Patent Application
20120052450
The present invention relates generally to a system and method for optimizing the control and performance of a boiler, and more specifically to a system and method for optimizing the control and performance of an oxy-fired or an air-fired pulverized coal boiler.
A boiler typically includes a furnace in which fuel is burned to generate heat to produce steam. The combustion of the fuel creates thermal energy or heat, which is used to heat and vaporize a liquid, usually being water into steam. The generated steam may be used to drive a turbine to generate electricity. Pulverized coal is the fuel consumed to provide thermal energy for production of about 50% of the world's electric supply. In an air-fired pulverized coal boiler, atmospheric air is fed into the furnace and mixed with the pulverized coal for combustion. In an oxy-fired pulverized coal boiler, concentrated levels of oxygen are fed into the furnace and mixed with pulverized coal for combustion.
Pulverized coal boilers include multi-variable controls applicable to steam temperature controls, combustion controls, and backend controls. Backend controls are controls applied to one or more flue gas treatment systems used to treat boiler combustion gases. Such flue gas treatment systems may include but are not limited to selective catalytic reduction (SCR), selective non-catalytic reduction (SNCR), and flue gas desulfurization (FGD). Such flue gas treatment systems have a large number of known input variables that drastically affect the performance and efficiency of the treatment system and affect the emission levels achievable by the treatment system.
Boiler operation is typically monitored by measuring temperatures at metal structures in and around the inside surface of the furnace. One disadvantage of this known method is that it does not provide enough measurement data to analyze the flame and determine detailed characteristics of the flame in the furnace. (The flame is also referred to herein as “fireball” and is used interchangeably throughout this specification.) Detailed characteristics of the flame can provide further insights into combustion dynamics and is useful for improving combustion efficiency and reducing emissions.
Steam temperature control in pulverized coal boilers relies heavily on desuperheater water spray. Desuperheater water spray however, is thermally inefficient because it causes losses in both mass and energy, and has hard constraints with regard to both maximum spay flow and maximum change rate. The latter may result in ineffective steam temperature control if not properly addressed. In addition, current steam temperature control systems are mostly reactive and use feedbacks of current temperature measurements at a few points in the steam path to control steam temperature at turbine inlets. Such temperature measurement feedbacks can make steam temperature control sluggish, especially during load changes.
During load ramping periods, steam temperature variations may potentially damage pressure parts such as main steam headers, pipes, turbine inlet valves, and turbine walls. This is of particular concern when the steam pressure is high, e.g. in the case of supercritical and ultra-supercritical fossil power generation units. Another factor that sometimes restricts faster load ramping rates in coal-fired boilers is the delay between a fuel feed supply change demand to an actual fuel feed supply increase/decrease to the furnace. Physical renovations to a boiler system to avoid such drawbacks are expensive and difficult to implement.
The present invention relates to a multivariable advanced process control (“APC”) system and method for using the same for improving or optimizing the performance of regulatory control systems of a boiler, and more specifically, a pulverized coal boiler. In one embodiment, the APC system includes at least one camera positioned within or in communication with the interior of the furnace of a boiler. As such, camera is operable receive one-dimensional, two-dimensional and/or three-dimensional images of combustion in the furnace and to transmit signals indicative of at least one parameter of a fireball within the furnace to a controller in communication with the camera. The controller includes one or more predictive modeling modules. The at least one camera captures measurement data pertaining to fireball characteristics including but not limited to temperature, radiant energy, and emissivity in the furnace. The controller is operable to generate multi-dimensional images of the at least one parameter of the fireball within the furnace. Software executing on the controller optimizes and controls combustion systems, steam system, and backend control systems based, at least in part, on the multi-dimensional images.
In reference to
The camera 130 is in communication with a controller 140. The controller 140 is operable to receive signals from the camera 130. The controller 140 may comprise a processor based computing device, such as a server. Software executing on the controller 140 processes image data captured by the camera 130 and transmitted to the controller 140. The image data and processed images are used to calculate the fireball 120 characteristics, such as temperature profile, heat flux, and emissivity. The image data can be converted into multi-dimensional combustion temperature maps, emissivity maps, and profiles of other variables of the combustion. Based, at least in part, on the data collected by the at least one camera 130, software executing on the controller generates control adjustments to control one or more of the boiler components to improve combustion efficiency and/or reduce emissions. The control adjustments are transmitted to and received by one or more components 150 that form part of the boiler system, steam system, or backend systems. For example, in the embodiment illustrated in
In reference to
The cameras 230 are in communication with a controller 240. The controller includes a central processing unit (CPU) and assorted hardware and software for operating the system 200 in accordance with the present invention. Several software modules 242, 244, 246, 248 are executing on the controller 240. The controller 240 includes an image processing, analysis and inference module 242. Module 242 receives an image and/or a stream of images from the one or more cameras 230 and processes the images. The processed images are used to calculate fireball characteristics, such as but not limited to, temperature profile, heat flux, and/or emissivity, without using a process models or an optimizer. The module 242 generates multi-dimensional temperature maps, heat flux maps, and emissivity maps.
