TABLE OF CONTENTS	PAGE

I.	BACKGROUND	1

II.	SAMPLING DESIGN BASIS	1

A.	Testing Rationale	1

B.	Sampling Design	1

C.	DFS TSCA Testing	2

III.	PCB FEED (DFS TSCA TESTING)	2

IV.	EMISSIONS	3

A.	DFS Exhaust Blower Duct	3

B.	Common Stack	3

C.	LIC Exhaust Blower Duct	4

D.	Ambient Air	4

V.	DESTRUCTION AND REMOVAL EFFICIENCY	4

VI.	EXTRANEOUS SOURCES OF PCBS	4

VII.	CONCLUSIONS	5

LIST OF TABLES	PAGE

1	Sampling Time Intervals	6

2	PCB Emissions Summary at the DFS Exhaust Blower Duct	7

3	Dioxin/Furan Emissions Summary at the DFS Exhaust Blower Duct	9

4	Summary of PCDD/PCDF Emissions at the DFS Exhaust Blower Duct	11

5	Destruction and Removal Efficiency Summary for the DFS Exhaust Blower
Duct	12

6	PCB Emissions Summary at the Common Stack	13

7	Destruction and Removal Efficiency Summary at the Common Stack	15

8	PCB Emissions Summary at the LIC Exhaust Blower Duct	16

9	PCB Ambient Concentration Summary	18

LIST OF APPENDICES

A	ANCDF Test Plan/US EPA Letter

B	PDARS and Rocket Feed Summary

C	Sampling Report

D	Process Sampling Report

E	QA/QC Summary

F	Calibration Package

G	Field Logs and Master Sample Checklist

H	Analytical Data Packages

	H-1	STL Package No. G4C180344 (PCDDs/PCDFs and PCBs in Exhaust Gas)

	H-2	RESERVED (Data Package for Process Samples)

	H-3	RESERVED (Data Package for Wipe Samples)

	H-4	RESERVED (Data Package for Natural Gas Samples)

LIST OF ACRONYMS AND ABBREVIATIONS

%	percent

ACGIH	American Council of Government Industrial Hygienists

AEHA	Army Environmental Hygiene Agency

ANCDF	Anniston Chemical Agent Disposal Facility

ATB	agent trial burn

ATBP	agent trial burn plan

CAA	Clean Air Act

DFS	Deactivation Furnace System

DRE	destruction and removal efficiency

EPA	Environmental Protection Agency

ft/sec	feet per second

g/s	gram per second

HDC	Heated Discharge Conveyor

lbs/hr	pounds per hour

LIC	Liquid Incinerator

MDB	Munitions Demilitarization Building

mg/m3	milligrams per cubic meter

ND	non-detect

ng/dscm	nanograms per dry standard cubic meter

ng/m3	nanograms per cubic meter

NIST	National Institute of Standards and Technology

O2	oxygen

PAS	Pollution Abatement System

PCB	polychlorinated biphenyl

PCDD	polychlorinated dibenzo-p-dioxin

PCDF	polychlorinated dibenzofuran

PDARS	process data acquisition and recording system

PFS	PAS Filtration System

ppm	parts per million

RCRA	Resources Conservation and Recovery Act

rockets/hr	rockets per hour

STB	surrogate trial burn

TEF	toxic equivalent factor

TEQ	toxic equivalent quotient

TOCDF	Tooele Chemical Agent Disposal Facility

TSCA	Toxic Substances Control Act

US	United States

WHO	World Health Organization

I.	Background

Anniston Chemical Agent Disposal Facility (ANCDF) commenced rocket
processing on August 9, 2003, using the ANCDF Deactivation Furnace
System (DFS) GB Agent Trial Burn (ATB) Plan (ATBP), Revision 3a.  ANCDF
is required to test for the presence of polychlorinated biphenyls (PCBs)
under the Toxic Substance Control Act (TSCA), Resource Conservation and
Recovery Act (RCRA), and Clean Air Act (CAA) Permits.  The shipping and
firing tubes for the M55 rockets contain PCBs.  The amount of PCBs in
the shipping and firing tubes vary depending on agent lot as documented
in the 1986-87 United States (US) Army Environmental Hygiene Agency
(AEHA) Study.  The permit limits are as follows:

Emissions of PCB’s from the DFS shall not exceed 6.45E-07 grams per
second (g/s) (CAA) and a maximum of 0.2 nanograms toxicity equivalent
quotient per dry standard cubic meters (ng TEQ/dscm), adjusted to 7
percent (%) excess oxygen (O2) and

Each test of the ATB shall demonstrate a minimum PCB destruction and
removal efficiency (DRE) of 99.9999% (TSCA).

