A method for determining the behavior of exhaust gases

Abstract

Method for determining the behavior of pollutants

Description

Description A method for determining the behavior of exhaust gases Field of invention The present invention relates to a method for determining the behavior of exhaust gases of a gas turbine engine. The present invention also relates to a gas turbine engine operable by such a method. History of invention It is well known that industrial plants not only produce large amounts of energy but also exhaust gases, such as nitrogen oxides (NOx) and carbon dioxide (CO), which can be harmful to humans and the environment. Therefore, many efforts have been made to reduce these pollutants. Hence, monitoring exhaust gases from industrial plants is essential. Furthermore, depending on the size of an industrial plant, and applicable regulations, continuous monitoring of exhaust gas levels (primarily NOx) is a legal requirement for some industrial combustion processes. Continuous monitoring of plant exhaust gases may be performed by an automatic exhaust gas monitoring system (AMS), which is a continuous direct method of measuring exhaust gases, or by a predictive emission monitoring system (PEMS), which uses specific process parameters to calculate (predict) exhaust gas levels. Of the two methods, PEMS has a significantly lower implementation cost and less operational complexity. However, PEMS models typically require "major training and on-site calibration", and are usually only capable of being calibrated in a specific plant under operating conditions / environment (see below). Such a system is described for example in EP 1 864 193 B11. Because PEMS models are typically produced by third parties who may not have the same detailed product knowledge as an original equipment manufacturer (OEM), current PEMS models typically rely heavily on on-site “training” (i.e., neural network tuning) and calibration, and the model varies from site to site even for the same plant configuration. While an initial model may exist, incorporating some basic combustion characteristics, the model is essentially an experience-based relationship between a process parameter and exhaust gas levels. Such models often require regular recalibration as an engine naturally deteriorates over time. The accuracy of the model is questionable under environmental and operating conditions beyond its calibration range. A first object of the present invention is to provide a method for determining the behavior of the exhaust gases of a gas turbine engine, whereby the aforementioned drawback can be reduced, and in particular, it enables efficient and reliable operation of the gas turbine engine. A second object of the invention is to provide a gas turbine engine that can operate reliably and with reduced exhaust gas levels. These objects may be achieved by a method and a gas turbine engine according to the subject matter of the independent claims. Summary of the invention Accordingly, the present invention provides a method for determining the behavior of exhaust gases of a gas turbine engine. The method further includes the following steps: determining the behavior parameters of the exhaust gases from the gas turbine engine for at least one first selected variable state of the gas turbine engine using a model, which represents the behavior state of the gas turbine engine, and determining the behavior of the exhaust gases from the gas turbine engine using the determination of the parameters. Due to the inventive method, the PEMS model is able to predict the exhaust gas production for a wide range of inlet conditions. Furthermore, the scope of the inventive method allows the predicted exhaust gas levels to be used in optimizing the actual exhaust gas levels. This allows, for example, a balance between the combustion chamber pressure dynamics and low exhaust gases, e.g. NOX. Furthermore, the accuracy of the model is higher over a wider range of inlet and operating conditions than conventional systems, especially if the model is based on chemical kinetics principles. Furthermore, the PEMS model is relatively independent of the specific site configuration and requires minimal "on-site" tuning and training. Advantageously, the model also tolerates changes in engine performance over long periods due to, for example, component failure or the effects of changes in the installation (e.g. fuel, filter, burner). The model is used to predict the behavior of the exhaust gases under selected conditions. This makes these conditions unique, which are then used in the PEMS model.Using a 'unique feature' instead of the model itself means that a complex (and hence time-consuming) model can be used for combustion modelling, but the final site-based version can be a much simpler and faster-to-implement software. Even if a word such as state variable, processing unit, or fuel delivery system is used in the singular or in the form of specific numbers in the claims and specifications of the patent scope (declaration), it should not be limited to the singular or specific number. Also, having more than one or more of the above words or structures should also be within the scope of the invention. In this context, a state of behavior is understood to mean a result of a specific condition or state of the gas turbine engine or a part thereof, such as operating during a sequence of the gas turbine engine or operating in a high temperature environment, etc., and an exhaust gas behavior is understood to mean a result of a specific condition of the gas turbine engine or a part thereof relating to the exhaust gases of this gas turbine engine or part thereof. This may be a one-dimensional quantity, for example an exhaust gas level, or a multi-dimensional pattern depending on more than one quantity or at least on different types of quantities, such as an exhaust gas pattern, for example the change in exhaust gas level over time or relative to combustion temperature. Furthermore, a state variable is understood to mean a variable indicative of a selected or specific state of the gas turbine engine, such as a temperature in a selected region of the gas turbine engine or its parts or a specific time during operation of the gas turbine or its parts.The state variable may be a measured value or an inferred / obtained value that is derived from a measured value. A parameterization means a single parameterization case. A multiple parameterization case is considered a single parameterization. In addition, a parameterization may also be called a parametric equation. In addition, the model is preferably a mathematical