Submitted
to
OMB
for
Review
on
05­
05­
2005
CHAPTER
1
HDIUT
Economic
Assessment
This
chapter
contains
our
economic
analysis
of
the
costs
associated
with
implementing
a
manufacturer­
run,
in­
use
NTE
testing
program
for
heavy­
duty
diesel
engines
and
vehicles.
The
cost
of
the
program
is
dependent
on
several
key
variables.
One
of
these
is
the
number
of
vehicles
tested
under
the
Phase
1
and
2
testing
schemes.
This,
of
course,
depends
on
how
many
vehicles
pass,
or
fail,
the
vehicle
pass
criteria
at
various
points
under
the
tiered
testing
procedures.
Also
important
is
just
how
each
manufacturer
will
chose
to
design
and
conduct
the
test
program,
how
many
portable
emission
measurement
systems
(
PEMS)
will
be
purchased,
and
the
availability
of
test
vehicles.
Obviously,
it
is
difficult
to
project
these
variables
for
an
all
new
program.
However,
based
on
our
experience
with
in­
use
emissions
testing,
including
the
development
and
use
of
a
portable
emission
measurement
device
for
compliance
testing,
we
have
identified
a
set
of
reasonable
testing
scenarios
that
allow
us
to
estimate
the
potential
costs
associated
with
the
program.

As
part
of
the
comments
on
the
proposed
rule,
one
manufacturer
provided
comment
on
a
number
of
the
values
and
assumptions
used
in
the
analysis
We
have
considered
each
of
these
and
revised
the
analysis
where
appropriate.
In
most
cases
these
changes
increased
costs,
leading
to
what
EPA
believes
to
be
a
more
conservative
assumption.
The
most
notable
change
is
our
decision
to
eliminate
scenario
1,
and
to
assume
all
engine
family
testing
events
involve
longer
trips
and
overnight
travel
as
laid
out
in
our
original
Scenario
2.

This
chapter
is
divided
into
several
parts.
Section
1.1
contains
a
brief
outline
of
the
methodology
used
to
estimate
the
associated
costs
is
presented.
Section
1.2
develops
the
fixed
and
variable
cost
components
associated
with
the
program.
Section
1.3
summarizes
the
component
costs
into
estimates
of
the
overall
cost
of
the
program.
All
costs
are
estimated
in
Dec
2004
dollars.
In
most
cases
our
estimates
in
2003
were
increased
by
2.7
percent
based
on
the
CPI
inflator.

1.1
Overview
All
costs
are
divided
into
either
fixed
or
variable
cost
components.
Fixed
costs
are
associated
with
the
direct
expense
of
purchasing
the
requisite
portable
emission
measurement
system
(
PEMS)
units.
Variable
costs
depend
primarily
on
the
number
of
families
and
vehicles
tested.
They
include
the
direct
costs
for
vehicle
recruiting,
labor
for
on­
site
testing,
instrument
calibration
and
maintenance,
travel,
data
analysis,
and
reporting
expenses.
Variable
costs
also
include
indirect
costs
associated
with
overhead
and
general
and
administrative
(
G&
A)
expenses.

To
explore
the
range
of
possible
costs,
we
assessed
a
range
of
costs
associated
with
different
testing
intensities
under
Phase
1
or
Phase
2
of
the
program
(
i.
e.,
the
amount
or
number
of
vehicle
tests
that
might
occur
under
the
two
phases
of
testing).
Finally,
we
combined
this
information
to
show
a
range
of
possible
costs
and
a
single
point
estimate
by
assuming
a
specific
mix
of
the
above
Submitted
to
OMB
for
Review
on
05­
05­
2005
testing
variables.
The
results
are
presented
for
all
heavy­
duty
engine
manufacturers
as
annualized
costs,
total
costs
for
the
first
five
years
of
the
program,
and
costs
over
30
years.
Submitted
to
OMB
for
Review
on
05­
05­
2005
1
Chapter
2
contains
additional
information
on
the
status
and
development
of
portable
emission
measurement
systems.
1.2
Cost
Components
1.2.1
Fixed
Costs
Fixed
costs
for
this
program
are
primarily
associated
with
buying
PEMS
units.
Some
of
the
fixed
cost
components
have
significant
uncertainties
associated
with
them.
Portable
measurement
devices
are
already
commercially
available
that
can
measure
all
the
gaseous
pollutants
required
by
the
program.
These
systems
cost
approximately
$
70,000
per
unit.
The
commenter
mentioned
above
indicated
that
costs
were
30
percent
greater.
Based
on
our
experience
and
investment
in
developing
portable
particulate
matter
(
PM)
measurement
technology,
we
estimate
that
systems
capable
of
measuring
both
gaseous
and
PM
emissions
will
cost
an
additional
$
30,000
per
unit.
1
The
commenter
mentioned
above
indicated
interactions
with
vendors
indicating
costs
3­
4
times
greater.
EPA
has
elected
to
very
conservatively
estimate
costs
ranging
from
$
100,000­$
180,000.
Taking
the
average,
we
estimate
the
capital
cost
for
portable
measurement
devices
that
measure
both
gaseous
and
PM
emissions
is
approximately
$
140,000
per
unit.

As
described
in
Chapter
2
of
this
document,
we
assume
that
engine
manufacturers
will
initially
purchase
PEMS
units
with
the
capability
to
measure
gaseous
pollutants
in
time
to
coincide
with
the
initiation
of
the
pilot
program
during
the
2005
calendar
year.
Add­
on,
modular
devices
that
measure
PM
emissions
will
be
purchased
later
in
2006
as
this
technology
becomes
available.
Regardless
of
the
purchasing
strategy,
we
assume
all
equipment
purchases
occur
at
the
beginning
of
2005
in
order
to
simplify
the
analysis.

