Final Report on the World Trade Center (WTC) Dust
                         Screening Method Study
                             August 17, 2005
                               Prepared By:
                   U.S. Environmental Protection Agency,
                 Office of Research and Development (ORD),
              Research Triangle Park, NC and Washington, D.C.
                                   and
               U.S. Environmental Protection Agency, Region 2
                           New York, New York
The use of trade names does not imply endorsement and is for illustrative purposes only.

-------
ACRONYMS:
ATSDR      Agency for Toxic Substances and Disease Registry
AVG        Average
COPC        Contaminant of Potential Concern
EPA         U.S Environmental Protection Agency
EPIC        Environmental Photographic Interpretation Center
ERT         U S EPA's Emergency Response Team
HEPA        High Efficiency Particulate Air
LI           Long Island
MMVF       Man Made Vitreous Fibers
MQO        Measurement Quality Objective
ND          Non-Detect
NEIC        U S EPA's National Enforcement Investigations Center
NERL        U S EPA's National Exposure Research Laboratory
NJ           New Jersey
NYCDOMH  New York City Department of Health and Mental Hygiene
ORD        U S. EPA's Office of Research and Development
PM          Particulate Matter
PM2 s        Particulate Matter Smaller than 2 5 microns
QAPP        Quality Assurance Project Plan
SD          Standard Deviation
SEM         Scanning Electron Microscopy
USGS        U S Geological Survey
WTC        World Trade Center

-------
TABLE OF CONTENTS:

EXECUTIVE SUMMARY                                                4
I  INTRODUCTION AND BACKGROUND                                  5
II. METHOD DEVELOPMENT                                            6
III METHOD VALIDATION STUDY                                       9
IV RESULTS AND DISCUSSION                                         12
V CONCLUSIONS                                                     21
VI. REFERENCES                                                      22
VII. CONTRIBUTORS                                                  22
VIII ACKNOWLEDGEMENTS                                           23
IX APPENDICES                                                      24
      A. Quality Assurance Project Plan (QAPP) for the World Trade Center Screening
      Method Study (Under Separate Cover)
      B. Data gathered by U S EPA NERL during method development
      C  Data gathered by U S EPA NERL post method development
      D- Analytical method/protocol used during study
      E  Report from USEPA contractor on Screening Method Study Results (including SEM
      calibration data)
      F  Statistical analysis and interpretation of test results
TABLES:

TABLE 1 • Average, Standard Deviation and Range of Results                    14



FIGURES:

FIGURE la: ORD-Modeled WTC Plume Dispersion                           7
FIGURE Ib EPIC Analysis of Deposition Boundaries                          8
FIGURE 2 USGS Spiking Material Results                                 11
FIGURES 4 Albany Spiking Material Results                               12
FIGURE 4 Average Slag Wool in background and spiked samples               15
FIGURE 5 Average Slag Wool in background, spiked and impacted samples       16
FIGURE 6 Average of Elements of Concrete in background and spiked samples    17
FIGURE 7 Average of Gypsum in background and spiked samples               18
FIGURE 8- Map of the origin of samples analyzed                            19

-------
                              EXECUTIVE SUMMARY

The September 11, 2001 attack on the World Trade Center (WTC) covered a large area with dust
and debns  To assist in determining if residual contamination exists in the indoor environment,
the U S  Environmental Protection Agency (EPA) initiated a study to sample indoor
environments that may have been impacted by the WTC collapse  A critical component of this
study is determining whether sampled dust originated from the collapse of the WTC or instead is
urban dust originating from other sources  This report descnbes work performed to develop and
validate a screening method for indoor dust that can be used to determine whether dust sampled
is from the collapse of the World Trade Center towers

Dispersion models, monitoring, photos, interviews, and satellite data were reviewed to discern
areas that were likely impacted by WTC emissions  and those that were not (US EPA 2002,
2004)  A total of 117 samples were collected from both impacted and non-impacted areas A
subset of these samples were analyzed by EPA's National Exposure Research Laboratory
(NERL) and National Enforcement Investigations Center (NEIC), and United States Geological
Survey (USGS) to evaluate the slag wool levels in the dust and develop an analytical method
The analytical method that was developed screens for three materials that are believed to be
present in large quantities in WTC dusts  slag wool, elements of concrete, and gypsum This
method involves the use of Scanning Electron Microscopy (SEM) to determine the quantity of
each of the materials present

Five commercial  laboratories, along with the three above listed government labs, were recruited
to test the screening method Thirty-two dust samples, consisting of both confirmed background
samples and a confirmed background dust spiked with varying amounts of confirmed WTC dust,
were sent out to the eight labs  The labs were provided the samples "blind" They did not know
which samples were background dust and which were non-impacted dust spiked with WTC dust
In addition to the thirty-two samples, one of the five commercial laboratories also received
twenty-eight background samples to increase the available data characterizing background
locations

The data reported by these laboratories indicated the following

       1) Five of the eight laboratories were able to reasonably measure the slag wool
       concentrations in non-impacted dust spiked with confirmed WTC dust

       2) A substantial amount of variability in slag wool measurements was found within labs
       and between labs Despite this variability, slag wool measurements appear to be sensitive
       enough to distinguish WTC dust (defined as 4 Albany) spiked at the 10% level from
       background dust.

       3) The levels of gypsum and elements of concrete in the spiked samples were
       indistinguishable from the levels in the background samples This suggests that, while
       these components may have been elevated in dust samples collected near the WTC site in
       September 2001 (as found by USGS in their studies on WTC dust), they are also
       commonly found in the indoor environment and would not be useful as WTC signature

-------
      components

      4) Analysis of samples during method development showed elevated levels of slag wool
      in samples from several impacted locations compared to slag wool levels measured at
      background locations
I. INTRODUCTION AND BACKGROUND

The objective of this effort was to develop and validate a means of determining whether dust
sampled as part of EPA's planned sampling program contains residual contamination attributable
to the collapse of the WTC towers The tested screening method is a critical component of the
sampling program as it will be used for two primary purposes: 1) to determine the geographic
extent of the dust remaining from the collapse impact, and 2) along with the results from
contaminants of potential concern (COPC) testing, to determine the need for a clean-up of the
sampled areas

The USGS has published two reports that provided the basis for the initial hypothesis that a
WTC collapse signature is comprised of three marker components slag wool, gypsum and
elements of concrete  The first report discusses the analysis and interpretation of indoor and
outdoor WTC dust samples collected near Ground Zero, days and weeks after September 11,
2001 (Meeker et al, 2005). From this work, we see that the WTC dust samples are dominated
by gypsum, concrete, and man-made vitreous fibers (MMVF), mainly slag wool  It is on the
basis of these key results that gypsum, elements of concrete, and slag wool  were identified as
candidates for a WTC signature The second report discusses the analysis of EPA supplied
samples taken from several indoor locations well outside of the WTC impacted area
(background)  These samples were taken between September of 2004 and April of 2005  Slag
wool was absent from many of these background samples, but Lowers et al (2005a) state that the
samples do have gypsum present, which they speculate might be due to the presence of wall
board in the  sampled apartments  Because of the lack of slag wool in these samples, USGS
concluded that these samples did not contain WTC dust  USGS also concluded that perhaps slag
wool  is the single most critical of the three WTC dust constituents when distinguishing WTC
dust from other common dusts

Other studies also identified MMVF and gypsum  as predominant components of WTC dust In a
study of air and settled dust quality in apartments  in Lower Manhattan, the  Agency for Toxic
Substances and Disease Registry (ATSDR) and the New York City Department of Health and
Mental Hygiene (NYCDOMH) found significantly more MMVF and gypsum in samples taken
from Lower Manhattan apartments as compared to samples taken from apartments in areas above
59th Street (NYCDOMH/ATSDR, 2002)  They also concluded that gypsum was  seen at a higher
percentage level in the Lower Manhattan dust samples as compared to the comparison area
samples   In a comprehensive study of the composition of settled dust in the Deutsche Bank
building at 130 Liberty Street, R J Lee identified numerous hazardous contaminants that were
present in the dust at levels much higher than in background office buildings, and among those
substances identified in  their "WTC signature" were mineral wool and gypsum (R J Lee, 2004)

-------
If the WTC building collapse signature components of slag wool, gypsum, and elements of
concrete are not present, then one could conclude that WTC building collapse dust is not present
However, since these components might be present in typical New York City dust, and as slag
wool is a component of insulating materials in currently constructed buildings, it is possible that
a test might show them to be present even though WTC dust never impacted the sampled area
A 'screening test' will, by its design, result in some fraction of such false positives (i e a
location without residual WTC dust that tests positive for the above components)   However, an
appropriate 'screening test' would result in very few, if any, false negatives (i e a location with
residual WTC dust that tests negative for the above components)

II. METHOD DEVELOPMENT

Sample Collection
EPA acquired 117 dust samples during the time period of September 2004 to April 2005
Twenty-one 'impacted' samples were taken by the EPA at two buildings that were part of the
Deutsche Bank complex located at 130 Liberty Street and 4 Albany Street Both affected
buildings were uninhabited and slated for demolition  Fifty samples were taken from locations
well beyond the impacted zone (based on modeling, monitoring and photo analysis, these
samples are considered to be 'background' dust) Forty-six samples were taken from locations
that were possibly impacted, but were a bit farther from the WTC site than the known 'impacted'
samples  None of these forty-six samples were used in the method validation study, but several
were evaluated during both the method/protocol development phase and post-study  In addition,
one impacted sample was obtained from the USGS This sample was a composite sample of
outdoor and indoor WTC dust collected in September of 2001

A standard method utilizing a High Efficiency Paniculate Air (HEP A) vacuum  collector was
used by EPA to collect most bulk dust samples  Information on this method is provided in the
Quality Assurance Plan (QAPP) for this study (Appendix A)  Some bulk  dust samples were
collected from residential and commercial vacuum cleaner bags

Modeling and satellite photography were used to determine sampling locations  for the collection
of the 117 samples  Figures la and Ib (EPA 2002, EPIC 2004) are examples of modeling and
photographic analysis used to distinguish non-impacted or background locations.  Figure la
shows ORD-modeled WTC Plume Dispersion on September 11, 2001 at 12 noon  The values
indicated by red are hourly PM2 5 concentrations (in ug/m3) measured at pre-existing NJ and NY
State-operated PM monitoring stations m northern New Jersey and New York City  Red, orange,
and yellow shading represent most likely areas of plume dispersion (red = estimated dilution to
100th to SOOth and dark blue = dilution to < one millionth of pollutant concentration at WTC
source).  As seen in this figure, the plume very rapidly diluted to concentrations less than 1/1000
(which is the yellow area) of the initial source strength at Ground Zero. Figure  1 b shows the
boundaries of collapse deposition debris as determined by aerial photographs This  photograph
was taken on September 13, and shows the four areas of "confirmed", "probable", "possible",
and "no dust" from the collapse. These areas were used in the determination of strata used in the
design for the overall  sampling program.

-------
              CAIMST Sttrfsm* Wind Reid ar»d CAtPUFP Wuroe DNutien
                              ta vaiujTMs Munse «t recovery «wt«>
                             X   V

Figure la: ORD-modeled WTC Plume Dispersion on September 11, 2001 at 12 noon.
(Source: Exposure and Human Health Evaluation of Airborne Pollution from the World
Trade Center Disaster (External Review Draft). U.S. Environmental Protection Agency,
Washington, D.C., 2002.)

-------
                                                                   Mapping Roaiifts from Saptember
                                                                     13, 2001 aerial photographs
                                                                   s  c
                                                                          Possible
                                                                           Vehicle tracks a
                                                                             possible dust
                                                                         os      a.«
                                                                      12 Ste^eiTi^f 13. 2001.  Iffiyye
                                                                 »Mosawof fo'A«i tenhatisifi and portions
                                                                 
-------
Preliminary Analysis of Collected Samples for Slag Wool
Most of the collected samples were analyzed for slag wool content by the EPA's National
Exposure Research Laboratory (NERL) Scanning Electron Microscopy (SEM) Laboratory.  This
analysis was performed as part of the EPA's development of a protocol for sample preparation
and analysis and for preliminary sample characterization.  These samples were not analyzed for
elements of concrete or gypsum as an analytical method for these components had not yet been
developed  The data acquired during this method/protocol development effort are presented in
Appendix B  Caution should be used with these data as it was obtained while the method was
being developed Post-study data acquired by NERL are also presented in Appendix C.

In evaluating the method development data acquired by NERL (Appendix B), there appears to be
a distinction between samples taken in impacted areas versus background samples Eighteen of
the 21 samples from impacted areas had slag wool at concentrations of greater than 100,000 slag
wool fibers per gram of dust, with a range of 69,000 to 13,400,000, while all of the samples from
background areas had concentrations less than 100,000 fibers/gram, ranging from no slag wool
detected (in 12 of 47 samples) to 92,800 fibers/gram of dust

Based on this preliminary work, the USGS, the EPA's Office of Research and Development
(ORD), the EPA's National Enforcement Investigations Center (NEIC), and experts five
commercial testing laboratories (denoted labs A-H in Appendix E), worked together to develop
an analytical method to identify the presence and concentration of the screening constituents (i e.
slag wool, gypsum and elements of concrete) in indoor dust  This method was reviewed by the
WTC Expert Technical Panel's signature subcommittee and is presented in Appendix  D The
composition of this technical panel can be found at http //www epa.gov/wtc/panel

III. METHOD VALIDATION STUDY

Study Design
The basis for the WTC dust screening method discussed above is as follows  if a unit has been
impacted, those materials that are found in WTC dust will be found in the dust collected from the
unit  The materials under consideration  are:  1) slag wool, 2) elements consistent with concrete
and 3) gypsum. The study described herein was intended to validate the WTC dust screening
method by  demonstrating the following things
    1) that the above described materials are reasonable markers for WTC dust (by showing that
       these markers distinguish WTC-laden dust from background dust),
    2) that WTC dust at a diluted concentration can be distinguished from background, and
    3) that the analytical method works well enough and is able to be carried out by enough
       analytical laboratories to 1) evaluate the above materials as markers and 2) distinguish
       WTC dust from background dust

The first of these three objectives was partially addressed in method development work, which
focused on slag wool   As described in the previous section, slag wool was found to be elevated
in locations deemed "impacted", while slag wool was not detected or detected at low
concentrations in "background" areas

-------
Five independent laboratones and three government laboratories participated in this method
validation phase  One government laboratory analyzed only a small portion of the samples, but
this lab was critical in the method development  Each laboratory attended a two day session
during which the method was further developed and discussed, and the protocol was adapted to
suit each laboratory's equipment.

Following this session, the laboratories received dust samples consisting of both confirmed
background samples (10 samples plus duplicates for a total of 20) and confirmed non-impacted
dust spiked with varying amounts of confirmed WTC dust (6 spiked samples plus duplicates for
a total of 12).  Specifically, a sample that was characterized and confirmed as non-impacted
(designated in Appendix B as NE Queens maid service) was split, and the splits were spiked at
levels of 1, 5, and 10% total mass with two different characterized and confirmed WTC dusts
These spiked samples were then homogenized as documented in the QAPP for this study
(Appendix A). The two spiking dusts were 1) a composite sample of predominantly outdoor dust
collected in September of 2001 by USGS, and 2) dust collected by the U S EPA from the
Deutsche Bank building at 4 Albany Street in September of 2004   The 4 Albany  Street building
borders the south side of the WTC complex.  Six spiked samples were prepared for each
laboratory, these were split so that each laboratory received 12 spiked samples. Each laboratory
also received 10 non-impacted background samples that were also split, resulting  in a total of 20
background samples. Thirty-two samples in all were sent for  analysis to the eight labs

In addition to the 32 samples, one of the five commercial laboratones also received 28
background samples to increase the available data characterizing background locations.

The labs were provided the 32 samples "blind",  they did not know which samples were pure
background dust, and which were the spiked dust  To ensure sufficient results for spiked
samples, the government laboratory that was only  able to analyze a small portion of the samples
was asked to analyze only the 12 spiked samples  Again, they were not told the identity of these
samples (Lab C)   The labs had five weeks to analyze all samples  The final data from all
laboratories, including the data for the additional 28 background samples, were reviewed,
evaluated and analyzed by the EPA and the EPA's prime contractor This prime contractor's
from this analysis is  presented m Appendix E

Composition of Spiked Samples
The USGS performed an analysis of the spiked, homogenized samples pnor to the samples being
sent to the labs  The measured levels were in the approximate range for the spiking percent (1, 5,
and 10%) based on the undiluted concentration level of each WTC dust and, in all but one case,
each percent level was fully distinguishable from the others (Figures 2  and 3)  The variability in
the measured levels was expected due to the difficulty in homogenizing dusts that have large
particle size distributions, and the fact that components of WTC dust will vary within a sample
because of the nature of the source  Given these difficulties and the measurement results, these
dusts were determined to be reasonably homogeneous

As seen in Figures 2 and 3, the level of slag wool differs between the two WTC dusts, with the
pure dust that was collected from 4 Albany Street in 2004 more than an order of magnitude
lower than the dust collected by the USGS in September of 2001. The pure dust from 4 Albany
                                          10

-------
Street had slag wool levels at 500,000 fibers/gram of dust versus approximately 11,000,000
fibers/gram of dust for the USGS collected sample.  There are likely explanations for this large
difference in slag wool levels. The USGS sample was a composite of multiple outdoor samples
and one indoor sample taken during September of 2001. The 4 Albany was an indoor sample
was taken three years post 9/11 in September of 2004. As this 4 Albany sample was taken
exclusively inside of a building, it was not only diluted by three years accumulation of urban
background dust, but was also characteristic of dust that had penetrated the shell of a building as
opposed to that deposited on the ground outside.
                                USGS Spiking Material
1,400,000-
•+*
0
2
o>
u.
200,000
0
^
t
j
j
0
Q
f5>
^

s
J1234S678910
% Spiking
  Figure 2: USGS Spiking Material Results. Analysis was conducted by USGS prior to being
             sent to labs for study.  Pure dust averaged approx. 11,000,000 fibers/gram.
                                  (Figure provided by USGS)
                                         11

-------
  3
  Q
  "5

  2
  o
  0)
  .a
90,000


80,000


70,000


60,000


50,000


40,000
       20,000
       10,000
                      4 Albany Street Spiking Material
                X
                       23456
                                     % Spiking
                                                                       10     11
    Figure 3: 4 Albany Street Spiking Material Results. Analysis was conducted by USGS
             prior to being sent to labs. Pure dust averaged approx. 500,000 fibers/gram.
                                   (Figure provided by USGS)
IV. RESULTS AND DISCUSSION

Development of Study Results
The final report from the prime contractor with all raw analytical and calibration data can be
found in Appendix E. A summary of the study results that includes the data from the 28
additional background samples analyzed by a single commercial laboratory is provided in Table
I, as well as Figures 4-7. A map of the origin of the samples analyzed during this study is shown
in Figure 8.

All background sample data used in Table I and Figures 4-7 are from the Greater NY City area.
Background samples taken in Research Triangle Park, North Carolina are not included as they
are not representative of NY City background dust. Data for all background sample results may
be found in Tables 3 and 4 of the Versar report in Appendix E. It should be noted that the
Research Triangle Park samples show higher slag wool levels than NY City area background
samples.  This is due to the presence of slag wool containing ceiling tiles in the building
sampled. Note also that Table I indicates two average values for background slag wool.  These
values reflect the inclusion and exclusion of two samples collected in New Jersey (NJ) and Long
Island (LI) that were extremely high in slag wool fibers, likely due to their insulation,
fireproofmg or ceiling tiles. Based on these results it is likely that some false positive results will
occur in buildings  with slag wool-based ceiling tiles, fireproofmg or insulation. .
                                          12

-------
Three of the commercial laboratones, designated as labs E, F and G, reported analytical data that
are not consistent with other five labs Generally, these labs were not able to distinguish
differences between the three spiking levels  In addition, these labs did not meet the
measurement quality objectives (MQOs) for the spiked samples put forth in the QAPP for this
study (Appendix A Section A.7.1)  Thus, the data from these three labs are not considered in the
results presented in Table I and Figures 4-7  The statistical analysis performed to make this
determination is presented in Appendix F In addition, Lab H was not considered when
determining concrete and gypsum levels  as their data were at least two times higher than the
sample average without these data (Table I and Figures 6 and 7)

In discussions with the commercial laboratories, it was determined that some labs did not have
the personnel or the equipment to perform the required analysis in the given timeframe, thus,
data quality became an issue  Additionally, labs that had less experience with slag wool analysis
felt that a clearer definition, in addition to that provided in the catalog developed by USGS in
Lowers et al., 2005b, of slag wool was needed to distinguish it from other mineral wools
Finally, labs that were unable to automate the gypsum and concrete analysis expressed their
belief that the method was too  long and complicated for accurate quantitative dust analysis  All
laboratory comments will be taken into consideration in when finalizing the protocol
                                            13

-------


Slag Wool
Average
(fibers/g
dust)

Elements of
Concrete
(% Area)



Gypsum
(% Area)



Background
(Greater NY Area)
AVG + SD
35,950 + 74,300
17,740+15,835*
Range of Samples
ND* - 369,230
ND* - 60,000**

AVG + SD
156 + 57
Range of Samples
6 - 30.5



AVG±SD
95 + 34
Range of Samples
4-165


USGS Spiked (Collected
9/01)
1%
94,000 + 25,740
5%
452,510+100,640
10%
870,280 + 310,420
1%
20 + 6

5%
19 + 7
70%
16 + 2
1%
9±6
5%
7 + 3
70%
6 + 05
4 Albany Spiked
(Collected 9/04)
1%
17,270 ±7,880
5%
52,510 + 26,140
10%
88,540+18,300
1%
15 + 1

5%
18±4
10%
16 + 3
1%
9 + 4
5%
5 + 2
70%
7 + 2
  •   **ND=Non Detect (Zero slag wool fibers)
  •   *Two extremely high values from NJ and LI removed

Table 1: Avg, Standard Dev., and Range of Results for Background and Spiked Samples
             (Data Summarized from Tables 1,2,3 and 4 of Appendix E).
                                       14

-------
  1000000 -i
   100000 —
in
a
•o


2
5

2
o
g



D)
a

CO
    10000
     1000
      100
                                   High Slag Wool

                                 Values in NJ and LI
                      Background
Spiked Samples
   Figure 4: Average Slag Wool (Fibers/Gram of Dust) in background and spiked samples.

                           (Data from Tables 3 and 4 Appendix E)
                                            15

-------
  10000000 3
   1000000 :
(A
3
•o
ro
5
"5
1
o
o
(A
100000
     10000
      1000
       100
            • Background
            A 4 Albany 1%
            $4 Albany 5%
            84 Albany 10%
            «» USGS 5%
            &USGS 10%
            '$• Impacted
                       High Slag Wool
                     Values in NJ and LI
                       0
                        Background
                                      Impacted
Spiked Samples
  Figure 5: Average Slag Wool (Fibers/Gram of Dust) in background, spiked and impacted
  samples. Impacted samples are locations that are shown in satellite pictures to have been
  affected by WTC Collapse Dust. Slag wool results for impacted samples were derived
  during method development and were not  part of this method validation; they are provided
  for comparative purposes. These impacted samples range from 0.1 to 1.6 miles from the
  WTC site (see Figure 8 for sample origin location). Data from Appendix B (Impacted) and
  Tables 3 and 4 of Appendix E (Background and Spiked).
                                          16

-------
  30
  25-
  20
ra
4)
  15
  10
0
$
                     T
                   Background
                                Spiked Samples
  Figure 6: Average of Elements of Concrete (% Area) in background and spiked samples.

                           (Data from Tables 1 of Appendix E)
                                          17

-------
  18 r
  16 •
  14 •
  12
  10
                      I
                      •
19
O
0
o
0
                     T
   o-i-
                   Background
                              Spiked Samples
          Figure 7: Average of Gypsum (% Area) in background and spiked samples.

                           (Data from Tables 2 of Appendix E)
                                          18

-------
               DESIGNATION
                                                           LOCATION
     3,4
     5,6
     7
     B"
     10
     il',HS3<1),HS(2)
                               Stony Brook, LI
                               West End Ave between 72nd and 73rd Streets
                               30th Ave between 21st and 23rd, Queens
                               70th Street between 20th and 21st Ave, Brooklyn
                               79thi St between York and East End Ave, Manhattan
                               :92nd Street between Columbus and CFW, Manhattan
                               iColumbia'lviedicai'Center.w; 'i 68th St', Manhattan	
                              ''teaneck,NJ	
     14,15
     16
                               89th Street between Amsterdam and Columbia, Manhat
                               :8fJth Street between' Riverside and East End Ave , Ma
                               West End Ave between 1 0Stn and 1 6eih sireets
                               :Research Tiiangje Park, NC
     24,25,26
CMC(1),CMC(2)
vvoa;i),wGsi;2)"
USCCi),USC(2)''
                                     Chttenden Avenue, Manhattan
                                    ' Colombia Medical'Center , w! 'i 68th SI , Manhattan
                                    .Nassau County, LI
                                    :Federai Courthouse, White Plains. NY
12,13
28	
27	
FP(i'JiFP(2J"
                                     Northern Manhattan, Above 70th Street
                                     Nassau County
                                     NE Queens
                                    'long Beachi is land, NJ
                                     Federal Courthouse, Central islip, NY
Figure S: Map  of the origin of the samples  analyzed during this  study
                  (Reference Appendix D for sampling data).
                                              19

-------
Discussion
Slag wool appears to be an indicator for WTC dust and can be distinguished from background
dust at all three spiking levels for the USGS dust and at the 10% level of the 4 Albany Street
dust The 4 Albany Street dust is considered to be WTC impacted dust but as noted earlier, the 4
Albany dust likely had lower levels of slag wool due to the fact that it was an indoor dust that
was not sampled until three years after the WTC collapse

Levels of gypsum and elements of concrete have no discemable relationship to the level of WTC
dust There does not appear to be a distinguishable difference between levels of concrete and
gypsum in background dust and the samples spiked with WTC dust, despite USGS analysis of
WTC dust from 2001 (Meeker, 2005) showing elevated levels of these components.  This is
likely due to the fact that while these components may seem high in WTC dust, they are also
high m general background dust as they are common building materials

While method development (Appendix B and summarized in Section II above) work showed that
dusts from known impacted locations generally had slag wool levels above 100,000 fibers/gram,
several samples taken within this impacted zone and analyzed during method development
showed lower levels of slag wool  Two likely explanations can be offered for these results
First, as the data in Appendix B was acquired during method development, it must be viewed as
such, and second, multiple cleanings of the inhabited areas since September 11, 2001 may have
removed  residual WTC collapse contamination. The majority of these samples were taken in
fully inhabited buildings, from locations within the buildings that can be characterized as either
'accessible5 or 'infrequently accessed' areas.  These terms are described in the final draft EPA
sampling program, and they  denote areas that are accessed by people over the course of time,
such as counter tops or rugs (accessible) or underneath furniture (infrequently accessed)  For
this reason alone, it is encouraging that a substantial amount of the dust sampled in late 2004 and
beyond had high levels of slag wool

While there was ample evidence of higher levels of slag wool associated with the WTC dust and
lower levels associated with background, there is high variability in slag wool measurements
within and between labs  Estimates of within lab relative standard deviations based on analysis
of duplicate samples of the 4 Albany Street data are 55%, 24% and 14% for the 1%, 5% and 10%
dilution levels, respectively  Estimates of between lab relative standard deviations based on the
4 Albany Street data are 64%, 70% and 29% for the 1%, 5% and 10% dilution levels,
respectively (looking at results from analysis of the same spike level samples by multiple labs)
Causes of the high levels of variability may include1
             •  Procedures to homogenize the spiked samples did not result in complete
                mixing and distribution of fibers; they instead resulted in a 'reasonably'
                homogeneous sample given the large size variation of the dust components
             •  Components  of both non-impacted/background and WTC dusts will vary
                within a sample because of the inherent nature of the dust samples Thus, the
                samples received by the labs may vary in content
             •  Operator experience with the target components appeared to be an issue -
                post-study discussion indicated that labs representatives with less familiarity
                with slag wool expressed a belief that further guidance as to  its definition was
                needed
                                          20

-------
              •   The variability in the mass of dust used for the analysis, as the protocol allows
                 for a range, not a specific mass, to be used  This range is essential due to the
                 extreme differences in slag wool levels possible between background and
                 spiked samples

Finally, it is noted that Table I indicates two average values for background slag wool. These
values reflect the inclusion and exclusion of two samples (and their duplicates) collected in New
Jersey (NJ) and Long Island (LI) that were extremely high in slag wool fibers, likely due to their
insulation, fireproofing or ceiling tiles.  Similarly, it was earlier noted that samples taken from a
North Carolina building due also to slag wool used in ceiling tiles were not included in the
interpretative analyses  Based on these  results, it is likely that some false positive results will
occur in buildings with slag wool-based ceiling tiles, fireproofing or insulation

V. CONCLUSIONS
The interlaboratory results indicate that the better performing labs are capable of distinguishing
the difference between  1,5 and 10% 4 Albany Street dust Also, despite the high levels of within
sample and within lab variability, the method using slag wool appears to be sensitive enough to
distinguish 10% 4 Albany Street dust from background dust  Additional evaluation of the data
will be performed to further understand the variability  Measures will be taken (i e standards
will be sent regularly to each lab) during EPA's planned sampling program to evaluate the
accuracy and precision of the laboratones

In summary, the data developed in this study support the following findings

        1) Five of the eight laboratones  were able to reasonably measure the slag wool
       concentrations in background dust spiked with confirmed WTC dust

       2) High levels of variability in slag wool measurements, both within labs and between
       labs, were observed in the data  Despite this variability, the slag wool method appears to
       be sensitive enough to distinguish WTC dust from background dust at the 10% level
       (defined as, 4 Albany Street)

       3) The levels of gypsum and elements of concrete m the spiked samples were
       indistinguishable from the levels in the background samples  This observation suggests
       that, while these components may have been elevated in dust samples collected near
        September 2001, as found by USGS in their studies on WTC dust, they are also
       commonly found in the indoor environment and would not be useful  as WTC signature
       components

       4) Analysis of samples during method development generally showed slag wool levels m
        samples from impacted locations to be greater than slag wool levels in samples from
        background locations
                                            21

-------
VI. REFERENCES:
Lowers, H A , G.P Meeker, and I K Brownfield (2005a) Analysis of Background Residential
Dust for World Trade Center Signature Components Using Scanning Electron Microscopy and
X-ray Microanalysis  U S Geological Survey Open File Report 2005-1073
http //pubs usgs gov/of/2005/1073/

Lowers, H A , Meeker, G P , I K Brownfield. (2005b) World Trade Center Dust Particle Atlas
U S  Geological Survey Open-File Report 2005-1165 http //pubs uses gov/of/2005/1165/

Meeker, G P , A M Bern, H A Lowers, and I K. Brownfield. (2005) Determination of a
Diagnostic Signature for World Trade Center Dust using Scanning Electron Microscopy Point
Counting Techniques  URL http //pubs usgs gov/of/2005/1031/
U S. Geological Survey Open File Report 2005-1031

NYCDOHMH/ATSDR (2002) New York Department of Health and Mental Hygiene and
Agency for Toxic Substances and Disease Registry  Final Technical Report of the Public Health
Investigation To Assess Potential Exposures to Airborne and Settled Surface Dust in Residential
Areas of Lower Manhattan URL: http://www epa gov/wtc/panel/ATSDRFinal-report-
lowermanhattan-02.pdf Agency for Toxic Substances and Disease Registry, US Department of
Health  and Human Services, Atlanta, GA

R J Lee (2004)  Signature Assessment 130 Liberty Street Property Expert Report WTC Dust
Signature Prepared for Deutsche Bank  May, 2004. R.J Lee Group, Inc  350 Hochberg Road,
Monroeville, PA  15146

US EPA (2004) Mapping the Spatial Extent of Ground Dust and Debns from the Collapse of the
World Trade Center Buildings, DRAFT in peer review,  EPA/600/X-03/018, URL
http./Avww epa aov/wtc/panel/backdocs html
United  States Environmental Protection Agency, Office of Research and Development,
Washington, D.C , July 2004, 35pp.

