December
3,
2004
Mr.
Steve
Spengler,
Chief
Water
Division
Mississippi
Department
of
Environmental
Quality
P.
O.
Box
39289
Jackson,
Mississippi
39289
Dear
Steve:

Recently,
staff
members
of
the
Mississippi
Department
of
Environmental
Quality
(
MDEQ)
and
the
U.
S.
Environmental
Protection
Agency
Region
IV
(
EPA)
met
to
discuss
water
quality
and
permit
compliance
issues
related
to
the
Mississippi
Phosphates
Corp.
(
MPC)
facility
in
Pascagoula,
Mississippi.
Discussion
topics
and
concerns
were
relayed
to
MPC
in
a
meeting
on
October
21,
held
shortly
after
the
EPA/
MDEQ
meeting.
One
specific
issue
raised
by
the
agencies
is
"
Does
MPC's
phosphogypsum
storage
and
process
water
management
system
meet
the
categorical
wastewater
standards
for
phosphate
fertilizer
facilities
found
at
40
CFR
418?"
MPC
has
research
this
matter
and
concludes
it
does.
This
letter
details
MPC's
rationale
for
this
belief.

If
further
information
is
required
on
this
topic,
please
call
me
or
Randy
Palmertree
at
(
228)
712­
3336.

Sincerely,

Trey
Fleming
Director
of
Environmental
Programs
cc:
Jim
Perkins
Randy
Palmertree
John
Sparks
Jimmy
Smith
Ed
McCraw
John
Milner
(
Brunini)
2
DISCUSSION
OF
WATER
BALANCE
IN
MPC's
GYPSUM
STACK/
PROCESS
WATER
SYSTEM
INTRODUCTION
The
MPC
facility
is
located
on
Bayou
Casotte
in
Pascagoula,
Mississippi.
The
facility
manufactures
sulfuric
acid,
phosphoric
acid
and
diammonium
phosphate
(
DAP).
Phosphogypsum
is
also
produced
as
a
byproduct.
Federal
regulations
(
40
CFR
61
Subpart
R)
dictate
that
all
such
phosphogypsum
be
stockpiled
onsite
and
capped
(
i.
e.
covered),
if
necessary,
at
the
end
of
the
stack's
useful
life
to
minimize
radon
emissions
to
meet
emission
limits
codified
at
40
CFR
61.202.
Radon
emissions
from
gypsum
originate
from
the
presence
of
low
concentrations
of
naturally­
occurring
radium
in
the
feedstock
phosphate
rock.
This
radium
decays
to
radon,
a
radioactive
gas.

Associated
with
the
phosphogypsum
stack
is
process
water/
wastewater.
This
includes
water
that
is
used
in
various
plant
processes
which
picks
up
contaminants
such
as
ammoniacal
nitrogen
(
N),
phosphorus
(
P),
and
fluoride
(
F).
This
water
also
absorbs
process
heat,
and
prior
to
reuse
is
cooled
by
circulation
through
various
ditches
and
ponds
in
the
gypsum
stack
system.

As
gypsum
is
placed
in
the
stack,
it
contains
residual
amounts
of
N,
P
and
F.
The
stack
is
designed
to
be
a
closed
system,
and
any
rainfall
that
falls
on
it
is
captured
for
reuse.
If
more
rain
falls
on
the
stack
than
can
be
reused
in
a
reasonable
amount
of
time,
the
water
is
treated
and
discharged
so
that
the
system
does
not
fill
up
and
perhaps
overflow.
This
is
consistent
with
the
philosophy
embodied
in
40
CFR
418.

