United
States
Office
of
Air
Quality
Environmental
Protection
Planning
and
Standards
Agency
Research
Triangle
Park,
NC
27711
EPA­
454/
R­
00­
035s
September
2001
AIR
Final
Report
­
Volume
I
of
II
Testing
of
a
4­
Stroke
Diesel
Cycle
Oil­
fired
Reciprocating
Internal
Combustion
Engine
to
Determine
the
Effectiveness
of
an
Oxidation
Reduction
Catalyst
System
for
Reduction
of
Hazardous
Air
Pollutant
Emissions
FINAL
REPORT
TESTING
OF
A
4­
STROKE
DIESEL
CYCLE
OIL­
FIRED
RECIPROCATING
INTERNAL
COMBUSTION
ENGINE
TO
DETERMINE
THE
EFFECTIVENESS
OF
AN
OXIDATION
CATALYST
SYSTEM
FOR
REDUCTION
OF
HAZARDOUS
AIR
POLLUTANT
EMISSIONS
VOLUME
I
OF
II
Prepared
for:

Terry
Harrison
(
MD­
l
9)
Work
Assignment
Manager
SMTG,
EMC,
EMAD,
OAQPS
U.
S.
Environmental
Protection
Agency
Research
Triangle
Park,
NC
27711
September
200
1
Submitted
by:

PACIFIC
ENVIRONMENTAL
SERVICES,
INC.
5001
S.
Miami
Blvd.,
Suite
300
Research
Triangle
Park,
NC
27709­
2077
(
919)
941­
0333
FAX
(
919)
941­
0234
DISCLAIMER
Pacific
Environmental
Services,
Inc.
(
PES)
prepared
this
document
under
EPA
Contract
No.
68­
D­
01
­
003,
Work
Assignment
No.
l­
04.
PES
reviewed
this
document
in
accordance
with
its
internal
quality
assurance
procedures
and
approved
it
for
distribution.
The
contents
of
this
document
do
not
necessarily
reflect
the
views
and
policies
of
the
U.
S.
EPA.
Mention
of
trade
names
does
not
constitute
endorsement
by
the
EPA
or
PES.
TABLE
OF
CONTENTS
VOLUME
I
1.0
INTRODUCTION
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
l­
l
2.0
SUMMARY
OF
RESULTS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
1
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
EMISSIONS
TEST
LOG.
......................................
2­
l
ENGINE
PARAMETERS
AND
.................................
2­
3
FTIRS
AND
CEMS
MEASUREMENTS
..........................
2­
3
GCMS
MEASUREMENTS
....................................
2­
6
POLYNUCLEAR
AROMATIC
HYDROCARBON
(
PAH)
MEASUREMENTS..
.........................................
2­
9
DESTRUCTION
OF
ORGANIC
COMPOUNDS
BY
THE
CATALYST
2­
l
1
PARTICULATE
MATTER
MEASUREMENTS
...................
2­
15
FUEL
OIL
ANALYSIS
.......................................
2.15
.

3.0
SOURCE
DESCRIPTION
AND
OPERATION
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3­
1
3.1
ENGINE
DESCRIPTION
......................................
3
­
1
3.2
ENGINE
OPERATION
DURING
TESTING
.......................
3­
4
4.0
SAMPLING
LOCATIONS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
u
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
1
5.0
SAMPLING
AND
ANALYSIS
METHODS.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
l
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
LOCATION
OF
MEASUREMENT
SITES
AND
SAMPLE/
VELOCITY
TRAVERSE
POINTS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
l
DETERMINATION
OF
STACK
GAS
VOLUMETRIC
FLOW
RATE.
.
.
5­
3
DETERMINATION
OF
STACK
GAS
OXYGEN
AND
CARBON
DIOXIDE
CONTENT
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
4
DETERMINATION
OF
STACK
GAS
MOISTURE
CONTENT
.
.
.
.
.
.
.
5­
4
DETERMINATION
OF
NITROGEN
OXIDES
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
6
DETERMINATION
OF
CARBON
MONOXIDE
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
6
DETERMINATION
OF
TOTAL
HYDROCARBONS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
7
DETERMINATION
OF
METHANE
AND
NON­
METHANE
HYDROCARBONS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
7
ii
TABLE
OF
CONTENTS
(
Concluded)

Pane
5.9
DETERMINATION
OF
GASEOUS
ORGANIC
HAP
USING
FTIRS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
8
5.10
DETERMINATION
OF
ORGANIC
HAPS
BY
DIRECT
INTERFACE
GCMS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
9
5.11
DETERMINATION
OF
POLYCYCLIC
AROMATIC
HYDROCARBONS
BY
CARB
429
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
12
5.12
DETERMINATION
OF
PARTICULATE
MATER
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
14
5.13
DETERMINATION
OF
FUEL
OIL
COMPOSITION
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
14
6.0
QUALITY
ASSURANCE/
QUALITY
CONTROL
PROCEDURES
ANDRESULTS....................................................
6­
l
6.1
FTIRS
QA/
QC
PROCEDURES
.................................
6­
l
6.2
CEMS
QA/
QC
PROCEDURES
.................................
6­
5
6.3
GCMS
QA/
QC
PROCEDURES
................................
6­
12
6.4
CARB
429
QA/
QC
PROCEDURES
.............................
6­
19
6.5
DATA
QUALITY
ASSESSMENT
..............................
6­
29
APPENDIX
A
APPENDIX
B
VOLUME
II
APPENDIX
C
APPENDIX
D
SUBCONTRACTOR
TEST
REPORT
­
COLORADO
STATE
UNIVERSITY
ENGINES
AND
ENERGY
CONVERSION
LABORATORY,
"
EMISSIONS
TESTING
OF
CONTROL
DEVICES
FOR
RECIPROCATING
INTERNAL
COMBUSTION
ENGINES
IN
SUPPORT
OF
REGULATORY
DEVELOPMENT
BY
THE
U.
S.
ENVIRONMENTAL
PROTECTION
AGENCY
(
EPA)
PHASE
3:
FOUR­
STROKE,
DIESEL
INTERNAL
COMBUSTION
ENGINES"

SUBCONTRACTOR
TEST
REPORT
­
EMISSION
MONITORING,
INC.
"
RESULTS
OF
DIRECT
INTERFACE
GCMS
TESTING
CONDUCTED
ON
A
2­
STROKE
LEAD
BURN
ENGINE"

SUBCONTRACTOR
TEST
REPORT
­
EASTERN
RESEARCH
GROUP,
INC.
"
CARB
METHOD
429:
SAMPLE
ANALYSIS"

CARB
METHOD
429
FIELD
DATA
.
.
.
111
LIST
OF
TABLES
Table
2.1
Table
2.2
Table
2.3
Table
2.4
Table
2.5
Table
2.6
Table
2.7
Table
2.8
Table
2.9
Table
2.10
Table
2.11
Emissions
Test
Log
...........................................
2­
2
Summary
of
Engine
and
Exhaust
Gas
Parameters
...................
2­
4
Stack
Concentrations
of
Detected
FTIRS
and
CEMS
Compounds
.......
2­
5
Stack
Concentrations
of
Detected
GCMS
Compounds
................
2­
7
Summary
of
Stack
Gas
and
Sampling
Parameters
CARB
429
Catalyst
Inlet
and
Outlet.
......................................
2­
10
Emission
Rates
of
Detected
PAHS
at
Catalyst
Inlet
.................
2­
12
Emission
Rates
of
Detected
PAHS
at
Catalyst
Outlet
................
2­
17
Mass
Flow
Scenarios
.........................................
2­
13
Removal
Efficiencies
of
Detected
Organic
Compounds
.............
2­
l
5
Method
IS0
8
178­
l
Particulate
Matter
Mass
Flow
Data
..............
2­
l
6
Summary
of
Fuel
Oil
Analysis
..................................
2­
17
Table
3.1
Table
3.2
Table
3.3
Table
3.4
Table
3.5
Engine
and
Catalyst
Specifications
Caterpillar
3508
EUI
(
4­
stroke,
diesel
cycle,
oil­
fired).
.....................................
I
...
3­
2
Summary
of
Nominal
Engine
Parameters
..........................
3­
3
Target
Engine
Operating
Conditions
During
Testing
.................
3­
5
Summary
of
Engine
Parameters
­
Caterpillar
3508
EUI
...............
3­
6
Summary
of
Engine
Parameters
During
Baseline
Runs.
...............
3­
7
Table
5.1
Summary
of
Sampling
and
Analysis
Methods
.......................
5­
2
Table
5.2
FTIRS
Analyzer
Specifications
..................................
5­
8
Table
5.3
Summary
of
Fuel
Oil
Analysis
Methods
..........................
5­
l
5
Table
6.1
Table
6.2
Table
6.3
Table
6.4
Table
6.5
Table
6.6
Table
6.7
Table
6.8
Table
6.8
Table
6.10
Table
6.11
Detection
Limits
of
FTIRS
and
CEMS
Compounds
..................
6­
6
Types
and
Frequencies
of
CEMS
Analyzer
Calibrations.
..............
6­
8
Summary
of
Fuel
Factor
Values
................................
6­
l
1
Summary
of
CEMS
Analytical
Detection
Limits
...................
6­
12
Summary
of
GCMS
Continuing
Calibrations
......................
6­
14
GCMS
Analyte
Spike
Recoveries
...............................
6­
15
Detection
Limits
of
GCMS
Compounds
at
Catalyst
Inlet
.............
6­
17
Detection
Limits
of
GCMS
Compounds
at
Catalyst
Outlet
............
6­
l
8
CARB
429
Sample
Train
­
Summary
of
Temperature
Sensor
Calibration
Data
.............................................
6­
20
CARB
429
Sample
Train
Summary
of
Pitot
Tube
Calibration
Data
.....
6­
21
CARB
429
Sample
Train
Summary
of
Dry
Gas
Meter
and
Orifice
Calibration
Data
.............................................
6­
22
Page
iv
LIST
OF
TABLES
(
Concluded)

Page
Table
6.12
Surnmary
of
CARB
429
Blank
Results
...........................
6­
25
Table
6.13
Summary
of
CARB
429
Surrogate
Recoveries
.....................
6­
26
Table
6.14
Detection
Limits
of
PAH
Compounds
at
Catalyst
Inlet
...............
6­
27
Table
6.15
Detection
Limits
of
PAH
Compounds
at
Catalyst
Outlet
.............
6­
28
Table
6.16
Summary
of
Engine
and
Method
Performance
.....................
6­
3
1
LIST
OF
FIGURES
Figure
1.1
Test
Program
Organization
and
Major
Lines
of
Communication
.
.
.
.
.
.
.
.
l­
3
Figure
4.1
Figure
4.2
Figure
4.3
Figure
4.4
Figure
5.1
Figure
5.2
Figure
5.3
Page
Inlet
Sample
Port
Locations
for
Velocity,
CARB
429,
FTIRS,
CEMS,
and
GCMS
Sampling
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
2
Inlet
Traverse
Point
Locations
for
Velocity
and
CARB
429
Sampling
.
.
.
.
4­
3
Outlet
Sample
Port
Locations
for
Velocity,
CARB
429,
FTIRS,
CEMS,
and
GCMS
Sampling
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
4
Outlet
Traverse
Point
Locations
for
Velocity
and
CARB
429
Sampling
.
.
4­
5
Schematic
Diagram
of
EECL
FTIRWEMS
Sampling
and
Analysis
System
..............................................
5­
6
Schematic
of
GCMS
Sampling
and
Analysis
System
................
5­
l
1
Schematic
Diagram
of
CARB
429
PAH
Sampling
Train
.............
5­
13
1.0
INTRODUCTION
The
United
States
Environmental
Protection
Agency
(
EPA)
is
investigating
Reciprocating
Internal
Combustion
Engines
(
RICE)
to
characterize
engine
emissions
and
catalyst
control
efficiencies
of
hazardous
air
pollutants
(
HAP).
This
document
describes
the
results
of
HAP
and
particulate
matter
(
PM)
emissions
testing
conducted
on
a
Caterpillar
3508
EUI
diesel
cycle,
oil­
fired,
4­
stroke
engine.
Early
in
1998,
several
industry
and
EPA
representatives
agreed
that
the
Caterpillar
3508
EUI
engine
at
the
Colorado
State
University's
(
CSU)
Engine
and
Energy
Conversion
Laboratory
(
EECL)
is
adequately
representative
of
existing
and
new
diesel
cycle
engines.
The
group
agreed
that
a
matrix
of
test
results
from
testing
conducted
at
the
EECL
could
be
used
to
develop
Maximum
Achievable
Control
Technology
(
MACT)
standards
for
RICE.
The
group
further
agreed
that
an
oxidation
catalyst
installed
on
the
Caterpillar
3508
EUI
could
be
used
to
determine
the
effectiveness
of
oxidation
catalysts
for
these
engines,
and
that
the
EPA
could
use
the
results
from
testing
at
CSU
as
the
basis
for
developing
the
MACT
standard
for
diesel
cycle
oil­
fired
engines.

PES
conducted
emission
testing
to
measure
pollutant
concentrations
in
the
exhaust
gas
both
up
and
downstream
of
an
oxidation
catalyst.
Englehard
Corporation
manufactured
the
catalyst
and
CSU
personnel
installed
it
on
the
engine.
Several
sampling
and
analysis
methods
measured
HAP
emissions
before
and
after
the
oxidation
catalyst.
Fourier
transform
infrared
spectroscopy
(
FTIRS)
measured
formaldehyde,
acetaldehyde,
and
acrolein.
Benzene,
toluene,
ethyl
benzene,
(
o,
m,
p)­
xylenes,
styrene,
hexane,
and
1,3­
butadiene,
were
measured
using
a
direct­
interface
gas
chromatograph
with
a
mass
spectrometer
detector
(
GCMS).
Continuous
emission
monitors
(
CEMS)
measured
oxygen,
(
O,),
carbon
dioxide
(
CO,),
nitrogen
oxides
(
NO,),
carbon
monoxide
(
CO),
total
hydrocarbons
(
THC),
methane
(
CH,)
j
and
non­
methane
hydrocarbons
(
NMHC).
Naphthalene
and
polycyclic
aromatic
hydrocarbons
(
PAHs)
[
acenaphthene,
acenapthylene,
anthracene,
benzo(
a)
anthracene,
benzo(
a)
pyrene,
benzo(
b)
fluoranthene,
benzo(
e)
pyrene,
benzo(
k)
fluoranthene,
benzo(
g,
h,
i)
perylene,
chrysene,
dibenzo(
a,
h)
anthracene,
fluoranthene,
fluorene,
indeno(
1,2,3­
cd)
pyrene,
2­
methylnapthalene,
perylene,
phenanthrene,
and
pyrene]
were
measured
by
California
Air
Resources
Board
(
CARB)
Method
429.

PM
testing
was
conducted
using
a
dilution
sampling
system.
A
sample
of
the
exhaust
gas
was
extracted
from
the
stack,
and
diluted
with
clean,
dry
air
then
passed
through
a
series
of
filters.
Particle
mass
was
determined
gravimetrically.
Fuel
oil
samples
were
collected
and
analyzed
to
determine
the
concentrations
of
target
metals
(
beryllium,
cadmium,
chromium,
lead,
maganese,
mercury,
nickel,
and
selenium).

PES
employed
four
subcontractors
for
this
effort.
The
CSU
EECL
provided
the
Final
Report
­
Caterpillar
3508
EUI
l­
l
September
200
1
facility
and
the
engine
for
the
test
program,
operated
the
engine
at
predefined
conditions,
and
recorded
engine
operational
data
during
the
testing.
In
addition,
CSU
EECL
personnel
operated
two
FTIRS
sampling
and
analysis
systems
and
two
CEMS
systems
that
measured
pollutants
and
diluents
in
the
exhaust
gas.
Emissions
Monitoring,
Inc.,
(
EMI)
of
Raleigh,
North
Carolina
provided
emissions
testing
services
and
two
direct­
interface
GCMS
sample
extraction
and
analysis
systems.
Eastern
Research
Group
(
ERG)
of
Morrisville,
North
Carolina,
prepared
filter
media
and
XAD­
2@
sorbent
resin
traps
and
analyzed
the
CARB
Method
429
samples
for
PAHs
using
Low
Resolution
Mass
Spectrometry
(
LRMS).
Galbraith
Laboratories,
Inc.
of
Knoxville,
Tennessee
provided
ultimate,
proximate,
and
metals
analysis
of
the
fuel
oil
samples
collected
by
PES.

Under
a
separate
work
EPA
assignment,
ERG
personnel
operated
an
EPA­
owned
dynamic
spiking
system
for
the
validation
of
the
FTIRS
systems
for
formaldehyde,
acetaldehyde,
and
acrolein.
Sierra
Instruments,
Inc.,
of
Monterey,
California
conducted
the
PM
testing
on
the
engine
in
conjunction
with
the
testing
conducted
by
PES.

The
test
program
organization
and
major
lines
of
communication
employed
during
this
project
are
presented
in
Figure
1.1.
The
balance
of
this
report
contains
the
following
Sections:

Section
2.0
Summary
of
Results
Section
3
.
O
Source
Description
and
Operation
Section
4.0
Sampling
Locations
Section
5
.
O
Sampling
and
Analysis
Methods
Section
6.0
Quality
Assurance/
Quality
Control
Procedures
and
Results
The
appendices
of
this
report
contain
the
engine
test
report
submitted
to
PES
by
CSU,
the
test
report
outlining
the
results
of
the
testing
for
the
GCMS
compounds
submitted
to
PES
by
EMI,
and
the
analytical
report
for
PAH
compounds
submitted
to
PES
by
ERG.
Also
included
are
PES
field
data
sheets
and
calibration
data
associated
with
the
PAH
sampling.

Final
Report
­
Caterpillar
3508
EUI
l­
2
September
200
1
PES
Project
Manager
Dennis
A.
Falgout
(
703)
471­
8383
Test
Plan
PES
Testing
PES
Analysis
PES
Subcontractor
CSU
EECL
Subcontractor
Eastern
Research
Group,
Inc.
I
Subcontractor
Eastern
Research
Subcontractor
Eastern
Research
Group,
Inc.

Figure
1.1.
Test
Program
Organization
and
Major
Lines
of
Communication
Draft
Final
Reports
PES
I
Subcontractor
CSU
EECL
I
i
Subcontractor
Emissions
Monitoring,
Inc.
I
Final
Report
­
Caterpillar
3508
EUI
l­
3
September
200
1
2.0
SUMMARY
OF
REXULTS
This
section
provides
summaries
of
the
stack
gas
parameters
and
HAP
emissions
measured
during
the
test
program.
Testing
of
the
Caterpillar
3508
EUI
engine
was
conducted
August
3
1
through
September
2,
1999
at
CSU's
Engines
and
Energy
Conversion
Laboratory
in
Fort
Collins,
Colorado.
The
following
sub­
sections
present
the
test
times
and
durations,
engine
and
stack
gas
parameters,
HAP
concentrations
before
and
after
the
oxidation
catalyst,
PM
concentrations
and
metal
content
in
the
fuel
oil.

The
measurements
that
were
made
of
fuel
flow
consumption
during
each
test
run
were
determined
by
EECL
personnel
to
be
inaccurate.
Pollutant
emission
data
is
presented
on
a
concentration
basis,
corrected
to
a
reference
oxygen
concentration
of
15%.
The
removal
efficiency
of
HAP
by
the
catalyst
is
calculated
using
pre­
and
post­
catalyst
concentrations
of
each
compound,
corrected
to
15%
oxygen.

2.1
EMISSIONS
TEST
LOG
During
the
test
period,
the
test
team
conducted
twenty­
five
test
runs.
These
test
runs
consisted
of
ten
5­
minute
Quality
Control
(
QC)
runs,
ten
33
­
minute
sampling
runs
for
collection
of
FTIRS,
CEMS
and
GCMS
data,
three
CARB
Method
429
runs,
and
two
5­
minute
baseline
runs.
PM
sampling
runs
were
conducted
just
before
and
just
after
each
33­
minute
FTIRS/
CEMS/
GCMS
sampling
run.
Table
2.1
presents
the
emissions
test
log.
The
test
log
summarizes
the
date
and
time
of
each
run
and
the
sampling
methods
used
during
that
particular
run.
Additional
discussions
of
the
engine
operating
parameters
may
be
found
in
Section
3.0
of
this
document.

In
Table
2.1,
the
sampling
runs
are
presented
in
the
order
of
their
conduct.
In
the
tables
that
follow
Table
2.1,
the
sampling
runs
are
presented
in
numerical
order.
The
test
team
decided
to
arrange
the
test
order
so
that
making
small
changes
in
engine
operation
could
accomplish
changes
from
condition
to
condition
rather
than
large
changes.
The
approach
reduced
both
the
time
between
test
runs
needed
to
effect
the
change
and
the
time
the
engine
needed
to
stabilize
after
the
change.
The
effect
on
the
test
program
was
that
we
did
not
conduct
the
engine
load
tests
in
the
order
in
which
the
Quality
Assurance
Project
Plan
(
QAPP)
presents
them.
To
maintain
consistency
with
the
QAPP,
we
did
not
change
the
numbers
denoting
the
engine
test
conditions.
The
reader
should
note
that
no
Runs
designated
5,6,7,
and
8
were
conducted
on
this
engine.
These
run
designations
describe
conditions
that
were
applicable
to
the
other
two
engines
tested
during
the
HAP
characterization
project.

Final
Report
­
Caterpillar
3508
EUI
2­
l
September
200
1
TABLE
2.1
EMISSIONS
TEST
LOG
Date
813
l/
99
Run
Time
1006­
1011
Run
ID
Run
1
QC
Sampling
Methodology
813
l/
99
1
1047­
1120
1
Run
1
I
FTIRS/
CEMS/
GCMS/
PM
813
1
I99
1
1335­
1340
1
Run
14
QC­
2
I
813
1
I99
I
1358­
143
1
I
Run
14
I
FTIRWEMWGCMZYPM
813
l/
99
1
1645­
1650
1
Run
13
QC
I
813
1
I99
1
1659­
1732
1
Run
13
I
FTIRS/
CEMS/
GCMS/
PM
813
1
I99
1
1831­
1836
1
Run
10
QC
I
813
1
I99
1
1859­
1932
1
Run
10
I
FTIRWEMWGCMWM
813
1
I99
1
1859­
2059
1
PAH
1
(
Run
10)
I
CARB
Method
429
813
l/
99
1
2220­
2225
1
Run
9
QC
I
813
1
I99
1
2243­
2316
1
Run
9
I
FTIRS/
CEMS/
GCMS/
PM
9/
I/
99
1
1055­
1100
1
Run
4
QC
I
9/
l/
99
1
1116­
1149
1
Run
4
I
FTIRS/
CEMS/
GCMS/
PM
9/
l/
99
1
1309­
1314
1
Baseline
No.
1
I
9/
l/
99
1
1340­
1345
I
Run
11
QC
I
9/
l/
99
I
1517­
1550
I
Run
11
I
FTIRS/
CEMS/
GCMS/
PM
9/
l/
99
1
1624­
1629
1
Run
12
QC
I
9/
l/
99
1
1637­
1710
1
Run
12
I
FTIRS/
CEMS/
GCMS/
PM
9/
l/
99
I
1830­
1835
I
Run
2
QC
I
­

9/
l/
99
1
1845­
1918
1
Run
2
I
FTIRWCEMWGCMW'M
911199
1
1912­
2124
1
PAH
2
(
Run
2)
CARB
Method
429
912199
1
0949­
0954
1
Run
3
QC
I
912199
1
1005­
1038
1
Run
3
I
FTIRS/
CEMS/
GCMS/
PM
912199
I
1038­
1218
I
PAH
3
(
Run
3)
I
CARB
Method
429
912199
I
1250­
1255
I
Baseline
No.
2
I
Final
Report
­
Caterpillar
3508
EUI
2­
2
September
200
1
2.2
ENGINE
PARAMETERS
Table
2.2
summarizes
some
engine
and
exhaust
gas
parameters
measured
and/
or
calculated
during
the
test
program.
The
EECL's
Data
Acquisition
System
(
DAS),
monitored
and
recorded
approximately
200
engine
operating
parameters,
and
gas
temperatures,
and
concentrations
of
02,
CO,,
and
moisture
at
the
catalyst
inlet
and
exhaust.
(
The
test
report
generated
by
CSU
EECL
is
presented
in
Appendix
A).

2.3
FTIRS
AND
CEMS
MEASUREMENTS
Table
2.3
summarizes
the
in­
stack
and
corrected
concentrations
of
the
FTIRS
target
compounds
(
formaldehyde,
acetaldehyde,
and
acrolein)
and
the
CEMS
target
compounds
(
carbon
monoxide,
nitrogen
oxides,
THC,
methane,
and
NMHC).

EECL
operated
two
FTIRS
sampling
and
analysis
systems
to
quantify
concentrations
of
the
target
compounds.
Exhaust
gas
samples
were
extracted
from
locations
upstream
and
downstream
of
the
oxidation
catalyst,
conditioned,
and
transported
to
a
Nicolet
Rega
7000
FTIRS
(
upstream
location)
and
a
Nicolet
Magna
560
FTIRS
(
downstream
location).
The
upstream
FTIRS
also
measured
the
moisture
content
in
the
exhaust
gas.
Moisture
measurements
by
the
downstream
FTIRS
were
determined
by
EECL
personnel
to
be
inaccurate.
A
carbon
balance
method
calculated
the
moisture
concentration
at
the
downstream
sampling
location.

EECL
reported
formaldehyde,
acetaldehyde,
and
acrolein
values
upstream
of
the
catalyst
during
every
run.
Inspection
of
EECL's
FTIRS
detection
limit
(
DL)
data
for
these
compounds
showed
that
reported
formaldehyde
concentrations
were
approximately
10
times
the
formaldehyde
DLs,
acetaldehyde
concentrations
were
between
1
and
2
times
the
acetaldehyde
DLs,
and
acrolein
concentrations
were
less
than
the
acrolein
DLs.
PES
changed
a
reported
value
to
"
Not
Detected"
(
ND)
if
the
reported
value
was
less
than
the
DL
value
reported
by
EECL.
Run
by
run
detection
limit
values
of
the
FTIRS
compounds
are
presented
in
Table
6.1.

Downstream
of
the
catalyst,
EECL
reported
formaldehyde
values
for
ten
of
the
thirteen
runs.
Concentrations
for
the
other
three
runs
were
all
reported
as
zeros.
At
the
downstream
location,
only
one
(
Run
1)
of
the
ten
runs
where
formaldehyde
values
were
reported
was
greater
than
the
formaldehyde
DL.
PES
changed
the
reported
formaldehyde
values
downstream
of
the
catalyst
to
ND
for
every
run
except
for
Run
1.
EECL
reported
acetaldehyde
values
for
every
run
downstream
of
the
catalyst.
The
reported
downstream
Final
Report
­
Caterpillar
3508
EUI
2­
3
September
200
1
TABLE
2.2
SUMMARY
OFENGINEAND
EXHAUST
GAS
PARAMETERS
rptn
­
revolutions
per
rrinute
t#~
­
reciprocal
of
%
Excess
Air
f
t­
lb
­
foot­
pounds
"
F
­
degrees
Fahrenheit
bhp
­
brake
horsepower
%
vd
d.
b.
­
%
volume
dry
basis
Final
Report
­
Caterpillar
3508
EUI
2­
4
September
200
1
TABLE
2.3
STACK
CONCENTRATIONS
OF
DETECTED
FllRS
AND
CEMS
COMPOUNDS
ppmvd
­
parts
per
Milton
by
volume,
dry
basis
ppnwd
Q
15%
02
­
parts
per
rrittii
by
voluma,
dry
basis,
normaitzed
to
15%
oxygen
ppnww
­
parts
per
niitiin
by
volume,
wet
basis
Final
Report
­
Caterpillar
3508
EUI
2­
5
September
200
1
concentrations
were
between
one
and
three
times
the
values
of
the
DLs.
These
values
were
approximately
twice
the
acetaldehyde
concentrations
reported
upstream
of
the
catalyst.
EECL
reported
zeros
for
acrolein
for
all
runs
downstream
of
the
catalyst.
PES
changed
the
reported
acrolein
concentrations
for
all
runs
downstream
of
the
catalyst
to
ND.

Table
2.3
also
shows
in­
stack
and
corrected
concentrations
of
the
CEMS
compounds.
EECL
personnel
operated
two
CEMS
sampling
and
analysis
systems.
Engine
exhaust
gas
samples
were
extracted
from
locations
upstream
and
downstream
of
the
catalyst.
These
samples
were
filtered
and
dried
(
except
that
the
methane/
non­
methane
sample
was
not
dried),
then
transported
to
the
CEMS
analyzer
racks.
The
CEMS
detected
all
target
compounds
at
both
the
inlet
and
the
outlet
locations.
Table
6.1
presents
the
CEMS
detection
limits
for
each
run.

2.4
GCMS
MEASUREMENTS
Table
2.4
presents
the
in­
stack
concentrations
and
concentrations
corrected
to
15%
oxygen
of
the
GCMS
compounds
(
1,3­
butadiene,
hexane
benzene,
toluene,
ethyl
benzene,
(
o,
m,
p)­
xylenes,
and
styrene).
EM1
personnel
operated
two
Inficon
Portable
Gas
Chromatographs
with
Mass
Spectrometer
Detectors.
Gas
samples
for
GCMS
analysis
were
extracted
from
the
up
and
downstream
locations
through
a
heated
probe
and
quartz
fiber
filter,
then
transported
via
a
heated
Teflon@
sample
line
to
a
Peltier
condenser
that
continuously
removed
moisture.
The
sample
was
then
co­
mixed
with
an
internal
standard
mixture
(
in
a
constant
ratio
of
10:
1)
in
the
GC
sampling
loop
for
1
minute
before
injection
into
the
GCMS.
After
purging
the
sample
loop
for
1
minute,
the
GCMS
injected
the
sample
onto
the
separatory
column
to
resolve
the
target
compounds
for
quantization.
A
PC­
based
DAS
supported
each
GCMS
to
calculate
peak
areas
of
the
target
compounds.

The
large
quantity
of
small,
highly
adsorptive
soot
particles
found
at
both
the
inlet
and
outlet
sampling
locations
caused
analytical
problems.
The
soot
adsorbed
a
portion
of
the
target
analytes
(
i.
e.,
benzene
and
toluene).
This
effect
was
documented
by
analyte
spiking
(
See
Section
6).
EM1
took
special
precautions
including
modification
of
the
sampling
equipment
and
modification
of
sampling
procedures
to
minimize
the
effects
of
the
soot
on
the
measurement
results.
(
See
detailed
discussion
in
Section
5)
In
spite
of
these
efforts
to
minimize
collection
of
particulate
matter,
concentration
results
for
benzene
and
toluene
are
consistently
biased
low.

Benzene
and
toluene
were
the
only
compounds
detected
upstream
of
the
catalyst.
Their
concentrations
varied
from
30
to
140
ppb.
These
are
near,
or
below
the
lowest
GCMS
calibration
point
of
100
ppb.
Benzene
was
the
only
compound
detected
at
the
catalyst
outlet
at
concentration
levels
ranging
from
30
to
150
ppb.
These
concentration
levels
are
near,
orB
below
the
lowest
GCMS
calibration
point
of
100
ppb.

Final
Report
­
Caterpillar
3508
EUI
2­
6
September
200
1
TABLE
2.4
STACK
CONCENTRATIONS
OF
DETECTED
GCMS
COMPOUNDS
Pun
ID
Run1
Run2
Run3
Run4
Run9
Run10
Run11
Run12
Run13
Run14
PAH
1
PAH
2
PAH
3
Catalyst
Inlet
PPbd
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
I,
3­
Butadiene
iexane
ppbd
@
15%
02
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
iwbd
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
N/
A
ppbti
@
15%
02
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
ppbd
40
50
120
90
33
35
53
58
33
40
35
50
N/
A
jenzene
ppbut
@
15%
02
22
30
66
46
19
20
30
33
19
23
20
30
N/
A
PPbd
140
loo
0
loo
loo
100
loo
IO0
loo
loo
loo
loo
N/
A
­
0luene
ppb\
rd@
15%
02
79
100
0
loo
loo
loo
loo
loo
loo
loo
loo
loo
N/
A
ppbd
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
Ithyl
Benzene
n/
p­
Xylene
Zyrene
ppbti
@
15%
02
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
ppbd
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
­
Xy
lene
ppkd
@
15%
02
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
PPbd
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
ppbuj
@
15%
02
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
PPbd
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
N/
A
ppbuj
@
15%
02
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
ppbvd­
parts
perbilliinbyvolume,
drybasis
ppbvd@
15%
02­
parts
perbilliin
byvokrme,
drybask,
noralizedto
lS%
ox)
gen
ND
­
Referto
Table6.7forIxm­
bywndetection
limits
atthecatatystinlet,
andTable68forrun­
by­
rundetection
limits
atthecatalptoutkt.

N/
A
­
Notavailable.
GCMS
dataws
notcollectedduringthis
samplingrun.

