BACKGROUND   DOCUMENT
        STANDARDS APPLICABLE TO OWNERS AND OPERATORS
    OF HAZARDOUS WASTE TREATMENT, STORAGE, AND DISPOSAL
      FACILITIES UNDER RCRA, SUBTITLE C, SECTION 3004
            Proposed Additions to Standards For
                Hazardous Waste Incineration
                (40 CFR 264.342 and 264.343)"'
This document (ms. 1941.33) provides background information
     and support for EPA's hazardous waste regulations
            U.S. ENVIRONMENTAL PROTECTION AGENCY
                        January 1981

-------
                     025216
CONTENTS
I.


II.
III.
IV.












V.
VI.
VII.
Introduction and Background
A. Content of the Background Document
B. RCRA Mandate for the Regulation
C. Key Definitions
Need for Additions to the Final Incinerator
Standards
Role of Risk Assessment in Regulating
Incinerators
Rationale for the Proposed Regulation
A. Emission Limits for Hazardous Combustion
By- Products
B. Variance to the Destruction and Removal
Efficiency
1. Variance Based on Risk Assessment
2. Limitations of the Risk Assessment
Approach
3. Use of Atmospheric Dispersion
Modeling for Incinerators Emitting
Hazardous Wastes
4. Comparison of Regulations to Regulation
Under the Clean Air Act
5. Use of the Linearized Multi-Stage Model
for Cancer Induction
6. Examples and Sample Calculations for
the Variance
7. Use of the Carcinogen Risk Assessment
Strategy
C. Emission Limit for Metals, Hydrogen Halides
and Elemental Halogens
Text of the Proposed Standards
References
Appendix
Page
1
1
3
4
8
12
18
18

27
28

30
35
49
50
58
64
72
74
78
80

-------
I.     INTRODUCTION AND BACKGROUND



       A.   Content of the Background Document



     This is one of a series of documents providinq support and



background information for regulations issued under Section 3004



of the Resource Conservation and Recovery Act of 1976.  Each



Background Document describes a regulation as oriqinallv oro-



posed, summarizes and responds to comments received that relate



to that original proposal, and indicates the Aaencv's rationalp



for final regulations.



     On December 18. 1978, the Aqencv proposed standards for



incineration of hazardous waste (43 PR, at 59008).   As a result



of that proposal, extensive comments were received.  The Agenov



issued a limited set of Interim Status standards on Mav 19,



1980, and responded to some of those comments.  Those standards,



Part 265, Subpart O-Incinerators,  were issued as interim final



standards, subject to comment.



     The Agency has now promulgated Interim Final Standards for



incinerators.  These are the maior technical requirements which



provide the basis for issuing permits under Part 122 of the



regulations.  These standards are  discussed in another backaround



document presenting the rationale for the final incineration



standards, including response to the comments received on the



proposed regulations.



     The Phase II regulations relv on a basic performance standard



(a destruction and removal efficiency at 99.99*) with facilitv



specific operating conditions set  to attain the performance
                               -1-

-------
standard.   The basis  for predicting  compliance with  the perfor-



mance standard is  trial  burns.   These  burns define operating



conditions  associated  with achievement of  the performance standard



The operating conditions thus developed  then become  part of the



permit and  are the basis for continuous  compliance monitoring.



The engineering  judgement of the permitting official is applied



to define acceptable  ranges of  these operating conditions and in



the composition  of the wastes to which they may be applied.



When sufficient  alternative data are available to make these



same determinations,  the permitting  official may waive the



requirement  for  a  trial  burn.



     This new proposal is designed to  complement the Interim



Final Standards.   The  proposed  requirements allow permit writers



to make variances  (e.g., greater or  less than 99.99% ORE) to the



basic performance  standards.



     Specifically,  this  proposal adds  the  following  to §264.343



Performance  Standards;



     (1)    A procedure for a case-by-case  variance from the



           minimum ORE of 99.99% when  protection of human health



           requires a  more stringent standard, or allows a less



           stringent standard.



     (2)    A provision which requires  that the mass emission



           rate  of  hazardous by-products or products of incomplete



           combustion  produced  during  combustion can be no more



           than  0.01%  of the total mass  feed rate of the principal



           organic  hazardous constituents  in the waste.
                               -2-

-------
     (3)    A case-bv-case determination for limitations  on

           emissions of  toxic metals,  hvdroqen halites and

           elemental haloqens.

     The Aqency believes that incineration of oraanic hazardous

waste is the primary near-term alternative to land disposal.

Incineration is capable  of providing safe destruction of wastes.

Larqe volumes of liquid  orqanic wastes not suitable for  land

disposal can be reduced  to safe gaseous emissions and smaller

amounts of residues (ash, scrubber sludges, etc).  Incineration

can thus minimize or eliminate the lonq term impact on human

health and the environment of many hazardous wastes.

     B.    RCRA Mandate  for the Regulation

     The Congress of the United States, in Section T004  of

Subtitle C of the Resource Conservation and Recovery Act (RCRA)

of 1976 (PL 94-580), required that the Administrator of  the

U.S. Environmental Protection Aqency:

     "...promulgate regulations establishinq such performance
     standards, applicable to owners and operators of facilities
     for the treatment,  storage, or disposal of hazardous waste
     identified or listed under this Subtitle, as may be necessary
     to protect human health and the environment.  Such  standards
     shall include, but  need not be limited to, requirements
     respecting -...

     (3)   treatment, storage, or disposal of all such  wastes
           received by the facility oursuant to such operating
           methods, techniques, and practices as may be
           satisfactory  to the Administrator;

     (4)   the location, design, and construction of such
           hazardous waste treatment,  disposal, or storaae
           facililties;"
           (emphasis added).

     The term "treatment" is defined in Section 1004(34) of

the Act to mean:
                               -3-

-------
     "...any method, technique, or orocess, including
     neutralization, designed to chanqe the phvsical, chemical,
     or biological character or comoosition of anv hazardous
     waste so as to neutralize such waste or so as to render
     such waste non-hazardous, safer for transport, amenable
     for storage, or reduced in volume..."

     One objective of  incinerating hazardous waste is normallv

to chanqe the physical form or chemical composition of the

waste so as to render  it non-hazardous.  Incineration mav also

render the waste "safer for transport, amenable for recoverv,

amenable for storage, or reduced in volume."  Therefore, incine-

ration is a treatment process within the meaning of the Act,

and the Agency is mandated to produce operating, location,

design, and construction regulations for the incineration of

hazardous waste adequate to protect human health and the

environment.

     C.   Key Definitions

     The definitions given in Part 260 of the Regulations promul-

gated on May 19, 1980  (45 PR at 33066) should aid the reader

in understanding this document.  Some of those definitions are

provided here for the readers' convenience.  Chanaes from the

definitions proposed on December 18. 1978 (43 PR at 58946) are

discussed if they are relevant to the incineration standards.

1.   "Disposal" means  the discharge, deposit, injection,

     dumping, spilling, leaking, or placing of. anv solid

     waste or hazardous waste into or on any land or water

     so that such solid waste or hazardous waste or any

     constituent thereof may enter the environment or

     be emitted into the air, or discharged into anv

     waters, including groundwater.


                               -4-

-------
2.    "Disposal Facility"  means a facility or  oart  of  a  facilitv



     at which hazardous  waste  is intentionallv  placed into or



     on any land or water,  and at which waste will remain



     after closure.



3.    "Facility" mens all  contiguous land, and structures,



     other aoputenances  and improvements on the land, used



     for treating, storing, or disposing of hazardous waste.



     A facility may consist of several  treatment,  storaae,



     or disposal operational units (e.g., one or more land-



     fills, surface impoundments, or combinations  of  them).



4.    "Fugitive Emissions" means air contaminant emissions



     from non-point emission sources, or other  than those



     from stacks, ducts,  or vents.



5.    "Hazardous Waste" means hazardous  waste  as defined in



     §261.3 of the Regulations promulgated On May  19, 1980



     (45 FR at 33119).



6.    "Hazardous Combustion Bv-Products" (products  of  incomplete



     combustion) are hazardous organic  constituents formed  in  an



     incinerator from incomplete combustion of POHC's to which



     the emission rate limit in 
-------
     (2)   utilizinq the criteria in paragraph (c)  of this
           Section.
7.    "Incinerator" means an enclosed device using controlled
     flame combustion, the orimary ouroose of which is to
     thermally break down hazardous waste.  Examoles of
     incinerators are rotarv kiln, fluidized bed, and liauid
     injection incinerators.
8.    "Principal Organic Hazardous Constituents (POHC's)" are
     the one or more organic constituents in a waste to he
     incinerated to which the Destruction and Removal effi-
     ciency (DRE) standard in §264.341(a) aonlies.   POHC's will
     be designated by the Regional Administrator:
     (1)   prior to a trial burn  (defined under $264.344)
     (2)   based on the results of the waste analysis
           performed under $264.345, and
     (3)   utilizinq the criteria in oaragraoh (c) of this
           Section.
9.   "Treatment" means any method, technique, or nrocess,
     including neutralization, designed  to change the
     physical, chemical, or biological character or
     composition of any hazardous waste  so as to neutralize
     such waste or  so as to render such  waste non-hazardous,
     safer for transport, amenable for recoveryf amenable
     for  storage,  or  reduced  in  volume.
10.  "Trial burn"  means an experimental  burn of  a hazardous
     waste in  an  incinerator  in  order  to evaluate  the
                                -6-

-------
capability of an incinerator of that desiqn to achieve



a specified performance (destruction and removal



efficiency) and to establish the operatinq conditions



(temperature, air flow, etc.) necessary to achieve that



performance for the incinerator oermit.
                          -7-

-------
II.    Need for Additions  to  the  Final  Incinerator Performance
         Standards

     EPA has recognized  that  incineration of hazardous wastes

is one of the primary waste management  techniques which is

preferable to land disposal.   In  preparing  the final standards,

the Agency determined that the  following performance standards

were state-of-the-art and  were  fully  supportable based on existing

data:

     1.    a minimum Destruction  and  Removal Efficiency (ORE)

           of 99.99% based on one or  more Principal Organic

           Hazardous Constituents (POHC's)

     2.    a minimum removal  efficiency of  99% for hydrogen

           choride gas when chlorine  was present in the feed

           in excess of  0.5%, and

     3.    a maximum particulate  emission standard of 180

           milligrams per  dry standard  cubic meter, corrected

           to 12% carbon dioxide  in the exhaust gas.

     However, as these performance standards were being developed

and finalized, the Agency  became  increasingly aware of the limita-

tions of these standards.  Specifically, major limitations of

these standards are as follows:

     1.    The ORE of 99.99%  is a percentage removal or destruc-

           tion standard and  does not provide for regulation of

           the absolute  quantities of POHC's which can be emitted.

           By basing the exit emissions to  the environment on a

           percentage of the  rate of  waste  being fed to the

           incinerator,  a  large incinerator capable of a high
                                -8-

-------
    feed rate would be allowed to emit larqer quantities



    of unburned POHC's.   A small incinerator unit  on the



    other hand, which could handle onlv a small  waste



    feed would be restricted to a smaller quantitv of



    emitted.  Thus the DRR value of 99.99% is based on a



    minimum technoloqy capability and not on the imoact. of



    waste emissions on the environment and human health.



    Also comments received on the 197B proposed  requlations



    expressed a need to develop standards which  reflect



    the deqree of hazard that wastes reoresent.  Wastes



    which represent a low level of hazard to human health



    and the environment should not have as strinaent



    regulatory requirements as hiqhly danqerous  wastes.



2.   The ORE of 99.99% does not account for Hazardous



    Combustion by-products (HCBP) which are known  to he



    formed in the combustion of manv waste substances.



    Documentation on the formation and emissions of HCBP' s



    is increasinq at a rapid rate throuqh onaoino  labora-



    tory test work and field samplinq (5,24). HCRP's



    may in some situations be more hazardous to  the environ-



    ment and human health than the wastes beinq  fed to



    incinerators (POHC's).  The Interim Final Standards



    provide no regulatory control over HCBP's.



3.   Both the ORE of 99.99% and the particulate standard



    of 180 mg/DSCM do not regulate the emissions of inor-



    ganic hazardous constituents such as heavv metals,



    hvdroqen halides and elemental halogens other  than HC1.
                        -9-

-------
           The ORE has been documented only for oqanic compounds



           which can be reduced by oxidation to CO?,  H^O and



           other relatively harmless simple compounds.  Metals



           cannot be destroyed bv oxidation or other chemical



           means and can he emitted in a number of forms such as:



                solids in the incinerator ash



             -  vapors in the exhaust qases



             -  particulates in the exhaust qases



             -  solid or liquid phase in the scrubber effluent.



           In these situations the DRE for orqanic coupounds is



           of no value and the particulate standard of 180 mo/D^CM



           may allow emissions of sufficient quantities of metals



           as solids to endanqer human health and the environment.



           The particulate standard is of no value in the case



           of emissions of vaporized metals.



     4.    The control of emissions of elemental haloqens and



           hydroqen halides other than HC1, is also lackina in



           the Interim Final performance standards.  HC1 is con-



           trolled with the 99% removal requirement hut EP^ was



           unable to find data on emission control svstem perfor-



           mance for the other haloqens and hvdroaen halides.



           Thus no technoloqy based removal standard was



           established.



EPA in identifying these limitations of the final oerformance



standards for incinerators has developed a pronosal to be adde^



to the performance standards in S264.343 which will accomplish



the following:





                               -in-

-------
        Provide for  a  variance  to  the  ORE  of.  <*9.99% which  diresctlv



        links  the  allowable  emission rates of POHCs and  HCBP's



        to anticioated human health  impacts



        Place  a bench-mark limit on HCBP's similar to  the  ORE of



        99.99%



     -  Allow  emission limits to be established  for metals,  ele-



        mental haloqens,  and hydroqen  halides based on their imoact



        on human health.



