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

-------