Final
Report
­
Caterpillar
3508
EUI
2­
7
September
200
1
TABLE
2.4
(
Concluded)

STACK
CONCENTRATIONS
OF
DETECTED
GCMS
COMPOUNDS
Run
LD
Run1
Run2
Run3
Run4
Run9
Run10
Run11
Run12
Run13
Run14
PAH
1
PAH
2
PAH
3
Catalyst
Outlet
ppbd
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
I,
3­
Butadiene
ppbuj
@
15%
02
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
Hexane
PPbd
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
ppb\
rd
@
15%
02
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
Benzene
Toluene
PPbd
0
100
100
100
IO0
100
ND
ND
100
0
100
100
N/
A
ppbti@
l5%
02
0
0
loo
loo
loo
0
ND
ND
0
0
0
100
N/
A
ppbd
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
ppbul@
15%
02
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
ppbd
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
Ethyl
Benzene
ppbui
@
15%
02
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
m/
pXylene
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
ppb\
rd
@
15%
02
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
Styrene
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
ppbu4
@
15%
02
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
PXylene
PPbd
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
ppbui
@
15%
02
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
N/
A
ppbvd­
park
perbillbnbyvolume,
dtybasis
ppbvd
@
I
15%
02­
parts
perbillbn
byvolume,
d~
basis,
nortnallzed
to
lS%
oxygen
ND
­
Referto
Table6.7forrun­
by­
rundetectbn
limits
atthecatalyst
lnlet,
andTable6.8forrun­
byiundetectbn
limits
atthecatalystoutlet.

N/
A
­
Not
available.
GC
M
S
data
was
not
collected
during
this
sampling
run.

Final
Report
­
Caterpillar
3508
EUI
2­
8
September
200
1
2.5
POLYNUCLEAR
AROMATIC
HYDROCARBON
(
PAH)
MEASUREMENTS
PES
used
CARB
Method
429
to
collect
samples
of
the
engine
exhaust
for
determination
of
PAHs
up
and
downstream
of
the
catalyst.
A
sample
of
the
exhaust
gas
was
extracted
through
a
glass
nozzle,
heated
glass­
lined
probe,
a
heated
quartz
filter,
and
a
chilled
sorbent
trap
containing
XAD­
2
sorbent
resin.
The
resin
was
extracted
and
combined
with
the
front­
half
train
rinses
and
the
filter
and
analyzed
for
PAH
content
by
ERG
using
Low
Resolution
Mass
Spectrometry.
PES
did
three
CARB
429
sampling
runs.
Each
run
consisted
of
simultaneous
sampling
at
both
the
upstream
and
downstream
locations.
The
first
PAH
run
was
conducted
at
Run
Condition
No.
10,
and
the
second
PAH
run
was
conducted
at
Run
Condition
No.
2.
The
last
PAH
run
was
conducted
at
Run
Condition
No.
3.
Table
2.5
presents
stack
gas
and
sample
train
parameters
for
the
CARB
429
testing.

The
dangerous
characteristics
(
5
inches
Hg
static
pressure,
800
"
F
stack
temperature,
2,000
ppm
NO,
concentration)
of
the
exhaust
gas
and
inadequate
access
to
one
of
the
sample
ports
mandated
that
sampling
be
conducted
through
only
one
sampling
port
upstream
of
the
catalyst.
The
first
sampling
run
at
the
inlet
was
aborted
after
50
minutes
of
sampling.
The
static
pressure
of
the
engine
exhaust
pushed
the
glass
liner
from
the
heated
probe.
PES
modified
the
sample
probe
by
replacing
the
glass
liner
with
a
stainless
steel
liner.
The
technicians
tightened
the
compression
nuts
and
ferrules
on
the
stainless
steel
liner
more
than
was
possible
on
the
glass
liner.
This
modification
enabled
collection
of
samples
for
the
remaining
CARB
429
sampling
runs.
Both
of
the
remaining
runs
were
120
minutes
in
duration.
PES
could
not
conduct
velocity
traverses
upstream
of
the
catalyst.
When
the
pitot
tube
was
inserted
into
the
duct
for
velocity
traverses,
the
reading
on
the
manometer
was
greater
than
10
inches
H,
O.
This
phenomenon
was
likely
due
to
the
high
static
pressure.
PES
did
a
velocity
traverse
downstream
of
the
catalyst.
This
traverse
was
used
to
calculate
gas
velocities
upstream
of
the
catalyst,
to
set
sampling
rates,
and
to
estimate
isokinetic
sampling
ratios
at
the
inlet.
The
estimated
isokinetic
sampling
ratios
for
the
CARB
429
runs
upstream
of
the
catalyst
were
68.4,
104.5,
and
92.5
percent
for
PAH
Runs
1,2,
and
3.

Downstream
of
the
catalyst,
PES
conducted
three
CARB
429
sampling
runs.
Each
sampling
run
was
120
minutes
in
duration.
Velocity
traverses
were
conducted
before
and
after
each
run,
and
the
results
used
to
set
sampling
rates
and
calculate
isokinetic
sampling
ratios.
The
isokinetic
sampling
ratios
were
97.1,96.2,
and
101.7
for
PAH
Runs
1,2,
and
3,
respectively.

Table
2.6
presents
the
mass
emission
rates
of
detected
PAH
target
compounds
upstream
of
the
catalyst.
Since
velocity
traverses
were
conducted
in
conjunction
with
the
PAH
sampling,
mass
flow
rates
for
the
PAH
compounds
could
be
calculated.
Several
of
the
PAH
target
compounds
(
acenapthylene,
benzo(
b)
fluoranthene,
benzo(
k)
fluoranthene,
chrysene,
fluoranthene,
fluorene,
napthalene,
phenanthrene,
and
pyrene)
were
detected
during
all
three
runs.
Benzo(
a)
anthracene
was
detected
during
Runs
PAH
2
and
PAH
3.
Acenapthene,
anthracene,
benzo(
g,
h,
i)
perylene,
benzo(
a)
pyrene,
dibenzo(
a,
h)
anthracene,

Final
Report
­
Caterpillar
3508
EUI
2­
9
September
200
1
TABLE
2.5
SUMMARY
OF
STACK
GAS
AND
SAMPLING
PARAMETERS
CARB
429
CATALYST
INLET
AND
OUTLET
Run
ID
PAH
1
PAH
2
PAH
3
Date
8131
I99
9/
l
/
99
g/
2/
99
Average
Catalyst
Inlet
Sampling
Duration,
minutes
Average
Sampling
Rate,
dscfm
a
Sample
Volume,
dscf
b
Exhaust
Gas
Temperature,
"
F
Stack
Pressure,
inches
Hg
O2
Concentration,
%
by
Volume
CO2
Concentration,
%
by
Volume
Moisture,
%
by
Volume
Exhaust
Gas
Volumetric
Flow
Rate:

acfm
`
Id
dscfm
`
ld
Stack
Gas
Velcocity,
ft/
s
d
lsokinetic
Sampling
Ratio,
%

Sampling
Duration,
minutes
Average
Sampling
Rate,
dscfm
a
Sample
Volume,
dscf
b
Exhaust
Gas
Temperature,
"
F
Stack
Pressure,
inches
Hg
O2
Concentration,
%
by
Volume
CO2
Concentration,
%
by
Volume
Moisture,
%
by
Volume
Exhaust
Gas
Volumetric
Flow
Rate:

acfm
'

dscfm
a
Stack
Gas
Velcocity,
ft/
s
lsokinetic
Sampling
Ratio,
%
50
120
120
0.755
0.621
0.555
37.745
74.575
66.598
800
800
800
29.84
29.84
29.84
10.7
11.1
10.1
7.5
6.8
7.8
8.4
9.3
9.3
0.644
59.639
800
29.84
10.6
7.4
9.0
2979
3035
3061
3025
1,140
1,150
1,160
1150
142
145
146
144
68.4
104.5
92.5
80.4
Catalyst
C
rttet
120
120
120
0.504
0.503
0.539
60.470
60.399
64.685
761
760
762
25.30
25.38
25.38
10.70
11.20
10.06
7.65
6.87
7.74
8.1
8.2
6.9
0.515
61.851
761
25.38
10.65
7.42
7.7
3,380
3,410
3,410
3400
1,140
1,150
1,160
1150
71.9
72.4
72.5
72.3
97.1
96.2
101.7
98.3
a
Dry
standard
cubic
feet
per
minute
at
68"
F
(
20"
C)
and
1
atm.

b
Dry
standard
cubic
feet
at
68"
F
(
20"
C)
and
1
atm.

'
Actual
cubic
feet
per
minute
at
exhaust
gas
conditions.

d
Calculated
from
standard
flow
conditions
observed
at
catalyst
outlet
(
SEE
text)
C
Final
Report
­
Caterpillar
3508
EUI
2­
10
September
200
1
TABLE
2.6
EMISSION
RATES
OF
DETECTED
PAHS
AT
CATALYST
INLET
Run
ID
PAH
1
PAH
2
PAH
3
Date
8131
I99
9/
l
/
99
g/
2/
99
Average
a
Acenaphthene
pg/
bhp­
hr
NDb
ND
ND
<
0.15
plb/
hr
ND
ND
ND
<
0.22
Acenaphthylene
pg/
bhp­
hr
13
20
35
28
plb/
hr
28
31
48
39
Anthracene
pg/
bhp­
hr
ND
6.3
ND
<
3.2
plb/
hr
ND
9.6
ND
<
4.9
Benzo(
a)
anthracene
pg/
bhp­
hr
ND
1.0
1.2
<
1.1
plb/
hr
ND
1.6
1.6
<
1.6
Benzo(
b)
fluoranthene
pg/
bhp­
hr
0.34
1.6
1.7
1.6
plblhr
0.75
2.4
2.3
2.3
Benzo(
k)
fluoranthene
pg/
bhp­
hr
0.14
0.43
0.41
0.42
plb/
hr
0.31
0.66
0.55
0.61
Benzo(
g,
h,
i)
perylene
pg/
bhp­
hr
ND
ND
ND
<
0.15
plb/
hr
ND
ND
ND
c
0.22
Benzo(
a)
pyrene
pg/
bhp­
hr
ND
ND
ND
<
0.15
plb/
hr
ND
ND
ND
C
0.22
Chrysene
pg/
bhp­
hr
1.4
2.8
3.3
3.0
plb/
hr
3.0
4.2
4.5
4.4
Dibenz(
a,
h)
anthracene
pg/
bhp­
hr
ND
ND
ND
c
0.15
plb/
hr
ND
ND
ND
<
0.22
Fluoranthene
pg/
bhp­
hr
3.5
16
25
21
plb/
hr
7.7
24
34
29
Fluorene
pg/
bhp­
hr
41
43
56
49
@
b/
hr
90
65
75
70
Indeno(
l,
2,3­
cd)
pyrene
pg/
bhp­
hr
ND
ND
ND
<
0.15
plb/
hr
ND
ND
ND
<
0.22
Naphthalene
pg/
bhp­
hr
179
215
381
298
plb/
hr
390
328
517
422
Phenanthrene
pg/
bhp­
hr
57
71
118
94
plb/
hr
124
107
160
134
Pyrene
pg/
bhp­
hr
14
25
35
30
vlb/
hr
30
38
48
43
*
Average
of
Runs
PAH
2
and
PAH
3;
Run
PAH
1
was
aborted
after
50
minutes
because
the
glass
liner
separated
from
the
nozzle.
PAH
1
data
are
presented
for
information
only.

b
ND
indicates
that
the
compound
was
not
detected.
Averages
include
detection
limits.

Table
6.14
presents
detection
limits
of
PAHs
at
the
catalyst
inlet.

Final
Report
­
Caterpillar
3508
EUI
2­
11
September
200
1
TABLE
2.7
EMISSION
RATES
OF
DETECTED
PAHS
AT
CATALYST
OUTLET
Run
ID
Date
Acenapt
hene
pg/
bhp­
hr
pi
b/
hr
PAH
1
PAH
2
8131199
9/
l/
99
ND
ND
ND
ND
PAH
3
912199
Average
ND
<
0.15
ND
<
0.25
Benzo(
a)
anthracene
ND
indicates
that
the
compound
was
not
detected.
Averages
include
detection
limits.

Table
6.15
presents
detection
linits
of
PAHs
at
the
catalyst
inlet.

Final
Report
­
Caterpillar
3508
EUI
2­
12
September
200
1
and
indeno(
1,2,3­
cd)
pyrene
were
not
detected
during
any
of
the
sampling
runs
upstream
of
the
catalyst.
The
table
presents
the
results
of
each
sampling
run.
Run
PAH
1
was
aborted,
so
the
averages
reported
are
the
averages
of
the
results
for
the
second
and
the
third
run.
For
those
compounds
that
were
not
detected,
the
average
is
the
average
of
the
in­
stack
mass
flow
rate
using
analytical
detection
limits
reported
by
ERG.
Table
6.14
presents
the
in­
stack
detection
limits
at
the
catalyst
inlet
for
each
compound
on
a
run­
by­
run
basis.

Table
2.7
presents
the
mass
emission
rates
of
detected
PAH
target
compounds
at
the
catalyst
outlet.
Napthalene
and
phenanthrene
were
detected
during
all
three
runs,
and
acenapthylene
and
fluorene
were
detected
during
two
of
the
three
runs.
Fluoranthene
was
detected
during
the
third
run
only.
None
of
the
remaining
PAH
compounds
were
detected
downstream
of
the
catalyst.
For
these
compounds,
the
(
3­
r@
average
detection
limit
is
presented
in
the
average
column.
Table
6.15
presents
the
in­
stack
detection
limits
at
the
catalyst
outlet
for
each
compound
on
a
run­
by­
run
basis.

2.6
DESTRUCTION
OF
HAP
BY
THE
CATALYST
There
are
five
possible
HAP
concentration
(
or
in
the
case
of
PAH
compounds,
mass
flow
rate)
combinations
that
can
occur
across
the
oxidation
catalyst.
Table
2.8
presents
these
combinations,
and
notes
whether
a
destruction
efficiency
is
reported.
Out
of
the
five
possible
combinations,
there
are
two
instances
where
the
destruction
efficiency
of
the
target
pollutant
is
reported.
If
pollutant
emissions
into
the
catalyst
(
Qin)
is
greater
than
pollutant
emissions
exiting
the
catalyst
(
QOJ,
%
DE
is
calculated.
If
the
pollutant
is
detected
entering
the
catalyst,
but
is
not
detected
exiting
the
catalyst,
%
DE
is
estimated
using
the
concentration
of
PAH
mass
flow
rate
at
the
inlet,
and
the
concentration
or
PAH
mass
flow
rate
corresponding
to
the
analytical
detection
limit
at
the
outlet.

TABLE
2.8
MASS
FLOW
SCENARIOS
Scenario
No.
Result
DE
Reported?

1
Qin
>
0;
Qout
'
0;
Qin
'
Qmt
YES
II
2
I
Qin
'
0;
Qout
=
ND
I
YES
II
3
I
Qin
<
Qout
I
NO
II
4
I
Qin
=
ND;
Qwt
'
0
I
NO
I
5
Qin
=
ND;
Q,",
=
ND
NO
Final
Report
­
Caterpillar
3508
EUI
2­
13
September
200
1
_
.
lll­­­­
­
_"
ll._
PES
calculated
the
catalyst
destruction
efficiency
of
several
target
compounds.
These
data
are
presented
in
Table
2.9.
Formaldehyde
is
the
only
FTIRS
compound
for
which
catalyst
removal
efficiencies
are
calculated.
Since
formaldehyde
was
detected
downstream
of
the
catalyst
during
Run
1,
the
formaldehyde
removal
efficiency
for
Run
1
is
calculated
with
quantified
mass
emissions
rates.
For
all
other
runs,
the
removal
efficiency
of
formaldehdye
is
estimated
using
the
value
of
detection
limit
values
at
the
catalyst
outlet.
Acetaldehyde
removal
efficiencies
are
not
calculated,
since
the
mass
flow
rate
of
acetaldehyde
downstream
of
the
catalyst
exceeds
the
upstream
mass
flow
for
every
run.
Acrolein
removal
was
not
calculated
because
acrolein
was
detected
neither
upstream
nor
downstream
of
the
catalyst.

PES
calculated
the
removal
of
carbon
monoxide
and
total
hydrocarbons,
but
none
of
the
remaining
CEMS
compounds.
The
mass
flow
rates
of
NOx
into
and
out
of
the
catalyst
were
essentially
the
same.
Toluene
was
the
only
GCMS
compounds
for
which
removal
efficiencies
were
calculated.
Since
toluene
was
not
detected
downstream
of
the
catalyst,
toluene
detection
limits
values
were
used
to
estimate
toluene
removal
efficiency.
Removal
efficiencies
for
benzene
were
not
calculated,
since
the
calculated
mass
flow
rates
downstream
of
the
catalyst
usually
exceeded
the
flow
rates
upstream
of
the
catalyst.

PES
calculated
removal
efficiencies
for
most
of
the
PAH
compounds
using
the
data
from
Runs
PAH
2
and
PAH
3.
Since
the
Run
PAH
1
was
aborted
at
the
inlet,
no
PAH
removal
efficiencies
have
been
calculated
for
this
run.
On
the
remaining
two
runs,
fluorene,
napthalene,
and
phenanthrene
were
detected
both
upstream
and
downstream.
of
the
catalyst.
Acenaphthylene
and
flouranthene
were
detected
at
both
locations
during
Run
PAH
3.

2.7
PARTICULATE
MATTER
MEASUREMENTS
Under
contract
to
the
Engine
Manufacturer's
Association,
Sierra
Instruments
conducted
testing
to
determine
the
mass
flow
rates
of
total
condensible
particulate
matter
upstream
and
downstream
of
the
catalyst.
EECL
included
the
results
of
these
tests
in
its
test
report
to
PES.
PES
has
reproduced
those
data
in
Table
2.9.
At
the
time
that
this
report
was
written,
PES
had
not
received
a
test
report
describing
the
testing
or
test
procedures.
Sierra
used
a
BG­
1
Micro­
Dilution
Test
Stand
(
which
they
manufacture)
to
extract,
dilute,
and
collect
entrained
particulate
matter
up­
and
downstream
of
the
catalyst.

2.8
FUEL
OIL
ANALYSES
PES
collected
three
samples
of
the
fuel
oil
that
was
used
to
fire
the
Caterpillar
engine.
One
sample
was
collected
each
day.
Galbraith
Laboratories,
Inc.
in
Knoxville,
Tennessee
did
ultimate
and
proximate
analysis
of
each
sample,
and
analyzed
each
sample
for
the
target
metals.
Table
2.10
presents
the
results
of
these
analyses.

Final
Report
­
Caterpillar
3508
EUI
2­
14
September
200
1
TABLE
2.9
REMOVAL
EFFICIENCIES
OF
DETECTED
ORGANIC
COMPOUNDS
I'
I
Values
preceeded
by
"
9'
indicate
that
the
value
of
the
Detectii
Limit
at
the
catalyst
outlet
was
used
to
estimate
removal
efficiency.
"­"
indicates
that
the
compound
was
not
detected
at
the
catalyst
inlet,
therfore
removal
effiiiency
was
not
calculated.

Final
Report
­
Caterpillar
3508
EUI
2­
15
September
200
1
TABLE
2.10
METHOD
IS0
8178­
l
PARTICULATE
MAITER
MASS
FLOW
DATA
Run
ID
Run
1
Run
2
Run
3
Run
4
Run
9
Run
10
Run
11
Run
12
Run
13
Run
14
Engine
Load,
bhp
988
692
615
878
988
989
988
988
989
989
Catalyst
Inlet
Test
Aa,
g/
bhp­
hr
0.05
0.07
0.06
0.07
0.05
0.10
0.04
0.10
0.05
0.09
Test
Bb,
g/
bhp­
hr
0.05
0.08
0.06
0.05
0.06
0.08
0.04
0.05
0.05
0.09
Awrage
0.05
0.08
0.06
0.06
0.06
0.09
0.04
0.08
0.05
0.09
Catalyst
Outlet
Test
Aa,
g/
bhphr
0.08
0.10
0.10
0.08
0.10
0.12
0.08
0.09
0.07
0.17
Test
Bb,
g/
bhphr
0.07
0.05
0.10
0.08
0.12
0.12
0.08
0.11
0.08
0.15
AIRrage
0.08
0.08
0.10
0.08
0.11
0.12
0.08
0.10
0.08
0.16
a
Test
Run
A
was
conducted
prior
to
the
33­
minute
FTIRSICEMSIGCMS
run
b
Test
Run
B
was
conducted
atier
the
33­
minute
FTlRS/
CEMS/
GCMS
run
Final
Report
­
Caterpillar
3508
EUI
2­
16
September
200
1
TABLE
2.11
SUMMARY
OF
FUEL
OIL
ANALYSES
C­
FO­
1
C­
FO­
2
C­
FO­
3
S/
31/
99
9/
l/
99
912199
2005
2135
1545
Average
Carbon,
%
w/
w
I
87.03
I
87.17
I
87.14
I
87.11
Hydrogen,
%
w/
w
I
13.19
I
13.33
I
13.44
I
13.32
Nitrogen,
%
w/
w
~~
I
<
0.5
I
<
0.5
I
<
0.5
I
<
0.5
Oxygen,
%
w/
w
I
<
0.5
I
<
0.5
I
<
0.5
I
<
0.5
Sulfur,
%
w/
w
I
0.04
I
0.04
I
0.06
I
0.05
Water,
%
w/
w
I
0.0073
I
0.0078
I
0.0065
I
0.0072
Ash,
%
w/
w
I
<
0.009
I
<
0.009
I
<
0.008
I
<
0.009
Heat
of
Combustion,
Btu/
ib
I
19347
I
18,668
I
18,888
I
­
18,968
Beryllium,
ppmw
I
<
l
I
cl
I
<
l
I
cl
Cadmium,
ppmw
I
<
l
I
<
l
I
<
l
I
~
Cl
Chromium,
ppmw
I
co.
1
I
co.
1
I
<
0.1
I
co.
1
Lead,
ppmw
I
<
l
I
0.4
I
0.54
I
~
<
0.6
Manganese,
ppmw
I
<
l
I
<
l
I
<
l
I
<
I
Mercury,
ppmw
I
<
0.57
I
<
0.61
I
<
0.62
I
<
0.60
Nickel,
ppmw
~
Selenium,
ppmw
<
l
<
l
­
cl
<
l
<
0.6
<
0.6
<
0.6
<
0.6
Final
Report
­
Caterpillar
3508
EUI
2­
17
September
200
1
3.0
SOURCE
DESCRIPTION
AND
OPERATION
This
section
presents
discussions
of
the
candidate
engine
and
the
catalyst
that
EPA
selected
for
the
test
program.
The
sections
that
follow
describe
the
engine
and
the
operation
of
the
engine
during
testing.

3.1
ENGINE
DESCRIPTION
The
Caterpillar
3508
EUI
stationary
internal
combustion
engine
is
an
eight­
cylinder,
4­
stroke,
diesel
cycle,
internal
combustion
engine
with
a
manufacturer's
sea
level
rating
of
775
brake­
horsepower
(
bhp)
at
1800
rpm.
The
pistons
are
6.7
inches
in
diameter
with
a
7.5­
inch
stroke.
Air
is
delivered
to
the
engine
via
a
pressurized
air
delivery
system;
air
manifold
pressures
are
controlled
by
the
EECL
process
control
system.
Engine
loading
is
controlled
by
a
computer­
controlled
water
brake
dynamometer.
Before
the
test
program
EECL
installed
an
oxidation
catalyst,
manufactured
by
Engelhard,
on
the
engine.
EECL
aged
the
catalyst
under
its
normal
operating
condition
(
i.
e.,
burned
in
the
catalyst)
before
the
test
program.
This
procedure
ensured
that
the
catalyst's
HAP
destruction
efficiency
approximated
the
HAP
destruction
efficiency
of
mature
catalysts
installed
on
4­
stroke
diesel
engines
in
industry.
Table
3.1
presents
specifications
of
the
engine
and
the
catalyst.
Table
3.2
presents
nominal
engine
operating
parameters.

The
compression
ignition
(
Diesel'
cycle)
engine
is
similar
to
the
spark
ignition
(
Otto*
cycle)
engine,
except
that
the
compression
ratio
is
higher,
and
air
alone,
rather
than
a
combustible
mixture,
is
admitted
into
the
cylinder
chamber
on
the
intake
stroke.
The
rapid
compression
of
the
air
during
the
compression
stroke
raises
its
temperature
higher
than
the
ignition
temperature
of
the
fuel.
During
the
first
part
of
the
expansion
stroke,
the
fuel
is
injected
into
the
cylinder
chamber
at
a
rate
such
that
the
combustion
maintains
constant
pressure
in
the
cylinder.
The
exhaust
stroke
pushes
the
combustion
products
from
the
chamber.

'
Named
for
Rudolf
Diesel,
who
began
design
of
the
compression
ignition
engine
in
1892.

*
Named
for
Nikolaus
A.
Otto,
who
built
a
highly
successful
four­
stroke
spark
ignited
engine
in
1876.
The
name
of
the
cycle
of
events
during
the
operation
of
the
engine
gradually
came
to
be
known
as
the
Otto
cycle.

Final
Report
­
Caterpillar
3508
EUI
3­
l
September
200
1
TABLE
3.1
ENGINE
AND
CATALYST
SPECIFICATIONS
Caterpillar
3508
EUI
(
4­
stroke,
diesel
cycle,
oil­
fired)

Engine
Classification
I
Four­
Stroke,
Diesel
Cycle
I
Caterpillar
3508
EUI
Manufacturer
and
Type:

Number
of
Cylinders:
I
8
Bore
and
Stroke:
I
6.7
in.
x
7.5
in.

Nominal
Engine
Speed:
I
1800
rpm
Catalyst
Classification
I
CO/
Odor
Control
Manufacturer:
I
Engelhard
Date
of
Manufacture:
.
Model
Number:
Serial
Number:
Item
Number:
unknown
unknown
unknown
unknown
'

Catalyst
Material:

Element
Size:

Number
of
Elements:
unknown
12
in.
x
16
in.
x
3.5
in.

4
Final
Report
­
Caterpillar
3508
EUI
3­
2
September
200
1
TABLE
3.2
SUMMARY
OF
NOMINAL
ENGINE
PARAMETERS
Parameter
Nominal
Value
Acceptable
Range
Designation
Torque,
&
lb
2880
f
2%
of
value
Primary
Speed,
rpm
I
1800
I
f
2%
of
value
I
Primary
Jacket
Water
Temperature
Outlet,
"
F
I
196
I
f
5%
of
value
I
Primary
Oil
Temperature
Outlet,
"
F
I
215
I
f
5%
of
value
I
Primary
Air
Manifold
Temperature,
"
F
I
150
I
f
5%
of
value
I
Primary
Air
Manifold
Pressure,
in.
Hg
I
5"
above
atm.
I
f
5%
of
value
I
Primary
Exhaust
Manifold
Pressure,
in.
Hg
Varies
with
AMP
f
5%
of
value
Primary
Injection
Timing
I
21'
BTDC
I
f
5%
of
value
I
Primary
Overall
Air/
Fuel
Ratio
I
3O:
l
I
f
5%
of
value
I
Primary
Inlet
Air
Humidity­
Absolute,
lb
H,
O/
lb
Air
Fuel
Flow,
gal/
hr
(
lb/
lx)
0.015
42
(
310)
f
10%
of
value
f
5%
of
value
Primary
Primary
Oil
Pressure
Inlet,
psi
I
67
I
f
5%
of
value
I
Secondary
Air
Flow,
scfm
Average
Exhaust
Temperature,
"
F
2150
f
5%
of
value
1000
f
5%
of
value
Secondary
Secondary
ft­
lb
­
foot­
pounds
rpm
­
revolutions
per
minute
"
F
­
degrees
Fahrenheit
in.
Hg
­
inches
mercury
column
BTDC
­
Before
Top
Dead
Center
lb
H,
O/
lb
Air
­
pounds
water
vapor
per
pound
of
air
gal/
hr
­
gallons
per
hour
lb/
hr
­
pounds
per
hour
psi
­
pounds
per
square
inch
scfm
­
standard
cubic
feet
per
minute
Final
Report
­
Caterpillar
3508
EUI
3­
3
September
200
1
3.2
ENGINE
OPERATION
DURING
TESTING
As
stated
in
Section
2
of
this
document,
four
types
of
test
runs
were
conducted
during
the
test
program:
quality
assurance
runs,
sampling
runs
for
FTIRS/
CEMS/
GCMS/
PM,
CARB
429
sampling
runs,
and
daily
baseline
runs.
The
operation
of
the
engine
during
these
various
runs
is
discussed
on
the
following
pages
and
in
the
following
tables.

Table
3.3
presents
the
test
matrix
for
the
Caterpillar
engine.
The
test
matrix
originally
presented
in
the
Quality
Assurance
Project
Plan
was
estimated
based
upon
the
manufacturer's
data.
When
the
engine
was
installed
and
operated
at
the
EECL,
the
estimates
were
found
to
be
inaccurate.
Therefore,
the
test
matrix
was
revised
to
represent
nominal
engine
operating
conditions.
Run
Conditions
5,6,7,
and
8
are
conditions
which
call
for
changes
in
the
air/
fuel
ratio.
These
conditions
are
not
applicable
to
the
Caterpillar
engine,
since
there
is
no
mixture
of
air
and
fuel.
These
conditions
were
applicable
during
testing
of
the
2­
stroke
and
4­
stroke
spark
ignition
engines.

During
the
test
program,
the
five
engine
operating
parameters
expected
to
have
the
greatest
impact
on
pollutant
formation
were
varied.
These
parameters
were:
engine
speed
(
measured
in
revolutions
per
minute
or
r­
pm),
engine
torque
(
measured
in
foot­
pounds
or
ft­
lb),
injection
timing
(
the
location
of
the
cylinder,
relative
to
top
dead
center,
at
the
time
of
fuel
injection),
intercooler
air
temperature
(
measured
in
degrees
Fahrenheit),
and
jacket
water
outlet
temperature
(
also
measured
in
degrees
Fahrenheit).
Table
3.4
presents
engine
parameters
that
were
recorded
during
each
test
run
and
their
percent
deviation
from
the
target
values.
The
target
engine
operating
parameters
were
met
for
every
run,
except
for
the
engine
equivalence
ratio.
Actual
equivalence
ratios
were
less
than
the
target
equivalence
ratios
for
every
run,
which
means
that
engine
excess
air
was
greater
than
the
target
for
every
run.

Table
3.5
presents
engine
parameters
during
baseline
test
points,
and
the
deviation
of
the
parameters
from
the
nominal
engine
parameters.
The
testing
was
conducted
over
a
period
of
three
days.
During
that
period
the
engine
did
not
run
continuously,
but
was
shut
down
each
night.
Test
accuracy
required
that
the
overall
engine
operation
did
not
change
over
the
three­
day
period.
The
stability
of
the
engine
over
this
period
was
shown
by
operating
the
engine
at
the
baseline
condition
for
one
5­
minute
period
on
the
second
and
third
day
of
testing.
Changes
to
the
baseline
parameters
would
have
indicated
a
change
in
the
overall
operating
characteristics
of
the
engine.
Distinguishing
between
emission
rate
changes
attributable
to
changes
in
the
independent
variables
and
emission
rate
changes
attributable
to
random
changes
in
the
performance
of
the
engine
would
have
been
impossible.

