I
I
United States Office of Air Quality EPA-450/2-78-041
Environmental Protection Planning and Standards OAQPS No. 1.2-115
Agency Research Triangle Park NC 27711 October 1978
_
Guideline Series
Measurement of
Volatile Organic
| Compounds
I
I
I
I
I
I
-------
I
EPA-450/2-78-041
| OAQPS No. 1.2-115
I
I
Measurement of Volatile
Organic Compounds
I
I
I
I
I Emissions Measurement Branch
Emission Standards and Engineering Division
I
I
I
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
_ October 1978
-------
I
I
I
I
OAQPS GUIDELINE SERIES
The guideline series of reports is being issued by the Office of Air Quality Planning and Standards (OAQPS)B>
provide information to state and local air pollution control agencies; for example, to provide guidance on the
acquisition and processing of air quality data and on the planning and analysis requisite for the mamtenance«f
air quality. Reports published in this series will be available- as supplies permit-from the Library Services Off»
(MD35), U.S. Environmental Protection Agency. Research Triangle Park, North Carolina 2771 1; or, for a nominal
fee, from the National Technical Information Service, 5285 Port Royal Road, Springfield, Virginia 22161.
I
I
I
I
Publication No. EPA-450/2-78-041
(OAQPS No 1.2-115)
n
I
I
I
I
I
I
-------
I
I PREFACE
« Emphasis on the control of volatile organic compounds through the
State Implementation Plans, new source performance standards, and
I national emission standards for hazardous air pollutants has
created a need for standardized test procedures. In setting national
| performance standards for new sources and national emission standards
_ for hazardous air pollutants, the Environmental Protection Agency has
followed a policy of establishing a reference method for each regulated
source category and pollutant. Under the State Implementation Plan
process, however, test methods and erocedures are defined by the States.
| Thus, the case-by-case approach used by the Environmental Protection
Agency for national standards could conflict with State established
methods. In addition, the case-by-case approach does not pro-
vide sufficient guidance to the States in their efforts to develop
regulations for a large number of sources and organic compounds.
The purpose of this document, therefore, is to provide guidance
to the States on the measurement of volatile organic compounds from a
diversity of sources and pollutants that is consistent with the methodology
being applied by the Environmental Protection Agency as it develops
regulations for specific sources and pollutants.
I
I
I
I
I
-------
I
CONTENTS
Page No.
METHODS FOR DETERMINING VOLATILE ORGANIC COMPOUNDS "
AS CARBON IN STATIONARY SOURCES 1
INTRODUCTION 1 1
RETIONALE FOR SELECTING ORGANIC CARBON 1
RECOMMENDED REFERENCE METHOD 6
ALTERNATE METHODS 8 I
SCREENING METHODS 9
REGULATORY LANGUAGE 10 I
ATTACHMENT 1. REFERENCE METHOD FOR DETERMINATION OF TOTAL
GASEOUS NONMETHANE ORGANIC EMISSIONS AS CARBON - AUTOMATED I
ANALYZER VERSION 11
1. Principle and Applicability 11
2. Range and Sensitivity 11
3. Interferences 11
4. Apparatus 11
5. Reagents 12 |
6. Analyzer Performance Specifications 14
7. Procedure 15
8. Calculations ; 17
9. References 18 *
ATTACHMENT 2. DETERMINATION OF TOTAL GASEOUS NONMETHANE ORGANIC
EMISSIONS AS CARBON: MANUAL SAMPLING AND ANALYSIS PROCEDURE 19
1. Principle and Applicability 19
2. Apparatus 19 |
3. Reagents 28
4. Procedure 30 M
5. Calculations 38
6. Bibliography 40
ATTACHMENT 3. ALTERNATE TEST METHOD FOR DIRECT MEASUREMENT OF
TOTAL GASEOUS ORGANIC COMPOUNDS USING A FLAME IONIZATION ANALYZER. 42 I
1. Principle and Applicability 42
2. Range and Sensitivity 43 |
3. Interferences 43
4. Apparatus 45 _
5. Reagents 46
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
CONTENTS (CONTINUED)
Page No.
ATTACHMENT 3.
6. System Performance Specifications 48
7. Procedure 49
8. Calculations 52
9. References 54
TECHNICAL REPORT DATA SHEET 55
-------
I
I METHODS FOR DETERMINING VOLATILE ORGANIC COMPOUNDS
AS CARBON IN STATIONARY SOURCES
Introduction
Volatile organic compound (VOC) emission control regulations are
being developed by EPA and by State and local agencies to meet the
j oxidant control needs. In some cases, the regulations are in
terms of the volatile organic content of solvents. In other cases,
I they cover organic volume or mass concentrations, mass emission
rates, and control equipment efficiencies. Regardless of the
I approach taken in the regulation, consideration must be given to the
_ expression of emission limits in terms of what can be measured, and
to the cost and practicality of the test methods.
I One concept of volatile organic emission measurement is
the determination of organic carbon mass concentration.
| The rationale for selecting this concept and conceptual approach
for writing regulations in terms of volatile organic carbon are dis-
cussed herein, and two specific test methods are presented to implement
the recommended approach.
Rationale for Selecting Organic Carbon
I In considering volatile organic compound test methods one must
recognize that organic emissions normally consist of a
mixture of compounds and that there is presently no detection
technique having an inherent, quantitative response to the
total molecular structure of the mixture. Several detection
techniques respond to organic compounds; however, the response
can vary widely from compound to compound and may, there-
I
fore, not be proportional to the total organic mass or volume in
a mixture.
-------
I
I
2
Such is the case of the flame ionization detector (FID), the
most commonly used detector for organic measurements. The FID re-
sponse can vary from compound to compound because it is a function
of the number of carbon atoms, the type of bonds, and the elements
present in the organic molecules. Thus, if the volatile organic I
emission limit is expressed in terms that require the measurement
of the total molecular structure of the organic emissions, the vari-
able response of the flame ionization detector must be overcome by
one of the methods described below.
1. Gas Chromatograph/Flame Ionization Detector. This method I
involves the separation of the organic components into dis-
crete compounds using gas chromatography (GC). The compounds 8
are identified, and the FID is calibrated for each of the
identified compounds. The compounds are then measured
individually, and the total mass concentration is determined
by adding the individual mass concentration values; methane
can be identified and excluded from the results. This |
method may be eractical where only two or three compounds are
emitted, such as in maleic anhydride plants; but if it is
applied to sources that emit numerous organics, the time fl
and expense would be formidable. For example, over 20
peaks were noted in a preliminary study of emissions from the |
manufacture of nitrobenzene. .
2. Direct Flame Ionization Detector with Emission Stream Character!- "
zation. This method involves direct measurement with an FID analyBr,
I
-------
I
I
3
with prior characterization of the gas stream and knowledge
| that the detector responds predictably to the organic
_ components in the stream. If present, methane will, of
course, also be measured.
In practice, this method can be applied to the determina-
tion of the mass concentration of the total molecular struc-
ture of the organic emissions under the following limited con-
ditions: (1) where only one compound is known to exist, (2)
when the organic compounds consist of only hydrogen and carbon,
(3) where the relative percentage of the compounds is known or
can be determined, and the FID response to the compounds is
known; (4) where a consistent mixture of compounds exists be-
fore and after emission control and only the relative concen-
trations are to be assessed, or (5) where the FID can be cali-
brated against mass standards of the emissions (solvent emis-
sions, for example).
In the case of volatile organic solvents, accurate measure-
ments by direct FID analyzers without calibration with solvent
standards are seldom possible because these solvents are often a
mixture of multiple unknown compounds. Even if the emissions
can be separated and identified using a GC, accurate determina-
tion of the average FID response is often impractical. In addi-
tion, the emissions may be altered as they pass through a control
I device: for example, they may be partially oxidized in an
incinerator or selectively retained in an adsorber. In such cases
the measurement is more difficult and cannot be corrected with
I
-------
I
I
4
solvent standards; therefore even the determination of per-
cent control efficiency can be a difficult problem. I
Another applicable measurement technique involves an oxida-
tion-reduction analysis; however, the results of this technique |
are in terms of organic carbon and not a mass concentration of the «
total molecular structure. In this approach, the nonmethane organic
compounds are separated from other carbon compounds and are then fl
oxidized to COp. The resultant CCL is subsequently reduced to
methane, which is then measured with an FID. The C0? from the |
combustion step can also be measured with a nondispersive _
infrared (NDIR) analyzer; however, the NDIR is not as sensi- "
tive as the FID and is therefore limited to high concentration
levels. One limitation to the oxidation-reduction analysis
is that the equipment required is somewhat complex and is I
unlikely to be made available in a portable form.
Consideration of the various measurement approaches
indicates that organic emission regulations expressed in
terms of the measurement of organic carbon could be applied
to a wide range of volatile compounds. The measurement of I
organic carbon can be used to assess directly the efficiency of con-
trol devices such as incinerators or adsorbers. By performing volume-
trie flow rate measurements, one can then determine organic carbon emis-
sion rates. Organic carbon content can also be related to volatile
I
I
I
-------
I
I
5
content of solvents or surface coatings.
