United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-451/R-92-002
September 1992
Air/Superfund
AIR/SUPERFUND
NATIONAL TECHNICAL
GUIDANCE STUDY SERIES
Assessing Potential
Indoor Air Impacts
for Superfund Sites

-------
          ASSESSING POTENTIAL INDOOR
       AIR  IMPACTS FOR SUPERFUND SITES
            Contract No. 68D00124
             Work Assignment 1-76
                Prepared for:

     U.S.  Environmental  Protection  Agency
Office  of Air Quality Planning  and Standards
Research Triangle Park, North Carolina   27711
                September  1992
                Submitted by:

    Pacific  Environmental Services,  Inc.
        560 Herndon  Parkway,  Suite  200
          Herndon,  Virginia   22070

-------
                               DISCLAIMER
     This report was prepared for the U.S. Environmental  Protection
Agency by Pacific Environmental Services, Inc., Herndon,  VA,  under
Contract No. 68-00-0124, Work Assignment No. 1-76.  The contents are
reproduced herein as received from the contractor.  The mention of
product names or trademarks are not intended as endorsements  of the
products or their use.  The opinions, findings, and conclusions ex-
pressed are those of the authors and do not necessarily reflect those of
the U.S. Environmental Protection Agency.
                                   n

-------
                            TABLE OF CONTENTS

 DISCLAIMER	ii
 TABLES	iv
 FIGURES	iv
 1.0   INTRODUCTION	1-1

      1.1  BACKGROUND	1-1
      1.2  PURPOSE	1-3
      1.3  SCOPE	1-3

 2.0   INDOOR AIR  IMPACT ASSESSMENT PROCEDURES	f -.  ..   2-1

      2.1  GENERAL CONCEPTS   	   2-1
      2.2  SCREENING LEVEL MODELING   	   2-3
          2.2.1  SIMPLE CONSERVATIVE MODELING  	   2-3
          2.2.2  SCREENING LEVEL 2 MODELING  	   2-5
      2.3  ON-SITE EXTERIOR MONITORING 	   2-7
          2.3.1  DESIGNING THE MONITORING PLAN	2-8
          2.3.2  BUILDING INSPECTION  	 2-11
          2.3.3  ESTIMATION OF INDOOR AIR CONCENTRATIONS  	 2-12
      2.4  INDOOR MONITORING  	 2-13
          2.4.1  SCREENING LEVEL MONITORING  	 2-14
          2.4.2  REFINED MONITORING FOR INDOOR AIR
                  IMPACTS OF SITES	2-16
          2.4.3  MAKING LONG-TERM ESTIMATES FROM
                  SHORT-TERM MONITORING DATA   ..... 	 2-26

3.0  HAZARD ANALYSIS	3-1

     3.1  INCREMENTAL INDOOR AIR RISKS  	  3-1
     3.2  RISK COMMUNICATION	3-3
     3.3  OVERVIEW OF RISK ASSESSMENT GUIDANCE  	  3-7
          3.3.1  EXPOSURE	3-8
          3.3.2  TOXICITY ASSESSMENT  	  3-9
          3.3.3  SOURCES OF TOXICITY VALUES 	 3-10
          3.4.4  QUANTIFYING RISKS  	 3-11
          3.3.5  DATA LIMITATIONS AND INTERPRETATIONS	3-13

     APPENDICES
       A - PREDICTIVE SCREENING TECHNIQUES
       B - MONITORING METHODS
       C - CASE STUDIES
                                  in

-------
                             LIST OF TABLES
Table                                                             Page

 2-1     Hypothetical Monitoring Data 	  2-25
 2-2     Hypothetical Data Corrected for Ambient Air	2-25
 2-3     Hypothetical Data Corrected for Soil Gas Ratio	2-25
                             LIST  OF  FIGURES


Figure

 3-1     Illustrative Decision Tree ....
                                   iv

-------
                            1.0  INTRODUCTION

 1.1    BACKGROUND
       The  Comprehensive  Environmental Response, Compensation, and
 Liability  Act  of  1980  (CERCLA  or "Superfund") and  its reauthorization in
 the Superfund  Amendments  and Reauthorization Act (SARA) of 1986 estab-
 lishes a national  program for  responding to releases of hazardous
 substances into the environment.  The overarching  mandate of the
 Superfund  program  is to  protect human health and the environment from
 current and  potential  threats  posed by these releases.
       Occupants of buildings proximate to a site where such releases
 have occurred  may  potentially  be exposed to the released chemicals from
 their  transport into the  indoor environment.  Typical pathways for entry
 of site contaminants into a building include transport through the
 ambient air, the use of  contaminated groundwater,  seeps of non-aqueous
 liquids and  groundwater  through building exteriors, and intrusion of
 contaminated soil  gases.   In some cases, the resultant indoor air
 concentrations of  site related chemicals may be high enough to warrant
 immediate  corrective action, while in others they  may be inconsequen-
 tial.
       Responses taken  to  assess whether or not an  indoor air impact has
 occurred and the significance of that impact have  varied from predictive
 modeling to  on-site monitoring of the indoor air quality.  Both the
 modeling and monitoring techniques used have been  quite varied.
 Predictive modeling techniques used have ranged from quite simple
 screening  tools to complex approaches.  Although guidance exists for
 selection  of air emission and dispersion models and information docu-
ments  are  available for indoor air modeling, no established guidance
 exists to  assist with proper selection of predictive models for assess-
 ing potential  indoor air  impacts.  Likewise, monitoring has varied from
 collecting grab samples to complex multi-building comparative studies.
Although guidance exists  for selection and use of monitoring techniques
                                   1-1

-------
for ambient air  and  soil  gases  and  information documents are available
for indoor air monitoring,  no established guidance exists to assist with
proper selection of  methods to  assess the potential site related indoor
air impact.
      The indoor environment, even  for buildings not impacted by outdoor
pollutants, is recognized to be highly polluted.  Many indoor studies
have shown that the  normal  indoor environment contains a wide variety of
pollutants at higher concentrations than found outdoors.  The sources of
these pollutants are believed to be primarily consumer products.  Indoor
pollution can occur  from  cleaners and waxes, paints, pesticides,
adhesives, cosmetic  and personal care products, hobby supplies, furnish-
ings and clothing, building materials, and heating and cooling systems,
among others.
      Many of the chemicals emitted by the materials are the same as
chemicals released from Superfund and hazardous waste sites which
severely complicates any  attempt to determine the amount of any indoor
air pollutant that is directly  attributable to external sources.
Further complicating the  situation  is the fact that buildings interact
with the outdoor environment.   The  rate at which ambient air infiltrates
a building is a function  of several factors including wind speed,
indoor-outdoor temperature  differences, height of the building, and
leaks in the building shell, as well as vented equipment, such as
bathroom and kitchen fans or oil and gas furnaces and fireplaces.  Air
pressures below ambient can develop in the lower stories of a building.
Although these pressure differences are small, typically between 1 and
10 pascals, they can result in  the  building literally sucking in soil
gases through cracks and  openings in below grade walls.
      Site contaminant location in  relation to building and contaminant
source strength may  also  affect the rate of chemical entry into the
building.  Contaminant may  arrive in the ambient air, in groundwater
leaking into the building,  from gases migrating laterally from the site,
from nearby contaminated  soils, and from contaminated groundwater plumes
passing near or beneath the building.  The amount of chemical arriving
at the building then becomes a  complex function of meteorological
conditions and soil  properties.
                                   1-2

-------
      Determining  the  impact  a nearby contaminated site may be having on
 the  indoor  air  quality of  a specific building can, thus, be seen to be a
 difficult task.

 1.2   PURPOSE
      This  manual  for  assessing  indoor air impacts was developed to be
 used for buildings proximate  to  Superfund sites.  The procedures and
 methods may also be  applied to other assessments of hazardous wastes and
 hazardous materials.   The  purpose is to provide information on methods
 and techniques  that  can be used  to assess the potential or actual impact
 a Superfund site may have  on  the indoor air quality of buildings
 proximate to the site  in order that the risk to the occupants from this
 source may  be estimated.   The manual is designed to provide the tools
 needed to assist in  the identification and resolution of indoor air
 quality concerns at  Superfund site.

 1.3   SCOPE
      This  manual  focuses  on contaminant entry into the indoor environ-
 ment via subsurface  gases  and ambient air.  It does not address contami-
 nant entry  via direct  intrusion of contaminated groundwater or indoor
 uses of contaminated groundwater such as showering.  It also does not
 address radiologically contaminated sites, which,  although an important
 consideration for some sites, is believed to be adequately covered by
 existing guidance for  radon.
      It provides summary  level  information on predictive modeling,  on-
 site monitoring, and risk  characterization.  It contains information
 from indoor air studies that have been conducted using both modeling and
monitoring  approaches.  This information was utilized to provide general
 information on the selection and use of models,  monitoring methods,  and
assumptions useful in  assessing site impacts on indoor air quality at
levels ranging from  scoping to detailed monitoring.
      Individuals having different levels of scientific training and
experience  are likely  to use this manual.   Because assumptions and
judgement are required  in many parts of the assessment, the individuals
conducting  the evaluations are key elements in the process.  The manual
                                   1-3

-------
is not intended to instruct non-technical persons how to perform
technical evaluations, nor to allow professionals trained in one
discipline to perform work in another.
      The manual cannot address all circumstances.  Users must exercise
technical and management judgement, and should consult with appropriate
regional and headquarters staff when encountering unusual or particular-
ly complex technical issues.  The procedures described should be viewed
as flexible and can and should be tailored to specific circumstances and
information needs of individual sites, and not as a rigid approach that
must be conducted at every site.
                                   1-4

-------
              2.0    INDOOR AIR IMPACT ASSESSMENT PROCEDURES

       This  Chapter presents  general  procedures  for  assessing potential
 Superfund site  impact on indoor  air  in  structures proximate to the  site.
 It  is  recognized  that the contaminant source  type,  location, proximity
 to  the structure(s)  and transport pathways, as  well as the geology,
 hydrogeology, structural  characteristics, and use of the structure(s)
 differ widely for all  potential  situation that  may  be encountered.
 Thus,  this  discussion,  of necessity, is generic.  Although the proce-
 dures  should be generally applicable, they do not constitute a "how-to"
 or  "cookbook" set of instructions.   It  is expected  that proper applica-
 tion will require substantial expertise and professional .judgement.

 2.1    GENERAL CONCEPTS
       The objective  is  to assess the potential  exposure, and resultant
 risk,  for occupants  of  buildings proximate to a Superfund sites to site
 contaminants via  the indoor  air pathway.  To accomplish this objective,
 various techniques ranging from simple screening models to complex
monitoring may be used.   The confidence level  in the results increases
with increasing sophistication.
       In recognition that conducting monitoring for buildings of a
 sophistication level sufficient to segregate site impacts from other
potential sources can be  expensive,  the approach outlined is based on a
graduated approach to scope the possible extent and magnitude of the
possible impact before  on-site detailed monitoring is  executed.   The
approach follows these  general steps:
      1.    Conduct simple conservative modeling to estimate the poten-
            tial magnitude of the exposure.
      2.    If step  1 suggests a potential  problem,  conduct more sophis-
            ticated modeling to provide more realistic estimates of the
            potential exposure.
                                  2-1

-------
       3.     If  Step  2  indicates  a  significant exposure potential
             exists,  conduct  exterior monitoring at the site to confirm
             model  predictions  and  make better estimates of indoor expo-
             sures.   This  step  may  be preceded by screening level indoor
             monitoring.
       4.     When deemed necessary, conduct monitoring at the building
             site at  a  level  adequate to provide the best estimate of
             site impact on the indoor air quality and collect data to
             estimate exposure  and  resultant risk over longer time peri-
             ods.
       Step 1, simple conservative  modeling, is suggested in virtually
all cases because  it provides  overall perspective and can reduce the
potential for conducting  unnecessary monitoring and reduce the possibil-
ity of focusing efforts based  on complaints which may or may not
represent the locations of most concern.   Step 2,  sophisticated model-
ing, is suggested  as a follow-up to Step 1 because simple models
typically over predict exposures.  Executing Step 2 reduces the likel-
ihood of conducting monitoring when, in fact,  no exposure of concern may
be occurring.
      Step 3, exterior monitoring  at a specific building location,  is
suggested as the primary  on-site technique for assessing the site
related indoor air impacts on  specific structures.   These techniques
allow reasonable estimates of  indoor air concentration without the
necessity of dealing with building specific air flow patterns and indoor
pollutant sources.  If it is considered desirable to first conduct
indoor monitoring  at the  screening level,  it should be recognized the
indoor data are useful for order of magnitude  estimates only.   This
monitoring cannot, except in unusual cases, provide pollutant concen-
trations due to site impacts because of the variability in types and
strengths of indoor sources.
      Step 4, detailed building monitoring, is suggested only when it is
necessary to obtain actual indoor  air site impact data for a specific
building.  It is expected that this monitoring will  typically only
provide marginal improvements  in the exposure  estimates obtainable by
                                   2-2

-------
 following Step 3.   Indoor air monitoring itself is complex and requires
 a high level  of professional expertise to obtain reasonable results.
 Except in unusual  cases, site impacts cannot be estimated unless
 exterior monitoring is executed simultaneously.
       The above general  procedures are applicable principally to indoor
 air impacts associated with transport of organics (volatile and semi-
 volatile) by  soil  gases  and ambient air.   Impacts of particulates
 transported to the building may,  however,  be best resolved with direct
 indoor monitoring.  This is suggested because,  once inside the struc-
 ture,  dusts tend to settle and then be resuspended by traffic and
 cleaning activities.   Potential  risks from dusts may then be a combina-
 tion of inhalation,  ingestion, and dermal  contact paths.

 2.2    SCREENING LEVEL  MODELING
       The use  of screening models  corresponds to Step 1 and 2 above.
 They are used  to obtain  order-of-magnitude estimates  of the potential
 indoor air concentrations.   The  simple modeling approach  described  in
 Section  2.2.1  should be  viewed only as a  scoping tool.  The modeling
 tools  discussed in Section 2.2.2 can be,  and have been, used to estimate
 the risks  due  to the indoor air pathway in  remedial  investigations.

 2.2.1  Simple Conservative  Modeling
       Modeling  may be  used  to  examine  the  possibility  that  site contami-
 nants  may  be transported to  buildings  in  sufficient concentration and
 rate to  significantly  impact the indoor air.  This  is  a particularly
 useful  step when more  than  one structure or type  of structure  may be
 impacted due to  a  fairly large contaminated zone,  such as a  groundwater
 plume  or dispersion in ambient air.   Information  may already be  avail-
 able to determine  whether  or not a particular transport mechanism is of
 concern and no modeling  would  be needed for these mechanisms.
      A minimal  amount of  site specific information is needed  to run the
 simple  models.   For emissions  from contaminated groundwater,  the
chemical concentration in the  groundwater and depth to the top  of the
water table will suffice.  Similar emission related information  is
needed  for contaminants  transported  through ambient air.   Because these
                                   2-3

-------
modeling  results should be  only used  to determine whether or not more
sophisticated modeling is needed,  missing  data may be supplied using
realistically conservative  estimates.
       It  is  recommended that  the air  quality models SCREEN and/or
TSCREEN be used  for estimating  the ambient air concentrations at the
building(s)  being considered.   These  models will calculate the short-
term maximum concentrations at  various  distances from the source.
Because buildings typically have air  exchanges rates of 0.5 to 1 per
hour,  it  is  recommended that  averaging  periods less than 3 hours not be
used.  The use of the  24-hour maximum concentration should provide a
conservative estimate  even for  developmental toxicants.  Professional
judgement should be used when estimating possible impacts for very
short-term releases that could  occur  from  sudden releases or remedial
actions.
       It  is  recommended that the simple  Fickian diffusion model (Appen-
dix A, Section A.1.1)  be used to estimate  the contaminant flux from
contaminated groundwater and contaminated  soil.  The important variables
are chemical  concentration and  distance  from the building floor to the
contamination  because  diffusivities in  air and Henry's Law constants
each vary by less  than an order of magnitude for VOCs.  Assume 100
percent of the soil  gas  rising  under  the building footprint enters the
building, for  all  building types,  and the  building air exchange rate is
0.5 per hour.
      A first  cut  approximation  can be made without even running the
model by assuming  diffusivities  in air  =  10"5, Henry's  Law constants =
10"1  (dimensionless), air filled and total  porosities  both equal 0.4,
building air changes per hour =  0.5, and 100 percent of the soil  gas
rising under the  footprint of the  building enters.   Under these assump-
tions,
                                   2-4

-------
 Where
             IAC   =      Indoor  air  concentrations, M9/m3
             CL     =      Concentration  in  groundwater, /jg/L
             L      =      Depth to water table, M.

 Note that  this  is  the  upperbound concentration  for the most volatile
 component  in very  porous  completely dry soil and  should be expected to
 yield  excessively  high  results  for  less volatile  compounds and for
 tightly  packed  or  wet  soils.  The equation  is of  little utility when the
 pollutant  source is within about 2  meters of the  building or when direct
 intrusion  of contaminated groundwater  is  occurring.
       For  contaminated  soil, the relationship given in Equation 11 of
 Appendix A,  Section A.1.3 may be used  to  estimate the soil gas concen-
 tration.   For contaminated soil immediately adjacent (ie, within 1
 meter) to  the building, the relationships of Appendix A Section A.3.1
 are recommended to calculate the indoor air concentration.
       Note in particular that it is not recommended to estimate a
 reduction  in soil gas flux entering the building based on the ratio of
 assumed area of floor cracks to floor  area.  Based on current under-
 standing, this  is a questionable assumption (See Appendix A,  Section
A.3).

 2.2.2  Screening Level 2 Modeling
       Indoor air concentrations calculated using the more sophisticated
modeling techniques provide the best estimate that can currently be made
without sampling at the structure(s) in question.  However,  use of these
models requires both professional  modeling expertise and site-specific
information.  The better the site-specific information,  the more likely
are the model predictions to accurately represent the concentrations in
the structure.   Of all  the terms in the various equations of Appendix A,
the effective diffusivity (DTeff) and soil  permeabilities (KJ  are  likely
to have the most uncertainty.   This is because not only are soil
properties quite variable vertically and horizontally over regions as
small  as a single house, soil  moisture contents (and thus air filled
porosity) can vary widely.  It is  not unusual  for Kv  to  vary  by 3  orders
                                  2-5

-------
 of magnitude  across  a site the size of a typical  residential  lot.   While
 the use  of "average"  or "typical"  values for these  parameters may yield
 fairly accurate  results,  it should not be expected  that  any current
 model will  exactly  represent the real  world.
      As  stated  in Appendix A,  Section A.2,  air modeling  should  be
 conducted  in  accordance with the guidance referenced.  For emissions
 from groundwater and  contaminated  soils,  any of the models of Appendix A
 Sections A.I  and A.3.3  may be used provided  that  appropriate consider-
 ation is given to the contaminant  concentration in the soil and  site
 conditions  affecting  the  effective diffusivity.
      Specifically, the model  of Section  A.1.1 may be used for groundwa-
 ter when the  refinement of Section A.1.2  is  made.  The model may also be
 used for contaminated soil  when  corrected  for soil adsorption (Section
 A.1.3).  Both uses of the model  should be  corrected using the consider-
 ations of Section A.1.6 when neat  liquids  (non-aqueous phase), or soil
 concentrations exceeding  the level  that  calculated soil gas concentra-
 tions are greater than  compound  vapor  pressures are present.
      The model  presented in section A.1.4 has a wide range of applica-
 bility for  contaminated soils  and  groundwater.  It is the preferred
 model for calculating soil  gas flux for most  applications.  However, the
 model is not  suitable for situations  in which non-aqueous phase  liquids
 or  high chemical  concentrations,  as explained above, are present.   For
 these applications, the model  presented  in Section A.1.6 is most
 appropriate.
      All of the  Models of Appendix A, Section A.I, estimate only the
 contaminant flux  (mass  of contaminant  impacting on the structure per
 unit area of below surface  walls per unit  time).  They do not, there-
 fore, directly yield an estimate of indoor air concentration.  To obtain
 the  indoor  air concentration, divide the calculated flux by the air
exfiltration rate for the building  (see Section A.3.2).  It is explicit-
 ly  recommended that the calculation be based  on 100 percent of the  flux
entering the building (ie.,  no correction  made for building resistance
to  intrusion, such as the fraction  of  floor  area cracked).
      The Johnson and Ettinger models  presented in Appendix A, Section
A.3.3,  appear to  provide  a  good  representation of both contaminant
                                   2-6

-------
 transport and the effects of building characteristics  on  soil  gas  entry
 into a building.   Properly coupled with  appropriate  calculations of the
 contaminant source gas phase concentration  and  equations  for  building
 air  exchange rates,  the models show promise of  becoming the best
 available short  of detailed numerical  simulations.   The use of this
 model  is  considered  appropriate for indoor  air  impact  assessments.

 2.3    ON-SITE EXTERIOR MONITORING
       In  this section  the value of,  ajid  approach to, the  use  of monitor-
 ing  in the  immediate vicinity of a potentially  impacted building and how
 this data may be  combined with models  to yield  an estimate of  the  indoor
 impacts from Superfund chemicals is  discussed.  It is  expected  that this
on-site monitoring would  be  conducted
as a consequence of the modeling
described  above  indicating  the  potential for significant impact.   It is
recognized that  on-site  monitoring may also be indicated by other
considerations.
      The  accuracy of the values calculated from modeling is a function
of the quality of input  data.   Generally, several assumptions will have
been made  in these calculations that affect the accuracy.  The principal
assumptions are:
      1.    Soil gas concentration at the soil-building interface  is
            zero.
      2.    The  effective vertical diffusivity is accurately represented
            by the data  input for total and air filled porosity.
      3.    Effective vertical diffusivities are the same throughout the
            entire soil  column extending an infinite distance from the
            building.
      4.    Soil gas concentrations at the contaminant source are accu-
            rately represented by the equilibrium relationships given in
            Appendix A for chemicals dissolved  in groundwater or ad-
            sorbed to soil.
      5.    No adsorption or biological reaction occurs as the gas dif-
            fuses to the building.
      None of these are  likely to be strictly true.   For example, if
assumption 1  were strictly true,- no chemicals would  enter the building.
                                   2-7

-------
Exterior monitoring provides the data needed for the best estimate of
chemical concentrations entering the building.  These data are necessary
if a reasonable estimate of Superfund site chemicals impacting the
building are to be made.  Note that because the typical indoor air
contains such a large number of chemicals that are also found at
Superfund sites, simply knowing the indoor air concentrations usually
only establishes the upper bound for site impact but does not establish
that the chemicals are actually from the Superfund site.  Reasonable
exceptions to th'is might be for chemicals not found in non-impacted
buildings (eg., vinyl chloride) or chemical concentration exceeding the
upper end of the typical range.
      Since assumption 3 is typically not true, especially in the region
near the building, it should be expected that soil  gas concentrations,
and possibly compositions, will vary from location to location around
the building.  Thus, the monitoring plan must be designed to take this
into account in so far as practical.
      The building itself exerts an influence on the surrounding soil.
Because of the slight negative pressures (ie., 1 to 10 Pa) developed due
primarily to thermal gradients and wind effects, the building may draw
in gases from the surface as well as from several  meters laterally from
the building walls.  Thus, in permeable soils, soil gas concentrations
adjacent to basement walls, in particular, may be lowered due to
dilution from surface gas. (Note that this also implies that volatile
chemicals applied adjacent to the building, such as pesticides applied
to shrubbery, may be rapidly drawn into the building.)

2.3.1 Designing The Monitoring Plan
      This section assumes that impacts on the building indoor air occur
from both transport through ambient air and intrusion of soil gas.  The
plan should be developed considering the likely magnitude of the impact
from each source.   The procedures for ambient air would quantify
chemical concentrations at the building arising from the Superfund site
as well  as any other upwind emission sources.  If ambient air transport
is of concern and other potential emission sources exist, upwind-
                                   2-8

-------
 downwind sampling may be necessary.   For guidance  see  Volume  II  of the
 Air/Superfund NTGS series (EPA-4501-89-002a).
       Ambient Air
       Monitoring ambient air for  Superfund  site  impacts  is only  valid
 when  the wind direction is  from the  site toward  the  building  being
 monitored.   Monitoring equipment  should  be  placed  in a location  on the
 windward side of the building  as  free  as practical from  localized
 sources  of  emissions,  such  as  lawn mowers or automobiles, and wind
 shields,  such as out-buildings  or shrubbery.  Monitor  intakes should
 normally be about five feet above ground level (about midpoint of
 building ground  floor) and  about  5 to  15 feet from the building  wall.
 Duplicate monitors are recommended.  Air samples should  be collected
 over  a period of not less than  one-hour  and preferably over an eight
 hour  period.   Wind speed and direction should be recorded continuously
 or at, at least,  30 minute  intervals.
      The sampling and analytical methods to be used are dependent on
 the known or  suspected emissions  from the Superfund site.  For most
 applications,  EPA methods TO-1, TO-2, or TO-14 will be adequate  for
 organics.   Particulates (inhalable) may  be collected using Method  IP-10.
 Support  in  selecting the methodologies should be obtained from the
 appropriate EPA  Division.
      Soil  Gases
      As  stated  previously,  it  is to be  expected that soil gas concen-
 trations  will  vary with location  around  the building due to,  among
 several  effects,  soil  inhomogeneity.  Thus,  it is necessary to measure
 the soil  gases at  various points  around  the building.  It is  recommended
 that soil gas  probes be used to measure  soil gas concentrations.   The
 use of flux chambers is  not  recommended.  [Flux chambers located within
one to two  meters  of the  building may give significantly negatively
biased results if  building underpressurization is exerting an effect on
soil  gas  flow  rates  and  flow directions.  Low permeability zones near
the surface,  frozen  ground, or wet surface soils (eg.,  from recent
rains), may also result  in low flux chamber results.   Flux chambers
located further from the  building may or may not realistically represent
                                   2-9

-------
 the  diffusive  flux reaching the region  of  the  building  subfloor  due  to
 inhomogeneity  of soil  properties around the  building.]
       It  is  recommended that at least two  soil probes be  installed on
 each side of the building.   Statistical  considerations  presented  in
 Appendix  C of  Volume  II of  the  Air/Superfund NTGS series  indicate this
 would  be  adequate to  determine  the  average concentration  within  20
 percent at a 95  percent confidence  level.
       The objectives  of soil  gas monitoring are to determine the  soil
 gas  concentrations and the  soil  permeabilities in the zone very near the
 bottom of the  building foundation.  The  probes should be  installed so
 that the  probe tips are between  one-half and one meter  of the building
 floor  or  basement wall.   For slab-on-grade and crawl-space type con-
 structions,  the  probes should be installed on an angle  to go under the
 building.
       It  is  recommended that  slightly undersized pilot  holes for the
 probes be made with an auger.  At this time a sample of soil from near
 the  bottom of  the pilot hole  should be obtained and retained for
 analysis  needed  to estimate  the  effective diffusivity of soil near the
 building.  (Analyses would  include bulk  density,  moisture content, and
 soil characterization,  eg.,  sandy loam).
      A volume of soil  gas  equal  to at least two probe volumes should be
 drawn through  the probes  before  samples  are taken.  Gas flow rates
 should be  low  (eg., 10 to 100 cm3/min.)  to  reduce  the  possibility of
 establishing unwanted  pressure gradients.
      The  exact  sampling  and analytical  methods used will depend on the
 contaminants expected  from  the Superfund site.   Assistance should be
 obtained  from  the appropriate EPA Division.  The following should be
 considered when  selecting these  methods.
      •     The  indoor air  concentrations will  likely never exceed five
percent of the soil gas concentration.  Thus, very low detection limits
 are  not required.
      •     A  portable  GC will be useful to determine the magnitude of
 soil  gas concentrations.  This or other on-site measurements will likely
be necessary if  adsorption tube  techniques  (eg.,  T01 and T02) are used
to ensure the  capacity  of the tubes is not  exceeded.
                                  2-10

-------
       •      The SUMMA canister technique  of  TO-14  may  likely be the best
 technique where prior data on chemical  compositions  and  concentrations
 are  not  available.
       After soil  gas  samples are collected,  the  probe  should be used to
 determine the  soil  permeability to  gas  flow.  The  method is  quite
 simple.   In brief,  a  regulated flow of  compressed  air  is passed through
 the  probe into the  soil  while measuring the  flow rate  and probe pres-
 sure.  Pressure gauges (0-125Pa)  and flow meters with  range  capabilities
 of 5 to  450 cm3/nrin are  needed.  The permeability,  which  assumes Darcy
 flow,  is  calculated from:
                               K-
                                     0V-
where,
      Kv    =     Permeability, M2
      Q     =     air flow rate, M3/s
      H     =     viscosity of air,  1.83 x 10s Kg/m-s
      r     =     internal radius of probe, m
      Pa    =     pressure in pascals

Determining the soil permeability is important as it is an indicator as
to whether soil gas intrusion is by  diffusion or convection.
2.3.2 Building Inspection
      A general inspection of the building should be conducted.  The
purpose of the inspection is to assess building construction details
that can be used to judge reasonableness of parameters used to calculate
final indoor air concentrations.  Thus, of particular interest are
building size (area of building footprint and area of below grade
walls), construction type (slab-on-grade, crawl-space, or basement,
basement wall  construction type - poured concrete or hollow block - ,and
number of stories),  presence of obvious cracks in floors or walls in
contact with soil, and details useful for estimating building air
exchange rates (see Appendix A,  Section A.3).  The inspection forms
given in EPA 400/1-91/-033 and EPA 400/3-91/003 are recommended.

                                  2-11

-------
 2.3.3  Estimation  of Indoor Air  Concentrations
       It  can  be assumed  that  the  soil  gas concentrations near the
 building  are  at steady-state.   That  is,  soil gas is diffusing to the
 zone of building  influence at the  same rate that it is entering the
 building.   Based  on the  empirical  evidence for radon intrusion and
 predictions of the  Johnson and  Ettinger  Model (see Appendix A, Section
 A.3),  this  should result  in an  estimate  of contaminant entry that is
 conservative, but not  strongly  so.
       Data  analysis should begin with  a  review of soil  permeabilities.
 In regions where  permeabilities are  greater than 10~8 cm2, soil gas
 flows  into  the building  primarily  by convection and at a rate directly
 proportional  to the permeability.  At  permeabilities much less than 10"8
 cm2,  intrusion is primarily by diffusion and is  independent  of the
 permeability.  Thus, for  permeabilities  of 10"8 cm2, or greater, the
 empirical relationship for radon presented in Appendix A, Section A.3.1,
may be used to estimate the indoor air concentration.   A more precise
 estimate  for  any  permeability may  be obtained using the Johnson and
 Ettinger  model.
      When using  this model in this  context, it is important to remember
that the  "source" is at the location soil-gas samples  were collected
 (ie,  LT = 0.5m)  and the effective diffusivity is  for the  region  between
 sampling  points and  the building walls.  Also, the entire area of the
building  in contact with  soil (floor plus below grade  walls) should be
used in the calculation.
            It is possible to perform  internal consistency checks based
on indoor air concentrations calculated  by either method.  Using the
calculated indoor concentrations, building air exfiltration  rate, and
area of exposed walls, calculate the estimated flux.  Calculate the soil
gas concentrations  at the actual source  (eg. top of water table for
groundwater sources) using the relationships of Appendix A,  Section A.I.
Using the measured  soil gas concentrations and these data,  calculate the
effective diffusivity using the appropriate emission model.   This
calculated diffusivity should be reasonable when compared to the
diffusivity calculated from known or reasonable estimates of soil
properties.
                                  2-12

-------
 2.4    INDOOR  MONITORING
       There are  various  levels  at which  indoor  air monitoring to  assess
 Superfund  site impact can be conducted.   It should be clearly understood
 at the  outset, however,  what the utility  of each  level  is and what
 conclusions may  be  reasonably drawn from  the data.
       The  most simplistic approach is to  conduct  monitoring only  inside
 the  structure.   Unless monitoring is being conducted for a specific
 compound from the Superfund site that is  not found in non-impacted
 indoor  environments  (eg., vinyl chloride), it matters little what
 monitoring technique is  used so long as the technique has a detection
 limit  for  the target compounds  in the mid-to upper-end of the typical
 range  for  non-impacted buildings.  This is because only two outcomes are
 possible.  Either the concentrations are  above the maximum reported
 concentrations for non-impacted structures, in which case it can be
 reasonably concluded that some part (unknown) of  that concentration is
 due  to  the Superfund site, or the concentrations  are below this maximum,
 in which case all that can reasonably be concluded is that if Superfund
 site chemicals are entering the structure, they are not causing the
 concentrations to exceed the typical  range.  In the latter case, it
 specifically  cannot be concluded that none of the target chemicals
 identified are from the Superfund site.   Thus,  monitoring only the
 indoor  air can,  at best, demonstrate that the Superfund site has a major
 or minor impact on the indoor air quality.  At worst, it provides
 information that the site may be contaminating the structure but not
 enough  information to prove the impact is or is not occurring.
      Another indoor only monitoring technique that has been used is to
monitor at the suspected impacted structure and at "control" structures
 known or suspected not to be impacted.   The objective is to obtain
 "typical"  concentrations for non-impacted structures for use as a
 "background"  correction.   Because the number of structures typically
monitored does not provide a good statistical  sample of the population,
data obtained  this way generally provide marginal, if any,  improvement
over using the means and ranges from larger studies.
      Combining  indoor air monitoring with monitoring the potential
pathway(s)  (eg.,  soil gases  and ambient  air)  for Superfund site impacts
                                  2-13

-------
 is the preferred method  to  assess the  impact of the site on the indoor
 air quality of  specific  structures when  it is necessary to quantitate
 the impact through  indoor air monitoring.  The procedures, discussed in
 more detail below,  allow refined estimates of the instantaneous concen-
 trations and data which  can be used to make estimates of the long-term
 concentration averages for  use in risk assessment.

 2.4.1 Screening Level Monitoring
      Screening level monitoring can be used to ascertain whether or not
 very high levels (relative  to typical) of indoor air pollutants exist
 and to provide estimates of the indoor air concentrations to guide
 design of more refined indoor monitoring.  Before this type monitoring
 is conducted, decisions  should be made as to the appropriate follow-up
 action to be taken  depending on the concentration levels determined by
 monitoring and this information communicated to the building occupants
 or other appropriate people (see Chapter 3).   It is extremely important
 that everyone understand the actions to be taken if the data prove to be
 inconclusive with regards to site impact (ie.,  when concentrations are
within the "typical" range for indoor pollutants).
      The use of a  portable gas chromatograph (EPA Method IP-1A) is
 suggested in this application for organic compounds.  The instrument can
 be field calibrated and has detection limits adequate for screening
 level  results.  It  has the advantage that preliminary results are
 immediately available.  Alternatively, samples  may be collected using
 EPA Method IP-1A or TO-14 (SUMMA canisters).  The canister method offers
the advantages of positive compound identification,  lower detection
limits,  and a wider range of compounds identified, with the disadvantag-
es of higher costs  and delayed analytical results.  The use of adsorp-
tion tubes (eg;  EPA TO-1, TO-2,  IP-IB) are not  recommended for initial
screening unless a  rapid response instrument, such as a portable GC, is
available to determine approximate concentrations.  Adsorptive capaci-
ties of the tubes varies with different compounds and it is easy to
underload or overload the tubes in unknown environments.
      When particulate matter from the site is  considered a possible
route for indoor air exposure, samples of both  airborne and settled dust
                                  2-14

-------
 should  be  collected.   The  latter  is needed because the  larger particles
 deposit throughout  the structure  and may be  intermittently resuspended
 by  traffic or  cleaning.  Airborne dust may be monitored using an
 impactor with  filter  pack  assembly or continuous monitor  (EPA method  1P-
 10).  EPA  does  not  currently have guidance for collecting deposited
 dusts.   Procedures  developed by OSHA can be  used [29CFR 1910.132  (a)].
      Air  samples should be collected over a period of at least one hour
 in  the  living  area  of each story of the building and the basement, if
 applicable.  Samples  should be collected at  between three and six feet
 above the  floor, near the center of the room, and at least one foot from
 any object.  Deposited dust samples should be collected from such areas
 as  refrigerator tops  and window sills.
      If soil gas intrusion is suspected, it is preferred that the
 building windows and  doors be kept closed for the preceding 24 hours to
 allow establishment of normal pressure gradients.  If particulate
 monitors are used,  it  is preferred that indoor smoking be discouraged
 for several hours before and during monitoring because of the high
 particle count from smoke.  Any obvious potential sources of VOCs should
 be noted.  These include spilled furnace oil, gasoline storage cans and
 power equipment in attached garages,  paint cans,  etc.   Inquiries should
 be made to determine  how much VOC containing products, such as furniture
 polishes and hairsprays, have been used in the past 24 hours.
      If intrusion by soil gas is of primary concern and there is reason
 to suspect ambient air in the vicinity of the building may contain
 significant levels of non-site related pollutants of concern, ambient
 air monitoring should be conducted concurrent with indoor monitoring.
 If indoor and outdoor monitoring are  both conducted over at least a 4-
 hour period, it is reasonable to subtract out the outdoor concentrations
 to obtain an estimate of indoor concentrations due to  indoor sources and
 potential site chemical impact.
      If the only route of concern is  infiltration  of  site contaminants
 in the ambient air,  indoor monitoring  is  not  required.   The average
outdoor concentration near the building may be used as a good approxima-
tion of the indoor concentrations related to  site emissions.
                                  2-15

-------
2.4.2 Refined Monitoring For Indoor Air  Impacts Of Sites
      The objective of the monitoring discussed below is to determine
with high accuracy the actual impact from a Superfund site on the indoor
air of a specific structure.  The discussion is predicated on the
assumption that site impacts are from soil gas intrusion and organic and
particulate matter transported via the air pathway.  With only slight
modification in the procedures, the impacts of groundwater or non-
aqueous phase liquids seeping directly into the structure may be
determined.  However, in these cases, it is expected that indoor air
concentrations would be sufficiently high that decisions could be made
using screening techniques discussed in the previous section.
      The procedures allow estimation of concentrations in all stories
of the building and effective air exchange rates in the case of poorly
mixed buildings.  Explicitly not included is short-circuiting - pollut-
ants exiting directly to outside without mixing with indoor air.  Short-
circuiting can occur when fireplaces or forced air combustion devices
(eg.,oil  and gas furnaces)  are in operation on the lowest floor of the
building.  Additional causes may be clothes dryers or bathroom fans
vented to outside.
      The procedures are not prescriptive nor detailed enough to
encompass all possible cases.  Professional expertise and judgement are
required especially in such areas as potential  pathways to consider,
target compound selection,  sampling and analysis methods, and exact
positioning of the monitors.
      The procedures also include collecting data useful for making
long-term predictions of indoor concentrations based on short-term
testing.   [Professional  judgement is needed in deciding if this informa-
tion is needed for the particular assessment].   This is not an exact
science and,  although the analysis should provide reasonable results,
the accuracy will  depend on how close conditions are on the test day(s)
to average conditions for the structure under test.  It is suggested
that monitoring be conducted on several  different days under conditions
that approximate the range  of normal  meteorological and building operat-
ing conditions for the particular building site.
                                  2-16

-------
 Data  To Be Collected
       The following  data  needs  listing  is  provided  to give orientation
 to  types of information and  equipment needed.
       •     Outdoor  air concentrations  for all target species.   This
             includes upwind-downwind monitoring of  the Superfund site in
             some  cases.