US EPA (2002) Exposure and Human Health Evaluation of Airborne Pollution from the World
Trade Center Disaster, Draft in peer review, EPA/ URL'
bUp..//cfpub..ej3a.gw^                                United States Environmental
Protection Agency, Office of Research and Development, Washington, D C , October 2002
VII. CONTRIBUTORS TO THIS STUDY:

Principal Investigators:
Jacky Ann Rosati    U S EPA, ORD
David Friedman     U S EPA, ORD

Contributors (in alphabetical order):
Nancy Adams       U S EPA, ORD
Gma L Andrews    U S. EPA, ORD
Lara Autry          U S. EPA, ORD
                                        22

-------
Amy Bern          US EPA,NEIC
Isabella Brownfield   USGS
Ten Conner         U S EPA, ORD
EvanEnglund       US EPA, ORD
Pat Evangehsta      U S EPA, Region 2
Henry Kahn         US EPA, ORD
Matthew Lorber     U.S EPA, ORD
Heather Lowers      USGS
Mark Maddalom     U S EPA, Region 2
Lisa Matthews       US EPA, ORD
Greg Meeker        USGS
TimOppelt         US EPA, ORD
Joachim Pleil        U S EPA, ORD
Dennis Santella      US EPA, Region 2
Raj Smghvi         U S EPA, Region 2, ERT
Stanley Stephanson   U S EPA, Region 2
Shirley Wasson      U S. EPA, ORD
Steve Wilson        USGS

Supporting Contractors:
Prime Contractors
Alion Scientific
Lockheed-Martin
Versar
Subcontractors
RJ Lee Group, Inc
EMSL Analytical Inc.
MVA Scientific Consultants
Reservoir Environmental
MAS, Inc

Other Contributors:
John Holland  SEE   US EPA, ORD

VIII. ACKNOWLEDGEMENTS:
The U S  EPA would like to acknowledge the U.S General Services Administration (GSA),
NY/NJ Port Authority, the U S National Park Service, Deutsche Bank, and Columbia University
(K Crowley) for allowing samples to be collected at their facilities

In addition, the EPA would like to acknowledge the residents of NY and NJ who allowed us to
sample in their homes, and the maid services that collected vacuum cleaner bags for use in this
study. Finally, the EPA would like to thank the Sierra Club for helping to recruit sampling
locations
                                        23

-------
IX. APPENDICES

 APPENDIX A: QUALITY ASSURANCE PROJECT PLAN FOR THE WORLD TRADE
               CENTER (WTC) SCREENING METHOD STUDY
      (Due to formatting - this document will be provided under separate cover)
                                  24

-------
               APPENDIX B: DATA ACQUIRED BY EPA NERL DURING METHOD
                                         DEVELOPMENT
Samples Collected at Background Locations
Residential
West End Ave between 72nd and 73rd Streets, Manhattan

30th Avenue between 21st and 23rd St, Queens

E 79th Street between York and East End Ave, Manhattan
Chittenden Avenue, Manhattan

92nd Street between Columbus and CPW, Manhattan
80th Street between Riverside and West End Ave, Manhattan
Edison, NJ
Stony Brook, LI

70th Street between 20th and 21st Ave. Brooklyn

Teaneck NJ
Long Beach Island, NJ
West End Avenue between 105th and 106th Streets, Manhattan
Edison, NJ
88th Street between Amsterdam and Columbia, Manhattan
North East Queens (Maid Service)
 Nassau County, Long Island (Maid Service)

 Bus/ness
 Port Authority Bldg, Port of Newark, NJ
slag wool fibers/

  gram of dust
       2 53E+04
       5 47E+04
       2 80E+04
       2 26E+04
       4 93E+04
       1 53E+04
       2 87E+04
       2 42E+03
       1 46E+04
       0 OOE+00
       1 79E+04
       2 90E+04
       4 09E+04
       4 77E+04
       0 OOE+00
       0 OOE+00
       1 77E+04
       412E+03
       8 35E+03
       0 OOE+00
       5 74E+03
       0 OOE+00
       0 OOE+00
       0 OOE+00
       5 37E+03
       1 02E+04
       1 27E+04
       0 OOE+00
       1 63E+04
       6 43E+03
       0 OOE+00
       1 65E+04
       0 OOE+00
       0 OOE+00
       1 95E+04

       3 86E+04
       3 45E+04
       7 32E+04
       5 09E+04
        1 85E+04
       6 60E+04
Average of Duplicates
(slag wool fibers/gram
dust)
       2 20E+04
       4 43E+04
                                                 25

-------
Columbia Medical Center, W 168" St, Manhattan

Edison, NJ
Federal Courthouse, Quarropas St, White Plains
Federal Courthouse, (slip, Long Island
8 58E+04
0 OOE+00
1 33E+04
9 09E+04
9 28E+04
9 OOE+04
Samples Collected at Known Impacted Locations
Business
290 Broadway, Manhattan
Broadway between Maiden Lane and John Street, Manhattan
Deutsche Bank Bldg, 130 Liberty Street, Manhattan
Deutsche Bank Bldg, 4 Albany Street, Manhattan

USGS Composite Sample Collected Sept 2001

Samples Collected at Locations with Unknown Impact
Residential
John Street between Gold and Pearl, Manhattan
South End Avenue between Albany and Liberty, Manhattan
River Terrace, Manhattan
40th Street between Tunnel Exit St and 2nd Ave. Manhattan
Orange Street between Henry and Hicks, Brooklyn
24th Street between 8th and 9th Ave, Manhattan
Montague between Montague Terrace and Hicks Street, Manhattan
Houston and Mulberry Streets, Manhattan
Bus/ness
Port Authority Bldg, Columbia St, Brooklyn
6.92E+04
881E+04
1 64E+05
1 95E+05
8.35E+04
1 33E+05
2 79E+05
471E+06
5.77E+06
6 60E+06
1 18E+07
1 22E+07
1 13E+05
2 06E+05
214E+05
2 25E+05
2 28E+05
2 78E+05
6 36E+05
1 67E+06
1 34E+07
1 26E+04
917E+03
0 OOE+00
291E+03
1 11E+04
3 32E+03
5 03E+03
6 30E+03

2 06E+05
9 89E+04
1 30E+05
1 94E+05
1 12E+04
3 06E+05
                                                                                    1 20E+05
                                                                                    619E+06
                                                                                    2 30E+05
                                                 26

-------
Governor's Island                                                     5 07E+04
                                                                   5 75E+05
                                                                   8 79E+04
Vanck Street, Manhattan                                               9 57E+04

Samples Collected Outside of NY City
Business
Research Triangle Park, NC                                            5 OOE+04
                                                                   8 96E+04
                                                  27

-------
       APPENDIX C: DATA ACQUIRED BY EPA NERL POST-STUDY

Samples Collected at Background Locations
Residential
Composite -North East Queens (Maid Service)                           1 06E+04
                                                                1 49E+04    1 28E+04
Business
Port Authority - Port of Newark, N J                                     9 77E+03

Samples Collected at Impacted Locations
Business
Governor's Island                                                  1 93E+04
                                                                6 39E+05
                                                                1 21E+06
Port Authority Bldg, Columbia St, Brooklyn                               1 22E+05
                                               28

-------
   APPENDIX D: PROTOCOL USED FOR THE SCREENING METHOD STUDY

       Protocol for Preparation and Analysis of Residential and Office Space
   Dust by Polarized Light Microscopy and Scanning Electron Microscopy with
                    Energy Dispersive X-Ray Spectroscopy

                              June 27, 2005
                               Prepared by:
                   U.S. Environmental Protection Agency
    National Enforcement Investigations Center/ National Exposure Research
           Laboratory/National Homeland Security Research Center
                 Denver, CO and Research Triangle Park, NC
The use of trade names does not imply endorsement and are used for illustrative purposes only
                                    29

-------
                                         Contents


1.0   Purpose	31
2.0   Scope/Application	31
  2.1    Limitations of the Method and Future Considerations	31
3.0   Definitions	31
4.0   Summary of Method	31
5.0   Interferences	32
6.0   Safety	32
7.0   Apparatus and Materials	32
8.0   Reagents	33
9.0   Sample Storage	33
10.0  Quality Control	33
  10.1   Calibration	34
11.0  Procedure	34
  11.1   Weighing and Splitting	34
  11.2   Ashing	35
  11.3   Sieving	35
  11.4   Preparation of Sample for Polarized Light Microscopy	35
  11.4   Mounting Sample on SEM Sample Stubs	35
12.0  Analysis	37
  12.1   Analysis by Polarized Light Microscopy	37
  12.2   Analysis by SEM/EDS	35
     12.2.1   Screening for Slag Wool                                                35
     12.2.2   EDX Screening for Gypsum/Anhydrite                                    3 5
     12.2.3   X-Ray Mapping for Gypsum                                            36
     12.2.4   X-Ray Mapping for Ca-rich Particles                                     37
     12.2.5   Particle Analysis for Gypsum and Concrete                                37
13.0  Data Analysis and Calculations	38
14.0    References                                                                .39
15.0    Appendix                                                                  40.
                                         30

-------
1.0    Purpose

       This document describes sample preparation and analytical screening procedures for bulk samples of dust
       collected from residential and commercial office environments  These methods are collectively referred to
       as the protocol

2.0    Scope/Application

       The protocol describes polarized light microscopy (PLM) and scanning electron
       microscopy (SEM) with energy dispersive spectrometry (EDS) to screen bulk dust
       samples for mineral slag wool, particles consistent with concrete compositions, and
       gypsum  The analysis methods include operating parameters and particle identification
       criteria.

       2.1    Limitations of the Method and Future Considerations

               This protocol provides a means of analyzing for particles consistent with those found in dust
               present after the collapse of the World Trade Center (WTC) in New York City Components of
               WTC Dust have been documented and catalogued by the U S Geological Survey Denver
               Microbeam Facility and the images and charactenstics shall be used in identification of particles
               0)

               The x-ray mapping procedure in sections 1223 and 1224 and the calculations presented in
               section 130 only determine the maximum percentage of non-gypsum, calcium-nch particles,
               which may  include non-concrete matenals The particle analysis procedure presented in section
                12 2 5 is the preferred procedure for determining the percentages of gypsum and concrete particles
                in the sample

               The x-ray mapping and image analysis procedure relies heavily on the thresholds for backscattered
                electron images Binary (particles white and background black) backscattered electron images
                (BEI) should be used to reduce errors in setting thresholds in Photoshop

3.0    Definitions

        1   PLM - Polarized Light Microscopy
        2   SEM - Scanning Electron Microscope
        3   EDS - Energy Dispersive Spectrometry
        4   SEI - Secondary Electron Image
        5   BEI - Backscattered Electron Image
        6   Mineral Wool - lightweight vitreous fibrous material composed of rock wool and slag wool and used
            especially for heat and sound insulation
        7   Rock Wool - a man-made vitreous fiber (MMVF) component of mineral wool containing magnesium,
            aluminum, silicon, and calcium  Sodium and potassium may also be present Iron oxide is typically 3-
            12% by weight
        8   Slag Wool - a man-made vitreous fiber (MMVF) component of mineral wool containing magnesium,
            aluminum, silicon, and calcium  Sodium and potassium may also be present Iron oxide is typically
            less than 2% by weight
        9   HEPA - High-Efficiency-Particulate-Air Filter

 4.0    Summary of Method

         1.  Weigh sample  to nearest 0.0005 g
                                                 31

-------
       2  Split the sample, archive half and keep half for analysis
       3  Ash half of the sample for analysis
       4  Sieve the ashed sample to ISO um
       5.  Split the <150 um ashed portion  Archive three quarters of the sample  Keep one
          quarter for PLM and SEM/EDS analysis
       6  Weigh the quarter and place it in enough isopropanol to get a 10-20 mg per mL
          dilution  Apply an aliquot to a glass slide, let dry, and add 1 55 (or 1 605) refractive
          index oil. Analyze by PLM for mineral wool.
       7  Prepare a sample for SEM/EDS analysis using the same dilution prepared for PLM.
       8  Apply an aliquot of the sample to an aluminum sample stub with a carbon adhesive
          tab covered by a piece of polycarbonate filter (13-mm diameter or punched out of a
          larger filter to fit the size  of the stub)
       9  Identify fibers by EDS and  record the occurrence of fibers > 25 um in length at 100 x
          magnification to get a statistical representation of fiber compositions
       10. Prepare 10-fold dilution of the suspension from step 7 and apply an aliquot to a
          polycarbonate/adhesive tab substrate affixed to an aluminum sample stub
          Alternatively, a lighter loading can be prepared by filtering the diluted suspension
          through a 25-mm diameter, 0.4-um pore size, polycarbonate filter and affix this to a
          carbon adhesive tab affixed to an aluminum sample stub
       11 Collect x-ray maps of 10  fields at 500 x magnification for major elements, especially
          Ca, S, and Fe and use Adobe Photoshop or similar software to determine the area
          percent of gypsum and Ca-nch particles  Fe-nch particles may also be identified in
          this step
       12 Perform particle analysis  via computer-controlled SEM/EDX analysis

5.0    Interferences

       Interferences include possible contamination of samples by airborne dust or through improperly cleaned
       glassware and sieves  Interferences are minimized by performing all procedures involving dry dust in a
       clean room, cleaning countertops and glassware thoroughly before proceeding and placing particle-free
       wipes on all working surfaces  To  avoid cross-contamination, properly clean all glassware, sieves, and
       tools  between samples

6.0    Safety

       Respirable particles which may present a health hazard may exist in the sample  Bulk samples may release
       respirable particles during handling  All procedures involving dry dust samples will be performed under a
       negative flow High-Efficiency-Particulate-Air Filter (HEPA) hood  Samples handled outside of the HEPA
       hood  will be covered with aluminum  foil or placed in sealed glass jars

7.0    Apparatus and  Materials

       1  HEPA negative flow hood
       2  Forceps
       3  Kimwipes
       4  Stainless steel spatula
       S  Weighing paper
       6  Programmable furnace [not required for validation study]
       7  Ceramic crucibles with lids [not required for validation study]
                                             32

-------
       8   Analytical balance (accuracy to 0 OOOS g)
       9   Retsch ultrasonic sieve shaker (AS200 Basic), or similar [not required for validation study]
       10  Sample sieves, 3-inch diameter (recommended), 150-nm (100-mesh) opening, with lid and bottom pan
           similar [not required for validation study]
       11  SEM aluminum sample stubs
       12  Conductive carbon adhesive tabs
       13  Eppendorf pipette, 10-uL capacity
       14  Disposable pipette tips
       15  1 - 10 mL pipette
       16  Glass vials for sonicating dust in isopropanol suspension (holds 10-mL volume)
       17  Razor blade
       18  Ultrasonic bath
       19  50 mL glass beaker
       20  Polycarbonate filters (25-mm diameter, 0 4-nm pore size)
       21  Polycarbonate filters (13-mm diameter, 0 4-um pore size), or borer to cut larger filters to SEM stub
           size
       22  11 -mm diameter cork borer
       23  Milhpore filter apparatus for use with 25 mm filters
       24  125 mLNalgene bottles
       25  Hand-held vacuum pump
       26  High-vacuum carbon evaporator with rotating stage
       27  Glass   etri dishes with lids
       28  Adobe Photoshop Software, or similar
       29  Glass petrographic slides
       30  Glass cover slips
       31  Polarized light microscope for mineral identifications
       32  Scanning Electron Microscope with the following attributes
                a   Resolution  5 run (at 25 k V, WD=10 mm - system dependent) or better
                b   Accelerating Voltage 10to20kV
                c   Minimum magnification range 50x to 200,000x
                d   SEI (secondary electron image)
                e   6EI (backscattered electron image)
                f   Energy dispersive x-ray detector and analyzer for EDS analysis
                g   Ability to collect x-ray maps or particle analysis software (preferably both)


8.0    Reagents

        1    Isopropanol, reagent grade [CAS No 67-63-0]
        2   1 55 or 1 605 Refractive Index Oil

9.0    Sample Storage

        Dust samples will be stored in an air-tight container, such as a sealed glass jar  Samples placed in reagents
        will be labeled appropriately and stored according to laboratory safety standards  Samples prepared for
        analyses will be stored in a protective container, such as a plastic case or covered  etn dish, to prevent
        contamination

10.0  Quality Control

        Quality control is implemented by thoroughly cleaning glassware and spatulas, keeping working surfaces
        clean, and preventing cross contamination.  During ashing, particles may be suspended if slow heating is
        not achieved Following the ashing program as outlined will minimize flashing, which can cause particles
        to become airborne  Covered crucibles will be used to prevent contamination caused by flashing  Used
        Eppendorf pipette tips and weighing papers will be discarded and new tips and papers will be used for each
                                                   33

-------
       sample

       Duplicate samples shall be prepared to determine the precision of the analysis In addition, sample blanks
       shall be prepared  These blanks are checks for cross contamination during handling of the samples Blanks
       shall be prepared at the same time and in the same manner as samples

       10.1    Calibration

               Calibration of the EDS system must be completed at least once at the beginning and again at the
               end of each analytical session  Backscattered electron image (BEI) calibration should be
               performed at the beginning of the session and anytime the backscattered image brightness and/or
               contrast is adjusted

               EDS calibration for both qualitative and quantitative (not required by this method but could be
               useful for identification of particle type) analysis is accomplished by the analysis of a polished
               carbon-coated reference standard The recommended material is USGS BIR1-G basalt glass
               mounted in epoxy in a brass tube, polished, and carbon coated using a carbon evaporator (2, 3)

               The calibration reference matenal should be analyzed at the same operating conditions to be used
               for the analysis including beam current, accelerating voltage, working distance, detector dead
               tune, and sample tilt (= 0°)  For BIR1 -G the analysis should be performed with a beam size of 10-
               20 um or equivalent area raster  All calibration spectra will be saved with the corresponding data
               set The calibration data will be used for inter- as well as intra-laboratory comparisons This
               calibration is in addition to, and not a substitute for the normal EDS calibration recommended by
               the EDS manufacturer which will be performed at regular intervals as specified by the EDS
               manufacturer

               Backscattered electron detector calibration can be performed on the same BIR1-G material by
               adjusting the detector brightness and contrast to achieve the following conditions  The epoxy  on
               the BIR1-G reference material will be at 0 in a 256 grayscale image and the brass mounting tube
               will be at 256 The BIR1-G basalt glass should fall at approximately 130-140 gray scale units

11.0  Procedure


       11.1    Weighing and Splitting

               Weighing and splitting should be performed  under a negative flow  HEPA hood
               If the fan speed is set too high, loss of particles may occur.  The fan speed may
               need to be adjusted to prevent the loss of fine particles

               Obtain  an analytical balance  with an  accuracy of 0 0005 g and preweigh a clean
               piece of weighing paper  Transfer  the  dust from the  sample vial  to the weighing
               paper and determine the weight of the dust   Split the  sample with a clean razor
               blade using the  cone-and-quarter method  If there are large clumps  of organic
               fibers, such as hair or lint, temporarily  remove the hair with a pair of forceps and
               tap the forceps lightly with another  tool over a piece of weighing paper to remove
               fine particles.  Center the fine fraction on the paper and split the sample into four
               equal parts using a razor blade  Collect opposite corners  (!/2 of the sample) for
               analysis and archive the other half  Quarter  the  larger organic fiber bundles the
               same way, keeping half to  proceed  to  the ashing  step  and  half for archival
               purposes
                                               34

-------
       Place the two quarters for ashing into a preweighed crucible  Weigh the split and
       record the results

11.2   Ashing

       Place the ceramic crucibles containing the samples into a furnace

       The furnace program should proceed as follows
       1   Increase temperature by 1 "C/minute until sample reaches 250 °C
       2   Hold temperature at 250 °C for 4 hours
       3   Increase temperature by 1 °C/minute until sample reaches 480 °C
       4   Hold temperature at 480 °C (sufficient for decomposing orgamcs) for 8 hours  Do not exceed
           500 °C
       5   Shut off furnace
       6   Allow sample to cool before removing from furnace
       7   Weigh the ashed sample to the nearest 0 0005 g and record the result

11.3   Sieving

       Sieve the sample through a 150-um sieve using a Retsch ultrasonic sieve shaker,
       or similar  Three-inch diameter sieves are recommended to minimize sample loss
       from particles being  trapped in the sieve  The ultrasonic shaker will be operated
       at 20-minute intervals at the following settings. 20,  40, 60, 70, 80, then  back
       down to 50 and 20. This will provide amplitudes ranging from 0 to 1 5 mm

       Transfer the large and small  fractions to clean pieces of weighing paper and
       weigh to the nearest 0 0005 g   Archive the fraction greater than 150-um


11.4   Preparation of Sample for Polarized Light Microscopy

       Split the less than 150-um sample fraction using the cone and quarter method Collect one comer
       for analysis and archive the other three quarters  Weigh the quarter  split to the nearest 0 0005 g
       and place it into a glass vial  Make a suspension of 10-20 mg dust per mL of isopropanol  The
       amount of isopropanol needed will vary depending on the amount of dust, the target dilution is 10-
       20 mg per mL

       Cut  an Eppendorf pipette  tip with a  razor blade to  increase the  opening  to
       approximately  1 mm

       Place the suspension in an ultrasonic bath for one minute, then remove the suspension from the
       ultrasonic bath and shake it gently to suspend all particles  Collect a 10-uL aliquot of the mixture
       using an Eppendorf pipette with the modified tip and transfer to a glass slide  Prepare 4 such
        slides  Allow them to dry, then add a drop of 1 55 (or 1 605) refractive index oil

 11.5   Preparation of Sample  for SEM Analysis

        Prepare  the  SEM  substrate  on  aluminum  stubs  using  04-um  pore  size
        polycarbonate filters, carbon  adhesive tabs.  Using an  11 mm filter punch and
        placing  the filter between  two filter separators, punch a circle  the size of the
                                        35

-------
              carbon tab into the filter  Place carbon adhesive tab affixed to an aluminum stub
              on the dull side of the 11-mm polycarbonate filter such that the shiny side of the
              filter exposed  If available, a 13-mm diameter polycarbonate filter may be used in
              place of the punched out 11-mm filter

              Collect a 10-uL aliquot  of the mixture from the PLM sample preparation using
              the  Eppendorf pipette  with  the  modified  tip  and  transfer  to a  prepared
              polycarbonate/adhesive tab substrate  This will yield a loading on a 12-mm SEM
              stub of about  100-200 ug, which  is a moderately heavy loading.  Adjust the
              number of aliquots as needed to obtain the target loading

              Prepare a 10-fold  dilution of the above suspension to get a suspension of 1-2 mg
              dust per mL of isopropanol  Sonicate the suspension in an ultrasonic bath for one
              minutes.  Remove the suspension  and gently  shake it to suspend all  particles.
              Wait one minute to allow the coarse particles to settle  Collect a 10-^iL aliquot of
              the suspended mixture using an Eppendorf pipette with the modified tip and
              transfer to a prepared polycarbonate/adhesive tab substrate   This will  yield a
              loading on  a 12-mm  SEM stub of about 10-20  ug, which  is  a  light loading
              Adjust the number of aliquots as needed to obtain the target loading.