I Costs, logistics, and other practicalities of source testing
M may, under limited conditions, make other test methods more
* desirable for routine compliance determinations. Three distinct
I categories of test methods are therefore recommended for use with
volatile organic compound regulations expressed in terms of organic
J carbon: a reference method, alternate methods, and screening methods,
_ These categories are described as follows:
1. Reference Method. This method would be applicable to
all regulated sources and would be accurate in reference
to the emission standard. OAQPS recommends that the
reference method be based on the oxidation-reduction
method of analysis to measure organic carbon.
" 2. Alternate Methods. These are methods not necessarily
demonstrated to be equivalent to the reference method,
but demonstrated to the satisfaction of the control
agency to produce results adequate for determining com-
pliance, in specific applications. Methods involving
direct measurement with flame ionization detectors would
be primary candidates for alternate methods.
3. Screening Methods. These alternative methods may produce
biased or imprecise results, but they have been demonstrated
to the satisfaction of the control agency to be
| adequate for determining compliance, provided that
I
I
any bias or imprecision is taken into account. These
-------
I
I
methods are normally characterized by portability of .
equipment, procedural simplicity, and low cost. Methods *
based on thermal conductivity or low-cost portable FID
analyzers would be candidates for approval as screening methods.
Recommended Reference Method p
A reference method that involves indirect measurement of volatile
organic carbon by an oxidation-reduction is recommended. This neces-
sitates that the emission limits be expressed in terms of organic
carbon. If the emission limits and a universal reference method are
both based on organic carbon, the volatile organic standard will be I
expressed in clear, unambiguous terms. No other known practical test
method could accomplish this objective for a wide range of volatile
organic compounds.
A draft "Reference Method for Determination of Total Gaseous
Nonmethane Organic Emissions as Carbon" is included as Attachment 1.
The method requires a system for separating total nonmethane organics
from other carbon compounds, converting the total nonmethane organics
to methane, and analysis of the methane by a flame ionization detector.
Other than requiring this general equipment, the method provides per-
formance specifications designed to assure correct performance of
the separation-detection system. Depending on the organic carbon
concentration, the method may allow the use of the NDIR to detect I
the CO,, formed by the initial oxidation step.
The concept upon which the method is based has been utilized
for many years in Los Angeles County, where it has been demon- I
strated to be valid and effective for compliance determinations.
I
-------
I
I
7
There is at least one private laboratory that is able to perform
' the analysis on a fee basis. The oxidation-reduction analysis
concept for measurement of gaseous nonmethane organic compounds
is also offered commercially by at least one instrument vendor.
I Although the availability of instruments is admittedly very
limited, there is no serious technical impediment that would pre-
vent additional vendors from designing and producing acceptable
instruments.
Because of the somewhat limited use potential for the Los
Angeles laboratory-oriented procedure and the present limited
production of commerical instruments for field use, the OAQPS
I recommendation of the oxidation-reduction reference method
for the definition of organic emissions is recognized as
leading the technology. Wide acceptance of the organic carbon
reference method will, however, procide the needed inducement
for additional vendors to enter the market and thereby increase
| the supply and variety of organic carbon analysis instruments.
A draft of the Los Angeles procedure is included as Attach-
ment 2. Although the detail is considerable, some agencies may
I wish to assemble the laboratory apparatus. The Emission Measure-
ment Branch of OAQPS is working to refine the operating details
| of the procedure; as information becomes available, it will be
incorporated into the method.
I
I
I
-------
Alternate Methods
I
I
An "Alternate method" may have only limited applicability I
or other deficiencies that prevent its designation as a reference
method. Such methods may offer practical advantages and produce
results adequate for determining compliance in certain applications. I
Where such methods are applicable, the results may be accepted in
lieu of reference method results. I
As a specific example, methods based on direct measurement with M
flame ionization detectors are often practical for hydrocarbon
compounds, and, in addition, the equipment is widely available. Ij
An FID analyzer will be less costly and may be less complicated
for field application that an oxidation-reduction analyzer; therefore, |
for those applications where such methods can be made to produce .
accurate results, approval of them as alternate methods is "
desirable.
The Office of Air Quality Planning and Standards, EPA,has drafted
an "Alternate Test Method for Direct Measurement of Total Gaseous Organic |
Compounds Using a Flame Ionization Detector," which is included as
Attachment 3. This method outlines the known characteristics and limi-
tations of FID techniques and provides procedures needed to assure its
proper operation. The method does not and is not intended to
indicate specific applications where the method can (or cannot) be
used or correction factors to be applied to the results. Such deter-
minations must be made on a case-by-case basis founded on knowledge of
I
I
-------
I
I
I
9
the contents of the stream under test and the limitations of the
detector.
Screening Methods
In addition to the limitations associated with alternate
| methods, screening methods may also lack precision. In spite
_ of such shortcomings, screening methods may play an
important role in any volatile organic control program. As a practical
matter, the cost of applying a reference or alternate method to all
or even a majority of the regulated effluent streams in a jurisdic-
tion may be unreasonable; therefore, less expensive, simpler testing
techniques will be needed.
To date, OAQPS has made effective use of an explosimeter to detect
vapor leaks in gasoline marketing operations. In addition, an inexpensive
hydrocarbon monitor using a solid-state ionization detector was designed
I by OAQPS and has been used successfully as an emissions breakthrough
detector on the exit of a carbon adsorber. More recently, OAQPS has
initiated a test program associated with the development of new source
standards, using portable analyzers to detect leaks occurring in unit
operations in the petroleum industry. Two analyzers will be employed
in this program, one involving a combustion/thermal-conductivity-type
detector and the other a low-cost FID.
| Another example of a screening method would be the case of an FID
« analyzer applied to an unknown gas stream. In such case there is often
enough information available to provide a rough estimate of the analyzer
I accuracy, but a more exact determination would be prohibitive. In such an
I
-------
I
I
event the FID may be used as an alternative method for determining H
compliance, provided that sufficient buffer is included to account
for the possible inaccuracy. I
Regulatory Language
Examples of how general regulations may be expressed in terms of |
the reference method that measures organic carbon concentration are _
as follows:
1. To regulate concentration:
"Emissions of organic carbon shall not exceed
grams carbon per cubic meter." |
2. To regulate mass rate: _
"Emissions of organic carbon shall not exceed *
grams carbon per hour" or "Emissions of organic car-
bon shall not exceed grams carbon per kilogram
of solvent used." I
To protect the analytical instrument from contamination from
particulates and condensation, a filter and heated sample line (tern-
perature defined) must also be included in the emission regulations.
I
I
I
I
I
-------
n 8/29/77
ATTACHMENT 1. REFERENCE METHOD FOR DETERMINATION OF TOTAL GASEOUS
NONMETHANE ORGANIC EMISSIONS AS CARBONAUTOMATED ANALYZER VERSION
1. Pri nci pie and Appl i cab i 1 i ty
1.1 Principle. Conditioned stack gas is transported to and
analyzed by a semiportable gas chromatograph (GC) equipped with a flame
I ionization detector (FID). The total gaseous nonmethane organic (TGNMO)
I
fraction is separated by means of various GC columns from the other
constituents, oxidized to C02 and then reduced to methane (CH.) before
I it is introduced to the FID. In this manner, the variable response of
the FID associated with different types of organics is eliminated, and a
| count of TGNMO carbon atoms is obtained.
M 1.2 Applicability. The method is applicable to the semicontinuous
measurement of total gaseous nonmethane organics in source emissions.
V 2. Range and Sensitivity
2.1 Range. Signal attenuators shall be available so that a
I minimum signal response of 10 percent of full scale can be produced
when analyzing calibration gas or sample.
2.2 Sensitivity. The detector sensitivity shall be equal to or
better than 2.0 percent of the full scale setting, with a minimum full
scale setting of 10 ppm (methane or carbon equivalent).
| 3. Interferences
None.
4. Apparatus
4.1 TGNMO analyzers are available commercially or can be constructed
from available components by a qualified instrument laboratory. The
I
I
-------
I
12
primary components of the analyzer are an FID preceded by a
GC column to achieve the necessary separation of TGNMO from other I
carbon compounds. Oxidation and reduction catalysts then convert
the TGNMO to CH. prior to detection. The analyzer shall be |
accompanied by an instruction manual (supplied by the manufacturer _
if the analyzer was commercially produced) describing proper operation
and maintenance procedures. In addition to the specific procedures
required by this method, the analyzer shall be demonstrated prior to
initial use to be capable of proper separation, oxidation, and |
reduction. As a minimum this demonstration shall include measurement _
of a known TGNMO concentration present in a mixture that also contains
similar amounts of CH., CO^s and CO. Certification of such demon-
stration by the manufacturer is acceptable.
