GB2424197A

GB2424197A - A method of adaptively controlling the regeneration of a lean nox trap

Info

Publication number: GB2424197A
Application number: GB0505272A
Authority: GB
Inventors: Robert Anthony Marshall; Trevor Charles Taylor
Original assignee: Ford Global Technologies LLC
Current assignee: Ford Global Technologies LLC
Priority date: 2005-03-15
Filing date: 2005-03-15
Publication date: 2006-09-20
Anticipated expiration: 2025-03-15
Also published as: GB0505272D0; GB2424197B

Abstract

A method is disclosed for adaptively controlling the regeneration of a lean NOx trap to compensate for ageing of the lean NOx trap. The method comprises of supplying reductant to the lean NOx trap based on a reductant profile predicted to correspond with the activity state of sites within the lean NOx trap so as to match the reductant supply to the activity state of the lean NOx trap. As the lean NOx trap ages the reductant profile is iteratively modified to ensure that the reductant supply profile is continuously adapted to match the changing state of the lean NOx trap. The embodiments disclose, for example, a three stage purging cycle. Initially a high level of hc concentration (step 1) is delivered based on engine operating conditions (this purges the high active sites), then a medium HC concentration for a longer duration (step 2) and then a lower hc concentration for the longest duration(step 3). This regime is intended to completely purge the trap and ensure the lower activity sites are fully utilised. The adaptive control accounts for catalytic ageing, indicated by an earlier than expected end of purge measured by an o2 sensor downstream, by applying different levels of adaption Level 1 - reduce the rate at which reductant is supplied Level 2 - decrease the duration of step 1 Level 3 - replace step 1 with levels of step 2 (i.e let step 2 last longer) First, second level and third level adaption is then applied to step 2. When no further NOx purge adaptation is possible then NOx conversion is maintained by reducing the lean time. Once the lean time reaches a minimum NOx trap failure is indicated.

Description

A METHOD FOR ADAPTIVELY CONTROLLING THE

REGENERATION OF A LEAN NOX TRAP

This invention relates to motor vehicles and in particular to the regeneration of a Lean NOx trap used to reduce emissions from an internal combustion engine fitted to a motor vehicle.

It is well known that an engine operated with air fuel mixtures leaner than stoichiometric requires a different type of after-treatment device to the conventional three way catalyst to convert engine out exhaust Nitrogen Oxides (NOx) emissions. One such after-treatment device is a Lean NOx Trap (LNT). This works by storing NOx under lean conditions and then releasing and converting it to harmless compounds under rich conditions during a process known as regeneration or NOx purge.

The start of a NOx purge is normally controlled either by modelling how much NOx has been stored in thern lean NOx trap or by measuring NOx break through into the tail pipe with a NOx sensor. In either case these determinations

S

* S. S control the frequency of NOx purges. A purge event is terminated when it is determined that the NOx trap is sufficiently regenerated to give the required efficiency from a model of lean NOx trap regeneration or from the detection of reductant breakthrough by sensing the air/fuel ratio Lambda' downstream from the lean NOx trap.

It is further known to vary the regeneration to take account of two factors which are commonly referred to as lean NOx trap ageing these are: 1) Thermal Ageing, which causes loss of Platinum group metal (PGM) surface area resulting slower reaction times because of the longer path between the PGM sites and the NOx storage sites; and 2) Poisoning, mainly by sulphur from fuel and oil, which occupy the NOx storage sites and will not be regenerated at normal lean NOx trap operating temperatures.

When a lean NOx trap is fresh the apparent mass of NOx stored is greater than when it is aged from use on an engine. However, in reality it is not the mass of stored NOx that changes with ageing but the amount of NOx that can be purged from the system, in other words the system gradually fails because it cannot be fully emptied. Since the actual storage capacity is largely unaffected by ageing it is possible, but not always practical, to restore much of the fresh capacity to an aged lean NOx trap by modifying the NOx purge cycle. A characteristic of lean NOx trap ageing is the rate of both NOx release and regeneration slows down, but not necessarily by the same degree. If NOx purge cycle is not adapted correctly to compensate for the change in reaction rates it will result in loss of useable storage capacity and reductant break through, either of hich will risk non-compliancy with emission regulations. S...

S S...

Prior regeneration methods compensate for the reduction in apparent storage capacity by gradually reducing the amount of reductant supplied to the lean NOx trap during regeneration to prevent hydrocarbon and carbon monoxide breakthrough and by reducing the time period during which the reductant is supplied. The reductant is provided by running the engine rich and the amount of reductant supplied is a product of how far the air/fuel ratio is rich of stoichiometric (known as the depth of purge') and the time the engine operates at this rich air/fuel ratio.

It is a disadvantage of this approach that the reductant supplied is not matched to the actual storage condition of the lean NOx trap and so it can result in a ILarger than necessary amount of fuel being used and in addition this type of methodology can actually increase the rate at which the lean NOx trap is perceived to be aging because insufficient time is provided to regenerate remote sites or less active sites within the lean NOx trap which then become inactive.

It is an object of this invention to more accurately control the regeneration of a lean NOx trap and to adaptively control the regeneration during the life of the lean NOx trap.

