United States        Office of Air Quality         EPA «50/5-82-ooa
Environmental Protection   Planning and Standards       OCTOBER 1992
Agency           Research Triangle Park NC 27711

Air



HAZARDOUS AIR  POLLUTANT



PRIORITIZATION  SYSTEM



           HAPPS

-------
              ARGONNE  NATIONAL LABORATORY
                9700  South  Cass Avenue,
               Argonne,  Illinois  60439
                HAZARDOUS AIR POLLUTANT
                 PRIORITIZATION SYSTEM
                        (HAPPS)
                          by

             A.E.  Smith and D.J.  Fingleton
       Energy and  Environmental Systems Division
                     October 1982
                     prepared for ~
              Pollutant Assessment  Branch
         Standards and Air Strategies Division
     Office of Air Quality Planning and Standards
         U.S. Environmental Protection Agency
     Research Triangle Park, North  Carolina  27711

    Under Interagency Agreement No. AD-89-F-1-344-0
            Project Officer:  Robert Schell
U.S. Envjrr.
Chicago, l_i

-------
                              DISCLAIMER
     This report has been reviewed by the Office of Air Quality Planning
and Standards, U.S. Environmental Protection Agency, and approved for
publication as received from the Argonne National Laboratory.  Approval
does not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection Agency, nor does mention
of trade names or commercial products constitute endorsement or
recommendation for use.

-------
                                   CONTENTS

1  PURPOSE AND RATIONALE	1
   1.1  SCOPE AND LIMITATIONS 	   1
   1.2  RATIONALE FOR HAPPS FACTORS 	   5
   1.3  RATIONALE FOR CRITERIA	10
   1.4  FACTOR GROUPS AND WEIGHTS	42
        1.4.1  General	42
        1.4.2  Groups	45
   1.5  Incergroup Weights	49
2  PRIORITIZATION METHODOLOGY 	  54
REFERENCES	55
APPENDIX A:  Tables, Worksheets,  and Abbreviations Used in RTECS	  58
                                   TABLES
1.1   Factors in HAPPS and the ORNL Procedure	6
1.2   Criteria for Oncogenicity	12
1.3   Criteria for Mutagenicity	18
1.4   Criteria for Reproduction and Developmental Toxicity	21
1.5   Criteria for Acute Lethality	24
1.6   Criteria for Effects Other than Acute Lethality 	  31
1.7   Criteria for Production Volume	35
1.8   Criteria for Vapor Pressure 	  37
1.9   Criteria for Bioaccumulation	39
1.10  Criteria for Existing Standards 	  40
1.11  Utility of Normalization	43
1.12  Groups of Factors	46
1.13  Intergroup Weights	50
1.14  Sensitivity Analysis	52

                                   FIGURE

1  Scales for Equivalent Volumetric and Mass Concentration Units	26
                                     111

-------
                          1   PURPOSE  AND RATIONALE

       This  section  discusses  the  purpose  of  the Hazardous  Air Pollutant
Prioritizacion  System  (HAPPS) and  the rationale used  in developing the
system.   Subsections 1.2,  1.3,  1.4,  and  1.5  provide,  respectively, the
rationale for choosing the particular set of eight  factors,  the rationale for
the specific criteria and  weights  within  each factor,  the rationale for
grouping the  factors  and  assigning the  intragroup  weights, and the rationale
for the intergroup weights used in  the final  ranking.   These  rationales can be
properly understood only  when  the  scope of HAPPS and  the  limitations imposed
by this scope are taken  into account.   Subsection 1.1  discusses these limita-
tions .
       Section 2 provides instructions for using  HAPPS.   Appendix A contains a
set of worksheets and tables  for documenting  a  prioritization.  Both Section 2
and the Appendix have  been placed  after  the  explanatory  material in Sec. 1 to
aid in copying  the  working  material  in the appendix.  Section 1 assumes some
familiarity  with HAPPS  and  readers   may want  to  scan  the  instructions and
materials in Sec. 2 and Appendix A  prior to reading  Sec.  1,

1.1  SCOPE AND LIMITATIONS
       The  strategies  and Air  Standards Division (SASD)  of U.S. SPA's Office
of Air Quality Planning  and  Standards periodically  selects new substances for
assessment  to determine whether regulatory development  under  the Clean Air Act
should  begin.    Ideally,  a  full  range  of  toxicological and  epideiniological
information  coupled  with detailed estimates of current emissions  and  human
exposure would be available  to  aid in such decisions.  Hovever, such complete
information  is   seldom  available  and early  assessment  is often made  on the
basis  of  incomplete  and/or  dated  information.   This  is particularly true of
EPA's  hazardous  air  pollutant  assessment  effort,   since  a   large  number of
organic and inorganic substances are  potential  candidates for study.  However,
even  if  only a  small number  of  potential candidates  existed, the  resources
involved  in producing complete  scientific  information  preclude the develop-
ment  of such  information for  each  substance  until there  is some  certainty
that  regulation  is  appropriate.   Thus,  a  procedure for  initially  prioriti-
zing  substances on  the basis  of limited,  readily  available  information

-------
is  needed  so  that  resources  for  detailed studies  might be  allocated  effi-
ciently, that is allocated first to detailed studies of substances anticipated
to be significant  air  pollution problems and  later  to  substances anticipated
to be lesser problems.   HAPPS provides a means of producing such a prioritiza-
tion.
       It is recognized that  a  prioritization  with  limited data as opposed to
extensive,  detailed data might produce substantially different results.  HAPPS
is only intended to provide a reasonable prioritization to aid EPA in deciding
which substances  to study first based  on readily  available  information.   As
such HAPPS is  intended  to  be  used  as part of EPA's internal planning process.
It  is  important to recognize  that even substances ranked very high by HAPPS
might never be  regulated.   Many subjective decisions must be made and detailed
objective studies  done  and evaluated between  the time  a substance  is ranked
highly  by  HAPPS and a  decision  is made to regulate that  substance  as an air
pollutant.   In  other words, HAPPS must be viewed a's an initial, tentative step
within  the context of the  overall  regulatory  program.   Even the  initial
prioritizacion  produced by HAPPS  will  be subject  to  additional  screening by
experts to eliminate any obvious anomalies.  In addition, the methodology will
be  applied  periodically to  incorporate  new information  as  it  becomes avail-
abl;..   Such  periodic reviews  could result in changes in  the relative rankings
of  various  substances,  reflecting  the  new information.   There may  also be
progrsmaiatic  reasons   for  overriding the  indications -given  by  HAPPS.   For
example,  it  might  occur  that some  particular class of  compounds  like heavy
metals  is  receiving special  attention  throughout  air  programs,  a considera-
tion  which could  lead  to  alterations to a. list produced  by HAPPS.  Regulatory
decisions will not be  made on the basis  of a  substance's ranking by the HAPPS
procedure.   Health assessments,  exposure  assessments,  and  other information
must  all be  evaluated   prior  to  making  the  decision to regulate.  It might be
found  that  the health  effects  associated  with a  substance  ranked  highly by
HAPPS were  not serious enough,  that  control  of the substance  was technically
infeasible,  or that the likelihood  of  exposure was not  sufficiently great to
justify regulation.
        To be  useful,  a procedure must be tailored  to the expertise  and exper-
ience of potential  users.  No macter how  complete or precise  a  procedure  is in
theory,  it  is useless  if its  application requires detailed  or specialised

-------
knowledge not available Co Che user group.  With Chese points in mind,  discus-

sions with EPA personnel defined  several  guidelines  which HAPPS would  need to
follow Co be useful in the prioritization effort:

       •  Generally, readily available summary documents or compu-
          terized data bases will be used; searches of primary
          sources are precluded by the intended use of the proce-
          dure and the inability to justify allocating significant
          resources to this preliminary prioritization step.

       •  The methodology is purposely designed not to utilize
          expert judgment in prioritizing chemicals.   Such expert
          judgment is more appropriate as part of subsequent regu-
          latory decision making than to the preliminary prioriti-
          zation stage where HAPPS is used.

       •  The preliminary nature of the prioritization means that
          the procedure should be simple enough to permit a single
          user to prioritize several substances per day and prefer-
          ably ten or more.

       •  Personnel using the procedure should have only limited
          expertise in toxicology or related subjects and only
          limited familiarity with some of the sources of emissions
          of the substances being ranked.  Hence,  the procedure
          could not rely on decisions requiring expert judgment or
          special knowledge related to these areas.
       •  A particular set of substances should receive the same
          ranked order when prioritized by two different persons.
          Thus, insofar as possible, the procedure should be objec-
          tive and the sources of data should be identical  for  all
          users.  Complete agreement between two different users of
          HAPPS is unlikely because the goal of complete objectiv-
          ity could not be attained; some factors still require the
          use of informed, as distinct from expert, judgment in
          choosing between criteria.

       •  The system should be sufficiently flexible to permit  up-
          dates for a substance  for which additional data becomes
          available in the standard references.
       •  The procedure is only  intended  to produce reasonable,
          initial rankings.  Detailed studies will take place after
          the prioritization to  develop sufficient information  to
          determine whether or not regulation is required.  Ques-
          tions of data interpretation, the validity of data and
          similar technical items are left for experts to decide at
          later stages in  the assessment  process.

These guidelines place considerable constraints upon the  procedure by  restric-
ting  the  source  of data  to be used and by limiting  the effort  in prioritizing
a  single compound.   In  particular,  searching the  literature  and contacting
other  workers involved  in prioritizing  hazardous  compounds led  to  an  early
recognition  that the most suitable  summary  reference  appeared  to be the

-------
Registry of Toxic  Effects  of Chemical  Substances  (RTECS).*   RTECS is  a
concise, easily used summary of toxic effects and is kept current  by  continual
updates  in a computerized  format and  by  quarterly  updates  in  microfiche copy.
RTECS also contains data related to most  of the criteria used  for  ranking sub-
stances  in the  scoring  procedure  developed by Oak  Ridge National Laboratory.
As discussed below, the Oak  Ridge  procedure  provided  the basis  for HAPPS thus
making RTECS an excellent  match to the data requirements of HAPPS.  That RTECS
was  the most  easily  accessible  reference  had  a  significant  impact  on  the
choice  of  factors  and the  structure  of  criteria for individual factors.   In
addition,  the  need  for consistency  between  users  and  the  expected  expertise
of users limit the types  of decisions users  should  be expected  to make  and
make  a  straightforward  approach with  little chance  for  individual  deviation
desirable.
       Many  ranking or scoring  procedures  for  prioritizing  chemicals  exist
(see, for  example,  Refs.  2-25).   Rather  than  develop an entirely new  system,
consultation with EPA indicated  the desirability of  using  a  draft EPA multi-
media ranking  procedure  as  the basis  for HAPPS.   That  scoring procedure  had
been  developed  by  Oak Ridge National Laboratory (ORNL)21  for EPA's  Office of
Pesticides  and  Toxic  Substances  (OPTS)  and itself drew heavily on  existing
scoring  systems.  Using an existing procedure  as a basis for  HAPPS  was consi-
dered to be  efficient in   that a significant amount of  duplicative work could
be  avoided.   Much  thought and developmental  work  had  already gone into  the
ORNL procedure and  its   predecessors  in  choosing  factors,  criteria,  and
weights.   After an  initial review,  it was clear that much of what had  already
been  done  was  relevant  to air programs  and  might  need  only minor revision or
expansion.   It  should be noted that the ORNL procedure was still in  draft form
at the  time  HAPPS  was developed.   At the date  of this report, the ORNL proce-
dure  has not been  completed  and  applied  because objectives other  than priori-
tization have  become  priority items in OPTS programs.   Thus,  the ORNL proce-
dure  was used  as  the principal basis  for  HAPPS  even  though still  in draft
form.    In  choosing  the  specific   factors  and criteria,  additional  existing
scoring systems,  frequently those  used in  developing  the  ORNL procedure
itself,  were consulted and used in  developing HAPPS.
        However,  the ORNL  procedure differed in  scope and  purpose  from HAPPS.
The  ORNL  procedure  was   intended  to consider  multimedia exposures  through
various routes  (air,  water,  consumer usage,  etc.) and was  also  intended to

-------
present  data  from  the  literature and  data submitted  in  compliance with  the
8(a)-Level-A rule  of  the Toxic Substances  Control  Act  (TSCA) in a  form  suit-
able for review  and final  ranking by experts,  both intentions making the ORNL
procedure inappropriate  under  the HAPPS guidelines.  In developing HAPPS,  the
ORNL procedure was  used  to select some  of the factors, criteria, and weights
and then tailored  to  the specific concerns of air  programs.  Since  the output
of  HAPPS will  provide  program  planning  information  for  the  Office  of  Air
Programs, it was not necessary  to consider  total human  exposure; consideration
of  exposure  through the air  route only is  sufficient  for such planning pur-
poses.

1.2  RATIONALE FOR  HAPPS FACTORS
       HAPPS prioritizes substances by  scoring them in  eight  factors chosen to
reflect  the concerns  of  air  programs  and issues  deemed important  by EPA.
Table 1.1 presents  the eight  factors used  in HAPPS  and  the  twenty-five factors
used in  the ORNL procedure.   The broader range of  impacts  on humans, animals,
plants,  and  the  environment  in the ORNL procedure  shows clearly when the  two
sets of  factors  are compared.
       Among all the  aspects  of human health,  the  Office  of  Air Programs felt
that  carcinogenesis should receive  special attention  in  accordance with  the
public's concern with carcinogens.  The ORNL  procedure already contained  two
factors, oncogenicity  and  mutagenicity, (items 1 and 2 in Table 1.1) related
to  carcinogenesis  and these  were  retained  in  HAPPS.    The  oncogenicity  factor
contains both  malignant and  benign  tumors.   Mutagenicity is related to car-
cinogenicity  and evidence  of  mutagenic  potential  is  frequently used  as   an
indicator  in  screening  for carcinogens as in the  Ames test.  Although  not  a
current  major  concern for air  programs,  retention  of  the  separate  factor  for
mutagenicity was reasonable   for  two  reasons  even  though  it  is  later grouped
together with  oncogenicity under the carcinogenicity group.  First, oncogeni-
city and mutagenicity are  distinct effects; all mutagens  are not carcinogens;
however, most  carcinogens  are mutagens.  In addition,  there  is  only a  limited
amount  of  data  available  from  carcinogenicity testing, making the  use  of  the
more extensive  surrogate data from mutagenicity testing desirable in order to
maximize the  number of  substances that  could  be scored using  data  reasonably
related  to  carcinogenic  potential.  Second, data are available  separately  for
each  factor in  RTECS,   the principal  data  source, tasking scoring  a compound

-------
                  Table 1.1  Factors in HAPPS and the ORNL Procedure
                                       Procedure
Item
                               ORNL3
                                                                HAPPS
       Component
          Factor
          Factor
  1
  2
  3

  4
  5

  6
  7
  8
  9
 10
 11

 12
 13
 14
 15

 16
17-21
 22

 23
 24
 25
 26
Chronic Toxicity
Acute Toxicity
Environmental Exposure
Occupational Exposure
Consumer Exposure
Oncogenicity
Mutagenicity.
Embryo-Fetotoxicity

Reproductive Effects
Terrestrial Animals

Aquatic Animals
Plants, Fungi, Bacteria
Terrestrial Animals
Aquatic Animals
Plants, Fungi, Bacteria
Production Volume

Environmental Release"
Transport & Transformation
Bioconcentration
Weighted Quantity
  Processed13
Weighted Quantity
  in Products^
(Five Separate Factors)0
Weighted Quantity
  in Products'5
Number Exposed
Frequency of Exposure
Intensity of Exposure
Oncogenicity
Mutagenicity

Reproductive and Develop-
  mental Toxicity
Effects Other than Acute
  Lethality
Acute Lethality
                                                      Potential for Airborne
                                                        Release
                                                      Bioaccumulation
                                                      Existing Standards
aAdapted from Ref. 21.
"Requires information available from manufacturer compliance with 8(a)-Level-A rule of
 TSCA.

-------
unambiguous if two  factors are  used while  confusion might result  with a
composite factor based  on both  oncogenicity  and mutagenicity data.
       Oncongenicity and mutagenicity were separated from other toxic effects
because the Office of Air Programs  is  primarily concerned with carcinogenicicy
rather  than the  toxic  effects  considered  under the  other  toxicity-related
factors:   reproductive and  developmental  effects  (items  3  and  4),  effects
other than acute lethality (item 5),  and  acute  lethality (item 8).  It is also
generally recognized  chat oncogens and some  mutagens  have no thresholds
whereas  the effects dealt  with in  the  factors  for  items  5 and  8  normally
exhibit thresholds  giving the  separation  a  reasonable  conceptual  basis.
Assignment of appropriate weights to the carcinogenicity and  toxicity groups,
as discussed later, was used to combine  all toxic effects including oncogeni-
city  and  mutagenicity  in the  final  prioritization.    Finally, RTECS  provides
separate  data  for  oncogenicity,  mutagenicity,  reproductive and developmental
effects,  and the  two toxic  effects  factors  so that  the factors used  in HAPPS
match the available data,.thereby reducing the  likelihood of error.
       The  ORNL factors  for  items  3  and  4 were combined into a single factor
for reproductive and developmental toxicity in HAPPS.  Increasing concern for
developmental effects has been shown  in  recent  years as evidence accumulates
revealing the high sensitivity of human embryos,  fetuses, and  young to certain
substances.  Reproductive effects could  have long-term impacts on the popula-
tion  and  might  be considered  severe.   However,  data  in RTECS  does  not dis-
tinguish  between the two  types of effects and hence  the separate ORNL factors
were  combined.   As discussed above,  these effects were separated  from tine two
factors most directly related  to carcinogenicity which is currently the
principal  focus of  air programs.   Although  it has  been theorized that repro-
ductive  and  developmental effects exhibit thresholds they were also separated
from  the two factors  for toxic effects  because  this  separation matches the
form  of the data as presented in RTECS.
        Since HAPPS  emphasizes  human  health,  the  ORNL  factors  for aquatic
animals and  for plants, fungi,  and bacteria  (items 6,7,9,  and  10  in Table 1.1)
were  eliminated  as being poor  indicators of  human  health effects.   In  addi-
tion, no readily available  source of  information was  found  for  toxicity
effects  in  plants,  fungi, and  bacteria making  the related  factors  inappropri-
ate  for  HAPPS  even if  they  were relevant  to the  major  concern with  human

-------
health.   Factors for  toxic effects  in  terrestrial  animals  including humans
(items 5  and  8)  but distinct  from  oncogenic,  mutaganic,  and reproductive and
developmental effects were  retained.   Effects in nonhuman terrestrial species
were  included as  being  reasonable  indicators  of  potential  effects in humans.
Such  inclusion,  at  least for  higher  mammals, is  in accordance with  standard
toxicological procedures  that  use animal studies to anticipate health effects
or  toxicity in humans.   Such  procedures  frequently  involve  tests on higher
mammals  and results  of  such  toxicity  studies  are  used  in scoring  the two
toxicity  factors in HAPPS.
       The  factor for production  volume  (item 11) was modified for HAPPS.  In
assessing exposure  via  the  ambient  air,  it is desirable to know how much of  a
substance becomes  airborne.   Such  detailed  information is  not available for
most  substances at  the  prioritization phase  of study so production volume was
considered  as a  surrogate   following  the  ORNL  procedure.    Because  airborne
release  was  the  particular  interest in  HAPPS,   the  factor was  modified to
include  the consideration  of  vapor pressure  in  estimating  the potential for
airborne  release,  since,  other conditions being  the same,  a substance with  a
high  vapor  pressure will  become airborne more readily than a substance with  a
low vapor pressure.  (Vapor pressure  was also  considered in  the ORNL procedure
in one of the factors related  to  occupational  exposure.)
       Of   the  five ORNL  factors  related  to environmental  exposure  (items
12-16),  four  were eliminated  and one  was  retained.  The factors  for  environ-
mental  release   (item  12)  and for the  weighted  quantities  processed and in
products  (items  15  and  16)  require data that  is  to  be submitted  in compliance
with  the  8(a)-Level-A  rule  of TSCA and may  be of use in the  future.   However,
these  factors are aimed at  assessing overall  environmental exposure and  human
exposures  through  routes  in  addition  to ambient air giving  them a  broader
scope  than  appropriate  for air  programs.   Thus,  these  factors were  dropped
from  HAPPS.   The factor  for transport and transformation (item 13)  was  consi-
dered  to  be important  in view of possible  chemical  transformations and  resi-
dence  time  in the  atmosphere.   However,  no summary sources of  relevant  data
were  found  and  no  sources  seem likely to become  available  in  the  near  future
thus  eliminating  this factor from consideration  under the guidelines.
       The  factor  for  bioaccumulation  (item  14)  was  retained  in HAPPS.  Air
pollutants  can  either be deposited mechanically or absorbed directly  by  food

-------
or  be  deposited on  soil  and absorbed  later  by  plants.   Such  deposited sub-
stances can then either  pass up  the  food chain through animals to humans with
possible  biological  magnification or  be ingested  directly.   In  both cases,
toxic  levels can result  directly  or  build up in the body over time.  Data for
this factor  was available  in summary  form and consequently was  retained in
HAPPS.
       Overall,  then,  three factors  were  used  as surrogates  for exposure:
production volume  and  vapor  pressure  combined  to  estimate  the  potential for
airborne  release and bioaccumulation.   Information for these three factors is
readily available  but  is not readily  available  for other indicators of ulti-
mate fate  and  exposure like  residence  time  in  the atmosphere and atmospheric
reactions.   In  addition,  these  latter  two indicators  were  considered  too
detailed  for the preliminary nature of  this prioritization.
       The five ORNL factors  relating  to  occupational exposure were deleted in
consultation with  SASD, because occupational exposure does not fall within the
ambit  of  the Clean Air Act.   In addition, only one of  the five factors, level
of  potential occupational exposure,  does not depend  upon data to be gathered
under  the 8(a)-Level-A rule  of TSCA.   This ORNL actor scores compounds based
on  exposure concentrations experienced  by workers.  While  some of  the criteria
used for  the  factor relate  to  the  ease  with which workers  could contact the
airborne  chemical, the information required  would not generally be available
and  the  factor was dropped  from HAPPS.   However,  vapor pressure could be used
for scoring  liquids  under  this  criterion in  the  ORNL procedure  and was
retained  as one   of  the  indicators  of  the  potential for  airborne  release.
       The  four ORNL factors for consumer exposure (items 22-25) were dropped
from EAFPS.  These factors  are  oriented  toward exposures  due  to use of  house-
hold and  consumer  products and  would include many  routes  of  exposure  inappro-
priate to air  programs.   Although indoor air pollution problems due to  use of
household  and  consumer products  would logically  come  under  the criteria, no
summary  source of the data required  by the  criteria for  these  factors was
found.    In  addition,  indoor exposure  is  not  the focus  of  the air  programs
office.   Thus, these  criteria were  dropped and the factors  for potential for
airborne  release and bioaccumulation  are the  only factors related  specifically
to  the potential for human exposure  via the ambient air.

