EPA
United States
-^Environmental Protection
;°^incy
Office of. Solid Waste
Washington DC 20460
April 1989
Hazardous Waste
GUIDANCE ON PICTCONTROtS
FOR HAZARDOUS WASTE
INCINERATORS
VOLUME V OF THE HAZARDOUS
WASTE INCINERATION GUIDANCE SERIES
-------
GUIDANCE ON PIC CONTROLS FOR
HAZARDOUS WASTE INCINERATORS
DRAFT FINAL REPORT
VOLUME V OF THE HAZARDOUS WASTE
INCINERATION GUIDANCE SERIES
U.S. Environmental Protection Agency
Office of Solid Waste
Waste Treatment Branch
401 M Street, SW
Washington, D.C. 20460
Work Assignment Manager: Mr. Shiva Garg
April 3, 1989
-------
This guidance was developed by the Office of Solid Waste, UVS. Environmental
Protection Agency with the assistance of Midwest Research Institute in partial
fulfillment of Contract No. 68-01-7287. Major contributors were Carlo
Castaldim", Drew Trenholm, John Pitcher, and. Shiva Garg. Contributions were
also made by Robert Holloway and the Incinerator Permit Writers Workgroup,
including Gary Gross, Y.J. Kim, Sonya Stelmack, and Betty Willis.
-i-
-------
HAZARDOUS WASTE INCINERATION GUIDANCE SERIES
Volume I Guidance Manual for Hazardous Waste Incinerator Permits, Mitre
Corp., 1983. NTIS #PB 84 100577.
Volume II Guidance on Setting Permit Conditions and Reporting Trial Burn
Results, EPA/625/6-89/019. Acurex, 1989.
Volume III Hazardous Waste Incineration Measurement Guidance Manual, MRI,
1989.
Volume IV Guidance on Metals and Hydrogen Chloride Controls for Hazardous
Waste Incinerators, 1989.
Volume V Guidance on PIC Controls for Hazardous Waste Incinerators, 1989.
Volume VI Proposed Methods for Measurements for CO, 02, THC, HC1, and Metals
at Hazardous Waste Incinerators, MRI, September 1988.
-ii-
-------
TABLrOF CONTENTS"
,v^^:^^s»^
SEaiON - j- PAGE
1.0 INTRODUCTION 1-1
2.0 AUTHORITY 2-1
3.0 SUMMARY OF GUIDANCE 3-1
3.1 Overview of CO and THC Controls 3-1
3.2 Recommended CO and THC Emission Limits 3-4
3.2.1 Tier I CO and THC Limits 3-4
3.2.2 Tier II CO and THC Limits 3-4
3.2.3 Formats for Monitoring Compliance with
the CO Permit 3-8
3.2.4 Monitoring During the Trial Burn 3-9
3.2.5 Monitoring THC Over the Life of the
Permit 3-10
3.2.6 Compliance Monitoring 3-10
3.3 Hazardous Waste Feed Cutoff ., 3-13
3.3.1 Maintaining Combustion Temperatures 3-13
3.3.2 Restarting Waste Feed. 3-14
3.4 Implementation of Risk-Based Approach to Establish
Tier II CO Limits 3-14
4.0 RATIONALE FOR PIC CONTROLS..... 4-1
4.1 Use of CO Limits to Ensure Good Combustion
Conditions. 4-2
4.2 Supporting Information on CO as a Surrogate
for PICs... 4-4
4.3 Alternate Formats for Compliance Monitoring with
the CO Limits 4-6
4.3.1 Methods for Specifying CO Limits 4-6
4.3.2 Rationale for Oxygen and Humidity
Corrections 4-9
4.4 Rationale for Recommending the Time-Weighted
Average CO Level for Tier II Permits '... 4-9
4.5 Equivalence of CO Mass Emissions Under CO Formats.. 4-10
4.5.1 Constraining Permitted Instantaneous
Peak CO To Trial Burn Level 4-13
-iii-
-------
TABLE OF CONTENTS
SECTION ' PAGE
4.5.2 Constraining Time (t) To A Specific
Limit. 4-13
4.5.3 Constraining The Base CO Limit To A Percent
Of The Time Average 4-14
4.6 Derivation of the Tier I CO Limit 4-15
4.7 Derivation of Tier II Risk-Based THC Limits 4-20
4.8 Derivation of Tier II Technology-Based THC Limits.. 4-21
4.8.1 Limitations of Risk Methodology 4-22
4.8.2 Basis for THC Limit of 20 ppmv 4-23
4.9 Methods for Monitoring THC 4-24
5.0 REFERENCES 5-1
-iv-
-------
~^ TABLE OF CONTENTS
APPENDICES
SECTION PAGE
APPENDIX A - TEST METHODS A-l
APPENDIX A.2 PERFORMANCE SPECIFICATIONS FOR CONTINUOUS
EMISSION MONITORING OF CARBON MONOXIDE AND OXYGEN
IN HAZARDOUS WASTE INCINERATORS, BOILERS, AND
INDUSTRIAL FURNACES A-3
1.0 APPLICABILITY AND PRINCIPLE A-3
1.1 Applicability A-3
1.2 Principle A-3
2.0 DEFINITIONS. A-3
2.1 Continuous Emission Monitoring System
(CEMS) A-3
2.2 Continuous A-4
2.3 Monitoring System Types A-4
2.3.1 Extractive ..* A-4
2.3.2 In-situ. A-4
2.3.3 Cross-stack A-4
2.4 Span A-5
2.5 Instrument Range A-5
1 2.6 Calibration Drift... A-5
2.7 Response Time A-5
2.8 Accuracy A-5
2.8.1 Calibration Error A-6
2.8.2 Relative Accuracy A-6
3.0 INSTALLATION AND MEASUREMENT LOCATION
SPECIFICATIONS. A-6
3.1 CEMS Measurement Location A-6
3.2 Reference Method (RM) Measurement Location
and Traverse Points A-7
-------
TABbE-OF-CONTENTS "
• -APPENDICES - •
(CONTINUED)
/
SECTION PAGE
4.0 MONITORING SYSTEM PERFORMANCE SPECIFICATIONS....... A-8
4.1 CEMS Span Values A-9
4.2 System Measurement Range A-10
4.3 Response Time ,.. A-10
4.4 Calibration Drift A-ll
4.5 Calibration Error A-ll
4.6 Relative Accuracy A-ll
5.0 PERFORMANCE SPECIFICATION TEST PERIOD A-12
5.1 Pretest Preparation A-12
5.2 Calibration Error and Response Time Tests A-12
5.3 Calibration Drift Test Period A-12
5.4 RA Test Period A-12
6.0 PERFORMANCE SPECIFICATION TEST PROCEDURES A-13
%
6.1 Response Time A-13
6.2 ^Calibration Error Test A-13
6.2.1 Procedure A-13
6.2.2 Calculations A-14
6.3 Zero and Span Calibration Drift A-14
6.4 Relative Accuracy Test Procedure A-15
6.4.1 Sampling Strategy for RM Tests A-15
6.4.2 Correlation of RM and CEMS Data ... A-15
6.4.3 Number of RM Tests A-16
6.4.4 Calculations... A-16
7.0 EQUATIONS A-16
7.1 Arithmetic Mean A-16
-vi-
-------
TABLE OF CONTENTS ™
APPENDICES
(CONTINUED)
SECTION PAGE
7.2 Standard Deviation A-17
7.3 Confidence Coefficient A-17
7.4 Calibration Error A-18
7.5 Relative Accuracy........ A-18
8.0 REPORTING.. .* A-19
9.0 REFERENCES. A-19
APPENDIX A.3 - MEASUREMENT OF TOTAL HYDROCARBONS IN HAZARDOUS
WASTE INCINERATORS, BOILERS, AND INDUSTRIAL
FURNACES A-20
1.0 APPLICABILITY AND PRINCIPLE A-20
1.1 Applicability A-20
1.2 Principle A-20
2.0 DEFINITIONS. A-20
2.1 Measurement System........... A-20
2.1.1 Sample Interface A-20
2.1.2 Organic Analyzer A-21
2.1.3 Data Recorder A-21
2.2 Span Value.. A-21
2.3 Calibration Gas A-21
2.4 Zero Drift A-21
2.5 Calibration Drift ...... A-21
2.6 Response Time A-22
2.7 Calibration Error , A-22
-vii-
-------
APPENDICES
(CONTINUED)
SECTION PAGE
3.0 APPARATUS A-22
3.1 Organic Concentration Analyzer.... A-22
3.2 Sample Probe A-22
3.3 Sample Line..., A-23
3.4 Calibration Value Assembly A-23
3.5 Participate Filter A-23
3.6 Recorder A-23
4.0 CALIBRATION AND OTHER GASES A-23
4.1 Fuel A-24
4.2 Zero Gas A-24
4.3 Low-Level Calibration Gas A-24
. 4.4 Mid-Level Calibration Gas A-24
4.5 H1gh-Level Calibration Gas A-24
5.0 MEASUREMENT SYSTEM PERFORMANCE SPECIFICATIONS A-24
5.1 Zero Drift A-24
5.2 Calibration Drift A-25
5.3 Calibration Error A-25
6.0 PRETEST PREPARATIONS... A-25
6.1 Selection of Sampling Site A-25
6.2 Location of Sample Probe A-25
6.3 Measurement of System Preparation A-25
-viii-
-------
-—• --- APPENDICES __
-— (CONTINUED)'
SECTION PAGE
_.. 6.4 ..Calibration Error Test A-25
6.5 Response Time Test A-26
7.0 EMISSIONS MEASUREMENT TEST PROCEDURE. A-26
7.1 Organic Measurement A-26
7.2 Drift Determination A-27
8.0 ORGANIC CONCENTRATION CALCULATIONS A-27
9.0 BIBLIOGRAPHY A-27
APPENDIX B - TECHNICAL BACKGROUND DATA FOR THC SCREENING
LIMITS.:~. B-l
APPENDIX C - HYDROCARBON CONVERSION FACTOR . C-l
* APPENDIX 0 - SAMPLE CASES — CO PERMIT DEVELOPMENT D-l
1.0 TIER I LIMITS ... D-l
1.1 Rolling Average CO Permit Format D-l
1.2 Cumulative Time Above Limit Format.. D-2
2.0 TIER II LIMITS 0-3
2.1 Rolling Average CO Permit Format D-3
2.2 Cumulative Time Above Limit Format D-4
-------
TABLE OF CONTENTS
APPENDICES
(CONCLUDED)
SECTION
PAGE
APPENDIX E - TECHNICAL BACKGROUND DATA.FOR THC EMISSION LIMIT
OF 20 PPMV
1.0 Allowable THC Emissions Under the Risk-Based
Approach
2.0 Existing Data Based on THC Emissions When
Burning Hazardous Waste
3.0 Calculated Risk Posed by 20 ppmv from Hazardous
Waste Incinerators
4.0 References.
E-l
E-l
E-3
E-15
E-19
APPENDIX F - LIST OF ACRONYMS.
F-l
-x-
-------
TABLE OF CONTENTS
LIST OF FIGURES
SECTION
PAGE
Figure la
Figure Ib
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Diagram of Procedure for Establishing CO/THC
Limits
Diagram of Procedure for Establishing Tier II
CO/THC Permit Under Risk-Based Alternative
CO and THC Versus Benzene Concentration
CO and THC Versus Vinyl Chloride Concentrations
CO and THC Versus Methyl Chloride Concentrations...
Trial Burn Hour With Highest Hourly Rolling
Average
Alternate Permit Format
3-2
3-3
4-5
4-7
4-7
4-11
4-11
-XI-
-------
TABLE OP CONTENTS
LIST OF TABLES
t
SECTION
PAGE
Table 1 Emission Rate Screening Limits for Total
Hydrocarbons 3-7
Table 2 H-l Values (m) Versus Stack Parameters 3-18
Table 3 Classification of Land Use Types 3-20
Table 4 CO Data from Research Tests 4-17
Table 5 CO Data from Trial Burns 4-18
Table A-l Performance Specifications of CO and Oxygen
Monitors A-9
Table A-2 CEMS Span Values for CO and Oxygen Monitors A-9
Table A-3 Values... 1 A-17
Table B-l Maximum Dispersion Coefficients Used to
Determine the Screening Limits B-4
Table B-2 Weighted Unit Risk Value for PICs B-5~
Table B-3 Noncarcinogens Emission Concentrations, RACs
and Actual Maximum Ambient Air Concentrations
for Reasonable Worst Case Dispersion
Coefficient B-7
Table C-l Weighted Average Molecular Weight Calculation C-3
Table E-l Allowable THC Levels Using Site Specific
Risk Assessment E-2
Table E-2 Incinerator CO/THC/Data From Research Tests E-4
Table E-3 Summary of Total Hydrocarbon Emission (THC)
Data from Industrial Boilers E-6
Table E-4 THC and CO Emissions from Cement Kilns
(CO Fire Tests Only) E-7
Table E-5 Industrial Boiler CO and THC Test Data E-8
Table E-6 Risk Determination for Site Specific
Incinerators E-16
Table E-7 Calculation of Risk from Modeled Facilities at
GOP Emissions of THC = 20 ppm E-17
-xii-
-------
ii.Rw^SSStii^.V™.
SECTION 1.0
INTRODUCTION
The U.S. Environmental Protection Agency (EPA) plans to propose amendments to
the Subpart 0, Part 264 hazardous waste incinerator rules. The Agency is
concerned that the existing standards may not adequately control products of
incomplete combustion (PICs). This guidance document is designed to assist
the permit writer in exercising his/her authority under Section 3005(c)(3) of
the Resource Conservation and Recovery Act (RCRA), as amended by the Hazardous
and Solid Waste Amendments of 1984 (HSWA), to develop permit requirements as
may be necessary to ensure that PIC emissions do not pose an unacceptable risk
to human health and the environment.
This document recommends ways of implementing controls for PIC emissions
consistent with the proposed rule. EPA believes that requiring incinerators
to operate at high combustion efficiency is a prudent approach to-minimize the
potential health risk from PICs. Given that stack gas carbon monoxide (CO)
concentration is a conventional indicator of combustion efficiency and a
conservative indicator of combustion upset (i.e., poor combustion conditions
possibly leading to higher PIC emissions), the Agency believes that
controlling CO is a prudent and reasonable approach to minimize the potential
risk from PICs. Accordingly, the proposed rule would limit stack gas CO
concentrations to 100 ppmv. EPA believes, however, that CO is a conservative
indicator of PIC emissions. That is, when CO levels are less than 100 ppmv,
PIC emissions are at levels that do not pose unacceptable health risk. When
CO levels are high, however, PIC emissions may or may not pose unacceptable
health risk. Consequently, the proposed rule would provide a waiver to allow
incinerators to operate at higher CO levels (i.e., higher than 100 ppmv)
provided that .the applicant demonstrates that PIC emissions do not pose
unacceptable health risk using a prescribed risk assessment, procedure. As an
alternative approach to this risk assessment procedure, however, EPA is
requesting comments on limiting total hydrocarbon (THC) concentrations to a
level that is consistent with good operating practices.
1-1
-------
.... .
EPA emphasizes that permit writers choosing to include permit provisions based
on this guidance must accept and respond to critical comment with an open--
mind, just as the Agency solicits public comment on the proposed regulations
with an open mind. In addition, permit writers must justify in the
administrative record supporting the permit, any decisions based on the
guidance. The administrative record to the proposed amendments to the
incinerator rules presents the basis for the proposed controls. Key parts of
this record are discussed in Chapter 4 and Appendix B of this document, and
could serve to justify the permit writer's use of the guidance. The key
point, however, is that in using the guidance, permit writers must keep an
open mind, accepting and responding to comment, and justifying use of the
guidance, or parts thereof, on the record, just as the Agency will respond to
comment on Its proposed rules and ultimately any final rule.
A summary of the guidance and rationale for the CO and THC limits under the
two-tiered permit setting approach are discussed in Sections 3.0 and 4.0,
respectively. Appendix A specifies recommended monitoring procedures.
Appendix B details technical background data and calculations used in
developing specific portions of the guidance. Appendix C specifies the
parameters for converting THC as reported to units for use 1n the THC
Screening Limits. Examples of developing CO permit limits using this guidance
are presented in Appendix D. Appendix E provides supporting information on
the alternative approach to the Tier II risk-based THC limits. Appendix F
provides a list of acronyms used throughout this document.
1-2
-------
.SECIKML.2JL
AUTHORITY
Section 3005(c)(3) of the Resource Conservation and Recovery Act (RC.RA), as
amended by the Hazardous and Solid Waste Amendments of 1984 (HSWA), provides
authority to EPA to establish permit conditions for hazardous waste facilities
beyond the scope of existing regulations. It states, "[e]ach permit...shall
contain such terms and conditions as the Administrator or State determines
necessary to protect human health and the environment." This language has
been added verbatim to EPA's hazardous waste regulations at 40 CFR 270.32
(b)(2) by the Codification Rule published at 50 FR 28701-28755 on July 15,
1985.l It is also listed as a self-implementing HSWA provision at 40 CFR
271.l(j) in 51 FR 22712-23 (September 22, 1986).
Because this guidance is implemented under HSWA's omnibus authority, it may be
put into effect immediately in all States, regardless of their authority
status. EPA has authority to implement this guidance in authorized States
until those States have revised their own requirements and such revisions have
been approved by EPA. The schedule for revising State requirements is given
in 40 CFR 271.21(6)(2), as revised in 51 FR 33722.
f • .
At present, EPA does not have the authority to reopen existing permits to
implement this guidance. After the PIC controls are promulgated, however,
existing permits may be reopened to provide permit conditions in conformance
with the new standards.
The preamble to this regulation provides EPA's legal interpretation and
discusses its impact on State authorization (50 FR 28728-33).
2-1
-------
fettV^-iiWi 4%-Sir- "-
-------
SUM4ARY OF GUIDANCE
This guidance document recommends controls on flue gas concentrations of
carbon monoxide (CO) and total hydrocarbons (THC) to limit emissions of
products of incomplete combustion (PICs). This section presents an overview
of the procedures to develop CO and THC permit limits (Section 3.1),
summarizes the recommended CO and THC limits (Section 3.2), lists additional
provisions of the guidance (Section 3.3), and discusses specific
implementation procedures (Section 3.4). Supporting documentation on
measurement requirements and. methods, and information needed to implement
specific portions of this guidance are presented in Appendices A and B,
respectively.
3.1 OVERVIEW OF CO AND THC CONTROLS
The CO and THC controls are based on a two-tiered approach. See Figures la
and Ib. Under Tier I, the applicant can demonstrate compliance by meeting the
recommended de minimis CO limit of 100 ppmv on an hourly rolling average.
Under Tier II, the de minimis CO limit would be waived if the applicant
demonstrates that THC emissions are not likely to pose unacceptable health
risk. This guidance provides two alternative approaches to demonstrate that
THC emissions are acceptable: (1) a risk-based approach based on site-
specific risk assessment (see Section 3.4); or (2) a technology-based approach
where the applicant demonstrates that THC levels do not exceed a good
operating practice-based level of 20 ppmv. As discussed in Section 4.8, the
Agency believes the technology-based approach is preferable.
The CO limits can be implemented in either of two formats: (1) an hourly
rolling average format; or (2) a cumulative hourly time-above-a-level
format. See Section 3.2.3. The cumulative hourly time-above-a-level format
is designed to allow hourly average CO emissions equivalent to the hourly
rolling average format. This alternative format is provided to minimize the
cost of instrumentation needed to monitor, analyze, and record CO levels.
3-1
-------
FIGURE 1a
DIAGRAM OF PROCEDURE FOR ESTABLISHING CO/THC LIMITS
Select
Preferred Format
for CO Limit
Rolling Average
Not to Exceed
Accumulative
'
Submit Incinerator
Trial Burn Data
1
Corrected
CO/THC Data
(
Does
Incinerator
^Comply with de Minimis
CO Limit
Tier 1]
Define CO
Permit Limit
(Tier 1)
ro
See Rgure 1b
Select
Permit Alternative under
Tier II
Tier II
Risk Assessmen
is Required
Technology-Based
*NOTE: Technology-based alternative is preferred.
Define CO/THC
Permit Limit
(Tier II)
-------
FIGURE 1b
DIAGRAM OF PROCEDURE FOR ESTABLISHING TIER I
CO/THC PERMIT UNDER RISK-BASED ALTERNATIVE
Corrected CO. THC.
CO
I
OJ
and Risk ^"""^^ ^-xDoes\^
Tier II Risk Assessment Data ^^ Are THC \^ Yes . ./" '?f '"f™?" \y«s
' "*^f^ ''V.nnr.Inn llnitlt "b ^"'V Meet I HC '
Assessment >-C screening Limits ^^ ^-^^ . ™cei "r- ^
Is Required ^^plicable^-^ • ^^SS"V^^
"|" No ^-Uo
1 /x
Perform ^ IS ^^- 'r'a':
^ * -c ^^ Modeling \. Yes . . u Terrain
Site— SD6C rIC ^k«r n x j X— — — — -^ In — rlOIISft , ^- "- '•>• "'-
^nv v^v/v/iiiv ^^^^^ Performed ^x^ ^^ nwuae ".^ rcmni wi
Risk Assessment ^^x^ln-House ^^ Analysis Uses GE
^r"No • Rolling or Fail
Complex Terrain
*• Request Applicant No PAT Determines Require
Risk Assessment Screening Model to Reduce
Emissions
Pass Fail Yes or Deny
PAI Runs Model A
J Pass v j Fail
i
i
tjefine CO/THC?
^ Permit Limits
(Tier II.
Option 1)
iter
MS
Pass
Ybefine CO/TH^
1 Cl INK. LMIIKb
(Tier II.
^ Option 2) ;
-------
3.2 RECOMMENDED CO AND THC EMISSION LIMITS
3.2.1 Tier I CO Limits
Under Tier I, the Incinerator should be operated so that CO levels 1n the
stack gas, corrected to seven percent oxygen, dry basis, do not exceed 100
ppmv on an hourly rolling average basis (or equivalent limits established
under the cumulative hourly time-above-a-level format). The permit should
limit CO to 100 ppmv even if lower levels were achieved during the trial burn.
The applicant should demonstrate compliance with the Tier I CO limit using the
highest hourly rolling average CO level recorded during the trial, burn for
destruction and removal efficiency (ORE) or under test burns with equivalent
conditions.
3.2.2 Tier II CO and THC Limits
Under Tier II, the 100 ppmv CO limit may be waived by a demonstration that THC
emissions are "acceptable" under either a risk-based approach or a technology-
based approach. These alternative "approaches are discussed below.
3.2.2.1 Technology-Based Approach for Evaluating THC Emissions
*
Under the technology-based approach, the applicant would demonstrate that THC
levels in the stack gas do not exceed a level ~ 20 ppmv (hourly rolling
average, corrected to seven percent oxygen, dry basis and reported as propane)
— considered to represent good operating practice. The Agency has used the
risk assessment methodology discussed below (Section 3.2.2.2) to show that a
THC concentration of 20 ppmv would be protective of public health under
reasonable, worst-case scenarios.
If the applicant demonstrates during the trial burn that the highest hourly
average THC concentration does not exceed 20 ppmv, the permit writer should
select the time-weighted average CO level during the trial burn as the CO
permit limit. THC should be monitored continuously during the trial burn in
3-4
-------
accordance with methods specif Ted^trr"Appendix A ,and Reference 1." The perm ft
writer is also encouraged to require continuous monitoring of THC over the
life of.-the permit. See Section 3.2.5. The permit should limit THC to 20
ppmv even if the highest hourly average concentration during the trial burn
was lower.
The highest hourly rolling average THC concentration should be used for this
evaluation regardless of wheir this average occurred during the trial burn.
For example, under the current trial burn guidelines, three separate test runs
under equivalent incinerator operating conditions are required for evaluation
of compliance with RCRA standards. The highest hourly average THC emission
rate that occurred during any one of the three test runs should not exceed 20
ppmv even if the highest hourly average CO level occurred during one of the
other two test runs.
As discussed in Section. 4.8, the Agency believes this technology-based
approach 1s preferable to the risk-based approach discussed below. -After
developing the risk-based approach for the proposed rule the Agency realized
that the approach had serious limitations. Consequently, the Agency recently
developed the technology-based approach and is requesting public comment on it
in the proposed rule as an alternative to the risk-based approach. Based on
the public comment and further analysis, the Agency may promulgate this
approach in lieu of the risk-based approach.
3.2.2.2 Risk-Based Approach for Evaluating THC Emissions
Under the risk-based approach, the applicant would document THC levels in the
stack gas and demonstrate that the THC emissions do not pose unacceptable
health risk using prescribed procedures. To make this demonstration, the
applicant can either: (1) show that THC emissions do not exceed conservative
Screening Limits established as a function of stack height and type of terrain
in the vicinity of the facility (Option 1); or (2) conduct site-specific
dispersion modeling and risk assessment to show that the incremental lifetime
cancer risk to the most exposed individual (MEI) does not exceed 10" or 1 in
100,000 (Option 2).
3-5
-------
- - .
The THC~~Scfeening Limits are back-calculated from the T6~ risk-specific dose
(based on~a~ calculated," conservative unit risk value as described in
Appendix C) for THC using reasonable, worst-case dispersion coefficients. See
2
Table 1. Those dispersion coefficients are developed to be conservative by a
factor of 2 to 10.3 Thus, site-specific dispersion modeling under Option 2
can result 1n substantially higher allowable THC emissions than allowed by the
Screening Limits. , -
As with the technology-based approach, the highest hourly rolling average THC
emissions that occur during the trial burn should be used for this
evaluation. If the permit writer requires continuous monitoring over the life
of the permit as recommended, he/she should limit THC to the time-weighted
average concentration that occurs during the trial burn. This is more
conservative than limiting emissions to the highest hourly level during the
trial burn (which would allow THC emissions over the life of the permit at the
highest level that occurred during only a fraction of the trial burn).
To establish the permit limit for CO, the time-weighted average CO level
during the trial burn should be used.
Detailed step-by-step procedures for implementing the risk-based approach are
discussed in Section 3.4.
The measured THC concentration in ppmv .must be converted to the
Screening Limits units of mg/s using equation 6 shown in Step B on p.
3-21 of this document.
The permit writer must determine whether the THC Screening Limits are
appropriate (i.e., conservative) for the specific facility in
question. Although the Limits are derived from dispersion analyses of
reasonable, worst-case facilities, the limits may not be fully
protective 1n every situation. A particular facility may, in fact, have
poorer dispersion than the reasonable, worst-case facilities used to
develop the Limits. Site-specific dispersion modeling will be necessary
If the criteria stated in Section 3.4 of the text are not met.