       •     Indoor air concentrations for  all target species on  all
             building levels.

       •     Surface  sampling for deposited  particulates if particulate
             transport from the  site is  considered.

       •     Soil  gas concentrations for all target volatile compounds

       •     Indoor and outdoor  temperatures

       •     Complete building inspection

       •     Physical volume of  each level of the building
       •     Effective air exchange rate for the building
       •     Wind  speed and direction
       •      Indoor-outdoor pressure difference
       •      Barometric pressure
       •      Soil  permeability to vapor  flow
      .The  last 6  items are required only when long-term estimates of
indoor  air concentrations are needed.   Monitoring should be conducted
over  a  period not less than four hours  and preferably at least eight
hours.

Soil Gas Monitoring
      When Superfund site impact via transport of pollutants through
soil is of concern,   soil  gas monitoring at the structure being investi-
gated is required.  The general  procedures were described in a previous
section.   In this case,  however, although the actual concentration is
                                  2-17

-------
 important, the concentrations of compounds relative to each other.
 rather than actual concentrations, are of utmost importance.   (The
 reasons for this will be discussed under Data Analysis).
      Although soil gas concentrations should not vary significantly
 over periods as short as one day, it is considered good practice to
 collect an integrated sample over the same time period as ambient air
 and indoor air sampling.  This is particularly important if measurable
 rainfall or significant changes in barometric pressure occur within 12
 hours before or during sampling.  Rainfall and decreasing barometric
 pressure may increase soil gas flow into the building.  Increasing
 barometric pressure may decrease soil gas intrusion.
      Measuring soil permeabilities to vapor flow is not essential to
data analysis.  However, it provides extra data that may be useful.  If
 soil gas relative compositions vary significantly at different loca-
tions, soil permeabilities may be used to estimate relative contribu-
tions from each location.

Ambient Air Monitoring
      When Superfund site impact via transport through ambient air is of
concern, ambient air monitoring is required.   Monitoring at the poten-
tially impacted structure should, however, be conducted in all cases to
correct for target species from all  sources that may enter by this
pathway.
      Determining the site related fraction of the total amount of
chemicals measured at a monitoring station adjacent to a potentially
impacted structure can range from quite simple to complex.  The simplest
case is when no other significant source of target species exists
between the structure and the site and beyond the site.  The most
complex is the opposite situation,  but little real  difference is
presented if the other source is beyond the Superfund site.  In these
complex cases, monitoring upwind and downwind (between the site and the
secondary pollutant source and relative to the direction toward the
structure being tested) of the Superfund site as well  as at the struc-
ture would be required.  Significant sources   of target species between
the structure being tested and the Superfund  site would require a second
                                  2-18

-------
 monitoring station between the Superfund site and the secondary source.
 Resolution of these cases would require significant professional
 expertise  in  both monitoring and dispersion modeling.  A general  guide
 to  the  requirements can be found in Section 4 of Volume II  of the
 Air/Superfund NTGS Series.  Although these cases may not be typical,
 special  circumstances may warrant their evaluation so as not to overes-
 timate  the contribution of this pathway to the indoor air concentra-
 tions.
      For  the ambient air mode of transport,  determining the actual
 concentrations  of target species is important.   And,  because the
 concentrations  will,  in general, be low,  methods with low detection
 limits  are needed.   In general,  methods such  as  TO-1  or TO-14 will
 suffice  for determining the organics but  it is  recommended  that assis-
 tance be obtained from the appropriate  EPA Division.   Particulate
 methods  used  should be capable of collecting  integrated samples in
 various  size  ranges.   It is preferred that the  same monitoring methods
 be  used  outdoors  and  indoors.
      It is preferred that ambient air  sampling  begin at least one hour
 and preferably  two  hours before  indoor  air monitoring begins  and
 continue until  at least 30 minutes before  indoor monitoring  is complete.
 Because most  buildings have hourly air  exchange  rates in the  0.5 to 1
 range, air  entering the building in  the period before indoor  sampling
 remains  in  the  building for a  substantial  time.   Conversely,  outdoor air
 concentration fluctuations in  the  final 30 minutes, unless very large,
 have virtually  no effect on the  average indoor air concentration
 measured.   Because  indoor air  is reasonably well  mixed,  concentrations
 inside rise and fall  in an exponential  relationship (Introduction to
 Indoor Air Quality: A Reference  Manual, EPA/400/3-91/003, page 7):
      C,.T   -     Co +  
-------
      Ambient  air  monitors  should  be  located between 5 and 15 feet from
the  building with  sampler intakes  about five to six feet above ground
level.   The location  should be  such that the effects of windshields
(eg., shrubbery) is minimized.  Local sources, such as lawn mowers,
should  be moved to a  downwind location if possible.  Wind speed and wind
direction should be recorded continuously during monitoring.
      Ambient  air  monitoring to assess the air transport pathway should
obviously only be  conducted when the  prevailing wind is from the
Superfund site toward the monitoring  location.  Calm winds or signifi-
cant wind direction fluctuations can  result in significant negative bias
in the  results.  These are relatively unimportant if this pathway is not
of concern.  Wind  speed and direction should still be monitored for use
in air  exchange rate  and building  under-pressure calculations.
      It should be noted that in the  special case where Superfund site
impact  is only from the air transport pathway, indoor monitoring is not
necessary (except  for deposited particulates).  This is because the
average  indoor air concentrations  of  target species attributable to the
Superfund site will be equal to their average concentration in the
outside  air.   If short-term effects,  such as from remedial  actions, are
being evaluated, the building air  exchange rate may need to be deter-
mined and indoor air concentrations evaluated using the equation above
(or suitable modification dependent on fluctuations in the outdoor
concentrations).

Indoor Air Monitoring
      Indoor air monitoring is conducted to obtain the total  concentra-
tions of target species arising from ambient air infiltration, soil gas
intrusion, and indoor sources.   Information should be collected that
allows characterization of building dynamics such that reasonable
estimates can  be made under conditions different from those existing
during monitoring.   This will  allow more realistic risk assessment
estimates to be made.
      If soil   gas intrusion is suspected,  best monitoring conditions
exist when the indoor temperature  is at least 10°F higher than outside
and windspeeds are steady and exceed about five miles per hour.   Under
                                   2-20

-------
 these  conditions  reasonable  building  air  exchange  rates  and  under-
 pressurizations develop.   The  worst condition  for  monitoring  is when  the
 indoor temperature  is  lower  than  the  outdoor temperatures  and winds are
 calm.   Under  these  conditions,  the lowest  level(s)  of  the  building have
 a slight  over-pressurization and  soil  gas  entry may be restricted or
 eliminated.   Monitoring should  be avoided  if significant precipitation
 or barometric pressure fluctuations have occurred  in the preceding 12
 hours.
       Because air circulation  patterns  in  the  building vary with forced
 air circulation rates, outside  wind speed  and  direction, indoor-outdoor
 temperature differences, and indoor thermal gradients, several activi-
 ties should be completed before locating monitors.  Estimate the volume
 of the  structure by measuring outside dimensions,  as well  as each room
 in the  building.  Set up equipment to measure  the  effective air exchange
 rate.   It is  highly recommended that this  be done  using a  tracer gas
 method.   Procedures for this are  given  in  EPA Method IP-4.   In this
 procedure, a  known quantity of  tracer,  such as sulfur  hexafloride, is
 released  into the building, well mixed, and the concentration decrease
measured  as a function of time.   It is  preferable to monitor the tracer
 simultaneously in as many rooms of the  building as possible.  Using the
equation  presented in the above section on Ambient Air monitoring and
 assuming  the concentration of tracer gas in the outdoor air is zero,

      CI,T =  Cj0 e ~V1 ;  T  = time since tracer  release.

      But Cio  is the mass of  tracer released divided by the effective
air volume of the  building and v is the infiltration rate,  Q, divided by
the effective air  volume,  V.    Thus,
                               _ msrSF,     -JZ
                          Li.T	—  e
                                  2-21

-------
Thus, a plot  of  Ln  Cf T versus  T yields  a  straight  line with an  inter-
cept at T = 0 (equal  to the  first  term  on the  right side of the equa-
tion) from which  the  effective  air volume of the building can be
estimated and slope  (equal to  -Q/V, building air exchange rate) from
which the infiltration  flow  rate can be estimated.  Furthermore, since
      Q = L [AAT  +  B/i2]  °-5 from EPA 400/3-9/003, page 8, the informa-
tion collected may  be  used to estimate  infiltrations under various
meteorological conditions.   In  the above equation,
      L     =    leakage area  of  building
      A     =    Stack coefficient
      AT    =    Indoor-outdoor temperature difference
      B     =    wind  coefficient
      /i     =    average wind  speed

      The above EPA document provides the appropriate values for A and
B.  Thus, L,  the  leakage area may  be estimated.  Obviously, indoor and
outdoor temperatures and wind speed should be monitored during this
testing.  The relative  tracer decay rates in different rooms of the
building are  indicative  of air  movement and may be used to locate the
samplers.
      If this level  of  sophistication is considered unnecessary for the
particular application,  the effective leakage area can be determined
using the fan pressurization-depressurization method and the infiltra-
tion rate calculated from the above equation.  Although the method is
less accurate, primarily due to uncertainty in the actual air volume of
the building,  it  is simpler to  conduct.  Air exchange rates should not
simply be estimated for  indoor  air monitoring of the level  discussed
here.
      It is also  important to measure the pressure difference between
the inside and outside of the building.  This pressure difference will
likely be in  the  range of zero  to ten pascals.   It will  vary with height
in the building.  Building under-pressurization is effected by indoor-

                                  2-22

-------
 outdoor temperature difference and wind speed.   Thus,  it is important
 that these variables also be measured.   The indoor pressure monitoring
 location should be between three and six feet  above floor level  in  the
 lowest  level  of the building.   (See Appendix A,  Section  A.3.3,  Equation
 24.)
      Monitors  for target species should be located on each level of the
 building.   Preferred locations are in the living spaces  (living  rooms,
 dens, bedrooms).   They should  be located in the  breathing zone,  three to
 six  feet above  floor level,  and approximately  in the center of  the  room.
 Monitors should not be located near obstructions or obvious sources of
 pollutants.
      Selection of monitoring  methods depends on the target species
 selected.   In most cases,  indoor air methods IP-1A (canisters)  or IP-IB
 will  suffice  for  organics  although specific methods may  have  to  be used
 for  some species.   Airborne  particlates  should be collected using EPA
 method  IP-9 when  airborne  transport of Superfund site particulates is of
 concern.   In this  case, deposited  dusts  should also be collected [OSHA
 procedure, 29CFR  1910.132(a)].
      Indoor air  samples should  be collected over a period  of at least
 four hours; eight  hours of sampling  is preferred.   During this period,
 minimal  indoor  activity should  occur.  Pollutant  generating activities
 such as  housecleaning, furniture polishing,  and  indoor smoking should be
 discouraged.  Windows  and  exterior doors  should  remain closed to the
 maximum  extent  practical.

 Data Analysis
      A  hypothetical case  is presented in Tables  2.1, 2.2,  and 2.3 to
 assist with following  the  text below.  For  this  example,  monitoring data
 are given for the ambient  air adjacent to the building, upwind back-
ground ambient  air, soil gas, and  the indoor air  at  two levels in the
building.  The monitoring  data collected for the  three target compounds
are tabulated in Table 2.1.
      The first step should be to  consider the impact due to pollutants
 in ambient air.   It can be assumed that the average  target  species
concentration in the indoor air attributable to this pathway is the same
                                   2-23

-------
as the average target species concentration in the ambient air.
Subtract the total average concentrations of target species in ambient
air from the average concentrations measured at each indoor monitoring
location (in Table 2.1,column 1  is subtracted from columns 5 and 6 to
yield the results in columns 5 and 6 of Table 2.2).  The resulting
indoor concentrations are the sum of target species concentrations
resulting from indoor sources and soil gas intrusion.
      If Superfund site impact via the ambient air pathway is of
concern, determine the average target species concentrations in the
ambient air from the upwind monitoring and modeling results and tabulate
these as "Indoor air concentrations due to Superfund Site Target Species
in Ambient Air" (column 3 is subtracted from column 2,  both in Table
2.1, to yield column 3 of Table  2.2) .  Note that if this pathway is of
concern and appropriate corrections are not made for non-site sources,
the total target species concentration measured in the  ambient air must
be considered to be from the Superfund site.  These concentrations will
be the same for all  indoor monitoring locations.
      Using data from the soil gas monitoring and the indoor air
concentration residuals obtained by subtracting the total average
ambient air target species concentrations (column 2 of  either Table 2.1
or 2.2) from the total  average concentration measured at each monitoring
location, the site contribution due to soil gas intrusion can be
calculated.   Because neither the soil gas flow rate nor the source
strength of all indoor sources is known, the calculation is based on the
relative concentrations of target species in soil gas and the indoor air
residuals above.   If it is assumed that soil gas components are not
differentially absorbed indoors,  then they should appear at all loca-
tions in the same ratios.
      Thus,  first divide the soil gas concentrations of all target
species by the highest measured concentration in the soil gas (all
concentrations are,  thus,  expressed as a fraction between zero and one).
In this example,  all  soil  gas concentrations in column  3 of Table 2.1
are divided by 100 to yield the ratios in column 2 of Table 2.3.
                                  2-24

-------
 C
 o
 re
 o
 o
 Q.
CM


 0)
.0
 nS
CM
ra S-
c
0 -r-
Cf~
fc_
o o
o z:
n) S-
c
O -r-
C C
O 0
o z:
C V)
-i— ra
. ^
O f—
C "-
o o
O OO
S-
•i- -a
< c
•M O
C S-
eu DI
^ u
<: ca
S- 01
-o
-4-> 1 	
C -i-

C
nj
c
•r™
E
nj
4-9
c:
o


in
CM



o
^jl





CD
O
t-H





in
o
o
o




f-H
0

o





LU
0
t—



in
~



o
CM





O
in






l-H
O
o
o



CM
o
0

o





LU
0
O-



UD
•^



«5
^1





O
tp







in
o
CD





l-H
•
O

cu
c
CO
IM
C
0)
CQ

                                      C
                                      at
                                      s_
                                      o
 O)
4->
 O
 tu

 s_
 o
o

 rs
-I-)
 nj
Q
                                     (0
                                     O
                                     O
                                     Q.
CM


CM

 O)

jQ
 a)
CM

ns S-
O
O •«-
C. C.
o o
o z:
t
nj S-
O
O -r-
0 O
o z:
C 
"~ 
O • i—
O O t»
T3 C
c o u_
i— i O OO
S- CD
•r- C
-o
+-> 1 —
C •!-
CU 3
•r- CQ
is
C
«J
C
•f—
E
nj
.4-3
C
o
o


en
^[
CM




en
en
CO




o
0





in
o
o
•
o






i-H
0

o





LU
O
t—



00
en
^
,_!



00
en
en
f-H




O
in





l-H
O
0
•
o





CM
O
0

o





LU
t_l
Q_





in
^





in
^




o
vo






in
o
*
o







l-H
•
O

0)
C
cu

C
cu
CQ


                                                                                CJ
                                                                                O
                                                                                OO
                                            S-
                                            o
                                                                                -o
                                                                                a>
                                                                                s-
                                                                                o
                                                                                ra
                                                                               o
                                                                                O
                                                                               •f—
                                                                               -t->
                                                                                CO
                                            o
                                            a.
co

CM

 O)

JD
 OS
m
C\J
,—
*f~ ^_
0 O
OO +J
X C
rO O
z: z:
VI
ea
1—1
.f— ^.
0 O
00 4->
•f—
X C
(T9 O
z: z:
CM
S-
o
+-> o
C 4->
O re
z: o:
l-H
S-
o
•«-» o
C -t->
O n)
z: a:
0)
O
c/i nj
ra C£.
t n
•
i— O
•r- C
O 0
00 O
C
re
C
•r—
E
(O
4^
C
0
o



en
*!•
CM




en
en
•
CO




en
•*
CM



en
en
CO






0
o
«
rp-4






LU
O
1—




in
^j-
CM
l-H



in
en
en
*
t-H



vo
en
en
CM


to
en
en
CO






0
in
•
o






LU
O
O_




^^
en
^»"
^



^
en
co
•
CM





in
i^




in
f>I






<*^
VO
*
o


CO
C
CO
N
C
CO
CO


                                                             2-25

-------
       Now divide the appropriate  indoor  air  residual  concentrations at
 each monitoring  location  by  the soil gas  fraction  for that compound.   In
 this example,  for monitor location  1,  the 3.99  for TCE was divided by
 1.00,  the 1.998  for  PCE was  divided by 0.5,  and  the 4.5  for benzene was
 divided  by 0.6 to yield the  results in column 3  of Table 2.3.
       Identify the lowest non-zero quotient  from this  operation for each
 monitoring location  and multiply  all soil  gas fractions  by these
 numbers.   In this example, 3.99 for TCE  is the  lowest  ratio for monitor
 location  1.  Thus, the results of this step, given in column 5 of Table
 2.3, are  obtained by multiplying  3.99  by  1 for TCE, 0.5  for PCE, and 0.6
 for benzene.   The results are the maximum potential concentrations at
 each sampling  location from  soil  gases.
       These concentrations should be added to the  values calculated
 above  for site impact via the ambient  air pathway  (column 3 of Table
 2.2) to obtain the total  site related  concentrations  to which occupants
 may be exposed (summation  not shown).

 2.4.3 Making Long-term Estimates  from  Short-term Monitoring Data
      The  data above could be used for a  preliminary  risk characteriza-
 tion.  However,  it is preferred that prediction of longer-term average
 concentrations be made for risk evaluation purposes.  General procedures
 to make this estimate are presented below.
      First estimate the  building air  infiltration rate and air exchange
 rate under the appropriate long-term meteorological conditions.  Using
 the building leakage area estimated from  the tracer gas testing (or the
 fan depressurization test), estimate the  long-term average air infiltra-
 tion rate  from:
      Q =  L[AAT + B/u2]0-5
 using appropriate long-term average values for wind speed and indoor-
outdoor temperature differences.  Unless  long-term meteorological
conditions are similar to those existing  during monitoring, divide Q by
the measured physical volume of the building (rather than the estimated
effective volume) to obtain average air exchanges per unit time.
      Secondly, estimate  ambient  air concentrations at the building for
Superfund  site related chemicals due to the air transport mechanism.
                                  2-26

-------
 The ambient air concentrations determined during monitoring can be used
 to  calibrate an air dispersion model  for use in this estimation.
       Thirdly,  estimate the soil  gas  flow rate into the building  under
 the long-term meteorological  conditions.  Estimate the soil gas flow
 rate into the building during the on-site monitoring from:
       ^soil  ~ "   i/ soil
       where,
       Q      = estimated building  air  infiltration rate
       Cf     = calculated indoor air average concentration of chemicals
             in  soil  gas
       Csojt   =  chemical concentration in soil  gas

 Now calculate K from
       ^soii  = K Pa
       where,
             Pa  = the indoor-outdoor pressure difference  measured
       during monitoring
            .K   = a  constant  comprised  of building specific terms
       If  soil gas intrusion is  primarily by diffusion,  the  Qsoil calcu-
 lated  can be used for the  long-term average.   However,  if soil  gas
 intrusion is primarily by  convection, Qsoil  is  proportional  (See Johnson
 and  Ettinger,   Environmental  Science  and Technology,  Vol. 25, No. 8,
 1991,  page  1449)  to  the building  under-pressurization,  as indicated  by
 the  above equation for K.  Thus,  by using  Pa calculated  for the building
 under-pressurization  from  equation  24 of Appendix A,  Qsoil under differ-
 ent meteorological conditions can be  estimated  from the  above rela-
 tionship.  The  measured building  under-pressurization during monitoring
 should be used  to calibrate Equation  24  for site  specific conditions.
 This implicitly  assumes soil gas  concentrations remain constant.  This
may need adjustment  if the source strength  varies  with periods short
 compared to  the  averaging  time.
      The above  procedure  allows  reasonable  estimates to  be made of
 long-term indoor  concentrations from  short  term data.  Using the risk
 assessment procedures  outlined  in Section 3, reasonable  assessments of
 the indoor air  impact  of a site can be made.
                                  2-27

-------
                           3.0  HAZARD ANALYSIS

       Inhalation  of contaminants  in indoor  air  is  but  one  of  the
 exposure  pathways considered  in risk assessments.   Explicit guidance  for
 the  conduct  of  risk assessments is  given  in "Risk  Assessment  Guidance
 for  Superfund:  Volume  1  Human Health Evaluation Manual".   The Manual  has
 three  parts:
       (1)    baseline risk  assessment given  in Part A,  EPA/540/1-89/002
       (2)    development  of preliminary remediation goals given in  Part
             B,  Publication 9285.7-01B
       (3)    remedial alternatives risk evaluation  given in Part C,
             publication  9285.7-01C
       This chapter  considers  the potential  impacts on  indoor  air to.be
 from soil gas,  chemical  volatilization from contaminated groundwater  use
 and ambient  air infiltration.   The  procedures are  useful for  exposures
 occurring during  baseline,  remedial  actions, and post  clean-up.
 Information  is provided  to estimate risks for cancer and non-cancer
 effects and  for short-term and  long-term impacts.   It  is explicitly
 assumed that adequate  information has  been  developed on indoor air
 chemical concentrations  and the likely exposed population.

 3.1    INCREMENTAL INDOOR AIR  RISKS
      The risk assessment  procedures detailed in RAGS/HHEM are predicat-
 ed on evaluating the incremental risks  due  to contaminants originating
 at the Superfund site.   Modeling predications can  be,  and are, used to
 estimate emissions from  specific site  sources and  potential chemical
 concentrations at various  on-site and  off-site locations.  These models
 do not incorporate chemical releases from non-site sources, although
 such releases may be modeled  to provide comparative background informa-
 tion.  Likewise, monitoring data are adjusted for  background  to assess
 incremental  site impacts.  For example, ambient air monitoring is
conducted up-wind and down-wind of  the  site, up-gradient and  down-
gradient wells are used for releases to ground water, and contaminant
                                   3-1

-------
concentrations  in  background soils are used to determine site contribu-
tions to the  total.
      For consistency  it  is necessary that equivalent procedures be
followed when assessing potential site impacts on indoor air quality.
Since it has  been  well established that indoor air typically contains a
wide variety  of chemicals found at Superfund sites, and, in many cases,
in concentrations  high enough to be of concern, risk estimates based on
the total measured indoor air concentrations may reasonably be expected
to frequently result in substantial overestimates of site impacts.   Risk
estimates based on predictive modeling would not be expected to include
non-site impacts and therefore are in agreement with current EPA
guidance.
      Risk estimates based on combinations of exterior monitoring (eg.,
ambient air and soil gases) may contain some extraneous risk due to non-
site related chemicals.  For example, ambient air measurements made
outside the building being evaluated may contain chemicals transported
from non-site locations such as industrial operations, auto exhaust,
roadway paving, and pesticide applications.  Soil gases may also contain
chemicals from  materials deposited on or in the ground within a few
meters of the structure such as pesticide application or fuel oil
spills.
      Risk estimates based only on indoor air monitoring would be a
combination of  possible site related impacts and probable non-site
related impacts from the many potential indoor sources.  As such, they
would not follow the basic guidelines (ie, assess site related impacts).
To be useful, these data must be corrected for background in so far as
practical.  EPA guidance provides that if background risk might be  a
concern, it should be calculated separately from site-related risk
(RAGS/HHEM Part A, Section 5.7).
      Current EPA  guidance, therefore, is that efforts to distinguish
between site-related and non-site related impacts on indoor air would be
required. The study design must, therefore, provide a way to distinguish
among site related indoor impacts, pollutants from indoor sources,  and
background concentrations in ambient air.  Only the site related
potential risks need to be calculated (non-site related risks may be
                                   3-2

-------
calculated  but  this  is  not  necessarily required).  No current EPA
guidance  provides  for summing the potential risks  from site related  and
non-site  related  impacts.

3.2   RISK  COMMUNICATION
      Section 8.6  of RAGS/HHEM Part A states:
      "The  results of the baseline evaluation should not be taken as a
characterization of absolute risk.  An important use of the risk and
hazard  index estimates  is to highlight potential sources of risk at  a
site so that they may be dealt with effectively in the remedial  process.
It is the responsibility of the risk assessment team to develop conclu-
sions about the magnitude and kinds of risk at the site and the major
uncertainties affecting the risk estimates. It is not the responsibility
of the risk assessment  team to evaluate the significance of the risk in
a program context, or whether and how the risk should be addressed,
which are risk management decisions."
      The ultimate user of the risk characterization results will  be the
RPM or other risk manager for the site.
      It is important that the indoor air impact assessment contain  a
discussion of:
      •     confidence that the key site-related contaminants  were
            identified and discussion of contaminant concentrations
            relative to background concentration ranges;
      •     level  of confidence in the quantitative toxicity information
            used to estimate risks and presentation of qualitative
            information  on the toxicity  of substances not included in
            the quantitative assessment;
      •     level  of confidence in the exposure estimates for  key  expo-
            sure pathways and related exposure parameter  assumptions;
      •     the magnitude of the  cancer  risks  and non-cancer hazard
            indices relative to the Superfund  site remediation goals in
            the NCP (e.g.,  the cancer risk range of 10"4 to 10"7  and  non
            cancer hazard index of 1.0);
      •     the major factors driving the site risks  (e.g.,  substances,
            pathways,  and pathway combinations);
                                  3-3

-------
      •     the major  factors  reducing  the certainty  in the results and
            the significance of  these uncertainties (e.g., adding risks
            over  several  substances  and  pathways);
      Assessment  of potential  risks  due  to the  indoor air pathway for
buildings located proximate to Superfund sites  may be initiated as a
result of individual property  owner  requests,   community requests,
preliminary assessments,  baseline  investigations, or assessment of
remedial alternatives.   In all cases, it is  important to bear in mind
that good communication with the affected people  is critical to their
acceptance of the results and  recommendations.  RPMs are encouraged to
work with the risk assessor and  community relations coordinator to
develop the appropriate means  for  communication.
      It is important  in  estimating  risks due to  exposure to indoor air
to recognize that pollutant sources  other than  Superfund sites normally
exist in buildings.  In fact,  prior  studies  indicate the cancer risk
factor for private residences  not  proximate  to  Superfund sites may be
10"4  or  higher.  While  this  fact  may be  communicated to  building occu-
pants, as well as other general  risk information, it must be understood
that, in general, people view  and  accept voluntary and involuntary risks
differently.  Thus, the potentially  affected population may express
different perceptions  and acceptabilities for equivalent risks arising
from exposure to  chemicals used  in routine household activities and the
same chemicals originating from  a  Superfund  site.
      Experience  has shown that  it is important to develop a management
plan prior to initiation of an indoor air impact  assessment.  While the
plan will vary with site-specific  considerations, it should be adequate
to communicate the sequence of actions to be taken to evaluate exposures
and risks,  the methodology that will be employed, the levels of risks
upon which decisions will be made, and actions  to be taken based upon
those decisions.  An illustrative  decision tree is presented in Figure
3-1.
      In this illustration, potential exposures,  and resultant risks,
are first estimated based on the use of reasonable predictive models.
If the simple screening model predicts significant exposures, then a
more sophisticated model is run to give a better, and more realistic,
                                   3-4

-------
               Figure 3-1.  Illustrative Decision Tree
                     Conduct Modeling to E*tIm*U Exposure
             1. Conduct corw«rvctlv« tcf*«nlng mod«llno
             2. If modal pf«dlcts »Kjmfic«nt impact, th*n conduct
               •opni*tlc*t*d modeling
               1
     Model Predict* Expowre
     Risk Above X
            Communjctte to Reiident* No
            Additional Risk Projected
Results Indicate risks are:
High
Moderate
Low
Confirm Remits by Monitoring Inside
and Outside Building
                   1
Confirm Model Prediction! by Ambient Air
•nd/or Soil Gas Monitoring
             i
                   Monitoring Results Confirm Risk Exists
                    Inform Residents of Actions to be Taken
                                              3-5

-------
 estimate of the indoor concentrations likely to exist.  If the simple
screening model, which is conservative, does not indicate concentrations
could be high enough to cause concern, there is no reason to run the
more sophisticated models, which would only reduce the estimated concen-
trations.   If the modeling predicts exposures which would pose risks
above a predetermined value, on-site monitoring, varying with the level
of risk estimated, is conducted.  If the on-site monitoring confirms the
model predictions, then predetermined actions may be taken based on the
estimated risk level.
      Whatever decision tree is developed for a particular site should
be communicated to the potentially affected public before the plan is
executed.  The meaning of the cancer risk factors (eg., 1 xlO"4) and
hazard indices should also be communicated.
      Consider, for example, the following case which was developed for
a specific chemical at a specific site.  Information developed during
the screening phase indicated the potential for significant indoor
exposures to a chemical with estimated excess lifetime cancer risks to
humans as detailed below.

       Apportioned Lifetime Risk from a Four-year Exposure
AGE
0 to 5
6 to 9
10 to 13
14 to 17
18 to 21
100 ppb/v
2.3 E-2
5.8 E-3
4.6 E-3
3.5 E-3
2.3 E-3
10 ppb/v
2.3 E-3
5.8 E-4
4.6 E-4
3.5 E-4
2.3 E-4
1 ppb/v
2.3 E-4
5.8 E-5
4.6 E-5
3.5 E-5
2.3 E-5
0.2 ppb/v
4.6 E-5
1.2 E-5
9.3 E-6
7.0 E-6
4.6 E-6
0.1 ppb/v
2.3 E-5
5.8 E-6
4.6 E-6
3.5 E-6
2.3 E-6
These data suggest that children, especially very young children and
neonates, may be unusually sensitive to the chemical.  In this case no
children resided in the buildings on a full-time basis but were present
on a regular basis as frequent visitors or for daycare purposes.  Thus,
                                   3-6

-------
 it would be relatively simple for parents to take measures to reduce
 exposure time for their children.
       Before on-site monitoring took place,  it was determined that
 temporary corrective actions could be taken, if needed,  and effective-
 ness  confirmed within twelve days after the  monitoring data was  avail-
 able.   It was also determined that permanent remedial action,  if needed,
 could  be taken,  and its effectiveness determined within  60 days.
       The following decision scheme was then designed for  actions to be
 taken  depending  on the actual  concentrations determined  by on-site
 monitoring:
    0.2 to 10
     ppb/v
 If the concentration were to be found to be within
 this  range, permanent remedial action would be
 taken to bring the levels below 0.2 ppb/v.  The
 action would be  instituted and effectiveness con-
 firmed within sixty days.
    10 to 100
     ppb/v
If the concentration were to be found to be within
this range, temporary corrective actions would be
instituted to bring the levels below 10 ppb/v.
This action would be instituted and confirmed with-
in twelve days.  The temporary action would be
followed by a permanent remedial action to bring
the levels below 0.2 ppb/v.  The permanent action
would be instituted and effectiveness confirmed
within sixty days.
3.3 OVERVIEW OF RISK ASSESSMENT GUIDANCE
      The purpose of the balance of this chapter is to provide an
overview of the Superfund guidance (RAGS Part A) for the indoor air
pathway.  As such, it must be understood that it does not contain the
detail necessary for untrained personnel to conduct a risk assessment.
Risk assessment data, even in cases where accepted cancer or non-cancer
                                   3-7

-------
risk factors are available,  should be reviewed with a trained profes-
sional for site specific adjustments that might be necessary.

3.3.1 Exposure
      The first step considered is exposure of the affected population.
The fact that certain concentration of pollutants exist in the indoor
air does not in and of itself constitute a risk.  The magnitude,
frequency, and duration of that exposure for the affected individuals
(or grouping such as adults  and children) must be defined.  The average
exposure can be expressed as a function of time or of body weight.
Exposure normalized for time and body weight is termed "intake" and is
expressed in units of mg chemical/Kg body weight-day.  Thus, there are
three categories of variables for intake:
      •     chemical concentrations
      •     population (contact rate, exposure frequency and duration,
            and body weight)
      •     averaging time
      The intake variable values should be selected so that the combina-
tion of all intake variables results in an estimate of the reasonable
maximum exposure for the pathway.  "Reasonable" is not based solely on
quantitative information and requires professional judgement.  EPA
guidelines for "reasonable" generally use the 95th or 90th percentile
values for contact rate, exposure frequency, and duration.  If statis-
tical data are not available, use reasonable conservative estimates.
Professional  judgement is required to estimate the exposure time and
exposure duration for exposure to indoor air.  Both can vary with
building type (home, school, workplace) and age.
      Some chemicals can produce an effect after a single or very short-
term exposure to relatively low concentrations.  These chemicals include
acute toxicants such as skin irritants and neurological poisons, and
developmental toxicants.  At sites where these types of chemicals are
present,  it is important to assess exposure for the shortest time period
that could result in an effect.  For acute toxicants this is usually a
single exposure event or a day, although multiple exposures over several
days also could result in an effect.  For developmental toxicants, the
                                  3-8

-------
time period of concern  is  the  exposure event.  This is based on the
assumption that a single exposure  at the critical time in development is
sufficient to produce an adverse effect.   It should be noted that the
critical time referred  to  can  occur in almost any segment of the human
population (i.e., fertile  men  and  women, the conceptus, and the child up
to the age of sexual maturation).
      When evaluating longer-term  exposure to non-carcinogenic toxi-
cants, intakes are calculated  by averaging intakes over the period of
exposure (i.e., subchronic or  chronic daily intakes).  For carcinogens,
intakes are calculated  by  prorating the total cumulative dose over a
lifetime.  This distinction relates to the currently held scientific
opinion that the mechanism of  action for each category is different.
The approach for carcinogens is based on the assumption that a high dose
received over a short period of time is equivalent to a corresponding
low dose spread over a  lifetime (EPA 1986b).  This approach becomes
problematic as the exposures in question become more intense but less
frequent, especially when there is evidence that the agent has shown
dose-rate related carcinogenic effects.  It is necessary to consult a
toxicologist to assess  the level of uncertainty associated with the
exposure assessment for carcinogens.

3.3.2       Toxicity Assessment
      Health impacts of indoor air chemicals may be divided into two
broad classes: non-carcinogenic effects and carcinogenic effects.
However,  there are additional considerations within each broad class.
      A reference dose,  RfD, is the toxicity value most often used for
evaluating non-carcinogenic effects resulting from exposures at Superf-
und sites.   Note that carcinogens may exhibit non-carcinogenic effects
and that RfDs for these chemicals should also be sought.   Various  types
of RfDs are available.   These include route of exposure (e.i., oral  and
inhalation),  critical  effect (i.e., developmental  and others), and
length of exposure event (chronic,  subchronic,  or single event).   A
chronic RfD is an estimate of a daily exposure level  for the human
population,  including sensitive subpopulations,  that is unlikely to
result in an  appreciable risk of deleterious effects during a lifetime.
                                   3-9

-------
Chronic  RfDs  should  be  used  to  evaluate  potential non-carcinogenic
effects  associated with exposure  of  from seven years to lifetime.
Subchronic  RfDs  (RfDs)  should be used to evaluate effects from exposure
periods  of  two weeks  to seven years.  Developmental RfDs (RfDdT)  are
used to  evaluate  the  potential  effects on a developing organism follow-
ing a single  exposure event.  General use of the RfDs assumes a thresh-
old type mechanism in which  no  adverse effect is observed below the RfD.
However, RfDs are generally  considered to have an uncertainty spanning
an order of magnitude or more and  are not viewed as a strict scientific
demarcation between toxic and non-toxic  levels.
      Carcinogenic effects are  considered to have no threshold and,
thus, calculations are  based on the  presumption that any exposure
carries  a finite probability of a  carcinogenic response.  For Superfund
purposes, the relationship between dose  and response is considered to be
linear in the low-dose  region and  represented by the slope factor.  This
factor generally represents the upper 95th percent confidence limit on
the probability of a  response per  unit intake of a chemical  over a
lifetime.  Most carcinogenic slope factors (CSF) are based on adminis-
tered dose, however,  some data  are based on the absorbed dose.  Care
should be exercised in  the selection and application of these two
different slope factors.  Toxicity values for carcinogenic effects may
also be  expressed in  terms of risk per unit concentration of the
chemical in the medium  where human contact occurs.

3.3.3       Sources of  Toxicity Values
      EPA's Integrated  Risk Information System (IRIS) is a data base
containing up-to-date health risk and regulatory information for
numerous chemicals.   It contains only those RfDs and slope factors that
have been verified by appropriate work groups.   Information  in IRIS
supersedes all other  sources.  Only  if information is not in IRIS should
other sources be consulted.   IRIS is only available on-line.  For
information on how to use the database, contact IRIS User Support
(513/569-7254).
      The Health Effects Summary Tables  (HEAST) summarize interim, and
some verified, RfDs and slope factors and contains toxicity  information
                                  3-10

-------
 for  specific  chemicals.   HEAST  also  provides  references  to  supporting
 toxicity  information.  HEAST  is  published  quarterly  and  can be  obtained
 from the  Superfund  Docket (202/382-3046)
       The Agency  for  Toxic  Substances  and  Disease  Registry  (ATSDR)  is
 developing toxicological  profiles  for  275  hazardous  substances  identi-
 fied at Superfund sites.  The profiles contain general toxicity informa-
 tion and  levels of  exposure associated with various  endpoints.   Health
 effects in humans are discussed  by exposure route  (e.g., oral,  inhala-
 tion,  and dermal) and duration  (eg.  acute, intermediate  and chronic).
 Information on the  status of  a  particular  profile  can be obtained by
 contacting ATSDR  or the National Technical Information Service  (NTIS) at
 703/487-4650  or 800/336-4700.
       EPA's Environmental Criteria and Assessment  Office (ECAO)  may be
 contacted at  513/569-7300 for general  toxicological  information  as well
 as for technical  guidance on the evaluation of chemicals without
 toxicity  values.  ECAO will respond  to contractor  requests only  upon
 identification of the RPM or regional  risk assessment contact.   ECAO
 should be contacted before  using references other  than those cited in
 IRIS  or HEAST to  see if more current information is  available.  Any
 derivation  of toxicity values should be done only  in conjunction with
 the  regional risk assessment contact,  who will submit the derivation to
 ECAO  for  approval.

