UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

                    WASHINGTON, D.C. 20460
                      16 DEC 198
                                                    OFFICE OF
                                           SOLID WASTE AND EMERGENCY RESPONSE
MEMORANDUM
SUBJECT:  ffjCDD,.Emissions for Municipal Waste. Combustors

FROM:     Michael B. Cook
          Dioxin Management C

TO:       Addressees
     We have recently completed an assessment of
tetrachlorodibenzo-p-dioxin  (TCDD) emissions from another
municipal waste combustion (MWC) facility sampled early
this year.  This is the sixth MWC plant sampled by the
Agency in its continuing program to evaluate the health risks
associated with emissions of TCDD from combustion facilities.

     EPA assessed TCDD emissions from the first five MWC
facilities sampled in a report dated November, 1981, entitled
"Interim Evaluation of Health Risks Associated with Emissions
of Tetrachlorinated Dioxins from Municipal Waste Resource
Recovery Facilities",  The report concluded that ". . . the'
levels of TCDD's from the five municipal waste combustors . .
do not present a public health hazard for residents living
in the immediate vicinity."

     The emissions of TCDD from the sixth plant were higher
than had previously been found.  We have nonetheless concluded
(p. 11) that ". . . in light of:  . . . conservative assump-
tions  . . . steps being taken . . ., the Agency does not
believe that this most recently sampled MWC represents.a
significant health concern ..."

     I have attached a copy of the November 1981 interim
evaluation and the recent assessment for .your use.   These
assessments may be shared, with interested members of the
public^                                                .

Atta.ciime.nfcs

Addressees:

Regional Dioxin Coordinators
Regional Solid Waste Branch Chiefs
"^gional Division Directors
  ffice of Pesticides and Toxic Substances
  :ate  of Virginia

-------
 12/8/83

 ASSESSMENT  OF  EMISSIONS  FROM A  RECENT  MUNICIPAL WASTE  COMBUSTOR
 Background
      In  the late  1970's  concern was  raised  in the United
 States regarding  the possible emission of trace amounts of
 highly toxic organic pollutants as a consequence of  large
 scale combustion.   Following suggestive  findings in  this
 country,  which essentially  confirmed reports from overseas
 where emissions testing  had first identified the presence
 of  chlorinated dibenzo-p-dioxins  (CDDs)/ particular  atten-
 tion was  directed to municipal  waste combustors  (MWCs).
      In  response,  the  Environmental  Protection Agency  (EPA)
 conducted a program which performed  sampling and analysis
 at  five  separate  MWCs.   The focus of these  studies was the
 emission of tetrachlorodibenzo-p-dioxins (TCDDs), with an
'emphasis on the specific isomer,  2,3,7,8tetrachlorodiben2o-p~
 dioxin  ( 2,3,3, 8-TCDD)..   This latter  compound is known  to
 be  quite toxic, even at  very low doses,  as  demonstrated in
 animal studies.  Documented evidence of  its presence in
 emissions evoked  special concerns.
      In  November,  1981,  the Agency published a report  entitled
 "Interim Evaluation of Health Risks  Associated with  Emissions
 of  Tetrachlorinated Dioxins From  Municipal  Waste Resource
 Recovery Facilities".   (EPA, 1981) The report presented
 upper limit estimates  of what the health risks might be to

-------
people living in the vicinity of the MWCs and concluded:



        "These estimates suggest that the present  emissions



        levels of TCDDs from the five municipal waste



        combustors described in this report  do not present a



        public health hazard for residents living  in the



        immediate vicinity.   In addition, the health risk



        estimates presented  in the assessment indicate that



        as long as emission  levels of TCDDs  do not greatly



        exceed the emissions measured at the five  US sites



        evaluated in this interim assessment, there should



        be no reason for concern.  This conclusion is valid



        for all toxicological effects (including reproductive



        and cancer) for which the available  animal and human



        data have been analyzed."



     In the pasc few months, data have been  generated by  the



Agency on the emission of TCDDs (and other pollutants) from



a sixth MWC.  These data and the supporting  contractor report



on the sampling and analysis are currently undergoing the normal



review procedure within the  Agency.