              Alternatively, prepare a lightly loaded sample using  the filtration method  as
              follows  Use a Millipore filter apparatus for use with 25-mm filters for filtration.
              Place a few drops of isopropanol on the fritted glass surface and place the 25-mm
              polycarbonate filter (0 4-um pore size) on the isopropanol. Attach  the top of the
              apparatus and add a few milliliters of isopropanol to the filter so that no part of it
              is exposed  to  air   Sonicate the suspension (diluted as described in previous
              paragraph) in an  ultrasonic bath for one minute.  Remove the suspension and
              gently shake it to suspend all particles   Wait one minute to allow the coarse
              particles to settle.  Collect 1  mL of the suspended mixture using  a pipette and
              filter it through the polycarbonate filter   Actual amounts for filtration will  vary
              based on sample loading. The goal  is to have a loading on a 12-mm SEM stub of
              about  10-20 ug, or about 5-10 percent area coverage, which is a  light loading
              Adjust the volume of the aliquot to filter as needed to obtain the target loading

              Place the filter  on a carbon adhesive tab on a standard SEM aluminum mount
              The filter needs to be completely flat on the SEM stub.  This can be achieved by
              forming the wet filter into a gentle U-shape using forceps and the side of the
              forefinger, then placing the bottom curve of the filter onto the center of the carbon
              adhesive tab and slowly releasing the sides so they lay flat. Tnm the edges of the
              filter using a razor blade

              After drying, coat the samples on the polycarbonate or polycarbonate/adhesive tab substrates with
              carbon using a carbon evaporator with a rotating stage  Transfer the stubs to the SEM in a clean,
              covered container
12.0   Analysis
                                           36

-------
12.1    Analysis by Polarized Light Microscopy

        Polarized light microscopy will be conducted using the general techniques outlined in EPA
        600/R93/116 (4)  For this procedure, four slides (prepared as descnbed in section 11 4) will be
        analyzed  The fraction of fibers with refractive index greater than 1 55 (or 1 605) will contain
        mineral wool, which includes both slag wool and rock wool, and possibly some E-type glass and
        ceramic fibers The fraction of fibers with refractive index less than 1  55 (or 1 605) will contain
        primarily soda-lime glass fibers  For the validation study, numbers of fibers greater than and less
        than 1 55 (1 605) refractive index will be counted  Dispersion staining and becke line techniques
        may be used  Fiber point counting will be performed at 100 x magnification

        If more than 20 mineral wool fibers are found, continue counting and recording all of the fibers
        above and below the index oil refractive index  Report both raw fiber counts per refractive index
        category and number of fibers from  each category per gram of sample Continue on to step  1221
        to determine the ratio of slag wool to other fibers with refractive index greater than 1 55 (or 1 605)
        using EDS as described below

        If less than 20 mineral wool fibers are found on each slide, count the number of slag wool fibers
        using SEM/EDS and report as number of fibers per gram of sample

12.2   Analysis by SEM/EDS


        12.2.1  Screening for Slag Wool

                 Operating conditions for the JEOL 6460-LV SEM are 15 kV, 0 5-5-nA beam current, 10-
                 mm working distance (system dependent), and zero degree tilt

                 Place the more concentrated sample deposited directly on the polycarbonate/adhesive tab
                 substrate into the SEM Use the backscattered electron mode at lOOx magnification to
                 quickly distinguish carbon fibers from inorganic fibers (carbon fibers may be visible, but
                 not as bright in a BEI)  Identify all inorganic fibers over 25  um in length (smaller fibers
                 cannot be reliably detected at the 1 OOx operating magnification)  When an inorganic fiber
                 is found, identify the composition of the particle by EDS Slag wool is the primary fiber
                 of interest  Record all inorganic fiber results as number of fibers for each fiber type

                 For the samples with high fiber loading, as determined by PLM as descnbed in section
                 121, count fibers per type  until a statistical representation of the ratios of fiber
                 compositions in the sample is achieved  Report the ratio (by fiber number) of slag wool
                 fibers to total MMVF fibers corresponding to the high RI Use this ratio to correct the
                 total number for high RI fibers counted by PLM to number of slag wool fibers present

                 For the samples with low fiber loading, as determined by PLM as descnbed in section
                 121, scan the entire stub to determine the number of fibers per type  Report the slag
                 wool fiber results as the number of slag wool fibers/gram of sample

        12.2.2  EDS Screening for Gypsum/Anhydrite

                 Place the more concentrated sample deposited directly on the polycarbonate/adhesive tab
                 substrate in the SEM  Choose a random field at lOOx magnification and perform an EDS
                 analysis on the entire field  Look for the  presence of sulfur in this field  If sulfur is
                 present, continue to Section 12 2 3 or 12 2 5 for analysis of gypsum and concrete by
                 mapping or particle analysis  If it is not present, repeat the analysis on another random
                 field If sulfur is still not present, mark the sample as non-detect (ND) for sulfur
                                            37

-------
12.2.3  X-Ray Mapping for Gypsum

        Place a more dilute sample, deposited directly on the polycarbonate/adhesive tab
        substrate or prepared by filtration, in the SEM Collect binary backscattered electron
        images (particles white and background black, shadow off) and secondary electron
        images for 10 non-overlapping, random fields at 500 x magnification  Collect x-ray
        maps for Na, Mg, Al, Si, S, Ca, and Fe at each of these fields  Fields containing MMVF
        will not be used for this analysis Operating parameters for the SEM are the same as
        those for analyzing slag wool  Acquisition parameters for x-ray mapping using the
        NORAN System Six Software are time constant 14 (mapping mode, 11333 cps), 10-20 %
        deadtime, 256 x 256 image resolution, 20 second frame time, and 100 frames collected
        (about 40 minutes total acquisition time)  Secondary electron images will be used for
        reference only  Save all of the maps and electron images in TIFF format

        Open the backscattered electron image and the Ca and S x-ray maps in Adobe Photoshop
        Make sure that all of the element maps are the same size and resolution by choosing
        Image Size from the Image Menu and changing the pixel size or the resolution as needed
        The presence of gypsum can be determined by overlapping the Ca and S maps

        Perform the following functions in Adobe PhotoShop  (A macro is in development to
        perform the following functions to decrease user time and human errors in adjusting the
        threshold )

        1   Convert each of the three images to grayscale (Image —> Mode —» Grayscale)
        2   Perform an auto contrast and brightness on each image and map to increase the scale
            of colors (Image —> Adjustments —» Auto Levels)
        3   Threshold each element map, Ca and S (do not analyze the backscattered electron
            image at this time), by going to the Image Menu and choosing Adjustments —>
            Threshold  Adjust the threshold to 128 The background will be black and the
            particles white
        4   Invert the image (Image—* Adjustments —>Invert) to make the background white and
            the particles black
        5   Copy the S map and paste it over the Ca map in a separate layer in the file and
            change the opacity (located in the Layers window) to 50 % for the S map layer  The
            black areas are gypsum/anhydnte
        6   Display a histogram of the image in expanded mode by selecting the Histogram tab
            on the Navigator Window (or under the Image Menu in some versions of
            Photoshop)  Place the cursor over the line for the black area and record the
            percentile for the black area This is the percentage of particles containing Ca and S
            in the entire field

        NOTE If a binary backscattered electron image is obtained during data collection, then
        steps 7-11 may be deleted The Invert function will, however, need to be applied to make
        the particles black and the background white before continuing to step 12

        7   Begin analysis of the backscattered electron image  Select the particles by going to
            the Select Menu and choosing Color Range Go to the selection pulldown menu and
            choose Highlights
        8   Fill the selection with black by going to the Edit Menu —» Fill and choosing black
            from the color pulldown menu
        9   Select the inverse areas by going to the Select Menu and selecting Inverse
        10  Fill the selection with white by going to the Edit Menu —> Fill and choosing white
            from the color pull down menu
        11  Deselect the area by clicking on the image
        12  Perform the Threshold and Histogram functions for the backscattered electron image
                                   38

-------
            as outlined in 3 and 6  Record the histogram result for the backscattered electron
            image

        Determine the area percent of gypsum by performing the calculations in Section 13 0

12.2.4  X-Ray Mapping for Ca-Rich Particles

        Analysis of components of concrete will be performed on the same fields as the
        gypsum/anhydnte analysis At this time, only a method for the determination of the area
        percent of Ca-nch particles is presented  See Section 2 1 for discussion

        Perform the following steps on the Ca x-ray map Tiff file in Adobe Photoshop

        1   Convert the Ca x-ray map to grayscale (Image -» Mode —> Grayscale)
        2   Perform an auto contrast and brightness on the map to increase the scale of colors
            (Image —» Adjustments —» Auto Levels)
        3   Threshold the Ca map by going to the  Image Menu and choosing Adjustments —>
            Threshold Adjust the threshold to 128  The background will be black and the
            particles white
        4.  Invert the image (Image—» Adjustments —>Invert) to make the background white and
            the particles black
        5   Display a histogram of the image Place the cursor over the line for the black area
            and record the percent lie for the black area  This is the area percent coverage of
            particles containing Ca in the entire field

        Determine the maximum area percent coverage of non-gypsum, Ca-nch particles by
        performing the calculation in Section 13 0

12.2.5 Particle Analysis for Identification of Gypsum and Concrete.

        Place the more dilute sample, deposited directly on the polycarbonate/adhesive tab
        substrate or prepared by filtration, in the SEM Particle analysis will be used to identify
        gypsum and concrete particles

        Perform particle analysis at 500 x magnification  All other operating parameters for the
        SEM are the same as those used to analyze for slag wool (Section 1221) A binary
        backscattered electron image should be used in particle analysis mode   Particle analysis
        parameters should be set to analyze all particles in the field greater than 0 5 um and to
        separate touching particles  For particles greater than 5 um, scan the entire particle, spot
        analysis is adequate for smaller particles The x-ray spectrum and counts for all particles,
        and an image of particles > 20 um long, will be recorded and saved  Other particle
        parameters to be reported will include the maximum, minimum, and average diameters,
        the aspect ratio, and area of each particle

        It will  be necessary to review data collected by automated software to ensure data
        integrity  An Excel spreadsheet, in conjunction with images and x-ray data, may be used
        for this purpose  Particles should be sorted into one of three categories Ca-S (gypsum),
        Ca-nch, and Other  Aid in identification of particles may by facilitated by referencing
        the U S Geological Survey's WTC Dust Particle Atlas (1)  A particle classification
        protocol will be developed based on the data from the validation study

        The number of particles analyzed will be determined using the results of the validation
        study  For the study, the area percent of each component should be within 10% relative
        error or better Typically, data for 1000 - 1200 particles should be acquired
                                    39

-------
                Results for particle analysis will be recorded as area percent gypsum and area percent concrete
                particles for each field and average area percent for the each component in the sample

13.0   Data Analysis and Calculations

                Table 60       To determine the concentration of slag wool in fibers/gram, perform the
                        following calculations

                Determine the number of fibers with RI > 1 55 (or 1 605)

                # fibers identified - mg of sample on slide * 1000 = fibers/gram on slide

                Determine the percentage of fibers with the composition of slag wool with RI > 1 55 (or 1 605)

                Fibers/gram on slide * # fibers identified as slag wool	= fibers slag wool/gram on slide
                Total number of fibers identified by EDS with RI > 1 55 (or 1 605)

                Back calculate to the number of fibers per gram of the original sample

                Fibers slag wool/g on slide x  g after sieving * g sample after ashing =  Total f/g of sample
                g before sieving  x g sample before ashing

                Table 61       To determine the area percent of gypsum/anhydrite from the x-ray mapping
                        procedure, perform the following calculations

                Determine the area percent of gypsum/anhydrite in each field of view

                % of black area in Ca-S map overlay x IQO  = area % gypsum
                % of black area in BSE image

                Calculate the average percentage of gypsum/anhydnte for the sample

                (area % gypsuml^ + (area % gypsum^ +   = Avg area % gypsum
                        number of fields

                Table 62       To determine the maximum area percentage of Ca-nch particles, which includes
                        concrete particles, from the x-ray mapping procedure, perform the following calculations

                Determine the area percent of non-gypsum Ca-nch particles in each field of view

                (% black area Ca map") - (% black area Ca-S mapl = % non-gypsum Ca-nch particles
                        % black area on BSE image

                Calculate the average percentage of non-gypsum Ca-nch particles  for the sample

                (area % Ca-nch particlesV + (area % Ca-nch particles')?; +   = Avg area % Ca-nch particles
                                number of fields

                Table 63       Calculate the area percent for gypsum and concrete by summing the areas of
                        each particle in for each particle type and dividing by the  total area analyzed

                area gypsum 1 + area gypsum 2 +    x 100 = area percent gypsum (do likewise for concrete)
                        total area analyzed

                Rules for concrete and gypsum classification are currently being developed
                                                  40

-------
14.0   References
             Lowers, Heather A , Meeker, Gregory P, Brownfield, Isabelle K, 2005  World Trade Center
             Dust Particle Atlas U S Geological Survey Open-File Report 2005-1165  On the web at
             litto //pub" uses gov/of/2QOS/l 165/
             Meeker, G P, Taggart, J E, and Wilson, S A , 1998  A Basalt Glass Standard for Multiple
             Microanalytical Techniques  Proceedings Microscopy and Microanalysis 1998  Microscopy
             Society of America
             A polished and carbon coated calibration reference sample of BIR1-G may be obtained by
             contacting Stephen Wilson, U S Geological Survery, MS 973, Denver Federal Center, Denver,
             CO, 80225, svvilson^usgs gov
             Perkins, R L and Harvey, B W, 1993, TEST METHOD  Method for the Determination of
             Asbestos in Bulk Building Materials, EPA/600/R-93/116
                                                41

-------
      15.0   Appendix:
DATA SHEETS
Determination of Slag Wool Fibers in Dust- PLM with Dispersion Staining

Sample ID                                                Project
                                                       Analyst
Circle One    Original      Duplicate     Triplicate                Date
General Sample Appearance.
Homogeneous?
Structure #
                   Rl Fluid
              _155_
                                    Dispersion Staining
                                                             Becke Line
                                                                                            Fiber
                                                                                           non-MW
                                                                                                      chrvsotlle
                                                                                                               Comments
                                                        42

-------
SEM Sheet
                                                                     Reference ASTM - D5755-03
Report Number1 _
Sample Number:.
File Name.	
Sample Description:
Preparation Date:
Analysis Date:
Computer Entry Date:
Sample weiaht:
Dilution Volume:
Volume Aliquot:
Magnification:
Bv:
Bv:
Bv:







qrams
ml
uL
X
   Structure #
                  Field #
Fiber Type   Length (Microns) Width (Microns)
Image
                                                                                        EDS
                                                 43

-------
TJGbTET
                                                           Ota

-------
APPENDIX E:  REPORT FROM THE U.S. EPA CONTRACTOR ON THE SCREENING
                                     METHOD STUDY
                                       Versar
                                       6850 Versar Center
                                     Springfield, VA 22151

Ms  Jacky Rosati
US Environmental Protection Agency
E-305-03 109 TW Alexander Drive
Research Triangle Park, NC 27711                      July 21,2005
Dear Ms Rosati

       Attached is a preliminary report based  on analytical data thus far received, for dust
samples collected primarily m the New York City area  Most of the samples were taken in areas
that, it is believed, were not  affected by paniculate matter generated during the World Trade
Center (WTC) collapse (i e , background samples)  Some of the samples were  spiked with one
or the other of two dusts that are believed to have  originated from the WTC collapse.  The
analytical  protocol was developed by the government, specifically for this project, and was
modified as the project developed.  The purpose of the testing was  to determine if the  spiked
background dusts could be distinguished from those samples that were not spiked

       Three parameters were measured to make this  determination- (1) slag wool fiber content,
(2) calcium-rich particle content, and (3) gypsum particle content

       The analytical data indicate that

    •   With respect to calcium-nch particles and gypsum particles, spiked samples cannot readily be distinguished
        from background samples
    •    With respect to slag wool content in the samples spiked with the first of the two WTC dusts, spikes at the
        10% level may be statistically identifiable as WTC-contammation, although spikes at or below the 5% level
        are probably not identifiable
    •    With respect to slag wool content, samples spiked with 5% and 10% of the second of the two WTC dusts
        are easily identifiable as WTC-contammated  Even at the 1% spike level, samples may be statistically
        identifiable

        The attached preliminary report will explain the above conclusions in more detail  However, it must be
    noted that all of the analytical data from the eight laboratories that performed the analysis has not yet been
    received  Nevertheless, it is believed that the above conclusions will not likely change once those additional
    data are incorporated

                                                  Sincerely,
                                                   Stephen M Schwartz, P E, Q E P
                                                   Project Manager
                                          45

-------
                                      Versar

                                 Preliminary Report
                                            of
                Analysis of New York City Area Dust Samples
    Purpose:
    The objective of this study is to determine if New York City area dusts that are contaminated with
varying levels of dusts known to originate from the collapse of the World Trade Center (WTC) can be
distinguished from background dusts that are believed not to be contaminated with WTC dusts

    Project Summary:

    In the initial portion of the testing, 10 dust samples from New York City areas that are believed
not to be contaminated with dusts originating from the collapse of the WTC were used  These are
referred to as the first set of background samples  An additional background dust sample was spiked at
1, 5, and 10 percent levels (by weight) with dust believed to  have originated from the WTC collapse
An additional background sample was spiked at  1, 5, and 10 percent levels with a second dust sample
that is believed to have originated  from the WTC collapse  Therefore, a set of  16 samples was
generated

    •   10 different background dusts
    •   3 samples, each consisting of one  background dust  sample spiked with one source of WTC
        dust at 1, 5, and 10% levels
    •   3 samples, each consisting  of one background dust sample spiked with a second source of
        WTC dust at 1, 5, and 10% levels

    Initially, 32 samples were sent to each of eight analytical laboratories (three U S government, and
five private)  The 32 samples consisted of two identical sets (i e,  duplicates) of the 16  samples
discussed above  The pnvate  laboratones did not know that there were duplicate samples   Further,
they did not know which, if any, of the samples contained WTC  spikes

    Subsequently, a second set of 28 different background samples was analyzed to obtain a better
understanding of the variability of background dusts  These  28 samples were sent to only one of the
five private laboratones

    It was ultimately  agreed that each of the laboratones would  perform  the following
    three Scanning Electron Microscopy-based (SEM) analyses on each  of the
    samples they received (see  Methodology and Data Analysis section)

    •   Slag  wool fiber content (in  number of fibers per gram of dust)  Slag wool was a significant
        component of the WTC insulation material
    •   Calcium-rich particle content (in area percent concentration in the SEM field) Such particles
        are assumed to be indicative of cement/concrete-like  particles
    •   Gypsum particle content (in area percent concentration in the  SEM field)  Such particles are
        assumed to be indicative of "dry wall" (i e , gypsum-containing wall board)

    Conclusions:

-------
   A number of conclusions can be drawn from the analytical results thus far
   obtained.  It is not expected that data that are subsequently received will
   substantially change these conclusions. It must be noted that there are several
   caveats that affect the quality of the data  Those are discussed later in this report

   2   With respect to calcium-rich particles and gypsum particles, spiked samples
       cannot readily be distinguished from background samples

Tables 1 and 2 present the analytical data thus far available for calcium-rich and
       gypsum content respectively  Analysis was performed using SEM and x-ray
       mapping (XRM) techniques The shaded areas represent the samples spiked
       with 1, 5, and 10% WTC dust. The others areas are background samples
       Sample designations followed by "(1)" and "(2)" are duplicate samples
       (Samples received by the laboratories had random identification numbers, so
       that the laboratories did not know if any samples were duplicates, nor did they
       know if any samples contained WTC dust) In addition, Table 3 is the
       analysis of a subsequent 28 background samples, analyzed by only laboratory
       "B"  Analysis of calcium-rich and gypsum particles for this sample set is
       shown on Table 3

The average of all background samples (including the second set of 28 samples) for
       calcium-rich particles is 22 3 area percent, with a high value of 66 5% and a
       low value of 4 2%  The average for the spiked samples is 20 7%, with the
       highest value being 25 9% The 1, 5, and 10% spiked samples do not show
       any trend with respect to calcium-rich particle content (i e., they do not show
       any increase as the spike level increases).

The average of all background samples (including the second set of 28 samples) for
       gypsum particles is 11 7 area percent, with a high value of 56 5% and a low
       value of 0 1 %  The average for the spiked samples is 9 3%, with the highest
       value being 32 8%  The 1,5, and 10% spiked samples do not show any trend
       with respect to gypsum particle content

    3.  With respect to slag wool content in the samples spiked with the  first of the
       two WTC dusts, spikes  at the 10% level may be statistically identifiable as
       WTC-contammation, although spikes at  or below the 5%  level  are probably
       not identifiable

Table 4 presents all the analytical data thus far available for SEM slag wool fiber
       analysis (as the number of slag wool fibers per gram of dust). The shaded
       areas represent samples that are spiked at the 1, 5, and 10% levels with WTC
       dust  Table 3 also presents additional slag wool fiber background-only sample
       data (next to last column)  It can be seen from Figure 1 that for those spiked
       samples designated as "DB" that at the 5% spike level, the slag wool
       concentrations probably do not exceed one standard deviation above the
       average slag wool background concentration (including the Table 3
                                     47

-------
          background data) However, at the 10% spike level, the slag wool
          concentration typically exceeds one standard deviation (see Figure 2), but
          never exceeds two standard deviations above the average background sample
          concentration The average background concentration is about 27,400 fibers
          per gram.  The standard deviation is about 40,100 fibers per gram '

   It should be noted that there is a trend showing a clear increase in slag wool fiber
          concentration from the 1% to the 10% spike level (see "DB" sample shaded
          area on Table 4)  However, the numerical values of those concentrations, as
          noted above, are still less than two standard deviations above the average
          concentration.

       4.  With respect to  slag wool content, samples spiked with 5% and 10% of the
          second of the two WTC dusts are easily identifiable as WTC-contaminated
          Even at the 1% spike level, samples may be statistically identifiable.

   The slag wool content data for the samples spiked with the WTC dust shown in Table
          4 as "USGS" are easily identifiable  As can be seen in Figures 4 and 5,
          samples spiked with the USGS WTC dust at the 5 and 10% levels are
          essentially all more  than two standard deviations above the average
          background sample concentration.  (Average plus two standard deviations
          would be about 108,000 fibers per gram.2)  At the 1% spike level though,
          WTC dust is more difficult to identify because the slag wool concentrations
          are mostly between  one and two standard deviations above the average
          background sample (see Figure 3)

       5   With respect to slag wool  content,  clearly, there is a large difference between
          the two WTC dust spikes used  In the "DB"-spiked samples, as noted above,
          it is expected to be more  difficult to determine a significant slag wool fiber
          concentration  difference from background   The "USGS"-spiked samples
          clearly had significantly more slag wool fiber content than the "DB"  samples

       6   Examining Tables 1, 2, and 4 and the Figures, it can  be seen that the analyses
          for the duplicate samples rarely replicate one another  However, the variation
          between duplicate sample values (i e,  intralab) is about half of the variation
          between individual laboratory values (mterlab).3
1 Background concentration data for this analysis excluded several samples that were known to have high
slag wool content, specifically the Cl-RTP samples (see Table 4), and samples C2,3,4,5,6 (see Table 3)
2 Ibid
3 For slag wool fiber analysis, the average difference between the  analyses of duplicates (i e, intralab
differences) is about 50% of one standard deviation of the between-laboratones analyses (i e, inUrlab
differences) For both calcium-rich and gypsum particle analysis the average intralab difference is 20% of
the interlab difference
                                        48

-------
       Methodology and Data Analysis:

       The analytical protocol was developed specifically for this project by  one of the
   government laboratories, and modified by all laboratory participants at a meeting held
   for that purpose  All laboratory participants held weekly conference calls  as the
   analytical  program  was proceeding  to  discuss  general  issues with the protocol
   Additional modifications were made to the protocol based on those conference calls

       The original protocol included analysis by Polarized Light Microscopy (PLM), so
   data are also available for PLM analysis  The PLM analyses were curtailed because it
   became obvious that PLM could not adequately differentiate between fiber types
   Further,  total  fiber  concentrations were also determined, both by PLM and SEM
   methods, but those data are not presented in this report

       Caveats:

       There are a few factors that may contribute to data uncertainty  Nevertheless, it is
unlikely that these factors will alter the above major conclusions  Some of these factors
are as follows:

       1      As noted earlier, not all of the analytical data have been received
       2      Dust  samples were  collected  by  several methods   Evaluation  of the
              sampling methodology was not part of the study
       3      To  determine fiber concentration, fibers  were  counted  using an SEM
              Different laboratories diluted samples to different levels before counting,
              introducing some variability of results
       4      Laboratory equipment capabilities and personnel skills varied
                                         49

-------
TABLE 1: SEM X-Ray Mapping - Calcium-Rich Area Percent
Sample
Designations
AP5(1)
AP5(2)
CMC(1)
CMC(2)
HS3(1)
HS3(2)
WGS(1)
WGS(2)
MW(1)
MW(2)
DB1%(1)
DB1%(2)
DB5%{1)
DB5%(2>
DB10%(1)
DB10%(2)
C1-RTP(1)
C1-RTP(2)
USGS1%{1)
USGS1%(2)
USGS5%{1)
USGS5%(2)
USGS10%(1)
USGS10%(2)
USC(1)
USC(2)
FP(1)
Laboratory Letter Codes
A



























B
234
224
277
341
103
178
228
199
122
120
18.2
131
13,4
20,5
14.4
126
132
162
16.5
210
148
146
17.0
179
123
94
130
C










156

230


18.9



26.6

26.4





D
204
221
21 8
21 5
141
133
132
163
142
109
14.0
141
161
129
15,2
149
11 3
11 9
174
111
142
162
15.8
121
11 1
95
11 6
E
144
168
67
204
64
148
139
57
126
83
183
15.4
108
4,1
8.6
10.8
75
56
124
57
11 9
10 .7
109
98
195
66
59
F
11 6
98
79
101
153
65
77
74
76
57
8,9
9.4
9.0
75
8.0
81
61
42
7.6
74
6.7
83
8.9
83
54
71
62
G
307
396
551
389
291
440
584

493

55,9

400

50.7

571







435

424
H
458
489
604
552
631
497
534
523
463
498
502
522
490
39,6
406
48.8
665
61 0
432
419
536
51.8
400
45.1
461
403
709
                                                                          Location Key:

                                                                      Chittenden Avenue, Manhattan

                                                                       Columbia Medical Center, W68*1 Street, Manhattan

                                                                      Teaneck, NJ

                                                                      Nassau County, LI

                                                                       West End Ave Between 105* and 106* Streets, Manhattan

                                                                      4 Albany Street Spiked into NE Queens background dust
                                                                      4 Albany Street Spiked into NE Queens background dust
                                                                      4 Albany Street Spiked into NE Queens background dust
                                                                      4 Albany Street Spiked into NE Queens background dust
                                                                      4 Albany Street Spiked into NE Queens background dust
                                                                      4 Albany Street Spiked into NE Queens background dust
                                                                      Research Triangle Park, NC
                                                                      Research Triangle Park, NC
                                                                      USGS Dust Spiked into NE Queens background dust
                                                                      USGS Dust Spiked into NE Queens background dust
                                                                      USGS Dust Spiked into NE Queens background dust
                                                                      USGS Dust Spiked into NE Queens background dust
                                                                      USGS Dust Spiked into NE Queens background dust
                                                                      USGS Dust Spiked into NE Queens background dust
                                                                      Federal Courthouse, White Plains, NY

                                                                      Federal Courthouse, Central I slip, LI

-------
FP(2)
MUNYCK1)
MUNYCK2)
MUNYC2(1)
MUNYC2(2)





105
254
197
204
176





103
171
14.0
199
144
103
152
31 9
277
136
82
63
80
89
73

556

454

61 5
396
366
568
578
                                                                                Northern Manhattan, Above 70th Street
                                                                                Northern Manhattan, Above 70th Street
Samples spiked with WTC dust, at 1, 5. and 10% levels are shaded. All others are
background samples
                                                                      51

-------
TABLE 2: SEM X-Ray Mapping - Gypsum Area Percent
Sample
Designations
AP5(1)
AP5(2)
CMC(1)
CMC(2)
HS3(1)
HS3(2)
WGS(1)
WGS{2)
MW(1)
MW(2)
DB1%(1)
DB1%(2)
DB5%(1)
DB5%(2)
DB10%{1)
DB10%[2)
C1-RTP(1)
C1-RTP(2)
Laboratory Letter Codes
A


















USGS1%<1) 1
USGS1%(2)
USGS5%{1 )
USGS5%(2)



USGS10%(1)
USGS10%(2) !
USC(1)
USC(2)


B
80
203
43
69
59
149
29
61
38
54
72
71
7.3
61
65
50
85
87
63
54
77
25
63
4.8
48
62
C










13.8

3.4


8.7



152

9.8




D
144
11 3
48
30
92
11 0
54
47
70
53
5.7
5.2
55
5.5
7.8
4.8
97
82
5.8
4.1
5.7
4.1
7.1
4.8
52
42
E
09
1 8
02
1 0
03
23
02
02
02
01
3.0
1.1
0.7
0.1
0.5
0.€
02
03
1.0
0.2
09
0.5
1.2
0.7
1 2
02
F
25
1 6
1 1
1 0
57
1 5
04
03
07
1 1
0.6
1.3
1.2
1.6
1.0
1 9
1 3
08
0.9
09
1.1
24
1 1
1.4
07
24
G
341
31 3
261
308
440
290
190

232

22.0

291

25.7

245







249


H
261
337
224
176
429
405
422
391
376
41 6
28.0
30.0
24.3
28.9
27.0
28.5
534
504
29.4
29.2
29.3
21.7
30.9
328
271
324
                                                                           Location Key:

                                                                   Chittenden Avenue, Manhattan

                                                                   Columbia Medical Center, W68*1 Street, Manhattan

                                                                   Teaneck, NJ

                                                                   Nassau County, LI

                                                                   West End Ave Between 105* and 106* Streets, Manhattan

                                                                   4 Albany Street Spiked into NE Queens background dust
                                                                   4 Albany Street Spiked into NE Queens background dust
                                                                   4 Albany Street Spiked into NE Queens background dust
                                                                   4 Albany Street Spiked into NE Queens background dust
                                                                   4 Albany Street Spiked into NE Queens background dust
                                                                   4 Albany Street Spiked into NE Queens background dust
                                                                   Research Triangle Park, NC
                                                                   Research Triangle Park, NC
                                                                   USGS Dust Spiked into NE Queens background dust
                                                                   USGS Dust Spiked into NE Queens background dust
                                                                   USGS Dust Spiked into NE Queens background dust
                                                                   USGS Dust Spiked into NE Queens background dust
                                                                   USGS Dust Spiked into NE Queens background dust
                                                                   USGS Dust Spiked into NE Queens background dust
                                                                   Federal Courthouse, White Plains, NY
                                                       52

-------
FP(1)
FP(2)
MUNYC1(1)
MUNYCK2)
MUNYC2(1)
MUNYC2(2)






11 6
44
105
30
55
42






54
61
92
55
61
60
03
06
1 2
1 4
92
07
1 2
1 5
09
1 0
25
1 8
245

268

31 0

565
400
241
263
308
295
                                                                             Federal Courthouse, Central I slip, LI

                                                                             Northern Manhattan, Above 70th Street
                                                                              Northern Manhattan, Above 70th Street
Samples spiked with WTC dust, at 1, 5. and 10% levels are shaded. All others are
background samples
                                                                 53

-------
Table 3: New York City Background Dust Samples
Sam
pie
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
EPA Sample ID
HS1 -06-01"
HS1 -06-02
AP2-07-01
AP2-07-02
AP3-08-01
AP3-08-02
HS2-09-01
AP4-10-01
AP7-14-01
CMC-17-01
HS3-18-01
WGS6557
WGS5826-1
PT1 52W88
PT1 52W88-2ndFI
CY321W80
MW924WEAve
C2**
C3
C4
C4 (no date)
C5
C6
N-01S
SEM XRM
Calcium-
Rich (area
%)
58
163
124
99
105
173
256
131
175
305
143
102
181
175
158
147
21 1
79
168
137
175
167
100
123
Gypsum
(area %)
58
40
85
72
47
76
96
11 5
60
104
90
42
62
84
56
122
101
61
34
38
88
72
109
78
SEM (Heavv
Slag Wool
(fibers/g)*
35,565
230,769
32,432
7,692
12,500
<3.636
7,605
42,857
3,333
4,651
11,858
34,826
15,564
17,021
19,305
30,888
8,097
46,703
170,309
160,772
488,372
74,236
280,762
369,231
Loading)
Total
Fibers
(fibers/g)
104,603
523,077
113.514
130.769
212.500
21.818
22.814
485,714
23,333
23,256
71,146
44,776
54,475
46,809
42,471
34,749
28,340
102,890
321 ,696
227,760
790,698
148,472
415,039
523,077
Particle Count
Slag
Wool
9
15
6
2
2
0
2
3
1
1
3
7
4
4
5
8
2
11
18
24
21
17
23
24
Total
Fibers
25
34
21
34
34
6
6
34
7
5
18
9
14
11
11
9
7
24
34
34
34
34
34
34
                                                                                 Stony Brook, LI