 3.3.4       Quantifying Risks
      Following the above procedures,  chemical intakes and appropriate
 RfDs  and  slope factors will have been  obtained.  Before proceeding with
 calculations,  it  is highly  recommended that the information be consoli-
dated into  tabular format to ensure the proper data are used in the
calculations.   Specific checks should  be made to ensure:
      •     All  RfDs and  CSFs are  based on inhalation as the route of
            exposure
      •     All  RfDs and CSFs are  expressed in the same units as used
            for chemical   intake.   Toxicity values obtained from IRIS are
            generally expressed as ambient air concentrations (ie.,
            mg/m3) instead of  administered  dose (ie., mg/Kg-day).
                                   3-11

-------
      •     Non-cancer  effects  are  considered for carcinogens.
      •     The proper  RfDs  are used  for the exposure scenarios.  Do not
            use chronic RfDs  for short-term exposures (except as initial
            screening values) and do  not convert short-term exposures to
            equivalent  lifetime values to compare with chronic RfDs.
            Use only subchronic or  shorter-term toxicity values that
            compare well with the estimated exposure duration.
      •     All intakes and  toxicity  values are expressed as adminis-
            tered dose.  A few  chemicals listed in IRIS (eg., trichloro-
            ethylene) have the  CSF  expressed as absorbed dose.
      •     Averaging times  (AT)  for  non-carcinogenic risks are equal to
            the exposure duration (ED).  ATs for carcinogenic effects
            are always  70 years.
      The table must have the data  separated such that carcinogenic and
non-carcinogenic risks  can be summed  separately.  (It is best to present
these in separate tables).  Also, the table must have the data for non-
carcinogenic risks presented  such that short-term, subchronic, chronic,
and developmental risks, as  appropriate for the scenario being consid-
ered, can be summed individually.
      For carcinogenic  effects,  multiply the appropriate calculated
intakes by the slope factors  for the  chemicals.  For non-carcinogenic
effects, divide the appropriate calculated intakes by the RfDs for the
chemicals.  Note that this procedures gives, for carcinogens only,  an
upper-bound estimate of the  lifetime  incremental risk of developing
cancer.  For non-carcinogens, this  yields a non-cancer hazard quotient
(HQ), the value of which is not a measure of statistical probability of
non-cancer effects.
      For carcinogenic  effects,  sum only the individual risks calculated
using the slope factors.  This  sum  represents the total upper-bound
incremental lifetime cancer risk.
      For non-carcinogenic effects, several separate summations can be
made.  If more than one exposure duration (ie, chronic, subchronic,
short-term, or developmental) was considered, each must be summed
separately.  Do not sum all non-cancer HQs either by summing individual
data or summing totals  for chronic, subchronic, etc.  Furthermore,  sum
                                  3-12

-------
 only those Hqs corresponding to exposures that will  be occurring
 simultaneously.   These sums are designated Hazard Indices  (HI).
       At this point,  one or more tables  will  exist which contain
 individual  and total  cancer risk factors and  non-cancer hazard quotients
 and  hazard indices.   Understanding the meaning of and  limitations  of
 these  results is discussed in the following section.

 3.3.5   Data Limitations and Interpretations

       3.3.5.1     Carcinogenic Effects
       It was  assumed  in the preceding that the cancer  risk could be
 calculated  by multiplying intake by the  slope  factor.   This  is valid  in
 the  low-dose  portion  of the dose response curve.   For  estimated  risks
 above  0.01,  this assumption may no longer be valid and  risk  may  be
 calculated  using the  one-hit equation:

       Risk  =  1 - exp  [(I)(AF)]
       where all  terms are as previously  defined.

       Unless  the total  risk number calculated  is dominated by the risks
 from one or two  chemicals,  the  total cancer risk may be overstated.
 This can occur because  slope factors are  upper  95th percentile estimates
 of potency  and upper  95th percentiles of  probability distributions are
 not strictly  additive.
       The total  risk  summation  procedures  gave  equal  weight  to class A,
 B, and C carcinogens  as well  as  to  slope  factors derived from human or
 animal data.  The  calculation,  therefore,  may overestimate the risk from
 some chemicals.
      The summation procedure explicitly  assumes the actions of  the
carcinogens are  independent.  This  assumption ignores possible synergis-
tic or antagonistic effects  among  chemicals and assumes similarity in
mechanisms and metabolism.   These  assumptions are made because data to
quantitatively assess mixture interactions are generally not available.
                                  3-13

-------
      3.3.5.2     Non-Carcinogenic  Effects
      The hazard quotient calculated assumes no effect below HQ = 1.
The level of concern does not  increase linearly as HQ approaches or
exceeds unity.  The RfDs are not based on the same severity of toxic
effects and the slope of the dose-response curve can vary widely among
different chemicals.  However, for  HQ >1, there is concern for potential
non-cancer effects which increases  as HQ increases.
      The hazard indices (HI)  calculated assume that simultaneous
subthreshold (ie, HQ< 1) exposures  to several chemicals could result in
an adverse health effect and that the magnitude of the concern is
proportional to the sum of the HQs.  Although an HI exceeding unity may
be of concern, the level of concern is higher if individual Hqs exceed
unity.  Since the His are derived from Hqs for compounds that may not be
expected to induce the same type of effect or that do not act by the
same mechanism, they can overestimate the potential for effects.  If the
HI is greater than unity as a consequence of summing several  Hqs of
similar value, it would be appropriate to segregate the compounds by
effect and by mechanism of action and to derive separate his  for each
group.  This analysis is not simple and should be performed by a
toxicologist.
                                  3-14

-------
APPENDIX A. PREDICTIVE SCREENING TECHNIQUES

-------
                            TABLE OF CONTENTS
APPENDIX A.  PREDICTIVE SCREENING TECHNIQUES

             A.I  Models to Estimate Contaminant Flux	A-l
                  A. 1.1    Farmer Model	A-2
                  A.1.2    Correction for Transport Resistance of
                           Capillary Fringe 	  A-4
                  A.1.3    Correction for Adsorption to Soil  . .  .  A-7
                  A. 1.4    FAVN Model	A-8
                  A.1.5    Emissions From Contaminated Soil ....  A-9
                  A.1.6    Emissions from Soils Containing High
                           Chemical Concentrations  	 A-12
             A.2  Airborne Pollutants 	 A-14
             A.3  Relating Emissions to  Indoor Air
                    Concentrations  	 A-16
                  A.3.1    Empirical Relationship Based 	 A-16
                  A.3.2    Typical Simplified Assumptions Used  .  . A-18
                  A.3.3    Incorporating Building Impacts on
                           Infiltration Rate	A-20
                  A.3.4    Impact of Contaminated Outdoor Air .  .  . A-29
             A.4  Sources of Data for Model Parameters	A-31

-------
              APPENDIX A.  PREDICTIVE SCREENING TECHNIQUES

      Conducting  sampling of the indoor environment at each and every
potentially  impacted structure proximate to all sites would be an
enormous undertaking.  Furthermore, simply having indoor air quality
(IAQ) data for one day or even a week would likely be of limited
usefulness since  the IAQ at other times would not be known.  It is
reasonable to expect that chemical intrusion into a building would be
different at different times of the year and would also be effected by
occupant activities such as heating or cooling the building among
others.  Thus, single point data alone gives little indication as to
whether the  instant IAQ was typical for that particular structure.
      Modeling provides an opportunity to estimate the IAQ for a variety
of site conditions and structure types.  It should be recognized,
however, that model results inevetiably depend on the quality of input
data available.   Seldom is all site specific data needed by the models
actually available.  The simplifying assumptions used to substitute for
this data likely  will not be exact for any particular site. Thus, while
model results can provide reasonable estimates, one should not expect
highly accurate results.  Because modeling can give  estimates for a
range of scenarios, it can provide information to assess the likely
level of site impact on the structures and provide an informed basis
upon which to select areas for actual  monitoring.
      Presented here are some of the models available for estimating the
rate at which subsurface chemicals and chemicals in ambient air can be
transported into  the region near a building, models to estimate the
infiltration of those chemicals into the building, and models to
estimate the concentration of the chemicals in the indoor environment.

A.I   MODELS TO ESTIMATE CONTAMINANT FLUX
      Except in the case of direct groundwater intrusion into a build-
ing,  chemicals in the nearby and underlying soils  must volatilize into
the soil gas and migrate to the below grade walls  before they can
infiltrate a building.  Models in this  section describe several  ap-
proaches,  varying in complexity,  for estimating this source strength.

                                  A-l

-------
 A.1.1   Farmer Model
      The Farmer model  (1)  is  a  fairly simple  screening tool.  It was
 originally developed  for estimating  emission  rates  from covered  land-
 fills without internal  gas generation.   It  is the basis for this type
 model currently recommended in  EPA's  Superfund Exposure Assessment
 Manual  (SEAM).  The SEAMS model differs  from  the Farmer model  in that
 the SEAM model assumes  completely dry soil, contains  an explicit term
 for surface area, and an explicit term for  estimating  soil  gas concen-
 tration from waste composition.
      The Farmer model  can be used for calculating  emissions from chemi-
 cals dissolved in groundwater and for contaminated  soils.   It explicitly
 assumes that the source pollutant concentration (ie.,  chemical concen-
 tration in groundwater  or soil) is not decreased by transport of the
 chemical to the surface and the depth  to the top of the pollutant source
 remains constant.  The model considers the  flux rate of chemicals to be
 a result of Fickian diffusion of the  vapor  through  the  soil  matrix.
      By assumming complete equilibrium  is  established  between chemicals
 in the soil  gas and the groundwater and  soil and ignoring all attenuat-
 ing factors such as biodegradation, it is possible  to  estimate the
 chemical concentration  in each phase.  For  the typically low chemical
 concentrations in groundwater or soil  (the  high concentration case  is
 discussed later), the vapor concentration of the chemical in the soil
 pore vapor can be estimated assuming  the chemical  equilibrium follows
 Henry's Law and is unaffected by other components of the system.
      The soil  gas concentration at the contaminant source  (eg.  water
 table surface)  is calculated from:
where,
      Cg     =     Chemical concentration in vapor phase  (g/cm3)
      Ct     =     Liquid phase concentration of chemical  (g/cm3)
      H     =     Henry's Law constant (atm - m3/mole)
      R     =     Universal gas constant (8.2 x 10"5  atm - m3/mole-°K)
      T     =     Soil temperature ( °K)
                                  A-2

-------
       [Note:  Some  tabulations  of  the  Henry's  Law constants  use  different
 units.   Dimensionless  constants  (H/RT)  are common.   It  is  important  that
 proper  units  be  used.]
      Once  in the  vapor  phase, the  chemical diffuses  through  the  soil  at
 a rate  dependent on  the  soil porosity,  pore space geometry, the chem-
 ical's  air  diffusion coefficient, and concentration gradient  between the
 source  and  point of  exit from  the soil.  The  effective  diffusion
 coefficient (Ds) is  calculated from:

                             D P10/3
                        D. = -2fi5i-          (2)
                                PT
      where,
            DA    =     Vapor phase diffusion coefficient  in air
                        (cm2/sec)
            Pa    =     Air filled soil porosity (unitless)
            PT    =     Total soil porosity (unitless)
      The steady-state flux (J)  is then calculated from:
      where,
            C2    =     Vapor phase concentration at point of soil exit
                        (9/cm3)
            L     =     Distance from source to point of exit (cm)

      Typically, C2 is set to zero to maximize the flux.
      Expressed in complete form, the modeling equation is:
      Using the above units, J is calculated in grams per second per
square centimeter at the point of exit (ground surface or building
floor).   This only provides an estimate of the rate at which chemicals

                                  A-3

-------
are reaching the building.  Estimates of the amount of this flux
actually entering the building are needed to estimate indoor air
concentrations.  This is considered later in this Appendix.
      Because the equation ignores all possible attenuating factors, it
is likely this model overpredicts the contaminant flux.   However, due to
its simplicity, this approach provides a simple method to estimate, at a
screening level, the likely maximum rate at which chemicals would be
transported to a building.

A.1.2 Correction for Transport Resistance of Capillary Fringe 2
      There is not a clean separation between saturated soil in the
groundwater and air filled soil above.  There exists a capillary fringe
above the top of the groundwater table in which the soil pores are also
filled with water (in the upper region of the fringe, the soil moisture
content is below saturation but still elevated in comparison to the soil
above the fringe).  Because water in the fringe does not move laterally
with the bulk groundwater, chemical concentrations in the fringe are not
necessarily the same as in the bulk groundwater.  Contaminants In the
bulk groundwater must diffuse through this fringe before diffusing
through the air-filled fraction of the soil pores in the unsaturated
zone.  Because molecular diffusivities in water are characteristically
four orders of magnitude smaller than through air (eg.,  air filled soil
pores), the concentration of volatile components can be expected to be
lower at the top of the fringe as compared to the concentration at the
fringe-groundwater interface.  The presence of even a thin capillary
fringe should impede transport of volatile components from the groundwa-
ter.  Note that the Farmer model  essentially assumes this fringe is
small compared to the distance from the top of the groundwater table and
that the volatile component concentration is the same at the top of the
fringe as in the bulk groundwater.  This is a conservative assumption
that maximizes the emission flux.
      The depth of a capillary fringe is dependent primarily on the size
of soil particles.  As the particle size decreases, the depth of the
fringe increases.  Thus, for sandy soils a small fringe would be
expected, whereas for a clayey soil a substantial capillary fringe is
likely.  The correction would normally be considered only in cases where

                                   A-4

-------
 a  significant  capillary  fringe  is  expected  and when estimates made using
 the  Farmer  model  indicate  a  moderately  high  impact on  indoor air.  If
 the  Farmer  model  predicts  a  low impact,  there is  no reason to make this
 correction  (except  to  estimate  even  lower potential impacts).  If the
 Farmer model predicted high  potential impacts, application of this
 correction  would  likely  result  in  a  substantially reduced estimated
 impact but  would  likely  still indicate  on-site soil gas monitoring was
 needed.
      If the fringe is considered  to provide transport resistance, the
 total effective diffusivity  is  represented  by:
D- = L
                             L
/-
                                 dX
                                                (5)
      where,
            DT    =     Total effective diffusivity
            L     =     Distance from water table - capillary fringe
                        interface to point of surface exit of volatile
                        component (eg., ground surface or basement
                        floor)
            X     =     A vertical coordinate
                               ia/3     D  (P  (x)10/3)
        De(X)  = DA    T  *    -  + ^-^^- - }-          (6)
            Dw    =     Diffusivity of contaminant in water
            H     =     Dimensionless Henry's law constant (concentra-
                        tion in air per concentration in water)
      Note that equation 6 is similar to equation 2, but explicitly
corrects for the variable water-and air-filled porosities across the
capillary fringe.
      Above the capillary fringe, equation 5 is easily solved since PT-
PH(x)  from equation 6 becomes the same Pa  from equation  2  and the second
term on the right side of equation 6 can be assumed to be very small in
comparison to the first term.
      The problem then can be reduced further to:

                                  A-5

-------
                    D
                                        -i
                     T _
                          D
                                                  (v:
                            cap
      where,
            D,
             cap
and:
         a.
Height of capillary fringe
Effective diffusivity through capillary fringe,
f
J .
0
dx
D6(x)
-i
'1>T
IDO(X)
, f
i
dx
9eU>
                                                  -i
                                                             (8)
      where,
            LT    =     Height of the saturated portion of the capillary
                        fringe
      This is readily solved if high resolution spatial data are
available for moisture profile within the capillary fringe.  Field
measurements of soil moisture content with, for example, a piezometer
would be required for accurate determinations. For this modeling
approach, it is essential that the thickness of the saturated portion of
the fringe (LT)  be determined fairly accurately since this is the region
where it is anticipated that most of the resistance occurs.

      Using equation 6,
                      4
                                  LTH
                       A.U)
                                                 (9)
      One approach to solving the last integral in equation 8 when
adequate water-filled porosity data are not available for the region LT
to Lc,  is to assume constant water-filled porosity equal  to the average
in the saturated and unsaturated zones.
                                   A-6

-------
      The  calculated  value of DT  should  be  substituted  for  Ds in equa-
 tion 3  to  apply  the correction to the Farmer model  for transport from
 groundwater.

 A.1.3 Correction  for  Adsorption to Soil
      When making estimates of emission  flux based  on chemical  concen-
 trations in soils, it is  necessary to consider  their adsorption to the
 soil particles and how the strength of that adsorption  affects  the
 concentration in  the  soil  gas.   In this  case, the compound  will be
 distributed among soil  particles,  soil pore moisture,  and soil  gas.
      Equilibrium between  the solid and  liquid  phases for the compound
 can be  expressed  by:  3/4
                         C0 = K,Cr          (10)
      where,
            Cs    =     Concentration  in  soil
            CL    =     Concentration  in  soil moisture
            Kd    =     Slope  of  adsorption  isotherm
      Kd primarily represents adsorption to organic matter and can be
represented by:
            K,    =     K,
                         oc
      where,
            K.
             'OC
             oc
Organic carbon in soil to water partition coef-
ficient
Fraction of organic carbon in soil
From equation 1, Cg = HCL  (using the dimensionless Henry's Law con-
stant).
Thus,
                        C  =
                             _
                             k
                             &.
                     (11)
                               OC  OC
                                   A-7

-------
      This correction  is typically not  included in models based on
pollutant concentrations in groundwater because the concentration in the
liquid phase can be directly determined and related to the concentration
in the soil gas as previously described.   It is used when modeling
emissions using bulk soil concentrations.  The correction is only
applicable to cases where the chemical concentration in the soil is low
(defined as a concentration for which the  soil gas concentration
calculated from Equation 11 is less than that calculated from the vapor
pressure of the pure compound).
      This correction would normally be used only in cases where the
soil has a high organic matter content and when the simple screening
model (Farmer) does not provide clear indication of significant poten-
tial for indoor air impacts.  In these cases, application of the more
complex model may well show that there is  a low potential for impacts
and on-site monitoring is not required.

A.1.4 FAVN Model *
      Modeling presented in the preceeding parts of this Appendix is
used to estimate instaneous surface chemical flux based on instaneous
chemical concentrations in the soil or groundwater.  The FAVN model can
be used to calculate the long-term average vapor flux of a volatile
compound from contaminated soil or groundwater when the time to com-
pletely evaporate the compound exceeds the exposure period for the
calculation.  The equation upon which it is based was obtained from
EPA's Superfund Exposure Assessment Manual5 (equation 2-19).
        FAVN
             longteim
_Cg
~T
                                 2tDTHCv
                                           1/2
                                               (12)
      where,
FAVN
    iongterm
                              Time-average vapor flux to the surface
                              over period t (mg/m2-day)
                              Bulk contaminant concentration in soil
                              (mg/m3)
                              Averaging period (days)
                                   A-8

-------
             L           =     Distance from basement bottom or ground
                              surface to top of contaminated zone (m)
             DT          =     Effective diffusivity of contaminant
                              through soil pores  (m2/d)
             H           =     Dimensionless Henry's Law coefficient
             CH          =     Contaminant concentration in pore water
                              (mg/m3)
      When calculating the flux from groundwater, with a clean soil
layer between the top of the water table and the  surface, DT should  be
calculated as given in Section 3.1.2 to correct for resistance in the
capillary fringe.
      The following relationships may be of use in application of the
model:
      •      from equation 10,
            from equation 1,

                                CB

            from equation 2,

                       D P 10/3
            DT = Ds =   A * -  ,  in the unsaturated zone
A. 1.5 Emissions From Contaminated Soil 6
      For a soil column with an initial uniform vertical chemical
distribution, the top layers of soil are depleted of volatile chemicals
by the process of volatilization and diffusion, forming a decontaminated
zone at the top.  Over time, the size of the uncontaminated zone
increases and the size of the contaminated zone decreases.  Hence,  the
instantaneous emission rate, which is proportional  to the depth of  the
contaminated zone, decreases with time.  The modeling equations in  this
section are particularly suited to cases in which the soil contaminant

                                  A-9

-------
concentrations are low, the contaminant(s) have moderate to low volatil-
ities, and average emission rates over long time periods are desired.
The model cannot be used if chemical concentrations are so high that the
soil gas concentration calculated from Equation 11 exceeds that calcu-
lated from the vapor pressure of the pure compound (see next section for
that case).
      Several authors7' 8' &9 have  solved  second-order  differential
equations to simulate the transport of vapors through soil and waste.
The analytical solutions to these equations generally differ due to the
initial and boundary conditions that are applied to the equations.
These conditions are designated according to the specific situation
being modeled, i.e., the nature and distribution of contamination at the
site of interest, and the variation in source concentrations with time.
The following assumptions are made in the derivation of the following
equations:
      The chemicals are uniformly distributed throughout the soil
      column; no transport of chemicals by water movement occurs; the
      total porosity is equal to the sum of air-filled and moisture-
      filled porosity; soil properties are constant in both time and
      space; diffusion in air is the rate-controlling step, with all
      other partitioning occurring instantly; adsorption is reversible;
      and the fraction of organic carbon in soil is constant in time and
      space.

      In a one-dimensional, homogeneous,  porous medium, the generalized
mass conservation equation without convection for chemicals undergoing a
first-order biodegradation can be expressed as7:
                     at
where,
      Ca    =     Chemical concentrations in soil pores, g/m3;
      /i     =     Net degradation rate, day;
      z     =     Soil depth below the surface, m;
      t     =     Time, sec; and

                                  A-10

-------
       a     =     The effective diffusion coefficient, m/sec, given as:
                 a. =
                        H      H
Where,
      DG     =     Effective air diffusion coefficient, m2/s;
      DL     =     Effective water diffusion coefficient, m2/s;
      Da,    =     Air diffusivity,  m2/s;
                   Water diffusivity,  m2/s;
      9      =      Total  soil  porosity,  9 = 9a + 0m, cm3/cm3;
      9a     =      Air-filled  soil  porosity,  cm3/cm3;
      T      =      Exposure interval,  sec.;
      9m     =      Moisture-filled  soil  porosity,  6m =  wpb, cm3/cm3;
      w      =      Moisture content, g water/g soil;
      p      =      Soil  particle density,  g/cm3;
      pb     =      Dry  bulk density of soil,  pb =  (l-9)p,  g/cm3;
      Kd     =      Distribution  coefficient,  cm3/g; and
      H      =      Henry's  Law constant,  dimensionless.
      The initial  and  boundary  conditions used here  are kept the same as
those used by the  USEPA8 in the development of advisory levels  for PCB
cleanup:
      Initial Concentration:  Ca = (H/Kd)Cso  9 t  = 0,  z > 0
      Boundary Condition 1:   Ca = (H/Kd)Cso  @ t  > 0,  z - «°
      Boundary Condition 2:   Ca = 0         @  t > 0,  z  = 0
      Cso =  concentration in bulk soil
The solution to Equation 13 using  the stated conditions is:
                    rr           S 7        1
C               —  I    \ /""   * or ft   ( H t-\ 1/2 fl-|*£          I "\ A \
             a ~  \^^I cso * erj-\~^ I« c;   o            114;
                                if-f
                                 L 2
where erf (n) is the error  function  of  the  argument  n.
      Under the stated boundary  conditions,  Equation 14 can be used to
estimate the instantaneous  mass  flux of chemical  vapors at the soil-air
interface, Na,  as a function of  time:
                                  A-ll

-------
                     ac
      Using Equation  15, the time-average mass  flux, Na, over the
interval t has been estimated to be  (USEPA  1986):
                N~ = _   G*° _ . e-^         (16)
                 a
      If this general equation is simplified by assuming that contami-
nants move predominantly in the vapor phase(ie., unsaturated soil -
second term in a disappears); the chemical biodegradation rate is
negligible (last term in Equation 16 =  1); the total soil porosity is
equal to the air-filled soil porosity and, thus, the effective diffusiv-
ity, DG,  is represented by 6a4/3Dai, Equation 16 becomes identical to the
solution developed by the USEPA8.
      It should be noted that the solution presented above applies to
cases with an infinite depth source of  constant concentration (see
stated boundary conditions).  Thus, although the zone of contamination
is constantly receeding from the initial upper boundary, some chemical
remains at all times.  This differs from the simpler models which assume
a constant depth to the contamination.  However, applying these condi-
tions to chemical sources of finite depth, may results in an overesti-
mate of soil concentrations, emission rates, and ambient air concentra-
tions, depending on averaging times used, and is therefore conservative
but not overly so.

A.1.6 Emissions from Soils Containing High Chemical Concentrations6
      This model is presented for use when the chemical concentration in
the soil  is sufficiently high that the  soil gas concentration calculated
from equation 11 exceeds that calculated from the vapor pressure of the
compound.
      The vapor concentration in the soil pores cannot exceed the
saturation vapor concentration.  Therefore,
                                  A-12

-------
                                      - pMW
where:
      Csv    =      Saturation  vapor  concentration  (g/m3);
      p      =      Vapor pressure  (atm);
      MW     =      Molecular weight  (g/gmole);
      R      =      Gas  constant  (atm-m3/gnu>le-°K);  and
      T      =      Temperature (°K).
The chemical concentration  in soil  corresponding  to  the  saturation  point
vapor concentration  can be  estimated  from  the  knowledge  of  vapor
pressure  and air-soil  partitioning  in  accordance  with  Equation  11:

                        c   =  —C           (17)
                        C        ^           \-i-' >
where:
      Css   »      soil  concentration corresponding to the saturation
                   vapor concentration  (ng/g)
For soil concentrations greater than Css,  chemical  vapor emission rates
can be estimated by assuming  that at any  time the soil  and chemical
approach their steady-state concentrations,  (H/Kd)(Css/z):
                               ff  C
                               H  -SS.         (18)
                       dz     Kd
and a linear soil concentration profile exists in the layer of chemical-
depleted soil (USEPA 1986).  Based on the assumed linear soil concentra-
tion profile and a simple mass balance, the layer of chemical-depleted
soil is:

                                .  2£*
              	:	12	T-T,	          (19)
               —  ' "^-C   - C  )  - —C6  - —-^-C
                   j  ss    so'    «}  L at    ~>  v   ss
                   1              44  AJ
                                  A-13

-------
Substituting  Equation  18  into  Equation  19  and  integrating  the  resulting
equation over the  time interval  0  to  t, gives:
          z  =
                              -£-ea) + cLHBa} + Pb(cso - css)
The chemical vapor emission  rate  under  saturation  condition  is:
                                  dc.       HCSO
                          HA = D--^  = DG	2£
                           A    G dz     G K^z
or
          / n • /**  \  / O « /"*   _L A /"•  A  ^*  _i_Oy^*t7«-N*/^  1
          \DG Cas>  ^2PbCso + VaCss -fT + ^mCLH Pb Css>
                               	—^	, at z =0 (20)
If 9m = 0, 9T = 8a,  and  DG = 9a4/3Dai,  Equation  20  is  identical  to  the
solution developed  by the USEPA  (1986).  The  current equation  allows
estimates based on  actual soil moisture information rather than  assuming
completely dry soil as  would be  the  case using the above assumptions.

A.2   AIRBORNE POLLUTANTS
      Under appropriate meteorological conditions, airborne pollutants
from many sources,  including a Superfund site, may simultaneously  impact
the indoor air quality  of structures proximate to a superfund  site.
Considerable guidance currently  exists for the selection and use of
appropriate emission and dispersion  models.   Table A-l lists the
principal sources of information.  Additional assistance may be  obtained
from the EPA Office of  Air Quality Planning and  Standards and  from
modelers in each of the 10 EPA Regional offices.
                                  A-14

-------























c_D
Z
i— •
_J
LU
O
O
J£

TABLE A-1
ES AIRBORNE POLLUTANT
INFORMATION SOURC






LU
O
Z
LU
rr
LU
LL
LU
CC


















!•• '; DESCRIPTION
'•'.:• .
'- -.":- •
-":" -V
:';';; :
.
•"-- •

LU
H
r-

^_
o
o
en
CO
A
O
in
•«•
<
Q.
UJ













CO
O
tho conduct of Air Pathway'' Analys
1
CD
£
1
o
o
a
2
o
c
o
o>
CO
§
:c
«
O


CO
11
CO 'fS
a. a
u. O
< ^
NTGS Series:
Volume 1, Application of
Analysis for Superfund Ap

CM
o
o
en
CO
A
0
in
^l-
<
CL
UJ





CO
«
"co
T3
=

C
o
0
c
3
C
Q

£
o
ring and modeling techniques for c
Provides descriptions of monito
w
c
UJ
w
<
«
c
"5
CO
CD
m
NTGS Series:
Volume II, Estimation of
sions at Superfund Sites

CO
o
0
en
op
A
o
in
•*
<
Q.
UJ












c
Q

ing air emissions from remedial act
Provides procedures for estimat
CO
o
ir Emissions
uperfund Sit
< to
NTGS Series:
Volume III, Estimation of
from Cleanup Activities at

,_
o
o
co-
co
,1
o
t
in
<
Q.
111











m
o
*co
"O
i

g techniques for common uncontro
Provides description of modelini


nent Manua
CO
|| Superfund Exposure Asses
pr
r~
CM
CD
CO
r^
CM
O
in
-
Screening Procedures for E
Quality Impact of Stationa
CO .£>
 '
S °°
UJ rn
•T "?
< «
0. O
uj in
. . . . ^
r- a CM 1.
_ CM _ <
0 O ° uj
> 0 > *"















"3
•o
o
O
W
1
•*-
Provides a detailed discussion o
c
o
'co
o
0.
CO
s
0
CO

Industrial Source Complex
Model User's Guide

n
0
° §
en *
'J- •*
0 "
in ^
•- <
•*— /y
^ i\
o
>w
•i E
s e
a e
!*
*" -o
CO CO
Tiaximum short-term concentration
sources of toxic releases. Downio
PC-based model for estimating i
from Superfund sites and other
below).

•—
ning Toxic /
o
01
TSCREEN:A Model for Scr
Pollutant Concentration
fy
CM
O
CN
CO O
co ¥•
en *
o o
«— m
? 6
< en
Q_ CQ
UJ O-
o "
-c E
** 0
w 3
E o
» n —
— CL «
•9 •- f
Poo
°- » E
C? §0
'•S S 52
•32 —
0 ~ X
I I
J»5
S) — O
0 0 0
"•5. t: 3
S- o o
i model. The FDM is particularly a(
spears to be a better predictor of p
ind deposition than the Industrial si
S (see below).
Provides guidance on use of the
involve particle deposition. It fl|
involving fugitive dust sources «
Downloadable from SCRAM BB!


o
•o
o
to
Q
o
>
'&
£
o
jz
o
o
•o
°5
0
JO
o
CO
3

o
CM
O
en co
co in
4 ^
^- CN
o •*
in .-
T 6
< CO
Q_ CD
UJ O.

•o
C
rt-terrn area source models (FDM,
models (ISCLT, COM, VALLEY) >
Analyzes and compares five she
three long-term (rector average)


.2 CO
CO fi
E co
ll
O 3
m •*—
Area Source Dispersion Al(
|| Emission Sources at Super


CO
•— {«^
"O O ^
t: Z *"**
S »H>
00 S S
— O ^
•£ •£ s
0 °- rt\
P qi O)
"5 "o !T
CQ H S
-o >
O :=
C 
-------
A.3   RELATING  EMISSIONS TO  INDOOR AIR CONCENTRATIONS
      In the case  studies presented  in Appendix C, a wide range of
assumptions were made  for estimating the percentage of the contaminant
flux that actually entered the building.   In this section we present
estimating techniques  based  on both  empirical relationships developed
from monitoring data and models.  The models presented here are analytic
rather than numerical.  Numeric models are also available and have been
used for modeling  radon intrusion.lt should also be noted that inade-
quate field data currently exists to validate the modeling connecting
soil gas flux rate to  indoor air concentrations.

A.3.1 Empirical Relationship Based on Radon Data<2)
      Soil is believed to be the dominant  source of radon in indoor air.
In the United States,  the mean indoor radon concentration in the living
space of single-family dwellings is  believed to be about
55 Bq/m3.<10'11'12)  This conclusion is based on measurements in 1,270
homes in which there was no  basis for expecting elevated indoor concen-
trations.
      The mean concentration of radon in basements is about twice the
mean value for above-ground  living spaces.(13'U)  Alter and Oswald(13>
reported 9,000 long-term (1-month to 1-year) measurements in basements
yielding an average concentration of 520 Bq/m3 and 12,000 nonbasement
indoor measurements yielding an average concentration of 260 Bq/m3.
Cohen (14> reported an average concentration of 300Bq/m3 from 10,000
short-term (2 days to  1 week) measurements in basements and an average
concentration of 135 Bq/m3 from 34,000 short-term measurements  in living
spaces.   Although these data show higher indoor concentrations than are
believed to exist for the population as a whole because of the large
number of measurements made  in houses with suspected elevated indoor
concentrations <12), the data should be reliable as an indicator of
relative basement/living space concentrations.
      Cohen<14> reported the following data from annual  average indoor
radon measurements in the living spaces of houses across the country:
                                  A-16

-------
Substructure Type


Basement
Slab-on-grada
Crawl space
Number of Houses


266
85
84
Average"2Rn cone.
In Living Space

-------
      It is important  to  understand that the soil gas concentration
referred to here  is that  found or calculated for the immediate vicinity
of the building (defined  by Nazaroff, Jury, and Rogers (2)  as a volume
extending 2 meters laterally  from the basement walls and 1 meter below
the floor).  It is not  the soil gas concentration that would be calcu-
lated from Equation 1  of  this Appendix.

A.3.2 Typical Simplified  Assumptions Used
      Estimations of indoor air concentrations are typically based on:
where,
      Cjn    =     indoor air concentration
      E     =     contaminant infiltration rate
      Q     =     building ventilation rate

      The assumptions used to derive both E and Q are important in
arriving at a realistic value of Cin.   Note,  for example,  that  the
equation predicts that as the building ventilation rate approaches zero
(minimum is the soil gas intrusion rate unless the building is complete-
ly sealed in which case the expression is indeterminant), the indoor air
concentration approaches the concentration in the soil gas.
      As a first approximation, some modelers assume that soil gas
enters only by diffusion into the building and that
                        E = J»A*F         (22)
where,
      J     =     Contaminant flux estimated from source model (g/m2s)
                  (See Section 3.1)
      A     =     Area of building floor (m2)
      F     =     Fraction of floor through which soil gas can enter.
and

                                  A-18

-------
                        Q =  ACH .y         (23;
                             3600
where,
      ACH   =     building  air  changes per hour
      V     =     volume  of building  (m3)

      Mueller et al(15> reports  that typical ACH for single family
residences range from 0.5 to 1.5, with new or retrofitted energy-
efficient structures generally  ranging from 0.5 to 0.8 but with some as
low as 0.2.  For general  case application, choosing any reasonable value
in this range has only a  modest effect on the final predicted indoor
contaminant concentration and is easily within the likely error for the
contaminant flux rate.
      As a first assumption  for E, it is frequently assumed that for
contaminant entry into buildings with dirt floors or ventilated crawl
spaces, a reasonable conservative assumption is that F = 1 (ie, all soil
gas enters the building).   Measured data for radon(16) indicates 0.7 < F
> 1.0 for single family homes with ventilated crawl spaces.
      For buildings with  slab floors, a typical assumption mode is that
F = 0.001.  Data used to  support this is obtained from Carlos et al (17>,
who reports that the average California home has an open area of 2- to
10-cm2 per m2 of  floor  space and Grumund  et  al  <18> who concluded that
the area of cracks to total  floor space is 0.01 to 0.1 percent.  Some
modelers assume soil gas  enters only through an assumed 0.5 cm-wide
crack around the building perimeter at the slab/wall interface (note
that this approach results  in decreasing F as the floor area increases).
The technical  literature  does not support the use of the assumptions
based on percent cracked  area of the floor for the attenuation and their
use is not recommended.   For screening purposes, it is recommended that
F be set to 1  in all cases.
      With these assumptions, it is a simple matter to calculate Cjn
using the flux rate calculated  by any method from Section A.I.  Models
                                  A-19

-------
for the more complex case in which pressure coupling to the building
must be considered are given in the following section.