     The existence of these  "data and their qualitative indi-



cation of the presence of TCDDs in the emissions from this



MWC, however, have raised public concerns.  Therefore, in



order to give some perspective to these findings,  the Agency



is issuing this assessment employing the same procedures  used

-------
in the November, 1981 document.  These procedures incor-
porate a series of conservative assumptions which the
Agency believes tend to overestimate the risks due to TCDD
emissions.  If this "worst case" assessment projects risks
which are so low as to not present a health concern to people
living in the vicinity, then there is additional assurance
that the actual risk from the TCDD emissions should not form
a health concern to nearby residents.
     In sum, the purpose of this document is to project the
results from a sixth MWC on the scale generated by the results
from the five MWCs which were assessed in 1981, thereby providing
a basis for interpreting the significance of the new data and
'the efforts already underway to modify conditions at the plant.
     Note that the present document is being issued before the
final report on the sampling and analysis that underlie this
effort have been thoroughly reviewed.  Consequently, the
conclusions of this assessment are subject to changes that
might be necessitated by changes in the final report.
Overv Jew
     This document presents an assessment of the health
implications associated with the emission of TCDDs from a
recently sampled MWC.  The assessment is based on stack
emission data which were used to estimate the level of exposure
that people living near this facility might encounter, and

-------
on estimates of the health hazards that might be associated
with these emissions.
     Exposure information on the TCDDs was obtained by field
sampling of stack emissions, followed by chemical analysis
using gas chromatography and mass spectrometry (GC/MS).   The
actual amount of TCDDs from the stacks that would reach  people
living in the areas surrounding the plant was expected to be
so small that it would not be detectable by available
analytical techniques.  Therefore, the Agency used a mathema-
tical air dispersion model to estimate the ground level
concentration levels of TCDDs to which people were likely to
be exposed.
     Estimates of the risk to human health from these TCDDs
emission were obtained by extrapolating from animal data on
the carcinogenic and reproductive effects of 2,3,7,8-TCDD.
While the toxicity information on the other isomers of TCDD
is limited, there is reason to believe that none of. the  other
isomers are as toxic as 2,3,7,8-TCDD.
Hazard Assessment
     The reader is referred to other sources which discuss
the toxic properties of TCDDs in detail (Huff, 1980).  The
present document makes use of the same hazard assessment as
was used in the November, 1981 document.
Dose~Response Assessment
     The reader is referred to other sources for a discussion
                              4

-------
of the dose-response assessment that the Agency associates with



2,3,7,8-TCDD (EPA, 1980), based on a lifetime feeding study in



rats (Kociba, et al, 1978).   The present document makes use



of the same dose-response assessment as was used in the



November, 1981 document.



Exposure Assessment



     Table I contains information on the MWC facility and the



TCDD emissions detected there.



     In the present estimates,  the relation between the emission



data and the maximum concentration to which people in the



surrounding area are likely  to be exposed has been obtained



through a theoretical air dispersion model, PTMAX (EPA, 1977).



This computer program calculates the location and magnitude of



the maximum short term (1 hour) concentration in the area around



the stack.  The necessary input data are contained in Table I.



In order to obtain a maximum annual average ground level con-



centration, a reasonable assumption was made that the maximum



annual average concentration is 1/40 of the maximum hourly



concentration (Tikvart, 1981).   These results are found in



Table II.



Toxicity and Exposure Assumptions



     Ideally, there would be sufficient information compiled



during the Hazard, Dose-Response, and Exposure Assessments to



directly combine the data in a Risk Characterization step

-------
(Rational-Research Council,  1983).   However,  many unanswered



questions relating to the toxicity  of and exposure to these



TCDD emissions remain.  Since there are insufficient  data to



answer these questions definitively,  and because some type



of answers is needed in order to characterize the risk to



people breathing the emissions,  the Agency has adopted a



series of assumptions which are  designed to represent



"reasonable worst cases".  These are the same assumptions



used in the November, 1981 document.   Some of these un-



answered questions and related assumptions are presented



be low:



   Question 1



       What are the toxicological properties of the 21 TCDD



           isomers, other than 2,3,7,8-TCDD?