model. According to one embodiment of the invention, the method comprises the following steps: determining the parameters of the behavior of the exhaust gases from the gas turbine engine for at least one selected first variable state and several selected second variable states. Thus, the production of exhaust gases for a wide range of input conditions can be predicted. Advantageously and in particular, the method comprises: performing the determination of the parameters of the behavior of the exhaust gases from the gas turbine engine by performing the determination of individual parameters for at least one selected first variable state and for each selected second variable state separately. Thus, the different state variables can be examined independently of the determination of other state variable parameters, and thus multiple sets of information. The determination of the individual parameters can be considered as the determination of sub-parameters of the complex determined parameters. Preferably, at least one case of determining the parameters of the behavior of the exhaust gases from a gas turbine engine is an image in a two-dimensional state space. This allows the dependencies of the two state variables to be represented in a simple and straightforward manner. In a preferred embodiment of the invention, the method comprises: describing the behavior of exhaust gases from a gas turbine engine using another gas turbine engine variable. Thus, the conditions of the gas turbine engine or its components can be easily defined. This other gas turbine engine variable may be any variable suitable to a person skilled in the art, such as exhaust temperature, combustion temperature, amount of unburned hydrocarbons (in UHC), or one or more exhaust gas parameters, preferably an exhaust gas level of any suitable exhaust gas material or combination of materials or, in particular, the exhaust gas NOx level or the exhaust gas CO level. In addition, at least one first state variable of the selected state represents an input of the model. In other words, the first state variable of the selected state is an input variable of this model. In addition, another state variable, which represents a behavior of the exhaust gases from the gas turbine engine, represents an output of the model. In other words, the other state variable is an output variable of the model. As a result, the defined dependencies can be found. Thus, the model represents the first state variable of the gas turbine engine as an input variable on another gas turbine engine variable as an output variable. The output variable may be, for example, the amount of NOx exhaust gases or a unique characteristic of the exhaust gases under the selected conditions. As mentioned above, at least one selected first state variable is used as a model input. Advantageously, the implementation of the method of the subject invention comprises: implementing the model by varying at least one selected first state variable, while keeping the other model inputs constant. This allows only one selected variable to be investigated or under certain defined conditions. As a result, the implementation of the model of a specific behavior state of the gas turbine engine can be realized depending on a specific combination of the at least one selected first state variable and other state variables. The method then comprises: determining the parameters using the behavior state variation modeling of the at least one selected first state variable. In addition, the method includes: determining the parameters using a discontinuity resulting from the change of at least one selected first state variable and the modeled behavior state, in particular, a modeled value of a modeled behavior state variable. Hence, the model results can be easily represented. This is easily done when the method includes: determining the parameters by an approximation of the discontinuity with a continuous function. Thus, the behavior of the exhaust gases is transformed into a simple mathematical function (usually a polynomial expression). Commonly implemented PEMS models are usually complex models and require specialized software and significant processing time. However, by converting the model output into relatively simple mathematical expressions using the proposed innovative method, a much faster implementation model can be designed using simpler software. The results of the mathematical function or the unique characteristic of the exhaust gases can be displayed in a graphic image, which provides results that are easy for the operator to read. In addition, the mathematical function can be verified using test and data (small constant adjustments are allowed) to more accurately match the predicted and calculated values. Preferably, the model is a kinetic model and, in particular, the model is a physical kinetic model or, most preferably, a chemical kinetic model. The chemical kinetic model of a particular gas turbine engine or its combustion system will be capable of predicting the production of exhaust gases for a wide range of inlet conditions. Advantageously, the at least one selected first state variable is a variable selected from the group consisting of: a directly measured parameter, an estimated parameter, or a thermodynamically derived parameter. Thus, a wide range of different variables or exact values ​​of variables can be considered and used to determine the state of behavior of the gas turbine engine or its components. In addition, another state variable can also be selected from this group. A derived thermodynamic parameter is a parameter based on measured values ​​and constants obtained from factory testing of a gas turbine engine. These constants describe the operating conditions of the components (such as the stage 1 compressor turbine intake capacity and compressor turbine loading) and are based on the thermodynamic principles of gas turbine engine operation. On site, the measured and derived constants are fed into a control algorithm and state variables or parameters that cannot be directly measured. For example, it is known to use a calculated turbine inlet temperature to control the operation of a gas turbine engine other than full power. The at least one selected first state variable may be any parameter practical to a person skilled in the art. However, in the case of a directly measured parameter, preferably one is selected from the group consisting of: an ambient inlet pressure, an engine inlet pressure, an ambient inlet temperature, an engine inlet temperature, a compressor inlet temperature, a compressor inlet pressure, a compressor