-------
                                      10
       A  factor for  existing  standards  (item  26),  not  found in  the ORNL
procedure, was added  to HAPPS.   The factor was based on the MITRE procedure^1-1
and is intended to be scored based on standards set by the Occupational  Safety
and Health Administration  (OSHA).   Establishment  of an OSHA standard requires
a  finding  of  potential  toxic effect.  Such  a  finding  was  considered to be an
important  indicator  that  a  substance  might  need  to  be  considered  in more
detail even though  the  OSHA standards are  intended to apply in the workplace
where  concentrations  are  likely  to  exceed ambient levels.   It  was  also felt
that existing standards would be useful for prioritization, especially  if data
for  the  other factors  was sparse  or if  two  or  more  substances  were  scored
relatively close.

1.3  RATIONALE FOR CRITERIA
       This section  discusses the  reasons  for choosing the  criteria and the
associated weights within  the individual factors.   As already noted,  the draft
ORNL procedure^! provided  the principal model  for HAPPS.  However, in develop-
ing the specific criteria, several other procedures were frequently consulted.
References 2,  9,  17,  18,   20, 22,  and 23  were  found to be particularly  useful
and  include an  earlier  version  of the ORNL procedure itself (Ref. 20)  as well
as  several systems  used  in the  development of that procedure.   These referen-
ces were  selected after review of Refs. 2-24 as being most nearly suitable for
HAPPS.   Thus, the  criteria  finally  used were chosen  from among several sets
available  in  the   literature  with  possible  additions  and  modifications   to
conform  to the guidelines  for  HAPPS and to make  them useable with  the RTECS
data  base.   With  regard  to other  systems,  it should also  be noted that the
Multimedia Environmental  Goals  (MEGS)^ was  considered as  a  basis  for  HAPPS.
Review indicated that  for ambient  air  all  the  parameters  used by MEGS were
already  represented  in  HAPPS  in  a form  more  appropriate  for prioritizing
compounds.   MEGS  establishes "estimated  permissible  concentrations"  (EPC's)
and  "minimum  acute  toxicity levels"  (MATE's)   from  threshold  limit  values
(TLV's),  National  Institute  for  Occupational  Safety and Health (NIOSH)  stan-
dards,  and toxicity data  available  in RTECS and/or other  standard  references.
The  SPC's and MATE's  are  oriented  towards  establishing  ambient and  emission
limits for sources  rather  than toward comparisons  between different  pollu-
tants.   In establishing ambient and  emission  limits, MEGS  provides  a tool  for

-------
                                      11
detailed  evaluation  of  the  impacts  of particular  sources,  not  a means  of
comparing different substances  emitted by sources with many different  emission
characteristics.   Thus,  the  use  of  MEGS as a  basis  for HAPPS was  rejected.

       Oncogenicity.   Table 1.2 compares the criteria used for oncogenicity in
HAPPS and the ORNL procedure.   It is  anticipated that most of the  data used in
scoring a compound using  these criteria will relate  directly  to  carcinogeni-
city but the definition of the  factor and the data in RTECS include neoplastic
and equivocal effects  as  well.   It was  considered reasonable  to  use  all  data
related to  tumerogenic effects during the  prioritization leaving  distinctions
between types of effects  and  their relationship to cancer for later considera-
tion by experts.
       The  HAPPS  criteria involve several  modifications  of  the ORNL  criteria
for this  factor.   First,  since Air Program's  principal  concern  is with human
health, criteria  explicitly  recognizing this concern were  added  and  weighted
more heavily  than  criteria related to evidence  based  only on  animals.   Addi-
tional weight was  also assigned  to oncogenetic  effects  in humans  if caused by
inhalation  giving  the  highest  weight to the route of exposure  of interes.t to
air  programs  (items  1 and 2).   The  distinction based on  inhalation  was  not
made for  the  criteria  based  on evidence in  animals,  because positive studies
by any route  of administration in nonhuman species were considered as reason-
able  indicators that  additional  study would  be warranted.   As   in  the  ORNL
procedure,  evidence  in two or  more animal  species was considered  a reason for
greater concern than evidence   in a  single species  (items  3 and 4).   A cri-
terion determined by a  substance's status under  the  National Toxicology
Program's (NTP) Carcinogenesis Testing Program was introduced into HAPPS (item
8).   Selection for  testing  under this program  requires  a determination that
concern over  a substance's carcinogenic potential  is justified  and  that  the
degree of concern  is greater than  that associated with substances not selected
for  testing.   Both of these determinations  were considered sufficient reason
for  air  programs  to consider  looking at a substance  in  more  detail  but were
not  considered as  important   as  actual data.   The  ORNL criterion  based on
determining  that  a  substance  was a  precursor  to cancer was  dropped because
summary data  would not be found.   Both of the criteria requiring professional
judgment  were dropped,  because such  expertise  could  not  be assumed under the

-------
12
























X
4»J
«r4
U
• *4
C
41
ac
o
o
c
o

w
0


CO
• 1-1
U
Hj
4J
• p^
Vrf
U


CN
•
^4

0)
i— *
_2
CO
H



























CO
J
Z

o
















&
CO
,p4
^4
41
4J
• vl
lj
O








41
h
O
U
CO





















co
Cu
0.
^
K























J3
to
t*4
kl
41
4J
• ft
k.
o










X
w
co u
•0 £.
C 00
0 •*
U 0)
0) 3
to


X
CO £
E 90
U CU
ft. 7



e
41
*-<





^4
o

o

4J

u
o

CD
C
CO
i
5;





i^





•
e i
O co
• *4 »-4 i»4
4J to to
to j: E
t— i C -H
to '^ C
£ C to
C 0
•* C 4)
M
xx o
•O ^ S

co co >-
c c . o
CO !0 C
E E 0 0

^ Z AJ ^*


r



r** u^
* *
o o







in •* co






— i CM e*i



CO
0)

o
41
a
a •
4)
f— 1 4J
CO 3
S 0
•«4 M
C
to x
c
4) tQ
0 X
E ^











•
01

3
0


>^
c
CO

x
.a

to
41
• H
U
41
a.
to

























u
-c •*
1 4! U
tB C CO 4J
41 41 CO g
•-t G 4) .O «
0 O 1- -< 1-

o. . 
CO CO ••<
3 01 4J
CO U-4 CO
0 1 90 1
JJ W 1 01 1
3 a • c

e c 41 41
CO O 41 4J CJ
« E 
Z J3 •— . 
4J
4>
^-4
a


u

• w


o
iu

T3
41 •
e <

a.
c o.


•o e
41 .^

e c
41 0)
CO >
41 •**
— 1 U 00
CN a
4>
41 U
41 CO
ce to
tO '^
0 'C 4)
U 4) 4J
•^•4 W
•O t-i O
41 U

-------
                                      13
guidelines for HAPPS.   The criterion based on tnutagenicity testing  was  dropped
for  two  reasons.   The data  in RTECS could not be  used to identify those
particular  tests  for  mutagenicity  which, like  the  Ames  test,  are  standard
protocols for assessing carcinogenic potential.  More  important, even  if  such
tests  for mutagenicity could  be identified in RTECS,  it would have been
improper  to  use  them  to score a  substance in  both the  oncogenicity  and muta-
genicity  factors because  such  a procedure would amount to  a double  counting.
The  HAPPS procedure  corresponds  to the  presentation of  the  data  in RTECS,
avoiding  both  the  possibility  of error  involved in  using some mutagenicity
data to score the mutagenicity factor and other mutagenicity data to  score  the
oncogenicity  factor  and,  more  importantly,   the  error  of double  counting.
Mutagenicity  tests  carried  out  to  screen  for  carcinogens are  included  in
scoring  the  mutagenicity  factor  and hence  affect  the overall  scoring  of  a
substance  in the carcinogenicity group  which depends  upon the two  separate
factors  for  oncogenicity  and mutagenicity.  Thus, mutagenicity screening  for
carcinogenicity is taken into account in the  final  ranking of a  substance  even
though such  tests are  not explicitly singled  out in  a  specific  criterion under
oncogenicity.   The  criterion for negative evidence  was  retained  but  with  no
judgment as to  the adequacy of the evidence being required of the user.
Finally,  a criterion  for "no data" was added  to HAPPS.
       Weights  were  assigned in  HAPPS  to be  reasonably  consistent with  the
weights  assigned in the ORNL  procedure  and thus  to  retain as much  as possible
of  the  expert  judgment  as to  the  relative   importance  of the criteria  that
went  into developing  the  ORNL  weights.   Generally  speaking,  an  attempt  was
made  to  match each HAPPS  criterion  to  a  similar ORNL  criterion allowing  for
the  differences between the two procedures.  Where reasonable matches could be
made,  the matched criteria  were used as benchmarks  and assigned weights  equal
to  the corresponding  ORNL  scores.   The  weights of  unmatched  HAPPS criteria
lying  between  two  matched criteria were obtained by interpolation  and  the
weights  of  unmatched  HAPPS criteria  lying beyond  the range of the  ORNL  cri-
teria  were  obtained  by  extrapolation.    In  some  instances,  this  procedure
produced  inconsistencies  in  the  relative weights  of  various  HAPPS  criteria.
These  instances are  noted  in  the  following discussions  of   the  individual
factors.   To guarantee complete  consistency would have  required  changing the
relative  scores (weights)  assigned to  the  benchmark  criteria; retention of

-------
                                      14
these relative weights was  considered  more important than  the  achievement  of
complete internal consistency.
       The HAPPS  and  ORNL criteria  for  items  3,  4,  and  9  in Table  1.2  were
felt to correspond  reasonably well and these weights were  used  as  benchmarks
in developing the other weights  in HAPPS.   Since  the ratios of weights  within
a  factor  and not  the  difference between  them  give  relative measures  of the
importance of  the corresponding  criteria,  a constant  scaling factor  can  be
applied to the individual weights  without  changing  the  relative  importance  of
the criteria.  For oncogenicity,  a factor  of  1/3  was chosen.   Scaling in this
way compresses the scale  and thus maximizes the effect of additional secondary
weight  throughout  the range of primary  weights.   In general,  the  scaled
weights in HAPPS are related to the corresponding  ORNL scores by
       (Weight in HAPPS)  - (Scaling Factor) x (ORNL Score).
Applying this equation with  a scaling  factor  of 1/3, the weight for item 3  in
HAPPS  is  3 (=» 9/3) and  the  weight for item 4  is  2  (» 6/3).   (Note  that the
ratio of the weights of  the  two  criteria  in HAPPS is the same as the ratio of
the  scores of  the  corresponding  criteria  in  the ORNL  procedure   (3/2  - the
ratio  of  the weights  in  HAPS =  1.5 • 9/6 =  the  ratio of the  ORNL  scores).
       HAPPS contains  two  criteria (items  1 and 2)  designed  to emphasize the
primary concern of  air programs  with human health  and exposure by inhalation.
Both of these criteria would correspond  to a single criterion (item 3)  in the
ORNL procedure so that assigning weights to them required extropolation beyond
the  initial  correspondences  between the  HAPPS and  ORNL  criteria  established
above.   In  the  ORNL procedure,  evidence of  oncogenicity  in humans  would
receive  a score  of 9  (item 3)  and  evidence  of  oncogenicity  in  one  animal
species would  receive  a  score  of 6.   Thus, the  score  (or  weight)  attributed
to  evidence  in  humans  is 502 greater  than that attributed  to evidence in one
animal species.   It was felt  reasonable to use half  as great an increase  (25%)
in  HAPPS  as  a factor for determining  the  increase  in weight to be associated
with evidence of oncogenicity in  humans by  a  noninhalation route  (item 2) in
comparison  with  evidence  from  two  or more  animal   species  (item  3).   This
procedure  gives additional weight  to human data reflecting  the concerns of air
programs  but  less additional weight  than the ORNL procedure gives  to evidence
 in  a second  animal  species, the  lesser  weight reflecting the  fact  that the
expert  opinion  embodied  in  the  ORNL  procedure considered human  evidence or

-------
                                      15
evidence in  a second animal  species  equally important.  The  same  factor was
also applied  to extropolate  from  evidence  in humans  by noninhalation (item 2)
to evidence  in humans by  inhalation  (item  1).   Thus, the weight for item 2 in
HAPPS is
       (Weight for item 3) x (Extropolation Factor)
       - 3 x 1.25 - 3.75 * 4
and the weight for item 1 in HAPPS is
       (Weight fo'r icetn 3) x (Extropolation Factor) x (Extropolation Factor)
       - 3 x 1.25 x 1.25 * 4.68 *5.
       The HAPPS  criterion  for item  8 (scheduled  for  testing)  was  considered
to  correspond most  closely  to  the  ORNL  criteria requiring  expert  judgment
(items 6 and  7),  since  it was felt that  expert judgment is frequently used in
developing testing schedules.   The weight  for  this criterion was chosen to be
the geometric  mean of  the ORNL scores for  the  two  criteria requiring expert
judgment taking the scaling factor into account.  Thus,
       (HAPPS Weight for item 8) - (Geometric Mean of ORNL Scores for
       items 6 and 7) x (Scaling Factor)
       = V3 x 1 K  (1/3) = 0.57 » 1.
       "Finally,  the  criterion for no data  (item  10)  was assigned a weight of
zero,  the  same  as negative  evidence.   The  references do  not  reliably dis-
tinguish between  no  data and negative results  so equal weights for these  two
criteria were  considered reasonable.  Strictly  speaking,  this weight assign-
ment means  that  there  is  less need for concern about a substance for which no
data  is  available  in   the  standard  sources  (weight m 0)  than  there  is  for
concern  about  a  substance which has been  scheduled  for testing (weight *  1).
However,  in practice,   the  results of  a prioritization will  be reviewed  and
substances  with   a significant lack  of  data  will be  identified  and handled
separately.   In  terms  of the initial ranked list needed to  start a program of
detailed  evaluation,  HAPPS  emphasizes substances for  which  the most data is
available  and,  hence,   presumably  substances  for which additional evaluation
will  be easiest  in the  early stages  of  the  evaluation program.   Program-
matically,  this  would  permit  time for data  to  be gathered on  data-deficient
substances  identified exogenous to  HAPPS  as  being  of  potential  concern.
Furthermore,  results of prioritization by HAPPS could  still be  of  use in  this

-------
                                      16
effort  by indicating the relative need  for  concern  based on the  limited
available data.
       In addition to  the primary weights just discussed, HAPPS  assigns
secondary weights  for  the  oncogenicity  factor.   Given the  criteria,  it is
possible for a substance to satisfy two or more criteria simultaneously.  For
example, there could  be  evidence  of oncogeoicity  in humans by the noninhala-
tion route and evidence  from  a single animal study.  These secondary weights
serve the purpose of  giving a  substance with data  satisfying  several criteria
a total higher score than a substance  with  data  satisfying  fewer criteria when
the  highest  weighted  criterion  satisfied by  both  substances  is  the  same.
Furthermore,  the  secondary weights  have  been  assigned so  that  a substance
could never  receive  a total score  higher  than  a  substance which satisfied a
single more  heavily weighted  primary  criterion  simply by  satisfying multiple
secondary  criteria.    This relationship  between  the  primary  and secondary
weights is based on the view that  while concommitant  evidence  should have  some
positive influence on the  final  score for a substance, the additional weight
should not be  sufficient to raise a substance to  the  score of the next higher
ranked criterion.  For  example,  a substance  with  evidence of oncogenicity in
humans  by  a noninhalation  route  and  evidence  of  oncogenicity  in two animal
species  should not  receive a  higher  score  than a substance with evidence of
oncogenicity in  humans  by the inhalation  route.   That the secondary weights
achieve this aim can be confirmed  by noting that,  except for a substance whose
primary weight is 5, the maximum sum of  the secondary weights  is 0.55  (« 0.5 +
0.05),  because  the  HAPPS  criteria  for  items  3   and  4 cannot be  satisfied
simultaneously.  This maximum  sum for the secondary weights  is less  than the
minimum  difference  of one  between  any  two primary  weights.   In making  this
test it is not necessary to consider the maximum possible secondary weight for
a  substance  whose  primary weight is  the  maximum  possible, a substance whose
primary  weight  is  5  in the oncogenicity example, because there  are no  more
highly weighted  criteria.   The addition  of  secondary weight can never raise
the  total  weight of  a substance with  the maximum  primary weight above  the
weight associated with  a  substance  satisfying  a more heavily weighted  single
criterion, because there is no such  criterion.
       As used  in HAPPS,  additional  secondary  weight  is  assigned only  when
different  criteria are  satisfied;   additional  weight  is   not  assigned  when

-------
                                      17
several data satisfy  the  same  criterion.   It was felt that some  sense  of  the
quality of the data was necessary  in  assessing  whether multiple  data  satisfy-
ing the same criterion were better  than data  based  on  a  single  study  satisfy-
ing only one criterion.   For example,  a substance which  has positive  evidence
in two animal tests of poor quality  should  probably not  be  ranked higher  than
another  substance  which  has  positive  evidence in  only one  animal  test  of
superior quality.  HAPPS  would, however,  rank  the former compound higher  than
the latter if additional  secondary weight were  assigned  but since the quality
of the  studies  involved  is  not  reported  in  the  readily  available literature,
such occurrences could not be checked  for in  the HAPPS  procedure.  Therefore,
it was  decided, based  on considerations  of  this   type  of  situation and  the
unavailability of data on study quality,  not to assign  additional weight based
solely on meeting the same criterion with  data from  multiple studies.

       Mutagenicity.    Table  1.3   compares  the  HAPPS  and ORNL   criteria  for
mutagenicity.   Many  of the differences between the two  sets of  criteria  are
similar to the  differences already discussed  for oncogenicity  and will  not be
fully discussed  in  this  subsection.   Although  information specific  to  humans
is  likely to be  unavailable,  evidence  of mutagenicity in mammalian  test
systems was given  more  weight in both  HAPPS  and the ORNL  procedure  than
evidence in nonmammalian systems.   In HAPPS, evidence obtained  from inhalation
studies  in mammals  (item 1)  receives additional  weight  as  being  directly
related to the  route  of  administration  of interest  in  air programs.  The ORNL
criteria (items 9 and 10) requiring expert judgment  were dropped from HAPPS as
were  the  corresponging criteria  for  oncogenicity.   As   for the  oncogenicity
factor, a criterion was added to HAPPS to reflect whether a substance has been
scheduled  for  or  is  undergoing mutagenicity testing under  the  National Toxi-
cology  Program.   Data  in RTECS  do  not  unambiguously identify  tests showing
germinalcell DNA interactions as required by the ORNL criteria for items 7 and
8.  Thus,  this determination, although desirable, was dropped  from HAPPS while
the remainder  of the ORNL criteria specifying  the  type  of test was retained.
Finally, the criterion for no data (item  13) was added.
       With  these  changes,  the HAPPS criteria  were  expanded  according  to the
following  scheme.   Evidence of mutagenicity from  mammalian test systems  was
considered  a better  indicator  of potential  effects  in  humans  than  evidence

-------
13
1



























^
tf4
y

c
01
ac
9)
y
JC

b
0
IM

eg
.^
u
OI
u
• *-(
u
^


ft
*
— «

01
i— i
.C
eg
H





































eg

z
oi
O










>









r;
eg
I*
Oi
AJ
• t»4
U
o










u
O
y
VJ


























en
0.
a.
^
^




































,£
eg
• H
b
01
AJ
• «-i
p
u
















u
CO U
•3 J=
S DC
O "*
y oi
oi 3
en

X
eg .=
E 90
•H *f-4
14 0)
e. 3



S
C ^J 1
CO 00 W
•H 0)0;
^* ij jj
CO C
^ 0 >H
S i-
eg AJ <
E -HZ
> Q
41
1-1 C -*
O •« 1-4
£ 01
5 oi y
C 1
0) C r-4
u eg
o e E
Q eg **"4
V4 • AJ C
O 99 01
AJ 4) 10
41 CO k-i
e o> o u
O AJ 2£ 0


CT* ^O




.
93
CO 93 CO
4) 01 U C
AJ AJ AJ C eg
eg -H
sec •-* • —<
9j 91 as r* AJ eg
•rf -H .H • eg 93 E
§01 £
AJ eg
S E E 4) eg e
E E E AJ S e 2
3J CQ CQ £ CQ Q
E E ECO •-< c
93 C -*
O O Q -H eg o
> > >- -- O E U
—4 .»4 AJ  E AJ
> > . -H E — ' eg .^
S > S > E >
e c c «g e
••* • — ' -H C E C O C
e u .* .-i c -H
Oi o 01 eg Q
C *H C —^ 4J W 4J O 4)
OAjQegbAJt-4 ^b
03 j= 0 -H o -HO
AJ—UCS>S >E
co eg co »»^
9]^3 egc« b C w* C w*
4IC 0)0 O-i-l O 03 •-( O M
—4 <*4 i— 1 C AJ AJ
O 41 O CO 4) O 91
4-i^Aj>%3 c jyoj c j*cu
^^3 
^c cy u &. AJ 4) cu
AJ a. y E
AJ C U-l 4) 60 4)
CO -»4 C O C. T3 >
cu O n 3 -^
O AJ < AJ 3 "I AJ
> Z -c C 91 1 eg |
-4 O a 0) 4) -J 1 90 1

