Regulations Requiring Onboard Diagnostic
Systems on 2010 and Later Heavy-Duty
Engines Used in Highway Vehicles Over
14,000 Pounds; Revisions to Onboard
Diagnostic Requirements for Diesel
Highway Vehicles Under 14,000 Pounds
Final Technical Support Document
United States
Environmental Protection
Agency
-------
Regulations Requiring Onboard Diagnostic
Systems on 2010 and Later Heavy-Duty
Engines Used in Highway Vehicles Over
14,000 Pounds; Revisions to Onboard
Diagnostic Requirements for Diesel
Highway Vehicles Under 14,000 Pounds
Final Technical Support Document
Assessment and Standards Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency
v>EPA
United States EPA-420-R-08-019
Environmental Protection _. , „„„
Agency December 2008
-------
Final Technical Support Document; HDOBD Final Rule
Table of Contents
1. Introduction 1
2. Technological Feasibility 1
2.1 Update on ZrO2 NOx Sensor Development 1
2.1.1 Current Technology 1
2.1.2 Future Improvements 4
2.1.3 Heavy-duty NOx Detection for 2010 Technology Engines 5
2.2 Diesel Particulate Filter Monitoring 5
2.2.1 Alternative to a PM Threshold for DPF Monitor 5
2.2.2 Monitoring Area for Alternative Approach 6
2.2.3 Engine Data Supporting Alternative DPF Monitor Approach 7
3. Costs 10
3.1 Cost Analysis for Engines Used in Over 14,000 Pound Applications 11
3.1.1 Variable Costs 11
3.1.2 Fixed Costs 14
3.1.3 Total Costs 34
3.2 Cost Analysis for 8,500 to 14,000 Pound Diesel Applications 35
3.3 Updated 2007/2010 HD Highway Costs Including OBD 41
References 45
-------
Final Technical Support Document; HDOBD Final Rule
List of Tables
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
1. Data from 5.9L Diesel Engine with DPF Modified to Simulate a Failed Part 9
2. Estimated OBD Hardware Costs for Diesel and Gasoline Engines Used in Vehicles Over
14,000 Pounds 12
3. Total OBD Variable Costs for Diesel and Gasoline Engines Used in Vehicles Over 14,000
Pounds 14
4. R&D Costs for OBD Algorithm Development and Application - Diesel Engines for Over
14,000 Pound Applications 16
5. R&D Costs for OBD Algorithm Development and Application - Gasoline Engines for Over
14,000 Pound Applications 19
OBD R&D Test Cell Costs - Diesel Engines for Over 14,000 Pound Applications 21
OBD R&D Test Cell Costs - Gasoline Engines for Over 14,000 Pound Applications 22
OBD R&D Test Cell Demand per Manufacturer - Diesel Engines for Over 14,000 Pound
Applications 23
OBD R&D Test Cell Demand per Manufacturer - Gasoline Engines for Over 14,000 Pound
Applications 23
Summary of OBD R&D Costs - Diesel and Gasoline Engines for Over 14,000 Pound
10
Applications
13.
14.
24
11. Cost for OBD Certification Demonstration Limit Parts - Diesel Engines for Over 14,000
Pound Applications 26
12. OBD Certification and Production Evaluation Testing Costs - Diesel Engines for Over
14,000 Pound Applications 28
OBD Certification and Production Evaluation Testing Costs - Gasoline Engines for Over
14,000 Pound Applications
Total OBD Fixed Costs - Diesel and Gasoline Engines for Over 14,000 Pound
Applications
29
33
15. Total Estimated OBD Costs - Diesel and Gasoline Engines for Over 14,000 Pound
34
35
37
38
Applications
16. Total Estimated OBD Costs per Engine for Over 14,000 Pound Applications
17. R&D Costs for OBD Algorithm Development and Application -
18. OBD R&D Test Cell Costs - Diesel Applications Under 14,000 Pounds
19. OBD Certification and Production Evaluation Testing Costs - Diesel Applications Under
14,000 Pounds 40
20. Total Estimated OBD Costs - Diesel Applications Under 14,000 Pounds 41
21. Costs of the 2007/2010 Heavy-duty Highway Program 42
22. Updated 2007/2010 Program Costs Including New OBD-Related Costs 43
23. Consumer Price Index Data 44
-------
Final Technical Support Document; HDOBD Final Rule
1. Introduction
This document contains technical details in support of the final requirements for onboard
diagnostic (OBD) systems on highway applications over 14,000 pounds. The details of these
requirements are not covered in this document and can be found in the preamble to the final
regulations contained in the docket for the rule.1
The details presented in this document support statements in the technological feasibility
and costs sections of the preamble for this rule. As such, this document is broken into two
sections: technological feasibility and costs. Note that many of our technological feasibility
arguments are presented in the preamble and the Summary and Analysis of Comments
document which are also contained in the docket for this rule. The preamble to the rule
contains only a brief summary of our cost estimates while the details behind our cost
estimates are presented here.
2. Technological Feasibility
2.1 Update on ZrO2 NOx Sensor Development
2.1.1 Current Technology
a. Manufacturers
Zirconium Oxide NOX sensors have been developed to measure modal NOX emissions
from lean burn engines. Currently there are three companies that are selling these devices.
They are as follows: NGK Automotive Ceramics, lonotec, and Ceramatec.
b. Measurement Principle
Typical NOX sensor design consists of two internal cavities and three oxygen pumping
cells designed to measure both oxygen (air to fuel ratio measurement) and NOX
concentrations. The most common commercial sensor used today is based on zirconia (ZrC^
partly or fully stabilized with ytteria (Y2O3). The presence of oxygen vacancies in the
material makes the mobility of the oxygen ion 02" possible. The resulting conductivity is
very low at room temperatures, but reaches values of a wet electrolyte when the sensor is
heated up to < 600°C. An oxygen sensor can be constructed if the solid electrolyte is
provided with porous electrodes separating two gas chambers. At higher cell temperatures
the solid electrolyte conducts oxygen ions, thus an oxygen concentration difference between
the two chambers results in a voltage signal. The half cell reactions are as follows:
-------
Final Technical Support Document; HDOBD Final Rule
4e + O
2O
2-
2(test) "
4e" + O
2(ref)
This voltage signal is described in a very good agreement by the Nernst equation:
E =
'RT
In
P,
°2(re/)
) \Po.
Where R is the ideal gas constant, T is absolute temperature, F is the Faraday constant,
PO2(ref)is the partial pressure of the reference gas and PO2(test)is the partial pressure of the
sample gas.
In general, the measurement concept consists of:
1) Lowering the oxygen concentration of a measuring gas to a predetermined level in the
first internal cavity, in which NOX does not decompose, and
2) Further lowering the oxygen concentration of the measuring gas to a predetermined
level in the second internal cavity, in which NOX decomposes on a measuring
electrode and the oxygen generated is detected as a sensor signal.
Figure 3 shows a cross-sectional view of the NOX sensor element. Each part in the
sensing element functions as follows:
First Internal Cavity
The first internal cavity connects a measuring gas stream through the first diffusion path
under a predetermined diffusion resistance. There is an oxygen pumping cell and an oxygen
sensing cell inside the first internal cavity.
The first oxygen pumping cell consists of a pair of first pumping (+) and (-) electrodes
First pumping
firs
st d
d
{
i
! H
1
t internal cavi
T
i-
l
\
^
'First pumping / S"\ t
I X J^7
\ /
N
aater WQ
/ /
id d if fusion xith
S«j;md internal cavity
^ f ,*,..n. ^punir.ing *^. n.*> >
I sJ -i
' 1 7
1
m¥V^"
1
i
"t
|
|_
Kyjr
./40Q
BTl
\
s
mV
_^^
Air Diet —
-f
XD . , , , '
KQfiQrwnc© siK-ifxxi©
•
400 m¥
Figure 3. Cross-sectional view of NOX sensing element.2
-------
Final Technical Support Document; HDOBD Final Rule
on the ZrC>2-l layer, in order to lower the oxygen concentration to a predetermined level.
The first pumping electrode (+) is platinum and the (-) electrode is a platinum/gold alloy to
reduce NOX reduction catalytic activity.
The oxygen sensing cell consists of the first pumping (-) electrode in the first internal
cavity and a reference electrode in an air duct. This allows monitoring of the oxygen
concentration in the first internal cavity by generated electromotive force and feedback to the
first oxygen pumping cell.
Second Internal Cavity
The second internal cavity connects to the first internal cavity through the second
diffusion path under a predetermined diffusion resistance. There are two different oxygen
pumping cells and an oxygen sensing cell inside the second internal cavity.
The second oxygen pumping cell consists of the second pumping (-) electrode in the
second internal cavity and the first pumping (+) electrode on the ZrCVl layer, in order to
further lower the oxygen concentration to a predetermined level. The second pumping (-)
electrode is also made of a platinum/gold alloy.
The oxygen sensing cell consists of the second pumping (-) electrode and the reference
electrode in the air duct to monitor the oxygen concentration in the second internal cavity by
generated electromotive force and feedback to the second oxygen pumping cell.
The NOX sensing cell consists of a measuring electrode in the second internal cavity and
the reference electrode in the air duct. The measuring electrode is rhodium and has a NOX
reduction catalytic activity. Therefore, NOX decomposes on the measuring electrode and the
oxygen generated is detected as an oxygen pumping current in the NOX sensing cell. The
sensor signal is in proportion to the NOX concentration in the measuring gas.2
c. Measurement Range
ZrC>2 NOX sensors are currently available in the 0 - 500 ppm, 0 - 1500 ppm, and 0 -
2000 ppm range. Reported accuracy is in the ± 10% range for readings in the 100 to 2000
ppm range and ±10 ppm for readings in the 0 to 100 ppm range.
d. Interference
ZrC>2 NOX sensor interference has been limited to ammonia (NHa). Sensitivity to
has been shown to be up to 65% of the amount of NHs present in the sample gas. This
is converted to NOX in the internal cavities of the sensor and then measured.3 This
phenomenon may only plague urea SCR applications, where over dosing of urea could lead
to NHs slip. In addition, urea SCR feedback control studies have shown that the NHa
interference signal is discernable from the NOX signal and can, in effect, allow the design of a
better feedback control loop than a NOX sensor that doesn't have any NHs cross-sensitivity.
-------
Final Technical Support Document; HDOBD Final Rule
The signal conditioning method developed, resulted in a linear output for both NH3 and NOX
from the NOX sensor downstream of the catalyst.3
e. Durability
Durability data for diesel applications is limited. NGK has reported data for 1000 hours
of testing (60,000 mile equivalent) on a 2.5 L diesel engine. This data showed that the aged
sensor achieved ± 20 ppm (or ± 7% measurement error) NOX accuracy for a 300 ppm NOX
sample on a 0 to 2000 ppm range sensor. This is almost equivalent to the accuracy of the
fresh sensor in this concentration range.4
Twenty-five NGK NOX sensors in the 0 to 2000 ppm range have undergone 6,000 hours
of aging on a 12 L Caterpillar C-12 engine. Five of these sensors are in the engine out
location, 10 are located downstream of the DPF and upstream of the SCR catalyst, and 10 are
located downstream of the clean-up catalyst. NOX sensors are compared every 1,000 hours
and are independently calibrated every 2,000 hours.
Typical sensor NOX exposure varies by location. On average, the 15 sensors located
upstream of the SCR catalyst were exposed to NOX concentrations in the 100 to 600 ppm
range. This is close to the expected range of engine out exhaust emissions for a 2010 engine,
but the range maximum is on the low side. The 10 sensors located downstream of the
cleanup catalyst were exposed to NOX concentrations in the 10 to 200 ppm range. For testing
out to 2,000 hours, of the pre-catalyst sensors, 12 degraded by 3 to 4%, while the remaining
three degraded by 5 to 7%. Of the post NOX catalyst sensors, all 10 had minimal degradation.
For those sensors that degraded a similar amount, degradation was linear.
For testing out to 4,000 hours, of the pre-catalyst sensors, 11 degraded by 4 to 6%, while
the remaining four degraded by 7 to 8%. Of the post NOX catalyst sensors, 9 had minimal
degradation while one degraded 7%. For those sensors that degraded a similar amount,
degradation was linear. Overall relative error ranged from 8% at engine-out concentrations
to minimal degradation at lower concentrations.5
For testing out to 6,000 hours, of the pre-catalyst sensors, 5 degraded by 5 to 6%, while
the remaining ten degraded by 7 to 12%. Of the post NOX catalyst sensors, 8 had minimal
degradation while two degraded 10%. For those sensors that degraded a similar amount,
degradation was linear. Overall relative error ranged from 10% at engine-out concentrations
to minimal degradation at lower concentrations.6
2.1.2 Future Improvements
As with any maturing technologies, it is expected that improvements will be made to
sensor accuracy and durability in the near future. Requests by engine manufacturers have
been made to instrument manufacturers to develop sensors that have improved accuracy in
the 0 to 100 ppm range. Instrument manufacturers are complying with these requests and it
-------
Final Technical Support Document; HDOBD Final Rule
is expected that NOX sensors in the 0 to 100 ppm range with a zero hour ± 5 ppm accuracy
and aged ±10 ppm accuracy will be available by the middle of 2010.
2.1.3 Heavy-duty NOx Detection for 2010 Technology Engines
a. Future NOx Emission Levels
It is expected thatNOx concentrations downstream of an emission control system on an
engine meeting the 2010 NOX standard will be in the 0 to 50 ppm range, on average,
depending on engine speed, load, and the state of the emission control system (ECS).
