2
C02/N2
C3H8/Air
C3H8/Air
C3H8/Air
C3H8/Air
C3H8/Air
C3H8/N2
C3H8/N2 .
C3H8/N2
C3H8/N2
C3H8/N2
C3H8/N2
C3H8/N2
C3H8/N2
gas at STP.
Nominal
concentration
50 ppm
100 ppm
500 ppm
1000 ppm
1500 ppm
2500 ppm
3500 ppm
330 . ppm
340 ppm
350 ppm
300 ppm
400 ppm
800 ppm
0.5 percent
1.0 percent
1.5 percent
2.0 percent
2.5 percent
3.0 percent
3.5 percent
4.0 percent
7.0 percent
14.0 percent
3 ppm
10 ppm
50 ppm
100 ppm
500 ppm
100 ppm
250 ppm
500 ppm
1000 ppm
2500 ppm
5000 ppm
1 percent
2 percent
For availability, contact:
Office of Standard Reference Materials
Chemistry Building, Room B311
NBS, Gaithersburg, Maryland 20899
(301) 975-6776. (FTS 879-6776)
-------
Section 3.0.4
Rev. 6/9/87
Page Gl-1
4.1 PROCEDURE Gl; ASSAY AND CERTIFICATION OF A COMPRESSED
GAS STANDARD WITHOUT DILUTION
4.1.1 Applicability
This procedure may be used to assay the concentration of a candidate compressed
gas (cylinder) pollutant standard, based on the concentration of a compressed gas
(cylinder) reference standard of the same pollutant compound, and certify that the
assayed concentration thus established for the candidate standard 1s traceable to
the reference standard. The procedure employs a pollutant gas analyzer to compare
the candidate and reference gas concentrations by direct measurement—without dilu-
tion of either gas--to minimize assay error.
4.1.2 Limitations
1. The concentration of the candidate gas standard must be between 0.3 and 1.3
times the concentration of the reference gas standard.
2. The analyzer must have a calibrated range capable of directly measuring both
the candidate and the reference gas concentrations.
3. The analyzer's response (or transformed response) must be linear with respect
to concentration.
4. The balance gas in both the candidate and reference standards must be Identi-
cal, unless it can be shown that the analyzer is Insensitive to any difference 1n
the balance gases.
5. A source of clean, dry zero gas is required.
4.1.3 Assay Apparatus
Figure Gl illustrates the relatively simple assay apparatus. The configuration
is designed to allow convenient routing of the zero gas and undiluted samples of the
reference gas and candidate gases, in turn, to the analyzer for measurement, as
selected by three-way valves VI and V2. Pressure regulators and needle valves (V3
and V4) control the individual gas flows. The pollutant concentrations are deliv-
ered to the analyzer via a vented tee, which discharges excess flow and Insures that
the assay concentrations sampled by the analyzer are always at a fixed (atmospheric)
pressure. A small, uncalibrated rotameter monitors the vent flow to verify that the
total gas flow rate exceeds the sample flow rate demand of the analyzer so that no
room air is admitted through the vent. Valves VI and V2 could be replaced by a sin-
gle four-way valve (with 3 inputs and 1 output) or by manually moving the output
connection to each of the gases as needed. See also subsection 4.0.3.
4.1.4 Analyzer
See subsection 4.0.7.1. The pollutant gas analyzer must have a linear response
function and a calibrated range capable of measuring the full concentration of both
the candidate and the reference gas standards directly, without dilution. It must
-------
Section 3:0.4
Rev. 6/9/87
Page Gl-2
Vent
Zero Gas
(Air or N2)
Supply
1
Rotameter
To Analyzer
Three-Way Valve
Pressure
Regulators
Reference
Gas
Candidate
Gas
Figure Gl. Suggested assay apparatus for Procedure 61,
-------
Section 3.0.4
Rev. 6/9/87
Page Gl-3
have good resolution (readability), good precision, a stable response, and low out-
put signal noise. In addition, the analyzer must have good specificity for the
pollutant of Interest so that 1t has no detectable response to any contaminant that
may be contained 1n either the candidate or reference gas. If the candidate and
reference gases contain dissimilar balance gases (air versus nitrogen or different
proportions of oxygen 1n the balance air, for example), the analyzer must be proven
to be Insensitive to the two different balance gases. This may be accomplished by
showing no difference 1n analyzer response when measuring pollutant concentrations
diluted with Identical flow rates of the two balance gases.
The analyzer should be connected to a suitable, precision chart recorder or
other data acquisition device to facilitate graphical observation and documentation
of the analyzer responses obtained during the assay.
4.1.5 Analyzer Calibration
4.1.5.1 Multipoint calibration. See subsections 4.0.7.2 and 4.0.7.4.
4.1.5.2 Calibration range. The calibrated range of the analyzer must Include
both the candidate and reference gas concentrations, such that the higher concen-
tration does not exceed 97 percent of the upper range limit, and the lower con-
centration 1s not below 25 percent of the upper range limit (assuming a lower
range limit of zero). Within these limits, select a calibrated analyzer range
that will produce the highest analyzer responses.
4.1.5.3 Linearity. The direct ratio assay technique used In this procedure
requires that the analyzer have a linear response to concentration (see subsec-
tion 4.0.7.5). High-concentration-range analyzers of the type that are required
for this procedure may not be Inherently linear, but they usually have a predic-
table, non-11 near response characteristic that can be mathematically transformed
to produce a sufficiently linear response characteristic suitable for use 1n this
procedure. Any such response transformation should be verified by using 1t for
the multipoint calibration. Caution should be exercised In using a transformed
response curve because physical zero or span adjustments to the analyzer may
produce unexpected effects on the transformed characteristic.
4.1.5.4 Zero and span adjustment. See subsections 4.0.7.3 and 4.0.7.4. Prior
to carrying out the assay/certification procedure, check the calibration of the
analyzer and, 1f necessary, adjust the analyzer's zero and span controls to re-
establish the response characteristic determined at the most recent multipoint
calibration. Allow the analyzer to stabilize for an hour or more after any zero
or span adjustment. If there 1s any doubt that a transformed response character-
istic 1s still linear following a zero or span adjustment, verify linearity with
a multipoint calibration (subsection 4.0.7.2) using at least 3 known pollutant
concentrations, Including zero.
4.1.6 Assay Gases
4.1.6.1 Candidate gas standard. See subsections 4.0.6 and 4.1.2.
4.1.6.2 Reference gas standard. See subsections 4.0.2, 4.1.2, and 4.0.6.4. Se-
lect a reference standard such that the concentration of the candidate gas Is not
-------
Section 3.0.4
Rev. 6/9/87
Page Gl-4
more than 30 percent above nor less than 70 percent below the concentration of
the standard.
4.1.6.3 Zero gas. See subsection 4.0.8. The zero gas should match the balance
gas used 1n the cylinder concentrations.
4.1.7 Assay Procedure
1. Verify that the assay apparatus 1s properly configured, as described in sub-
section 4.1.3 and shown 1n Figure Gl.
2. Verify that the linearity of the analyzer has been checked within the last 3
months (see subsections 4.0.7.2, 4.0.7.5, and 4.1.4), that the zero and span are
adjusted correctly (subsection 4.0.7.3), that the candidate and reference gas
concentrations are within 25 and 97 percent of the upper range limit of the cali-
brated measurement range of the analyzer, and that the analyzer 1s operating
stably.
3. Adjust the flow rates of the three gases (reference, candidate, and zero) to
approximately the same value that will provide enough flow for the analyzer and
sufficient excess to assure that no ambient air will be drawn Into the vent.
4. Conduct a triad of measurements with the analyzer. Each triad consists of a
measurement of the zero gas concentration, a measurement of the reference gas
concentration, and a measurement of the candidate gas concentration. Use valves
VI and V2 to select each of the three concentrations for measurement. For each
measurement, allow ample time for the analyzer to achieve a stable response read--
Ing. Record the stable analyzer response for each measurement, using the same
response units (volt, millivolts, percent of scale, etc.) used for the multipoint
calibration and any transformation of the response readings necessary for linear-
ity. Do not translate the response readings to concentration values via the
calibration curve (see the footnote following Equation Gl). Do not make any
zero, span, or other physical adjustments to the analyzer during the triad of
measurements.
5. Conduct at least 2 additional measurement triads, similar to step 4 above.
However, for these subsequent triads, change the order of the three measurements
(e.g. measure reference gas, zero gas, candidate gas for the second triad and
zero gas, candidate gas, reference gas for the third triad, etc.).
6. If any one or more of the measurements of a triad is Invalid or abnormal for
any reason, discard all three measurements of the triad and repeat the triad.
7. For each triad of measurements, calculate the assay concentration of the
candidate gas as follows:
Cc = cr RC " RZ Equation Gl
-------
Section 3.0.4
Rev. 6/9/87
Page Gl-5
where: Cc = Assay concentration of the candidate gas standard, ppm or
percent;
Cr = Concentration of the reference gas standard, ppm or percent;
Rc = Stable response reading of the analyzer for the candidate
gas, analyzer response units;*
Rz = Stable response reading of the analyzer for the zero gas,
analyzer response units;*
Rr = Stable response reading of the analyzer for the reference
gas, analyzer response units.*
*Analyzer response units are the units used to express the direct response
readings of the analyzer, such as volts, millivolts, percent of scale, etc.
DO NOT convert these direct response readings to concentration units with the
multipoint calibration curve or otherwise adjust these readings except for
transformation necessary to achieve response linearity.
8. Calculate the mean of the 3 (or more) valid assays. Calculate the percent
difference of each assay from the mean. If any one of the assay values differs
from the mean by more than 1.5%, discard that assay value and conduct another
triad of measurements to obtain another assay value. When at least 3 assay val-
ues all agree within 1.5% of their mean, report the mean value as the certified
concentration of the candidate gas standard. For newly-prepared reactive stan-
dards, a reassay at least 7 days later is required to check the stability of the
standard; see subsection 4.0.6.2.
4.1.8 Stability Test for Newly-Prepared Standards
See subsections 4.0.6.1 and 4.0.6.2.
4.1.9 Certification Documentation
See subsections 4.0.4 and 4.0.5.
4.1.10 Recertification Requirements
See subsections 4.0.6.3 and 4.0.6.4.
-------
Section 3.0.4
Rev. 6/9/87
References
4.2 References.
1. Code of Federal Regulations, Title 40, Part 60, "Standards of Performance for
New Stationary Sources," Appendix A, Method 20 (1982).
2. Standards of Performance for New Stationary Sources; Quality Assurance Re-
quirements for Gaseous Continuous Emission Monitoring Systems Used for Compliance
Determination, promulgated in the Federal Register, June 4, 1987, pp. 21003-
21010.
3. "A Procedure for Establishing Traceability of Gas Mixtures to Certain Nation-
al Bureau of Standards Standard Reference Materials. EPA-600/7-81-010. Joint
publication by NBS and EPA, May 1981. Available from the U.S. Environmental
Protection Agency, Environmental Monitoring Systems Laboratory (MD-77), Research
Triangle Park, NC 27711.
4. Code of Federal Regulations, Title 40, Part 50, "National Ambient Air Quality
Measurement Methodology".
5. Code of Federal Regulations, Title 40, Part 58, "Ambient Air Quality Surveil-
lance," Appendixes A and B.
6. Shores, R. C. and F. Smith, "Stability Evaluation of Sulfur Dioxide, Nitric
Oxide, and Carbon Monoxide Gases in Cylinders. NTIS No. PB 85-122646. Available
from the National Technical Information Service, 5285 Port Royal Road, Spring-
field, VA 22161.
7. Method 6A and 6B, "Determination of Sulfur Dioxide, Moisture, and Carbon
Dioxide Emissions from Fossil Fuel Combustion Sources," Quality Assurance Hand-
book for Air Pollution Measurement Systems. Volume III, Section 3.13.8, July
1986.Available from the U.S.Environmental Protection Agency, Center for Envi-
ronmental Research Information, Cincinnati, OH 45268.
8. "List of Designated Reference and Equivalent Methods." Current edition
available from the U.S. Environmental Protection Agency, Environmental Monitoring
Systems Laboratory, Quality Assurance Division (MD-77), Research Triangle Park,
NC 27711.
-------
Section No. 3.0.7
Date November 5, 1985
Page 1
7.0 CALCULATION AND INTERPRETATION OF ACCURACY FOR CONTINUOUS
EMISSION MONITORING SYSTEM (CEMS)
This section contains a discussion on the accuracy calcula-
tions required in Appendix F and their interpretation. The
goals of Appendix F, Procedure 1, are to (1) assess CEMS accur-
acy, (2) indicate when a CEMS is out-of-control and correction is
required, and (3) specify criteria for unacceptable CEMS data.
The quarterly accuracy assessments required in Appendix F provide
a mechanism for identifying and correcting CEMS's that are
out-of-control. This results in an increase in acceptable CEMS
data. Increasing acceptable CEMS data strengthens decisions made
with regard to compliance.
The following subsections discuss the meaning, interpreta-
tion, calculation, and reporting of accuracy data.
7.1 Meaning of Accuracy
Accuracy is the measure of the closeness of a measurement to
its "true value." Although the true value is not known, it can
be approximated by the use of an appropriate standard of refer-
ence, for example, an NBS-SRM (National Bureau of Standards
Standard Reference Materials), a primary standard. Secondary
standards are also used as an approximation to "truth," although
errors may be introduced in this process.
The preferred measure of accuracy depends on the situation.
If the magnitude of the difference tends to be dependent on the
true value, T, then the percentage difference is preferable. If
it is desired to follow or observe the pattern of the differences
over time, then the signed difference or signed percentage
difference is preferable.
In the context of accuracy data based on Appendix F, three
types of audits for CEMS accuracy assessment are specified:
Relative Accuracy Test Audits (RATA), Relative Accuracy Audits
(RAA), and Cylinder Gas Audits (CGA). The procedure for the RATA
and the RAA are the same as for the Relative Accuracy Test
described in the applicable EPA performance specification (e.g.,
Performance Specification 2 for S02 and NO , and Performance
Specification 3 for 02 and CO2), with the exception that the RAA
requires three rather than nine sets of measurements, and the
accuracy is based on the average of the three sets of data. In
addition, EPA performance audit samples must be analyzed
concurrently with the RATA samples to demonstrate and document
the proficiency and accuracy of the analytical system. The same
person must conduct the RATA and the EPA audit sample analysis.
Thus, the RATA approximates "truth" by the reference method test
results, which are in turn checked for analytical accuracy by EPA
audit sample analyses. The EPA audit sample analysis must agree
-------
Section No. 3.0.7
Date November 5, 1985
Page 2
within 5 percent of the audit concentration on each of two S09
audit samples or within 10 percent of the audit concentration on
each of two NO audit samples.
In Appendix F, each GEMS must be audited at least once each
calendar quarter. Successive audits shall occur no closer than
two months apart. The audits must be conducted as follows:
1.
The RATA must be
calendar quarters.
the Performance
conducted at least once every four
The RATA is conducted as described in
Specifications in Appendix B (e.g.,
In
Performance Specification 2 for S02 and NO ).
addition, the appropriate performance audit xsamples
received from EPA are analyzed as described in the
applicable Reference Methods (e.g., Methods 6 for S00 and
7 for NO ). z
*£
If applicable, a CGA may be conducted in three of the four
calendar quarters. A CGA is conducted by challenging the
GEMS's (both pollutant and diluent monitors, if appli-
cable) with an audit gas of known concentration at two
points within the following ranges:
Audit
point
1
2
Audit range
Pollutant monitors
20 to 30% of span value
50 to 60% of span value
Diluent monitors for--
co2 o2
5 to 8% by
volume
10 to 14% by
volume
4 to 6% by
volume
8 to 12% by
volume
A separate audit gas cylinder must be used for audit
points 1 and 2. No dilution of the gas from the audit
cylinder is allowed when challenging the GEMS. Challenge
the GEMS three times at each point, and use the average
of the three responses in determining accuracy. The
monitor should be challenged at each point for a
sufficient period of time to assure absorption-
desorption of the CEMS sample transport surfaces has
stabilized. Each monitor is audited in its normal
sampling mode, i.e., pass the audit gas through all
filters, scrubbers, conditioners, and other monitor
components used during normal sampling and as much of the
sampling probe as is practical. At a minimum, the audit
gas should be introduced at the connection between the
probe and the sample line. Audit gases must be certified
-------
Section No. 3.0.7
Date November 5, 1985
Page 3
by comparison with gaseous NBS-SRM or NBS/EPA approved
CRM (Certified Reference Material) following EPA
Traceability Protocol No. 1. Procedures for preparation
of CRM's are described in Reference 2. Procedures for
preparation of EPA Traceability Protocol No. 1 gases are
described in Reference 3. The difference between the
actual concentration of the audit gas and the concen-
tration indicated by the monitor is used to assess the
accuracy of the CEMS.
3. The RAA may be conducted three of the four calendar
quarters. To conduct a RAA, follow the procedures
described in the applicable Performance Specification in
Appendix B for the Relative Accuracy Test, except that
only three sets of measurement data are required.
Analysis of EPA performance audit samples is required for
the RAA. The relative difference between the mean of the
reference method values and the mean of the CEMS values
(in terms of the standard) are used to assess the
accuracy of the CEMS.
The performance of RATA's, RAA's, and CGA's provides an
independent check of the CEMS accuracy. These independent audits
serve to document that the CEMS is providing quality data.
Examples of audit calculations are given in the subsection that
follows.
In summary, an accuracy assessment is a measure of the
deviation of a measurement obtained under standard operational
procedures from a known reference measurement. There is no
reason to expect that accuracy will remain constant over each
quarter because of changes in calibration gases, analysts, and
environment.
7.2 Example Calculations and Interpretation for Accuracy
7.2.1 Relative Accuracy Test Audit Calculations - Example data
from a RATA on a S02/02 CEMS are shown in Table 7.1.
The SO2 and 02 CEMS data shown in Table 7.1'were corrected to
a dry basis using Equation 7-1:
CEMS_ Equation 7-1
CEMS ^ = ppm, wer
ppm, dry _
ws
where
B = moisture fraction of the CEMS gas sampled.
-------
TABLE 7.1
Section No. 3.0.7
Date November 5, 1985
Page 4
RELATIVE ACCURACY TEST AUDIT DATA FOR S02 AND C>2 CEMS
Run
number
1
2
3
4
5
6
7
8
9
Avg
S02
RM
ppm
500
505
510
510
500
500
510
505
510
S02
CEMS , ,
ppm
475
480
480
480
480
500
510
505
520
°2
RM
%Q
3.0
3.0
3.0
2.9
2.9
3.0
3.0
2.9
2.9
°2
CEMS.,
3.1
3.1
3.0
2.9
3.0
3.1
3.1
3.0
3.0
S°2
RM,,
ng/J
422.4
426.6
430.8
428.4
420.0
422.4
430.8
424.2
428.4
426.0
so2
CEMS ,,
ng/Ja
403.5
407.7
405.4
403.2
405.4
424.7
433.3
426.6
439.3
413.1
so2
Diff,
ng/J
18.9
18.9
25.4
25.2
14.6
-2.3
-2.5
-2.4
-10.9
9.43
RM
, = reference method data, dry basis,
, = monitor data, dry basis.
CEMS
The S02 and 0« CEMS and RATA data in Table 7.1 were converted
to the units of tne applicable standard using Equation 7-2:
E = CF
20.9
20.9 - percent
Equation 7-2
where
Percent 0,
E = pollutant emission, ng/J (Ib/million Btu),
3
C = pollutant concentration, ng/dsm (Ib/dscf),
F = factor representing a ratio of the volume of dry
flue gas generated to the calorific value of the
fuel, dsm /J (dscf/million Btu), and
oxygen content by volume (expressed as percent),
dry basis.
Note; For the calculations shown in Table 7.1, ppm of S02 wa^
converted to ng/J using a conversion fagtor of 2.66 x 10
ng/scm/ppm and an F factor of 2.72 x 10 dsm /J.
-------
Section No. 3.0.7
Date November 5, 1985
Page 5
For complete explanation of the equations and calculations, see
40 CFR; Part 60; Appendix A; Method 19; 5. Calculation of
Particulate, Sulfur Dioxide, and Nitrogen Oxides Emission Rates.
After the data are converted to the units of the standard, the
Relative Accuracy (RA) is calculated by using the equations in
Section 8 of Performance Specification 2. For convenience in
illustrating the calculation, these equations (7-3 through 7-8)
are also shown here.
The average difference, d, is calculated for the S02 monitor
using Equation 7-3:
_ - n
d = ~ 2 (Xi " Yi) 2 di Equation 7-3
n . , n . ,
= - (84.9) = 9.43 ng/J
9
where
n = number of data points,
X. = concentration from reference method (RM, in Table
1 7.1), ng/J, a
Y. = concentration from the CEMS (CEMS, in Table 7.1),
1 ng/J, d
d. = signed difference between individual pairs, X. and
Y±, ng/J, and
Zd. = algebraic sum of the individual differences, d.,
ng/J. 1
The standard deviation S. is calculated using Equation 7-4:
Vi rn 2 i n 2!
S di ( 2 di> Equation 7-4
n"1U-i n i-i J
-------
V
'1 1 21
- 2344 - - (84.9)
8 9 I
Section No. 3.0.7
Date November 5, 1985
Page 6
= 13.9 ng/J.
The 2.5 percent error confidence coefficient, CC, is calcula-
ted using Equation 7-5:
CC = t
0.975
Equation 7-5
= 2.306
= 10.68 ng/J.
where tQ g75 = t-values in Table 7.2 for n = 9.
TABLE 7.2. VALUES OF t FOR 95 PERCENT PROBABILITY0
.Area = 0.95
\rea = 0
na
2
3
4
5
6
.025^
to.
12.
4.
3.
2.
2.
"^.^
-rtfirrffflll
^0.975
975
706
303
182
776
571
*^"^
^"""^x. .^Area = 0.025
ItmT
0
a
n t0.975
7 2.447
8 2.365
9 2.306
10 2.262
11 2.228
t0.975
na
12
13
14
15
16
*0
2
2
2
2
2
.975
.201
.179
.160
.145
.131
The values in this table are already corrected for
n-1 degrees of freedom. Use n equal to the number of
individual values.
-------
Section No. 3.0.7
Date November 5, 1985
Page 7
The RA for the RATA is calculated using Equation 7-6:
RA =
Iccl
x 100
Equation 7-6
RM
|9.43|
. 68|
x 100 = 4.72%
426
where
RA = relative accuracy, %,
ld|= absolute value of the mean differences .from Equa-
tion 7-3, ng/J,
|CC|= absolute value of the confidence coefficient from
Equation 7-5, ng/J, and
KM = average reference method value or applicable stan-
dard, ng/J.
7.2.2 Relative Accuracy Audit Calculations - Example data from
an RAA on an S02/O2 CEMS are shown in Table 7.3.
TABLE 7.3 RELATIVE ACCURACY AUDIT DATA FOR S02 AND C>2 CEMS
Run
number
1
2
3
Avg
so2
ppm
500
505
510
so2
CEMS,,
ppm
475
480
480
°2
>
3.0
3.0
3.0
°2.
CEMSd,
3.1
3.1
3.0
so2
RM
ng/J
422.4
426.6
430.8
426.6
so2
CEMS ,,
ng/JQ
403.5
407.7
405.4
405.5
RM, = reference method data, dry basis.
CEMS, = monitor data, dry basis.
-------
Section No. 3.0.7
Date November 5, 1985
Page 8
The S02 and 02 OEMS data shown in Table 7.3 were corrected to
a dry basil using Equation 7-1. The SO2 and 02 GEMS and RAA data
were converted to the units of the applicable standard using
Equation 7-2.
The accuracy (A) for the RAA is calculated using Equation 7-7.
A = Cm " Ca x 100
Equation 7-7
405.5 - 426.6
426.6
x 100 = - 4.95%
where
A = accuracy of the CEMS, %,
Cm '
average CEMS response during audit in units
of applicable standard, and
average audit value of the three reference
method runs in units of the applicable standard.
7.2.3 Cylinder Gas Audit Calculations - Example data from a CGA
on an SO2/O2 CEMS are shown in Table 7.4.
TABLE 7.4 CYLINDER GAS AUDIT DATA FOR S02 AND C>2 CEMS
Audit
number
1
2
Reading
No.
1
2
3
Avg
1
2
3
Avg
S02
CGA
ppm
212
212
208
210.7
398
399
403
400.0
S02 A
CEMS,, Diff,
ppm %
218
219
225
220.7 4.75
409
416
414
413 3.25
°2
°2 A
CGAd, CEMSd, Diff,
5.0
5.0
5.1
5.03
9.1
9.1
8.9
9.03
5.2
5.3
5.2
5.23 3.98
8.9
8.9
8.9
8.90 -1.44
-------
Section No. 3.0.7
Date November 5, 1985
Page 9
CGA, = cylinder gas audit value, dry basis.
CEMS, = average of the three monitor values, dry basis.
The S00 and Q~ CEMS data shown in Table 7.4 were corrected to a
dry basis using Equation 7-1. The accuracy (A) for the GCA is
calculated using Equation 7-8.
c - C
A = m a x 100 Equation 7-8
Ca
= 220.7 - 210.7 x 1QO = 4>75%
210.7
where
A = accuracy of the CEMS component, %,
C = CEMS component mean response for three values
during audit with CGA in units of the appropriate
concentration, and
C = audit value of the cylinder gas in units
of appropriate concentration.
-------
Section No. 3.0.7
Date November 5, 1985
Page 10
7.3 Reporting Requirements
At the reporting interval specified in the applicable regu-
lation, a report of each CEMS accuracy audit must be submitted in
the form of a Data Accuracy Report (DAR). One copy of the DAR
must be included for each quarterly audit along with the report
of emissions required under the applicable regulation. As a
minimum, the DAR must contain the following information:
1. Source owner or operator name and address.
2. Identification and location of monitors in the CEMS.
3. Manufacturer and model number of each monitor in the
CEMS.
\
4. Assessment of CEMS data accuracy and date of assessment
as determined by a RATA, RAA, or CGA, including the RA
for the RATA, the A for the RAA or CGA, the reference
method results, certified values for the cylinder
gases, the CEMS responses, and the CEMS accuracy
calculation results. If the accuracy audit results
show the CEMS to be out-of-control, the CEMS operator
shall report both the audit results showing the CEMS to
be out-of-control and the results of the audit
following corrective action showing the CEMS to be
operating within specifications.
5. Results from the EPA performance audit samples.
6. Summary of all corrective actions taken when the moni-
tor was determined out-of-control.
An example of a DAR form is shown in Figure 7.1.
-------
Section No. 3.0.7
Date September 23, 1985
Page 11
Period ending date Year
Company name
Plant name Source unit no.
CEMS manufacturer Model no.
CEMS serial no. , CEMS type (e.g., in situ)
CEMS sampling location (e.g., control device outlet)
CEMS span values as per the applicable regulation, S0_ ppm
0_ percent, NO ppm, CO- percent
^ ' X " £
I. Accuracy assessment results (Complete A, B, or C below for each CEMS
or for each pollutant and diluent analyzer, as applicable.) If the
quarterly audit results show the CEMS to be out-of-control, report the
results of both the quarterly audit and the audit following the
corrective action showing the CEMS to be operating properly.
A. Relative accuracy test audit (RATA) for
(e.g., S02 in ng/J).
1. Date of Audit
2. Reference methods (RM's) used (e.g., Methods 3 and 6).
3. Average RM value (e.g., ng/J, mg/dsm , or percent
volume).
4. Average CEMS value .
5. Absolute value of the mean difference IdI
6. Confidence coefficient |CC|
7. Percent relative accuracy (RA) percent.
8. EPA performance audit results:
a. Audit lot number (1) (2)
b. Audit sample number (1) (2)
c. Results (mg/dsm-5) (1) (2)
d. Actual value (mg/dsm3)* (1) (2)
e. Relative error* (1) (2)
*To be completed by the Agency.
Figure 7.1 Example format for data assessment report (DAR).
-------
B. Cylinder gas audit (CGA) for
1. Date of audit
2. Cylinder ID number
3. Date of certification
4. Type of certification
5. Certified audit value
6. CEMS response value
7. Accuracy
Audit
point 1
Section No. 3.0.7
Date September 23, 1985
Page 12
(e.g., SCL in ppm).
Audit
point 2
(e.g., EPA Protocol 1
or CRM).
(e.g., ppm).
(e.g., ppm).
percent.
C. Relative accuracy audit (RAA) for
1. Date of audit
2. Reference methods (RM's) used
3. Average RM value
4. Average CEMS value
5. Accuracy
6. EPA performance audit results:
a. Audit lot number (1)
b. Audit sample number (1)~
c. Results (mg/dsnr)* _ (1)
d. Actual value (mg/dsnr)* (1)
e. Relative error* (1)
(e.g., S02 in ng/J).
(e.g.. Methods 3 and 6)
(e.g., ng/J).
percent.
(2)
(2)"
(2)"
(2)
(2)
*To be completed by the Agency.
Figure f.l (continued)
-------
Section No. 3.0.7
Date September 23, 1985
Page 13
D. Corrective action for excessive inaccuracy.
1. Out-of-control periods.
a. Date(s)
b. Number of days
2. Corrective action taken
3. Results of audit following corrective action. (Use format of
A, B, or C above, as applicable.
II. Calibration drift assessment.
A. Out-of-control periods.
1. Date(s)
2. Number of days
B. Corrective action taken
Figure 7-1 (continued)
-------
Section No. 3.0.7
Date September 23, 1985
Page 14
7.4 References
I. Standards of Performance for New Stationary Sources:
40 CFR 60, Appendix F - Quality Assurance Procedures,
Procedure 1 - Quality Assurance Requirements for Gaseous
Continuous Emission Monitoring Systems Used for Compliance
Determination.
2. A Procedure for Establishing Traceability of Gas Mixtures
to Certain National Bureau of Standards Standard Reference
Materials. Joint publication by NBS and EPA,
EPA-600/7-81-010. Available from the U. S. Environmental
Protection Agency, Quality Assurance Division (MD-77),
Research Triangle Park, North Carolina 27711.
3. Traceability Protocol for Establishing True Concentrations
of Gases Used for Calibration and Audits of Continuous
Source Emission Monitors (Protocol Number 1). June 1978,
Section 3.0.4 of the Quality Assurance Handbook for Air
Pollution Measurement Systems, Volume III, Stationary
Source Specific Methods. EPA-600/4-77-027b. August
1977. U. S. Environmental Protection Agency, Office of
Research and Development Publications, 26 West St. Clair
Street, Cincinnati, Ohio 45268.
-------
Section No. 3.0.9
Date 6/1/86
Page 1
9.0 CONTINUOUS EMISSION MONITORING (CEM) SYSTEMS GOOD OPERATING
PRACTICES
Continuous emission monitoring (CEM) systems are required to be
installed in facilities specified by the EPA Standards of Performance
for New Stationary Sources (SPNSS) and by other Federal and state
regulations. The systems are used to continuously monitor the effec-
tiveness of air pollution control techniques and to determine if
source compliance standards are being met.
This section of Volume III is intended to provide guidance for
technical personnel in air pollution control agencies and in industry
who are responsible for CEM programs. Guidelines are given to aid
agency personnel in evaluating operation and quality assurance
practices associated with permanently installed CEM systems. The
guidelines may also be useful to operators of CEM systems in
developing quality assurance and quality control procedures that meet
agency minimum requirements. Section 3.0.9 does not address the use
of continuous monitors in mobile testing vans or as portable
compliance monitors. However, much of the information presented here
is relevant to these applications.
CEM systems have been developed to monitor pollutant gases, such
as SO2 and NO, and the so-called diluent gases, C02 and 02, present
in the exhaust gas streams of combustion sources. Systems have also
been developed to monitor flue-gas opacity. A system is defined as
the total equipment required for the determination of flue-gas
opacity, a gas concentration, or the emission rate. A system is
normally composed of a sample interface, the pollutant and diluent
analyzers, and the data recording subsystem. The system is used to
generate emission data that are representative of the total emissions
from the facility.
The sample interface is the portion of the monitoring system that
protects the analyzer from the effects of the effluent. In
extractive systems, the interface consists of the probe assembly,
sampling lines, and conditioning subsystems. The sample is normally
taken from a single point in the stack or duct and then transported
to the analyzer. A conditioning system is often used to remove
particulate matter from the sample and to dry the sample before it
enters the analyzer.
In-situ monitors have been developed to measure the stack gas
concentrations, without transporting the gas itself. Gas measure-
ments are made either at a point or along a path of known length
within the flue. For in-situ path monitoring, the interface may
consist of optical windows and blower assemblies used to keep the
-------
Section No. 3.0.9
Date 6/1/86
Page 2
windows clean. For point in-situ designs, it may consist of ceramic
thimbles and support housings. These different approaches taken
toward the measurement of effluent gases will be discussed later in
this section.
The SPNSS require data obtained from a CEM system to be
representative, accurate, and precise. In contrast to EPA
certification procedures for ambient air monitors, source emission
monitors are not categorically approved by model or manufacturer.
Instead, installed systems are approved on a case-by-case basis
through the procedures established in the Performance Specifications
for Continuous Emission Monitoring Systems in Stationary Sources
(40CFR60 Appendix B).i After an installed monitoring system is found
to meet these specifications, it is expected that it will be properly
maintained at the same or better level of performance.
The proper operation and maintenance of a CEM system is
imperative if the data are to be used for regulatory purposes. The
responsibility for the system lies with the owner, and in general,
ownership lies with the plant or industrial facility. The generation
of valid data from a CEM system through proper operation and
maintenance procedures must therefore come from plant personnel or
through services contracted by the plant. It is, however, the plant
personnel or their contractors who will actually operate and maintain
these systems.
9.1 CEM Operation/Maintenance Programs - Levels of Quality Control
A maintenance program for a CEM system should be part of a
larger, plant instrumentation quality assurance (QA) program.
Quality control practices within the QA program are those activities
performed to assure that accurate and precise data are generated from
the monitoring system. Daily operation checks, preventive mainte-
nance routines, and audits are quality control activities that can be
used for this purpose.
There are four levels of quality > control that should be estab-
lished for a CEM system:
Level 1. Operation Checks (daily checks, observations, and
adjustments)
Level 2. Routine Maintenance (periodic preventive maintenance)
Level 3. Performance Audits
Level 4. Corrective Maintenance
Operation checks are performed on a routine basis, generally
daily, to see that the equipment is operating properly. These
-------
Section No. 3.0.9
Date 6/1/86
Page 3
procedures will include daily zero and calibration checks, checks of
reference signals from control panels, and checks of flow rates,
pressures, vacuum levels.
Routine maintenance is performed at regular intervals.
Activities include the replacement of filters, lamps, motor bearings,
or other parts. Detailed service checks of electronic and optical
systems may also take place at this time to uncover incipient prob-
lems in the instrumentation. Depending on the system, the replace-
ment and check intervals may vary from 30 days to a year or more.
Performance audits, which provide a check of the system's
operation, identify problems, identify the need to improve preventive
maintenance procedures, or alert the operator to the need for
corrective maintenance.
Corrective maintenance is performed to bring the monitoring
system into operation after a breakdown in the system occurs. It is
also termed nonroutine maintenance, the unscheduled need to repair a
faulty system.
9.2 Gas GEM Systems - Operation Practices
The day-to-day operation of a GEM system is not difficult once
the instruments are turned on and operating properly; generally all
that needs to be done is to periodically check the zero and the span
of the instruments in the system. This check may be conducted either
manually or automatically by using calibration gases or optical
filters. However, routine and corrective maintenance practices vary,
depending upon the methods of analysis and the overall design of the
system. For this reason, it is important to understand the special
demands of different monitoring systems. Extractive systems have
different maintenance requirements than in-situ systems. Within the
categories of extractive systems, or in-situ systems, the different
types of analyzers will require servicing dependent on the principle
by which they analyze the pollutant.
To help understand maintenance requirements, this section will
present an overview of the various analysis principles used in the
commercial systems.2 Table 9.1 summarizes these principles.
9.2.1 Extractive Monitoring Systems - This section will first
discuss the design of extractive systems and then the various ana-
lytical techniques used to measure the gas concentration.
9.2.1.1 Extractive System Design. A complete extractive system
consists of a sample probe and conditioning system, analyzer, and
-------
Section No. 3.0.9
Date 6/1/86
Page 4
TABLE 9.1. PRINCIPLES OP DETECTION USED IN
CONTINUOUS EMISSION MONITORS
Extractive System
In-Situ Systems
Gaseous Emission
Monitors
Gaseous Emission
Monitors
Opacity
Monitors
Absorption Spectroscopy
Nondispersive Infrared
Differential Absorption
Luminescence Methods
Fluorescence
Chemiluminescence
Flame Photometry
Electroanalytical Methods
Polarography
Electrocatalysis
Paramagnetic Methods
Absorption Spectroscopy
Nondispersive Infrared
(Gas filter-correlation)
Differential Absorption
Second Derivative
Spectroscopy
Electroanalytical Methods
Electrocatalysis
Visible
Light Scattering
and Absorption
data recording system. Sampling probes and conditioning systems are
today commonly purchased from the analyzer vendor rather than
assembled from miscellaneous parts by plant technicians. Also, a
number of companies specialize in marketing hybrid systems (complete
extractive systems composed of components supplied by different
vendors).
There are two approaches taken in extractive system design. One
is to condition the gas near the analyzer; the other, to condition
the gas as close as possible to the stack or duct. In the first
approach, a probe is inserted into the flue gas and the gas is drawn
through a coarse particulate filter into a heated sampling line. The
sample line may extend to over 60 m to a control room or
environmental enclosure, where the gas is conditioned. The condi-
tioning system cools the gas and removes water vapor by some type of
refrigeration, dilution, or permeation device. Usually a fine filter
is placed just before the analyzer to prevent small particles from
entering the analyzer. Diaphragm pumps, rotary vane pumps, or air
aspirators are used to transport the sample from the probe to the
analyzer.
-------
Section No. 3.0.9
Date 6/1/86
Page 5
In the second approach of sample extraction, gas is conditioned
at the stack or duct. Filters, chillers, or dilution systems are
located at the sampling site, and in the case of some dilution
designs, the probe itself does the conditioning. This approach
allows a low moisture sample to be transported to the analyzer. Long
sections of heat-traced or insulated lines may therefore be avoided.
Calibration gases are used in both approaches to check the
performance of the system. The gases are injected as close to the
probe as is technically feasible. Also, blow-back devices are often
installed to clean the coarse particulate filters. As the system
operates, these filters may eventually plug up. A burst of high
pressure air "blown back" through the filter reduces plugging and
provides for continued operation.
Extractive systems are normally constructed from components that
are familiar to plant mechanics. Valves, filters, tubing, tube
fittings, solenoids, etc., are commonly encountered. These compo-
nents must be maintained if the system is to provide continuous data.
9.2.1.2 Extractive Analyzers - Spectroscopic Absorption Techniques.
Two basic absorption spectroscopic techniques are utilized in
commercially available extractive analyzers: (1) non-dispersive
infrared spectroscopy and (2) differential absorption spectroscopy.
Nondispersive infrared spectroscopy utilizes infrared light in a
limited range of the electromagnetic spectrum. The light is not
scanned or "dispersed" as with scanning laboratory spectrometers. In
general, the light is filtered to select light wavelengths that will
be absorbed by the molecules that are to be measured. The light
passes through a gas cell that contains the flue gas extracted from
the stack. A portion of the light from the lamp passes through a
cell containing a reference gas that does not absorb the filtered
light. A detector senses the amount of light absorption in the
sample cell relative to the signal from the reference cell. Through
proper calibration, the detector responses are electronically con-
verted to pollutant concentration readings. A variant of this tech-
nique, called gas filter correlation spectroscopy, uses a reference
cell that absorbs 100% of the light in the molecular absorption
region of the pollutant.
Infrared analyzers have been developed to measure gases such as
S02/ NO, NO2f HC1, C02, and CO. The commercially available monitors
differ primarily in the design of the detector and the level of
rejection of interfering gases.
-------
Section No. 3.0.9
Date 6/1/86
Page 6
Differential absorption spectroscopy also takes advantage of the
characteristics of molecules to absorb light of certain wavelengths.
Instead of using a sample cell and reference cell as do the common
infrared systems, differential absorption spectrometers use a measur-
ing "wavelength" corresponding to a region of the spectrum where the
molecule absorbs light energy. The reference wavelength corresponds
to a region where there is little or no absorption.
Most extractive differential absorption systems operate in the
ultraviolet (UV) region of the spectrum, although it is possible to
use the technique in the infrared region. The gases may be measured
hot in the UV without removing water vapor, although it is generally
advisable to dry the sample.
S02 is commonly measured using this technique. A technique has
also been devised to measure flue gas NO by injecting oxygen into the
sample chamber, sealing it, and monitoring the production of NO2 from
NO at an N02 absorption wavelength.
9.2.1.3 Extractive Analyzers - Luminescence Techniques.
Luminescence is the emission of light from a molecule or atom that
has been excited in some manner. Three luminescence techniques are
used in the field of source monitoring: (1) fluorescence, (2) chemi-
luminescence, and (3) flame photometry.
Ultraviolet fluorescence is used to measure SO2- Ultraviolet
light in the region of 210 nm is used to excite an S02 molecule. The
molecular excited state persists for a few nanoseconds, during which
time some of the energy is lost in vibrational transitions. The
molecule eventually returns to its unexcited state with the release
of light at a longer wavelength (near 350 urn). This light is then
detected by a photomultiplier tube, resulting in a measurement of the
SO2 concentration in the sample gas.
Fluorescence monitors can be affected by changes in the flue-gas
composition (%O2, %C02). This is caused by the de-excitation of
excited SO2 molecules through the process of quenching. For this
reason, fluorescence analyzers are most successful in flue-gas analy-
sis when they are coupled with a dilution system, thereby providing a
relatively constant background composition.
Chemiluminescence is used in flue-gas analysis to measure NO and
NO2 concentrations. In this application of chemiluminescence,
excited NO2 molecules are produced by reacting ozone with the flue
gas NO. The excited NO2 product (NO2*) de-excites to its ground
state with the release of light energy. The light is measured with a
photomultiplier tube. Quenching effects also occur in this method,
-------
Section No. 3.0.9
Date 6/1/86
Page 7
but dilution of the sample through the introduction of the reactant
ozone gas stream minimizes the effect.
Since the light is produced only through the reaction of ozone
with NO, NO2 must first be reduced to NO before it can be measured.
A catalytic reaction chamber is used when a measurement of both NO
and N02 (NOx) must be obtained.
Flame photometry can be used to measure compounds that contain
sulfur. In this technique, the compounds are "burned" in a hydrogen
flame, leading to the formation of excited diatomic sulfur molecules,
82*. The conversion of the high energy 82* molecules to the lower
energy ground state, 82, occurs with the emission of light. The
intensity of this light is measured and related to the concentration
of sulfur species in the sample. The flame photometric method does
not discriminate between different sulfur-containing compounds, so
scrubbers or gas chromatographic columns may be required if more than
one species is present in the sample.
9.2.1.4 Extractive Analyzers - Electroanalytical Techniques. Two
principal electroanalytical techniques have been developed for the
measurement of flue gases. These are polarography and electro-
catalysis. A clean, dry sample must be supplied to an analyzer
ope.rating by the polarographic method. The electrocatalytic
technique can, however, be applied to both extractive and in-situ
measurement methods.
Polaroqraphic analyzers are, basically, diffusion-controlled
electrochemical cells. The cells are constructed much like
batteries, with a sensing electrode, electrolyte, and counter-
electrode. The main difference is the addition of a thin-film
membrane, through which the pollutant must diffuse to initiate the
electrochemical reactions and current flow. The current across the
cell is proportional to the rate of diffusion of the pollutant into
the cell and is also proportional to the pollutant concentration.
Polarographic analyzers have been developed to measure gases such
as S02, NO, 02, and C02. Different choices of electrodes and elec-
trolytes are made for each gas. As with batteries, the electrolyte
will eventually be consumed, and the cell will need to be replaced or
recharged.
Electrocatalytic analyzers have been developed for the
measurement of ©2 and S02. This technique uses a solid electrolyte
instead of liquid electrolytes generally associated with
electrochemical cells. A platinum film, coated on the solid surface,
-------
Section No. 3.0.9
Date 6/1/86
Page 8
catalyzes a reaction that allows molecules to migrate through the
solid and generate a measurable flow of electrons.
In oxygen electrocatalytic analyzers, a zirconium oxide disc;
coated with a thin film of platinum, is heated to 850°C. A reference
gas of about 21% oxygen is maintained on one side of the solid, and
the sample gas is on the other side. Oxygen ions are generated at
the platinum surface and then migrate through vacancies in the
heated, solid electrolyte. Electrons are released in the process as
the system attempts to equalize the oxygen concentration.
An electrocatalytic analyzer has also been developed for the
measurement of SO2. This system uses a potassium sulfate crystal and
requires the simultaneous measurement of the sample oxygen
concentration.
9.2.1.5 Extractive Analyzers - Paramagnetic Techniques. Oxygen
exhibits paramagnetic behavior by being attracted to a magnetic
field. This behavior has been utilized in the design of several
different types of extractive flue-gas analyzers. For example, in
thermomaqnetic oxygen analyzers, a magnet causes 02 to flow through a
tube and cool a resistor. The resistance is then related to 02 con-
centration. In maqnetodynamic systems, 02 disturbs a magnetic field
around a torsion pendulum, and in paramagnetic pressure analyzers, a
magnetic field causes a pressure imbalance that can be measured.
9.2.2 Recommended Maintenance - Extractive Monitoring Systems
9.2.2.1 Operation Checks (Daily Checks). Operation checks of an
extractive monitoring system should be performed each day by a
qualified and trained instrument operator. The operator should be
familiar with the system and be able to recognize a problem from
discrepancies found during the operation check procedure. Many
extractive monitoring systems are designed to automatically perform
daily zero and calibration checks and internal self checks without
operator intervention. Unfortunately, this can reduce the level of
operator attention to the system. Small problems, as a result, may
go undetected and very quickly lead to large problems. On the other
hand, "intelligent" systems, which monitor key system parameters and
report out-of-control conditions at remote panels, have helped to
alleviate such situations.
A daily operation check of an extractive monitoring system should
start with a check of the strip chart record and/or other data
recording devices. The operator should mark the exact time on the
chart for calibration purposes, and write down the date, his or her
name, and the chart recorder settings. This should all be written
-------
Section No. 3.0.9
Date 6/1/86
Page 9
directly on the chart. The paper in the recorder and printer (if
applicable) should be checked to see if the supply is sufficient for
the next 24-h run.
Indicator lights on the system or monitor control panel should be
checked next. It is advised that a record of the system status be
recorded in ink in a hardbound logbook. All maintenance, unscheduled
repairs, or system modifications should be described in the logbook.
This book will serve as an invaluable tool in tracking the long-term
performance of the system and will enable other technicians or
servicemen to become familiar with the system.
The system indicator lights will notify the operator of out-of-
range conditions occurring in the system or of other problems felt to
be important by the system designer. If a problem occurs, it should
be attended to immediately since subsequent data will be otherwise
suspect. Some systems contain reset buttons, installed to override
the indicator lights. These should not be used until after the
problem is resolved. The values of other systems indicators, such as
vacuum or pressure gauges, sample flow rates, and lamp and detector
reference levels (if applicable), should also be recorded at this
time.
A calibration check should be performed next.3 This involves
injecting a zero (or low-level) gas and a high-level gas (calibration
gas) into the sample line. It is recommended that the gas be
injected at a point where as many of the conditioning system
components as possible can be checked. In some systems, this can be
done at the probe itself, providing the advantage of checking the
system for sample line losses.
Gas injected from gas cylinders may pressurize the system.
Consequently, the flow rate of the cylinder gas into the analyzer
sample cell may differ from the flow rate of the extracted stack gas.
Also, cylinder gas is dry gas; it does not contain moisture. These
two factors may adversely affect the calibration process if flow
rates and moisture content of gases entering the analyzer sample cell
are not similar. It is also important to note that if the system is
pressurized, leaks in the system may not be detected. Ambient air,
will not enter the system as it otherwise might if a vacuum is used
to draw a sample into the analyzer when leaks are present.
Gases used for the daily checks should first be validated against
certified calibration gases or be certified themselves. Certified
gases should have their concentration established through EPA
-------
Section No. 3.0.9
Date 6/1/86
Page 10
Traceability Protocol No. 1.3 Using gases of uncertain concentration
can result in gross miscalibration of the system.
The monitor should first be checked with the zero (or low-level)
gas, and the instrument reading should be noted in the logbook. The
high-level gas should next be injected into the system and the
reading likewise noted. The readings should also be recorded appro-
priately on the strip chart record. The differences noted between
the cylinder gas value and the monitor readings are used to assess
the low-level and high-level calibration drift.
The instrument operator may not have to "rezero" and
"recalibrate" the system every 24 h when the values are checked.
Small values for drift may be due merely to system noise. It is
recommended as a minimum that the system be adjusted when the drift
exceeds twice the limits of the drift performance specification.4
For example, if the performance specification, is 2.5% for an
instrument span value of 1000 ppm, adjustments should be made when
drift exceeds 50 ppm. For systems with lower span values, the drift
tolerance will be accordingly less. (The span value is given in the
Code of Federal Regulations for source categories affected by
continuous monitoring regulations. The span value is defined as "The
upper limit of a gas concentration measurement range that is
specified for affected source categories in the applicable subpart of
the regulation." i)
The operator should record values from the instrument meter, the
strip chart, and digital printer. If a microprocessor controller is
used to check and/or adjust monitor data automatically to the
appropriate values, it must be programmed to record the unadjusted
values first. If a strip chart recorder is used in conjunction with
the microprocessor, the system should be programmed so that the
adjusted values will appear on both the strip chart and the printer
output. This may be difficult since the microprocessor adjustment is
often only done numerically by the program, i.e., the analyzer itself
is not physically adjusted. In such a case, the meter readings and
strip chart readings may differ significantly from the microprocessor
output. Data interpretation in such cases may become difficult.
To assist in performing the daily operation checks, a data sheet
has been provided in Figure 9.1. The figure is meant to serve as a
guide for the inspector or operator in developing a data sheet
applicable to a specific system.
9.2.2.2 Routine Maintenance (30-day Checks). Routine maintenance
should be initially performed on an extractive system at least every
30 days. With experience/ this time period can be either increased
-------
Section No. 3.0.9
Date 6/1/86
Page 11
Example Format for Extractive Gas Monitoring System
Daily QC Check Sheet
Plant f flJvCl/Tder /^ m^ fi^UW Date S/.ao/5?6 Time f .' 3-
Analvzer I.D. MgQ<;i;itf rrh +*4-7tj
Span Value IOQO ooivi
Calibration Gas Value c]Li'S
Zero Gas Value (air, N2, oth
Hours Ooeratinq in Period
er) a\C
Part 1 Indicators
Indicator Lights/Gauges
Sample pressure/vacuum
Sample flow
Lamp
Detector
Name Joe. JonCS
Phone 919 - 5HT- 0*6,3
Zero Offset value ^>O ^>prvi
Date Certified q/i<-f/S3
n
Paper Status: Strip Char
Printer 34
Status
IO psi.
0.3SJL/^
OK
OK
Part 2 Calibration Check
Unadjusted Readings
Zero (low-level) gas
Calibration (high-level) gas
Stack Concentration
Time
q ; 50 a,/^
10: co
10'. 10
Part 3 Zero and Span Adjustment (if
Control limit i 50
Adjusted Readings
Zero (low-level) gas
Calibration (high-level) gas
Stack Concentration
t (5K
hrs -rp(i'«««te»
Problem/Action taken
Louj/ (Boost "fct> 1 2- ps i_
^
-
-
Meter
!(.
Mff
^^8
Strip Chart
If*
BUS
5HS
Digital Printer
at
-------
Section No. 3.0.9
Date 6/1/86
Page 12
or decreased, and depending on the system, maintenance intervals of
varying frequency may be established for subsystems or for individual
components.
In extractive monitoring systems, most maintenance lies with the
sample conditioning systems rather than with the analyzers.
Particulate matter and water vapor are usually removed from the gas
stream before the gas reaches the analyzer. The filters that remove
the particulate matter must be periodically cleaned or replaced.
Condensed water in condensing-type moisture removal systems must be
drained.
The plumbing associated with extractive systems is prone to
corrosion and leaks; therefore, the system should be periodically
checked for leaks. Fittings, valves, and gas regulators should also
be checked. Solenoid valves have been commonly used to automate
extractive systems. These valves are prone to failure and should be
checked frequently to ensure they move freely and on command. The
use of motorized or air-activated rotary valves instead of solenoid
valves may also help to reduce the frequency of valve failures. Care
should be taken to avoid over-design or over-automation of a system.
The more valves there are, the more valves there are to check. Spare
valves should be kept in the parts inventory.
Electrical cables and heat-traced lines should also be checked
frequently. In a plant environment, damage can occur from
construction projects or through normal plant operations. The
ambient atmosphere, particularly near flue-gas leakage or stack down-
wash areas, may cause electrical insulation to deteriorate rapidly.
Acid gases circulating near the stack may corrode both electrical
fittings and the plumbing of the extractive system.
The pumps and chillers used in extractive systems work 24 h/day.
At some time, the motor brushes will wear out, a pump diaphragm will
break, or a part will require oiling. Rather than treating such
events as problems, they should be anticipated by establishing a
regular schedule of overhaul and maintenance.
The overall cleanliness of an extractive system is also
important. The particulate matter in the flue-gas can migrate into
unexpected places. If a system is located outside, near the stack,
sensitive components should be installed in dust-free cabinets. The
system should be cleaned if fly ash settles on it, and in no case
should cabinets be opened when fly ash is circulating in the ambient
air.
-------
Section No. 3.0.9
Date 6/1/86
Page 13
The extractive analyzers contain components which have limited
lifetimes. Lamps and bulbs generally have given performance periods.
They should be replaced before this period is up, since a weakened
bulb can often produce spurious results. Analyzers are often
designed to operate best over a given range of lamp intensity. When
the intensity drops too low, the detector will not be able to respond
as accurately as it should to the incoming signal. Detectors may
also have to be replaced, but this is not common.
Many instruments have test points on the back panel or on circuit
boards. These test points are checked with a voltmeter or an
oscilloscope to indicate certain limiting values. The instrument
operator should perform these electronic tests routinely to check for
electronic integrity.
Figure 9.2 gives an example of a checklist designed for a routine
maintenance procedure. Again, this list is suggested to help the
inspector or operator design his or her own list. It is not uncommon
for instrument operators to spend a year or more in designing a
system maintenance schedule. Vendor instruction manuals are often
lacking in this regard, so points of maintenance may have to be
determined through experience. The system logbook is an invaluable
tool in developing such a schedule.
9.2.2.3 Performance Audits. Performance audits should be conducted
on extractive monitoring systems at appropriate intervals. EPA
40CFR60 Appendix F - Quality Assurance Procedure I5 requires that an
audit be performed at least once every quarter for monitoring systems
used for determining compliance with emission standards. This
frequency is recommended to identify CEM systems that may be
generating biased results.
The performance audit is essentially an independent check of the
system, and can vary, depending upon the resources of the owner of
the system. For CEM systems installed to demonstrate compliance with
emissions standards, EPA requires an audit at least once each
quarter, using one of the followingS:
• Relative Accuracy Test Audit (RATA)
A repeat of the relative accuracy test procedures as
defined in Appendix B Performance Specifications.1
• Cylinder Gas Audit (CGA)
Challenging the monitoring system with cylinder gas of
known concentration (certified gases).
-------
Section No. 3.0.9
Date 6/1/86
Page 14
Example Format for Extractive Gas Monitoring System
30-Day Maintenance Check Sheet
Plant
Unit
~B&\\f!r 1*3. - OutU»±-
System I.D.
Analyzer I.D.
Analyzer I.D.
Analyzer I.D.