AJ <-* eg 60 3 C C
c •-• ^ js -a ,00
rt > 4) 3 — < 01
y AJ •••I eg co AJ
C C 1 U AJ 00 9)
O -H r-i 4) -J • cfl 0) 3
•J 9lD.eg9j-ovt4 cr
AJOICCOCAJ o 4)
O C *^ 3 O CO O l^ "C
cO (2 £ ^^ * ^ "0 2» Q" ^1


 4) 41
f-i > eg
C 3 — 4 AJ
.* -o AJ eg
cy eg Tj
4) .CM
c y 4) O
O en z Z




V^i
-^ CM O O
* i i • • •
o coo







• p4
10
01
b
eg

eg
•fn4
Vj
OI
AJ
• *4
y

en
a.
a-
:S
V
ij

M-J
O

93
C
0
AJ
eg
u
• H
U-l
• f-4
y
OI
a.
eo

01
AJ
OI

c_
E
O
y

• w
S
u
0
VA4

•a

^
Oi
AJ

O

"oo
c
'
"O
4)

—4 e
eN 4)
en
4)
4) 0.
eg
§ '£ •
V- 4) <
«-l U

13 U --4
v v •?
AJ C
C. AJ O)
eg co c.
^j C "«
< z <
eg .a

-------
                                      19
from  nonmammalian  test  systems.    Thus,  mammalian  tests  were given  higher
weights than nonmammalian tests just as in the ORNL procedure.  Within each of
these groups, in vivo tests carried  on in the living body of an organism were
weighted higher than in vitro  tests  conducted outside  living organisms.   This
is an  extension  of the  separation of  in  vivo and  in vitro mammalian tests in
the ORNL  procedure (items 2  and  7).  Finally,  HAPPS  gives  additional  weight
when evidence of mutagenicity is available in more than one test system except
for  in  vivo  mammalian  tests  for  which  a distinction based  on  inhalation or
noninhalation  route  of  administration  was   thought  to  better  represent  the
concerns of  air programs.  Distinctions  based on both the number of tests and
the route  of administration  within  the  in vivo  mammalian  category would have
been  too  detailed  for  this screening  level  of analysis.   Assignment of addi-
tional weight for  multiple tests  is  consistent with  the treatment of in vitro
tests with no germinal cell interaction in the ORNL procedure (items 7 and 8).
       As  shown  on Table 1.3,  items  2,  7,  8, and 12  were used as benchmarks
in developing the  weights in HAPPS.  The  weight  for item  11  (scheduled for
testing) was  assigned as  the geometric mean of the two ORNL criteria requiring
expert  judgment,   just  as was  done   for  oncogenicity.   Thus,  the  weight is
2  (Y3  x 2 = 2.4 **2).   For the four criteria listed as  items 3-6, interpolation
in equal  multiplicative  steps was used.   For five  steps  between 6 and 9, the
factor  is  about  1.0844  (6 x  1.08445  « 9).   Using  this factor and rounding to
the nearest  tenth, the weights  given  in Table  1.3 were obtained.  For example,
the  weight  for  the criterion for  two or more in  vivo  nonmammalian  tests
(item  5)  is  6 x  1.0844^ » 7.055  * 7.1, because it  lies  two steps above the
benchmark  criterion  (item 7).   The weight for item 1  was  obtained by extrapo-
lation  using the  same  factor as  was  used  for oncogenicity:  9 x  1.25  «sll.
It should  be  noted that  the same  factor  of  1.25  could  be derived by using  one-
half  the increase  of 50Z applied by  ORNL  to account  for additional in vitro
tests  or mammalian tests of  mutagenicity (see items 2,7,  and 8) thus indicat-
ing  that  a  consistent   approach  was  applied in developing  the  ORNL scores.
        The  type  of inconsistency  noted in the general discussion shows  up for
mutagenicity criteria.    Several  examples,  but  not  a complete  list, follow.
HAPPS  and  ORNL  increase  weights by a  factor  of  1.5  as evidence becomes  avail-
able  in a second nonmammalian test  system in vitro  (items 7  and 8).  It would
be desirable to have  the same factor  apply  to the  criteria for  nonmammalian

-------
                                      20
in vivo tests (items 5 and 6)  and  to  mammalian  in  vitro  tests  (items  3  and 4)
but  in these  cases the  factor  involved  is  only about  1.1  (as7.1/6.5 and
«3.3/7.7)  which  is,  of course,  simply  the interpolation  factor  used to
determine the weights.   Similarly,  it would be desirable  for  the ratios
between the  weights for  criteria  which  are the same  except  for the  type of
species involved to be equal.   Thus,  the ratio of the  weights  of items 4 and 8
which correspond to evidence from one in vitro test in nonmannnals and  mammals,
respectively, should ideally be equal  to the  ratio of the  weights  for items 3
and 7 which  correspond to evidence from  two or  more  in vitro  tests  in mammals
and nonmammals, respectively.  However,  the first  ratio  is  about 1.9  0*7.7/4)
while  the  second  is only  about 1.4  (*8.3/6).   The  weights  could have  been
assigned  to  avoid  this  type  of  inconsistency  but only if experts  had  been
available  to provide a  reasonable set  of  weights for  the components   to be
matched.   In view of the unavailability of the  needed experts,  the procedure
used  here  of  retaining  the expert  opinion embodied  in the  ORNL  scores  and
interpolating  and  extropolating  to express the  needs of air  programs  and to
more  fully utilize  the data in RTECS  to spread  out the  final  scores  was felt
to provide a reasonable approach.
       As  is the  case for  oncogenicity-  HAPPS  assigns  secondary weights  for
mutagenicity  when  more  than one  criterLon  is satisfied  by a  substance.   The
previous  discussion for  oncongenicity  is  equally  applicable  to mutagenicity
and  a similar confirmation can  be  made  chat additional secondary  weight  can
never  cause  a  substance  with  a particular  primary  weight to receive a greater
total  score  than  a  substance  with  the next  higher  primary  weight.   Additional
weight  is  not  assigned  under mutagenicity when  a  substance has multiple data
satisfying the same criterion.

       Reproductive and Developmental Toxicity.     As  noted in  Sec.  1.2,  the
ORNL factor for  embryotoxicity and  fetotoxicity  and  the ORNL factor  for
reproductive effects were combined into  the single HAPPS factor for reproduc-
tive  and developmental toxicity.  Since the structure of  the ORNL criteria for
both  factors is  very similar, Table  1.4  lists  them  together  in the interests
of  simplicity for  comparison  with  the HAPPS criteria.  For  this  factor,  the
rationale  for  the  changes between  HAPPS  and the ORNL procedure is the same as
that  for  the two  previous  factors.   Additional  weight is assigned to  evidence

-------
21








^
4J
If4
U

x
o
H

i— i
to
4J
C
01
s
a
c

01
>
0)
o

•c

OJ

OI
.>
u
3
TT
O
kl
a
01
a:

ki
0

03
• H
kl
01
u
• H
kl
0


-3-

^

OI
l»H
X!
0)
H



















03

Z
as
O

















•^
• H
^j
(U
4J
• *•(
^j
o









0)
ki
o
u
CO


















in
a.
a.
•y




























Xi
0]
kl
41
4-1
• *4
kl
CJ








>^
^
Q XJ
•c x;
C 00
O —i
O 01
oi S
C£



>•,
kl 4J
o! x:
S 00
• 1-1 ••**
k, 4)
a. 3



S
CJ
u
)— 1

i
U

O-
o
0
5


^4
o

09
C
0}
E

X





^s





«
C 1
O "3
• •4 1*4 ^
U W Q
as x: S
•-J C •«
03 .« e
•i § "
•- C 01
kl
XX O
XI J3 E

CO 03 1-4
c c • o
OJ fQ C
£ S O O
•2 ~ w H






r*» in
• •
O 0









U*^ +jf fO







~H c^ en


9
0)
•H
u

^
3)
(— 1
aj
s

c


4)

03
• H
kl
a











•
01
4J
3
0
W
x
C
0]

>^
XI

01
01
••4
U
0)
a.
01





























1 -<
O «3
ki • u C
CO O- 4J 03  O E ^ u "^
4) 4) 00 3 01
a. -o T3 y a «<-i
0 4) 1» 3 — * O
> CO "-5 kl J-l k.
^ -H as 4-1 3 a
03 X X! •-< « X>
E 3 • 03 E C u
•H -HO 4J C 03 «3 O C
C U O O W 4J 4)
(8 C4) til 'H <8 oj-op
O U & CO On ^ 4) 00
oi u c 03 oo co T;
C COI SllW-i O«33
(5 MT5 W^-iO ZXl"-i





SO «^ CO "^





^
o

to
01
• »4
U
OI
a.
03 1 1 1
1 1 1

oj OI
E u
.- 3
c; o
re u

0) X
e e
O 03






m
•
0










< 0)





0




*
oo
c
• H
4J
CO CU
01 U
4J C
01
i- -o
o •*
U-l >
0)
•o
01 01
i— 4 > 03
3 •-< u
"O 4J 03
4) 03 *O
x: oo
u oi c
w z z





i/*i
o o o

o o o









-< O O







ao a* o
•—4


p
O


1
03

5-*\
u
•^4
U
•H
x
0
J-*
0
4J
0!
^4-J

•a
e
03

>•,
4J
• F^
o
• «H
X
0

o

^4
i
0)
kl
o
UJ

»^
fsl

•
U-l
4J

C
• »•<

03
p
O
4J
O
03
14^

O

4J



kl
<4«4

T3
4i
jj
a
93
•w
<
CQ


•
03
• H
4)

• H
kl
U

u
• H
01
x:
4^

kl
0
U^

01
U
^
3
4j
U
3
W
u
00

Vd
ct]
f-^
*f^
E
• ^*
en
cu
>
03
x:

x:
u
• H
r-
3
14-1
O

Si

o
XI

CO
4J
Cj
4)
U-l
14-1
01

41

• <^
4-1
U
3
"O
o
kl
c.
01
kl


w
ex
a.
^
=
4)
£.
^

^M
o

00
fi
o
•i-4
4J
03
U
• M
vu
• w4
O
4)
a
03

0)
4J
OI
v— 4
a.
£
O
u
• M
g
O
vu

T3
01
C
01
4J
kl
0
CO

e
,r*

•V~
OI
4J
c
01
CO
4)
t.4
e.

OS
• »4
kl
0)

.^4
kl
u

4J
CO
O
s
XI












































•
^

X
• *4
•o
c

a.
c.
<

c


c
01
^
• H
00

41
kl


03
•••4
kl
01
u
• H
ki
O







•
•
CO
4J
(J
0)
vw
VM
4)

01

• iJ
4J
U
3
•o
0
kl
0.
01
ki

kl
0


kl
o

u
cd
U-l
^J
Z
06
C

0)
x:
4J

C

^
p-^
e
o

91
kl
03
01
a.
c.
S3

e
o

U
o>
4J
.^,4
kl
U

co

£.
r-
o

-------
                                      22
of effects in humans and to evidence of effects caused by inhalation in huaians
to  emphasize  the interests  of  air programs.   Three ORNL  criteria requiring
expert judgment  (items  5,  6, and  7)  were  omitted.   A criterion  related  to  a
substance's testing status and a criterion for scoring substances with no data
were added.
       The assignment of weights  for  this  factor  parallels  that  for oncogeni-
city.   An overall  scaling factor of  1/3  was applied and  items 3, 4,  and  9
served as  the  benchmarks.   An increase of 25% was  used  to  extropolate to the
criteria  for  humans (item  2)  and for  humans by  inhalation  (item 1).   This
increase is one half the increase of 50% assigned by ORNL for positive results
in  an additional  species  (compare ORNL scores of  9  and  6  for items  3 and 4,
respectively).   The  HAPPS  weights for  items  1 and  2  can be extropolated from
the  HAPPS weight for  item 3 by using an  extrapolation factor for  a 25%
increase just as was done for oncogenicity.  Thus, in HAPPS
       (Weight for item 2) » (Weight for item 3)  x (Extropolation Factor)
                           - 3 x 1.25 - 3.75 w 4
and
       (Weight for item 1) » (Weight for item 3)  x (Extropolation Factor)
                           x (Extropolation Factor)
                           * 3 x 1.25 x 1.25 - 4.68 * 5.
As  for  oncogenicity,  the  criterion for compounds  scheduled for  testing  (item
8) was weighted at the geometric mean of the ORNL criteria  (items 5, 6, and 7)
requiring  expert  judgment  and taking  the  scaling factor  into  account.   Thus
       (Weight for item 8  in HAPPS) " (Geometric Mean of ORNL Scores for
                                      Items 5, 6, and 7)  x  (Scaling Factor)
                                    - (4 x 3 x 1)1/3 x (1/3) * 0.76 *  1.
Secondary  weights  were  also assigned  in  the same  fashion  and  subject to the
same  Limitation  as  was  done for  oncogenicity and  mutagenicity.   These latter
two  effects,  however, are  generally  believed to  exhibit  either no  or  very low
thresholds  while it  has  been  theorized  that reproductive and  developmental
effects  exhibit  thresholds.   However,  the thresholds may be  very  low or even
zero  so  the system of secondary weights was used rather  than a system  based on
lowest effective  dose as  might  be done for effects exhibiting thresholds when

-------
                                      23
dose-response data is  available.   In this regard, the  criteria  for  oncogeni-
city,  mutagenicity,  and,  with  the  above caveat  in mind,  reproductive  and
developmental toxicicy  should be  compared to  the  criteria for effects  which
exhibit  thresholds.   In  these  factors,  no secondary weights  are used; a
substance's  rank  depends  only  upon the  lowest  recorded  dose  producing  the
effect under consideration.

       Acute Lethality.   Table  1.5  compares the ORNL  and HAPPS  criteria  for
the  acute  lethality factor.   Both  systems  base their  rankings  on  data  for
lethal dose  and  lethal concentration.   Rather than  retain  the  standard  toxi-
cological  term  "acute  toxicity"  used in  the ORNL procedure,  the name of  the
factor was  changed  to  "acute lethality" in  HAPPS to reflect this use  of data
on  lethal  doses.   This  change emphasizes  to the nonexpert that  only  data  on
lethal doses should be  used  in scoring  this  factor  and, in fact,  HAPPS speci-
fies that only data so identified in RTECS be used.   Data on chronic  lethality
and  nonlethal effects  both chronic  and  acute are to be considered in  scoring
under  the  factor  for  effects  other than  acute  lethality.   If  only  chronic
effects were considered  in the factor for  effects other than  acute lethality,
then the two factors would correspond relatively  well to the traditional  acute
and  chronic  categories  in toxicology.    Since both  factors are  scored  using
distinct sets of  toxicological  data as  reported  in  RTECS,  the  approach  taken
in  HAPPS  corresponds  to  the available  data.  An appropriate  balance  between
the  two  factors was obtained by  the assignment  of  their  relative weights  in
developing an overall score for the toxicity group.
       The  major difference between  the two  procedures was  the  deletion
of  criteria  for the intravenous, subcutaneous,  and  intraperitoreal  routes  of
administration  from HAPPS.  While these routes of administration are  important
in  detailed toxicological  evaluation,  they were  considered  less  closely
related  to exposure through  the  ambient  air  than  the  inhalation,  oral,  and
dermal  routes.    In  addition,  these  later  three routes  were  the only  three
considered  by  the Iteragency Testing Committee^  and in the Michigan Critical
Materials Register*7 which were used in developing  the  ORNL procedure.  ORNL,
however, opted  to be more  detailed  in their  approach.
       A  second difference  between the  two procedures  does  not show  up  on
Table  1.5  and concerns the definition  of the period of exposure which defines

-------
24
































^

^^
03
"3
1*
jj

2
—

^

u
3



-3
"^
U
^
w
• •4
^
, ,
s~'


'J*"l
*
«*

w
— *
^
"^
."*
?"*


















































a

z
z
O
















































^
^,
i.
^
™








1












'










1


41

y
«
.3
a
a
a

^
c


a

Is
a

41
U
3
a
a
a.
x
u















"3
•
u

A
U

3
^

•o
c
^

V
u
2
a


^
^
™
a
a

X
'J*









1
1

























—
•3
5
u
y
a





•*
a
w
O




e
a

u a
fl 9
"•4 y
«» >-<
~
e
**


u
a
u
40



•*
a
a
a
a






~-«
9
U
O






i


%

S — *
3 C

w
4
^
41
jg
w

' «
,fl


1




•
4)
• •4
U

*<
VJ






>^
W w
2 —
•~ ~
i. 3



=
^
-«
O O
a o o o •
o a o a ><

1 1 NX 1 1 1 1 1 /\ •— 1
«i a a >
a a -I
u
3

o a —
a a a  «
a a a a a n 4

1 in m I i I 1 I in u
\y\^ u*imaayN>a
' ^ a •- z

fl
M
5J
a z
a
a a a
a a a - a
__ a a a a a

u^ i ^-. i i i i i a
X. 'v- -3 a a a —
'^ m a a /v
vrt a
m



a
a a a
a a a a a
u*^ i/> m in ^n — • a
X, I X- ' ' ' ' ' "*
m ^i a a /N
-x"l O



a a a a a
.5 • .3 .3 .5

c c c c c


! ccaceacccce
• sasaasa^aao
S|'-3S--S2S33

™^j*;^;^^^^™™S





'-"•"-O^O^OOCOOO
•sj " 1 ,-"»  O *™










*
•*« *
a
x u
"* U*
™ 3
C *
U O
a. a
a x
< 4)
c — »
.•* 4
«
e c
u o
> u

* QO ^
u
U CM
U ^
at u
u
4) -O
U C
• ^ D
U "
01
W3 3
SU O
£U • U
•< •" C
=: o «:

4* U 3 — i
— « y
u *j «l3 *
3 M
u u a •—
a u '•«.

a 41 a c
S a — < 3 •-<
a -s o« a
.- •- ecu
u «* ••« 4) C
« a « > — •
v a a ^
— « -u u
•— •- u X
•-a a = »j

41 C —

a" g a ^
~- a u
•u 10 u rj u
uS ^s 1) —
^ oo a
— « « .— a u
C. a 41 J w
a ii 3 s
a a a u
u a -• * TJ a
30 a s
3 3 ac
S u c- o =
^ a 4) 3a *^
O vu, 2 a
*• j= a 3
> u U
•as •- x
il c. 3 a j
C 3. MJ 41
u i 50^ ** a —•
h j; ao • u «
a e -v. g ao a u
••S 3 30 •
• ••"31 -0
U f*H » ^
. s * | — — 3
_ .-, — « a * i
-*( ^ w ei 3 ?!,
^3 JS ^ W 5 5J '—'
• ^ C ^ «) ^ ^
*< w >~ O a •- 3 a
41 e -35
2C 41 a 4) Sfl
a .. 4) 
— - s i, —
a ••• a i. 41 i-
— i. O < — SI
< y a s •- a
« -3 U -3 a •*•

-------
                                      25
an acute  exposure  for the  inhalation  route.   The  ORNL procedure and  its
predecessors  follow  the   standard   toxicological definition  by  considering
inhalation exposures as acute  if  they  are  under about 4 hours in duration but
do allow  the  use of professional  judgment in  evaluating  studies  with  longer
exposure  times.   Expecting such  judgments  would, however, violate  the  HAPPS
guidelines and in HAPPS,  inhalation  exposure  times  up to 24-hours in duration
would  still  be  scored  under  the  factor  for acute lethality.   This  value  was
chosen because it covers  the  complete  range of the  current short-term ambient
air  quality  standards   and  because  it  is  frequently used in  discussing  the
health effects of air pollution.
       As shown  in  Table  1.5,  only  inhalation doses   for  gases  given in ppmv
could be scored using the ORNL scale.  RTECS gives exposure data for solids in
units of mg/nH which cannot  be meaningfully converted to ppmv,  a natural unit
for  gases because  equal  volumes of  different  gases at the same conditions of
temperature and pressure contain equal numbers of moles.  Reference 25 gives a
toxicity  ranking for  dusts  and  mists  in mass  concentration  units and  the
corresponding volumetric  concentrations  for particular gases.   For example, a
particular mass  concentration of  a  certain solid might  be considered "highly
toxic" while a particular volumetric concentration of a certain gas would also
be considered "highly  toxic."   Only  two  pairs of corresponding concentrations
were  available.   In Fig.  1,   these  pairs  have been  plotted  on log-log  graph
paper  and  the  straight line  labeled "Acute Lethality"  has been drawn through
them  and  extropolated.   The  line so generated was  used to establish the mass
concentrations measured  in mg/nr*  coresponding  to the ORNL volumetric concen-
trations  in  ppm's.   In this  way,  the  HAPPS scale for solids was constructed.
Log-log paper was chosen  for  the extropolation because the toxicity scales are
generally close  to  being  logarithmic rather than linear as might be expected,
since  the exposure  ranges  involved in  toxicological experiments  can  cover
several orders of magnitude.   Also  shown in Fig. 1  is a  line representing the
correspondence between  volumetric  concentrations of gases and mass concentra-
tions  of  solids  for use in ranking existing standards.  The development of the
points  through  which this  line  is   drawn  is  described later  in this section.
Although  developed  heuristically,  three  of  the  points lie on  a straight  line
that  is  exactly  parallel  to  the  line  drawn for acute lethality.  This paral-
lelism  shows  some  degree  of  underlying  consistency  between  the approach used

-------
                     26
                                                                            e
                                                                            o
                                                                           •H
                                                                            4J
                                                                            C
                                                                            01
                                                                           CJ
                                                                            33
                                                                            en
                                                                           •a
                                                                            u
                                                                            4J
                                                                            IU
                                                                            o
                                                                            c
                                                                            01
                                                                            O
                                                                            U-l