3-6
-------
TABLE 1
EMISSION RATE SCREENING LIMITS FOR TOTAL HYDROCARBONS (mg/sec)
TERRAIN-
ADJUSTED
EFFECTIVE
STACK HEIGHT
NONCOMPLEX TERRAIN
COMPLEX TERRAIN
On)
4
6
8
10
12
14
16
18
20
22
24
26
28
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
105
110
115
120
URBAN LAND USE
54
61
69
77
88
99
110
130
140
160
180
200
230
260
340
430
540
700
880
1,100
1,300
1,500
1,700
1,900
2,200
2,500
2,800
3,200
3,600
4,100
4,600
5,300
RURAL LAND USE
28
32
36
42.
51
62
77
86
120
150
190
250
309
400
630
960
1,300
1,800
2,300
3,100
4,100
4,900
5,800
6,900
8,200
9,700
12,000
14,000
16,000
20,000
23,000
28,000
13
19
27
40
49
60
69
77
85
94
100
120
130
140
180
220
270
330
410
500
620
690
770
860
970
1,100
1,200
1,400
1,500
1,700
1,900
2,100
Note 1: Applicability of these THC Screening Limits depends on the
Incinerator meeting specific criteria outlined 1n Section 3.4,
Step A.
Note 2: See Section 3.4 for direction in selecting appropriate values for
effective stack height, terrain, and land use.
Note 3: For effective stack heights not shown in this table, use
interpolation to calculate applicable Screening Limit. Results
should be rounded off to two significant figures.
Source: Referenced.
3-7
-------
3.2.3 Formats for Monitoring Compliance with the CO Permit
After the CO limits under either Tier I or Tier II are set as specified in
Sections 3.2.1 and 3.2.2, the requirements for monitoring compliance with the
CO limits should be established in the permit under either of the following
formats:
Format A. Hourly rolling average format:
• The permitted level . is the arithmetic mean of the 60 most
recent minute average values recorded by the monitoring system
(100 ppmv for Tier I); or
Format B. Cumulative hourly time-above-a-level format, where the
permitted levels consist of:
• A maximum instantaneous peak CO limit which cannot be exceeded
at any time; and
• A lower limit that can be exceeded only up to a specified
period (e.g., 5 or 15 minutes) in any clock hour of operation.
To establish Format B CO limits and the time of exceedance in any hour the
permit writer should consider the preferences of the applicant. The guiding
requirement, however, is that the total permitted mass CO emission rate should
not exceed that allowed under the hourly rolling average format in any hour of
operation (e.g., 6,000 ppmv-minutes for Tier I)'. A few examples of suggested
methods to ensure equivalence of mass emissions between the two formats are
discussed 1n Section 4.5. Note that the applicant may elect to accept a
maximum Instantaneous peak CO limit, for example, that is higher or lower than
what occurred during the trial burn. The trial burn CO levels, thus, are used
to establish the total permitted mass CO emission rate that cannot be exceeded
under Format B, not necessarily to establish the specific CO and time
limits. Additional requirements for monitoring compliance are discussed in
Section 3.2.5.
3-8
-------
The Agency notes that some reviewers of this guidance document have recently
-suggested thatr the geometric mean should be-used-rat her than the-ar-ithmetic-^
,_^•^.**S*?gj^via-A>i^a^ _, ,_.<_,,*. , :,-.'-: .•-..-.>—s^-~»i-—»•-»-•- • .•••V • .-,', - ; '- '-
mean to establish the hourly rolling average given that CO (and THC) emissions
are distributed lognormally. When CO (and THC) levels are relatively steady
with only an occasional spike, the geometric mean can be substantially lower
(e.g.," a "factor of ~3)"""than" thearithmetic7 mean. The Agency continues to
believe, however, that the arithmetic mean should be used because: (1) CO is
used as a surrogate for combustion efficiency, a parameter that is based on an
instantaneous measure of CO; (2) we are allowing for inevitable spikes in CO
levels that occur even when facilities are well designed and operated by
allowing the CO levels to be averaged; (3) we believe that the flexibility
provided by an arithmetic average is sufficient to enable the vast majority of
incinerators to routinely meet the recommended 100 ppmv CO limit; and (4) for
facilities that cannot easily meet the 100 ppmv CO limit, we are recommending
a waiver that would allow higher CO levels provided that THC levels do not
exceed 20 ppmv.
3.2.4 Monitoring During the Trial Burn
CO should be monitored in accordance with procedures defined in Appendix A.2
and reported on a dry basis corrected for seven percent oxygen in the stack
gas. During the trial burn, CO emissions should be reported as follows:
• For compliance with Tier I, Format A: hourly rolling average levels
and the highest hourly average (HHA);
• For compliance with Tier I, Format B: instantaneous levels, the
highest instantaneous level, and the HHA;
• For compliance with Tier II, Format A: hourly rolling average
levels, the HHA, and the time-weighted average level;
^
For compliance with Tier II, Format B: instantaneous levels, the
highest instantaneous level, and the time-weighted average level.
3-9
-------
THC should be monitored"in accordance with procedures discussed in Section 4.9
wand?—Appendix A.3. When demonstrating compliance with the 20 ppmv good
operating practice-based limit, THC should be reported as propane, on a dry
basis, and corrected to seven percent oxygen. During the trial burn, THC
levels should be reported as follows:
* For compliance with the 20 ppmv limit: hourly rolling average
levels and the highest hourly average (HHA);
• For compliance under the risk-based approach: hourly ro.lli-ng
average levels, the HHA, and, 1f THC monitoring will be required
over the life of the permit (see Section 3.2.5), the time-weighted
average level.
3.2.5 Monitoring THC Over the Life of the Permit
The Agency believes that THC should be monitored continuously over the life of
the permit when CO levels exceed 100 ppmv (i.e., under Tier II). This is
because there does not appear to be a correlation between CO and THC or PIC
emissions when CO levels exceed 100 ppmv — THC levels may or may not be
high. The concern is that although THC levels may be low during the trial
burn when CO 1s high, changes 1n combustion conditions within those allowed by
the permit could result in high THC levels.
3.2.6 Compliance Monitoring
Monitoring for compliance with the CO permit limits under both formats
Includes the following requirements:
1. The CO limits for either Tier I or Tier II are based on an oxygen
concentration of 7 percent in the stack gas. When the stack gas oxygen
content differs from 7 percent, measured CO levels should be corrected
continuously for the amount of oxygen in the stack gas according to the
formula:
3-10
-------
CCL = OL x 14
C m
where COC is the corrected concentration of CO in the stack gas, C0m is
the measured CO concentration according to guidelines specified in
Appendix A, and Y is the measured oxygen concentration on a dry basis in
the stack. Oxygen should be measured at the same stack location that CO
is measured.
2. When oxygen enriched air is used for incineration, a different
correction factor is necessary to account for the reduced volume of
gas. In such cases, the corrected CO concentration should be calculated
as follows:
CO. = C0m x 14 (2)
c m £ _ y
Where E is the oxygen enriched concentration in the total combustion air
(e.g. 30 percent), and Y is the measured oxygen concentration in the
stack gas on a dry basis.
For oxygen enriched incinerators the Regional Administrators may select
a different correction procedure and specify it in the facility permit.
3. Compliance CO monitoring for facilities using the hourly rolling average
limits in the permit (Format A) requires instrumentation that
continuously calculates hourly rolling averages, and that continuously
adjusts the oxygen correction factor to record CO levels based on 7
percent oxygen.
4. Under Tier II, THC monitoring during the trial burn is required and
continuous THC monitoring over the life of the permit is highly
recommended. Compliance monitoring for THC will require continuous
hourly rolling average instrumentation with continuous adjustment of
concentrations to seven percent oxygen in a manner that 1s consistent
with the requirements for CO compliance monitoring.
3-11
-------
5, All continuous morntpring systems for measuring CO and THC emissions
should complete -a minimum of one cycle of sampling and analyzing for
each successive 15 second period and- one cycle of data recording for
each successive 1 minute period. At each successive minute, the 60 most
recent 1 minute averages should be used to calculate .and record a 1-hour
rolling average.
Both the one minute average and the most recent 60 minute average are
calculated as an arithmetic average (as follows):
n
Avg » 1 Y X-j where
n 1-1
(3)
n = number of observations
X.j » individual observations
6. For facilities using Format B for CO monitoring (cumulative hourly time-
above-a-level) the oxygen correction factor need not be determined
continuously. Rather, the appropriate correction factor may be
determined initially during the trial burn (or by data in lieu of a
trial burn) and the CO monitor calibrated accordingly to correct for
7 percent oxygen. For compliance, the stack gas oxygen correction
factor, determined according to equation (1) above, should be determined
at least annually thereafter, unless specified otherwise in the
permit. Annual determination of the oxygen correction factor is deemed
appropriate in most cases because the concern is whether duct in-leakage
has substantially changed over time. The fact that excess oxygen level
also changes with types of waste and feedrate should be considered in
establishing the correction factor initially. The 02 correction
factors, so determined, are "hardwired" into the CO monitoring system to
continuously monitor compliance with the corrected CO limit.
/
7. The CO and THC limits are on the basis of a dry stack gas. When
instruments that measure CO and THC on a wet basis are used, a
correction factor should be used to convert the measured value to a dry
basis. This correction factor for humidity should initially be
determined during the trial burn, and annually thereafter unless
specified more frequently in the permit.
3-12
-------
3.3 HAZARDOUS WASTE FEED CUTOFF
The CO and THC recorders should be linked to an automatic waste feed cutoff to
engage the cutoff when a CO or THC limit is exceeded. Recommendations are
discussed below on: (1) requiring that combustion temperature be maintained
while residues remain in the combustion chamber after a cutoff; and (2) when
the hazardous waste feed should be allowed to resume.
3.3.1 Maintaining Combustion Temperatures
To minimize emissions of toxic pollutants when the hazardous waste feed is
cutoff, combustion chamber temperature specified in the permit and the air
pollution control equipment operation should be maintained as long as the
wastes remain in the combustion chamber. For incinerators with a secondary
combustion chamber, temperatures should be maintained in the secondary chamber
and the permit writer should use his/her engineering judgement to determine if
temperatures should be maintained in the primary combustion chamber as well.
Adequate auxiliary burner capacity may be needed to maintain the temperature
in the combustion chamber(s) and allow destruction of the waste materials and
associated combustion gases left in the incineration system after the waste
feed is automatically cutoff. The safe startup of the burners using auxiliary
. fuel requires approved burner safety management systems for prepurge, pilot
lights, and induced draft fan starts. If these safety requirements preclude
immediate startup of auxiliary fuel burners and such startup is needed to
maintain temperatures (I.e., if the combustion chamber temperatures drop
precipitously after waste feed cutoff), the auxiliary fuel may have to be
burned continuously on "low fire" during nonupset conditions. After an
automatic cutoff, hazardous waste should not be used as auxiliary fuel unless
the waste 1s exempt under existing § 264.340(b) or (c) from the emissions
standards because the waste is hazardous solely because it is ignitable,
corrosive, or reactive, or it contains insignificant levels of toxic
constituents.
3-13
-------
"3.3.2 Restarting Waste Feed
CO and THC monitoring should continue after a waste feed cutoff as long as
waste remains in the combustion chamber.
When the automatic waste feed is triggered by a CO exceedance, the permit
should allow the hazardous waste feed to restart after an automatic feed
cutoff when the instantaneous CO level meets the hourly rolling average
limit. We considered whether to allow a restart only after the hourly rolling
average no longer exceeded the limit or after an arbitrary 10 minute time
period to enable the operators to stabilize combustion conditions. We do not
believe that either of these alternatives are appropriate. It may take quite
a while for the hourly rolling average to come within the limit while the
event that caused the exceedance may well be over even before the CO monitor
reports the exceedance. Consequently, it appears reasonable to allow restarts
after the instantaneous CO levels meet the hourly rolling average limit.
When the automatic waste feed cutoff is triggered by a THC exceedance, we
recommend that a more conservative approach be used to allow restarts given
that the THC monitor is a better surrogate for toxic organic emissions than
CO. The permit should allow a restart after the hourly rolling average THC
level has been reduced to 20 ppmv or less.
3.4 IMPLEMENTATION OF RISK-BASED APPROACH TO ESTABLISH TIER II CO LIMITS
This section presents a step-by-step approach to implement the site-specific,
risk-based approach to determine if THC emissions are likely to pose
substantial health risk. If not, the Tier I CO limits of 100 ppmv should be
waived and the permit should allow (higher) CO levels based on the trial
burn. Note that these procedures do not pertain to the preferred approach of
waiving the 100 ppmv Tier I CO limit: demonstrating that THC levels do not
exceed a good operating practice-based level of 20 ppmv.
The major steps depicted in Figure Ib (page 3-3) are:
• Step A - collect necessary input data
3-14
-------
Step-B---determine compliance with the THC Screening Limits (Table 1)
• Step C - conduct site-specific dispersion modeling and risk assessment,
and
Step D - determine the CO and THC permit limits.
Each of these steps 1s described below.
Step A; To determine compliance with risk-based THC limits and to establish
CO permit limits under Tier II, the permit writer needs to gather from the
applicant information on the source(s) and facility in order to:
(1) Calculate indnerator(s) terrain-adjusted effective stack height;
(2) Define terrain (i.e. complex or non-complex);
(3) Determine land use classification (i.e. urban or rural) if the terrain
1s non-complex; and
(4) Determine applicability of THC Screening Limits.
The THC Screening Limits of Table 1 classify facilities 1n terms of terrain-
adjusted effective stack height, noncomplex versus complex terrain, and" urban
versus rural land use. Information needed from the applicant to make these
determinations for any given hazardous waste incineration facility includes:
a. Reference Information—facility name, address, etc.
b. Stack-related parameters—height of the stack above grade,
exhaust gas temperature and velocity, and inner diameter of the
stack.
c. Facility-specific information—the stack location shown on USGS
7.5 min topographical map(s) for a 5 km radius centered on the
facility; location and dimensions (length, width, and height)
of major buildings; distance of these buildings from the stack
if distance is less than 5 times the building height or 5 times
3-15
-------
the projected building width; latitude/4ongitude~and=..Un4vei5sal-
Transverse Mercator (DIM) coordinates for each stack, and
position of stack(s) relative to property line.
The USGS map(s) and stack(s) information submitted by the applicant provide
the means to make the following determinations:
Terrain-Adjusted Effective Stack Height
The terrain-adjusted effective stack height is the sum of physical stack
height plus plume rise minus the maximum terrain rise above ground level
within 5 km radius of the stack as shown by the equation on this page.
If the facility has more than one hazardous waste incinerator stack, the
worst-case stack must be determined to use the THC Screening Limits. The
worst case stack 1s determined by applying the following equation to each
stack:
K = HVT (4)
where: K = an arbitrary parameter accounting for relative influence
of stack height, and plume rise,
(m* K/sec)
H - stack height (m)
V =* flow rate (mVsec)
T = exhaust temperature (K)
The stack that has the lowest value of K should be used as the worst-case
stack.
To determine the terrain-adjusted effective stack height, use the
following expression
' He = HA + H! - R (5)
3-16
-------
where He is the terrain-adjusted effective stack height, HA is the
physical stack height, H! is plume rise factor from Table 2 defined by
the gas flowrate and the gas temperature, and R is the maximum terrain
rise within 5 km from the stack. In cases where the above formula yields
a value of less than 4 m f or the terrain-adjusted effective stack height,
a minimum of 4 m shall be used. The tables have been calculated such
that the THC Screening Limits given therein are conservative for any
stack height of 4 meters or less.
Note that the physical stack height used to determine the effective stack
height cannot be any greater than Good Engineering Practice (GEP) for the
facility. GEP stack height is defined as the greater of 65 m or, Hg = H
+ 1.5 L, where:
Hg = GEP stack height measured from ground level elevation at the base of
the stack,
H = height of nearby structure measured from ground level elevation at
the base of the stack,
U= the lesser dimension of the height or projected width of a nearby
structure [see 40 CFR 51.1 (11)1.
For this analysis, all buildings within a distance from the stack of 5 times
the building height or 5 times the projected building width, whichever is
greater, should be considered.
• Terrain Characteristics
The terrain surrounding the stack is examined for a 5 km radius to
determine whether the facility lies in noncomplex (i.e.., rolling or flat)
or complex terrain. If the terrain rise, within 5 km, is greater than
the physical stack height, the facility is considered to be in complex
terrain for purposes of this analysis. If this terrain rise is between
10% and 100% of the physical stack height, the terrain will be classified
3-17
-------
TABLE 2. H-l VALUES («) VERSUS STACK PARAMETERS
CO
i
h-«
oo
Flow rate
(m3/sec)
< 0.5
0.5-0.9
1.0-1.9
2.0-2.9
3.0-3.9
4.0-4.9
5.0-7.4
7.5-9.9
10.0-12.4
12.5-14.9
15.0-19.9
20.0-24.9
25.0-29.9
30.0-34.9
35.0-39.9
40.0-39.9
50.0-59.9
60.0-69.9
> 69.9
Exhaust temperature (K)
< 325
0
1
1
1
2
2
3
3
4
5
6
7
8
9
10
It
14
16
18
325-
349
0
1
1
1
2
2
3
4
5
5
6
8
9
10
12
13
15
18
20
350-
399
0
1
1
2
3
3
4
5
7
8
9
11
13
15
17
19
22
26
29
400-
449
1
1
2
3
4
5
6
8
10
12
13
17
20
22
25
28
33 „
38
42
450-
499
' !
i :
2
4
5
6
7
10
12
14
16
20
24
27
31
34
40
45
49
500-
599
1
1
2
4
6
7
8
11
14
16
19
23
27
31
35
39
44
50
54
600-
699
1
2
3
5
7
8
10
13
16
19
22
27
32
37
41
44
50
56
62
700-
799
1
2
3
5
7
9
11
14
18
21
24
30
35
40
44
48
55
61
67
800-
999
1
2
3
6
8
10
11
15
19
22
26
32
38
42
46
50
57
64
70
1,000-
1,499
1
3
4
6
8
10
12
17
21
24
28
35
41
45
50
54
61
68
75
> 1,499
1
2
4
7
9
11
13
18
23
- 27
31
38
44
49
54
58
66
74
81
Source: Reference 2.
-------
as "rolling."" -"Flat" terrain signifies terrain rise of less than 10% of
the physical stack height. Worksheet 1 of the Metals/HCl Guidance
Document (Reference 2) can be used as a guide to determine terrain
characteristics.
i
Land Use Characteristics
Next, the land use characteristics in a 3 km radius of the stack is
assessed. Topographic maps, zoning and/or aerial photographs can be used
to Identify land use types. However, this approach can be time consuming
and cumbersome. As an alternative, a simplified procedure is given in
Appendix I of "Guidance on Metals and Hydrogen Chloride Controls for
Hazardous Waste Incinerators" (Reference 2). This procedure (Auer 1978)
is consistent with the "EPA Guideline on Air Quality Modes" and makes use
of 12 classifications of land use shown in Table 3.
If the urban land use types within a 3 km radius of the stack, are less
than 50 percent-of the total area based on a planimeter (or 30 percent if
based on a visual estimate), the land use characteristic is rural. If
the urban land use types are greater than 30 percent by visual estimate
or 50 percent based on the planimeter measurement, the more conservative
(lower) value between the urban and rural Screening Limits should be
used, or the standard Auer land use technique applied.
Applicability of THC Screening Limits
The final determination in the data gathering step concerns the
applicability of the THC Screening Limits (Table 1). If the facility
meets any of the following criteria, the THC Screening Limits may not be
conservative and should not be used. In that case, site-specific
modeling (or the screening model) should be used (see Step C). If the
screening limits are applicable, proceed to Step B.
The facility is located in a narrow valley less than 1 km wide; or
The facility has a stack taller than 20 m and is located such that the
3-19
-------
Table 3
Classification of Land Use Types
Type1
11
12
Cl
Rl
R2
R3
R4
Al
A2
A3
A4
A5
Description Urban or
Heavy Industrial
Light/Moderate Industrial
Commercial
Common Residential
(Normal Easements)
Compact Residential
(Single Family)
Compact Residential
(Multi-Family)
Estate Residential
(Multi-Acre Plots)
Metropolitan Natural
Agricultural
Undeveloped
(Grasses/Weeds)
Undeveloped
(Heavily Wooded)
Water Surfaces
Rural Designation2
Urban
Urban
Urban
Rural
Urban
Urban
Rural
Rural
Rural
Rural
Rural
Rural
EPA, Guideline on Air Quality Models (Revised), EPA-450/2-78-027, Office
of Air Quality Planning and Standards, Research Triangle Park, North
Carolina, July 1986.
Auer, August H. Jr., "Correlation of Land Use and Cover with
Meteorological Anomalies," Journal of Applied Meteorology, pp. 636-643,
1978.
3-20
-------
----- .
w^^ height within 1 km of the facility;
or
• If stack height is less than 2.5 times building height and either the
distance from the stack to the building is less than 5 times the building
height or if the distance from the building to the stack is less than 5
times the maximum projected building width, site-specific analysis is
required because of the potential downwash complication at the MEI
receptors; or
• On-site receptors are of concern, and the stack height is less than or
equal to 10 m.
Step B; This step determines compliance with the THC Screening Limits.
• Conversion of THC Concentration (ppmv) to Mass Emissions (mq/s)
THC levels from the trial burn will be reported as ppmv propane and need
to be converted to the mg/s units used for the THC Screening Limits.
This conversion is accomplished with the following equation:
THC, mg/s = (THC ppmv propane) x (Stack gas flow) x 0.028 (6)
where:
• THC = concentration as measured by the THC method, (see Appendix
A.3), ppmv propane;
• Stack gas flow = in dry standard cubic meters per minute measured by
EPA Reference Method 5 (or Modified EPA Method 5) during the ORE
trial burn; and
Constant factor 0.028 (See Appendix C).
3-21
-------
Compare THC Emission Rate with Applicable Screening Limit
Read the THC emission rate Screening Limit (see Table 1) that corresponds
to the appropriate terrain-adjusted effective stack height, terrain, and
land use. Compare the maximum hourly average THC emission rate during
the ORE trial burn with the applicable Screening Limit. If the
calculated terrain-adjusted effective stack height falls between two
values shown in Table 1 (e.g., 19 m falls between 18 and 20 m)
interpolate between the corresponding THC values to obtain the applicable
Screening Limit.
As stated in Section 3.2.2,
• If the maximum hourly average THC emission rate (mg/s) exceeds the
Screening Limit, proceed to Step C. (Tier II, Option 2)
• If the maximum hourly average THC emission rate is less than or equal to
the Screening Limit, then set the CO limit or THC as specified in Step D
(also Section 3.2.2). There is no need for additional evaluation of PIC
risk.
For facilities with multiple on-site stacks, compare the Screening Limit for
the worst-case stack with the total THC emission rate from all incinerators
(i.e., all emissions are assumed to be emitted from the worst-case stack).
Step C: This step involves site-specific dispersion modeling. The permit
writer must determine whether to require the applicant to conduct the modeling
and to demonstrate that the incremental lifetime cancer risk to the maximum
exposed individual (MEI) does not exceed 1 in 100,000 (10~5), or to conduct
the modeling (and risk assessment) in-house. The permit writer may choose to
conduct the modeling in-house for a number of reasons. If the facility is
located in flat terrain, the permit writer can use EPA's GEMS ,dispersion
modeling program to predict the maximum annual average ground level
concentration. As discussed below, GEMS is a readily available, user-friendly
program that incorporates the regulatory dispersion model Industrial Source
Complex Model, Long Term (ISCLT)1* for flat terrain. Although the permit
3-22
-------
"writer will not normally attempt to model facilities in rolling or complex
terrain using EPA's regulatory models given the complexity of the modeling,
the permit writer may decide to use a conservative screening model for such
terrain. The screening model (see discussion below) does not require the use
of site-specific meteorological data. Thus, for example, the screening model
would be useful when representative meteorological data are not readily
available.
Step C is discussed below in two sections: procedures for modeling conducted
in-house, and procedures for when the applicant conducts the modeling.
Modeling Conducted In-House
A. For flat terrain, use the ISCLT available through EPA's Graphical
Exposure Modeling System (GEMS) to determine the maximum annual average
dispersion coefficient. See Appendix II of the Metals Guidance Document
for specific input requirements (Reference 2). Use the GEMS dispersion
coefficient and the measured THC emission rate to determine if the
acceptable ambient level of 1.0'ug/m3 1s exceeded. (This ambient level
is based on the acceptable risk to the MEI of 10""5 and the THC unit risk
of 1.0 x 10~5 m3/yg. See Appendix B for technical background data and
development of this THC unit risk).
Thus, the following equation applies for compliance determination:
GEMS Dispersion Coefficient (yg/m3/g/s) X THC Emissions (g/s) < 1.0 (ug/m3) (7)
Where the THC unit risk value is 1 x 10"5 ma/yg, as shown in Appendix B.
If potential THC risk is unacceptable, THC emissions must be reduced.
B. For rolling and complex terrain, the permit writer may use a screening
model. See Appendix V of the Methods/HCl Guidance Document
(Reference 2).
* Regulatory model refers to EPA's recommended dispersion model as provided
by EPA's "Guidelines on Air Quality Models (Revised)," July 1986.
3-23
-------
Tfie" screening model, in lieu of regulatory modeling, is particularly
useful for facilities where: (1) the Screening Limits were not
appropriate (i.e., they may not be conservative); (2) the Screening
Limits were too conservative, or (3) the cost of regulatory modeling is
substantial. Specific situations where the screening model is useful are
where:
• The facility has multiple stacks with substantially different
release specifications (e.g., stack heights differ by > 50 percent,
exit temperatures differ by > 50 K, or exit flow rates differ by
more than a factor of 2).
• The terrain does not reach stack height within 1 km of the
incinerator, when the stack is greater than 20 m high and in complex
terrain.
• There are no representative meteorological data available for the
site under consideration.
• The distance to the nearest facility boundary is greater than the
distance.shown in the table below for land use type and the terrain-
adjusted effective height of the stack under consideration.
Terrain-Adjusted
Effective Stack Height Distance (m)
Range (m) Urban Rural
1 to 9.9200 200
10 to 14.9 200 250
15 to 19.9 200 250
20 to 24.9 200 350
25 to 30.9 200 450
31 to 41.9 200 550
42 to 52.9 250 800
53 to 64.9 300 1,000
65 to 112.9 400 1,200
113+ 700 2,500
The permit writer should use the dispersion coefficient predicted by the
screening model and the actual THC emission rate (and metals emission rates as
discussed above) to determine if the 10~5 risk level is exceeded:
3-24
-------
If risk is acceptable, establish CO and THC limits as described in
Step D.
If risk is unacceptable, proceed to the Applicant Conducts Modeling
section below.
Applicant Conducts Modeling
A. The applicant needs to draft a dispersion modeling plan for a site-
specific analysis consistent with the EPA Guideline on Air Quality Models.
The following documentation should be provided with the draft modeling
plan:
The rationale for the selection of the meteorological monitoring
station, including a map showing alternative stations considered in
the region.
• A site layout map showing the locations of all sources, and building
dimensions for adjacent structures.