4*3^
Date_
Name
Phone
tf3.7 /?5
Time
- 543 - 8k 3
Gas_
Gas.
Gas
Required Haintenance Checks
Extractive System
Probe filter
Fine filter
Condensation system drain
Heat trace continuity
Pump - bearing noise
Plumbing leak check - vacuum
- pressure
Cable integrity
Cleanliness
Corrosion levels - probe
Solenoid performance
Regulator pressures
Air-operated valves
Air purge/blouback l
Zero gas
Calibration gas
Status
o£^^xSWete*"fc
&ra.u
JDrruV^ p^pert^
L\'*\e. (juarm -to-toycK
S(-v=,k-t ITLSpl "TO
NA
"Pressurised, 3o psi.
3^t£taOUSe. te^W-t. CQUJ&P&
pcSHcl-e aett'ii^ o»-v ccUsiei
Ply OA'n OLCCumuloM^ f><\.
Wo rust
Set£S^i IS^
(pCOp^x.
1 1 CO (XSL.
S&C p9L
(pao p»L
Action
)VlOK>e.
'R*^SJ"U*c^tf "eXt
^TXn^odL i^iCL'^OCLLU.A
^LOCXit^iA^ C^QX'L, i
Mo^e
C^ec^t- ^-o^t. 36 £^.(Xys
OleD«_*^ ^d
C^ecL*\fid. — c*Tdc»"tcX.
p^OttfC^t ^€. GO^G^
—
T^fA^t-d
Mo*^€
Ncv^C,
Wone
NJfcnt
Analyzers
Lamp
Sensor
Test points
Chopper motor
Optical window status
U)o noise
C\«o^
-
j\ct±. to l.foO o-i^
-
-
Jo
Operator Signature
Date
Supervisor Signature
D
ate
Figure 9.2. Example format for extractive gas monitoring
system 30-day maintenance check sheet.
-------
Section No. 3.0.9
Date 6/1/86
Page 15
• Relative Accuracy Audit (RAA)
An audit similar to the RATA except that only three sets
of measurement data (instead of nine) are taken.
Appendix F requires that at least one of the quarterly audits be
a RATA, and either the CGA or RAA can be used for the other three
quarters. If the relative accuracy between the audit and analyzer
values exceeds 20% for the RATA, the instrument is viewed as being
out of control. If the relative accuracy between the audit and
analyzer values exceeds 15% for the CGA or RAA, the instrument is
viewed as being out of control. A RATA, CGA, or RAA must be
conducted after repairs are made to out-of-control systems.
Additional techniques can also be used during an audit. For
example, portable gas monitoring instruments can be used to check the
stack gas concentrations rapidly. Although the portable monitor
itself may not meet the performance specifications that the CEM
system meets, it can give valuable information during an audit.
The main idea behind the performance audit is to provide an
independent assessment of the monitoring system accuracy. Daily
calibration drift determinations and routine maintenance do not
necessarily guarantee that data will be accurate. An independent
assessment using an appropriate auditing technique can, however,
provide an indication of data validity. Figure 9.3 gives the
Appendix F example format for an audit "Data Assessment Report."
9.2.2.4 Corrective Maintenance (Problems and Troubleshooting).
Maintenance problems with extractive monitoring systems usually occur
in the gas transport and gas conditioning components. Valves,
fittings, tubing, and filters in the presence of acid gases,
submicron particulate matter, and continuous vibration are likely to
have limited life unless they are routinely maintained. Lack of
routine maintenance or lack of foresight will result in the need for
corrective maintenance. The need for corrective maintenance can be
avoided by establishing good quality assurance and quality control
programs.
The extractive system gas analyzers normally will have few
problems unless they are located in a severe environment or if the
gas conditioning system fails. If the conditioning system fails,
acid gases can condense in the sample cell and particulate matter can
settle in the system to plug the probe or sample lines, or the
analyzer itself.
-------
Section No. 3.0.9
Date 6/1/86
Page 16
Example Format for Data Assessment Reports
Period Ending Date 3 3o
Year I9SS
Company Name Acme. fkwie.f
Plant Hame
Source Unit Ho.
Model Mo.
CEM System Manufacturer M'e£LSUf&-hsr>\
CEM System Serial Ho. A 8106.3 CEM System Type (e.g., in-situ) £x.4racH-w g.
CEM System Sampling Location (e.g., control device outlet) &SP Out-let-
CEM System Span Values, as per the applicable regulation, SO2
O2 percent, HOx ppm, C02
loon
ppm,
percent
I. Accuracy Assessment Results. Complete A, B, or C below for each CEN system or
for each pollutant and diluenc analyzer, as applicable.) If the quarterly audit
results show the CEM System to be out of control, report the results of both the
quarterly audit and the audit following the corrective action showing the CEM
System to be operating properly.
A. Relative accuracy test audit (RATA) for
1. Date of Audit
2
3
4
5
(e.g., S02 in ng/J)
Reference methods (RMs) used 3onA(f (e.g.. Methods 3 and 6)
Average RM value -434-.(a (e.g., ng/J, mg/dsm, or percent volume).
Average CEM value H5 I.a ng/J .
Absolute value of mean difference Idl l(c,.T3 •
Confidence coefficient ICCI 3.~l• 3D .
7. Percent relative accuracy (RA) IO. I 3
8. EPA performance audit results:
a. Audit lot number (1)
b. Audit sample number (1)
c. Results (mg/dsm3) (1)
d. Actual value (mg/dsm3)* (1)
e. Relative error* (1)
B. Cylinder gas audit (CGA) for
1. Date of Audit
percent.
-C.1*
(2)
(2)
(2)
(2)
(2)
. 3
Je.gr., SC<2 in ppm)
Audit point 1
2. Cylinder ID number
3. Date of certification
4. Type of certification
5. Certified audit value
6. CEMS response value
7. Accuracy
* To be completed by the Agency
Audit point 2
L. 3.01
t 5
(e.g., EPA protocol 1 or
CRM).
(e.g., ppm).
(e.g., ppm).
percent.
(continued)
Figure 9.3. Example format for data assessment report.
-------
Section No. 3.0.9
Date 6/1/86
Page 17
C. Relative accuracy audit (RAA) for SOja. in noi jJ~ _ (e.g., SO2 in ng/J)
1. Date of audit
2. Reference methods (RMs) used 3aifdla (e.g. . Methods 3 and 6)
3. Average RM value 32.%. 4 (e.g., ng/J).
4. Average CEM value A^3 . 8 _ .
5. Accuracy _ 3(5 % _ percent.
6. EPA performance audit results:
a. Audit lot number (1) ^fe/6" (2)
b. Audit sample number (1) 'lOXZ (2)
c. Results (mg/dsn>3) (1) jt. fc (2)
d. Actual value (mg/dsm3)« (1) 12£. & (2)
e. Relative error* (1) ±(*.°( 4 (21
D. Corrective action for excessive inaccuracy.
1. Out-of-control periods.
a. Date(s) 1/»iJtS- 7/3 1 /Sf
b. Number of days _ |? _ .
2. Corrective action taken 1$-ep\aLB\ [CLrtf> 77 3 1 /&**»"
Adjusted, resisters <^I3 asrt. R\K on
recaJULbrnXtd.
3. Results of audit following corrective action. (Use format of A, B, or C,
above, as applicable. )
II. Calibration Drift Assessnent.
A. Out-of-control periods.
a. Date(s) 7/fo /fS - 1 /3A /Sf 1/31- 1/5\/S5'
b. Number of days 4 _ •
B. Corrective action taken 'ftg-ge.rQg^ gyrf rgQX^t^rztS^. On
-stlLI drif^ because of-
77?
Operator Signature Date Supervisor Signature ' D
ate
To be completed by the Agency
Figure 9.3. (continued).
-------
Section No. 3.0.9
Date 6/1/86
Page 18
Most problems inherent to extractive analyzers will be electronic
in nature. Procedures recommended in the maintenance manuals can be
used to troubleshoot circuit boards and components. Extractive
analyzers are often compact enough so that in the case of severe
electronic problems, they can be returned to the instrument vendor
for repair.
The system operator will often be warned of problems by a loss of
signal, inconsistent readings, or poor calibration response.
Approaches to resolving problems largely depend on the skill of the
operator; however, some general guidelines can be given:
• Loss of signal or abnormally low values - check
conditioning system for plugging, leaks, pump failure.
• Noisy, erratic signals - check for electronic problems,
electrical supply problems, weak lamps, moisture
condensation, particulate matter in analyzer.
• Loss of linearity - check for sample cell contamination,
leaking calibration manifold, incorrect gas cylinder
values.
• Slow response - check for leaks, water in lines, measuring
cell failures.
Failure of the conditioning system and consequent drawing of
unconditioned sample gas into the analyzer is one of the worst
situations that can occur with an extractive monitoring system. Once
unconditioned gas enters the system, it may take months before the
system will again operate properly. Therefore, redundancy and fail-
safe devices should be built into the system from the start.
Table 9.2 lists some common problems that occur with extractive
monitoring systems. Those listed range from the physical problems
often associated with the conditioning system to those associated
with the analyzers. Many of these problems are due to poor system
design resulting from a failure on the part of the system
manufacturer to understand the constraints imposed by the plant
environment and stack gas conditions. These problems also may be the
result of inadequate specifications provided by the user at the time
of purchase. It is difficult to foresee problems and it may often
require a redesign of the system before the frequency of corrective
maintenance is minimized.
-------
Section No. 3.0.9
Date 6/1/86
Page 19
TABLE 9.2. EXTRACTIVE SYSTEM PROBLEMS
Common Physical Problems
Possible Corrective Actions
A. Conditioning Systems
Probe plugging
Probe/filter corrosion
Probe breakage (due to vibration or
embrittlement from chloride)
Condensation in sample lines
Inadequate water removal
Dirt in sample lines, plugged valves,
plugged sample lines
Leaks in sample lines/fittings
Pump failure
Install blowback system, increase blow-
back frequency probe shield. Relocate.
Change probe design. Change system
design. Enter probe at downward angle.
Relocate probe. Obtain corrosion-
resistant alloy for probe construction.
Support probe.
alloy.
Shorten. Select resistant
Resize heaters. Don't let heat go off on
heat trace. Use backup power. Avoid
shorts - don't loop lines. Install
thermal conductivity sensor if
continuing problem. Remove water at
stack probe. Filter at lower
temperature (acid may be condensing) -
increase temperature or heat.
Improve chiller design. Connect two
chillers in series. Back up chiller
with Permapure dryer (but heat front end
of Permapure). Dilute the gas stream to
lower moisture content.
Decrease pore size of probe filter. In-
crease sample flow rate to fine filter.
Increase diameter of line. Use clear
Teflon tubing to detect areas of
accumulation. Redesign to reduce number
of valves. Use redundant filters.
Reduce number of fittings as much as
possible. Detect leaks by pressurizing
system and using soap bubble indicator.
Check for leaks in gas regulators.
Don't wrench down on compression
fittings too severely. Don't use glue,
paint, glyptal, etc., to cover leaks -
rebuild system if necessary.
Perform routine maintenance - check brushes
periodically. Check diaphragms of
diaphragm pumps.
(continued)
-------
Section No. 3.0.9
Date 6/1/86
Page 20
TABLE 9.2. (continued)
Common Physical Problems
Possible Corrective Actions
B. Analyzers
Internal corrosion/damage
Poor response time
(False positive zero values or
poor calibration check values)
Excessive drift
Component failures
Lamps, fan, chopper motors
Electronic problems
Loose circuit boards, poor
contacts
Ground loops and noise
Large voltage drops when plant
equipment is started. Spikes
in strip chart record.
Static electrical charges
Burned out electronic circuits
from lightning strikes
Check moisture removal system for failure.
Build redundancy in system. Add extra
chiller. Put thermal conductivity
sensor in line to stop pump when
moisture breaks through. When moisture
breaks through, dismantle sample cell,
clean, and dry. May have to replace
entire cell in some models. Clean and
dry all sample lines.
Check sample line length. Shorten line or
increase flow rate. Some analyzers have
slow response times. Increase time for
calibration gas flow during daily
checks.
Check fouling of sample cell for dirt or
water. Electrical problems. Passiva-
tion of cell surfaces. Lamp weakening -
light levels too low. Detector prob-
lems. Electronic problems. Erratic
power supply.
Check component wear
regular schedule.
Check and replace on
Check for vibration problems. Install
circuit board clamps. Check for SO2
corrosion in exposed units
Trace and rewire.
Install transient suppressor, dedicated
power transformer or constant
voltage/isolation transformer for
monitoring system.
Connect probe case to dedicated earth
ground.
Add phenolic gaskets between metal stack
and probe. Add surge arrestors at
junction box.
(continued)
-------
Section No. 3.0.9
Date 6/1/86
Page 21
TABLE 9.2. (continued)
Common Physical Problems
Possible Corrective Actions
B. Analyzers
Electronic problems (cont'd)
No output from instrument, no
calibration cycle, etc.
Improper instrument response -
faulty calibration, improper or
no output
Check fuses,
Check electronics. Check to see that cards
and components are secure. Use trouble-
shooting guide supplied by vendor to
check electronic test points. Replace
appropriate components or replace cards.
Check software for errors in programming
- particularly in calibration adjustment
routines.
9.2.3 In-situ Gas Monitoring Systems
9.2.3.1 In-situ System Design. In-situ gas analyzers measure
pollutant and diluent gases as they exist in the stack or the flue.
There are two basic types of in-situ monitoring systems: path and
point.
Path monitors generally consist of two units, placed opposite
each other across a duct or stack. The systems use electro-optical
techniques in which either infrared or ultraviolet light is beamed
through the flue gas. Absorption of light energy at specific wave-
lengths provides a quantitative measurement of different molecular
species. Such instruments can be designed to pass the light either
once or twice through the gas. The once-through systems are known as
single-pass monitors and the twice-through systems as double-pass
monitors.
In single-pass monitors, the stack units consist of a transmitter
and a receiver. The transmitter contains an infrared or UV lamp that
beams light to the receiver unit directly across from it on the other-
side of the stack. The receiver unit senses the transmitted light
energy and analyzes it to provide an indication of the gas concentra-
tion. The transmitter and receiver units are protected from the flue
gases by windows, over which a curtain of air is blown. The air
prevents particulate matter from soiling the windows, cools the parts
of the unit mounted on the stack, and prevents the condensation of
corrosive materials on the cooler instrument windows. The purge air
-------
Section No. 3.0.9
Date 6/1/86
Page 22
is generally provided by blowers that draw in filtered ambient air.
The use of plant air for this purpose is not advised.
Double-pass in-situ path monitors consist of a transceiver unit
and a retroreflector unit. Light is transmitted and received by the
transceiver. The transmitted light is bounced off the retroreflector
located on the opposite side of the flue and returns to be analyzed.
The light passes over a path twice the length of that of a single-
pass unit/ and in so doing, may be considerably weakened upon
returning to the transceiver unit. For this reason, most commercial
path in-situ gas monitors are single-pass designs rather than double-
pass. Double-pass in-situ monitors also use purge air systems to
protect window interfaces.
Path in-situ monitor systems come with a number of accessories
that are needed to protect them from the often hostile environment
encountered at the installation site. For stack-mounted systems,
protective hoods and covers are necessary to protect the transmitter/
transceiver,receiver/reflector units from rain, wind, and temperature
fluctuations. Lightning protection is frequently necessary. Also,
anti-vibration systems are often required to prevent the optical and
electrical components from shaking loose. A constant voltage trans-
former dedicated to the monitor is also frequently necessary to avoid
large plant voltage transients from affecting the sensitive elec-
tronics of the transceiver units.
Frequently, pipes are used to maintain the alignment between the
two cross-stack instrument units. A modification of the support pipe
provides a system in which an added outer pipe can rotate to shut out
flue gas from entering the light path. Ambient air can be used to
purge out any remaining flue gas in the closed tubes so that a zero
reference reading can be obtained by the monitor. Calibration gas
cells are generally slipped in the light path at this time to obtain
an upscale calibration reading.
Point monitors measure the flue gas over a small distance
relative to the larger duct or stack diameter. This distance may be
less than 5 cm or, in some cases, on the order of 1/3 to 1 m in
length. The length of the probes available for these systems are
generally fixed, projecting the measurement section of the probe from
1 m to 2 or 3m into the stack.
If the pollutant gases are greatly stratified, the fixed, one-
point measurement may not adequately represent pollutant emissions.
The same problem, of course, exists for the single-point probes of
extractive systems.
-------
Section No. 3.0.9
Date 6/1/86
Page 23
In-situ point monitors normally use a ceramic thimble to protect
the measuring cavity from particulate matter. Pollutant or diluent
gases diffuse through the thimble, which excludes the particles.
Because thimbles may become plugged or glazed over in wet scrubber
applications, they must periodically be checked or replaced. Point
monitors that use ceramic thimbles do not require blowers as do the
path monitors, but still may require protective hoods, lightning
protection, and constant voltage transformers.
Another type of in-situ point monitor utilizes an open probe to
support a retroreflector. Much like a double-pass path monitor,
light from a transceiver hits the retroreflector and returns to be
measured. The measurement path, however, will be on the order of
only a meter or less, classifying the instrument as an in-situ point
monitor, since it measures only a relatively small distance in the
flue gas. In this case, purge air blowers are required to keep the
retroreflector and transceiver windows clean.
In-situ monitors provide an alternative to extractive systems
since they can avoid the requirements for gas conditioning systems.
There are trade-offs, however, since the in-situ analyzers installed
directly on the stack may experience severe environmental conditions.
Table 9.3 summarizes some of the advantages and disadvantages of in-
situ and extractive systems.
9.2.3.2 In-situ Analyzers - Absorption Spectrometers and
Electrocatalytic Systems. The basic principles of operation of in-
situ systems are similar to those of the extractive analyzers.
Absorption spectroscopy and electrocatalytic methods are the two most
common techniques employed.
The absorption spectroscopic techniques used in in-situ monitors
are
• differential absorption
• gas filter correlation
• second derivative spectroscopy.
The most common differential absorption systems are single-pass
path monitors that use a diffraction grating to distinguish between
measuring and reference wavelengths in the UV region of the spectrum.
SO2 and NO are measured by this technique, although by changing the
optical system, it is possible to measure other gases. Filters can
also be used to distinguish between measuring and reference
wavelengths in in-situ analyzers. CO2 is measured by infrared light
with this method.
-------
Section No. 3.0.9
Date 6/1/86
Page 24
TABLE 9.3. COMPARISON OP EXTRACTIVE AND IN-SITD SYSTEMS
Extractive Systems
In-Situ Systems
Path
Point
Advantages
May be zeroed and
calibrated with cylinder
gases
Analyzers can easily be
located in controlled
environments
Can time share analyzer
Maintenance may not
require special training
Linearly averages stack
concentrations to possibly
reduce stratification
effects
Does not alter sample
Has fewer separate compo-
nents than extractive
systems
Rapid response
Does not alter sample
May be zeroed and calibrated
with cylinder gases
Has fewer separate compo-
nents than extractive
systems
Disadvantages
System (other than
analyzer) may require
frequent maintenance
Probe plugging possible
in dirty gas streams
May alter sample
Long sample lines reduce
response time
Zero and calibration
gases expensive
Can monitor at only one
location
Can monitor at only one
location
Difficult to repair on site Difficult to repair on site
May lose light levels in
dirty gas streams or in wet
scrubber applications
Requires temperature
compensation
Generally cannot be cali-
brated with cylinder gases.
Audits are expensive: must
perform reference method
testing
Special training or vendor
service often needed
May clog in dirty gas
streams or in wet scrubber
applications
Requires temperature compen-
sation
Special training or vendor
service often needed
-------
Section No. 3.0.9
Date 6/1/86
Page 25
Gas filter correlation spectroscopy is a nondispersive
spectroscopic technique that has been used in single-pass in-situ
analyzers as well as in ambient air analyzers. The method requires a
gas cell to be filled with the pollutant or diluent gas that is to be
measured. Light, usually in the infrared, is transmitted through the
flue gas into the receiver unit, where it passes through the gas
filter cell and then through a neutral density filter (or no filter
at all). The gas filter essentially "filters out" the spectroscopic
regions in which the pollutant molecule absorbs light energy. This
100% filter thus gives a reference signal to which the light passing
through the neutral filter can be compared. The advantage of this
method, as well as with the differential absorption technique, is
that low levels of particulate matter will not affect the
measurement. Gas filter correlation instruments have been designed
to measure SO2/ NO, C02, and CO.
Second derivative spectroscopy is employed in a point in-situ
monitor that has been quite widely marketed. Using an oscillating
lens, UV absorption peaks of S02 and NO are scanned 45 times per
second. This scanning creates a signal that is related to the second
derivative of the absorption peak, taken with respect to wavelength.
Using the Beer-Lambert Law, it can be shown that this signal is
proportional to the concentration of the pollutant in the flue gas.
The instrument has a measurement cavity, generally 5 cm in length,
that is protected by a ceramic thimble. The system can be calibrated
by using either gas cells or cylinder gases.
Electrocatalytic analyzers used for in-situ measurements are
exclusively point monitors. Currently, this technique has been
applied for the measurement of 02 and SO2. As discussed in Section
9.2.1.4 for extractive analyzers, solid electrolytes can be used to
generate a measurable flow of electrons. Here, a ceramic thimble
keeps the measurement side of the solid electrolyte free of
particulate matter. Calibration gases can be injected into the
measurement cavity to check the instrument operation.
9.2.4 Recommended Maintenance - In-situ Monitoring Systems
9.2.4.1 Operation Checks (Daily Checks). The daily operation checks
associated with in-situ gas monitors are similar to those for
extractive gas monitoring systems. The operation checks should be
performed by a trained and qualified operator who has been given
responsibility for the system. The monitoring system will have a
better record of performance if the operator checks the system daily-
automatic zero and calibration procedures can create a sense of false
confidence that can lead to system failures.
-------
Section No. 3.0.9
Date 6/1/86
Page 26
The daily operation checks should start with a review of the
previous 24 hours of strip chart recordings and computer printouts.
Discrepancies should be noted on the check sheet and instrument
logbook. Again, any system faults identified by the indicator lights
should be resolved immediately.
Many in-situ systems use a gas cell containing pollutant gas at a
known concentration for calibration. In the optical in-situ ana-
lyzers, the cell is put into the path of the light beam, and the
signal is attenuated by a specific amount. This method may not, how-
ever, always check the complete optical system over which the flue gas
is measured, although the use of pipes that close out the stack gas
from the light path have, in some cases, alleviated the problem. Also,
the gas cell pollutant concentrations are not normally certified by
independent laboratories (their concentrations may, however, be inde-
pendently verified). Degradation of cell concentrations by leakage,
adsorption, or internal reactions can cause discrepancies in the
calibration data.
It is possible to calibrate some in-situ monitors with certified
cylinder gas. This can be done with in-situ point monitors by
flooding the volume within the ceramic thimble with calibration gas
or with zero gas. A "flow-through" gas cell can be used in either
single-pass or double-pass monitors. By flowing gas of a certified
concentration through a fixed cell in the instrument, a calibration
"traceable" to NBS or other certified gases can be obtained.
However, problems associated with the optical path used in the
calibration sequence may still be present. It should also be noted
if the temperature compensation circuit is disconnected in any of
these calibration sequences. Problems in these circuits may be
overlooked in such cases.
The optical alignment of the components of in-situ systems is
very important for proper operation. Alignments should be checked at
appropriate intervals. Many of the systems do not, however, incor-
porate alignment sights in their design. In such cases, detailed
electronic or optical checks may need to be performed to optimize the
system alignment.
Note again that actual calibration adjustments should not be
performed arbitrarily. Control limits should be set to avoid
calibrating against normal system fluctuations. The Federal
monitoring requirements specify that that the calibration be adjusted
whenever the zero (or low-level) or the high-level calibration value
exceeds two times the limit of the applicable performance specifica-
tions. * Performance specifications are given in Appendix B of Part
60 of the U.S. Code of Federal Regulations - Title 40. If either
-------
Section No. 3.0.9
Date 6/1/86
Page 27
calibration value exceeds four times the applicable performance
specification during any calibration drift check, all of the data
collected since the last check are considered to be data collected
during a CEM system out-of-control period and are rejected.
Similarly, if the system exhibits drift (either low-level or high-
level) greater than the performance specification for five conse-
cutive 24-h periods, the system is again viewed as being out of
control, and subsequent data are rejected until the system problems
are corrected.5
Figure 9.4 gives a format for developing a daily QA check sheet
for in-situ systems.
9.2.4.2 Routine Maintenance (30-day Checks). Routine maintenance
should be initially performed on in-situ monitors, at a minimum,
every 30 days. Although many equipment vendors recommend routine
maintenance at periods of two or three months, it is advisable to
gain experience with the system before waiting these longer periods.
Routine maintenance for in-situ systems should consist of
ensuring that key components of the system are clean and operational.
In general, windows, filters, and desiccants should be cleaned and/or
replaced.
In the cleaning procedures for windows or optics, great care
should be taken. Lens tissue or clean, soft rags should be used with
a solution of alcohol and water. In severe cases, mild detergent may
be necessary to clean windows exposed to the flue gas. Sensitive
optical components such as diffraction gratings should never be
touched or cleaned in the field. Fingerprints or traces of cleaning
materials can severely affect their performance, so special
techniques must be used.
A systematic procedure should be instituted for checking the
electrical performance to compare it with the original factory or
start-up performance. This normally involves using a digital volt-
meter and oscilloscope to check the analyzer at various test points.
These test points should include a check of lamp voltages, power
supply voltages, and detector outputs. The procedures involved in
these checks will generally require a well-trained serviceman or
electronic technician.
Many in-situ analyzers use chopper motors in the transceiver
assembly to modulate the light beam or switch instrument functions.
These motors should be checked for bearing noise or for excessive
vibration. For motors that automatically move mirrors or gas cells
-------
Section No. 3.0.9
Date 6/1/86
Page 28
Example Format for In-sLtu Gas Monitoring System
Daily QC Checks
Plan: C$L\ Vfljrtfgf / kCjfrle. Date 1/28/85 Time \O\\loCLJT\
Unit fipULe/v *•' C/UtipJ"
Gas Monitsred 5fc>j /C.OA
*nal/2er 1.0. CrOSiS^T-^rV"! 1— A5O
Scan Value 1 OoO rarn -^Pp/IE* to.
Calibration Value BYa. ia%C^>
icxiir.-ier cis, -;js ceUil
Zero Gas (air, N?, other) W/4
Part 1 Indicators
Indicator L-Tr.ta S:a
Mam. -RoteH= Exwrnee^
Phone fiq-5t-q-.au.
Offset Value /O j^pno
So*.
Paoer Status:
Str ip chart OK
Printer ^l^
•us 1 Problem/Ac ti an taken
Power failure Q^
Blower failure ; ...
Lamp failure OK j
0lcty "ind" at (u-^artKO Mote, te ol«n^ new:t tuett
Alignment Q^.
*lacrn UluVlu'n U»t«>«
Oth'r i . /Vo«'.5&
Gas cell 12 || ; 36
Calibration qas
{ if aoclicabU)
'
,
a 1 13. pp^vt i? a.
| ^ia pp^*i "^sio ^oa-
i
! IA. ^ la.o i A. j
i
i 385 pf»vi 31^ 37S
Part 3 Zero and Span Adjustment (if outside of control Liaits)
Zerocontrollimitt 5 Don
Ad jus fed Read ini a | Tin*
z«f<> i ia;i5 p<
Gas cell » I , 1^-. 3^
Meter ! 5:.-.? Cr.art Digital Printer
^ 0 ! 0 0
\ iia. • lia 913-
Gas cell ll-O
Calibration ^a? ! - I
5tac'< ::r;c »-.- : i- icr. , : :
a==:ic.«:*; i:iti
Hft>tYt Cl Vt I V<^|- %1 /.M"/
Operator ai-qr.dt-rs "^'e
! 3^3. : 3U-3. 3H-3-
at 4^_ jU,\ir
„,.„.,„ .^ N ,,
Figure 9.4. Example format for in-situ gas monitoring system daily
QC checks.
-------
Section No. 3.0.9
Date 6/1/86
Page 29
into position, the operations should be manually checked to see that
the movements are smooth and complete.
Observations made during the routine maintenance period should be
noted in the system logbook to provide the base data necessary to
optimize the maintenance procedures. After six months to one year of
operation, trends in performance should become apparent. These
trends may indicate that some maintenance procedures may need to be
improved or that others may need to be performed more frequently or
less frequently.
Figure 9.5 gives a suggested format for the development of an in-
situ system maintenance check sheet.
9.2.4.3 Performance Audits. The discussion in Section 9.2.2.3 for
performance audits on extractive systems is also applicable to in-
situ systems. However, in the case of path in-situ monitors that
come under the requirements of Appendix P, it is not possible to
conduct a cylinder gas audit (CGA) unless a flow-through cell is
incorporated in the system. RATA and RAA audits can, however, still
be conducted on these systems. Path monitors should be checked quite
frequently to see if they are giving representative data, since, as
mentioned previously, the methods used for internal zero and
calibration may not check the complete optical system. The in-situ
point monitors can, in most cases, be checked with cylinder gas, but
audits independent of the monitoring system should also be performed
at regular intervals.
9.2.4.4 Corrective Maintenance (Problems and Troubleshooting). In-
situ monitors avoid many of the problems associated with extractive
systems; however, in-situ systems are not devoid of problems. In-
situ systems installed directly on a stack or duct are subjected to
severe environments. Temperature cycling, acid gases, and vibration
can damage the sensitive optical components and alter the optical
alignment that is important to these systems. The electrical
components located in the mounted system assemblies are subjected to
the same type of abuse and can fail if not adequately protected.
The often complex nature of in-situ systems requires a higher
level of troubleshooting skill. It is advisable that the operator
responsible for a CEM system ' first be trained by the instrument
vendor, preferably in a formal training session at the vendor's
facility.
The basic problems that occur are often a result of the environ-
ment, such as vibration and ambient temperature variation. The
solutions to these problems are very specific to a given installation
-------
Section No. 3.0.9
Date 6/1/86
Page 30
Example Format Cor In-situ Monitor
30-day Maintenance Check Sheet
Plant CcL.\vafY\0C /Acsr\€. fhwef Date H/Sk./?^) Time /0;DS QJVI
unit TboiLp/i Wo- i - Ou*i»i: Name "Robert £rcu*ef>r
1.T
••oni'.ot 1.0. Ho. CjrCf¥Jtf>cb\ L2.50 Pr.one 1 lirtj/
& (jDmtenaxJwi.
—
Vi'3uo.liy dlir-ty
Claecktct OC.
Action
CoHec*€d/ ne«io«^C*
t Icjcaji ct^d>ve^
lUipedi el«tr\
Ouiperf cteouxv
LUip«d* ci-eo-i^.
CX->rf«* cJeox\
(Wifxa* cJwa*v_
OK.
OK
Clax^ei ptvsefRA£tito\
CWeoLned
C.i«ui«e)
"Ref3tac«d5
I\JO OLtVtiorv
Status
OK
g^Biue n Red
C>K
OK
RcLsjx^g
Ac t ion
—
—
-
"R«pU«jr,«GL brushes S/ia-/J5
"Robei-t SrviLoepr 4/'S>-/8-5 CAc — ^ 6u
-------
Section No. 3.0.9
Date 6/1/86
Page 31
and may require re-installation or the construction of environ-
mentally controlled enclosures around the analyzers.
Table 9.4 lists some of the problems associated with in-situ
systems and recommendations for corrective action.
9.3 Opacity CEM System Operation Practices
9.3.1 System Design - Many types of instruments have been developed
that can measure the opacity of a gas in a stack or duct. These
instruments are called opacity monitors, or alternatively,
transmissometers. They are in-situ monitors and can be designed
either as single-pass or as double-pass units.
9.3.1.1 Single-pass Systems. Single-pass opacity monitors are quite
simple in design. Light emitted from a lamp passes through the stack
and is sensed by a detector on the other side of the stack. Blowers
located on each side help keep the windows of the apparatus clean.
One problem associated with the design is that of calibration.
The electronic and optical components of the system cannot be checked
unless the stack or duct is free of particulate matter. This can be
achieved if the industrial process is shut down and provisions are
made so that residual dust does not recirculate in the stack. Since
most industrial processes cannot shut down just to zero and calibrate
an opacity monitor, instrument designers have used optical light
fibers or zero-pipes to provide this capability. The light fibers
pass from the lamp to the detector around the outside of the stack.
The zero-pipes pass through the stack and can be purged with air to
provide a zero reading for the instrument.
9.3.1.2 Double-pass Systems. In double-pass opacity monitors, light
crosses the stack and is returned by using a retroreflector. The
retroreflector returns it to the main analyzer housing, where a
detector then senses the light. This instrument can be zeroed and
calibrated by flipping a mirror up into the light path to approximate
a clean-stack condition. The mirror close to the transceiver
prevents the beam from crossing the stack and merely sends the light
through the clean interior to give a "pseudo-zero" reading. A
calibration filter can then be flipped into the path of light to give
an upscale calibration reading. Rotating choppers have also been
used for zero and calibration procedures.
Blowers again are used to keep clean the optical surfaces exposed
to the stack gas. Filters are needed before the blower so that clean
air will pass through the analyzer.
-------
Section No. 3.0.9
Date 6/1/86
Page 32
TABLE 9.4. IN-SITU GAS ANALYZER PROBLEMS
Problem
Possible Corrective Action
Excessive dirt buildup on windows or
thimbles
Cyclic drift in signal unrelated to
plant performance - due to ambient
temperature changes. Signal
becomes erratic from high
temperatures
Optics misalignment/electrical noise
due to stack or duct
Signal becomes erratic at high
opacities
Misaligned system
Probe/seal leaks (in-stack monitors)
Lightning strikes
Static charge buildup burning out
circuit boards
Lamp burnout/degradation
Gas cells unreliable
Spurious readings during plant
start-up, shutdown, etc.
Improper temperature compensation
Increased response time
Clean blower filters on path systems. In-
crease blower capacity. Rotate ceramic
thimbles on point systems or replace.
Insulate protective hood or install
temperature conditioning system about
monitor.
Mount assemblies independently from duct.
Use flexible bellows for duct
connection. Dampen mountings. Relocate
monitor.
Relocate monitor after control device.
Water droplets from scrubber may also
cause this problem. In such cases,
analyzer may have to be located before
the scrubber.
Realign, check, and tighten system.
Return to vendor for repair.
Add phenolic gaskets. Add surge arresters.
Relocate monitor.
Run copper cables to earth ground.
Replace on regular schedule. Check lamp
power supply for high incidence of lamp
failure.
Cell leakage, losses to walls - recalibrate
cells or replace.
Install transient suppressor, dedicated
transformer for monitoring system.
Adjust circuits, recalibrate,
or replace boards.
or repair
For in-stack in-situ monitors, check
ceramic thimble for plugging. Replace
if necessary.
-------
Section No. 3.0.9
Date 6/1/86
Page 33
Any transmissometer, be it single-pass or double-pass, will have
similar components, including a light source, a detector, and various
optical components. The light source may be specifically selected
for the system or may merely be a lamp normally used in an automobile
headlight assembly. Detectors are chosen so that they will sense the
light emitted by the source. This light is normally in the visible
and infrared regions of the spectrum. Optical components include
focusing lenses, mirrors, and optical filters. Filters are used to
help the instrument respond only to light in the visible region of
the spectrum - the so-called photopic region, to which the human eye
is sensitive. Filters normally prevent infrared light from being
sensed by the detector.
Mirrors, of course, direct the light from point to point inside
the instrument, so that the operations of measurement, zeroing, and
calibrating can take place. Lenses focus the light so that a well-
collimated beam is sent across the stack and through the instrument.
In addition to these components, chopper wheels and light modulating
wheels are often used to direct or modify the light beam.
Various accessories can be purchased for the opacity monitor.
These include integrators, beam combiners, "reasons" panels, and so
on. The main instrument panel for the monitor is normally installed
in the control room of the plant. Here, the instrument outputs are
observed and recorded.
Transmissometers are generally installed to meet requirements of
pollution control agencies. The Federal EPA has developed speci-
fications for these instruments.1 There are basically two types of
specifications: design specifications and performance specifica-
tions. The design specifications detail how the instrument is to be
constructed.
Monitors that satisfy the design specifications are purchased and
installed in a location according to EPA guidelines.1 The installed
monitor undergoes the performance specification test procedures.
These procedures check the system for zero and calibration drift over
a one-week period. Passing this test indicates that the monitor is
acceptable for the opacity monitoring.
However, testing should not stop at this point. As with all
instruments, problems can occur over long periods of operation.
Blower filters must be checked and regularly cleaned, burned-out
lamps replaced, and the integrity of the data must be consistently
audited. The continued success of a monitoring program depends
heavily on how well the instruments are maintained.
-------
Section No. 3.0.9
Date 6/1/86
Page 34
9.3.2 Recommended Maintenance - Opacity Monitoring Systems
9.3.2.1 Operation Checks (Daily Checks). Operation checks should be
conducted daily. These checks can generally be performed at the
remote control unit located in the station control room. It is not
necessary to go into the plant to check the transmissometer itself,
unless a fault lamp indicates a problem has developed.
At this level, the activities include noting the status of the
system indicator (fault) lights and recording the daily zero and span
values on an appropriate form. The day's strip chart record should
be examined to check for trends or problems that might not be
identified by the system fault lights. At this time, the strip chart
should be annotated. Figure 9.6 gives examples of the type of
information that should be noted.
1OVU
1700
1600
1500
1400
1 1300
1200
1100
1000
onn
C
*
^
i
I
i
^|
^
e,
^
-i
i
j
•
=3
^l
/
^
^i
s5
1,
IP
a —
1:15 am
4/
"~Ch
FS
Of
20/84
art speed 5 cm/h
= 100% Opacity
fset=10%
r —
-10 0 10 20 30 40 50 60 70 80 90 100
Average opacity, six minutes
Figure 9.6. Strip chart
annotations.
The first level of quality control serves to alert the operator
to problems or necessary adjustments. If the window indicator or
-------
Section No. 3.0.9
Date 6/1/86
Page 35
zero compensation light (where applicable) warns that the windows are
dirty, someone will have to go to the duct or stack and clean the
windows of the transceiver and retroreflector assemblies. If the
span reading or zero reading has drifted to an unacceptable degree/
the span or zero potentiometers may have to be adjusted to bring the
readings back to the proper values, or other corrective action may be
necessary. Care should be taken in such adjustments, however. If
there is only a small difference from the proper values, the
difference may be due just to random noise in the system. Also, if
the drift is large, problems may be developing that cannot be solved
by merely adjusting zero and span potentiometers. For these reasons,
"control limits" are commonly established to aid in deciding whether
to adjust the system. The specific limits should be established at a
level that (1) ensures the monitoring data will be sufficiently
precise for its intended use and (2) is achievable.
To make the daily check thorough and complete, log sheets are
often developed by the operator. Figure 9.7 should be regarded as a
starting point for developing such a form.
Part 1 of Figure 9.7 provides a checklist for the indicator
lights on the remote control panel. Items incorporated in Part 1
will depend on the specific instrument system. Depending on the
status of the lights, the daily operation check may develop into more
than just a check. A blower failure or dirty window alarm will
require a visit to the transmissometer for corrective action.
Note also that Parts 2 and 3 of Figure 9.7 require data from the
meter, the strip chart, and the digital printer. If the monitoring
system presents data in all three of these formats, the data obtained
from each should be identical. If they are not, the correct one must
be determined. In poorly maintained systems, improperly zeroed
meters or recorders can create discrepancies. Also, meter readings
often differ from the computer printout because the transmissometer
and computer are improperly connected.
Zero compensation values should also be recorded so that they can
be evaluated at a later time. After several weeks, these data can be
drawn from the daily logs to evaluate the rate of window soiling.
As a part of the quality assurance program, the form should be
signed by the person performing the checks. At appropriate
intervals, the operator's supervisor should review and initial the
logs to see that assigned responsibilities are being carried out.
9.3.2.2 Routine Maintenance (30-day Checks). The second level of
quality control for transmissometer systems involves establishing a
-------
Section No. 3.0.9
Date 6/1/86
Page 36
Example Format for Transmissometer Daily QC Check Sheet
Plant LaJyoyY+er /Acvvre ffcujer Date ^/SiS'/KfT Time /'. /5"pK>n
Unit uQilfff No. 1 OtrOe-t Name 'Birjljpr't. Eri«u-ieer
Transmissometer 1.0. Ho. CrDSsfecH Tl.ll Phone fl<^- SfR-IS^M-
Span filter value ?R.p-^ Stack-exit correlation value (l*/lr) Q.9/
Hours ooeratinq in period: ^'f/3'V Paoer status: strip char; <3K
printer OK
P»rt 1 Indicator Lights | Status Problen/Accion taken
Power failure o^
Blower failure ....
L?(S
Lamp failure olKi«.
Other _ j
Meter Strip char; {Digital printer
Part 2 unadjusted Readings | Ti:"e * °P ° ' °? | D j < 0? | D
2eco iiaopm 0.6 o.ooa. e>. a. | o.eoi 1 O-3 [o.oot
sp«n r.a6 at.5 o.iaf AU.I (0.131 | at.j. jo./sa.
Stack opacity ,130 Jv.S' O.07? ' /fe . / 1 O.O7fe 1 /(p. A. |<5. 077
Zero compensation 1' 36 I.A. 0 005
Part 3 Adjusted Zero and Span Readings (if outside of control Units)
Zero control limit » f£ ^ Op, Span control limit « £ * ^P
Meter Strip chart Digital printer
Time 4 Op o A Op D J Op D
Zero 1:35 0 f
Zero compensation ^ . qg / ^ O.OOS' windows cleaned? Yes Q *° &
Comments/observations:
'Rc'jjgf^j S^ci^tear" l/J-f/SS^ fJU*— <^\ OAX\.^S <^\x'^\li5')
Operator Signature Date Supervisor S^igr.atiUaJ Date
Figure 9.7. Example format for transmissometer daily QC check sheet.
-------
Section No. 3.0.9
Date 6/1/86
Page 37
schedule for routine maintenance. A basic schedule may be provided
in the instruction manual, but such schedules do not address problems
specific to an installation. Over a period of time, after the
operator is comfortable and experienced with the system, a
maintenance routine can be developed. It may take a year or two to
perfect the schedule, but once developed, it should be adhered to.
It is recommended that, at least initially, a 30-day maintenance
routine be established. Housekeeping of the transmissometer equip-
ment located on the stack is the predominant form of maintenance per-
formed at this interval. Cleaning of outer surfaces, windows, parts,
and filters is the major activity. Once the operator or serviceman
is at the instrument site, the alignment, desiccants, and blower
motors can also be checked.
After a few system failures have been experienced, a replacement
schedule can be developed for parts having a lifetime of a year or
more. For example, transmissometer lamps will generally have a
lifetime of over two years (20,000 h). Rather than allowing the lamp
to burn out and cause a system failure, it can be replaced routinely
at the end of every two-year period. Blower motors also require
maintenance. Since the motors are "on" continuously, at some point
the bearings are going to wear out. The period of time it takes for
this to happen will depend on several things - such as the amount of
particulate matter and acid in the ambient air and the amount of
stack vibration. These factors will aggravate wear to cause a
problem eventually. When breakdown does occur, the period for motor
maintenance can then be established.
In the transmissometer check sheet given in Figure 9.8, there are
two parts, one for required maintenance actions, the other for
required maintenance checks. The serviceman or operator should
perform the basic housekeeping chores of cleaning and then check for
other problems. If a problem is observed or if one appears to be
developing, corrective action should then be taken.
The form should be modified to fit both the plant maintenance
routines and the characteristics of the actual system. Within a year
after the system has passed Performance Specification 1, the operator
should use the logbook and daily logs to devise a schedule of
periodic preventive maintenance that best fits the installation.
9.3.2.3 Performance Audits. The third level of activity that a
company should institute for its opacity monitoring system involves
conducting performance audits.
-------
Section No. 3.0.9
Date 6/1/86
Page 38
Example Format Cor Transmissometer 30-day Maintenance Sheet
Plant (la.\ypLr^e.r /Actvie fewer Date M/ia./S'Sj Time /( '• \S ajr\
unit Boiler- Wo. 1 - Outlet Name "Robert Sncti^eer"
Transmissometer I.D. Mo. CrPSStech "Tl^l
Part 1 Required Maintenance Actions
Clean outer surfaces
Hood
Transceiver assembly
Retroref lector assembly
*
Purge air system
Clean inner surfaces
Hood
Transceiver assembly t'jnlatchedl
Retroref lector assembly i^-j latched)
Preseparator - air purging system
Clean transceiver windows
Clean retroref lector windows
Clean fiber-optic cables (if applicabls)
Replace purge air filter (or clean and
replace)
Tighten hose cla.Ttps, cables, mountings
Phone • £T;<1'-5HC1- ISI.M-
Status
Dosl-
J>ir^
Di'rt^
D,r-t^
PLUn
OK
OK
Diri^
D/rgT^' '
^';Ht
NA
92-%
Part 2 Required Maintenance Checks
Status
Check alignment (?) Misaligned
Check desiccants Blue (3 Red LJ
Check cables - continuity, pinching/ T/Txx>=c«iver coJOK-
cuts, corrosion leoje
Check hoses - continuity, pinching/
cuts, corrosion OK
Check security seals 0K.
Blower motor - bearing noise £,^
nob^ydl" £wci/T>eer' H/iays?> Cx/i-.^c^
Action
^rlS^J-eo^ed
U^iped1
Luif>^»
UJ.perf
W.ped
(\JOioe
(\^sne.
Cu.pcd
C ^Ga.^&A
CI«L.r»:C*
7.ep^^
Ac t ion
(V) Realigned
"^•^ ^^
-
-
_
6" ^Ml<'
Operator Signature Date , Supervisor Sjignatu^je/ Date
Figure 9.8. Example format for transmissometer 30-day maintenance
check sheet.
-------
Section No. 3.0.9
Date 6/1/86
Page 39
Audit devices have been developed for most of the double-pass
transmissometer systems and for some of the single-pass systems.
Basically, the devices designed for double-pass systems consist of
holders that can be attached onto the transceiver. The holder
consists of a slot for holding calibration filters and a short-range
retroreflector. The device and the transceiver basically constitute
a "mini-transmissometer" that can accommodate audit calibration
filters. The device also contains an iris, which allows the reflec-
tance of the mirror to be adjusted so that it will correspond to a
simulated, stack zero. It is best to make this adjustment initially
when the system is set up for the performance specification
calibration and response time test. After the instrument is
installed, it provides a method of checking the simulated, instrument
zero. This "audit zero," of course, may not necessarily be identical
to a clean-stack zero, but comparing it to the instrument's internal
"simulated zero" provides a good quality control check.
An audit does not just involve obtaining data from audit devices.
An audit is a check of the performance of the entire transmissometer
system. Indicator lamp status, stack-exit correlation corrections,
alignment, and other functions of the system should all be checked at
this time by the auditor. The most common problems uncovered in
audits are errors associated with the calculation of stack-exit
corrected opacity values. An auditor can easily uncover such a
problem through a well-designed program.
A system of auditing procedures has been developed for double-
pass transmissometers.6 This system, designed for use by control
agency personnel, gives detailed step-by-step instructions for
conducting an audit using the calibration filter audit devices. A
generalized form, based on the work detailed in reference 6, is
presented in Figure 9.9.
Figure 9.9 indicates the type of information that can be obtained
during an audit. Using a ^calibration audit device, both the
transceiver optics and electronics are evaluated. Using a reference
signal source, the data handling system can be evaluated. In the
case of a double-pass system, these procedures check only the
transceiver assembly and data handling system. Since the retro-
reflector assembly is not involved in the checks, the audit evaluates
only part of the system. It is possible that a system can pass an
audit without problems, but. the cross-stack opacity readings may
still be inaccurate if misalignment or window fouling problems occur
at the retroreflector side of the instrument.
-------
Section No. 3.0.9
Date 6/1/86
Page 40
Exaaple Format for Transmissometer Performance Audit Data Sheet
Plane CaJvorVW /Acwif. ftxugr Date l/'Wf5 Tine ,
Unit 'BpJer Mn. I - Outtefc
Tr>n»iiioncir 1.0. »0
.Cro.
TlJJ
A. Stack-e«it correlation
1. Emission outlet pathLengtn, ln
2. Monitor pathlength, /, I &_2_*e:a:i
:alculated ratio I*,
S. ault indicator lamps OM CFP
Power
Otrtf window
Air purge
Alarm
C. Internal zero a.id span che
1. Internal span value
2. Internal span value 'crrr»cc»d co JCJCJt exi-) ^?-£ *
Opx«t-< l-'PuMnmrtfl'v't
3. Remote control unit ^etar readings (corrected)
Zero raadina Q.C A
Span read I .ig ^"» «
4. Scrip chart readings (corrected)
Zaro raadinq ^-^ — ^
Span reading ..—. , ^i V i
9. Oiqttal printout icorrecced;
Zero reading ,.—., , J^»^ \
Span reading —*^- fc *
Averaging time \£. GiS" ,3. C^ \ op
Before cleaning _.
After cleaning r .<^_
Time cleaned Jt'C-5
2. Alignment chack-stacus — ^
3. Other mater performance indicators 'if dpplicdM*.' ^A
Raf signal £**
AGC
(. Recneck of instrument intarnaL calibrations
Tine //*'/6 L/J»/
2«ro reading ^-^
Span readi.-.g 3±LJc*
Zaro compansation ii
II. Calibration Device Audit
A. Audit device lero
L. «ater 0.0 *0p
Strip char: Q.^L *Cp
Print output &Q "^Op
Low-rang* filter
Filter number
Filter opicity (urccer*cte4) /C. A
Filter opacity fcorr«ec*i> 9.. 3
Figure 9.9.
Example format for transmissometer performance
audit data sheet.
-------
Section No. 3.0.9
Date 6/1/86
Page 41
II. (
c
[
III.
IV.
_J
alibracioo Device Audit (Continued)
I. Filter nuaiber A£~_/*fF
2. filter opacity (uncorreccedj 30. ¥
3. filter opacity ('corrected^ /#, X
t. High-range filter
2. Filter pacity funccr recced/1 •*//. 3.
3 . filter pac i t y f corrected 1 3S-3
finte^r ted corrected opacity values frost systeta cuepuc - computer printout
or eiyui alenc)
| Tiaie i | Time j T:=e j
I ! r.Cfp~.\ //.f i l;si> 1C.7 1 /•>>- '
2 . /.•« j ii? • ! a: /a. 30.1 '. W •
1 i f.tc ! //.-' : j.'jy j j/.j ! 3:sa
4 • ?:3o \ il.'f ' 3Wf <3c.3 ' v:.^7
5 i Afjv i //. f ' *+:3S £C.<, \ ^;vc,
: \ ••
RKOrdlnq STIOKI clcctranic Audit Utinq RtCinaei Si«ul>
A. sti?-:tv.i t«st
?.«ftrt^.c«
>:J?.IL Strip 0.-)i:«i
?•«: Ko. I.T14 2r .IT'.*) raadin^ r«adin9
; -^f ^ — ^f-
, :o.o it.q wr
, is.c sa.3 s.s.(.
s //.o -jf.tf h\«r*\ o>'ic< loce-n'
2-rrc bet^ren 3-^/0 rl^rr a«rf .-to;(fet( /xr/tteat
\ Opacity
•*.,
General Syste* Coaaents i'.':ousff>ceepi^7 ciservacicns, cperacor sAili -eve is,
recommended corrections, ;jrprovenen cs ;
before oudjJc . 5.*% *'<& M£. rftuUv n^lec-tci m at
dtevtc*. r&uUnfS k- (^IdFftlS irtQ OfXLCL~C
-------
Section No. 3.0.9
Date 6/1/86
Page 42
Sufficient data should be taken to calculate mean errors,
confidence intervals, and the like. Detailed methods for these
calculations have been documented.2
9.3.2.4 Corrective Maintenance (Problems and Troubleshooting). Even
in the best of systems, problems will eventually develop. Quality
control technique and quality assurance programs can help to minimize
problems, but they cannot eliminate them. Table 9.5 lists some
common and uncommon problems associated with transmissometer systems.
Faults and component failures are generally immediately obvious.
Some problems can be so subtle that they will be detected only during
a period of corrective maintenance or during a performance audit.
But then again, some otherwise' obvious problems are not recognized
for long periods of time and are discovered only during the
independent performance audit.
Problems that are site-specific or instrument-specific may
develop at an installation. The problem and the corrective action
taken should be described in the logbook. If a recurring problem is
preventable, the method used to prevent it or the means of corrective
action should be incorporated into the preventive maintenance
schedule.
Troubleshooting is an acquired skill. A good instrument
technician can piece together a number of clues and rapidly identify
a problem. If the monitoring system is new and unfamiliar, it may
take longer, but the time required should decrease as experience is
gained.
One of the most useful troubleshooting tools for transmissometers
is the strip chart recorder. A common tendency is to eliminate the
strip chart recorder with the rationalization that a computer can be
programmed to identify system faults. This can be done, but
computers and programming have their own faults. In any case, a
recorder provides a good backup to any data handling system, since it
is easier to detect trends from the analog output. At a glance,
noise levels and system performance can generally be evaluated. The
strip chart record can indicate a number of the problems that have
been identified in Table 9.5. Figure 9.10 illustrates the type of
strip chart traces that might be observed.7
The problems that can be detected vary from environmental ones,
such as temperature effects and vibration, to instrumentation
problems, such as sticking zero mirrors.
Figure 9.10 presents only a few examples of clues that can help
the instrument serviceman. Microprocessors incorporated in modern
-------
Section No. 3.0.9
Date 6/1/86
Page 43
TABLE 9.5. Opacity Monitoring System Problems
Problem
Corrective Action
Part 1. Common Operator
Problems
Rezeroing and calibrating without
cleaning windows.
Excessive zeroing and calibrating
(attempts to adjust random noise).
Wrong stack-exit correlation value
set in instrument.
Alarm goes off when system goes
through span check.
Lack of correspondence between
(1) meter opacity and optical
density readings;
(2) meter scales; or
(3) meter, strip chart, and
digital printout values.
Failure to clean windows when fault
light indicates cleaning required.
Failure to clean retroreflector
window.
Improper use or no use of combiner
. equations for multiple monitor
system.
Part 2. Physical Problems
Blower motor bearings freeze-bearing
noise.
Clean windows first.
Set statistical criteria for making
adjustments.
Correct calculation. Flange-to-flange
distance ofjten mistakenly used rather
than inside stack diameter.
Reprogram system. A common programming
error - although this is sometimes done
intentionally to check the alarm system.
Recalibrate system. Most systems can be
adjusted to produce consistent readings.
It may take some work, but a good
operator will see that it is done.
Clean windows.
Correct calculations.
Replace bearings - develop better
preventive maintenance program.
Excessive dirt buildup on windows.
Cyclic drift in signal unrelated to
plant performance - due to ambient
temperature changes.
Monitor reads 100% opacity for long
period of time; protective shutter
in place.
Clean filters.
Insulate protective hood or install
temperature conditioning system about
monitor.
Reset shutter or troubleshoot purge air
supply (once activated and in place,
some shutters have to be manually
reset).
(continued)
-------
Section No. 3.0.9
Date 6/1/86
Page 44
TABLE 9.5. (Continued)
Problem
Corrective Action.
Part 2. Physical Problems
(Continued)
Excessive duct or stack vibration
causing misalignment and
electrical noise.
Improper alignment.
Part 3. Electrical Problems
Ground loops and noise.
Large voltage drops when plant
equipment is started; spikes in
strip chart record.
Static electrical charges.
Lightning strikes burning electronic
circuits.
System stuck in simulated zero or
span calibration mode.
Lamp out, erratic lamp output (a
fault light should indicate this
problem).
Mo output from instrument, no
calibration cycle, etc.
Improper instrument responses -
faulty calibrations, improper or
no outputs.