                                                                            en
                                                                            a)
                                                                            ^H
                                                                            S3
                                                                            y
                                                                            C/3
                                                                            eo
(Aiudd)  UOIJBJ4U80UO

-------
                                      27
for acute  lethality and  the  approach used  for  existing standards  and  lends
support to  the extropolation used  for the  "Acute  Lethality" line.   Another
approach to  ranking both gases and  vapors  would have been to have  converted
the volumetric  concentrations  for gases to  mass concentration units.   Since
such conversions involve the molecular weight of the  gas,  this  approach  could
lead to cases where two gases with equal lethal  volume concentrations but with
different  molecular weights would  be given  different weights on a mass concen-
tration scale.   On  the  ORNL scale  for volumetric  concentrations  they  would
receive the  same  rank.   In addition, Ref. 25 indicated  that  the  mass concen-
tration scale  should be  used for  dusts and mists.    To  avoid inconsistencies
with the established volumetric concentration  scale for gases and  in view of
the recommended use of the mass concentration scale, separate  scales for  gases
and solids were included  in HAPPS.
       A final difference concerns the relative  weights assigned  doses greater
than those  corresponding to item  6  in Table 1.5.   In HAPPS, these exposure
levels are all  scored  as  having a weight of zero, the same weight assigned to
negative evidence.   For  the inhalation  route,   the  dosages  associated  with
these items are all higher than 50 ppm, a value  rarely exceeded in the ambient
air unless  under  accidental conditions.   In the absence of  other  indicators
and since the  effects  being  considered  exhibit  thresholds,  these  high dosages
were thus  considered as poor indicators of  a need for concern by air programs.
Little  need  for  additional study seemed  warrented  if  the threshold  was
unlikely to  be exceeded.   Death  from  nonthreshold  effects  like  cancer,  fetal
death, and  death  from developmental  disorders  are  considered in the factors
for oncogenicity,  mutagenicity,  and reproductive and  developmental toxicity.
Criteria ac  the corresponding dosage  levels  for  the oral and  dermal routes of
administration were  also  given weights  of  zero,  because  they  are obviously of
less concern to air programs than the inhalation route.
       The  remaining differences   between  HAPPS  and  the ORNL procedure were
made  for  reasons   similar  to those  discussed  above in  connection  with  other
factors.   Given the concerns of  air programs with human health  and exposure
via  the  ambient air,  extra weight  was  assigned to  substances  documented as
being lethal  to humans by inhalation  at the dosages  deemed  to  be of interest
(items 1-6).   The ORNL criterion  for  low  or no  biological  activity was  modi-
fied  to  require  negative evidence  in order to remove any  need to  make an

-------
                                      28
expert judgment as to what constituted low biological activity.  In HAPPS, low
biological  activity  is  operationally  defined  as  dosages above  those  corre-
sponding  to  item  6.   As  with  the  previous three factors, a  criterion  for no
data was added.
       The criteria  for  animals at  low  and moderate doses (items 3 and 6) are
            •
identical  in HAPPS  and  the ORNL  procedure  and  served  as  benchmarks.   The
criteria  listed  under item  10 were also  considered sufficiently  related to
serve as  a third benchmark.   Because  HAPPS combined the three  ORNL criteria
for the highest dose  ranges  (items  7,  8,  9),  retention of the ORNL score of 6
for the benchmark  item 6 would have created a  large  gap  at  the  bottom of the
HAPPS scale between the criterion for animals at moderate doses (item 6 with a
score of  6)  and  the  criterion  for high doses (item  7 with a  score  of 9).  To
avoid distortion  from such  a  large gap,  a scale  factor of  1/3 was  used to
reduce the HAPPS weight  for item 6 to 2.0.  The weights for items 4 and 5 were
determined  by  interpolation between  the  benchmark items using a  factor of
1.1447 (6  x  1.1447^  =  9)  appropriate  to three equal steps.  Also, as with the
previous  factors,  the  weights  for the criteria  for  human noninhalation expo-
sures  (item 2)  and  human  inhalation  exposures  (item  1) were  determined by
extropolation using a factor of 1.25.
       As  noted  for  the  mutagenicity  factor,  the interpolation and extrapola-
tion  from  the benchmarks  leaves a certain  degree of  inconsistency in the HAPPS
weights.   For example, the ratio of the weights for human exposures by  differ-
ent routes  of  administration should  be  the same regardless of the dose level.
However,  in  the  low dose range the ratio  is  about  1.3  («4.7/3.7) for  items 1
and 2 while  in the moderate dose range  the  ratio  is about 1.1 (=s2.6/2.3) for
items 4 and 5.  Similarly, the additional  importance attached  to human  data in
comparison  to animal  data  should be  independent of  dose range  for  a  given
route of  administration.   However,  for inhalation exposures,   the ratio of the
weights for  humans and animals at low exposure  levels is  around  1.6  (**4.7/3.0)
(items 1  and 3)  and  around  1.3 (= 2.6/2.0) at moderate doses  (items 4  and 6).
As  previously noted,  this type of inconsistency was  deemed less  important  than
retention  of  the  relative  importance  of the  benchmarks;  and  despite   this
inconsistency, the ratios are  fairly close to being  equal.
       Secondary  weights were  not assigned  for  satisfying more  than one
criterion  under acute  lethality.  Since these  lethal  effects  generally  exhibit

-------
                                      29
thresholds, the  best indication of the  need for  further assessment of  a.
substance was considered to be the  lowest  documented  dose.   In other words,  a
substance causing  death at 1  ppm  and  1000  ppm was  not considered  a  better
candidate for  further  assessment  than a  substance   causing death  at  1  ppm.
Both substances should be and, under HAPPS are, given equal  weights, assuming
that both sets  of data  corresponding  to  1 ppm are  for the same species  and
route.   This  situation  is  different than for  the  nonthreshold effects, onco-
genesis and mutagenesis, for which  all  the criteria  satisfied  by the data for
a  substance  are considered in  developing an  overall score for  a  particular
factor.

       Effects Other than Acute Lethality.     Unfortunately,   little  data  is
available for  scoring this factor.   Ideally,  this  catchall  factor  could  be
split  into  several categories such  as  acute  (nonlethal),  subacute (lethal  and
nonlethal),   and  chronic (lethal  and  nonlethal)  with,  perhaps,  additional
breakdowns  into  toxic and  irritant  effects.    It  is  precisely some  of these
chronic and  irritant  effects  which correspond  to  the  types of  exposures  and
effects frequently  of concern in dealing with  populations  repeatedly exposed
to  low concentrations  over  long  times.   In  dealing with  toxic  pollutants,
however, air programs is primarily interested in serious health effects.  Very
little data  on  chronic  exposures is available  due to the  time and  expense  of
conducting  controlled chronic  experiments.   Since  the  data  necessary to make
distinctions between  various  types  of  effects are not  readily available,  all
effects other than acute lethality  are  scored  under  a single factor in HAPPS.
HAPPS  is  also  an extension of the  ORNL procedure  which specifically requires
chronic  exposures with  repeated  doses over  a period  of  time like  weeks  or
months.   In  fact,  the QRNL  factor is  titled  to indicate that chronic toxicity
is  being  scored.   The  extension in HAPPS to  include acute  nonlethal effects
makes  fuller  utilization of the data available in RTECS.   It  should be noted
that the terminology used in HAPPS is employed to  match the material presented
in  RTECS  which  itself  does  not  always  employ the  terminology of standard
toxicological protocols  and that the studies  reported in RTECS  were not always
done under  standard protocol.   For  example,  lethality should not occur during
standard  chronic  tests  so  that the term chronic lethalitv should not be used

-------
                                      30
for standard tests.   However,  since it  is used to describe some  tests  reported
in RTECS, it was considered appropriate to use it in HAPPS.
       Another  important  difference  between  the  two procedures  involves  the
absence  of  the severity  subfactor  used  in the  ORNL procedure.   A  substance
is scored by the dose at  which  an  effect is observed and separately  according
to the  severity of  that  effect,   for  example,  severe incapacitation  or  mild
incapacitation.  Then the  overall ORNL score  for chronic toxicity is  obtained
by multiplication of the dose score and the severity score.   HAPPS uses only a
single  score  which reflects  dose but  not severity.   No  standard  list  of
effects categorized by severity was  found  although  typical  effects  in similar
categories  of  severity  are given  in  Ref.  17.   Assignment  of effects  to  the
severity categories would require  expert  judgment and would  probably  vary with
the operator.   Thus, assignment of  severity  rankings was  not  acceptable  for
inclusion  in HAPPS.   Even  if  a  standard  list of  the  relative severity  of
different  effects  were  available,  the  primary  literature   would  need  to  be
consulted  to determine  the nature  of  the effect in a  majority of  the cases
reported.   RTECS distinguishes  between toxic  effects and irritant  effects  and
ranks the  latter according  to  the  degree of irritant severity.   However, only
dose data with  no  indication of the  type of effect,  let  alone the severity of
the effect,  is  given  for  toxic  data.   Such data could only be found  by refer-
encing  the  primary  literature,  a  procedure  precluded by  the  guidelines  for
HAPPS  as being too detailed  at  this preliminary  stage.   In  summary,  all
compounds  have the  potential   to  elicit deliterious health effects  at  some
level;  however,  this  methodology  emphasizes effects  related to carcinogenesis
and  lethal  effects as being most  important to air  programs  at  this  time  and
does not  address  the character or severity of  other effects.   This  procedure
was  considered reasonable given the  preliminary nature  of  ranking  by HAPPS,
the  judgemental nature  of ranking these other  effects,  and the improbability
that such effects would occur at ambient concentrations.   Dropping the  ranking
based  on severity  also  makes  the  schemes  used  in  HAPPS for  the  factors  for
acute  toxicity and  for  other effects consistent in  structure and application.
        Having  noted these  differences,  the HAPPS criteria  for  effects other
than  acute  lethality and the ORNL criteria  for the dose component  of  chronic
toxicity in terrestrial  animals are  compared  in Table 1.6   The ORNL procedure
did  not provide a  scale  for scoring inhalation data given  in ppm.  Since such

-------
                                   31
      Table 1.6  Criteria for Effects Other than Acute Lethality
Item
1
2
3
4
5
6
7

8

9


Primary
Weight
7
6
5
4
3
2
1

0

0


Species
Human
Human
Human
Human
Animal
Animal
Human/
Animal
Human/
Animal

Human/
Animal
HAPPS3
Exposure Route and Dose°
Inhalat ion
Gas Solid6 Oral Dermal
,<1 %<10
a a
1-10 10-100
1-10 1-10
10 MOO >10 >10

Negative or insignificant
evidence.

No data.

ORNLa'c
,d
Score Dose
-
-
-
-
3
-
-
-
-
<1
2 1-10
1

Low or no
biologica
activity.


MO

1



aMost criteria given in shortened form; complete specifications of
 the HAPPS criteria are given in Appendix A.

''Dosage units:  Inhalation - ppmv for gases, mg/m^ for solids
                Oral - mg/kg
                Dermal - mg/kg

CORNL procedure considers only chronic toxicity.
dAdapted from Ref. 21.

eDerived from gas (ppm) scale by using acute lethality line in Fig. 1.

-------
                                      32
data is Che primary concern of air programs, an inhalation scale was developed
for HAPPS.  Reference  to  the  ORNL scales  for chronic toxicity (see Table 1.5)
shows  that at  low and moderate dose  ranges, the  numerical  values  of the
concentrations  defining  the  ranges for ranking inhalation  and  oral exposures
are identical.   For inhalation and dermal  exposureSj  the  defining concentra-
tions  are  identical or  comparable.   The  assumption was made  that  a similar
equivalence was reasonable  for  both chronic  and acute  nonlethal exposures.
The HAPPS  scale for inhalation data for gases  in volumetric  units (ppm) thus
assigned defining  concentrations  in  ppm's  numerically  equal  to  the ORNL doses
(see  items  5,  6,  and  7).   The  corresponding mass  concentrations  for solids
were  taken  from the acute  lethality line   in  Fig.  1 again assuming  that the
correspondence developed  for acute data should be reasonably valid for chronic
data as we 11.
       HAPPS  distinguishes  between  effects  in  humans  and  those  in animals
except at  high  doses  above 10 ppm for  gases  and  100 rag/m^  for  solids for in-
halation exposures  and above  10  mg/kg  for  oral and dermal exposures (item 7).
Extra  weight  is  assigned  to  human data  to  reflect the air programs' charge to
protect human health.   The distinction was  not made at  the high dose levels,
because it  was  felt that human  exposures  at  these high  levels  would probably
be  experienced  only accidentally and even  then probably only in occupational
settings.   Both of these  types of exposures were felt not to be indicative of
the  situation that would be  regulated  by  air programs.   Thus,  human data at
high  doses  was  not considered a  better  indicator of increased  need for  addi-
tional study  than  animal  data  at  high doses.
       At  lower dose  levels  (items 1-6),  HAPPS always weights human data more
heavily than  animal data  whereas  the ORNL  procedure  weights both types of data
equally.   This  additional weight  reflects  air  programs' primary concern with
human  health  but,   as  just  discussed,  was  not assigned  to  human data at high
dose  levels.   It was  recognized  that  information on human toxicity is  rarely
available  for this factor.   However,  where  such  information is available or
where  it  becomes available  in the future  it was  felt  that  the  significance of
human  toxicity  data should be emphasized  by ranking it  above animal toxicity
data  at dose  levels near  those expected in the ambient air.  Similarly,  in the
HAPPS  system, human data at  moderate doses  (items  3 and 4)  is ranked  above
animal data  in  the  lowest  dose   range  (item 5).   This  differs from the ORNL

-------
                                      33
procedure and from the HAPPS assignment of weights for acute lethality in both
of which  the range in which  the  dose  lies is the primary  determinant  of the
weight assigned.  Except where acute lethality is concerned, it was considered
more  appropriate  to weight  evidence  from huraans  more heavily  than  evidence
from  animals at the  low  and  moderate dose  levels.    Such  a  weighting  seemed
reasonable  in  view of the  many different  types  of effects and levels  of
severity  of  effects which were  subsumed  under this  factor.   For example,  if
dose  were  the  primary determinant of weight,  mildly  incapaciating effects  at
low  doses would  be weighted  more  heavily  than  severly  incapacitiating but
nonlethal effects at moderate doses.  Modification of  such an  assignment would
require consideration of  the  severity  of  the effects.  Since  no way to incor-
porate relative  severity  was  available to HAPPS,  it  was  decided to give more
weight  to effects  in humans  regardless  of  the  dose involved  as  being more
valid  indicators of  the  need for  additional study  in  a  program concerned
primarily  with  human  health.   Exactly  as   for  acute  lethality,  additional
weight  was   assigned  for  inhalation effects  in  humans as  being most closely
related to the goals of air programs.
       Since  the ORNL  procedure  did  not distinguish  between  data  based  on
human  exposures  and data based on  animal  exposures,  the  choice of benchmarks
was  somewhat more arbitrary  for  this  factor  than  for those  discussed previ-
ously.   By using items 5, 6,  and  7  as  benchmarks and extrapolating,  the ORNL
weights  were consistently applied  to  animal data at  high,  moderate,  and low
exposure  levels.  With these  benchmarks,  the HAPPS weights are easily obtained
by extrapolation using the 1.25 factor and rounding the results  to  the nearest
integer  for  convenience.   For example,  the weight for item 3, two  steps  above
benchmark  item 5 is 5 (3 x 1.252 « 4.68 a 5).
       Additional  weight  was  not assigned when  a substance satisfied several
criteria  simultaneously.   The reasons  given  for acute lethality  also apply to
this  factor for other  effects.   Finally, & criterion for  no data has  again
been  included  in HAPPS.

       Potential  for  Airborne  Release.   As noted  in  Sec.  1.1,  this factor in
HAPPS contains two subfactors, production volume  and vapor pressure, each of
which either is  or  is associated with  a separate  factor in  the ORNL procedure.
Both  of  these  subfactors  are being  used  as rough  indicators of  potential

-------
                                      34
exposure  and  do not  reflect  other  factors  like the  fraction  of production
emitted  to the atmosphere and  the number  of people  potentially  exposed.
However,   all  attempts  in the  literature  reviewed to  improve  upon the use of
these  particular  subfactors  appeared  to  require  data which  was   generally
unavailable or was available under  the 8(a)-Level-A rule of TSCA  and  thus not
specific  to  exposure via  the  air.   When actually  implemented,  the systems
reviewed frequently relied on  default values for many  of the substances
scored.   Other proposals have been  made such as  using  the  labor intensiveness
or price  per weight  as  indicators of  how valuable a  substance is  and  thus how
well its  release might be controlled.   However,  the  indications were  that the
data  required  for  such  efforts are not readily  available and that  such
approaches suffer from  as much,   if  not  more,  uncertainty than is associated
with figures for production volume.   In addition,  measures  of  market  economics
such as  downward trends  in  production  volume were not  considered  in  the HAPPS
decision  process because  such data  are not readily available and were consi-
dered  too detailed  for  this effort.   Thus, production volume seems  to be the
only simple,  easily  accessible surrogate  for exposure.  In HAPPS, production
volume  information  was  supplemented  by  scoring a  substance  by  its  physical
state:    solid,  liquid,  or gas  and  within  liquids  by vapor  pressure  as an
additional measure  of  the  potential  for  a substance  to be released  into the
atmosphere and  thus  as  an  additional  rough  indicator  of  the quantity poten-
tially released.
       Table  1.7  compares  the  two  sets  of criteria  for  scoring production
volume.   Little need be said to  compare  the  sets  because  the  criteria are
essentially  identical except  for  slight differences in rounding and the
addition  of the criterion for no  data (item 8)  to HAPPS.   The  relative weights
between  various  criteria are  the  same in the  two systems and  the additional
criterion for  no data has  again  been weighted  at the  same  level as  the  lowest
weighted  criterion used when data is available.
       Both  sets  of criteria are  structured   so  that  substances  with  high
production volumes  are  considered more likely  candidates  for  additional  study
than substances with  low production volumes.   Of course,  some substances  like
sucrose  with very high production volumes but with no adverse  effects could be
ranked higher  than  a  substance  with  little  production  but  with   very  high
toxicity.  However,  the reverse situation could occur when the factors related

-------
35











03
0)
c
5
^
0


c
o
"jj
u
3
T3
O
ti
(X

w
o
U-i

Q
• •^
w
0)
u
••*
w
u


r^»
*
«•*

-s
^*-»
jS
»«4

v^
^
^^



U
>^
**x.
&c
^

vC
O




*••
ao

gj
3



E
01
4J
i-"

vC
O
O m

ON ^D
/\ ON
^«





O
en
O cs
m i
sr o
/\ in






O 00
_4





0
O —i
O vn
C 1
— o
/N. O
o
.•^




o
t^! f^)
in CM
•* 1
/\ O
l/N
^





O
-i CO







—1 CM


ON ON
ON \O ON ON ON
ON O C ON
1 in 1 • O 1
VO 1 NO 1
O ON ON V
m ON f^ CT*
m




in in
in co sj 
•H <*4
ac
c
01 0)
> i^
• ^4 CC
ac
m
0) .M
u u
a o>
^j
<8 —
• *4 \*
1)
u M
•-i a.
1- 0.
0 <

^i
tn oi
O J=

-2















































•
^M4
CM

•
1)
3£

£
u
vu

•o
41
jj
a.
«
*c
<^
y

-------
                                      36
co coxicity were scored  and  even  if a substance like sucrose were prioritized
at  a  relatively high  level,  review of  the  list  prior  to  additional  study
should eliminate most such anomalies.
       The  physical  state  of  a  substance  may  affect  the  potential  of that
substance  for  being  released  to  the atmosphere and thus  the  quantity of the
substance  actually  released.   The  subfactor  for vapor pressure  really  looks
at  the  physical state  and,  for  liquids,  the  vapor pressure  of  a substance.
Criteria for vapor pressure  are presented  in  the ORNL procedure as guidelines
for use  in scoring the  level  of  potential occupational  exposure  when actual
exposure concentration data is unavailable.  As modified for use in HAPPS, the
scale for scoring vapor pressure  is  identical to one used in the MITRE scoring
procedure (Ref. 9).  During testing  of HAPPS,  it was found that vapor pressure
data  was  unavailable  for some substances  for  which  boiling  point  data was
available.  A  toiling  point  scale equivalent  to the MITRE boiling point scale
was added  to  HAPPS for  use  when  vapor pressure data was  unavailable.  • Table
1.8 compares the two sets of criteria.   The weights in HAPPS reflect the fact
that gases are generally more difficult to contain than liquids and solids and
hence, other things being equal,  will be  released in greater quantities.  Two
liquids with equal  production volumes will be  emitted  in proportion to their
vapor  pressures  if all  other  factors are  equal.   In  the  ORNL procedure for
scoring  occupational   exposure,   the  type  of  process   is  categorized by  the
degree  of containment:   open,  controlled  release,  or enclosed.    Such con-
siderations would also be relevant  to determining the quantity  released  to the
ambient  air but the  required data  would  not  normally be  available  and such
determinations would probably  require expert knowledge  of the specific proces-
ses  involved.   Such considerations  would be  unsuitable for HAPPS and are too
detailed  for a preliminary  prioritization.   Gases were ranked  above all other
forms  of  substances  in HAPPS  as  being  the  most difficult to contain and thus
the most  likely to be  released in large quantities.  Solids  are weighted equal
in  importance  to highly  volatile liquids,  not  at  the  lowest  level of  impor-
tance  where their very   low vapor  pressures  would  place  them.  However, the
factor of  true concern is  the  quantity of material  released  and vapor  pressure
is  not the  only consideration.  The  ranking for  solids  in HAPPS was based upon
consideration  of  the  importance  of  solid particulate matter air pollution and
EPA's  assessment  of how  important  they  considered  solids compared to various

-------
37























CO
11
u
3
co

u
ex

^4
0
o.
CO
"*
M
o
U-l

CO
• •4
i.
4>
4J
• *4
u
c^.5


00
•
•H

u
2
CO
f4







































•O
.a"
j
*y
ex
0




















/— s
oo

I
a.
^

A
•a
3
cy
'•J











































O
M
to

a.