According to the invention there is provided a method for controlling the regeneration of a lean NOx trap having a number of active sites arranged to receive a supply of exhaust gasses from an internal combustion engine, the method comprising the steps of estimating the number and the relative activity levels of the active sites within the lean NOx trap, determining a reductant supply profile based upon the number and relative activity levels of the active sites to produce effective regeneration without excessive reductant breakthrough and supplying a reductant to the lean *::::* NOx trap during a regeneration event based upon the determined reductant supply profile. S. *S * S S * *

The reductant may be supplied to the lean NOx trap by operating the engine with an air/fuel ratio rich of stoichiometric or may be supplied by direct injection of a reductant to the lean NOx trap or upstream from the lean NOx trap.

The degree to which the engine is operated with an air/fuel ratio rich of stoichiometric may be derermined by the reductant supply profile.

The reductant supply profile may be such that the air/fuel ratio of the engine is varied during the regeneration event bounded by upper and lower air/fuel ratio limits.

The reductant supply profile may be such that the air/fuel ratio of the engine is changed between at least two levels of richness during the regeneration event.

The active sites may be grouped into two or more groups of sites having differing relative activity levels and the reductant supply profile is determined based upon the number of active sites in each of the groups of active sites.

The reductant supply profile may comprise a like number of supply phases as there are groups of active sites.

The air/fuel ratio at which the engine is operated during each phase may be based upon the relative activity of the sites being regenerated during that phase.

The air/fuel ratio at which the engine is oerated during each phase may be related to the relative activity s..., level of the sites being regenerated so that the air/fuel S...

ratio is lower for higher activity level sites than it is for lower activity level sites.

The duration of each phase may be dependent upon the number of sites to be regenerated and the relative activity level of the sites being regenerated.

The active sites may be grouped into three groups comprising of high activity level sites, lower activity level sites and low activity level sites.

In which case, the reductant supply profile may comprise the supply of reductant at a high rate during a first supply phase for a first period of time to regenerate the high activity level sites, the supply of reductant at a second lower rate during a second supply phase for a second period of time to regenerate the lower activity sites and the supply of reductant during a third supply phase at a third rate which is lower than the second rate for a third period of time to regenerate the low activity level sites.

The rate of reductant supply during at least one of the periods of time may vary during the respective period of time.

The method may further comprise predicting a NOx release rate from the number of sites and the relative activity levels of the sites and using the predicted NOx release rate to determine the reductant supply profile so as to match the supply of reductant to the predicted NOx release rate from the lean NOx trap and prevent excessive reductant breakthrough.

The step of estimating the number of active sites and activity levels of the active sites may be basedat least partially upon the thermal history of the lean NOx trap. S.. * . *S**

The step of estimating the number of active sites and activity levels of the active sites may be based at least partially upon a prediction for the number of active sites lOst due to sulphur poisoning. * . * S..

The method may further comprise adapting the reductant supply profile over time to compensate for ageing of the lean NOx trap.

Adapting the reductant supply profile over time may comprise increasing the air/fuel ratio at which the engine :s operated during the first supply phase so as to reduce 35:he amount of reductant supplied until the time between regeneration events is reduced.

Adapting the reductant supply profile over time may further comprise reducing the duration of the first supply phase until the time between regeneration events is reduced.

Adapting the reductant supply profile over time may further comprise increasing the duration of the second supply phase to return the time between regeneration events to its original value.

Adapting the reductant supply profile over time may further comprise merging the first supply phase with the second supply phase to form a modified second supply phase while supplying the same amount of reductant during the regeneration event.

Adapting the reductant supply profile over time may further comprise reducing the amount of fuel supplied during the modified second supply phase by reducing the duration of the second supply phase and increasing the duration of the third supply phase to compensate for the loss of reductant during the modified second supply phase. * S

Adapting the reductant supply profile over time may further comprise reducing the amount of reductant supplied during the regeneration event while increasing the duration of the regeneration event. *5 * S * *5S

The invention will now be described by way of example with reference to the accompanying drawing of which:- Fig.l is a table showing the relationship between temperature and lean NOx trap efficiency for a new lean NOx trap, after the lean NOx trap has been thermally aged, and the effect of sulphur poisoning on a lean NOx trap; Fig.2 is a graph of the data given in Fig.1; Fig.3 is a plot of air/fuel ratio versus time for a new lean NOx trap for two regeneration events according to the invention; Figs.4 to 10 are plots similar to that shown in Fig.3 but showing how the regeneration of the lean NOx trap is adaptively controlled; Figs.lla to lic are respectively plots of Hego voltage versus time, air/fuel ratio versus time and stored NOx versus time for a new lean NOx trap undergoing regeneration; and Figs.12a to 12c are respectively plots of Hego voltage versus time, air/fuel ratio versus time and stored NOx versus time for an aged lean NOx trap undergoing regeneration.

With reference to Figs.1 and 2 there is shown typical relationships between temperature and lean NOx trap efficiency for new and aged lean NOx traps. It can be seen that the effect of thermal ageing on a lean NOx trap is to increase the temperature at which the lean NOx trap becomes active. That is to say, the light-off temperature increases as the lean NOx trap ages.