The applicant must include a discussion on how a follow-up run will be
used to perform a more refined analysis around the area of maximum annual
average off-site concentrations.5 In addition, special receptors should
be used --to define the distance to the fenceline for each wind direction
sector if the initial model runs show that the maximum inputs occur
within the first kilometer from the source.
The permit writer should send the draft modeling plan and supporting
documentation to the Regional Meteorologist or PAT for review. The
application must revise the modeling plan based on recommendations of the
Regional Meteorologist or PAT.
On-site concentrations should be considered if individuals reside on-site
(e.g., military bases, universities).
3-25
-------
4Gant=~must—-pp0v-ide—the! iriodeling results and risk
analysis for review. The model output should include a full printout of
the input data or the full input should be appended to the results.
The permit writer should then send the results of the modeling to the
Regional Meteorologist or PAt for review of the results for conformity to
the modeling plan and to determine if the results are valid. The permit
writer should then use the dispersion coefficient and THC emission rate
(and metals emission rate) to determine if the MEI risk exceeds 10" .
• If the risk is acceptable, establish CO and THC limits as described
in Step D.
If the risk is not acceptable, it must be reduced (e.g., reduce THC
emission rates by improving combustion performance).
Step D; To establish the CO and THC permit limits, see Section 3.2.2.2,
Section 3.2.3, Section 3.2.4, and Section 3.2.5.
3-26
-------
RATIONALE FOR PIC CONTROLS
Current RCRA regulations control organic emissions from hazardous waste
incinerators by the destruction and removal efficiency (ORE) standard of 40
CFR 264.242(a). This standard limits stack emissions of principal organic
hazardous constituents (POHCs) to 0.01 percent (0.0001 percent for dioxin-
containing waste) of the quantity of the POHC in the waste. This standard,
however, does not limit the emission rate of products of incomplete combustion
(PICs). Hazardous waste combustion, like all combustion processes, always
produces PICs ~ partially destroyed constituents in the waste or organic
compounds synthesized in the hot combustion gas from the "soup" of organic
compounds available. The combustion of halogenated hydrocarbons to produce
only carbon dioxide, water, and acid gases (e.g., HC1, HBr) is theoretical and
could occur only under perfect conditions. Real-world combustion systems —
incinerators, fossil-fuel fired steam generators, diesel engines, etc. —
virtually always produce PICs. PIC emissions from hazardous waste combustion
include relatively high concentrations of nontoxic compounds like methane and
low concentrations of compounds that can be as or more toxic than the
constituents in the waste.
The health risk posed by PIC emissions depends on the quantity and toxicity of
the individual components of the emissions, and the resultant ambient levels
to which persons are exposed. Estimates of risk to public health resulting
from PICs, based on available emissions data, indicate that PIC emissions from
well designed and operated incinerators do not pose significant health
risks. However, only limited information about PICs is available. PIC
emissions are composed of thousands of different compounds, some of which are
in very minute quantities and cannot be detected and quantified without very
elaborate and expensive sampling and analytical (S&A) techniques. Such
elaborate S&A work is not feasible in trial burns for permitting purposes and
could only be attempted in research tests. In addition, reliable S&A
procedures simply do not exist for some types of PICs (e.g., water-soluble
compounds). The most comprehensive analysis of PIC emissions from a hazardous
4-1
-------
waste incinera'tor identified and quantified only approximately 70 percent of
organic emissions. Typical research-oriented field tests identify a much
lower fraction -- from 1 to 60 percent. Even if all the organic compounds
emitted could be quantified, there are inadequate health effects data
available to assess the resultant health risk. The Agency believes, that, due
to the above limitations, additional testing will not, in the foreseeable
future, be able to. prove quantitatively whether PICs do or do not pose
unacceptable health risk.
Considering the uncertainties about PIC emissions and their potential risk to
public health, the Agency believes that it is prudent to take reasonable steps
to minimize PIC emissions. Consequently, the Agency recommends that
incinerators operate at high combustion efficiency as evidence by low CO
levels. In cases where CO levels exceed the recommended de minimis limit of
100 ppmv, the Agency has provided a waiver provision where higher CO levels
are allowed when the applicant shows either: (1) that total hydrocarbon (THC)
emissions do not exceed 20 ppmv; or (2) that THC emissions do not pose
unacceptable health risk using prescribed risk assessment procedures.
4.1 USE OF CO LIMITS TO ENSURE GOOD COMBUSTION CONDITIONS
Generally accepted combustion theory holds that low CO flue gas levels are
indicative of an incinerator (or any combustion device) operating at high
combustion efficiency. In fact, combustion efficiency correlates with the CO
level by definition (see Section 4.2). Operating at high combustion
efficiency helps ensure minimum emissions of unburned (or incompletely burned)
organic compounds". In a simplified view of combustion of hazardous waste, the
first stage is Immediate thermal decomposition of the POHCs in the flame to
form other, usually smaller, compounds, also referred to as PICs. These PICs
are also rapidly decomposed to form CO.
A demonstration that THC levels do not exceed 20 ppmv is the preferred
approach to allow CO levels greater than 100 ppmv given the serious
limitations of the risk assessment methodology. See Section 4.8 of the
text.
4-2
-------
The second stage of combustion- involves the oxidation of CO to C02 (carbon
dioxide). The CO to C02 step is the slowest (rate controlling) step in the
"combustion-process because CO is considered to be more thermally stable
(difficult to oxidize) than other intermediate products of combustion of
hazardous waste constituents. Since fuel is continuously being fired, both
combustion stages are occurring simultaneously.
Using this view of waste combustion, the "destruction" of a POHC does not, and
the destruction of PICs may not correlate with flue gas CO levels. As
discussed below, some data, in fact, show a slight apparent correlation
between CO and chlorinated PICs, and a fair correlation between CO and total
hydrocarbons (THC). Although the question of whether there is a correlation
between CO and toxic hydrocarbons and THC may be open to discussion, the data
does indicate a clear relationship between these parameters. When CO is low,
concentrations of both toxic hydrocarbons and THC are low. This is consistent
with the thinking that low CO is an indicator of the status of the CO and C02
conversion process — the last, rate-limiting process. Since oxidation of CO
to C02 occurs after destruction of the waste constituents and resulting PICs,
the absence of CO is a useful indication of PIC (and POHC) destruction. Note,
however, that the presence of high levels of CO may not indicate the presence
of high levels of toxic hydrocarbons or THC. Thus, the Agency considers CO to
be a conservative indicator of PIC emissions.
The presence of high levels of CO in the flue gas is a useful indication of
inefficient combustion and, at some elevated CO level, an indication of the
failure of the PIC (and POHC) destruction process. Because the Agency does
not know the precise CO level that is indicative of significant failure of the
PIC destruction process, we recommend limiting CO levels to levels indicative
of high combustion efficiency. In fact, the critical CO level may be
dependent on site-specific and event-specific factors (e.g., fuel type, air-
to-fuel ratios, rate and extend of change of these and other factors that
affect combustion efficiency). Limiting CO levels to minimize PIC emissions
1s reasonable because: (1) it is a widely practiced approach to monitor and
improve combustion efficiency; and (2) most well-designed and operated
incinerators can easily be operated in conformance with the Tier I CO limit of
100 ppmv.
4-3
-------
4.2 SUPPORTING INFORMATION ON CO AS A SURROGATE FOR PICs
Several "types of information are available which indicate CO emissions may
relate to PIC emissions. Combustion efficiency is directly related to CO by
the following equation:
percent CO
Combustion Efficiency (CE) = z x (100) (8)
percent C02 + percent CO
"s
CE has been used as a measure of completeness of combustion. EPA's
regulations for Incineration of PCBs require that combustion efficiency be
maintained above 99.9 percent. As combustion becomes less efficient or less
complete, at some point, the emission of total orgam'cs will increase and
smoke will eventually result. It is probable that some quantity of toxic
organic compounds will be present in these organic emissions. Thus, CE or CO
levels provide an Indication of the potential for total organic emissions and
possibly toxic PICs. Data are not available, however, to correlate
quantitatively these variables with PICs in combustion processes. - —• -
Several studies have been conducted to evaluate CO monitoring as a measure of
performance of hazardous waste combustion (References 3 through 9). Though
correlations with destruction efficiency of POHCs have not been found, the
data from these studies generally show that as combustion conditions
deteriorate, both CO and total hydrocarbon (THC) emissions increase. These
data support the relation between CO and increased organic emissions discussed
above. In one of these studies, an attempt was made to correlate the
concentrations of CO and THC with the concentrations of four common PICs
(benzene, toluene, carbon tetrachloride, and trichloroethylene) in stack gases
of full scale incinerators (Reference 3). Figure 2 shows the data for CO and
THC versus benzene, one of the most common PICs. There is considerable
scatter in the data indicating that parameters other than CO and THC affect
the benzene levels; However, the data suggest that, when CO and THC levels
are low, benzene (PIC) levels are low. The data do not suggest, however,
that when CO and THC levels are high, benzene levels are always high.
Therefore, CO is a conservative indicator. Similar trends were observed for
4-4
-------
•— s
L.
.*.
C
c —,
O M
11
c 3
V O
o .c
0~
41
N
1
2.6-
2.4-
2.2-
2.0-
1.8-
1.6-
1.4-
1.2-
1.0-
0.8-J
0.6-
0.4-
0.24
I
0-f
0
•}
• • • • ;" i
A Total Hydrocarbon Concentration (ppm) -j j
° I f , ' * -!| 90 1
' '• ' I
\ 1
' ' i '
' J
•
A •
/— X40THC Point, )
r+— -(44 CO Points)
y A» • A A
fL^» * . A . ' * _
mm 9 . • — _ ^ A •
200 400 ' 600
j;
I ';
i '
i
-
i •
•Carbon Monoxide Concentration (ppm)
Figure 2. CO and THC vs. benzene concentration.
Source: Reference 3.
-------
-r'
toluene and carbon tetrachloride, but not for trichloroethylene. In another
of these studies, similar results were observed for chlorobenzene and"
methylene chloride (Reference 4).
A third study provided data on the relationship between CO and THC and the
PICs, vinyl chloride and methyl chloride (Reference 3). Figure 3 and Figure 4
are plots of CO and THC concentrations versus vinyl chloride and methyl
chloride concentrations, respectively. Both figures visually display
increasing PIC concentration levels as CO and THC levels increase. Nine other
PICs were examined in this study; however, their concentrations were much
lower than the concentrations for vinyl chloride and methyl chloride, and no
clear trends relative to CO or THC were evident.
4.3 ALTERNATE FORMATS FOR COMPLIANCE WITH THE CO LIMITS
Three aspects of the format for monitoring compliance with a CO limit were
considered in the development of this guidance. They are the method of
specifying a level, the correction to a specified oxygen concentration, and a
correction for moisture. Each of these is discussed below.
4.3.1 Methods for Specifying CO Limits
Three alternative methods to specify a level for the limit are:
* A level never to be exceeded.
• A level that can be exceeded only for a specified time.
• An average level over a specified time that is never to be exceeded.
The first method requires immediate shutdown of an incinerator when the limit
is exceeded, regardless of how long the CO levels remain high. Short-term CO
excursions or peaks (a few minutes duration) are typical of incinerator
operation and can occur during routine operations; e.g., when a burner is
adjusted. It is possible that shut down and startup of the incinerator may
4-6
-------
AToral Hydrocarbon Concentration (ppm)
50 100
^^9
-a
Q.
a.
*— •
e
«
0
v
|
-S
5
u
•x'
>""
•
1
e
o
I
•s
o
2
o
U
a
2
U
"x
1
; IOC
9C
80
70
60
50
40
30
20
10
0
• ' L «:
-
_
A •
...
• A
- : " -
_
" - • A
- • A
A ' * • A
•AA «A
• A
• "S-^ — i f 1 < • ' . r 1.1.
0 400 800 1200 1600 2000 2400 2800
•Carbon Monoxide Concentration (ppm)
Figure 3. CO and THC versus vinyl chloride concentrations.
Source: Reference 4.
A Total Hydrocarbon Concentration (ppm)
5000
* •
4000
3000
2000
1000
o
3 50 100 150 200 250 300 350 40
1 i ' I « I »- i • — i r j 1 1 T
• .
•A • A
• A
m
• * A
• A
0
24002800"
•Carbon Monoxide Concentration (ppm)
Figure 4. CO and THC versus methyl chloride concentrations
Source: Reference 4.
4-7
-------
cause higher emissions than those associated with these short term CO peaks.
Therefore, a never-to-exceed limit would impede .incineration operations-while.,
providing little reduction in health risk.
The second method, allowing the CO limit to exceed the de minimis limit for a
specified cumulative time within a determined time frame (e.g., x minutes in
an hour), solves the problem associated with the first method. Incinerators
would not be shut down by a single CO peak of high intensity yet they would be
restricted from operation with several short interval CO* peaks, or a single
long duration peak.
The third method, allowing the CO level never to exceed an average level
determined over a specified time, also avoids the problem of shutting off the
waste feed each time an instantaneous CO peak occurs. A time-weighted average
value (i.e., integrated area under the CO curve for a given time period
divided by that time period) also provides a direct quantitative measure of
mass emissions of CO, which is the recommended surrogate parameter for PIC
emissions. For this- reason, the use of a rolling average is the preferred
format. A combination of the first and second method, structured to ensure
that mass CO emissions per unit time are equivalent to the time-weighted
average method, is offered as an alternative format. This format is termed
the cumulative time-above-a-level format, and is coupled with an instantaneous
CO limit.
The CO monitoring system needed for the time-weighted average format requires
Instrumentation that continuously calculates hourly rolling averages and that
continuously adjusts the oxygen correction factor to record CO levels based on
7 percent oxygen. The instrumentation costs of such a system consisting of a
continuous oxygen monitor, a data logger, and microprocessor could be up to
$91,000 and would require increased sophistication and operating costs over
simpler systems. Use of the alternative CO format will reduce the cost of
instrumentation. Compliance requires instrumentation that cumulates time
above the de minimis limit in every clock hour, at the end of which it is
recalibrated (manually or electronically) to restart afresh. Oxygen, also,
would not have to be measured continuously in this format; instead, an
appropriate oxygen correction, determined during the trial burn, is
4-8
-------
We^ysffiiTlir^^ to the CO readings. Subsequently,
oxygen correction values would be determined annually or at more frequent
intervals specified in the permit.
4.3.2 Rationale for Oxygen and Humidity Corrections
The CO (and THC) limits specified for either format are on: a dry gas basis and
corrected to 7 percent oxygen. The oxygen correction normalizes the CO data
to a common base, recognizing the variation among the different technologies
as well as modes of operation using different quantities of excess air. In-
system leakage, the size of the facility and the type of waste feed are other
factors that cause oxygen concentration to vary widely in incinerator flue
gases. Seven percent oxygen was selected as the reference oxygen level
because it is in the middle of the range of normal oxygen levels for hazardous
waste incinerators and it also is the reference level for the existing
partlculate standard under Section 264.343(c).
The correction for humidity normalizes the CO data from the different types of
CO monitors (e.g., extractive vs. in-situ). Evaluation of possible variation
of stack gas oxygen and moisture levels indicates that the above two
corrections, when applied, could change the measured CO levels by a factor of
two in some cases.
4.4 RATIONALE FOR RECOMMENDING THE TIME-WEIGHTED AVERAGE CO LEVEL FOR TIER II
PERMITS
Under the Tier II permit approach, the Agency recommends that the CO limit
should be based on the time-weighted average CO level during the trial burn.
This 1s consistent with ORE testing where organic emissions are sampled over a
trial burn run and evaluated as an average emission rate. In addition, the
Agency does not believe it would be appropriate to allow an incinerator to
operate over the life of the permit at elevated (i.e., higher than 100 ppmv)
CO levels based on the highest average level that occurred during one hour of
the trial burn.
4-9
-------
4.5 EQUIVALENCE OF CO MASS EMISSIONS UNDER CO FORMATS"
Under Format B for establishing CO limits, the three parameters — highest
Instantaneous CO peak, base CO limit, and time the base limit can be exceeded
in any hour — should be limited so that the mass emission rate of CO under
Format B is the same as would be allowed under Format A. This equivalence of
CO mass emissions is illustrated in Figures 5 and 6. _ Figure 5 illustrates a
hypothetical continuous CO emission trace recorded on a strip chart during a
ORE trial burn hour. Figure 6 illustrates the maximum CO emissions permitted
under both format options. The rolling average permit format is based on
compliance with a single hourly rolling average limit, COHA. This value is
the time average for the entire trial burn (100 ppmv for Tier I). The time-
weighted average CO is computed by integration of the area under the CO
emissions curve averaged for the duration of the trial burn period. Data
loggers interfaced with the CO monitor can easily compute hourly rolling
averages such that with any instantaneous CO reading the time-average CO is
computed over the previous one hour period. The alternative permit format
requires that the following three limits be specified:
• The highest instantaneous CO peak (C0_) allowed,
• A base CO limit (C060_t) which can be exceeded only for a cumulative time
in any clock hour, and
• A cumulative time (t) in any clock hour that the limit C060_t can be
exceeded.
The cumulative time t in excess of the C060_t limit essentially allows the
incinerator to operate continuously at C0p during that specified time in each
clock hour.
It is apparent that in order to ensure compliance with the hourly rolling
average CO mass emission (COHA x 60) under any operating condition allowed by
the alternate format, the following expression must be satisfied:
4-10
-------
CO
(pp.m)
Time (minutes) t + 60
Figure 9. Trial Burn Continuous CO Eaiaaiona Tract
Source: MRI.
CO
(ppm)
COHA
C060-t
CO
Time (minutes)
Figure §. Alternate Perait Fonrat
t + 60
Source: MRI.
4-11
-------
60 COHA~= ^eOXOeTj-^+^GOp — -CO^j^^^^^t— —^.^^^^
Once CO- and t have been defined the base CO value, C060_t, can be calculated
from equation 1 as follows:
C060.t * COHA - t (COp - COHA) (10)
60-t :
For Her I, equation 2 becomes:
C060_t 3 100 - t '(COp - 100) (11)
60-t
based on compliance with the 100 ppmv hourly rolling average.
This equation which equates the shaded and cross-etched areas 1n Figure 6, can
be satisfied by an Infinite number of C0pt t, and C060_t permit limit
combinations. For example, the higher the permitted peak or time in excess of
the base CO limit the lower the C060_t limit has to be. The converse is also
true. Therefore, there are several options available to the permit writer.
However, because there are an infinite number of combinations for C0p, t, and
C06o_t ^at w^^ satisfy the equivalence of the CO mass emissions during the
trial burn with the permitted levels, no one combination enables the
incinerator to comply with all available scenarios. This is an important
consideration, because once these limits are selected, compliance is based on
meeting each limit separately rather than the corresponding rolling average
from which these limits were determined.
The following paragraphs give some examples of the permit options available.
Selection of the most appropriate option is left to the permit writer's
judgment after considering the applicant's request. Some guidance on making
this selection is also discussed.
4-12
-------
4.5.1 Constraining Permitted-Instantaneous Peak CO To Trial Burn Level
Under-this_, option the highest instantaneous peak recorded during the trial
burn is selected as the highest permitted peak CO (CCL). In the hypothetical
case shown in Figure 5, this peak corresponds to C0pl. The time t can be set
arbitrarily or based on an equivalent tjrial burn time. Selection of an
arbitrary time can be established according to criteria specified in Section
4.5.2. Calculation of an equivalent time can be performed using the following
approach. Because the CO trace is likely to have more than one peak,
specifying only the highest peak would require a reduced cumulative time than
that recorded during the trial burn. That is:
tiCOpl, + t2COp2 + t3COp3 + t^COp,, = tCOpl or
*= *' * ^r X tlCO>1 (12)
Where tt is the time corresponding to the highest peak and n is the total
number of peaks. This expression assumes that all peaks are similar in shape
(e.g., triangles as shown in Figure 5). Once C0p and t have been defined the
base CO value, C060_t, can be calculated from equation 10.
4.5.2 Constraining Time (t) to a Specific Limit
This option defines an arbitrary time (e.g., 6 minutes) where the level C060_t
can be exceeded. Selection of an arbitrary time can be based on the following
criteria:
• A maximum CO mass emission (ppmv-minutes) allowed in any one hour during
nonsteady combustion conditions (e.g., 3,000 ppmv-minutes)
• Site-specific data on typical CO cycles and duration of cycles caused by
"transient puffs" in primary combustion chambers (e.g., rotary kilns).
• Site-specific data on contaminated solid waste feedrate.
Applicant's desired operating flexibility.
4-13
-------
The highest permitted G0~peak-(C0p) can be set based on the highest ORE trial
burn peak or based on equivalence of peak areas using- equation 12.
For example, for fixed time t = 6 minutes and C0_ equal to the actual trial
burn maximum recorded, the base limit is calculated as follows:
- COHA - C0p " COHA
However, this option can result in C054_min levels that are excessively
stringent because a peak of arbitrary duration is assumed. Thus, this option
is most appropriate when the highest trial burn peak is less than 20 times
COHA. If the permitted C0_ is based on equivalence of CO mass emissions
during CO excursions, the following equation applies: .
n-1
CO
P 1
Where t is the arbitrary set time (e.g., 6 minutes)
Thus the peak CO limit can be higher or lower than the actual trial burn
maximum. Again the equation assumes that all peaks are identical in shape.
For Tier I, the value for C060_t is calculated with equation 11.
4.5.3 Constraining The Base CO Limit To A Percent Of The Time Average
The last option available is based on establishing an arbitrary limit for the
C06o_t level and adjusting the height and duration of the allowed peak so that
the total mass emissions are not exceeded past the trial burn recorded
level. For example, if
C060.t = 0.90(COHA)
the remaining area, 0.1 COHA, must be equated to the peak area or
4-14
-------
- t (co - CO) ...... ^ -
O.IO.COHA - HA
(60-t)
Where t or C0_ can be equated to trial burn levels using,,either equations 12
or 14.
4.6 DERIVATION OF THE TIER I CO LIMIT
The Tier I CO limit of 100 ppmv was selected for a number of reasons: (1) it
is within the range of CO levels (i.e., 0-200 ppmv) that represent high
combustion efficiency; (2) available field test data indicate that PICs are
not emitted at levels that pose unacceptable risks when CO does not exceed
100 ppmv; (3) the 100 ppmv level is consistent with the combustion efficiency
of 99.9 percent currently required by EPA's PCS incineration regulations under
the Toxic Substance Control Act (TSCA); (4) it is the CO limit proposed for
boilers and furnaces burning hazardous waste (see 52 FR 16997, May 6, 1987);
and (5) it is a level that the majority of well designed and operated
incinerators can meet. These reasons are discussed below.
The current TSCA rule for the incineration of PCB-laden wastes requires a
minimum combustion efficiency (CE) of 99.9 percent. Combustion efficiency of
99.9 percent translates to CO emissions levels of 80 to 125 ppmv corrected to
7 percent 02, depending on the fuel C/H ratio. The intent of the PCS
combustion efficiency rule is to minimize emissions of potentially toxic
organics. Therefore, the proposed 100 ppmv CO level for hazardous waste
destruction is consistent with the intent of the regulations governing the
incineration of PCB wastes.
\
CO emission data from hazardous waste incineration research and trial burn
tests also confirm the relationship between CE greater than 99.9 percent and
CO levels less than 100 ppmv. The combustion efficiencies in all cases where
data were available were calculated to exceed 99.9 percent, except for the
test runs where CO exceeded the proposed CO limit. Two other data sets were
used to evaluate the recommended level.
4-15
-------
The first data set was from the research tests of eight incinerators cited
earlier (Reference 7). These data are displayed in Table 4 and show that
these incinerators easily complied with the recommended 100 ppmv 'limit with
two exceptions. One run at Site 2 had a maximum hourly average of 120 ppmv.
Information was not available to evaluate why CO levels were higher for this
test run; however, the three other runs at this site all showed easy Table 4
compliance with the recommended limits. All three runs at Site 3 showed CO
levels clearly higher than the recommended limit. This incinerator operated
with a relatively high baseline CO level and also exhibited frequent CO spikes
as drums of volatile waste were fed to the rotary kiln. It is likely that
either the operating conditions for this Incinerator would have to be modified
to comply with the recommended limit or the facility would be permitted under
Tier II 1f THC levels do not exceed 20 ppmv. Table 4 also shows corrected CO
values and calculated combustion efficiencies. The combustion efficiencies
are all above 99.9 percent, except the test runs where CO exceeded the
recommended CO limits.
The second data set is displayed in Table 5. It consists of available CO data
compiled from results of trial burns conducted during permitting of hazardous
waste Incinerators (Reference 10). Sufficient information was not available
to calculate maximum 60-min averages or to correct for 02 concentration in a
few cases. The values shown are averages for each test run (unspecified
length of run) or the range of data in one case. These data support that
Tests 1, 3, 5, 6, 8, 11, and 13 may not have complied with the recommended
Tier I limit. 'Thus, it is estimated that up to 50 percent of these
Incinerators may have failed the recommended Tier I CO limit. Information was
not available to evaluate the reasons CO levels were higher for some of these
incinerators versus others. Reduction of these CO levels may involve
relatively simple changes in some cases. Significant changes in operating
conditions may be required in other cases to meet the Tier I CO limit.
In general, the data reviewed suggest that most hazardous waste incinerators
can achieve the recommended Tier I CO .limit. However, a number of
incinerators will likely have to modify their operating conditions or obtain
CO limits by using Tier II requirements.
4-16
-------
TABLE 4. CO DATA"FROM RESEARCH TESTS
Site
Plant B
American
Cyanamid
DuPont
Mitchell
Ross
TWI
Upjohn
Zapata
Run
1
2
3
1
2
4
5
1
2
3
1
2
3
4
1
2
3
1
2
3
4
1
2
3
4
2
3
4
S-CO, ppm
Average
23
< 5
10
9
96
28
38
530
330
680
< 5
< 5
19
< 5
8
14
7
7
< 5
< 5
< 5
11
12
12
10
36
7
15
e 7% o2
Range
14-37
< 5
< 5-33
2-65
14-570
11-45
13-400
25-2,000
25-2,100
40-2,500
< 5-7
< 5-7
< 5-700
< 5-22
< 5-27
7-25
< 5-15
< 5-120
< 5-23
< 5
< 5-6
9-14
10-14
11-13
9-11
< 5-620
5-10
4-33
Maximum rolling
average CO^ ppjn
60 rain
27
< 5
14 -
U-
120C
a
41
650°
510C
910C
< 5
< 5
40
< 5
9
17
9
13
< 5
< 5
< 5
13
12
12
10
68
8
22
•~sef!lj!-?!tt,-&i*-fr?r. •;.-.' . . ij^4~ —-f
THC,
Average
-. . .-. * : —
< 1
< 1
< 1
< r-
< i
< i
1.0
75.9
47.6
58.1
< 1
< 1
0.6
< 1
0.9
1.0
2.5
1.9
1.7
0.8
8.9
6.0
3.9
1.9
< 1
< 1
ppm
Range
< 1
< 1
< 1-1.9
< 1-1.6
< 1
< 1
< 1-1.1
45.0-140
36.2-85.8
39.4-86.3
< 1
< 1
0.2-1.8
< 1
0.8-2.3
< 1-2.3
2.0-2.9
1.7-2.1
1.3-2.2
0.3-2.1
7.1-11.9
4.5-9.0
3.1-6.2
< 1-40.9
< 1
< 1-2.9
Combustion
efficiency
99.96
> 99.99
99.99
99.87
99.87
99.96
99.94
99.43
99.64
99.25
> 99.99
> 99.99
99.97
> 99.99
99.99
99.98
99.99
99.99
> 99.99
> 99.99
> 99.99
99.99
99.99
99.99
99.99
b
b
b
a A 60-min average could not be calculated due to a short sampling run.