*
















4

•

c


•"• H
^
1
CN
^.
CN
1
O
0
"~*

o
o
••4
/\



-o

»H
o
to



CO
CO
u



01

0
o
to


.J
2
H
0
U •— N
u
50«
C ^•w'
•



c

















H
mt
^
0
Q













U
kl
3
a
go /-N


c


















\





U £0
4 """
IH E

^ 'SM'
C

-3
>



jj
A
5C
• -w
4«
II XI

II XI



1 1 X 1




1 1 X 1







1 1 1






1 1 1






1 1 O CO vo — i





e
O
o «
« -. 0
o v o
CO Ol -H 1
v « /\
a, v a,
CQ , &
o
CO




•
CO
"O J-J
M •-! CO
CO — ' "O
0 0
CO O
z






0
0 0
o — i sr
-1 V CN
XN ex v
a. > a.
> V >
-3-
CN








*^ f*") O CN ™* 1 — .

2
!


S
£ 1
^
— . JN t* ^ u-i vo r»
—



c
• H
C
Ol
>
• f^
ao
Ol
m
CO
• *4
t^
01
iJ Ol
• H JJ
M 3
to o u
OH -H
Cu tO 4J
< (X Ul
=: o- J2
< 3
C S W
• H
a; cij
1) J2
03 OJ CQ
CQ CQ
OJ U-l
-* O "3
0) U
L* CO CO
C 3
01 O
C •-! CO
O co
-2 O u
k- •-! C
.«H I4_t >^
CO "^ O
o a.
U Ol
o a. so
"*-l <0 C
•1-4
_4 4) —1
CO iJ -H
•••* 4J O
ij -H .0
e a
US Ol
tj O »C
O U -u
O.
. * M
? t 2
•HO .0
u y-i co
O — •

C CU CO
•-< .u e
P 3
T3 O
 01
O •* M >"
Ci. 60 3 U
S ec e£
C «J eo
cj — i Oi E
u . k c
o a < a. w
3 w >n
4J -H X 1-
W ••* C "3
t. O T) C. H
O C ^3 w
u U > 6.
ti a a. a
e C e. "-I -s
c -z. < — <
to £i y TJ

-------
                                      38
liquids  in  terms of  the  quantity released,  other  things being  equal.   With
this ordering of  the  criteria,  the  actual  weights were assigned.  Once again,
a  scaling  factor of  1/3  was  used  in HAPPS.   The  identical  criteria corre-
sponding to items 3 and 4 were used as benchmarks to the HAPPS weight  for item
3 was 3 (10/3 - 3.3 « 3) and the weight for item 4 was 2 (8/3 = 2.7 «2).  For
item 4,  the  weight  was truncated to 2 rather  than  rounded  to  3 to keep three
of the criteria  (items  2,  3,  and 4)  from receiving equal weights.  The weight
for solids (item 2) also received a weight of 3 equal to the weight for highly
volatile liquids as discussed in the previous paragraph.  The weight for gases
was obtained by  increasing  the  weight  for  highly volatile liquids by  the same
factor as ORNL used  to  go  between moderate and high volatility liquids.  This
factor can be found from
       (ORNL Score for High Volatility Liquids) = (ORNL Factor) x (ORNL Score
                                                  for Moderate Volatility
                                                  Liquids)
         or 10 =  (ORNL Factor) x 8 and
                  (ORNL Factor) = 10/8.
The  HAPPS  weight can be  calculated  taking  the  scaling  factor into  account:
       (HAPPS Weight  for Gases) = (ORNL Score for High Volatility Liquids)
                                x (ORNL Factor) x (Scaling Factor)
                                - 10 x (10/8) x (1/3) - 4.16 * 4.
As  had  been  done for  several  other factors,  the  weight  for  low volatility
liquids  which  corresponded  to two separate ORNL  criteria (items 5 and 6) was
taken as  the  geometric  mean of the corresponding ORNL criteria.  Thus, item  5
received a HAPPS  weight of  1:
       (Weight for item 5 in HAPPS) = (Geometric Mean of ORNL Scores for
                                      items 5 and 6) x  (Scaling Factor)
                                      = VTTl x  (1/3) * 0.87 •"* 1.
As  was  the  case  for  previous  factors,  the criterion for no data was  weighted
at  the lowest  level of  importance.
       The  principal  difference  between  HAPPS and the ORNL procedures  is  that
the  scores  for production volume  and vapor pressure  are multiplied  together  in
HAPPS  to obtain  a  score for  the  potential  for  airborne  release rather  than
being  used  separately  as  individual  scores  for  separate  factors.   Combining

-------
                                      39
the subfactors provided a convenient way of placing the source-related surro-
gates  for  exposure into  a  single  factor.    As  discussed in  Sec.  1.4,  this
source-related factor  is  combined  with  the  receptor-related  bioaccumulation
factor in scoring  the  exposure  group.   The individual scores  were multiplied
rather than added  for  several  reasons.16   First,  at the factor level, it was
deemed desirable  to avoid the problem of weighting  the individual subfactors.
Subjective weighting decisions  are  always  required when  scores  are  added in
procedures like HAPPS.   (The interfactor weightings used  in HAPPS are discus-
sed in Sec. 1.4.)  Second,  use  of the  multiplicative  method normally  provides
a wider  range of  scores than  does addition.   The  wider range, even  when
normalized,  tends to avoid equally weighted  substances.

       Bioaccumulation.   The  criteria for  bioaccumulation  are  compared  in
Table 1.9.  No summary  of bioconcentration  factors  seem  to be available so the
criteria  based on this parameter were not included  in HAPPS.  Instead, the
criteria were based  on  the octanol/water partition  coefficient  which is
related  to the tendency of  a substance  to  accumulate  in  fat  rather than  water
and  hence  to  accumulate  in .-tniirais.   The  criteria were  based upon   the  fact
that  higher  values  of the cctanol/water  partition coefficient generally
correspond to  a  substance with  a greater tendency  to dissolve in and accumu-
late  in  fat.   There are  so1 e  exceptions  to  this,   particularly for values of

                  Table 1.9  Criteria for Bioaccumulation
HAPPSa'b
Item
1
2
3
4
5
Weight
10
8
6
1
0
LoglOp
>6
6-4
4-2
s<2
No data
Score
10
8
6
1
-
ORNLa > c
Bioconcentracion
Factor L
>4000
4000-1000
1000-300
<300
-

°£io?
>6
6-4
4-2
<2
-
         aP  is  the octanol/water partition coefficient.
         ^Criteria given  in shortened form, complete specifications
          of the HAPPS criteria are given  in Appendix A.
         cAdapted from Ref. 21.

-------
                                      40
        greater  than  6  but  the  data  to  correct  for  such exceptions  is  not
readily  available.    HAPPS  uses  the  same  criteria  and  weights  as  the  ORNL
procedure uses for items 1-4 and, as with previously discussed factors and for
similar reasons, adds a  criterion  for  no  data at the lowest importance level.
       Existing Standards.   Table  1.10  compares  the  criteria used  in HAPPS
with those used  in  the  MITRE  system^  upon which this HAPPS factor was based;
the  factor  for existing  standards  was not used  in  the ORNL  procedure.   The
criteria  for gases  are the  same  except  at  the low  end  (item 1)  where the
criterion  for ranking  carcinogens  was  eliminated  from HAPPS.    Since HAPPS
assigns weight for carcinogenic activity under the factor for oncogenicity, to
include it  again in the factor  for  existing  standards could have resulted in
double counting.
       A new sec  of criteria for solids  was  introduced in HAPPS because both
solids  and  gases are air  contaminants.    None  of the  systems  reviewed had  a
model  set  of criteria  for  scoring  standards for  solids.   The HAPPS criteria
for solids were developed by examination of the current OSHA standards.  As an
initial attempt,  the  range of  OSHA  standards  for  solids  was divided  in  a
manner  proportional to  the  division of  the  OSHA standards for  gases  by the
MITRE criteria eliminating one substance with an extremely small time-weighted

                 Table  1.10  Criteria  for  Existing Standards
Item
1
2
3
4
5
6
7

Weight
6
5
4
3
2
1
0
HAPPSa
Gas
(ppm)
^5
5-10
10-25
25-100
100-200
>200


Solid
(mg/np)
$.25
.25-. 5
.5-1
1-5
5-10
>10
No standard

Score
5
4
3
2
1
0
-
MITRE3. b
Gas
(ppm)
<5 or carcinogen
5-10
10-25
25-100
100-200
>200
-
        aMost  criteria  given in shortened  fora;  complete specifications
         of the  HAPPS criteria are  given in Appendix  A.
        bAdapted from Ref.  10.

-------
                                      41
average  (TWA)  from the  process.   The  criteria  so construted  ran  from <.01
mg/m^  for item  1  to >.4  mg/m^ for  item 6.   However,  they were  found to
deemphasize the  importance of  solids  with  respect  to gases.  Very few solids
had  standards  in the  low  end  of  the  scale  so  constructed;  only four solid
substances would have  received  weights of  four  or  more  given  the current
standards while many gases would receive  such weights.   In order  not to
deemphasize solids with  respect  to gases,  these  initial  criteria for solids
were  redefined  in a  sequence of  steps.  In  the first  step, the  range of
current  standards  less than  1  mg/m^ was divided into  four ranges  corre-
sponding to items 1-4 in the table.   The division  was done  in  such a way that
approximately equal numbers  of standards fell into  each  of the  four ranges.
For  standards  above  5 mg/aH,  a  different  approach was used.   In this first
step, the  second concentration class  (corresponding  to  item 2 in Table 1.10)
covered a  concentration  range  whose  width  from its lower concentration limit
of  0.1  mg/m^  co  £ts  Upper concentration limit  of 0.5  mg/ta^ corresponded to
a  factor  of 5.   This  initial division point  between the  first and second
concentration classes is shown as  the open circle near the  line  for existing
standards  in Fig.  1.   Similarly,  the  width  factors  corresponding to items 3
and 4 were  2  and 5,  respectively.   These width  factors suggested a  repeating
sequence of 5,2,5 as  one moved from  item 2 to  item 3  to  item 4.   It  should be
noted  that  this 5,2,5  sequence  of width  factors  is destroyed by  the final
adjustment of the boundary between the  first and second concentration classes
from  0.1  mg/m^  upward to  0.25 mg/m-^  as described below.   A  tentative width
factor  for item 5 was chosen to  be 2, the next factor suggested by this
series.  Then,
       (Upper Bound for item 5) * (Lower Bound  for  item 5) x (Width Factor)
                                " (Upper Bound  for  item 4) x (Width Factor)
                                * 5 x 2 * 10
and item 6 would then correspond to any standards exceeding the  upper  bound  of
10 mg/m-' established  for  item  5.   The  corresponding  values for  gas  concentra-
tions  in ppm's  and  for  solid  concentrations  in mg/m^  were  then plotted  on
log-log graph paper (see  the line for existing standards in Fig.  1).   As  noted
in  the  discussion of  the factor for  acute  lethality,  three  of  these  points
were  found  to  lie  on  a  straight line parallel  to the line already established
for acute  lethality.   As a last step,  this  standards  line  was  used to  adjust

-------
                                      42

Che boundary  between  items  1  and 2 upward from 0.1 to 0.25  mg/m^  so  that  the
adjusted point would lie on the line and correspond to 5  ppmv on the MITRE  and
HAPPS  gas  scales.   Similarly,  the  boundary between  items  3  and 4  (corre-
sponding to  10 ppmv)  could  also have  been  adjusted upward  from  1.0  to  1.2
mg/ar.   Such  a small  change  was not considered worth making  considering  the
preliminary  nature of the use of HAPPS  and the  heuristic  method  used  to
develop the scale for solids.
       The weights  used in the  MITRE  system  were  modified  slightly for HAPPS.
Even at the highest concentration levels (item 6),  the existence of a  standard
was  felt  to  indicate  a positive  finding  of  adverse human health  impact.   It
was desired that the existence of no standard should receive  even less weight.
The  criterion for no  standard  was  given  zero weight and the  weights  for  all
the other criteria were incremented by one over the corresponding MITRE scores
to avoid having any two HAPPS criteria having identical  weights.

1.4  FACTOR GROUPS AND WEIGHTS

       1.4.1  General
       After  scoring  a substance in  each  of the  factors, HAPPS  departs sub-
stantially  from the ORNL  procedure although  there  are  still  some points  of
similarity such as the grouping  together  of related factors.   These differ-
ences  arise  mostly out of  the differences in the  scopes and  purposes  of  the
two  systems.   In  view  of  these differences,  HAPPS will  not be compared to  the
ORNL procedure  in discussing the groups of factors nor in discussing the final
ranking procedure.
       HAPPS  ranks substances  by  proceeding through three  levels, beginning
with  the most detailed  level and aggregating at successive levels to provide a
final  single  rank for  each  substance.   The first, most detailed level, scores
substances in each  of  the eight  factors chosen as described above in Sec. 1.2.
These  scores  are  chosen by comparing  the  available data against the  criteria
described  in  Sec.  1.3.  At the second  stage,  the  scores in  certain groups of
closely  related  factors  are   combined  to give  group scores.    Section 1.4.2
describes  the groups and  the relative weights of  the  factors  in  them.

-------
43






















c
0

•u

N

»»«
CO
£
u
o
z

U-4
0

^,
^J
"•^
5
—
*>H
*
—rf

11
•o
4J da O
U 3 W
0) O O
IJ 1. 
wH '-^ «~1 — H




U"\ W-l
• . in O
— < -H CN «N






ii a ii u





•H -^ ^^ «<





X X X X




O >O
• O 0
— < O CN -H


+ * +

— 1
60
•ri
CU



X



o;

0
CJ
C/5



41
r^ 1—4 «4 «H





X X X X




u-i O
u-i O
O — t -H





OS i '<-> 1—4
(5 I — , 1— 1 M 1— 1
C^J






! 1 "
T3
1
UM
! O 03
^J

3 (V
"C W
i, •*•»
N —
!!!_'•- 3
a. = — S
1 .x u a u

!r- v; = o
^ C




















g
cs

^1
0
JJ
u
flj
u.

^4
o
U-l

o
CN
1
0

•o
c
09
0
o
«J
u.

^
0

o
^H
1
0

U-l
0

ffl
01
u
o
o
CD

IUI
o

01
eo
c
eg
1.

a
01
c
3
X
en


























































•
^
01

• nH
3
c-
O!
)»4

C
0

ij
u
01
u
^
c
u

c














•
4J
X
01
c
• H
c
o
• ••4
to
CO
3
O
to
• 1-d
•o

01
01
CO

• *
CN

M
O
u
u
0

-------
                                      44
Finally, the  group scores are  combined  Co  give the overall  rank  of the sub-
stance  as  described in  Sec.  1.5,  the  overall ranks  of  different substances
giving them a numerical prioritization.
       At  each  stage  or  level,  HAPPS  normalizes  the score  or   rank  to  the
maximum value  that could  be  obtained.   Thus,  the  maximum score  or rank
at any  level  will  be one.   This normalization procedure was adopted primarily
to aid in the assignment of the interfactor and intergroup weights required in
the second  and  third  levels of the  procedure.   The relative score or rank of
different  substances  remains  the same  before and  after  normalization so  the
important  information   is  unaltered.    However,   the  assignment   of  relative
weights  becomes much  easier  when  the  factors  being  combined have  all been
normalized  to one  (or  to  some  other single  value).  For example,   suppose that
two closely related factors were being grouped and  that it was desired to give
each  factor equal  weight  in  prioritizing  substances, that  is,  both factors
were  considered as equally  important indicators  of  the  need  for additional
study in relation  to the group.  Table 1.11 illustrates the utility of normal-
ization  in this  situation.    A substance  satisfying  the  midrange  criterion
for the  first  factor and  the  highest priority criterion for  the second  factor
(Case  I) should  be ranked  the same  as  a  substance  satisfying  the highest
priority  criterion for the  first  factor and the  midrange  criterion for  the
second  factor (Case II).   The table illustrates  the  problem avoided by nor-
malization.    This illustration  assumes  that  equal  weights  (=1)  have been
assigned  to each  factor.   When normalized  factor scores  are used, a  score of
1.0 corresponds to the highest criterion and a score  of about 0.5 corresponds
to the midrange criterion for  both  factors.   In  the two cases  considered, both
substances  would   receive  the   same  group  score  (= 1.5)  and hence  the same
priority  based on this group  alone just  as  they should.    However, if the
scores  for  the two factors range over  different sets  of values,   say  0-10 for
the first  factor  and 0-20 for  the second factor,  then the unnormalized  scores
corresponding  to  the highest  and  the midrange criteria depend upon  the  factor
being  considered  with  the result that  the  two substances are no  longer given
equal  group  scores.    This  situation  could  be  corrected  in the example by
assigning  a weight of  0.5  to  Factor 2 but  then unequal weights  would  corre-
spond  to equal  rankings of importance,  obscuring  for  someone  interpreting the
results  of  a priorization the relative importance  assumed  for  factors and

-------
                                      45
groups.   Normalization thus  aids  in  assigning  the relative  interfactor  and
intergroup weights;  elements considered  equally important  could  be assigned
equal weights when normalized scores were used without bothering to adjust the
weights for differences in the scales of individual elements.  Of course, both
normalized and  unnormalized  weights  would  provide  the  same  ranked  list  of
substances if the  unnormalized  group  scores  were corrected appropriately,  but
normalization was used in HAPPS to aid the clarity of presentation and to make
the assignment of weights as simple as possible.

       1.4.2  Groups
       Certain  factors  are  either closely  related  or are  surrogates  for the
same effect of  real  interest.   For example,  oncogenicity and mutagenicity are
closely  related and  potential  for airborne  release  and  bioaccumulation  are
surrogates for  human  exposure.   Scores for each factors are combined together
into group scores  prior  to  final  scoring of a substance.  This procedure is a
matter of  convenience only.   The  same  rankings  could  be obtained by applying
an  appropriately  chosen  set  of  weights  to  the  eight  factors  individually
without  going  through the intermediate  step of scoring  within  groups.   How-
ever,  use  of  groups makes it  easier  to assign relative weights both within a
particular group and  between different groups.
       Within a  single  group,  only a few factors need to be considered  at one
time.   The number of decisions  as to the relative weights  to be assigned to
various  factors is  thus  reduced  to  a  level  where  the  process  becomes more
managable.   In addition, the  judgments required are  between  related factors
like  oncogenicity  and mutagenicity,  a  situation  in  which, for  example,  the
decision  to  weight them both equally  or to  give one  factor five or ten times
the  weight  of  the  other  is  relatively  easy compared  to a situation in which
relative weights must be  assigned  to disparate elements  like  oncogenicity and
production volume.  Of course,  the use  of  groups only postpones the  assignment
of weight to disparate elements until  the  intergroup weights must be assigned,
but  because  of  the  grouping  of  similar  factors  into  groups,  the  number of
decisions  to be  made has  been reduced and  the  comparisons  to  be made are
between  the  dissimilar  groups.   Problems encountered when  considering similar
and  dissimilar  elements at the  same  time are avoided because the  similar items
have been combined  in the groups.

-------
                                      46
       In  proceeding  in this  fashion,  HAPPS  assumes that  linear  expressions
are acceptable  for prioritizating  substances,  that is,  there  are no  inter-
actions between  factors  and/or groups.   For  example,  HAPPS assumes  that  the
oncogenicity  scores   for  two  substances  satisfying  the  same  criterion  for
oncogenicity should be  the  same  even if one  substance  is  toxic  at  low  levels
and produced  in  high volumes  while  the other has  substantial evidence  of no
toxic potential and is produced  in  very small volumes.   In fact,  interactions
between factors may be very important in making decisions and are neglected in
HAPPS.   Linear  systems  like  HAPPS  cannot  account  for  interactions between
factors or groups;  the weights  simply  tell how important an element is com-
pared  to   other  elements  for  a  specific  set of scores  for these  elements.
Thus,  HAPPS is  an  approximation to  an ideal  approach but an  approximation
which  is  reasonable  given the purpose  of the system and  the available  data.
       The  groups  of factors  actually  used  in HAPPS are  shown  in  Table 1.12
along  with the relative  weight  of  each  factor within  a  group.   Two of  the
groups (items  2 and 5  in  the table)  contain only a  single factor and need not
be  discussed  further.   The  remaining three groups  are  discussed below.

       Carcinogenicity Group.  As  discussed  in  connection with  the oncogeni-
city and mutagenicity factors  in Sec. 1.2, carcinogenic potential provides the

                        Table  1.12  Groups of Factors
Item
1
Group
Carcinogenicity
Factor
Name
Oncogenicity
Mutagenicity

Weight
1
1/4.40
                   Reproductive and   Reproductive and       -
                   Developmental      Developmental
                   Toxicity           Toxicity
                   Toxicity           Acute Lethality        1
                                      Effects Other than     1
                                      Acute Lethality
                   Exposure           Potential  for Air-    10
                                      borne Release
                                      Bioaccumulation        1
                   Standards          Existing Standards

-------
                                      47
basis for EPA's concern with oncogenicity and mutagenicity.  Higher scores for
the two factors separately  are  intended  to  reflect  increased concern that the
substance being  scored  is  a  human  carcinogen acting  through  inhalation.   It
was thus reasonable to combine these two closely related factors into a single
group.   Since  not all  mutagens are  carcinogens,   somewhat less  weight  was
attached to evidence based on mutagenicity than to evidence based on oncogeni-
city  even  though  some  evidence for oncogenicity  could  relate  to noncancerous
tumors.   Put  another  way,  the strongest evidence  for  mutagenicity was consi-
dered to be a  less  reliable indicator  of a substance's carcinogenic potential
than  the strongest evidence for oncogenicity.  For both factors, the strongest
evidence corresponds  to  a  normalized  score  of  1.0.   For  onccgenicity,  for
example, a  normalized  score of 1.0 would require  evidence of oncogenicity in
humans  by  both inhalation  and  noninhalation  routes, evidence  in  two  or more
animal  species, and scheduling  for  carcinogenesis  testing  under the NTP.  The
weighting factor  of 1/4.40  ( = 0.23)  was  chosen so that the  strongest evidence
from  mutagenicity (normalized score =  1.0)  would receive  less  weight in the
group than  evidence  of noncogenicity in one  animal  species (normalized score
» 2/6.25 * 0.32) but more weight than if the only evidence of oncogenicity was
scheduling  for testing under  the NTP (normalized score = 1/6.25 » 0.16).  (The
criteria and  scores can  be  checked  by  reference to  the  tables  in  Sec 1.3;
normalization  factors  can  be checked by  reference  to the   coring sheets in
Appendix A.)  Any  factor between 0.16 and 0.32 could have  been chosen; the one
selected was  0.23 (*  »0.16  x 0.32),  the geometric mean of  the  two scores of
interest.   The geometric  mean was  used to  keep  the  relative  ratios of the
weights the same,  since it  is  the ratio  of weights,  not the  difference between
them  that  is  the  measure  of their relative importance.  (Values quoted  in the
text  ha.ve  been rounded for presentation and a check  of  the ratios using the
text  values will  show a slight inequality.)