outlet pressure or delivery pressure, respectively, a compressor outlet temperature or delivery temperature, respectively, a turbine inlet pressure, a turbine inlet temperature, an exhaust temperature, a gas turbine engine temperature, a fuel flow, a fuel composition, a fuel temperature or a fuel split ratio of the main to the pilot. Using the directly measured parameters, the actual conditions of the gas turbine engine or its components can be accurately detected and taken into account. Alternatively, if at least one selected parameter is thermodynamically derived, it is selected from the group consisting of: a combustor outlet pressure or a combustor discharge pressure, respectively, a combustor outlet temperature or a combustor discharge temperature, respectively, a combustion temperature, a compressor delivery air percentage, or a calculated mass flow. By using the compressor inlet mass flow and / or the compressor delivery air percentage from one of its flow chambers as a derived thermodynamic parameter, a more accurate modeling of the gas turbine engine combustion system can be performed. It can serve as a diagnostic tool for direct measurement instruments and allow additional combustor operating parameters to be calculated (e.g. flame temperature, air to fuel ratio). Furthermore, by using parameter sensors, direct measurements can be eliminated, saving space, installation costs, and installation effort. In addition, the basic design characteristic of the combustion system may also be considered. This may be, for example, the geometry of the components, such as a combustion chamber component, or calculated or derived combustion chamber characteristics. In case, for example, there is more than one compressor or more than one turbine all the above-mentioned values ​​may be related to, for example, the pressure or temperature between two adjacent compressors / turbines. In a preferred configuration, the thermodynamic behavior model of key gas turbine engine components enables accurate monitoring. In particular, the model maps a gas turbine engine combustion system and thus focuses on a gas turbine engine system that is highly relevant in terms of gas turbine engine exhaust gases. A key component may be any component practical to a person skilled in the art, but preferably a key component is a component selected from the group consisting of: a primary combustion zone, a fully stirred reactor, a main flame, a pilot flame, a plug flow or mass flow integration reactor, a mass flow gap, a flow resistance. In other words, the model - chemical kinetics - describes a specific gas turbine combustion system by creating a network model of the combustion system in which key parts of the combustion process are separate model components. Examples include: the primary combustion zone consisting of a series of fully stirred reactor model elements, where the main and pilot flames are modeled separately (allowing the split main / pilot effect to be investigated); the downstream parts are characterized by the plug flow of the reactor model elements; dilution and mixing are represented by the appropriate mass flow integration model elements and the boundary conditions are provided by the model boundary elements. Furthermore, the method comprises: using the determination of the gas turbine engine exhaust gas behavior parameters to predict the exhaust gas behavior for predetermined values ​​of at least one selected first state variable. Hence, by using the exhaust gas 'unique feature', gas turbine operation prediction can be performed in a much simpler and faster manner compared to using conventional complex PEMS models. The predictions from the model and the subsequent determination of parameters are used to run and operate the gas turbine engine. Therefore, the various mathematical functions obtained from running the model and determining the parameters are incorporated into a software package for use in a processor unit of the gas turbine engine. Therefore, the predicted exhaust gas levels are fed into the engine control system. Then, when the gas turbine engine is started up on site, the effects of changes in each parameter are combined using mathematical functions to calculate the predicted exhaust gas production. Based on the predictions, the exact values ​​of the gas turbine engine variables are adjusted to the values ​​resulting in the output or exhaust gases of the behavior recommended by the model prediction. This is done by including the means or parts of the gas turbine engine affecting the exact value, such as temperature, pressure, fuel ratio, etc. This may be any part or system practical for a person skilled in the art, such as a fuel or air valve, a position of a blade or vane, a cooling system for coolant inlet, etc. Preferably the main and pilot fuel split ratio is affected. Advantageously, the method comprises: using a predicted exhaust gas behavior or unique characteristic from the model and determining parameters to control the exhaust gas level during operation of the gas turbine engine. As discussed above, this may be accomplished by manipulating the main / test fuel split. This will allow the PEMS model to optimize the exhaust gas level by varying the main and test fuel split ratios (e.g., to ensure that the exhaust gas level remains within controlled values). The subject method of the invention is to use a chemical kinetic model of a specific gas turbine engine or its combustion system to obtain a 'unique characteristic' of the exhaust gases for directly measured, inferred, and derived thermodynamic parameters, which are then validated using test and performance data over a wide range of operating conditions. This means that the PEMS model is relatively independent of the specific site configuration and requires minimal on-site tuning and training. The innovative PEMS model can be implemented both as a standalone software package and can be incorporated into the engine control system. The present invention also relates to a gas turbine engine comprising at least one processing unit and operable with at least one prediction obtained using the method of the invention. It is proposed that the model be implemented in at least one processing unit to predict and / or control the behavior of exhaust gases and, in particular, the level of exhaust gases from a gas turbine engine. Because of this, the PEMS model will be able to predict the exhaust gas production for a wide range of inlet conditions. Furthermore, it is possible to use the predicted exhaust gas levels in optimizing the actual exhaust gas levels. This, for example, achieves a