Final
Report
­
Caterpillar
3508
EUI
3­
4
September
200
1
TABLE
3.3
TARGET
ENGINE
OPERATING
CONDITIONS
DURING
TESTING
Operating
Conditions
Tested:

Condition
1
Condition
2
Condition
3
Condition
4
Condition
5
Condition
6
Torque
Air/
Fuel
Injection
Intercooler
Speed
Jacket
Water
("
A
of
Equivalence
Timing
Air
Temperature
@
pm)
maximum)
Ratio
Temperature
(
9)
("
BTDC)
(
em
09
H
H
N
S
S
S
H
L
N
S
S
S
L
L
N
S
S
S
L
H
N
S
S
S
Operating
Condition
Not
Applicable
For
This
Engine
Operating
Condition
Not
Applicable
For
This
Engine
Condition
7
I
Operating
Condition
Not
Applicable
For
This
Engine
Condition
8
Operating
Condition
Not
Applicable
For
This
Engine
Condition
9
H
H
N
S
L
S
Condition
10
H
H
N
S
H
S
Condition
11
H
H
N
S
S
L
Condition
12
H
H
N
S
S
H
Condition
13
H
H
N
L
S
S
Condition
14
H
H
N
H
S
S
H
=
1800
H
=
100
H
=
23
H
=
160
H=
206
L
=
1600
L=
70
1
N
=
0.58
s=
21
s=
150
S=
196
I
I
I
L=
19
I
L=
120
I
L=
186
Final
Report
­
Caterpillar
3508
EUI
3­
5
September
200
1
lABIE3.4
suMMARYmmGlNEP­­~
3508m
­
F&
m
10
­
lam
13
­
­
PAJi2
PAH3
1799
1600
1800
16tJl
0%
0%

2018
2019
2016
2016
0.1%
0.2%

21.0
21.0
21.0
21.0
0.0%
0.0%

150
149
130
130
3.4%
3.3%

195
195
196
196
4.7%
4.6%

691
615
0.49
0.54
Fan1
FaRl2
Rm3
fam4
Iam9
Rm
II
lam
12
lam
14
PAHI
1800
1800
0%
1800
1800
0%

2019
2019
2016
2016
0.1%
0.1%
0.2%
0.2%
0.1%

21.0
21.0
O.
O?
h
21.0
21.0
0.0%
21.0
21.0
O.
C%
21.0
21.0
O.
O?
h
21.0
21.0
O.
U?!
21.0
21.0
O.
oD!
19.0
19.0
O.
U?~
23.0
23.0
0.0%
21.0
21.0
0.0%
21.0
21.0
O.
oo/
o
150
150
4.1%
Mud
InjectionTining,
"
BTDC
Target
151
150
0.2%
151
150
0.1%
140
140
0.0%
257
160
­
0.5%
150
150
­
0.1%
151
150
0.2%
149
150
­
0.2%
150
150
0.0%
159
140
3.2%

194
1%

­
1.1%
195
196
4.5%
195
196
4.5%
195
196
­
0.7%
194
196
­
0.9%
185
186
­
0.7%
2cE
206
0.1%
194
196
­
1.0%
194
196
4.9%
194
196
­
l.
u?
h
988
692
615
878
988
989
988
988
989
989
Ec+
i&
me
Ratio,
Q
0.52
0.49
0.53
0.57
0.49
0.51
0.51
0.51
0.51
0.52
0.51
rpn­
rmh.&
ims
per
minute
It­
lb­
W
pods
Final
Report
­
Caterpillar
3508
EUI
3­
6
September
200
1
TABLE
3.5
SUMMARY
OF
ENGINE
PARAMETERS
DURING
BASELINE
RUNS
un
ID
Baseline
1
ngine
Speed,
rpm
Actual
Target
Detiation
1799
1800
­
0.05%

ngine
Torque,
ft­
lb
Actual
Target
Deviation
2884
2880
0.14%

Actual
jection
Timing,
"
BTDC
Target
Detiation
Actual
ltercooler
Air
Outlet
Temperature,
'
Target
Detiation
acket
Water
Temperature,
"
F
Actual
Target
Detiation
194
196
­
0.30%

Gl
Temperature,
"
F
Actual
Target
Detiation
231
215
2.37%

,
ir
Manifold
Pressure,
in.
Hg
rlet
Air
Humidity,
lb
H,
O/
lb
air
Actual
Target
Dedation
Actual
Target
Detiation
29.6C
29.6C
0.01%

O.
OlE
0.015
6.67%

)
il
Pressure,
psig
Actual
Target
Deviation
66.4f
67.0(

­
0.82%

ixhaust
Temperature,
"
F
Actual
Target
Detiation
230
215
2.27%

rpm
­
rewlutions
per
minute
ft­
lb
­
foot­
pounds
"
BTDC
­
degrees
Before
Top
Dead
Center
lb
H20
/
lb
air
­
pounds
water
Mpor
per
pound
of
air
in.
Hg
­
inches
of
mercury
psig
­
pounds
per
square
inch,
gauge
"
F
­
degrees
Fahrenheit
Final
Report
­
Caterpillar
3508
EUI
3­
7
September
2001
4.0
SAMPLING
LOCATIONS
Figures
4.1
and
4.2
present
schematic
drawings
of
the
pre­
catalyst
exhaust
gas
piping
and
CARB
429
traverse
point
locations
for
the
Caterpillar
3508
EUI
engine.
Figures
4.3
and
4.4
present
a
schematic
drawing
of
the
12
inch
(
ID)
post­
catalyst
exhaust
gas
piping
and
traverse
point
locations
for
the
3508
EUI
engine.
Sample
locations
for
the
testing
of
this
engine
are
also
shown.

The
exhaust
piping
upstream
of
the
catalyst
consisted
of
an
8­
inch
internal
diameter
(
ID)
pipe
that
connected
the
engine
exhaust
to
the
catalyst.
The
sampling
location
before
the
catalyst
consisted
of
several
sets
of
sampling
ports
used
for
isokinetic
sampling
and
extraction
of
sample
gas
for
the
FTIRS,
CEMS
and
GCMS
systems.
CAR$
429
sampling
before
the
catalyst
was
conducted
using
one
3­
inch
ID
sample
port.
The
sample
port
was
located
36
inches
(
4.5
diameters)
downstream
of
the
turbo
charger
exhausts.
The
port
was
located
26
inches
(
3.25
diameters)
upstream
of
the
the
nearest
disturbance,
which
was
a
90"
bend.
The
sample
port
was
fitted
with
a
ball
valve
and
high
pressure
couplings
to
enable
sample
traveres.
Sampling
was
conducted
through
one
port
using
a
three­
point
sample
matrix,
as
shown
in
Figure
4.2.
Lack
of
safe
access
to
the
second
port
precluded
a
traverse
at
this
location.

Multiple
ports
for
sample
gas
extraction
were
installed
on
an
exhaust
header
after
the
catalyst.
The
common
header
was
12
inches
in
diameter
and
directed
exhaust
gases
from
the
engines
tested
to
the
atmosphere.
CARB
429
sampling
and
velocity
traverses
were
conducted
through
two
3­
inch
ID
ports.
The
ports
were
on
perpendicular
diameters
and
located
8
1
inches
(
6.75
diameters)
downstream
of
the
nearest
flow
disturbance,
which
was
a
"
tee"
union.
Exhaust
gases
from
the
outlet
of
the
catalyst
made
a
90"
turn
into
the
exhaust
header.
The
third
leg
of
the
tee
was
disconnected
and
capped
off.
The
ports
were
located
225
inches
(
18.75
diameters)
upstream
of
the
nearest
flow
disturbance,
which
was
a
90"
bend
in
the
header.
Velocity
and
sample
traverses
were
conducted
using
a
12­
point
sample
matrix,
as
shown
in
Figure
4.4.

Final
Report
­
Caterpillar
3508
EUI
4­
l
September
200
1
Oxidation
Catalyst
FTIRS
Port
GC/
MS
Port
Expansion
Joint
(
typ.)

I
I
t
Hydraulic
Back
Pressure
Valve
A­
Frame
___)

Note:
Drawing
is
not
to
scale
Sampling
Platform
PM
Port
r1
v
0
b
Caterpillar
3508
EUI
Diesel
Engine
8"

f
26
'

c
PAH
Sampling
Location
Dynomometer
Figure
4.1
Inlet
Sample
Port
Locations
for
Velocity,
CARB
429,
FTIRS,
CEMS,
AND
GCMS
Sampling
1
Final
Report
­
Caterpillar
3508
EUI
4­
2
September
200
1
8
"

Traverse
Point
Number
:
3
Distance
from
Inside
wall
(
inches
)

15116
4
7
l/
16
Figure
4.2
Inlet
Traverse
Point
Locations
for
Velocity
and
CARB
429
Sampling
Final
Report
­
Caterpillar
3508
EUI
4­
3
September
200
1
p­­­­­
8+­­­­
4
1
225"

l­
l
t
FTIRS
Hydraulic
Back
Pressure
+
Valve
Oxidation
Catalyst
+
Ll
Ll
t
t
t
Port
PAH
Sampling
PM
Port
Ports
Note:
Top
View
GUMS
Port
Ll
Flow
from
Engine
Figure
4.3
Outlet
Sample
Port
Locations
for
Velocity,
CARB
429,
FTIRS,
CEMS
and
GCMS
Sampling
Final
Report
­
Caterpillar
3508
EUI
4­
4
September
200
1
12
"
6
5
`:

Traverse
Point
Distance
from
Number
Inside
wall
(
inches
)

1
II2
2
1
314
3
3
I/
2
4
8
7116
5
10
l/
4
6
11
I/
2
Figure
4.4
Outlet
Traverse
Point
Locations
for
Velocity
and
CARB
429
Sampling
Final
Report
­
Caterpillar
3508
EUI
4­
5
September
200
1
5.0
SAMPLING
AND
ANALYSIS
METHODS
This
section
discusses
the
various
sampling
and
analysis
methods
employed
by
PES,
EMI,
EECL,
ERG
and
Sierra
Instruments
to
quantify
the
HAP
emissions
before
and
after
the
oxidation
catalyst.
PES
selected
the
sampling
and
analysis
procedures
that
would
provide
the
information
required
during
the
planning
stages
of
the
project.
The
methods
were
selected
to
provide
the
required
data
in
the
most
economical
fashion,
while
providing
the
quality
required
by
the
Emissions
Standards
Division
(
ESD).

PES
divided
these
methods
into
two
categories
based
upon
quality
control
procedures
employed.
Type
I
methods
were
typical
source
test
methods,
designed
by
EPA
to
be
portable,
field
test
procedures.
PES
and
the
subcontractors
followed
QA
and
calibration
procedures
described
in
40
CFR
60,
Appendix
A
(
or
other
references
as
appropriate)
for
these
methods.

Type
II
methods
were
those
that
used
permanently
installed
instruments
housed
in
a
temperature­
controlled
environment
and
operated
in
the
same
fashion
as
continuous
monitors
used
by
industry
to
show
compliance
with
emission
regulations.
Because
these
instruments
are
maintained
in
a
laboratory­
type
environment
(
the
control
room
at
EECL),
fewer
QA
activities
and
calibrations
adequately
show
their
continuing
accuracy.
The
only
significant
change
to
the
quality
assurance
activities
was
that
fewer
instrument
calibrations
were
done
to
quantify
instrument
drift.
Historical
calibration
data
for
the
instruments
shows
their
stable
operation
over
extended,
e.
g.,
24­
hour,
periods.
Multipoint
calibrations
were
conducted
(
including
the
sampling
system
bias
checks)
on
these
instruments
once
at
the
beginning
of
each
engine
test.

Table
5.1
summarizes
the
parameters
measured,
the
sampling
methods,
the
classification,
and
measurement
principle.
The
text
that
follows
presents
brief
descriptions
of
the
sampling
and
analysis
procedures
used.

5.1
LOCATION
OF
MEASUREMENT
SITES
AND
SAMPLE/
VELOCITY
TRAVERSE
POINTS
PES
used
EPA
Method
1,
"
Sample
and
Velocity
Traverses
for
Stationary
Sources,"
to
select
the
measurement
sites
for
velocity
traverses
and
CARB
429
sampling
up
and
Final
Report
­
Caterpillar
3508
EUI
5­
l
September
200
1
TABLE
5.1
SUMMARY
OF
SAMPLING
AND
ANALYSIS
METHODS
Parameter
sample
Point
Location
Velocity
and
Volumetric
Flow
Oxygen
and
Carbon
Dioxide
Moisture
Nitrogen
Oxides
Carbon
Monoxide
Formaldehyde,
Acetaldehyde,
Acrolein
1,3­
Butadiene,
Hexane,
Benzene,
Toluene,
Ethyl
benzene,
Xylenes,
Styrene
Methane
Non&
Methane
Hydrocarbons
Total
Hydrocarbons
Polycyclic
Aromatic
Hydrocarbons
Measurement
Test
Method
QA
Category
Principle
EPA
Method
1
Type
I
Linear
Measurement
EPA
Method
2
Type
I
Differential
Pressure
Paramagnetic
and
EPA
Method
3A
Type
II
Non­
dispersive
Infrared
Analyzers
EPA
Method
4
Type
I
Gravimetric
GRI
Protocol'
Type
I
FTIRS
Analyzer
Carbon
Balance*
Type
I
Stoichiometry
Chemiluminescent
EPA
Method
7E
Type
II
Analyzer
EPA
Method
10
Type
II
GFC/
NDIR
Analyzer
GRI
Protocol
Type
II
FTIRS
Analyzer
Gas
Chromatograph
Alternate
Method
17
Type
I
w/
Mass
Spectrometer
Detector
EPA
Method
25A
(
modified)
Type
II
GC­
FID
Analyzer
EPA
Method
25A
(
modified)
Type
II
GC­
FID
Analyzer
EPA
Method
25A
Type
II
FID
Analyzer
CARB
429
Type
I
Low
Resolution
GCMS
Particulate
Matter
IS0
8178­
1
Type
I
Micro
Dilution
Gravimetric
'
Measurement
of
Select
Hazardous
Air
Pollutants,
Criteria
Pollutants,
and
Moisture
Using
Fourier
Transform
Infrared
(
FTIR)
Spectroscopy.
Presented
as
an
Appendix
to
Fourier
Transform
Infrared
Spectroscopy
(
FTIRS)
Method
Validation
at
a
Natural
Gas­
Fired
Internal
Combustion
Engine
(
G&
95/
027
l),
Gas
Research
Institute,
December
1995.

2
Derivation
of
General
Equation
for
Obtaining
Engine
Exhaust
Emissions
on
a
Mass
Basis
Using
the
"
Total
Carbon"
Method.

Final
Report
­
Caterpillar
3508
EUI
5­
2
September
200
1
TABLE
5.1
(
Concluded)

SUMMARY
OF
SAMPLING
AND
ANALYSIS
METHODS
Parameter
Test
Method
QA
Category
Measurement
Principle
Fuel
Oil
Composition
(
Ultimate
Analysis)
ASTM
D
5291
(
C,
H,
N)
ASTM
4239
(
S)
ASTM
D
1744
(
Moisture)
ASTM
D482
Ash
Type
I
(
See
Text)

Fuel
Oil
Metals
Analysis
SW­
846
3051
(
Prep)
SW­
846
6010B
SW­
846
7000
SW­
846
7470A
SW­
846­
7471A
Type
I
(
See
Text)

downstream
of
the
catalyst.
PES
used
the
cyclonic
flow
check
procedure
outlined
in
Method
1
to
evaluate
the
suitability
of
the
inlet
location
for
isokinetic
sampling.
The
measurement
sites
are
discussed
in
Section
4.0.

5.2
DETERMINATION
OF
STACK
GAS
VOLUMETRIC
FLOW
RATE
,

During
the
PAH
runs,
Method
2
was
used
in
direct
support
of
the
CAIXB
429
sampling.
The
mass
flow
rates
of
the
PAH
compounds
and
the
run­
by­
run
detection
limits
are
calculated
using
the
results
of
these
velocity
traverses.
PES
used
EPA
Method
2,
`
Determination
of
Stack
Gas
Velocity
and
Volumetric
Flow
Rate
(
Type
S
Pitot
Tube),"
to
determine
stack
gas
velocity
during
CARB
429
sampling.
The
test
crew
used
a
Type
S
pitot
tube,
constructed
according
to
specifications
of
Section
2.1
of
Method
2
and
having
a
coefficient
(
C,)
of
0.84.
The
pitot
tube
was
connected
to
an
inclined/
vertical
manometer
and
the
Ap
measured
at
each
traverse
point.
Stack
gas
temperature
was
measured
using
a
Type­
K
thermocouple.
The
average
stack
gas
velocity
was
calculated
from
the
average
of
the
square
roots
of
the
Ap
values,
the
average
stack
gas
temperature,
the
stack
gas
molecular
weight,
and
the
absolute
stack
pressure.
The
volumetric
flow
rate
is
the
product
of
velocity
and
the
stack
cross­
sectional
area
of
the
duct
at
the
sampling
location.
PES
conducted
a
velocity
traverse
using
the
standard
pitot
tube
before
each
run
and
adjusted
the
sampling
rate
of
the
CARB
429
train
based
upon
these
data.
PES
employed
this
approach
with
the
approval
of
the
WAM.
Access
to
the
sampling
locations
was
severely
restricted
due
to
the
short
runs
of
exhaust
piping
and
the
profusion
of
sampling
probes
required
during
each
sampling
run.

Final
Report
­
Caterpillar
3508
EUI
5­
3
September
200
1
5.3
DETERMINATION
OF
STACK
GAS
OXYGEN
AND
CARBON
DIOXIDE
CONTENT
EECL
used
EPA
Method
3A,
"
Determination
of
Oxygen
and
Carbon
Dioxide
Concentrations
in
Emissions
from
Stationary
Sources
(
Instrumental
Analyzer
Procedure),"
to
measure
oxygen
and
carbon
dioxide
content
of
the
exhaust
gas
during
testing.
EECL's
sample
gas
extraction
and
transport
system
extracted
a
gas
sample
from
the
exhaust
gas
stream.
The
sample
was
conditioned
to
remove
moisture
and
entrained
particulate
matter
and
directed
the
Rosemount
NGA­
2000
gas
analysis
system.
Oxygen
was
measured
using
the
paramagnetic
detection
principle.
Carbon
dioxide
was
measured
using
a
non­
dispersive
infrared
(
NDIR)
analyzer.
The
oxygen
and
carbon
dioxide
monitors
were
calibrated
with
a
pre­
purified
zero
gas
and
three
upscale
gas
standards
corresponding
to
approximately
30,55,
and
85
percent
of
the
instruments'
analytical
ranges.
EECL
used
only
EPA
Protocol
gas
standards
certified
by
the
gas
manufacturer.
The
calibration
gases
that
were
used
and
the
calibration
responses
of
the
instruments
are
discussed
in
Section
6.0
of
this
document.
A
schematic
diagram
of
the
FTIRWEMS
sampling
and
analysis
system
is
presented
in
Figure
5.1.

5.4
DETERMINATION
OF
STACK
GAS
MOISTURE
CONTENT
PES
and
EECL
used
three
methods
to
determine
the
moisture
concentration
in
the
exhaust
gas
before
and
after
the
catalyst.
Method
4
was
used
in
direct
support
of
the
CARB
429
sampling
during
the
PAH
runs.
During
the
CEMS/
GCMS/
FTIRS
runs,
moisture
was
measured
using
the
FTIRS
upstream
of
the
catalyst,
and
by
a
carbon
balance
calculation
downstream
of
the
catalyst.
During
the
testing,
EECL
personnel
determined
that
the
moisture
concentrations
after
the
catalyst,
as
measured
by
the
Nicolet
Magna
560
FTIRS
analyzer,
were
about
6
percent
higher
that
actual.
EECL
calculated
the
moisture
concentration
after
the
catalyst
using
a
carbon
balance
method.

PES
used
EPA
Method
4,
"
Determination
of
Moisture
Content
in
Stack
Gases,"
to
measure
the
flue
gas
moisture
content
during
the
CARB
429
sampling.
The
gas
sample
was
extracted
from
the
exhaust
pipe
and
pulled
through
an
impinger
train
chilled
by
an
ice
bath.
The
field
technicians
weighed
the
impinger
train
(
including
the
XAD@­
2
sorbent
trap)
before
and
after
sampling.
PES
then
calculated
the
quantity
of
water
collected
in
the
train
and
the
moisture
content
of
the
stack
gas.

EECL
used
methodology
described
in
the
document
"
Measurement
of
Select
Hazardous
Air
Pollutants,
Criteria
Pollutants,
and
Moisture
Using
Fourier
Transform
Infiared
(
FTIRS)
Spectroscopy"
to
measure
moisture
concentrations
upstream
of
the
catalyst.
This
document
is
called
the
GRI
Protocol
in
this
report,
and
is
presented
as
Appendix
B
of
a
report
published
by
the
Gas
Research
Institute:
"
Fourier
Transform
Inpared
Spectroscopy
(
FTIRS)
Method
Validation
at
a
Natural
Gas­
Fired
Internal
Combustion
Engine."
A
sample
Final
Report
­
Caterpillar
3508
EUI
5­
4
September
200
1
.
CHjNMHC
Analyzer
­
­
­
I
I
Heated
Sample
Line
I
I
t
I
I
I
.
_
_.
.
.
­.
.
­
­.
.
..­
­
­
.
.
ti
Miratech
Oxidation
I
I
I
I
.
­

88
(
Acq&
if$
m
1
I
I
Calibration
Gas
Cylinders
:
Catalyst
I
c
.
.._­..._­...­­...­­...­­..
i
Exhaust
Flow
Heated
Sample
Line
­
1
I­
­
THC
Analyzer
I
I
­{
CHjNMHC
Analyzer
I­
­
i
Figure
5.1
Schematic
Diagram
of
EECL
FTIRWCEMS
Sampling
and
Analysis
System
Final
Report
­
Caterpillar
3508
EUI
5­
5
September
200
1
of
the
gas
was
extracted
from
the
exhaust
and
directed
to
a
Nicolet
Rega
7000
FTIRS
analyzer
to
measure
the
moisture
concentration
in
the
exhaust
gas.
The
gas
sample
was
filtered
to
remove
entrained
particulate
matter
and
transported
to
the
analyzer
via
a
heated
Teflon@
sampling
line.
Further
discussion
of
the
FTIRS
sampling
and
analysis
method
may
be
found
in
the
report
generated
by
the
EECL
and
the
GRI
protocol.

Because
the
FTIRS
analyzer
downstream
of
the
catalyst
did
not
measure
the
moisture
concentration
accurately,
EECL
used
a
carbon
balance
method
to
calculate
the
moisture
present
in
the
gas
stream
downstream
of
the
catalyst.
The
method
used
is
discussed
in
the
EECL
report
in
Appendix
A.

5.5
DETERMINATION
OF
NITROGEN
OXIDES
EPA
Method
7E,
"
Determination
of
Nitrogen
Oxide
Emissions
from
Stationary
Sources
(
Instrumental
Analyzer
Procedure),"
determined
nitrogen
oxide
content
of
the
exhaust
gases.
These
tests
also
provided
the
data
needed
to
do
the
EPA
Method
301
validation
of
the
FTIRS
for
NO,
emissions
from
this
source.
A
gas
sample
was
extracted
from
the
exhaust
gas
stream,
conditioned
to
remove
moisture,
and
the
nitrogen
oxide
concentration
determined
by
an
instrumental
analyzer.
The
measurement
principle
for
oxides
of
nitrogen
is
chemiluminescence.
The
NO,
monitor
was
calibrated
with
a
pre­
purified
zero
gas,
and
three
upscale
gas
standards
corresponding
to
approximately
30,55,
and
85
percent
of
the
instruments
analytical
ranges.
EECL
used
EPA
Protocol
gas
standards
certified
by
the
gas
manufacturer.
The
calibration
gases
that
were
used
and
the
calibration
responses
of
the
instruments
are
discussed
in
Section
6.0
of
this
document.
A
schematic
diagram
of
the
FTIRSKEMS
sampling
and
analysis
system
is
presented
in
Figure
5.1.

5.6
DETERMINATION
OF
CARBON
MONOXIDE
EPA
Method
10,
"
Determination
of
Carbon
Monoxide
Emissionsfiom
Stationary
Sources,"
measured
the
CO
concentration
of
the
exhaust
gas
during
the
testing.
These
tests
also
provided
the
data
needed
to
do
the
EPA
Method
30
1
validation
of
the
FTIRS
sampling
and
analysis
system
for
CO
emissions
from
this
source.
A
gas
sample
was
extracted
from
the
exhaust
gas
stream,
conditioned
to
remove
moisture,
and
the
carbon
monoxide
concentration
determined
by
an
instrumental
analyzer.
The
measurement
principle
for
carbon
monoxide
is
GFUNDIR.
The
CO
monitor
was
calibrated
using
a
pre­
purified
zero
gas
and
three
upscale
gas
standards
corresponding
to
approximately
30,55
and
85
percent
of
the
instrument's
analytical
range.
All
gas
standards
used
for
calibrations
were
prepared
according
to
EPA
Protocol
and
certified
by
the
gas
manufacturer.
The
calibration
gases
that
were
used
and
the
calibration
responses
of
the
instruments
are
discussed
in
Section
6.0
of
this
document.
A
schematic
diagram
of
the
FTIRSKEMS
sampling
and
analysis
system
is
presented
in
Figure
5.1.

Final
Report
­
Caterpillar
3508
EUI
5­
6
September
200
1
5.7
DETERMINATION
OF
TOTAL
HYDROCARBONS
EPA
Method
25A,
Determination
of
Total
Gaseous
Organic
Concentration
Using
a
Flame
Ionization
Analyzer,
determined
the
total
hydrocarbon
concentrations
at
the
inlet
and
the
outlet
of
the
catalyst.
At
the
catalyst
inlet,
EECL
used
a
Therm0
Environmental
Instruments
(
TECO)
Model
5
1
Total
Hydrocarbon
Analyzer.
The
analyzer
consisted
of
a
heated
compartment
to
prevent
condensation
of
organic
compounds,
and
a
Flame
Ionization
Detector
(
FID)
to
measure
THC
concentrations.
At
the
catalyst
outlet,
EECL
used
a
Rosemount
Analytical
NGA­
2000
FID
Hydrocarbon
Analyzer.
This
analyzer
also
used
an
FID
to
measure
the
concentrations
of
THC
in
the
gas
stream.
The
FID
detector
consists
of
a
burner
in
which
a
regulated
flow
of
a
sample
gas
passes
through
a
flame
sustained
by
regulated
flows
of
a
fuel
gas
and
air.
The
hydrocarbon
components
of
the
sample
stream
ionize
in
the
flame.
The
positive
ions
that
are
produced
are
collected
by
an
electrode
causing
current
to
flow
through
a
measuring
circuit.
The
ionization
current
is
proportional
to
the
rate
at
which
carbon
atoms
enter
the
burner,
and
is
therefore
a
measure
of
the
concentration
of
hydrocarbons
in
the
sample.

5.8
DETERMINATION
OF
METHANE
AND
NON­
METHANE
HYDROCARBONS
A
modification
of
EPA
Method
25A,
"
Determination
of
Total
Gaseous
Organic
Concentration
Using
a
Flame
Ionization
Analyzer,"
determined
the
methane
and
non­
methane
concentrations
at
the
inlet
and
the
outlet
of
the
catalyst.
Gas
samples
extracted
from
each
gas
stream
were
transported
to
MSA
1030H
Methane/
Non­
Methane
Analyzers.
These
analyzers
are
single­
purpose
gas
chromatographs
that
separate
methane
from
the
other
organic
compounds
in
the
sample
by
passing
the
sample
through
a
separation
column.
The
methane
elutes
from
the
column
first
and
is
directed
to
the
flame
ionization
detector.
Then,
the
analyzer
reverses
the
flow
through
the
column
and
the
remaining
organic
compounds
are
back
flushed
to
the
same
detector.
The
analyzer
sums
the
two
fractions
to
yield
the
concentration
of
total
organic
compounds.
Because
this
unit
is
a
gas
chromatograph,
it
cannot
measure
methane
and
non­
methane
concentrations
continuously.
During
testing,
each
analyzer
determined
concentrations
once
every
five
minutes.
This
frequency
is
sufficient
for
testing
on
RICE
because
the
operating
conditions
were
maintained
within
close
constraints.
Each
analyzer
was
calibrated
before
each
week
of
testing
using
methane
and
propane
calibration
standards
corresponding
to
approximately
30,50,
and
85
percent
of
the
instrument
span.
The
calibration
gases
that
were
used
and
the
calibration
responses
of
the
instruments
are
discussed
in
Section
6.0
of
this
document.
A
schematic
diagram
of
the
FTIRWEMS
sampling
and
analysis
system
is
presented
in
Figure
5.1.

Final
Report
­
Caterpillar
3508
EUI
5­
7
September
200
1
5.9
DETERMINATION
OF
GASEOUS
ORGANIC
HAP
USING
FTIRS
EECL
used
two
FTIRS
systems
that
met
the
sampling
and
analysis
requirements
set
forth
in
the
GRI
Protocol.
GRI
validated
extractive
FTIRS
systems
successfully
for
an
analysis
of
emissions
from
natural
gas­
fired
RICE.
The
extractive
FTIRS
continuously
extracts
a
sample
gas
from
the
stack,
transports
the
sample
to
the
FTIRS
system,
and
does
spectral
analysis
of
the
sample
gas.
The
computer
system
analyzes
sample
gas
spectra
for
target
analytes
continuously
and
archives
them
for
possible
later
reanalysis.
Table
5.2
presents
specifications
of
the
FTIRS
analyzers.

TABLE
5.2
FTIRS
ANALYZER
SPECIFICATIONS
Parameter
Pre­
catalyst
Manufacturer
and
Type
Nicolet
Rega
7000
Post­
catalyst
Nicolet
Magna
560
Spectral
Resolution
Detector
Type
Cell
Type
Cell
Temperature
Cell
Pressure
0.5
cm
­
I
0.5
cm"

MCT­
A
MCT­
A
4.2
Meter
­
Fixed
Path
Length
2.0
Meter
­
Fixed
Path
Length
185
"
C
165
"
C
600
Torr
600
Torr
Cell
Window
Material
Zinc
Cellinide
KE3r
The
sampling
and
measurement
system
consists
of
the
following
components:

l
heated
probe;

l
heated
filter;

l
heat­
traced
Teflon@
sample
line;

l
Teflon@
coated,
heated­
head
sample
pump;

l
FTIRS
spectrometer;
and
l
QA/
QC
apparatus.

EECL
validated
each
sample
extraction
and
analysis
system
for
formaldehyde,
acetaldehyde,
and
acrolein
before
testing.
The
results
of
the
FTIRS
validation
are
discussed
in
Section
6.0.
The
basic
sampling
procedure
consisted
of
EECL
taking
an
initial
Final
Report
­
Caterpillar
3508
EUI
5­
8
September
200
1
interferogram
of
the
stack
gas
with
the
FTIRS
measurement
and
analysis
system
before
each
test
to
describe
the
sample
matrix.
This
measured
the
concentrations
of
moisture
and
the
target
pollutants
and
allowed
for
adjustments
to
the
cell
pathlength
and
the
spectral
analysis
regions
if
the
concentrations
differed
from
expectations.
Sample
conditioning
was
not
necessary
at
the
EECL
test
site.

After
QA/
QC
procedures
and
initial
adjustments
were
completed
for
a
given
test
day,
a
gas
sample
was
drawn
continuously
through
the
heated
FTIRS
cell
while
the
system
collected
spectral
data.
The
FTIRS
systems
collected
data
simultaneously
with
the
other
continuous
monitors
and
with
the
manual
train
sampling
for
PAHs
during
CARB
429
runs.
The
spectrometer
collected
one
complete
spectrum
of
the
sample,
as
an
interferogram,
per
second
and
averaged
interferograms
over
l­
minute
periods.
The
FTIRS
computer
converted
these
time­
integrated
interferogram
into
conventional
wave
number
spectra,
analyzed
for
the
target
compounds,
and
archived
the
data.
Sample
collection
was
33
minutes
in
duration,
coinciding
with
the
test
runs.

5.10
DETERMINATION
OF
ORGANIC
HAP
BY
DIRECT
INTERFACE
GCMS
The
sampling
and
analytical
procedures
used
during
this
testing
program
followed
those
detailed
in
EPA
Alternate
Method
17
"
Determination
of
Gaseous
Organic
Compounds
by
Direct
Interface
GUMS".
The
instrument
was
calibrated
on­
site
using
a
lo­
ppm
manufacturer's
certified
compressed
gas
mixture
consisting
of
nine
target
analytes
(
benzene,
toluene,
o,
m,
p­
xylenes,
styrene,
ethyl
benzene,
1,3­
butadiene,
and
hexane)
in
nitrogen
balance.
Calibrations
were
conducted
at
10
ppm,
3
ppm,
1
ppm
and
100
ppb
to
provide
enough
calibration
points
for
all
of
the
target
analytes.
The
10
ppm
standard
was
diluted
with
VOC
free
nitrogen
using
an
EM1
calibration
gas
manifold
with
three
mass
flow
meters.
The
mass
flow
meters
were
calibrated
in
the
field
on
the
day
of
use
by
comparison
to
a
digital
bubble
meter
with
a
NIST
traceable
calibration.
Also,
an
independent
1
ppm
standard
was
used
to
verify
the
calibration
validity
and
calibration
gas
dilution
technique
as
required
by
the
method.

Effluent
gas
samples
were
withdrawn
at
a
constant
flow
rate
(
1.5­
liters/
minute,
dry
basis)
from
a
single
point
located
approximately
at
the
midpoint
of
the
exhaust
pipe.
The
estimated
gas
residence
time
through
the
sampling
system
at
this
flow
rate
is
less
than
1
minute.
The
response
time
required
to
equilibrate
a
step
change
in
sample
concentration
was
approximately
6
minutes.
During
this
test
program,
the
catalyst
inlet
and
outlet
GCMS
measurement
systems
collected
effluent
simultaneously
from
each
location
for
a
period
not
less
than
10
minutes
before
acquisition
by
the
GCMS
instrumentation.
A
test
run
consisted
of
collection
of
four
lo­
minute
samples
from
each
location.

Figure
5.2
presents
a
schematic
of
the
GCMS
measurement
system(
s)
used
during
the
test
program.
Each
sampling
system
consisted
of
a
heated
probe,
heated
0.3
micron
quartz
Final
Report
­
Caterpillar
3508
EUI
5­
9
September
200
1
fiber
filters,
heated­
head
sample
pump,
and
a
heated
Teflon@
sampling
line
to
transport
the
gas
to
the
control
units.
The
control
units
included
flow
measurement
and
flow
control
capabilities
and
Peltier
cooled
condensers
with
continuous
condensate
removal
to
dry
the
sample
gas.
All
of
the
sample
gases
were
directed
through
the
condensers
(
i.
e.,
the
condenser
bypass
was
not
used)
and
the
condensers
were
operated
at
36­
39
"
F.
This
resulted
in
a
sample
moisture
content
of
less
than
2%.
The
control
units
also
included
provisions
to
ensure
that
the
sample
gas
at
the
GCMS
inlet
probe
was
at
atmospheric
pressure
so
that
the
quantity
of
internal
standards
co­
injected
with
the
sample
was
unaffected.