As an example, a 5.9 L Cummins ISB meeting the 2010 NOX standard for the FTP (0.13
g/hp-hr) and SET (0.12 g/hp-hr) using a NOX adsorber based ECS will have average NOX
emissions ranging from 0 to 60 ppm.7 Data from the APBF-DEC Heavy-Duty NOX
Adsorber/DPF Project: Heavy Duty Linehaul Platform reported NOX emissions downstream
of the ECS in the range of 0 to 200 ppm for an engine emitting NOX in the range of 0.05 to
0.5 g/hp-hr NOX over 2000 hours.8 It is important to note that the average NOX emissions are
less than 40 ppm for this engine and ECS. Therefore it is important to note that NOX spikes
larger than the average will have to be dealt with accordingly by the OBD system.
b. Current NOx Sensor Detection Limits
Current NOX sensors have a stated accuracy of ± 10 ppm in the zero to 100 ppm range
for a 0 to 2000 ppm range. Accuracies for some sensors have been reported as high as ± 30
ppm. With this in mind, current NOX sensor technology should be able detect NOX emissions
that exceed the standard by 2 to 3 times the 2010 limit.
c. Future NOx Sensor Detection Limits
If NOX sensor manufacturers are able to develop the proposed 0 to 100 ppm range sensor
with ± 5 ppm accuracy, it should be possible to accurately measure emissions increases as
low as 1.5 times the 2010 NOX emission standard. With sensor development underway, this
sensor should be available by early to mid 2006 for evaluation.
2.2 Diesel Particulate Filter Monitoring
2.2.1 Alternative to a PM Threshold for DPF Monitor
Given that sensors which can directly measure exhaust PM will not be available for
commercial sale in 2010, an alternative to a PM threshold monitor is provided. This
alternative is a differential pressure (or delta pressure) sensor-based approach which can be
used to determine the inherent functionality of the DPF. When the engine is operated under
moderate-to-high speed and load conditions, the exhaust flowing through a DPF creates a
pressure drop across the device. This pressure drop is measured by a differential pressure
-------
Final Technical Support Document; HDOBD Final Rule
(dP) sensor, and is influenced by exhaust gas conditions (mass flow rate and temperature),
the level of soot loading on the filter, and the structural integrity of the filter (e.g. cracks
and/or substrate damage which would allow exhaust gasses to bypass the filter media). By
comparing the dP observed whenever the engine is operated in the monitored region to the
expected dP for a nominal, clean filter under similar operating conditions, a judgment can be
made as to whether the DPF is performing properly (i.e. exhaust gas is flowing through the
filter media, not bypassing it). If a dP value is observed which is lower than the expected dP
value minus some detectable change in the dP signal, it is likely that some failure of the DPF
has occurred and PM emissions have increased. The "detectable change in dP" that must be
detected is determined by running the engine at a single mode on the Supplemental
Emissions Test (SET), recording the dP value with a clean DPF, and multiplying this dP
value by 0.5; the result of this calculation becomes the detectable change in dP value. For a
given operating point within the monitored region, the observed dP is continuously compared
to the result of the "expected dP" minus the "detectable change in dP"; if the observed dP
drops below the result of this calculation for more than 5 seconds, a malfunction is present
and a DTC would be stored.
2.2.2 Monitoring Area for Alternative Approach
At lower engine speed and load conditions (which result in lower exhaust mass flow, and
hence, a lower dP across the DPF), the observed dP is low (less than 4 kPa in our testing on a
5.9L engine, see Table 1), and if the sensitivity of a dP sensor is +/- IkPa, the system is
unable to reliably distinguish between good DPF and a damaged DPF with lower flow
restriction. Since the dP sensor-based monitor would not be reliable under conditions where
that exhaust mass flow is low, the monitored area will defined using SET modes where
significant engine air flow conditions exist and a meaningful dP signal is available. We
believe that DPF failures can be accurately and reliably identified using this approach.
Figure 4 illustrates an example of the engine speed and load operating conditions under
which this alternative DPF monitor will function and how the enable conditions for the
monitor are defined using SET modes A100, A75, B50, and C50.
-------
Final Technical Support Document; HDOBD Final Rule
1000
900 -
800
700
- 600
500
400
300
200
100
S)
Monitors 1 Area
800
1000
1200 1400 1600 1800 2000
Engine Speed (RPM)
2200
2400
2600
2800
•Torque A SET modes ™ O DPF Monitoring Area
Figure 4. Example of Engine Speed-Load Area Where Alternative DPF Monitor Is
Active
2.2.3 Engine Data Supporting Alternative DPF Monitor Approach
A Cummins 5.9L ISB engine was used to determine the affect that a simulated DPF
failure would have on the dP sensor output. The DPF used for this test was a 200 cell per-
square-inch, wall-flow design, 9 inches in diameter and 11 inches in length. To simulate a
"failed" part (e.g.. one in which a portion of the rear face and volume is missing due to
excessive temperature excursions during an un-controlled regeneration event), a 5 inch hole
was machined in the rear face of the filter to a depth of 5.5 inches (50% of the filter length).
In addition, the channel plugs on half of the remaining rear face were machined off to
increase the percentage of channels that were "open." In total, 18% of the filter volume was
removed and 33% of the filter face area had open channels. The photo in Figure 5 shows the
rear face of the modified DPF.
The baseline case was run using a clean and completely intact DPF (i.e., the part before
modification to simulate a "failed" part) and the engine was operated at the SET modes,
where engine and exhaust data were recorded. The modified or "failed" DPF was then
installed, run under the same test conditions, and the results are summarized in Table 1.
Using the "B50" mode of the SET on the baseline test, a "detectable change in dP" value of
1.7 kPa was established (i.e. dP @ B50 = 3.38 kPa; 3.38*0.5 = 1.7). Subtracting this 1.7 kPa
-------
Final Technical Support Document; HDOBD Final Rule
detectable dP change from the dP observed at each mode in the baseline test, an observed dP
value at which a malfunction can be detected is established. As the test data show, for the
particular DPF failure that was simulated, a malfunction could be detected if the engine were
operated at or above the "B75" engine speed and load conditions. At test conditions where
the malfunction would have been detected (i.e. B75, B100, C75, and C100), the post-DPF
PM levels were 1.8-to-3.1 times the 0.01 g/bhp»hr PM emission standard (i.e., below our
threshold of the PM standard + 0.04). For DPF failures which result in a larger change in the
dP signal (e.g. more of the DPF volume is missing and/or more channels are open), the
engine speeds and loads at which a malfunction can be detected become lower as well.
Figure 5. Rear Face of DPF Modified to Simulate Missing Substrate Due to Melting
During Un-Controlled Regeneration
8
-------
Final Technical Support Document; HDOBD Final Rule
Table 1. Data from 5.9L Diesel Engine with DPF Modified to Simulate a Failed Part
SET Mode
A75 - baseline
A75 - modified
A100 - baseline
A100 - modified
B50 - baseline
B50 - modified
B75 - baseline
B75 - modified
B100 - baseline
B100 - modified
C50 - baseline
C50 - modified
C75 - baseline
C75 - modified
C1 00 -baseline
C1 00 -modified
Engine Speed
(RPM)
1619
1619
1947
1947
1947
2275
2275
2275
Torque
(N-m)
675
877
445
668
876
407
610
811
Mass exh flow
(kg/hr)
551.09
555.07
677.36
685.11
552.07
555.96
721.67
723.36
862.25
864.30
673.30
675.87
862.27
868.35
991.89
992.28
EGT
(°C)
453
450
484
481
375
375
427
421
457
460
n/a
353
414
n/a
476
477
PM-DPF-ln
(g/bhp'hr)
n/a
n/a
n/a
n/a
0.047
0.047
0.032
0.032
0.039
0.039
n/a
n/a
0.077
0.077
0.044
0.044
PM-DPF-Out
(g/bhp'hr)
0.002
0.011
0.005
0.021
0.001
0.028
0.002
0.018
0.001
0.022
n/a
0.021
0.002
n/a
0.001
0.031
DPFdP
(kPa)
3.91
2.46
5.41
3.95
3.38
2.14
5.29
3.25
6.78
4.28
4.21
2.60
6.17
4.13
8.09
5.53
Calculated
Change in dP
(base - mod)
1.5
1.5
1.2
2.0
2.5
1.6
2.0
2.6
Fault Detected ?
(change in dP >
dP©B50*0.5)
No
No
No
/es
/es
No
/es
Ves
-------
Final Technical Support Document; HDOBD Final Rule
3. Costs
This section provides the details behind the cost analysis done in support of our
over!4,000 pound OBD program and our changes to the existing under!4,000 pound heavy-
duty diesel OBD requirements. Details associated with the new requirements and changes to
existing requirements can be found in the preamble to the rulemaking and are not presented
here. As a result, there may be details within this report that can be understood only by
reading the associated preamble for the rulemaking.
The final cost analysis differs from the draft cost analysis (i.e., the analysis done in
support of the proposed rule) in three ways. First, we have included costs for aging limit
parts to their OBD thresholds. We inadvertently did not include those costs in the draft
analysis. Discussion of this can be found in the Summary and Analysis of Comments
document in Section VLB. These newly added costs are presented in Section 3.1.2.b of this
document. Second, while in the proposal we estimated lower warranty costs beginning in
2013, we have delayed that until 2016 in the final rule. This is discussed in Section VIA of
the Summary and Analysis of Comments document and in Section 3.1.1 of this document.
Third, we have adjusted all costs to 2007 dollars - the draft analysis used 2004 dollars - by
using the Consumer Price Index. As a result, all costs presented here are slightly higher than
in the draft analysis although we have not changed the analysis with the exception of this
adjustment for inflation and, as mentioned previously, the addition of costs for aging of limit
parts and delay of lower warranty costs.
This analysis breaks estimated costs into two primary categories: variable costs and fixed
costs. Variable costs are those costs associated with any new hardware required to meet the
requirements, the associated assembly time to install that hardware, and any increased
warranty costs associated with the new hardware. Variable costs are additionally marked up
to account for both manufacturer and dealer overhead and carrying costs. The
manufacturer's carrying cost was estimated to be four percent of the direct costs to account
for the capital cost of the extra inventory and the incremental costs of insurance, handling,
and storage. The dealer's carrying cost was estimated to be three percent of their direct costs
to account for the cost of capital tied up in inventory. We adopted this same approach to
markups in the heavy-duty 2007/2010 rule and our more recent Nonroad Tier 4 rule based on
industry input.9
Fixed costs considered here are those for research and development (R&D), certification,
and production evaluation testing. The fixed costs for engine R&D are estimated to be
incurred over the four-year period preceding introduction of the engine. The fixed costs for
certification include costs associated with demonstration testing of OBD parent engines
including the "limit" parts used to demonstrate detection of malfunctions at or near the
applicable OBD thresholds. The demonstration testing costs are estimated to be incurred one
year preceding introduction of the engine while the production evaluation testing is estimated
to occur in the same year as introduction. Importantly, none of the fixed costs estimated here
10
-------
Final Technical Support Document; HDOBD Final Rule
consider the recent California Air Resources Board approved requirements for over 14,000
pound OBD.10
We present all of these costs in the year during which we estimate they will be incurred
by manufacturers over the 30 year time period following publication of the final rule. We
then calculate a 30 year net present value of those cost streams using both a three percent and
a seven percent discount rate to reflect the time value of money at both ends of the most
likely range.
We present all costs in 2007 dollars. We refer to both near term costs and long term costs.
The near term costs represent those costs when warranty costs are estimated to be the highest.
The long term costs consider the effects of a reduction in warranty costs. For warranty costs,
we have estimated a three percent near term rate for warranty claims and a one percent long
term rate for warranty claims.
3.1 Cost Analysis for Engines Used in Over 14,000 Pound Applications
3.1.1 Variable Costs
The variable costs we have estimated represent those costs associated with various
sensors that we believe would have to be added to the engine to provide the required OBD
monitoring capability. Our variable (i.e., hardware) cost estimates are summarized in Table
2.
11
-------
Final Technical Support Document; HDOBD Final Rule
Table 2. Estimated OBD Hardware Costs for Diesel and Gasoline Engines Used in Vehicles Over 14,000
Pounds
(2007 dollars)
Diesel
201 0-201 2 Model Year
New Hardware
ECU upgrade
Purge solenoid for evap leak check
Pressure sensor for evap leak check
Subtotal
Assembly labor (hours)
Assembly labor cost
Assembly labor overhead at 40%
Cost to Mfr
Warranty cost - near term at 3% claim rate
Mfr. Carrying cost at 4%
Cost to Buyer -201 0-201 2
201 3+ Model Year
New Hardware
MIL and wiring
Subtotal (2010+2013)
Assembly labor (hours)
Assembly labor cost
Assembly labor overhead at 40%
Cost to Mfr
Warranty cost - long term at 3% claim rate
Mfr carrying cost at 4%
Cost to Buyer -201 3-201 5
2016+ Model Year
Cost to Mfr
Warranty cost - long term at 1% claim rate
Mfr carrying cost at 4%
Cost to Buyer — long term
$ 33
$
$
$ 33
0.10
$ 3
$ 1
$ 38
$ 4
$ 2
$ 43
$ 11
$ 44
0.20
$ 7
$ 3
$ 53
$c;
O
$0
£
$ 60
$ 53
$ 2
$ 2
$ 57
Gasoline
$ 11
$ 11
$ 11
$ 33
0.30
$ 10
$ 4
$ 47
$ 4
$ 2
$ 53
$ 11
$ 44
0.40
$ 13
$ 5
$ 62
$C
-J
$ 2
$ 70
$ 62
$O
^
$ 2
$ 66
For the 2010 model year, we believe that both diesel and gasoline engines would have to
upgrade their engine control computers, or engine control units, to accommodate the
increased computing capacity required for the proposed OBD. We have estimated this cost
at $33 per engine for diesel engines and $11 for gasoline engines, inclusive of supplier
markup. We have estimated a different cost because we believe that the gasoline engines are
using computers similar, if not in fact identical to, their under 14,000 pound counterparts.