- Combustion efficiency could not be calculated because no C02 data
were available.
c Exceeds recommended limits.
Source: Reference 7.
4-17
-------
TABLE 5. CO DATA FROM TRIAL BURNS
Average CO for each test run, ppn f 1% 0,
*
Test 1 23456789 10
11 12
1 > 420 180 230 1 0.4 - 2 8 11
!
213<7<7<7<7<7
3 990 1,300 190 320 1,000 1,400 57 230
I1
4<1 1 < I < 1 < 1
5 190 62 58 27 24 22 18
6 140 68 25
•P*
iL . 7 "0 - 0
00
8 16a 58° l,000a
9 18 19 27 22 35 22 15
10 19 4 5 11
11 250 710 870
12 80 50 49 37
13 30-2,000a'b 40-2,000a»b 50-2,000a'b 40-2,000a'b 50-270a'b 0-1,800a'b 250-500a'b 10-800a'b30-2,000a'b 30-2,000a'b
14 17 16 15 17 16 16 20 18 17 15 16 16
a Uncorrected ppm, sufficient data on 0- was not reported.
b Range of data; average value was not reported.
Source: Reference 10.
-------
This manual recommends use of Tier I or Tier II CO,:-limits in all permits
versus current practice of case-by-case limits based soley .otv_.t.h'e levels that
occurred during trial burn. The Tier I limit should be .specified in the"
permit even when the CO levels during the trial burn are lower. For example,
if the results of a trial burn showed the incinerator operated at a maximum
1-hour average of 20 ppmv, the permit limit should still allow a 1-hour
average of 100 ppmv. This raises the issue of whether there would be a
significant deterioration in ORE between trial burn conditions and operation
at 100 ppmv CO. (As discussed above, available data indicated that PIC
emissions do not pose significant risk when CO levels are 100 ppmv or less).
This issue was carefully considered and the recommendation is based on three
reasons. First, ORE will not be reduced below the levels specified in
§264.343(a)(l) when £0 levels are increased to 100 ppmv (see discussion
below). Second, many incinerators run very efficiently during a trial burn
and have CO levels less than 10 ppmv. It may not be possible to achieve that
high degree of efficiency on a consistent basis. Specifying such low CO values
may result in numerous hazardous waste feed cutoffs due to CO exceedances that
inhibit the operator's ability to efficiently operate the facility without any
environmental benefit. Third, the emission of PICs from incinerators has not
been shown to increase linearly at such low CO levels. In fact, the trial
.burn data indicate that total organic emissions are consistently low when CO
emission levels are less than 100 ppmv.
Two studies were identified that provide data related to this issue. The
first study generated data from combustion of a 12-component mixture in a
bench scale facility (Reference 11). Results from bench scale operations do
not necessarily represent full scale incineration and should be evaluated with
caution. However, carbon monoxide levels 1n this study ranged from 15 to
522 ppmv without a significant correlation on the destruction efficiency for
the compounds investigated.
The second study was conducted on a pilot scale combustor (Reference 8). Test
runs were conducted with average CO concentrations ranging from 30 to
700 ppmv. When the CO concentration was less than 220 ppmv, no apparent
decrease in the destruction efficiency was noticed. Test runs with CO
4-19
-------
concentrations greater than 220 ppmv showed signs of a decrease in destruction
efficiency. K;_:_
These studies indicate that no measurable change in ORE occurs for CO levels
up to the recommended Tier I permit limit. A decrease in ORE may occur,
however, at higher CO levels.
4.7 DERIVATION OF TIER II RISK-BASED THC LIMITS
Facilities unable to meet specific CO limits prescribed in Tier I can use the
Tier II alternative. Tier II establishes CO and THC limits based on a site-
specific risk assessment for THC emissions during jthe trial burn. In lieu of
conducting site-specific dispersion modeling to determine whether THC
emissions may pose unacceptable health risk, applicants and permit writers may
use conservative THC Screening Limits (shown in Table 1). The Screening
Limits are based upon an acceptable risk to the MEI of 1 x 10~5 using
reasonable, worst-case dispersion analyses and a conservative potency for
inhalation health effects of PICs historically identified in stack gases from
hazardous waste combustion. See Appendices B and C.
Detailed discussion of the development of the dispersion coefficients is
presented in the Metals/HCl Guidance Document (Reference 2). In brief,
calculation of these conservative dispersion coefficients was based on the
evaluation of several factors known to influence the relationships between
releases (emissions) and ground level concentration, including: (1) the rate
of emissions; (2) the release specifications of selected facilities used in
the analysis (especially stack elevation, and combustion gas velocity and
temperature, which together define the facility's "effective stack height");
(3) local terrain; and (4) local meteorology. From a survey of 154 existing
facilities a sample of 24 facilities with relatively low stack heights,
divided among the three basic categories of terrain types (flat, "rolling",
and complex) was used in the dispersion modeling analysis. In addition, 11
generic (hypothetical) stacks that spanned the range of stack release
parameters (including a stack lower than good engineering practices) were
assumed to be located at each of the 24 sites. EPA recommended dispersion
4-20
-------
predict the maximum annual average ground level dispersion
coefficient (ug/m3/g/s) for each stack. The highest conservative dispersion
coefficients (yg/m3/g/s.) were identified for any terrain-adjusted effective
stack height within the range of those modeled (i.e., 4 m to 120 m). The
analysis determined that there was no significant difference in dispersion
coefficients (under the severe conditions modeled) between flat and. rolling
terrain. Thus, those terrain types were merged together and termed noncomplex
terrain. In addition, there was no significant difference in coefficients
and, thus, Screening Limits for urban versus land use in complex terrain.
Thus coefficients were not distinguished between land use classifications in
complex terrain.
Appendix B lists the conservative dispersion coefficients used to calculate
the THC Screening Limits and provides more detailed information on derivation
of the conservative unit risk.
4.8 DERIVATION OF TIER II TECHNOLOGY-BASED THC LIMIT
The Agency is concerned that the risk-based approach to determine whether THC
emissions may pose unacceptable risk may have serious limitations. 'These
concerns are discussed below. In addition, the risk-based approach could
allow incinerators to operate at very high THC levels indicative of upset
combustion conditions. As shown 1n Appendix E, THC levels as high as 1780
ppmv could be allowed by the risk assessment methodology. Given these
concerns, the Agency believes that it is preferable to waive the 100 ppmv CO
limit only when the good operating practice-based THC limit of 20 ppmv is not
exceeded. We believe that the development of a risk-based approach is a step
in the right direction, however, and that the approach described in Section
3.2.2.2 and 4.7 and Appendices B and C is the best available risk assessment
approach. Accordingly, we have used the risk methodology, notwithstanding its
limitations, to demonstrate .that a THC limit of 20 pprav appears to be
protective of public health using reasonable, worst-case scenarios.
-------
4.8.1 Limitations of Risk Methodology
The Agency's primary concern with the risk methodology is that, given the
limitations of our knowledge about the types and concentrations of organic
compounds emitted from the combustion of various wastes under various
combustion conditions, and the fraction of organic emissions actually detected
by a flame ionization detector (FID), our data base may be too limited to
conduct s1te.-specific, quantitative risk assessments for PIC emissions. The
uncertainties of the risk methodology may be compounded when it is applied to
devices operated under poor combustion conditions ~ when CO exceeds 100 ppmv.
The vast majority of our data on the types and concentrations of PIC emissions
from incinerators, boilers, and industrial furnaces were obtained during test
burns when the devices were operated under good combustion conditions. CO
levels were often well below 50 ppmv. Under Tier II applications, CO levels
can be 100 to 1,000 ppmv or higher (there is no upper limit on CO). The
concern is that we do not know whether the types and concentrations of PICs at
these elevated CO levels, indicative of combustion upset conditions, are
similar to the types and concentrations of PICs in our data base. It could be
hypothesized that as combustion conditions deteriorate, the ratio of semi- and
nonvolatile compounds to volatile compounds may increase. If so, this could
have serious impacts on the proposed risk assessment methodology. First, the
proposed generic unit risk value for THC may be under-stated when applied to
THC emitted under poor combustion conditions. This is because semi- and
nonvolatile compounds comprise only 1% of the mass of THC in our data base but
pose BQ% of the cancer risk. Thus, if the fraction of semi- and nonvolatile
compounds increases under poor combustion conditions, the cancer risk may also
increase. To put this concern in perspective, the Agency notes, that the
proposed THC unit risk value is 1 x 10"5 m3/ug. This unit risk is 100 times
greater (i.e., more potent) than the unit risk for the quantified PICs with
the lowest unit risk (e.g., tetrachloroethylene), but 1,000 times lower then
the unit risk for PICs such as dibenzoanthracene, and 10,000 to 1,000,000
times lower than the unit risk for various chlorinated dioxins and furans.
Second, if the fraction of semi- and nonvolatile THC increases under poor
combustion conditions, the fraction of THC in the vapor phase when entering
4-22
-------
the THC detector may be lower than the 75% assumed when operating^undej^goad,-
combustion conditions. If so, the correction factor for the condensed3
organics would be greater than the 1.33 factor proposed (see Appendix C).
The Agency is currently conducting emissions testing to improve the data base
in support of the risk assessment methodology. The Agency is concerned,
however, that the testing that is underway and planned may not provide
adequate information to fully address all the issues. In addition, the Agency
is concerned that its stack sampling and analysis procedures and its health
effects data base are not adequate to satisfactorily characterize the health
effects posed by PICs emitted under poor combustion conditions.
A final concern with the risk assessment methodology is that it does not
consider health impacts resulting from indirect exposure. The methodology
considered human health impacts only from direct inhalation. Indirect
exposure via uptake through the food chain, for example, has not been
considered because the Agency has not yet developed procedures for quantifying
indirect exposure impacts for purposes of establishing regulatory emission
limits.
4.8.2 Basis for THC Limit of 20 ppmv
The Agency has selected a THC limit of 20 ppmv as representative of a THC
level distinguishing between good and poor combustion conditions. The value
is within the range of values reported in the Agency's data base for hazardous
waste incinerators and boilers and industrial furnaces burning hazardous waste
under good combustion conditions, and the level appears to be protective of
human health based on risk assessments for 30 incinerators using the risk
assessment methodology described in Sections 3.2.2.2 and 4.7. See Appendix E.
Although the available data indicate that a few devices may not be able to
meet a THC limit of 20 ppmv, the data clearly indicate that the vast majority
of devices can meet a 20 ppmv limit. It appears that many hazardous waste
incinerators can typically achieve THC levels of 5 to 10 ppmv when operating
generally at low CO levels. When incinerators emit higher THC levels, CO
levels typically exceed 100 ppmv, indicative of poor combustion conditions.
4-23
-------
d.gta-_^e_gn_1ndustr1 al boi lers^ap4i^e}LepJLJ^llBS..^.tn(itcates.-that good
.operating practice THC levels are generally on the order of 5 to 20 ppmv.
Given that the Agency would prefer to set a single THC limit for all regulated
devices, a limit of 20 ppmv appears to be appropriate considering that the
human health risk at that level appears to be well within acceptable levels.
The determine whether a 20 ppmv THC limit would be protective, the Agency
assumed that 30 incinerators in the data base were emitting THC at that
concentration and used site-specific, regulatory dispersion modeling (i.e.,
consistent with EPA's Guideline on Air Quality Models) to predict ambient THC
concentrations to which the maximum exposed individual would be exposed. The
Agency then assumed that the THC had unit risk value of 1 x 10~5 m3/ug» the
value used in the risk-based methodology. The MEI risk levels ranged from
10"6 to 10~7, well within the 10~5 risk-level considered acceptable for
purposes of this rule.
4.9 METHODS FOR MONITORING THC
t
The Agency recommends use of a flame ionization detector (FID) to monitor
THC. Several hazardous waste incinerators have been equipped with FIDs to
continuously monitor THC. Some of these instruments have been in operation
for up to 10 years. In addition, FIDs are routinely used during trial burns
to monitor THC.
Two variations of FID methods are in use: heated and unheated. With the
heated method, sampling lines and the FID itself are heated to 150 C or higher
to maximize detection of organic emissions by minimizing condensation of
organic compounds. This method is described in Appendix A.3. With the
unheated system, neither the sampling lines nor the instrument are heated.
The Agency has obtained data on THC emissions from various types of
boilers burning various types of fossil fuels (not hazardous waste),
however, that indicate that some.boilers may not be able to meet a 20 ppm
THC limit. See Appendix E. The Agency is reviewing that data and
obtaining additional information to determine if an alternative limit may
be appropriate for boilers.
4-24
-------
—^__—"=», '
Condensate traps- are used tcP^cpjnditioln1' "^the''
condensed moisture (and organic compounds). With. thisT unheated~rneth6d~,"more~
semi volatile organic compounds are condensed out and not detected, and water-
soluble volatile compounds may also be lost to the condensate trap(s).
At this time, the Agency recommends use of the unheated FID system with
condensate trap(s) for continuous monitoring of THC over the life of the
permit. We previously preferred the heated method because, under the risk-
based THC analysis to waive the 100 ppmv CO limit, it was important to detect
as large a fraction of the organic emissions as possible. The site-specific
risk assessment is based to a large extent on the mass emission rate of THC.
We now, however, prefer the unheated system for two reasons: (1) there is
considerable uncertainty about the reliability and validity of results using a
heated system; and (2) given that the Agency now prefers use of the
technology-based 20 ppmv THC limit to waive the 100 ppmv CO limit, attempting
to detect the major portion of organic emissions using a heated system is not
as important.
We understand that a heated FID system can pose a number of problems: (1) the
sample extraction lines may plug due to heavy particulate loadings and
condensed organic compounds; and (2) semi and nonvolatile compounds may adsorb
on the inside of the extraction lines causing unknown effects on
measurements. The unheated system should not pose these problems because the
•gas conditioning system and condensate traps should remove the particulates
and condensable compounds at the beginning of the extraction system. We
understand further that a number of the FID systems currently installed for
continuous monitoring use condensate trap(s) even though, in some cases, the
sampling line may be heated. The Agency is currently evaluating field
experiences with FID systems and the practicability of continuous monitoring
with a heated system.
Although an unheated FID system monitoring a conditioned gas will detect only
the volatile fraction of organic compounds, the Agency believes this is
adequate for the purpose of determining whether the facility is operating
under good operating conditions. Available data indicate that when emissions
of semi and nonvolatile organic compounds increase, volatile compounds also
4-25
-------
increase.8 Thus, volatile compounds appear to be a good surrogate for the
semi and nonvolatile compounds that are often of greater concern because of
their health effects. Given, however, that the good operating practice-based
THC limit of 20 ppmv was based primarily on test burn data using (we believe)
heated FID systems, the Agency considered whether to lower the recommended THC
limit when an unheated system is used for compliance monitoring. Limited
available field test data indicated that a heated system would detect two to
four times the mass of organic compounds than a conditioned system. We
believe, however, that the 20 ppmv THC limit is still appropriate when a
conditioned system is used because: (1) the data correlating heated vs
conditioned systems are very limited; (2) the data on THG emission rates from
devices are limited (and there apparently is confusion in some cases as to
whether the data were taken with a heated or conditioned system); and (3) the
risk methodology is not sophisticated enough to demonstrate that a THC limit
of 5 or 10 ppmv using a conditioned system rather than a limit of 20 ppmv is
needed to adequately protect public health.
The THC monitoring method specified in Appendix A.3 of this document may be
appropriately modified for an unheated, conditioned system by ignoring
references to a heated sampling line and heated inlet to the FID, and by use
of condensate trap(s) and other conditioning methods.
Midwest Research Institute, Measurements of Particulates, Metals, and
Orqanics at a Hazardous Waste Incinerator, November 15, 1988 (Draft Final
Report).~
4-26
-------
^SECTION 5.0
REFERENCES
1. Midwest Research" Institute, Proposed Methods for Measurements of CO2*
THC. HC1. and Metals for Hazardous Waste Incinerators, EPA Contract
No. 68-01-7287, September 1988.
2. U.S. EPA. Guidance on Metals and Hydrogen Chloride Controls for Hazardous
Waste Incinerators. March, 1989, Draft Final Report.
3. Graham, J. L., D. L. Hall, and B. Dellinger, "Laboratory Investigation of
Thermal Degradation of a Mixture of Hazardous Organic Compounds," Envi.
. Sci. Technol., Vol. 20, No. 7, pp. 703-710, July 1986.
4. Taylor, P. H., and B. Dellinger, "Thermal Degradation Characteristics of
Chloromethane Mixtures," Envi. Sci. Techno!., April 1988.
5. Kramlich, J. C., M. P. Heap, W. R. Seeker, and G. S. Samuelson, "Flame-
Mode Destruction of Hazardous Waste Compounds," 20th Symposium
(International) on Combustion. The Combustion Institute; 1991; 1984.
6. LaFond, R. K., J. L. Kramlich, and W. R. Seeker, "Evaluation of
Continuous Performance Monitoring Techniques for Hazardous Waste
Incinerators," APCA Journal. 35, (6): 658; June 1985.
7. Trenholm, A., P. Gorman, and G. Jungclaus, "Performance Evaluation of
Full-Scale Hazardous Waste Incinerators, Vol. 2 - Incinerator Performance
Results."EPA-600/2-84-181b, PB 85-129518, November 1984.
8. Waterland, L. R., "Pilot-Scale Investigation of Surrogate Means of
Determining POHC Destruction." Final Report for the Chemical
Manufacturers Association, Acurex Corporation, Mountain View, California,
July 1983.
9. Change, D. P., et al., "Evaluation of a Pilot-Scale Circulating Bed
Combustor as a Potential Hazardous Waste Incinerator," APCA Journal. 37,
(3): 266; March 1987. ~~
10. U.S. EPA, "Permit Writer's Guide to Test Burn Data - Hazardous Waste
Incineration," EPA/625/6-86/012 September 1986.
11. Hall, D. L.f B. Dellinger, J. L. Graham, and W. A. Rubey, "Thermal
Decomposition Properties of a Twelve Component Organic Mixture,"
Hazardous Waste & Hazardous Materials. 3, (4): 441; 1986.
12. EPA, Guideline on Air Quality Models (Revised), EPA-450/2-78-027, Office
of A1r Quality Planning and Standards, Research Triangle Park, North
Carolina, July, 1986.
5-1
-------
13. Auer, August H. Jr., "Correlation of Land Use^-and Cover with
Meteorological Anomalies," Journal of Applied Meteorology, pp. 636-643,
1978.
5-2
-------
APPENDIX A—TEST METHODS
This appendix provides general guidance on the combustion effluent gas (i.e.,
stack gas) measurements necessary to implement the guidance on CO permit
Timits^
Compliance with the Tier II CO limits will require all Tier I measurements
with the addition of THC measured continuously during the ORE trial burn (or
during test burns under conditions equivalent to the ORE trial burn). The
maximum hourly average THC emission rate would be used for Tier II
assessment. Continuous CO and 02 measurements must be made concurrently with
the ORE trial burn and must be of the same duration as each ORE test run.
Three replicate test runs are required under existing practices for ORE trial
burn measurements.
Methods for Moisture and Oxygen Measurements
If the CO and/or the (oxygen) continuous emission monitors measure the
effluent gas on a wet basis, then moisture measurements will be required
during the the trial burn and at intervals specified by the permit writer
thereafter. Moisture measurements are made using Reference Method 4—
Determination of Moisture Content in Stack Gas, or in conjunction with
Reference Method 5—Determination of Particulate Emissions from Stationary
Sources; both methods are published in 40 CFR 60, Appendix A.
When the alternate time-above-a-CO-level format is used, oxygen measurements
need not be continuous for the life of the permit. Rather, they are performed
annually or on a more frequent basis as specified in the permit. For these
intermittent measurements, Reference Method 3—Gas Analysis for Carbon
Dioxide, Oxygen, Excess Air, and Dry Molecular Weight (40 CFR 60) is the
method used. Method 3 presents several optional procedures. The method to be
used is single-point, integrated sampling (multipoint integrated sampling at
the permit applicant's option) with analysis for oxygen by ORSAT.
Alternatively, Reference Method 3A—Determination of Oxygen and Carbon Dioxide
A-l
-------
Concentrations in Emissions from Stationary Sources (Instrumental Analyzer
Procedure) may be used.
Quality Assurance Guidance for Continuous Emission Monitoring Systems
The continuous emission monitoring system (CEMS) performance specification
presented in Appendix A.2 is intended for evaluating the performance of the
monitor upon installation. It is the responsibility of the owner/operator to
assure proper calibration, maintenance, and operation of the CEMS on a
continual basis. The owner/operator should establish a QA program to evaluate
and monitor performance on a continual basis. The following QA guidelines are
presented:
1. Conduct a daily calibration check for each monitor. Adjust the
calibration if the check indicates the instrument's calibration drift
exceeds the specification established in Appendix A.2.
2. Conduct a daily system audit. During the audit, review the calibration
check data, inspect the recording system, inspect the control panel
warning lights, and inspect the sample transport/interface system (e.g.,
flowmeters, filters), as appropriate.
«3. Conduct a quarterly calibration error test at the span midpoint.
4. Repeat the entire performance specification test every second year.
The following appendices provide summaries of these methods; specifically:
A.2 Performance Specifications for Continuous Emission Monitoring of Carbon
Monoxide and Oxygen in Hazardous Waste Incinerators, Boilers* and
Industrial Furnaces; and
A.3 Measurement of Total Hydrocarbons in Hazardous Waste Incinerators,
Boilers, and Industrial Furnaces.
A-2
-------
APPENDIX A.2--PERFORMANCE SPECIF.ICAIIQNS^.EQfl.CQNIIKUOUS^MlSSlflK-
MONITORING OF~CMBON MONOXIDE AND OXYGEN IN HAZARDOUS WASTE
INCINERATORS, BOILERS, AND INDUSTRIAL FURNACES
1.0 Applicability and Principle
1.1 Applicability.
This specification is to be used for evaluating the acceptability of carbon
monoxide (CO) and oxygen (02) continuous emission monitoring systems (CEMS)
installed on hazardous waste incinerators, boilers, and industrial furnaces.
' * ,
This specification is intended to be used in evaluating the acceptability of
the CEMS at the time of or soon after installation and at other times as
specified in the regulations. This specification is not designed to evaluate
the CEMS performance over an extended period of time nor does it identify
specific routine calibration techniques and other auxiliary procedures to
assess CEMS performance. The source owner or operator, however, is
responsible to calibrate, maintain, and operate the CEMS.
1.2 Principle.
Installation and measurement location specifications, performance and
equipment specifications, test procedures, and data reduction procedures are
included in this specification. Relative accuracy (RA) tests, calibration
error (Ec) tests, calibration drift (CD) tests, and response time (RT) tests
are conducted to determine conformance of the CEMS with the specification.
2.0 Definitions
2.1 Continuous Emission Monitoring System (CEMS).
The CEMS is comprised of all the equipment used to generate data and includes
the sample extraction and transport hardware, the analyzer(s), and the data
recording/processing hardware (and software).
A-3
-------
2.2 Continuous.
A continuous monitor is one in which the sample to be analyzed passes the mea-
surement section of the analyzer without interruption and which evaluates the
detector response to the sample at least once each 15 sec. and records the
average of these observations each and every minute.
The hourly rolling average is the arithmetic mean of sixty (60) most recent 1- '
minute average values recorded by the continuous monitoring system.
2.3 Monitoring System Types.
There are three basic types of monitoring systems: extractive, cross-stack,
and 1n-situ. Carbon monoxide monitoring systems generally are extractive or
cross-stack, while oxygen monitors are either extractive or in-situ.
,2.3.1 Extractive.
Extractive systems use a pump or other mechanical, pneumatic, or hydraulic
means to draw a small portion of the stack or flue gas and convey it to the
remotely located analyzer.
2.3.2 In-situ.
In-situ analyzers place the sensing or detecting element directly in the flue
gas stream and thus perform the analysis without removing a sample from the
stack.
2.3.3 Cross-stack.
Cross-stack analyzers measure the parameter of interest by placing a source
beam on one side of the stack and either the detector (in single-pass
Instruments) or a retro-reflector (in double-pass instruments) on the other
side and measuring the parameter of interest (e.g., CO) by the attenuation of
the beam by the gas in its path.
A-4
-------
2.4 Span.
The upper limit of the gas concentration meansurement range specified Section
4.1.
2.5 Instrument Range.
The maximum and minimum concentration that can be measured by a specific in-
strument. The minimum is often stated or assumed to be zero (0) and the range
expressed only as the maximum. If a single analyzer is used for measuring
multiple ranges (either manually or automatically), the performance standards
expressed as a percentage of full scale apply to all ranges.
2.6 Calibration Drift.
Calibration drift is the change in response or output of an instrument from a
reference value over time. Drift is measured by comparing the responses to a
reference standard over time with no adjustment of instrument settings.
2.7 Response Time.
The response time of a system or part of a system is the amount of time be-
tween a step change in the system input (e.g., change of calibration gas)
until the data recorder displays 95 percent of the final value.
2.8 Accuracy.
Accuracy 1s a measure of agreement between a measured value and an accepted or
true value and is usually expressed as the percentage difference between the
true and measured values relative to the true value. For this performance
specification the accuracy 1s checked by conducting a calibration error (Ec)
test and a relative accuracy (RA) test.
A-5
-------
2.8.1 Calibration Error.
Calibration error is a measure of the deviation of a measured value at the
analyzer mid range from a reference value.
2.8.2 Relative Accuracy..
Relative accuracy is the comparison of the CEMS response to a value measured
by a reference test method (RM). The applicable reference test methods are
Method 10—Determination of Carbon Monoxide from Stationary Sources and
Method 3—Gas Analysis for Carbon Monoxide, Oxygen, Excess Air, and Dry Molec-
ular Weight; these methods are found in 40 CFR 60, Appendix A.