Several
equipment
modifications
were
made
for
the
test
on
the
Caterpillar
engine.
High
soot
loading
was
expected
at
the
inlet
sampling
location.
Therefore,
two
0.3­
micron
quartz
fiber
filters
were
installed
in
series
on
the
sampling
system
upstream
of
the
catalyst
to
prevent
breakthrough
of
the
particulate
matter
and
subsequent
coating
of
the
internal
surfaces
of
the
measurement
system.
A
single
0.3­
micron
quartz
fiber
filter
was
used
in
the
downstream
of
the
catalyst.
A
short
length
of
unheated
stainless
steel
tubing
was
installed
at
the
inlet
sampling
location
to
cool
the
exhaust
stream
from
1100
"
F
to
about
300­
400
"
F.
A
three­
way
valve
with
an
atmospheric
vent
was
installed
at
the
upstream
location
sampling
probe
because
of
the
elevated
static
pressure
at
this
sampling
location
and
concerns
that
this
could
adversely
affect
system
calibration
checks.
The
three­
way
valve
was
very
useful
because
it
also
allowed
sampling
of
ambient
air
between
test
runs
during
the
extended
periods
required
to
achieve
stable
engine
operation
and
during
other
test
delays.
Finally,
a
cooling
system
comprised
of
a
fan
and
metal
foil
flexible
duct
was
used
to
convey
relatively
cool
ambient
air
from
the
floor
of
the
engine
room
to
the
electronic
control
sections
of
the
inlet
and
outlet
sampling
system
probe
boxes.
This
prevented
overheating
of
electronics
and
pump
motors
caused
by
the
elevated
ambient
temperatures
and
infrared
radiation
at
the
sampling
locations
during
the
tests.

Final
Report
­
Caterpillar
3508
EUI
5­
10
September
200
1
Heated
Probe
250'

.5
lpm
constant
rate
sampling)

___
_­__
_­­­.­
­

Excess
Samnle
Atmosnheric
Vent
____
­­­
­­
By
Mass
Flow
Meter
Mass
Flow
Meter
I
Condenser
_
Flow
Control
l­
71
Flow
r­
l
Flow
Condenser
System
ondensate
Drain
\
Contr
1
Box
Heated
to
125
°
F
(
or
at
least
5
°
F
above
saturation
temperature
of
sample
gas
)
Connection
Line
50
cc/
min
during
GC­
MS
sample
acquisition)

GC­
MS
Analyzer
Figure
5.2
Schematic
of
GCMS
Sampling
and
Analysis
System
Final
Report
Caterpillar
3508
EUI
5­
l
1
September
2001
.
5.11
DETERMINATION
OF
POLYCYCLIC
AROMATIC
HYDROCARBONS
BY
CARB
429
PES
used
CARB
Method
429,
"
Determination
of
Polycyclic
Aromatic
Hydrocarbon
(
PAH)
Emissions
porn
Stationary
Sources,"
to
quantify
PAH
concentrations
and
emission
rates
before
and
after
the
catalyst.
Sample
run
times
were
120
minutes
in
duration.
Runs
PAH
1,
PAH
2,
and
PAH
3
were
conducted
with
the
engine
and
test
conditions
10,2,
and
3,
respectively.
Figure
5.3
presents
a
simplified
schematic
diagram
of
the
CARB
429
sample
train.
The
CARB
429
sampling
train
was
modified
to
enable
sampling
under
difficult
test
conditions.
At
the
inlet
location
a
ball
valve
was
placed
inclined
at
the
end
of
the
probe.
The
valve
was
used
to
restrict
the
static
pressure
and
to
ensure
that
the
sampling
train
was
not
over
pressurized.
The
outlet
of
the
ball
valve
was
connected
to
the
inlet
of
the
condenser
with
a
heated
Teflon@
sample
line.
A
heated
sample
line
between
the
sample
probe
and
the
condenser
was
also
used
downstream
of
the
catalyst.
The
sample
line
was
used
to
facilitate
sample
traverses
in
the
horizontal
duct.

PES
field
technicians
recovered
the
CARB
Method
429
sample
train
as
described
by
CARB
Method
429.
Method
429
specifies
that
sample
recovery
rinses
be
done
with
acetone,
hexane,
and
methylene
chloride.
PES
collected
blank
samples
of
reagent
grade
water,
acetone,
hexane,
methylene
chloride,
unused
filters,
and
XAD@­
2
resin
cartridges
used
during
the
test
program.
The
sample
recovery
apparatus
consisted
of
pre­
cleaned
Teflon@'
or
glass.
Field
technicians
did
three
acetone
rinses,
three
hexane
rinses,
and
three
methylene
chloride
rinses
of
each
sample
train
component
from
the
nozzle
to
the
front
half
of
the
filter.
They
also
rinsed
the
back
half
of
the
filter
holder,
the
connector,
and
the
condenser
three
times
with
acetone.
They
soaked
the
back
half
of
the
filter
holder,
connector,
and
condenser
three
times
with
acetone,
hexane,
and
methylene
chloride,
for
five
minutes
each
time.
PES
provided
pre­
cleaned
amber
glass
sample
bottles
with
Teflon@
seals
for
the
recovery
of
solvent
rinses.

After
sampling
and
recovery,
the
CARB
429
sample
fractions
were
stored
on
ice
and
transported
by
PES
personnel
from
Fort
Collins,
Colorado,
to
PES'
laboratory
facilities
in
Research
Triangle
Park,
North
Carolina.
The
sample
bottles
were
examined
for
breakage
and
sample
loss.
The
samples
were
then
transferred
by
PES
personnel
to
ERG
laboratory
facilities
in
Morrisville,
North
Carolina,
for
sample
extraction
and
analysis.
ERG
extracted
the
sample
fractions
for
each
PAH
sampling
run
with
methylene
chloride,
then
combined
the
extracts.
The
6
extracts
(
3
inlet
samples
and
3
outlet
samples)
were
concentrated
to
a
volume
of
about
15
ml
using
a
Kuderna­
Danish
flask,
then
evaporated
to
dryness
using
a
nitrogen
blowdown
apparatus.
The
extracts
were
each
reconstituted
with
1
ml
hexane
before
analysis
using
a
gas
chromatograph
with
a
low
resolution
mass
spectrometer.

Final
Report
Caterpillar
3508
EUI
5­
12
September
200
1
/
bygone
(
Optional)

Heated
Probe,
S­
type
Pitot
&
Temp.
Sensor
t
Temp.
Readout
Pitot
Manometer
Filter
Assembly
­
Transfer
Line
Thermocou
Sorbent
Module
(
water
cooled)

Impingers
in
Ice
Bath:

Thermocouple
Main
Valve
Orifice
Dr
Gas
9
eter
Pump
Figure
5.3.
Schematic
Diagram
of
CARB
429
PAH
Sampling
Train
Final
Report
Caterpillar
3508
EUI
5­
13
September
200
1
5.12
DETERMINATION
OF
PARTICULATE
MATTER
Particulate
sampling
was
conducted
simultaneously
with
the
FTIRS/
CEMS/
GCMS
and
CARB
429
testing
both
upstream
and
downstream
of
the
catalyst.
PM
sampling
was
conducted
by
Sierra
Instruments
using
a
BG­
1
Micro­
dilution
test
stand.
Sierra
employed
procedures
outlined
in
International
Organization
for
Standardization
(
ISO)
Method
8
178­
1
"
Reciprocating
Internal
Combustion
Engine
­
Exhaust
Emission
Measurement
­
Part
1:
Test­
bed
Measurement
of
Gaseous
and
Particulate
Exhaust
Emissions
".

5.13
DETERMINATION
OF
FUEL
OIL
COMPOSITION
PES
personnel
collected
one
sample
of
the
fuel
oil
used
to
fire
the
engine
on
each
day
of
sampling.
These
samples
were
analyzed
by
Galbraith
Laboratories,
Inc.
The
sample
analysis
consisted
of
ultimate
and
proximate
analyses
and
an
analysis
to
measure
the
concentrations
of
the
target
metals.
Table
5.3
summarizes
the
analytical
methods
used
for
these
determinations.

Galbraith
Laboratories
used
methods
from
the
American
Society
for
Testing
and
Materials
(
ASTM)
to
determine
the
carbon,
hydrogen,
nitrogen,
sulfur,
moisture,
and
ash
content
of
the
fuel
oil.
Galbraith
Laboratories
also
determined
the
heat
content
of
the
fuel
oil.
From
this
information,
PES
calculated
a
fuel
factor,
F,,
and
the
heating
rate
of
the
engine
during
each
of
the
test
runs.

Galbraith
Laboratories
used
analysis
methods
published
by
EPA's
Office
of
Solid
Waste
(
OSW)
to
measure
the
content
of
the
target
metals.
An
aliquot
of
the
fuel
oil
was
acid
digested,
and
the
digestate
analyzed
to
determine
metals
content.

Final
Report
Caterpillar
3508
EUI
5­
14
September
200
1
TABLE
5.3
SUMMARY
OF
FUEL
OIL
ANALYSIS
METHODS
Parameter
Analysis
Method
Carbon,
Hydrogen,
Nitrogen
ASTM
D529
l­
96
Sulfur
Moisture
Ash
ASTM
D4239
ASTM
D1744
ASTM
D482
I
Sample
Preparation
(
Metals)
SW­
846
305
1
I
Cadmium,
Chromium,
Lead,
Manganese,
Nickel,
SW­
846
6010B
(
ICP­
AES)'

II
Mercury
I
SW­
846
7470A
(
CVAAS)
4
SW­
846
7471A
(
CVAAS)
(
I
3
Inductively
coupled
plasma­
atomic
emission
spectrometry.

4
Cold­
vapor
atomic
absorption
spectrophotometry.

Final
Report
Caterpillar
3508
EUI
5­
15
September
200
1
6.0
QUALITY
ASSURANCE/
QUALITY
CONTROL
PROCEDURES
AND
RESULTS
This
section
summarizes
the
specific
QA/
QC
procedures
that
PES,
EECL,
EMI,
and
ERG
personnel
employed
during
the
performance
of
this
source
testing
program.
PES'
quality
assurance
program
was
based
upon
the
procedures
and
guidelines
contained
in
the
"
Quality
Assurance
Handbook
for
Air
Pollution
Measurement
Systems,
Volume
III,
Stationary
Source
Specific
Methods,"
EPA/
600/
R­
94/
038c,
and
in
the
test
methods.
These
procedures
ensure
the
collection,
analysis,
and
reporting
of
reliable
source
test
data.

6.1
FTIRS
QA/
QC
PROCEDURES
EECL
calibrated
the
FTIRS
instruments
before
each
engine
test
series
and
at
the
beginning
and
end
of
each
test
day.
The
calibration
procedures
employed
were
consistent
with
procedures
found
in
the
following
documents:

Gas
Research
Institute
Report
Number
GRI­
95/
027
1
entitled,
"
Fourier
Transform
Infrared
(
FTIRS)
Method
Validation
at
a
Natural
Gas­
Fired
Internal
Combustion
Engine"

This
report
was
prepared
for
the
Gas
Research
Institute
by
Radian
Corporation.
Included
as
appendices
are
two
additional
documents,
which
also
have
relevance
in
the
test
program:

"
Measurement
of
Select
Hazardous
Air
Pollutants,
Criteria
Pollutants,
and
Moisture
Using
Fourier
Transform
Infrared
(
FTIRS)
Spectroscopy"
­
Prepared
by
Radian
Corporation
for
the
Gas
Research
Institute.

"
Protocol
for
Performing
Extractive
FTIRS
Measurements
to
Characterize
Various
Gas
Industry
Sources
for
Air
Toxics"
­
Prepared
by
Radian
Corporation
for
the
Gas
Research
Institute.

Final
Report
­
Caterpillar
3508
EUI
6­
l
September
200
1
6.1.1
FTIRS
Svstem
PreDaration
Both
FTIRS
sampling
systems
(
before
and
after
the
catalyst)
were
subjected
to
an
EPA
Method
301
validation
process
for
formaldehyde,
acetaldehyde,
and
acrolein.
The
validation
process
quantified
the
precision
and
accuracy
of
each
FTIRS
analyzer
for
these
compounds.
Besides
the
validation
program,
EECL
personnel
did
the
following
calibration
procedures
before
each
engine
test
series.

1.
Source
Evaluation
­
Initial
source
data
were
acquired
to
verify
concentration
'
ranges
of
target
compounds
and
possible
interferences.
This
was
completed
before
and
during
the
Method
301
validation
process
for
formaldehyde,
acetaldehyde,
and
acrolein,
and
during
the
test
program
for
moisture.

2.
Sample
System
Leak
Check
­
A
leak
check
was
done
on
the
portions
of
the
system
between
the
sample
filter
and
the
pump
outlet.
A
rotameter
was
connected
to
the
discharge
side
of
the
sample
pump.
The
indicated
sample
flow
rate
was
recorded
while
the
sample
system
was
operating
at
typical
temperatures
and
pressures
(
the
sample
pump
pulled
a
slight
vacuum
on
the
suction
side).
The
inlet
was
closed
off
just
downstream
of
the
sample
probe.
A
rotameter
monitored
the
flow
rate.
A
leak
rate
of
4%
or
less
of
the
standard
sampling
rate
of
500
ml/
min
indicated
an
acceptable
leak
check.

3.
Analyzer
Leak
Check
­
Both
FTIRS
analyzers
were
checked
to
ensure
that
they
were
operating
at
normal
operating
temperatures
and
pressures.
The
operating
pressures
were
recorded.
The
automatic
pressure
control
device
was
disabled
and
the
inlet
to
the
FTIRS
was
closed.
The
cell
was
evacuated
to
20%
or
less
of
the
normal
operating
pressure.
After
the
cell
was
evacuated,
it
was
isolated
and
the
cell
pressure
was
monitored
with
a
dedicated
pressure
sensor.
The
leak
rate
of
the
measurement
cell
must
be
less
than
10
Torr
per
minute
for
1
minute
for
the
analyzer
leak
to
be
considered
acceptable.

4.
Cell
Pathlength
Determination
­
The
FTIRS
cell
pathlengths
were
to
be
determined
using
the
procedure
outlined
in
the
Field
Procedure
Section
the
document
entitled
"
Protocol
for
Performing
Extractive
FTIRS
Measurements
to
Characterize
Various
Gas
Industry
Sources
for
Air
Toxics."
Because
each
FTIRS
was
a
fixed
pathlength
unit
(
i.
e.,
the
pathlengths
were
not
adjustable)
measurements
of
the
cell
pathlengths
were,
deemed
unnecessary.
The
cell
pathlengths
specified
by
the
manufacturer
were
used
in
the
measurement
algorithms.

Final
Report
­
Caterpillar
3508
EUI
6­
2
September
200
1
6.1.2
FTIRS
Daily
Calibrations
and
OA
Procedures
Before
each
day
of
testing,
EECL
personnel
calibrated
each
FTIRS
system
following
the
procedures
outlined
below.

1.
Instrument
Stabilization
­
Each
of
the
following
components
were
checked
for
proper
operation
to
ensure
the
stability
of
the
operation
of
the
FTIRS
instruments:

4
Instrument
heaters
and
temperature
controllers.

W
Pressure
sensors
and
pressure
controllers.

C>
Sample
system
(
pump,
filters,
flow
meters,
and
water
knockouts).

2.
The
FTIRS
analyzers
were
purged
with
conditioned
air
for
a
minimum
of
30
minutes
before
conducting
and
analysis
of
the
background
spectrum.
During
periods
when
the
instruments
were
in
stand­
by
mode
(
i.
e.,
between
sampling
runs
or
between
test
days),
they
were
maintained
at
normal
operating
temperatures
and
purged
with
conditioned
air.

3.
Background
Spectrum
Procedures
­
Each
instrument
was
allowed
to
stabilize
while
being
purged
with
Ultrahigh
Purity
(
UHP)
nitrogen
for
10
minutes.
The
FTIRS
spectra
were
monitored
during
this
time,
until
CO
and
HZ0
concentrations
reached
a
steady
state.
The
following
procedures
were
then
done:

a>
The
interferogram
signal
was
checked
using
signal
alignment
software.

b)
A
single
beam
spectrum
was
collected
and
inspected
for
irregularities.

C>
Using
the
single
beam
spectrum,
the
detector
was
checked
for
non­
linearity,
and
corrected
if
necessary.

d)
The
instrument
alignment
procedure
was
done.

9
A
background
spectrum
consisting
of
256
scans
was
collected.

4.
Analyzer
Diagnostics
­
Analyzer
diagnostics
were
done
by
analyzing
a
diagnostic
cylinder
containing
109
ppm
CO.
The
standard
was
an
EPA
Protocol
gas.
EECL
used
CO
because
it
has
distinct
spectral
features
that
are
sensitive
to
variations
in
system
operation
and
performance.
The
standard
was
introduced
directly
into
each
instrument,
and
instrument
readings
were
allowed
to
stabilize
for
five
minutes.
The
accuracy
and
precision
of
each
instrument
were
calculated.
The
pass/
fail
criterion
for
accuracy
and
precision
was
10%
of
the
concentration
of
the
Final
Report
­
Caterpillar
3508
EUI
6­
3
September
200
1
standard
gas.
A
second
diagnostic
standard
consisting
of
a
blend
of
CO*,
CO,
CH4,
and
NOx
was
analyzed
using
the
same
procedure.
Each
instrument
met
the
precision
and
accuracy
requirements.
Analyzer
diagnostic
data
are
presented
in
the
report
generated
by
EECL
5.
Indicator
Check
&
Sample
Integrity
Check
­
An
indicator
check
was
done
by
analyzing
an
indicator
standard.
A
10.66
ppm
formaldehyde
standard
was
introduced
directly
into
each
instrument.
The
instrument
readings
were
allowed
to
stabilize
and
a
5­
minute
data
set
was
collected.
The
indicator
standard
was
then
introduced
into
the
sample
system
at
the
sample
probe,
just
upstream
of
the
filter.
The
instrument
readings
were
allowed
to
stabilize
and
a
5­
minute
set
of
data
was
collected.
The
accuracy,
precision,
and
recovery
were
calculated
based
on
equations
in
the
document
entitled
"
Protocol
for
Performing
Extractive
FTIRS
Measurements
to
Characterize
Various
Gas
Industry
Sources
for
Air
Toxics",
prepared
by
Radian
International
for
the
Gas
Research
Institute.
The
pass/
fail
criterion
for
accuracy,
precision,
and
recovery
is
100
f
10%
of
the
known
standard
(
recovery
shall
be
100
f
10%
of
the
instrument
reading
when
the
indicator
gas
was
introduced
directly
into
the
instrument.)
Each
instrument
met
these
criteria.
Indicator
check
and
sample
integrity
data
sheets
are
included
with
the
EECL
report.

6.1.3
Backeround
Assessment
During
data
acquisition
procedures,
the
baseline
absorbance
was
continually
monitored.
If
at
any
time
the
baseline
spectrum
changed
by
more
than
0.1
absorbance
units,
the
instrument's
interferometer
was
realigned
and
a
new
background
spectrum
collected.

6.1.4
Post
Test
Checks
Upon
completion
of
the
daily
test
program
steps
4
and
5
of
the
pre­
test
calibration
procedures
were
repeated.
Both
of
the
FTIRS
analyzers
met
all
of
the
acceptance
criteria
for
the
calibration
and
QA
procedures.
Post
test
calibration
data
sheets
are
included
in
the
EECL
report.

6.1.5
FTIRS
Validation
Before
the
initiation
of
testing
of
the
engine,
both
FTIRS
sampling
and
analysis
systems
were
validated
for
formaldehyde,
acrolein,
and
acetaldehyde.
The
validation
was
conducted
by
personnel
from
ERG,
using
procedures
outlined
in
EPA
Method
301
"
Field
Validation
of
Pollutant
Measurement
Methods
from
Various
Waste
Media."
The
validation
was
conducted
using
a
dynamic
spiking
the
sample
gas
with
known
concentrations
of
formaldehyde,
acrolein,
and
acetaldehyde.
The
spike
gas
consisted
of
a
compressed
gas
cylinder
containing
a
mixture
of
acrolein
and
acetaldehyde.
Formaldehyde
was
added
to
the
Final
Report
­
Caterpillar
3508
EUI
6­
4
September
200
1
mixture
by
injecting
a
stock
formalin
solution
onto
a
heated
block
at
a
fixed
rate.
The
acrolein/
acetaldehyde
gas
standard
was
used
as
a
carrier
gas
for
the
vaporized
formaldehyde.
The
three­
component
mixture
was
injected
into
each
FTIRS
sampling
system
at
a
point
upstream
of
each
system's
filter.
Further
discussions
of
the
validation
procedures
employed
may
be
found
in
the
report
generated
by
EECL.

6.1.6
FTIRS
Detection
Limits
Table
6.1
presents
the
in­
stack
detection
limits
for
formaldehyde,
acetaldehyde,
and
acrolein
as
reported
by
CSU
EECL.
These
detection
limits
have
been
used
to
calculate
the
run­
by­
run
mass
detection
limits
for
each
of
the
target
pollutants.

6.2
CEMS
QA/
QC
PROCEDURES
The
following
paragraphs
describe
the
CEMS
quality
assurance
procedures
that
EECL
personnel
used
during
the
test
program.
The
calibration
and
QC
frequencies
far
exceeded
those
required
for
permanently­
installed,
compliance
analyzers,
but
are
less
than
those
specified
for
compliance
tests
by
EPA
(
40
CFR
60,
Appendix
A).
EECL
operates
their
CEMS
in
a
way
that
is
more
similar
to
permanently­
installed
analyzers.

6.2.1
Analvzer
Calibration
Gases
EECL
used
EPA
Protocol
calibration
gases.
The
calibration
gases
were
manufactured
by
Scott
Specialty
Gases.
For
this
program,
EPA
Protocol
1
calibration
gases
(
RATA
Class)
were
used.
Formaldehyde
and
acetaldehyde/
acrolein
standards
with
concentration
ranges
between
5
­
20
ppm
were
obtained
for
FTIRS
calibrations.
These
gases
are
not
available
as
EPA
Protocol
Gases,
so
EECL
specified
the
highest
quality
available.
Scott
supplied
certification
sheets,
which
may
be
found
in
the
Appendices
of
EECL's
test
report.

6.2.2
Resaonse
Time
Tests
Response
time
tests
were
done
on
each
sample
system
before
initiation
of
the
engine
test
program.
The
response
time
tests
were
done
before
the
FTIRS
validation
process
for
each
sampling
system.
The
response
time
of
the
slowest
responding
analyzer
(
Questar
Baseline)
was
determined.
Response
time
tests
conducted
at
the
EECL
indicated
sampling
system
response
times
of
1:
10
minutes.
This
is
the
time
for
the
Rosemount
Oxygen
Analyzer
(
the
slowest
responding
continuous
analyzer)
to
stabilize
to
response
output
of
the
analyzer.
The
Questar
Baseline
Industries
CH4/
Non­
CH4
analyzers
have
a
minimum
cycle
time
of
4:
50
minutes.
The
overall
response
time
for
these
analyzers
when
their
cycle
is
started
1:
10
minutes
after
a
sample
source
change
is
550
minutes.
When
the
methane/
non­
methane
analyzer
cycle
time
was
initiated
at
a
sample
source
change,
the,
overall
response
time
was
Final
Report
­
Caterpillar
3508
EUI
6­
5
September
200
1
TABLE
6.1
DETECTION
LIMITS
OF
FI'IRS
AND
CEMS
COMPOUNDS
ppmev­
partsper~
byvdume,
wetbasis
parts
per
million
by
volum?.
dry
basis
ppnwd
Q
15%
02
­
parts
per
niBion
by
vohnm,
dry
basis,
corrected
to
15%
oxygen
F&
al
Report
­
Ckerpillar
3508
EUI
6­
6
September
200
1
9:
00
minutes.
The
response
time
was
tested
to
assure
that
the
analyzers'
response
was
for
exhaust
gas
entering
the
sample
system
from
each
of
the
test
point
conditions.

6.2.3
Analvzer
Calibrations
Zero
and
mid­
level
span
calibration
procedures
were
done
on
the
CO,
CO*,
OS,
NOx,
and
THC
analyzers
before
each
test
day.
Zero
and
span
drift
checks
were
performed
upon
completion
of
each
data
point
and
upon
completion
of
each
test
day.
A
zero
and
mid­
level
gas
was
introduced
individually
directly
to
the
back
of
the
analyzers
before
testing
for
carbon
monoxide,
carbon
dioxide,
oxygen,
total
hydrocarbons,
Methane/
Non­
Methane,
and
oxides
of
nitrogen.
The
analyzers
output
response
was
set
to
the
appropriate
levels.
Each
analyzer's
stable
response
was
recorded.
From
this
data
a
linear
fit
was
developed
for
each
analyzer.
The
voltages
for
each
analyzer
were
recorded
and
used
in
the
following
formula:

Y
=
mx+
b
Where:
b
=
Intercept
m
=
Slope
X
=
Analyzer
or
transducer
voltage
Y
=
Engineering
Units
After
each
test
point
and
upon
completion
of
a
test
day,
calibrations
were
conducted
by
reintroducing
the
zero
and
span
gases
directly
to
the
back
of
the
analyzers.
The
analyzers'
stabilized
responses
were
recorded.
No
adjustments
were
made
during
testing
or
during
the
final
calibration
check.
Initial
calibration
values
and
all
calibration
checks
were
recorded
for
each
analyzer
during
the
daily
test
program.

The
before
and
after
calibrations
checks
were
used
to
determine
the
zero
and
span
drift
for
each
test
point
for
the
CO,
CO,,
OZ,
THC,
methane/
non­
methane,
and
NOx
analyzers.
The
zero
and
span
drift
checks
for
all
test
points
and
all
test
days
were
less
than
~
2.0%
of
the
span
value
of
each
analyzer
used
during
the
daily
test
program.
The
calibration
data
sheets
are
presented
in
the
test
report
generated
by
EECL.
Table
6.2
presents
the
types
and
frequencies
of
the
analyzer
calibrations
conducted
by
EECL.

6.2.4
Analvzer
Linearitv
Check
Analyzer
linearity
checks
were
done
before
beginning
the
test
program.
The
oxygen,
carbon
monoxide,
total
hydrocarbon,
methane/
non­
methane,
and
oxides
of
nitrogen
analyzers
were
"
zeroed"
using
either
zero
grade
nitrogen
or
hydrocarbon
free
air.
The
analyzers
were
allowed
to
stabilize,
their
output
was
recorded
and
then
spanned
using
the
mid­
level
calibration
gases.
The
analyzers
were
allowed
to
stabilize
a
second
time
and
their
output
was
recorded.
From
this
data
a
linear
fit
was
developed
for
each
analyzer.
The
voltage
for
each
analyzer
was
recorded
and
used
in
the
following
formula:

Final
Report
­
Caterpillar
3508
EUI
6­
7
September
200
1
TABLE
6.2
TYPES
AND
FREQUENCIES
OF
CEMS
ANALYZER
CALIBRATIONS
Calibration
Type
Gas
Calibration
Gas
Calibrant
Concentration
(
units
Frequency
Injection
of
%
of
span
(
l))
Point
Validation
Criterion
ACE
t2)
02,
co2,
co,
NO,

Methane/
Non­
Methane
Hydrocarbons
0
to
0.25,
40
to
60,
80
to
100
0
to
0.1,
25
to
35,
45
to
55,
80
to
90
­
3%
of
analyzer
span
for
each
gas
Before
each
Directly
into
,
engine
test
the
analyzer
<
5%
of
respective
cal.
gas
value
ZSD
t3)
02,
co2,
co,
NO,

Methane/
Non­
Methane
Hydrocarbons
0
to
0.25,
40
to
60
or
80
to
100
(`)

25
to
35,
45
to
55
Before
and
after
each
test
run
Directly
into
the
analyzer
All
errors
<
3%
of
span
All
errors
<
3%
of
span
SSB
t4)
NO,

MethaneMon­
Methane
Hydrocarbons
0
to
0.25,
40
to
60
or
80
to
90
(`
I
0
to
0.25,
25
to
35,
45
to
55
or
80
to
90
(`)
Before
and
after
each
Both
directly
test
day
into
the
Both
errors
analyzer
and
into
the
inlet
x5%
of
Before
and
after
each
of
the
sample
analyzer
spar
line
test
day
(*)
­
The
span
must
be
1.5
to
2.5
the
concentration
expected
for
each
pollutant
t2)
­
Analyzer
calibration
error
check
t3)
­
Zero
and
span
drift
check
t4)
­
Sampling
system
bias
check
t5)
­
Whichever
is
closer
to
the
exhaust
gas
concentration
Final
Report
­
Caterpillar
3508
EUI
6­
8
September
200
1
where:
Y
=
mx+
b
b
=
Intercept
m
=
Slope
X
=
Analyzer
or
transducer
voltage
Y
=
Engineering
Units
Using
the
linear
fit,
the
linear
response
of
the
analyzer
was
calculated.
Low­
level
and
high­
level
calibration
gases
were
individually
introduced
to
the
analyzers.
For
each
calibration
gas,
the
analyzers
were
allowed
to
stabilize
and
their
outputs
were
recorded.
Each
analyzer's
linearity
was
acceptable.
The
predicted
values
of
a
linear
curve
determined
from
the
zero
and
mid­
level
calibration
gas
responses
agreed
with
the
actual
responses
of
the
low­
level
and
high­
level
calibration
gases
within
&
2.0%
of
the
analyzer
span
value.
The
methane/
non­
methane
analyzers'
linearity
was
acceptable
as
the
predicted
valued
agreed
with
the
actual
response
of
the
low­
level
and
high­
level
calibration
gases
within
&
5.0%
of
the
actual
calibration
gas
value.
This
procedure
was
done
for
one
range
setting
for
each
analyzer.
The
Linearity
Check
data
sheets
are
presented
the
test
report
generated
by
EECL.

6.2.5
NO,
Converter
Check
EECL
did
NO,
converter
checks
before
the
test
program
began.
A
calibration
gas
mixture
of
known
concentration
between
240
and
270
ppm
nitrogen
dioxide
(
NO,)
and
160
to
190
ppm
nitric
oxide
(
NO)
with
a
balance
of
nitrogen
was
used.
The
calibration
gas
mixture
was
introduced
to
the
oxides
of
nitrogen
(
NO,)
analyzer
until
a
stable
response
was
recorded.
The
converter
was
considered
acceptable
if
the
instrument
response
indicated
a
90
percent
or
greater
NO,
to
NO
conversion.
The
NO2
Converter
Check
data
sheets
are
presented
in
the
test
report
generated
by
EECL.

6.2.6
Sample
Line
Leak
Check
The
sample
lines
were
leak­
checked
before
the
engine
test
program.
The
leak
check
procedure
was
done
for
both
pre­
catalyst
and
post­
catalyst
sample
trains.
The
procedure
was
to
close
the
valve
on
the
inlet
to
the
sample
filter
found
just
downstream
of
the
exhaust
stack
probe.
With
the
sample
pump
operating,
a
vacuum
was
pulled
on
the
exhaust
sample
train.
Once
the
maximum
vacuum
was
reached,
the
valve
on
the
$
ressure
side
of
the
pump
was
closed,
thus
sealing
off
the
vacuum
section
of
the
sampling
system.
The
pump
was
turned
off
and
the
pressure
in
the
sample
system
was
monitored.
The
leak
test
was
acceptable
as
the
vacuum
gauge
reading
dropped
by
an
amount
less
than
1
inch
of
mercury
over
a
period
of
1
minute.
The
Sample
Line
Leak
Check
data
sheets
are
presented
the
test
report
generated
by
EECL.

Final
Report
­
Caterpillar
3508
EUI
6­
9
September
200
1
6.2.7
Sample
Line
Inteeritv
Check
A
sample
line
integrity
check
was
done
before
and
upon
completion
of
each
test
day.
The
analyzers'
response
was
tested
by
first
introducing
a
mid­
level
calibration
gas
directly
to
the
NOx
analyzer.
The
analyzer
was
allowed
to
stabilize
and
the
response
recorded.
The
same
mid­
level
calibration
gas
was
then
introduced
into
the
analyzer
through
the
sampling
system.
The
calibration
gas
was
introduced
into
the
sample
line
at
the
stack,
upstream
of
the
inlet
sample
filter.
The
analyzer
was
allowed
to
stabilize
and
the
response
recorded.
The
analyzer
response
values
were
compared
and
the
percent
difference
did
not
to
exceed
&
5%
of
the
analyzer
span
value.

The
sample
line
integrity
check
was
to
be
done
for
both
the
NO,
and
methane/
non­
methane
analyzers.
Due
to
time
constraints,
EECL
performed
the
integrity
check
for
the
NO,
analyzers
only.
The
SSB
procedure
was
done
for
the
methane/
non­
methane
analyzers
before
and
upon
completion
of
the
test
program.
The
Sample
Line
Integrity
Check
data
sheets
are
presented
in
the
test
report
generated
by
EECL.

6.2.8
Carbon
Balance
Check
One
of
the
methods
used
to
calculate
mass
emissions
was
a
carbon
balance
calculation
developed
by
Southwest
Research
Institute
specifically
for
the
American
Gas
Association.
The
calculations
consist
of
a
theoretical
O2
calculation
based
upon
measured
exhaust
stack
constituents
and
fuel
gas
composition.
The
theoretical
exhaust
O2
is
then
compared
to
the
measured
exhaust
0,.
The
percent
difference
between
the
actual
and
theoretical
O2
measurements
was
within
55
%
of
the
measured
O2
reading.
The
O2
balance
was
done
for
every
1
­
minute
average
and
the
33­
minute
averaged
valued
for
each
test
point.