The
MPC
facility
was
opened
in
the
late
1950'
s,
and
began
producing
gypsum
byproduct.
Facility
enlargements
have
increased
phosphate/
gypsum
production
over
time
to
the
current
annual
average
rate
of
about
1,200
tons/
day
P2O5.
The
West
(
old)
gypsum
stack
was
used
from
the
opening
of
the
facility
until
about
2002.
In
anticipation
of
the
West
stack
reaching
its
capacity,
MPC
began
permitting
a
new
(
East)
gypsum
stack
in
1995.
Permits
were
issued
for
the
new
stack
in
1997,
construction
was
completed
in
1998,
and
the
system
was
first
used
in
2002.

Shortly
after
the
East
stack
permits
were
issued,
MPC
and
MDEQ
began
negotiating
the
closure
of
the
West
stack.
A
closure
plan
was
approved
by
MDEQ
in
2001,
and
closure
work
began
the
following
year.
Most
of
the
West
stack
is
currently
capped
and
grassed,
and
the
final
portion
of
the
cap
is
scheduled
for
completion
by
Summer
2005.

MPC
has
used
a
double
liming
system
to
treat
wastewater
for
a
number
of
years.
Prior
to
2002,
this
was
done
in
a
series
of
mix
tanks
and
settling
ponds
located
on
the
north
end
of
the
West
gypsum
stack.
This
tank/
pond
system
was
a
batch
operation
that
required
5­
7
days
for
each
stage
in
the
two­
stage
liming
operation
to
process
and
discharge
wastewater.
In
2002,
MPC
installed
and
began
operation
of
a
mechanical
system
(
fully
tank­
based)
that
used
the
same
chemistry
as
the
old
system,
but
is
much
3
more
efficient
and
easy
to
operate.
The
new
system
is
continuous
rather
than
batch
and
can
process
more
than
twice
as
much
water
as
the
old
system
(
1.5
million
gallons
per
day
versus
0.7
million
gallons
per
day).
The
30­
year
average
rainfall
for
the
MPC
site
is
about
66
inches.

CATEGORICAL
STANDARD
The
categorical
wastewater
standard
for
phosphate
fertilizer
facilities,
found
at
40
CFR
418.15,
reads:

The
following
standards
of
performance
establish
the
quantity
or
quality
of
pollutants
or
pollutant
properties
which
may
be
discharged
by
a
new
source
subject
to
the
provisions
of
this
subpart:
(
a)
Subject
to
the
provision
of
paragraphs
(
b)
and
(
c)
of
this
section,
the
following
limitations
establish
the
quantity
or
quality
of
pollutants
or
pollutant
properties,
controlled
by
this
section,
which
may
be
discharged
by
a
point
source
subject
to
the
provisions
of
this
subpart
after
application
of
the
best
available
demonstrated
control
technology:
There
shall
be
no
discharge
of
process
wastewater
pollutants
to
navigable
waters.
(
b)
Process
wastewater
pollutants
from
a
calcium
sulfate
storage
pile
runoff
facility
operated
separately
or
in
combination
with
a
water
recirculation
system
designed,
constructed
and
operated
to
maintain
a
surge
capacity
equal
to
the
runoff
from
the
25­
year,
24­
hour
rainfall
event
may
be
discharged,
after
treatment
to
the
standards
set
forth
in
paragraph
(
c)
of
this
section,
whenever
chronic
or
catastrophic
precipitation
events
cause
the
water
level
to
rise
into
the
surge
capacity.
Process
wastewater
must
be
treated
and
discharged
whenever
the
water
level
equals
or
exceeds
the
midpoint
of
the
surge
capacity.
(
c)
The
concentration
of
pollutants
discharged
in
process
wastewater
pursuant
to
the
limitations
of
paragraph
(
b)
shall
not
exceed
the
values
listed
in
the
following
table:

­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­
Effluent
characteristic
Effluent
limitations
(
mg/
l)
Maximum
Average
of
for
any
1
Daily
values
for
day
30
consecutive
days
shall
not
exceed
­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­
Total
phosphorus
(
as
P)
105
35
Fluoride
75
25
TSS
150
50
­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­­

In
reading
Part
418,
it
is
important
to
note
that,
while
Paragraph
a)
clearly
puts
forth
a
"
no
discharge"
requirement
for
subject
facilities,
Paragraph
b)
equally
clearly
contemplates
circumstances
where
subject
facilities
may
treat
and
discharge
process
4
wastewater.
Therefore,
the
"
no
discharge"
requirement
is
not
an
absolute
one,
and
just
because
a
facility
treats
and
discharges
wastewater
from
time
to
time
does
not
mean
it
is
in
violation
of
Part
418.