We
estimate
that
portable
emission
measurement
devices
have
a
product
life
cycle
of
five
years
for
the
purposes
of
the
program.
After
that
time
they
are
assumed
to
be
replaced
with
brand
new
equipment.
Also,
we
assume
there
is
no
salvage
value
for
units
that
may
remain
in
service
for
other
less
rigorous
or
less
important
duties
after
five
years,
although
this
could
occur
in
some
instances.
Alternatively,
some
manufacturers
may
chose
to
replace
or
rebuild
component
parts
of
a
PEMS
unit
rather
than
replace
the
entire
unit
after
five
years.
To
the
extent
this
may
occur,
we
assume
such
a
maintenance
strategy
will
cost
approximately
the
same
as
the
replacement
strategy.
The
annualized
cost
of
a
single
PEMS
unit
can
now
be
calculated
by
using
the
above
values
and
assuming
a
typical
capital
recovery
rate
of
seven
percent
per
annum.
The
result
is
an
annualized
cost
of
$
34,145
per
unit.

The
total
annualized
fixed
cost
for
the
program
depends
on
how
many
PEMS
units
each
manufacturer
will
purchase,
the
fraction
of
time
the
equipment
is
used
for
the
in­
use
testing
program,
and
the
number
of
manufacturers
subject
to
the
requirements.
Each
of
these
cost
components
is
addressed
separately
below.

We
assume
that
each
manufacturer
will
purchase
three
units.
We
chose
this
number
to
Submitted
to
OMB
for
Review
on
05­
05­
2005
illustrate
the
average
equipment
cost
of
the
program
recognizing
that
two
units
is
adequate
to
perform
more
than
the
needed
amount
of
tests
for
even
the
largest
manufacturer
if
its
program
is
designed
so
that
testing
can
be
conducted
in
an
orderly,
efficient
manner.
We
recognize
that
manufacturers
with
a
limited
number
of
engine
families
may
need
only
one
unit
with
a
backup
unit.
Manufacturers
with
a
large
number
of
families
may
prefer
an
additional
unit
for
backup.

Our
final
assumption
in
estimating
the
total
annual
fixed
cost
of
the
program
is
that
13
engine
manufacturers
will
participate
in
the
program.
This
is
the
number
of
companies
that
certified
heavy­
duty
diesel
engines
in
the
2005
model
year.
The
manufacturers
are
shown
below.

Caterpillar,
Inc.
Cummins
Engine
Company,
Inc.
DaimlerChrysler
AG
John
Deere,
Inc
Detroit
Diesel
Corporation
General
Engine
Products
General
Motos
Corporation
Hino
Motors,
Ltd.
International
Truck
and
Engine
Corporation
Isuzu
Motors
Limited
Mack
Mitsubishi
Motors
Corporation
Volvo
Powertrain
Corporation.

Using
the
above
information,
the
total
annualized
fixed
cost
of
the
program
$
1,331,665,
as
shown
in
Table
1­
1.

1.2.2
Variable
Costs
Variable
costs
are
grouped
into
two
broad
categories:
cost
per
vehicle
test
and
cost
per
family.
This
approach
allows
us
to
more
easily
account
for
tests
that
must
be
repeated
on
the
same
vehicle
in
order
to
obtain
a
valid
test
result.
Repeat
testing
can
occur
in
the
laboratory
due
to
equipment
or
vehicle
malfunctions,
and
operator
error.
We
expect
that
similar
problems
may
occur
in
field
testing,
and
assume
that
these
issues
can
generally
be
remedied
at
the
testing
location.
Further,
a
vehicle
may
be
tested
a
second
day
under
the
terms
of
the
program
if
less
than
three
hours
of
non­
idle
operation
are
recorded
during
the
first
"
shift
day"
of
testing.
Obviously,
multiple
tests
on
the
same
vehicle
do
not
directly
affect
other
costs
associated
with
testing
an
engine
family,
e.
g.,
vehicle
recruiting.
Therefore,
these
costs
are
estimated
separately
in
our
analysis.

Also,
as
noted
earlier,
many
of
the
costs
of
the
program
vary
by
the
relative
availability
of
test
vehicles
(
i.
e.,
how
difficult
it
is
to
access,
instrument,
and
test
vehicles
at
a
job
site),
and
the
type
Submitted
to
OMB
for
Review
on
05­
05­
2005
and
amount
of
travel
required
to
conduct
the
test
campaign
(
e.
g.,
overnight
versus
local
travel).
In
order
to
reasonably
bracket
these
cost
elements,
we
constructed
two
testing
scenarios
that
differ
in
the
above
characteristics.
These
scenarios
are
based
on
our
direct
experience
in
conducting
in­
use
testing
of
heavy­
duty
trucks
with
portable
emission
measurement
systems
under
the
1998
consent
decrees,
our
continued
development
of
portable
measurement
systems,
and
a
recently
awarded
EPA
contract
to
conduct
a
large
scale,
in­
use
testing
pilot
program
in
Kansas
City,
Missouri
for
passenger
cars
(
USEPA
2003).
The
two
testing
scenarios
are
described
in
the
following
section
by
first
identifying
some
of
the
key
elements
shared
by
both
scenarios
and
then
presenting
the
specifics
of
each
scenario
separately.