On the other hand the effect of sulphur poisoning is to reduce the efficiency at higher temperatures. It is therefore possible to deduce from the efficiency of a lean NOx trap at various temperatures whether it has thermally aged and whether it is poisoned with sulphur and to discriminate between these two ageing effects. This information can be used to determine the regeneration strategy that needs to be applied to regenerate the lean NOx trap in the most effective manner.

During a regeneration event the reductants cause six main reactions to take place: 1. They reduce the oxidized Platinum Group Metals, (PGM) sites back to metal or a lower oxidation state; 2. Oxygen is removed from the lean NOx trap environment shifting the equilibrium, making the stored NOx thermally unstable in the presence of PGM's; 3. Water gas shift of CO and H2O over Rh to produce H2.

This has greater mobility to reach remote storage sites; 4. Nitrate on storage sites close to PGM's react with reductants to release NOx; 5. Reduced Rh sites react with NO and reductants in the conventional manner to produce N2; and 6. Any Oxygen storage component (OSC) present is partially reduced.

The ageing of a lean NOx trap will affect the reactions 3, 4, 5 and 6. * * * *** 0

With a new lean NOx trap a NOx purge cycle consists of making the engine run rich for a short period to supply a measured amount of reductants sufficient to react with all the stored NOx with minimal excess reductant break through.

Initially there is a rapid release of stored NOx from sites that are easy to access, as these site become depleted the rate of NOx release slows down. Therefore the supply of reductants in the NOx purge profile has to be matched to the released NOx to prevent either NOx or CO/HC break through.

This is done by having the engine run as close to its rich stability limit for a few engine cycles and then gradually becoming less rich until an upper air/fuel ratio vaThe is reached. The amount of reductant is calculated from the air and fuel mass flow rate multiplied by the number of engine cycles so that the NOx purge is not a fixed time but is a function of the engine speed and load.

The rich or lower air/fuel ratio limit is determined by engine combustion stability and the upper air/fuel ratio limit is the leanest mixture that can be reliably used for regeneration and is estimated from catalyst chemistry and the residual oxygen in the exhaust gas. It will be appreciated that if reductant is supplied at an air/fuel ratio just rich of stoichiometric then the preferential reaction may not be decomposition and conversion of stored NOx but reduction of PGM oxides and their reoxidation from the OSC. The minimum reductant level has to be greater than this oxidation overhead.

The NOx purge or regeneration is terminated when either a lean NOx trap model determines that the lean NOx trap is empty or when an oxygen sensor located downstream from the lean NOx trap switches from lean to rich.

When aged the lean NOx trap requires a different NOx *::::* purge profile to when it was fresh. This is because the number of storage sites rapidly accessible to regeneration will have reduced due to PGM sintering and interactions between the storage medium and the washcoat. Simply reducing the duration or depth of NOx purge to compensate for loss of regeneration sites as previously proposed will not give optimum performance since it is the rate of reaction that has slowed down. The purge has to be managed carefully to allow time for the reductants to reach sites that have become less accessible.

During a purge, NOx is rapidly released from those sites in contact with PGM's and the greater the separation to storage site to a PGM site the slower will be the release of NOx. Therefore to obtain optimum performance the degree of initial richness need to be reduced and the delivery time 10 - increased. This will compensate for the smaller and slower rate of release of usable NOx storage. Since the duration of the NOx purge cycle is increased there is now greater competition for reductants from the OSC, if present, and this has to form part of the calculation for NOx regeneration.

The method proposed uses a prediction of the NOx release rate from the number of sites and the relative activity levels of the sites and uses the predicted NOx release rate to determine a reductant supply profile so as to match the supply of reductant to the predicted NOx release rate from the lean NOx trap and prevent excessive reductant breakthrough.

The step of estimating the number of active sites and activity levels of the active sites is based partially upon the thermal history of the lean NOx trap which provides an indication of thermal ageing and partially upon a prediction for the number of active sites lost due to sulphir poisoning. S...

S S

This prediction of active sites can be done by using a look up table determined from experimental results for a similar lean NOx trap using the thermal history and the operating history of the engine or can be obtained using real time measurements of the lean NOx trap efficiency to determine the nature of the ageing and combining this with a look up table relating lean NOx trap efficiency to temperature (as shown in Fig.2) and NOx production over time during a regeneration event.

This invention also provides a method for controlling the adaptation of a NOx purge in order to maintain optimum regeneration efficiency with minimal break through of reductants throughout the life of the lean NOx trap.

- 11 - The method used for age compensation of NOx purge uses an adjustment of both richness and duration of purge in relation to oxygen sensor switching. The objective of this method is to maximize the amount of NOx that can be stored after regeneration and minimize the mass of the reductant used to effect the NOx purge.

The base NOx purge calibration is set up for a fresh system. When fresh the lean NOx trap will have the largest quantity of storage sites and the quickest regeneration time. The purge when the lean NOx trap is fresh will have the shortest time and the richest depth. The ageing process is slow compared to frequency of lean and NOx purge cycles so the strategy can make small adjustments to the NOx purge richness and duration and test against the rolling average of regeneration frequency. The objective is to make the regeneration frequency a minimum therefore maximizing lean NOx trap NOx storage while at the same time minimizing reductant break through.