Therefore, those computer upgrades should cost little, if anything. For diesel engines, we
believe that the OBD requirements will result in a more substantial upgrade to existing
computers. Also for the 2010 model year, we believe that gasoline engines would have to
add both a purge solenoid and a pressure sensor for the evaporative system monitoring
requirement. We have estimated the cost of both of these items at $11 a piece inclusive of
supplier markup. We believe that the other sensors needed by the OBD system on both
diesel and gasoline engines will already be on the engines for either emissions control and/or
protection of the engine (e.g., temperature sensors used to protect against condensation
formation caused by overcooling of the EGR gases—engine protection—can also be used to
monitor the effectiveness of the EGR cooler—OBD). The result is a manufacturer cost
subtotal of $33 for both diesel and gasoline engines in the 2010 model year. Note that we
12
-------
Final Technical Support Document; HDOBD Final Rule
have not included costs for a malfunction indicator light (MIL) and associated wiring in the
2010 timeframe since we are not requiring a dedicated MIL until the 2013 model year.
We have estimated that adding these sensors and actuators will require increased
assembly time. We have estimated these times at one-tenth of an hour for diesel engines and
one-third of an hour for gasoline engines (i.e., six minutes for each newly added part). We
have estimated a labor rate of $33 per hour for this assembly along with overhead at 40
percent. This results in an estimated cost to the manufacturer of $38 and $47 for diesel and
gasoline engines, respectively, in the 2010 model year.
We have included a warranty cost recovery estimating a three percent warranty claim rate
in the near term. We have also included a four percent manufacturer carrying cost to cover
increased insurance and inventory costs incurred by the manufacturer.11 Including these
costs results in an end cost to the buyer of roughly $43 and $53 for diesel and gasoline
engines, respectively, in the 2010 model year.
For the 2013 model year, we have included costs associated with the dedicated MIL and
its wiring. These costs were estimated at $11 per engine inclusive of supplier markup.
Following the same process for assembly costs (another one-tenth of an hour per engine),
warranty costs (three percent claim rate since many engines will be complying with the rule
beginning in 2013), and carrying costs, we have estimated the 2013 model year hardware
cost to the buyer at roughly $60 and $70 for diesel and gasoline engines, respectively.
For the 2016 model year, we have reduced the costs associated with warranty by
estimating a one percent claim rate since all engines will have complied with the rule for
several years. Including the carrying costs, we have estimated the 2016 and later, or long-
term, hardware cost to the buyer at $57 and $66 for diesel and gasoline engines, respectively.
To determine the fleetwide estimated hardware costs, or total variable costs, we looked at
the projected over 14,000 pound sales data from our 2004 model year certification database
which showed projected US sales less projected California sales of 614,500 for diesel
engines and 39,400 for gasoline engines. In the 2010 through 2012 model years, we
estimated 50 percent of engines would comply with the proposed OBD requirements based
on our proposed phase-in schedule. For model years 2013 and later, we will have 100
percent compliance. Applying the estimated hardware costs presented in Table 2 to the
appropriate projected sales in each model year through 2035, estimating a two percent
growth in sales based on 2004 sales, results in a 30 year net present value (NPV) cost of $620
million and $47 million for diesel and gasoline engines, respectively, using a three percent
discount rate. These costs, including a NPV at a seven percent rate, are shown in Table 3.
13
-------
Final Technical Support Document; HDOBD Final Rule
Table 3. Total OBD Variable Costs for Diesel and Gasoline Engines Used in Vehicles Over 14,000
Pounds
(2007 dollars)
Year
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
NPV@
NPV@
CY
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
3%
7%
Diesel
Projected Sales|
639,103
651,393
663,684
675,974
688,265
700,555
712,846
725,136
737,426
749,717
762,007
774,298
786,588
798,879
811,169
823,459
835,750
848,040
860,331
872,621
884,912
897,202
909,493
921,783
934,073
946,364
958,654
970,945
983,235
995,526
$/engine|
$43
$43
$43
$60
$60
$60
$57
$57
$57
$57
$57
$57
$57
$57
$57
$57
$57
$57
$57
$57
$57
$57
$57
$57
$57
$57
% complying]
50%
50%
50%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
Subtotal
$14,852,000
$15,117,000
$15,382,000
$43,646,000
$44,385,000
$45,125,000
$43,356,000
$44,055,000
$44,754,000
$45,454,000
$46,153,000
$46,852,000
$47,551,000
$48,251,000
$48,950,000
$49,649,000
$50,349,000
$51,048,000
$51,747,000
$52,446,000
$53,146,000
$53,845,000
$54,544,000
$55,244,000
$55,943,000
$56,642,000
$685,964,000
$363,769,000
Gasoline
Projected Sales]
40,976
41,764
42,552
43,340
44,128
44,916
45,704
46,492
47,280
48,068
48,856
49,644
50,432
51,220
52,008
52,796
53,584
54,372
55,160
55,948
56,736
57,524
58,312
59,100
59,888
60,676
61,464
62,252
63,040
63,828
$/engine|
$53
$53
$53
$70
$70
$70
$66
$66
$66
$66
$66
$66
$66
$66
$66
$66
$66
$66
$66
$66
$66
$66
$66
$66
$66
$66
/o complying!
50%
50%
50%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
Subtotal
$1,164,000
$1,185,000
$1,205,000
$3,244,000
$3,299,000
$3,354,000
$3,248,000
$3,301,000
$3,353,000
$3,405,000
$3,458,000
$3,510,000
$3,563,000
$3,615,000
$3,667,000
$3,720,000
$3,772,000
$3,825,000
$3,877,000
$3,929,000
$3,982,000
$4,034,000
$4,086,000
$4,139,000
$4,191,000
$4,244,000
$51,463,000
$27,315,000
Total Hardware
Costs
$16,016,000
$16,302,000
$16,587,000
$46,890,000
$47,684,000
$48,479,000
$46,604,000
$47,356,000
$48,107,000
$48,859,000
$49,611,000
$50,362,000
$51,114,000
$51,866,000
$52,617,000
$53,369,000
$54,121,000
$54,873,000
$55,624,000
$56,375,000
$57,128,000
$57,879,000
$58,630,000
$59,383,000
$60,134,000
$60,886,000
$737,427,000
$391,084,000
3.1.2 Fixed Costs
We have estimated fixed costs for research and development (R&D), certification, and
production evaluation testing. The R&D costs include the costs to develop the computer
algorithms required to diagnose engine and emission control systems, and the costs for
applying the developed algorithms to each engine family and to each variant within each
engine family. The certification costs include the costs associated with testing of durability
data vehicles (i.e., the OBD parent engines), the costs associated with generating the "limit"
parts that are required to demonstrate OBD detection at or near the applicable emissions
thresholds, and the costs associated with generating the necessary certification
documentation. Production evaluation testing costs consist of the costs associated with the
three different elements of production evaluation testing.
a. Research & Development Costs
We have broken the estimated R&D costs into three separate categories. The first of
these is the cost for developing computer controlled diagnostic algorithms. These costs are
estimated to be incurred once per manufacturer since once an algorithm is developed, it can,
practically speaking, be used over and over again with only minor changes, if any, to
improve upon the original. The second R&D cost is that for applying the manufacturer's
developed algorithm to each of its engine families. Each engine family may have a different
14
-------
Final Technical Support Document; HDOBD Final Rule
number of cylinders or different emissions control architecture (e.g., different combinations
of aftertreatment devices) and the algorithm may have to be adapted for each of these engine
families. Consequently, this cost is estimated to be incurred once for each of the engine
families expected to be sold. The third R&D cost is that for applying the algorithm that has
been adapted for each engine family to every variant within each engine family. Variants
within engine families have different horsepower and/or torque characteristics and, therefore,
the adapted algorithm would have to be fine tuned to each of the engine family's variants.
These costs are estimated to be incurred once for each of the remaining variants within each
family (i.e., one variant will use the adapted algorithm while the remaining variants will
require further fine tuning).
We have estimated separate development and separate application costs for the different
types of monitors—system monitors, rationality monitors, and comprehensive component
monitors. System monitors are generally the most difficult monitors and for the most part are
those monitors for which an emissions threshold exists. Nonetheless, most system monitors
are not correlated to an emissions threshold and are, instead, functional monitors that can
detect a malfunctioning component prior to emissions exceeding the applicable thresholds.
For such monitors, manufacturers generally forego the more costly emissions correlation
work and rely on the functional check alone which saves both time and money.
We have estimated that an engineer and a technician would be involved in most of the
development work since much of the work will entail testing on an engine test bed. We have
estimated that an engineer costs $110,000 a year while a technician costs $66,000 a year, and
that they each work 48 forty hour weeks per year. Table 4 shows these R&D costs for diesel
engines. The total costs shown represent industry totals for ten manufacturers.
15
-------
Final Technical Support Document; HDOBD Final Rule
Table 4. R&D Costs for OBD Algorithm Development and Application - Diesel Engines for Over 14,000
Pound Applications
(2007 dollars)
Fixed Costs - Diesel
A. Algorithm Development Costs
System Threshold Monitors
Engineers
Technician $
Subtotal
System Functional Monitors
Engineers
Technician $
Subtotal
CCM Rationality Monitors
Engineers
Technician $
Subtotal
CCM Continuity Monitors
Engineers
Technician $
Subtotal
Total
| weeks/monitor]
30
15
20
5
15
1
2
0
Cost/monitor] # of monitors] Total/Mfr
Total
$69,000
$21,000
$90,000 9 $810,000 $8,100,000
$46,000
$7,000
$53,000 41 $2,173,000 $21,730,000
$34,000
$1,000
$35,000 50 $1,750,000 $17,500,000
$5,000
$0
$5,000 80 $400,000 $4,000,000
S5,133,000 S51, 330,000
B. Application Costs to each Family
System Threshold Monitors
Engineers
Technician $
Subtotal
System Functional Monitors
Engineers
Technician $
Subtotal
CCM Rationality Monitors
Engineers
Technician $
Subtotal
Total
[weeks/monitor |
5
10
5
10
3
1
Cost/monitorl # of monitors| Total/Familyl #familie
$11,000
$14,000
$25,000 9 $225,000
$11,000
$14,000
$25,000 41 $1,025,000
$7,000
$1,000
$8,000 50 $400,000
$1,650,000
s/mfr Total/Mfr| Total |
6.5 $1,463,000 $14,630,000
6.5 $6,663,000 $66,630,000
6.5 $2,600,000 $26,000,000
$10,726,000 $107,260,000
C. Application Costs to remaining Variants
Total
I I
| TotalA/ariant| #variants/family| #familie
$413,000 4
s/mfr Total/Mfr| Total |
6.5 $10,738,000 $107,380,000
For diesel engines, using industry input and our own engineering analysis, we have
estimated that there will be roughly 50 system monitors. Of these, we treated 9 as threshold
monitors with the remainders being functional monitors.21 Based on industry input, we have
also estimated that there will be an additional 50 rationality monitors and 80 circuit
continuity monitors.
a The 9 threshold monitors for diesel engines, based on our engineering judgment, would be: fuel system
pressure high; fuel system injection timing too advanced; fuel system injection timing too retarded; EGR low
flow; EGR slow response; EGR low cooling; variable valve timing (WT) above target; WT below target;
WT slow response; NOx catalyst system conversion; NOx catalyst system reductant delivery; NOx adsorber
performance; DPF filtering performance; NOx sensor slow response; and, NOx sensor offset. We have
estimated that 50 percent of engines would do threshold monitoring for fuel systems (based on changes in the
final rule to the electronic unit injector provisions). Similarly, we have estimated 50 percent of engines to be
SCR equipped with 50 percent being NOx adsorber equipped. We have also estimated 50 percent to be EGR
equipped with 50 percent being WT equipped. Using these factors on the list of threshold monitors results in 9
monitors for the "average" diesel engine.
16
-------
Final Technical Support Document; HDOBD Final Rule
We believe that algorithm development will be more resource intensive than will
algorithm application (on a per monitor basis). For algorithm development of system
threshold monitors, we have estimated 30 engineer-weeks of development per monitor and
15 technician-weeks per monitor while for system functional monitors we have estimated 20
and 5 weeks of development per monitor, respectively. For rationality monitors, we have
estimated 15 engineer-weeks and only one technician-week since determining the proper
rationality—the engineer's job—can be difficult but testing and verifying that it works—the
technician's job—should not be difficult. For circuit continuity monitors, we have estimated
only two engineer-weeks and no technician weeks since these monitors are relatively straight
forward (open circuit/short circuit).
Multiplying by the engineer and technician labor rates and the number of monitors results
in total costs of $51 million which will be incurred during the four year period leading up to
implementation (i.e., during the years 2006 through 2009). These costs are shown in Table
4A.
