`
`SAE TECHNICAL
`PAPER SERIES
`
`2004-01-1292
`
`Technical Advantages of Urea SCR for Light-Duty
` and Heavy-Duty Diesel Vehicle Applications
`
`Christine Lambert and Robert Hammerle
`Ford Research & Advanced Engineering
`
`Ralph McGill
`Oak Ridge National Laboratory
`
`Magdi Khair and Christopher Sharp
`Southwest Research Institute
`
`Reprinted From: Diesel Exhaust Emission Control
`(SP-1860)
`
`400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-5760 Web: www.sae.org
`
`2004 SAE World Congress
`Detroit, Michigan
`March 8-11, 2004
`
`BASF-2006.001
`
`
`
`Downloaded from SAE International by James Kite, Tuesday, April 26, 2016
`
`All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
`transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise,
`without the prior written permission of SAE.
`
`For permission and licensing requests contact:
`
`SAE Permissions
`400 Commonwealth Drive
`Warrendale, PA 15096-0001-USA
`Email: permissions@sae.org
`Fax:
`724-772-4891
`Tel:
`724-772-4028
`
`For multiple print copies contact:
`
`SAE Customer Service
`Tel:
`877-606-7323 (inside USA and Canada)
`Tel:
`724-776-4970 (outside USA)
`Fax:
`724-776-1615
`Email: CustomerService@sae.org
`
`ISBN 0-7680-1319-4
`Copyright © 2004 SAE International
`
`Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE.
`The author is solely responsible for the content of the paper. A process is available by which discussions
`will be printed with the paper if it is published in SAE Transactions.
`
`Persons wishing to submit papers to be considered for presentation or publication by SAE should send the
`manuscript or a 300 word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE.
`
`Printed in USA
`
`BASF-2006.002
`
`
`
`Downloaded from SAE International by James Kite, Tuesday, April 26, 2016
`
`Technical Advantages of Urea SCR for Light-Duty
` and Heavy-Duty Diesel Vehicle Applications
`
`2004-01-1292
`
`Christine Lambert and Robert Hammerle
`Ford Research & Advanced Engineering
`
`Ralph McGill
`Oak Ridge National Laboratory
`
`Magdi Khair and Christopher Sharp
`Southwest Research Institute
`
`the nitrate
`(low oxygen) conditions,
`rich
`During
`compound releases NOx that is then reduced to N2.
`
`The reduction of NOx with either ammonia (NH3) or urea
`has been used extensively for stationary source emission
`control. NOx reduction is possible due to the high
`selectivity of the NH3 and NOx reaction to form elemental
`N2. The typical reaction scheme using urea is as follows:
`
`Copyright © 2004 SAE International
`
`ABSTRACT
`
`The 2007 emission standards for both light-duty and
`heavy-duty diesel vehicles remain a challenge. A level of
`about 90% NOx conversion is required to meet the
`standards. Technologies that have the most potential to
`achieve very high NOx conversion at low temperatures of
`diesel exhaust are lean NOx traps (LNTs) and Selective
`Catalytic Reduction (SCR) of NOx using aqueous urea,
`typically known as Urea SCR. The LNT has the
`advantage of requiring no new infrastructure, and does
`not pose any new customer compliance
`issues.
`However, Urea SCR has high and durable NOx
`conversion in a wider temperature window, a lower
`equivalent fuel penalty, and lower system cost. On a
`technical basis, Urea SCR has the best chance of
`meeting the 2007 NOx targets. This paper reviews the
`results of some demonstration programs for both light-
`and heavy-duty applications.
`
`
`
`INTRODUCTION
`
`Reduction of NOx in the lean exhaust gas of diesel
`engines is not trivial. The 2007 NOx standards represent
`a 90+% reduction from previous standards as shown in
`Table 1. Also shown are NMOG (Non-Methane Organic
`Gases), NMHC
`(Non-Methane Hydrocarbon), CO
`(Carbon Monoxide) and PM
`(Particulate Matter)
`standards. In general, the two leading technologies for
`very high NOx conversion are Lean NOx Traps (LNTs)
`and Selective Catalytic Reduction of NOx using aqueous
`urea (Urea SCR).
`
`LNTs contain an oxidative component such as Pt that
`oxidizes engine-out NO to NO2 and a storage component
`(typically an alkali metal salt) that forms a nitrate
`compound during
`lean (excess oxygen) conditions.
`
`urea decomposition:
`
`heat
`
`CO
`
`H2N
`
`NH2
`
`urea
`NOx reduction:
`4NO
`+ 4NH3 +
`6NO2+ 8NH3
`
`2NH3 NO+
`
`O2
`
`NO2
`
`7N2
`
`+
`6H2O
`4N2
`+
`12H2O
`2N2
`+
`
`3H2O
`
`
`
`+
`
`
`
`Urea is the preferred means of delivering ammonia
`onboard a vehicle because it is safely transported and
`can be easily injected as an aqueous solution. Urea has
`a very low toxicity and is widely used as a fertilizer. The
`mixture with the lowest freeze point (eutectic) with about
`32.5 wt% urea in water is preferred.