balance between the dynamics of the combustion chamber pressure and low exhaust gases, e.g. NOX. Furthermore, the reliability of the accurate model over a wider range of inlets and operating conditions is higher than for conventional systems, especially if the model is based on chemical kinetics principles. Furthermore, the PEMS model is relatively independent of the specific site configuration and requires minimal on-site tuning and training. Advantageously, the model accommodates changes in engine performance over long periods due, for example, to component failure or the effects of changes in the installation (e.g. fuel, filter, burner). The model is used to predict the behavior of the exhaust gases under selected conditions. This leads to the identification of the ‘unique characteristic’ of these conditions, which is subsequently used in the PEMS model.Using a 'unique feature' instead of the model itself means that a complex (and hence time-consuming) model can be used for combustion modelling, but the final site-based version can be a much simpler and faster-to-implement software. In another embodiment of the invention, it is proposed that the gas turbine engine includes at least one fuel supply system, wherein a main and pilot fuel split ratio of the at least one fuel supply system is adjustable in relation to said method for infiltrating the exhaust gas surface based on a model prediction. The aforementioned features, features and advantages of the use of this invention and the manner in which they are realized are clearly and distinctly explained in connection with the following description of exemplary configurations which are illustrated in connection with the figures. Brief description of the shapes The present invention will be described with reference to the figures, in which: Figure 1: A schematic and cross-sectional view of a gas turbine engine operable by the method of the invention, Fig. 2: A flow chart of a chemical kinetics model and a parameterization in a gas turbine engine processor unit of Fig. 1 shows the implementation, 3-: Traces and simple elements of possible analysis of the chemical kinetics model are shown in Figure 2; Figure 4: Schematic of the gas turbine engine from Figure 1 showing the engine position from where the state variables are derived from the chemical kinetics model of Figure 2, Figure 5: A plot depicting the dependence of NOX emission on a combustion chamber inlet temperature obtained with the chemical kinetics model in terms of 22 Figure 6 is a plot of the dependence of an exhaust gas NOx on an original / experimental split ratio obtained with the chemical kinetics model of Figure 2 and Figure 7: Flowchart showing two possible operating modes of a PEMS model monitoring and controlling the exhaust gases of the gas turbine engine from Figure 1. Detailed description of the depicted configurations The terms upstream and downstream refer to the direction of air flow and / or working gas flow from the gas turbine engine 10 unless otherwise noted. When used and unless otherwise specified, the terms axial, radial, and circumferential refer to the axis of rotation 300 of the gas turbine engine 10. Figure 1 shows an exemplary gas turbine engine 10 in a cross-sectional view. Gas turbine engine 10 includes, in series order, an inlet 22, a compressor section 24, a combustion section 266, and a turbine section 28, which are generally arranged in series order and generally in the direction of a longitudinal or rotational axis 30. Gas turbine engine 10 also includes a shaft 32 rotatable about the axis of rotation 30 and extending longitudinally from gas turbine engine 10. Shaft 32 is power-coupled to turbine section 28 of compressor section 24. In operation of the gas turbine engine 10, air 34 drawn from the air inlet 22 is compressed by a compressor section 24 and delivered to a combustion section or burner section 26. The burner section 26 includes a combustion system 14 with a plenum burner 36, one or more combustion chambers 38 defined by a double-walled canister 40, and at least one burner 42 mounted in each combustion chamber 38. The combustion chamber 38 and the burner(s) 42 are located within the burner plenum 36. Compressed air from the compressor section 24 enters a diffuser 44 and from the diffuser 44 to the burner plenum 36 where a portion of the air enters the burner(s) 42 and is mixed with a gaseous or liquid fuel. The air / fuel mixture is subsequently combusted and the combustion gas 466 or working gas from the combustion is directed through a transfer duct 48 to the turbine section 28. Turbine section 28 includes a plurality of blade-carrying production disks 50 or turbine wheels connected to shaft 32. In the present example, turbine section 28 includes four disks 50, each carrying an array of annular turbine blades 52. However, the number of blade-carrying production disks 50 can vary, i.e., only one production disk 50 or more than one production disk 50. In addition, stator stages or rows of turbine blades 54 are positioned between turbine blades 52. Each stator stage carries an annular array of guide vanes 56, which are connected to a stator 58 of gas turbine engine 10. Inlet guide vanes or nozzle guide vanes 60 are disposed between the outlet of combustor 38 and the guide turbine blades 52. Combustion gas 46 enters turbine section 28 from combustion chamber 36 and drives turbine blades 52 which in turn rotate shaft 32. Guide vanes 56, 60 act on turbine blades 52 to optimize the angle of combustion or working gas 46. Compressor section 24 includes a series of axial guide vanes 62 and rotor blades 64 with turbine blades 52 or blades 56. In addition, the gas turbine engine 10 includes a processing unit 18, which is depicted outside the gas turbine engine 10 for better presentation. To determine the behavior of the exhaust gases of a gas turbine engine 10 or to monitor and control the exhaust gases from the gas turbine engine 10, such as NOx and / or CO, the processing unit 188 includes an implemented model of a predictive exhaust monitoring system PEMS 104. The PEMS model 104 is based on a combination of a chemical kinetics model 102 and a parameterization 100 or, in other words, a parametric equation or model 102, which is described with reference to FIGS. 2 to 7. The PEMS model 104 is realized by modeling selected systems of the gas turbine engine 10, such as a combustion system 14 of the gas turbine engine 10. The modeling is performed using a chemical kinetics model 102 that uses the same input parameters or state variables P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT (see below) as