For algorithm application to each engine family, we have estimated that the majority of
the work will entail testing and, therefore, it will be done by the technician. For system
threshold monitors and functional monitors, we have estimated five engineer-weeks and 10
technician weeks. For rationality monitors, we have estimated three engineer-weeks and one
technician-week because adapting these algorithms should be more straight forward than
adapting system monitors. For circuit continuity monitors, we have estimated no costs for
applying algorithms since these should be directly applicable to any engine.
These algorithm application costs will be incurred on each engine family. Our 2004
model year database shows a total of 65 diesel engine families meant for over 14,000 pound
vehicles. The database also shows 10 heavy-duty diesel engine manufacturers for an average
of 6.5 engine families per manufacturer. Multiplying the estimated weeks by the appropriate
engineering and technician labor rates, the number of monitors, the number of engine
families per manufacturer, and the number of manufacturers results in total costs of $107
million dollars. These costs are shown in Table 4B. These costs will be incurred on some
engine families during the four years leading up to the 2010 model year (i.e., one engine
family per manufacturer) and on the remaining families during the four years leading up to
the 2013 model year.
To estimate the costs for fine tuning the adapted algorithm to the remaining variants
within each engine family, we have considered this to take roughly one-quarter the effort
required for the initial engine family application. Therefore, the $413,000 cost per variant is
estimated as one-quarter of the $1.65 million per family cost to apply the algorithm to the
engine family. The variant based application costs are estimated to be incurred by those
remaining variants within the engine family (i.e., these costs are not incurred on the variant
for which the initial application work was done). Based on input from industry, we have
estimated that there is an average of five variants per engine family. As a result, the variant
application cost will be incurred on four variants per engine family. Multiplying the cost per
variant by the number of remaining variants, the average number of engine families per
manufacturer and again by the number of manufacturers results in another $107 million
17
-------
Final Technical Support Document; HDOBD Final Rule
dollars in total costs. These costs are shown in Table 4C. These costs will be incurred on
some engine families during the four years leading up to the 2010 model year (i.e., four
variants within one engine family per manufacturer) and on the variants of the remaining
families during the four years leading up to the 2013 model year.
We have used this same process for estimating the R&D costs for gasoline engines which
are shown in Table 5. We have used many of the same estimates for gasoline engines as for
diesel engines with the exception that we have estimated only eight system threshold
monitors for gasoline engines.13 As shown in Table 5 A, we have estimated that the algorithm
development costs for gasoline engines will be zero since the manufacturers of gasoline
engines (only Ford and General Motors have certified gasoline engines for over 14,000
pound vehicles) have been complying with OBD requirements for over 10 years on their
under 14,000 pound vehicles. We believe that the algorithms used in under 14,000 pound
vehicles will be directly applicable to over 14,000 pound vehicles with only some adapting of
those algorithms. The costs for adapting the existing algorithms to each engine family are
shown in Table 5B where we have estimated the costs at $5 million. Note that our 2004
model year certification database shows two over 14,000 pound engine families certified by
General Motors and none certified by Ford. We have estimated that Ford will certify an
engine family in future model years and, therefore, have estimated an average of 1.5 engine
families per manufacturer. Table 5C shows the costs for applying algorithms to each
remaining variant within the engine family. Again, as with diesel, we have estimated this
cost at one-quarter the cost of first adapting an algorithm to the engine family. These efforts
are estimated to result in another $5 million. All of these gasoline engine costs will be
incurred in a manner analogous to that described above for diesel engines.
b The eight threshold monitors for gasoline engines, based on our engineering judgement, would be: fuel
system too rich; fuel system too lean; multiple cylinder random misfire; secondary air system low flow; catalyst
conversion; EGR low flow; variable valve timing (VVT) above target; WT below target; WT slow response;
and primary exhaust gas sensor slow response. As with diesel engines, we have estimated 50 percent to be
EGR equipped with 50 percent being WT equipped.
18
-------
Final Technical Support Document; HDOBD Final Rule
Table 5. R&D Costs for OBD Algorithm Development and Application - Gasoline Engines for Over
14,000 Pound Applications
(2007 dollars)
|A. Algorithm Development Costs
System Threshold Monitors
Engineers
Technician $
Subtotal
System Functional Monitors
Engineers
Technician $
Subtotal
CCM Rationality Monitors
Engineers
Technician $
Subtotal
CCM Continuity Monitors
Engineers
Technician $
Subtotal
Total
|B. Application Costs to each Family
System Threshold Monitors
Engineers
Technician $
Subtotal
System Functional Monitors
Engineers
Technician $
Subtotal
CCM Rationality Monitors
Engineers
Technician $
Subtotal
Total
| weeks/monitor)
30
15
20
5
15
1
2
0
[weeks/monitor |
5
10
5
10
3
1
Cost/monitor) # of monitors)
$69,000
$21,000
$90,000
$46,000
$7,000
$53,000
$34,000
$1,000
$35,000
$5,000
$0
$5,000
Cost/monitor) # of monitors)
$11,000
$14,000
$25,000 8
$11,000
$14,000
$25,000 42
$7,000
$1,000
$8,000 50
Total/Mfr| Total |
$0 $0
$0 $0
$0 $0
$0 $0
$0 $0
Total/Familyl #families/mfr|
$200,000 1.5
$1,050,000 1.5
$400,000 1.5
$1,650,000
Total/Mfr| Total |
$300,000 $600,000
$1,575,000 $3,150,000
$600,000 $1,200,000
$2,475,000 $4,950,000
|C. Application Costs to remaining Variants
Total
I I
| Total/Variant)
$413,000
# variants/family) #families/mfr|
4 1.5
Total/Mfr| Total |
$2,478,000 $4,956,000
Closely associated with the costs shown in Table 4 and Table 5 would be costs associated
with operating and maintaining the test cells required for testing and evaluating the OBD
systems and associated algorithms. To determine these costs we projected that two types of
test cell work would be done. The first would be actual emissions testing using a certified
emissions test cell. The other would be performance and/or endurance testing done in a
development test cell where OBD monitors could be evaluated against functional criteria
rather than emissions criteria and where operating hours can be amassed far more cost
efficiently than by using a certified emissions test cell. The costs associated with these
different test cells were estimated at $770 per hour for an emissions test cell and $110 per
hour for an endurance test cell. We also estimated that 90 percent of the test cell time for
OBD development work would be done in an endurance test cell with the remaining 10
percent being done in an emissions test cell.
Table 6 shows the test cell costs we have estimated for diesel engines. Note that these
costs represent the costs associated with operating existing test cells for the sake of meeting
the OBD requirements. We are not projecting that any new test cells would have to be built.
As shown in Table 6, we have estimated the test cell demand for algorithm development of a
system threshold monitor at three weeks. Algorithm development of a system functional
monitor was estimated to require two weeks of test cell time while a rationality monitor was
estimated at one week. We have estimated no test cell demand for circuit continuity
19
-------
Final Technical Support Document; HDOBD Final Rule
monitors. We have used the same base estimates for the test cell demand associated with
applying algorithms to individual engine families except that we have estimated the demand
to be only 30 percent of that required for algorithm development. The same is true for
applying engine family algorithms to individual variants except here we have estimated the
demand to be only 10 percent of that required for initial algorithm development.
Table 6 shows how these costs are incurred in preparation for compliance in the 2010
model year and the 2013 model year. As stated above, 90 percent of the test cell demand—
i.e., the total test weeks—would be met using an endurance test cell at $110 per hour while
the remaining 10 percent of the demand would be met using an emissions test cell at $770
per hour. Note that there would be no test cell demand for algorithm development beyond
that incurred for 2010 since the same algorithms would be used for 2010 and later model
years. Table 6A shows an estimated cost for test cell operation of $1.9 million per
manufacturer or $19 million for the industry in preparation for the 2010 model year. These
costs would be incurred over the four year period leading up to the 2010 model year. For the
2013 model year when 100 percent compliance is required, the cost is estimated at $4.3
million per manufacturer or $43 million total to be spread over the four year period leading
up to the 2013 model year. The 2013 costs are shown in Table 6B.
Table 7A and Table 7B show the analogous information for gasoline engines complying
in the 2010 and 2013 model years, respectively. The table shows that we have estimated no
costs—development or test cell—for developing monitoring algorithms for gasoline engines
since the same algorithms as are used on under 14,000 pound vehicles can be used for over
14,000 pound vehicles. The test cell costs for gasoline engines are estimated at $1.6 million
for 2010 model year compliance and $780 thousand for 2013 model year compliance. As
with the diesel costs, these costs are expected to be incurred over the four year period leading
up to the first year of compliance.
Table 8 and Table 9 summarize the estimated test cell demand per manufacturer for
meeting the 2010 and the 2013 requirements. These summaries estimate that testing is
conducted during 48 weeks in a given year.
20
-------
Final Technical Support Document; HDOBD Final Rule
Table 6. OBD R&D Test Cell Costs - Diesel Engines for Over 14,000 Pound Applications
(2007 dollars)
A. R&D Test Cell Costs - Diesel
Monitor Algorithms testwks
System monitor - threshold 3
System monitor -functional 2
Rationality monitor 1
Subtotal
$ per year for 4 years
Cost for 2010
# of monitors) | total test wks|
9
41
50
27.0
82.0
50.0
Costs/mfr|
$190,000
$576,000
$351,000
$1,117,000
$279,250
# mfrs|
10
10
10
Total
$1,900,000
$5,760,000
$3,510,000
$11,170,000
$2,792,500
Monitor Application to each engine family
factor 30% |
System monitor- threshold 0.9
System monitor -functional 0.6
Rationality monitor 0.3
Subtotal
$ per year for 4 years
# of monitors) #families/mfr I total test wks|
9 1.0
41 1.0
50 1.0
8.1
24.6
15.0
Costs/mfr|
$57,000
$173,000
$105,000
$335,000
$83,750
# mfrs|
10
10
10
Total |
$570,000
$1,730,000
$1,050,000
$3,350,000
$837,500
Monitor Application to each engine family variant
factor 10% |
System monitor- threshold 0.3
System monitor -functional 0.2
Rationality monitor 0.1
Subtotal
$ per year for 4 years
Total R&D Test Cell Costs
$ per year for 4 years
# of monitors| #families/mfr additional variants] total test wks|
9 1.0 4.0
41 1.0 4.0
50 1.0 4.0
10.8
32.8
20.0
Costs/mfr|
$76,000
$230,000
$140,000
$446,000
$111,500
$1,898,000
$474,500
# mfrs|
10
10
10
Total]
$760,000
$2,300,000
$1,400,000
$4,460,000
$1,115,000
$18,980,000
$4,745,000
B. R&D Test Cell Costs - Diesel
Monitor Algorithms testwks
System monitor - threshold 3
System monitor -functional 2
Rationality monitor 1
Subtotal
$ per year for 4 years
Costs for 201 3
# of monitors) | total test wks|
0
0
0
Costs/mfr|
# mfrs|
10
10
10
Total
$0
$0
$0
Monitor Application to each engine family
factor 30%
System monitor- threshold 0.9
System monitor -functional 0.6
Rationality monitor 0.3
Subtotal
$ per year for 4 years
# of monitors| #families/mfr | total test wks|
9 5.5
41 5.5
50 5.5
44.6
135.3
82.5
Costs/mfr|
$313,000
$950,000
$580,000
$1,843,000
$460,750
# mfrs|
10
10
10
Total
$3,130,000
$9,500,000
$5,800,000
$18,430,000
$4,607,500
Monitor Application to each engine family variant
factor 10% |
System monitor- threshold 0.3
System monitor -functional 0.2
Rationality monitor 0.1
Subtotal
$ per year for 4 years
Total R&D Test Cell Costs
$ per year for 4 years
# of monitors| #families/mfr additional variants] total test wks|
9 5.5 4.0
41 5.5 4.0
50 5.5 4.0
59.4
180.4
110.0
Costs/mfr|
$417,000
$1,267,000
$773,000
$2,457,000
$4,300,000
$1,075,000
# mfrs|
10
10
10
Total]
$4,170,000
$12,670,000
$7,730,000
$24,570,000
$43,000,000
$10,750,000
21
-------
Final Technical Support Document; HDOBD Final Rule
Table 7. OBD R&D Test Cell Costs - Gasoline Engines for Over 14,000 Pound Applications
(2007 dollars)
A. R&D Test Cell Costs - Gasoline
Monitor Algorithms
Cost for 2010
test wks
# of monitors
I
total test wks | Costs/mfr
#mfrs
Total
System monitor - threshold
System monitor - functional
Rationality monitor
Subtotal
$ per year for 4 years
Monitor Application to each engine family
factor
System monitor - threshold
System monitor - functional
Rationality monitor
Subtotal
$ per year for 4 years
30%| # of monitors) # families/mfr|
0.9
0.6
0.3
42
50
1.0
1.0
1.0
total test wks| Costs/mfr| # mirs]
7.2
25.2
15.0
$51,000
$177,000
$105,000
$333,000
$83,250
Total |
$102,000
$354,000
$210,000
$666,000
$166,500
Monitor Application to each engine family variant
factor
System monitor - threshold
System monitor - functional
Rationality monitor
Subtotal
$ per year for 4 years
10%
0.3
0.2
0.1
#ofmonitors| # families/mfr| additional variants) total test wks| Costs/mfr|
42
50
1.0
1.0
1.0
4.0
4.0
4.0
9.6
33.6
20.0
$67,000
$236,000
$140,000
$443,000
$110,750
# mfrs|
Total |
$134,000
$472,000
$280,000
$886,000
$221,500
Total R&D Test Cell Costs
$ per year for 4 years
$776,000
$194,000
$1,552,000
$388,000
B. R&D Test Cell Costs - Gasoline
Monitor Algorithms test wks
System monitor - threshold 3
System monitor - functional 2
Rationality monitor 1
Subtotal
$ per year for 4 years
Costs for 201 3
# of monitors | total test wks| Costs/mfr
0
0
0
# mfrs Total
2 $0
2 $0
2 $0
Monitor Application to each engine family
factor 30°/
System monitor - threshold 0.9
System monitor - functional 0.6
Rationality monitor 0.3
Subtotal
$ per year for 4 years
# of monitors # families/mfr| total test wks| Costs/mfr
8 0.5 3.6 $25,000
42 0.5 12.6 $89,000
50 0.5 7.5 $53,000
$167,000
$41,750
# mfrs Total
2 $50,000
2 $178,000
2 $106,000
$334,000
$83,500
Monitor Application to each engine family variant
factor 1 0°/
System monitor - threshold 0.3
System monitor - functional 0.2
Rationality monitor 0.1
Subtotal
$ per year for 4 years
Total R&D Test Cell Costs
$ per year for 4 years
# of monitors # families/mfr| additional variants total test wks| Costs/mfr
8 0.5 4.0 4.8 $34,000
42 0.5 4.0 16.8 $118,000
50 0.5 4.0 10.0 $70,000
$222,000
$55,500
$389,000
$97,250
# mfrs Total
2 $68,000
2 $236,000
2 $140,000
$444,000
$111,000
$778,000
$194,500
22
-------
Final Technical Support Document; HDOBD Final Rule
Table 8. OBD R&D Test Cell Demand per Manufacturer - Diesel Engines for Over 14,000 Pound
Applications
A. R&D Test Cell
Monitor Algorithms
Monitor Application to
Monitor Application to
Total
Cells needed per mfr
Demand - Diesel
each engine family
each engine family variant
For 2010
Total test wks|
159.0
47.7
63.6
270.3
Cells needed per mfr per each of 4 years
B. R&D Test Cell
Monitor Algorithms
Monitor Application to
Monitor Application to
Total
Cells needed per mfr
Demand - Diesel
each engine family
each engine family variant
CVS cell test wks|
15.9
4.8
6.4
27.0
0.6
0.1
Endurance cell test wks
143.1
42.9
57.2
243.3
5.1
1.3
For 201 3
Total test wks|
-
262.4
349.8
612.2
Cells needed per mfr per each of 4 years
CVS cell test wks|
-
26.2
35.0
61.2
1.3
0.3
Endurance cell test wks
-
236.1
314.8
550.9
11.5
2.9
Table 9. OBD R&D Test Cell Demand per Manufacturer - Gasoline Engines for Over 14,000 Pound
Applications
A. R&D Test Cell Demand - Gasoline
For 2010
Total test wks| CVS cell test wks| Endurance cell test wks
Monitor Algorithms
Monitor Application to each engine family
Monitor Application to each engine family variant
Total
Cells needed per mfr
Cells needed per mfr per each of 4 years
47.4
63.2
110.6
4.7
6.3
11.1
0.2
0.1
42.7
56.9
99.5
2.1
0.5
B. R&D Test Cell Demand - Gasoline
For 2013
Total test wks| CVS cell test wks| Endurance cell test wks
Monitor Algorithms
Monitor Application to each engine family
Monitor Application to each engine family variant
Total
Cells needed per mfr
Cells needed per mfr per each of 4 years
23.7
31.6
55.3
2.4
3.2
5.5
0.1
0.0
21.3
28.4
49.8
1.0
0.3
These R&D costs—algorithm development, algorithm application, and test cell—are
summarized in Table 10 for both diesel and gasoline engines. The net present value of the
estimated R&D costs through 2035 is $298 million using a three percent discount rate and $252
million using a seven percent discount rate.