4.2 Sample Conditioning or Interface System (see Figure 1).
Probe with filter, 6.4 mm O.D. Teflon sample line, Teflon-coated
I
I
diaphragm pump, and Teflon flow control valves. A heating system
capable of maintaining all components at 120°C or greater shall be
included. The pump shall be sized so that the sample residence time
from the probe to the instrument will not exceed 15 seconds.
4.3 Potentiometric Recorder (optional). Strip chart recorder
with a voltage output compatible with the analyzer.
5. Reagents
5.1 Combustion Gas. Air containing less than 2 ppm organics
(methane or carbon equivalent). I
Mention of trade names on specific products does not constitute I
endorsement by the Environmental Protection Agency. *
I
-------
1
HEATED 13
1 PROBE (~ 100 °C) HEATED
. TEFLON
FILTER x / / SAMPLE LINE
[GI^S WOOL) JX / (-100 °C)
V-.J±L^L__ /
cc4;v -----_-_' -~hr= -/
1 xl
STACK ' ///
WALL ^X'
IHEAT
TEFLON-C
DIAPHRAGM
(LEAKLE
1
1
HEATED TGNMO
ANALYZER v
(~100°C) \
1 \
1
1
1
' 'N\
l^oYi
ED ^^^ A^-V /
DATED ^^ dL/~\\, /
PUMP ' s~^
*t-. ?
^_I3
r Jr^
CHROMATOGRAPHIC
SEPARATION
TGNMO
Optional j-
(CO,C02,CH^ OXIDATION
i CATALYST
r -L_ , T
REDUCTION ] 1
CATALYST I C09
^nr r
V REDUCTION
CH CATALYST
4
\ ^"4
FLAME 10N1ZATION
DL'TECTOR
t Optional
RESULTS - ^0,C09,CH/)TGf?MO
HEATED
FLOW CONT
/ VALVE
/ (~-100°C
' EX
/ B
'^
~=^=-^y=3
-* SPAN GASES
^~ ZEKO GAS
"* CARRIER GAS
-«- HYDROGEN
-<- AIR
FIGURE 1. On-site Application of TGNMO Analyzer
1
EXCESS SAMP
BLEED VALV
-------
I
I
5.2 Fuel. Hydrogen or a mixture of hydrogen and inert gas
containing less than 1 ppm organics (methane or carbon equivalent).
5.3 Carrier Gas. Helium, nitrogen, air or hydrogen containing
less than 1 ppm organics (methane or carbon equivalent).
5.4 Zero Gas. Air containing less than 1 ppm organics (methane
or carbon equivalent).
5.5 Calibration Gases (2). Gas mixture standards with known
propane (C,HQ) concentrations corresponding to ranges of 5-10 ppm and
5-10 percent (methane or carbon equivalent) are prepared and certified
by a gas manufacturer. The mixture shall consist of CgHg, CO, C02,
and CH, in nitrogen. The gas manufacturer must recommend a maximum
shelf life for each cylinder so that the C-^Hg concentration does not
change more than +^5 percent from its certified value. The date of
gas cylinder preparation, certified C3Hg, CO, C02, and CH^ concentratiop
and recommended maximum shelf life must be affixed to the cylinder
before shipment from the gas manufacturer to the buyer. These gas
mixture standards are to be used to prepare a chromatograph calibration*
curve as described in Section 7.2.
5.6 Span Gas. The calibration gas corresponding to 5 to 10 percenB
(methane or carbon equivalent) is used to span the analyzer.
6. Analyzer Performance Specifications 8
6.1 Linearity: +_ 5 percent of the expected value for full scale
settings up to the maximum percent absolute (methane or carbon
equivalent) calibration point. The analyzer shall be demonstrated prioB
I
I
-------
I
15
to initial use to meet this specification through a 5-point (minimum)
I calibration. There shall be at least one calibration point in each
of the following ranges: 5-10, 50-100, 500-1,000, 5,000-10,000, and
50,000-100,000 ppm (methane or carbon equivalent). Certification
of such demonstration by the manufacturer is acceptable. An additional
linearity performance check (see Section 7.2.1) must be made before
each use.
6.2 Zero Drift. One percent full scale per test period.
I 6.3 Span Drift. One percent full scale per test period.
7. Procedure
7.1 Sampling
7.1.1 Assemble the system as shown in Figure 1. Locate the
analyzer in a suitable environment. Take particular care that sample
8 will be introduced to the system under the same conditions of pressure
and flow rates as are used in calibration. For specific operating
instructions for the TGNMO analyzer, refer to the operation manual.
I 7.1.2 Adjust the sampling system and analyzer heating system to
provide a minimum temperature of 120°C and allow the system to warm up.
8 7.1.3 Perform a leak check as follows before sampling: Recheck
« to confirm that all fittings are tight. With the sample probe plugged,
open the flow control valve and the excess sample bleed valve. Use
I leak detection fluid or immerse the tubing leading from the bleed valve
in a jar of water to check that sample flow has ceased. At the con-
I elusion of the sampling tests, recheck for leaks.
I
I
-------
I
I
16
7.1.4 Begin Actual Sampling. Set the signal attenuation
to yield a minimum response of 10 percent of full scale, unless the
stack concentration is less than 1 ppm. Adjust the flow and bleed
valves to minimize sample line residence time. Perform the analysis
a minimum of four times. Report the average of the final four
readings. The analyzer cycle time is normally 10 to 15 minutes.
7.1.5 At the conclusion of the sampling tests, but at least once
every day, introduce zero and span gas to the analyzer to determine
zero and span drifts. If the analyzer has drifted beyond the allowable*
performance specification, the tests shall be considered invalid.
7.2 Calibration I
7.2.1 Calibration Curve. Maintain a record of performance of
each item. Determine the linearity of the analyzer for TGNMO as
follows: With the signal attenuation at the most sensitive setting, I
introduce zero gas and adjust the respectivezeroing controls to
indicate a reading of less than 1 percent of full scale. With the |
signal attenuation at the least sensitive setting, introduce the span
gas and adjust the span control to indicate the proper value on the
analyzer readout. Repeat these two steps until adjustments are no
longer necessary. Calculate a predicted response for the 5-10 ppm
calibration gas. Introduce that calibration gas and note the value |
obtained. If this value is not within +5 percent of its predicted _
value, then the analyzer may need repairs, or one or both of the
calibration gases may need replacement. In any event, this linearity
performance specification shall be met before the analyzer is placed
in actual use.
I
-------
I
I
7.3 Catalyst Performance Check. These checks should be
performed on a frequency established by the amount of use of the
analyzer, and the nature of the organic emissions to which it is
I exposed. To confirm that the oxidation catalyst is functioning in
the correct manner, the operator must turn off or bypass the reduction
catalyst while operating the analyzer in an otherwise normal fashion.
If oxidation is adequate, the only gas that will then reach the
detector will be C09, to which the FID has no response. If responses
I2
are noted, then the oxidation catalyst must be replaced. To confirm
I
the operation of the reduction catalyst, reverse the above procedure.
If CC" in the calibration gases is not reduced to CH^ as it should be,
I then the reduction catalyst must be replaced.
8. Calculations
| 8.1 Determine concentrations of TGNMO (propane equivalent)
_ directly from the calibration curves. Multiply this number by 3 to
obtain ppm TGNMO (methane or carbon equivalent).
8.2 Conversion to mass concentration values for TGNMO as
carbon is made as follows:
13
mg TGNMO/m as carbon = ppm TGNMO (methane or carbon equivalent)
x 0.499
where:
II ppm TGNMO (methane or carbon equivalent) = ~- x 41'57 9"mo1e
10b ill
I
I
0.499 mg/m3 as carbon.
-------
18
I
where:
Molecular weight of carbon =12
Standard conditions: 20°C, 1 atm.
9. References
9.1 Albert E. Salo, Samuel Whitz, and Robert D. MacPhee.
"Determination of Solvent Vapor Concentrations by Total Combustion
Analysis: A Comparison of Infrared With Flame lonization Detectors." I
Presented at the 68th Annual Meeting of the Air Pollution Control
Association, Boston, Ma. Paper No. 75-33.2. June 15-20, 1975. |
9.2 Instruction Manual, Byron Model 401 Total Emission Analyzer, «
Byron Instruments, Inc., 520 1/2 S. Harrington Street, Raleigh, N.C. 2760.
I
I
I
I
I
I
I
I
I
I
-------
I
19
I ATTACHMENT 2. DETERMINATION OF TOTAL GASEOUS NONMETHANE
ORGANIC EMISSIONS AS CARBON: MANUAL SAMPLING AND
ANALYSIS PROCEDURE
1. Principle and Applicability
1.1 Principle. An emission sample is anisokinetically
drawn from the stack through a heated filter and a chilled
condensate trap by means of an evacuated gas collection tank.
Total gaseous non-methane organics (TGNMO) are determined by
combining the analytical results obtained from independent
analyses of the condensate trap and evacuated tank fractions.