-

The profile and duration of the NOx purge, which is se's basically a rich operation of the engine to produce S...

reductants, CC, H2 and HC, is matched to the release rate and quantity of NOx stored and is based upon the activity levels of the sites with the lean NOx trap. Such that all the stored NOx is released and converted without any excess reductants breaking through or any NOx slip to the tail pipe.

Within the lean NOx trap the activity level of all of the active sites is not uniform and will vary from highly active sites to low activity sites. Although this distribution of activity levels could be used to control the profile for the NOx purge this would be more difficult to achieve and so it is more convenient to group the sites based upon their activity level and to provide figures for the number of sites in each activity level group. For the - 12 - purposes of this example three activity level groupings are used but it will be appreciate that more or less groupings could be used if required.

When the lean NOx trap is new there will be a number of highly active sites, a number of moderately active sites and a number of sites of low activity. The value for each of these groupings is stored in an electronic device and is corrected over the life of the lean NOx trap to allow for ageing. In general as the lean NOx trap ages the number of highly active sites will firstly reduce converting to moderate activity sites until none remain and then the number of moderately active sites will reduce converting to low activity sites.

The supply of reductant is chosen to match the three groupings and so initially there will be three phases of reductant supply, a first phase to supply reductant to the highly active sites which uses an air/fuel ratio that is at or close to the rich stability limit of the engihe, a second phase to regenerate the moderately active sites and uses a slightly less rich mixture to supply reductant at a lower rate and a third phase to regenerate the low activity sites which uses an air/fuel ratio that is at or close to the upper air/fuel ratio limit.

One adaption method comprises the following steps:- *.S.,. * S

Firstly, to reduce how rich the initial rich step of the purge is, that is to say the depth of purge, until the purge frequency just starts to rise from its average value.

Secondly, to increase the length of the initial purge step until the purge frequency is a minimum. This is compared to the original purge frequency. If smaller this becomes the new setting and a new historical rolling average - 13 - value is established. If greater then the strategy tracks back to the original setting.

If the second step returns a new value, then a third step is to repeat the reduction of initial rich step until the purge frequency is a minimum.

The steps 1 to 3 are then repeated until stability is reached within a given tolerance band.

Initially, the purge profile consists of three air/fuel ratio steps or ramps from an initial rich value to a final holding value in order to replicate the distribution of active sites in the lean NOx trap. It will be appreciated that in order to regenerate highly active sites it is desirable to supply reductant at a high rate for a short period of time but for low activity sites it is required to supply reductant for a relatively long period of time at a slow rate in order to prevent HO break through while maximising the number of sites that are regeneraed. * S. * * .

S

The adaptation method continues as the lean NOx trap S...

ages and the three steps above are repeated on each level until the rich setting of the first phase reaches the value of its neighbour at which point the steps combine. This point is now treated as the initial rich step and the order of logic for age compensation is repeated.

S

*5S*sS

S S

Near the end of the lean NOx trap life the upper air/fuel ratio value will be reached. At this point no further ageing compensation can be applied through purge adaption. However, compensation can still be applied through adjustment of the lean period to change NOx efficiency and fuel economy.

The size and update frequency of each step is predetermined in the calibration to make sure the system - 14 - remains stable. The number of iterations of each level in the order of logic is set in the calibration before the lower levels are enabled. When the number of iteration in the final level is complete the cycle returns to the top level and the learning process starts again. This is done at a number of speed/load points frequently used by the customer. This can be pre-determined at the calibration stage and fixed or can use a frequency algorithm and be developed for individual users from their driving profile.

The end of NOx purge is determined either by a model of lean NOx trap regeneration or by an oxygen sensor switching from a lean condition to rich.

Referring now to Figs. 3 to 10 there are shown plots of air/fuel ratio versus time for a lean NOx trap regenerated according to this invention as the lean NOx trap ages. On each plot an air/fuel ratio (Lambda) equivalent to stoichiometric is indicate by a horizontal line s-s when the plot line is above this s-s line the mixture is weak of stojchiometrjc and when it is below the line the mixture is rich of stoichiometric For a given amount of stored NOx there is a calculatable reductant requirement to bring about a stoichiometric regeneration. This amount is independent of activity of the storage site.

The activity of the storage sites determines the rate at which the reductant is required. If the reductant is supplied too fast the reductant passes through the lean NOx trap before it can react, but if it is supplied too slow the NOx passes out of the lean NOx trap before it finds sufficient reductant to react with.

To prevent break through, the reductant flow to rate needs to be matched to the NOx release rate. Since it is not - 15 - possible to rapidly change the exhaust mass flow into the lean NOx trap during a purge, as this is coupled to engine operation, the alternative approach is to change the reductant concentration or the rate at which reductant is supplied.

This is done by changing the exhaust air/fuel ratio either in a series of steps or as a progressive ramp, see fig.3 where three phases or steps of fuel supply are used.