23
-------
Final Technical Support Document; HDOBD Final Rule
Table 10. Summary of OBD R&D Costs - Diesel and Gasoline Engines for Over 14,000 Pound Applications
(2007 dollars)
Year
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
NPV@
NPV@
CY
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
3%
7%
Diesel
R&D-Algorithms| R&D-Application| R&D-Test Cell) Subtotal R&D
$12,833,000 $8,255,000 $4,745,000 $25,833,000
$12,833,000 $8,255,000 $4,745,000 $25,833,000
$12,833,000 $8,255,000 $4,745,000 $25,833,000
$12,833,000 $53,658,000 $15,495,000 $81,986,000
$45,403,000 $10,750,000 $56,153,000
$45,403,000 $10,750,000 $56,153,000
$45,403,000 $10,750,000 $56,153,000
$47,702,000 $185,131,000 $54,206,000 $287,038,000
$43,468,000 $153,500,000 $45,796,000 $242,763,000
Gasoline
R&D-Algorithms| R&D-Application| R&D-Test Cell) Subtotal R&D
$1,651,000 $388,000 $2,039,000
$1,651,000 $388,000 $2,039,000
$1,651,000 $388,000 $2,039,000
$2,477,000 $583,000 $3,060,000
$826,000 $195,000 $1,021,000
$826,000 $195,000 $1,021,000
$826,000 $195,000 $1,021,000
$8,947,000 $2,106,000 $11,052,000
$7,876,000 $1,853,000 $9,730,000
Total R&D
$27,872,000
$27,872,000
$27,872,000
$85,046,000
$57,174,000
$57,174,000
$57,174,000
$298,090,000
$252,493,000
24
-------
Final Technical Support Document; HDOBD Final Rule
b. Certification and Production Evaluation Testing Costs
As noted above, the certification costs include the costs associated with testing of durability
data vehicles (i.e., the OBD parent engines), the costs associated with generating the "limit" parts
that are required to demonstrate OBD detection at or near the applicable emissions thresholds,
and the costs associated with generating the necessary certification documentation.
Cost of OBD Limit Parts
We look first at the costs associated with generating limit parts for certification
demonstration testing. These are the parts used to demonstrate OBD detection at or near the
applicable emissions thresholds. Such parts can be very difficult to generate because of the
difficulties associated with deteriorating parts just the right amount—not so much that the
thresholds are grossly exceeded thereby making the demonstration test somewhat meaningless
and not so little that emissions remain well below the thresholds.
Table 11 shows the costs we have estimated for the limit parts needed for diesel engine
demonstration testing. To arrive at these costs, we estimated the part costs of aftertreatment
devices based on our 2007/2010 highway heavy-duty rule and our recent nonroad Tier 4 rule.
However, since those costs represented costs of new parts being mass produced, we doubled the
costs here to represent the higher costs associated with orders to suppliers consisting of only one
or two parts. Fuel system costs were estimated to include costs for injectors, pressure regulators,
etc. The exhaust gas sensor costs estimate NOx sensors and estimate that these are ordered (and
costed) in sets of two. We estimated the costs for a typical light-heavy, medium-heavy, and
heavy-heavy engine assuming 6, 8, and 14 liter displacements, respectively. We sales weighted
these costs using the projected sales data from our 2004 model year certification database
excluding California sales and excluding those engines certified for use in vehicles under 14,000
pounds. We have estimated that two parts would be needed to account for possible errors and/or
the need for parts to demonstrate both a high and a low failure (e.g., EGR flow high/EGR flow
low). For variable valve timing (VVT) costs, we have estimated these based on input from
industry and not based on our prior analyses which did not consider costs for VVT systems. As
shown in Table 11, multiplying through and including the percent of engines we expect will need
the particular limit parts, results in limit parts cost of $21,300 for each diesel engine undergoing
demonstration testing.
25
-------
Final Technical Support Document; HDOBD Final Rule
Table 11. Cost for OBD Certification Demonstration Limit Parts - Diesel Engines for Over 14,000 Pound
Applications
(2007 dollars)
Light-heavy
Diesel Engines 14-19.5K Medi
Displacement (liters) 6
2004 Projected Sales less CA sales 21 ,695
NOx Adsorber $1,646
SCR $1,646
DPF $2,744
Fuel system $1,372
Exhaust gas sensors $220
Turbo $615
EGR System $406
VVT $1,646
Total for Limit Parts
Parts
jm-heavy Heavy-heavy Sales Weighted (inc
8 14
361,393 231,434 614,522
$2,195 $3,622 $2,700
$2,195 $3,622 $2,700
$3,512 $6,147 $4,500
$1,372 $1,646 $1,500
$220 $220 $200
$626 $692 $700
$483 $724 $600
$1,646 $1,646 $1,500
needed Percent
errors) needing part
2 50%
2 50%
2 100%
2 100%
2 100%
2 100%
2 50%
2 50%
Fleet weighted
$2,700
$2,700
$9,000
$3,000
$400
$1,400
$600
$1,500
$21,300
We have not estimated costs associated with generating limit parts for gasoline engines
because we do not expect that over 14,000 pound engines will be used for certification
demonstration. Instead, we expect that manufacturers will demonstrate their OBD systems using
an engine or vehicle in the under 14,000 pound range and then provide documentation in their
certification package showing how their over 14,000 pound engine is represented by the under
14,000 pound demonstration as allowed by the proposed program. While this may also be the
case for some diesel engine manufacturers, we have chosen to be conservative in our estimates
by assuming that all diesel demonstrations will be over 14,000 pounds.
We have also estimated costs for aging these limit parts to the point where emissions are
near the OBD thresholds. These costs were not included in our draft cost analysis but have been
included in the final analysis (see the Summary and Analysis of Comments document Section
VLB). To estimate these costs, we have considered the final requirement for aging of
aftertreatment devices which requires half life aging in early years and full life aging in later
years. On average, we have estimated this at 80 percent of uselife life given a belief that
manufacturers will be able to demonstrate that, for OBD, such aging is representative of full life.
We have also used a sales weighted full useful life of 335,000 miles and sales weighted MPG of
7, and an endurance test cell cost of $110/hour to arrive at the following costs:
Fuel costs: (335,000 miles x 80%)/7 MPG x $3/gallon = $115,000 per parent engine
Test cell costs: (335,000 miles x 80%)/30 MPH x $110/hr = $981,000 per parent engine.
The total costs for limit part aging being $1.1 million per parent engine. We have added
these costs to the costs of limit parts in the final analysis. As a result, in the final cost analysis,
limit part costs consist of the limit part hardware and the limit part aging.
Note that we believe this to be an overestimation of the costs associated with aging of limit
parts. We expect that manufacturers will use other means to "age" the parts to the OBD
thresholds (e.g., rapid aging, oven baking, etc.) and will not actually age the parts on an engine
dynamometer for 335,000 miles. However, such dynamometer aging is one potential method,
26
-------
Final Technical Support Document; HDOBD Final Rule
albeit a worst case method, and provides a practical means of estimating the costs. In the end,
the costs we have estimated are worst case and will almost certainly be lower.
We have estimated that these costs for limit parts will be incurred every three years going
forward. In 2010, one engine family per manufacturer will have to be demonstrated and in 2013
we expect another two engine families per manufacturer to undergo demonstration testing (for
diesels). We would then expect engine families to be carried-over for three years at which time
another three engines would be demonstrated, etc. This is an over simplification of the carry-
over provisions of our certification program, but it serves our purpose here and does not under
estimate the costs but rather impacts only when those costs are incurred. We use this same
simplifying assumption throughout our analysis of certification and production evaluation testing
costs as is shown in Table 12 which shows all our estimated certification and production
evaluation testing costs for diesel engines and Table 13 which shows the analogous costs for
gasoline engines.