A.3.3  Incorporating Building Impacts on Infiltration Rate
       It is well recognized that buildings can develop negative pres-
sures  (relative to ambient pressure) as a result of temperature differ-
ences  and wind effects.  The American society of Heating, Refrigerating,
and Air-Conditioning Engineering (ASHRAE) adopted equation 24
calculate this pressure difference, AP:
                                                              (19)
                                          to
                   AP =
                                 AP,
                         (24)
where,
      PV

       w'  L

      N

and   P.,
where,
      V
      P
      h
*K
static pressure over the building, Pa
pressure difference due to thermal gradient, Pa
area of building on windward and leeward side,
m2
empirical exponent = 0.65 (Wadden & Scheff,
1983}
0.6008 V2,   V = wind velocity in meters  per
second;
0.0342 Ph (Tr TJ/V,

wind velocity, M/S
atmospheric pressure, Pa
distance from neutral pressure plane, m
outdoor and indoor temperatures, respectively,
      Evaluation indicates that pressure differences due to thermal
gradients are generally small (eg. about 1.5 Pa for a 25'F difference)
and that wind effects dominate in equation 24.  Typical values of AP are
                                  A-20

-------
 1  to  10 Pa.   It should also be recognized that there are many other
 potential  sources in building that can lead to negative (and, in some
 cases,  positive) pressures.  Among those resulting in negative pressures
 are bathroom and kitchen fans, attic fans, forced air combustion devices
 (eg.,  oil  and gas fired furnaces), fireplaces, and leaking air supply
 ducts  located in crawl spaces.  Systems resulting in positive pressures
 are typically found in certain types of commercial buildings, however,
 it should  be recognized that leaking air return ducts located in crawl
 spaces  or  attics may cause the indoor pressure to be higher than
 Equation 24  would predict.
       Soil-gas entry due to depressurization of basements and crawl -
 spaces  has been experimentally demonstrated by a number of re-
 searchers16'18'19'20.   The current level  of understanding  is that both
 diffusion  and convection contribute to vapor intrusion  and specific  site
 characteristics determine the significance of each.   A  number of
 researchers  <16.18«19.2°.21>  nave  attempted  to model the transport for
 radon.   Johnson and Ettinger22 have adapted this work and extended it to
 the case of  chemical  vapors.   This model is described in this section.
 For details  of the  derivation one  should consult the original paper.
      There  are three basic derived equations for this  model.  They
 correspond to:
       1)     Contaminant source is  infinite (with respect to modeling
             time of interest)  and  vapor infiltration is through
             cracks/opening in the  foundation
      2)     Soil  gas  transport into the building is  substantially higher
             through relatively permeable walls (eg.,concrete block
             construction  below grade)  than through foundation cracks  and
             openings.
      3)     Contaminant is located near the  building and decreases with
             time (ie,  this case  provides an  average  when the contaminant
             source  would  become  depleted over the  averaging period
             used).
      The model  equation  corresponding  to case 1  above  (section  A.3.3.1)
 is probably  the  most  useful  for  general  application.  In this case it
would be assumed  that  the distance from the  source to the  building does
                                   A-21

-------
not change with time  (ie., Lt in Equation 25 is a constant) and does not
change in composition over the time of interest for the calculation.
The equation would be used for structures with crawl spaces and slab
floor construction with solid (eg., poured concrete) below grade walls.
      The modeling equation corresponding to case 2 above  (section
A.3.3.2) is applicable to construction which uses hollow concrete block
construction below grade (including cases where the large voids are
filled with poured concrete).  This is a fairly common construction
technique for houses with basements.  Soil gas entry into homes with
this type construction is discussed in EPA's guidance for radon reduc-
tion techniques'24'.  The model formulation presented assumes an infinite
source at constant depth as does the first equation.
      The model equation corresponding to case 3 above (section A.3.3.3)
is applicable to cases where a long term average concentration is needed
and it is reasonably expected that significant changes will occur in the
mass of chemical in the soil (and, therefore, the source to building
distance will change significantly) over the time period of interest.
The model should not be applied when the contaminant is near (within 2
meters) and intrusion into the building is dominated by convective flow.
One of the other two model equations should be used to make estimates
for the time this condition exists.  This third model equation may then
be applied for the time period that intrusive flow is diffusion dominat-
ed.
      The models below require an estimate of the source vapor concen-
tration, Csource< because what is calculated is the ratio of the concen-
tration inside the building to the soil gas concentration at the source
(ie the "«" in the equations).  Two main approaches are used in vapor
transport modeling; in the first Csource is assumed to be proportional to
the residual level in the soil, and in the second Csource is  independent
of the residual level, but is a function of composition.  The former is
applicable in the limit of "low" residual levels where compounds are
sorbed to the soil, dissolved in the soil moisture, and present in the
vapor space; the latter is applicable for "high" residual  levels where
free-phase liquid or precipitate is trapped in the soil interstices.  It
is important to note that if one chooses an incorrect model for predict-
                                  A-22

-------
ing Csource,  then it is possible to over- or under-predict the actual
Csource va^ue by orders of magnitude.  Section A.I provides a limited
discussion  of  models  to  estimate  Csource for both cases.
       In the following equations,  the  symbol  "«" (alpha)  is  the  ratio  of
the contaminant  concentration in  the indoor air to  the soil  gas  concen-
tration at  the source (ie.,  oc = C
                                  building' source
A.3.3.1     Model  for  Infinite Source,  Vapor Infiltrating  Through  Cracks
      The derived  equation  is:
                                 x exp
                                      |  QHlllLcitC*
           exp
«i^) . \ *"*•  \ . [j^l Lp (_£«&-)_
:"*A««eJ   [ObalUi^rl   (Q.oilLT\ [   (DC"C*Acr«*)
                                                               (25)
where,
      a     =
      DTeff -
      AD
       building
      LT


      Qsoil


       crack
      Dcrack
        bui Iding'  source'
       overall  effective diffusion coefficient,  cm2/s
       cross-sectional  area through which contaminants
       may pass (can be approximated by area of floor
       and below grade  walls),  cm2
       building ventilation rate,  cm3/s
       distance from contaminant source to building
       foundation,  cm
       volumetric flow  rate of  soil gas into the build-
       ing
       thickness of foundation,  cm
       effective vapor-pressure  diffusion coefficient
       through  the  crack,  cm2/s
       area of  cracks/openings  through which vapors can
       pass,  cm2
      Interesting features of the equation are:
                                  A-23

-------
       (a)    (Qsoil Lcrack/DCraCk Acrack)  " *•   In this 1imit COnVBCtion is
the  dominant transport mechanism through  the basement  (building) floor
and  walls.
then,

                                    neffA
                                    L>T  AB
                                     effA
                                     T  Ag
If the  source lies directly beneath the  foundation  (LT -* 0),  then  «-*
Qsoii/Qbuiiding' which  1S  the proper result for convection-dominated
transport  of a vapor stream with concentration Csource.   If the source  is
"far" from the basement (i.e., DTeff AB/QSOJI LT -*0),  then transport  is
limited by diffusion from the source to  foundation,  and «  -* (DTEffAB/-
Qbuiiding ^T)-   Note that  these results are independent of the cracked area
of the  floor and walls.  This is because contaminant vapors are swept
into the building as fast as they are transported to the soil adjacent
to the  floor and walls.
      (b)    (Qsoii Lcrack/DCrack Acrack) -0-   In this  limit  diffusion 1s  the
dominant transport mechanism through the basement floor and walls.  When
Dreff AB Lcrack/DCrack Acrack LT * *» then diffusion  through  the  floor and
walls is the rate limiting mechanism, and  there  is  a vapor concentration
"buildup"  below the building or basement.
      (c)    Q building "*^.   This limit corresponds to a perfectly  sealed
(no ambient  air exchange) basement, Qsoil must also  approach zero,  and
the model  predicts that « -»•!; this is, the indoor contaminant vapor
concentration approaches the contaminant vapor concentration in the soil
gas.
      Sample calculations were presented for this equation using the
following  parameters:
      AB     =     7m x  10m + 2(2m x 10m) = 138 m2 =  138 x  104 cm2
      Lcrack -      6 in.  = 15 cm
                                   A-24

-------
       ^building =   7mxlOmx3mx0.5 volume exchanges/  h  =  105 m3/h
                   = 2.9 x  104 cm3/s
       Dair =       0.087 cm2/s (benzene)
       DH2° =       0.087 1.0  x  10"5 cm2/s
       H =         0.18 cm3 of H20/cm3 of air (benzene)
       6m =         0.07 g of  H20/cm3 of soil
       £T =         0.38 cm3/cm3 of soil
       Pb =         1.7 g/cm3
       AP =         1.0 Pa = 10 g/cm-s2
and  estimating Q soil from
                   Q.
                    soil
                           Hln[2ZcracJt/rcraeJt]
which  is  an analytical solution  for flow to a cylinder of length  Xcrack
and radius rcrack  located  a depth Zcrack below ground surface;  this  is an
idealized model  for soil  gas flow  to cracks located at floor/wall  seams.
ky is the  soil permeability to vapor  flow (cm2) and /t  is  the vapor
viscosity (g/cm-s).  For this sample problem, Zcrack =  2 m,  Xcrack is
taken  to  be the  total  floor/wall seam perimeter distance (34 m),  and  for
consistency rcrack is given by
       rcrack ~ ^"B/*crack
where  the ratio  ;;= Acrack/AB,  so  that 0 < TJ< 1.  For reference,  T]=  0.01
corresponds to rcrack =  4.1 cm for the values  of AB and  Xcrack given  above;
rcrack = 1 cm corresponds  to rj =  0.0025.
      The results in  graphical form are:
                                   A-25

-------
10   l«  10   10  10   10   10
                                                          Ifl".
       Pcnne*bility(an:l
10"
 LT  =o
#"" W"*  I04  M*  104
  PtnnubQity [an1]

 LT  = 100  cm
                                                              	 n • O-Oi
                                                              	 H.CLOOI
to4  io4  W  w4  w*
                                    Penne»Wiiy[cnizJ

                                   Ly   =  1000 CHI
        For a contaminant source  adjacent to the building (LT =0)  ,  « is
  proportional to the soil permeability to vapor flow .K  at ^ »10"*cm2
                                "7
                                                                        '6
  (permeable soils).  In this example,  0.001 <  « <0.01 for IO"   < K^ tlO
  which happens to fall in the  range   of values typically reported for
  radon studies (see section A. 3.1).   The results also predict that, for
  practical purposes, it can be concluded that  the effect of crack size on
  contaminant vapor intrusion rates will  be relatively insignificant in
  the limit of convective-dominated transport.
        For LT - 100 cm and 1000 cm, the  dependence  of « on  ^  is a
  sigmoidal -shaped curve, where «  becomes independent of ^  for both
  "large" and "small" soil permeabilities.   For less permeable soils,  soil
  gas flow rates are so low that vapor intrusion is governed entirely by
  the relative rates of diffusion  through the soil and foundation.   As the
  soil becomes more permeable,  the "sweeping" of contaminant vapors into
  the building by soil gas flow increases the intrusion rate.  At some
  point, however, the ability of the contaminant to diffuse from the
  contaminant vapor source to the  region  of soil gas flow limits the rate
  of contaminant vapor transport into  the building.   For highly permeable
                                     A-26

-------
 soils,  therefore,  ex becomes independent  of the  soil  permeability to
 vapor  flow  and  only weakly dependent  on  foundation properties.   « and
 the  intrusion rate become less dependent on  foundation  properties as the
 distance  to the contaminant source  increases.   This  can be seen from the
 convergence of  the curves at both high and low  values of the  soil
 permeability, and  the 77 = 0.01 and  77  = 0.001  predictions as LT  increas-
 es.  That is, the  further the source  is  from  the  building,  the  less it
 matters what the building foundation  is.   In  the  limit  of infinite
 separation  between the source and the building, diffusion through soil
 becomes the limiting transport mechanism and  the  building substructure
 does not  matter at all.
A.3.3.2     Modeling  Using  Relatively  Permeable  Foundation  Walls
      Garbesi  and  Sextro(20) conducted  a field study of a building with
basement walls  constructed  of hollow concrete blocks  which  were subse-
quently backfilled with  concrete  and coated with  asphalt sealant on  the
exterior.  No  evidence of cracks  at the wall-floor interface was ob-
served.  Their  experimental work  showed good pressure coupling between
the basement and soil which they  interpreted to  imply the entire wall
area had a measurable permeability and soil gas  could infiltrate over
this entire area.
      The Johnson  and Ettinger model for this case is:
  a =
       exp
                          D
                            'ff
                                    exp
exp  -
-l
                   (26)
      This equation is similar to equation 25 except Acrac(c is replaced
by the area of the basement walls and floor, AB, Dcrack  is  replaced by
Df,  the effective diffusion coefficient through the porous foundation
                      crack
floor and walls, and Lc    is replaced by the foundation/wall  thickness,
LF.   While eqs 26 and 25 appear similar,  they can predict quite differ-
ent results.  Equation 26 is independent of the area of cracks/openings
                                  A-27

-------
because intrusion is assumed to occur uniformly over the floor/wall
area.  For a given Qsoil, therefore, the soil gas velocity through the
floor/walls is lower for the permeable floor/wall case.  The impact of
this is that eq 26 may predict that transport through the foundation is
diffusion dominated, while for the same conditions equation 25 would
predict that it is convection dominated.

A.3.3.3  Model for Source Depletion Over Time
      Equation 25 provides a screening estimate of indoor vapor concen-
trations, but does not account for depletion of the contaminant vapor
source.  This is reasonable when short-term exposures are being estimat-
ed and it provides a conservative (upper bound) estimate for long-term
exposures.  It is not appropriate, however, when more realistic long-
term exposure estimates are desired and it is unlikely that the source
will remain constant for a long period of time. This model  formulation
considers the depletion due to transport.
      In this approach it is recognized that the separation between
contaminant source and the building increases with time due to source
depletion.  It is assumed that the rate at which a steady-state vapor
concentration profile is established is much greater that the rate at
which depletion occurs.  Implicit in this approach is the assumption
that depletion occurs first from the layers of contaminant closest to
the building floor and walls, and the mass of contaminant incorporated
in the soil disappears, beginning at the edge closest to the building.
This is a reasonable assumption for diffusion-dominated transport to the
building-soil interface, but not valid for convection-dominated trans-
port from contaminated soil adjacent to a building floor.
      For time periods less than the time required for total depletion
of the contaminant,  the average emission rate is given by:

                                  [(p2
The corresponding long-term average attenuation coefficient <«> is then

                                  A-28

-------
      (a) = 	H^AtfeAfl   I £j_  [(p2 + 2t|rt)i/2 _  p]          (28)
             ^building^source^ \ ^"cj
       While  this  approach  is  more sophisticated  than  equation  25,
 increasing sophistication  usually increases  the  amount  of  site-specific
 information  required.   More sophisticated  screening models  are usually
 also based on  additional assumptions,  and  one must be careful  to  ensure
 that these assumptions  are valid  for specific site characteristics.
       For equation  27 and  28,  the following  definitions  apply:
       pb   = soil density, g/cm3
       CR   = average residual  contaminant  level   in soil, g/g
       AHC  = thickness of  the contaminant  layer, cm
       T    = time
       L°T  =  initial contaminant-building floor  separation,  cm
                           = DgC.
                                  source'
              P  =
exp  -

A.3.4  Impact of Contaminated Outdoor Air
      All of the preceding discussion in this chapter has assumed that
the outdoor air is contaminant free.  To be conservative, we have also
ignored other building characteristics that would affect the estimated
indoor pollutant concentration.  In the following, we attempt to remedy
both assumptions.  This discussion is based on modeling equations
presented in EPA/400/3-91/003, "Introduction to Indoor Air Quality - A
Reference Manual."  (This manual is a rich source of information.)
                                  A-29

-------
      The generalized mass balance equation for  indoor air is:

dC,. = (1 - Fb)  i/C0  +  S  -  im/C,. - \_ - flFC,.          (29)
dt                   kV         kV    kV

where,
      C,. =  indoor concentration (mass/volume);
      Fb =  fraction of  outdoor concentration intercepted by the build-
            ing envelope and not mixing with indoor air (dimensionless
            fraction);
      v  =  air exchange rate (I/time);
      C0 =  outdoor  concentration (mass/volume);
      S  =  indoor source generation rate (mass/time);
      V  =  actual indoor volume;
      kV =  effective indoor volume where k is a dimensionless fraction;
      A  =  decay rate (mass/time);
      q  =  flow rate through air cleaning device (volume/time);
      F  =  efficiency of the air cleaning device (dimensionless frac-
            tion);
      T  =  time; and
      m  =  mixing factor.

      Fb is  included because not  all  outdoor air contaminants that move
into a structure reach the inside.   Some fraction, Fb, is  intercepted by
the cracks and crevices  in the building envelope which decreases the
amount that actually reaches the indoor air.
      KV, the effective  indoor volume,  is included because the actual
indoor volume available  for contaminants dispersal depends on the degree
of air circulation.  K is less than one when there is no forced mixing
and the degree of circulation depends on thermal gradients indoors.
      X, the decay rate, is included to allow accounting for losses due
to indoor chemical reactions, and adsorption of contaminants on indoor
surfaces.  Removal of contaminants by mechanical means through air
cleaning devices is dependent on flow through the device and F, the
fractional efficiency for removal of the particular contaminant.
                                  A-30

-------
      The  mixing  factor  m is  the  ratio  of  the  concentration  of  the
exiting  air  to  the  concentration  of  the indoor air.   When  the two
concentrations  are  the same,  the  air is said to  be completely mixed,  and
m equals one.   When contaminants  are exhausted directly  from their
source,  m  will  be >1.  The  complete  mixing  assumption  is often  used to
estimate average concentrations over time  periods greater  than  1 hour.
      Equation  29 was developed for  ambient air  infiltration and
additional contaminant sources and sinks inside  the structure.  However,
it appears that, with little  error,  since  soil gas comprises a  very
small fraction  of the indoor  air, the source term S could  be replaced
with the equivalent term from soil gas  modeling.
      For  the simple case where X, F, and  Fb equal  zero and m = 1,
equation 29  reduces to :
            dt
                       - c.)
kV
and the equilibrium concentration  (when dC,./dt = 0) would be given by:
            C, = C0 +   S  .
                      ki/V               (30)

then if C0 = 0 and K = 1, equation 30 is identical to equation 21.

A.4   SOURCES OF DATA FOR MODEL PARAMETERS
      The models presented in this Appendix require a variety of  input
parameters for chemical  and soil properties. Listed below are sources of
information for many of  the more important parameters.
Model Parameter
Henry's Law constant
Diffusion Coefficient
Soil Porosity
Reference Number(s)
1, 25, 26, 27, 28
1, 5, 25, 29, 31
5, 30
                                  A-31

-------
REFERENCES

1     Air/Superfund National Technical Guidance Study Series, Volume
      II - Estimation of Baseline Air Emissions at Superfund Sites.
      EPA-450/l-89-002a, August  1990.

2     Nazaroff, W.W., W.A. Jury, V.C. Rogers, "Enclosed-Space Vapor
      Models - Technical Panel Report," 14 February 1991. Prepared for
      Department of the Army as  a review of the Rocky Mountain Arsenal
      Human Health Exposure Assessment prepared by Ebasco Services, Inc.
      under contract DAAA15-88-D-024, September, 1990.

3     Karickhoff, S.W., D.S. Brown and T.A. Scott. 1979. Sorption of
      Hydrophobic Pollutants on  Natural Sediments and Soils. Water
      Research 13:241-248.

4     Karickhoff, S.W. 1981.  Semi-Empirical Estimation of Sorption of
      Hydrophobic Pollutants on  Natural Sediments and Soils.  Chemo-
      sphere 10:833-846.

5     Superfund Exposure Assessment Manual, EPA 540/1-88/001, U.S. EPA
      Office of Remedial Response, Washington, D.C.. April 1988.

6     The section is adapted from Appendix X of the Draft Remedial
      Investigation Report for the Redwing Carriers, Inc-Saraland
      Apartments Site, Alabama,  February 1992.

7     Jury, W.A., W.F. Spencer,  and W.J. Farmer. 1983. Behavior assess-
      ment model for trace organics in soil. I.  Model  description. J.
      Environ.  Qual. 12(4):558-564.

8     U.S. Environmental Protection Agency (USEPA). Office of Research
      and Development.  Exposure Assessment Group. 1986.  Development of
      advisory levels for poly-chlorinated biphenyl (PCB) cleanup.
      Washington, D.C. EPA/600/6-86-002.

9     Bomberger, D.C., J.L. Gwinn, W.R. Mabey, D. Tuse and T.W. Chou.
      1982.  Environmental Fate  and Transport at the Terrestrial-Atmo-
      spheric Interface. Symposium on models for predicting fate of
      chemicals in the environment.  American Chemical Society, Division
      of Pesticide Chemistry 184th National Meeting, Kansas City, MO.

10    Nero, A.V., M.B. Schwehr, W.W. Nazaroff, and K.L.  Revzan (1986).
      Distribution of Airborne Radon-222 Concentrations in U.S. Homes,
      Science.  234, 992-997.

11.   Cohen,  B.L. (1986).  A national survey of 222Rn in U.S. homes and
      correlating factors, Health Physics. 51, 175-183.
                                  A-32

-------
12.   Nero, A.J. Gadgil, W.W. Nazaroff,  and K.L. Revzan  (1990).   "Indoor
      Radon and Decay Products:  Concentrations, Causes, and Control
      Strategies", Report DOE/ER-0480P,  U.S. Department  of Energy,
      Office of Health and Environmental Research, Washington, D.C.,
      November 1990.

13.   Alter, H.W. and R.A. Oswald  (1987).  Nationwide distribution of
      indoor radon measurements:  A preliminary data base, Journal of
      the Air Pollution Control Association. 37, 227-231.

14.   Cohen, B.L. (1987).  "Radon:  A Homeowner's Guide  to Detection and
      Control", Consumer Reports Books,  Consumers Union, Mount Vernon,
      New York, p. 61.

15    Mueller Associates, Syscon Co., and Brookhaven National Laborato-
      ry, "Handbook of Radon in Buildings," Hemisphere Publishing Co.,
      New York, NY., 1988.

16    Nazaroff, W. W. and S. M. Doyle, "Radon Entry into Homes Having a
      Crawl space," Health Physics. 48:   265-281.

17    American Society of Heating, Refrigerating and Air Conditioning
      Engineering (ASHRAE), ASHRAE Handbook: 1981 Fundamentals. New
      York, 1981.

18    Turk, B. H..;  Prill, R. J.; Fisk,  W. J.; Grimsrud, D. T.; Moed,
      B. A.; Sextro, R.  G. In Proceedings of the 79th Annual Meeting of
      the Air Pollution Control Association; Minneapolis, MN; Air
      Pollution Control  Association:  Pittsburgh, PA, 1986; Paper No.
      86-43.2.

19    Nazaroff, W. W.; Lewis, S. R.; Doyle, S. M.; Moed, B. A.; Nero,
      A. V., Environmental Science and Technology. 1987, 21, 459-466.

20    Garbesi, Karina, and R. G. Sextro, "Modeling and Field Evidence of
      Pressure-Driven Entry of Soil Gas  into a House through Permeable
      Below-Grade Walls," Environmental  Science and Technology. 23:
      1481-1487.

21    Loureiro, C. 0.; Abriola, L. M.; Martin, J. E.; Sextro, R. G.
      Environmental  Science and Technology. 1990, 24:  1338-1348.

22    Johnson, P.  C., and R.  A. Ettinger, "Heuristic Model  for Predict-
      ing the Intrusion  Rate of Contaminant Vapors Into Buildings,"
      Environmental  Science and Technology. 1991, 25:  1445-1452.

23    U. S.  EPA,  "Introduction to Indoor Air Quality - A Reference
      Manual," EPA/400/3-91/003, July 1991.

24    U. S.  EPA,"Radon Reduction Techniques for Detached Houses,
      Technical  Guidance," EPA/625/5-87/019,  January 1988.
                                  A-33

-------
25    U. S. EPA,  "Hazardous  Waste Treatment, And Disposal Facilities
      (TSDF)   --  Air  Emission Models.  EPA-450/3/87-026.

26    Thibodeau,  L.J.,  and S.T. Hwang,  "Landfarming of Petroleum
      Wastes  - Modeling the  Air Emission  Problem. Environmental
      Progress,  l(l):42-46,  1982.

27    Macay,  D.,  and  P.J. LeinOnen,  "Rate of Evaporation of Low-
      Solubility  Contaminants from Water  Bodies to Atmosphere.
      Environmental Science  and Technology, 9(13): 1178-1180,
      1975.

28    Mackay, D.  and  W.Y. Shiu, "A review of Henry's Law Constants
      for Chemicals of  Environmental Interest". Journal of Physi-
      cal Chemistry Reference Data,10(4):1175-1199, 1981.

29    Perry,  R.H., and  C.H.  Chilton, Chemical Engineer's Handbook.
      Sixth Edition,  McGraw-Hill Book  Company, Inc. New York, NY.

30    Brady,  N. C. The  Nature and Properties Of Soils. Eighth
      Edition, McMillian Publishing  Company, Inc., New York.

31    Lyman,  W.J., Reehl, W. F., Rosenblatt, D. H., Handbook of
      Chemical Property Estimation Methods. McGraw-Hill Book
      Company, Inc. New York. 1982
                                  A-34

-------
APPENDIX B. MONITORING METHODS

-------
                            TABLE OF CONTENTS
APPENDIX B.   MONITORING METHODS
             B.I  Methods for Use Outside Buildings	B-l
                  B.I.I    Organics in Ambient Air	B-l
                  B.I.2    Particulars	B-4
                  B.I.3    Organics in Soil  Gases	B-14
                  B.I.4    Soil  Permeability to Gas  Flow	B-16
             B.2  Methods for Use Inside Buildings	B-17
                  B.2.1    Pollutant Measurement Methods  	 B-17
                  B.2.2    Air Infiltration  Rates	B-19
             B.3  Indoor Air Pollutants	B-21

-------
                     APPENDIX B.  MONITORING METHODS

      In  this  section  information  is  provided on monitoring methods
 relevant to an  indoor air  impact  assessment and types of compounds
 typically found  in  indoor  environments.  These methods include those
 applicable to measuring  pollutants outside the building which may enter,
 as well  as indoor monitoring  methods.  The general  applicability and
 procedures are described.   References  are provided  for details of the
 methods.
      The information  provides only a summary of the relevant methods.
 Successful application requires skilled professionals and rigorous QA/QC
 programs.

 B.I   METHODS FOR USE  OUTSIDE  BUILDINGS
      Pollutants outside  a  building may enter through many openings in
 the structure.  This  includes organics and particulates in the ambient
 air and  gases in the  soil  in  the  immediate vicinity of the building.
 Outdoor  measurements  needed to obtain data required in the various
 approaches of Section 2  include:
      •    Temperature
      •    Wind velocity
      •    Barometric pressure
      •    Organics concentrations  in ambient air
      •    Particulate matter in ambient air
      •    Chemical concentrations  in soil  gases
      •    Soil  permeability to gas flow
     Methods for the latter four  are described here.  It should be noted
 that  in  some cases it may be  preferable to use an indoor method outdoors
to achieve consistency in measurements.

B.I.I    Organics in Ambient  Air
     Because buildings have a relatively high rate of air exchange with
the outside air (air exchange rates of 0.5 to 1 building volume per hour
are typical for a building with windows and doors closed),  pollutant
concentrations in ambient air can be significant in an assessment.
                                   B-l

-------
Ambient air may contain many pollutants not associated with the nearby
Superfund site (e.g., auto exhaust, lawn chemicals, industrial emis-
sions, etc.).  It is important to recognize that, unless an assessment
is to be made for background risks also, the sampling method and
analytical finish need only address known or suspected site related
pollutants.
     Whichever of the following sampling methods is selected, the
equipment should be located so as to minimize collection of pollutants
from nearby sources such as automobiles, gasoline powered equipment, and
oil storage tanks.  It should be located on the windward side of the
building away from windshields such as trees and bushes.  The intake
should be about 5 feet above ground level (approximately at the midpoint
of the ground level story of the building) and 5 to 15 feet away from
the building.
     Ambient air sampling should begin between 1- and 2-hours before
indoor air sampling begins.  This is recommended because concentration
fluctuations during this period will be reflected inside the building
for several hours since the building effectively acts as an equalization
chamber.  Likewise, for monitoring periods exceeding 4 hours, consider-
ation may be given to terminating ambient air sampling up to 30 minutes
before terminating indoor sampling.
     Very complex monitoring situations can occur if ambient air
transport of Superfund site emissions is of concern.  In these cases
additional sources emitting compounds of interest may exist upwind of
the Superfund site as well  as between it and the building being investi-
gated.  In these cases multiple monitoring stations and site specific
modeling could be required to distinguish the impact of the site from
the other sources.  Information on various monitoring techniques for
these complex situations can be found in Section 4.3 of EPA 450/1-89-
002a (volume II of the Air/Superfund NTGS Series).
     As a special case,  it should be recognized that if the ambient air
pathway is the only pathway of concern for Superfund site impact from
non-particulates on the indoor air quality, there is no reason to
conduct indoor monitoring.   That is because the average long-term
concentration of Superfund site related chemicals (except possibly
                                   B-2

-------
participates)  would  be  the  same  as  the outdoor average at the building
location.

B.I. 1.1  Methods
      In  general  EPA  Methods TO-1, TO-14,  IP-1A, or  IP-IB will encompass
the  types  of organics normally associated with a Superfund site.
However, certain  specific types  of  compounds  (e.g., pesticides) may
require  specific  methods.  All "TO-x" methods can be found in "Compendi-
um of Methods  for the Determination of Toxic Organic Compounds  in
Ambient  Air,"  EPA 600/4-84-041 and  supplements EPA  600/4-87-006 and EPA
600/4-89-018.  All "IP-x" methods can be  found in "Compendium of Methods
for  the  Determination of Air  Pollutants in  Indoor Air," EPA 600/4-90-
010.  It should  be noted that the methods in the latter reference are
not  currently  certified and should  not be regarded  as officially
recommended or endorsed by EPA.  One source of these and other methods
is the U.S. EPA Air  Methods Database prepared by the EPA Environmental
Response Team.   Information about this PC-based software package can be
obtained by calling  1-800-999-6990.
     TO-14 and IP-1A Methods
     Both of these methods use pre-evacuated SUMMA® passivated stainless
steel containers  to  collect whole air samples.  The canisters can be  -
used to  collect grab samples or  time integrated samples.   For most
efforts  related to indoor air monitoring, time integrated samples are
needed.
     For time  integrated samples a flow restricting device must be used
to control sampling  rate.  A mass flow controller is preferred for this
use because a  critical orifice flow restrictor results in decreasing
flow rate into the canister as the canister pressure approaches atmo-
spheric.  Greater sampling volumes can be obtained by using a metal
bellows  type pump to boost final  canister pressures to as high as 30
psi.
     This method  allows convenient integration of samples over a
specified time period, remote sampling and central  analysis,  ease of
shipping and storing samples,  unattended sample collection,  and collec-
tion  of sufficient volumes to allow assessment of measurement precision
and/or analysis of samples by several  analytical  systems.   Contamination
                                   B-3

-------
is a critical issue with canister-based systems because the canister is
the last element  in the sampling train.  Thus, as with all sampling
systems, care must be used  in the selection, cleaning, and handling of
both canisters and associated sampling apparatus.
     Although the method has been used for many VOCs and SVOCs (i.e.,
compounds with vapor pressures greater than 10"7  mm  Hg  at  25'C),  it  has
been validated for less than 40 compounds (See Table B.I).  Furthermore,
minimal documentation is currently available demonstrating VOC stability
in subatmospheric pressure  canisters.
     One of the primary advantages of the canister methods is that the
sample volume that can be collected is not limited by the "break
through" capacity as are methods using solid sorbents.  This is particu-
larly advantageous when sampling in areas of unknown concentrations or
where concentrations are varying during sampling.
     Canister samples are analyzed using a high resolution gas chromato-
graph coupled to one or more appropriate detectors.   Although non-
specific detectors such as  flame ionization (FID), electron capture
(ECO), or photoionization (PID) may be less expensive per analysis and
in some cases more sensitive than specific detectors, they suffer from
variable specificity and sensitivity, non-positive compound identifica-
tions, and lack of resolution of co-eluting compounds.  It is strongly
recommended that specific detectors be used for positive identification
and primary quantification.  The recommended detectors are mass spec-
trometer operating in either the selected ion monitoring mode (for use
when a specific set of compounds are to be determined) or the SCAN mode
for identification of all compounds.
     Methods TO-1 and IP-IB
     These methods are based on the collection of VOCs on Tenax® solid
adsorbent.  Integrated sampling over periods up to 12 hours can be
performed.  The air to be sampled is drawn through the Tenax® loaded
cartridge by small, portable pumps.  As with any adsorption method,
knowledge of flow rate and volume sampled are important data.
                                   B-4

-------
                                            TABLE  B.I

                         VOLATILE ORGANIC  COMPOUND DATA SHEET8
  COMPOUND (SYNONYM)

  Freon 12 (Oichlorodifluororaethane)
  Methyl chloride {Chloromethane)
  Freon 114 (1.2-Oichloro-l.l .2,2-
   tetrafluoroethane)
  Vinyl chloride (Chloroethylene)
  Methyl bromide (Bromomethane)
  Ethyl chloride (Cnloroethane)
  Freon 11 (Trichlorofluoromethane)
  Vtnylldene chloride  (1,1-Oichloroethene)
  Dichlororaethane (Methylene chloride)
  Freon 113 (1.1.2-TrJchloro-1.2,2-
   trifluoroethane)'
  1,1-Olchloroethane  (Ethylidene chloride)
  cis-l,2-0ich1oroethylene
 Chloroform (Trichloronethane)
  1,2-Olchloroethane  (Ethylene dichloride)
 Methyl chloroform (1,1.1-Trichloroethane)
 Benzene (Cyclohexatriene)
 Carbon tetrachloride  (Tetrachlorcmethane)
 1,2-Oichloropropane  (Propylene
   dichloride)
 Trichloroe thy lene (Trichloroethene)
 cl s-l,3-01chloropropene (cis-1,3-
   dichloropropylene)
 trans-l.3-D1chloropropene (cls-1,3-
   Oichloropropylene)
 1,1.2-Trichloroethane (Vinyl trichloride)
 Toluene (Methyl benzene)
 1,2-Olbromoethane (Ethylene dibromide)
 Tetrachloroe thy lene (Perch loroe thy lene)
 Chlorobenzene  (Phenyl chloride)
 Ethyl benzene
 n-Xylene  (1,3-Oimetnylbenzene)
•p-Xylene  (1,4-Oimethylxylene)
 Styrene (Vinyl  benzene)
•1,1,2,2-Tetrachloroethane
 o-Xylene  (1.2-Oiraethylbenzene)
.1.3,5-Trimethylbenzene (Mesitylene)
 1,2,4-Trinethylbenzene (Pseudocumene)
.Bi-01 Chlorobenzene (1,3-Oichlorobenzene)
 Benzyl  chloride («-Chlorotoluene)
 o-01chlorobenzene (1,2-Dichlorobenzene)
 p-Oichlorobenzene (1,4-Olchlorobenzene)
 1,2,4-Trichlorobenzene
 Hexachlorobutadiene (1.1,2,3.4,4-
   Hexach1oro-l,3-butadiene)
                                          FORMULA

                                         C12CF2
                                         CH3C1
                                         C1CF2CC1F2

                                         CH2-CHC1
                                         CH38r
                                         CH3CH2C1
                                         CC13F
                                         C2H2C12
                                         CH2C12
                                         CF2C1CC12F

                                         CH3CHC12
                                         CHC1-CHC1
                                         CHC13
                                         C1CH2CH2C1
                                         CH3CC13
CC14
CH3CHC1CH2CI

C1CH«CC12
CH3CC1-CHC1

C1CH2CH-CHC1

CH2C1CHC12
                                        BrCH2CH2Br
                                        C12C-CC12
                                        C6HSC1
                                       1.3-(CH3)2C6H4
                                       1,4-(CH3)2C6H4
                                        CHC12CHC12
                                       1.3.S-(CH3)3C6H6
                                       1.2,4-(CH3)3C6H6
                                       1.3-C12C6H4
                                       1,2-C12C6H4
                                       1.4-C12C6H4
                                       1,2.4-C13C6H3
MOLECULAR
WEIGHT
120.91
50.49
170.93
62.50
94.94
64.52
137.38
96.95
84.94
187:38
98.96
96.94
119.38
98.96
133.41
78.12
153.82
112.99
131.29
110.97
BOILING
POINT (*C)
-29.8
-24.2
4.1
-13.4
3.6
12.3
23.7
31.7
39.8
47.7
57.3
60.3
61.7
83.5
74.1
80.1
76.5
96.4
87
76
KLTING
POINT CC)
-158.0
-97.1
-94.0
-1538.0
-93.6
-136.4
-111.0
-122.5
-95.1
-36.4
-97.0
-80.5
-63.5
-35.3
-30.4
5.5
-23.0
-100.4
-73.0

CAS
NUMBER

74-87-3

75-01-4
74-83-9
75-00-3

75-35-4
75-09-2

74-34-3

67-66-3
107-06-2
71-55-6
71-43-2
56-23-5
78-87-5
79-01-6

                                                           110.97
112.0
133.41
92.15
187.88
165.83
112.56
106.17
106.17
106.17
104.16
167.85
106.17
120.20
120.20
147.01
126.59
147.01
147.01
181.45
113.8
110.6
131.3
121.1
132.0
136.2
139.1
138.3
145.2
146.2
144.4
164.7
169.3
173.0
173.3
180.5
174.0
213.5
•36.5
•95.0
9.8
•19.0
•45.6
•95.0
•47.9
13.3
•30.6
36.0 '
•25.2
•44.7
•43.8
24.7
39.0
17.0
53.1
17.0
79-00-5
108-88-3
106-93-4
127-18-4
108-90-7
100-41-4


100-42-5
79-34-5

108-67-8
95-63-6
541-73-1
100-44-7
95-50-1
106-46-7
120-82-1
a  - EPA  600/4-90-010
                                              B-5

-------
     The method offers  some advantages over the canister methods but has
significant  limitations  and overall  is complex and difficult to use.
The primary,  if not only, advantage  is that the method has been validat-
ed for more  compounds than the canister methods (See Table B.2).  Note
that although small, portable (even  personal sampling) pumps may be
used, mass flow controllers are recommended to ensure flow stability.  A
10% deviation in  initial and final flow rates may invalidate data
collected using cartridges with that pump.
     Limitations  of the  method include:
     •   Breakthrough volumes of some compounds, such as vinyl chloride,
         are very low.   The lowest breakthrough volume limits the volume
         of  sample that  can be collected.  (See Table B.3).
     •   High benzene backgrounds from Tenax® are common
     •   Sensitive to high humidity
     •   Contamination with compounds of interest (e.g., benzene,
         toluene, chloroform, methylene chloride, etc.) commonly encoun-
         tered problem
     •   Variable desorption rates for compounds adsorbed on adsorbent
     •   Necessary to know approximate range of contaminate concentra-
         tions
     •   Overall  accuracy of method  is unknown.
     The analytical finish for this method is similar to that for the
canister method above.   The primary differences are that the compounds
must be thermally desorbed from the Tenax® before injecting it into the
gas chromatograph and that each cartridge can be analyzed only once.  It
is strongly recommended, therefore, that the mass spectrometer be
operated in the SCAN mode.
                                   B-6