`
`There have been successful light-duty (LD) and heavy-
`duty (HD) testing of both LNT and Urea SCR on US
`emission cycles. Table 2 summarizes the most recent
`results.
`
`
`
`HNCO
`heat
`H2O
`
`+
`
`NH3
`
`CO2
`
`+
`
`NH3
`
`BASF-2006.003
`
`
`
`Downloaded from SAE International by James Kite, Tuesday, April 26, 2016
`
`Table 1
`
`2007 standards for LD and HD diesel vehicles.
`
`
`LD Tier 1 diesel (100k mi)a
`LD Tier 2 Bin 5 (120k mi)
`
`0.090 g/mi
`
`NMOG/NMHC
`
`CO
`
`NOx
`
`PM
`
`0.31-0.56 g/mi
`
`4.2-7.3 g/mi
`
`0.97-1.53 g/mi
`
`0.10-0.12 g/mi
`
`4.2 g/mi
`15.5 g/hp-hrb
`15.5 g/hp-hrb
`
`0.07 g/mi
`
`0.01 g/mi
`
`2.0 g/hp-hr
`0.20 g/hp-hrc
`
`0.10 g/hp-hr
`
`0.01 g/hp-hr
`
`HD 2004 MY
`
`0.5 g/hp-hr
`0.14 g/hp-hrc
`HD 2007 MY
`a Tier 1 standards vary according to vehicle weight.
`b HD CO standard is carried over from 1998 MY.
`c 2007 NMHC and NOx standards to be phased in from 2007-2010.
`
`Summary of reported results from recent diesel technology testing with fresh catalysts.
`
`Table 2
`
`NOx Conv.
`
`Tailpipe NOx Ref.
`
`Application Technology
`
`PC
`
`PC
`
`PC
`
`PC
`
`LDT
`
`LDT
`
`HDT
`
`HDT
`
`HDT
`
`HDT
`
`HDT
`
`Urea SCR
`
`Urea SCR
`
`LNT (DPNR)
`
`LNT (DPNR)
`
`Urea SCR
`
`LNT
`
`Urea SCR
`
`Urea SCR
`
`Urea SCR
`
`Urea SCR
`
`LNT
`
`Cycle
`
`FTP-75
`
`US06
`
`FTP-75
`
`US06
`
`FTP-75
`
`FTP-75
`
`HD FTP
`
`SET
`
`HD FTP
`
`SET
`
`HD FTP
`
`90+%
`
`90+%
`
`80+% [2]
`
`80+% [2]
`
`82%
`
`72%
`
`85%
`
`86%
`
`90%
`
`90%
`
`95%
`
`0.05 g/mi
`
`0.05 g/mi
`
`0.05 g/mi
`
`0.14 g/mi
`
`0.17 g/mi
`
`not reported
`
`0.86 g/hp-hr
`
`0.85 g/hp-hr
`0.22 g/hp-hra
`0.18 g/hp-hr
`
`0.13 g/hp-hr
`
`LNT
`
`SET
`
`94%
`
`0.12 g/hp-hr
`
`HDT
`Notes:
`All results obtained with very low sulfur diesel fuel (< 30 ppm).
`PC = Passenger Car
`LDT = Light-Duty Truck
`HDT = Heavy-Duty Truck
`DPNR = Diesel Particulate - NOx Reduction
`FTP-75: Federal Test Procedure for LD vehicles; cold-start three-bag cycle.
`US06: High speed, high load cycle for LD vehicles; part of Supplemental FTP.
`HD FTP: Heavy-Duty Federal Test Procedure; transient dynamometer cycle.
`SET: Supplemental Emission Test for HD; steady-state points.
`a Composite results for HD FTP (1/7 cold, 6/7 hot start).
`b Results obtained with catalyst preconditioning.
`
`1
`
`1
`
`3
`
`3
`
`4
`
`5
`
`6
`
`6
`
`7
`
`7
`8b
`8b
`
`
`
`
`
`
`
`for vehicle
`The case against using Urea SCR
`applications is twofold: (1) an infrastructure is required
`to deliver the reductant onboard the vehicle and (2)
`customer compliance is required to maintain adequate
`reductant for continuous, high NOx conversion. These
`issues make Urea SCR difficult to implement, but not
`impossible. In Europe, Urea SCR is the technology
`
`chosen by HD manufacturers to meet Euro IV and Euro
`V standards [9, 10]. Urea SCR does not have a direct
`fuel penalty because the engine can be tuned for
`optimum fuel consumption rather than minimum engine-
`out NOx, and durability is confirmed for over 500,000 km
`(>300,000 mi) [9]. Development of an aqueous urea
`infrastructure for LD diesels is more uncertain, but also
`
`BASF-2006.004
`
`
`
`Downloaded from SAE International by James Kite, Tuesday, April 26, 2016
`
`not impossible, especially if an approach like co-fueling
`of diesel and urea simultaneously is adopted [11,12].
