Air
Emission
Research
Plan
for
Dairy
Preliminary
2­
25­
04
Purpose:
The
purpose
of
this
research
project
is
to
provide
quality­
assured
air
emission
data
from
representative
dairy
farms
in
the
U.
S.,
to
U.
S.
EPA,
in
the
effort
to
determine
which
farms
might
fall
under
regulatory
authority.
Following
sound
scientific
principles,
this
project
will
collect
new
data
and
aggregate
existing
emissions
data
from
previous
studies.
These
data
will
serve
as
the
beginning
of
a
database
to
which
new
data
can
be
added
as
emissions
and
against
which
control
technologies
can
be
compared.

Objectives:
New
data
will
be
collected
to
create
tools
(
i.
e.
look
up
tables,
charts,
or
models)
to
meet
the
following
objectives.

 
Determine
whether
individual
dairy
farms
are
likely
to
emit
particulate
matter
(
both
total
suspended
particulate
[
TSP],
particles
smaller
than
10
and
2.5
microns
[
PM10
&
PM2.5]),
and
volatile
organic
compounds
(
VOC)
in
excess
of
applicable
Clean
Air
Act
(
CAA)
thresholds.
 
Determine
whether
individual
dairy
farms
are
likely
to
emit
ammonia
(
NH3)
and
hydrogen
sulfide
(
H2S)
in
excess
of
applicable
Comprehensive
Environmental
Response,
Compensation
and
Liability
Act
(
CERCLA)
reporting
requirements.

BACKGROUND
Dairy
operations
are
naturally
ventilated
buildings
with
different
manure
handling
systems.
Measurement
of
the
emission
from
these
operations
is
to
be
conducted
with
a
series
of
measurement
systems
that
provide
a
concentration
measurement
along
a
path
that
would
be
representative
of
the
emission
plume
from
the
building.
In
order
to
estimate
the
emission
rate
it
is
necessary
to
couple
the
concentration
with
a
measurement
of
the
wind
flow
through
the
building
or
facility.

Manure
storage
sites
could
be
either
liquid
(
lagoons
or
slurry
store)
or
piles
of
solid
materials.
These
sites
represent
a
different
source
area
for
emissions
than
buildings
and
will
have
to
be
considered
separately
in
the
measurement
scheme.

The
protocols
that
are
developed
for
these
studies
are
based
on
the
following
assumptions.

1.
The
buildings
are
naturally
ventilated
and
require
a
measurement
method
that
captures
the
entire
plume
leaving
the
building.
Mechanically
ventilated
facilities
are
beginning
to
enter
the
industry.
2.
Manure
storage
is
separate
from
the
building
and
will
have
to
be
measured
as
a
distinct
entity
as
part
of
the
farm
emission
factor.
3.
The
primary
emission
sources
are
the
housing
and
feeding
areas
and
manure
storage.
4.
There
is
a
large
diversity
among
dairy
operations
across
the
United
States
and
although
there
are
similar
characteristics
in
general
structure,
the
difference
in
building
design,
management,
and
climate
require
measurements
of
facilities
that
represent
these
factors.
5.
Measurements
will
be
conducted
at
facilities
which
represent
a
diversity
of
systems
in
three
general
areas;
California
and
southern
US,
Northeast
US,
and
Upper
Midwest.

SITE
SELECTION
Milk
production
facilities
include
cattle
(
dry
cows,
lactating
cows,
and
replacement
heifers)
and
calves.
The
partially
open
barns
range
from
those
with
windows
and
flaps
to
fully
open
free
stalls.
The
buildings
are
most
typically
naturally
ventilated
except
for
some
mechanically­
ventilated
freestall
and
tie
stall
houses.
The
naturally
ventilated
barns
range
from
partially
open
barns
with
windows
and
flaps
to
fully­
open
free
stalls.
External
manure
storages
generally
consist
of
either
earthen
basins
that
store
undiluted
manure
collected
from
the
barn,
or
anaerobic
treatment
lagoons
that
treat
manure
that
is
diluted
by
a
factor
of
about
5:
1.
Manure
collection
systems
generally
are
either
scrape
or
flush.
Four
dairy
sites
that
consider
climate,
and
types
of
ventilation,
manure
collection,
and
manure
storage
have
been
identified
by
the
dairy
industry
for
collecting
the
comprehensive
air
emission
data
required
by
this
study
(
Table
1).
Final
site
selections
will
also
depend
on
site­
specific
factors
including:
representativeness
of
facility
age,
size,
design
and
management,
and
cow
diet
and
genetics.
The
facility
should
be
isolated
from
other
potential
air
pollution
sources
and
have
potential
for
testing
mitigation
strategies.
Producers
should
be
willing
to
make
changes
and
keep
extra
records
to
facilitate
a
quality
study.