3.0 Installation and Measurement Location Specifications
3.1 CEMS Measurement Location.
The best or optimum location of the sample interface for the monitoring system
is determined by a number of factors, including ease of access for calibration
and maintenance, the degree to which sample conditioning will be required, the
degree to which it represents total emissions, and the degree to which it
represents the combustion situation in the firebox. The location should be as
free from in-leakage influences as possible and reasonably free from severe
flow disturbances. The sample location should be at least two equivalent duct
diameters downstream from the nearest control device, point of pollutant gen-
eration, or other point at which a change in the pollutant concentration or
emission rate occurs and at least 0.5 diameters upstream from the exhaust or
control device. The equivalent duct diameter is calculated as per 40 CFR 60,
Appendix A, Method 1, Section 2.1.
The sample path or sample point(s) should include the concentric inner 50 per-
cent of the stack or duct cross section. For circular ducts, this is 0.707 x
diameter and a single-point probe, therefore, should be located between
0.141 x diameter and 0.839 x diameter from the stack wall and a multiple-point
probe should have sample inlets in this region. A location which meets both
the diameter and the cross-section criteria will be acceptable.
A-6
-------
"if these criteria are not achievable or if the location is otherwise less than
optimum, the possibility"of stratification should be investigated. To check
for stratification, the oxygen concentration should also be measured as veri-
fication of oxygen in-leakage. For rectangular ducts, at least nine sample
points located at the center of similarly shaped, equal area divisions of the
cross section should be used. For circular ducts, 12 sample points (i.e., six
points on each of the two perpendicular diameters) should be used, locating
the points as described 1n 40 CFR 60, Appendix A, Method 1. Calculate the
mean value for all sample points and select the point(s) or path that provides
a value equivalent to the mean. For these purposes, 1f no single value is
more than 15 percent different from the mean and if no two single values are
different from each other by more than 20 percent of the mean, then the gas
can be assumed homogeneous and can be sampled anywhere. The point(s) or path
should be within the Inner 50 percent of the area.
Both the oxygen and CO monitors should be Installed at the same location or
very close to each other. If this is not possible, they may be installed at
different locations 1f the effluent gases at both sample locations are not
stratified and there 1s no 1n-leakage of air between sampling locations.
3.2 Reference Method (RM) Measurement Location^andI Traverse Points.
•Select, as appropriate, an accessible RM measurement point at least two equiv-
alent diameters downstream from the nearest control device, the point of pol-
lutant generation, or other point at which a change in the pollutant concen-
tration or emission rate may occur, and at least a half equivalent diameter
upstream from the effluent exhaust or control device. When pollutant concen-
tration changes are due .solely to oxygen in-leakage (e.g., air heater leak-
ages) and pollutants and diluents are simultaneously measured at the same
location, a half diameter may be used in lieu of two equivalent diameters.
The CEMS and RM locations need not be the same.
Then select traverse points that assure acquisition of representative samples
over the stack or duct cross section. The minimum requirements are as fol-
lows: Establish a "measurement line" that passes through the centroidal area
A-7
-------
and in the direction of any expected-'S-trati^icalrtOTri^I^
with the CEM measurements, displace the line up to 30 cm (or 5 percent of the
equivalent diameter of the cross section, whichever is less) from the cen-
troidal area. Locate three traverse points at 16.7, 50.0, and 83.3 percent of
the measurement line. If the measurement line is longer than 2.4 m and pol-
lutant stratification is not expected, the tester may choose to locate the
three traverse points on the line at 0.4, 1.2, and 2.0 m from the stack or
duct wall. This option must not be used at points where two streams with dif-
ferent pollutant concentrations are combined. The tester may select other
traverse points, provided that they can be shown to the satisfaction of the
Administrator to provide a representative sample over the stack or duct cross
section. Conduct all necessary RM tests within 3 cm (but no less than 3 cm
from the stack or duct wall) of the traverse points.
4.0 Monitoring System Performance Specifications
Table A-l summarizes the performance standards for the continuous monitoring
systems. Each of the items is discussed in the following paragraphs. Two
sets of standards for CO are given—one for low range measurement and another
for high range measurement since the proposed CO limits are dual range. The
high range standards relate to measurement and quantification of short
duration high concentration peaks, while the low range standards relate to the
overall average operating condition of the incinerator. The dual-range
specification can be met either by using two separate analyzers, one for each
range, or by using dual range units which have the capability of meeting both
standards with a single unit. In the latter case, when the reading goes above
the full scale measurement value of the lower range, the higher range
operation will be started automatically.
A-8
-------
TABLE A-l. PERFORMANCE SPECIFICATIONS OF CO AND OXYGEN"MONITORS
CO monitors
Parameter
Low range
High range
Oxygen
monitors
Calibration drift
24 h
Calibration
error"
Response time
Relative accuracy
< 5% FSa
< 5% FS
< 1.5 min
< 5% FS
< 5% FS
< 1.5 min
< The greater
of 10% of
RM or 20 ppm
< 0.5% 0
< 0.5% 02
< 1.5 min
< The greater
of 20% of
RM or 1.0% 02
? FS means full scale measurement range.
Expressed as the sum of the mean absolute value plus the 95% con-
fidence interval of a series of measurements.
4.1 CEMS Span Values.
The span values shown below in Table A-2 are to be established for the contin-
uous emission monitoring system.
TABLE A-2. CEMS SPAN VALUES FOR CO AND OXYGEN MONITORS
CO monitors Oxygen
Low range High range monitors
(ppm) (ppm) (%)
Tier 1 rolling
average format
Tier 1 alternate
format
Tier 2 rolling
average format
Tier 2 alternate
format
200
200
2 x permit
limit
2 x permit
limit
3,000
3,000
3,000
1.1 x permitted
peak value
25
25
25
25
A-9
-------
4.2 System Measurement
In order to measure both the high and low concentrations consistently with- the
same or similar degree of accuracy, system measurement range maximum span
specifications are given for both the low and high range monitors. The system
measurement range chosen is based upon the permitted level and the span value
presented in Section 4.1.
The owner or operator must choose a measurement range that includes zero and a
high-level value. The high-level value is chosen by the source owner and
operator as follows:
1. For the low range CO measurements, the high level value is set between
1.5 times the permit limit and the span value specified in Section 4.1.
2. For the high range CO measurement, except for Tier II alternate format,
the high level value is set between 2000 ppm, as a minimum, and the span
value specified in Section 4.1.
3. For the high range CO measurement under Tier II using the alternate type
B format, the high level value is set at the span value specified in
Section 4.1.
4. For oxygen the high level value is set between 1.5 times the highest
level measured during the trial burn and the span value specified in Sec-
tion 4.1.
The calibration gas, or gas cell values used to establish the data recorder
scale should produce the zero and high level values.
4.3 Response Time.
The mean response time for the CO monitor(s) and oxygen monitors should not
exceed 1.5 min to achieve 95 percent of the final stable value.
A-10
-------
^4.4 Calibration Drift.
The CEMS calibration must not drift or deviate from the reference value of the
gas cylinder or gas cell by more than 5 percent full scale in 24 h for the CO
low range and the CO high range. For oxygen the calibration drift must be
less than 0.5 percent 02 in 24 h. The calibration drift specification must
not be exceeded for six out of the seven test days required during the test
(see Section 5 for the test procedures).
4.5 Calibration Error.
''.'.''-
/
The calibration error specification evaluates the system accuracy at the mid-
point of the measurement range by the calibration error test described in Sec-
tion 6. The test determines the difference between the measured value and the
expected value at this midpoint.
The calibration error of the CEMS must not exceed 5 percent full scale for
CO. The calibration error of the oxygen CEMS must not exceed 0.5 percent 02.
4.6 Relative Accuracy.
The relative accuracy (RA) of the carbon monoxide CEMS must not exceed 10 per-
cent of the mean value of the reference method (RM) test data or 20 ppm CO,
whichever is greater. Note that during the relative accuracy test, the CO
level may exceed the full scale of the low range monitor. When this occurs,
the mean CEMS measurement value should be calculated using the appropriate
data from both the low range and high range monitors.
The relative accuracy of the oxygen CEMS must not exceed 20 percent of the
mean value of the RM test data or 1 percent oxygen, whichever is greater.
A-ll
-------
Jf
5.0 Performance Specification Test Period
5.1 Pretest Preparation.
Install the GEMS, prepare the RM test site according to the specifications in
Section 3, and prepare the CEMS for operation according to the manufacturer's
written Instructions.
5.2 Calibration Error and Response Time Tests.
Prior to Initiating the calibration drift tests conduct the calibration error
test and the response time test according to the test procedures established
in Section 6. The carbon monoxide and oxygen (if applicable) monitoring
systems must be evaluated separately.
5.3 Calibration Drift Test Period.
The monitoring system should be operated for some time before attempting drift
checks because most systems need a period of equilibration and adjustment
before the performance 1s reasonably stable. At least one week (168 h) of
continuous operation is recommended before attempting drift tests.
While the facility is operating at normal conditions, determine the magnitude
.of the calibration drift (CD) once each day (at 24-h intervals) for seven
consecutive days according to the procedure given in Section 6. The carbon
monoxide and oxygen (if applicable) monitoring systems must be evaluated
separately.
5.4 RA Test Period.
Conduct the RA test according to the procedure given in Section 6 while the
facility is operating at normal conditions. The RA test may be conducted dur-
ing the CD test period. The RA test may be conducted separately for each of
the monitors (carbon monoxide and oxygen, if applicable) or may be conducted
as a combined test so that the results are calculated only for the corrected
CO concentration (i.e., CO corrected to 7 percent oxygen); the latter approach
1s preferred.
A-12
-------
6.0 Performance Specification Test Procedures
6.1 Response Time.
The response time tests apply to all types of monitors, but will generally
have significance only for extractive systems. The entire system'is checked
with this procedure including sample extraction and transport (if applicable),'
sample conditioning (if applicable), gas analyses, and the data recording.
Introduce zero gas into the system. For extractive systems, the calibration
gases should be introduced at the probe as near to the sample location as
possible. For 1n-situ systems, introduce the zero gas at the sample interface
so that all components active in the analysis are tested. When the system
output has stabilized (no change greater than 1 percent of full scale for
30 s), switch to monitor stack effluent and wait for a stable value. Record
the time (upscale response time) required to reach 95 percent of the final
stable value. Next, introduce a high level calibration gas and repeat the
above procedure (stable, switch to sample, stable, record). Repeat the entire
procedure three times and determine the mean upscale and downscale response
times. The slower or longer of the two means is the system response time.
6.2 Calibration Error Test
6.2.1 Procedure.
The procedure for testing calibration error is to set the instrument zero and
span with the appropriate standards and then repeatedly measure a standard in
the middle of the range. In order to minimize bias from previous analyses,
the sequence of standard introduction should alternate between high and low
standards prior to the mid-level standard (e.g., high, mid, low, mid, high,
mid, low, mid, etc.) until six analyses of the mid-level standard are ob-
tained, with three values obtained from upscale approach and three values ob-
tained from downscale approach.
A-13
-------
The differences between the measured instrument output and the expected output
of the reference-^standards*are~used as the data points.
6.2.2 Calculations.
Summarize the results on a data sheet. For each of the six measurements made,
calculate the arithmetic difference between the midpoint reference value and
the measured value. Then calculate the mean of the differences standard
deviation, confidence coefficient, and calibration error using Equations 2-1,
2-2, 2-3, and 2-4 presented in Section 7.
6.3 Zero and Span Calibration Drift.
The purpose of the calibration drift (CD) checks is to determine the ability
of the CEMS to maintain its calibration over a specified period of time. The
performance specifications establish a standard related to span drift. Each
drift test 1s conducted seven times and the system(s) are allowed to exceed
the limit once during the test.
»
During the drift tests, no adjustment of the system is permitted except those
automatic internal adjustments which are part of the automatic compensation
circuits Integral to the analyzer. If periodic automatic adjustments are made
to the CEMS zero and calibration settings, conduct the daily CD test
Immediately before these adjustments, or conduct it in such a way that the CD
can be determined (calculated). Subsequent CEMs operation must include the
same system configuration as used during the performance testing.
Select a reference gas with a CO or 02 concentration between 80 and
!
100 percent of the full-scale measurement range of the analyzer; ambient air
(20.9 percent 02) may be used as the reference gas for oxygen. The zero gas
• should contain the lowest concentration recommended by the manufacturer.
Prior to the test, calibrate the instrument. At the beginning of the test,
Introduce the selected zero and span reference gases (or cells or filters).
After 24 h and at 24-h Intervals thereafter, alternately introduce both the
zero and span reference gases, wait until a stable reading is obtained and
record the values reported by the system. Subtract the recorded CEMS response
A-14
-------
from the reference value. Repeat this procedure for 7 days, obtaining eight
values of zero and span gas measurements (the initial values and seven 24-h
readings). The difference between the established or reference values for the
zero and span and the measured values may not exceed the specifications in
Table 4.1 more than once, and the average value must not exceed the
specification. .
6.4 Relative Accuracy Test Procedure
6.4.1 Sampling Strategy for RM Tests.
•
Conduct the RM tests in such a way that they will yield results representative
of the emissions from the source and can be correlated to the CEMS data.
Although it is preferable to conduct the confirm that the pair of results are
on a consistent moisture, temperature, and diluent concentration basis. Then,
compare each Integrated CEMS value against the corresponding average RM
value. Make a direct comparison of the RM results and CEMS integrated average
value. When oxygen monitoring is required by the regulation to calculate
carbon monoxide normalized to 7 percent 02, the RM test results should be
calculated and compared on this basis. That is, the CO concentrations
normalized to 7 percent 02 should be calculated using the RM test data and
these results should be compared to the CO concentration normalized to
7 percent 02 measured by the CEMS.
6.4.2 Correlation of RM and CEMS Data.
Correlate the CEMS and the RM test data as to the time and duration by first
determining from the CEMS final output (the one used for reporting) the
integrated average pollutant concentration during each pollutant RM test
period. Consider system response time, if important, and confirm that the
pair of results are on a consistent moisture, temperature, and diluent
concentration basis. Then, compare each integrated CEMS value against the
corresponding average RM value. Make a direct comaprison of the RM results
and CEMS integrated average value. When oxygen monitoring is required by the
regulation to calculate carbon monoxide normalized to 7 percent 02, the RM
test results should be calculated and compared on this basis. That is, the CO
A-15
-------
concentrations—noma44-ze<^o-^^eFeenfe^%=^eu-1^b the RM
test data and these results should be compared to the CO concentration
normalized to 7 percent 02 measured by the CEMS.
6.4.3 Number of RM Tests.
Conduct a minimum of nine sets of all necessary RM tests.
Note: The tester may choose to perform more than nine sets of RM tests. If
this option is chosen, the tester may, at their discretion, reject a maximum
of three sets of the test results so long as the total number of test results
used to determine the RA is greater than or equal to nine, but they must
report all data including the rejected data.
6.4.4 Calculations.
Summarize the results on a data sheet. Calculate the mean of the RM values.
Calculate the arithmetic differences between the RM and the CEMS output
sets. Then calculate the mean of the difference, standard deviation, confi-
dence coefficient, and CEMS RA, using Equations 2-1, 2-2, 2-3, and 2-5.
7.0 Equations
7.1 Arithmetic Mean.
Calculate the arithmetic mean of the difference, d, of a data set as follows:
n
* - 7f 1-1 d1 (Eq* 2-
Where n = number of data points
n
Ed*- algebraic sum of the individual differences
1-1
A-16
-------
When the mean of the differences-of—pairs—of—dat-a—ts™
correct the data for moisture, if applicable.
7.2 Standard Deviation.
Calculate the standard deviation, Sd, as follows:
L/2
/
*
n
Z H 2
1=1 d1 -
n
z .
1=1 d1
_ n
Z
n - 1
(Eq. 2-2)
7.3 Confidence Coefficient.
Calculate the 2.5 percent error confidence coefficient (one-tailed), CC, as
follows:
CC= '0.975
. 2-3)
Where t0>975 = t-value (see Table A-3).
TABLE A-3. VALUES
na '0.975
2 12.706
3 4.303
4 3.182
5 2.776
6 2.571
na
7
8
9
10
11
'0.975
2.447
2.365
2.306
2.262
2.228
na
12
13
14
15
16
'0.975
2.201
2.179
2.160
2.145
2.131
The values in this table are already cor-
rected for n-1 degrees of freedom. Use
n equal to the number of individual values.
A-17
-------
Calculate the calibration error (Ec) of a set of data as follows:
For carbon monoxide:
Ec = l?i *
-------
8.0 Reporting
At a minimum (check with the appropriate regional office, or state, or local
agency for additional requirements, if any) summarize in tabular form the
results of the response time tests, calibration error tests, calibration drift
testsT and the relative accuracy tests. Include all data sheets,
calculations, charts (records of CEMS responses), cylinder gas concentration
certifications, and calibration cell response certifications (if applicable),
necessary to substantiate that the performance of the CEMS met the performance
specifications.
9.0 References
Jahnke, J. A., and G. J. Aldina, Handbook: Continuous Air Pollution Source Mon-
itoring systems, U.S. Environmental Protection Agency Technology Transfer,
Cincinnati, Ohio 45268, EPA-625/6-79-005 (June 1979).
Gaseous Continuous Emission Monitoring Systems - Performance Specification Guidelines
for_so2, NO , co2, 02, and TRS, U.S. Environmental Protection Agency OAQPS/ESED,
Research Triangle Park, North Carolina 27711, EPA-450/3-82-026 (October 1982).
Quality Assurance Handbook for Air Pollution Measurement Systems: Volume I.
Principles, U.S. Environmental Protection Agency ORD/EMSL, Research Triangle
Park, North Carolina 27711, EPA-600/9-76-006 (December 1984).
Michie, R. M. Jr., et a 1.,'Performance Test Results and Comparative Data for
Designated Reference Methods for Carbon Monoxide, U.S. Environmental Protection
Agency ORD/EMSL, Research Triangle Park, North Carolina 27711,
EPA-600/S4-83-013 (September 1982).
Ferguson, B. B., R. E. Lester, and W. J. Mitchell, field ^valuation of carbon
Monoxide and Hydrogen Sulfide Continuous Emission Monitors at an Oil Refinery, U.S.
Environmental Protection Agency, Research Triangle Park, North Carolina 27711,
EPA-600/4-82-054 (August 1982).
A-19
-------
APPENDIX A.3--MEASUREMENT OF TOTAL HYDROCARBONS IN HAZARDOUS WASTE
INCINERATORS, BOILERS, AND INDUSTRIAL FURNACES
NOTE; See Discussion in Section 4.9 of the guidance document regarding use
of an unheated, conditioned gas THC monitoring system.
1.0 Applicability and Principle
1.1 Applicability.
This method applies to the measurement of total hydrocarbons, as a surrogate
measure for total gaseous organic concentration of the combustion gas
stream. The concentration is expressed in terms of propane.
*
1.2 Principle.
A gas sample 'is extracted from the source through a heated sample line, and
heated glass fiber filter to a flame ionization detector (FID). Results are
reported as volume concentration equivalents of the propane.
2.0 Definitions
2.1 Measurement System.
The total equipment required for the determination of the gas concentration.
The system consists of the following major subsystems.
2.1.1 Sample Interface.
That portion of the system that is used for one or more of the following
sample acquisition, sample transportation, sample conditioning, or protection
of the analyzer from the effects of the stack effluent.
A-20
-------
--• -=*"
2.1.2 Organic Analyzer.
That portion of the system that senses organic concentration and generates an
""' output proportional to the gas concentration.
2.1.3 Data Recorder
That portion of the system that records a permanent record of the measurement
values.
2.2 Span Value.
For most incinerators a 50 ppmv propane span is appropriate. Higher span
values may be necessary if propane emissions are significant. For conve-
nience, the span value should correspond to 100 percent of the recorder scale.
2.3 Calibration Gas.
A known concentration of a gas in an appropriate diluent gas.
2.4 Zero Drift.
The difference in the measurement system response to a zero level calibration
gas before and after a stated period of operation during which no unscheduled
maintenance, repair, or adjustment took place.
2.5 Calibration Drift.
The difference in the measurement system response to a mid-level calibration
gas before and after a stated period of operation during which no unscheduled
maintenance, repair, or adjustment took place.
A-21
-------
_,___ ,--*'"
2.6 Response Time. ~---'~-™"-- - .__••; •^' • • "~ ~ •" " " " -•--"---==-
The time interval from a step change in pollutant concentration at the inlet
to the emission measurement system to the time at which 95 percent of the
corresponding final value is reached as displayed on the recorder.
2.7 Calibration Error.
The difference between the gas concentration indicated by the measurement
system and the known concentration of the calibration gas.
3.0 Apparatus
An acceptable measurement system includes a calibration value, gas filter and
heated pump proceeding the analyzer. All components in contact with the
sample gas (probe, calibration valve, filter, and sample lines), as well as
all parts of the flame lonization analyzer between the sample inlet and the
flame ionization detector (FID) must be heated to 150-175°C. This includes
the sample pump if it 1s located on the inlet side of the FID.
The essential components of the measurement system are described below:
.3.1 Organic Concentration Analyzer.
A heated flame ionizatlon analyzer (FIA) capable of meeting or exceeding the
specifications in this method.
3.2 Sample Probe.
Stainless steel, or equivalent, three-hole rake type. Sample holes shall be
4 mm in diameter or smaller and located at 16.7, 50, and 83.3 percent of the
equivalent stack diameter. Alternatively, a single opening probe may be used
so that a gas sample is collected from the centrally located 10 percent area
of the stack cross section.
A-22
-------
3.3 Sample Line.
Stainless steel or Teflon* tubing to transport the sample gas to the
analyzer. The sample line should be heated to between 150 and 175°C.
3.4 Calibration Valve Assembly.
A heated three-way valve assembly to direct the zero and calibration gases to
the analyzers is recommended. Other methods, such as quick-connect lines, to
route calibration gas to the analyzers are applicable.
3.5 Partlculate Filter.
An in-stack or an out-of-stack glass fiber filter is recommended if exhaust
gas particulate loading is significant. An out-of-stack filter must be
heated.
3.6 Recorder.
A strip-chart recorder, analog computer, or digital recorder for recording
measurement data. The minimum data recording requirement is one measurement
value per minute.
Note; This method is often applied in highly explosive areas. Caution and
care should be exercised in choice of equipment and installation.
4.0 Calibration and Other Gases
Gases used for calibration, fuel, and combustion air (if required) are
contained in compressed gas cylinders. Preparation of calibration gases shall
be done according to the procedure in Protocol No. 1, listed in
Reference 9.2. Additionally, the manufacturer of the cylinder should provide
* Mention of trade names or specific products does not constitute endorsement
by the Environmental Protection Agency.
A-23
-------
a recommended shelf life for each calibration gas cylinder over which the
concentration does not change more than ±2 percent from the certified value.
4.1 Fuel. " ' '"- '" ' "' "
A 40 percent H2/60 percent He or 40 percent H2/60 percent N2 gas mixture is
recommended to avoid an oxygen synergism effect that reportedly occurs when
oxygen concentration varies significantly from a mean value.
4.2 Zero Gas. •
High purity air with less than 0.1 parts per million by volume (ppmv) of
organic material methane or carbon equivalent or less than 0.1 percent of the
span value, whichever is greater.
4.3 Low-Level Calibration Gas.
Propane calibration gas (in air or nitrogen) with a concentration equivalent
to 20 to 30 percent of the applicable span value.
4.4 H1d-Leve1 Calibration Gas.
Propane calibration gas (in air or nitrogen) with a concentration equivalent
to 45 to 55 percent of the applicable span value.
4.5 High-level Calibration Gas.
Propane calibration gas (in air or nitrogen) with a concentration equivalent
to 80 to 90 percent of the applicable span value.
5.0 Measurement System Performance Specifications
5.1 Zero Drift.
Less than ±3 percent of the span value.
A-24
-------
•
5.2 Calibration Drift. ..~ ™ -, . , ___•
Less than ±3 percent of the span value.
5.3 Calibration Error.
Less than ±5 percent of the calibration gas value.
6.0 Pretest Preparations
6.1 Selection of Sampling Site.
The location of the sampling site is generally specified by the applicable
regulation or purpose of the test, i.e., exhaust stack, inlet line, etc. The
sample port shall be located at least 1.5 m or 2 equivalent diameters upstream
of the gas discharge to the atmosphere.
6.2 Location of Sample Probe.
Install the sample probe so that the probe is centrally located in the stack,
pipe, or duct and is sealed tightly at the stack port connection.
»
6.3 Measurement System Preparation.
Prior to the emission test, assemble the measurement system following the
manufacturer's written instructions in preparing the sample interface and the
organic analyzer. Make the system operable.
6.4 Calibration Error Test.
Immediately prior to the test series (within 2 h of the start of the test)
introduce zero gas and high-level calibration gas at the calibration valve
assembly. Adjust the analyzer output to the appropriate levels, if neces-
sary. Calculate the predicted response for the low-level and mid-level gases
A-25
-------
based on a linear response line between the zero and high-level responses.
Then introduce low-level and mid-level call brat ion gases- - successively^ to-the
measurement system. Record the analyzer responses for low-level and mid-level
calibration gases and determine the differences between the measurement system
responses and the predicted responses. These differences must be less than
5 percent of the respective calibration gas value. If not, the measurement
system is not acceptable and must be replaced or repaired prior to testing.
No adjustments to the measurement system shall be conducted after the
calibration and before the drift check (Section 7.3). If adjustments are
necessary before the completion of the test series, perform the drift checks
prior to the required adjustments and repeat the calibration following the
adjustments. If multiple electronic ranges are to be used, each additional
range must be checked with a mid-level calibration gas to verify the
multiplication factor.
6.5 Response Time Test.
Introduce zero gas into the measurement system at the calibration valve
assembly. When the system output has stabilized, switch quickly to the high-
level calibration gas. Record the time from the concentration change to the
measurement system reponse equivalent to 95 percent of the step change.
Repeat the test three times and average the results.
7.0 Emissions Measurement Test Procedure
7.1 Organic Measurement.
Begin sampling at the.start of the test period, recording the time and any
required process information as appropriate. In particular, note on the
recording chart periods of process interruption or cyclic operation.
A-26
-------
Immediately following the completion of the test period and hourly during the
test period, reintroduce the zero and mid-level calibration gases, one at a
time, to the measurement system at the calibration valve assembly. (Make no
adjustments to the measurement system until after both the zero and
calibration drift checks are made.) Record the analyzer response. If the
drift values exceed the specified limits, invalidate the test results
preceding the check and repeat the test following corrections to the
measurement system. Alternatively, recalibrate the test measurement system as
in Section 6.4 and report the results using both sets of calibration data
(i.e.,, data determined prior to the test period and data determined following
the test period).
8.0 Organic Concentration Calculations
Determine the average organic concentration in terms of ppmv propane. The
average shall be determined by the integration of the output recording over
the period specified in the applicable regulation.
9.0 Bibliography
-Measurement of Volatile Organic Compounds—Guideline Series, U.S.
Environmental Protection Agency, Research Triangle Park, North Carolina,
Publication No. EPA-450/2-78-041, pp. 46-54 (.June 1978).