A definite problem. One solution is to
mount transceiver and retroreflector
assemblies independently from this stack
or duct. Make connections between duct
and monitor using a flexible bellows.
Another is to move the monitor to a less
vibration-prone location.
Realign, check, and tighten system.
Trace and rewire.
Install transient suppressor, dedicated
power transformer for monitoring system.
Connect transmissometer case to dedicated
earth ground.
Add phenolic gaskets between metal stack
and transmissometer. Add surge
arrestors to junction at junction box.
Check solenoids and motors in transceiver
clean or replace.
Check modulator and motor; adjust or re-
place motor if necessary. Check lamp
and replace; when replacing, keep glass
surfaces of lamp clean; avoid finger-
prints and clean with lens solution
before turning on.
Check fuses (hope that it is this simple).
Check electronics. Check to see that cards
and components are secure. Use
troubleshooting guide supplied by vendor
to check electronic test points.
Replace appropriate components or
replace cards.
-------
Section No. 3.0.9
Date 6/1/86
Page 45
instrument systems are today being programmed to directly alert the
operator about such problems. However, a computerized system can be
programmed to identify only a set number of malfunctions. If
problems occur that are not identified by such a system, a strip
chart recorder can be a valuable tool in uncovering them.
Suspect
Excessive drift
Window fouling
Straight line trace
sustained for several
hours
Cyclic pattern/
temperature sensitivity l?l
Spikes-
electrical instability
Source: Larkin, R.. 1977. Ref.7.
Drift, gradually
increasing or
decreasing
Bad zero check
Figure 9.10. Possible strip chart
traces indicating problems.
9.4 References
1. Code of Federal Regulations. Title 40 Protection of
Environment. Part 60 Appendix B. 40 CFR 60 App. B, 1985.
the
2. Jahnke, J.A. and Aldina, G.J. Continuous Air Pollution Source
Monitoring Systems. EPA 625/6-79-005, June 1979.
3. Traceability Protocol for Establishing True Concentrations of
Gases Used for Calibration and Audits of Continuous Source
Emission Monitors (Protocol Number 1). June 1978, Section 3.0.4
of the Quality Assurance Handbook for Air Pollution Measurement
Systems, Volume III, Stationary Source Specific Methods.
EPA-600/4-77-027b. August 1977. U.S. Environmental Protection
Agency, Office of Research and Development Publications, 26 West
St. Clair Street, Cincinnati, OH 45268.
-------
Section No. 3.0.9
Date 6/1/86
Page 46
Code of Federal Regulations. Title 40 Protection of the
Environment. Part 60.13, Monitoring Requirements. 40 CFR 60.13,
1985.
Standards of Performance for New Stationary Sources: 40 CFR 60,
Appendix F - Quality Assurance Procedures, Procedure 1 - Quality
Assurance Requirements for Gaseous Continuous Emission Monitoring
Systems Used for Compliance Determination.
Purcell, R.Y. and Rosenquest, J.M. Field Performance Audit
Procedures for Opacity Monitors. EPA CEM report series No.
5-271-7/82 (see also, Performance Audit Procedures for Opacity
Monitors. EPA 340/1-83-010), 1982.
Larkin, R., Jaye, F., and Steiner, J. Resource Manual for
Implementing the NSPS Continuous Monitoring Regulations. Manual
4 - Source Operating and Maintenance Procedures for Continuous
Monitoring Systems. EPA 340/l-78-005d, 1978.
Additional References
Entropy Environmentalists, Inc. Assessment of the Adequacy of
the Appendix F Quality Assurance Procedures for Maintaining CEMs
Data Accuracy: Status Report #1. EPA 600/4-83-047, 1983.
Entropy Environmentalists, Inc. Guidelines for the Observations
of Performance Specification Tests of Continuous Emission
Monitors. EPA 340/1-83-009, 1983.
Jahnke, J.A. APTI Course SI: 476A Transmissometer Systems -
Operation and Maintenance. An Advanced Course. EPA 450/2-84-004,
September 1984.
Logan, J.T. and Rollins, R. Quality Assurance for Compliance
Continuous Emission Monitoring Systems: Evaluation of Span Drift
for Gas CEMs, Proceedings - APCA/ASQC Specialty Conference:
Quality Assurance in Air Pollution Measurements. October 10-14,
1984, Boulder, CO.
Osborne, M.C. and Midgett, M.R. Survey of Continuous Source
Emission Monitors: Survey No. 1 NOx and SO2. EPA 600/4-77-022,
1977.
Peeler, J.W. A Compilation of SO2 and NOy Continuous Emission
Monitor Reliability Information. EPA 340/1-83-012, 1983.
-------
Section No. 3.0.9
Date 6/1/86
Page 47
7. Peeler, J.W. Performance Audit Procedures for S02r NOx/ CO2, and
02 Continuous Emission Monitoring Systems. EPA 340/1-85-015,
1983.
8. Reynolds, W.E. Development and Evaluation of SO2 CEM QA
Procedures. Proceedings - APCA/ASQC Specialty Conference:
Quality Assurance in Air Pollution Measurements. October 10-14,
1984, Boulder, CO.
9. Wayne, A. Opacity Monitoring Quality Assurance Programs in U.S.
EPA Region VIII. Proceedings-APCA/ASQC Specialty Conference:
Quality Assurance in Air Pollution Measurement. October 10-14,
1984, Boulder, CO.
-------
Section No. 3.0.10
Date November 26, 1985
Page 1
10.0 GUIDELINE FOR DEVELOPING QUALITY CONTROL PROCEDURES
FOR GASEOUS CONTINUOUS EMISSION MONITORING SYSTEMS
10.1 Introduction
This guideline describes the minimum content for a quality
control plan to satisfy the requirements of Section 3 of
Appendix F, Procedure 1 to 40 CFR Part 60. Source owners or
operators may wish to add other items to assure the generation
and reporting of valid data from their continuous emission
monitoring systems (GEMS's).
Appendix F, Procedure 1 requires written procedures for each
of the following activities:
1. Calibration of the CEMS.
2. Calibration drift determination and adjustment of the
CEMS.
3. Preventive maintenance of the CEMS (including maintain-
ing a spare parts inventory).
4. Data recording, calculations, and reporting for emis-
sions and QA data.
5. Accuracy audit procedures including sampling and
analysis methods.
6. Program of corrective action for the malfunctioning
CEMS.
Figure 1 is a flow chart showing the requirements in
Appendix F, Procedure 1 for quality assurance and in Part 60.13
for monitoring requirements. This flow chart is included to show
how these requirements for CEMS's interact.
10.2 Calibration of the CEMS
Calibration refers to the adjustment of the CEMS response
relative to specified standards such as gas cells or calibration
gases, or relative to independent effluent measurements.
Appendix F, Procedure 1 requires that sources have written
procedures for CEMS calibration. Sources may develop their own
written procedures; alternatively, they may specify applicable
sections of the instrument manual as their written procedures.
There are no currently promulgated regulations that require
either specific calibration frequencies or specific criteria for
initiating calibration procedures. Sources may therefore choose
their own frequency or criteria for calibration based on
operating experience or manufacturer's recommendations.
-------
Section No. 3.0.10
Date November 26, 1985
Page 2
START
DAILY DRIFT CHECK
PERFORM DAILY
CAL. DRIFT
DETERMINATION
5TH
N / CONS.
DAY CD >
(2MSPEC)
PASS
DECLARE CEMS OUT-
OF-CONTROL
BEGINNING AT END
OF 5TH 24-HR.
PERIOD
DECLARE CEMS
OUT-OF-CONTROL
BEGINNING AT END
OF PREVIOUS
SUCCESSFUL CD
CHECK
PERFORM
CORRECTIVE ACTION
REPEAT CD CHECK;
WAIT UNTIL TIME
FOR NEXT DAILY
CD CHECK
PERFORM
CORRECTIVE ACTION
REPEAT CD CHECK;
WAIT UNTIL TIME
FOR NEXT DAILY
CD CHECK
COLLECT DATA
FOR DAR
DECLARE CEMS BACK IN CONTROL
BEGINNING AT COMPLETION OF
LAST SUCCESSFUL CD CHECK
INVALIDATE DATA
FROM BEGINNING TO
END OF O.O.C. PERIOD.
DESCRIBE CORRECTIVE
.ACTION FOR DAR
Figure 1. Flow Chart for Required QC Procedures.
-------
Section No. 3.0.10
Date November 26, 1985
Page 3
WAIT UNTIL TIME
FOR NEXT DAILY
CAL DRIFT CHECK
TIME
FOR REQUIRED
PREVENTIVE
MAINTENANCE
TIME
FOR OTRLY
ACCY. CK.
1
CONDUCT REQUIRED
PREVENTIVE
MAINTENANCE
QUARTERLY
ACCURACY
AUDIT
RECORD ACCURACY RESULTS FOR DAR. IF CEMS
FAILED INITIAL ACCURACY TEST THEN:
(A) DECLARE CEMS O.O.C. AND INVALIDATE
CEMS DATA FROM END OF SAMPLING FOR
INITIAL TEST UNTIL END OF SAMPLING
FOR SUCCESSFUL ACCURACY TEST
(B) REPORT DATA FROM INITIAL AND
SUCCESSFUL ACCURACY TESTS
(C) DESCRIBE CORRECTIVE ACTION
CHANGE QAP
OR
REPAIR CEMS
OR
REPLACE CEMS
2ND
CONSECUTIVE
QTR. WITH
UNACCEPTABLE
ACCY. RESULTS
CODE:
£ - "EQUAL TO OR LESS THAN"
> . "GREATER THAN"
CD - CALIBRATION DRIFT
CONS - CONSECUTIVE
DAR - DATA ASSESSMENT REPORT
N - NO
O.O.C. - OUT-OF-CONTROL
QCP - QUALITY CONTROL PROCEDURES
SPEC - DRIFT LIMITS IN PERFORMANCE
SPECIFICATIONS 2 OR 3
Y . YES
Figure 1. (continued)
-------
Section No. 3.0.10
Date November 26, 1985
Page 4
For calibrations based on external gas cells, sufficient time
should be allowed for the cell and/or analyzer cabinet to reach
normal operating temperature; accordingly, it is recommended that
procedures be incorporated into the QC program that ensure
sufficient time for the monitor response to stabilize before it
is compared to the cell's named value. Some in-situ analyzers
partially or totally disable temperature compensation circuitry
during cell-type calibrations. In these cases, it is recommended
that additional procedures addressing the calibration of this
circuitry be incorporated into the QC program.
For analyzers calibrated using calibration gases as the
reference, the written procedures should specify (1) at what
point in the sampling system the calibration gases are to be
introduced and (2) either the specific gas flow rate to be used
or how the flow rate is determined. Although current continuous
emission monitoring (CEM) regulations do not require establishing
the traceability of calibration gases to higher standards, it is
strongly recommended that procedures be established and included
within the QC program for verifying the concentrations of cali-
bration gases. One acceptable procedure is EPA's traceability
protocol 1 (Reference 1).
In cases where a portable CEMS is to be used as the reference
for adjusting the installed CEMS, written procedures should
specify calibration and operating procedures for the portable
CEMS, including the portable CEMS sampling location.
The written calibration procedures for the installed CEMS may
be incorporated into one or more of the following sections of a
QC program:
1. A Stand-Alone "Calibration" Section of the QC Program.
In this case, the frequency of calibration or the
criteria for initiating calibration activities should be
clearly specified.
2. Preventive Maintenance. Within the section delineating
the preventive maintenance procedures, calibration may
be specified as a routine maintenance activity to be
performed at regular, specified intervals. Alterna-
tively, calibration may be specified on an as-needed
basis with stated criteria for the implementation of
calibration activities.
3. Corrective Action. Calibration procedures may be
included within the section delineating corrective
action activities to be performed at the discretion of
CEMS repair personnel in response to an out-of-control
CEMS.
Regardless of how the calibration procedures are incorporated
into the QC program; it is recommended that the individual or
-------
Section No. 3.0.10
Date November 26, 1985
Page 5
group responsible for GEMS calibration be identified within the
written QC plan.
10.3 Calibration Drift and Adjustment of the CEMS
Calibration drift (CD) refers to the difference between the
CEMS output reading and a reference value after a period of
operation during which no unscheduled maintenance, repair, or
adjustment took place. Daily zero (or low level) and span drift
checks are required by 40 CFR 60.13; these checks are to be used
to fulfill the calibration drift check requirement of Appendix F,
Procedure 1. Appendix F, Procedure 1 requires written procedures
that specify how the zero (or low level) and span calibration
drift determinations are to be performed. These procedures must
be consistent with the monitor vendor's prescribed method for
checking CD.
Table 10.1 presents CD criteria and the corresponding
required source responses. Sources may choose to establish more
stringent criteria for adjustment of CEMS for zero (or low level)
and/or span calibration drift. It is recommended that the CD
criteria selected for adjustment of the CEMS be incorporated into
the written instructions for the calibration drift check
procedures, so that the need for adjustment based on calibration
drift may be determined immediately.
Corrections for excessive drift may consist of any
adjustments or activities that the operator or technician deems
necessary to correct for the observed drift. These activities
typically consist of routine checks and adjustments of
calibration gas flow rates and pressures, verification of proper
sample cell temperatures, verification of the status of monitor
specific auxiliary monitoring parameters, and adjustment of zero
and/or span potentiometers. Written procedures should be
available for performing these routine activities and should
include criteria for determining that adjustments have been
successful.
10.4 Preventive Maintenance of the CEMS
Preventive maintenance is comprised of activities designed to
detect and prevent the development of monitoring problems. These
activities typically include both routine maintenance procedures
and maintenance, repairs, or adjustments performed on an as-
needed basis. An example of as-needed preventive maintenance
would be the repairing of the protective covering of an
extractive sample line following damage resulting from an
accident during the construction activities. If the sample line
itself were not damaged, the repair would be considered
preventive maintenance and would not consititute corrective
action for a malfunctioning CEMS. The importance of this type of
-------
Section No. 3.0.10
Date November 26, 1985
Page 6
TABLE 10.1. CEMS CALIBRATION DRIFT CRITERIA
Parameter
Criterion*
Action Required
Zero (or low)
level cali-
bration drift
CD > 2 x (Spec)**
CD > 2 x (Spec) for
5 consecutive 24-hour
periods
CD > 4 x (Spec)
Adjust CEMS for calibra-
tion drift
CEMS out-of-control period
begins at end of 5th day
the CD exceeds 2 x (Spec);
perform corrective action
and repeat CD check
CEMS out-of-control period
begins at the time corres-
ponding to the completion
of the last acceptable CD
check preceding the CD
check which exceeds
4 x (Spec); perform
corrective action and
repeat the CD check
Span cali-
bration drift
CD > 2 x (Spec)**
CD > 2 x (Spec) for
5 consecutive 24-hour
periods
CD > 4 x (Spec)
Adjust CEMS for calibration
drift
CEMS out-of-control period
begins at end of 5th day
the CD exceeds 2 x (Spec);
perform corrective action
and repeat CD check
CEMS out-of-control period
begins at the time corres-
ponding to the completion of
the last successful CD check
preceding the CD check that
exceeds 4 x (Spec); perform
corrective action and repeat
the CD check
*Spec refers to the applicable performance specification in
Appendix B.
**This is the minimum criterion for adjustment of the CEMS. More
stringent criteria, which may be preferred by many sources, are
also acceptable.
-------
Section No. 3.0.10
Date November 26, 1985
Page 7
maintenance is recognized; however, it is neither practical nor
necessary to develop written procedures for such needed
activities.
Written procedures must be available for routine maintenance
activities. These procedures should specify what procedures are
to be conducted and the frequency with which the various
activities are to be performed. The QC program should specify
the individual or office responsible for ensuring that the
preventive maintenance procedures are conducted at the
appropriate frequencies and the individual or group who will
perform the actual routine maintenance procedures.
The applicable regulations do not specify the minimum level
of routine preventive maintenance. It is suggested that, at a
minimum, the initial procedures should incorporate the vendor's
recommendations regarding preventive maintenance activities and
frequencies. These procedures may later be adjusted to reflect
actual operating experience with individual CEMS installations.
A list of spare parts for the CEMS must be included in the
written QC plan. At a minimum, those spare parts recommended by
the monitor vendor should be available. The QC program should
specify the individual or office who is responsible for
maintaining the listed spare parts inventory.
10.5 Data Records, Calculations, and Reporting for the CEMS
The QA/QC program must address recordkeeping, calculations,
and reporting of emissions and quality assurance data. The
requirements for these activities are contained in the subparts
of 40 CFR 60 that specify the use of CEM. A Data Assessment
Report (DAR) must be provided with emissions reports required by
the applicable subpart of 40 CFR 60. The DAR must contain, at a
minimum:
1. The name and address of the source owner or operator.
2. Identification and location of each monitor in the CEMS.
3. The manufacturer and model number of each monitor in the
CEMS.
4. Quarterly accuracy results, including dates, CEMS
responses, and either reference method results or
certified gas values; if either a RATA or a RAA was
performed, the results from the EPA performance audit
sample analysis must also be included.
5. A summary of corrective actions taken when the monitor
was determined to be out-of-control.
-------
Section No. 3.0.10
Date November 26, 1985
Page 8
For emissions data, a list or diagram should be provided
indicating the offices or individuals responsible for (1)
retrieving the data from the CEMS, (2) calculating emissions
rates from the CEMS data, (3) compiling emissions reports, and
(4) reviewing and/or approving emissions reports. Formulas and
example calculations should be provided for emission rate
calculations. Similar information should be provided for
emissions data from alternative monitoring methods that may be
necessary during CEMS out-of-control periods.
A list or diagram should also be provided indicating the
offices or individuals responsible for (1) collecting quality
assurance (QA) data, (2) performing applicable calculations of
QA/QC results, (3) recording the QA/QC results in appropriate
logs (as applicable), (4) preparing the DAR, and (5) approving
and/or reviewing the DAR. Formulas and example calculations
should be provided for all required QA data calculations.
10.6 Accuracy Audit Procedures Including Sampling and Analysis
Methods
Appendix F, Procedure 1 requires that each CEMS be audited at
least once each calendar quarter. Three audit techniques are
acceptable:
1. Relative accuracy test audits (RATA's);
2. Cylinder Gas Audits (CGA's); and
3. Relative accuracy audits (RAA's).
In addition, other alternative audit procedures may be used as
approved by the Administrator.
If the CEMS does not demonstrate acceptable accuracy during
the quarterly audit, then corrective actions must be initiated,
and the CEMS must be declared out-of-control from the time cor-
responding to the completion of the sampling for the unsuccessful
audit until the completion of the sampling for a successful
follow-up audit. If the CEMS demonstrates unacceptable accuracy
for two consecutive quarters, then the QA program must be
revised, or the CEMS must be modified or replaced.
Table 10.2 presents the specific requirements and the
corresponding CEMS performance criteria for each of the three
acceptable audit techniques.
The QC program must include written sampling and analysis
procedures to be used during the required quarterly accuracy
audits. At a minimum, these procedures must describe the methods
to be used to conduct a RATA. Applicable sections of Appendix A
(Reference Methods) and Appendix B (Performance Specifications)
-------
Section No. 3.0.10
Date November 26, 1985
Page 9
TABLE 10.2. REQUIREMENTS AND CRITERIA FOR
APPENDIX F, PROCEDURE 1 AUDIT TECHNIQUES
Technique
Requirements
Performance Criteria
RATA
RAA
CGA
Conduct as per
applicable performance
specification (PS) in
Appendix B (e.g., PS 2
for S09 and NO )
£» Ji
Analyze appropriate
performance audit
samples from EPA
RA must not exceed 20% or
10% of applicable standard,
whichever is greater
For SO,
standards from
Conduct as per appli-
cable PS in Appendix
B except only 3 runs
are required
Use relative difference
between the mean
reference method values
and the mean of the CEMS
responses to assess the
accuracy of the CEMS data
Challenge both pollu-
tant and diluent chan- H
nels (if applicable)
of CEMS three times at
the two points specified
in Procedure 1
Use gases that have been
certified by comparison
to NBS SRM's or NBS/EPA
approved gas manufacturer's
CRM's
Operate analyzer in normal
sampling mode
Use average difference between
actual gas value and concentra-
tion indicated by CEMS to
access accuracy
0.20 to 0.30 Ib/lO" Btu,
RA must not exceed 15% of
the standard
For S09 standards below
0.20 IB/10 Btu, RA must
not exceed 20% of the
standard
Inaccuracy must not exceed
+ 15% or 7.5% of the appli-
cable standard, whichever
is greater
Inaccuracy must not exceed
+ 15%
-------
Section No. 3.0.10
Date November 26, 1985
Page 10
may be cited where possible to describe audit procedures. The
written procedures should specify individuals or groups respon-
sible for audit program oversight, sampling, analysis, and accu-
acy assessment calculations. If the source chooses to conduct
RAA's and/or CGA's during quarters when RATA's are not required,
the QC plan should include written procedures for these audit
techniques. Again, applicable sections of Appendix A, Appen-
dix B, and/or instrument operation manuals may be cited where
possible.
Sources may choose to have an outside contractor perform some
or all of the accuracy audit activities. Since contractor selec-
tion may be subject to competitive bidding, the QC program need
not specify a particular contractor. However, the specific
activities for which the contractor will be responsible should be
listed.
10.7 Program of Corrective Action for the Malfunctioning CEMS
Appendix F, Procedure 1 specifies that corrective action must
be performed when a CEMS is out-of-control. Appropriate
corrective action will depend on the nature of the CEMS
malfunction. At a minimum, written procedures must be available,
to be applied as necessary, for instrument start-up and trouble
shooting. Appropriate sections of instrument operation and/or
repair manuals may be referenced - to fulfill this requirement.
Where possible, it is recommended that additional quality
assessment procedures be provided to verify proper operation of
the CEMS following repair or adjustment.
A list should be provided to indicate what alternative
methods are to be used for monitoring emissions during CEMS
out-of-control periods in order to fulfill the minimum data
availability requirements of the applicable subpart. Written
procedures should be available for operation of these alternative
methods.
A list or chart should be provided to indicate the offices or
individuals (1) to be contacted when a CEMS out-of-control period
occurs, (2) to approve the corrective action (if applicable), and
(3 j to be responsible for determining when alternative monitoring
procedures are to be employed. Criteria should be provided for
determining when the CEMS is out-of-control. As a minimum, these
must include the Appendix F, Procedure I criteria for excessive
drift and excessive inaccuracy.
10.8 References
1. Traceability Protocol for Establishing True Concentrations
of Gases Used for Calibration and Audits of Continuous
Source Emission Monitors (Protocol Number 1). June 1978,
Section 3.0.4 of the Quality Assurance Handbook for Air
-------
Section No. 3.0.10
Date November 26, 1985
Page 11
Pollution Measurement Systems, Volume, ill, Stationary
Source Specific Methods. EPA-600/4-77-027b. August 1977.
U. S. Environmental Protection Agency, Office of Research
and Development Publications, 26 West St. Clair Street,
Cincinnati, Ohio 45268.
-------
Section No. 3.14
Date July 1, 1986
Page 1
Section 3.14
METHOD 7A - DETERMINATION OF NITROGEN OXIDE
EMISSIONS FROM STATIONARY SOURCES
(Grab Sampling - Ion Chromatographic Method)
OUTLINE
Number of
Section Documentation pages
SUMMARY 3.14 2
METHOD HIGHLIGHTS 3.14 8
METHOD DESCRIPTION
1. PROCUREMENT OF APPARATUS
AND SUPPLIES 3.14.1 10
2. CALIBRATION OF APPARATUS 3.14.2 14
3. PRESAMPLING OPERATIONS 3.14.3 6
4. ON-SITE MEASUREMENTS 3.14.4 7
5. POSTSAMPLING OPERATIONS 3.14.5 11
6. CALCULATIONS 3.14.6 6
7. MAINTENANCE 3.14.7 2
8. AUDITING PROCEDURES 3.14.8 6
9. RECOMMENDED STANDARDS FOR
ESTABLISHING TRACEABILITY 3.14.9 1
10. REFERENCE METHOD 3.14.10 3
11. REFERENCES 3.14.11 2
12. DATA FORMS 3.14.12 12
-------
Section No. 3.14
Date July 1, 1986
Page 2
SUMMARY
A gas sample is extracted from the sampling point in the stack.
The sample is collected in an evacuated 2-liter round bottom boro-
silicate flask containing 25 ml of dilute sulfuric acid-hydrogen
peroxide absorbing reagent. The nitrogen oxides, NO and NO^, react
with the absorbing reagent to form nitrate ion which is analyzed by
ion chromatography (1C). The method does not respond to nitrous
oxide, N20.
The reactions that describe absorption of the NO are distinct
for NO and N02._ The common feature of the reactions isxthe formation
of nitrate, N03~, as nitric acid, HN03.
The absorption of NO involves an oxidation-reduction reaction
where the oxidizing agent is the acidic hydrogen peroxide solution.
The two half reactions are:
3H2°2 + 6H+ + 6e~ = 6H2°
and
2NO + 4H-0 = 2NO ~ + 9H+ + 6e~;
£ O
and the overall reaction is:
• — +
O KT^ _i_ O U f\ _ O VT/^ _i_ O tl _i_ O U i
zNO -f on—O-t = ^JNO— + 2.n + 2H~<
The absorption of N02 presumably involves the reaction with water
to form nitric acid and NO. NO, reacts with water to form nitric
acid and nitrous acid, HN02:
2N02 + H20 = HN03 + HN02.
The nitrous acid is unstable and decomposes:
3HNO2 - 2NO + HN03 + H20.
The observed reaction is the sum of the two reactions above:
3NO2 + H2O = 2HN03 + NO.
Absorption of N02 proceeds faster than absorption of NO because
N0« is more soluble in solution, where reaction occurs. In this
respect, it should be noted that absorption of NO is quickened as a
consequence of reaction with oxygen also present within the flask:
2ND + 02 = 2N02.
If the gas being sampled contains insufficient oxygen for the
conversion of NO to N02, then oxygen should be introduced into the
flask by one of three methods: (1) before evacuating the sampling
-------
Section No. 3.14
Date July 1, 1986
Page 3
flask, flush it with pure cylinder oxygen, and then evacuate the
flask to 75 mm (3.0 in.) Hg absolute pressure or less; or (2) inject
oxygen into the flask after sampling; or (3) terminate sampling with
a minimum of 50 mm (2.0 in.) Hg vacuum remaining in the flask, record
this final pressure, and then open the flask to the atmosphere until
the flask pressure is almost equal to atmospheric pressure.
Method 7A is applicable to the measurement of nitrogen oxides
emitted from stationary sources. It may be used as an alternative to
Method 7 (as defined in 40 CFR Part 60.8(b)) to determine compliance
if the stack concentration is within the analytical range. The
analytical range of the method is from 125 to 1250 mg NO , expressed
as NO?' per drv standard cubic meter (65 to 655 ppm). Higher
concentrations may be analyzed by diluting the sample. The lower
detection limit is approximately 19 mg/m (10 ppm), but may vary
among instruments.
The method description which follows is based on the method that
was promulgated on'December 8, 1983.
Section 3.14.10 contains a copy of Method 7A, and blank data
forms are provided in Section 3.14.12 for the convenience of the
Handbook user.
Note: Because of similarities between Method 7A and Method 7
sampling equipment and procedures, in most cases only the differences
in Method 7A are presented in detail in this section (3.14). How-
ever, all tasks are shown in the activity matrices and data sheets
needed to perform Method 7A are included, whether or not differences
occur in the written descriptions. Other Method 7A procedures are
referenced to the corresponding description in Section 3.6, Method
7. This is done for both time savings to the reader and cost savings
to the Government.
-------
Section No. 3.14
Date July 1, 1986
Page 4
METHOD HIGHLIGHTS
Section 3.14 (Method 7A) describes the required procedure for
sampling and analyzing of nitrogen oxides emissions from stationary
sources. For the method, a grab sample is extracted from a point in
the stack, and collected in a previously evacuated flask containing a
sulfuric acid-hydrogen peroxide absorbing solution. With the
exception of nitrous oxide, the nitrogen oxides are oxidized to
nitrate which is analyzed by ion chromatography (1C). Results are
expressed as concentrations of nitrogen dioxide (N02). The
applicable regulation should be consulted to determine whether
additional measurements, such as velocity or CU determinations, are
required.
The absorbing reagent for EPA Method 7A has a sulfuric acid con-
centration one-tenth that of EPA Method 7. In all other respects,
the sampling train and sampling procedures of EPA Method 7A are iden-
tical to those of EPA Method 7. Sample preparation involves only
dilution to reach a measurable concentration range for the ion
chromatograph.
Ion chromatography is a relatively recent analytical develop-
ment. The reader is referred,2to the literature for detailed
descriptions of the subject. Small, et al., developed the
technique using the principles of ion exchange chromatography and
conductimetric detection. Previous attempts to use this type of
detection were unsuccessful because of the presence of the background
electrolyte used for elution of the ionic species. Small, et al.,
used a novel combination of resins to separate the ions of interest
and neutralize the eluent from the background.
The aqueous sample is introduced into a fixed-volume sample loop
by using a plastic syringe. Once injected, the sample is carried
through a separation column at different rates according to their
relative affinities for the resin and the eluent and are therefore
separated into discrete bands. The separated ions are then passed
through a post-separation suppressor device, a source of hydrogen ion
(H ), which converts the eluent ions into a less conducting weak acid
while converting the analyte ions into a highly conducting form.
This permits the use of a conductivity cell as a very sensitive
detector of all ionic species.
Gjerde, et al., described a modified ion chromatographic method
that eliminates the need for a suppressor device. Anions are
separated on a column containing an anion-exchange resin with a low
exchange capacity. Because of the low capacity, a very dilute
solution of an aromatic organic acid salt may be used as the eluent.
The conductance of the eluent is sufficiently low that no suppression
is needed.
For Method 7A, either suppressed or non-suppressed 1C may be
used. The basic ion chromatograph will have the following
components:
-------
Section No. 3.14
Date July 1, 1986
Page 5
(a) sample injection device,
(b) anion separation column,
(c) anion suppressor column, either packed bed or fiber
type (not required for non-suppressed 1C),
(d) conductivity detector, and
(e) recorder.
Two critical aspects of Method 7A are (a) the measurement of the
gaseous sample volume, and (b) the preparation of the calibration
standards for the ion chromatograph. Analysts are advised to observe
specified procedures carefully at these points of the method.
Analysts performing the method should be well trained in the use of
the ion chromatograph.
Collaborative testing of EPA Method 7A has not been performed.
However, from a technical standpoint, it can be expected that EPA
Method 7A will exhibit accuracy and precision as good as, if not
better than, EPA Method 7.
The four blank data forms at the end of this section may be
removed from the Handbook and used in the pretest, test, and posttest,
operations. Each form has a subtitle (e.g., Method 7A, Figure 3.1)
to assist the user in finding a similar completed form in the method
description (Section 3.14.3). On the blank and filled-in forms, the
items/parameters that can cause the most significant errors are
designated with an asterisk.
1. Procurement of Apparatus and Supplies
Section 3.14.1 (Procurement of Apparatus and Supplies) gives
specifications, criteria, and design features for the required
equipment and materials. The sampling apparatus for Method 7A has
the same design features as that of Method 7. Section 3.14.1 can be
used as a guide for procurement and initial checks of equipment and
supplies. The activity matrix (Table 1.1) at the end of the section
is a summary of the details given in the text and can be used as a
quick reference.
2. Pretest Preparations
Section 3.14.2 (Calibration of Apparatus) addresses the required
calibration procedures and considerations for the Method 7A sampling
equipment (same as Method 7) and analytical equipment (the ion chrom-
atograph). Required accuracies for each component are also included.
A pretest sampling checklist (Figure 3.1 in Section 3.14.3) or a
similar form should be used to summarize the calibration and other
pertinent pretest data. The volume of each collection flask must be
determined with stopcock in place. This volume measurement is
required only on the initial calibration, provided the stopcock is
not changed. The calibration section may be removed along with the
-------
Section No. 3.14
Date July 1, 1986
Page 6
corresponding sections from the other methods and made into a
separate quality assurance reference manual for use by personnel
involved in calibration activities.
Section 3.14.3 (Presampling Operations) provides the tester with
a guide for equipment and supplies preparation for the field test.
With the exception of the preparation of certain reagents, these are
the same as for Method 7. A pretest preparation form (Figure 3.2,
Section 3.14.3) can be used as an equipment checkout and packing
list. The flasks may be charged with the absorbing reagent in the
base laboratory. The method of packing and the use of the described
packing containers should help protect the equipment, but neither is
required by Method 7A.
Activity matrices for the calibration of equipment and the pre-
sampling operations (Tables 2.2 and 3.1) summarize the activities.
3. On-Site Measurements
Section 3.14.4 (On-Site Measurements) contains step-by-step
procedures for sample collection and for sample recovery. Sample
collections are the same as for Method 7; sample recovery proce-
dures differ slightly from Method 7 in that the sample pH does not
have to be checked and adjusted. The on-site checklist (Figure 4.3,
Section 3.14.4) provides the tester with a quick method of checking
the on-site requirements. When high negative stack pressures are
present, extra care should be taken to purge the leak-tested sample
system and to be sure the flask is £ 75 mm (3 in.) Hg absolute
pressure prior to testing. Also, the 16-hour sample residence time
in the flask must be observed* Table 4.1 provides an activity
matrix for all on-site activities.
4. Posttest Operations
Section3.14.5TPostsampling Operations) gives the posttest
equipment procedures and a step-by-step analytical procedure for
determination of NO , expressed as N02- Posttest calibration is not
required on any of the sampling equipment. The posttest operations
form (Figure 5.1, Section 3.14.5) provides some key parameters to be
checked by the tester and laboratory personnel. The step-by-step
analytical procedure description can be removed and made into a
separate quality assurance analytical reference manual for the
laboratory personnel. Analysis of calibration standards is conducted
in conjunction with the analysis of the field samples. Strict
adherence to Method 7A analytical procedures must be observed.
Section 3.14.6 (Calculations) provides the tester with the
required equations, nomenclature, and significant digits. It is
suggested that a calculator be used, if available, to reduce the
chances of calculation error.
Section 3.14.7 (Maintenance) provides the tester with a guide for
a maintenance program. This program is not required, but should
reduce equipment malfunctions. Activity matrices (Tables 5.1, 6.1,
-------
Section No. 3.14
Date July 1, 1986
Page 7
and 7.1) summarize all postsampling, calculation, and maintenance
activities.
5. Auditing Procedure
Section 3.14.8 (Auditing Procedure) provides a description of
necessary activities for conducting performance and system audits.
When Method 7A is used to demonstrate compliance with an EPA poll-
utant emission standard, a performance audit is required to be
conducted of the analytical phase of the method. The data processing
procedures and a checklist for a systems audit are also included in
this section. Table 8.1 is an activity matrix for conducting the
performance and system audits.
Section 3.14.9 (Recommended Standards for Establishing
Traceability) provides the primary standard to which the analysis
data should be traceable.
6. References
Section 3.14.10 contains the promulgated Method 7A; Section
3.14.11 contains the references cited throughout the text; and
Section 3.14.12 contains copies of data forms recommended for Method
7A.
-------
Section No. 3.14
Date July 1, 1986
Page 8
PRETEST SAMPLING CHECKS
(Method 7A, Figure 3.1)
Date Calibrated by
Flask Volume
Flask volumes measured with valves? yes no
Volume measured within 10 ml of actual volume?* yes no
Temperature Gauge
Was a pretest temperature correction used? yes no
If yes, temperature correction (within 1°C (2°F)
of reference values for calibration and within +_ 2°C
(4°F) of reference values for calibration check).
Vacuum Gauge
Was gauge calibrated against a U-tube mercury manometer (if it
was a mechanical gauge)?* yes no not applicable
Barometer
Was the pretest field barometer reading within 2.5 mm (0.1 in.) Hg
of the mercury-in-glass barometer? yes no
*Most significant items/parameters to be checked.
-------
Section No. 3.14
Date July 1, 1986
Page 9
PRETEST PREPARATIONS
(Method 7A, Figure 3.2)
Apparatus check
Probe
Glass liner
clean
Heated properly*
Leak checked
Collection Flask
Clean
Leak checked
Temperature
gauge
Evacuation System
Leak- free pumps
Manifold and
tubing
U-tube manometer
Barometer
Reagents
Water
Absorbing solu-
tion*
Sample Recovery
Dropper or burette
Sample bottles
Pipette, 25-ml
Acceptable
Yes
No
Quantity
required
Ready
Yes
No
Loaded
and packed
Yes
No
*Most significant items/parameters to be checked.
-------
Section No. 3.14
Date July 1, 1986
Page 10
ON-SITE MEASUREMENTS
(Method 7A, Figure 4.3)
Sampling
Volume of 25 ml of absorbing solution** placed in flask?
Flask valve stopper in purge position?
Sampling train properly assembled?
Leak free?* Stopcock grease used?
Type?
Flask evacuated to £75 mm (3 in.) Hg pressure?
Leakage from manometer observation?*
(e.g., maximum change in manometer of £10 mm (0.4 in.)
Hg/min) .
Initial pressure of flask recorded?*
Initial temperature of flask recorded?
Probe purged before sampling?
Sample collected properly?*
Flask shaken for 5 min after collection and disassembly from
train?*
Samples properly labeled and sealed and stored for shipment?
Sample Recovery
Samples allowed to remain in flasks for minimum of 16 h?*
Final flask temperature and pressure recorded?*
Sample transferred to leak-free polyethylene bottle?
Flask rinsed twice with 5-ml portions of water and rinse
added to bottle containing sample?
* Most significant items/parameters to be checked.
** Note that absprbing solution for Method 7A is different from
that of Method 7.
-------
Section No. 3.14
Date July 1, 1986
Page 11
POSTTEST OPERATIONS
(Method 7A, Figure 5.1)
Reagents
Sodium nitrate dried at 105° to 110°C for a minimum of 2 hours
before use?
Stock standard solution (sodium nitrate) less than 1 month old?
Sample Preparation
Has liquid level noticeably changed?*
Original volume Corrected volume
Analysis
Standard calibration curve prepared?*
All calibration points within 7 percent of linear calibration
curve?*
Reagent blanks made from absorbing solution or eluent solution?
Same injection volume for both standards and samples?
Duplicate sample values agree within 5 percent of their mean?
All analytical data recorded on checklist and laboratory form?
* Most significant items/parameters to be checked.
-------
Section No. 3.14.1
Date July 1, 1986
Page 1
1.0 PROCUREMENT OF APPARATUS AND SUPPLIES
A schematic of the sampling train used for Method 7A is shown in
Figure 1.1. The train and sampling procedures are identical to those
for Method 7. The sample recovery procedures and equipment are also
identical, with the exception that there is no need to check and
adjust the pH of the samples. The analytical procedures and
equipment involved are different.
Specifications, criteria, and/or design features are given in
this section to aid in the selection of equipment or any components
that are different from those in Section 3.6.1. Procedures and
limits (where applicable) for acceptance checks are also given.
Alternative grab sampling systems or equipment capable of measuring
sample volume to within 2% and collecting a sufficient sample volume
to allow analytical repeatability to within 5% may be acceptable,
subject to approval.
During the procurement of equipment and supplies, it is suggested
that a procurement log be used to record the descriptive title of the
equipment, identification number (if applicable), and the results of
acceptance checks. An example of a procurement log is shown in Fig-
ure 1.2. A blank copy of this form is provided in Section 3.14.12
for the convenience of the Handbook user. Calibration data generated
in the acceptance check are to be recorded in the calibration log
book.
The following equipment is that which is specified in Method 7A
and has not already been described in Section 3.6.1 for Method 7.
Table 1.1 at the end of this section summarizes quality assurance
activities for the procurement and acceptance of all apparatus and
supplies for Method 7A including the equipment described in Section
3.6.1.
1.1 Analysis
For the analysis, the following equipment is needed. Alternative
instrumentation (and corresponding procedures) will be allowed,
provided the calibration precision discussed in Section 3.14.2 and
acceptable accuracy can be met.
1.1.1 Volumetric Pipets - Class-A volumetric pipets are required.
For making up the calibration standards, pipets of the following
sizes are needed: one 1-ml, one 2-ml, one 4-ml, one 6-ml, and one
10-ml. Enough 5-ml pipets are needed for preparing calibration
standards, blanks, and samples.
1.1.2 Volumetric Flasks - Two Class-A 50-ml volumetric flasks are
needed for each sample, and one Class-A 50-ml volumetric flask is
needed for each standard and each blank. Also required are Class-A
200-ml and Class-A 1000-ml sizes. Additional volumetric flasks
(50-ml) may be required for audit samples and for dilution of samples
having concentrations in excess of the highest standard.
-------
PROBE
FILTER
FLASK VALVE
FLASK
FLASK SHIELOu .\
SQUEEZE BULB
MP VALVE
PUMP
THERMOMETER
Figure 1.1. Method 7A evacuated flask sampling train.
•o o w
0> 0> fl>
«Q ft O
-------
Item description
Qty-
Purchase
order
number
/OSS'
Vendor
Date
Ord.
i/ilesr
Rec.
I/IS/6S
Cost
Disposition
£.& /1 1<>*'&'£&£L '
f/2~ Z'/Q S"
Comments
Figure 1.2. Example of a procurement log.
13 a w
0) 0) 0)
IQ ft O
(D
-------
Section No. 3.14.1
Date July 1, 1986
Page 4
1.1.3 Analytical Balance - One analytical balance that weighs to 0.1
mg and a set of Class-S calibration weights to check the accuracy of
the balance (+_ 0.3 mg) upon receipt are needed. The balance should
be serviced or returned to the manufacturer if agreement cannot be
met.
1.1.4 Ion Chromatograph - The ion chromatograph should, at a mini-
mum, have the components described below.
Sample Injection Device - This device must be capable of deliver-
ing a reproducible volume of sample to the ion chromatograph.
Columns - The ion chromatograph should have an anion separator
column capable of giving duplicate results within 5 percent of mean
value and of resolving the nitrate ion from sulfate ion and from
other species present. Both the Dionex HPIC-ASC fast run anion
column for suppressed 1C and the Wescan 269-029 Anion/R Column for
non-suppressed 1C have been demonstrated to give acceptable separ-
ation. If suppressed 1C is to be used, an anion suppressor column is
required. The Dionex AFS anion fiber suppressor (recommended) or
ASC-1 general purpose suppressor may be used. Suppressor columns are
generally produced as proprietary items; however, one can be made in
the laboratory using the resin available from BioRad Company, 32nd
and Griffin Streets, Richmond, California.
Pump - The pump must be capable of maintaining a steady eluent
flow as required by the system.
Flow Gauges - These must be capable of measuring the specified
eluent flow rate. It is recommended that the gauge be calibrated
upon receipt.
Conductivity Detector with Temperature Compensation - It should
be capable of giving responses that can be integrated with a precis-
ion of +_ 5 percent. It is recommended that the detector be cali-
brated according to manufacturer's procedures prior to initial use.
Recorder - It should be compatible with the output voltage of the
detector.
1.2 Reagents
Unless otherwise indicated, it is intended that all reagents
conform to the specifications established by the Committee on
Analytical Reagents of the American Chemical Society (ACS), where
such specifications are available; otherwise, use the best grade
available.
1.2.1 Sampling - To prepare the absorbing solution, cautiously add
2.8 ml concentrated H2S04 to a 100-ml flask containing water (see
specifications in Subsection 1.2.3 below), and dilute to volume with
mixing. Add 10 ml of this solution, along with 6 ml of 3%
-------
Section No. 3.14.1
Date July 1, 1986
Page 5
hydrogen peroxide that has been freshly prepared from 30% hydrogen
peroxide, to a 1-liter flask. Dilute to volume with water (see
Subsection 1.2.3), and mix well. The absorbing solution must be used
within 1 week of its preparation and, if possible, within 24 hours.
Store in a dark-colored bottle. Do not expose to extreme heat or
direct sunlight. Refrigerate the 30% hydrogen peroxide solution.
Note; The H2S04 content of this absorbing solution is 10 times less
than that used for Method 7. The solution is prepared in this manner
to avoid interference from sulfate ions during the analysis by 1C.
1.2.2 Sample Recovery - Use ASTM D1193-82, Type III water (see
Subsection 1.2.3) for sample recovery and in making various
solutions. At the option of the analyst, the KMnO. test for
oxidizable organic matter may be omitted whenever high concentrations
of organic matter are not expected to be present.
1.2.3 Analysis - For the analysis, the following reagents are
required.
Water - Water should be used which conforms with ASTM specifi-
cation D1193-82, Type III. Type III water is prepared by distilla-
tion, ion exchange, reverse osmosis, or a combination thereof,
followed by polishing with a 0.45 ym membrane filter. The specifica-
tions for Type III water are shown below.
Specifications for ASTM D1193-82, Type III Water
Total matter, max., (mg/L) 1.0
Electrical conductivity, max., 1.0
(ymho/cm) at 25 C
Electrical resistivity, min., 1.0
(ymho/cm) at 25
pH at 25°C 6.2 to 7.5
Minimum color retention time 10
of KMnO., (min)
Maximum soluble silica, (vg/L) 10
Note; Mention of "water" anywhere in this Section (3.14) refers to
ASTM D1193-82, Type III water as described above. By using water
from the same source for making reagents, calibration standards, and
eluents for the ion chromatograph, the effects of trace quantities of
nitrate in the water will be negated with regard to sample analysis.
Therefore, a water blank correction is not necessary in the develop-
ment of the calibration curve.
-------
Section No. 3.14.1
Date July 1, 1986
Page 6
Sodium Nitrate - Dry an adequate amount of sodium nitrate
at 105 to 110 C for a minimum of 2 hours just prior to preparing tRe
standard solution. (The analyst should note that potassium nitrate,
KNOo, is used in EPA Method 7; KNO~ is an acceptable alternative for
Method 7A. ) J
Stock Standard Solution, 1 mg N0?/ml - To prepare, dissolve
exactly 1.847 g of dried NaNO3 (or 2.198 g of dried KNCU) in water,
and dilute to 1 liter in a volumetric flask; mix well. Tnis solution
is stable for 1 month and should not be used beyond this time.
The use of old solution may cause results to be biased high.
Solutions are readily contaminated by microorganisms that feed on
nitrate ion. Unquantified loss of nitrate ion from the standard
solution causes the high bias.
Working Standard Solution, 25 y g NO^/ml - Dilute 5 ml of the
standard solution to 200 ml with water in a volumetric flask, and mix
well.
Eluent Solution - Use an eluent appropriate to the column type
and capable of resolving nitrate ion from sulfate and other species
present. The following eluent s have been demonstrated to give
acceptable separation:
Suppressed 1C — 0.0024M Na2C03/0.003M NaHCXU. To prepare, weigh
1.018 g of sodium carbonate (Na^COg) and 1.008 g of sodium
bicarbonate (NaHC03), and dissolve in 4 liters of water.
Non-Suppressed 1C — 0.007M p-hydroxybenzoic acid, pH 8.4. To
prepare, weigh 3.867 g p-hydroxybenzoic acid, and dissolve in 4
liters of water. Adjust to pH 8.4 with lithium hydroxide.
Quality Assurance Audit Samples - Same as required by Method 7
(Section 3.6.8).
-------
Section No. 3.14.1
Date July 1, 1986
Page 7
TABLE 1.1. ACTIVITY MATRIX FOR PROCUREMENT OF APPARATUS
AND SUPPLIES
Apparatus/
supplies
Probe
Collection
flask
Flask valve
Temperature
gauge
Vacuum line
tubing
Vacuum gauge
Vacuum pump
Squeeze bulb
Volumetric
pipettes
Acceptance criteria
Borosilicate glass
stainless steel, or Tef-
lon tubing capable of
removing moisture
condensation
Two-liter borosilicate
glass round bottom, short
neck w/24/40 standard
taper opening
Borosilicate glass T-bore
stopcock w/24/40 standard
taper male joint (joint
connection to be made by
glassblower)
Dial- type, capable of
measuring from -5 to
+50°C within 1C
Capable of withstanding
75 mm absolute pressure
U-tube manometer, open
end, 1 m with 1-mm divi-
sions
Pump capable of pulling
vacuum of 75 """ Hg or
less
Rubber, one way
1-, 2-, 4-, 5-, 6-, 10-,
25-ml Class-A glass and
graduated 5~ml
Frequency and method
of measurement
Upon receipt, visually
check for cracks or
flaws and heating capa-
bility
Upon receipt, visually
check, and leak check
Visually check upon
receipt
Visually check upon
receipt, and compare
against Hg-in-glass
thermometer
Upon recer.pt, visually
check and leak check
Visually check upon
receipt
Upon receipt, check with
suitable pressure gauge
Visually check upon
receipt
As above
Action if
requirements
are not met
Return to sup-
plier, and
note in pro-
curement log
As above
As above
As above
As above
As above
As above
As above
As above
(continued)
-------
Section No. 3-1^.
Date July 1, 1986
Page 8
Table 1.1 (continued)
Apparatus/
supplies
Acceptance criteria
Frequency and method
of measurement
Action if
requirements
are not met
Stopcock
grease
High vacuum high temper-
ature chlorofluorocarbon
grease
As above
As above
Barometer (or
consult lo-
cal weather
station)
Capable of reading atmos-
pheric pressure to
+2.5 mm Hg
Visually check; cali-
brate against mercury-
in-glass barometer
As above
Storage bottle
Polyethylene, 100-ml, or
greater capacity, screw
cap
Visually check upon
receipt
Return to sup-
plier and note
in procurement
log
Wash bottle
Polyethylene or glass
Visually check label
upon receipt
As above
Analytical
balance
Capable of measuring
to +0.1 mg
Check with standard
weights upon receipt
and before each use
Replace or
return to man-
ufacturer
Volumetric
cylinders
50-ml (Class-A) with
1-ml divisions
As above
As above
Ion Chroma-
tograph
1. Columns
1. Capable of giving
nitrate ion peaks
with baseline
separation; capable of
giving duplicate results
within 5 percent of mean
value
1. Check during
analyses
1. Consult op-
erator's manu-
al ; regenerate
suppressor
column; clean
separator
column; check
performance
of components
below; replace
column(s) if
above actions
are unsuccess-
ful
(continued)
-------
Table 1.1 (continued)
Section No. 3.14.1
Date July 1, 1986
Page 9
Apparatus/
supplies
Acceptance criteria
Frequency and method
of measurement
Action if
requirements
are not met
2. Pump
2. Capable of delivering
eluent at constant and
repeatable flow rate
3. Flow
control
3. Capable of giving
repeatable indications
of eluent flow rate
4. Conduc-
tivity
detector
5. Recorder
4. Capable of giving
responses which can be
manually or electron-
ically integrated within
a precision of 5 percent
5. As above, if used to
record responses for
manual integration
2. Check during analyses
by monitoring flow rate
3- Check calibration
and repeatability upon
receipt
4. Calibrate according
to manufacturer's in-
structions prior to use
5. Check during
analyses
2. Consult oper-
ator's manual;
oil, clean, re-
repair, replace,
or return to man-
ufacturer; check
tubing of
ion chroma-
tograph for
leaks or ob-
structions;
check flow meter
performance
3. Consult oper-
ator's manual;
adjust, repair,
replace, or re-
turn to manu-
facturer ;
check pump per-
formance
4. Consult opera-
tor's manual;
Repair, replace,
or return to
manufacturer
5. Consult opera-
tor's manual;
adjust speed
Water
Meets ASTM D1193-82;
Type III
Check each lot, or
specify type when
ordering
Replace, or re-
turn to manu-
facturer
(continued)
-------
Table 1.1 (continued)
Section No. 3.14.1
Date July 1, 1986
Page 10
Apparatus/
supplies
Sulfuric
acid
Hydrogen
peroxide
Sodium nitrate
Sodium carbon-
ate
Sodium bicar-
bonate
p-Hydroxy-
benzoic acid
Acceptance criteria
Concentrated, ACS re-
agent grade
30# aqueous solution,
ACS reagent grade
(store refrigerated)
ACS reagent grade
ACS reagent grade
ACS reagent grade
ACS reagent grade
Frequency and method
of measurement
As above
As above
As above
As above
As above
As above
Action if
requirements
are not met
As above
As above
As above
As above
As above
As above
-------
Section No. 3.14.2
Date July 1, 1986
Page 1
2.0 CALIBRATION OF APPARATUS
Calibration of apparatus is one of the most important functions
in maintaining data quality. It is highly recommended that a labor-
atory log book of all calibrations be maintained. Calibration proce-
dures for the collection flasks, field barometer, thermometers,
vacuum gauge, and analytical balance used in Method 7A are the same
as those described for Method 7 (see Section 3.6.2) and are not
duplicated in this section; a form, however, for use in the analy-
tical balance calibration is shown in Figure 2.1. Detailed calibra-
tion procedures for the ion chromatograph system are described in
this section. Table 2.2 at the end of this section summarizes the
quality assurance activities for all calibrations in Method 7A
including those described in Section 3.6.2.
2.1 Ion Chromatograph System
For Method 7A, the calibration of the ion chromatograph (1C)
system, except for the initial calibration of the conductivity
detector, is conducted in conjunction with analysis of the field
samples. Specifically, the field samples are analyzed twice in
between three analyses of the ion chromatograph calibration stan-
dards; the exact sequence is discussed in detail in Section 3.14.5.
The three analyses of the calibration standards are used to prepare a
calibration curve that is used to determine a calibration factor for
calculating the concentration of nitrogen oxides in the field
samples. It is, however, highly recommended that the analyst conduct
a preliminary calibration of the 1C any time the system is set up for
analysis of NO field samples. For this reason, the full discus-
sion of the analysis of calibration standards and preparation of the
calibration curve is presented in this section. Also addressed in
this section are preliminary considerations in preparing the 1C
system for use and other considerations for ensuring quality data.
2.1.1 Preliminary Considerations
Conductivity Detector - Prior to its initial use, the conductiv-
ity detector of the ion chromatograph must be calibrated by the
method described in the operator's manual.
Recorder - A strip chart recorder compatible with the output
voltage range of the conductivity detector may be used to record the
ion chromatogram. Manual measurement techniques that can be used for
quantitation of the chromatogram include (a) peak height, (b) peak
area by triangulation, (c) peak area by multiplying peak height times
the peak width at half-height, (d) peak area by cutting out the peak
from the chromatogram and weighing it on an analytical balance, and
(e) peak area by planimetry.
The use of an electronic integrator, if available, is recommended
for greater accuracy and precision. The electronic integrator can be
used in the peak area mode when the integration parameters are set up
-------
Section No. 3.14.2
Date July 1, 1986
Page 2
Balance name
Number
Classification of standard weights
"£'
Date
/Z5&-
0.5000 g
(}. 5£0<4-
1.0000 g
10.000 g
50.0000 g
100.0000 g
/£>O • OOO 4-
Analyst
Figure 2.1. Analytical balance calibration form
-------
Section No. 3.14.2
Date July 1, 1986
Page 3
properly. The key integration parameters for peak area determination
concern the identification of the beginning and end of a peak and the
placement of the baseline under the peak. Analysts should carefully
read the operator's manual and understand the selection and set up of
the integration parameters for their particular integrator. The
electronic integrator can also be used in the peak height mode
provided that the peaks are symmetrical and an acceptable standard
calibration curve can be generated without any calibration point
deviating from the line by more than 7 percent (see Subsection 2.1.3
of this section).
Sample Injection Device Contamination Check - The analyst is
encouraged to check the sample injection device for contamination by
injecting water before the calibration standards are analyzed.
Contaminants will appear as peaks on the chromatogram. Repeated
injections of water should be used to remove contaminants from the
sample injection device. If certain peaks remain after several
injections of water then the water may be contaminated and should be
replaced.