27
-------
Final Technical Support Document; HDOBD Final Rule
Table 12. OBD Certification and Production Evaluation Testing Costs - Diesel Engines for Over 14,000 Pound Applications
(2007 dollars)
Year
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
NPV@
NPV@
CY
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
3%
7%
Certification Demonstration Testing Related
# of parent Costs for Limit
test engines Parts
0
0
0
10 $11,173,000
0
0
20 $22,346,000
0
0
30 $33,519,000
0
0
30 $33,519,000
0
0
30 $33,519,000
0
0
30 $33,519,000
0
0
30 $33,519,000
0
0
30 $33,519,000
0
0
30 $33,519,000
0
0
$164,018,000
$92,794,000
DDV Testing
Costs Total DDV Costs
$553,000 $11,726,000
$1,106,000 $23,452,000
$1,660,000 $35,179,000
$1,660,000 $35,179,000
$1,660,000 $35,179,000
$1,660,000 $35,179,000
$1,660,000 $35,179,000
$1,660,000 $35,179,000
$1,660,000 $35,179,000
$8,122,000 $172,140,000
$4,595,000 $97,388,000
Certification Documentation Related
#of #
parent remaining
families families
0 0
0 0
0 0
10 0
0 0
0 0
20 45
0 0
0 0
30 35
0 0
0 0
30 35
0 0
0 0
30 35
0 0
0 0
30 35
0 0
0 0
30 35
0 0
0 0
30 35
0 0
0 0
30 35
0 0
0 0
Cert
Documentation
Costs
$55,000
$233,000
$261,000
$261,000
$261,000
$261,000
$261,000
$261,000
$261,000
$1,297,000
$735,000
Production Evaluation Testing Related
PE Testing - Scan Tool PE
# of engine # of
families for Gr
testing PE Costs te
0
0
0
0
10 $23,000
0
0
55 $126,000
0
0
65 $149,000
0
0
65 $149,000
0
0
65 $149,000
0
0
65 $149,000
0
0
65 $149,000
0
0
65 $149,000
0
0
65 $149,000
0
$706,000
$382,000
I Testing - Monitors
OBD
Dups PE Costs (incl
5ted vehicle rental)
0
0
0
0
10 $211,000
0
0
20 $280,000
0
0
30 $348,000
0
0
30 $348,000
0
0
30 $348,000
0
0
30 $348,000
0
0
30 $348,000
0
0
30 $348,000
0
0
30 $348,000
0
$1,773,000
$996,000
PE Testing - Ratios
#of
monitoring
groups
tested
0
0
0
0
30
30
30
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
PE Costs
$8,000
$8,000
$8,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$15,000
$221,000
$121,000
PE Costs -
Total
$242,000
$8,000
$8,000
$421,000
$15,000
$15,000
$512,000
$15,000
$15,000
$512,000
$15,000
$15,000
$512,000
$15,000
$15,000
$512,000
$15,000
$15,000
$512,000
$15,000
$15,000
$512,000
$15,000
$15,000
$512,000
$15,000
$2,700,000
$1,499,000
Total
Certification &
PE Testing
Costs
$11,781,000
$242,000
$8,000
$23,693,000
$421,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$176,137,000
$99,623,000
28
-------
Final Technical Support Document; HDOBD Final Rule
Table 13. OBD Certification and Production Evaluation Testing Costs - Gasoline Engines for Over 14,000 Pound Applications
(2007 dollars)
Year
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
NPV@
NPV@
CY
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
3%
7%
Certification Demonstration Testing Related
#of
parent
test Costs for DDV Testing Total DDV
engines Limit Parts Costs Costs
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Certification Documentation Related
# of parent # remaining
families families
0 0
0 0
0 0
0 0
0 0
0 0
0 3
0 0
0 0
0 3
0 0
0 0
0 3
0 0
0 0
0 3
0 0
0 0
0 3
0 0
0 0
0 3
0 0
0 0
0 3
0 0
0 0
0 3
0 0
0 0
Cert
Documentation
Costs
$8,000
$8,000
$8,000
$8,000
$8,000
$8,000
$8,000
$8,000
$39,000
$22,000
Production Evaluation Testing Related
PE Testing - Scan Tool PE Testing - Monitors PE
# of engine # of OBD
families for Groups
testing PE Costs tested
0 0
0 0
0 0
0 0
2 $5,000 2
0 0
0 0
1 $2,000 1
0 0
0 0
3 $7,000 3
0 0
0 0
3 $7,000 3
0 0
0 0
3 $7,000 3
0 0
0 0
3 $7,000 3
0 0
0 0
3 $7,000 3
0 0
0 0
3 $7,000 3
0 0
0 0
3 $7,000 3
0 0
$33,000
$18,000
*
mon
PE Costs (incl gr
vehicle rental) te
$42,000
$21,000
$49,000
$49,000
$49,000
$49,000
$49,000
$49,000
$49,000
$246,000
$138,000
Testing - Ratios
of
itoring
Dups
sted PE Costs
0
0
0
0
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
6 $2,000
$32,000
$18,000
PE Costs -
Total
$49,000
$2,000
$2,000
$25,000
$2,000
$2,000
$58,000
$2,000
$2,000
$58,000
$2,000
$2,000
$58,000
$2,000
$2,000
$58,000
$2,000
$2,000
$58,000
$2,000
$2,000
$58,000
$2,000
$2,000
$58,000
$2,000
$311,000
$175,000
Total Certification
& PE Testing
Costs
$49,000
$2,000
$10,000
$25,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$350,000
$197,000
29
-------
Final Technical Support Document; HDOBD Final Rule
Focusing first on Table 12, the limit parts costs are first incurred in 2009 in advance of the
2010 model year. The limit parts cost estimate for diesels shown in Table 11 ($21,300 per
engine) plus the aging component of $1.1 million per engine is incurred on one engine family for
each of 10 engine manufacturers for a total cost that year of $11 million. This process is carried
forward every three years as discussed above. As noted, for gasoline engines, Table 13 shows no
limit parts costs or demonstration testing costs.
OBD Certification Demonstration Testing Costs
For costs associated with the actual demonstration testing of OBD parent engines (diesel
only), we have estimated that two OBD threshold monitors can be demonstrated during a given
day of testing in an emissions test cell. With our estimate of 9 threshold monitors per engine,
this means 9 days of testing in an emissions test cell that costs $770 dollars per hour or $6,100
per day to operate. The OBD parent engine, or durability data vehicle (DDV), demonstration
testing costs were then calculated by multiplying the test days per engine (9) by the dollars per
day ($6,100) and again by the number of demonstration engines being demonstrated for the
given model year. The result in 2009 is $553,000 for all 10 engine manufacturers which is
incurred one year in advance of implementation because they are certification costs. These costs
change depending on the number of engine families undergoing demonstration testing.
OBD Certification Documentation Costs
For certification documentation costs, we have estimated that a certification documentation
package for an OBD parent engine would cost $5,500 while it would cost $2,700 for a non-OBD
parent engine (i.e., an OBD child rating). We consider this to be a conservative estimate since
most child ratings would very likely incur no costs since it would be part of an OBD group
represented by the OBD parent engine and should, therefore, require no further certification
documentation. Our certification database for the 2004 model year showed 65 diesel engine
families and three gasoline engine families in the over 14,000 pound range. Multiplying the
expected number of OBD parent engines and child engines being certified for each given year by
the estimated costs to generate the certification documentation packages results in the costs
shown in Table 12 and Table 13.
OBD Production Evaluation Testing Costs
Also shown are costs for production evaluation (PE) testing. The required production
evaluation testing consists of three elements. The first of these is testing to ensure that
engines/vehicles comply with the standardization requirements of the OBD rule. This is done by
connecting a scan tool to a production vehicle to ensure that the onboard systems communicate
properly to an off board device (e.g., a scan tool). We would expect this testing to be done as
vehicles roll off the vehicle assembly line. The second element of PE testing is testing to ensure
that the OBD monitors are functioning properly. This is done by implanting or simulating
malfunctions and determining whether or not the OBD monitors run and detects them. This
testing does not involve any actual emissions testing. We would expect this testing to be done on
one to three production vehicles but required test beyond one vehicle could be done on
30
-------
Final Technical Support Document; HDOBD Final Rule
production engines rather than production vehicles. The third element of PE testing is testing to
ensure that OBD monitors are running and making diagnostic decisions with sufficient frequency
in the real world. This is done by scanning the stored OBD information contained in actual in-
use vehicles and noting the performance ratios for various non-continuous monitors. Since the
production evaluation testing is a post-certification requirement, the costs would be incurred
either as new engines/vehicles are rolling off the assembly line or during the six to 12 months
following introduction into commerce.
OBD Production Evaluation Testing Costs - Standardization Requirements
To estimate the PE testing costs for verifying the standardization requirements, we have
conservatively estimated that the actual test would take four hours and that for each engine
family sold the maximum of 10 vehicles would be tested. We have also conservatively estimated
that the testing would be done by an engineer at $110,000 per year rather than the more likely
choice of a technician at $66,000 per year. Multiplying the number of engine families by the
number of vehicles tested per family, the hours per test, and the engineer's cost per hour results
in the yearly estimated costs. This cost—shown as "PE testing - scan tool" in the tables—is
estimated at $23,000 for diesel engines in 2010 and $5,000 for gasoline engines in 2010. These
costs would be incurred on newly introduced OBD-compliant engine families. Therefore, we
have estimated costs for testing the engine families from which the OBD parent engine has been
chosen. We have also included costs for future model years assuming that most engines undergo
enough changes over a three year period to nullify the ability to carry-over from a prior year's
certification. When that occurs, we would expect the PE scan tool testing to be done.
OBD Production Evaluation Testing Costs - Monitor Verification
To estimate the PE testing costs for verifying monitors, we have first been conservative by
estimating that each manufacturer would conduct the testing for each of three OBD groups. This
overestimates these costs because some manufacturers will only have to conduct the testing on
one, and others on two, OBD groups because they do not sell enough different engine families to
require testing of three. We have also estimated that, as allowed by the proposed rule, the first
OBD group tested would have to be tested using a production vehicle while the remaining OBD
groups tested would use a production engine. We have estimated the time required to conduct
the testing at three weeks and that the testing would be done by an engineer costing $110,000 per
year. We have also estimated that it would cost $11,000 to rent or otherwise acquire a vehicle
for testing while acquiring an engine would not cost the engine manufacturer anything. Lastly,
we have estimated travel costs at $3,300 dollars for testing done on a production vehicle while
travel costs for testing on production engines would be zero. The certification and production
engine testing cost tables show—in the columns under "PE testing - monitors"—the number of
OBD groups undergoing this testing in given years. The 10 shown for 2010 represent one engine
tested from each OBD compliant engine family by each of 10 manufacturers. In 2013, we
require all engine families to comply but only up to two new engine families must undergo
certification demonstration testing and, consequently, PE testing for monitors. For simplicity, as
stated elsewhere, we have estimated three new parent engines per manufacturer undergo
certification demonstration testing every three years and, consequently, they undergo PE testing
for monitors.
-------
Final Technical Support Document; HDOBD Final Rule
OBD Production Evaluation Testing Costs - Performance Ratios
To estimate the PE testing costs for evaluating in-use performance ratios, we have first
conservatively estimated that every OBD monitoring group would have to test the maximum of
15 vehicles. An OBD monitoring group is defined first by emissions control architecture (i.e.,
combination of EGR, turbo, and aftertreatment devices) and secondly by application type (i.e.,
line haul, urban delivery, other). We have estimated that each manufacturer would have two
emissions control architectures and engines sold into each of the three application types. As a
result, there would be six monitoring groups per each of 10 different manufacturers for 60
monitoring groups being tested. This is true except for the 2010 to 2012 model years when,
since only one engine family is compliant, we have assumed only one emissions control
architecture and, therefore, only three OBD monitoring groups for each of 10 manufacturers for
30 total. We have also estimated that the test itself—simply connecting a scan tool and
downloading the performance ratio data—would take half an hour to complete by a technician
costing $66,000 per year. We have been conservative in our estimate by including costs for this
testing in every year even though we would expect that data could be carried over from one year
to the next once we are sure that monitors are indeed running at sufficient frequency in-use.
Table 14 shows the cost streams presented above for all fixed costs. The fixed costs
consist of R&D, certification, and production evaluation testing costs. Also shown are the 30
year net present values at a three percent discount rate which are $463 million for diesel, $11
million for gasoline and $475 million for the entire industry. The total fixed costs are also shown
on a per engine basis using the projected sales shown in Table 3.
32
-------
Final Technical Support Document; HDOBD Final Rule
Table 14. Total OBD Fixed Costs - Diesel and Gasoline Engines for Over 14,000 Pound Applications
(2007 dollars)
Year
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
NPV@
NPV@
CY
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
3%
7%
Diesel
R&D| Cert/PE Testing]
$25,833,000
$25,833,000
$25,833,000
$81,986,000 $11,781,000
$56,153,000 $242,000
$56,153,000 $8,000
$56,153,000 $23,693,000
$421,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$287,038,000 $176,137,000
$242,763,000 $99,623,000
Subtotal]
$25,833,000
$25,833,000
$25,833,000
$93,767,000
$56,395,000
$56,161,000
$79,846,000
$421,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$463,175,000
$342,386,000
Projected Sales]
639,103
651,393
663,684
675,974
688,265
700,555
712,846
725,136
737,426
749,717
762,007
774,298
786,588
798,879
811,169
823,459
835,750
848,040
860,331
872,621
884,912
897,202
909,493
921,783
934,073
946,364
958,654
970,945
983,235
995,526
$/engine
$40
$40
$39
$139
$82
$80
$112
$1
$0
$47
$1
$0
$45
$1
$0
$43
$1
$0
$41
$1
$0
$40
$1
$0
$38
$1
$0
$37
$1
$0
Gasoline
R&D| Cert/PE Testing |
$2,039,000
$2,039,000
$2,039,000
$3,060,000
$1,021,000 $49,000
$1,021,000 $2,000
$1,021,000 $10,000
$25,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$11,052,000 $350,000
$9,730,000 $197,000
Subtotal]
$2,039,000
$2,039,000
$2,039,000
$3,060,000
$1,070,000
$1,023,000
$1,031,000
$25,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$11,402,000
$9,926,000
Projected Sales]
40,976
41,764
42,552
43,340
44,128
44,916
45,704
46,492
47,280
48,068
48,856
49,644
50,432
51,220
52,008
52,796
53,584
54,372
55,160
55,948
56,736
57,524
58,312
59,100
59,888
60,676
61,464
62,252
63,040
63,828
$/engine
$50
$49
$48
$71
$24
$23
$23
$1
$0
$0
$1
$0
$0
$1
$0
$0
$1
$0
$0
$1
$0
$0
$1
$0
$0
$1
$0
$0
$1
$0
Total Fixed Costs
$27,872,000
$27,872,000
$27,872,000
$96,827,000
$57,465,000
$57,184,000
$80,877,000
$446,000
$17,000
$35,465,000
$570,000
$17,000
$35,465,000
$570,000
$17,000
$35,465,000
$570,000
$17,000
$35,465,000
$570,000
$17,000
$35,465,000
$570,000
$17,000
$35,465,000
$570,000
$17,000
$35,465,000
$570,000
$17,000
$474,577,000
$352,312,000
33
-------
Final Technical Support Document; HDOBD Final Rule
3.1.3 Total Costs
Combining the variable cost streams shown in Table 3 and the fixed costs streams shown in
Table 14 results in the total estimated costs for the over 14,000 pound proposed OBD
requirements. The results are shown in Table 15. As shown, the 30 year net present value at a
three percent discount rate is estimated at $1.2 billion with the majority of those costs being for
new hardware in the form of more powerful engine and emissions control system computers.