the final PEMS model 104. However, the results of the model 102 are converted into mathematical functions that can be used to predict the production of exhaust gases without the need to run the chemical kinetics model 102 online. The process of creating such a PEMS 104 model is depicted in the flow chart of Figure 2. To create a chemical kinetics model 102 describing a specific combustion system 14 of a gas turbine engine 10, an appropriate network model of the combustion system 14 is created (step 102a). Thus, key parts or components 12 of the combustion process are identified. These key components 122 are represented by separate model components or analysis elements. This is a simple example for the two key components 12 shown in FIG. 3 (see also FIG. 1). The key component 122 may, for example, be a primary combustion zone 16 that is comprised of a series of fully stirred reactor (PSR) model elements 66, in which the main and pilot flames 68 are modeled separately (allowing the effect of the main / pilot split to be examined). Another key component 122 may be located in a downstream combustion zone 70 and may be comprised of plug flow reactor (PFR) model elements 72. Other key components 12 include, for example, appropriate mass flow integration model elements, such as dilution and mixing of the various gas streams being modeled, for example air or combustion products, or fuel, or a mass flow gap element or flow resistance element. Boundary conditions include model boundary elements (for example, the volume fraction of selected components or the duration of various processes available) (not shown). As shown in Figure 3, at least one selected MCI initial state variable is used as an input to the model 102. While this is defined as a mass flow input, it includes the selected pressure P, the selected temperature T, the fuel mixture flow QH and QF, where the input “mass flow” is a derived thermodynamic parameter. Other selected initial state variables are P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, TFIRE, QF, QH, QT, SPLIT. These variables are P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT or a directly measured parameter, an inferred parameter or a derived thermodynamic parameter. Figure 4 shows which engine positions have which conditions and are present and the state variables P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT Chemical kinetic model 102 can be obtained. Ambient air 344 has ambient pressure and ambient temperature (not specifically depicted with reference numerals) before entering gas turbine engine 10. At inlet 22 air 34 passing through a filter housing 74 has engine inlet pressure P0 and engine inlet temperature T0. Air entering the compressor from compressor section 244 has compressor inlet pressure P1 and compressor inlet temperature T1. P2 is the compressor outlet pressure, also referred to as the combustion chamber inlet pressure P2 or the compressor delivery pressure P2, and T2 is the compressor outlet temperature, also referred to as the combustion chamber inlet temperature T2 or the compressor delivery temperature T2. Variables related to the fuel 76 and the fuel delivery system 20 are a fuel flow QF, a fuel composition QH, a fuel temperature QT, and a split ratio of the main and test fuel.Combustion gas 46 exiting combustion section 26 has a combustion chamber outlet pressure P3 and a combustion chamber outlet temperature T33. Fluid flowing through passages 78 located between a compressor turbine 80 and a power turbine 82 of turbine section 28 has a turbine passage pressure P4 and a turbine passage temperature T4. An exhaust gas 84 exiting gas turbine engine 10 has an exhaust pressure P5 and an exhaust temperature T55. The directly measured parameter could be engine inlet pressure P0, engine inlet temperature T0, compressor inlet temperature T1, compressor inlet pressure P1, compressor outlet pressure P2, compressor outlet temperature T2, turbine inlet pressure P4. All of these variables may be measured by one or more sensors not shown. The abbreviation used for gas turbine temperature depends on the type of gas turbine engine and may be TOP, TMAX or TLIMIT and is calculated based on some of the measured values ​​above (not shown). Variables that may be inferred based on the relationship parameters may be compressor inlet pressure P1, fuel flow QF, fuel composition QH, and exhaust temperature T5. The derived thermodynamic values ​​use a combination of direct measurements and component characteristics obtained from internal engine factory testing and may be the combustion chamber outlet pressure P3, combustion chamber outlet temperature T3, engine ignition temperature TFIRE, compressor flow, or calculated mass inlet flow MCI. In addition, the basic design characteristics of the combustion system (used in the chemical kinetics model) may also be considered. The geometry of the combustion chamber components or the combustion chamber properties may be calculated or derived. P2, T2, SPLIT, QH, QF, TFIRE, P2B and MCI are the initial input parameters to the PEMS model or the parametric model used 100, respectively. Subsequently, in step 102B, the model 102 is run under a range of input conditions, or in other words, the model 102 is run by varying at least one selected first state variable MCI, where the other input or other input variables or selected second state variables P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT of the model 102 are held constant. This may be done individually for several or all variables P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT. By varying each of the parameters / boundary variables P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT while keeping all other input parameters P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT constant (as far as is reasonably practicable), the effect of each of the parameters P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT on the production of exhaust gases within the parameter modeling range is obtained. The parameters / variables P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT that need to be changed are variables that: have values ​​directly available at the equivalent site (P0, P1, P2, P4, T0, T1, T2, T4, T5, QF, QH, QT, SPLIT), can be inferred from a relationship with other directly measured parameters (P1, P3, T3, QF, QH, T5), or are a thermodynamic value derived using a combination of direct measurement and component characteristics obtained from internal factory testing of 10 gas turbine engines (P3, T3, TFIRE, P2B, MCI). In step 102C, the 'unique characteristics' of the exhaust gases 