-------
                                    TABLE B.2

                 COMPOUNDS  IDENTIFIED AND QUANTIFIED BY AUTOMATED
               GC-MS-DS PROCEDURE WITH TYPICAL DETECTION LIMITS IN
                                 FULL SCAN MODE8
                                           Quantitatlon         Detection  Limits
 Compound                                    Mass  fm/z)          	fnq)	

 perfluorotoluene  (internal  standard)*          217                     0.3
 benzene*                      '                 78                     2.6
 methyl benzene*                                 91                     2.0
 1,2-dimethylbenzene*                           106                     0.5
 1,3,5-trimethylbenzene                         120                     2.5
 ethylbenzene*                                  91                     1.6
 ethylbenzene*                                  104                     1.7
 (1-methylethyl) benzene                        105                     1.1
 butylbenzene                                   91                     0.7
 l-methyl-4-(l-methylethyl)  benzene             119                     4.0
 chlorobenzene*                                 112                     1.7
 bromobenzene                                   156                    14.1
 1,2-dichlorobenzene*                           146                    12.4
 l-ethenyl-4-chlorobenzene                      138                     2.0
 trichloromethane                      .         83                     2.7
 tetrachloromethane*                            82                     2.1
 bromochloromethane*                            130                     2.1
 bromotrichloromethane*                         163                     1.6
 dibromomethane*                                174                     4.5
 tribromomethane*                               171                     8.5
 1,1-dichloroethane*                            63                     5.7
 1,2-dlchloroethane                             62                     3.8
 1,1,1-trichloroethane*                         99                     1.7
 1,1,2-trichloroethane*                         85                     2.1
 1,1,1,2-tetrachloroethane                      31                     0.9
 1,1,2,2-tetrachloroethane                      83                     6.5
 pentachloroethane*                             167                     1.8
 1,1-dichloroethane*                            961                     6.9
 trichloroethene*                               132                     0.8
 tetrachloroethene                              166                     2.6
 bromoethane*                                  .108                     7.8
 1,2-dibromoethane*                             107                     3.3
 1-chloropropane*                               42                     1.7
 2-chloropropane*                               43                     3.4
 1,2-dichloropropane                            63                     4.0
 1,3-dichloropropane                            76                     9.6
 1,2,3-trichloropropane                         753                     4.7
 l-bromo-3-chloropropane                        158                     1.6
3-chloro-l-propene                             41                     1.6
                                       B-7

-------
Compound

1,2-dibromopropane*
2-chlorobutane
1,3-dichlorobutane
1,4-dichlorobutane
2-3-dichlorobutane*
l,4-dichloro-2-butane (cis)
3,4-dichloro-l-butane
tetrahydrofuran
1,4-dioxane
l-ch!oro-2,3-epoxypropane
2-chloroethoxyethene
benzaldehyde*
acetophenone
benzonitrile

ISOMER GROUPS
                             TABLE B.2 --  continued
                     Quantitation
                      Mass  (m/z)

                         121
                         57
                         55
                         55
                         90
                         752
                         75
                         72
                         88
                         71
                         631
                         77
                         105
                         103
1,3- & OR 1,4-dimethylbenzene
1,2- & OR 1,3-dichlorobenzene*
2- & OR 3- & OR 4-chloro-l-
  methylbenzene*

SURROGATE GROUPS AND INTERNAL STANDARDS
4-bromofluorobenzene
chlorobenzene-ds
1,4-dichlorobenzene
1,4-di f1uorobenzene
(BFB)
                         106
                         146

                         126
 95
117
150
114
               Detection Limits
               	(ng)

                     14.4
                      3.5
                      0.5
                      8.2
                      5.1
                      1.9
                      6.5
                      1.2
                      3.9
                      8.1
                      8.2
                      5.9
                      2.9
                      1.3
                      0.5
                      1.3

                      0.5
*   Compounds used  to calibrate GC-MS-DS  on a daily  basis  either by  direct
injection or on spiked adsorbent tubes.

a - EPA 600/4-90-010
                                     B-8

-------
                                     TABLE8 B.3

                BREAKTHROUGH  VOLUMES6 AND SAFE SAMPLING VOLUMES6 FOR
                                TENAX-GC AND TENAX-TA
 Acetaldehyde
 Acrolein
 Acrylonitrile
 Ally! chloride
 Benzene
 Benzyl chloride
 Bromobenzene
 Carbon tetrachloride
 Chlorobenzene
 Chloroform
 Chloroprene
 Cresol
 p-Dichlorobenzene
 1,4-Dioxane
 Ethylene dibromide
 Ethylene dichloride
 Ethylene oxide
 Formaldehyde
 Hexachlorocyclo-
   pentadiene
 Methyl bromide
 Methyl chloroform
 Methylene chloride
 Nitrobenzene
 Perch!oroethylene
 Phenol
 Propylene oxide
 Trichloroethylene
 Vinyl chloride
 Vinylidene chloride
 Xylene
Tenax-GC
Tenax-TA
breakthrough breakthrough
volume6 volume6
38*C
0.6
4
-
-
19
300
300
8
150
8
-
440
510
-
60
-
-
-
-
0.8
-
3
-
-
-
3
21
0.6
-
200
20*C
0.6
5
8
8
36
440

"27
184
13
26
570
820
58
77
29
0.5
0.6
2000
0.8
9
5
520
100
300
3
45
.06
4
177
35'C
0
2
3
3
15
200

13
75
5
12
240
330
24
35
12
0.3
0.2
900
0.4
4
2
240
45
140
1
17
.03
2
79
  Tenax-GC
safe sampling
    volumec
    38'C

    0.3
    1.7
    8.2
    130
    130
    3.5
    6.5
      4

    191
    221
     87
     26
    0.4

    1.5
    1.5
    8.5
    .03

     89
                                                                    Tenax-TA
                                                                  safe sampling
                                                                       volume0
                                                                   20'C    351C
  2
  3
  3
 14
175

 11
  5
  5
 10
230
290
 23
 30
 12
                800
  3
  2
200
 40
120
  1
 18

  2
 70
  1
  1
  6
 80

  5
  2
  2
  5
 95
130
 10
 14
  5
360


  2

 95
 18
 55

  7


 32
"EPA 600/4-90/010
Breakthrough volumes expressed as liters/gram of  sorbent.
°Safe  sampling  volume  = {[Breakthrough  volume  (L/g)]/1.5)  x  0.65 grams  of
sorbent.
dBreakthrough volumes for other  chemicals can be  extrapolated  on  the basis of
boiling points for chemicals  in  the  same  chemical  class.
                                        B-9

-------
     Other Methods for Specific  Organic  Classes
     Other site related organic  compounds of potential interest include
aldehydes and ketones, pesticides,  and polynuclear aromatic hydrocar-
bons.
     Aldehyde and ketone concentrations  may be determined by either EPA
Methods TO-5 or IP-6A.  The methods  have different sampling techniques
but similar analysis techniques  (both'use high pressure liquid chroma-
tography (HPLC)). In Method TO-5, ambient air is drawn through a midget
impinger sampling train containing  10 ml of a 2N HC1/0.05% 2,4-dini-
trophenylhydrozine (DNPH) reagent.   Aldehydes and ketones react with the
DNPH to form stable derivatives  which can be separated using HPLC.
Method IP-6A uses a prepackaged  silica gel cartridge coated with
acidified DNPH.  The relatively  high pressure drop across the cartridges
(»30 inches of water) limits flow rates  to about 1.5 liters/minute.
Some experimentation with HPLC operating conditions is necessary for
good compound separation.  Table B.4 gives sensitivities versus sampled
volume.
     Pesticides may be determined using  EPA Methods 608 (EPA 600/4-82-
057) or Method IP-8.  The latter is  preferred for the current purpose
and is briefly described.  Ambient air is drawn through a cartridge
filled with polyurethane foam (PUF)  at a rate of 1 to 5 liters/minute.
Concentrations of 0.01 to 50 /ig/m3 can be determined  with sampling
periods of 4 to 24 hours.  The PUF is recovered and extracted with an
ether/hexane mixture.  The extract is concentrated and analyzed using
gas-liquid chromatography with electron  capture detector.  Compounds
listed in Table B.5 have been determined with this method.
     Polynuclear aromatic hydrocarbons (PAHs) may be determined using
EPA Method TO-13 or IP-7.  Method TO-13  is a high volume method applica-
ble only to outdoor environments.  IP-7  is a low volume method but has
detection capabilities adequate  for  indoor air impact assessments.  The
method is applicable to PAHs adsorbed on particulates as well as in the
vapor phase.   It also can be adapted for use with a PM-10 cut-point
sampling nozzle.
     In Method IP-7, air is drawn through a filter backed by either an
XAD-2 or PUF filled cartridge.   Up to 30m3 can  be sampled at a rate of
                                  B-10

-------
                                                  TABLE B.4

                        SENSITIVITY  (ppb,  v/v) OF SAMPLING/ANALYSIS  USING
                                ADSORBENT CARTRIDGE  FOLLOWED  BY  HPLC
                                                          Sample Volume, I
Compound

Formaldehyde
Acetaldehyde
Acrolein
Acetone
Propionaldehyde
Crotonaldehyde
Butyraldehyde
Benzaldehyde
Isovaleraldehyde
Valeraldehyde
o-tolualdehyde
tn-tolualdehyde
p-tolualdehyde
Hexanaldehyde
2.5-dimethyIbenzaIdehyde
10
20
30
40
50
100    200
300
400    500   1000
1.45
1.36
1.29
1.28
1.28
1.22
1.21
1.07
1.15
1.15
1.02
1.02
1.02
1.09
0.97
0.73
0.68
0.65
0.64
0.64
0.61
0.61
0.53
0.57
0.57
0.51
0.51
0.51
0.55
0.49
0.48
0.45
0.43
0.43
0.43
0.41
0.40
0.36
0.38
0.38
0.34
0.34
0.34
0.36
0.32
0.36
0.34
0.32
0.32
0.32
0.31
0.30
0.27
0.29
0.29
0.25-
0.25
0.25
0.27
0.24
0.29
0.27
0.26
0.26
0.26
0.24
0.24
0.21
0.23
0.23 •
0.20
0.20
0.20
0.22
0.19
0.15
0.14
0.13
0.13
0.13
0.12
0.12
0.11
0.11
0.11
0.10
0.10
0.10
0.11
0.10
0.07
0.07
0.06
0.06
0.06
0.06
0.06
0.05
0.06
0.06
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.03
0.03
0.03
0.04
0.03
0.04
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.02
0.03
0.03
0.03
0.03
0.03
0.02
0.02
0.02
*0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
Note:   ppb values are measured at 1 atra and 25'C;  sample cartridge Is eluted with 5 ml acetonitrile,
       and 25 ml are injected onto HPLC column.


Note:   Maximum sampling flow through a DNPH-coated Sep-PAK* cartridge is about 1.5 L per minute.
                                                   B-ll

-------
                  TABLE  B.5
PESTICIDES DETERMINED BY GAS CHROMATOGRAPHY/
     ELECTRON CAPTURE DETECTOR (GC-ECD)
Aldrin
BHC (a-and /3-Hexa-
 chlorocyclohexanes)
Captan
Chlordane, technical
Chlorothalonil
Chlorpyrifos
2,4,-D esters
fi.fi,-DDT
fi.fi,-DDE
Dieldrin
Dichlorvos (DDVP)
Dicofol
2,4,5-Tri chlorophenol
Folpet
Heptachlor
Heptachlor epoxide
Hexachlorobenzene
Lindane (r-BHC)
Methoxychlor
Mexacarbate
Hirex
trans-Nonachlor
Oxychlordane
Pentachlorobenzeiie
Pentachlorophenol
Ronnel
                    B-12

-------
 up  to  20  liters  per minute without significant  breakthrough.  Because
 some of the  PAH  collected by the filter may  volatilize  and  be collected
 on  the adsorbent,  the two should not be analyzed  separately.  The  filter
 and adsorbent  are  extracted in Soxhlet extractors,  cleaned  up with
 silica gel column  chromatography and analyzed using gas chromatography
 (with  FID or MS  detector) or by HPLC.   In  general,  MS operated  in  SCAN
 mode is preferred  because of the positive  compound  identifications.

 B.I.2     Particulates
     Methods are available to determine total suspended participates and
 respirable (<10/xm)  particles.  In many cases only the latter will  be of
 concern.   However,  in some cases consideration  may  be given to  the
 larger particulate.   Those particles that  penetrate the building
 envelope  will  rapidly deposit on indoor surfaces.   Exposure to  these
 particulates may occur through dermal  contact,  ingestion, or inhalation
 if  resuspended by  occupant traffic or  during such normal activities as
 cleaning.
     Total suspended  particles may be  determined  by the High-Volume
 Method given in  Section  2.2 of "Quality Assurance Handbook for  Air
 Pollution Measurement System:   Volume  II - Ambient  Air  Specific Meth-
 ods," EPA 600/4-77-027a.   Ambient air  is drawn  into a covered housing
 through a filter (nominal  8 x 10 inch),  with a  collection efficiency of
 at  least  99  percent for  particle size  0.3pi, at 1.1  to  1.7 m3/min.
 Particles up to  50 jwn (aerodynamic diameter) collect on the filter
 surface.   Air containing  up to 750 #g/m3 may be sampled for up to 24
 hours at this rate.   Particulate concentration  is determined by dividing
 air volume sampled into  the weight of  particulate collected by  the
 filter.   If  should be noted that because of the high air volume sampled,
 the method is not applicable to  indoor air sampling.
     Method  IP-10A can be  used to determine the concentration of
 particulate  in both the <2.5/im and 2.5 to  10/wn  ranges.  The method
contains  both a  fixed  site monitor and a personal sampler.  Only the
fixed site monitor is  recommended here for outdoor  ambient air monitor-
 ing.  The monitor operates  on  the principal of  impaction.  The  horizon-
tal  slotted  inlet prevents  very  large  particles from entering.  Either a
one-stage or two-stage impactor  assembly can be used to trap particles
                                  B-13

-------
above  10/im or 2.5/zm,  respectively.   Participate passing the impactor
section  is collected  on a filter  for subsequent weighing.  Note that
particles in both size ranges cannot be collected simultaneously with a
single monitor.  It is possible to connect two sampling heads to the
same pump assembly to simultaneously collect in both size ranges.  The
limit  of detection for a 12 hour  sampling period at the specified
sampling rate of 4 liters per minute is 4/ig/m3.
     The sampler does not use a critical orifice to control flow rate.
Thus,  a feedback control from a mass flow meter is required to vary pump
speed  in order to maintain flow constant with ±5 percent.

B.I.3    Organics in  Soil Gases
     Knowledge of the chemical concentrations in soil gases near a
building will be important when site impact via the underground pathway
is of concern.  These situations  can arise when soil gases migrate
directly from the site (e.g., landfills), the nearby soil is contaminat-
ed, or the underlying groundwater is contaminated.  As discussed in
Section 2, it is important in these cases to determine concentrations in
soil gas very near the building (i.e., in the region between 0.5 and 1
meter from the underground walls  and floor).  Gases in this soil region
are most likely to diffuse through building cracks or be drawn into the
building as a result of building  underpressurization.
     Because soil properties are  inhomogeneous,  both vertically and
horizontally, it is expected that concentrations will vary around the
building.  Therefore, soil gas concentrations must be measured at
several locations around the building.  Following the approach used in
Appendix C of Volume  II of the Air/Superfund NTGS Series (EPA-450/1-89-
002a), it is estimated that a minimum of two sampling points should be
used on each side of the building (typical residential building).  It is
preferred that one additional sample be collected from directly under
the building (approximate center).   It may not be practical to collect
this sample for slab-on-grade or  basement constructions because a hole
would have to be made in the floor.  Efforts to collect this sample
should be made, if possible.  Buildings of this type generally have a
very permeable zone (1 to 4 inches in depth) immediately below the slab
                                  B-14

-------
 due  either  to  the  use  of  a  gravel  bed  underlying  the  slab or gaps formed
 by soil  not making  continuous  contact  with  the  slab.
      The preferred  technique  for  collecting soil  gas  samples is the soil
 vapor (ground)  probe.   The  installation  of  probes  is  simple.  A small
 diameter pipe  (=0.5  inch  OD)  is inserted into the  ground to the desired
 depth either by hammering or  driving it  down a  slightly undersized pre-
 augured  hole.   The  use  of pre-augured  holes is  recommended if soil
 permeabilities,  described below,  are to  be  measured.  Retractable
 centering probe tips should be used to keep soil  out  of the probe.  The
 probe should be withdrawn a minimum of two  inches  to  expose soil at the
 probe tip.
      A minimum  of two probe volumes of soil  gas should be extracted
 before samples  are  taken.   Because soil  gas concentrations must be
 fairly elevated if  indoor air  impact is  to  be of  significant concern,
 low  detection limits are  not  typically required.   (Soil gas intrusion
 rates  into  a building are typically less  than 5 percent of ambient air
 infiltration rates.)  However,  positive  compound  identification and
 accurate  concentrations relative  to other measured compounds are needed.
      It  is  preferred that integrated samples be collected over at least
 one  hour  when indoor air  monitoring is not  being also conducted.  If
 indoor air  monitoring is  also being conducted,  integrated samples over
 the  same  time period should be collected.   (Although  soil gas concentra-
 tions  should not vary significantly over  this time period,  some varia-
 tion  is possible and it is best to eliminate this uncertainty.)
      In most cases, canister sample collection and analysis procedures
 such  as those given in EPA Methods TO-14  or  IP-1A are preferred because
 prior  knowledge of actual  concentrations  is not required and concerns
 over  breakthrough volumes are eliminated.   However, professional
judgement is required in  the selection of technique based on known or
 suspected contamination at the site.   For example, it may be possible to
adapt  IP-7  for pesticides or IP-8 for  PAHs  if soil contamination in the
 immediate vicinity of the building is  known or suspected.   Sampling flow
rates, and,  thus, sample volumes,  would need to be kept low to  reduce
concern about drawing in surface air.
                                  B-15

-------
B.I.4    Soil  Permeability to  Gas  Flow
     The permeability of soil  to gas  flow can  be  an  important indicator
of the rate at which  soil  gas  can  enter  a building.   If the permeability
is low (i.e.,  below 10~8cm2),  it is unlikely  building underpressures
have much effect  on the  rate of intrusion and  gases  will enter primarily
by diffusion.  In this case, soil  gas concentrations  below the building
may be higher  than estimated from  modeling equations.  As the permeabil-
ity increases, however,  building pressure becomes more important and may
result in gases several  meters  from the  building  being drawn inside.
This effect reduces the  soil gas concentration  (relative to model
assumptions) at the soil-building  interface  but increases the intrusion
rate thus resulting in higher  indoor concentrations.  In essence, this
is the reason  predictive models assume a soil  gas concentration of zero
at the basement floor-soil  interface.
     Permeabilities are  easy to measure  using  the soil gas probes
discussed above.   Permeabilities should  not  be measured until after soil
gas samples have  been  obtained  because the injected  gas will distort
concentration measurements.  Permeabilities  are determined by measuring
the gas pressure  in the  probe  as a metered flow of air is passed through
the probe into the soil.   A cylinder of  compressed air, flow meters with
ranges from about 5 to 500 cm3/min., and a differential  pressure gauge
with a range of 0 to  125 pascals are needed.   Probe  pressure should be
measured at three or more  flow  rates to  obtain a good average.  The
permeability is calculated, assuming Darcy flow,  from
         KV =    Q u
                4wr2Pa
where,
         KY = Permeability, m2
          Q = Air Flow rate, m3/s
          H = Viscosity  of air, 1.83 x 10s Kg/m-s
          r =  Internal radius  of probe,  M
         Pa =  Probe pressure in pascals
     The probe tip internal diameter should  be measured to within ±0.005
inches before  inserting  into the soil and after extraction.  Method
reference is contained in  DMSA  Action LTD.,  "Review  of Existing Informa-
tion and evaluation for  possibilities of research and development of
                                  B-16

-------
 instrumentation  to determine future levels of radon ^at a proposed
 building site.", Report  INFO-0096, Atomic Energy Control Board, Ottawa,
 Canada  (1985).

 B.2  METHODS FOR USE  INSIDE BUILDINGS
     Critical measurements to be made indoors include those to determine
 pollutant concentrations and those to determine air exchange rates.
 Both are discussed here.

 B.2.1    Pollutant Measurement Methods
     The air inside a typical building may contain hundreds of compounds
 (over 3800 compounds have been identified in tobacco smoke alone).  To
 assess the potential impacts from a nearby contaminated site, it is only
 necessary to monitor for those compounds identified with the site.
 Because these compounds are also common in building materials and
 consumer products, their detection in the indoor air does not necessari-
 ly indicate their origin.
     Discussion  of appropriate sampler locations is included in Section
 2.  Principal considerations in locating samplers are:
     •   Air exchange rates in the area should be large compared to
         sampling volume.
     •   Samplers should be located in the normal breathing zone (i.e.,
         3 to 5  feet above floor level).
     •   At least one sampling location should be established on each
         level  of the building.
     •   Samplers should not be located near obstructions,  air supply or
         return  registers,  or local  pollutant sources.
     •   Exterior windows and doors should be kept closed as much as
         possible.  All  indoor doors should be open.
     •   Sampling is best conducted under conditions  that yield building
         underpressurizations of several pascals.  Low  rates of air
         infiltration and soil  gas convective intrusion may result from
         low building underpressurizations.
     In general   it is recommended that indoor air monitoring use methods
listed in the "Compendium of Methods for the Determination  of Air
                                  B-17

-------
Pollutants  in  Indoor Air,"  EPA-600/4-90/010.   In particular  it  is
recommended that:
     •   Method  IP-1A, Appendix  B,  use  of  portable gas chromatograph, be
         used  for  initial screening investigations and to assist in
         calculating proper sample  size for methods based on use of
         solid adsorbents.
     •   Method  IP-1A, canister,  be used for collection of most target
         VOCs.   Analytical  finish by GC-MS in  SIM or SCAN mode.
     •   Method  IP-IB, Tenax®  cartridge, be used only for compounds not
         quantifiable by  IP-1A.   Analytical finish by GC-MS  in SCAN
         mode.
     •   Method  IP-6A, solid adsorbent,  be used for aldehydes/ketones.
         Method  TO-5 may  also  be  used.
     •   Method  IP-7, PUF/XAD  solid adsorbents, be used for  PAHs.  Note
         that  if the PUF  adsorbent  is used, the method may also be used
         to determine pesticides.   GC-MS in SIM or SCAN mode is the
         preferred analytical  finish.
     •   Method  IP-8, PUF adsorbent,  be  used for pesticides.  The low
         sample  volume collected  in this method may be inadequate for
         PAH quantitation in indoor and  ambient air.  GC-MS  in SIM or
         SCAN mode is the preferred analytical finish.
     •   Method  IP-lOa, size specific impactor, be used for  airborne
         particulate matter.   The single stage impactor collecting
         particles up to  10/fm  aerodynamic diameter is preferred.  The
         stationary microenvironmental  exposure monitor (MEM) is pre-
         ferred.  Collected particulates may be subjected to additional
         analysis if desired.
     Specific circumstances may also warrant collection of deposited
dusts/chemicals.  The OSHA Method,  described in 29 CFR 1910.132(a), may
be used to collect deposited material for characterization.  The method
cannot be used for air inhalation estimates.  Collection procedure is
simple.  A Whatman 42 filter is moistened with an appropriate solvent
and used to wipe a known area  (generally about 100 cm2).   The filter is
folded, with exposed side in,  and folded again to form a 90-degree angle
at the center of the filter.   The filter is placed in a clean vial  for
transport to the laboratory.   A blank filter that is also moistened with
                                  B-18

-------
 the  solvent should be submitted in  a separate vial.  Clean gloves,
 impervious  to the solvent used and  potential contaminants, should be
 worn during all  phases of collection.

 B.2.2   Air  Infiltration Rates
      Knowledge of air infiltration  rates  is not necessary to determine
 site related impacts  on indoor air  quality for the time period over
 which  sampling occurred.   However,  if any estimate is to be made of
 likely concentrations at  other times,  knowledge of the infiltration rate
 and  building air changes  per unit time during the sampling period are
 required.
      Preferred techniques to make these measurements are IP-4A and IP-4B
 which  are both tracer gas methods.   Indirect measurements can be made
 using  the fan pressurization/depressurization method (ASTM-779).
 Because  the ASTM method significantly  affects building air infiltration
 rates,  it cannot be used  immediately before (=12 hours) or during indoor
 monitoring.
     Method IP-4A can be  used  to obtain the integrated air exchange rate
 over the monitoring period.  This method  uses constant emitting perfluo-
 carbon  (PFT)  sources  which must be  placed in the building at least 8
 hours  before monitoring begins.  One source is required for about every
 500  ft2 of living area.  They  are temperature  sensitive and  care must be
 used in placement.  Samples are collected on activated charcoal
 spherules inside capillary adsorption  tubes over the entire monitoring
 period.  The  tubes are  analyzed by  GC-ECD.  The method includes cleanup
 for  other compounds that  respond to  the ECD detector.
     The method  is applicable  to small and large buildings and can be
 used to determine  interzonal flows  as  well as  exchange rates.   The
method assumes the effective exchange  volume of the building is equal to
the physical  volume.  Poor mixing within the building may effect this
assumption.
     Method  IP-4B  is  a  tracer  concentration decay method.   In  this
method, a tracer, such  as SF6,  is injected into  the  structure,  thorough-
ly mixed and  its concentration measured over time.   The concentration
decreases due  to air  exchange  with outside air.   Tracer concentrations
can be measured  on-site with a GC-ECD  or SF6 specific detectors.   Grab
                                  B-19

-------
samples or time-integrated samples for off-site analysis can also be
obtained.
     Adequate mixing  in the indoor environment is critical to the
success of this method.  Local fans and/or heating system fans should be
used to obtain tracer concentration throughout the structure within 5
percent of the average.  It is preferred that measurements be made in
several locations on each floor of the building.  Tracer injections can
be made at about one-hour intervals to obtain a series of infiltration
measurements.  If sufficient difference exists in meteorological condi-
tions during these measurements,  it is possible to characterize the
effects of temperature, wind,  and structural factors on the infiltration
rates.
     Tracer concentration is assumed to follow
                            C = C0 exp (-IT)
     where,
         C   =  Concentration at time T
         C0  =  Concentration at T = o
         I   =  Air exchange rate
         T   =  Time
The air exchange rate can be obtained from the slope of a LnC versus
time plot.
     When using either Method IP-4A or 4B, measurements should be made
of indoor and outdoor temperatures, wind velocity, barometric pressure,
and building underpressurization.
     In the ASTM Method, a large fan is mounted on a door connected to
the outside.   The fan is used to draw air out of and force air into the
building at several differential  pressures.  Flow rates are determined
from a fan calibration curve.   The effective leakage area is calculated
from
                            L - Q/(2APp)-°-5
     where,
         L   =  Leakage area
         Q   =  Air flow rate
         AP  =  Pressure difference across building shell
         p   =  Density of air
                                  B-20

-------
      Infiltration rates and air exchange  rates  can  then  be  estimated
 using
            Q = L(AAT + B/i2)0"5 from  page 8 of  EPA  400/3-91/003
     where,
          Q    =  Air infiltration rate, CFM
          L    =  Effective leakage area, in2
          AT   =  Indoor-outdoor temperature difference,  °F
          /x    =  Wind speed,  mph
          A    =  Stack coefficient
          B    =  Wind coefficient
 Values  for A and B are in the above document.

 B.3   INDOOR  AIR POLLUTANTS
     Literally hundreds of compounds  have been  detected  in  indoor air.
 Over 3,800 compounds have been detected in tobacco  smoke alone.  The
 sources of the compounds are many and  include building materials,
 furnishings,  cleaners and waxes,  paints,  pesticides, hobby  supplies,
 combustion devices,  and personal  care  products.   It has  been estimated
 (Tancrede et al,  "The Carcinogenic  Risk of Some Organic  Vapors  Indoors:
 A Theoretical  Study",  Atmospheric Environment,  Vol. 21,  No. 10,  1987)
 that cancer  risks from exposure to  indoor air probably well exceeds 1 x
 lO'4.
     Many of the compounds typically  associated with a contaminated site
 are also  present in  typical  non-impacted  indoor environments.  This
 significantly  complicates monitoring  efforts to differentiate site
 contributions  to indoor air  concentrations.  The data currently  avail-
 able indicates  that  compounds  detected and concentrations of those
 compounds is  a  function of many factors including building  type  (e.g.,
 office, residence,  school, etc.), age, and location.
     Although  many  studies of  indoor  air  exposures have been made (most
 notable is The  Total  Exposure  Assessment  Methodology (TEAM) Study - see
 EPA 600/6-87-002Q  for  summary  and analysis), it does not appear  to be
 feasible at  this  time  to  tabulate the averages and ranges for individual
 compounds in all  types  of potentially impacted structures in a way that
would be useful  for  determining  site  impacts from indoor air concentra-
 tions.  The  TEAM  studies  do,  however, have a substantial amount  of
                                  B-21

-------
information on averages and ranges for many compounds in a number of
different building types.  This may be considered the best currently
available source of information on typical concentrations.
     Examples of some specific compound measurements are contained in
EPA 400/3-91-003, "Introduction to Indoor Air Quality:   A Reference
Manual."  EPA 600-R-92-02, "Indoor Air Quality Data Base for Organic
Compounds," provides the best current summary of data on organics from
studies over the last decade.  This data base contains  information on
over 220 organic compounds.  The compounds most frequently reported in
the studies reviewed were:  formaldehyde, tetrachloroethylene,  1,1,1-
trichloroethane, trichloroethylene, benzene,  p-dichlorobenzene,  toluene,
ethylbenzene, xylene, decane, and undecane.  It should  be noted  that not
every study monitored for the same compounds  or used similar methods.
                                  B-22

-------
APPENDIX C.  CASE STUDIES

-------
                            TABLE OF CONTENTS


APPENDIX C.  CASE STUDIES
             C.I  Buildings on Landfill Sites 	  C-l
                  C.I.I    Quality Assurance/Qua!ity Control  .  .   .  C-5
                  C.I.2    Results	C-7
                  C.I.3    Conclusion Reached	C-9
                  C.I.4    Comments on The Study	C-9
             C.2  Buildings Near Landfill	C-ll
                  C.2.1    Nature and Extent of the Problem .... C-13
                  C.2.2    Prevention of Gas in Structures  .... C-20
                  C.2.3    Comments on the Study	C-20
             C.3  Buildings Over Groundwater Plume (A)	C-21
                  C.3.1    Nature and Extent of the Problem .... C-22
                  C.3.2    Indoor Air Modeling Results  	 C-30
                  C.3.3    Comments on Case Study	C-30
             C.4  Buildings Over Groundwater Plume (B)	C-35
                  C.4.1    Nature and Extent of Problem	C-37
                  C.4.2    Conclusions	C-41
                  C.4.3    Comments on Case Study	C-42
             C.5  Buildings Over Groundwater Plume (C)  	C-44
                  C.5.1    Nature and Extent of the Problem .... C-44
                  C.5.2    Results	C-48
                  C.5.3    Conclusion	C-50
                  C.5.4    Comments on Study	C-51
             C.6  Buildings Near Sludge Disposal  Pit	C-53
                  C.6.1    Nature and Extent of the Problem .... C-53
                  C.6.2    Residential Air Monitoring  	 C-57
                  C.6.3    Comments on Indoor Air Effort	C-58
             C.7  Fugitive Dust Case Study	C-60
                  C.7.1    Nature and Extent of the Problem .... C-60
                  C.7.2    Comments on the Study	C-66
             C.8  Buildings Over Waste Oil  Pit	C-67
                  C.8.1    Nature and Extent of Problem	C-69

-------
                                APPENDIX  C
                               CASE  STUDIES

      The  potential  impacts  of Superfund  sites on the  indoor  air  quality
 of  nearby buildings  have been investigated  in a number of cases.   This
 Appendix  documents  a sample of such  cases.  The examples cover a  range
 of  site conditions,  potential  types  of indoor air  impacts, phases  in  the
 Superfund program,  and  judgements of investigating personnel.
      It should  be noted that,  in most cases, additional work beyond that
 presented here  has  been conducted.   Therefore, this Appendix should be
 read  in the  context  that it presents partial results  of investigations
 and should not  be construed as the  final analysis of  the potential site
 impact.   The purpose is to  provide  an overview of the modeling and
 monitoring techniques that  have been used and the relative success of
 these techniques.
      It is not  intended as  a  revie of all possible case types.   It is
 also not  intended as documentation that  indoor air impacts have been
 shown to  occur.  It  is  hoped  that sufficient detail is provided to allow
 site investigations  to  avoid  some of the major pitfalls in assessing
 indoor air impacts.

 C.I  BUILDINGS ON LANDFILL  SITES
     The  Landfill extends over approximately ten acres of land.  It
 received  both municipal  and industrial wastes from about 1910 to 1967.
The wastes included drummed industrial wastes and uncontained liquid
wastes.    Prior to 1964,  open burning took place  at the site.   After
closure,  when the wastes were  covered by approximately two feet of soil,
the property was subdivided and sold for both residential  and commercial
development.    Based on  verbal reports of methane buildup  in  buildings
currently  located on the site, a study was conducted in three commercial
buildings  in late 1985.  The results of the study indicated the presence
                                 C-l

-------
 of methane  in  one  of the  buildings  at  levels well  below the lower
 explosive limit  for  methane  in  air,  even  in the  floor cracks where
 levels were  most highly concentrated.   However,  the concern existed that
 the methane  entering buildings  acts  as  a  carrier gas for toxic species.
 Preliminary  soil gas analysis conducted by the regulated party's
 contractor utilizing a portable chromatograph indicated the presence of
 eleven toxic chemical species.
     Therefore,  indoor air quality  sampling for  volatile organic
 chemicals was  performed at the  site.   It  took place between approximate-
 ly 9 AM  and  5  PM on  a day in September, 1990.
     The purpose of  this study was  to collect data of a quality suffi-
 cient to document  the presence of any toxic volatile organic chemicals
 in buildings constructed on the former  landfill  site which may pose a
 health risk  to their occupants.
     The buildings sampled were selected  on the  basis of an on-site
 reconnaissance visit conducted jointly  by EPA and State personnel as
well  as existing data.  They are single-family residences referred to
 here as Building A and Building B.   Both  homes are of split-foyer design
 and have garages adjacent to finished portions of their lower levels.
Aside from automobile storage, miscellaneous small gasoline engines (and
 the fuel for such) are commonly kept in such indoor areas. The garage in
 Building A,  in particular, smelled strongly of fuels and solvents (HNu
 readings were 2  ppm).  Virtually all of the tentatively identified
compounds detected are known components of gasoline.  Both homes are
 inhabited by cigarette smokers.  Although no smoking took place during
the sampling period, information on  the prior 24 hours was not obtained.

     Sampling and analytical  methodologies were tailored for purposes of
 identifying  and quantifying the compounds listed in Table C-l.
Compounds of interest were selected  based on preliminary data developed
during the soil gas  sampling conducted  in November and March and April
1989.   EPA personnel performed both  the field sampling activities and
the analytical  work.
                                 C-2

-------
                                TABLE C-l
                            Targeted Compounds
                                 A - Tenax
       Benzene
       Ethylbenzene
       Tetrachlorocthylene
       Toluene
      Trichloroethylene
       1,1,  1-trichloroethane
       m, p-xylene
      o-xylene
                               B - Spherocarb
      Trans-1, 2-dichloroethylene
      Vinyl Chloride
     Sampling and analysis of indoor air conformed to Method TO-1,
Method  for  the Determination of Volatile Organic Compounds in Ambient
Air using Tenax Adsorption and Gas Chromatography/Mass Spectrometry
(GC/MS)  and Method TO-2,  Method for the Determination of Volatile
Organic  Compounds in  Ambient Air by Carbon Molecular Sieve Adsorption
and GC/MS,  from the Compendium of Methods for the Determination of Toxic
Organic  Compounds in  Ambient Air,  EPA-600/4-84-041,  May 1987.
     One sampling station was located on each of the lower two floors of
each building.   Each  station on the lower floors consisted of primary
and secondary tubes in  series plus a duplicate (of each type of tube).
Each station  on  the upper floors,  consisting of single tubes of each
type, was located upstairs on the  main living level  in the kitchen.   The
lower-level  station in  each building was expected to produce the highest
readings of any  of the  stations.   Because,  the possibility of break-
through of  the  species  through the cartridges,  with  resultant loss of
accuracy, was greatest  at this point,  the series tube configurations
were utilized there to  gauge breakthrough.    A single cartridge of each
                                 C-3

-------
 type was  also  set  up  outdoors  to monitor ambient concentrations, for
 comparison with  those measured  indoors.
     Samples were  collected over an 8-hour period, utilizing personal
 constant-flow  pumps to draw ambient air through cartridges containing
 adsorbents.  Certain  nonpolar  volatile organic compounds having boiling
 points  in the  range of approximately 25 to 95 *C were captured on Tenax
 GC  (poly  2,6-Diphenyl  phenylene oxide).  Other more volatile compounds
 having  boiling points ranging  from -26 to +50'C (which will pass through
 Tenax adsorbent) were collected on a carbon molecular sieve (CMS)
 adsorbent, Spherocarb.
     A  sample  volume  requirement of approximately 10 liters dictated a
 flow rate of approximately 21 ml/min over the 8-hour sampling period.
 Sampling cartridges were positioned in the "breathing zone" (approxi-
mately  three feet  above floor  level).
     Occupants were requested to eliminate, insofar as is possible, the
opening of any doors  or operation of ventilation/exhaust fans (clothes
driers  act as exhaust  fans) for a minimum of 24 hours preceding the
sampling period.  Windows remained closed for 24 hours preceding
 sampling.  Door openings were kept to a minimum during the sampling
period.
     During the eight  hour period while the pumps were running, methane
levels were measured  utilizing an Organic Vapor Analyzer (OVA)  in the
gas chromatograph mode, with the output going to a strip-chart recorder.
An Hnu  (which does not respond to methane) was used for total  non
methane readings.  The highest methane readings obtained were 8 ppm
 (directly above a cat  litter pan).
     Temperature, barometric pressure and relative humidity (RH) were
measured periodically  inside the buildings throughout the eight-hour
sampling period.  Temperature readings were obtained with an Ertco
thermometer,  pressure with a Taylor aneroid barometer and relative
humidity with a Bacharach Instrument sling psychrometer.  Inside
temperatures  were fairly constant at 22-23'C and RH was constant at 42-
44 percent.   The test  report did not give barometric pressure.
                                 C-4

-------
 C.I.I    Quality  Assurance/Quality  Control
     Cartridge Conditioning
     Cartridges  were  thermally  conditioned  in  a Tekmar Model  5100
 Thermal  Conditioner two  days  prior to  sampling with  a nitrogen  purge
 rate of  approximately 100  ml/min.   Tenax cartridges  were conditioned for
 14 hours  at 250°C.  Spherocarb  cartridges were conditioned  for  15 hours
 at 399°  C.
     After the Tenax  cartridges were conditioned, they were  injected
 with 5 micro!iters of a  surrogate  solution, composed of deuterated
 Benzene  (d6), Toluene (d8) and  bromofluorobenzene in methanol,  using the
 flash vaporization technique.   The cartridges were then refrigerated
 prior to  field sampling  activities.