`Such an approach requires no extra action by the
`customer other than normal refueling. In fact, refueling a
`vehicle equipped with either SCR or LNT emission
`control technologies can be transparent to the end user.
`
`This paper will focus on the technical advantages of
`Urea SCR over LNT that include wider temperature
`window for very high NOx conversion, higher resistance
`to sulfur poisoning, better thermal durability, lower fuel
`economy penalty, lower system HC emissions, lower
`greenhouse gas emissions, and lower system cost.
`
`
`
`EXPERIMENTAL
`
`LABORATORY TESTING & AGING CONDITIONS:
`UREA SCR
`
`Fundamental catalyst activity data were obtained at Ford
`using a laboratory-scale flow reactor system. A round
`sample core with dimensions of 1" diameter and 1.5"
`length was taken from a washcoated cordierite monolith
`obtained from a supplier. The catalyst was a base
`metal/zeolite
`type
`that did not contain vanadium.
`Simulated diesel exhaust gas flowed through the sample
`core at a space velocity of 30k h-1, measured at standard
`conditions. The composition of the feedgas is shown in
`Table 3. The composition of the inlet NOx was varied
`from 0 to 80% NO2 (balance NO).
`
`Table 3
`
`Composition of simulated diesel exhaust gas for SCR
`activity measurement.
`
`Component Concentration
`
`O2
`H2O
`CO2
`NOx
`
`NH3
`N2
`
`14%
`
`4.5%
`
`5%
`
`350 ppm
`
`350 ppm
`
`Balance
`
`
`
`tube
`temperature was maintained with a
`Catalyst
`furnace. An FTIR (infrared) spectrometer was used at
`the reactor outlet to measure NOx and NH3. Data were
`taken at catalyst temperatures from approximately 170 to
`600°C in order to cover the normal operating range
`expected on a diesel vehicle. NOx conversion was
`allowed
`to equilibrate before
`raising
`the catalyst
`temperature to the next level.
`
`SCR catalysts were aged to represent 120k miles via two
`modes:
`low
`temperature
`sulfur exposure and
`hydrothermal aging. For aging with sulfur, the monolith
`
`core was installed into a tube furnace that was set at
`350°C. N 2, O2, H2O and CO2 at the appropriate levels for
`diesel exhaust were flowed through the catalyst at a
`space velocity of 30k h-1. SO3 was generated from SO2
`over a Pt catalyst and added to the gas stream at a level
`to represent 120k miles of exposure over 24 hours for a
`range of diesel vehicles from PC to LDT. A future fuel
`sulfur level of 10 ppm was assumed for this calculation.
`After aging, NOx performance was measured without
`any attempts
`to remove sulfur
`from
`the catalyst.
`Hydrothermal aging was performed at 670°C and 30k h -1
`with N2, O2, H2O, CO2 and SO2 at the appropriate levels.
`An aging time of 64 hours was chosen to represent well
`over 120k miles based on Ford's projection of
`accumulated
`time at high
`temperature
`if a diesel
`particulate
`filter
`is periodically regenerating
`in
`the
`system.
`
`LABORATORY TESTING & AGING CONDITIONS: LNT
`
`Monolith cores of LNT samples were tested at Ford in a
`similar manner to the SCR catalysts except that the
`reactor feed gas was operated in a cyclic fashion of lean
`and rich durations. The feed gas concentration and the
`duration of lean and rich stages were controlled. A
`typical cycle for the LNT performance measurement
`contained 25 s of lean condition and 5 s of rich condition.
`The lean and rich feedgas conditions are shown in Table
`4. Total gas flow was 30k h-1, measured at standard
`conditions.
`
`Table 4
`
`Composition of simulated diesel exhaust gas for LNT
`activity measurement.
`
`Component
`
`Concentration
`(lean)
`
`Concentration
`(rich)
`
`CO
`
`H2
`C3H6
`NO
`
`O2
`CO2
`H2O
`N2
`Lambda
`
`
`
`500 ppm
`
`167 ppm
`
`300 ppm C1
`500 ppm
`
`10%
`
`5%
`
`5%
`
`Balance
`
`1.96
`
`4%
`
`1.3%
`
`5000 ppm C1
`500 ppm
`
`1%
`
`5%
`
`5%
`
`Balance
`
`0.90
`
`LNTs were aged under the conditions that may occur
`onboard a vehicle when removing sulfur from the catalyst
`(deSOx). The deSOx operation is typically done at high
`temperature
`(> 600°C) and predominantly
`rich
`conditions. For laboratory aging of LNTs, an inlet
`temperature between 600-700°C was chosen.