     Section IV of this Background Document  explains  the rationale



for this proposal and  provides  examples of how thev would  be



applied.
                               -11-

-------
III.  Role of Risk-Assessment in Regulating Incineration



     Many of the comments on the proposed incinerator regulations



(FR December 18, 1978) obiected to the proposal on two manor



grounds:  1) the universally applicable specific design and



operating requirements were too inflexible and not iustifiable



and 2) the performance requirements did not allow variances to



reflect case-by-case situations.  EP^ concluded that these com-



ments had merit.  The differences in wastes and incinerator



designs argue that operating requirements can not be effectivelv



established on a national basis and that performance standards



should be tailored to case-by-case situations to better ensure



protection of human health and also avoid overlv stringent require-



ments.  The Interim Final standards (IF) reflect this approach in



that operating and design standards were drooped and performance



standards promulgated.   However, the IF standards do not orovide



for a case-by-case variance to the basic performance standard.



The question of how a variance to the performance standards



would be determined is a manor Question.



     The use of a risk assessment approach for emissions from



incinerators has been adopted by EPA as the best available method-



ology to link incinerator performance requirements to human



health impacts.  This will be based on an evaluation of risk to



human health posed by the emissions from the incinerator stack.



The proposed standard defines the basic factors involved in this



assessment, including stack emissions, dispersion modeling,



consequent human exposure, and the health effects of the exposure.



It does not define a specific methodoloy.
                                -12-

-------
     If a risk assessment indicates that a more restrictive



emission limit is needed to protect human health, the permit



writer may lower the mass emission rate either bv requiring a



hiqher destruction and removal efficiencv or specifying a lower



waste feed rate or both.  In a like manner, the nermit writer



may approve a lower destruction and removal efficiencv (or hiqher



feed rate) if a risk assessment indicates that no significant



impact on human health will result.



     The factors of stack emission rates and dispersion models



can be addressed utilizing data from enqineerinq calculations,



the trial burn (stack emissions), and existinq air dispersion



models developed under the Clean Air Act.  The determination of



health risk can be based on available EPA "dose response models"



for a certain number of carcinoqens.  For other substances, the



estimation of health effects will be less direct, and will depend



on scientific judqments based on the best health effects data



available, or usinq benchmarks of acceptable exposure such as



threshold limit values (TLV's).  The methodoloqv for conducting



risk assessments is described more fully in Section IV of this



Background Document.



     The Aqency recognizes that in many instances it mav not



be possible to conduct an acceptable risk assessment.  When a



risk assessment is not possible or is not requested bv the anoli-



cant or required by the oermit writer, the performance standard



of 99.99 percent DRE will be the basis for oermittinq.  SPA



expects that over time, data will be developed to expand risk



assessment capabilities.





                               -13-

-------
     The integration of  the variance procedure into the permitt-



ing process could occur  in several ways.   It could be carried



out prior to the trial burn, after the trial burn but before



issuance of a draft permit for public comment, or as a consequence



of public comment.



     A risk assessment to support a variance could be required by



the permit writer or requested by the applicant prior to conduct



of a trial burn.  This would be advantageous in that the appli-



cant would know prior to the trial burn whether he would need



to demonstrate a performance other than 99.99% ORE.  Thus,



he may be able to avoid  having to repeat a trial burn, although



the POHC's and hazardous combustion by-products will not be



finally determined until the trial burn is complete.  In most



instances, the trial POHC's, designated by the permit writer



from waste analysis data included with the trial burn plan,



will also be the final POHC's.  Hazardous  combustion by-products



present a more difficult problem.  While a prediction may be



made, the trial burn may indicate different or additional by-



products than those predicted.  Should this occur, a risk



assessment for those new hazardous by-products would have to



be performed after the trial burn.



     If the applicant were requesting that a risk assessment for



a variance be performance, this request would be included in a



"variance assessment plan" submitted as a  part of the trial burn



plan.  It would include  a description of the proposed methodology



to be used in the assessment.   In reviewing the variance assessment



plan, the permit writer  would  accept or, require modification



                               -14-

-------
of, the risk assessment methodology, and would designate the
trial POHC's and trial hazardous combustion by-products to he
included in the risk assessment.  The risk assessment would be
performed and the results submitted to the oermit writer to
complete the trial burn plan.
     In cases where the applicant does not request a variance
assessment, but instead it is required bv the permit writer uoon
review of the trial burn plan, the applicant would be requested
to amend the plan with a methodology for the risk assessment.
Then the process would proceed as described above.  In either
case, the performance of the risk assessment prior to conduct of
a trial burn, would add a step to the trial burn aoolication
process.  That steo would require that a new part he added to
the trial burn plan.  The permit writer would approve that oart
of the plan, and the applicant would comolete the assessment and
submit it to the permit writer to comolete the trial burn olan.
In addition, the applicant or permit writer miqht decide to
provide an opportunity for public comment on the results of the
risk assessment and the variances determined by the oermit writer
prior to conduct of the trial burn.
     In cases where a waiver of the trial burn is requested in
Part B of the permit application, the same procedure would be
followed reqardinq a risk assessment variance.  It would then
mean that Part B would not be considered final until a determi-
nation of need for, and, where aopropriate, completion of, a
risk assessment were made.
                               -15-

-------
     The second method, for includinq a risk assessment into



the permittinq process is to conduct the assessment following



the trial burn.  A probable basis for requirinq a risk assess-



ment at this time would be that the data from the trial burn



revealed hazardous by-products which were not predicted.  A



risk assessment at this juncture could be requested by the



applicant in his submittal of trial burn results in Part B of



the permit application, required by the permit writer upon re-



view of those results prior to issuance of a draft permit,



or requested by the public as a part of their review of the



permit application or draft permit.  Should this occur, the



applicant would he requested to submit a methodoloqv for the



risk assessment, and upon approval, conduct the assessment and



submit the results to the permit writer, esentiallv as a modifi-



cation of the permit.  If review of the results causes the permit



writer to exercise the variance and alter the performance



standard, a repeat of the trial burn may be necessitated.  If



so a new trial burn plan would be required, in essence return-



ing to the beqinninq of the permit application process.



     In a similar manner limitations on the emission of toxic



metals, elemental halogens and hydrogen halides can be



established.  A preliminary trial burn Plan would include a



proposed methodology for assessing acceptable risks associated



with metals and non-organic haloaens (excent HC1), when apolicable.



Upon approval, this assessment would be comoleted and the data



submitted to the permit writer to complete the trial burn nlan.



Based on these data the permit writer would establish emission

-------
limits on metals and non-orqanic halogens to be achieved



in the trial burn.   In this instance,  as well as in the variance



procedure, the permit writer or applicant may orovide opportunity



for public comment on the results of the risk assessment and



selected performance standards, orior to conduct of the trial



burn.   Thus EPA will determine emission 1imitations of these



inorganic materials in a manner identical to the variance orocedure



for the ORE.
                               -17-

-------
IV.  Rationale for the  Proposed Regulation
A.   Emission Limits  for  Hazardous Combustion By-Products
     Questions raised by  commenters  to the proposed reuglations
published December 18,  1978, led the Agency to recognize that in
additon to defining hazardous constituents in the waste burned
in an incinerator, it is  important to define hazardous combustion
by-products formed during incineration.  Many hazardous wastes
may simply break down and recombine  in an incinerator into other
forms of hazardous organic ocmpounds if combustion is not complete.
Thus, even though the principal organic hazardous constituents
(POHC's) in the waste feed may be destroyed in accordance with
the destruction and removal efficiency standard, the stack gases
may contain other hazardous constituents formed during incineration.
     The Agency has continued to collect evidence that hazardous
combustion by-products  of incineration are a concern.  EPA has
sponsored laboratory  studies at the  University of Dayton of the
thermal decomposition of  complex organic halogen compounds which
illustrate the potential  for formation of these by-products.(24>
Some of the experimental  results are shown in figures 1 through
4 and Table 1.
     Thermal decomposition profiles  were obtained for the tetra-,
penta-, and hexachlorobiphenyl  isomers of PCB, in a flowing air
environment at two  (2)  seconds  residence time, and at the exposure
temperatures shown  in Figure 1.  There was little difference  in
thermal stability noted among the  three  isomers, except the hexa-
form appears to break more sharply between 650° and 700°C.
                                -18-

-------
     Figure 2 illustrates the production of by-product trichlo-

rodibenzofuran from tetrachlorobiphenyl at temperatures of 650°C

up to approximately 750°C.  Figure 3 illustrates the production

of four by-product compounds from the thermal decomposition of

2, 2', 4, 5, 51 -pentachlorobiphenyl in a nitrogen atmosphere.

Figure 4 similarly illustrates the production of several poly-

chlorinated benzene compounds from 2, 21, 4, 4', 5, 5' -hexa-

chlorobiphenyl.  Figures 2-4 clearly demonstrate the gene-

ration of by-product compounds at temperature levels commonly

used in incinerators.

     Table I shows results from the University of Dayton work

with three selected isomers of PCB's (Polychlorinated biphenyls).

It is noted that a variety of benzenes, biphenyls, dibenzofurans

and other chlorinated ocmpounds were detected as decomposition

by-products.

     In order to protect human health and the environment, it  is

essential that a performance standard be applied to these sub-

stances as well.   In some cases the combustion by-products produced

may be more toxic  than the unburned POHC's.(5)  This proposed  amend-
                                               \
ment to the final  regulation requires that  the mass emission  rate

of hazardous combustion by-products must not exceed 0.01 percent

of the total mass  feed rate of POHC's fed to the incinerator.

The rationale for  this standard is as follows:  if  the  combustion

by-product were introduced to an  incinerator as a principal organic

hazardous constituent  (POHC) in the feed, then  it would be subject

to the ORE standard of 99.99%.  Thus the combustion by-products

should be controlled to  the  same  level  as  the  POHC's.


                                -19-

-------
K>
O
I
H
O
a
s
     H
                      100
                  10
                    2
                    2
                    <
                    5
                    UJ
                    or
2
UJ
U
ft
UJ  i n
n  I U
                    O

                    UJ
                      O.I
                      001
                                                             a—
                                                   lf -2.1)0 SCC  IN  Allt
                                          I
                                                      I
                                                             1
                                                        I
                         0
                          50
                                                                                   I
                                                                                           I
         2.2.J.5-

         lErRACIILOROBIPIIfNYL



         2.2.4.5.5* -

         PINIACIIIOROBIPIIENYI
                                                                                              2.2.4.4.5.5-

                                                                                              HtXACIIlOUOBIPIIENYl
                                                                                                  I
                                                                                                                I
                         500    55O    600    650     700    750

                                        EXPOSURE  TEMPERATURE  . °C
80O   850
900
950    1000
                        Source:   Duvall,  D. S.;  Letter  on Research Results Utilizing  the TDAS;  University of
                                  Dayton;  to R. A.  Games, October 1979.

-------
    130
r
a
a
e.
c
u
    0.1
    3.01
2,2,5.5-


TETRACHLOROBIPHENYL




       IN  AIR
                          500
                                           300
                                                            1DOO
                                  EXPOSURE

                                  «?SRAT'JRE

                                    CO
Source:   Duvall,  S. D.; Letter on Research Results Utilizing  the

          IDAS;  University of Dayton;  to  R. A.  Carnes, October  1979,


                                FIGURE   2
                         -21-

-------
     100
3»
•4


I
§



(J
     a.;
    :.ai
                                     -C-
                 2,2.4,5,5- PENTACHLCROBIPHENYL

                               IN

                          I N1T30GENI
                tr -2.00 SEC
                           500
                                            ESC
                                                            '.:00
                                     CC!

 Source:   Duvall, D. S.; Letter on  Research Results Utilizing the

           IDAS; University of Dayton:  to R. A. Carnes, October 1979.



                           FIGURE   3
                            -22-

-------
    :oo
c
II
u
LI
01
-
I
3
1.3
    0.1
    2.0:
                                  2,2.4.4.5,5-

                                  HEXACHLORCBIPHENYL
                                        IN  AIR
                         soo
                                           300
                               TTJffSSATOSS
                                   CO
Source:   Duvall,  D.  S.; Letter on Research Results Utilizing  the

          TDAS;  University of Dayton;  to  R. A.  Games, October  1979,


                             FIGURE   4
                              -2?-

-------
                          TABLE   I

      ORGANIC COMPOUNDS  IDENTIFIED  FROM THE DECOMPOSITION

      OF SELECTED PCB 7SOMERS  IN  FLOWING AIR AT 725°C FOR

                A RESIDENCE  TIME  OF 2  SECONDS
                                             Isomers
     Compounds

trichlorobenzene
biphenyl
tetrachlorobenzene
monochlorobiphenyl
chlorinated compound  MW204+
dichlorobiphenyl
pentachlorobenzene
chlorinated compound  MW230+
trichlorobiphenyl
dichlorodibenzofuran
tetrachlorobiphenyl
pentachlorobiphenyl
trichlorodibenzofuran
hexachlorobenzene
chlorinated compound  MW264+
uetrachlorodibenzofuran
hexachlorobiphenyl
heptachlorobiphenyl
pentachlorodibenzofuran
chlorinated compound  MW288+

•f = tentative  identification

- = not  found
                              •50'i  cc«    95'  4 55'   22*.44'  5.5'
                              2., z.  i j / 3     & i *• / * / 3 / J    & i *• i ** i **  / J i J
                                            2 isomers
                              2 isomers
                              2 isomers
                                 +
                              2 isomers
                              2 isomers
                                 +
+

•f
                                            3 isomers
                                            2 isomers
Source:  Duvall, D. S.; Letter on Research Results Utilizing the IDAS;
        University of Dayton; to R. A. Carnes, October 1979.
                               -24-

-------
     In addition, any combustion by-products detected will be



subject to an assessment of their impact on human health and the



environment in the same manner as the POHC's.  The assessment



method is explained in detail in Section IV-B.