106, 106 of the model 102 are obtained in terms of the variation of the input parameters or selected first and second state variables P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT, respectively. The results of these variations may be graphically displayed and referred to as the 'unique characteristic' parameters 106, 106. The two 'unique characteristics' 106, 106 are shown in Figures 5 and 6, which each show a graph of the dependence of an exhaust gas NOx on a combustor inlet temperature T2 (Figure 5) and an experimental / original contribution split (Figure 6) obtained with the chemical kinetics model 1022. As can be seen in Figures 5 and 6, the behavior of the exhaust gases of the gas turbine engine 10 is described using another state variable OUT of the gas turbine engine (10), particularly in terms of the level of NOx exhaust gases. In other words, the other state variable OUT, which represents the behavior of the exhaust gases of the gas turbine engine 10, is an output of the model 102 (see also Figure 3). Subsequently, in step 100A (obtaining transfer standards based on mathematical functions of exhaust gas levels) the parameterization is performed or the parametric model is implemented. This is done by using the change in the state of the modeled behavior of at least one first state variable P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT and in particular, by using a discontinuity resulting from the change in the at least one selected first state variable P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT and the modeled behavior state, in particular, a modeled value of a modeled behavior state variable and in particular, by setting the parameters 100 with an approximation of the discontinuity with a continuous function. In other words, the unique characteristic output gases 106, 106 or their graphical images, respectively, are converted into relatively simple mathematical functions (usually polynomial expressions). An example of such a mathematical function might be the following function, which is the OUT variable of the NOx exhaust gas in dependence on the compressor delivery pressure PCD or the compressor outlet pressure P2: NOx= 12.26 + (4.93 E-31* P25) + (3.157 E-18* P25) - (1.88 E-24* P24) - (8.267 E-13* P22) - (4.58 E-38* P25) – (0,0000034 * P2) In a further step 100B the predicted exhaust gases are compared with the measured data. The functions may be simplified if necessary. In addition, the mathematical functions can be validated using test data and small adjustments of the constant values ​​allowed in order to match the predicted values ​​and calculate more accurately. Determining the parameters 100 of the behavior of the exhaust gases of the gas turbine engine 10 by determining the individual parameters 100 for at least one of the selected first state variables P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT and for each selected second state variable separately P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT. Additionally, one or each determination of parameters 100 of the behavior of exhaust gases 10 of a gas turbine engine is an image in a two-dimensional space state. In general, the method of the invention includes the steps of: determining the parameters 100 of the behavior of the exhaust gases of the gas turbine engine 10 for at least one selected first state variable P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT of the gas turbine engine 10 using the model 102, which represents the behavior state of the gas turbine engine 10 and determining the behavior of the exhaust gases of the gas turbine engine 10 using the determination of the parameters 100. Mathematical functions are used to create a PEMS algorithm or model 104 (Figure 2). Various mathematical functions involving the dependencies of the state variables P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT, OUT are incorporated into a software package. This can be seen in Figure 7, which is a flow chart showing two possible operating modes of the PEMS model monitoring and controlling the exhaust gases from the gas turbine engine 10. Then, when the gas turbine engine 10 is started up on site, the effects of changes in each parameter P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT are combined using mathematical functions to calculate predicted exhaust gas production (step 108). Thus, the determination of the parameters 100 of the exhaust gas behavior of the gas turbine engine 10 is used to predict the exhaust gas behavior for predetermined values ​​of at least one selected first state variable P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT. The resulting model 102 or PEMS model 104 may be implemented in the processing unit 18 of the gas turbine engine 10 (left side of FIG. 7 ) or may be a stand-alone software package (right side of FIG. 7 ). The predicted exhaust gas levels may be stored in a database (step 110) or they may be fed to the engine control system (step 112) which may be used to control the exhaust gas levels of the gas turbine engine 10. This may be accomplished by adjusting the main and pilot fuel ratios to the SPLIT fueling system 20 in such a manner as to affect the exhaust gas levels based on the predictions of the model 102 (step 114). As a result, the invention controls and predicts exhaust gas levels (primarily NOx and CO) using unique exhaust gas characteristics derived mathematically based on chemical kinetic models of measured input parameters and derived thermodynamic parameters. It should be noted that the term "consisting of" does not exclude other elements or steps, and "a" does not exclude a plurality. Also, elements described in connection with different configurations may be combined. It should also be noted that symbols in the claims should not be construed as limiting the scope of the claims. Although the invention has been shown and described in detail with preferred configurations, the invention is not limited to the disclosed embodiments, and other modifications thereof may be made by a person skilled in the art without departing from the scope of the invention. Complaint 1. A method for determining the behavior of exhaust gases of a gas turbine engine (10), characterized by the steps of: - determining the parameters (100) of the behavior of exhaust gases of the gas turbine engine (10) for at least one selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT) of the gas turbine engine (10) using a model (102), which represents the behavior of the gas turbine engine (10), wherein the model (102) obtains results that are converted into mathematical functions - determining the behavior of exhaust gases of the gas turbine engine (10) using mathematical functions. 