After sampling is completed, the organic contents of the
I condensate trap are oxidized to carbon dioxide which is
quantitatively collected in an evacuated vessel; a portion
I of the carbon dioxide is reduced to methane and measured by
a flame ionization detector (FID). A portion of the sample
collected in the gas sampling tank is injected into a gas
chromatographic (GC) column to achieve separation of the
nonmethane organics from carbon monoxide, carbon dioxide
I and methane; the nonmethane organics are oxidized to carbon
M dioxide, reduced to methane, and measured by a FID.
1.2 Applicability. This method is applicable to the
I measurement of total gaseous nonmethane organics in source
emissions.
| 2. Apparatus
_ 2.1 General. TGNMO sampling equipment can be constructed
* by a laboratory from commercially available components and
I
I
-------
20
I
I
components fabricated in a machine shop. The primary com-
ponents of the sampling system are a heated filter, condensate I
trap, flow control system, and gas sampling tank. (Figure 1).
The primary components of the analytical system are an |
oxidation system for recovery of the sample from the condensate _
trap and a TGNMO analyzer. The TGNMO analyzer is a FID
preceded by an oxidation catalyst, a reduction catalyst, and
a GC column with backflush capability (Figure 2). The system
for the removal and conditioning of the organics captured in
the condensate trap consists of a heat source, oxidation
catalyst, Non-Dispersive Infrared (NDIR) analyzer and an inter-
mediate gas collection tank (Figure 3).
2.2 Sampling.
2.2.1 Probe. 1/8" stainless steel tubing heated to I
approximately 120°C.
2.2.2 Filter Holder. Stainless steel with a stainless I
steel or glass frit filter support and a Teflon gasket. The
holder design shall provide a positive seal against leakage
from the outside or around the filter. The holder shall be I
attached at the outlet of the probe.
2.2.3 Filter Heating System. Any heating system capable I
of maintaining a temperature around the filter holder during
sampling of 120 + 14°C (248 + 25° F), or such other temperature
as specified by an applicable subpart of the standards or I
approved by the Administrator for a particular application.
A temperature gauge capable of measuring temperature to within
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
21
<
ts
<
V),
Ol
_
u _i
-------
22
SAMPLE
o
3
cj
<
a.
<
K
K
111
C/l
Z
LU
a
z
o
o
cc
o
u_
K
<
«V
>-
CO
z
S
_j
o
o
<3
CARRIER GAS
SEPARATION
COLUMN
CONTROL VALVE ( ^
CO
I
CH4
C02
CONTROL VALVE
OXIDIZING
CATALYST
i
WATER
TRAP
BACKFLUSH
CARRIER
GAS
TGNMO
BACKFLUSH
CONTROL VALVE
REDUCTION
CATALYST
FID
1 f
COMBUSTION AIR
FUEL
Figure 2. TOTAL GASEOUSNONMETHANE ORGANIC (TGNMO) ANALYZER SCHEMATIC
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
23
DC
O
<0
O t-
<
ec
<
C9
Z
O
Z
O
o
O
<
ec
ui
>
o
o
Ul
ec
a
Z
o
o
O)
CO
o
u.
*
u
u.
o
-------
24
3° C (5.4° F) shall be installed so that the temperature
I
I
I
around the filter holder can be regulated and monitored
during sampling.
2.2.4 Condensate Trap. The condensate trap shall be
constructed of 316 stainless steel; construction details of a
suitable trap are shown in Figure 4.
2.2.5 Flow Control System.
2.2.5.1 Needle Valve. To regulate sample gas flow rate.
2.2.5.2 Rate Meter. Rotameter, or equivalent capable
of measuring flow rate to within +_10 percent of the I
selected flow rate of about 80 cc/min. Other flow control systems
capable of maintaining a constant sample rate of 80 cc/min +_
10 percent may be used subject to the approval of the Administrator.
2.2.6 Gas Collection Tank. Stainless steel or aluminum tank
with a minimum volume of 6 liters. The tank is fitted with a I
vacuum gauge, a leak!ess valve, and a t-connector for conducting
leak checks. I
2.3 Analysis. For analysis, the following equipment is needed.
2.3.1 Condensate Recovery and Conditioning Apparatus
(Figure 3). I
2.3.1.1 Heat Source. A heat source sufficient to heat
the condansate trap to a "cherry red" color. An electric j|
muffle-type furnace or bunsen burner may be used. _
2.3.1.2 Oxidizing Catalyst. A platinum and quartz
catalyst constructed from a 44-inch length of 1/4" tubing of I
I
-------
I
I
I
I
I
I
I
I
I
I
25
CONNECTOR
EXIT TUBE. 6mm (% in) 0.0.
X 0.71mm (0.028 in) WALL
6mm EXIT TUBE CONTAINS 3mm (1/8 in) SS WOOL PLUGS
OIMITHER SIDE OF A 6mm ('/. in) QUARTZ WOOL FILTER
I
I
I
I
I
I
I
I
N0.40HOLE
WELDED JOINTS
INLET TUBE, 6mm (% in) O.D.
X 0.71 mm (0.028 in) WALL
CONNECTOR
CRIMPED AND WELDED GAS-TIGHT SEAL
BARREL 19mm (% in) O.D. X 140mm (5-'/i in) LONG
1.5mm (1/16 in) WALL
BARREL PACKING. SS WOOL PACKED TIGHTLY
AT BOTTOM, LOOSELY AT TOP
HEAT SINK (NUT, PRESS-FIT TO BARREL)
WELDED PLUG
MATERIAL: TYPE 316 STAINLESS STEEL
Figure 4 CONDENSATETRAPZ
-------
I
26
I
70 percent Ni-30 percent Cr alloy packed as follows:
First 4 inchesempty. I
Next 4 inches--8-10 mesh alumina coated with 0.5 percent
finely divided platinum.
Next 28 inches8 mesh quartz chips. I
Last 8 inches8-10 mesh alumina 0.5 percent platinum coated.
Other catalyst systems capable of meeting the catalyst |
efficiency criteria of this method (Section 4.4.2) may be used _
subject to the approval of the Administrator.
2.3.1.3 Water Trap. Any leak proof moisture trap capable
of removing moisture from the gas stream may be used. A
condensate trap designed according to the specifications of |
Figure 4 without packing in the exit tube will suffice. _
2.3.1.4 NDIR Detector. Detector capable of indicating the
COp level in the zero to five percent range; required to monitor
the combustion progress of the organic matter in the condensate
trap. I
2.3.1.5 Pressure Regulator. Stainless steel needle
valve required to maintain the NDIR detector at a constant I
pressure.
2.3.1.6 Intermediate Collection Tank. Stainless steel
or aluminum collection vessel. Tanks with nominal volumes of I
2 and 6 liters are recommended. The end of the tank is fitted
with a t-connector, vacuum gauge, and leakless valve. I
2.3.1.7 Calibration Injection Port. Injection port valve
and sample loop for injection of calibration standards required
I
I
-------
I
I
27
to check the combustion efficiency of the condensate recovery
g system.
2.3.2 Total Gaseous Nonmethane Organic (TGNMO) Analyzer.
Semicontinuous GC/FID analyzer capable of: (1) separating CO,
COp, CH., and gaseous nonmethane organics, (2) oxidizing the
nonmethane organic fraction to CO,,, reducing the C02 to methane,
and quantifying the methane. The analyzer shall be demonstrated
prior to initial use to be capable of proper separation, oxidation,
reduction, and measurement. As a minimum, this demonstration
shall include measurement of a known TGNMO concentration present
in a mixture that also contains CH. , CO, and COp. (see paragraph
I 4.4.1) In addition, the analyzer shall meet the following per-
formance specifications:
I 2.3.2.1 Linearity. j^5 percent of the expected value for
each full scale setting up to the maximum percent absolute
(methane or carbon equivalent) calibration point. The analyzer
I shall be demonstrated prior to initial use to meet this specifica-
tion through a 5-point (minimum) calibration. There shall be at
| least one calibration point in each of the following ranges:
5-10, 50-100, 500-1,000, 5,000-10,000, and 40,000-100,000 ppm
(methane or carbon equivalent). Certification of such demonstra-
I tion by the manufacturer is acceptable. An additional linearity
performance check (see Section 4.4.1.1) must be made before each
| use.
« 2.3.2.2 Zero Drift. One percent full scale per analysis of
an emission test series.
I
I
-------
28
2.3.2.3 Span Drift. One percent full scale per analysis of
1
Mention of trade names or specific products does not
I
I
an emission test series. I
2.3.2.4 The following components have been found to be
acceptable for use in the TGNMO System: I
2.3.2.4.1 Oxidation Catalyst. Type 316 stainless steel
0.25 inch OD tubing x 14 inches long packed with Hopcalite 25 -
30 mesh; operated at 850° C.