     Pump Calibration
     Prior to sampling,  the pumps  were calibrated using a Mini-Buck
 commercial primary standard flow meter with optical  sensing of  bubble
 passage.  Following sample collection, calibrations were again  checked
with the Mini-Buck.   The criterion  for the difference between the
 initial and final flow rates  is no  more than 15%.  Where the difference
 exceeds 15%,  the data  is reported  as an estimated value.  Four  pumps
exhibited a flow rate difference of more than 15%.

     Sample Storage and Transportation
     -  Cartridges were stored  in  individual containers following
conditioning for transportation and storage.  The individual containers
were sealed in jars containing granulated charocal to minimize  the
possibility of cross-contamination.
     -  Cartridges were stored  in  a clean laboratory freezer maintained
at approximately -20 to 0* C following conditioning,  and were returned
to said freezer following sample collection while awaiting analysis.
     -  Cartridges were transported to and from the field in an ice-
filled cooler.
                                 C-5

-------
      Blanks
      One cartridge of each type used in  the sampling  process accompanied
 the  sampling cartridges to the  field and was handled  in precisely the
 same manner as the sample cartridges,  except that  they remained  in their
 containers  to serve as field  blanks.
      A  laboratory blank of each type used remained  in the freezer and
 was  subject to the same criteria as  the  field blanks.
      The results  of the analysis of  the  Tenax trip  blank showed  it to be
 contaminated with two compounds,  1,1,1-trichloroethane and benzene.  The
 Tenax laboratory  blank was found to  be contaminated with four compounds,
 1,1,1-trichloroethane,  benzene,  toluene  and 1, 2, 3-trichlorobenzene.
 All  of  the  sample cartridge data for benzene,  with  two exceptions, met
 the  criterion and were considered  valid  from the standpoint of blank
 contamination.  All  of the Tenax cartridge  data for toluene met the
 criterion for blank contamination.
      No  contamination was  found  on either the trip  or laboratory
 Spherocarb  cartridges.

      Breakthrough
      Breakthrough of compounds  through the  adsorption medium was
 assessed by a comparison of the  levels found  on the primary-secondary
 cartridge pairs.   Secondary, or  backup cartridges must meet the criteri-
 on of containing  less  than 20%  of the  amount  of each species found on
 the  primary cartridge  of each pair.  At  Building B, two compounds failed
 the  criterion.    At  Building A,  all  the  components  failed the criterion
with  the exception  of benzene.   None of  the  Spherocarb primary-secondary
 pairs exhibited any  breakthrough.

      Precision
      Each sampling  event consisted of the collection of an additional
 set  of parallel samples collected simultaneously at different flow
rates.  Agreement  between  parallel samples  should generally be within ±
25%  if the  concentration is at  least 10  times  greater than the minimum
detection level.   None of  the targeted Tenax  compounds detected at this
level failed  the  criterion.  Methylene chloride was the only compound
detected on  Spherocarb which failed  the  criterion.

                                 C-6

-------
C.I.2   Results
     Much of the chemical concentration data from this study were
qualified because of:
     •  Four pumps not meeting flow-rate acceptance criterion
     •  Poor recovery on two cartridges
     •  Blank contamination for 1,1,1-Trichloroethane and benzene
     •  Breakthrough on primary cartridges using Tenax (precision
        criteria were met however).
     Summarized results are in Table C-2.
                                C-7

-------
TABLE C.2
Maximum Concentrations of Targeted Compounds Found (1) (PPBV/V)
A - Tenax

benzene
ethylbenzene
tetrachloroethylene
toluene
trichloroethylene
1,1, 1-trichloroethane
m,p-xylene
Building B
5.9
2.6
0.34
19*
0.04*
3*
8.4®
Building A
17*
7.7*
0.52
56*
ND
5.2*
25®*
Ambient
ND(3)
ND
ND
ND
ND
ND
ND
* - Estimated Value
B - Spherocarb
trans- 1 ,2-dichloroethylene
vinyl chloride
ND
ND
ND
ND
ND
ND
(1) - 18 other compounds were detected; the highest was 15 ppb V/V for isopropylbenzene
(2) - Reported as total o,m,p-xylenes
(3) - ND - not detected in any sample
                                   C-8

-------
 C.I.3    CONCLUSION  REACHED
     The test  report  stated,
     "In summary, all  eight  of the  targeted compounds detected in the
 homes  studied  here  have  potential sources apart from the underlying
 landfill  and are, in  fact, not uncommon  in a household environment.
 Therefore,  it  is  impossible  to determine with complete certainty which
 compounds  are  actually migrating  into the homes from past landfill ing
 and which  are  present  merely  as a result of typical household activi-
 ties.
     Finally,  whether  or not  these  levels constitute a health hazard
 must be  determined  by  a  risk  analysis."
     The complete test report  was submitted to the residents along with
 a cover  letter that states,  in part:
     For Building A;
     "The  levels of chemicals  detected in your home were comparable to
 those  found in the  average home.
     At  this time we do  not know the origin of these chemicals.   As we
 are planning to complete  the Remedial Investigation of the landfill  in
 the summer of  1992,  we hope, at that time to determine the source of
 these  indoor pollutants."
     For Building B;
     "The  levels of chemicals  detected in your home were found to be
 slightly higher than those normally found in an average home.  However,
 the levels detected in your home do not pose an immediate health  threat.
     At this time we do not know the origin of these chemicals.   As  we
 are planning to complete  the Remedial Investigation of the landfill  in
the summer of  1992,  we hope,  at that time to determine the source of
these  indoor pollutants.
     The chemicals identified  in your home at slightly elevated  levels
are:  benzene,  ethylbenzene, toluene, and xylenes.   Apart from the
landfill, possible sources of  these indoor air pollutants are numerous."

C.I.4   Comments On  The Study
     This case  provides a good example of some of  the pitfalls of indoor
air monitoring.  The study met  its primary objective which was to
"document the  presence of any  toxic  volatile  organic chemicals in

                                C-9

-------
buildings	which may pose a health risk to their occupants." However,
because the  indoor air concentrations of chemicals of the same identity
as those associated with the Superfund site were similar to typical
indoor concentrations, no conclusive statements about the source of the
chemicals could be made.  This would be true even if all sampling and
analysis had gone perfectly.
     In retrospect, because the study design was such the contribution
of site chemicals to the indoor air concentrations could not be deter-
mined,  it would have been preferable to state the purpose as: Determine
whether or not indoor air contaminants were of types and concentrations
to suggest significant intrusion of site related chemicals.  With this
as the purpose, it would have been easier to develop a risk communica-
tion plan that had specific follow-up steps depending on the contaminant
concentrations determined. Building occupants could, then,  have been
advised before monitoring took place what actions would be taken
depending on the concentrations found in this screening study.
                                C-10

-------
 C.2   BUILDINGS NEAR LANDFILL
      The Landfill  is in South King  County,  Washington.   Puget  Sound  is
 slightly more than a mile to the  west.   Residential  areas  surround the
 site,  with  the exception of a commercial  strip  along Highway 99  to the
 west  and a  stand  of second-growth mixed  conifer-hardwood forest  on the
 north.   Two elementary schools and  a  community  college  are within one
 mile  of  the site.   Interstate 5 (1-5) borders the  site  on  the  east.
 Approximately one  mile east of 1-5  is the Green River,  which meanders
 north, becomes the Duwamish River,  and enters Puget  Sound.  Figure C-l
 shows  the location of the landfill  and the  landmarks in its vicinity.
      From 1945 to  1966,  the site  was  operated as a gravel pit.   The  pit
 originally  was adjacent  to  a peat bog lake  located northeast of  the
 center of the present  landfill.   As the  pit was mined,  water was drawn
 from  the lake to wash  silt  and  clay from the gravel  and sand,  then the
 water was returned to  the lake.   Silt and clay  built up on the lake
 bottom.   Near the  end  of the operation of the gravel  mine the  barrier
 between  the  lake and the gravel pit was broken, allowing the silty lake
 water to  flow into the gravel  pit.  As a result, a clay/silt layer
 underlies much, but not  all,  of the landfill.
      In  January 1966, the 60-acre site was leased and used as  a landfill
 for nonputrescible waste, which includes organic material that decompos-
 es slowly, such as demolition debris and wood wastes.   The landfill
 received demolition debris  from commercial haulers and wood wastes and
yard trimmings.  Records  beginning in 1980 indicate  that some  industrial
wastes also were deposited.   Information included in  EPA's Emergency and
Remedial  Response  Information System (ERRIS) files indicates that the
landfill  may have  received  industrial  liquid and sludge wastes  before
1980.
                                C-ll

-------
   LANDFILL
SCALE w FEET
     500   1000
FIGURE C-1
Location Map
                                  C-12

-------
      During the course of operations at the  landfill,  an  estimated  3
 million cubic yards of solid waste were deposited  covering  approximately
 40 acres up to about 130 feet deep in places.   The east side  of  the
 landfill rises above the adjacent property;  the landfill  surface  slopes
 downward to the northwest corner.  The landfill  was  closed  in October
 1983.   The entire site was covered with 6  to 24 inches of a silt/sand
 material when operations ceased;  it now appears as an  open grassy area
 with  scattered shrubs and a few areas of exposed soil.  Approximately
 102  acres of land east of 1-5 currently drain directly into the  solid
 waste.

 C.2.1    Nature And Extent Of The  Problem
      Potential  problems related to solid waste  landfills  similar  to the
 Landfill  mainly derive from the combustible  gas that is a byproduct of
 decomposition,  and leachate,  i.e.,  water or  other  liquid  that has come
 into  contact  with  the waste.   Methane,  carbon dioxide, and small  amounts
 of hydrogen  sulfide  are normally  generated by the  decomposition of
 landfill  wastes.   Small  amounts of toxic volatile  organics are commonly
 found as well.   These gases can be released  to  the atmosphere or migrate
 offsite  by means  of  underground pathways.  Leachate can flow or seep
 offsite  to contaminate surface water  or  it can  infiltrate the groundwa-
 ter underlying  the landfill  if no artificial or natural barriers exist.
 The Landfill  is  a  natural  drainage  basin from which no surface water
 exits.   However, depending on  subsurface conditions, groundwater
 contamination  is possible.
     The  landfill was  closed  in the fall of  1983 and extensive testing
 of gas and water  in  the  landfill  and  its vicinity began.   Samples of
 leachate  and groundwater  from  monitoring wells  in and around the
 landfill  and gas samples  from  gas  probes indicated the presence of
 organic  and inorganic  contaminants with  a high potential  for offsite
migration.  In May 1986,  the EPA  placed the site on its National
 Priority  List for cleanup  and  a remedial investigation was initiated.
     Initial remedial measures were taken to control the  offsite
migration of gas and prevent possible emissions  from the  landfill
                                C-13

-------
 surface;  gas  extraction wells  were  installed  offsite  and gas migration
 control wells were installed on  the  perimeter of  the  landfill.
 Combustible Gas
      Combustible  gas,  primarily  methane  (CH4), was detected at concen-
 trations  up to 10,000  ppm in the basements  of homes near the Landfill in
 1985.  A  series of gas probes  were  installed  around the site perimeter
 to monitor gas concentrations.   Gas  was  found in  many of the probes.
 Measures  were taken to seal cracks  in  building foundations.  Numerous
 shallow and some  deep  probes were installed offsite.  A program to
 monitor homes for gas  was established.   Data  from the offsite probes and
 the monitoring program indicated that  the gas had migrated up to
 approximately 2,600 feet  from  the landfill  and was seeping into homes at
 that  distance.  Additional probes were installed  farther from the site.
 The sealing programs within homes and  businesses  was expanded, and some
 families  were evacuated.
      In September 1985,  and again in September 1986, a series of onsite
 gas migration control  wells around the perimeter  of the landfill were
 installed as  initial remedial  measures.  Designed to prevent offsite gas
 migration by  withdrawing  and venting the gas, these wells are connected
 by manifold piping  to  motor blowers  and  flares.   As part of final
 landfill  closure,  portions of  this system will be replaced by permanent
 equipment that will remain in  operation  for the indefinite future.
 Leachate
      It is estimated that  over 50 million gallons per year of surface
 water (drainage and direct precipitation) enter the landfill.   Measure-
 ments of water levels  in  leachate monitoring  wells indicate that
 stormwater discharge from drainage pipes produces rapid and significant
 increases in  water  levels within the solid waste.   Since there is no
 surface runoff from the landfill, leachate must eventually enter the
 groundwater system  if  it  does  not remain in the landfill.   Leachate
 flows to  the  southeast corner  of the landfill and then downward and then
 east  and west  away  from a mound  located  beneath the Sand Aquifer sink.
 Based on chloride concentrations, leachate  is calculated to be indistin-
guishable from background groundwater  at a maximum of 3,000 feet from
                                C-14

-------
the landfill boundary  to  the  east and southeast and 1,500 feet from the
landfill boundary  to the  west.
Offsite Gas Monitoring
     Data recorded  from the monitoring probes were computerized and used
to generate gas concentration isopleths (maps showing gradients of gas
concentrations) for the areas surrounding the Landfill.
     The isopleths  for February 1986, shown in Figures C-2 and C-3,
represent the status of landfill gas migration just after the Phase I
onsite gas migration control system was put into continuous operation.
Methane was detected northwest of the landfill in concentrations over
75%,  north in concentrations over 25%, and east and southwest in
concentrations over 40%.  Methane was detected in the deep soil  zone in
all  directions from the landfill and in concentrations over 90% in some
areas.
                                C-15

-------
       LANDFILL
    A* M« Of protttl (no
    »*»» may not M shown
0    SCO    1000
 FIGURE C-2
Shallow Gas Concentration
Isopleth, February, 1986
                                        C-16

-------
       LANDFILL
 Note AJ s«ts of (xo«$ and
    wefts may not be shown
  SCALE IN FEET
0    500    1000
 FIGURE  C-3
Deep Gas Concentration
Isopleth, February, 1986
                                                 C-17

-------
      Field  measurements were taken  using  the  following  instruments:
      •   Hnu PI-101  photoionization  detector  (total non-methane volatile
         organic  compounds)
      •   MSA 361  portable detector (combustible gas, hydrogen sulfide,
         and oxygen)
      •   Fyrite carbon  dioxide  analyzer
Gas samples were collected  in  Tedlar  bags  and shipped.  Gas samples were
analyzed  for Hazardous Substances List volatile organic compounds (HSL
VOC)  by  gas chromatography/mass spectrometry  (GC/MS) in accordance with
USEPA Method 624 guidelines.
      This analysis  found that  subsurface gas collected from the onsite
gas extraction wells and flare manifolds contained a wide variety of
substances,  including  23 USEPA HSL  VOCs.   The compounds found most
frequently  and in the  highest concentrations in the onsite subsurface
gas included  ethylbenzene,  vinyl chloride, total xylenes, toluene, and
benzene.  The maximum  concentrations  of these compounds were in the low
parts-per-million (ppm)  range.
                                TABLE C-3
             ESTIMATED ONSITE/OFFSITE ATTENUATION  OF  PRIMARY
               HSL VOCS DETECTED IN SUBSURFACE GAS SAMPLES

CAS#
100-41-4
1 33O-20-7
108-88-3
71-43-2
1 00-42-5
75-01-4

Target
Compound
Ethylbenzene
Xylenet (total)
Toluene
Benzene
Styrene
Vinyl Chloride
MAXIMUM CONCENTRATION
On-Site
(ppb)
16,610
29.195
24,044
1,384
508
31,215
Off-
Site
(ppb)
127
106
68
185
134
275
Atten.
(%)
99.2
99.6
99.7
86.6
73.6
99.1
MEAN CONCENTRATION
On-Site
(ppb)
2.825
3.419
1.920
318
41
2.607
Off-
Site
(ppb)
44
19
18
28
18
35
Atten.
(%)
98.4
99.4
99.1
91.2
56.1
98.8
Attenuation (%) = Onsite Concen. (Dob) - Offsite Concen. (ppb) x 100

Oniite Concen. (ppb)

                                C-18

-------
      Vinyl  chloride  was  frequently associated with the BTX-group
 compounds  in  onsite  subsurface gas, but during sampling offsite, it was
 found  at only two  of the gas monitor probe locations, both south of the
 landfill.
      The analysis  concluded that offsite migration of at least some of
 the contaminants consistently identified in landfill gas has occurred,
 possibly in all directions away from the landfill.  The basis for this
 conclusion  is the  presence of the BTX-group compounds in gases found
 onsite and  offsite.   However, offsite concentrations of most BTX-group
 compounds were attenuated by more than 98 percent (Table C-3).  Further-
 more, no BTX  compounds were found beyond 2,300 feet from the landfill
 perimeter.  Because  BTX-group compounds are also present in gas from
 other sources, their presence in a particular sample of offsite gas is
 not conclusive evidence  that the gas originated from the landfill;  it is
 only further  evidence that the gas may have originated in the landfill.
 Mechanisms for Gas Movement
     Where gas migration potential  exists,  the inherent slight pressure
 of gas being  generated by decomposition, fluctuations in barometric
 pressure, and fluctuations in water table may create mechanisms for gas
 movement.
     The low  gas generation pressures in the landfill initially move the
 gas away from the points of origin through the paths of least resis-
 tance.  The generation pressures are estimated to be in the range of two
 inches of water column, based upon measurements in other landfills.
 Changes in atmospheric pressure of -9.5 inches of water column are
 common during the winter months as high pressure  ridges give way to low
 pressures during storms.  This can create a relatively strong pressure
differential  between the air or gas within  the soil  or landfill  and the
 air at the ground surface.
     When barometric pressure is falling,  the air or gas within  the soil
or landfill tends to flow toward the ground surface  through the  paths  of
 least resistance.   When barometric  pressure is rising,  air from  the
ground surface tends to move down  into the  soil  both vertically  and
                                C-19

-------
horizontally.  The combined effect of the generation pressures and the
barometric changes is  a pumping action within the soils.

C.2.2   Prevention Of  Gas  In Structures
     Several residences adjacent to control well installations typically
recorded levels of methane above 10,000 ppm despite repeated efforts to
seal the foundations of those structures until the control wells were
turned on.  According  to gas monitoring data, all gas was effectively
removed from these structures within one day of the startup of the
offsite control wells.  Further, gas has remained out of structures in
the vicinity of the landfill since the implementation of the offsite
control wells.  Combustible gas above 100 parts per million has not been
recorded in a structure in the vicinity since November 1986.

C.2.3   Comments On The Study
     The study demonstrated that landfill gas migrated from the site
independently from contaminated groundwater.  It demonstrates the diffi-
culty in preventing soil gas intrusion by retrofit patching of existing
buildings and the effectiveness of soil  gas extraction wells.  The data
suggests that vinyl chloride is migrating in the direction of groundwa-
ter flow rather than with near surface gases.
     Because of the high concentrations of methane detected in the
initial phase of the investigation, few data were obtained related to
other gases that might also be migrating.  This initial  oversight, while
understandable, significantly reduced the data needed for an accurate
baseline assessment.
                                C-20

-------
 C.3   BUILDINGS OVER GROUNDWATER PLUME (A)
      The site is located at the southern end  of San  Francisco  Bay.   In
 1982  and 1983,  it was discovered that three facilities  (identified  here
 as A,  B,  and  C)  in the city had leaking underground  tanks contaminating
 soil  and groundwater.    All  three were placed on the National  Priorities
 List  (NPL).   Facility C has since been removed from  the NPL and  is  being
 regulated under  the Resource Concentration and Recovery Act (RCRA).
 Although each site has its  own  source of pollution,  the off-site
 groundwater pollution  areas have merged and the sites and are  being
 treated  as one unit.
      The area is heavily populated with commercial,  light industrial,
 and residential  use.   According to the 1980 Census information, there
 are approximately 4,000 people  living in the general area of the off-
 site  groundwater contamination:
      The off-site area is determined  by the extent of the groundwater
 contamination of volatile organic compounds.   The plume of contaminated
 groundwater extends  approximately one and three-quarters mile north of
 the site.  This  area  is composed of light industry and residential
 housing.  The former junior  high school is located west to northwest of
 the site.
     The  closest  residential  neighborhood begins about 1/4 mile north of
 the site.  According to the  1980 Census,  of the approximately 1,500
 housing  units, 54% are  single family  homes and 24% have 10 or more
 units.  The area  consists of modest single family homes interspersed
with small apartment complexes.  Within 3/4 mile of the site is a mobile
 home park which  is  primarily occupied  by senior citizens.
     A former elementary school  houses approximately 200 children per
day.   These children attend  either a  state child development program, a
state preschool,  or a YMCA day-care program and a Head Start Program,
all  operating at  the school  facility.   There  is at least a half acre of
grassy land on the side of the  facility that  is not included in the
fenced off playground.  This  is  used  as a  neighborhood playground.
                                C-21

-------
C.3.1   Nature  and  Extent of Problem
     Chlorinated volatile organic compounds  (VOCs) are the major
contaminants  tested  for and found in soils and groundwater.  Highest
concentrations  were  found around the leaking underground tanks.  The
tanks and much  of the contaminated soil were removed from the sites in
1983 and 1984.  Soils with a total VOC concentration as high as 15,700
parts per million (PPM) were excavated.  Soil boring samples were
obtained in 1988 at  the A and B Facilities.  The data are in Table C-4.

Groundwater
      The VOC contamination in groundwater is currently monitored by 29
wells on the A  site  and 30 wells on the B site, in addition to 83 off-
site wells.  There are 7 extraction wells on the A site and 6 extraction
wells on the B  site.  There are 23 off-site extraction wells.  The on-
site extraction wells have been in operation since 1985,  and the off-
site wells began operation in 1986 and 1988.
      Three major water-bearing zones (aquifers) - defined as the A-,
B-, and C-aquifer zones - exist at the site.  The A-aquifer is the
shallowest and  the C-aquifer is the deepest of these three zones.  The
B-aquifer is further divided into the Bl through B5 zones.  The approxi-
mate depths (below ground level) at which these zones occur at the A
site are as follows:  A:  10-28 ft; Bl: 28-50 ft; B3: 70-90 ft; B4: 90-
110 ft;  and B5: 110-123 ft.   A regional aquitard, the B-C aquitard, is
reportedly located at the depth range from 100-150 feet below ground
level.   The C-aquifer, which supplies most of the municipal  water in the
region,  is located below the regional aquitard.
      The aquifer zones appear to consist of mostly discontinuous layers
and lenses of fine to coarse sand, gravels, and often a substantial
proportion of clay and silt.  These predominantly discontinuous layers
and lenses are separated and/or isolated by low-permeability clays and
silts (aquitards).
                                C-22

-------
o
      CO
      CO
      en
      o
      ts>
     00
     o
     o




























<
I-H
t— 1
o
2











"§
a
c
o
re
4->
0)
o
o
U
'6
w








o
c
'5
m
a>
TJ
"to
c.









0
C
jo
0
T>
1
o







k.
TO
'5
CT






U
re
t
M
3
W







Aquitard



o
o
epnsqns






TJ
C








TJ








CN
CO
TJ
C




TJ
c



0>
TJ
J5
6
_c
•*
10
CO







CO
TJ
C








CO
CN
1
0
o
d



„.
0
o
d
TJ
c



•Dichloro-
e
31
u o>
TJ
C







CO
LO
O
d







CN
CO
o
TJ



CN
CO
CN
O
d



•oethylene
o
^
_o
'C
1-

c







CO
d
TJ
c








T>
C




0
c

0)
c
loroethyle
.c
u
re
4-»
0)
I-
n
c







CO
TJ








TJ
C




CO
CN
TJ







CN
"^
CO
o
d
V


®
richloroethan
i-
i
*
-
CD
0
0
O
TJ
C






8
o
TJ
C







05
CN
r—
TJ
C









C

«
c
^
,2-Dichloroett

U)
C
re
4-*
LO
O
cp
TJ
C





(^
CN
T—
d
i
TJ
C







O
d
CO
TJ






CO
CO
1
o
d
V



oethylene
•r
^
(j
H
if)
CN
0
O
0
TJ
C





^.
8
d
i
TJ
C







O
in
CO
TJ









c




loroethylene
JZ
u
k.
V

4-»







4^
C








CN
1
TJ









C




hlorobenzene
o
0
1
s
                                                                                                  c
                                                                                                  o
                                                                                                  •a
                                                                                                  o
                                                                                                  •8
                                                                                                  S -o

                                                                                                  S £
                                                                                                  o «
                                                                                                  •
                                                   C-23

-------
       Based on the well  monitoring results  at the site,  it  appears  the
 bulk  of the VOC contamination is confined to  the  A-,  Bl-  B2-,  and B3-
 aquifers.   However,  low concentrations  of VOCs have  been  detected in the
 B4-aquifer.   On-site VOC concentrations in  the fourth quarter  of 1990 at
 the A and  B Facilities are given in Table C-5 and C-6.

       Off-site VOC concentration in groundwater for  the fourth quarter
 of 1990 are given  in Table C-7.   The data were from  predetermined wells
 that  were  not necessarily in  the middle of  the plume.  The maximum
 concentration of trichloroethylene,  1,  2-dichloroethylene, 1,  1-dichlor-
 oethane, and tetrachloroethylene exceeded drinking water  standards.

 Soil  Gas
       On March 12,  1991,  the  soil  surface emission of organic vapors was
 analyzed using a flux chamber at off-site locations just  to the north of
 Facility A  and near  the  former Junior High  School.  The soil moisture
 content  and  ambient  temperature  during  this sampling  period were not
 typical  for  this area and would  tend to under-represent typical organic
 vapor  emission from  the  soil.  Another  sampling round was conducted on
 April  28,  1991.  One sampling location  north  of Facility A was repeated
 while  the rest of  the samples were  taken  on the former elementary school
 property.   The data  are  given in Table  C-8.   [Note: The ATSDR draft
 report  on this site  states that  data collected using  the neutral
 pressure isolation flu chamber may  be inappropriate for estimating flux
 into buildings.]

 Indoor Air Monitoring
       Indoor  air monitoring has  been conducted only at Facility B.   On
May 16, 1991,  a  crude analysis for  organic compounds was made using a
hand-held detector.   All  readings were  below  the  limit of detection (0.5
to 1.0 ppm).   On May 18,1991,  sampling  was conducted  inside and outside
the building.  One-hour  samples  were collected at each location five
feet above ground.   The  results  are  in  Table  C-9.
                                C-24

-------
U

LU
_1
OQ
      cc
      UJ
Q
Z
D
O
CC
(D
Z

CO
Z
o

<
cc


UJ
o
Z
o
o
o
o

UJ
t
CO
 I

O
               O

ration
c
c
o
o
o
o
'5
CT


S
w
00





m


0}
? I- "°
.= C o
•^ 5 -5
5*1
w







q
c-i
8
01
i





g
i
u>
00
i

§







| Dichloromethane
'
?
i





o
i
R
i

1






o
c
a
Dichloroflourome







o
i-

?

in
0






e
1 ,2-Dichloroethai
'

-------
(O
 I
U

LU
_l
m
      cc
      LU


      <


      Q
      Z
      D
      O
      CC.
CO
•z.
g
i-

cc


LU
O
•z.
o
o
o
o

LU

CO
 I

o
                CO
                U
                2



ncentration (ppbl
0
o
Q>
'5
CT




m

5





<




M
III
a *








CO
•0
c
o
1
TD





O
CM
•o
C





o
o
CM





,1,1 -Trichloroethane
r-

0
in
to
i
c
O
CM
•6
c





•O





6
o
CM
r—





CO
T—
o
V
k>»
U.


c

o
o
CO
c





CO
CO
c





in
d





inyl chloride
>

CD
•6
C
0
O
q
CN
in
•6
c




o
CO
OJ
•6





o
 T) •
                                                C-26

-------
UJ
_i
CD

f
       cc.
       UJ

       <


       Q
       Z
       D
       O
       CC
05
Z
O

<
cc


LU
U
Z
o
o

U
o

111
      u.
      LL
      o






c
o
'+-•
2
*->
CO
u
c
o
0
CO
*»•
cr
<



3
CD
2




CN
CO




»—
CO
<



V)
0 w "S
111
i« ^i n
Q > 2
u 








^f
r»'
i
TJ
C
TJ
C



O
«—
TJ
C



O3
op*
TJ
C
9
•o





b
t






omethane
c
_o
u
0

^t
o
TJ
C
CO
o
TJ
C

IT)
CO
TJ
C



CO
CN
TJ
|





b
O
f—






§
o
s-
o
o
sz
o


TJ
C
TJ
C




TJ
C




0
*—
i
TJ
C
CO
tN
TJ





jj
tn






hloroethane
o
5
*
r-

*t
cp
TJ
C
TJ
C



fc^
r^*
i
•D
C



0
o
r^
TJ
c
(^
TJ





U
O
o
CN






richloroethane
H-
r—

LT>
? —
TJ
O
O
CN
-ri
W


TJ
C




O
CO
1
TJ
C
CO
CN
TJ
C





b
o
CN
T"





n
c
0
(U
til

CO
CN
TJ
C
TJ
C




TJ
C




tn
CM
TJ
c
^
«
TJ





u
CO






hloroethylene
o
b
T—

't
T—
TJ
C
TJ
C




T)
C




O
CO
1
T)
C
co
TJ





ja
CO






hloroethylene
u
b
i
CNJ

00
CO
TJ
C
o
o
»r
CN

TJ
C
O
o
t
CN
TJ
C



O
o
*f
CM
TJ
C
O
CO
up
TJ
C





U1






oethylene
U-
O
JT
o
*c
(-


TJ
C
TJ




TJ
C




T)
C
CN
TJ
C





u
UD






loroethylene
^
u
TO
*-»
^















>
0
>
G
s.
8
*
0)
o
i
c
re
i:
*•"
w 
*~ - j
eg o
2 c a)
*•; 5 .c
Jn — ~
« JO
0) -C O
TJ 'C •*-
5^d d
2S5 5
ro re «
s - 1 1
3-1 .2-j £
« " -5 " -"5
5 S o 5 o
a ffl TJ « TJ
o £ SI £ S
c o o o o
„ S: ca *: Q.
11 « 2 « S
0 O Q.CJ Q.
Z « ^ U -Q
                                                  C-27

-------
       cc
       o
       0_
op

6
LU
_l
m
LU
H

op

LL
LL
O
       (O


       y


       LU
       I
       U

       y

       z
       DC

       O
              c
              o
              o
              a

              CO
              u.
              co
              a
       c
       CO
       £
       CO
       V)

       o


       2

       c
       0)
       u
       c
       o
       u








en

o
CD


m
|~
'o
re
LL
£
0)
X
er Junior
School
E
0
LL

CO
£
'o
re
LL
















C






c




c




c







ichloromethane
Q








c






c




c




c






o
ichlorodiflouromethan
Q








c






c




•o
c




c







,1-Dichloroethane
t—





00
co
o
00
d





CO
ID
d



CO
r-
d
•6




CO
CO*
6






,1,1 -Trichloroethane
i—





iD
co

CO
d





in
en
d



CO
TJ




O)
CD
d
d






CO
o









c






c




c




•a







,1-Dichloroethane
r-








C






T3




C




co
U)
d
•o





Q)
s-1 ,2-Dichloroethylen
'o








c






•o
c




•o




•O





0)
C

ans-1 ,2-Dichloroethyl
4-1






O
CO
o
TJ






C




•D




•a







richloroethylene
1—








c






c




•a
c




c







etrachloroethylene
H-








c






c




en
d
•6




•o
c






,2-Dichlorobenzene
T-



























2
E
c
g
o
4-<
T3
(D
= not detected abovi
= not tested
•0 4-
c c
                                                      C-28

-------
CD








1
0
LU
i
m
^
^~












CJ

Lt_
LU
I
H
^
QC
1-
LU
CD
2
^
Z

(/)
^
O
2
LU
I
O
u
z
(3
OC
0
c
o
ss
5
Q)
Q.
U)
C
ro
a.
c
a>
'&
0)
ro
U)
c
o
4^
2
4->
C
0)
c
0
u
^







a
c
'J5
'5
CO
0)
2
<2
D
O




a
c
'•6
(i
c
•o
"Z













(O
LU
LU
at
c
13
u
c
u.


or
u
c
a>
£
o
it
O
Q>
a>
2
o
4-<
CO


C
o
a
a
3
CO
c
0
1
TJ
8
CL








O
o
oo



c




c

c


H




s



c







Dichloromethane

i



c




4-<

4-*


C




€



c







Dichlorodiflouro-
methane

o
CO



c




c

4-*
C.


c




c



c







1 ,2-Dichloroethane

IT)
in



CO
d




CO
o

CO
o


CM
CO




0
CO



6







1,1,1-Trichloro-
ethane

i



c




4-»

C


4^




4->



«







CO
o
0)
LL

IT)



T3
C




C

C


•o
c




•o
c



•o
c







Vinyl chloride

,



•o
c




•a
c

T>


•o
C




•o



•o







1 ,2-Dichloroethylene

0
CO



CM




CM
O

O>
O
0


6




tn
o



6
V







Trichloroethylene

OT



T3




CM
O

0


T5




1



r—
o







Tetrachloroethylene

o
o
CO


•o
c




•a
c

•o
c


c




•o



•o
c







Dichlorobenzene
























13
E

c
o
u
M
V
•o
0)
8
C
J "O
D Q)
U W
•^ 09
3 o
c C
II II
3 4^

                                                                                      4)
         c
         0)
         N
         C
         0)


         8
        5
        ^:
         u
        '•5

*-«'?4
                                                                                      u
                                                                                     T3
                                                                                      a>
                                                                                      a.
                                                                                      a
                                                                                     =re£
                                                                                     O  .
                                               C-29

-------
C.3.2  Indoor Air Modeling Results
       Estimates of  indoor air concentrations were made, using models,
for current residences, the former elementary school, and possible
future apartment complexes.  The Farmer model (see Appendix A of this
report) was used to estimate emission flux at ground surface.  Based on
assumptions about building construction (area of the flooring through
which  gas could infiltrate is the dominant factor) and building air
exchange rates, the indoor air concentrations were then calculated.  The
modeling results are given in Tables C-10 through C-12. [The tables show
that identical groundwater concentrations and contaminant vapor phase
concentrations were used for the average and plausible worst cases.]

Interpretation of Results
       The Agency for Toxic Substances and Disease Registery (ATSDR)
reviewed the above data (as well as other data not presented here) and
estimated cancer and non-cancer risks based on their standard methodolo-
gy.  The analysis indicated that only current off-site residences (not
including the former elementary school) have any increased risk from
exposure to chemicals volatilizing from the groundwater plume.  The
upperbound lifetime excess cancer risk was estimated to be 1 x 10"4  (low
increased risk) for average case conditions, and 2 x 10~3  (moderate
increased risk) for the plausible maximum conditions.  The summary of
the ATSDR report states that the site is an indeterminate public health
hazard because of potential human exposure to concentrations of chemi-
cals that upon long-term exposure could cause adverse health effects.