       Assumption 1



           The carcinogenic properties and reproductive



              effects of all TCDDs  are taken to be the same



              as that of 2,3,7,8-TCDD.






   Question 2



       How can the toxicological effects in humans be assessed



           in the absence of data in humans?



       Assumption 2



           The Agency has established methods (EPA, 1978) to



           address this question which include the following:

-------
           a.   Use of the no-threshold assumption for



               carcinogenicity.



           b.   Use of the most sensitive/  valid animal study.



           c.   Use of the linearized multi-stage model to



               generate an estimate of the upper limit of



               the excess cancer risk at  low doses.   The



               actual risk could be nearly any number



               between this upper limit and some lower



               number (possibly  zero).-



           d.   Conversion of animal dose  to human equivalent



               dose by. use of relative body surface  area.
Question 3
    Given the concentration and composition of TCDDs



        measured in emissions from the stack,  what are



        the resulting air concentrations and compositions



        at ground level to which people would  be exposed?



    Assumption 3



        The computerized PTMAX air dispersion  model and



           the factor used to convert to the annual con-



           centration is assumed to adequately represent



           the transport of the emissions to ground level.



           In lieu of a definitive analysis of atmospheric



           conditions, the result from the worst of six



           atmospheric classes modeled by the  computer is

-------
           assumed to be applicable.   The composition of



           emission products found at  ground  level is



           taken to be identical to the composition (but



           not the concentration) in the stack.



Question 4



    How do the TCDD contaminants in the air behave when



        they are breathed by humans?  (The TCDDs in the



        stack gases are generally associated  with parti-



        culate matter from which they are difficult to



        remove in the laboratory).



    Assumption 4



        Seventy-five percent of the inhaled particulates



           are assumed to be retained in the  body (ICRP,



           1968).  Further, 100% of the TCDDs (gaseous



           or particulate-bound) are treated  as  being



           biologically available to exhibit  a toxic



           response.



Question 5



    How often, for how long, and at what level will people



        be exposed?



    Assumption 5



        People are assumed to be exposed continuously to



            the maximum annual average ground level con-



            centration 24 hours/day, under the worst



            atmospheric conditions, for a 70  year lifetime,

-------
Health Risk Characteration



     Within the limitations of the assumptions discussed in



the previous section/  Table III contains the results of the



health risk characterizations for the upper limit of excess



cancer and for reproductive effects resulting from lifetime



exposure to the maximum annual average concentration of TCDDs



which are likely to be generated at the MWC.  The details of



these calculations are contained in the Appendix.



     The cancer risk is characterized by an "estimated



upper limit of excess cancer risk", which is expressed as a



probability.  For example, the upper limit of excess cancer



risk for the MWC, based on maximum total TCDDs, is 4.6 x 10~6,



This figure can be interpreted as the upper limit of the



excess cancer risk (probability) for an individual living



at the point of maximum annual average concentration of TCDDs



(resulting from emissions from the MWC) for 24 hrs/day, under



the worst atmospheric conditions, for a 70 year lifetime.



Alternatively expressed, this is a upper limit of risk of



46 in a 1,000,000 or 1 in 22,000.  That is, based upon the



assumptions above, the excess risk of contracting cancer is



likely to be something less than 1 in 20,000.  Again,



this is not a prediction of the risk but simply a



statement that the risk is not likely to exceed this level.

-------
     For comparison, the highest upper limit of excess cancer



risk reported in the November, 1981 document for total TCDDs



was 8 x 10~6.



     The reproductive effects risk is characterized in this



assessment by a "confidence ratio", which is the ratio of



the lowest level tested in animals divided by the anticipated



exposure level in humans.  Note that if this lowest dose



tested is seen as a "no effect level" (this point is currently



the subject of some scientific dispute), then the confidence



ratio would become the more familiar "margin of safety".



     For comparison, the lowest confidence ratio reported



in the November, 1981 document was total TCDDs was 30,000.





Conclusion



     The information in Tables I and II indicates that



compared to the situations at the five MWCs evaluated in



1981, the roost recently sampled MWC, when sampled, was emitting



greater amounts of TCDDs, resulting in higher ground level



exposures at the point of maximum impact (approximately .6



km from the stack under the worst atmospheric conditions).