In addition to the reductant required to release and convert the stored NOx there is an overhead associated with changing the environment in the lean NOx trap from oxidizing to reducing. Part of this overhead is a discrete quantity, that is to say, the change of oxidation state of the PGM but the majority is continuous for the duration of the purge as in the reduction of excess exhaust gas oxygen. This means that per unit mass of NOx conversion a short deep rich purge will require less fuel than long shallow rich purge.

The characteristic of lean NOx trap thermal ageing is to slow the speed of regeneration reactions. This requires the purge to be longer.

To maintain lean NOx trap storage capacity but avoid excessive break though and fuel consumption the purge adaption needs a higher level of management than a simple change in purge duration.

As previously mentioned, known purge adaption techniques concentrate only on those storage sites that easily regenerate. This is done by either making the short deep rich purge even shorter or by reducing the lean time so that the purge frequency is increased. As a consequence the working storage capacity of the lean NOx trap is prematurely reduced.

- 16 - The following method for adapting a purge addresses this issue.

1) From Base NOx purge setting, a purge adaption is triggered by a NOx Conversion Model and /or by an earlier than expected end of purge as indicated by the output from an oxygen sensor located downstream from the lean NOx trap; 2) A first level of adaption is applied to increase the Step 1 lambda value, that is to say increase the air/fuel ratio or reduce the rate at which reductant is supplied; 3) A second level of adaption is to decrease the duration of Step 1; 4) A third level of adaption is to add/subtract Step 1 fuel quantity change to Step 2. Typically this will be an addition so that Step 2 duration will be increased; 5) The first and second level adaption is applied to Step 2; p... * * *1pI

6) The third level of adaption is to add/subtract the Step 2 fuel quantity change to Step 3. Typically this will be an addition so that the Step 3 duration will be increased; p. * . * ,I.

a 7) The NOx purge adaption fuel penalty for a given conversion efficiency is compared to that derived from shortening the lean time and the most fuel efficient option is chosen; 8) The steps 1 to 7 is repeated continuously for the life of the LNT; - 17 - 9) When no further NOx purge adaption is possible due to all purge being delivered in Step 3, then the NOx conversion is maintained by reducing the lean time; and 10) When the lean time is reduced to a minimum then further ageing of the lean NOx trap will result in a drop in conversion efficiency and when the conversion efficiency drops to a trigger level a MIL lamp is set to signal that a lean NOx trap failure has occurred.

In Fig.3 the reductant supply profile for a new lean NOx trap is shown (Case 1) . It can be seen that the reductant supply profile comprises of three phase or steps.

In the first phase (step 1) a very rich air/fuel ratio close to the engine rich stability limit is supplied for a short period of time to regenerate highly active sites, then in the second phase (step 2) a less rich air/fuel ratio is used for a longer period of time to regenerate moderately active sites and this ramps up to an upper air/fuel ratio used to regenerate low activity sites. It will be appreciated that throughout the period when a rich air/fuel ratio is being supplied all sites within the lean NOx trap can be potentially regenerated.

In Fig.4 (Case 2) a reduction in the rich air/fuel * ** ratio used in step 1 has resulted in the frequency of the :.: purge events from being reduced as insufficient reductant S...

has been supplied to the lean NOx trap. In effect the :.. reduction in the richness of the air/fuel ratio was too high * 30 and compensation is required. S..

In Fig.5 (Case 3) the air/fuel ratio (Lambda) in step 1 * has been reduced slightly to return the purge event frequency to that shown in Fig.3 so as to restore conversion efficiency.

- 18 - In Fig.6 (Case 4) the duration of step 1 is incrementally reduced until the purge event frequency is reduced.

In fig.7 (Case 5) a second level of adaption is applied and the step 1 duration is increased to a time between that shown in Figs.5 and 6.

In Fig.8 a third level of adaption is shown (Case 6) The fuel differencefrom step 1 before adaption (Case 1) to after adaption (Case 5) is added or subtracted from the second phase (step 2) . At this stage of adaption the total fuel quantity used to regenerate the lean NOx trap is unchanged but it is redistributed between steps 1 and 2.

In Fig. 9 a third adaption occurs (Case 7) and the fuel difference from step 2 is added or subtracted to step 3.

Because the air/fuel ratio of step 3 is fixed at the upper air/fuel ratio limit the fuel difference has to be accommodated by increasing the length of the pure time.

The method operates so as to maintain the lean time as it was for case 1.

It will therefore be appreciated that unlike previous control methodologies this method acts so as to increase the : duration of the purge as the lean NOx trap ages because this gives the less active sites more time to be regenerated.

In Fig.1O it can be seen that step 1 has been merged * 30 with step 2 and the adaption method continues to reduce the * S. amount of reductant being supplied by reducing the lambda of the combined steps 1 and 2 while increasing the purge time. ***

When the adaption method can no longer maintain the desired efficiency the frequency of the purge events is shortened until a fuel economy limit is reached. From this point further lean NOx trap ageing will result in a loss in - 19 - efficiency and when an efficiency limit is reached a malfunction indicator lamp is illuminated to indicate lean NOx trap failure.