     Hazardous combustion by-products which are subject to the



proposed standard will be designated by the permit-writer.  This



designation will either take place during the course of the trial



burn based on the analysis of the incinerator emissions, or the



owner or operator may present data in the trial burn plan from



laboratory, pilot or full scale burns where hazardous combustion



by-products have been identified.  In cases where a trial burn



waiver is requested, this predictive approach is the only means



of identifying combustion by-products.  EPA has research facilities



which may be used to assist owners and operators in this area.



These facilities are discussed in more detail in the Background



Document on Subpart 0 - Interim Final Standards.





Proposed Regulatory Language



§264.342  Designation of principal organic hazardous



constituents and hazardous combustion by-products.



(a)  Principal organic hazardous consituents (POHCs) and hazardous



combustion by-products must be treated to the extent required by



the performance standards specified in §264.343.



(b)(i) For each waste feed to be burned, one or more POHC's and



hazardous combustion by-products will be specified from among



those constituents listed in Part 261, Appendix VIII of this Chapter,



This specification will be based on the degree of difficulty of
                               -25-

-------
incineration of the organic constituents of the waste feed and its
combustion by-products, their concentration or mass, considerina
the results of waste analyses and trial burns or alternative data
submitted with Part B of the facility's permit aoplication.
Organic constituents or by-products which represent the greatest
degree of difficulty of incineration will be those most likelv to
be designated as POHCs or hazardous combustion by-products.  Consti-
tuents are more likely to be designated as POHCs or hazardous
combustion by-products if they are present in large Quantities or
concentrations.
(ii) Trial POHCs will be designated for performance of trial burns
in accordance with the procedure specified in Sl22.?7(b) for
obtaining trial burn permits.  Trial hazardous combustion by-
products may be designated under the same orocedures.
*****
§264.343  Performance standards.
*****
(d)  Incinerators burning hazardous waste must destroy hazardous
combustion by-products designated under S264.342 so that the
total mass emission rate of these by-products emitted from the
stack is no more than .01% of the total mass feed rate of POHCs
fed into the incinerator.
                               -26-

-------
B.   Variance to the Destruction and Removal Efficiency.



     In the December 18, 1978 proposed standards for incinerators



no variance procedure to the destruction efficiency was proposed.



Among the many comments received on the 1978 proposal were sug-



gestions that a variance procedure should be established to



account for the differences in the degree of hazard of waste



emissions and to reflect specific site-by-site differences.



     EPA agrees that a variance procedure is desirable and some-



times necessary, in order to ensure adequate protection of public



health.  The destruction and removal efficiency of 99.99% minimum,



although the most feasible and defendable state-of-the-art



standard, suffers from a significant short coming.  The approach



is a percent removal approach, and therefore, allows varying



amounts of actual emissions (mass per unit time) depending on the



composition of the waste (concentration of hazardous constituents)



and the rate of feed of the waste.  The 99% removal requirement



for hydrogen chloride suffers from the same shortcomings.



     For example, if 10,000 Ibs of a waste were burned, one pound



of a toxic component would be discharged assuming a 99.99% DRE.



In the case of highly toxic components such as some of the dioxin



isomers, such a quantity of mass emissions may be unacceptable,



depending on the mass feed rate of the dioxin containing waste.



The "pure" DRE approach in the example would permit the owner or



operator to discharge one pound over an unspecified period of time



A large incinerator could be discharging large quantities of



hazardous materials over a period extending from minutes to



years while remaining in compliance with the 99.99% DRE.






                               -27-

-------
     1.   Variance Based on Risk Assessment



     In order to meet the  RCRA mandate of protection of human



health and the environment, a procedure  is provided in the proposed



regulation for a variance  to the DRE based on an assessment of



risk to human health.   No  particular methodology is provided in



these proposed regulations for performing the risk assessment.



However, the Agency  is  providing below a sample approach to risk



assessment which it  believes will satisfy the requirements of



the proposal.  The reader  is cautioned that the sample procedure



presented is but one example of how a risk assessment would be



conducted.



     The conceptual  approach to risk assessment which is presented



as an example in this background discussion is a determination of



individual incremental  risk at the point of greatest ground level



concentration of emissions from the incinerator.  The actual



presence of  individuals at this point, or the number of individuals



is not a factor in the  determination.  This approach is conserva-



tive in protection of health.



     This is a relatively  simplified approach to risk assessment.



It assumes,  in essence, that any individual is exposed to the



greatest ambient concentration of hazardous constituents, regard-



less of where that may  be.  It avoids the difficult and often



disputed estimates of actual total population exposure to diffe-



rent concentrations. A total population exposure analysis could



be performed as a part  of  risk analysis  if desired.  The conside-



ration of population likely  to be exposed can be considered in
                                -28-

-------
making a judgement of an acceptable risk.  The proposed regulation

does not suggest an acceptable risk value.

     The cancer induction model is based on life time average

exposure.  Thus, short term variations in concentrations need not

be determined for cancer induction.  There is a strong history

to support the regulatory and technical basis for the carcingen

risk assessment approach (6,7,8,9,10),  Tne reader should analyze

the referenced material for more details.

     The overall approach to conducting a carcinogen-based risk

assessment is as follows:

     0 Based on data from the trial burn  (or alternate data)
       the mass emission rate of POHC's, hazardous combustion
       by-products, and toxic metals from the incinerator stack
       is determined or calculated.

     0 Appropriate air dispersion models are applied to these
       emissions to predict the ground level ambient concen-
       trations .

     0 Using the greatest level of ambient concentration, a
       cancer risk assessment model (dose response model) is
       applied to determine the risk to an individual of deve-
       loping cancer.

     0 The Regional Administrator (and the public through
       hearings) makes a judgement as to whether this level of
       risk is acceptable.  If it is not, then an acceptable
       level of risk is determined.  Based on this risk, the
       calculations are made in reverse to determine the maximum
       permissible level of stack emissions.

     0 In order to achieve this level of stack emissions, the
       Regional Administrator either can  impose a more stringent
       level of destruction and removal efficiency and/or can
       limit the mass feed rate to the incinerator.

     A risk assessment may not be needed  in every case.  The

impetus for a risk assessment can come either from the Regional

Administrator, the permit applicant, or the public.  The burden
                               -29-

-------
of conducting the risk assessment is on the permit applicant in
all cases.  Further, any risk assessments for the purpose of
lowering the ORE (making it less stringent) must come from the
permit applicant.  When a risk assessment requested by the Regional
Administrator indicates that a 99.99% ORE provides an acceptable
level of risk, the ORE will remain at 99.99% unless there is a
specific request from the permit applicant.
     2.    Limitations of the Risk Assessment Approach
     There are a number of limitations to the application of the
risk assessment variance approach.  They are discussed in the
following paragraphs:
     (i)   The capability to conduct a quantitative risk assess-
ment at the present time is most clearly defined for certain
carcinogenic substances.  Currently, actual dose response data
for inhalation of carcinogens exist for 21 substances (Table II).
These have been evaluated by the EPA Cancer Assessment Group (CAG)
for carcinogenic potency for inhalation.  The data exist in the
technical literature to develop dose response curves for virtually
all of the approximately 150 known and suspected carcinogens.
The EPA Cancer Assessment Group (CAG) expects to develop dose
response data for all of the compounds for which adequate data
exist.
     (ii)  In doing an assessment of risk using cancer data, the
user should be aware that the development of cancer data is a new
science and is subject to degrees of uncertainties and even disa-
greement among those working in the field.  Specifically:
                                -30-

-------
                             Table II
         THE CARCINOGENIC ASSESSMENT GROUP'S PRELIMINARY
           CARCINOGENIC POTENCY ESTIMATES ON COMPOUNDS
                     INCLUDED IN AIR PROGRAM
Compound
Potency Slope (ug/m3)-!
Acrylonitrile

Allyl Chloride

Arsenic

Benzene

Beryllium

Cadmium

Diethyl-Nitrosamine (DEN)

Dimethyl-Nitrosamine (DMN)

Ethylene Dibromide

Ethylene Dichloride

Ethylene Oxide

Formaldehyde

Manganese

Nickel

N-nitroso-N-ethylurea (NEU)

N-nitroso-N-methylurea (NMU)

Perchloroethylene

TCDD

Trichloroethylene

Vinyl Chloride

Vinylidene Chloride
8.50 x ID"5

2.66 x 10~6

3.00 x 10-3

7.40 x ID"6

2.70 x lO'1

1.90 x 10-3

7.18 x ID"2

4.35 x ID"3

5.90 x 10-4

1.20 x lO-5

1.20 x lO"4

6.53 x lO-5

4.0 x ID"4

1.80 x 10-3

6.66 x 10-3

0.67

7.60 x 10~6

121.428

8.80 x 10-7

1.43 x ID"3

5.93 x ID"5
                               -31-

-------
     1.    The degree of certainty for carcinogenic effects is



           different for each compound.



     2.    The value of the potency  (BH) is different for each



           element or compound and directly reflects the degree



           of carcinogenicity.



     3.    The experimental or human exposure data has been



           obtained from different exposure routes, ingestion,



           inhalation, and skin absorption.  Methods of converting



           exposure data from one exposure route to another have



           been developed using reasonable assumptions.  However,



           these transformations introduce additional uncertainities



           into exposure data.  It is important to note that the



           potency (Bjj) slopes for the same element or compound



           may be different for each of the exposure routes.



     Since the field is in a  rapid state of development anyone



applying cancer induction data to assess environmental and health



impacts should insure that the data  is current.



     The cautions above should not be taken as an excuse to avoid



proceeding with a risk assessment based on the cancer effects of



chemicals, as these effects are very real.



     (iii) A variance analysis may also be conducted for substances



which do not manifest carcinogenic effects.  In this case the



threshold assumption may be used in deriving a criterion.  This



assumption is based on the premise that a physiological reserve



capacity exists within the organism which is thought to be depleted



before clinical disease ensues.  Alternatively, it may be assumed
                               -32-

-------
that the rate of damage will be insignificant over the life span



of the organism.  The Threshold Limit Value (TLV)  for worker



exposure to chemical substances published by the American Confe-



rence for Government Industrial Hygienists (ACGIH) is an example



of a threshold approach for non-carcinogens.  The  TLV could be



modified by a "safety factor" to derive an ambient exposure



concentration value.  There are two major drawbacks to the use



of the TLV.  These are:



     1.   The general population contains subgroups more suscep-



          tible than workers, e.g., the old, young children, and



          people with illnesses.



     2.   Workers are exposed typically for eight  hours a day.



          General populations are exposed on a continuous 24-hour



          basis.



     Assessments using the TLV approach therefore  will require



judgment on the part of the permitting official.  The advantages



of using OSHA type standards are:



     (1)  they apply to a wide variety of toxic substances,



     (2)  the TLVs are largely inhalation based,



     (3)  TLVs are continually updated ^D,



     (4)  TLVs are derived directly from experimental human



          and animal studies,



     (5)  they are already a part of the law,



     (6)  a methodology exists for handling mixtures of compounds



          for which TLVs exist (26).  This methodology is for the



          workplace and would be modified for application to



          ambient usage.






                               -33-

-------
The history of each TLV should be examined to assess the extent

to which it has assured worker safety in past application.

     An alternative is to use the multi Media Environmental Goals

(MEGs) which have been produced by EPA for use in performing

environmental assessments.  Some 650 compounds have been analyzed

and ambient concentrations  for most of them have been derived

for air, land and water(25).

     The carcinogen-based risk assessment approach is based on

a non-threshold concept of  disease induction.  The threshold

concept considers a level of environmental contamination below

which there is no adverse effect.  The TLV is based on this

concept:

     "Threshold Limit Values refer to airborne concen-
     trations of substances and represent conditions
     under which it is believes that nearly all workers
     may be repeatedly exposed day after day without
     adverse effeet."(25)

     A concept for criteria setting, including the TLV, has been

proposed by EPA(25).

          (iv) Another factor which complicates risk assessment

is the determination of an  acceptable risk.  This determination

is as much a political/social decision as a technical one.  For

many of the risk assessments the estimated risk will be in a

range of fairly clear acceptability or unacceptability.  However,

for those cases in the "gray area", the judgment of the permitting

official and the reaction of the public in the public hearing

process will impact the determination of acceptability.
                               -34-

-------
     3.   Use of Atmospheric Dispersion Modeling for Incinerators
            Emitting Hazardous Wasted

     Dispersion modeling will be used in this application to

estimate the maximum allowable emissions of each hazardous sub-

stance such that the incinerator impact does not exceed any of

the specified ambient air concentrations related to a specified

risk level.  That is, each incinerator is modeled to determine

emission limits that are specific to that incinerator.   In cases

where several incinerators emit hazardous substances in the same

area, the incinerators should be modeled simultaneously in order

to account for the combined impact of these sources.

     The purpose of this section is to provide an overrview of

the role of modeling in reviewing permit applications for hazar-

dous wastes incinerators, indicate the extent to which procedures

are available to do that modeling, and outline some of the data

requirements for the models.  This discussion points out that

procedures for dispersion modeling are available, are supplemented

with guidance, and are applicable to hazardous waste incinerators.

Persons involved in a modeling analysis pertaining to the incine-

ration of hazardous wastes should be thoroughly familiar with

both the modeling guidelines and the user's manual of the model(s)

selected.  In addition, those conducting the analysis should

have sufficient expertise in air quality modeling to make the

judgments required in the modeling exercise.

     Dispersion modeling will be carried out using the recommen-

dations of the EPA Guideline on Air Quality Models(12)   The guideline

recommends specific models for various situations.  The guideline
                               -35-

-------
also makes recommendations  concerning  the source and meteorolo-

gical data to be used  in  these models.  The guideline is expected

to be updated intermittently.  In  all  cases the most recent

version of the guideline  should be used.