Traceablilitg Protocol for Establishing True Concentrations of Gases Used for
Calibration and Audits of Continuous Source Emissions Monitors (Protocol No.
1), U.S. Environmental Protection Agency, Environmental Monitoring and Support
Laboratory, Research Triangle Park, North Carolina (June 1978).
Gasoline Vapor Emission Laboratory Evaluation—Part 2, U.S. Environmental
Protection Agency, Office of Air Quality Planning and Standards, Research
Triangle Park, North Carolina, EMB Report No. 75-6AS-6 (August 1975).
A-27
-------
-------
APPENDIX B
TECHNICAL BACKGROUND DATA FOR THC RISK-BASED SCREENING LIMITS
This appendix provides background data on the risk-based Tier II approach to
establish CO limits based on conservative site-specific risk assessment using
THC as a surrogate for PICs. To demonstrate that THC emissions do not pose
significant risk, the applicant should either show that THC does not exceed
conservative Screening Limits, or conduct site-specific dispersion modeling
and risk assessment to show that the lifetime cancer risk to the MEI does not
exceed 10~5.
Conservative THC Screening Limits (shown in Table 1) have been developed to
reduce the burden on applicants and permit writers. These limits were
calculated using the maximum dispersion coefficients shown in Table Bl, and
the conservative, carcinogenic unit risk for THC. (The coefficients in Table
Bl are the same as those used to calculate the feed rate and emission
Screening Levels for metals, see Reference 2). According to the equation:
1 x 10-5
THC,mg/s = . X 1000
(dispersion coefficient x pg/m3) (unit risk, 1.0 x 10~5 m3/yg)
9/s •
Where:
«fc
1 x 10"5 = acceptable risk level to the MEI
dispersion coefficient = value from Table Bl, based on a nominal 1 g/s THC
emission rate
unit risk = conservative unit risk of 1.0 x 10~5 m3/yg,
based on the acceptable risk and calculated as
described below
1000 = conversion factor from g to mg
B-l
-------
=TfiTs™equaliorTsimpTTfTesTofne Fcfllow fng:
1000
THC, mg/sec
dispersion coefficient from Table B-l
The conservative unit risk 1.0 x 10"s ra3/yg for THC emissions is also used
when a site-specific risk assessment is conducted. This value was calculated
from data on organic compounds potentially emitted in stack gases during
hazardous waste combustion. All compounds (including dioxins and furans)
Identified historically during tests of hazardous waste combustion in
incinerators, boilers, and kilns were considered for these calculations along
with other toxic compounds for which health effects data were available.
Table B2 lists all the compounds used to derive the conservative unit risk.
All of the compounds listed and their emission concentrations are based on
data from tests of hazardous waste combustors except;
the compounds with a O.lng/1 entry for emission concentration (a nominal
detection limit) are those that have not been detected from hazardous
waste combustion, but for which health effects data are available;
• formaldehyde data were not available for hazardous waste combustion,
therefore, the concentrations were based on data for municipal waste
incinerators; and
• Cl and C2 hydrocarbon data were not available for hazardous waste
combustion, therefore, the concentrations were based on data from fossil
fuel combustion.
All of the emission concentrations where more than one data point was
available were calculated as the upper 95th percent of the range of available
data.
The compound specific unit risks in Table B2 are the carcinogenic inhalation
exposure limits based on the acceptable risk level of 1 x 10~ MEI.
Noncarcinogenic compounds have zero unit risk but are included on the list to
account for their contribution to the total mass of emissions. The potential
health effects associated with these compounds is discussed below.
B-2
-------
For toxic substances not known to display carcinogenic properties, "there"
appears to be an identifiable exposure threshold below which adverse health
effects usually do not occur. Therefore, protection against the adverse
health effects of a noncarcinogen is likely to be achieved by preventing total
exposure levels from exceeding the threshold dose. The Agency has therefore
conservatively defined reference air concentrations (RACs) for noncarcinogenic
compounds that are defined in terms of a fixed fraction of the estimated
threshold concentration. RACs are derived from oral Reference Doses (RfDs)
for the compounds of concern.
The following equation is used to convert oral RfDs to RACs:
RAC (mg/m3) » RfP ("gAg-bw/day) X body weight X Corrective Factor X Background Level Factor
m air breathed/day
Where: - RfD 1s the oral reference dose
- Body weight is assumed to be 70 kg for an adult male
- Volume of air breathed by an adult male 1s~assumed to be 20 m3 per
day.
- Correction factor for route-to-route extrapolation (going from the
oral route to the inhalation route) 1s 1.0
- Factor to fraction the RfD to- the intake resulting from direct
Inhalation of the compound emitted from the source is 0.25 (i.e.,
an Individual 1s assumed to be exposed to 75 percent of the RfD
from other sources).
Table B3 presents the RACs, emission concentrations, and actual maximum air
concentrations for the noncarcinogenic compounds of interest. The maximum
ambient air concentrations were calculated using a conservative stack flowrate
of 15,000 dscfm and dispersion coefficient of 76 (ug/m3/g/s). The stack
flowrate 1s a reasonable flowrate for a large hazardous waste incinerator, and
the dispersion coefficient is the worst case value from Table 81 for complex
terrain and an effective stack height of 4 m. As shown in this table, the
actual maximum ambient air concentrations for each compound do not exceed the
RACs and therefore, would not cause adverse health effects.
B-3
-------
TABLE Bl ^
CONSERVATIVE DISPERSION COEFFICIENTS USED TO DETERMINE THE|
SCREENING LIMITS (ug/m3/g/s)
Terrain - Adjusted
Effective Stack
Height
<•)
4
6
8
10
12
14
16
18
20
22
24
26
28
30
35
40
45
50
55 '
60
65
70
75
80
85
90
95
100
105
110
115
120
Noncomplex
Urban land use
18
16
14
12'
11
9.7
8.6
7.6
6.7
6.0
5.3
4.7
4.1
3.7
2.9
2.2
1.8
1.4
1.1
0.89
0.72
0.64
0.56
0.50
0.44
0.39
0.34
0.30
0.26
0.23
0.20
0.18
Terrain
Rural land use
35
30
26
23
19
15
12
10
8.1
6.3
5.0
3.9
3.1
2.4
1.5
1.0
0.72
0.54
0.41
0.31
0.24
0.20
0.17
0.14
0.12
0.099
0.083
0.070
•0.059
0.049
0.041
0.035
Complex
Terrain
76
52
35
24
20
16
14
13
11
10
9.2
8.3
7.5
6.8
5.5
4.4
3.6
2.9
2.4
1.9
1.6
1.4
1.2
1.1
0.99
0.89
0.79
0.71
0.63
0.56
0.50
0.45
Source: Versar, Guidance on Metals and Hydrogen Chloride Controls for
Hazardous Waste Incinerators, U.S. Environmental Protection
Agency Office of Solid Waste, September 1988.
B-4
-------
TABLE 82. WEIGHTED UNIT RISK VALUE FOR PICs
Compound (CAS No.)
Risk ..
:- specific,
dose (ug/nr)
95th
percent! le
cone; (ng/L)
Weight
fraction
Unit
rLsk
(nrVpg)
Weighted
unit risk
(m-Vug)
CARCINOGENS
Aery I amide (79-06-01)
Acrylonitrile (107-13-1)
Aldrin (309-00-2)
Aniline (62-53-3)
Benzjalanthracene (56-55-3)
Benzene (71-43-2)
Benzidine (92-87-5)
Benzolalpyrene (50-32-8)
Bis(2-chloroethyl) ether (111-44-4)
Bis(chloromethyl) ether (542-88-1)
Bis(2-ethylhexyl) phthalate
1,3-Butadiene (106-99-0)
Carbon tetrachloride (56-25-5)
Chlordane (57-74-9)
Chloroform (67-66-3)
Chloromethane (74-87-3)
Chloromethyl methyl ether (107-30-2)
DOT (50-29-3)
Dibenzo[a,h]anthracene (53-70-3)
1,2-Dibromo-3-chloropropane (96-12-8)
1,2-0!bromoethane (106-93-4)
1,2-Olchloroethane (107-06-2)
1,1-Oichloroethylene (75-35-4)
Dieldrln (60-57-1)
Diethylstilbestrol (56-53-1)
Otmethylnitrosamine (62-75-9)
2,4-Oinitrotoluene (121-14-2)
Oioxane (123-91-1)
1,2-Oiphenylhydrazine (122-66-7)
Epichlorohydrin (106-89-8)
Ethylene oxide (75-21-8)
Formaldehyde (50-00-0)
Heptachlor (76-44-8)
Heptachlor epoxide (1024-57-3)
2,3,7,8-HeptachIorod i benzo-p-d i ox i n
Other-HeptachIorod i benzo-p-diox i n
HexachIorobenzene (118-77-1)
HexachIorobutadiene (87-68-3)
a-Hexachlorocyclohexane (319-84-6)
b-HexachIorocycIohexane (319-85-7)
q-Hexachlorocyclohexane (58-89-9)
HexachIorocycIohexane, technical
2,3,7,8-HexachIorod!benzo-p-dioxIn
Other-HexachIorod i benzo-g-dIox i n
HexachIoroethane (67-72-1)
3-Methytcholanthrene (56-49-5)
Methylene chloride (75-09-2)
4,4-MethyIene-bis-2-chIoroaniIine (101-14-4)
MethyIhydraz i ne (60-34-4)
2-Nitropropane (79-46-9)
N-Nitrosodf-n-butylamine (924-16-3)
N-N i trosod i eThy I am i ne (55-18-5)
R-Nitrosodimethyl amine (62-75-9)
R-Ni trosopyrroli d i ne (930-55-2)
PCBs
9.1E-03
1.5E-01
2.0E-03
1.4E+00
1.1E-02
1 .2E+00
1.5E-04
3.0E-03
3.0E-02
1 .6E-04
4.2E+01
3.6E-02
6.7E-01
2.7E-02
4.3E-01
2.3E+00
3.7E-03
1 .OE-01
7.1E-04
1 .6E-03
4.5E-02
3.8E-01
2.0E-01
2.2E-03
7. IE-OS
7.0E-04
1.1E-00
7.1E-00
4.5E-02
8.3E+00
1 .OE-01
8.0E-01
7.7E-03
3.8E-03
2.0E-04
2.0E-02
2.0E-02
5. OE-01
5.6E-03
1 .9E-02
2.6E-02
2.0E-02
5.0E-06
5.0E-04
2.5E+00
3.7E-03
2.4E+00
2.1E-01
3.2E-02
3.7E-04
6.3E-03
2.3E-04
7.1E-04
1 .6E-02
8.3E-03
0.1
0.1
0.1
0.1
1.1
4,500
0.1
0.1
0.1
0.1
25.9
0.1
130
0.1
1,400
450
0.1
0.1
0.1
0.1
0.1
440
18
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
780
0.1
0.1
0.0026 '
0.0026 '
6.2 <
0.1
0.1
0.1
0.1
• 0.1
0.0034
0.008
0.1
0.1
2,800
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
1.57E-06
.57E-06
.57E-06
.57E-06
.72E-05
7.09E-02
.57E-06
.57E-06
.57E-06
.57E-06
4.06E-04
1 .57E-06
2.04E-03
1.57E-06
2.20E-02
7.06E-03
.57E-06
.57E-06
.57E-06
.57E-06
.57E-06
5.90E-03
2.82E-04
.57E-06
.57E-06
.57E-06
.57E-06
.57E-06
.57E-06
.57E-06
.57E-06
.22E-02
.57E-06
.57E-06
M4E-03
U4E-03
J.72E-05
.57E-06
.57E-06
.57E-06
.57E-06
.57E-06
5.39E-08
.25E-07
.57E-06
.57E-06
».33E-02
.57E-06
.57E-06
.57E-06
.57E-06
.57E-06
.57E-06
.57E-06
.57E-06
1.1E-03
6.8E-05
4.9E-03
7.4E-06
8.9E-04
8.3E-06
6.7E-02
3.3E-03
3.3E-04
6.2E-02
2.4E-07
2.8E-04
1.5E-05
3.7E-04
2.3E-05
4.3E-03
2.7E-03
9.7E-05
1.4E-02
6.3E-03
2.2E-04
2.6E-05
5.0E-05
4.6E-03
1.4E-01
1.4E-02
8.8E-05
1.4E-06
2.2E-04
1 .2E-06
1 .OE-04
1.3E-05
1.3E-03
2.6E-03
5.0E-02
5. OE-04
4.9E-04
2.0E-05
1 .8E-03
5.3E-04
3.8E-04
5.1E-04
2.0E+00
2.0E-02
4.0E-06
2.7E-03
4.1E-06
4.7E-05
3.1E-04
2.7E-02
1 .6E-03
4.3E-02
1.4E-02
6.1E-04
1.2E-03
1.72E-09
1 .07E-10
7.68E-09
1.16E-11
1 .54E-08
5.86E-07
1 .05E-07
5.23E-09
5.23E-10
9.72E-08
9.75E-11
4.36E-10
3.04E-08
5.81E-10
5.11E-07
3.07E-08
4.24E-09
1.52E-10
2.21E-08
9.80E-09
3.49E-10
1 .82E-07
1.41E-08
7.21E-09
2.21E-07
2.24E-08
1.38E-10
2.21E-12
3.49E-10
1.89E-12
1.57E-10
1 .53E-07
2.04E-09
4.13E-09
2.04E-09
2.04E-11
4.77E-08
3.14E-11
2.80E-09
8.25E-10
5.96E-10
8.00E-10
1 .07E-07
2.51E-09
6.27E-12
4.24E-09
1 .80E-07
7.37E-11
4.86E-10
4.23E-08
2.49E-09
6.82E-08
2.21E-08
9.57E-10
1.88E-09
B-5
-------
TABLE 82. (CONCLUDED)
Compound (CAS No.)
2,3,7,8-Pentachlorodlbenzo-p-dioxin
Othar-Petanch I orod I banzo-p-d"! ox i n
Pentachloronitrobenzene T82-68-8)
Pronamlde (23950-58-5)
Reserplne (50-55-5)
2,3,7,8-Tetrachlorodlbenzofuran
2,3 ,7,8-Tetrach 1 orod i benzo-p-d t ox 1 n
Other-Tetrach 1 orod i benzo-p-3i ox i n
1 ,1,2,2-Tatrachloroethane~ (79-34-5)
Tetrachloroethylene (127-18-4)
Thlourea (62-56-6)
Toxaphene (8001-35-2)
1,1,2-TrIchloroethane (79-00-5)
Trtchloroethylene (79-01-6)
2,4,6-Trlchlorophenol (88-06-2)
Vinyl chloride (75-01-4)
Total Noncarcinogens
Risk
specific.
dose (ug/nr5)3
4.0E-07
4.0E-05
1.4E-01
2.2E+00
3.3E-03
2.0E-06
, 2.0E-07
2.0E-05
1.7E-01
2.1E+01
1 .8E-02
3.1E-02
6.3E-01
7.7E+00
1 .8E+00
1.4E+00
O.OE+00
95th
percent! le
cone. (ng/L)
0.0016
0.021
0.1
0.1
0.1
0.002
0.0071
0.061
12
220
0.1
0.1
19
130
100
9.7
53,000
Weight
fraction
2.57E-08
3.36E-07
1.57E-06
1 .57E-06
1.57E-06
3.12E-08
1.11E-07
9.53E-07
1 .88E-04
3.39E-03
1.57E-06
1 .57E-06
2.95E-04
2.05E-03
1.64E-03
1 .52E-04
8.27E-01
Unit
risk
(m^/yg)
2.5E+01
2.5E-01
7.3E-05
4.6E-06
3.0E-03
5.0E+00
5.0E+01
5.0E-01
5.8E-05
4.8E-07
5.5E-04
3.2E-04
1 .6E-05
1 .3E-06
5.7E-06
7.1E-06
NA
Weighted
unit risk
(nrVjig)
6.27E-07
8.23E-08
1.14E-10
7.21E-12
4.75E-09
1.57E-07
5.57E-06
4.78E-07
1 .09E-08
1 .64E-09
8.63E-10
5.06E-10
4.73E-09
2.65E-09
8.94E-09
1.08E-09
O.OE+00
Weighted unit risk value
64,000
1.OOE+00 - 9.45E-06
Rounded off to 1.0X10~D
a/
b/
at 10~5 level.
from Table 83.
B-6
-------
TABLE 83
NONCARCINOGENS EMISSION-CONCENTRAT,10NS,.,RACs AND ACTUALS-MAXIMUM...;
"AMI ENT AI R-'CONCENTRATrONS5' Fmr'RFASONABLiE""WORST-CASE-DT SPERSTON COEFF 1C I ENT
COMPOUND
NONCARCINOGENS
Acetonltrfle (75-05-8)
Acetophenone (98-86-2)
Acroleln (107-02-8)
Allyl alcohol (107-18-6)
Bromotnethane (74-83-9)
2-chloro-1,3-buad!ene (126-99-8)
Cresols (1319-77-3)
Df-n-butyl phthalate (84-74-2)
o-Oichlorobenzene (95-50-1)
p-Olchlorobenzene (106-46-7)
01 ch lorodi f 1 uoromethane (75-71-8).
2,4-Olchlorophenol (120-83-2)
Di ethyl phthalate (84-66-2)
Dtmethoate (60-51-5) ,
2,4-Olnltrophenol (51-28-5)
Diphenylanifne (122-39-4)
Endosulfan (115-29-7)
Endrln (72-20-8)
Formic acid (64-18-6)
Hexach 1 orocyc 1 opentad ! ene (77-47-4)
Isobutyl alcohol (78-83-1)
Methomyl (16752-77-5)
Methoxychlor (72-43-5)
Methyl ethyl ketone (78-93-3)
Methyl parathlon (298-00-0)
Nitrobenzene (98-95-3)
Pentachlorobenzene (608-93-5)
Pentach 1 oropheno 1 ( 87-86-5 )
Phenol (108-95-2)
N-phenylenediaraine (108-45-2)
Phenyl mercuric acetate (62-38-4)
Pyridlne (110-86-1)
Selenoureax(630-10-4)
Strychnine (57-24-9)
REFERENCE AIR
CONCENTRATIONS
(yg/m3)
1.0E+01b
1 .OE+02a
2.0E+013
5.0E+OOa
8.0E-01b
3.0E+OOb
5.0E+013
1 .OE+023
1.0E+013
1.0E+018
2.0E+02a
3.0E+003
8.0E+023
8.0E-01a
2.0E+003
2.0E+01a
5.0E-02a
3.0E-01a
2.0E-»-03a
5.0E+003
3.0E+023
2.0E+013
5.0E+013
8.0E-i-01b
8.0E-013
8.0E-01b
8.0E-013
3.0E+01a
3.0E+01a
5.0E+003
8.0E-023
1 .OE+003
5.0E-OOa
3.0E-013
ACTUAL
" MAXIMUM
951 PERCENT ILE AMBIENT AIR
EM 1 SS 1 ON CONCENTRAT 1 ON
CONC. (ng/L) (ug/m3)
i
0.26 1.48E-04
0.1 5.70E-05
0.1 5.70E-05
0.1 5.70E-05
2.1 1.20E-04
0.1 5.70E-05
0.1 5.70E-05
0.1 5.70E-05
95 5.42E-02
86.3 4.92E-04
1.22 6.95E-04
0.5 2.85E-04
31 1.77E-02
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
33..
0.
0.
0.
9..
33.
0.
0.
5.70E-05
5.70E-05
3.70E-05
5.70E-05
5.70E-05
5.70E-05
5.70E-05
5.70E-05
5.70E-05
5.70E-05
I 1.89E-02
5.70E-05
5.70E-05
5.70E-05
5.30E-03
1 .89E-02
5.70E+05
5.70E-05
0.1' 5.70E-05
0.1 5.70E-05
0.1 5.70E-05
SAFTEY
FACTOR3
6.7E+04
1 .8E+06
3.5E+05
"8.8E+04
6.7E+02
5.3E+04
8.8E+05
1 .8E+06
1.8E+02
2.0E-H32
2.9E+05
1.1E+04
4.5E+04
1.4E+04
3.5E+04
3.5E+05
8.8E+02
5.3Et-03
3.5E+07
8.8E+04
5.3E+06
3.5E+05
8.8E+05
4.2E+03
1.4E+04
1.4E+04
1.4E+04
5.7E+03
1 .6E+03
8.8E+04
1.4Et03
1 .8E+04
8.8E+04
5.3E+03
B-7
-------
- ' TABLE 83 (CONCLUDED)
NONCARCIMOGENS EMISSION CONCENTRATIONS, RACs AND ACTUAL MAXIMUM
AMBIENT AIR CONCENTRATIONS FOR REASONABLE WORST CASE DISPERSION COEFFICIENT
COMPOUND
REFERENCE AIR
CONCENTRATIONS
(wg/iri3)
951 PERCENT ILE
EMISSION
CONC. (ng/L)
ACTUAL
MAXIMUM
AMBIENT AIR
CONCENTRATION
(yg/m3)
SAFTEY
FACTOR3
NONCARCINOGENS (CONCLUDED)
1,2,4,5-Tetrachlorobenzene (95-94-3) 3.0E-018
2,3,4,6-Tetrachlorophenol (58-90-2) 3.0E+018
Tetraethyl lead (78-00-2) I.OE-048
Toluene (108-88-3) 3.0E+02b
1,2,4-Trlchlorobenzene (120-82-1) 2.0E+01b
Trlchlorofluoromethane (75-69-4) 3.0E+023
2,4,5-TrIchlorophenol (95-95-4) 1.0E+02a
C1 Hydrocarbons 2.7E+06
C2 Hydrocarbons 2.7E+06
Total Noncarclnogens
0.1
0.1
0.1
550.5
77
0.1
143.6
9600
17000
27700
5.70E-05
5.70E-05
.70E-05
.14E-01
.39E-02
.70E-05
8.19E-02
5.47E+00
9.69E+00
"5.3E+03
5.3E+05
1 .SE-t-00
9.6E+02
4.6E+02
5.3E+06
1.2E+03
4.9E+05
2.8E+05
a Verified oral RfD.
** Inhalation study used as basis for verified oral RfD.
c The safety factor Is the ratio of the Reference Air Concentration (RAC) to the calculated maximum
ambient air concentration. A ratio much greater than 1.0 Is indicative of the margin of safety
available before ambient air concentration would have a health concern.
B-8
-------
APPENDIX C
HYDROCARBON CONVERSION FACTOR
Section 3.4, Step B, specified the equation to convert THC from ppm, measured
as propane, to mg/s for comparison to the THC Screening Limits. In
equation 6:
THC, mg/s = (THC ppm propane) x (Stack gas flow) x 0.028
the constant factor 0.028 is derived from the following equation:
(6.9 x 10"'t)(45.3)
(0.75)(1.5)
where:
6.9 x 10~ = factor to convert units;
45.3 = weighted average molecular weight of the generic list of
carcinogenic and noncarcinogenic compounds listed 1n Tables B-2, B-
3, Appendix B; see Table C-l; g/g-mole.
0.75 = dimensionless factor to adjust the measured THC for the
potential loss of heavy organlcs in the sampling system. This
factor is based on a conservative analysis of the fraction of the
total organic mass emitted from combustion devices that is
nonvolatile and, therefore, has potential to condense or be adsorbed
1n the sampling system precluding detection by the THC monitor; and
1.5 = ratio of response for propane (3.0) to the weighted average
response of the . generic list of compounds (2.15) to a flame
ionization detector, dimensionless' '.
C-l
-------
Appendix B discusses the procedure used to
emissions. The development of a unit risk for the THC emissions was
predicated on several key assumptions regarding the chemical make-up of the
hydrocarbons detected by the flame ionization monitor (FID) and by the
relative gas concentrations of the individual organic compounds. THC
emissions are grouped in carcinogenic and noncarcinogenic fractions. The
carcinogenic fraction consists of a list of all PICs emitted at the 95
percentile of the range of emissions reported from the burning of hazardous
wastes. In spite of this apparent conservativism, test programs have only
been able to identify the molecular make-up of approximately 60 percent of THC
emissions.
As the hydrocarbons burn in the THC monitor using flame ionization
detection, they produce ions which set up a minute current between the
burner tip and the collector electrode. This current is related to the
mass of carbon atoms into the flame. Thus, equal volume concentrations
of methane, ethane, and propane would produce relative constant responses
of 1, 2, and 3 units respectively. Unfortunately, the type of carbon
bonding also has a significant impact on the instrument response. The
response (relative to methane taken as 1.0) varies from 0 for
formaldehyde and formic acid to 0.4 for methylamine (CH3NH2) and 0.76 for
dichloromethane (CH2Cl2). In general, each chlorine atom makes a
reduction in response by 0.12 units. The weighted average response of
all the known carcinogens and noncarcinogens listed in Tables 82 and 83
has been calculated at 2.15 units. (See Appendix B)
C-2
-------
• WEIGHTED AVERAGE
Compound,. .