The calculated fireball characteristics are available for online monitoring in the form of horizontal or vertical planer profiles that may be presented in one-dimensional, two-dimensional or three-dimensional form. In some embodiments of the present invention an interface is in communication with the controller 240 and the profiles are displayed on the interface. In further reference to
In further reference to
The controller 240 is further in communication with one or more backend system components 280 used to treat the combustion products or flue gas from the boiler system. Backend components are applied to the boiler system to treat combustion gases exiting through the flue. Backend components may include but are not limited to selective catalytic reduction (SCR), selective non-catalytic reduction (SNCR), and flue gas desulfurization (FGD) components. Based on the fireball characteristics, the control module 248 generates control adjustments for the one or more backend control components 280 to control and optimize their performance. The control adjustments are communicated to the one or more backend control components 280 and control the same to optimize the efficiency of system 200. The embodiment shown in
In some embodiments of the present invention, the boiler system may use separate process models for the boiler system components and the backend components. In other embodiments, the present invention may employ a single integrated process model for the whole boiler island. In some embodiments, the software executing on the controller 240, sometimes referred to as the model-based multivariable APC system, may use material models to estimate components' and/or plant's life and use them in the cost function of its optimization problem to extend the life of components and/or increase time between required maintenance.
In one embodiment of the present invention, the one or more cameras 230 capture images of the fireball 220. This image data is transmitted to and received by the controller 240. Based in part on the received image data, software executing on the controller 240 estimates the location of the center of the fireball 220. In some embodiments, the controller 240 incorporates temperature mapping in estimating the center of the fireball 220 in the furnace 210. Based on this identified location, the controller 240 determines the elevation of the fireball 220 in the furnace 230. Tilt controls in the furnace may be employed to adjust the elevation of the fireball 220 in the furnace 230 based in response to commands from the controller 240. By adjusting the elevation of the fireball 220 in the furnace, the system can control the temperature of the steam and control NOx/CO emissions. For example, the system can reduce the temperature of steam in the system by directing the center of the fireball 220 way from one or more heating coils.
In yet other embodiments, image and temperature mapping data is received by the controller 240. Based, at least in part, on this data, the controller 240 is capable of detecting the location of the fireball 220 in the furnace 230, and is capable of detecting a deviation of the fireball 220 from the center of the furnace 230. To the extent the fireball 220 has deviated from the center of the furnace 220, the location of the fireball 220 can be affected by altering the direction of air being fed into the furnace in response to commands sent from the controller 240. By centering the fireball 220 in the furnace 230, a more efficient operation is achieved.
In some embodiments, it is preferred to achieve a uniformly distributed temperature inside the furnace to improve control of NOx and CO emissions and particulate matter based on different properties of the coal fuel. Image date captured by the one or more cameras is processed by the controller 240 to determine the temperature distribution in the furnace 230. Based on this distribution, the controller 240 transmits commands to affect the air/oxidant flue gas to provide a more optimally distributed in-furnace temperature and O2 level distribution, thereby improving control of backend emissions.
In reference to
In reference to
In one embodiment, the present invention the control system includes a rapid load change controller (RLC). The first element in the RLC control scheme is the furnace radiant measurement system. This sensor system includes both furnace flame intensity sensor components and water wall heat flux sensor components. The radiant sensor system provides a novel input to the RLC control system's predictive steam properties (TSP) model. It is important that the radiant measuring system accurately measures the furnace flame and energy field. In order to do this, the system generates a data fusion of camera image 2-D data and point furnace wall heat flux probes. The radiant measurement system data is analyzed to compute furnace flame and energy release features. The objective is to establish a robust set of independent radiant characteristics with known high correlations to water/steam side boiler process conditions. Reference plant field data will be collected in order to determine the minimum number of radiant characteristics that form a reliable system basis set for flame/furnace state characterization.
The second element in the RLC is the Turbine Steam Parameters (TSP) predictive model. This predictive model uses the RLC furnace radiant characteristics to predict the ultimate turbine inlet steam parameters for the current fuel firing condition. It is the rapid accurate prediction from the TSP predictive model that makes more rapid control of the field firing during load change possible. The delay between the fuel delivery to the furnace and the radiant energy release is brief, for example a few seconds. The radiant energy release is an accurate instantaneous measure of the energy input to the furnace waterwalls. Therefore using this prompt radiant response to fuel firing as a model input yields a superior prediction over methods using fuel mass flow and heating value estimation. The turbine steam property predictor will also use the existing plant process measurement values to generate an adaptive error signal.
Although the present invention has been disclosed and described with reference to certain embodiments thereof, it should be noted that other variations and modifications may be made, and it is intended that the following claims cover the variations and modifications within the true scope of the invention.
The present application claims the benefit under 35 U.S.C. §119(e) of Provisional Patent Application Ser. No. 61/377,637 entitled FLAME IMAGING SYSTEM BASED CONTROL/OPTIMIZATION FOR AN OXY-FIRED PULVERIZATION COAL BOILER filed Aug. 27, 2010, the disclosure of which is incorporated herein by reference in its entirety. The present application claims the benefit under 35 U.S.C. §119(e) of Provisional Patent Application Ser. No. 61/377,676 entitled OPTIMIZATION AND CONTROLS IN PULVERIZED COAL BOILERS BY USING IMAGE-BASED TECHNIQUES OF TEMPERATURE MEASUREMENTS filed Aug. 27, 2010, the disclosure of which is incorporated herein by reference in its entirety.
| Number | Date | Country | |
|---|---|---|---|
| 61377637 | Aug 2010 | US | |
| 61377676 | Aug 2010 | US |