FACTORS
AFFECTING
WATER
BALANCE
Relevant
to
this
language,
MPC's
calcium
sulfate
(
phosphogypsum)
storage
pile
(
or
stack)
is
operated
in
combination
with
a
water
recirculation
system.
The
system
has
historically
been,
and
is
currently,
designed
and
operated
to
maintain
the
required
surge
capacity
(
the
mandated
surge
capacity
has
changed
over
time).

First,
it
must
be
made
clear
that
the
MPC
facility
is
a
net
water
consumer.
Substantial
quantities
of
process
water
are
lost
to
evaporation
via
use
in
processes,
absorption
as
water
of
hydration
in
gypsum,
and
evaporation
from
the
open
water
storage
areas
on
the
stack.
This
basic
principal
applies
to
all
phosphate
facilities,
and
is
the
fundamental
basis
for
the
categorical
standard.

If
not
for
the
collection
of
rainwater
contaminated
by
exposure
to
phosphogypsum,
as
required
by
Part
418,
MPC
would
have
to
import
fresh
water
in
order
to
operate.
However,
because
we
cannot
preclude
rainwater
from
active
gypsum
disposal
areas,
this
water
must
be
captured,
incorporated
into
the
process
water
system,
and
either
evaporated,
consumed,
or
treated/
discharged.
Thus,
rainfall
frequency
and
intensity
are
the
governing
factors
in
whether
water
must
be
discharged
from
the
gypsum
stack/
process
water
system.
Further,
if
rainfall
were
delivered
to
the
plant
in
equal
daily
increments,
the
MPC
facility
would
never
have
to
discharge.
Conversely,
it
is
because
rainfall
comes
in
uneven,
often
excessive,
increments
(
i.
e.,
chronic
or
acute
conditions)
that
MPC
must
periodically
treat
and
discharge
water.

The
amount
of
rainfall
collected
is
directly
proportional
to
the
acres
of
land
on
which
the
rain
falls.
The
amount
of
open
acreage
in
the
gypsum
stack
system
at
MPC
is
not
arbitrary.
The
most
fundamental
design
characteristic
in
MPC's
system
is
geotechnical
stability.
To
explain,
the
entire
area
of
the
MPC
facility
is
geologically
underlain
by
a
large
clay
layer,
which
is
saturated
with
water.
As
a
load
(
gypsum)
is
placed
on
this
clay,
internal
pore
water
pressure
builds
up
within
the
clay
and
water
is
forced
out
of
it.
If
this
dewatering
process
occurs
too
rapidly,
the
clay
becomes
unstable,
and
slope
failure
can
occur.
This
is
essentially
a
big
landslide
down
the
side
of
the
gypsum
stack,
and
is
to
be
avoided
at
all
costs.
This
characteristic
has
been
evaluated
in
great
detail
at
MPC,
over
a
number
of
years.
A
detailed
evaluation
of
geotechnical
stability
is
contained
in
a
report
prepared
by
Dames
and
Moore
Consultants
as
part
of
the
process
to
permit
the
new
gypsum
stack
and
submitted
to
MDEQ
and
EPA
in
1996.
MPC
has
an
ongoing
stability
monitoring
program
at
both
the
old
and
new
stacks
to
ensure
this
does
not
happen.
If
measurements
indicate
pore
water
pressure
is
building
excessively,
then
the
placement
of
gypsum
is
relocated
for
a
period.
Our
geotechnical
consultants
have
calculated
the
required
acreage
over
which
to
spread
gypsum
to
avoid
this
phenomenon.
As
more
open
acreage
is
otherwise
detrimental
to
the
MPC
facility
(
more
rainwater
collected
results
in
more
wastewater
treatment
expense),
we'd
prefer
to
minimize
open
acreage.
However,
5
because
of
geotechnical
stability,
there
are
lower
bounds
on
the
acreage
necessary
to
operate
the
facility.