Table
1.
Recommended
types
and
locations
of
dairy
facilities
to
be
monitored
in
this
study.
Region
Site
type
Ventilation
Manure
collection
Manure
storage
Midwest
Freestall
Natural
Flush
Lagoon
Northea
st
Freestall
Natural
Scrape
Basin
West
Open*
freestall
Natural
Flush
Lagoon
South
Open
freestall
Natural
Scrape
Basin
*
Cattle
are
free
to
walk
outside
in
open
freestall
barns.

MEASUREMENT
PROTOCOLS
Naturally
Ventilated
Buildings
To
achieve
the
most
representative
measurements
of
the
emissions
of
the
gases,
it
is
recommended
that
a
FTIR
system
be
used
to
quantify
the
concentration
of
NH3,
CO2,
and,
at
levels
above
50
ppb,
H2S
in
various
paths
through
the
atmosphere.
A
variation
of
the
horizontal
gradient
method
called
radial
plume
mapping
utilizing
multiple
paths
through
the
airflow
from
the
building
measures
the
concentrations.
The
FTIR
method
is
selected
because
of
the
extreme
turbulence
adjacent
to
the
building
and
the
lack
of
a
defined
plume
in
this
area
of
the
facility.
A
scanning
system
rotates
among
the
paths
to
provide
a
serial
measurement
of
the
paths
utilizing
horizontally
and
vertically
located
retro­
reflectors.
A
computer
calculates
the
concentration
gradients
in
real­
time.
FTIR
measurements
would
be
coupled
to
two
sonic
anemometers
positioned
at
two
locations
along
the
length
of
the
building.
This
will
provide
the
wind
flow
measurements
needed
to
estimate
the
flux
from
the
measured
concentrations.

Particulate
load
would
be
sampled
using
a
series
of
particle
samplers
located
with
a
sampling
height
of
5
m
adjacent
to
one
of
the
sonic
anemometer
towers.
These
units
would
be
designed
to
collect
2.5
µ
m,
10
µ
m
and
TSP
values
Volatile
organic
compounds
(
VOCs)
would
be
sampled
at
the
same
position
as
the
particulate
samples
for
the
building
emissions.
VOC
emissions
from
the
manure
storage
would
be
sampled
with
a
system
located
both
upwind
and
downwind
of
the
manure
storage
system.
These
units
would
be
positioned
at
a
height
of
2
m.

Mechanically
Ventilated
Buildings
Mechanically
ventilated
buildings
have
begun
to
be
used
in
the
dairy
industry.
If
warranted
by
current
or
future
use,
a
mechanically
ventilated
facility
will
be
included
in
this
project.
An
on­
site
instrument
shelter
(
OSIS)
will
house
the
equipment
for
monitoring
pollutant
concentrations
at
representative
air
inlets
and
outlets
(
primarily
by
air
extraction),
barn
airflows,
and
operational
processes
and
environmental
variables.
Sampling
will
be
conducted
for
24
months
with
data
logged
every
60
s.
Data
will
be
retrieved
with
network­
connected
PCs,
formatted,
validated,
and
delivered
to
EPA
as
hourly
averages
for
subsequent
calculations
of
emission
factors.
A
multipoint
air
sampling
system
in
the
OSIS
will
draw
air
sequentially
from
representative
locations
(
including
ambient)
at
the
barns
and
deliver
selected
streams
to
a
manifold
from
which
on­
line
gas
monitors
draw
their
sub
samples.
The
pollutants
targeted
for
measurement
will
be
evaluated
as
follows:
 
Ammonia
will
be
measured
using
chemiluminescence
or
photoacoustic
infrared.
 
Hydrogen
sulfide
will
be
measured
with
pulsed
fluorescence.
 
Carbon
dioxide
will
be
measured
using
photoacoustic
infrared.
 
TSP
will
be
measured
using
an
isokinetic
multipoint
gravimetric
method.
 
PM2.5
will
be
measured
gravimetrically
with
a
federal
reference
method
for
PM2.5
at
least
for
one
month
per
site.
It
will
be
shared
among
sites.
 
PM10
concentrations
will
be
measured
in
real
time
using
the
tapered
element
oscillating
microbalance
(
TEOM)
at
representative
exhaust
locations
in
the
barn
and
ambient
air.
 