II.	Sampling Design Basis

Testing Rationale

ANCDF completed preliminary testing for TSCA in November 2003 and
completed the GB ATB in November 2003 on the DFS.  During the November
2003, sampling for PCBs was conducted at two locations:  the DFS exhaust
blower duct and at the common stack.  The testing at the DFS exhaust
blower duct was consistent with all surrogate trial burns (STB);
however, the Tooele Chemical Agent Disposal Facility (TOCDF) reported
that flexible heated transfer lines were a potential source of PCB
contamination.  ANCDF employs flexible heated transfer lines to sample
at the horizontal duct locations.  Though not an ideal location, the
common stack was identified as the only location where sampling could
occur without using a flexible heated transfer line.  The PCBs from the
DFS exhaust blower duct location met the TSCA Permit conditions.  The
common stack results suggest that the low flow conditions at the common
stack created results that are not defensible.  Also, the Liquid
Incinerator (LIC) was on-line processing natural gas only to increase
flow to the common stack.  During the LIC natural gas ATB in November
2003, low levels of PCBs had been detected.  These PCBs on natural gas
could have contributed to higher stack results.  ANCDF asserts that the
point of compliance is at the DFS exhaust blower duct where all TSCA
Permit conditions were met; however, on March 2, 2004, the US
Environmental Protection Agency (EPA) requested that ANCDF undertake
additional testing for PCBs.

B.	Sampling Design

The March 2004 letter from the US EPA requested that ANCDF conduct two
preliminary runs with simultaneous sampling at the LIC exhaust blower
duct when processing natural gas; sampling at the DFS exhaust blower
duct while feeding rockets, and sampling at the common stack.  The
letter went on to specify three runs while sampling at the DFS and the
common stack with the LIC in “bottle up” condition.  In a meeting
with the US EPA on March 12, 2004, the testing regime was changed to
conduct three runs with the LIC in “idle” mode processing natural
gas and sampling at the LIC exhaust blower duct, the DFS processing
rockets while sampling at the DFS exhaust blower duct, and at the common
stack simultaneously.  In addition, ambient air was sampled at the air
intake to the Munitions Demilitarization Building (MDB).  Appendix A
contains the ANCDF Test Plan and the field memorandum where the US EPA,
the Anniston Field Office, and Westinghouse Anniston agreed to the
changes.  The basis for the change is that the common stack flow is too
low to meet the requirements of the US EPA Method 2 when the LIC is in
“bottle up” condition.  The US EPA representative had been to the
National Institute of Standards and Technology (NIST) certified
laboratory in Birmingham, AL and discussed the calibration of the pitot
tubes used at ANCDF with the NIST laboratory personnel.  The pitot tubes
are certified to a velocity of 9.6 feet per second (ft/sec); however,
the flow at the common stack, based on the November data, was projected
to be less than 5.0 ft/sec.  The sampling regime was changed and
approved by US EPA to do three runs simultaneously at four locations: 
LIC exhaust blower duct, DFS exhaust blower duct, common stack, and
ambient air at the MDB air intake.   

C.	DFS TSCA Testing

On March 15, 16, and 17, 2004, ANCDF conducted four TSCA test runs.  A
SW-846 Method 0023A sampling train was used at each of the four
locations.  Run 1 samples were archived and not analyzed, except for the
field blank train.  The test runs were executed at the DFS exhaust
blower duct, LIC exhaust blower duct, and the common stack.  In
addition, an ambient air sample was taken for each run to measure PCB
contribution at the air intake of the MDB, the source for combustion air
for the DFS and LIC.  The sampling train at the common stack was an all
glass train in a vertical stack configuration.  The sampling trains at
the DFS and LIC exhaust blower duct, and ambient locations incorporated
a flexible heated transfer line in a horizontal duct configuration.  In
addition field blanks were taken at the DFS exhaust blower duct for
Run 1, the LIC exhaust blower duct for Run 2, and at the common stack
for Run 3.

Samples of the DFS scrubber brine was taken every 30 minutes during each
of the runs and composited into one sample per run at the on-site
laboratory.  Heated Discharge Conveyor (HDC) ash was sampled for each
run.  The cyclone ash drum did not contain sufficient volume of ash to
sample at the end of the fourth run.  Attempts were made to evacuate the
cyclone, per the procedure, and no ash was deposited in the cyclone
drum.  Thus, no sample of cyclone ash was generated from this test.  The
process sample results are in Appendix H-2.

Sampling times and durations are found in Table 1 by location.