Separation of Nitrate, NO^ - To ensure accurate results from the
ion chromatographic analysis, baseline separation of the nitrate ion
(N03~) peak from the other ion peaks should be achieved. For_Method
7A, the separation of the N03~ peak from the sulfate ion (S0.~) peak
is of major concern. The SO4~ originates primarily from the sulfuric
acid absorbing reagent. A second source of S04~ in a sample may be
sulfur dioxide present in the effluent stream sample. Figures 2.2a
and 2.2b show two chromatograms, one having overlapping N0~~ and SO."
peaks, and the other having baseline separation of the NO., and sof=
peaks. The sulfuric acid concentration in the absorbing reagent used
for Method 7A is 10 times less than that for Method 7 to minimize the
problem of adequately separating N03~ from C>0~.
The analyst is encouraged to check the performance of the ion
chromatograph system before_analyzing samples in order to ensure
baseline separation of N03~ is attainable. A test for baseline
separation of NO3 can be made by preparing a performance check
sample and analyzing during the recommended preliminary calibration
as follows:
1. Pipet 10.0 ml of the 25 yg N02/ml working standard solution
into a 50-ml volumetric flask.
2. Into the same volumetric flask, pipet 5 ml of absorbing
reagent.
3. Dilute with water to the mark.
4. Analyze this performance check sample with calibration
standards in the same manner as described for field samples (see
Subsections 5.1.4, 2.1.2, and 2.1.3).
-------
Section No. 3.14.2
Date July 1, 1986
Page 4
so,
Figure 2.2a. Example chromatogram having
overlapping peaks.
SO,
NO.
Figure 2.2b. Example chromatogram showing
baseline separations of peaks,
-------
Section No. 3.14.2
Date July 1, 1986
Page 5
The analyst should check the chromatogram of the performance
check sample for baseline separation. If the baseline separation is
marginal for the performance check sample and the samples have N0~~
concentrations close to that of the highest standard (5 ug N0?/ml7,
the analyst should closely monitor subsequent field sample
chromatograms to ensure that results are not adversely affected by
deterioration of the ion chromatograph column or varying performance
of the ion chromatograph.
The final aspect of the performance check involves a precision
assessment. The result from the analysis of the performance check
sample should agree within 5 percent of the value for the 5 ug N0~/ml
calibration standard data point.
2.1.2 Preparation of Calibration Standards - The preparation of the
calibration standards is perhaps the most critical aspect of the
Method 7A analysis, since the quality of sample results will only be
as good as the quality of the calibration. The steps observed in the
preparation of the calibration standards are detailed below.
Stock Standard Solution
1. Dry approximately 5 g ACS-grade sodium nitrate (NaNO~) in an
oven at 105 to 110 C for at least 2 hours prior to use.
Drying of the NaNOo is necessary to prevent NO results from
being biased high because of absorbed moisture.x
2. Calibrate the analytical balance using a 2-g Class-S calibra-
tion weight (see Figure 2.1 for an example form). The
balance reading should agree within 2 mg of the Class-S
calibration weight. Corrective actions should be taken if
this agreement is not achieved.
3. Allow the dried NaN03 to cool to room temperature in a desic-
cator. When the reagent has cooled, weigh out 1.847 g to
+0.002 g. Cooling is required to prevent weighing errors
originating from convection currents. Storage of the NaNO~
in the desiccator ensures that moisture will not be adsorbed.
4. Place weighed NaN03 in a 1-liter Class-A volumetric flask and
dissolve in exactly 1 liter of water. Label the flask
accordingly:
NaN03(aq)
StocR Standard
for EPA Method 7A
(1 mg N02/ml)
Date
Analyst's Initials
-------
Section No. 3.14.2
Date July 1, 1986
Page 6
The solution is stable for one month and should not be used
beyond that time. After about one month, there is increased
risk that the reagent will be contaminated by microorganisms
that feed on nitrate. The use of such contaminated reagents
will cause NO results to be biased high.
a
Working Standard Solution
5. Pour about 25 ml of stock standard solution into a clean,
dry beaker.
6. Using a 5-ml Class-A pipet, pipet 5 ml of stock standard
solution into a 200-ml Class-A volumetric flask. Dilute to
the calibration mark with water, and mix well.
This solution is the Working Standard; its nitrate content
represents a concentration of 25 yg N02/ml. The working
standard solution is prepared fresh for each set of
analyses.
Calibration Standards
7. Prepare a series of five calibration standards by pipetting
1.0, 2.0, 4.0, 6.0, and 10.0 ml of working standard solution
(25 yg/ml) into a series of five 50-ml Class-A volumetric
flasks. The standard masses will equal 25, 50, 100, 150,
and 250 yg N02, respectively. Dilute to the mark with either
water or eluent solution, and mix well.
The choice of diluent is determined by practical considera-
tions. If the "water dip" (see Figure 2.2) is expected to
interfere with the nitrate peak of the chromatogram, then
eluent should be used as the diluent since this will
minimize the "water dip." Note: Whichever diluent is used,
it is important for the analyst to use. the same diluent for
the field samples, the calibration standards, and the blank,
as specified in the Federal Register.
2.1.3 Preparation and Validation of the Calibration Curve - Method
7A specifies the determination of a calibration factor, S, which is
used to calculate the concentration of NO in the field samples. S
is defined as the reciprocal of the slope of the calibration curve,
which is determined by preparing or calculating a linear regression
plot of the standard masses of the calibration standards (yg) versus
instrument response (peak height or area). Determination of S does
not take into account the y-intercept, if present, of the calibration
curve.
The first subsection that follows describes the calibration pro-
cedures and the determination of the calibration factor as specified
in Method 7A. The second subsection offers an alternative approach,
acceptable to the Administrator, for conducting the calibration
-------
Section No. 3.14.2
Date July 1, 1986
Page 7
calculations that utilize the non-zero y-intercept, if present. This
approach is based on the calibration procedures of Method 7D and
involves the determination of a calibration equation. A data form
which can be used with both approaches is presented in Figure 2.3.
Determination of the Calibration Factor (S) - The determination
of the calibration factor, S, involves the three steps presented
below.
1. Analyze each of the calibration standards (25, 50, 100, 150,
and 250 yg N0~) three times using the ion chromatograph. Document
chromatograms fsee Subsection 5.1.4) and record the results on the
analytical data form for calibration standards (Figure 2.3). Average
the three responses for each of the five standards.
2. Use the average response for the five calibration standards
to calculate the slope of the calibration curve, graphically, by
least squares, or by linear regression. To calculate the slope
graphically, plot the instrument response (peak height or area count)
on the y-axis against the corresponding NG>2 standard concentration
value on the x-axis. Draw a "best-fit" line between the points and
determine the slope of the line. Least squares (a method acceptable
to the Administrator) can be hand calculated and is shown in Figure
2.3. To calculate the slope by linear regression, use the NO.-
standards as the independent variable (x-axis) and the corresponding
instrument response as the dependent variable (y-axis).
3. The calibration factor, S, is calculated as the reciprocal of
the slope of the calibration curve, determined from the "best-fit"
line or the linear regression equation. Any y-intercept is ignored.
4. The calibration factor, S, and therefore, the curve must be
validated. Using the calibration factor for calculation, the pre-
dicted sample mass for each calibration standard is compared with the
known value for that standard. The predicted sample mass must not
deviate from the known standard concentration by more than 7%. The
quantity "yg N02 Predicted" is calculated using the calibration
factor (S) and the detector response (H), in millimeters or integra-
tor response, as shown in Equation 2-1.
Equation 2-1
yg N02 = S (yg/mm) x Detector (mm)
Predicted Response
H
The deviation of each predicted sample mass from the known mass is
calculated using Equation 2-2.
Equation 2-2
Deviation = ^-^2 Predicted - yg N02 Standard x 1QQ%
(%) y g N02 Standard
-------
Plant
Date
/ /Z
Location
Analyst .
Section No. 3.14.2
Date July 1, 1986
Page 8
Was an integrator used?
yes
Was the intercept (I) used for calculations? yes S ho
Were all points within 7 percent of calculated value? */ yes
Sample
Identifier
Std I
Std 2
Std 3
Std 4
Std 5
Sample
Mass
(yg NO,)
25
50
100
150
250
Integrator Response
or Peak Height (mm)
H
1
4,1
/ZS"
2.5*3
381
6>0.\
2
6.4-
IZ-1
Z+3
39. /
5W
3
6.2
/3-0
zs.r
30.fr
57. <*
Avg
6-23;
IZM
ZS".23
3'0. (r0
61, &t
Predicted
Sample Mass
(yg NO,)
52. //
/^Z, 72-
/r/./r
>^f-3, 7r
no
Deviation
/ — / ^ /
t * ^f^f
+ 4.23
^™ ^ ^*^
+-4.77
-^.5Z>
Predicted Sample Mass using Least Squares to Calculate Calibration Factor (S)
with Zero Intercept
S = S1H1
S2H2
S3H3
S5H5
H
(/2.0Q)
S = 4:07/3 y g N02/mm
Predicted Sample Mass (yg N02)
yg N0_ = H x S = (^.Z3 ) x
Equation 2-1
Predicted Sample Mass using Linear Regression to Calculate Calibration Factor (S)
and Non-Zero Intercept (I)
y = mx + b; m =
x = - (y - b); - = S =
.5 b
1
m
m
y = H; and b = I (Intercept) =
Predicted Sample Mass (yg N0_)
yg N02 = S(H - I)
yg N02 at 25 yg standard =
Equation 2-
Figure 2.3. Analytical data form for analysis of calibration standards.
-------
Section No. 3.14.2
Date July 1, 1986
Page 9
This calculation is performed for each calibration standard using the
average of the three response measurements. If any point (known con-
centration of standard) deviates from the line (predicted concentra-
tion) by more than +7 percent, that standard should be remade and
reanalyzed.
Linear regression using a hand-held calculator is recommended to
obtain the slope (and equation) for the calibration curve. Inexpen-
sive calculators are available which have linear regression programs
that are quick and simple to use. Graphical techniques are relative-
ly simple matters when all the calibration data points lie on or
close to the line. However, when deviations from linearity occur,
the placement of the "best-fit" line becomes ambiguous because the
data points are not evenly distributed.
Determination of the Calibration Equation - As discussed pre-
viously, Method 7A directs that the calibration factor, S, be used to
calculate the field sample analytical results. in cases where the
calibration curve does not pass through the origin, the procedure of
Method 7A could give biased results for both the field samples and
the linearity check since the equation for the calibration curve will
contain an intercept term not taken into account in the calculations.
Accordingly, this section offers an alternative calibration approach
adapted from Method 7D. The approach involves determination of a
calibration equation which takes into account both the slope of the
calibration curve and any y-intercept term and which is used in
calculating the NO concentration of field samples.
H
Derive the linear calibration equation or curve using linear
regression. The calibration equation should be expressed in the
following form:
y = m x + b Equation 2-3
where
m = slope of the linear calibration curve, which is equal to
the reciprocal of the calibration factor, 1/S, and
b = y-intercept of linear calibration curve which will be
referred to as "I" for purposes of later calculations.
As discussed in the previous section, Method 7A requires that none
of the calibration data points deviate from the calibration curve by
more than 7 percent of the concentration at that point. Method 7A
(Section 5.2.3) states that deviations can be determined by
multiplying the calibration factor S times the peak height response
for each standard. When the calibration equation with intercept is
used, the quantity " yg NO, Predicted" is computed using the following
equation:
-------
Section No. 3.14.2
Date July 1, 1986 •
Page 10
. yg N02 = S (yg/mm) /Detector (mm) - I (mm)\ Equation 2-4
Predicted I Response I
\ H /
As before, calculation of the % deviation from the line is
accomplished using Equation 2-2. If any deviation is greater than
7%, the corresponding standard should be remade and reanalyzed. If
this does not result in improved results, other approaches are
discussed in the following subsection "Other Considerations."
2.1.4 Other Considerations - Method 7A requires that if any calibra-
tion standard point deviates from the standard calibration curve by
more than 7%, then that corresponding calibration standard is to be
remade and reanalyzed. This corrective action may not always reduce
the calibration point deviations below 7%. Some potential causes for
deviation of the calibration points from the calibration curve
include (a) improper pipetting procedures used to prepare calibration
standards, (b) improper technique for manual sample injection into
the ion chromatograph, (c) inaccurate measurement of the ion
chromatograph response, and (d) non-linear detector response. Table
2.1 shows the precisions for calibration operations for Method 7A.
TABLE 2.1. TARGET PRECISIONS FOR
CALIBRATION OPERATIONS OF METHOD 7A
Operation Precision Target (%)
Pipetting 1
Introduction of Samples <1
into Ion Chromatograph
Measurement Response
o Peak Height 1-4
o Triangulation 4
o Height X Width at
Half-Height 3
o Electronic Integration <0.5
Pipetting Procedure and Pipetting Errors - In preparing the
calibration standards, pipetting is the most critical step. Serious
errors can originate from poor pipetting technique. In general,
errors will appear as high biased NO results. The correct pipetting
procedure is described below. x
-------
Section No. 3.14.2
Date July 1, 1986
Page 11
The pipet should be inspected before use and checked to ensure
that the tip is not chipped. The pipet should be replaced if a chip
is observed.
The pipet should be rinsed with the reagent to be pipetted and
checked for cleanliness before use as follows. Approximately 2 ml of
reagent is drawn into the pipet, which is then rotated and tilted in
order to expose the inner surface to the solution. The rinse solu-
tion is then allowed to drain freely from the pipet into a beaker
assigned for waste. If the pipet is clean, the analyst will observe,
after about 10 seconds, that all the rinse solution will have drained
from the pipet with the exception of a small quantity remaining in
the tip. If this is not observed, either the pipet should be
cleaned, or another pipet should be obtained. The rinse and check
for cleanliness should be performed at least once.
For the actual pipetting, reagent is drawn into the pipet until
the liquid meniscus is above the calibration mark. The pipet is then
withdrawn from the solution and the end is wiped with a laboratory
tissue. Next, the pipet is brought to a vertical position and its
tip is brought to touch the inside of the beaker assigned for waste.
The liquid in the pipet is then allowed to drain slowly until the
meniscus coincides with the calibration mark.
The pipet is then transferred to the appropriate container and,
with the pipet in a vertical position and its tip touching the inside
wall of the container, the liquid is allowed to drain freely into the
container. The pipet's tip should be kept in contact with the wall
for roughly 10 seconds after the liquid has apparently drained. The
pipet is then removed from the container without disturbing the small
amount of liquid remaining in the tip.
It is important to recognize that Class-A pipets are calibrated
in a manner which accounts for the drainage time and the liquid
remaining in the tip. If dirty pipets are used or if the proper
draining technique is not observed, NO results will be biased high.
Low biases will occur if the liquid xremaining in the pipet tip is
blown out into the receiving container. The significance of these
biases depends on the size of the pipet involved. For example, the
error with a dirty 25-ml pipet may be undetectable, while the error
for a 1-ml pipet can easily exceed 10 percent.
The precision of the pipetting operation can be checked gravimet-
rically using water. The technique involves pipetting a known volume
of water into a tared container and determining the weight of the
water. The precision of the pipetting operation is estimated from
the results of several repetitions.
The procedure for manually injecting a sample into the ion chrom-
atograph can be a source of error when analyzing calibration stan-
dards, field samples, and QA samples. For fixed loop injection
systems,.considerable variation can result from injecting the sample
-------
Section No. 3.14.2.
Date July 1, 1986
Page 12
into the loop too fast, resulting in the sample loop not being
completely filled. A slow, deliberate injection of the sample into
the loop will completely fill the loop. The precision of the
injection procedure can be checked by performing repetitive analyses
on a single sample.
Chromatogram Quantitation - The choice of quantitation methods
for the ion chromatograms can also be a source of error when analy-
zing calibration standards, field samples, and QA samples. As shown
in Table 2.1, measurement of the detector response by manual methods
has a higher degree of imprecision compared to measurement by elec-
tronic integration. Method 7A states that peak height measurement
can be used provided the peaks are symmetrical and the required 7%
deviation of calibration points from the standard calibration curves
can be met. The peak height measurement method, even with symmetri-
cal peaks, may not produce a linear standard caliration curve because
the peak width of the higher concentration standards will typically
be wider than the peak width of the lower concentration standards.
Figure 2.4 shows the difference in the linearity of ion chromato-
graphic calibration curves using the peak area mode and the peak
height mode. The dead volume of the ion chromatograph system,
particularly suppressed ion chromatograph systems, can also affect
the peak width. Quantitation by peak area measurement will eliminate
the biases caused by widening peaks provided the peak area
measurement is done properly. The use of an electronic integrator in
the peak area mode for ion chromatograms with baseline separation of
the nitrate peak will produce the most precise calibration curves and
subsequent accurate analyses of field samples and QA samples.
-------
Response
Section No. 3.14.2
Date July l, 1935
Page 13
Peak Area
Approach
Peak Height
Approach
150
2,00
2.50 300
100
P g N02
Linear and non-linear ion chroma-
tographic calibration curves.
-------
Section No. 3.14.2
Date July 1, 1986
Page 14
TABLE 2.2. ACTIVITY MATRIX FOR CALIBRATION OF EQUIPMENT
Apparatus
Acceptance limits
Frequency and method
of measurement
Action if
requirements
are not met
Collection
flask
Measure volume within
10 ml
On receipt, measure with
graduated cylinder
Recalibrate
Barometer
Reading agrees within
2.5 mm (0.1 in.) Hg of
mercury-in-glass
barometer
Upon receipt and before
each field test
Repair
or return
Thermometer
Reading agrees within
1°C (2 F) of mercury-
in-glass thermometer
As above
As above
Vacuum gauge
(mechanical
only)
Reading agrees within
2.5 mm (0.1 in.) Hg
of mercury U-tube mano-
meter
As above
As above
Analytical
balance
Weight within 2 mg of
standard weights
(Class S)
Use standard weight
before preparation of
working solution
Repair or
return to
manufacturer
Ion chroma-
tograph
Calibration curve should
be linear; data points
for calibration stan-
dards must not deviate
from the linear calibra-
tion curve by more than
+7 percent
With each set of field
samples; calibration
standards prepared from
sodium nitrate
Interpret data
using another
technique: e.g
if using peak
height, change
to peak area;
conduct addi-
tional analy-
ses of cali-
bration stan-
dards; cali-
brate conduc-
tivity detec-
tor; consult
operator's
manual
-------
Section No. 3.14.3
Date July 1, 1986
Page 1
3.0 PRESAMPLING OPERATIONS
This section addresses the preparation and packing of supplies
and equipment needed for the sampling. The pretest preparation form
(Figure 3.1) can be used as an equipment checklist. Many presampling
operations for Method 7A are identical to those for Method 7. This
section will only discuss the operations that are different; however
all quality assurance activities for Method 7A presampling operations
are summarized in Table 3.1 at the end of this section including
those described in Section 3.6.3. See Section 3.0 of this Handbook
for details on preliminary site visits.
3.1 Apparatus Check and Calibration
Previously used equipment should be visually checked for damage
and/or excessive wear before each field test. Items should be
repaired or replaced (as applicable) if judged to be unsuitable for
use. A pretest sampling checks form (Figure 3.1) summarizes equip-
ment calibration. The pretest preparations form (Figure 3.2) can be
used as an equipment check and packing list. The completed form
should be dated, signed by the field crew supervisor, and filed in
the operational log book. The replacement of worn or damaged items
of equipment should be initiated. Procedures for performing the
checks are given herein; a check is placed in the proper row and
column as the check/operation is completed. Each team will have to
construct its own checklist according to the type of sampling train
and equipment it uses.
3.2 Reagents
Unless otherwise indicated, it is intended that all reagents
conform to the specifications established by the Committee on
Analytical Reagents of the American Chemical Society (ACS), where
such specifications are available; otherwise, use the best available
grade. See Subsection 1.2.3 of Section 3.14.1 for water specifica-
tions.
3.2.1 Sampling - The absorbing reagent is prepared by adding 2.8 ml
of concentrated sulfuric acid (H^SO.) to a 100-ml flask containing
water and diluting to volume witn mixing. Add 10 ml of this
solution, along with 6 ml of 3% hydrogen peroxide (H202) that has
been freshly prepared from a 30 percent solution, to a 1-liter
flask. Dilute to volume with water, and mix well. Prepare fresh
absorbing solution weekly, and avoid exposure to extreme heat or to
direct sunlight, as these will cause the hydrogen peroxide to
decompose. If the reagent must be shipped to the field, it is
advisable that the absorbing reagent be prepared fresh on-site.
3.2.2 Analysis - The following reagents are needed for analysis and
standardization:
-------
Section No. 3.14.3
Date July 1, 1986
Page 2
Date ^/<£5 /0S Calibrated by
r "^
Flask Volume
Flask volume measured with valves?
-------
Section No. 3.14.3
Date July 1, 1986
Page 3
Apparatus check
Probe
Glass liner
clean
Heated properly*
Leak checked
Collection Flask
Clean
Leak checked
Temperature
gauge
Evacuation System
Leak- free pumps
Manifold and
tubing
U-tube manometer
Barometer
Reagents
Water
Absorbing solu-
tion*
Sample Recovery
Dropper or burette
Sample bottles
Pipette, 25 ml
Acceptable
Yes
/
^
X
x
X
X
X
X
x
X
X
^
X
X
X
No
Quantity
required
3
14
T
z
2-
1
/ likr
1 lifer
2
14-
1 T
2-
Ready
Yes
X
X
<^
,
iX
X
X
X
^
^
^
No
Loaded
and packed
Yes
i^-"
^
•/
X
X
x
x
X
^^
^s"^
X
No
*Most significant items/parameters to be checked,
Figure 3.2. Pretest preparations.
-------
Section No. 3.14.3
Date July 1, 1986
Page 4
Stock standard solution - Dissolve exactly 1.847 g of dried
sodium nitrate (NaN03) [or 2.198 g of dried potassium nitrate (KNOq)]
in water, and dilute to 1 liter in a volumetric flask; mix well.
Prepare fresh after 1 month.
Working standard solution - Dilute 5 ml of the standard solution
to 200 ml with water in a volumetric flask, and mix well. Note; One
ml of the working standard solution is equivalent to 25 g of
nitrogen dioxide.
Eluent solution - Weigh 1.018 g of sodium carbonate (NaCO,~) and
1.008 g of sodium bicarbonate (NaHCOg), and dissolve in 4 liters of
water. Other eluents may be used (see Subsection 1.4.3).
-------
Section No. 3.14.3
Date July 1, 1986
Page 5
TABLE 3.1. ACTIVITY MATRIX FOR PRESAMPLING PREPARATION
Characteristic
Acceptance limits
Frequency and method
of measurement
Action if
requirements
are not met
Apparatus Check
Probe
1. Clean; glass liner
inert to oxides of
nitrogen
2. Heating properly if
equipped with heating
system
3. Leak free
1. Before each test
2. As above
3. Pressure <380 mm
(15 in.) Hg
Replace
Replace or
repair
Replace or
repair
Collection
flask
Clean; volume within
10 ml
Before each test,
clean with strong de-
tergent and hot tap
water, and rinse with
tap water and then
ASTM Type III water;
periodically clean
with grease remover
Repeat cleans-
ing of flask
and/or meas-
ure volume
Evacuation
system
Vacuum of 75 nun
(3 in.) Hg absolute
pressure in each flask;
leakage rate <10 mm
(0.4 in.) Hg/min
Before each test, check
for leaks using Hg-
filled U-tube manometer
Correct leaks
Absorbing
Reagent
Sulfuric acid
concentrated
Final concentration:
0.28 ml/liter
Prepare fresh absorbing
solution weekly; use
graduated pipette
Make up new
solution
Hydrogen perox-
ide, 3%
6 ml/liter
Water
Deionized distilled
to ASTM specifications
D 1193-82, Type III
Prepare fresh
for each anal-
ysis period
(continued)
-------
TABLE 3.1. (continued)
Section No. 3-14.3
Date July 1, 1986
Page 6
Characteristic
Acceptance limits
Frequency and method
of measurement
Action if
requirements
are not met
Analytical
Reagents
Stock standard
solution
1. 1.84? +0.001 g
NaNO- ACS reagent
grade into a 1-liter
volumetric flask
(Class-A)
2. Stored for less
than 1 month
1. On makeup of solution
use analytical balance
2. Date solution
1. Make up new
solution
2. As above
Working standard
solution
5 ml of stock solution
into 200-ml volumetric
flask (Class-A)
On makeup of solution,
use Class A pipet and
proper technique
As above
Eluent solution
1.018 g + 0.001 g of
NaCO- and 1.008 g
+ 0.002 g of NaHCO,
in 4 liters 5
On makeup of solution,
use analytical balance
As above
Packing Equip-
ment for Ship-
ment
Probe
Rigid container lined
with polyethylene foam
Prior to each shipment
Repack
Collection
flasks and
valves
Rigid container lined
with polyethylene foam
As above
As above
Evacuation
system, tem-
perature
gauges,
vacuum lines,
and reagents
Sturdy case lined with
polyethylene foam
As above
As above
Evacuation
pump
Shipping container
or housing designed
for travel
As above
As above
-------
Plant fTCsKji rorier r/^nt
Sample location £Sr 0<>i~H£--l> y Bretf-e^ & 1
Operator &&C?
City
Date
C,O^/f^^>a^ ^ Jvfcr>ir-s>Hf*
2/Z7/&ST
Barometric pressure (PK,^>- 2?- 84~
in. Hg
Sample
number
M-\
M-Z
Sample
point
location
6-l\
Jj>~ /O
CIO
Sample
time
24-hr
£733
074S-
Probe
temperature ,
°F
•Z./0
2-/0
Flask
and valve
. number
f-/3
Be -to
Volume
of flask
and valve (V^) ,
ml
2-0/3
Zo/O
r
Initial pressure
in. Hg
Leg A±
/3.6
13.7
Leg B±-
/3V
/3.8 '
/3V.
P ^
2^ . 5**?*
Z.3f
2.4^
Initial temperature
^(ti)
73
73
°R(Ti)
^53
5-35
Pi = pbar - (D
-------
Section No. 3.14.4
Date July 1, 1986
Page 1
4.0 ON-SITE MEASUREMENTS
The on-site activities include transporting equipment to the test
site, unpacking and assembling the equipment, confirming duct meas-
urements and traverse points (if volumetric flow rate is to be
determined), determining the molecular weight of the stack gas,
sampling for oxides of nitrogen, and recording the data. These
activities are the same as for Method 7 (Section 3.6.4), with the
exception of a portion of the sample recovery procedures as described
below. Blank data forms can be found in Section 3.14.12 for the
convenience of the Handbook user. Table 4.1 at the end of this
section summarizes the quality assurance activities relative to all
on-site measurements in Method 7A, including those described in
Section 3.6.12.
4.1 Sampling
On-site sampling procedures for Method 7A are the same as those
for Method 7. See Subsection 4.3 of Section 3.6.4 for detailed
descriptions of sampling procedures. For convenience, examples of
completed field data forms for Method 7 are reproduced in this
section (Figures 4.1A and 4.IB); blank copies are provided in Section
3.14.12.
4.2 Sample Recovery
Sample recovery procedures should be performed as described for
Method 7 (Section 3.6.4), with the exception that the steps for
checking and adjusting the pH of the sample should be deleted (note
changes in Figures 4.2A, 4.2B, and 4.3).
A 16-hour minimum sample absorption period is required as in
Method 7. Samples should be recovered within 4 days of collection.
As currently written, the method states that the samples should be
stored no more.than 4 days between collection and analysis. However,
a recent study utilizing samples from nitric acid plants and power
plants indicates that the storage period between recovery and
collection may be extended to 30 days.
-------
Plant
Sample location
Operator
City
Date
Barometric pressure (Pbar)
mm Hg
Sample
number
^-/
M>-t
Af>-3
Sample
point
location
*-'/
&-10
c-io
Sample
time
24-hr
0733
074-^
O&oi
Probe
temperature ,
°C
100
loo
loo
Flask
and valve
number
£%r&
K-/6
££-<&
Volume
of flask
and valve (Vp) ,
ml
2013,
ZO/0
ZOO'd
Initial pressure
in. Hg
Le.g Ai
37Z
373
57^.5-
Leg Bi
371
370.^
370
^ia
17. i
/(*.7
n.7
Initial temperature
°C (t±)
Z2. 2
zi.z.
^3.r
°R(Ti)b
2fs:z
2-14. 7.
Z?6.S~
Pi = pbar
•O D W
0) 0) (D
U3 ft O
0> (D ft
H-
W LI O
Figure 4.IB. Nitrogen oxide field data form (metric units)
O
*
CO
00
en
-------
Plant
rs
s*Je-r
Date
Sample recovery personnel £?. O(c&r Barometric pressure, (P, )
Person with direct responsibility for recovered samples /y. £ .
in. Hg
Sample
number
A/M
frp-Z.
Final pressure,
in. Hg
Leg Af
/fr
1.2
2.0
Leg Bf
o.(.
o.$
/.o
pfa
2-1 M
J27£f
J$~-W
Final temperature,
°F (tf)
73
72-
73
°R (Tf)b
£33
S"32
Sample
recovery
time,
24-h
/3Z2.
1 $4-o
Liquid
level
marked
^
^
Samples
stored
in locked
container
^
LiX*"^
P — D «/A -4- T) ^ **P ss *•
pf ~ pbar (Af + Bf'* xf fcf
460°F.
Lab person with direct responsibility for recovered samples
Date recovered samples received 3// IvS Analyst £'.
All samples identifiable?
Remarks
All liquids at marked level?
I
Signature of lab sample trustee
Figure 4.2A. NOX sample recovery and integrity data form (English units)
V O W
tu 0) (D
«Q rt O
(D (D ft
M •
-*
GO
l-i •
VD t->
00 it^
-------
Plant
ro
Date
Sample recovery personnel
Barometric pressure, (Pbar)
mm Hg
Person with direct responsibility for recovered samples
Sample
number
AA/
A^'Z
AP-3
Final pressure,
' mm Hg
Leg Af
4*.(,
?>o.<£~
5Z?.8
Leg B£
/^.Z
^0.3
pfa
702
7^>7
Final temperature,
°C (tf)
22.?
2Z.Z
22. 7-
°K (Tf)b
MS.?
2<7£~.2
Sample
recovery
time,
24-h
1322
1330
131-1
Liquid
level
marked
•x
^
Samples
stored
in locked
container
^
^
bar
= t
273°C.
Lab person with direct responsibility for recovered samples
Date recovered samples received ^ //^S"" Analyst £•".
All samples identifiable?
Remarks
All liquids at marked level?
Signature of lab sample trustee
"P.
Figure 4.2B. NO sample recovery and integrity data form (metric units)
•d a w
o> 0) a>
tQ ct O
a> a> ft
H-
01 Q O
C 3
GJ
VO
00
-------
Section No. 3.14.4
Date July 1, 1986
Page 6
Sampling
Volume of 25 ml of absorbing solution** placed in flask?
Flask valve stopper in purge position?
Sampling train properly assembled?
Leak free?* \/ Stopcock grease used?
Type? C,
Flask evacuated to 75 nun (3 in. ) Hg pressure?
Leakage from manometer observation?* _ ^ / //?.
[e.g., maximum change in manometer of £ 10 mm (0.4 in.)
Hg/min]
Initial pressure of flask recorded?*
Initial temperature of flask recorded?* _
Probe purged before sampling? _ S
Sample collected properly?* _ •_
Flask shaken for 5 min after collection and disassembly
from train?*
Samples properly labeled and sealed and stored for shipment?
Sample Recovery
Samples allowed to remain in flasks for minimum of 16 h?*
Final flask temperature and pressure recorded?*
Sample transferred to leak-free polyethylene bottle? >^
Flask rinsed twice with 5-ml portions of water, and rinse
added to bottle containing sample? _ t^
* Most significant items/parameters to be checked.
** Note that absorbing solution for Method 7A is different from
that of Method 7.
Figure 4.3. On-site measurements.
-------
Section No. 3.14.
Date July 1, 1986
Page 7
Table 4.1. ACTIVITY MATRIX FOR ON-SITE MEASUREMENTS
Characteristic
Acceptance limits
Frequency and method
of measurement
Action if
requirements
are not met
Apparatus
assembly
Assemble using Fig.
1.1; no leakage
Before sample collec-
tion, visually and
physically inspect all
connections
Check for
leaks; repair
system;
repeat test
Operational
check
Maximum vacuum of
75 mm (3 in.) Hg
absolute pressure
Leakage rate £10 mm
(0.4 in.) Hg/min
Before sample collec-
tion, use Hg-filled
U-tube manometer
As above
Check system
for leaks;
check vacuum
pump
Check all
joints and
valves for
source of leak
Sample
recovery
Shake flask for 5 min
Let flask set for a
minimum of 16 h, but
no more than 4 days
Shake flask for 2 min
Determine flask pres-
sure and temperature
Mark sample level on
container
Record data on data
form (Fig. 4.2)
During each sample
collection, use mano-
meter and Celsius
thermometer
Reject sample,
rerun test
Sample logistics
Properly label all
containers, etc.
Record all data on
field data forms
Visually check each
sample
As above
Complete the
labeling
Complete the
data records
-------
Section No. 3.14.5
Date July 1, 1986
Page 1
5.0 POSTSAMPLING OPERATIONS
The postsampling operations include checks on (a) the apparatus
used in the field to quantify sample volumes (volume, temperature,
and pressure measurements), and (b) analyses of the samples collected
and forwarded to the base laboratory. If the laboratory receives the
samples in the sample flasks, laboratory personnel will have to com-
plete the sample recovery procedures referred to in Section 3.6.4.
The postsampling checks on the sample collection train are the
same as for Method 7 (Section 3.6.5). The analytical procedures for
Method 7A are different from Method 7 and are discussed below.
Figure 5.1 is a checklist for all Method 7A posttest operations.
Table 5.1 at the end of this section summarizes the quality assurance
activities for all postsampling operations for Method 7A including
those described in Section 3.6.5.
5.1 Analysis (Base Laboratory)
Calibration of the ion chromatograph, including preparation of
the calibration standards and preparation of the field samples is of
primary importance to a precise and accurate analysis. For Method
7A, the calibration of the 1C is conducted in conjunction with analy-
sis of the field samples (and quality assurance samples). This
section presents the steps for analysis of the field samples
including preparation of samples, field blanks, and use of quality
assurance samples. The relationship between analysis of the field
samples and preparation of the calibration curve is addressed.
However, because a calibration and performance check of the 1C prior
to conducting any NO analyses is highly recommended, the detailed
discussion of the TC calibration is presented in Section 3.14.2.
Therefore, the analyst should use Section 3.14.2 in association with
this section (3.14.5) in conducting the analysis. In particular, the
analyst is encouraged to review the discussion of pipetting errors
(see Subsection 2.1.4). Upon completion of each step of the
preparation of the calibration curve and of each sample analysis, the
data should be entered on the proper data form.
5.1.1 Preparation of Field Samples - Check the level of the liquid
in the sample container and confirm whether any sample was lost
during shipment; note this on a data form such as that shown in
Figure 5.1. If a noticeable amount of leakage has occurred, either
void the sample or use methods subject to the approval of the
Administrator to correct the final results. Immediately before
analysis prepare each field sample. The following steps detail
sample preparation operations.
1. With the aid of a funnel, transfer the contents of the samp-
ling flask to a 50-ml Class-A volumetric flask.
2. Add approximately a 5-ml portion of water to the sampling
flask, replace the stopcock (ensuring that it is in the
-------
Section No. 3.14.5
Date July 1, 1986
Page 2
Reagents
Sodium nitrate dried at 105° to 110°C for a minimum of
2 hours before use?
Stock standard solution (sodium nitrate) less than 1 month old?
Sample Preparation
Has liquid level noticeably changed?* /Vp
Original volume Corrected volume
Analysis
Standard calibration curve prepared?* •_
All calibration points within 7 percent of linear calibration
curve?*
Reagent blanks made from absorbing solution? ^
Same injection volume for both standards and samples?
Duplicate sample values agree within 5 percent of their mean?
All analytical data recorded on checklist and laboratory form?
* Most significant items/parameters to be checked.
Figure 5.1. Posttest operations.
-------
Section No. 3.14.5
Date July 1, 1986
Page 3
closed position), and rinse the interior by shaking and
rotating the flask. Transfer the rinse to the volumetric
flask. Repeat the rinse with another 5-ml portion of water,
and add this rinse to the volumetric flask also.
3. Reassemble the sampling flask and place the stopcock in the
closed position to prevent contamination during storage prior
to reuse.
4. Using water, dilute the contents of the volumetric flask to
the mark. Mix the contents of the flask well.
5. Using a 5-ml Class-A pipet, pipet a 5-ml aliquot of the sample
into another 50-ml Class-A volumetric flask. This aliquot is
diluted to the mark with either water or eluent solution. Mix
the contents of the flask well.
The diluent used must be the same as that used for the calibra-
tion standards. (See Subsection 2.1.2 Preparation of Calibration
Standards.)
5.1.2 Preparation of Reagent Blank - The reagent blank is prepared
in essentially the same manner as the field samples. The difference
in procedure occurs at the first step. In preparing the reagent
blank, 25 ml of absorbing reagent is transferred to a 50-ml Class-A
volumetric flask. A 25-ml pipet may be used for measuring and
dispensing the reagent solution; however, the use of a graduated
cylinder will give results of acceptable accuracy and precision.
After introducing the absorbing reagent into the volumetric flask,
add water to the mark, and mix the contents of the flask well. The
remaining steps for preparing the reagent blank are identical to
those of Step 5 under Preparation of Field Samples.
The reagent blank is used to adjust the analytical results of the
field samples for matrix effects of the absorbing reagent and the
water. (The sample matrix is simply the medium that contains the
substance to be analyzed, which in this case is nitrate.) Because
ion chromatography involves separation of the ions prior to detection
and quantification, the potential for the sample matrix to interfere
with the analysis is small. For Method 7A, matrix effects can arise
from the presence of (a) nitrate contaminant in either the absorbing
reagent or the water, or (b) a contaminating substance appearing on
the chromatogram at about the same time as the nitrate peak. In
practice, the ion chromatogram should exhibit no significant response
at that point where nitrate should appear. Nevertheless, since data
are adjusted for the reagent blank, quality results can be obtained
even if contamination exists. The presence of contamination,
however, indicates the need for greater quality control in connection
with reagent integrity.
5.1.3 Quality Assurance Audit Samples - The quality of analytical
results can be assessed by analyzing nitrate standard solutions
-------
Section No. 3.14.5
Date July 1, 1986
Page 4
prepared by an independent laboratory. For such standard solutions,
or quality assurance audit samples, the concentrations are known to
the control agency (the auditor) but are unknown to the analyst.
Subsection 3.3.5 of the Federal Register promulgation of Method
7A (see Section 3.14.10) requires the analysis of quality assurance
audit samples as described in Method 7. This means that when Method
7A is used to demonstrate compliance with an EPA pollutant emission
standard (specified in 40 CFR Part 60), a performance audit must be
conducted on the analytical phase of the method. Nitrate samples in
glass vials must be obtained for this performance audit from the
Quality Assurance Management Office at each EPA Regional Office or
from the responsible enforcement agency. The addresses of the EPA
Regional Quality Assurance Coordinators are shown in Table 5.1 of
Section 3.0.5 of this Handbook.
The concentration of each audit sample measured by the analyst
must agree within 10 percent (relative error) of the actual audit
concentration. The relative error is calculated using the following
equation:
Equation 5-1
RE = Cd " Ca x 100
Ca
where
C, = Determined audit sample concentration, mg/dscm, and
C = Actual audit sample concentration, mg/dscm.
5.1.4 Analysis of Calibration Standards, Reagent Blank, Field
Samples, and Quality Assurance Samples - Field samples should be
recovered within 4 days of sample collection. As currently
written, the method states that the samples should be stored no more
than 4g days between collection and analysis. However, a recent
study utilizing samples from nitric acid plants and power plants
indicates that the storage period between recovery and collection may
be extended to 30 days. Sample analysis using an ion chromatograph
is a straightforward operation provided that the instrument has been
properly set up (see Section 3.14.2). All samples (calibration
standards, reagent blank, field samples, and quality assurance
samples) should be introduced into the ion chromatograph using the
same procedure. Sample introduction involves filling a constant
volume sample loop using a syringe or automatic sampling device.
Sample loops give extremely repeatable injection volumes; however,
the volumes that identify sample loop capacity are not necessarily
accurate. Nevertheless, accurate results can be obtained without
having accurately known sample loop volumes, provided that the same
sample loop is used for injecting field samples and calibration
standards. With this procedure, any inaccuracy in the injection
volume is accounted for by the calibration.
-------
Section No. 3.14.5
Date July 1, 1986
Page 5
Ion chromatographic analysis of calibration standards, field
samples, reagent blank, and quality assurance samples are performed
in five phases during the same day, alternating between the calibra-
tion standards and unknown samples to account for instrument cali-
bration drift. These phases are shown in the schedule below. When
Method 7A is used to demonstrate compliance with an EPA pollutant
emission standard, the quality assurance audit samples described in
Subsection 5.1.3 must be analyzed with the field samples.
Phase Activity
1 First analysis of all calibration standards.
2 First analysis of all field samples, reagent
blank, and quality assurance samples, if
applicable.
3 Second analysis of all calibration standards.
4 Second analysis of field samples, reagent blank,
and quality assurance samples, if applicable.
5 Third analysis of all calibration standards.
The calibration standards are analyzed in triplicate; the field
samples, reagent blank, and quality assurance samples in duplicate.
Replication of analyses increases the accuracy and precision of the
results. Each chromatogram obtained from the analysis should be
documented with the following information:
• sample identification,
• injection point,
• injection volume,
• nitrate retention time,
• sulfate retention time,
• eluent flow rate,
• detector sensitivity setting, and
• recorder chart speed.
Figure 5.2 shows an example chromatogram having acceptable documen-
tation. The injection volume, eluent flow rate, detector sensitivity
setting, and the recorder chart speed need to be documented only once
for the series of chromatograms if these analytical parameters remain
constant over the course of the Method 7A analysis.
Retention time is the elapsed time between when the sample is
introduced into the ion chromatograph and when the peak of interest
-------
Section No. 3.14.5
Date July 1, 1986
Page 6
Field Sample: AP-1
Chart Speed: 1 cm/min
Flow Rate: 1.5 ml/min
Detector: 30 yS full scale
Injection: 50 yl
N03 3.3 minutes
Inject
Figure 5.2. Example of chromatogram having adequate documentation.
-------
Section No. 3.14.5
Date July 1, 1986
Page 7
occurs. Peaks on the chromatogram may be qualitatively identified by
retention time. Retention times can be easily computed from chroma-
tograms provided that the injection point is indicated clearly and
the chart speed is known. Identification of the injection point is
necessary because a chromatogram's trace will not show when injection
occurred.
Record the results for the calibration standards, the field
samples, and reagent blank on the appropriate analytical data form
(Figures 5.3 and 5.4, respectively). As discussed in Subsection
2.1.3 and shown in Figure 5.3, the percent deviation from the cali-
bration curve of the average response value for each calibration
standard must be calculated and must be within 7 percent. A detailed
discussion of preparation of the calibration curve and calculation of
the calibration factor (S) is found in 3.14.2. The example data in
Figure 5.3 shows the use of linear regression to calculate S and a
non-zero intercept; the example data in Figure 2.3 shows calculation
of S with a zero intercept using least squares. Equation 2-1 or 2-4
along with Equation 2-2 (repeated below) are used to calculate the
percent deviation using either a zero intercept (Eq. 2-1) or a
non-zero intercept (Eq. 2-4).
Equation 2-1
vig N02 = S (yg/mm) x Detector (mm)
Predicted Response
H
Equation 2-4
yg N02 = S (ug/mm) /Detector (mm) - I (mm)\
Predicted I Response )
x H
Equation 2-2
Deviation = "9 NO2 Predicted - yg NO2 Standard x 10Q%
(%) vi g N02 Standard
For the analyses of the field samples, average the two response
values .of each sample (see Figure 5.4). The calculated average
should have units consistent with those of the calibration curve, for
example, units of peak height, peak area, etc. The pair of response
values for each sample must each agree within 5 percent of their mean
for the analysis to be valid. For this computation, the following
equation is used:
Equation 5-2
Deviation (%) = Instrument Response - Mean Response x 1QQ%
Mean Response
-------
Plant
Date
retftr rl**rh
I,
Location
Analyst
Section No. 3.11.5
Date July 1, 1986
Page 8
Was an integrator used?
yes
no
Was the intercept (I) used for calculations? * yes no
Were all points within 7 percent of calculated value? >/ yes
Sample
Identifier
Std 1
Std 2
Std 3
Std 1
Std 5
Sample
Mass
(vg NO,)
25
50
100
150
250
Integrator Response
or Peak Height (mm)
H
1
0.3
/4.3
u,^
30-0
57, 2
yg NOp/mm
Predicted Sample Mass (yg N0_)
yg N02 = H x S = (
Equation 2-1
Predicted Sample Mass using Linear Regression to Calculate Calibration Factor (S)
and Non-Zero Intercept (I)
y = mx + b; m =
; b =
= 4.30/3 ;
m
m
y = H; and b = I (Intercept) = 2
Predicted Sample Mass (yg N02)
yg N02 = S(H - I)
yg N02 at 25 yg standard = 4-WI ( &17 - ^.
Equation 2-4
Figure 5.3. Analytical data form for analysis of calibration standards.
-------
Section No. 3.1^-
Date July 1, 1986
Page 9
Date samples received 3//
i />
Plant ,
Date samples analyzed 3
Run number(s) AP~l l 3.
1 I / & S
~l l 3. STk
Location
May} /Wio> Analyst
Calibration factor (S) 4-. 3O/ Intercept (I), if applicable ;%. . ftt
Reagent blank values: 0.0 1st, 0-0 2nd, 0-0 Avg
Field
Sample
Number
Af-,
4/-Z
Analysis
Number
'£
2-tfp~
Instrument
Response
(mm)
28.7
30.2-
23.7
Mean
Instrument
Response
(mm)
^cl. sr
Deviation of two samples, (%) = 100 x
Percent
Deviation
Z.2-
Mean
Instrument
Response
Blank
Corrected
(H)
V.sr
13.1
Dilution
Factor
(F)
/
Mass of
Field
Sample
//4.4
Al ~ A2 (must be less than 5X)
= 100
Mass of field sample
without intercept
(US N02)
Mass of field sample
with intercept
(Vg N0)
= S x H x F
= S (H - I) F
Figure 5-^« Analytical data form for analysis of field samples.
-------
Section No. 3.14.5
Date July 1, 1986
Page 10
The reagent blank is analyzed at the same time as the field sam-
ples. The average blank corrected instrument response (H) is deter-
mined by subtracting the blank value from the average instrument
response for each sample. The blank corrected instrument response
(H), the dilution factor (F), and the calibration factor (S) [with
intercept (I) if necessary] are then used to calculate the mass
of N02 per sample as shown in Figure 5.4.
-------
Section 3-14.5
Date July 1, 1986
Page 11
Table 5.1. ACTIVITY MATRIX FOR SAMPLE ANALYSIS
Characteristic
Acceptance Limits
Frequency and method
of measurement
Action if
requirements
are not met
Calibration
standards
Data points for cali-
bration standards must
not deviate from the
linear calibration
curve by more than
+1%
Conduct for all analy-
ses of field samples
and calibration stan-
dards
Remake and reana-
lyze standards for
data points that
do not meet cri-
teria; interpret
data using another
technique (e.g.
peak area instead
of peak height);
strictly observe
.pipetting tech-
nique; use cali-
bration factor
with y-intercept
for calculations;
calibrate conduc-
tivity detector
Field sample
Results from dupli-
cate analyses must
be within 5 percent
of mean value
Conduct for all
analyses of field
samples
No results exceeding
value for calibration
standard having larg-
est concentration
Applicable to all
analyses of field
samples; determined
by visual inspection
Repeat duplicate
analysis, and
strictly observe
correct pipetting
technique; seek
assistance with
analytical tech-
nique
Dilute blank and
and affected field
sample with equal
volumes of water
and repeat analy-
ses of both
Performance
audit of
analytical
phase
See Section 3-14.8
See Section 3.14.8
See Section 3.14.8
Data
recording
All pertinent data
recorded on Figs. 5 • 1,
5.2, 5-3, and 5-4
Visually check
Supply missing
data
-------
Section No. 3.14.6
Date July 1, 1986
Page 1
6.0 CALCULATIONS
Calculation errors due to procedural or mathematical mistakes can
be a large component of total system error. Therefore, it is recom-
mended that each set of calculations be repeated or spot-checked,
preferably by a team member other than the one who performed the
original calculations. If a difference greater than typical round-
off error is detected, the calculations should be checked step-by-
step until the source of error is found and corrected. A computer
program is advantageous in reducing calculation errors. If a
standardized computer program is used, the original data entry should
be checked, and if differences are observed, a new computer run
should be made. Table 6.1 at the end of this section summarizes the
quality assurance activities for calculations.
Calculations should be carried out at least one extra decimal
figure beyond that of the acquired data, and should be rounded after
final calculation to two significant digits for each run or sample.
All rounding of numbers should be performed in accordance with the
ASTM 380-76 procedures. All calculations are then recorded on a form
such as the one in Figure 6.1A.
6.1 Nomenclature
The following nomenclature is used in the calculations:
P_ = final absolute pressure of flask, mm (in.) Hg,
P. = initial absolute pressure of flask, mm (in.) Hg,
P .. = standard absolute pressure, 760 mm (29.92 in.) Hg,
Tf = final absolute temperature of flask, °K (°R),
T± = initial absolute temperature of flask, °K (°R),
T .. = standard absolute temperature, 293°K (528°R),
V = sample volume at standard conditions, dry basis, ml,
sc
Vf =s volume of flask and valve, ml,
V = volume of absorbing solution, 25 ml,
O
H = sample peak height or area (blank should be subtracted
out), mm,
F = dilution factor (required only if additional sample
dilution was needed to reduce the concentration into
the range of calibration),
C = sample concentration of NO as N00, mg/dscm,
X ft
-------
Section No. 3.14.6
Date July 1, 1986
Page 2
S = calibration factor, y g/mm, and
I = intercept term from calibration equation, mm.
6.2 Calculations
The following four Subsections outline the procedures for calcu-
lating the concentration of nitrogen oxides in samples. Subsection
6.2.1 presents the equation for calculating the sample volume on a
dry basis at standard conditions.
Subsection 6.2.2 presents the equation for calculating the sample
concentration of nitrogen oxides as it appears in Method 7A. This
equation utilizes the calibration factor, S, determined during the
calibration of the ion chromatograph (see Subsection 2.1.3 of Section
3.14.2). Subsection 6.2.3 offers an alternative approach acceptable
to the Administrator for calculating the sample concentration of
nitrogen oxides utilizing the calibration factor, S, and the inter-
cept term, I, from the calibration equation. This equation is
determined following the procedures outlined in Method 7D for
calibration of the ion chromatograph (see Subsection 2.1.3 of Section
3.14.2).
Subsection 6.2.4 presents a simple equation for converting sample
concentration to parts per million. Examples of nitrogen oxide
calculation forms are presented at the end of each section and should
be used with the appropriate calculation methodology.
6.2.1 Sample Volume - Calculate the sample volume on a dry basis at
standard conditions [760 mm (29.92 in.) Hg and 293 K (528 R)] by
using the following equation.
T (v - v \ /P P v Equation 6-1
V = <3t-d* f a' ' f •!
vsc —^— —
std
P
= K.(Vf - 25 ml) /If
1 f \T
where
°K
K. = 0.3858 for metric units, or
mm Hg
K. = 17.64 °R for English units.
in. Hg
6.2.2 Sample Concentration Using the Calibration Factor, S - Calcu-
late the sample concentration on a dry basis at standard conditions
using the calibration factor, S, as shown in Equation 6-2 when the
calibration factor S was calculated with no intercept. See Figures
-------
Section No. 3.14.6
Date July 1, 1986
Page 3
6.1A and 6.IB for examples of calculation forms for English and
metric units, respectively.
Equation 6-2
HSF x 104
V
sc
where
4
10 =1:10 dilution times conversion factors of
mg 10 ml
3 3
10 y g m
6.2.3 Sample Concentration Using the Calibration Equation and
Factor, S - Calculate the sample concentration on a dry basis at
standard conditions using the calibration factor and the intercept
term for the calibration equation as shown in Equation 6-3. See
Figures 6.1A and 6.IB for examples of calculation forms for English
and metric units, respectively.
r „ (H-I) SF * 104 Equation 6-3
2~^
where
K2 = 1 for metric units, or
K9 = 6.243 x 10~8 dscm/mg for English units.
^ dscf/lb
4
10 =1:10 dilution times conversion factors of
mg 10 ml
3 X 3
10 yg m .
6.2.4 Sample Concentration in Parts-Per-Million - If desired, the
concentration of N02 may be calculated as ppm N02 at standard
conditions using Equation 6-4 as shown below.
Equation 6-4
ppm N02= K3 C
where
K3 = 0.5228 —ppm N02 for metric units, or
mg NO^/dscm
Kq = 8.375 x 106 PPnL_N02 for English units.
Ibs N02/dscf
-------
Section No. 3.14.6
Date July 1, 1986
Page 4
Sample Volume
vf = 2-0'L 3 ml, Pf = 2-7 . 64 in. Hg, Tf = £T3 3 °R
_L ™"™* *™~ "*~~ ™^ JL ~~ ~~ ^~
p p v , -, x. Equation 6-1
Vo^ = 17.64 (V, - 25) {_£ _ ^±_\ = I 7 $ 0 ml
SC I.
i) /If _ M -'_/7g^
\ ^p '^ »
v if i±/
Sample Concentration
(No Intercept Used)
H = __ . __ mm, S = ____ yg/mm,
F ' --- ' vsc ' ---- ml
Equation 6-2
4
C = 6.243 x 10"8 HSF x 10 = . x 10"5 Ibs NO0/dscf
V ~ z
sc
(With Intercept Used)
H = _2-3 . _/Omm, I = _ Z-'. f5fmm, S = 4._5'£ / yg/mm,
F = _ 1-0, Vgc = _/7 ^0 ml
Equation 6-3
4
C = 6.243 x 10~8 (H-I)SF x 10 = 3 . 0 f x 1Q-5 lbg N0 /dscf
v ^
sc
Sample Concentration in ppm
Equation 6-4
ppm N02 = 8.375 x 10 C = _ 2, 5" 5
Figure 6.1A. Nitrogen oxide calculation form (English units).
-------
Section No. 3.14.6
Date July 1, 1986
Page 5
Sample Volume
vf = 2-9. L $ m1' pf " 7 0 t. Q mm Hg, Tf = £ ? 5f. £ °K
P
± - _ / 6.. 0 mm Hg, T± = _/•£ 5_. 5TWK
Var, = 0.3858 (V, - 25) /If _ li | = _/ 7 fc> 5_. ml
SC I I m m
Tf T±.
Sample Concentration
(No Intercept Used)
H = _ _• _ _ WMr S - _ _ _ _ v g/mm
F = ___, V =____ m^
Equation 6-2
4
C = HSF x 10 = _. x 103 mg N02/dscm
sc
(With Intercept Used)
H*y "2. / f\ T 7 Q £~ r* ri *flt /^ / /
= (* 2 • / U mm, I = £•-•. / j mm, s = 7~. 3 u i yg/mm,
Equation 6-3
C = (H-I)SF x 10 = ^) . 4 ^ rx 103 mg N09/dscm
V
SC
Sample Concentration in ppm
Equation 6-4
ppm NO-, = 0.5228 C = ^-D T ppm N00
« ""^ *~" ~~ """" «b
Figure 6.IB. Nitrogen oxide calculation form (metric units).