`
`
`
`BASF-2006.005
`
`
`
`Downloaded from SAE International by James Kite, Tuesday, April 26, 2016
`
`HC SCR
`
`Urea SCR
`
`LNT
`
`FTP-75
`
`200
`
`US06
`
`300
`
`400
`
`500
`
`600
`
`Catalyst Tem peratur e (°C)
`
`
`
`100
`
`90
`
`80
`
`70
`
`60
`
`50
`
`40
`
`30
`
`20
`
`10
`
`NOx Conversion (%)
`
`0
`100
`
`Fig. 1. Typical performance curves for slightly aged catalysts on LD
`test cycles. Typical underfloor temperature ranges for PC and LDT
`are shown for reference (see shaded areas). Space velocity was
`about 30k h-1, as measured at standard conditions, corresponding to
`a catalyst volume of about equal to the engine swept volume.
`
`MODELING TOOLS
`
`The Ford proprietary computer simulation tool, SIMTWC,
`is used
`to evaluate and compare
`the expected
`performance of aftertreatment configurations, saving
`experimental time and effort. A detailed description of
`this model has already been published and is readily
`available
`[13].
` SIMTWC
`is
`integrated
`into
`the
`Matlab®/Simulink® environment to provide a flexible
`means to build and configure a wide variety of simulated
`exhaust aftertreatment systems. It includes a library of
`exhaust
`system
`components
`that
`range
`from
`thermocouples and flanges to pipes (single and dual
`wall) and catalysts. Inputs to the model include
`geometric specifications of
`the exhaust system
`components, performance
`characteristics of
`the
`catalysts, exhaust flow rate, temperature and emission
`species levels.
`
`RESISTANCE TO SULFUR POISONING
`
`Sulfur is found in both diesel fuel and lube oils and
`generally has a negative effect on catalyst functionality.
`An LNT adsorbs sulfur in a similar way as it adsorbs
`nitrates. Sulfates are more stable and more difficult to
`remove than the nitrates. Sulfur must be removed
`periodically from the LNT (deSOx) in order to maintain
`high NOx conversion, and the time beween deSOx
`operations has been estimated to be about 3000 miles
`for a LD application. Urea SCR catalysts are inherently
`more resistant to sulfur since they do not possess the
`same type of storage component as an LNT. In Figure
`2, the laboratory NOx reduction performance of an SCR
`catalyst core that was exposed to 120k miles of flowing
`SO3 is shown. The exact amount of sulfur exposure (g/l
`catalyst) was calculated based on 10 ppm fuel sulfur and
`a catalyst volume of twice engine displacement. By
`using the most durable formulation available and a
`combination of NO and NO2 in the feedgas, the effect of
`sulfur on performance is minimized.
`
`
`
`
`
`RESULTS
`
`that Urea SCR has a wider
`found
`It has been
`temperature operating window, greater resistance to
`sulfur poisoning, greater thermal durability, lower fuel
`penalty, lower system HC emissions, lower greenhouse
`gas emissions, lower system cost and lower usage cost
`than LNT.
`
`TEMPERATURE OPERATING WINDOW
`
`The temperature operating windows at 30k h-1 for the
`best available LNT and Urea SCR catalysts as tested by
`Ford are shown in Figure 1. These performance data
`are for slightly aged catalysts. A typical HC SCR curve
`for diesel fuel and a Pt catalyst is shown for reference.
`Urea SCR has the widest operating window for >90%
`NOx conversion. Ammonia was used as the reductant
`here, but similar performance has been verified at Ford
`on engine exhaust systems using aqueous urea as the
`reductant. LNT has a narrower window that is shifted to
`temperatures higher than those that generally occur on
`the LD FTP-75 test. This means that the LNT must be
`heated by at least 100°C to give high efficiency on that
`test cycle. HD diesels typically operate with slightly
`higher
`temperature windows
`(250-500°C).
` LNT
`efficiency falls off drastically above 400°C.
`
`
`
`BASF-2006.006
`
`
`
`Downloaded from SAE International by James Kite, Tuesday, April 26, 2016
`
`4k mi SCR
`120k mi SCR
`4k mi LNT1
`120k mi LNT1
`120k mi LNT2
`
`100
`
`90
`80
`
`70
`60
`
`50
`
`40
`30
`
`20
`10
`
`NOx Conversion (%)
`
`4k mi
`120k mi
`
`100
`
`90
`
`80
`
`70
`
`60
`
`50
`
`40
`
`30
`
`20
`
`10
`
`NOx Conversion (%)
`
`0
`150
`
`200
`
`250
`
`300
`
`350
`
`400
`
`450
`
`500
`
`550
`
`600
`
`Catalyst Inlet Temperature (°C)
`
`
`
`Fig. 2. Resistance of SCR catalyst to poisoning by sulfur. The
`catalyst was exposed to SO3 at 350°C to simulate 120k mi exposure
`from diesel fuel containing 10 ppm sulfur. NOx conversion was
`measured with no attempt to remove the sulfur. Feedgas NOx
`contained equimolar NO and NO2 with total flow at 30k h-1, measured
`at standard conditions.