        Toxicity Group.    Two  of the  factors   rank  substances  by  their toxic
effects:   acute  lethality  and effects  other  than acute  lethality (item 3 in
Table 1.12).   As  discussed  previously, both of these  factors deal  with  tradi-
tional  toxicological  data  except  for the  effects  of  special  interest dealt.
with  under  oncogenicity,  mutagenicity, and  reproductive  and developmental
toxicity.   Generally  speaking, the acute  lethality  factor will score data on

-------
                                      48
acute exposures while Che factor for other effects will  score  data  on  chronic
exposures  although  nonlethal acute  effects  would  also  be  scored  under  the
latter factor.  Even  with  this  potential mixing of acute and  chronic  effects
in the  factor for  other effects,  it  is  reasonable  to group both factors
together into  a toxicity group  which summarize the need for concern based  on
data  from  standard  toxicological  tests.   Were  some  of the  data  from  other
types of tests, say  epidemiological  studies of human  populations,  the grouping
is still  sensible,  as  both factors measure  the degree of  concern based  on
evidence of toxic  effects in humans  or other  species.
       Equal  weights were  assigned to both acute  lethality and to other
effects in  deriving  an  overall score  for  the toxicity group.   Since most  of
the  concentrations  used in  the experiments  which provide  the  data used  in
scoring the  two  constituent  factors  exceed   the concentrations  likely to  be
encountered in the ambient  air,  the  weights were chosen based on consideration
of the  lowest  concentration  ranges  scored.  Furthermore, since the principal
interest of Air Programs is  in  effects  on  human health caused  by  air contami-
nants,  the considerations   were restricted to  inhalation  effects  in  humans.
From  the viewpoint  of the  need  for  additional  assessment, it  was felt  that  a
substance  producing acute,  lethal  effects  in humans at  very  low  doses  should
be of  as much concern as a  substance  producing other  effects  (probably chro-
nic)  in  humans at  very low doses,  that  is,  that  the two   factors  should  be
equally weighted.   Use  of  weights of one  accomplishes  this  goal.   Inspection
of Table 1.5  shows  that  the  normalized  score  for acute lethality  in humans in
the  very  low dose range is  1.0 (=  4.7/4.7)  for inhalation.   Likewise,  Table
1.6  shows  that the  normalized  score  for other effects  in humans in the  very
low  dose  range is also 1.0  (=  7.3/7.3)  for  inhalation.   Thus, the choice  of
equal weighting factors (=1) does provide  the intended equal relative  weights
to the two factors for the species,  concentration range,  and  exposure route of
greatest interest in air programs.

       Exposure Group.   The factors measuring  potential  for airborne  release
and  bioaccumulation  are surrogates  for human  exposure.   As  already  pointed
out,  they  leave much  to be  desired  in  terms  of being  reliable indicators but
within  the  constraints  placed  on HAPPs, they  appeared  to be the  only  reason-
ably  available indicators;  detailed exposure analyses must await  further study

-------
                                      49
in  cases  where  Che  need for  concern  indicated  in  prioritizing by  HAPPS  is
substantiated  by  additional  preliminary  information  and  expert  judgment.
       In weighting  the  two  factors in the  exposure  group,  the potential for
airborne release  was considered to  be  more indicative of the  need  for addi-
tional study  than the potential to bioaccumulate.    Airborne  release is more
directly related  to  the  charge  of  air  programs;  a substance  which accumulated
in humans but  whose  exposure medium was  drinking water would  not come within
the purview of air programs.   Specifically,  a reasonable weighting scheme was
considered  to  be  one which  would  rank  a substance  with  a high ability  to
bioaccumulate approximately equivalent, in terms of public exposure potential,
to  a  liquid substance  with  a moderate  production volume and a low  vapor
pressure.  It was also felt reasonable to consider as approximately equivalent
a  substance with a high ability to  bioaccumulate and a gas  with a relatively
low production volume.   In both cases, a high potential to bioaccumulate would
correspond  to  a  normalized score  of one  (see  item  1  in Table  1.9).   In the
first case,  the  normalized  score for potential  for  airborne release  would be
0.10  (»  4  x 1/40) where moderate  production volume  has been chosen  as item 4
in  Table  1.7  and  a  liquid  with low vapor pressure   corresponds  to  item 5  in
Table 1.8.   To  make  this  normalized score  weigh equally  with the normalized
score of  1.0  for bioaccumulation would require  multiplication  by a  weighting
factor of  10.0 (10.0 x  0.10  =  1).   Similarly,  the  normalized  score  for air-
borne release  in the second case would be 0.10 (=  1  x 4/40)  where  item 7 in
Table 1.7 corresponds  to low production  volume and  item 1 in Table 1.8 corre-
sponds,  to gases.   This  score  would  also  need to  be multiplied by  10.0  to
receive  equal  weight  with  the  score  of  the high  potential  bioaccumulator.
Although the use  of  two  independent,  reasonable ways of determining a weight-
ing factor would  not ordinarily result in  equal  estimates of that factor, such
was the case here and 10.0 was  taken as the  weighting  factor for  the potential
for airborne release  factor in  the  exposure  group.

1.5   Intergroup Weights
       The  final step in the  HAPPS procedure  combines the  normalized  scores
for  the  five groups  into  an overall score  or  rank.   This combining  requires
that  the groups  be weighted to  indicate their  relative  importance.  As pointed
out previously,  this  task  was difficult  because it  required that decisions be
made  as  to  the relative  importance of dissimilar elements.  Different  indivi-
duals could reasonably be expected  to disagree on  the relative  importance cr

-------
                                      50
weight to be assigned to  a.  particular  group.   Procedures  like  decision  analy-
sis capable of assisting in such tasks  and of ensuring  internal  consistency do
exist but could  not  be  applied to HAPPS  within available  resources.  Rather,
Air Programs  developed  a set  of  weights  believed to approximate fairly well
the  importance  given  to  various groups in  the past when decisions were
required as to whether or not further assessment  of a substance  was  warranted.
A  sensitivity  analysis  on the weights  indicated  that  shifts in the  priority
levels of substances were small for practical purposes.  The  remainder of this
section discusses this process in greater  depth.
       Table 1.13 lists  the relative  weights  of the five groups.  Three  of  the
groups  (items  1, 2,  and  3)  deal  directly with data related  to health.   Of
these three, toxicity was considered  to be the least  important  because most of
the  concentrations  used  in  developing  the   data  for toxicity  would  exceed
ambient levels.  Although the same is probably true of  the  concentrations used
in developing  the  data  for  the  carcinogenicity  and the reproductive and  the
developmental  toxicity  groups,  the  effects  considered  in  these   latter  two
groups  probably exhibit no  thresholds;  for carcinogens  this  lack of  a
threshold is  almost  certainly true.   It  was  still  felt, however, that  tradi-
tional  toxicity  data  by itself should  provide sufficient  justification for  a
closer  look at  a  compound.    Making  carcinogenicity twice  as important as
toxicity was felt to reasonably balance these two  considerations. Even  though
carcinogens are  currently  a  major  concern within  EPA,  reproductive  and  devel-
opmental  toxicity  was  weighted  as being   of  the  same importance as  carcino-
genicity, because  both  groups deal  with  severe  health effects  that  may well
occur at ambient levels of concentration.

                        Table  1.13  Intergroup Weights
Item
1
2
3
4
5
Group
Carcinogenicity
Reproductive and Develop-
mental Toxicity
Toxicity
Exposure
Standards
Weight
2
2
1
5
0.5

-------
                                      51
       The exposure group (item 4)  was assigned a weight  of 5  to  make  it  equal
in importance to  the three  primary health effects  groups  together.  EPA
considers  both  exposure and  health  effects  in  making  regulatory  decisions.
For example,  it  is  unlikely that even a  potent  carcinogen  would  be regulated
unless there were significant exposure via the ambient  air.   Conversely,  it  is
unlikely  that  a  widely  distributed  substance exposing  many  people would  be
regulated  in the absence of  severe  health effects.   A  substance  ranked at  the
top of  all three health-related groups  (the normalized  score for  each  group
*  1)  would receive  an accumulated  unnormalized  score of  5  for these  three
groups with the  weights  shown in Table 1.13  (2x1  +2x1 + 1x1* 5).   A
substance  ranked at  the  top of the  exposure  group would receive  an unnormal-
ized  score of 1 for group  before  the  assignment  of the relative weight.
Choosing  a weight  of 5 would make  the  score for the  exposure group  equal  in
importance to that of  the  three  health-related groups  together.   Although  the
choice of  weights made  in  HAPPS  clearly  cannot  reflect all the nuances invol-
ved in  considering health effects  and  exposure  for regulatory  purposes,  the
equal weights given  to  the  exposure  group and the three  health effects groups
together  w2re considered to  provide  a reasonable  approximation to the type  of
thinking  cfone  in the past, particularly  in  terms of determining  the  order  in
which sub. tances should receive additional study.
       The  assignment  of  tha  weight to  the  standards group  was  perhaps  the
most  arbitrary  assignment of  weights.    This  group  was considered important
because  it i-dicated a past concern  with  human  health.  However,  the concen-
trations  invcJved  and  the  exposure conditions assumed in  setting  these  stan-
dards are  significantly different  from those experienced  in  exposure  via  the
ambient  air.   It was fell  that  a similar degree of concern might be appropri-
ate for  tvc substances one of which had a  top score of one  in either the acute
lethality  or  other effects factors  and  no data  (score = 0)  in  the other  and
the second of which had a  top  score  of  one  in the  existing standards factor.
In this  situation,  the toxicity group would  have a normalized score of 0.5 (=
[1 xl +  1 x 0]/2) and  the standards  group would have  a normalized  score of 1.
To make  the standards group equal  in importance  to  the  toxicity group in  the
final ranking thus required the assignment of a weight of 0.5  to the standards
group.   Again,   it  should  be emphasized   that both the toxicity  group and  the
standards group  are  given low weights  because  the concentrations  needed  to

-------
                                    52






elicit the responses dealt with under these groups are generally higher




than those encountered under ambient conditions.




     Since there was subjectivity in the assignment of the intergroup




weights, a sensitivity analysis was conducted using different values for




the intergroup weights.  The analysis indicates that the assignment of




intergroup weights is not all that critical in determining the rank of a




substance within reasonable bounds.  Thus, given Air Programs' own




uncertainty about what weights are best, the overall rank of a substance




is sufficiently accurate for their purposes.

-------
                                     53
                        2  PRIORITIZATION METHODOLOGY

       The  first  task  that  must be  completed  in  a HAPPS  analysis is  daca
collection.  All the  information necessary  to  identify  the  criteria  satisfied
for each  of eight  factors must  be  assembled.    Seven basic  reference  sources
are needed  to  complete a HAPPS  analysis.   Toxicological information,  stand-
ards,  and  some  physical  and  structural properties are  obtained  from RTECS.^6
The status  of  a compound within  the  National Toxicological Program (NTP)  is
available  from  the NTP  Carcinogenesis Testing  Program list of  Chemicals  on
Standard Protocol*? or NTPs  Annual  Plan.28  Production volume data is obtained
from  the  SRI  Chemical Economies Handbook.^9    The  state  of  matter  (solid,
liquid,   or gas) for  the chemical  is obtained  from The Merek Index^O  while
vapor pressure  information   is  from  the  Handbook of Chemistry  and Physics,31
and octonal/water partition  coefficients  are from Let et al.^2
       The  procedure  for completion  of 'a  HAPPS analysis  is  presented  in a
worksheet  format  that  leads the  analyst  step-by-step from data  collection
through  prioritization  of any  substance.   Worksheets  1-9  are  used  for data
collection.  Each  worksheet  explains  what information  is needed  and where  to
find  it.   Worksheets  10-17  are used  in conjunction  with  Tables  1-9  and Work-
sheets  1-9 to  assign a  normalized  weight for  each of the eight factors.
Worksheet  18 helps the  analyst  combine the  eight factors  into  five groups and
calculate  normalized  group weights.   The  final prioritization  is  accomplished
using Worksheet 19.
       To  facilitate their use apart from this document, Tables  and Worksheets
used  in the HAPPS analysis are presented  in the Appendix and are numbered from
1-9 and 1-19, respectively.   Abbreviations used in RTECS also are presented  in
the Appendix in Tables A-l to A-3.

-------
                                     54
                                 REFERENCES
 1.   Lewis,  R.J.,  and  R.L.  Tatken,  eds.,  1979 Ed., Registry of Toxic Effects of
     Chemical Substances, VoLs.  I  and  II,  U.S. Dept. of Health and Human
     Services,  National  Institute  for  Occupational Safety  and Health,
     Cincinnati, Oh.  (Sept.  1980).

 2.   As till,  B.D.,  at  al., Sequential  Testing for Chemical Risk Assessment,
     Health,  Safety,  and Human Factors Laboratory, Eastman Kodak Company.

 3.   Babcock, L.R., Jr., and N.L.  Nagda, Popes - Ranking Air Pollution
     Sources by Population Exposure, EPA-600/2-76-063  (1976).

 4.   Cleland, J.G.,  and  G.L.  Kingsbury, Multimedia Environmental Coals for
     Environmental Assessment, Vol.  I,  U.S.  Environmental  Protection Agency
     Report  No. EPA-600/7-77-136a,  Research  Triangle Park, N.C. (Nov. 1977).

 5.   Council  on Environmental Quality.  TSCA Interagency Testing Committee.
     FR 42 No.  197  (Oct. 12,  1977).

 6.   Carroll, J.W., Formulation  and Assessment of Air Pollutant Abatement
     Strateaies and Priorities,  Task 1:  Air Pollutant Prioritization
     Methodology,  GCA-TR-73-14-G (1974).

 7.   Carroll, J.W.,  and  N.F.  Suprenant, Implementation of  the CCA Prioritiza-
     tion Methodology  for Selected Chemicals, final report, GCA-TR-76-10-G
     (1976).

 8.   Environmental  Protection Agency,  Pesticide Chemical Active Ingredients;
     Proposed Registration  Standards Ranking Scheme, FR 45 No. 222 (Nov. 14,
     1980).

 9.   Fuller,  B., et  al., Preliminary Scoring of Organic Air Pollutants,
     PB-264442, MITRE  Corp., McLean, Va.,  METREK Div.  (1976).

10.   Fuller,  B., et  al., Scoring of Organic  Air Pollutants , MTR-7248, Rev.  1
     (1976).

11.   Final Report  of  NSF Workshop  Panel to Select Organic  Compounds Hazardous
     to the Environment  (Oct. 1975).

12.' Fong, C.V., and  R.J. Clerann,  Hazard Evaluation of Hew Chemicals,
     Approaches to Level I  Test  Selection, MTR-79W00347, MITRE Corp., McLean,
     Va. (1979).

13.   General Procedures  for Scoring Air and  Water Pollutants, draft report,
     Clement Associates, Inc., Washington, B.C.

14.   Gevertz, J.N.,  and  E.  Bild, Chemical Selection Methods:  An Annotated
     Bibliography, EPA 560/TIIS-80-001 (1980).

-------
                                      55
15.    Griesemer,  R.A.,  and C.  Cueto,  Jr.,  Towards a Classification Scheme for
      Degrees of Experimental  Evidence for the Cardnogenicity of Chemicals
      for Animals,  reprinted from Molecular and Cellular Aspects  of Carcinogen
      Screening Tests.

16.    Margler,  L.W.,  M.B.  Rogozen,  R.A.  Ziskind,  and R.  Reynolds,  Rapid
      Screening and Identification of Airborn Carcinogens of Greatest Concern
      in California,  JAPCA, 23(11):1153-1157 (Nov.  1979).

17.    Michigan Critical Materials Register 1980,  Michigan Dept. of Natural
      Resources,  Lansing,  Mich.,  Publication No.  4833-5324.

18.    Michigan Air Priority Chemicals List 1980,  Michigan Dept. of Natural
      Resources,  Lansing,  Mich.,  Publication No.  4833-5324.

19.    Preliminary List of Chemical Substances for Further Evaluation, TSCA,
      Interagency Testing  Committee (1977).

20.    Ross, R.H., and P. Lu, Chemical Scoring System Development,  draft report,
      Oak Ridge National Laboratory,  Oak Ridge, Tenn.  (Dec.  1980).

21.    Ross, R.H., and P. Lu, Chemical Scoring System Development,  draft report,
      Oak Ridge National Laboratory,  Oak Ridge, Tenn.  (June  1980).

22.    Ross, R.H., and J. Welch, Proceedings for the EPA Workshop  on the
      Environmental Scoring of Chemicals,  ORNL/EIS-158,  EPA-560/11-80-010
      (1980).

23.    Scoring Chemicals for Health and Ecological Effects Testing} TSCA-ITC
      Workshop, unnumbered Enviro-Control, Inc.,  report (no  date).

24.    Welch, J.L. and R.H. Ross,  An Approach to Scoring of Toxic  Chemicals for
      Environmental Effects, paper presented at the First Annual  Meeting of
      the Society of Environmental Toxicology and Chemistry, Arlington, Va.
      (Nov. 1980).

25.    Kohan, A.M., A Summary of Hazardous Substance Classification Systems,
      Office of Solid Waste Management Programs,  report No.  SW-171, Washing-
      ton, D.C. (1975).

26.   Lewis, R.J., and R.L. Tatken, eds.,  Registry of Toxic Effects of Chemical
      Substances, October  1981, microfiche Ed. No. 210-81-8101, U.S. Dept. of
      Health and Human Services,  Cincinnati, Oh.  (1981).

27.    National Toxicology Program, Chemicals on Standard Protocol, Carcinogen-
      esis Testing Program, National Toxicology Program, Sethesda, Md. (1982).

28.    National Toxicology Program, Annual Plan, National Toxicology Program,
      Bethesda, Md. (1982).

-------
                                        56
29.   Stanford Research Institute, Chemical  Economics Handbook, Stanford
      Research Institute, Menlo Park, Calif., (1982).

30.   Windholz, M., S. Budavari, L.Y. Stroumtsos, and M.N. Fertig, eds.,  The
      Merck Index, Merck and Co., Inc., Rahway, N.J. (1976).

31.   Weast,  R.C., ed., Handbook of  Chemistry and Physics, The Chemical Rubber
      Co., Cleveland, Oh. (1971).

32.   Leo, A., C. Hanscle, and D. Elkins, Partition  Coefficients  and Their
      Uses,' Chemical Reviews,  71:525-616 (1971).

-------
                         57
                     APPENDIX A
TABLES,  WORKSHEETS, AND ABBREVIATIONS USED IN RTECS

-------
     Table 1  Criteria and Associated Weights for Oncogenicitya
Primary Secondary
Index Criteria13 Weight Weight
1 Evidence of oncogenicity in humans by 5.0
inhalation route.
2 Evidence of oncogenicity in humans by 4.0 0.7
noninhalat ion route.
3 Evidence of one
ogenicity in two or more 3.0 0.5
        animal species by any route of adminis-
        tration.0
4
5
6
7
Evidence of oncogenicity in one animal
species by any route of administration.0
Compound scheduled for or currently
undergoing oncogenicity testing.
Negative or equivocal results from
oncogenicity testing.
No data.
2.0
1.0
0.0
0.0
0.3
0.05
0.0
0.0
aMost available data will relate to carcinogenicity .
     Table A-l in the appendix for a complete list of abbreviations
 used in RTECS ,  Table A-2 for species abbreviations, and Table A-3
 for route of administration abbreviations.
clf the data satisfy the criteria for Index 3 then Index 4 should
 not be considered.

-------
          Table 2  Criteria and Associated Weights for Mutagenicity


                                                          Primary   Secondary
Index                   Criteria3                         Weight      Weight

  1     Evidence of mutagenicity (in vivo) in at least     11.0
        one mammalian test species by the inhalation
        route. b

  2     Evidence of mutagenicity (in vivo) in at least "     9.0        0.7
        one mammalian test species by the non-inhala-
        tion route.

  3     Evidence of mutagenicity (in vitro) in two or       8,3        0.5
        more mammalian test species. ci^

  4     Evidence of mutagenicity (in vitro) in one mam-     7.7        0.4
        malian test species. ^

  5     Evidence of mutagenicity (in vivo) in two or        7.1        0.25
        more nonmammalian test species by any route of
        administration.6

  6     Evidence of mutagenicity (in vivo) in one non-      6.5        0.2
        mammalian test species by any route of admini-
        stration.6

  7     Evidence of mutagenicity (in vitro) in two or       6.0        0.15
        more nonmammalian test species. ^

  8     Evidence of mutagenicity (in vitro) in one          4.0        0.1
        nonmammalian test species. *

  9     Compound scheduled  for or currently undergoing      2.0        0.05
        mutagenicity testing.

 10     Negative or equivocal results from mutagenicity     0.0        0.0
        testing.

 11     No data.                                            0.0        0.0

aSee Table A-l in the appendix for  a complete list of abbreviations used
 in RTECS, Table A-2 for  species abbreviations, and Table A-3 for route of
 administration abbreviations.

