Method of monitoring an SCR catalyst

Abstract

A method of monitoring an SCR catalyst in which an area factor (a) of the SCR catalyst is ascertained by means of an observer. It is concluded that there is a fault in the SCR catalyst when a comparison shows that the area factor (a) has gone below a threshold value (S).

Claims

1. A method of monitoring an SCR catalyst (12), the method comprising: ascertaining, via an observer (22), an area factor (a) of the SCR catalyst (12); and determining (36) that there is a fault in the SCR catalyst (12) when a comparison (34) shows that the area factor (a) has gone below a threshold value (S), wherein the area factor (a) is ascertained (32) in that a difference (Δc.sub.NO.sub.x.sub.+NH.sub.x) between a modeled exhaust gas signal (c.sub.NO.sub.x.sub.+HN.sub.x.sup.mod) upstream of the SCR catalyst (12) and a measured exhaust gas signal (c.sub.NO.sub.x.sub.+HN.sub.x.sup.mess) downstream of the SCR catalyst (12) is ascertained and the difference is amplified by the observer (22), wherein the modeled exhaust gas signal (c.sub.NO.sub.x.sub.+HN.sub.x.sup.mod) and the measured exhaust gas signal (c.sub.NO.sub.x.sub.+HN.sub.x.sup.mess) are each a cumulative signal of a nitrogen oxide signal and an ammonia signal, and wherein the area factor (a) is taken into account for the modeled exhaust gas signal (c.sub.NO.sub.x.sub.+HN.sub.x.sup.mod).

2. The method according to claim 1, wherein the area factor can assume a value between 0 and 1 and the area factor (a) gives a reduction in a surface area of the SCR catalyst resulting from aging or damage.

3. The method according to claim 1, wherein, for the modeled exhaust gas signal (c.sub.NO.sub.x.sub.+HN.sub.x.sup.mod), the area factor is taken into account in a reaction equation for a NOx reaction, a NH3 adsorption and a NH3 desorption of the SCR catalyst.

4. The method according to claim 2, wherein the amplification depends on at least one value selected from the group consisting of a temperature (T) of the SCR catalyst (12), a temperature gradient of the SCR catalyst (12), an exhaust gas mass flow rate upstream of the SCR catalyst (12), a nitrogen oxide mass flow rate upstream of the SCR catalyst (12) and a modeled ammonia mass flow rate downstream of the SCR catalyst (12).

5. The method according to claim 1, wherein the observer (22) includes an integrator (23).

6. The method according to claim 1, wherein the comparison (34) of the area factor (a) with the threshold value (S) is made only after a teach-in phase for the observer (22) has elapsed.

7. The method according to claim 1, wherein the area factor (a) is initialized to the threshold value (S).

8. A non-transitory, machine-readable storage medium containing instructions that when executed by a computer cause the computer to monitor an SCR catalyst (12), by: ascertaining, via an observer (22), an area factor (a) of the SCR catalyst (12); and determining (36) that there is a fault in the SCR catalyst (12) when a comparison (34) shows that the area factor (a) has gone below a threshold value (S), wherein the area factor (a) is ascertained (32) in that a difference (Δc.sub.NO.sub.x.sub.+NH.sub.x) between a modeled exhaust gas signal (c.sub.NO.sub.x.sub.+HN.sub.x.sup.mod) upstream of the SCR catalyst (12) and a measured exhaust gas signal (c.sub.NO.sub.x.sub.+HN.sub.x.sup.mess) downstream of the SCR catalyst (12) is ascertained and the difference is amplified by the observer (22), wherein the modeled exhaust gas signal (c.sub.NO.sub.x.sub.+HN.sub.x.sup.mod) and the measured exhaust gas signal (c.sub.NO.sub.x.sub.+HN.sub.x.sup.mess) are each a cumulative signal of a nitrogen oxide signal and an ammonia signal, and wherein the area factor (a) is taken into account for the modeled exhaust gas signal (c.sub.NO.sub.x.sub.+HN.sub.x.sup.mod).