It will be appreciated that in use the method is performed using an electronic processor or electronic control unit programed to perform the control of the air/fuel ratio of the engine and to implement the adaption method. Pseudo coding for a program to effect this adaptive method in an electronic processor is as follows:- [Primary Loop} If NOx (efficiency < threshold) and (mu light off) then go to Adapt purge procedure If purge fuel economy < threshold then Turn on mu light End if loop End if loop End primary loop [Adapt purge Procedure] Calculate fuel hit, using modified profile procedure Calculate fuel hit, increased frequency Set new purge= one with best fuel economy End Procedure [Calculate fuel hit, modified profile procedure] Set Start purge lambda segment 1= old start purge lambda segment 1 S...

Adapt purge segment (1st segment) For 2nd to last purge time segment If lambda<> lean value then Start purge lambda = lambda_value (from exit of previous segment) * S5 Adapt_purge_segment End if loop Next segment End Procedure (Adapt purge segment procedure] - 20 - Start loop Increase lambda in purge segment Measure new efficiency If new efficiency better or equal than present_efficiency then Set present_efficiency = new_efficiency Store lambda value End if Set present conversion efficiency=stored conversion efficiency End loop when new efficiency not better than present efficiency or lambda>lambda purge limit If lambda>lambda purge limit then set lambda=lean lambda End Procedure [Calculate fuel hit, increased frequency procedure] Set New lean duration= stored lean duration Start loop Set New lean duration = New_lean_duration * factor If conversion efficiency improved then Store New_lean_duration Endif Loop until (conversion efficiency not improved) or (fuel economy worse than for modified profile) End Procedure With reference to Fig.lla there is shown a plot of Hego or oxygen sensor voltage supplied by a Hego sensor located at the outlet of a new lean NOx trap during a regeneration * ** S or purge event in accordance with this invention. It can be seen that the voltage is initially zero indicating an oxidizing environment and then rises sharply to a plateau indicating that the environment is stoichiometric before rising again indicating that a reducing environment is breaking through. When the Hego voltage reaches a calibratable trigger voltage the purge is terminated.

Fig. lib shows the air fuel ratio in the new lean NOx trap during the same period of time and clearly indicates - 21 - that the air/fuel ratio is not constant throughout the purge event but is profiled to fit the expected activity of the lean NOx trap. At the start of the purge during a first phase of reductant supply a very rich mixture close to an engine operating rich stability limit is used to regenerated all of the highly accessible high activity sites and then during a second phase of reductant supply this is reduced to a less rich mixture in order to regenerate the slightly less accessible or moderate activity sites and then finally during a third phase of reductant supply the air/fuel ratio is further weakened to a holding value at or close to the maximum air/fuel ratio that regeneration will occur in order to regenerate any remaining remote or low activity sites.

Fig.llc shows the volume of NOx stored in the new lean NOx trap and it can be seen that NOx is liberated quickly during the early stages of the purge event and then is liberated less quickly towards the end of the event as by then all of the highly active and moderately active sites have been regenerated and so it is only remote sites or sites of low activity that are being regenerated.

Fig.12a there is shown a plot of Hego voltage supplied by a Hego sensor located at the outlet of an aged lean NOx trap during a regeneration or purge event in accordance with this invention. It can be seen that the voltage is initially zero indicating an oxidizing environment and then rises sharply to a plateau indicating that the environment :.. is stoichiometric before rising again indicating that a reducing environment is breaking through. When the Hego voltage reaches a calibratable trigger voltage the purge is : . terminated. However, in this case it will be noticed that * the change from a stoichiometric environment to a reducing S.....

occurs much earlier during the purge event than is the case with a new lean NOx trap.

- 22 - Fig.12b shows the air fuel ratio in the aged lean NOx trap during the same period of time and clearly indicates that the air/fuel ratio is not constant throughout the purge event but is profiled to fit the expected activity of the lean NOx trap. However, due to the fact that in an aged lean NOx trap there are no longer any high activity sites the reductant supply profile has been modified by the adaption method so that at the start of the purge a very rich mixture close to an engine operating rich stability limit is no longer supplied but instead a less rich mixture is used which is similar in air/fuel ratio to that used in the second phase of fuelling for a new lean NOx trap in order to regenerate the less accessible or less active sites. In effect the first phase of reductant supply has been adapted out as the number of highly accessible or highly active sites has reduced and so the reductant supply profile now only has two supply phases with in this case a ramped change linking them although this could be a step change. In the second of these two remaining phases the air/fuel ratio is further weakened to a holding ralue at or close to the maximum air/fuel ratio that regeneration will occur as was used in the third phase for a new lean NOx trap.

Fig.l2c shows the volume of NOx stored in the aged lean NOx trap. It can be seen that as before NOx is liberated quickly during the early stages of the purge event and then S..

is liberated less quickly towards the end of the event. The quantity of NOx stored is the integration of the area under the NOx release curve. This profile changes from fresh to aged lean NOx trap and the objective is to keep the area constant whenever practical to do so. When the lean NOx trap is aged the height of the peak is less but the length * is longer so the integration area can be close to that of a fresh lean NOx trap. When a lean NOx trap is severely aged the area is reduced because it is not practical to run such - 23 - a long purge as would be required to regenerate the very slow storage sites.