1.2.2.1
General
Description
of
Testing
Scenarios
Our
testing
scenarios
share
a
number
of
key
assumptions
which
we
believe
provide
a
reasonable
representation
of
how
manufacturers
are
likely
to
design
and
conduct
testing
under
this
program.
Alternatively,
if
an
engine
manufacturer
decides
to
contract
for
testing
services,
we
expect
the
service
provider
to
similarly
design
and
conduct
the
testing
campaign.
One
of
our
most
basic
methodological
assumptions
is
that
field
testing
will
usually
consist
of
a
multi­
day
campaign
where
a
minimum
of
five
vehicles
are
tested.
This
number
was
chosen
for
several
reasons.
First,
it
captures
the
type
of
back
 
to­
back
vehicle
testing
likely
to
be
employed
in
order
to
facilitate
efficient
testing.
Second,
it
represents
the
minimum
number
of
test
vehicles
for
Phase
1
testing
under
the
program.
Third,
and
finally,
it
simplifies
the
analysis
when
evaluating
the
potential
costs
of
higher
testing
intensities
associated
with
the
maximum
number
of
vehicles
that
may
be
tested
under
Phase
1
(
10
total
vehicles)
and
Phase
2
(
20
total
vehicles).
These
later
testing
levels
are
simple
multiples
the
Phase
1
minimum
testing
scenario,
i.
e.,
two
times
or
four
times,
respectively.
Other
key
assumptions
are
described
below.

S
Vehicle
recruiting
and
pre­
screening
of
prospective
test
vehicles
will
be
done
by
telephone
or
other
means
prior
to
the
test
site
visit.

S
All
portable
measurement
systems
will
be
inspected
and
pre­
calibrated
at
the
manufacturer's
facility
prior
to
deployment
in
the
field.

S
Field
testing
will
be
conducted
by
two
people.
One
is
an
engineer
and
the
other
a
qualified
technician.
Both
are
capable
of
installing
and
operating
the
portable
measurement
systems,
screening
vehicles
for
OBD
trouble
codes
and
MIL
indications,
performing
maintenance
on
the
portable
systems,
etc.
The
technician
is
also
capable
of
performing
all
required
inspections
of
the
vehicle's
mechanical,
electrical,
and
emission
control
systems;
as
well
as
performing
allowable
maintenance
and
the
setting
of
adjustable
engine
parameters
as
required.

S
The
test
engineer
and
technician
will
coordinate
their
activities
to
optimize
their
productivity.
For
example,
the
engineer
may
acquire
and
enter
a
vehicle's
history
and
vital
statistics
into
an
electronic
database
concurrently
while
the
technician
performs
vehicle
inspection
and
allowable
maintenance.

S
Test
vehicles
for
an
engine
family
are
obtained
from
two
independent
sources,
as
required
by
the
regulation.
It
is
assumed
that
the
sources
are
located
relatively
close
Submitted
to
OMB
for
Review
on
05­
05­
2005
to
each
other
to
minimize
travel
distances
between
test
sites.

S
The
test
sites
are
not
necessarily
in
relatively
close
proximity
to
a
manufacturer's
technical
center,
test
center,
maintenance
facility,
or
other
similar
location
thus
increasing
personnel
travel
and
field
logistics
costs.

S
In
many
cases
test
vehicles
will
depart
and
not
return
to
the
same
location
the
same
day,
necessitating
an
overnight
stay.

S
Field
personnel
have
access
to
the
test
vehicles
and
a
service
location
before
and
after
the
shift
day
as
needed
to
install
and
remove
the
portable
measurement
devices.
Special
arrangements
with
the
vehicle
owner/
operator
may
be
necessary.

S
Two
portable
emission
measurement
systems
will
be
deployed
simultaneously
during
a
test
site
visit
when
possible,
i.
e.,
two
vehicles
will
be
tested
at
the
same
time.

S
All
necessary
tools,
spare
parts,
testing
supplies,
office
supplies,
etc.
will
be
taken
to
or
otherwise
supplied
at
the
site
of
testing
to
avoid
unnecessary
delays.

It
is
also
possible
that
test
vehicles
may
be
operated
on
multi­
day
driving
routes,
i.
e.,
longhaul
operation.
We
do
not
think
this
will
be
a
standard
practice
for
all
vehicles/
engines,
but
may
happen
for
engines
used
in
over
the
road
trucks
and
buses.
As
alluded
to
above,
we
have
incorporated
this
assumption
into
the
analysis
for
all
families.

There
are
also
some
common
on­
site
work
activities.
These
categories
are
described
below
in
their
approximate
order
of
occurrence.

S
Vehicle
History/
Documentation.
Obtaining
all
relevant
information
not
available
prior
to
the
field
visit
or
verifying
the
accuracy
of
information
previously
obtained.

S
Vehicle
Set­
to­
Specification.
Inspecting
the
vehicle,
performing
allowable
maintenance,
and
setting
all
adjustable
engine
control
parameters.

S
PEMS
Installation.
Install
the
portable
measurement
system
onboard
the
test
vehicle.
Warm
the
instrument
to
operating
temperature,
verify
proper
operation,
perform
final
calibrations
and
span
instruments,
etc.

S
PEMS
Data
Acquisition.
Download
the
measurement
data
set;
perform
on­
site
data
verification
and
initial
quality
assurance;
and
record
and
store
all
other
relevant
test
information.

S
PEMS
Removal.
Verify
proper
operation,
perform
post­
test
calibration,
and
remove
system
from
the
test
vehicle.

S
Miscellaneous
Time.
Non­
routine
labor
for
repairing
or
replacing
parts
of
the
portable
emission
measurement
systems
or
test
vehicles,
obtaining
supplies
not
at
the
test
site,
etc.

Now
that
the
overall
components
of
the
testing
scenarios
have
been
identified,
the
specific
scenario
we
have
analyzed
will
be
presented.
We
believe
this
is
conservative
since
it
assumes
all
families
are
used
in
overnight
operation.
Submitted
to
OMB
for
Review
on
05­
05­
2005
Limited
Vehicle
Access
and
Overnight
Travel
Under
this
scenario,
test
vehicles
are
not
readily
available
and
the
testing
technician
and
engineer
must
sometimes
work
around
a
test
vehicle's
normal
daily
work
shift.
This
means
that
the
work
day
for
the
testing
personnel
includes
the
time
the
vehicle
is
being
driven
over
its
normal
work
route.
In
these
instances,
we
made
a
worst
case
assumption
that
the
testing
personnel
remain
"
idle"
on
the
job
site
or
that
they
are
on
the
vehicle,
with
this
time
charged
to
the
in­
use
testing
program.
Also,
the
test
sites
are
located
far
enough
away
from
the
manufacturer's
facility,
or
employees
home
base,
that
a
single
round
trip
to
and
from
the
job
site
and
overnight
travel
is
required.
However,
the
sites
are
still
close
enough
to
one
another
that
travel
between
the
two
locations
is
not
restrictive
or
prohibitive.