Louisiana
phosphate
facilities
were
exempted
from
most
of
the
Part
418
standards
in
1981
because
of
this
similar
geotechnical
phenomenon.

Once
the
"
open
acreage"
variable
is
established
(
it
is
now
276
acres),
we
then
use
references
to
determine
a
25­
year,
24­
hour
storm
event,
the
surge
capacity
required
by
Part
418.
For
the
Pascagoula
area,
this
is
10.2
inches.
This
amount
of
rainfall,
applied
to
the
open
acreage,
is
76.4
million
gallons
of
storage
capacity
in
the
system,
plus
additional
"
freeboard"
for
a
safety
factor
to
account
for
wave
action
in
the
water
storage
system
and
prevent
dike
failure
due
to
erosion
from
high
winds.
Open
acreage
and
the
25­
year,
24­
hour
storm
event
then
define
the
amount
of
storage
required
in
the
system.
Because
all
of
the
footprint
cannot
hold
surge
(
sideslopes,
paved
areas,
etc.),
the
pond
acreage
must
also
be
calculated.
The
locations
in
which
water
is
stored
comprise
a
known
acreage
(
not
all
open
acreage
is
used
for
water
storage
­
some
has
"
dry"
gypsum).
MPC
uses
daily
gauge
board
readings
to
monitor
changes
in
available
surge
capacity.

Part
418
prohibits
discharge
from
the
system
unless
chronic
or
catastrophic
precipitation
events
cause
the
water
level
to
rise
into
the
regulatory
surge
capacity.
If
such
precipitation
occurs,
Part
418
allows
MPC
to
begin
treatment/
discharge,
if
it
chooses.
If
the
midpoint
of
surge
is
reached
(
only
a
5.1
inch
rainfall
can
be
held),
then
MPC
is
required
by
Part
418
to
begin
treatment
and
discharge.
MPC's
system
allows
for
the
daily
computation
of
available
surge
capacity,
so
the
determination
of
can't
treat/
may
treat/
must
treat
is
easily
determined.

The
only
other
variable
affecting
water
balance
in
the
system
is
production
rate
(
increased
production
increases
water
consumption).
This
is
really
the
only
variable
over
which
MPC
has
some
control
­
area
is
governed
by
natural
soil
stability
and
rainfall
is
uncontrollable.
However,
plant
operating
rates
are
driven
by
market
conditions,
and
uneconomical
plant
operations
can't
be
regularly
used
as
a
water
balance
tool.

The
calculations
in
Attachment
I
show
a
gross
water
balance
with
all
the
major
inputs
and
outputs
in
the
water
balance.
These
results,
computed
using
conservative
values,
clearly
show
the
facility
has
a
negative
water
balance
on
an
annual
average
basis.
These
calculations
clearly
show
the
facility
is
designed
as
a
negative
water
balance,
or
"
zero­
discharge",
facility,
and
thus
meets
the
requirements
of
40
CFR
418.