-------
Section No. 3.14.6
Date July 1, 1986
Page 6
Table 6.1. ACTIVITY MATRIX FOR CALCULATIONS
Characteristics
Sample volume
calculation
Sample mass
calculation
Sample concen-
tration
Calculation
check
Document and
report re-
sults
Acceptance limits
All data available;
calculations correct
within round-off error
As above
As above
Original and checked
calculations agree
within round-off error
All data available;
calculations correct
within round-off error
Frequency and method
of measurement
For each sample, exam-
ine the data form
As above
As above
For each sample, per-
form independent cal-
culation using data on
Figs. 4.1, 4.2, and
4.3
For each sample, exam-
ine the data form
Action if
requirements
are not met
Complete the
data, or void
the sample
As above
As above
Check and
correct all
data
Complete the
data, or void
the sample
-------
Section No. 3.14.7
Date July 1, 1986
Page 1
7.0 MAINTENANCE
The normal use of emission-testing equipment subjects it to cor-
rosive gases, extremes in temperature, vibration, and shock. Keeping
the equipment in good operating order over an extended period of time
requires knowledge of the equipment and a routine maintenance program
which should be performed quarterly or upon improper functioning of
the apparatus. As for Method 7, it is suggested that the vacuum pump
be disassembled and cleaned yearly. A summary of the components with
maintenance procedures is presented in Table 7.1 at the end of this
section. These procedures are not required, but are recommended to
increase the reliability of the equipment.
7.1 Pumps
Several types of pumps are used in the present commercial samp-
ling trains. The two most common are the fiber vane pump with
in-line oiler and the diaphragm pump. The fiber vane pump requires a
periodic check of the oiler jar. The oil should be translucent.
During the yearly disassembly or if the fiber vane pump starts to run
erratically, the head should be removed and the fiber vanes changed.
The diaphragm pump will show a leak when the diaphragm needs
changing. If the diaphragm.pump runs erratically, it is usually due
to a bad diaphragm (causing leakage) or to malfunctions in the
valves. The valves should be cleaned annually by complete
disassembly of the pump.
7.2 Shipping Containers
Since the majority of the sampling train is glassware, the ship-
ping containers are very important for protection and safety. All
shipping containers should be inspected quarterly for their
condition, and repaired or modified to assure the safety of the
equipment.
7.3 Ion Chromatograph
Maintenance activities and schedules for ion chromatographs are
make and model specific. It is therefore recommended that the
analyst consult the operator's manual for instructions relative to
maintenance practices and procedures.
Guard columns, while not required, are recommended for use with
the ion chromatograph in order to extend column lifetime.
-------
Section No. 3.1^-7
Date July 1, 1986
Page 2
Table 7.1. ACTIVITY MATRIX FOR MAINTENANCE
Apparatus
Acceptance criteria
Frequency and method
of measurement
Action if
requirements
are not met
Fiber vane pump
Oil translucent; pump
leakless and capable
of pulling a vacuum of
less than 75 mm (3
in.) Hg absolute
pressure
Check oiler jar
periodically; remove
head and change fiber
vanes
Replace as
needed
Diaphragm pump
Leak free, valves
functioning properly,
and capable of pulling
a vacuum of < 75
-------
Section No. 3.14.8
Date July 1, 1986
Page 1
8.0 AUDITING PROCEDURE
An audit is an independent assessment of data quality. Indepen-
dence is achieved if the individual(s) performing the audit and their
standards and equipment are different from the regular field team and
their standards and equipment. Routine quality assurance checks by a
field team are necessary to generate good quality data, but they are
not part of the auditing procedure. Table 8.1 at the end of this
section summarizes the quality assurance functions for auditing.
19 20 21
Based on the results of collaborative tests ' ' of Method 7,
two specific performance audits are recommended:
1. Audit of the analytical phase of Method 7A.
2. Audit of data processing.
It is suggested that a systems audit be conducted as specified by the
quality assurance coordinator, in addition to these performance
audits. The two performance audits and the systems audit are
described in detail in Subsections 8.1 and 8.2, respectively.
8.1 Performance Audits
Performance audits are made to evaluate quantitatively the qual-
ity of data produced by the total measurement system (sample collec-
tion, sample analysis, and data processing). It is recommended that
these audits be performed by the responsible control agency once
during every enforcement source test. A source test for enforcement
comprises a series of runs at one source. The performance audit of
the analytical phase is subdivided into two steps: (1) a pretest
audit which is optional, and (2) an audit during the field sampling
and/or analysis phase which is required. No audit is recommended at
this time for the sample collection phase.
8.1.1 Pretest Audit of Analytical Phase (Optional) - The pretest
audit described in this section can be used to determine the pro-
ficiency of the analyst, the quality of the standard solutions in
the Method 7A analysis, and the ability to perform the computations
correctly. It should be performed at the discretion of the agency
auditor, the laboratory supervisor, source test company, or quality
assurance officer. The analytical phase of Method 7A can be audited
with the use of aqueous potassium or sodium nitrate samples. Aqueous
sodium nitrate samples may be prepared using the same procedure
described in Section 3.14.2 for calibration standard preparation.
The pretest audit provides the opportunity for the testing
laboratory to check the accuracy of its analytical procedure. This
audit is especially recommended for a laboratory with little or no
experience with the Method 7A analysis procedure described in this
Handbook.
As an alternative to preparing their own audit samples for a
pretest audit, a testing laboratory may, 30 days prior to the time of
-------
Section No. 3.14.8-
Date July 1, 1986
Page 2
the planned pretest audit, make a request to EPA's Environmental
Monitoring Systems Laboratory, Quality Assurance Division, Source
Branch, Mail Drop 77A, Research Triangle Park, North Carolina 27711
for known quality control samples. These samples are aqueous
potassium nitrate samples (and not sodium nitrate samples).
The relative error for each of two samples should be within 10
percent of true value. The relative error (RE) is an indication of
the bias that may be associated with the analytical phase of Method
7A. Calculate RE using Equation 8-1.
RE = Cd " Ca x 100
Ca
Equation 8-1
where
C, = Determined audit sample concentration, mg/dscm, and
C = Actual audit sample concentration, mg/dscm.
O
8.1.2 Audit of Analytical Phase of the Field Test (Required) - As
stated in Sections 3.3.9 and 4.4 of 40 CFR 60, Appendix A, Method 7
(49 FR 26522, 06/27/84), when the method is used for enforcement
testing, the analyst must analyze two audit samples along with the
field samples. The testing laboratory should notify the respon-
sible agency requiring the performance test of the intent to test at
least 30 days prior to the enforcement source test. The responsible
agency will provide two audit samples to be analyzed along with the
field samples from the enforcement source test. The purpose of this
audit is to assess the data quality at the time of the analysis. If
EPA is the agency requiring the performance test, the testing
laboratory should notify the Quality Assurance Management Office in
the respective EPA Regional Office. The addresses of the EPA
Regional Quality Assurance Coordinators are shown in Table 5.1 of
Section 3.0.5 of this Handbook.
The two audit samples and the compliance samples must be concur-
rently analyzed in the same manner to evaluate the technique of the
analyst, the standards preparation, and computation skills. (Note:
It is recommended that known quality control samples be analyzed
prior to the compliance and audit sample analysis to indicate any
problems. One source of these samples is the Source Branch listed in
Subsection 8.1.1.) The same analyst, analytical reagents, and
analytical system shall be used both for compliance samples and the
EPA audit samples; if this condition is met, auditing of subsequent
compliance analyses for the same enforcement agency within 30 days
may not be required. An audit sample set may not be used to validate
different sets of compliance samples under the jurisdiction of
different enforcement agencies, unless prior arrangements are made
with both enforcement agencies.
-------
Section No. 3.14.8
Date July 1, 1986
Page 3
Calculate the concentrations in mg/dscm using the specified
sample volume in the audit instructions. (Note: Indication of
acceptable results may be obtained immediately by reporting the audit
results in mg/dscm and compliance results in total mg N02/sample by
telephone to the responsible enforcement agency.) Include the
results of both audit samples, their identification numbers, and the
analyst's name with the results of the compliance determination
samples in appropriate reports to the EPA Regional Office or the
appropriate enforcement agency. Include this information with
subsequent compliance analyses for the same enforcement agency during
the 30-day period.
The concentration of each audit sample measured by the analyst
shall agree within 10 percent of the actual concentration. If the
10-percent specification is not met, reanalyze the compliance samples
and audit samples, and include initial and reanalysis values in the
test report.
Failure to meet the 10-percent specification may require retests
until the audit problems are resolved. However, if the audit results
do not affect the compliance or noncompliance status of the affected
facility, the Administrator may waive the reanalysis requirement,
further audits, or retests and accept the results of the compliance
test. While steps are being taken to resolve audit analysis prob-
lems, the Administrator may also choose to use the data to determine
the compliance or noncompliance status of the affected facility.
Other applications of Method 7A (i.e., Performance Specification
Tests) should follow agency recommended or required procedures.
8.1.3 Audit..Qf2Rai;a Processing - Calculation errors are prevalent
in Method 7. ' ' Data processing errors can be determined by
auditing the recorded data on the field and laboratory forms. The
original and audit (check) calculations should agree within round-off
error; if not, all of the remaining data should be checked. The data
processing may also be audited by providing the testing laboratory
with specific data sets (exactly as would appear in the field), and
by requesting that the data calculation be completed and that the
results be returned to the agency/organization. This audit is useful
in checking both computer programs and manual methods of data
processing.
8.2 Systems Audit
A systems audit is an on-site qualitative inspection and review
of the total measurement system (sample collection, sample analysis,
etc.). Initially, a systems audit is recommended for each enforce-
ment source test, defined here as a series of three runs at one
source. After the test team gains experience with the method, the
frequency of audit may be reduced--for example, to once for every
four tests.
The auditor should have extensive background experience in
source sampling, specifically with the measurement system being
audited. The functions of the auditor are summarized below:
-------
Section No. 3.14.8
Date July 1, 1986
Page 4
1. Inform the testing team of the results of pretest audits,
specifying any area(s) that need special attention or improvement.
2. Observe procedures and techniques of the field team during
sample collection.
3. Check/verify records of apparatus calibration checks and
quality control used in the laboratory analysis of control samples
from previous source tests, where applicable.
4. Record the results of the audit, and forward them with
comments to the team management so that appropriate corrective action
may be initiated.
While on site, the auditor observes the source test team's
overall performance, including the following specific operations:
1. Setting up and leak testing the sampling train.
2. Preparing the absorbing solution (if performed on-site) and
adding it to the collection flasks.
3. Collecting the sample.
4. Sample absorption procedures, sample recovery, and
preparation of samples for shipment.
Figure 8.1 is a suggested checklist for the auditor.
-------
Section No. 3.14.8
Date July 1, 1986
Page 5
Yes
No
Comment
/
/
Presampling preparation
1. Plant operation parameters variation
2. Calibration of the flask and valve volume—three
determinations
3. Absorbing reagent preparation
5.
On-site measurements
Leak testing of sampling train
Preparation and introduction of absorbing solution
into sampling flask
Postsampling
(Analysis and Calculation)
6. Control sample analysis
7. Sample aliquotting techniques
8. Ion chromatographic technique
a. Preparation of standard nitrate samples
(pipetting)
b. Calibration factor (+_ 7 percent for all
standards)
c. Duplicate sample values within 5 percent
of their mean
d. Adequate peak separation
9. Audit results (+ 10%)
a. Use of computer program
b. Independent check of calculations
Comments
Figure 8.1. Method 7A checklist to be used by auditors.
-------
Section No. 3.1U.8
Date July 1, 1986
Page 6
Table 8.1. ACTIVITY MATRIX FOR AUDITING PROCEDURE
Audit
Acceptance Limits
Frequency and method
of measurement
Action if
requirements
are not met
Performance
audit of
analytical
phase
Measured RE of the
audit samples shall
be within 10% for
both audit results
Frequency; Once during
every enforcement source
test*
Method; Measure QA sam-
ples and report values
to responsible agency
Review operating
technique and/or
calibration check
Data
processing
errors
Original and checked
calculations agree
within round-off
error
Frequency; Once during
every enforcement
source test
Method; Independent
calculations starting
with recorded data on
Figures 4.1 and 5.1
Check and correct
all data for the
audit period rep-
resented by the
sampled data
Systems
audit—
observance
of tech-
nique
Operational tech-
nique as described
in this section of
the Handbook
Frequency; Once during
every enforcement source
test until experience
gained, then every
fourth test
Method; Observation of
techniques assisted by
audit checklist,
Fig. 8.1
Explain to team
their deviations
from recommended
techniques, and
note on Fig. 8.1
*As defined here, a source test for enforcement of the NSPS comprises a series
of runs at one source. Source tests for purposes other than enforcement (e.g.,
a research project) may be audited at a lower frequency.
-------
Section No. 3.14.9
Date July 1, 1986
Page 1
9.0 RECOMMENDED STANDARDS FOR ESTABLISHING TRACEABILITY
To achieve data of desired quality, two essential considerations
are necessary: (1) the measurement process must be in a state of
statistical control at the time of the measurement, and (2) the
systematic errors, when combined with the random variation (errors or
measurement), must result in an acceptable uncertainty. As evidence
in support of good quality data, it is necessary to perform quality
control checks and independent audits of the measurement process; to
document these data; and to use materials, instruments, and measure-
ment procedures that can be traced to an appropriate standard of
reference.
Data must be routinely obtained by repeat measurements of stan-
dard reference samples (primary, secondary, and/or working standards)
and the establishment of a condition of process control. The working
calibration standards should be traceable to standards of higher
accuracy.
Class-S weights (made to NBS specifications) are recommended for
the analytical balance calibration. See Section 3.6.2 for details on
balance calibration checks.
Class-A volumetric flasks and pipets (made to NBS specifications)
should be used in the preparation and transfer of solutions.
Audit samples (as discussed in Section 3.14.8) must be used to
validate test results for compliance determination purposes and are
recommended as an independent check on the measurement process when
the method is performed for other purposes.
-------
10.0 REFERENCE METHOD*
Section No. 3.14.10
Date July 1, 1986
Page 1
METHOD 7A—DETERMINATION or NITROGEN
OXIDE EMISSIONS FROM STATIONARY
SOURCES—ION CHROMATOCRAPHIC METHOD
1. Applicability and Principle.
1.1 Applicability. This method applies to
the measurement of nitrogen oxides emitted
from stationary sources: it may be used as
an alternative to Method 7 (as defined in 40
CFR Part 60.8(b» to determine compliance
if the stack concentration is within the ana-
lytical range. The analytical range of the
method is from 125 to 1.2SO mg NO./m* as
NOi (65 to 655 ppm). and higher concentra-
tions may be analyzed by diluting the
sample. The lower detection limit is ap-
proximately 19 mg/m' (10 ppm), but may
vary among Instruments.
1.2 Principle. A grab sample is collected
in an evacuated flask containing a diluted
'sulfuric acid-hydrogen peroxide absorbing
solution. The nitrogen oxides, except ni-
trous oxide, are oxidized to nitrate and
measured by ion chromatography.
2. .Apparatus.
2.1 Sampling. Same as in Method 7, Sec-
tion 2.1.
2.2 Sampling Recovery. Same as in
Method 7. Section 2.2. except the stirring
rod and pH paper are not needed.
2.3 Analysis. For the analysis, the follow-
ing equipment is needed. Alternative instru-
mentation and procedures will be allowed
provided the calibration precision In Section
5.2 and acceptable audit accuracy can be
met.
2.3.1 Volumetric Plpets. Class A; 1-. 2-. 4-.
5-ml (two for the set of standards and one
per sample). 6-. 10-. and graduated 5-ml
sizes.
2.3.2 Volumetric Flasks. 50-ml (two per
sample and one per standard), 200-ml, and
1-llter sizes.
2.3.3 Analytical Balance. To measure to
within 0.1 mg.
2.3.4 Ion Chromatograph. The ion chro-
matograph should have at least the follow-
ing components:
2.3.4.1 Columns. An anion separation or
other column capable of resolving the ni-
trate ion from sulfate and other species
present and a standard anion suppressor
column (optional). Suppressor columns are
produced as proprietary items: however, one
can be produced in the laboratory using the
resin available from BioRad Company, 32nd
and Griffin Streets, Richmond. California.
2.3.4.2 Pump. Capable of maintaining a
steady flow as required by the system.
2.3.4.3 Flow Gauges. Capable of measur-
ing the specified system flow rate.
2.3.4.4 Conductivity Detector.
2.3.4.5 Recorder. Compatible with the
output voltage range of the detector.
3. ReagenU.
Unless otherwise indicated. It is Intended
that all reagents conform to the specifica-
tions established by the Committee on Ana-
lytical Reagents of the American Chemical
Society, where such specifications are avail-
able: otherwise, use the best available grade.
3.1 Sampling. An absorbing solution con-
sisting of sulfuric acid (H.SO.) and hydro-
gen peroxide (HiOi) is required for sam-
pling. To prepare the absorbing solution, i
cautiously add 2.8 ml concentrated H.SO. to
a 100-ml flask containing water (same as
Section 3.2), and dilute to volume with
mixing. Add 10 ml of this solution, along
with 6 ml of 3 percent H.O. that has been
freshly prepared from 30 percent solution.
to a 1-llter flask. Dilute to volume with
water and mix well. This absorbing solution
should be used within 1 week of Its prepara-
tion. Do not expose to extreme heat or
direct sunlight.
3.2 Sample Recovery. Deionized distilled
water that conforms to American Society
for Testing and Materials specification D
1193-74, Type 3. Is required for sample re-
covery. At the option of the analyst, the
KMnO. test for oxidizable organic matter
may be omitted when high concentrations
of organic matter are not expected to be
present.
3.3 Analysis. For the analysis, the follow-
ing reagents are required:
3.3.1 Water. Same as In Section 3.2.
3.3.2 Stock Standard Solution. 1 mg NO,/
ml. Dry an adequate amount of sodium ni-
trate (NaNOi) at 105 to 110'C for a mini-
mum of 2 hours just before preparing the
standard solution. Then dissolve exactly
1.847 g of dried NaNO, In water, and dilute
to 1 liter In a volumetric flask. Mix well.
This solution Is stable for 1 month and
st.ould not be used beyond this time.
3.3.3 Working Standard Solution. 25 \t%l
ml. Dilute 5 ml of the standard solution to
200 ml with water in a volumetric flask, and
mix well. •
3.3.4 Eluent Solution. Weight 1.018 g of
sodium carbonate (Na.CO,> and 1.008 g of
sodium bicarbonate (NaHCO,). and dissolve
in 4 liters of water. This solution is 0.0024 M
Na.COj/0.003 M NaHCO, Other eluents ap-
propriate to the column type and capable of
resolving nitrate ion from sulfate and other
species present may be used.
3.35 Quality Assurance Audit Samples.
Same as required in Method 7.
4. Procedure.
4.1 Sampling. Same as in Method 7, Sec-
tion 4.1.
4.2 Sample. Recovery. Same as in
Method 7, Section 4.2. except delete the
steps on adjusting and checking the pH of
the sample. Do not store the samples more
than 4 days between collection and analysis.
* Federal Register, Volume 48, No. 237, December 8, 1983.
-------
Section No. 3.14.10
Date July 1, 1986
Page 2
4.3 Sample. Preparation. Note the level
of the liquid in the container and confirm
whether any sample was lost during ship-
ment: note this on the analytical data sheet.
If a noticeable amount of leakage has oc-
curred, either void the sample or use meth-
ods, subject to the approval of the Adminis-
trator, to correct the final results. Immedi-
ately before analysis, transfer the contents
of the shipping container to a 50-mJ volu-
metric flask, and rinse the container twice
with 5-ml portions of water. Add the rinse
water to the flask, and dilute to the mark
with water. Mix thoroughly.
Plpet a 5-ml aliquot of the sample into a
50-ml volumetric flask, and dilute to the
mark with water. Mix thoroughly. For each
set of determinations, prepare a reagent
blank by diluting 5 ml of absorbing solution
to 50 ml with water. (Alternatively, eluent
solution may be used in all sample, stand-
ard, and blank dilutions.)
4.4 Analysis. Prepare a standard calibra-
tion curve according to Section 5.2. Analyze
the set of standards followed by the set of
samples using the same injection volume for
both standards and samples. Repeat this
analysis sequence followed by a final analy-
sis of the standard set. Average the results.
The two sample values must agree within 5
percent of their mean for the anlaysis to be
valid. Perform this duplicate analysis se-
quence on the same day. Dilute any sample
and the blank with equal volumes of water
if the concentration exceeds that of the
highest standard.
Document each sample chromatogram by
listing the following analytical parameters:
Injection point, injection volume, nitrate
and sulfate retention times, flow rate, detec-
tor sensitivity setting, and recorder chart
speed.
4.5 Audit Analysis. Same as required in
Method 7.
5. Calibration.
5.1 Flask Volume. Same aa in Method 7,
Section 5.1.
5.2 Standard Calibration Curve. Prepare
a series of five standards by adding 1.0. 2.0,
4.0. 6.0. and 10.0 ml of working standard so-
lution (25 fig/ml) to a series of five 50-ml
volumetric flasks. (The standard masses will
equal 25, 50. 100. 150. and 250 »ig.) Dilute
each flask to volume with water, and mix
well. Analyze with the samples as described
in Section 4.4 and subtract the blank from
each value. Prepare or calculate a linear re-
gression plot to the standard masses in pg
(x-axis> versus their peak height responses
in millimeters (y-axis). (Take peak height
measurements with symmetrical peaks: in
all other cases, calculate peak areas.) From
this curve, or equation, determine the slope,
and calculate its reciprocal to denote as the
calibration factor, S. If any point deviates
from the line by more than 7 percent of the
concentration at that point, remake and re-
analyze that standard. This deviation can be
determined by multiplying S times the peak
height response for each standard. The re-
sultant concentrations must not differ by
more than 7 percent from each known
standard mass (Le., 25, 50. 100. 150. and 250
tig).
5.3 Conductivity Detector. Calibrate ac-
cording to manufacturer's specifications
prior to initial use.
5.4 Barometer. Calibrate against a mer-
cury barometer.
5.5 Temperature Gauge. Calibrate dial
thermometers against mercury-in-glass
thermometers.
5.6 Vacuum Gauge. Calibrate mechanical
gauges, if used, against a mercury manome-
ter such as that specified in Section 2.1.6 of
Method 7.
5.7 Analytical Balance. Calibrate against
standard weights.
6. Calculations
Carry out the calculations, retaining at
least one extra decimal figure beyond that
of the acquired data. Round off figures
after final calculations.
6.1 Sample Volume. Calculate the sample
volume V. (in ml) on a dry basis, corrected
to standard conditions, using Equation 7-2
of Method 7.
6.2 Sample Concentration of NO. as NO,.
Calculate the sample concentration C (in
mg/dscm) as follows:
HSP x 10«
C= Eq. 7A-1
"•c
Where:
H = Sample peak height, mm
S = Calibration factor. »xg/inm
F = Dilution factor (required only if sample
dilution was needed to reduce the con-
centration into the range of calibration)
104 = 1:10 dilution times conversion factor
of
mg
10'us
10* ml
If desired, the concentration of NOt may
be calculated as ppm NOi at standard condi-
tions as follows:
ppm NOj - 0.5228 C Eq. 7A-2
Where:
0.5228 - ml/mg NO,.
-------
Section No. 3.14.10
Date July 1, 1986
Page 3
7. Bibliography.
1. Mulik. J. D. and E. Sawicki. Ion Chro-
matographic Analysis of Environmental Pol-
lutants. Ann Arbor. Ann Arbor Science Pub-
lishers. Inc. Vol. 2. 1979.
2. Sawicki, E.. J. D. Mulik. and E. Wittgen-
stein. Ion Chromatographic Analysis of En-
vironmental Pollutants. Ann Arbor, Ann
Arbor Science Publishers. Inc. Vol. 1.1978.
3. Slemer. D. D. Separation of Chloride
and Bromide from Complex Matrices Prior
to Ion Chromatographic Determination. An-
alytical Chemistry 52(12:1874-1877). Octo-
ber 1980.
4. Small. H.. T. S. Stevens, and W. C.
Bauman. Novel Ion Exchange Chromatogra-
phic Method Using Conductimetric Deter-
mination. Analytical Chemistry. 47(11:1801).
1975.
S. Yu. King K. and Peter R. Westlln. Eval-
uation of Reference Method 7 Flask Reac-
tion Time. Source Evaluation Society News-
letter. 4(4). November 1979. 10 p.
-------
Section No. 3.14.11
Date July 1, 1986
Page 1
11.0 REFERENCES
1. Federal Register, Volume 48, No. 237, December 8, 1983.
Method 7A - Determination of Nitrogen Oxide Emissions From
Stationary Sources.
2. Small, H. T., S. Stevens, and W. C. Bauman. Novel Ion
Exchange Chromatographic Method Using Conductimetric
Determination. Analytical Chemistry, 47(11):801, 1975.
3. Johnson, E. L. and R. Stevenson. Basic Liquid
Chromatography. Varian Associates, Inc., 1978.
4. Yost, R. W., L. S. Ettre, and R. D. Conlon, Practical
Liquid Chromatography, An Introduction. Perkin-Elmer,
1980.
5. Smith, F. C., Jr., and R. C. Chang. The Practice of Ion
Chromatography. John Wiley and Sons, Inc., New York,
1983.
6. Stevens, T. S. and M. A. Langhorst. Agglomerated Pellicular
Anion-Exchange Columns for Ion Chromatography. Analytical
Chemistry, 54 (6):950, 1982.
7. Stevens, T. S., G. L. Jewett, and R. A. Bredeweg. Packed
hollow fiber suppressors for ion Chromatography.
Analytical Chemistry, 54 (7):1206, 1982.
8. Mulik, J. D., and E. Sawicki. Ion Chromatography.
Environmental Science and Technology, 13 (7):804, 1979.
9. Stevens, T. S., J. C. Davis, and H. Small. Hollow Fiber Ion
Exchange Suppressor for Ion Chromatography. Analytical
Chemistry, 53 (9):1488, 1981.
10. Stevens, T. S. Packed fibers and new columns speed,
simplify ion Chromatography. Industrial Research and
Development, September 1983.
11. Gjerde, D. T., J. S. Fritz, and G. Schmuckler. Anion
Chromatography with Low-Conductivity Eluents. Journal of
Chromatography, 186 (509), 1979.
12. Jupille, T., D. Surge, and D. Togami. Ion Chromatography
uses only one column to get all the ions. Research and
Development 26 (3):135, 1984.
13. Jenke, D. Anion Peak Migration Ion Chromatography.
Analytical Chemistry, 53 (9):1535, 1981.
-------
Section No. 3.14.11
Date July 1, 1986
Page 2
14. Skoog, D. A., and D. W. West. Fundamentals of Analytical
Chemistry, Second Edition. Holt, Rinehart and Winston,
Inc., New York, 1969.
15. Yu, King D. and Peter R. Westlin. Evaluation of Reference
Method 7 Flask Reaction Time. Source Evaluation Society
Newsletter, 4(4), November 1979. 10 p. (Sees. Ill, 114,
and 301(a) of the Clean Air Act, as amended (42 U. S. C.
7411, 7414, and 7601(a))).
16. Steinsberger, S. C. (Entropy Environmentalists, Inc.).
Unpublished results of NO sample stability study. June
1987. x
17. Siemer, D. D. Separation of Chloride and Bromide from
Complex Matrices Prior to Ion Chromatographic Determina-
tion. Analytical Chemistry. 52 (12):1874-1877, October
1980.
18. Eubanks, D. R., and J. R. Stillian. Care of Ion
Chromatography Columns. Liquid Chromatography. 2 (2):74,
1984.
19. Hamil, Henry F. et. al. The Collaborative Study of EPA
Methods 5, 6, and 7 in Fossil Fuel Fired Steam Generators.
Final Report, EPA-650/4-74-013, May 1974.
20. Hamil, H. F., and R. E. Thomas. Collaborative Study of
Method for the Determination of Nitrogen Oxide Emissions
from Stationary Sources (Nitric Acid Plants).
EPA-650/4074-028, May 1974.
21. Hamil, Henry F. Laboratory and Field Evaluations of EPA
Methods 2, 6, and 7. Final Report, EPA Contract No.
68-02-0626, Southwest Research Institute, San Antonio,
Texas, October 1973.
-------
Section No. 3.14.12
Date July 1, 1986
Page 1
12.0 DATA FORMS
Blank data forms are provided on the following pages for the
convenience of the Handbook user. Each blank form has the customary
descriptive title centered at the top of the page. However, the
section-page documentation in the top right-hand corner of each page
has been replaced with a number in the lower right-hand corner that
will enable the user to identify and refer to a similar filled-in
form in a text section. For example, Form M7A-1.2 indicates that the
form is Figure 1.2 in Section 3.14.1 of the Method 7A section.
Future revisions of these forms, if any, can be documented by 1.2A,
1.2B, etc. Eleven of the blank forms listed below are included in
this section. Four are in the Method Highlights subsection as shown
by the MH following the form number.
Title
Procurement Log
Analytical Balance Calibration Form
Analytical Data Form for Analysis of
Calibration Standards
3.1 (MH) Pretest Sampling Checks
3.2 (MH) Pretest Preparations
4.1A AND 4.IB Nitrogen Oxide Field Data Form (English
and metric units)
4.2A and 4.2B NO Sample Recovery and Integrity Data
Form (English and metric units)
4.3 (MH) On-site Measurements
5.1 (MH) Posttest Operations
5.4 Analytical Data Form for Analysis of
Field Samples
6.1A and 6.IB Nitrogen Oxide Calculation Form (English and
metric units)
8.1 Method 7A Checklist to be Used by Auditors
-------
PROCUREMENT LOG
Item description
«ty.
Purchase
order
number
Vendor
Date
Ord.
Rec.
Cost
Disposition
Comments
Quality Assurance Handbook M7A-1.2
-------
ANALYTICAL BALANCE CALIBRATION FORM
Balance name
Number
Classification of standard weights
Date
0.5000 g
1.0000 g
10.000 g
50.0000 g
100.0000 g
Analyst
Quality Assurance Handbook M7A-2.1
-------
ANALYTICAL DATA FORM FOR ANALYSIS OF CALIBRATION STANDARDS
Plant
Date
Location
Analyst
Was an integrator used?
yes
no
Was the intercept (I) used for calculations? yes no
Were all points within 7 percent of calculated value? yes
Sample
Identifier
Std 1
Std 2
Std 3
Std 4
Std 5
Sample
Mass
(u§ NO.,)
25
50
100
150
250
Integrator Response
or Peak Height (mm)
H
1
2
3
Avg
Predicted
Sample Mass
(yg N00)
no
Deviation
(%)
Predicted Sample Mass using Least Squares to Calculate Calibration Factor (S)
with Zero Intercept
S = S1H1
H 2
Hl '
Sv
—
S =
3redicted
u g NO- =
+ S2H2 + S^H^ + S^ + S^H^
2222
H H2^ + H^ * Hf + H^
)( ) + ( )( ) + ( )( ) + ( )( ) * ( )( )
P P P 5 P
( r + ( r + ( )2 + ( )2 + ( )2
yg N0_/mm
Sample Mass (yg NOp)
H x S = ( ) x ( ) = Equation 2-1
Predicted Sample Mass using Linear Regression to Calculate Calibration Factor (S)
and Non-Zero Intercept (I)
y = mx + b; m =
m
m
y = H; and b = I (Intercept) =
Predicted Sample Mass (yg N0_)
yg N02 = S(H - I)
yg N02 at 25 yg standard = (
Equation 2-
Quality Assurance Handbook M7A-2.3
-------
NITROGEN OXIDE FIELD DATA FORM (ENGLISH UNITS)
Plant
Sample location
Operator
City
Date
Barometric pressure (P, )
in. Hg
Sample
number
Sample
point
location
Sample
time
24-hr
Probe
temperature ,
OF
Flask
and valve
number
Volume
of flask
and valve {Vp) ,
ml
Initial pressure
in. Hg
Leg A±
Leg Bi
Pia
Initial temperature
°F(ti)
°R(Ti)b
= p
bar
460°F.
Quality Assurance Handbook M7A-4.1A
-------
Plant
Sample location
Operator
NITROGEN OXIDE FIELD DATA FORM (METRIC UNITS)
City
Date
Barometric pressure (P. )
oar
mm Hg
Sample
number
Sample
point
location
Sample
time
24-hr
Probe
temperature ,
°C
Flask
and valve
number
Volume
of flask
and valve (VF) ,
ml
Initial pressure
in. Hg
Leg A±
Leg B±
*ia
Initial temperature
°C(ti)
°R(Ti)b
pi = pbar
273°c-
Quality Assurance Handbook M7A-4.1B
-------
NO SAMPLE RECOVERY AND INTEGRITY DATA FORM (ENGLISH UNITS)
Plant Date
Sample recovery personnel
Barometric pressure, (Pbar)
in. Hg
Person with direct responsibility for recovered samples
Sample
number
Final pressure,
in. Hg
Leg Af
Leg Bf
Pfa
Final temperature,
°F (tf)
°R (Tf)b
Sample
recovery
time,
24-h
Liquid
level
marked
Samples
stored
in locked
container
f = pbar - - ^'f = fcf + 460°F-
Lab person with direct responsibility for recovered samples
Date recovered samples received Analyst
All samples identifiable?
Remarks
All liquids at marked level?
Signature of lab sample trustee
Quality Assurance Handbook M7A-4.2A
-------
N0x SAMPLE RECOVERY AND INTEGRITY DATA FORM (METRIC UNITS)
Plant
Date
Sample recovery personnel
Barometric pressure, (P, )
oar
Person with direct responsibility for recovered samples
mm Hg
Sample
number
Final pressure,
mm Hg
Leg Af
Leg Bf
V
Final temperature,
°c (tf)
°K (Tf)b
Sample
recovery
time,
24-h
Liquid
level
marked
Samples
stored
in locked
container
= Pbar - (Af
tf + 273°C.
Lab person with direct responsibility for recovered samples
Date recovered samples received Analyst
All samples identifiable?
Remarks
All liquids at marked level?
Signature of lab sample trustee
Quality Assurance Handbook M7A-4.2B
-------
ANALYTICAL DATA FORM FOR ANALYSIS OF FIELD SAMPLES
Date samples received
Plant
Date samples analyzed
Run number(s)
Location
Calibration factor (S)
Reagent blank values:
1st,
Analyst
Intercept (I), if applicable
2nd, Avg
Field
Sample
Number
Analysis
Number
Instrument
Response
(mm)
Mean
Instrument
Response
(mm)
Deviation
(yg N02)
Mean
Instrument
Response
Blank
Corrected
(H)
Dilution
Factor
(F)
Mass of
Field
Sample
(yg N02)
Deviation of two samples, (%) =
A - A i
100 x 1 2\ (must be less than
Al + A2
= 100
Mass of field sample
without intercept
N02)
Mass of field sample
with intercept
(yg N02)
= S x H x F
= S (H - I) F
Quality Assurance Handbook M7A-5.4
-------
NITROGEN OXIDE CALCULATION FORM (ENGLISH UNITS)
Sample Volume
Vf = ml, Pf = . in. Hg, Tf = °R
P± = . in. Hg, T± = °R
Equation 6-1
Vsc = 17.64 (V£ -25) If _ li =
ml
Sample Concentration
(No Intercept Used)
H = . mm, S = vg/mm,
F . Vsc = ml
Equation 6-2
4
C = 6.243 x 10"8 HSF x 10 = . x 10"5 Ibs N00/dscf
V ^
sc
(With Intercept Used)
H = . mm, I = . mm, S = vg/mm,
F ' ' Vsc = ml
Equation 6-3
4
C = 6.243 x 10"8 (H-DSF x 10 = ^ x 1Q-5 lbs N0 /dscf
V z
sc
Sample Concentration in ppm
ppm N02 = 8.375 x 10 C = ppm N02
- Equation 6-4
D
Quality Assurance Handbook M7A-6.1A
-------
NITROGEN OXIDE CALCULATION FORM (METRIC UNITS)
Sample Volume
'f
ml, Pf = .
mm Hg, T. = _
i • _
V = 0.3858 (V- - 25)
SC I
_ mm Hg, Tf
. °K
. ml
Equation 6-1
Sample Concentration
(No Intercept Used)
H = . mm, S = yg/mm
F - < vsc • ml
C =
HSF x 10
sc
4
x 10 mg NO0/dscm
(With Intercept Used)
H = . mm, I
F = V =
r ' vsc
c a (H-I)SF x 10
Vsc
Equation 6-2
= . mm, S = y g/mm,
ml
Equation 6-3
3
_ . x 10 mg N02/dscm
Sample Concentration in ppm
ppm N02 = 0.5228 C = ppm N02
Equation 6-4
Quality Assurance Handbook M7A-6.1B
-------
METHOD 7A CHECKLIST TO BE USED BY AUDITORS
Yes
No
Comment
Presampling preparation
1. Plant operation parameters variation
2. Calibration of the flask and valve volume three
de terminations
3. Absorbing reagent preparation
On-site measurements
4. Leak testing of sampling train
5. Preparation and introduction of absorbing solution
into sampling flask
Postsampling
(Analysis and Calculation)
6. Control sample analysis
7. Sample aliquotting techniques
8. Ion chromatographic technique
a. Preparation of standard nitrate samples
(pipetting)
b. Calibration factor (+_ 7 percent for all
standards)
c. Duplicate sample values within 5 percent
of their mean
d. Adequate peak separation
9. Audit results (+_ 102)
a. Use of computer program
b. Independent check of calculations
Comments
Quality Assurance Handbook M7A-8.1
-------
Section No. 3.15
Date July 1, 1986
Page 1
Section 3.15
METHOD 7D - DETERMINATION OF NITROGEN OXIDE
EMISSIONS FROM STATIONARY SOURCES
(Alkaline-Permanganate - Ion Chromatographic Method)
OUTLINE
Number of
Section Documentation pages
SUMMARY 3.15 3
METHOD HIGHLIGHTS 3.15 8
METHOD DESCRIPTION
1. PROCUREMENT OF APPARATUS
AND SUPPLIES 3.15.1 18
2. CALIBRATION OF APPARATUS 3.15.2 20
3. PRESAMPLING OPERATIONS 3.15.3 6
4. ON-SITE MEASUREMENTS 3.15.4 10
5. POSTSAMPLING OPERATIONS 3.15.5 13
6. CALCULATIONS 3.15.6 5
7. MAINTENANCE 3.15.7 3
8. AUDITING PROCEDURES 3.15.8 6
9. RECOMMENDED STANDARDS FOR
ESTABLISHING TRACEABILITY 3.15.9 1
10. REFERENCE METHODS 3.15.10 9
11. REFERENCES 3.15.11 2
12. DATA FORMS 3.15.12 11
-------
Section No. 3.15
Date July 1, 1986
Page 2
SUMMARY
For EPA Method 7D , an integrated, metered sample is extracted
via a heated probe positioned at a point within the duct or stack.
The sample is passed through a series of 3 restricted orifice
impingers each containing an absorbing solution of sodium hydroxide
( NaOH ) and potassium permanganate ( KMnO. ) . The absorbing solution
reacts with nitrogen oxides in the effluent gas to form nitrate ion,
~
N0«, and nitrite ion, N02 • Nitrogen oxides (NO ) are the sum of
nitric oxide (NO) and nitrogen dioxide (N02) whichxare usually at a
ratio of 19 to 1 by weight, respectively, in the emission stream.
The collected sample is required to sit for 36 hours prior to
analysis in order for the N02~ to react completely to N03~. Ion
chromatography is then used to quantify the N03~ wnich is
functionally related to the NO concentration of the effluent sample.
a
The absorbing solution also reacts with carbon dioxide, C07, in
the effluent sample. Therefore, EPA Method 3 determinations of C02
must be conducted with Method 7D in order to correct the Method 7D
volumetric data for the volume of C02 absorbed.
Ammonia, NH3, interferes with Method 7D by causing NO
be biased high. Method 7D results can be corrected for"-
using data from concurrent determinations of NH,, .
results to
the bias
Collection of the NO is presumed to involve oxidation - reduction
reactions where the NO is oxidized sequentially to NO~ ~and then to
*T^N ml_ — > 1_ — ^ .f — . — — — -1_ J — . __ — .^ ~_ — _L.1_. » ^ ,•. _._ ,_. J_ J ... _» ... *£ »T^X _. __ v A ™*
NO
The half reactions for the formation of N0~ ar
Mn0
+ 2H20
3e =
3NO + 60H
3NO,
+ 3H20
and the overall reaction is:
3NO + Mn04 + 20H
3N0
40H
3e~
- MnO,
* *) • J**Vx« » 11 A V
The half reactions for the formation of NO,
6e"
80H~
are:
3N0
'2
2MnO,
60H~ = 3N0
4H00
2
6e~ = 2Mn0
J * • ~XA« A W • W W — **fc-»**^//^
and the overall reaction is:
3N0
2Mn0
2Mn0
20H
Reaction S-l
Reaction S-2
Reaction S-3
Reaction S-4
Reaction S-5
Reaction S-6
NO
A * 4*l~lLl\*/ * • A A OV"/ — ^/*^V^Q • *** ** A xx A
The sum of Reactions S-3 and S-6 describes the reaction of NO to
*
3 '
NO + MnO,
'4 = N°3
0
2
Reaction S-7
-------
Section No. 3.15
Date July 1, 1986
Page 3
The rate of the reaction of NO to NO-" is controlled by the
solubility of NO. It takes approximately 36 hours for the reaction
of N02~ to N03" to reach completion; the factors controlling this
reaction are unknown.
Absorption of N02 is also presumed to involve an oxidation
-reduction reaction. In contrast to NO, N02 is rather reactive;
thus, it is reasonable to show N02 reacting directly to N0o~* Tne
half reactions are:
3N02 + 60H~ = 3N03~ + 3H20 + 3e~ Reaction S-8
Mn04~ + 2H20 + 3e~ = Mn02 + 40H~ Reaction S-9
and the overall reaction is:
3NO0 + MnO.," + 2OH~ = 3NO~ + MnO,, + H00 Reaction S-10
24 322
The absorption of C02 involves the simple acid-base reaction with
OH~ to form bicarbonate Ion, HC03~:
C02 + OH" = HC03~ Reaction S-ll
In the strongly basic absorbing solution, the bicarbonate ion reacts
further to carbonate ion, C03~:
HC03~ + OH" = H20 + C03= Reaction S-12
Method 7D is applicable to the measurement of NO emitted from
sources in the following categories: x
(a) fossil-fuel-fired steam generators subject to 40 CFR
Part 60, Subpart D;
(b) electric utility steam generating units subject to 40 CFR
Part 60, Subpart Da; and
(c) nitric acid plants subject to 40 CFR Part 60, Subpart G.
It may be used as an alternative to Method 7 [as defined in 40
CFR Part 60.8(b)] to determine compliance if the stack concentration
is within the analytical range. The lower limit ofgdetectability
(with NO defined as N02) is approximately 13 mg N02/m (7 ppm N02)
when sampling is conducted at a flow rate of 500 cc/min for 1 hour.
The method's upper analytical limit has not been established;
however, results of field evaluations have shown that NO can be
collected quantitatively at concentrations of 1,782 mg N02^m (932
ppm N02) when sampling is conducted at a flow rate of 500 cc/min for
1 hour.
-------
Section No. 3.15
Date July 1, 1986
Page 4
The method description which follows is based on the method that
was promulgated on September 27, 1984.
Section 3.15.10 contains a copy of Method 7D, and blank data
forms are provided in Section 3.15.12 for the convenience of the
Handbook user.
-------
Section No. 3.15
Date July 1, 1986
Page 5
METHOD HIGHLIGHTS
Section 3.15 contains the required procedure for sampling and
analyzing emissions of nitrogen oxides from stationary sources using
Method 7D. For the method, an integrated sample is taken from a
point in the duct or stack using a heated probe constructed of
borosilicate glass, stainless steel, or Teflon™. The effluent sample
stream is passed through a series of three restricted orifice
impingers, each containing 200 ml of a 4.0% (w/w) KMn04 and 2.0%
(w/w) NaOH solution, termed "alkaline permanganate solution." The
alkaline permanganate solution quantitatively removes NO , C02, and
SO- from the effluent sample stream and converts (provided the
samples are allowed to sit for at least 36 hours) these to ions:
N03 , C03~, and SO.", respectively. Sampling is conducted at a
measured flow rate between 400 and 500 cc/min for 60 minutes. The
measured flow rate is on a moisture- and CO.-free basis, and
consequently, when the method is applied to effluents from combustion
processes, the measured flow rate will be less than the sampling flow
rate. In addition, sampling for C02 must be conducted using Method 3
in conjunction with Method 7D in oraer to correct the volumetric data
for the volume of C02 absorbed.
After acquisition, the sample is allowed to sit for a minimum of
36 _hours to ensure that the N02 has been quantitatively reacted to
NO3 . Sample preparation entails destruction of the excess
permanganate and filtration of the solid, manganese reaction product,
manganese dioxide (Mn02). NO as N03~ is quantified using ion
chromatography (1C). x
Ion chromatography is a relatively recent analytical develop-
ment. The reader is referred12to the literature for detailed
descriptions of the subject. Small, et al., developed the
technique using the principles of ion exchange chromatography and
conductimetric detection. Previous attempts to use this type of
detection were unsuccessful because of the presence of the background
electrolyte used for elution of the ionic species. Small, et al.,
used a novel combination of resins to separate the ions of interest
and neutralize the eluent from the background.
The • aqueous sample is introduced into a fixed-volume sample loop
by using a plastic syringe. Once injected, the sample is carried
through a separation column at different rates according to their
relative affinities for the resin material and are therefore
separated into discrete bands. The separated ions are then passed
through a post-separation suppressor device which converts the eluent
ions into a less conducting weak acid while converting the analyte
ions into a highly conducting form. This permits the use of a
conductivity cell as a very sensitive detector of all ionic species.
12
Gjerde, et al., described a modified ion chromatographic method
that eliminates the need for a suppressor device. Anions are
-------
Section No. 3.15
Date July 1, 1986
Page 6
separated on a column containing an anion-exchange resin with a low
exchange capacity. Because of the low capacity, a very dilute
solution of an aromatic organic acid salt may be used as the eluent.
The conductance of the eluent is sufficiently low that no suppression
is needed.
For Method 7D, either suppressed or non-suppressed 1C may be
used. The basic ion chromatograph will have the following
components:
(a) sample injection device,
(b) anion separation column,
(c) anion suppressor column, either packed bed or fiber type
(not required for non-suppressed 1C),
(d) conductivity detector, and
(e) recorder.
The critical aspects of the method are (a) the measurement of the
gaseous sample volume, and (b) the preparation of the calibration
standards for the ion chromatograph. Analysts are advised to observe
specified procedures carefully at these points of the method.
Analysts performing the method should be well trained in the use of
the ion chromatograph.
Collaborative testing has been performed for Method 7D and the
results exhibit accuracy and precision similar to that of Method 7.
The approporiate blank data forms at the end of this section may
be removed from the Handbook and used in the pretest, on-site, and in
posttest operations. Each form has a subtitle to assist the user in
finding a similar filled-in form in the method description. On the
blank and filled-in forms, the items/ parameters that can cause the
most significant errors are designated with an asterisk.
1. Procurement of Apparatus and Supplies
Section 3.15.1 (Procurement of Apparatus and Supplies) gives
specifications, criteria, and design features for the required
equipment and materials. The sampling apparatus of Method 7D has
design features similar to those of Method 6. Section 3.15.1 can be
used as a guide for procurement and initial checks of equipment and
supplies. The activity matrix (Table 1.1) at the end of the section
is a summary of the details given in the text and can be used as a
quick reference.
2. Pretest Preparations
Section 3.15.2 (Calibration of Apparatus) describes the required
calibration procedures and considerations for the Method 7D sampling
-------
Section No. 3.15
Date July 1, 1986
Page 7
equipment (essentially the same as Method 6) and for the ion
chromatograph (the same as for Method 7A). Required accuracies for
each component are also included. A pretest checklist (Figure 2.5,
Section 3.15.2) or a similar form should be used to summarize the
calibration and other pertinent pretest data. The calibration
section may be removed along with the corresponding sections from the
other methods and made into a separate quality assurance reference
manual for use by personnel involved in calibration activities.
Section 3.15.3 (Presampling Operations) provides the tester with
a guide for equipment and supplies preparation for the field test.
With the exception of the preparation of certain reagents, these are
the same as for Method 6 and Method 3. A pretest preparation form
(Figure 3.1, Section 3.15.3) can be used as an equipment checkout and
packing list. The method of packing and the use of the described
packing containers should help protect the equipment, but neither is
required by Method 7D.
Activity matrices for the calibration of equipment and the
presampling operations (Tables 2.1 and 3.1) summarize the activ-
ities.
3. On-Site Measurements
Section 3.15.4 (On-Site Measurements) contains step-by-step
procedures for sample collection and for sample recovery. Sample
collection is similar to Method 6, with the exception that the
alkaline permanganate solution is placed in restriced orifice
impingers and the C02 content of the stack gas must be determined to
correct the sample volume for the C02 removed by the sampling train.
The on-site measurement checklist (Figure 4.4, Section 3.15.4)
provides the tester with a quick method of checking the on-site
requirements. Table 4.1 provides an activity matrix for all on-site
activities.
4. Posttest Operations
Section3.15.5fPostsampling Operations) gives the posttest
equipment procedures and a step-by-step analytical procedure for
determination of NO , expressed as N02. The posttest operations form
(Figure 5.4, Section 3.15.5) provides some key parameters to be
checked by the tester and laboratory personnel. The step-by-step
analytical procedure description can be removed and made into a
separate quality assurance analytical reference manual for the
laboratory personnel. Analysis of a control sample is required prior
to the analysis of the field samples. This analysis of independently
prepared, known standards will provide the laboratory with quality
control checks on the accuracy and precision of the analytical
techniques. Strict adherence to the Method 7D analytical procedures
must be observed.
Section 3.15.6 (Calculations) provides the tester with the
required equations, nomenclature, and significant digits. It is
-------
Section No. 3.15
Date July 1, 1986
Page 8
suggested that a calculator be used, if available, to reduce the
chances of calculation error.
Section 3.15.7 (Maintenance) provides the tester with a guide for
a maintenance program. This program is not required, but should
reduce equipment malfunctions. Activity matrices (Tables 5.1, 6.1,
and 7.1) summarize all postsampling, calculation, and maintenance
activities.
5. Auditing Procedure
Section 3.15.8 (Auditing Procedures) provides a description of
necessary activities for conducting performance and system audits.
When Method 7D is used to demonstrate compliance with an EPA
pollutant emission standard, a performance audit must be conducted on
the analytical phase of the method. The data processing procedures
and a checklist for a systems audit are also included in this
section. Table 8.1 is an activity matrix for conducting the audits.
Section 3.15.9 (Recommended Standards for Establishing
Traceability) provides the primary standard to which the analysis
data should be traceable. The primary standard is sodium nitrate
(NaNO3).
6. References
Section 3.15.10 contains the promulgated Method 7D; Section
3.15.11 contains the references cited throughout the text; and
Section 3.15.12 contains copies of data forms recommended for Method
7D.
-------
Section No. 3.15
Date July 1, 1986
Page 9
PRETEST SAMPLING CHECKS
(Method 7D, Figure 2.5)
Date Calibrated by
Meter box number
Dry Gas Meter*
Pretest calibration factor (Y) = (within 2% of
average factor for each calibration run).
Rotameter
Pretest calibration factor (Y ) or setting =
(between 400 and 500 cc/minl.
Dry Gas Meter Thermometer
Was a pretest temperature correction made? yes no
If yes, temperature correction (within 3°C (5.4°F) of
reference values for calibration and within 6 C (10.8°F) of
reference values for calibration check).
Barometer
Was the pretest field barometer reading correct? yes no
(within 2.5 mm (0.1 in.) Hg of mercury-in-glass barometer).
*Most significant items/parameters to be checked.
-------
Section No. 3.15
Date July 1, 1986
Page 10
PRETEST PREPARATIONS
(Method 7D, Figure 3.1)
Apparatus check
Probe
Type liner
Glass
Stainless
steel
Other
Heated properly*
Leak checked
Filter
Glass wool
Other
Glassware
Restricted
orifice
impinger
Size
Type
Meter System
Leak- free pumps*
Rate meter*
Dry gas meter*
C02 Measurement
Orsat
Fyrite
Reagents
Water
Alkaline per-
manganate*
Silica gel
Other
Barometer
Drying tube
Acceptable
Yes
No
Quantity
required
Ready
Yes
No
Loaded
and packed
Yes
No
*Most significant items/parameters to be checked.
-------
Section No. 3.15
Date July 1, 1986
Page 11
ON-SITE MEASUREMENTS
(Method 7D, Figure 4.4)
Sampling
Impinger contents properly selected , measured, and placed in
impingers?*
Impinger Contents/Parameters*
1st: 200 ml of KMn04/NaOH
2nd: 200 ml of KMn04/NaOH
3rd: 200 ml of KMn04/NaOH
Drying tube: 6-16 mesh indicating type silica gel
Probe heat at proper level?*
Crushed ice around impingers?
Pretest leak check at 250 mm (10 in.) Hg?
Leakage rate?
Check of rotameter setting?
Probe placed at proper sampling point?
Flow rate constant at approximately 450 cc/min?*
C02 concentration measured?*
Posttest leak check at 250 mm (10 in.) Hg?*
Leakage rate?*
Sample Recovery
Contents of impingers placed in polyethylene bottles?
Fluid level marked?*
Sample containers sealed and identified?*
*Most significant items/parameters to be checked.
-------
Section No. 3.15
Date July 1, 1986
Page 12
POSTTEST OPERATIONS
(Method 7D, Figure 5.4)
Reagents
Potassium nitrate dried at 105 to 110 C for a minimum of 2 hours
before use?
Stock standard solution (potassium nitrate) less than 2 months
old?
Sample Preparation
Has liquid level noticeably changed?*
Original volume Corrected volume
Analysis
Standard calibration curve prepared?*
Reagent blanks made from absorbing solution?
Same injection volume for both standards and samples?
Duplicate sample values agree within 5 percent of their mean?
Audit sample analytical results within 10 percent of true value?
All analytical data recorded on checklist and laboratory form?
* Most significant items/parameters to be checked.
-------
Section No. 3.15.1
Date July 1, 1986
Page 1
1.0 PROCUREMENT OF APPARATUS AND SUPPLIES
The procurement of appropriate apparatus and supplies enables
quality results to be obtained from Method 7D. This section
provides the user with information which complements the two sections
of Method 7D, entitled "Apparatus" and "Reagents." The information
is offered in the form of guidance and includes the following:
o Procedures for use in checking whether apparatus conforms with
the requirements of the Method 7D and corrective actions for
when it does not (Table 1.1 at the end of this section
summarizes these procedures and also contains recommended
corrective actions).
o Background information which can explain why specific appar-
atus and reagents are required, and therefore, what limits may
exist for alternatives or deviations.
o Practical information pertinent to the selection and use of
apparatus and reagents.
o Safety considerations.
Persons responsible for the initial procurement of apparatus and
supplies may find a procurement log helpful in ensuring that all the
necessary items are acquired and are in good working order. A
procurement log can be used to record the descriptive title of the
equipment, the quantity, an identification number (if appropriate),
and the results of acceptance checks. An example procurement log is
shown by Figure 1.1, a blank copy of this form is contained in
Section 3.15.12 for the Handbook user. Calibration data obtained
during acceptance checks also should be recorded in a calibration log
book; see Section 2.0.
1.1 Sampling Apparatus
Figure 1.2 shows the sampling train for Method 7D. It should be
noted that this sampling train is very similar to that used for
Method 6. Several of the components and their use are identical,
including:
o Needle Valve
o Drying Tube
o Vacuum Pump
o Parts of the Metering System
This subsection addresses the specifications needed for procurement
purposes for all components of the sampling train and associated
apparatus.
1.1.1 Sampling Probe - Method 7D specifies that sampling probes are
to be constructed of borosilicate glass. The method also states that
-------
Item description
Qty.
Purchase
order
number
Vendor
Date
Ord.
Rec.
Cost
Disposition
Comments
7Z4/3/
I/4/&
J/ZI/B+
•^J '
/i
Z//0/B4
Figure 1.1. Example of a procurement log.