`
`THERMAL DURABILITY
`
`The deSOx operation for an LNT is done typically at 600-
`700°C under predominantly rich conditions, leading to
`the issue of thermal durability. A recent study looked at
`LNT
`thermal degradation at 510°C, not
`the most
`demanding of conditions [14]. In another study, repeated
`deSOx operations were performed at 700°C, leading to a
`steady decline in LNT performance [15].
`
`A comparison of Urea SCR and LNT thermal durability is
`shown in Figure 3. SCR performance is relatively
`unaffected by the hydrothermal treatment and in fact
`shows a slight improvement in the high temperature
`region. Again a combination of NO and NO2 is used in
`the SCR reactor feedgas. Thermal aging of a low
`temperature LNT formulation (LNT1) causes a dramatic
`decrease in performance at 200°C from about 90% to
`65% NOx conversion. Another aged LNT formulation
`(LNT2) shows similar low temperature performance to
`aged LNT1 with better high temperature performance.
`However, the performance of aged LNT2 is inferior to the
`aged SCR.
`
`
`
`
`
`
`
`0
`150 200 250 300 350 400 450 500 550 600
`Catalyst Inlet Temperature (°C)
`
`
`
`Fig. 3. Degreened (4k) and aged (120k) performance comparison of
`urea SCR and LNT in steady state flow reactors. The SCR catalyst
`was hydrothermally aged under lean conditions at 670°C to represent
`120k LD vehicle miles. LNTs were aged under deSOx conditions.
`Total flow was 30k h-1, measured at standard conditions.
`
`FUEL ECONOMY PENALTY
`
`LNT technology is expected to have a higher fuel
`economy penalty because rich conditions are required to
`remove and reduce NOx and SOx stored on the trap,
`and raise the temperature of the LNT during low
`temperature (< 300°C) operation. Urea SCR has no
`significant fuel penalty because it can provide high NOx
`conversion at low temperatures and does not use fuel to
`reduce NOx. An equivalent fuel penalty may be
`calculated to account for the energy used in the
`manufacture of urea.
`
`The expected performance curves for LNT and Urea
`SCR vs fuel penalty are shown in Figure 4 for the LD
`FTP-75. A similar curve for HC SCR using diesel fuel
`over a Pt catalyst is shown for reference. The NOx
`conversions are optimistic projections
`for the best
`formulations available without any type of cold start
`strategy. It is assumed that the NOx catalyst is located
`upstream of the PM filter, and similar feedgas conditions
`were used for each technology. A penalty of 3% is
`included for all three systems to account for the
`increased backpressure and periodic regeneration of a
`diesel PM filter. An equivalent fuel penalty for use of
`urea based on the energy required for manufacture is
`estimated to be approximately 0.2%. This is insignificant
`compared to the 7% penalty projected for LNT to achieve
`80% NOx conversion. Similarly, a recent study at
`Cummins found a 7% fuel penalty on a hot FTP-75
`simulation for 84% NOx conversion with an LNT, and
`11.6% fuel penalty for 89% NOx conversion on an FTP-
`72 cycle without LNT preconditioning [5].
`
`BASF-2006.007
`
`
`
`Downloaded from SAE International by James Kite, Tuesday, April 26, 2016
`
`Urea SCR +
`filter
`
`SYSTEM HC EMISSIONS
`
`
`
`LNT +
`filter
`
`HC SCR + filter
`
`For 2007, the LD NMHC (Non-Methane Hydrocarbon)
`emission standards must be met in addition to the NOx
`standards. For LNTs, this can be a challenge due to the
`frequent rich conditions and excess fuel used to heat the
`catalyst. On the other hand, Urea SCR catalysts for
`diesel vehicles do not contain precious metal, so an
`oxidation function must be added to the system to
`oxidize HC. In Table 5, data are shown from the recent
`LD diesel
`technology demonstrations described
`previously in Table 2 that reported NMHC.
`
`It is apparent that, at least with fresh catalysts, both
`systems can provide enough HC oxidation to achieve the
`FTP-75 standard. More challenging is the US06 cycle
`with its high speeds and loads. The DOC (Diesel
`Oxidation Catalyst) + Urea SCR system has essentially
`no NMHC emissions and 0.05 g/mi NOx, well below the
`0.14 g/mi standard. The DPNR system delivers 0.14
`g/mi NOx and 0.19 g/mi NMHC.