It will be appreciated that a purge event as shown in Figs. ha to c and 12a to c will be preceded and followed by periods of lean engine operation and for a typical engine a duty cycle of 2 seconds of purge to 60 seconds of lean running is typical.

It will be appreciated that for convenience the invention has been described in relation to a reductant supply profile that has only three phases when the lean NOx trap is new but it will be appreciated that more than three phases could be used if required. The object of this profiled reductant supply is to match the reductant supply with the distribution of active sites in the lean NOx trap so as to regenerate the lean NOx trap as effectively as possible using the minimum mass of reductant irrespective of the aged condition of the lean NOx trap. It will be appreciated that other reductant supply profiles that perform this function may also fall within the scope of this invention.

Although the invention has been described by way of example in which reductant is supplied by changing the : *. air/fuel ratio such that the air/fuel ratio is lower than a stoichiometric air/fuel ratio that is to say a rich mixture or rich lambda is used it will be appreciated that a source of reductant could be stored and then supplied directly to * 30 the lean NOx trap or be added to the exhaust gas flow 0* upstream from the lean NOx trap without necessarily changing :. the air/fuel ratio of the engine (See US-A-2004098972 and FP-A-1331373 for examples of reductant injection systems) In this case the supply of reductant is changed in a similar manner to that described above except that it would be the rate at which the reductant is supplied that would be changed and not the air/fuel ratio of the engine.

- 24 - That is to say when reductant is required to regenerate highly active sites it would be supplied at a fast rate and the rate of supply would be lower for moderate activity S sites and even lower for the low activity sites. As with the other examples given the supply of reductant is based upon a reductant supply profile and it adaptively changed to account for lean NOx trap ageing.

It will be appreciated by those skilled in the art that although the invention has been described by way of example with reference to one or more embodiments it is not limited to the disclosed embodiments and that modifications to the disclosed embodiments or alternative embodiments could be constructed without departing from the scope of the invention. * ** S.. * *SSS * .

* S* S *. S. * S S * . **S * I S **S I.,...

I S

Claims

- 25 - Claims 1. A method for controlling the regeneration of a lean NOx

trap having a number of active sites arranged to receive a supply of exhaust gasses from an internal combustion engine, the method comprising the steps of estimating the number and the relative activity levels of the active sites within the lean NOx trap, determining a reductant supply profile based upon the number and relative activity levels of the active sites to produce effective regeneration without excessive reductant breakthrough and supplying a reductant to the lean NOx trap during a regeneration event based upon the determined reductant supply profile.
2. A method as claimed in claim 1 wherein the reductant is supplied to the lean NOx trap by operating the engine with an air/fuel ratio rich of stoichiometric.
3. A method as claimed in claim 2 wherein the degree to which the engine is operated with an air/fuel ratio rich of stoichiometric is determined by the reductant supply profile.
4. A method as claimed in claim 3 wherein the reductant supply profile is such that the air/fuel ratio of the engine is varied during the regeneration event bounded by upper and lower air/fuel ratio limits. S. ** * S S

* 30
5. A method as claimed in claim 3 or in claim 4 *5S wherein the reductant supply profile is such that the :, air/fuel ratio of the engine is changed between at least two * 555 levels of richness during the regeneration event.
6. A method as claimed in any of claims 1 to 5 wherein the active sites are grouped into two or more groups of sites having differing relative activity levels and the - 26 - reductant supply profile is determined based upon the number of active sites in each of the groups of active sites.
7. A method as claimed in claim 6 wherein the reductant supply profile comprises a like number of supply phases as there are groups of active sites.
8. A method as claimed in claim 7 wherein the air/fuel ratio at which the engine is operated during each phase is based upon the relative activity of the sites being regenerated during that phase.
9. A method as claimed in claim 8 wherein the air/fuel ratio at which the engine is operated during each phase is related to the relative activity level of the sites being regenerated so that the air/fuel ratio is lower for higher activity level sites than it is for lower activity level sites.
10. A method as claimed in any of claims 7 to 9 wherein the duration of each phase is dependent upon the number of sites to be regenerated and the relative activity level of the sites being regenerated.
11. A method as claimed in any of claims 6 to 10 wherein the active sites are grouped into three groups comprising of high activity level sites, lower activity * .* * level sites and low activity level sites. ** *. * S S * I

* 30
12. A method as claimed in claim 11 wherein the SI, reductant supply profile comprises the supply of reductant at a high rate during a first supply phase for a first period of time to regenerate the high activity level sites, the supply of reductant at a second lower rate during a second supply phase for a second period of time to regenerate the lower activity sites and the supply of reductant during a third supply phase at a third rate which - 27 - is lower than the second rate for a third period of time to regenerate the low activity level sites.
13. A method substantially as described herein with reference to the accompanying drawing.

13. A method as claimed in claim 12 wherein the rate of reductant supply during at least one of the periods of time varies during the respective period of time.

14. A method as claimed in any of claims 7 to 13 wherein the method further comprises predicting a NOx release rate from the number of sites and the relative activity levels of the sites and using the predicted NOx release rate to determine the reductant supply profile so as to match the supply of reductant to the predicted NOx release rate from the lean NOx trap and prevent excessive reductant breakthrough.