9. An electronic control device (19) configured to monitor an SCR catalyst (12) by ascertaining, via an observer (22), an area factor (a) of the SCR catalyst (12); and determining (36) that there is a fault in the SCR catalyst (12) when a comparison (34) shows that the area factor (a) has gone below a threshold value (S), wherein the area factor (a) is ascertained (32) in that a difference (Δc.sub.NO.sub.x.sub.+HN.sub.x) between a modeled exhaust gas signal (c.sub.NO.sub.x.sub.+HN.sub.x.sup.mod) upstream of the SCR catalyst (12) and a measured exhaust gas signal (c.sub.NO.sub.x.sub.+HN.sub.x.sup.mess) downstream of the SCR catalyst (12) is ascertained and the difference is amplified by the observer (22), wherein the modeled exhaust gas signal (c.sub.NO.sub.x.sub.+HN.sub.x.sup.mod) and the measured exhaust gas signal (c.sub.NO.sub.x.sub.+HN.sub.x.sup.mess) are each a cumulative signal of a nitrogen oxide signal and an ammonia signal, and wherein the area factor (a) is taken into account for the modeled exhaust gas signal (c.sub.NO.sub.x.sub.+HN.sub.x.sup.mod).

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 shows a schematic of an SCR catalyst system in which an SCR catalyst can be monitored by means of a working example of the method of the invention.

(2) FIG. 2 shows an observer which is used in a method according to a working example of the invention.

(3) FIG. 3 shows a flow diagram of a method in a working example of the invention.

(4) FIG. 4 shows a flow diagram of a method in another working example of the invention.

(5) FIG. 5 shows a flow diagram of a method in yet another working example of the invention.

(6) FIG. 6 shows, in two diagrams, the progression of an area factor and of a temperature of an SCR catalyst against time in one working example of the method of the invention.

(7) FIG. 7 shows, in a diagram, monitoring values of an area factor in a working example of the method of the invention.

DETAILED DESCRIPTION

(8) FIG. 1 shows an exhaust gas conduit 10 in which multiple catalysts, namely a diesel oxidation catalyst (DOC) 11, an SCR catalyst (SCRF) 12 disposed on the diesel particulate filter, and a further SCR catalyst 13, are disposed in succession. Between the diesel oxidation catalyst 11 and the first SCR catalyst 12 is disposed a first dosage module 14. Between the two SCR catalysts 12, 13 is disposed a second dosage module 15. The two dosage modules 14, 15 are set up to meter an aqueous urea solution into the exhaust gas conduit 10 upstream of each of their respective SCR catalysts 12, 13. Nitrogen oxide sensors 16, 17, 18 are disposed between the diesel oxidation catalyst 11 and the first dosage module 14 between the first SCR catalyst 12 and the second dosage module 15 and downstream of the second SCR catalyst 13. An electronic control device 19 controls the dosage modules 14, 15 and receives sensor data from the nitrogen oxide sensors 16, 17, 18. In the electronic control device 19, methods can proceed according to different working examples of the invention.

(9) In all the working examples described hereinafter, the method of the invention uses an observer structure as shown in FIG. 2. This is described hereinafter for the monitoring of the first SCR catalyst 12. An ammonia concentration to be dosed into the SCR catalyst 12 is calculated from an actuation of the first dosage module 14. A nitrogen oxide concentration ca, upstream of the SCR catalyst 12 can be either measured by means of the first nitrogen oxide sensor 16 or ascertained by means of a model. These two input parameters are used by the electronic control device 19 for control 20 of the first SCR catalyst 12. In addition, they are supplied to a kinetic model 21 of the first SCR catalyst 12. The second nitrogen oxide sensor 17, owing to its cross-sensitivity to ammonia, gives an exhaust gas signal c.sub.NO.sub.x.sub.+NH.sub.x.sup.mess corresponding to the sum total of the concentration c.sub.NO.sub.x.sup.mess of nitrogen oxides and the concentration c.sub.NH.sub.x.sup.mess of ammonia downstream of the first SCR catalyst 12. In the kinetic model 21, these two individual concentrations are modeled, and the modeled nitrogen oxide concentration c.sub.NO.sub.x.sup.mod and the modeled ammonia concentration c.sub.NH.sub.x.sup.mod are used to create a modeled cumulative signal c.sub.NO.sub.x.sub.+NH.sub.x.sup.mod. Subsequently, a difference Δc.sub.NO.sub.x.sub.+NH.sub.x between the measured exhaust gas signal c.sub.NO.sub.x.sub.+NH.sub.x.sup.mess and the modeled cumulative signal c.sub.NO.sub.x.sub.+NH.sub.x.sup.mod is created.