`
`GREENHOUSE GAS (GHG) EMISSIONS
`
`Diesel vehicles emit CO2 at a rate that is about 20-30%
`lower than comparable gasoline vehicles based on
`European vehicle data. Increased CO2 emissions can
`result
`from excess
`fuel
`injection
`to enhance
`the
`operation of the aftertreatment system. This is directly
`related to a decrease in fuel economy as explained
`above. Also lean NOx aftertreatment devices often do
`not completely reduce NOx to N2 and can form nitrous
`oxide (N2O) as a byproduct. It is well known that N2O is
`a GHG that has about 270 times the global warming
`potential of CO2. Therefore, even if N2O is formed in
`small amounts, it can have a potentially large impact.
`Both LNTs and Urea SCR catalysts can form N2O in
`varying amounts depending upon operating conditions
`and catalyst formulation. The total CO2 penalty may be
`calculated using an "equivalent CO2 penalty" from N2O
`formed on a driving cycle as shown below, where the
`base emission is without excess fuel injection:
`
`Total CO2 penalty = CO2 (g/mi) / base CO2 (g/mi) + N2O
`(g/mi) x 270 / base CO2 (g/mi)
`
`A breakdown of CO2 penalties for LNT and Urea SCR on
`a LD vehicle (FTP-75) is shown in Table 6. The LNT
`produces more GHG than a Urea SCR system operating
`at equivalent high NOx efficiency.
`
`
`
`
`
`3% Fuel
`Penalty
`from
`PM
`Filter
`
`100
`90
`80
`70
`60
`50
`40
`30
`20
`10
`0
`
`FTP NOx Conversion (%)
`
`0
`
`8
`6
`4
`2
`FTP Fuel Economy Penalty (%)
`
`10
`
`
`
`Fig. 4. Estimated fuel penalties of lean NOx reduction technologies vs
`NOx conversion for the LD FTP-75.
`
`The expected performance curves for LNT and Urea
`SCR vs fuel penalty are shown in Figure 5 for the HD
`FTP. Here it is assumed that the NOx catalyst is located
`downstream of the PM filter since cold-start is not heavily
`weighted for HD emission testing. An equivalent fuel
`penalty for use of urea is estimated to be approximately
`0.7%. Estimated fuel penalties for HD truck in Europe
`for an LNT system are 5% for Euro IV and 7% for Euro V
`[10]. Fuel consumption penalties for high NOx efficiency
`(>90%) on US HD cycles are around 7% [16].
`
`Urea SCR +
`filter
`
`LNT +
`filter
`
`3% Fuel
`Penalty
`from
`PM
`Filter
`
`100
`90
`80
`70
`60
`50
`40
`30
`20
`10
`0
`
`HD FTP NOx Conversion (%)
`
`0
`
`2
`
`8
`6
`4
`Fuel Economy Penalty (%)
`
`10
`
`12
`
`Fig 5. Calculated fuel penalties of lean NOx reduction technologies for
`HD engines vs NOx conversion.
`
`BASF-2006.008
`
`
`
`Downloaded from SAE International by James Kite, Tuesday, April 26, 2016
`
`Table 5
`
`Typical HC emissions on various test cycles.
`
`Application
`
`NOx Technology
`
`
`
`LD PC
`
`LDT
`
`LD PC
`
`
`
`
`
`DOC + Urea SCR
`
`DOC + Urea SCR
`
`LNT (DPNR)
`
`
`
`LD PC
`
`DOC+Urea SCR
`
`Cycle
`
`FTP-75
`FTP-75
`
`FTP-75
`
`FTP-75
`
`US06
`US06
`
`US06
`LNT (DPNR)
`LD PC
`* 2007 LD Tier 2 Bin 5 and SFTP emission standards.
`
`NMHC
`
`NMHC + NOx
`
`Ref.
`
`0.090 g/mi*
`0.012 g/mi
`
`0.066 g/mi
`
`0.07 g/mi
`
`---
`---
`
`---
`
`---
`
`---
`---
`
`---
`
`0.14 g/mi*
`0.05 g/mi
`
`0.33 g/mi
`
`
`
`1
`
`4
`
`3
`
`
`
`1
`
`3
`
`Table 6
`CO2 penalties for LNT and Urea SCR operating at high NOx efficiency on the LD FTP-75.
`
`
`
`GHG source
`
`LNT deNOx
`
`LNT deSOx
`
`LNT heating
`LNT N2O make
`LNT Total
`
`
`
`CO2 penalty (%)
`2.2
`
`0.2
`
`4.5
`
`0.9
`
`7.8
`
`Catalyst configurations assumed for HD applications are
`shown in Figure 7. The cost estimate for Urea SCR
`includes: aqueous urea injection system and SCR
`catalyst. The cost estimate for LNT includes: secondary
`fuel injection system and LNT catalyst. Again the PM
`filter is not included in the cost estimates and is assumed
`to be catalyzed and identical for both NOx systems.