     The Guideline on  Air Quality  Models discusses both screening

techniques and refined modeling techniques(12).*  The screening

techniques are simple  calculations and  tend to be based on

conservative assumptions.   Thus, if  screening  shows that an allow-

able concentration is  not exceeded,  then a more refined technique

will give the same conclusion and  a  more refined analysis is not

required.  If on the other  hand, a screening results in a concen-

tration in excess of the  allowable concentration, it is desirable

to use a more refined  technique for  estimating atmospheric concen-

trations of hazardous  substances.  The  Guideline has been published

in the Federal Register by  EPA and has  been tested in the courts

to some extent (12,13,14).

     Model Selection

     Three factors are most significant in selecting an air

quality model:   (1) The nature of  the  pollutant (i.e., inert vs.

chemically reactive);  (2) the nature of the emission source or

sources (e.g., point source); and  (3)  averaging time (i.e., the
     * Some proposed  revisions  to  the  Guidelines on Air Quality
Models are discussed  in  Regional Workshops on Air Quality
Modeling:  A  Summary  Report   (Draft).   These proposals principally
reflect additional  refined models  which might be recommended and
greater refinement  of guidance  on  various modeling issues.
However, until  revised guidelines  are  issued (expected in Spring
1981), the 1978 guideline should be  followed.
                                -36-

-------
time period over which concentrations are to be averaged).



Other considerations are sometimes important; for example,  the



occurrence of an unusual meteorological phenomenon, terrain



feature or source characteristic will often require the use of a



specialized model.  Nevertheless, these factors are useful in



narrowing the choice of models.



     Hazardous pollutants should generally be considered chemi-



cally unreactive.  This is a conservative assumption in that the



maximum concentration at the point of exposure is derived using



this approach.  Even in cases where the pollutant does decompose



in the atmosphere, it is appropriate to use an inert pollutant



model in conjunction with using a half-life approach to simulate



chemical disappearance of the pollutant.  Chemical removal may



be considered only if the applicant can demonstrate to the



satisfaction of the Agency that the products of atmospheric



reaction have no effect on the health or welfare of man.  Other-



wise, the conservative assumption - that no chemical removal



occurs - should be used.



     For modeling purposes, incinerators should clearly be



considered as point sources.  In most cases, only one source is



to be modeled, but in some cases, it is appropriate to model the



combined impact of several sources using a multi-source model.



     Finally, the averaging time of interest is one year in this



example which is oriented toward limiting the occurrence of cancer



Carcinogenic effects are a function of cumulative exposure to a



compound, and so  it is appropriate to use a long-term average
                               -37-

-------
concentration  (i.e.,  a one-year  average)  to estimate long-term



exposure to a  compound.   For pollutants which exhibit other forms



of toxicity such as TLV's,  other averaging times, appropriate to



the type of health effect,  should be  selected.



     The Guideline on Air Quality Models^2) specifically addresses



only those pollutants for which  a National Ambient Air Quality



Standard has been set.   However,  the  hazardous pollutants being



considered here are analogous  to S02  and  should be modeled as



being chemically inert or as having first-order decay.  Thus,



the techniques used for  incinerators  emitting carcinogens should



be in accordance with the guideline recommendations for esti-



mating annual  average concentrations  of SC>2 resulting from one



or more point  sources.



     When just one incinerator is being considered, the guideline



recommends several suitable screening  techniques.  A useful



discussion of  many of these techniques is provided in Volume 10



(Revised) of the Guidelines for  Air Quality Maintenance Planning



and Analysis,  entitled Procedures for  Evaluating Air Quality



Impact of New  Stationary Sources.(15)  This document provides



step-by-step approaches  for making screening estimates of con-



centrations for cases of flat  terrain  with no significant meteo-



rological complications  and for  more  complex situations.  The



Guideline on Air Quality Models^2) also  references several other



documents which discuss  screening techniques, some of which are



useful for situations not discussed in Volume 10.



     If a refined modeling  technique  is to be used and if one



incinerator is being  considered,  the  guideline recommends using





                               -38-

-------
the Single Source {CRSTER) Model.(16)  Alternatively, other



models with other output formats may be used.



     Screening techniques are not generally available for multi-



source situations.  In the multisource situation direct use of



a refined multisource model is necessary.  Recommended models



of this type for estimating annual average concentrations include



the Climatological Dispersion Model (CDM/CDMQC) for urban cases(17



For rural cases, models described in the Summary Report of the



Regional Workshops on Air Quality Modeling are recommended(18) .



     Data Collection



     If refined modeling techniques are used, it is necessary to



obtain several types of data.  The Guideline on Air Quality Models



discusses four types of data required by air quality models:



source data, meteorological data, receptor locations, and back-



ground concentration.  Source data are primarily used to estimate



emissions rates and  plume rise.  It may be necessary to model more



than one operating condition.  The meteorological data includes



wind speed, wind direction, atmospheric stability, and mixing



height.  These data  must be representative of the meteorological



conditions at the source.  Five years of data should be used to



insure the data are  representative.  Receptor locations must be



be carefully chosen  so that the maximum concentrations is esti-



mated.  The guideline gives specific recommendations on locating



receptors.  Background concentrations are important when the sum



of the background concentration plus the source impact is not



allowed to exceed a  given concentration.
                               -39-

-------
     One additional type of necessary data is information on



whether special circumstances exist that will affect atmospheric



dispersion.  For example,  it is necessary to determine if the



plume is affected by complex terrain, lake/sea breezes, fumiga-



tion, aerodynamic downwash, or deposition.  If so, it may be



necessary to use a model specially designed for those circum-



stances.



     Recommended Procedures



     The following procedure for determination of emission limi-



tations on hazardous substances is generally recommended:



     (1)  A screening analysis will be performed to estimate



          highest atmospheric concentrations of all compounds



          designated as hazardous that are to be emitted by



          the incinerator  (this analysis will assume that the



          incinerator destroys 99.99% of the POHC's introduced



          into the incinerator).



     (2)  Using the health effects information identified for



          hazardous compounds emitted, the increase in cancer



          risk caused by the highest concentrations estimated in



          step (1) for each hazardous substance is calculated.



          If more than one substance is  involved, increases are



          summed to find a total increase in cancer risk.



     (3)  If the total increase in cancer risk caused by the



          incinerator is less than or equal to an acceptable



          increment, then  99.99% destruction and removal may be



          considered adequate.  If the total increase in cancer
                                -40-

-------
          risk is greater than the acceptable increment, then



          steps (1)  and (2)  should be repeated using different



          values of acceptability.  Additional modeling may be



          desirable also.



     (4)  If a more refined  modeling technique also shows an



          increase in cancer risk that exceeds the acceptable



          increment, maximum allowable emission limits must be



          determined.  These limits may be calculated by assuming



          a linear relationship between emissions and risk.



     (5)  After a maximum allowable emission rate is determined



          from the air modeling and health risk computations, the



          permit writer must exercise his best engineering judg-



          ment to determine  how the emission limit will be



          controlled. He may do this by requiring 1} a demonstra-



          tion of a higher ORE, 2) specifying in the permit a



          maximum feed rate  limit for one or more wastes or their



          components, or 3)  a combination of both 1) and 2).



     It should be noted that in some situations the guideline



does not recommend both a screening technique and a more refined



model.  First, there are situations where applicable refined



models are not available.  When possible, an attempt should be



made to develop refined techniques; however, in many cases



screening techniques will be the only viable option.  Second,



there are also situations, most notably when two or more sources



affect the same area, when no suitable screening techniques



exist.  If one of the types  of modeling techniques is not



possible, the above procedure should be shortened accordingly.






                               -41-

-------
     Figures 5-8 give  flow diagrams of the analyses necessary to

determine emissions limitations for incinerators of hazardous

wastes.  Four possible  situations exist:  a single incinerator

emits one hazardous substance  (Fig. 5), one incinerator emits

several hazardous substances  (Fig. 6), multiple incinerators

emit the same single hazardous substance  (Fig. 7), and multiple

incinerators emit several hazardous materials  (Fig. 8).  The

analyses for the cases  of a single incinerator are essentially

equivalent to the procedure described previously.  Note that

"estimated concentration" is  shorthand for the highest annual

average concentration  chosen  from a number of  receptors.  The

analysis for the case  of several incinerators  emitting the same

one substance is also  similar  to the procedure discussed above

except that no screening analysis is performed.

     As Figure 8 illustrates,  the analysis for the case where

several incinerators emit several hazardous substances should be

conducted somewhat differently from other analyses to facilitate

location and determination of  the maximum risk.  In these ana-

lyses, emissions (units of, e.g., g/sec)  are used in the model

to estimate ambient concentrations (e.g.,  g   x sec =  g ) ,*
                                          sec   m3     m3

and concentrations are  then converted into risk factors (e.g.,

g__ x risk units j. mj. =  risk units) .  In the multiple incinerator,
^3         g

multiple substance case, the  emission rate for each substance

should be multipled by  the concentration-to-risk conversion factor
     sec
     m3  are  the  units  of dilution  (X/Q).


                               -42-

-------
                  Use Screening
                  Technique to
                    Estimate
                  Concentrations
                                        r
      \/	
  Use Model
 To Estimate
 Concentrations
                    Translate
                 Concentrations
                  Into Risks
Add Risks


                                                        \/
  Translate
Concentrations
  Into Risks
                    Refined
                     Model
                    Availabl
                                                          No
                                                    Allocate
                                                 Risk Reductions
                                                     Among
                                                  Substances
Figure 6.  Recommended Procedure for
One Incinerator Emitting Several
Hazardous Substances.
                                      -43-

-------
                                  Use Multiple
                                  Source Model
                                  to Estimate
                                Concentrations
                                   Translate
                                 Concentrations
                                   Into Risk
                                                        99.99%
                                                       Removal
                                                     is Adequate
                                         No
                                    Allocate
                                Risk Reductions
                                 Among Plants
Figure  7.  Recommended  Procedure for
Several Incinerators  Emitting  One
Hazardous Substance
                                         -44-

-------
                                      Convert
                                     Emissions
                                     to "Risk
                                     Emissions"
                                     Add "Risk
                                     Emissions"
                                    for each plant
                                     Use Multiple
                                     Source Model to
                                     Estimate Increase
                                       in Risk
                                    Allocate Risk
                                     Reductions
                                    Among Plants
                                     and among
                                     Substances
Figure 8.  Recommended Procedure for
Several Incinerators Emitting Several
Hazardous Substances
                                            -45-

-------
                                   Use Screening
                                   Technique to
                                     Estimate
                                   Concentration
                                      Translate
                                   Concentration
                                     Into Risk
                        99.99%
                       Removal
                      Is Adequat
                                        fined
                                      Model
                                      Avail-
                                        able 2
                                    Use Model
                                   To Estimate
                                   Concentration
                                    Translate
                                  Concentration
                                    Into Risk
Figure  5-  Recommended Procedure
for One Incinerator Emitting One
Hazardous Substance
                        99.99%
                       Removal
                       Adequate
Acceptable
    7    >
   Scale
   Down
Emissions
                                          -46-

-------
for the substance.  This generates quantities labeled here as

"risk emissions" (e.g.,  g  x risk units • m3 = risk units • m3) .**
                        sec         g                 sec

Multiplying the emissions by a risk weighting places the emissions

of all substances on a common basis.  The "risk emissions" for

each plant can then be added to give a total "risk emission" for

the plant.  These "risk emissions" may be used in the model to

estimate risk directly (e.g., risk units • m3 x sec = risk units).
                                    sec         m3

Using this approach, the task of locating the receptor with the

highest overall risk burden is greatly simplified.***  Using the

total "risk emissions" as input (in place of emission rate) the

multisource model output will display risk factors which have

already been summed for all species and all incinerators.

     The "risk-emission" approach is not required for simpler

cases since there is no ambiguity about the location of the

maximum risk.  (In the case of one incinerator emitting several

substances/ the maximum concentrations of the substances can all

be expected to occur at the same location).
**   As an example:  assume that the emission rate is 2 g/sec
and that a concentration of 10"^ g/m3 causes a risk factor of
10~7.  Then the "risk emission" would be 2 x 10-7/10~6 =0.2
risk units-m3/sec.

***  Location of highest risk is difficult using the conventional
approach.  It is likely that different substances will have their
maximum concentrations at different locations.  The location of
greatest total risk may not correspond to the location of any of
the individual substance maxima.  Thus determining the maximum
total risk would necessitate a tedious process of converting
concentration to risk factors at a large number of receptors.
                               -47-

-------
     One other  noteworthy feature  of  Figures  5-8  is  that once



modeling is performed  to estimate  risks  based on  99.99% destruc-



tion, no further  modeling is necessary to  determine  emissions



limitations.  Modeling essentially provides a linear relation-



ship between  emissions and concentrations.  As a  result, a given



percentage reduction  in concentration  (and associated risk) is



achieved by reducing  emissions  by  the  same percentage.  A word



of caution, however;  if two or  more incinerators  are controlled



to different  degrees,  the location of  the  maximum concentration



may change.   In such  cases, it  is  advisable to confirm, possibly



via an additional model simulation, that the proposed emission



reductions will in  fact result  in  risks  a_t all locations being



within acceptable limits.



     It is important  to note that  the  specific location of the



point or points of  maximum concentration is not important to the



strategy proposed in  this document.  It  is sufficient only to



know (1) that they  do  exist and (2) what the worse case concentra-



tion(s) are or  could  be.   With  these  two facts, the  health impact



(risk) can be determined  regardless of whether a  receptor is in



fact located  at the point of maximum concentration.



     One exception  to  the above strategy might be envisioned.  If



an incinerator  were located in  a remote  area where extensive air



modeling could  demonstrate that no person(s) would be significantly



impacted under  any  meteorological  condition (at the  point of



maximum concentration)  then a less restrictive emission rate and



accompanying  higher ambient air concentration could  be allowed.
                               -48-

-------
     4.    Comparison of Regulations to Regulations Under the
           Clean Air Act

     It  is useful to place these regulations in the context of

general  approaches to managing atmospheric contamination and in

the context of comparable regulations under the Clean Air Act.