C.3.3  Comments on Case Study
       1.    As noted by ATSDR, use of the flux chamber to check emission
flux predictions of the model  at ground surface may be valuable, but
flux chamber data may be inappropriate as flux source data for infiltra-
tion into a building due to negative pressure in the building.
       2.    The modeling report states that convective flow due to
building vacuum (ie. stack effect) is generally more important than
simple diffusion into the building.  Only diffusion into the buildings
                                C-30

-------
U
LU
CQ
U
o
LL
O
Z
g
^
cc
\-
z
LU
o
z
o
0
cc
J<
cc
o
o
o
z

LL
o
z
o
h-



LU
CO
0
cc
LL
r-
LU
Q
CC
(D
Z
O
Q
CO
LU
U
Z
LU
Q
CO
LU
CC

(J
z
Q 0
UJ CJ
5 <
P £E

9
UJ

r»
CO

27E-01
CM
O
UJ
CO
o>
CM
LU
r»

n
9
Ul
CO



97E-03
^
-309
LO
CO
9
UJ
LO
N
0)
9
UJ
CO
00

9
UJ
n



37E-02
t
o
9
UJ
CM
CO
P)
9
UJ
O
09
03

O
CM


O
CM

richloro-
thylana
H- «
9
UJ
CD
n
n

CM
9
UJ
S3

9
UJ
LO
cn
CM



9SE-03
CM

9
UJ
8
"
P)
9
UJ
o
LO
O

^



--

ans-1,2-Di-
hloroethylene
^ o
CM
9
UJ
r~
CO


68E-01
*
CO
9
UJ
o
o
09
9
LU
0




35E-01
P)
o
9
Ul
8
0)

9
UJ
O
0)
ri

CM



CM

P)
i
LL
§
UJ
p^
n.
n

9
UJ
CO
09
9
Ul
LO
^
0
UJ
CD
CM
n
9
LU
CM



P)
9
UJ
CM
fN
O
9
Ul
O
03
^
CM
9
Ul
o
03
CM

P>
<6


n
o

,1,1-Trl-
hloroathana
•- o
                                       C-31

-------
O
LU
CO

f
O
o
U-
O
Z
o
ENTRATI
O
Z
0
o
cc
cc
0
0
Q
z
LL
0
Z
o
\-
F
LU





1
SCHOOl
^.
CC
^
^—
Z
UJ
UJ
UJ
cc
UJ
cc
o
LL






d
Q 0
UJ U
t—
ESTIW
INDOOR >>

UJ
i
x
u.
<
I

O

o
o
u
Vl
VAPOR PH










O
z
o
u
cc
UJ
GROUNOW/
S E
£ ? *2
lit
n

IK
= 2

? vJ
5 I
< 3

In
il«

II

5 >-5

Q •£

- =s
* « ^

1 -^ o E
o*
E
1 I

I11
_J
O
2
Ul
U


9
UJ
q
r>
eo-308
^
.58E-O9
ID
CO
9
Ul
V

fM
9
Ul
(S|

2.71E-02
in
9
UJ
O
*~

9
UJ
to
*~

2

r)
o
«- j:
*" *
n
1.61E-0
18E-O4
"*
O
Ul
CT)
CO
(SI
_
UJ
JJ
"'
n
9
UJ
r»
a>

1.97E-03

9
UJ
CD
CO

(Si
9
Ul
n

4.37E-02
0
9
UJ
(N
CO
P)
9
UJ
O
CO
00

o
CN

O
(S|

richloro-
thylene
1- 0

9
UJ
CD
CD
6OE-O4
r*
O
UJ
*~

UJ
ID
0)


9
UJ
(SI
CO
ui

6.82E-O4

c
o
0
u.
o
9
UJ
CSI
03
61E-03
^"
.15E-09
"
O
UJ
CD
O)
CM
r)
9
Ul
(SI

7.21E-O3
CD
9
UJ
8
N
(S
9
UJ
O
CD
(SI

r>
CD'

r>
CD
I
,1,1-Trl-
iloroetha
•- o
                                          C-32

-------












CN
i
u
LU
CO
1-









CJ
o
LL
O
Z
0
r-
^
CC
f—
Z
LU
O
Z
o
u
cc
<
cc
o
o
Q
Z

LL
O
Z
o
h-






co
LU
U
Z
LU
Q
(/O
LU
CC
LU
r-
w
1
Z
0
LU
CC
D
H
D
LL



LU

o
Q 0
uj u
<£
5 <
w o
UJ O
Z

p
^
cc
X

u.
O
5
UJ
X
0
(J
Z
o
o
Ul
CC
O
Q.
>







O
O
(J
cc
UJ

0
Z
o
cc
o
i E -
S 3 CO

2 g >
C 5 3


01 CO
s 1

*E?
"S c CM
5 x E
1? * C»
CL 2 —

* "S
ft f^
« E

< S
1 i co
si£
«I«
1 co
1 1
< ^


i •*

-• Si
t j a E o
I " 1 'e g
i o S
E
i 1


•g
U!
o
0
UJ
I
o
o
UJ
CO
CD
,_


CM
O
Ul
•9
"*
§
Ul
CO
CM
*-

o>
o
Ul
0)

co
CM
O
Ul
03
en
in
S
Ul
CD
cn

8
Ul
0
r-
O
UJ
in


in
cri



in
en

1,1-Dichloro-
ethylena
CM
o
Ul
in
fx
CM


CO
o
Ul
CM
^
8
Ul
in
CD
•-

o
Ul
01

*
s
111
CM
CM
8
UJ
CM
CN

8
Ul
o
ls-
CM
O
UJ
0
CM

2



CO

Tetrachloro
ethylene
o
Ul
CO
o
in


O
Ul
O
•~
8
Ul
03

co

8
Ul
in
rx
CO
r-
O
Ul
S
CM
O
Ul
3

8
Ul
CM
CO
8
Ul
o
0)
03

O
CO
in



o
CD
in

Trichloro-
ethylena
o
Ul
0
CM
CM


8
Ul
o
in
8
Ul
CD

*-

O
Ul
CO
03
CO
8
Ul
co'
8
Ul
CD
8
Ul
fx,
0
*•
O
Ul
o
en
CD

O
CM



O
CM

Vinyl chloride
0
LU
in

-------
was considered.   No  explanation  for  not  considering convection was
given.
      3.    Whereas  gas was considered to  infiltrate buildings through
the entire  floor  area  for current private  residences, infiltration
through only  a 0.5 cm  crack around the perimeter was allowed for the
former elementary school and future  residences.  For example, for the
school, this  assumes only 0.08 percent of  the soil gas flux enters. For
future residences the  assumption is  0.2  percent flux entry.  The
assumption  that gas  entry rates  are  a direct function of the percent
cracked area  is not  supportable  and  could  result in significant under-
predictions of indoor  air levels.
      4.    The modelling approach made no attempt to correct for the
capillary fringe  above the groundwater.  While this is appropriate for
an initial  simple screening approach, it should be recognized that this
could lead  to large overestimations  of contaminant flux to the surface.
      5.    Incorrect Henry's  Law Constants were used leading to  a
potential overestimation of flux by  a factor of about 50.  (Non-dimens-
ional  values used in a dimensional equation).
      Because of  the opposing influence of item 2 through  5, the
relevance of the  model predictions is uncertain.  Recognizing that the
initial modeling  approach was simplistic,  in retrospect it would have
been preferred to conduct more refined modeling before releasing the
data or conducting on-site monitoring.
                                C-34

-------
 C.4   BUILDINGS OVER GROUNDWATER PLUME (B)
       The site (Figure C-4),  a former printed  circuit  board  manufactur-
 ing facility,  is located at the southern  end of San  Francisco  bay.   The
 facility operated from 1970 to 1985.   The U.S.  Environmental Protection
 Agency (EPA)  placed the site  on the  National Priorities  List (NPL)  on
 February 1,  1990.
       The primary site of both wet and dry  manufacturing  processes  was
 the "wet floor" building where waste  waters containing heavy metals and
 organic  compounds were released to the "wet floor" and then discharged
 to  the neutralization  sump.   Other solvents and wastes were placed  in
 drums  and disposed of  off-site.   It  is still unclear exactly how the
 chemicals discharged to the "wet  floor" or  the neutralization  sump  were
 released to the soil and then  the ground  water.
       In 1985  manufacturing stopped.   As  part of the plant closure
 process,  soil  and  ground-water investigations were initiated.  The
 investigation  showed chlorinated volatile organic compounds (VOCs)  and
 metals (copper,  lead and nickel) are  the major contaminants found in the
 soil and ground-water  samples.  The highest concentration of organic
 contaminants in the soils was  found around  the sump and the "wet floor".
 The sump and surrounding contaminated  soil  and the contaminated soil
 beneath  the "wet  floor"  were removed.
       Seven wells  on the site  and 24  off-site wells (Figure C-4)
 currently monitor  the  VOC contamination in  the groundwater.  The ground
water monitoring  indicates that the plume extends north underneath
 Interstate 101  (1/8 mile) and west of  the site approximately a quarter
mile.  Remediation of  the groundwater  began around 1986.   Currently
there are 3 extraction wells on the site and 4 extraction wells off-
site.
      The site  consists  of a number of high-ceiling one-story buildings
now occupied by other  businesses.  Light industry and businesses also
rent the other  four buildings that were once part of the  manufacturing
complex.
                                C-35

-------
       KEY
   •  Monliorinf well
   A  Extraction weU
     Conumintted pound wttcr pturac
Figure  1.   Contaminated ground-water plume.
                                              FIGURE C-4
                             C-36

-------
       According to 1990 census  information, approximately  1000  people
 live  in  the 10 block area surrounding  the plume of groundwater  contami-
 nation connected with the site.   The bulk of the population  resides  on
 the south  side of Highway 101, with 1990 census information  showing  only
 10 people  living on  the north side of  Interstate Highway 101.
       Land in  the area surrounding the site has been used  for residen-
 tial,  commercial,  and agricultural purposes for many years.  The  areas
 surrounding the site are zoned for a mixture of multiple family and  two
 family residential units,  and general  and limited industrial use.
 Within the four block vicinity of the  site, there are 409  housing units
 with many  of these being 14  unit  complexes.
       The  area north of Highway 101 is zoned light industrial and
 planned  community designation.  This type of zoning is designed for
 research and development,  office  space, and light manufacturing.  No new
 housing  is  allowed under this zoning.  Within this area approximately 20
 acres  of land  are used  for commercial farming.   Two schools are located
 northwest  within  a half mile of the site.  The nearest city park  is one-
 half mile  southwest  of  the site.
       Four  major  water-bearing zones (aquifers)-defined as the A-, B-,
 intermediate -  and C- zones-exist at the site.   The A-aquifer is the
 shallowest  and  the C-aquifer is the deepest of these three zones.  The
 approximate  depths (below  ground  level) at which these zones occur at
 the site are as  follows: A:  10-20 feet; B:  30-40 feet; and intermediate:
 60-75  feet.  A  regional  impermeable zone, the B-C aquitard, is reported-
 ly located  at  a depth range from  100-150 feet below ground level.  The
 C-aquifer, which  supplies  most of the municipal  water in the region, is
 located below the regional aquitard.

 C.4.1  Nature and Extent  of Problem
      Between August  1985  and August 1986,  the  neutralization sump and
 some surrounding soil (70  cubic feet)  was excavated and sequentially
 backfilled to grade  level with pea gravel,  well-compacted clay,  and
 asphalt.   In September  1986,  about 255  cubic yards  of soil  were excavat-
ed under the "wet floor".
      In May 1987, soil  samples were taken  at  five  feet depth from
borings at the sump  and  soil  around  it  and  analyzed for organic  contami-

                                C-37

-------
 nants  {"wet floor"  soil  was  not  analyzed).   Results  are  shown  in  Table
 C-13.   Results  of the post-remediation analyses  showed the  presence and
 levels  of various organic  contaminants not  originally found in  the
 neutralization  sump area.
                                TABLE  C-13
             ORGANIC CONTAMINANTS IN ON-SITE SUBSURFACE SOIL

1, 1-Dichloroethane
1 , 1 , 1-Trichloroethane
1 , 1-Dichloroethylene
t-l,2-Dichloroethylene
Trichloroethylene
Sump
Before
(4/85)
nd
10
7.3
nd
nd
Sump
After
(5/87)
nd-24
nd-69
nd
nd-110
41-220
" Wet "Wet
Floor" Floor"
Before After
(10/86)
na
na
na
na
nd-100
na
na
na
na
na
nd = not detected above detection limits
na = not analyzed
      Since  1986, 7 on-site monitoring wells have been installed  to
characterize the vertical and horizontal  extent of contamination.  The
most contaminated wells are located near  to or downgradient, north to
northwest, from the sump excavation.  The migration of contaminants
through the major water-bearing aquifers  is being monitored.  The two
shallowest aquifers (A and Bl) are contaminated with dichloroethane,
1,1,1-trichloroethane, 1,1-dichloroethylene, trans-1, 2-d-
ichloroethylene, and trichloroethylene at concentrations exceeding state
or federal drinking water standards (Table C-14).
      Three extraction wells were installed on-site: two draw from the
A-aquifer and one draws from the B-aquifer.  The untreated ground water
is discharged by permit into the sanitary sewer system.
                                C-38

-------
                                      TABLE C-14
                   ORGANIC CONTAMINANTS IN ON-SITE MONITORING WELLS
       Data  are taken from the quarterly monitoring well  report, fourth
       quarter, 1990.  The historical  range of contaminant concentrations for
       each  aquifer is given with the  highest concentration currently measured
       in  that  aquifer given in parenthesis.


1,1- & 1,2-Dichloroethane
1,1,1 -Trichloroethane
1 ,1-Dichloreothyiene
t-1 ,2-Dichloroethyiene
Trichloroethylene
Drinking Water
Standards

0.5'
200'-"
6"
10C
5'-b
Aquifer Concentration (ppb)
A B Intermediate
nd- 13 (4)
nd- 45 (6.2)
nd- 17(1.4)
nd- 36(7.03)
nd-1 22(20)
nd- 20(nd)
nd-1 3,000(1. 8)
nd- 1 ,000 (nd)
nd- 14(4.3)
1.3-7,500(26)
nd
nd
nd
nd
nd
nd = not detected above detection limits
•California MCL
"Federal MCL
'proposed California MCL

      Starting  in late  1986,  24  off-site monitoring wells were  installed to
      assess the extent and  severity of the migration of contaminants  from the
      site.  The two shallowest  aquifers (A and  B) were found to  be  contami-
      nated with organic compounds.  Six chlorinated organic compounds
      identical to those found at the  site, have been detected  in the  ground
      water (Table C-15).  Concentrations of dichloroethane, 1,1,1-trichloro-
      ethane,  1,1-dichlorethylene, trans-l,2-dichloroethylene,  and trichloro-
      ethylene, exceed drinking  water  standards.
            Four extraction wells were installed off-site: two  draw  from the
      A-aquifer and two draw  from the  B-aquifer.  The untreated ground water
      is discharged by permit into the City sanitary sewer system.
                                      C-39

-------
                                      TABLE C-15
                  ORGANIC CONTAMINANTS IN OFF-SITE MONITORING WELLS
      Data  are  taken  from the quarterly  monitoring  well  report,  fourth
      quarter,  1990  (10).   The historical  range of  contaminant  concentrations
      for each  aquifer is given with  the highest concentration  currently
      measured  in  that aquifer given  in  parenthesis.


1,1- & 1,2-Dichloroethane
1,1,1 -Trichloroethane
1 , 1 -Dichloreothylene
t, 1 ,2-Dichloroethyien8
Trichloroethylene
Drinking Water
Standards

0.5'
200'*
6'
10e
5'-k
Aquifer Concentration (ppb)
A B intermediate
nd- 330(64)
nd- 2500(1 40)
nd- 420(47)
nd- 580(580)
nd- 1599(520)
nd- 310(310)
nd- 1 100(12)
nd- 350(140)
nd- 550(550)
nd- 320(200)
nd
nd-1.0(nd)
nd
nd
nd
nd = not detected above detection limits
•California MCL
"Federal MCL
"proposed California MCL
            The concentrations  of  site-related contaminants  in ground  water
      are at levels that would  be  of potential public health  concern  if
      domestic use of the contaminated ground water occurred.  No municipal
      wells currently exist  near the site.  There are several private  wells  in
      the area or directly within  the contaminated ground-water  plume,  but
      most of these are not  active.  Of those wells that are  still  active, the
      water pumped from these wells is not currently used  for domestic
      purposes.  At this time,  sufficient water for municipal use is  available
      from aquifers having higher  quality water and water  yield.  Additional-
      ly, regulatory barriers exist to prohibit installation  of  shallow
      private wells.
            A theoretical model (Farmer-See Appendix A) was  used to estimate
      the contaminant soil gas  flux and the air concentrations within  single-
      family residences located above the contaminated plume.  Air  concentra-
      tions for an "average  case scenario", "plausible maximum scenario", and
      a "most plausible case" were calculated.  These scenarios  use different
                                      C-40

-------
 assumptions  for area of infiltration (crawlspace or crack around
 perimeter of building),  the fraction of air that infiltrates  from the
 crawlspace,  and/or the air exchange rate of the  home or business.
       Based  on  the concentrations of organic contaminants accumulating
 in  a  house as developed  by this model,  the  risk  of developing cancer
 from  inhaling volatilized organic contaminants for current residents
 living  above the contaminated ground water  was calculated.  The exposure
 was assumed  to  be 24 hours per day,  365 days per year for 40  years .for
 residents.   The exposure was assumed to be  8 hours per day, 260 days per
 year  for  40  years off-site workers.
      The  lifetime excess cancer risk from  the inhalation of  volatiliz-
 ing organic  contaminants from the shallow ground water was  estimated
 from  the  addition of the individual  risks to the three potential
 carcinogens,  1,1-dichloroethylene,  1,1-dichloroethane and trichloroeth-
 ylene.
      Lifetime  excess  cancer risk for off-site residents  ranged from 1.8
 x 10"8 ("no increased risk") for the least conservative estimate  and 2.9
 x 10"5 ("no apparent increased risk") for upperbound worst-case condi-
 tions.  Lifetime  excess  cancer risk  for off-site workers  ranged from 2.1
 x 10"9 ("no increased risk") for the least conservative estimate  and 3.5
 x 10"6 ("no increased risk") for upperbound  worst-case conditions.
      Potential risks  were also assessed  for the  noncarcinogens (trans-
 1,2-dichloroethylene and 1,1,1-trichloroethane)  for residents above the
 contaminated ground-water plume.  The analyses indicated  that the
 inhalation of organic  contaminants from off-site ground water would not
 result in adverse  noncarcinogenic effects.

 C.4.2  Conclusions
      Based on  information  reviewed, ATSDR and CDHS concluded that the
 site is not an  apparent  public  health hazard.  As noted in the human
exposure pathways  section  above, off-site resident and worker exposure
 is predicted  by an air-model but the exposure is  at a level below that
of health concern.
      Future  significant  exposure to ground-water contaminants is
unlikely if the ground-water extraction  and treatment system reduces
concentrations of  site related  contaminants to below levels of health
                                C-41

-------
 concern,  no wells currently  in  place  are  used  for domestic purposes,  and
 future  drinking water wells  are not placed  in  areas of known contamina-
 tion  if ground-water remediation does  not clean up contaminants to
 drinking  water standards.

 C.4.3   Comments on Case  Study
      As  we reviewed this  case  study,  we were  struck by the similarity
 to  the  preceding case study.  The two  sites are both located at the
 southern  end of San  Francisco Bay and  are only a few miles apart.  They
 appear  to have similar hydrology and geology.  Based on the reported
 chemical  concentrations  in the  ground  water plumes for the two sites,
 the groundwater appears  to be substantially more contaminated at the
 site discussed in  this section.   Based on the relative lifetime excess
 cancer  risks for current off-site residents for the two cases,  it
 appears that for this  second case indoor air model  calculation used a
 diffusion  only method  and a flux  attenuation factor based on an infil-
 tration area represented by a 0.5 cm crack around the building perime-
 ters.   We  note that  this is extremely more restrictive (yields lower
 predicted  indoor concentrations) than the assumptions used for current
 residents  in the preceding example but is similar to that used for the
 former  elementary  school and future residences in that case example.
 This assumption  is likely overly restrictive and insupportable in both
 case examples.
      For  illustration purposes  only,  we have calculated indoor air
 concentrations for off-site structures above the ground water plume for
 the current  site using the same  parameters used for current off-site
 residences  in  the previous case  example.  We also ignore corrections for
 the capillary  fringe  (as did the study authors) because only a screening
 level result is  desired.  [This  correction should be made if a better
 estimate  is  needed.]   Because we did not have access to sufficient data
 to calculate the geometric mean  concentrations in the ground water, as
was used  in  the  previous case, we have used one-half the maximum values
                                C-42

-------
                                        TABLE C-16
                       HYPOTHETICAL CALCULATIONS  FOR INDOOR AIR
                       IN STRUCTURES ABOVE THE GROUND WATER PLUME
Compound
1 , 1 -Dichloroethylene
Trichloroethylene
1 , 1 -Dichloroethane
Trans 1 ,2-Dichloroethylene
1,1,1 -Trichloroethane
Sunnyvale
Off-site cone.
(max) in
ground water
4.3
120
2.5
11
6.3
Mountain
Viow"1
Off-site Cone.
(max) in
ground water
47
520
64
580
140
Modeled Indoor Air Concentrations (ug/m3)
Sunnyvale
AVG
4.24
5.50
0.087
0.46
0.87
Plausible
Max
16.4
21.3
0.34
1.77
3.37
Mountain View
AVG
23
12
1.1
12
9.7
Plausible
Max
90
46
4.3
47
37
(1) - For calculation purposes, one-half this value was used in model
                                        C-43

-------
 given  in  Table  C-15.   The  results  are  in  Table C-16.   These calculated
 results  indicate  that  under  these  modeling  assumptions,  indoor air near
 the  site  would  be substantially  more contaminated than in the previous
 case.   Thus, it  is  likely these estimates  are much too  high.

 C.5    BUILDINGS OVER GROUNDWATER PLUME  (C)
       The site  is located  in  a suburb of  Casper, Wyoming.  A residential
 subdivision is  located  approximately one-quarter mile northeast of an
 industrial park.   An oil refinery  is located west-southwest of the
 subdivision on  the opposite  side of the North Platte River.
       In  the mid-1980's, a contaminated water plume was discovered
 underlying part of the  residential subdivision.  Because the residences
 all  use wellwater, an  indoor  air study comparing periods with and
 without showers operating  was conducted.
      Air sampling was  performed in five  residences with contaminated
 wells, five homes  with  uncontaminated wells and five Casper area homes
 as controls.  Each home was  sampled during  a 5-hour period with the
 shower operated at least 10 minutes and on  a subsequent day, another 5-
 hour period without shower operation.  In addition, 5-hour basement and
 5-hour outdoor air samples were  collected.  Samples were obtained with
 low-flow air pumps and  Tenax collectors and were analyzed for volatile
 organics.  The study was designed  to compare shower vs. non-shower
 concentrations and to compare results among the three study groups to
 see  if significant differences existed in indoor concentrations.

 C.5.1       Nature and Extent of Problem
      Organic contaminants found in groundwater samples were used to
 divide the residents into  two groups, contaminated (Group 1) and
 uncontaminated  (Group 2).  In the  residential air study,  these two
 groups were compared to background homes  in the Casper area (Group 3).
      Residential  sampling included indoor  air with shower (IAS), indoor
 air without shower (IA), basement  air (BA), and outdoor air (OA)
 collected in the  backyard  of each  home.   In addition, a questionnaire
was administered  at each home to identify potential alternative sources
 of airborne contaminants.
                                C-44

-------
       The  air/Tenax analysis was  designed  for  quantitation  of HSL
 volatile compounds.   Tenax tubes  were  used for sample  collection.  The
 Tenax  tubes  can  retain most volatile HSL compounds  (lowest  boiling,
 i.e.,  chloroethane  excluded),  but do not adequately  retain  many  light
 hydrocarbons and low boiling compounds, i.e.,  saturated  hydrocarbons 
-------
       O3

       C

       cc

       -z.
       UJ
       O
       •z.
       o
       o

       r-
U

UJ
_i
CO
      o
      U

      y
      z
CC
o

UJ
z
cc
o
CQ
CC
      O

      Z
      o
      cc.
      t

      o
      UJ
      UJ

      CD
      Z





a
o
0










a
cS







Outdoor
c
E
o
V)
CD



S
C 0
O .C
W yj
*
c S
o o
« £
f -
°- i

0
TJ
3
O
Basement

o

c o
o ^
2 to
o
Q.
C $
0 0
111 -C
£-
°- $




<*J
co" *
CO —
3

1 I






m

^


^
rZ «



Benzene
CO
• in

1 I







t 1


1°

! 1

i !



. .




1 i
o
•o
•c
z
Carbon Tetrac
I i

1 I







1 I


1 1

1 I

in
d *•;
3



: i




; i

0
c
.c
1,1-Dichloroe
1 I

1 1







1 1


1 1

1 1

§3



l i




i :



Ethylbenzene
I i

I I







1 I


1 !

1 I

1 1



! i




*~



2-Hexanone
6 •*:
7 o
3

I !







1 1


1 1

1 1

3



co •-;
3



i°"



Methylene
Chloride
I i

I !






v—
^2
3

O <">
3

1 1

23
f J-
3



in" >"
n —
3


0)
3



Tetrachloro-
ethylene


: ;






(N
0 CO
3

i°'

! 1

0)
"° 9
3



CO
' ui
3


CO
"9 «•>
3



Trichloro-
olhylene
?5
3

i :






f*s
in «
o> ro
3
en
6 °\
3

«2

in
in



iri PI
O) »-



CO t
'

1 1






(*^
!*• "?
3
P>
r-.'
CM
en i;
CO (si
O
to
d


r^
en ^;
^ *~



in en
3



Toluene
                                                  C-46

-------
T3
 03
 C
 O
 o
O

LLJ

CO
cn ^~
(j M
C
O
•£ ^
'•6 ^j
3






(0
Q.
3
O
<5

















o
o
•5
3
0
1
isement
m





$
c o
o - to
d






(O
4 oi









Benzene
1 I



d "
3



9 d

3








° d
3






3







O
T3
•c
O
Carbon Tetrach
! 1



I 1




I 1

1 1








: :







i ;








c
03
1,1-Dichloroeth
1 i



1 I




1 I

1 I








! !







1 1









Ethylbenzene
1 i



1 I




1 i

CO ^j
3 *~








i !







cn ^_
3 *~









2-Hexanone
m "



n
in •
3 *~



: i

i i







CO
r^" ^
CO 01
3






in
co q
3









Methylene
Chloride
I I



I I




D "J

i0'







in
6 
-------
       In order  to  calculate  airborne concentration, the flow rates were
 adjusted to conditions  of  standard temperature and pressure.  The
 compound weights were blank  corrected by subtracting mean weights of all
 field  blank analytes from  the reported field sample value prior to
 dividing by standard liters  of air to calculate the flow rate.
       Descriptive  statistics, including mean and range of concentration
 were calculated for each sample.  When calculating mean values, one-half
 the instrument detection limit was used for non-reported values.  The
 data were lognormally distributed and samples were randomly selected.
       Data were analyzed statistically using the Mann-Whitney (MW) U-
 Test for independent groups.  The data were then logarithmically
 transformed and analyzed using the more powerful  t-test for basement
 samples and Analysis of Variance (ANOVA) for indoor and outdoor samples.
 One tailed probability values (p) are reported as an indication of
 statistical significance.
       The MW test  converts the original data to ranks and compares two
 independent groups of data.  Ranks are used due to the fact that the
 small  size and high variability of the data do not met the assumptions
 of normality or homogeneity  of variance necessary to use parametric
 statistical analysis.  The MW test closely approximates the parametric
 t-test.  It is less powerful than the t-test in finding a difference
 between two means  if one exists.
       A questionnaire was completed by each participant to identify the
 following potential sources  of airborne organic contaminants:  cigarette
 smoking, use of water operated appliances,  use of organic chemicals,
 house  cleaning, and open windows during sampling.   The results indicated
 that Group 2 had more smokers, greater use of appliances,  more house
 cleaning, and more open windows during sampling.   Participants were
 asked  not to use chemicals during the sampling period.

 C.5.2  RESULTS
 Shower Compared to Non-Shower Exposure
      Samples collected with and without a shower from homes in Group 1
were compared to determine whether groundwater contaminants entered
 indoor air via showering.  As expected, PCE, TCE,  and TCA levels were
                                C-48

-------
 slightly  higher  in  IAS  samples  compared  to  IA samples.  A  statistical
 comparison  of  the data  indicated  that  the results were not significant
 at  the  (ANOVA) p<0.10 level.  The highest significance level was found
 for toluene at (MW)  p<0.34.   IA and  IAS  samples also contained virtually
 the same  amounts of  contaminants  when  comparisons were made for Groups 2
 and 3.
 Airborne  Contaminant with  Shower  Exposure in the Three Groups
      Airborne contaminant exposure  in IAS  samples were assessed in the
 three study groups to determine the  contribution of contaminants from
 drinking  water to residential air.   Mean values were higher in Group 1
 homes,  compared to Groups  2 and 3, for PCE, TCE, and TCA.  However, the
 values  were not found to be statistically significant at the (ANOVA)
 p<0.10  level.  The highest level  of  significance was found for TCE
 between Groups 1 and 2  (MW) (p<0.20).  Significance levels for other
 comparisons ranged from (MW) p<0.27  to (MW) p<0.42.  The lowest levels
 of  PCE  and  TCA were  found  in Group 3 homes.  TCE levels were very
 similar in  Groups 2  and 3.
 Basement  Air Samples
      None  of Group  2 homes had basements, consequently,  comparisons
were made between Groups 1 and  3.   PCE concentration was higher in the
Group 1 samples at a significance level of  (t-test) p<0.05.  Mean TCE
concentration was higher in Group 1, but not significantly (t-test
p<0.17).  The other contaminants were not tested statistically due to
the small difference between mean values.
Ambient Air Samples
      Outdoor air concentrations of all contaminants were consistently
lower in Group 1  compared  to Group 3 indicating that they did not
contribute  to indoor levels in Group 1.  TCE and PCE were not found in
Group 1 outdoor samples, but were found in Group 3 outdoor samples.
Toluene and benzene levels were lower in  Group 1 samples  at significance
levels of (MW)  p<0.26 and  (MW) p<0.20,  respectively.  TCA levels  were
slightly lower in Group 1  samples  compared to Group 3 samples.
                                C-49

-------
C.5.3  Conclusion
       Contaminants  found  in  previous  well water samples from homes in
Group  1 were  also present  in indoor air.  Mean levels of PCE, TCE, and
TCA were  higher in  Group  1 homes  compared to Group 2 and 3 homes  (not
statistically significant  at p<0.1 level).  The fact that outdoor
concentrations of these contaminants  were relatively low, even though
the primary wind direction was  found  to be from the southwest emanating
from the  industrial park toward the subdivision, suggests an indoor
source of contaminants.  Showering was investigated as a potential
source.   TCE,  PCE,  and TCA mean concentrations were found to be slight-
ly, but not significantly higher  in IAS as compared to IA samples
collected in  Group  1 homes.
       PCE concentrations in  basement  samples were significantly higher
(t-test;  p<0.05) in Group 1  homes compared to Group 3 homes.  The source
of PCE in basements may have been vapor directly from contaminated
groundwater or from appliances that use water, such as hot water heaters
or washing machines, located  in the basement.  If groundwater were the
source, TCE levels would also be expected to be elevated.  TCE values
were higher in Group 1 homes, but not at as significant a level  (t-test
p<0.17) as PCE.  A possible  explanation for this result is that PCE
partitions from water to air more readily than TCE and, consequently,
may volatilize from groundwater and accumulate in residential basements
at a higher rate than TCE.   The air/water partitioning of the two
contaminants  was determined  using Henry's Law constants at constant
atmosphere:   PCE,  1.1 x 103;  TCE,  5.5  x 102.  The larger of the Henry's
Law constant, the greater the equilibrium concentration of the compound
in air compared to its concentration  in water.   It should also be noted
that the  sample size for the basement sample comparison is quite small,
three homes per group.  Very small sample size makes a statistical
comparison of the data difficult and consequently,  makes the data
comparison inconclusive.
      The influence of other potential sources of airborne contaminants
were investigated.  Prior to sampling, all participants agreed to avoid
use of household or other chemical during the sampling period.   Group 2
homes had increased numbers  of cigarette smokers,  use of water operated
                                C-50

-------
 appliances,  use  of organic  chemicals  and  house  cleaning  activity  during
 sampling.   These factors  would  be  expected  to increase the  airborne
 levels  of  the  same volatile organic contaminants  found in well water,  as
 well  as increased  volatile  organics from  cigarette  smoke and organic
 cleaners and solvents.  However, the  measured indoor air contaminant
 levels  were  not  consistently elevated in  Group  2  homes compared to
 Groups  1 and 3.
      Cigarette  smoking may not have  been an important factor  in
 contributing to  the airborne contaminant  levels presented in Table 2-17.
 According  to the Surgeon  General Report,  1981 (Wynder, E. and  Hoffman,
 D.),  none  of the airborne contaminants  identified in the study were
 found as "major  toxic agents" in cigarette  smoke.  The report  does say
 that  the list  presented was  incomplete  and  adds that cigarette smoke may
 contain "such  carcinogens as volatile chlorinated olefins".  This group
 may contain  PCE  and TCE.
      The  fact that more windows were open  during sampling  in  Group 2
 homes may  have lowered airborne indoor  contaminant concentration.  This
 would negate the effects of  the previously  discussed factors that tend
 to increase  contaminant levels.
      The  levels of contaminants found  in homes in all  three study
 groups were  typical of indoor concentrations found in enclosed living
 spaces.
      Outdoor mean contaminant levels in the three groups and 24-hour
 samples were also  found to be lower than those in the urban areas of
 four major U.S. cities. (Houston,  St.  Louis, Denver, and Riverside).

 C.5.4 Comments on Study
      This study was fairly complex in that it monitored at three type
 homes, with and without showers operating, as well as ambient air.  It
 comes to the conclusion that the indoor air quality in  all  cases is
within the "typical" range for residences.  It should be noted that
 because none of the Group 2 homes  had  basements, whereas group 1 and
Group 3 did, and the possibility of soil gas intrusion  from the contami-
nated groundwater plume existed for Group 1 homes, the  Group 2 homes
would not appear to be an adequate control group.   However,  the design
                                C-51

-------
has left open the question whether or not soil gas is infiltrating the
homes even at these low levels by presenting an inconclusive statistical
result indicating a significant difference from the controls.
      Extraction wells were later installed at this site to control
groundwater migration.
                                C-52

-------
 C.6    BUILDING NEAR SLUDGE DISPOSAL  PIT
       The site is located in  a  rural  section of Brunswick County,  North
 Carolina  (Figure  C-5).
       The site was used  by sludge  hauling and oil  spill  cleanup  compa-
 nies  for  the  disposal  of septic tank sludge, oil sludge,  and  other waste
 materials from 1969 to  1976.  These  wastes were disposed  in shallow
 (approximately two-to  eight-feet deep) unlined pits or directly  on the
 land  surface  at the site.   A  residential subdivision now  exists  in the
 area  of the former disposal areas.   Typical dwellings in  the  subdivision
 consist of manufactured  homes situated on one or two-acre lots,  each
 having a  private  domestic  well.  There are approximately  150  residential
 lots  in the area,  70 of  which were occupied in 1990.
       In  May  1976,  the owner was informed that an  oil disposal pit
 violated  North  Carolina  statutes and must be cleaned up  immediately.
 This  pit  was  approximately 60 feet long, 20 feet wide, and two to  four
 feet  deep.  At  the  time,  it was  estimated that approximately  2,000 to
 3,000 gallons of  black oil were  contained within the pit.  The owner
 pumped the oil  from the  pit and  then covered the pit with  soil.
 Documentation pertaining to the  chemical composition of materials
 disposed  in the pit, the fate of the liquid removed from  the  pit,  and
 the quantities  and  characteristics of the material  buried on  site  is not
 available.

 C.6.1 Nature and  Extent of Problem
      In August 1976, the  failure of an earthen berm allowed  approxi-
mately 20,000 gallons of black oil  to escape from an unlined  pit on the
property and flow into an on-site creek.   The oil  remaining in the  pit
was pumped and transferred to tank trucks and hauled away.  The bottom
sludges and some oil-stained soils  were excavated and disposed in  the
County Landfill.  The remaining thicker sludges,  which could  not be
                                C-53

-------
                    SITE  LOCATION  MAP
               SOURCE: USCS OUAtXUNGLE ACME.
                    NOflTH CAROCXA 1514.
CONTOURS AND ELEVATIONS
     IN METERS
                   SCALE
FIGURE C-5
                              C-54

-------
 pumped,  were mixed with sand and  buried  on  site;  however,  the  burial
 location was not documented.
       In 1980,  the property changed  hands through foreclosure  and
 subdivided for  residential  development.  Family A purchased  two  lots  in
 the  subdivision in 1982 and 1983  and in  July  1983,  uncovered creosote,
 oil  spill  residue,  tank bottom  sludges,  and septic tank  sludges  in  the
 yard.
       State of  North  Carolina personnel  sampled and analyzed soils  at
 the  site in response  to the discovery.   The investigation  confirmed the
 presence of these materials in  soil  close to  the  surface at  locations on
 all  sides  of the house.   The well was contaminated  with phenols  (100
 mg/1), dimethyl  benzene,  methyl phenol,  one unknown organic  compound,
 creosote,  and septic  tank sludge.  The shallow well  was subsequently
 condemned  and the house was connected to a neighbor's well system.
       In September 1983,  EPA and  the Region IV Field Investigation Team
 (FIT)  conducted  an  investigation  of  the  property.   The investigation
 included an electromagnetic survey of the area; initial ambient  air
 monitoring  under the  home;  and collection and laboratory analysis of
 surface  water and sediment  samples,  five subsurface soil samples, and
 six groundwater  samples  from off-site and on-site wells.   All readings
 of ambient  air were negative.  Based on  surface water and  sediment
 sampling and  analysis,  no contamination of  Chinnis  Branch was detected.
 Groundwater analysis  for  inorganic compounds  indicated the presence of
 seven priority pollutant  constituents.  Groundwater analyses for organic
 compounds  (purgeable, extractable, pesticide/PCBs,  and other chlorinated
 compounds)  indicated  the  presence of 81 different compounds, many of
which are commonly associated with creosote or coal-tar derivatives.
 Soils analyses identified a  total  of 24 inorganic constituents in the
 subsurface  soil  samples,  10  of which are target compound list (TCL)
parameters.   Soils analyses  also  indicated the presence of 46 organic
compounds in  the  subsurface  samples,  14 of which are TCL parameters, and
many of which are polynuclear aromatics associated with asphaltic and
coal-tar derivatives.
      In February 1984, EPA  used ground penetrating  radar  (GPR) to
further delineate the site  boundaries.   The  GPR survey revealed two
                                C-55

-------
anomalous areas  that  roughly  correspond  to  the location  of  two  surface
impoundments  illustrated on historical maps.  The pits were estimated to
be six feet deep toward the center  and three feet deep toward the edges.
      In March 1984,  an Immediate Removal Action (IRA) at the site was
requested by  the EPA.  The IRA at the property consisted of the excava-
tion and removal of approximately 1,770  tons of oil sludge  and  soils
which were transported to a hazardous waste landfill  in  Pinewood, South
Carolina.  Soil cleanup activities  were  completed in  April  1984.
      In May  1984, EPA installed nine groundwater monitoring wells at
the site and  conducted sampling and analysis of groundwater samples for
volatile organic compounds.   Relatively  high concentrations (in the ppm
range) of benzene, ethyl-benzene, toluene, and xylenes (BETX) were
detected in the groundwater samples from the shallow  aquifer.  Groundwa-
ter sampling  indicated that the deep aquifer had not  been affected by
the contamination in the shallow zone.   Neither the upgradient nor down-
gradient deep wells contained detectable volatile organic compounds.
EPA recommended that these monitoring wells be regularly monitored.
      In 1988, the groundwater monitoring wells were  resampled and
samples analyzed for purgeable organics  and some inorganic  parameters
(metals, nutrients, etc.).  As was  the case with the  EPA 1984 data, the
1988 monitoring well data indicated that gasoline or waste  petroleum
product (especially benzene,   toluene, and xylenes)  were still prevalent
at the site.   However, the concentration of purgeable organics detected
during the two sampling episodes varied  significantly.  In  addition, the
1988 data indicated the possibility of low level  benzene, ethyl-benzene,
and xylenes in a deep well which would indicate that  an aquifer underly-
ing the surficial aquifer had now been affected.
      In 1989, EPA determined that  the level and extent of  on-site
contamination warranted a more thorough  investigation and assessment.
Consequently, a Remedial Investigation (RI) was undertaken  through a
CERCLA action.  The purpose of the  initial RI was to assess the nature
and distribution of contaminants at the  site and to provide the data
necessary for developing a Feasibility Study (FS) and ultimately
conducting a Remedial  Action  (RA).
                                C-56

-------
       The  field  investigation of the  initial  RI was  conducted from
 January  1990  through  April  1990,  and  included soil gas  surveys,  surface
 and  subsurface soil sampling,  monitor well  installation,  groundwater
 sampling (of  monitor  wells  and residential  wells), hydraulic  conductivi-
 ty testing, and  identification of potential ecological  and  human
 receptors.  Three  separate  study areas were addressed during  the initial
 RI.  Area  1 was  comprised of  the Family A Property,  which is  the
 original site.   Area  2,  located approximately 1.5 miles from  Area 1  in a
 similar  rural residential area,  was identified during the document
 review.  Historical records indicated that  a  waste oil disposal  pit
 existed  within Area 2.   Area  3,  located adjacent to  Area  1, was  identi-
 fied in  historical records as  a potential waste disposal  site based  on
 historical aerial  photographs  of the  area.  Area 2 was not  well  docu-
 mented and could not  be  located even  after  extensive interviewing and a
 thorough reconnaissance  of the  area.   Therefore, no  field investigation
 was initiated.  A  majority of  the  field investigation activities
 subsequently  centered in Areas  1  and  3.
      The initial  RI  report concluded  that  the extent of  contamination
 which poses a risk to human health or  potential ecological  receptors is
 limited to the area in the vicinity of the  Family A residence.   Contami-
 nation has impacted surface and  subsurface  soils, the shallow aquifer,
 and creek sediments in this area.  The predominant media  and  compounds
 contributing to public health  risk include: benzene and lead  in ground-
 water,  and carcinogenic polynuclear aromatic compounds (PAHs)  and  lead
 in surface soils.