2.3.2.4.2 Reduction Catalyst. Type 316 stainless steel
0.25 inch OD tubing x 7 inches long packed with 10 percent nickel |
on chromasorb W, 60-80 mesh; operated at 400° C. Method of pre- .
paration: 100 grams chromasorb W, 10 grams nickelous nitrate,
75 ml water, evaporated to dryness then heated in air for 4
hours to convert to nickel oxide. After packing the tubing,
reduce overnight at 450° C and 30 ml/min H2 to nickel metal. |
2.3.2.5.3 Separation Column. Type 316 Stainless steel
0.125 inch OD tubing x 18 feet long packed with Porapak Q 60/80
mesh; operated isothermally at 80° F.
2.3.3 Mercury Manometer. U-tube mercury manometer
capable of measuring pressure to within 1.0 mm Hg in the 0 - g
900 mm range. _
2.3.4 Barometer. Mercury, aneroid, or other barometer
capable of measuring atmospheric pressure to within 2.5 mm
(0.1 inch Hg).
3. Reagents J
3.1 Sampling.
I
constitute endorsement by the EPA. |
I
-------
29
I 3.1.1 Filter. Glass fiber filter without organic binder,
exhibiting at least 99.95 percent efficiency (<_ 0.05 percent
penetration) on 0.3 micron dioctyl phthalate smoke particles.
I The filter efficiency test shall be conducted in accordance with
ASTM standard method D 2986-71. Test data from suppliers
I quality control program are sufficient for this purpose.
_ 3.1.2 Crushed Dry Ice.
" 3.2 Analysis.
3.2.1 (TGNMO) Analyzer.
3.2.1.1 Carrier Gas. 5 percent Op in N« containing less
I than 1 ppm organics.
3.2.1.2 Fuel Gas. 40 percent hydrogen in nitrogen con-
taining less than 1 ppm organics.
3.2.2 Condensate Recovery and Conditioning Apparatus.
3.2.2.1 Carrier Gas. 5 percent 02 in N2 containing less
than 1 ppm organics.
3.2.2.2 Oxygen. Oxygen containing less than 1 ppm organics,
3.3 Calibration.
3.3.1 (TGNMO) Analyzer.
3.3.1.1 Calibration Gases (3). Gas mixture standards with
known propane (C,H_) concentrations corresponding to ranges of
5-10 ppm, 50-10 percent and 20-25 percent methane or carbon
equivalent are prepared and certified by a gas manufacturer.
The Mixture shall consist of C_Hfl, CO, C0«, and CH. in nitrogen.
The gas manufacturer must recommend a maximum shelf life for
each cylinder so that the C-HR concentration does not change
I
I
-------
30
more than +_ 5 percent from its certified value. The date of I
gas cylinder preparation, certified C,Hg, CO, CO-, and CH. concen- M
trations and recommended maximum shelf life must be affixed to
the cylinder before shipment from the gas manufacturer to the I
buyer. These gas mixture standards are to be used to prepare a
chromatograph calibration curve as described in Section 4.4.1.1. |
3.3.1.2 Span Gas. The calibration gas (Section 3.3.1.1) _
corresponding to 20 to 25 percent is used to span the analyzer.
3.3.1.3 Oxidation Catalyst Check. The calibration gas
(Section 3.3.1.1) corresponding to 20 to 25 percent is used to
check the oxidation catalyst. g
3.3.1.4 Reduction Catalyst Check. A gas standard with a
known concentration of 5 percent (nominal) C0« in nitrogen is *
used to check the reduction catalyst.
3.3.2 Condensate Recovery and Conditioning Apparatus. Gas
mixture standards (2) with known propane (C,Hg) concentrations
in nitrogen corresponding to ranges of 5-10 ppm and 5-10 percent
(methane or carbon equivalent) are prepared and certified by a
gas manufacturer. These gas mixture standards are to be used to
check the operation of the condensate trap oxidation system as
described in Section 4.4.2.
4. Procedure
4.1 Sampling I
4.1.1 Pretest Preparation. The sample tank shall be cali-
brated according to the procedure described in paragraph 4.4.3.
Check filters visually against the light for irregularities,
I
I
-------
I
I
flaws, or pinhole leaks. Either in the laboratory or in the field
| evacuate the sample tank to a vacuum of 755 mm mercury (measured
by a mercury U-tube manometer). Record the temperature, bara-
metric pressure, tank vacuum measured with the manometer, and
the vacuum indicated on the tank gauge.
4.1.2 Assemble the system as shown in Figure 1. Immerse
the condensate trap in dry ice and start the filter and probe
heaters.
4.1.3 Leak check procedures.
4.1.3.1 Gas Sampling Tank Leak Check. Leak check the gas
sampling tank immediately after the tank is evacuated. Once
the tank is evacuated, allow the tank to sit for 30 minutes. The
tank is acceptable if no change in tank vacuum (measured by the
I mercury manometer) is noted.
4.1.3.2 Pretest Leak Check. A pretest leak check is
recommended, but not required. If the tester opts to conduct
I the pretest leak check, the following procedure is used. After
the sampling train has been assembled (including cooling of
I condensate trap and heating of filter) plug the probe tip.
H Attach the vacuum line of the leak check apparatus (Figure 5)
to the T-connector of the evacuated tank; open the valve on this
V connector (not the sample flow control valve to the evacuated
tank) and evacuate the sample train to a vacuum of 625 mm Hg.
| Shut the valve on the pump side of the manometer and allow the
H sampling train to sit for 10 minutes. A leak rate in excess of
0.5 mm Hg for this 10 minute period is unacceptable. When the
I
I
-------
32
CONTROL
VALVE
CONTROL
VALVE
BYPASS
VALVE
VACUUM
LINE
MERCURY
MANOMETER
VACUUM
PUMP
LEAK CHECK
Figure 5. APPARATUS
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
33
leak check is completed, slowly release the vacuum 1n the train
by unpluging the probe, close the T-connector valve, and plug
this connector to assure a leak free system.
| 4.1.3.3 Post Test Leak Check. A leak check is mandatory
M at the conclusion of each test run. After sampling is completed,
plug the end of the sampling probe and attach the vacuum line of
B the leak-check apparatus (Figure 5) to the evacuated tank
t-connector. Assure that the flow valve to the evacuation pump
| (valve between manometer and pump) is closed. First open the
_ t-connector valve to the manometer and then open the flow control
valve to the evacuated tank. Record the clock time and tank
vacuum. After 10 minutes note the tank vacuum. A leak rate in
excess of 0.5 mm Hg for this ten minute period is unacceptable
I and the sampling run shall be voided. After completing the leak
check, close the evacuated tank flow control valve and the
t-connector flow control valve. Disconnect the leak check
apparatus and plug the t-connector to assure a leakproof seal
during shipping. Unplug the probe tip.
j| . 4.1.4 Sample Train Operation. Place the probe into the
stack such that the probe tip is located at a pre-selected location.
For stacks under negative pressure, assure that the sample port
H is sufficiently sealed to prevent leakage of ambient air around
the probe. Record the clock time, sample tank gauge vacuum, and
barometric pressure. Assure that the flow control needle valve
is closed. Begin sampling by opening the evacuated tank flow
I valve all the way. Open the flow control needle valve until
I
I
-------
34
I
I
the rotameter indicates the desired setting; maintain a constant
flow rate (+ 10 percent) throughout the duration of the sampling
period. Record the gauge vacuum, rotameter setting, and filter
temperature at 5 minute intervals. Select a total sample time |
greater than or equal to the minimum sampling time specified in m
the applicable subpart of the standard; end the sampling when
this time period is reached or when a constant flow rate can fl
no longer be maintained. When the sampling is completed, close
the evacuated tank valve and remove the probe from the stack. |
Record the final readings. Conduct the post test leak check _
according to the procedures of paragraph 4.1.3.3.
If the sampling must be stopped before obtaining the minimum
sampling time specified in the applicable subpart because a
constant flow rate cannot be maintained, proceed as follows: I
After removing the probe from the stack, conduct the post
test leak check. After the leak check is completed, remove the
evacuated tank from the sampling train (without disconnecting
other portions of the sampling train) and connect another
evacuated tank to the sampling train. Proceed with the sampling;
after the minimum total sampling time is exceeded, end the test.
\.2 Sample Recovery.
Disconnect the condensate trap at the filter and at the flow M
metering system. Tightly seal the ends of the condensate trap;
keep the trap packed in dry ice until analysis is conducted. Seal
the connection at the evacuated tank to assure a leak proof seal
during shipping. After the evacuated tank has cooled to ambient I
I
I
-------
I
35
conditions, attach the U-tube manometer to the t-connector,
fl open the valve, and record the tank vacuum, ambient temperature,
barometric pressure, and indicated gauge vacuum. Close the flow
| valve and reseal the t-connector to assure a leak proof seal
_ during shipping. Assure that the test run number is properly
* indentified on the condensate trap and evacuated tank(s).