The management of atmospheric contamination can proceed by either

the air  quality management approach or the emissions management

approach.   In the air quality management approach, the regulations

specify  a target air concentration, and modeling is used on a

case-by-case basis to determine the emissions limitations neces-

sary to  avoid violating the target air concentration.  In the

emissions management approach, the regulations directly specify

emissions limitations (e.g., pounds of emissions per ton of

manufactured product) without regard to the case-by-case impact

on air quality.

     Regulatory actions pursuant to the Clean Air Act provide

examples of both types of management approaches.  Examples of the

air quality management approach include the National Ambient Air

Quality Standards (NAAQS) and the program for Prevention of Sig-

nificant Deterioration (PSD).  The NAAQS are concentrations not

to be exceeded more than once per year in any location, and the

increments under the PSD program represent maximum allowable

increases in concentration for areas meeting NAAQS.  Examples of

the emissions management approach include New Source Performance

Standards (NSPS) and National Emissions Standards for Hazardous

Air Pollutants (NESHAP).  The NSPS are emission standards for

pollutants having an established NAAQS, and the NESHAP are emission
                               -49-

-------
standards that apply without  regard  to  the case-by-case impact

on air quality.

     These proposed regulations  represent a mix of the air quality

management approach and  the emissions management approach.  That

is, sources of hazardous pollutants  must meet an emission standard

based on a 99.99% removal  but also may  be evaluated for air

quality impact on a case-by-case basis  to assess the need for

more stringent emissions limitations.   These regulations differ

from regulations under the Clean Air Act in that these regulations

are concerned with the combined  effects of several pollutants

rather than the effect of  each pollutant individually.  Thus,

these regulations could,  for  example, limit the combined increase

in risk to cancer rather than limiting  concentrations of individual

chemicals.  However, these regulations  are similar to the regula-

tions for PSD in that the  concern is with the degradation of

air quality beyond the existing  base line and that if more than

one source locates in an area the combined impact must be

considered.

     5.   Use of the Linearized  Multistage Model for Cancer
            Induction

     Definitions

     0    Carcinogenic potency;  The proportionality constant,

          BH, between the  lifetime average daily exposure concen-

          tration to an  agent, C, and the incremental lifetime

          risk of cancer due  to  that exposure alone, R.  It is

          defined by the equation R  = BjjC.  The numerical value

          of BH is determined by the human or animal data on
                               -50-

-------
          the carcinogenic effect of the agent.  The units of



          BH must be compatible with the units of C, so that



          the product of B^ and C is a dimensionless quantity,



          since R is a probability and has no units.  BH is the



          slope of a linear potency curve which passes through



          the intercept (C=0, R=0).



     0    Threshold;  A certain exposure, usually expressed as



          a concentration in air or water, below which a given



          adverse effect does not occur.



     0    Threshold Limit Value (TLV);  An air concentration of



          an agency established by the American Council of Govern-



          ment Industrial Hydienists, below which continued



          exposure would not result in adverse impact on health



          of workers.  Therefore, TLV is a threshold concentration.



     Induction



     Defining the health impacts of exposure to a given hazardous



substance relies on the use of a dose response model for cancer



induction.  The dose response model has been developed by the



Cancer Assessment Group of EPA and others.  It has been used in



other ongoing and planned EPA regulations, including National



Emissions Standards for Hazardous Air Pollutants (NESHAPS),



groundwater quality standards, and for other purposes.



     In quantitatively assessing the public health risk of air



emission sources, the ambient air concentration of each toxic



chemical is one of the most critical parameters.  For toxic



effects besides cancer and mutagenesis, most authorities agree
                               -51-

-------
that a threshold concentration  exists below which no response



occurs.   If the chemicals  have  caused cancer in animal experiments,



they generally are  regarded as  potential human carcinogens, and



the risk  is proportional to the long-term average concentration,



the proportionality constant  of this risk being called BH-



Refer to  Table 2 for potency  (slope) values developed to date.



     The  lifetime cancer incidence  in the general U.S. population



from all  causes is  about 0.25,  and  the extra risk due to expo-



sure to a chemical, R  (called the incremental cancer risk), is



equal to  R = BH * C, where C  is the lifetime average concentration



of that chemical.   In  cases where several chemicals are present,



the risks may be assured to be  additive, so that the total risk



can be represented  by:



     R =    BHi Ci



     R is proportional  to  the total dose a person receives in his/



her lifetime, and is expressed  in terms of lifetime average



daily exposure.  According to this  model, a person exposed to a



given concentration for just  n  years out of an assumed lifetime



of 70 years, will have  a risk of only n - 70 times the risk



experienced by a person exposed to  the same degree for a whole



lifetime.  If the exposure changes  during the lifetime, then the



time-weighted average  is the  appropriate quantity to use for C^.
                                -52-

-------
                             Table II
         THE CARCINOGENIC ASSESSMENT GROUP'S PRELIMINARY
           CARCINOGENIC POTENCY ESTIMATES ON COMPOUNDS
                     INCLUDED IN AIR PROGRAM
Compound                                 Potency Slope (ug/m3)-!
Acrylonitrile                            8.50 x 10~5

Allyl Chloride                           2.66 x 10~6

Arsenic                                  3.00 x 10~3

Benzene                                  7.40 x 1Q-*

Beryllium                                2.70 x lO"1

Cadmium                                  1-90 x 10'3
                                                    *y
Diethyl-Nitrosamine  (DEN)                7.18 x 10~z

Dimethyl-Nitrosamine  (DMN)               4.35 x 10~3

Ethylene Dibromide                       5.90 x 10~4

Ethylene Dichloride                      1.20 x 10~5

Ethylene Oxide                           1.20 x 10~4

Formaldehyde                             6.53 x lO"5

Manganese                                4.0  x  10~4

Nickel                                   1-80 x 10~3

N-nitroso-N-ethylurea (NEU)              6.66 x 10~3

N-nitroso-N-methylurea  (NMU)             0.67

Perchloroethylene                        7.60 x  10~6

TCDD                                    121.428

Trichloroethylene                        8.80 x  lO"7

Vinyl  Chloride                           1-43 x  10~3

Vinylidene  Chloride                       5.93 x 10"5


                                -53-

-------
     The potency  factors, BH, are estimated from epidemiological
data where possible and  from chronic animal carcinogenic experi-
ments when appropriate human evidence  is not available.  Methods
for deriving BH are discussed in the Water Quality Methodology
Paper.(8,10)
     In the regulation of emissions from incinerators, an
approach presented as an example is to  first establish an accep-
table or target lifetime individual risk level, and from that
calculate what air concentration limit  must prevail in order to
keep the lifetime risk below the target level.  By means of air
dispersion modeling, the upper limit ambient concentrations are
converted into maximum emission rates  from the incinerator stack.
This, in turn, is translated into incinerator waste feed rates.
     Because each chemical  has its own  characteristic carcino-
genic potency, BH, the total risk will  be determined by a potency-
weighted sum of ambient  air concentrations.  This sum dictates
the critical parameters  affecting the  incinerator operation.  A
technique is discussed in the previous  section.
          Acceptable Risk
     Two quantitative measures of risk  have been used by the
Agency in evaluating carcinogenic hazards to populations exposed
to the agent.  These are:   (1) the individual lifetime cancer
risk, which is defined as the probability that an exposed person
will die of cancer, as opposed to other causes, as a result of
exposure, and  (2) the number of cancer  cases per year which can
be attributed  to  the exposure.  The individual risk depends on
the carcinogenic  effectiveness of the  compound, which is called

                               -54-

-------
its potency, and the concentration of the agent in the exposure
medium, whereas the number of cases depends on the individual
risk and the size of the exposed population.
     In deciding what risk is acceptable from a public health
protection point of view, the EPA regulatory offices have concen-
trated on the individual risk.  For example, the Pesticides Office
is considering a lifetime risk of 10~6 as acceptably low in the
case of nitrosamine contamination of pesticide products.  The
Water Quality Office is requiring the reporting of hazardous
material spills into navigable water that could be used as a
source of drinking water if the risks are greater than 10~6.   In
the Food and Drug Administration regulations of animal feed addi-
tives that could cause residues of carcinogenic substances in
edible meat, a risk of less than 10~6 is considered safe enough
to require no use restriction.  The water quality criteria for
the protection of human health were based on a risk range of 10~7
to 10-5.(10)
     The attitude of many scientists and policy makers is that a
risk of less than 10~7 is usually too small to justify the
resources required to issue and enforce a regulation.  A risk  of
greater than 10~4 is usually considered serious enough either  to
take regulatory action or to require a determination that the
costs of control are prohibitively large.  Within the range of
roughly 10~7 to 10"4 the acceptability of a risk is usually a
result of cost-benefit balancing.  The Agency has not made a judg-
ment on what constitutes an "acceptable" risk level.  The reader
is referred to the Water Quality Criteria documents (10) and other
sources for more background.
                               -55-

-------
     For noncarcinoqenic substances it is qenerally believed that
risks are zero if the exposure is less than a certain low concen-
tration, or threshold.
     For these two classes of compounds the following approaches
could be used:   (1) carcinoqens with no currently available
potency value could be assiqned a value based on their structural
similarity to chemicals for which a potencv value is known, and
(2) noncarcinoqens could be treated as threshold pollutants and
an acceptable ambient air concentration could be established as a
certain fraction  (say one-tenth) of the Threshold Limit Value
(TLV).  Adopt inq  the unmodified TLV as the acceptable concentra-
tion for general  population exposure is not advised because it is
desiqned for healthy people (factory workers) who voluntarily
assume the risk  of exposure in order to work at their nobs.
Multi-media environmental qoal values  (MRG's) could also be used.(
     For chemicals without an established TLV, the Procedure
outlined in the  health methodoloqy of  the Water Quality Criteria
documents (^ u) could be followed.  It is a procedure for sett inq
acceptable limits based on toxicity information.
          Multistage Mod el
     The mathematical model chosen for extrapolation of cancer
risks from the hiqh doses used in animal experiments to low doses
of environmental  exposure is the linearized multistaqe model.
This approach is  described in the Water Qualitv Criteria document
procedures.(10)   Althouqh that document describes the procedure
in detail, the major features are reoeated here.
                                -56-

-------
     The model is a general curve-fitting procedure which can fit



a monotonically increasing function of dose to the animal tumor



incidence data.  The assumption is made that the tumor incidence



is linearly related to the dose with no threshold.  This is in



accord with the assumption of other regulatory agencies, as



manifested by the Interagency Regulatory Liaison Group (IRLG)



guidelines for the evaluation of carcinogenic risk.*19)  For some



compounds, a threshold at low doses might exist.  If this were



the case, then the extrapolation procedure used here is regarded



as giving a reasonable upper limit of the risk at low doses (i.e.,



is conservative).  Future research on mechanisms of carcinogenic



action might result in a more definitive quantitative statement



of risk.  The structure of the proposed regulation would allow



use of any technically acceptable approach.



Fundamental Cancer Model Assumptions



     The linear non-threshold model assumes that the lifetime



total dose is the basis for the risk estimation.  In the animal



experiments which form the basis for the procedures the dose



rate is usually constant throughout the lifetime.  If the same



lifetime dose were given within a short fraction of the lifetime,



then the result would be the same under the linear assumption.



However, some evidence exists that large amounts given within a



short time cause more damage, at least for acute  toxicity effects,



than the same amount spread out over a much larger time.  There



is some evidence that these non-linear dose rate  effects also



occur in carcinogenesis experiments, at least  in  the one case of
                               -57-

-------
vinyl chloride.  However, the  information on this point is very
sparse.  The only generality that can be made is that as long as
the dose rate for environmental exposures is less than the equi-
valent dose rate in the lifetime animal bioassay with constant
dosing, the potency factors derived according to these procedures
can be used.  But if the dose  rates become significantly larger,
then calculated risks might be too small.  The correction factors
for these short range high dose situations are unknown.  It is
unlikely that environmental doses would be higher than experimental
doses.
     The instances of pulsed doses to populations which are large
enough to exceed the animal dose rate are expected to be very
rare, since animal doses are typically many orders of magnitude
larger than environmental doses.
     6.   Examples and Sample  Calculations for the Variance
     This section presents data and examines calculations to
illustrate how the variance (override) approach may work.  In
general, conservative approaches have been adopted to illustrate
certain points.  Different acceptable risk levels and model
assumptions have been selected to illustrate the sensitivity of
the approach.
     For assessment of carcinogens the process begins with the
types of data illustrated in Table 2.  "The Carcinogenic Assess-
ment Group's Preliminary Carcinogenic Potency Estimates on
Compounds included in Air Program".  The data are used in the
following equation:
                               -58-

-------
     Concentration  =       -Risk
                         Potency (slope)

Where:

     Concentration  =  The lifetime average exposure level to a
                       carcinoqenic substance.

              Risk  =  The incremental increase in cancer deaths.

           Potency  =  The estimated slope of the straight 1ine
                       which passes throuqh the intercept ooint
                       of the dose-risk curve.

By using this relationship, calculation of the long-term air

concentrations for anv risk level can be done auicklv.   As seen

in Table 2, a wide range of potency slopes have been determined

all the way from trichlorethylene with a value of 8.8 x 10~7

(ug/m3)"1 to TCDD with a potency of about 121 (ua/m3)"1.

     The next steos are illustrated in Table 3.  Usina the potency

slope (BH) for a number of organic comounds and five metals, the

air concentration (C) in mq/m3 is calculated for a risk of 1 x in~s(

Next a dispersion model is selected, assumptions are made for the

model and the allowable stack emission rate (E) is determined.  In

the last column the maximum allowable waste feed rate is calculated

based on the maximum allowable emission rate (5).  This calculation

is straight-forward and is also shown in Tables III and IV for

different air dispersion model conditions and risk levels.