1.2.2.2
Variable
Cost
Per
Test
As
described
above,
the
cost
to
perform
an
individual
vehicle
test
is
based
on
a
Phase
1
testing
scheme
where
a
minimum
of
five
vehicles
must
be
tested.
Test
costs
consist
of
direct
labor,
labor
overhead,
other
direct
costs,
and
general
and
administrative
overhead.
Each
of
these
cost
components
is
described
below.

1.2.2.2.1
Direct
Labor
The
cost
of
direct
labor
for
each
scenario
is
estimated
by
applying
an
hourly
compensation
rate
by
labor
category
to
the
various
activities
associated
with
the
field
testing
campaign.
Table
1­
2
display
the
work
flow
for
the
scenario,
broken
down
by
activity,
labor
type,
and
number
of
hours
spent
preforming
each
activity.
These
depictions
reflect
the
assumptions,
work
activities,
and
optimization
of
the
work
flow
as
described
in
Section
1.2.2.1.
The
time
spent
in
the
various
work
tasks
are
estimated
based
on
EPA's
direct
experience
in
conducting
in­
use
testing
with
portable
emission
measurement
devices
and
on
the
recently
awarded
Kansas
City,
Missouri
test
program.

The
test
campaign
is
completed
in
five
days.
The
lack
of
vehicle
flexibility
leads
to
long
work
days
in
this
scenario.
The
total
direct
labor
is
55
and
56
hours
for
the
engineer
and
technician,
respectively.

We
developed
an
hourly
estimate
of
employee
compensation
from
information
published
by
the
Bureau
of
Labor
Statistics,
Office
of
Compensation
and
Working
Conditions,
Employer
Costs
for
Employee
Compensation
(
BLS
2003).
When
increased
by
the
CPI,
Table
1­
3
shows
our
estimate
of
$
32
and
$
29
per
hour
for
an
engineer
and
technician,
respectively.
These
hourly
compensation
rates
are
"
total
compensation,"
and
include
wages
and
salaries,
paid
leave,
supplemental
pay,
and
insurance.
By
comparison,
these
labor
rates
compare
well
with
the
contract
labor
costs
we
incur
in
conducting
our
in­
use
testing
under
the
consent
decrees.
Finally,
we
assume
that
labor
exceeding
40
hours
per
week
is
paid
as
overtime,
i.
e.,
1.5
times
the
normal
hourly
rate.
Submitted
to
OMB
for
Review
on
05­
05­
2005
For
convenience,
Table
1­
3
also
shows
an
hourly
compensation
rate
for
a
manager
using
the
same
literature
source
as
described
above.
This
labor
category
will
be
used
in
Section
1.2.2.3,
where
variable
costs
are
discussed.

The
resulting
direct
labor
cost
per
test
can
now
be
calculated
based
on
the
above
labor
hours
and
hourly
compensation
rates.
As
shown
in
Table
1­
4,
the
resulting
per
test
cost
is
$
773.
Submitted
to
OMB
for
Review
on
05­
05­
2005
1.2.2.2.2
Labor
Overhead
We
assume
that
all
direct
labor
is
burdened
at
100
percent
of
the
total
compensation
rate.
For
simplicity,
overtime
pay
is
also
burdened
at
this
same
overhead
rate.

1.2.2.2.3
Other
Direct
Costs
A
number
of
other
costs
not
related
to
labor
that
are
"
consumed"
during
in­
use
testing
include
office
supplies,
DVDs,
calibration
gases,
and
fuel
for
the
flame
ionization
detector
(
FID).
Based
on
our
experience
with
using
portable
measurement
systems,
we
estimate
that
calibration
gases
will
cost
about
$
75
per
test
and
all
other
supplies
will
cost
about
$
25
per
test.
Therefore,
we
estimate
that
a
total
of
$
100
per
test
for
other
direct
costs.

1.2.2.2.4
Repeat
Tests
Some
in­
use
tests
will
be
voided
due
to
operator
error
and
test
equipment
malfunctions.
Other
tests
will
be
repeated
if
less
than
three
hours
of
non­
idle
vehicle
operation
are
recorded
during
the
first
day
of
testing.
At
our
National
Vehicle
Fuel
and
Emissions
Laboratory
in
Ann
Arbor,
Michigan,
we
experienced
a
test
void
rate
for
laboratory­
based,
non­
research
testing
of
approximately
four
percent
over
the
last
two
years.
We
expect
a
somewhat
higher
void
rate
for
field
testing.
Also,
as
noted,
some
tests
will
be
repeated
do
to
the
three
hour
non­
idle
requirement.
Overall,
we
assume
a
combined
repeat
test
rate
of
10
percent
for
this
analysis.

1.2.2.2.5
General
and
Administrative
Overhead
Certain
costs
are
incurred
for
common
or
joint
objectives
and
therefore
cannot
be
identified
specifically
with
a
particular
project
or
activity.
We
assume
these
general
and
administrative
costs
to
be
6.5
percent
of
all
other
costs.

1.2.2.2.6
Summary
of
Variable
Cost
per
Test
Table
1­
5
summarizes
the
various
direct
cost
elements
described
above.
As
shown,
the
resulting
total
variable
costs
per
test
are
$
1,928.