15. A method as claimed in any of claims 1 to 14 wherein the step of estimating the number of active sites and activity levels of the active sites is based at least partially upon the thermal history of the lean NOx trap.

16. A method as claimed in any of claims 1 to 15 wherein the step of estimating the number of active sites and activity levels of the active sites is based at least partially upon a prediction for the number of active sites lost due to sulphur poisoning.

* S. S

S S * S..

17. A method as claimed in any of claims 2 to 16 wherein the method further comprises adapting the reductant * 30 supply profile over time to compensate for ageing of the lean NOx trap. S. * **.

18. A method as claimed in claim 18 wherein adapting the reductant supply profile over time comprises increasing the air/fuel ratio at which the engine is operated during the first supply phase so as to reduce the amount of - 28 - reductant supplied until the time between regeneration events is reduced.

19. A method as claimed in claim 18 wherein adapting the reductant supply profile over time further comprises reducing the duration of the first supply phase until the time between regeneration events is reduced.

20. A method as claimed in claim 19 wherein adapting the reductant supply profile over time further comprises increasing the duration of the second supply phase to return the time between regeneration events to its original value.

21. A method as claimed in claim 20 wherein adapting the reductant supply profile over time further comprises merging the first supply phase with the second supply phase to form a modified second supply phase while supplying the same amount of reductant during the regeneration event.

22. A method as claimed in claim 21 whereii-i adapting the reductant supply profile over time further comprises reducing the amount of fuel supplied during the modified second supply phase by reducing the duration of the second supply phase and increasing the duration of the third supply phase to compensate for the loss of reductant during the modified second supply phase. alt. * . * p..

23. A method as claimed in claim 22 wherein adapting the reductant supply profile over time further comprises reducing the amount of reductant supplied during the regeneration event while increasing the duration of the regeneration event.

SI.... a

24. A method substantially as described herein with reference to the accompanying drawing.

Amendments to the claims have been filed as follows Claims 1. A method for controlling the regeneration of a lean NOx trap having a number of active sites arranged to receive a supply of exhaust gasses from an internal combustion engine, the method comprising the steps of estimating the number and the relative activit7 levels of the active sites within the lean NOx trap, grouping the active sites into a number of groups based upon the activity level of the sites, determining a reductant supply profile based upon the number and relative activity levels of the active sites in each group to produce effective regeneration without excessive reductant breakthrough and supplying a reductant to the lean NOx trap during a regeneration event based upon the determined reductant supply profile.

2. A method as claimed in claim 1 wherein the reductant supply profile comprises a like number of supply phases as there are groups of active sites.

3. A method as claimed in claim 2 wherein the air/fuel ratio at which the engine is operated during each phase is based upon the relative activity of the sites being regenerated during that phase.

4. A method as claimed in claim 2 or in claim 3 wherein the duration of each phase is dependent upon the number of sites to be regenerated and the relative activity level of the sites being regenerated.

5. A method as claimed in any of claims 1 to 4 wherein determining a reductant supply profile further comprises predicting a NOx release rate from the number of sites and the relative activity levels of the sites in each group and using the predicted NOx release rate to determine the reductant supply profile for each group so as to match the supply of reductant to the predicted NOx release rate from the lean NOx trap and prevent excessive reductant breakthrough.

6. A method as claimed in any of claims 1 to 5 wherein the step of estimating the number of active sites and activity levels of the active sites is based upon at least one of the thermal history of the lean NOx trap and a prediction of the number of active sites lost due to sulphur poisoning.

7. A method as claimed in any of claims 1 to 6 wherein the method further comprises adapting the reductant supply profile over time to compensate for ageing of the lean NOx trap.

8. A method as claimed in claim 7 wherein the active sites are grouped into three groups comprising of high activity level sites, moderate activity level sites and low activity level sites and the reductant supply profile comprises the supply of reductant at a high rate during a I first supply phase for a first period of time to regenerate the high activity level sites, the supply of reductant at a second lower rate during a second supply phase for a second period of time to regenerate the moderate activity sites and the supply of reductant during a third supply phase at a third rate which is lower than the second rate for a third period of time to regenerate the low activity level sites and adapting the reductant supply profile over time comprises increasing the air/fuel ratio at which the engine is operated during the first supply phase so as to reduce the amount of reductant supplied until the time between regeneration events is reduced.

9. A method as claimed in claim 8 wherein adapting the reductant supply profile over time further comprises reducing the duration of the first supply phase until the time between regeneration events is reduced and then increasing the duration of the second supply phase to return the time between regeneration events to its original value.

10. A method as claimed in claim 9 wherein adapting the reductant supply profile over time further comprises merging the first supply phase with the second supply phase to form a modified second supply phase while supplying the same amount of reductant during the regeneration event.

11. A method as claimed in claim 10 wherein adapting the reductant supply profile over time further comprises reducing the amount of fuel supplied during the modified second supply phase by reducing the duration of the second supply phase and increasing the duration of the third supply phase to compensate for the loss of reductant during the modified second supply phase.

12. A method as claimed in claim 11 wherein adapting the reductant supply profile over time further comprises reducing the amount of reductant supplied during the regeneration event while increasing the duration of the regeneration event.