4.3 Analysis
4.3.1 TGNMO Analyzer. Heat the catalysts to their operating
| temperatures and set the carrier gas and fuel flow rates. Conduct
_ the calibration check required in paragraph 4.4.1.1 and the
catalyst performance checks required in paragraph 4.4.1.2 prior
to analyzing the test samples.
4.3.2 Condensate Trap. Return the condensate trap to the
Jj laboratory and hook it into the recovery and conditioning system
_ (Figure 3). Set the oven for the oxidizing catalyst at 850° C
and the trap heating furnace at 600° C. Set the gas directing
valve to permit flow of 5 percent 02/N2 through channel A to the
condensate trap at a rate of 80 cc/min; at the same time set the
I oxygen flow through channel B at 20 cc/min (1:4 ratio). After
two minutes, switch the gas directing valve to permit the oxygen
to flow via channel A directly through the condensate trap and
the 5 percent CL/N2 carrier gas to flow through channel B. When
the NDIR indicates that COp is no longer being emitted from the
combustion system, shut off the collection flask from the system
and cease combustion. Record the collection flask pressure after
combustion is completed (P.) and then pressurize the flask to
860 mm Hg (nominal) with nitrogen and record the final pressure
I
-------
36
record the tank vacuum (Pt). Pressurize the tank with
nitrogen and record the final tank pressure (P. }, temperature
I
I
(Pf). Remove a syringe sample from the flask and inject this
into the TGNMO analyzer. Record the analyzer response (ppm C) |
for triplicate samples.
4.3.3 Gas Sampling Tank. Using a U-tube mercury manometer,
I
l<(W«V^J«*ll %A I I w I * W 1 * ** I I W * I I IM t VV( I I It h/ISaWMfWII^ II I / J 1*^1111^^1 VI VMlb
f
and barometric pressure. Remove a syringe sample from the tank
and inject this into the TGNMO analyzer. Record the analyzer
response (ppm C) for the non-methane organic fraction for triplicate
samples.
4.4 Calibration. Maintain a record of performance of
each item. I
4.4.1 TGNMO Analyzer.
4.4.1.1 Calibration Curve. Determine the linearity of
the analyzer for TGNMO as follows: With the signal attenuation
at the most sensitive setting, introduce zero gas and adjust
the respective zeroing controls to indicate a reading of less I
than 1 percent of full scale. With the signal attenuation at
the least sensitive setting, introduce the span gas and adjust
the span control to indicate the proper value on the analyzer m
readout. Repeat these two steps until adjustments are no longer
necessary. Calculate a predicted response for the 5-10 ppm I
calibration gas. Introduce that calibration gas and note the
value obtained. If this value is not within j^5 percent of its |
predicted value, then the analyzer may need repairs, or one or
both of the calibration gases may need replacement. In any
I
I
-------
I
I
event, this linearity performance specification shall be met
| before the analyzer is placed in actual use.
_ 4.4.1.2 Catalyst Performance Check. These checks should be
performed on a frequency established by the amount of use of the
analyzer and the nature of the organic emissions to which it is
exposed. To confirm that the oxidation catalyst is functioning in
the correct manner, the operator must turn off or bypass the reduction
catalyst while operating the analyzer in an otherwise normal
fashion. Inject the calibration gas (paragraph 3.3.1.3) into the
system. If oxidation is adequate, the only gas that will then
reach the detector will be C02» to which the FID has no response.
If a response is noted, the oxidation catalyst must be replaced.
To confirm the proper operation of the reduction catalyst, inject
a sample of the C(L calibration gas (Section 3.3.1.4) into the
system. If the C(L is not reduced to CH. as it should be, then
the reduction catalyst must be replaced or regenerated.
4.4.2 Condensate Trap Oxidation Catalyst. Inject syringe
samples of the calibration gases listed in Section 3.3.2 into the
I sample port of the condensate trap combustion system (Figure 3).
m Proceed with a normal analysis (i.e., collection of the CCL in
the flask followed by analysis of triplicate aliquots using
I the TGNMO analyzer) and compare results to the actual concentration.
Repair the system if the results (average of triplicate aliquots)
| deviate by greater than +_ 5 percent from the calibration gas value.
« 4.4.3 Gas Sampling Tank. The volume of the gas sampling
tanks used must be determined. Prior to putting each tank in
I
I
-------
I
38
service, determine the tank volume by weighing the tanks empty |
and then filled with water; weigh to the nearest 0.5 gm and M
record the results.
4.4.4 Intermediate Collection Flask. The volume of the inter- M
mediate collection flasks used to collect CO- during the analysis
of the condensate traps must be determined. Prior to putting |
each flask in service, determine the volume by weighing the «
flasks empty and then filled with water; weigh to the nearest "
0.5 gm and record the results.
4.4.5 Condensate Trap Leak Check. Prior to each use, check
each condensate trap for leaks by pressurizing with N« to approximately J
50 psig and immersing in water.
4.4.6 Rotameter. The rotameter need not be calibrated but
should be cleaned and maintained according to the manufacturer's
instruction.
5. Calculations I
5.1 Sample Volume. For each test run calculate the gas
volume sampled:
/pt PII\ *
Vs . 0.36 Vf ^ - T- j I
\ * ri '
5.2 Noncondensible TGNMO. For each collection tank, determine I
the concentration of TGNMO (ppm C):
5.3 Condensible TGNMO. For each condensate trap determine the
concentration of TGNMO (ppm C):
I
I
-------
1
1
1
^B
1
1
1
1
1
1
|
I
1
1
1
1
1
1
39
3 X V- X P. n
r - " T v v C
Cc ' Vs X P. x kf1 Lcpk
5.4 Total Gaseous Nonmethane Organics (TGNMO). To
determine the TGNMO concentration for each test run, use the
following equation:
ct = c + cc
5.5 Control Device Efficiency. To determine the TGNMO control
device efficiency for each test run, use the following equation:
r - r
r _ Lti LtO v inn
t - - V A iV/U
Cti
where:
C « Noncondensible TGNMO calculated concentration, ppm
carbon equivalent.
C = TGNMO analyzer measured concentration for gas
collection tank, ppm propane.
C = Condensible TGNMO (condensate trap) calculated concentra
tion, ppm carbon equivalent.
C = TGNMO analyzer measured concentration for intermediate
collection flask, ppm propane.
C. = Total gaseous nonmethane organic (TGNMO), ppm carbon
equivalent.
C. = TGNMO at control device outlet, ppm carbon equivalent.
C. . = TGNMO at control device inlet, ppm carbon equivalent.
E = Control device efficiency, percent.
Pf = Final pressure of intermediate collection flask
(nominal 860 mm Hg.), mm Kg, absolute.
-------
I
40
I
P. = Pressure of intermediate collection flask at completion _
of combustion, mm Hg, absolute.
Pt = Gas sample tank pressure prior to sampling, mm Hg,
absolute.
P. = Gas sample tank pressure after sampling, but prior to Jj
pressurizing, mm Hg, absolute.
P. = Final gas sample tank pressure after pressurizing,
mm Hg, absolute.
T. = Gas sample tank temperature prior to sampling, °K.
T. = Gas sample tank temperature at completion of sampling,
I
T. = Gas sample tank temperature after pressurizing, °K.
f
3
V = Gas collection tank volume, M
-3
Vf = Intermediate collection tank volume, M
V = Gas volume sampled, dscm
m = Total number of injections of non-condensible TGNMO
during analysis (where j = injection number, 1 . . . m) I
n = Total number of injections of condensible TGNMO during
analysis (where k = injection number, 1 . . . n)
0.36 = 273°K/760 mm Hg I
Standard Conditions = Dry, 760 mm Hg, 273°K.
6. Bibliography
6.1 Albert E. Salo, Samuel Witz, and Robert D. MacPhee. "Deter-
mination of Solvent Vapor Concentrations by Total Combustion Analysis:
A comparison of Infrared with Flame lonization Detectors." Presented I
at the 68th Annual Meeting of the Air Pollution Control Association,
Boston, Ma. Paper No. 75-33.2. June 15-20, 1975. §
I
I
-------
I
41
I
_ 6.2 Albert E. Salo, William L. Oaks, Robert D. MacPhee.
"Measuring the Organic Carbon Content of Source Emissions for Air
Pollution Control." Presented at the 67th Annual Meeting of the Air
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Pollution Control Association, Denver, Colorado. Paper No. 74-190
June 9-13, 1974.
-------
June 6, 19||
42
ATTACHMENT 3. ALTERNATE TEST METHOD FOR DIRECT MEASUREMENT OF TOTAL |
GASEOUS ORGANIC COMPOUNDS USING A FLAME IONIZATION ANALYZER
I
INTRODUCTION
Performance of this method should not be |
attempted by persons unfamiliar with the
performance characteristics of the flame
ionization detector, nor by those who are I
unfamiliar with source sampling.