III.	PCB Feed (DFS TSCA Testing)

ANCDF fed an average of 33.0 rockets per hour (rockets/hr) for Run 2,
31.78 rockets/hr for Run 3, and 30.67 rockets/hr for Run 4.  The
concentration of PCBs fed varies depending on the agent lot.  The
rockets fed are recorded on the Certificate of Destruction required by
the Chemical Agent Accountability Procedure.  Appendix B summarizes the
agent lots and the number of rockets from each lot.  Further, rockets
have one of three types of firing tubes:  chopped, matted, and
matted/coated.  The request for rocket feed from Anniston Army Depot was
for rockets that were representative of the worst-case PCB concentration
in the stockpile.  The highest concentration of PCBs is in the
matted/coated shipping and firing tubes.

The PCB concentration associated with chopped and matted shipping and
firing tubes was derived from the 1986-87 AEHA Study.  The average PCB
concentration of matted and chopped tubes based on 2,169 samples was
1,815 parts per million (ppm).  Of these shipping and firing tubes,
2,000 were matted and 169 were chopped.  The 1,815 ppm value includes
multiple entries of 0 ppm when the sample analysis was less than the
detection limit; thus, the 1,815 ppm is biased low.  The PCB
concentration of matted and coated is based on the same study (See
Tables E-2 through E-4 of the 1986-7 AEHA Study) reported mean values
of 15,200, 1,864, and 6,441 ppm yielding an average value of 7,835 ppm.
 The 1986-87 AEHA data represents the official database for PCB use. 
ANCDF used the mean of the mean of 6,441 ppm.  US EPA approved the use
of the 6,441ppm concentration in the March 2, 2004 letter (See Appendix
A).  The aggregate feed rate of rockets is in the Process Data
Acquisition and Recording System (PDARS) and the specific PCB
concentration per run is on Tables 5 and 7 (See Appendix B).  The
average concentration of PCBs fed during the March 2004 testing was 2.8
pounds per hour (lbs/hr).

IV.	Emissions

A.	DFS Exhaust Blower Duct

The PCBs that are reported for the DFS exhaust blower duct include the
total homologues and those congeners that are on the World Health
Organization (WHO) list as having toxic equivalency factors (TEFs).  The
total PCB emissions are 1.87E-08, 1.87E-08, and 2.46E-08 g/s for Runs
2, 3, and 4, respectively, and are in compliance with the CAA Permit
limit of 6.45E-07 g/s (See Table 2).  The weighted PCB concentrations
are 0.006, 0.006, and 0.006 ng TEQ/dscm corrected to 7% O2 for Runs 2,
3, and 4, respectively.  Polychlorinated dibenzo-p-dioxin (PCDD) and
polychlorinated dibenzofuran emissions (PCDF), shown on Table 3, are
0.008, 0.008, and 0.008 ng TEQ/dscm at 7% oxygen for Runs 2, 3, and 4,
respectively.  The combined PCB and PCDD/F TEQ weighted emissions are
0.014, 0.014, and 0.014 ng TEQ/dscm for Runs 2, 3, and 4,
respectively, and are in compliance with the TSCA Permit limit of 0.2 ng
TEQ/dscm corrected to 7% O2.  Table 4 contains PCDD/PCDF data.  All are
“non-detect” and in compliance with the ANCDF RCRA Permit.  The DRE
was 99.999995, 99.999995, and 99.999993% for Runs 2, 3, and 4,
respectively, as shown on Table 5.  The average DRE at the DFS duct was
99.999994%, and is in compliance with the TSCA approval of 99.9999%.

B.	Common Stack

The PCBs reported for the common stack include the total homologues and
those congeners that are on the WHO list as having TEFs.  The total PCB
emissions are 4.66E-08, 9.27E-08, and 2.48E-08 g/s for Runs 2, 3, and 4,
respectively, and are in compliance with the CAA Permit limit of
6.45E-07 g/s (See Table 6).  The weighted PCB concentrations are 0.006,
0.008, and 0.008 ng TEQ/dscm corrected to 7% O2 for Runs 2, 3, and 4,
respectively, and are in compliance with the TSCA Permit limit of 0.2 ng
TEQ/dscm corrected to 7% O2.  The DRE calculated at the common stack
based on the rocket feed to the DFS was 99.999987, 99.999974, and
99.999993% for Runs 2, 3, and 4, respectively, as shown on Table 7.  The
flow at the common stack was measured by a micromanometer as required by
US EPA Method 2 for low flow sources.  The velocity at the stack
averaged 6.2 ft/sec.  The velocities were 7.05, 5.94, and 5.63 ft/sec
for Runs 2, 3 and 4, respectively.  This flow is below the pitot tube
certification of 9.6 ft/sec at the NIST certified wind tunnel (See
Appendix A).