An
initial
characterization
study
of
barn
volatile
organic
compounds
(
VOCs)
will
be
conducted
on
one
day
during
the
first
month
at
the
first
site
(
site
1).
While
total
nonmethane
hydrocarbons
(
NMHC)
are
continuously
monitored
using
a
dual­
channel
FID
analyzer
(
Method
25A)
along
with
building
airflow
rate,
VOCs
will
be
sampled
with
replication
at
two
barns
using
Silcosteel
canisters,
and
all­
glass
impingers
(
EPA
Method
26A).
Each
sample
will
be
evaluated
using
concurrent
gas
chromatography
 
mass
spectrometry
(
GC­
MS)
and
GC/
FID
for
TO
15
and
other
FID­
responding
compounds.
VOC
mass
will
be
calculated
as
the
sum
of
individual
analytes.
The
20
analytes
making
the
greatest
contribution
to
total
mass
will
be
identified
during
the
initial
characterization
study.
A
sampling
method
that
captures
a
significant
fraction
of
the
VOC
mass
will
be
chosen
for
the
remainder
of
the
study.
 
The
Method
26A
sampling
train
is
suitable
for
collecting
samples
for
analysis
of
formaldehyde
and
acetaldehyde
using
NCASI
94.02,
requiring
only
the
addition
of
spectrophotometry
for
the
detection
of
formaldehyde.
These
compounds
will
be
measured
during
the
initial
characterization
study
and,
if
not
found,
will
not
be
analyzed
during
subsequent
measurements.
 
Total
VOC
mass
may
be
estimated
(
scaled)
by
multiplying
the
total
carbon
as
determined
by
Method
25A
by
the
molecular
weight/
carbon
weight
ratio
derived
from
GC­
MS
or
GC­
FID
speciation.
This
should
account
for
the
VOCs
that
are
not
identified
by
GC
methods
due
either
to
sampling
bias
or
the
analytical
procedures
used,
although
some
error
is
anticipated
due
to
the
imprecise
response
of
the
Method
25A
FID
to
oxygenated
compounds.
Acceptance
of
a
scaling
factor
will
depend
on
whether
the
Method
25A
analyzer
response
is
reasonable
based
on
the
manufacturer's
stated
response
factors,
bench­
scale
verification,
or
judgmental
estimation
of
the
mass
of
unaccounted
for
VOCs.
 
By
the
middle
of
the
second
month,
FASS
will
report
results
of
the
initial
VOC
characterization
to
EPA
with
recommendations
on
the
appropriateness
and
validity
of
the
selected
methodologies.
 
Quarterly
VOC
samples
using
the
selected
VOC
sampling
method
will
occur
at
all
sites,
along
with
continuous
Method
25A
monitoring
at
site
1
throughout
the
study.

 
Method
25A
measurements
will
be
corrected
from
an
"
as
carbon"
basis
to
a
total
VOC
mass
basis
by
multiplying
them
by
the
mean
molecular
weight
per
carbon
atom
established
by
GC­
MS
evaluations
during
applicable
intervals
of
time.

Manure
Storage
Systems
Micrometeorological
techniques
will
be
used
to
estimate
emissions
of
NH3,
H2S,
and
a
limited
number
of
VOCs
from
manure
storage
systems
and
storages.
Fundamentally,
this
approach
will
use
optical
remote
sensing
(
ORS)
downwind
and
upwind
of
the
storage
coupled
with
3­
dimensional
(
3D)
wind
velocity
measurements
at
heights
of
2
and
12
m.
The
concentrations
of
NH3
and
the
various
hydrocarbons
will
be
made
using
open
path
Fourier
transform
infrared
spectroscopy
(
FTIR).
Measurements
of
H2S
(
and
NH3)
will
be
made
using
collocated
open
path
UV
differential
optical
absorption
spectroscopy
(
UV­
DOAS)
systems.
A
team
of
two
persons
with
two
scanning
FTIR
systems,
two
single­
path
UV­
DOAS
systems,
and
two
3D
sonics
with
supplementary
meteorological
instruments
will
move
sequentially
from
farm
to
farm.

Each
of
two
ORS
systems
will
be
oriented
parallel
to
the
storage
side
and
approximately
10
m
from
the
storage
edge.
Each
monostatic
FTIR
system
will
scan
five
retroreflectors;
three
mounted
at
1m
height
equally
dividing
the
length
of
the
open
path
along
the
storage
side
and
two
mounted
on
a
tower
at
heights
of
6
and
12
m
located
at
the
corners
down
the
adjacent
sides
of
the
storage,
resulting
in
scan
lines
down
each
of
the
four
sides
of
the
storage.
Two
bistatic
single­
path
UVDOAS
systems
will
be
located
at
a
nominal
2
m
height
within
2
m
laterally
of
the
FTIR
scan
lines
on
the
two
sides
of
the
storage
oriented
most
closely
with
prevailing
winds.