(10) The illustrative model equations of the kinetic model 21 for the NOx and NH3 concentration beyond the SCR catalyst and for the loading level θ are represented in the formulae that follow. The symbols mean:

(11) c.sub.NOx,US: concentration upstream of the SCR catalyst

(12) K.sub.NOx: frequency factor for NOx reaction

(13) K.sub.Des: frequency factor for NH3 desorption

(14) K.sub.Ads: frequency factor for NH3 adsorption

(15) E.sub.ADs: activation energy for NH3 adsorption

(16) E.sub.Des: activation energy for NH3 desorption

(17) E.sub.NOx: activation energy for NOx reaction

(18) θ: NH3 loading level

(19) m.sub.NH3: current amount of NH3 stored in the SCR catalyst results from the balance of the amounts of NH3 flowing in and out

(20) m.sub.NH3.sup.MAX: maximum amount of NH3 stored in a new SCR catalyst

(21) R: gas constant

(22) T: catalyst temperature

(23) a: area factor

(24) ε: correction factor to reflect the nonlinearity of desorption

(25) c.sub.NH3: NH3 concentration in the gas phase (derived from the dosage)

(26) c.sub.NOx: NOx concentration in the gas phase

(27) t.sub.V: dwell time of the exhaust gas in the catalyst

(28) Equations to describe the reaction kinetics:

(29) Reaction rate of NH3 adsorption:

(30) $r_{\mathrm{Ads}}={\mathrm{aK}}_{\mathrm{Ads}}{c}_{\mathrm{NH}3}\left(1-\theta \right)e^{\frac{-E_{\mathrm{Ads}}}{\mathrm{RT}}}$

(31) Reaction rate of NH3 desorption:

(32) $r_{\mathrm{Des}}={\mathrm{aK}}_{\mathrm{Des}}\theta e^{\frac{-E_{\mathrm{Des}}\left(1-s\theta \right)}{\mathrm{RT}}}$

(33) Reaction rate of NOx reaction:

(34) $r_{\mathrm{NOx}}={\mathrm{aK}}_{\mathrm{NOx}}{c}_{\mathrm{NOx}}\theta e^{\frac{-E_{\mathrm{NOx}}}{\mathrm{RT}}}$

(35) Equations for modeled NOx and NH3 concentration

(36) $c_{\mathrm{NOx}}^{\mathrm{mod}}\left(t\right)=c_{\mathrm{NOx},\mathrm{US}}{e}^{-{\mathrm{aK}}_{\mathrm{NOx}}{\theta }_{t_V}{e}^{\left(-\frac{E_{\mathrm{NOx}}}{\mathrm{RT}}\right)}}$$C_{\mathrm{NH}3}^{\mathrm{mod}}\left(t\right)=\frac{K_{\mathrm{Des}}}{K_{\mathrm{Ads}}}\frac{\theta }{1-\theta }{e}^{E_{\mathrm{ADs}}-E_{\mathrm{Des}}\left(1-s\theta \right)}$$\mathrm{with}$$\theta =\frac{m_{\mathrm{NH}3}}{a{m}_{\mathrm{NH}3}^{\mathrm{MAX}}}$

(37) The modeling is based on a reaction kinetics approach in which, among other reactions, the NOx reaction and the NH3 adsorption and desorption are also represented with the aid of the Arrhenius equation. In order to arrive at the equations shown in the example, some simplifications were made.

(38) The area factor described above is a factor between 0 and 1 that describes the aging state of the catalytically active surface. If the factor is 1, the surface corresponds to that of a new SCR catalyst. As the factor decreases, the catalytically active surface area is reduced. With area factor=0, an inactive, i.e. uncoated, SCR catalyst is thus represented. The surface area can be reduced by aging and poisoning effects.

(39) The formulae detailed above show, by way of example, the influence of the area factor on the system characteristics. A reduction on the one hand increases the modeled NH3 concentration and on the other hand reduces the NOx conversion, which leads to an elevated NOx concentration. The area factor also affects possible further reactions, for example that of NH3 oxidation and that of hydrolysis.

(40) This representation of the aging or effective surface area can be implemented in all reaction kinetics models (according to Arrhenius) irrespective of their exact execution.

(41) What this means for the observer approach described is that the two starting parameters (NOx and NH3 concentration) and their sum total can be used to conclude the catalytic surface area. This is especially advantageous since NOx sensors that emit a cumulative signal of NOx+NH3 concentration are usually used.

(42) This is amplified by an observer 22 by multiplication by a factor. The factor is chosen depending on parameters of the SCR catalyst 12, including its temperature, its temperature gradient, an exhaust gas mass flow rate upstream of the SCR catalyst 12 and a nitrogen oxide mass flow rate upstream of the SCR catalyst 12, and a modeled ammonia mass flow rate downstream of the SCR catalyst 12. If enablement conditions are not fulfilled for the monitoring of the SCR catalyst 12, the amplification factor can also be set at 0. The result of the amplification is multiplied by a factor of −1 and sent to an integrator 23. This gives an area factor a that can assume values in the range from 0 to 1. The area factor a is used firstly for correction of the kinetic model 21 and secondly sent to an on-board diagnosis evaluation 24. It is ascertained therein, by comparison of the area factor a with threshold value, whether the SCR catalyst 12 is intact the purposes of the OBD, i.e. is a WPA catalyst, or whether it should be regarded as faulty for the purposes of the OBD, i.e. is a BPU catalyst.