°If the route of administration  is  specified the  test is in vivo.

clf the route of administration  is  not specified  the  test is in vitro.
dlf the data satisfy the  criteria for Index 3 then Index 4 should not be
 considered.

elf the data satisfy the  criteria for Index 5 then Index 6 should not be
 considered.
     the data  satisfy the criteria in Index 7 then Index 8 should not be
  considered.

-------
       Table 3  Criteria and Associated Weights for Reproductive
                and Developmental Toxicity
Index
1

2

3


4


5


6

7
Criteria3
Evidence for reproductive or developmental
effects in humans by inhalation route.
Evidence for reproductive or developmental
effects in humans by noninhalat ion route.
Evidence for reproductive or developmental
effects in two or more animal species by
any route of entry. ^
Evidence for reproductive or developmental
effects in one animal species by any route
of entry. k
Compound scheduled for or currently under-
going testing for reproductive and develop-
mental effects.
Negative or equivocal results from testing
for reproductive or developmental effects.
No data.
Primary
Weight
5.0

4.0

3.0


2.0


1.0


0.0

0.0
Secondary
Weight


0.7

0.5


0.3


0.05


0.0

0.0
aSee Table A-l in the appendix for a complete list of abbreviations used
 in RTECS, Table A-2 for species abbreviations, and Table A-3 for route
 of administration abbreviations.

^If the data satisfy the criteria for Index 3 then Index 4 should not be
 cons idered.

-------
    Table 4  Criteria and Associated Weights for Acute Lethality
Route of Exposure and Criteriaa
Index
1
2
3
4
5
6
7
8
9
Species
Human
Human
Animal
Human
Human
Animal
Human/
An ima 1
Human/
Animal
Human/
Animal
Inhalation
Gas(ppm) Solid(mg/mJ)
X<5
-
X<5
550
Negative or
or animals.
No data.
X<50
-
X<50
50500
insignificant

Oral
(mg/kg)
-
X<5
X<5
-
550
results

Dermal Primary
(tng/kg) Weight
-
X<5
X<5
-
5200
in humans

4.7
3.7
3.0
2.6
2.3
2.0
0.0
0.0
0.0
aSee Table A-l in the appendix for a. complete list of abbreviation-
 used in RTECS, Table A-2 for species abbreviations, and Table A-3
 for route of administration abbreviations.

-------
              Table 5  Criteria and Associated Weights for
                       Effects Other than Acute Lethality
Route of Exposure and Criteria3
Index
1
2
3
4
5
6
7
8
9
Species
Human
Human
Human
Human
Animal
Animal
Human/
Animal
Human/
Animal
Human/
Animal
Inhalation Oral Dermal
Gas(ppm) Solid(mg/m3) (mg/kg) (mg/kg)
xa
-
i10
Negative or
or animals.
No data.
xao
xa xa
io100 X>10 X>10
insignificant results in humans

Primary
Weight
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
0.0
aSee Table A-l in the appendix for a complete list of abbreviations
 used in RTECS, Table A-2 for species abbreviations, and Table A-3
 for route of administration abbreviations.

-------
Table 6  Criteria and Associated Weights
         for Production Volume (PV)
Criteria
Index
1
2
3
4
5
6
7
8
106 kg/yr
PV>450
2301000
5io
-------
     Table 7  Criteria and Associated Weights
              for Vapor Pressure (VP)
Index
1
2
3
4
5
6
Criteria3'13
VT(mmHg) bp(*C)
Gas
Solidc
VPHOO bp<80
24100
No data.
Primary
Weight
4.0
3.0
3.0
2.0
1.0
1.0
aVapor pressure data should be reported at 25°C
 and 760 mm Hg.

°If vapor pressure data is unavailable use the
 substance boiling point (bp), at 760 ram Hg, as
 a substitute.

CA substance should be considered a solid if its
 melting point is greater than 25*C.

-------
Table 8  Criteria and Associated
         Weights for Bioaccurau-
         lation
Index
Criteria
Primary
Weight
  1     Log  P>6.0        10.0
  2     4.0
-------
Table 9  Criteria and Associated Weights
         for Existing Standards
Index
1
2
3
4
5
6
7
Criteria
Gas ( ppm)
X<5
5200
No
Solid
(mg/m^)
X<0 . 25
0.2510.0
standard.
Primary
Weight
6.0
5.0
4.0
3.0
2.0
1.0
0.0
aBased on OSHA time-weighted-average (TWA)
 standards or threshold limit values (TLV)
 when TWAs are not available.

-------
                      WORKSHEET 1.  CENTRAL INFORMATION*
Chemical Name:

RTECS Number:5

CAS NUMBER:0
MOLECULAR WEIGHT:d 	 g/g-mol

MOLECULAR FORMULA:6
MELTING POINT:f	: *C   BOILING POINT:f	"C

SUBSTANCE PHYSICAL STATE:8  Gas       Liquid       Solid       Unknown

RTECS Edition:h 	

MERCK INDEX Edition:*	

HANDBOOK OF CHEMISTRY AND PHYSICS Ed it ion:J
aAll data taken from the Registry of Toxic Effects of Chemical Substances
 (RTECS).  The latest microfiche edition of RTECS should be used for the
 analysis and can be obtained from the Superintendent of Documents, U.S.
 Government Printing Office, Washington, D.C.


bUsing the RTECS data field RTECS Accession Number, enter the RTECS number.


cUsing the RTECS data field CAS NUMBER enter the Chemical Abstract Service
 Registry Number.


dUsing the RTECS data field MOLECULAR WEIGHT, enter the molecular weight.


eUsing the RTECS data field MOLECULAR FORMULA, enter the molecular formula.


%sing the most recent edition of the Merck Index, enter the melting/boiling
 points of the substance under analysis.  If the substances is not in  the Merck
 Index use the most recent  edition of the Handbook of Chemistry and Physics
 (Tables of Physical Constants of Organic/Inorganic Compounds) for the melting/
 boiling points. •

SUsing the information on physical characteristics from the Merck Index  and
 the melting/boiling point  information determine the physical state of the sub-
 stance under analysis.

hEnter the edition of RTECS used for this analysis.

i-Enter the edition of the Merck Index used for this analysis.

JEncer the edition of the Handbook of Chemistry and Physics used for this
 analvsis.

-------
                         WORKSHEET 2.   ONCOGENICITY
Using Che RTECS  data  field TUMORIGENIC DATA, record  the  route of administra-
tion and  test  species for all  citations  that show the Toxic  Effect  (TFX)  as
TFX:CAR, TFX:NEO, or TFX:ETA.
             ROUTE OF       TEST             ROUTE OF       TEST
          ADMINISTRATION   SPECIES        ADMINISTRATION   SPECIES
Using  the  latest edition  of  the Carcinogenis  Testing  Program:   Chemicals on
Standard  Protocol3   from  the  National  Toxicology Program,  record  the status
of  the  substance  in  the Carcinogenesis Testing  Program.


     Q Not on  list  (should not  be  scored on Worksheet  10)

     Q Scheduled for  or currently  undergoing  testing

     Q Negative  or  equivocal results

     Edition  Usedb:
 aA copy can be obtained  from:   Technical Information Section,  Carcinogenesis
  Testing  Program,  National  Toxicology  Program,  Landow  Building,  Room  A306,
  Bethesda,  MD  20205.
 bEnter  the edition of  the  Carcinogenis Testing Program:Chemicals on  Standard
  Protocol  used for this analysis.

-------
                         WORKSHEET 3.   MUTAGENICITY
Using the RTECS data field MUTATION  DATA,  record  the test species and route of
administration (if applicable)  for all citations.  Below the heading MUTATION
DATA, each mutation data  line includes, in sequence, the mutation test system
utilized, the species of  the tested organism (and where applicable,  the route
of  administration  or  cell  type),  and the  exposure  concentration or  dose.
           TEST        ROUTE OF            TEST        ROUTE OF
          SPECIES   ADMINISTRATION        SPECIES   ADMINISTRATION
Using  the  latest  edition  of  the  National Toxicology  Program  Annual Plan,3
record the status of the  substance  in  the  Mutagenicity Testing  Program.
Tables  in  the section on  cellular  and genetic toxicology should be used  for
the  following  analysis  (See for example,  pages 35-73,  Tables 2-8, of  Che  NTP
Aur.ual
     n Not on list (should not be scored on Worksheet 11)

     Q Scheduled for or currently undergoing testing

     Q Negative or equivocal results

     Edition
 aA copy  can be obtained from:   Technical  Information Section,  National
  Toxicology  Program, Landow Building, Room A306, Bethesda, MD 20205.
 ''inter  the  edition  of  the National  Toxicology  Program Annual Plan used  for
  this analysis.

-------
            WORKSHEET 4.   REPRODUCTIVE  AND  DEVELOPMENTAL  TOXICITY
Using  the  RTECS data  field REPRODUCTIVE  EFFECTS  DATA, record  the  rouce of
adminiscration and  test  species  for all  citations.
             ROUTE OF       TEST             ROUTE  OF        TEST
          ADMINISTRATION   SPECIES        ADMINISTRATION    SPECIES
Using  the National  Toxicology  Program  Annual  Plan3,  record  the  status  of
Che  substance  in the Teratogenic  Testing  Program.   Tables in  the  section  on
reproductive  and developmental  toxicology  should be  used for  the  following
analysis  (sse  for  example,  pages  125-134,  Tables  17-20,  of the NTP  Annual
Plan28).


     Q Not on list (should not be scored on Worksheet 12)

     Q Scheduled for or currently undergoing testing

     Q Negative or equivocal results

     Edition Usedb:
 aA copy  can be obtained from:   Technical Information Section, National
  Toxicology  Program, Landow Building, Room A306, Bethesda, MD 20205.

 nEnter  :ne  edition  of  the National  Toxicology Program Annual  Plan  used  for
  th is anaivs is.

-------
                        WORKSHEET 5.   ACUTE LETHALITY
Using the RTECS  data  field TOXICITY AND DATA  REFERENCES,  record  che  route of
administration,  test  species,  dose,  and  length  of  exposure  for all  cita-
tions (other  than TFX:CAR,  TFX:NEO,  or TFX:ETA)  that  are of the  form  LDLo,
LD^Q, LCj^Q,  and LC5Q that give the route of administration as ihl, orl,  or skn
and length of exposure less than or equal to 24 hr.a


                 ROUTE OF       TEST                   LENGTH OF
              ADMINISTRATION   SPECIES      DOSEb>c     EXPOSURE
 aAny  data  for  LDLo>  LD5Qj LC^o' or ^5Q with exposure  times greater than 24 hr
  should  be  included in  Worksheet  6  (Effects  Other  Than  Acute Lethality).

 ^Concentrations  for solid substances  should  be recorded  as  mg/m3 and cannot
  be  converted to  ppm.    If  the physical  state of  the  substance is unknown,
  record the  units  given in  RTECS and  do not  attempt  to  convert  to  ppm.
  Occasionally,  the dose  for  solids  will be given  in  ppia's  (usually when the
  melting  point is  near  25*C), the  dose then  should be  recorded in ppm's and
  not  converted to  mg/m^.
 cFor  gases and vapors,  concentrations in mg/m^ should be  converted to  ppm by:

  ppm  = —rj£— x concentration  (mg/m^)   @  25*C and  760 mo  Hg

  where MW  = molecular weight  (Worksheet  1).

-------
               WORKSHEET 6.  EFFECTS OTHER THAN ACUTE LETHALITY
Using  Che  RTECS data  field TOXICITY  AND DATA  REFERENCES,  record  the  route
of administration,  test  species, and  dose for  all citations  (other  than
TFX:CAfi, TFX:NEO, or TFX:ETA) that are of the form TDLo> or TCLo that give the
route  of  administration as  ihl,  orl, or  skn.    Citations  of  the  form  LD^QJ
LDjQ, LCLo> or LG50  cl*ac 8^-ve  che route  °f administration as ihl, orl, or skn
and length of exposure greater than 24 hr should be included here.
                         ROUTE OF       TEST
                      ADMINISTRATION   SPECIES      DOSEa>b
 Concentrations  for  solid  substances  should  be  recorded  as mg/ai^ ancj  cannot be
 converted  to ppm.   If  the physical state of the substance is unknown,  record
 the  units  given  in RTECS and do not  attempt  to  convert  to ppm.  Occasionally,
 the  dose  for solids  will be given  in ppm's  (usually when the melting point is
 near 25*C), the  dose  Chen should  be recorded  in ppm's  and- not converted Co
 'DFor gases and vapors,  concentrations in mg/m^ should be  converted  to  ppm  by:

  ppm = - "*y?,  x concentration (mg/nr)   @  25*C  and  760  mm  Hg

  where MW * molecular weight (Worksheet  1).

-------
                   WORKSHEET la.  PRODUCTION VOLUME (PV)
Using the SRI Chemical Economics Handbook find the total U.S. production for
the substance under  analysis.

PV - 	 x 106 kg/yr   or  PV = 	 x 106 Ib/yr
Edition Used3:
aEnter  the edition  of  the SRI chemical  Economics  Handbook used  for this
 analysis.

-------
                        WORKSHEET 7b.  VAPOR PRESSURE  (VP)


Using Che information  from Worksheet 1, record the physical  state (at 25*C and
760 mm Hg^  and boiling  point (BP) of the substance under analysis.

  D Gas

  D Solid

  Q Liquid                                         BP - _ °C
     Unknown
If the physical state is  given  as  gaseous,  solid, or unknown, continue on to
Worksheet 8.

If the substance is a liquid  use  the  data  from  the Handbook of Chemistry and
Physics-^ to  calculate the vapor pressure (VP), at 25*C  and  760 mm Hg.a  Enter
constants  A  and B, and  the temperature range  for  which the equation and
constants are valid.
Temperature   Range:	"C

VP  »	mmHg  @  25 "C and  760 mmgHg

If  25*C  does not  fall  within the  temperature range above continue on to
Worksheet 8.
aVP (mm Hg) - antilog10  [(-7.3285 x  10~4 x A)+B]

where A = molar heat of vaporization
      B * constant.

Constants  A  and  B are obtained from  the Handbook  of  Chemistry  and  Physics^*
pages  D-151 to  D-170 for organic  compounds  and  pages D-171  to D-177 for
inorganic compounds.

-------
                         WORKSHEET 8.   BIOACCUMULATION


Using Table XVII  from Leo et al.32,  find  the  MOLECULAR FORMULA (Worksheet l)
of the substance under analysis in the column headed EMPIRICAL FORMULA.  Find,
in the  column headed SOLVENT,  the  line corresponding  to octanol.   Match the
chemical name  of  the substnace under analysis with the chemical  name in the
column headed  NAME.   Enter,  in the  space provided above,  the  value from the
column headed  LOGP OCT.   If more  than one value of the LOGP  OCT is  given for
the octonal solvent system use the average of the LOGP OCT values.  If octonal
is not  one of  the solvents listed for  the  given  MOLECULAR  FORMULA and NAME
enter the average LOGP OCT estimated  from the other solvent systems.

-------
                      WORKSHEET 9.   EXISTING STANDARDS


Using  Che RTECS  data  field STANDARDS AND  REGULATIONS,  record  che entry
for  the  OSHA time  weighted  average (OSHA STANDARD-air:TWA) in  the space
provided  below.   If  OSHA  Standard-air  TWA  is  not given use  the RTECS data
field REVIEW, record  the entry  for  the  threshold  limit value (THRESHOLD LIMIT
VALUE-air:) in the space provided below.


STANDARDS AND REGULATIONS:  OSHA STANDARD-air:TWA « 	 ppm  (gas)a.b

                                                   	mg/m3  (solid)3.b
REVIEW: THRESHOLD LIMIT VALUE-air:  	 ppm (gas)a>b

                                   	 mg/m3  (solid)aib
 aStandards  for solid  substances  should  be  recorded as  mg/m3  and cannot  be
  converted  to  ppm.   If the physical  state of the substance is unknown,  record
  the  standard  in the  units  given in RTECS  and  do  not attempt  to  convert  to
  ppm.  Occasionally,  the standard for  solids will be given  in  ppm's  (usually
  when  the melting point is near 25*C), the standard then should be recorded  in
  ppm's and  not converted to mg/m3.
 "For  gases and  vapors, standards  in mg/m3  should  be converted  to  ppm  by:

  ppm - 2.4v^.5 x standard (ng/m3)  $ 25'C and 760 mm Hg

  where MW - molecular  weight (Worksheet 1).

-------
                WORKSHEET 10.   ONCOGENICITY  FACTOR  SCORE
    The completion of Worksheet 10 require*  the  uae  of Worksheet 2.

1.  Circle  Che  index  number*,  in  the  t«ble below,  that corretpond  co Che
    criteria that  »re satisfied by Che data  from Worksheet 2.

2.  Starting at Index  1,  read down  Che  Index  column until  the fint circled
    index number  i«  encountered.   Record the Primary  Weight  anociated vith
    that index number in the corre*pondiog Criteria  Weight column.

3.  Continue reading down  the Index column  to  each successive circled  index
    number and record  the  Secondary  Weight  aitociated  with  each index number
    in the corresponding Criteria  Weight  column.

A.  SUB all  value*  recorded in the  Criteria Weight column  and record in the
    •pace labeled  £(Criceria Weight)  below.

5.  Divide  Che  reiulcs from  Seep 4  by  6.25 Co  obtain  the  Normalized Factor
    Score for Oncogenicicy (ONCOnorm).


              Criteria and Aisociated Weight* for Oncogenicicy4
Index
1
2
3
4
5
6
7
Criteria*1
Evidence of oncogenicity in huaan* by
inhalation route.
Evidence of oncogenicity in human* by
aoninhalat ion route.
Evidence of oncogenicity in £vo or more
animal ipecie* by any route of adminis-
tration. c
Evidence of oncogenicity in one animal
ipecie* by any route of adminiccrac ion.c
Compound icheduled for or currencly
undergoing oncogenicity testing.
Negative or equivocal result* from
oncogenicity testing.
Mo data.
?rimary
Weight
5.0
4.0
3.0
2.0
1.0
0.0
0.0
Secondary
Weight

0.7
0.5
0.3
0.05
0.0
0.0
Criteria
Weight








    _                        .  w.*^
    2,(Criteri* Weight)                     ONCOnorm


'Host available data vill relate to carcinogenicity.

bSee Table A-l in the appendix for a cooplete Hit  of abbreviation* used  in
 RTECS, Table A-2 for specie* abbreviation*, and  Table*A-3  for  rouce of admini-
 ttration abbreviation*.

clf the data iati*fy the criteria for Index 3 then  Index 4  *hould not be  consi-
 dered.

-------
              WORKSHEET  11.   MUTAGENICITY FACTOR SCORE
         The completion of Worksheet  11  requires  Che  use  of Worksheet 3.

     1.   Circle Che index numbers,  in Che  Cable  belov,  chat correspond  to  the
         criteria that  are satisfied  by  the  data  from Worksheet 3.

     2.   Starting at  Index 1,  read  dovn  the  Index column  until Che  first circled
         index number is encountered.  Record  the Primary Weight  associated with
         Chat index number in  the corresponding  Criteria  Weight column.

     3.   Continue reading dovn the  Index colunn  to each successive  circled index
         number and record the Secondary Weight  associated with each  index number
         in the corresponding  Criteria Weight  column.

     4.   Sum all values recorded in the  Criteria  Weight column and  record  in the
         space labeled £( Criteria Weight)  below.

     5.   Divide the results from Step 4  by 12.65  to obtain the Normalized Factor
         Score for Mutagenicity (MUTnorm).


                   Criteria and Associated Weights for  Mutagenieicy


                                                          Primary  Secondary   Criteria
Index                   Criteria'                         Weight      Weight      Weight

  1     Evidence of mutagenicity (in  vivo) in  at  least     11.0
        one mammalian test species  by the  inhalation
        route.b

  2     Evidence of mutagenicity (in  vivo) in  at  least      9.0         0.7
        one mammalian test species  by the  non-inhala-
        tion route.

  3     Evidence of nutagenicity (in  vitro)  in two or      8.3         0.5
        more mammalian  test species.c>d

  4     Evidence of nucagenicity (in  vitro)  in one mam-    7.7         0.4
        ma Inn cesc species.^

  5     Evidence of nutagenicity (in  vivo) in  tvo or        7.1         0.25
        more nonaaomalian test species by  any  route
        of administration.'
6


7

8

9

10

11
Evidence of mutagenicity (in vivo) in one non-
maomalian test species by any route of admini-
stration.'
Evidence of mutagenicity (in vitro) in tvo or
•ore nonmamalian test species, f
Evidence of mutagenicity (in vitro) in one
nonmaomalian test species. f
Compound scheduled for or currently undergoing
outagenicity testing.
Negative or equivocal results from mucagenicity
testing.
(to data.
6.5


6.0

4.0

2.0

0.0

0.0
0.2


0.15

0.1

0.05

0.0

0.0
_________________ i t? *< < 	
                      Weight)    '  '-•"-1            MOTnoro


•See Table A-l in the appendix for  a complete list of abbreviations used in RTECS,
 Table A-2 for species abbreviations, and Table A-3 for route of administration
 abbreviations.

blf the route of administration is  specified the test is in vivo.

clf the route of administration is  aot specified :he test is in vicro.

d!f the data  satisfy the criteria for Index 3 then Index 4 should  aot be consi-
 dered.

•If the data  satisfy the criteria for Index 5 then Index 6 should  aot b« consi-
 dered.

'If the data  satisfy the criteria ia Index 7 then Index 8 should not be consi-
 dered.

-------
 WORKSHEET 12.   REPRODUCTIVE AND  DEVELOPMENTAL  TOXICITY FACTOR SCORE
         The completion  of Worksheet  12  require* the uie Worksheet 4.

     1.   Circle  Che  index nuabers,  in the  table below,  that  correapond to  the
         criteria that (re aaciafied  by  the data from Worksheet 4.

