1
MEMO
To:
Martha
Segall,
EPA
CC:
Ashley
Allen,
EPA
Paul
Shriner,
EPA
From:
Terry
Haas,
Tetra
Tech
Scott
Daly,
Tetra
Tech
Date:
October
19,
2004
Spatial
Analysis
of
316(
b)
Phase
III
Freshwater
Facility
Locations
to
303(
d)
Impaired
Waters
Summary
The
purpose
of
this
analysis
was
to
determine
the
occurrence
and
frequency
of
316(
b)
Phase
III
freshwater
facility
withdrawals
from
303(
d)
impaired
waterbodies.
Using
a
Geographic
Information
System
(
GIS)
proximity
analysis
it
can
be
said
that
at
least
55%
of
all
Phase
III
facilities
withdrawing
from
a
freshwater
stream
or
river
exist
within
watersheds
that
contain
impaired
stream
segments.

Method
The
method
used
was
a
GIS
proximity
analysis,
supported
by
review
of
the
attribute
data
associated
with
the
GIS
data
files.
Spatial
data
from
the
P3Master.
xls
file
based
on
survey
results
for
Phase
III
freshwater
facilities
was
overlain
with
spatial
data
representing
all
303(
d)
impaired
waters
in
the
United
States.

Using
buffer
distances
of
100
meters,
500
meters,
one
mile,
and
two
miles,
spatial
selection
was
used
to
select
the
closest
303(
d)
impaired
stream
segment
near
each
facility,
and
the
name
of
the
selected
303(
d)
waterbody
was
compared
with
the
name
of
the
waterbody
identified
in
survey
data
to
verify
that
the
spatial
selections
were
correct.
If
the
name
of
the
spatially
selected
303(
d)
waterbody
matched
the
name
that
the
facility
identified
as
the
intake
water
source,
then
it
is
assumed
that
the
cooling
water
intake
withdraws
water
from
the
303(
d)
waterbody.

More
details
of
the
GIS
analysis
method
are
provided
in
Appendix
A.

Data
Sources
303(
d)
reach
data
based
on
2002
data
was
obtained
from
the
U.
S.
EPA,
Current
Section
303(
d)
Listed
Waters
­
Linear
Features,
http://
www.
epa.
gov/
waters/
data/
downloads.
html.
These
data
are
provided
in
Decimal
Degrees,
NAD
1983
datum.

Latitude
and
longitude
data
for
the
Phase
III
freshwater
facilities
was
obtained
from
the
P3Master.
xls
file.
It
is
assumed
that
the
latitude
and
longitude
values
provided
for
the
intakes
are
also
based
on
the
NAD
1983
datum.
2
Environmental
Science
Research
Institute's
(
ESRI)
U.
S.
States
shapefile
data
was
used
as
a
background
reference
layer.

Results
The
100­
meter
analysis
selected
facilities
with
geographical
data
within
100
meters
of
a
303(
d)
listed
reach.
The
results
of
the
100­
meter
analysis
showed
that
55
out
of
322
Phase
III
freshwater
facilities
appear
to
be
located
within
100
meters
of
a
303(
d)
waterbody
(
Table
1).
Of
these,
one
facility
does
not
match
the
waterbody
name
between
the
facility
data
table
and
the
303(
d)
data
tables.
Therefore,
approximately
98.2%
of
the
facilities
withdraw
water
from
an
impaired
stream
of
the
same
name.
Table
1
summarizes
the
results
of
the
buffer
distance
analyses
at
100
meters,
500
meters,
1
mile,
and
2
miles.

Table
1.
Summary
Results
for
GIS
Buffer
Analysis.

Buffer
Distance
Number
of
Facilities
Within
Distance
of
303(
d)
Segment
Number
of
Facilities
Within
Distance
of
303(
d)
Segment
with
Matching
Waterbody
Name
Percent
Matching
Percent
of
Total
Freshwater
Facilities
Within
Distance
of
303(
d)
Segment
with
Matching
Waterbody
Name
100
meters
55
54
98.18
16.7
500
meters
141
138
97.87
42.85
1
mile
201
178
88.56
55.27
2
miles
232
192
82.76
59.62
The
500­
meter
analysis
selected
facilities
with
geographical
data
within
500
meters
of
a
303(
d)
listed
segment.
The
results
of
the
500­
meter
analysis
showed
that
141
out
of
the
322
freshwater
facilities
appear
to
be
located
within
500
meters
of
a
303(
d)
listed
waterbody.
Three
of
these
do
not
have
waterbody
name
matches
between
the
facility
data
table
and
the
303(
d)
data
tables,
so
the
actual
number
of
facilities
with
intakes
from
303(
d)
listed
waters
might
range
down
to
138,
or
so.
Thus,
using
the
500­
meter
spatial
analysis
provides
about
97.9%
of
facilities
that
withdraw
water
from
a
303(
d)
listed
stream
of
the
same
name.

The
one­
mile
analysis
selected
facilities
with
geographical
data
within
one
mile
of
a
303(
d)
listed
segment.
The
results
of
the
one­
mile
analysis
showed
that
201
out
of
322
Phase
III
freshwater
facilities
appear
to
be
located
within
one
mile
of
a
303(
d)
listed
waterbody.
Of
these,
23
facilities
do
not
have
a
waterbody
name
match
between
the
facility
data
table
and
the
303(
d)
data
tables.
Therefore,
approximately
88.6%
of
these
facilities
withdraw
water
from
an
impaired
stream
of
the
same
name.