 C.6.2 Residential Air Monitoring
      Air samples were collected  in February  1990 in the  crawl space and
 inside the residences  of Family A and Family B.  Sampling was  conducted
 at these two residences since they are situated on or near the former
waste disposal pits.  Unlike the Family A residence,  the Family B
 residence is not located above or adjacent to identified  contaminated
 areas.
      A total  of five  residential air samples were collected  using EPA
method  TO-14 from within the crawl spaces and interiors of the two
                                C-57

-------
homes.  The  air  samples were analyzed for the 34 TCL VOCs listed in
Table 2.18.  Only three compounds, chloromethane, methylene chloride,
and 1,1,1-trichloroethane were detected.  Methylene chloride was
detected  inside  the Family B residence at a concentration of 11 ppbv.
Low levels of chloromethane (16 ppbv) and 1,1,1-trichloroethane (1.5
ppbv) were detected in the crawl space beneath the Family B residence.
No VOCs were detected within or beneath the Family A residence.
      Of  these three VOCs, 1,1,1-trichloroethane was the only compound
detected  in  sampling in the vicinity of the site and it was only
detected  in  low//g/kg concentrations in background soil samples.  None
of these  VOCs were detected in known contaminated areas.   These
observations suggest that the source(s) of VOCs in the Family B resi-
dence are not related to the site.

C.6.3 COMMENTS ON INDOOR AIR EFFORT
      The fact that no VOCs were detected in or under the Family A resi-
dence is counter to published information on typical  residential indoor
air quality. These data are suspect.
      Perhaps of more significance is what was not sampled and analyzed
for.   The soils analysis demonstrated the site is contaminated with
creosote,  polynuclear aromatic hydrocarbons, phenols,  and other hazard-
ous compounds.  Even low concentrations of some of these can result in
substantially elevated risk numbers.  Vapor pressures are high enough to
consider their transport.through soil gases especially considering
contamination was found in the immediate vicinity of the residence.
                                C-58

-------
   TABLE C.18



TARGET COMPOUNDS
COMPOUNDS
CHLOROMETHANE
BROMOMETHANE
VINYL CHLORIDE
CHLOROETHANE
METHYLENE CHLORIDE
1,1-DICHLOROETHENE
1,1-DICHLOROETHANE
CHLOROFORM
1,2-DICHLOROETHANE
1,1,1-TRICHLOROETHANE
1.1.1 -TRICHLOROETH ANE
CARBON TETRACHLORIDE
DIBROMOCHLOROMETHANE
1 ,2-DICHLOROPROPANE
TRANS-1 ,3-DICHLOROPROPENE
TRICHLOROETHENE
BENZENE
DIBROMOCHLOROMETHANE
1 ,1 ,2-TRICHLOROETHANE
CIS-1 ,3-DICHLOROPROPENE
BROMOFORM
1,1,2.2-TETRACHLOROETHANE
TETRACHLOROETHENE
TOLUENE
CHLOROBENZENE
ETHYL BENZENE
TOTAL XYLENES
ACETONE
CARBON DISULFIDE
VINYL ACETATE
METHYL ETHYL KETONE
METHYL ISOBUTYL KETONE
METHYL BUTYL KETONE
STYRENE
1,2-DICHLOROETHENE (TOTAL)




















      C-59

-------
C.7   FUGITIVE DUST CASE STUDY
      Primary zinc smelting operations at this facility (Figure C.6)
since the turn of the century emitted large quantities of zinc, cadmium,
lead, and copper into the atmosphere in the vicinity of the plant.
Significant concentrations of these heavy metals have been measured in
the soil within a large area surrounding the plant.  The plant area is
currently a Superfund and a RCRA site.  The facility is currently
operating under new ownership as a secondary zinc smelter.  A NAMS
reference monitor is actively monitoring the area for attainment of the
lead NAAQS.  The facility has been in compliance with the standard for
the past several years.

C.7.1 Nature and Extent of Problem
      Investigations were conducted at two homes by the State in
response to citizen requests.  Two dust sampling methods used were based
on protocols of the Center for Disease Control,  Lead Poisoning Preven-
tion Branch, Division of Environmental  Hazards and Health Effects,
Center for Environmental Health and Injury Control (CDC).   The first
method used a vacuum pump and filter cassette to pick up dust from a one
square foot area.   A plastic template was used to measure the area to be
sampled - except on window sills where the area was estimated.  The
second method, based on the Kellogg Dust Protocol Number 2,  was wipe
sampling.  A Whatman Number 42 filter paper dampened with isopropyl
alcohol  was used,  although the original  method specified alcohol swabs.
      Field blanks were taken at each residence by momentarily opening
the cassette.   The wipe sample field blank filter was removed from the
transport tube,  moistened with alcohol,  then immediately returned to the
transport tube.   A new disposable latex glove was used to handle each
filter.
                                C-60

-------
PREVAILING WIND
                             X = AIR MONITORS
                  MINE TAILINGS
                 1  SMELTER
                 X  COMPLEX
                                  LOCATIONAL MAP

                                       FIGURE C-6
                           C-61

-------
      The  U.S.  Public  Health  Service/EPA  Toxicological Profile Document
 for  lead uses  a value  of  300  /jg/g  lead  in household dust, as a baseline-
 value to estimate  overall population  lead exposure from various media.
 The  urban  Particulate  Standard  1648 from  the National  Institute for
 Standards  and  Technology  (NIST), which  is a composite  urban dust sample,
 has  values of  6,550 jig/g  for  lead, 75 /ig/g for cadmium and 4,760 /zg/g
 for  zinc.  The  lead values obtained from  the household dust in the two
 homes were well  above  the national averages.
      An indoor monitoring study was developed to determine the extent
 of exposure to  lead and heavy metals  inside the residents homes.
 Twenty-five homes  were selected randomly  in areas where there is a large
 probability of  deposition of  airborne lead.  Atmospheric dispersion
 modeling with  ISCLT was used  to determine these areas. The homes were
 vacuumed sampled with mini rams outfitted  with HEPA filters.
      Results from the indoor study demonstrated that most of the homes
 in the area had  very high levels of lead.  However, since a health
 threat had not yet demonstrated to exist  due to the high levels of lead
 found in the residential homes, ATSDR was contacted to determine whether
 the  environmental  contamination is a public health hazard for the
 residents.
      ATSDR proposed an exposure study to determine the extent of the
 health hazard,  if  any, in the town.  Although the study has been
 completed, the  results have not yet been  reported.
      The study:
      •     Measured blood lead and urinary cadmium levels  among  a
            representative group of residents
      •     Compared the levels of lead and cadmium found  in  residents
            to levels  found  in a comparison community while controlling
            for other  risk factors known to influence exposure  to these
            heavy metals
      •     Performed  a standardized  panel of  medical  tests on  both
            groups

      A Superfund  Removal  Action was scheduled to be performed at those
homes in with elevated lead levels.  The  objective was to clean the
homes and eradicate them of lead, thus reducing the probable  health
                                C-62

-------
 threat.   However,  Superfund is  reluctant  to  perform this  Remedial  Action
 until  adequate evidence is  available  to  identify  the source  of  lead dust
 currently infiltrating  the  homes  so that  recurrence can be prevented.
 Data obtained in  the  25 home sampling program  showed that lead  levels
 inside the homes  were much  higher than in the  outside soil.  Although
 this suggests an  airborne pathway for the indoor  air contamination, the
 data did  not  conclusively establish whether  the contamination was  from
 soil erosion,  the  smelter,  or other sources.
      The problem,  therefore, has become  one of determining  the actual
 current source of  lead  dust infiltrating  the homes.   Based on current
 information,  it can be  demonstrated with  reasonable scientific certainty
 that:
     1.   Past primary Zn smelting practices have  contributed to the
          burden of  hazardous  substances in soil and dust
     2.   Present  secondary  Zn recovery processes  are  adding  to the
          burden of  hazardous  substances in soil and dust
     3.   Both  1. and  2. are  "non-de minimus"
     4.   Other common anthropogenic sources are responsible  for little
          of the hazardous substance burden

     The  problem, thus, becomes one of determining  whether the lead dust
 is from the old primary smelter operations and is being tracked into
 homes or  carried in by wind erosion,  or whether it  is from wind dispen-
 sion of the current secondary smelting operations.
     There are  five basic identification methods that can be used:
geostatistical, elemental composition  coupled with multi-variate analy-
 sis, chemical  speciation, individual  particle analysis, and lead isotope
ratios.   In the geostatistical approach, the spatial variability of the
substance of  interest is determined and displayed on isoplethic plots.
while this provides an easily understandable presentation from which one
can infer source locations,  a large number of highly representative
samples are needed.
     The elemental composition method  requires determination of multiple
elements from both potential sources  and receptors.   The data are then
analyzed using multi-variate techniques such as ratio discriminant
                                C-63

-------
 functions and factor  analysis.  While  one can potentially  learn  a  lot
 from a few samples and  infer  sources by comparison of  source and
 receptor patterns, the  method  is  not sensitive for subtle  sources  and
 post-release chemical transformation can pose analysis difficulties.
 Also, the method depends on having  a fairly complete picture of  all
 potential sources.
     Chemical speciation methods  depend on determining the actual
 chemical compound or  "species"  in source and receptors.  Very selective
 analytical techniques are required.  The method suffers significantly if
 post release transformations take place.  For lead, many source com-
 pounds may weather to lead sulfate making source identification diffi-
 cult or impossible.
     Individual  particles may be  analyzed microscopically and by various
 analytical techniques to determine elemental and chemical composition.
This approach appears to combine  the best features of the chemical
 speciation and multi-variate methods.  For example, particles from lead-
based paint would be associated with carbonates,  chromates, titanium,
and zirconium whereas lead particles from batteries would contain
antimony.
     The use of lead isotope ratios can potentially be a very powerful
technique, especially for the current problem.  Lead ores from different
geological ages  contain different ratios of the four lead isotopes
206Pb", 207Pb, 208Pb to  the minor isotope  204Pb.   Because of this,  different
economically important lead ore bodies have different lead isotope
fingerprints as  shown in Table 2.19.  These ratios can easily be
determined using either thermal ionization mass spectroscopy (MS) or
 inductively coupled argon plasma MS.
                                C-64

-------
                                TABLE C.19
               LEAD ISOTOPE RATIOS IN WORLD-WIDE LEAD ORES
Source
Rosetta, S. Africa
Ivigtut, Greenland
Broken Hill, Australia
Bunker Hill, Idaho
Coeur D'Alene Group
Pine Point, NWT
Tintic, Utah
Casapalca, Peru
Metalline Falls, WA
Missouri Leads
206/204
12.5
14.8
16.0 -
16.2 -
16.1 -
18.2 -
18.5 -
18.6 -
19.5
Ratio


16.2
16.4
16.7
18.6
18.7
19.0

21.0 -24.0
            It  is suspected  that  lead ores used during primary smelter
operations  were from a  limited number of geologically well-defined
sources and the secondary smelting operations process lead from a highly
average group  of world-wide  sources.  Thus, the lead isotope ratios will
be substantially different  from  these operational periods.  If this is
the case, then the isotope  ratios in  household dusts and collected
human body  fluids will  provide convincing evidence for the current
source of the  lead dust.  Furthermore, it is likely the primary Zn
smelter emissions were  reflective of sulfide ore feedstocks containing
Zn, Cd Pb,  As, Se,  Cu,  Ga,  In, Ag, and the slag was rich in iron and
manganese.  Secondary zinc smelter operations, on the other hand, likely
reflect compositions of diverse  ferrous alloy feed materials and contain
high levels of elements not  associated with primary zinc smelting such
as Be, V,  Co,  Ni,  Mo, Pd, Pt, and lanthanides.
     Based on the scientific evidence and available analytical  methodol-
ogies, the  investigative course of action to be taken is:

1.  Review existing site characterization data;  identify source areas;
    review site history (what ores were smelted).   Review other informa-
    tion such  as RCRA records, Air Enforcement records,  OSHA records -
    (the smelter has been cited in the past for worker exposure in the
    plant  and  subsequent high blood lead levels).
                                C-65

-------
2.  Develop piggyback sampling strategy for residential areas; obtain
    samples of sources  (slag, primary smelter emissions, secondary feed
    materials, waste piles).
                        Types of Samples Desired

    Sources

    1.  Primary Zn smelter stack emissions
    2.  Slag piles
    3.  Pre-1980 air filters
    4.  Post-1980 (present) air filters
    5.  Fugitive sources of dust from secondary Zn plant (piles, bins,
        etc.)
    6.  Raw materials received
    7.  Products made
    8.  Present day stack emissions, baghouse dust, etc.

    Receptors

    1.  House dust
    2.  Soil from yards, public areas, etc.
    3.  Hillside soil
    4.  Park, etc. dust

3.  Apply a combination of the following approaches:

    a.  Elemental composition + MVA
    b.  Lead isotope ratio analysis
    c.  Individual particle analysis


C.7.2   Comments on Study
     The investigative procedures outlined are currently underway.  The

value of this case example is that it illustrates that investigative
techniques are frequently available to distinguish impacts from specific
current sources from other current and past emission sources.  Although
this is a very specific case involving smelters, the fingerprinting

investigative approach may be applicable to other investigations of

Superfund site impacts on indoor air quality.
                                C-66

-------
 C.8   BUILDINGS OVER WASTE OIL PIT
      The  Site is located in Mobile County,  Alabama.   The  5.1  acre  site,
 about 11  miles north of Mobile,  Alabama,  is bounded  to  the  east  by U.S.
 Highway 43,  by a gas pipe line easement to  the  north, an  undeveloped  lot
 to the west,  and a residential  development  to the  south.
      According to available records,  in March 1961 a trucking terminal
 was constructed for intrastate and interstate commercial  trucking  opera-
 tions.  The  Site was used for parking,  maintaining,  and cleaning trucks
 and trailers.   According to a salesman  at the Site,  from  May  1966  until
 late  1969, tank trailers were regularly and routinely washed  at  the
 Site.  The washwater was then discharged  in the  rear of the terminal
 property  where the ground level  was  low.
      The  salesman states that he was  told that  in late  1965 (before he
 worked at the  Site)  levees  were  built around the terminal to contain
 materials and  an aboveground  asphalt  tanker on wheels was placed at the
 back  of the  property for storing asphalt  products to be returned to
 customers.
      In February 1971,  the  terminal operator was encouraged to move its
 operations.   In  March  1971, an offer  to buy the property was accepted.
 Figure C-7 shows the terminal  layout  as of  May 1972,  including the
 configuration  of a dike  built in  1971.   In  September 1971, six geotechn-
 ical   borings,  were drilled  at the  Site  to depths ranging from 40 feet to
 55 feet.  The  data collected  from  these borings were used to design the
 foundations of the Site.  Logs of  these borings were compiled on a
drawing that was part  of the  building architectural plans.  These boring
 logs  show the  stratigraphy  underlying the Site consists of a sand
 stratum,  up to 7.5 feet  thick, overlying  a  silty clay stratum, which
attains a thickness  of up to  34  feet.   Underlying the silty clay is
another sand stratum of  which the  thickness was not determined by the
borings.  To date,  a drawing  showing  the  locations of these borings has
not been  found.
                                C-67

-------
C-68

-------
      In  1973,  the  property was sold to the  property's  current  owner
which constructed  an  apartment complex (see Figure  C-8)  consisting of  13
buildings  on  the Site.   Approximately 160 people  live  in the apartment
complex.

C.8.1    Nature  and  Extent  of Problem
      In  1977,  the  owner  first became aware  of  tar-like substances
seeping  to the  surface.  The Department  of  Housing  and Urban Development
(HUD), which  inspected the apartments on an annual  basis,  inspected the
seepage  and recommended  to the managing  agent  that  the seepage be
excavated  to  a  depth  of  1^ feet and the  area be filled in  with sand.
This was done periodically as the  seepage appeared.  HUD continued to
inspect  the apartments on  an annual  basis.   The managing agent reported
the presence of the tar-like substance to the  Department of Health.
They continued  to excavate the property  and fill with  sand when any
seepage  occurred.  Continued appearance  of  the tar-like  substance
prompted the manager  of  the  apartment complex  to report  the matter to
the State  in the fall of 1984.
     The State  investigated  the Site  and  reported the  matter to the U.S.
Environmental Protection Agency (USEPA)  during the  same  year.  The State
identified 18 individual seeps  where  a tar-like substance was oozing to
the surface and described  the material as a tar or  asphalt with a
vitreous luster. The material  was described  as very sticky with a
petroleum odor  and at cooler temperatures was  very  plastic and viscous.
The investigation included six  hand augured  holes ranging from 1.1 feet
to 5.0 feet in depth.  Samples  were taken from two  holes in the interior
grassy courtyard area of the apartments  and  one hole in  the southwest
corner of the complex in a grassy area inside  the paved  driveway.   One
analysis was taken from these  three soil  samples although it is unclear
                                C-69

-------
: - 0
  FIGURE
                            C-8
C-70

-------
 if this  was an analysis of a composite sample.   The sample  contained  the
 following compounds  and concentrations:

         Compounds                  Concentration  (ug/g)
         Naphthalane                        135.4
         Acenaphthene                      101.9
         Fluorene                           128.6
         Phenanthrene                       33.9
         Anthracene                          53.7
         Benzene                             0.05
      Preliminary sampling  by the  USEPA in  April  and May of  1985 detected
 concentrations of naphthalene,  1,2,4  trichlorobenzene and other com-
 pounds in the  tar-like  material and soil.   In February 1990, the USEPA
 added the Site to the NPL  with  an MRS score of 30.83.
      Compounds reported to have been  detected in the subsurface 4 feet
 from  the southeast corner  of the  courtyard  area  included butyl ate
 (20,000  ppb),  vernolate (90,000 ppb)  and a  mixture  of compounds normally
 found in coal-tar products.   Compounds reported  to  have been detected in
 the subsurface 4 feet from the east-central  courtyard included butyl ate
 (2,000,000  ppb), vernolate (300,000 ppb),  eptam  (30,000 ppb), and coal-
 tar products at a total  concentration of (70,000 ppb).
     No  organic compounds  were reported to  have  been detected in a
 composite surface sample of scrapings of white material from nine areas
 in the complex.  However,  this sample was  reported  to contain 37,000,000
 ppb sodium, with a Ph of 8.7.
     Samples SA-17,  SA-18,  and SA-20  were  surface samples of tar-like
 seep material.  Sample  SA-17  from the east  side  of  the Site was reported
 to contain  vernolate (1,000,000 ppb),  1, 2,  4-trichlorobenezene
 (9,300,000  ppb), coal tar  products (total concentration of 380,000 ppb),
carbon disulfide (5,000-ppb)  and  two  unidentified terpenes (20,000 ppb).
A petroleum based product  was also reported  to have  been detected in
this sample.   Sample SA-18 in the  southeastern corner of the courtyard
was reported to contain butylate  (70,000 ppb) and vernolate (300,000
ppb) along with some coal  products (total concentration of 760,000 ppb).
A petroleum based product  was also reported  to have  been detected in
this sample.   East of the  ditch by the playground, Sample SA-20 was

                                C-71

-------
reported to contain coal-tar products at a total concentration of
(2,160,000 ppb), xylene  (2,000 ppb)  and a petroleum product.  The USEPA
samples of April and May of 1985  found no contamination at the sampling
points located outside or downstream of the reported position of the
former levee.
     A Phase I RI was conducted in February and March of 1991.  Phase II
was conducted in September of 1991.  The investigation was quite
thorough and reporting all details is beyond the present scope.
     Given in Table C-20 is a summary of the number of compounds found
on-site.
                               TABLE C-20
        SUMMARY  OF  NUMBER OF CONSTITUENTS DETECTED  IN  EACH MEDIUM
Media Type/
Sample Location
Alluvial Groundwater
Surficial Groundwater
Soils
Ditches
Tar-like Material
Total
Detected
Constituents
41
66
110
68
69
Total
Constituents
with
Unqualified
Detections
27
37
40
20
34
Total
Constituents
with More than
One Unqualified
Detection
15
24
32
13
19
     Table C-21 gives organic chemical concentration data from the
monitoring wells at the Site.
                                C-72

-------






























r-
1
O
UJ
03








































CD
'I
5
c
Q
II

_
"o
a
c

o

"c
o
5

II
5
5

VI
c
o
CO
£
-Q
"el <
gj =
II

0
o
0}
c
o
3
a
_J
Q
—
c
o
o
03

o
cc
II
UJ
cc

c
n
a
c
0
o

0}
c
o
a
0)
II
f2
u
'5
0










Z
o
J—
ui "J
0^
II) *-
ti U ^J
Q 2 (n
O ~™N
U





























O O O O O O «— O ^" O 00 O •— LOroOOOOOOOOT-OOOO^l"O
•— OOOOcoOO*.OOCoOOOOtOOOOOMCScr)
co co ro co J CS «— «~ CJ »— ^-r-CO




















--
                 a.

                 O
                 o
                 •z.
                 o
                 m
                 cc

                 o
                 o
                 cc
                 a
      UJ
      x
      a.
cc.
O
                      X
                      UJ
                      X


                   uj X

                   O uj

                   iu —

                   O w
                   < CQ
              CL


              O
                               $5
                               _i ca
                               < cc.
                               x <
                               H O
                               X O
                               a. cc
           X X
                   ui
X

X
a.
                                       X
                                       UJ
                               x a    x


                               ||H|§S
                               UJ CC Q UJ O O
                               CM O 7 CM Z I-

                               ^ J " a £ S
                                       ca > <
        o w x
        < ca u
           g_



           X
           UJ
           I
        Q «V


        V M

        •
H X J
£?£
X


X
Q.
                                                                 UJ
 1    fc
i£    2
: -7    N
X 5   .    X S
UJ ', UJ    UJ  .
X Ci Z    X -J
-J •. uj    -JO
                                                              (E X

                                                              3x

                                                              5^
                O
                z
                "I I £i fc
                "t < " 3
                CM u ca ca
                                                                            o
                                                           ui

                                                           <

                                                           O
                                                                      u ca
                      o. uj
                      UJ >
0
a
a
c
a
o

6
                                           UJ UJ
                                           OL CC
                                                                       UJ  UJ
                                                                       cc  cc
pooppopooqoqpoppooooopcorococococofpco
                   C-73

-------
•a
 0)
 c
 o
o
o
UJ
_l
QQ
o
0

=
«5
o
o
II
—
o
£
o
c

o
'c
o
5
II

5

M
C
o
s
o
.O
"°l<
ll
.2
Q
II
_J
Q
c
o
i
X
o
ir
II
UJ
CC
0
JD
(0
r
o
o
$
GO
C
o
0
0
w
II
w
5
1
                   o
                      0
                             OOOfOOOOOOOOOOOOOOOOOOOOOOOO
                             pr-f-;(^qoqoqq'-;'-_oqqoqqqqqqoqqq
                                                                                    O)OU1
                                                                                    T 00
                                                                                           TJ- CQ CO
                                                        C-74

-------
 Indoor Air Modeling
      No actual  monitoring was  conducted  inside  the  structures.  All
 estimates  are based on modeling.   [The modeling approach  for  soil gas
 intrusion  is described in Appendix A, Sections  A.1.5  and  A.1.6].
      In this case,  the modeling  approach to  estimate  soil gas flux was
 based on the assumption that the  contaminants are initially uniformly
 distributed vertically and horizontally  throughout  the soil.  The model
 then  allows equilibrium to be  established among chemicals adsorbed on
 soil,  chemicals dissolved in soil  water,  and chemicals in the soil gas.
 The soil gas diffuses  to the surface resulting  in an  ever-expanding zone
 of clean soil  from  the soil surface downward.   It appears that the model
 assumptions result  in  an average  flux over a ten year period.  These
 source  modeling equations were based on  and  conform to those given in
 EPA/600/6-86-002, "Development of  Advisory Levels for Poly-chlorinated
 Biphenyl (PCB)  Cleanup."
      The above  calculation yields  only the flux (mass of chemical
 exiting  the soil surface per unit  area per unit of time, eg. g/m2-s).
 To calculate the rate  of entry into the  structures, it was necessary to
 make  estimations of:
      •  the ratios  of  diffusive to  convective flow of soil gas,
      •  the area of the  structure  through  which soil gas could enter,
      •  the air exchange  rate for  the structure.
      The ratios of  convective to diffusive flow was calculated from the
 Peclet  number,  a dimensionless parameter,  that  considers the diffusiv-
 ity,  soil permeability,  and vacuum  created by the structure.  The
 calculation  indicated  the  diffusion mechanism dominated (see comments
 section below).  The area  through which gas  could enter the structure
was estimated as 0.1 percent of the floor  area,  based on literature data
 for homes on slabs.  The  structure  air exchange rate was estimated to be
 about 0.7 air exchanges per hour, based on average wind speeds and
 indoor-outdoor  temperature difference.
      In addition, outdoor  air concentrations can contribute to indoor
air concentrations.   Since the buildings are surrounded by the Grassy
Area,  indoor air concentrations can be affected by vapors from the
Grassy Area.  Thus,  the total  indoor air concentration was estimated by
                                C-75

-------
 summing  the modeled  indoor air  and  100%  of  the  Grassy  Area  air  concen-
 trations  (indoor  exposure,  excluding  seeps).  Seeps  of tar-like material
 may also  contribute  to  the indoor air concentrations.   Thus, total
 indoor air concentration  including  vapor emissions from seeps was
 estimated by  summing  modeled  indoor air  concentrations,  Grassy  Area air
 concentrations  (99.66%),  and  emissions from seeps of tar-like material
 (0.34%)  (indoor exposure,  including seeps).  This conservatively assumes
 that all  tar  seeps are  located  in the Grassy Area and  contribute 100% to
 indoor air.   The  results,  excluding contribution from  seeps, are given
 in Table  C-22.
     Data from all sources was  used to complete a risk  assessment.  The
 results for the current exposure are  summarized in Tables C-23  and C-24.
 (Risks for future exposure scenarios, which assume use  of on-site wells,
 are not included  in the table,  but  are discussed below.
     Total current cancer  risks including exposure to  seeps were
 estimated to  be between 7 x 10"5 and 2 x  10"4 for the various receptor
 populations.  The most  important pathway contributing to the risk is
 indoor inhalation of  vapors containing PAHs from seeps  of tar-like
materials.  In particular, benzo(a)anthracene, benzo(b)fluoranthene, and
chrysene contribute about 90% of the  risk from indoor  inhalation.
     Under the Future I scenario including  seeps, the cancer risks range
between 4 x 10"4  and  2 x 10~3.    Ingestion of water from the alluvial
 aquifer accounts for most of the cancer risk,  approximately 75% of the
total  risk.   Arsenic  and beryllium constitute the greatest proportion of
the risk, with aldrin, BEHP and dieldrin also contributing to the total
risk.
     The Future II scenario including seeps poses risks ranging from 3 x
 10"3 to 1  x 10"2.   Inhalation of indoor air vapors during showering
account for approximately 100% of the risk,  and the risk is primarily
due to chloroform from the on-site  surficial aquifer.
                                C-76

-------
    CO
    z
    O
    h-

    cc
    l-
    z
    UJ
    o
(N  Z
eg  o

i  °
LLJ  £
CO
    oc
    o
    o
    Q
    Z

    Q
    LU
    _l
    UJ
    Q
    O






c
a
i:
V
u
c
o
u
"o










9 £ i-0
- (3^-2
c —
l}|i
ujS
Oi on
||

a 	 	
III
m
a en
_ 3
^
l=\
a "~
si
i!li

u -^.
Itf
- • 1.
Contaminant!
?
O
in
1

s
•

3
ff>
§
CM
O
§

6.248-06
45.50471
6.078-02
S
id
a
ACENAPHTHENE
?
S
CM
1

O
ft

CM
O
e
to
CM
O
O
3

1.248-05
0.130183
8.588-04
o
§
Si
ACETONE
CM
§
n
(D

8
O

8
|
8
S

4.508-06
1
1
5
O
CM
CO
in
ALDRIN
?
to
CM
§
01

CM
O
O
p

8
O
O
s

5.778-06
128.2497
S
CO
S
o
CO
N
ANTHRACENE
^
CO
5
n
c\j

8
O
CM
in

8
i
(vi
O
8

s
in
1811.575
4.830-05
in
o
e
in
8
BENZO(A)ANTHRACENE
CM
£
CO
CM
m
p

CM
O
C
CO

CM
O
O
r-
to
S
c
in

CM

0
c

o
3
0
0
CT
ri

4.806-06
9384 613
3.948-O4
S
CM
0
CHLORDANE. ALPHA
CM
Ol
tn
CO
CM
*

5
fe

o
V
0
o

4.808-06
295.7S52
3.948-04
o
c
p
0
CHLORDANE, GAMMA
o
n
i-
0
I

8
«

8
CO
8


7.306-06
3.14816
9
in
CM
O
§
s
CHLOROBENZENE
?
to
o
1

o
at
q

O
ri
8
Cvi

1.048-05
2.073532
oj
S
CO
CM
O
CHLOROFORM
                                              C-77

-------


n
OJ
D
,C
C
o
O
s^
CJ
«N
1
O
LU
m
t-






o
f-
oc
f— •
2
LU
O
Z
o
o
cc
^
oc
0
o
Q
^
Q
LU
_J
LU
Q
O









C

0
C
tJ
3









- 0 * ~
c *—
1 fl"e

UJ -2
&«
It

•'-3
=? • <
J| ^ W
H S 3.

I"5
a en
3~
f-T
1 c'V
a
•i
, |j
i a

i"f
"•^
If!
j
2
9
u

*-
c
*
^-
&
CM

3
2

8
•
2i
CM

8
X
^
9
fi
n
1811575
1

8
£
CM
§




1 CHRYSENE
2
$
0
CM
*
r~

0
*
CO

8




3
A
*"
?
3
^
1
9

3
2
*
8
P
c
X
"*•
i
5
9
in

0
r*.
••^

3
%

8




8
&
*
3
1
ri
E9GICTO
I
n
0
^
n
«




| DICAMBA
CM
tn

0
$
•^

8
e>

8
4
Q
«

8
5
a
9
r?
"
414.8168
1

3
fi
to
8




V
(M
*

0
1
•n

8
X
Ol
ri

8




8
f?
•*
9
1!
*r
1129.943
3
2

a

8
g
03
9
$

rj

CM
o
£
CM
s
<^
n
72.12009
?
1
CM
S
g
Ol
S




1 FLUORENE
^
i
CM
^
c?
*-

8
rj

8
^
y
•»

8
:?
n
9
1
»
43.51676
8
£
CO
m
s
1
in
5


CC
n
	
HEPTACHL
-
m
rj
,_
a


8
§

8
*



8


9
5
*
5.174339
|

3
0
K
§
.
Cj
cc
O
I
C-78

-------


TJ
0)
a
.£
•*-*
c
o
O
"^
C\J
OJ
O
LU
m
j_





O
H
oc
H

LU
0
•z.
o
0
DC
DC
O
O
Q
Z
Q
LU
|
LU
Q
O








C
c
o
• c
(5
o
















1M||
-
§^_
•
3II-E

SlCji
ii
• "S •=

i ; fi
t- S ~-

t»
a en
^~
I"-*
= e1^
a
=f
= Ii

i-t

|ii

c
c
E
c
,3
OJ
C
C4
(M
4

8
i
8
+
§
*

8
$
*
?
S
in
232625
2.400-04
S
X
o
8


OL
m
CM
CN*
"
O

W7
8
V
8
+
£
•»

8
$
0
§
c
in
SZ9TCZ
3.050-05
8
X

a


m
o
m
csj
to

n
1
r-
8
a
8
O



8


9

tn
23.2625
\
8
2
n
a

(UNOANE)
S
5?
m
r>
f,
"
n
«

8
S
8
+
*
tfl
D

8


9

in
14389.85
2.800-O6
8
*

to
UJ
CDJPYREN
| INDENO(1.2,3-
0
t~-
to
o
i
o

o
V
S
^
?
*

0
(9
*
«
£
a
38.39342
2.280-01
S
&

«">
S

§
1 METHOXYCHt
^
1
CO
,_
i
*
8
*
8
o



8
Si
h-
«
0
to
0243466
1.910-03
0
§
o
R

i
!->
METHYL ETW
o
CT

0
3


8
S
m
9
Z
"
0
1.310-01
8
*
r-
3

UJ
o
1
UJ
S
._
*
r>
_
5

o
«
8
o



o
$
rj
9
§
in
0.764199
0.000057
8
S

1


MOLINATE
ft
i
f-
C?
0
2
n
o
-i*
to
09
S
+
in
•"

CM
O
N
rJ
9
1
in
10.42455
^
S
5

a


|| NAPHTHALEN
ft
(M

O
OJ
„
3
*
g
•*•
a
*~

o
fi

9
•8
in
I
6
g
S

g


PEBULATE
_
£

-------
TJ
 Q)
 a
_c
•4-*
 C
 o
O
CM
 I
O

UJ
_J
m
cc
h-
2
HI
O
2
O
o

DC
CC
O
O
Q
2

Q
111
_J
HI
Q
O





1
•o
m
Ja
•
u
3
"o
(A




8 - • §•£
15 sf t
- 0 * —
ill!
>->,§£
?f
U
111
311
•
11
fi
II!
a
s!
• ?-s-
^$1*
-51
S5
|-g
• - T
Ift
•t
C
a
E
a
3

9
K
h~
?
2
rs
3
+
S
S
•t-
§
n
8
o

7.009-06
I
S
1.12» + 03
8


UJ
2
a
fc
                                   C-80

-------
                    TABLE C-23




SUMMARY OF PATHWAY-SPECIFIC CARCINOGENIC RISKS

Exposure Scenario
RME Receptor (Wortt-Caee Scenario)
Sum of Adult
Cancer Rltk
Sum of
9-Year-OW
Cancer Rltk
Sum of
4-Year-OW
Cancer Rltk
Sum of 9-and
4-Year-OW
Cancer Risk
CURRENT EXPOSURE: No known documented cases of exposure at these levels
Target Area E
Ingestion of soil
Dermal contact with soil
Inhalation of vapor
Inhalation of particulates
Grassy Area
Ingestion of soil
Dermal contact with soil
Inhalation of vapors
Total Indoor Exposure
Inhalation of vapors (including seeps)
Inhalation of vapors (excluding seeps)
Ditch
Ingestion of sediments
Dermal contact with sediments
7x10"'
7 X 1 O'6
6 x 10'*
1 x 10"*
1 X 10'7'
1 X 10-"
9 X 10-"
5X 10*
2X10"7
0
0
2 X 108
2 X 10'6
6 X 10'"
1 x 10'9
6X 10"'
7 X 10"'
1 X 1 0"'
5X 10*
2X10'7
2X 10'7
3X 10"s
5 X 108
2 X 1 0'6
7 x 1 0'9
2x10'
5 X 1 0'7
2 X 10'*
1 X 10"
6X 10*
2 X 10'7
6 X 10'7
2 X 20'*
6x10-'
4 x 1 0'6
1 x 1 0'6
3 x.10-'
6 X 1 0'7
3 X 1 0'*
2X10"'
1 X 10-*
3X10'7
8 X 10'7
5X 10'*
EXPOSURE TO SEEPS OF TAR-LIKE MATERIAL: A««ume« no removal of teep* for RME Receptor*
Ingestion of tar-like material
Dermal contact with tar-like material
Inhalation of vapors
TOTAL CURRENT EXPOSURE
(EXCLUDING SEEPS)
TOTAL CURRENT EXPOSURE
(INCLUDING SEEPS)
4X 10'*
7X 107
5 X 10'*

7 X 106
5X 10"'
1 X 10*
5 X 10'*
3 X 1 0'6
9 X 10'6
2X 10'7
2 X 10*
5 X 10'9
4 X 10'6
1 X 10"4
3 X 1 0'7
3X 10-*
1 X 10-5
7 X 10-*
2 X 10"4
                         C-81

-------
                   TABLE C-24
SUMMARY OF PATHWAY-SPECIFIC TOTAL HAZARD INDICES

Exposure Scenario
RME Receptor (Worst-Case Scenario)
Adult
Nine-Year Old
Four- Year Old
CURRENT EXPOSURE: No known documented case* of exposure at theee level*
Target Area E
Ingastion of soil
Dermal contact with soil
Inhalation of vapor
Paniculate inhalation
Grassy Area
Ingestion of soil
Dermal contact with soil
Inhalation of vapor
Paniculate inhalation
Northern Ditch
Ingestion of sediments
Dermal contact with sediments
Indoor Exposure
Inhalation of vapor*, total indoor (including seep*)
Inhalation of vapor*, total indoor (excluding *eep*)
4 X 10 J
3 x 10'
2 X 10J
9 X lO*
3 X 10J
6 X 10-3
5 X 10'
0
0
1 X 10*
1 X10J
1 X 10'1
1 X 10°
3 X 10J
1 X 10'
3 X 1 0 3
7X 10-3
9X10''
3 X 10J
1 X 10'
2 X 10*
2X 10J
4x 10'
2 X 10°
4 X 10J
2 X 10 '
2 X 10'
2X 10'1
1 X 10^
7 X 104
1 X 10'1
2 X 10*
3X 10J
EXPOSURE TO SEEPS OF TAR-LIKE MATERIAL: A«ume« no removal of seep* for RME Receptor
Ingestion of tar-like material
Dermal contact with tar-like material
Inhalation of vapor*
TOTAL CURRENT EXPOSURE
(EXCLUDING SEEPS)*
TOTAL CURRENT EXPOSURE
(INCLUDING SEEPS)2
9 X 10'
7X10^
1 X 10J
4X 10'1
4X10'1
2X 10-4
2X 10"3
2X 103
1X10°
1X10°
9 X 10"4
3 X 10J
2X 10-3
2 X 10°
2 X 10°
                     C-82

-------