6.2.10
Fuel
Factor
Oualitv
Assurance
Checks
Besides
the
CEMS
calibration
and
QC
checks,
carbon
dioxide
and
oxygen
.
measurements
were
validated
by
calculating
the
fuel
factor,
F,,
using
the
following
equation:

F,'
20.9
­%
02
o/
oco2
The
values
of
F,
at
the
inlet
and
the
outlet
for
each
sampling
run
are
presented
in
Table
6.3.
For
distillate
fuel
oil
combustion,
the
value
of
F,
should
be
approximately
1.35.
The
F,
values
were
within
10%
of
the
expected
F,
for
all
of
the
runs
conducted.
Based
upon
these
results,
the
integrity
of
the
CEMS
sample
stream
was
not
compromised
due
to
leaks
in
the
sampling
system.

Final
Report
­
Caterpillar
3508
EUI
6­
10
September
200
1
TABLE
6.3
SUMMARY
OF
FUEL
FACTOR
VALUES
Run
Number
Inlet
F,

1
1.44
2
1.45
Outlet
F,

1.36
1.41
I
3
1.38
1.40
I
1.41
1.35
1.34
I
1.36
I
13
1.34
II
14
1.35
II
II
PAHl
I
1.36
I
1.33
II
PAH
2
PAH
3
Final
Report
­
Caterpillar
3508
EUI
6­
11
September
200
1
6.2.11
CEMS
Detection
Limits
For
each
of
the
sample
runs,
the
mass
detection
limits
of
the
CEMS
were
presented
previously
in
Table
6.1.
For
each
run,
the
detection
limit
was
calculated
using
analytical
detection
limit
data
supplied
by
EECL.
Table
6.4
summarizes
these
values.

TABLE
6.4
SUMMARY
OF
CEMS
ANALYTICAL
DETECTION
LIMITS
II
Parameter
I
Inlet
Detection
I
Outlet
Detection
Limit
Limit
II
Oxygen
I
0.01
%
volume
I
0.01
%
volume
II
Carbon
Dioxide
I
0.25
%
volume
I
0.15
%
volume
II
Nitrogen
Oxides
I
0.1
ppm
I
0.1
ppm
II
Carbon
Monoxide
I
10
PPm
I
2
PPm
ir­
Methane
I
2
PPm
I
2
PPm
Non­
methane
Hydrocarbons
2
PPm
0.2
ppm
Total
Hvdrocarbons
0.04
ppm
0.04
ppm
6.3
GCMS
QA/
QC
PROCEDURES
Each
day
the
GCMS
measurement
system
was
tuned
according
to
the
criteria
identified
in
the
method.
Achieving
the
criteria
for
a
valid
mass
spectral
tune
and
achieving
the
internal
standard
relative
mass
abundances
during
each
GCMS
run
(
see
Tables
3
and
4
of
Alternate
Method
17)
verifies
continuing
instrument
performance
and
ensures
that
the
QA/
QC
criteria
of
the
method
are
achieved.
Achieving
the
criteria
for
a
valid
tune
also
allows
searches
of
the
NIST
Mass
Spectral
library
and
verification
of
compounds
that
are
not
contained
in
the
instrument
specific
calibration.

Daily
system
calibrations
were
conducted
to
check
both
the
validity
of
the
initial
instrument
calibration
and
the
effectiveness
of
the
sampling
system
to
transport
the
target
analytes
to
the
instrumentation.
Daily
system
calibration
checks
were
conducted
at
1
ppm
using
the
blended
mixture
of
test
program
analytes.
For
system
calibrations,
the
gas
standard
was
directed
through
the
entire
sampling
system
(
including
the
filter).
The
certified
standard
Final
Report
­
Caterpillar
3508
EUI
6­
12
September
200
1
used
for
this
procedure
was
independent
from
that
used
to
generate
the
initial
calibration.
All
target
analytes
met
the
method
system
calibration
criteria
of
20%
except
for
toluene
on
September
1,
1999
that
was
24%.
Immediately
following
the
system
continuing
calibration,
nitrogen
was
allowed
to
flow
through
both
measurement
systems
and
a
system
blank
was
acquired.
No
analytes
were
detected
in
any
of
the
system
blank
analyses.
The
results
of
the
continuing
calibration
checks
are
presented
in
Table
6.5.

Periodic
analyte
spiking
was
conducted
at
1
ppm
on
August
3
1,
100
ppb
on
September
1,
and
at
multiple
concentration
levels
on
September
2
to
determine
the
extent
of
the
soot
adsorption
for
benzene
and
toluene.
The
analyte
spike
was
delivered
at
a
ratio
of
exactly
1
part
of
spike
gas
to
9
parts
of
effluent
as
measured
by
calibrated
mass
flowrneters.
The
analyte
spike
mixture
was
injected
into
the
sampling
system
immediately
upstrp
of
the
particulate
filters
in
both
measurement
systems
so
that
the
combined
effect
of
the
effluent
matrix
and
the
diesel
soot
could
be
evaluated.

From
the
analyte
spiking
procedures
conducted
on
August
3
1,
it
was
observed
that
the
soot
adsorbed
about
60%
of
these
compounds
at
the
1
ppm
concentration
level.
The
adsorption
was
fairly
consistent
with
time
at
this
spike
level.

Because
the
1
ppm
analyte
spike
concentration
was
about
10
times
greater
than
the
effluent
concentration
of
benzene
and
toluene,
analyte
spikes
at
100
ppb
were
conducted
on
September
1.
The
purpose
was
to
document
more
fully
the
extent
of
the
adsorption
at
applicable
concentration
levels.
Filter
changes
at
the
beginning
of
the
day
combined
with
sampling
ambient
air
between
runs
minimized
the
adsorption
of
the
benzene
and
toluene
and
yielded
acceptable
recoveries
at
the
100
ppb
level
after
the
first
few
runs.
At
the
end
of
the
day
however,
soot
build­
ups
again
reduced
the
analyte
spike
recoveries
to
the
levels
seen
the
previous
day.

Analyte
spiking
done
on
September
2
was
conducted
at
multiple
concentration
levels.
Recoveries
for
benzene
were
within
the
acceptable
30%
tolerance.
Toluene
recoveries
ranged
from
50­
60%
of
the
expected
values.
Table
6.6
presents
the
results
for
the
analyte
spiking
procedures
conducted
during
this
test
program.

One
is
tempted
to
correct
actual
measurement
data
based
on
the
results
of
analyte
spikes.
Such
corrections
may
sometimes
decrease
the
accuracy
of
the
data.
The
following
factors
should
be
considered:

1.
Calculation
of
analyte
spike
recoveries
may
be
adversely
affected
due
to
actual
changes
in
the
underlying
native
concentrations
for
compounds
present
during
the
spiking
procedure.
In
these
tests,
our
recovery
calculations
assume
that
the
benzene
concentration
preceding
the
spike
is
the
same
as
the
native
benzene
concentration
during
the
spike.
However,
the
benzene
concentration
was
Final
Report
­
Caterpillar
3508
EUI
6­
13
September
200
1
TABLE
6.5
SUMMARY
OF
GCMS
CONTINUING
CALIBRATIONS
Compound
Expected
Value
(
PPW
August
31,1999
September
1,1999
September
2,1999
Result
VW
Result
c­
w
Result
W)
(
PPN
Diff.
(
PPN
Diff.
(
PPN
Diff.

Catalyst
Inlet
Hexane
1.03
1.02
­
0.97
0.95
­
7.77
1
.
oo
­
2.91
Benzene
1.04
0.88
­
15.4
0.94
­
9.62
1
.
oo
­
3.85
Toluene
I
1.01
1
0.81
1
­
19.8
1
0.77
1
­
23.8
1
1.01
I
0
Ethyl
Benzene
I
1.04
I
0.91
I
­
12.5
I
1.01
I
­
2.88
1
1.01
I
­
2.88
m/
p­
Xylene
I
2.06
1
1.86
1
­
9.71
I
1.93
I
­
6.31
1
2.00
1
­
2.91
Styrene
1.04
1.02
­
1.92
0.83
­
20.2
1
.
oo
­
3.85
o­
Xylene
1.03
0.91
­
11.7
1.01
­
1.94
1
.
oo
­
2.91
1,3
­
Butadiene
1.03
Hexane
1.03
Benzene
1.04
Toluene
1.04
Ethyl
Benzene
1.04
m/
p­
Xylene
2.06
Styrene
1.04
o­
Xylene
1.03
Catalyst
Outlet
1.00
­
2.91
1.09
5.83
1.01
­
1.94
1.01
­
1.94
1.01
­
1.94
1.01
­
1.94
1.01
­
2.88
1.03
­
0.96
1.01
­
2.88
1.00
­
3.85
1.09
4.81
1.00
­
3.85
1.01
­
2.88
1.10
5.77
1.01
­
2.88
2.02
­
1.94
2.14
3.88
2.03
­
1.46
1.00
­
3.85
1.07
2.88
1
.
oo
­
3.85
1
.
oo
­
2.91
1.03
0
1
.
oo
­
2.91
i
*

­­
I
Final
Report
­
Caterpillar
3508
EUI
6­
14
September
200
1
TABLE
6.6
GCMS
ANALYTE
SPIKE
RECOVERIES
I
Benzene
Toluene
Inlet
Recovery
W)
Outlet
Recovery
VW
Inlet
Recovery
rw
Outlet
Recovery
(
0
OO
August
31,1999
(
1
ppm
standard)

Run1
I
39
I
88
I
28
I
60
Run
14
I
42
I
88
I
30
I
60
Run
13
I
42
I
88
I
30
I
60
Run
10
I
42
I
88
I
29
I
60
Run9
I
42
I
88
I
29
I
60
September
1,1999
(
100
ppb
standard)

Run4
I
81
I
104
I
74
I
58
Run
11
I
83
I
107
I
26
I
48
Run
12
I
83
I
I
26
I
Run2
I
57
I
89
I
69
I
48
September
2,1999
(
100
ppb
standard)

Run3
I
71
I
117
I
54
7
58
Run3
76
126
61
48
Run3
71
120
67
48
Final
Report
­
Caterpillar
3508
EUI
6­
15
September
200
1
observed
to
vary
significantly
among
the
four
samples
analyzed
for
certain
test
ll.
UlS.

2.
The
analyte
spike
procedure
used
equipment
that
is
not
part
of
the
method
and
thus
may
have
added
errors.
Calculation
of
spike
recoveries
includes
measurement
errors
associated
with
the
sample
flow
system
(
rotameter
or
mass
flow
meter
that
is
normally
used
only
for
assuring
constant
rate
sampling)
and
the
mass
flow
meter
used
to
monitor
the
spike
rate.
Both
measurements
contribute
some
additional
uncertainty
due
to
their
accuracy
and
precision.
The
sample
flow
meter
is
also
susceptible
to
changes
in
composition
and
molecular
weight
of
the
sample
gas.

3.
The
1
ppm
analyte
spikes
were
significantly
greater
than
the
sample
concentrations
and
thus
may
affect
the
dynamic
equilibrium
between
the
benzene
and
soot
in
the
exhaust
stream.

4.
The
100
ppb
analyte
spikes
involve
the
use
of
a
second
calibration
standard,
different
from
the
calibration
standard
used
for
the
fundamental
multi­
point
calibration.
This
too,
may
have
introduced
another
source
of
uncertainty.
(
Note:
the
tolerance
allowed
for
audit
of
the
calibration
curve
using
the
1
ppm
gas
is
20%.)

5.
Both
the
unspiked
and
the
spiked
concentration
measurements
are
influenced
by
noise
or
imprecision
associated
with
the
mass
spectrometer
measurements
particularly
at
very
low
concentration
levels.
(
Note
that
the
tolerance
for
the
daily
continuing
calibration
checks
is
20%
and
that
replicate
injections
are
performed
to
minimize
the
effects
of
noise
in
constructing
the
fundamental
calibration
curve.)

6.
The
limitations
of
the
analyte
spiking
procedure
require
that
the
normal
acceptable
tolerance
be
70%
to
130%
recovery
of
a
1
ppm
spike.
The
spike
criterion
is
less
restrictive
than
the
+
20%
tolerance
for
the
continuing
calibration
checks
because
it
reflects
the
limitations
stated
above.
The
spike
criterion
applied
to
a
100
ppb
spike
becomes
very
restrictive.
(
i.
e.,
30
ppb
for
a
100
ppb
spike
versus
300
ppb
for
a
1
ppm
spike)

6.3.1
GCMS
Detection
Limits
Tables
6.7
and
6.8
present
the
GCMS
Detection
Limits
at
the
pre­
catalyst
and
the
post­
catalyst
sampling
locations.
PES
used
the
analytical
detection
limits
supplied
by
EM1
to
calculate
the
run­
by­
run
mass
detection
limits.

Final
Report
­
Caterpillar
3508
EUI
6­
16
September
200
1
­
I
"
l
r
.
1".
LI
I
­
u
,
­
x
­
.
TABLE
6.7
DETECTION
LIMITS
OF
GCMS
COMPOUNDS
AT
CATALYST
INLET
Run
11)
1
Run1
1
Run2
1
Run3
I
Run4
I
Run9
I
Run10
I
Run11
1
Run12
I
Run13
I
Run14
I
PAH
1
I
PAH
2
I
PAH
3
1
H
B
T
fwmvd
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
No
data
,3­
Butadiene
pg/
bhp­
hr
2000
3000
2000
2000
4000
3000
3000
4000
2000
3000
3000
3000
Nodata
plb/
hr
5000
5000
3000
4000
8000
6000
7000
8000
4000
7000
7000
5000
No
data
wmvd
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
No
data
lexane
pg/
bhp­
hr
1000
1000
1000
1000
2000
1000
2000
2000
1000
2000
1000
1000
No
data
plb/
hr
2000
2000
1000
2000
4000
3000
4000
4000
2000
4000
3000
2000
No
data
ppmvd
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
No
data
enzene
pg/
bhp­
hr
100
100
100
100
100
100
100
100
100
100
100
100
No
data
plb/
hr
200
200
100
200
300
200
300
300
200
300
300
200
No
data
wmvd
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
No
data
oluene
pg/
bhp­
hr
100
100
100
100
200
100
200
200
100
200
200
200
No
data
plb/
hr
200
200
200
200
400
300
400
400
200
400
400
300
No
data
wmvd
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
No
data
thyl
Benzene
pg/
bhp­
hr
100
200
100
100
200
100
200
200
100
200
200
200
No
data
plb/
hr
300
300
200
200
500
300
400
400
300
400
400
300
No
data
wmvd
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
No
data
/
p­
Xylene
pg/
bhp­
hr
300
300
300
200
400
300
400
400
200
400
400
400
No
data
plb/
hr
600
500
400
400
900
700
900
900
500
900
800
600
No
data
ppmvd
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
No
data
tyrene
pg/
bhp­
hr
1700
2400
1800
1400
2800
2000
2600
2700
1600
2700
2400
2600
No
data
plb/
hr
3700
3600
2400
2800
6100
4400
5700
5800
3400
5800
5300
4000
Nodata
wmvd
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
No
data
,
Xylene
pg/
bhp­
hr
300
300
300
200
400
300
400
400
200
400
400
400
No
data
plb/
hr
600
500
400
400
900
700
900
900
500
900
800
600
No
data
E
m
Final
Report
­
Caterpillar
3508
EUI
6­
17
September
200
1
A
I
TABLE
6.8
DETECTION
LIMITS
OF
GCMS
COMPOUNDS
AT
CATALYST
OUTLET
tun
ID
Run1
Run2
Run3
Run4
Run9
Run10
Run11
Run12
Run13
Run14
PAH
1
PAHZ
PAH3
wmvd
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
No
data
,3­
Butadiene
pg/
bhp­
hr
500
1000
1000
1000
1000
500
500
1000
500
1000
500
1000
No
data
plb/
hr
1000
1000
1000
1000
2000
1000
1000
2000
1000
2000
1000
1000
No
data
ppmvd
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
No
data
lexane
pg/
bhp­
hr
500
1000
370
1000
500
500
500
500
500
500
500
1000
No
data
plb/
hr
1000
1000
500
1000
1000
1000
1000
1000
1000
1000
1000
1000
No
data
wmvd
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.08
0.02
0.02
0.02
0.02
No
data
lenzene
pg/
bhp­
hr
100
100
100
100
100
100
300
800
100
100
100
100
No
data
vlb/
hr
200
200
100
200
300
200
600
1700
200
300
300
200
No
data
wmvd
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
No
data
oluene
pg/
bhp­
hr
100
100
100
100
200
100
200
200
100
200
200
200
No
data
plb/
hr
200
200
200
200
400
300
400
400
200
400
400
300
No
data
ppmvd
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
No
data
thy1
Benzene
pg/
bhp­
hr
100
200
100
100
200
100
200
200
100
200
200
200
No
data
plb/
hr
300
300
200
200
500
300
400
400
300
400
400
300
No
data
wmvd
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
No
data
I/
p­
Xylene
pglbhp­
hr
300
400
300
200
400
300
400
400
200
400
400
400
No
data
plb/
hr
600
600
400
400
900
700
900
900
500
900
800
600
No
data
wmvd
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
No
data
tyrene
pg/
bhp­
hr
500
1000
1000
1000
1000
500
500
500
500
500
500
1000
No
data
plb/
hr
1000
1000
1000
1000
2000
1000
1000
1000
1000
1000
1000
1000
No
data
ppmvd
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
No
data
.
Xylene
pglbhp­
hr
500
1000
300
200
500
500
500
500
500
500
500
1000
No
data
plb/
hr
1000
1000
400
400
1000
1000
1000
1000
1000
1000
1000
1000
No
data
Final
Report
­
Caterpillar
3508
EUI
6­
18
September
200
1
6.4
CARB
429
QA/
QC
PROCEDURES
The
following
text
describes
the
QA/
QC
procedures
employed
by
PES
and
ERG
during
the
PAH
sampling
and
analysis.

6.4.1
Calibration
of
CARB
429
Sampliw
Apparatus
Because
no
mechanism
exists
for
an
independent
measurement
of
emissions
from
the
source,
careful
preparation,
checkout,
and
calibration
of
the
sampling
and
analysis
equipment
is
essential
to
ensure
collection
of
high
quality
data.
PES
maintains
a
comprehensive
schedule
for
preventive
maintenance,
calibration,
and
preparation
of
the
source
testing
equipment.

6.4.1.1
Barometers.
PES
used
aneroid
barometers
calibrated
against
a
station
pressure
value
reported
by
a
nearby
National
Weather
Service
Station
and
corrected
for
elevation.

6.4.1.2
Temperature
Sensors.
The
responses
of
the
Type
K
thermocouples
used
in
the
field
testing
program
were
checked
using
Calibration
Procedure
2e
as
described
in
the
Quality
Assurance
Handbook.
The
response
of
each
temperature
sensor
was
recorded
when
immersed
in
an
ice
water
bath,
at
ambient
temperature,
and
in
a
boiling
water
bath;
each
response
was
checked
against
an
ASTM
3F
reference
thermometer.
Table
6.9
summarizes
the
results
of
the
thermocouple
checks
and
the
acceptable
levels
of
variance.
Digital
temperature
readouts
were
checked
for
calibration
using
a
thermocouple
simulator
having
a
range
of
O­
2400
"
F.

6.4.1.3
Pitot
Tubes.
PES
used
Type
S
Pitot
tubes
or
Standard
Pitot
tubes
constructed
according
to
EPA
Method
2
specifications.
Type
S
Pitot
tubes
were
calibrated
against
the
dimensional
criteria
described
in
Method
2
using
Calibration
Procedure
2a
as
described
in
the
Quality
Assurance
Handbook,
Volume
III,
1994.
Type
S
Pitot
tubes
meeting
these
criteria
are
assigned
a
pitot
coefficient
(
C,,)
of
0.84.
Standard
Pitot
tubes
were
checked
for
dimensional
criteria
using
Calibration
Procedure
2b
as
described
in
the
Quality
Assurance
Handbook,
Volume
III,
1994.
Standard
Pitot
tubes
meeting
these
criteria
were
assigned
a
pitot
coefficient
(
C,)
of
0.99.
Table
6.10
summarizes
the
results
of
the
pitot
tube
checks
and
the
acceptable
levels
of
variance.

6.4.1.4
Differential
Pressure
Gawes.
PES
used
Dwyer
inclined/
vertical
manometers
to
measure
differential
pressures
including:
velocity
pressure,
static
pressure,
and
orifice
meter
pressure.
PES
chose
manometers
having
sufficient
sensitivity
to
measure
pressures
over
the
entire
range
of
expected
values
accurately.
Manometers
are
primary
standards
and
require
no
calibration.

Final
Report
­
Caterpillar
3508
EUI
6­
19
September­
200
1
­­

TABLE
6.9
CARB
429
SAMPLE
TRAIN
SUMMARY
OF
TEMPERATURE
SENSOR
CALIBRATION
DATA
Temp.
Sensor
I.
D.
Usage
RMB­
15
Dry
Gas
Meter
Inlet
RMB­
15
Dry
Gas
Meter
Outlet
I
MB­
1
1
Dry
Gas
Meter
Inlet
MB­
1
1
Dry
Gas
Meter
Outlet
SH­
1
I
Impinger
Exit
SH­
5
Impinger
Exit
Temperature,
"
R
Absolute
Difference
493
493
534
535
668
668
T
0
0.19
0
492
492
0
534
534
0
670
668
0.30
492
492
0
534
534
0
668
668
0
492
492
0
536
536
0
668
668
0
492
493
0.20
531
531
0
667
667
0
EPA
Criteria
%

­
1.5
­
1.5
­
1.5
­
1.5
e1.5
ti1.5
­
1.5
­
1.5
­
1.5
e1.5
ck1.5
e1.5
­
1.5
e1.5
e1.5
e1.5
­
1.5
e1.5
Final
Report
­
Caterpillar
3508
EUI
6­
20
September
200
1
TABLE
6.10
CARB
429
SAMPLE
TRAIN
SUMMARY
OF
PITOT
TUBE
CALIBRATION
DATA
Measurement
EPA
Criteria
aI
<
10'
Pitot
HPP­
1
0"
Pitot
HPP­
2
0'

a2
I
<
10"
I
0"
I
1"

2
A
I
I
3
l/
32
I
3
l/
32
W=
Atany
I
~
0.125
in.
I
0.0160
I
0.0122
W=
Atan8
I
~
0.0.3
125
in.
I
0.0160
I
0
I
0.1875
in.
s
D,
<
0.375
in.
I
318
318
Ai2D,
1
1.05
5
P*/
D,
5
1.50
I
1.29
I
0.29
I
Acceptable
?:
I
YES
I
YES
I
Assigned
Coeffkient:
I
0.84
I
0.84
Final
Report
­
Caterpillar
3508
EUI
6­
21
September
200
1
6.4.1.5
Drv
Gas
Meter
and
Orifice.
The
CARB
Method
429
dry
gas
meters
and
orifices
were
calibrated
according
to
Calibration
Procedure
5
in
the
Quality
Assurance
Handbook.
This
procedure
requires
direct
comparison
of
the
dry
gas
meter
to
a
reference
dry
test
meter.
PES
calibrates
its
reference
dry
test
meter
annually
against
a
wet
test
meter.
Before
its
initial
use
in
the
field,
the
metering
system
was
calibrated
at
several
flow
rates
over
the
normal
operating
range
of
the
metering
system.
Individual
meter
calibration
factors
(
y)
cannot
differ
from
the
average
by
more
than
0.02,
and
the
results
of
individual
meter
orifice
factors
(
AH&
cannot
differ
from
the
average
by
more
than
0.20.
After
field
use,
the
metering
system
calibration
was
checked
at
the
average
flow
rate
and
highest
vacuum
observed
during
the
test
period.
The
results
of
the
post­
test
meter
correction
factor
check
cannot
differ
by
more
than
5%
from
the
average
meter
correction
factor
obtained
during
the
initial,
or
thereafter,
the
annual
calibration.
Table
6.11
presents
the
results
of
the
dry
gas
meter
and
orifice
calibrations.
All
dry
gas
meters
and
orifices
used
in
this
test
program
met
the
method
calibration
requirements.

TABLE
6.11
CARB
429
SAMPLE
TRAIN
SUMMARY
OF
DRY
GAS
METER
AND
ORIFICE
CALIBRATION
DATA
Meter
Box
No.
Dry
Gas
Meter
Correction
Factor
(
y)
Meter
Orifice
Coefficient
(
AH&

Pre­
test
Post­
test
%
Diff.
EPA
Criteria
Average
Range
EPA
Criteria
MB­
10
0.999
0.999
0.0
<
5%
1.89
1.83
­
1.94
1.69
­
2.09
RMB­
15
1.001
0.997
­
0.40
­
6%
1.87
1.79
­
1.98
1.67
­
2.07
6.4.2
Reagents
and
Glassware
Prenaration
Before
field
testing,
PES
pre­
cleaned
all
sample
train
glassware
following
the
procedures
in
CARB
Method
429.
Specifically,
the
glassware
was
cleaned
according
to
the
following
protocol.

1.
2.
3.
4.
5.
Wash
in
hot
soapy
water
with
Alconox.
Rinse
three
times
with
tap
water.
Rinse
three
times
with
reagent
(
i.
e.,
deionized)
water.
Soak
in
10%
(
v/
v)
nitric
acid
(
HNO,)
solution
for
a
minimum
of
4
hours.
Rinse
three
times
each
with
pesticide­
grade
acetone,
hexane,
and
methylene
chloride,
and
allow
to
air
dry.

Final
Report
­
Caterpillar
3508
EUI
6­
22
September
200
1
After
preparation
of
the
glassware,
the
openings
were
sealed
with
Teflon
tape
to
prevent
contamination,
and
the
glassware
wrapped
and
packed
for
transport
to
the
EECL.
ERG
prepared
the
XAD­
2@
sorbent
resin
traps.
ERG
then
pre­
spiked
the
traps
with
surrogates
and
capped
them
with
glass
balls
and
sockets.
Impinger
water
used
was
organic­
free,
reagent
grade.
Pesticide­
grade
acetone,
hexane,
and
methylene
chloride
were
used
as
recovery
solvents.

6.4.3
On­
site
Measurements
The
on­
site
QA/
QC
activities
included:

6.4.3.1
Measurement
Sites.
Before
sampling,
PES
checked
the
dimensions
of
the
exhaust
duct
to
assure
that
the
port
locations
complied
with
Method
1
criteria.
PES
confirmed
the
distances
to
upstream
and
downstream
disturbances
and
test
port
locations.
PES
also
measured
inside
stack
dimensions
through
perpendicular
ports
to
assure
uniformity
of
the
stack
cross
sectional
area.
PES
measured
the
inside
stack
dimensions,
stack
wall
thickness,
and
sample
port
lengths
to
the
nearest
0.1
inch.

6.4.3.2
Velocitv
Measurements.
PES
assembled,
leveled,
zeroed,
and
leak­
checked
all
velocity
measurement
apparatus
before
and
after
each
sampling
run.
The
stack
static
pressure
was
determined
at
a
single
point.
PES
selected
a
point
of
average
velocity
pressure
found
during
the
pre­
test
velocity
traverse.

6.4.3.3
Moisture.
During
sampling,
the
exit
gas
temperature
of
the
last
impinger
in
each
sampling
train
was
maintained
below
68
°
F
to
ensure
condensation
of
stack
gas
water
vapor.
The
moisture
gain
in
the
impinger
train
due
to
flue
gas
moisture
was
determined
gravimetrically
using
a
digital
top­
loading
electronic
balance
with
a
resolution
of
0.1
g.

6.4.4
Analytical
Oualitv
Assurance
PES
and
ERG
personnel
employed
several
methods
to
ensure
the
quality
of
the
PAH
analytical
data.
These
methods
included
analysis
of
reagent
blanks,
a
laboratory
method
blank,
and
field
blanks.
In
addition,
the
XAD­
2
sorbent
traps
were
spiked
with
isotopically
labeled
internal
standards.
The
recovery
efficiency
of
the
internal
standards
is
used
to
evaluate
method
performance.
The
results
of
these
QA
checks
are
discussed
in
the
following
paragraphs.

6.4.4.1
Blank
Analyses.
During
the
field
testing,
PES
personnel
collected
blanks
of&
e
CARB
429
sampling
train
reagents
to
quantify
contamination
levels.
Field
blank
trains
were
assembled,
transported
to
each
sampling
site,
and
leak
checked.
The
field
blank
trains
were
then
returned
to
the
PES
field
laboratory,
where
they
were
recovered
in
the
same
manner
as
the
trains
used
for
sampling.
The
field
blank
train
impingers
and
connecting
glassware
were
the
same
components
used
during
actual
sampling.
Since
the
sampling
glassware
is
cleaned
Final
Report
­
Caterpillar
3508
EUI
6­
23
September
200
1
after
each
run
and
reused,
analysis
of
field
blank
trains
is
used
to
find
out
if
poor
cleanup
technique
caused
cross­
contamination
between
sampling
runs.
Per
CARB
Method
429,
PES
did
not
correct
any
of
the
PAH
results
for
blank
results.
The
results
of
the
reagent
and
field
blank
analyses
are
presented
in
Table
6.12.
The
levels
of
any
unlabeled
analyte
quantified
in
the
blank
train
must
not
exceed
20
percent
of
the
level
of
that
analyte
in
the
sampling
train.

6.4.4.2
Internal
Standard
Recoveries.
Table
6.13
presents
the
recovery
efficiencies
of
isotopically
labeled
surrogate
compounds.
Recovery
efficiency
gives
a
measure
of
the
capture
efficiency
and
the
efficiency
of
the
solvent
extraction
for
specific
compounds.
Recoveries
for
each
of
the
internal
standards
must
be
greater
than
50
percent
and
less
than
150
percent
of
the
known
value.
This
criterion
is
used
to
assess
method
performance.
Because
this
is
an
isotope
dilution
technique,
it
should
be
independent
of
internal
standard
recovery.
Lower
recoveries
do
not
necessarily
invalidate
the
analytical
results
for
PAH,
but
they
may
result
in
higher
detection
limits.

6.4.5
CARB
429
Detection
Limits
Tables
6.14
and
6.15
present
the
in­
stack
detection
limits
of
each
PAH
compound
before
the
catalyst
and
after
the
catalyst.

Final
Report
­
Caterpillar
3508
EUI
6­
24
September
200
1
TABLE
6.12
SUMMARY
OF
CARB
429
BLANK
RESULTS
Compound
Laboratory
Blank
Result
(
pg)
Reagent
Blank
Result
(
W'
Inlet
Field
Outlet
Blank
Field
Blank
Result
(
pg)
Result
(
pg)

Naphthalene
ND
0.847
1.081
0.965
Acenaphthylene
I
ND
I
ND
I
ND
I
Acenaphthene
I
ND
I
ND
I
ND
I
ND
Fluorene
Phenanthrene
I
ND
1
ND
1
ND
1
0.029
Anthracene
I
ND
I
ND
I
ND
I
0.069
Fluoranthene
I
ND
I
ND
I
ND
I
ND
Pyrene
I
ND
I
ND
I
ND
I
ND
Benzo(
a)
anthracene
I
ND
I
ND
I
ND
I
ND
Chrysene
I
ND
I
ND
I
ND
I
ND
Benzo(
b)
fluoranthene
I
ND
I
ND
I
ND
I
ND
Ben.
zo(
k)
fluoranthene
I
ND
I
ND
I
ND
I
ND
Benzo(
a)
pyrene
I
ND
I
ND
I
ND
I
ND
Indeno(
1,2,3­
cd)
pyrene
I
ND
I
ND
I
ND
I
ND
1
Dibenz(
a,
h)
anthracene
I
ND
I
ND
I
ND
I
ND
Benzo(
g,
h,
i)
perylene
ND
1
ND
ND
1
ND
'
pg
­
microgram
NOTE:
The
reagent
blank
value
is
the
sum
of
separate
analyses
hexane,
acetone,
methylene
chloride,
and
distilled
water
blank
samples.

