UNITED
STATES
ENVIRONMENTAL
PROTECTION
AGENCY
OFFICE
OF
AIR
QUALITY
PLANNING
AND
STANDARDS
EMISSION
STANDARDS
DIVISION
RESEARCH
TRIANGLE
PARK,
NC
27711
August
12,
2003
SUBJECT:
Use
of
Metallic
Cobalt
in
Magnetic
Tape
Manufacturing
Process
at
Sony
Magnetic
Products
Inc.,
Dothan,
Alabama
FROM:
H.
Lynn
Dail,
Project
Lead
Coatings
and
Consumer
Products
Group
(
C539­
03)

TO:
Docket
No.
A­
91­
31,
Magnetic
Tape
Manufacturing
Operations
E­
Docket
OAR­
2003­
0161
Sony
Magnetics
Products
Inc.
of
America,
located
in
Dothan,
Alabama,
produces
a
specialized
magnetic
tape.
This
tape,
an
Advanced
Metal
Evaporative
(
AME)
media
incorporates
pure
cobalt
metallization
to
produce
their
trademarked
Advanced
Intelligence
Tape
(
AIT).
The
tape
is
used
as
a
computer
back
up
storage
tape
capable
of
storing
150
Gigabytes
on
an
8
mm
cassette.

The
magnetic
recording
layer
of
AME
media
contains
pure
evaporated
cobalt
and
is
deposited
on
the
tape
under
vacuum.
When
an
electron
bean
is
introduced
into
a
mass
of
metallic
cobalt,
the
metal
essentially
melts
into
a
liquid
that
is
magnetized
and
allows
the
cobalt
to
"
stick"
to
the
tape
as
a
surface
coating.
During
the
process,
no
metallic
cobalt
dust
or
vapor
emissions
are
occurring.

The
deposition
of
the
cobalt
onto
the
tape
in
this
process
is
inefficient.
Approximately
10%
cobalt
vapor
sticks
to
tape
and
the
remaining
amount
is
deposited
into
the
vacuum
chamber.
The
accumulation
in
the
vacuum
chamber
can
be
recovered
and
recycled.
Therefore,
a
cleaning
procedure
is
implemented
to
recover
the
remaining
cobalt
slag
and
dust
when
the
process
is
shut
down.
The
cleaning
process,
however,
does
not
recover
all
unused
cobalt.

Cleaning
the
equipment
is
a
health
hazard
to
employees
unless
OSHA
guidelines
for
personnel
protection
and
safety
are
followed.
The
cobalt
must
be
removed
from
all
parts
of
the
machine.
The
employees
don
protective
gear,
enter
the
production
area
and
initiate
a
cleaning
process.

The
opportunity
for
metallic
cobalt
emissions
occur
during
the
cleaning
process.
However,
emissions
mainly
consist
of
particulate
cobalt
instead
of
vapor.
The
emission
potential
is
very
low
and
the
possibility
only
occurs
during
the
cleaning
process.
There
are
three
pieces
of
production
equipment
that
require
cleaning.
The
largest
portion
of
the
equipment
opens
up
like
a
"
clam­
shell"
to
allow
access
for
cleaning.
The
two
large
halves
are
cleaned
in
place
in
the
production
area.
One
particular
removable
component
of
the
equipment
is
disassembled
from
the
2
line,
and
moved
into
a
cleaning
booth.
The
cleaning
booth,
of
which
Sony
operates
two,
is
constructed
with
a
bag
house.

Sony
reports
that
most
of
the
cobalt
accumulation
occurs
during
cleaning
and
is
metallic
cobalt
slag
adhered
to
machinery
or
small
amounts
of
dust
deposited
on
the
floor,
walls
and
equipment.
Any
airborne
particles
are
captured
by
the
bag
house.
Tests
conducted
by
Sony
indicate
that
the
respirable
particulate
matter,
particles
smaller
than
5
microns
in
size,
comprises
about
5
to
15%
of
the
mass.
The
remaining
85
to
95%
of
cobalt
is
a
heavy
metal
and
essentially
settles
onto
the
floor
within
a
reported
15
foot
radius
around
the
equipment
as
cleaning
is
underway.
Sony's
in­
house
testing
shows
that
beyond
15
feet,
no
cobalt
has
been
detected.
Sony
has
conducted
other
particle
detection
tests
and
found
no
cobalt
emissions
outside
the
buildings
where
the
process
equipment
is
installed.