Table III shows that, under the conditions prevailing at the



time of the test, the emissions represented a risk approx-



imately 6-fold greater than that seen at any of the MWCs



included in the 1981 survey.
                              10

-------
     A preliminary inquiry, into the design and operation of



this MWC has revealed a number of conditions that could be



contributing to the increased level of emissions.  Discussions



are already underway with responsible parties in the public  ,



and private sectors to determine appropriate corrective



measure that will likely lead to reduction into the emissions



of TCDDs.  In light of:



a.  the conservative assumptions made in this



    current assessment,



b.  the steps being taken to ameliorate the situation, and



•c.  the relatively short time span anticipated before these



    corrective measures are in place, the Agency does not



    believe that this most recently sampled MWC represents



    a significant health concern to people living in its



    vicinity.



    The Agency will continue to work with all parties concerned



to see to it that the planned changes in the facility and its



operations are carried out expeditiously and that a subsequent



re-sampling and analysis of the emissions is conducted



effectively and efficiently.
                              11

-------
                           TABLE I

         PLANT PARAMETERS AND EMISSION RATES FOUND AT
             A MUNICIPAL WASTE COMBUSTOR IN 1983

Parameters
Stack height                                    27.4 meters
Stack diameter                                  1.22 meters
Stack temperature                                271 °C
Flue gas flow rate^3)                             12 m-Ysec
Flue gas velocity                               11.4 m/sec
Ambient temperature                                4 °C

Average Emission Rate
Total TCDDs                              2.9 x 10~6 gram/sec

   (2,3,7,8-TCDD constitutes 21% of the total)

(a)	total from both stacks averaged over four tests
                              12

-------
                           TABLE II

      MAXIMUM ANNUAL AVERAGE GROUND LEVEL CONCENTRATION
  OF TCDDs CALCULATED1 AT A MUNICIPAL WASTE COMBUSTOR (MWC)
                           IN 1983.

Facility          Pollutant            Concentration

MWC              Total TCDDs             5.1 x 10~4 nanograms/m3

1 	 These values were generated through the air dispersion
      model PTMAX (EPA, 1977) with a correction factor of
      1/40 to convert to maximum annual average.  [Acknowledge-
      ment of the assistance by OPTS (Kinerson) and OANR
      (McGinnity)].
                              13

-------
                            TABLE III


   HEALTH RISK CHARACTERIZATION1 AT A MUNICIPAL WASTE COMBUSTOR
        IN 1983, BASED ON TOTAL TCDDs IN STACK EMISSIONS.

                        UPPER LIMIT OF           CONFIDENCE RATIO
  FACILITY            EXCESS CANCER RISK2    FOR REPRODUCTIVE EFFECTS3,

  Municipal Waste
    Combustor             4.6 x 10~5)

  1 	 These results cannot be effectively interpreted indepen-
        dent of the underlying conservative assumption upon
        which they are based.  (See text for further details)

        a.  All of the 22 possible TCDD isomers are assumed to
            have carcinogenic and reproductive effects properties
            equal to those of 2, 3 ,7,8-TCDD.

        b.  The established procedures for extrapolating from
            high dose to low dose and from animals to man are
            assumed to be appropriate.

        c.  The air dispersion model (with worst atmospheric
            assumptions) is assumed to be an effective method
            for extrapolating from concentrations in the stack
            emissions to concentrations to which people will be
            exposed.

        d.  The majority of inhaled particulate matter is assumed
            to be retained in the body and all of the particulate
            bound TCDD is assumed to be bioavailable.

        c.  Exposure to the annual maximum average ground
            level concentration is assumed to occur continously
            for 70 years.