Final
Report
­
Caterpillar
3508
EUI
6­
25
September
200
1
TABLE
6.13
SUMMARY
OF
CARB
429
SURROGATE
RECOVERIES
Surrogate
Compound
Naphthalene­
d8
Acenaphthylene­
d8
Acenaphthene­
d
10
Fluorene­
dl0
Phenanthrene­
d
10
Anthracene­
d
10
Fluoranthene­
dl0
Pyrene­
d
10
Benzo(
a)
anthracene­
d
12
Chrysene­
d
12
Benzo(
b)
fluoranthene­
d
12
Benzo(
k)
fluoranthene­
d12
Benzo(
a)
pyrene­
d
12
Indeno(
1,2,3­
cd)
pyrene­
d
12
Dibenz(
a,
h)
anthracene­
d14
Benzo(
g,
h,
i)
perylene­
d12
Lab
Blank
v4
68
119
100
98
95
140
101
105
136
103
124
123
116
119
119
111
Field
Blanks
PAH
Run
1
PAH
Run
2
PAH
Run
3
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
cw
VW
cw
WQ
cw
VW
c­
w
cw
21
39
67
51
42
52
63
40
23
57
93
11
49
10
53
8
56
73
66
92
8
80
86
65
78
103
93
69
95
106
89
75
81
92
81
56
83
72
79
155
82
80
ND
79
14
89
12
76
84
101
83
74
85
95
78
70
84
102
66
76
75
99
67
45
130
100
44
80
93
102
68
71
82
63
78
47
78
60
66
79
113
78
98
60
96
73
66
71
137
82
91
64
97
74
64
0
106
49
ND
52
ND
37
44
96
84
62
75
76
80
43
49
103
78
63
72
77
78
46
48
86
67
55
57
59
57
30
Final
Report
­
Caterpillar
3508
EUI
6­
26
September
200
1
TABLE
6.14
DETECTION
LIMITS
OF
PAH
COMPOUNDS
AT
CATALYST
INLET
Run
ID
Date
Acenaphthene
ug/
bhp­
hr
ulblhr
PAH
1
PAH
2
PAH
3
8131199
911199
g/
2/
99
0.18
0.13
0.17
0.40
0.20
0.23
Average
0.15
0.22
II
Acenaphthylene
ug/
bhp­
h
ulb/
hr
0.18
0.40
0.13
0.20
I
0.17
0.23
I
0.15
0.22
II
Anthracene
ug/
bhp­
hr
0.18
0.13
0.17
0.15
ulb/
hr
0.40
0.20
0.23
0.22
Benzo(
a)
anthracene
ug/
bhp­
hr
0.18
0.13
0.17
0.15
ulb/
hr
0.40
0.20
0.23
0.22
Benzo(
b)
fluoranthene
pg/
bhp­
hr
0.18
0.13
0.17
0.15
ulblhr
0.40
0.20
0.23
0.22
I
Benzo(
k)
fluoranthene
ug/
bhp­
h
ulb/
hr
1
0.18
0.40
I
0.13
0.20
I
0.17
0.23
I
0.15
0.22
I
Benzo(
g,
h,
i)
perylene
ug/
bhp­
hr
0.18
0.13
0.17
0.15
ulb/
hr
0.40
0.20
0.23
0.22
pg/
bhp­
hr
0.18
0.13
0.17
0.15
Benzo(
a)
pyrene
ulb/
hr
0.40
0.20
0.23
0.22
Chrysene
ug/
bhp­
hr
0.18
0.13
0.17
0.15
ulb/
hr
0.40
0.20
0.23
0.22
'
Dibenz(
a,
h)
anthracene
ug/
bhp­
hr
0.18
0.13
0.17
0.15
ulblhr
0.40
0.20
0.23
0.22
Fluoranthene
ug/
bhp­
hr
0.18
0.13
0.17
0.15
ulb/
hr
0.40
0.20
0.23
0.22
0.15
Fluorene
ug/
bhp­
hr
0.18
0.13
0.17
ulb/
hr
0.40
0.20
0.23
0.22
I
Indeno(
l,
2,3­
cd)
pyrene
ug/
bhp­
h
ulb/
hr
1
0.18
0.40
I
0.13
0.20
I
0.17
0.23
I
0.15
0.22
I
Naphthalene
Phenanthrene
ug/
bhp­
hr
0.18
0.13
0.17
0.15
ulblhr
0.40
0.20
0.23
0.22
pg/
bhp­
hr
0.18
0.13
0.17
0.15
ulblhr
0.40
0.20
0.23
0.22
I
Pvrane
ug/
bhp­
h
0.18
0.13
0.17
I
0.15
I
II
.
J""'
plb/
hr
1
0.40
0.20
I
0.23
I
0.22
11
NOTE:
The
reported
analytical
detection
limit
for
each
compound
was
0.1
pg
per
run.

Final
Report
­
Caterpillar
3508
EUI
6­
27
September
200
1
TABLE
6.15
DETECTION
LIMITS
OF
PAH
COMPOUNDS
AT
CATALYST
OUTLET
Run
ID
Date
I
PAH
1
813119Sj
pg/
bhp­
hr
Acenapthene
plb/
hr
pg/
bhp­
hr
Acenaphthylene
plb/
hr
vg/
bhp­
hr
Anthracene
plb/
hr
Benzo(
a)
anthracene
pg/
bhp­
hr
plb/
hr
Benzo(
b)
fluoranthene
pg/
bhp­
hr
plb/
hr
pg/
bhp­
hr
Benzo(
k)
fluoranthene
plb/
hr
Benzo(
g,
h,
i)
perylene
pg/
bhp­
hr
plb/
hr
pg/
bhp­
hr
Benzo(
a)
pyrene
plb/
hr
Chrysene
pg/
bhp­
hr
plb/
hr
Dibenz(
a,
h)
anthracene
pg/
bhp­
hr
plb/
hr
pg/
bhp­
hr
Fluoranthene
plb/
hr
vg/
bhp­
hr
Fluorene
plb/
hr
Indnnnf
1
7
%
cd)
pyrene
pg/
bhp­
hr
ulb/
hr
PAH
2
9/
l
I99
PAH
3
912199
Average
I
I
m
.
*­
w.
.
­
\
.
,
­
,
­
0.11
0.17
0.18
0.15
0.25
0.25
0.24
0.25
0.11
0.17
0.18
0.15
0.25
0.25
0.24
0.25
0.11
0.17
0.18
0.15
0.25
0.25
0.24
0.25
0.11
0.17
0.18
0.15
0.25
0.25
0.24
0.25
0.11
0.17
0.18
0.15
0.25
0.25
0.24
0.25
0.11
0.17
0.18
0.15
0.25
0.25
0.24
0.25
0.11
0.17
0.18
0.15
0.25
0.25
0.24
0.25
0.11
0.17
0.18
0.15
0.25
0.25
0.24
0.25
0.11
0.17
0.18
0.15
0.25
0.25
0.24
0.25
0.11
0.17
0.18
0.15
0.25
0.25
0.24
0.25
0.11
0.17
0.18
0.15
0.25
0.25
0.24
0.25
0.11
0.17
0.18
0.15
0.25
0.25
0.24
0.25
0.11
0.17
0.18
0.15
0.25
0.25
0.24
0.25
pg/
bhp­
h
ulb/
hr
0.11
0.17
0.18
0.15
0.25
I
0.25
I
0.24
I
0.25
Phenanthrene
Pyrene
pg/
bhp­
hr
0.11
0.17
0.18
0.15
plb/
hr
0.25
0.25
0.24
0.25
WW­
hr
0.11
0.17
0.18
0.15
plb/
hr
0.25
0.25
0.24
0.25
NOTE:
The
reported
analytical
detection
limit
for
each
compound
was
0.1
pg
per
run.

Final
Report
­
Caterpillar
3508
EUI
6­
28
September
200
I
6.5
DATA
QUALITY
ASSESSMENT
EPA
used
the
Data
Quality
Objective
(
DQO)
Process
to
plan
the
test
program.
The
DQO
Process
consists
of
seven
distinct
steps.

1.
State
the
problem.
2.
Identify
the
decision.
3.
Define
inputs
to
the
decision.
4.
Define
the
study
boundaries.
5.
Develop
the
decision
rule.
6.
Specify
tolerable
limits
on
decision
errors
7.
Optimize
the
design
for
obtaining
data.

The
DQO
outputs
for
this
test
program
were
presented
in
the
Quality
Assurance
Project
Plan.
The
problem
was
defined
in
the
QAPP
and
is
restated
below.

EPA
believes
that
there
is
a
need
to
conduct
emission
tests
on
a
subset
of
engines
of
differing
designs
to
evaluate
the
following
issues:

0
the
effectiveness
of
after­
combustion
control
systems
on
HAP
emissions,
and
0
the
effectiveness
of
combustion
modifications
(
engine
operating
parameters)
on
HAP
emissions.

EPA
then
developed
a
decision
statement.
The
decision
statement
defined
the
process
that
would
be
used
to
answer
the
stated
problem.
The
decision
statement
is
restated
below:

IfEPA
can
identify
a
range
of
engine
operating
conditions
for
a
defined
set
of
engines
with
specified
after­
combustion
treatment
systems
and
a
list
ofpollutants
of
interest,
and
EPA
collects
data
to
determine
emissions
of
those
pollutants
for
each
engine
operated
at
each
engine
operating
condition,
then
EPA
can
make
a
determination
of
the
control
effectiveness
of
after­
combustion
and
combustion
modifications.
In
addition,
EPA
can
obtain
information
on
HAP
emissions
throughout
the
engine
operating
range.

PES,
EECL,
and
EM1
conducted
the
test
program
on
the
Caterpillar
3508
EM1
diesel
cycle,
oil­
fired,
4­
stroke,
reciprocating
internal
combustion
engine.
The
Engelhard
oxidation
catalyst
was
designed
to
provide
the
information
required
by
the
decision
statement.
Based
upon
the
inputs,
EPA
will
make
decisions
that
will
be
used
to
regulate
this
engine
subcategory.
Inputs
to
the
decision
were
defined,
agreed
to,
and
documented
in
the
QAPP.
These
inputs
consisted
of
agreement
on
a
finite
list
of
engines
to
test,
the
after­
combustion
control
systems
to
test,
the
range
of
engine
operating
conditions,
the
catalyst
conditioning
process,
the
target
list
of
pollutants,
and
the
sampling
and
analysis
methods,
and
sample
durations.

Final
Report
­
Caterpillar
3508
EUI
6­
29
September
200
1
During
conduct
of
the
test
program,
there
were
deviations
from
the
QAPP.
Deviations
to
the
QAPP
have
been
discussed
in
Section
3.0
for
deviations
in
engine
operation,
and
Section
5.0
for
deviations
in
Sampling
and
Analysis
procedures.

Table
6.16
presents
a
summary
of
engine
and
sample
method
performance
compared
to
the
QAPP
requirements.
Outlier
and
data
validation
issues
have
been
discussed
in
previous
sections.
Based
upon
the
engine
and
method
performance,
the
data
quality
is
evaluated
on
a
run­
by­
run
basis
for
suitability
in
the
assessment
of
pollutant
emissions
and
destruction
efficiency
of
HAP
by
the
catalyst.

Five
engine
parameters
were
varied
during
the
test
program.
The
parameters
were
changed
so
that
emissions
data
and
HAP
destruction
efficiency
could
be
evaluated
at
a
variety
of
engine
operating
conditions.
These
conditions
are
expected
to
simulate
the
range
of
engine
operating
conditions
in
industry.
Table
6.16
identifies
the
number
of
engine
parameters
that
were
within
the
tolerances
prescribed
in
the
QAPP.
The
target
engine
operating
conditions
were
estimates
based
upon
manufacturer's
recommendations.
There
are
differences
between
these
recommendations
and
the
nominal
engine
operating
parameters
of
the
Caterpillar
engine
located
at
the
EECL.
When
testing
was
conducted,
some
of
the
prescribed
engine
parameters
could
not
be
met.
The
fact
that
a
pre­
set
engine
parameter
could
not
be
met
is
considered
to
be
minor.
The
testing
was
conducted
over
a
range
of
engine
operating
conditions,
and
these
operating
conditions
are
documented.

The
remainder
of
the
table
assesses
data
quality
using
a
three­
tiered
system.
A
(
J
+)
indicates
that
all
method
performance
parameters
defined
in
the
QAPP
and/
or
the
sampling
method
were
met.
A
(
J>
indicates
that
at
least
90%
of
the
method
performance
parameters
were
met.
The
QAPP
specified
no
detection
limits
for
FTIRS
or
GCMS.
The
calculated
detection
limits
are
reasonable
for
this
test
program.
A
(
J
­)
indicates
that
fewer
than
90
%
of
the
method
performance
parameters
were
met.

Final
Report
­
Caterpillar
3508
EUI
6­
30
September
200
1
TABLE
6.16
SUMMARY
OF
ENGINE
AND
METHOD
PERFORMANCE
Run
ID
1
2
3
4
9
10
11
12
13
14
PAH
1
PAH
2
PAH
3
Engine
Parameters
Met
516
516
516
516
6J6
516
516
616
616
516
516
5f6
516
Catalyst
Inlet
FTIR
QA
Requirements
J+
J+
J+
J+
J+
J+
I/+
J+
J+
J+
J+
J+
J+

FTIR
Detection
Limits
a
J
J
J
J
J
J
J
J
J
J
J
J
J
CEMSQARequirements
J+
J+
J+
J+
J+
J+
J+
J+
J+
J+
J+
J+
J+

CEMS
Detection
Limits
'
J
J
J
J
J
J
J
J
J
J
J
J
J
GCMSQARequirements
J­
J­
J­
J
J­
J­
J­
J­
J­
J­
J­
J­
No
Data
GCMSDetectionLimits
J+
J+
J+
J+
J+
J+
J+
J+
J+
J+
J+
J+
No
Data
PAH
QA
Requirements
­
­
­
­
­
­
­
­
­
­
J­
J+
J+

PAH
Detection
Limits
J­
J+
J+

Catalyst
Outlet
FTIR
QA
Requirements
J
J
J
J
J
J
J
J
J
J
J
J
J
FTIR
Detection
Limits
*
J
J
J
J
J
J
J
J
J
J
J
J
J
CEMSQARequirements
J+
J+
J+
J+
J+
J+
J+
J+
J+
J+
J+
J+
J+

CEMS
Detection
Limits
'
J
J
J
J
J
J
J
J
J
J
J
J
J
GCMS
QA
Requirements
J
J
J
J
J
J
J
J
J
J
J­
J+
No
Data
GCMS
Detection
Limits
J+
J+
J+
J+
J+
J+
J+
J+
J+
J+
J+
J+
No
Data
I
PAHQARequirements
­
­
­
­
­
­
­
­
­
­
J+
J+
J+

PAH
Detection
Limits
J+
J+
J+

Assesement
of
Data
Quality
Catalyst
Inlet
Mass
Flow
J
J
J
J
J
J
J
J
J
J
J
J
J+

Catalyst
Outlet
Mass
Flow
/
J
J
J
J
J
J
J
J
J
J
J+
J+

HAP
Destruction
Efficiency
J
J
J
J
J
J
J
J
J
J
J
J
J
l
Neither
FTIRS
nor
CEMS
detection
limits
were
specified
in
the
QAPP.

Final
Report
­
Caterpillar
3508
EUI
6­
31
September
200
1
APPENDIX
A
SUBCONTRACTOR
TEST
REPORT
COLORADO
STATE
UNIVERSITY
ENGINES
AND
ENERGY
CONVERSION
LABORATORY
EMISSIONS
TESTING
OF
CONTROL
DEVICES
FOR
RECIPROCATING
INTERNAL
COMBUSTION
ENGINES
IN
SUPPORT
OF
REGULATORY
DEVELOPMENT
BY
THE
U.
S.
ENVIRONMENTAL
PROTECTION
AGENCY
(
EPA)

PHASE
3:
FOUR­
STROKIE,
DIESEL
INTERNAL
COMBUSTION
ENGINES
COLORADOSTATEUNIVERSITY
EMISSIONS
TESTING
OF
CONTROL
DEVICES
FOR
RECIPROCATING
INTERNAL
COMBUSTION
ENGINES
IN
SUPPORT
OF
REGULATORY
DEVELOPMENT
BY
THE
U.
S.
ENVIRONMENTAL
PROTECTION
AGENCY
(
EPA)

PHASE
3:
FOUR­
STROKE,
DIESEL
INTERNAL
COMBUSTION
ENGINES
Prepared
for:

PACIFIC
ENVIRONMENTAL
SERVICES
L
Submitted
by:

Engines
&
Energy
Conversion
Laboratory
Department
of
Mechanical
Engineering
Colorado
State
University
May
24,200O
Statement
of
Confidentiality
This
report
has
been
submittedfor
the
sole
and
exclusive
use
o/
Pacific
Environmental
Services,
and
shall
not
be
disclosed
or
provided
to
any
other
entity,
corporation,
or
third
part
for
purposes
beyond
the
specific
scope
or
intent
of
this
document
without
the
express
written
consent
of
Colorado
State
University.
TABLE
OF
CONTENTS
1.0
2.0
3.0
4.0
INTRODUCTION
1.1
Overview
1.2
Background
TEST
PROGRAM
2.1
Objective
2.2
Incentives
2.3
Work
Plan
DEVIATIONS
TO
TEST
PROGRAM
3.1
FTIR
Validation
3.2
FTIR
Post
Catalyst
Water
Analysis
3.3
Baseline
Engine
Operating
Conditions
3.4
Four­
Stroke
Engine
Test
Matrix
TEST
SAMPLING
PROCEDURES
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.9
4.10
General
Test
Procedures
Test
Specifics­
Data
Collection
Test
Specifics­
Engine
Stability
Test
Specifics­
Data
Collection
Hardware
Test
Specifics­
Data
Collection
Process
Test
Specifics­
Emissions
Analyzer
General
Test
Procedures
Test
Specifics­
Emissions
Analyzer
Checks
and
Calibrations
Test
Specifics­
FTIR
Calibration
Procedures
Test
Specifics­
FTIR
Validation
Procedures
Test
Specifics­
General
Calibration
Statement
of
Confidentiality
l­
l
l­
2
2­
1
2­
1
2­
2
3­
1
3­
2
3­
3
3­
3
4­
l
4­
l
4­
2
4­
5
4­
5
4­
6
4­
11
4­
15
4­
20
4­
22
This
report
has
been
submitted
for
the
sole
and
exclusive
use
of
Pacific
Environmental
Services,
and
shall
not
be
disclosed
or
provided
to
any
other
entity,
corporation,
or
third
part
for
purposes
beyond
the
specific
scope
or
intent
of
this
document
without
the
express
written
consent
of
Colorado
State
University.
_
`
LII.
I1"*
III*
UI~","
l"
lll*"
llil
I
_
­
w.
nl*
mmlll
COLORADO
STATE
UNIVERSITY
APPENDIX
Appendix
A
Appendix
B
Appendix
C
Appendix
D
Appendix
E
Appendix
F
Appendix
G
Appendix
H
Appendix
I
Engine
Test
Data
Daily
Baseline
Data
Points
Test
Point
QC
Checks
Test
Points
Reference
Method
Analyzers
Calibrations
FTIR
Calibration
FTIR
Validation
Calibration
Gas
Certification
Sheets
Baseline
Methane/
Non­
Methane
Analyzer
Appendix
J
Pressure
and
Temperature
Calibrations
Appendix
K
Equipment
Certification
Sheets
Appendix
L
Dynamometer
Calibration
Appendix
M
Dynamometer
Calibration
Procedure
Appendix
N
Fuel
Analysis
Appendix
0
Fuel
Analysis
Calculations
­
Fuel
Specific
F
Factor
Appendix
P
Computing
Air/
Fuel
Ratio
from
Exhaust
Composition
Appendix
Q
"
Reciprocating
Internal
Combustion
Engines
­
Exhaust
Emission
Measurement"
Appendix
R
Annubar
Flow
Calculations
Appendix
S
Additional
Calculations
Appendix
T
Exhaust
Piping
Schematic
Appendix
U
Catalyst
Schematic
and
Information
Appendix
V
Diesel
Load
Cell
Calibration
Statement
of
Confidentiality
This
report
has
been
submitted
for
the
sole
and
exclusive
use
of
Pacific
Environmental
Services,
and
shall
not
be
disclosed
or
provided
to
any
other
entity,
corporation,
or
third
part
for
purposes
beyond
the
specific
scope
or
intent
of
this
document
without
the
express
written
consent
of
Colorado
State
University.
COLORADO
STATE
UNIVERSITY
1
.
O
INTRODUCTION
1.1
OVERVIEW
Natural
gas
fueled
and
diesel
fueled
reciprocating
engines
represent
a
large
portion
of
the
horsepower
in
operation
within
the
oil
and
gas
industry
and
power
generation
markets.
Criteria
pollutants
and
Hazardous
Air
Pollutants
(
HAPS)
issues
are
of
major
concern
for
both
two­
stroke
and
four­
stroke
engine
operators.
Current
Environmental
Protection
Agency
(
EPA)
and
natural
gas
industry
funded
test
programs
are
directed
toward
evaluating
emission
levels
from
existing
engines,

determining
formation
mechanisms
for
the
exhaust
gas
constituents
of
interest,
and
developing
new
technologies
to
reduce
the
emissions
levels
of
these
constituents.
The
investigation
of
the
application
of
commercially
available
techniques
designed
to
address
the
HAPS
emissions
from
reciprocating
internal
combustion
engines
(
RICES)
will
allow
the
EPA
to
quantify
the
effectiveness
of
current
commercially
available
control
devices.
These
devices
have
been
identified
as
having
the
potential
*

to
reduce
HAPS
emissions
from
stationary
RICE
sources.
Information
gained
through
this
program
will
assist
the
EPA
in
the
regulatory
development
effort.

Accurate
information
on
emission
levels
from
operational
facilities
is
difficult
to
obtain.
Based
upon
a
recommendation
from
the
Internal
Combustion
Coordinating
Rulemaking
Committee
(
ICCR)
to
the
EPA,
testing
is
being
conducted
on
industrial
class
engines
at
the
Industrial
Engine
Test
Facility
operated
by
Colorado
State
University.
Testing
is
being
conducted
on
both
two­
stroke
and
four­

stroke,
natural
gas
and
diesel
fueled
industrial
class
engines.
The
test
program
for
four­
stroke,
diesel
fueled
internal
combustion
engine
has
been
performed
during
Phase
Three
of
this
test
program.
The
results
of
Phase
Three
testing
are
contained
within
this
document.

1.2
BACKGROUND
The
1990
Amendments
to
the
Clean
Air
Act
include
provisions
that
significantly
impact
the
operation
of
stationary
reciprocating
internal
combustion
engines.
Of
the
ten
titles
to
these
amendments,
four
have
direct
bearing.
They
are
as
follows:

Title
I
­
Attainment
of
Air
Quality
Standards
Defines
ambient
air
quality
standards,
defines
non­
attainment
areas
based,
imposes
emissions
reductions
to
achieve
attainment
per
specified
timeline
per
reasonably
available
control
technology
(
IWCT).

Emissions
Testing
l­
1
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADO
STATE
UNIVERSITY
Title
III
­
Hazardous
Air
Pollutants
Defines
189
pollutants
classified
as
hazardous
air
pollutants
(
HAPS),
specifies
thresholds
in
tons
per
year
(
TPY)
for
any
one
of
these
pollutants
or
a
combination
of
these
compounds,
introduces
maximum
achievable
control
technology
(
MACT)
for
sources
triggering
thresholds.

Title
V
­
Operating
Permits
I
Imposes
requirement
to
obtain
federal
operating
permits
for
major
sources,
imposes
requirement
to
provide
annual
certification
of
compliance,
defines
emissions
fees
based
on
actual
emissions.

Title
VII
­
Enforcement
Establish
mechanisms
to
enhance
and
strengthen
enforcement
of
CAA,
establishes
criminal
penalties,
gives
authority
to
issue
administrative
orders
(
fines
/
penalties)
without
going
to
federal
court
for
certain
violations.

Because
of
the
significant
economic
and
operational
impacts
of
the
CAAA
and
subsequent
rulemakings
by
the
EPA
and
state
agencies,
reciprocating
internal
combustion
engine
research
efforts
are
focused
on
reduction
and
monitoring
of
emissions
from
these
sources.
Specifically,
much
of
the
work
performed
to
date
has
focused
on
the
reduction
of
NO,
emissions.
These
efforts
have
developed
control
strategies
for
NO,
reductions
by
either
altering
the
combustion
process
or
by
means
of
exhaust
gas
after­
treatment.
Currently,
none
of
these
strategies
focus
on
the
formation
/

reduction
of
air
toxins.

The
EPA
in
conjunction
with
the
RICE
Work
Group
of
the
ICCR
process
has
determined
that
additional
HAPS
emissions
data
is
necessary
to
support
the
regulatory
development
process.
In
a
RICE
Emissions
Test
Plan
Document
dated
November
1997,
a
five
component
test
plan
to
acquire
additional
HAPS
emissions
test
data
was
set
forth.
The
five
components
include
the
following:

Engines,
Fuels,
and
Emissions
Controls
to
be
tested
Matrix
of
Operating
Conditions
to
be
tested
Pollutants
to
be
Measured
During
Testing
Test
Methods
to
Quantify
Emissions
Prioritization
Ten
HAPS
pollutants
are
included
in
the
test
plan
for
diesel
engines.
These
compounds
are:

formaldehyde,
acetaldehyde,
acrolein,
the
BTEX
compounds
(
benzene,
toluene,
ethylbenzene,

xylene),
naphthalene,
l­
3
butadiene,
and
naphthalene.

Insight
gained
through
the
test
program
will
provide
information
on
the
engine
operating
conditions
that
affect
the
formation
/
reduction
mechanisms
of
HAPS.
The
investigation
of
the
application
of
Emissions
Testing
l­
2
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
lntemal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADOSTATEVNIVERSITY
commercially
available
techniques
designed
to
address
the
HAPS
emissions
from
RICES
will
allow
the
EPA
to
quantify
the
effectiveness
of
current
commercially
available
control
devices.
These
devices
have
been
identified
as
having
the
potential
to
reduce
HAPS
emissions
from
stationary
RICE
sources.
Information
gained
through
this
program
will
assist
the
EPA
in
the
regulatory
development
effort.

Emissions
Testing
l­
3
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADO
STATE
UNIVERSITY
2.0
TEST
PROGRAM
2.1
OBJECTIVE
The
objective
of
this
program
is
to
evaluate
commercially
available
catalyst
technologies
which
have
been
identified
as
having
the
potential
to
control
both
formaldehyde
and
other
Hazardous
Air
Pollutants
(
HAPS)
as
well
as
existing
criteria
pollutants
from
reciprocating
internal
combustion
engines
(
RICE).
The
specific
internal
combustion
engine
class
tested
under
the
Phase
Three
test
program
was
the
four­
stroke,
diesel­
fueled,
internal
combustion
engine.
The
catalyst
hardware
was
evaluated
according
to
the
1
&
point
test
matrix
developed
by
the
EPA,
and
the
Reciprocating
Internal
Combustion
Engine
(
RICE)
Work
Group
of
the
ICCR
process.
Investigation
of
catalyst
performance
during
operation
at
various
engine
operating
conditions
provides
insight
into
the
effectiveness
of
catalysts
at
various
conditions.
The
information
gained
through
the
test
program
will
assist
the
EPA
in
regulatory
development
efforts
for
control
of
HAPS
emissions
and
criteria
pollutants
from
RICE
sources.

2.2
INCENTIVES
Title
III
of
the
1990
Clean
Air
Act
Amendments
requires
the
development
of
Maximum
Achievable
Control
Technology
(
MACT)
standards
for
major
sources
of
Hazardous
Air
Pollutants
(
HAPS)

emissions.
A
MACT
major
source
is
defined
as
one
that
emits
greater
than
10
tons
per
year
of
any
single
HAP
or
25
tons
per
year
for
all
HAPS.
For
most
source
categories
(
RICE
included),
the
MACT
standards
will
require
that
existing
major
sources
apply
HAPS
emissions
control
technologies
that
reduce
emissions
to
a
level
achieved
by
the
best
performing
existing
sources.
In
some
cases,

depending
upon
the
cost
of
the
control
technology
and
the
amount
and
toxicity
of
the
HAPS
removed,

more
stringent
standards
may
be
set.
The
MACT
standards
for
RICES
are
scheduled
to
be
promulgated
by
the
year
2000.

Of
the
HAPS
listed,
the
EPA
in
conjunction
with
the
Internal
Combustion
Coordinating
Rulemaking
Committee
(
ICCR)
have
identified
compounds
which
may
be
present
in
the
exhaust
of
reciprocating
internal
combustion
engines.
Existing
test
data
from
natural
gas
engines
indicates
that
the
only
HAP
compound
present
in
the
exhaust
of
RICES
at
levels
approaching
10
tons
per
year
is
formaldehyde.

Currently,
commercially
available
technologies
which
may
have
the
potential
ability
toward
reducing
HAPS
emissions
from
RICES
are
after­
treatment
technologies
(
catalyst).

Commercially
available
aftertreatment
technologies
(
catalysts)
for
the
control
of
organic
compound
emissions
are
currently
in
operation
on
RICES.
The
performance
of
these
technologies
for
control
of
volatile
organic
compounds
(
VOCs)
and
products
of
incomplete
combustion
has
been
documented.

However,
the
information
on
the
effectiveness
of
these
technologies
for
reducing
organic
HAPS
Emissions
Testing
2­
l
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADOSTATEUNIVERSITY
emissions
is
very
limited.
Determining
the
effectiveness
and
longevity
of
exhaust
catalysts
will
aid
the
EPA
in
evaluating
current
technologies
for
control
of
HAPS
emissions
from
RICE
sources
as
well
as
providing
information
in
support
of
regulatory
development
by
the
EPA
for
these
sources.

2.3
WORK
PLAN
Pacific
Environmental
Services
(
PES)
serves
as
the
prime
contractor
responsible
for
providing
information
to
the
EPA.
CSU
is
a
subcontractor
to
PES.
Testing
was
conducted
at
the
Colorado
State
University`
s
Engines
and
Energy
Conversion
Laboratory.
The
engine
and
catalyst
type
tested
is
described
in
Table
1.

TABLE
1
ENGINE
AND
CATALYST
TYPE
I
Number
of
Cylinders
I
8
II
II
Bore
and
Stroke
I
6.7"
X7.5"
li
II
Engine
Speed
I
1800
RPM
II
Catalyst
Classification
Manufacturer
Oxidation
Type
Engelhard
"

Element
Size
12"~
16"
x3
S"

Number
of
Elements
4
i
The
test
matrix
as
defined,
is
described
in
Table
2
with
engine
baseline
conditions
shown
in
Table
3.

Deviations
fi­
om
the
described
test
conditions
are
detailed
in
Section
3
of
this
report.
Each
test
point
consisted
of
collecting
thirty­
three
minutes
of
data.
The
raw
data
was
averaged
into
thirty­
three
one­

minute
data
points.
The
data
points
were
then
averaged
to
provide
the
results
for
the
single
test
point.
The
results
are
presented
in
tabular
form
in
Appendix
A
of
this
report.

Emissions
Testing
2­
2
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADOSTATE
UNIVERSITY
TABLE
2
ENGINE
OPERATING
CONDITIONS
DURING
TESTING
CATERPILLAR
3508
EUI
(
DIESEL­
FIRED)
US
EPA
ICCR
RICE
I­
LAP
EMISSION
TESTING
Operating
Speed
Torque
Fuel/
Air
Timing
Intercooler
Jacket
Conditions
to
@
pm)
(%
of
Equivalen
Air
Water
be
Tested:
baseline)
ce
Ratio
Temp.
Temp.

?
un
1
H
H
N
S
S
S
?
un
2
H
L
N
S
S
S
?
un
3
L
L
N
S
S
S
?
un
4
L
H
N
S
S
S
?
un
5
Operating
Condition
Not
Applicable
For
This
Engine
?
un
6
Run
7
Operating
Condition
Not
Applicable
For
This
Engine
Operating
Condition
Not
Applicable
For
This
Engine
Run
8
Run
9
Run
10
Run
11
Run
12
Run
13
Run
14
Run
15
Run
16
Operating
Condition
Not
Applicable
For
This
Engine
H
H
N
S
L
S
H
H
N
S
H
S
H
H
N
S
S
L
H
H
N
S
S
H
H
H
N
L
S
S
H
H
N
H
S
S
H
H
N
S
S
S
H
H
N
S
S
S
L=
1200
L
=
70
N
=
0.58
s
=
21
s=
130
S=
180
H
=
1800
H=
lOO
L
=
0.53
L=
19
L=
120
L=
155
H
=
0.74
H
=
23
H=
140
H
=
205
*
Defined
as:
ti
Fuel
/
Air,,,,,
=

Fue1
'
AirSmkhiometric
Emissions
Testing
2­
3
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
TABLE
3
CATERPILLAR
3508
BASELINE
CONDITIONS
Engine
Operating
Parameters
Nominal
Value
Acceptable
Range
Designation
Engine
Torque
2714
ft­
lb.
f
2%
of
value
Primary
Engine
Speed
Jacket
Water
Temperature
Outlet
1500
RPM
16O"
F­
210
°
F
+
2%
of
value
+
5%
of
value
Primary
Primary
Engine
Oil
Temperature
Outlet
Air
Manifold
Temperature
160
°
F
to
210
°
F
186
°
F
+
5%
of
value
*
5%
of
value
Primary
Primary
1
Air
Manifold
Pressure
(
AMP)
139"
Hg
above
Atm.
1
f.
5%
of
value
1
Primary
I
I
Exhaust
Manifold
Pressure
I
Varies
with
AMP.
I
&
5%
of
value
1
Primary
I
Ignition
Timing
21"
BTDC
+
5%
of
value
Primary
I
Overall
Air:
Fuel
Ratio
I
25:
l
I
f
5%
of
value
1
Primary
I
Inlet
Air
Humidity­
Absolute
~
I
.0015
lb
H,
O/
lb
Air
I
+
10%
of
value
1
Primary
I
38.8
Gal/
Hr,
272
Ib./
hr
k
5%
of
value
I
Primary
Engine
Oil
Pressure
Inlet
~
1
59­
70
lb.
I
If:
5%
of
value
1
Secondary
­
1
I­
Inlet
Air
Flow
I
1700­
I
800
SCFM
I
+_
5%
of
value
1
Secondary
1
I
Average
Engine
Exhaust
Temperature
I
1111
°
F
I
f
5%
of
value
I
Secondary
I
Note:
Based
on
Engine
Manufacturers
Specification
Data
sheets.