CARCINOGENS
Aery I amide
Acrylonitrile
Aldrin
Aniline
Benzo(a)anthracene
Benzene
Benzidine
Benzo(a)pyrene
Bis(2-chloroethyl> ether
Bis
-------
TABLE C-1
WEIGHTED AVERAGE MOLECULAR WEIGHT CALCULATION
Compound
Average
other-Hexachlorodibenzo-p-dioxin
Hexach toroethane
3-Hothylchotanthrene
Hethylene chloride
4,4-Hethylen«-bis-2-chloroanilin«
Methyl hydrazine
2-Hitropropane
H-Hitrosodi-N-butylaraine
N-Hitrosodiethylamine
H-Nitrojopyrrolidine
PCBS
2,3,7,8-pentachlorodibenzo-p-dioxin
other*pentachlorodibenzo-p-dioxin
Pentach loroni trobenzene
Pronomide
Reserpine
2,3,7,8-Tetrachlorodibenzofuran
2,3,7,8-Tetrachlorodibenzo-p-dioxin
other-Tetrachlorodibenzo-p-dioxin
1,1,2,2-Tetrachloroethane
Tctrachloroethylene
Thiourea
Toxaphene
1 ,1 ,2-Trichloroethan«
Trichloroethylene
2,4,6-Trichlorophenol
Vtnyt chloride
HU
391
236.74
268.34
84.94
267.16
46.07
89.09
158.24
102.14
96.09
292
356.5
356.5
295.36
256.13
608.7
306
322
322
167.86
165.85
76.12
413.81
133.42
131.4
197.46
62.5
95X Emission
Cone. (ng/L)
0.00711
0.1
0.1
1755.3
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.00125
0.0255
0.1
0.1
0.1
0.00141
0.00157
0.0598
17
297
0.1
0.1
36.7
81.8
0.1
14
Fraction
Present
1.83E-07
2.57E-06
2.57E-06
4.52E-02
2.57E-06
2.57E-06
2.57E-06
2.57E-06
2.57E-06
2.57E-06
2.57E-06
3.22E-08
6.56E-07
2.57E-06
2.57E-06
2.57E-06
3.63E-08
4.04E-08
1.54E-06
.4.38E-04
7.65E-03
2.57E-06
• 2.57E-06
9.45E-04
2.11E-03
2.57E-06
3.60E-04
Average
MU
7.16E-05
6.09E-04
6.91E-04
3.84E+QO
6.88E-04
1.19E-04
2.29E-04
4.07E-04
2.63E-04
2.47E-04
7.52E-04
1.15E-05
2.34E-04
7.60E-04
6.59E-04
1.57E-03
1.11E-05
1.30E-05
4.96E-04
7.35E-02
1 .27E+00
1.96E-04
1.07E-03
1.26E-01
2.77E-01
5.08E-04
2.25E-02
Response
Factor
11.50
1.28
21.00
0.90
11.80
0.40
3.00
7.25
3.25
3.25
12.00
11.50
11.50
5.00
10.95
30.25
11.75
11.50
11.50
1.20
1.10
0.40
9.76
1.50
1.50
5.40
1.75
Response
Factor
2.11E-06
3.30E-06
5.41E-05
4.07E-02
3.04E-05
1.03E-06
7.72E-06
1.87E-05
8.37E-06
8.37E-06
3.09E-05
3.70E-07
7.55E-06
1.29E-05
2.82E-05
7.79E-05
4.27E-07
4.65E-07
1.77E-05
5.25E-04
8.41E-03
1.03E-06
2.51E-05
1.42E-03
3.16E-03
1.39E-05
6.31E-04
HOH-CARCIHOGENS
Acetonftrile
Acetophenone
Acrolein
AllyI alcohol
Bronxxnethane
2-chloro-1,3-butadIene
Cresols
Oi*n-butyl phthalate
o-Dichlorobenzene
p-Dichlorobenzene
0 ich lorodi f luorometharte
2,4-Ofchlorophenol
Dicthyl phthalate
41.05
120.15
56.06
58.08
94.95
88.54
108.13
278.34
147.01
147.01
120.92
162
222.23
0.26
0.1
0.1
0.1
2.13
0.1
0.1
0.1
95
86
1.22
0.5
31
6.69E-06
2.57E-06
2.57E-06
2.57E-06
5.48E-05
2.57E-06
2.57E-06
2.57E-06
2.45E-03
2.22E-03
3.14E-05
1.29E-05
7.98E-04
2.75E-04
3.09E-04
1.44E-04
1.50E-04
5.21E-03
2.28E-04
2.78E-04
7.17E-04
3.60E-01
3.27E-01
3.80E-03
2.09E-03
1.77E-01
1.30
7.00
1.90
2.30
0.95
3.60
6.40
13.50
1.83
1.83
0.60
1.83
1,83
8.70E-06
1.80E-05
4.89E-06
5.92E-06
5.21E-05
9.27E-06
1.65E-05
3.A8E-05
4.48E-03
4.07E-03
1.88E-05
2.36E-05
1.46E-03
C-4
-------
TABLE C.-1 !-=: '
WEIGHTED AVERAGE MOLECULAR WEIGHT CALCULATION
Compound
Average
Dimethoate
2,4-Dinitrophenol
Oipheny lamina
Endosulfan
Endrin
Formic acid
Hexachlorocyclopentadiene
Isobutyl alcohol
Methomyl
Hethoxychlor
Methyl ethyl ketone
Methyl parathion
Nitrobenzene
Pentachlorobenzene
Pentachlorophenol
Phenol
N-phenylenediamine
Phenylmercupic acetate
Pypidine
Selenourea
Strychnine
1,2,4,5-Tetrachlorobenzene
2,3,4,6-Tetpachlorophenol
Tetpaethyl lead
Toluene
1 ,2, 4-Trichlorobenzene
Trichlopofluopomethane
2,4,5-Trichlopophenol
C1 Hydrocarbons
C2 Hydrocarbons
MW
229.28
184.11
169.22
406.95
380.93
46.02
272.77
74.12
162.2
345.65
72.1
263.23
123.11
250.34
266.35
94.11
108.14
336.75
79.1
123.02
334.4
215.89
231.89
323.45
92.13
181.46
137.38
197.45
16
30
95% Emission
Cone. (ng/L)
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
33.2
0.1
0.1
0.1
9.3
33.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
551
77
0.1
144
9600
17000
Fraction
Present
2.57E-06
2.57E-06
2.57E-06
2.57E-06
2.57E-06
2.57E-06
2.57E-06
2.57E-06
2.57E-06
2.57E-06
8.55E-04
2.57E-06
2.57E-06
2.57E-06
2.39E-04
8.52E-Q4
2.57E-06
2.57E-06
2.57E-06
2.57E-.06
2.57E-06
2.57E-06
2.57E-06
2.57E-06
1.42E-02
1.98E-03
2.57E-06
3.70E-03
2.47E-01
4.38E-01
Average
MW
5.90E-04
4.74E-04
4.36E-04
1.05E-03
9.81E-04
1.18E-04
7.02E-04
1.91E-04
4.18E-04
8.90E-04
6.16E-02
6.78E-04
3.17E-04
6.45E-04
6.38E-02
8.02E-02
2.78E-04
8.67E-04
2.04E-04
3.17E-04
8.61E-04
5.56E-04
5.97E-04
8.33E-04
1.31E+00
3.60E-01
3.54E-04
7.30E-01
3.95E+00
1.31E+01
Response
Factor
3.25
5.40
11.25
8.76
10.76
0.00
5.00
3.75
7.75
13.64
1.83
8.00
6.00
6.00
1.83
1.83
5.25
6.00
5.00
0.40
20.50
6.00
5.40
4.00
7.00
1.83
0.64
1.83
1.00
2.00
Response
Factor
8.37E-06
1.39E-05
2.90E-05
2.26E-05
2.77E-05
O.OOE+00
1.29E-05
9.65E-06
.2.00E-05
3.51E-05
1.56E-03
2.06E-05
1.54E-05
1.54E-05
4.38E-04
1.56E-03
1.35E-05
1.54E-05
1.29E-05
1.03E-06
5.28E-05
1.54E-05
1.39E-05
1.03E-05
9.92E-02
3.63E-03
1.65E-06
6.77E-03
2.47E-01
8.75E-01
38842.4
4.53E+01
2.15E+00
Note: Methane was assumed for C1 Hydrocarbons
Ethane was assumed for C2 Hydrocarbons
Tetra CB was assumed for PCBs
C-5
-------
-------
APPENDIX D ..-..-=sfe^
SAMPLE CASES—CO PERMIT DEVELOPMENT
This appendix discusses the development of permit limits for CO emissions from
two hypothetical hazardous waste incinerators. The' primary objective is to
demonstrate the application of the guidance 1n developing CO permit limits
using both the Tier I and Tier II approaches.* Because two permit formats are
allowed, under either Tier I br tier II approaches, the sample cases discussed
below Illustrate permit development using both formats.
1.0 TIER I LIMITS
1.1 Rolling Average CO Permit Format
Thls.sample case assumes that the applicant has opted for the Tier I, 100 ppmv
hourly rolling average CO limit dry corrected to 7 percent 02.
The following ORE trial burn data (or data from a test burn under conditions
equivalent to the DRE trial burn) summarizes the CO emission results:
Test Run 123
CO Emission
(ppmv dry 7% 02)
• Highest Hourly Rolling Average 35 45 70
(C°HHA)
• Highest Peak 120 550 1,010
Note that there are three test runs as specified in the measurement
requirements discussed 1n Appendix A. In spite of some high CO peaks (one in
excess of 1,000 ppmv) the Incinerator was able to meet the Tier I 60-minute
rolling average limit of 100 ppmv (dry, corrected to 7 percent 02). Also note
that the maximum recorded hourly average is only 70 ppmv.
Therefore, the CO permit limit to cover both test conditions for this
Incinerator will be 100 ppmv for an hourly rolling average.
D-l
-------
1.2 Cumulative Time-Above-Limit-Format "
This sample case assumes that the applicant has chosen to obtain a CO permit
limit based on the alternate format which allows a maximum time (cumulative)
in any clock hour with operation in excess of a base CO limit (C060_t) and no
exceedances above an absolute peak (C0p) at any time.
The incinerator has shown compliance with the Tier I limit of 100 ppmv on the
hourly rolling average basis. The emission data for each of the trial burn
test runs are summarized as follows:
CO Emissions
(ppmv dry 7% 02) Run 1 Run 2
Highest Hourly Rolling Average 80 40
Highest Peak 300 200
The permit under Tierj allows a 100-ppmv hourly rolling average even when the
trial burn levels are lower. Thus, under the alternative permit format, the
6,000 ppmv-min is the basis for the permit. Since this is the only criterion
for permit setting under Tier I, any combination of C060_t» t, and C0p is
permitted as long as the total CO mass emission rate is 6,000 ppmv-min. One
permit option is based on the highest trial burn peak which, in the example,
is 1,200 ppmv. The time, t, above a base CO limit can be arbitrarily selected
to be 6 cumulative minutes in any one clock hour. Therefore, the base CO
limit is calculated as follows:
COs, - 100 - 6 t500 - 10°1 = 56 ppmv
60-6
A lower permitted CO peak or reduced time above the base level would result in
higher base CO limits. For example, a permitted peak of '300 ppmv and four
cumulative minutes results in a C056 ~ 86 ppmv. It is important to note that
selection of the right combination of C0p, t, and C060_t limits is very intial
since compliance with the permit will be based on the ability to meet all
three limits independently regardless of the corresponding rolling average CO.
D-2
-------
2.0 TIER II LIMITS
2.1 Rolling Average CO Permit Format
The applicant seeks a permit for a hourly rolling average CO limit. The
applicant has reported THC emission data using separate burn tests performed
prior to the ORE trial burn. These separate burn tests were performed under
conditions equivalent to those investigated during the ORE trial burn. Thus,
the data are applicable to risk assessment for CO permit setting. ORE
compliance was established during the trial burn.
The following table summarizes the CO and THC emission results for this
incinerator.
Test Run 123
Duration (min) 120 120 120
CO Emisions:
(ppmv dry, 7% 02)
• Time-weighted Average Over Test Run 100 300 350
Duration
• Highest Peak 500 1,500 800
Maximum Hourly Average THC
• Concentration (ppmv) 10 20 16
As anticipated by the applicant, the incinerator has not met the Tier I CO
limit.
However, the THC concentration did not exceed the good operating practice-
based limit of 20 ppmv. Thus the permitted hourly rolling average CO is 250
ppmv,
(100 + 300 + 350),
3
D-3
-------
dry corrected to 7 percent 02, which is the time-weighted average for the
entire trial burn. The permitted THC limit is 20 ppmv.
2.2 Cumulative Tlme-Above-Limlt-Format
For this sample case the applicant has chosen to obtain a permit using the
alternate CO limit format. On the basis of preliminary emissions data taken
on this incinerator the applicant anticipates difficulty,in achieving the CO
limits specified in Tier I. Therefore, the trial burn sampling protocol has
included THC continuous monitoring according to stated guidance.
The applicant reports the following data during a successful ORE trial burn:
Test Run 1 2 3
Duration (min) 120 120 120
CO Emissions: .
(ppmv dry, 7%, 02)
• Highest Peak 2,200 300 1,200
• Time-weighted Average over Test Run 400 200 350
Period
Maximum Hourly Average THC
• Concentration (ppmv) 12 8 18
The incinerator has obviously not met the Tier I limits because the CO is in
excess of the allowed 100 ppmv. Therefore, the applicant would apply for a CO
permit under the Tier II approach.
Given that the highest hourly average THC concentration was below the proposed
alternate limit of 20 ppmv, the permit limits will be specified based on the
arithmetic average of the time-weighted average COHA in each run, or 317 ppmv;
(400 + 200 + 350)/3.
Evaluation of the strip chart data on CO emission during that one trial burn
hour with the highest hourly average indicates three major CO peaks as
follows:
D-4
-------
COpi = 800 ppmv, tj = 3 min.
C0p2 = t, 200 ppmv, t2 = 1.5 min
C0p3 = 2,200 ppmv, t3 = 0.5 min
where the times t^, t2» and t% are established from strip chart data.
The permitted time in excess of a given CO level is calculated using Equation
(4) Section 4.4.1 as follows:
t = 0.5 + 1 [(800)3 + (1,200)1.5] = 2.4 minutes
2,200
Where the allowed peak is 2,200 ppmv. The base CO level (C060_t) is given by
Equation (2) as follows:
C057>6 = 317 - 2.4 [2,200 - 317] = 238 ppmv
57.6
The time, t, may be changed from this level provided that there is sufficient
justification. For example, the applicant may envision incinerator operation
with more frequent peaks or longer duration peaks. Thus 1f a longer time is
warranted, the absolute maximum CO level must be reduced accordingly. For a
time of 6 minutes in any clock hour (i.e., 10 percent of the time) the
permitted peak CO level will be:
C0p = (2,200) 2.4 = 880 ppmv
6
and the base limit is:
C054_m1n = 317 - eiSStL^JiLL = 254 ppmv
54
The alternative to this permit option is to allow both 6 minutes in excess of
a level C060_t and the trial burn highest peak, C0p, of 2,200 ppmv. In this
0-5
-------
case, the"v¥Tue C054_min must be reduced to equate the permitted-mass with the
trfal burrTmas s as ~f6TTowsT
C054-min s 317 ' 6f2,200 - 317] = 108 ppmv
54
In summary the three permit options allowed for this sample case are:
Option 12 3
COD 2,200 880 2,200 ppmv
C060 t 240 250 110 ppmv
t(>C060_t) 2.4 6 6 min
THC 20 20 20 ppmv
All CO emission limits are rounded to two significant figures.
Since the Incinerator was operating with an average 10 percent 02 in the stack
gas during the test run, the enforced peak CO limits would translate to 1700
or 750 ppmv, dry basis as measured by the CO monitor and 190, 200 and 86 ppmv
for Options 1, 2 and 3 base limits respectively.
D-6
-------
APPENDIX E
TECHNICAL BACKGROUND DATA FOR THC EMISSION LIMIT
OF 20 PPMV
This Appendix provides information in support of the THC emission limit of 20
ppmv as the preferred approach to waive the 100 ppmv CO limit of Tier I. This
limit 1s Intended to be consistent with good operating practice (GOP) for
Incinerators, boilers, and Industrial furnaces.
This Appendix presents the following support information:
• Allowable THC Emissions Under the Risk-Based Approach;
• Existing Data Base on THC Emissions from Incinerators. Boilers,
Industrial Furnaces, and Municipal Waste Combustors; and
• Calculated Risk Posed by 20 ppmv THC from Incinerators.
1.0 ALLOWABLE THC EMISSIONS UNDER THE RISK-BASED APPROACH
The risk-based approach to waive the 100 ppmv CO limit (see Sections 3.2.2.2
and 4.7 of the guidance document) could allow THC levels of greater than 100
ppmv 1n most cases and as high as 1800 ppmv in some cases. Table E-l
Illustrates the THC levels that would be allowed at eight facilities (five
Incinerator, two boilers, and one cement kiln selected from the existing data
base) using the proposed site-specific risk assessment. These THC levels are
very significant and are well above actual THC emissions measured at these
facilities, and recorded levels from many other combustion sources operating
under good combustion conditions, as shown below.
E-l
-------
TABLE E-l
ALLOWABLE THC LEVELS USING
SITE-SPECIFIC RISK ASSESSMENT
FACILITY
(PERCENT)
STACK GAS GAS
FLOWRATE DISPERSION
(dscfra) (yg/ra /g/sec)
ALLOWABLE THC PPM AT
CORRECTED TO STACK
7% 02 CONDITIONS
Incinerators:
DuPont
Upjohn
TWI
Zapata
Dow (Test 1-3)
Dow (Test 4)
9.7
NA
13.6
11.0
10.9
11.2
12,890
1,872
5,544
777
13,700
15,000
0.4336
1.727
0.2935
2.148
0.375
0.5847
339
472
1,780
1,294
414
250
275
475
942
920
299
175
Boilers:
Site G
Site I
8.7
2.6
4,400
12,600
6.603
1.61
60
58
53
76
Cement K11n:
Lone Star NA
76,500
0.04145
484
484
Note: Allowable THC (corrected to 7 percent 02).1s based on acceptable
risk =* 1 X 10"5, and unit risk = 1 X 10'5 m3/yg. The following
expression 1s used:
THC (ppm 97% 02) »
14
(21 -
15.374 X 10"
DF X dscfm
Where DF = dispersion coefficient and dscfm = gas flowrate.
E-2
-------
2.0 EXISTING DATA BASED ON THC EMISSIONS WHEN BURNING HAZARDOUS WASTE
Tables E-2, £-3, and E-4 summarize test data on incinerators, boilers, and
cement kilns burning hazardous wastes. The data on hazardous waste
incinerators comprises the current data base developed under research test
programs. The"¥ighest-'"test" average ~tHC from these facilities was only 89
ppmv, measured on a dry basis with a fully heated extractive system and flame
ionization detector (FID) continuous monitor' and corrected to 7 percent
oxygen. Most of the average THC emission levels are clearly well below 20
ppmv, and is in most cases below 5 ppmv. Test data from trial burns also
shows that THC emissions are typically very low, e.g., less than 20 ppmv.
The data base on industrial boilers cofiring hazardous waste also shows that
THC emissions from this thermal treatment and energy recovery practice are
typically very low. Test results from research test program on eleven boilers
operating under normal combustion conditions are summarized in Table E-3. The
data indicate that, with the exception of a firetube boiler (SITE B) burning
natural gas at very low steam load, all THC emissions were well below the
recommended THC limit of 20 ppmv.
A large emission data base on industrial boilers burning fossil fuels only was
also evaluated to determine the measured THC emission levels from these
devices. The data, summarized in Table E-5, contains THC emission levels
measured while the boilers were operating under "as-found" conditions and when
combustion modification techniques were Implemented to reduce levels of nitric
oxides (NOX). All the THC data were obtained with a heated sampling system
and monitor calibrated with propane. A review of the data indicates that 16
of 29 boilers emitted more than 20 ppmv corrected to 7 percent 02. A summary
of the THC data by boiler fuel shows that natural gas fired boilers typically
emit more THC than other types. The average THC for all gas fired boilers was
71 ppm compared with 27 ppmv for distillate oil and only 7.6 ppmv for
pulverized coal.
The data base on cement kilns, summarized in Table E-4, also shows that THC
emission levels from this source category are typically less than 20 ppmv as
propane corrected to 7 percent 02. Several of the cement kiln test programs
E-3
-------
TABLE E-2
INCINERATOR CO/THC/DATA FROM RESEARCH TESTS
SITE ID
Plant B
Ross
Upjohn
Zapata
A. Cyanamid
RUN NO.
1
2
3
4
5
Average
1
2
3
Average
1
2
3
Average
1
2
3
4
Average
1
2
3
4
5
Average
02
(PERCENT)
11.8
10.3
10.7
14.3
10.1
11.4
10.4
10.8
10.7
10.6
8.1
8.3
8.4
8.3
8.2
12.0
11.8
11.9
11.0
10.3
12.4
_
12.7
13.0
12.1
AVERAGE CO
(AS MEASURED)
14.8
< 1.0
6.9
7.2
4,300
866
4.8
9.1
4.7
6.2
10.5
11.2
9.9
10.5
1,275
22.2
7.5
8.8
328.4
6.7
19.3
_
13.8
14.3
13.5
ppn DRY)*
(ppn 7% 02)
22.5
1.3
9.4
15.0
5,523
1,114 '
6.3
12.5
6.4
8.4
11.4
12.3
11.0
11.6
1,394
34.5
11.4
13.5
363.5
8.8
31.4
_
23.3
25.0
22.1
AVERAGE THC ppa DRY)*
(AS MEASURED (§ 7X02)
< 1
< 1
< 1
< 1
341
69
< 1
0.9
1.0
1.0
8.9
6.0
3.9
6.3
71.0
1.9
< 1
< 1
18.7
< 1
< 1
_
< 1
< 1
1.0
1.5
1.3
1.4
2.1
438
89
1.3
1.2
1.4
1.3
9.6
6.6
4.3
6.9
77.7
3.0
1.5
1.5
20.9
1.3
1.6
_
1.7
1.8
1.6
HIGHEST RECORDED VALUES
(ppa DRY 07*02)
CO THC
34.2 1.5
1.3 1.3
14.5 2.6
17.6 2.3
6,935.8 671.7 /
-
9.8 1.3
21.3 3.2
11.8 3.1
-
7.3 7.9
7.6 6.0
6.7 .4.1
-
1,717.2 235.2
612.9 63.6
13.4 1.5
28.2 4.5
-
40.2 2.1
60.7 4.2
_ _
43.0 1.7
45.0 1.9
_
m
•£>
-------
TABLE E-2
(CONCLUDED)
INCINERATOR CO/THC/DATA FROM RESEARCH TESTS
SITE ID
Mitchell
DuPont
TWI
DOW
RUN NO.
1
2
3
Average
1
2
3
Average
1
2
3
4
Average
1
2
3
4
Average
02
(PERCENT)
9.4
10.5
• 9.9
9.9
9.2
9.6
10.3
9.7
12.4
13.0
13.2
15.6
13.6
10.1
11.1
11.5
11.2
11.0
AVERAGE CO
(AS MEASURED)
1.4
1.8
< 1
1.4
666
422
624
571
4.3
0.9
1.2
0.6
1.8
1
NA
1
10
4.0
ppra DRY)*
(PPM 7% 02)
1.7
2.4
1.2
1.8
790
518
816
708
7.0
1.6
2.2
1.6
3.1
1.3
NA
1.5
14
5.6
AVERAGE THC
(AS MEASURED
< 1
< 1
0.6
0.9
75.9
47.6
58.1
60.5
2.5
1.9
1.7
0.8
1.7
2.5
2.3
2.1
2.9
2.5
Dpra DRY)*
(§ 7X02)
1.2
1.3
0.7
1.1
90.1
58.5
76.0
74.8
4.1
3.3
3.1
2.1
3.1
3.3
3.2
3.0
4.2
3.4
HIGHEST RECORDED VALUES
(pptn DRY 07*02)
CO THC
6.1 1.3
4.1 1.4
16.3 2.3
1,364.4 166.1
1,854.4 105.4
1,975.7 112.9
.
107.9 4.7
24.2 3.7
4.1 3.9
5.2 5.4
110.5 11.6 :i
11.2 i
3.7 14.7 a
1,028.6 230.0
-
m
en
Sources: MRI "Performance Evaluation of Full-Scale Hazardous Waste Incinerators. Volume 2. Incinerators
Performance Results," EPA-600/2-84-181b, PB85-129518, Nov. 1984.
MRI "Total Mass Emissions from a Hazardous Waste Incinerator," MRI Project No. 8671-L(1), May 1987
* All THC data are measured propane with the exception of DOW Site where THC was measured as methane.
Heated extraction system and heated THC monitor was used. The THC data for this Site was converted to
propane using the following equation:
THC (propane) = THC (methane)/3 (to account for the FID response factor)
, V.
-------
SUMMARY OF TOTAL HYDROCARBON EMISSION (THC) DATA
FROM INDUSTRIAL BOILERS
BOILER IDENTIFICATION/TYPE
Wood-F1red Stoker-SUe A
F1retube-S1te B
Wastetube-S1te C
Converted Watertube-S1te D
Package Watertube-S1te E
Watertube-S1te F
F1retube-S1te G
Watertube-S1te H
Watertube-S1te I
F1retube-S1te J
Watertube-S1te K
PRIMARY
FUEL
Wood
Natural Gas
Natural Gas
No. 6 011
No. 6 011
Natural Gas
None
Pulverized
Coal
Natural Gas
Distillate
011
No. 6 011
WASTE
CHARACT
Creosote
Paint Waste
Phenol Waste
Chi. *
Solvents
Spiked MMA
Spiked Paint
Chlor. Org.
Spiked Met.
AC
N1tr. Waste
Spiked Solv.
Chi. Spike
on
02
PERCENT
10.4
9.6
7.4
12
6.6
8.9
9.5
6.3
2.5
5.4
4
CO
PPM S
7% 02
832
46
16
84
91
96
122
114
93
68
89
THC*
(PPM §
7% 02)
13.22
57.56
0.00
NA
NA
0.71
0.39
0.39
4.47
NA
NA
AVERAGE
151
- Methyl Methacrylate
* Measured as propane with heated extractive system and unheated monitor.
Source: EPA, Engineering Assessment Report Hazardous Waste Coflring In
Industrial Boilers, Volume I, EPA 600/2-4-1772, PB85-187838/AS
E-6
-------
—TABLE E^-4 -- ;^__ ___'- ""*"'
"THC^ANrCO EMISSTONS-FROM^CEMENT-ieiLIIS^HAZARDOUS^WAftr
SITE ID
B
D
E*
F
G
Keystone Bath, PA
K1ln 1
K1ln 2
Citadel /Lafarge
Oemopolis, AL
Lafarge
New Branfels, TX
Ashgrove
.
Lafarge
Lebec, CA
RUN
NO.
1
1
1
1
1
1
1
2
3
9/25
1
2
3
1
2
3
4
5
6
1
2
3
4
02
PERCENT
8.9
5.6
5.4
7.2
11.8
8.0
8.0
14
14.8
17
9.7
8.9
9.5
8.0
8.0
8.0
8.0
8.0
8.0
9
9.8
8.9
8.5
CO
(ppm 7% 02)
220
587
421
40
496
14
40
968
1,210
1,215
NA
NA
NA
Preh/Bypass
1142/84
1416/195
1256/494
1093/292
1305/93
1396/284
79
38
359
112
THC
(AS MEASURED)
7.0
1.2
156.3
1.7
6.3
0.7
1.1
6.6
6.6
11.7
11.0
10.0
9.0
7.0
25.0
9.0
6.0
5.0
22.0
1.7
1.7
2.3
1.7
THC
(PPM 7% 02)
8.1
1.1
140.3
1.7
9.6
0.7
1.2
13.3
15.0
41.0
13.6
11.6
11.0
7.5
26.9
9.7
6.5
5.4
23.7
. 1.9
2.1
2.7
1.9
Source: EPA, Background Document on Boilers and Industrial Furnaces
Hlustlck, Memo to Shiva Garg, titled "Summary of Total Hydrocarbon
Measurements 1n Cement Kilns", dated October 20, 1988.
NOTES: 1)
2)
All THC measurements taken as methane with the exception of
Keystone, LaFarge (TX), and Ash Grove. Reported methane values
were corrected to propane by dividing by 3 to account for the
relative response of the FID monitor.
All THC measurements were made with and unheated FID lines with
the exception of Site E and Ash Grove.
* EPA Method 25, unheated probe, methane basis.