Note that the per engine costs shown in Table 15 use the engine sales estimates shown in Table 3
without accounting for any phase-in (i.e., the costs have been divided by the total new engine
sales rather than dividing by the fraction of new engine sales that are compliant).
Table 15. Total Estimated OBD Costs - Diesel and Gasoline Engines for Over 14,000 Pound Applications
(2007 dollars)
Year
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
NPV@
NPV@
CY
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
3%
7%
Diesel
Variable]
$14,852,000
$15,117,000
$15,382,000
$43,646,000
$44,385,000
$45,125,000
$43,356,000
$44,055,000
$44,754,000
$45,454,000
$46,153,000
$46,852,000
$47,551,000
$48,251,000
$48,950,000
$49,649,000
$50,349,000
$51,048,000
$51,747,000
$52,446,000
$53,146,000
$53,845,000
$54,544,000
$55,244,000
$55,943,000
$56,642,000
$685,964,000
$363,769,000
Fixed]
$25,833,000
$25,833,000
$25,833,000
$93,767,000
$56,395,000
$56,161,000
$79,846,000
$421,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$35,455,000
$512,000
$15,000
$463,175,000
$342,386,000
Subtotal
$25,833,000
$25,833,000
$25,833,000
$93,767,000
$71,247,000
$71,278,000
$95,228,000
$44,067,000
$44,400,000
$80,580,000
$43,868,000
$44,070,000
$80,209,000
$45,966,000
$46,168,000
$82,307,000
$48,063,000
$48,266,000
$84,405,000
$50,161,000
$50,364,000
$86,503,000
$52,259,000
$52,461,000
$88,601,000
$54,357,000
$54,559,000
$90,699,000
$56,455,000
$56,657,000
$1,149,139,000
$706,155,000
Gasoline
Variable!
$1,164,000
$1,185,000
$1,205,000
$3,244,000
$3,299,000
$3,354,000
$3,248,000
$3,301,000
$3,353,000
$3,405,000
$3,458,000
$3,510,000
$3,563,000
$3,615,000
$3,667,000
$3,720,000
$3,772,000
$3,825,000
$3,877,000
$3,929,000
$3,982,000
$4,034,000
$4,086,000
$4,139,000
$4,191,000
$4,244,000
$51,463,000
$27,315,000
Fixed]
$2,039,000
$2,039,000
$2,039,000
$3,060,000
$1,070,000
$1,023,000
$1,031,000
$25,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$10,000
$58,000
$2,000
$11,402,000
$9,926,000
Subtotal
$2,039,000
$2,039,000
$2,039,000
$3,060,000
$2,234,000
$2,208,000
$2,236,000
$3,269,000
$3,301,000
$3,364,000
$3,306,000
$3,303,000
$3,363,000
$3,463,000
$3,460,000
$3,520,000
$3,621,000
$3,617,000
$3,677,000
$3,778,000
$3,774,000
$3,835,000
$3,935,000
$3,931,000
$3,992,000
$4,092,000
$4,088,000
$4,149,000
$4,249,000
$4,246,000
$62,865,000
$37,241,000
Total Costs
$27,872,000
$27,872,000
$27,872,000
$96,827,000
$73,481,000
$73,486,000
$97,464,000
$47,336,000
$47,701,000
$83,944,000
$47,174,000
$47,373,000
$83,572,000
$49,429,000
$49,628,000
$85,827,000
$51,684,000
$51,883,000
$88,082,000
$53,939,000
$54,138,000
$90,338,000
$56,194,000
$56,392,000
$92,593,000
$58,449,000
$58,647,000
$94,848,000
$60,704,000
$60,903,000
$1,212,004,000
$743,396,000
Table 16 shows these costs on a per engine basis by combining the per engine costs shown
in Table 3 (variable costs) and Table 14 (fixed costs).
34
-------
Final Technical Support Document; HDOBD Final Rule
Table 16. Total Estimated OBD Costs per Engine for Over 14,000 Pound Applications
(2007 dollars)
Year
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
CY
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
Total $/engine
Diesel) Gasoline
$40
$40
$39
$139
$125
$123
$155
$61
$60
$107
$58
$57
$102
$58
$57
$100
$58
$57
$98
$57
$57
$96
$57
$57
$95
$57
$57
$93
$57
$57
$50
$49
$48
$71
$77
$76
$75
$70
$70
$70
$68
$67
$67
$68
$67
$67
$68
$67
$67
$68
$67
$67
$67
$67
$67
$67
$67
$67
$67
$67
3.2 Cost Analysis for 8,500 to 14,000 Pound Diesel Applications
We have used the same approach as described above for estimating costs associated with the
8,500 to 14,000 pound OBD requirements. Since we have had OBD requirements for many
years on such vehicles and engines the costs described here are incremental to past requirements.
For hardware costs, we anticipate no new costs since all sensors and actuators should already be
present and the computers should already be capable of handling the demands of OBD. We have
estimated some new R&D costs for the DPF monitor since our current DPF monitoring
requirement is to detect only a catastrophic failure while the final requirement would be more
difficult. This requirement will begin in the 2010 model year and the R&D associated with it
will be incurred over the years leading up to 2010.
We have estimated that five manufacturers will be making diesels in the 8,500 to 14,000
pound market. We have also used the same engineering and testing related costs for the under
14,000 pound requirements as used above for the over 14,000 pound requirements. This is being
conservative since most testing related costs, especially official emissions testing in a
35
-------
Final Technical Support Document; HDOBD Final Rule
certification test cell, is generally less costly on a chassis dynamometer than on an engine
dynamometer.
The analogous tables to those presented above are presented here. Table 17 shows the R&D
costs for OBD algorithm development and application. We have estimated no costs for
algorithm development since the same algorithm should suffice in both the over and under
14,000 pound categories. We have estimated costs for algorithm application since the engines
will be expected to operate slightly differently and, hence, we would expect some application
costs to be incurred. We have estimated costs for one new threshold monitor for the new DPF
monitoring requirement. As for the threshold monitor for NMHC catalyst monitoring, we
believe that manufacturers will test out of this monitor and conduct a functional monitor instead.
We have also estimated four and a half (on average) functional monitors associated with DPF
and NMHC catalyst monitoring, and for nine continuity monitors associated with DPF and
NMHC catalyst monitoring. We have also estimated costs for two engine families per
manufacturer with two variants each. The total costs are estimated at $5.3 million to be spread
over the four year period prior to the 2010 implementation date for the new monitoring
requirements.
36
-------
Final Technical Support Document; HDOBD Final Rule
Table 17. R&D Costs for OBD Algorithm Development and Application -
Diesel Applications Under 14,000 Pounds
(2007 dollars)
|A. Algorithm Development Costs
System Threshold Monitors
Engineer $
Technician $
Subtotal
| weeks/monitor) Cost/monitorl # of monitors)
30 $69,000
15 $21,000
$90,000 1
Total/Mfr| Total]
$90,000 $450,000
System Functional Monitors
Engineer $
Technician $
Subtotal
CCM Rationality Monitors
Engineer $
Technician $
Subtotal
CCM Continuity Monitors
Engineer $
Technician $
Subtotal
Total
|B. Application Costs to each Family
System Threshold Monitors
Engineer $
Technician $
Subtotal
System Functional Monitors
Engineer $
Technician $
Subtotal
CCM Rationality Monitors
Engineer $
Technician $
Subtotal
Total
20 $46,000
5 $7,000
$53,000 5
15 $34,000
1 $1,000
$35,000 5
2 $5,000
0 $0
$5,000 9
(weeks/monitor | Cost/monitorl # of monitors!
5 $11,000
10 $14,000
$25,000 1
5 $11,000
10 $14,000
$25,000 5
3 $7,000
1 $1,000
$8,000 5
$239,000 $1,195,000
$158,000 $790,000
$45,000 $225,000
$532,000 $2,660,000
Total/Family) #families/mfr|
$25,000 2
$113,000 2
$36,000 2
$174,000
Total/Mfr| Total|
$50,000 $250,000
$226,000 $1,130,000
$72,000 $360,000
$348,000 $1,740,000
|C. Application Costs to remaining Variants
Total
| | | Total/Variantl
$44,000
# variants/family | #families/mfr|
2 2
Total/Mfr| Total|
$176,000 $880,000
The R&D testing costs associated with the R&D effort that we have estimated are shown in
Table 18. These costs are estimated at $860,000 to be spread over the four year period prior to
the 2010 implementation date, and $260,000 to be spread over the four year period prior to the
2013 implementation date, for the new monitoring requirements.
37
-------
Final Technical Support Document; HDOBD Final Rule
Table 18. OBD R&D Test Cell Costs - Diesel Applications Under 14,000 Pounds
(2007 dollars)
A. R&D Test Cell Costs - Diesel
Monitor Algorithms test
System monitor - threshold
System monitor - functional
Rationality monitor
Subtotal
$ per year for 4 years
Cost for 2010
wks| # of monitors |
3
2 5
1 5
| total test wks|
9.0
4.5
Costs/mfr
$63,000
$32,000
$95,000
$23,750
# mfrs Total
5
5 $315,000
5 $160,000
$475,000
$118,750
Monitor Application to each engine family
factor 30%| # of monitors #families/mfr|
System monitor - threshold
System monitor - functional
Rationality monitor
Subtotal
$ per year for 4 years
0.9 2.0
0.6 5 2.0
0.3 5 2.0
| total test wks |
5.4
2.7
Costs/mfr|
$38,000
$19,000
$57,000
$14,250
#mfrs| Total
5
5 $190,000
5 $95,000
$285,000
$71,250
Monitor Application to each engine family variant
factor
System monitor - threshold
System monitor - functional
Rationality monitor
Subtotal
$ per year for 4 years
Total R&D Test Cell Costs
$ per year for 4 years
10%| # of monitors #families/mfr|
0.3 2.0
0.2 5 2.0
0.1 5 2.0
additional variants] total test wks|
1.0
1.0
1.0
1.8
0.9
Costs/mfr
$13,000
$6,000
$19,000
$4,750
$171,000
$42,750
# mfrs Total
5
5 $65,000
5 $30,000
$95,000
$23,750
$855,000
$213,750
B. R&D Test Cell Costs - Diesel
Monitor Algorithms test
System monitor - threshold
System monitor - functional
Rationality monitor
Subtotal
$ per year for 4 years
Costs for 201 3
wks| # of monitors |
3 1
2 0
1 0
| total test wks|
3.0
Costs/mfr
$21,000
$21,000
$5,250
# mfrs Total
5 $105,000
5 $0
5 $0
$105,000
$26,250
Monitor Application to each engine family
factor
System monitor - threshold
System monitor - functional
Rationality monitor
Subtotal
$ per year for 4 years
30%| # of monitors #families/mfr|
0.9 1 2.0
0.6 2.0
0.3 2.0
| total test wks |
1.8
Costs/mfr
$13,000
$13,000
$3,250
# mfrs Total
5 $65,000
5
5
$65,000
$16,250
Monitor Application to each engine family variant
factor 10%| # of monitors #families/mfr|
System monitor - threshold
System monitor - functional
Rationality monitor
Subtotal
$ per year for 4 years
Total R&D Test Cell Costs
$ per year for 4 years
0.3 1 2.0
0.2 2.0
0.1 2.0
additional variants] total test wks|
4.0
4.0
4.0
2.4
Costs/mfr|
$17,000
$17,000
$51,000
$12,750
#mfrs| Total]
5 $85,000
5
5
$85,000
$255,000
$63,750
38
-------
Final Technical Support Document; HDOBD Final Rule
For certification costs, we have first estimated costs for limit parts for certification
demonstration at $5,500. We have estimated the costs of aging limit parts at $293,000 per parent
engine using the same methodology as that described in section 3.1.2.b for aging of limit parts on
engines meant for over 14,000 pound applications. Here we have used values appropriate to
under 14,000 pound applications such as 110,000 mile useful life and weighted MPG of 14.
Table 19 shows the estimated costs for demonstration testing. Note that we have not estimated
costs for certification documentation since all 8,500 to 14,000 pound diesel applications are
already generating and submitting OBD certification documentation. We have also estimated no
costs for production evaluation testing since we do not have requirements for such testing in our
under 14,000 pound OBD program. We have estimated costs for a total of 10 engine families
with only one per manufacturer being demonstrated every three years, on average. The 30 year
net present value costs for certification demonstration testing are estimated at $10 million and
$6.2 million at a three percent and a seven percent discount rate, respectively.
The total costs for the 8,500 to 14,000 pound diesel applications are shown in Table 20. The
per vehicle numbers assume a two percent sales growth rate using an estimated sales number of
470,000 in 2006; entries of $0 represent costs less than $1 per vehicle. The 30 year net present
value of total costs are estimated at $16 million and $12 million at a three percent and a seven
percent discount rate, respectively. Importantly, these costs represent the incremental costs of
the additional OBD requirements, as compared to our current OBD requirements, for 8,500 to
14,000 pound diesel applications and do not represent the total costs for OBD.