BASIC
DIFFERENCES
IN
COASTAL
MISSISSIPPI
AND
CENTRAL
FLORIDA
While
comparisons
of
the
MPC
facility
and
other
phosphate
facilities
(
notably
in
Florida)
are
made
by
regulatory
agencies,
we
must
point
out
here
there
are
several
factors
that
are
fundamentally
different
between
Central
Florida
and
the
Mississippi
Gulf
Coast:

1)
Rainfall.
Data
collected
for
over
thirty
years
at
the
MPC
site
indicate
a
long­
term
annual
average
rainfall
for
the
facility
to
be
about
66
inches/
year.
This
is
consistent
6
with
data
available
on
the
National
Oceanographic
and
Atmospheric
Administration
(
NOAA)
website.
In
contrast,
NOAA
data
indicates
long­
term
annual
average
rainfall
for
the
Florida
phosphate
belt
of
44­
48
inches/
year.
This
40%
difference
equates
to
150
million
gallons
of
additional
water
to
be
managed
by
MPC
in
an
average
year
than
the
same
size
producer
in
the
Florida
phosphate
belt.
(
See
Attachment
II.)
2)
Geotechnical
stability/
Subsurface
geology.
As
discussed
above,
Florida
facilities
can
stack
gypsum
more
steeply,
and
therefore
can
store
more
gypsum
within
a
given
footprint.
3)
Solar
radiation.
Solar
radiation
is
the
driving
force
for
natural
evaporation,
a
key
component
in
the
facility
water
balance.
According
to
NOAA
data,
the
Central
Florida
phosphate
belt
receives
9%
more
solar
radiation
than
the
Mississippi
Gulf
Coast.
Therefore,
in
addition
to
the
rainfall
differential
and
subsurface
geology,
the
Florida
facilities
have
a
third
natural
advantage
over
MPC
to
produce
favorable
water
balances.
(
See
Attachment
II.)
4)
More­
protective
operating
standard.
In
1997,
Hurricane
Danny
deposited
about
20
inches
of
rainfall
on
MPC,
causing
MPC
to
bypass
partially
treated
wastewater
to
Bayou
Casotte
in
order
to
prevent
an
uncontrolled
and
perhaps
catastrophic
release.
Similarly,
MPC
had
two
bypass
events
in
Spring
1998
due
to
El
Nino
rainfall
totaling
40
inches
for
the
January­
March
quarter.
Several
months
after
the
El
Nino
bypasses,
MPC
and
MDEQ
staff
met
to
discuss
the
bypasses
and
controllable
factors
which
may
be
used
in
the
future
to
reduce
the
likelihood
of
bypassing.
MDEQ
indicated
that
MPC
should
use
a
100­
year
storm
event
(
13.4
inches)
as
the
operating
benchmark,
rather
than
the
25­
year,
24­
hour
storm
event
(
10.2
inches).
Both
parties
recognized
that
doing
this
would
increase
the
frequency
of
NPDES
discharge,
but
felt
that
more
frequent
discharge
of
fully­
treated
water
was
preferred
to
bypassing.
MPC
has
used
the
100­
year
benchmark
as
the
operating
standard
until
EPA
pointed
out
in
these
recent
discussions
with
MDEQ
that
it
was
a
violation
of
the
Part
418
standard
to
do
so.
It
has
now
reverted
to
the
25­
year
benchmark.
5)
West
stack
closure.
As
discussed
earlier,
the
West
phosphogypsum
stack
neared
the
end
of
its
useful
life
in
the
mid­
1990'
s,
and
a
new
stack
was
permitted
in
1997.
Working
with
MDEQ,
MPC
began
capping
the
old
stack
in
2002
to
remove
acreage
from
the
wastewater
management
system
covered
by
Part
418,
as
rainwater
running
off
the
completed
cap
does
not
contribute
to
the
facility
water
balance.
However,
to
transition
from
an
operating
West
stack
(
with
an
uncontaminated
East
stack
area)
to
a
fully­
closed
West
stack
(
and
fully
operating
East
stack)
required
a
3­
year
transition
period.
During
this
transition
period,
contaminated
acreage
changed
regularly,
but
it
rose
from
about
300
acres
(
pre­
closure),
up
to
450
acres
(
Winter
02­
03),
and
is
currently
276
acres.
This
temporary
rise
in
acreage
increased
the
amount
of
water
MPC
had
to
handle
during
this
period,
and
coupled
with
the
100­
year
storm
operating
benchmark,
most
certainly
increased
discharge
frequency.
7
CONCLUSION
Given
all
the
water
balance
variables
discussed
in
this
letter,
MPC
has
a
facility
designed
for
"
no
discharge".
That
is,
under
typical
conditions,
the
facility
can
clearly
demonstrate
a
negative
water
balance.
The
only
controllable
variable
(
acreage)
has
been
minimized
to
the
as
much
as
possible
given
the
geotechnical
criteria.
The
facility
is
maintained
so
that
it
can
easily
reach
the
holding
capacity
mandated
by
Part
418.
But
for
the
period
it
operated
using
the
100­
year
surge
as
a
benchmark
(
done
at
the
behest
of
MDEQ),
the
facility
has
operated
to
maintain
a
25­
year,
24­
hour
surge
capacity,
using
the
can't
treat/
may
treat/
must
treat
criteria
of
Paragraph
b).
Therefore,
MPC
is
confident
the
facility
complies
with
40
CFR
418,
regardless
of
differences
in
discharge
profiles
with
Florida
facilities.