E-7
-------
__>•*-•' TABLE E-5
SUMMARY OF BOILER CO AND THC EMISSIONS AT NORMAL OPERATION AND
LOW NOx CONDITIONS
BOILER
TYPE
FIretube
FIretube
FIretube
FIretube
FIretube
FIretube
FIretube
FIretube
FIretube
FIretube
FIretube
FIretube
FIretube
FIretube
FIretube
FIretube
FIretube
FIretube
FIretube
FUEL
gas
gas
gas
gas
•
gas
gas
gas
gas
gas
No 2 Oil
No 2 Oil
No 2 Oil
No 2 Oil
No 2 Oil
No 2 OH
No 2 OH
No 2 Oil
No 6 OH
No 5 Oil
TEST RUN
37-8
37-4
37-2
38-2
38-1
38-4
39-1
39-3
40-1
40-6
41-6
41-1
41-5
47-1
47-5
48-4
48-1
48-3
49-1
49-3
58-2
58-1
58-5
33-3
33-6
55-1
56-1
57-1
59-6
59-5
59-8
64-1
65-1
65-2
65-4
73-1
34-11
34-8
35-1
DESCRIPTION
OR OPERATION
As Found
High Load
Low Load
As Found
High Load
Low Load
As Found
Low Load
As Found
Low Load
As Found
High Load
Low Load
As Found
Low Load
As Found
High Load
Low Load
As Found
Low Load
As Found
High Load
Low Load
As Found
Low Load
As Found
As Found
As Found
As Found
High Load
Low Load
As Found
As Found
High Load
Low Load
As Found
As Found
Low Load
As Found
NORMAL OPERATION .
AVERAGE
CO
(§ 7% 02)
0
0
0
0
0
121
0
158
428
132
52
0
0
0
10
17
11
108
8
0
5
0
0
0
0
0
0
0
0
0
0
9
0
16
13
0
AVERAGE
THC As C3Hg
(i 7% 02)
4
5
8
12
17
24
15
41
72
67
37
6
2
7
10
13
147
86
8
7
132
25
10
7
6
8
3
16
15
2
7
LOW NOx OPERATION .
AVERAGE
CO
(8 7% 02)
2000
483
0
41
0
16
46
21
2000
0
0
20
0
AVERAGE
THC AS CH4
(S 7% 02)
93
12
5
43
1
5
160
63
7
23
3
COMMENTS
Low NOx with low air
Low NOx with low air
»»»»*»#*»»»»»**»*»»»**»*
Low NOx with high air
Low NOx with low air
»»**»*»*»**»***«*»*»»««*
»»*»»»»**»»*»»»*«*»»»««&
Low NOx with low air
Low NOx with low air
Low NOx with high air
»»*#»»»#»*#*»»*»»»*»»*»»
Low NOx with low air
»«***»»»»»*********»**»»
»#*»»»**»***»»»*»***»»*»
' Low NOx with low air
Low NOx w/low air & low
temp
E-8
-------
TABLE E-5
SUMMARY OF BOILER CO AND THC EMISSIONS AT NORMAL OPERATION AND
LOW NOx CONDITIONS
(CONTINUED)
BOILER
TYPE
Firetube
Firetube
F i retube
F i retube
Firetube
F i retube
Firetube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Wa-tertube
Watertube
Watertube
FUEL
No 6 oi 1
No 5 oil
No 5 oil
No 5 oil
No 5 oil
Stok Coal
Stok Coal
gas
gas
gas
gas
gas
gas
gas
gas
gas
gas
gas
TEST RUN
35-3
36-2
36-1
36-4
44-4
44-1
44-3
45-7
45-1
45-3
46-7
51-1
42-1
43-1
101-2
104-1
109-1
113-2
122-1
12-20
12-22
12-28
13-4
13-10
13-3
140-2
143-3
146-1
14-1
14-6
14-9
153-1
154-1
DESCRIPTION
OR OPERATION
Low Load
As Found
Low Load
High Load
As Found
High Load
Low Load
As Found
High Load
Low Load
As Found
Low Load
As Found
As Found
As Found
As Found
As Found
As Found
As Found
As Found
As Found
Low Load
High Load
As Found
Low Load
High Load
As Found
As Found
Low Load
As Found
High Load
Low Load
As Found
Low Load
NORMAL OPERATION
AVERAGE
CO
(§ 7* 02)
0
88
0
21
0
0
0
10
17
0
0
0
0
252
840
52
380
0
0
0
112
187
137
789
77
103
156
0
0
0
0
0
250
18
AVERAGE
THC As CjHg
(§ 7* 02)
1
1
32
0
11
17
6
7
13
8
10
195
LOW NOx OPERATION
AVERAGE
CO
(8 7* 02)
88
0
16
88
0
0
4
12
2000
789
75
9
0
18
AVERAGE
THC AS CH4
(9 1% 02)
4
18
0
8
7
195
-
COMMENTS
Very few boi lers of
this type
Very few boi lers of
this type
Burner tuned to zero CO
THC recorded after
burner tune
»»»»*»*»«»»»**»»»»»»»»*»
Low NOx w/low air
Low NOx same as baseline
CO reduced to zero with
burner
Low NOx is same as Low
Load
******»**»******»»**»***
E-9
-------
TABLE E-5
SUMMARY OF BOILER CO AND THC EMISSIONS AT NORMAL OPERATION AND
LOW NOx CONDITIONS
(CONTINUED)
BOILER
TYPE
Watartube
Watertube
Watortubo
Watartube
Watortube
Watertube
Watortube
Watertuba
Watartube
Watertube
Watertube
Watortube
Watertube
Watertube
FUEL
gas
'
gas
gas
gas
gas
gas
gas
gas
gas
gas
gas
gas
gas
gas
TEST RUN
15-1
15-6
15-8
180-2
184-1
185-3
190-3
207-1
210-1
24-2
24-3
24-4
25-3
25-4
25-6
30-14
30-11
30-13
4-1
4-2
4-5
5-1
5-2
67-6
67-2
69-1
75-7
75-5
75-2
77-11
77-10
77-5
80-11
80-13
80-9
DESCRIPTION
OR OPERATION
As Found
High Load
Low Load
As Found
High Load
Low Load
As Found
As Found
Low Load
High Load
As Found
Low Load
As Found
High Load
Low Load
As Found
High Load
Low Load
As Found
High Load
Low Load
As Found
High Load
As Found
High Load
As Found
As Found
High Load
Low Load
As Found
High Load
Low Load
As Found
High Load
Low Load
NORMAL OPERATION
AVERAGE
CO
(§ 7% 02)
8
29
' 2000
0
16
0
8
19
0
68
41
25
0
0
32
0
0
0
131
2000
0
126
46
0
0
72
0
0
0
0
0
0
AVERAGE
THC As C3Hg
(i 7% 02)
3
1
0
0
9
11
12
1
0
1
8
10
15
2
3
2
1
1
LOW NOx OPERATION
AVERAGE
CO
(8 7t 02)
2000
249
93
1556
25
21
79
2000
126
0
0
0
AVERAGE
THC AS CH4
(« 7% 02)
1
19
8
6
10
2
COMMENTS
Low NOx with low air
Low NOx w/ low & staged
air
Low NOx w/flue gas
rec i rcu 1 ated
Low NOx is staged
combustion
Low NOx w/low air
Low NOx same as high load
E-10
-------
TABLE E-5
SUMMARY OF BOILER CO AND THC EMISSIONS AT NORMAL OPERATION AND
LOW NOx CONDITIONS
(CONTINUED)
BOILER
TYPE
Watertube
Water-tube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
FUEL
No 2 oil
No 2 oil
No 2 oi 1
No 2 oil
No 2 oil
No 2 oil
No 2 oil
No 2 oil
No 2 oi 1
No 6 oil
No 6 oil
No 6 oil
No 6 oil
No 6 oil
No 6 oil
No 6 oil
No 6 oil
No 6 oil
No 6 oil
No 6 oil
No 6 oil
No 6 oil
No 6 oil
TEST RUN
102-6
107-2
160-1
52-5
53-1
54-5
65-1
65-2
65-3
66-1
66-4
7-10 -
7-5
7-9
10-2
10-7
111-1
116-1
119-1
126-2
170-3
171-1
171-6
176-2
186-1
195-1
1-12
1-8
200-3
204-1 ^
21-4
21-5
21-6
22-1
22-4
29-5
29-4
DESCRIPTION
OR OPERATION
As Found
As Found
As Found
As Found
As Found
As Found
As Found
High Load
Low Load
As Found
Low Load
As Found
High Load
Low Load
As Found
Low Load
As Found
As Found
Low Load
As Found
As Found
Low Load
Low Load
As Found
As Found
As Found
As Found
Low Load
As Found
As Found
Low Load
High Load
As Found
As Found
Low Load
As Found
High Load
NORMAL OPERATION
AVERAGE
CO
(8-7% 02)
70
317
0
37
0
0
0
0
0
0
0
0
0
0
0
0
90
0
0
0
0
0
23
0
0
0
0
0
0
8
0
114
0
0
0
AVERAGE
THC As C3Hg
(8 7* 02)
•
1
3
16
15
9
2
0
2
4
9
6
2
1
LOW NOx OPERATION
AVERAGE
CO
(8 7f 02)
117
86
15
369
6
0
7
0
0
12
35
46
19
68
121
34
109
0
0
0
0
AVERAGE
THC AS CH4
(8 7* 02)
0
5
2
0
6
3
7
17
6
•
1
COMMENTS
Burner tuned to 86 CO
Low NOx with low air
Low NOx with staged air
Low NOx with comb, low
air 4 adj.
Low NOx with low air
Low NOx with staged air
Low NOx with FGR and
staged air
Low NOx with Staged air
Low NOx with gas recirc.
E-ll
-------
TABLE E-5
SUMMARY OF BOILER CO AND THC
LOW NOx CONDITIONS
(CONTINUED)
BOILER
TYPE
Watertube
Watertuba
Watertuba
Water-tube
Wotertube
Watartube
Watertuba
Watertuba
Watertube
Watertuba
Watertube
Watartube
Watertube
Motor-tube
FUEL
No 6 on
No 5 oil
No 6 oil
No 6 oil
No 5 oil
No 6 oil
No 6 oil
No 6 oil
PV coal
PV coal
PV coal
PV coal
PV coal
PV coal
TEST RUN
29-5
2-5
3-2
3-5
3-6
63-6
63-11
63-15
68-2
68-3
68-5
6-6
6-2
6-5
70-2
70-3
70-6
8-5
8-2
8-4
9-1
9-3
9-4
131-4
139-4
156-2
157-1
157-3
169-1
26-1
26-7
26-9
31-1
32-4
32-2
32-3
DESCRIPTION
OR OPERATION
Low Load
As Found
High Load
As Found
Low Load
As Found
High Load
Low Load
As Found
High Load
Low Load
As Found
High Load
Low Load
As Found
High Load
Low Load
As Found
High Load
Low Load
As Found
Low Load
High Load
As Found
Low Load
As Found
Low Load
High Load
As Found
As Found
High Load
Low Load
As Found
As Found
High Load
Low Load
NORMAL OPERATION
AVERAGE
CO
(i 7* 02)
0
209
0
0
27
0
38
0
0
0
10
22~
62
0
0
0
0
0
99
0
27
0
86
40
0
0
0
0
0
0
AVERAGE
THC As C3Hg
(8 7% 02)
2
18
17
17
.
4
4
7
5
2
2
2
LOW NOx OPERATION
AVERAGE
CO
(i 7% 02)
209
0
0
7
0
0
0
98
0
AVERAGE
THC AS CH4
(§ 7% 02)
21
5
1
0
COMMENTS
Low NOx w/low & staged
air
Low NOx w/low & staged
air
i
Low NOx with staged air
Low NOx with staged air
Cyclone furnace. Very
few boil.
E-12
-------
TABLE E-5
SUMMARY OF BOILER CO AND THC, EMISSIONS AT NORMAL OPERATION AND
LOW NOx CONDITIONS
(CONTINUED)
BOILER
TYPE
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
Watertube
FUEL
PV coal
ft
PV coal
PV coal
Stok coal
Stok coal
Stok coal
Stok coal
Stok coal
Stok coal
Stok coal
Stok coal
Stok coal
Stok coal
Stok coal
Ref. gas
Oi 1 & gas
Oi 1 & gas
TEST RUN
71-3
71-1
72-4
72-3
78-1
134-2
165-1
167-2
167-4
16-12
16-6
16-8
17-6
17-10
17-8
18-3
18-13
18-20
19-6
19-5
19-7
20-6
20-7
20-9
27-1
27-10
27-8
28-2
28-11
28-7
149-1
23-1
23-^2
23-6
74-1
74-4
DESCRIPTION
OR OPERATION
As Found
High Load
As Found
High Load
As Found
As Found
As Found
High Load
Low Load
As Found
Low Load
High load
As Found
Low Load
High Load
As Found
Low Load
High Load
As Found
High Load
Low Load
As Found
High Load
Low Load
As Found
High Load
Low Load
As Found
High Load
Low Load
As Found
As Found
High Load
Low Load
As Found
Low Load
NORMAL OPERATION
AVERAGE
CO
(8 71 02)
0
-0
0
0
0
0
19
58
27
0
- 0
0
0
0
0
28
0
285
27
41
29
95
102
77
0
0
0
0
0
0
0
12
10
25
0
0
AVERAGE
I Ml^ AS VVIIQ
(8 71 02)
2
6
4
13
7
4
1
4
4
4
7
9
7
3
4
0
LOW NOx OPERATION
AVERAGE
CO
(8 71 02)
0
0
28
47
0
0
110
5.6
18
0
0
306
9
0
AVERAGE
THC AS CH4
(8 7* 02)
14
2
6
6
6
0
COMMENTS
Cyclone furnace. Very
few bo! 1 .
Low NOx with staged air
Low NOx with low air
Low NOx with low air
Low NOx with low air
Low NOx with staged air
Low NOx w/low air
E-13
-------
TABLE E-5
(CONCLUDED)^
SUMMARY OF BOILER CO AND THC EMISSIONS AT NORMAL OPERATION AND
LOW NOx CONDITIONS
NOTES: (1) THC emissions were measured using a heated extractive and
conditioning system (350° F) and a heated FID analyzer (Beckman
402).
(2) All original THC data are reported as methane. These data were
converted to propane basis by dividing the methane values by a
factor of 3 to account for FID instrument response.
(3) Normal boiler operation includes as-found boiler operation and
operation at high and low steam loads.
(4) Blank entries for emissions indicate no data reported due
primarily to instrument malfunction.
(5) Asterisks in the comment column highlight normal boiler
operating conditions when the THC was greater than 20 ppm as propane
(@ 7% 02) and the corresponding CO was less than 100 ppm (corrected
to 7% 02).
SOURCE: Cato, G.A., et al "Field Testing: Application of Combustion
Modifications to Control Pollutant Emissions from Industrial Boilers
Phase II,11 EPA-600/-76-086a, April 1976.
E-14
-------
a^FMNfflOfR-tgr^ metfratur rather than propane. The THC
emission levels from these test sites (B,D,E,F,G, "Keystone, CrtadeJ»J"anci,
Lafarge) were converted to a propane basis by dividing by a factor of 3 to
account for the relative FID responses. Test result from Site E are of
dubious quality because the measurements were performed with a noncontiguous
test method.
THC emissions from Municipal Waste Combustion (MWC) units have also been
reported to be low, less than 12 ppmv (as propane corrected to 7 percent 02)
in all cases. Examples of these emission levels are given in test results at
the following three MWCs:
Mass Burn Unit - Marion County, Oregon
Spray Dryer and Fabric Filter
Continuous THC - range 0.5 to 2.0 ppmv (cold THC)
RDU - Biddeford, Maine
Spray Dryer and Fabric Filter .
Continuous THC - range 0.7 to 11.5 ppmv (cold THC)
*
Mass Burn Unit - Quebec City, Canada
Spray Dryer and Fabric Filter
Continuous THC - range 4 to 5 ppmv (cold THC and hot THC (Ratfisch RS 5))
Large utility steam generators typically emit THC at less than 1 ppmv.9
Combustion modification for low NOX operation typically increases THC
emissions from these large sources to less than 5 ppmv.
3.0 CALCULATED RISK POSED BY 20 ppmv THC FROM HAZARDOUS WASTE INCINERATORS
In order to determine the level of risk posed by the good operating practice
based 20 ppmv THC limit, several calculations were carried out using the risk
evaluation procedure presented in Section 3.2.2.2 and 4.7 of the Guidance
Reference: Lim, et al., 1980.
E-15
-------
Document. The results of these calculations are shown in Tables E-6 and
.E-7~^TaWe,,E^;i1st£;^ average THC,
the highest THC recorded, and the proposed 20 ppmv limit for five incinerator
facilities. The risk calculated on the basis of 20 ppmv limit is less than
10"5 in all cases. Table E-7 lists the risks calculated for 24 hypothetical
incinerators under several air dispersion scenarios. In all cases, the risk
to the MEI posed by an emission rate of 20 ppm (@ 7 percent 02) is less than
the allowable of 10" .
4.0 REFERENCES
Lim, Waterland, Castaldini, Chiba, and Higgenbotham, "Environmental Assessment
of Utility Boiler Combustion Modification Nox Controls: Volume 1. Technical
Results," EPA-600/7-80-075a, April 1980.
MRI "Performance Evaluation of Full-Scale Hazardous Waste Incinerators. Volume
2. Incinerators Performance Results," EPA-600/2-84-181b, PB85-129518, Nov.
1984.
\
MRI "Total Mass Emissions from a Hazardous Waste Incinerator," MRI Project No.
8671-L(1), May 1987.
EPA, Engineering Assessment Report Hazardous Waste Cofiring in Industrial
Boilers, Volume I, EPA 600/2-4-1772, PB85-187838/AS.
EPA, Background Document on Boilers and Industrial Furnaces Hlustick, Memo to
Shiva Garg, titled "Summary of Total Hydrocarbon Measurements in Cement
Kilns", dated October 20, 1988.
Field testing: Application of Combustion Modifications to Control Pollutant
Emissions from Industrial Boilers Phase I and II, EPA-650/2-74-078-a and EPA-
600/2-76-086a.
E-16
-------
TABLE E-6
RISK DETERMINATION FOR SITE-SPECIFIC INCINERATORS
•'- ' . ' " .- -. . • "- ' ' . - ' '
INCINERATION
FACILITY
Upjohn.
Zapatab
DuPont
TWI
DOM
(Runs 1-3)
(Run 4)
STACK
FLOW
RATE
(DSCM/M)
53
22
365
157
388
425
EFF.
STACK
HEIGHT
(M)
24
11
38
. 21
32
32
DISP
COEF
GEMS
0/ort
1.727
2.148
0.4336
0.2935
0.375
0.5847
AVG
02
(*)
NA
11.0
9.7
13.6
10.9
11.2
AVG THC
AS MEAS
PROPANE
(ppm)
6.3
18.2
60.5
1.7
2.3
2.9
HIGHEST
THC AS
MEAS
PROPANE
(ppm)
8.9
71
75.9
2.5
7.5
2.9
AVG
THC
(rag/sec)
9.3
11.2
618.3
7.5
25.0
34.5
HIGHEST
THC
(mg/sec)
13.2
43.7
775.7
11.0
81.5
34.5
AVG CO
AS MEAS
(ppm)
10.5
378.3
570.7
1.8
10C
MEIa
RISK
AT AVG
THC
1.6E-07
2.4E-07
2.7E-06
2.2E-08
9.4E-08
2.0E-07
MEIa
RISK AT
HIGHEST
THC
2.3E-07
9.4E-07
3.4E-06
3.2E-08
3.1E-07
2.0E-07
NOTE 1: Risk determined for actual dispersion coefficient using GEMS.
NOTE 2: THC is milligrams/sec calculated from following equation:
THC (ppm) * stack gas flowrate (DSCM/M) * 0.028
(a) Based on the weighted unit risk for THC = l.OE-5 mg/)g.
(b) Only test run No. 5 had high THC = 341 ppm. This data point could not be verified and is considered
suspect. Four other test runs showed THC levels less than detection limit.
(c) Measured by EPA mobile laboratory.
-------
TABLE E-7
CALCULATION OF RISK FROM MODELED FACILITIES AT 60P EMISSIONS OF THC = 20 pp«v
INCINERATOR SITE
Flat Terrain
Sources
Site 1
Site 2
Site 2»
Site 3
Site 4
Site 5
Site 6
Site 7
Site 8
Rol 1 ing Terrain
Sources
Site 9
Site 10
Site 11
Site 12
Site 13
Site 14
Site 15
Site 16
Complex Terrain
Sources
Site 17
Site 18
STACK PARAMETERS
TERRAIN HEIGHT
HEIGHT DIAMETER TEMPERATURE VELOCITY EFFECTIVE ABOVE
(METERS) (METERS) ( 'KELVIN) (M/S) STK HT(M) STK HT(M) .
7.3 0.3 325 19.2 11 10
9.8 0.5 339 .5.5 11 2
9.9 0.8 339 2.1 TO 2
24.4 0.5 361 4.9 .25 -1
10.4 0.5 811 6.1 16 8
25.9 0.5 305 17.4 29 -21
27.4 0.2 422 18.6 29 -3
4.9 0.8 1,255 1.2 8 7
7.3 0.5 922 6.4 14 5
9.1 0.5 589 12.5 18 14
7.6 0.8 433 4.0 12 29
6.1 0.8 1,505 2.4 13 17
12.2 0.5 355 21.3 20 29
15.2 0.2 322 18.9 17 15
7.9 0.4 951 10.7 15 16
12.2 0.4 755 13.7 20 21
9.1 0.5 367 5.5 11 45
8.5 0.5 366 2.4 9 52
8.2 0.3 314 35.4 14 131
STACK
FLOW
RATE
DSCFS
43.3
33.0
32.4
39.8
15.3
11.6
14.2
4.95
14.0
43.2
48.0
8.32
121.8
19.0
14.6
23.6
30.4
13.3
82.4
DISPERSION
FACTOR
uG/M3
G/SEC
13.72
8.85
7.55
3.32
2.60
0.91
2,72
10.48
7.74
6.29
6.36
5.60
4.59
3.93
2.92
3.58
9.95
13.62
11.02
ME!
RISK
FOR THC
20 PPM
4.6E-6
2.3E-6
1 .8E-6
1 .OE-6
3.1E-7
8.3E-7
3.0E-7
4.1E-7
8.5E-7
2.1E-6
2.4E-6
3.7E-7
4.4E-6
5.9E-7
3.4E-7
6.6E-7
2.4E-6
1.4E-6
7. OE-6
THC
EMISSION
IN PPM
FOR MjEI
RISK 1D~?
'
" 43
88
105
194
643
242
662
493
• 236
94
84
549
46
342
600
303
85
141
28
... 1 . .
) ;;
m
i
i-*
oo
i \
-------
TABLE E-7 (CONCLUDED)
CALCULATION OF RISK FROM MODELED FACILITIES AT GOP EMISSIONS OF THC - 20 pprav
INCINERATOR SITE
Site 19
Site 20
Site 21
Site 22
Site 23
Site 24
STACK PARAMETERS
TERRAIN HEIGHT
HEIGHT DIAMETER TEMPERATURE VELOCITY EFFECTIVE ABOVE
(METERS) (METERS) ('KELVIN) (M/S) STK HT(M) STK HT(M)
5.2 0.9 1,311 2.4 13 76
9.1 0.5 394 3.0 10 157
10.1 0.3 348 21.6 14 57
10.7 0.2 361 12.2 12 87
7.9 0.4 951 10.7 15 91
8.5 0.3 589 13.1 13 314
STACK
FLOW
RATE
DSCFS
12.1
15.5
45.5
10.9
14.6
16.3
DISPERSION
FACTOR
uG/M3
G/SEC
17.77
7.83
4.95
13.70
9.90
4.98
MET
RISK
FOR THC
20 PPM
1.7E-6
9.5E-7
1 .8E-6
1 .2E-6
1.1E-6
6.3E-7
THC
EMISSION
IN PPM
FOR MEI
RISK 10~5
119
211
114
171
177
315
m
-------
s 1 ess than
10~s in all cases. Table E-7 lists the risks calculated for 24 hypothetical
incinerators under several air dispersion scenarios.In all cases, the risk
to the MEI posed by an emission rate of 20 ppm (@ 7 percent 02) is less than
the allowable of 10~5.
4.0 REFERENCES
Lim, Waterland, Castaldini, Chiba, and Higgenbotham, "Environmental Assessment
of Utility Boiler Combustion Modification Nox Controls: Volume 1. Technical
Results," EPA-600/7-80-075a, April 1980.
MRI "Performance Evaluation of Full-Scale Hazardous Waste Incinerators. Volume
2. Incinerators Performance Results," EPA-600/2-84-181b,'PB85-129518, Nov.
1984.
i
MRI "Total Mass Emissions from a Hazardous Waste Incinerator," MRI Project No.
8671-L(1), May 1987.
EPA, Engineering Assessment Report Hazardous Waste Cofiring in. Industrial
Boilers, Volume I, EPA 600/2-4-1772, PB85-187838/AS.
EPA, Background Document on Boilers and Industrial Furnaces Hlustick, Memo to
Shiva Garg, titled "Summary of Total Hydrocarbon Measurements in Cement
Kilns", dated October 20, 1988.
Field testing: Application of Combustion Modifications to Control Pollutant
Emissions from Industrial Boilers Phase I and II, EPA-650/2-74-078-a and EPA-
600/2-76-086a.
E-20
-------
~ —„,.__ APPENDIX F
LIST OF ACRONYMS
CC Confidence Coefficient
CD Calibration Drift - ^
CE Combustion Efficiency
CEM Continuous Emission Monitor
CEMS Continuous Emission Monitoring System
C/H Fuel Carbon/Hydrogen Ratio
CO Carbon Monoxide
d Arithmetic Mean of Difference Between Values
di Difference Between Values in a Data Set
ORE Destruction and Removal Efficiency
dscfm Dry Standard Cubic Feet Per Minute
E Oxygen Enrichment Level
EC Calibration Error
EPA Environmental Protection Agency
FID Flame lonization Detector
FS Full Scale
GEMS Graphics Exposure Modeling System
GEP Good Engineering Practice
He Terrain-Adjusted Effective Stack Height
HSWA Hazardous and Solid Waste Amendments of 1984
ISLCT Industrial Source Complex Model, Long Term
MEI Maximum Exposed Individual
MW Weighted Average Molecular Weight of the Generic List of Compounds
PAT Permit Assistance Team
PCB Polychlorinated biphenyls
F-l
-------
PIC Product of Incomplete Combustion
POHC Principal Organic Hazardous Constituent
QA Quality Assurance Guideline
RA Relative Accuracy
RAC Reference A1r Concentration -
RCRA Resource Conservation and Recovery Act
RfD Oral Reference Dose
RM Reference Test Method
RSD Risk-Specific Dose
RT Response Time
Sd Standard Deviation
THC Total Hydrocarbon
TSCA Toxic Substances Control Act
USGS United States Geological Survey
UTM Universal Transverse Mercator
F-2
------- |