•d O W
0> 0) d>
(Q rt O
0) 0> rt
00
vO (-•
oo cn
-------
Section No. 3.15.1
Date July 1, 1986
Page 3
PROBE END PACKED.
RESTRICED ORIFICE IMPINGERS
SILICA GEL
DRYING TUBE
Figure 1.2. Method 7D.sampling train.
-------
Section No. 3.15.1
Date July 1, 1986
Page 4
probes made of either stainless steel or Teflon™ are acceptable.
Quartz probes (for example Vycor™) may be used for sampling when
effluent temperatures exceed 480 C (900°F).
The function of the probe is rather simple: to transport a
representative effluent sample, cleaned of particulate matter, to the
impinger train. To perform this function, the probe should:
(a) hold a filter to remove particulate matter, including
sulfuric acid mist;
(b) be constructed of a material that is unreactive toward NO ;
Jv
(c) be free from leaks;
(d) be sufficiently long to enable samples to be acquired from
the specified points(s) within the stack or duct;
(e) have provisions for being heated in order to prevent
condensation of water vapor in the effluent sample; and
(f) be designed to connect to the inlet of the impinger train.
The three materials identified above are unreactive toward NO . The
appropriate length for the probe is determined primarily Sy its
intended application which will depend upon regulatory requirements
and the dimensions of the stack or duct where the measurements are to
be made.
Sampling probes are generally provided with electrical heating
systems consisting of a nichrome wire which is wrapped around the
probe. The probe and heating system are, for protection, placed
within a tightly fitting tube, referred to as a sheath. The heating
system should be capable of preventing condensation of water vapor in
the effluent sample stream during sampling. Condensation is not
desired, because water absorbs N02 and lowers NO results.
Additionally, if a stainless steel probe is used, condensation will
promote corrosion which shortens probe lifetime and makes cleaning
difficult.
It is recommended that probes be performance checked before
initial use in the field to ensure that condensation can be pre-
vented. The probe should first be visually checked for cracks or
breaks and then checked for leaks according to the procedure
described in Section 3.15.3 of this Handbook. Then the probe heating
system should be checked as follows:
1. Connect the probe (without filter) to the inlet of the pump.
2. Electrically connect and turn on the probe heater for 2 or 3
minutes. If functioning properly, it will become warm to the touch.
-------
Section No. 3.15.1
Date July 1, 1986
Page 5
3o Start the pump, and adjust the needle valve until a flow rate
of between 400 and 500 cc/min is achieved.
4= Check the probe. It should remain warm to the touch. The
heater must be capable of maintaining the exit air temperature at a
minimum of 100°C (212 F) under these conditions. If it cannot, the
probe should be rejected. Any probe not satisfying the acceptance
check should be repaired, if possible, or returned to the supplier.
The connection between the probe's outlet and the impinger train
may be a simple fitting or an additional length of tubing. Any
connection should be leak-free. In addition, the connection should
be constructed of borosilicate glass, stainless steel, or Teflon? and
therefore, like the probe be unreactive toward NO . Lastly, if
tubing is used, provisions should exist to prevent condensation of
water vapor upstream of the impinger train during sampling. A
heating system for the connection is not required if the probe's
heating system can supply enough heat to the effluent sample.
1.1.2 Restricted Orifice Impingers - The sampling train requires the
use of three restricted orifice impingers connected in series.
Figure 1.3 shows one of these impingers, which are commercially
available.
13
Impinger design is important to obtain quality results. The
restricted orifice impinger is specifically designed to promote the
complete collection of NO, which is relatively unreactive. Two
design features are important; (a) the length of the liquid column,
and (b) the size of the impinger's stem tip. The impingers used for
Method 7D are narrower than Greenburg-Smith impingers in order to
provide a greater depth of absorbing reagent and, hence, to increase
the reaction time of the sample gas in the absorbing reagent.
Because of the narrow opening of the stem tip, the effluent sample is
introduced into the absorbing reagent as smaller bubbles. Smaller
bubbles promote the reaction of NO because of their greater surface-
to-volume ratio and thus, greater exposure to the absorbing reagent.
Impingers with stem tips restricted to less than 1.5 mm internal
diameter are easily plugged by reaction products. The problem
typically affects only the first impinger of the sampling train
because: (a) most of the NO and C02, and (b) all the sulfur dioxide,
if present, are reacted there. If plugging occurs, the problem may
be minimized by making the length of the capillary tubing shorter;
plugging also can be minimized by keeping stem tips clean. Reaction
products in the stem tips can be removed by immersion in either 3
percent by volume hydrogen peroxide solution [3% (v/v) H909 (aq)] or
3M hydrochloric acid solution [HC1 (aq)]. CAUTION: Chlorine (Cl..,)
gas is evolved during the use of the HC1; therefore, cleaning
operations should be conducted in a fume hood. The H.-02 solution is
identical to the absorbing solution used for Method 6.
-------
45/50
DIMENSIONS: mm
35
CAPILLARY
TUBING:
1.5 I.D.
Section No. 3.15.1
Date July 1, 1986
Page 6
28/12
Figure 1.3. Restricted orifice impinger.
-------
Section No. 3.15.1
Date July 1, 1986
Page 7
It is recommended that each impinger upon receipt be checked vis-
ually for damage, such as breaks or cracks, and for manufacturing
flaws, such as poorly shaped connections.
Other nonspecified collection absorbers and sampling flow rates
may be used, subject to the approval to the Administrator, but
collection efficiency must be shown to be at least 99% for each of
three test runs and must be documented in the emission test report.
For efficiency testing, an extra absorber must be added and analyzed
separately and must not contain more than 1% of the total NO .
a
1.1.3 Vacuum Pump - The vacuum pump should be capable of maintaining
a flow rate of approximately 400 to 500 cc/min for pump inlet vacuums
up to 250 mm (10 in.) Hg with the pump outlet near standard pressure,
[i.e., 760 mm (29.92 in.) Hg]. The pump must be leak free when
running and pulling a vacuum (inlet plugged) of 250 mm (10 in.) Hg.
Two types of vacuum pumps are commonly used: either a modified
sliding fiber vane pump or a diaphragm pump. For safety reasons, the
pump should be equipped with a three-wire electrical cord.
To check the pump for leaks, install a vacuum gauge in the pump
inlet line. Plug the inlet line, and run the pump until the vacuum
gauge reads 250 mm (10 in.) Hg of vacuum. Clamp the pump outlet
line, and turn off the pump; the vacuum reading should remain stable
for 30 seconds.
1.1.4 Volume Meter - The dry gas meter must be capable of measuring
total volume with an accuracy to within 2%, calibrated at the
selected flow rate (between 400 and 500 cc/min), and must be equipped
with a temperature gauge (dial thermometer, or eguivalent) capable of
measuring the gas temperature to within 3°C (5.4 F).
A new dry gas meter may be checked for damage visually and by
performing a calibration according to Section 3.5.2 of this Hand-
book. Any dry gas meter that is damaged, behaves erratically, or
does not give readings within 4-2% of the selected flow rate for each
calibration run is unsatisfactory. Also upon receipt, the meter
should be calibrated over a varying flow range to see whether there
is any effect on the calibration.
Dry gas meters that are equipped with temperature compensation
must be calibrated over the entire range of temperatures that the
meter encounters under actual field conditions. The calibration must
contain at least one data point at each 10°F interval. All
temperatures that are to be used in the field must be within 2% of
the calibrated value.
The wet test meter used to check the dry gas meter should be
calibrated using the primary displacement technique explained in
Section 3.5.2 of this Handbook. The wet3 test meter must have a
capacity of at least 0.0003 m /min (0.1 ft /min) with an accuracy of
+2%; otherwise at the higher flow rates, the water will not be level
and this possibly will result in an incorrect reading.
-------
Section No. 3.15.1
Date July 1, 1986
Page 8
1.1.5 Rotameter - A rotameter, or its equivalent, with a range of 0
to 1 L/min isused to monitor the sampling flow rate. The rotameter
is checked against the calibrated dry gas meter with which it is to
be used or against a wet test meter. It is recommended that it be
within 5% of the manufacturer's calibration curve. The rotameter
flow setting of about 450 cc/min should be determined.
Changes in pressure, density, and viscosity of the sample gas
will affect the calibrated sample rate. However, since sampling is
performed at a constant rate, which need not be isokinetic, these
changes do not affect the sample volume measured by the dry gas
meter.
1.1.6 Needle Valve - A metering valve with conveniently sized
fittings is required in the sampling train to adjust and control the
sample flow rate. It is recommended that the needle valve be placed
on the vacuum side of the pump.
1.1.7 Drying Tube - The drying tube should be packed with 6- to
16-mesh indicating-type silica gel, or equivalent, to dry the sample
gas and to protect the pump and the meter. A drying tube can be made
by filling a 10-mm polyethylene tube with silica gel and packing
glass wool in each end to hold the silicia gel and protect the
sampling system. Plastic tubing can be utilized in any connections
downstream of the impinger train without affecting the sampling
results. The drying tube should have a minimum capacity of 30 to 50
g of silica gel and should be visually checked before use for proper
size and for damage.
If the silica gel has been used previously, it must be dried at
175°C (350 F) for 2 hours. New silica gel may be used as received.
Other types of desiccants may be used subject to approval of the
Administrator.
1.1.8 Metering System - For ease of use, the metering system—which
contains the dry gas meter, thermometer(s), vacuum pump, needle
valve, and rotameter--can be assembled into one unit (meter box).
After a meter box has been either constructed or purchased, then
positive and negative pressure leak checks should be performed.
The positive pressure leak check, similar to the procedure
described in Method 5 (Section 3.4) of this Handbook, is performed as
follows:
1. Attach rubber tubing and inclined manometer, as shown in
Figure 2.1 of Section 3.4.2 of this Handbook.
2. Shut off the needle valve, and apply positive pressure to the
system by blowing into the rubber tubing until the inclined manometer
or magnehelic gauge reads from 12.5 to 17.5 cm (5 to 7 in.) H0.
-------
Section No. 3.15.1
Date July 1, 1986
Page 9
3. Pinch off the tube, and observe the manometer for 1 minute.
A loss of pressure indicates a leak of the apparatus in the meter
box.
After the meter box apparatus has passed the positive leak check,
then the negative leak check should be performed as follows:
1. Attach the vacuum gauge at the inlet to the drying tube, and
pull a 250 mm (10 in.) Hg vacuum.
2. Pinch or clamp the outlet of the flow meter. This can be
accomplished by closing the optional shutoff valve if employed.
3. Turn off the pump. Any deflection noted in the vacuum
reading within 30 seconds indicates a leak.
4. Carefully release the vacuum gauge before releasing the flow
meter end.
If either of these checks detects a leak that cannot be cor-
rected, the meter box must be rejected and/or returned to the
manufacturer.
The dry gas meter must be equipped with a temperature gauge (dial
thermometer or equivalent). It is recommended that upon receipt this
be checked visually for damage, such as dents or a bent stem. The
thermometer should read within 3 C (5.4 F) of the true value when
checked at two different ambient temperatures against a
mercury-in-glass thermometer that conforms to ASTM E-l No. 63C or
63F. The two ambient temperatures used to calibrate the thermometer
must differ by a minimum of 10 C (18 F). Damaged thermometers that
cannot be calibrated are to be rejected.
1.1.9 Barometer - A mercury, aneroid, or other barometer capable
of measuring atmospheric pressure to within 2.5 mm (0.1 in.) Hg may
be used. However, in many cases, the barometric pressure can be
obtained from a nearby National Weather Service Station, in which
case the station value (which is the absolute barometric pressure)
should be requested. The tester should be aware that the pressure is
normally corrected to sea level by the weather station; the uncor-
rected readings should be obtained. An adjustment for differences in
elevation of the weather station and the sampling location is applied
at a rate of -2.5 mm Hg/30 m (-0.1 in. Hg/100 ft) of elevation
increase, or vice versa for elevation decrease.
Accuracy can be ensured by checking the field barometer against a
mercury-in-glass barometer or its equivalent. If the field barometer
cannot be adjusted to agree with the mercury-in-glass barometer, it
is not acceptable.
1.1.10 Vacuum Gauge - At least one 760-mm (29.92-in.) Hg gauge is
necessary to leak check the sampling train. An acceptable vacuum
-------
Section No. 3.15.1
Date July 1, 1986
Page 10
gauge, when checked in a parallel leakless system with a mercury U-
tube manometer at 250-mm (10-in.) Hg vacuum, will agree within 25 mm
(1.0 in.) Hg.
1.2 Sample Recovery Apparatus
1.2.1 Wash Bottles - Two 500-ml polyethylene or glass wash bottles
are needed for quantitative recovery of collected samples.
1.2.2 Storage Bottles - One 1-L polyethylene bottle is required to
store each collected sample. An additional polyethylene bottle is
necessary to retain a blank for each absorbing solution used in
testing. Wash and storage bottles should be visually checked for
damage. CAUTION: Each storage bottle seal should be checked prior
to use to ensure that leakage will not occur.
1.2.3 Funnel and Stirring Rods - The analyst may find a glass funnel
and glass stirring rods are helpful in transferring the absorbing
reagent to and from the restricted orifice impingers. The flow of
absorbing reagent can be controlled by pouring along the glass
stirring rod.
1.3 Apparatus for Sample Preparation and Analysis
1.3.1 Magnetic Stirrer with Magnetic Stirring Bars - The magnetic
stirrer and stirring bars are used for the removal of excess
permanganate ion. The stirring bars should be Teflon™-coated owing
to the corrosiveness of the alkaline-permanganate solution. Stirring
bars having dimensions 25 mm by 10 mm are recommended. Smaller
stirring bars can be expected to be less efficient because of the
resistance offered by the absorbing reagent, which is relatively
viscous.
Manual stirring is acceptable; however, because it is tedious and
laborious, this is not recommended.
1.3.2 Filtering Flask - One filtering flask having a 500-ml capacity
is needed to filter the liquid sample after the excess permanganate
ion has been removed.
1.3.3 Buchner Funnel - The Buchner funnel is used with the filtering
flask for the filtering operations. A convenient size funnel is one
with a 75-mm internal diameter. The analyst may wish to attach a
section of Teflon™tubing to the funnel's spout in order to prevent
loss of sample via the side arm of the flask during filtration. The
use of a trap located between the filtering flask and the vacuum
source is also recommended in order to prevent inadvertant sample
losses. Upstream tubing connections for the trap should be Teflon™.
1.3.4 Filter - Whatman GF/C glass microfiber discs are used in the
Buchner funnel. For the funnel size recommended above, the \
-------
Section No. 3.15.1
Date July 1, 1986
Page 11
applicable disc diameter is 7.0 cm. This filter is specified because
it performs well on materials having a small particle size. The
material filtered from the sample is primarily manganese dioxide
which exists in very small particles when suspended in water.
In general, other types of filters are unsuitable owing either to
clogging or to their inability to filter the manganese dioxide
particles effectively. The analyst should note that particulate
matter must be removed from the sample in order to avoid damage to
syringes and the ion chromatograph.
1.3.5 Vacuum Source - A vacuum source is needed for the filtering
operations. Either a water aspiration system or a vacuum pump can be
used. If a vacuum pump is used, it should be protected by a trap
installed at an upstream position.
1^3.6 Funnel and Stirring Rods - The analyst may find a funnel and
glass stirring rods are helpful in transfering the sample aliquot to
the Erlenmeyer flask prior to removal of the excess permanganate ion.
1.3.7 Volumetric Flask - One volumetric flask having the Class-A
designation and a 250-ml capacity is needed for each sample and
blank. As a practical matter, samples should be stored in the flasks
for_ a minimum period of time owing to the fact that hydroxide ions
(OH~) will attack the glass and can also cause frozen ground-glass
fittings.
1.3.8 Pipettes - A 50-ml Class-A pipette is needed for taking a
sample aliquot. A 5-ml pipet is usually used for adding (not
quantitatively) hydrogen peroxide to the sample aliquot in order to
remove excess permanganate ion. Because hydroxide ion (OH~), which
is present in the sample, can attack glass, it is recommended that
analysts rinse pipettes with water immediately after use on samples.
If Quality Assurance Audit Samples are to be analyzed, additional
pipettes (Class-A) may be needed.
1.3.9 Erlenmeyer Flasks - Erlenmeyer flasks having a 250-ml capacity
are used for operations involving the removal of excess permanganate
ion in the samples.
1.3.10 Ion Chromatograph - An ion chromatograph (1C) is used for
analyzing the samples. The instrument should, at a minimum, have the
components described below.
Columns - The 1C should be equipped with an ion separator column
capable of resolving nitrate ion (N03~) from sulfate ion (S04=),
which may be found in samples acquired at fossil-fuel-fired steam
generators. In addition, it should be capable of detecting and
resolving nitrite ion (NO-')- Either suppressed or nonsuppressed
IC's may be used provided that performance meets the above criteria.
Suppressed IC's should be equipped with an acid (H ) suppressor
column in addition to the anion separator column. Suppressor columns
-------
Section No. 3.15.1
Date July 1, 1986
Page 12
(fiber preferred over packed bed) are generally produced as
proprietory items; however, an acceptable column can be made using
the resin available from BioRad Company, 32nd and Griffin Streets,
Richmond, California.
Pump - The pump must be capable of maintaining a steady eluent
flow as required by the system.
Flow Gauges - These must be capable of measuring the specified
eluent flow rate. It is recommended that the gauge be calibrated
upon receipt.
Conductivity Detector - It is recommended that the detector be
calibrated according to manufacturer's procedures prior to initial
use.
Recorder - It should be compatible with the output voltage of the
detector.
1.3.11 Analytical Balance - One analytical balance that weighs to
0.1 mg and a set of Class-S calibration weights to check the accuracy
of the balance (+0.3 mg) upon receipt are needed. The balance should
be serviced or returned to the manufacturer if agreement cannot be
met.
1.4 Reagents - Unless otherwise indicated, it is intended that all
reagents conform to the specifications established by the Committee
on Analytical Reagents of the American Chemical Society (ACS), where
such specifications are available; otherwise, use the best grade
available.
1.4.1 Sampling - For sampling, the following are needed.
Absorbing solution - The absorbing solution is prepared by dis-
solving 40.0 g potassium permanganate (KMnO.) and 20.0 g sodium
hydroxide (NaOH) in 940 ml of water. The solution's concentration is
4.0 percent (w/w) KMnO., 2.0 percent (w/w) NaOH. CAUTION: Extreme
care should be taken in Handling the KMnO, reagent and the absorbing
solution. KMnO. is a strong oxidant and is incompatible with
substances containing carbon such as paper, fabric, and human
tissue. It is recommended that eye protection be worn when handling
the absorbing solution. Skin exposed to the absorbing solution
should be washed with plenty of water and until the exposed area no
longer exhibits a soapy feeling.
Water - Water should be used which conforms with ASTM specifica-
tion D1193-82, Type III. Type III water is prepared by distillation,
ion exchange, reverse osmosis, or a combination thereof, followed by
polishing with a 0.45 ym membrane filter. The specifications for
Type III water are shown below.
-------
Section No. 3.15.1
Date July 1, 1986
Page 13
Specifications for ASTM D1193 - 82, Type III Water
Total matter, max., (mg/L) 1.0
Electrical conductivity, max., 1.0
(umho/cm) at 25 C
Electrical resistivity, min., 1.0
(umho/cm) at 25 C
pH at 25°C 6.2 to 7.5
Minimum color retention time 10
of KMn04, (min)
Maximum soluble silica, (yg/L) 10
Note: Mention of "water" anywhere in this Section (3.15) refers to
ASTM D1193-82, Type III Water as described above. By using water
from the same source for making reagents, calibration standards, and
eluents for the ion chromatograph, the presence of trace quantities
of nitrate in the water will be negated. Therefore, a water blank
correction is not necessary in the development of the calibration
curve.
Stopcock Grease - An acetone insoluble, heat stable, silicone
grease must be used when sealing of ground-glass connections is
required.
1.4.2 Analysis - For analysis, the following reagents are required.
Water - See Subsection 1.4.1 above.
Hydrogen Peroxide - Five (5) percent (v/v) hydrogen peroxide (H_O2)
is used which is prepared by mixing 1 part 30% (v/v) H202 with 5
parts water.
Reagent Blank - The reagent blank may be prepared by dissolving
2.4 g KMnO. and 1.2 g NaOH in 96 ml water. Alternatively, the blank
may be prepared by diluting 60 ml of the absorbing reagent to 100 ml
using water.
Potassium Nitrate (KNO^) Standard Solution - The following
procedure is observed to prepare the KNO~ standard solution.
1. Dry an adequate amount of KN03 at 110°C for about 2 hours; then
transfer to a desiccator, ana allow to cool to laboratory
temperature.
2. Using an analytical balance, accurately weigh 9 to 10 g of the
dried KN03 to the nearest 0.1 mg.
-------
Section No. 3.15.1
Date July 1, 1986
Page 14
3. Transfer the KN03 to a suitable container, such as a beaker,
dissolve the KN03 in water, and transfer all of the KN03 solution to
a 1-L volumetric flask.
4. Dilute the KN03 solution to the 1-L mark with water.
The N03~ concentration of the standard solution is calculated from
the mass of KN03 using the following relationship:
NO ~ /Mass of KNOJV/106 vg\/ L W 62.01 g/mol NO " \
Concentration =\ (g) )\ g /\1Q3 mlMl01.10 g/mol KNO3 f
(yg/ml)
Method 7D states that the KN03 standard solution is stable for 2
months without preservative at laboratory conditions. Novice
analysts should note that certain microbes feed on N03 solutions
with the consequence for Method 7D being that NO results will be
biased high. For this reason, standard solutions snould be disposed
of after 2 months.
Eluent Solution - For IC's involving the suppressedgtechnique, an
eluent solution being 3 x 10 M NaHC03 and 2.4 x 10 M Na2C03 has
proved adequate for Method 7D applications. This eluent is prepared
by taking 1.008 g NaHC03 and 1.018 g Na2C03 and dissolving them in
4 L water.
Other eluents may be used provided that they are capable of
resolving N0~~ from SO." and other ions which may be present in
samples.
Quality Assurance Audit Samples - Quality Assurance Audit Samples
are required to be analyzed in conjunction with field samples when
Method 7D is used to demonstrate compliance with EPA's New Source
Performance Standards in 40 CFR Part 60. The audit samples for
Method 7D are essentially the same as those described in Method 7,
Section 3.3.9. Because the analytical range for Method 7D differs
from that for Method 7, analysts requesting audit samples should
specify that samples be appropriate for Method 7D.
-------
Section No. 3-15-1
Date July 1, 1986
Page 15
Table 1.1. ACTIVITY MATRIX FOR PROCUREMENT OF APPARATUS AND SUPPLIES
Apparatus and
supplies
Acceptance limits
Frequency and method
of measurement
Action if
requirements
are not met
Sampling probe
with heating
system
Capable of maintaining
100°C (212°F) exit
air at flow rate of
500 cc/min
Visually check and
run heating system
checkout
Repair, or
return to
supplier
Restricted
orifice
impingers
Standard stock glass;
ensure that dimensions
conform with specifi-
cations
Visually check upon
receipt for breaks
or leaks
Return to manu-
facturer
Vacuum pump
Capable of maintaining
flow rate of 400 to
500 cc/min; leak free
at 250 mm (10 in.) Hg
Check upon receipt
for leaks and capacity
As above
Dry gas meter
Capable of measuring
total volume within
2% at a flow rate of
500 cc/min
Check for damage upon
receipt, and calibrate
(Sec. 3.15.2) against
wet test meter
Reject if dam-
aged , behaves
erratically,
or cannot be
properly
adjusted
Wet test meter
Capable of measuring
total volume within
2% at a flow rate of
500 cc/min
Upon assembly, leak
check all connections,
and check calibration
by liquid displacement
As above
Rotameter
Within 5% of manufac-
turer's calibration
curve (recommended)
Check upon receipt for
damage, and calibrate
(Sec. 3.15.2) against
wet test meter
Recalibrate,
and construct
a new calibra-
tion curve
Drying tube
Minimum capacity of
30 to 50 g of silica
gel
Visually check upon
receipt for damage and
proper size
Return to
supplier
(continued)
-------
Table 1.1 (continued)
Section No. 3.15.1
Date July 1, 1986
Page 16
Apparatus and
supplies
Acceptance limits
Frequency and method
of measurement
Action if
requirements
are not met
Thermometers
Within 1°C (2°F)
true value in the range
of 0°C to 25°C
(32° to 77°F)
for impinger and within
3°C (5^°F) for
dry gas meter thermom-
eter
Check upon receipt for
damage (i.e., dents and
bent stem), and
calibrate (Sec. 3.15.2)
against mercury-in-
glass thermometer
Return to
supplier if
unable to
calibrate
Barometer
Capable of measuring
atmospheric pressure
to within 2.5 mm
(0.1 in.) Hg calibrate
Check against mercury-
in-glass barometer or
equivalent (Sec. 3.5-2)
Determine cor-
rection factor,
or reject if
difference is
more than
2.5 mm
Vacuum gauge
Wash bottles
Storage
bottles
Pipettes and
volumetric
flasks
0 to 760 mm (0 to
29.92 in.) Hg range,
^25 mm (1.0 in.) Hg
accuracy at 250 mm
(10 in.) Hg
Check against U-tube
mercury manometer
upon receipt
Adjust, or re-
turn to
supplier
Polyethylene or glass,
500-ml
Visually check for
damage upon receipt
Replace, or re-
turn to
supplier
Polyethylene, 1-L
Visually check for dam-
age upon receipt, and
be sure that caps seal
properly
As above
Glass, Class-A
Upon receipt, check for
stock number, cracks,
breaks, and manufac-
turer flaws
As above
Water
Must conform to ASTM-
D1193-82, Type III
Check each lot or spec-
ify type when ordering
As above
(continued)
-------
Table 1.1 (continued)
Section No. 3.15.1
Date July 1, 1986
Page 17
Apparatus and
supplies
Acceptance limits
Frequency and method
of measurement
Action if
requirements
are not met
Stopcock
grease
High vacuum, high temp-
erature chlorofluoro-
carbon grease
Visually check upon
receipt
Return to
supplier, and
note in procure-
ment log
Analytical
balance
Capable of measuring
to +0.1 mg
Check with standard
weights upon receipt
and before each use
Replace, or
return to manu-
facturer
Ion Chroma-
tograph
1. Columns
2. Pump
3. Flow
gauges
1. Capable of giving
nitrate ion peaks with
baseline separation
1. Check during
analyses
2. Capable of deliv-
ering eluent at con-
stant and repeatable
flow rate
2. Check during
analyses by monitor-
ing flow rate
3. Capable of giving
repeatable indications
of eluent flow rate
3. Check calibration
and repeatability
upon receipt
1. Consult op-
erators ' manual;
regenerate sup-
pressor column;
clean separator
column; check
performance of
components below;
replace column(s)
if above actions
are unsuccessful
2. Consult op-
erator's manual;
oil, clean, re-
repair, replace,
or return to man-
ufacturer; check
tubing of ion
chromatograph for
leaks or ob-
structions;
check flow meter
performance
3. Consult oper-
ator 's manual;
adjust, repair,
replace, or
return to
manufacturer
(continued)
-------
Table 1.1 (continued)
Section No. 3.15.1
Date July 1, 1986
Page 18
Apparatus and
supplies
4 . Conduc-
tivity
detector
5 . Recorder
Hydrogen per-
oxide
Potassium
nitrate
Sodium carbonate
Sodium bicarbon-
ate
Sodium hydroxide
Potassium
permanganate
Acceptance limits
4. Capable of giving
responses which can be
manually or electron-
ically integrated
within a precision of
5 percent
5. As above, if used
record responses for
manual integration
30/li aqueous solution,
ACS reagent grade
ACS reagent grade
ACS reagent grade
ACS reagent grade
ACS reagent grade
ACS reagent grade
Frequency and method
of measurement
4. Calibrate accord-
ing to manufacturer's
instructions prior to
use
5. Check during
analyses
Check each lot, or
specify type when
ordering
As above
As above
As above
As above
As above
Action if
requirements
are not met
4. Consult
operator ' s
manual ; repair
replace, or
return to
manufacturer
5. Consult
operator's
manual; adjust
speed
Replace or
return to
manufacturer
As above
As above
As above
As above
As above
-------
Section No. 3.15.2
Date July 1, 1986
Page 1
2.0 CALIBRATION OF APPARATUS
Calibration of the apparatus is one of the most important
functions in maintaining data quality. The detailed calibration
procedures included in this section were designed for the equipment
specified in Method 7D and described in the previous section. Table
2.1 at the end of this section summarizes the quality assurance
functions for calibration. All calibrations should be recorded on
standardized forms and retained in a calibration log book.
The calibration procedures and considerations addressed in this
section are those which are unique to Method 7D. The sampling phase
of Method 7D involves the use of equipment that is essentially the
same as that used for Method 6. The analysis phase of Method 7D
entails the use of an ion chromatograph, an instrument that also is
used for Method 7A. The Handbook user should note that: (a) the
standard used for Method 7A is sodium nitrate (NaN03, while for
Method 7D the standard used is potassium nitrate (KNo3); and (b)
sulfate ion (SO,") peaks in ion chromatograms for Method 7D will have
a lesser tendancy to overlap and therefore to interfere with nitrate
(N03~) peaks because SO." will exist at a lower concentration because
it originates only from sulfur oxides in the effluent.
2.1 Metering System
2.1.1. Wet Test Meter - The wet test meter must be calibrated and
have the proper capacity. For Method 7D, the wet test meter should
have a capacity of at least 1 L/min. No upper limit is placed on the
capacity; however, the wet test meter dial should make at least one
complete revolution at the specified flow rate for each of the three
independent calibrations.
Wet test meters are calibrated by the manufacturer to an accuracy
of +2%. Calibration of the wet test meter must be checked initially
upon receipt and yearly thereafter.
The following liquid positive displacement technique can be used
to verify and adjust, if necessary, the accuracy of the wet test
meter to +2%:
1. Level the wet test meter by adjusting the legs until the
bubble on the level located on the top of the meter is centered.
2. Adjust the water volume in the meter so that the pointer in
the water level gauge just touches the meniscus.
3. Adjust the water manometer to zero by moving the scale or by
adding water to the manometer.
4. Set up the apparatus and calibration system as shown in
Figure 2.1.
-------
Section No. 3.15.2
Date July 1, 1986
Page 2
MANOMETER
THERMOMETER
AIR INLET
WATER
LEVEL
GAUGE
VALVE
2000-ml LINE
TYPE-A
VOLUMETRIC
FLASK
Figure 2.1. Calibration check apparatus for wet test meter.
-------
Section No. 3.15.2
Date July 1, 1986
Page 3
a. Fill the rigid-wall 5-gal jug with water to below the
air inlet tube. Put water in the impinger or saturator,
and allow both to equilibrate to room temperature (about
24 hours) before use.
b. Start water siphoning through the system, and collect
the water in a 1-gal container, located in place of the
volumetric flask.
5. Check operation of the meter as follows:
a. If the manometer is reading <10 mm (0.4 in.) H-O, the
meter is in proper working condition. Continue to step
6.
b. If the manometer reading is >10 mm (0.4 in.) H.-0, the
wet test meter is defective. If the wet test meter is
defective, and if the defects(s) (e.g., bad connections
or joints) cannot be found and corrected, return it to
the manufacturer for repair.
6. Continue the operation until the 1-gal container is almost
full. Plug the inlet to the saturator. If no leak exists, the flow
of liquid to the gallon container should stop. If the flow
continues, correct for leaks. Turn the siphon system off by closing
the valve, and unplug the inlet to the wet test meter.
7. Read the initial volume (V.) from the wet test meter dial,
and record on the wet test meter calibration log, Figure 2.2.
8. Place a clean, dry volumetric flask (Class-A) under the
siphon tube, open the pinch clamp, and fill the volumetric flask to
the mark. The volumetric flask must be large enough to allow at
least one complete revolution of the wet test meter with not more
than two fillings of the volumetric flask.
9. Start the flow of water, and record the maximum wet test
meter manometer reading during the test after a constant flow of
liquid is obtained.
10. Carefully fill _the volumetric flask, and shut off the liquid
flow at the 2-L mark. Record the final volume shown on the wet test
meter.
11. Steps 7 through. 10 must be performed three times.
Since the water temperature in the wet test meter and reservoir
has been equilibrated to the ambient temperature and the pressure in
the wet test meter will equilibrate with the water reservoir after
the water flow is shut off, the air volume can be compared directly
with the liquid displacement volume. Any temperature or pressure
-------
test meter serial number
4 5 ~~
Date 2-/2.1/&4-
Satisfactory leak check?
. ~7sf ^ r~
Ambient temperature of equilibrate liquid in wet test meter and reservoir /y- /""
Range of wet test meter flow rate 0~~12-0 (—/^ir\
Volume of test flask Vg = -2- > 00 L-
Test
number
1
2
3
Manometer
reading, a
mm H2O
r
5~
^~m s s
(+1%).
Signature of calibration person
V O OT
o>a>a>
(Q ft O
0) 0> rt
H-
t>-<-iO
C 3
Figure 2.2. Wet test meter calibration log.
w
M •
VD (-•
oo cn
CT» •
to
-------
Section No. 3.15.2
Date July 1, 1986
Page 5
difference would be less than measurement error and would not affect
the final calculations.
The error should not exceed +^1%; if this error magnitude is
exceeded, check all connections within the test apparatus for leaks,
and gravimetrically check the volume of the standard flask. Repeat
the calibration procedure, and if the tolerance level is not met,
adjust the liquid level within the meter (see the manufacturer's
manual) until the specifications are met.
2.1.2 Sample Metering System - The sample metering system, con-
sisting of the drying tube, needle valve, pump, rotameter, and dry
gas meter, is initially calibrated by stringent laboratory methods
before it is used in the field. The calibration is then rechecked
after each field test series. This recheck requires less effort than
the initial calibration. When a recheck indicates that the calibra-
tion factor has changed, the tester must again perform the complete
laboratory procedure to obtain the new calibration factor. After the
meter is recalibrated, the metered sample volume is multiplied by the
calibration factor (initial or recalibrated) that yields the lower
gas volume for each test run.
Initial Calibration - The metering system should be calibrated
when first purchased and at any time the posttest check yields a
calibration factor that does not agree within 5% of the pretest
calibration factor. A calibrated wet test meter (properly sized,
with +1% accuracy) should be used to calibrate the metering system.
The metering system should be calibrated in the following manner
before its initial use in the field.
1. Leak check the metering system (drying tube, needle valve,
pump, rotameter, and dry gas meter) as follows:
a. Temporarily attach a suitable rotameter (e.g., 0-40 cm /
min) to the outlet of the dry gas meter, and place a
vacuum gauge at the inlet to the drying tube.
b. Plug the drying tube inlet. Pull a vacuum of at least
250 mm (10 in. ) Hg.
c. Note the flow rate as indicated by the rotameter.
d. A leak of <0.02 L/min must be recorded or leaks must be
eliminated.
e. Carefully release the vacuum gauge before turning off
pump.
2. Assemble the apparatus, as shown in Figure 2.3, with the wet
test meter replacing the drying tube and impingers; that is, connect
the outlet of the wet test meter to the inlet side of the needle
-------
THERMOMETER
MANOMETER
AIR INLET
Figure 2.3. Sample metering system calibration setup.
•a o OT
0) 0) (D
(Q rt o
a> a> rt
H-
a\ Q o
c 3
VD H"
00 CJ1
CT» •
to
-------
Section No. 3.15.2
Date July 1, 1986
Page 7
valve and the inlet side of the wet test meter to a saturator which
is open to the atmosphere. Note: Do not use a drying tube.
3. Run the pump for 15 minutes with the flow rate set at
450 cc/min to allow the pump to warm up and to permit the interior
surface of the wet test meter to become wet.
4. Collect the information required in the forms provided,
Figure 2.4A (English units) or 2.4B (metric units), using sample
volumes equivalent to at least five revolutions of the dry test
meter. Three independent runs must be made.
5. Calculate Y. for each of the three runs using Equation 2-1.
Record the values in the form (Figure 2.4A or 2.4B).
D \
m \
+ 1 O £ I t 4- -L-O ^rtO^ 'M-ioO,
13.6/ (t, +
Yi ='Vw Pm + 13-6/ (td + 460WF or 273WC) Equation 2-1
V, P (t + 460°F or 273°C)
. d m w
where:
Y. = ratio for each run of volumes measured by the wet
test meter and dry gas meter, dimensionless
calibration factor,
3 3
V = volume measured by wet test meter, m (ft ),
Wf
P = barometric pressure at the meters, mm (in.) Hg,
D = pressure drop across the wet tost meter, mm (in.) H20,
t, = average temperature of dry gas meter, °C (°F),
3 3
V, = volume measured by the dry gas meter, m (ft ), and
t = temperature of wet test meter, °C (°F).
6. Adjust and recalibrate or reject the dry gas meter if one or
more values of Y± fall outside the interval Y _+0.02Y, where Y is the
average for three runs. Otherwise, the Y (calibration factor) is
acceptable and is to be used for future checks and subsequent test
runs. The completed form should be forwarded to the supervisor for
approval, and then filed in the calibration log book.
Posttest Calibration Check - After each field test series,
conduct a calibration check as in Subsection 2.1.2 with the following
exceptions:
-------
Date
2-/2-Z-
Calibrated by
Barometer pressure, P
u
Meter box number ££ / Wet test meter number /Of -/}•
in. Hg Dry gas meter temperature correction factor
Wet test
meter
pressure
drop
.b
ft3
/. 0$&
/.on
/.0(,l
Dry test meter
gas volume
(vd),b ft3
Initial
72 5". M3
n72.
730.02!
733. /£&
Wet test
meter
gas temp
(tw),
°F
72-
72-
72-
Inlet
gas
temp
°F
76
60
80
Average
gas temp
| >C
oF
77
&(
32-
Time
of run
(0),d
rain
^g
£6
66
Average
ratio
(Yt),6
/.O/J
1,017-
/,o^o
(Yri),f
OMt,
0.110
O.W4-
3 Dm exPresse(^ as negative number.
Volume passing tbrougb meter. Dry gas volume is minimum for at least five revolutions of the meter.
c The average of t. and t. if using two thermometers; the actual reading if using one thermometer.
H di do
. The time it takes to complete the calibration run.
e With Y defined as the average ratio of volumes for the wet test and the dry test meters, Y. = Y ^0.02 Y
calibration and Y = Y +0.05 Y for the posttest checks; thus,
for
Vw (td + 460°F) [pm + (Dm/13.6)]
Vd (tw
(P)
?q. 1) and
With Y defined as the average ratio of volumetric measurement by wet test meter to rotameter.
Tolerance Y = 1 +0.05 for calibration and Y +0.1 for posttest checks.
9 (fcw + 460°F> (pm>
(Eq. 3)
and Yr =
o,
. 4)
Figure 2.4A. Dry gas meter calibration data form (English units).
•a O w
0) 0) 0)
tQ rt- O
0) •
to
-------
Date 2 /2~2y&f Calibrated by
ft
Barometer pressure, Pm =
Meter box number
in. Hg
/ Wet test meter number /Of •
Dry gas meter temperature correction factor /1//4 °C
Wet test
meter
pressure
drop
0>m>.a
mm H~0
6.4
£.4
£.4
Rota-
meter
setting
(R8),
cc/min
4-50
4-50
450
Wet test
meter gas
volume
L
£?.75B
Z?.?gf
30.04^
Dry test meter
gas volume
(Vd),b L
Initial
/0s:fr3/
14$. $1*2-
X6/.6/?
Final
/3ST6/0
X70.377
2JI.W
Wet test
meter
gas temp
(tw),
°C
Z?-
2,2-
Z-2-
Inlet
gas
temp
Average
gas temp
(td),c
°C
Z£
68
(08
Average
ratio
(V'6
/.Ol^
/,o/e
I.OZ+
(Yri),f
AW
o.m
1.004-
Dra exPressed as negative number.
Volume passing through meter. Dry gas volume is minimum for at least five revolutions of the meter.
The average of t, and t, if using two thermometers; the actual reading if using one thermometer.
di do
The time it takes to complete the calibration run.
e'With Y defined as the average ratio of volumes for the wet test and the dry test meters,
calibration and Y* = Y +0.05 Y for the posttest checks; thus,
Y +0.02 Y for
Vw (td + 273°C)
(Dm/13.6)]
Yl -
(t
(Pm)
(Eq. 1)
and
i. on . (Eq-2)
With Y defined as the average ratio of volumetric measurement by wet test meter to rotameter.
Tolerance Yr = 1 +0.05 for calibration and Y +0.1 for posttest checks.
v~ (t, + 273°C) P^ + (D_/13.6) ,™J Y,
»O D w
0) 0> (D
-------
Section No. 3.15.2
Date July 1, 1986
Page 10
1. The leak check is not conducted because a leak may have been
corrected that was present during testing.
2. Three or more revolutions of the dry gas meter may be used.
3. Only two independent runs need be made.
4. If a temperature-compensating dry gas meter was used, the
calibration temperature for the dry gas meter must be within 6 C
(10.8 F) of the average meter temperature observed during the field
test series.
When a lower meter calibration factor is obtained as a result of
an uncorrected leak, the tester should correct the leak and then
determine the calibration factor for the leakless system. If the new
calibration factor changes the compliance status of the facility in
comparison to the lower factor, either include this information in
the report or consult with the Administrator for reporting proce-
dures. If the calibration factor does not deviate by >5% from the
initial calibration factor Y (determined in Subsection 2.1.2), then
the dry gas meter volumes obtained during the test series are accept-
able. If the calibration factor does deviate by >5%, recalibrate the
metering system as in Subsection 2.1.2; for the calculations, use the
calibration factor (initial or recalibration) that yields the lower
gas volume for each test run.
2.2 Thermometer
The thermometer(s) on the dry gas meter inlet used to measure the
metered sample gas temperature should be initially compared with a
mercury-in-glass thermometer that meets ASTM E-l No. 63C or 63F
specifications:
1. Place the dial type or an equivalent thermometer and the
mercurv-in-glass thermometer in a hot water bath, 40 C to 50°C (104
to 122 F). Compare the readings after the bath stabilizes.
2. Allow both thermometers to come to room temperature. Compare
the readings after the thermometers stabilize.
3. The dial type or equivalent thermometer is acceptable if (1)
values agree within 3 C (5.4 F) at both points (steps 1 and 2 above)
or (2) if the temperature differentials at both points are within 3 C
(5.4 F) and the temperature differential is taped to the thermometer
and recorded on the meter calibration form (Figure 2.4A or 2.4B).
4. Prior to each field trip, compare the temperature reading of
the mercury-in-glass thermometer at room temperature with that of the
thermometer that is part of the metering system. If the values or
the corrected values are not within 6 C (10.8 F) of each other,
replace or recalibrate the meter thermometer.
-------
Section No. 3.15.2
Date July 1, 1986
Page 11
2.3 Rotameter
Method 7D recommends (optional) that the tester calibrate the
rotameter prior to each test. Before being sent to the field, the
rotameter should be cleaned and maintained according to the
manufacturer's instructions. For this reason, it is recommended
(optional) that the calibration curve and/or rotameter markings be
checked upon receipt and then routinely checked with the posttest
metering system check. The rotameter may be calibrated as follows:
1. Ensure that the rotameter has been cleaned as specified by the
manufacturer and is not damaged.
2. Use the manufacturer's calibration curve and/or markings on
the rotameter for the initial calibration. Calibrate the rotameter
as described in the metering system calibration of Subsection 2.1.2,
and record the data on the calibration form (Figure 2.4A or 2.4B).
3. Use the rotameter for testing if the pretest calculated
calibration is within 450 _+25 cc/min. If, however, the calibration
point is not within 5%, "determine a new flow rate setting, and
recalibrate the system until the proper setting is determined.
4. Check the rotameter calibration with each posttest metering
system check. If the rotameter check is within 10% of the 450 cc/min
setting, the rotameter is acceptable. If, however, the check is not
within 10% of the flow setting, disassemble and clean the rotameter,
and perform a full recalibration.
2.4 Barometer
The field barometer should be adjusted initially and before each
test series to agree within 2.5 mm (0.1 in.) Hg with a mercury-in-
glass barometer or with the pressure value reported from a nearby
National Weather Service Station and corrected for elevation. The
tester should be aware that the pressure readings are normally
corrected to sea level. The uncorrected readings should be
obtained. The correction for the elevation difference between the
weather station and the sampling point should be applied at a rate of
-2.5 mm Hg/30 m (-0.1 in. Hg/100 ft) elevation increase, or vice
versa for elevation decrease.
The calibration checks should be recorded on the pretest sampling
form (Figure 2.5).
2.5 Analytical Balance
The analytical balance used to weigh the reagents for the nitrate
stock standard should be calibrated by the following procedure:
1. Zero the balance.
-------
Section No. 3.15.2
Date July 1, 1986
Page 12
Date Z-Z^- Calibrated by
' / "
Meter box number £.£.' — /
Dry Gas Meter*
Pretest calibration factor = /.&ZQ (within 2% of average factor
for each calibration run).
Rotameter
Pretest calibration factor (Y ) or setting = /,OQ (between
400 and 500 cc/min).
Dry Gas Meter Thermometer
Was a pretest temperature correction made? yes X no
If yes, temperature correction (within 3°C (5.4°F) of
reference values for calibration and within 6 C (10.8 F) of
reference values for calibration check).
Barometer
Was the pretest field barometer reading correct? X yes no
(within 2.5 mm (0.1 in.) Hg of mercury-in-glass barometer).
*Most significant items/parameters to be checked.
Figure 2.5. Pretest sampling checks.
-------
Section No. 3.15.2
Date July 1, 1986
Page 13
2. Place a 5-g Class-S weight on the balance. Record the
balance reading for the 5-g weight.
3. Place a 10-g Class-S weight on the balance. Record the
balance reading for the 10-g weight.
4. The balance readings for the 5-g and 10-g weights must be
within 2 mg of the actual weights.
5. If the balance readings are greater than +2 mg either of the
actual weights, repair the balance or contact the balance
manufacturer.
2.6 Ion Chromatograph System
2.6.1 Performance Check of the Ion Chromatograph - Method 7D states
that the instrument used for analysis should provide adequate
resolution of NOo" and should be able to resolve and detect nitrite
ion (N02 ). It is recommended that the instrument be performance
checked prior to initial use to ensure that the instrument can meet
the above criteria.
Method 7D does not quantify the criteria for acceptable instru-
ment performance. The numerical limits and procedures given below
are offered from a purely technical viewpoint. Their observance
should ensure that the instrument conforms with the method, but
should not be interpreted as a requirement. Considerations for
preparing the ion Chromatograph for analysis follow.
Conductivity Detector - Prior to its initial use, the con-
ductivity detector of the ion Chromatograph (1C) should be calibrated
by the method described in the operator's manual. Following this
calibration it is highly recommended that the analyst conduct a
preliminary calibration of the 1C. Because the 1C calibration is
conducted concurrently with the field sample analysis when performing
Method 7D, the full discussion of the calibration procedure is
presented in Section 3.15.5.
Recorder - A strip chart recorder compatible with the output
voltage range of the conductivity detector may be used to record the
ion chromatogram. Manual measurement techniques that can be used for
quantitation of the chromatogram include (a) peak height, (b) peak
area by triangulation, (c) peak area by multiplying peak height times
the peak width at half-height, (d) peak area by cutting out the peak
from the chromatogram and weighing it on an analytical balance, and
(e) peak area by planimetry.
The use of an electronic integrator, if available, is recommended
for greater accuracy and precision. The electronic integrator can be
used in the peak area mode when the integration parameters are set up
properly. The key integration parameters for peak area determination
concern the identification of the beginning and end of a peak and the
-------
Section No. 3.15.2'
Date July 1, 1986
Page 14
placement of the baseline under the peak. Analysts should carefully
read the operator's manual and understand the selection and set up of
the integration parameters for their particular integrator. The
electronic integrator can also be used in the peak height mode
provided that the peaks are symmetrical and an acceptable standard
calibration curve can be generated.
Sample Injection Device Contamination Check - The analyst is
encouraged to check the sample injection device for contamination by
injecting water before the calibration standards are analyzed.
Contaminants will appear as peaks on the chromatogram. Repeated
injections of water should be used to remove contaminants from the
sample injection device. If certain peaks remain after several
injections, then the water may be contaminated and should be
replaced.
Separation of Nitrate, NO^ - To ensure accurate results from
the ion chromatographic analysis, baseline separation of the N03~
peak from the other ion peaks should be achieved. A source of SO."
in a sample may be sulfur dioxide present in the effluent stream
sample. Figures 2.6a and_ 2.6b show two chromatograms, one having
overlapping N0« and SO." peaks, and the other having baseline
separation of the N03 ana SO." peaks.
The analyst is encouraged to check the performance of the ion
chromatograph system before analyzing samples in order to ensure that
baseline separation of NO3 is attainable.
The ion chromatograph can be performance checked using a solution
containing NO~ and_ SO. for_ compliance purposes or a solution
containing N03 , N0o~' ^nc* SO." if the nitrate is to be quantified.
A solution that will provide rigorous conditions involves the use of.
KNO~ working standard solution £described in Section 3.15.4, page 3)
and NO2~ (if applicable) and SO." solutions, the preparation of which
are addressed below.
The S0.~ solution is prepared as follows: Weigh out 0.231 of
sodium suitate (Na2SO.), and transfer it to a beaker. Dissolve the
Na2SO. in water, quantitatively transfer the solution to a 250-ml
volumetric flask, and finally, dilute to the mark with water.
The concentration of the solution is 625 vg SO."/ml. Sodium
sulfate (Na2SO.) is a component of the pusher solution in the Orsat
apparatus used for Method 3. It is not special and has been chosen
because of its probable availability. Other SO." reagents can be
used.
If the nitrite is to be quantified, then separation of the
nitrate peak should also be checked. To prepare the N02 stock
solution, first weigh out 52.5 mg sodium nitrite (NaN02) and
-------
Section No. 3.15.2
Date July 1, 1986
Page 15
so,
WATER DIP
Figure 2.6a. Example chromatogram having overlapping peaks.
Figure 2.6b. Example chromatogram showing baseline
separation of peaks.
-------
Section No. 3.15.2
Date July 1, 1986
Page 16
transfer it to a beaker. Dissolve the NaN02 in water, quantitatively
transfer it to a 250-ml volumetric flask, and finally, dilute to the
mark with water. To prepare the N02~ working solution, pipet 10.0 ml
of the stock solution into a 100-ml volumetric flask, and dilute to
volume_with water. The concentration of the working solution is 14
yg N02~/ml.
To prepare the performance check solution, pipet 10 ml of the KN03
working standard solution, 8 ml of the SO." solution, and 1 ml of thS
N02~ working solution (if applicable) into a 200-ml volumetric flask,
ana dilute to the mark with water.
The_concentration of N03~ in the performance check sample is 7.5
yg NO3~/ml, which corresponds to a NO level around the emission
standard for coal-fired boilers subject to 40 CFR Part 60, Subparts D
or Da. This correspondence also is based on the assumptions that
sampling is conducted for one hour at 500 ml/min and that the
effluent sample is 12% (v/v) C02.
The sulfate concentration of the performance check sample is 25 yg
SO. /ml, which corresponds to an S02 level of roughly 1000 ppm (for a
1-nour sample acquired at 500 ml/min and containing 12% (v/v) C02).
This concentration level should be more than adequate for situations
involving the application of Method 7D to sources subject to 40 CFR
Part 60, Subpart D; thus, it is recommended that analysts decrease
the SO." concentration in proportion to the S02 levels expected for
the effluent. For example, if the effluent concentration of S02 were
500 ppm, 5 ml (rather than 10 ml) of the SO." solution would be used
in preparing the performance check sample. For applications upstream
of flue gas desulfurization systems at sources subject to 40 CFR Part
60, Subpart Da, the opposite situation may exist, and it is
recommended that the concentration of SO." be increased accordingly.
The_N02~ concentration of the performance check solution is 0.07
yg N02~/ml. This corresponds to 6 ppm N02 for a one-hour sample
acquired at 500 ml/min and containing 12% (v/v; C02.
The performance check solution should be analyzed with the
calibration standards during the initial check of the ion
chromatograph's calibrations. The same experimental conditions
should be observed for the solution and the standards. Figure 2.7 is
an example chromatogram showing where the N02 , N03 , and SO." can be
expected to elute.
2.6.2 Preparation of Calibration Curve - Method 7D gives general
instructions for preparing the calibration curve for the ion
chromatograph. Accordingly, the method requires that:
(a) at least four calibration standards be prepared;
-------
Section No. 3.15.2
Date July 1, 1986
Page 17
5.5 minutes
N03 3.7 minutes
NO_ 1.4 minutes
Inject
Figure 2.7. Chromatogram showing resolution of nitrite,
nitrate, and sulfate peaks.
(b) the concentration range of the calibration standards cover the
concentration range of the samples being analyzed;
(c) the calibration standards be prepared from the KN03 standard
solution using pipettes having volumes 1.0 ml or greater;
(d) the calibration standards be analyzed and the results be
interpreted in the same manner as for the samples being
analyzed;
(e) the results of the analyses of the calibration standards (in
units of either peak height or peak area) should be plotted
versus the standards' concentrations (in units of yg N03~/ml);
(f) the plotted points define a linear relation;
(g) the calibration equation be determined from the points using
linear regression; and
-------
Section No. 3.15.2
Date July 1, 1986
Page 18
(g) the calibration equation be determined from the points using
linear regression; and
(h) the calibration standards be analyzed twice in order to
compensate for any drift in the response of the ion
chromatograph.
The method leaves to the analyst details including:
(a) the concentration values for the individual calibration
standards;
(b) the degree of linearity of the calibration curve that will
ensure quality results; and
(c) the procedure to be used to compensate results for the ion
chromatograph's drift.
Concentration values for calibration standards - The step-by
step-procedures for preparing the calibration standards and preparing
the calibration curve are given in Section 3.15.5.
-------
Section No. 3-15.2
Date July 1. 1986
Page 19
Table 2.1. ACTIVITIY MATRIX FOR CALIBRATION OF EQUIPMENT
Apparatus
Acceptance Limits
Frequency and method
of measurement
Action if
requirements
are not met
Wet test meter
Capacity of at least 2
L/min and an accuracy
within l.Q%
Calibrate initially and
then yearly by liquid
displacement
Adjust until
specifications
are met, or
return to
manufacturer
Dry gas meter
Y = 'Y+0.02Y at a
flow rate of about
450 cc/min
Calibrate vs. wet test
meter initially and
when the posttest check
is not within Y+0.05
Repair and
then recali-
brate, or
replace
Dry gas meter
thermometer
Within 3°C (5-t F)
of true value
Calibrate each initially
as a separate component
against a mercury-in-
glass thermometer; after
train is assembled
before each field test,
compare with mercury-in-
glass thermometer
Adjust, deter-
mine a con-
stant correc-
tion factor
or reject
Rotameter
Clean and maintain ac-
cording to manufactur-
er's instructions (re-
quired) ; calibrate to
+5# (recommended)
Initially and after each
field trip
Adjust and re-
calibrate, or
reject
Barometer
Within 2.5 mm
(0.1 in.) Hg of
mercury-in-glass bar-
ometer or of weather
station value
Calibrate initially
using a mercury-in-glass
barometer; check before
and after each field
test
Adjust to
agree with
certified
barometer
Analytical
balance
Weight within 2 mg of
standard weights
(Class-S)
Use standard weight be-
fore preparation of
working solution
Repair or
return to
manufacturer
(continued)
-------
Section No. 3.15.2
Date July 1, 1986
Page 20
Table 2.1. (continued)
Apparatus
Acceptance Limits
Frequency and method
of measurement
Action if
requirements
are not met
Ion chromato-
graph
Calibrate prior to each
set of sample analyses
With each set of field
samples; calibration
standards prepared from
potassium nitrate
Interpret data
using another
technique;
e.g., if using
peak height,
change to peak
area; analyze
additional
calibration
standards;
calibrate
conductivity
detector;
consult opera-
tor's manual
-------
Section No. 3.15.3
Date July 1, 1986
Page 1
3.0 PRESAMPLING OPERATIONS
The quality assurance activities for presampling preparation are
summarized in Table 3.1 at the end of this section. See Section 3.0
of this Handbook for details on preliminary site visits.