                                                                                 West End Ave between 72nd and 73rd Streets

                                                                                 30th Ave between 21st and 23rd, Queens

                                                                                 70th Street between 20th and 21st Ave, Brooklyn
                                                                                 79th St between York and East End Ave, Manhattan
                                                                                 92nd Street between Columbus and CPW, Manhattan
                                                                                 Columbia Medical Center, W 168th St, Manhattan
                                                                                 Teaneck,  NJ
                                                                                 Nassau County, LI
                                                                                 Nassau County LI
                                                                                 88th Street between Amsterdam and Columbia, Manhattan
                                                                                 88th Street between Amsterdam and Columbia, Manhattan
                                                                                 80th Street between Riverside and East End Ave, Manhattan
                                                                                 West End Ave between 105th and 106th Streets
                                                                                 Research Triangle Park, NC
                                                                                 Research Triangle Park, NC
                                                                                 Research Triangle Park, NC
                                                                                 Research Triangle Park, NC
                                                                                 Research Triangle Park, NC
                                                                                 Research Triangle Park, NC
                                                                                 Edison, NJ
                                                         54

-------
25
26
27
28

Aver
age

Stan
dard
Devi
ation

Coeff
. Of
Varia
nee

Nevms Ct
E Curtis Ave**
LBI
Mixture







160
79
73
191

14.9

5.5

0.4

96
90
163
11 8

8.1

3.0

0.4

<4,367
5,173
<3,636
7,194

84,709

128,759

1.5

91,703
24,138
61,818
35,971

168,837

200,808

1.2

0
2
0
2







21
7
17
10







                                                                                              Edison, NJ
                                                                                              Edison, NJ
                                                                                              Long Beach Island, NJ
                                                                                              NE Queens
* A fiber count of one fiber was used to calculate the analytical sensitivity for non-
detects
" Internal laboratory duplicates were run on these samples The result shown is the
average of the two duplicates ("<" samples were assumed to be 0}
                                                                     55

-------
TABLE 4:  SEM - Slag Wool Fiber Count/Gram of Sample
Sample
Designations
AP5(1)
AP5(2)
CMC(1)
CMC(2)
HS3(1)
HS3(2)
WGS(1)
WGS(2)
MW(1)
MW(2)
DB1%(1>
DB1%(2)
DB5%(1)
DB5%(2)
DB10%(1)
DB10%{2)
C1-RTP{1)
C1-RTP(2)
USGS1%(1)
USGS1%(2)
USGS5%(1)
USGS5%(2)
USGS10%(1)
USGS10%(2)
USC(1)
USC(2)
FP(1)
FP(2)
Laboratory Letter Codes
A
non-det

non-det

16.393

5,900

12,232

5,747
34,826
72,562
67,797
104.575
84,746
246.914

98.039

600,000

1,218.855

73,394

18,519

B
3.663
<3636
3,448
<3875
7,299
7,692
34,221
10.753
18,939
3,717
10.909
17.422
29.197
25,271
66.421
77.778
159,011
173,585
109,091
83,032
404,332
343,284
840,231
1.366,470
56,025
41.199
18,051
16,470
C










5,451
9.133
32.385
33,646
74.837
57.644


50.293
50.160

364,813
531,277
521,212




D
non-det
6,980
11,800
9,620
19,000
18,600
26,400
18,100
18,700
31,800
29.900
27,300
50.800
35,800
113,000
95.100
269,000
165,000
119,000
104,000
681,000
146,000
1,620.000
238,000
91,800
40.700
16.300
31,800
E
<249
<667
<282
309
<286
<667
<256
6,990
1,320
893
<2,000
3,770
31,000
6,900
108,000
20.400
168,000
21,900
366.000
18,700
227,900
191,000
1.410.000
271,000
33,700
7,890
1,100
3,920
F
<500
500
<4,500
667
2,750
5,060
1,630
<30,500
1,000
<45,500
1.920
12,500
1.700
14,700
7.000
34.100
38,000
160,000
79.800
79,500
433,000
197.000
629,000
372,000
15,600
48,400
12,400
30,500
G
2,470
13,910
5,780
6.100
<6.320
7,370
9,480
3,520
13,630
18,080
7.650
1.320
6,230
13.040
12,900
25.210
84,650
39,930
9.200
25.370
66,450
73.330
144,120
33,040
<3,230
3,540
11,920
<1,181
H
<7.386
<7,698
<7,241
<6.289
<7,576
34,813
16,077
18,399
17,301
<9,497
15,924
16.038
107.143
70,472
114,638
96.696
188,088
318.143
90,992
137,363
672,926
347,904
734.767
413,153
29,268
74,212
28,249
25,489
                                                                                        Location Key:

                                                                             Chittenden Avenue, Manhattan

                                                                             Columbia Medical Center, W68* Street, Manhattan

                                                                             Teaneck, NJ

                                                                             Nassau County, LI

                                                                             West End Ave Between 105*1 and 106th Streets, Manhattan
                                                                                 »
                                                                             4 Albany Street Spiked into NE Queens background dust
                                                                             4 Albany Street Spiked into NE Queens background dust
                                                                             4 Albany Street Spiked into NE Queens background dust
                                                                             4 Albany Street Spiked into NE Queens background dust
                                                                             4 Albany Street Spiked into NE Queens background dust
                                                                             4 Albany Street Spiked into NE Queens background dust
                                                                             Research Triangle Park, NC
                                                                             Research Triangle Park, NC
                                                                             USGS Dust Spiked into NE Queens background dust
                                                                             USGS Dust Spiked into NE Queens background dust
                                                                             USGS Dust Spiked into NE Queens background dust
                                                                             USGS Dust Spiked into NE Queens background dust
                                                                             USGS Dust Spiked into NE Queens background dust
                                                                             USGS Dust Spiked into NE Queens background dust
                                                                             Federal Courthouse,  White Plains, NY

                                                                             Federal Courthouse. Central  Islip, LI
                                                       56

-------
MUNYCK1)
MUNYCK2)
MUNYC2(1)
MUNYC2(2)
10.840

41,298

7.220
3.745
28.777
48.507




14,400
20.200
66,500
45,500
14,900
1.960
1.390
24,200
13,100
<22,300
17,800
30,500
<2,545
<1,228
<12,453
2,330
6,803
41,118
123,106
59,473
Samples spiked with WTC dus^ at 1, 5, and 10% levels are shaded. All others are background samples.

  For data analysis purposes
      •  Non-det = Non-detect - zero slag wool fibers were noted in the sample
      •  <# indicates that the value was less than the detection limit of the respective laboratory
         by the square root of 2.
                                                                                            Northern Manhattan, Above 70  Street
                                                                                            Northern Manhattan, Above 70* Street
When this result was reached, the value was divided
                                                                       57

-------
Table 5: SEM - Slag Wool Fiber Count
Sample Designations
AP5(1)
AP5(2)
CMC(1)
CMC(2)
HS3(1)
HS3(2)
WGS(1)
WGS(2)
MW(1)
MW(2)
DB1%m
DB1%(2)
DB5%(1)
DB5%(2)
DB10%(1)
DB10%(2)
C1-RTP(1)
C1-RTP(2)
USGS1%(1)
USGS1%(2)
USGS5%(1)
USGS5%(2)
Laboratory Letter Codes
A
0

0

3

1

2

1
7
8
12
16
15
20

15

99

B
1
0
1
0
2
2
9
3
5
6
3
5
8
7
18
21
45
46
30
23
112
92
C










1
2
7
7
12
12


9
11

62
D
0
3
5
4
8
8
11
8
8
14
13
12
22
16
48
42
116
72
54
47
194
65
E
0
0
0
1
0
0
1
7
6
2
0
4
7
8
•V3
9
16
30
27
23
25
21
F
0
1
0
0
1
3
1
0
1
0
1
2
1
2
2
3
4
4
11
4
>20
19
G
2
6
4
5
0
4
6
3
6
22
7
1
6
11
10
25
22
17
11
H
0
0
0
0
0
4
2
2
2
0
2
2
6
10
13
12
24
37
10
22 I 15
64
43
27 ; 39
                                                              Location Key:

                                                              Chittenden Avenue, Manhattan




                                                              Teaneck, NJ

                                                              Nassau County, LI
                                                              4 Albany Street Spiked into NE Queens background
                                                              dust
                                                              4 Albany Street Spiked into NE Queens background
                                                              dust
                                                              4 Albany Street Spiked into NE Queens background
                                                              dust
                                                              4 Albany Street Spiked into NE Queens background
                                                              dust
                                                              4 Albany Street Spiked into NE Queens background
                                                              dust
                                                              4 Albany Street Spiked into NE Queens background
                                                              dust
                                                              Research Triangle Park, NC
                                                              Research Triangle Park, NC
                                                              USGS Dust Spiked into NE Queens background dust
                                                              USGS Dust Spiked into NE Queens background dust
                                                              USGS Dust Spiked into NE Queens background dust
                                                              USGS Dust Spiked into NE Queens background dust
                                               58

-------
USGS10%(1)
USGS10%(2)
USC(1)
USC(2)
FP(1)
FP(2)
MUNYCK1)
MUNYC1(2)
MUNYC2(1)
MUNYC2(2)

181

6

3

1

7


45
45
13
11
5
5
2
1
8
13

124
129









450
105
39
18
7
14
6
9
28
20

38
19
6
6
3
4
4
2
3
24

19
16
13
4
2
1
3
0
3
3

18
9
0
2
2
0
0
0
0
1

41
49
3
8
3
3
1
5
13
7

                                                                                 USGS Dust Spiked into NE Queens background dust
                                                                                 USGS Dust Spiked into NE Queens background dust
Samples spiked with WTC dust, at 1,5. and 10% levels are highlighted in yellow
                                                                  59

-------
 ADDENDUM TO VERSAR REPORT: SEM CALIBRATION DATA

                             MEMORANDUM

TO. Jacky Rosati

CC. David Friedman

FROM Stephen Schwartz

DATE  August 5, 2005

SUBJECT  BIR-1G Sample Analyses
       Identical mounted and polished reference samples, each designated BIR-1G, were
sent to each of the five private laboratories participating in the analyses of dust samples
from New York City and elsewhere  The samples were analyzed by Scanning Electron
Microscopy/Energy  Dispersive X-ray  Spectrometry  (SEM/EDX) to determine their
elemental content  The purpose of the study was to determine the variation within and
between each of the laboratories, and to assess their ability to identify elements using this
technology

       Each of the five laboratories  analyzed their BIR-1 G sample between 4 and 11
times, as convenient (there was no requirement for a specific number of analyses).  The
average elemental concentration data for each laboratory is presented in the attached
table    For calcium (Ca),  magnesium (Mg), silicon (Si), and  oxygen (O)4,  which
constitute  over 80% by  weight of the elemental  composition, the standard deviation
within  each laboratory, for each element, was typically much less than 10% (i.e,  the
coefficient  of variation)  Likewise, the coefficient of variation between laboratories for
Ca, Mg, Si, and 0, as shown on the attached table, was also much less  than a 10%  (The
graphic presentations of the EDX spectra within and between laboratories also appear to
be extremely similar.

       Therefore, it can be concluded that  each of the laboratories was easily able to
achieve excellent precision, by SEM/EDX, in quantifying the elements that were present
in larger concentrations
4 Some of the laboratories reported results as the weight percent of the elemental oxides, specifically Na2O,
MgO, A12O3, SiO2, CaO, TiOj, FeO, K2O, and MnO2 Oxide values were converted to individual elemental
values (e g, A12O3 is about 53% Aluminum, and 47% oxygen by weight)

                                       60

-------
      AVERAGE ELEMENTAL CONCENTRATION REPORTED FOR BIR-1G SAMPLES (Weight Percent of Sample)
Lab

A
B
C
D
E
F
G
H

AVERAGES

Standard Dev.
(%of
Average)

Sodium





044

1 14
1 84

1 14

61 40

Magnesium





502

582
510

531

833

Aluminum





811

875
788

825

548

Silicon





21 60

2476
2220

2285

735

Calcium





966

768
1030

921

1483

Titanium





068

052
064

061

1358

Iron





968

556
837

787

2673

Potassium





NR

000
003

002

141 42

Manganese





NR

000
021

011

141 42

Oxygen





4484

4578
4343

4468

265

NR -NotReported
                                             61

-------
 APPENDIX F: STATISTICAL ANALYSIS AND INTERPRETATION OF TEST
                RESULTS- LABORATORY QUALIFICATION
Slag wool fiber content as a discriminator for residual WTC contamination in
indoor dust sample: Interpretation of multi-laboratory test results

Introduction

Eight laboratories were each challenged with a number of blinded dust sample aliquots to
determine number of slag wool fibers per gram These samples included background
dusts from various locations in NYC and a series of samples of common household dust
spiked with different levels of WTC collapse dust collected in 2001 or in 2004 The
purpose of this endeavor was to assess whether or not the method developed can be used
by qualified laboratories to discriminate between WTC and non-WTC impacted dust
samples.

In the following discussion, individual laboratories are evaluated and ranked for validity
and precision, and then the top performers are further evaluated as groups to determine
the expected confidence level for the slag wool content of any individual and randomly
assigned)sample.

Laboratory Qualification

Validity: Assessment of validity was conducted by analysis of a series of spiked samples
where the expected response ratios are known. The challenge samples consisted of a large
volume of non-impacted background dust collected in 2004 from locations in Northeast
Queens over ten miles from the WTC site This dust was subsequently spiked with 1,5,
and 10% WTC dusts by weight using either bulk collapse dusts collected in September
2001 immediately following the disaster (designated  as USGS dust), or nominally
undisturbed dusts collected in 2004 in the abandoned Deutsche Bank (DB) complex that
borders the south side of the WTC complex (designated as 4 Albany dust).

Using units of (# slagwool fibers)/(gram of dust), preliminary analyses showed a mean
value of 12,200,000 (s d  1,697,056) for USGS dust, 579,667 (s d 173,782) for 4
ALBANY dust, and a nominal background level of 7,190 Based on these data, the
expected values of slope expressed as [(# slagwool fibers)/(gram dust)]/[% spike level]
are 121,928 and 5,725, respectively for USGS and 4 Albany spikes Specifically, each lab
was furnished two samples each of 1, 5,10% spikes from both 4 Albany and  USGS series
for a total of 12 spiked samples plus a series of 20 additional background samples
collected from random locations all over the greater NYC area

Scatterplots and linear least squares regressions were constructed for each lab and for
each of the two spike series Preliminary inspection showed no apparent violations of
underlying  assumptions required for regression analysis (primarily homogeneity of
variance), as such no lognormal transformation was performed. Using a forward selection
strategy, it was found that a higher order polynomial  model does not statistically improve
the linear fit, this  is expected as the sample set is designed as a linear progression Also, a
simpler but more general "runs" test for each linear regression confirmed these results
Data handling and manipulation was performed with Microsoft Excel SP-2, statistical
analyses were performed with SAS 9.1.3 XP-Pro (proc rsreg/lackfit, proc reg, proc
                                       62

-------
mixed, and proc unwanate); graphing, ANOVA, and various other statistical results were
performed or verified with GraphPad Prism 3 03 The linear regression results are given
in Table 1, a summary of SAS proc rsreg/lackfit results are given in Table 2

Table r Summary of linear least squares regression results
Lab


A
B
G
D
E
F
G
H
DB spskes (4
slope

8126
6541
6554
8538
6936
1523
1630
9695
Albany)
95% C!
<*/•)
1760
971
691
1515
3632
1288
610
2645

sig siope
p-valus
0,0099
0.0025
0.0007
00049
0 *t28S
G.3027
0.0559
0,0215

f2

0,8421
G91SO
09573
08881
04769
G2SS8
05404
0.7705
USGS spikes
(2001 dusts)
slope 95% Cl ssg

124500
1 1 3300
52380
91340
74240
45370
7750
49510
{+•'-* P-
488
23460
5914
58230
49310
S40SO
4607
21790

slope
valtfs
0.0025
0.0085
00030
0,19t8
0.2104
00298
0,1873
0.0855

f2

1 0000
08537
O.S632
0,3608
0,3571
07325
0,4 144
05634
                                                                                 note: n =3
Table 2 Summary of "lack of fit" tests
Lab
A
B
C
D
E
F
G
H
A B.C.and D
A,B,C,D and H
r2
linear
00421
0.9190
Q.S573
0.888 1
04769
OZ588
0.6404
0.7705
07550
07027
r2
quad
00590
00564
OOOQ1
0.0706
0.0197
00445
00069
0 1455
0.0032
0.0023
p-v«lue
lin«ar
00149
0.0058
0 003$
0,0040
0.5904
03386
0,10:S
0.0 » 35
00001
0.000 1
pvaiue
quad
G2732
&079S
0944J
0 1037
07547
G6913
0.8240
01C70
06038
06473
 Based on these summaries the linear model is appropriate  Laboratories A and B
 demonstrate excellent performance across the board  each has r2 values > 0 80, significant
 positive slopes with p < 0 05, and slopes with the expected magnitude  We caution that
 Lab A only has three points for the USGS spike results (yellow highlights, Table 1)
 Fields highlighted in blue indicate potential problem areas If only the 4 Albany spike
 series are considered, then Labs C and D can be added to the preferred performer group
 This is reasonable because the range covered here is more likely to reflect the range of
 concern for unknown samples Although the Lab H results demonstrate a lower r2 value
 and a larger 95% CI for slope, this is caused by a single outlying point. As such, there is
 no reason to exclude Lab H from the analysis Because Labs E, F, and  G fail in more than
 one category in both spiked data sets, they are not included in the remaining study
 analysis
                                        63

-------
       From the above discussion, we can construct two groups of laboratories based on
the estimated validity of their results: "Best", consisting of Labs A, B, C, and D and
"very good" consisting of the best group plus Lab H. In the following series of figures,
the individual and composite linear regression results for the groups are demonstrated
graphically.
   120000 •
   100000 -
    80000 •
    60000
    40000
    20000
                             Spiked Samples - 4 Albany
                                   "Best" Group
&  LabB
D  Lab A
4  LabC
*  LabD
                123456789
                                   % spike level
                                                     10    11
                                       64

-------
120000  -
100000  -
 80000
 60000
 40000
 20000
                        Spiked Samples - 4 Albany
                         "Best" Group Combined
                          (Labs A, B, C, and D)
           Slope: 7440+-903
           r2:    0.7550
      01234567
                             % spike level
                                      8     9    10    11
120000
100000
 80000
 60000
 40000
 20000
O  LabB
D  Lab A
.4  Lab C
4  LabD
*  LabH
       0    1
                        Spiked samples - 4 Albany
                           "Very Good" Group
                         (Labs A, B, C, D and H)
                 456
                   % spike level
                                                 8     9    10    11
                                  65

-------
120000
100000 -
 80000 •
 60000
 40000
 20000
                        Spiked samples - 4 Albany
                      "Very Good" Group Combined
                         (Labs A, B, C, D and H)
           Slope: 7891 +- 970
           r2:   0.7027
      0123456789    10    11
                             % spike level
120000
100000
 80000
 60000
 40000
 20000
          O LabF
          C Lab G
          O LabE
      0     1
                        Spiked samples - 4 Albany
                           "Outlying" Group
                           (Labs E, F and G)
456
  % spike level
                                O
8    9    10    11
                                 66

-------
                               Spiked samples - 4 Albany
                              "Outlying" Group Combined
                                   (Labs E, F and G)
      120000 -
      100000 -
       80000
       60000
       40000
       20000
             0     1
456
  % spike level
1U    11
Precision: Up until this point, validity has been assessed only with those samples for
which there is some prior knowledge of content. For assessing precision, however, one
can use all of the samples (including unknowns) because each laboratory received
aliquots of the same set of 32 samples. Furthermore, the sample structure is such that
these 32 samples are comprised of 16 paired samples allowing within laboratory
precision estimates as well. Although there are a number of statistical options for
proceeding, an analysis of variance (ANOVA) and intra-class correlation coefficients
(ICC) are pragmatic for these circumstances as samples and laboratories are used in
groups. Preliminary analyses of "within" and "among" laboratory results indicate that the
underlying distributions (considering all 32 sample results) are not normal based on the
Shapiro-Wilk (S-W) test, and that natural log transformation of the data should used to
perform analysis of variance. The only exception is the USGS data set where only three
pairs of samples are reported and thus the natural space numbers did not require
transformation. Table 3 shows the results for the ICC analyses within laboratories, and
also for the groups (Labs A, B, C, D) and (Labs A, B, C, D, H) aggregated. The variance
components and p-values for the S-W normality test are also given. The lower part of
Table 3 gives the aggregated results for the background samples only; Laboratories A and
C did not contribute to these statistics but it is expected that they would perform
similarly.
                                       67

-------
Table 3 Summary statistics for mtra-class correlation coefficients

 Ml available Pairs

 IntraClass Cotretaiion Calculations: from SAS proc mixed
 Log Space data
Lab

A
AA
B
C
0
E


H

A.B.C.snd D

A,e,C,D and H
A natural space
Ali NYC Backgrou
nobs

6
5
32
10
31
32
32
32
32

80

112
Int

7.6570E-G1
1.331QE+09
2.7493E+GO
1 9870£+00
1,396e£*03
5.5988£+00
2.8559E+00
1 0962E+C3
2.064 7£* 00

2.0491E+00

2.0820E*00
Res

5.4910E-Q1
2 1025E+08
1.7470E-01
7 3Q50£'Q1
2.7860E-01
1 7150F+QG
1 3172E+GG
67860E-01
3.2810E-01

3.02SDE-01

2.8740E-01
!CC

0.5824
08536
09403
0.7312
08337
0.7655
06844
0,6176
08629

0.8714

0.8787
S-W p-valus

02786
0.767S
O.i 059
QG342
0.8847
06012
OSOS7
03571
0.1571

0.3296

0 '-540
nd Pairs
 intraCiass Correlation Calculations: from SAS proc mixed
 Log Space data
Ubx

A.S.C.and D

A(8,C,D and H
nobs

36

54
int

64570E-01

6-SSOOE'QI
fics

3.5070E-01

3.4290E-Q1
!CC

06430

06706
S W p value

02657

0.2571
*Laboratones A and C did not report paired New York City background data

From this exercise, we see that all of the individual laboratories demonstrate reasonable
ICCs (generally above 0 6) Furthermore, the laboratory groups chosen to demonstrate
good validity show ICCs greater than 0.87 when all data are considered. When only the
New York City background samples are analyzed, the ICCs are somewhat lower. These
results can be interpreted to mean that about 35% of the variance is attributable to
variability in the pooled laboratory analyses, and the remainder to true differences among
the background samples

As a further assessment of inter-laboratory precision, the between laboratory ANOVA
shows no reason to reject the null hypothesis (Ho = no difference, in natural log space)
among Laboratories A, B, D, and H  Laboratory C was left out of this analysis because
they reported no background data at all
                                       68

-------
Evaluation of Unknown Samples



In the previous section we qualified a group of laboratories for measurement of

unknowns based on spiked samples (validity), and comparative precision measures based

on ICC and ANOVA. We now assume that these laboratories are statistically similar and

combine their spike results into a single response graph. Based on these results, we

calculate 95% confidence intervals and 95% prediction bands as illustrated in the figures

below.
    140000
    120000-
    100000
  01


  I
     80000-
  u-  60000-

  o


  I

  01  40000-

  V)


     20000-
                         Splked samples - 4 Albany

                       Labs A, B, C, D, and H combined
          0123456789   10   11   12
    140000-
    120000-


  1
  •o

  E 100000-
  2
  Dl
  ^

  8. 80000-

  12

  £
  E  60000-1
   01 40000
20000-




   0
                           Spiked samples - 4 Albany

                                 Lab B only
                     upper 95% prediction band
                                                lower 95% prediction band
                               4567

                               percent 4 Albany spike
                                                    8    9
                                                              10    11
                                         69

-------
The major effort here is to estimate the performance of the aggregate laboratory group
(A, B, C, D, and H) with respect to the group of samples from the greater New York City
area designated as "background" or "non-WTC impacted". The composite behavior of
these samples is illustrated below with respect to the analytical laboratories. The graph
indicates no consistent (high or low) percent bias from the cross laboratory means. This
confirms the conjecture made earlier that these laboratories are statistically similar. We
caution that Laboratory C  did not provide any background data at all and could not be
directly  included here, however, it is assumed that it would behave like the others.

                    Sample # vs percent difference by labs
                       (Slag Wool Fibers per gram dust)
                            NYC  Background only
          1   234567
8   9 10 11 12 13 14  15 16  17 18
  sample!
The next step is to assess how an individual (presumably unknown) dust sample assay
relates to the amount of spiked 4 Albany dust percentage. Given the graph and underlying
statistics of the above figure entitled "Spiked Samples - 4 Albany, Labs A. B, C, D and H
Combined), one can calculate the x-value in % spiked 4 Albany equivalent and the 95%
confidence interval for the prediction for any unknown sample measurement from any
laboratory. This is essentially the use of the prediction band graph above in reverse. As
such the prediction of "x" and the CI take the following form:

      Xpredicted = (Ybar - a)/(b)

      Cl = Xprcd,cted ± [t(RSE)/b] *  {1/m + 1/n + [(Ybar - ybar)2/(b2(n-l)sx2)}'2

                                       70

-------
where Ybar is the mean laboratory measurement, a and b are the intercept and slope of the
regression, t is the critical t-value for n-2 degrees of freedom, RSE is the residual
standard error, m is the # of replicate measurements, n is the number of calibration points,
ybaris the mean of the regression y data, and sx is the standard deviation of the x values of
the regression data.
The reported slag wool results for the background samples can now be interpreted  Table
5 shows the results for each background sample measurement across all participating
laboratories as a prediction of the percent equivalent 4 Albany spike level and half of the
95% confidence interval associated with the measurement  There are a total of 63
measurement results in the table
Table 5' Results for each background sample across all participating laboratories as a
prediction of the percent equivalent 4 Albany  spike level and ± 95% confidence interval.

I  Lab*     A                 B                D                 H
I Sample  Percent   CI +.   percent   CI +-   Percent    CI +-   Percent    CI +-
:    *

:    1       -1  35     5 43
    3      -135     543
    4
    5       0 72     5 29
    6
    7      -061     5.38
    8
    9       0 20     5 33
    10
    25      7 95     5 21
    26
    27      0 99     5 28
    28 .....
    29      662     534
    30
    31      388     518
 •   32

 * Laboratory C was left out of this analysis because they reported no background data
 We note that negative entries above are only statistical constructs  Of the 63 background
 measurements  in this table, 7 (or about 11%) exceed the 4 Albany 5% spike level, 2 of
 the 63 measurements exceed the 10% 4 Albany spike level  If the upper confidence
 limits are considered, 42 out of 63 (67%) exceed the 5% spike level and 7 of 63 (1 1%)
                                         71
-089
-1 03
-092
-101
-0.43
-038
298
001
1 05
-088
575
387
0.93
073
-044
-0.88
2.29
4.79
540
541
5.40
541
537
536
520
534
528
540
5 17
518
528
5.29
5.37
540
522
5 17
-1.35
-0.47
0.14
-014
1 05
1.00
199
094
102
268
10.28
3.80
071
' 2.68
047
1 21
7.07
441
543
537
533
535
528
528
523
528
528
521
531
518
530
	 52l"
531
" 5"27 	
518
5 17
-069
-066
-071
-079
-068
306
068
098
0.84
-050
235
805
223
1.88
-0.49
386
1425
618
538
538
539
539
538
520
530
528
529
537
522
5.21
522
5.24
5 37
"518
5.63
517

-------
exceed the 10% spike level  For instance for sample 25 at lab A the percent equivalent of
the fiber measurement is 7 95% and the upper confidence limit is 7 95% + 521% =
13 16%

As a further exercise, we calculated the same statistics for data within only one laboratory
(choosing Laboratory B as the example), these results do not include scatter in the
regression from the other qualified laboratories Here we find some improvement we see
3 of 18 values (16 7%) exceed the 5% 4 Albany dust level and 0 of 18 values exceed the
10% 4 Albany dust level For the upper confidence levels, 6 of 18 values exceed the 5%
4 Albany dust level and 1 of 18 values exceed the 10% 4 Albany dust level.