`Geometric volumes of the Urea SCR catalyst and the
`LNT are assumed to be equal and twice the engine
`displacement. A means of heating the LNT may be
`required to achieve the needed cold-start performance
`on the HD FTP and is not included here. The estimated
`cost of the HD LNT system is about three times that of
`the Urea SCR system.
`
`
`
`GHG source
`
`Urea manufacture
`Urea SCR N2O make
`
`
`
`
`Urea SCR Total
`
`CO2 penalty (%)
`0.2
`
`2.0
`
`
`
`
`
`2.2
`
`
`
`SYSTEM COST
`
`System costs for Urea SCR and LNT systems can be
`estimated using historical
`cost
`information
`for
`washcoated catalysts and recent precious metal prices.
`Details
`for exact costs are proprietary
`to each
`manufacturer and are not discussed here; however,
`general
`comparisons
`can be made.
` Those
`knowledgeable can do similar calculations based on their
`own assumptions.
`
`In Figure 6, catalyst configurations assumed for LD
`applications are shown. The cost estimate for Urea SCR
`system includes: diesel oxidation catalyst, aqueous urea
`injection system and SCR catalyst. The cost estimate for
`LNT includes: diesel oxidation catalyst, secondary fuel
`injection system and LNT catalyst. The PM filter is not
`included for simplicity and is assumed to be identical for
`both systems. Geometric volumes of the Urea SCR
`catalyst and the LNT are assumed to be equal and twice
`the engine displacement. Estimated cost of the LD LNT
`system is about twice that of the Urea SCR system.
`
`
`
`BASF-2006.009
`
`
`
`Downloaded from SAE International by James Kite, Tuesday, April 26, 2016
`
`
`aqueousaqueous
`
`ureaurea
`
`
`
`DOCDOC
`
`SCR
`
`PM
`filter
`
`fuel
`
`
`
`DOCDOC
`
`LNT
`
`PM
`filter
`
`
`dieseldiesel
`
`engineengine
`
`
`
`
`dieseldieseldieseldiesel
`
`
`
`engineengineengineengine
`
`Fig. 6. Configurations used for LD system cost comparison.
`
`aqueous
`urea
`
`PM filter
`
`SCR
`
`fuel
`
`PM filter
`
`LNT
`
`
`dieseldiesel
`
`engineengine
`
`
`dieseldiesel
`
`engineengine
`
`
`
`
`
`Fig. 7. Configurations used for HD system cost comparison.
`
`The reasons why LNT systems are significantly more
`expensive than Urea SCR systems are:
`
`(1) The LNT contains precious metal while the Urea SCR
`catalyst contains base metal. The ratio of average
`precious metal prices to average base metal prices over
`the last year is roughly 2000:1.
`
`(2) The LNT precious metal cost is much greater than
`the urea injection system cost. A simple, air-assisted,
`reductant injection system has been developed by Ford
`that has a lower projected cost than previous systems
`[17]. The point at which the injection system equals the
`cost of the LNT occurs at an engine displacement much
`smaller than current PC diesel engines.
`
`(3) The LNT system cost / Urea system cost ratio is
`higher for HD than for LD because while the LNT and
`Urea SCR catalysts costs scale with engine size, the
`
`urea injection system cost stays largely fixed. Since the
`LNT contains precious metal and the SCR catalyst
`contains base metal, the cost differential between the
`two technologies becomes more disparate as engine
`size increases.
`
`USAGE COST
`
`Urea SCR – Light-Duty Usage Costs
`
`For a Tier 2 Bin 5 LD diesel vehicle, it is assumed that
`90% NOx conversion is required to meet the 0.07 g/mi
`NOx standard at 120k miles. This allows one to estimate
`the consumption of a 32.5 wt% aqueous urea solution at
`approximately 2700 mpg. An average cost of $1.50/gal
`for diesel fuel may be assumed, but the projected cost of
`aqueous urea for vehicle use is more uncertain. A
`recent study on potential HD urea consumption
`estimates that average future urea price could be as low
`as $1/gal [18]. For the LD case, no such studies
`currently exist because of the relative uncertainty in the
`future US LD diesel market. If a slightly higher and more
`conservative urea cost is assumed ($1.50/gal), then the
`total usage cost over 120k miles can be estimated for a
`range of fuel consumption rates as shown below in Table
`7. The usage cost for an LNT system is also shown,
`assuming a 7% fuel economy penalty for high NOx
`conversion as discussed in a previous section.
`
`Table 7
`
`Lifetime usage costs for a LD diesel vehicle with high
`efficiency NOx emission control.