3.1 Apparatus Check and Calibration
Figure 3.1 or a similar form is recommended to aid the tester in
preparing an equipment checklist, status report form, and packing
list.
3.1.1 Sampling Train - The schematic for the NO sampling train is
given in Figure 1.2. Commercial models of tftis system are avail-
able. Each individual or fabricated train must be in compliance with
the specifications in Method 7D, Section 3.15.10.
3.1.2 Probe - The probe should be cleaned internally by brushing
first with tap water, then with water, and finally with acetone.
Allow the probe to dry in the air. In extreme cases, the glass or
stainless steel liner can be cleaned with stronger reagents; the
objective is to leave the liner free from contaminants. The probe's
heating system should be checked to see whether it is operating
properly. The probe must be leak free when sealed at the inlet or
tip and checked for leaks at a vacuum of 250 mm (10 in.) Hg with the
meter box. Any leaks should be corrected. The liner should be
sealed inside the metal sheath to prevent diluent air from entering
the source since most stacks are under negative pressure.
3.1.3 Restricted Orifice Impingers and Glass Connectors - All
glassware should be cleaned with detergent and tap water, and then
with water. Any items that do not pass a visual inspection for
cracks or breakage must be repaired or discarded.
3.1.4 Drying Tubes - Drying tubes should be packed with 6- to
16-mesh silica gel and sealed at both ends.
3.1.5 Valve and Rotameter - Prior to each field trip or at any sign
of erratic behavior, the flow control valve and rotameter should be
cleaned according to the maintenance procedure recommended by the
manufacturer.
3.1.6 Pump - The vacuum pump and oiler should be serviced as
recommended by the manufacturer, every 3 months, or every 10th test
(whichever comes first), or upon erratic behavior (nonuniform or
insufficient pumping action).
3.1.7 Dry Gas Meter - A dry gas meter calibration check should be
made in accordance with the procedure in Section 3.15.2. An
acceptable posttest check from the previous test is sufficient.
-------
Section No. 3.15.3
Date July 1, 1986
Page 2
Apparatus check
Probe
Type liner
Glass *X
Stainless
steel
Other
Heated properly*
Leak checked
Filter
Glass wool
Other
Glassware
Restriced
orifice
impinger
Size
Type
Meter System
Leak- free pumps*
Rate meter*
Dry gas meter*
C00 Measurement
OrSat t/"
Fyrite
Reagents
Water
Potassium
permanganate*
Silica gel
Other
Barometer
Drying tube
Acceptable
Yes
iX
tX"
ix^
^
i/
iX
X'
^
i/
i/
«X"
X
^
No
Quantity
required
+
^^i^ll tooyt.
14-
•2.
/
1
2^
2 j*l
2- J*l
Z&
/
6>
Ready
Yes
l^
^
^
(^
iX
\^
i^
^
i/'
^
(^
t^
NO
Loaded
and packed
Yes
^
^
^
iX
i^
^
^
^
^
^
No
* Most significant items/parameters to be checked.
Figure 3.1. Pretest preparations.
-------
Section No. 3.15.3
Date July 1, 1986
Page 3
3.1.8 Thermometers - The thermometers should be compared with the
mercury-in-glass thermometer at room temperature prior to each field
trip.
3.1.9 Barometer - The field barometer should be compared with the
mercury-in-glass barometer or with a National Weather Service Station
reading prior to each field trip.
3.1.10 COg Analysis - Method 3 sampling apparatus should be leak
checked, ana the reagents should be checked to ensure freshness (see
Section 3.2 of this Handbook).
3.2 Reagents for Sampling
The following reagents are needed during the sampling phase of
Method 7D:
3.2.1 Water - Deionized distilled water should conform to ASTM
specification D1193-82, Type III (see Subsection 1.4.1 for detailed
specifications).
3.2.2 Potassium Permanganate/Sodium Hydroxide (KMnO./NaOH) Solution -
Dissolve 40.0 g of KMN04 and 20.0 g of NaOH in 940 ffll of water.
3.3 Packaging Equipment for Shipment
Equipment should be packed in rigid containers to protect it
against rough handling during shipping and field operations.
3.3.1 Probe - The inlet and outlet of the probe must be sealed and
protected from breakage. A suggested container is a wooden case
lined with polyethylene foam or other suitable packing material; the
case should have separate compartments for individual devices. The
case should be equipped with handles or eye hooks that can withstand
hoisting, and should be rigid to prevent bending or twisting during
shipping and handling.
3.3.2 Impingers, Connectors, and Assorted Glassware - All impingers
and glassware should be packed in a rigid container and protected by
polyethylene foam or other suitable packing material. Individual
compartments for glassware help to organize and protect each item.
The impinger train may be charged and assembled in the laboratory if
sampling is to be performed within 24 hours.
3.3.3 Drying Tubes and Volumetric Glassware - A rigid container lined
with polyethylene foam material protects drying tubes and assorted
volumetric glassware.
3.3.4 Meter Box - The meter box, which contains the valve, rotameter,
vacuum pump, dry gas meter, and thermometer(s), should be packed in a
rigid shipping container unless its housing is strong enough to
-------
Section No. 3.15.3
Date July 1, 1986
Page 4
protect components during travel. Additional pump oil should be
packed if oil is required for operation. It is advisable to ship a
spare meter box in case of equipment failure.
3.3.5 Wash Bottles and Storage Containers - Storage containers and
miscellaneous glassware should be packed in a rigid foam-lined
container. Samples being transported in the containers should be
protected from extremely low ambient temperatures (below freezing).
-------
Section No. 3.15-3
Date July 1, 1986
Page 5
Table 3.1. ACTIVITY MATRIX FOR PRESAMPLING OPERATIONS
Operation
Acceptance limits
Frequency and method
of measurement
Action if
requirements
are not met
Probe
1. Probe liner free
of contaminants
2. Probe leak free at
250 inm (10 in.) Hg
3. No moisture conden-
sation
1. Clean probe inter-
nally by brushing with
tap water, then deion-
ized distilled water,
then acetone; allow to
dry in air before test
2. Visually check for
cracks before test
3. Check out heating
system initially and
when moisture appears
during testing
1. Retrace
cleaning pro-
cedure and
assembly
2. Replace
3. Repair
or replace
Restricted
orifice impin-
gers and glass
connectors
Clean and free of
breaks, cracks, etc.
Clean with detergent,
tap water, and then
with deionized dis-
tilled water
Repair or
discard
Flow control
valve and
rotameter
Clean and without sign
of erratic behavior
(ball not moving)
Clean prior to each
field trip or upon
erratic behavior
Repair or
return to
manufacturer
Vacuum pump
Maintain sampling rate
of 400 to 500 cc/min
at a vacuum up to
250 mm (10 in.) Hg
Service every 3 mo. or
upon erratic behavior;
check oiler jars every
10th test
As above
Dry gas meter
Clean and within 2%
of calibration factor
Calibrate according
to Section 3.15.2;
check for excess oil
if oiler is used
As above
CO- analyzer
Leak-free and fresh
reagents
Leak check, and check
reagents
As above
(continued)
-------
Section No. 3-15-3
Date July 1, 1986
Page 6
Table 3.1 (continued)
Operation
Acceptance limits
Frequency and method
of measurement
Action if
requirements
are not met
Reagents
Sampling
Requires all ACS grade
reagents
Prepare and store in
sealed containers
Prepare new
reagent
Sample recovery
Requires water on
site
Quantity sufficient to re-
cover sample after testing
and clean impingers prior
to testing
Prepare new
reagent
Package Equip-
ment for Ship-
ment
Probe
Protect with poly-
ethylene foam
Prior to each shipment
Repack
Impingers,
connectors,
and assorted
glassware
Pack in rigid con-
tainers with poly-
ethylene foam
As above
As above
Drying tubes,
volumetric
glassware
Sturdy container
lined with foam
As above
As above
Meter box
Meter box case and/or
container to protect
components, pack spare
meter box and oil
As above
As above
Wash bottles
and storage
containers
Pack in rigid foam-
lined container
As above
As above
Samples
Protect from extreme
cold (below freezing)
As above
As above
-------
Section No. 3.15.4
Date July 1, 1986
Page 1
4.0 ON-SITE MEASUREMENTS
On-site activities include transporting the equipment to the test
site, unpacking and assembling, sampling for nitrogen oxides, and
recording the data. The quality assurance activities are summarized
in Table 4.1 at the end of this section.
4.1 Transport of Equipment to the Sampling Site
The most efficient means of transporting the equipment from
ground level to the sampling site (often above ground level) should
be decided during the preliminary site visit or by prior corres-
pondence. Care should be taken to prevent damage to the equipment or
injury to test personnel during the moving. A laboratory area should
be designated for preparing the absorbing reagents, charging the
impingers, and sample recovery.
4.2 Preliminary Measurements and Setup
Method 7D outlines the procedure for determining the concentra-
tion of nitrogen oxides in the gas stream. The accuracy of the
equipment that has been transported to the sampling site and that may
have been handled roughly can be determined by making a one-point
check of the rotameter reading against the dry gas meter reading at
the test site. Use Equation 3 in Figure 2.4A or 2.4B, and substitute
dry gas meter readings in place of wet test meter readings (i.e., V,
= V ). Y . should be between 0.9 and 1.1; if not, the meter box hai
losr its rate or volume calibration. The tester can still use the
meter box, but the data should not be released for decision making
until a post-test recalibration has been made. If the dry gas meter
calibration factor did change, the dry gas meter volumes may have to
be corrected. Record the test identification number on the sampling
data form, Figure 4.1.
4.3 Sampling
The on-site sampling includes the following steps:
1. Preparation and/or addition of the absorbing reagents to the
impingers.
2. Setup of the sampling train.
3. Connection to the electrical service.
4. Preparation of the probe (leak check of entire sampling train
and addition of particulate filter).
5. Check of rotameter setting.
6. Insertion of the probe into the stack.
-------
Plant name
Location
Operator
ro
oyer'
/Y&.
City
Date
Section No. 3.15-4
Date July 1, 1986
Page 2
Sample no.
A?-/
Probe length/material
Meter box no. c^5 —/
Probe setting
Meter factor (Y)
Sampling point location(s)
Rotameter setting
nv*
Bar press mm (in.) Hg
Rotameter check?
2-tf. 4-1
Initial leak check?
CO- concentration
(1) Af. g (2)
Final leak check?
IbWfa
l(t7.ZZ(,
Total
non
Sample flow
rate setting,
cc/min (ft^/min)
4Zb
4$o
4Q
4^0
4^0
4$0
4- so
t&>
4ZO
4-^0
4^o
4*^0
Sample volume
metered,,j(V )
L (ft3) m
— •
2. 14J*
2-.14V
2.Z&-
2-ZW
2-ZZ1-
2,2^
2-Z^-~
Z.Z^°I
Z.Ztt
Z^W>
2-^4
Z.Z.44-
V
m
avgZ.Z^7
Percent
deviation, a
%
-O.Z2.
-0.41
+O.Z7
-0.31
-H?-04-
-0.^1
+0.13
-+0.&0
+o.?>\
-0,13
to.l3>
-O.3/
Avg
dev "0. 42
Dry gas
meter temp,
§C <°F)
—
72
74
74
75-
7b
76
76
7g
70
7?
76
7?
Avg
76-4-
Percent deviation = m " m avg x 100 (must be less than 10 percent).
V avg
Figure 4.1. Field sampling data form for NO .
-------
Section No. 3.15.4
Date July 1, 1986
Page 3
7. Sealing of the port.
8. Check of the temperature of the probe.
9. Sampling.
10. Measuring the C02 concentration.
11. Recording of the data in Figure 4.1.
A final leak check of the train is always performed after sampling.
4.3.1 Preparation and/or Addition of Absorbing Reagents to
Collection System - Absorbing reagents can be prepared on site, if
necessary, according to the directions in Subsection 1.4.1.
1. Use a pipette or a graduated cylinder to introduce 200 ml of
alkaline permanganate (KMnO./NaOH) solution into each of the three
impingers.
2. Place in the sampling train a drying tube that has new or
regenerated silica gel.
4.3.2 Assembling the Sampling Train - After assembling the sampling
train as shown in Figure 1.2, perform the following:
1. Adjust probe heater to operating temperature. Place crushed
ice and water around the impingers.
2. Leak check the sampling train just prior to use at the
sampling site (not mandatory) by temporarily attaching a rotameter
(capacity of 0 to 40 cc/min) to the outlet of the dry gas meter and
placing a vacuum gauge at or near the probe inlet. Plug the probe
inlet, pull a vacuum of at least 250 mm (10 in.) Hg, and note the
flow rate indicated by the rotameter. A leakage rate £ 2% of the
average sampling rate is acceptable. Note; Carefully release the
probe inlet plug before turning off the pump. It is suggested (but
not mandatory) that the pump be leak checked separately, either prior
to or after the sampling run. If prior to the run, the pump leak
check shall precede the leak check of the sampling train. If after,
the pump leak check shall follow the train leak check. To leak check
the pump, proceed as follows. Disconnect the drying tube from the
probe impinger assembly. Pull a vacuum of 250 mm (10 in.) Hg. Plug
or pinch off the outlet of the flow meter, and then turn off the
pump. The vacuum should remain stable for at least 30 seconds.
3. Place a loosely packed filter of glass wool in the end of the
probe, and connect the probe to the first impinger.
4.3.3 Rotameter Setting Check (Optional) - After leak checking the
sampling train, disconnect the probe from the first impinger, and
connect the filter (optional). The filter is a tube containing
-------
Section No. 3.15.4
Date July 1, 1986
Page 4
approximately 20 g of 5-Angstrom molecular sieve to remove the NO
from the ambient air. Start the pump, and adjust the flow to the
rotameter setting to be used during the sampling run. After the flow
has stablized, start measuring the volume sampled, as recorded by the
dry gas meter and the sampling time. Collect sufficient volume to
measure accurately the flow rate, and calculate the flow rate. The
average flow rate must be less than 500 cc/min for the sample to be
valid; therefore, it is recommended that the flow rate be checked as
above prior to each run. Record the sampling rate on the data form.
4.3.4 Sampling (Constant Rate) - Sampling is performed at a constant
rate of between 400 and 500 cc/min as indicated by the rotameter
during the entire sampling run. The procedure is as follows:
1. Record the initial dry gas meter readings, barometer reading,
and other data as indicated in Figure 4.1. Double check the dry gas
meter reading.
2. Position the tip of the probe at the sampling point, connect
the probe to the first impinger, and start the pump. Warning; If
the stack is under a negative pressure of >250 mm (10 in.) H20 while
disconnected from the impinger, the probe should be positioned at the
sampling point, the sample pump turned on, and then the probe
immediately connected to the impinger to prevent the impinger
solutions from being siphoned backwards.
3. Adjust the sample flow to the preselected flow rate (400 to
500 cc/min) as indicated by the rotameter.
4. Maintain a constant rate within 10% during the entire
sampling run, and take readings (dry gas meter, temperatures at dry
gas meter, and rate meter) at least every 5 minutes.
5. Refer to emission standards for minimum sampling time and/or
volume. (For example, the Federal standard for fossil fuel-fired
steam generators specifies a minimum sampling time of 60 minutes; for
relative accuracy tests, when the S02 concentration is greater than
1200 ppm, the sampling time should be 30 minutes since S0>2 depletes
the absorbing solution.) A quick calculation should be made after
half the sampling time to guarantee that the sampling rate will not
exceed 500 cc/min.
6. During sampling, measure the C02 content of the stack gas
near the sampling point using Method 3. The single-point grab
sampling procedure is adequate, provided the measurements are made at
least three times (near the start, midway, and before the end of a
run) and provided the average C02 concentration is computed. An
Orsat (which is highly recommended) or Fyrite analyzer may be used
for this analysis. The results should be recorded on the data form
(Figure 4.1).
-------
Section No. 3.15.4
Date July 1, 1986
Page 5
7. Turn off the pump at the conclusion of each run, remove the
probe from the stack, and record the final readings. Warning; Again,
if the stack is under negative pressure, disconnect the probe first,
and turn off the pump immediately thereafter.
8. Conduct a leak
( mandatory ) .
check, as described in Subsection 4.3.2
9. Calculate the sampling rate. The sample volume ( V ) for
each point should be within 10% of the average sampling volume for
all points, and the average sampling rate for the test should be less
than 500 cc/min. If the average sampling rate exceeds 500 cc/min,
the sample collection efficiency may be affected.
4.4 Sample Recovery
Method 7D requires transfer of the impinger contents and the
connector washings to a polyethylene storage container. This trans-
fer should be done in the "laboratory" area to prevent contamination
of the test sample.
After completing the final leak check, disconnect the impingers,
and transport them to the cleanup area. Cap off the impinger section
with the use of polyethylene or equivalent caps before transport to
the cleanup area. Transfer the contents of the impingers into a
labeled, leak-free polyethylene sample bottle. Rinse the three
impingers a couple of times and the connecting tubes once with 3- to
15-ml portions of water. Add these washings to the same sample
bottle, and mark the fluid level on the side. Place about 100 ml of
the absorbing reagent (KMn04/NaOH) in a polyethylene bottle, and
label it for use as a blank during samp?.e analysis (once for each
test). An example of a sample label is shown in Figure 4.2.
Plant City
Site Sample type
Date Run number
Front rinse [J Front filter LJ Front solution LJ
Back rinse LJ Back filter LJ Back solution 1 ]
• •
Solution Level marked J2
j_i
Volume: Initial Final 2
Cleanup by £
Figure 4.2. Example of a sample label.
-------
Section No. 3.15.4
Date July 1, 1986
Page 6
4.5 Sample Logistics (Data) and Packing Equipment
The sampling and sample recovery procedures are followed until
the required number of runs are completed. Log all data on the
Sample Recovery and Integrity Data Form, Figure 4.3. If the
impingers, and connectors are to be used in the next test, they
should be rinsed with water, and a new drying tube should be inserted
into the sampling train. At the completion of the test:
~1. Check all sample containers for proper labeling (time, date,
location, number of test, liquid level, and any other pertinent
documentation). Be sure that a blank has been taken.
2. Record all data collected during the field test in duplicate
by using carbon paper or by using data forms and a field laboratory
notebook. One set of data should be mailed to the base laboratory,
given to another team member, or given to the Agency. Hand carrying
the other set (not mandatory) can prevent a very costly and
embarrassing mistake.
3. Examine all sample containers and sampling equipment for
damage, and pack them for shipment to the base laboratory, being
careful to label all shipping containers to prevent loss of samples
or equipment.
4. Make a quick check of the sampling and sample recovery
procedures using the data form, Figure 4.4.
-------
Section No. 3.15.4
Date July 1, 1986
Page 7
Plant
ro der /A**/- Sampling location
A/0 . 3
Field Data Checks
Sample recovery personnel
Person with direct responsibility for recovered samples
Sample
number
1
2
3
Blank
Sample
identification
number
M-l
Date
of
recovery
3/75/04
Liquid
level
marked
i^
Stored
in locked
container
^^
Remarks
fw
Signature of field sample trustee
Laboratory Data Checks
Lab person with direct responsibility for recovered samples
Date recovered samples received
Analyst ^coft
Sample
number
1
2
3
Blank
Sample
identification
number
AP-I
Date
of
analysis
3//B/84
/ i
Liquid
level
marked
^-
Stored
in locked
container
t_^
Remarks
n
Signature of lab sample trustee
Figure 4.3. Sample recovery and
-------
Section No. 3.15.4
Date July 1, 1986
Page 8
Sampling
Impinger contents properly selected, measured, and placed in
impingers?* ^
Impinger Contents/Parameters*
1st: 200 ml of KMn04/NaOH
2nd: 200 ml of KMn04/NaOH
3rd: 200 ml of KMn04/NaOH r
Drying tube: 6- to 16-mesh silica gel ^_
Probe heat at proper level?* _
Pretest leak check at 250 mm (10 in.) Hg?
Leakage rate? 0- OO4-
Check of rotameter setting?
Probe placed at proper sampling point?
Flow rate constant at approximately 450 cc/min?*
CO,, concentration measured?* £-
Sample Recovery
Contents of impingers placed in polyethylene bottles?
Fluid level marked?* """^
Sample containers sealed and identified?*
* Most significant items/parameters to be checked.
Figure 4.4. On-site measurements.
-------
Section No. 3.15-4
Date July 1, 1986
Page 9
Table 4.1. ACTIVITY MATRIX FOR ON-SITE MEASUREMENT CHECKS
Activity
Acceptance limits
Frequency and method
of measurement
Action if
requirements
are not met
Preparing and/
or adding
absorbing
reagents
Add 200 ml of
NaOH to the impingers
Add 40.0 g of KMnO. and
20.0 g of NaOH to §40 ml
of water
Reassemble
collection
system
Assembling
the sampling
train
1. Assemble to speci-
fications in Fig. 1.2
2. A leakage rate
of <2% of the average
sampling rate
1. Before each sampling
run
2. Leak check before
sampling (recommended) by
attaching a rotameter to
dry gas meter outlet,
placing a vacuum gauge
at or near probe inlet,
and pulling a vacuum
of > 250 mm (10 in.) Hg
1. Reassemble
2. Correct
the leak
Sampling (con-
stant rate)
1. Within 10% of
constant rate
2. Minimum accepta-
ble time is 60 min
and sampling rate
less than 500 cc/min
3. Less than 2% leak-
age rate at 250 mm
(10 in.) Hg
4. Determine CO,
content
1. Calculate % devia-
tion for each sample
using equation in
Fig. 4.1
2. Make a quick calcu-
lation prior to comple-
tion and an exact calcu-
lation after completion
3. Leak check after
sampling run (mandatory);
use same procedure
as above
4. Measure C0_ content
using Method 3
1. Repeat
the sampling,
or obtain ac-
ceptance from
a representa-
tive of the
Administrator
2. As above
As above
As above
(continued)
-------
Section No. 3.15.4
Date July 1, 1986
Page 10
Table 4.1 (continued)
Activity
Acceptance limits
Frequency and method
of measurement
Action if
requirements
are not met
Sample logistics
(data) and
packing of
equipment
1. All data are re-
corded correctly
2. All equipment ex-
amined for damage and
labeled for shipment
3. All sample con-
tainers properly
labeled and packaged
1. Visually check upon
completion of each run
and before packing
2. As above
3. Visually check upon
completion of test
1. Complete
the data form
2. Redo test
if damage
occurred
during testing
3. Correct
when possible
-------
Section No. 3.15.5
Date July 1, 1986
Page 1
5.0 POSTSAMPLING OPERATIONS
The postsampling operations for Method 7D include an apparatus
check, a barometer check, sample preparation, and sample analysis by
ion chromatogr aphy . The procedures for the apparatus check and the
barometer check are the same as in Method 6. These procedures are
detailed in Method 6 in Section 3.5.5 and are not discussed here.
The procedures for sample preparation and sample analysis are
described here. Table 5.1 provides a checklist summarizing the
postsampling procedures.
5. 1 Sample Preparation
Sample preparation should not be started until the required 36-
hour conversion time has elapsed for complete conversion of N02~ to
N03~. When using Method 7D for relative accuracy testing of
continuous emission monitors, the sample can be prepared immediately
if the nitrite in the sample is quantitated using the procedures
described in Subsection 5.2. The liquid level in the sample
container should be checked to determine if sample has been lost
during shipment. If a loss has occured, the appropriate steps should
be taken to correct for the loss ( see Subsection 5.1.1). The sample
is prepared for ion chromatography by precipitating the excess
permanganate as manganese dioxide (Mn02). A 5% (v/v) hydrogen
peroxide (H202) solution is used to reduce the permanganate to MnO,,.
The Mn02 precipitate is removed by vacuum filtration and the filtered
solution is volumetrically diluted prior to chromatographic analysis.
5.1.1 Sample Loss Determination and Correction - Before preparing the
sample, it must be allowed the full 36 -hour conversion time. Compare
the liquid level in the sample container to the mark on the
container. If a noticeable amount of sample has been lost, use the
following procedure for correcting the sample volume:
1. Mark the new or final level of liquid on the sample container.
2. Transfer the sample to a 1-liter volumetric flask (V ), and
rinse the container with water. soin
3. Fill the sample container with water to the initial sample
level marked on the container after sampling. Transfer the water
to a graduated cylinder, and determine the original sample
volume (
4. Fill the sample container with water to the final sample level.
Transfer the water to a graduated cylinder, and determine the final
sample volume (vsoln )•
*f
JS
^\^f J- Alp ^\JA,lli
by using Equation 5-1:
5. If V . is less than Vgoln , correct the sample volume (V )
-------
Section No. 3.15.5
Date July 1, 1986
Page 2
Vsoln ' - Vsoln Vsolni Equation 5-1
v
where: soln,.
i i
V = sample volume to be used for calculations, ml,
SO xll
V , = volumetric flask volume, ml,
soln
V = initial sample volume placed in sample container, ml, and
so J. n.
V n = final sample volume removed from container, ml.
soln.p
6. Both the corrected and uncorrected values should be submitted
in the test report to the Agency.
5.1.2 Permanganate Precipitation and Filtration - After the required
36-hour conversion period for the sample has elapsed and the sample
container has been checked for sample loss, the sample can be
quantitatively transfered to a 1-liter volumetric flask. (If the
correction for sample loss has already been made, the sample should
already be in a 1-liter volumetric flask.) The procedure for
precipitating the excess permanganate is as follows:
1. Dilute the sample in a 1-liter volumetric flask (V ) to
volume with reagent water, and mix well.
2. Take a 50-ml aliquot (V ) of the sample from the 1-liter
volumetric flask, and transfer3 the aliquot to a 250-ml Erlenmeyer
flask containing a Teflon-coated stirring bar. If the NO
concentration is low, a 100-ml aliquot may be taken to increase th§
instrument response.
3. Stir the sample as fast as possible without splashing any of
the sample out of the flask.
4. Add a 5% H202 solution in 5-ml portions while stirring until
the permanganate color disappears.
5. Stop stirring and allow the precipitated manganese dioxide to
settle. If the solution is clear, then enough H2<32 has been added.
If the permanganate color persists in the solution, then continue the
H202 addition in 5-ml portions until a clear solution is produced
after settling.
6. Assemble the Buchner funnel and filter flask. The spout of
the Buchner funnel may be fitted with a length of Teflon tubing to
minimize the probability of sample loss by aspiration during
filtration.
-------
Section No. 3.15.5
Date July 1, 1986
Page 3
7. Place a piece of GF/C filter paper (or an equivalent type of
filter paper) in the Buchner funnel. Wet the paper with water, and
seal the filter by applying a vacuum to the flask.
8. Quantitatively transfer the precipitated sample solution to
the filter, and filter the solution. Wash the Erlenmeyer flask and
the solid material on the filter with water four times, and collect
the washings with the filtered solution.
9. Quantitatively transfer the filtered solution from the filter
flask to a 250-ml volumetric flask (vvJ* Dilute to volume with
water.
10. Prepare a reagent blank by repeating steps 2 through 9 on a
diluted sample of the alkaline-permanganate absorbing solution.
Dilute 60 ml of the absorbing solution to 100 ml with water, and use
50 ml in step 2.
5.2 Sample Analysis by Ion Chromatography
For Method 7D, the basic components and the operation of the ion
chromatograph are the same as for Method 7A. Discussions of the ion
chromatograph can be found in both Section 3.15.2 and Section
3.14.2. The analyst should be familiar with the operator's manual
for his particular ion chromatograph system. In this section, the
preparation of calibration standards, the use of quality assurance
audit samples, the analysis procedure, and the data reduction and
reporting are described.
5.2.1 Preparation of Calibration Standards - The accuracy of the ion
chromatographic analysis, as in any analysis, depends directly on the
accuracy of the prepared calibration standards. The use of proper
pipetting procedures, described in Method 7A, Section 3.14.2, and a
properly dried, reagent grade standard are necessary to obtain
quality results from the analysis. The preparation of the N0^~
calibration standards is as follows: 3
1. Dry approximately 15 g of potassium nitrate (KNO~) in an oven
at 105 to 110 C for 2 hours. (Sodium nitrate can also be used
provided the difference in the formula weight is considered in the
subsequent calculations.) Allow the dried KNO- to cool to room
temperature in a desiccator before weighing.
2. Calibrate an analytical balance with a 5-g Class-S calibra-
tion weight and a 10-g Class-S calibration weight to within 2 mg.
Accurately weigh 9 to 10 g of dried KN03 to within 0.1 mg.
3. Dissolve in reagent water, and dilute to_ 1 liter in a
Class-A volumetric flask. Calculate the exact N03~ concentration
using the following formula:
-------
Section No. 3.15.5
Date July 1, 1986
Page 4
yg NO ~/ml = g of KNCU x 103 x 62'01 Equation 5-2
d 3 101.1
The stock standard solution should be stable for 2 months if precau-
tions, such as refrigeration, are used to prevent decomposition by
nitrate-utilizing microorganisms.
4. Prepare a fresh working standard solution for each set of
analyses by pipetting 5 ml of stock standard solution into a 200-ml
Class-A volumetric flask. Dilute to volume with water.
5. Prepare a series of four calibration standards from the fresh
working standard solution. Pipet 1.0 ml, 3.0 ml, 5.0 ml, and 10.0 ml
into a series of four 100-ml Class-A volumetric flasks. Dilute to
volume with reagent water. The concentration of the_calibration
standards made from a 9.7823 g KN03/liter (6000 yg N03~/ml) stock
standard solution would be 1.5, 4.5, 7.5, and 15.0 yg Nof~/ml.
o
The calibration standard concentrations cited above are used in
the example used in Figure 5.1, the analytical data form for analysis
of calibration standards.
The calibration standards for nitrite quantitation are prepared
when Method 7D is used for relative accuracy testing of continuous
emission monitors. A stock N02~ standard solution is (1) prepared
with NaNO2 of known purity or (2) analyzed before use. Do not oven
dry the NaN02. Dissolve 52.5 mg of NaN02 in water and dilute to
volume in a 250-ml Class-A volumetric flasR. A series of four
calibration_standards with N02 concentrations of 1.4, 4.2, 7.0, and
14.0 yg N02~/ml are prepared By pipetting 10, 3.0, 5.0, and 10 ml of
stock N02 standard into four 100-ml Class-A volumetric flasks. The
N02~ calibration standards are diluted to volume with water.
5.2.2 Quality Assurance Audit Samples - The quality of analytical
results can be assessed by analyzing nitrate standard solutions
prepared by an independent laboratory. For such standard solutions,
or quality assurance audit samples, the concentrations are known to
the control agency (the auditor) but are unknown to the analyst.
Subsection 3.2.6 of the Federal Register promulgation of Method
7D (see Section 3.15.10) requires the analysis of quality assurance
audit samples as described in Method 7. This means that when Method
7D is used to demonstrate compliance with an EPA pollutant emission
standard (specified in 40 CFR Part 60), a performance audit must be
conducted on the analytical phase of the method. Nitrate samples in
glass vials must be obtained for this performance audit from the
Quality Assurance Management Office at each EPA Regional Office or
from the responsible enforcement agency. The addresses of the EPA
Regional Quality Assurance Coordinators are shown in Table 5.1 of
Section 3.0.5 of this Handbook.
-------
Plant
Date
Location
Analys t
Section No. 3.15.5
Date July 1, 1986
Page 5
A/o. 3>
ate
Standard
identifier
Std 1
Std 2
Std 3
Std 1*
Standard
concentration (x)
(yg/ml N03 )
/.r
4
-------
Section No. 3.15.5
Date July 1, 1986
Page 6
The concentration of each audit sample measured by the analyst
must agree within 10 percent (relative error) of the actual audit
concentration. The relative error is calculated using the following
equation:
where
RE
= Cd -
Equation 5-3
'a x 100
C, = Determined audit sample concentration, mg/dscm, and
C = Actual audit sample concentration, mg/dscm.
Cl
5.2.3. Ion Chromatographic Analysis of Calibration Standards, Reagent
Blank, Field Samples, and Quality Assurance Samples - The selection
of the ion Chromatographic conditions depends on the particular ion
chromatograph system available to the analyst. The selection of
eluents for ion chromatography depends on the method of detection
used. For suppressed ion chromatography the following conditions
have been used successfully:
1. A 0.003M NaHCO«/0.0024M ^
dissolving 1.008 g of NaHCO,
diluting to 4 liters. *
eluent solution is prepared by
1.018
of Na2C03 in water and
2. The full-scale detection range is set at 3
conductance), and a 0.5-ml sample loop is used.
MHO (units of
3. A flow rate of 2.5 ml/min gives a NO
approximately 15 minutes depending on the type
retention time of
f column used.
Non-suppressed ion chromatography and ion-pairing_chromatography
may also be used provided baseline separation of N0~ and SO.
aration and detection of NO
are obtained (see Figure
sep-
5.2).
Packed-bed suppression columns are not recommended for quantifying
N0«~ when using Method 7D for relative accuracy testing.
The recommended procedure
as follows:
for the ion Chromatographic analysis is
1. Establish a stable baseline. Inject a sample of water, and
observe the chromatogram to see whether any NO3 elutes. Repeat the
water injection until N0« is not observed on tne chromatogram. If,
after 5 injections, a N03 peak is still seen, the water source
should be checked for contamination.
2. Inject samples as follows:
• calibration standards;
-------
Section No. 3.15.5
Date July 1, 1986
Page 7
Field Sample: AP-1
Flow Rate: 1.5 ml/min
Detector: 30 yS full scale
Injection: 50 y1
Chart Speed: 1 cm/min
N03 3.3 minutes
Inject
Figure 5.2.
Example of a properly documented chromatogram having
adequate baseline separation.
-------
Section No. 3.15.5
Date July 1, 1986
Page 8
• field samples, reagent blank, and quality assurance samples
(duplicate injections); and
• finally the calibration standards again.
The injection volumes for all the standards and samples should be the
same.
3. The chromatograms should be documented with the sample
identification, injection point, injection volume, nitrate retention
time, eluent flow rate, detector sensitivity setting, and recorder
chart speed (see Figure 5.2).
4. Manually measure the N03~ peak height or determine the N03~
peak area with an electronic integrator.
5.2.4 Data Reduction and Reporting - The details of the data reduc-
tion procedure are discussed in Section 3.15.6. The procedures for
calculating a response factor from the calibration standards by
linear regression and for calculating the % deviation of the known
concentration of each standard from the predicted value are as
follows:
1. Use the analytical data form (see Figure 5.1) for calculat-
ing the linear regression equation based on the calibration
standards.
2. Record the calculated concentrations for the four
calibration standards (x) on the data sheet._ Determine the average
value for the instrument response (y) for N03~ (peak height or area
under the peak) from the three determinations for each of the four
calibration standards.
3. Plot the average values for the instrument response for the
calibration standards against the corresponding calculated concen-
trations of the calibration standards. Draw a smooth curve through
the points without forcing the curve through zero. The curve should
be linear.
4. Determine the slope (m) and the intercept term (b or I) for
the linear calibration curve by linear regression. Many scientific
calculators are capable of performing linear regression.
5. It is recommended that calculations be performed to
determine the percent deviation of the known concentration value for
each calibration standard from the concentration predicted by the
calibration curve. To do this, first calculate the predicted
concentration for each calibration standard (P) using the following
equation:
Equation 5-4
isin ~ ^ Average Instrument Response (y) - Intercept (I)
wu3 ; = Calibration Curve Slope (m)
-------
Section No. 3.15.5
Date July 1, 1986
Page 9
Then calculate the percent deviation of each calibration standard (x)
from the predicted value as shown in Equation 5-5.
Equation 5-5
% Deviation = P (»* N03"/ml> " * <* N03"ml> x 100
x (yg N03~/ml)
If any standard deviates from the standard curve by more than +7%,
the problem should be investigated.
The concentration of the field samples, the reagent blank, and
the quality assurance samples are calculated by the same procedure
used to calculate the predicted values for the calibration stan-
dards. Use the data form shown in Figure 5.3 for the analysis of
field samples. The procedure is as follows:
1. Determine the instrument response factor for the sample and
calculate the sample concentration using Equation 5-4. Calculate the
average value for the two determinations made on each sample.
2. It is recommended that calculations be performed to determine
the percent deviation of the concentration measured for each individ-
ual sample from the average concentration value calculated for the
two determinations made on each sample. The deviation can be
calculated using the following equation:
Equation 5-6
% Deviation = SamPle Concentration - Average Concentration 10_
Average Concentration
The percent deviation for a sample should be within 5% of the
average value before the analysis is considered valid.
The data reduction_procedures described above for NO ~ analysis
can be used for NO2 analysis when using Method 7D tor relative
accuracy testing of continuous emission monitors.
The main parameters of the analytical procedures may be checked
during or after the analysis, using the posttest operations form
(Figure 5.4).
-------
Section No. 3-15-5
Date July 1, 1986
Page 10
Date samples received
Plant
Date samples analyzed
Run number(s)
/ 2, .
Location
MT"
Calibration curve slope (m) 15.
Analyst S .
Intercept term (I) — £>, 2,4-3 ^
Field
sample
number
AP-I
AP-I
AP-3
Field
Blank
Analysis
number
1st
2nd
1st
2nd
1st
2nd
1st
2nd
Instrument
response (y)
(mm or area counts)
7? *,*
/fttntr*
fa^ (W
(p~Z-mw
72~to**
/O wn\
Stn^~>
JT w w
Concentration of
analysis samgle
(yg/ml N03 )
S.8
^7
4.?
4-0>
5:3
^.z
0,31
O.W
Average
Concentration of
analysis samgle
(yg/ml N03 )
s = ^. ?r
s = 4^r
s = 5:er
B = 6,3
-------
Section No. 3.15.5
Date July 1, 1986
Page 11
Reagents
Potassium nitrate dried at 105° to 110°C for a minimum
of 2 hours before use?
Stock standard solution (potassium nitrate) less than 2 months
old?
Sample Preparation
Has liquid level noticeably changed?*
Original volume Corrected volume
Analysis
Standard calibration curve prepared?* ^
All calibration points within 7 percent of linear calibration
curve (optional)? Y€&
Reagent blanks made from absorbing solution? Yes.
Same injection volume for both standards and samples? ^
Duplicate sample values agree within 5 percent of their mean?
Audit sample analytical results within 10 percent of true value?
^££
All analytical data recorded on checklist and laboratory form?
* Most significant items/parameters to be checked.
Figure 5.4. Posttest operations.
-------
Section No. 3-15.5
Date July 1, 1986
Page 12
Table 5.1. ACTIVITY MATRIX FOR SAMPLE ANALYSIS
Characteristics
Acceptance Limits
Frequency and method
of measurement
Action if
requirements
are not met
Sample Preparation
1. Conversion
time
2. Sample loss
3. Permanganate
precipitation
Permanganate
filtration
36 hour minimum
Determine sample age
Noticeable amount
Absence of purple
permanganate color
Absence of solids
in the filtrate
Compare sample level
to mark on container
Between each 5 ml
portion of 5% ^2Q2
solution
After filtration is
complete
Hold sample in
container for
36 hours min-
imum time
Correct by pro-
cedure in
Section 5.1.1
Continue adding
5 ml portions
of 5% H202
Refilter
Calibration Stan-
dards Preparation
1. ACS grade KNO_
15 g dry KNO-
2. Stock standard
solution
3. Calibration
standards
9 to 10 g of KNO-
accurately weighed
to 0.1 mg; dilute
to 1 liter; store
refrigerated
Standard range to
cover sample range;
maximum allowed
deviation of indi-
vidual standard
from the predicted
value is +1%
(optional!
Oven dry at 105 to
110 C for 2 hours;
cool in desiccator
Calibrate analytical
balance
Use recommended volumes
of stock standard solu-
tion; calculate devia-
tion (optional) using
Equation 5-5
High bias will
occur if stan-
dard contains
moisture; redry
KNO_
Biases will
occur with
poor pipetting
or improper
storage; re-
make standard
Invalid anal-
ysis; remake
and rerun
calibration
standards
(continued)
-------
Section No. 3.15-5
Date July 1. 1986
Page 13
Table 5.1. (continued)
Characteristics
Acceptance Limits
Frequency and method
of measurement
Action if
requirements
are not met
Ion Chromatograph
Analysis
1. Sample injection
device
2. Sample analysis
3. Chromatogram
documentation
Quality
assurance
Absence of KNO,
on chromatogram of
water injection
Individual sample
replicates within
5% of average
(optional)
Include sample
identification,
injection point,
injection volume,
NO, retention
time, eluent flow
rate, detector
sensitivity
setting, and
speed chart
Analytical results
must be within 10%
of actual value
Inject reagent water
up to four times
Calculate deviation
(optional) using
Equation 5~6
Visually check
Check water source
for contamination
Invalidate anal-
ysis; reanalyze
samples
Supply missing
"information
Report results to
agency with sample
identification
Invalidate anal-
ysis; repeat prep-
aration of sample;
prepare new stan-
dards
-------
Section No. 3.15.6
Date July 1, 1986
Page 1
6.0 CALCULATIONS
Calculation errors due to procedural or mathematical mistakes can
be a large component of total system error. Therefore, it is recom-
mended that each set of calculations be repeated or spotchecked,
preferably by a team member other than the one who performed the
original calculations. If a difference greater than typical round-
off error is detected, the calculations should be corrected. A
computer program is advantageous in reducing calculation errors. If
a standardized computer program is used, the original data entry
should be checked, and if differences are observed, a new computer
run should be made. Table 6.1 at the end of this section summarizes
the quality assurance activities for calculations.
Calculations should be carried at least one extra decimal figure
beyond that of the acquired data, and should be rounded after final
calculation to two significant digits for each run or sample. All
rounding of numbers should be performed in accordance with the ASTM
380-76 procedures. All calculations are then recorded on a form such
as the ones shown in Figure 6.1A and 6.IB, following the nomenclature
list.
6.1 Nomenc1ature
The following nomenclature is used in the calculations:
V = dry gas volume as measured by the dry gas meter,
dcm (dcf),
Y = dry gas meter calibration factor, dimensionless,
P. = barometric pressure, mm (in.) Hg,
P td = standard absolute pressure, 760 mm (29.92 in.) Hg,
T = average dry gas meter absolute temperature, °K (°R),
T = standard absolute temperature, 293°K (528°R),
S cQ
V , f.d\ = dry gas volume measured by the dry gas meter corrected
^ ' to standard conditions, dscm (dscf),
S = analysis of sample, yg N03 /ml,
B = analysis of blank, yg N03 /ml,
m = mass of NO as N0~ in sample, yg,
X ^
C = concentration of NO as N09, dry basis, mg/dscm
(Ib/dscf), and
X = CO, correction factor.
-------
Section No. 3.15.6
Date July 1, 1986
Page 2
6.2 Calculations
The following are the equations used with example calculation
forms (Figures 6.1A and 6.IB) to calculate the concentration of
nitrogen oxides in the samples.
6.2.1 Sample Volume - Calculate the sample volume on a dry basis at
standard conditions (760 mm (29.92 in.) Hg and 293 K (528 R) using
Equation 6-1.
T p v p Equation 6-1
V / 4.,-x - V XY std bar = K.XY m bar
m(std) m 1
where: m std m
X = correction factor for C00 collection, 100
100 - %C09(v/v),
KI = 0.3858 K for metric units, or
mmHg
K. = 17.64 °R for English units.
in. Hg
6.2.2 Total yg N0~ Per Sample - Calculate the total \i g of NO2 per
sample using Equation 6-2.
m = (S - B) 250 x x 46'01 = 3710 (S - B)
50 62.01
Equation 6-2
where:
250 = volume of prepared sample, ml,
46.01 = molecular weight of NO ~,
62.01 = molecular weight of NO ~,
1000 = total volume of KMnO. solution, ml, and
50 = aliquot KMnO^ / NaOH solution, ml.
6.2.3 Sample Concentration - Calculate the sample concentration on a
dry basis at standard conditions using Equation 6-3.
C = K, 5
Vm Equation 6-3
where:
-3
K2 = 10 mg/vig for metric units, or
K = 2.205 x 10~9 IbVyg f°r English units.
-------
Section No. 3.15.6
Date July 1, 1986
Page 3
Sample Volume
bar = _^^_'^_ /__ in. Hg, T = j?" "5 £> . $~~°R
v i +A\ ' 17-64 X Y Vm Pbar = _/_._/_ 0 £" dscf Equation 6-1
/ 4.^x
m(std)
Tm
Total y g NO2 Per Sample
S = S_._/ _^_ yg/mi, B = 0_*_3 7_ yg/ml
m = 3710 (S-B)= / f ^^ 6ygof NO0 Equation 6-2
"" ~~ """" " ~""™" ^— ^
Sample Concentration
C = 2.205 x 10"9 = _£/_£ &_ b_ x 10~5 Ib/dscf
m(std) Equation 6-3
Sample Concentration in ppm
ppm N02 = 8.375 x 10 C = 3_ 5_ 2.^ ppm N02 Equation 6-4
Figure 6.1A. Nitrogen oxide calculation form (English units).
-------
Section No. 3.15.6
Date July 1, 1986
Page 4
Sample Volume
v =0.0 2
m —
J_
m3, y =
, X
bar
±_ ?• nun Hg, T = ^ f ^.
~^^^~ ^^-M» III '^^^™ — L1- "l^ •" ^"^~ ^^^^
V / 4.^,x = 0.3858 X Y Vm Pbar = 0.0
m(sta) - ^ -
m
/ 3 / dscm Equation 6-1
Total ug N02 Per Sample
S =
B = J?-_2 f_ vg/ml
m = 3710 (S - B) =
N0
Equation 6-2
C = 10
4-3 m
'm(std)
Sample Concentration
= G 3 t> •_/_ mg N02/dscm
Sample Concentration in ppm
Equation 6-3
ppm N02 = 0.5228 C = _ 5_ ^_ 2- ppm
NO,
Equation 6-4
Figure 6.IB. Nitrogen oxide calculation form (metric units).
-------
Section No. 3.15.6
Date July 1, 1986
Page 5
Table 6.1. ACTIVITY MATRIX FOR CALCULATIONS
Characteristics
Sample volume
calculation
Sample mass
calculation
Sample concen-
tration
Calculation
check
Document and
report re-
sults
Acceptance limits
All data available;
calculations correct
within round-off error
As above
As above
Original and checked
calculations agree
within round-off error
All data available;
calculations correct
within round-off error
Frequency and method
of measurement
For each sample, exam-
ine the data form
As above
As above
For each sample, per-
form independent cal-
culations
For each sample, exam-
ine the data form
Action if
requirements
are not met
Complete the
data, or void
the sample
As above
As above
Check and
correct all
data
Complete the
data, or void
the sample
-------
Section No. 3.15.7
Date July 1, 1986
Page 1
7.0 MAINTENANCE
The normal use of emission-testing equipment subjects it to cor-
rosive gases, extremes in temperature, vibration, and shock. Keeping
the equipment in good operating order over an extended period of time
requires knowledge of the equipment and a program of routine-main-
tenance which is performed quarterly or after 2830 L (100 ft ) of
operation, whichever is greater. In addition to the quarterly main-
tenance, a yearly cleaning of the entire meter box is recommended.
Maintenance procedures for the various components are summarized in
Table 7.1 at the end of the section. The following procedures are
not required, but are recommended to increase the reliability of the
equipment.
7.1 Pump
In the present commercial sampling train, several types of pumps
are used; the two' most common are the fiber vane pump with in-line
oiler and the diaphragm pump. The fiber vane pump requires a
periodic check of the oiler Jar. Its contents should be translucent;
the oil should be changed if not translucent. Use the oil specified
by the manufacturer. If none is specified, use SAE-10 nondetergent
oil. Whenever the fiber vane pump starts to run erratically or
during the yearly disassembly, the head should be removed and the
fiber vanes changed. Erratic operation of the diaphragm pump is
normally due to either a bad diaphragm (causing leakage) or to
malfunctions of the valves, which should be cleaned annually by
complete disassembly.
7.2 Dry Gas Meter
The dry gas meter should be checked for excess oil or corrosion
of the components by removing the top plate every 3 months. The
meter should be disassembled and all components cleaned and checked
whenever the rotation of the dials is erratic, whenever the meter
will not calibrate properly over the required flow rate range, and
during the yearly maintenance.
7.3 Rotameter
The rotameter should be disassembled and cleaned according to the
manufacturer's instructions using only recommended cleaning fluids
every 3 months or upon erratic operation.
7.4 Sampling Train
All remaining sampling train components should be visually
checked every 3 months and completely disassembled and cleaned or
replaced yearly. Many items, such as quick disconnects, should be
replaced whenever damaged rather than checked periodically.
Normally, the best procedure for maintenance in the field is to use
another entire unit such as a meter box, sample box, or umbilical
-------
Section No. 3.15.7
Date July 1, 1986
Page 2
cord (the hose that connects the sample box and meter box) rather
than replacing individual components.
7.5 Ion chromatograph
Maintenance activities and schedules for ion chromatographs are
make and model specific. It is therefore recommended that the
analyst consult the operator's manual for instructions relative to
maintenance practices and procedures.
Guard columns, while not required, are recommended for use with
the ion chromatograph in order to extend column lifetime.
-------
Section No. 3.15-7
Date July 1. 1986
Page 3
Table 7.1. ACTIVITY MATRIX FOR EQUIPMENT MAINTENANCE CHECKS
Apparatus
Acceptance limits
Frequency and method
of measurement
Action if
requirements
are not met
Sample train
control con-
sole
No erratic behavior
Routine maintenance
performed quarterly;
disassemble and clean
yearly
Replace parts
as needed
Fiber vane pump
In-line oiler free of
leaks
Periodically check
oiler jar; remove
head and change fiber
vanes
Replace as
needed
Diaphragm pump
Leak-free valves
functioning properly
Clean valves during
yearly disassembly
Replace when
leaking or mal-
functioning
Dry gas meter
No excess oil, corro-
sion, or erratic rota-
tion of the dial
Check every 3 mo. for
excess oil or corrosion
by removing the top
plate; check valves and
diaphragm yearly and
whenever meter dial runs
erratically or whenever
meter will not calibrate
Replace parts
as needed, or
replace meter
Rotameter
Clean and no erratic
behavior
Clean every 3 mo. or
whenever ball does not
move freely
Replace
Sampling train
No damage
Visually check every
3 mo.; completely dis-
assemble and clean or
replace yearly
If failure
noted, use an-
other entire
meter box,
sample box,
or umbilical
cord
Ion chroma-
tograph
See owner's manual
See owner's manual
See owner's
manual
-------
Section No. 3.15.8
Date July 1, 1986
Page 1
8.0 AUDITING PROCEDURE
An audit is an independent assessment of data quality. Indepen-
dence is achieved if the individual(s) performing the audit and their
standards and equipment are different from the regular field team and
their standards and equipment. Routine quality assurance checks by a
field team are necessary to generate good quality data, but they are
not part of the auditing procedure. Table 8.1 at the end of this
section summarizes the quality assurance functions for auditing.
Based on the results of collaborative tests ' ' of Method 7,
two specific performance audits are recommended:
1. Audit of the analytical phase of Method 7D.
2. Audit of data processing.
It is suggested that a systems audit be conducted as specified by the
quality assurance coordinator in addition to these performance
audits. The two performance audits and the systems audit are
described in detail in Subsections 8.1 and 8.2, respectively.
8.1 Performance Audits
Performance audits are made to evaluate quantitatively the quality
of data produced by the total measurement system (sample collection,
sample analysis, and data processing). It is recommended that audits
be performed by the responsible control agency once during every
enforcement source test. A source test for enforcement comprises a
series of runs at one source. The performance audit of the
analytical phase is subdivided into two steps: (1) a pretest audit
which is optional, and (2) an audit during the field sampling and/or
analysis phase which is required. No audit is recommended at this
time for the sample collection phase.
8.1.1 Pretest Audit of Analytical Phase (Optional) - The pretest
audit described in this section can be used to determine the
proficiency of the analyst, the quality of the standardization of
solutions in the Method 7D analysis, and the ability to perform the
computations correctly. It should be performed at the discretion of
the agency auditor, the laboratory supervisor, source test company,
or quality assurance officer. The analytical phase of Method 7D can
be audited with the use of aqueous potassium or sodium nitrate
samples provided to the testing laboratory before the enforcement
source test. Aqueous potassium or sodium nitrate samples may be
prepared by the procedure described in Section 3.15.5 for calibration
standard preparation.
The pretest audit provides the opportunity for the testing labora-
tory to check the accuracy of its analytical procedure. This audit
is especially recommended for a laboratory with little or no experi-
ence with the Method 7D analysis procedure described in this
Handbook.
-------
Section No. 3.15.8
Date July 1, 1986
Page 2
As an alternative to preparing their own audit samples for a
pretest audit, a testing laboratory may make a request 30 days prior
to the time of the planned pretest audit to EPA's Environmental
Monitoring Systems Laboratory, Quality Assurance Division, Source
Branch, Mail Drop 77A, Research Triangle Park, North Carolina 27711
for known quality control samples. These EPA audit samples are
aqueous potassium nitrate samples and not aqueous sodium nitrate
samples. ,
The relative error for each of two samples should be within 10
percent of true value. The relative error (RE) is an indication of
the bias that may be associated with the analytical phase of Method
7D. Calculate RE using Equation 8-1.
RE _ Cd " Ca x 10Q Equation 8-1
Ca
where:
C, = determined audit sample concentration, mg/dscm, and
C = actual audit sample concentration, mg/dscm.
8.1.2 Audit of Analytical Phase during Field Test (Required) - As
stated in Sections 3.3.9 and 4.4 of 40 CFR 60, Appendix A, Method 7
(49 FR 26522, 06/27/84), when the method is used for enforcement
testing, the analyst must analyze two audit samples along with the
field samples. The testing laboratory should notify the responsi-
ble agency requiring the performance test of the intent to test at
least 30 days prior to the enforcement source test. The responsi-
ble agency will provide two audit samples to be analyzed along with
the field samples from the enforcement source test. The purpose of
this audit is to assess the data quality at the time of the
analysis. If EPA is the agency requiring the performance test, the
testing laboratory should notify the Quality Assurance Management
Office in the respective EPA Regional Office. The addresses of the
EPA Regional Quality Assurance Coordinators are shown in Table 5.1 of
Section 3.0.5 of this Handbook.
The two audit samples and the compliance samples should be con-
currently analyzed in the same manner to evaluate the technique of
the analyst, standards preparation, and computation skills. (Note:
It is recommended that known quality control samples be analyzed
prior to the compliance and audit sample analysis to indicate any
problems. One source of these samples is the Source Branch listed
previously in Subsection 8.1.1.) The same analyst, analytical
reagents, and analytical system must be used both for compliance
samples and the EPA audit samples; if this condition is met,
auditing of subsequent compliance analyses for the same enforcement
agency within 30 days may not be required. An audit sample set may
not be used to validate different sets of compliance samples under
-------
Section No. 3.15.8
Date July 1, 1986
Page 3
the jurisdiction of different enforcement agencies, unless prior
arrangements are made with both enforcement agencies.
Calculate the concentrations in mg/dscm using the specified
sample volume in the audit instructions. (Note: Indication of
acceptable results may be obtained immediately by reporting the audit
results in mg/dscm and compliance results in total mg N02 /sample by
telephone to the responsible enforcement agency.) Include the
results of both audit samples, their identification numbers, and the
analyst's name with the results of the compliance determination
samples in appropriate reports to the EPA regional office or the
appropriate enforcement agency. Include this information with
subsequent compliance analyses for the same enforcement agency during
the 30-day period.