Conclusions

The conclusions are based solely on the analytical data provided from the laboratory test
and a few analyses of the 100% WTC spike samples From validity estimates based on
expected slopes and data scatter of WTC spiked samples, five of eight laboratories (A, B,
C, D, and H) were used for further analysis  Intra-class correlation coefficients (with
natural log transformation) for individual labs and for the group  of five demonstrate
similar and reasonable values (>0 7) when all available data are considered One-way
ANOVA analysis of Laboratories A, B, D, and H results provides no evidence to reject
the null hypothesis (that the results are from the same distribution)  Laboratory C was
not included here because of insufficient reported data but, based on spike sample
statistics, it is likely that that they too would fall into this category

Under the practical constraints that the five laboratories are used at random with  one
analysis per unknown sample, we cannot expect statistical discrimination at the 1% or 5%
4 Albany spike equivalent level because the upper 95% prediction bounds exceeds the
5% spike equivalent level across the board  Reasonable discrimination is possible at the
10% 4 Albany spike equivalent level  because the lower bound onlO% equivalent
measurements is  approximately equal to the mean at 5% Albany spike equivalent level
                                       72

-------
 Al. Quality Assurance Project Plan for World Trade
         Center (WTC) Screening Method Study
                         Project # WTC-1
                            Revision # 3
                           August 8, 2005
                            Prepared by:
                U.S. Environmental Protection Agency
                 Office of Research and Development
                 Research Triangle Park, NC  27711
Approvals given below indicate that technical and administrative reviews have been conducted,
and reviewer comments for the document preceding the signature date have been resolved
Dr Jacky A Rosati, EPA Principal Investigator, ORD, NHSRC            Date
David Friedman, EPA Principal Investigator & Project Officer, ORD, IOAA    Date
Shirley Wasson, EPA QA Representative, ORD, NRMRL                 Date

-------
ACRONYMS
COPCs       Contaminants of Potential Concern
DQI         Data Quality Indicator
DQO         Data Quality Objective
EPA         U S Environmental Protection Agency
ERT         Emergency Response Team (EPA)
HEPA        High Efficiency Particulate Air
HVAC       Heating, Ventilation and Air Conditioning
IOAA        Immediate Office of the Assistant Administrator
MQI         Measurement Quality Indicator
NERL        National Exposure Research Laboratory (EPA)
MQO        Measurement Quality Objective
NHSRC      National Homeland Secunty Research Center (EPA)
NRMRL      National Risk Management Research Laboratory (EPA)
ORD         Office of Research and Development (EPA)
PLM         Polarized Light Microscopy
QA          Quality Assurance
QC          Quality Control
QAPP        Quality Assurance Project Plan
SEM         Scanning Electron Microscopy
SOP         Standard Operating Procedure
TEM         Transmission Electron Microscopy
USGS        US Geological Survey
WTC         World Trade Center

-------
                            A2. TABLE OF CONTENTS

 A PROJECT MANAGEMENT                                                   5
A3 Distnbutton List                                                              5
A4 Project Organization                                                           5
A4 1 Responsibilities and Roles                                                     5
A4 1 1 Analytical Laboratory Responsibilities                                         6
A4.1 2 Sampling Responsibilities                                                   6
A4.2 Reporting Relationships                                                       7
A5 PROBLEM DEFINITION AND BACKGROUND                                  9
A5.1 Introduction                                                                9
A5.2 Method Development                                                        10
A.5 3 Screening Method Study                                                     11
A6. Project/Task Description                                                       12
A 6 1 Sampling                                                                  12
A 6 2 Sample Preparation                                                          12
A 6 3 Analytical Study                                                           12
A 6 4 Tasks and Timeframes                                                       13
A 7 Quality Objectives and Criteria                                                 13
A7.1 Data Quality Objectives  (DQOs)                                               13
A 7 2 Measurement Quality Objectives (MQOs)                                       13
A8 Special Training/Certifications                                                  14
A9 Documents and Records                                                        14
B DATA GENERATION AND ACQUISITION                                      15
Bl  Sampling Process Design                                                      15
B2 Sampling Methods                                                           15
B3 Sample Handling and Custody                                                 15
B3 1 Homogemzation                                                            -15
B4 Analytical Methods                                                           16
B5 Quality Control                                                               16
B6 Instrument/Equipment Testing, Inspection and Maintenance                          18
B7 Instrument/Equipment Calibration and Frequency                                   18

-------
B8 Inspection/Acceptance of Supplies and Consumables                               18
B9 Non-Direct Measurements                                                    18
B10 Data Management                                                          18
C ASSESSMENT AND OVERSIGHT                                             19
D DATA VALIDAT1ON AND USABILITY                                        19
APPENDICES                                                                20
REFERENCES                                                                20
                                   FIGURES

Figure 1 Project Organizational Chart                                               7

                                    TABLES

Table 1 Intralaboratory Measurement Quality Objectives (MQOs)                       13
Table 2 Interlaboratory Measurement Quality Objectives (MQOs)                       14
Table 3 Mass Concentrations of Spiked Samples                                     16

-------
A. PROJECT MANAGEMENT

Presented herein is the site Quality Assurance Project Plan (QAPP) for the World Trade Center
Screening Method Validation Study  The QAPP has been developed in accordance with the
United States Environmental Protection Agency (EPA) guidance for Quality Assurance Project
Plans (EPA QA/G-5, EPA/240/R-02/009) December 2002 This plan is based on information
currently available and may be modified on site in light of field observations/results and other
acquired information  Any modifications or deviations from this QAPP shall be approved by EPA
and documented


A3.  Distribution List

The Distribution List documents who shall receive copies of the approved QAPP and any
subsequent revisions or amendments to the QAPP The US EPA shall distribute the QAPP to all
project team members and shall ensure that the project team members are familiar with any and all
QA issues A complete copy of the QAPP and any subsequent revisions shall be maintained on
file at the U S EPA RTF office and shall be available upon request. The following personnel will
receive copies of the approve QAPP for the WTC Screening Method Study (contact information
for these people can be found in A4)-

EPA
    1. Jacky Rosati, EPA, ORD, NHSRC,  Pnncipal Investigator
    2 David Friedman, EPA, ORD, OAA, Principal Investigator
    3 Rajeshmal Smghvi, EPA/ERT, WAM - Sampling Contract
    4 Ten Conner, EPA, ORD, NERL ,WAM - SEM contract
    5. Shirley Wasson, EPA, ORD, NRMRL, APPCD - Quality Assurance

Other Government Agencies
    6 Greg Meeker, USGS - IAG Manager

Contractors
    7 Cindy Klemman, Lockheed Martin -Sampling Contractor
    8  Bob Willis, Alion Sciences - SEM Contractor
    9 Steve Schwartz, Versar - Analysis  Prime Contractor
    10 Keith Rickabaugh, RJ Lee Group, Inc - Analysis Subcontractor
    11 Rich Brown, MVA Scientific Consultants - Analysis Subcontractor
    12. Garth Freeman, MAS, Inc - Analysis Subcontractor
    13 John Newton, EMSL Analytical Inc - Analysis Subcontractor
    14 Jeannie Orr, Reservoir Environmental,  Inc - Analysis Subcontractor

A4. Project Organization

A4.1 Responsibilities and Roles

-------
Dr Jacky Rosati of the EPA, ORD, NHSRC and David Fnedman, ORD, OAA shall have the
oversight authority for all work conducted for this project and shall act as backup and work
assignment manager on the Versar analytical contract, respectively  Dr Rosati has prepared the
Validation Study QAPP, and Shirley Wasson, EPA, ORD, NRMRL, APPCD will perform the QA
review of this QAPP  Dr Rosati and Mr Fnedman shall provide technical assistance to ensure
that sample collection and analysis work is completed efficiently, and in compliance with the
applicable Scope of Work and all applicable rules.

A4.1.1 Analytical Responsibilities

Steve Schwartz, Versar, shall arrange and oversee the analysis work by analytical laboratories
Mr Schwartz and his subcontractors shall adapt, adopt and follow the Quality Assurance Project
Plan prepared by EPA for all sample analysis activities performed for this project  All appropriate
data, original field forms/data sheets, shall be collected and completed in accordance with the
instructions contained in the contract and provided to EPA

The analytical laboratories retained by Versar include RJ Lee Group, MVA Scientific, MAS,
EMSL Analytical, and Reservoir Environmental  Dr Rosati will arrange and oversee the analysis
work by the U S government analytical laboratories (Greg Meeker, USGS, Ten Conner, EPA,
NERL) All analytical and government laboratones shall conduct the required analysis within the
requested turnaround time, input the relevant analytical data into the appropnate spreadsheets, and
other similar duties as descnbed in its contract Scope of Work, IAG or as requested by Dr. Rosati
All data for this project are considered confidential and only the EPA is authorized to allow for
their release

The pnmary contractor for the environmental sampling shall follow the QAPP that they have
prepared entitled "Generic Quality Assurance Project Plan for WTC Residue Sampling New York
City, NY, March 2005 (Appendix A).  All appropriate data, original field forms/data sheets, and
cham-of-custody forms shall be collected and completed in accordance with the instructions
contained in the contract and provided to EPA  All samples shall be handled as instructed by
EPA, to include sample sieving, ashing, splitting, archiving and distnbuting

A4.1.2 Sampling Responsibilities

Raj Smghvi, EPA, ERT Work Assignment Manager shall oversee the sample collection performed
by Lockheed Martin (Cindy Kleinman), as directed by Dr Jacky Rosati. All sampling
appointments shall be arranged by Raj Smghvi, EPA/ERT for Lockheed Martin (Cindy
Kleinman)  All samples are to be archived and stored in a safe location as  described in Section
31

Lockheed Martin, the pnmary contractor for the environmental sampling shall follow the QAPP
that they have prepared entitled "Generic Quality Assurance Project Plan for WTC Residue
Sampling New York City, NY, March 2005 (Appendix A)  All appropriate data, original field
forms/data sheets, and chain-of-custody forms shall be collected and completed in accordance
with the instructions contained in the contract and provided to EPA.  All samples shall be handled
as instructed by Dr Rosati, to include sample sieving, ashing, splitting, archiving and distributing
All data shall be considered confidential  Only the EPA is authonzed to release any data collected
for this project

-------
A4.2 Reporting Relationships

The Project Organizational Chart (Figure 1) shows the reporting relationships between all of
organizations involved in this project, including the lead organization (i.e., EPA) and all
contractors and subcontractors.
                                   Dt -Us. (ft
                                       J
           l-s U,kl
        Kasfe huul .s ni-h' i
     »\ • \ s.i
   U-eUwJ Mirtai i
         :B5JF »-•>  |
                                 .
JIG&CT
              s>
             '5 i> i
           VVoik .
                                                           MX" (.
                                                            5jJ!(T.U 'v't
                                                -.Hi
                            Figure 1: Project Organizational Chart
 Contact Information for the above listed personnel (alphabetical):

-------
Rich Brown
MVA Scientific Consultants
5500 Oakbrook Parkway - Suite 200
Norcross, GA 30093
770-662-8509

Ten Conner
U S EPA National Exposure Research Laboratory
109 TW Alexander Drive, D205-03
Research Triangle Park, NC 27711
919-541-3157

Garth Freeman
MAS, Inc
3945 Lakefield Court
Suwanee, GA 30024
678-687-5990

David Friedman
U S EPA Headquarters
Anel Rios Building
1200 Pennsylvania Avenue, N W
8101R
Washington, DC 20460

Cindy Kleinman
Lockheed Martin
USEPA Facilities
Raritan Depot
2890 Woodbndge Avenue
800MS800
Edison, NJ 08837-3679
732-321-4252

Greg Meeker
USGS
Bldg53Sl, Mail Stop
Denver Federal Center
Denver, CO 80225
303-303-236-3188

John Newton
EMSL Analytical Inc
108 Haddon Avenue
Westmont, NJ 08108
856-858-4800

Jeannie Orr

-------
Reservoir Environmental, Inc
2059 Bryan Street
Denver, CO 80211
303-964-1986

Keith Rickabaugh
R J Lee Group, Inc
350 Hochberg Road
Monroeville, PA 15146
724-325-1776

Jacky Rosati
U S  EPA National Homeland Security Research Center
109 TW Alexander Drive, E305-03
Research Triangle Park, NC 27711
919-541-9429

Steve Schwartz
Versar, Inc
6850 Versar Center
Springfield, VA 22151
703-642-6787

Rajeshmal Singhvi
US  EPA ERT, Region 2
Raritan Depot
2890 Woodbndge Avenue
101MS101
Edison, NJ 08837-3679
732-321-6761

Bob Willis
Ahon Scientific
109 TW Alexander Drive, E205-06
Research Triangle Park, NC 27711
919-541-2809

AS. PROBLEM DEFINITION AND BACKGROUND


 A5.1 Introduction
The objective of this effort is to develop and evaluate a means of determining whether dust
sampled as part of EPA's future sampling program contains residual contamination attributable to
the collapse of the WTC towers  The tested screening method is a critical component of the
sampling program as it will be used, along with the results from contaminants of potential concern
(COPC) testing, to determine the need for cleanup

The USGS has published two reports which provide the basis for the WTC dust signature adopted
in this sampling program The first report discusses the analysis and interpretation of indoor and
                                                                                   9

-------
outdoor WTC dust samples collected near Ground Zero, days and weeks after September 11, 2001
(Meeker, et al, 2005)  From this work, we see that the WTC dust samples are dominated by
gypsum, concrete, and man-made vitreous fibers (MMVF), mainly slag wool   It is on the basis of
these key results that gypsum, elements of concrete, and slag wool were identified as candidates
for a WTC signature The second report discusses the analysis of EPA supplied samples taken
from several indoor locations well outside of the WTC impacted area (background)  These
samples were taken between September of 2004 and April of 2005  Slag wool is absent from
many of these background samples, but Lowers et al (2005a) state that the samples do have
gypsum present, which they speculate might be due to the presence of wall board in the sampled
apartments  Because of the lack of slag wool in these samples, it was concluded that these
samples did not contain WTC dust  It was also concluded that perhaps slag wool is the single
most critical of the three WTC dust constituents when distinguishing WTC dust from other
common dusts.

Other studies also identified MMVF and gypsum as predominant components of WTC dust  In a
study of air and settled dust quality in apartments in Lower Manhattan, the Agency for Toxic
Substances and Disease Registry (ATSDR) and the New York City Department of Health and
Mental Hygiene (NYCDOMH) found significantly more MMVF and gypsum in Lower Manhattan
apartments as compared to comparison areas above 59th St (NYCDOMH/ATSDR, 2002).
Meanwhile, no MMVF was found in comparison locations. They also concluded that gypsum was
seen at a higher percentage in dust m Lower Manhattan samples as compared to the comparison
area samples   In a comprehensive study  of the composition of settled dust in the Deutsche Bank
building at 130 Liberty St, R J  Lee identified numerous hazardous contaminants that were present
in the dust at levels much higher than in background office buildings, and among those substances
identified in their "WTC signature" were mineral wool and gypsum (R J  Lee, 2004).

 If the WTC building collapse signature components of slag wool, gypsum, and elements of
concrete are not present, then one could conclude that WTC building collapse dust is not present
However, since these components might be present in typical New York City dust, as slag wool is
a component of insulating materials in currently constructed buildings, it is possible that a test
might show them to be present  even though WTC dust never impacted the sampled area. A
'screening test' will, by its design, result in some fraction of such false positives (a location
without residual WTC dust that tests positive for the above components)   However, an
appropnate 'screening test' would result in very few,  if any, false negatives (a location with
residual WTC dust that tests negative for the above components).

A5.2. Method Development

EPA acquired 117 dust samples dunng the time period of September 2004 to April 2005.  Twenty-
one 'impacted' samples were taken by the EPA at two buildings that were part of the Deutsche
Bank complex located at 130 Liberty Street and 4 Albany Street  Both buildings were uninhabited
and slated for demolition  Fifty samples were taken from locations well beyond the impacted
zone, these samples are considered to be 'background' dust Forty-six samples were taken from
locations that were possibly impacted but  were a bit farther from the WTC site than the known
'impacted' samples  None of these forty-six samples  are used in the study, but several were
evaluated during the development of the analytical method. In addition, one impacted sample was
obtained from the USGS. This sample was a composite sample of outdoor and indoor WTC dust
collected in September of 2001

                                                                                    10

-------
While a standard method using a HEPA vacuum collector was used by EPA to collect most bulk
dust samples (Appendix C), some bulk dust samples were collected from residential and
commercial vacuum cleaner bags.  Whether collected by HEPA vac or by acquiring vacuum
cleaner bag, all samples are handled (sieved, split and stored) as described in the sampling QAPP
in Appendix C  Many of the above samples were analyzed for slag wool content by the EPA's
National Exposure Research Laboratory (NERL) Scanning Electron Microscopy (SEM)
Laboratory  This analysis was performed as part of the EPA's development of a standard protocol
for sample preparation and analysis (Appendix B), to determine the sample status (background or
impacted) and content.

There appears to be a clear distinction between samples taken in impacted areas versus
background samples  All of the impacted samples had slag wool at concentrations of greater than
100,000 fibers per gram of dust, with a range of 113,000 to 13,400,000, while all of the
background samples had concentrations less than 100,000 ppm, ranging from no slag wool
identified inlO samples to 92,800 fibers of slag wool per gram of dust Based on this preliminary
work, the USGS, the EPA's Office of Research and Development, the EPA's National
Enforcement Investigations Center (NEIC), and a number of experts from the commercial testing
laboratory community, worked together to develop an analytical method to identify the presence
and concentration of the screening constituents (slag wool, gypsum and elements of concrete) in
indoor dust  This method was reviewed by the WTC Expert Technical Panel's signature
subcommittee and is presented in Appendix B The composition of this technical panel can be
found at http //www.epa gov/wtc/panel

A.5.3 Screening Method Study

The hypothesis that is the foundation for the WTC dust screening method is as follows  If a unit
has been impacted, those materials that are found in WTC dust (markers) will be found in the dust
collected from the unit.  The materials under consideration are. 1) slag wool, 2) elements
consistent with concrete, and 3) gypsum Since slag wool is a major component of WTC collapse
dust, if a sample does not contain 'significant' levels of this marker, the unit would not be
considered to contain WTC residuals The other markers will be used to distinguish samples
containing non-WTC slag wool from those containing WTC slag wool.  It is expected that data
from this study will  define the term 'significant level'

Five independent laboratories and three government laboratories will participate in this final
method validation phase  One government laboratory will analyze only a small portion of the
samples, but this lab was critical in the method development. Each laboratory attended  a two day
session during which the method was further developed and discussed, and procedures  to adapt the
method to suit each  laboratory's equipment were determined  Following this session, the
laboratories received dust samples consisting of both confirmed background samples (10 samples
plus duplicates) and a confirmed non-impacted dust spiked with varying amounts of confirmed
WTC dust (6 spiked samples plus duplicates) The spiked dust contains known quantities
(concentrations) of the screening materials The labs were provided the samples "blind", thus, they
did not know which samples were pure background dust, and which were the spiked dust

While the goal was to validate a method of differentiating between samples of dust that contain
residues from the WTC collapse from those that do not, since the three primary materials (slag
                                                                                      11

-------
wool, and elements of concrete and gypsum) identified above are all normally found in dusts
present in the New York area, it is possible that the proposed screen may yield some percentage of
false positive identifications of WTC dust

A6. Project/Task Description

A.6.1 Sampling
Dispersion models, photos, interviews, and satellite data were reviewed to discern areas that were
likely impacted by WTC emissions and those that were not. Samples to be used to study the
above discussed protocol were collected from within both of these areas (background and
impacted)  Samples were analyzed by EPA's NERL and USGS for content verification, and
confirmation of background or impacted status by evaluating levels of slag wool in the dust
collected by SEM.

A.6.2 Sample Preparation
WTC dust was spiked into confirmed non-impacted dust at three levels (1, 5, and 10% of total
mass) and homogenized The dusts were all characterized by the USGS and U S EPA NERL
prior to spiking  The two spiking dusts were 1) a composite sample from USGS of predominantly
outdoor dust collected in September of 2001, and 2) dust collected by the U S EPA from the
Deutsche Bank building at 4 Albany  Street in September of 2004  The 4 Albany Street building
borders the south side of the WTC complex  The USGS performed an analysis of the spiked
samples prior to the samples being sent to labs. The spiked samples showed varied levels of slag
wool, this was expected due to the difficulty in homogenizing dust containing large fibers, and the
fact that components of WTC dust will vary within a sample because of the nature of the source
Despite this variability, the measured levels were in the approximate range
expected for the spiking percent (1, 5, and 10%) and, in all but one case, each percent level was
fully distinguishable from the other in all but one case

Analysis by USGS, NEIC and NERL determined that the levels of slag wool differs between the
two WTC dusts, with the pure dust from 4 Albany  Street more than an order of magnitude lower
in slag wool than that provided by USGS (approximately 500,000 fibers/gram of dust vs
approximately 11,000,000 fibers/gram of dust, respectively)  There are likely two explanations for
this significant difference in slag wool levels. The USGS sample was a composite of multiple
outdoor samples and one indoor sample taken during September of 2001  The 4 Albany sample
was taken three years post 9/11 in September of 2004 This sample was taken exclusively inside
of a building, thus, the dust was not only diluted by three years of urban background dust, but was
also characteristic of dust that had penetrated the shell of an unopened building as opposed to that
dropping on the ground outside

A.6.3 Analytical Study
Five independent laboratories were recruited for a final test of the screening method  Each
laboratory attended a two day session during which the method was further developed and
discussed, and procedures  to adapt the method to suit each laboratory's equipment was
determined  These laboratories have received 32 of the collected dust samples consisting of both
confirmed background samples and confirmed background samples spiked with varying amounts
of confirmed WTC dust (Sample Distribution Table shown in Section B 5)  The spiked dust
contained known quantities (concentrations) of the screening materials and reasonable
homogeneity was confirmed by USGS The labs were provided the samples "blind", thus, they did

                                                                                     12

-------
not know which samples were pure background dust, and which were background dust samples
spiked with WTC dust  The labs have several weeks to analyze all dust samples. They have been
asked to provide data as to the quantity of screening materials present in the dust in a standardized
format (Appendix B). The final data from all laboratories will be evaluated to determine if they
were able to distinguish background samples from WTC spiked samples  In addition, criteria such
as time for analysis, and mtra- and interlaboratory variability will be considered when determining
validity of the method
A.6.4 Tasks and Timeframes

Sample Collection
Method/Protocol Development
Screening Study
Completion of Reports
Peer Review
September 2004-May 2005
February 2005-June 2005
June 2005-August 2005
August 2005
August 2005-September 2005
A7. Quality Objectives and Criteria

A7.1 Data Quality Objectives (DQOs)

The data quality objectives for this project are based on data acquired in the methods development
stage of this study. An inter-lab data quality objective was determined using the variability within
each dust sample for slag wool  fibers, one of the markers for WTC dust  It was determined that
data could be acquired with a relative certainty of + 35%. An mtra-lab data quality objective was
determined using the vanability within each dust sample for slag wool as well  It was determined
that data could be acquired with a relative certainty of ± 30%

A7.2 Measurement Quality Objectives (MQOs)

The accompanying tables (Tables 1 and 2) list Measurement Quality Objectives (MQOs) for this
mtralaboratory (within lab) and interlaboratory (within sample) variability. Accuracies and
precision were taken from preliminary data and manufacturer's specifications.
Measurement
Parameter
Individual dust
sample mass
Fibers/Concrete
Particles/Gypsum
Particles
Fibers
Analysis
Method
Microbalance
SEM
PLM
MQO for
Accuracy
+/- 5%
+/- 30%
+/- 30%
MQO for
Precision
+/- 5%
+/- 30%
+/- 30%
MQO for
Completeness
85%
85%
85%
   Table 1 Measurement Quality Objectives (MQOs) for mtralaboratory Vanability (within lab)
                                                                                      13

-------
Measurement
Parameter
Individual dust
sample mass
Fibers/Concrete
Particles/Gypsum
Particles
Fibers
Analysis
Method
Microbalance
SEM
PLM
MQO for
Accuracy
+/- 5%
+/- 30%
+/- 30%
MQO for
Precision
+/- 5%
+/- 30%
+/- 30%
MQO for
Completeness
85%
85%
85%
Table 2  Measurement Quality Objectives (MQOs) for Intel-laboratory Variability (within sample)
Intralaboratory MQOs will be calculated based on results within each lab for the 32 samples
Interlaboratory MQOs will be calculated based on the composite result for each lab and compared
with other labs for the 32 samples  Both sets of MQOs will be compared with the target MQOs
listed above  Accuracy will be based on how close the labs are to the calculated overall laboratory
mean for each sample or set of samples (i e each spiking percentage or background set of
samples) and precision will be based on the duplicate results within each lab (relative %
difference)
AS Special Training/Certifications

All laboratories and analysts chosen for this work will have training in both Polarized Light
Microscopy (PLM) and Scanning Electron Microscopy (SEM)

A9 Documents and Records

Documents generated (or to be generated) during this project and responsible party
    1) Sampling Access Agreement - EPA
    2) Sampling Information Sheet - EPA
    3) Sampling and Sample Handling QAPP - Lockheed Martin
    4) Analytical  Method/Protocol to be used in study - EPA
    5) Screening Study QAPP - EPA
    6) Prime Contractor Report on Screening Study - Versar
    7) EPA Report (separate from Prime Contractor Report) on Screening Study - EPA

The Screening Study QAPP will be distributed as indicated in Section A3 of this document. Dr
Rosati will distribute the QAPP to Versar and the government labs, USGS and EPA NERL as well
as EPA, ERT  Versar will be responsible for distributing the QAPP to all analytical
subcontractors (RJ Lee, MVA, MAS, EMSL and Reservoir) and EPA, ERT will be responsible for
distributing the QAPP to its sampling contractor, Lockheed Martin.

Sampling and analytical data will be reported to Dr  Rosati by EPA/ERT and Versar, respectively,
on a weekly basis  These data will be presented in a spreadsheet format  The final data shall be
presented to Dr Rosati in a report format, both electronically (including a final data spreadsheet)
and hard copy
                                                                                     14

-------
B. DATA GENERATION AND ACQUISITION

Bl. Sampling Process Design

Dispersion models, photos, interviews, and satellite data were reviewed to discern areas that were
likely impacted by WTC dust and those that were not Impacted samples were collected very
close to the WTC site, and background samples were collected from areas distinctly outside of
those that were 'likely' impacted Samples were collected from federal buildings, office buildings
and private residences on a volunteer basis A pre-sampling survey of building and sampling
areas, including photos of sampling areas (if permitted by building owners) and notes on building
usage, to identify conditions that might compromise samples (e g , smoking or cooking areas) was
developed  Additionally, an access agreement was signed by the unit occupant/owner and an
information sheet regarding the sampling was provided by the sampling contractor to the
occupant/owner (Appendix D and E, respectively)

B2. Sampling Methods

Sampling followed the approved "Generic Quality Assurance Project Plan for WTC Residue
Sampling New York City, NY, March 2005 (Appendix A)  In each building identified for
sampling, dust samples were collected from at least three areas  1) one sample from a track-in
area near a building entrance, preferably in a carpeted area, 2) two samples from relatively
undisturbed areas (e g, on top of bookcases, under furniture), and 3) other areas showing visible
accumulation of settled dust, including HVAC ducts  A standard method (REAC SOP 2040 -
Collection of Indoor Dust Samples  from Carpeted Surfaces for Chemical Analysis using a Nilfisk
GS-80 Vacuum Cleaner) using a HEP A vacuum was used by EPA to collect bulk dust samples
(Appendix C)

B3. Sample Handling and Custody

Once samples were collected, they  were sieved to 150 microns  Sieving was performed by the
method in REAC SOP 2040 - Collection of Indoor Dust Samples from Carpeted Surfaces for
Chemical Analysis using a Nilfisk GS-80 Vacuum Cleaner (Appendix C).

 Once samples were sieved, they were ashed as described in Protocol for Preparation and Analysis
of Residential and Office Space Dust by  Polarized Light Microscopy and Scanning Electron
Microscopy with Energy Dispersive X-Ray Spectroscopy, May 18, 2005 (Appendix B)  0 25 mg
of each ashed sample was archived Ashed samples were disseminated as instructed by EPA (Dr.
Rosati)  As not all collected  samples were used in the study, remaining samples will be sealed and
stored in a limited access area  All samples are accompanied by cham-of-custody forms  To
ensure that these important samples are properly collected, tracked, stored, and  distributed, quality
assurance (QA) procedures were in place prior to any sample collection (Appendix C)

B3.1 Homogenization

Homogemzation of spiked samples was performed by the USGS Standards Laboratory after
ashing occurred   First, the background material was transferred to a 16 ounce glass  container and
an  expanded metal mixing card inserted   The container was sealed, and placed on a roller mixer
where the contents were mixed for a total of eight hours   Each spiking material was then
                                                                                     15

-------
transferred to individual four ounce glass containers.  A laminated customized paper mixing card
was inserted and the container was reseated and placed on a horizontal roller mixer  Once pre-
mixing of the two dust types was complete, three concentrations of spiking material were prepared
according to Table 3 as follows
Mass Target
Cone Wt. %
1
5
10
background
material, g
29.7
285
270
spiking
material, g
0.3
1 5
30
                       Table 3: Mass concentrations of spiked samples
The appropnate amount of background and spiking material were weighed into plastic weighing
boats and then transferred to pre-labeled four ounce glass bottles.  A laminated customized paper
mixing card was inserted and the container sealed and placed on a horizontal roller mixer  The
samples were blended for a total often hours After blending a total of six ahquots, ~0 5g was
removed from each container using a spatula and transferred to individual one ounce vials  The
six samples from each concentration underwent SEM analysis at the USGS to assure reasonable
sample homogeneity has been accomplished. The samples were then shipped to EPA, ERT for
archiving and distribution

B4. Analytical Methods

The USGS, the EPA's National Exposure Research Laboratory (NERL) and EPA's National
Enforcement Investigations Center (NEIC) developed a method to screen for the three key
materials slag wool, elements of concrete, and gypsum  This method involves the use of
polarized light microscopy (PLM) or scanning electron microscopy (SEM) to determine the
quantity of each of the materials present (Appendix B)

Data will be reported on the standardized sheets in the appendix of this  protocol   Data to be
reported includes
   •   Slag wool (fibers/gram of dust) and length/width
   •   Elements of concrete (area %)
   •   Gypsum (area %)

B5. Quality Control

Quality control is addressed in Section 10 0 of the standard protocol in Appendix B entitled
"Protocol for Preparation and Analysis of Residential and Office Space Dust by Polarized Light
Microscopy and Scanning Electron Microscopy with Energy Dispersive X-Ray Spectroscopy"
In addition to the items referred to in this section, several measures have been taken to ensure the
quality of this study
                                                                                      16

-------
   •   Standard protocol - a standardized protocol will be used for all sample preparation and
       analysis. This was determined to be a necessity during the scoping part of this work.  This
       standardized protocol should help to minimize interlaboratory variability

   •   Duplicates- duplicates of all samples were provided for analysis  Analysis of these
       duplicates will allow us to determine intralaboratory variability when using a standardized
       protocol

   •   Blind samples - all labs received the same 32 samples, and all samples were coded so that
       this was a 'blind'  study, no lab knows what they are analyzing nor will they be able to
       compare their results with other laboratories

Sample Distribution Table (note, each of the eight laboratories has been given a letter A-H)
Sample
Designations
AP5(1)
AP5(2)
CMC(1)
CMC(2)
HS3(1)
HS3(2)
WGSd)
WGS(2)
MW(1)
MW(2)
DB1%(1)
DB1%(2)
DB5%(1)
DB5%(2)
DB10%(1)
DB10%(2)
C1-RTP(1)
C1-RTP(2)
USGS1%(1)
USGS1%(2)
USGS5%(1)
USGS5%(2)
USGS10%(1)
USGS10%(2)
USC(1)
USC(2)
FP(1)
FP(2)
MUNYdd)
MUNYCK2)
Laboratory Letter Codes
A






























B






























C






























D






























E






























F






























G






























H






























                                                                                         17

-------
MUNYC2(1)
MUNYC2(2)
















Samples spiked with WTC dust, at 1, 5, and 10% levels are shaded.  All others are
background samples. Total of 20 background samples (10 samples + 10 duplicates) and
twelve spiked samples (6 samples + 6 duplicates)
 B6. Instrument/Equipment Testing, Inspection and Maintenance

    Vacuum cleaners used for sampling were maintained as described in Appendix A  Generic
    Quality Assurance Project Plan for WTC Residue Sampling New York City, NY, March 2005.