Emissions
Testing
2­
4
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADO
STATE
UNIVERSITY
3.0
DEVIATIONS
TO
TEST
PROGRAM
Testing
on
the
four­
stroke,
diesel
fueled
IC
engine
was
conducted
between
August
3
1,
1999
and
September
2,
1999.
Prior
to
initiation
of
the
12
point
test
matrix,
a
validation
procedure
was
performed
on
the
two
FTIR
analyzers.
The
analyzers
were
validated
for
formaldehyde,
acrolein,
and
acetaldehyde.
Modifications
to
the
baseline
engine
operating
conditions
were
made
prior
to
the
beginning
of
the
test
matrix.
The
variances
from
the
original
test
program
are
described
below:

3.1
FTIR
VALIDATION
A
validation
procedure
was
performed
on
the
two
FTIR
analyzers
on
August
26,
1999.
The
validation
procedure
was
conducted
in
basic
accordance
with
procedures
outlined
in
EPA
Method
301
­"
Field
Validation
of
Pollutant
Measurement
Methods
from
Various
Waste
Media".
Validation
procedures
for
aldehydes
utilized
an
analyte
spiking
technique
as
specified
in
Method
301.

Validation
procedures
for
NO,,
CO,
and
moisture
were
not
performed.
The
validation
for
the
criteria
pollutants
will
use
the
data
collected
during
the
test
program
to
perform
the
validation
procedures.
Comparative
sampling
to
the
appropriate
EPA
reference
methods,
(
Method
7E
&
20,
Method
10,
and
Method
4,
respectively),
for
these
compounds
was
performed
by
comparing
FTIR
analyzer
data
to
reference
methods
data
generated
during
the
test
program.
Deviations
from
the
described
procedures
are
as
follows:

Analyte
Spiking:
The
validation
for
the
target
aldehyde
compounds
was
carried
out
by
means
of
dynamic
analyte
spiking
of
the
sample
gas.
The
sample
stream
of
the
exhaust
gas
was
spiked
with
all
of
the
specific
analytes
simultaneously.
This
change
had
no
impact
on
the
test
procedure
or
results.

Formaldehyde:

Formaldehyde
spike
gas
was
generated
by
volatilization
of
a
formalin
solution
prepared
from
a
stock
formalin
solution
of
37%
formaldehyde
by
weight.
The
solution
was
vaporized
by
means
of
a
heated
vaporization
block.
The
vaporized
formalin
solution
then
mixed
with
a
carrier
gas
and
flowed
into
the
sample
exhaust
stream.
Carrier
gas
flow
rate
was
measured
by
a
mass
flow
meter
equipped
with
readout.
The
carrier
gas
was
to
be
Nitrogen;
however,
since
it
was
determined
to
perform
the
validation
process
for
all
aldehyde
compounds
simultaneously,
the
Acetaldehyde/
Acrolein
blend
calibration
gas
was
used
as
a
carrier
gas
for
the
vaporized
values
were
not
adjusted
for
the
statistical
bias.
Formaldehyde.
The
final
emissions
Acetlyaldehyde/
Acrolein
Emissions
Testing
3­
1
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
Acetlyaldehyde
and
acrolein
spiked
samples
were
generated
from
a
certified
gas
standard
(
Scott
Specialty
Gases,
+
2%
analytical
accuracy)
which
contain
both
analyte
species
and
a
sulfur
hexaflouride
(
SF6)
tracer
gas.
The
gas
flow
rate
was
measured
by
a
mass
flow
meter
equipped
with
readout.
The
validation
of
Acetaldehyde/
Acrolein
was
conducted
in
conjunction
with
the
Formaldehyde
validation.
The
final
emissions
values
were
not
adjusted
for
the
statistical
bias.

3.2
FTIR
POST
CATALYST
WATER
ANALYSIS
The
analytic
method
on
the
Nicolet
Magna
560
FTIR
analyzer
gave
water
measurements
that
were
excessively
high
for
post­
catalyst
emissions
measurements.
The
spectra
for
H20,
provided
by
Nicolet,
on
the
Magna
560
calculated
water
content
to
be
approximately
6%
higher
than
actual
exhaust
gas
concentrations.
Carbon
balance
calculations
for
each
one­
minute
data
point
agreed
with
the
HZ0
readings
from
the
Rega
7000
FTIR
pre­
catalyst
emissions
measurement
at
all
test
conditions.

The
measurements
agreed
within
+
0.5%
to
+
l%
water
content.
The
carbon
balance
calculations
for
the
post
catalyst
water
content
agreed
with
the
pre­
catalyst
measurements
within
+
0.5%
to
+
l%
water
content
at
all
test
conditions.
The
carbon
balance
calculations
are
based
upon
the
pre­
catalyst
and
post­
catalyst
reference
method
analyzers.
Since
the
pre­
catalyst
and
post­
catalyst
measurements
were
made
with
separate
analyzers,
the
variability
in
the
Hz0
calculation
could
be
caused
by
variability
in
emissions
analyzers.

Water
content
in
the
exhaust
is
dependent
upon
the
actual
combustion
process
within
the
engine's
combustion
chambers.
Since
water
is
one
of
the
major
products
of
combustion,
as
the
combustion
process
varies,
so
will
the
water
content
in
the
exhaust.
Changes
in
engine
operating
parameters
over
the
sixteen­
point
test
matrix
caused
changes
in
the
products
of
combustion,
water
being
one
of
these
products.
As
the
actual
combustion
process
was
being
modified
based
on
the
varying
engine
operating
conditions
at
each
test
point,
the
water
content
in
the
exhaust
changed
with
these
variations.

The
changes
in
the
water
content
were
calculated
by
the
carbon
balance
method
and
detected
by
the
FTIR
analyzer.
Based
on
the
agreement
between
the
pre­
catalyst
FTIR
measurements
and
the
carbon
balance
calculation
for
water
content,
at
every
test
condition,
and
between
the
pre­
catalyst
and
post­

catalyst
calculations,
the
water
content
from
the
pre­
catalyst
FTTR
measurements
were
used
to
convert
the
wet
FTIR
measurements
to
dry
measurements.
As
both
FTIR
analyzers
passed
the
validation
process
and
passed
all
QC
checks,
the
variation
in
water
readings
fi­
om
the
Nicolet
Magna
560
analyzer
has
no
impact
on
the
results
of
the
testing
conducted
during
Phase
Three
of
the
overall
test
program.

Emissions
Testing
3­
2
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORALIO
STATE
UNIVERSITY
3.3
BASELINE
ENGINE
OPERATING
CONDITIONS
Baseline
engine
operating
conditions
as
described
in
the
Scope
of
Work
are
presented
in
Table
3
of
this
report.
These
conditions
were
estimates.
After
running
the
engines,
these
values
were
found
to
be
inaccurate.
Deviations
from
the
Baseline
engine
operating
conditions
as
presented
are
as
follows:

TABLE
4
CATERPILLAR
3508
BASELINE
CONDITIONS
Engine
Operating
Parameters
Nominal
Value
Acceptable
I
Designation
Range
I
Engine
Torque
I
2880
ft­
lb.
I
k
2%
of
value
1
Primary
I
I
Engine
Speed
I
1800
RPM
I
k
2%
of
value
1
Primary
I
1
Jacket
Water
Temperature
Outlet
I
196
°
F
I
+
5%
of
value
\
Primary
I
I
Engine
Oil
Temperature
Outlet
I
215
°
F
I
+
5%
of
value
1
Primary
­
1
I
Air
Manifold
TemDerature
I
150
°
F
I
+
5%
of
value
I
Primary
­
1
I
Air
Manifold
Pressure
(
AMP)
I
5"
Ha
above
Atm.
I
+
5%
of
value
I
Primary
I
I
Exhaust
Manifold
Pressure
I
Varies
with
AMP.
I
+_
5%
of
value
I
Primary
I
Injection
Timing
Overall
Air
Fuel
Ratio
21
"
BTDC
3O:
l
Jf:
5%
of
value
+
5%
of
value
Primary
Primary
I
Inlet
Air
Humidity­
Absolute
I
.
Ol5
lb
HzOllb
Air
I
+
lO%
of
value
\
Primary
I
1
Engine
Fuel
Flow
I
42
GallHr,
310
Ib./
hr
I
+_
5%
of
value
I
Primary
I
I
Engine
Oil
Pressure
Inlet
I
67
PSI
I
k
5%
of
value
I
Secondary
1
I
Inlet
Air
Flow
I
2150
SCFM
I
+
5%
of
value
I
Secondary
~
I
I
Average
Engine
Exhaust
Temperature
I
1000
°
F
I
+
5%
of
value
I
Secondary
I
Humidity
Ratio:

The
baseline
humidity
ratio
was
stated
as
0.0015lb.
H20/
lb.
air.
This
is
a
misprint.

Documentation
should
be
corrected
to
show
0.015lb.
H,
O/
lb.
air
as
baseline
humidity
ratio.

3.4
TWO­
STROKE
ENGINE
TEST
MATRIX
The
four­
stroke
engine
sixteen
point
test
matrix
and
associated
engine
operating
conditions
as
Emissions
Testing
3­
3
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
lntemal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADO
STATE
UNIVERSITY
described
in
the
Scope
of
Work
are
presented
in
Table
2
of
this
report.
During
testing
discrepancies
between
the
CSU
"
Scope
of
Work"
and
the
QAPP
in
relation
to
engine
operating
conditions
were
identified.
The
QAPP
referenced
engine­
operating
data
in
relation
to
field
engines
originally
proposed
in
the
ICCR
process
and
not
the
engines
at
Colorado
State
University.
Deviations
from
the
engine
operating
conditions
described
in
the
sixteen­
point
test
matrix
are
referenced
to
the
CSU
"
Scope
of
Work'.
Deviations
from
the
described
engine
operating
conditions
are
as
follows:

Global
Deviation
in
Engine
Operating
Conditions
Speed:

The
baseline
speed
condition
was
changed
to
1800
rpm
as
indicated.
This
value
was
used
for
the
high
speed
points.
The
value
used
for
the
low
speed
points
was
1600
r­
pm.

Intercooler
Air
Temperature:

Following
the
changes
to
the
baseline
conditions,
higher
temperatures
were
found
to
be
more
adequate.
As
a
result,
the
test
point
values
were
increased
to
S
=
150",
L
=
140",
and
H
=
160
°
F.

Jacket
Water
Temperature:

Following
the
changes
to
the
baseline
conditions,
higher
temperatures
were
found
to
be
more
adequate.
As
a
result,
the
test
point
values
were
increased
to
S
=
196*,
L
=
186*,
and
H
=
206
°
F.

Test
Point
Specific
Variances
Only
deviations,
which
were
not
previously
described
in
the
"
Global
Deviation"
section,
will
be
described.

Test
Point
1,
Test
Point
9,
Test
Point
10,
Test
Point
13,
and
Test
Point
14:

During
these
points,
the
Heated
GC
/
FID
monitors
were
being
repaired.
CSU,
PES,
and
EPA
agreed
to
continue
the
testing
without
the
monitors.
*

Emissions
Testing
3­
4
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADO
STATE
UNIVERSITY
4.0
TEST
SAMPLING
PROCEDURES
Engines
&
Energy
Conversion
Laboratory
Industrial
Engine
Test
Facility
Colorado
State
University
The
Industrial
Engine
Test
Facility
was
installed
at
the
Engines
&
Energy
Conversion
Laboratory
to
facilitate
research
on
environmental
and
technological
issues
related
to
industrial
class
engines.

4.1
GENERAL
TEST
PROCEDURES
Procedures
have
been
established
to
ensure
accurate
and
repeatable
results.
Testing
criteria
established
for
the
test
facility
ensures
that
the
data
collected
has
a
high
degree
of
accuracy
and
can
be
repeated
if
warranted.
However,
since
the
Industrial
Engine
Test
Facility
was
designed
to
allow
for
several
different
industrial
engine
types
to
be
tested
in
a
laboratory
environment,

testing
procedures
differ
somewhat
from
field
test
procedures
and
are
unique
to
this
facility.
The
sampling
procedure
and
calibration
procedures
are
described
under
their
respective
sections
of
the
TEST
SPECIFICS
portion
of
this
report.

4.2
TEST
SPECIFICS
­
DATA
COLLECTION
The
data
collection
process
has
been
standardized
to
afford
accurate
and
repeatable
results
throughout
a
test
program.
The
high
degree
of
accuracy
which
can
be
obtained
at
the
Industrial
Engine
Test
Facility
is
due
to
the
sophisticated
level
of
instrumentation
utilized
at
the
facility.

To
ensure
accurate
and
repeatable
results,
a
specific
outline
of
the
data
collection
process
has
been
developed
for
the
Industrial
Engine
Test
Facility.

A
standard
data
point
collected
at
the
EECL
consists
of
engine
operating
data
being
gathered
over
either
a
three­
minute
or
five­
minute
period
and
averaged.
It
has
been
determined,
based
on
previous
tests,
that
3­
5
minutes
provides
an
acceptable
time
period
required
for
an
appropriate
data
set
to
be
collected
and
an
average
for
each
parameter
calculated.

For
the
work
conducted
under
this
test
program,
a
test
point
consisted
of
a
series
of
data
points
taken
in
succession
and
averaged.
The
data
was
gathered
in
l­
minute
averages
over
a
33­
minute
test
period.
Using
a
data
set
consisting
of
thirty­
three,
one­
minute
data
points
would
highlight
any
large
fluctuations
in
load
and
other
parameters
that
would
have
a
significant
effect
on
emissions
data.
No
fluctuations
in
data
occurred
during
any
test
points.
This
demonstrated
that
the
engine
was
operating
at
a
steady
condition
and
the
data
recorded
in
the
individual
data
points
was
repeatable.

Emissions
Testing
4­
1
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
Table
5
provides
information
on
the
nominal
number
of
samples
collected
under
each
data
point
/

test
run
scenario
for
the
LBET.

TABLE
5
SAMPLING
SPECIFICATIONS
'
Measured
Number
of
Samples
Collected
­
Parameters
1
Minute
30
Minute
Data
Point
Test
Run
Engine
30­
60
900
­
1800
Operation
Emissions
30
­
60
900­
H300­

I
CEMS
I
Emissions
45
­
50
1350
­
1500
FTIR
4.3
TEST
SPECIFICS
­
ENGINE
STABILITY
For
data
taken
during
testing
to
be
reliable,
the
engine
was
operated
in
a
state
of
equilibrium
at
each
test
point.
The
engine
control
system
allowed
for
engine
operation
data
to
be
monitored
so
that
engine
stability
could
be
easily
recognized.
The
stability
of
each
specific
engine's
operation
was
not
only
determined
on
a
point­
by­
point
basis,
but
also
on
a
daily
basis.
Since
combustion
parameters
for
each
engine
type
will
vary,
engine­
operating
parameters
were
used
to
determine
engine
stability.
Procedures
used
for
determining
acceptable
engine
stability
are
as
follows:

Engine
Stability:
Engine
Start
Up
Procedures
Prior
to
the
beginning
of
data
collection
each
day,
the
engine
was
warmed
up
until
thermal
equilibrium
state
was
achieved.
This
was
nominally
determined
when
the
engine
coolant
water
systems
and
lubricating
oil
reached
a
steady
state
temperature.
Once
steady
state
operation
was
achieved,
a
daily
"
baseline"
data
point
was
gathered.
The
length
of
time
required
to
obtain
steady
state
operation
was
highly
dependent
upon
the
ambient
temperature
and
the
temperature
of
the
engine
when
started.
Due
to
the
dependence
on
these
factors,
there
was
no
pre­
determined
warm­
up
time.

Engine
Stability:
Daily
Baseline
Data
Point
The
Scope
of
Work
for
the
project
required
that
a
specified
number
of
test
points
be
collected
on
the
engine.
The
data
collection
process
encompassed
multiple
days
of
testing.
To
ensure
that
the
Emissions
Testing
4­
2
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
.
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADOSTATE
UNIVERSITY
engine
was
operating
in
a
similar
manner
on
each
test
day,
a
set
of
engine
baseline
data
was
collected.
An
initial
set
of
engine
baseline
data
(
one
five­
minute
data
point)
was
collected
prior
to
the
first
data
point.
On
the
ensuing
test
days,
baseline
data
points
were
collected
to
verify
the
data
collection
for
that
day.
The
primary
engine
operating
parameters
of
the
data
point
must
compare
within
a
specified
acceptable
range
of
the
values
of
the
primary
engine
operating
parameters
on
the
original
"
baseline"
data
set
for
engine
stability
and
to
the
baseline
operating
conditions
specified
in
Table
4.
If
primary
engine
operating
parameters
did
not
compare
to
within
the
predetermined
range,
corrective
measures
were
taken
to
isolate
and
correct
the
cause
of
the
unacceptable
values
for
the
primary
engine
operating
parameters.
Both
CSU
and
PES
representatives
initialized
the
daily
"
baseline"
data
set.
All
baseline
data
points
were
acceptable
during
the
test
program.
The
primary/
secondary
engine
operating
parameters,
acceptable
ranges,
and
their
nominal
values
for
a
"
baseline"
data
set
are
presented
in
Table
3.

Engine
Stability:
Pre­
Data
Point
Test
Procedures
Prior
to
initiating
a
test
run,
a
pre­
test
run
data
point
was
gathered.
The
data
point
was
five­
minutes
in
length.
For
each
pre­
test
run
data
point,
the
average
value,
minimum
value,
maximum
value,
and
standard
deviation
were
obtained
for
all
engine
operation
and
emissions
parameters.
Primary
engine
operating
parameters
specified
at
a
test
condition
must
agree
with
the
test
condition
value
within
+/­
2%
to
+/­
lo%
of
the
requested
value
dependent
upon
the
engine
parameter.
The
relative
standard
deviations
of
the
primary
operating
variables
were
below
1
.
O%
for
engine
operating
parameters
and
below
3.0%
for
the
engine
emissions
parameters.
The
primary
engine
operating
parameters
and
their
nominal
values
for
a
"
pre­
test
run"
data
point
are
presented
below
in
Table
6.

If
primary
engine
operating
parameters
did
not
agree
with
the
requested
test
condition
values
within
the
predetermined
range,
corrective
measures
was
taken
to
isolate
and
correct
the
cause
of
the
unacceptable
values
for
the
primary
engine
operating
parameters.
All
pre­
test
run
data
points
were
acceptable
for
the
test
program.
Both
CSU
and
PES
representatives
initialized
each
"
pre­
test
run"
data
point.

Engine
Stability:
Test
Run
Stability
A
test
run
consisted
of
a
set
of
one­
minute
averaged
data
points
taken
consecutively
over
a
33­
minute
time
period.
For
each
data
point,
the
average
value
for
each
primary
engine
operating
parameter
must
compare
to
within
the
acceptable
range
of
the
specified
target
value
at
the
test
condition
for
engine
stability
and
the
data
collection
process
to
be
valid
for
the
specific
test
condition.
If
primary
engine
operating
parameters
did
not
compare
to
within
the
predetermined
range,
the
data
point
was
invalid,
and
corrective
measures
were
taken
to
isolate
and
correct
the
Emissions
Testing
4­
3
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
cause
of
the
unacceptable
values
for
the
primary
engine
operating
parameters.

Engine
stability
was
maintained
throughout
the
data
collection
process
for
each
test
run.
The
relative
standard
deviation
of
the
primary
operating
variables
was
below
1
.
O%
for
engine
operating
parameters
and
below
3.0%
for
the
engine
emissions
parameters
at
each
data
point.

Both
CSU
and
PES
representatives
initialized
each
data
point
of
a
test
run.
The
tabular
format
of
the
primary
engine
operating
parameters,
designation,
and
the
acceptance
criteria
is
presented
in
Table
6:

TABLE
6
TEST
POINT
­
ENGINE
STABILITY
Engine
Operating
Parameters
Acceptable
Range
Standard
Deviation
Designation
Engine
Torque
f
2%
of
value
Il.
0
Primary
Engine
Speed
f
5%
of
value
Il.
0
Primary
Jacket
Water
Temperature
Outlet
f
5%
of
value
<
1.0
Primary
Engine
Oil
Temperature
Outlet
f
5%
of
value
5
1.0
Primary
Air
Manifold
Temperature
f
5%
of
value
2
`
1.0
Primary
Air
Manifold
Pressure
sfr
5%
of
value
5
1.0
Primary
Exhaust
Manifold
Pressure
*
5%
of
value
5
1.0
Primary
Ignition
Timing
f
5%
of
value
5
1.0
Primary
Overall
Air/
Fuel
Ratio
f
5%
of
value
Il.
0
Primary
Inlet
Air
Humidity­
Absolute
f
10%
of
value
5
1.0
Primary
1
Engine
Fuel
Flow
SCFH
/
Gal./
Hr.
1
f
5%
of
value
I
Il.
0
I
Primary
1
I
Engine
Oil
Pressure
Inlet
I
f
5%
of
value
I
5
3.0
1
Secondary
1
I
Inlet
Air
Flow
I
f
5%
of
value
1
23.0
1
Secondary
1
I
Average
Engine
Exhaust
Temperature
1
f
5%
of
value
I
53.0
I
Secondary
I
NO,
Emissions
(
PPM)
I
f
5%
of
value
1
5
3.0
1
Secondary
\

I
CO
Emissions
(
PPM)
I
f
5%
of
value
I
23.0
1
Secondary
I
I
THC
Emissions
(
PPM)
I
f
5%
of
value
I
5
3.0
1
Secondary
I
I
co2
(%)
I
f
5%
of
value
I
223.0
I
Secondary
I
I
02
(%)
I
f
5%
of
value
I
I3.0
1
Secondary
I
I
Exhaust
Air
Flow
I
f
5%
of
value
I
5
3.0
I
Secondary
I
Emissions
Testing
4­
4
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADOSTATEUNIVERSITY
4.4
TEST
SPECIFICS
­
DATA
COLLECTION
HARDWARE
The
design
of
the
test
facility
provides
a
platform
for
accurate
and
versatile
performance
and
emission
research
on
industrial
engines.
Control
and
measurement
systems
installed
on
the
Industrial
Engine
Test­
Beds
are
as
follows:

Caterpillar
3508
EUI:
Four­
Stroke,
Diesel
Fueled
Analyzer
Rack
for
NO,,
CO,
C02,
02,
&

Nicolet
Rega
7000
Fourier
Transform
Servomex
CO;!
&
02
Nicolet
Magna
560
Fourier
Transform
Infrared
(
FTIR)
Spectrometer
for
aldehydes
and
speciated
hydrocarbons
4.5
TEST
SPECIFICS
­
DATA
COLLECTION
PROCESS
The
data
collection
process
consisted
of
acquiring
information
from
the
various
control
and
monitoring
systems.
The
engine
control
and
monitoring
system
(
ECMS)
collected
all
engine
operating
and
emissions
parameters
(
criteria
pollutants
only).
All
engine
operating
parameters
were
direct
measurements
of
the
ECMS,
while
emissions
parameters
(
criteria
pollutants)
were
passed
by
communication
link
from
a
computer
dedicated
to
emissions
hardware
control
and
monitoring.
All
emissions
parameters
measured
with
an
FTIR
were
collected
and
stored
on
a
computer
dedicated
to
individual
FTIR
operation.

Emissions
Testing
4­
5
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADOSTATEVNIVERSITY
After
engine
stability
had
been
confirmed,
the
data
collection
process
for
a
test
run
condition
commenced.
The
data
collection
process
was
performed
as
follows:

Data
Collection
Process:

1.)
Verification
of
engine
stability
confirmed,
accepted,
and
initialized
by
PES
and
CSU
representatives.
2.)
Proper
file
names
are
assigned
to
all
data
acquisition
hardware.
3.)
Commence
acquisition
of
data
point
for
specified
test
condition
4.)
At
completion
of
data
point,
electronic
files
are
saved
and
hard
copies
are
printed
out.
5.)
PES
and
CSU
representatives
initialize
hard
copies
verifying
acceptable
data
point.
6.)
Move
engine
operation
to
next
test
condition.

TEST
SPECIFICS
­
EMISSION
ANALYZER
GENERAL
TEST
PROCEDURE23
Introduction
The
following
general
test
procedures
and
calibration
checks
guaranteed
the
integrity
of
our
sampling
system
and
the
accuracy
of
our
data.
The
testing
was
conducted
in
basic
accordance
with
approved
Environmental
Protection
Agency
(
EPA)
test
methods
as
described
in
the
Code
of
Federal
Regulations,
Title
40,
Part
60,
Appendix
A.

General
Procedure
Exhaust
oxygen
and
oxides
of
nitrogen
concentrations
from
the
engine
were
determined
in
basic
compliance
with
EPA
Method
20,
"
Determination
of
Nitrogen
Oxides,
Sulfur
Dioxide,
and
Diluent
Emissions
From
Stationary
Gas
Turbines"
and
EPA
Method
7E,
"
Determination
of
Nitrogen
Oxides
Emissions
From
Stationary
Sources
(
Instrumental
Analyzer
Procedure)".
The
sampling
procedure
for
CO
concentrations
was
based
on
EPA
Method
10,
"
Determination
of
Carbon
Monoxide
Emissions
from
Stationary
Sources."
EPA
Method
25A,
"
Determination
of
Total
Gaseous
Organic
Concentration
Using
a
Flame
Ionization
Analyzer"
was
the
sample
procedure
used
to
determine
THC
emission
concentrations.
A
modified
EPA
Method
18A
was
used
for
the
sampling
procedures
for
Methane/
Non­
Methane
Analysis.
The
method
for
calculating
mass
emissions
levels
was
based
upon
an
EPA
Method
19
"
Determination
of
Sulfur
Dioxide
Removal
Efficiency
and
Particulate
Matter,
Sulfur
Dioxide,
and
nitrogen
Oxides
Emission
Rates"
calculation.
Mass
based
emissions
were
evaluated
using
EPA
Method
19
(
F­

Emissions
Testing
4­
6
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADOSTATEUNIVERSITY
factor
technique).
Calibration
and
test
procedures
are
detailed
under
their
respective
sections
of
the
TEST
SPECIFICS
portion
of
this
report.

Sampling
System
Dedicated
analyzers
were
used
to
determine
the
NO,,
CO,
THC,
CO2,
and
02
emissions
level
on
a
dry
basis
for
both
pre
and
post
catalyst
emissions.
Dedicated
analyzers
were
used
to
determine
the
Methane/
Non­
Methane
emissions
on
a
wet
basis
for
both
pre
and
post
catalyst
emissions.

FTIR
analyzers
were
used
to
determine
aldehyde
emissions
on
a
wet
basis
for
both
pre
and
post
catalyst
emissions.
Refer
to
Table
8
for
the
analyzers
and
the
methods
of
analysis.

Exhaust
gas
was
extracted
from
the
engine
exhaust
system
through
a
3/
8"
stainless
steel
multi­

point
probe.
Sample
points
were
located
in
accordance
with
procedures
described
in
EPA
Method
1.
Exhaust
gas
then
passed
through
a
heated
3­
way
sample
valve
and
glass
wool
filter
assembly.
The
sample
was
transported
via
heat­
traced
Teflon
sample
lines
and
a
heated
sample
distribution
manifold.
Sample
for
the
"
dry"
gas
analyzers
then
passed
through
a
4­
pass
minimum
contact
condenser
specifically
designed
to
dry
the
sample.
The
"
dry"
sample
then
entered
a
stainless
steel
sample
pump.
The
discharge
of
the
pump
passed
through
3/
8"
Teflon
tubing
to
a
Balston
Microfibre
coalescing
filter,
moisture
sensor,
and
then
to
the
sample
manifold.
The
sample
manifold
was
maintained
at
a
constant
pressure
by
means
of
a
pressure
bypass
regulator.
A
flowmeter,
placed
in
line
at
the
exhaust
of
each
analyzer,
monitored
exact
sample
flows.

Heated
sample
flow
for
all
"
wet"
measurement
analyzers
will
be
provided
by
means
of
a
heated
sample
distribution
manifold
prior
to
sample
gas
entering
the
"
dry"
gas
analyzer
platform.
Each
heated
analyzer
had
a
dedicated
sample
pump
and
heat
traced
line
from
the
main
sample
train
to
the
analyzer.

Emissions
Testing
4­
7
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
TABLE
1
CURRENT
INSTRUMENTATION
Post
Catalyst
Emissions
Manufacturer
and
Model
1
Parameters
Detection
Principle
(
Range
Rosemount
NGA­
2000
CLD
Analyzer
NO
or
NO,
Thermal
reduction
of
NO2
to
NO.
Chemiluminescent
reaction
NO
with
03.
Variable
to
10000
PPM
Rosemount
NGA­
2000
CO
NDIR
Analyzer
Rosemount
NGA­
2000
COZ
NDIR
Analyzer
Rosemount
NGA­
2000
THC
FID
Analyzer
NDIR
with
Gas
Filter
Variable
to
2000
Correlation
PPM
N­
DIR
Variable
to
20%

Flame
Ionization
Variable
to
10000
PPM
Rosemount
NGA­
2000
PMD
Analyzer
O2
Paramagnetic
Variable
to
100%
Questar
Baseline
1030H
Heated
GC
/
FID
C&
Non­
CH4
Gas
Chromatograph
Flame
Ionization
Variable
to
5000
PPM
Nicolet
Magna
560
Multiple
See
Attached
FTIR
analysis
utilizing
a
Medium
range
LR
source.
Variable
Sierra
BG­
1
Micro­
Gas
extraction
Mass­
flowrate
based
Variable
Dilution
Test
Stand
through
extraction:
constant
particulate
temperature
dilution
filter
Mettler
Toledo
AG425
Particulate
Precision
Digital
Scale
filter
mass
Precision
strain
measurement
.
Ol
mg
to
41g
.
l
mg
t0
210g
Emissions
Testing
4­
8
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
CoLoR4Do
STATE
~~
I~
IVERSITY
TABLE
7
(
continued)

CURRENT
INSTRUMENTATION
Pre
Catalyst
Emissions
reaction
NO
wi
temperature
dilution
Emissions
Testing
4­
9
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
TABLE
8
COMPONENTS
MEASURED
BY
NICOLET
FTIR
Component
Formula
H20
co
co2
NO
NO2
N20
NH3
NOx
CI­­
I4
C2H2
C2H4
C2H6
C3H6
H2C0
CH30H
C3H8
I­
C4H10
N­
W­
ho
CH3CH0
so2
THC
Component
Name
Water
Carbon
Monoxide
Carbon
Dioxide
Nitric
Oxide
Nitrogen
Dioxide
Nitrous
Oxide
Ammonia
Oxides
of
Nitrogen
Methane
Acetylene
Ethylene
Ethane
Propene
Formaldehyde
Methanol
Propane
Iso­
Butylene
Normal­
Butane
Acetaldehyde
Sulfur
Dioxide
Total
Hydrocarbons
Emissions
Testing
4­
10
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADOSTATEVNIVERSITY
4.1
TEST
SPECIFICS
­
EMISSION
ANALYZERS
CHECKS
AND
CALIBRATIONS
The
following
instrument
checks
and
calibrations
guaranteed
the
integrity
of
our
sampling
system
and
the
accuracy
of
our
data.

Analyzer
Calibration
Gases
Standard
calibration
gases
used
at
the
facility
are
Scott
Specialty
Gases
EPA
Protocol
Gas
Standard
calibration
gases
with
a
rt
1
.
O%
or
+
2.0%
accuracy.
For
this
program,
EPA
Protocol
1
calibration
gases
(
RATA
Class)
were
used.
Manufacturer
supplied
certification
sheets
were
available
during
the
testing
procedure
and
copies
of
the
current
inventory
of
gases,
which
were
used
for
calibration
and
integrity
checks
on
the
reference
method
and
FTIR
analyzers,
are
provided
within
this
document.

EPA
Protocol
1
gases
(
Rata
Class)
were
used
to
calibrate
the
reference
method
analyzers
and
FTIR
analyzers.
Formaldehyde
standards
with
a
concentration
range
between
5
­
10
PPM
were
obtained.
Acetylaldehyde/
acrolein
standards
were
also
acquired.
Any
calibration
standards
which
were
not
EPA
Protocol
1
gases,
were
the
highest
quality
standard
available.

Analyzer
Specifications
Vendor
instrument
data
concerning
interference
response
and
analyzer
specifications
were
available
during
the
test
program.
Information
supplied
by
the
manufacturer
on
the
factory
specification
sheets
will
be
furnished
if
requested.

Response
Time
Tests
(
Prior
to
initiation
of
engine
test
program)

Response
time
tests
were
performed
on
each
sample
system.
The
response
time
tests
were
performed
prior
to
the
FTLR
validation
process
for
each
sampling
system.
The
response
time
of
the
slowest
responding
analyzer
(
Questar
Baseline)
was
determined.
Response
time
tests
conducted
at
the
EECL
indicated
sampling
system
response
times
of
I:
15
minutes.
This
is
the
time
for
the
Rosemount
Oxygen
Analyzer
(
slowest
responding
analyzer
which
continuously
monitors)
to
stabilize
to
response
output
of
the
analyzer.
The
Questar
Baseline
Industries
CHJNon­
CH4
analyzers
have
a
minimum
cycle
time
of
4:
50
minutes.
The
overall
response
time
for
these
analyzers
when
their
cycle
is
started
1:
15
minutes
after
a
sample
source
change
is
5:
55
minutes.
When
the
CH4/
Non­
CH4
analyzer
cycle
time
was
initiated
at
a
sample
source
change,
the
overall
response
time
is
9:
05
minutes.
The
response
time
was
tested
to
assure
that
the
analyzers'
response
was
for
exhaust
gas
entering
the
sample
system
from
each
of
the
test
point
conditions.