C.	LIC Exhaust Blower Duct

The PCBs reported for the LIC exhaust blower duct include the total
homologues and those congeners that are on the WHO list as having TEFs. 
The total PCB emissions are 1.09E-08, 8.47E-09, and 3.19E-08 g/s for
Runs 2, 3, and 4, respectively, and are in compliance with the CAA
Permit limit of 6.45E-07 g/s (See Table 8).  The weighted PCB
concentrations are 0.005, 0.005, and 0.006 ng TEQ/dscm corrected to 7%
O2 for Runs 2, 3, and 4, respectively, and are in compliance with the
TSCA Permit limit of 0.2 ng TEQ/dscm corrected to 7% O2.  No DRE can
be calculated for the LIC since it was operating on natural gas only
with no agent being fed.

D.	Ambient Air

The PCBs reported for the ambient air samples include the total
homologues and those congeners that are on the WHO list as having TEFs. 
The total PCB concentrations are 26, 24, and 37 ng/dscm with an average
concentration of 20 ng/dscm.  The weighted PCB concentrations are
0.005, 0.005, and 0.005 ng/dscm for Runs 2, 3, and 4, respectively. 
The results are on Table 9.

V.	Destruction and Removal Efficiency

The ANCDF TSCA approval for an incinerator is 99.9999% DRE.  The
regulation states that DRE is to be determined by dividing the
difference between the total PCBs fed and total PCBs emitted by the
total PCBs fed, all on a lbs/hr basis, and expressing the result as a
percent.  As discussed above, the DRE met the 99.9999% TSCA requirement
at the DFS exhaust blower duct and common stack.

EXTRANEOUS SOURCES OF PCBs

ANCDF has executed a variety of actions to determine if there are
sources external to the combustion process that may contain PCBs. 
Samples were taken of the Pollution Abatement System Filtration System
(PFS) prefilters, PFS condensate, chemical products that are used for
water treatment in the Pollution Abatement System (PAS), and wipe
samples around the sampling ports at the LIC, DFS, and common stack. 
Results received from the off site laboratory using SW-846 Method 8082
for PCB determination have been non-detect (ND) for the PFS prefilter
and condensate as well as the water treatment chemicals.  Results are
pending for the sample port wipes.

The American Council of Government Industrial Hygienists (ACGIH)
threshold limit value for PCB's with 54% chlorine is 0.25 milligrams
per cubic meter (mg/m3) (or 250,000 nanograms per cubic meter (ng/m3))
for a twelve-hour shift exposure as a time weighted average.  The
results for the MDB ambient air concentration over the four-hour time
period were varied with 0.0000365 mg/m3 (or 36.5 ng/m3) being the
highest.  As a conservative measure, this four-hour sampling period can
be extrapolated to assume that all twelve hours would be at the same
concentration level.  From the comparison, there is no concern for the
health of employees to this ambient level of PCB in the air around the
facility.  It should also be noted that the ambient PCB concentrations
detected at the MDB air intake are within backgound levels recorded by
the US EPA in the surrounding community.

ANCDF conducted an internet literature search for PCBs.  One US EPA
research article cited forest fires as a PCB contributor
(www.elsevier.com/locate/atmosen).  Both Talledega National Forest and
Pelham range were conducting controlled burns in March 2004.  The
ambient air is the source for combustion air for the furnaces.

The US EPA took wipe samples of the natural gas distribution system at
ANCDF.  The natural gas feed line to the boilers that supply hot water
to the facility were sampled.  These boilers have been in
near-continuous operation since December 2000; thus, should be
representative of the highest use of natural gas at ANCDF.

VII.	Conclusions

All TSCA approval requirements were met for the DFS exhaust blower duct
location and common stack.  The data from the common stack does not meet
US EPA Method 2 requirements, as the stack velocity continues to be
below the level that the NIST certified wind tunnel can calibrate. 
ANCDF requests that the US EPA grant interim operations consistent with
the rocket rate approved by ADEM, i.e., 50% then 75%.  The submission of
all the laboratory data to US EPA and their approval of the report will
then allow ANCDF to process rockets at 100% of the feed rate.