     Table 4 illustrates the impact of different conditions

imposed in the air modeling exercise.  In this table the risk is

the same (1 x 10~5) but the stability class is more stringent

and the effective stack height has been lowered from 150 meters

to 30 meters.  The allowable emission rate and thus the allowable
                               -59-

-------
                          TABLE 3 :  ESTIMATION OF ALLOWABLE MASS FEED BASUU ON A RISK Of IxlO"5,
                                                    ASSUMING A DUE OF 99.99%
                                                                             E**
Allowable Haste
Bli* C
Compounds Potency Slope Concentration(mg/in3)
(mg/m3)-! Associated with 10"^ risk
Aery Ion it rile
Allyl Chloride
Dimethyl-Nitrosamine
N-nilroso-N-Me thy 1 urea
Manganese
Nickel
Beryllium
Cadmium
TCDU
Tr ichloretylene
8.5 x ID"2
2.66 x 10-3
4.35
670
0.40
1.8
270
1.9
121428.57
B. 80 x 10-4
1.18
3.76
2.30
1.49
2.50
5.56
3.70
5.26
8.23
1.14
x ID"4
x 10-3
x 10-6
x 10"8
x 10~5
x 10-6
x 10"8
x ID"6
x Ifl— 1 1
x 10-2
Allowable E
Stack Emission rate (g/hr)
Rate (ing/sec)
59
1880
1.15
7.45 x 10-3
12.50
2.78
1.85 x ID"2
2.63
4.12 x ID"5
5700
212.4
6768
4.14
2.68 x 10-2
45
10
6.66 x 10"2
9.47
1.48 x 10-4
20520
Input (kg/hr)
Assuming 99.99%
Efficiency
2124
67680
41.4
0.26B
450
100
0.666
94.70
1.48 x
205200
*   These are preliminary estimates and arc subject to change

**  Assumptions:  I.  Sunny summer afternoon, wind speed measured at 10 meters is 4 in/sec (stability class  B).

                  2.  Effective stack height is 150 meters.

                  3.  Open flat country.

                   4.   Single point source.

-------
                          TABLE 4  :   ESTIMATION OP ALLOWABLE MASS FEED BASED ON RISK APPROACH,
                                               ASSUMING A ORE OF 99.99% AND A RISK OF 1x10-5
                                                                             E**
                                                                                            Allowable Waste
Compounds
Acrylonitrile
Allyl Chloride
Dime thy 1-Nitrosamine
N-nitroso-N-Methylurea
Manganese
Nickel
Beryllium
Cadmium
TCDD
Trichloretylene
Bll* C
Potency Slope Concentration(inq/m3)
(mg/m3)-l Associated with 10~5 risk
8.5 x 10-2
2.66 x ID-3
4.35
670
0.40
1.8
270
1.9
121428
8.80 x 10-4
1.18 x
3.76 x
2.30 x
1.49 x
2.50 x
5.56 x
3.70 x
5.26 x
8.23 x
1.14 x
10-4
10-3
10~6
10-8
10-5
10~6
10-B
10-6
10-11
10-2
Allowable
Stack Emission rate
Rate (9/hr)
11.3
361
0.22
1.34 x 10-3
2.4
0.53
3.6 x 10-3
0.50
7.9 x 10"6
1094
Input (kg/hr)
Assuming 99.99%
Efficiency
113
3610
2.2
1.34 x
24
5.3
3.6 x
5.0
7.9 x
10,940



10-2


10-2

10-4
1
*   These are preliminary estimates and are subject to change

**  Assumptions:  I.  Wind speed is 4 in/sec (stability class A)

                  2.  Effective stack height is 30 meters.

                  J.  Open flat country.

                  4.  Single point source.

-------
feed rate at a 99.99%  ORE  has  been dramatically reduced by a

factor of nearly  20  times.   These examples  illustrate the dramatic

impact of stack height on  ground level concentrations of emissions.

     Table 5 shows the same  calculations using the same air

dispersion model  but decreasing the  risk level to 1 x 10~6.  This

change in risk level causes  both the allowable emissions rate and

the maximum feed  rate  at 99.99% ORE  to drop by a factor of 10.

This illustrates  the sensitivity of  the procedure to the selection

of an acceptable  risk  level.

     As examples  of  how this analytical process would impact a

real world incinerator, the  restriction on  the destruction of

TCDD containing waste  from 2,4, 5-T  production is determined as

follows:

          The waste  contains 300 ppm of TCDD

          From Table III,  the  allowable feed rate at 99.99%
          ORE is  7.9 x lO-4  kg/hr

          7.9 x 10~4 kg/hr (TCDD) 7  300 ppm =2.63 kg/hr of waste

          Thus an incinerator  would  be restricted to a maximum
          feed rate  of 2.6 kg/hr (5.810/hr) to result in a risk
          of 1 x  10-5.

     An incinerator  burning  this waste would be restricted to a

very low feed rate of  the  2,4,  5-T waste.   If the owner or operator

of the incinerator could demonstrate that a higher ORE could be

achieved then a higher feed  rate could be allowed.  In cases

such as this the  permitting  official would  have the option of

either requiring  a higher  DRE,  restricting  the feed rate of the

waste or a combination of  both.
                                -62-

-------
                          TABLE 5 :   ESTIMATION OK ALLOWABLE MASS FEED BASED ON A RISK OF lxlO~6
                                                     ASSUMING A ORE OP 99.99%
                                                                             E**
Allowable Waste
 Input (kg/hr)
Compounds
Aery Ion it rile
Allyl Chloride
Dime thy 1-Nitrosamine
N-nitroso-N-Me thy 1 urea
Manganese
Nickel
Beryllium
Cadmium
TCDD
Trich lore ty lone
BH*
Potency Slope
(mg/m3)-l
8.5 x l
-------
     7.   Use of the Carcinogen Risk Assessment Strategy
     This discussion explains how the risk assessment approach
may be used from two directions:
     (a)  Starting with  an acceptable risk statement and working
backward to a stack emission limit  (i.e., a limit on hazardous
waste feed to the incinerator).
     (b)  Initially stipulating emission rates and deriving expo-
sure concentrations which are related to an acceptable risk
statement.
     The approach adopted for the purpose of the standard whether
manipulated in the "forward" or "backward" mode is a conservative
approach from the point  of view of  exposure of individuals to
hazardous materials in the ambient  air.  It is conservative in
the sense of once having chosen an  acceptable risk limit, e.g.,
1 x 10~6, each individual is assumed to be exposed to that degree
regardless of where in the immediate area that individual might
spend most of his time.  A complete dispersion model analysis of
the area surrounding a hazardous waste incinerator would clearly
result in different risk exposures  depending on proximity to the
incinerator.  The dispersion model  calculation is related basi-
cally to a zone of maximum concentration downwind from an incine-
rator stack and the basis for the entire risk assessment analysis
is therefore, this zone  of maximum  exposure.  Thus, the conserva-
tive nature of the approach.  Each  individual is assumed to be
exposed at a risk level  in this example of 1 x 10~6 even though a
more detailed analysis,  both nearer to and farther from the
incinerator stack could  possible result in a lower exposure risk.

                                -64-

-------
     As a part of the overall risk assessment, an analysis of
total population exposure could be performed.  Such an analysis
would yield the probability of health damage, e.g., cancer
induction for the people in the general surrounding area, or
even the Nation as a whole based on this one source of hazardous
material.  Such analyses are time consuming and difficult to
perform.  The result of such analyses do not enhance the protec-
tive aspect of the risk assessment approach, since the individual
incremental risk approach is conservative.
     To further understand how the override approach would affect
decisions for some typical cases, the reader is referred to Table 6
which presents results of several sample calculations.  For these
examples an acceptable risk level of 1 x 10-6 (probability of
increased risk of contracting cancer for an individual during a
70 year lifetime) has been chosen to perform calculations.  Other
risk levels could have been used.  The approach to override or
underride the ORE of 99.99% is suggested for use only where dose-
response or threshold data are already available.
     Table 6 presents four manufacturing processes representing
several wastes streams (column 1) and typical or average discharge
quantities for these wastes.  From available dose-response data,
a choice of an acceptable risk level will result in associated
ambient air concentration (column 3), and from this a stack
emission rate of the chemical may be estimated (column 4).  The
calculated incinerator feed rate is then derived (column 5).
                               -65-

-------
                               TABLE 6
        SAMPLE CALCULATIONS COMPARING ALLOWABLE INCINERATOR




FEED RATES AT IxKT6 RISK WITH TYPICAL PLANT WASTE PRODUCTION RATES
(1)
Manufacturing
process
waste
I Vinyl Chloride
Monomer plant
wastes^ 19)

Electronic
components
Manufacturing
Solvents(2°)
""
Textile
Processing Wool
Scouring Sludge
(21)

f (2)
Typical or
Average Plant
Waste Compo-
nent (MT/yr.)
544-1,1,2 Tri-
chloroethane
326-Ethylene
dichloride
1.2-1,1,1,
Trichloroe thane
1.06 Perchlo-
roethylene
0.6-Manganeso
0.04-Nickel
0.0003-Arsenic
(3)
Air Concen-
tration Asso-
ciated with
lxlO~6 Incre-
mental Risk
(mg/m3)
1.13xlO-3
8.3 xlO-5
1.13x10-3
1.3xlO-4
2.5x10-6
5. 56xlO-7
3.3x10-7
(4)
Modeled
Emission
Rate (gm/hr)
108
8.0
108
12.6
0.24
0.53
0.032
(5)
Calculated
Incinerator
Feed Rate
( MT/yr . )
99.99% ORE
8510
631
8510
993
0.0019*
4.2x10-4*
2.6x10-4*
(6)
Limitations for
on-site Incinera-
tion
None

None

If 100% of metals
were emitted, on-
site incinera-
tion would be
prohibited

-------
                                         TABLE  6  (CONT.)
                        SAMPLE CALCULATIONS COMPARING ALLOWABLE INCINERATOR

                FEED RATES AT lxlQ-6 RISK WITH TYPICAL PLANT WASTE PRODUCTION RATES
(1)
Manufacturing
process
waste
Petroleum
Refining (Total
of 17 waste
streams( 1)

F (2)
Typical or
Average Plant
Waste Compo-
nent (MT/yr.)
0.0275-Arsenic
0.32 Nickel

(3)
Air Concen-
tration Asso-
ciated with
IxlO"6 Incre-
mental Risk
(mg/m3)
3.3x10-7
5.6x10-7

(4)
Modeled
Emission
Rate (gm/hr)
0.032
0.053

(5)
Calculated
Incinerator
Feed Rate
(MR/yr.)
99.99% ORE
2.6xlO-4*
4.2x10-4*

(6) 1
Limitations for
on-site Incinera-
tion

If 100% of metals
were emitted, the
total waste stream
could not be in-
cinerated at
lxlO~6 risk
* For these calculations it was conservatively assumed that
  100% of the metal feed passed through into the emissions.

     (a)  Air Model Assumptions:

          1.  Effective Plume Height - 30 meters
          2.  Stability Class A
          3.  Wind Speed - 4 m/s

-------
     The conclusions  in column  (6) were developed based on a



comparison of the calculated  incinerator  feed rate (column 5)



with the quantity of  waste component  in the typical waste stream



(column 2) at a destruction and removal efficiency of 99.99%.



     In the case of vinyl chloride monomer production using a



1 x 10~6 incremental  increase in cancer risk and 99.99% ORE of



the waste components, the 1,  1, 2 Trichlorethane waste component



would not be restrictive.  Similarly, the average quantities of



solvents from electronic component manufacturing could be incin-



erated with no restrictions.



     The third and fourth cases shown in  Table 6 indicate signi-



ficant problems.  In  these examples,  the  metals content of the



textile sludge and the petroleum refining wastes would prohibit



incineration of the total quantity of these wastes assuming that



100% of the metals in the waste were  emitted to the atmosphere.



However, if only a small amount of these  metals were emitted and



most (say 99.99%) were retained either in the ash or in the



scrubber then these wastes could be incinerated on-site with no



restrictions.



     The variance approach may  be viewed  in another way.  Incin-



erators will be required to meet the  99.99% ORE for a wide



variety of wastes.  It is instructive to  examine the application



of the ORE to some typical wastes.  Table 7 shows the results of



applying ORE = 99.99% to the  same wastes  for which the previous



calculations were performed to  yield  allowable incinerator feed



rate, starting with predetermined acceptable risk levels.  In
                                -68-

-------
                                    TABLE  7

RISK EVALUATION FOR 30,000 METRIC  TON/YEAR  INCINERATOR BURNING HAZARDOUS WASTE
(1)
Waste
30,000 MT/yr -
Petroleum
Re fin ing ( 1)
Vinyl Chloride
(19)
Monomer
Textile^21)
Wool Scouring
Sludge

Electronic
Components
Waste Solvents
(20)
(2)
Hazardous
Components
MT/yr .
0.48-Arsenic
.561-Nickel
1428-1,1,2
Trichloroe thane
857-ethylene
dichloride
0.038-Nickel
0.003-Arsenic
0.62-Manganese
8760-Trichloro-
ethylene
1830-Perchloro-
ethylene
(3)
Emission Rate*
g/hr
3.1 x 10-7
3.6 x 10~6
18.1
10.9
0.24
0.02
3.9
111.0
23.0
(4)
Ambient Concen-
tration from Air
Model Calculation
mg/rn-*
3.2 x ID"12
3.8 x 10-8
1.9 x ID'4
1.1 x ID"4
2.5 x 10-6
2 x 10-7
4.1 x lO-5
1.1 x lO-3
2.4 x 10~4
(5)
Incremental Risk
from Dose Response
Calculation

9.6 x ID"12
6.8 x lO"8
1.7 x ID"7
1.3 x ID"6
4.5 x 10~6
6 x 10-7
1.6 x lO-5
1 x 10-6
1.8 x ID"6
  *  For Organics emission rate is based  on 99.99%  ORE;  for metals a 95% removal
     rate is assumed.