1.2.2.3
Variable
Cost
Per
Engine
Family
As
with
the
previous
section,
the
cost
per
engine
family
is
based
on
a
Phase
1
testing
scheme
where
a
minimum
of
five
vehicles
must
be
tested.
This
overall
cost
is
composed
of
a
number
of
individual
expenses
such
as
paying
the
test
vehicle's
owner/
operator
an
incentive,
recruitment,
travel,
instrument
pre­
calibration,
data
analysis,
and
reporting.
Each
of
the
cost
elements
are
described
below.
Submitted
to
OMB
for
Review
on
05­
05­
2005
1.2.2.3.1
Vehicle
Incentives
We
generally
offer
a
vehicle's
owner
an
incentive
in
the
form
of
a
government
bond
and
free
vehicle
repairs
as
part
of
our
in­
use
test
programs.
Sometimes
the
owner
cooperates
without
such
an
incentive,
as
most
often
occurs
in
our
in­
use
testing
under
the
consent
decrees.
For
the
purposes
of
this
analysis,
we
assume
that
a
cash
incentive
of
$
150
per
vehicle
will
be
paid
to
the
owner
by
the
engine
manufacturer.
This
is
the
average
cost
of
the
incentive,
with
some
owners
being
offered
more
some
less,
and
some
cooperating
without
an
incentive.
Therefore,
the
total
incentive
for
an
engine
family
tested
under
the
Phase
1
minimum
requirements
is
$
750.

1.2.2.3.2
Direct
Labor
Each
engine
family
will
incur
costs
in
three
main
labor
categories:
vehicle
recruitment,
instrument
pre­
calibration,
and
data
analysis
and
reporting.
We
expect
that
manufacturers
will
rely
heavily
on
their
existing
customer
relationships
to
recruit
appropriate
test
vehicles
from
fleets
or
individual
owners.
Alternatively,
they
will
create
new
lines
of
communication
with
their
customers.
A
significant
amount
of
pre­
screening
and
vehicle
history
will
also
be
associated
with
vehicle
recruitment.
We
assume
that
with
a
heavy
emphasis
on
existing
customer
relationships
and
data
bases,
recruiting
the
requisite
five
test
vehicles
will
average
about
$
300
per
engine
family.

Prior
to
being
deployed
in
the
field,
each
portable
measurement
system
will
be
carefully
examined
at
the
manufacturer's
facility
to
ensure
proper
operation.
Based
on
our
experience
with
portable
emission
measurement
systems,
we
estimate
that
pre­
calibrating
each
unit
will
require
0.5
and
1.5
hours
of
engineer
and
technician
time,
respectively.
Using
the
total
compensation
rates
previously
described
in
Table1­
4,
this
would
cost
$
60
per
unit.
Since
it
is
assumed
that
testing
will
be
conducted
using
two
portable
systems,
the
total
direct
labor
cost
for
pre­
calibrating
the
instruments
is
$
120
per
engine
family.

The
last
category
of
direct
labor
per
engine
family
is
primarily
for
final
data
analysis
and
quality
assurance
(
beyond
that
which
is
conducted
in
the
field),
reporting
results,
and
archiving
information.
We
assume
that
engine
manufacturers
will
develop
a
number
of
automated
methods
to
perform
many
of
these
functions
to
minimize
labor
requirements.
Our
direct
labor
estimates
are
basically
taken
from
another
EPA
report
that
was
prepared
to
support
a
new
pilot
program
aimed
at
developing
new
in­
use
data
collection
methods
for
nonroad
diesel­
powered
equipment
(
USEPA
2004).
That
program
will
also
collect,
analyze,
and
report
emissions
data
using
portable
emission
measurement
systems.
For
the
purposes
of
this
analysis,
we
doubled
the
time
per
test
for
managerial
oversight,
since
the
original
estimate
was
developed
to
reflect
an
emission
factor
style
program,
while
the
in­
use
program
potential
compliance
implications.

Table
1­
6
presents
the
estimated
labor
hours
for
each
data
analysis
and
reporting
activity,
the
cost
per
test,
and
the
cost
per
engine
family.
The
cost
per
test
is
based
on
a
labor
rate
of
$
32
per
hour
for
an
engineer
and
$
48
per
hour
for
a
manager.
These
labor
classifications
and
Submitted
to
OMB
for
Review
on
05­
05­
2005
compensation
rates
were
previously
discussed
in
Section
1.2.2.2
and
presented
in
Table
1­
3.
The
total
cost
of
post­
data
analysis
and
reporting
is
estimated
to
be
$
751
per
engine
family.

The
resulting
total
direct
labor
for
the
three
categories
described
above
is
$
1,171
per
engine
family.

1.2.2.3.3
Travel
As
discussed
above,
we
assume
the
test
sites
are
located
far
enough
away
from
the
manufacturer's
facility,
or
employees
home
base,
that
a
single
round
trip
to
and
from
the
job
site
and
overnight
travel
is
required.
More
specifically,
we
assume
that
commuting
at
the
beginning
and
end
of
the
work
week
is
four
hours
one
way
for
a
total
of
eight
hours.
Again,
using
the
total
compensation
rates
previously
presented
in
Table
1­
4
of
$
32
and
$
29
per
hour
for
an
engineer
and
technician,
respectively,
the
travel­
related
direct
labor
cost
with
100%
overhead
is
$
976
per
engine
family.

The
vehicle
expenses
consist
of
the
round
trip
travel
to
the
testing
locations
described
above,
and
daily
travel
to
and
from
the
test
sites
as
well
as
itinerant
travel,
e.
g.,
lunch
and
dinner.
The
round
trip
travel
expense
associated
with
the
eight
hours
of
commuting
time
is
estimated
by
using
an
assumed
average
speed
of
50
miles
per
hour
and
a
mileage
fee
of
$
0.40
per
mile.
The
result
is
$
160.
The
itinerant
travel
distances
are
assumed
to
be
30
miles
for
Day
1
and
45
miles
each
day
for
the
next
three
days,
or
a
total
of
165
miles.
Using
the
assumed
vehicle
reimbursement
fee,
this
amounts
to
$
226.