 B7. Instrument/Equipment Calibration and Frequency

 All microbalances used in this study shall calibrated annually  Scanning Electron Microscopes
 (EDS system) shall be calibrated on a daily basis as discussed in Appendix B entitled "Protocol
 for Preparation and Analysis of Residential and Office Space Dust by Polarized Light Microscopy
 and Scanning Electron Microscopy with Energy Dispersive X-Ray Spectroscopy, Section 10 1

 B8. Inspection/Acceptance of Supplies  and Consumables

 Inspection and acceptance of all consumables used during sampling will be performed as
 described in Appendix A Generic Quality Assurance Project Plan for WTC Residue Sampling
 New York City, NY, March 2005

 Inspection and acceptance of all consumables used during sample analysis will be performed by
 the analytical and government laboratones  Consumables are listed under apparatus and materials
 in Appendix  B entitled "Protocol for Preparation and Analysis of Residential and Office Space
 Dust by Polarized Light Microscopy and Scanning Electron Microscopy with Energy Dispersive
 X-Ray Spectroscopy.
 B9. Non-Direct Measurements

 Not applicable

 BIO. Data Management

 Data from this study will be reported by the subcontractors and government labs to the prime
 contractor on standardized data sheets found in Appendix B "Protocol for Preparation and
 Analysis of Residential and Office Space Dust by Polarized Light Microscopy and Scanning
 Electron Microscopy with Energy Dispersive X-Ray Spectroscopy, May 18, 2005." Data will be
 compiled and analyzed by the prime contractor in a electronic spreadsheet, and a report of this
 data will be written
                                                                                     18

-------
C. ASSESSMENT AND OVERSIGHT

   •   All work will be overseen by the principal investigators of this study. Weekly conference
       calls will be held with EPA, the prime contractor and all subcontractors to assess progress
       and discuss any issues or problems that rtiay have arisen Data is to be submitted to the
       prime contractor by the subcontractors on a weekly basis as this data is obtained  Due to
       the rapid nature of this study, no interim reports are required.

   •   All stakeholders (EPA, USGS and contractors) will be provided a copy of the QAPP and
       will review it for correctness

   •   All sampling and sample preparation performed by Lockheed Martin will be under direct
       oversight of quality assurance personnel and audits will be conducted as noted in Appendix
       A (Sampling QAPP)

   •   All contracting laboratories are required to employ standard QA practices and all  work
       should be performed under the oversight of in-house quality assurance personnel

   •   All data will undergo evaluation and review by the prime contractor prior to being
       assembled into a report to the EPA  The EPA will perform its own assessment and
       evaluation of the study data and issues, and will assemble this information into an overall
       EPA report

   •   The study will undergo a formal EPA peer review once it has been completed  Peer
       reviewers will be provided all data and reports, and will be given 6 weeks to perform a full
       evaluation of the study  In addition, all data and reports will also be provided to the WTC
       Technical Panel for their review and comments

D. DATA VALIDATION AND USABILITY

       •  All data shall be provided by the subcontractors to the prime contractor on the
           spreadsheets  provided in Appendix B "Protocol for Preparation and Analysis  of
           Residential and Office Space Dust by Polarized Light Microscopy and Scanning
           Electron Microscopy with Energy Dispersive X-Ray Spectroscopy, May 18, 2005.",
           Section 15 0
       •   The prime contractor will review and verify the data before compiling it into a report
       •   EPA will evaluate the prime contractor report, along with the compiled data to
           determine'
              o  whether the MQO's presented in Table I of this QAPP were met
              o  whether the study described herein demonstrated the following
                     •   that slag wool, gypsum  and elements of concrete are reasonable markers
                        for WTC dust (by showing that these markers distinguish WTC-laden
                        dust from background dust),
                     •   that WTC dust at a diluted concentration can be distinguished from
                        background, and
                     •   that the  analytical method works well enough, and is able to be earned
                        out by enough analytical laboratories to  1) evaluate the  above materials
                        as markers and 2) distinguish WTC dust from background dust
                                                                                       19

-------
      •  EPA will prepare a final report documenting all data, analysis and conclusions based
         on the above evaluation
APPENDICES

   A Genenc Quality Assurance Project Plan for WTC Residue Sampling New York City, NY,
      March 2005
   B Protocol for Preparation and Analysis of Residential and Office Space Dust by Polarized
      Light Microscopy and Scanning Electron Microscopy with Energy Dispersive X-Ray
      Spectroscopy, May 18, 2005
   C REAC SOP 2040 - Collection of Indoor Dust Samples from Carpeted Surfaces for
      Chemical Analysis using aNilfisk GS-80 Vacuum Cleaner
   D Access Agreement
   E Information Sheet

REFERENCES

Lowers, H A, G.P Meeker, and IK Brownfield  (2005a)  Analysis of Background Residential
Dust for World Trade Center Signature Components Using Scanning Electron Microscopy and X-
ray Microanalysis  U S Geological Survey Open File Report 2005-1073

Lowers, H A, Meeker, G P , I K Brownfield (2005b) World Trade Center Dust Particle Atlas.
U S  Geological Survey Open-File Report 2005-1165. http //Eubs.usgs^Qv/of/2Q05/].165/

Meeker, G P , A.M Bern, H A Lowers, and I K. Brownfield (2005)  Determination of a
Diagnostic Signature for World Trade Center Dust using Scanning Electron Microscopy Point
Counting Techniques  U S Geological Survey Open File Report 2005-1031.

NYCDOHMH/ATSDR  (2002) New York Department of Health and Mental Hygiene and
Agency for Toxic Substances and Disease Registry. Final  Technical Report of the Public Health
Investigation To Assess Potential Exposures to Airborne and Settled Surface Dust in Residential
Areas of Lower Manhattan  Agency for Toxic Substances  and Disease Registry, US Department
of Health and Human Services, Atlanta, GA.

R J Lee (2004)  Signature Assessment 130 Liberty Street  Property Expert Report WTC Dust
Signature  Prepared for  Deutsche Bank  May, 2004 R J Lee Group, Inc  350 Hochberg Road,
Monroeville, PA   15146
                                                                                   20

-------
           APPENDIX A: SAMPLING QAPP
          Al. GENERIC QUALITY ASSURANCE PROJECT PLAN FOR
       WORLD TRADE CENTER (WTC) RESIDUE SAMPLING NEW
                         YORK CITY. NEW YORK
                U.S. EPA Work Assignment No 0-089
              Lockheed Martin Work Order No • EACOO089
                 U S  EPA Contract No • EP-C-04-032
                          Prepared For
United States Environmental Protection Agency/Environmental Response Team
                           Edison, NJ
                           March 2005
                                                             21

-------
                      A2. TABLE OF CONTENTS


A    PROJECT MANAGEMENT
     Al    TITLE PAGE          .        	                . .  1
     A2    TABLE OF CONTENTS  .       	     .      ...   2
     A3    DISTRIBUTION LIST	                .  .     3
     A4    PROJECT ORGANIZATION      ....                .3
     A5    PROBLEM DEFINITION   .               	       3
     A6    PROJECT DESCRIPTION AND SCHEDULE      ...     .      .4
     A7    DATA QUALITY OBJECTIVES AND CRITERIA FOR MEASUREMENT
          OF DATA       	     	4
     A8    TRAINING AND CERTIFICATION      ..        	4
     A9    DOCUMENTS AND RECORDS    	         ...  5

B    DATA GENERATION AND ACQUISITION  	   5
     Bl    SAMPLING PLAN DESIGN        .       .   .     ....        5
     B2    SAMPLING METHODS .                              .       6
     B3    SAMPLE HANDLING AND CUSTODY       	       .      6
     B4    ANALYTICAL METHODS ....                    ..6
     B5    QUALITY CONTROL ...                           .7
     B6    INSTRUMENT/EQUIPMENT TESTING, INSPECTION AND
     MAINTENANCE   	             ...          	  7
     B7    INSTRUMENT/EQUIPMENT CALIBRATION AND FREQUENCY .   ...  7
     BS    INSPECTION/ACCEPTANCE OF SUPPLIES AND CONSUMABLES   .   7
     B9,    NON-DIRECT MEASUREMENTS	7
     BIO   DATA MANAGEMENT	     ..      ..    .7

C    ASSESSMENT/OVERSIGHT  ..   .           	    8
     Cl    ASSESSMENT AND RESPONSE ACTIONS	      	8
     C2    REPORTS TO MANAGEMENT.        .        .               8

D    DATA VALIDATION AND US ABILITY      .    .     .     .      .9
     DI    DATA REVIEW, VERIFICATION AND VALIDATION     .      .     9
     D2    VERIFICATION AND VALIDATION METHODS             .   .    9
     D3    RECONCILIATION WITH USER REQUIREMENTS ..    .            9

REFERENCES    	    9


TABLE 1 - FIELD SAMPLING SUMMARY ...        . .        .     .      10
0089-DQAPP-031805
                                                                22

-------
A     PROJECT MANAGEMENT

This project generic Quality  Assurance Project Plan (QAPP) was prepared in accordance with
EPA Requirements for Quality Assurance Project Plans (QAPPs), EPA QA/R-5 and the Response
Engineering and Analytical Contract (REAC) Program QAPP,

A3     DISTRIBUTION LIST

The following personnel will receive copies of the approved QAPP for the World Trade Center
(WTC) Residue Sampling, Work Assignment (WA) No 0-089

       1.     Rajeshmal  Smghvi, Environmental Protection Agency/Environmental Response
             Team (EPA/ERT)
             Work Assignment Manager (WAM)
       2.     Jacky Rosati, EPA, Research Triangle Park North Carolina (NC)
       3     Eletha Brady-Roberts, National Homeland Security Research Center (NHSRC)
       4     Jeffrey Bradstreet, REAC Air Response Section Leader/Task Leader
             (TL)/Quality Control (QC)
             Coordinator
       5     Deborah Killeen, REAC Quality Assurance Officer (QAO)
       6     Dennis Miller, REAC Program Manager

A4     PROJECT ORGANIZATION

The following individuals will participate in the project

             EPA/ERT
             Rajeshmal Singhvi -WAM
             Jacky Rosati - Project Coordinator
             Jeff Catanzanta - Technical Auditor
             Eletha Brady-Roberts - NHRSC Quality Assurance (QA)
             REAC
             Jeffrey Bradstreet -TL/QC Coordinator
             Miguel Trespalacios - Senior Air Sampling Scientist
             Michael Hoppe - Environmental Scientist/Sampler
             TBD - Environmental Scientists/Samplers
             Richard Magan - Technician/Sampler
             Deborah Killeen - QAO

 Laboratories that will receive residue samples for chemical marker/signature identification on
 this project include1

             Department of Environmental Sciences and Engineering, University of
             North Carolina, United States Geological Survey (USGS) Denver
             Microbeam Laboratory, and Other commercial laboratones to be
             determined

 The REAC TL/QC Coordinator for the project is the pnmary point of contact with the EPA/ERT
 WAM. The XL is responsible for the completion of the Work Plan (WP) and QAPP, project
 team organization, and supervision of all project tasks, including reporting and deliverables The
 EPA NHSRC will provide oversight and guidance in the field through the WAM

 A5,    PROBLEM DEFINITION
                                                                                 23

-------
The World Trade Center (WTC) attacks on 11 September 2001 caused the airborne release of two
types of dusts  those related to the building collapse and fine participate matter from the subsequent
fires There are concerns among residents of New York City (NYC)about the potential health
effects of WTC dusts that might remain in buildings in NYC. The goal of this study is to collect
dust samples from areas near the WTC and distant from the WTC (background NYC dusts)
These dust samples will be used to validate chemical markers or signatures for WTC dust - as
compared to background dust - for a larger sampling effort to identify indoor areas still
contaminated with WTC dust. The markers or signatures for WTC dust are being developed by
laboratories at EPA, USGS, and several universities  By sampling a number of contaminated and
uncontammated sites and by utilizing recently collected samples, the WTC signatures can be
validated or improved for the larger sampling study to delineate contaminated areas

The residue from the collapse of the WTC  Towers may contain heavy metals (primarily lead,
arsenic, and mercury), polynuclear aromatic hydrocarbons (PAHs), and other contaminants that
EPA designated laboratories are investigating to associate specific analytes with the WTC.

A6.    PROJECT DESCRIPTION AND SCHEDULE

The purpose of this project is to collect dust samples from designated buildings that may be from
the collapse of the WTC towers  or from background sites for  comparison EPA will identify
contaminated  and uncontammated buildings in NYC and obtain access for REAC personnel to
conduct sampling  To the extent possible, contaminated buildings  that are within both the dust
and the fire plumes will be selected, so that current dust  samples for these two types of emissions
can be collected  EPA may identify several groups of buildings over time, as permission for access is
obtained  EPA will provide REAC personnel with the  street address  and  the name and phone
number for a contact person in each building identified for surveying

This activity will involve conducting scoping surveys of buildings  identified by the EPA in the
NYC  Area,  preparing sampling plans, collecting samples,  splitting samples  for multiple
laboratories, shipping samples, and archiving samples for up  to two  years for future analysis and
report preparation

The schedule of activities and reports is as follows

             WP                                    6 October 2004
             Draft Generic QAPP                     6 October 2004
             Final Generic QAPP                     18 March 2005
             Draft Building Survey Form               15 October 2004
      •      Collect Residue Samples                  As scheduled by EPA
      •      Prepare and Send Sample Aliquots         4 Days after sampling
      •      final Report                             10 Days after sampling

A7     DATA QUAIJTY OBJECTIVES AND CRITERIA FOR MEASUREMENT OF
DATA

The focus  of this project is to collect dust samples that may be  contaminated with materials from
the destruction of the WTC towers or are potentially uncontaminated  This QAPP covers the
collection, storage and  shipment of the samples to EPA designated laboratories  The specific
chemical markers or signatures associated with WTC dust samples  are investigatory and will be
used to further define the project Once  defined, the specific chemical markers or signatures will
be used to further define the project

A8     TRAINING AND CERTIFICATION


                                                                                   24

-------
The training of all field personnel involved with sampling activities is intrinsic to their position
and required responsibilities  They will have the following documented training

       n     Occupational Safety and Health Administration (OSHA) 40-hour and 8-hour
       refresher in Hazardous
             Waste Operations (20 CFR1910 120)
                                                                                     25

-------
       •      Department of Transportation (DOT) hazardous materials shipping
             First Aid and Cardiopulmonary Resuscitation (CPR) training

A9     DOCUMENTS AND RECORDS

The RE AC Program QAPP serves as the basis for this generic QAPP The most current approved
version is available to all REAC technical personnel as an uncontrolled copy on the REAC Local
Area Network (LAN) Documents and records that will be generated during this project include.

             WP
             Draft Generic QAPP
             Final Generic QAPP
             Field logbooks
             Site maps
             Photos of Sampling Locations
             Chain of Custody forms
             Final Reports

The Final  Report will provide a description of the project, field procedures, sample preparation
procedures, difficulties encountered  and will include validated final copies of chain of custody
forms. All documentation  will be recorded  in accordance  with REAC standard operating
procedure  (SOP)  #2002, Sample  Documentation   and  REAC  SOP   #4001,  Logbook
Documentation  The final report will be prepared using REAC SOP #4021, Preparation of Final
Reports

B,     DATA GENERATION AND ACQUISITION

Bl     SAMPLING PLAN DESIGN

Judgmental sampling will be used to select sample locations that are most likely to represent WTC
residue or background dust  This will be based on historical information, visual inspection and best
professional judgment of the WAM  and sampling team. This type of sampling is used to identify
contaminants present  in areas potentially having the highest concentration of contaminants.
Additional samples may be collected when requested by the WAM and EPA NHSRC personnel

During the sampling of EPA specified buildings, dust samples will be collected from each of the
buildings up to 20 in accordance with  the EPA approved generic QAPP.  Sampling will likely be
performed at two types of areas in each building  a high traffic area (to characterize tracked-in  dust)
and a lowtraffic area (to represent settled indoor  dust)  Two low traffic areas will be specified  for a
total of three areas that will be sampled. The desired high traffic area is to be an area near a  main
entrance, preferably carpeted  The desired low-traffic areas include areas infrequently cleaned,
such as the top of elevator housing, under refrigerators, behind file cabinets, above ceiling tiles, on
high shelves, or m other areas that show visible  dust accumulation and are infrequently disturbed
Sampling  will not be restricted to  carpeted areas as the intent of the sampling is to obtain the
desired residue. If vacuum  sampling is not  possible or preferred, sweep sampling will be  used to
collect the residue  Sampling will not be conducted m areas that would likely contain chemicals in
dusts that would interfere with the  analysis for the WTC markers  The  following areas  will be
avoided m the sampling effort

       •     Areas with significant cigarette or cigar smoke, incense, or
             burning candles
             Areas near major outdoor combustion sources (e g, power plants)

Due to the inability to obtain triplicate samples (once an area is sampled, little residual remains),
three samples will be collected in the same general area, for a total of nine samples from a building
(i e, three sample areas times three samples in each general area). The proximity of the samples in

                                                                                     26

-------
each general area will be determined as a result of visual inspection in the field and discussions
with the WAM
                                                                                           27

-------
B2      SAMPLING METHODS

Vacuum sampling will be performed in accordance with modified REAC SOP #2040, Collection of
Indoor Dust Samples From Carpeted Surfaces for Chemical Analysis Using a Nilfisk GS-80
Vacuum Cleaner This method may afford collection of samples large enough for analysis of both
purported organic and inorganic signatures Although the method specifies the size and shape of the
areas to be sampled and the mass to be collected, the sample collection procedure will vary to
accommodate the site-specific conditions and ensure that an adequate sample is obtained If it is
not feasible to use the vacuum method of sampling, samples will be collected in bulk by sweeping
the residue into a pan or sample bag in accordance with  modified ERT/REAC SOP #2011 Chip,
Wipe and Sweep Sampling The sample handling and data collection requirements specified in
modified REAC SOP #2040 will be followed

The area to be sampled is not measured before sampling, but after the sample is collected  This is a
modification of both REAC SOP #2040  and  ERT/REAC #2011.  REAC SOP #2040 is further
modified in that samples will also be collected  from non-carpeted surfaces, the amount of sample
collected will be visibly checked and dust weight calculations will not be performed Sweep
sampling utilizes a dedicated, hand held sweeper brush to acquire the sample from an area The
area sampled is measured after sampling

Sample Volume,  Container, Preservation  and Holding  Time. The collected samples are
placed into  appropriately sized  glass jars or zip-lock plastic bags  Storage of the samples
collected by sweep  or vacuum are maintained in a refrigerated unit at 4 ± 2 degrees Celsius (°C)
after sieving

Sampling Equipment Decontamination. The nozzles, wands and hoses are decontaminated
after use with a bottle brush, to remove any accumulated dust in the hose and nozzle When the
nozzle is clean, it is removed and sprayed with reagent grade methanol and allowed to air dry on a
clean  surface The wand and hose are then cleaned with the bottle brush To continue a new
polylmer and collection bag for the collection of another sample is installed

B3     SAMPLE HANDLING AND CUSTODY

In the field, sampling data are recorded on a Vacuum Sampling Work Sheet or in a dedicated
project logbook Cham of custody (COC) records will be used to document the collection of dust
samples by vacuum or bulk.  All COC records will receive a peer review in the  field prior to
shipment of the samples in accordance with REAC SOP #4005, Chain of Custody Procedures

All samples will be  delivered to the REAC facility and sieved in accordance with modified REAC
SOP 2040, Collection of Indoor Dust Samples From Carpeted Surfaces for Chemical Analysis
Using a Nilfisk GS-80 Vacuum Cleaner and modified ERT/REAC SOP #2011 Chip, Wipe and
Sweep Sampling The samples will be sieved through a No 100 sieve (150 microns [firnl)  After
sieving, the samples will either be transferred to jars, which will  be placed into Ziplock* storage
bags,  or directly into Ziplock™ storage bags, and then placed into a holding refrigerator with the
corresponding COC record.

Scnbe* spreadsheet formats will be used for sample management  REAC is required by contract to
use Scnbe" to track and  log the samples  In addition a unique sample numbering system has been
established to each sample, which identifies the site identification,  event number and  the sample
number Additional information is provided with the sample number to identify  whether it is a
sieved (S) or coarse (C) fraction and the weight of the fraction in grams, e g, SO. 1

The samples collected by REAC personnel  will  be shipped to the  designated laboratory for
analysis in accordance with REAC SOP #2004, Sample Packaging and Shipment One of the four
aliquots of each sample will be retained and stored by REAC staff for up to two years  in a secure
refrigerator,


                                                                                   28

-------
B4     ANALYTICAL METHODS
Once specific chemical markers or signatures have been defined, EPA personnel in consultation
with NHSRC will be able to determine which analyses will be appropriate  The laboratories
specified under Section A4 are conducting the investigatory work

B5     QUALITY CONTROL

This QAPP covers the collection, storage and shipment of the samples to EPA designated
laboratories for analysis Quality control for the field and storage procedures are as follows-

       •     Field documentation on Field Sampling Worksheets or in logbooks
             Documentation of temperature for the dedicated secure refrigerator

Duplicate samples will not  be taken due to the nature of the sampling method Once an area is
vacuumed,  little residual sample remains Quality control for the laboratory procedures will be
specified by the EPA/NHSRC

B6.     INSTRUMENT/EQUIPMENT TESTING, INSPECTION AND MAINTENANCE

The Nilfisk vacuums used in the collection  of the residue samples will be maintained  in
accordance with established specifications On a quarterly basis, the parts  of the Nilfisk vacuum
cleaners are inspected for cracks and breaks  An inventory of available  supplies is conducted
every three months.

B7.   INSTRUMENT/EQUIPMENT   CALIBRATION  AND

FREQUENCY  The instrument/equipment calibration frequency is

not applicable to this QAPP

B8         INSPECTION/ACCEPTANCE OF SUPPLIES AND

CONSUMABLES

REAC personnel are responsible for the procurement, inspection, and acceptance of supplies and
consumables for this WA. The vacuum cleaner filters purchased by REAC personnel must meet
the  requirements specified  by  the  manufacturer  The REAC TL and Group Leaders are
responsible for ensuring that the correct filters  and sampling bags are specified in the purchase
orders and  verifying upon receipt that the correct parts have been shipped  It is the responsibility
of the EP AVERT to provide adequate facilities, equipment and supplies for REAC to perform all
field related tasks for this WA

B9     NON-DIRECT

MEASUREMENTS This section is not

applicable to this QAPP BIO

DATA MANAGEMENT

The QAPP is identified by  the footer located on the bottom left hand comer of the page The file
identification represents the structure and the filename  The filename starts with the 3-digit WA
number preceded by a "zero", then the deliverable type (D  or N) to identify the document as a
deliverable or non-deliverable followed by the document type For amended or revised documents,
the letters "A" and "R" for  amended and revised, respectively, and the appropriate amendment or
                                                                                  29

-------
revision number (eg  1.2,3  ) are added after the document type  After the document type and
revision/amendment code (if any), a six-digit code based on the month, day and year (mmddyy) is
added to indicate the date the document was delivered to the client.

Field sampling data will initially be recorded on field data sheets and in field notebooks Samples
will be identified by the field assigned sample number  Paper versions of all dehverables (Work
Plan, Generic QAPP and Final Reports) will be provided to the ERT WAM and stored in the
REAC Central Files Electronic versions of ail dehverables will be saved on the REAC archive
drive in accordance with Administrative Procedures (AP) #34, Archiving Electronic Files All data
dehverables for this WA will be posted to the ERT-Information Management System (IMS) web
site as either a Scribe* electronic data deliverable (EDO) or in portable document format (pfd)
Submission of the deliverable to the appropriate ERT-IMS website will be considered delivery to
the WAM as of the date and time such dehverables are received on the website

Field log books will also be archived once the project is completed and the Work Assignment 0-089
is closed All SOPs referenced in this QAPP are available on the REAC LAN.

C.     ASSESSMENT/OVERSIGHT

Cl.     ASSESSMENT AND RESPONSE ACTIONS

The REAC TL, Air Response Section Leader, QAO and QC Coordinator are responsible for QC
assessments and corrective action for this WA  These personnel have the authority to issue stop
work orders  The tasks associated with this  QAPP are assessed through the use of peer reviews,
technical reviews and/or technical system audits, and management system reviews Peer review
enables  the  reviewers to identify  and correct reporting errors before  reports are  submitted
Technical reviews are conducted by those immediately responsible for overseeing or performing the
work (self-assessments) An independent assessment or  technical audit will be performed by Jeff
Catanzanta  Management system reviews  establish compliance  with prevailing management
structure, policies and procedures, and ensures that the required data are obtained

Peer reviews are conducted on project dehverables to ensure a technical review with respect to
content, completeness and the overall quality of the deliverable prior to submittal to the EPA/ERT
The responsibilities of the review team and  the sequence in which the deliverable is reviewed, is
outlined in REAC AP #22, Peer Review of REAC Deliverable* The REAC QAO will audit data
dehverables on a biannual basis to determine compliance with the peer review procedures

The EPA/ERT WAM  for this task will be present and will have the responsibility for verifying that
the proper SOPs and sampling procedures are followed  If any technical issues or deficiencies are
identified, they will be reported to the REAC TL for immediate resolution or corrective action
Any changes in scope of work will be documented on a Field Change Form and approved by the
WAM

C2     REPORTS TO MANAGEMENT

Monthly technical reports will be prepared for this WA when hours have been charged on a monthly
basis. These reports will detail the accomplishments for the past month, any problems encountered,
solutions to rectify the problem, contacts and meetings, goals for the next month, and an estimate
of the of the total labor hours and costs for the next reporting period  The monthly technical
reports are submitted to the EPA/ERT Project Officer and WAM

On a quarterly basis, the REAC QAO provides a report to the REAC Program manager and the
ERT QA Manager that summarizes the quality assurance (QA) activities on a quarterly penod
These reports include results of performance evaluation samples, system audits (internal  and
external),  summary of non-cpnformance  and corrective  actions,  preparation  of SOPs for
analytical and operational activities, training, contacts/meetings and other QA activities


                                                                                    30

-------
REAC Report
Monthly Progress
Quarterly Q A Reports
Recipients
EPA/ERT Project Officer and WAM
EPA/ERT Project Officer and WAM
31

-------
D     DATA VALIDATION AND USABILITY

Dl     DATA REVIEW, VERIFICATION AND VALIDATION

For field activities, it is necessary to determine  whether the samples  were collected using the
sampling design specified in  element Bl, whether the samples were  collected  according to a
specific method or SOP as specified in element B2, and whether the collected samples have been
recorded and handled properly as in element B3  Field sampling worksheets and field notes will
be reviewed by the RE AC TL for completeness The COC records will be reviewed to ensure that
the field information has been accurately reflected on the COC records

D2     VERIFICATION AND VALIDATION METHODS

Verification occurs at eacii level in the field to ensure that appropnate outputs are being generated
routinely Records produced  electronically or maintained as hard  copies are subject to data
verification During field activities, records associated with sample collection such as field data
sheets, COC records, logbook documentation, or electronic devices to  log samples are verified
Naming conventions for the initial samples and  samples fractions produced during sieving are
verified by the RE AC TL  Chain  of custody records are verified along with refrigerator and
freezer logs to ensure the integrity of the samples

There is no analytical data  being generated under this WA, therefore, procedures for verifying
and validating data, including the chain of custody for data throughout the life cycle is not
applicable

D3      RECONCILIATION WITH USER REQUIREMENTS

Responsibility lies with the EPA, thus, this element is not applicable to this QAPP

REFERENCES

Response Engineering and  Analytical Contract  2003  Quality Assurance Project Plan for the
Response, Engineering, and Analytical Contract, Revision 0 0

U S Environmental Protection Agency  1990, Quality Assurance/Quality Control Guidance for
Removal Activities, EPA/540/G-9/004, Office of Emergency and Remedial Response

U S. Environmental Protection Agency 2001  EPA Requirements for Quality Assurance Project
Plans (QAPPs), EPA/240/B -01/003, Office of Environmental Information.
                                                                                   32

-------
    TABLE I  Field Sampling Summary
World Trade Center (WTC) Residue Sampling
              March 2005
Analybca
1
Parameter
Dust/Settled Partculate
Dust/Settled Paniculate
Sampling
Method
Nilfisk GS-80
Vacuum
Cleaner
Sweep
Preservation
Up to 2 years at 4
degrees C +1-2 degrees
C
Up to 2 years at 4
degrees C +/-7 degrees C
Total Samples
Up to 9 per Building
Up to 9 per Building
Maximum
Number
Samples
9
9
                                                                 33

-------
                         APPENDIX B: PROTOCOL
  Protocol for Preparation and Analysis of Residential and Office Space Dust by
   Polarized Light Microscopy and Scanning Electron Microscopy with Energy
                       Dispersive X-Ray Spectroscopy

                               June 27, 2005
                               Prepared by:
                    U.S. Environmental Protection Agency
    National Enforcement Investigations Center/ National Exposure Research
            Laboratory/National Homeland Security Research Center
                 Denver, CO and Research Triangle Park, NC
The use of trade names does not imply endorsement and are used for illustrative purposes only.