-------
this case, an average  incinerator capacity  (off-site) was deter-
mined from a survey of  23  incinerators(23)  to be 30,000 wet
Mt/yr.  The assumption  is  made  that the  incinerator will operate
at full capacity over  the  period of one  year.  The quantity of
POHC discharged under  these burn conditions is then modeled for
dispersion in the ambient  air to the  receptor.  The resultant
risk level (column 5)  is then calculated based on the ambient
air concentration at the receptor determined through dispersion
modeling.

Proposed Regulatory Language
§264.343 Performance Standards
(e)  After consideration of the factors  listed in paragraph (g)
of this Section, the Regional Administrator may, on a case-by-case
basis, establish performance standards which are either more or
less stringent than those  required by paragraphs (a) and (d) of
this Section based on  a finding that:
     (1)  More stringent standards are necessary because the
     emission rates achieved by the application of the perfor-
     mance standards otherwise  required  by  this Section may
     pose an unacceptable  risk  to human  health and the environ-
     ment, or
     (2)  Less stringent standards will  achieve emission rates
     which do not pose  an  unacceptable risk to human health and
     the environment.
(g)  The findings under paragraphs (e) and  (f) of this Section
will be made after evaluating the following data, which the
Regional Administrator  may require from  the permit applicant:
                                -70-

-------
(1)  Emissions of POHC's, hazardous combustion by-nroducts,
metals, and hydroqen halides, includinq:
     (i)  Mass emission rates from the stack, and
    (iij  Concentration in the qas stream exitina the stack;
(2)  Air dispersion estimates for these substances,
including:
     (i)  Heteoroloqical data,
    (ii)  Descriotion of the air disoersion models,
   (iii)  Assumptions underlyinq the air dispersion models
          used;
{3)  Expected human and environmental exoosure, includinq:
     (i)  Topoqraohic considerations,
    {ii)  Population distributions,
   (iii]  Population activities, and
    (iv)  Modes,  intensity and duration of exposure;
(4)  Consequences of exoosure, includinq:
     (i)  Dose-response  curves for carcinoqens,
    (ii)  Health  effects based on human or animal studies
          for other toxic constituents,
   (iii)  Potential for  accumulation of toxic  constituents
          in the  human body,  and
     (iv)  Statements of  exoected risk to  individuals or
          populations.
                           -71-

-------
C.   Emission Limits on "Metals/ Hydrogen Hal ides and Elemental
       Halogens

     The proposed regulations included toxic metals and halogen

compounds in the destruction efficiencv reauirement.  Commenters

objected on the basis that non-organic components cannot be

thermally destructed and that 99.99 percent removal in the flv

ash and bottom ash is not feasible.

     The destruction and removal efficiencv approach could he

applied to metals and non-organic halogen compounds, because it

considers removal of waste constituents in the emission control

system and ash.  Thus, metals and non-organic halogens emitted

could potentiallv be controlled in this wav and included in a

destruction and removal efficiency calculation.  However, the

Agency elected not to apply a ORE standard to metals and non-

organic halogens in the final regulation because the Agencv does

not have test data to indicate what specific removal levels are

achievable, except in the case of hydrogen chloride emissions.

     In the case of hydrogen chloride sufficient data is avail-

able to determine that air pollution control equipment can

consistently remove 99% of the HC1 contained in incinerator com-

bustion gases.  This determination is reflected in the interim

final regulation in §264.343{c).  This is further discussed in

the Interim Final Incinerator Background Document.

     The Agency also considered whether metals and non-organic

halogens were adequately addressed through standards developed

under the Clean Air Act.  The onlv existing standard applicable

to hazardous waste incinerators addresses beryllium, which is
                               -72-

-------
controlled through a National Emission Standard for Hazardous



Air Pollutants (NESHAP).   A NESHAP standard for mercurv applies



to sludge incinerators but not hazardous waste incinerators.



(See the discussion under II, B - other Federal Requlations).



     For metals, other than beryllium, and for non-organic halo-



gens, this proposed regulation reauires that emission limits he



set on a case-by-case basis by assessing the risk to human health



using the same criteria established for assessing a variance to



the basic DRE requirement.  For metals for which EPA has deve-



loped dose response models, health effect assessments using



those models could be made.  For other metals or for non-organic



halogens, emission assessments could he made using available



health effects assessment data including TLV's or MEG's.  (See



IV-B - Variance to the DRE for a complete discussion of the



methodology of health effect assessments.



Proposed Regulation Language



§264.343 Performance-Standards



(f)  After consideration of the factors listed in paragraph (g)



of this Section, the Regional Administrator mav, on a case-bv-



case basis, stipulate performance standards for metals, hv^rogen



halides, and elemental halogens, based on a finding that such



standards are necessary to limit the emission rates of. these



constituents to levels which do not pose an unacceptable risk to



human health and the environment.
                               -73-

-------
V.   Text of the Proposed Standards
§264.342  Designation of principal organic hazardous
constituents and hazardous combustion by-products.
(a)  Principal orqanic hazardous constituents (POHC's) and
hazardous combustion by-products must be treated to the extent
required by the performance standards specified in S264.343.
(b) (i)   For each waste feed to be burned, one or more POHC's
and hazardous combustion by-products will be specified from
among those constituents listed in Part 261, Appendix vui of
this Chapter.  This specification will be based on the deqree
difficulty of incineration of the orqanic constituents of the
waste feed and its combustion by-products, their concentration
or mass, considering the results of waste analvses and trial
burns or alternative data submitted with Part B of the facility's
permit application.  Orqanic constituents or by-oroducts which
represent the greatest deqree of difficulty of incineration will
be those most likely to be designated as POHCs or hazardous
combustion by-oroducts.  Constituents are more likelv to be
designated as POHCs or hazardous combustion bv-oroducts if thev
are present in large quantities or concentrations.
(ii) Trial POHC's will be desiqnated  for performance of trial
burns in accordance with the procedure specified  in Sl22.27(b)
for obtaining trial burn permits.  Trial hazardous combustion
by-products may be desiqnated under the same Procedures.
*****
                                -74-

-------
§264.343 is amended by redesiqnattnq paraqraoh (d)  as paraqranh
(h) and addinq new paraqraphs (d),  (e), (f), and (q)  as follows:
§264.343  Performance ^standards.
*****
(d)  Incinerators burning hazardous waste must destroy hazar-
dous combustion by-products desiqnated under §264.342 so that the
total mass emission rate of these by-products emitted from the
stack is no more than .01% of the total mass feed rate of POHC's
into the incinerator.
(e)  After consideration of the factors listed in oaraqraoh (q)
of this Section/ the Reqional Administrator mav, on a case-bv-case
basis, establish performance standards which are either more or
less strinqent than those required  by oaraqraphs (a)  and (d) of
this Section based on a findinq that:
     (1)  More strinqent standards  are necessarv because th«?
     emission rates achieved by the application of the Perfor-
     mance standards otherwise required by this Section mav
     pose an unacceptable risk to human health and the environ-
     ment, or
     (2)  Less strinqent standards  will achieve emission rates
     which do not nose an unacceptable risk to human health
     and the environment.
(f)  After consideration of the factors listed in paraqraph (q)
of this Section, the Reqional Administrator mav, on a case-bv-
case basis, stipulate performance standards for metals, hvdroqen
halides, and elemental haloqens,  based on a findinq that such
                               -75-

-------
standards are necessary to limit the emission rates of these
constituents to levels which do not pose an unacceptable risk
to human health and the environment.
(g)  The findings under paragraphs (e) and (f) of this Section
will be made after evaluating the following data, which the
Regional Administrator mav require from the permit applicant:
     (1)  Emissions of POHC's, hazardous combustion bv-products,
     metals, and hydrogen halides, includina:
          (i)  Mass emission rates from the stack, and
         (ii)  Concentration in the gas stream exiting the stack;
     (2)  Air dispersion estimates for these substances,
     including:
          (i)  Meteorological data,
         (ii)  Description of the air dispersion models,
        (iii)  Assumptions underlying the air dispersion models
               used;
     (3)  Expected human and environmental exposure, including:
          (i)  Topographic considerations,
         (ii)  Population distributions,
        (iii)  Population activities, and
         (iv)  Modes,  intensity and duration of exposure;
     (4)  Consequences of exposure, including:
          (1)  Dose-response curves for carcinogens,
         (ii)  Health  effects based on human or animal studies
               for other toxic constituents,
        (iii)  Potential for accumulation of toxic constituents
               in the  human bodv, and

                               -76-

-------
         (iv)   Statements of expected risk to individuals or
               populations.
(h)   For purposes of permit  enforcement, comoliance with the
operating requirements specified in the t>ermit (under 52fi4
will be regarded as compliance with this Section.  However,
evidence that compliance with those permit conditions is insuffi-
cient to ensure compliance with the performance requirements of
this Section may be "information" justifvinq modification; revo-
cation, or reissuance of a permit under 5122.15 of this Chapter.
*****
                               -77-

-------
                             References


 1.    Assessment of Hazardous Waste Practices in the Petroleum
      Refining Industry, NTIS No. PB-259-097.

 2.    Ferguson, T. L.; Bergman, F. J.; Cooper, G. R.; Li, R. T. ;
      and Homea, F. I.; Determination of  Incinerator Operating
      Conditions Necessary for Safe Disposal of Pesticides.  EPA
      600/2-75-041, NTIS No. PB 251-131/AS.

 3.    Polychlorinated Biphenyls, Criteria Modifications,
      40 CFR 761.40 Incineration, May 31, 1979.

 4.    Wisconsin "Administrative Rules for Air Pollution Control"
      Act 250 of 1965 as amended and Act  348 of 1965 as amended.

 5.    Lustenhouwer, J.W.A.; K Olie; and O. Hatrzinger; Chlori-
      nated Dibenzo-p-dioxins and Related Compounds in Incinerator
      Effluents, In Press Chemosphere.

 6.    U.S. Environmental Protection Agency, National Emission
      Standards for Identifying, Assessing and Relating Airborne
      Substances Posing Risk of Cancer, Federal Register
      Wednesday, October 10, 1979.

 7.    Interim Procedures and Guidelines for Health Risk and
      Economic Impact Assessments of Suspected Carcinogens;
      U.S. EPA, Federal Register, Vol. 41, p. 21402, May 25, 1978.

 8.    Scientific Basis for Identification of Potential Cacinogens
      and Estimation of Risk; Interagency Regulatory Liaison
      Group (IRLG), Federal Register, Vol. 44, No. 131, Friday,
      July 6, 1979.

 9.    Designation of 1822 Hazardous Substances, U.S. EPA,
      Federal Register, Vol. 45, No. 133, Wednesday July 9, 1980
      p. 46094.

10.    Water Quality Criteria Documents Summaries, U.S. EPA,
      Federal Register, Vol. 45, No. 231  November 28, 1980.

11.    TLVs Threshold Limit Values for Chemical Substances and
      Physical Agents in the Workroom Environment with Intended
      Changes for  1980, Adopted by American Conference of
      Governmental Industrial Hygienists.

12.    U.S. EPA, OAQPS Guidelines Series,  Guidelines on Air
      Quality Models, OAQPS 1.2-080 April 1978.
                                 -78-

-------
13.  Cincinnati Gas and Eletric Company, U.S. EPA, 578,
     p. 2d 660, 6th Circuit, 1978.

14.  Cleveland Electric Illuminating Company, U.S. EPA 572
     2d 1150, 6th Circuit 1978, cert, denied, 436 U.S. 911
     c 1978.

15.  Budney L. J., Guidelines for Air Quality Maintenance and
     Planning and Analysis, Volume 10 (Revised);  Procedures
     for Evaluating Air Quality Impact of New Stationary Sources,
     EPA 450/4-77-001, October 1977.

16.  User's Manual for a Single Source (CRSTER) Model, EPA
     #450/2-77-013, July 1977.

17.  Busse, A.D. and J.R. Zimmerman, User's Guide for the
     Climotological Dispersion Model, EPA # RA-73-024,
     December 1975.

18.  Regional Workshops and Air Quality Modeling:  A Summary
     Report (Draft).

19.  Identification of Potential Carcinogens and Estimates of
     Risk; Report of the Interagency Regulatory Liaison Group,
     Work Group on Risk Assessment, Journal of the National
     Cancer Institute, Volume 63, No. 1 July, 1979.

20.  Assessment of Industrial Hazardous Waste Practices; Organic
     Chemicals, Pesticides, and Explosive Industries NTIS
     No. PB 251-307, 1976.

21.  Assessment of Industrial Hazardous Waste Practices -
     Electronic Components  Manufacturing Industly.  NTIS
     No. PB-265-532, 1977.

22.  Assessment of Industrial Hazardous Waste Practices Textiles
     Industry.  NTIS No. PB-258-953.

23.  Off-Site Hazardous Waste Management Capacity, Booz - Allen
     and Hamilton Inc., Bethesda Md., August 4, 1980 (Draft
     Report).

24.  Duvall, D.S.; University of Dayton, Letter on Research
     Results Utilizing the  TDAS; to R.A. Carnes, U.S. EPA,
     October 17, 1979 (Draft).