This
scenario
requires
overnight
travel.
There
will
be
approximately
four
days
of
full
per
diem
expenses
and
a
partial
day
of
meals
on
the
fifth
day
(
Table
1­
3).
We
assume
it
costs
$
100
per
night
for
lodging
and
$
40
per
day
for
meals.
This
results
in
per
diem
expenses
of
$
1200
for
the
full
week.
Combining
this
with
the
travel­
related
labor
of
$
976
nd
the
vehicle
cost
of
$
226
from
above,
the
total
travel
cost
is
$
2,402
per
engine
family.

1.2.2.3.4
Labor
Overhead
We
assume
that
all
direct
labor
is
burdened
at
100
percent
of
the
total
compensation
rate.

1.2.2.3.5
General
and
Administrative
Overhead
We
assume
general
and
administrative
expenses
to
be
6.5
percent
of
all
other
costs.

1.2.2.3.6
Summary
of
Variable
Cost
per
Engine
Family
Table
1­
7
summarizes
the
various
cost
elements
discussed
above.
As
shown,
the
resulting
total
variable
costs
per
engine
family
are
$
5,851.
Submitted
to
OMB
for
Review
on
05­
05­
2005
1.3
Costs
of
the
Program
Now
that
the
basic
fixed
and
variable
cost
inputs
have
been
developed,
we
will
use
that
information
to
identify
a
range
of
total
annual
costs
for
the
program.
This
range
reflects
the
testing
scenario,
as
described
in
Section
1.2.2.1,
and
three
different
levels
of
testing
intensity
that
may
occur
under
the
Phase
1
and
2
requirements,
which
are
described
below.
We
will
also
develop
a
single
point
estimate
of
the
program's
annual
cost.
Finally,
we
will
use
this
point
estimate
to
present
total
costs
for
the
first
five
years
of
the
program,
and
costs
over
30
years.
These
costs
are
presented
for
the
entire
industry.

The
first
level
of
testing
intensity
is
the
minimum
number
of
vehicles
that
must
be
tested
under
Phase
1
of
the
program
to
demonstrate
if
a
designated
engine
family
passes
the
NTE
criteria,
i.
e.,
five
vehicle
tests
per
family.
This
is
the
basis
upon
which
the
variable
cost
components
were
developed
in
Section
2.1,
and
is
referred
to
as
Phase
1
minimum.
The
second
level
of
testing
is
the
maximum
number
of
vehicles
that
could
be
required
under
Phase
1,
i.
e.,
10
vehicle
tests
per
family.
This
is
referred
to
as
Phase
1
maximum.
The
third
level
of
testing
is
a
worst
case
where
a
manufacturer
must
complete
Phase
2
testing
for
an
engine
family.
At
its
maximum,
Phase
2
requires
up
to
20
vehicle
tests
per
family.
This
is
referred
to
as
Phase
2
maximum.

Overall,
our
methodology
for
estimating
the
costs
associated
with
the
three
testing
levels
is
simple.
We
assume
that
a
manufacturer
will
complete
each
level
of
testing
in
discrete
steps.
For
example,
after
completing
Phase
1
minimum,
the
test
results
for
the
engine
family
will
be
thoroughly
evaluated
at
the
manufacturer's
technical
center.
If
one
or
more
of
the
vehicles
do
not
pass
the
testing
criteria,
the
manufacturer
is
assumed
to
return
to
the
field
to
continue
testing
five
more
vehicles,
i.
e.,
Phase
1
maximum.
For
the
purposes
of
this
analysis,
this
means
that
the
variable
cost
of
Phase
1
maximum
testing
is
twice
the
cost
of
performing
Phase
1
minimum
testing.
Similarly,
the
variable
cost
of
Phase
2
maximum,
i.
e.,
20
vehicles,
is
twice
the
cost
of
Phase
1
maximum,
i.
e.,
10
vehicles.
The
fixed
cost
of
testing
is
constant
for
each
of
the
three
testing
intensities,
since
the
cost
of
purchasing
the
portable
measurement
systems
does
not
change
with
the
number
of
tests
performed.

1.3.1
Variable
Costs
by
Level
of
Testing
Intensity
As
noted
above,
fixed
costs
do
not
change
by
the
number
of
tests
performed,
although
variable
costs
do
vary
by
testing
intensity.
Therefore,
the
first
step
in
estimating
the
range
of
annual
costs
is
to
determine
variable
cost
per
family
for
each
of
the
testing
levels.
This
is
presented
in
Table
1­
8.

The
next
step
is
to
find
the
range
of
annual
costs
for
all
manufacturers,
i.
e.,
all
engine
families.
Under
the
program,
we
may
generally
select
up
to
25
percent
of
an
engine
manufacturer'
families
for
testing
each
year.
In
the
2005
model
year,
there
were
71
heavy­
duty
diesel
engine
families
certified.
Hence,
we
may
select
up
to
18
engine
families
per
year.
Using
this
value,
the
resulting
range
of
annual
costs
for
all
manufacturers
is
shown
in
Table
1­
9.
Submitted
to
OMB
for
Review
on
05­
05­
2005
1.3.2
Total
Annual
Costs
by
Level
of
Testing
Intensity
Table
1­
10
summarizes
the
fixed
and
variable
cost
estimates
and
presents
the
total
annual
cost
for
each
testing
level.
The
low
end
of
the
range
is
about
$
1.6
million
per
year
for
Phase
1
minimum,
and
the
high
end
of
the
range
is
$
2.1
million
for
Phase
2
maximum.