-------
                                     Contents
1 0    Purpose   	            	  36
2.0    Scope/Application    ...      ....        	         ....    36
  21     Limitations of the Method and Future Considerations      	36
3 0    Definitions   ...       . .          	       • •        36
40    Summary of Method   	     ..       	     	       -37
5 0    Interferences    	     	     	        	  37
6.0    Safety  	   37
70    Apparatus and Materials	      	38
8.0    Reagents  	39
9.0    Sample Storage  	39
100   Quality Control   	    39
   101   Calibration	        	        	    39
110   Procedure	        	       ...           .   . 40
   111   Weighing and Splitting  .    .        .       ...              ...          40
   11 2   Ashing       . .       .    ..      	        40
   113   Sieving	      .       ..         -             ...    41
   11 4   Preparation of Sample for Polarized Light Microscopy...                 ..    41
   114   Preparation of Sample for SEM Analysis  ..       	            ..   41
120   Analysis  ...        	       -     -    -        •     9
   121   Analysis by Polarized Light Microscopy	      9
   122   Analysis by SEM/EDS  	      	     -10
     12.21   Screening for Slag Wool                                                10
     12.22   EDX Screening for Gypsum/Anhydnte              .                      11
     1223   X-Ray Mapping for Gypsum                                            11
     1224   X-Ray Mapping for Ca-nch Particles                                     12
     1225   Particle Analysis for Gypsum and Concrete          .                    .13
 130   Data Analysis and Calculations	      ....       ..     .       .14
 140     References                                                                15
 150    Appendix                                                                  16
                                                                                  35

-------
1.0   Purpose

      This document describes sample preparation and analytical screening procedures for bulk
      samples of dust collected from residential and commercial office environments.  These
      methods are collectively referred to as the protocol

2.0   Scope/Application

      The protocol describes polarized light microscopy (PLM) and scanning electron
      microscopy (SEM) with energy dispersive spectrometry (EDS) to screen bulk dust
      samples for mineral slag wool, particles consistent with concrete compositions, and
      gypsum. The analysis methods include operating parameters and particle identification
      cntena

      2.1    Limitations of the Method and Future Considerations

             This protocol provides a means of analyzing for particles consistent with those
             found in dust present after the collapse of the World Trade Center (WTC) in New
             York City  Components of WTC Dust have been documented and catalogued by
             the U S  Geological Survey Denver Microbeam Facility and the images and
             characteristics shall be used in identification of particles (1)

             The x-ray mapping procedure in sections 1223 and 1224 and the calculations
             presented in section 130 only determine the maximum percentage of non-
             gypsum, calcium-nch particles, which may include non-concrete materials  The
             particle analysis procedure presented in section 12 2 5 is the preferred procedure
             for determining the percentages  of gypsum and concrete particles in the sample

             The x-ray mapping and image analysis procedure relies heavily on the thresholds
             for backscattered electron images  Binary (particles white and background black)
             backscattered electron images (BE1) should be used to reduce errors in setting
             thresholds in Photoshop

3.0   Definitions

      1   PLM - Polarized Light Microscopy
      2   SEM - Scanning Electron Microscope
      3   EDS - Energy Dispersive Spectrometry
      4   SEI - Secondary Electron Image
      5   BEI - Backscattered Electron Image
      6   Mineral Wool - lightweight vitreous fibrous material composed of rock wool and slag
          wool and used especially for heat and sound insulation
      7   Rock Wool - a man-made vitreous fiber (MMVF) component of mineral wool
          containing magnesium, aluminum, silicon, and calcium  Sodium and potassium may
          also be present Iron oxide is typically 3-12% by weight
      8   Slag Wool - a man-made vitreous fiber (MMVF) component of mineral wool
                                                                                    36

-------
          containing magnesium, aluminum, silicon, and calcium. Sodium and potassium may
          also be present. Iron oxide is typically less than 2% by weight
       9.  HEPA - High-Efficiency-Particulate-Air Filter

4.0    Summary of Method

       1  Weigh sample to nearest 0.0005 g
       2  Split the sample, archive half and keep half for analysis
       3  Ash half of the sample for analysis
       4.  Sieve the ashed sample to 1 SO um.
       5  Split the <150 um ashed  portion Archive three quarters of the sample Keep one
          quarter for PLM and SEM/EDS analysis
       6  Weigh the quarter and place it in enough isopropanol to get a 10-20 mg per mL
          dilution. Apply an aliquot to a glass slide, let dry, and add 1  55 (or 1 605) refractive
          index oil  Analyze by PLM for mineral wool.
       7  Prepare a sample for SEM/EDS analysis using the same dilution prepared for PLM
       8  Apply an aliquot of the sample to an aluminum sample stub with a carbon adhesive
          tab covered by a piece of polycarbonate filter (13-mm diameter or punched out of a
          larger filter to fit the size of the stub)
       9  Identify fibers by EDS and record the occurrence of fibers > 25 um in length at 100 x
          magnification to get a statistical representation of fiber compositions
       10 Prepare 10-fold dilution  of the suspension from step 7 and apply an aliquot to a
          polycarbonate/adhesive tab substrate affixed to an aluminum sample stub
          Alternatively, a lighter loading can be prepared by filtering the diluted suspension
          through a 25-mm diameter, 0 4-um pore size, polycarbonate filter and affix this to a
          carbon adhesive tab affixed to an aluminum sample stub
       11. Collect x-ray maps of 10 fields at 500 x magnification  for major elements, especially
          Ca, S, and Fe and use Adobe Photoshop or similar software to determine the area
          percent of gypsum and Ca-nch particles  Fe-nch particles may also be identified in
          this step
       12 Perform particle analysis via computer-controlled SEM/EDX analysis

 5.0   Interferences

       Interferences include possible contamination of samples by airborne dust or through
       improperly cleaned glassware and sieves. Interferences are minimized by performing all
       procedures involving dry dust in a clean room, cleaning countertops and glassware
       thoroughly before proceeding and placing particle-free wipes on all working surfaces To
       avoid cross-contamination, properly clean all glassware, sieves, and tools between
       samples

 6.0    Safety

        Respirable particles which may present a health hazard may exist in the sample  Bulk
        samples may release respirable particles during handling  All procedures involving dry
        dust samples will be performed under a negative flow High-Efficiency-Particulate-Air
                                                                                     37

-------
       Filter (HEP A) hood  Samples handled outside of the HEP A hood will be covered with
       aluminum foil or placed in sealed glass jars

7.0    Apparatus and Materials

       1   HEPA negative flow hood
       2   Forceps
       3   Kimwipes
       4   Stainless steel spatula
       5   Weighing paper
       6   Programmable furnace [not required for validation study]
       7   Ceramic crucibles with lids [not required for validation study]
       8.  Analytical balance (accuracy to 0 0005 g)
       9   Retsch ultrasonic sieve shaker (AS200 Basic), or similar [not required for validation
          study]
       10 Sample sieves, 3-inch diameter (recommended), 150-um (100-mesh) opening, with
          lid and bottom pan similar [not required for validation study]
       11 SEM aluminum sample stubs
       12 Conductive carbon adhesive tabs
       13 Eppendorf pipette, 10-uL capacity
       14 Disposable pipette tips
       15. 1 - 10 mL pipette
       16 Glass vials for sonicating dust in isopropanol suspension (holds 10-mL volume)
       17 Razor blade
       18 Ultrasonic bath
       19 50 mL glass beaker
       20 Polycarbonate filters (25-mm diameter, 0 4-um pore size)
       21 Polycarbonate filters (13-mm diameter, 0 4-um pore size), or borer to cut larger filters
          to SEM stub size
       22 11-mm diameter cork borer
       23 Milhpore filter apparatus for use with 25 mm filters
       24 125 mLNalgene bottles
       25 Hand-held vacuum pump
       26 High-vacuum carbon evaporator with rotating stage
       27 Glass petn dishes with lids
       28 Adobe Photoshop Software, or similar
       29 Glass petrographic slides
       30 Glass cover slips
       31 Polarized light microscope for mineral identifications
       32 Scanning Electron Microscope with the following attributes
              a.  Resolution 5 nm (at 25 kV, WD=10 mm - system dependent) or better
              b  Accelerating Voltage. 10 to 20 kV
              c  Minimum magnification range  50x to 200,000x
              d  SEI (secondary electron image)
              e  BEI (backscattered electron image)
              f  Energy dispersive x-ray detector and analyzer for EDS analysis
                                                                                     38

-------
              g.  Ability to collect x-ray maps or particle analysis software (preferably both)

8.0    Reagents

       1  Isopropanol, reagent grade [CAS No. 67-63-0]
       2  1 55 or 1.605 Refractive Index Oil

9.0    Sample Storage

       Dust samples will be stored in an air-tight container, such as a sealed glass jar  Samples
       placed in reagents will be labeled appropriately and stored according to laboratory safety
       standards  Samples  prepared for analyses will be stored in a protective container, such as
       a plastic case or covered petn dish, to prevent contamination

10.0  Quality Control

       Quality control is implemented by thoroughly cleaning glassware and spatulas, keeping
       working surfaces clean, and preventing cross contamination  During ashing, particles
       may be suspended if slow heating is not achieved  Following the ashing program as
       outlined will minimize flashing, which can cause particles to become airborne  Covered
       crucibles will be used to prevent contamination caused by flashing  Used Eppendorf
       pipette tips and weighing  papers will be discarded and new tips and papers will be used
       for each sample

       Duplicate samples shall be prepared to determine the precision of the analysis  In
       addition, sample blanks shall be prepared  These blanks are checks for cross
       contamination during handling of the samples. Blanks shall be prepared at the same time
       and in the same manner as samples

        10.1   Calibration

              Calibration of the EDS system must be completed at least once at the beginning
              and again at the end of each analytical session Backscattered electron image
              (BEI) calibration should be performed at the beginning of the session and anytime
              the backscattered  image brightness and/or contrast is  adjusted

              EDS calibration for both qualitative and quantitative (not required by this method
              but could be useful for identification of particle type) analysis is accomplished by
              the analysis  of a polished carbon-coated reference standard  The recommended
              material is USGS  BIR1-G basalt glass mounted in epoxy in a brass tube, polished,
              and carbon coated using a carbon evaporator (2, 3)

              The calibration reference material should be analyzed at the same operating
              conditions to be used for the analysis including beam current, accelerating
              voltage, working distance, detector dead time, and sample tilt (= 0°)  For BIR1-G
                                                                                      39

-------
             the analysis should be performed with a beam size of 10-20 urn or equivalent area
             raster  All calibration spectra will be saved with the corresponding data set.  The
             calibration data will be used for inter- as well as mtra-laboratory comparisons
             This calibration is in addition to, and not a substitute for the normal EDS
             calibration recommended by the EDS manufacturer which will be performed at
             regular intervals as specified by the EDS manufacturer

             Backscattered electron detector calibration can be performed on the same BIR1-G
             material by adjusting the detector brightness and contrast to achieve the following
             conditions  The epoxy on the BIR1-G reference material will be at 0 in a 256
             grayscale image and the brass mounting tube will be at 256  The BIR1-G basalt
             glass should fall at approximately 130-140 gray scale units

11.0   Procedure

       11.1   Weighing and Splitting

             Weighing and splitting should be performed under a negative flow HEPA hood
             If the fan speed is set too high, loss of particles may occur  The fan speed may
             need to be adjusted to prevent the loss of fine particles

             Obtain an analytical balance with an accuracy of 0 0005 g and preweigh a clean
             piece of weighing paper  Transfer the dust from the sample vial to the weighing
             paper and determine the weight of the dust  Split the sample with a clean razor
             blade using the cone-and-quarter method  If there are large clumps of organic
             fibers,  such as hair or lint, temporarily remove the hair with a pair of forceps and
             tap the forceps lightly with another tool over a piece of weighing paper to remove
             fine  particles.  Center the fine fraction on the paper and split the sample into four
             equal parts using a razor blade  Collect opposite comers ('/z of the sample) for
             analysis and archive the other half Quarter the larger organic fiber bundles the
             same way, keeping half to proceed to the ashing step and half for archival
             purposes

             Place the two quarters for ashing into a pre-vveighed crucible Weigh the split and
             record the results

       11.2   Ashing

             Place the ceramic crucibles containing the samples into a furnace.

             The  furnace program should proceed as follows
             1  Increase temperature by 1 °C/rmnute until sample reaches 250 °C
             2  Hold temperature at 250 °C for 4 hours
             3  Increase temperature by 1 °C/rmnute until sample reaches 480 °C
             4. Hold temperature at 480 °C (sufficient for decomposing organics) for 8 hours
                Do not exceed 500 °C
                                                                                     40

-------
       5   Shut off furnace
       6   Allow sample to cool before removing from furnace.
       7   Weigh the ashed sample to the nearest 0 0005 g and record the result

11.3   Sieving

       Sieve the sample through a 150-um sieve using a Retsch ultrasonic sieve shaker,
       or similar  Three-inch diameter sieves are recommended to minimize sample loss
       from particles being trapped in the sieve. The ultrasonic shaker will be operated
       at 20-minute intervals at the following settings' 20, 40, 60, 70, 80, then back
       down to 50 and 20. This will provide amplitudes ranging from 0 to 1  5 mm

       Transfer the large and small fractions to clean pieces of weighing paper and
       weigh to the nearest 0 0005 g  Archive the fraction greater than 150-um

11.4   Preparation of Sample for Polarized Light Microscopy

       Split the less than 150-um sample fraction using the cone and quarter method
       Collect one comer for analysis and archive the other three quarters Weigh the
       quarter split to the nearest 0 0005 g and place it into a glass vial  Make a
       suspension of 10-20 mg dust per mL of isopropanol.  The amount of isopropanol
       needed will vary depending on the amount of dust; the target dilution  is 10-20 mg
       per mL.

       Cut an Eppendorf pipette tip with a razor blade to increase the opening to
       approximately 1 mm

       Place the suspension in an ultrasonic bath for one minute,  then remove the
       suspension from the ultrasonic bath and shake it gently to  suspend all particles
       Collect a 10-uL aliquot of the mixture using an Eppendorf pipette with the
       modified tip and transfer to a glass slide.  Prepare 4 such slides   Allow them to
       dry, then add a drop of 1 55 (or 1 605) refractive index oil

 11.5   Preparation of Sample for SEM Analysis

       Prepare the SEM substrate on aluminum stubs using 0 4-um pore size
       polycarbonate filters, carbon  adhesive tabs  Using an 11 mm filter punch and
       placing the filter between two filter separators, punch a circle the size of the
       carbon tab into the filter Place carbon adhesive tab affixed to an aluminum stub
       on the dull side of the 11-mm polycarbonate filter such that the shiny  side of the
       filter exposed  If available, a 13-mm diameter polycarbonate filter may be used in
       place of the punched out 11-mm filter

       Collect a 10-u.L aliquot of the mixture from the PLM sample  preparation using
       the Eppendorf pipette with the modified tip and transfer to a prepared
       polycarbonate/adhesive tab substrate  This will yield a loading on a 12-mm SEM
                                                                               41

-------
             stub of about 100-200 ug, which is a moderately heavy loading  Adjust the
             number of aliquots as needed to obtain the target loading

             Prepare a 10-fold dilution of the above suspension to get a suspension of 1-2 mg
             dust per mL of isopropanol  Sonicate the suspension in an ultrasonic bath for one
             minutes Remove the suspension and gently shake it to suspend all particles
             Wait one minute to allow the coarse particles to settle  Collect a 10-u.L aliquot of
             the suspended mixture using an Eppendorf pipette with the modified tip and
             transfer to a prepared polycarbonate/adhesive tab substrate.  This will yield a
             loading on a 12-mm SEM stub of about 10-20 ug, which is a light loading
             Adjust the number of aliquots as needed to obtain the target loading.

             Alternatively, prepare a lightly loaded sample using the filtration method as
             follows  Use a Millipore filter apparatus for use with 25-mm filters for filtration
             Place a few drops of isopropanol on the fritted glass surface and place the 25-mm
             polycarbonate filter (0 4-um pore size) on the isopropanol  Attach the top of the
             apparatus and add a few milhhters of isopropanol to the filter so that no part of it
             is exposed to air.  Sonicate the suspension (diluted as described in previous
             paragraph) in an ultrasonic bath for one minute. Remove the suspension and
             gently shake it to suspend all particles  Wait one minute to allow the coarse
             particles to settle Collect 1 mL of the suspended mixture using a pipette and
             filter it through the polycarbonate filter Actual amounts for filtration will  vary
             based on sample loading The goal is to have a loading on a 12-mm SEM stub of
             about 10-20 ug, or about 5-10 percent area coverage, which is a light loading
             Adjust the volume of the aliquot to filter as needed to obtain the target loading

             Place the filter on a carbon adhesive tab on a standard SEM  aluminum mount
             The filter needs to be completely flat on the SEM stub  This can be achieved by
             forming the wet filter into a  gentle  U-shape using forceps and the side of the
             forefinger, then placing the bottom curve of the filter onto the center of the carbon
             adhesive tab and slowly  releasing the sides so they lay flat  Trim the edges of the
             filter using a razor blade.

             After drying, coat the samples on the polycarbonate or polycarbonate/adhesive tab
             substrates with carbon using a carbon evaporator with a rotating stage Transfer
             the stubs to the SEM in a clean, covered container

12.0   Analysis

       12.1   Analysis by Polarized Light Microscopy

             Pol an zed light microscopy will be conducted using the general techniques
             outlined in EPA 600/R93/116 (4)  For this procedure, four slides (prepared as
             described in section 114) will be analyzed  The fraction of fibers with refractive
             index greater than 1 55 (or 1 605) will contain mineral wool, which includes both
             slag wool and rock wool, and possibly some E-type glass and ceramic fibers The
                                                                                      42

-------
      fraction of fibers with refractive index less than 1.55 (or 1.605) will contain
      primarily soda-lime glass fibers. For the validation study, numbers of fibers
      greater than and less than 1 55 (1.605) refractive index will be counted
      Dispersion staining and becke line techniques may be used.  Fiber point counting
      will be performed at  100 x magnification

      If more than 20 mineral wool fibers are found, continue counting and recording
      all of the fibers above and below the index oil refractive index  Report both raw
      fiber counts per refractive index category and number of fibers from each
      category per gram of sample. Continue on to step 122 1 to determine the ratio of
      slag wool to other fibers with refractive index greater than 1.55 (or 1 605) using
      EDS as described below

      If less than 20 mineral wool fibers are found on each slide, count the number of
      slag wool fibers using SEM/EDS and report as number of fibers per gram of
      sample

12.2  Analysis by SEM/EDS

       12.2.1 Screening for Slag Wool

             Operating conditions for the JEOL 6460-LV SEM are 15 kV, 0 5-5-nA
             beam current, 10-mm working distance (system dependent), and zero
             degree tilt

             Place the more concentrated sample deposited directly on the
             polycarbonate/adhesive tab substrate into the SEM.  Use the backscattered
             electron mode at lOOx magnification to quickly distinguish carbon fibers
              from inorganic fibers (carbon fibers may  be visible, but not as bright in a
              BEI).  Identify all inorganic fibers over 25 urn in length (smaller fibers
              cannot be reliably detected at the lOOx operating magnification)  When an
              inorganic fiber is found, identify the composition  of the particle by EDS.
              Slag wool is  the primary fiber of interest. Record all inorganic fiber
              results as number of fibers for each fiber  type

              For the samples with high fiber loading, as determined by PLM as
              described in section 121, count fibers per type until a statistical
              representation of the ratios of fiber compositions in the sample is
              achieved  Report the ratio (by  fiber number) of slag wool fibers to total
              MMVF fibers corresponding to the high RI  Use this ratio to correct the
              total number for high RI fibers counted by PLM to number of slag wool
              fibers present

              For the samples with low fiber loading, as determined by PLM as
              described in  section 12.1, scan the entire stub to determine the number of
                                                                               43

-------
      fibers per type  Report the slag wool fiber results as the number of slag
      wool fibers/gram of sample

12.2.2 EDS Screening for Gypsum/Anhydrite

      Place the more concentrated sample deposited directly on the
      polycarbonate/adhesive tab substrate in the SEM  Choose a random field
      at lOOx magnification and perform an EDS analysis on the entire field
      Look for the presence of sulfur in this field  If sulfur is present, continue
      to Section 12 2 3 or 12 2 5 for analysis of gypsum and concrete by
      mapping or particle analysis  If it is not present, repeat the analysis on
      another random field  If sulfur is still not present, mark the sample as non-
      detect (ND) for sulfur

12.2.3 X-Ray Mapping for Gypsum

      Place a more dilute sample, deposited directly on the
      polycarbonate/adhesive tab substrate or prepared by filtration, in the SEM
      Collect binary backscattered electron images (particles white and
      background black, shadow off) and secondary electron images for  10 non-
      overlapping, random fields at 500 x magnification  Collect x-ray maps for
      Na, Mg, Al, Si, S, Ca, and Fe at each of these fields  Fields containing
      MMVF will not be used for this analysis Operating parameters for the
      SEM are the same as those for analyzing slag wool. Acquisition
      parameters for x-ray mapping using the NORAN System Six Software are
      time constant 14 (mapping mode, 11333 cps), 10-20 % deadtime, 256 x
      256 image resolution, 20 second  frame time, and 100 frames collected
      (about 40 minutes total acquisition  time) Secondary electron images will
      be used for reference only  Save all of the maps and electron images in
      TIFF format

      Open the backscattered electron image and the Ca and S x-ray maps in
      Adobe Photoshop  Make sure that  all of the element maps are the same
      size and resolution  by choosing Image Size from the Image Menu and
      changing the pixel size or the resolution as needed  The presence of
      gypsum can be determined by  overlapping the Ca and S maps

      Perform the following functions in  Adobe PhotoShop  (A macro is in
      development to perform the following functions to decrease user time and
      human errors in adjusting the threshold )

      1   Convert each of the three images to grayscale (Image —» Mode —>
          Grayscale)
      2   Perform an auto contrast and  brightness on each image and map to
          increase the scale of colors (Image —> Adjustments —» Auto Levels)
      3   Threshold each element map, Ca and S (do not analyze the
                                                                      44

-------
         backscattered electron image at this time), by going to the Image Menu
         and choosing Adjustments —* Threshold. Adjust the threshold to 128
         The background will be black and the particles white
      4  Invert the image (Image—* Adjustments —»Invert) to make the
         background white and the particles black
      5  Copy the S map and paste it over the Ca map in a separate layer in the
         file and change the opacity (located in the Layers window) to 50 % for
         the S map layer The black areas are gypsum/anhydrite
      6  Display a histogram of the image m expanded mode by selecting the
         Histogram tab on the Navigator Window (or under the Image Menu in
         some versions of Photoshop)  Place the cursor over the line for the
         black area and record the percentile for the black area. This is the
         percentage of particles containing Ca and S in the entire field.

      NOTE. If a binary backscattered electron image is obtained during data
      collection, then steps 7-11  may be deleted  The Invert function will,
      however, need to be applied to make the particles black and the
      background white before continuing to step 12.

      7  Begin analysis  of the backscattered electron image Select the
         particles by going to the Select Menu and choosing Color Range  Go
         to the selection pulldown menu and choose Highlights
      8  Fill the selection with black by going to the Edit Menu —» Fill and
         choosing black from the color pulldown menu.
      9. Select the inverse areas by going to the Select Menu and selecting
         Inverse
      10 Fill the selection with white by going to the Edit Menu —» Fill and
         choosing white from the color pulldown menu
      11 Deselect the area by clicking on the image
      12 Perform the Threshold and Histogram functions for the backscattered
         electron image as outlined in 3 and 6. Record the histogram result for
         the backscattered electron image

      Determine the area percent of gypsum by performing the calculations in
      Section 13.0.

12.2.4 X-Ray Mapping for Ca-Rich Particles

      Analysis of components of concrete will be performed on the same fields
      as the gypsum/anhydrite analysis  At this time, only a method for the
      determination of the area percent of Ca-rich particles is presented See
      Section 2 1  for discussion.