  2 	 Using linearized multi-stage extrapolation model (EPA, 1978).
                           Lowest dose tested  =  1 ng/kg-d
  3 	 Confidence Ratio = Estimated human dose   Estimated human dose

a 	 For comparison, the highest value from the five previously
        tested MWCs was 8 x 10~6.
b 	 For comparison, the lowest value from the five previously
        tested MWCs was 30,000.
                                14

-------
                              APPENDIX

                      DETAILS  OF  CALCULATIONS
               (using  MWC  maximum data  as  an example)

        The  PTMAX  model  was  run using the  input  parameters  in

   Table I.   For general purposes,  however, the  emission

   rate  entered into the model was  1 g/s.  This  permitted easy

   scaling  to whatever specific emission rate  might be  of interest,

   since the model is  linear in mass emission  rate.

        Specifically,  for  the  MWC:

      a.  PTMAX showed that  with  1  g/s  the maximum hourly con-

         centration of  7  x  10"^  g/m3 was  obtained for  atmospheric

         stability class  I  (or A); that is, "unstable".

      b.  Applying the correction factor to estimate the annual

              maximum  average  concentration, we  obtain

                   7 x 10~6  g/m3  /  40 = 1.75 x 10~7 g/m3

      c.  Table I  indicates  that  a  total of 2.9  x 10"^  g/s  was

               the observed  emission rate  at the MWC.   Applying

               the factor  from b, we obtain as annual maximum

               average concentration:

(2.9  x 10~6  g/s)  (1.75 x 10~7  g/m3)=5.1 x  10~13  g/m3 =  5.1  x 1CT4 ng/ro3

                                                      (See  Table  II)
                                 15

-------
     The estimated upper limit of excess risk of cancer was



obtained using the unit risk factor developed by th'e Agency's



Cancer Assessment Group (GAG) for 2,3,7,8-TCDD (EPA, 1981).



   d.  Upper limit of excess cancer risk = (unit risk factor) (cone.)



             = [.091 (ng/m3)-1]  [5.1 x 10~4 ng/m3]



             = 4.6 x 10-5 (See Table III)



     The confidence ratio makes use of the data from the three-



generation reproduction study in the rat conducted by Murray



(Murray, 1979) and compares the lowest dose in that study  (1



ng/kg-d) to the estimated human dose derived from breathing



the dispersed emissions.



   f.  Estimated human dose =



                (Cone) x (Breathing rate) x (75% retention) /



                (Body mass)



            = (5.1 x 10~4 ng/m3) (20 m3/d) (.75) / (70 kg)



            = 1.1 x lO-4 ng/kg-d



   g.  Therefore, using results from f,



         Confidence Ratio =



                (Lowest dose in animals) / (Estimate human dose)



                          = (1 ng/kg-d) / (1.1 x 10~4 ng/kg-d)



                          = 9,100 (See Table III)
                              16

-------
                          References






EPA, 1977.   Guidelines for Air Quality Maintenance Planning



     and Analysis.   Vol. 10 (Revised):  Procedures for Eval-



     uating Air Quality Impact of New Stationary Sources",



     Pub. No. 450/4-77-001.



EPA, 1978.   Water Quality Criteria Development Guidelines,



     Cancer Assessment Group.



EPA, 1980.  "Risk Assessment on 2,4,5-T, 2,4,5-Trichlorophenoxy



     Acid,  and 2,3,7,8-TCDD",  Cancer Assessment Group, Sept. 12,



EPA, 1981.  "Interim Evaluation of Health Risks Associated with



     Emissions of Tetrachlorinated Dioxins  from Municipal Waste



     Resource Recovery Facilities," Office  of the Deputy Admini-



     strator, November.



Huff, J.E., J.A. Moore, R. Saracci, and L.  Tomatis, 1980.



     "Long  Term Hazards of Polychlorinated  Dibenzodioxins and



     Polychlorinated Dibenzofurans", Env. Health Perspectives



     26, 221-240.



ICRP (International Commission on Radiation Processes), 1968.



     Report of Committee IV, Publication No. 10.



Kociba, R.J., et al, 1978. Tox. Appl. Pharm., 46, 279-303.



Murray, F.J. et al, 1979. Tox. Appl. Pharm. , _5£, 241-252.



National Research Council, 1983.  Risk Assessment in the



     Federa1 Government;  Managing the Process, National



     Academy Press.



Tikvart, J. 1981.  Office of Air, Noise and Radiation,



     personal communication to Randolph Chrismon, 1981.



                              17

-------