25.  Multimedia Environmental Goals for Environmental Assessment;
     Vol.  I, II, III, and IV, U.S. EPA,  1978 and 1979, EPA-600/
     -7-77-136 A and B, EPA-600/-7-79-176 A and B respectively
     and Volume 1 Supplement, EPA 600/-7-80-041.
                               -79-

-------
these chemicals are potentially human carcinogens.  (Chemicals



regulated as carcinogens by the Occupational Safety and Health



Administration (OSHA) and the Consumer Product Safety Commission



(CPSC) are also on this list but are not noted as such since



they have been evaluated as being carcinogens by one of the



other organizations previously mentioned).  CAG evaluated the



studies upon which IARC, NTP, or FDA relied and agreed with all



the NTP and FDA evaluations that the chemicals presented a



potential human cancer risk.  The CAG agreed with most of lARC's



evaluations.  There are inconsistencies between the CAG and IARC



evaluations for a few chemicals because the CAG considered infor-



mation not available to or not otherwise used by IARC, and



because there are differences in the criteria used in making the



qualitative evaluations.



     The list is not a comprehensive listing of all chemicals



having substantial or strong evidence of carcinogenicity, chemi-



cals which do not now appear on the list will be added.  A



continuing review of evaluations by organizations such IARC, NTP,



FDA, OSHA, and CPSC may result in periodic revisions to the



present list.



     The CAG evaluates substances for possible carcinogenicity



according to the procedures outlined in the Agency's Interim



Guidelines for Carcinogen Risk Assessment found in Interim Proce-



dures and Guidelines for Health Risk and Economic Impact Assess-



ments of Suspected Carcinogens (41 Fed. Reg. 21402, May 25, 1976),



These guidelines are consistent with the Interagency Regulatory

-------
Liaison Group's Scientific Bases for Identification of Potential
Carcinogens and Estimation of Risks (Jouirnal of the National
Cancer Institute (>3_ (1): 243-268 1979,  44 Fed. Reg. 39858, July 5,
1979), and the Regulatory Council Statement on Regulation of
Chemical Carcinogens (44 Fed. Reg. 760037, October 17, 1979).
     Evidence concerning the carcinogenicity of chemical
substances is of three types:  (1) epidemiologic evidence derived
from long-term bioassays on animals; and (3) supportive or
suggestive evidence derived from studies of chemical-structure
or from short-term mutagenicity, cell  transformation or other
tests that are believed to correlate with carcinogenic activity.
     The CAG evaluates all available evidence on the carcinogeni-
city of a chemical before reaching a conclusion based on the
"weight of the evidence," about the chemical's human carcinogenic
potential.  Conclusions about the overall weight of evidence
involve a consideration of the quality and adequacy of the data
and the kinds of responses induced by the suspect carcinogen.
The best evidence that an agent is a human carcinogen comes from
epidemiologic studies in conjunction with confirmatory animal
tests.  Substantial evidence is provided by animal tests that
demonstrate the induction of malignant tumors in one or more
species or of benign tumors that are generally recognized as
early stages of malignancies.  Suggestive evidence includes indi-
rect tests of tumorigenic activity, such as mutagenicity, in
vitro cell transformation, and initation-promotion skin tests in
mice.  Ancillary data that bear on judgments about carcinogenic

-------
potential, e.g., evidence  from systematic studies that relate
chemical structure to carcinogenicity, are alos considered.
     Substances were placed on the CAG list only if they had
been demonstrated to induce malignant tumors in one or more ani-
mal species or to induce benign tumors that are generally recog-
nized as early stages of malignancies, and/or if positive epide-
miologic studies indicated they were carcinogenic.  Although the
CAG has determined that there is substantial evidence of carcino-
genicity for each chemical substance on  the list, the data varies
to some extent with respect to the scope and quality of the
studies.
     No uncommonly, CAG reports are updated because new evidence
becomes available.  Because of this, it  is important that the
most recent CAG evaluation be consulted.
     Some of the reports prepared by CAG are subject ot confiden-
tiality claims.  Because of these claims (primarily under the
Federal Insecticide, Fungicide, and Rodenticide Act) some reports
may not be released.   Therefore, all requests for CAG reports and
related documentation  must be submitted  through EPA's Freedom of
Information Office  (A-101), Washington,  D.C.  20460, and should
be marked CAG/LOC.

-------
Benzo(b)fluoranthene (IARC)

Benzo(j)fluoranthene (IARC)@

Beryllium and Beryllium Compounds (CAG/ IARC)

N,N-Bis(2-Chloroethyl)-2-Napthylamine (Chlornaphazine) (IARC)**

Cadmium and Cadmium Compounds (CAG/ IARC)

Carbon Tetrachloride (CAG, IARC)

Chlorambucil (IARC)**

Chloroalkyl Ethers
  Bis(2-chloroethyl)ether  (BCEE) (CAG) (IARC)@
  Bis(chloromethyl)ether (BCME) (CAG/ IARC)
  Chloromethyl methyl ether (CMME), technical grade (IARC)

Chlordane (CAG, NCI)

Chlorinated Ethanes
  1,2-Dichloroethane [Ethylene Chloride, Ethylene Dichloride (EDC)]
      (CAG, IARC, NCI)
  Hexachloroethane  (CAG)
  1,1,2,2-Tetrachloroethane (CAG)
  1,1,2-Trichloroethane (CAG, NCI, IARC)@

Chlorobenzilate (CAG)

Chloroform (CAG, IARC)

Chromium Compounds, Hexavalent (CAG,  IARC)

Chrysene (IARC)@

Citrus Red No. 2 (IARC)

Coal Tar and Soot (CAG, included in lARC's soots, tars, and oils
 designation)

Coke Oven Emissions  [Polycyclic Organic Matter  (POM)]  (CAG)

Creosote (CAG)

Cycasin (IARC)
     ** Used as a drug.

     '? Evaluated by IARC as not having sufficient evidence of
carcinogenicity.

-------
              CHEMICALS HAVING SUBSTANTIAL EVIDENCE
                       OF CARCINOGENICITY3
2-Acetylaminoflourene (See references)

Acrylonitrile (CAG, IARC)

Aflatoxins (IARC)*

Aldrin (CAG, NCI)

4-Aminobiphenyl  (IARC)

Amitrole (IARC)

Aramite (IARC)

Arsenic and Arsenic Compounds (CAG, IARC)

Asbestos (CAG, IARC)

Auramine and the manufacture of Auramine (IARC)

Azaserine (IARC)**

Benz(c)acridine  (IARC)@

Benz(a)anthracene  (IARC)

Benzene (CAG, IARC)

Benzidine (CAG,  IARC)

Benzo(a)pyrene (IARC)
     a This is not a comprehensive list of all chemicals having
substantial evidence of  carcinogenicity.  Other chemicals will be
added.  No attempt has been made to select chemicals based upon
appropriateness  for regulation by EPA.  The list is intended to
be a basis for selection by the various program offices according
to their specific needs.

     * Fungal toxin, not an industrially manufactured product.

     ** Used as  a drug.

     @ Evaluated by IARC as not having  sufficient evidence of
carcinogenicity.

-------
           CHEMICALS HAVING EVIDENCE OF CARCINOGENICITY





     In response to requests from several EPA offices, the



Carcinogen Assessment Group (CAG), Office of Health and Environ-



mental Assessment in EPA's Research and Development Office has



prepared a list of chemical substances for which substantial or



strong evidence exists showing that exposure to these chemicals/



under certain conditions, causes cancer in human,s or can cause



cancer in animal species which in turn, makes them potentially



carcinogenic in humans.



     The list was initially prepared in response to the needs of



the OFfice of Pesticides and Toxic Substances (OPTS) to develop



labeling regulations under section 6 of TSCA and the Office of



Solid Waste (OSW) to develop hazardous waste regulations under



section 3001 of RCRA.  It is anticipated that it will serve



other purposes within th Agency according to the needs of the



program offices.



     The sources of information used in selecting agents as



candidates for the list are of two types:  chemicals which the



Carcinogen Assessment Group previously has evaluated and has



determined pose a potential human cancer risk; and chemicals,



the carcinogenicity of which the CAG reviewed because one or more



of three organizations — the International Agency for Research



on Cancer (IARC), the National Cancer Institute Bioassay Program



which has been reorganized into the National Toxicology Program



(NTP), and the Food and Drug Administration (FDA) of the U.S.



Department of Health and Human Services — had concluded that

-------
Cyclophosphamide (IARC)**

Daunomycin (IARC)**

DDT (Dichlorodiphenyltrichloroethane) (CAG)

Diallate (CAG) (IARC)@

Dibenz(a,h)acridine (IARC)

Dibenz(a,j)acridine (IARC)

Dibenz(a,h)anthracene (IARC)

Dibenzo(a,e)pyrene (IARC)

Dibenzo(a,h)pyrene (IARC)

Dibenzo(a, j)pyrene (IARC)

1,2-Dibromo-3-chloropropane (DBCP) (CAG, IARC, NCI)

1,2-Dibromoethane  [Ethylene Bromide, Ethylene Dibromide (EDB)]
  (NCI, CAG, IARC)

3,3'-Dichlorobenzidine (DCB) (CAG, IARC)

Dieldrin (CAG)

Diepoxybutane (IARC)

1,2-Diethylhydrazine (IARC)

Diethylstilbestrol (DBS)  (IARC)**

Dihydrosafrole (IARC)

3,3'-Dimethoxybenzidine  (o-Dianisidine) (IARC)

p-Dimethylarainoazobenzene  (IARC)

7,12-Dimethylbenz(a)anthracene  (See references)

3,31-Dimethylbenzidine (o-Tolidine) (IARC)

Dimethylcarbamoyl Chloride  (IARC)
     ** Used as a drug.

     @ Evaluated by IARC as not having sufficient evidence of
carcinogenicity.

-------
1,1-Dimethylhydrzine (IARC)

1,2-Dimethylhydrazine (IARC)

Dimethyl Sulfate (IARC)

2,4-Dinitrotoluene (CAG, NCI)

1,4-Dioxane (NCI)

1,2-Diphenylhydrazine (CAG)

Epichlorohydrin  (CAG)

Ethylene Bis Ditiocarbaraate (EBDC) (CAG)

Ethyleneimine (Aziridine) (IARC)@

Ethylene Oxide (CAG, IARC)

Ethylenethiourea (CAG, IARC)

Ethyl Methanesulfonate (IARC)

Formaldehyde (CAG)

Glycidaldehyde (IARC)

Heptachlor (CAG, NCI)

Hexachlorobenzene (CAG, IARC)

Hexachlorobutadiene (CAG)

Hexachlorocyclohexane (HCH)
     HCH (CAG)
     HCH (CAG)
     HCH (Lindane) (CAG)
    Technical HCH (CAG)

Hydrazine (IARC)

Indeno(l,2,3-cd)pyrene (IARC)

Iron Dextran (IARC)**@
     ** Used as a drug.

     @ Evaluated by IARC as not having sufficient evidence of
carcinogenici ty.

-------
Isosafrole (IARC)

Kepone (Chlordecone) (CAG, NCI)

Lasiocarpine (IARC, NCI)

Melphalan (IARC)**

Methapyrilene (FDA)**

3-Methylchoanthrene (See references)

4,4'-Methylenebis(2-Chloroaniline) (MOCA) (IARC)

Methyl Iodide (CAG, IARC)

Methyl Methanesulfonate  (IARC)

N-Methyl-N1-nitro-N-nitrosoguanidine (IARC)

Methylthiouracil  (IARC)**

Mustard Gas (IARC)

1-Naphthylamine,  technical grade  (CAG)

2-Naphthylamine  (IARC)

Nickel and Nickel Compounds (CAG, IARC)

Nitrogen Mustard  and its hydrochloride (IARC)

Nitrogen Mustard  N-oxide and its hydrochloride (IARC)

5-Nitro-o-toluidine (NCI)

4-Nitroquinoline-l-oxide (See references)

Nitrosamines
  N-Nitrosodiethanolamine (IARC)
  N-Nitrosodiethylamine  (DENA)  (CAG, IARC)
  N-Nitrosodimethylamine (DMNA) (CAG, IARC)
  N-Nitrosodi-n-butylamine (IARC)
  N-Nitrosomethylethylamine (IARC)
  N-Nitrosodi-n-propylamine (IARC)
  N-Nitrosomethylethylamine (IARC)
  N-Nitrosomethylvinylamine (IARC)
     ** Used as a drug.
     @ Evaluated by  IARC as not having sufficient evidence of
carcinogenicity.

-------
  N-Nitroso-N-Ethylurea (NEU) (CAG, IARC)
  N-Nitroso-N-Methylurea (NMU) (CAG, IARC)
  N-Nitroso-N-methylurethane (IARC)
  N-Nitrosomorpholine (IARC)
  N-Nitrosonornicotine (IARC)
  N-Nitrosopiperidine (IARC)
  N-Nitrosopyrrolidine (IARC)
  N-Nitrososarcosine (IARC)

Pentachloronitrobenzene (PCNB) (CAG)

Phenacetin (IARC)**

Polychlorinated Biphynyls (PCBs)  (CAG, IARC)

Pronamide (CAG)

1,3-Propane Sultone (IARC)

3-Propiolactone (IARC)

Propylthiouracil (IARC)**

Reserpine (NCI)**

Saccharin (FDA)***

Safrole (CAG, IARC)***

Selenium Sulfide (NCI)

Streptozotocin (IARC)**

2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)  (CAG)

Tetrachloroethylene (Perchloroethylene)  (CAG, NCI)

Thioacetamide (IARC)

Thiourea (IARC)

o-Toluidine Hydrochloride (NCI)

Toxaphene (CAG, IARC, NCI)

Trichloroethylene (CAG, NCI)
     ** Used as a drug.

     *** Used as a food.

-------
2,4,5-Trichlorophenol (NCI)



Tris(l-aziridinyl)phosphine sulfide (thio-TEPA)  (IARC,  NCI)**



Tris(2,3-dibromopropyl)phosphate (IARC, NCI)



Trypan Blue, commercial grade (IARC)



Uracil Mustard (IARC)**



Urethane (IARC) (Ethyl carbamate; ethyl ester of carbamic  acid)



Vinyl Chloride (CAG, IARC)



Vinylidene Chloride (CAG)
     ** Used as a drug.

-------