The
Alabama
Department
of
Environmental
Management
(
ADEM)
permits
Sony
to
emit
9
tons
per
year
(
tpy)
or
metallic
cobalt.
It
has
been
the
practice
for
Sony
to
report
lost
inventory
as
emissions.
The
EPA
2000
Toxic
Release
Inventory
(
TRI)
data
showed
1555
pounds
per
year,
presumed
to
be
a
combination
of
inventory
loss
and
possible
emissions.
The
ADEM
provided
EPA,
Sony's
inventory
reports
for
a
period
covering
4/
2002
to
4/
2003.
The
reports
indicated
a
2.1
tpy
inventory
loss.

It
is
apparent
that
using
a
value
that
incorporates
inventory
loss
with
possible
emissions
may
artificially
elevate
the
emissions
and
result
in
a
false
indication
of
a
risk.
It
became
imperative
to
test
the
perceived
cobalt
emission
stream
to
determine
if
any
cobalt
could
be
lost
to
processes
other
than
emissions,
thus
reducing
the
overall
reported
cobalt
emissions.

Inventory
records
for
2003
indicated
a
2.1
ton
loss
of
Co
inventory.
Upon
testing
of
four
major
components
associated
with
the
magnetic
tape
production
process
using
Co,
Sony's
testing
has
accounted
for
1.4
tons
per
year
of
Co
not
lost
through
emissions,
yet
accounted
as
solid
waste.
The
emission
amount
of
0.7
tpy
is
more
representative
of
true
emission
amounts.
Sony
confirms
through
its
data
that
nearly
60
to
75%
of
it's
cobalt
inventory
loss,
reported
as
emissions,
is
actually
lost
in
the
waste
streams
other
than
airborne
emissions.
That
indicates
that
1.40
tpy
of
Co
is
accounted
for
in
the
metallic
slag,
unrecovered
slag,
baghouse
dust,
floor
ventilation
filters
and
possibly
wash
water
and
protective
gear
decontamination
wash
water
waste.
With
Sony's
confirmation
of
this
activity,
we
are
able
to
refine
and
reduce
the
emissions
quantities
of
cobalt
from
what
has
been
reported
in
the
2000
TRI,
down
to
a
level,
0.175
tons/
year,
that
has
resulted
in
a
hazard
index
for
this
site
that
is
well
below
EPA's
threshold.

Testing
has
confirmed
there
were
four
areas
where
amounts
of
Co
accumulated
as
either
metallic
slag
or
dust
and
where
not
emitted.
The
first
areas
tested
were
2
bag
houses
where
200
lbs
Co/
house
was
found.
Next,
testing
involving
the
in­
floor
ventilation
openings
produced
an
accumulation
of
50
lbs
Co/
year.
The
lion's
share
of
Co
loss
has
been
discovered
as
a
metallic
slag
of
solid
Co,
adhered
to
metallic
splash
protectors
that
were
a
part
of
production
equipment.
These
splash
protectors
are
used
and
disposed
of
as
solid
waste
with
the
Co
metallic
slag.
Sony's
current
estimates
indicate
the
majority
of
Co
loss
is
at
this
point.
The
final
area
where
Co
was
3
accounted
for
as
other
than
emissions,
was
the
cleaning
areas.
When
the
equipment
is
cleaned,
molten
Co,
accumulated
on
equipment
and
during
production,
releases
from
the
equipment
as
solid
chunks
or
flakes
and
thus
becomes
solid
waste.
These
fallen
chunks
and
flakes
collect
under
the
structure
supporting
the
production
equipment,
used
in
the
cleaning
process.
The
Co
flakes
"
pile
up"
until
removed.
As
a
side
note,
the
cleaning
wash
water
was
tested
to
have
less
than
10
lbs
Co/
year
accumulation
and
this
was
determined
insignificant
factor
in
delineating
Co
emissions/
inventory
loss.

Sony's
Environmental
Engineer
is
preparing
the
final
written
report
to
be
sent
to
me
by
Monday,
August
18,
2003.
The
bottom
line
that
I
am
reporting
to
you
is
the
following
facts:

1.
Sony's
Co
emissions
have
been
analyzed
and
better
Co
loss
tracking
indicates
that
emission
numbers
about
66%
less
than
previously
reported
based
on
inventory
losses
alone.

2.
Nearly
15%
of
Co
loss
was
discovered
in
bag
house
filters
and
in­
floor
ventilation
filters.

3.
Over
50%
of
the
Co
loss
is
solid
waste
in
the
form
of
molten
slag.

We'll
talk
more
when
the
final
written
report
from
Sony
hits
my
desk
next
week.