Table 1:	Sampling Time Intervals

Run	Date	1st Port	2nd Port



Start	Stop	Start	Stop

DFS Exhaust Blower Duct

1(1)	3/15/04	1240	1440	1530	1730

2	3/16/04	0740	0940	0955	1155

3	3/16/04	1502	1702	1712	1912

4	3/17/04	0800	1000	1025	1225

Common Stack

1(1)	3/15/04	1240	1504	1530	1754

2	3/16/04	0740	0940	0955	1155

3	3/16/04	1502	1702	1712	1912

4	3/17/04	0800	1007(2)	1025	1225

LIC Exhaust Blower Duct

1(1)	3/15/04	1240	1440	1530	1730

2	3/16/04	0740	0940	0955	1155

3	3/16/04	1502	1702	1712	1912

4	3/17/04	0800	1000	1025	1225

Ambient Air

1(1)	3/15/04	1240	1500	1530	1750

2	3/16/04	0740	0940	0955	1155

3	3/16/04	1502	1702	1712	1912

4	3/17/04	0800	1000	1025	1225



Footnotes:

Run 1 was conducted as a backup test run for all parameters.  All
samples associated with the run were archived.

Sampling was paused from 0933 to 0941 due to micromanometer adjustments.

Abbreviations:

LIC	Liquid Incinerator

DFS	Deactivation Furnace SystemTable 4:	Summary of PCDD/PCDF Emissions
at the DFS Exhaust Blower Duct

Parameter	Units	Run 2	Run 3	Run 4	Average	RCRA/CAA

Permit Limit

Total 2,3,7,8-TetraCDD	g/s	< 7.42E-12 [ND]	< 7.61E-12 [ND]	< 8.23E-12
[ND]	< 7.75E-12 [ND]	5.54E-11

Total 2,3,7,8-PentaCDD	g/s	< 1.57E-11 [ND]	< 1.47E-11 [ND]	< 1.53E-11
[ND]	< 1.52E-11 [ND]	2.77E-10

Total 2,3,7,8-HexaCDD	g/s	< 3.13E-11 [ND]	< 3.17E-11 [ND]	< 3.19E-11
[ND]	< 3.17E-11 [ND]	7.57E-10

Total 2,3,7,8-HeptaCDD	g/s	< 1.25E-11 [ND]	< 1.17E-11 [ND]	< 1.42E-11
[ND]	< 1.28E-11 [ND]	5.93E-10

Total OctaCDD	g/s	< 1.67E-11 [ND]	< 1.47E-11 [ND]	< 1.56E-11 [ND]	<
1.56E-11 [ND]	1.26E-09

Total 2,3,7,8-TetraCDF	g/s	< 7.60E-12 [ND]	< 6.78E-12 [ND]	< 6.22E-12
[ND]	< 6.87E-12 [ND]	5.54E-11

Total 2,3,7,8-PentaCDF	g/s	< 2.03E-11 [ND]	< 2.14E-11 [ND]	< 2.20E-11
[ND]	< 2.12E-11 [ND]	5.68E-10

Total 2,3,7,8-HexaCDF	g/s	< 3.22E-11 [ND]	< 3.22E-11 [ND]	< 3.26E-11
[ND]	< 3.23E-11 [ND]	1.16E-09

Total 2,3,7,8-HeptaCDF	g/s	< 2.03E-11 [ND]	< 2.12E-11 [ND]	< 1.78E-11
[ND]	< 1.98E-11 [ND]	6.69E-10

Total OctaCDF	g/s	< 1.51E-11 [ND]	< 1.65E-11 [ND]	< 1.69E-11 [ND]	<
1.62E-11 [ND]	5.54E-10

Total CDD/CDF	ng TEQ/dscm @ 7% O2	< 0.008 [ND]	< 0.008 [ND]	< 0.008 [ND]
< 0.008 [ND]	0.20



Abbreviations:

<	less than	@	at

CAA	Clean Air Act	CDD	chlorinated dibenzo-p-dioxin

CDF	chlorinated dibenzofuran	g/s	grams per second

ND	not detected	ng/dscm	nanograms per dry standard cubic meter

O2	oxygen	PCDD	polychlorinated dibenzo-p-dioxin

PCDF	polychlorinated dibenzofuran	RCRA	Resource Conservation and
Recovery Act

TEQ	toxic equivalent quotient

ANCDF TSCA Sampling and Analysis Report, Revision 0	March 2004



  PAGE  i 

  PAGE  6 

ANCDF TSCA Sampling and Analysis Report, Revision 0	March 2004



11

ANCDF

Toxic Substances Control Act Sampling and Analysis Report

Revision 0, March 2004

Submitted to:	National Program Chemicals Division

Submitted by:            Chemical Materials Agency                      


Aberdeen Proving Ground, MD 21010-5401