      Perform the following steps on the Ca x-ray map Tiff file in Adobe
      Photoshop
                                                                       45

-------
       1   Convert the Ca x-ray map to grayscale (Image —> Mode —» Grayscale)
       2   Perform an auto contrast and brightness on the map to increase the
          scale of colors (Image —» Adjustments —» Auto Levels)
       3   Threshold the Ca map by going to the Image Menu and choosing
          Adjustments —» Threshold Adjust the threshold to 128.  The
          background will be black and the particles white
       4   Invert the image (Image—> Adjustments —»Invert) to make the
          background white and the particles black
       5.  Display a histogram of the image  Place the cursor over the line for
          the black area and record the percentile for the black area This is the
          area percent coverage of particles containing Ca in the entire field

       Determine the maximum area percent coverage of non-gypsum, Ca-nch
       particles by performing the calculation in Section 130

12.2.5  Particle Analysis for Identification of Gypsum and Concrete.

       Place the more dilute sample, deposited directly on the
       polycarbonate/adhesive tab substrate or prepared by filtration, in the SEM
       Particle analysis will be used to identify gypsum and concrete particles

       Perform particle analysis at 500 x magnification. All other operating
       parameters for the SEM are the same as those used to analyze for slag
       wool (Section 1221) A binary backscattered electron image should be
       used in particle analysis mode. Particle analysis parameters should be set
       to analyze all particles m the field greater than 0 5 urn and to separate
       touching particles  For particles greater than 5 urn, scan the entire
       particle, spot analysis is adequate for smaller particles  The x-ray
       spectrum and counts for all particles, and an image of particles > 20 um
       long, will be recorded and saved  Other particle parameters to be reported
       will include the maximum, minimum, and average diameters, the aspect
       ratio, and area of each particle

       It will be necessary to review data collected by automated software to
       ensure data integrity  An Excel spreadsheet, in conjunction with images
       and x-ray data, may be used for this purpose  Particles  should be sorted
       into one of three categories. Ca-S (gypsum), Ca-nch, and Other.  Aid in
       identification of particles may by facilitated by referencing the U.S
       Geological Survey's WTC Dust Particle Atlas (1)   A particle
       classification protocol will be developed based on the data from the
       validation study

       The number of particles analyzed will be determined using the results of
       the validation study  For the study, the area percent of each component
       should be within 10% relative error or better.  Typically, data for  1000 -
       1200 particles should be acquired.
                                                                        46

-------
              Results for particle analysis will be recorded as area percent gypsum and area
              percent concrete particles for each field and average area percent for the each
              component m the sample


13.0  Data Analysis and Calculations

       1.  To determine the concentration of slag wool in fibers/gram, perform the following
       calculations:

              Determine the number of fibers with RI > 1 55 (or 1 605)

              # fibers identified - mg of sample on slide * 1000 = fibers/gram on slide

              Determine the percentage of fibers with the composition of slag wool with RI >
              1 55 (or 1 605)

              Fibers/gram on slide x # fibers identified as slae wool	= fibers slag wool/gram on slide
              Total number of fibers identified by EDS with RI > 1 55 (or 1 605)

              Back calculate to the number of fibers per gram of the original sample

              Fibers slag wool/g on slide *  g after sieving x E sample after ashing = Total f/g of sample
              g before sieving * g sample before ashing

       2  To determine the area percent of gypsum/anhydnte from the x-ray mapping
       procedure, perform the following calculations:

              Determine the area percent of gypsum/anhydnte in each field of view

              % of black area in Ca-S map overlay * 100= area % gypsum
              % of black area in BSE image

              Calculate the average percentage of gypsum/anhydnte for the sample

              (area % gypsum)(i + (area % gypsumVi +   = Avg area % gypsum
                     number of fields

       3  To determine the maximum area percentage of Ca-rich particles, which includes
       concrete particles, from the x-ray mapping procedure, perform the following calculations

              Determine the area percent of non-gypsum Ca-nch particles in each field of view:

     f/o black area Ca map") - f% black area Ca-S map") = % non-gypsum Ca-nch particles
     % black area on BSE image

              Calculate the average percentage of non-gypsum Ca-nch particles for the sample-

              (area. % Ca-nch particlesVi + (area % Ca-nch oarticleslp +     Avg  area % Ca-nch particles


                                                                                         47

-------
                          number of fields

      4. Calculate the area percent for gypsum and concrete by summing the areas of each
      particle in for each particle type and dividing by the total area analyzed.

             area gypsum 1 + area evpsum 2 +   x 100  = area percent gypsum (do likewise for concrete)
             total area analyzed

             Rules for concrete and gypsum classification are currently being developed

14.0  References

      1      Lowers, Heather A, Meeker, Gregory P, Brownfield, Isabelle K, 2005  World
             Trade Center Dust Particle Atlas U S Geological Survey Open-File Report 2005-
             1165  On the web athttp //pubs uses uoy/of/2005/J J65/
      2      Meeker, G P , Taggart, J E , and Wilson, S A, 1998  A Basalt Glass Standard for
             Multiple Microanalytical Techniques. Proceedings Microscopy and
             Microanalysis 1998 Microscopy Society of America.
      3      A polished and carbon coated calibration reference sample of BIR1-G may be
             obtained by contacting Stephen Wilson, U.S Geological Survery, MS 973,
             Denver Federal Center,  Denver, CO,  80225, svyijson^iis^s goy
      4      Perkins, R.L and Harvey, B.W., 1993, TEST METHOD: Method for the
             Determination of Asbestos in Bulk Building Materials, EPA/600/R-93/116
                                                                                     48

-------
   15.0  Appendix: DATA SHEETS
Determination of Slag Wool Fibers In Dust- PLM with Dispersion Staining
Sample ID
Circle One
General Sample
Homogeneous?
Structure #

Project
Analyst
Original Duplicate Triplicate Date
Appearance
Y
Rl Fluid
156

1 60S

Dispersion Staining
>RI

RI


-------
SEM Sheet
                                                                     Reference ASTM - D5755-03
Report Number: _
Sample Number:.
File Name:	
Sample Description.
Preparation Date:
Analysis Date:
Computer Entry Date:
Sample weiaht:
Dilution Volume:
Volume Aliquot:
Magnification:
Bv:
Bv:
Bv:







grams
mL
uL
X
   Structure #
                  Field #
Fiber Type
Length (Microns) Width (Microns)
Image
                                                                                         EDS
                                                                                                 50

-------
              Ai
                                             Ctte

*y^ Rvttdcs in ^B

-------
                         APPENDIX C: REAC SOP 2040
                  STANDARD OPERATING PROCEDURES
                                                                   SOP 2040
                                                                   PAGE of 10
                                                                   REV 00
                                                                   DATE
      05/17/02
         COLLECTION OF INDOOR DUST SAMPLES FROM CARPETED SURFACES FOR
             CHEMICAL ANALYSIS USING A NILFISK GS-80 VACUUM CLEANER
                                   CONTENTS

1 0    SCOPE AND APPLICATION*

2 0    METHOD SUMMARY*

3 0    SAMPLE PRESERVATION, CONTAINERS, HANDLING, AND STORAGE*

4 0    INTERFERENCES AND POTENTIAL PROBLEMS

50    EQUIPMENT/APPARATUS*

      5 1    Sampling Equipment*
      5 2    Sieving Equipment*

6 0    REAGENTS*

7 0    PROCEDURES*

      71    Preparation*
      7 2    Calibration Procedures*
      7 3    Field Operations*
      7 4    Sieving Procedures*
      7 5    Sampling Tram Decontamination*

8 0    CALCULATIONS*

90    QUALITY ASSURANCE/QUALITY CONTROL

10 0   DATA VALIDATION*
                                                                            52

-------
11 0    HEALTH AND SAFETY




120    REFERENCES*




130    APPENDICES*




These sections affected by Revision 0 0




       SUPERCEDES SOP #2040, Revision 0 0,11/18/98, USEPA Contract 68-C4-0022
                                                                                          53

-------
                    STANDARD OPERATING PROCEDURES
                                                                                SOP       2040
                                                                                PAGE    54 of 10
                                                                                REV         00
                                                                                DATE   05/17/02
         COLLECTION OF INDOOR DUST SAMPLES FROM CARPETED SURFACES FOR
              CHEMICAL ANALYSIS USING A NILFISK GS-80 VACUUM CLEANER
1 0     SCOPE AND APPLICATION

       The purpose of this Standard Operating Procedure (SOP) is to define the procedures for the collection of
       carpet-embedded dust samples that can be analyzed for lead, pesticides, or any other chemicals or elements
       This procedure is applicable for the collection of samples on a variety of carpeted surfaces This SOP may
       be modified to include the collection of dust adhering to floor surfaces but is not intended for the collection
       of dust containing asbestos fibers

       These are standard (i e,  typically applicable) operating procedures which may be varied or changed as
       required, dependent upon site conditions, equipment limitations or limitations imposed by the procedure
       In all  instances, the ultimate procedures employed should be documented and associated with the final
       report

       Mention of trade names or commercial products does not constitute United States Environmental Protection
       Agency (U S EPA) endorsement or recommendation for use

2 OMETHOD SUMMARY

       Sample collection is performed utilizing the Nilfisk GS-80 vacuum cleaner equipped with a high efficiency
       particulate air (HEPA) filter  A diagram of the Nilfisk GS-80 dust sampling apparatus is presented in
       Figure 1, Appendix A Soil and other particulate matter with aerodynamic diameters of approximately 5
       microns (^m) and larger that are embedded within the carpet are collected, sieved and submitted to the
       laboratory for analysis
3 0     SAMPLE PRESERVATION, CONTAINERS, HANDLING AND STORAGE

       Following collection  of a sample into a dedicated  collection bag, the bag is removed from the vacuum
       cleaner and placed into  a 32-ounce(oz) glass jar or a zip-lock plastic bag    Storage of the samples at
       ambient temperature is appropriate for samples that will be analyzed only for metals  Samples for organic
       analysis should  be maintained at approximately 4 ± 2 degrees Celsius (°C)

4 OINTERFERENCES AND POTENTIAL PROBLEMS

       There are no known interferences with this method

5 OEQUIPMENT/APPARATUS

       5 1    Sampling Equipment

                      •      Nilfisk Model GS-80 vacuum cleaner
                      •      Two-meter folding ruler or similar device
                      •      Masking tape

                                                                                             54

-------
                   STANDARD OPERATING PROCEDURES
                                                                               SOP       2040
                                                                               PAGE   55 of 10
                                                                               REV         00
                                                                               DATE   05/17/02
        COLLECTION OF INDOOR DUST SAMPLES FROM CARPETED SURFACES FOR
              CHEMICAL ANALYSIS USING A NILFISK GS-80 VACUUM CLEANER
                             Deiomzer or distilled water
                             Methanol, ACS grade
                             Kimwipes ™ or equivalent
                             Vacuum collection bags
                             Bottle brush
                             Scrub brush
                             Plotlmes
                             32-ounce glass jars or Ziploc  plastic bags
                             Disposable gloves

       5 2    Sieving Equipment

                      •       100-mesh  sieve, 150-Dn mean  diameter,  as  specified in ASTM D  422,
                             consisting of the cover, sieve and receiver pan
                      •       Sieve shaker for mechanical  sieving (CSC Scientific, Catalog Number 18480,
                             Thomas Scientific, Catalog Number 8324-A10) or equivalent
                      •       Analytical  balance, capable of weighing 0 1 milligrams (mg) and a range of 0 1
                             mgto 1000 grams (g)
                      •       Disposable gloves
                      •       Disposable dust mask
                      »       Clean aluminum foil
                      •       Kimwipes ™ or equivalent
                      •       Camel hair brush (TFisher Scientific, Catalog Number 03-655) or equivalent

6 0     REAGENTS

       Methanol and deiomzer/distilled water are required for sampling tram cleaning and decontamination

7 0     PROCEDURES

       7 1 Preparation

               The overall sampling strategy should be designed to address the goals of the study Users should
               consider factors such as foot traffic volume, types of activities, and proximity to potential sources
               The sampling strategy should be descnbed in the Work Plan (WP), Quality Assurance Project Plan
               (QAPP), or Sampling and Analysis  Plan (SAP)  prepared prior to the sampling event The ideal
               sampling locations are those areas that conform to the overall sampling strategy  For example,
               protocol may require the selection of a carpeted area for sampling where small children play or are
               likely to play
                                                                                             55

-------
                    STANDARD OPERATING PROCEDURES
                                                                                   SOP        2040
                                                                                   PAGE   56 of 11
                                                                                   REV          00
                                                                                   DATE   05/17/02
         COLLECTION OF INDOOR DUST SAMPLES FROM CARPETED SURFACES FOR
              CHEMICAL ANALYSIS USING A NILFISK GS-80 VACUUM CLEANER
                       1 Determine the extent of the sampling effort, the sampling methods to be employed, the
                       amount of dust needed to reach the desired detection limit and the types and amounts of
                       equipment and supplies needed

               2 Obtain and organize the necessary sampling and monitoring equipment

               3       Decontaminate or pre-clean equipment, as specified in Section 7 5, and ensure that it   is
                       in working order

               4 Prepare schedule and coordinate with staff, client, regulatory agency, as appropnate

               5       Perform a general site survey pnor to site entry in accordance with the site-specific
                       Health and Safety Plan

               6 Measure the area to be sampled and outline it using masking tape or other appropriate
                       methods Draw a diagram of the room(s) where the sample(s) were taken, locating the
sampled area(s)

        7 2Cahbration Procedures

               The Nilfisk GS-80 vacuum cleaner has no flow devices that require calibration pnor to sampling
               The sampling train shall be thoroughly inspected to ensure that it has been cleaned, properly
               assembled, and complete

        7 3 Field Operations

  1                     Prior to collecting a sample at a specific location, complete a Vacuum  Sampling Work
    Sheet (Figure 2, Appendix A) recording all required information and sketch the area to be sampled
  2                     Select a sampling area according to the data collection design  outlined in the WP, QAPP
    or SAP  Typically, three rooms per floor are selected for sampling in each building Each sample is collected
    with a dedicated sampling train that has  been properly assembled, cleaned, and  decontaminated to ensure
    sample integrity   The size/weight of each sample is dependent on the goals and objectives of the sampling
    event, the analyses requested, and the desired method detection levels (MDLs)  A  100-g sample is highly
    desirable if multiple analyses (metals, pesticides, etc ) are requested  A minimum 5- to 10-g sample is required
    for metal analysis only

               3       Using the 2-meter folding ruler or any  other measuring device, outline and mark the
                       recommended 1-square meter (m2) portion of the carpet to be sampled

               4       Begin collecting sample at one corner of the delineated sample area, moving the sampler
                       back  and forth four times over a strip running in  a straight line between the defined

                                                                                                 56

-------
  PsKpoittft Hwjfuasw-y onrt Atx&fwui Cortiiw.*


                   STANDARD OPERATING PROCEDURES
                                                                                   SOP        2040
                                                                                  PAGE   57 of II
                                                                                  REV          00
                                                                                  DATE   05/17/02
        COLLECTION OF INDOOR DUST SAMPLES FROM CARPETED SURFACES FOR
             CHEMICAL ANALYSIS USING A NILFISK GS-80 VACUUM CLEANER
                      sampling area edges  The width of the strip is defined by the width of the sampling
                      nozzle  After completing the first stnp, angle over to the second strip gradually on the
                      next pass, again completing four double passes

              5       Continue sampling  the delineated area until an adequate  sample is collected  Visual
                      observation is  used to determine if enough sample  has  been  collected  from the
                      recommended 1-m2 area or if a larger area is required  If sampling a larger area, measure
                      the area accurately and document accordingly

6                     Wearing surgical gloves, be sure to tap with your hand on the nozzle inlet to dislodge any
  dust remaining in the nozzle or the hose  This procedure will ensure complete sample recovery  Turn off the
  vacuum cleaner and allow to sit undisturbed for at least 30 seconds  Unsnap the two vacuum container clips to
  access the inside of the container Remove the polylmer and the vacuum collection bag within it  Seal off the
  polyliner with the vacuum collection bag inside, and transfer to a properly labeled 32-oz glass jar or plastic bag
  depending on the analysis(es) to be performed  Document the sample  information on the Vacuum Sampling
  Work Sheet and pack properly for shipment to the laboratory

7                     Remove the hose and the nozzle, and install a new polyliner and collection bag for the
  collection of additional samples

8                     Decontaminate the vacuum components using the steps outlined in Section 7 5

       7 4      Sieving Procedures

               Prior to submitting  dust samples to the  laboratory for analysis, the samples are sieved through a
               100-mesh sieve using the following procedure

   1                   Select a clean working area in a facility equipped with a fume hood (a 4-foot by 4-foot
       area is sufficient) Weigh the receiver pan on an analytical balance and record the weight

  2                   Wearing clean surgical gloves and a dust mask, retrieve the vacuum collection bags from
       the 32-ounce glass jars used to transport the bags  from the field to the laboratory

   3                   Empty  the entire contents of  the bag into  the 100-mesh sieve with the receiver pan
       attached  Remove the plastic adaptor (blue  ring) from the  collection bag inlet and shake  the bag as
       necessary to ensure all the contents have been transferred into the sieve

  4                   Place the  cover  on the  sieve and  manually or mechanically shake the  sieve  for a
       minimum of 5 minutes and  a maximum  of 10 minutes until all the fine dust particles are collected  in the
       bottom receiver pan  If manual shaking is performed, follow the instructions given in American Society for
       Testing and Materials (ASTM) D-422 "Conduct the sieving operation by means of a lateral and vertical

                                                                                                 57

-------
                   STANDARD OPERATING PROCEDURES
                                                                                  SOP        2040
                                                                                  PAGE    58 of 11
                                                                                  REV          00
                                                                                  DATE   05/17/02
        COLLECTION OF INDOOR DUST SAMPLES FROM CARPETED SURFACES FOR
              CHEMICAL ANALYSIS USING A NILFISK GS-80 VACUUM CLEANER
       motion of the sieve, accompanied by a jarring action in order to keep the sample moving continuously over
       the surface of the sieve  Continue sieving until not more than 1 mass percent of the residue on a sieve
       passes that sieve dunng 1 minute of sieving"

                      If mechanical shaking is performed, set up the recommended sieve shaker on an even and
                      stable  surface   Proceed  with the sieving  operation following directions in the
                      manufacturer's manual

              5       Re-weigh the receiver pan using an analytical balance  The difference in weight  is the
                      weight of the sieved sample  If total weight of material is desired, the coarse material
                      remaining on top of the sieve must be collected on a pre-weighed sheet of aluminum foil,
                      re-weighed and the weight added to the weight of the sieved sample

              6       Transfer the sieved sample from the receiver pan to an 8-oz wide-mouth glass jar   Use a
                      camel hair brush to ensure complete transfer of the sample Cap the glass jar securely

              7       Document each sample  Each sample must be provided with the following information
                      identification number, date of sampling, location, analysis requested Each sample must
                      be recorded onto a Cham of Custody form before delivery to the analytical laboratory

              8       Before processing the next  sample, thoroughly wipe clean the cover, sieve and receiver
                      pan using a Kimwipe™ and deiomzed/distilled water  Let dry prior to sieving additional
                      samples

       7 5     Sampling Tram Decontamination

              To decontaminate the sampling trains,  move them to a well-ventilated area and perform the
              following

 3                    Assemble one  of the sampling trains to be used  as  the decontamination unit for
   decontaminating the nozzles, hoses, and wands  This unit must be equipped with a clean polylmer and dust bag

!                      With the vacuum cleaner turned on, decontaminate the nozzles, wands, and hoses using a
   bottle brush to remove any accumulated dust in the hose and nozzle Be sure to tap the nozzle with your hand  to
   remove any visible dirt that has accumulated, and  use the scrub brush to remove any hair or fibers entangled on
   the nozzle's brush  When the nozzle is considered to be clean, remove and spray with reagent grade methanol
   and allow to air dry on a clean  surface The wand and hose are then cleaned with the bottle brush  Tap your
   hand on the wand inlet while cleaning with the bottle brush to remove any visible dirt Repeat this procedure  to
   decontaminate any remaining nozzles, wands, and hoses

              2       Remove the used dust bag from the decontamination unit and wipe clean the inside  of the

                                                                                               58

-------
                    STANDARD OPERATING  PROCEDURES
                                                                                  SOP        2040
                                                                                 PAGE   59 of 11
                                                                                 REV          00
                                                                                 DATE   05/17/02
         COLLECTION OF INDOOR DUST SAMPLES FROM CARPETED SURFACES FOR
              CHEMICAL ANALYSIS USING A NILFISK GS-80 VACUUM CLEANER
                       container with deiomzed/distilled water  Spray the inside of the containers with methanol
                       and allow to air dry  When decontaminating in between residential homes, cleaning the
                       inside of the containers with deiomzed/distilled water is sufficient
8 OCALCULATIONS
        The dust weight calculations for the final sieved dust fraction is performed in accordance with ASTM
        Method D-422 Dividing the final dust weight by the area sampled (expressed in m2) provides dust loading
        in grams per squared meter ( g/m2 )  When the analysis results are received, the loading of analyte in
        micrograms per square meter of carpet area (ug/m2) can be calculated  in the  same way The analysis
        provides concentrations in milligrams/kilogram (mg/kg) or micrograms/kilogram (Cg/kg)  If total (gross)
        dust loading of the sampled area needs to be  calculated, the total  dust weight before sieving must be
        obtained The total dust weight is divided by the area sampled to obtain total dust loading in g/m2

9 0     QUALITY ASSURANCE/QUALITY CONTROL

        There are no specific quality assurance activities which apply to the implementation of these procedures
        However, the following general QA procedures apply

        1       All data must be documented on field data sheets or within site logbooks

        2      All instruments  must be operated in accordance with operating instructions as supplied by the
               manufacturer, unless otherwise specified in the work  plan  Equipment checkout and calibration
               activities must occur prior to sampling/operation and they must be documented

10 0    DATA VALIDATION

        The information recorded dunng sampling will be used in conjunction with the analytical data during
        validation
11 0    HEALTH AND SAFETY

        When working with  potential  hazardous matenals, follow U S  EPA, Occupational  Safety and Health
        (OSHA) and corporate health and safety procedures

120REFERENCES

        American  Society For Testing  And Matenals  2000   Standard Practice for Collection of Dust from
        Carpeted Floor for Chemical Analysis, Designation D 5438-00, Repnnted from the Annual Book of ASTM
        Standards, Philadelphia, PA
                                                                                               59

-------
                 STANDARD OPERATING PROCEDURES
                                                                     SOP      2040
                                                                     PAGE   60 of 11
                                                                     REV        00
                                                                     DATE   05/17/02
        COLLECTION OF INDOOR DUST SAMPLES FROM CARPETED SURFACES FOR
            CHEMICAL ANALYSIS USING A NILFISK GS-80 VACUUM CLEANER
      American Society For Testing And Matenals  1998  Standard Test Method for Particle Size Analysis of
      Soils, Designation D 422-63, Repnnted from the Annual Book of ASTM Standards, Philadelphia, PA

      Instructions for Use-Nilfisk Model GS 80, Nilfisk of America, Inc .Malvern, PA (1987)

130   APPENDICES

      A - Figures
                                   APPENDIX A
                                      Figures
                                     SOP #2040
                                     May 2002
                                                                                60

-------
     STANDARD OPERATING PROCEDURES
                                    SOP:   2040
                                    PAGE:  61ofll
                                    REV:    0.0
                                    DATE:  05/17/02
COLLECTION OF INDOOR DUST SAMPLES FROM CARPETED SURFACES FOR
   CHEMICAL ANALYSIS USING A NILFISK GS-80 VACUUM CLEANER
      NILFISK
      ^O Qf\
      GooO
                FIGURE 1. GS-80

              Dust Sampling Apparatus
                                          61

-------
         STANDARD OPERATING PROCEDURES
                                                         SOP       2040
                                                         PAGE   62 of 11
                                                         REV        00
                                                         DATE   05/17/02
COLLECTION OF INDOOR DUST SAMPLES FROM CARPETED SURFACES FOR
    CHEMICAL ANALYSIS USING A NILFISK GS-80 VACUUM CLEANER
                  V.S BPM
                                              P»g.	of	
                   Rnpentt E8(lMcrfB| AMtytfetl CoMmt
                     Vienna Stnpftef Work hk«M

                      L««n< Mmn Ckm. Eton. Nl
                                 OPA/SftTCWAM
                                  KE AC Tttt Uriel.
                                               U»l Otto-

                                               OAu
                 FIGURE 2 Vacuum Sampling Work Sheet
                                                                    62

-------
   ftnptttaa HmjtnwsN.1 x<) oral Anrt-vV'oJ Cws>sw.»
                  STANDARD OPERATING PROCEDURES
                                                                         SOP      2040
                                                                         PAGE   63 of 11
                                                                         REV        00
                                                                         DATE   05/17/02
        COLLECTION OF INDOOR DUST SAMPLES FROM CARPETED SURFACES FOR
             CHEMICAL ANALYSIS USING A NILFISK GS-80 VACUUM CLEANER
SIEVE CLEANING

SCOPE AND APPLICATION

Sieves must be cleaned before each dust sample is separated into fractions  Most of the "near-
mesh size" particles can usually be removed from the apertures by inverting the sieve and gently
tapping the frame of the sieve  For sieves with apertures less than 1 millimeter (mm) (e g, 100-
mesh, 150 micron [Dn] sieve), the most effective method for cleaning the apertures is the use of
an ultrasonic bath

EQUIPMENT/APPARATUS

•      Ultrasonic bath, capable of holding a standard sieve
•      Magnifying glass
•      Source of air, standard hair dryer or compressed air
•      Spray bottle

REAGENTS

•      Ultrasonic cleaner or laboratory-grade detergent that leaves no interfering residues
•      Deionized (DI) water, Type II water or equivalent
       Methanol, American Chemical Society (ACS) grade or equivalent

PROCEDURE

The following cleaning procedure will be used to clean sieves prior to use and after each sample

1      Place the sieve into an ultrasonic bath containing detergent and DI water and sonicate for approximately 10
       minutes

2      Remove the sieve from the ultrasonic bath and nnse well with DI water
                                                                                     63

-------
                 STANDARD OPERATING PROCEDURES
                                                                      SOP       2040
                                                                      PAGE    64 of 11
                                                                      REV         00
                                                                      DATE   05/17/02
        COLLECTION OF INDOOR DUST SAMPLES FROM CARPETED SURFACES FOR
            CHEMICAL ANALYSIS USING A NILFISK GS-80 VACUUM CLEANER
3     Spray the sieve with methanol

4     Dry the sieve using a standard hair dryer or a compressed air source

5     Visually inspect the sieve to ensure that there are no remaining particles present in the apertures  A
      magnifying glass may be used to aid in this process

6     Repeat steps 1 through 5 prior to sieving subsequent samples
                                                                                  64
Rev 0, 3/11/05

-------
                           APPENDIX D: ACCESS AGREEMENT


                                      REQUEST FORM
                     for the U.S. Environmental Protection Agency's Effort
                        to Develop and Validate a WTC Dust Signature
                    and to Characterize Background Dust in New York City

Name of Occupant:

Address:

Apartment Number (if applicable)

Contact Phone Numbers


REQUEST

I have read the fact sheet on the U S Environmental Protection Agency's Program to develop and
validate a WTC dust signature and to characterize background dust in New York City and considered
the information provided to me by the U S Environmental Protection Agency (EPA) Having
considered the information regarding the sampling program, I would like to participate

AGREEMENT

On behalf of myself and any other occupants, I agree to the following

I consent to employees, authorized representatives and contractors of the EPA having access to the above
referenced space for as long as necessary to conduct dust sampling activities

1 agree to obtain any required permission for sampling activities from my building management  I also
agree to inform the EPA Project Monitor at least one business day prior to the scheduled work of any
building rules that are applicable to the program, including time restrictions, appropriate entrances to the
building, and elevator usage

I understand that the sampling will be performed by contractors retained by EPA I also understand that the
contractors performing sampling activities are required to maintain insurance coverage for commercial
general liability, workers  compensation, dishonest acts of their employees and environmental impairment
liability related to this work The contractors are required to maintain such insurance at all times that they
are conducting sampling activities  The contractors are responsible for damage or loss of property

 I understand that the activities will require access to interior spaces and the use of electricity  Sampling
 activities will be performed throughout the entire space

 I understand that the program will employ various methods of dust removal from surfaces, including, but
 not limited to, vacuuming and wet wiping
                                                                                             65

-------
COMMUNICATION OF RESULTS

I understand that I will receive a copy of the sampling results for the residence once an analysis has
been completed and the data are quality assured Depending upon analysis and review time, these
results may not be available until up to six months after sampling.

I understand that results provided for locations sampled under the signature study will only indicate
whether WTC dust signature components are present, absent or inconclusive Results provided for
locations sampled under the background study will indicate the presence of WTC signature dust, as
well as the presence and levels of the contaminants of potential concern (COPC)

I understand that an explanation of the findings will be included in these results along with the name
and contact information for a U S. EPA toxicologists/nsk assessor  This person will be able to answer
questions regarding data interpretation and health-related issues

I understand that monitoring data in EPA's database for this effort will be made available to the public,
but the identity of the participants and the specific location of the sampling will be kept confidential

AUTHORIZED SIGNATURE

I certify that I am authorized to grant this request on behalf of all  the occupants of the above specified
space, and I grant this request and agree to its terms
Signature                                 Date
Name and Title (PRINT)
Signature of U S  EPA Representative        Date
                                                                                           66

-------
                          APPENDIX E: INFORMATION SHEET

                    The U.S. Environmental Protection Agency's Program
                        to Develop and Validate a WTC Dust Signature
                    and to Characterize Background Dust in New York City


       The September 11, 2001 attack on the WTC covered a large area with dust, debris, and
combustion by-products In order to determine if residual contamination exists, and to identify areas
that may be in need of clean up, the US EPA has undertaken studies both to identify a unique WTC
dust signature, and to characterize typical indoor dust from NY City. In order to complete these
studies, the EPA is seeking to acquire samples of urban dust from buildings both inside and outside of
the area of lower Manhattan that was impacted by the WTC collapse You are being asked to
participate in the study checked below

WTC DUST SIGNATURE STUDY
The U.S Environmental Protection Agency has initiated a study to define signatures for WTC dusts
The purpose of this study is to develop and validate one or more "signatures" in indoor dust that can be
used to determine whether dust sampled is from the collapse of the World Trade Center towers or not.
A "signature" is a chemical or physical characteristic of a material that can be used to identify that
specific material and discriminate between the material sought (WTC dust, in this case) and other
similar materials (NYC urban dusts). The signature materials are not necessarily related to health
concerns  The signature could be something harmless but unique to the WTC source, measured only to
identify the origin of other chemicals of concern that occur in the same sample  The WTC signatures,
if they can be developed, will support analysis to discriminate between normal indoor dusts and WTC-
generated dusts

Samples from approximately 20 buildings are needed for validation of the proposed  signatures
Samples will be collected from approximately 10 buildings in the area that is suspected to be affected
by WTC emissions, and samples will be obtained from 10 buildings that are not suspected of being
affected

SAMPLING METHODS
Dispersion models, photos, interviews, and satellite data will be reviewed to discern areas that were
likely impacted by WTC emissions In each building identified for sampling, dust samples will be
collected from at least three areas 1) one sample from a track-in area near a building entrance,
preferably in a carpeted area, 2) two samples from relatively undisturbed areas  (e g  , on top of
bookcases, under furniture), and 3) other areas  showing visible accumulation of settled dust, including
HVAC ducts  A standard method using a HEPA vacuum collector will be used by EPA to collect bulk
dust samples  Samples will be sealed and stored under refrigeration  in a limited access area

To ensure that these important samples are properly collected, tracked, stored, and distnbuted,
comprehensive quality assurance (QA) procedures will be in place prior to any sample collection
There will be a pre-sampling survey of building and sampling areas, to include photos of sampling
areas (if permitted by building owners) and notes on building usage, to identify  conditions that might
compromise samples (e g , smoking or cooking areas)

                                                                                          67

-------
Dust samples from background and affected locations will be made available to researchers involved in
developing and evaluating WTC signatures, as well as researchers characterizing typical NY City dust
When the results of this work are complete, EPA will develop and release reports on these studies

COMMUNICATING RESULTS
Publicly released results will not be provided by name or specific location, thus a resident's privacy
will always be preserved.  The occupant will receive a copy of the sampling results for their residence
once an analysis has been completed and the data are quality assured. Depending upon analysis and
review time, this may take up to six months  An explanation  of the findings will be included in these
results along with the name and contact information for a U S EPA toxicologists/nsk assessor  This
person will be able to answer questions regarding data interpretation and health-related issues  Finally,
results provided for residences sampled under the signature study will only indicate whether WTC dust
signature components are present, absent or inconclusive
                                                                                            68

-------