Process parameter estimation in controlling emission of a non-particulate pollutant into the air

• US 20060045804A1
• Filed: 12/06/2004
• Published: 03/02/2006
• Est. Priority Date: 08/27/2004
• Status: Active Grant

• Current Assignee: General Electric Technology GmbH (GE Aerospace)
• Original Assignee: Alstom Technology Ltd. (Alstom SA)
• Inventors: Piche, Stephen, Boyden, Scott A.
• Granted Patent: US 7,860,586 B2
• Time in Patent Office: Days
• US Class Current: 422/62
• CPC Class Codes: B01D 2257/302 Sulfur oxides
• CPC Class Codes: B01D 2257/404 Nitrogen oxides other than dinitrogen oxide
• CPC Class Codes: G05B 13/027 using neural networks only
• CPC Class Codes: G05B 13/048 using a predictor
• CPC Class Codes: Y10T 436/12 Condition responsive control

Abstract

A parameter value estimator is provided for a process performed primarily to control emission of a particular non-particulate pollutant, such as NO_x and SO₂, into the air. The process has multiple process parameters (MPPs) including a parameter representing an amount of the particular non-particulate pollutant emitted. The parameter value estimator includes either a neural network process model or a non-neural network process model. In either case the model represents a relationship between one of the MPPs, other than the parameter representing the amount of the emitted particular non-particulate pollutant, and one or more other of the MPPs. Also included is a processor configured with the logic, e.g. programmed software, to estimate a value of the one MPP based on a value of each of the one or more other MPPs and the one model.

Claims

1. A parameter value estimator for a process primarily performed to control emission of a particular non-particulate pollutant into the air, the process having multiple process parameters (MPPs) including a parameter representing an amount of the particular non-particulate pollutant emitted, comprising:

one of a neural network process model and a non-neural network process model representing a relationship between one of the MPPs, other than the parameter representing the amount of the emitted particular non-particulate pollutant, and one or more of the other MPPs; and

a processor configured with the logic to estimate a value of the one MPP based on a value of each of the one or more other MPPs and the one model.

2. The parameter value estimator according to claim 1, wherein:

the particular non-particulate pollutant is one of NO_x and SO₂.

3. The parameter value estimator according to claim 1, wherein:

the value of each of the one or more other MPPs is either (i) a value measured during the performance of the process or (ii) a value estimated based on one or more other values of the multiple MPPs measured during the performance of the process.

4. The parameter value estimator according to claim 1, wherein:

the estimated value of the one MPP is an estimated first value of the one MPP;

the value of each of the one or more other MPPs is a first value of each of the one or more other MPPs; and

the processor is further configured with the logic to estimate a second value of the one MPP based on the estimated first value of the one MPP, a second value of each of the one or more other MPPs and the one model.

5. The parameter value estimator according to claim 1, wherein:

the processor is further configured to at least one of (i) estimate the value of the one MPP in real time during performance of process and (ii) estimate the value of the one MPP periodically.

6. The parameter value estimator according to claim 1, wherein the one MPP is either:

a pH level of matter applied in the process to absorb the particular non-particulate pollutant and thereby control its emission into the air;

a purity of a by-product produced in performing the process;

an amount of oxygen dissolved in matter applied in the process to absorb the particular non-particulate pollutant and thereby control its emission into the air;

an amount of ammonia applied in the process to absorb the particular non-particulate pollutant that is emitted into the air with the particular non-particulate pollutant that is not absorbed;

or

an amount of applied ammonia in exhausted reduced NOx flue gas, the particular non-particulate pollutant being NOx, and the one or more other MPPs include an amount of the applied ammonia.

7. The parameter value estimator according to claim 1, wherein:

the processor is configured with logic including estimate generator logic and estimator logic;

the processor estimates the value of the one MPP by;

executing the estimate generator logic to compute a value of the one MPP based on the value of each of the one or more other MPPs and the one model; and

executing the estimator logic to determine the estimated value of the one MPP based on the computed value of the one MPP and a measured value of the one MPP; and

the processor is further configured with the logic to update the one model based on the determined estimated value of the one MPP.

8. The parameter value estimator according to claim 7, wherein:

the updating includes updating the represented relationship between the one MPP and the one or more other MPPs based on the determined estimated value of the one MPP.

9. The parameter value estimator according to claim 7, wherein:

the estimator logic includes a Kalman filter for filtering the computed and the measured values of the one MPP to determine the estimated value of the one MPP.

10. The parameter value estimator according to claim 1, wherein:

the process is a selective catalytic reduction (SCR) process that applies ammonia to remove NO_x from the NO_x laden flue gas and thereby control emissions of NO_x, and exhausts reduced NO_x flue gas; and

the one MPP is an amount of the applied ammonia in the exhausted reduced NO_x flue gas and the one or more other MPPs includes an amount of the applied ammonia.