1.3.3.
Total
Annual
Cost
Point
Estimate
Our
point
estimate
assumes
the
overall
program
will
reflect
the
average
of
weighted
at
80
percent
of
the
Phase
1
minimum
average
cost,
10
percent
of
the
Phase
1
maximum
average
cost,
and
10
percent
of
the
Phase
2
maximum
average
cost.
This
reflects
our
belief
that
most
of
the
engine
families
will
be
designed
and
built
in
full
conformance
with
the
applicable
NTE
standards.
But
also
that
the
program
will
identify
some
level
of
potential
nonconformance.
Table
1­
11
summarizes
the
Phase
1
minimum
and
Phase
2
maximum
costs
from
the
previous
table
for
convenience
and
presents
our
point
estimate
of
the
total
annual
cost
for
all
manufacturers.
The
point
estimate
is
$
1.68
million
per
year.
Submitted
to
OMB
for
Review
on
05­
05­
2005
1.3.4.
Total
Costs
Over
5
and
30
Years
We
developed
an
estimate
of
the
total
program
costs
over
both
5
and
30
years
using
the
annual
point
estimate
costs
from
Table
1­
12
and
a
discount
rate
of
seven
percent
per
annum.
As
shown,
the
5
year
cost
is
about
$
6.89
million
and
the
30
year
cost
is
about
$
20.85
million.
Submitted
to
OMB
for
Review
on
05­
05­
2005
Chapter
1
References
1.
U.
S.
Environmental
Protection
Agency.
2003.
Characterizing
Exhaust
Emissions
from
Light­
Duty
Gasoline
Vehicles
in
the
Kansas
City
Metropolitan
Area.
ERG.
EPA
Contract
Number
GS­
10F­
0036K.
Office
of
Transportation
and
Air
Quality,
Assessment
and
Standards
Division,
Ann
Arbor,
Michigan.
Awarded
February
2004.

2.
Bureau
of
Labor
Statistics.
2003.
Employer
Costs
for
Employee
Compensation­
June
2003.
USDL:
03­
446.
U.
S.
Department
of
Labor,
Washington,
D.
C.

3.
U.
S.
Environmental
Protection
Agency.
2004.
Mobile
Source
Emission
Factors:
Populations,
Usage
and
Emissions
of
Diesel
Nonroad
Equipment
in
EPA
Region
7.
Agency
Form
Number
0619.11,
Supporting
Statement,
Part
A.
Office
of
Transportation
and
Air
Quality.
March
2004.
EPA
Edocket
No.
OAR­
2003­
0225­
0003.
Submitted
to
OMB
for
Review
on
05­
05­
2005
Table
1­
1.
Total
Annualized
Fixed
Costs1
Cost
per
PEMS
Unit
($)
Annualized
cost
of
PEMS
Unit
($)
Annual
Cost
per
Manufacturer
($)
Number
of
Manufacturers
(#)
Total
Annual
Cost
($)

140,000
34,145
102,435
13
1,331,655
1
2004
dollars.
Table
1­
2.
Limited
Vehicle
Access
and
Overnight
Travel
Day
1
Day
2
Day
3
Day
4
Day
5
Activity
Hrs
Labor
Type
Activity
Hrs
Labor
Type
Activity
Hrs
Labor
Type
Activity
Hrs
Labor
Type
Activity
Hrs
Labor
Type
Travel
4
B1
V1,
V2
Warm­
Up
1
T
V4
Set­
to­

Spec
1
T
V5
Set­
to­

Spec
1
T
Travel
4
B
V1,
V22
History
2
E
Shift
Wait
Time
8
B
V4
History
1
E
V
5
History
1
E
V1,
V2
Setto
Spec
2
T
V3
Set­
to­

Spec
1
T
V4
Install
V3
Warm
1.5
B
V5
Install
1.5
B
V1,
V2
Install
3
B
V3
History
1
E
Shift
Wait
Time
8
B
Shift
Wait
Time
8
B
Misc.
Time
1
B
V1,
V2
Data
Acquisition
1
B
V3,
V4
Data
Acquisition
1
B
V5
Remove
.75
B
V1,
V2
Remove
1.5
B
V3,
V4
Remove
1.5
B
V5
Data
Acquisition
.5
B
V3
Install
1.5
B
Misc.
Time
1
B
Misc.
Time
1
B
Misc.
Time
1
B
Totals
10
T
15
T
14
T
13
T
4
T
10
E
14
E
14
E
13
E
4
E
1
T=
Technician,
E=
Engineer,
B=
Both
2
V
=
Vehicle
(
identifier).
Table
1­
3.
Labor
Compensation
Rates
BLS
Category1
In­
Use
Testing
Category
Total
Compensation2
($/
Hour)

Technical3
Engineer
32
Precision
Production,
Craft,
and
Repair4
Technician
29
Executive,
administrative,
and
managerial5
Manager
48
1
BLS
2003,
inflated
by
2.7%.

2
Total
compensation
includes
wages
and
salaries,
paid
leave,
supplemental
pay,
and
insurance.
Rounded
to
the
nearest
whole
dollar.
June
2003
dollars.

3
Table
11,
Private
industry,
goods­
producing
and
service­
producing
industries,
by
occupational
group;
All
workers,
goods­
producing
industries;
White­
collar
occupations;
Professional
specialty
and
technical.

4
Table
11,
Private
industry,
goods­
producing
and
service­
producing
industries,
by
occupational
group;
All
workers,
goods­
producing
industries;
Blue­
collar
occupations;
Precision
production,
craft,
and
repair.

5
Table
11,
Private
industry,
goods­
producing
and
service­
producing
industries,
by
occupational
group;
All
workers,
goods­
producing
industries;
White­
collar
occupations;
Executive,
administrative,
and
managerial.
Table
1­
4.
Direct
Labor
Costs
Per
Vehicle
Test
1
(
Based
on
Phase
1
Minimum
Five
Vehicle
Tests
per
Engine
Family)
2
Labor
Rate
Type
Hours
Per
Family
(
5
tests)
Hourly
Compensation
($
per
hour)
Total
Cost
($/
5
tests)
Cost
Per
Test
($/
test)

Engineer
Technician
Engineer
Technician
Regular
40
40
32
29
2440
488
Overtime2
15
16
48
44
1424
285
Total
­­
­­
­­
­­
3864
773
1
2004
dollars.