Emissions
will
be
determined
from
the
difference
in
upwind
and
downwind
concentration
measurements
using
two
different
methods­
an
Eulerian
Gaussian
approach
and
a
Lagrangian
Stochastic
approach.
The
Lagrangian
approach
is
based
on
an
inverse
dispersion
analysis
using
a
backward
Lagrangian
stochastic
method
(
bLS).
This
approach
will
be
used
to
estimate
NH3
emissions
from
concentration
measurements
made
using
the
FTIR
and
UV­
DOAS
systems
and
the
H2S
emissions
from
concentration
measurements
made
using
the
UV­
DOAS
systems.
The
emission
rate
for
NH3
will
be
the
ensemble
average
of
the
estimated
emissions
for
each
of
the
five
FTIR
scans
with
a
corresponding
error
of
the
emission
estimate.
The
Eulerian
approach
is
based
on
a
computed
tomography
(
CT)
method
using
Eulerian
Gaussian
statistics
and
a
fitted
wind
profile
from
the
two­
3D
sonics.
Measurements
of
air
and
storage
temperatures,
wind
speed
and
direction,
humidity,
atmospheric
pressure,
and
solar
radiation
will
be
also
be
conducted.

The
bLS
and
CT
emission
estimates
will
be
quality
assured
using
tests
of
instrument
response,
wind
direction
and
wind
speed,
stability,
turbulence
intensity,
differences
between
the
storage
and
the
surrounding
surface
temperatures,
differences
in
the
mean
and
turbulent
wind
components
with
height,
and
the
temporal
variability
in
emission.
Emission
estimates
using
the
CT
method
will
be
qualified
by
the
measured
fraction
of
the
estimated
plume.

To
estimate
VOC
emissions
from
lagoons,
samples
of
the
lagoon
liquid
will
be
collected
and
analyzed
for
VOCs,
and
the
EPA
model
WATER9
will
be
used
to
estimate
emissions
based
on
measured
VOC
concentrations,
pH,
and
other
factors.

Alternate
Techniques
1.
For
the
circuit
rider
system,
an
instrumental
system
such
as
the
DustTrak
by
TSI
could
be
used
for
continuous
particle
data
for
PM2.5
and
PM10.
These
systems
provide
optical
light
scattering
measurements
of
the
concentration
in
mg/
m3
and
cost
about
$
5000
per
point
including
an
environmental
shelter.

2.
A
radial
plume
mapping
approach
could
be
applied
to
the
manure
storage
systems
using
a
TDL
system
that
has
been
approved
by
EPA
for
use
in
the
aluminum
industry
in
a
single
path
mode.
1
upwind
and
3
downwind
paths
provide
the
same
type
of
data
as
the
FTIR
except
for
a
single
compound.
The
single
laser
is
scanned
via
fiberoptic
cables
to
the
individual
paths
with
a
complete
scan
taking
40
seconds.
It
provides
a
fast,
direct
measurement
of
the
flux
of
ammonia
from
these
manure
systems.
A
single
4­
channel
system
costs
$
68,000.

3.
It
is
recommended
that
one
short­
term
(
2­
week)
measurement
of
each
facility
be
made
with
a
LIDAR
system
to
measure
and
quantify
the
plume
dynamics
of
particles,
water
vapor,
and
ammonia
surrounding
the
facility.
This
recommendation
is
made
because
the
short­
term
measurements
will
be
made
at
different
times
throughout
the
year
and
will
be
placed
at
a
series
of
heights
based
on
experience.
These
associated
data
of
the
plume
structure
will
provide
evidence
of
representativeness
of
the
micrometeorological
measurements
for
the
emission
rates.

4.
It
is
recommended
that
each
building
site
be
instrumented
with
temperature
and
associated
sensors
to
provide
a
continuous
measurement
record
of
the
microclimate
within
and
adjacent
to
the
building.
These
systems
can
be
linked
with
sensors
to
measure
and
record
animal
activity
and
floor
temperature.
A
similar
system
would
be
located
to
measure
the
microclimate
of
the
manure
storage
system
and
would
include
air
temperature,
wind
speed,
wind
direction,
surface
temperature,
and
relative
humidity
of
the
manure
storage
system.
The
continuous
record
from
these
manure
storage
units
and
buildings
would
provide
a
reference
for
the
short­
term
measurements
made
with
the
FTIR
systems.
Costs
Table
2.
Estimated
costs
of
various
monitoring
configurations.
The
costs
do
not
include
university
overhead
charges.
Scenario
Total
Study
Cost
Two
barns
(
24
months)
$
545,000
Micromet
system
on
manure
storage/
lagoon
$
301,000
*
Setup
and
operation
of
the
non­
profit
organization,
FASS
charges,
outside
auditor
fees,
and
science
advisor
fees.
The
administrative
costs
are
estimated
to
be
$
1.2M
for
the
overall
project.
Assuming
three
species,
the
likely
cost
per
species
will
be
$
1.2M/
3=$
0.4M.