1. Principle and Applicability
1.1 Principle. The sample is drawn from the source, through
a heated sample line and glass fiber filter to a flame ionization I
analyzer (FIA). Ions formed in the combustion of a specific hydro-
carbon compound in a FL - 0^ flame establish a current that is
proportional to the mass flow rate of that hydrocarbon to the
flame. This current is collected at two polarized electrodes,
and is read out on a potentiometric recorder and compared with a I
calibration curve based on propane (C3Hg), or an organic solvent,
as appropriate. The results are reported as equivalents of I
methane (CH.) or carbon, or in terms of an organic solvent.
1.2 Applicability. This method is applicable for the
determination of the true carbon mass concentration, and/or an
indicated volume or mass concentration (expressed in terms of
carbon or of an assumed organic compound, e.g., methane equivalent)
of gaseous organic compounds present in an emission stream. It m
can also be used to measure the mass concentration of an organic
solvent if stable mass standards of the solvent can be generated. I
I
I
I
-------
I
I
_ The measurement will not exclude methane, so a supplemental
measurement of methane may be necessary.
2. Range and Sensitivity
2.1 Range. Signal attenuators shall be available so that
a minimum signal response of 10 percent of full scale can be pro-
duced when analyzing calibration gas or sample.
2.2 Sensitivity. The detector sensitivity shall be equal
to or better than 2.0 percent of the full scale setting, with a
minimum full scale setting of 10 ppm (methane or carbon equivalent)
I 3. Interferences
3.1 Nonorganic Gases. There is no response to nitrogen,
carbon monoxide, carbon dioxide, or water vapor, however, the
analyzer response to organics will be affected by the composition
of the background or carrier gas. It is, therefore, required
I that the calibration gases be contained in air, which is most
likely to be the same carrier gas as that of the actual sample.
2
Investigation of a reported oxygen synergism has shown
I
that a 40/60 mixed fuel (40 percent Hp» 60 percent He) is required
if the oxygen content of the emission stream varies more than a
I few percent from its mean value. Mixed fuel will also be re-
quired if the oxygen content of the emission stream varies more
I than a few percent from the oxygen content of the calibration gases.
3.2 Organic compounds. Acetylenic compounds give a slightly
higher response than aliphatic compounds. Carbon atoms bound to
oxygen, nitrogen, or halogens give a reduced or zero response.
I
I
-------
44
3
Table 1 illustrates these effects in terms of the relative
response of one FIA to various hydrocarbons. The response is
shown as effective carbon number (ECN), as follows:
Instrument response caused
ECN = by atom of given type
Instrument response caused
by aliphatic carbon atom
These values are true for one mode of operation of a specific
detector under specific conditions (e.g., mixed Np, H« fuel). It
has been reported that these numbers may vary widely for different
operating conditions and for different detectors. Variations of
as much as 25 percent have been observed in studies of the types
of organics associated with automotive emissions. The variation
was observed to decrease with decreasing sample flow rate, but
with an accompanying decrease in sensitivity.
TABLE 1. APPROXIMATE EFFECTIVE CARBON NUMBERS
(FROM BECKMAN INSTRUMENTS)
Type of Atom
Carbon
Carbon
Carbon
Carbon
Carbon
Carbon
Oxygen
Occurrence
In Aliphatic Compound
In Aromatic Compound
In Olefinic Compound
In Acetylenic Compound
In Carbonyl Radical
In Nitrite
In Ether
Effective
Carbon Number
+1.0
+1.0
+0.95
+1.30
0.0
+0.3
-KO
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
TABLE 1. APPROXIMATE EFFECTIVE CARBON NUMBERS
CKMAN INSTRl
(Continued)
I
I
(FROM BECKMAN INSTRUMENTS)
I
I
I
I
I
I
M From this information it can be seen that the accuracy of
this method for a given source will be largely dependent on the
particular makeup of organic emissions from the source.
3.3 Other effects. Significant changes in viscosity of the
| emission gas from that of the calibration gas will affect the mass
_ rate of organics to the detector. If this phenomena is expected
to occur, a corrective technique must be devised.
If the instrument is calibrated with organic solvent standards,
and then used to measure emissions of that solvent, their response
variations have been calibrated out.
4. Apparatus
Type of Atom
Oxygen
Oxygen
Oxygen
Chlorine
Chlorine
Nitrogen
Occurrence
In Primary Alcohol
In Secondary Alcohol
In Tertiary Alcohol, Ester
As two or more chlorine atoms
on single aliphatic carbon atom
In Olefinic Carbon Atom
In Amine
Effective
Carbon Number
-0.6
-0.75
-0.25
-0.12 each
+0.05
Value similar to
that for oxygen atom
in corresponding
alcohol
4.1 Commercially available heated FIA. The analyzer should
-------
I
I
be demonstrated, preferably by the manufacturer, or his repre-
sentative, to meet or exceed manufacturer's specifications and
those described in this method. The entire sampling and analysis
system as encountered by gaseous organics must be capable of being
maintained in the temperature range of 350 to 440°F, or less,
consistent with the emission regulation.
4.2 Sample conditioning or interface system. Probe with I
filter, Teflon* sample line, Teflon-coated diaphragm pump or
stainless steel bellows pump and Teflon flow control valves,
capable of being maintained in the temperature range of 350 to I
400°F, or less, consistent with the emission regulation.
4.3 Potentiometric Recorder (optional). Strip chart |
recorder with a voltage output compatible with the FIA. m
5. Reagents
5.1 Fuel. A hydrogen and helium mixture containing less I
than 2 ppm organics (methane or carbon equivalent).
5.2 Combustion Air. High purity air with less than 2 ppm |
organics (methane or carbon equivalent). Required only if the .
emission stream does not contain sufficient oxygen.
5.3 Zero Gas. Less than 0.1 ppm organics (methane or
carbon equivalent).
* Mention of trade names on specific products does not constitute
endorsement by the Environmental Protection Agency.
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
47
CL.
UJ
00
oo
O
QL
L ^
d:
V
O.
f
O
z:
O
OQ
O
2:
O
a.
UJ
CVJ
LU
a:
Z3
t!3
-------
48
I
I
I
5.4 Calibration Gases (2). Gas mixture standards with known
concentrations corresponding to ranges of 5 to 10 ppm and 5 to 10
percent (methane or carbon equivalent) are prepared and certified
by a gas manufacturer. The mixture will normally consist of I
C-Hg in air. Other organic(s) can be used, if appropriate. The
gas manufacturer must recommend a maximum shelf life for each I
cylinder so that the concentration does not change more than +_ 5
percent from the certified value. The date of gas cylinder pre-
paration, certified propane concentration and recommended maximum I
shelf life must be affixed to the cylinder before shipment from
the gas manufacturer to the buyer. These gas mixture standards |
are to be used to prepare a calibration curve as described in
Section 7.2. *
5.5 Span Gas. The calibration gas corresponding to 5 to I
10 percent (methane or carbon equivalent) is used to span the
analyzer. |
5.6 Organic Solvent. Either a sample obtained from the
solvent source, or a sample distilled from paint, ink, etc. in
accordance with ASTM Procedure D3272-73T. Required only if
unaltered solvent emissions are being measured, mass calculations
in terms of the solvent are necessary, and the relative response |
factor of the FIA to the solvent is unknown. .
6. System Performance Specifications
6.1 Linearity. +_ 5 percent of the expected value for full
scale settings up to the maximum percent absolute (methane or
I
I
I
-------
49
I
carbon equivalent) calibration point. The analyzer shall be
demonstrated prior to initial use to meet this specification
through a 5-point (minimum) calibration. There shall be at
least one calibration point in each of the following ranges:
5-10, 50-100, 500-1,000, 5,000-10,000, and 50,000-100,000 ppm
(methane or carbon equivalent). Certification of such demon-
stration by the manufacturer is acceptable. An additional
linearity performance check (see Section 7.2.1) must be made
before each use.
I 6.2 Zero Drift. One percent full scale per test period.
6.3 Span Drift. One percent full scale per test period.
| 7. Procedure
« 7.1 Sampling.
7.1.1 Assemble the systems as shown in Figure 1. Locate
the FIA in a suitably protected environment. Take particular
care that sample will be introduced to the FIA under the same
| conditions of pressure and flow rates as are used in calibration.
_ For specific operating instructions for the FID, refer to
manufacturer's manual.
7.1.2 Adjust the sample conditioning and analyzer heating
systems to provide a temperature of 350 to 400°F, or less, con-
sistent with the emission regulation, and allow the systems to
warm up.
7.1.3 Perform a leak check as follows before sampling.
Recheck to confirm that all fittings are tight. With the sample
I
I
-------
50
I
probe plugged, open the flow control valve and the excess sample
bleed valve. Use leak detection fluid or immerse the tubing I
leading from the bleed valve in a jar of water to check that
sample flow has ceased. At the conclusion of the sampling tests,
recheck for leaks. I
7.1.4 Begin Actual Sampling. Set the signal attenuation
to yield a minimum response of 10 percent of full scale unless |
the stack concentration is less than 1 ppm. Adjust the flow .
and bleed valves to minimize sample line residence time. Com-
pare instrument readings with the calibration curve to obtain I
emission concentrations based on the calibration gas.