2. A method according to claim 1, wherein the method comprises: - Determining the parameters (100) of the behavior of the exhaust gases of the gas turbine engine (10) for at least one selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT) and several selected second state variables, and in particular, - Performing parameter determination (100) of the behavior of the exhaust gases of the gas turbine engine (10) by performing individual parameter determination (100) for at least one selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT) and for each selected second state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT) separately. 3. A method according to claim 2, wherein the at least one determined parameter (100) is the behavior of exhaust gases of a gas turbine engine (10) of an image in a two-dimensional space state. 4. A method according to any one of the preceding claims, wherein the method comprises: - describing the behavior of the exhaust gases of the gas turbine engine (10) using another variable state (output) of the gas turbine engine (10), preferably an exhaust gas level, in particular a NOX exhaust gas level or a CO exhaust gas level. 5. A method according to any one of the preceding claims, wherein at least one selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT) represents an input to the model (102) and / or wherein another state variable (output), which represents a behavior of the exhaust gases of the gas turbine engine (10), is an output to the model (102). 6. A method according to any one of the preceding claims, wherein the method comprises: - using at least one selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT) as input to the model (102), - executing the model (102) by varying at least one selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT), where the other inputs to the model (102) are held constant, and thus - The determination of parameters (100) is performed using the modeled behavior state from the change of a minimum selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT). 7. A method according to any one of the preceding claims, wherein the method comprises: - determining parameters (100) using a discontinuity resulting from the change of a minimum first selected state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT) and the modeled behavior state, in particular, a model value of a modeled behavior state variable and in particular, - Determination of parameters (100) with an approximation of discontinuity with a continuous function. 8. A method according to any one of the preceding claims, wherein the model is a kinetic model, in particular, wherein the model is a physical kinetic model or a chemical kinetic model (102). 9. A method according to any one of the preceding claims, wherein a minimum first state variable selected from a group (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT) is: a directly measured parameter, an estimated parameter or a derived thermodynamic parameter. 10. A method according to any one of the preceding claims, wherein a selected minimum first state variable (P0, P1, P2, P4, T0, T1, T2, T4, T5, QF, QH, QT, SPLIT) is a variable selected from the group of directly measured parameters comprising: inlet ambient pressure, an engine inlet pressure (P0), ambient inlet temperature, engine inlet temperature (T0), compressor inlet temperature (T1), a compressor inlet pressure (P1), compressor outlet pressure (P2), compressor outlet temperature (T2), and turbine inlet pressure (P4), turbine inlet temperature (T4), exhaust temperature (T5), gas turbine engine temperature, fuel flow (QF), a fuel composition (QH), fuel temperature (QT), or a main and pilot fuel split ratio (SPLIT). 11. A method according to any one of the preceding claims, wherein a selected minimum first state variable (P3, T3, TFIRE, P2B, MCI) is a selected variable of the group of derived thermodynamic parameters of combustor outlet pressure (P3), combustor outlet temperature (T3), engine temperature (TFIRE), compressor delivery air flow percentage (P2B) or a calculated mass flow (MCI). 12. A method according to any preceding claim, wherein the model depicts a thermodynamic behavior of key components (12) of a gas turbine engine (10) and in particular, a combustion system (14) of the gas turbine engine (10), and in particular, wherein said key component (12) is a component selected from the group consisting of: a primary combustion zone (16), a fully stirred reactor (66), a main flame (68), a pilot flame (68), a plug flow reactor (72), a mass flow merger, a mass flow splitter or a flow resistance. 13. A method according to any one of the preceding claims, wherein the method comprises: - Using the determination of parameters (100) of the behavior of exhaust gases of a gas turbine engine (10) to predict the behavior of exhaust gases for predetermined values ​​of at least one selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT). 14. A gas turbine engine (10) comprising at least one processing unit (18) and operable with at least one prediction obtained using the method of any one of claims 1 to 13, characterized in that the model (102), which represents the behavior of the gas turbine engine (10) and obtains results that are converted into mathematical functions by determining parameters (100), and in the at least one processing unit (18) for predicting and / or controlling a level of exhaust gases of the gas turbine engine (10) using mathematical functions. 15. A gas turbine engine according to claim 14, characterized in that it comprises at least one fuel system (20), wherein the main and pilot fuel split ratio (SPLIT) of the at least one fuel system (20) is adjustable such that affect the level of exhaust gases according to the prediction of model (102). Abstract of the invention A method for determining the behavior of exhaust gases The present invention relates to a method for determining the behavior of exhaust gases of a gas turbine engine (10). In order to provide reliable operation of a gas turbine engine (10), the method includes the steps of: determining parameters (100) of the behavior of the exhaust gases of the gas turbine engine (10) for at least one selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT) of the gas turbine engine (10) using a model (102), which represents the behavior of the gas turbine engine (10), and determining the behavior of the exhaust gases of the gas turbine engine (10) using the determination of the parameters (100). Figure 2