The concentration of each audit sample measured by the analyst
must agree within 10 percent of the actual concentrations. If the
10-percent specification is not met, reanalyze the compliance samples
and audit samples, and include initial and reanalysis values in the
test report.
Failure to meet the 10-percent specification may require retests
until the audit problems are resolved. However, if the audit results
do not affect the compliance or noncompliance status of the affected
facility, the Administrator may waive the reanalysis requirement,
further audits, or retests and accept the results of the compliance
test. While steps are being taken to resolve audit analysis prob-
lems, the Administrator may also choose to use the data to determine
the compliance or noncompliance status of the affected facility.
Other applications of Method 7D (i.e., Performance Specification
Tests) should follow agency recommended or required procedures.
8.1.3 Audit of Data Processing -g Calculation errors are prevalent
in performing this method. ' ' Data processing errors can be
determined by auditing the recorded data on the field and laboratory
forms. The original and audit (check) calculations should agree
within round-off error; if not, all of the remaining data should be
checked. The data processing may also be audited by providing the
testing laboratory with specific data sets (exactly as would appear
in the field), and by requesting that the data calculation be
completed and that the results be returned to the
agency/organization. This audit is useful in checking both computer
programs and manual methods of data processing.
8.2 Systems Audit
A systems audit is an on-site qualitative inspection and review
of the total measurement system (sample collection, sample analysis,
etc.). Initially, a systems audit is recommended for each
enforcement source test, defined here as a series of three runs at
one source. After the test team gains experience with the method,
the frequency of auditing may be reduced--for example, to once for
every four tests.
-------
Section No. 3.15.8
Date July 1, 1986
Page 4
The auditor should have extensive background experience in source
sampling, specifically with the measurement system being audited.
The functions of the auditor are summarized below:
1. Inform the testing team of the results of pretest audits,
specifying any area(s) that need special attention or improvement.
2. Observe procedures and techniques of the field team during
sample collection.
3. Check/verify records of apparatus calibration checks and
quality control used in the laboratory analysis of control samples
from previous source tests, where applicable.
4. Record the results of the audit, and forward them with
comments to the team management so that appropriate corrective action
may be initiated.
While on site, the auditor observes the source test team's overall
performance, including the following specific operations:
1. Setting up and leak testing the sampling train.
2. Preparing the absorbing solution (if performed on-site) and
adding it to the sampling train.
3. Collecting the sample at a constant rate less than 500 cc/min.
4. Determination of C02 content.
5. Sample documentation procedures, sample recovery, and prepara-
tion of samples for shipment.
Figure 8.1 is a suggested checklist for the auditor.
-------
Section No. 3-15-8
Date July 1, 1986
Page 5
Yes
/
No
Comment
OPERATION
PRESAMPLING PREPARATION
1. Knowledge of process conditions
2. Calibration of pertinent equipment, in particular,
dry gas meter and rotameter, prior to each field
test
ON-SITE MEASUREMENTS
3- Leak-testing of sampling train after sample run
4. Preparation of absorbing solution and its addition
to impingers
5. Constant sampling at less than 500 cc/min
6. Measurement of CO- content
7. Recording of pertinent process conditions during
sample collection
8. Maintaining the probe at a given temperature
POSTSAMPLING
9. Control sample analysis - accuracy and precision
10. Sample aliquotting techniques
11. Ion chromatographic technique
a. Preparation of standard nitrate samples
(pipetting)
b. Calibration factor (+7 % for all standards,
optional)
c. Duplicate sample values (+5% of mean, optional)
d. Adequate peak separation
12. Audit results (+10%)
a. Use of computer program
b. Independent check of calculations
COMMENTS
Figure 8.1. Method 7D checklist to be used by auditors.
-------
Section No. 3.15.8
Date July 1, 1986
Page 6
Table 8.1. ACTIVITY MATRIX FOR AUDITING PROCEDURE
Audit
Acceptance limits
Frequency and method
of measurement
Action if
requirements
are not met
Performance
audit of
analytical
phase
Measured RE of the
audit sample should
be less than 10% for
both audit results
Frequency; Once during
every enforcement source
test*
Method; Measure refer-
ence samples and compare
their true values
Review opera-
ting technique
and calibration
check
Data
processing
errors
! Original and checked
calculations agree
within round-off
error
Frequency: Once during
every enforcement source
test*
Method; Independent
calculations starting
with recorded data on
Figures 4.1 and 5-1
Check and cor-
rect all data
for the audit
period repre-
sented by the
sampled data
Systems
audit—
observance
of technique
Operational tech-
nique as described
in this section of
the Handbook
Frequency; Once during
every enforcement source
test* until experience
gained, then every
fourth test
Method; Observation of
techniques assisted by
audit checklist, Fig. 8.1
Explain to
team their devi-
ations from
recommended
techniques and
note on
Fig. 8.1
*As defined here, a source test for enforcement of the NSPS comprises a series of
runs at one source. Source tests for purposes other than enforcement (e.g., a
research project) may be audited at a lower frequency.
-------
Section No. 3.15.9
Date July 1, 1986
Page 1
9.0 RECOMMENDED STANDARDS FOR ESTABLISHING TRACEABILITY
To achieve data of desired quality, two essential considerations
are necessary: (1) the measurement process must be in a state of
statistical control at the time of the measurement, and (2) the
systematic errors, when combined with the random variation (errors or
measurement), must result in an acceptable uncertainty. As evidence
in support of good quality data, it is necessary to perform quality
control checks and independent audits of the measurement process; to
document these data; and to use materials, instruments, and
measurement procedures that can be traced to an appropriate standard
of reference.
Data must be routinely obtained by repeat measurements of stan-
dard reference samples (primary, secondary, and/or working stan-
dards) and the establishment of a condition of process control. The
working calibration standards should be traceable to standards of
higher accuracy.
Class-S weights (made to NBS specifications) are recommended for
the analytical balance calibration. See Section 3.15.5 for details
on balance calibration checks.
The dry gas meter must be calibrated against a wet test meter
that has been verified by an independent method as described in
Section 2.1.1.
Audit samples (as discussed in Section 3.15.8) must be used to
validate test results for compliance determination purposes and are
recommended as an independent check on the measurement process when
the method is performed for other purposes.
-------
10.0 REFERENCE METHOD*,**
Method 7D—Determination of Nitrogen
Oxide Emission* From Stationary Source*
Alkaline-Permanganate/Ion
Chromatographic Method
1. Applicability. Principle. Interferences.
Precision. Bias, and Stability.
1.1 Applicability. The method is
•applicable to the determination of NO.
emissions from fossil-fuel fired steam
generators, electric utility plants, nitric acid
plants, or other sources as specified in the •
regulations. The lower detectable limit is
similar to that for Method 7C. No upper limit
has been established: however, when using
the recommended sampling conditions, the
method has been found to collect NO,
emissions quantitatively up to 1782 mg/NO,/
m». as NOi (932 pm NO,).
' 1.2 Principle. An integrated gas sample is
extracted from the stack and collected in
Hlkaline-potassium permanganate solution:
NO, (NO+NOi) emissions are oxidized to
NOi-. Then N0>- is analyzed by ion
chroma tography.
1.3 Interferences. Possible interferences
are SOs and NH>. High concentrations of SOi
could interfere because SO» consumes MnO<-
(as does NO,) and. therefore, could reduce
the NO, collection efficiency. However, when
sampling emissions from a coal-fired electric
utility plant burning 2.1-percent sulfur coal
with no control of SOi emissions, collection
efficiency was not reduced, in fact
calculations show that sampling 3000 ppm
SO) will reduce the MnO«- concentration by
only 5 percent if all the SOi is consumed in
the first impinger.
NH> is slowly oxidized to NO>- by the
absorbing solution. At 100 ppm NH> in the
gas stream, an interference of 6 ppm NO, (11
mg NOt/m1) was observed when the sample
was analyzed 10 days after collection.
Therefore, the method may not be applicable
to plants using NH» injection to control NO,
emissions unless means are taken to correct
the results. An equation has been developed
to allow quantitation of the interference and
is discussed in Citation 4 of the bibliography.
1.4 Precision and Bias. The method does
not exhibit any bias relative to Method 7. The
within-laboratory relative standard deviation
for a single measurement was approximately
6 percent at 200 to 270 ppm NO,.
1.5 Stability. Collected samples are stable
for at least 4 weeks.
2. Apparatus.
2.1 Sampling and Sample Recovery. The
sampling train is the same as in Figure 7C-1
of Method 7C. Component parts are the same
as in Method 7C. Section 2.1.
2.2.7
2-2.8
2.2.9
Section No. 3.15.10
Date July 1, 1986
Page 1
2.2 Sample Preparation and Analysis.
2.2.1 Magnetic Stirrer. With 25- by 10-mm
Teflon-coated stirring bars.
2.23. Filtering Flask. 500-ml capacity with
sidearm.
2.2.3 Buchner Funnel. 75-mm ID. The
spout equipped with a 13-mm ID by 90-mm
long piece of Teflon tubing to minimize
possibility of aspirating sample solution
during filtration.
Z2.4 Filter Paper. Whatman GF/C. 7.0-cm
diameter.
2.2.5 Stirring Rods.
Volumetric Flask. 250-mI.
Pipettes. Class A.
Erlenmeyer Flasks. 250-ml.
Ion Chroma lograph. Equipped with
an anion separator column to separate NOt-.
a H* suppressor, and* necessary auxiliary
equipment Nonsuppressed and other forms
of ion chroma tography may also be used
provided that adequate resolution of NOj- is
obtained. The system must also be able to
resolve and detect NOi-.
3. Reagents.
Unless otherwise indicated, all reagents
should conform to the specifications
established by the Committee on Analytical
Reagents of the American Chemical Society,
where such specifications are available:
otherwise, use the best available grade.
3.1 Sampling.
3.1.1 Water. Deionized distilled to
conform to ASTM specification D 1193-74.
Type 3 (incorporated by reference— see
§ 60.17).
3.1.2 Potassium Permanganate. 4.0 Percent
(w/w). Sodium Hydroxide. 2.0 Percent (w/w).
Dissolve 40.0 g of KMnO. and 20.0 g of NaOH
in 940 ml of water.
3.2 Sample Preparation and Analysis.
3.2.1 Water. Same as in Section 3.1.1.
3i2 Hydrogen Peroxide. 5 Percent. Dilute
30 percent H>Ot 1:5 (v/v) with water.
3.2.3 Blank Solution. Dissolve 2.4 g of
KMnOi and 1.2 g of NaOH in 96 ml of water.
Alternatively, dilute 60 ml of KMnO./NaOH
solution to 100 ml.
3.2.4 KNOj Standard Solution. Dry
KNO> at 110 ' C for 2 hours, and cool in a
desiccator. Accurately weigh 9 to 10 g of
KNOj to within 0.1 mg. dissolve in water, and
dilute to 1 liter. Calculate the exact NOj-
concentration from the following relationship
ug NOj-/m)=g of KNOi XIO'X •
62.01
101.10
*Method 7C is reproduced in this section in addition to Method 7D
since the latter refers extensively to Method 7C and Method 7C
is not reproduced elsewhere in this Handbook.
** Federal Register, Volume 49, No. 189, September 27, 1984.
-------
Section No. 3.15.10
Date July 1, 1986
Page 2
This solution is stable for 2 months without
preservative under laboratory conditions.
3.2.5 Eluent, 0.003 M NaHCOj/0.0024 M
Na,CCs. Dissolve 1.008 g NaHCOs and 1.018 g
NajCOj in water, and dilute to 4 liters. Other
eiuents capable of resolving nitrate ion from
sulfate and other species present may be
used.
3.2.6 Quality Assurance Audit Samples.
This is the same as in Method 7, section 3.3.8.
When requesting audit samples, specify that
they be in the appropriate concentration
range for Method 7D.
4. Procedure.
4.1 Sampling. This is the same as in
Method 7C. Section 4.1.
4.2 Sample Recovery. This is the same as
in Method 7C. Section 4.2.
4.3 Sample Preparation for Analysis. Note
the level of liquid in the sample container,
and determine whether any sample was lost
during shipment. If a noticeable amount of
leakage has occurred, the volume lost can be
determined from the difference between
initial and final solution levels, and this value
can then be used to correct the analytical
result. Quantitatively transfer the contents to
a 1-liter volumetric flask, and dilute to
volume.
Sample preparation can be started 36 hours
after collection. This time is necessary to
ensure that all NOr- is converted to NOj-.
Take a SO-ml aliquot of the sample and
blank, and transfer to 250-ml Erlenmeyer
flasks. Add a magnetic stirring bar. Adjust
the stirring rate to as fast a rate as possible
without loss of solution. Add 5 percent
HjOa in increments of approximately S ml
using a 5-ml pipette. When the KMnO. color
appears to have been removed, allow the
precipitate to settle, and examine the
supernatant liquid. If the liquid is clear, the
HjO? addition is complete. If the
KMnO. color persists, add more H»Oi, with
stirring, until the supernatant liquid is clear.
Note.—The faster the stirring rate, the less
volume of HjO» that will be required to
remove the KMnO«.) Quantitatively transfer
the mixture to a Buchner funnel contaiing
GF/C filter paper, and filter the precipitate.
The spout of the Buchner funnel should be
equipped with a 13-mm ID by 90-mm long
piece of Teflon tubing. This modification
minimizes the possibility of aspirating sample
solution during filtration. Filter the mixture
into a 500-ml filtering flask. Wash the solid
material four times with water. When
filtration is complete, wash the Teflon tubing,
quantitatively transfer the filtrate to a 250-ml
volumetric flask, and dilute to volume. The
sample and blank are now ready for
NCs analysis.
4.4 Sample Analysis. The following
chromatographic conditions are
recommended: 0.003 M NaHCOj/0.0024 M
NaiCOi eluent solution. (3.2.5). full scale
range 3 pMHO: sample loop, 0.5 ml: flow rate,
2.5 ml/min. These conditions should give a
NOi- retention time of approximately 15
minutes (Figure 7D-1).
/ I
1M.M
Flfim KM. t«"
Establish a stable baseline. Inject a sample
of water, and determine if any NOi- appears
in the chromatogram. If NO>- is present,
repeat the water load/injection procedure
approximately five times: then re-inject a
water sample, and observe the
chromatogram. When no NOi- is present, the
instrument is ready for use. Inject calibration
standards. Then inject samples and a blank.
Repeat the injection of the calibration
standards (to compensate for any drift in
response of the instrument). Measure the
NOj peak height or peak area, and determine
the sample concentration from the calibration
curve.
4.5 Audit analysis. This is the same as in
Method 7. Section 4.4
5. Calibration.
5.1 Dry Gas Metering System (DGM).
5.1.1 Initial Calibration. Same as in
Method 0. Section 5.1.1. For detailed
instructions on carrying out this calibration, it
is suggested that Section 3.5.2 of Citation 3 in
the bibliography be consulted.
5.1 2 Post-Test Calibration Check. Same
as in Method 6. Section 5.1.2.
5.2 Thermometers for DGM and
Barometer. Same as in Method 6. Section 5.2
and 5.4, respectively.
5.3 Calibration Curve for Ion
Chromatograph. Dilute a given volume (1.0 ml
or greater) of the KNCs standard solution to a
convenient volume with water, and use this
solution to prepare calibration standards.
Prepare at least four standards to cover the
range of the samples being analyzed. Use
pipettes for all additions. Run standards as
instructed in Section 4.4. Determine peak
height or area, and plot the individual values
versus concentration in »ig NCy-/ml. Do not
force the curve through zero. Draw a smooth
curve through the points. The curve should be
linear. With the linear curve, use linear
regression to determine the calibration
equation.
-------
6. Calculations.
Carry out calculations, retaining at least
one extra decimal figure beyond that of the
acquired data. Round off figures after final
calculation.
6.1 Sample Volume. Dry Basis. Corrected
to Standard Conditions. Same as in Method
7C. Section 6.1.
62 Total Hg NO, Per Sample.
Section No. 3.15.10
Date July 1, 1935
Page 3
1000 46.01
m-(S-B)X250x x
50 62.01
Where:
m«=Mass of NO,, as NOi. in sample, fig.
S«= Analysis of sample, fig NCs-/ml.
B«> Analysis of blank, ^g NCS-/ml.
250— Volume of prepared sample, ml.
46.01 - Molecular weight of NOi-.
62.01 «= Molecular weight of NO>-.
1000<=Total volume of KMnO« solution, ml.
50-Aliquot KMnO./NaOH solution, ml.
6.3 Sample Concentration.
m
C=Kj ———
3710 (S-BT fEq. 7D-1)
Where:
C«= Concentration of NO, as NO>, dry basis,
mg/dscm.
V«i««>«=Dry gas volume measured by the dry
gas meter, corrected to standard
conditions, dscm.
6.4 Conversion Factors.
1.0 ppm NO=1.247 mg NO/m'at STP.
1.0 ppm NOj-1.912 mg NOj/m'at STP.
Ift3-2.832xl0-lm'.
7. Quality Control.
Quality control procedures are specified in
Sections 4.1 J (flow rate accuracy) and 4.5
(audit analysis accuracy) of Method 7C.
&.~Bibliogrophy.
1. Margeson. ).H.. W.J. Mitchell, J.C. Suggs.
and M.R. Midgett. Integrated Sampling and
Analysis* Methods for Determining NO,
Emissions at Electric Utility Plants. U.S.
Environmental Protection Agency. Research
Triangle Park. N.C. Journal of the Air
Pollution Control Association. 32.1210-1215.
1982.
2. Memorandum and attachment form J.H.
Margeson. Source Branch. Quality Assurance
Division. Environmental Monitoring Systems
Laboratory, to The Record. EPA. March 30.
1983. NHi Interference in Methods 7C and 70.
3. Quality Assurance Handbook for Air
Pollution Measurement Systems. Volume
III— Stationary Source Specific Methods. U.S.
Environmental Protection Agency. Research
Triangle Park. N.C. Publication No. EPA-600/
4-77-027b. August 1977.
4. Margeson. J.H.. et al. An Integrated
Method for determining NO. Emissions at
Nitric Acid Plants. Manuscript submitted to
Analytical Chemistry. April 1984.
-------
Section No. 3.15.10
Date July 1, 1986
Page 4
Method 7C—Determination of Nitrogen
Oxide Emissions From Stationary Sources
AUialine-Pernianganote/Colorimetric
Method
1. Applicability, Principle. Interferences,
Precision, Bias, and Stability.
1.1 Applicability. The method if
applicable to the determination of NO,
emissions from fossil-fuel Tired steam
generators, electric utility plants, nitric acid
plants, or other sources as specified in the
regulations. The lower detectable limit is 13
mg NO./m', as NO« (7 ppm NO.) when
sampling at 500 cc/min for 1 hour. No upper
limit has been established: however, when
using the recommended sampling conditions,
the method has been found to collect NO,
emissions quantitatively up to 1,782 mg NO,/
m', as NO, (932 ppm NO.).
1.2 Principle. Ac integrated gas sample is
extracted from the stack and collected in
alkaline-potassium permanganate solution;
NO, (NQ+NOj) emissions are oxidized to
NOT- and NOi-. The NOr- is reduced to
NOr- with cadmium, and the NOr- is
analyzed colorimetrically.
1J Interferences. Possible interferences
are SOj and NH». High concentrations of SO»
could interfere because SO* consumes MnO<-
(as does NOJ and. therefore, could reduce
the NO, collection efficiency. However, when
sampling emissions from i coal-fired electric
utility plant burning 2.1-percent sulfur coal
with no control of SOj emissions, collection
efficiency was not reduced. In fact.
calculations show that sampling 3000 ppm
SO, will reduce the MnO«- concentration by
only 5 percent if all the S0> is consumed in
the first impinger.
NH> is slowly oxidized to NOr- by the
absorbing solution. At 100 ppm NH» in the
gas stream, an interference of 6 ppn NO. (11
mg NOj/m5) was observed when the sample
was analyzed 10 days after collection.
Therefore, the method may not be applicable
to plants using NH» injection to control NO,
emissions unless means are taken to correct
the results. An equation has been developed
to allow quantilation of the interference and
is discussed in Citation 5 of the bibliography.
1.4 Precision and Bias. The method does
not exhibit any bias relative to Method 7. The
within-laboratory relative standard deviation
for a single measurement is 2.8 and 2-fl
percent at 201 and 268 ppm NO,, respectively.
1.5 Stability. Collected samples are stable
for at least 4 weeks.
2. Apparatus.
2.1 Sampling and Sample Recovery. The
Sampling train is shown in Figure 7C-1. and
component parts are discussed below.
Alternative apparatus and procedures are
allowed provided acceptable accuracy and
precision can be demonstrated.
RESTRICTED omncE WPINGEMS
'ROBE (END PACKED
WITH GLASS WOOL
SURCl TANS
Figure 7C-1. 110 sampling tr»ir
-------
2.1.1 Probe. BorosiUcate glass tubing.
sufficiently heated to prevent water
condensation and equipped with an to-slack
or out-stack filter to remove participate
matter (• plug of glass woo) is satisfactory for
this purpose). Stainless steel or Teflon tubing
may also be used for the probe. (Note.
Mention of trade names or specific products
does not constitute endorsement by the UJ5.
Environmental Protection Agency.)
2.1.2 Impingers. Three restricted-orifice
glass impingers. having the specifications
given in Figure 7C-2. are required for each
sampling train. The impingers must be
connected in series with leak-free glass
connectors. Stopcock grease may be used, if
necessary, to prevent leakage. (The impinjrcrs
can-be fabricated by « glass blower until they
become available commercially.)
IM/II
I4IAO.
DIMINSIONI: M
I
11
f\9ur» K'l. 'fattrictM crlfiu
2.1 J Class Wool Stopcock Creese.
Drying Tube. Valve. Pump. Barometer, and
Vacuum Gauge and Rotameter. Same as in
Method 6, Sections 2.1.3. 2.1.4. 2.1.6. 2.1.7,
2.1.6. 2.1.11. and 2.1.12. respectively.
Section No. 3.15.10
Date July 1, 1986
Page 5
2.1.4 Rate Meter. Rotameter. or
equivalent, accurate to within 2 percent at the
selected flow rate between 400 and 500 cc/
min. For rotameters, a range of 0 to 1 liter/
min is recommended.
2.1.5 Volume Meter. Dry gas meter
capable of measuring the sample volume,
under the sampling conditions of 400 to 500
cc/min for 60 minutes within an accuracy of 2
percent.
2.1.6 Filter-To remove NO. from ambient
air. prepared by adding 20 g of a 5-angstrom
molecular sieve to a cylindrical tube. e.g.. a
polyethylene drying tube.
2.1.7 Polyethylene Bottles. 1-liter, for
sample recovery.
2.1.8 Funnel and Stirring Rods. For sample
recovery.
2.2 Sample Preparation and Analysis.
2.2.1 Hot Plate. Stirring type with 50- by
10-mm Teflon-coated stirring bars.
2.2.2 Beakers. 400-. 600-. and 1000-ml
capacities.
2.2.3 Filtering Flask. 500-ml capacity with
side arm.
2.2.4 Buchner Funnel 75-mm ID. with .
spout equipped with a 13-mm ID by 90-mm
long piece of Teflon tubing to minimize
possibility of aspirating sample solution
during filtration.
2.2.5 Filter Paper. Whatman CF/C. 7.0-cm
diameter.
2.2.6 Stirring Rods.
2.2.7 Volumetric Flasks. 100-, 200- or 250-.
500-. and 1000-ml capacity.
2.2.8 Watch Classes. To cover 600- and
1.000-ml beakers.
2.2.9 Graduated Cylinders. 50- and 250-ml
capacities.
2^.10 Pipettes. Class A
2.2.11 pH Meter. To measure pH from 0.5
to 12.0
2-2.12 Burette. 50-ml with a micrometer
type stopcock. (The stopcock is Catalogue
No. 6225-1-05. Ace Class. Inc.. Post Office
Box 996. Louisville. Kentucky 50201.) Place a
glass wool plug in bottom of burette. Cut off
burette at a height of 43 cm from the top of
plug, and have a glass blower attach a glass
funnel to top of burette such that the
diameter of the burette remains essentially
unchanged. Other means of attaching the
funnel are acceptable.
2.2.13 Glass Funnel. 75-mm ID at the top.
2.2.14 Spectrophotometer. Capable of
measuring absorbance at 540 nm. One-cm
cells are adequate.
2.2.15 Metal Thermometers. Bimetallic
thermometers, range 0 to 150 *C
£2.18 Culture Tubes. 20- by 150-mm.
Kimax No. 45046.
2.2.17 Parafllm "M." Obtained from
American Can Company, Greenwich.
Connecticut 06830.
2.2.18 CO> Measurement Equipment
Same as in Method 3.
3. Reagents.
Unless otherwise indicated, all reagents
should conform to the specifications
established by the Committee on Analytical
Reagents of the American Chemical Society,
where such specifications are available;
otherwise, use the best available grade.
-------
3.1 Sampling.
3.1.1 Water. Deionized distilled to
conform to ASTM specification D 1193-74,
Type 3 (incorporated by reference—see
§ 60.17).
3.1.2 Potassium Permanganate, 4.0 percent
(w/w). Sodium Hydroxide. 2.0 percent (w/w).
Dissolve 40.0 g of KMnO. and 20.0 g of NaOH
in 940 ml of water.
3.2 Sample Preparation and Analysis.
3.2.1 Water. Same as in Section 3.1.1.
3.2.2 Sulfuric Acid. Concentrated H,SO..
3.2.3 Oxalic Acid Solution. Dissolve 48 g
of oxalic acid |(COOH)>-2HiO) in water, and
dilute to 500 ml. Do not heat the solution.
3.2.4 Sodium Hydroxide. 0.5 N. Dissolve
20 g of NaOH in water, and dilute to 1 liter.
3.2.5 Sodium Hydroxide. 10 N. Dissolve
40 g of NaOH in water and dilute to 100 ml.
3.2.6 Ethyienediamine Tetraacetic Acid
(EDTA) Solution. 6.5 Percent. Dissolve 6.5 g of
EDTA (disodium salt) in water, and dilute to
100 ml. Solution is best accomplished by
using a magnetic stirrer.
3.2.7 Column Rinse Solution. Add 20 ml of
6.5 percent EDTA solution to 960 ml of water.
and adjust the pH to 11.7 to 12.0 with 0.5 N
NaOH.
3.2.8 Hydrochloric Acid (HC1). 2 N. Add
86 ml of concentrated HC1 to a 500-ml
volumetric flask containing water, dilute to
volume, and mix well. Store in a glass-
stoppered bottle.
3.2.9 Sulfanilamide Solution. Add 20 g of
sulfanilamide (melting point 165 to 167 *C) to
700 ml of water. Add. with mixing, 50 ml
concentrated phosphoric acid (85 percent),
and dilute to 1000 ml. This solution is stable
for at least 1 month, if refrigerated.
3.2.10 N-(l-Naphthyl)-Ethylenediamine
Dihydrochloride (NEDA) Solution. Dissolve
0.5 g of NEDA in 500 ml of water. An aqueous
solution should have one absorption peak at
320 run over the range of 260 to 400 run.
NEDA. showing more than one absorption
peak over this range, is impure and should
not be used. This solution is stable for at
least 1 month if protected from light and
refrigerated.
3.2.11 Cadmium. Obtained from Matheson
Coleman and BelL 2909 Highland Avenue.
Norwood. Ohio 4521Z as EM Laboratories
Catalogue No. 2001. Prepare by rinsing in 2 N
HC1 for 5 minutes until the color is silver-
grey. Then rinse the cadmium with water
until the rinsings are neutral when tested
with pH paper. CAUTION: Hi is liberated
during preparation. Prepare in an exhaust
hood away from any flame.
3.2.12 NaNOt Standard Solution. Nominal
Concentration. 100 n g NOr-/ml. Desiccate
NaNOi overnight. Accurately weigh 1.4 to 1.6
g of NaNOt (assay of 97 percent NaNOt or
greater), dissolve in water, and dilute to 1
liter. Calculate the exact NOr- concentration
from the following relationship:
purity, % 46.01
ug NOi-/ml=g of NaNOtX XlO'X ——
100 o9.01
Section No. 3.15.10
Date July 1, 1935
Page 6
This solution is stable for at least 6 months
under laboratory conditions.
3.2.13 KNOj Standard Solution. Dry KNO»
at 110 *C for 2 hours, and cool in a desiccator.
Accurately weigh 9 to 10 g of KNQ, to within '
0.1 mg. dissolve in water, and dilute to 1 liter.
Calculate the exact NOi- concentration from
the following relationship:
NCWml-g of KNO.X10'
62.01
101.10
This solution is stable for 2 months without
preservative under laboratory conditions.
3.2.14 Spiking Solution. Pipette 7 ml of the
KNOi standard into a 100-ml volumetric
flask, and dilute to volume.
3.2.15 Blank Solution. Dissolve 2.4 g of
KMnO« and 1.2 g of NaOH in 96 ml of water.
Alternatively, dilute 60 ml of KMnO./NaOH
solution to 100 mi.
3.2.16 Quality Assurance Audit Samples.
Same as in Method 7, Section 3.3.9. When
requesting audit samples, specify that they be
in the appropriate concentration range for
Method 7C.
4. Procedure.
4.1 Sampling.
4.1.1 Preparation of Collection Train. Add
200 ml of KMnO./NaOH solution (3.1.2) to
each of three impingers. and assemble the
train as shown in Figure 7C-1. Adjust probe
heater to a temperature sufficient to prevent
water condensation.
4.1.2 Leak-Check Procedure. A leak-check
prior to the sampling run should be carried
out: a leak-check after the sampling run is
mandatory. Carry out the leak-check(s)
according to Method 6. Section 4.1.2.
4.1.3 Check of Rotameter Calibration
Accuracy (Optional). Disconnect the probe
from the first impinger. and connect the filter .
(2.1.6). Start the pump, and adjust the
rotameter to read between 400 and 500 cc/
mm. After the flow rate has stabilized, start
measuring the volume sampled, as recorded
by the dry gas meter (DOM), and the
sampling time. Collect enough volume to
measure accurately the flow rate, and
calculate the flow rate. This average flow
rate must be less than 500 cc/min for the
sample to be valid: therefore, it is
recommended that the flow rate be checked
as above prior to each test
4.1.4 Sample Collection. Record the initial
DGM reading and barometric pressure.
Determine the sampling point or points
according to the appropriate regulations, e.g..
Section 60.46(c) of 40 CFR Part 60. Position
the tip of the probe at the sampling point.
connect the probe to the first impinger, and
start the pump. Adjust the sample flow to a
value between .400 and 500 cc/min.
CAUTION: HIGHER FLOW RATES WILL
PRODUCE LOW RESULTS. Once adjusted.
maintain a constant flow rate during the
entire sampling run. Sample for 60 minutes.
For relative accuracy (RA) testing of
continuous emission monitors, the minimum
sampling time is 1 hour, sampling 20 minutes
-------
at each traverse point. [Note.—When the SO,
concentration is greater than 1200 ppm. the
•ampling time may have to be reduced to 30
minutes to eliminate plugging of the impinger
orifice with MnO,. For RA tests with SO,
greater than 1200 ppm, sample for 30 minutes
(10 minutes at each point)]. Record the DCM
temperature, and check the flow rate at least
every 5 minutes. At the conclusion of each
run. turn off the pomp, remove probe from the
•tack, and record the final reeding*. Divide
the campie volume by the sampling Ume to
determine the average flow rate. Conduct a
leak-check a* in Section 4.1.2. If a leak is
found, void the test nia, or we procedures
acceptable to the Administrator to adjust the
sample volume for the leakage.
4.1.5 CO, Measurement. During sampling.
measure the CO, content of the slack gas
near the sampling point using Method 3. The.
single-point grab sampling procedure is
adequate, provided the measurements are
made at least three times—near the start
midway, and before the end of a run and the
average CO, concentration is computed. The
Orsat or Fyrite analyzer may be used for this
analysis.
4-2 Sample Recovery. Disconnect the
impingers. Pour the contents of the hnpirrgers
into a 1-liter polyethylene bottle using a
funnel and a stirring rod (or other means) to
prevent spillage. Complete the quantitative
transfer by rinsing the impingers and
connecting tubes with water until the rinsings
that are clear to light pink, and add the
rinsings to the bottle. Mix tbe sample, and
mark the solution level. Seal and identify the
sample container.
4J Sample Preparation for Analysis.
Prepare a cadmium reduction column as
follows: Fill the burette (&2.12) with water.
Add freshly prepared cadmium slowly with
tapping until no further settling occurs. The
height of the cadmium column should be 39
cm. When not in use. store the column under
rinse solution (3.2.7). (Note.—The column
should not contain any bands of cadmium
fines. This may occur if regenerated column
is used and will greatly reduce the column
lifetime.)
Note the level of liquid in the sample
container, and determine whether any sample
was lost during shipment If a noticeable
amount of leakage has occurred, the volume
lost can be determined from the difference
between initial and final solution levels, and
this value can then be used to correct the
analytical result. Quantitatively transfer the
contents to a 1-liter volumetric flask, and
dilute to volume.
Take a 100-ml aliquot of the sample and
blank (unexposed KMnO./NaOH) solutions,
and transfer to 400-ml beakers containing
magnetic stirring bars. Using a pH meter, add
concentrated H,SO. with stirring until a pH
of 0.7 is obtained. Allow the solutions to
stand for 15 minutes. Cover the beakers with
watch glasses, and bring the temperature of
the solutions to 50 "C. Keep the temperature
below 60 *C Dissolve 4.8 g of oxalic acid in a
minimum volume of water, approximately SO
ml. at room temperature. Do not heat the
solution. Add this solution slowly, in
increments, until the KMnO. solution
Section No. 3.15.10
Date July 1, 1986
Page 7
becomes colorless. If the color is not
completely removed, prepare some more of
the above oxalic acid solution, and add until
a colorless solution is obtained. Add an
excess of oxalic acid by dissolving 1.6 g of
oxalic acid in SO ml of water, and add 6 ml of
this solution to the colorless solution. If
suspended matter is present, add
concentrated HfSO4 until a clear solution is
obtained.
Allow the samples to cool to near room
temperature, being sure that the samples are
still clear. Adjust the pH to 11.7 to 12-f) with
10 N NaOH. Quantitatively transfer the
mixture to a Buchner funnel containing CF/C
filter paper, and filter the precipitate. Filter
the mixture into a SOO-ml filtering flask. Wash
the solid material four times with water.
When filtration is complete, wash the Teflon
tubing, quantitatively transfer tbe filtrate to a
SOO-ml volumetric flask, and dilute to volume.
The samples are now ready for cadmium
reduction. Pipette a 50-ml aliquot of the
sample into a 150-ml beaker, and add a
magnetic stirring bar. Pipette in 1.0 ml of 6.5
percent EDTA solution, and mix.'
Determine the correct stopcock setting to
establish a flow rate of 7 to H ml/min of
column rinse solution through the cadmium
reduction column. Use a 50-ml graduated
cylinder to collect and measure the solution
volume. After the last of the rinse solution
has passed from the funnel into the burette,
but before air entrapment can occur, start
adding the sample, and collect it in a 250-ml
graduated cylinder. Complete the
quantitative transfer of the sample to the
column as the sample pusses through the
column. After the last of. the sample has
passed from the funnel into the burette, start
adding 60 ml of column rinse solution, and
collect the rinse solution until the solution
just disappears from the funnel.
Quantitatively transfer the sample to a 200-ml
volumetric flask (250-ml may be required).
and dilute to volume. The samples are now
ready for NOr-4- analysis. (Note.— Both the
sample and blank should go through this
procedure. Additionally, two spiked samples
should be run with every group of samples
passed through the column. To do this.
prepare two additional 50-ml aliquots of the
sample suspected to have tbe highest NOr-
concentration, and add 1 ml of the spiking
solution to these aliquots. If the spike
recovery or column efficiency (see 62.1) is
below 05 percent, prepare a new column, and
repeat the cadmium reduction).
4.4 Sample Analysis. Pipette 10 ml of
sample into a culture tube. (Note.—Some test
tubes give a high blank NOr- value but
culture tubes do not.) Pipette in 10 ml of
sulfanilamide solution and 1.4 ml of NEDA
solution. Cover the culture tube with
parafilm. and mix the solution. Prepare a
blank in the same manner using the sample
from treatment of the unexposed KMnO«/
NaOH solution (3.1.2). Also, prepare a
calibration standard to check the slope of the
calibration curve. After a 10-minuie color
development interval, measure the
absorbance at 540 nm against water. Rend pg
NOr-/ml from the calibration curve. If the
-------
absoruance it greater than that of the highest
calibration standard, pipette less than 10 ml
of sample and enough water to make the total
sample volume 10 ml. and repeat the
analysis. Determine the No? concentration
using the calibration curve obtained in
Section 5.3.
4.5 Audit Analysis. This is the same as in
Method 7, Section 4.4.
5. Calibration.
5.1 Dry Gas Metering System (DCM).
5.1.1 Initial Calibration. Same as in Method
6. Section 5.1.1. For detailed instructions on
carrying out this calibration, it is suggested
that Section 3.5.2 of Citation 4 in the
bibiography be consulted.
5.1.2 Post-Test Calibration Check. Same
•s in Method & Section 5.1.2.
5.2 Thermometers for DCM and
Barometer. Same as in Method 6. Sections 5.2
and 5.4. respectively.
5.3 Calibration Curve for
Spectrophotomeier. Dilute 5.0 ml of the
NaNCs standard solution to 200 ml with
water. This solution nominally contains 25 Kg
NCWml. Use this solution to prepare
calibration standards to cover the range of
0.25 to 3.00 fig NOr-/ml. Prepare a minimum
of three standards each for the linear and
slightly nonlinear (described below) range of
the curve. Use pipettes for all additions.
Run standards and a water blank as
instructed in Section 4.4. Plot the net
absorbance vs ngNO»-/ml. Draw a smooth
curve through the points. The curve should be
linear up to an absorbance of approximately
1.2 with a slope of approximately 0.53
absorbance units/ pig NOi-/ml. The curve
should pass through the origin. The curve is
slightly nonlinear from an absorbance of 1.2
to 1.6.
6. Calculations.
Carry out calculations, retaining at least
one extra decimal figure beyond that of the
acquired data. Round off figures after final
calculation.
6.1 Sample volume, dry basis, corrected to
standard conditions.
Section No. 3.15.10
Date July 1, 1986
Page 8
6.2 Total >ig NOs Per Sample.
6il Efficiency of Cadmium Reduction
Column. Calculate this value as follows:
E
- y)'2°0
"
s x i-Q
SOT
(Eq. 7C-2)
= 269.6 (x - y)
s
m*
„
-
(£0. 7C-1)
Where:
V.ua>— Dry gas volume measured by the dry
gas meter, corrected to standard
conditions, dscm.
Vw«=Dry gas volume as measured by the dry
gas meter, don.
Ye Dry gas meter calibration factor.
XK Correction factor for CO> collection.
100
Where:
E= Column efficiency, unitless.
x= Analysis of spiked sample, fig NO—/inl.
y= Analysis of unspiked sample, jig NO—/
ml.
200= Final volume of sample and blank after
passing through the column, ml
9= Concentration of spiking solution, ng
NOj/ml.
1.0= Volume of spiking solution added, ml.
46.01 = us NCs-/»imole.
62.01= >ig NOs-/fimole.
6.2.2 Total fig NCv
(S-B) 500 1000 (2X10T (S-B)
- X200X — x - = -
E 50 100 E
"(Eq. 7C-3)
Where:
m = Mass of NO., as NO,, in sample, >ig.
S~ Analysis of sample. u.g NCWml.
B=Analysis of blank. fig NO— /ml.
500»=Total volume of prepared sample, ml.
50 =» Aliquot of prepared sample processed
through cadmium column, ml.
100= Aliquot of KMnO./NaOH solution, ml.
1000=Total volume of KMnOJNaOH
solution ml.
6.3 Sample Concentration. .
C-K,
Where:
C= Concentration of NO, as NO:, dry basis,
mg/dscm.
" IOO-*COi(v/v)
PtarK Barometric pressure, mm fig.
P^c Standard absolute pressure, 760 mm Hg.
TK«» Average dry gas meter absolute
temperature, *K.
T^B Standard absolute temperature, 293 'K.
K, .03858 'K/mmHg.
6.4 Conversion Factors.
1.0 ppm NO=1.247 mg NO/m' at STP.
1.0 ppm NOs =1.912 mg NO2/m'at STP.
1 ftj«2.832xl0'1mj.
7. Quality Control.
Quality control procedures are specified in
Sections 4.1.3 (flow rate accuracy): 4.3
(cadmium column efficiency): 4.4 (calibration
curve accuracy); and 4.5 (audit analysis
accuracy).
8. Bibliography. ,
1. Marge'son. |.H.. W.J. Mitchell. ).C. Sungs.
and M.R. Mtdgett. Integrated Samplins and
Analysis Methods for Determining NO,
Emissions at Electric Utility Plants. U.S.
Environmental Protection Agency. Research
Triangle Park, N.C. Journal of the Air
Pollution Control Association. J21210-1215.
1982.
-------
Section No. 3.15.10
Date July 1, 1986
Page 9
2. Memorandum and attachment from J.H.
Margeson. Source Branch. Quality Assurance
Division. Environmental Monitoring System*
Laboratory, to The Record. EPA. March 30.
1983. NHi Interference in Methods 7C and 7D.
3. Margeson. |.H.. ).C Suggs, and M.R.
Midgett. Reduction of Nitrate to Nitrite with
Cadmium. Anal. Chem. 52:1955-57.1960.
4. Quality Assurance Handbook for Air
Pollution Measurement Systems. Volume
III—Stationary Source Specific Methods.
August 1977. U.S. Environmental Protection
Agency. Research Triangle Park. N.C.
Publication No. EPA-4>00/4-77-027b. August
1977.
5. Margeson. J.H.. et al. An Integrated
Method for Determining NO, Emissions at
Nitric Acid Plants. Manuscript submitted to
Analytical Chemistry. April 1984.
-------
Section No. 3.15.11
Date July 1, 1986
Page 1
11.0 REFERENCES
1. Federal Register, Volume 49, No. 189, September 27, 1984.
Method 7D - Determination of Nitrogen Oxide Emissions From
Stationary Sources, Alkaline-Permanganate/Ion Chromato-
graphic Method.
2. Margeson, J. H., J. E. Knoll, M. R. Midgett,,G. B. Oldaker
III, K. R. Loder, P. M. Grohse, and W. F. Gutknecht.
Integrated Method for Determining NO Emissions at Nitric
Acid Plants. Analytical Chemistry, §6, 2607, 1984.
3. Small, H. T., S. Stevens, and W. C. Bauman. Novel Ion
Exchange Chromatographic Method Using Conductimetric
Determination. Analytical Chemistry, 47, 11:801, 1975.
4. Johnson, E. L. and R. Stevenson. Basic Liquid
Chromatography. Varian Associates, Inc., 1978.
5. Yost, R. W., L. S. Ettre, and R. D. Conlon, Practical
Liquid Chromatography, An Introduction. Perkin-Elmer,
1980.
6. Smith, F. C., Jr., and R. C. Chang. The Practice of Ion
Chromatography. John Wiley and Sons, Inc., New York,
1983.
7. Stevens, T. S. and M. A. Langhorst. Agglomerated Pellicular
Anion-Exchange Columns for Ion Chromatography. Analytical
Chemistry, 54, 6:950, 1982.
8. Stevens, T. S., G. L. Jewett, and R. A. Bredeweg. Packed
Hollow Fiber Suppressors for Ion Chromatography.
Analytical Chemistry, 54, 7:1206, 1982.
9. Mulik, J. D., and E. Sawicki. Ion Chromatography.
Environmental Science and Technology, 13, 7:804, 1979.
10. Stevens, T. S., J. C. Davis, and H. Small. Hollow Fiber Ion
Exchange Suppressor for Ion Chromatography. Analytical
Chemistry, 53_, 9:14,88, 1981.
11. Stevens, T. S. Packed Fibers and New Columns Speed,
Simplify Ion Chromatography. Industrial Research and
Development, September 1983.
12. Gjerde, D. T., J. S. Fritz, and G. Schmuckler. Anion
Chromatography with Low-Conductivity Eluents. Journal of
Chromatography. 186, 509, 1979.
-------
Section No. 3.15.11
Date July 1, 1986
Page 2
13. Margeson, J. H., W. J. Mitchell, J. C. Suggs, and M. R.
Midgett. Integrated Sampling and Analysis Methods for
Determining NO Emissions at Electric Utility Plants.
Journal of the Air Pollution Control Association, 32, 1210,
1982.
14. Eubanks, D. R., and J. R. Stillian. Care of Ion Chroma-
tography Columns. Liquid Chromatography, 2^, 2:74, 1984.
15. Hamil, Henry F., et. al. The Collaborative Study of EPA
Methods 5, 6, and 7 in Fossil Fuel Fired Steam Generators.
Final Report, EPA-650/4-74-013, May 1974.
16. Hamil, H. F., and R. E. Thomas. Collaborative Study of
Method for the Determination of Nitrogen Oxide Emissions
from Stationary Sources (Nitric Acid Plants).
EPA-650/4-074-028, May 1974.
17. Hamil, Henry F. Laboratory and Field Evaluations of EPA
Methods 2, 6, and 7. Final Report, EPA Contract No.
68-02-0626, Southwest Research Institute, San Antonio,
Texas, October 1973.
-------
Section No. 3.15.12
Date July 1, 1986
Page 1
12.0 DATA FORMS
Blank data forms are provided on the following pages for the
convenience of the Handbook user. Each blank form has the customary
descriptive title centered at the top of the page. However, the
section-page documentation in the top right-hand corner of each page
of other sections has been replaced with a number in the lower
right-hand corner that will enable the user to identify and refer to
a similar filled-in form in a text section. For example, Form
M7D-1.1 indicates that the form is Figure 1.1 in Section 3.15.1 of
the Method 7D section. Future revisions of these forms, if any, can
be documented by 1.2A, 1.2B, etc. Twelve of the blank forms listed
below are included in this section. Four are in the Method
Highlights subsection as shown by the MH following the form number.
Form Title
1.1 Procurement Log
2.2 Wet Test Meter Calibration Log
2.4A and 2.4B Dry Gas Meter Calibration Data Form (English and
Metric Units)
2.5 (MH) Pretest Sampling Checks
3.1 (MH) Pretest Preparations
4.1 Field Sampling Data Form for NO
X
4.2 Sample Label
4.3 Sample Recovery and Integrity Data
4.4 (MH) On-Site Measurements
5.1 Analytical Data Form for Analysis of Calibration
Standards
5.3 Analytical Laboratory Data Form for Analysis of
Field Samples
5.4 (MH) Posttest Operations
6.1A and 6.IB Nitrogen Oxide Calculation Form (English and
Metric Units)
8.1 Method 7D Checklist to be Used by Auditors
-------
PROCUREMENT LOG
Item description
Qty-
Purchase
order
number
Vendor
Date
Ord.
Rec.
Cost
Disposition
Comments
Quality Assurance Handbook M7D-1.1
-------
wet test meter serial number
WET TEST METER CALIBRATION LOG
Date
Range of wet test meter flow rate
Volume of test flask Vs =
Satisfactory leak check?
Ambient temperature of equilibrate liquid in wet test meter and reservoir
Test
number
1
2
3
Manometer
reading, a
mm H2O
Final
volume (Vf) ,
L
Initial
volume (V^) ,
L
Total
volume, (Vm)
L
Flask
volume (V ) ,
L
Percent
error, °
%
Must be less than 10 mm (0.4 in.) H2O.
Calculations:
= Vf *
% error = 100 (V - V )/V0 =
(+1%).
Signature of calibration person
Quality Assurance Handbook M7D-2.2
-------
DRY GAS METER CALIBRATION DATA FORM (ENGLISH UNITS)
Date
Calibrated by
Meter box number
Wet test meter number
Barometer pressure, P =
in. Hg Dry gas meter temperature correction factor
Wet test
meter
pressure
drop
'a
in. UjO
Rota-
ineter
setting
'
ft3/min
Wet test
meter gas
volume
'
o
°F
Average
gas temp
'C
°F
Time
of run
(9),d
min
Average
ratio
(Yi),6
(Y_ ),f
i
D expressed as negative number.
Volume passing through meter. Dry gas volume is minimum for at least five revolutions of the meter.
c The average of td and td if using two thermometers; the actual reading if using one thermometer.
d i o
The time it takes to complete the calibration run.
e With Y defined as the average ratio of volumes for the wet test and the dry test meters, Y^^ = Y +0.02 Y for
calibration and YA = Y +0.05 Y for the posttest checks; thus,
'w (td •*" 460°F) [Pm + (pn/13-6)]
Vd (tw + 460°F) (Pm)
Y, + Y2 + Y3
(Eq. 1) and
Y =
(Eq. 2)
With Y defined as the average ratio of volumetric measurement by wet test meter to rotameter.
Tolerance Y = 1 +J0.05 for calibration and Y +0.1 for posttest checks.
V
460
F) [pm + (Dm/13.6)J
ri
0 (tw + 460°F) (Pm) (Rs)
(Eq. 3)
and
(Eq. 4)
Quality Assurance Handbook M7D-2.4A
-------
Date
DRY GAS METER CALIBRATION DATA FORM (METRIC UNITS)
Calibrated by Meter box number Wet test meter number
Barometer pressure, P =
in. Hg Dry gas meter temperature correction factor
Wet test
meter
pressure
drop
(Dm>'a
mm H2O
Rota-
meter
setting
Us),
cc/min
Wet test
meter .gas
volume
'
o
°C
Average
gas temp
-------
Plant name
Location _
Operator
FIELD SAMPLING DATA FORM FOR NO
City
Date
Sample no.
Probe length/material
Meter box no.
Probe setting
Meter factor (Y)
Sampling point location(s)
Rotameter setting
Initial leak check?
CO- concentration
(1)
(2)
Bar press mm (in.) Hg
Rotameter check?
Final leak check?
(3)
avg
Sampling
time,
min
Total
Clock
time
24 h
Dry gas
meter
readings
L (ft3)
Total
Sample flow
rate setting,
cc/min (ft^/min)
Sample volume
metered,,(V )
L (ft3) m
V
m
avg
Percent
deviation, a
%
Avg
dev
Dry gas
meter temp,
°C (°F)
Avg
Percent deviation = m m avg x 100 (must be less than 10 percent).
V avg
Quality Assurance Handbook M7D-4.1
-------
SAMPLE LABEL
Plant
Site
Date
Front rinse Fi
Back rinse I 1 B£
Solution
Volume: Initial
Cleanup by
Samp
Run
-ont filter |
ick filter |
L
F
City
le type
number
Front solution
Back solution
evel marked
inal
L
—
n
n)
£
«
Quality Assurance Handbook M7D-4.2
-------
SAMPLE RECOVERY AND INTEGRITY DATA
Plant
Sampling location
Field Data Checks
Sample recovery personnel
Person with direct responsibility for recovered samples
Sample
number
1
2
3
Blank
Sample
identification
number
Date
of
recovery
Liquid
level
marked
Stored
in locked
container
Remarks
Signature of field sample trustee
Laboratory Data Checks
Lab person with direct responsibility for recovered samples
Date recovered samples received
Analyst
Sample
number
1
2
3
Blank
Sample
identification
number
Date
of
analysis
Liquid
level
marked
Stored
in locked
container
Remarks
Signature of lab sample trustee
Quality Assurance Handbook M7D-4.3
-------
ANALYTICAL DATA FORM FOR ANALYSIS OF CALIBRATION STANDARDS
Plant
Date
Location
Analyst
Standard
identifier
Std 1
Std 2
Std 3
Std 4
Standard
concentration (x)
(yg/ml N03 )
Integrator Response,
Peak Height or Area
(y), (mm)
1
2
Avg
Predicted
standard
concentration
(P)
(yg/ml NO- )
Deviation
(*)
Equation for Linear Calibration Curve, Average Response as a Function of Standard
Concentration
y = mx + b = (
) x
where:
y = instrument response (mm or area count) =
m = calibration curve slope
mm or area count
yg NO ~/ml
x = standard concentration (y g NO- /ml) =
b = I = intercept term (mm or area count) =
Predicted Standard Concentration (P)
P (yg N03"/ml) =
Average Instrument Response (y) - Intercept (I)
Calibration Curve Slope (m)
P (for first standard) =
yg N03"/ml
Deviation
Deviation (%)
Deviation
(of first set
of standards)
P (yg NO ~/ml) - x (yg N(
x (yg N03~/mL)
x 100# =
Quality Assurance Handbook M7D-5.1
-------
ANALYTICAL LABORATORY DATA FORM FOR ANALYSIS OF FIELD SAMPLES
Date samples received
Plant
Location
Calibration curve slope (m)
Date samples analyzed
Run number(s)
Analyst
Intercept term (I)
Field
sample
number
Field
Blank
Analysis
number
1st
2nd
1st
2nd
1st
2nd
1st
2nd
Instrument
response (y)
(mm or area counts)
Concentration of
analysis sample
(yg/ml NO^ )
.s .
Average
Concentration of
analysis sample
(yg/ml NO- )
S =
S =
S =
B =
Deviation
(*)
Concentration of
Analysis Sample
(Vg NO-"/ml)
Concentration
(of first sample)
Deviation
Deviation
(of first
standard set)
Instrument Response (y) - Intercept (I)
Calibration Curve Slope (m)
Sample Concentration - Average Concentration
Average Concentration
x 100% =
Quality Assurance Handbook M7D-5.3
-------
NITROGEN OXIDE CALCULATION FORM (ENGLISH UNITS)
Sample Volume
Vm =
Pbar
dcf, Y =
in. Hg, T =
m
X = .
°R
Vm(std) = 17'64 X Y m bar '
Tm
dscf
Equation 6-1
Total yg NO2 Per Sample
S = ug/ml, B = yg/ml
m = 3710 (S-B) = vg of N02
Equation 6-2
Sample Concentration
C = 2.205 x 10
-9 m
x 10~5 Ib/dscf
V
m(std)
Equation 6-3
Sample Concentration in ppm
ppm N02 = 8.375 x 10 C = . ppm N02
Equation 6-4
Quality Assurance Handbook M7D-6.1A
-------
V • 0.0
m
NITROGEN OXIDE CALCULATION FORM (METRIC UNITS)
Sample Volume
m3, Y =
>bar mm Hg, Tm = _ ._ °K
Vm(std) = °*3858 x Y
0.0
dscm Equation 6-1
m
Total yg N02 Per Sample
ug/ml, B = wg/ml
m = 3710 (S-B) = vgof N02
Equation 6-2
C = 10
-3
m
'm(std)
Sample Concentration
mg N02/dscm
Equation 6-3
Sample Concentration in ppm
ppm NO2 = 0.5228 C = ppm
NO,
Equation 6-4
Quality Assurance Handbook M7D-6.1B
-------
METHOD 7D CHECKLIST TO BE USED BY .AUDITORS
Yes
No
Comment
OPERATION
PRESAMPLING PREPARATION
1. Knowledge of process conditions
2. Calibration of pertinent equipment, in particular,
dry gas meter and rotameter, prior to each field
test
5.
6.
7.
ON-SITE MEASUREMENTS
Leak- testing of sampling train after sample run
Preparation of absorbing solution and its addition
to impingers
Constant sampling at less than 500 cc/min
Measurement of CO- content
Recording of pertinent process conditions during
sample collection
8. Maintaining the probe at a given temperature
POSTSAMPLING
9. Control sample analysis - accuracy and precision
10. Sample aliquotting techniques
11. Ion chromatographic technique
a. Preparation of standard nitrate samples
(pipetting)
b. Calibration factor (+7 % for all standards,
optional)
c. Duplicate sample values (+5# of mean, optional)
d. Adequate peak separation
12. Audit results (+10%)
a. Use of computer program
b. Independent check of calculations
COMMENTS
OU.S. GOVERNMENT PRINTING OFFICE: 1991-548-18 Tt 0522
Quality Assurance Handbook M7D-8.1
------- |
|
|