(43) For derivation of the illustrative observer, the area factor a is introduced as an additional state which is invariable and hence its derivative is 0.

(44) $\frac{\mathrm{da}}{\mathrm{dt}}=0$

(45) This is supplemented as an additive term by the observer term with the observer amplification L (Luenberger observer) which is generally present in this formula

(46) $\frac{\mathrm{da}}{\mathrm{dt}}=0+L\left(c_{\mathrm{NOx}+\mathrm{NH}3}^{\mathrm{mess}}\left(t\right)-c_{\mathrm{NOx}+\mathrm{NH}3}^{\mathrm{mod}}\left(t\right)\right)$

(47) L is, in the case of the observer, for the area factor (−1)*f.sub.Obs(f.sub.Obs here is a the observer amplification factor, which can be altered as a function of particular conditions; see below), which, after integration in block 23 over a period t.sub.Eval, leads to the following calculation method for a:
a=∫.sub.0.sup.t.sup.Eval−f.sub.Obs(c.sub.NO.sub.x.sub.+NH.sub.x.sup.mess(t)−c.sub.NO.sub.x.sub.+NH.sub.x.sup.mod(t))dt

(48) In a further execution, in the case of use of a multigas sensor that can give a separate output for NOx and NH3, the observer can react separately to variances in NOx and NH3. For example, the amplification can be determined individually. This is also true of the dependences of the amplification on further parameters that are detailed below.

(49) In one working example of the method of the invention which is shown in FIG. 3, after a start 30 of monitoring, a test 31 is first conducted as to whether the enablement conditions for the monitoring are fulfilled. These enablement conditions include the enablement of the dosage module 14 and that of nitrogen oxide sensors 16, 17 involved in the monitoring. If enablement is approved, a calculation 32 calculates a trust factor in the observer 22 by which the difference Δc.sub.NO.sub.x.sub.+NH.sub.x should be amplified. Using this trust factor, there is a recalculation 33 of the area factor a which is given to the kinetic model 21 and to the OBD evaluation 24. In the OBD evaluation 24, the area factor a is compared to a threshold value. If it has been found beforehand in the test 31 that the conditions for the monitoring are not met, the observer 22 is frozen and the comparison 34 is made by means of the last stored value of the area factor a. On the basis of the comparison 34, a decision is made as to whether the SCR catalyst 12 is intact 35, i.e. is a WPA catalyst, or is defective 36, i.e. is a BPU catalyst.

(50) A second working example of the method of the invention is shown in FIG. 4. In this method, when the test 31 shows that the enablement conditions are not met, the monitoring is stopped until enablement is approved. After the calculation 33 of the area factor a, it is not compared immediately with the threshold value in step 34. Instead, a further test 40 is first carried out as to whether a change in the area factor a exceeds a change threshold value or whether monitoring is still in a teach-in phase for the area factor a. Only when neither of these conditions is fulfilled is the method continued with step 34. Otherwise, a teach-in phase for the area factor a is initialized and started 41. If a test 42 shows that the teach-in phase has been completely ended, the method is continued with step 34. Otherwise, it returns to the start 30.

(51) A third working example of the method of the invention is shown in FIG. 5. In a departure from the second working example, this dispenses with steps 40 and 41 and instead envisages merely the test 42 as to whether a teach-in phase has ended between the calculation 33 of the area factor and the comparison 34 of the area factor with the threshold value. If this condition is not met, the method returns to the start 30.

(52) FIG. 6 shows the progression of the area factor a and the temperature T of the SCR catalyst 12, in each case for a WPA catalyst and for a BPU catalyst, with time t. Each point in this diagram indicates the start of a teach-in phase. In this teach-in phase, the area factor a is in each case initialized to the threshold value S, which in the present case is 0.45. Proceeding from the initialization value, the area factor a (WPA) of the WPA catalyst subsequently assumes values above the threshold value S, and the area factor a (BPU) of the BPU catalyst assumes values below the threshold value S. Illustrative individual values of the area factor a for the WPA catalyst and for the BPU catalyst are each shown in FIG. 7 after the teach-in phase has elapsed.