Claims

CLAIMS 1. A method for determining an emission behaviour of a gas turbine engine (10), characterised by the steps of: - creating (102a) a model (102) of a system of the gas turbine engine (10); - running (102b) the model (102) by varying at least one selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT)used as an input of the model (102); - obtaining (102c) an output (OUT) of the model (102)) representing an emission behaviour of the gas turbine engine (10); - parameterising (100) the emission behaviour of the gas turbine engine (10) for the at least one selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT) of the gas turbine engine (10) by converting the output (OUT) into mathematical functions and - determining the emission behaviour of the gas turbine engine (10) by using the mathematical functions.

2. A method according to claim 1, wherein the method comprises the step(s) of: - parameterising (100) the emission behaviour of the gas turbine engine (10) for the at least one selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT) and several different selected second state variables and specifically, - performing the parameterisation (100) of the emission behaviour of the gas turbine engine (10) by performing individual parameterisations (100) for the at least one selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT) and for each selected second state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT) separately.

3. A method according to claim 2, wherein at least one individual parameterisation (100) of the emission behaviour of the gas turbine engine (10) is a depiction in a two-dimensional state space.

4. A method according to any one of the preceding claims, wherein the output of the model (102) comprises a further state variable (OUT) of the gas turbine engine (10), preferably an emission level, specifically an emission level of NOx or an emission level of CO.

5. A method according to any one of the preceding claims, wherein, the at least one selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT) represents an input of the model (102) and / or wherein a further state variable (OUT), which reflects an emission behaviour of the gas turbine engine (10), represents an output of the model (102).

6. A method according to any one of the preceding claims, wherein the method comprises the step(s) of: - comparing (102b) the results of the mathematical functions with measured data.

7. A method according to any preceding claim, wherein the method comprises the step(s) of: - determining the parameterisation (100) by using a discretisation resulting from the variation of the at least one selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT) and the modelled state behaviour, especially, a modelled value of a variable of the modelled state behaviour and specifically, - determining the parameterisation (100) with an approximation of the discretisation with a continuous function.

8. A method according to any preceding claim, wherein the model is a kinetic model, specifically, wherein the model is a physical kinetic model or a chemical kinetic model (102).

9. A method according to any preceding claim, wherein the at least one selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT) is a variable selected out of the group consisting of: a directly measured parameter, an inferred parameter or a thermodynamically derived parameter.

10. A method according to any preceding claim, wherein the at least one selected first state variable (P0, P1, P2, P4, T0, T1, T2, T4, T5, QF, QH, QT, SPLIT) is a variable selected out of the group of directly measured parameters consisting of: an ambient inlet pressure, an engine inlet pressure (P0), an ambient inlet temperature, an engine inlet temperature (T0), a compressor inlet temperature (T1), a compressor inlet pressure (P1), a compressor exit pressure (P2), a compressor exit temperature (T2), a turbine interduct pressure (P4), a turbine interduct temperature (T4), an exhaust temperature (T5), a gas turbine engine operating temperature, a fuel flow (QF), a fuel composition (QH), a fuel temperature (QT) or a main and pilot fuel split ratio (SPLIT).

11. A method according to any preceding claim, wherein the at least one selected first state variable (P3, T3, TFIRE, P2B, MCI) is a variable selected out of the group of thermodynamically derived parameters consisting of: combustor exit pressure (P3), combustor exit temperature (T3), an engine firing temperature (TFIRE), a compressor delivery air percentage bleed (P2B) or a calculated mass flow (MCI).

12. A method according to any preceding claim, wherein the model maps a thermodynamic behaviour of key components (12) of the gas turbine engine (10) and specifically, of a combustion system (14) of the gas turbine engine (10), and specifically, wherein the key component (12) is a component selected out of the group consisting of: a combustion primary zone (16), a perfectly-stirred reactor (66), a main flame (68), a pilot flame (68), a plug-flow reactor (72), a mass flow merger, a mass flow splitter or a flow resistance.

13. A method according to any preceding claim, wherein the method comprises the step of: - using the parameterisation (100) of the emission behaviour of the gas turbine engine (10) to predict an emission behaviour for predetermined values of the at least one selected first state variable (P0, P1, P2, P3, P4, T0, T1, T2, T3, T4, T5, P2B, MCI, TFIRE, QF, QH, QT, SPLIT).

14. A gas turbine engine (10) comprising at least one processing unit (18) and being operatable with at least one prediction obtained by using to the method according to any one of claims 1 to 13, characterised in that the model (102), which reflects a state behaviour of the gas turbine engine (10) and that obtains results that are converted into mathematical functions by a parameterisation (100), is implemented in the at least one processing unit (18) to predict and / or control an emission level of the gas turbine engine (10) by using the mathematical functions.

15. A gas turbine engine according to claim 14, characterised by at least one fuel supply arrangement (20), wherein a main and pilot fuel split ratio (SPLIT) of the at least one fuel supply arrangement (20) is adjustable in such a way to influence the emission level according to the prediction of the model (102).

Images