Calibration
(
Daily)

Emissions
Testing
4­
11
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADOSTATE
UNIVERSITY
Zero
and
mid­
level
span
calibration
procedures
were
performed
on
the
reference
method
analyzer
prior
to
each
test
day.
Zero
and
span
drift
checks
were
performed
upon
completion
of
each
data
point
and
upon
completion
of
each
test
day.
This
procedure
is
referenced
as
ZSD
(
zero
and
span
drift
check)
in
the
CSU
"
Scope
of
Work".
A
zero
and
a
mid­
level
gas
was
introduced
individually
directly
to
the
back
of
the
analyzers
before
testing
for
carbon
monoxide,
carbon
dioxide,
oxygen,
total
hydrocarbons,
Methane/
Non­
Methane,
and
oxides
of
nitrogen.
The
analyzers'
output
response
was
set
to
the
appropriate
levels.
Each
analyzer's
stable
response
was
recorded.
From
this
data
a
linear
fit
was
developed
for
each
analyzer.
The
voltage
for
each
analyzer
were
recorded
and
used
in
the
following
formula:

Y=
MX+
B
Where:
B
=
Intercept
M
=
Slope
X=
Analyzer
or
transducer
voltage
Y
=
Engineering
Units
After
each
test
point
and
upon
completion
of
a
test
day,
calibration
checks
were
conducted
by
re­
introducing
the
zero
and
span
gases
directly
to
the
back
of
the
analyzers.
The
analyzers'
stabilized
responses
were
recorded.
No
adjustments
were
made
during
testing
or
during
the
final
calibration
check.
Initial
calibration
values
and
all
calibration
checks
were
recorded
for
each
analyzer
during
the
daily
test
program.

The
before
and
after
calibrations
checks
were
used
to
determine
a
zero
and
span
drift
for
each
test
point
for
the
CO,
CO*,
02,
THC,
CH4/
Non­
CH4,
and
NO,
analyzers.
The
zero
and
span
drift
checks
for
each
test
point
and
each
test
day
were
less
than
+
2.0%
of
the
span
value
(
specific
range
setting)
of
each
analyzer
used
during
the
daily
test
program.
The
calibration
data
sheets
are
presented
in
Appendix
E
of
this
document.

Linearity
Check
(
Prior
to
initiation
of
engine
test
program)

Prior
to
initiation
of
the
test
program,
analyzer
linearity
checks
were
performed.
This
procedure
is
referenced
as
ACE
(
analyzer
calibration
error
check)
in
the
CSU
"
Scope
of
Work".
The
oxygen,
carbon
monoxide,
total
hydrocarbon,
methane/
non­
methane
and
oxides
of
nitrogen
analyzers
were
"
zeroed"
using
either
zero
grade
nitrogen,
or
hydrocarbon
free
air.
The
analyzers
were
allowed
stabilize
and
their
output
recorded.
The
analyzers
were
then
"
spanned"
using
the
mid­
level
calibration
gases.
The
analyzers
were
allowed
to
stabilize,
and
their
output
recorded.
From
this
data
a
linear
fit
was
developed
for
each
analyzer.
The
voltage
for
each
analyzer
were
recorded
and
used
in
the
following
formula:

Emissions
Testing
4­
12
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADOSTATELNIVERSITY
Y=
MX+
B
Where:
B
=
Intercept
M
=
Slope
X=
Analyzer
or
transducer
voltage
Y
=
Engineering
Units
Using
the
linear
fit,
the
linear
response
of
the
analyzer
was
calculated.
Low
level
and
high
level
calibration
gases
were
individually
introduced
to
the
analyzers.
For
each
calibration
gas,
the
analyzers
were
allowed
to
stabilize
and
their
outputs
were
recorded.
Each
analyzers'
linearity
was
acceptable
if
the
predicted
values
of
a
linear
curve
determined
from
the
zero
and
mid­
level
calibration
gas
responses
agreed
with
the
actual
responses
of
the
low
level
and
high
level
calibration
gases
within
+_
2.0%
of
the
analyzer
span
value.
The
methane/
non­
methane
analyzers'

linearity
was
acceptable
as
the
predicted
values
agreed
with
the
actual
response
of
the
low
level
and
high
level
calibration
gases
within
~
50%
of
the
actual
calibration
gas
value.
This
procedure
was
performed
for
one
range
setting
for
each
analyzer.
The
Linearity
Check
data
sheets
are
presented
in
Appendix
E
of
this
document.

NO2
Converter
Check
(
Prior
to
initiation
of
engine
test
program)

Prior
to
initiation
of
the
test
program,
NO2
converter
checks
were
performed.
A
calibration
gas
mixture
of
known
concentrations
between
240
and
270
PPM
nitrogen
dioxide
(
NO2)
and
160
to
190
PPM
nitric
oxide
(
NO)
with
a
balance
of
nitrogen
was
used.
The
calibration
gas
mixture
was
be
introduced
to
the
oxides
of
nitrogen
(
NO,)
analyzer
until
a
stable
response
was
recorded.

The
converter
is
considered
acceptable
if
the
instrument
response
indicated
a
90
percent
or
greater
NO2
to
NO
conversion.
The
NO*
Converter
Check
data
sheets
are
presented
in
Appendix
E
of
this
document.

Sample
Line
Leak
Check
(
Prior
to
initiation
of
engine
test
program)

The
sample
lines
were
leak
checked
before
the
engine
test
program.
The
leak
check
procedure
was
performed
for
both
pre­
catalyst
and
post­
catalyst
sample
trains.
The
procedure
involved
closing
the
valve
on
the
inlet
to
the
sample
filter
located
just
downstream
of
the
exhaust
stack
probe.
With
the
sample
pump
operating,
a
vacuum
was
pulled
on
the
exhaust
sample
train.
Once
the
maximum
vacuum
was
reached,
the
valve
on
the
pressure
side
of
the
pump
was
closed,
thus
sealing
off
the
vacuum
section
of
the
sampling
system.
The
pump
was
turned
off
and
the
pressure
in
the
sample
system
was
monitored.
The
leak
test
was
acceptable
as
the
vacuum
gauge
reading
dropped
by
an
amount
less
than
1
inch
of
mercury
over
a
period
of
1
minute.
The
Sample
Line
Leak
Check
data
sheets
are
presented
in
Appendix
E
of
this
document.

Emissions
Testing
4­
13
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADOSTATEUNIVERSITY
Sample
Line
Integrity
Check
(
Daily)

A
sample
line
integrity
check
was
performed
prior
to
and
upon
completion
of
each
test
day.
This
procedure
is
referenced
as
SSB
(
Sampling
System
Bias
Check)
in
the
CSU
"
Scope
of
Work".
The
analyzer's
response
was
tested
by
first
introducing
the
mid
level
calibration
gas
directly
to
the
NO,
analyzer.
The
analyzer
was
allowed
to
stabilize
and
the
response
recorded.
The
same
mid
level
calibration
gas
was
then
introduced
to
the
analyzer
through
the
sampling
system.
The
calibration
gas
was
introduced
into
the
sample
line
at
the
stack,
upstream
of
the
inlet
sample
filter.
The
analyzer
was
allowed
to
stabilize
and
the
response
recorded.
The
analyzer
response
values
were
compared
and
the
percent
difference
did
not
to
exceed
+
5
%
of
the
analyzer
span
value
(
range
setting).

The
SSB
procedure
was
to
be
performed
for
both
the
NO,
and
methane/
non­
methane
analyzers.
It
was
determined
to
perform
the
integrity
check
for
the
NO,
analyzers
only.
The
SSB
procedure
was
performed
for
the
methane/
non­
methane
analyzers
prior
to
and
upon
completion
of
the
test
program.
The
Sample
Line
Integrity
Check
data
sheets
are
presented
in
Appendix
E
of
this
document.

Diesel
Fuel
Analysis
and
Flow
Measurement
Two
independent
laboratories,
Southern
Petroleum
Labs
in
Houston,
TX
and
BG
Products,
Inc.
in
Wichita,
KS
analyzed
the
diesel
fuel.
Both
laboratories
performed
ASTM
tests
D240,
D5291,
D4294,
D975,
and
D13
19
for
two
different
samples.
The
parameters
provided
by
these
tests
include
the
fuel
carbon,
hydrogen,
nitrogen,
and
sulfur
content
by
weight,
the
lower
heating
value,
cetane
index,
flash
point,
and
specific
gravity.
Southern
Petroleum
Labs
also
performed
a
GC
PIANO
test
on
the
diesel
fuel,
Method
GPA
2186,
which
provided
detailed
speciation
and
average
molecular
weight.
Appendix
N
­
Fuel
Analysis
provides
a
summary
of
the
fuel
analysis
results
as
well
as
the
analysis
reports
provided
by
Southern
Petroleum
Labs
and
BG
Products,

IIlC.

The
fuel
analysis
is
an
important
input
for
data
reduction,
in
particular
for
the
evaluation
of
brake
specific
emissions.
Emissions
analysis
is
based
on
the
technique
described
in
the
Code
of
Federal
Regulations,
Title
40,
Part
60,
Appendix
A,
Method
19
for
dry
combustion
product
measurements.
Our
data
analysis
program
is
based
on
the
use
of
gaseous
fuel.
Rather
than
change
the
program,
fuel
properties
are
expressed
relative
to
the
vapor
phase,
using
the
ideal
gas
law
and
the
fuel
molecular
weight
from
the
GC
PIANO
analysis.

Emissions
Testing
4­
14
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
CoLouDo
STATE
UNIVERSITY
Evaluation
of
diesel
fuel
flow
was
performed
by
measuring
the
change
in
weight
of
the
fuel
tank
over
time.
The
weight
of
the
fuel
tank
was
measured
using
an
Interface
load
cell,
Model
1210HQ­
SK­
B.
The
fuel
tank
was
suspended
from
the
load
cell,
which
was
suspended
from
an
overhead
beam.
The
response
of
the
load
cell
was
nonlinear,
necessitating
a
precise
calibration
of
the
load
cell.
A
calibration
report
was
generated
and
is
provided
in
Appendix­
V.
The
differential
weight
fuel
flow
measurement
technique
is
typically
used
for
measuring
diesel
fuel
flow
rates
because
a
significant
fraction
of
the
delivered
fuel
is
recirculated
back
to
the
fuel
tank.
Thus,
a
direct
fuel
flow
measurement
of
the
fuel
delivered
to
the
engine
would
not
indicate
the
amount
of
fuel
consumed
by
the
engine.
The
desired
quantity
is
the
fuel
flow
rate
delivered
to
the
engine
minus
the
recirculation
flow
rate.
This
is
the
quantity
that
is
evaluated
from
the
differential
weight
measurement.

4.8
TEST
SPECIFICS:
FTIR
CALIBRATION
PROCEDURES
Calibration
was
performed
on
the
FTIR
instrument
prior
to
each
phase
of
the
test
program
and
at
the
beginning
and
end
of
each
test
day.
The
calibration
procedures
described
within
this
document
are
consistent
with
procedures
found
in
the
following
documents:

"
Measurement
of
Select
Hazardous
Air
Pollutants,
Criteria
Pollutants,
and
Moisture
Using
Fourier
Transform
Infrared
(
FTIR)
Spectroscopy"
­­
Prepared
by
Radian
International
for
the
Gas
Research
Institute.

"
Protocol
for
Performing
Extractive
FTIR
Measurements
to
Characterize
Various
Gas
Industry
Sources
for
Air
Toxics"
­
Prepared
by
Radian
International
for
the
Gas
Research
Institute.

Both
documents
are
contained
with
the
Gas
Research
Institute
Report
Number
GRI­

95/
0271
entitled,
"
Fourier
Transform
Infrared
(
FTIR)
Method
Validation
at
a
Natural
Gas­
Fired
Internal
Combustion
Engine"
­
Prepared
by
Radian
lntemational
for
the
Gas
Research
Institute.

Instrument
Description
Dedicated
FTIR
analyzers
and
sampling
conditioning
systems
were
used
to
measure
pre­
catalyst
and
post­
catalyst
exhaust
emissions.
A
description
of
each
unit
is
presented
in
Table
9:

Emissions
Testing
4­
15
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
lntemal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADOSTATEUNIVERSITY
TABLE
9
FTIR
EQUIPMENT
DESCRIPTION
Pre
Catalyst
Analyzer
Manufacturer
and
Type
Spectral
Resolution
Nicolet
Rega
7000
OScm"

H
Detector
Type
I
MCT­
A
II
Each
unit
and
the
associated
test
method
have
been
designed
for
measurement
of
raw
exhaust
gases
from
internal
combustion
engines.
Dedicated
temperature
controllers
maintained
the
sample
lines
and
cells
at
the
appropriate
the
design
temperature.
Pressure
was
controlled
by
an
MISS
pressure
controller
on
each
system.
Sample
flow
to
each
analyzer
was
between
8
­
15
liters/
minute.
The
units
utilized
a
high­
energy
mid­
range
IR
source
and
are
equipped
with
a
modulating,
potassium
bromide
beamsplitter
with
MCT­
A
liquid
nitrogen
cooled
detectors.
The
cells
have
been
equipped
with
specific
optical
windows
to
prevent
signal
degradation
from
damaged
optics
due
to
moisture
and
corrosive
gases
present
in
the
exhaust
stream.

Pre
Engine
Test
Calibration
Prior
to
initiation
of
an
engine
specific
test
program,
the
FTIR
sampling
systems,
both
pre
and
post
catalyst
sample
trains
underwent
an
EPA
Method
301
validation
process.
The
validation
process
was
to
verify
the
sample
and
analytical
system
performance
in
relation
to
precision
and
accuracy
of
data
collected.
Additional
calibration
procedures
prior
to
testing
of
the
engine
were
as
follows:

Emissions
Testing
4­
16
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADO
STATE
UNIVERSITY
Source
Evaluation
­
Acquired
initial
source
data
to
verify
concentration
ranges
of
target
compounds
and
possible
interferants.
This
was
accomplished
prior
to
and
during
the
Method
30
1
validation
process
Sample
System
Leak
Check
­
Sample
system
leak
checks
were
performed.
The
leak
check
procedure
encompassed
the
sample
train
from
the
sample
filter
to
the
pump
outlet.

A
dedicated
rotameter
has
been
installed
on
the
discharge
side
of
the
sample
pump.
With
the
sample
system
operating
at
typical
temperatures
and
pressures
(
sample
pump
will
pull
a
slight
vacuum
on
the
suction
side),
the
sample
flow
rate
from
the
rotameter
was
recorded.
The
inlet
to
the
sample
filter
located
just
downstream
of
the
sample
probe
was
closed
and
the
flow
rate
through
the
rotameter
was
monitored.
The
flow
rate
through
the
rotameter
went
to
zero.
The
leak
checks
were
determined
to
be
acceptable,
as
the
leak
rate
was
less
than
4%
of
the
standard
sampling
rate
or
500ml/
min,
whichever
is
less.

Sample
system
leak
check
data
sheets
are
provided
in
Appendix
F
of
this
document.

Analyzer
Leak
Check
­
With
the
FTIR
analyzers
operating
at
normal
operating
temperatures
and
pressures,
the
operating
pressures
were
recorded.
The
automatic
pressure
controllers
were
then
disabled,
and
the
inlet
valves
to
the
FTIR
analyzers
were
closed.
The
measurement
cells
were
then
evacuated
to
20%
or
less
of
their
normal
operating
pressure.
After
the
measurement
cells
were
evacuated,
each
measurement
cell
was
then
isolated
and
the
cell
pressure
monitored
with
a
dedicated
pressure
sensor.
The
leak
rate
of
each
measurement
cell
was
less
than
10
Torr
per
minute
for
a
one­
minute
period.
The
analyzer
leak
rate
was
determined
to
be
acceptable.
Analyzer
leak
check
data
sheets
are
provided
in
Appendix
F
of
this
document.

Cell
Pathlength
Determination
­
The
cell
pathlength
was
to
be
determined
using
the
measurement
procedures
as
outlined
in
the
Field
Procedure
Section
of
the
document
entitled
"
Protocol
For
Performing
Extractive
FTIR
Measurements
To
Characterize
Various
Gas
Industry
Sources
For
Air
Toxics",
prepared
by
Radian
International
for
the
Gas
Research
Institute.
Because
the
units
are
fixed
pathlength
(
non­
adjustable)

measurement
cells,
which
are
stationary
units
dedicated
to
a
specific
task,
the
pathlength
determination
process
was
determined
not
to
be
necessary.
The
units
are
"
as
specified"

from
the
manufacturer,
and
have
passed
all
validation
and
calibration
procedures
at
this
fixed
nathlentih.

Emissions
Testing
4­
17
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
C~
L~~
DOSTATE
UNIVERSITY
Daily
Calibration
Procedures
­
Pre
Test
The
following
daily
calibration
procedures
were
performed
prior
to
the
initiation
of
each
day's
testing.

1.)
Instrument
Stabihtion
­
To
ensure
the
FTIR
instruments
were
operating
in
a
stable
manner,
verification
of
the
operation
of
the
following
components
at
the
beginning
of
each
day
was
performed:

a.)
All
instrument
heated
devices
and
temperature
controller
were
at
operating
temperature
and
performing
properly.

b.)
Pressure
sensor
and
pressure
controllers
were
at
operating
conditions
and
performing
properly.

c.)
Sample
systems
(
pumps,
filters,
flow
meters,
and
water
knockouts)
were
functioning
properly.

2.)
Instruments
were
operated
on
a
conditioned
air
source
for
a
minimum
of
30
minutes
prior
to
conducting
background
spectrum
procedures.
When
the
instruments
were
in
standby
mode,
between
test
days,
the
analyzers
and
all
components
were
kept
at
normal
operating
temperatures.
The
analyzers
operated
on
a
conditioned
air
at
all
times
when
not
involved
with
data
acquisition.

3.)
Background
spectrum
procedures
­
After
purging
with
a
conditioned
air
source
for
a
minimum
of
30
minutes,
the
instruments
were
allowed
to
stabilize
by
flowing
an
ultra
high
purity
N2
gas
through
the
measurement
cell
for
a
minimum
of
ten
minutes.
During
the
stabilization
process,
the
FTIR
spectra
were
monitored
until
the
concentrations
of
CO
and
HZ0
were
reduced
and
normal
steady
state
background
levels
had
been
achieved.
The
following
procedures
were
then
performed:

a.)
Check
for
proper
interferogram
signal
using
alignment
software
b.)
Collect
a
single
beam
spectrum
and
inspect
for
irregularities
c.)
Check
the
single
beam
spectrum
for
detector
non­
linearity
and
correct
if
necessary
d.)
Perform
an
instrument
alignment
procedure
e.)
Collect
a
background
spectrum
­
The
background
spectrum
was
comprised
256
scans,
which
was
equal
to
or
greater
than
the
number
of
scans
used
for
sample
analysis.

4.)
Analyzer
Diagnostics
­
Perform
an
analyzer
diagnostic
procedure
by
analyzing
a
diagnostic
standard.
The
standard
was
an
EPA
Protocol
1
CO
gas
standard
at
concentration
levels
indicative
of
the
emissions
source,
109
ppm.
A
CO
standard
was
recommended
due
to
the
distinct
spectral
features,
which
are
sensitive
to
variations
in
system
operation
and
performance.
The
standard
was
introduced
directly
into
the
instrument.
The
instrument
readings
were
allowed
to
stabilize
and
a
Emissions
Testing
4­
18
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
US.
EPA.
COLORADOSTATEUNIVERSITY
five­
minute
set
of
data
was
acquired.
The
calculated
accuracy
and
precision
based
on
equations
from
the
document
entitled
"
Protocol
for
Performing
Extractive
FTIR
Measurements
To
Characterize
Various
Gas
Industry
Sources
for
Air
Toxics",

prepared
by
Radian
International
for
the
Gas
Research
Institute,
was
acceptable.
The
pass/
fail
criteria
for
accuracy
and
precision
was
+
10%
of
the
known
standard
for
the
instrument
to
be
acceptable.
Each
instrument
meets
this
criterion
for
all
daily
calibrations.
Analyzer
diagnostic
data
sheets
are
provided
in
Appendix
F
of
this
document.

5.)
Additional
Analyzer
Diagnostic
­
An
additional
diagnostic
check
was
performed
to
ensure
system
operation
and
performance.
A
second
diagnostic
standard
comprised
of
a
multi­
gas
composition
was
analyzed
by
the
same
procedure.
The
gas
consisted
of
COZ,
CO,
CH4,
and
NO,
in
concentrations
similar
to
exhaust
gas
composition.
The
same
pass/
fail
criteria
was
used
to
evaluate
each
analyzer's
performance
when
analyzing
the
multi­
gas
standard.
Each
instrument
meets
this
criteria
for
all
daily
calibrations.
Analyzer
diagnostic
data
sheets
are
provided
in
Appendix
F
of
this
document.

6.)
Indicator
Check
&
Sample
Integrity
Check
­
An
indicator
check
procedure
was
performed
on
each
analyzer
by
analyzing
a
certified
indicator
standard.
The
standard
was
either
a
NIST
traceable,
EPA
Protocol
1
gas
standard,
or
highest
grade
standard
available
of
a
surrogate/
analyte
gas
concentration
at
levels
indicative
of
the
emissions
source.
A
formaldehyde
standard
(
concentration
of
10.66
ppm)
was
used
due
to
the
fact
that
formaldehyde
represents
a
sampling
challenge
because
of
its
solubility
in
water.
The
standard
was
introduced
directly
into
the
instrument.
The
instrument
readings
were
allowed
to
stabilize
and
a
five­
minute
set
of
data
was
acquired.
Next,
the
indicator
standard
was
introduced
into
the
sample
system
at
the
sample
filter
located
just
downstream
of
the
sample
probe.
The
instrument
readings
were
allowed
to
stabilize
and
a
five­
minute
set
of
data
was
acquired.
The
calculated
accuracy
and
precision
based
on
equations
from
the
document
entitled
"
Protocol
For
Performing
Extractive
FTIR
Measurements
To
Characterize
Various
Gas
Industry
Sources
For
Air
Toxics",
prepared
by
Radian
lntemational
for
the
Gas
Research
Institute.
The
pass/
fail
criteria
for
accuracy,
precision,
and
recovery
was
+,
10%
of
the
known
standard
(
recovery
was
­
t
10%
of
the
instrument
reading
with
the
indicator
gas
introduced
directly
into
the
instrument.)
for
the
instrument
to
be
acceptable.
Each
instrument
meets
this
criteria
for
all
daily
calibrations.
Indicator
check
and
sample
integrity
check
data
sheets
are
provided
in
Appendix
F
of
this
document.

Emissions
Testing
4­
19
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
lntemal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADOSTATEUNIVERSITY
Daily
Calibration
Procedures
­
Background
assessment
The
baseline
absorbance
was
continually
monitored
during
data
acquisition
procedures.
If
it
was
determined
by
PES,
ERG,
and
CSU
personnel
that
the
baseline
had
changed
by
more
than
0.1
absorbance
units,
the
instrument
interferometer
was
realigned
and
a
background
spectrum
collected.

Daily
Calibration
Procedures
­
Post
Test
Upon
completion
of
the
daily
test
program
steps
4
­
6
of
the
pre
test
calibration
procedures
were
repeated.
All
analyzers
met
all
acceptance
criteria
and
calibration
procedures.
All
post
test
calibration
data
sheets
are
presented
in
Appendix
F
of
this
document.

4.9
TEST
SPECIFIC
­
FTIR
VALIDATION
PROCEDURES
To
ensure
the
accuracy
of
data
collected
during
testing,
the
test
program
required
procedures
to
evaluate
instrument
performance.
Prior
to
collecting
test
data,
a
validation
procedure
was
performed
on
each
FTIR
sample
train,
both
pre­
catalyst
and
post­
catalyst,
for
the
diesel
fueled
engine
classification.
The
specific
sample
trains
are
as
follows:

1.)
Pre­
catalyst
emissions
sample
trains
from
the
exhaust
of
diesel
fueled
engines.

2.)
Post­
catalyst
emissions
sample
trains
from
the
exhaust
of
diesel
fueled
engines.

Each
sample
train
was
validated
for
formaldehyde,
acetaldehyde,
and
acrolein:

Instrument
Description
Refer
to
FTIR
calibration
procedures
for
FTIR
instrument
description.

Procedures
The
validation
procedure
was
conducted
in
basic
accordance
with
procedures
outlined
in
Method
301­"
Field
Validation
of
Pollutant
Measurement
Methods
from
Various
Waste
Media".

Validation
procedures
for
aldehydes
utilized
an
analyte
spiking
technique
as
specified
in
Method
301.
Validation
procedures
for
criteria
pollutants
and
moisture
will
use
comparative
sampling
to
the
appropriate
EPA
reference
methods.
Paired
sampling
was
not
performed
under
the
validation
procedure.
The
paired
samples
will
be
generated
from
FTIR
analyzer
data
and
reference
method
analyzer
data
collected
during
the
test
program.
The
procedures
for
the
validation
process
are
as
follows:

Emissions
Testing
4­
20
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.

_
I
_
"
"
­
­
­
COLORADO
STATE
UNIVERSITY
Analyte
Spiking:

The
process
was
carried
out
by
means
of
dynamic
analyte
spiking
of
the
sample
gas.
The
sample
stream
of
the
exhaust
gas
was
spiked
with
the
specific
analyte
after
the
sample
probe,
and
before
the
sample
filter.
Spike
levels
for
the
specific
aldehydes
were
determined
and
the
spike
gas
concentrations
were
generated
for
the
specific
aldehydes
using
the
following
methods:

Formaldehyde:
Formaldehyde
spike
gas
was
generated
by
volatilization
of
a
formalin
solution
prepared
from
a
stock
formalin
solution
of
37%
formaldehyde
by
weight.
The
solution
was
injected
into
a
heated
vaporization
block.
The
vaporized
formalin
solution
was
mixed
with
an
acetylaldehyde/
acrolein
carrier
gas
and
carried
into
the
sample
exhaust
stream.
Carrier
gas
flow
rate
was
measured
by
a
mass
flow
meter
equipped
with
readout.

AcetlyaldehydeIAcrolein:

Acetlyaldehyde
and
acrolein
spike
samples
were
generated
from
a
certified
gas
standard
(
Scott
Specialty
Gases,
52%
analytical
accuracy)
which
contained
both
analyte
species
and
a
sulfur
hexaflouride
(
SF6)
tracer
gas.
Carrier
gas
flow
rate
was
measured
by
a
mass
flow
meter
equipped
with
readout.

Analyte
specific
spike
gas
was
introduced
to
the
FTIR
sample
train
upstream
of
the
sample
system
filter.
The
spike
gas
was
introduced
at
a
known
flow
rate.
The
spike
gas
flow
was
controlled
by
a
three­
way
solenoid
valve,
which
directed
gas
either
into
the
sample
stream
or
diverted
the
spike
gas
to
the
atmosphere.
This
allowed
for
uninterrupted
flow
of
the
analyte
spike
gas
source
during
the
validation
procedures.

The
formaldehyde
and
acetylaldehyde/
acrolein
validation
runs
were
conducted
simultaneously.
The
validation
test
runs
consisted
of
24
test
runs,
12
spiked
and
12
unspiked
runs,
which
were
paired
and
grouped
further
into
six
groups
of
2
spiked/
unspiked
pairs
to
simulate
the
"
quad
train"
approach
used
for
Method
301
statistical
calculations.
Samples
were
one
minute
in
duration.
Measurement
procedures
for
acquiring
the
spiked/
unspiked
pairs
are
as
follows:

1.1
2.1
3.1
4.1
Verify
stable
engine
operation
Begin
measurement
of
the
unspiked
native
exhaust
stack
gas.
Upon
completion
of
acquiring
the
unspiked
sample,
initiate
spike
gas
flow
into
sample
stream.

Let
system
equilibrate.

Emissions
Testing
4­
2
1
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
5
.)
Begin
measurement
of
the
spiked
exhaust
gas
sample.

6.)
Upon
completion
of
acquiring
the
spiked
sample,
divert
spike
gas
flow
to
atmosphere.

7.)
Let
system
equilibrate.

8.)
Repeat
items
2
through
7.

This
procedure
was
performed
twelve
times
to
acquire
the
appropriate
number
of
spiked/
unspiked
pairs.
To
ensure
stable
engine
operation
during
the
validation
procedure,
engine
operating
data
was
collected
during
the
spiking
process.

4.10
TEST
SPECIFIC
­
GENERAL
CALIBRATION
To
ensure
the
accuracy
of
data
collected
during
testing,
the
test
procedure
required
that
all
instrumentation
be
routinely
calibrated.
Calibrations
and/
or
calibration
checks
were
performed
within
one
week
before
initiation
of
testing,
and
upon
completion
of
the
entire
test
program
to
ensure
that
no
"
drift"
has
occurred.
The
devices
calibrated
included
the
dynamometer
50004b.
load
cell
and
amplifier,
all
thermocouples,
pressure
transducers,
and
all
pressure
transmitters.

Dynamometer
Load
Cell
and
Amplifier
(
Daily)

The
5000
pound
load
cell
and
amplifier
was
calibrated
prior
to
the
engine
test
section.
The
calibration
procedure
is
outlined
in
a
document
contained
in
Appendix
M
of
this
document.

Calibration
of
the
load
cell
and
amplifier
were
then
be
verified
by
applying
the
full
range
of
load
without
any
adjustments
to
the
offset
or
gain
of
the
instrumentation.
Calibration
checks
were
performed
on
a
daily
basis
prior
to
starting
the
engine
to
identify
and
correct
any
drift
in
the
load
cell
or
amplifier.
These
checks
used
the
same
procedure
as
the
calibration
verification.
If
the
daily
calibration
check
showed
an
indicated
load
that
exceeded
f
1
.
O%
of
the
torque
applied
by
the
standard
weights,
the
full
calibration
procedure
was
performed.
The
dynamometer
was
within
acceptable
limits
during
the
test
program.
Dynamometer
calibration
data
sheets
are
provided
in
Appendix
L
of
this
document.

Thermocouples
(
Within
one
week
prior
to
initiation
of
each
engine
test
program)

K­
type
insertion
thermocouples
are
used
throughout
the
Large
Bore
Engine
Testbed
with
compensation
performed
through
the
engine
control
and
data
acquisition
hardware.
The
thermocouples
were
calibrated
using
a
Ronan
X88
portable
calibrator
calibrated
within
+
l
.
O
°
F
of
N.
I.
S.
T.
standard
by
an
independent
laboratory.
The
thermocouple
signal
was
zeroed
and
the
gain
adjusted
at
full
span
until
the
value
displayed
by
the
NetCon
5000
matched
the
setting
of
the
Ronan
X88
within
+
2.0
°
F.
Once
the
zero
and
gain
have
been
set
a
minimum
of
two
mid­
point
Emissions
Testing
4­
22
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
COLORADO
STATEUNIVERSITY
temperatures
were
checked
to
verify
the
calibration.
Thermocouple
calibration
data
sheets
are
provided
in
Appendix
J
of
this
report.

Pressure
Transducers
(
Within
one
week
prior
to
initiation
of
each
engine
test
program)

3­
way
valves
have
been
installed
on
the
pressure
transmitters
to
allow
calibration
without
removing
the
sensor
from
the
system.
The
Model
320
Beta
calibrator
used
for
transducers
calibration
provides
an
accuracy
of
0.05%
of
reading
or
0.02%
of
full
span
and
is
calibrated
to
N.
I.
S.
T.
standards
by
an
independent
laboratory.
The
transducer
was
zeroed
and
the
gain
adjusted
at
full
span
until
the
value
displayed
by
the
NetCon
5000
was
within
+
1
.
O
psig
of
the
pressure
supplied
by
the
pressure
calibration
standard.
A
minimum
of
two
midpoints
was
checked
to
verify
calibration.
Pressure
transducer
calibration
data
sheets
are
provided
in
Appendix
J
of
this
report.

Pressure
Transmitters
(
Within
one
week
prior
to
initiation
of
each
engine
test
program)

Pressure
which
were
critical
to
control
and
emissions
calculations
were
measured
using
RosemountQ
305
1C
transmitters.
The
calibration
was
performed
at
the
transmitter
and
no
adjustments
are
made
to
the
current
loop.
A
known
pressure
was
supplied
to
the
sensing
port
of
the
transmitter
using
the
Model
320
Beta
calibrator.
The
current
transmitter
was
then
zeroed
and
spanned
at
the
full
range
value
of
the
system.
A
minimum
of
two
mid­
span
points
was
checked
to
verify
calibration.
Pressure
transmitter
calibration
data
sheets
are
provided
in
Appendix
J
of
this
report.

Emissions
Testing
4­
23
Pacific
Environmental
Services
Of
Control
Devices
for
Reciprocating
Internal
Combustion
Engines
In
Support
of
Regulatory
Development
By
the
U.
S.
EPA.