7.1.5 At the conclusion of the sampling tests, but at |
least once every day, introduce zero and span gases to the _
analyzer to determine zero and span drifts. If the analyzer "
has drifted beyond the allowable performance specification,
the tests shall be considered invalid.
7.2 Calibration and Solvent Standards. |
7.2.1 Calibration Curve. Maintain a record of perform-
ance of each item. Determine the linearity of the analyzer as
follows: With the signal attenuation at the most sensitive
setting, introduce zero gas and adjust the respective zeroing
controls to indiate a reading of less than 1 percent of full I
scale. With the signal attenuation at the least sensitive
setting, introduce the span gas and adjust the span control to I
indicate the proper value on the analyzer readout. Repeat these
I
I
I
-------
I
51
I
_ two steps until adjustments are no longer necessary. Calculate a
" predicted response for the 5 to 10 ppm calibration gas. Introduce
that calibration gas and note the value obtained. If the value
is not within +_ 5 percent of its predicted value, then the
I analyzer may need repairs, or one or both of the calibration gases
may need replacement. In any event, this linearity performance
specification shall be met before the analyzer is placed in
actual use.
7.2.2 Preparation of Solvent Standard Gas Mixtures. (Optional--
I see Sections 1.2 and 5.6). Assemble the apparatus shown in Figure 2.
Evacuate a 50-liter Tedlar or aluminized Mylar bag that has passed
a leak check (described in Section 7.2.2.1) and meter in about
50 liters of air. Measure the barometric pressure, the relative
pressure at the dry gas meter, and the temperature at the dry gas
meter. Mhile the bag is filling use the 10 yl syringe to inject
10 yl of the solvent through the septum on top of the impinger.
I This gives a concentration of approximately 200 yg/liter.
In a like manner, use the other syringe to prepare dilutions having
approximately 40 and 20 yg/liter concentrations. To calculate the
specific concentrations, refer to Section 8.1. These gas mixture
standards may be used for a few days from the date of preparation,
| as determined by repetitive analysis for concentration degradation.
M (Caution: Contamination may be a problem when a bag is reused if the
new gas mixture standard is a lower concentration than the previous
I gas mixture standard.)
I
I
-------
52
I
I
7.2.2.1 Solvent Standards Bag Leak Checks. While performance
of this section is required subsequent to bag use, it is also advised I
that it be performed prior to bag use. After each use, make sure
a bag did not develop leaks as follows: to leak check, connect a I
water manometer and pressurize the bag to 5-10 cm H20 (2-4 in. hLO).
Allow to stand for 10 minutes. Any displacement in the water mano-
meter indicates a leak. Also, check the rigid container for leaks I
in this manner. (Note: an alternative leak check method is to
pressurize the bag to 5-10 cm HLO or 2-4 in. HLO and allow to stand |
overnight. A deflated bag indicates a leak.) For each sample bag »
in its rigid container, place a rotameter in line between the bag
and the pump inlet. Evacuate the bag. Failure of the rotameter to
register zero flow when the bag appears to be empty indicates a leak.
8. Calculations |
All measurements or calculations must be corrected for CH., if _
required by the emission regulation. "
8.1 Carbon or Surrogate Organic Compound Concentration.
8.1.1 Volume concentration [ppm]. To determine emission concen-
trations of total gaseous organics (wet basis) on a CH. or carbon |
equivalent basis, multiply the recorded emission values by the number _
of carbon atoms in a molecule of calibration gas. In some instances
it will be required to report emissions on the basis of the calibra-
tion gas, in which case no calculations are necessary.
8.1.2 Mass concentration [mg/m ]. To convert volume concentra- |
tion to mass concentration, proceed as follows: _
I
I
-------
I
I
53
I
I
8.1.2.1 Establish Standard Conditions. Find the volume
| occupied by 1 mg. mole of ideal gas at these conditions. Then
2
find the number of mg. moles in 1 m (at saturation).
8.1.2.2 Determine the molecular weight of the assumed organic
I compound in which the emission is to be expressed.
8.1.2.3 Use the values obtained in 8.1.2.1 and 8.1.2.2 to
determine the mass concentration (mg/m ) at saturation. Divide this
C }
number by 10 to find the mg/m equivalent to 1 ppm.
8.1.2.4 Multiply the result obtained in 8.1.2.3 by the volume
concentration obtained in 8.1.1. The result is the mass concen-
tration expressed in terms of the compound whose molecular weight
was determined in 8.1.2.2.
8.2 Organic Solvent Concentration
8.2.1 Solvent Standards Concentrations. Calculate each solvent
standard concentration prepared in accordance with Section 7.2.1.2
as follows: 3
I. mo) 10 y_q
I r = ul mg
c = Ml "'tf Equation 1
c .. u 293 m
where:
C = Solvent standard concentration, yg/1.
B = Number of yl of solvent injected.
I
Y = Dry gas meter, calibration factor.
I
d. = Density of the solvent at 293°A.
I
V = Gas volume measured by dry gas meter in liters.
P = Absolute pressure of the dry gas meter, mm Hg.
T = Absolute temperature of the dry gas meter, °A.
-------
I
54
8.2.2 Solvent Emission Concentrations The emission values I
in pg/ml are taken from the solvent standards response curve. Mo
further calculations are required.
I
9. References
1. Method 108, 43101-02-71T, "Tentative Method for Continuous
Analysis of Total Hydrocarbons in the Atmosphere (Flame lonization |
Method)." Methods of Air Sampling and Analysis, Intersociety _
Committee, Amer. Pub. Health Assn., Washington, D. C., 1972.
2. M. Johnson, "Oxygen Synergism in the Model 400 FIA."
Beckman Instruments, Inc., Fullerton, Ca. Oct., 1970.
3. Instruction Manual 82132-A, "Model 402 Hydrocarbon |
Analyzer." Beckman Instruments, Inc., Fullerton, Ca. Feb., 1971.
4. "Air-Hydrocarbon Monitoring Instrumentation," Lawrence
Berkeley Laboratory, Univ. of Ca., Berkeley, Ca. Nov., 1973.
5. A. J. Andreatch and R. Feinland, "Continuous Trace Hydro-
carbon Analysis by Flame lonization." Anal. Chem. 32 (8) 1021-4.
July, 1960.
6. R. A. Morris and R. L. Chapman, "Flame lonization Hydro-
carbon Analyzer." J. Air. Pol. Cont. Assoc. 11 (10) 467-9. Oct., 1961.
7. F. M. Black, L. E. High, and J. E. Sigsby, "The Application
of Total Hydrocarbon Flame lonization Detectors to the Analysis of I
Hydrocarbon Mixtures from Motor Vehicles, With and Without Catalytic
Emission Control." Water. Air. Soil Pollut. 5 (1) 53-62. Oct., 1975. I
I
I
I
I
-------
1
m TECHNICAL REPORT DATA
(Please read Instructions on the "everse bcjore completing)
j. F*>ORT NO. 2.
Ep-450/2-78-041
t. TITLE ANDSUBTITLE
MBisurement of Volatile Organic Compounds
1. AUTHOR(S)
Epssion Measurement Branch
). PERFORMING ORGANIZATION NAME AND ADDRESS
EHjssion Measurement Branch (MD-13)
Ejironmental Standards and Engineering Division
U. S. Environmental Protection Agency
Rwearch Triangle Park, North Carolina 27711
I2.HONSORING AGENCY NAME AND ADDRESS
Dm for Air Quality Planning and Standards (MD-10)
Office of Air, Noise, and Radiation
U«S. Environmental Protection Agency
Rmearch Triangle Park, North Carolina 27711
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
October, 1978
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
1
BSTRACT
I This document discusses the rationale of total volatile organics stationary
rce emission measurement through the determination of organic carbon mass
concentration. A conceptual approach for writing emission regulations in terms
o^volatile organic carbon is recommended, and drafts of two specific test methods
aB presented for regulation implementation. The methods are the measurement of
tRal gaseous nonmethane organics as carbon by the chromatographic oxidation/
reduction procedure, and the relative organic measurement derived by direct appli-
cBion of the flame ionization analyzer.
I
I
I
7. KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Mr Pollution
Analyzing
Imp! ing
ganic Compounds
Gas Sampling
8. IK I HIBUTION STATEMENT
lease Unlimited
b. IDENTIFIERS/OPEN ENDED TERMS
Stationary Sources
Volatile Organic
Compounds
Analytical Strategy
Organic Vapors
Environmental Assessment
19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS (This page)
Unclassified
c. COS AT I Field/Group
13B
21. NO. OF PAGES
54
22. PRICE
:PA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
I
-------
5 co n
.
^ CD
O Ctl ^J"
m ^ TJ m TI ~v
?"§ a ? 8 g
3 Q-
?! -n"
------- |