The
two­
mile
analysis
selected
facilities
with
geographical
data
within
two
miles
of
a
303(
d)
listed
segment.
The
results
of
the
two­
mile
analysis
showed
that
232
out
of
322
Phase
III
freshwater
facilities
appear
to
be
located
within
two
miles
of
a
303(
d)
listed
waterbody.
Of
these,
40
facilities
do
not
have
a
waterbody
name
match
between
the
facility
data
table
and
the
303(
d)
data
tables.
Approximately
82.8%
of
these
facilities
withdraw
water
from
an
impaired
stream
of
the
3
same
name.

Overall,
the
500­
meter
spatial
analysis
captures
60%
more
facilities
than
the
100­
meter
analysis
while
maintaining
the
same
percentage
of
matched
facilities.
In
addition
the
one­
mile
buffer
incorporated
22%
more
facilities
than
the
500­
meter
buffer
while
the
2­
mile
buffer
accounted
for
seven
percent
more
than
the
1­
mile
analysis.

Limitations
As
wider
selection
buffers
are
used,
the
percentage
of
waterbody
name
matches
decreases.
It
is
important
to
note
that
this
analysis
selects
only
the
closest
303(
d)
stream
segment
to
the
facility
and
overlooks
other
stream
segments
with
proximity
to
the
facility.
The
wider
the
buffer,
the
greater
the
chance
to
include
an
impaired
stream
segment
that
is
not
directly
relevant
to
the
facility
or
its
withdrawals.
In
other
words,
the
facility
may
withdraw
from
an
unlisted
waterbody
(
or
at
least
unlisted
within
the
buffer
radius)
and
a
segment
on
an
entirely
different
waterbody
may
be
the
closest
impaired
segment,
leading
to
a
mismatched
waterbody
name
in
the
analysis.

Also,
it
is
possible
that
some
facilities
withdraw
water
from
farther
than
the
buffered
distance.
For
example,
Table
1
shows
that
the
two­
mile
buffer
matched
14
more
waterbodies
than
the
onemile
buffer.
The
same
result
may
be
observed
for
a
facility
with
intakes
located
more
than
two
miles
from
the
facility.
Other
possible
reasons
include
incorrect
source
data
errors
in
facility
and
303(
d)
segment
locations.
4
Appendix
A:
GIS
Analysis
Details
Spatial
selection
of
303(
d)
water
bodies
lying
within
a
specified
distance
of
a
cooling
water
intake
or
facility
is
a
simple
GIS
analysis
procedure.
This
is
done
using
the
ArcView
Select
By
Theme
procedure.
However,
correlating
which
waterbody
is
related
to
which
intake
is
more
complicated.

The
P3Master.
xls
data
table
includes
fields
WATERBODY1,
WATERBODY2,
and
WATERBODY3
that
provide
the
names
of
waterbodies
that
the
intakes
withdraw
water
from.
The
list_
impairments.
txt
data
table
for
the
303(
d)
data
contains
a
field
WATER_
BODY_
NAME
that
identifies
the
303(
d)
waterbody.
Matching
the
names
between
these
two
data
sets
provides
confirmation
that
the
intake
withdraws
water
from
the
303(
d)
waterbody.

After
the
Select
by
Theme
spatial
selection
(
as
an
example,
select
303(
d)
reaches
within
500
meters
of
intakes
is
provided
here)
the
303(
d)
waterbodies
were
correlated
to
the
intakes
as
follows:
1.
The
selected
303(
d)
waterbodies
were
exported
as
a
new
shapefile
named
rchs_
win_
500m.
shp.
2.
This
shapefile
was
added
to
the
View
in
ArcView.
3.
A
script
was
run
called
"
Find
Nearest
Feature"
that
finds
the
nearest
intake
to
each
of
the
reaches
in
the
new
shapefile
rchs_
win_
500m.
shp.
This
resulted
in
a
table
that
associates
each
reach
with
an
intake,
listing
the
Rch_
Code,
the
Facility
Name,
and
distance
between
them.
This
table
is
named
"
NF
from
Rchs_
win_
500m.
shp
to
Intakes"
(
Nearest
facility
table).
4.
The
Nearest
facility
table
was
joined
to
the
intake
table
by
the
Facility
Name
field.
5.
The
Nearest
facility
table
was
joined
to
the
reach
table
by
the
Rch_
Code
field.
6.
The
Nearest
facility
table
was
joined
to
the
list_
impairments.
txt
table
on
the
Entity_
ID


List_
ID
fields.
This
puts
all
associated
data
together.
These
joins
are
shown
schematically
in
Figure
A­
1.
7.
The
combined
table
was
exported
as
a
dbf
file,
then
imported
it
into
Excel
and
cleaned
up
by
removing
unnecessary
fields
and
rearranging
fields
in
a
logical
manner.
8.
The
waterbody
names
from
fields
from
the
intake
table
(
WATERBODY1,
WATERBODY2,
and
WATERBODY3
t)
were
then
compared
with
the
corresponding
names
from
the
WATER_
BODY_
NAME
field
from
the
list_
impairments.
txt
table
to
see
if
they
match.
9.
The
final
processed
Excel
spreadsheet
was
brought
into
Microsoft
Access
as
a
table,
and
a
query
was
run
to
see
how
many
DISTINCT
facilities
have
intakes
from
the
303(
d)
water
bodies.

Figure
A­
1.
Schematic
diagram
of
table
relationships
to
show
how
the
water
body
name
fields
can
be
compared
between
the
facility
intake
data
and
the
303(
d)
data.