39
-------
Final Technical Support Document; HDOBD Final Rule
Table 19. OBD Certification and Production Evaluation Testing Costs - Diesel Applications Under 14,000 Pounds
(2007 dollars)
Year
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
NPV@
NPV@
CY
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
3%
7%
Certfication Demonstration Testing Related
# of parent test Costs for Limit
engines Parts
0
0
0
5 $1,493,000
0
0
5 $1,493,000
0
0
5 $1,493,000
0
0
5 $1,493,000
0
0
5 $1,493,000
0
0
5 $1,493,000
0
0
5 $1,493,000
0
0
5 $1,493,000
0
0
5 $1,493,000
0
0
$8,595,000
$5,202,000
DDV Testing Total DDV
Costs Costs
$277,000 $1,770,000
$277,000 $1,770,000
$277,000 $1,770,000
$277,000 $1,770,000
$277,000 $1,770,000
$277,000 $1,770,000
$277,000 $1,770,000
$277,000 $1,770,000
$277,000 $1,770,000
$1,595,000 $10,189,000
$965,000 $6,168,000
Production Evaluation Testing Related
PE Testing - Scan Tool PE Testing - Monitors PE Testing - Ratios
# of engine PE Costs
families for # of OBD (incl vehicle # of monitoring PE Costs -
testing PE Costs Groups tested rental) groups tested PE Costs Total
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
0 00
Total
Certification &
PE Testing
Costs
$1,770,000
$1,770,000
$1,770,000
$1,770,000
$1,770,000
$1,770,000
$1,770,000
$1,770,000
$1,770,000
$10,189,000
$6,168,000
40
-------
Final Technical Support Document; HDOBD Final Rule
Table 20. Total Estimated OBD Costs - Diesel Applications Under 14,000 Pounds
(2007 dollars)
Year
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
NPV@
NPV@
CY
2006 $1
2007 $1
2008 $1
2009 $1
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
R&D Cert/PE Testing HE
,533,750
,533,750
,533,750
,597,500 $1,770,000
$63,750
$63,750
$63,750 $1,770,000
$1,770,000
$1,770,000
$1,770,000
$1,770,000
$1,770,000
$1,770,000
$1,770,000
3% $5,918,000 $10,189,000
7% $5,371,000 $6,168,000
irdware Total
$1,533,750
$1,533,750
$1,533,750
$3,367,500
$63,750
$63,750
$1,833,750
$1,770,000
$1,770,000
$1,770,000
$1,770,000
$1,770,000
$1,770,000
$1,770,000
$0 $16,107,000
$0 $11,539,000
Projected
Sales
470,000
479,400
488,800
498,200
507,600
517,000
526,400
535,800
545,200
554,600
564,000
573,400
582,800
592,200
601 ,600
611,000
620,400
629,800
639,200
648,600
658,000
667,400
676,800
686,200
695,600
705,000
714,400
723,800
733,200
742,600
$/vehicle
$3
$3
$3
$7
$0
$0
$3
$0
$0
$3
$0
$0
$3
$0
$0
$3
$0
$0
$3
$0
$0
$3
$0
$0
$3
$0
$0
$2
$0
$0
3.3 Updated 2007/2010 HD Highway Costs Including OBD
Table 21 shows the cost estimates for the 2007/2010 heavy-duty highway program. As
shown, the 30 year net present value cost at a three percent discount rate was estimated at $70
billion with $25 billion of that being engine related costs (these costs are in terms of 1999
dollars).
41
-------
Final Technical Support Document; HDOBD Final Rule
Table 21. Costs of the 2007/2010 Heavy-duty Highway Program
(All Costs in SMillions; 1999 Dollars)
Calendar
Year Year
1 2006
2 2007
3 2008
4 2009
5 2010
6 2011
7 2012
8 2013
9 2014
10 2015
11 2016
12 2017
13 2018
14 2019
1 5 2020
16 2021
1 7 2022
1 8 2023
1 9 2024
20 2025
21 2026
22 2027
23 2028
24 2029
25 2030
26 2031
27 2032
28 2033
29 2034
30 2035
NPV @ 3%
NPV @ 7%
Gasoline
Vehicles &
Diesel Engines Engines HD2007
HD2007 FRM FRM
-$80 $0
$1 ,266 $0
$1 ,321 $46
$1 ,072 $80
$1 ,520 $81
$1 ,225 $82
$1,133 $83
$1,157 $78
$1,180 $79
$1,141 $80
$1,156 $82
$1,159 $83
$1,182 $84
$1 ,205 $85
$1 ,226 $86
$1 ,247 $87
$1 ,268 $89
$1 ,288 $90
$1 ,307 $91
$1 ,326 $92
$1 ,344 $93
$1 ,362 $94
$1 ,380 $95
$1 ,398 $97
$1,415 $98
$1 ,432 $99
$1,450 $100
$1,467 $101
$1,484 $102
$1,500 $104
$23,721 $1,514
$14,369 $877
Diesel Fuel
$880
$1 ,786
$1 ,809
$1 ,904
$2,014
$2,128
$2,160
$2,192
$2,225
$2,258
$2,292
$2,327
$2,362
$2,397
$2,433
$2,469
$2,506
$2,544
$2,582
$2,621
$2,660
$2,700
$2,741
$2,782
$2,824
$2,866
$2,909
$2,953
$2,997
$3,042
$45,191
$26,957
Total Costs -
Engines, Fuel
$799
$3,052
$3,177
$3,056
$3,615
$3,434
$3,376
$3,427
$3,484
$3,480
$3,530
$3,568
$3,628
$3,687
$3,746
$3,804
$3,863
$3,921
$3,980
$4,039
$4,098
$4,157
$4,217
$4,276
$4,337
$4,397
$4,459
$4,521
$4,583
$4,646
$70,427
$42,203
Source: EPA420-R-00-026; Table V.D-1 & Appendix VI-B; December 2000.
The updated 2007/2010 program costs are shown in Table 22, which now include the new
OBD-related costs. The 2007/2010 program costs of $88 billion (2007 dollars) far outweigh the
OBD related costs of $1.2 billion. Note that the 2007/2010 program costs of $70 billion were
generated using 1999 dollars which, in 2007 dollars, is just under $88 billion. We have adjusted
the 1999 dollars to 2007 dollars using the Consumer Price Index. The CPI data are shown in
Table 23.
42
-------
Final Technical Support Document; HDOBD Final Rule
Table 22. Updated 2007/2010 Program Costs Including New OBD-Related Costs
(All costs in $Millions; 2007 dollars)
Calendar Diesel Engines Diesel Engines
Year Year HD2007 FRM >14KOBD
1 2006 -$100 $26
2 2007 $1,576 $26
3 2008 $1,644 $26
4 2009 $1,334 $94
5 2010 $1,892 $71
6 2011 $1,525 $71
7 2012 $1,410 $95
8 2013 $1,440 $44
9 2014 $1,469 $44
10 2015 $1,420 $81
11 2016 $1,439 $44
12 2017 $1,442 $44
13 2018 $1,471 $80
14 2019 $1,500 $46
15 2020 $1,526 $46
16 2021 $1,552 $82
17 2022 $1,578 $48
18 2023 $1,603 $48
19 2024 $1,627 $84
20 2025 $1,650 $50
21 2026 $1,673 $50
22 2027 $1,695 $87
23 2028 $1,717 $52
24 2029 $1,740 $52
25 2030 $1,761 $89
26 2031 $1,782 $54
27 2032 $1,805 $55
28 2033 $1,826 $91
29 2034 $1,847 $56
30 2035 $1,867 $57
NPV@ 3% $29,522 $1,149
NPV@ 7% $17,882 $706
Gasoline
Vehicles & Gasoline Diesel
Engines Engines >14K Applications
HD2007FRM OBD 8500-1 4KOBD Diesel Fuel
$0 $2 $2 $1,095
$0 $2 $2 $2,223
$57 $2 $2 $2,251
$100 $3 $3 $2,370
$101 $2 $0 $2,507
$102 $2 $0 $2,648
$103 $2 $2 $2,688
$97 $3 $0 $2,728
$98 $3 $0 $2,769
$100 $3 $2 $2,810
$102 $3 $0 $2,853
$103 $3 $0 $2,896
$105 $3 $2 $2,940
$106 $3 $0 $2,983
$107 $3 $0 $3,028
$108 $4 $2 $3,073
$111 $4 $0 $3,119
$112 $4 $0 $3,166
$113 $4 $2 $3,213
$114 $4 $0 $3,262
$116 $4 $0 $3,311
$117 $4 $2 $3,360
$118 $4 $0 $3,411
$121 $4 $0 $3,462
$122 $4 $2 $3,515
$123 $4 $0 $3,567
$124 $4 $0 $3,620
$126 $4 $2 $3,675
$127 $4 $0 $3,730
$129 $4 $0 $3,786
$1,884 $63 $16 $56,243
$1,091 $37 $12 $33,550
Total Costs -
Engines, OBD,
Fuel
$1 ,025
$3,828
$3,982
$3,904
$4,573
$4,349
$4,301
$4,312
$4,384
$4,416
$4,440
$4,489
$4,601
$4,638
$4,710
$4,821
$4,859
$4,933
$5,043
$5,081
$5,153
$5,264
$5,303
$5,379
$5,492
$5,531
$5,608
$5,723
$5,764
$5,843
$88,876
$53,279
43
-------
Final Technical Support Document; HDOBD Final Rule
Table 23. Consumer Price Index Data
Series Id: CUUROOOOSAO.CUUSOOOOSAO
Not Seasonally Adjusted
Area: U.S. city average
Item: All items
Base Period: 1982-84=100
Year
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
Jan
161.6
164.3
168.8
175.1
177.1
181.7
185.2
190.7
198.3
202.416
Feb
161.9
164.5
169.8
175.8
177.8
183.1
186.2
191.8
198.7
203.499
Mar
162.2
165
171.2
176.2
178.8
184.2
187.4
193.3
199.8
205.352
Apr
162.5
166.2
171.3
176.9
179.8
183.8
188
194.6
201.5
206.686
May
162.8
166.2
171.5
177.7
179.8
183.5
189.1
194.4
202.5
207.949
Jun
163
166.2
172.4
178
179.9
183.7
189.7
194.5
202.9
208.352
Jul
163.2
166.7
172.8
177.5
180.1
183.9
189.4
195.4
203.5
208.299
Aug
163.4
167.1
172.8
177.5
180.7
184.6
189.5
196.4
203.9
207.917
Sep
163.6
167.9
173.7
178.3
181
185.2
189.9
198.8
202.9
208.49
Oct
164
168.2
174
177.7
181.3
185
190.9
199.2
201.8
208.936
Nov
164
168.3
174.1
177.4
181.3
184.5
191
197.6
201.5
210.177
Dec
163.9
168.3
174
176.7
180.9
184.3
190.3
196.8
201.8
210.036
Annual
163
166.6
172.2
177.1
179.9
184
188.9
195.3
201.6
207.342
HALF1
162.3
165.4
170.8
176.6
178.9
183.3
187.6
193.2
200.6
205.709
HALF2
163. 7|
167. 8|
173. 6|
177. 5|
180. 9|
184. 6|
190. 2|
197. 4|
202. 6|
208.976J
Source: www.bls.gov/cpi.
44
-------
Final Technical Support Document; HDOBD Final Rule
References
1 Docket ID Number EPA-HQ-OAR-2005-0047 which can be found at www.regulations.gov.
2 Kato, N.; Hamada, Y.; Kurachi, H. Performance of Thick Film NOX sensor on Diesel and
Gasoline Engines. Soc. Automot. Eng. Tech. Pap. Ser. 1997, No. 970858.
3 Schar, C.; Onder, C.; Geering, H. Control of a Urea SCR Catalytic Converter System for a
Mobile Heavy Duty Diesel Engine. Soc. Automot. Eng. Tech. Pap. Ser. 2003, No. 2003-01-0776.
4 Kato, N.; Kokune, N.; Lemire, B.; Walde, T. Long Term Stable NOX Sensor with Integrated In-
Connector Control Electronics. Soc. Automot. Eng. Tech. Pap. Ser. 1999, No. 1999-01-0202.
5 Orban, I; Naber, S.; Sharp, C.; Khair, M.; McGill, R. Long-Term Aging of NOX Sensors in
Heavy-Duty Engine Exhaust. SAE Tech. Pap. Ser. 2005, No. 2005-01-3793.
6 Orban, J.; Naber, S.; Sharp, C.; Khair, M.; McGill, R. Long-Term Aging of NOX Sensors in
Heavy-Duty Engine Exhaust. SAE Tech. Pap. Ser. 2005, No. 2005-01-3793.
7 Schenk, C.; McDonald, J.; Laroo, C. High-Efficiency NOX and PM Exhaust Emission Control
for Heavy-Duty On-Highway Diesel Engines - Part Two. SAE Tech. Pap. Ser. 2001, No. 2001-
01-3619.
8 May, M.; Adelman, B. APBF-DEC Heavy Duty NOX Adsorber/DPF Project: Heavy-Duty
Linehaul Platform Final Project Meeting. Presented at the Heavy-Duty Final Project Meeting,
Rosemont, IL, April 27, 2004.
9 66 FR 5002 and 69 FR 38958, respectively.
10 Board approved as of July 21, 2005 (see 13 CCR 1971.1).
11 "Final Regulatory Analysis: Control of Emissions from Nonroad Diesel Engines," EPA420-R-
04-007, May 2004.
45
------- |