2
Based
on
Phase
1
testing
five
vehicles.

3
Overtime
paid
for
work
exceeding
40
hours/
week
at
1.5
times
the
regular
pay
rate.
Table
1­
5.
Variable
Costs
Per
Test
Vehicle1
($/
test)

Scenari
o
Direct
Labor2
Labor
Overhead
3
Other
Direct
Costs4
Voided
Tests5
General
and
Administrativ
e6
Total
773
773
100
165
118
1,928
1
2004
dollars.
2
See
Table
1­
4.
3
100
percent
of
direct
labor.
4
General
supplies,
PEMS
maintenance,
calibration
gases,
FID
fuel,
etc.
5
Assumes
10
percent
of
tests
are
void
(
i.
e.,
0.10
*
(
direct
labor,
labor
overhead,
and
other
direct
costs)).
6
6.5
percent
of
all
costs.
Table
1­
6.
Post­
Test
Data
Analysis
and
Reporting
Variable
Cost
Per
Family1
Activity
Hours/
Test
(
hrs)
2
Cost/
Test
($)
Cost
Per
Engine
Family
($/
5
Tests)

Manager
Engineer
Manager
Engineer
Manager
Engineer
Total
QA
Measurements
0.056
3.0
2.69
96.00
13
480
493
Load
Database
0.056
0.278
2.69
8.90
13
45
58
Analysis,

Write
Report,
Archive
0.056
1.167
2.69
37.34
13
187
200
Total
0.168
4.445
8.07
142.24
40
711
751
1
2003
dollars.

2
See
USEPA
2004.
Table
1­
7.
Summary
of
Variable
Costs
Per
Engine
Family1
($/
family)

Direct
Labor
Labor
Overheard2
Incentive
Travel
General
and
Administrative3
Total
1,171
1,171
750
2402
357
5,851
1
2004
dollars.

2
100%
of
direct
labor.

3
6.5%
of
all
costs.
Table
1­
8.
Summary
of
Variable
Costs
Per
Engine
Family
by
Level
of
Testing
Intensity1
($)

Phase
1
Minimum2
Phase
1
Maximum3
Phase
2
Maximum4
Scenario
Vehicle
Testing5
Engine
Family
Total
Vehicle
Testing5
Engine
Family
Total
Vehicle
Testing5
Engine
Family
Total
9,640
5,851
15,491
19,280
5,851
25,131
38,560
5,851
44,411
1
2004
dollars.

2
Phase
1
minimum
=
5
test
vehicles.

3
Phase
1
maximum
=
10
test
vehicles.

4
Phase
2
maximum
=
20
test
vehicles.

5
Cost
per
vehicle
from
Table
1­
5
multiplied
by
the
number
of
vehicles
tested.
Table
1­
9.
Total
Annual
Variable
Costs
for
All
Manufacturers
by
Level
of
Testing
Intensity1
($)

Scenario
Phase
1
Minimum2
Phase
1
Maximum3
Phase
2
Maximum4
Cost
Per
Engine
Family
#
Families
Per
Year5
Total
Cost
Per
Engine
Family
#
Families
Per
Year
Total
Cost
Per
Engine
Family
#
Families
Per
Year
Total
15,491
18
278,838
25,131
18
452,358
44
18
799
1
2004
dollars.

2
Phase
1
requires
that
a
minimum
of
5
test
vehicles.

3
The
maximum
number
of
vehicles
tested
in
Phase
1
is
10.

4
The
maximum
number
of
vehicles
tested
through
Phase
2
is
20.

5
25%
of
a
total
of
71
engine
families
certified
in
the
2005
model
year.
Table
1­
10.
Total
Annual
Costs
for
All
Manufacturers
by
Level
of
Testing1
Intensity
($
Thousands)

Scenario
Phase
1
Minimum2
Phase
1
Maximum3
Phase
2
Maximum4
Fixed
Cost
Variable
Cost
Total
Fixed
Cost
Variable
Cost
Total
Fixed
Cost
Variable
Cost
Total
1,332
279
1,611
1,332
452
1,784
1,332
799
2,131
1
2004
dollars.
2
Phase
1
minimum
=
5
vehicle
tests.
3
Phase
1
maximum
=
10
vehicle
tests.
4
Phase
2
maximum
=
20
vehicle
tests.
Table
1­
11.
Total
Annual
Cost
Point
Estimate
for
All
Manufacturers1
($
Thousands)

Phase
1
Minimum
Phase
2
Maximum
Phase
2
Maximum
Point
Estimate2
Scenario
Fixed
Cost
Variable
Cost
Total
Fixed
Cost
Variable
Cost
Total
Fixed
Cost
Variable
Cost
Total
Fixed
Cost
Variable
Cost
Total
1,332
279
1,611
1,332
452
1,784
1,332
799
2,131
1,332
348
1,680
1
2004
dollars.

2
Assumes
a
80/
10/
10
split
between
Phase
1
minimum,
Phase
1
maximum,
and
Phase
2
maximum,
respectively.
Table
1­
12.
Total
Program
Cost
Over
5
and
30
Years1
(
Based
on
Point
Cost
Estimate)
(
thousands
of
$)

Years
Annualized
Fixed
Costs
Annual
Variable
Costs
Total
Annual
Costs
2005­
2034
1332
348
1,680
30
Year
NPV
in
2005
16,530
4319
20,849
1st
5
Year
NPV
in
2005
5461
1423
6,888
1
2004
dollars.