11. The parameter value estimator according to claim 1, wherein:

the process is a wet flue gas desulfurization (WFGD) process that distributes limestone slurry, applies the distributed limestone slurry to remove SO₂ from SO₂ laden wet flue gas and thereby control emissions of SO₂, and exhausts desulfurized flue gas;

the one MPP is a pH level of the applied limestone slurry; and

the one or more other MPPs includes at least one of an amount of SO₂ in the SO₂ laden wet flue gas, an amount of SO₂ in the exhausted desulfurized flue gas, and the distribution of the applied limestone slurry.

12. The parameter value estimator according to claim 1, wherein:

the process is a wet flue gas desulfurization (WFGD) process that (i) applies oxidation air to limestone slurry, (ii) distributes the oxidized limestone slurry, (iii) applies the distributed limestone slurry to remove and crystallize SO₂ from SO₂ laden wet flue gas and thereby control emissions of SO₂ and produce gypsum as a by-product, and (iv) exhausts desulfurized flue gas;

the one MPP is either a quality of the produced gypsum or an amount of dissolved oxygen in the oxidized limestone slurry; and

the one or more other MPPs includes at least one a pH level of the applied limestone slurry, a distribution of the applied limestone slurry and an amount of the applied oxidation air.

13. A method of estimating a parameter value for a process performed primarily to control emission of a particular non-particulate pollutant into the air, the process having multiple process parameters (MPPs) including a parameter representing an amount of the particular non-particulate pollutant emitted, comprising:

determining a value of each of one or more of the MPPs; and

estimating a value of another one of the MPPs, other than the parameter representing the amount of the emitted particular non-particulate pollutant, based on (i) the determined value of each of the one or more MPPs and (ii) one of a neural network process model and a non-neural network process model representing a relationship between the one other MPP and the one or more MPPs.

14. The method according to claim 13, wherein:

the value of each of the one or more MPPs is determined by either (i) measuring the value during the performance of the process or (ii) estimating the value based on one or more other values of the multiple MPPs measured during the performance of the process.

15. The method according to claim 13, wherein the estimated value of the one other MPP is an estimated first value of the one other MPP and the value of each of the one or more MPPs is a first value of each of the one or more MPPs, and further comprising:

estimating a second value of the one other MPP based on the estimated first value of the one other MPP, a second value of each of the one or more MPPs, and the one model.

16. The method according to claim 13, wherein:

the value of the one other MPP is at least one of (i) estimated in real time during performance of process and (ii) estimated periodically.

17. The method according to claim 13, wherein the one other MPP is either:

a pH level of matter applied in the process to absorb the particular non-particulate pollutant and thereby control its emission into the air;

a purity of a by-product produced in performing the process;

an amount of oxygen dissolved in matter applied in the process to absorb the particular non-particulate pollutant and thereby control its emission into the air;

an amount of ammonia applied in the process to absorb the particular non-particulate pollutant that is emitted into the air with the particular non-particulate pollutant that is not absorbed;

or

an amount of applied ammonia in exhausted reduced NO_x flue gas, the particular non-particulate pollutant being NO_x, and the one or more MPPs include an amount of the applied ammonia.

18. The method according to claim 13, wherein:

the process is a selective catalytic reduction (SCR) process that applies ammonia to remove NO_x from the NO_x laden flue gas and thereby control emissions of NO_x, and exhausts reduced NO_x flue gas; and

the one other MPP is an amount of the applied ammonia in the exhausted reduced NO_x flue gas and the one or more MPPs include an amount of the applied ammonia.

19. The method according to claim 13, wherein:

the process is a wet flue gas desulfurization (WFGD) process that distributes limestone slurry, applies the distributed limestone slurry to remove SO₂ from SO₂ laden wet flue gas and thereby control emissions of SO₂, and exhausts desulfurized flue gas;

the one other MPP is a pH level of the applied limestone slurry; and

the one or more MPPs include at least one of an amount of SO₂ in the SO₂ laden wet flue gas, an amount of SO₂ in the exhausted desulfurized flue gas, and the distribution of the applied limestone slurry.

20. The method according to claim 13, wherein:

the process is a wet flue gas desulfurization (WFGD) process that (i) applies oxidation air to limestone slurry, (ii) distributes the oxidized limestone slurry, (iii) applies the distributed limestone slurry to remove and crystallize SO₂ from SO₂ laden wet flue gas and thereby control emissions of SO₂ and produce gypsum as a by-product, and (iv) exhausts desulfurized flue gas;

the one other MPP is either a quality of the produced gypsum or an amount of dissolved oxygen in the oxidized limestone slurry; and

the one or more MPPs include at least one of a pH level of the applied limestone slurry, the distribution of the applied limestone slurry and an amount of the applied oxidation air.

21. The method according to claim 13, wherein estimating the value of the one other MPP includes computing a value of the one other MPP based on the value of each of the one or more MPPs and the one model, and determining the estimated value of the one other MPP based on the computed value of the one other MPP and a measured value of the one other MPP, and further comprising:

updating the one model based on the determined estimated value of the one other MPP.

22. The method according to claim 21, wherein:

the updating includes updating the represented relationship between the one other MPP and the one or more MPPs based on the determined estimated value of the one other MPP.

23. The method according to claim 21, wherein:

the estimated value of the one other MPP is determined by filtering the computed and the measured values of the one other MPP with a Kalman filter.