     2.   Starting »t Index 1,  read  down the Index column  until  the fir*t circled
         index nusber ia  encountered.   iecord  the Primary Weight  aaaociated  vith
         that index  amber in  the corresponding Criteria Weight column.

     3.   Continue reading dovn  the  Index  column  to each *uccei*ive circled  index
         number and  record the  Secondary Weight aiaociated vith  each index  number
         in the corresponding  Criteria Weight column.

     4.   Sun til valuei  recorded  in the Criteria Weight column  and record  in  the
         •pace labeled ^(Criteria Weight) below.

     5.   Divide the  results  froa  Step 4 by  6.25  to  obtain the  Normalized  Factor
         Score for Reproductive and  Developmental Toxicity (RDTnorm).


    Criteria and Aaaociated Weight!  for Reproductive and Developmental Toxicity
Index
                        Criteria*
  Primary   Secondary   Criteria
  Weight      Weight      Weight
1

2

3
Evidence for reproductive or developmental
effecta in humana by inhalation route.
Evidence for reproductive or developmental
effecta in humans by noninhalat ion route.
Evidence for reproductive or developmental
5.0

4.0

3.0


0.7

0.5
        effecti in two or  more  animal  species by
        any route of entry.^

        Evidence for reproductive or developmental     2.0        0.3
        effecti in one animal  species  by  any route
        of entry.b

        Compound scheduled for  or currently under-     1.0        0.05
        going testing for  reproductive and develop-
        mental effects.

        Negative or equivocal  results  from testing     0.0        0.0
        for reproductive or developmental effects.

        No data.                                      0.0        0.0
         2,(Criteria Weight)
                                   T  6.25 •
SDTnorm
'See Table A-l in the appendix  for  a complete liat of abbreviations vised in RTECS,
 Table A-2 for apecies abbreviations,  and Table A-3 for route of administration
 abbreviations.

'If the data satisfy the  criteria  for  ladex 3 then Index 4 should not be considered.

-------
               WORKSHEET 13.   ACUTE  LETHALITY  FACTOR  SCORE
    The completion  of  Worksheet 13 require* the use  of Worksheet 5.

1.  Circle the  index  numbers,  ia  the  table  belov,  chjt correspond  Co  the
    criteria  thie are  tatisfied by the data from Worksheet 5.

2.  Starting  at Index 1,  read  down  the Index column until  the first circled
    index number if encountered.  Record  the Primary  Weight  associated vith
    that index  number  in Che (pace labeled Criteria  Weight below.

3.  Divide the value obtained in Step 2 by A.7  to obtain  the Normalized
    Factor Score for Acute Lethality (AlZTHnora).
              Criteria  and Asiociated Weight* for Acute  Lethality
Index
1
2
3
4
5
6
7
8
9
Specie*
Huaan
Human
Animal
Huaan
Human
Animal
Human/
Animal
Human/
Animal
Human/
Animal
Route
of Exposure
Inhalation
Ca«(ppm) Solid(mg/m3)
X<5
-
X<5
5OC<50
-
550
negative
human* or
No data.
X<50
-
X<50
50500
and Criteria*
Oral Dermal
(mg/kg) (mg/kg)
-
X<5 X<5
X<5 X<5
-
550 X>200
or in«ignif icant re*ult* in
animal*.


Primary
Weight
4.7
3.7
3.0
2.6
2.3
2.0
0.0
0.0
0.0

              Criteria Weight
ALETHnorm
       'See Table A-l  in  the appendix for a complete lilt  of  abbreviations
        used in STICS,  Table A-2 for specie* abbreviation*, and Table A-3 for
        route of adminiitration abbreviation*.

-------
 WORKSHEET  14.   EFFECTS OTHER THAN ACUTE  LETHALITY  FACTOR SCORE
    The couplet ion  of Worksheet 14 require! Che use  of  and Worksheet 6.

1.  Circle Che index nunbers, in Che table below,  that  correspond Co Che
    criteria that are satisfied by the data froa Worksheet 6.

2.  Starting at  Index 1, read down the Index column  until Che firsc circled
    index number is encountered.  Record the Primary Weight associated with
    that index number in the space labeled Criteria  Weight below.

3.  Divide the value obtained in Step 2 by 7.0 to  obtain the Normalized
    Factor Score for Effect* Other Chan Acute Lethality (NLZTHnonn).
    Criteria  and Associated Weights for Effects  Other  than Acute Lethality
Route of Exposure and Criteria*
Index
1
2
3
4
5
6
7
8
9
Species
Human
Hunan
Hunan
Human
Animal
Animal
Human/
Animal
Human/
Animal
Human/
Animal
Inhalation Oral Dermal
Gaa(ppm) Solid(mg/m3) (mg/kg) (og/kg)
X<1
-
110
Kegacive
in humans
Mo data.
X<10
X<1 X<1
10100 X>10 X>10
or insignificant results
or animals.

Primary
Weight
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
0.0
           Criteria  Weight
                               T 7.0  «
HLETHnorm
    •See Table A-l  in  th« appendix for a complete lisc  of  abbreviations
     used in RTECS,  Table A-2 for species abbreviations, and Table A-3 for
     route of administration abbreviations.

-------
    WORKSHEET 15.   POTENTIAL FOR AIRBORNE  RESEASE  FACTOR  SCORE

    The  completion  of Vorkiheet  15  require* the  use  of Worksheet*  7i  and
7b.

                             Production Volume (PV)
1.  Circle the  index number,  in  the  table  below,  that correspond*  to  the
    ?V interval  that  the PV data from Work»heec 7a fall*  into.
                                                            •
2.  Stad  the  Primary Weight  correiponding  to the  index number  circled  in
    Step 1 and record  it in the space labeled PV Weight below.


                   Criteria and Associated Weight! for  Produc-
                   tion Volume (PV)
Criteria
Index
1
2
3
4
5
6
7
8
10° kg/yr
PV>450
230100
So data
Primary
We ight
4.0
3.0
3.0
2.0
1.0
1.0 •
5.  Divide  the Product  cf PV Weight  and VP Weight  by 40.0 to  obtain the
    normalized Factor Score  for  Potential for Airborne Release  (AIRJOnorm).
                                        T 40.0
         PV Weight         VP Weight      '  v'                AIRBOnona
                 'Vapor prexure data thould be reported at  25*C
                  and 760 OB Bg.
                 "If vapor pre**ure data  i* unavailable use  the
                  substance boiling point (bp), at 760 mm Hg,  a*
                  a (ubttitute.
                 CA subcance ihould be con*idered a (olid if  it*
                  •citing point  is greater thin 25'C.

-------
                  WORKSHEET 16.   BIOACCUMULATION FACTOR SCORE
    The completion of Worksheet 16 requires the use of Worksheet 8.

1,   Circle the index numbers, in the table below, that correspond to the
    Log,0  P interval that the Log10  P data from Worksheet 8 falls into.

2.   Read the Primary Weight corresponding to the index number circled in
    Step 1 and record it. in the space labeled Bioaccumulation Weight below.

3.   Divide Bioaccumulation Weight by 10.0 to obtain the Normalized Factor
    Score for Bioaccumulation (BIOAnorm).
                       Table 8  Criteria and Associated
                                Weights for Bioaccumu-
                                lation

Index
1
2
3
4
5

Criteria
Log10P>6.0
4.0
-------
                 WORKSHEET 17.   EXISTING  STANDARD  FACTOR SCORE
    The completion of Worksheet  17  requires  the use of Worksheet 9.

1.  Circle  the  index  numbers,  in  the  table below,  that correspond  to  the
    TWA (or TLV, if TWA unavailable) interval that the TWA (or TLV) data from
    Worksheet 9  falls into.

2.  Read  the  Primary  Weight  corresponding  to  the  index number  circled  in
    Step 1 and record it  in  the  space labeled  Standard Weight below.

3.  Divide  Standard  Weight by  6.0 to obtain  the Normalized Factor Score
    for Existing Standard  (ESTDnorm).
                 Criteria and Associated  Weights  for Exist-
                 ing Standards

Index
1
2
3
4
5
6
7
Criteria
Gas (ppm)
X<5
5200
No standard.
Solid
(mg/m3)
X<0 . 25
0.2510.0

Primary
Weight
6.0
5.0
4.0
3.0
2.0
1.0
0.0
- 6 n T-
          Standard Weight
ESTDnorm
                 aBased on OSHA time-weighted-average  (TWA)
                  standards or threshold limit  values  (TLV)
                  when TWAs are not available.

-------
          WORKSHEET 18.   CALCULATE NORMALIZED  GROUP SCORE
                        CAXClSOCEHIClTf CROC? (CAR)

    Record the Normalized  Factor Scorei (HFS)  for Oncogenicity (ONCOnorm from
    Workfheet  10) «nd  Mutagenicity (MUTnorn from Worksheet  11)  and  complete
    the requefted mathematical  procedures  to  calculate  the  normalized  group
    •core  for  CAR (CAKnorm).
                                         7 4.40     r 1.23  -
        ONCOnorm        I    MUTnorm      '      M   '           CAtaorm
               REPRODUCTIVE AMD DEVELOPMENTAL TOX1CITY GROUP (REPRO)
                                                               i
1.   Record the NFS  for  Reproductive  and  Developmental  Toxicity  (RDTnorm from
    Worksheet 12).   Because  there  ii only  one factor  in  the REPRO  group,
    RDTnorm  it  equal  to  the  nornalized  group  score  for REPRO  (REPROnorm).
                            RDTnorm
                                                REPROnorm
                            TOXICin GROUP  (TOX)

    Record  the  NTS  for  Acute  Lethality  (ALETHnorm  froa  Worksheet  13)  and
    Effects  Other than  Acute  Lethality  (NLETHnorm froo Worksheet  14  and
    complete the requested  mathematical procedures to calculate the normalized
    group score for TOX (TOXnona),
                                                     T 2.0
           ALETHnona                NLETHnorm         '   '        TOXnorm
                            EXPOSURE  GROUP  (EXPO)

    Record the NFS  for Potential for Airborne  Release  (AIRBOnorm  from Work-
    sheet  15)  and Bioaccuaulation (BIOAnonn  from Worksheet  16)  and complete
    the  requested mathematical  procedures  to calculate  the  normalized group
    score for EXPO (LXPOnora).
    I  ^   AIRBOnorn    * 10-°j  *    BIOAnorm     I ?
11.0
         EXPOnorm
                           STANDARDS CROUP (STD)

5.  Record the NFS  for  Existing  Standards (ESTDnorm from Worksheet 17).  Be-
    cause there is only one  factor  in the  STD group, ESTDnorm  is equal to the
    normalised group score for STD (STDnorm).


      	H   -.	 «  STDnorm
              ESTDnorm

-------
                   WORKSHEET 19.  SUBSTANCE PRIORITIZATION
    The completion of Worksheet 19 requires the use of Worksheet 18.

1.   Record  the  values for CARnorm, REPROnorm,  TOXnorm,  EXPOnonn,  and STDnorm
    from Worksheet 18 in the appropriate space below.

2.   Calculate the Normalized Substance Rank.
      2 x
          CARnorm
j *  V2  X REPROnorray   I TOXnorm '   (5  X EXPOnonn I"1"
                0.5 x
                      STDnorra /
               T 10.5
                        Normalized  Substance Ranka
aValues for Nortna'i^ac Substance Rank will range from 0.014 to 1.0.

-------
  Table A-l.  Comprehensive List of All Abbreviations
              Used in RTECSa>b
ALR - allergenic effects
AQTX- Aquatic Toxicity
asn - Aspergillus nidulans (mold)
ast - Ascites tumor
BCM - blood clotting mechanism effects
bcs - Bacillus subtilis (bacteria)
bfa - body fluid assay
BLD - blood effects
bmr - bone marrow
BPR - blood pressure effects
brd - bird (domestic or lab)
bvd - wild bird species
C   - continuous
CAR - carcinogenic effects
cat - cat
cc  - cubic centimeter
end - child
ckn - chicken
CL  - ceiling concentration
clr - Chlamydomonas reinhardi (protoza)
CNS - central nervous system effects
COR - corrosive effects
GRIT DOC - NIOSH criteria document
ctl - cattle
CUM - cumulative effects
CVS - cardiovascular effects
cyt - cytogenetic analysis
D   - day
dck - duck
DDP - drug dependence effects
DEF - definition
dlt - dominant lethal test
dog - Drosophila melanogaster (insect)

-------
                  Table A-l.  (Cont'd)
dnd - DNA damage
dni - DNA inhibition
dnr - DNA repair
dns - unscheduled DNA synthesis
dog - dog
dom - domestic
DOT - Department of Transportation
dpo - Drosophila pseudo-obscura (insect)
emb - emb ryo
EPA - Environmental Protection Agency
esc - Escherichia coli (bacteria)
ETA - equivocal tumorigenic agent
eug - Eugiena gracilis (protoza)
eye - administration into eye (irritant)
EYE - eye effects (systemic)
fbr - fibroblast
frg - frog
GIT - gastrointestinal tract effects
GLN - glandular effects
gm - gram
gpg - guinea pig
grb - gerbil
grh - grasshopper
H - hour
ham - hamster
hla - HeLa cell
hma - host mediated assay
hmi - Haemophilus influenzae (bacteria)
htnn - human
hor - horse
I - intermittent
ial - intraaural
IARC - International Agency for Research on Cancer
iat - intraarterial

-------
               Table A-l.  (Cont'd)
ice - intracerebral
lev - intracervical
idr - intradermal
idu - intraduodenal
ihl - inhalation
imm - immersion
imp - implant
ims - intramuscular
inf - infant
ipc - intraplacental
ipl - intrapleural
ipr - intraperitoneal
IRDS- primary irritation dose
irn - intrarenal
IRR - irritant effects (systemic)
isp - intraspinal
itr - intratracheal
ivg - intravaginal
ivn - intravenous
kdy - kidney
kg  - kilogram
kip - Klebsiella pneumoniae (bacteria)
L   - liter
LC5Q- lethal concentration 50 percent kill
L^Lo" lowest published lethal concentration
    - lethal dose 50 percent kill
    - lowest published lethal dose
leu - keukocyte
Ing - lung
Ivr - liver
lym - lymphocyte
M   - minute
nH  - cubic meter
mam - mamal (species unspecified)

-------
               Table A-l.  (Cont'd)
man - man
mg  - milligram
MGN - multigeneration
mky - monkey
ml  - milliliter
MLD - mild irritation effects
mma - microsomal mutagenicity assay
MMI - mucous membrane effects
mmo - mutation in microorganisms
mmol- millimole
mmr - mammary gland
mnt - micronucleus test
MOD - moderate irritation effects
mol - mole
mppcf - million particles per cubic foot
mrc - gene conversion and mitotic recombination
msc - mutation in somatic mamallian cells
MSK - musculo-skeletal effects
MTDS- mutation dose
MTH - mouth effects
mul - multiple routes
mus — mouse
NEO - neoplastic effects
ng  - nanogram
nmol- nanomole
nsc - Neurospora crassa (mold)
nse - non-standard exposure
DBS.- obsolete (trade name)
ocu - ocular
omi - other microorganisms
oin - other insects
open- open irritation test
orl - oral
OEM - Other Regulated Material (DOT)

-------
                  Table A-l.   (Cont'd)
OSHA - Occupational Safety and Health Administration
oth - other cell types
otr - oncogenic transformation
ovr - ovary
par - parenteral
pg - picogram
pgn - pigeon
pic - phage inhibition capacity
pig ~ pig
Pk - peak concentration
pmol - picomole
PNS - peripheral nervous system effects
post - after birth
ppb - parts per billion
pph - parts per hundred
ppm - parts per million
ppt - parts per trillion
pre - prior to copulation
preg - pregnant
PSY - psychotropic effects
PUL - pulmonary system effects
qal - quail
rat - rat
RBC - red blood cell effects
rbt - rabbit
rec - rectal
REGS - standards and regulations
rns - rinsed with water
RPDS - reproductive effects dose
RTECS - Registry of Toxic Effects of Chemical Substances
S - second
sal - salmon
sat - Salmonella typhimurium (bacteria)
see - sister chromatid exchange

-------
                  Table A-l.  (Cont'd)
WBC - white blood cell effects

wmn - woman
Y - year

% - percent
aFrom RTECS microfiche edition, October 1981 (Lewis and
 Tatken 1981).
      recent editions of RTECS may use abbreviations not
 included on this list.  Any abbreviation can be iden-
 ified by using the Key to Abbreviations in the Appendix
 of the RTECS edition being used.

-------
    Table A-2.   Species Abbreviations Used  in RTECSa>b

                        Mammalian
cat - cat, adult
ctl - cattle
chd - child
dog - dog, adult
dom - domestic animal (goat,  sheep)
grb - gerbil
gpg - guinea pig, adult
ham - hamster
hor - horse, donkey
hmn - human
inf - infant
mam - mammal (species unspecified in reference)
man - man
mky - monkey
mus - mouse
Pig - pig
rbt - rabbit, adult
rat - rat
sql - squirrel
wmn - woman
                      Nonmammalian
asn - Aspergillus nidulans (mold)
bcs - Bacillus subtilis (bacteria)
brd - bird (any domestic or laboratory bird reported but
      not otherwise identified)
bwd - bird (wild bird species)
ckn - chicken
clr - Chlamydomonas reinhardi (protoza)
dck - duck
dmg - Drosophila melanogaster (insect)
dpo - Drosophila pseudo-obscura (insect)
esc - Escherichia coli (bacteria)

-------
                  Table A-2.   (Cont'd)
eug - Eugiena gracilis (protoza)
frg - frog
grh - grasshopper
hmi - Haemophilus influenzae (bacteria)
Kip - Klebsiella pneumoniae (bacteria)
nsc - Neurospora crassa (mold)
pgn - pigeon
qal - quail (laboratory)
sal - salmon
sat - Salmonella typhimurium (bacteria)
slw - silkworm
smc - Saccharmyces cerevisiae (yeast)
srm - Serratia raarcescens (bacteria)
ssp - Schizosaccharomyces pombe (yeast)
tod - toad
trk - turkey

aFrom RTECS microfiche edition, October 1981 (Lewis and
 Tatken 1981).
      recent editions of RTECS may use abbreviations not
 included on this list.  Any abbreviation can be identi-
 fied by using the Key to Abbreviations in the Appendix
 of the RTECS edition being used.

-------
Table A-3.  Route of Administration Abbre-
            viations Used in RTECSa>b
          eye - eyes
          ial - intraaural
          iat - intraarterial
          ice - intracerebral
          icv - intracervical
          idr - intradermal
          idu - intraduodenal
          ihl - inhalation
          imp - implant
          ims - intramuscular
          ipc - intraplacental
          ipl - intrapleural
          ipr - intraperitoneal
          irn - intrarenal
          isp - intraspinal
          itr - intratracheal
          ivg - intravaginal
          ivn - intravenous
          mul - multiple
          ocu - ocular
          orl - oral
          par - parenteral
          rec - rectal
          skn - skin
          scu - subcutaneous
          unk - unreported
aFrom RTECS microfiche edition, October
 1981 (Lewis and Tatken 1981).
''More recent editions of RTZCS may use ab-
 breviations not included on this list.
 Any abbreviation can be identified by us-
 ing the Key to Abbreviations  in the Ap-
 pendix of the RTECS edition being used.

-------
                                      TECHNICAL REPORT DATA
                              (Please read Instructions on the reverse before completing]
1. REPORT NO.
 EPA  450/5-82-008
                                                                3. RECIPIENT'S ACCESSIOf*NO.
4. TITLE AND SUBTITLE
 Hazardous Air Pollutant Prioritization System  (HAPPS)
5. REPORT DATE
   October  1982
                                                                6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
 A. E.  Smith  and D.  J. Fingleton
                                                                8. PERFORMING ORGANIZATION REPORT NO.

-------
                                    TECHNICAL REPORT DATA
                             (Please read Inunctions on the reverse before completing)
 1, REPORT NO.

  EPA 450/5-82-008
                                                            3. RECIPIENT'S ACCESSION>NO.
 4. TITLE ANDSUBTITLE

  Hazardous Air Pollutant Prioritization  System (HAPPS)
              5. REPORT DATE
                October  1982
                                                            6. PERFORMING ORGANIZATION CODE
 7. AUTHOR(S)
                                                            8. PERFORMING ORGANIZATION REPORT NO.
  A.  E. Smith and  D.  J.  Fingleton
 9. PERFORMING ORGANIZATION NAME AND ADDRESS
   Argonne National  Laboratory
   9700 South Cass Avenue
   Argonne, Illinois  60439
                                                             10. PROGRAM ELEMENT NO.
              11. CONTRACT/GRANT NO.
               Interagency Agreement No.
               AD-89-F-1-344-0  '
 12. SPONSORING AGENCY NAME AND ADDRESS
   Pollutant Assessment Branch
   Office of Air Quality Planning  and  Standards
   U.S.  Environmental  Protection Agency
   Research Triangle  Park, N.C.  27711
              13. TYPE OF REPORT AND PERIOD COVERED
               Final
              14. SPONSORING AGENCY CODE
 15. SUPPLEMENTARY NOTES
 16. ABSTRACT
         This report  presents a preliminary screening  technique by which  a large
 number of potentially hazardous compounds can be numerically ranked  using readily
 available information on health effects and release to  the ambient air.   Factors
 considered are oncogenicity, mutagenicity, reproduction and developmental toxicity,
 acute lethality,  effects other than  acute lethality,  production volume,  vapor pressure,
 bioaccumulation  and  existing standards.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS  C.  COS AT I Field/Group
   Prioritization  System
Hazardous Air Pollutants
 8. DISTRIBUTION STATEMENT
  Unlimited
                                               19. SECURITY CLASS (ThisReport)
                                                Unclassified
                                                                          21. NO. OF PAGES
                               91
                                               20. SECURITY CLASS (Thispage}
                                                Unclassified
                                                                          22. PRICE
EPA Form 2220-1 (9-73)

-------