`
`diesel fuel
`consumption
`(mpg)
`
`Urea SCR
`Usage
`Cost ($)
`
`LNT
`Usage
`Cost ($)
`
`Cost
`Differential
`($)
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`
`
`7267
`
`6067
`
`5210
`
`4567
`
`4067
`
`3667
`
`7704
`
`6420
`
`5503
`
`4815
`
`4280
`
`3852
`
`437
`
`353
`
`293
`
`248
`
`213
`
`185
`
`It is interesting to estimate the "break-even points" where
`Urea SCR and LNT may cost the customer the same
`amount to operate. Assuming aqueous urea costs
`$1.50/gal, the LNT has to have a fuel penalty of about
`1.5% to have the same usage cost as the Urea SCR
`system. Conversely, if the LNT fuel penalty is 7%,
`aqueous urea would have to cost $7-10/gal to have the
`same usage costs.
`
`The real-world cost differential between use of Urea SCR
`and LNT is variable according to how the customer
`actually drives the vehicle. The estimates made here are
`based on 0.7 g/mi NOx engine-out over the FTP-75 that
`would require 90% NOx conversion to meet the Tier 2
`
`BASF-2006.010
`
`
`
`Downloaded from SAE International by James Kite, Tuesday, April 26, 2016
`
`Bin 5 standard. Other driving cycles would result in
`different engine-out NOx
`levels
`requiring different
`conversion levels, but our conclusion is that the Urea
`SCR system will in general cost less to operate than an
`LNT system at 90% NOx conversion.
`
`Urea SCR – Heavy-Duty Usage Costs
`
`ratio of aqueous urea
`For HD applications, a
`consumption relative to fuel consumption of around 5%
`is quite common [6]. A similar result is reported for a
`NOx conversion efficiency of 71% [19]. Recent work
`with more advanced SCR formulations, as experienced
`in the APBF-DEC program, disclose that as high as 90%
`NOx efficiency may be achieved in HD applications using
`Urea SCR technology. This reduction is realized with a
`more efficient urea consumption of approximately 3% of
`the fuel consumption [20].
`
`For an average urea consumption over the lifetime of a
`HD vehicle, we may assume a ratio of urea/fuel
`consumption of 4%. Additionally, we assume that the
`vehicle
`lifetime
`is
`the EPA mandated durability
`requirement of 435k miles and that the vehicle fuel
`mileage is 6 mpg. Based on these assumptions, the fuel
`used over the lifetime of the vehicle is 72500 gallons and
`the urea consumed is 2900 gallons. Using the most
`likely scenario/pathway for urea distribution published in
`a recent study [18], the following table is suggested:
`
`Table 8
`
`Retail cost of aqueous urea [18].
`
`Aq. Urea Costs
`
`$/gal Remarks
`
`Production
`
`0.21
`
`Average of 0.12 – 0.30
`reported in reference*
`
`Distribution
`
`0.70 Most likely distribution
`pathway
`
`Dealer Mark-up
`
`0.08
`
`Average of 0.05 – 0.10
`reported in reference*
`
`Total
`Estimated
`
`
`0.99
`
`
`
`Based on the above table and other assumptions made
`here, the cost of a lifetime supply of urea for a HD
`vehicle is about $3000. In comparison, an LNT system
`with a 7% fuel penalty results in approximately $7600 in
`extra fuel costs. Additionally, the use of Urea SCR may
`allow for additional cost savings, because fuel economy
`is maximized through injection timing advance. A 6%
`fuel consumption advantage for a MY2007 HD Urea
`SCR and CDPF system over a MY2004 HD engine is
`noted in a recent publication [21]. A similar study reports
`a few percent improvement in BSFC (Brake Specific
`Fuel Consumption) after engine recalibration [6]. If a
`conservative 3% advantage in fuel consumption is
`assumed for Urea SCR, a lifetime fuel savings of 2175
`gal is realized, or about $3250. This savings in fuel not
`
`only more than pays for the additional price of the urea, it
`also brings the total lifetime usage cost differential
`between LNT and Urea SCR to $7350.
`
`CONCLUSIONS
`
`Urea SCR can provide the 90+% NOx conversion
`required for 2007 LD and HD NOx standards with the
`following technical advantages over LNTs:
`
`• wider operating temperature window
`greater durability
`
`•
`
`lower fuel economy penalty
`
`lower HC emissions
`
`lower greenhouse gas emissions
`
`lower system cost
`
`lower usage cost
`
`•
`
`•
`
`•
`
`•
`
`•
`
`
`
`ACKNOWLEDGMENTS
`
`This paper was prepared in part with the support of the
`U.S. Department of Energy, under Award No. DE-FC26-
`01NT41103.
` However, any opinions,
`findings,
`conclusions, or recommendations expressed herei