Lastly,
though
it
has
no
bearing
on
compliance,
we
think
MDEQ
and
EPA
should
understand
that
treating
wastewater
at
MPC
costs
a
lot
more
money
than
holding/
evaporating
water.
If
there
was
a
way
to
reduce
the
necessity
of
treating
process
wastewater,
MPC
would
be
doing
it
for
economic
reasons.
MPC
treats/
discharges
wastewater
only
as
it
understands
it
is
compelled
to
pursuant
to
its
NPDES
permit
and
federal
and
state
regulations.
8
ATTACHMENT
1
FACILITY
WATER
BALANCE
CALCULATION
Constants
Survey
Data
Average
Rainfall
80.0
in/
yr
Ponded
Acreage
63%
Reservoir
Evaporation*
44.0
in/
yr
Non­
ponded
Acreage
37%

Evapotranspiration**
31.9
in/
yr
Total
Exposed
Acreage
276
Plant
Water
Accumulation
Water
Attributed
to
Rainfall
493,591,942
gal/
yr
Water
Evaporated
(
Reservoir)
207,749,296
gal/
yr
Water
Evaporated
(
Evapotranspiration)
44,229,166
gal/
yr
Total
Water
Evaporated
251,978,461
gal/
yr
Water
to
be
Consumed
241,613,480
gal/
yr
460
gpm
Plant
Water
Consumption
T
H20
/
T
P2O5
Gal
H2O
/
T
P2O5
Gypsum
Dihydrate
0.95
228
Sponge
Water
1.89
453
Emissions
0.20
48
Phosacid
1.09
261
Heat
Load
Evaporation
1.75
420
Sulfuric
Acid
(
0.11)
(
26)
Phosphate
Rock
(
0.19)
(
47)
DAP
Scrubber
Return
Water
(
0.31)
(
74)

Pump
Seal
Water
(
0.50)
(
120)
Total
Plant
Water
Consumption
4.77
1143
Water
Consumption
@
1000
TPD
P2O5
794
gpm
Water
Consumption
@
1200
TPD
P2O6
953
gpm
Water
Balance
All
units
are
in
terms
of
gpm
1000
TPD
P2O5
1200
TPD
P2O5
460
Water
Required
to
be
Consumed
460
794
Plant
Water
Consumption
953
(
334)
Net
Water
Gain
(
Loss)
(
493)
*
Synthesis
of
a
Serially
Complete
and
Homogeneous
Evaporation
Data
Set
for
the
Southeastern
Region
of
the
United
States,
May
2004
**
Evaluation
of
Post­
Closure
Dewatering
Behavior,
September
2000
9
ATTACHMENT
II
LETTER
FROM
T.
FLEMING
TO
A.
ZIMMER
9/
9/
04
